sensors and electronic instrumentation advances · sensors and electronic instrumentation advances:...

152
Sensors and Electronic Instrumentation Advances: Proceedings of the 2 nd International Conference on Sensors and Electronic Instrumentation Advances 22-23 September 2016, Barcelona, Castelldefels, Spain Edited by Sergey Y. Yurish, Amin Daneshmand Malayeri

Upload: lamnhu

Post on 13-Jul-2018

273 views

Category:

Documents


9 download

TRANSCRIPT

Page 1: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

Sensors and Electronic Instrumentation Advances:

Proceedings of the 2nd International Conference

on Sensors and Electronic Instrumentation Advances

22-23 September 2016, Barcelona, Castelldefels, Spain

Edited by Sergey Y. Yurish, Amin Daneshmand Malayeri

Page 2: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2

Sergey Y. Yurish and Amin Daneshmand Malayeri, Editors Sensors and Electronic Instrumentation Advances SEIA’ 2016 Conference Proceedings Copyright © 2016 by International Frequency Sensor Association (IFSA) Publishing, S. L.

E-mail (for orders and customer service enquires): [email protected]

Visit our Home Page on http://www.sensorsportal.com

All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (IFSA Publishing, S. L., Barcelona, Spain).

Neither the authors nor International Frequency Sensor Association Publishing accept any responsibility or liability for loss or damage occasioned to any person or property through using the material, instructions, methods or ideas contained herein, or acting or refraining from acting as a result of such use.

The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identifies as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

ISBN: 978-84-608-9963-1 BN-20160920-XX BIC: TBM

Page 3: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

3

Contents

Foreword ............................................................................................................................................................... 6

Oxygen Gas Sensing in Combustion Process: from Basics to Application ...................................................... 8 P. Shuk

Solid Electrolyte Gas Sensors based on Voltammetric Modes ........................................................................ 11 Prof. Ulrich Guth

Sensors in the Era of Internet of Things and Industrial Internet: Innovations in Materials, Transducers, and Data Analytics ...................................................................................................................... 12 R. A. Potyrailo

Ion Beam Nanofabrication and Its Applications in Precision Metrology and Bio-sensing ......................... 15 Dr. Zongwei Xu

Trace Gas Measurements with Zirconia Sensors: An Overview .................................................................... 17 M. Schelter, J. Zosel, V. Vashook, U. Guth, W. Oelßner and M. Mertig

Field Effect Based Gas Sensors, from Basic Mechanisms to the Latest Commercial Device Designs ......... 19 M. Andersson, P. Möller, H. Fashandi, J. Eriksson, D. Puglisi, J. Huotari, J. Puustinen, J. Lappalainen and A. Lloyd Spetz

Application of New Ultrasonic Instrumentation to Real-time Monitoring and Analysis of Binary Gases Mixtures .................................................................................................................................................... 22 M. Battistin, S. Berry, P. Bonneau, O. Crespo-Lopez, C. Deterre, M. Doubek, G. Favre, G. Hallewell, S. Katunin, D. Lombard, A. O’Rourke, A. Madsen, S. McMahon, K. Nagai, C. Rossi , B. Pearson, D. Robinson, A. Rozanov, E. Stanecka, M. Strauss, V. Vacek, R. Vaglio and J. Young

A Universal Experimental and Computational Framework for Decoding Complex Gas Mixtures ............ 28 Unab Javed, Kannan Ramaiyan, Cortney R. Kreller, Eric L. Brosha, Rangachary Mukundan and Alexandre V. Morozov

Dopant Effects on the Response to Acetone and Ethanol of SnO2-based Gas Sensor for Breath Monitoring ........................................................................................................................................ 31 Y. Sadaoka, M. Mori, T. Ueda, H. Mitsuhashi and M. Nakatani

Small Sensor System (S3) Equipped with Nanowires Sensors as an Interdisciplinary e-Sensing Device for the Microbiological and Quality Control ...................................................................................... 33 V. Sberveglieri, M. Soprani and E. Núñez Carmona

Design of Optical-fibre Refractometric Sensors for Liquid Helium .............................................................. 36 Santa Junnuen Miron-Carrasco, Dora Mariela Martínez-Gonzalez, Sergei Khotiaintsev

Integration of VLC and a-SiC:H Technology for Indoors Navigation ........................................................... 38 P. Louro, M. A. Vieira, M. Vieira

Comparison of Pre-processing on Different Kind of Images Produced by Optical Sensor System ........... 41 Muhammad Waseem Tahir, N. A. Zaidi, R. Blank, P. P. Vinayaka, M. J. Vellekoop and W. Lang

An Image Depth Extractable CMOS Image Sensor with a Variable Bit-resolution Infrared Pixel ........... 44 Youngchul Son and Minkyu Song

CCD Camera Characterization by Means of a Custom Integrating Sphere ................................................. 47 Andreas W. Winkler, Bernhard G. Zagar

Reflective Multimode Interference Structures for Liquids Remote Detection ............................................. 52 L. R. Villarreal Jiménez, S. Enríquez Sías, C. Elizondo González, R. Domínguez-Cruz, G. Romero-Galván and Ma. José Erro Betrán

Simultaneous Electrocatalytic Voltammetric Determination of Dihydroxybenzene Isomers Using a Cobalt-phthalocyanine Modified Pencil Electrode............................................................................ 55 A. Ciucu, M. Buleandra, A. Rabinca and I. Stamatin

Page 4: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

4

Nanoplasmonic Sensing of Pb-acid and Li-ion Batteries ................................................................................ 57 David Johansson, Jenny Andersson, Björn Wickman, Fredrik Björefors, Adam Sobkowiak, Bengt Kasemo

pH Measurements Of nano-liter Solutions Using THz Technology ............................................................... 60 Y. Zhou, K. Akimune, K. Hamada, K. Sakai, T. Kiwa, K. Tsukada

Voltammetric Detection of ∆9-THC Using Carbon Screen Printed Electrode in Aqueous Media: Improvements in Forensic Analysis .................................................................................................................. 62 M. A. Balbino, I.C. Eleotério, B. R. McCord, and M. F. Oliveira

Novel Screen-printed Electrode Modified with Lead Film for Highly Sensitive and Selective Adsorptive Stripping Voltammetric Determination of Cobalt and Nickel in the Form of Dioximate Complexes ..................................................................................................................................... 64 A. Bobrowski, A. Królicka, M. Maczuga, J. Zarębski

How the Derivative Curve of Metal Oxide Sensor Response Gives Access to Features which Improve the Accuracy of Odour Quantification ................................................................................... 67 M. D. Ahmadou, E. Losson, M. Siadat, M. Lumbreras

Real-time Pressure Measuring System for the Medical Device ...................................................................... 69 D. G. Kim, S. Seo, K. Cha, S. Jeong, J. K. Choi and T.H. Song

Calibration Techniques for Skew Redundant Inertial Measurement Units .................................................. 72 M. V. Gheorghe

Measuring Machine for Mechanical Properties of Garments ........................................................................ 79 Lukas Pfarr and Bernhard Zagar

Stress Induced Birefringence Spectroscopy ..................................................................................................... 82 S. Schallmeiner, F. W. Dietachmayr and B. G. Zagar,

Health Checking System Using Wearable Health Devise and PIR Sensors .................................................. 85 T. Miyazaki, F. Shinohara, T. Horiuchi, Y. Ohira, H. Yamamoto, and M. Nishi

Power Consumption Considerations of an Agricultural Camera Sensor with Image Processing Capability ......................................................................................................................................... 87 Gábor Paller and Gábor Élő

Entropy-Based Markers of EEG Background Activity of Stroke-Related Mild Cognitive Impairment and Vascular Dementia Patients .................................................................................................. 92 Noor Kamal Al-Qazzaz, Sawal Ali, Siti Anom Ahmad, Md. Shabiul Islam, Javier Escudero

Modeling of Non-Dispersive Infrared Gas Sensors ......................................................................................... 95 F. W. Dietachmayr, and B. G. Zagar

Novel High Reliable Si-Based Trace Humidity Sensor Array for Aerospace and Process Industry ........ 100 Shuyao Zhou, Biswajit Mukhopadhyay, Piotr Mackowiak, Oswin Ehrmann, Kevin Kröhnert, Robert Gernhardt, Klaus-Dieter Lang, Michael Woratz, Peter Herrmann, N. Volkmer, Olaf Pohl, Volker Noack, Ha-Duong Ngo

Combinatorial Sensing of Catalytic Materials Using Terahertz Chemical Microscope ............................ 104 Yuji Hino, Yuki Kawakami, Kenji Sakai, Toshihiko Kiwa, and Keiji Tsukada

Amplification of Anodic Stripping Voltammetric Signals Recorded Using Screen Printed Electrodes in Weak Magnetic Fields ............................................................................................................... 107 A. Królicka, A. Bobrowski, M. Czarnota

Sensor Development of Electronic Tongue System for Taste Identification ............................................... 110 Rohini Mudhalwadkar

Analysis of Cocaine Using a Chemically Modified Electrode with Vanadium Hexacyanoferrate Film by Cyclic Voltammetry ............................................................................................................................ 112 I. C Eleotério, M. A Balbino, J. Magalhães, A. S Castro, B. R McCord, Marcelo F Oliveira

Multifunctional Sensor with Frequency Output Based on SOI TFT Double-gate Sensing Element ....... 114 V. N. Mordkovich, A. V. Leonov, A. A. Malykh and M. I. Pavluyk

Page 5: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

5

Terahertz-Time-of-Flight Method for Evaluating Cosmetic Samples Penetrating into Skin Samples ..... 117 T. Arisawa, T. Morimoto, K. Sakai, T. Kiwa and K. Tsukada

UWB Sensor Based Localization of Person with the Changing Nature of His/Her Movement ................ 120 Dušan Kocur, Daniel Novák,

3D Position Estimation with Capacitive Sensors for Touchless Interaction ................................................ 123 L. Haslinger and B. G. Zagar

Multi-functional Detectors of Ionizing Radiation on the Base of Anion-defective Alumina ..................... 128 S. V. Zvonarev, V. S. Kortov, S. V. Nikiforov

Design and Characterization of a 2D Array of MEMS Microphones for Acoustical Imaging ................. 130 A. Izquierdo, J. J. Villacorta, L. del Val and L. A. Suárez

GSM-GPS Based Security System and Its Implementation as Anti-Theft System in Automobiles .......... 133 Mandakinee Bandyopadhyay, Nirupama Mandal 2 and Subrata Chatterjee

Magnetometer Design for Copters .................................................................................................................. 138 V. Korepanov and F. Dudkin

Thermoelectric Single-photon Detector on the Base of W/(La,Ce)B6/W/Al2O3 Multi-layer Sensor ......... 140 A. S. Kuzanyan, A. A. Kuzanyan, V. R. Nikoghosyan, V. N. Gurin and M. P. Volkov

Ultrasonic Acoustic Beam Modulation in Water by Pre-fractal Geometries ............................................. 142 Juan Carlos Melgarejo, Sergio Castiñeira-Ibáñez, Daniel Tarrazó-Serrano, Constanza Rubio, Pilar Candelas, Antonio Uris

Development of Physical Computing Education Systems for Technical Colleges using Free Software .................................................................................................................................................... 144 T. Horiuchi, T. Miyazaki, Y. Yodo, Y. Yokoyama, H. Yamamoto and M. Masaaki

Optical Processor Design for Data Error Detection and Correction Using a (9,5 ) Binary Code Generator and the Syndrome Decoding Process .................................................................................. 147 M. A. Vieira, M. Vieira, P. Louro

MUX/DEMUX Si/C Device with Five Channel Separation in the Visible Range ...................................... 150 M. Vieira, M. A. Vieira, P. Louro

Page 6: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

6

Foreword

On behalf of the SEIA’ 2016 Organizing Committee, we introduce with pleasure these proceedings devoted to contributions from the 2nd International Conference on Sensors and Electronic Instrumental Advances 2016 held in Barcelona, Castelldefels, Spain. The conference is organized by the International Frequency Sensor Association (IFSA) and Asian Society of Applied Mathematics and Engineering (ASAME) in technical cooperation with IFSA Publishing and Excelera, S.L. The two-day program provides an opportunity for researchers interested in various applications of sensing and measurement to discuss their latest results and exchange ideas on the new trends. The main objective of the conference is to encourage discussion on a broad range of sensor related topics and to stimulate new collaborations among the participants. Since its inception, the Conference has enjoyed an atmosphere of high quality technical presentations on a broad range of multidisciplinary topics related to sensors, transducers, measurements and their applications. Extending the tradition that began in 2015 in Dubai, UAE, this Conference on sensors attracts over 50 researchers and practitioners in the sensor, field, sensor technology and application from around the world including 4 keynote speakers from a distinguished researchers of industry and academia from USA, Germany and China), who were invited to overview the progress in selected research trends. This year, we had more than 80 submissions from 25 countries (15 European and 10 non-European countries), from which 44 abstracts (31 oral and 13 posters) were selected for presentation at the Conference covering theory, design, device technology, and applications of sensors and sensing systems. To accommodate this range of interests, the 2nd SEIA’ 2016 Conference was organized in two dedicated poster sessions and oral tracks, including a special session on ‘Gas Sensors’. The proceedings contains all full-page papers of both: oral and poster presentations and keynote presentations. We hope that these proceedings will give readers an excellent overview of important and diversity topics discussed at the conference, including ‘Chemical Sensors’, ‘Gas Sensors’, ‘Nanosensor’, ‘Optical and Image Sensors’, ‘Sensing Methods and Technologies’ and ‘Measuring Systems’. Selected, extended papers will be submitted by the author(s) to the IFSA Sensors & Transducers journal based on the proceeding’s contributions. We thank all authors for submitting their latest work, thus contributing to the excellent technical contents of the Conference. Especially, we would like to thank the individuals and organizations that worked together diligently to make this Conference a success, and to the members of the International Program Committee for the thorough and careful review of the papers. It is important to point out that the great majority of the efforts in organizing the technical program of the Conference came from volunteers.

Prof., Dr. Sergey Y. Yurish Dr. Amin Daneshmand Malayeri

Page 7: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

7

Keynote Presentations

Page 8: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

8

Oxygen Gas Sensing in Combustion Process: from Basics to Application

P. Shuk

Rosemount Liquid and Gas Analysis, Emerson Process Management, 2400 Barranca Pkwy, Irvine, CA 92606, USA Tel.: +1 (949) 757-8530, fax: +1 (949) 474-7250

E-mail: [email protected] Abstract: Two major comprehensive oxygen gas sensing technologies, zirconia potentiometric and TDLS reviewed for the oxygen measurements in the combustion environment with many theoretical aspects and operation basics details. A comprehensive technologies review is supported with the latest developments trends. Performance and applications options for oxygen measurements in the process using zirconia potentiometric O2-probe and tunable diode laser (TDL) O2 system across the duct or as in-situ O2-probe at power generation, chemicals production, heating, process control, safety, and quality are discussed. Special attention is given to the technologies application limits and analyzer’s system requirements. Keywords: Combustion analysis, Flue gas, Oxygen gas sensing, Electrochemical sensor, Zirconia, Tunable diode laser spectroscopy (TDLS), Application options. 1. Introduction

Oxygen is vitally important not only to the existence of the human life but also very critical for any combustion process at chemicals production, power generation, heating, process control, safety, and quality. The oxygen is measured to optimize the combustion process efficiency, to control the product formation and quality and to minimize runaway combustion leading to the explosion [1]. Oxygen excess in the combustion is measured in the flue gas to controll the ideal safe minimum of oxygen and process efficiency (Fig. 1).

0

5

10

15

20

25

30

-4 -2 0 2 4 620

30

40

50

60

70

80

Gas

con

cent

rati

on (

%)

Oxygen concentration (%)

N2

CO2

COO

2

% Deficiency % Excess

H2O

Combustioncontrol with

~2...6% O2 excess

Gas

con

cent

rati

on (

%)

Fig. 1. Combustion flue gas concentration diagram.

The oxygen sensors used under ambient conditions include Clark type cells, paramagnetic and optical sensors. The high temperature oxygen sensors can be based on the potentiometric or amperometric cells or resistive semiconductors. Potentiometric sensors measure the equilibrium voltage of the cells according to the Nernst equation and would require the reference. The amperometric oxygen sensor with the oxygen

diffusion barrier limiting the current and fixed applied voltage can be only successfully applied in very clean environment and was discussed in many details elsewhere.

The two most comprehensive oxygen gas sensing technologies Zirconia potentiometric and Tunable Diode Laser Spectroscopy (TDLS) are dominating the most of in-situ and extractive combustion applications. 2. Zirconia Potentiometric Oxygen Technology

Zirconia potentiometric oxygen gas sensing technology widely used nowadays in O2 analyzers for different industrial combustion applications and in automotive Lambda sensors was invented in 1961 by Möbius and Peters [2] and Weissbart and Ruka [3] and was rapidly accepted in the power industry. The first industrial zirconia oxygen analyser for the process application was developed early 70th by Westinghouse Electric Co (Rosemount Analytical Inc.), and was based on the advanced platinum cermet electrode and zirconia solid electrolyte technologies developed initially for the solid oxide fuel cells.

All industrial zirconia oxygen sensors are based on an electrochemical cell with oxygen ion conducting yttrium stabilized zirconia (YSZ) solid electrolyte and two platinum based electrodes printed and sintered on the opposite sides of the zirconia ceramic and exposed to the process and reference gases.

Differential oxygen chemical potentials on the oxygen sensor electrodes would develop electromotive force (EMF), E, according to the Nernstian equation:

Ref'

2

Process"

2

)(

)(ln

4 Op

Op

F

RTE (1)

with R universal gas constant, T the process temperature in Kelvin (K) and F the Faraday number.

Page 9: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

9

By using fixed oxygen partial pressure on the reference electrode, e.g., air with p(O2)= 2.1×104 Pa, the sensor signal of the thermally balanced oxygen sensor will be only depending on the sensor temperature. Zirconia potentiometric oxygen gas analysis was established by H.-H. Mӧbius [4] and is permitting measurements of chemically bonded oxygen in the reducing conditions with very low equilibrium concentration of oxygen (<1ppb) released at high temperature (Fig. 2).

0

500

1000

1500

700 800 900 1000 1100 1200 1300

E (

mV

)

T (oC)

10-1

10-3

10-5

103

10

C

c(O2)

(%)

10-20

10-15

10-4 =1ppm10-3

10-2

0.10.418

10-5

10-10

CO2

c(CO)/c(CO2)

Fig. 2. Oxygen sensor concentration diagram.

Zirconia O2-probe (Fig. 3) was applied with some very good success in the combustion control of power generation boilers, hot stoves for steelmaking, heating and combustion exhaust gas control of coke ovens for steelmaking, lime and cement kilns combustion control, incinerator combustion control, combustion control of heating furnaces for oil refinery and petrochemical industry and many other applications.

O -sensor2

Fig. 3. In-situ O2-Analyzer.

There have been numerous attempts to miniaturize oxygen sensor with innovative reference like sealed chamber and oxygen pump or glass sealed metal/metal oxide references electrode but the evaluation couldn’t confirm these designs reliability.

2. Tunable Diode Laser Spectroscopy (TDLS)

Tunable diode laser spectroscopy (TDLS) as an innovative optical measurement technique utilizing semiconductor lasers to detect a variety of gases in the near infrared (IR) range [5] has become an established method for non-intrusive measurements in combustion

environment [6]. TDLS is highly distinguishing from the conventional process photometry by the ability of the laser to be scanned across the narrow oxygen absorption peaks (Fig. 4) many times per second by trimming the current through the laser. With typical scan in the range of 0.2…0.3 nm laser would provide much better resolution and selectivity. Wavelength-modulation spectroscopy (WMS) has been increasingly applied to the measurements improving sensitivity and selectivity especially in the noisy combustion environment [7]. Sensitivity can be also father improved by using specific microstructures such as Hollow-Core Photonic Bandgap Fibres (HC-PBFs) or high Q cavities that provide longer absorption path lengths for the gases. TDL O2 analyzer would provide near real time measurements with no cross interference to other gases in the combustion environment.

Fig. 4. Oxygen absorption spectrum at 760 nm.

With the increasing temperature oxygen absorption intensity will be reduced bringing some challenges in the weaker intensity interpretation and temperature variation compensation contributing to signal to noise ratio reduction. Temperature and pressure variation in the process has to be compensated and might be bringing an additional error in the oxygen measurements.

For the alignment reliability across the gas stream flexible bellows on the installation flanges were implemented in some TDL O2-Analyzers. Calibration and validation of TDL O2 analyzer on line is still waiting for a conclusive answer. Implemented in some TDL instrumentation a reference cell is very short to provide reliable in-situ validation. IR light reflection at high temperature combined with wide background radiation from the fire box and process windows fouling might bring additional challenge in the application and might reduce the service life of the analyzer. Fiber coupled laser solution increases the optical noise (x5...7) and would reduce the sensitivity with detection limit to be ~500 ppm. References [1]. P. Shuk, Process Zirconia oxygen analyzer - state of

art, Technisches Messen, Vol. 77, Issue 1, 2010, pp. 19-23.

Page 10: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

10

[2]. H. Peters and H.-H. Möbius, Procedure for the gas analysis at elevated temperatures using galvanic solid electrolyte elements (Germ), DD-Patent 21673, 1961.

[3]. J. Weissbart and R. Ruka, Oxygen gauge, Rev. Sci. Instrum., Vol. 32, 1961, pp. 593-595.

[4]. H.- H. Möbius, Basics of oxygen gas potentiomitric analysis (Germ), Z. Phys. Chem., Vol. 230, Issue 5-6, 1965, pp. 396-416.

[5]. E. D. Hinkley and P. L. Kelley, Detection of air pollutants with Tunable Diode Lasers, Science, Vol. 171, Issue 3972, 1971, pp. 635–639.

[6]. R. K. Hanson, Applications of quantitative sensor laser to kinetics, propulsion and practical energy systems. Proceedings of the Combustion Institute, Vol. 33, 2011, pp. 1-40.

[7]. G. B. Rieker, J. B. Jeffries and R. K Hanson, Calibration-free wavelength-modulation spectroscopy for measurements of gas temperature and concentration in harsh environments, Appl. Opt., Vol. 48, Issue 29, 2009, pp. 5546–5560.

Page 11: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

11

Solid Electrolyte Gas Sensors based on Voltammetric Modes

Prof. Ulrich Guth

Kurt-Schwabe-Institute for Measurements and Sensortechnique, Waldheim, Germany E-mail: [email protected]

Abstract: Solid electrolyte cells widely used as gas sensors for emission monitoring in exhaust will be reviewed in details. Most of such electrochemical sensors work in steady state mode, e.g. in potentiometric or amperometric mode. Non-steady methods with definite change of potential and mixed potential sensors open new opportunities to obtain beside oxygen also the concentration of carbon monoxide (CO), nitric oxides (NOx) and hydrogen (H2). By varying the potential linearly with time and back to the starting potential a specific response in the form of peaks is obtained (cyclovoltammery). Reduction peaks (oxygen) and oxidation peaks (hydrogen) can be clearly distinguished so that the simultaneous determination of both gases e.g. in bio gas are possible. Mixed potential sensors having two different catalytically active electrodes on zirconia are used to determine one combustible beside oxygen, e.g. CO. By varying the operating temperature in a triangle mode two or more components like CO and H2 beside oxygen can be determined simultaneously.

Page 12: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

12

Sensors in the Era of Internet of Things and Industrial Internet:

Innovations in Materials, Transducers, and Data Analytics

R. A. Potyrailo GE Global Research, Niskayuna, NY, USA

Tel.: (518) 387-7370 E-mail: [email protected]

Summary: Modern monitoring scenarios of gases and liquids for industrial safety and productivity, medical diagnostics, environmental surveillance, and other Internet of Things and Industrial Internet applications demand new sensing performance capabilities of higher reliability, better accuracy, lower power consumption; all of these in unobtrusive form factors and at low cost. A new generation of gas and liquid sensors based on the multivariable response principles are under development in our laboratories to meet these new demanding sensing requirements. Multivariable sensors provide several partially or fully independent responses from an individual device to allow quantitation of several environmental parameters of interest, rejection of interferences, and correction for environmental instabilities. This innovative multivariable sensing approach is attractive in numerous scenarios when manual inspection or the use of “classic” analytical instruments would be canceled by given application requirements. Keywords: Monitoring of gases and liquids, Multivariable sensing principles, Internet of Things, Industrial Internet, High reliability, High accuracy, Low power consumption.

1. Introduction

Gas and liquid sensors provide tunable sensitivity,

continuous monitoring capability, small power consumption, and operation without consumables. Unfortunately, existing sensors also often have several performance limitations such as inability to preserve detection accuracy in the presence of interferences and sensor drift, especially noticeable in detection of low analyte levels. As a result, these limitations often can revoke advantages of sensors in their intended practical applications. Thus, field uses of gas and liquid sensors are often most successful when their poor selectivity is not important, concentrations of measured analyte gases or liquids are high enough to make the sensor drift unnoticed, or frequent recalibration is acceptable.

We introduce new generation of gas and liquid sensors such as multivariable sensors into applications where manual inspection or the use of “classic” analytical instruments is prohibitive due to the application requirements.

2. Diversity of Applications Modern applications of sensors demand their high reliability and accuracy, low operation power, unobtrusive form factors, low cost, and standard and secure communication [1-4]. Sensors that meet these requirements are attractive for implementations for Internet of Things and Industrial Internet demanding applications as summarized in Fig. 1. The Internet of Things is the network of everyday objects with embedded networked sensors to increase the value of these objects [5]. The Industrial Internet is the integration of machinery with networked sensors [6]. The use of sensors in these applications provides multi-billion-dollar market opportunities [1, 2, 5, 6].

Fig. 1. Examples of developments of multivariable gas and liquid sensors for (A) Internet of Things and (B) Industrial

Internet applications. Examples of measured gases and volatiles of interest for these applications include environmental background (e.g. O2, CO2, H2O), atmospheric pollutants (e.g. CO2, CO, O3, H2S, NH3, NOx, SO2, CH4, industrial fumes, waste odors), breath biomarkers (e.g. NO, H2S, NH4, acetone, ethane, pentane, isoprene), and public/homeland safety hazardous volatiles (e.g. toxic industrial chemicals, chemical warfare agents, explosives). Examples of measured liquids of interest for these applications include

Page 13: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

13

potable/waste/industrial water, bioprocess streams, lubricating oils, fuels, and some others. 3. Solving Unmet Monitoring Needs with

Multivariable Sensors Multivariable sensors provide several partially or

fully independent responses from an individual device to allow quantitation of several environmental parameters of interest, rejection of interferences, and correction for environmental instabilities [3, 4]. Design criteria for multivariable sensors bring a fundamentally new sensing philosophy that allows response to interferences with these new sensors but in different directions vs. analyte response (Fig. 2A). This new sensing philosophy is the major departure from the existing sensing approach where conventional sensors cannot differentiate between the analyte and interferences (Fig. 2B).

Fig. 2. Multivariable sensors as a new generation of gas and liquid sensors. (A) New sensing philosophy with multivariable sensors that allows response to interferences but in different directions than the analyte response. (B) Existing sensing approach where a conventional sensor does not differentiate between the analyte and interferences.

For multivariable gas sensing, we have employed several types of sensing materials, including dielectric and conducting polymers, carbon nanotubes, graphene, metal oxides, and metal nanoparticles. These sensing materials can operate not only at ambient room temperature but also at elevated temperatures, attractive for in situ monitoring of high-temperature industrial processes [4]. To identify diverse effects of sensing materials to different gases, we employed three transducer types such as electrical, photonic, and electromechanical transducers [4]. Diverse data analytics tools, originally developed for sensor arrays,

were adapted for analysis of data from multivariable sensors [4].

A comparison of different types of multivariable gas sensors developed in our laboratories and by other research teams is presented in Fig. 3 [4].

Multivariable sensors based on electrical and other transducers have been developed for measurements of diverse liquids of interest for industrial applications [7].

Fig. 3. Comparison of different types of reported multivariable sensors. The areas of each segment are

proportional to the number of publications [4]. 4. From Sensor Ideas to Products To bring a new sensor idea from laboratory to a practical application, this sensor must demonstrate a new value to users by outperforming existing sensors and other technical solutions, for example manual inspection or “classic” analytical instruments. Sensor commercialization process can be described using technology readiness levels (TRLs) and manufacturing readiness levels (MRLs) [8-9]. Examples of important tasks in commercialization of new sensors are outlined in Fig. 4. The goals of these tasks are to eliminate technical risks, develop manufacturing infrastructure and supply chain, to build a commercial roadmap, and to assemble a commercial team to deliver new sensors to customers.

Fig. 4. Tasks in commercialization of new sensors.

Page 14: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

14

Development of new sensor products requires money and time investments. Money investment is typically 1 : 10 : 100 : 1000 when advancing from proof-of-concept, to working prototype, to pilot scale production, and to product launch [10]. Typical time investments are several years [11-12]. 5. Outlook The key metric for acceptance of new sensors in diverse applications will remain to be the cost-to-benefit ratio to the users. The value of multivariable sensors will continue to increase with their ability to discriminate and quantify multiple gases and liquids in the presence of known and unknown interferences and to correct for multiple environmental effects without the added system hardware complexity. References [1]. Markets for Sensors in the Internet of Things 2014-

2021, Report #Nano-750, 2014 www.nanomarkets.net [2]. Markets for Sensors in the Industrial Internet, Report

#Nano-752, 2014, www.nanomarkets.net [3]. R. A. Potyrailo, C. Surman, N. N. Nagraj, A. Burns,

Materials and Transducers Toward Selective Wireless Gas Sensing, Chemical Reviews, Vol. 111, 2011, pp. 7315–7354.

[4]. R. A. Potyrailo, Multivariable sensors for ubiquitous monitoring of gases in the era of Internet of Things and

Industrial Internet, Chemical Reviews, DOI:10.1021/acs.chemrev.6b00187, 2016.

[5]. O. Vermesan, P. Friess, P. Guillemin, S. Gusmeroli, H. Sundmaeker, A. Bassi, et al., Internet of Things Strategic Research Roadmap, in Internet of Things: Global Technological and Societal Trends, O. Vermesan and P. Friess, Eds., River Publishers, 2011, pp. 9-52.

[6]. P. C. Evans, M. Annunziata, Industrial Internet: Pushing the Boundaries of Minds and Machines, www.ge.com/docs/chapters/Industrial_Internet.pdf, 2012.

[7]. R. A. Potyrailo, C. Surman, D. Monk, W. G. Morris, T. Wortley, M. Vincent, et al., RFID Sensors as The Common Sensing Platform for Single-Use Biopharmaceutical Manufacturing, Measurement Science and Technology, Vol. 22, 2011, art. No 082001.

[8]. Technology Readiness Assessment (TRA) Guidance, Department of Defence, Prepared by the Deputy Under Secretary of Defense for Science and Technology (DUSD(S&T)), www.acc.dau.mil, 2005.

[9]. Manufacturing Readiness Level (MRL) Deskbook, OSD Manufacturing Technology Program, www.dodmrl.com/MRL_Deskbook_V2.pdf, 2011.

[10]. G. M. Whitesides, Cool, or simple and cheap? Why not both?, Lab on a Chip, Vol. 13, 2013, pp. 11–13.

[11]. R. A. Potyrailo, W. G. Morris, A. M. Leach, L. Hassib, K. Krishnan, C. Surman, et al., Theory and Practice of Ubiquitous Quantitative Chemical Analysis Using Conventional Computer Optical Disk Drives, Applied Optics, Vol. 46, 2007, pp. 7007-7017.

[12]. General Electric TrueSense Personal Water Analytics, The Prism Awards for Photonics Innovation Winners, www.photonicsprismaward.com/winners.aspx, 2011.

Page 15: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

15

Ion Beam Nanofabrication and Its Applications in Precision Metrology

and Bio-sensing

Dr. Zongwei Xu Centre of MicroNano Manufacturing Technology, Tianjin University, China

E-mail: [email protected] Abstract: Nanomanufacturing is vital and acts as the bridge and foundation for the design and application of nanotechnology, such as, nano-optics and bio-sensing, etc. Ion beam nanofabrication technology has an important status in nanomanufacturing technology for its advantages of direct writing and high precision.

The nanoscale effect of multi-parameters’ coupling mechanism involved in focused ion beam (FIB) nanofabrication during nanostructure patterning was discovered, which effectively conquered the key problem of FIB fabrication of nanostructure arrays with high density and aspect ratio, and realized the nanofabrication of nano-photomask that has a line width of 32 nm with a length-width ratio of more than 100:1. Nano-star functional structure was developed and applied in precision metrology. By measuring the developed nano-star structures, the measurement capabilities in lateral and longitudinal directions for optical microscopy were evaluated.

Novel nanofabrication methods were proposed to fabricate nanostructures with gap size less than 10 nm, including FIB nanofabrication combining with subsequent film coating, film over nanosphere surfaces (FON) with ion sputter coating, which were applied to nanofabricate surface-enhanced Raman scattering (SERS) substrate for bio-sensing. The configuration of the Au-Polystyrene spheres can be regulated to hexagonal close packing with nanoscale V-shaped slits with less than 10 ~ 20 nm gap pattern. Nanoscale Au clusters with a clear outline were covered the surface of PS spheres. The developed hierarchical nanostructures substrate showed good SERS performance.

Page 16: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

16

Regular Papers

Page 17: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

17

Trace Gas Measurements with Zirconia Sensors: An Overview

M. Schelter 1, J. Zosel 1, V. Vashook 1, U. Guth 1,2, W. Oelßner 1 and M. Mertig 1,2

1 Kurt-Schwabe-Institut für Mess- und Sensortechnik e.V. Meinsberg, Kurt-Schwabe-Strasse 4, 04736 Waldheim, Germany

2 Technische Universität Dresden; Physikalische Chemie, Mess- und Sensortechnik, 01062 Dresden, Germany Tel.: + 49 34327 608 120, fax: + 49 34327 608 131

E-mail: [email protected] Summary: Since zirconia-based solid electrolyte sensors are economic, space-saving and long-term stable devices, they could be an useful alternative for mass spectrometry systems for the determination of traces of volatile substances in gas mixtures. Here we give an overview on sensor related parameters of zirconia-based solid electrolyte sensors, such as selectivity and detection limit. Furthermore, we show two possibilities for increasing the selectivity of these sensors and also that it is possible to decrease the lower limit of detection into the vol.-ppb range that corresponds to a lowering of about 4 orders of magnitude compared to commercially available state-of-the-art solid electrolyte sensors. Keywords: Solid electrolyte sensor, Sensitivity, Selectivity, Coulometry, Cyclic voltammetry.

1. Introduction The determination of traces of volatile substances in gas mixtures is highly informative - not only in noninvasive health evaluation like human breath gas analysis [1], but also for monitoring the quality of air [2] as well as for safety and security aspects of several technical processes. Commercially available measuring systems show good detection limits, but are expensive and often have huge dimensions due to the use of mass spectrometers [3]. In contrast to this, zirconia-based solid electrolyte gas sensors (SESs) are economic, space-saving and long-term stable devices for the determination of traces of volatile substances in gas mixtures. Here we give an overview on sensor related parameters of solid electrolyte devices, such as selectivity and detection limit. Furthermore, we present possibilities for increasing the selectivity and for lowering the limit of detection of coulometrically operated SES.

2. Zirconia-based Solid Electrolyte Sensors SESs are advantageous devices for trace gas

analysis due to their high sensitivity and excellent long-term stability. Since stabilized zirconia, e.g. yttria-stabilized zirconia (YSZ), feature high and selective oxygen ion conductivity and low electronic conductivity, they are preferred materials for SESs. Equipped with platinum electrodes and heated to temperatures between 650 °C and 800 °C, these materials are broadly used as potentiometric, amperometric or coulometric sensors.

Operated in the modes shown in Table 1, SESs show hardly any selectivity due to the equilibration of oxidizable and reducible gases and oxygen on the hot platinum electrodes. Consequently, the measuring signal gives no information about number, type and concentrations of gases in gas mixtures.

Table 1. Selected parameters of solid electrolyte sensors.

Measuring principle Potentiometric Amperometric Coulometric

schematic measurement configuration:

measuring signal: 2

2

( )ln

( )

I

c II

p ORTU

zF p O 2( )dI p O 2( )FI n O z F

detection limits (for O2): 10-30 … 1 vol.-ppm 10-4 … 1 atm 10-5 … 1 atm general selectivity: low low low

Page 18: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

18

The calibration-free and temperature-independent operation based on Faraday’s law makes coulometric SESs most advantageous candidates for reliable and practicable trace gas sensors [4]. Different measures are shown to improve the selectivity and to level down the detection limits of this sensor type. With the shown adjustments, it is possible to measure traces of H2 in the vol.-ppb range in gases where also other oxidizable and reducible components like CH4 and O2 are present.

3. Improving the Selectivity of Solid Electrolyte Sensors

The potentiodynamic operation of a SES is one possibility to increase its selectivity. Fig. 1 proves that the determination of hydrogen is not influenced by oxygen in the measuring gas, if the SES equipped with platinum electrodes is operated with cyclic voltammetry.

Fig. 1. Cyclic voltammograms with a SES on measuring gases with the same hydrogen concentration but different oxygen concentrations each; sensor temperature: 700 °C; potential scan rate: 10 mV/s.

Another possibility to increase the selectivity of a SES is to operate it downstream of a gas chromatograph (GC). Fig. 2 shows that it is possible to quantify hydrogen, oxygen and methane concentrations with very high selectivity.

4. Lowering the Limit of Detection of Coulometric Solid Electrolyte Sensors

A newly developed low-noise current measuring unit, an adjusted temperature control of the YSZ cell and optimized operating conditions of the cell with regard to a high ratio of ionic to electronic conductivity ensure an elevated signal-to-noise-ratio of a coulometric SES [5]. Under these conditions, it was possible to decrease the lower limit of detection into the vol.-ppb range that corresponds to a

lowering of about 4 orders of magnitude compared to commercially available state-of-the-art sensors.

Fig. 2. Chromatogramms of a GC equipped with a potentiostatically operated coulometric SES; GC columns: silicagel + molecular sieve 13X (50 °C const.); sample volume: 1 ml; sensor temperature: 700 °C; polarization voltage applied between “YSZ|Pt, measuring gas” electrode and “YSZ|Pt,air” electrode of the SES: -450 mV. Acknowledgements

The investigations were carried out partially within the project SEVERA (support code 03KB067B) funded by the German Ministry of Environment, Nature Conservation, Building and Nuclear Safety (BMU). The authors are responsible for the content of this paper and are thankful for the funding. References [1]. A. Amann, G. Poupart, S. Telser, M. Ledochowski,

A Schmid, S. Mechtcheriakov. Applications of breath gas analysis in medicine, International Journal of Mass Spectrometry, Vol. 239, Issue 2-3, 2004, pp. 227-233.

[2]. S. Król, B. Zabiegała, J. Namieśnik. Monitoring and analytics of semivolatile organic compounds (SVOCs) in indoor air, Analytical and Bioanalytical Chemistry, Vol. 400, Issue 6, 2011, pp. 1751-1769.

[3]. M. S. El-Shahawi, A. Hamza, A. S. Bashammakh, W. T. Al-Saggaf, An overview on the accumulation, distribution, transformations, toxicity and analytical methods for the monitoring of persistent organic pollutants, Talanta, Vol. 80, Issue 5, 2010, pp 1587-1597.

[4]. V. Vashook, J. Zosel, U. Guth. Oxygen solid electrolyte coulometry (OSEC), Journal of Solid State Electrochemistry, Vol. 16, 2012, pp. 3401-3421.

[5]. M. Schelter, J. Zosel, W. Oelßner, U. Guth, M. Mertig. A Solid Electrolyte Sensor for Trace Gas Analysis, Sensors and Actuators B: Chemical, Vol. 187, 2013, pp. 209-214.

Page 19: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

19

Field Effect Based Gas Sensors, from Basic Mechanisms to the Latest

Commercial Device Designs

M. Andersson 1,2, P. Möller 1, H. Fashandi 1, J. Eriksson 1, D. Puglisi 1, J. Huotari 2, J. Puustinen 2, J. Lappalainen 2 and A. Lloyd Spetz 1,2

1 Div. of Applied Sensor Science, Linköping University, SE-581 83 Linköping, Sweden 2 Team Lappalainen, Oulu University

Tel.: + 46 723 282327 E-mail: [email protected]

Summary: This contribution treats the latest developments in the understanding of basic principles regarding device design, transduction mechanisms, gas-materials-interactions, and materials processing for the tailored design and fabrication of SiC FET gas sensor devices, mainly intended as products for the automotive sector. Keywords: SiC, Field Effect Transistors, Gas sensors, Automotive, Emissions monitoring, Sensor product development.

1. Introduction

Following steady maturation of the SiC material and corresponding processing procedures over the past thirty years, the last decade has seen the introduction of the first SiC based Field Effect Transistor devices for power electronics on the market [1]. The wide band gap and chemical inertness of SiC also render the material interesting in low voltage, high temperature (> 200 °C) applications, e.g. in the fabrication of gas sensors for exhaust and flue gas emissions monitoring in the automotive sector. On the brink of more large-scale commercialization this contribution therefore treats the latest developments in SiC FET devices for use in gas sensing applications, from the understanding of basic mechanisms to product design principles.

1.1. FET Based Gas Sensors – Basic Design

The basic design of SiC FET based gas sensors is displayed in Fig. 1(a). By applying as gate contact in traditional MOSFETs (Metal Oxide Semiconductor Field Effect Transistor) a material (commonly a catalytic metal) which interacts with one or more gaseous substances such that the gate to substrate electric field is modulated, FET based gas sensors can be fabricated.

Fig. 1. Basic cross-sectional design of a field effect transistor based gas sensor.

1.2. FET Based Gas Sensors – Basic Principles

When, for example, exposing a Pt gate MOSFET structure to hydrogen, atomic hydrogen is generated through dissociative adsorption on the catalytic metal surface. Adsorbed hydrogen atoms subsequently diffuse through the thin Pt film and adsorb on top of oxygen atoms in the surface of the oxide, see Fig. 2(a). The resulting hydroxide groups give rise to a change in polarization of the metal/ oxide interface, affecting the gate-to-semiconductor electric field. Two more principles by which this electric field can be modulated is through the adsorption on the oxide surface of ions, e.g. oxygen anions, and a change in the gate metal work function. The change in gate-to-semiconductor electric field in turn modulates the number of charge carriers in the top part of the semiconductor and thus the Ids/Vds-characteristics of the device (Fig. 2(b)). Either the change in drain current upon gas exposure when supplying a constant drain voltage or vice versa, is taken as the sensor output [2]. 2. Basic Mechanism and Process

Developments 2.1. Transducer Platform

Over the last few years certain design rules for the transistor device, when operated as a transducer for gas sensing, have been established. From long-term reliability studies, a depletion-type device has been concluded to be the preferred choice regarding transistor design. Furthermore, to ensure high sensitivity to even low concentrations the gate dielectrics should be as thin as possible, see Fig. 3, but no less than about 35 nm and preferably include a high-k material, whereas the gate contact width/length ratio should be as large as possible, within limits.

Page 20: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

20

Fig. 2. (a) Change in the number of mobile charge carriers in the top part of the semiconductor upon gas exposure, (b) Resulting change in I/V characteristics

Fig. 3 Increase in sensor response (sensor voltage signal change, ΔV) upon a decrease in gate insulator thickness

(dins), as exemplified by the exposure of the sensor devices to different concentrations of ammonia.

2.2. Gate Insulator Surface

In order to tune the transducer platform to fulfil certain requirements on the final sensor product in terms of operating temperature and gas sensitivity/ dynamic range, basic principles for the choice of top insulator material have been investigated. Studying the dynamic adsorption/ desorption characteristics of NO and NO2 on conceptually different oxide materials at different temperatures it has been possible to tune the maximum sensitivity of the final NOx sensor close to the desired 600 °C operating temperature (see Fig. 4). With the use of pulsed laser deposition of gate insulator and gas-sensitive materials under different conditions it has also been possible to tune the nano- and micro-structure of these materials, from amorphous to crystalline, and from dense, basically epitaxial, to porous thin-films.

2.3. Gate Materials

By the choice of gas-sensitive gate material or the addition of promoters, it has furthermore been shown possible to tune both the selectivity towards different gases and reduce the interference of certain other substances on the sensor signal. As an example, the addition of certain monolayer metal oxides to Pt thin-films have eliminated the cross-sensitivity of NH3 sensing to CO in some applications.

Fig. 4. Response of an optimized SiC-FET sensor device to different concentrations of NO2, for which it can be sen that

the maximum sensitivity is achieved for an operation temperature close to the desired 600 °C.

3. Product Development

From the developed principles and understanding of basic sensor mechanisms, the tuning of sensor properties by the combination of transducer design and operation, different materials and the choice of

Page 21: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

21

processing methods and parameters new SiC FET products may more efficiently be brought to the market. In combining depletion-type devices with the suitable insulator and gate materials, utilizing optimized deposition/ fabrication methods and parameters two new SiC FET sensors, for the monitoring of intake-oxygen concentrations and tailpipe NOx emissions in the automotive sector, have recently entered the “proof-of-concept”-stage for commercialization.

4. Conclusions

With the latest developments regarding possibilities to predict the outcome when tuning

certain design parameters and processes, the SiC FET based platform currently is on the verge of more large-scale commercialization of sensor products. References [1]. M. Östling, R. Ghandi, C.-M. Zetterling, SiC power

devices – present status, applications and future perspective, in Proceedings of the 23rd Intl Symposium on Power Semiconductor Devices & IC's, May 23-26, 2011, San Diego, CA.

[2]. M. Andersson, R. Pearce, and A. Lloyd Spetz, New Generation SiC based Field Effect Transistor Gas Sensors, Sens. Actuat. B, 179, 2013, pp. 95-106.

[3]. PLD reference.

Page 22: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

22

Application of New Ultrasonic Instrumentation to Real-time Monitoring

and Analysis of Binary Gases Mixtures

M. Battistin 3, S. Berry 3, P. Bonneau 3, O. Crespo-Lopez 3, C. Deterre 4, M. Doubek 6, G. Favre 3, G. Hallewell 7, S. Katunin 8, D. Lombard 3, A. O’Rourke 4, A. Madsen 4, S. McMahon 9, K. Nagai 10, C. Rossi 1*, B. Pearson, D. Robinson 11,

A. Rozanov 7, E. Stanecka 12, M. Strauss 2, V. Vacek 6, R. Vaglio 3 and J. Young 2 1 INFN Genova, via Dodecaneso 33, 16146 Genova, Italy

2 Department of Physics and Astronomy, University of Oklahoma, Norman, OK 73019, USA 3 CERN, 1211 Geneva 23, Switzerland

4 DESY, Notkestraße 85, Hamburg 22607, Germany 6 Czech Technical University, Technická 4, 166 07 Prague 6, Czech Republic

7 Centre de Physique des Particules de Marseille, 163 Avenue de Luminy, 13288 Marseille Cedex 09, France 8 B.P. Konstantinov Petersburg Nuclear Physics Institute (PNPI), 188300 St. Petersburg, Russia

9 Rutherford Appleton Laboratory, Science & Technology Facilities Council, Chilton, OX11OQX, UK 10 Department of Physics, Oxford University, Keble Road, Oxford OX1 3RH, UK

11 Department of Physics, Cambridge University, J. J. Thomson Avenue, Cambridge, CB3 0HE, UK 12 Institute of Nuclear Physics PAS, ul. Radzikowskiego 152, 31-342 Kraków, Poland

* Tel.: +390103536236 E-mail: [email protected]

Abstract: Custom ultrasonic instruments have been developed for continuous monitoring and real-time measurement of composition and flow in binary gas mixtures. These characteristics are derived from measurements of sound transit time along two opposite directions - parallel or tilted to the gas flow direction. The flow rate is then calculated from the transit time difference while the average is used to compute sound velocity by comparison with a sound velocity/composition database. Five devices are integrated in the Detector Control System of the ATLAS experiment at the CERN Large Hadron Collider. Three instruments monitor C3F8 and CO2 coolant leaks into the nitrogen-purged envelopes of the inner silicon tracker; respectively with precisions better than ±2 x 10-5 and ± 10-4. Two further instruments are used to monitor the new thermosiphon C3F8 evaporative coolant recirculator. One of these measures C3F8 vapour return flow to the condenser while the other tracks air ingress into the condenser. The precision of these instruments highlights their potential in other applications requiring continuous binary gas composition monitoring. Keywords: Ultrasonic binary gas analysis, Ultrasonic instrumentation, ATLAS Detector Control System, Sonar, thermosiphon, Evaporative cooling system.

1. Introduction

Custom ultrasonic (“sonar”) instruments have been

developed for real-time monitoring and measurement of binary gas mixtures in the ATLAS experiment, one of the major experiments at the CERN Large Hadron Collider (LHC). ATLAS is a particle physics detector consisting of a series of concentric sub-detectors arranged around one of the LHC proton beam collision points [1]. The innermost sub-detector - the ATLAS Inner Detector (ID) - tracks charged particles in a solenoidal magnetic field using gas-based tracking elements and high resolution silicon detectors in microstrip and pixelated configurations. The silicon sub-detectors are closest to the beam collisions and must be kept at low temperature (-10 ⁰C or lower) to mitigate radiation damage. They are evaporatively cooled using C3F8 (octofluropropane) and CO2. Three of five ultrasonic instruments monitor coolant leaks into the nitrogen-purged envelopes of these silicon tracking detectors. Two further instruments form part of the control system of the new thermosiphon C3F8

evaporative cooling recirculator plant [2], being commissioned to replace the present compressor-driven system. One (angled acoustic path) instrument was built to measure the C3F8 vapour flow returning to the thermosiphon condenser while another detects and eliminates (“degasses”) ingressed air from the condenser. The locations of the five instruments are shown in Fig. 1.

2. The Instrument and its Operating Principle

The instruments contain facing pairs of 50 kHz capacitive ultrasonic transducers within flanged envelopes through which gas flows. When transmitting, a transducer is excited by (3000V) square pulses, generated from low-voltage precursors in a dsPIC33F microcontroller. A 40 MHz transit time clock is started synchronously with the leading edge of the first transmitted pulse. The acoustic signal received by the other transducer is passed through amplifiers and a comparator, stopping the clock when a user-

Page 23: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

23

defined threshold is crossed. The micro-controller also reverses the transmission direction for the transit time to be measured in both directions. These transit times

are used with geometrical parameters to evaluate the gas flow rate.

Fig. 1. Installations in the ATLAS Inner Detector. Three sonars monitor the N2 envelopes of the (“Pixel, “SCT” & “IBL”) silicon sub-detectors, while a degassing sonar and an angled flowmeter monitor the thermosiphon recirculator.

The ultrasonic instruments can be used for simultaneous gas flow measurement and composition analysis, exploiting the physical phenomenon whereby - at known temperature and pressure - the sound velocity in a binary gas mixture depends on the molar concentrations of its components. For binary gas analysis, the sound velocity is calculated from the average of the transit times in opposite directions and then compared in real-time to a stored concentration vs. sound velocity database for a given temperature and pressure. This database can be generated from theoretical models and/or from previously-made calibration mixtures [3]. The instruments are integrated in the Detector Control System (DCS) of the ATLAS experiment. A schematical diagram of the wire protocol eletrocnics is shown in Fig. 2. The main parameters of the sonar system (transit time counter, temperature and pressure readout, valve and pump driver and 4-20 mA DAC) are controlled by dedicated modules. All the modules are in turn controlled by the main controller and communication is done through the main controller using a standard TCP/IP Modbus proticol over Ethernet. 2. Application and Commissioning of the

Instruments in the ATLAS Thermosiphon Cooling Recirculator

Two instruments are directly included in the new

ATLAS thermosiphon C3F8 coolant circulation system

- as shown in Fig. 1. The thermosiphon takes advantage of the typical configuration of the LHC experiments located around 100 m underground. It is based on the physical phenomenon of natural circulation: C3F8 is condensed at a lower temperature/pressure but at the highest elevation in the system. In ATLAS the driving force of the circuit is given by the 92 m liquid column, starting from the condenser.

Fig. 2. Diagram of the 2-wire protocol eletronics.

Page 24: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

24

The C3F8 refrigerant exits the detector at -25 °C and 1.67 barabs and returns to condenser against gravity due to pressure differential. The thermsiphon system must provide 60 kW of on-detector cooling capacity, and guarantee a vapour pressure of 1.67 barabs at the end of the on-detector cooling loops for stable operation of the silicon tracking detector components in the high radiation zone near the LHC beams. To achieve 60 kW cooling capacity 1.2 kg·s-1 of C3F8 must be circulated and condensed. A diagram of the thermosiphon plant is shown in Fig. 3. The purpose of the new cooling system is to replace the compressors of the present

cooling plant recirculator, removing “active” (reciprocating) components from the primary cooling loop to a more reliable system requiring less maintenance.

The degassing sonar (Fig. 1) represents the key tool for detection and elimination of air leaks into the thermosiphon plant. The instrument is placed on the top of the thermosiphon condenser, which is the element at the lowest temperature and pressure of the cooling system (-60 °C, 309 mbarabs for pure C3F8)

Fig. 3. Schematic and thermodynamic phase diagram of the new thermosiphon cooling system. and therefore the point where any air leaks will accumulate. The ingress of air must be avoided since an increase in condenser pressure will erode the pressure differential needed to circulate the 1.2 kg·s-1 C3F8 vapour flow back to the condenser against gravity.

During the commissioning of the TS with C3F8 circulation through a dummy load in July 2016 several tests were performed on the degassing (DG) sonar. Fig. 4 illustrates results with condenser containing liquid C3F8 with a starting pressure 410 mbarabs. Around 12.00 on July 12 2016 20 liters of N2 were injected into the lowest point of the vapour side of the TS recirculator (around 90 metres below the condenser) to simulate an air ingress, without the risk of icing. A rise in apparent air concentration of 2.6 %

in the vapour seen by the DG sonar was observed over the next 24 hours as the N2 gas migrated through C3F8 vapour in the chicane and baffle structure of the condenser toward the DG sonar. This increase was relative to a baseline level of around 26 % from previous N2 injections and the unvented accumulation from small leaks encountered and repaired over several weeks of running. Also shown in Fig. 4 are the temperature of the sonar tube itself (thermally insulated from the condenser beneath) and its pressure, which reflects that of the condenser by communication. The rise in apparent “air” concentration is not affected by the regular day/night temperature variations of the sonar tube, and is indicative of the robustness of the “air in C3F8” concentration-finding algorithm.

Page 25: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

25

Fig. 4. Effect on concentration (in green), speed of sound (in black), of adding 20 litres of N2 to the vapour side of the TS system (July 12, 2016). Temperature and pressure in the DG sonar are shown respectively in purple and red.

The precise measurement of high C3F8 vapour return flows (up to 1.2 m.s-1) to the ATLAS thermosiphon condenser is ensured by the angled ultrasonic flowmeter (Fig. 1 and Fig. 5).

The device was constructed in stainless steel with the acoustic path crossing the gas flow at 45°. Quarter turn ball valves in the acoustic tubes can be closed to allow ultrasonic transducer replacement without interrupting the main gas flow. The instrument has been calibrated against an anemometer in air (Fig. 5) and has demonstrated a flow precision of ± 2.3% full scale for flows up to 10.5 m.s-1[3]. Operation of the device in high returning flows of C3F8 vapour is planned soon.

Fig. 5. Calibration in air of the 45° angled ultrasonic flowmeter prior to installation in the ATLAS thermosiphon recirculator (vs. an Amprobe TMA10A anemometer) [3].

3. Measurements in the Triple Sonar

Instrumentation The triple sonar instrumentation (see Fig. 1)

monitors coolant leaks from the SCT, Pixel and IBL

detector into their separate N2-purged anti-humidity environments.

Fig. 6 illustrates the change in effective C3F8 concentration in the N2-purged anti-humidity environment surrounding the ATLAS Pixel silicon tracker subsystem, following the start-up of the detector cooling on January 28 2016. The sensitivity to molar concentration changes was better than 2.10-5. An increase in C3F8 concentration was seen, from < 5.10-5 to around 1.3.10-3, allowing identification of leaking cooling sub-circuits.

Fig. 6. C3F8 concentration sampled from the N2-purged volume of the ATLAS Pixel detector before and following evaporative cooling system turn-on: January 28, 2016 [4].

The four segments of the much larger N2-purged environmental volume of the ATLAS SCT silicon sub-detector are sequentially monitored in a 16-hour supercycle by aspiration through a single sonar instrument. Sample extraction points (“End-caps “A” &“C”); “Barrel levels” “1” & “7”) are chosen to maximize sensitivity to expected localized hydrostatic pooling of heavy C3F8 vapour.

Fig. 7 illustrates the apparent C3F8 concentration preceding and following the SCT cooling system re-start on January 27th 2016. This restart occurred while zone “Barrel 7” was being measured: the sharp spike here is due to the higher evaporation pressure at which the evaporative cooling was initially operated.

Page 26: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

26

Fig. 7. Increase in the C3F8 concentration sampled at four points from the N2 purged volumes of the SCT detector following turn on of its C3F8 evaporative cooling

on January 27, 2016 [4].

After the cooling system start-up a significant change in concentration was also visible in zone “Barrel 1”, as in the Pixel detector (Fig. 6). The sampling sequence continues to “End Cap C”, then to “End Cap A”, then “Barrel level 7”. No significant

increase was seen for the zones “Endcap C” and “Endcap A”. In the latter zone, however, there was already a high apparent C3F8 concentration due to known dry air ingress from the ATLAS ID external envelope purge system at this time. Since binary gas analysis is a rapid hypothesis-dependent diagnostic, contamination increases in a heavy “search” gas can be mimicked by higher concentrations of a lighter contaminant. Warnings given by a continuously-sensitive sonar instrument can however indicate the need for further investigation with more expensive multi-gas sensitive instrumentation, including gas chromatography.

Fig. 8 illustrates the time history of the molar concentration of C3F8 coolant leaks into the N2

environment of the SCT and Pixel sub-detectors over a 5-day period in May 2016.

For the SCT, the highest C3F8 concentraton is seen at the lowest of the points monitoring the SCT N2

environmental volume (“Barrel level 1”) where more heavy C3F8 vapour leaking from the evaporative cooling circuits has had more time to pool after several months of operation. The increased apparent C3F8 concentration, particularly in the “Endcap A” zone is also in part due to a change of the ATLAS ID external envelope purge gas from nitrogen to heavier CO2.

Fig. 8. Variation of the C3F8 concentration in the N2 environment surrounding the ATLAS Pixel and SCT sub-detectors recorded for a period of 5 days in May 2016.

4. Conclusions

We have developed instruments that have demonstrated high-precision real-time measurements both for flow measurement and binary gas composition analysis. A series of ultrasonic instruments are included in the control of the new thermosiphon cooling system and for leak detection at the ATLAS detectors. Several measurements were performed during the commissioning of the plant. For flow measurement, a precision of < 2.3 % for flows up to 10 m.s-1 was measured. For the binary gas analysis, the precision of mixture determination also depends on the difference between the molecular weights of the two components: a smaller difference results in an increased uncertainty in the mixture determination. Nevertheless the precision is satisfying in all the tested applications. In the case of C3F8 (m.w. = 188) leaks into N2 (m.w. = 28) envelopes of the ATLAS Pixel and SCT silicon tracker subdetectors

a mixture precision is ±0.002 % was demonstrated for C3F8 concentration for the 0 – 1 % range, of most interest in leak detection. References [1]. G. Aad et al., The ATLAS experiment at the CERN

Large Hadron Collider, JINST 3.08: S08003, 2008. [2]. M. Battistin et al., The thermosiphon cooling system of

the ATLAS experiment at the CERN Large Hadron Collider, Int. J. Chem. React. Eng., Vol. 13, Issue 4, 2015, pp. 511–521.

[3]. R. Bates et al., A Custom Online Ultrasonic Gas Mixture Analyzer With Simultaneous Flowmetry, Developed for the Upgraded Evaporative Cooling System of the ATLAS Silicon Tracker, IEEE Trans. Nucl. Sci., 61, 2014, p. 20159.

[4]. M. Alhroob et al, Custom real-time ultrasonic instrumentation for simultaneous mixture and flow analysis of binary gases in the CERN ATLAS

Page 27: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

27

experiment, Nuclear Instrument and Methods in Physics Research A, April 2016, in press.

[5]. R. Bates et al., The cooling capabilities of C2F6/C3F8 saturated fluorocarbon blends for the ATLAS silicon tracker, JINST 10: P03027, 2015.

Page 28: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

28

A Universal Experimental and Computational Framework for Decoding

Complex Gas Mixtures

Unab Javed 1, Kannan Ramaiyan 2, Cortney R. Kreller 2, Eric L. Brosha 2, Rangachary Mukundan 2 and Alexandre V. Morozov 1

1 Department of Physics & Astronomy and Center for Quantitative Biology, Rutgers University, Piscataway, NJ, USA

2 Los Alamos National Laboratory, Los Alamos, NM, USA E-mail: [email protected], [email protected]

Summary: An array of sensors is employed to identify and quantify gases in complex mixtures of unknown composition. The sensors use dense metal electrodes with a porous ceramic electrolyte which conducts O2− ions. These mixed-potential sensors utilize the different rates of reactions on each electrode surface to set up a potential difference. Here, we employ a first-principles approach to develop a computational model of the sensors, which enables robust predictions of gas concentrations from the observed voltage output. Keywords: Mixed-potential sensors, Sensor arrays, Physical models of sensor output, Sensor cross-sensitivity. 1. Introduction

As the requirements for fuel-efficient cars increase, it is essential that vehicles be equipped with more advanced detectors for better combustion and emissions control. The on-board diagnostics (OBD) used in sparkignition vehicles include oxygen λ-sensors to detect oxygen levels in emissions. A similar OBD set-up is sought to measure the gases NOx, NH3

and C3H8 in diesel engine exhaust and treatment systems.

In a mixture with several gases present, ideal detectors would have perfect selectivity and sensitivity for each constituent gas. However, this is difficult to accomplish and quickly becomes impractical as the number of gases increases. In previous work we have shown that combinatorial sensor arrays can be used to detect multiple gases using relatively few sensors, much like the olfactory system [1, 2].

Here, we describe both the sensors used for detecting diesel exhaust gases and a computational model for the sensor array response. The model is based on the fundamental electrochemistry of the sensor, utilizing a quantitative description of gas-sensor interactions.

2. Sensor Design

LANL has developed patented pre-commercial prototype mixed-potential sensors, which incorporate dense electrodes and porous electrolyte. These sensors have been shown to exhibit preferential selectivity and sensitivity to the target gases, although perfect response to each particular gas in the mixture has been elusive [2].

The sensors employed in this work have the composition La(0.8)Sr(0.2)CrO(3)|YSZ|Pt and AuPd|YSZ|Pt, referred to below as the Cr and Au

sensors for brevity. One such sensor is shown in Fig. 1. YSZ, or yttriadoped zirconia, is a ceramic material that conducts at temperatures between 400-700 C. The top panel of Fig. 1 shows the electrodes of the sensor, printed onto the ceramic substrate and coated with the porous YSZ electrolyte. The bottom panel shows the opposite side of the sensor, where the Pt-heater is printed onto the sensor to keep the temperature range between 400-700C. It has been shown that the sensor exhibits the greatest sensitivity around 400 C which then tapers off, to limited response around 700C [3].

Fig. 1. The two sides of a planar Cr sensing element.

Thus sensor response varies greatly with the temperature. We exploit this observation here by running two Cr sensors at different temperatures, and treating the resulting responses as those of “independent” sensors. Fig. 2 demonstrates how the Cr sensor response is modified with temperature.

3. Model of Sensor Response

The gases in the system arrive at the electrode surfaces where they are oxidised or reduced, with O2− ions in the YSZ electrolyte acting as the charge carriers. We assume that the following reaction takes place on one of the electrodes:

Page 29: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

29

(1) where Arepresents all reactants, Ball products, and z1is the number of elections (a similar reaction takes place on the other electrode). The reaction rates on each electrode determine the observed potential difference. As the reaction rates reach limiting values, the system can be treated using the Butler-Volmer equation, which reduces to the Tafel equation in the high overpotential limit. For a single electrode, this is

(2) where ie, β, R, F, Ve, VE0are the current density, the charge transfer coefficient, the gas constant, Faraday’s constant, the voltage difference between the electrode and electrolyte, and the thermodynamically determined voltage, respectively; i0e is the limiting value of ie in the Ve = VE0 case; n is a constant determined by the number of electrons taking part in the rate-determining reaction, and the sign in the exponent is determined by the direction of the reaction.

Fig. 2. Cr sensor response to the C3H8-NO binary mixture at temperatures 450 C (red) and 470 C (blue), as a function of the NO concentration on the log scale. The [C3H8]/[NO] ratio in the mixture is 3.1.

At steady state, the difference in the Ve’s for the two electrodes determines the measured voltage, V. By solving for V using Eq. 2, and assuming that concentrations of O2and Bremain constant, we get V=V0+AlnK, (3) where V 0and A are factors which may in principle depend on several parameters, such as relative concentrations of gases in the mixture, and K is the equilibrium constant corresponding to the reactions that take place on both electrodes. 4. Binary Gas Mixtures

The simplest case to analyze is a binary gas mixture, where one of the NO2, NH3or C3H8gases is

diluted by NO to the ratio α=[X]/[NO], where Xis one of the three gases and square brackets indicate concentrations. For binary mixtures, Eq. 3 can be rewritten as V=V0(α)+Aln[NO], (4) where V0(α)=a+blnα(aand bare constants to be determined from data), and A is also assumed to be constant. A plot of V0vs. lnαis shown in Fig. 3 for the [C3H8]/[NO] binary mixture. Clearly, the predicted log-linear dependence is consistent with the data.

Using Eq. 4, a, band Acan be determined for each mixture and each sensor using either least-squares or Bayesian fit to the observed voltage values. These calibrated models can then be used to predict mixing ratios (α values) for binary mixtures of unknown composition exposed to the same sensor. Table 1 shows our preliminary results for predicting α = [C3H8]/[NO]. The accuracy of our approach is apparent, with all the predictions within 10% of the actual mixing ratios.

Fig. 3. Plot of V0vs. lnαfor α=[C3H8]/[NO]. The sensor

used is the Cr sensor at 450C.

Table 1. Predictions of α=[C3H8]/[NO]using two singlesensor setups. Cr450 refers to the thin Cr sensor operating at 450C, and Au475 refers to the Au sensor

at 475C.

αexact αpredCr450 αpredAu475 1.0 1.09±0.23 1.13±0.20 1.6 1.48±0.25 1.51±0.24 2.1 1.93±0.27 1.94±0.27 3.1 3.02±0.43 3.02±0.40 4.7 4.76±1.19 4.90±0.88

Inference of gas concentrations in more complex gas mixtures, and even simultaneous inference of the total gas concentration and the mixing ratio in binary mixtures require input from multiple sensors. We are currently in the process of extending our approach to the multiple-sensor, multiple-gas scenario. In particular, in the case of complex gas mixtures the

Page 30: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

30

parameters of the model will be calibrated separately for each sensor, and then the combined output from the array of sensors will be used to predict relative concentrations of each constituent gas.

Overall, our goal is to develop a computational and experimental approach to rapid and robust predictions of relative concentrations of potentially harmful gases that appear in diesel exhaust, as well as diesel exhaust treatment systems. Such a system would be invaluable in monitoring and minimizing atmospheric pollution caused by diesel engines.

References [1]. Tsitron, J. et al., Decoding complex chemical mixtures

with a physical model of a sensor array, PLoS Comp. Biol., 7, 2011, p. e1002224.

[2]. Tsitron, J. et al. Bayesian decoding of the ammonia response of a zirconia-based mixed-potential sensor in the presence of hydrocarbon interference, Sens. Act. B: Chem., 192, 2014, pp. 283–293.

[3]. Kreller, C. et al. Mixed-Potential NOx and NH3 sensors fabricated by commercial manufacturing methods, ECS Transactions, 64.1, 2014, 105–113.

Page 31: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

31

Dopant Effects on the Response to Acetone and Ethanol

of SnO2-based Gas Sensor for Breath Monitoring

Y. Sadaoka 1, M. Mori 1, T. Ueda 2, H. Mitsuhashi 2 and M. Nakatani 2 1 The Cooperative Center of Scientific and Industrial Research, Ehime University,

Matsuyama, Ehime, 790-8577, Japan 2 Gas Detection Systems, New Cosmos Electric Co. Ltd. Yodogawa-ku, Osaka, 532-0036, Japan

E-mail: [email protected] Summary: the detection of acetone in the human breath is promising for non-invasive diagnosis and painless monitoring of diabetes. The acetone concentration in the health varies from 0.3 to 0.9 ppm in healthy people to more than 1.8 ppm for diabetics. The concentration of ethanol in the breath is ranging from 1 to 100 ppm. A contamination with a higher level’s ethanol induced a lack of reliable detection of acetone. To determine the concentration of acetone in the breath containing ppm level’s or more ethanol, the use of two gas sensors in which one of them having a high selectivity for acetone and other having comparable level’s sensitivity for acetone and ethanol is considered. In this study, sensing characteristics of SnO2 based gas sensors with additive (V, Mo, Ti, Fe, Zn, La, and Pb) for acetone and ethanol were examined under a high humidity condition above 80 %RH. Keywords: Breath analysis, Acetone, Ethanol, SnO2 based gas sensor, Doping effects.

1. Introduction

Because breath tests are among the least invasive methods for monitoring a person’s disease state or exposure to drug or an environmental pollutant, interest in breath analysis for clinical diagnosis has increase in recent years. The breath contains several hundred volatile organic compounds with concentrations ranging from ppt to ppm. In particular, the detection of acetone in the human breath is promising for non-invasive diagnosis and painless monitoring of diabetes. The acetone concentration in the health varies from 0.3 to 0.9 ppm in healthy people to more than 1.8 ppm for diabetics. The concentration of ethanol in the breath is ranging from 1 to 100 ppm depending an amount of adoption/ drinking. Recently, a high sensitivity and selectivity to acetone for some semiconducting oxide gas sensors are reported and these sensors are also responded to ethanol, while the sensitivity for ethanol is lower than that for acetone. The response behavior of a commercially available gas sensor having a higher selectivity for acetone is shown in Fig. 1.

1

2

3

4

5

1

1.5

2

2.5

0.01 0.1 1 10

ratio (acetone/ethanol)

C/ppm

Eou

t/V

acetone

ethanol

ratio

Fig. 1. Response to acetone and ethanol of a commercially

available semiconducting gas sensor for acetone.

A contamination with a higher level’s ethanol obstructs the determination of acetone’s level. For example, the response for 1 ppm of acetone is comparable to 14.3 ppm of ethanol. To ensure the estimation of the concentration of acetone in the breath containing ppm level’s or more ethanol, the use of two gas sensors in which one of them having a high selectivity for acetone and other having comparable level’s sensitivity for acetone and ethanol is considered.

Polycrystalline semiconducting oxides such as SnO2 are widely used in the solid-state gas sensors. SnO2-based sensors are less conductive when exposed to air. This is because electrons from SnO2 are captures by adsorbed oxygen molecules. The ethanol sensing mechanism of SnO2 can be explained by the release of electrons from oxygen ions when they react with reducing ethanol molecules. H2O and oxygen molecules, main adsorbates in ambient air, both affect SnO2 conductance and sensitivity to ethanol and acetone. Many investigations of SnO2–based sensors proved that both conductivity and sensitivity depend on the relative humidity. The sensor response of oxide semiconducting gas sensors decreases with increasing humidity in general. Many studies have been carried out on improving of sensor responses by additives. The addition to SnO2 was useful for the improvement of sensor response for volatile organic compounds. In this study, sensing characteristics for acetone and ethanol by additives (V, Mo, Ti, Fe, Zn, La, and Pb) were examined under a high humidity condition above 80 %RH at RT.

2. Experimental

SnO2 powders dispersed in a mixture of ethylene glycol and water are dropped on a substrate with comb-type electrodes and heated at 600 oC, following a

Page 32: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

32

prescribed amount of acidic solution with dopant-ion is dropped on the surface, finally the elements were heated at 600 oC again for 8 hrs in ambient air. Sensor response for acetone, ethanol, and mixture of acetone and ethanol with concentration ranging below 20 ppm at 83 %RH/air was examined. The concentration of acetone and ethanol was controlled with a permeater. The total flow rate remained a constant. 3. Results

For all examined sensors, the conductance is increased with an increase in the concentration of acetone and ethanol. Concentration dependence of the conductance at 500 oC for Mo-doped SnO2 sensor is summarized in Fig. 2. The conductance changes are well expressed with a relationship

G = G0 + m1 [C]m2, where [C] is the total concentration of acetone and ethanol in ppm. The conductance is a function of the total concentration of acetone and ethanol and could be expressed as the relationship with G0=1.01E-6, m1=8.46E-6, m2=0.405, and R2=0.98098.

10-6

10-5

10-4

0.01 0.1 1 10

G-Mo-EA1G-Mo-EA2G-Mo-AEG-Mo-AcetoneG-Mo-EtOHMo-all

Con

duc

tanc

e/S

EA1: Ethanol: 3.37 ppm, Acetone: 0-12.1 ppmEA2: Ethanol: 0.42 ppm, Acetone: 0-7.05 ppmAE1: Acetone: 3.75 ppm, Ethanol: 0-3.37 ppm

air-base

(CAcetone

+ CEthanol

)/ppm

Fig. 2. Conductance changes with the concentration

of acetone, ethanol, and a mixture of acetone and ethanol at 500 oC and 83 %RH at RT for Mo-doped SnO2 sensor.

The sensing characteristics, Go, m1 and m2 are

strongly depended on the doped species. The characteristics for ethanol at 560 oC and 83 %RH at RT are summarized in Fig. 3.

The estimated characteristics and deriations are summarized in Table. 1. It seems that the doped SnO2 sensors are preferable to detect the sum of acetone and ethanol ranging below 30 ppm. 4. Conclusion

To monitor the changes of the concentration of both acetone and ethanol in breath, the use of two gas

sensors, in which one of them having a high sensitivity and selectivity for acetone and other having a comparable level’s sensitivity for ethanol and acetone is proposed. The SnO2–based gas sensors with additive are sensitive to both acetone and ethanol. The response is experimentally expressed by the relationship, G = Go + m1[C]m2 for acetone and ethanol. The ratio, G(1ppm)/G(air), is depended on doped species, i.e., Mo > La > Zn > Ti > Pb > Fe > V at 530 oC and 83 %RH.

0

2

4

6

8

10

0

0.2

0.4

0.6

0.8

1

V Mo Ti Fe Zn La PbG

o &

m1

m2

Dopant

G = Go + m1*[C]m2

Go

m1

m2

x105

Fig. 3. Sensing characteristics for ethanol of M-doped SnO2

sensors at 560 oC at 83 %RH.

Table 1. The characteristics for total concentration of acetone and ethanolat 560 oC and 83 %RH at RT.

additive G0 m1 m2 R2

V 6.17E-06 4.94E-06 0.4726 0.9819

Mo 5.25E-06 2.14E-05 0.4941 0.9732

Ti 5.21E-06 1.16E-05 0.4928 0.9703

Fe 7.50E-06 1.03E-05 0.5762 0.9923

Zn 5.29E-06 1.27E-05 0.7053 0.9879

La 1.17E-05 3.54E-05 0.4527 0.9456

Pb 3.10E-05 5.09E-05 0.4120 0.9478

1

2

3

4

5

6

V Mo Ti Fe Zn La Pb

G(1

ppm

)/G

(air

)

Dopant

Fig. 4. Sensitivity for 1 ppm acetone and ethanol at 560 oC and 83 %RH at RT.

Page 33: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

33

Small Sensor System (S3) Equipped with Nanowires Sensors

as an Interdisciplinary e-Sensing Device for the Microbiological and Quality Control

V. Sberveglieri 1, M. Soprani 2 and E. Núñez Carmona 1,3

1 CNR-INO, Sensor Lab, Via Branze 45, 25124 Brescia, Italy 2 University of Modena and Reggio Emilia, Department of Life Science,

Via J. F. Kennedy 17, 42124, Reggio Emilia, Italy 3 University of Brescia, Department of Information Engineering, Via Branze, 38 - 25123 Brescia, Italy

E-mail: [email protected] Summary: Nowadays a fast and economic device for the early detection of microbial contamination and quality assurance is need to reduce the number of food-borne related hospitalizations by year. These paper present applications crossing the food quality to arrive at the human cross-contamination. The S3 Device has been constructed at the Sensor Lab as well as the classical MOX sensors and the new technologies with nanowire. It is linked to the necessity of a portable, low power consumption and user-friendly device. The aim of this study was to investigate and show the broad spectrum of potential applications of the S3 device in food quality control and microbial cross-contamination diagnosis. The S3 technique was combined with classical microbiological and chemical techniques, like GC-MS with SPME, for the control of river wastewater contamination. In the analysed case, the novel S3 device was proficient to identify the target by means of the alterations in the pattern of volatile organic compounds (VOCs), recreated by principal component analysis (PCA) of the sensor-achieved responses. The reached results powerfully support the use of S3 device in industrial laboratories. Keywords: Small sensor system (S3), Nanowire sensors, MOX, Electronic nose, GC-MS-SPME, Microbial quality control.

1. Introduction

The increasing concerns strongly linked to the pollution on water and food safety recall the need of supervising all aspects in real time and in turn, led to a remarkable effort in terms of investigation and funding for the improvement of sensors device dedicated to numerous applications. Applications that have in common the necessity of a device that give results in a few minutes, portable, low power consumption and user friendly.

To promote this technology to industrial application, metal oxide gas/odour sensors became representative candidates in areas like food-chain, environment control, automotive, indoor and outdoor air quality check and checking, humans applications.

Scientific groups worldwide are investigating them giving due significance to the various aspects of gas/odour sensing properties [1, 2].

In this contextual, technologies for food safety control play a crucial role in avoid food contamination and as such, food and beverage producers have been continually searching for reliable rapid analytical methods for the detection or control the microbial presence at the earliest stage or on-line [3].

In these years, the incessant development of nanotechnology run to the production of quasi-one dimensional (Q1D) structures in a multiplicity of morphologies such as nanowires, core–shell nanowires, nanotubes, nanobelts, hierarchical structures, nanorods, nanorings. In specific, metal

oxides (MOX) are an increasing interest for both fundamental and applied science [4].

In addition, nanowires may display physical properties, which are considerably different from their coarse-grained poly-crystalline complement subsequently of their nano-sized sizes. Surface effects dominate due to the increase of their specific surface, which leads to the improvement of the surface related properties, such as catalytic activity or surface adsorption: key properties for superior chemical sensors construction.

High degree of crystallinity and atomic sharp terminations make nanowires very promising for the improvement of a new generation of gas sensors decreasing instabilities, typical in polycrystalline systems, related with grain coalescence and drift in electrical properties.

The research studies published over the last years regarding nanowire gas sensors, and principally on chemical/gas sensors, indicate that significance in the synthesis of nano-structures and investigations into their sensing properties for manufacturing of the gas sensing devices appreciably increased.

The aim of this work was to investigate and to show the wide-ranging of potential uses of the S3 device in the detection of river wastewater contamination. The S3 technique was coupled with classical microbiological and chemical techniques, like gas chromatography with mass spectroscopy (GC-MS) with SPME technique or classical microbiological analysis.

Page 34: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

34

2. Materials and Methods The analyses were performed with the novel Small

Sensor System (S3) (Sensor Lab, Brescia, Italy http://sensor.unibs.it).

The sensor's array is located in a thermally controlled sensor chamber where are placed: 3 RGTO sensors and 3 Nanowires sensors (Table 1).

Table 1. Sensor array configuration.

Sensor Operating

temperature (°C)

Composition Description

1 245 SnO2-MoO2 RGTO blend of

Tin Molybdenum oxides

2 280 ZnO Nanowire sensor

made of Zinc oxide

3 375 SnO2 Nanowire sensor

made of Tin oxide

4 400 SnO2 // Ag RGTO Tin oxide

catalysed with silver

5 500 ZnO Nanowire sensor

made of Zin oxide

6 500 SnO2-WO3 RGTO blend of

Tin and Tungsten oxides

The sample headspace (HS) for the S3 analysis was created in vials (20 ml). For this study, the S3 was equipped with a HT280T auto-sampler, which supply a 40 loading position carousel, in order to increase the number of replicates. The vial conditioning was performed using the shaking oven at 40 °C for 10 minutes. It allows reaching the VOCs equilibrium concentration in the vial’s headspace.

The vial’s headspace (2 mL) was adsorbed in static headspace and injected into the carrier flow at the 4 ml/min of speed. Appropriately adapted gas chromatography injector (kept at 40°C to prevent any condensation) using synthetic chromatographic air with a continuous flow rate of 10 ml/min was applied to perform the sensor baseline and the recovery time was 28 min. The data analysis was run by means of Principal Component Analysis (PCA) (Fig. 1). 2.2. Detection of River Wastewater

Contamination Three different kinds of samples have been used:

drinking water directly collected from the bottle, drinking water directly collected from the tap and contaminated water.

The pathogenic microorganisms (Table 2) were isolated directly from rivers’ wastewater with the membrane filtration technique. River's wastewater was filtered with a 45 µm cut-off filter, in vacuum and sterile conditions. The real pathogenic microorganism’s concentration was founded to be 500 CFU/ml in the river's wastewater .

Fig. 1. Working process flow chart.

Table 2. Pathogenic Microorganisms Culture Mediums and Temperature/Time ratio.

Pathogenic

Microorganisms Culture Medium

Temperature /Time Ratio

Escherichia coli MacConkey

Agar 35 °C / 24 h

Salmonella typhimurium

Rappaport 41 °C / 24 h

Vibrio cholerae TCBS 36 °C / 18-48 h Pseudomonas aeruginosa

Pseudomonas Agar

36 °C / 48 h

The individual solutions of pathogenic microorganisms were mixed together to obtain a single solution, 500 CFU/mL.

Finally, 5 mL of the solution containing the pathogenic microorganisms were placed in the vials. 3. Results

The Score Plot graph in Fig. 2 shows that S3 generates two different clusters when a comparison is make for the two different samples, contaminated an uncontaminated ones so it is able to recognize the contaminated water and the drinking water.

The PCA score PLOT was acquired using the feature extraction algorithm (MatLab), FFT, Fast Fourier transform of signal. If P1=0 then the FFT is normalized at the max point of BEFORE step otherwise it isn’t normalized.

3. Conclusions

Classical chemical and microbiological techniques requires high tech lab skills, high trained lab staff and a consistent amount of money to carry out the analysis.

On the other hands, S3 device will improve the capacity to detect and recognize the set of VOCs. It can be used in cooperation to classical techniques.

Page 35: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

35

Figure 2. PCA “Score Plot” about Drinking water (blue and green dots) and 500 CFU/mL pathogenic microorganisms’ solution (black dots).

This study depose that the S3 device, once created the specific database, is able to detect in the first stages the microbial contamination and follow up they development in food matrix (also compared with the commercial tools equipped only with traditional MOX sensors). References [1]. Loutfi, A., Coradeschi, S., Manib, G. K., Shankarb, P.,

Balaguru Rayappan, J. B. Electronic noses for food quality: A review, Journal of Food Engineering, 144, 2015, pp. 103–111.

[2]. Compagnone, D., Faieta, M., Pizzoni D., Di Natale, C., Paolesse, R., Van Caelenbergd, T., Beheydt, B., Pittia, P. Quartz crystal microbalance gas sensor arrays for the quality control of chocolate, Sensors Actuat B-Chem, B, 207, 2015, pp. 1114–1120.

[3]. Gobbi, E., Falasconi, M., Zambotti, G., Sberveglieri, V., Pulvirenti, A., Sberveglieri, G. Rapid diagnosis of Enterobacteriaceae in vegetable soups by a metal oxide sensor based electronic nose, Sensors Actuat. B-Chem., 207, 2015, pp. 1104–1113.

[4]. Comini, E., Baratto, C., Faglia G., Ferroni, M., Vomiero A., Sberveglieri, G. Quasi-one dimensional metal oxide semiconductors: Preparation, characterization and application as chemical sensors. Prog. Mater. Sci., 54, 1, 2009, pp. 1–67.

Page 36: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

36

Design of Optical-fibre Refractometric Sensors for Liquid Helium

Santa Junnuen Miron-Carrasco 1, Dora Mariela Martínez-Gonzalez 1, Sergei Khotiaintsev 2

1Posgrado en Ingeniería, 2Facultad de Ingeniería, Universidad Nacional Autónoma de México, Avenida Universidad 3000, Ciudad de México, México, c.p. 04510

E-mail: [email protected], [email protected], [email protected] Abstract: We present an optical-fibre refractometric transducer designed for the detection of liquid helium in storage tanks and elements of cryogenic installations. The transducer parameters were optimized for the discrimination between the liquid helium and the gaseous helium or air above it. The predicted signal is of 2.2 dB, which, although relatively small, allows for the detection of liquid–helium and gas interface, and for the measurement of liquid helium level. Keywords: Refractive index sensing, Optical fibre sensors, Cryogenic fluids, Liquid hydrogen, Liquid helium. 1. Introduction

The use of cryogenic fluids such as liquid nitrogen (LN), liquid oxygen (LOX), liquid hydrogen (LH2) and liquid helium (LHe) progressively increases in scientific research, industry and medicine. Therefore, there is an interest in developing new and better sensors for cryogenic fluids. The refractometric optical fibre sensors, in particular, offer several competitive advantages over traditional electrical sensors, thanks to low thermal conductivity, immunity to electromagnetic interference and small size of the optical fibres [1] - [6].

However, the refractive index, n, of the cryogenic fluids is quite small. For example, nLN=1.21, nLH2=1.11 and nLHe=1.026. The small refractive index presents difficulty in discriminating between the cryogenic fluid and gas above it [7]. 2. Refractometric Transducer Design for the

LHe

We considered a hemispherical transducer described in our previous works [8], [9], with the difference that we designed and optimized this transducer for the detection of the LHe with the lowest refractive index among the liquid substances.

The generic configuration of this transducer is shown in Fig. 1.

The transducer consists of a small transparent optical detection element (1) and a pair of multimode optical fibres (2) that link the element with the optical transmitter and receiver. When the sensing element is in the air, the two optical fibres are optically coupled due to internal reflection of light at the spherical surface of the element.

In a liquid, the difference between the refractive index of the sensor material and that of the external medium is smaller. It reduces the total internal reflection (or the reflexion vanishes completely). This allows one to perceive the changes in the kind of external liquid medium from air to liquid and vice versa.

Fig. 1. Schematic of the refractometric transducer.

For reliable discrimination between a cryogenic fluid such as LHe and a gaseous medium, an optical transmission function of the transducer, T(n), has to have the largest possible negative gradient in the range from n=1 to nLHe.

In order to achieve this goal, we carried on an iterative optimization, which consisted of cycles of numerical analyses of the transducer by means of ray-tracing of respective mathematical model. More specifically, we analysed the transducer response to the external refractive index. The mathematical model accounted for the transducer’s geometry; in particular the form of transducer’s working surface and the position of the two optical fibers with respect to the transducer axis z, as well as optical parameters and material constants of all array elements. The analysis by ray-tracing was followed by a variation of one of transducer’s parameters. These cycles were repeated successively. The criterion for the optimality was the maximum signal =I(air)/I(n), I(air) and I(n) is the light intensity at the sensor output when the sensor is in the air and liquid, respectively.

The results are presented in Fig. 2. The analysis showed a very strong effect of

dimensionless diameter of the optical fibre core, d, on the signal in the vicinity of n=1. Also, the position of the two optical fibres with regard to transducer axis z had a strong effect on the signal . While the numerical aperture NA in a range 0.2-0.5 had a

Page 37: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

37

moderate effect on the sensitivity . For the transducer and optical fibre core of fused silica glass (SiO2, n=1.45 at 20 K), the optimal parameters found in this work were: = 0.707, d=0.05, NA=0.45. These parameters yielded the signal LHe = 2.2 dB.

Fig. 2. Optical transmission T vs. the external refractive index n for distinct combinations of parameters.

3. Conclusions

The transducer described in this paper was designed and optimized for discrimination between the LHe and the gaseous medium such as the air or gaseous helium. The optimization yielded the predicted signal LHe = 2.2 dB. Such a signal is relatively small, nevertheless, it allows one to discriminate between the LHe and the gaseous medium, and detect the interface between the LHe and the gas above it, as well as measure the level of LHe in storage tanks and elements of cryogenic installations.

Acknowledgments

The authors acknowledge the support of this work by the General Directorate for Academic Staff Affairs

(DGAPA) of the National Autonomous University of Mexico (UNAM) [grants PAPIIT IT102515 and PAPIME PE10161], and the Institute of Engineering and the Faculty of Engineering of the UNAM [joined grant Sistemas de Detección Temprana de Riesgos en Edificios Históricos]. References [1]. J. Villatoro, J. Zubia, New perspectives in photonic

crystal fibre sensors, Optics and Laser Technology, Vol. 78, 2016, pp. 67-75.

[2]. A. N. C. Martinez, M. Komanec, T Nemecek, S. Zvanovec, & S. Khotiaintsev, Fiber optic refractometric sensors using a semi-ellipsoidal sensing element, Applied Optics, Vol. 55, No. 10, 2016, pp. 2574-2579.

[3]. M. Komanec, T. Martan, T Nemecek, & S. Zvanovec, Multimode fiber tapers for reproducible refractometric liquid detection, Optical Eng., Vol. 54, No. 4, 2015, p. 047102-1-047102-6.

[4]. P. Zubiate, C. R. Zamarreño, I. Del Villar, I. R. Matias, and F. J. Arregui, High sensitive refractometers based on lossy mode resonances (LMRs) supported by ITO coated D-shaped optical fibers, Optics Express, Vol. 23, No. 6, 2015, pp. 8045-8050.

[5]. I. Del Villar, A. B. Socorro, J. M. Corres, F. J. Arregui, and I. R Matias, Refractometric sensors based on multimode interference in a thin-film coated singlemode-multimode-single-mode structure with reflection configuration, Applied Opt., Vol. 18, 2014, pp. 3913-3919.

[6]. G. Salceda-Delgado, D. Monzon-Hernandez, A. Martinez-Rios, G. A. Cardenas-Sevilla, and J. Villatoro, Optical microfiber mode interferometer for temperature independent refractometric sensing, Optics Lett., Vol. 37, No. 11, 2012, pp. 1974-1976.

[7]. C. Yang, S. Chen, and A. Kazemi, Fiber-optical liquid level sensor under cryogenic environment, Proc. SPIE, Vol. 4204, 2001, pp. 206-215.

[8]. K. E. Romo Medrano, S. Khotiaintsev, An optical-fiber discrete liquid-level sensor for liquid nitrogen, Measurement Sci. Technol., Vol. 17, 2006, pp. 998-1004.

[9]. S. Khotiaintsev, V. Svyryd, H. Mejia del Puerto, Fiber-optic liquid-interface sensor for liquid hydrogen, Sensors and Materials, Vol. 21, No. 1, pp. 13-23.

Page 38: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

38

Integration of VLC and a-SiC:H Technology for Indoors Navigation

P. Louro 1,2, M. A. Vieira 1,2, M. Vieira 1,2,3

1Electronics Telecommunication and Computer Dept. ISEL, R. Conselheiro Emídio Navarro, 1949-014 Lisboa, Portugal

2 CTS-UNINOVA, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal. 3 DEE-FCT-UNL, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal

1 Tel.: +351919310252, fax: +351218317144 1 E-mail: [email protected]

Summary: In this paper a photodetector working as an active optical filter device is used to detect modulated visible optical signals for applications based on Visible Light Communication (VLC). The proposed application demonstrates the viability of indoor positioning using VLC technology established by the modulation of indoor ultra-bright RGB white LEDs. The signals of the internal red and blue chips of the white LEDs were modulated at specific frequencies and the generated photocurrent was measured by a pin-pin photodetector based on a-SiC:H/a-Si:H. This device operates as a visible optical filter with controlled wavelength sensitivity through the use of adequate optical biasing light. Thus it is able to detect different wavelengths which allow the detection of the individual components of the tri-chromatic white LED. This possibility is the basis for the indoor location algorithm. We demonstrate the possibility of decoding four transmission optical channels supplied by two different wavelengths of white LEDs modulated under different bit sequences. The identification of the signals received by the photodetector allows the location identification of the photodetector position and supplies indoor navigation. Keywords: amorphous SiC technology, Visible light communication, indoors positioning, Fourier transform.

1. Introduction

VLC technology makes use of the visible part of

the light spectrum to modulate specific wavelengths and encode and transmit information [1]. The most common optical sources are the widely used white LED lamps that can also be easily modulated, fulfilling the VLC requirements. An interesting application of VLC technology is for indoor positioning and navigation resources. Its use can be extended from in-house navigation to guide users inside large buildings [2] to location detection of products inside large warehouses.

We propose a communication system operating in the visible range using 4 ultra-bright white RGB and a photodetector device based on 2 stacked multilayered a-SiC:H/a-Si:H structures that act as optical filters [3]. LEDs enable four transmission optical channels supplied by the modulation of different frequencies of the internal red and blue chips. The chance of tuning the spectral device sensitivity is analyzed and discussed using several optical bias conditions that induce different modulations of the electrical field along both front and back structures, amplifying or cutting specific wavelengths. This enables the identification of the transmitted individual input channels and allows the photodetector location identification. The assignment of the identified signal to the location is the basis of the proposed position algorithm. A decoding strategy based on the evaluation of the output photocurrent complex Fourier coefficient for the detection of optical signals is presented and discussed.

2. Experimental Setup

The experimental setup used for the detection of the signals emitted by the white LEDs included four white LED lamps framed at the corners of a square as if assembled on the ceiling (Fig. 1). At a fixed distance the photodetector device was centered inside this square.

Fig. 1. Configuration of the experimental setup showing the position of the white LEDs and the wavelength

and frequency of the modulation chip.

3. Photodetector Configuration

Fig. 2 shows the simplified cross-section structure of the device used to detect the transmitted information. It is a multilayer heterostructure composed by two pin structures built on a glass substrate and sandwiched between two transparent electrical contacts. The front pin a-SiC:H photodiode is responsible for the device sensitivity in the short wavelengths of the visible range

Page 39: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

39

due to its minor thickness (200 nm) and higher bandgap (2.1 eV). The back pin a-Si:H structure works in the complimentary past of the visible range, collecting the long wavelengths.

Fig. 2. Simplified cross-section view of the photodetector.

4. Optoelectronic Characterisation

The characterization of the optical sources was done through the measurement of the output spectra of each biased chip junction of the RGB white LED with the driving current. In Fig. 3 it is plotted the normalized output spectra of the red and blue chips of the RGB white LEDs.

400 450 500 550 600 650 7000,0

0,2

0,4

0,6

0,8

1,0

= 13 nm

0 = 626 nm

0 = 470 nm

= 22 nm

Rel

ativ

e in

tens

ity

Wavelength (nm)

Fig. 3. Output spectra of the red and blue chips.

The output spectra cover the wavelengths assigned to the blue, green and red regions, with wavelengths centered, respectively at 470 nm and 626 nm. The full width half height (FWHH) is 22 nm for the blue chip and 13 nm for the red chip, which is in agreement with the usual design of these chips adjusted for the white color perception.

5. Results

In order to analyse the photocurrent signal when the red and blue chips of the tri-chromatic white LED are transmitting a different signal, the internal LEDs were pulsed using different time dependent biasing currents. The location identification is based on the analysis of the device photocurrent, which results from the optical

excitation induced by the optical signals. Thus it is important for the system to be able to detect the combination of two, three of four optical signals. The output photocurrent signals measured under different optical bias conditions are displayed in Fig. 4.

Results show that the signal measured under back illumination is similar to the signal measured without background illumination, which is due to the presence of both red and blue wavelengths that exhibit opposite behaviours under back illumination. The red light quenches the signal and the blue one amplifies it. On the other hand, the photocurrent under front illumination results in an amplified signal due to the high amplification factor of the red light.

0 1 2 3 40,0

0,4

0,8

1,2

1,6

2,0

2,4

2,8

470 nm

626 nm

Back

Front

Dark

#5

Pho

tocu

rren

t (A

)

Time (ms)

Fig. 4. Photocurrent measured without and under front and back optical bias when the device is at the central position.

6. Decoding Strategy

As proof of concept we have developed a computer program to detect the position of the sensor relative to the LEDs based on the measured photocurrents. The program uses two simple steps. In the first step, taking advantage of the colour filtering properties of the photodiode, front biasing is used to detect the red wavelength and back biasing to detect the blue wavelength. Because signals of two different frequencies may share the same light wavelength an additional step is necessary to detect which frequencies are present. The modulus of the complex Fourier coefficient of the photocurrent is calculated for each of the relevant frequencies and compared with a predefined threshold value. 7. Conclusions

A novel wireless transmission system based on VLC technology was designed and its viability analyzed using tri-chromatic white LEDs. The performance of the transmission system was assessed by using tri-chromatic white LEDs pulsed with different frequencies, each assigned to a red or blue wavelength. Results show that the output photocurrent signal can be related to the input optical signals

Page 40: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

40

informing about detector position and enabling navigation tools.

References

[1]. K. Panta and J. Armstrong, Indoor localisation using white LEDs, Electron. Lett., 48, 4, 2012, pp. 228–230.

[2]. T. Tanaka and S. Haruyama, New position detection method using image sensor and visible light leds, in Proceedings of the Second International Conference on Machine Vision (ICMV '09), 2009, pp. 150–153.

[3]. P. Louro, M. Vieira, M. A. Vieira, S. Amaral, J. Costa, M. Fernandes, Sensors & Actuators: A Physical, 172, 1, 2011, pp. 35-39.

Page 41: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

41

Comparison of Pre-processing on Different Kind of Images Produced

by Optical Sensor System

Muhammad Waseem Tahir, N. A. Zaidi, R. Blank, P. P. Vinayaka, M. J. Vellekoop and W. Lang

University of Bremen, Institute of Micro Sensors, Actuators and Systems (IMSAS) Otto Hahn Allee, NW 1 Building, 28359 Bremen Germany

Tel.: + 49 42121862637, fax: + 49 4212184774 E-mail: [email protected]

Summary: Optical sensor system has used for the detection of mold. An air sample was collected on glass slide and then placed it under optical sensor system. Previously, this system only took dark-field images of spores. In latest version, it took three kind of images: dark-field, bright-field and autofluorescence. The image processing algorithms were used to enhanced the image quality of obtained images, which were further analyzed using computer vision algorithm for the detection of spores in images. In this paper, we presented the comparison of these images and highlight the advantage of one on other kind of images. We have observed that bright-field images are much better than other two kinds with respect to computational cost Keywords: Bright-field microscopy, Dark-field microscopy, Autofluorescence microscopy, Mold detection.

1. Introduction

Mold plays an important role of decomposition in

ecosystem. And it has more than 100,000 types. So, mold is present everywhere in our surroundings like house, office, vehicles etc. It is not harmful until it is in less concentrations. Problem only arises when it starts growing and spread in higher concentrations. Then it is threat for human life, food, documents etc.

Mold grows with the help of its airborn particles (spores). These are present in the air where mold is growing. We captured the images of these spores with the help of optical sensor system. These pictures were very challenging, as they were full of dust particles. Moreover, the other problems were illumination, cluttering and occlusion of spores.

To minimize these problems and to enhance the quality of images preprocessing is normally used. In this article, we presented a comparison of different kind preprocessing techniques for different types of images.

2. Setup It was very challenging task to develop a sensor system for very small size mold spores. Mold species like Aspergillus, Cladosproium and Eurotium have spores size between 1.5 µm to 7 µm. We have developed our first prototype as presented in [1]. It consisted of an 5 Mega pixel industrial camera from The Imaging-Source. And then a microscopic lens of 40X was fitted infront of it. We achieved the maximum resolution of 846 nm with this system. This resolution was dependent on pixel size and optics resolution. A dark-field microscopic technique was used to obtain the dark-field images and a ring of LEDs was used as light source. Ring was attached on the optics. First prototype

gave us only one kind of images but in our latest version as presented in [2], we used LED from bottom to obtain bright-field microscopic images and autofluorescence light source was used to obtain autoflourescence images.

3. Experiment and Results An air sampling unit from Umweltanalytik

Holbach GmbH was used to obtain the air sample of archives on glass slides. Then these glass slides were put inside our optical sensor system. The dark-field microscopic image, bright-field microscopic image and autofluorescence image were obtained as shown in Figs. 1, 2 and 3 respectively.

Two stage pre-processing was proposed for all images. In first stage of pre-processing, smoothing was done to remove salt and pepper noise on all images using averaging filter. Drawback of smoothing was that it made images blur. Then sharpening was done with the help of median filter of size 3 x 3 to regain the sharpeness of images.

In second stage parallely different algorithms (Contrast streching, histogram equalization and adaptive histogram equalization) were tried and their results are shown in Figs. 4, 5 and 6. It was observed that in dark-field microscopic images preprocessing algorithms make a significant improvement as shown in Fig. 4. And best pre processing technique for this kind of images was adaptive histogram equalization. In bright-field microscopic images there was no significant output from the preprocessing techniques as shown in Fig. 5. But in Autofluorescence images adaptive histogram equalization and contrast stretching both gave good results. So, to reduce the computational cost of system we can use the bright-field microscopic images for detection.

Page 42: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

42

Fig. 1. Dark-field microscopic image.

Fig. 2. Bright-field microscopic image.

Fig. 3. Autofluorescence image.

4. Conclusions In this paper, we compared result of different pre-processing techniques on different kind of images obtained from optical sensor system. It was observed

that bright-field microscopic images were better than dark-field and autofluorescence images with respect to computautional cost.

Fig. 4. Pre-processing results of dark-field microscopic image.

Fig. 5. Pre-processing results of Bright-field microscopic image.

Fig. 6. Pre-processing results of autofluorescence image.

References [1]. R. Blank, P. Vinayaka, M. Tahir, M. Vellecoop and W.

Lang, Optical Sensor System for the Detection of

Page 43: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

43

Mold, in Proceedings of the IEEE Sensors, 2015, pp. 677 - 682.

[2]. Roland Blank, Poornachandra P. Vinayaka, Muhammad W. Tahir, Joanne Yong, Michael

Vellekoop and Walter Lang, Comparison of several optical methods for an automated fungal spore sensor system concept, IEEE Sensors Journal, 16, 14, 2016, pp. 5596 – 5602.

Page 44: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

44

An Image Depth Extractable CMOS Image Sensor

with a Variable Bit-resolution Infrared Pixel

Youngchul Son and Minkyu Song Dept. of Semiconductor Science, Dongguk University, Seoul, Korea

E-mail: [email protected] Abstract: A variable bit-resolution CMOS image sensor(CIS) for an image depth extraction is discussed. In order to implement an image depth extraction, CIS pixel matrixes with RGB and Infrared (IR) color filters are used. Further, the single-slope ADC inside of CIS has a variable bit-resolution performance to compensate the light density between RGB and IR. For example, RGB signal has an 8-bit resolution, while IR signal has an 12-bit resolution. The proposed CIS has 4 different bit resolutions for IR pixel, such as 12-bit, 10-bit, 8-bit and 6-bit. The proposed ADC has a maximum resolution of 12-bit with the architecture of two-step single-slope (TS SS) type. The proposed image depth extractrable CIS has a 100 MHz clock, and it has been designed with 0.18 μm CIS technology. Keywords: CMOS image sensor, Image depth extraction, Variable bit-resolution IR pixel, Two-step single-slope ADC.

1. Introduction

Recently, Virtual Reality(VR) system is now

widely used in our life with new technology developments. It is expected that VR system is applied to game, entertainment, medical system, and so on [1]. Especially, the demands on 3-dimensional (3-D) recording equipment are rapidly increasing, because entertainment and medical systems with the VR are based on 3-D scene. In order to implement 3-D scene, VR does not need only 2-D imaging, but also 3-D depth map for vertical dimensions at the same time [2]. Thus a CMOS image sensor (CIS) is needed to satisfy the image depth requirements. Fig. 1 shows a scheme of CIS pixel matrix with RGB and IR color filter to implement the image depth extraction [3]. A conventional green filter of the A group is replaced to an IR filter for 3-D scene. As a result of imaging program, when brightness of a RGB imaging is 1/16 lower than brightness of an IR scene, contrast is almost same between RGB imaging and IR scene. That is, a requirement of ADC resolution for IR scene has a factor of 4-bit higher than ADC resolution for RGB imaging. If this method is applied to column-parallel CIS with a conventional ADC, there would be a problem at the fixed bit resolution [4-6]. When a column ADC converts the analog signal of RGB and IR pixels into digital bits, it is impossible for a conventional ADC to process a different resolution on each pixel to make a balanced scene. Therefore, as a first step of the image depth extractable CIS, a variable bit-resolution ADC which is able to change the bit resolution is proposed. For example, the proposed ADC converts a signal of RGB filter with an 8-bit resolution, while the proposed ADC converts a signal of IR filter with a 12-bit resolution. The proposed ADC has 4 different resolutions for RGB signals which are 8-bit mode, 6-bit mode, 4-bit mode and 2-bit mode. A two-step single-slope ADC (TS-SS ADC) for the column-parallel CIS has an advantage of simple

architecture composed of a counter, a comparator and a ramp generator. When the resolution of TS-SS ADC is increased, the conversion time of this ADC is also increased exponentially. To avoid this problem, the proposed ADC is based on the TS-SS architecture composed of a coarse ADC and a fine ADC [7]. The circuit design and implementation are discussed in Section II. Measurement results and conclusions are summarized in Section III and IV, respectively.

R G R G

B IR B IR

R G R G

B IR B IR

Row control

Evencolumn

Evencolumn

Oddcolumn

Oddcolumn

Evencolumn

Evencolumn

Oddcolumn

Oddcolumn

A group

Fig. 1. CIS Pixels with RGB+IR Filters for Image Depth Extraction.

2. CIS Scheme and Circuit Design

Fig. 2 shows the CIS block diagram that was designed in this paper. The CIS structure is composed of pixels, column ADCs, and digital control blocks. The pixel converts the amount of light into the corresponding voltage, which then becomes the input for the ADC. The pixel output voltage is transformed into a digital code in the block of ADC. The digital control block controls the pixel, ADC, and output interfaces, respectively. Further, a variable bit-resolution IR pixel for the image depth extractable CIS is also included.

Page 45: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

45

Fig. 2. Block Diagram of the Proposed CIS.

3. Experimental Results

Fig. 3 shows the chip layout and microphotograph of the designed CIS, and the core area of image sensor is about 4.14 mm2 (2.3 mm × 1.8 mm) with a total chip area of 6.25 mm2 (2.5 mm×2.5 mm). The chip in this study has been fabricated with a Towerjazz 0.18 µm CIS process, which uses an active pixel sensor (APS) with a 4-Tr structure. The odd column array is placed on the left side of the pixel array, and the even column array is placed on the right side of the pixel array. Further, to reduce fixed pattern noise, all the columns are designed to have the same repetitive pattern. The CIS has the pixel array of QCIF resolution which has 220×176 pixels.

(a) (b)

Fig. 3. (a) Chip Layout (b) Chip Microphotograph for CIS.

Fig. 4 shows the measured QCIF sample images with the condition of the 4 different bit resolutions. It achieves the frame rates of 30 frame/s at a main clock speed of 100 MHz. Fig. 4(a) shows the image of 8-bit resolution mode for RGB pixels, while the bit resolution of IR pixel is 12-bit. Fig.4 (d) shows the image of 2-bit resolution mode for RGB pixles, while that of IR is 6-bit. Tab. 1 shows the performance comparison among the bit resolution modes.

4. Conclusions

An image depth extractable CIS was discussed. To implement the technique, for example, RGB signals had 8-bit resolution, while IR signal had 12-bit resolution. Based on this technique, a high quality image depth map was obtained. Further, a two-step single-slope ADC was also discussed.

Acknowledgements

This work was supported by the MOTIE (Ministry of Trade, Industry, and Energy) Korea, under the NGPEP (New Growth Power Equipment Project) program supervised by the KEIT (Korea Evaluation Institute of Industrial Technology), and Exicon.

(a) (b)

(c) (d)

Fig. 4. Measured Image Depth Extraction with the variation of bit: (a) 8-bit, (b) 6-bit, (c) 4-bit, (d) 2-bit.

Page 46: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

46

Tab. 1. Performance Summary of the Prototype CIS.

References [1]. A. Rizzo, A. Hartholt, M. Grimani, A. Leeds and M.

Liewer, Virtual Reality Exposure Therapy for Combat-Related Posttraumatic Stress Disorder, IEEE Computer, Vol. 47, No. 7, 2014, pp. 31-37.

[2]. E. R. Fossum, CMOS Image Sensors: Electronic Camera-On-a-Chip, IEEE Transactions on Electron Devices, Vol. 44, No. 10, Oct., 1997, pp. 1689-1698.

[3]. K. Fife, A. E. Gamal and H.-S. P. Wong., A 3D Multi-Aperture Image Sensor Architecture, in Proceedings of

the IEEE Custom Integrated Circuits Conference, 2006, pp. 281-284.

[4]. T. Sugiki et al., A 60 mW 10 b CMOS image sensor with column-to-column FPN reduction, in Proc. IEEE ISSCC Dig. Tech. Papers, 2000, pp. 108-109.

[5]. S. Lim, J. Cheon, S. Ham, and G. Han, A new correlated double sampling and single slope ADC circuit for CMOS image sensors, in Proceedings of the Int. SoC Des. Conf., Oct. 2004, pp. 129–131.

[6]. M. F. Snoeij et al., Multiple-ramp column-parallel ADC architectures for CMOS image sensors, IEEE J. Solid-State Circuits, Vol. 42, No. 12, Dec. 2007, pp. 2968–2967.

[7]. J. Lee, H. Park, B. Song, K. Kim, J. Eom, K. Kim and J. Burm, High Frame-Rate VGA CMOS Image Sensor Using Non-Memory Capacitor Two-Step Single-Slope ADCs, IEEE Trans. on Circuits and Systems-I: Regular papers, Vol. 62, No. 9, Sep. 2015, pp. 2147–2155.

Page 47: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

47

CCD Camera Characterization by Means of a Custom Integrating Sphere

Andreas W. Winkler 1, Bernhard G. Zagar 1

1Institute for Measurement Technology, Johannes Kepler University Linz, Altenbergerstraße 69, 4040 Linz, Austria

Tel.: +43 732 2468 5929, Fax: + 43 732 2468 5933 E-mail: [email protected]

Summary: A monochrome CCD camera is used in a reflectometer in order to measure the luminance of light reflected by a specimen. This application requires a radiometric calibration of the camera. We use a self-built integrating sphere with an inner diameter of 500mm as a uniform light source in order to characterize the optical system in terms of image non-uniformities introduced by the optics and the fixed pattern noise of the image sensor. We provide an image formation model and propose a new approach for pixelwise compensation of the non-uniformities. Keywords: Integrating sphere, Camera calibration, Flat-field correction, Dark signal, Photo response.

1. Introduction

We use a CCD camera from Allied Vision

Technologies (model GC1600H) to measure the luminance of reflected light in a gonioreflectometer. The reflectometer is used to evaluate the bidirectional reflectance distribution function (BRDF) of metallic specimens [1]. The camera features a monochrome Sony ICX274AL progressive scan 2 megapixel CCD sensor, a 12 bit ADC, and is equipped with a Pentax TV zoom lens (8-48 mm).

The camera supplies the digital image , , measured in ADU (analog-to-digital units). Obviously,

, depends on , , the mean luminance of the pencil of light which is imaged onto pixel , . In an ideal camera, would be proportional to and the factor of proportionality would be the same for all pixels. A real camera sensor, however, has several imperfections which may be categorized in three main groups [2]: (i) fixed pattern noise (FPN), (ii) nonlinearities, and (iii) temporal noise [3].

The word noise in the term FPN may be, although widely adopted in the literature, somewhat misleading, because FPN means spatially varying but temporally constant (and thus deterministic) non-uniformities in acquired images [4]. In the remainder of this paper, the single word noise refers to temporal noise only. In related literature, the FPN is often further divided into the so-called dark signal non-uniformity (DSNU) and the photo-response non-uniformity (PRNU). The DSNU can be observed under dark conditions and the PRNU takes effect as soon as light strikes the image sensor.

In this study, we propose a method for radiometric image correction [5] with respect to the FPN by means of pixelwise evaluation of the DSNU and the PRNU, taking into account the nonlinearities of the sensor and the temporal noise. For this purpose, we use a custom integrating sphere as uniform light source.

1 Without the stabilization, the housing heats up to about

45°C during operation.

It should be emphasized that the proposed calibration procedure comprises the whole optical system, including the lens. Hence, the evaluation of the PRNU has to be repeated as soon as the lens or its parameters (the zoom level, the aperture, etc.) change.

As the characteristics of the camera depend on the temperature of the sensor and the electronics to a high degree [6], we have implemented a temperature stabilization. The stabilization uses a peltier element in a closed loop control to actively cool the aluminum housing of the camera. For all measurements in this paper, we have stabilized the housing temperature at 20 0.1 °C1. This temperature has the advantage

that it is close to the ambient temperature and hence a possible temperature gradient along the housing is minimized. It is also well above the dew point in order to avoid harmful water condensation.

1.1. Mathematical Notation We will denote images as , , as it is usual

practice in digital image processing literature [7] and we will use just , when the meaning is clear from the context.

Throughout this paper, we will make use of two different sample means. We call

⟨ ⟩1

,,

(1)

the spatial mean and

,1

, ; (2)

the temporal mean, where and are the dimensions of an image and ist the number of frames in an

Page 48: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

48

image series. Note that the former is a scalar value, whereas the latter is still an image.

We will also introduce a few random variables and we will make no difference in notation between a random variable and its realization2, whereby the difference is hopefully clear from the context.

2. An Image Formation Model It can be shown [1] that the illuminance , at

a pixel is described by

,4

cos , , (3)

when the optical system is approximated by a single thin lens. Here, , is an attenuation factor which is determined by the transmittance of the lens and distortions in the optical path, is the diameter of the entrance pupil, is the image distance and , is the angle between the principal ray and the optical axis.

Our model assumes that the photocurrent , generated in the sensor cells, is proportional to . From Eq. (3), in turn, it is clear that is proportional to . Hence, the two factors of proportionality can be combined to the single factor , , measured in A/(cd/m²), and the photo current can be written as

, , , . (4) This factor depends, besides the parameters of the optics noted in Eq. (3), on characteristics of the CCD sensor like the effective pixel area, the fill factor, the quantum efficiency, or the charge collection efficiency [2].

There are also thermally induced charges in the sensor cells, even if no light strikes the sensor. These charges are the source of the so-called dark current

, . The dark current is well understood and described in [6], for example. It depends exponentially on the sensor temperature.

During the exposure time of an acquisition, and are integrated and the mean numbers of generated dark and photo charges are

, ,

(5)

with being the elementary charge. The actual numbers and are subject to shot noise and can thus be modeled as Poisson distributed random

2 Some authors use capital letters for a random variable

and lowercase letters for a corresponding realization. 3 There are different kinds of sensor structures

available, the most common being (i) frame transfer,

variables with and being the Poisson parameters: ~ and ~ . Note that the Poisson

parameter denotes the mean as well as the variance of the distribution. For large values of the parameter, the Poisson distribution can be approximated by the normal distribution. The shot noise implies an inherent limit for the attainable signal-to-noise ratio (SNR) of the sensor.

Directly after the exposure, the collected charges are transferred off the sensor3 and are serially applied to a capacity. The capacity converts the charges to a voltage which is amplified and finally AD-converted. This process can be modeled by a bias image , and a factor , which denotes the conversion between the number of charges and the output signal in ADU. The bias image is modeled to contain additive noise. This noise is often referred to as read out noise [3].

The final image is composed of the bias image and the dark-, and the photo signal and can be written as

,

, , , . (6)

2.1. Image Correction

Assuming a constant temperature of the camera, Eq. (5) suggests that, for a given optical system, is just a function of , and is a function of the product

. Hence, Eq. (6) can be rewritten as (omitting the indices and in the following)

. (7)

The fact that is a bijection of the product is an important property, because it means that when we have the photo signal for an acquisition with known exposure time but unknown excitation and we have for the very same a given reference product

, , it holds

, (8) and finally

, . (9)

When the function which maps a photo signal to the corresponding reference product is denoted by

: ⟼ , , (10) we can rewrite Eq. (9) to

(ii) interline transfer and (iii) frame-interline transfer CCDs [2]. The sensor at hand is an interline transfer CCD.

Page 49: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

49

. (11)

In order to be able to evaluate this equation for a

given acquisition , estimates for the bias image and for the functions and are required.

Experiments verified that, as the physical model predicts, the dark signal is linear in and it can hence be mathematically modeled as

. (12)

The photo signal, however, is only approximately

linear [8]. Hence, its inverse is better modeled by a polynomial of order than by a simple proportional factor:

, . (13)

In our tests, a polynomial order 4 has proven

to provide sufficient accuracy.

3. Measurements The estimates and , 1… 4 of the

coefficients of and are determined by polynomial regression (Sections 3.1 and 3.3). The estimate of the bias image is a combination of the temporal mean and (Sec. 3.2), whereby is a byproduct of the determination of .

3.1. Evaluation of the Dark Signal The dark signal is evaluated by means of images

which were acquired with the lens covered so that no light strikes the sensor, i.e. is suppressed. Hence, we use

(14)

to estimate and by means of a polynomial regression on the basis of average images for 128 different exposure times between 10µs and 2s. For each , 256 acquisitions are averaged to compute

. The estimate , measured in ADU/s, is the slope

of the regression line and is also called the dark signal rate. Fig. 1 shows the dark signal for three typical pixels and Fig. 2 shows the histogram4 for the estimate

. Note that the histogram is cut off at 35ADU/s, whereby about 0.8% of the pixels have a higher dark signal rate. 495 pixels have an even higher than

4 This and all following histograms are normalized such

that the covered area equals 1.

100ADU/s. Pixels with such a high dark signal rate are commonly referred to as hot pixels. The mean is ⟨ ⟩ 8.7ADU/s.

Fig. 1. Dark signal for three typical pixels, the estimated regression lines are shown dashed.

Fig. 2. Histogram of the estimated dark signal rate.

A measure for the quality of the regression is the

mean squared error (MSE) , which is depicted in the histogram in Fig. 3. This histogram is cut off at

2ADU² and shows about 97.9 % of all values.

The mean is ⟨ ⟩ 0.84ADU².

Fig. 3. Histogram of the MSE.

Page 50: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

50

The MSE is useful to identify pixels with untypical dark signal behavior. These pixels with an unusually high MSE may have a quite noisy dark signal or the dark signal rates jump unpredictably between two or more discrete values. Fig. 4 shows three examples of untypical dark signal behavior. These imperfections of a few pixels are presumably caused by particular semiconductor effects. However, we could not yet find a satisfying explanation in the literature.

Fig. 4. Three examples for pixels with a high MSE and untypical dark signal behavior.

3.2. The Bias Image

An estimate of the bias image is given by the coefficient as a byproduct from the regression for the dark signal (Sec. 3.1). This estimate is, however, not suitable for pixels with a high MSE: It can be seen in Fig. 4 that the intercepts of the regression lines for these pixels do not correspond to what we would intuitively expect to be the bias .

Thus, we calculate a second estimate by averaging 256 2048 bias images which are acquired with the lens covered and at the minimally possible exposure time of 10µs. The final bias image estimate is then defined to be the composition

,if , 5

else, (15)

whereby the number 5 has been empirically determined as the threshold for the MSE, to whom the dark signal is highly unlikely to show untypical behavior.

The final bias image estimate is shown in Fig. 5. It shows a vertically striped pattern and has a darker region on the very left. These effects may be caused by the readout electronics and a non-uniform temperature distribution at the sensor.

The histogram of is shown in Fig. 6. The two lobes indicate the striped pattern and the shoulder on the left of the histogram corresponds to the darker region on the left edge of the bias image.

Fig. 5. The estimated bias image.

Fig. 6. Histogram of the bias image estimate .

3.3. Evaluation of the Photo Signal

The photo signal is evaluated by means of images which were acquired with the lens directed at the exit port of an integrating sphere (Sec. 4). The sphere provides a homogenous light field. Thus, it is assumed that each pixel is exposed to the same luminance

, . Usually, is not exactly known. However, for

most applications it is sufficient or even desired to utilize as an arbitrary normalization factor, as long as it is the same for the whole image. Here, we use

1 for reasons of simplicity. Hence, from Equations (7) and (13), we use

, , (16)

with

, , , (17) to calculate a polynomial regression on the basis of 128 mean images , for different exposure times

, between 10µs and , 10.02ms, whereby

, is chosen such that just no pixel saturates. For

each , 128 acquisitions are averaged to compute . The result of the regression for three typical pixels is shown in Fig. 7.

Page 51: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

51

Fig. 7. for three typical pixels, the estimated polynomials are shown dashed.

4. The Integrating Sphere

The integrating sphere with a diameter of 500 mm provides a uniform and fully diffuse light field at the 140 mm wide exit port (Fig. 1). The sphere is made of two metal-spun and powder-coated aluminum hemi-spheres. Inside it is additionally coated with white and diffusely reflecting barium sulfate with a reflectance of more than 97 % in the visible wavelength range. It is illuminated by six high power LEDs with three of them having a correlated color temperature (CCT) of 3000 K and a luminous flux of 300 l m@1050 mA and the other three having 5700 K and 360 l m@1050 mA. It is possible to set a desired CCT by means of driving the two LED types with appropriate currents. With all six LEDs at nominal current of 1050 mA, the luminance at the exit port is about 1.46 10 cd/m . The LEDs can temporarily be operated at up to 3A to increase light output even further.

For our experiments, all six LEDs have been operated at a constant current of 500 mA.

Fig. 8. The integrating sphere is used to evaluate the non-uniformities of the CCD sensor.

The performance of the sphere compares to those commercially available, while the manufacturing costs were just about one-tenth. 5. Conclusion

A CCD camera (including the lens) has been radiometrically characterized by means of a self-built 500mm integrating sphere. The image non-uniformities have been evaluated in terms of the DSNU and the PRNU and a method for pixelwise image correction using the measured data has been proposed. Acknowledgement

The authors gratefully acknowledge the partial financial support by the Austrian Research Promotion Agency and the Austrian COMET program. References [1] A. Winkler, B. Zagar, Building a gonioreflectometer -

a geometrical evaluation, in Proceedings of the IEEE International Instrumentation and Measurement Technology Conference (I2MTC’15), 2015, pp. 1900-1905.

[2] J. Nakamura, Image Sensors and Signal Processing for Digital Still Cameras, CRC Press, 2005.

[3] K. Irie, A. E. McKinnon et al., A Technique for Evaluation of CCD Video-Camera Noise, IEEE Transactions on Circuits and Systems for Video Technology, Vol. 18, No. 2, 2008, pp. 280-284.

[4]. J. Lukás et al., Digital Camera Identification from Sensor Pattern Noise, IEEE Transactions on Information Forensics and Security, Vol. 1, 2006, pp. 205-214.

[5]. G. Healey, R. Kondepudy, Radiometric CCD camera calibration and noise estimation, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 16, No. 3, 1994, pp. 267-276.

[6]. R. Widenhorn et al., Temperature dependence of dark current in a CCD, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, 2002, pp. 193-201.

[7]. R. C. Gonzalez, R. E. Woods., Digital Image Processing, 3rd Ed., Pearson Prentice Hall, 2008.

[8]. E. G. Stevens, Photoresponse Nonlinearity of Solid-State Image Sensors with Antiblooming Protection, IEEE Transactions on Electron Devices, Vol. 38, No. 2, 1991, pp. 299-302.

Page 52: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

52

Reflective Multimode Interference Structures

for Liquids Remote Detection

L. R. Villarreal Jiménez 1, S. Enríquez Sías 1, C. Elizondo González 1, R. Domínguez-Cruz 2, G. Romero-Galván 2 and Ma. José Erro Betrán 3

1 FIME. Universidad Autónoma de Nuevo León, Av. Universidad S/N, C.P. 66451Monterrey, NL., México. 2Universidad Autónoma de Tamaulipas, Ap. Postal 1460. Col. Arcoíris, Reynosa, Tam. México. 3UPNA Universidad Pública de Navarra. Campus Arrosadia - 31006 Pamplona-Iruña, España.

E-mail: [email protected] Summary: In this paper we show the preliminary experimental results of a fiber optic sensor using multimode interference effects in reflection scheme. The system, applied to detection of liquids via remote, is built by optical interrogator provided with a laser diode linked to Single Mode fiber (SMF). At the second end, the SMF is spliced with a MMI element which is fabricated with a segment of No-Core Multimode fiber (NC-MMF). We report a sentivity sensor is increased when the diameter of the NC-MMF is reduced at 80 µm by mechanical tapering technique. The sensor shows its able to detect the presence of liquid media operating at least 10 km of distance. Keywords: Fiber optic, Fiber optic sensors, Multimode interference effect, Refractometer.

1. Introduction Fiber optic sensors (FOS) mean a very useful and

important tool in the field of metrology sciences and engineering due to its great impact in medicine, civil engineering, biology and environment analysis. The relevant roll is due to compact size, low weight, freedom from corrosion and invulnerability to electrical interference [1]. Some techniques based in Bragg gratings (BGs) [2], Long Period Fiber Bragg gratings (LPFGs) [3], Surface Plasmon Resonance (SPR) [4], and Photonics Crystal Fibers (PCFs) [5] has been proposed to measure parameters such as temperature, stress and refractive index among others. In addition, the combination of elements built with PCF–FBG sensor has been able to measure in a remote operation for oil and gas sensing applications [6]. However, a simple, cheap and easy scheme used in fibers as novel technique is based on the effect of multimodal interference (MMI) [7], which consists of a segment of multimode fiber (MMF) spliced between two lengths single mode fiber (SMF). By propagating an optical field in this structure, self-images of the input signal are generated, which are reproduced at periodical intervals along the direction of propagation of the MMF. Due to the evanescent field of the modes propagated, it is possible interaction with the surrounding environment and thus able to measure any physical variable of interest as refraction index (RI) or temperature [8]. In this paper we describe a structure of MMI in reflective configuration to detect liquids. We report a sentivity sensor is increased when the diameter of the NC-MMF is reduced at 90 µm by mechanical tapering technique and capable to detect liquids media using 10 km fiber lengths.

2. Theoretical Considerations The studies of multimode interference (MMI)

effect was initially described in waveguides [7]. This phenomena consist in the construction of self-images from an input field at periodic distances due to constructive interference effects. However, the same effects can be performed in fiber optics. In this case, the structure of MMI element is showed in Fig. 1.

Fig. 1. Diagram of a Multimode Interference effect built in fiber optic (SMF- single mode fiber, NC-MMF-

multimode fiber).

As we can see, the MMI in fiber is built by a segment of No-Core Multimode fiber (NC-MMF) spliced between two Single mode fibers (SMF). The MMF element of MMI structure can support many guided modes. For an input field coupled to the MMF, it can reproduce a single self-image or multiple- images (called pseudo-images) at regular intervals along the MMF waveguide due to constructive interference between all guided modes. To reproduce a self-image, the phase difference between all guided modes must to be an integer multiple of 2π. In consequence, all modes are in phase (constructive interference) and the input field can be reproduced at the end of the MMF segment. Thus, the length where self-images are formed in NC-MMF segment can be calculated using [7]:

Page 53: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

53

with p= 0, 1, 2, 3… (1)

where Lπ is the beat length defined as:

4

3 (2)

with nMMF and DMMF are respectively the refractive index and diameter of the MMF core, with λ0 as the free-space wavelength.

3. Experimental Results and Discussion The experimental array (Fig. 2) consist in a optical

interrogator PXIe (National Instruments®) which is provided with a laser diode (spectral range from 1510 nm to 1590 nm) and a maximum output power of 0.25 mW. The laser source is coupled to 10 km of SMF-28 (Thorlabs®) segment which is spliced at the end by an MMI element. The MMI is built using a section of NC- MMF (FG125LA Thorlabs®, 125µm). For puntual measurements, we use a reflective configuration. In this case, at the end of NC-MMF was collocated a silver paint layer by sputtering technique (Fig. 2-a). Additionally a second MMI tapered was designed. If the diameter of the NC-MMF is reduced, the interaction of the evanescent field with the surrounding environment is increased (Fig.2-b).

Fig. 2. Experimental set-up to detect the presence of liquids

via MMI (M-mirror). We test our system to detect water in remote

configuration (Fig. 3). To explain the curves, consider the NC-MMF. In this kind of fiber, the core is exposed, i.e., no cladding is surrounds with air (transmitted peak is around 1536 nm). Then, the NC- MMF section plays the sensing role since this section has not a cladding and its exposed core acts as a sensing element when is surrounded with any medium. Under this situation, the water acts as the cladding for this fiber and the signal transmitted in NC-MMF. This effect is produced due to the MMF section is designed for specific length to reproduce only a pseudo-image of the input profile. The shift showed (1547 nm, red line, Fig. 3) for this pseudo-image will generate different transmitted intensity responses for each liquid which the sensing fiber was immersed (for water. In addition, the sensitivity is increased is due to reduction to diameter effective of the NC-MMF (1559 nm, green line, Fig. 3).

Fig. 3. Spectrum reflected of MMI for different surrounding conditions.

4. Conclusions

We report the preliminary experimental results of a

fiber optic sensor using multimode interference effects in reflection scheme. The system is assembled by optical interrogator and a laser diode, which is plugged to 10 km of SMF. In the second end, the SMF is spliced with a MMI structure which is fabricated with a segment of No-Core Multimode fiber. We report a sentivity sensor is increased when the diameter of the NC-MMF is reduced at 80 µm by mechanical tapering technique. The system shows its able to detect the presence of liquid media in remote operation. Further experiments will be showed in extended version of this paper and it suggest a use as tool for leak detection or quality control mechanism for purity measurements in liquids.

References

[1]. S. Yin et al., Fiber Optic Sensors, Second Edition, CRC Press, 2008.

[2]. A. Iadicicco, S. Campopiano, A. Cutolo, M. Giordano, A. Cusano, Self refractive index sensor by non- uniform thinned fiber Bragg gratings, Sensors and Actuators B: Chemistry, 120, 2006, pp. 231–237.

[3]. Y. Tan, W. Ji, V. Mamidala, K. Chow, S. Tjin, Carbon- nanotube-deposited long period fiber grating for continuous refractive index sensor applications, Sensor and Actuators B: Chemistry 196, 2014, pp. 260–264.

[4]. H. Liang, H. Miranto, N. Granqvist, J. W. Sadowski, T. Viitala, B. Wang, M. Ylipert-tula, Surface plasmon resonance instrument as a refractometer for liquids and ultrathin films, Sensors and Actuators B: Chemistry, 149, 2010, pp. 212–220.

[5]. Y. Wang, X. Tan, W. Jin, D. Ying, Y. L. Hoo, S. Liu, Temperature-controlled trans-formation in fiber types of fluid-filled photonic crystal fibers and applications, Optics. Letters., 35, 2010, pp. 88–99.

[6]. J. Johny, R. Prabhu and W. Fung, Investigation of structural parameter dependence of confinement losses in PCF–FBG sensor for oil and gas sensing applications, Optical Quantumm Electroncs, 48, 2016, p. 252.

Page 54: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

54

[7]. L. B. Soldano, E. C. M. Pennings, Optical multimode interference devices based on self-imaging: principles and applications, Journal of Lightwave Technology. 13, 1995, pp. 615–627.

[8]. J. E. Antonio López, D. López Cortes, M. A. Basurto Pensado, D. A. May Arrioja, J. J. Sánchez Mondragón, All-fiber multimode interference refractometer sensor, Proceedings SPIE, 7316, 2009, 73161F.

Page 55: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

55

Simultaneous Electrocatalytic Voltammetric Determination

of Dihydroxybenzene Isomers Using a Cobalt-phthalocyanine Modified Pencil Electrode

A. Ciucu 1, M. Buleandra 1 A. Rabinca 1 and I. Stamatin 2

1 University of Bucharest, Faculty of Chemistry, Department of Analytical Chemistry, 90-92 Panduri Avenue, 050663, Bucharest, Romania

2 University of Bucharest, Faculty of Physics, 3nano-SAE Research Centre, MG38-Magurele, Bucharest, Romania Tel.: + 40720122902, fax: + 40214102279

E-mail: [email protected] Abstract A novel assay for the electrochemical detection of hydroquinone (HQ), catechol (CC) and resorcinol (RS) based on pencil graphite electrode (PGE) modified with cobalt phthalocyanine (CoPc) has been investigated. The results indicated that the modification of the activated pencil graphite electrode (PGE*) with this CoPC results in amplification of the oxidation responses of dihydroxybenzene isomers in contrast to that on the unmodified PGE*. The electrochemical behavior of the modified electrode and the mechanism of the oxidation of dihydroxybenzene isomers were investigated using cyclic voltammetry (CV) and differential pulse voltammetry (DPV). This method has been successfully applied to the direct determination of dihydroxybenzene isomers in tea samples. Keywords: Simultaneous determination, Dihydroxybenzene isomers, Cobalt-phtalocyanine, Pencil graphite electrode, Tea samples. 1. Introduction

Dihydroxybenzene isomers are widely used in many fields, such as tanning, cosmetic, dye, chemical and pharmaceutical industries [1]. It is very important to develop simple, sensitive, and selective analytical methods for simultaneous determination of dihydroxybenzene isomers. Voltammetric methods have some specific advantages like simplicity, rapidity and low reagent consumption, which make them less expensive when compared with other analytical techniques [2, 3]. The present paper describes the voltammetric behavior of catechol (CC), hydroquinone (HQ) and resorcinol (RS) at the electrochemically activated cobalt-phthalocyanine modified pencil graphite electrode (CoPC-PGE*). Good analytical performance was achieved by a DPV method. The oxidation peak potentials were separated in neutral condition with 101 mV to HQ and catechol CC and 388 mV to CC and RS. The problems related to the surface contamination of this electrode have been resolved by using a disposable electrode. 2. Instrumentation

All voltammetric measurements were performed with an AUTOLAB electrochemical analyzer (PGSTAT 128N Ecochemie B.V., Netherlands). A three electrode cell consisting of a non-activated or activated pencil graphite electrode (PGE) un-modified or modified with cobalt-phtalocyanine (CoPC) as the working electrode, an Ag/AgCl (3.0 M KCl) as the reference and a platinum wire as the auxiliary electrode were employed.

3. Simultaneous Determination of Dihydroxybenzene Isomers DPV results showed three well-distinguished

anodic peaks for HQ, CC and RS at 102 mV, 203 mV and 591 mV, respectively (Fig. 1); therefore the simultaneous determination of the three isomers was possible at the surface of electrochemical pretreated modified electrode (CoPC-SPGE*) without previous separation. The detection limits are 0.10 M for HQ, 0.31 M for CC and 0.72 M for RS, respectively.

0.0E+00

3.0E-06

6.0E-06

9.0E-06

1.2E-05

-0.2 0 0.2 0.4 0.6 0.8 1E / V

I / A

HQ

CC

RS

HQ+CC+RS

BR pH 6.80

Fig. 1. Differential pulse voltammograms of 25 M HQ, 25 M CC, 25 M RS and a mixture of HQ, CC and RS

(each of 25 M) at CoPC-PGE* in BRB solution pH 6.80.

The applicability of the proposed method was evaluated for the quantitative determination of HQ, CC and RS in tea infusion samples. The standard additions method was used to avoid the possible effects due to matrix. In order to verify the accuracy

Page 56: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

56

of the proposed method, the content of HQ, CC and RS in tea infusion were also determined by HPLC. 4. Conclusions

A novel voltammetric approach for simultaneous and selective detection of multiple isomers is proposed. In this work, the CoPC modified PG electrode and DPV technique were first used to determine three dihydroxybenzene isomers simultaneously and quantitatively. The new developed DPV method is simple, rapid and enables the assessment of HQ, CC and RS in a mixture. Acknowledgements

The financial support of Romanian Grant PN-II-

ID-PCE-2011-3-0784 - contract no 251/2011 funded by UEFSCIDI is acknowledge.

References [1]. M. Buleandra, A. A. Rabinca, C. Mihailciuc, A. Balan,

C. Nichita, I. Stamatin, A. A. Ciucu, Screen-printed Prussian Blue modified electrode for simultaneous detection of hydroquinone and catechol, Sensors and Actuators B, 203, 2014, pp. 824–832.

[2]. I. Balan, I. G. David, V. David, A. I. Stoica, C. Mihailciuc, I. Stamatin, A. A. Ciucu, Electrocatalytic voltammetric determination of guanine at a cobalt phthalocyanine modified carbon nanotubes paste electrode, Journal of Electroanalytical Chemistry, 654, 2011, pp. 8–12.

[3]. D. Patrascu, I. G. David, V. David, C. Mihailciuc, I. Stamatin, J. Ciurea, L. Nagy, G. Nagy, A. A. Ciucu, Selective voltammetric determination of electroactive neuromodulating species in biological samples using iron(II) phthalocyanine modified multi-wall carbon nanotubes paste electrode, Sensors and Actuators B, 156, 2011, pp. 731–736.

Page 57: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

57

Nanoplasmonic Sensing of Pb-acid and Li-ion Batteries

David Johansson 1, Jenny Andersson 1, Björn Wickman 2,

Fredrik Björefors 3, Adam Sobkowiak 3, Bengt Kasemo 2

1 Insplorion AB, Medicinaregatan 8A, 413 90 Gothenburg, Sweden 2 Department of Physics, Chalmers University of Technology, Kemivägen 9, 412 96 Gothenburg, Sweden

3 Department of Chemistry – Ångström Laboratory, Uppsala University, Lägerhyddsvägen 1, 75121, Uppsala, Sweden

Tel.: + 46760460656 E-mail: [email protected]

Summary: The increasing sophistication and performance of batteries are connected with more complex chemical and physical battery processes and increase the need of more direct and informative measurements, both in the R&D phase and for monitoring and control during operation of vehicles. Todays potentiometric based measurement sensors are not sufficiently accurate for optimal battery sensing. To avoid the built in wide safety margins new, more informative monitoring signals are therefore desired or needed. In this study the optical technology NanoPlasmonic Sensing (NPS) has been used to in-situ monitor the charge and discharge processes of lead-acid and Li-ion batteries. The optical signals were found to correlate well with charging/discharging of both battery technologies. Keywords: Battery management system, Nanoplasmonic sensing, Lead acid, Li-ion, Electrical vehicles, SOC, SOH.

1. Introduction

The energy demand in the transport sector, and the pressure to reduce climate affecting gas emissions are constantly growing which put pressure on more efficient energy storage solutions with smaller impact on the environment. This goal can be met with a transition from fossil driven to electrical driven vehicles. The development and monitoring of new batteries and battery systems are highly important in this context. An important component is development of more precise and informative sensor systems.

1.1. Background

Todays potentiometric measurements of batteries do not provide sufficiently good estimations of State of Charge (SOC) of batteries. The open circuit voltage of batteries changes during the charge and discharge processes, over time, and with temperature, and in individual cells which makes voltage measurements an insufficient tool. Many of the fundamental processes such as charge transfer, chemical transformations, and temperature gradients, are not fully understood in batteries. Wide safety margins are therefore built into the management system of the battery, thus the full capacity is not utilized. A more accurate SOC sensor would enable an increased using range of the battery capacity, which should decrease the cost. It is also of interest to find new ways of local in-situ measurements of ageing mechanisms that take the chemical and structural changes into account, both for the built in sensor inside of commercial batteries, and for research studies during the product development. This is related to the State of Health (SOH) of the batteries.

In a battery the electrodes change their chemical composition during charge and discharge process, a

characteristic not directly considered in potentiometric measurements. The electrolyte, which act as a transport medium of the active material, also change its composition for some batteries, such as in Pb-acid batteries. The objective of this project is to use the optical technology nanoplasmonic sensing (NPS) to map the charge characteristics in Pb-acid and Li-ion batteries. 1.2. Nanoplasmonic Sensing

Nanoplasmonic sensing (NPS) is an optical

technology based on Localized Surface Plasmon Resonance (LSPR), wherein nanofeatures (e.g. gold nanoparticles) are used as sensing elements. The sensor is sensitive to changes in the refractive index in the volume within a few tens of nanometers from the nanoparticles. This allows sensitive local and real time in situ tracking of various surface processes such as adsorption/desorption phenomena, specific molecular interactions, structural changes, and temperature shifts. The plasmonic sensor structures in this study have been fabricated with the established fabrication method Hole-Mask Colloidal Lithography [1]. 2. Experimental Results

The first verifying experiments were on a lead acid electrode without plasmonic sensors. During the reaction the electrode converts between PbO2 and PbSO4, which changes the dielectric property of the material. In Fig. 1 the derivative of the variation of the optical reflectance is shown during the charge and discharge cycles. The optical signal correlates well with the corresponding current, with a high signal to noise level.

Page 58: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

58

Fig. 1. Derivative of the variation of optical reflectance from the lead film (blue line), illustrated together with the

corresponding current (green line).

In a lead acid battery the sulfuric acid concentration follows the charge condition; high sulphuric acid concentration at high charge level and vice versa, which will change the dielectric constant. Fig. 2 shows how the LSPR signal varies with varied dilution level of the sulfuric acid.

Fig. 2. NPS signal difference (peak shift) during dilution of

H2SO4

The experiments on Li-ion batteries were made in a closed package structure. A sensor with deposited

ITO-glass was used as current collector, which enabled optical access to the studied electrode material, LiFePO4. The plasmonic response upon charge and discharge in the electrode is shown in Fig. 3 together with the corresponding voltage. Notice the linear response of NPS signal in the area of operation (capacity between 20 % and 80 %). The change of optical signal is high compared to the potentiometric signal, which makes NPS a promising method for local, real time monitoring of charge capacity.

Fig. 3. NPS signal and voltage during charging (dashed lines) and discharging (solid lines) of a LiFePO4 electrode.

3. Conclusions

Optical measurements on electrode and electrolyte composition correlates well with charge/discharge processes. The NPS-technology has been verified both for Pb and Li-ion battery systems. Future studies will consider the possibility of local measurement of temperature, and to miniaturize the technology with all fiber-optic sensors, see Fig. 4. This will both enable a productification of the technology to an on-board sensor for real-time monitoring of the charge level and age mechanisms, and for fundamental studies of material changes during research and development of new batteries.

Fig. 4. Future on-board sensor design based on optical fibers, which enable local measurement of battery charge level.

Page 59: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

59

Acknowledgements

This work was financed by the Swedish Governmental Agency for Innovation Systems, the Swedish Energy Agency in the Battery Foundation Program, and the Swedish company Insplorion AB.

References [1]. H. Fredriksson, Y. Alaverdyan, A. Dmitriev, C.

Langhammer, D.S. Sutherland, M. Zäch, B. Kasemo, Hole-Mask Colloidal Lithography, Advanced Materials, Vol. 19, Issue 23, 2007, pp 4297-4302.

Page 60: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

60

pH Measurements Of nano-liter Solutions Using THz Technology

Y. Zhou, K. Akimune, K. Hamada, K. Sakai, T. Kiwa, K. Tsukada

Graduate School of Natural Science and Technology, Okayama University, 3-1-1 Tsushimanaka, Kitaku, Okayama 700-8530, Japan,

E-mail: [email protected] Summary: A terahertz chemical microscopy (TCM) has been proposed and developed to visualize the electric potential changes of water solution on the semiconductor-based plate named ‘sensing plate’. The TCM detects the change in the amplitude of THz waves from the sensing plate, which can be related to the electric potential of the water solutions on the sensing plate. Since the top surface of the sensing plate was made from the SiO2, the electric potential at the surface of the sensing plate could be change by changing the pH values of the solutions on the sensing plate. In order to detect the small amount of water solutions, micro-wells made of UV curable resin were patterned on the surface of the sensing plate with diameters range between 50 m to 500 m. Thus the solution of 16-nL were measured, which indicated that the pH measurements of nano-litter solutions could be possible using TCM. Keywords: THz, TCM, pH measurement, Ion sensing. 1. Introduction

The pH measurement is very important in many

liquid chemical processes, such as industrial, pharmaceutical, food production, and environment. In the medical point of view, the pH value of human body is from 7.2 to 7.4, to maintain the pH balance of human body is also important for monitoring health. Meanwhile, there are various ions in our body, such as sodium ion and potassium ion. Breaking of ion balances will lead to serious diseases. Therefore, the sensing systems for ions and pH values are necessary in medical diagnoses. In our group, a terahertz (THz) chemical microscopy (TCM) has been proposed and developed to visualize the electric potential changes of the water solution on the semiconductor-based plate named ‘sensing plate’. The THz wave is generated from the sensing plate by irradiating a femtosecond laser to sensing plate and the TCM can detect the change in the amplitude of radiated THz waves from the sensing plate, which can be related to the electric potential of the water solution on the sensing plate. The mechanism is very close to a LAPS (light-addressable potentiometric sensor) technique, however, although the LAPS generally detects the photocurrents follows in the semiconductor plate, the TCM detects the THz waves directly emitted to free space. So, TCM does not require any electrode on the sensing plate. Up to now, we have evaluated concentrations of various types of ions using TCM. Here, in order to detect the small amount of water solutions, micro-wells made from UV curable resin were patterned on the surface of the sensing plate with the well diameter of in the range between 50 m to 500 m. Thus the solution of 16-nL was measured.

2. Method

The sensing plate consists of Si and SiO2 layer on a sapphire substrate with thickness of 150 and 275 nm,

respectively. As femtosecond laser pluses irradiates to the sensing plate from the sapphire side, free carriers are excited and accelerated by the depletion field of the Si thin film, which was formed at the boundary of the Si and SiO2 layers. As the result of ultrafast carrier acceleration THz waves were generated.

When pH standard solution was dropped onto the sensing plate, the chemical reactions occurred at the surface of SiO2, which can be expressed by (1), (2)

HSiOHSiOH2 (1)

HSiOSiOH (2)

As the result, the surface potential of the sensing plate shift depending of the pH values of the solutions, which lead to the change in the depletion layer of the sensing plate. Thus, shift in the pH values of the solution on the sensing plate could be measured as the shift in the amplitude of THz waves.

In order to detect the small amount of water solutions, micro-wells made of UV curable resin were patterned on the surface of the sensing plate using a conventional photolithography. The size of the micro wells were in the range between 50 m to 500 m in diameter. Fig. 1 shows the schematic design of the wells fabricated on surface of sensing plate.

Fig. 2 shows the schematic of sensing plate in our experiment. 16-nL-solutions with different pH values were dropped by a jet pipette into the cavities with the diameters of 250 m. The volume of the droplets was calibrated by measuring the weight of 100- droplets and averaging. The laser spot was fixed at beneath the sample solutions to generate the THz waves.

3. Results

Fig. 3 (a) and (b) show the amplitude maps of radiated THz waves obtained by scanning the femtosecond laser across the sensing plate surface

Page 61: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

61

when the water solutions with the pH values of 10.01 and 6.87 were injected into the wells, respectively. The pH stabilized buffer solutions were used as the water solutions. The amplitude of THz waves were enhanced at beneath the well filled with the solutions. One can also see that the amplitude of THz waves for pH values shows the different amplitude. These results suggest that TCM can detect pH in solutions as the amplitude change of radiated THz waves even when the amount of the water solution is only 16 nL.

5 mm

Fig. 1. Wale on sensing plate.

Sensing plate

UV curable resin(NOA65)

F 250 m

3 m

THz radiation

Femtosecond Laser pulse

Sample solution

F 400 mcavity

Fig. 2. Schematic of a sensing plate.

Additionally, in this experiment, the reference electrodes were not used. However, since the volume of the water solution was extremely small and the potential of the water solution might be electrically charged up. Thus the pH values could be stably

measured without reference electrodes. The electric field distribution on the sensing plate is now underway.

0.0 0.2 0.4 0.6 0.8 1.00.0

0.2

0.4

0.6

0.8

1.0

TH

z A

mp

litu

de

(a

.u.)

Y a

xis

(mm

)

X axis (mm)

-2.0

-1.0

0.0

1.0

2.0(a) (b)

0.0 0.2 0.4 0.6 0.8 1.00.0

0.2

0.4

0.6

0.8

1.0

Y a

xis

(mm

)Y

axi

s (m

m)

X axis (mm)

Fig. 3. The maps of change in the THz amplitude with different pH solution of (a) 10.01 and (b) 6.86.

4. Conclusions

THz chemical microscopy has been proposed and developed to measure ion concentration and pH values of the water solutions. Micro-wells were fabricated on the surface of the sensing plate using conventional photolithograph technique. As results, the pH values of buffer solutions with the volume of only 16 nL could be measured. Acknowledgements

This work was partially supported by MEXT

Subsidy of the Acceleration of a Translational Research Network Program and KAKENHI Grant Number 16H03887. References [1]. Y. Zhou, et al., September. Presented at JSAP-OSA

Joint Symposia (2015) Nagoya. [2]. T. Kiwa, et al., Chemical sensing plate with a laser-

terahertz monitoring system, Applied Optics, Vol. 47, Issue 18, 2008, pp. 3324-3327.

[3]. T. Kiwa, et al., A terahertz chemical microscope to visualize chemical concentrations in microfluidic chips, Japanese Journal of Applied Physics, Vol. 46, Issue 11L, 2007, pp. L1052.

[4]. K. Akimune, et al., Multi-ion sensing of buffer solutions using terahertz chemical microscopy, Applied Physics Express, Vol. 7, Issue 12, 2014, pp. 122401.

[5]. T. Kuwana, et al., Label-free detection of low-molecular-weight samples using a terahertz chemical microscope, Applied Physics Express, Vol. 9, Issue 4, 2016, pp. 042401.

Page 62: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

62

Voltammetric Detection of ∆9-THC Using Carbon Screen Printed Electrode

in Aqueous Media: Improvements in Forensic Analysis

M. A. Balbino 1,2, I.C. Eleotério 1,2, B. R. McCord 2, and M. F. Oliveira 1 1 Universidade de São Paulo/ FFCLRP, Departamento de Química, Av. Bandeirantes,

14040901, Ribeirão Preto, SP, Brazil 2 Florida International University, Chemistry and Biochemistry Department, Sw 8th St, 33199, Miami, Fl, USA

Tel.: +551633153750, fax: + 551633154370 E-mail: [email protected]

Summary: Considerable efforts are notable for improving new and cheaper methods for detecting illicit drugs which show reliability and robustness in their results. Transducers and electrochemical sensors are reported in the literature for detecting several illegal drugs such as ∆9-THC, a psychoactive substance of marijuana. Screen printed electrodes are easy to operate, portable, inexpensive, and excellent sensitivity. They contain all electrodes necessary for voltammetric measurements in one device. In this work, we adopted a carbon screen printed electrode for voltammetric analysis of ∆9-THC using optimized parameters reported in the literature. The study was performed by cyclic voltammetry using potassium nitrate 0.15 mol L-1 as supporting electrolyte. The ∆9-THC standard solution was added to the surface electrode by drop coating method. The results showed versatility during the investigation, good sensibility and a low limit of detection of 1.0 µmol L-1. Keywords: Detection, Voltammetry, Screen printed electrodes, Delta-9-tetrahydrocannabinol, Versatility.

1. Introduction

Cannabis is a known and most illicit drug widely

used in the world. The primary psychoactive substance is named ∆9-tetrahydrocannabinol (∆9-THC), and their use can be associated with a development of psychosis. The presumptive test is performed after extraction of cannabinoids in organic media and an addition of colorimetric reagents such as Fast Blue B salt or Duquenois–Levine. A change of color is observed when ∆9-THC and other cannabinoids (Cannabinol, Cannabidiol) are present. However, these tests are vulnerable to false positive and false negative results [1].

Nowadays, new and cheap methods are developed for analyzing illicit drugs. The voltammetric technique is an attractive alternative for the development of new methods. Screen-printed electrode (SPE) is a device which contains three electrodes for voltammetric measurement. It has low dimensions, designated for working with microvolume of analytes. These advantages became the voltammetric analysis faster and cheap. Some studies for detection of illicit drugs such as cocaine, THC and ecstasy are reported in the literature using SPE devices [2]. The aim of this work is to verify the electrochemical behavior of Δ9-THC using a supporting electrolyte in aqueous solution. 2. Materials and Methods

A potentiostat Autolab model 128 N was used for all voltammetric measurements with the carbon screen printed electrode (CSPE) from DropSens model DRP-110. A commercial standard solution of Δ9-THC

1 mg/mL from Cerilliant was diluted in methanol. A potassium nitrate solution 0.15 mol L-1 was used in this study as supporting electrolyte.

The method for additioning Δ9-THC standard solution described by Novak et al. [3] was employed Using cyclic voltammetry for analysis of Δ9-THC. The influence of concentration was studied using a potential of −0.5 V for 60 seconds to preconcentrate Δ9-THC species on the electrode surface, a potential range from −0.8 V to 0.9 V (vs. Ag - SPE) at a scan rate of 100 mV s−1. 3. Results and Discussion

The voltammetric response of ∆9-THC standard

solution at 0.13 V vs. Ag-CSPE occurred after addition of 4 µmol L-1 on the surface electrode. Fig. 1 shows the cyclic voltammograms. This The anodic peak current (Ipa), and oxidation signal with one electrochemical process involving one-electron and one-proton can be attributed to the oxidation of the phenol group [31-35]. The Ipa values increase with linearity in proportion to the concentration (4- 20 µmol L-1). The analytical curve showed the values of r2 (0.997) and standard deviation, SD (0.01 µA). The following equation is:

Ipa = 0.27 µA + 0.04 µA/µmol L-1∆9-THC (1)

These results allowed a limit of detection (LOD) of 1.0 µmol L-1 and a limit of quantification (LOQ) of 3.3 µmol L-1 using the relations 3SD/m and 10SD/m (where m is the amperometric sensitivity of the curve).

Page 63: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

63

-0.8 -0.4 0.0 0.4 0.8

-3

-2

-1

0

1

2I /

A

E / V

9-THC (mol L-1)

22 20 18 17 16 13 12 10 4 SE

0 6 12 18 24

0.4

0.6

0.8

1.0

1.2

I pa/A

9-THC (mol L-1)

Fig. 1. Cyclic Voltammograms of ∆9-THC standard solution, analytical curve.

4. Conclusions The voltammetric analysis purposed detected Δ9-THC in lower concentrations. The analytical curve method can be useful for quantitative analysis of seized samples. The applications of CSPE saved the time of analysis and volume of chemicals, indicating a good alternative in forensic investigations.

Acknowledgements We would like to thank CAPES Foundation (Edital Pró-Forenses 25/2014) and FAPESP for financial support. References [1]. J. F. Kelly, K. Addanki, O. Bagasra, The non-

specificity of Duquenois-Levine field test of marijuana, The Open Forensic Science Journal, 5, 2012, pp. 4-8.

[2]. M. A. Balbino, E. N. Oyie, M. F. M. Ribeiro, J. W. Cruz Júnior, I. C. Eleotério, A. J. Ipólito, M. F. Oliveira, Use of screen-printed electrodes for quantification of cocaine and Δ9-THC: adaptions to portable systems for forensic purposes, Journal of Solid State Electrochemistry, Vol. 20, Issue 9, 2016, pp. 2435–2443.

[3]. I. Novak, M. Mlakar, S. Komorsky-Lovric, Voltammetry of Immobilized particles of cannabinoids, Electroanalysis, 25, 12, 2014, pp. 2631-2636.

Page 64: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

64

Novel Screen-printed Electrode Modified with Lead Film for Highly

Sensitive and Selective Adsorptive Stripping Voltammetric Determination of Cobalt and Nickel in the Form of Dioximate Complexes

A. Bobrowski, A. Królicka, M. Maczuga, J. Zarębski

AGH University of Science and Technology, Faculty of Materials Science and Ceramics, Department of Building Materials Technology, Al. Mickiewicza 30, 30-059 Kraków, Poland

Tel.: + 48 12 617 24 51, fax: + 48 12 617 38 99 E-mail: [email protected]

Summary: For the first time, a carbon screen-printed electrode modified in-situ with lead film (PbF-SPCE) was used for the adsorptive stripping voltammetric determination of Co(II) and Ni(II) in the form of DMG complexes, or Co(II) in the form of a complex with nioxime. Lead film was electrochemically generated in-situ on a commercially available three-electrode screen-printed electrochemical strip from a 0.2 M ammonia buffer solution (pH 8.7) containing 5·10-5 M Pb(NO3)2 and DMG or nioxime. The LOD of the elaborated AdSV procedure with nioxime is equal to 0.003 µgL-1 Co (for tacc = 120 s), which is significantly lower than the LOD values of the vast majority of anodic stripping procedures for other elements (e.g. Pb, Cd, Zn, Cu) at Bi, Sb and Sn films electrodes. Moreover, the separation of Ni(II) and Co(II) peaks in the presence of nioxime was even better at the PbF-SPCE than at the hanging mercury drop electrode. Keywords: Cobalt, Nickel, Lead film, Screen-printed electrode.

1. Introduction

Screen-printed electrodes (SPEs) are voltammetric

sensors that are expected to be widely used in environmental analysis. Thanks to mass production and relatively low cost, SPEs can be disposed of after a single use. Moreover, integrated screen-printed electrodes comprising 3-electrodes on one strip involve smaller amounts of samples. The application of SPEs as sensors in the stripping voltammetry of metal ions usually requires the modification of the carbon surface in SPEs with a thin metal film consisting of mercury, gold, silver, bismuth or other materials, which can improve selectivity and/or sensitivity of the determination [1,2].

In this paper, protocols for the simultaneous adsorptive stripping voltammetric determination of Co and Ni in the presence of dimethyldioxime (DMG) and a selective and extremely sensitive quantification of Co in the form of a complex with nioximate using SPEs plated with lead film (PbF-SPEs) are presented.

2. Experimental 2.1. Apparatus

All measurements were performed using an Autolab potentiostat (EcoChemie, The Netherlands) connected to the PC with GPES 4.9 software. The screen-printed electrodes (4 mm in diameter; carbon DRP C110) were provided by Dropsens, Spain. The working electrode used in all experiments was lead film plated in situ at a screen-printed carbon electrode (SPCE). A coil of platinum wire served as the counter electrode, and an Ag/AgCl electrode (3 M KCl)

(Metrohm, Switzerland) was used as the reference electrode. 2.2. Procedure

The electrochemical investigations were carried out by means of square-wave adsorptive stripping voltammetry (SW-AdSV). The lead film was plated in-situ from the solution containing ammonia buffer, 5·10-5 M Pb(NO3)2 and DMG or nioxime. Before each measurement, the SPE electrode was cleaned electrochemically by applying -1.4 V (20s), +0.5V (1s) and -0.5V (1s) potential steps. Immediately after cleaning, the potential of -0.65 V was applied for 120s to induce lead deposition and simultaneous adsorption of Co(II)-dioxime and/or Ni(II)-dioxime complexes. During all steps, the solution was stirred using a magnetic bar. 3. Results and Discussion

For the simultaneous AdSV determination of Ni(II) and Co(II) as the DMG complexes ex-situ and in-situ procedures of the generation of metal films were tested and the in-situ one was determined to be optimal (Fig. 1). As Fig. 1 presents, the formation of in-situ lead deposits on the SPE surface facilitates the adsorption of both Ni(II) and Co(II) complexes with DMG. Under the optimized conditions the PbF-SPEs provided well-defined peaks of Co(II) and Ni(II) with peak currents that depended linearly on the elements’ concentration in the ranges 5.9-41.3 µg l-1 Co(II) and 0.6-5.9 µg l-1 Ni(II) (tacc = 60 s) (Fig. 2). The DMG complexes adsorb very readily at the PbF-SPE and the

Page 65: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

65

main factor that affects their voltammetric response is accumulation time.

The proposed method was applied for the determination of Ni(II) in snow and drinking water samples and the obtained results were highly consistent with the certified reference values. Co(II) content in the examined snow sample was too low to be detected. However, the Co(II) recovery test performed after spiking the snow sample with a known amount of Co(II) demonstrated the good accuracy of the method.

-0.8 -0.9 -1.0 -1.1 -1.2 -1.3

E vs (Ag/AgCl) / V

a)

b)

c)

NiCo

Fig. 1. SW-AdSV curves of the solutions containing: 5·10-7 M Ni and 5·10-7 M Co, 1·10-5 M DMG, 0.2 M ammonia buffer, pH 8.2 at the bare SPE (a), at the PbF-SPE plated in-situ after an addition of 4·10-5 M Pb(II) to the investigated solution (b), PbF-SPE plated ex situ from 0.1 M acetic buffer (c) containing 4·10-5 M Pb(II). SWV instrumental parameters: Eacc = -0.8, tacc = 60 s.

-0.7 -0.8 -0.9 -1.0 -1.1 -1.2 -1.3

E / V

Ni

Co

Fig. 2. SW-AdSV curves for solutions containing increasing amount of Co(II) and Ni(II) (0.1 to 1 M)

recorded at the PbF-SPE electrode. Supporting electrolyte: 0.2 M ammonia buffer, 4·10-5 M Pb(II), 1·10-5 M DMG.

To elaborate a much more sensitive AdSV procedure of Co(II) detection at the PbF-SPE, DMG was replaced with a nioxime ligand. Surprisingly, the Co(II) signals produced during the reduction of its nioximate complex were approximately two times higher than those obtained when DMG was used. When both II) and Co(Ni(II) were present in the solution containing nioxime, the sensitivity of the Co(II) signal is almost one order of magnitude higher than that of the Ni(II) response (Fig. 3). In the supporting electrolyte containing 0.2 M ammonia buffer, 5·10-5 M nioxime, 5·10-5 M Pb(II), the dependence of peak current vs. Co(II) concentration was found to be linear from 0.03 to 12.4 µg L-1 for 120 s of accumulation at the potential of -0.65 V. The detection limit (LOD) for the accumulation time of 120 s, estimated from three times the standard deviation for the lowest investigated concentration of 0.03 µg L-1 Co(II), was equal to 0.003 µg L-1. The relative standard deviation for 0.3 µg L-1 Co(II) concentration was 5 % for 15 consecutive measurements. As a result of extremely low LOD it was possible to determine Co at the sub g l-1 level (Fig. 4).

-0.6 -0.8 -1.0 -1.2-2

-4

-6

-8

-10

I /

A

E / V

Fig. 3. SW-AdSV curve recorded in solution containing 3.5 g l-1 of Co(II) and Ni(II). Supporting electrolyte: 5·10-5 M nioxime, 5·10-5 M Pb(II), 0.2 M ammonia buffer (pH = 8.7). Edep = -0.65 V, tdep = 60s.

-1.0 -1.2

4

6

8

10

12

I /

A

E / V

-0.5 0.0 0.5 1.00

1

2

3

4

5

6

I p / A

c / g L-1

Fig. 4. The determination of Co(II) in fly ash extract by means of standard additions.

Page 66: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

66

4. Conclusions

The described procedures may help shorten and simplify the protocol for the determination of Ni(II) and Co(II) or Co(II) in the presence of high excess of Ni(II) using lead film electrodes (three-in-one setup, in-situ plating). One of the features that makes this possible is its insensitivity to dissolved oxygen (work in non-deaerated solution). It should be stressed that the LOD of the AdSV method of determination of Co(II) with nioxime is ca. two orders of magnitude lower than the LOD values of the vast majority of stripping procedures for other elements (e.g. Pb, Cd, Zn) at Bi, Sb and Sn films deposited not only on carbon screen-printed supports, but also on glassy carbon ones.

Acknowledgements

Financial support from the Polish National Science Centre (Project 2014/15/B/ST8/03921) is gratefully acknowledged.

References [1]. N. Serrano, A. Alberich, J. M. Dıaz-Cruz, C. Arino, M.

Esteban, Coating Methods. Modifiers and Applications of Bismuth Screen-Printed Electrodes. TrAC – Trendends in Analytical Chemistry Vo. 46, 2013, pp. 15-29.

[2]. X. Niu, M. Lan, H. Zhao, C. Chen, Y. Li, X. Zhu. Review: Electrochemical Stripping Analysis of Trace Heavy Metals Using Screen-Printed Electrodes. Analytical Letters, Vol. 46, Issue 16, 2013, pp. 2479-2502.

Page 67: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

67

How the Derivative Curve of Metal Oxide Sensor Response Gives Access

to Features which Improve the Accuracy of Odour Quantification

M. D. Ahmadou, E. Losson, M. Siadat, M. Lumbreras University of Lorraine, LCOMS, 7 rue Marconi, 57070 Metz, France

Tel.: + (33)387547097 E-mail: [email protected]

Summary: A lot of studies on the utilization of metal oxide gas sensors have been undertaken in order to give a good estimation of gas concentration. Some of them are focused on signal processing techniques applied to the sensor response in order to find robust features. When a good accuracy is needed in gas concentration measurements, one of the main problems is the lack of reproducibility of the sensor responses. In this work, we propose a new approach to compensate short term drift or instability of the sensor responses in order to improve the discrimination of different odour dilutions. For our application of continuous monitoring of different levels of odours, instabilities which appear on the sensor responses can be reduce by working on the derivative of them. So, from this derivative, new features can be defined to improve the accuracy of the odour estimation or gas concentration measurements. Keywords: Metal oxide gas sensor, odour quantification, Pine essential oil vapour, Feature extraction, Short term drift.

1. Introduction

Metal oxide gas (MOX) sensors are often proposed

to identify or quantify the gaseous mixtures because of their high sensitivity and reliability. But for a real application the choice of the operating conditions as well as the selection of features representing these sensor responses, must be considered carefully. For example, many baseline methods can be applied to define features which compensate short term drift [1]. To reduce instabilities of sensor response, different kind of normalization techniques have been proposed [2].

This work shows how the derivative of MOX sensor conductances gives access to features which improve the discrimination of different level of pine essential oil vapours dilutions (from 1% to 4%).

2. Experimental Methodology

A laboratory system has been developed in order to generate various essential oil vapours dilutions and to measure the gas sensor response by using the dynamic headspace technique with two stages: odour or gas exposition and cleaning with pure air [3]. For our application of continuous odour monitoring, a study of the optimized time for the gas exposition and the recovery phase has been undertaken [4]. This study shows that a time of 75 seconds for the gas exposition is necessary to obtain an inflexion point on the sensor responses, and that a time of 350 seconds for the cleaning phase is sufficient to reach a relatively good regeneration of the sensor layer. Numerous measurements were made for all the dilution range (1, 2, 3, 4 %) using different variations: repetitive, increasing (from 1 to 4 %), decreasing (from 4 to 1 %)

and also random. On the whole, we have collected 40 measurements for each pine essential oil dilution. 3. Experimental Results

In this work, we have focused on the responses of one gas sensor (TGS 2620) to highlight the advantage of using the derivative of these responses. This sensor was chosen for its good sensitivity to pine essential oil and for the velocity of his response [3]. Nevertheless this sensor has the same disadvantage than other MOX gas sensor: the slow recovery which introduce short term drift when the sensor is exposed to rapid changes of gas concentration. Fig. 1 shows the sensor response instabilities during the exposition of pine essential oil for different dilutions. The error bars, that are two standard deviation units in length, are helpful to visualise the dispersion of the measured data (40 experiments per dilution) around the mean value. So, by observing these error bars, it is easy to understand that it will be difficult to extract features from these sensor responses for an accurate estimation of odour quantification (especially for dilutions between 3 and 4 %). Obviously, different compensation or normalization methods allow to reduce short term drift that affect these data [1, 2], but no significant improvement has been noticed by processing our signals by these methods. Fig. 2 shows that a simple derivation of our responses gives access to data that allow a better discrimination of the selected dilutions despite of the important dispersion observed. So, it seems to be interesting to use these derivative curves to extract features that are more helpful for an accurate estimation of odour levels.

Page 68: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

68

Fig. 1. TGS2620 sensor responses obtained for different pine essential oil dilutions: 1, 2, 3, 4 %.

Fig. 2. Derivative curves of TGS2620 sensor responses obtained for each dilution.

4. Conclusions It is well known that derivation is an interesting tool for reducing drift, especially when it is caused by a linear trend of data. In our case, data obtained after derivation give access to features with more meaningful information. For instance, we see on the derivative curves (Fig. 2) that not only the peak heights but also the times at which the maxima occurs are related to the pine essential oil dilution. We try now to define new features extracted from the derivative curves in order to improve the odour quantification.

References [1]. S. Di Carlo, M. Falasconi. Drift Correction Methods

for Gas Chemical Sensors in Artificial Olfaction Systems: Techniques and Challenges, Interchopen, Vol. 32, Issue 13, 2011, pp. 1594–1603.

[2]. M. Macías, J. Agudo, A. Manso, C. Orellana, H. Velasco, R. Caballero. Improving Short Term Instability for Quantitative Analyses with Portable Electronic Noses, Sensors, Vol. 14, 2014, pp. 10514‑10526.

[3]. M. Siadat, E. Losson, D. Ahmadou, M. Lumbreras. Detection optimization using a transient feature from a metal oxide gas sensor array, Sensors & Transducers, Vol. 27, Special Issue, May 2014, pp. 340-346.

[4]. D. Ahmadou, E. Losson, M. Siadat, M. Lumbreras. Optimization of an electronic nose for rapid quantitative recognition, in IEEE Proceedings of 2nd

International Conference on Control Decision and Information Technologies, CoDIT’ Metz, France, 3-5 November 2014, pp. 736-741.

Page 69: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

69

Real-time Pressure Measuring System for the Medical Device D. G. Kim 1, S. Seo 1, K. Cha 1, S. Jeong 1, J. K. Choi 1 and T.H. Song 1

1 Daegu-Gyeongbuk Medical Innovation Foundation, Medical Device Development Center, Robot and Instrument Development Team, Cheombor-ro 80, Dong-gu, Daegu, Republic of Korea

Tel.: + 82537905577, fax: + 82537905519 E-mail: [email protected]

Summary: Foliculat Unit Transplantation (FUT) is a type of surgeon that needs skilled techniques. In this study, we proposed an Arduino-based pressure measuring system to train techniques for the hair transplantation. The proposed sensor is composed of simply control board and force sensing resistor (FSR) matrix. This study aims to allow the surgeon to apply to appropriate procedures of the hair transplant. The proposed pressure measuring system enable the surgeon to compare the outcome of the procedure to the hair transplantation. The system provides a result of the hair transplantation using effective visualizations. Moreover, this training simulator for the hair transplantation will help to reduce the appearance of hair transplant and accident scars.

Keywords: Hair transplantation, Pressure measuring system, Force sensitive resistor, Foliculat unit transplantation, Implanter pen. 1. Introduction

Follicular Unit Transplantation (FUT) is worlds ahead of the standard hair transplant procedure performed by many of today`s surgeons [1]. This surgery is complex and delicate procedures designed to restore naturally growing hair to the balding scap. Because of this, hair transplant surgeons are frequently looking for methods to help improve the technique and maximize results. One way hair restoration innovators continually improve the procedure is by inventing new harvesting and implantation tools [2]. The implanter is a pen-like device featuring a hollow needle attached to a tube and plunger apparatus. During a hair transplant procedure with the implanter, the hair restoration physician and/or technicians loads harvested follicular units into the implanter pen. Using forceps, the graft is placed gently into the hollow needle (at the end of the implanter) and handed back to the physician. The surgeon then inserts the needle at an appropriate angle in the scalp (making the incision site) and implants the graft by pressing down on the plunger. This procedure has many advantages such as adequate graft survival rates, and reduced trauma during graft handling, but it is important to remember that physician determines the outcome of the hair transplantation procedure. Even the most advanced tools can create poor results in the hands of a neophyte surgeon [3]. The implanter can produce satisfactory results when used by a trained surgeon who is familiar with the device.

In this paper, we propose a hair transplant training simulator using the implanter pen as shown in Fig. 1. We choose the Arduino board for with automiatical pressure measuring system due to its efficiency in handling real-time data rates and developing the low-cost system.

Fig. 1. Hair transplantation simulator system.

2. System

The purpose of the present study was to guide surgeons who is not familiar with the device appropriate pressure value to implant the graft for producing a satisfactory surgery. The pressuring measuring system is helpful to the surgeon that needs to create good results in the hands. Moreover, using this system, the surgeon who is not familiar with the device can train to be a proficient surgeon. The flowchart of the pressure measuring system is shown in Fig. 2.

The suggested pressure measuring system is implemented on a main controller, embedded with an Arduino Leonardo and two dedicated sensing ICs. For scanning the pressure data, the force sensing resistor (FSR) is used. FSR matrix has 160 force sensing nodes by using 16 columns and 10 rows and is designed for low force values. Combined with control board, this sensor can simultaneously detect force applied to each cell. It has an active area of 50.8 × 76.2 (mm).

Page 70: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

70

Fig. 2. Flowchart of pressure measuring system.

For displaying and recording the scanned pressure data, the application that visualized pressure mapping from the control board was developed [4]. The pressures that pressing down on the plunger were displayed on the application as shown in Fig. 3. If the pressure value is greater or less than the threshold, it is considered as unsatisfying transplantation results.

Fig. 3. FSR matrix and displayed pressure value.

3. Experiments and Results

A surgeon was recruited for this repeated measures prospective test. All study procedures were fully explained to the surgeon, who then had experience with hair implantation procedure.

Simple pressure values were evaluated using a monitoring system that was displayed the force

feedback value. The surgeon sat in a chair and the implanter pen was positioned as shown in Fig. 4.

Fig. 4. The test of hair transplant simulator.

The test was required to complete 10 trials. After trials, the surgeon was greatly satisfied with the performance of the system due to the real-time output on the monitor. According to 10 repetitions experimental results, the surgeon who participated in the experiment, are satisfied with the results. Then, by displaying the pressure change in a real-time graph as shown in Fig. 5, to inform the user, it is possible to improve the pressure required for immediate implanting.

Fig. 5. The graph of pressure change in the system.

Page 71: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

71

4. Conclusions

Simple relation test that evaluated by a surgeon showed good results.

In the future study, several quantitative assessment and visual scoring system will be used by several clinical experts to evaluate the proposed pressure measuring system.

Acknowledgements

This work is funded by the Ministry of Trade, Industry and Energy, Republic of Korea, under the medical device development assistance project with Project Number 10049767

References

[1]. R. M. Bernstein, W. R. Rassman, W. Szaniawski, and A. Halperin, Follicular transplantation, International Journal of Aesthetic Restorative Surgery, Vol. 3, 1995, pp. 119-132.

[2]. R. C. Shiell, A review of Modern Surgical Hiar Restoration Techniques, Journal of Cutaneous and Aesthetic Surgery, Vol. 1, Issue 1, 2008, pp. 12-16.

[3]. D. Stough, J. M. Whitworth, Methodology of follicular unit hair transplantation, Dermatologic Clinics, Vol. 17, Issue 2, 1999, pp. 297-306.

[4]. M. Romero, R. Figueroa, and C. Madden, Pressure sensing systems for medical devises. Medical Device and Diagnostic Industry, Vol. 22, Issue 10, 2000, pp. 124-124.

Page 72: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

72

Calibration Techniques for Skew Redundant Inertial Measurement Units

M. V. Gheorghe Ideal Aerosmith, Inc., 2205 W Lone Cactus Dr., Ste. 7, Phoenix, AZ, USA

Tel.: + 1.623.847.8933, Fax: + 1.623.587.5863 E-mail: [email protected]

Summary: Redundant inertial measurement units have been traditionally used in applications requiring high-reliability inertial space navigation and, more recently, have been shown to improve the performance of micro electromechanical systems-based inertial measurement units. Various redundancy schemes for inertial measurement units have been documented. Such schemes range from doubling the number of sensors arranged in classical 3 orthogonal axes configurations to positioning them in skewed axes configurations. In the latter case, the sensors are usually arranged based on cones and platonic solids configurations. Without loss of generality, this paper presents two calibration and correction techniques for dodecahedron-based skew redundant inertial measurement units. The advantages and disadvantages of the two techniques are analyzed and suggestions made for their possible application areas. The paper also analyzes the pros and cons of two different body frames, one that is better from a packaging standpoint and another one that is better from a redundancy standpoint, and it lays the groundwork for a future study on the usage of virtual body frames in the optimization of dodecahedron-based skew redundant inertial measurement units. Keywords: Accelerometer, Body frame, Gyroscope, IMU, Sensor frame, Skew redundant, SRIMU.

1. Introduction

This paper presents two calibration and correction techniques for skew redundant inertial measurement units (SRIMU). The techniques are presented for the dodecahedron configuration and body reference frame shown in Fig. 1, however they can also be applied to other redundant configurations and body frames.

Fig. 1. Dodecahedron configuration with standard body frame.

These techniques are discussed with their pros and cons and can be employed to determine calibration coefficients that can be used to perform corrections on raw sensor readings in real-life applications. Readings

corrected with these methods can be used directly, or fed as the measured inputs into Kalman filters.

The error models used herein are identical for accelerometers and gyroscopes, therefore the equations used in this paper refer to accelerations without loss of generality. 2. Prior Work

In [1] the author presents a method of calibrating

MEMS accelerometers in the body frame without using known stimuli. That method would require further work in order to be extended to gyroscopes and to address skew redundant configurations.

In [2] the author presents a method of aligning two different sensor frames, such as the accelerometer and magnetometer frames, using the relation between the earth’s gravitational and magnetic fields. While a similar relation exists between the earth’s gravitational and rotation fields it is doubtful that this relation can be applied to the alignment of the accelerometer and gyroscope sensor frames due to the fact that the rotational field is very weak (approximately 15°/hr.)

In [3] the authors present a calibration method for SRIMUs in dual triad and cone configurations. The presented error model isolates two orthogonal sub-frames. This approach increases the number of variables that need to be solved for.

In [4] the authors describe a calibration approach for a dodecahedron SRIMU. The error model uses a body reference frame like the one shown in Fig. 2. The body frame axes X2Y2Z2 bisect the angles of the ideal sensor pairs 2, 4, 1, 6 and 3, 5, respectively. The presented error model is appealing through its symmetry however it makes the assumption that the

Page 73: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

73

sensor misalignments are contained in the trihedral planes determined by the body reference frame system. Since, in real life, sensor misalignments are not confined to a 2-dimensional space this simplification renders the error model incomplete and can lead to substantial calibration errors due to unmodeled misalignment components.

Fig. 2. Dodecahedron configuration with alternate body frame.

In [5] the authors present a factorization-based

calibration approach that uses partial information about the motions and applies gravitational field constraints. Gyroscope stimuli are estimated using a camera affixed to the rate table and which observes objects placed at distance. This approach does not seem suitable for navigation-grade SRIMUs because they are expected to be enclosed in a thermal chamber with opaque walls during the calibration process. Additionally, it is not known whether rotation accuracies obtained using the camera approach are comparable with ones obtained through traditional methods, such as optical encoders which, for such applications, are commonly better than ±10arcsec/sec. 3. Body Frame Calibration 3.1. Body Frame Error Model

The body frame calibration method presented herein relies on the following error model:

(1)

Equation (1) can be expressed in a more compact form using extended matrices, as shown below:

(2) where ASX (3), is the 7xN extended matrix of uncorrected sensor readings; MBSX (4), is the 7x4 extended shape matrix combining the sensor scale factors, misalignments and biases; ABX (5), is the 7xN extended matrix of stimuli applied in the body reference frame; and N is the total number of samples and will be discussed later.

(3)

(4)

(5)

The number of sensors, in this case 6, determine the

number of rows in AS, BS and MBS, which are submatrices of ASX and MBSX. This observation can be used to generalize the error model for other redundant IMU configurations.

Another observation about this error model is that it does not make a distinction between cross-axis sensitivities, which are dictated by the sensors, and misalignments, which are dictated by mechanical assembly errors. Therefore, MBS combines both. However, this is not a limitation because the difference between cross-axis sensitivities and misalignments cannot be observed at the body frame level and the end goal is to calibrate the IMU as an assembly. The combined cross-axis sensitivities and misalignments are referred to simply as misalignments henceforth.

An additional characteristic about this error model is that each of the misalignments in MBS is aggregate value from multiple sensors. This may be a limitation in some SRIMU applications and will be discussed later.

Finally, N, the total number of samples will be expressed as follows:

(6) where P is the number of independent stimuli required to perform the calibration (i.e. accelerations for accelerometers and rates for gyroscopes); and Si, i = 1…P are the number of samples acquired for each one of the independent stimuli. The values for P and Si are discussed in the next sections.

Page 74: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

74

3.2. Body Frame Calibration Profile

The extended shape matrix can be computed from the known input stimuli and uncalibrated sensor readings using this equation:

(7)

In order for (7) to have a solution, both of the

following conditions must be met: P ≥ 24 and Si ≥ 1. The first condition is dictated by the number of

variables that need to be solved for (mx1 through mz6 and b1 through b6) while the second condition is related to sensor noise and it will be discussed in section 5.

In practice, (7) becomes overdetermined by making P > 24 and Si >> 1. Due to the least squares fitting properties, this approach typically yields more accurate calibration coefficients.

Body frame calibration profiles can be performed on 2-axis position and rate tables (see Fig. 3).

Fig. 3. Model 2002PG 2-axis rate and position table with thermal chamber.

Two notional body frame calibration profiles for the accelerometers and gyroscopes are shown in Table 1 and Table 2, respectively. 3.3. Body Frame Corrections

Once MBSX is computed, it can be used to perform body frame corrections based on raw sensor readings:

(8) where ÂBX contains body frame corrected readings and has the same layout as ABX (5).

As mentioned in section 3.1, the calibration coefficients contained in MBSX aggregate values from multiple sensors which may prevent the computation of parity vectors used in fault detection, isolation and

recovery (FDIR). This inference is based on the observation that parity vectors are obtained by combining in some fashion the corrected readings from multiple sensors, combination that may be hampered by the fact that individual sensor reading corrections are not available. However, no attempt is made in this paper to provide an analysis or formal proof.

Table 1. Notional body frame calibration profile for accelerometers.

No. Inner Axis

Position Outer Axis

Position Notes

1 0° 0° Position outer axis in 45° increments

from initial position

2 0° 45° 3 0° 90°

… 8 0° 315° 9 90° 0°

Position outer axis in 45° increments

from initial position

10 90° 45° 11 90° 90°

… 16 90° 315° 17 0° 0°

Position inner axis in 45° increments

from initial position

18 45° 0° 19 90° 0°

… 24 315° 0°

Table 2. Notional body frame calibration profile for gyroscopes.

No Inner Axis Pos.

Outer Axis Pos.

Inner Axis Rate

Outer Axis Rate

Notes

1 0° N/A 0°/s 0°/s Increment outer axis

rate by ±10°/s

from initial rate

2 0° N/A 0°/s 10°/s 3 0° N/A 0°/s -10°/s

… 8 0° N/A 0°/s 40°/s 9 0° N/A 0°/s -40°/s

10 90° N/A 0°/s 0°/s Increment outer axis

rate by ±10°/s

from initial rate

11 90° N/A 0°/s 10°/s 12 90° N/A 0°/s -10°/s

… 17 90° N/A 0°/s 40°/s 18 90° N/A 0°/s -40°/s 19 N/A 90° 0°/s 0°/s Increment

inner axis rate by ±10°/s

from initial rate

20 N/A 90° 10°/s 0°/s 21 N/A 90° -10°/s 0°/s

… 26 N/A 90° 40°/s 0°/s 27 N/A 90° -40°/s 0°/s

If the previous observation is proven to be true, this calibration scheme might be more useful for improving the IMU performance rather than providing redundancy, as it may be the case for MEMS-based IMUs.

Page 75: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

75

4. Sensor Frame Calibration 4.1. Sensor Frame Error Model

The sensor frame calibration presented herein relies on the following error model:

(9) Equation (9) can be expressed in a more compact form as shown below:

(10) where ASX (3), is the 7xN extended matrix of uncorrected sensor readings; MNSX (11), is the 7x7 extended shape matrix combining the sensor scale factors, misalignments and biases; and ÃSX (12), is the 7xN extended matrix of ideal sensor readings; and N is the total number of samples.

(11)

(12)

The number of sensors determine the number of

rows and columns in MNS and implicitly in MNSX and, respectively, the number of rows in ÃS and implicitly in ÃSX.

A first observation about this error model is that, like the body frame error model, it does not make a distinction between cross-axis sensitivities and misalignments. Thus, MNS combines both and its values will be referred to as misalignments hereafter. As mentioned before, this is not a limitation because the difference between cross-axis sensitivities and misalignments cannot be observed at the body frame level and the end goal is to calibrate the IMU as an assembly.

A second observation is that, unlike the body frame error model, MNS and BS contain entities on a per sensor basis, the advantage of which will be discussed later.

The third observation is that, like with the body frame calibration, N, the total number of samples, can be expressed as per (6). The values of P and Si for the sensor frame calibration will be discussed later.

4.2. Sensor Frame Calibration Profile

The extended shape matrix can be found from the known input stimuli and uncalibrated sensor readings using this equation:

(13)

In order for (13) to have a solution, the following

conditions must be met: P ≥ 42 and Si ≥ 1. The first condition is dictated by the number of

variables that need to be solved for while the second condition is related to sensor noise and it will be discussed in section 5.

In practice, (13) becomes overdetermined by making P > 42 and Si >> 1.

Dodecahedron-based SRIMU sensor frame calibrations can still be performed, in principle, on 2-axis rate and position tables however stimuli are provided at the sensor frame level rather than the body frame level.

Two notional sensor frame calibration profiles for the accelerometers and gyroscopes are shown for illustration in Table 3 and Table 4, respectively.

Table 3. Notional sensor frame calibration profile for accelerometers.

No. Inner Axis

Position Outer Axis

Position Notes

1 0° -37.28° *

Position outer axis in 45° increments from initial

angles marked with

asterisk

… 8 0° 277.62° 9 60° -79.19° *

… 16 60° 235.81° 17 120° -37.38° *

… 24 120° 277.62° 25 180° -79.19° *

… 32 180° 235.81° 33 240° -37.38° *

… 40 240° 232.62° 41 300° -79.19° *

… 48 300° 235.81°

Notwithstanding the fact that sensor frame calibrations can be performed in principle on 2-axis tables, there is a rate multiplication effect that can be seen in Table 4.

The 2-axis table’s rate capabilities may be exceeded in some cases due to this effect. In such instances one option is to reduce the rates experienced by the individual gyroscopes. The other option is to perform the calibration on a 3-axis rate table instead.

With 3-axis rate tables the inner and middle axes can be used to position the gyroscopes and the outer axis can be used to spin them. However, these tables (see Fig. 4) are substantially more expensive than the 2-axis ones. Therefore, the tradeoffs have to be carefully analyzed.

Page 76: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

76

4.3. Sensor Frame Corrections

Once MNSX is computed, it can be used to perform sensor frame corrections based on raw sensor readings:

(14)

Table 4. Notional sensor frame calibration profile for gyroscopes.

No. Inner Axis

Position

Middle Axis

Position

Outer Axis Rate

Notes

1 0° 52.62° 0°/s Alternate outer axis

rates and step in ±10°/s

increments

2 0° 52.62° 10°/s 3 0° 52.62° -10°/s

… 0° 52.62° 40°/s 0° 52.62° -40°/s 60° 10.81° 0°/s

Alternate outer axis

rates and step in ±10°/s

increments

60° 10.81° 10°/s 60° 10.81° -10°/s

… 60° 10.81° 40°/s 60° 10.81° -40°/s 120° 52.62° 0°/s

Alternate outer axis

rates and step in ±10°/s

increments

120° 52.62° 10°/s 120° 52.62° -10°/s

… 120° 52.62° 40°/s 120° 52.62° -40°/s 180° 10.81° 0°/s

Alternate outer axis

rates and step in ±10°/s

increments

180° 10.81° 10°/s 180° 10.81° -10°/s

… 180° 10.81° 40°/s 180° 10.81° -40°/s 240° 52.62° 0°/s

Alternate outer axis

rates and step in ±10°/s

increments

240° 52.62° 10°/s 240° 52.62° -10°/s

… 240° 52.62° 40°/s 240° 52.62° -40°/s 300° 10.81° 0°/s

Alternate outer axis

rates and step in ±10°/s

increments

300° 10.81° 10°/s 300° 10.81° -10°/s

… 300° 10.81° 40°/s 300° 10.81° -40°/s

Fig. 4. Model 2003HP 3-axis rate and position table with thermal chamber.

The corrected sensor frame readings can be transformed to body frame readings using the following equation:

(15)

where ÂB is the matrix of corrected readings in the body frame; MBS (16) is the shape matrix that defines the body frame to ideal sensor frame transformation for the configuration shown in Fig. 1; and ÂS, the subset of ÂSX (17), containing corrected sensor readings.

(16)

(17)

As mentioned before, the resulting calibration coefficients are on a per sensor basis which allows the computation of parity vectors and supports FDIR. Therefore, this method is appropriate for navigation applications where redundancy is important.

5. Noise Considerations

Sensor noise was not explicitly shown in either one of the error models however, when noise has a Gaussian distribution, it is addressed implicitly through the least squares fitting.

In practical terms, this is achieved by making Si >> 1 in (6).

Usually, Si have upper bounds determined with the help of Allan variance studies or simply limited by cost factors such as calibration time. Additionally, in the case of the gyroscopes, the numbers of samples may be adjusted such that they correspond to complete revolutions.

6. Practical Considerations

It was mentioned earlier that the error models are the same for accelerometers and gyroscopes. While this leads to a number of similarities in the way accelerometers and gyroscopes are calibrated, there are also a number of important differences which will be examined in this section.

First of all, accelerometers are usually excited using positioning versus the gravitational field. This way, accelerations of ±1G can be achieved. Depending on the application, G can be considered to be 9.80665m/s2 or its value can be computed from a geodetic model, based on latitude and elevation.

Another consideration for accelerometers is vibration isolation. Whether transmitted through the

Page 77: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

77

floor or caused by the acoustic noise inside the thermal chamber, special measures must be taken at times to prevent such vibrations from reaching the accelerometers.

By contrast, gyroscopes are excited by rotating them at prescribed rates instead of using the earth’s rotation field which is too weak for calibration purposes.

However, provided the gyroscopes can measure it, the earth’s rotation field could be used to verify the calibration accuracy. This can be achieved with 2-axis table that has its outer axis oriented east-west or, with no limitations, with a 3-axis table. In both cases the table is used to position each gyroscope in parallel and antiparallel orientations with regards to the earth’s rotation vector.

7. Future Work 7.1. Temperature Modeling Considerations

The sensors used in SRIMUs will likely exhibit a temperature dependency of their calibration constants. However, their temperature dependency is not captured in the error models presented herein.

A possible approach of decoupling the temperature dependency from the error model is suggested in [1] and was included below for the sake of completeness: Calibrate the SRIMU at M ≥ 4 temperatures; Derive the temperature dependency of each

calibration coefficient using a 3rd order polynomial;

If the model seems suitable, verify the performance of the correction scheme by testing at 3 or more temperatures different from the ones used for calibration in order to estimate the residual errors due to thermal modeling.

Based on the observed temperature dependency characteristics, it may be more beneficial to model it with linear interpolation or higher order polynomials.

Tradeoffs studies such as number of calibration temperature points, residual errors introduced by the selected interpolation model and propensity of higher order polynomials for sudden swings must be considered in determining the most appropriate model. 7.2. Enhanced SRIMU

The presented calibration methods can also be applied to alternate body frames such as the one presented in [4] and shown in Fig. 2.

In this arrangement, the body frame axes X2, Y2, and Z2 bisect the angles between the sensor pairs 1, 6, 2, 4, and 3, 5, respectively. This leads to the following body frame to ideal sensor frame shape matrix:

(18)

In this case, the contribution of the sensor axes 1 to

6 to the body frame axes X2, Y2, and Z2 is distributed more uniformly than in (16).

Also, [4] proposes parity vectors that impose a minimal computational burden, a definite advantage for microcontroller-based implementations.

However, the X2Y2Z2 body frame does not seem to be leading to an efficient packaging. This could be a considerable disadvantage in navigation applications where increases in volume and mass have negative consequences.

These characteristics are in contrast to those of the body frame shown in Fig. 1 which facilitates an efficient packaging and has a less than ideal shape matrix.

Considering these observations, future work will attempt to combine the advantages of the two body frames, while avoiding their disadvantages, by designating one of them as a virtual frame.

This is illustrated in Fig. 5 which shows a model of an SRIMU and two sets of body frames: the real one XYZ and the virtual one X2Y2Z2.

Fig. 5. Enhanced SRIMU real and virtual body frames. Acknowledgements

Many thanks to Ideal Aerosmith Inc., for facilitating the studies that led to the development of this paper, Mr. Jim Richtsmeier for reviewing the abstract and full-length papers, Mr. Dustan Larson for reviewing early work that let to this paper and to Ms. Jennifer Storm and Mr. Bryon Bischoff for preparing

Page 78: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

78

and consulting on the SRIMU 3D CAD model, respectively. References [1]. M. V. Gheorghe, Advanced Calibration Method for

3-Axis MEMS Accelerometers, unpublished. [2]. M. V. Gheorghe, Advanced Sensor Frame Calibration

for Tilt-Compensated Electronic Compasses, unpub-lished.

[3]. S. Y. Cho, C. G. Park, Calibration of a Redundant IMU, AIAA Guidance, Navigation, and Control Conference and Exhibit, 16-19 August 2004, Provi-dence, Rhode Island.

[4]. S. Sun, R. Liu, Error Calibration and FDI Technology of Gyros in Redundant IMU, in Proceedings of the First International Workshop on Database Technology and Applications, Wuhan, April 2009, pp. 398-401.

[5]. M. Hwangbo, J. S. Kim, T. Kanade, IMU Self-Calibration Using Factorization, IEEE Transactions on Robotics, Vol. 29, No. 2, April 2013, pp. 493-507.

Page 79: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

79

Measuring Machine for Mechanical Properties of Garments

Lukas Pfarr 1 and Bernhard Zagar 1

1 Johannes Kepler University, Institute for Measurement Technology, 4020 Linz, Austria Tel.: +43 732 2468 5921, Fax +43 732 2468 5933

E-mail: [email protected] Summary: Advanced software tools allow simulating the virtual fitting of clothes which is really close to reality. This ability comprises a broad number of advantages for both designers and industry. First and foremost the design process can be truly optimized in various aspects (speed-up of the design due to digital prototyping coupled with instant feedback, a priori information regarding the stress distribution within the garment, fitting for all kinds of body shapes). The virtual description of a garment takes the industry to a new level of flexibility.

The most important input data for the simulation are the mechanical and tribological properties of the garment. The key values can be named as: tensile, shear, bending and friction. Experimental values for the mechanical fabric properties can be derived from standard fabric characterization experiments. These methods have, however, not been designed for use in simulations [3]. The goal of the proposed paper is to develop a universal testing machine to precisely quantify all relevant garment parameters needed for cloth simulation. Keywords: garment properties, universal testing machine, fabric assessment, cloth simulation, physical behaviour

1. Introduction

As use of technology is rising in the clothing

industry, simulation of garments is one of the most powerful tools for the design process of the future. Software of companies like Lectra [1] or MIRALab [2] helps to shorten the time-lapse between an idea and the finished product enormously by creating a virtual description of the garment. Furthermore it provides a mould-breaking level of flexibility through digital prototyping coupled with instant feedback. As Fig. 1 shows, the designers even get information on the real-time stress distribution within the material, which will become especially beneficial for the sports clothing industry.

Fig. 1. Numerical fitting data while running [3].

1.1. Fundamentals for a Successful Simulation The accuracy of virtual garment simulations is on the one hand dependent on precise computational models and on the other hand on exact input parameters [4].

These input parameters are characterized by the mechanical and tribological properties of the textile. Tensile, shear, bending, friction and mass density are considered to be the crucial values for a correct description of the fabric behaviour [3].

1.2. Common Methods for Measuring Fabric Parameters

Today's standard measurement systems can be

named as KES-f [5] and FAST [6]. They were initially developed to replace outdated subjective fabric evaluation by an objective method with the main goal to reduce quality fluctuations of textiles.

Garment description itself is a highly complex topic due to its anisotropic and non-linear behaviour. The FAST system is limited to only one measured load and therefore less suited for the derivation of tensile and shear parameters [4]. Also derived parameters from KES-f are only accurate for one specified load (500 N/m) [4]. Existing characterization methods, which were not designed for the purpose of deriving parameters for virtual simulations, need to be rethought [4]. Scientific research shows that there is a need for a more sophisticated, loading machine, which meets the demands of the modern industry.

1.3. Improved Concept The concept of the proposed measuring machine is

shown in Fig. 2. It offers the possibility to perform multiple load/unload experiments with different applied forces and can also measure all other necessary properties in a reasonable amount of time. Its low cost construction in combination with good usability predestines it for a broad use in the industry.

Page 80: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

80

Fabric properties, especially tensile strength, vary significant with humidity and temperature. Therefore the proposed concept includes a precondition chamber to simulate the environment properly. This aspect ensures the repeatability of experiments as well as the validity of the measurements.

Fig. 2. Loading machine measuring stiffness, friction

coefficient and stress-strain in a single unit. 2. Determination of Parameters 2.1. Tensile and Shear Test

The goal is the ability to investigate all kinds of garments precisely. Elastic textiles on the one hand and tough textiles on the other hand, show highly different behaviour in a tensile test. To measure precisely over a wide range of forces, an adjustable measuring amplifier combined with a standard load cell are used. The force which clamps the textile, can be adopted proportional to the acting tensile force. This principle minimizes the damage of the probe in the clamp region and avoids measuring errors.

For the shear test, the relation of shear to the extension-compression in the bias-direction of a woven fabric is used [7]. A probe with 45° twisted warp- and weft-direction is tested like in a tensile test. This way a complete stress-strain profile can be measured accurately, while lowering the complexity of the systems kinematic dramatic.

2.2. Bending Test

The bending stiffness is an essential parameter for the cloth simulation, especially to determine the arrangement of the garment's folds. It is hard to capture accurately, as the span between the stiffest and the most flexible one is about 1:2940 [3].

The proposed machine implements the principle of [8] identical, where the textile is under the load of its own weight seen as cantilever. A camera captures an image of a laser line projected onto the probe. It's

possible to determine the 3D deflection line (shown in Fig. 3) via spatial triangulation out of which via a fourth-order differential equation the bending stiffness can be calculated from. 2.3. Coefficient of Friction

The parameter is measured between the friction partners garment – garment as well as garment – wet skin. The latter can be approximated with leather.

Fig. 3. Determination of the deflection curve [8].

No additional components are needed here, because the actuator of the tensile test is used to incline the plane of operation and the camera from the bending test detects the position of the experimental mass. An additional mode for dynamic lowering of the plane imitates a real-life situation and should deliver new sets of information on that point.

3. Conclusions

The proposed concept is developed to take the quality of garment measurements to a new level and to satisfy all needs of the future textile industry. The loading machine is seen as the missing link in a comprehensive process chain and will help to design the intelligent clothes of tomorrow.

Acknowledgments

This research was funded by the Austrian Research Promotion Agency (FFG) project 848215.

References [1]. Lectra Software, http://www.lectra.com [2]. MIRALab Software, http://www.miralab.ch

Page 81: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

81

[3]. Luible C., Study of mechanical properties in the simulation of 3D garments, Thèse de doctorat, Univ. Genève, 2008, no. SES 678.

[4]. Luible C., Suitability of standard fabric characteri-zation experiments for the use in virtual simulations, https://www.researchgate.net/publication/252543001_SUITABILITY_OF_STANDARD_FABRIC_CHARACTERISATION_EXPERIMENTS_FOR_THE_USE_IN_VIRTUAL_SIMULATIONS

[5]. Kawabata S., The standardization and analysis of hand evaluation (2nd Edition), The hand evaluation and

standardization committee, The Textile Machinery Society of Japan, Osaka, 1980.

[6]. De Boos A. and Tester D., FAST - Fabric Assurance by Simple Testing, Textile and Fibre Technology, Report No WT92, ISBN 0 643 06025 1, 1994.

[7]. Behre B., Dahlberg B. and Lindberg J., Mechanical Properties of Textile Fabrics, Part I – III, Textile Research Journal, Sage, Vol. 31, No. 2, 1961.

[8]. Indra C., Determination of Textile Parameters with Digital Image Processing, JKU Linz, 2016.

Page 82: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

82

Stress Induced Birefringence Spectroscopy

S. Schallmeiner 1, F. W. Dietachmayr 1 and B. G. Zagar 1

1 Johannes Kepler University, Institute for Measurement Technology, Altenbergerstr. 69, 4040 Linz, Austria Tel.: +43 732 2468 5920, fax: +43 732 2468 5933

E-mail: [email protected] Summary: The Fourier-transform infrared spectrometer is the most commonly used spectrometer. Its disadvantages are its high cost and stringent mechanical requirements. This proposed paper presents a novel approach on spectroscopy using stress induced birefringence. The presented setup requires only a planar light source, a loaded cantilever placed between crossed polarizers, and a camera. The spectrum of the transmitted light is calculated by using the Fourier-transform of the emerging fringe pattern due to the mechanical loading. Keywords: Spectroscopy, Stress induced birefringence, Photoelastic effect. 1. Introduction

Infrared (IR) spectroscopy is widely used in industrial and scientific contexts (e.g., in atmospheric sciences [1], in food industries for quality control [2], and in biomedicine [3]). The most commonly used IR-spectrometer is the Fourier-transform infrared (FTIR) spectrometer. Major advantages of FTIR spectrometers are higher luminous efficiency and better wavenumber precision than dispersive spectro-meters [4]. Conversely, disadvantages are their high cost and stringent mechanical requirements. This paper presents a novel approach on spectrometry based on stress induced birefringence (SIB) utilizing a mechanically loaded cantilever. In this preliminary work a setup for an SIB spectrometer for the visible spectrum is presented and analyzed. 2. Stress Induced Birefringence

If a mechanical stress is applied to specific, trans-parent materials, those materials become birefringent. Subsequently, the mechanical stress is assumed to in-duce a planar state of stress inside the material with principle stress directions 1 and 2. Linearly polarized light (in orthonormal directions 1 and 2) propagates with respective propagation velocities v1 and v2 given by

1 0 1 1 2 2

2 0 1 2 2 1

v = v +c σ +c σ

v = v +c σ +c σ, (1)

where v0 is the speed of light inside the unloaded probe, σ1 and σ2 are the principle stresses, and c1 and c2 are material constants [5].

The electromagnetic wave components of linearly polarized light passing through the medium under SIB can be evaluated along the two principal stress directions. The different velocities of those compo-nents cause a relative phase retardation δλ between the components:

1 2λ

Cbδ = σ σ

λ , (2)

where C is the photo-elastic constant, λ is the wave-length of light, and b is the propagation path in the medium.

3. Euler Beam Theory

To utilize SIB, a setup allowing a well-defined mechanical loading is required. In case of a cantilever the mechanical stress distribution is described by Euler beam theory. The bending stress σb at a point P of the cantilever with rectangular cross section is given by

3

12

yb z

y

y

y

Mσ = a

I

M = Fd

bhI =

, (3)

where F is the force loading the cantilever, d is the length of the lever, az is the distance from the neutral axis to P, b is the width of the cantilever, and h is the height of the cantilever (Fig. 1).

Fig. 1. Schematic of the cantilever.

4. Measurement Setup

The measurement setup consists of a planar light source (Motorola G2 mobile phone display), a polarizer and an analyzer in crossed configuration, a

Page 83: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

83

cantilever (polycarbonate) and a camera (AVT Prosilica GE1050C) as outlined in Fig. 2. The light is polarized by P1, passes through the cantilever where due to an applied force F the SIB causes a relative phase retardation between the ordinary and extraordi-nary wave components along the two principal stress directions. The relationship between the force F and the retardation δλ is given by Eq. (2) and Eq. (3) where σ1 equals σb and σ2 equals 0. Afterwards, the light passes through the analyzer P2 and is recorded by the camera.

Fig. 2. Experimental setup, planar light source, polarizer P1, cantilever, analyzer P2 and camera.

The light intensity I for monochromatic light at the camera’s face plate is given by

2 20sin 2 sin λI = I πδ , (4)

where I0 is the intensity of the emitted light, and φ is the angle between the transmission axis of the polarizer and the principle stress direction σ1 [6]. For a polychromatic line spectrum and an angle φ of 45° the light intensity after the analyzer can be expressed as

2sin ( )

(1 cos(2 ))2

I = I

I=

, (5)

where Iλ are the intensities at the wavelengths λ. By measuring the intensity for different δλ and using the Fourier-transform of Eq. (5) the spectrum of the transmitted light (i.e., the coefficients Iλ) can be calculated.

5. Results

Fig. 3a shows the measured pattern on the cantilever exposed to violet light. According to Eq. (3) the retardation increases linearly along line L1. Therefore, the spectrum of the transmitted light can be calculated by evaluating the Fourier-transform of the intensity along the line. The result is shown in Fig. 3b. As expected from inspection of the image the spectrum consists of blue and red components.

(a)

(b)

Fig. 3. (a) Measured pattern on the cantilever exposed to violet light; (b) calculated spectrum.

6. Conclusions

Using SIB and a cantilever we showed that the

spectrum of transmitted light can be obtained. The advantages of this method are its simple mechanical implementation and its low cost. Currently the setup can only be used for analyzing the visual region of the spectrum due to the high strain needed for better resolution in the infrared region. Further research will include analyzing other cantilever materials with higher photo elastic constants and higher tensile strength able to analyze also in the IR part of the spectrum. Acknowledgement

The authors gratefully acknowledge the partial financial support by the Austrian Research Promotion Agency and the Austrian COMET program.

Page 84: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

84

References [1]. B. J. Finlayson-Pitts, J. N. Pitts, Chemistry of the upper

and lower atmosphere, Academic Press, 2000. [2]. H. Huang, H. Yu, H. Xu, Y. Ying, Near infrared

spectroscopy for on/in-line monitoring of quality in foods and beverages: A review, Journal of Food Engineering, Vol. 87, Issue 3, 2008, pp. 303-313.

[3]. T. Jue, K. Masuda, Application of Near Infrared Spectroscopy in Biomedicine, Springer, 2013.

[4]. P. R. Griffiths, J.A. de Haseth, Fourier Transform Infrared Spectrometry, John Wiley & Sons, Inc., 2007.

[5]. L. Föppl, E. Mönch, Praktische Spannungsoptik, Springer, 1950.

[6]. H. Wolf, Spannungsoptik, Springer, 1961.

Page 85: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

85

Health Checking System Using Wearable Health Devise and PIR Sensors

T. Miyazaki 1, F. Shinohara 1 , T. Horiuchi 1 , Y. Ohira 1 , H. Yamamoto 2 and M. Nishi 2

1 National Institute of Technology, Nagano College, 716 Tokuma, 381-0041 Nagano, Japan 2 Shinshu University, Faculty of Engineering, 4-17-1 Wakasato, 380-8553 Nagano, Japan

3 Shinshu University, Faculty of Education, 6-Ro Nishinagano, 380-8544 Nagano, Japan Tel.: + 81262957065, fax: + 81262954950

E-mail: [email protected] Summary: As a method of specifying a movement of a person, we propose the low-power-consumption and low-cost supervising system using a PIR sensor which can be easily used, compared to a surveillance camera. A PIR sensor is a sensor which reacts to infrared displacement and detects the movement of a person. Movement of a person cannot be specified only by one PIR sensor, but it becomes possible by using more sensors. The purpose of our work is to develop the supervising system for checking the safety of elderly people, especially those who are living alone. This system enables the family members of the elderly to see the movement of the elderly using a PC or smart phone from remote places, and to manage the health condition from a wristband-type sensor that checks heart rate and body temperature. Keywords: PIR sensor, Arduino, Movement measuring, Wearable health monitor, Health monitoring system.

1. Introduction

Recently, due to the effects of separation of families,

decreasing birthrate and aging population, the number of elderly living alone is increasing. These elderly people are now in danger, because if they get hurt or become sick, they are not able to ask for help, and most of the time they are found too late. To prevent this kind of problem, one of the effective solutions is to be able to monitor the situation of the elderly. In this study, multiple sensors are used so that human movement can be monitored. The Arduino will be used as the processor because it is easy to make and use. An elderly safety supervising system can be developed by using the PIR sensors and monitoring the health of the elderly and indoor environment [1].

2. Health Monitoring System 2.1. Outline of the System

The block diagram of the system is shown in

Fig. 1. The PIR sensors can detect the movement of human. Also, in order to monitor the health of the elderly, a heartbeat sensor, which can detect and count heartbeats, and a temperature sensor for measuring the body temperature are a part of the system. In addition, the temperature sensor and humidity sensor are used to measure the temperature and humidity of the room. The output data of these sensors will be sent to and processed by the Arduino via XBee. By using XBee, wireless connection to Arduino can be achieved, so that it is easier to apply the system because the cabling is unnecessary. The communication protocol of XBee is based on the ZigBee standard, which can cooperate with Arduino, and wireless communication can be easily performed with low power consumption [2, 3]. Also, if there is a problem with the elderly’s health, family members will be informed by email.

Fig. 1. Block Diagram of the System.

2.2. PIR Sensor Network

PIR sensors are designed to be installed on the ceiling. Only one PIR sensor cannot specify the detailed human movement, because of the long range of detection. However, this can be achieved by using more than one PIR sensor. In the case of two PIR sensors, tree detecting area can be arranged. Therefore, by including the situation where both sensors have detected movement or only one of them has detected movement, there are three positions that can be specified. The detection range of a PIR sensor and the sensor by which the cover of the aluminum tape was attached is investigated.

Page 86: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

86

2.3. Health Monitor and Indoor Monitor

If there is only PIR sensor, family members cannot identify whether the elderly is stopping or fall down. To make it better, heartbeat sensor and temperature sensor are used to confirm the health status of the elderly. Considering the convenience of wearing, wristband equipped on the arm or leg is to be used as shown Fig. 2. Also, there is an additional function that alerts the elderlies about their abnormal health status by light and sound.

Fig. 2. Health Status Monitor.

The data of room temperature and humidity will be sent to Arduino by XBee, then uploaded to the server from Arduino, so that it can be checked from browser. If indoor environment is in a bad state for elderly people, the directions to adjust room temperature and humidity is urged will be notified by the wristband. The data of the temperature and humidity of the room, and the elderly’s heart rate and body temperature data which were measured by the sensor are sent to a management server.

3. Experiment and Results

To find out the characteristics of the PIR sensor, a real product of PIR, which is #555-28027 (PARALLAX), sensor is used to conduct the experiment. The First experiment is to determine the detection range of a PIR sensor sticking on a wall at different angle and distance. The cover, which is aluminum tape with a circular hole, is attached to a PIR sensor. The data was measured every 15 degrees from the position of 90 degrees of the center. The position is set every 1 m from the sensor to 5 m.

The second experiment confirmed that two PIR sensors can have three detection areas. The PIR sensors with the covers which diameters of holes are 8 mm are stuck on a wall between 1 m. The measurement positions are set 4 m from the wall, then sensitivity was measured. Because the detection angle is limited by 30 degree due to the cover, the diameter of detection area 4 m from the wall is calculated to be about 2 m. PIR sensors are arranged so that 1 m of detection areas are overlapped with another one.

The circle in Fig. 3 shows the detection area of two PIR sensors. As a result, three detection areas are

formed by two PIR sensors as we expected, and it is confirmed that the range of detection area can be controlled accurately.

Fig. 3. Detecting Areas of Two PIR Sensors. 4. Conclusions

We proposed the novel low-power and low-cost supervising system for the elderly’s safety. It is confirmed that the measuring system using PIR sensor and Arduino is very suitable to specify human movement. Also, according to this experiment result, it is known that the sensing range can be limited or changed by using the cover. Acknowledgements

This work was supported by JSPS KAKENHI Grant Number JP 26350356 and JP 16K00260.

References [1]. T. Hata, H. Masaki, T. Kito, T. Ito, A Smart Building

with PIR Sensor Networks: Counting Pedestrians with PIR Sensors, Enriched multimedia technical report 112(190), Institute of Electronics, Information and Communication Engineers, 2012, pp. 13-18.

[2]. M. Banzi, Getting Started with Arduino, 2nd Edition, Maker Media, Inc., 2011.

[3]. R. Faludi, Building Wireless Sensor Networks, O'Reilly Media, 2011.

Page 87: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

87

Power Consumption Considerations of an Agricultural Camera Sensor

with Image Processing Capability

Gábor Paller and Gábor Élő Széchenyi István University, Information Society Research & Education Group,

Egyetem tér 1. Győr, Hungary Tel.: +36 (96) 503-400

E-mail: paller.gabor,[email protected] Summary: This paper presents our experiments with image processing toolkits on microcontrollers through the use case of an agricultural camera sensor capturing images in multiple spectral ranges. Night animal population estimation requires frequent capture of infrared images and transferring these images to the server is not feasible due to bandwidth limitation and/or power consumption constraints. Hence image processing capability is needed in the sensor. The paper presents the common vole detection algorithm we developed and its power-aware implementation. We emphasize the need for more modular image processing frameworks that can be deployed on microcontrollers more easily. We also present our agricultural camera sensor platform that is suitable for various detection/observation tasks. Keywords: Agriculture, Infrared imaging, Image processing, Power efficiency. 1. Introduction

Agricultural sensor use cases include capturing

images for e.g. detecting drought, plant phenotype or diseases. These use cases require visible light [6, 7] or infrared imaging [1, 11]. While most applications require relatively simple sensors (e.g. capturing images several times a day), we present a case in this paper which requires more frequent sampling. This observation activity generates significant amount of data and transmitting this data from an isolated, battery-powered sensor operating far from the fixed network infrastructure is not a trivial task. This paper argues that in these use cases significant saving in power consumption can be achieved by implementing image processing capability in the sensor.

The energy consumption balance between data processing at the sensor endpoint vs. data processing at the server has already been analyzed in a general case [8]. In this paper we examine this question in a more special case, namely low-power microcontrollers as processing units, and limited communication options.

2. Common Vole Detection use Case AgroDat project, financed by the government of

Hungary intends to develop connected sensors for the agriculture. One of the more challenging use cases we identified is animal monitoring, specifically rodent tracking. Population outbreaks of certain rodent species can cause significant damage in crop production. More aggressive rodenticides are applied according to population estimation hence this estimation is an economically important task. Detection of wild animals during mowing operations reported by [10] requires similar technical solutions.

As common voles are night animals, the sensor used for population estimation must be able to detect these animals in the darkness. Previously the availability of long-wavelength infrared (LWIR) cameras was limited due to their high cost, therefore short-wavelength infrared (SWIR) cameras (like Kinect [2]) have been used for rodent tracking. SWIR cameras, however, have the disadvantage that the bait area needs to be illuminated by infrared light which limits their effective range. Relatively low-cost LWIR cameras appeared just recently.

We experimented with FLIR Lepton camera module whether small rodents can be detected reliably. The idea is that the rodents are attracted to a bait area which is surveyed by the infrared camera. The FLIR Lepton camera operates in the 8000-14000 nm wavelength range and has a resolution of 80×x60 pixels.

We made the following experiment. An animal similar to the common vole (Phodopus sungorus) was placed in a cage and images were captured with different distances between the camera and the animal. The background was lawn and other common foliage. The images were made in the night (Fig. 1.)

The infrared camera measures observed temperature values for each pixel. These temperature values are deduced from the infrared radiation observed in the viewport area corresponding to the pixel. In order to obtain an image with visible features, temperature range between the minimum and maximum temperature values in the input, raw image need to be mapped to intensity values (like 0-255 gray-scale) in the gray-scale image that acts as input to the image processing algorithm.

The small rodents we are looking for that are farther from the camera and therefore their observed size is smaller than a pixel size in this relatively low-resolution image look like colder than they actually are because the temperature of the elements of the foliage

Page 88: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

88

are calculated into the temperature measured for the pixel in question. The dynamic mapping of the temperature range in the raw input image to gray-scale representation means that as the warm object gets farther from the camera, features in the background get “brighter”.

Fig. 1. Small rodent similar to a common vole (Phodopus sungorus) in long-wavelength infrared image.

3. Vole Detection Algorithm

The goal of the vole detection is to identify images

where something relevant is captured. These images are then sent to the back-end server for further, more detailed analysis, eventually yielding the population estimate. Image processing is also important in case of extremely low-bandwidth wireless bearers like Sigfox where sending the image is not feasible. A Sigfox endpoint is able to send just 140 16-byte messages daily so sending the entire image is clearly not possible. The sensor has to extract some characteristic data (like the number of the vole-like objects identified in the image) and send this extracted data over the Sigfox network. The image itself is obtained by some other means (like auxiliary GSM network access providing only batch image upload or physical access to the mass storage (e.g. SD card) of the sensor).

The first version of the vole detection algorithm was implemented in OpenCV. The steps of the algorithm are the following. The greyscale image is transformed into a binary

image with a fixed threshold of 204. Contour tracing algorithm from the OpenCV

library is applied, then the resulting contours'

convex hull is filled. This step gets rid of spurious noise in the image resulting from the thresholding step.

Elements in the image are dilated then again contour traced.

Finally the enclosing circle of each resulting contour is calculated and these circles are compared to the circles obtained from the previous iteration. Largely overlapping circles are eliminated. If a circle is found moving and its size corresponds to the size of a vole, the image is stored and/or uploaded to the server. In order to ensure that the animal does not leave the

image when the next picture is taken but also moves significantly so that the circle representing the animal has sufficiently different location, we found that a frame rate of 1 Hz yields reliable results. Depending on the use case, this frame rate would be sustained continuously or just for a short period of time. We achieved good results by taking 5 consecutive pictures with 1 Hz frame rate then interrupting the image capturing/processing for 1 minute. Compared to the steady 1 Hz frame rate, this burst operation still identified the animals reliably because once they were in the bait area, they remained there for several minutes. On the other hand, the burst operation consumed significantly less power.

We prototyped the algorithm on an embedded Linux platform (BeagleBone Black/TI AM335x 1GHz ARM Cortex-A8) and we found good efficiency in recognizing relevant images. Unfortunately the high standby consumption of these embedded Linux platforms nullified any power consumption savings [3]. The project was therefore moved to a microcontroller unit (MCU) platform. Due to its high performance (internal floating-point unit (FPU), Cortex-M4 core, up to 168 MHz clock speed), large internal flash (512 Kbytes or 1 Mbytes, depending on subtype) and RAM memory (192 Kbytes) we chose the STM32F407 MCU and attempted to port OpenCV's 2 basic modules (core, imgproc) to the MCU. Even these modules required more flash space than the relatively large flash memory of this high-end MCU. The reason is OpenCV's heavily layered software architecture and its extensive usage of support libraries (e.g. libc, libm, libz, STL, etc.). Even though the actual image processing modules are relatively small, extracting them out of the OpenCV dependency network turned out to be too complicated.

We evaluated two additional image processing frameworks. CImg [4] is a C++ template library (hence it has dependency on STL) but it is missing morphological analysis tools needed for our vole detection algorithm. CVIPTools [5] is a quite exhaustive C library but the Linux version on which the STM32F407 port is based was last maintained in 2002. This version of CVIPTools does not support graphics processing units (GPU) either. Curiously, these features are advantages when it comes to using the library on an MCU as pure C implementation eliminates the need of STL support library and not

Page 89: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

89

even high-end MCUs have GPU. CVIPTools has the advantage that it depends only on the standard C library (libc). We satisfied this dependency by porting the Newlib library5 to the MCU. The flash image of the vole detection application with the relevant modules of CVIPTools and Newlib has the size of 126 Kbytes which fits conveniently into the MCU’s flash memory. This demonstrates that much more complex image processing algorithms can also be implemented on this platform.

While CVIPTools and OpenCV both offer plenty of algorithms and tools, the tool set is not exactly the same. The CVIPTools version, starting from the second step, employs a different processing. In the second step, after the greyscale-to-binary

conversion, a morphological dilating is performed followed by a morphological closing and an additional greyscale-to-binary thresholding operation.

Objects in the image are then labeled, yielding bounding boxes for contiguous objects.

The enclosing circles are calculated from these bounding boxes. Identification of the overlapping/moving circles is the same as in case of the OpenCV implementation. CVIPTools (on the STM32F407 MCU) and

OpenCV-based implementations (on BeagleBone Black) yield similar outputs and power consumption can be compared. The new, MCU-based implementation ported to CVIPTools consumes 0.0027 mAh when processing 5 consecutive pictures while the previous, embedded Linux-based implementation (OpenCV) needed 0.62 mAh. Moreover, the MCU is able to sleep with microamper-scale power consumption while the embedded Linux implementation consumes significant amount of power even when sleeping. In the previous iteration of the sensor [3], the sensor control logic was off-loaded to the GSM communication module (Telit GL865) that has user software execution feature due to the high power consumption of hardware responsible for the image processing function. The MCU-based implementation eliminated this more complex setup. In addition, the low power consumption in both computing and sleeping phases justifies the image processing capability in the sensor as significant saving is realized when only the relevant images are sent to the server.

We also tried to port CVIPTools and the vole detection algorithm to a much smaller microcontroller, an STM32L152RCT6. This MCU is optimized for ultra-low consumption application, has Cortex-M3 core, no FPU and up to 32 MHz clock speed. The MCU is also equipped with 256 Kbytes of flash memory and 32 Kbytes of RAM. Particularly the relatively small RAM is problematic for image processing applications but as our raw infrared image is just 9600 bytes, there was a hope that our vole detection algorithm fits into the RAM. The size of the application code (vole detection+relevant modules of CVIPTools and

5 https://sourceware.org/newlib/

Newlib) was 122 Kbytes which compares favorably with the total flash size of 256 Kbytes. No matter how hard we tried, however, the object labeling step required more memory than the about 29 Kbytes available for the C heap. Also, due to the lack of FPU support, (partial) processing of one image required 420 msec which indicates that even if there was enough memory, the desired frame rate of 1 sec would be hard to achieve.

4. The Camera Sensor

The experiments described in the previous sections led us to construct a multi-purpose agricultural camera sensor. The head unit of the sensor can be seen in Fig. 2. This head unit is usually mounted on a pole so that the vegetation or the bait area (in case of rodent sensor) can be observed. The sensor optionally contains 4 visible-light cameras, positioned 90 degrees from each other and 1 LWIR camera. Power supply of each of these cameras can be enabled separately, allowing the developer of the sensor application to switch on the cameras only when needed.

The sensor is equipped with multiple communication options that can also be deployed optionally. GSM modem provides the capability to perform bulk image upload. Low-power wide area (LPWAN) modem (Sigfox in the current version of the camera sensor) is used for delivering short messages in a power-efficient way – like sending the number of rodents detected in the bait area.

In order to demonstrate the need for multiple communication options, Fig. 3 and Fig. 4 depicts the power consumption of sending a small data item (60 bytes) by GSM/GPRS and Sigfox. The GSM/GPRS modem was Telit GL865, the Sigfox modem was Adeunis Si868. The Sigfox modem was controlled by an Atmel ATmega2560 MCU whose power consumption in this scenario was negligible, the Telit GL865 was controlled by its own, Python-based execution logic. The GSM/GPRS scenario included network registration, PDP context activation, data transmission and network un-registration procedures. Sigfox does not need registration, the power consumption diagrams show the sending of 4 messages as the 60 bytes payload fits only into 4 16-byte Sigfox messages. The result is that GSM/GPRS needs approximately 1 mAh power consumption while the Sigfox scenario requires 0.2 mAh. Also, GPRS maximum power consumption during the scenario is much higher which allows the Sigfox option to be implemented with smaller batteries. In order to transfer data relevant to an image over the extremely low-bandwidth Sigfox network, the sensor unit must extract relevant features from the image by means of image processing. A similar experience has been reported for other low-bandwidth networks operating in the license-free spectrum used to transfer image data [9].

Page 90: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

90

In case of LPWAN communication, there is an option that the relevant images are stored on an SD card in the sensor, available off-line (when the service personnel visits the camera sensor). The SD card can also store images for batch upload operations by means of the GSM modem, if that option is installed. Another option is a large, 4 Mbytes RAM that can act as a temporary memory for image processing operations on the large images that the visible-light cameras produce.

Fig. 2. Head unit of the camera sensor.

These optional features make the camera sensor a

versatile platform whose application areas span from simple foliage observation (with visible-light or LWIR camera) to more complex detection tasks requiring image processing. The STM32F407 MCU does have limitations with regards to complex image processing operations but the relatively powerful ARM core and the extensive feature set of CVIPTools does permit the implementation of reasonably sophisticated image processing. Also, the communication architecture that supports power-intensive but relatively high-bandwidth (cellular) and low-power wide area (Sigfox in our case) network support permits both short message sending with very small power consumption and bulk image uploads.

Fig. 3. Power consumption of sending a 60-byte packet by GSM/GPRS.

Fig. 4. Power consumption of sending a 60-byte packet by Sigfox.

5. Conclusions Sensors are often considered to be data capture devices which just transfer the data to more powerful nodes (“servers”) where the data is processed. Limited communication bandwidth or limited battery power may require more sophisticated data processing in the sensor. The common vole detection use case presented in this paper aimed to demonstrate that image processing frameworks with complex dependency structures and layered (as opposed to modular) architecture are often unsuitable for low-power environments. Also, the low-power and low-bandwidth communication options like Sigfox require that sensors communicate just the relevant features of the image and not the entire image. It is also often a requirement to transfer the images themselves for further processing on the server. This requires additional transfer mechanisms (off-line or high-power, high-bandwidth communication option) in addition to the low-power, low-bandwidth network. References [1]. A. Manickavasagan et al., Applications of Thermal

Imaging in Agriculture–A Review, Canadian Society for Engineering in Agricultural, Food and Biological Systems, 2005, pp. 05-002.

[2]. Z. Wang, S. A. Mirbozorgi, M. Ghovanloo, Towards a kinect-based behavior recognition and analysis system for small animals, in Proceedings of the Biomedical Circuits and Systems Conference (BioCAS), Atlanta, Georgia, USA, 22-24 Oct. 2015.

[3]. G. Paller, G. Élő, Energy-efficient operation of GSM-connected infrared rodent sensor, in Proceedings of the 5th International Conference on Sensor Networks SENSORNETS’16, Rome, Italy, 2016.

[4]. The Cimg Library (http://cimg.eu/) [5]. CVIPTools (http://cviptools.ece.siue.edu/) [6]. H.-E. Nilsson, Remote sensing and image analysis in

plant pathology, Canadian Journal of Plant Pathology, Volume 17, Issue 2, 1995.

[7]. J. Romeo, G. Pajaresa, M. Montalvo, J. M. Guerreroa, M. Guijarroa, J. M. de la Cruz, A new Expert System for greenness identification in agricultural images, Expert Systems with Applications, Vol. 40, Issue 6, May 2013, pp. 2275–2286.

Page 91: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

91

[8]. K. Kumar, Y.-H. Lu, Cloud Computing for Mobile Users: Can Offloading Computation Save Energy?, Computer, Vol. 43, Issue 4, pp. 51–56.

[9]. T. Wark et al., Transforming Agriculture through Pervasive Wireless Sensor Networks, IEEE Pervasive Computing, Vol. 6, Issue 2, April-June 2007.

[10]. K. A. Steen, A. Villa-Henriksen, O. Roland Therkildsen, O. Green, Automatic Detection of

Animals in Mowing Operations Using Thermal Cameras, Sensors, Vol. 12, Issue 6, 2012, pp. 7587-7597.

[11]. N. R. Falkenberg, G. Piccinni, J. T. Cothren, D. I. Leskovar, C. M. Rush, Remote sensing of biotic and abiotic stress for irrigation management of cotton, Agricultural Water Management, Vol. 87, Issue 1, 10 January 2007, pp. 23–31.

Page 92: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

92

Entropy-Based Markers of EEG Background Activity of Stroke-Related

Mild Cognitive Impairment and Vascular Dementia Patients

Noor Kamal Al-Qazzaz 1,5, Sawal Ali 1, Siti Anom Ahmad 2, Md. Shabiul Islam 3, Javier Escudero 4

1 Department of Electrical, Electronic and Systems Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia; UKM Bangi, Selangor 43600, Malaysia

2 Department of Electrical and Electronic Engineering, Faculty of Engineering, Universiti Putra Malaysia;UPM Serdang, Selangor 43400, Malaysia

3 Institute of Microengineering and Nanoelectronics (IMEN); Universiti Kebangsaan Malaysia; UKM Bangi, Selangor 43600, Malaysia

4 Institute for Digital Communications; School of Engineering; The University of Edinburgh; Edinburgh EH9 3FB; United Kingdom

5 Department of Biomedical Engineering, Al-Khwarizmi College of Engineering, Baghdad University, Baghdad 47146, Iraq

Tel.: + 6126480895 E-mail: [email protected]

Summary: The aim of the present study was to develop valuable and reliable indices of post-stroke dementia particularly vascular dementia (VaD) using entropy-based features extracted from the electroencephalography (EEG) background activity of 5 VaD patients, 15 stroke-related patients with mild cognitive impairment (MCI) and 15 control healthy subjects during a working memory (WM) task. EEG artifacts were removed using independent component analysis and wavelets (AICA-WT). Using ANOVA (p < 0.05), spectral entropy (SpecEn) was used to test the hypothesis that the EEG signal slows down in both VaD and MCI in comparison with control subjects whereas the permutation entropy (PerEn) and tsallis entropy (TsEn) features were used to test the hypothesis that the complexity in both VaD and MCI were reduced in comparison with control subjects. SpecEn reflected the slowing in the brain activity in VaD and MCI patients whereas PerEn and TsEn results in reducing the complexity in VaD and MCI patients. Therefore, EEG could be as a reliable index for inspecting the background activity in the identification of patients with VaD and stroke-related MCI.

Keywords: Electroencephalography, ICA-WT, Spectral entropy, Permutation entropy, Tsallis entropy, Mild cognitive impairment, Dementia. 1. Introduction

Cognitive impairment and dementia are common

after stroke. 30 % of stroke patients are prone to develop vascular dementia (VaD) within the first year of stroke onset. Mild cognitive impairment (MCI) is considered as a stage between early normal brain cognition and late severe dementia. The highest effect of a stroke is observed on the memory [1].

The early diagnosis of dementia will help dementia patients start an early treatment based on the symptoms. Therefore, developing markers to identify dementia in early stages is important to derive an optimal diagnostic index. Electroencephalogram (EEG) has been used as the full investigation of sensitive marker that helps in detecting cortical abnormalities associated with cognitive decline and dementia. EEG is a widely available, cost-effective, and non-invasive tool. The clinical EEG wave forms have an amplitude around 10-100 µv and frequency range of 1 to 100 Hz [1].

2. Methods and Materials In this paper, ICA-WT has been used to denoise the

EEG datasets. The spectral entropy (SpecEn),

permutation entropy (PerEn) and tsallis entropy (TsEn) [2-4] were extracted to examine the EEG background activity in MCI and VaD patients compared to control subjects.

2.1. Subjects and EEG Recording Procedure For this study, EEG were examined for 15 normal

records, 15 stroke-related MCI patients, and 5 VaD patients. The patients were recruited from the stroke ward and neurology clinic of the Pusat Perubatan Universiti Kebangsaan Malaysia (PPUKM). All experiment protocols were approved by the Human Ethics Committee of the National University of Malaysia. An information consent forms were also signed by the participants. The EEG activities were recorded using the NicoletOne systems. A total of 19 electrodes were used. In this study, an auditory WM task session was conducted. The session started with fixation cue when the subjects were asked to be motionless. A WM task was then performed, the subjects were asked to memorize five words for 10 s. Afterward, they were asked to remember these words with their eyes closed, and the EEG were recorded. After 60 s, the patients were asked to open their eyes and enumerate all words that they could remember.

Page 93: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

93

2.2. Denoising Stage In this paper, ICA is used to split a set of recorded

EEG signal into its sources. FastICA algorithm has been used due to its simplicity, fast convergence and efficiency to decompose the recorded EEG as in Equation 1:

x(t)=As(t) (1) where x(t) is the output vector, A is the mixing matrix, s(t) is the input vector. The artefactual independent components (ICs) were detected using statistical metrics and were arranged into new dataset to pass through the wavelet transform (WT) to denoise them. The corrected ICs were returned back to the EEG set to be reconstructed and used in the next stage.

2.3. Features Extraction In this study, 60 seconds, N=15360 samples,

6 windows of 10 second length (2560 samples) were used to extract features from the original EEG time series for each 19 channels. Entropies have been used to detect abnormalities in the EEGs’ of dementia patients. SpecEn for frequency range (0.5 to 64) Hz, has been used to distinguish dementia patients EEGs from the normal subjects and it is computed as in [5]. Given that the brain can investigate the complex dynamic information, entropies including PerEn with embedded dimension d=5 and time delay l=1 and TsEn with parameter q=2 have been used. PerEn also provides an alternate way of measuring similarity among patterns with respect to other types of complexity measurements, PerEn is calculated as in [4]. TsEn is non-additive, a description also referred to as non-extensive and it can be estimated as in [6].

3. Statistical Analysis

In order to perform ANOVA, feature results of the 19 channels from the EEG dataset were grouped into 5 recording regions that correspond to the scalp area. These are frontal (Fp1, Fp2, F3, F4, F7, F8, and Fz), temporal (T3, T4, T5, and T6), parietal (P3, P4, and Pz), occipital (O1 and O2), and central (C3, C4, and Cz). Statistical analyses were performed through ANOVAs in SPSS 22. Firstly, two-way ANOVA was applied on the SpecEn features, group factor (control, MCI, VaD) and the five groups of the scalp regions were the independent variables (IV), whereas the SpecEn was the dependent variable (DV). Secondly, two-way ANOVA was applied on the PerEn features, group factor (control, MCI, VaD) and the five groups of the scalp regions were the IV, whereas the PerEn was the DV. Finally, two-way ANOVA was applied on the TsEn features, group factor (control, MCI, VaD) and the five groups of the scalp regions were the IV, whereas the TsEn was the DV. The post-hoc comparison was performed through Duncan’s test. The significance for all statistical tests was set at p ˂ 0.05.

4. Results

In Table 1, the SpecEn, PerEn and TsEn values for the VaD, MCI patients and control subjects in the five scalp regions are given. It is evident that the EEG activity of the results showed that the VaD and MCI patients still had slightly lower values at all scalp regions, although the differences were not significant (p<0.05) when using SpecEn (SpecEnVaD < SpecEnMCI

< SpecEnControl). Further inspection of VaD and MCI patients are significantly more regular (less complex) as in (PerEnVaD<PerEnMCI<PerEnControl) and (TsEnVaD<Tsunamic<TsEnControl) with significant differences among the groups at all scalp regions (p<0.05).

Table 1. The average values (Mean ± SD) of EEGs for the VaD, MCI patients and the control subjects for the five scalp regions. Significant group differences are marked with an asterisk.

Regions Control MCI VaD p-

value

Frontal 0.77±0.08 0.75±0.09 0.75±0.05 0.12 Temporal 0.79±0.07 0.78±0.08 0.77±0.04 0.32 Parietal 0.78±0.05 0.76±0.08 0.75±0.05 0.17 Occipital 0.76±0.063 0.74±0.082 0.72±0.04 0.28 Central 0.81±0.054 0.79±0.08 0.79±0.04 0.6

Frontal 4.37±0.143 4.12±0.254 3.93±0.332 0.05* Temporal 4.19±0.231 4.07±0.247 4.06±0.187 0.008* Parietal 4.15±0.176 4.02±0.17 3.90±0.237 0.05* Occipital 4.19±0.178 4.08±0.17 3.97±0.244 0.001* Central 4.13±0.179 4.02±0.238 3.96±0.281 0.004*

Frontal 0.94±0.016 0.94±0.012 0.93±0.018 0.01* Temporal 0.95±0.004 0.95±0.007 0.94±0.011 0.001* Parietal 0.95±0.008 0.93±0.008 0.92±0.008 0.008* Occipital 0.95±0.002 0.94±0.012 0.94±0.012 0.001* Central 0.95±0.006 0.95±0.008 0.95±0.005 0.038*

Page 94: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

94

4. Conclusions

In the current ICA-WT technique has used to denoise the EEG datasets. Entropies have been used to detect abnormalities in the VaD and MCI patients compared to control subjects EEGs’. SpecEn reflected the slowing in the electrical brain activity in VaD and MCI patients whereas PerEn and TsEn results in reducing the complexity in VaD and MCI patients. As the EEG has been widely used as a potential screening technique in clinical practice due to its low cost and portability, it could become a reference in planning and customizing an optimal therapeutic program to address the changes associated with MCI and dementia. This study suggests that the entropy-based markers of EEG background activity in VaD and MCI patients might be helpful in providing useful diagnoses indexes for early dementia detection.

References [1]. N. K. Al-Qazzaz, S. H. B. Ali, S. A. Ahmad, K.

Chellappan, M. S. Islam, and J. Escudero, Role of EEG as Biomarker in the Early Detection and Classification of Dementia, The Scientific World Journal, Vol. 2014, 2014.

[2]. N. K. Al-Qazzaz, S. Ali, S. Islam, S. Ahmad, and J. Escudero, EEG Wavelet Spectral Analysis During a Working Memory Tasks in StrokeRelated Mild Cognitive Impairment Patients, in Proceedings of the International Conference for Innovation in Biomedical Engineering and Life Sciences, 2016, pp. 82-85.

[3]. R. Sneddon, W. Shankle, J. Hara, J. Fallon, and U. Saha, The Tsallis Entropy in the EEGs of Normal and Demented Individuals, in Proceedings of the 11th Joint Symposium on Neural Computation, University of Southern California, 15 May 2004, (Unpublished) http://resolver.caltech.edu/CaltechJSNC:2004.poster025.

[4]. F. C. Morabito, D. Labate, F. La Foresta, A. Bramanti, G. Morabito, and I. Palamara, Multivariate multi-scale permutation entropy for complexity analysis of Alzheimer’s disease EEG, Entropy, Vol. 14, 2012, pp. 1186-1202.

[5]. J. Escudero, R. Hornero, D. Abásolo, and A. Fernández, Blind source separation to enhance spectral and non-linear features of magnetoencephalogram recordings. Application to Alzheimer's disease, Medical Engineering & Physics, Vol. 31, 2009, pp. 872-879.

[6]. S. Robert, The Tsallis entropy of natural information [J], Physica A: Statistical Mechanics and Its Applications, Vol. 386, 2007, pp. 101-118.

Page 95: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

95

Modeling of Non-Dispersive Infrared Gas Sensors

F. W. Dietachmayr 1 and B. G. Zagar 1

1 Johannes Kepler University, Institute for Measurement Technology, Altenbergerstr. 69, 4040 Linz, Austria Tel.: +43 732 2468 5924, fax: +43 732 2468 5933

E-mail: [email protected] Summary: Non-dispersive infrared (NDIR) sensors are widely used to measure gas concentrations. However, there is currently no complete mathematical model for NDIR sensors available to speed-up the development process of new sensor alterations. This paper reports on an ongoing research to derive and verify accurate mathematical models for various components of NDIR sensors. In this paper, mathematical models for main components of an NDIR sensor are presented. They cover among others the thermal light source and the behavior of the thermopile detector. Finally, each mathematical model is verified by measurements. Keywords: NDIR, Gas sensors, Optical gas sensing, Mathematical model, Thermopile.

1. Introduction

Gas fraction measurement plays an important role

in various industrial and scientific processes. Gas sensors are used in petrochemical industries to ensure safety and measure species in products and processes [1]. In automotive industries, carbon dioxide concentration is both an important parameter for cabin air quality as well as a safety parameter when carbon dioxide is used as the refrigerant in the air conditioning system of a car [2]. Finally, measuring greenhouse gases in the atmosphere is an important task in atmospheric sciences [3].

There is a variety of different gas sensor technologies available: small low-cost devices such as electrochemical or semiconductor gas sensors, laboratory equipment such as gas chromatographs, and optical gas sensors based on molecular absorption. While laboratory equipment is expensive and is usually used for analyzing samples in a non-real-time setting, electrochemical, semiconductor, and optical gas sensors can be used in mass-market products. Electrochemical and semiconductor gas sensors suffer from limited lifetimes and drift, respectively, as well as cross-responses to other gases [4]. Optical gas sensors offer fast response times and, if carefully designed, almost zero cross-responses to other species. Therefore, optical gas sensors are widely used in applications requiring low maintenance and high specificity (e.g., automotive applications, heating, ventilation and air conditioning control, etc.).

One of the most widely used optical gas sensing principles is the non-dispersive infrared (NDIR) technique. In recent years, various research has been done regarding an increase of the flexibility of NDIR sensors (e.g., by using tunable infrared filters [5], varying the emitted spectrum of the infrared source [6], [7], etc.). Currently, there is no complete mathematical model for NDIR sensors available to estimate the impact of different sensor alterations on the sensor performance. Therefore, prototypes have to be built

and analyzed which is a time-consuming and expensive process.

After a short review of the NDIR sensor principle (Sec. 2) this paper presents mathematical models for the main components of an NDIR sensor: the infrared light source (Sec. 3), the measurement chamber (Sec. 4), and the detector (Sec. 5). Finally, the models are verified by measurements (Sec. 6).

2. NDIR Sensor Principle

Most gas species exhibit absorption bands in the near and mid infrared regions. The absorption bands consist of narrow lines due to molecular vibrations and rotations and are specific to the species [8]. Therefore, the abundance of a species can be calculated by measuring the absorption of infrared light transmitted through a sample volume (e.g., gas in a gas cell). Molecular absorption can be described by the Beer-Lambert Law [9]:

, (1)

where, IT, is the intensity of the light transmitted through the gas cell, I0 is the intensity of the incident light, α is the absorption coefficient of the gas, and z is the length of the optical path. A simple NDIR sensor consists of an infrared light source, a measurement chamber and a detector. 3. Infrared Light Source

There are mainly three different types of infrared light sources available: thermal light sources (e.g., micro-bulbs, IR emitting membranes), LEDs, and lasers. Most commercially available NDIR sensors use micro-bulbs as light sources as micro-bulbs have a relatively high spectral emission and their cost is low [10]. Subsequently, thermal light sources will be discussed in more detail.

Page 96: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

96

Using the law of conservation of energy the filament temperature T of a thermal emitter can be calculated:

d

d, (2)

where, I, is the electrical current flowing through the filament, R(T) is the resistance of the filament, and a(T) and b are parameters. Thermal emitters usually have a vacuum or low pressure environment around their filament which significantly reduces thermal conduction. Therefore, thermal conduction is neglected.

The parameters a(T) and b can be determined experimentally by measurements. For the calculation of the parameters samples of the micro-bulb HSL 5-115 by the company Heimann Sensors were used. Input voltage steps U of various amplitudes were applied to the micro-bulb and the electrical current through the filament I(T) was measured. Using Ohm’s law the resistance R(T) of the filament was calculated and, subsequently, the temperature of the filament was determined by evaluating the temperature dependence of the resistance

1 T

T , (3)

where, R0, is the resistance of the filament at temperature T0 . As tungsten is used as filament the parameters αT and βT were determined from the data for tungsten in [11].

The parameter a(T) was calculated from the steady-state temperature. Fig. 1 shows the parameter a(T) for multiple filament temperatures T for three samples of micro-bulbs. As the influence of the temperature on the parameter for the interesting temperature region is low a constant parameter a was used by calculating the mean value.

The parameter b was estimated by minimizing the mean square error between the measurements and a simulation based on equation (2). Fig. 2 shows the parameter b for multiple input voltages U for three samples of micro-bulbs. The scattering of the data points is reasonably low. Therefore, the mean value was calculated and used for the model.

Fig. 3 shows a comparison between the simulated filament temperature and the calculated filament temperature from measurements for three different input voltages. The simulation is in good agreement with the measurements.

The emission spectrum of a thermal emitter is usually assumed to follow the Planck emission curve for a black body with emissivity less than unity [1]. The radiant intensity I0 can then be written as

,

2 1

1, (4)

where, σ. is the Stefan-Boltzmann constant, h is Planck’s constant, c is the speed of light in vacuum, k

is Boltzmann’s constant and λ is the wavelength of the emitted light.

Fig. 1. Calculated parameter a(T) for multiple filament temperatures T. Three different micro-bulbs were measured.

Fig. 2. Estimated parameter b for multiple micro-bulb input voltages U. Three different micro-bulbs were measured. A

series resistor of 6.8 Ω was used for the measurements.

Fig. 3. Comparison between simulated filament temperature (solid line) and calculated filament temperature from measurement (x). A series resistor of 6.8 Ω and input

voltages of 2 V (blue), 3 V (red), and 4V (green) were used for the measurements.

900 1200 1500 1800

12

13

14

15

T in K

a(T

) in

W/K

4

Source 1Source 2Source 3

1.5 2.5 3.5 4.5

1.95

2.05

2.15

2.25

U in V

b in

nJ/

K

Source 1Source 2Source 3

0 0.2 0.4 0.6 0.8 1200

400

600

800

1000

1200

1400

1600

1800

t in s

T in

K

Page 97: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

97

If the thermal emitter has an envelope the transmission curve of the envelope has to be taken into account, too. This is done by integrating Equation (3) numerically over wavelength steps of λStep = 10 nm and, subsequently, multiplying the resulting intensities with the respective transmission factor of the envelope. Fig. 4 shows the transmission curve of the used micro-bulb HSL 5-115.

Fig. 4. Transmission curve of the envelope of the micro-bulb HSL 5-115 [12].

4. Measurement Chamber

In the measurement chamber the emitted light from the infrared light source gets partially absorbed by the gas species to be measured. The absorption is calculated by using the Beer-Lambert Law (Eq. (1)). The absorption coefficient is determined from line-by-line data from the HITRAN data base [13]. 5. Detector

There are mainly two different types of infrared detectors available: thermal detectors (e.g., bolometers, thermoelectric detectors, pyroelectric detectors) and photodetectors (e.g., photoelectric detectors, photo conductors). In this paper thermo-electric detectors (thermopiles) are discussed in more detail.

Fig. 5 depicts the thermal model of a thermopile detector. The model consists of three parts: the housing, the hot junction, and the cold junction, each with its own temperature T and thermal capacity C. The housing is connected to the hot junction via the thermal resistance KGH and is additionally connected to the surrounding environment via the thermal resistance KGA. The hot junction is additionally connected to the cold junction via the thermal resistance KHC. The cold junction of a thermopile is usually thermally well connected to a heat sink in order to keep the temperature of the cold junction stable. Therefore, the cold junction is assumed to be directly thermally connected to the surrounding environment via the thermal resistance KCA.

Using the law of conservation of energy the hou-sing temperature TG, the hot junction temperature TH, and the cold junction temperature TC of a thermopile detector can be calculated:

Fig. 5. Thermal model of a thermopile detector.

Gd G

d G G GH H GA A

Hd H

d H H GH G HC C,

Cd C

d C C HC H CA A

(5)

where CG, CH, CC are the thermal capacities of the housing, the hot junction, and the cold junction, re-spectively, PG, PH, PC are the input powers, and TA is the ambient temperature. The thermal resistances K1, K2, and K3 can be calculated by

1 GH .

3 HC (6)

As the temperature changes are usually low (mK or

lower) thermal radiation is neglected. The output voltage of the thermopile can be calcu-

lated by

out s H K , (7) where ss is the Seebeck’s constant of the thermopile.

For the estimation of the model parameters (i.e., the thermal capacities C and the thermal resistances K) the temperatures of the three model parts need to be measured. As this cannot easily be done a different approach was chosen.

Using Equations (5), (6), and (7) the transfer functions from the input powers to the output voltage can be calculated:

2 4 6 8 100

20

40

60

80

100

λ in μm

Tra

nsm

issi

on in

%

housingthermal cap. C

hot junctionthermal cap. C

cold junctionthermal cap. C

G

H

K

ambient temperature Ta

ambienttemperature Ta

K

K

K

K

GA

GH

HC

CA

T

T

T

G

H

K

P

P

P

G

H

C

Page 98: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

98

G out

H out

,

C out

(8)

with the parameters axy and bxy.

The parameters axy and bxy can be estimated from measurements. For the estimation of the parameters samples of the thermopile HIS E222 F1 F2 Gx (Heimann Sensors) were used. A laser beam was focused on the housing, the hot junction, and the heat sink of the cold junction to apply the input powers PG, PH, and PC, respectively. The wavelength of the laser beam was 450 nm and the optical power was changed between 410 µW and 840 µW based on a sine sweep with a frequency range of 10 mHz to 10 Hz with a sweep time of 500 seconds. The output voltage of the thermopile was measured and, subsequently, the parameters of the transfer functions were estimated in MATLAB using the transfer function estimation algorithm (command tfest) from the control system toolbox. The estimation algorithm uses prediction error minimization [14] and, for parameter initialization, the refined instrumental variable method [15]. Figs. 6 to 8 show a comparison between the measured step responses of the thermopile output voltage Uout to input power steps PG, PH, and PC, respectively, and the simulated step responses. The simulation is in good agreement with the measurements.

In order to guarantee the specificity of the NDIR sensor to only one gas species an optical bandpass filter is used. The effects of the filter can easily be taken into account by multiplying the incoming intensities from the light source with the respective transmission factor of the filter.

6. Measurement Results

To verify the derived mathematical models, a thermopile (HIS E222 F1 F2 Gx) was placed in front of a micro-bulb. The output voltage of the thermopile Uout was measured for different micro-bulb input voltages. Additionally, the thermopile output voltage was simulated using the mathematical models. Fig. 9 shows a comparison between the simulated thermopile output voltages and the measured output voltages. The simulation is in good agreement with the measurements. This proves that the suggested simulation models are suitable for simulating the sensor behaviour.

7. Conclusion

In this paper mathematical models for different parts of NDIR sensors were presented. All models are based on fundamental physical relations. Therefore, the models can easily be adapted to different thermal emitter and thermopile types. Measurements proved that the suggested models are suitable for describing NDIR sensor behavior. Acknowledgement

The authors gratefully acknowledge the partial financial support by the Austrian Research Promotion Agency and the Austrian COMET program.

Fig. 6. Comparison between simulated thermopile output voltage (red) and measured thermopile output voltage (blue)

for an input power step of PG of 840 µW.

Fig. 7. Comparison between simulated thermopile output voltage (red) and measured thermopile output voltage (blue)

for an input power step of PH of 7.1 µW.

0 5 10 15 200

0.2

0.4

0.6

0.8

1

t in s

Uou

t in V

0 0.02 0.04 0.06 0.08 0.10

0.5

1

1.5

t in s

Uou

t in V

Page 99: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

99

Fig. 8. Comparison between simulated thermopile output voltage (red) and measured thermopile output voltage (blue)

for an input power step of PC of 840 µW.

Fig. 9. Comparison between simulated thermopile output voltage (solid line) and measured thermopile output voltage (x). Micro-bulb input voltages of 2 V (blue), 3 V red), and

4V (green) were used for the measurements.

References [1]. J. Hodgkinson et al., Optical Gas Sensing: a review,

Measurement Science and Technology, Vol. 24, No. 1, 2013, p. 012004.

[2]. R. Frodl et al., A High-Precision NDIR CO2 Gas Sensor for Automotive Applications, IEEE Sensors Journal, Vol. 6, 2006, pp. 1697-1705.

[3]. P. Laj et al., Measuring atmospheric composition change, Atmospheric Environment, Vol. 43, No. 33, 2009, pp. 5351–5414.

[4]. J. Zosel et al., The measurement of dissolved and gaseous carbon dioxide concentration, Measurement Science and Technology, Vol. 22, No. 7, 2011, p. 072001.

[5]. N. Neumann et al., Mikromechanisches durch-stimmbares fabry–perot-filter für die ir-gasanalytik (micromechanical tunable fabry–perot filter for ir gas analysis), tm - Technisches Messen, 72, 1-2005, 2005, pp. 10–15.

[6]. D. A. Andrews et al., Gas analysis using an infrared source with temporally varying temperature, Measurement Science and Technology, Vol. 12, No. 8, 2001, pp. 1263–1269.

[7]. A. Graf, Software tailored non-dispersive infrared sensors, ser. Dresdner Beiträge zur Sensorik, TUDpress, 2009, Vol. Bd. 36.

[8]. P. F. Bernath, Spectra of Atoms and Molecules, Oxford University Press, Oxford, Jul. 2005.

[9]. W. Demtröder, Laserspektroskopie, Springer, Berlin, Jul. 2007.

[10]. S. D. Smith et al., Recent developments in the applications of mid-infrared lasers, leds, and other solid state sources to gas detection, in SPIE Proceedings, Symposium on Integrated Optoelectronic Devices, 2002, pp. 157–172.

[11]. P. D. Desai et al., Electrical resistivity of selected elements, Journal of Physical and Chemical Reference Data, Vol. 13, No. 4, 1984, p. 1069.

[12]. HSL Series: IR-Lamps, Datasheet, Heimann Sensor GmbH, Dresden, Apr. 2008.

[13]. L. S. Rothman et al., The hitran2012 molecular spectroscopic database, Journal of Quantitative Spectroscopy and Radiative Transfer, Vol. 130, 2013, pp. 4–50.

[14]. L. Ljung, Prediction error estimation methods, Circuits, Systems, and Signal Processing, Vol. 21, No. 1, 2002, pp. 11-21.

[15]. P. C. Young, The refined instrumental variable method, Journal Européen des Systèmes Automatisés, Vol. 42, No. 2-3, 2008, pp. 149-179.

0 10 20 30 40 50−0.25

−0.2

−0.15

−0.1

−0.05

0

t in s

Uou

t in V

0 5 10 15 200

0.1

0.2

0.3

0.4

0.5

0.6

t in s

Uou

t in V

Page 100: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

100

Novel High Reliable Si-Based Trace Humidity Sensor Array

for Aerospace and Process Industry

Shuyao Zhou1, Biswajit Mukhopadhyay1, Piotr Mackowiak1, Oswin Ehrmann1,4, Kevin Kröhnert1, Robert Gernhardt1, Klaus-Dieter Lang1,4, Michael Woratz2, Peter Herrmann2, N. Volkmer3, Olaf Pohl3,

Volker Noack3, Ha-Duong Ngo1,3 1Fraunhofer Institute of Reliability and Microintegration, Berlin, Germany

2ACI GmbH, Berlin, Germany 3University of Applied Sciences, Microsystems Engineering, Berlin, Germany

4Technical University Berlin, Germany E-mail: [email protected]

Summary: In this paper we present high reliable and accurate silicon-based trace humidity sensors for use in aerospace and process industry. The sensors have been realized by using simple MEMS technology in this work. One is a single sensor (sensor cell) for monitoring of humidity, and sensor array is able to measure the total trace humidity in the atmosphere. It’s very suitable for aerospace and process industry such pharmacy applications. Keywords: MEMS, Micro sensors, Humidity sensor.

1. Introduction

The coulometric sensor principle [1] is normally

used in trace humidity sensing (Fig. 1). The sensor concept using glass carrier with Pt-wire and sensing material (phosphoric pentoxide P2O5). It is based on the absorption of water vapour from the atmosphere by the highly hygroscopic phosphoric pentoxide (P2O5), which is located on the sensor between platinum electrodes. Realized sensors using glass carrier and thick film technology are robust but bulky and have some limitations such as reliability [1].

Fig. 1. Schematic view of a sensor on the market on glass carrier with Pt-wire and sensing material.

2. Sensor Concept, Design and Technology

The single sensor concept is shown Fig. 2. The sensing Pt-electrode is deposited on a 3D silicon surface etched in silicon substrate in KOH-solution. The idea behind it is to increase the active surface of the sensing electrode in compare to a planar concept. The other advantage of this 3D concept is to use the capillary force to get the sensing material into the silicon trenches. The very smooth trench surface is very suitable for a thin high quality Pt electrode. The Pt layer shows no defects and non-perfections such as

in standard sensors. Main disadvantage of this single sensor cell concept is the nonability to have exact information about the total humidity in the air. Just one part of the water in the air can be absorbed and converted into charges. Sensors on the market need therefore a calibration. Fig. 3 shows the array concept. The sensor surfaces are exposed to the air flow. By adjusting the sensors voltages the signal of the last sensor can be adjusted toward zero – it means that the humidity in air is fully absorbed by the sensors. So it’s very easy to measure the absolute humidity in the air without calibration. Another advantage of the array concept is a better redundancy.

The designed single sensor (sensor cell) has a size of 11.5 mm x 20 mm x 525 µm, and the sensor array has dimensions of 60 mm x 20 mm x 525 µm. The manufacturing technology is very simple using standard silicon MEMS processes. First the substrate is passivated by using double layers SiN and SiO. After that the double layer is structured as etching mask for following KOH etching. After KOH etching the mask is remove. Before Pt deposition by using sputtering a thin SiO was grown by using thermal oxidation. Purpose of this SiO layer is an isolation of Pt layer. Spray coating is used to make a lift-off mask for Pt deposition. After Pt sputtering the resist layer is removed.

Fig. 2. Schematic view – single silicon-based humidity sensor in this work.

Sensing material

Pt-wire

KOH-trench Sensing material

Pt-electrode

Page 101: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

101

The sensor current (charges) can be calculated by using the simple equation:

Va **0712104.1I , (1)

where a: absolute humidity and

V :the gas flow.

Fig. 3. Schematic design of the sensor array containing 4 single sensor cells for measurement of absolute humidity.

3. Fabrication

The sensor cells are fabricated by using standard silicon technology. First the alignment marks were etched in (100) silicon wafer. In the next step the deep trenches (100 µm) were etched by using KOH. After KOH etching the surface was passivated by a thick SiO2 layer. For lift-off a spray coating for photo resist was necessary. Pt layer of 100 nm was sputtered. Fig. 4 shows briefly process steps for fabrication of the sensor array in silicon.

a) SiO/SiN deposition

b) KOH etching

c) Passivation SiO

d) Spray coating

e) Sputtern Pt and lift-off

Fig. 4. Main process steps to fabricate the sensor array.

4. Characterization

Fig. 5 shows the top view of a realized sensor array with KOH-trenches and Pt-electrodes on top.

Fig. 5. Top view of the realized silicon-based sensor array without sensing material on top

(60 mm × 20 mm × 525 µm).

Tests have been done to characterize the realized sensors. A special package has been realized to meet the requirements for monitoring of humidity [2]. The whole sensor package is shown in Fig. 6.

Fig. 6. Package for the sensor array. The measurement set up can be seen in Fig. 7.

Fig. 7. Principle setup for characterization of the realized sensor arrays.

Flow

Sensor cells

Page 102: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

102

For the characterization nitrogen gas (N2) with 99,999 % purity is used as gas supply based on its ideal behavior which follows Dalton's law of partial pressure at standard atmosphere. The dry nitrogen gas flow through Teflon tubing in two directions separately. One of them flow through the cool box with ice, which keep the condition of low humidity approx. 1000 ppm. In this case the humidity concentration can be calculated according to DIN 4108 [3]. The reference flow is mixed with dry air to create an air steam with specific humidity concentration for the characterization. The Fig. 8 shows the used set up.

Fig. 8. Measurement setup for the sensor array. The principle of generation of dew point in cool

box is described in DIN 4108. The ratio of dry gas and moist gas is well-controlled by using multi gas controller and mass flow controller. The mixture of gases flow over the sensor array, which was installed in hygrometer, to participate in chemical reaction. Therefore, the performance of sensor array with implementation from quantity of water vapor can be tested and determined. The Tab. 1 shows the possible parameters set in this work.

Table 1. Parameters used in the characterization of the realized sensor arrays.

Parameter Value

Humidity measuring range 10-1000 ppmv Dew point range -60 ~ -20 ˚C Ambient pressure Approx..1000 mbar Ambient temperature 24-29 ˚C Flow rate 20-150ml/min Mixture of gas 99,999 % N2

H2O < 2 ppm/mol Cool box temperature ca. -2 ˚C Operating voltage 40 V

Coating of Sensor Material

The coating solution used in this work was phosphoric acid H3PO4 with 85 % concentration. For the characterization an amount of 0,5 µL phosphoric

acid was dosed on each sensor cell. The sensor cells with coating were then cured at 50 °C for 30 min. Fig. 9 shows sensor array with coating after curing. In order to get the sensing materials better into the trenches a wetting step in O2-plasma (5 min) was conducted prio coating.

Fig. 9. Sensor array with coating ready for measurement.

For the characterization a flow rate of 50ml/min was used. The electrical voltage was 40 V. Theory and measurement results are shown in Fig. 10. The electrolytic current behavior was a linear function of absolute humidity. At low absolute humidity till approx. 100 ppm, the agreement between theory and test results was excellent. Increasing the absolute humidity increases the difference of experimental current from theoretical current, which in accord with faraday’s law. However, the range of deviation is acceptable. Discrepancy can be explained in the non-perfect setup.

Fig. 11 shows the individual currents of individual sensor cells as functions of humidity. The current of the first sensor cell is the highest due to the higher absorbed humidity. The current of the last sensor cell is low indicating a very low humidity in the air flow above the last sensor cell. It shows that a complete electrolysis of water in the air is possible with the sensor array. Same effect could be achieved by adjusting the voltages for the individual sensor cells.

Fig. 10. Comparison of results of test and theory.

Nitrogen

Cool Box

Multi Gas Controlle

Mass Flow Controller

Hygr

Sensor array

Off gas

Sensorarray

Page 103: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

103

Fig. 11. Signals of individual sensor cells.

Fig. 12. Output signals of three sensors as function of humidity concentration up to 1000 Ppm.

Many sensor arrays have been characterized in order to test the repeatability and robustness of the

technology. The Fig. 12 shows the output signals of three sensors as function of humidity. It shows a good repeatability. 4. Conclusions

We present a novel reliable silicon-based sensor solutions for trace humidity. The single sensor cell can be used to monitor trace humidity as well as the sensor array. The sensor array can be used to monitor absolute water content in the air without calibration. The measurement results are verified and show quite good agreement with theory. The sensor package has been optimized by using simulation. The realized prototype has a geometry of 96 x 50 x 25 cm3. The simulation results show some limitation within the package. Improvements have been made to get better flow behavior over first sensor cell by modifying the flow path. The long flow path could be integrated in the package and can avoid backlash from air surrounded. Acknowledgements

The authors would thank the BMWI and ZIM for founding of this project. References [1]. R. Wernecke, Industrielle Feuchtemessung, Wiley-

VCH Verlag, 2003. [2]. H.-D. Ngo, T. Weiland et. all., Packaging solution for

a novel silicon-based trace humidity sensor using coulometric method, Paper accepted for EPTC conference, Singapore, Dec. 2016.

[3]. Dew Point Air compressed air note B210991EN-B-LOW-v1.

Page 104: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

104

Combinatorial Sensing of Catalytic Materials

Using Terahertz Chemical Microscope

Yuji Hino, Yuki Kawakami, Kenji Sakai, Toshihiko Kiwa, and Keiji Tsukada Graduate School of Natural Science and Technology, Okayama University,

2 3-1-1 Tsushimanaka, Kitaku, Okayama 700-8530, Japan, E-mail: [email protected]

Summary: Terahertz chemical microscope (TCM) has been developed to visualize the electrical and/or chemical potential on the ‘sensing plate’ as the distribution of the amplitude of THz pulses radiated from the sensing plate. In this study, the combinatorial search of membrane materials using TCM was proposed and evaluation of catalytic reactions of Pt thin film was demonstrated. Pt thin films with the thickness of 4 nm, 7 nm, 10 nm and 15 nm were fabricated on the same sensing plate. Each Pt thin film on the sensing plate was exposed to the dry air (80 % nitrogen gas and 20 % oxygen gas) and hydrogen gas (1 %) and air every three minutes, independently. The amplitudes shifts of radiated THz pulses could be related to the thickness of Pt thin films on the sensing plate. This result suggests that TCM is one of useful option for the combinatorial sensing of catalytic materials. Keywords: Fuel cell, Platinum (Pt), Terahertz chemical microscope (TCM), Engineering plastic, Combinatorial search.

1. Introduction

Globally, the majority of the world energy is produced from fossil fuels. However, when fossil fuels are combusted, a significant amount of by-product such as NOX, SOX, and CO2 are produced, which causes various environmental problems, such as global warning and air pollution. In this point of view, fuel cells have been attracting attention as one of the cleanest energy since fuel cells exhaust water only during the power generation. Moreover plenty amount of hydrogen gas are expected to be produced from oceans.

The performance of fuel cells is generally determined by the combination of parameters of catalytic electrodes, e.g. the thickness, roughness and materials. So to find optimum conditions and/or materials of the catalytic electrodes, the combinatorial sensing of catalytic electrodes is required.

In our group, A terahertz chemical microscope (TCM) has been proposed and developed to visualize the electrical or chemical potentials of the catalytic electrodes on the sensing plate [1-3], which consisted of SiO2 and Si thin films on a sapphire substrate. TCM can visualize the electrical and/or chemical potential on the sensing plate as the distribution of the amplitude of THz pulses radiated from the sensing plate.

Here, Pt thin films with the thickness of 4 nm, 7 nm, 10 nm and 15 nm were fabricated on the same sensing plate and the four gas flow channels were fabricated on the Pt thin films using a 3D-printing technique, thus the catalytic reactions of Pt thin films could be measured independently.

2. Experimental

Fig. 1 show a schematic of the sensing plate. When femtosecond laser pulses irradiates the sensing plate from the substrate side of the sensing plate, the carriers

are excited and accelerated by the electric field of the depletion region in the Si substrate [4]. As the result of carrier acceleration, THz pulses are generated and radiated to free space. When the chemical reactions occurred on the sensing plate, the electric potential on the sensing plate would changes, and simultaneously, the electric field of the depletion region changes. Thus the amplitude of THz pulses is changed. Because the THz pulses are radiated at exactly where the femtosecond laser irradiates, the distribution of the amplitude of THz, which can be related to the chemical reactions, are visualized.

SiO2

Si

Sapphire

Pt Pt

Femtosecond laser pulses

THz pulses

Fig. 1. Schematic of the sensing plate. In order to evaluate the catalytic reaction of the

cataltyci electrode, Pt thin films were fabricated on the same sensing plate. Fig. 2 (a) show the top view of the sensing plate. Pt thin films with four different thickness of approximately 4 nm, 7 nm, 10 nm and 15 nm were prepared on the same sensing plate.. Fig. 2 (b) shows the cross section view of the sensing plate. The each Pt thin films were covered by the independent gas flow channels, thus, the Pt films could be evaluated independently.

Page 105: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

105

Pt4 nm

Pt15 nm

Pt7 nm

Pt10 nm

15 mm

SiO2Si

Sapphire

Pt Pt

Gas entrance

Silicone rubberEngineering plastic substrate

(a)

(b)

Fig. 2. Schematic of the sensing plate (a) the top view and (b) the cross-section view.

3. Results and Discussion

In order to evaluate the potential, each Pt thin film on the sensing plate was exposed to the 80 %-oxygen gas (O2), 1 %-hydrogen gas (H2) and the 80 %-oxygen gas (O2) with the duration of three minutes. The pressures of all gases were balanced by pure nitrogen gas.

Fig. 3 shows temporal THz amplitude change of Pt (4 nm) during the gas exposure. The femtosecond laser was fixed at beneath the Pt thin film with the thickness of 4 nm. The amplitude of THz pulses increasing rapidly at the time is 3 min, where the gases were changed from O2 to H2. This result indicates that the THz amplitude change due to the catalytic reaction of Pt thin film could be observed. The amplitude was recovered at the time is 6 min, where the gases were changed from H2 to O2. However the amplitude was not decreased to the initial level. This may simply because the hydrogen gas were partially remaining in the gas flow channels and shift the electric potential of Pt thin film.

0 3 6 911.8

12.0

12.2

12.4

12.6

12.8

13.0

TH

z a

mpl

itude

(m

V)

Time (min)

O2 H2 O2

Time (min)

THz amplitude(mV)

12.0

12.2

12.4

12.6

12.8

13.0

11.80 3 6 9

Fig. 3. THz amplitude change of Pt (4nm) during gas exposures.

Fig. 4 shows the change in the amplitude of THz pulses when the gases changed from O2 to H2 for as a function of the thickness of Pt thin films. Clear dependence between the amplitude of THz pulses and the thickness of Pt thin films were observed. The other parameters such as the roughness of Pt thin films are need to be examined. However, this result suggests combinatorial sensing of catalytic electrodes could be possible using TCM.

Further development for faster scanning of the femtosecond laser is now underway to evaluate the catalytic reaction of the all films simultaneously.

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20

THz amplitude (mV)

Thickness (nm)Thickness (nm)

THz amplitude(mV)

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20

Fig. 4. THz amplitude changes as a function of the thickness of Pt thin films on the sensing plate.

4. Conclusions

The catalytic reactions of Pt thin films integrated on the single sensing plate were evaluated using TCM. Clear dependence between the amplitude of THz pulses and the thickness of Pt thin films were observed. Evaluations of other parameters such as the roughness of the thin films are still required. However this result suggests that TCM is one of useful option for the combinatorial sensing of catalytic electrodes. 5. Acknowledgements

This work was partially supported by MEXT Subsidy of the Acceleration of a Translational Research Network Program and KAKENHI Grant Number 16H03887. 6. References [1]. T. Kiwa et al., Chemical sensing plate with a laser-

terahertz monitoring system, Appl. Opt, 47, 18, 2008, pp. 3324-3327.

[2]. T. Kiwa et al., A Terahertz Chemical Microscope to Visualize Chemical Concentrations in Microfluidic Chips, Jpn. J. Appl. Phys., 46, 41-44, 2007, pp. L1052-L1054.

[3]. T. Kiwa et al., Laser terahertz emission system to investigate hydrogen gas sensors, Appl. Phys. Lett., 86, 2005, 261102.

Page 106: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

106

[4]. T. Kiwa, S. Oka, Y. Minami, I. Kawayama, M. Tonouchi, and K. Tsukada, Measurement of pH in Fluidic Chip Using a Terahertz Chemical Microscope,

IEE J Trans. Sens. Micromach., 129, 2009, pp. 221-224.

Page 107: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

107

Amplification of Anodic Stripping Voltammetric Signals Recorded Using

Screen Printed Electrodes in Weak Magnetic Fields

A. Królicka, A. Bobrowski, M. Czarnota AGH University of Science and Technology, Faculty of Materials Science and Ceramics, Department of

Building Materials Technology, Al. Mickiewicza 30, 30-059 Kraków, Poland Tel.: +48 12 617 45 65, fax: +48 12 617 38 99

E-mail: [email protected] Summary: An anodic stripping voltammetric procedure for the determination of trace amounts of heavy metals using bismuth film electrode was elaborated. Two precursors of bismuth film – Bi2O3 and Bi(III) ions – and a favorable orientation of the magnetic field were exploited to increase the sensitivity of Zn, In and Cd signals recorded at screen-printed electrodes modified with a bismuth layer. When using this configuration and an in-situ procedure, the measured signals were better shaped (and the method was much more sensitive) than those obtained for conventional bismuth-modified screen-printed electrodes. Keywords: Screen-printed electrodes, Bismuth film electrode, Magnetic field influence on plating.

1. Introduction

Electrochemical techniques are as sensitive as

spectrometric methods and may be used for the detection of toxic metal ions in various samples. Unlike spectrometric methods, electrochemical techniques are suitable for miniaturization and on-site applications. Using microfabrication techniques (e.g. chemical or physical vapor deposition coupled with photolithography) or screen-printing techniques it is possible to produce planar electrodes for electrochemical sensors.

In this paper, the applicability of anodic stripping voltammetry (ASV) coupled with bismuth film screen-printed electrodes (BiF-SPEs) for the determination of Zn(II), In(III) and Cd(II) is demonstrated. The bismuth electroactive layer was generated via the simultaneous electroreduction of two bismuth precursors: Bi2O3 dispersed in the electrode body and Bi(III) ions spiked into the tested solution.

2. Experimental 2.1. Apparatus and Instrumentation

Electrochemical measurements were performed using an Autolab PGSTAT204 (Nova 1.10.1.9 software) potentiostat with a standard three-electrode configuration. The screen-printed electrodes (4 mm in diameter; bismuth-oxide-modified DRP 110BI) were provided by Dropsens, Spain. The magnetic stirrer (Metrohm 801) was controlled with the NOVA software. The magnet built into the stirring device induced a magnetic field of up to 86 mT, while the field induced by the 7 mm long stirring flea was up to 3 mT. The voltammetric experiments were performed in the cuboid (50x50x29 mm) polystyrene vessel.

2.2. Procedure

Prior to the experiments the SPE electrode was immersed in the tested solution for 1 minute. The supporting electrolyte contained 0.08 M CH3COONH4, 1.5 mg l-1 Bi(III), 0.26 g l-1 NaCl, and had a pH of 6.3. The accumulation potential of -1.5 V was then applied for 300 s and the DP-ASV curve was recorded within the range of -1.5 V to 0.3 V. During accumulation, the solution was stirred using a magnetic bar. The parameters of the DPV mode were as follows: E = 25 mV, Es = 5 mV, teq = 5 s. 3. Results and Discussion

The commercial SPEs modified with bismuth oxide were chosen as a sensor for anodic stripping voltammetric studies. Bismuth oxide may be electrochemically converted to bismuth crystallites in highly alkaline media and in acidic media after the application of a negative potential. The ASV peaks recorded using this electrode were rather low, wide and often distorted by undesirable shoulders. After the addition of Bi(III) into the tested solution the geometry of peaks was far more defined [1] and this double-precursor-electrode was therefore selected as suitable for further experiments. The ASV signal of examined metals was improved by another 200 % - 600 % after adjusting the SPE location with the respect to magnetic field. Three factors were crucial: (v) vertical and (h) horizontal distance between the magnet center and the sensing area of SPE and, finally, the angle () between the magnet surface and the SPE strip (Fig. 1). Effect of different SPE arrangements on geometry of voltammograms of zinc, indium and cadmium is shown in Figs. 2–4.

Page 108: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

108

(a)

(b)

Fig. 1. (A) side and (B) top view of SPE arrangement. M – magnet.

0 10 200.0

0.1

0.2

Pea

k ar

ea

vertical distance / mm

Fig. 2. The influence of vertical location of the SPE on the zinc signal. h = 0.8 cm, = 0°, cZn(II) = 20 g l-1.

-1.0 -0.80

5

10

15

E / V

0 45 90 135 180

I /

A

A B

-1.0 -0.80

5

10

15

E / V

I /

A

Fig. 3. DP-ASV curves recorded in two horizontal SPE positions: (A) h = 0.8 cm, (B) h = 1.6 cm (B) and different

values. v = 0 cm, cIn(III) = 50 g l-1.

Increasing of the distance between the magnet and the SPEs’ surface both in vertical (Fig. 2) and horizontal directions (Fig. 3) caused a decrease in peak currents. The influence of the angle was more complex and depended on the SPE localization and target element. In the case of indium, when the SPE was immersed in the location expressed by v = 0 cm and h = 0.8 cm, did influence the In signal intensity

but only to a small degree (up to about 15%). If the location of the SPE was more distant (h = 1.6 cm), the 180 degrees rotation of the SPE produced a large 60% drop in the signal intensity. Unlike indium, zinc signal was almost not affected by the rotation of SPEs. Additionally, the SPE rotation angle can influence not only the signal intensity but also its symmetry. The SPE’s performance in terms of cadmium signal symmetry is highest at central locations. At farther, more peripheral locations, the SPE’s performance decreased and cadmium signals were wide and distorted by shoulders (Fig. 4B, 4C). Not only cadmium signals but also bismuth peaks became considerably wider at peripheral locations of SPE (Fig. 4D). The signal splitting indicates that crystallites formed at the peripheral SPEs locations during accumulation are non-identical and, therefore, different energy must be delivered to oxidize them. Furthermore, the comparison of oxidation signals shown in Fig. 4D, representing cadmium and bismuth oxidation, reveals that it is not the amount of bismuth which is crucial to obtaining well-shaped signals of cadmium, but, instead, the favourable morphology of bismuth deposits is indispensable. As shown in Fig. 4D when the oxidation signal of Bi is smallest at same time the cadmium signal is obviously highest.

-1.2 -1.0 -0.8

1

2

3

0 45 90 135 180

I /

A

E / V

Cd

-1.2 -1.0 -0.8

1

2

I /

A

E / V

CdA B

-1.2 -1.0 -0.80.5

1.0

1.5

2.0

I /

A

E / V

Cd

-1.6 -1.2 -0.8 -0.4 0.00

20

40

D

I /

A

E / V

a)

b)

c)

Cd

Bi

C

Fig. 4. DP-ASV curves recorded in three SPE positions: (A) h = 0.8 cm, v = 0.75 cm, (B) h = 0.8 cm, v = 1.5, (C)

h = 1.6 cm, v = 2.2 cm and different values. (D) voltammograms for = 0°, cCd(II) = 50 g l-1.

4. Conclusions

The dependence of Zn, In and Cd signal on the location of SPE strip in the magnetic field may be explained by assuming that in a favorable location of the SPE, the following:

Page 109: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

109

1) Transport of electroactive substances – Bi(III) that forms the films and low-temperature melting alloys with the determined metals is facilitated, or

2) The forces acting in the magnetic field prevent bismuth crystallites from leaving the surface of the electrode and at the same time affect orientation of crystallites relating to the surface of the electrode.

The results of investigations concerning the effect of the magnetic field on the electrodeposition of bismuth show that it is a very complex process [2] and it is difficult to unambiguously indicate which processes are responsible for the observed amplification of Zn, In and Cd signals. In addition, the majority of the effects reported in the published papers were observed for static magnetic fields of high strength (>1T), but in the case of our experiments a weak, rotating field (up to 86 mT) induced by the permanent magnet was applied.

Acknowledgements

Financial support from the Polish National Science Centre (Project 2014/15/B/ST8/03921) is gratefully acknowledged. References [1]. A. Królicka, A. Bobrowski, Employing a magnetic

field to amplify zinc signal obtained at bismuth film screen-printed electrodes generated using dual, Electrochimica Acta, Vol. 187, Issue 1, 2016, pp. 224-233.

[2]. M. Uhlemann, K. Tschulik, A. Gebert, G. Mutschke, J. Frohlich, A. Bund, X. Yang, K. Eckert, Structured electrodeposition in magnetic gradient Fields, The European Physical Journal Special Topics, Vol. 220, Issue 1, 2013, pp. 287–302.

Page 110: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

110

Sensor Development of Electronic Tongue System for Taste Identification

Rohini Mudhalwadkar

University Pune, Instrumentation and Control department, College of Engineering Pune, Pune, India E-mail: [email protected]; [email protected]

Summary: An attempt in the presented work is focused on development of indeginous and low cost sensors for electronic tongue system. The work is focused on an intelligent system for taste discrimination. The sensor array design is based on the conductivity measurement in the liquid samples. An intelligence in the system is included with an interface with LabVIEW (NI) and algorithm development for data treatment and taste comparison. The electronic tongue system is validated for food processing applications. Keywords: Sensor, Electronic Tongue, LabVIEW.

The electronic tongue system is developed on the basis of electrical conductivity measurement of sample. The sensor design and fabrication is shown in the Fig. 1 and Fig. 2 respectively. The sensor design is carried out with different materials such as Silver-Silver, Copper-Silver, Zinc-Silver, Lead-Silver and fabricated in PVC pipe, which makes it indigenous and low cost design.

(a) silver (b) copper

(c) Zinc (d) lead

Fig. 1. Sheets of metals used for sensor development.

Fig. 2. Photograph of designed sensors.

AC signal with frequency 1 to 3 kHz , which is generated by wein bridge is applied to the developed sensor array system. The developed sensor array is immersed into the sample solution, which results into an electrical current. The current is directly proportional to conductivity of the sample. Data acquisition is carried out with NI 9205 DAQ card and analysed with LabVIEW(NI) .The complete setup is shown in the Fig. 3.

Fig. 3. Photograph of the developed set up.

The samples are initially tested with Wavered conductivity analyser is used. The electrical conductivity results are used to understand the nature of the change and conductivity of the samples. The results are used for data validation. Fig. 4 and 5 show the results obtained by conductivity meter and developed sensor system. Both of the graphs shows similar nature, which is used as index for salty taste.

Fig. 6 shows the conductivity analysed with sweet taste sample such as glucose and Fig. 7 shows the sweet taste index with the developed system.

Page 111: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

111

Fig. 4. Measured Conductivity of NaCl with different

concentrations.

Fig. 5. Salty taste index with the developed system.

Fig. 6. Measured Conductivity of Glucose. Fig. 7. Taste index measurement for sweet taste by developed sensor system.

Conclusions

As seen from the experimental results, it is observed that the concentration of solution with the particular taste substance is proportional to conductivity. The conductivity of NaCl under same concentration is more than those of other standard solutions due to its dissociation property. The conductivity of the solutions is distinguished, the conductivity factor can be used as an index for the taste measurement. The sensor fabricated are very low in cost and easy in operation. The developed system can be able to work as an intelligent system and can be used for online applications. References

[1]. M. Ito, K. Wada, M. Yoshida, K. Wada, and T. Uchida,

Quantitative evaluation of bitterness of H1-receptor antagonists and masking effect of acesulfame potassium, an artificial sweetener, using a taste sensor, Sens. Mater., Vol. 25, No. 1, pp. 17–30, 2013.

[2]. J. Chandrashekar, K. L. Mueller, M. A. Hoon, E. Adler, L. Feng, W. Guo, and N. J. Ryba, T2Rs function as bitter taste receptors, Cell, Vol. 100, No. 6, Mar. 2000, pp. 703–711.

[3]. Y. Kobayashi, M. Habara, H. Ikezazki, R. Chen, Y. Naito, and K. Toko, Advanced taste sensors based on artificial lipids with global selectivity to basic taste qualities and high correlation to sensory scores, Sensors, Vol. 10, No. 4, Apr. 2010, pp. 3411–3443.

[4]. Legin, A. Rudnitskaya, Y. Vlasov, C. Di Nataleb, F. Davideb, and A. D’Amicob, Tasting of beverages using an electronic tongue, Sens. Actuators B Chem., 44, 1–3, Oct. 1997, pp. 291–296.

[5]. G. Verrelli, L. Lvova, R. Paolesse, C. Di Nataleb, and A. D’Amicob, Metalloporphyrin—Based electronic tongue: An application for the analysis of Italian white wines, Sensors, Vol. 7, No. 11, Nov. 2007, pp. 2750–2762.

[6]. Gardner and P. Bartlett, Sensors and Sensory Systems for an Electronic Nose, Kluwer, Boston, MA, USA, 1992.

[7]. P. M. Schweizer-Berberich, S. Vaihinger, and W. Göpel, Characterisation of food freshness with sensor arrays, Sens. Actuators B, Chem., 18, 1–3, Mar. 1994, pp. 282–290.

Page 112: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

112

Analysis of Cocaine Using a Chemically Modified Electrode

with Vanadium Hexacyanoferrate film by Cyclic Voltammetry

I. C Eleotério 1,2, M. A Balbino 1,2, J. Magalhães 1, A. S Castro 1, B. R McCord 2, Marcelo F Oliveira 1

1Universidade de São Paulo/ FFCLRP, Departamento de Química, Av. Bandeirantes, 14040901, Ribeirão Preto, SP, Brazil

2Florida International University, Chemistry and Biochemistry Department, Sw 8th St, 33199, Miami, Fl, USA

E-mail: [email protected] Summary: Hexacyanoferrates film comprises a viable strategy to develop electroanalytical methods regarding illicit drug analysis. The material includes many advantages such as a wide potential window, active redox low film thickness, preparation versatility, and low cost. With this in mind, we selected this electrode to determine cocaine abuse. More specifically, we studied the electrochemical behavior of cocaine by cyclic voltammetry. The anodic peak current at 0.83 V vs. Ag / AgCl (irreversible oxidation peak) increased linearly with the cocaine rise; the cocaine electrochemical oxidation occurred at the tertiary amino group .That electrode is a potentially suitable transducer in voltammetry to determine cocaine in confiscated samples. Keywords: Cocaine, Vanadium hexacyanoferrate, Cyclic voltammetry, Illicit drugs, Electroanalytical methods.

1. Introduction

Electroanalytical methods offer several advantages on the illicit drug analysis: they are versatile, fast, sensitive, inexpensive, and environmentally friendly (they involve limited use of chemicals). Recently developed, electrochemical devices efficiently identify drug of abuse through direct or indirect reactions between the substance and the electrode surface, which makes them potentially applicable in situ [1-3]. The large-scale use of this technology relies on scientific knowledge. Nowadays, many electrodes to develop this methodology is investigated [1-4]. Platinum disc electrode, carbon paste electrodes, glassy carbon electrode, chemically modified electrodes with film hexacyanoferrates are some of the examples reported in the literature [4]. The hexacyanoferrates film offers a broad potential window with redox active thin films, low cost, and versatile preparation [2-4]. Among these drugs has cocaine (Fig. 1), an illegal drug, which is used mainly alkaloid of Erythroxylum coca plant found in countries in South America such as Colombia, Peru, and Bolivia. The cocaine seizures over the years aroused the interest of forensic science in the investigation of the composition of this drug. Some drugs are chosen, whose purpose is to enhance the effects, two classic examples of adulterants more used for presenting stimulating effects and anesthetics are lidocaine and caffeine, respectively [2, 3].

2. Experimental 2.1. Reagents and Instrumentation

A stock solution of cocaine 0.001 mol L-1 (Sigma-Aldrich) was prepared in methanol (Merck). The film was made by mixing 0.01 mol L-1 potassium ferrocyanide with 0.01 mol L-1 ammonium

metavanadate 0.01 mol L-1 (Vetec®), and 3.6 mol L-1 sulfuric acid (Sigma-Aldrich). The analysis of cocaine was used 0.1 potassium sulfate (Vetec®) with 3.6 mol L-1 sulfuric acid (Sigma-Aldrich) such supporting electrolyte. The voltammetric experiments were conducted on a potentiostat model μAUTOLAB III (Eco Chemie) connected to a personal computer. The experiments were carried out in a three-electrode system. They consisting of a glassy carbon working electrode, platinum spiral wire auxiliary electrode, and as Ag/AgCl reference electrode arranged in a 5-mL electrochemical cell (Fig. 2), the solutions were deaerated with nitrogen for 15 min before voltammetric analysis.

Fig. 1. Chemical structures of the cocaine. 2.2. Voltammetric Analysis

The cyclic voltammetric measurements were conducted in the potential window from 0.45 to 1.5 V (vs. Ag/AgCl) ), at a scan rate of 100 mV s-1. Cocaine analysis at different concentrations was accomplished by the standard addition method. To this end, aliquots of the cocaine stock solution (from 10 to 30 µL) were added, to the electrochemical cell.

Page 113: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

113

Fig. 2. Electrode arrangement. 3. Results and Discussion

We were able to deposit the vanadium hexacyanoferrate film on a glassy carbon electrode by conducting consecutive cyclic voltammetry cycles at applied potentials ranging from 0.45 to 1.2 V (Fig. 3).

Fig. 3. Vanadium hexacyanoferrate film formation at the GC electrode.

We investigated the cyclic voltammetry behavior of the film on cocaine in terms qualitatives by successive additions of 10-µL aliquots of the standard cocaine solution (Fig. 4). The anodic peak current at 0.83 V vs. Ag / AgCl (irreversible oxidation peak) increased linearly upon rising cocaine; According to the literature, the electrochemical oxidation of cocaine occurs at the tertiary amino group [1, 2].

Fig. 4. Influence of cocaine concentration on the voltammetric response of the vanadium

hexacyanoferrate. 4. Conclusions

The glassy carbon electrode modified with a film of vanadium hexacyanoferrate has an electrocatalytic effect on cocaine oxidation. This electrode is a potentially suitable transducer in voltammetry to determine cocaine in confiscated samples. Acknowledgements

We would like to thank CAPES Foundation (Edital Pró-Forenses 25/2014) and FAPESP for financial support. References [1]. M. T. F. Abedul, J. R. B. Rodriguez, A. C. Garcia, P.

T. Blanco, Voltammetric determination of cocaine in confiscated samples, Electroanalysis, 3, 1991, pp. 409-412.

[2]. L. S. Oliveira, M. A. Balbino, M. M. T. Menezes, E. R. Dockal, M. F. Oliveira, Voltammetric analysis of cocaine using platinum and glassy carbon electrodes chemically modified with uranil Schiff base films, Microchemical Journal, 110, 2013, pp. 374-378.

[3]. L. S. Oliveira, A. P. P. Santos, M. A. Balbino, M. M. T. Menezes, J. F. Andrade, E. R. Dockal, H. M. Tristão, M. F. Oliveira, Voltammetric Determination of Cocaine in Confiscated Samples Using a Carbon Paste Electrode Modified with Different [UO2(X-MeOsalen)(H2O)]•H2O Complexes, Sensors, 13, 2013, pp. 7668-7679.

[4]. I. L. de Mattos, L. Gorton, Filmes de metal-hexacianoferrato: uma ferramenta em química analítica, Química Nova, 24, 2001, pp. 200-205.

Auxiliary electrode

Electrochemical cell (5 mL)

Reference electrode

working

Page 114: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

114

Multifunctional Sensor with Frequency Output Based on SOI TFT

Double-gate Sensing Element

V. N. Mordkovich 1, A. V. Leonov 1, A. A. Malykh 1,2 and M. I. Pavluyk 3 1 Institute of microelectronic technology and high purity materials RAS, Laboratory of Radiation Stimulated

Processes, b 6 Akademika Osipyana street, 142432, Chernogolovka, Moscow region, Russia 2 National University of Science and Technology “MISIS”, College of New Materials and Nanotechnologies,

b 4 Leninskiy prospekt, 119049, Moscow, Russia 3 ICC Milandr, JSC, b 5 Georgievsky prospekt, 124498, Moscow, Zelenograd, Russia

Tel.: + 74959815433, fax: + 74959815436 E-mail: [email protected]

Summary: The multifunctional microelectronic sensor with frequency output is considered. Primary transducer of this sensor is able to register different impacts. The presented sensor uses a scheme of bistable multivibrator. Sensitivity of the sensor is about 2.2 kHz/mT and 1.4 kHz/ºС for magnetic field and temperature measurements respectively. SOI primary transducer allow sensors based on it to have the extended operating temperature range. Keywords: SOI TFT, Double gate sensing element, Frequency output sensor, Extended temperature range.

1. Introduction

In a recent time, sensors with frequency output

(frequency-output sensor - FOS) are actively developing. The frequency form of signal provides the possibility of noise proof wireless connection with functional equipment units. Besides this type of signal actually represents the consequent digital code and possesses the advantages of digital systems.

In our work the multifunctional microelectronic FOS is considered. The primary transducer of this sensor is able to register different impacts, to control the operating frequency and provide the opportunity of creating sensors with an extended range of operating temperatures.

2. Double-gate Field-effect Sensing Element

The sensing element represents the double gate TF MOSFET with built-in n+-n-n+ channel and MISIM field effect control system integrated with traditional Si Hall element in the same structure [1]. Further it is denoted as double-gate field-effect sensing element – 2GFSE (Fig.1). It’s formed in a thin Si layer (500×500×0.2 μm) of the SOI structure with electron concentration 5·1014 cm-3. Both of the gate SiO2 oxides are 350 nm thick. The additional n+ contacts for Hall-effect measurements are located on the opposite lateral sides of n-Si operating layer. The main operating regime of the 2GFSE is the regime of electron accumulation near the Si-SiO2 interfaces. A unique peculiarity of the 2GFSE is the ability to register different external impacts, that are in any way affect the behavior of the flow of electrons in the channel (e.g. magnetic field due to the Hall-effect, the temperature and pressure due to changes in mobility of electrons, ionizing radiation due to accumulation of the positive charge and the corresponding change of

conductivity of accumulated areas in Si film). In this work 2GFSE is the basis of the multifunctional FOS of the magnetic field and temperature.

Fig. 1. Schematic illustration of 2GFSE’s structure. Vd – the supply voltage; Vg1, Vg2 – the voltage of the gates; Vh – the

induced Hall voltage. 3. Multifunctional Sensor with Frequency

Output

Multifunctional FOS with 2GFSE uses the scheme of bistable multivibrator [2] (Fig. 2). Consequent measurement of magnetic field and temperature is performed by the switch contacts of the 2GFSE using the K1-K4 electronic keys. The closure of the 1, 4, 7, 10 contacts implements a mode of the magnetic field measurement (the Hall-contacts of the 2GFSE are used), and the closure of the 2, 3, 5, 6, 8, 9 contacts implements a mode of the temperature measurement (using the contacts of the source and the drain, the 2GFSE is used as the thermal transistor).

In this FOS the induced by external impact signal leads to change of the gate voltages of FETs T5 and

Page 115: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

115

T6. In turn, this leads to a redistribution of the currents of BJTs T1-T4 and the rate of charge-discharge capacitors C1-C3 of the multivibrator, causing the shift of the operating frequency.

Fig. 2. Scheme of the multifunctional FOS on the base of the 2GFSE (both gates are under the same voltage Vg).

Field management in 2GFSE allows the change the operating frequency of the FOS with a slope near 20 kHz/V. In case of using 2GFSE as a Hall-type sensing element it is carried out due to the dependence of residual voltage on the Hall contacts (in the absence of a magnetic field) on the electrical operating modes of the sensing element. In case of using 2GFSE as a thermal transistor, which forms a voltage divider together with resistor R10, due to the dependence of the channel resistance of the 2GFSE on electrical operating modes of it.

The Fig. 3 shows that the dependences of the magnetic field and the temperature conversion into the changing of frequency of the FOS are linear. The sensitivity of magnetic field measurements can achieve 2.2 kHz/mT. The sensitivity of temperature measurements can achieve 1.4 кHz/ºC in the temperature range 25-150 ºС.

Significantly that 2GFSE is able to keep functionality under the extremаlly low temperatures (LHT) and extramally high (at least 375 ºC) temperatures [3]. This is not possible for conventional Si analogs. Thus the construction of the microelectronic FOS with separately located 2GFSE provides the opportunity of different impacts measurements in a wide temperature range. 4. Conclusions

The multifunctional microelectronic sensor with frequency output is considered. Primary transducer of this sensor is able to register different impacts and to control the operating frequency with 20 kHz/V slope.

The presented sensor uses the scheme of bistable multivibrator. Sensitivity of the sensor is about 2.2 kHz/mT and 1.4 kHz/ºС for magnetic field and temperature measurements, respectively. Wherein in case of magnetic field measurements the residual voltage between Hall-contacts, which is usually compensated by the circuit design or through the introduction of mathematical correction, used to control the operating frequency of the FOS.

Fig. 3. Sensory characteristics of multifunctional FOS on the base of 2GFSE. a) Magnetic field measurements; b)

Temperature measurements. Acknowledgements

This work was done with support of the Ministry of Education and Science of the Russian Federation (Agreement 14.576.21.0026, UI: RFMEFI57614X0026). References [1]. M. Baranochnikov, A. Leonov, V. Mordkovich, D.

Pazhin, M. Filatov, Some Features of Magnetometric and Sensors Devices Based on The Field Effect Hall Sensor, in Proceedings of the Applied Electromagnetic Symposium, Paris, France, 16-19 April 2012, pp. 455-459.

[2]. A. V. Leonov, A. A. Malykh, V. N. Mordkovich, M. I. Pavlyuk, Sensors with SOI FET primary transducer and frequency output, in Proceedings of the 17th International Conference on Sensors and

Page 116: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

116

Measurement Technology (SENSOR’15), Nuremberg, Germany, 19-21 May 2015, pp. 864-867.

[3]. A. V. Leonov, A. A. Malykh, V. N. Mordkovich, M. I. Pavlyuk, Field Controlled Si Hall Element with

Extended Operation Temperature Range from Liquid Helium Temperature up to 650K, Procedia Engineering, 120, 2015, pp. 1197-1200.

Page 117: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

117

Terahertz-Time-of-Flight Method for Evaluating Cosmetic Samples

Penetrating into Skin Samples

T. Arisawa, T. Morimoto, K. Sakai, T. Kiwa and K. Tsukada Graduate School of Natural Science and Technology, Okayama University,

3-1-1 Tsushimanaka, Kitaku, Okayama 700-8530, Japan, E-mail: [email protected]

Summary: A terahertz(THz)-time-of-flight(TOF) method was carried out for in-vitro measurements of the penetration depth of cosmetic liquid into skin samples. In our group, a sensing plate, which can generate terahertz pulses, has been developed. By irradiating the back side of the sensing plate using a femto-second laser pulses, the THz pulses with the pulse width of approximately 1 ps were radiated into a free space. When the artificial skin samples were put on the sensing plate, the generated THz pulses travel in the skin samples and are reflected at the boundary between the regions where the cosmetic liquid exists and does not exist. Since traveling time of the THz pulses are generally proportional to the traveling path, the distance between the boundary and the sensing plate can be estimated by evaluating the traveling time of the THz pulses. Here, as demonstration of the developed method, the penetration of liquid into the artificial dermis made from collagen were evaluated. Keywords: Terahertz, Time-of-flight method, in vitro measurement, Cosmetic, Femtosecond laser.

1. Introduction

The cosmetics including nano-particles are recently attracted since these kinds of cosmetics are generally well penetrated into the skin and enhance surface activity of the skin.

In order to evaluate the efficiency of cosmetic liquid, measurement of the penetration speed and/or penetration depth of the cosmetic liquid into the skin are important. Previously, this kind of measurements had been done using living animals. However, from the ethical point of view, evaluation of ‘not living animals’ such as artificial skin samples and cultivated cell sheets are now recommended.

In our group, a terahertz-time-of-flight method [1] has been applied to evaluate the penetration of liquid into ‘artificial’ samples.

In order to generate THz pulses, a sensing plate has been developed [2-5]. By irradiating the back side of the sensing plate using a femto-second laser pulses, the THz pulses with the pulse width of approximately 1 ps were radiated into a free space. When the artificial skin samples were put on the sensing plate, the generated THz pulses travel in the skin samples and are reflected

at the boundary between the regions where the cosmetic liquid exists and does not exist. Since traveling time of the THz pulses are generally proportional to the traveling path, the distance between the boundary and the sensing plate can be estimated by evaluating the traveling time of the THz pulses. Here, as demonstration of the developed method, the penetration of liquid into the artificial dermis made from collagen were evaluated.

2. Method

Fig. 1 shows the principle of our methods. The artificial skin was put on the sensing plate and the femto-second laser with the pulse width of approximately 100 fs were focused into the sensing plate. The sensing plate were fabricated by depositing Si and SiO2 films on the sapphire substrate. As the result of the laser irradiation, the carriers in the Si films were excited and accelerated by the dilation field in the Si films, thus the THz pulses with the pulse with of approximately 1 ps was generated in the Si film.

Fig. 1. Schematic of the sensing plate and the principle of TOF.

Page 118: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

118

Some of the part of the generated THz pulses were reflected at the boundary between the sensing plate and the artificial skin and observed (T1 pulses). The other THz pulses traveled in the artificial skin and were reflected at the boundary between the regions where the cosmetic liquid exists and does not exist (T2 pulses). Because T2 pulses has longer traveling path than the T1 pulses, the peaks of T2 pulses are observed after T1 pulses in time-domain waveforms. So by evaluating the difference between T1 and T2 peaks in time-domain waveform, the distance between the sensing plate and the boundary of the penetrated liquid can be estimated.

3. Results

Fig. 2 shows the time-domain waveforms of THz pulses from the sensing plate. As the artificial dermis made from collagen of the calf was used. The thickness of the skin was 3 mm. For the liquid samples, ion-exchange water was applied and dropped on the artificial skin with the different amounts. We have confirmed that the penetration depth of water droplet into the artificial skin was roughly proportional to the amount of dropped water. Strong T1 peak followed by T2 peak was observed. Because the boundary of liquid was not uniform and the refractive index of artificial skin generally have deviation in frequency domain, T2 peak was distorted.

In order to clarify the shift of T2 peak, the positons of T1 pulses at approximately 17 ps and T2 pulses at approximately 38 ps as a function of the amount of water droplet are summarized in Fig. 3. While the positions of T1 peaks were independent of the amount of water droplet, the positions of T2 peaks were shifted. This result indicate that T1 pulses were reflected at the boundary between the sensing plate and

the artificial skin where the boundary position was not changed by the amount of the water droplet, and T2 pulses were reflected at the boundary of liquid where the position of boundary is moved by the amount of the water droplet. Thus, the penetration depth could be evaluated by measurement the position difference between the T1 and T2 peaks in THz-time-domain waveform.

Fig. 2. Time-domain waveforms of THz for the artificial skin with water droplets.

Fig. 3. Positions of peaks T1 and T2 as a function of the amount of water droplet.

4. Conclusions

In this study, the penetration depth of water droplets into the artificial skin was measured using THz-TOF method. The difference in peak positions between T1 and T2 clearly showed. THz-TOF method is one of useful option for evaluating cosmetic liquid.

Acknowledgements

This work was partially supported by Industry- Academia Collaborative R&D from the Japan Science and Technology Agency (JST), and MEXT Subsidy of the Acceleration of a Translational Research Network Program.

Page 119: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

119

References [1]. J. Obradovic, et al., The use of THz time-domain

reflection measurements to investigate diffusion in polymers, Polymer, Vol. 48, Issue 12, 2007, pp. 3494-3503.

[2]. T. Kiwa, et al., Chemical sensing plate with a laser-terahertz monitoring system, Applied Optics, Vol. 47, Issue 18, 2008, pp. 3324-3327.

[3]. T. Kiwa, et al., A terahertz chemical microscope to visualize chemical concentrations in microfluidic

chips, Japanese Journal of Applied Physics, Vol. 46, Issue 11L, 2007, pp. L1052.

[4]. K. Akimune, et al., Multi-ion sensing of buffer solutions using terahertz chemical microscopy, Applied Physics Express, Vol. 7, Issue 12, 2014, pp. 122401.

[5]. T. Kuwana, et al., Label-free detection of low-molecular-weight samples using a terahertz chemical microscope, Applied Physics Express, Vol. 9, Issue 4, 2016, pp. 042401.

Page 120: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

120

UWB Sensor Based Localization of Person with the Changing

Nature of His/Her Movement

Dušan Kocur 1, Daniel Novák 1 1 Technical University of Košice, Department of Electronics and Multimedia Communications,

Park Komenského 13, 041 20 Košice, Slovakia Tel.: + 421 55 602 4233, +421 55 602 4234

E-mail: [email protected], [email protected] Summary: This paper deals with the problem of person detection, localization and tracking using UWB radar based on the nature of his movement. Here, an insight into the performance and operation of the methods is given. Time-based methods are used for the detection of the observable movement and spectral methods are used for the detection of the breathing.

Keywords: UWB radar, Static target, Moving target, Target detection, Target localization, Target tracking.

1. Introduction

At the beginning of the 21th century, the human society is facing quit a number of specific social trends. The increasing density of population in towns and town agglomerations, criminality growing and political tensions producing terrorism can be ranked among them.

Taking into account these facts the detection and positioning of human beings is very interesting especially for military and security operations. Here, reservoirs, power plants, and other critical infrastructures are extremely vulnerable to terrorist attack. Therefore, the request for monitoring of these critical environments and for the detection of unauthorized intrusion is still needful. At these events, the knowledge about the number of persons and their position in the operational area can be very useful for military or security teams to take the right decisions [1]. In the outlined operations of law enforcement troops, the persons to be detected and localized are often situated behind an obstacle (e.g. wall). That is the reason why conventional optical and infrared sensors cannot be applied for human beings localization and tracking.

As through-the-obstacles seeing sensors, short range high-resolution radars emitting electromagnetic waves with ultra-wide frequency band (UWB radars) using relatively low frequencies can be used with advantage [2]. Here, the ultra-wide frequency band provides the fine resolution of the radar systems. On the other hand, the electromagnetic waves emitted in the frequency band DC-5GHz can penetrate through standard non-metallic materials with acceptable attenuation. Therefore UWB radars exploiting this frequency band are capable to detect not only the targets located line-of-sight, but also the target situated behind a non-metallic obstacle.

2. Problem Statement

There are two fundamental approaches for human beings localization. The former is intent on the

localization of so-called moving persons, i.e. the persons moving within the monitored area in such a way that their co-ordinates are changing. In this case, the persons are detected based on observations of the time changes of the impulse responses scanned by the radar along the observation time (slow-time). The latter approach is devoted to the localization of so-called static persons, i.e. the persons situated but not moving (e.g. unconscious persons) within the monitored area (their co-ordinates are not changing). Respiration or heart beating could be given as the example of the mentioned motion activities of that kind. In this case, the persons can be localized based on the detection of their vital signs such as respiration or heart beating.

However in real-life scenarios, persons to be located are usually moving through monitored area with some stops. Therefore, the same person may be once regarded as static and once again as a moving person. Taking into account these facts, we will introduce in this paper UWB sensor signal processing method capable to localize a person for such scenarios (i.e. a person changing the nature of her/his motion).

3. Solution Our approach for the localization of a person

changing the nature of her/his motion will consist of two procedures of UWB radar signal processing. The former procedure is intent on moving person localization (Procedure 1), whereas the latter is dedicated to static person localization (Procedure 2). Both procedures consist of a set of signal processing phases such as signal preprocessing, background subtraction, target detection, time-of-arrival (TOA) estimation and target localization [3], [4]. The first signal processing procedure devoted to moving person localization contains also the target tracking phase following a phase of localization [3]. The essential difference between these procedures consists in the detection phase. In the first procedure, the detection is based on the application of a simple CFAR detector,

Page 121: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

121

where the decision statistics are formed from the samples of the processed impulse response. On the other hand, a two-stage detector (combination of CFAR detector and a constant threshold detector) is used for the static person detection. As the decision statics, signal energy of radar signals corresponding to a constant bistatic range has been used. Their estimate can be obtained by the application of Welch periodogram method [4]. Finally, the target position is obtained as the fusion of the results provided by the both signal processing procedures.

4. Experimental Results The performance of the outlined approach for

localization of a person changing the nature of his/her movement will illustrated by through-the-wall localization of a person moving with a stop at the position P3 in Fig. 6. For the measurement, the Msequence UWB sensor equipped with one transmitting (Tx) and two receiving (Rx) channels (antennas) has been used. The obtained results are given in Fig. 1 to Fig. 6. These results indicate, that the proposed approach has a potential to provide a quit efficient localization of a person moving with changing of nature of his/her motion.

Fig. 1. Radargram with subtracted background (Procedure 1 & 2, Rx1).

Fig. 2. TOA estimation and trace connection

(Procedure 1).

Fig. 3. Target track estimation for his movement from position P1 through P2 to P3 (Procedure 1).

Fig. 4. Welch periodogram estimation (Procedure 2, Rx1).

Fig. 5. Detector statistics and CFAR treshold

(Procedure 2, Rx1).

Fig. 6. Tracking of target similar to Fig. 3 with

localization of static person (joint Procedure 1 & 2).

Page 122: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

122

5. Conclusions

In this paper, we have dealt with the signal processing procedure for moving person by UWB sensors. Here, we have been focused on a special scenario, where the localized person can be considered as moving and static person as well. The obtained results have indicated the proposed approach can provide an efficient location of person for such scenarios.

References

[1]. Withington, P. et al., Enhancing Homeland Security with Advanced UWB sensors, IEEE Microwave Mag., Vol. 4, No. 3, 2003, pp. 51–58.

[2]. Sachs, J., Handbook of Ultra-Wideband Short-Range Sensing, Wiley-VCH, January 2013.

[3]. Kocur, D.- Rovňáková, J.- Švecová, M., Through Wall Tracking of Moving Targets by M-Sequence UWB Radar. In: Rudas, I. J.; Fodor, J.; Kacprzyk, J. (Eds.) Towards Intelligent Engineering and Information Technology, Springer, 2009; pp. 349–363.

[4]. Novák, D.-Kocur, D.: Multiple Static Person Locali-zation based on Respiratory Motion Detection by UWB Radar, in Proceedings of the 26th IEEE International Conference Radioelektronika’16, 2016, pp. 252-257.

Page 123: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

123

3D Position Estimation with Capacitive Sensors for Touchless Interaction

L. Haslinger and B. G. Zagar

Johannes Kepler University Linz, Institute for Measurement Technology, Altenberger Straße 69, 4040 Linz, Austria

Tel.: +43 732 2468 5923, fax: +43 732 2468 5933 E-mail: [email protected]

Summary: This contribution deals with preliminary studies for touchless interaction based on capacitive sensor technologies. An existing approach in combination with additional capacitance measurements is used for position estimation. For that a measurement system is utilized which is able to measure capacitances in a range of a few fF to several pF. The main focus is the 3D position estimation of a human finger through spatial capacitance distributions caused by its movements. A grounded metallic test object is used as an abstract model of a finger. The capacitance modeling is based on a 2D simplification by considering properties of symmetry. This new model, combined with complex data processing algorithms, should ensure precise finger gesture recognition in future research. Keywords: Modeling, 3D position estimation, Capacitive proximity sensors, Touchless interaction, Electric field studies. 1. Introduction

Numerous scientific publications in the context of touchless interaction, which are based on capacitive technologies, deal with machine learning methods [1, 2]. For an implementation of these algorithms empirical approximations, and a set of training data are often sufficient [3].

In contrast, this contribution uses a new approach [4] that leads to a deeper physical understanding and more accurate algorithms for touchless interaction. Based on the proposed capacitance model 3D position estimation can be implemented. Subsequently, tracking a sequence of positions should enable gesture recognition in future work. Here, a moderate spatial resolution of the measurement system is sufficient if the time resolution is high enough to recognize smooth motions. The small capacitance values and the rapid capacitance changes caused by finger movements render the system sensitive to interferences. Concerning this matter the measurement and online evaluation of the data is a huge challenge and already partly discussed in [5]. Furthermore, variety in finger forms as well as numerous possible environmental conditions and disturbances are not considered. The main target regarding touchless interaction is to recognize the user intention. In the case of incorrect interpretation an active feedback can inform the user how his behavior must be changed for a clear recognition.

Fig. 1 shows a schematic of the measurement set-up consisting of a sensing electrode with an area as × as, a shielding electrode and a human finger. The surrounding volume is assumed as free space with the permittivity 0. For measurements, a grounded brass cuboid with an area at × at and a height ht is used as an abstract model for an interacting finger, where the

6 The parallel-plate capacitor model and 1 mm distance between the plates yields 2 pF

coordinates xt, yt and zt describe the center position of its tip. In general the capacitance between a human body and ground is about 150 pF [6]. This value is essentially larger than the maximal expected capacitance measurement6 for C3D. Therefore, a ground connected brass cuboid is a favorable simplification for a human finger.

The excitation signal is applied to the sensing electrode, which is used to measure the capacitance C3D to ground. With a guard amplifier the shielding electrode is served to the same potential as the sensing electrode. Thereby, the effect or influence of parasitic capacitances is kept in check.

Fig. 1. Schematic of the measurement set-up.

Table 1 shows the values of the parameters in Fig. 1 which are used for all the presented measurement data.

Page 124: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

124

Table 1. Parameters for the measurement.

Description Parameter Value Width of electrode as 15 mm Width of cuboid at 15 mm Height of cuboid ht 55 mm

Due to symmetry properties the 3D problem in Fig. 1 is presented as planar 2D problem in section 'Capacitance Modeling'. This configuration is divided in further parts for a mathematical description of the spatial capacitance distribution. This partial solutions of the 2D problem are combined for a 3D capacitance model. Section 'Capacitance Measurement' contains details concerning the measurement system and the measured values are compared with the 3D capacitance model. Further, section 'Results' presents 3D position estimation as a base for touchless interaction with a human finger. Finally, section 'Conclusion' discusses advantages and disadvantages as well as possible improvements for the 3D position estimation. Moreover, an outlook presents open problems for further studies.

2. Capacitance Modeling

Since an analytical calculation of C3D in Fig. 1 is sophisticated, simplifications are useful. These simplifications are based on [7] and are explicitly shown in [4].

2.1. 2D Modeling

Fig. 2 illustrates the abstraction of the measure-ment set-up in Fig. 1 as a planar problem in the xz-plane for y = 0. It is necessary to scale quantities in Fig. 2 with the spatial dimension ay in y-direction to be consistent with Fig. 1.

Fig. 2. Capacitances of the planar problem.

In Fig. 2 the corresponding human finger from Fig. 1 is assumed by a grounded metallic cuboid as test object. Furthermore, Fig. 2 splits the total capacitance into the capacitance Cb between the base of the cuboid

7 A contour line of a function of two variables is a curve along

which the function has a constant value. It is a cross-section of the

and ground as well as the capacitances Csl and Csr between the left and right side of the cuboid and ground. The capacitance between the top side of the cuboid and ground is negligible. Therefore, the total capacitance C2D between the sensing electrode and the cuboid follows as

, (1)

where the xt and zt dependency of the capacitances is not denoted explicitly.

In [4] for an approximate calculation of Cb a compact electrostatic problem is defined. For this configuration a unique solution of the Laplace's equation

Δ , 0 (2)

for the potential φ is determined. Thereby, (2) describes a Dirichlet problem with given potential values at the boundary. By means of this potential distribution the electric charge Qs at the sensor electrode can be calculated. Consequently through the ratio

(3)

with the voltage Ust between sensing electrode and test object the capacitance Cb is determined. It can be shown that the analytical solution for Cb results in

πln

cosh coshcosh cosh

(4)

with parameters kb1, kb2, kb3 and kb4 given through

π

42 (5.a)

π

42 (5.b)

π

42 (5.c)

π

42 . (5.d)

Furthermore, with the help of conformal mapping it is possible to calculate Csl and Csr via the same formulas (4) and (5) as Cb.

2.2. 3D Modeling

Before the 2D model can approximatively be extended by a third space dimension some further investigations are necessary.

Fig. 3 shows the measured capacitance contour lines7 of C3D for zt = 2 mm and zt = 5 mm in the xy-plane. Here, it can be seen that the contours become circular with increasing distance. As a consequence of the rectangular forms of sensing electrode and test

three-dimensional graph of the function f(x,y) parallel to the xy-plane.

Page 125: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

125

object the contours differ from the circular form in the short range (see Fig. 3 for zt = 2 mm). For large distances the capacitance contours are nearly circular (see Fig. 3 for zt = 5 mm). z = 2 mm.

Fig. 3. Contours of C3D for zt = 2 mm and zt = 5 mm.

Therefore, an approximation for the 3D capacitance model C3D follows as

, , , . (6)

In the following this approach is used for short as well as for large distances and the validity is verified with measurements. 3. Capacitance Measurement

Fig. 4 shows the measurement system consisting of a 3-axis translation stage, the test object, the sensing electrode array and the extended OpenCapSense hardware with eight channels [5, 8]. The 3-axis system has an accuracy in the submillimeter range. The requirements for the capacitance measurement system was a small uncertainty (around ±15 fF) and a wide measurement range (a few fF to several pF) combined with a moderate temporal resolution (< 5 ms/sensor).

Fig. 4. Measurement system.

Fig. 5 illustrates the single and multi electrode structures (a) and (b) used as well as the grounded

brass cuboid which represents a human finger. The sensing electrodes are made of a double-sided PCB (printed circuit board) with an additional nanosilver imprinted PET foil at the top [9]. The sensor array (b) bases on a multilayer structure in order to circumvent interference artifacts and guarantee high symmetry. To achieve a precise 3D localization of the test object several capacitance sensors must be combined to increase the spatial resolution (e. g., eight sensing electrodes in Fig. 5 (b)). In the set-up each electrode is connected to a separate measuring channel. The output signal of an astable multivibrator excites the sensing electrode and C3D is proportional to the period of this signal. The combination of multiple sensing electrodes would decrease the accuracy for a simultaneous measurement of all channels, because the electrodes influence each other. Therefore, the sensing electrodes are evaluated sequentially. While the capacitance C3D

of one sensing electrode is measured the other sensing electrodes are served to the same electrical potential as the active one to eliminate the influence of parasitic capacitances between the active and inactive sensing electrodes.

Fig. 5. Single electrode structure (a), multi electrode structure (b) and brass cuboid (c).

Fig. 6 shows a comparison between the measurement and the 3D capacitance model (6) with the explicit expressions for Cb, Csl and Csr in [4] for zt = 5 mm in the xy-plane. In the observed part the 3D capacitance model is a good approximation for the measurement.

4. Results

With the measured capacitance values and (6) the center position of the cuboid's base can be calculated through an optimization problem for the measuring position. At every sample interval the problem is solved as quasi stationary and the minimization problem follows as

min, , , , , , (7)

with the measured capacitance values Cm,i and the modeled capacitance values C3D,i in (6) for the measuring channel i ∈ 1...8. The four largest capacitance values of the eight channels are always used for the calculation in (7) to guarantee an

extended hardware

test object

sensing electrode array

3-axis system

Page 126: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

126

overdetermined problem. If further capacitance values are bigger than 80% of the lowest value of the four these are also added to improve the estimation.

The left part of Fig. 7 illustrates the deviation between the x-coordinates of the estimation and the true value of the cuboid's base while the right part of Fig. 4 shows the deviation between the y-coordinates. The increasing deflection in the far field of the

electrodes in Fig. 4 is caused by neglections in the 3D capacitance model (6). Furthermore, there are also limitations of accuracy concerning the extended OpenCapSense hardware [5, 8]. The processing of the measured data with the new approach [4] results in merely slight deflections (< 5 mm) around the sensor electrodes.

Fig. 6. Comparison of measurement and model for C3D at zt = 5 mm.

Fig. 7. Deviation of the estimated position of the coboid's base at zm = 1 mm for xm and ym.

5. Conclusion

The approach in [4] can be a good base for touchless interaction with a human finger. As shown in Fig. 7 the absolute position error for estimation with the functional (7) is < 5 mm. A moderate spatial resolution is sufficient if the time resolution is high enough to recognize smooth motions. Here, wipe gestures (e.g. up and down, left and right, circular forms, etc.) as known from smartphone interaction are of interest for depicting a user's intention. If necessary for further studies adaption parameters can help to decrease the absolute position error. However, the main focus concerning touchless interaction is the recognition of user's intention.

Also an interesting point for future work is to investigate the quality of the assumption of a grounded metallic brass object as finger model and under which conditions this simplification holds.

Currently the optimization problem in (7) is solved offline. Here, methods must be developed to ensure

online finger gesture recognition in real-time in future research.

Acknowledgements

The authors gratefully acknowledge the partial financial support for the work presented in this paper by the Austrian Research Promotion Agency (FFG) un-der contract grant 846057 and WimTec Sanitärpro-dukte GmbH. Furthermore, we thank Simon Hehenberger for recording the measurement data. References [1]. A. Braun et al., Capacitive sensor-based hand gesture

recognition in ambient intelligence scenarios, in Proceedings of the 6th International Conference on PErvasive Technologies Related to Assistive Environments, Rhodes, Greece, 29-31 May 2013, Article No. 5.

Page 127: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

127

[2]. T. Große-Puppendahl et al., Classification of User Postures with Capacitive Proximity Sensors in AAL-Environments, in Ambient Intelligence, Vol. 7040, 2011, pp. 314-323.

[3]. T. Große-Puppendahl et al., Honeyfish - A High Resolution Gesture Recognition System based on Capacitive Proximity Sensing, in Embedded World Conference, Weka Fachmedien, 2012, pp. 1-10.

[4]. L. Haslinger, B. G. Zagar, Voruntersuchungen zur berührungslosen Gestenerkennung mittels kapazitiver Sensorik, in Proceedings of the 18th GMA/ITG Conference on Sensors and Measurement Systems, Nürnberg, Germany, 10-11 May 2016, pp. 210-217.

[5]. L. Haslinger et al., Capacitance Measurement System for Touchless Interaction, in Proceedings of the

Eurosensors 2016, Budapest, Ungarn, 4-7 September 2016.

[6] G. Durcansky, EMV-gerechtes Gerätedesign, Franzis Verlag GmbH, 1999.

[7]. G. Willem de Jong, Smart Capacitive Sensors, Dissertation, Delft University of Technology, 1994.

[8]. T. Große-Puppendahl et al., OpenCapSense: A Rapid Prototyping Toolkit for Pervasive Interaction using Capacitive Sensing, in IEEE International Conference on Pervasive Computing and Communications, IEEE, 2013, pp. 152-159.

[9]. C. Beisteiner, B. G. Zagar, Electrical and mechanical characterization of inkjet-printed functional structures, tm - Technisches Messen, Vol. 83, 2016.

Page 128: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

128

Multi-functional Detectors of Ionizing Radiation on the Base

of Anion-defective Alumina

S. V. Zvonarev, V. S. Kortov, S. V. Nikiforov Ural Federal University, 19 Mira Str., 620002, Ekaterinburg, Russia

Tel.: + 73433754594, fax: + 73433754415 E-mail: [email protected]

Summary: Possibilities to use anion-defective alumina as a material for detectors of ionizing radiation with different ranges of doses to be registered were discussed. Thermoluminescent dosimetry with the use of such detectors in the ranges of low, mean and high doses makes them applicable in environmental protection, medicine and industry. Keywords: Thermoluminescence, Dose response, Alumina, Oxygen vacancies.

1. Introduction

Luminescent dosimetry is an important application

of phosphors among numerous ways of using the latter in science and technology. At present there are over ten types of commercial thermoluminescent (TL) detectors made on the base of sulfates and sullfides doped with different impurities of alkali-haloid and alkali-earth compounds. They are used to measure low doses and mainly in dosimetry of the personnel working with ionizing radiations.

Anion-defective alumina as a material for dosimetric phosphors has a number of advantages. Luminescence centers in it are created by their own defects rather than by dopants [1]. Oxygen vacancies which occur at high-temperature treatment of the material under highly reducing conditions (in vacuum, with the presence of carbon) are among the defects. Methods to change concentrations of such centers, their charge state, and methods to form aggregate centers which create additional traps of charge carriers have been developed in [2]. Registration of low, mean and high doses is also possible. This work presents the results of study into dosimetric properties of detectors based on anion-defective alumina.

2. Experimental

Two kinds of samples – single crystals and ceramics of anion-defective alumina – were studied.

The samples under study were discs 5 mm in diameter and 1 mm in thickness. These were made of alumina single crystal grown in highly reducing atmosphere with the presence of carbon. According to the optical absorption data, the concentration of luminescent centers created by oxygen vacancies in the samples was 1.3•1017 cm-3.

The experimental ceramic samples were synthesized from the commercial high-purity α-Al2O3 (99.997 %) nanopowder with the particle size of 50-70 nm. The powder was pressed into pellets (compacts) 5 mm in diameter and 1.3 mm thick under

1000 kg/cm2. The pellet weighed 40 mg. The synthesis of ceramics was carried out in the vacuum (10-3 – 10-4 Torr) electric furnace at the temperature range from 1100 ⁰С to 1700 ⁰С and the annealing time varying from 30 min to 3 h.

The standard method was used to perform TL measurements with a FEU-142 photomultiplier at the linear temperature change from 300 to 820 K at the rate of 2 K/s. To find the dependence of the TL peak intensity on the absorbed dose, the samples were preliminarily exposed to different doses. 3. Results and Discussion

Fig. 1 shows TL curve and dose response of TL yield when irradiated anion-defective alumina single crystals are used. Highly-sensitive detectors TLD-500 are developed on the basis of these crystals. The oxygen vacancies formed in single crystals trap one or two electrons giving rise to F+ and F-centers correspondingly. These centers with impurities form complex defects which are traps of free electrons resulting from irradiation. Thermal ionization of such traps at the sample heating with the following recombination of delocalized electrons causes the main dosimetric thermoluminescent (TL) peak at 460 K. The peak intensity is proportional to the dose of X-ray, gamma- and beta-radiations in the range of 10-6-1 Gy. With the dose increasing higher than 1 Gy, the TL intensity of the main peak is saturated.

In the ultrafine ceramics the TL peak is shifted to the region of lower temperatures. Low-temperature shift of the dosimetric TL peak is typical for ultrafine phosphors due to a great contribution of the surface centers which are shallow traps. Ceramics with the highest TL intensity synthesized at 1700 ⁰С were chosen to measure dose dependence. Linear increase in TL intensity with a growing dose is observed in the range of 3•10-2 -3•102 Gy.

Page 129: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

129

Fig. 1. TL glow curve (a) and dose response (b) of anion-defective alumina single crystals.

It is also possible to increase capacity of the traps of dosimetric phosphor using deep traps of the charge carriers. In oxygen-deficient aluminum oxide the deep traps are present both in a single crystal and in ultrafine ceramics. The task of registering high doses consists in measuring TL yield of all deep traps. For this purpose the phototransfer luminescence and photo-thermostimulation luminescence methods are used. In both methods the detector exposed to a high dose is preliminary heated to (500-550) K to empty shallow traps. To register PTTL the detector is then exposed to intensive blue light for optical emptying the traps and for transfer of the electrons delocalized from the deep traps through the conduction band to the unfilled shallow traps. The following heating of the detector allows one to measure PTTL. Its intensity is

proportional to a number of electrons which filled the deep traps under high-dose exposure.

Table 1 features possibilities of the detectors made from anion-defective alumina in registering doses of different ranges and their related applications. The data in the table show that the dose ranges registered by using the materials under study may be of 11 orders of magnitude, which provides their multi-functionality. 4. Conclusions

Characteristics of thermoluminescence of anion-defective alumina as a promising material for detectors of ionizing radiations have been discussed. Highly-sensitive detectors are created on the base of single crystals. These detectors allow registering both low and high doses depending on the way of measurement. Ultrafine alumina ceramics made of nanopowder are promising for measurement of mean doses. A wide useful range of the detectors based on anion-defective alumina enables their multi-functional applications in science and technology. Acknowledgements

This work has been done a part of the government task (3.1016.2014/K) the Ministry of Education and Science of Russian Federation and was supported by the scholarship of the President of the Russian Federation for young scientists and graduate students engaged in advanced research and development in priority areas of modernizing the Russian economy number SP- 983.2015.2.

References [1]. M. S. Akselrod, V. S. Kortov, D. I. Kravetsky, V. I.

Gotlib, Highly sensitive thermoluminescent anion-defective α-A12O3:C single crystal detectors, Radiation Protection Dosimetry, Vol. 32, 1990, pp. 15-20.

[2]. V. S. Kortov, V. A. Pustovarov, T. V. Shtang. Defect evolution and photoluminescence in anion-defective alumina single crystals exposed to high doses of gamma-rays, Radiation Measurements, Vol. 85, 2016, pp. 627-638.

Table 1. Ranges of registering ionizing radiation and applications of detectors based on anion-defective alumina.

Structure of detectors

Kind of stimulation in dose measurement

Useful dose range, Gy

Applications

Single crystal Thermostimulation at

linear heating 10-6 – 1

Dosimetry of the personnel working with ionizing radiations Radiation environmental monitoring

Ceramics Thermostimulation at

linear heating 10-2 – 102 Emergency dosimetry in NPP, dose control in radiotherapy

Single crystal Photo-

thermostimulation 102 – 105

Sterilization of goods, modification of material properties, radiation technologies in industry

Page 130: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

130

Design and Characterization of a 2D Array of MEMS Microphones

for Acoustical Imaging

A. Izquierdo 1, J. J. Villacorta 1, L. del Val 2 and L. A. Suárez 3 1 University of Valladolid, Signal Theory Department, ETSII, Paseo Belén 15, 47011 Valladolid, Spain

2 University of Valladolid, Mechanical Engineering Department, EII, Paseo del Cauce 59, 47011 Valladolid, Spain

3 University of Burgos, Civil Engineering Department, EPS, Campus Milanera, Edificio Milanera C/ Villadiego s/n, 09001 Burgos, Spain

Tel.: + 34 983185801, fax: + 34 983423667 E-mail: [email protected]

Summary: This paper presents the characterization of each of the MEMS microphones which form a uniform planar array used for acoustical imaging. These microphones show an essentially flat frequency response, with a ±2 dB variation between all sensors. This paper also shows the design and the acoustic characterization of the planar array of MEMS microphones. Analyzing the acoustic array characterization it was observed that as the variations of the measured beampattern, with respect to the theoretical one, are limited, it is not necessary to apply calibration techniques to the array. Keywords: Characterization, Planar array, MEMS microphones, Acoustical imaging.

1. Introduction

An array is an arranged set of identical sensors, fed

in a specific manner. The beampattern of the array can be controlled by modifying the geometry of the array (linear, planar...), the sensor spacing and the beampattern, the amplitude and phase excitation of each sensor [1]. Microphone arrays are a particular case [2].

The authors of this paper have experience in the design [3] and development of acoustic ULAs (Uniform Linear Arrays), formed by acoustic sensors distributed uniformly along a line. In order to obtain spatial information in two dimensions, it is necessary to work with planar arrays with sensors distributed on a surface. Working with planar arrays leads to an increase in both system complexity and space required by the acoustic sensors and the associated hardware.

A typical system has four basic elements: sensors, signal conditioners, acquisition devices and signal processor. For the first three elements, system cost increases linearly with the number of channels, as each sensor needs a signal conditioner and an acquisition device.

Digital MEMS (Micro-Electro-Mechanical System) microphones include a microphone, a signal conditioner and an acquisition device incorporated in the chip itself. For this reason, an acquisition and processing system for an acoustic array based on MEMS microphones is reduced to two basic elements: MEMS microphone and a processing system. The integration of the microphone preamplifier and the ADC in a single chip significantly reduces costs, and also the space occupied by the system.

This paper presents the design and the acoustic characterization of a planar (2D) array of MEMS microphones that will be used to obtain acoustic

images of different targets and use them in different applications, such as a biometric system, or noise and vibration analysis.

Section 2 introduces MEMS microphones technology. Section 3 characterizes the frequency response of the microphones used. Section 4 defines the planar array geometry and Section 5 shows its acoustic characterization. Finally, Section 6 contains the conclusions and future research lines.

2. MEMS Microphones

The acronym MEMS refers to mechanical systems with a dimension smaller than 1 mm [4] manufactured with tools and technology arising from the integrated circuits (ICs) field.

The application of MEMS technology has allowed the development of high-quality microphones with high SNR (Signal to Noise Ratio), low power consumption and high sensitivity. MEMS microphones consist of two components, the acoustic sensor and the controller circuit [5]. 3. Characterization of each MEMS Acoustic

Sensor in Laboratory

For the implementation of the array, MP34DT01 microphones of STMicroelectronics, - digital MEMS microphones with PDM interface - were chosen.

MEMS sensor characterization was performed analyzing the frequency response of all MEMS sensors included in the array. A sinusoidal 4 ms pulse with a frequency changing between 2 and 18 kHz was generated using a reference loudspeaker. All

Page 131: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

131

measurements were performed in an anechoic chamber.

The frequency response of each MEMS sensor was obtained and normalized according to the loudspeaker’s response. Then, the average of the frequency responses was assessed. Fig. 1 shows all the responses.

It can be observed that the averaged frequency response is essentially flat, with a slight increase at high frequencies. This averaged response is bounded within a range of ±4 dB. Fig. 1 also shows that the frequency response of MEMS sensors varies in a range of ± 2 dB around the averaged value.

Fig. 1. Frequency responses and averaged response of the MEMS microphone.

4. Array Geometry

The array characterized in this paper is based on a Uniform Planar Array (UPA) of MEMS microphones. This is a square array of 64 (8x8) MEMS microphones which are spaced uniformly, every 2.125cm, in a rectangular Printed Circuit Board (PCB), with square gaps between the acoustic sensors, as it can be observed in Fig. 2, in order to make the array as light and portable as possible.

Fig. 2. Array of MEMS microphones.

This array was designed to work in an acoustic frequency range between 4 and 16 kHz. The 2.125 cm spacing corresponds to λ/2 for the 8 kHz frequency. 5. Array Acoustic Characterization

A reference loudspeaker placed in different positions was employed to obtain the beampatterns of

the MEMS array included in the system. Beamforming was carried out with a wideband FFT algorithm, focused on the loudspeaker position. Fig. 3 shows one of the obtained beampatterns.

The measured beampatterns are very similar to the theoretical ones, which assume that the acoustic sensors are omnidirectional and paired in phase. Nevertheless, a more detailed analysis of the measured beampatterns shows: i) there are more sidelobes with a level higher than -20 dB, and ii) there is a very small displacement of the sidelobes, which are closer. These effects are due to the fact that the gain of each microphone is slightly different for each frequency, as shown in Fig. 1. This is the same effect as applying windowing techniques to the beamforming weight vector, which modifies the level and the position of the sidelobes.

Fig. 3. Measured beampatterns for 16 kHz and steering angles [azimuth, elevation]:[5º, 10º].

Page 132: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

132

6. Conclusions

The averaged frequency responses of the MEMS microphones which have been employed to form the array are essentially flat, with a slight increase at high frequencies, and vary in a range of ± 2 dB around the averaged value.

Analyzing the acoustic array characterization it was observed that as the variations of the measured beampattern, with respect to the theoretical one, are limited, it is not necessary to apply calibration techniques to the array.

Acknowledgements

This work has been funded by the Spanish research project SAM: TEC 2015-68170-R (MINECO/FEDER, UE).

References [1]. Van Trees, H. Optimum Array Processing: Part IV of

Detection, Estimation and Modulation Theory, John Wiley & Sons, New York, USA, 2002.

[2]. Brandstein M.; Ward D., Microphone arrays, Springer, New York, USA, 2001.

[3]. del Val, L., Jiménez, M., Izquierdo, A., Villacorta J., Optimisation of sensor positions in random linear arrays based on statistical relations between geometry and performance, Appl Acoust, 2012, 73, 1, pp. 78–82.

[4]. Hsieh, C. T., Singapore, Y., et al., The introduction of MEMS packaging technology, in Proceedings of 4th International Symposium on Electronic, Materials and Packaging, Kaohsiung, Taiwan, Dec. 4-6, 2002, pp. 300-306.

[5]. Beeby, S., Ensell, G., Kraft, M., White, N., MEMS Mechanical Sensors; Artech House Publishers, Norwood, MA, USA, 2004.

Page 133: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

133

GSM-GPS Based Security System and Its Implementation as Anti-Theft

System in Automobiles

Mandakinee Bandyopadhyay 1, Nirupama Mandal 2 and Subrata Chatterjee 3

1Department of Electrical Engineering, Asansol Engineering College, Asansol, India 2Department of Electronics Engineering, Indian School of Mines, Dhanbad, India

3Department of Electrical Engineering, National Institute of Technical Teachers’ Training and Research, Kolkata, Kolkata, India

Email: [email protected] Summary: An integrated GPS-GSM based security system is developed to track and control vehicles from remote places which is cost-effective and reliable. In the present paper, an automotive localization system using GPS and GSM-SMS is developed and implemented as an anti-theft system in Automobiles. The presented application is a low cost solution for conveying automobile position and status, very useful in case of car theft situations in smart cities where cell phone is an essential, for monitoring adolescent drivers by their parents as well as in car tracking system applications. Keywords: GSM, GPS, Telemetric, Embedded Systems, Anti-theft, Real-time monitoring. 1. Introduction

From the invention of the first lock and key to the

introduction of RFID tags and biometric identification, anti-theft system [1] have evolved to match the introduction of new and improved techniques of automobile theft. An immobiliser is an electronic security device fitted to an automobile that prevents the engine from running unless the correct key (or other token) is present. This prevents the car from being "hot-wired" after entry has been achieved [2].

All Modern vehicle tracking systems [3-7, 9, 10] commonly use Global Positioning System (GPS) technology for locating the Vehicle, to achieve automatic Vehicle Location system that can transmit the location information in real time. The information is transmitted to Tracking server using GSM/GPRS modem on GSM network [8] by using SMS or using direct TCP/IP connection with tracking server through GPRS. Tracking server also has GSM/GPRS modem that receives vehicle location information via GSM network and stores this information in a database. This information is available to authorized users of the system via the Internet at a web-address.

This paper proposed to design a vehicle tracking system that works using GPS and GSM technology, which would be the cheapest source of vehicle tracking and it would work as anti-theft system.

It is an embedded system, which is used for tracking and positioning of any vehicle by using Global Positioning System (GPS) and Global system for mobile communication (GSM). A low cost anti-theft system of a car is designed and developed which is compatible as an additional feature to the present security system that will warn the owner of the vehicle by sending SMS when there has been an intrusion into the vehicle. Immediately, the car owner will send the appropriate command to the embedded platform and will be assured full security against theft and flexibility

in monitoring the car status via GSM Communication. The functional device for user to communicate with the embedded platform is GSM/ CDMA based mobile phone handset. In smart cities, where citizens are busy and monitoring of vehicles is becoming tougher day by day, this finds immense application.

2. Method of Approach Modern technology like GPS has been used in this

investigation. The GPS system works by receiving the radio frequency (RF) signals transmitted by the GPS satellites [11-13]. The block diagram of the system that represents the vehicle positioning and controlling, is shown in Fig. 1.

On being switched on, the main microcontroller searches for incoming SMS received by the GSM module. When it receives instruction to lock the car, it activates the door lock relay and starts monitoring if there is any signals from the sensors like Passive Infrared Sensor (PIR), 3-Axis Accelerometer and Magnetic Reed Switch.

Fig. 1. Block Diagram of the System.

Page 134: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

134

When the system is in monitoring mode, it looks for signal. If it detects any, the microcontroller sends an interrupt signal to the GPS data decoding auxiliary microcontroller the main controller sends command to the GSM module. The modem then sends the formatted data and intrusion type to the owner via SMS. If the owner thinks that the car has been stolen, he can send command to stop the engine of the car and lock the brake. After retrieving the car, when the owner sends command to unlock it, the monitoring system will be deactivated and the main microcontroller stays inactivated till new SMS is received for further processing. 3. Design and Analysis

A GSM modem is incorporated for user interface and vehicle monitoring purpose. Short Messaging Service or SMS has proven to be a very low-cost and popular mobile service all over the world. Thus it is used in this research for controlling and tracking the Automobiles.

The system hardware are shown in Fig. 2a and 2b. The system architecture is based on a multi-processor operation technique. Both the Atmega8 microcontrollers are configured to use the internal 8-Mhz oscillator. There is a main controller that handles the SIM300 GSM unit, 16X2 LCD display driver unit, digital PIR sensor (The viewing angle of the sensor is 100°), ADXL335 analog Accelerometer, Magnetic Reed Switch and a 12 V (JQC-3FC/T73) relay interface for the automobile control purpose.

Fig. 2a. Multi Controller Platform.

This design will continuously monitor a moving Vehicle and report the status of the Vehicle on demand.

The GPS unit - GTPA010 is comprised of an auxiliary atmega8 microcontroller for decoding the received NMEA format data into a user readable LAT., LON. Data format. A GSM modem is used to send the position (Latitude and Longitude) of the vehicle from a remote place which is provided to it by the GPS [9] modem. The same data is sent to the mobile at the other end from where the position of the vehicle is demanded. A 74HCT08N AND Gate is used as a switching circuit to feed the final acquired data and the received SMS data to the main controller. Two LEDs are employed to indicate the data collection & SMS sending acknowledgement. Three relays are connected to the controller for door, engine & brake activation- deactivation purpose. AVR Studio 6 is used as the main software.

Fig. 2b. Architecture of the System.

4. Operation Process

The idea of the system is very contemporary and

this kind of security system is of high demand in the society presently. The functional flowchart of this GSM-GPS security system and its implementation as an anti-theft system in automobiles is shown in Fig. 3.

When the system is switched ON, it takes at least one or two minutes to establish the connection of GSM Module, GPS and most importantly PIR sensor for auto calibration.

The owner sends an SMS lock n which activates the door lock relay and locks the car, sends an acknowledgment and further continuously monitors for any interruption in signals received from the sensors, namely, Passive Infrared Sensor (PIR), 3-Axis Accelerometer and Magnetic Reed Switch.

Page 135: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

135

Fig. 3. Functional Flowchart of GSM-GPS Security System.

The PIR sensor, 3-Axis Accelerometer and magnetic Reed Switch are used to detect Intrusion, crash and forceful door unlocking respectively. The PIR sensor gives a signal when human intrusion happens into the car though it is locked. It detects living being by the emitted infrared radiation. To detect sudden vibration or crash, the accelerometer is activated. Therefore when intrusion is detected or door lock is broken or crash is detected, then the owner will get an SMS and accordingly he can find out the position of the car by sending SMS “LOCK T”. A GSM unit is used to provide continuous data about the position (Latitude and Longitude) of the vehicle. This data is sent to the mobile at the other end as a return reply SMS.

Now, if owner finds that the car has been stolen, then he can send an SMS “LOCK B,” to activate brake relay for application of brakes and “LOCK E” to stop the engine of the car by another relay activation. Immediately owner will receive an acknowledgment through return SMS.

5. Experimental Results

When the system is switched on by applying +12 V, the on board LEDs will glow. The main and auxiliary

microcontroller will set up its I/O pins, baud rate and other types of primary settings. GSM modem & GPS unit will be activated. When the GSM modem finds a valid mobile network, the LCD unit will show “READY” as shown in Fig. 4.

Fig. 4. Initialization of the system. When the owner sends SMS to activate the lock, the

system receives a command SMS as shown in Fig. 5. The acknowledgment is displayed in the liquid crystal display (LCD) as shown in Fig. 6. When the car is locked through car door lock relay, a confirmation SMS is sent to the owner as shown in Fig. 7.

Page 136: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

136

Fig. 5. SMS received for activating the LOCK.

Fig. 6. Acknowledgment of Car Locking.

Fig. 7. Acknowledgment SMS is being sent to the owner.

Now the main controller starts to monitor the sensor and if any valid signal is received, the system will alert the owner through SMS. When PIR sensor detects the intrusion, it sends a message to the owner. When the door lock is broken by unauthorized person, then the owner will receive an SMS or if the car is crashed, the owner will receive an SMS.

Hence, when the car is stolen, then owner can find out the present position of the car by sending a command message “LOCK T”. If intrusion detection or door lock broken is taken place then the owner receives SMS from car and immediately the owner can send SMS command “LOCK E” and “LOCK B” one by one for stopping the engine and locking the brake of the car respectively. Simultaneously, the owner will get a return SMS which will confirm the same.

6. Discussion

After the system is powered on, it is advised to leave it for at least one or two minutes for GSM connection establishment, GPS fix and most importantly PIR sensor auto calibration.

The experimental flow chart is found to have quite good response. The results are satisfactory and show satisfactory repeatability. With the help of this flexible and upgradable microcontroller system, many more features would be added in future according to the purpose of the owner.

Human interference in the PIR sensor visible area should be avoided during the calibration process, otherwise the system can behave in an undesired manner. If any of the sensors detect intrusion or accident, it sends SMS immediately. It takes about 5 to 10 seconds to receive the SMS in the cell phone. Two consecutive commands must be sent after an interval of 10 to 15 seconds minimum.

Performance factors will highly vary according to the signal strength of the GSM communication, GPS satellite visibility etc. Unwanted service SMS can create disturbance in the system.

It is also useful in many other applications, such as, Asset Tracking system where companies need to track valuable assets for insurance or other monitoring purposes. They can now plot the real-time asset location on a map and closely monitor movement and operating status. This system can be implemented in banking sector, accident situations and small scale industrial purpose also. It is an essential in Smart Cities and an example of how car theft can be controlled. Implementation, no doubt, shall define smart governance. References [1]. Zhang Yu, Research on High Level Model and

Performance Estimation, PhD Thesis, Southeast University, 2007.

[2]. J. Teresko, Winning with wireless - Industry Week, 252, 6, Available ProQuest, 2003.

[3]. S. Ajaz, M. Asim, M. Ozair, M. Ahmed, M. Siddiqui, Z. Mushtaq, Autonomous Vehicle Monitoring & Tracking System, in Proceedings of the Student Conference on Engineering Sciences and Technology (SCONEST’05, 2005, pp. 1 – 4.

[4]. A. T. Hapsa, E. Y. Syamsudin, and I. Pramana, Design of Vehicle Position Tracking System Using Short Message Services and its Implementation on FPGA, in Proceedings of the Conference on Asia South Pacific Design Automation, Shanghai, China, 2005.

[5]. Fan, X., W. Xu, H. Chen, and L. Liu, CCSMOMS: A Composite Communication Scheme for Mobile Object Management System, in Proceedings of the 20th International Conference on Advanced Information Networking and Applications, Vol. 2, Issue 18-20, April 2006, pp. 235–239.

[6]. E. M. Tamil, D. B. Saleh and M. Y. I. Idris, A Mobile Vehicle Tracking System with GPS/GSM Technology, in Proceedings of the 5th Student Conference on Research and Development (SCORED), Permala Bangi, Malaysia, May 2007.

[7]. W. C. M. Hsiao and S. K. J. Chang, The Optimal Location Update Strategy of Cellular Network Based Traffic Information System, in Proceedings of the IEEE Intelligent Transportation Systems Conference, 2006.

Page 137: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

137

[8]. M. A. Al-Taee, O. B. Khader and N. A. Al-Saber, Remote monitoring of Automobile diagnostics and location using a smart box with Global Positioning System and General Packet Radio Service, in Proceedings of the IEEE/ACS AICCSA, May 13–16, 2007, pp. 385–388.

[9]. Oan Lita, Ion Bogdan Cioc and Daniel Alexandru Visan, A New Approach of Automobile Localization System Using GPS and GSM/GPRS Transmission, in Proceedings of the 29th International Spring Seminar on Electronics Technology (ISSE '06), 2006, pp. 115-119.

[10]. T. Krishna Kishore, T. Sasi Vardhan, N. Lakshmi Narayana, Automobile Tracking Using a Reliable Embedded Data Acquisition Sysytem With GPS and

GSM, International Journal of Computer Science and Network Security, Vol. 10 No. 2, 2010, pp. 286-291. N. Kamarudin and Z. M. Amin, Multipath error detection using different GPS receiver's antenna, in Proceedings of the 3rd FIG Regional Conf., Jakarta, Indonesia, October 3-7, 2004.

[11]. T. E. Melgard, G. Lachapelle, and H. Gehue, GPS Signal Availability in an Urban Area-Receiver Performance Analysis, in Proceedings of the IEEE Position Location and Navigation Symposium, 1994.

[12]. R. A. Nayak, M. E. Cannon, C. Wilson, G. Zhang, Analysis of Multiple GPS Antennas for Multipath Mitigation in Vehicular Navigation, Institute of Navigation National Technical Meeting/Anaheim, CA/January 26-28, 2000.

Page 138: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

138

Magnetometer Design for Copters

V. Korepanov and F. Dudkin

LLC Laboratory for Electromagnetic Innovations (LEMI), 5-A Naukova Str., 79060 Lviv, Ukraine Tel.: + 380322639163, fax: + 380322639163

E-mail: [email protected] Summary: The copters are the recent challenge and many instruments are attempted to be used with them, including magnetometers. The last is a pending problem, especially for component magnetometers, because of the complexity of reducing the data obtained in permanently rotating frame system to the data in geomagnetic frame system. Next problem is high interference level due to copterelectrical circuits and its moving metal parts during magnetometer operation. A possible approach to the first problem solution is presented. Keywords: Magnetometer, Fluxgate, Coordinate frame reducing.

1. Introduction

Recently, a new type of flying carriers – remotely

piloted vehicles or copters – is becoming more and more popular. They allow solving wide spectrum of tasks, including the search of small shallow targets such as UXO or archeological remnants. This implies new more stringent requirements to the onboard magnetometer for such a small apparatus, i. e. compactness, low weight and small power consumption. The regularly used magnetometers for aeromagnetic survey – scalar ones – are not applicable for so small vehicles and do not allow 3-component geomagnetic mapping. For that reason the researchers turn their attention to other more light but enough sensitive instruments for magnetic field measurement. A novelty of the given study is the attempt to replace the scalar magnetometers by vectorial ones and provision of corresponding measurement methodology. The experiments on the flying carriers with the use of vectorial magnetometers are very seldom, first of all because of the complexity of data interpretation [1]. It follows from the fact that the measurements taken in the rotating coordinate frame have to be reduced to a fixed frame, which is usually connected with a ground surface. This niche still waits for new experimental ideology, which will allow obtaining the final results with high accuracy. A possible realization of such a technology based both on recent achievements in vectorial magnetometry is described below.

2. Magnetometer Development

A detailed analysis shows that the best combination of sensitivity, weight and power consumption within all vectorial magnetometers have the flux-gate magnetometers (FGM). The existing FGMs parameters are analyzed and it is stated that, as to sensitivity threshold, modern FGMs are competitive with those used for the aeromagnetic survey, i.e.,

Overhouser or quantum (helium, cesium) scalar magnetometers. The lowest known value of the FGM sensor noise obtained today is ~ 1 pT at 1 Hz [1]. So, the FGM resolution allows their application for the given task. Such a high sensitivity is supposed to cover practically every detection task. This was encouraging for the development of a portable FGM version which should have as low as possible sensitivity threshold in miniaturized version. By this the most difficult to realize parameter of a stationary FGM – temporal stability of the baseline – is not of primary importance because the estimated FGM operation time is limited by a copter autonomous flight time (around ~ 3 hours maximum). During the FGM development much attention was paid to the influence of ambient temperature changes reducing on the magnetometer zero line drift, including the case when sensor is in the suspended state. Finally, it was obtained that the developed suspension mechanism practically does not add an essential drift at thermal tests, even more – in majority of tests the same sensor in suspended state showed smaller thermal dependence. The thermal tests results of the same flux-gate sensors in stationary and suspended state are given in the Table 1.

Table 1. Thermal tests results.

Sensor No X

nT/°C Y

nT/°C Z

nT/°C No 1,

not suspended 0,15 0,15 0,12

No 1, suspended 0,09 0,13 0,07 No 2,

not suspended 0,27 0,31 0,09

No 2, suspended 0,21 0,09 0,21 No 3,

not suspended 0,16 0,13 0,07

No 3, suspended 0,13 0,09 0,07

As it can be seen from this table, mean thermal drift value of the LEMI type magnetometers with

Page 139: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

139

suspended sensor is around 0.1 nT/ºC what is acceptable for onboard a copter application because the sharp temperature changes during short investigation time of a copter-portable experiment are very seldom. Moreover, as thermal tests results show, the obtained experimentally thermal drift is highly linear with temperature. Taking into account that each LEMI magnetometer has thermal sensor inside, the drift value can be still lowered by about one order of magnitude what will decrease it to a practically negligible value. 3. Frame Rotation Method

More important problem is the FGM sensor orientation which permanently changes during the flight. It is enough to say that in the Earth’s magnetic field as tiny as 1 arc second axis deviation of a perpendicularly oriented magnetic sensor can lead to zero shift in the range 0.14 - 0.32 nT, what may be already more than the expected measurement level of useful signals. The FGM axes reducing from the random position to the fixed frame system may be made using inverse transformations of the measured values to a chosen coordinate system by which some uncertainties may appear. They can be removed by the use of a tiltmeter which will give the X and Y axes deviations of FGM sensor from horizontal plane. Recently, the market of electronic components proposes new non-magnetic tiltmeters with resolution about few arcseconds. Then, applying known equations (for example for Euler rotation) modified for our case, it is possible to carry out the inverse transform for reduction of the FGM sensor axes being at arbitrary orientation to the fixed frame system (for example to geomagnetically oriented one).

The method idea was already successfully tested with sea bottom magnetometer which was installed for the fixed position [2]. We also used second magnetometer to determine the accurate values of the Earth’s magnetic field components and its variations what allowed us to calculate α, β, γ as precession, nutation and proper rotation angles (Euler’s angles). If the direction of the Earth magnetic field is known from a reference ground magnetometer then the direction of rotation axis can be calculated.

Here the situation is complicated by permanent displacement and rotation of the towed 3-component sensor by the copter. To overcome this, the quaternion representation for description of rotating magnetometer data can be used. It allows avoiding the uncertainty of the Euler angles application by description of rotating magnetometer data in 4D space with use of hypercomplex numbers calculating the

rotation matrix [M] around ort n. Then vector components measured by a rotating magnetometer can be calculated by equation:

Bm=[M]TBn , (1) where [M]T is transposed matrix, Bm and Bn are the magnetic field components measured by base and rotated magnetometers.

However such a presentation does not give possibility for calculation of vector components of anomalous magnetic field which can be associated with studied changes of conductivity in the lower sounded half-space (only the deviation of the absolute value for anomalous magnetic field may be found). Such components of the anomalous magnetic field provide not only localization of magnetic field source but calculation of effective conductivity of a studied half-space. It is connected with the rotation operation of a total magnetic field vector, which is sum of the Earth and anomalous magnetic field, relatively the Earth magnetic field vector of the base magnetometer. The maximum error of rotation angle calculation (Δθ, in degrees) can be estimated from simple relation:

Δθ≤ |ΔB|/|B0|, (2) where ΔB and B0 are vectors of the anomalous and the Earth’s magnetic fields. As an example, for |ΔB|=100 nT and |B0|=50,000 nT the angular error is less than 0.11o. 4. Conclusion

The problem solution for determination of the anomalous vector Cartesian components needs in development of a new approach what is the goal of the present study. The dedicated magnetometer was created and tested. Tests results showed encouraging progress which still needs finalization for routine application. References [1]. V. Korepanov, Yu. Tsvetkov, Gradient magnetometer

system for balloons, in Proceedings of the 17th ESA Symposium on European Rocket and Balloon Programmes and Related Research, Sandefjord, Norway, 30 May – 2 June 2005 (ESA SP-590, August 2005), pp. 443-44.

[2]. Prystay A. M., Monitoring of the Earth’s magnetic field on the sea bottom, Geophysical Journal, Vol. 26, Issue 3, 2004, pp. 96-101 (in Ukrainian).

Page 140: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

140

Thermoelectric Single-photon Detector on the Base

of W/(La,Ce)B6/W/Al2O3 Multi-layer Sensor

A. S. Kuzanyan 1,4, A. A. Kuzanyan 1, V. R. Nikoghosyan 1, V. N. Gurin 2 and M. P. Volkov 3 1 Institute for Physical Research NAS of Armenia, Material Science Laboratory,

1 Ter-Mikaelyan Av., 0203 Ashtarak, Republic of Armenia 2 Ioffe Physical-Technical Institute of the RAS, Solid State Physics, St. Petersburg, Russian Federation

3 Ioffe Physical-Technical Institute of the RAS, Physics of Dielectric and Semiconductors, St. Petersburg, Russian Federation

4 Institute of Physics and Technology National Research, Tomsk Polytechnic University, Tomsk, Russian Federation

Tel.: + 37494170406, fax: + 37423231172 E-mail: [email protected]

Summary: A new design of the sensor for thermoelectric single-photon detector is proposed. The results of modelling of kinetic processes in the sensor of thermoelectric single-photon detector after absorption IR – X-ray photons in different parts of the absorber for different geometries of the W/(La, Ce)B6/W/Al2O3 multi-layer sensor are presented. It is shown, that thermoelectric detectors with multi-layer sensor have high energy resolution, gigahertz count rate and can be a real competitor to superconducting detectors. Keywords: thermoelectric single photon detector, multi-layer sensor.

1. Introduction

Single-photon sources and single-photon detectors

are the most important components in the fields of quantum information and communications technology. Single-photon detectors are required in research in different areas of modern science as well. The operation principle of thermoelectric single-photon detector (TSPD) is based on photon absorption as a result of which a temperature gradient is generated on the edges of the thermoelectric sensor [1]. Materials, which can be used to prepare the absorber and the sensor, the achievable count rates and energy resolution are given in [2]. It can be concluded that the thermoelectric detector with single layer sensor may possess an energy resolution of 0.1 eV and a gigahertz level count rate. The results of computer simulation of heat distribution processes in a single layer sensor of TSPD were published in [3, 4]. The idea of multi-layer sensor of TSPD has recently been proposed and patented.

The results of computer simulation of heat distribution processes after single photon absorption in the W/(La, Ce)B6/W/Al2O3 multi-layer sensor of TSPD are presented in this work.

2. Results 2.1. Single Layer and Multi-layer Sensors of TSPD The scheme of the TSPD sensitive element is shown in Fig. 1. Such a sensor does not require either a separate power unit or a bias voltage. Therefore, it does not need additional leads for electronic circuitry. A matrix

detector built on them will have very simple engineering and electronic structure. The single layer sensor of the thermoelectric detector consists of three parts – the absorber, the heat sink and the connecting bridge made of thermoelectric material. The multi-layer sensor consists of a photon absorber deposited on the thermoelectric layer, which in turn is deposited on the electrically conductive layer of a heat sink. Such a ''sandwich'' is placed on a substrate, with electrical contacts to measure the potential difference ΔU, generated between the absorber and the heat sink.

Fig. 1. Detection pixels of TSPD: a – single layer, b – multi-layer. 1 – substrate, 2 –absorber,

3 – thermoelectric, 4 – heat sink.

2.2. Computing Technique The calculation were based on the heat conduction equation and were carried out by the matrix method for differential equations. Let us mention that tungsten (W) was selected as absorber and heat sink material. The operating temperature of the detector was taken equal to 9 K and 0.5 K for sensor with CeB6 and (La0.99Ce0.01)B6 thermoelectric layer respectively. The thermal processes modeling in details is presented in [4].

Page 141: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

141

2.3. Registration of Photons with Different Energy

The main parameter of multi-layer sensor is the thickness of thermoelectric layer. Fig. 2 shows the temporal dependence of temperature difference on the sensor after 100 eV photon absorption ΔT(t) for the four calculations. Graphs on Fig. 2 demonstrate that the dependences of ΔT(t) for 1, 0.5, 0.1 μm thicknesses of the thermoelectric layer (calculations 1M – 3M) do not differ significantly. Maximum of ΔT(t) of the 4M calculation (thickness 0.1 μm) is slightly lower, but it is offset by the rapid decline to the background value, which provides high count rate.

Fig. 2. ∆T(t) dependence for W/(La0.99Ce0.01)B6/W/Al2O3

sensors with different thickness of thermoelectric layer.

Graphs on Fig. 3 demonstrate the dependence of count rate R from thickness of thermoelectric layer z for 100 eV photon absorption in multi-layer sensor.

0.00 0.25 0.50 0.75 1.000

500

1000

1500

2000

2

zt, m

R, G

Hz

1

Fig. 3. Dependence of count rate R from thickness of thermoelectric layer z for sensor on the bases of:

1 – (La0.99Ce0.01)B6, 2 – CeB6.

The maximum of temperature difference (∆Tm) and potential difference (∆Um) appearing on the sensor of defined geometry after photon absorption linearly depends on the energy of the absorbed photon (Fig. 4).

0 2 4 6 8 10 120.0

0.2

0.4

0.6

0.8

1.0

1.2

E, V

T m

, K

0

20

40

60

80

100

120

Um,

V

Fig. 4. Dependence from photon energy E of ∆Tm and ∆Um parameters of W/(La0.99Ce0.01)B6/W/Al2O3 sensor.

3. Conclusions

Achievement of high count rate and independence of the waveform of time dependence of the response from the area of photon absorption can be noted among the advantages of the multi-layer sensor. Taking into account several features of the thermoelectric detector, such as simple design, high energy and position resolution, absence of strict requirements to operating conditions, we can state that thermoelectric detectors with multi-layer sensor can be a real competitor to superconducting nanowire detectors. Acknowledgements

This work was supported by the RA MES State Committee of Science and Russian Foundation for Basic Research (RF) in the frames of the joint research projects SCS 15RF-018 and RFBR 15-53-05047 accordingly.

References [1]. G. G. Fritz, K. S. Wood, D. Van Vechten, A. L.

Gyulamiryan, A. S. Kuzanyan, N. J. Giordano, T. M. Jacobs, H.-D. Wu, J. S. Horwits, A. M. Gulian. Thermoelectric single-photon detectors for X-ray/UV radiation, in Proc. SPIE, Vol. 4140, ‘X-Ray and Gamma-Ray Instrumentation for Astronomy XI’, San Diego, CA, December 2000, pp. 459-469.

[2]. A Kuzanyan, V A Petrosyan, A S Kuzanyan. Thermoelectric single-photon detector, Journal of Physics: Conference Series, Vol. 350, 2012, 012028.

[3]. A. S. Kuzanian, V. R. Nikoghosyan, A. A. Kuzanyan. CeB6 Sensor for Thermoelectric Single-Photon Detector, Sensors & Transducers, Vol. 191, Issue 8, August 2015, pp. 57-62.

[4]. A. S. Kuzanyan, V. R. Nikoghosyan, A. A. Kuzanyan. Modeling of Kinetic Processes in Thermoelectric Single Photon Detectors, in Proceedings of the SPIE Vol. 9504, ‘Photon Counting Applications 2015’, Prague, May 2015, 95040O, pp. 1-10.

Page 142: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

142

Ultrasonic Acoustic Beam Modulation in Water

by Pre-fractal Geometries

Juan Carlos Melgarejo 1, Sergio Castiñeira-Ibáñez 2, Daniel Tarrazó-Serrano 1, Constanza Rubio 1*, Pilar Candelas 1, Antonio Uris 1

1 Centro de Tecnologías Físicas, Universitat Politècnica de València, Camino de Vera s/n. 46022 Valencia, Spain

2 Departamento de Ingeniería Electrónica, Universitat de València, Avd. de la Universitat s/n. 46100 Burjassot, Valencia, Spain

Tel.: + 34963879521 E-mail: [email protected]

Summary: The main objective when designing an ultrasonic lens focuses on improving targeting the waves. On the other hand, it is well known the great scientific interest in structures that follow fractal geometries, having been used in multiple applications. The focusing and beam modulation capabilities of a set of annular scatterers with different lacunarities are studied in this work. The results have been obtained using the Finite Element Method (FEM). Keywords: Sound focusing, Ultrasonic lens, Cantor pre-fractals.

1. Introduction

Every structure with the symmetry properties of

scale invariance, even for a short range, is likely to possess fractal properties. It can be said that the whole structure resembles its internal parts [1]. It was found that several natural phenomena follow these patterns. For example, snowflakes or some tree leaves. Fractal structures have attracted the interest of the scientific community, due to its applications in different areas of science and technology [2].

Among the wave applications of fractal structures, different studies related to the acoustic field have been performed. Petri et al. [3] analyzed the vibration properties of a hierarchic system consisting on a Cantor sequence of piezoelectric and resin elements. Sapoval et al. [4] numerically investigated the acoustic properties of irregular cavities described as fractals. They proved that the geometric irregularity improved the modal density of the low frequencies, localizing many of the modes at the edge of the cavity and the attenuation properties of the cavity were modified. Lubniewski and Stepnowski [5] developed a simple method to identify the seabed using elements of fractal analysis. Castiñeira-Ibáñez et al. [6, 7] presented an acoustic barrier to control the noise formed by rigid cylinders in the fractal geometry of a Sierpinski triangle. Gómez-Lozano et al. [8] studied the transmission response in perforated plates with subwavelength holes. The transmission spectrums showed that every iteration with the Sierpinski Carpet has the characteristic peaks and dips associated with the lattice constant of each matrix pattern.

In other branches of acoustics other devices based on diffraction such as Root Primary Gratings (RPG) offers the possibility of modulate the acoustic beams.

Since the wave theory of different areas of physics have great similarity, many concepts used in these

related to the ability to guide the wave propagation studies are extrapolated to the ultrasonic range, and based on these studies, a new line for the construction of lenses is proposed. These lenses are able to modulate the ultrasonic beam in a different way compared to the conventional lenses, where its ability to focus and guide waves comes from the fact that the they are built with refractive materials with curved surfaces. Therefore, in this paper some new fractal acoustic lenses are numerically analyzed using the Finite Elements Method.

2. Numerical Model

In order to obtain a pre-fractal lenses, a 360 degree turning of the 8 objects around the z axis (see Fig. 1), having each object a with W and a height t has been simulated. These 16 radii have been obtained through the expression (1).

2/War ii (1)

Fig. 1. Pre-fractal lens (plane XY) of a Cantor sequence until the second stage (s=2).

Using an axysimmetric model is possible to build a 3D model by 2D simpler one. The results have been obtained using the Finite Elements Method (FEM). This method solves geometric complex structures with different acoustic phenomena. The model solves in the

Page 143: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

143

frequency domain, the wave equation in order to analyze the behavior of the acoustic waves when interacting with isolated scatterers under water. In order to solve the problem it is necessary to define the geometry, to define the boundary conditions and to make discrete the domain. 3. Results

Fig. 2 shows a comparison of the transverse section of the sound pressure level along the z-axes (perpendicular axis to the plane of the lens, XY) of the three lens considered. The three lenses have the same size, the same number of rings but the gaps are distributed in a different manner. The basic structure considered is based on a polyadic Cantor set [9]. Sample 1 is a prefractal lens with a Cantor sequence where the value of the normalized outer gap width in the first stage of growth, , was 0.044. Sample 2 was obtained by redistributing scattering center in Sample 1. The dispersing elements were moved a distance W/2 towards the rotation axis. In Sample 2, was not varied. Sample 3 was obtained by halving its value. In this case = 0.022.

Fig 2. Transverse section of sound pressure level along the z-axis for the pre-fractal lenses studied.

It can be seen that only by redistributing the solid rings the acoustic modulation of the beam can be obtained and focalizing behavior improves considerably. This capability of modulate the beam can

be interesting for certain medical applications such as HIFU. 4. Conclusions Different distributions of gaps of ultrasonic lenses based on fractal geometries has been studied and the modulation of the focus lens by means the variation of this parameter has been proved. . Acknowledgements This work has been supported by the Generalitat Valenciana (AICO/2015/119) and by Spanish MINECO (TEC2015-70939-R). References [1]. Mandelbrot, B. B., The Fractal Geometry of Nature,

WH Freeman and Co., San Francisco, CA, USA, 1982. [2]. Takayasu, H. Fractal in Physical Sciences, Manchester

University Press, Manchester, UK, 1992. [3]. Petri, A., Alippi, A., Bettucci, A., Cracium, F.,

Farrelly, F. Vibrational properties of a continuous self-similar structure, Phys. Rev. B, 1994, 49, pp. 15067–15075.

[4]. Sapoval, B., Haeberlé, O., Russ, S. Acoustical properties of irregular and fractal cavities. J. Acoust. Soc. Am., 1997, 102, pp. 2014–2019.

[5]. Lubniewski, Z., Stepnowski, A. Application of the fractal analysis in the sea bottom recognition, Arch. Acoust., 1998, 25, pp. 499–511.

[6]. Castiñeira-Ibáñez, S., Romero-García, V., Sánchez-Pérez, J. V., García-Raffi, L. M. Overlapping of acoustic bandgaps using fractal geometries, EPL 2010, 92, 24007.

[7]. Castiñeira-Ibáñez, S., Rubio, C., Romero-García, V., Sánchez-Pérez, J. V., García-Raffi, L. M. Design, manufacture and characterization of an acoustic barrier made of multi-phenomena cylindrical scatterers arrangen in a fractal-based geometry, Arch. Acoust., 2012, 37, pp. 455–462.

[8]. Gomez-Lozano, V., Uris, A., Candelas, P., Belmar, F. Acoustic transmission through perforated plates with fractal subwavelength apertures, Solid State Commun., 2013, 165, pp. 11–14.

[9] Jaggard, A. D., Jaggard, D. L. Cantor ring diffractals, Opt. Commun., 1998, 158, pp. 141-148.

Page 144: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

144

Development of Physical Computing Education Systems

for Technical Colleges using Free Software

T. Horiuchi 1, T. Miyazaki 1 , Y. Yodo 1 , Y. Yokoyama 1 , H. Yamamoto 2 and M. Masaaki 3 1 National Institute of Tech., Nagano College, 716 Tokuma, Nagano City 381-0041, Japan 2 Shinshu Univ., Faculty of Engineering, 4-17-1, Wakasato, Nagano City 380-8553, Japan 3 Shinshu Univ., Faculty of Education, 6-Ro, Nishinagano, Nagano City 380-8544, Japan

Tel.: +81-26-295-7105, fax: + 81-26-295-4950 E-mail: [email protected]

Summary: Necessity of physical computing education as well as conventional microcomputer education is advocated by higher educational facilities of the industrial field. We have been developed 3 physical computing education systems, suitable for a Japanese technical college (5-years system). The first system had developed 4 years ago. This system make student do experiments of various sensors and actuators using an Arduino microcomputer. It is used even now and the educational effect is raised. The second system is for the applying physical computing education to make 3D printer. We finished developing it by one previous fiscal year, and we used it in a class of 4th grade in last year, and also received the high evaluation from the students. And the last system is to combine Raspberry Pi microcomputer and an Arduino microcomputer effectively and make an original robot. This development is on the way at present, but we'll practice and estimate at the class in the latter period this year. Keywords: Physical computing education, Arduino microcomputer, Raspberry Pi microcomputer, 3D printer, Sensors.

1. Introduction

Necessity of a Physical Computing [1] (‘PhCom’

for short in the following.) education is pointed out recently, and the approach has started with several educational institutions. Since 2012, for the educational environment of PhCom suitable for technical college students, we have been developed three systems.

2. 1st System

This system consists of Arduino microcomputer and breadboard as hardware, and open-source Arduino Software (IDE) [2, 5] as a software development environment. As controlled parts, we selected temperature / light / distance / 3D acceleration sensors etc. as sensors, and LED, color LED, piezoelectric speaker, servomotor, stepping motor, character LCD and two-dimensional LCD etc. as actuators. And we prepared 50 sets of these parts in order to use 1 set for one student at a class. A picture of the teaching materials set is indicated on Fig. 1.

And we made them use this system at the class of 1st grade and 4th grade of a technical college. They could experiment smoothly because they already acquired “Processing” programming language.

We did a questionnaire to survey and heard a reaction from a student after the experiment. As a result, we got high response more than we expected [3]. In many questionnaire item, Fig. 2 shows a reply result of interested sensor. It shows target of interest is not same by different grade.

Fig. 1. One set of the Arduino experimental kit.

Fig. 2. Result of interests for sensor.

Page 145: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

145

3. 2nd System

The 2nd system is for applied PhCom education for the student who finished the 1st system. It's the first purpose to make a student design and make a plug-compatible Arduino.

First, we make a student learn a circuit diagram editor (BSch3v, free software) and CAD (CADLUSPCB, free software) for printed circuit board design, and make them do a circuit design and a substrate design of a plug-compatible Arduino. The substrate placed an order outside and a student made complete a plug-compatible (left side of Fig. 4) by soldering by him.

Next, we targeted to the 3D printer which is the famous product for which Arduino is used. By the RepRap project, open-hardware and open-software to make a 3D printer by oneself are opened.

First, we selected parts and calculated the expenses. But there was little supply of an inexpensive parts in Japan so we decided to buy the kit. We selected “Folger Tech Kossel 2020 Full 3D Printer Kit” of American Folgertech company. This kit is easy to make and the cost is low because it’s delta-type 3D printer. The student's number of the class was 16. So we bought 4 sets of this and made them put 1 for 4 students together (Fig. 3).

Fig. 3. Making 3D printer by students.

Fig. 4. Plug-compatible Arduino board (left side). ‘Mugbot’ developed by Tokyo City University (right-side).

By a student questionnaire after making, "It was impossible by myself, but the mutual strong point field could be shown with 4 people." "Many things could be acquired about the system, mechanism and software by making 3D printer." etc. comments say that high evaluation could obtained. 4. 3rd system

The 3rd system combines Arduino and Raspberry Pi[4][6] famous for a low-price microcomputer for education, and makes a student manufacture a original robot.

As a foundation of a robot, we selected the “Mugbot” (right-side of Fig. 4) developed by Tokyo City University. Because these hardware and software are all opened, it is most suitable for student's education. This uses some LEDs for eyes and the mouth and two servomotors for a rotation for head and neck. And also Japanese is vocalized from a speaker by speech synthesis. Arduino takes charge of control of LEDs and servomotors. On the other hand some servers (HTTP, DHCP, WebSocket etc.) is put in Raspberry Pi and it can be operated remotely from a web browser of a PC and a smart phone.

The purpose of our system is to compose cheaply system that can communicate with a man and a robot using voice and a countenance. For this, sensor (temperature, humidity and air pressure etc.), microphone and camera are added for original robot. Moreover free software for speech recognition and image recognition can be used.

This system is being designed at present, but we are planning to use and estimate it in the 4th grade's class in the latter period of current year. 5. Conclusions

The most advanced PhCom education could be performed now in technical colleges at low cost by use of 3 above mentioned systems. We think this is very useful for higher educational facilities of the world as well as Japan. We'd like to exhibit these outcomes to the world through a web from now on. Acknowledgements

This work was supported by JSPS KAKENHI Grant Numbers JP 26350356 and JP 16K00260. References [1]. Dan O. Sullivan, Tom Igoe, Physical Computing -

Sensing and Controlling the Physical World with Computers, Course Technology Ptr, 2004.

[2]. M. Banzi, Getting Started with Arduino 2nd Edition, O’Reilly Media Inc, 2011.

Page 146: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

146

[3]. Yusuke Yodo, Takashi Miyazaki, Taisuke Horiuch, Noriyuki Tanaka, Naruki Sirahama, Mutual Evaluation System Using Video and Web for Making Circuit Experiment, The 9th International Symposium on Advances in Technology Education (ISATE’15), Nagaoka, Japan, Vol. 9, 2015, pp. 122-127.

[4]. Simon Monk, Raspberry Pi Cookbook, 2nd Edition, O’Reilly Media Inc, 2016.

[5]. Arduino - Home (https://www.arduino.cc/). [6]. Raspberry Pi - Home (https://www.raspberrypi.org/).

Page 147: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

147

Optical Processor Design for Data Error Detection and Correction Using a (9,5 ) Binary Code Generator and the Syndrome Decoding Process

M. A. Vieira 1,2, M. Vieira 1,2,3, P. Louro 1,2

1 Electronics Telecommunication and Computer Dept. ISEL, R. Conselheiro Emídio Navarro, 1949-014 Lisboa, Portugal

2 CTS-UNINOVA, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal. 3 DEE-FCT-UNL, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal

1 Tel.: +351919310252, fax: +351218317144 1 E-mail: [email protected]

Summary: Based on a-SiC:H technology, we present an optical processor for data error detection and correction using a suitable (9,5) Hamming binary code generator and the syndrome decoding process. The optical processor consists of an a-SiC:H double p-i-n photodetector with two ultraviolet light biased gates. The relationship between the optical inputs (transmitted data) and the corresponding output levels (the received data) is established and decoded. Results show that under irradiation the device acts as an active filter. Under front irradiation the magnitude of the short wavelength is quenched and in the long wavelength range is enlarged, while the opposite happens under back lighting. Parity bits are generated and stored simultaneously with the data word. Parity logic operations are performed and checked for errors together. An all-optical processor for error detection and correction is presented to provide an experimental demonstration of this fault tolerant reversible system. Two original coloured string messages, having 4- and 5- bits, respectively, are analysed and the transmitted 7- or 9- bit string, the parity matrix, the encoding and decoding processes, are explained. The design of SiC syndrome generators for error correction is tested. Keywords: Amorphous SiC technology, Error detection and correction, Syndrome decoding process, Hamming binary code.

1. Introduction

In digital transmission systems, an error occurs when a bit gets altered between transmission and reception. Errors detection codes are generated as a function of the bits being transmitted. Such codes are appended to the data bits and transmitted together. The receiver evaluates the code based on the incoming bits and compares it with the incoming code to check for errors. Error coding uses mathematical formulas to encode data bits at the source into longer bit words for transmission [1]. The code word can then be decoded at the destination to retrieve the information. The extra bits, in the code word, provide redundancy that, according to the coding scheme used, will allow the destination to use the decoding process to determine if the communication system has introduced errors and, in some cases, correct them avoiding the need of data retransmission.

Tandem monolithic Si/C structures based on amorphous silicon technology can be reconfigurable to perform optoelectronic logic functions [2] due to their nonlinear magnitude-dependent response to each incident light wave, in the visible range. Based, in this properties we built an optical.

2. Experimental Details

The active device is a double pi’n/pin a-SiC:H photodetector produced by Plasma Enhanced Chemical Vapor Deposition (Fig. 1). It consists of a

p-i'(a-SiC:H)-n/p-i(a-Si:H)-n heterostructure with high resistivity 20 nm thick doped layers (>107 cm) packed between two transparent oxide layers (TCO) made of ITO.

Transmitted data[RGBV PRPGPB ]

p

Glass

i’

200 nm

a‐SiC:H

p

i

1000 nm

a‐Si:H

n nTCO

TCO

V

OpticalB

ias

B

G

R

Inputchannels

Front diodes Back diodes

V=-8VTCO

p n p n

TCO

G

R

i’

B

i

V=-8V

OpticalB

ias

MUX CODE

MUX PA

RITY

Fig. 1. Receiver configuration and operation.

The data is transmitted using four monochromatic (red, green, blue and violet; R,G,B,V;) pulsed communication channels that are mixed together, each one with a specific bit sequence and are absorbed in different regions depending on their wavelengths. The combined optical signal (multiplexed signal) is received and analysed by reading out the generated photocurrent under negative applied voltage, and 390 nm background lighting, applied either from the

Page 148: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

148

front or the back sides of the device. Here, for parity check, three or four red, green, blue and violet synchronous channels, respectively, were read in simultaneous with the data code. As an application, data was sent through one detector while error detection and correction bits were sent through the other.

3. Results

Fig. 2, show the MUX signal (solid lines) that arises from to the transmission of the four (Fig. 2a) or five (Fig. 2b) wavelength channels. The dotted line marks the generation of the synchronized parity MUX transmitted in simultaneous with the data code. The eight, or the sixteen, ordered levels of the parity bits are marked as horizontal dash lines in Fig. 2a and Fig. 2b, respectively. In the right side of figures the eight or the sixteen sublevels and their 3- or 4-bit binary codes are inserted, respectively. On the top the seven or the nine bit word [R,G,B,V (I), PR, PG, PB

(PV)] of the transmitted inputs guides the eyes.

0.0 0.5 1.0 1.50.0

0.2

0.4

0.6

0.8

1.0

1.1

1.3

pR p

G p

B

0 0 0

0 0 1

0 1 00 1 1

1 0 01 0 1

1 1 01 1 1p

7p

6

p5

p4

p3

p2

p1

p0

Cod

e w

ord

d15

d0

d3

d4

d7

d8

d11

d12

PBPG

PR

VB

GR

MU

X s

igna

ls

T ime (ms)

(a)

0.0 0.5 1.0 1.5 2.0 2.5 3.00.0

0.3

0.5

0.8

1.0

1.3

1.5

1.8

1 1 0 0

1 1 1 1

pR p

G p

B

d27

d4

d8

PVPBPGPRI

VBGR

d17

p0

p15

Cod

e w

ord

d0

d15

d31

MU

X s

igna

ls

Time (ms)

(b)

Fig. 2. Code and parity MUX signals under 390 nm front irradiation. On the top the transmitted channels [R G B V

PR PG PB] are shown. a) Four channel transmission; b) five channel transmission.

4. Decoding Algorithm In order to automate the process of recovering the

original transmitted data an algorithm was developed and implemented. The transmitted information is decoded by comparing the MUX signal from the code with the simultaneous parity MUX signal under front irradiation, as shown in Fig. 3. The decoding algorithm is based on a proximity search after each time slot is translated to a vector in multidimensional space. The vector components are determined by the signal currents I1 and I2, where I1 (d levels) and I2 (p levels) are the currents measured simultaneously under front optical bias for a 5-bit codeword (RGBVI) if five channels are used for the word transmission. The result is then compared with all vectors obtained from a calibration sequence (Fig. 2b) where to each code level, d (0-31), is assigned the correspondent parity level, p(0-15). The colour bits of the nearest calibration point are assigned to the time slot. An Eucledian metric is applied to measure the distances. We have tested the algorithm with different random sequences of the channels and we have recovered the original colour bits, as shown in the top of Fig. 3.

0.0 0.5 1.0 1.5 2.0 2.5 3

0.00.20.40.60.81.01.21.41.61.82.02.2

Parity levels

Word levels

PVPBPGPRI

VBGR

Pho

tocu

rren

t [A

]

Time [ms]

Fig. 3. Code and parity MUX/DEMUX signals under 390 nm front irradiation. On the top the decoded 9-bit word

code [R G B V I PR PG PB PV] are displayed. 5. Conclusions

Based on a-SiC:H technology, we presented an optical processor for data error detection and correction using a suitable (9,5) Hamming binary code generator and the syndrome decoding process.

Results show that by comparing the MUX signal due to the received data code word with the generated parity MUX signal, two consecutive levels in the data code (RGBVI) do not correspond to two near levels in the parity levels [PR PG PB PV] allowing information retrieval.

Page 149: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

149

References [1]. R. W. Hamming, Error detecting and error correcting

codes, Bell Syst. Tech. J., 29, 1960, pp. 147–160.

[2]. M. A. Vieira, M. Vieira, J. Costa, P. Louro, M. Fernandes, A. Fantoni, Double pin Photodiodes with two Optical Gate Connections for Light Triggering: A capacitive two-phototransistor model, Sensors & Transducers, 9, Special Issue, 2010, pp. 96-120.

Page 150: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

150

MUX/DEMUX Si/C Device with Five Channel Separation

in the Visible Range

M. Vieira 1,2,3, M. A. Vieira 1,2, P. Louro 1,2

1 Electronics Telecommunication and Computer Dept. ISEL, R. Conselheiro Emídio Navarro, 1949-014 Lisboa, Portugal

2 CTS-UNINOVA, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal. 3 DEE-FCT-UNL, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal

1 Tel.: +351919310252, fax: +351218317144 1 E-mail: [email protected]

Summary: In this paper we present a selector based on a multilayer a-SiC:H optical filter that requires appropriate near-ultraviolet steady states optical switches to select the desired wavelengths in the visible range. The selector filter is realized by using a two terminal double pi’n/pin a-SiC:H photodetector. Five visible communication channels are transmitted together, each one with a specific bit sequence. The combined optical signal is analyzed by reading out the photocurrent, under near-UV front steady state background. Data show that 25 current levels are detected and correspond to the thirty-two on/off possible states. Results show that the background works as a selector in the visible range, shifting the sensor sensitivity and allows the identification and decoding of the different input channels. A transmission capability of 60 Kbps using the generated codeword was achieved. Keywords: Amorphous SiC technology, MUX/DEMUX device, Visible light communication, Decoding algorithm. 1. Introduction

LEDs are a very effective lighting technology due

to high brightness, long life, energy efficiency and affordable cost. Their use as communication device with a photodiode as receptor has been experienced for many years. Due to the increasing LED lighting in homes and offices, the idea to use them for visible light communications (VLC) has come up recently [1]. To enhance the transmission capacity and the application flexibility of optical communication, efforts have to be considered, namely the use of WDM devices based on a-SiC:H light controlled filters, when different visible signals are encoded in the same optical transmission path. We present a selector based on a multilayer a-SiC:H optical filter that requires appropriate near-ultraviolet steady states optical switches to select the desired wavelengths in the visible range. A 1 x 5 wavelength division multiplexer SiC device with channel separation in the visible range is presented.

2. MUX/DEMUX Design and Operation The WDM device is shown in Fig. 1 and is a

double pi’n/pin a-SiC:H photodetector produced by Plasma Enhanced Chemical Vapour Deposition. The deposition conditions and optoelectronic characterization of the single layers and device as well as their optimization were described elsewhere [2, 3]. The device acts as an active filter, confining the short wavelength optical carriers in the front photodiode while the long ones are absorbed into the back photodiode (see arrow magnitudes). The medium

wavelength ranges are absorbed differently across both [4].

The data is transmitted using monochromatic (red, green, blue and violet; R,G,B,V;) pulsed communication channels (input channels; data code) that are mixed together, each one with a specific bit sequence and are absorbed in different regions depending on their wavelengths (see arrows in the picture). The combined optical signal (multiplexed signal) is received and analysed by reading out the generated photocurrent under negative applied voltage (-8 V), and 390 nm background lighting, applied from the front or the back sides of the device. In a four channel data transmission, the square wave of four modulated low power lights of red (R: 626 nm), green (G: 524 nm), blue (B: 470 nm) and violet (V: 400 nm) LEDs were used. In the five channel transmission an extra near infrared (I: 700 nm) LED was added.

Fig. 1. Receiver configuration and operation. In order to analyses the background effect in each

channel the five monochromatic input channels

Page 151: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

151

illuminated the device separately (transmitted data) at 12000 bps. The generated photocurrent was measured under front and back steady state irradiation. For each channel, the gain (αV

R,G,B,V), defined as the ratio between the photocurrent with and without applied background, was determined. Data measured confirms that the optical gain, under irradiation depends on the irradiated side and on the incoming wavelength acting the device as an active filter for the input channels. In Table 1 the optical gains of the input channels used in the transmission are displayed under 2800 µWcm-2

front (αF) and back (αB) irradiation.

Table 1. Front (αF) and back (αB) optical gains.

(nm) αF αB 400 0.90 11.60 470 1.07 1.96 524 3.55 0.57 626 4.70 0.45 700 5.55 0.40

3. Decoding Algorithm

In Fig. 2, the received data, i.e. the MUX code signal, due to the combination of four (Fig. 2a; 400 nm, 470 nm, 524 nm, 626 nm) or five (Fig. 2b; 400 nm, 470 nm, 524 nm, 626 nm, 700 nm) input channels are displayed under front irradiation. At the top, the decoded input channels (transmitted data) are shown to. Results show that, in Fig. 2a, the selection index for the 16-element look-up table (d0-d15, dotted levels) is a 4-bit binary code (RGBV) and in Fig. 2b the selection index for the 32-element look-up (d0-d31, dotted levels) is a 5-bit (RGBVI) binary code.

The algorithms to decode are relatively simple and the idea of the background acting as selector that chooses one or more of the 2n sublevels, with n the number of transmitted channels, and their n-bit binary code makes the communication reliable [5]. Results show that each one of the possible 2n on/off states corresponds to a well-defined level. In Fig. 2, all the on/off states are possible so, 2n ordered levels pondered by their optical gains (Table I) are detected and correspond to all the possible combinations of the on/off states [6]. So, by assigning each output level to an n digit binary code the signal can be decoded. A maximum transmission rate capability of 60 Kbps was achieved in a five channel transmission. 4. Conclusions

An a-SiC 1x5 WDM device, with channel separation in the visible range, was presented and its configuration and operation as wavelength selector explained. Encoding and decoding data was analyzed and the codewords generated using the MUX signals due to the data transmitted tested. An algorithm to

decode the transmitted information was presented. A transmitter capability of 60 kbps using the generated codeword was achieved.

0.0 0.5 1.0 1.50.0

0.5

1.0

1.5

d0

d3

d5

d7

d9

d11

d13

d15

400 nm470 nm

RGBV

V

=390 nm

0001

0

0011

0

0101

0

0111

0

1001

0

1011

0

1101

0

626 nm524 nm

1111

0

MU

X s

ign

al

T im e (m s)

(a)

0,0 0,5 1,0 1,5 2,0 2,5 3,00,0

0,5

1,0

1,5

2,0

00000

11110

11111

RGBVI

=390 nm

d31

d15

d0

Front

700 nm400 nm470 nm524 nm626 nm

MU

X s

igna

l (A

)

Time (ms)

(b)

Fig. 2. MUX/DEMUX signals under 390 nm front and back irradiation. On the top the transmitted channels are

displayed. a) Four channels transmission; b) Five channels transmission.

References [1]. D. O’Brien, H. L. Minh, L. Zeng, G. Faulkner, K. Lee,

D. Jung, Y. Oh, and E. T. Won, Indoor visible light communications: challenges and prospects, Proc. SPIE, 7091, 2008, 709106.

[2]. M. Vieira, P. Louro, M. Fernandes, M. A. Vieira, A. Fantoni and J. Costa, Advances in Photodiodes, InTech, Chap. 19, 2011, pp. 403-425.

[3]. M. A. Vieira, P. Louro, M. Vieira, A. Fantoni, and A. Steiger-Garção, Light-activated amplification in Si-C tandem devices: A capacitive active filter model, IEEE Sensor Journal, 12, No. 6, 2012, pp. 1755-1762.

[4]. M. Vieira, M. A. Vieira, P. Louro, J. Costa, M. Fernandes, A. Fantoni, and M. Barata, Multilayer architectures based on a-SiC:H material, Tunable wavelength filters in optical processing devices, J. Nanosci. Nanotechnol. Vol 11, No. 6, 2011, pp. 5299-5304.

Page 152: Sensors and Electronic Instrumentation Advances · Sensors and Electronic Instrumentation Advances: ... International Conference on Sensors and Electronic Instrumental Advances 2016

2nd International Conference on Sensors Engineering and Electronics Instrumental Advances (SEIA' 2016), 22-23 September 2016, Barcelona, Spain

152

[5]. M. A. Vieira, M. Vieira, V. Silva, P. Louro and M. Barata., Optoelectronic logic functions using optical bias controlled SiC multilayer devices, MRS Proceedings, Vol. 1536, 2013, pp. 91-96.

[6]. M A Vieira, M Vieira, V Silva, P Louro, J Costa, Optical signal processing for data error detection and correction using a-SiCH technology, Phys. Status Solidi C, 12, No. 12, 2015, pp. 1393–1400