Development of Sensors for AutomotivePEM-based Fuel Cells

DOE Agreement DE-FC04-02AL67616

DOE Hydrogen and Fuel Cells 2004 Annual Merit ReviewMay 26, 2004

UTC FC Series 200 - 50 kW PEM

MATERIALS

NEXTECH

Nancy Garland - DOE

Tom Clark – UTC FC

This presentation does not contain any proprietary or confidential information

2

Sensors for Automotive PEM Fuel Cells – Objectives

Sulfur

sensor

NH3, T, flow

rate sensors

NH3, T, flow

rate sensors

Sulfur

sensor

NH3, T, flow

rate sensors

NH3, T, flow

rate sensors

•Chemical sensors

–Process streams: before, in, and after reformer, before and in fuel cell stack: CO, H2, O2, H2S, NH3; Safety [H2].

•Physical Sensors

–Temperature, pressure, relative humidity, flow, P

Develop a technology and commercial supplier base for physical and chemical

sensors required to optimize the operation of PEM fuel cell power plants for

automotive applications with path to low cost (<$20 / sensor) at 500k qty.

3

Sensor Program Team Responsibilities

• Sensor development program utilizes a team approach

– UTRC for physical and chemical sensor evaluation and program coordination

– Illinois Institute of Technology (IIT) for chemical sensor evaluation

– Advanced Technical Materials (ATMI) for MEMS sensor development

– NexTech Materials for electrochemical and solid state sensor development

Team

Member

T P RH flow O2 CO H2 SO2 H2S NH3 Technological Expertise /

Responsibility

UTC FC X X X X X X X X X X Testing on S300

Breadboard

UTRC X X X X X X X X X X Testing in reformate

simulator

ATMI X X X X Develop Using MEMS

Silicon Microhotplate

IIT X X X X X X X X Testing in Benchmark

Facility

NexTech X X X X Develop Using Solid State

Electrochemical

4

Sensor Program Team Structure

• Continuous interaction among team members

• ATMI, NexTech develop sensors, IIT and UTRC test and aid in optimization

Sensor Suite

to DOE

Sensor Test in

S300 PEM Plant

MEMS Sensor

– ATMI

H2, H2S, SO2, NH3

Solid State/Chem.

Sensor – NexTech

CO, H2S, SO2, NH3

Physical Sensor

Evaluation –

UTRC

Benchmark

Sensor Test –

IIT

Sensor Test &

Refinement –

UTRC

Program Lead

And Evaluation

UTC FC/UTRC

Sensor Suite

to DOE

Sensor Suite

to DOE

Sensor Test in

S300 PEM Plant

Sensor Test in

S300 PEM Plant

MEMS Sensor

– ATMI

H2, H2S, SO2, NH3

MEMS Sensor

– ATMI

H2, H2S, SO2, NH3

Solid State/Chem.

Sensor – NexTech

CO, H2S, SO2, NH3

Solid State/Chem.

Sensor – NexTech

CO, H2S, SO2, NH3

Physical Sensor

Evaluation –

UTRC

Physical Sensor

Evaluation –

UTRC

Benchmark

Sensor Test –

IIT

Benchmark

Sensor Test –

IIT

Sensor Test &

Refinement –

UTRC

Sensor Test &

Refinement –

UTRC

Program Lead

And Evaluation

UTC FC/UTRC

Program Lead

And Evaluation

UTC FC/UTRC

5

Sensors for Automotive Fuel Cells Plan

Calendar Year2002 2003 2004 2005

2.3 Benchmark Facility Testing

1.0 Physical Sensor Evaluation

2.1 Electrochemical Sensor Development

2.2 MEMS Sensor Development

2.5 S300 Gasoline PEM Fuel Cell Testing

NexTech

IIT

ATMI

UTRC /UTC FC

Legend

2.4 Simulated ReformerStream Testing

3.0 Project Management and Reporting

Done except for Honeywell

Sensor Survey completed

6

Sensors Program Financial Status

•Total cost: $3.7MM; DOE cost: $3.0MM (80%) UTC Cost Share: $0.7MM (20%)

•Total expended to date: $1.6MM

•Duration: April 2002 – March 2005

0

500

1000

1500

2000

2500

3000

3500

4000

Nov-01 Feb-02 May-02 Sep-02 Dec-02 Mar-03 Jun-03 Oct-03 Jan-04 Apr-04 Aug-04 Nov-04 Feb-05 May-05

Date

PV

- $

k

UTRC IIT NexTech ATMI Total Plan Total Expenditure

7

H2 Safety Issues Associated with Project

• Use of H2 in laboratory environment

– Flammable gas detectors located in laboratory; relay opens and turnsoff power to solenoid valves on H2 supply at 10% of LEL

– LabView-based control program senses alarm, shuts off all othergases and purges all gas lines with N2

– All valves used in experiment are explosion-proof

– Pressure relief valves used in all piping to prevent over-pressurizationof components

• Sensor technology– Heated sensing elements can provide an ignition source; therefore the

detection element must be separated from the gas stream by a flash-arrestor (porous plate) to prevent ignition of the bulk gas

8

PEM Fuel Cell Gas Stream Simulators at UTRC & IIT

UTRC test rig with dual chambers

IIT test rig

Test chamber(25 - 450°C)Pressure: 1-4 atm

Both test rigs operate under

LabView control for 24/7

operation (data acquisition

and test matrix completion)

9

Sensor Evaluation Status at UTRC

• Physical Sensors

– Sensors for T, P, P, Relative Humidity (RH), and Flow evaluated in PEM fuel

cell simulator in near-condensing flow regime

• State-of-the-art physical sensors meeting program needs selected

• Chemical Sensors

– First round of sensor testing and qualification completed

– Multiple H2 sensors evaluated for sensitivity, selectivity, and performance

– Possible extension of the testing effort beyond April 2005 being considered in

order to accommodate field testing requested by Honeywell

Lei Chen and Brian Knight

10

Physical Parameter Sensors Results

Responsefluctuation due tocondensation

Most costeffective

Thermaldissipation

Flow

Improve recoveryfrom condensingflow regime

0 to 180 °C, 0-100% RH

Polymercapacitive

(Panametrics)

RH

May be massproduced andminiaturized

Silicon based ICcompatiblefabrication.

Strain gauge

(Druck)Pressure

Response timeneedsimprovement

0 to 250 °C,

-40 to 750 °C

ThermistorTemperature

DevelopmentNeeds

PositiveAttributes

OperatingPrinciple

Sensor

•UTRC researched and tested multiple physical

sensors; most promising tabulated below

11

• IIT evaluated over 70 H2 sensing technologies

• Tiered approach used to evaluate sensor technologies– Gas concentration, operating temperature, water vapor pressure

– Effect of pressure, other background gases

– Long-term testing

• Hydrogen Sensors (Reformer)-H2 Scan, Makel Engineering, ATMI, KSC NASA

• Hydrogen Sensors (Safety Application)-H2 Scan, Applied Sensors, Makel Engineering, ATMI, Figaro, TransducerTechnology, Inc., Argus Group, Nemoto Environmental Technology, AppliedNanotech

• Carbon Monoxide Sensor-NexTech Materials

Benchmark Testing of Viable Sensor

Technologies

Joseph R. Stetter, William R. Penrose, William Buttner, and Kapil Gupta

(Sensors currently available are listed in blue)

12

Literature search and review (for fuel cell sensors and H2/CO sensors)

Researched and short listed tentative companies, based on ourrequirement specifications, vendor products and application

Evaluated the survey responses and accordingly sent out formalinvitations for evaluation of sensors

Now acquiring sensors- NDAs/other formalities

Testing acquired sensors and updating Sensor research DatabaseIntelligent Optical Systems, NASA/KSC/ASRC, NGK

Contacted the companies and sent out sensor survey templates

Process for Selection of Viable Sensor

Technologies

SSTUF: Hydrogen Sensor Response (0.5 to 8%) in air

Single Data Run

- Sensitivity Curves obtained for different pressures at 22oC

- Automated Pressure Control, Flow Control and Concentration

- Capabilities also include Temperature Control and Humidity Control

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 50 100 150 200 250 300

Time (minutes)

1 ATM 2 ATM 3 ATM

PH2 Dependence

(next slide)

Example of Shared Sensor Testing User

Facility Data

Hydrogen Sensor Response (0 to 0.2 atm) in air

Sensor Sensitivity is often controlled by Partial Pressure

of H2 (not %H2)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 0.04 0.08 0.12 0.16 0.2

Partial Pressure Hydrogen (atm)

Sen

so

r R

esp

on

se

Ptotal = 3 atm

Ptotal = 2 atm

Ptotal = 1 atm

Results of IIT H2 Safety Sensor Testing in Air

15

• Targets

– [H2]: 0-10%; Temp: –30 to 80oC, Response time: < 1 s; Humidity: 10-98%; Selectivity from hydrocarbons; Accuracy: 5%; Lifetime:5 yrs

• Approach– Fundamental materials engineering and process control

– Optimization of operating conditions

• Accomplishments• Developed and tested alpha, beta systems

• Demonstrated performance against performance targets

• Delivered alpha prototypes for IIT, UTRC for evaluation

MEMS Sensor Development

Task 1a Safety Sensor in Ambient Air

Pd BarrierRare Earth Hydride

Pd BarrierRare Earth Hydride

BarrierBarrier

BarrierBarrier

Microhotptate Platform Microhotptate Platform

Ing-Shin Chen, Phil Chen, F. DiMeo, Jeff Neuner, Andreas Roehrl, Jim Welch

16

MEMS Sensor Development

Task 1a: Safety Sensor in Ambient Air

• Performance Demonstrated to date

– [H2]: 0-12.8%; Operating Temp: ~80oC,

– Response time: < 2 s @ 4%, 1.2s @ 6%

– Environment: 0–75% RH;

50

40

30

20

Sen

so

r R

esp

on

se(A

.U)

50403020100

Time(min)

0.01

0.1

1

10

[H2

] (%

)

50

40

30

20

Sensor Response in 75% Humid Air

Sensor Response in Dry Air H2 Concentration

t90= 35.3 s t90= 12.1 s t90= 12.0 s t90= 7.0 s t90= 5.3 s t90= 2.4 s t90= 1.2 s t90 = 1.8 s

t90 = 37.8 s t90 = 18.5 s t90 = 12.2 s t90 = 6.5 s t90 = 4.8 s t90 = 1.95 s t90 = 1.2 s t90 = 1.2 s

0.1% 0.2% 0.4%

0.8% 1.6% 3.2%

12.8%6.4%

17

• Targets

– [H2]: 1-100%; Temp: 70- 150oC; Response (T90 ):0.1-1 s; Environment: 1-3 atm total pressure, 10-30 mole % water, total H2, 30-75%, CO2, N2

Accuracy: 1-10 % full scale

• Approach– Materials modifications of safety sensor design

– Exploration of different transduction modes.

• Accomplishments• Fabricated new materials combinations

• Investigated new transduction methods

• Delivered alpha prototypes to UTRC

MEMS Sensor Development

Task 1b Pre Stack Monitor

Microhotptate Platform

New Top Layers

Piezo Resistive Transduction

BarrierBarrier

BarrierBarrier

18

MEMS Sensor Development

Task 1b Pre Stack Monitor

• Performance in Dry N2

– 0-4-40% H2

• 37 sec t90 0 - 4%

• 2 sec t90 4 - 40%

– 40 to 10% H2

• 31.8 sec 0-40%

• Performance in 70% RH

– Similar to dry N2

BQ-6-6B

420

400

380

360

340

320

300

280

SB

R(O

hm

s)

6050403020100

Time(min)

40

20

0

[H2

] (%

)

2.780

2.770

2.760HB

R(k

Ohm

s)

600

500

400

300

200

SB

R(O

hm

s)

50403020100

Time(min)

4020

0

[H2

] (%

)

2.94

2.92

2.90

2.88

2.86

2.84

HB

R(k

Oh

ms)

82

80

78

76

74

72

70

68

Re

sis

tan

ce

(Oh

ms)

6040200

Time(min)

40

20

0

[H2

] (%

)2930

2920

2910

2900

Rh

(Oh

ms)

Sensor A Dry

Sensor A in 40% Rh Sensor A Heater Resistance without Humidity Sensor A Heater Resistance with Humidity

H 2 Concentration

19

MEMS Sensor Development

Task 2 H2S Sensor Development

• Targets– Temp: 400oC; Range: 0.05 ppm -0.5 ppm; Response time: < 1 min at

0.05 ppm; Environment: H2,CO, CO2 H2O

• Approach– Ultra thin ( < 50nm) metal film deposition on micro hotplate platform

• Accomplishments– Demonstrated first sensor response to H2S

– 50 nm film responds to H2S• 160°C, 4% H2/N2,

• 20% RH,

• 180 ppm H2S

120

115

110

105

100

95

90

Resis

tance (

ohm

)

2520151050

Time (min)

43210

% H

2

150100

500

H2S

(ppm

)

'% H2' '% H2S' 'Sensor Response'

20

NexTech Materials Sensor DevelopmentMATERIALS

NEXTECH

Mixed Potential

V

Thin

GDC/YSZ

disc

Bianodes

Furnace

Seal

Fuel Cell

IDE Thick

Film

Sensor Platforms

Scott L. Swartz, Ph.D. (P.I.), Chris Holt, Todd G. Lesousky

21

MATERIALS

NEXTECHNexTech Sensor Development

Task 2.1.1 Miniature SOFC Fuel Cell Sensor

• NexTech’s SOFC sensor technology with electrodes engineered to respond to CO show reversible and quantitative response to CO in wet N2/H2.

• Future work will focus on schemes to improve sensitivity for 0-100ppm CO range and testing cross-sensitivity to alternate syngas components

22

MATERIALS

NEXTECH

• Metal oxide based chemi-resistor (not electrochemical sensor) exhibits reversible and quantitative response to H2S

• NexTech is currently evaluating various dopant schemes to reduce thetemperature of operation• Beta prototypes scheduled for early June

NexTech Sensor Development

Task 2.1.2 Hydrogen Sulfide Sensors

23

MATERIALS

NEXTECH

• Metal Oxide films show reversible response to H2S concentrations at 0.5 ppm in syngas (goal of 0.05 – 0.5 ppm).

•Future work will focus on measuring lower sulfur concentrations and cross-sensitivity to individual syngas components.

NexTech Sensor Development

Task 2.1.2 Hydrogen Sulfide Sensors

24

MATERIALS

NEXTECH

NexTech’s metal halide ammonia sensorshows very high sensitivity at lowtemperature

Future work will focus on improving hightemperature sensitivity and measuringcross-sensitivity to other syngascomponents.

Sensor responds reversibly in N2/H2

at 75ºC

NexTech Sensor Development

Task 2.1.3 Ammonia Sensor Response

25

Responses to Previous Year Reviewers’ Comments

• “..difficult to assess technical approach and progress”– Physical sensor evaluation completed

– H2 LEL sensor developed• Best response times <1 s, average ~14s; sensor drift rate < 0.16% / day

– Stack H2 sensor developed• Dynamic response up to 40% H2, H2 levels up to 70%, with humidity

• Fast response (T90<2 sec) with Pd

• New devices shows promise; minor cross sensitivity with CO; Drift <0.2% in 4% H2

– Multiple strategies identified for sensing CO in reducingenvironments; CO sensitivity established in humid environments

– Multiple strategies for sulfur• ATMI- 50 nm Metal Foil shows response to H2S

– NexTech

– H2S/SO2 sensor materials identified

– PPM level detection demonstrated

– Ammonia sensor easily packaged in a chemi-resistor format

Sensors for Automotive PEM-based Fuel Cells Project

Contractor and subcontractor PIs:

Name Affiliation Phone E-mailTom Clark UTPWR 860-727-2287 tom.clark@utpwr.com

Brian Knight UTRC 860-610-7293 knightba@utrc.utc.com

Frank DiMeo ATMI 203-794-1100 x4279 fdimeo@atmi.com

Joe Stetter IIT 312-567-3443 stetter@iit.edu

Scott Swartz NexTech 614-842-6606 x103 swartz@nextechmaterials.com

MATERIALS

NEXTECH

Team organization

DOE program manager and technical advisor:

Name Affiliation Phone E-mailNancy Garland DOE 202-586-5673 nancy.garland@ee.doe.gov

Robert Sutton ANL 630-252-4321 sutton@cmt.anl.gov