Loughborough UniversityInstitutional Repository

Cupric oxide-based p-typetransparent conductors

This item was submitted to Loughborough University's Institutional Repositoryby the/an author.

Citation: ISHERWOOD, P.J.M. and WALLS, M., 2014. Cupric oxide-basedp-type transparent conductors. Energy Procedia, 60, pp.129-134.

Additional Information:

• This is an Open Access Article. It is published by Elsevier under theCreative Commons Attribution-NonCommercial-NoDerivs 3.0 UnportedLicence (CC BY-NC-ND 3.0). Full details of this licence are availableat: https://creativecommons.org/licenses/by-nc-nd/3.0/. This paper wasalso presented at the E-MRS Spring Meeting 2014 Symposium Y �Ad-vanced materials and characterization techniques for solar cells II�, 26-30May 2014, Lille, France.

Metadata Record: https://dspace.lboro.ac.uk/2134/35038

Version: Published

Publisher: Elsevier c© The Authors

Rights: This work is made available according to the conditions of the CreativeCommons Attribution-NonCommercial-NoDerivatives 3.0 Unported (CC BY-NC-ND 3.0) licence. Full details of this licence are available at: http://creativecommons.org/licenses/by-nc-nd/3.0/

Please cite the published version.

Energy Procedia 60 ( 2014 ) 129 – 134

Available online at www.sciencedirect.com

ScienceDirect

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of The European Materials Research Society (E-MRS)doi: 10.1016/j.egypro.2014.12.354

E-MRS Spring Meeting 2014 Symposium Y “Advanced materials and characterization techniques for solar cells II”, 26-30 May 2014, Lille, France

Cupric oxide-based p-type transparent conductors

P J M Isherwood*, J M Walls CREST, School of Electronic, Electrical and Systems Engineering, Holywell Park,

Loughborough University, Loughborough, Leciestershire,LE11 3TU UK

Abstract

This study examines the impact of doping on the resistivity of sputtered cupric oxide (CuO), and investigates the effects of co-sputtering CuO with tin dioxide (SnO2). It was found that films sputtered from a 2 at. % sodium-doped target have resistivities of four orders of magnitude lower than equivalent undoped films. Addition of oxygen was found to reduce the resistivity further. The best films were found to have resistivities of 4.3x10-2 .cm. Co-sputtering with SnO2 was found to increase the band gap significantly, although it also caused an increase in the resistivity. All mixed oxide films were both amorphous and p-type. © 2014 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of the SiliconPV 2014 conference.

Keywords: P-type TCO; CuO; Cupric oxide; Amorphous TCO; Mixed oxide

1. Introduction

Metal oxide semiconductors find widespread use as transparent contact materials in solar cells [1]. However, the vast majority are n-type. The development of p-type transparent conducting oxide materials would be useful for a range of photovoltaic (PV) technologies such as bi-facial and multijunction cells, and as possible back contacts for cadmium telluride and organic solar cells [2], [3].

Various candidate p-type transparent conductors have been identified, but as yet few, if any, have been

* Corresponding author. Tel.: +44 (0)1509 635306

E-mail address: P.J.M.Isherwood@lboro.ac.uk

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of The European Materials Research Society (E-MRS)

130 P.J.M. Isherwood and J.M. Walls / Energy Procedia 60 ( 2014 ) 129 – 134

demonstrated to show the necessary low resistivity for widespread use [2]. This is because of the typical metal oxide band structure and high electronegativity of oxygen. This results in a narrow and flat valence band consisting primarily of oxygen 2p states [2], [4]–[6]. This means that metal oxides typically have a very high hole effective mass [2], [4]. Techniques for widening the valence band so as to increase hole mobility largely involve the use of metal cations which have valence electrons at the same or similar energy levels to the oxygen 2p states. Of these, the most commonly used are copper and tin [2].

Cupric oxide (CuO) is a native p-type material with a theoretical indirect band gap of 1 eV [7] and an experimental band gap of 1.2 eV [8]. Studies using density functional theory indicate that while successfully doping CuO to produce n-type behaviour is highly impractical, both sodium and lithium should act as effective p-type dopants [9].

This study aims to examine the effects of sodium doping on the resistivity of CuO, and to increase its band gap by co-sputtering with tin dioxide (SnO2).

2. Experimental details

Films were deposited onto pre-cleaned soda-lime glass by sputtering from 3 inch (7.62 cm) diameter ceramic targets using an AJA Orion 8 HV sputter coater. For all films, the deposition pressure was maintained at 1 mTorr (0.133 Pa) and the argon flow rate was 7 standard cubic centimetres per minute (SCCM).

Sodium-doped CuO was sputtered from a single pre-formed target using an AJA 600 series RF power supply. Power was kept to 120 W, giving a power density of 0.66 W/cm2. Sodium doping was 2 at. %. Oxygen input for doped films was varied from 0 to 6 SCCM. Undoped CuO films used for comparison were deposited from an intrinsic CuO target using the same procedure, with the exception that oxygen was not added.

Intrinsic CuO and SnO2 were co-sputtered from pre-formed targets using an AJA 600 series RF power supply and an Advanced Energy MDX 500 DC power supply respectively. Oxygen flow was kept to 2 SCCM, resulting in an oxygen partial pressure of 0.03 Pa. Depositions were carried out at room temperature (18 ºC). Power to each target was varied depending on the desired film composition. CuO target power was varied from 180 W to 72 W, and SnO2 target power was varied from 50 W to 12.5 W.

Film optical and electrical characterisation was carried out using an Ambios XP2 stylus profilometer to measure thickness, a Varian Cary 5000 spectrophotometer for transmission spectra and an Ecopia HMS 3000 Hall effect system for mobility, resistivity and carrier concentration measurements. Sheet resistance was measured using a four-point probe. Structural characterisation involved scanning electron microscopy (SEM) and X-ray diffraction (XRD), and compositional analysis was performed using a Thermo Scientific K-Alpha X-ray photoelectron spectroscopy (XPS) equipped with an Al-K X-ray source. Samples were etched for 60 seconds prior to analysis using a plasma etching ion source (EX06 ion source).. SEM analysis was carried out using a Carl Zeiss (Leo) 1530 VP field emission gun scanning electron microscope, using both InLens and secondary electron detectors. Aperture size was 30 m and operating Voltage was 5 kV. XRD was carried out using a Bruker D2-phaser desktop X-ray diffractometer equipped with a Cu-K X-ray source and LynxeyeTM detector. The beam slit was 1 mm wide and the anti-scatter plate was positioned 3 mm above the sample. The sample was rotated at 15 revolutions per minute. Experiments were run for a range of different times, from 15 minutes through to 12 hours depending on the sample.

Band gaps were estimated from film transmission data using the Tauc method, and plotted against XPS compositional data. Film carrier type was confirmed by exploiting the Seebeck effect using a multimeter with a heated probe.

3. Results and discussion

3.1. Sodium-doped cupric oxide

Intrinsic CuO was found to have a resistivity of 1.5 x 103 .cm. Sodium-doped films showed resistivities four orders of magnitude lower at 5 x 10-1 .cm, demonstrating that sodium is an effective dopant for CuO. Addition of oxygen to the deposition environment was found to reduce the resistivity further, down to a minimum of 4.3 x 10-2

.cm at an oxygen partial pressure of 6.2 x 10-2 Pa. Further increases in oxygen partial pressure were found to cause a slight increase in resistivity (Fig. 1). Unlike n-type metal oxides, increased oxygen partial pressure

P.J.M. Isherwood and J.M. Walls / Energy Procedia 60 ( 2014 ) 129 – 134 131

decreases the resistivity in CuO through the formation of copper vacancies and oxygen interstitials. Density functional theory modelling of CuO has shown both of these to be shallow acceptor states [9], meaning that they both contribute to an increased charge carrier concentration. Unfortunately measurements of the carrier concentration and mobility were inconclusive. This is most likely due to the films having a mobility below the measureable limit of the Hall device.

Fig. 1. Resistivity against oxygen partial pressure for sodium-doped CuO films SEM analysis showed that intrinsic CuO films are polycrystalline, with small but well-formed crystals. Doped

films, although still polycrystalline, were found to be less well-formed, with significantly smaller crystals and a large number of cracks and pinholes (Fig. 2). Oxygen partial pressure during deposition had no apparent impact on film quality. Therefore the reduced quality of the doped films is thought to be a result of distortion of the film crystal structure due to the incorporation of the sodium dopant. XRD analysis confirmed that both intrinsic and doped films are crystalline, and show a typical CuO peak profile (Fig. 3).

Fig. 2. SEM photomicrographs of sodium-doped CuO deposited at oxygen partial pressures of (a) 0.029 Pa and (b) 0.048 Pa

Fig. 3. XRD patterns of intrinsic and doped CuO. Both show the same two peaks at 2 values of 35.5º and 38.5º respectively

3.2. Copper-tin mixed oxides

Highly doped and partially conductive CuO has potential application as a back contact material and as an electron reflector, but the range of potential uses is limited by its narrow band gap. By co-sputtering CuO with SnO2 at

132 P.J.M. Isherwood and J.M. Walls / Energy Procedia 60 ( 2014 ) 129 – 134

different target power levels, it was found that the band gap could be altered between that of CuO (1.2 eV) and an upper limit of 2.75 eV. This change is non-linear and describes a curve, with the rate of change decreasing with increased SnO2 content (Fig. 4).

Fig. 4. Band gap against copper content (as a % of the total film metal content) for mixed oxide films. Pure SnO2 (0 % copper) and CuO (100 % copper) are included for comparison

Resistivity was found to increase rapidly with decreasing copper content, up to around 45% copper (as a

percentage of total metal content). Further reductions in copper content had little significant effect on resistivity (Fig. 5)

Fig. 5. Copper content (% of total film metal content) against resistivity for mixed oxide films XRD analysis indicated that all mixed-oxide films were amorphous, with only the two pure oxides showing

evidence of crystallinity (Fig. 6). This is likely due to disruption of the material structure resulting from the difference in size of the two metal ion types.

Fig. 6. XRD patterns of mixed oxide films with copper contents (% of total film metal content). All mixed oxide films show amorphous behaviour

SEM photomicrographs show that the films are quite rough, and display a variety of surface structures and

patterns (Fig. 7). This could be a result of incomplete or inhomogeneous mixing of the two materials, although as all electrical and optical measurements were consistent across the films this seems unlikely. Further structural and site-specific compositional analysis would be necessary to determine the precise cause.

P.J.M. Isherwood and J.M. Walls / Energy Procedia 60 ( 2014 ) 129 – 134 133

Fig. 7. SEM photomicrographs of mixed oxide films with copper contents (as a % of total metal content) of (a) 69 %, (b) 49 %, and (c) 24 % Non-linear band gap change as shown by this material is a known phenomenon in amorphous semiconductor

alloy systems, and has been extensively described in other mixed materials such as (In, Ga)N and Ga(As, N) [10], [11]. The cause of this behaviour is not entirely understood, but it is thought to be a result of competition for charge accumulation between the two cation species [12]. The precise nature of the curve shown by the material system is determined by a factor known as the bowing parameter [10], [12]. The band gap of a semiconductor alloy AxC1-xO for a given value of x is found using Equation 1:

)1()()1()()( xbxCEgxAxEgtotEg (1)

Where Eg(A) and Eg(C) are the band gaps of the two end-member semiconductors, and b is the bowing

parameter [10]. It was found that for the amorphous part of the CuO-SnO2 system, the bowing parameter was -1.5 eV. All mixed oxide films were found to demonstrate p-type semiconductor behaviour when tested using the Seebeck effect.

4. Conclusions

Doping CuO thin films with 2 at. % sodium was found to reduce the resistivity by four orders of magnitude. Addition of oxygen to the deposition environment was found to reduce the resistivity further, to a minimum of 4.3x10-2 .cm at an oxygen partial pressure of 6.2 x 10-2 Pa. Doped films were found to be crystalline, but show a greater degree of disruption and a lower overall structural quality than intrinsic films.

Co-deposition of CuO with SnO2 was found to cause a non-linear increase in the band gap, from 1.2 eV up to 2.75 eV. The curve described by this system was found to have a bowing parameter of -1.5 eV. All mixed oxide films were found to be amorphous and p-type. Resistivity initially increased significantly with decreasing copper content, but reducing the copper content below 45 % had no further significant impact.

Acknowledgements

The authors would like to acknowledge funding for the work through the EPSRC Supergen Supersolar Hub. They would also like to acknowledge the invaluable help and support of Ali Abbas for assistance with structural characterisation.

References

[1] C. G. Granqvist. Transparent conductors as solar energy materials: A panoramic review. Sol. Energy Mater. Sol. Cells 2007; 91, 17:1529–1598.

[2] G. Hautier, A. Miglio, G. Ceder, G.-M. Rignanese, X. Gonze. Identification and design principles of low hole effective mass p-type transparent conducting oxides. Nat. Commun. 2013; 4, 2292:1-7.

[3] H. Kawazoe, M. Yasukawa, H. Hyodo, M. Kurita, H. Yanagi, and H. Hosono. P-type electrical conduction in transparent thin films of CuAlO2. Nature 1997; 389:939–942.

[4] H. Kawazoe, H. Yanagi, K. Ueda, and H. Hosono. Transparent p-type conducting oxides: design and fabrication of pn heterojunctions. MRS Bull. 2000; 25, 08:28–36.

134 P.J.M. Isherwood and J.M. Walls / Energy Procedia 60 ( 2014 ) 129 – 134

[5] D. O. Scanlon, K. G. Godinho, B. J. Morgan, and G. W. Watson. Understanding conductivity anomalies in Cu(I)-based delafossite transparent conducting oxides: Theoretical insights. J. Chem. Phys. 2010; 132, 2:024707-1 - 024707-9.

[6] H. Yanagi, S. Inoue, K. Ueda, H. Kawazoe, H. Hosono, and N. Hamada. Electronic structure and optoelectronic properties of transparent p-type conducting CuAlO2. J. Appl. Phys. 2000; 88, 7:4159–4163.

[7] D. Wu, Q. Zhang, and M. Tao. LSDA+U study of cupric oxide: Electronic structure and native point defects. Phys. Rev. B 2006; 73, 23:1–6. [8] J. Chen, S. Z. Deng, N. S. Xu, W. Zhang, X. Wen, and S. Yang. Temperature dependence of field emission from cupric oxide nanobelt films.

Appl. Phys. Lett. 2003; 83, 4:746. [9] Y. Peng, Z. Zhang, T. Viet Pham, Y. Zhao, P. Wu, and J. Wang. Density functional theory analysis of dopants in cupric oxide. J. Appl. Phys.

2012; 111, 10:103708. [10] C. G. Van de Walle, M. D. McCluskey, C. P. Master, L. T. Romano, and N. M. Johnson. Large and composition-dependent band gap

bowing in In x Ga 1 x N alloys. Mater. Sci. Eng. 1999; B59:274–278. [11] T. S. Kim, T. V Cuong, C. S. Park, J. Y. Park, H. J. Lee, E. Suh, and C. Hong. Composition Dependence of the Band-Gap Energy of

GaAsN Alloys. J. Korean Phys. Soc. 2003; 43, 2:273–276. [12] N. Tit, I. M. Obaidat, and H. Alawadhi. Origins of bandgap bowing in compound-semiconductor common-cation ternary alloys. J. Phys.

Condens. Matter 2009; 21, 7:075802.