Potential Buffer Layers For Cu2ZnSnS4 (CZTS) Solar Cells from Numerical Analysis

M. I. Hossaina,d, P. Chelvanathana, M. M. Alamc, M. Akhtaruzzamana, K. Sopiana and N. Amina,b,c

aSolar Energy Research Institute (SERI), Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia. bDepartment of Electrical, Electronics and System Engineering, Universiti Kebangsaan Malaysia, Malaysia.

cCollege of Sciences, King Saud University, Riyadh 11451, Saudi Arabia. dQatar Environment and Energy Research Institute, Qatar Foundation, Doha, Qatar.

*Corresponding E-mail: nowshad@eng.ukm.my

Abstract— Recently Cu2ZnSnS4 (CZTS) absorber layer research shows extensive influential factors to replace expensive CIGS absorber layer. In this work, potential buffer layers for CZTS solar cells like ZnO, ZnS, ZnSe, and InS as replacement to conventional CdS buffer layer are numerically analysed. Among all the used structures, ZnS/CZTS structure shows an optimum efficiency of 13.71 % (with Voc = 0.78 V, Jsc = 31.97 mA/cm2, and fill factor = 55 %). This work explicitly reveals the most favorable CZTS layer thickness in the range of 1 μm to 2.2 μm, whereas buffer layer thickness lies down between 50 nm and 80 nm. It is not trivial to achieve promising efficiency for ZnO/CZTS structure as well. Temperature effects on the cells are also observed. The efficiency of ZnS/CZTS solar cells downsized from 13.71% (300k) to around 11.07% (350 K) with an absolute decreasing rate of (%k-1) = 0.05 whereas (%k-1) = 0.03 for ZnO/CZTS cell. Also, cells ideality is determined by theoretical modeling. The achieved results can lead to develop higher efficiency CZTS thin film solar cells.

IndexTerms—CZTS absorber layer, Layer thickness, Bandgaps, Buffer layers, SCAPS.

I.INTRODUCTION Inexpensive solar cells without material degradation are

essential to drive towards high throughput for PV commercialization. As CIGS and CZTS cells consist of the same governing physics, it becomes obvious to replace expensive and scarce Indium and Gallium with Zinc (Zn) and Tin (Sn), respectively. CZTS is a quaternary semiconductor of group I-II-IV-VI with effective light absorbing properties. CZTS is not a new material. It was first demonstrated in 1966 and a photovoltaic effect was confirmed in 1988 [1,2]. A very high absorption coefficient in the visible wavelength region makes this quaternary semiconductor favorable for PV device. In 2012, IBM developed an 11.1% efficient CZTS solar cell [3]. Its direct optical bandgap is in between 1.4 and 1.56 eV, and large absorption coefficient is >104 cm-1 [4]. Formation of native defects in the crystal structure is not uncommon due to quaternary properties [5]. Some previous works showed metallic droplet formation besides high defect densities. These limitations cause short diffusion length of the carriers [6]. Light soaking is found to improve the cell performance [7]. In this paper, effects of different operating temperatures are studied. Choosing a suitable buffer layer definitely improves

interface and enhances the blue wavelength transmission for n-buffer/p-CZTS device. The main scope of this investigation is to identify the best buffer layer for CZTS solar cells. Also, theoretical modeling is included to find the ideality of the cells from the analyzed open circuit voltage.

II.METHODOLOGY In this paper, the cell performance is investigated using SCAPS 2802 simulator. In this simulation, used materials’ feasible physical parameters are accentuated to explore the novel CZTS solar cells performances with various layer structures. All the relevant material parameters are incorporated into SCAPS for analysis purposes, and the changes in the values for efficiency, short circuit current, fill factor, open circuit voltage and effect of operating temperature are recorded and analyzed. The description and the baseline values of the physical parameters those are used in the simulation are shown in table 1. The materials properties used in this analysis are reviewed from the trustworthy source [8]. A basic schematic diagram is provided for proposed CZTS structure in fig. 1. Figure 2 shows the band edge diagram of p-n junction interface with various buffer materials. It is essential to determine the best junction without any band offset or spike.

In the second part of the work an empirical model is derived from the voltage dependent fill factor.

Table 1. Semiconductor physical properties

EPS Relative permittivity (relative),εr MUN Electron mobility, μn (cm2/Vs) MUP Hole mobility, μp (cm2/Vs) NA Acceptor carrier concentration, (1/cm3) ND Donor carrier concentration, (1/cm3) EG Bandgap, (eV) NC Conduction band effective density of

states, (1/cm3) NV Valance band effective density of states,

(1/cm3) CHI Affinity, χe (eV)

2013 IEEE Conference on Clean Energy and Technology (CEAT)

978-1-4799-3238-2/13/$31.00 ©2013 IEEE 450

Figure 1: Schematic structure of CZTS solar cell

Figure 2: Band edge diagram of CZTS solar cells with various buffer layers

The theoretical modeling is done in the following order. For short circuit current we know,

(1) where is the short circuit current, is the diode saturation current, is electron charge, is the boltzman constant, is the temperature, and is the voltage.

(2)

where is the open circuit voltage and others are as stated above. Maximum output voltage,

For ideal case scenario,

(4)

(5) Maximum output current,

(6)

Fill factor/ideality of the cell,

(7) Considering the best case where all the photon energy are converted into short circuit current, , fill factor becomes

(8)

(9)

,

(10)

(11) This derived empirical equation will certainly help to find the cells ideality from the simulated open circuit voltage result.

III.RESULTS AND DISCUSSIONS

A. CZTS layer thickness effects At the beginning, absorber layer thickness is varied

between 1 μm and 2.5 μm for the novel n-ZnS/p-CZTS hetero- junction cells. The simulated results show lower efficiency when the absorber layer thickness is in lower range. The efficiency reaches up to 13.71 % with 2.5 μm of absorber thickness.

The best possible buffer layer thickness of ZnS is between 30 nm and 50 nm as less thickness allows most of the photon energy to be transmitted with higher rate. An optimum 1.5 eV bandgap of CZTS absorbs the photons energy higher than the bandgap energy to produce higher Isc and Voc. The electrical performances of the CZTS cells are shown in fig. 3, where open circuit voltage and short circuit current improve with the absorber layer thickness. It is certainly related with the collection of longer wavelength of lights with lower photon energy in the absorber layer. However, the ratio of diffusion length and absorption thickness

should be maintained in a better way for higher efficiency,

Improvement indicator = (12)

2013 IEEE Conference on Clean Energy and Technology (CEAT)

978-1-4799-3238-2/13/$31.00 ©2013 IEEE 451

Lower Voc and Isc are resultant of the high recombination centers due to the defect states which causes trapping of electrons (or holes) which are influenced under illumination or bias voltage.

B. Performance analysis of ZnS/CZTS structure ZnS is a potential buffer layer due its higher bandgap and

ability to create better interface. ZnS thickness variation is done between 0.300 and 1.00 μm with 2.50 μm of CZTS. For ZnS bandgap of 3.5 eV and thickness of 250 nm, cell efficiency improves to 9.8% (fig. 4). Impressive high bandgap results in less absorption of photon and creates a better built in potential barrier at the interface to dissociate the excitons.

Figure 3: Simulated performance with absorber layer thickness

C. Performance analysis with various buffer layers Four other potential buffer layers such as InS, ZnSe, CdS,

and ZnO have been investigated besides ZnS. For all the cases, thickness is kept between 10 nm and 100 nm. For increasing buffer thickness, more photons are being absorbed by the n-type layer leading to a decrease in absorption as narrated earlier from spectral response. However, high bandgap of ZnO can compensate this effect. In case of ZnO, InS, and ZnSe, thickness around 50 nm can be set as optimum point as the efficiency decreases after 50 nm. Figure 5 summarizes the results with different buffers as achieved from the simulation. ZnS buffer layer based cell has been achieved as high efficiency as 13.71%, suggesting it to be a potential replacement of CdS. ZnO, ZnSe, and InS based cells show the

conversion efficiencies of 12.16%, 5.84%, and 8.5%, respectively.

Figure 4: Effects of ZnS buffer layer thickness

D. Effects of temperatures The effects of operating temperature on the performance of

CZTS/ZnS and CZTS/ZnO structure have been investigated (fig. 6). The efficiency of ZnS/CZTS solar cells downsized from 13.71% (300k) to around 11.07% (350 K) with a absolute decreasing rate of (%k-1) = 0.05 whereas

(%k-1) = 0.03 for ZnO/CZTS cell. ZnO/CZTS cells are found to be more stable for higher temperature. Operating temperature abruptly affects the cells performance in various standards. The concept of solar cells operation physics is enormously governed by temperature which is well aligned with the theoretical conception and practical condition. Similarly materials physical parameters like energy band gap (Eg), electron and hole mobilities, and carriers densities (Nc

and Nv) shows high temperature dependency due to the high acceleration rate of recombination, shrinkage of built in potential at the interface.

Beside this, unexploited photon energy inside the absorber layer definitely increases the cells temperature in a scattering way much beyond the surrounding temperature and such condition results in depreciating cell performance. Generally, temperature effects are determined by the increased short-circuit current (Isc) and the decrease in the open circuit voltage (Voc) for the cells. The change in energy bandgap with temperature is the primary reason for short circuit current variation. Due to the lower bandgap with higher temperature,

2013 IEEE Conference on Clean Energy and Technology (CEAT)

978-1-4799-3238-2/13/$31.00 ©2013 IEEE 452

the longer wavelength of lights will be accumulated in the cell, and will result in slightly better short circuit current. In contrast, linear reduction of (Voc) with higher temperature happens due to the bandgap shrinkage along with the deviation of effective density of states with temperature resulting in reduced the built in potential. Several studies have demonstrated the temperature effects on the solar cell operation [9,10].

Figure 5: Effects of various buffer layers

Figure 6: Effects of various operating temperatures

E. Cells ideality

CZTS cells ideality is also calculated from the achieved open circuit voltage using eqn. 11 (table 2). It is mainly a function of open circuit voltage. This empirical equation shows the voltage dependency as with higher open circuit voltage fill factor increases linearly. The results incorporated inside table show a clear ideality difference between simulated fill factor results and theoretically calculated values. It could be happen due to the ideal case consideration in the theoretical calculation. Generally, the ideality factor is heavily material properties dependent. By improving the films quality, the peak point can be controlled.

Table 2.Cells ideality

0.68 0.858 0.67 0.856 0.66 0.854 0.65 0.852

CONCLUSIONS Potential buffer layers for CZTS cells are investigated.

Though in feasible fabrication the conversion efficiency of 11.1% has been achieved, in this work ZnS/CZTS structure shows an optimum efficiency of 13.71 % (with Voc = 0.78 V, Jsc = 31.97 mA/cm2, and fill factor = 55 %). An empirical equation is derived which shows the scope of cells performance improvement by controlling the growth quality. Temperature effects are also included for the range of 300k to 350k. The achieved results can lead to develop higher efficiency CZTS thin film solar cells.

ACKNOWLEDGMENT The authors would like to acknowledge the Solar Energy

Research Institute, National University of Malaysia (UKM) and Qatar Environment and Energy Research Institute, Qatar Foundation.

REFERENCES R. Nitsche, D. F. Sargent, and P. Wild, "Crystal Growth of

Quaternary I(2)II-IV-VI(4) Chalcogenides by Iodine Vapor Transport". Journal of Crystal Growth, pp. 52–53, 1967.

K. Ito, T. Nakazawa, "Electrical and Optical Properties of Stannite-Type Quaternary Semiconductor Thin Films", Japanese Journal of Applied Physics 27: 2094, 1988.

T. Todorov, D. Mitzi, "Shedding light on new frontiers of solar cell semiconductors", IBM. Retrieved, 2012.

H. Wei, W. Guo, Y. Sun, Z. Yang, and Y. Zhang, “Hot-injection synthesis and characterization of quaternary Cu2ZnSnSe4nanocrystals,” Materials Letters, 2010.

T. Tanaka, A. Yoshida , D. Saiki, K. Saito, Q. Guo, M. Nishio, and T. Yamaguchi, “Influence of composition ratio on properties of Cu2ZnSnS4 thin films fabricated by co-evaporation,” Thin Solid Films, 2010.

2013 IEEE Conference on Clean Energy and Technology (CEAT)

978-1-4799-3238-2/13/$31.00 ©2013 IEEE 453

K. Tanaka, M. Oonuki, N. Moritake, and H. Uchiki, “Cu2ZnSnS4 thin film solar cells prepared by non-vacuum processing,” Solar Energy Materials & Solar Cells, pp. 583–587, 2009.

H. Katagiri, K. Jimbo, S. Yamada, T. Kamimura, W. S. Maw, T. Fukano, T. Ito, and T. Motohiro,” Enhanced Conversion Efficiencies of Cu2ZnSnS4-Based Thin Film Solar Cells by Using Preferential Etching Technique,” Applied Physics Express, 2008.

M. A. Olopade, O. O. Oyebola, and B. S. Adeleke, “Investigation of some materials as buffer layer in copper zinc tin sulphide (Cu2ZnSnS4) solar cells by SCAPS-1D,” Advances in Applied Science Research, pp. 3396-3400, 2012.

P. Singh, N. M. Ravindra, “Temperature dependenceofsolarcellperformance—an analysis,” Solar Energy Materials & Solar Cells pp. 36–45, 2012.

G. Landis, “Review of Solar Cell Temperature Coefficients for Space,” Proceedings in XIIISpace Photovoltaic Research and Technology Conference, pp. 385-400, 1994.

2013 IEEE Conference on Clean Energy and Technology (CEAT)

978-1-4799-3238-2/13/$31.00 ©2013 IEEE 454