European Journal of Scientific Research ISSN 1450-216X Vol.29 No.4 (2009), pp. 447-453 EuroJournals Publishing, Inc.

. 2009 http://www.eurojournals.com/ejsr.htm

Photovoltaic Properties of Si3N4 Layer on Silicon Solar Cell Using Silvaco Software
A. Lennie Dept. of Electrical, Electronics and Systems Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia 43600 Bangi Selangor, Malaysia E-mail: lenny_abdul@yahoo.com Tel: +60389216310; Fax: +60389216146 H. Abdullah Dept. of Electrical, Electronics and Systems Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia 43600 Bangi Selangor, Malaysia S.M.Mustaza Dept. of Electrical, Electronics and Systems Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia 43600 Bangi Selangor, Malaysia K.Sopian Research Institute Solar Energy (SERI) Universiti Kebangsaan Malaysia 43600 Bangi Selangor, Malaysia Abstract In this research, the silicon dioxide (SiO2) and silicon nitride (Si3N4) layer have been modelled and fabricated on silicon solar cell by using Silvaco Software Packaging. Basically the function of SiO2 and Si3N4 materials are an anti-reflective coating (ARC) on solar cell devices that based on refractive index of each material. The application of ARC is a good method to improve the solar cell construction. This fabrication carried out on high temperature during annealing process from 800C to 1050C when the ARC layer is 50 nm. The photovoltaic properties of Si3N4 layer have been compared with SiO2 layer to determine which material is suitable in fabricating single layer ARC. Short circuit current density Jsc, open circuit voltage Voc and pn junction were measured as a function of both layers material and the results of fill factor (FF) and power conversion efficiency, are discussed. For SiO2 results, the FF value are approximately 0.758 and efficiency, is nearly 9.43%. In annealing process, the temperature become higher resulted increasing of pn junction, but not to Voc and Jsc values. The Voc and Jsc were slowly decreased when temperature increased. As for Si3N4 result, the calculated FF in range of 0.7570.758 and

Photovoltaic Properties of Si3N4 Layer on Silicon Solar Cell Using Silvaco Software efficiency is around 9.559.57%. In annealing process, the temperature increasing constantly follows the increasing of pn junction and Jsc, meanwhile the Voc is decreased slowly. The optimum efficiency from SiO2 layer is 9.428% when temperature reaches at 950C, while the optimum efficiency of Si3N4 layer is 9.565% when temperature at 1050C. Solar cell simulation could be useful for time saving and cost consumption, and also cheaper and faster compared to physical experimental.

448

Keywords: Silicon solar cell, SiO2, Si3N4, Efficiency, Silvaco

1. Introduction
Traditional fuel energy resources are limited and generate harmful waste products in the atmosphere. Hence, alternative eco-friendly energy sources are receiving increasing attention. Of all substitute energy sources, solar power is most attractive because of its abundance and the consistency of sunlight (Hsu et al. 2009). In 1950s, Chapin et al. at Bell laboratories developed the first crystalline silicon solar cell with a solar-energy conversion efficiency of 6% (Chapin et al. 1954). However, the relatively high cost of manufacturing these silicon cells has prevented them from extensive use. Moreover, another disadvantage of silicon cells is the use of toxic chemicals during manufacturing. Accordingly, the development of low-cost organic solar cells is urgent. High quality anti-reflective coatings (ARC) have become a vital feature of high-efficiency silicon solar cells (Zhao et al. 1991). Several authors have conducted extensive studies in past few years to investigate the effects of various ARC on high efficiency monocrystalline and polycrystalline solar cells (Resnik et al. 1999). Several materials having the appropriate refractive index have been proposed and used for single or double layer AR coatings, such as TiO2 (n=2.2), MgF2 (n=1.35), ZnS (n=2.3), CVD SiN (n=1.6-2.4) and SiO2 (n=1.46). The thickness and refractive index play role in photoelectrical conversion by reducing the reflectivity losses. The refractive index of stoichiometric LPCVD Si3N4 is close to the calculated value of 2.3 for the top layer. Silvaco TCAD offers complete and well-integrated simulation software for all aspects of solar cell technology. TCAD modules required for solar cell simulation include: S-Pisces, Blaze, Luminous, TFT, Device3D, Luminous3D and TFT3D (ATLAS Users manual et al. 1998). S-Pisces is an advanced 2D device simulator for silicon based technologies that incorporates both drift-diffusion and energy balance transport equations. Large selections of physical models are available for solar cell simulation, which includes surface/bulk mobility, recombination, impact ionization and tunneling models. Blaze simulates 2D solar cell devices fabricated using advanced materials. It includes a library of binary, ternary and quaternary semiconductors. Blaze has built-in models for simulating state-of-theart multi-junction solar cell devices. Luminous and Luminous3D are advanced 2D and 3D simulator specially designed to model light absorption and photogeneration in non-planar solar cell devices. Exact solutions for general optical sources are obtained using geometric ray tracing. This feature enables Luminous and Luminous3D to account for arbitrary topologies, internal and external reflections and refractions, polarization dependencies and dispersion. Luminous and Luminous3D also allows optical transfer matrix method analysis for coherence effects in layered devices. The beam propagation method may be used to simulate coherence effects and diffraction.

2. Materials and Research Methodology

The solar cell that has been chosen for test is made in usual method in VLSI. An orientation silicon wafer of <100> with 50m thickness and 11014 atom/cm2 boron concentration was chosen. The p-n junction was developed by implant of phosphorus with 11016 atom/cm2 and energy is 30 eV. The diffuse time 60 min is constant, meanwhile the temperature is variable parameter. The deposition of anti-reflective coating also will be variable parameter. The ATHENA framework is used to design the

449

A. Lennie, H. Abdullah, S.M.Mustaza and K.Sopian

2D solar cell structure with area 50 10 m2. Then the voltage was applied on device by using ATLAS framework to compute open circuit voltage, Voc and short circuit current density, Jsc. In this simulation, the incident light has been fixed in 90 angle. The 90 incident light angle was applied on the top of silicon solar cell to trace the reflectance in silicon wafer. Figure 1 and 2 shows cross-section of the 1010 m2 solar cell structure with SiO2 and Si3N4 single layer when applying 90 incident light. In this simulation, the variable temperature that been chosen for analysis is 800C, 850C, 900C, 950C, 1000C and 1050C. The thickness of both layers has been fixed to 50 nm thick. Table 1 show the details of input data were used in modelling solar cell structure for both analyses.
Table 1: Input data were used in modelling solar cell structure for SiO2 and Si3N4 simulation.
SiO2 Simulation 100 50 10 m2 11016 atom/cm2 30 ev 60 min (1 hr) 800C, 850C, 900C, 950C, 1000C, 1050C 50 nm 90 Si3N4 Simulation 100 50 10 m2 11016 atom/cm2 30 ev 60 min (1 hr) 800C, 850C, 900C, 950C, 1000C, 1050C 50 nm 90

Parameter Orientation Device area Phosphorus doping Energy Diffuse time Diffuse temperature ARC thickness Incident light beam

3. Results and Discussion

The solar cell structure has been done modelling by using Silvaco TCAD Tools. The variable temperature that has been chosen in this simulation is 800C, 850C, 900C, 950C, 1000C and 1050C. After simulation has been done, the Athena framework will display the 2D device structure. Figure 1 and 2 shows cross-section of the 1010 m2 solar cell structure when heating in 950C on 50 nm thick of SiO2 and Si3N4 single layer, respectively. These figures shows two different contour region, which on left side is net doping contour on substrate, meanwhile on the right side is the contour when applying photogeneration rates. It is possible to study the details of photo generation of carriers in the solar cell device during light illumination. In these figures, the device has an opaque metal contact (cathode electrode) in the side of the structure. Once photogeneration rates are obtained, terminal currents can be evaluated to determine the quantum efficiency of the solar cell. Besides studying the photogeneration rates due to a normal incident light beam, the photogeneration rates due to an angled light beam can also be studied.
Figure 1: The cross-section of 1010 m2 solar cell structure with 50nm SiO2 single layer

Photovoltaic Properties of Si3N4 Layer on Silicon Solar Cell Using Silvaco Software
Figure 2: The cross-section of 1010 m2 solar cell structure with 50nm Si3N4 single layer

450

In this paper, the simulation only concentrated on normal incident light beam for different ARC material. So the photogeneration rates can be seen clearly from the red region then slowly changed the colour downward to yellow region in silicon substrate. The purple region on both side of solar cell represent the incident light beam could not go through above cathode area. The difference between SiO2 and Si3N4 layer figures is the distance of pn junction that clearly illustrated on figures where the vertical line below the ARC in silicon region. The figure on the left side shows the net doping colour follows the distance of pn junction. The net doping colour represents the phosphorus doping of implant n+ layer to create pn junction. The SiO2 layer has a pn junction 0.574181 m meanwhile Si3N4 layer has a tiny pn junction of 0.429013 m, both are from 50 nm ARC at 950C result as shown in Table 4 and 5. Silvaco TCAD tools can simulate the various aspects of solar cell characteristics. Typical characteristics include collection efficiency, spectral response, open circuit voltage, Voc and short circuit current density Jsc. The Jsc is extracted from the curve when the voltage is zero. On the other hand, the Voc can be extracted from the J-V curve when the current is zero. Figure 3 displayed the J-V characteristics for variable temperature simulation on 50 nm SiO2 and Si3N4 layer deposited on silicon solar cell. In overall the range of current density, Jsc for SiO2 layer is around 3.1203.129 mA/cm2 and Si3N4 layer is approximately to 3.1693.170 mA/cm2. This means the Jsc of Si3N4 layer is higher than SiO2 layer. Meanwhile the Voc of both layer are nearly to 0.40 V.
Figure 3: J-V curve for variable temperature simulation on SiO2 and Si3N4 layer
J-V curve 50 nm SiO2 on silicon solar cell for variable tem perature
0.5 0 0 0.1
800C 850C 900C 950C 1 000C 1 050C

J-V curve 50 nm Si3N4 on silicon solar cell for variable tem perature
0.5 0

Current Density, J (mA/cm 2)

Current Density, J (mA/cm 2)

-0.5 -1 -1 .5 -2 -2.5 -3 -3.5

0.2

0.3

0.4

0.5

-0.5 -1 -1 .5 -2 -2.5 -3 -3.5

0.1
800C 850C 900C 950C

0.2

0.3

0.4

0.5

1 000C 1 050C

Voltage, V (V)

Voltage, V (V)

451

A. Lennie, H. Abdullah, S.M.Mustaza and K.Sopian

In this project the first simulation has been done is simulation of variable temperature parameter from 800C to 1050C with interval 50C. Atlas simulator simulated the pn junction, Voc and Jsc, but the fill factor (FF) and power conversion efficiency () were calculated by using equations below. The fill factor, FF is the ratio of maximum power point, Pm divided by the Jsc and Voc, and that is J V Pm (1) FF = m m = J sc Voc J sc Voc Typically, the fill factor, FF is between 0.7 and 0.8 (Neamen et al.2006). The energy conversion efficiency of solar cell, is the comparison of maximum power point of cell, Pm to input light from source, Pin, P FF J sc Voc = m 100% = 100% (2) Pin Pin Earth can received the maximum power from sun is about 1000 W/m2. So, the input power Pin is 1000 W/m with an air mass 1.5 (AM1.5) spectrums (Weider et al. 1982). Tables 2 and 3 show the results of photovoltaic properties that have been simulated and also calculated for SiO2 and Si3N4 silicon solar cell.
Table 2: Photovoltaic properties of SiO2 on silicon solar cell with different temperature (T) when ARC thickness is 50 nm
pn junction (m) 0.415887 0.417483 0.425277 0.574181 0.904698 1.430930 Voc (V) 0.397788 0.397788 0.397775 0.397732 0.397652 0.397483 Jsc (mA/cm2) 3.12905 3.12901 3.12894 3.12832 3.12616 3.12004 FF 0.75722 0.75723 0.75738 0.75777 0.75821 0.75874

T (C) 800 850 900 950 1000 1050

(%)
9.425 9.425 9.426 9.428 9.426 9.410

Table 3:

Photovoltaic properties of Si3N4 on silicon solar cell with different temperature (T) when ARC thickness is 50 nm
pn junction (m) 0.387574 0.389204 0.397174 0.429013 0.576968 0.994429 Voc (V) 0.398097 0.398097 0.398096 0.398091 0.398064 0.397989 Jsc (mA/cm2) 3.16891 3.16892 3.16894 3.16905 3.16951 3.16975 FF 0.75715 0.75715 0.75717 0.75725 0.75759 0.75818

T (C) 800 850 900 950 1000 1050

(%)
9.552 9.552 9.552 9.553 9.558 9.565

This simulation were analyzed in different diffuse temperature while the ARC thickness is fixed to 50 nm. From the SiO2 results (in Table 2), FF value are approximately 0.758 and efficiency, is nearly 9.43%. In annealing process, the temperature become higher resulted increasing of pn junction, but not to Voc and Jsc values. The Voc and Jsc were slowly decreased when temperature increased. In this table, when the temperature reach at 950C the become optimum which has 9.428% efficiency. This can be concluded the SiO2 layer not recommended for high temperature in annealing process. As for Si3N4 result as in Table 3, the calculated FF in range of 0.7570.758 and efficiency is around 9.559.57%. In annealing process, the temperature increasing constantly follows the increasing of pn junction and Jsc while the Voc is decreased slowly. All the result illustrated in figure below that concluded the changes current density and photovoltaic efficiency of both layers.

Photovoltaic Properties of Si3N4 Layer on Silicon Solar Cell Using Silvaco Software

452

Figure 4: Graph of current density and photovoltaic efficiency for variable temperature on 50 nm SiO2 and Si3N4 layer
Graph of current density and efficiency on SiO2 and Si3N4 layer
3.18 3.17 9.55 3.16 9.60

Current Density,J

3.15

9.50 9.45

Efficiency (%)

(mA/cm2)

Jsc (SiO2) Jsc (Si3N4) Eff. (%) (SiO2) Eff. (%) (Si3N4)

3.14 3.13 3.12 3.11 9.35 3.10 3.09 800 850 900 950 1000 1050 9.30 9.40

Temperature (C)

In overall the maximum efficiency that can be reached on SiO2 and Si3N4 layers is 9.428% and 9.565%, respectively. These figures clearly show the current density and photovoltaic efficiency of SiO2 layer is decreasing, but for Si3N4, the both parameter is increasing follows the temperature. In this simulation, the Si3N4 material can be proved as an ARC function for solar cell application because it could enhance the Jsc and efficiency that could be improved on high temperature.

4. Conclusion
A theoretical study of the ARC on silicon solar cells is made (Bouhafs et al. 1998). The ability of the ATLAS device simulator to accurately a model solar cell characteristic has been shown (Micheal and Bates 2005, Micheal et al. 2005). The detailed outputs available to the solar cell designer allow for efficient and effective simulation and optimization of even most advanced solar cell designs. Using these tools, the SiO2 and Si3N4 layers silicon solar cell structure was designed by using ATHENA device simulator meanwhile the J-V characteristics for ARC analysis were showing by ATLAS framework. This simulation concentrated on high temperature during annealing process when ARC thickness is 50 nm deposited on top surface of silicon solar cell. For SiO2 results, the FF value are approximately 0.758 and efficiency, is nearly 9.43%. In annealing process, the temperature become higher resulted increasing of pn junction, but not to open circuit voltage, Voc and short circuit current density, Jsc values. The Voc and Jsc were slowly decreased when temperature increased. This can be concluded the SiO2 layer not recommended for high temperature in annealing process. As for Si3N4 result, the calculated FF in range of 0.7570.758 and efficiency is around 9.559.57%. In annealing process, the temperature increasing constantly follows the increasing of pn junction and Jsc, meanwhile the Voc is decreased slowly. The optimum efficiency from SiO2 layers is 9.428% when temperature reach at 950C, while the optimum efficiency from Si3N4 layers is 9.565% when temperature at 1050C. Solar cell simulation could be useful for time saving and cost consumption. This method also cheaper and faster compared to experimental. So the simulation has some advantages than physical experimental to made decision to fabricate a solar cell.

Acknowledgement
The authors would like to thank the Malaysian Ministry of Science, Technology & Environment for the grant of UKM-RS-06-FRGS0001-2007.

453

A. Lennie, H. Abdullah, S.M.Mustaza and K.Sopian ATLAS Users manual, 1998. Silvaco Software version 5.6.O.R. Silvaco International. 6th Edition. Vol. 1 & 2. Bouhafs, D., Moussi, A., Chikouche, A. & Ruiz, J.M. 1998. Design and simulation of antireflection coating systems for optoelectronic devices: Application to silicon solar cells. J. Solar Energy Materials and Solar Cells 52: 7993. Chapin, D.M., Fuller, C.S. & Pearson, G.L. 1954. J. Appl. Phys. 74, pp. 230. Hsu, C.W., Wang, L. & W.F. Su, 2009. Effect of chemical structure of interface modifier of TiO2 on photovoltaic properties of poly(3-hexylthiophene)/TiO2 layered solar cells, Journal of Colloid and Interface Science 329: 182187. Micheal, S. & Bates, A.D. 2005. The design and optimization of advanced multijunction solar cells using the Silvaco ATLAS software package. J. Solar Energy Materials and Solar Cells 87: 785794. Micheal, S., Bates, A.D. & Green, M.S. 2005. Silvaco Atlas as a solar cell modeling tool. IEEE. 719-721. Neamen, D.A. 2006. An Introduction to Semiconductor Devices, Mc Graw Hill, New York. Resnik, D., Vrtacnik D. & S. Amon, 1999. Optimization of SiO2/Si3N4 Antireflective Coating on P+NN+ Photodiode, IEEE Africon, 5th Africon Conf. in Africa, pp. 11531156. Weider, S. 1982. An Introduction to Solar Energy for Scientists and Engineers, John Wiley. Zhao J. & M.A. Green, 1991. Optimized antireflection coatings for high-efficiency Si solar cells, IEEE Trans. on ED 38: 1925-1934

References
[1] [2]

[3] [4]

[5]

[6] [7] [8] [9] [10]