Chapter 4 Result and Discussion
4.3 Investigation of Cu(In,Ga)Se 2 Films After Selenization
Figure 4-9 and Figure 4-15 shows SEM cross-section images of CIGS films of different plasma power and substrate temperature. We can find the MoSe2 between CIGS absorber layer and Mo from the cross-section images of different parameters. The MoSe2 layers assist CIGS/Mo to change Ohmic-type contact, which is better than Schottky-type contact of without MoSe2. The Ohmic-type contact can improve leakage current of solar cell [63]. Additionally, not only improve the electric properties but also improve adhesion strength of CIGS to Mo, because the orientation of MoSe2 is perpendicular to the Mo surface.
Figure 4-12 and Figure 4-18 shows GIXRD spectrum of CIGS thin films of different plasma power and substrate temperature. We can find (100) and (110) peak of MoSe2, which is perpendicular to the Mo surface.
Though, thickness of MoSe2 layers of different parameters is different, but the orientation of MoSe2 is more important than its thickness [64].
4.3.2 Cu(In,Ga)Se
2Films of Different Plasma Power
The sample of without plasma is denser than with plasma and the surface roughness of without plasma is smoother than with plasma in the Figure 4-10. The sample of without plasma has less void and smoother surface because the selenium thin films have some crystallization. The crystalline selenium gets more pre-diffusion time than amorphous selenium because diffusion coefficient of amorphous selenium is far larger than crystalline selenium. Therefore, the element In and Ga
63
inter-diffusion in the time of pre-diffusion. The surface roughness is decreased with increased plasma power due to crystalline selenium film.
We use AFM to measure surface roughness show in the Figure 4-11 and this data can demonstrate above mentioned.
The GIXRD spectrum is show in the Figure 4-12. In the figure, the CGS peak gradually increases with increased the plasma power, which is CGS separable phase, because the sample of without plasma have a little crystalline selenium. These crystalline selenium have more diffusion time cause to more uniform Cu(In,Ga)Se2 thin films. So the sample of without plasma has less CGS peak than 50W, 60W and 70W show in Figure 4-12.
Figure 4-13 shows representative peak position for (112) peak of different plasma power. According to other paper, (112) peak position will increase with increased the Ga concentration (Ga/Ⅲ). Because the atom radius of indium is larger gallium, so the lattice constants will decrease and increase peak position (2 ) of XRD as incorporation of the gallium element in the indium sites. The (112) peak position is 26.665 for Ga concentration is 0% and 27.769 for Ga concentration is 100%. So the trend of (112) peak position is the same trend of Ga concentration and the Ga concentration of with plasma is smaller than without plasma because consume Ga to generate the CIG separable phase. We use EDS to measure the element composition of CIGS thin films, which are depicted in Table 4-11. The trend of Ga concentration with plasma power is the same GIXRD spectrum. The FWHM for (112) peak of CIGS thin films of plasma power are depicted in Figure 4-14, and the FWHM represent degree of crystallization of CIGS films. From this figure, the FWHM is
64
decreased with increased the plasma power, but too high plasma power will critically damage the surface. And then we can see the sample of without plasma (0W) is larger than other sample because distort the normal lattice structure of CuInSe2 as the incorporation of the gallium element in the indium sites. Thickness of CIGS/MoSe2/Mo/Glass sample and RMS roughness of CIGS thin films are show in Table 4-10.
4.3.3 Cu(In,Ga)Se
2Films of Different Substrate Temperature
Figure 4-15 shows the SEM cross-section images of different substrate temperature. The CIGS thin film of 85℃ has many voids below the CIGS thin film. The voids influence film quality and cause to worst contact to back electrode. The reason may be the intensity of crystalline orientation (100) and (101) of selenium are same strength, cause to mismatch the lattice structure. The SEM top-view images are more and smoother with increased substrate temperature in Figure 4-16 because the crystalline selenium thin films need more time to dissociate. The results of AFM measurement is surface morphology and roughness show in Figure 4-17 and Table 4-12, which match trend of SEM top-view.
The GIXRD spectrum is show in the Figure 4-18. In the figure, the CGS peak gradually decreases with increased the substrate temperature, which is CIG separable phase. The crystalline selenium get more pre-diffusion time than amorphous selenium because diffusion coefficient of amorphous selenium is far larger than crystalline selenium. These crystalline selenium have more diffusion time cause to more uniform Cu(In,Ga)Se2 thin films. So the CGS peak will disappear at high substrate temperature. Therefore, the Ga concentration of CIGS thin films
65
gradually increases with increase substrate temperature. Additionally, the (112) peak position of CIGS thin films is more and larger with substrate temperature is more and higher show in Figure 4-19. The element compositions of EDS also appear this trend show in Table 4-13. We use the FWHM of (112) peak to determine the films quality of CIGS show in Figure 4-20. We can find the films quality of CIGS is more and worse with increased substrate temperature. The reason may be the (101) peak of selenium gradually increase with increase substrate temperature.
Figure 4-21 is show the depth analysis of Auger electron spectroscopy which is prepared with different substrate temperature, and we can find the Ga distributed concentration is more uniform with increased substrate temperature. The crystalline selenium thin films can cause to have much pre-diffusion time, so In and Ga element can uniformly diffuse in selenization process. In this figure can find the Se single to penetrate in the raised Mo single, which represent the MoSe2
layer is observed between Mo layer and Cu(In,Ga)Se2. Additionally, the Se single penetrate in the Mo single, which is decreased with increased substrate temperature. The result may be due to different crystallization of selenium thin film on the CIG precursor layer have different diffusion coefficient.
4.3.4 The Electric Properties of Solar Cell Device of Cu(In,Ga)Se
2Figure 4-22 compares the I-V curve for different plasma power.
Obviously, the short-circuit current of plasma power 60W is larger than without plasma and open-circuit voltage has insignificant difference. But the conversion efficiency of plasma power 60W slightly improves
66
compares to without plasma due to the effect of fill factor. The fill factor will change by the series-resistance and shunt-resistance. The shunt- resistance of without plasma is far larger compare to with plasma power 60W and series-resistance of without plasma insignificant difference compare to with plasma 60W show in Table 4-12. The difference of the shunt-resistance relates to leakage current and leakage current is increased with decreased shunt-resistance. In the Figure 4-23, we can find the quantum efficiency of sample of plasma power 60W is the highest compare to sample of different plasma power. And the short-circuit current of AM1.5G is given by
∫ (4-1)
, where Gλ is the spectral irradiance according to a reference distribution like the most commonly used AM 1.5. The short-circuit current of AM1.5G can decide a part of short-circuit current. Therefore, this is why the sample of plasma power 60W has the highest short-circuit current because the quality of absorption layer is better than without plasma.
The I-V curve of sample of different substrate temperature show in Figure 4-24. We can find open-circuit voltage, short-circuit current and fill factor at high substrate temperature (125℃) are higher than low substrate temperature. The reason of different open-circuit voltage may be the Ga concentration, and the Ga concentration is increased with increased substrate temperature. The band gap of CIGS absorption layer is increased with increase Ga concentration, which cause to open-circuit voltage is larger. Figure 4-25 shows EQE measurement of the CIGS solar cells prepared with different substrate temperature. Obviously, the
67
quantum efficiency of high substrate temperature (125℃) is higher than other sample in all region of wavelength. According to the function of (4.1), the short-circuit current is larger than other sample due to the quantum efficiency. Moreover, the quantum efficiency is the lowest among these curves, which substrate temperature is 85℃. This reason is the poor contact between Mo and CIGS absorption because the CIGS films generate many voids in the selenization process. We arrange these data of substrate temperature show in Table 4-14. We can find the conversion efficiency of high substrate temperature is larger compare to other sample. Because the roughness and film uniform of high substrate temperature are better than other sample, and cause to the coverage by buffer layer (CdS) is integral. The buffer layer is deposited on the CIGS absorption layer in order to avoid damaging the CIGS surface and producing the defect, possibly formed by sputtering of ZnO films. So the conversion efficiency may be increased with increased coverage of the buffer layer.
68
Figure 4-1 SEM cross-section images of deposited In/Cu3Ga film on the Mo/SLG by sputtering.
Figure 4-2 GIXRD spectrum of deposited In/Cu3Ga on the Mo/SLG by sputtering.
69
Figure 4-3 SEM top-view images of deposited selenium on the CIG precursor layer by APPECVD prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
(b)
(c) (d)
(a)
70
Figure 4-4 AFM images of deposited selenium on the CIG precursor layer by APPECVD prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
(a) (b)
(c) (d)
71
Figure 4-5 GIXRD spectrum of deposited selenium on the CIG precursor layer by APPECVD prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
Figure 4-6 SEM top-view images of selenium film of different substrate temperatures (a) 45℃ (b) 85℃ (c) 125℃.
(a) (b)
(c)
72
Figure 4-7 AFM images of deposited selenium on the CIG precursor layer by APPECVD prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
(a) (b)
(c)
73
Figure 4-8 GIXRD spectrum of deposited selenium on the CIG precursor layer by APPECVD prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
74
Figure 4-9 SEM cross-section images of CIGS thin films prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
(b)
(c) (d)
(a)
75
Figure 4-10 SEM top-view images of CIGS thin films prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
(The part of right half was 10K magnification images; the part of left half was 30K magnification images)
(a)
(b)
(c)
(d)
76
Figure 4-11 AFM images of CIGS thin films prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
(a) (b)
(c) (d)
77
Figure 4-12 GIXRD spectrum of CIGS thin films prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
Figure 4-13 Variation of the peak position (2 ) for (112) peak of CIGS thin films with different plasma power.
78
Figure 4-14 Variation of the FWHM for (112) peak of CIGS thin films with different plasma power.
Figure 4-15 SEM cross-section images of CIGS thin films prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
(a) (b)
(c)
(a)
79
Figure 4-16 SEM top-view images of CIGS thin films prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃. (The part of right half was 10K magnification images; the part of left half was 30K magnification images)
(d)
(c)
(a)
80
Figure 4-17 AFM images of CIGS thin films prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
(a) (b)
(c)
81
Figure 4-18 GIXRD spectrum of CIGS thin films prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
Figure 4-19 Variation of the peak position (2 ) for (112) peak of CIGS thin films with different substrate temperature.
82
Figure 4-20 Variation of the FWHM for (112) peak of CIGS thin films
with different substrate temperature.
(a)
83
Figure 4-21 Depth analysis of Auger Electron Spectroscopy (AES) prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
(b)
(c)
84
Figure 4-22 I-V curve measurement of the Cu(In,Ga)Se2 solar cells on the AM1.5G solar simulator prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
Figure 4-23 Extra quantum efficiency measurement of the Cu(In,Ga)Se2 solar cells prepared with different plasma power (a) without plasma (b) 50W (c) 60W (d) 70W.
85
Figure 4-24 I-V curve measurement of the Cu(In,Ga)Se2 solar cells on the AM1.5G solar simulator prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
Figure 4-25 Extra quantum efficiency measurement of the Cu(In,Ga)Se2
solar cells prepared with different substrate temperature (a) 45℃ (b) 85℃ (c) 125℃.
86
Table 4-1 Deposition selenium films on the glass without plasma prepared with different Se source temperature (0W 10SLM 550 torr).
Temperature(℃) 325 335 345
Thickness(nm) 570 840 1030
SEM Cross-section
SEM Top View
Table 4-2 Deposition selenium films on the glass without plasma prepared with different main gas flow rate (0W 335℃ 550 torr).
Main gas flow
rate (SLM) 5 10 15
Thickness (nm) 510 840 595
SEM Cross-section
SEM Top View
87
Table 4-3 Deposition selenium films on the glass without plasma prepared with different background pressure (0W 335℃ 10SLM).
Pressure (torr) 550 350 150
Thickness(nm) 840 1220 1640
SEM Cross-section
SEM Top View
Table 4-4 Deposition selenium films on the glass with plasma prepared with different Se source temperature (50W 20SLM 150torr).
Temperature(℃) 345 355 365
Thickness (μm) 1.55 1.94 2.87
SEM Cross-section
SEM Top View
88
Table 4-5 Deposition selenium films on the glass with plasma prepared with different main gas flow rate (50W 355℃ 150torr).
Main gas flow
rate (SLM) 15 20 25
Thickness(μm) 1.88 1.94 1.47
SEM Cross-section
SEM Top View
Table 4-6 Deposition selenium films on the glass without plasma prepared with different background pressure (50W 355℃
10SLM).
Pressure (torr) 550 350 150
Thickness(μm) 0.58 1.01 1.94
SEM Cross-section
SEM Top View
89
Table 4-7 Deposition selenium films on the glass without plasma prepared with different plasma power (355 ℃ 10SLM 150torr).
Plasma
Power(W) 50W 60W 70W
Thickness(μm) 1.94 1.52 1.22
SEM Cross-section
SEM Top View
Table 4-8 Thickness and RMS roughness of deposited selenium on the CIG precursor layer by APPECVD prepared with different plasma power.
Plasma Power Thickness (nm) RMS roughness (nm)
0W 2148.89 10.06
50W 2950.00 58.68
60W 3131.26 33.27
70W 2890.91 6.22
90
Table 4-9 Thickness and RMS roughness of deposited selenium on the CIG precursor layer by APPECVD prepared with different substrate temperature.
Substrate Temperature Thickness (nm) RMS roughness (nm)
45℃ 2950.00 58.68
91
Table 4-11 EDS results for atomic composition of the selenium thin films prepared with different plasma power.
Table 4-12 Thickness of CIGS/MoSe2/Mo/Glass sample and RMS roughness of CIGS films prepared with different substrate temperature.
92
Table 4-13 EDS results for atomic composition of the selenium thin films prepared with different substrate temperature. Ⅲ: elements of group III.
Table 4-14 Photovoltaic characteristics of CIGS-based solar cells including six devices fabricated with varied plasma power and substrate temperature. Voc: open-circuit voltage, Jsc: short-circuit current, FF: fill factor, Rshunt: shunt resistance and Rseries: series resistance.
Plasma
93
Chapter 5 Conclusions
We have successfully deposited selenium film on the CIG precursor layer by atmospheric pressure plasma, which can enhance chemical vapor deposition.
The selenium films are denser, smoother by APPECVD. The APPECVD contains smaller amount of Se than co-evaporation system. In this thesis, we have used rapid thermal process to selenizate the precursor layer, and successfully fabricated the absorber layer of Cu(In,Ga)Se2 thin films. All films showed a strong preferred (112) orientation and the MoSe2 layer was found between Cu(In,Ga)Se2 and Mo. The MoSe2 not only improved the electric properties but also improved adhesion strength of Cu(In,Ga)Se2 to Mo. Furthermore, the effect of increased substrate temperature raised band gap of Cu(In,Ga)Se2 thin films in our research due to different crystallization of selenium thin films. The Ga concentration (Ga/Ⅲ) of Cu(In,Ga)Se2 thin film is between 0.2 and 0.3, matching band gap which is between 1.104eV and 1.158eV. The effect of increased plasma power improved the crystallization of chalcopyrite structure of Cu(In,Ga)Se2 thin films. Those two effects mentioned-above possibly improve the conversion efficiency of solar cell device. Therefore, we complete solar cell, whose structure is Al/ITO/ZnO/CdS/CIGS/
Mo/SLG, covering 0.48 cm2. The highest conversion efficiencies of solar cell device without plasma and with plasma are respectively 4.694% and 5.031% when the substrate temperature is . The conversion efficiency
94
of solar cell device with plasma power 50W is 2.266% when the substrate temperature is 45℃. The conversion efficiency of solar cell device with plasma power 50W is 6.103% when the substrate temperature is 125℃, which FF=0.428, Voc=0.41V and Jsc=34.815 mA/cm2. Our study hopefully facilitates to enhance films quality and reduce the costs by APPECVD.
95
Chapter 6 Future works
In this thesis, some CIGS thin film samples has the CIG separate phase. We can decrease the CIG separate phase result in the region of Ga concentration smaller. Increase the investigation of selenization temperature and time of during process for rapid thermal process. Now, we are limited by the equipment, cannot to sharply increase the plasma power. So the effect of different plasma power is not significant. In the future, we want to do investigation of CIGS thin films by higher plasma power. The scan area will increase to the 10∙10cm2 in order to fabricate large area substrate. Additionally, selenium film need to improve the uniform by add the Shower head at the exit of carrier gas. We want to decrease the highest temperature of selenization by APPECVD and in-site selenization because the Se-radical have high chemical potential.
But the adjustment of priority is the wire of the equipment, and the wire need to bear high temperature (>350℃) and high voltage. We wish to prepare the CIGS solar cell, which have higher conversion efficiency than this thesis.
96
References
[1] W. Shockley, et al., “Detailed Balance Limit of Efficiency of pn Junction Solar Cells,” Journal of Applied Physics, vol. 32, pp.
510-519, 1961.
[2] P. Bermel, et al., “Improving Thin-Film Crystalline Silicon Solar Cell Efficiencies with Photonic,” Optics express, vol. 15, pp. 16986-17000, 2007.
[3] E. Alsema, et al., “Energy Requirements of Thin-Film Solar Cell Modules－a review,” Renewable and Sustainable Energy Reviews, vol. 2, pp. 387-415, 1998.
[4] M. A. Contreras, et al., “Progress Toward 20% Efficiency in Cu(In,Ga)Se2 Polycrystalline Thin-Film Solar Cells,” Progress in
Photovoltaics: Research and Applications, vol. 7, pp. 311-316, 1999.
[5] T. Negami, et al., “Cu(In,Ga)Se2 Thin-Film Solar Cells with an Efficiency of 18%,” Solar energy Materials & Solar Cells, vol. 67, pp.
331-335, 2001.
[6] H. S. Ullal, et al., “Polycrystalline Thin-Film Photovoltaic: Research, Development, and Technologies,” Proceedings of the 29th IEEE
Photovoltaic Specialists Conference, pp.472-477, 2002.
[7] NREL, “Konarka Power Plastic Reaches 8.3% Efficiency,”
pv-tech.org, 2011.
[8] F. Kang, et al., “Structure and photovoltaic characteristics of CuInSe2
thin films prepared by pulse-reverse electro-deposition and
97
selenization process,” Journal of Alloys and Compounds, vol. 478, pp.
L25-L27, 2009.
[9] Y. Hamakawa, et al., “Thin Film Solar Cells: Next Generation Photovoltaics and Its Applications,” 2004.
[10] S. Chen, et al., “Band-structure anomalies of the chalcopyrite semiconductors CuGaX2 versus AgGaX2 (X=S and Se) and their alloys,” Physical Review B, vol. 75, pp. 205-209, 2007.
[11] D. K. Suri, et al., “X-ray study of CuGaxIn1-xSe2 solid solutions,”
Journal of Applied Crystallography, vol. 22, pp. 578-583, 1989.
[12] T. F. Ciszek, et al., “Growth and properties of CuInSe2 crystals produced by chemical vapor transport with iodine,”
Journal of Crystal Growth, vol. 70, pp. 405-410, 1984.
[13] H. Neumann, et al., “Optical properties and electronic band structure of CuInSe2,” Solar Cells, vol. 16, pp. 317-333, 1986.
[14] S. M. Wasim, et al., “Transport properties of CuInSe2,” Solar Cells, vol. 16, pp. 289-316, 1986.
[15] R. Herberholz, et al., “Phase segregation, Cu migration and junction formation in Cu(In,Ga)Se2,” The European Physical Journal Applied
Physics, vol. 6, pp. 131-139, 1999.
[16] D. Schmid, et al., “A comprehensive characterization of the interfaces in Mo/CIS/CdS/ZnO solar cell structures,” Solar energy
materials and solar cells, vol. 41-42, pp. 281-294, 1996.
[17] S. H. Wei, et al., “Effects of Na on the electrical and structural properties of CuInSe2,” Journal of applied physics, vol. 85, pp.
7214-7218, 1999.
98
[18] C. Rincon, et al., “Defect physics of the CuInSe2 chalcopyrite semiconductor,” Journal of Physics and Chemistry of Solids, vol. 60, pp. 1865-1873, 1999.
[19] I. Repins, et al., “19.9%-effcient ZnO/CdS/CuInGaSe2 Solar Cell with 81.2% Fill Factor,” Progress in Photovoltaics: Research and
Applications, vol. 16, pp. 235-239, 2008.
[20] N.A.K. Abdul-Hussein, et al., “Electrical Properties of RF Sputtered Thin Films of CuInSe2 (I),” Crystal Research and Crystal
Technology, vol. 20, pp. 509-514, 1985.
[21] F. R. White, et al., “Growth of CuInSe2 on CdS using Molecular Beam Epitaxy,” Journal of Applied Physics, vol. 50, pp. 544-545, 1979.
[22] S. H. Yoon, et al., “ Preparation of CuInSe2 Thin-Films through Metal Organic Chemical Vapor Deposition Method by using Di-μ-Methylselenobis and Bis Copper(II) Precursors,” Thin Solid
Films ,vol. 515, pp. 1544-1547, 2006.
[23] A. A. Akl, et al., “Growth, Microstructure, Optical and Electrical Properties of Sprayed CuInSe2 Polycrystalline Films,” Materials
Research Bulletin, vol. 43, pp. 1539-1548, 2008.
[24] J. F. Guillemoles, et al., “Recrystallization of Electrodeposited Copper Indium Di-Selenide Thin-Films in an Atmosphere of Elemental Selenium,”
Advanced Materials, vol. 6, pp. 376-379, 1994.
[25] M. Kaelin, et al., “Low-cost CIGS solar cells by paste coating and
[25] M. Kaelin, et al., “Low-cost CIGS solar cells by paste coating and