- Email: [email protected]
Fabrication of Sb2Se3 thin film solar cells by co-sputtering of Sb2Se3 and Se targets
Solar Energy 193 (2019) 275–282
Contents lists available at ScienceDirect
Solar Energy journal homepage: www.elsevier.com/locate/solener
Fabrication of Sb2Se3 thin film solar cells by co-sputtering of Sb2Se3 and Se targets
T
Changhao Maa,1, Huafei Guoa,1, Xin Wanga, Zhiwen Chena, Qingfei Canga, Xuguang Jiaa, Yan Lic, ⁎ ⁎ Ningyi Yuana, , Jianning Dinga,b, a
School of Materials Science and Engineering, Jiangsu Collaborative Innovation Center for Photovoltaic Science and Engineering, Jiangsu Province Cultivation Base for State Key Laboratory of Photovoltaic Science and Technology, Changzhou University, Changzhou 213164, China b Institute of Intelligent Flexible Mechatronics, Jiangsu University, Zhenjiang 212013, China c Changzhou Institute of Industry Technology, Changzhou, Jiangsu Province 213164, China
ARTICLE INFO
ABSTRACT
Keywords: Sb2Se3 Co-sputtering Thin film Selenium supplement
Antimony selenide (Sb2Se3) has attracted scientific interest due to its many advantages such as its appropriate band gap, non-toxicity, element abundance, and high absorption coefficient (> 105 cm−1). However, selenium (Se) is easily lost at high vapor pressures, thus leading to Se vacancies and increased recombination centers in films. In this work, Sb2Se3 films with good crystallinity were fabricated by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. The crystal structural properties, surface morphology, and optical and electrical properties of films with different Se contents were explored. Finally, we fabricated solar cells with a structure of FTO/CdS/Sb2Se3/Au and achieved an efficiency up to 3.47%.
1. Introduction
7.6% was achieved by VTD method (Wen et al., 2018). Chen et al. prepared Sb2Se3 solar cells based on La-doped SnO2 and achieved an efficiency of 3.25% (Chen et al., 2019b). Wang et al. added CeO2 interlayer between CdS and Sb2Se3 and improved the efficiency to 5.14% (Wang et al., 2019). Tao et al. investigated the electronic transport mechanisms in Sb2Se3 thin-film solar cells and fabricated the solar cells with the efficiency of 6.24% (Tao et al., 2019). And they investigated the electrically-active defects in Sb2Se3 thin-film solar cells and achieved an efficiency of 5.91% (Hu et al., 2018b). Evaporation and sputtering are also alternative methods to vacuum preparation of Sb2Se3 thin films. Liu deposited Sb2Se3 films by thermal evaporation using an electron beam, and the fabricated photovoltaic device with a configuration of FTO/Sb2Se3/CdS/ZnO/ZnO:Al/Au achieved an efficiency of 2.1% (Liu et al., 2014). Chen et al. studied the effects of substrate temperature on material and photovoltaic properties of magnetron-sputtered Sb2Se3 thin films but the efficiency is just only 0.84% (Chen et al., 2019a). Many studies have shown that, during evaporation, Sb2Se3 can decompose into Sb, Se, and SbSe, which leads to Se loss. The resulting Se vacancies in the films increase the recombination density (Li et al., 2016), which has a large negative impact on the films and solar cells. Thus, many studies have focused on the selenization of Sb2Se3 thin films to fill the Se vacancies. For example, Li
Recently, Sb2Se3 has attracted the interest of researchers due to its many advantages. Unlike CdTe, Sb2Se3 is a non-toxic and earth-abundant material, and its appropriate band gap (~1.1 eV) (Chen et al., 2015) and high absorption coefficient (> 105 cm−1) (Zhou et al., 2014) make it promising in terms of photovoltaic materials. Furthermore, Sb2Se3 has a one-dimensional crystal structure. Ribbons of (Sb4Se6)n stack along the [0 0 1] direction through strong covalent Sb-Se bonds, whereas in the [1 0 0] and [0 1 0] directions, the (Sb4Se6)n ribbons are held together by van der Waals forces (Zhou et al., 2015). This crystal structure improves charge transport and reduces the occurrence of recombination (Chen et al., 2017a). This is a promising material for commercialization. Magnetron sputtering is a simple method to fabricate Sb2Se3 thin films which also has a good uniformity and a promising for the Sb2Se3 large-scale industrial production. Originally, Zhou et al. prepared Sb2Se3 thin films by the RTE method, then fabricated solar cells with a structure of FTO/CdS/ Sb2Se3/Au and achieved an efficiency of 5.6% (Zhou et al., 2015). Lee et al. investigated the effects of low temperature pre-annealing of the absorbers fabricated by thermal evaporation and achieved a maximum efficiency of 1.47% (Lee et al., 2018). Subsequently, an efficiency of Corresponding authors. E-mail addresses: [email protected] (N. Yuan), [email protected] (J. Ding). 1 Changhao Ma and Huafei Guo contributed equally to this work. ⁎
https://doi.org/10.1016/j.solener.2019.09.046 Received 23 July 2019; Received in revised form 11 September 2019; Accepted 12 September 2019 0038-092X/ © 2019 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
Solar Energy 193 (2019) 275–282
C. Ma, et al.
et al. prepared an Sb2Se3 thin film by the co-evaporation of Sb2Se3 and Se (Li et al., 2016), and Shongalova presented a sputtering method to prepare Sb2Se3 films followed by selenized annealing in a poisonous H2Se gas atmosphere (Shongalova et al., 2018). In this study, Sb2Se3 films with good crystallinity were fabricated by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. This method is simple and resolved the problem of excess vacancies caused by Se loss. We studied the effects of different Se contents on film properties and investigated film structure, morphology, and optical and electrical performance. The films with supplemental Se displayed a preferred orientation along the [2 2 1] direction, which provides easier carrier transport. Supplemental Se also improved the morphology and optical and electrical properties of the films. Solar cells with a structure of FTO/CdS/Sb2Se3/Au were fabricated and achieved efficiencies of up to 3.47%.
CdS with different Se sputtering powers. The diffraction peaks of the films correspond to orthorhombic Sb2Se3 by comparison with the standard PDF card (JCPDS#15-0861) and the diffraction peak marked with * at 26.5° was attributed to CdS. However, compared with the film prepared without Se co-sputtering, all the films with Se co-sputtering display a preferred orientation along the [2 2 1] direction. The films with a preferred orientation along [1 2 0] are formed by (Sb4Se6)n ribbons stacked horizontally parallel to the substrate, whereas the [2 2 1]oriented films are formed by vertically stacked oblique (Sb4Se6)n ribbons on the substrate. Carrier transport in the [2 2 1]-oriented grains is easier than in those oriented in the [1 2 0] direction, because the carriers in the [2 2 1] direction are transport between the covalently bonded (Sb4Se6)n ribbons, but in [1 2 0] direction carriers are transport between (Sb4Se6)n ribbons combined with van der Waals forces. Therefore, the preferred orientation of the film along [1 2 0] will significantly increase the series resistance and reduce the efficiency of the device (Zhou et al., 2015). The films with supplemental Se also exhibit more obvious peak in the [2 1 1] direction compare to the film without supplemental Se, which the grain is composed of titled (Sb4Se6)n ribbons to different angle vertically on the substrate. The narrow halfheight peak widths of the preferred orientation are 0.368°, 0.334°, 0.313° and 0.321° respectively to the film without supplemental Se and with the Se power is 5 W, 10 W and 15 W, which also show that the films prepared with Se co-sputtering with a preferred orientation along [2 2 1] have relatively good crystallinity compare with the film without supplemental Se. The film without supplemental Se exhibits a preferred orientation along [1 2 0], in addition to this, the orientation along [2 3 0] is also very obvious. In order to quantitate the difference of orientations of the diffraction peaks of Sb2Se3 thin films deposited on CdS with different power of Se, the texture coefficients (TC) of the diffraction peaks shown in Fig. 1(b) were calculated based on the following equation (1) (Li et al., 2017), where the I(h k l) was the observed peak intensity of (h k l) plane and I0(h k l) was the intensity value in the standard XRD pattern, N was the total number of reflections considered for the calculation.
2. Experimental 2.1. Fabrication of Sb2Se3 films Sb2Se3 films were deposited by RF magnetron co-sputtering with Sb2Se3 and Se targets (99.99%, Hefei Kejing Materials Technology, China). The background pressure of the chamber was 4.0 × 10−6 Torr, and the substrate temperature was 350 °C. Then, Ar was introduced at a rate of 25 sccm, and the sputtering pressure was maintained at 5.0 × 10−3 Torr. The Se content was controlled by adjusting the sputtering power of the Se target. The power of the Sb2Se3 target was 30 W, and the power of the Se target was set to 0 W, 5 W, 10 W, or 15 W. 2.2. Fabrication of Sb2Se3 solar cells The structure of the solar cells was FTO/CdS/Sb2Se3/Au. The 60 nm CdS films were deposited on FTO substrate (with sheet resistance of about 6 Ω sq−1, transmittance about 80%) by the chemical bath deposition method described in our previous work (Guo et al., 2019). Then, Sb2Se3 films approximately 550 nm in thickness were deposited by the sputtering procedure described above. Gold electrodes (0.025 cm2) with a thickness of 80 nm were prepared by evaporation.
TCh
k l=
I(h I0(h
k l) k l)
/
1 N
N
I(h I 0(h
k l) k l)
(1)
The film without supplemental Se exhibits [h k 0] preferred orientation such as [1 2 0], [2 3 0] and [2 4 0]. However, the films with Se co-sputtering exhibit [h k 1] preferred orientation such as [2 2 1] and [2 1 1]. In order to prove this change is not caused by the change of sputtering parameters, we compared the XRD patterns of the Sb2Se3 films deposited on CdS at different substrate temperature with different power of Se in Fig. 1(c). The films deposited at 300 °C and 350 °C without supplemental Se both exhibit a preferred orientation along [1 2 0]. However, the films with Se co-sputtering exhibit an obvious preferred orientation along [2 2 1]. When the substrate temperature is 250 °C, the films doesn’t exhibit a good crystallinity. Fig. 1(d) shows the Raman spectra of the Sb2Se3 films with different Se contents, which is another method to detect film composition. The obtained results were compared with other reported results (Liang et al., 2018; Tao et al., 2019). The peaks at 189 cm−1, and 210 cm−1 were assigned to the SbSe vibration. When the Se target power was increased to 15 W, a peak close to 256 cm−1 became apparent, indicating a relatively high Se content in this film (Salomé et al., 2014). Compared with the film without supplemental Se, the intensity of the films with Se co-sputtering increase slightly and the half-height peak widths decreased indicating that the improvement of crystalline. XPS was used to detect the element oxidation states of the Sb2Se3 films deposited on CdS with and without supplemental Se. The highresolution core level Sb3d and Se3d spectra are shown in Fig. 1(e) and (f). The C1s energy at 284.6 eV was used to correct the binding energies (BEs) in addition to the charge compensation by the flood gun associated with the spectrometer. Compared with the sample without
2.3. Characterization of films and solar cells The structures and crystallinities of the Sb2Se3 films deposited on CdS were detected by X-ray diffraction (XRD, Smart Apex II Duo), and Raman scattering analyses were performed using a micro-Raman spectrometer (Thermo Fisher) with a 532-nm excitation wavelength. Scanning electron microscopy (SEM, Hitachi) was used to observe the surface morphologies and cross-sections of different films, and the film element contents were determined by energy dispersive X-ray spectroscopy. Surface roughness was measured by atomic force microscope (AFM, Nanman), and optical performance was measured by UV–visible spectrophotometer (Shimadzu UV-2600). X-ray photoelectron spectroscopy (XPS) was performed on a photoelectron spectrometer (Kratos Axis Ultra DLD). Ultraviolet photoelectron spectroscopy (UPS) was measured with a monochromatic He I light source (21.2 eV) and a VG Scienta R4000 analyzer. The efficiency of the solar cells was measured by I-V tester (Cell tester) under AM1.5 global spectrum conditions with a radiance of 1000 W m−2. The external quantum efficiency (EQE) was detected by a photovoltaic characterization system (PV Measurements QEXL), and C-V was measured by a Keysight Precision LCR meter (Keysight E4980AL) in darkness at room temperature and 10 kHz, where the DC bias voltage was changed from −1 V to 0.5 V. 3. Results and discussion Fig. 1(a) shows the XRD patterns of the Sb2Se3 films deposited on 276
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 1. (a) The XRD pattern (b) The texture coefficients of the diffraction peaks of the Sb2Se3 thin films deposited on CdS with different power of Se. (c) The XRD pattern of Sb2Se3 thin films deposited on CdS with different power of Se under different substrate temperature. (d) The Raman spectra of the Sb2Se3 films with different Se content and the high-resolution core level spectra of (e) Sb3d and (f) Se3d for the films with 10 W Se and without supplemental Se.
supplemental Se, the films with Se co-sputtering have relatively highintensity doublets. The Sb3d5/2 and Sb3d3/2 spectra of the Se-supplemented films have peaks at 529.4 eV and 538.7 eV with a separation of 9.3 eV attributed to the charge state of Sb3+, which is similar to the results reported by others (Liang et al., 2017). The Se3d5/2 and Se3d3/2 peaks of the films are located at 53.8 eV and 54.6 eV with the separation of 0.8 eV, which are identical to that of metallic Se (Jin et al., 2013). Fig. 2(a–d) illustrates a significant difference between the surface morphologies of the Sb2Se3 films deposited on CdS with and without Se co-sputtering. The grains of the film without supplemental Se grows at different heights in different directions. This may be due to the loss of Se during sputtering and leading to the instability of the film during the sputtering process. However, the Se-supplemented films not only have good crystallinity but also the surface roughness is markedly reduced. (The wide range SEM images of the Sb2Se3 films deposited on CdS with different power of Se were shown in Fig. S1 in supplementary.) This result is supported by the AFM images of the different films shown in Fig. 2(e–h). The roughness of the film without Se was 61.5 nm, whereas Se target powers of 5 W, 10 W, and 15 W yielded Sb2Se3 film surface roughness of 33.5 nm, 30.5 nm, and 32.4 nm, respectively. This change
is due to the improvement of grain growth of the Sb2Se3 films compare with the film without supplemental Se. The films with smoother surface are more suitable for achieving better quality interface between the adjacent layers. This smooth surface was very helpful to reduce the recombination of light induced excitations at the interface of the absorber and metal electrode, which would be beneficial for the enhancement of cell performance (Lei et al., 2016). Although the growth of the grains was substantially improved, when Se was insufficiently supplemented, the produced film had relatively apparent holes, as shown in Fig. 2(b), which may have been caused by the loss of Se. On the other hand, as shown in Fig. 2(d), when the supplemental Se was excessive, more small grains were present in the film. Table 1 shows the chemical composition of the Sb2Se3 films fabricated with different Se target powers. The film without supplemental Se exhibited a Se loss compared with the normal proportion. The Se content increased with increasing power of Se. When the Se power was increased to 10 W, the resulting film had an excessive amount of Se. EDX elemental mapping images of the films without supplemental Se and with 10 W Se power are shown in Fig. 3. Previous work has shown that Sb-rich conditions lead to many defects with relatively low 277
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 2. The scanning electron microscope images of the Sb2Se3 films (a) without supplemental Se. (b) (c) and (d) the co-sputtering films with Se power of 5 W, 10 W and 15 W and the atomic force microscope images of the Sb2Se3 films (e) without supplemental Se. (f) (g) and (h) the co-sputtering films with Se power of 5 W, 10 W and 15 W.
formation energies such as Vse and SbSe. In contrast, under Se-rich conditions, more defects form with higher formation energies (Savory and Scanlon, 2019), these defects will be not easy to form. Therefore, more defects will occur when there is Se loss, and thus it is necessary to include supplemental Se by co-sputtering. The cross-sectional SEM images of solar cells without supplemental Se and with an Se co-sputtering power of 10 W are shown in Fig. 4(a) and (b). We can determine from the images that the Sb2Se3 film with
supplemental Se has better crystallinity and smaller roughness, whereas the film without supplemental Se exhibits disordered growth. Fig. 4(c) and (d) shows EDX mappings based on the cross-sections of the two samples. Compared with the Se-supplemented film in Fig. 4(d), the film without supplemental Se in Fig. 4(c) shows evident Se element loss, which may lead to Se vacancies in the film. In addition, from comparison of the Sb elemental mappings in Fig. 4(c) and (d), we suspect that Se loss may lead to the instability of Sb2Se3 and result in more 278
Solar Energy 193 (2019) 275–282
C. Ma, et al.
excessive elemental Se and the small grains as shown in Fig. 2(d) may result in a lower quality interface between Sb2Se3 and CdS and the degradation of the film quality, thus leading to the slightly decrease of the Jsc, Voc and FF. The diode ideality factor (A) and reverse saturation current density (J0) are shown in Table 2. Compare with the device without supplemental Se, the J0 and A of the devices with Se co-sputtering have a slight decline, showing that the reduction of the recombination loss and the improvement of the heterojunction in the devices. The EQE values of the devices are shown in Fig. 7(b), the spectral intensities of the cells with Se co-sputtering declined at shorter and longer wavelength ranges and peaked at 550 nm to 650 nm. The drop at short and long wavelength is mainly caused by the absorption by the CdS buffer layer and incomplete absorption of infrared photons and reduced collection efficiency of photo-generated carriers which are far from the depletion region. The devices with Se co-sputtering improved significantly. This improvement in performance may due to a reduced number of defects such as VSe, which can easily form during sputtering processes with large Se vapor pressures. Fig. 7(c) shows C−2V curves of the devices with different Se contents acquired in darkness at a frequency of 10 kHz and at room temperature. The built-in-voltages (Vbi) of the samples fabricated without Se, with 5 W, 10 W, and 15 W power of Se are 0.329 V, 0.483 V, 0.531 V and 0.510 V respectively. We calculated the depletion region width (Wd) are 178 nm, 184 nm, 190 nm and 187 nm respectively for the devices with different content of Se. The increasement of EQE at the long wave range can be caused by the enlargement of the depletion region width and thus leading to the more efficient charge collection through drifting (Chen et al., 2017b). Comparing with the sample without supplemental Se, the Vbi increased from 0.329 to 0.531 V. The samples with larger Vbi indicate a better contact and heterojunction quality at CdS/Sb2Se3 (Zeng et al., 2016) and the increase of Vbi can decrease the charge accumulation at electrode interfaces under illumination (Jiang et al., 2019). As Vbi is positively correlated to VOC, improved VOC has been observed in solar cells with Se co-sputtering. However, the increase of Voc is much smaller than the increase of Vbi, which could be caused by the Cd diffusion during sputtering (Hu et al., 2018a). Fig. 7(d) shows the statistical efficiency distribution charts of 16 solar cells with different Se contents. The PCEs of the devices without Se were within the range of approximately 0.60–0.89%, the 5 W Se power devices were within approximately 1.5–2.0%, and those of the 15 W Se power devices were within approximately 2.3–2.9%. In contrast, the PCEs of the devices with 10 W of Se power were within approximately 3.0–3.47%.
Table 1 The chemical composition of Sb2Se3 films of the normal proportion and with different power of Se. Sample Name
Sb
Se
Sb/Se
Normal proportion Without Se With 5 W Se With 10 W Se With 15 W Se
40 41.62 39.38 38.08 35.51
60 58.38 60.62 61.92 64.49
0.67 0.71 0.65 0.61 0.55
serious Sb diffusion. Fig. 5(a) shows the absorption spectra of the Sb2Se3 thin films deposited on CdS films and removed the CdS base background. Se cosputtering yielded a significant improvement in absorbance compared with the film without supplemental Se, and a sharp absorption edge was observed. The (αhυ)2-hυ plot of the Sb2Se3 thin films is shown in Fig. 5(b), illustrating that the band gap values of the films with supplemental Se are arranged from 1.26 to 1.28 eV. For the film without Se co-sputtering, the band gap value decreased slightly to 1.23 eV. This change could be caused by the crystallinity of the films. Fig. 6(a) and (b) show the UPS of the Sb2Se3 with different Se content. The energy level diagram pattern is shown in Fig. 6(c). The distance of the films without supplemental Se and with 10 W Se between the valence band maximum (VBM) and the Fermi energy were calculated to be 0.60 eV and 0.52 eV respectively. The VBM position of the film with Se cosputtering was closer to the Fermi energy. This indicates an increasement in hole concentration of the film with Se co-sputtering. Furthermore, the conduction band minimum (CBM) position of the co-sputtering film moved to the positive direction compare with the film without co-sputtering. So, the energy offset between the CBM of Sb2Se3 and the CBM of CdS is larger, suggesting that the photogenerated electron injection from the Sb2Se3 to CdS would be faster. The recombination between CdS and Sb2Se3 would be suppressed. Fig. 7(a) shows the J-V performances of the solar cells with different Se co-sputtering powers, and Table 2 lists the maximum efficiencies of the devices. With the appropriate increase in Se content the short-circuit current density (Jsc), open-circuit voltage (Voc) and fill factor (FF) improved significantly. We achieved an efficiency of up to 3.47% with Jsc, Voc, and FF of 19.2 mA/cm2, 0.350 V, and 51.6% respectively, when the Se power was 10 W. The decline of series resistance (Rs) and the increase of shunt resistance (Rsh) are obvious. The increment of Voc and Jsc with Se incorporation could be caused by the increasement of collection efficiency of photo-generated carriers and the improvement of CdS/Sb2Se3 interface. When the Se target power increased to 15 W, the
Fig. 3. EDX elemental mapping images of the Sb2Se3 films (a) without supplemental Se (b) with 10 W Se. 279
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 4. Cross-sectional SEM images of the solar cells (a) without supplemental Se (b) with the co-sputtering power of Se is 10 W and the EDX mapping based on the cross section of FTO/CdS/Sb2Se3 (c) without supplemental Se (d) with the co-sputtering power of Se is 10 W.
Fig. 5. (a) The absorption spectrum and (b) The (αhυ)2-hυ plot of the Sb2Se3 thin films with different Se co-sputtering power.
4. Conclusion
Se, the films prepared by co-sputtering exhibited better crystallinities and a preferred orientation along the [2 2 1] direction, which provides easier carrier transport. Supplemental Se decreased the number of Se vacancies in the films and improved their optical and electrical properties. We fabricated solar cells with a structure of FTO/CdS/Sb2Se3/Au and achieved an efficiency as high as 3.47%.
Sb2Se3 films were prepared by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. The crystal structural properties, surface morphologies, and optical and electrical properties of the films with different Se contents were explored. Compared with the film without supplemental 280
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 6. (a) and (b) are the UPS spectra of Sb2Se3 films with different Se deposited on CdS film. (c) is the Schematic of the energy alignment.
Fig. 7. (a) J-V curves (b) EQE (c) C−2-V plots and (d) The statistical efficiency distribution chart of the solar cells with different content of Se.
281
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Table 2 Device characteristics of the solar cells with different content of Se. Sample Name
Voc (V)
Jsc (mA/cm2)
Fill Factor (%)
Efficiency (%)
Rs (ohm)
Rsh (ohm)
A
J0 (mA/cm2)
Without Se With 5 W Se With 10 W Se With 15 W Se
0.251 0.311 0.350 0.321
9.1 15.9 19.2 18.1
39.1 41.4 51.6 51.1
0.90 2.04 3.47 2.97
459.7 366.2 222.2 202.2
2224.6 2363.9 3950.0 4048.5
1.68 1.65 1.52 1.56
3.71E-04 2.75E-04 1.12E-04 1.67E-04
Acknowledgement
Lee, D., Cho, J.Y., Heo, J., 2018. Improved efficiency of Sb2Se3/CdS thin-film solar cells: the effect of low-temperature pre-annealing of the absorbers. Sol. Energy 173, 1073–1079. Lei, H.W., Yang, G., Guo, Y.X., Xiong, L.B., Qin, P.L., Dai, X., Zheng, X.L., Ke, W.J., Tao, H., Chen, Z., Li, B.R., Fang, G.J., 2016. Efficient planar Sb2S3 solar cells using a lowtemperature solution-processed tin oxide electron conductor. Phys. Chem. Chem. Phys. 18 (24), 16436–16443. Li, Z., Zhu, H., Guo, Y., Niu, X., Chen, X., Zhang, C., Zhang, W., Liang, X., Zhou, D., Chen, J., 2016. Efficiency enhancement of Sb2Se3 thin-film solar cells by the co-evaporation of Se and Sb2Se3. Appl. Phys Express 9 (5), 052302. Li, Z., Chen, X., Zhu, H., Chen, J., Guo, Y., Zhang, C., Zhang, W., Niu, X., Mai, Y., 2017. Sb2Se3 thin film solar cells in substrate configuration and the back contact selenization. Sol. Energy Mater. Sol. Cells 161, 190–196. Liang, G.X., Zhang, X.H., Ma, H.L., Hu, J.G., Fan, B., Luo, Z.K., Zheng, Z.H., Luo, J.T., Fan, P., 2017. Facile preparation and enhanced photoelectrical performance of Sb2Se3 nano-rods by magnetron sputtering deposition. Sol. Energy Mater. Sol. Cells 160, 257–262. Liang, G.X., Zheng, Z.H., Fan, P., Luo, J.T., Hu, J.G., Zhang, X.H., Ma, H.L., Fan, B., Luo, Z.K., Zhang, D.P., 2018. Thermally induced structural evolution and performance of Sb2Se3 films and nanorods prepared by an easy sputtering method. Sol. Energy Mater. Sol. Cells 174, 263–270. Liu, X., Chen, J., Luo, M., Leng, M., Xia, Z., Zhou, Y., Qin, S., Xue, D.J., Lv, L., Huang, H., Niu, D., Tang, J., 2014. Thermal evaporation and characterization of Sb2Se3 thin film for substrate Sb2Se3/CdS solar cells. ACS Appl. Mater. Interf. 6 (13), 10687–10695. Salomé, P.M.P., Fernandes, P.A., Leitão, J.P., Sousa, M.G., Teixeira, J.P., Cunha, A.F.D., 2014. Secondary crystalline phases identification in Cu2ZnSnSe4 thin films: contributions from Raman scattering and photoluminescence. J. Mater. Sci. 49 (21), 7425–7436. Savory, Christopher N., Scanlon, D.O., 2019. The complex defect chemistry of antimony selenide. J. Mater. Chem. A 7 (17), 10739–10744. Shongalova, A., Correia, M.R., Teixeira, J.P., Leitão, J.P., González, J.C., Ranjbar, S., Garud, S., Vermang, B., Cunha, J.M.V., Salomé, P.M.P., Fernandes, P.A., 2018. Growth of Sb2Se3 thin films by selenization of RF sputtered binary precursors. Sol. Energy Mater. Sol. Cells 187, 219–226. Tao, J., Hu, X., Xue, J., Wang, Y., Weng, G., Chen, S., Zhu, Z., Chu, J., 2019. Investigation of electronic transport mechanisms in Sb2Se3 thin-film solar cells. Sol. Energy Mater. Sol. Cells 197, 1–6. Wang, X., Guo, H., Chen, Z., Ma, C., Fang, X., Jia, X., Yuan, N., Ding, J., 2019. Enhancement of Sb2Se3 thin-film solar cell photoelectric properties by addition of interlayer CeO2. Sol. Energy 188, 218–223. Wen, X., Chen, C., Lu, S., Li, K., Kondrotas, R., Zhao, Y., Chen, W., Gao, L., Wang, C., Zhang, J., Niu, G., Tang, J., 2018. Vapor transport deposition of antimony selenide thin film solar cells with 7.6% efficiency. Nat Commun 9 (1), 2179. Zeng, K., Xue, D.-J., Tang, J., 2016. Antimony selenide thin-film solar cells. Semicond. Sci. Technol. 31 (6). Zhou, Y., Leng, M.Y., Xia, Z., Zhong, J., Song, H.B., Liu, X.S., Yang, B., Zhang, J.P., Chen, J., Zhou, K.H., Han, J.B., Cheng, Y.B., Tang, J., 2014. Solution-processed antimony selenide heterojunction solar cells. Adv. Energy Mater. 4 (8). Zhou, Y., Wang, L., Chen, S.Y., Qin, S.K., Liu, X.S., Chen, J., Xue, D.J., Luo, M., Cao, Y.Z., Cheng, Y.B., Sargent, E.H., Tang, J., 2015. Thin-film Sb2Se3 photovoltaics with oriented one-dimensional ribbons and benign grain boundaries. Nat. Photon. 9 (6), 409.
This work was supported by the National Natural Science Foundation of china (Grant No. 91648109), the Key Research and Development Project of Jiangsu Province (BE2017006-3), the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (18KJD480002), Changzhou science and technology project (CJ20180008) and the Jiangsu Province Cultivation base for State Key Laboratory of Photovoltaic Science and Technology (SKLPST201705). Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.solener.2019.09.046. References Chen, C., Bobela, D.C., Yang, Y., Lu, S., Zeng, K., Ge, C., Yang, B., Gao, L., Zhao, Y., Beard, M.C., 2017a. Characterization of basic physical properties of Sb2Se3 and its relevance for photovoltaics. Front. Optoelectron. 10 (1), 18–30. Chen, Z., Guo, H., Ma, C., Wang, X., Jia, X., Yuan, N., Ding, J., 2019b. Efficiency improvement of Sb2Se3 solar cells based on La-doped SnO2 buffer layer. Sol. Energy 187, 404–410. Chen, S., Hu, X., Tao, J., Xue, J., Weng, G., Jiang, J., Shen, X., Chen, S., 2019a. Effects of substrate temperature on material and photovoltaic properties of magnetron-sputtered Sb2Se3 thin films. Appl. Opt. 58 (11), 2823–2827. Chen, C., Li, W., Zhou, Y., Chen, C., Luo, M., Liu, X., Zeng, K., Yang, B., Zhang, C., Han, J., 2015. Optical properties of amorphous and polycrystalline Sb2Se3 thin films prepared by thermal evaporation. Appl. Phys. Lett. 107 (4), 1301846. Chen, C., Wang, L., Gao, L., Nam, D., Li, D., Li, K., Zhao, Y., Ge, C., Cheong, H., Liu, H., Song, H., Tang, J., 2017b. 6.5% Certified efficiency Sb2Se3 solar cells using PbS colloidal quantum dot film as hole-transporting layer. ACS Energy Lett. 2 (9), 2125–2132. Guo, H., Chen, Z., Wang, X., Cang, Q., Jia, X., Ma, C., Yuan, N., Ding, J., 2019. Enhancement in the efficiency of Sb2Se3 thin-film solar cells by increasing carrier concertation and inducing columnar growth of the grains. Solar RRL 3 (3), 1800224. Hu, X., Tao, J., Weng, G., Jiang, J., Chen, S., Zhu, Z., Chu, J., 2018b. Investigation of electrically-active defects in Sb2Se3 thin-film solar cells with up to 5.91% efficiency via admittance spectroscopy. Sol. Energy Mater. Sol. Cells 186, 324–329. Hu, X., Tao, J., Chen, S., Xue, J., Weng, G., Kaijiang, Hu, Z., Jiang, J., Chen, S., Zhu, Z., Chu, J., 2018a. Improving the efficiency of Sb2Se3 thin-film solar cells by post annealing treatment in vacuum condition. Sol. Energy Mater. Sol. Cells 187, 170–175. Jiang, C., Tang, R., Wang, X., Ju, H., Chen, G., Chen, T., 2019. Alkali metals doping for high-performance planar heterojunction Sb2S3 solar cells. Solar RRL 3 (1). Jin, R., Liu, Z., Yang, L., Liu, J., Yanbin, X.U., Guihua, L.I., 2013. Facile synthesis of sulfur doped Sb2Se3 nanosheets with enhanced electrochemical performance. J. Alloy. Compd. 579 (10), 209–217.
282
Contents lists available at ScienceDirect
Solar Energy journal homepage: www.elsevier.com/locate/solener
Fabrication of Sb2Se3 thin film solar cells by co-sputtering of Sb2Se3 and Se targets
T
Changhao Maa,1, Huafei Guoa,1, Xin Wanga, Zhiwen Chena, Qingfei Canga, Xuguang Jiaa, Yan Lic, ⁎ ⁎ Ningyi Yuana, , Jianning Dinga,b, a
School of Materials Science and Engineering, Jiangsu Collaborative Innovation Center for Photovoltaic Science and Engineering, Jiangsu Province Cultivation Base for State Key Laboratory of Photovoltaic Science and Technology, Changzhou University, Changzhou 213164, China b Institute of Intelligent Flexible Mechatronics, Jiangsu University, Zhenjiang 212013, China c Changzhou Institute of Industry Technology, Changzhou, Jiangsu Province 213164, China
ARTICLE INFO
ABSTRACT
Keywords: Sb2Se3 Co-sputtering Thin film Selenium supplement
Antimony selenide (Sb2Se3) has attracted scientific interest due to its many advantages such as its appropriate band gap, non-toxicity, element abundance, and high absorption coefficient (> 105 cm−1). However, selenium (Se) is easily lost at high vapor pressures, thus leading to Se vacancies and increased recombination centers in films. In this work, Sb2Se3 films with good crystallinity were fabricated by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. The crystal structural properties, surface morphology, and optical and electrical properties of films with different Se contents were explored. Finally, we fabricated solar cells with a structure of FTO/CdS/Sb2Se3/Au and achieved an efficiency up to 3.47%.
1. Introduction
7.6% was achieved by VTD method (Wen et al., 2018). Chen et al. prepared Sb2Se3 solar cells based on La-doped SnO2 and achieved an efficiency of 3.25% (Chen et al., 2019b). Wang et al. added CeO2 interlayer between CdS and Sb2Se3 and improved the efficiency to 5.14% (Wang et al., 2019). Tao et al. investigated the electronic transport mechanisms in Sb2Se3 thin-film solar cells and fabricated the solar cells with the efficiency of 6.24% (Tao et al., 2019). And they investigated the electrically-active defects in Sb2Se3 thin-film solar cells and achieved an efficiency of 5.91% (Hu et al., 2018b). Evaporation and sputtering are also alternative methods to vacuum preparation of Sb2Se3 thin films. Liu deposited Sb2Se3 films by thermal evaporation using an electron beam, and the fabricated photovoltaic device with a configuration of FTO/Sb2Se3/CdS/ZnO/ZnO:Al/Au achieved an efficiency of 2.1% (Liu et al., 2014). Chen et al. studied the effects of substrate temperature on material and photovoltaic properties of magnetron-sputtered Sb2Se3 thin films but the efficiency is just only 0.84% (Chen et al., 2019a). Many studies have shown that, during evaporation, Sb2Se3 can decompose into Sb, Se, and SbSe, which leads to Se loss. The resulting Se vacancies in the films increase the recombination density (Li et al., 2016), which has a large negative impact on the films and solar cells. Thus, many studies have focused on the selenization of Sb2Se3 thin films to fill the Se vacancies. For example, Li
Recently, Sb2Se3 has attracted the interest of researchers due to its many advantages. Unlike CdTe, Sb2Se3 is a non-toxic and earth-abundant material, and its appropriate band gap (~1.1 eV) (Chen et al., 2015) and high absorption coefficient (> 105 cm−1) (Zhou et al., 2014) make it promising in terms of photovoltaic materials. Furthermore, Sb2Se3 has a one-dimensional crystal structure. Ribbons of (Sb4Se6)n stack along the [0 0 1] direction through strong covalent Sb-Se bonds, whereas in the [1 0 0] and [0 1 0] directions, the (Sb4Se6)n ribbons are held together by van der Waals forces (Zhou et al., 2015). This crystal structure improves charge transport and reduces the occurrence of recombination (Chen et al., 2017a). This is a promising material for commercialization. Magnetron sputtering is a simple method to fabricate Sb2Se3 thin films which also has a good uniformity and a promising for the Sb2Se3 large-scale industrial production. Originally, Zhou et al. prepared Sb2Se3 thin films by the RTE method, then fabricated solar cells with a structure of FTO/CdS/ Sb2Se3/Au and achieved an efficiency of 5.6% (Zhou et al., 2015). Lee et al. investigated the effects of low temperature pre-annealing of the absorbers fabricated by thermal evaporation and achieved a maximum efficiency of 1.47% (Lee et al., 2018). Subsequently, an efficiency of Corresponding authors. E-mail addresses: [email protected] (N. Yuan), [email protected] (J. Ding). 1 Changhao Ma and Huafei Guo contributed equally to this work. ⁎
https://doi.org/10.1016/j.solener.2019.09.046 Received 23 July 2019; Received in revised form 11 September 2019; Accepted 12 September 2019 0038-092X/ © 2019 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
Solar Energy 193 (2019) 275–282
C. Ma, et al.
et al. prepared an Sb2Se3 thin film by the co-evaporation of Sb2Se3 and Se (Li et al., 2016), and Shongalova presented a sputtering method to prepare Sb2Se3 films followed by selenized annealing in a poisonous H2Se gas atmosphere (Shongalova et al., 2018). In this study, Sb2Se3 films with good crystallinity were fabricated by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. This method is simple and resolved the problem of excess vacancies caused by Se loss. We studied the effects of different Se contents on film properties and investigated film structure, morphology, and optical and electrical performance. The films with supplemental Se displayed a preferred orientation along the [2 2 1] direction, which provides easier carrier transport. Supplemental Se also improved the morphology and optical and electrical properties of the films. Solar cells with a structure of FTO/CdS/Sb2Se3/Au were fabricated and achieved efficiencies of up to 3.47%.
CdS with different Se sputtering powers. The diffraction peaks of the films correspond to orthorhombic Sb2Se3 by comparison with the standard PDF card (JCPDS#15-0861) and the diffraction peak marked with * at 26.5° was attributed to CdS. However, compared with the film prepared without Se co-sputtering, all the films with Se co-sputtering display a preferred orientation along the [2 2 1] direction. The films with a preferred orientation along [1 2 0] are formed by (Sb4Se6)n ribbons stacked horizontally parallel to the substrate, whereas the [2 2 1]oriented films are formed by vertically stacked oblique (Sb4Se6)n ribbons on the substrate. Carrier transport in the [2 2 1]-oriented grains is easier than in those oriented in the [1 2 0] direction, because the carriers in the [2 2 1] direction are transport between the covalently bonded (Sb4Se6)n ribbons, but in [1 2 0] direction carriers are transport between (Sb4Se6)n ribbons combined with van der Waals forces. Therefore, the preferred orientation of the film along [1 2 0] will significantly increase the series resistance and reduce the efficiency of the device (Zhou et al., 2015). The films with supplemental Se also exhibit more obvious peak in the [2 1 1] direction compare to the film without supplemental Se, which the grain is composed of titled (Sb4Se6)n ribbons to different angle vertically on the substrate. The narrow halfheight peak widths of the preferred orientation are 0.368°, 0.334°, 0.313° and 0.321° respectively to the film without supplemental Se and with the Se power is 5 W, 10 W and 15 W, which also show that the films prepared with Se co-sputtering with a preferred orientation along [2 2 1] have relatively good crystallinity compare with the film without supplemental Se. The film without supplemental Se exhibits a preferred orientation along [1 2 0], in addition to this, the orientation along [2 3 0] is also very obvious. In order to quantitate the difference of orientations of the diffraction peaks of Sb2Se3 thin films deposited on CdS with different power of Se, the texture coefficients (TC) of the diffraction peaks shown in Fig. 1(b) were calculated based on the following equation (1) (Li et al., 2017), where the I(h k l) was the observed peak intensity of (h k l) plane and I0(h k l) was the intensity value in the standard XRD pattern, N was the total number of reflections considered for the calculation.
2. Experimental 2.1. Fabrication of Sb2Se3 films Sb2Se3 films were deposited by RF magnetron co-sputtering with Sb2Se3 and Se targets (99.99%, Hefei Kejing Materials Technology, China). The background pressure of the chamber was 4.0 × 10−6 Torr, and the substrate temperature was 350 °C. Then, Ar was introduced at a rate of 25 sccm, and the sputtering pressure was maintained at 5.0 × 10−3 Torr. The Se content was controlled by adjusting the sputtering power of the Se target. The power of the Sb2Se3 target was 30 W, and the power of the Se target was set to 0 W, 5 W, 10 W, or 15 W. 2.2. Fabrication of Sb2Se3 solar cells The structure of the solar cells was FTO/CdS/Sb2Se3/Au. The 60 nm CdS films were deposited on FTO substrate (with sheet resistance of about 6 Ω sq−1, transmittance about 80%) by the chemical bath deposition method described in our previous work (Guo et al., 2019). Then, Sb2Se3 films approximately 550 nm in thickness were deposited by the sputtering procedure described above. Gold electrodes (0.025 cm2) with a thickness of 80 nm were prepared by evaporation.
TCh
k l=
I(h I0(h
k l) k l)
/
1 N
N
I(h I 0(h
k l) k l)
(1)
The film without supplemental Se exhibits [h k 0] preferred orientation such as [1 2 0], [2 3 0] and [2 4 0]. However, the films with Se co-sputtering exhibit [h k 1] preferred orientation such as [2 2 1] and [2 1 1]. In order to prove this change is not caused by the change of sputtering parameters, we compared the XRD patterns of the Sb2Se3 films deposited on CdS at different substrate temperature with different power of Se in Fig. 1(c). The films deposited at 300 °C and 350 °C without supplemental Se both exhibit a preferred orientation along [1 2 0]. However, the films with Se co-sputtering exhibit an obvious preferred orientation along [2 2 1]. When the substrate temperature is 250 °C, the films doesn’t exhibit a good crystallinity. Fig. 1(d) shows the Raman spectra of the Sb2Se3 films with different Se contents, which is another method to detect film composition. The obtained results were compared with other reported results (Liang et al., 2018; Tao et al., 2019). The peaks at 189 cm−1, and 210 cm−1 were assigned to the SbSe vibration. When the Se target power was increased to 15 W, a peak close to 256 cm−1 became apparent, indicating a relatively high Se content in this film (Salomé et al., 2014). Compared with the film without supplemental Se, the intensity of the films with Se co-sputtering increase slightly and the half-height peak widths decreased indicating that the improvement of crystalline. XPS was used to detect the element oxidation states of the Sb2Se3 films deposited on CdS with and without supplemental Se. The highresolution core level Sb3d and Se3d spectra are shown in Fig. 1(e) and (f). The C1s energy at 284.6 eV was used to correct the binding energies (BEs) in addition to the charge compensation by the flood gun associated with the spectrometer. Compared with the sample without
2.3. Characterization of films and solar cells The structures and crystallinities of the Sb2Se3 films deposited on CdS were detected by X-ray diffraction (XRD, Smart Apex II Duo), and Raman scattering analyses were performed using a micro-Raman spectrometer (Thermo Fisher) with a 532-nm excitation wavelength. Scanning electron microscopy (SEM, Hitachi) was used to observe the surface morphologies and cross-sections of different films, and the film element contents were determined by energy dispersive X-ray spectroscopy. Surface roughness was measured by atomic force microscope (AFM, Nanman), and optical performance was measured by UV–visible spectrophotometer (Shimadzu UV-2600). X-ray photoelectron spectroscopy (XPS) was performed on a photoelectron spectrometer (Kratos Axis Ultra DLD). Ultraviolet photoelectron spectroscopy (UPS) was measured with a monochromatic He I light source (21.2 eV) and a VG Scienta R4000 analyzer. The efficiency of the solar cells was measured by I-V tester (Cell tester) under AM1.5 global spectrum conditions with a radiance of 1000 W m−2. The external quantum efficiency (EQE) was detected by a photovoltaic characterization system (PV Measurements QEXL), and C-V was measured by a Keysight Precision LCR meter (Keysight E4980AL) in darkness at room temperature and 10 kHz, where the DC bias voltage was changed from −1 V to 0.5 V. 3. Results and discussion Fig. 1(a) shows the XRD patterns of the Sb2Se3 films deposited on 276
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 1. (a) The XRD pattern (b) The texture coefficients of the diffraction peaks of the Sb2Se3 thin films deposited on CdS with different power of Se. (c) The XRD pattern of Sb2Se3 thin films deposited on CdS with different power of Se under different substrate temperature. (d) The Raman spectra of the Sb2Se3 films with different Se content and the high-resolution core level spectra of (e) Sb3d and (f) Se3d for the films with 10 W Se and without supplemental Se.
supplemental Se, the films with Se co-sputtering have relatively highintensity doublets. The Sb3d5/2 and Sb3d3/2 spectra of the Se-supplemented films have peaks at 529.4 eV and 538.7 eV with a separation of 9.3 eV attributed to the charge state of Sb3+, which is similar to the results reported by others (Liang et al., 2017). The Se3d5/2 and Se3d3/2 peaks of the films are located at 53.8 eV and 54.6 eV with the separation of 0.8 eV, which are identical to that of metallic Se (Jin et al., 2013). Fig. 2(a–d) illustrates a significant difference between the surface morphologies of the Sb2Se3 films deposited on CdS with and without Se co-sputtering. The grains of the film without supplemental Se grows at different heights in different directions. This may be due to the loss of Se during sputtering and leading to the instability of the film during the sputtering process. However, the Se-supplemented films not only have good crystallinity but also the surface roughness is markedly reduced. (The wide range SEM images of the Sb2Se3 films deposited on CdS with different power of Se were shown in Fig. S1 in supplementary.) This result is supported by the AFM images of the different films shown in Fig. 2(e–h). The roughness of the film without Se was 61.5 nm, whereas Se target powers of 5 W, 10 W, and 15 W yielded Sb2Se3 film surface roughness of 33.5 nm, 30.5 nm, and 32.4 nm, respectively. This change
is due to the improvement of grain growth of the Sb2Se3 films compare with the film without supplemental Se. The films with smoother surface are more suitable for achieving better quality interface between the adjacent layers. This smooth surface was very helpful to reduce the recombination of light induced excitations at the interface of the absorber and metal electrode, which would be beneficial for the enhancement of cell performance (Lei et al., 2016). Although the growth of the grains was substantially improved, when Se was insufficiently supplemented, the produced film had relatively apparent holes, as shown in Fig. 2(b), which may have been caused by the loss of Se. On the other hand, as shown in Fig. 2(d), when the supplemental Se was excessive, more small grains were present in the film. Table 1 shows the chemical composition of the Sb2Se3 films fabricated with different Se target powers. The film without supplemental Se exhibited a Se loss compared with the normal proportion. The Se content increased with increasing power of Se. When the Se power was increased to 10 W, the resulting film had an excessive amount of Se. EDX elemental mapping images of the films without supplemental Se and with 10 W Se power are shown in Fig. 3. Previous work has shown that Sb-rich conditions lead to many defects with relatively low 277
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 2. The scanning electron microscope images of the Sb2Se3 films (a) without supplemental Se. (b) (c) and (d) the co-sputtering films with Se power of 5 W, 10 W and 15 W and the atomic force microscope images of the Sb2Se3 films (e) without supplemental Se. (f) (g) and (h) the co-sputtering films with Se power of 5 W, 10 W and 15 W.
formation energies such as Vse and SbSe. In contrast, under Se-rich conditions, more defects form with higher formation energies (Savory and Scanlon, 2019), these defects will be not easy to form. Therefore, more defects will occur when there is Se loss, and thus it is necessary to include supplemental Se by co-sputtering. The cross-sectional SEM images of solar cells without supplemental Se and with an Se co-sputtering power of 10 W are shown in Fig. 4(a) and (b). We can determine from the images that the Sb2Se3 film with
supplemental Se has better crystallinity and smaller roughness, whereas the film without supplemental Se exhibits disordered growth. Fig. 4(c) and (d) shows EDX mappings based on the cross-sections of the two samples. Compared with the Se-supplemented film in Fig. 4(d), the film without supplemental Se in Fig. 4(c) shows evident Se element loss, which may lead to Se vacancies in the film. In addition, from comparison of the Sb elemental mappings in Fig. 4(c) and (d), we suspect that Se loss may lead to the instability of Sb2Se3 and result in more 278
Solar Energy 193 (2019) 275–282
C. Ma, et al.
excessive elemental Se and the small grains as shown in Fig. 2(d) may result in a lower quality interface between Sb2Se3 and CdS and the degradation of the film quality, thus leading to the slightly decrease of the Jsc, Voc and FF. The diode ideality factor (A) and reverse saturation current density (J0) are shown in Table 2. Compare with the device without supplemental Se, the J0 and A of the devices with Se co-sputtering have a slight decline, showing that the reduction of the recombination loss and the improvement of the heterojunction in the devices. The EQE values of the devices are shown in Fig. 7(b), the spectral intensities of the cells with Se co-sputtering declined at shorter and longer wavelength ranges and peaked at 550 nm to 650 nm. The drop at short and long wavelength is mainly caused by the absorption by the CdS buffer layer and incomplete absorption of infrared photons and reduced collection efficiency of photo-generated carriers which are far from the depletion region. The devices with Se co-sputtering improved significantly. This improvement in performance may due to a reduced number of defects such as VSe, which can easily form during sputtering processes with large Se vapor pressures. Fig. 7(c) shows C−2V curves of the devices with different Se contents acquired in darkness at a frequency of 10 kHz and at room temperature. The built-in-voltages (Vbi) of the samples fabricated without Se, with 5 W, 10 W, and 15 W power of Se are 0.329 V, 0.483 V, 0.531 V and 0.510 V respectively. We calculated the depletion region width (Wd) are 178 nm, 184 nm, 190 nm and 187 nm respectively for the devices with different content of Se. The increasement of EQE at the long wave range can be caused by the enlargement of the depletion region width and thus leading to the more efficient charge collection through drifting (Chen et al., 2017b). Comparing with the sample without supplemental Se, the Vbi increased from 0.329 to 0.531 V. The samples with larger Vbi indicate a better contact and heterojunction quality at CdS/Sb2Se3 (Zeng et al., 2016) and the increase of Vbi can decrease the charge accumulation at electrode interfaces under illumination (Jiang et al., 2019). As Vbi is positively correlated to VOC, improved VOC has been observed in solar cells with Se co-sputtering. However, the increase of Voc is much smaller than the increase of Vbi, which could be caused by the Cd diffusion during sputtering (Hu et al., 2018a). Fig. 7(d) shows the statistical efficiency distribution charts of 16 solar cells with different Se contents. The PCEs of the devices without Se were within the range of approximately 0.60–0.89%, the 5 W Se power devices were within approximately 1.5–2.0%, and those of the 15 W Se power devices were within approximately 2.3–2.9%. In contrast, the PCEs of the devices with 10 W of Se power were within approximately 3.0–3.47%.
Table 1 The chemical composition of Sb2Se3 films of the normal proportion and with different power of Se. Sample Name
Sb
Se
Sb/Se
Normal proportion Without Se With 5 W Se With 10 W Se With 15 W Se
40 41.62 39.38 38.08 35.51
60 58.38 60.62 61.92 64.49
0.67 0.71 0.65 0.61 0.55
serious Sb diffusion. Fig. 5(a) shows the absorption spectra of the Sb2Se3 thin films deposited on CdS films and removed the CdS base background. Se cosputtering yielded a significant improvement in absorbance compared with the film without supplemental Se, and a sharp absorption edge was observed. The (αhυ)2-hυ plot of the Sb2Se3 thin films is shown in Fig. 5(b), illustrating that the band gap values of the films with supplemental Se are arranged from 1.26 to 1.28 eV. For the film without Se co-sputtering, the band gap value decreased slightly to 1.23 eV. This change could be caused by the crystallinity of the films. Fig. 6(a) and (b) show the UPS of the Sb2Se3 with different Se content. The energy level diagram pattern is shown in Fig. 6(c). The distance of the films without supplemental Se and with 10 W Se between the valence band maximum (VBM) and the Fermi energy were calculated to be 0.60 eV and 0.52 eV respectively. The VBM position of the film with Se cosputtering was closer to the Fermi energy. This indicates an increasement in hole concentration of the film with Se co-sputtering. Furthermore, the conduction band minimum (CBM) position of the co-sputtering film moved to the positive direction compare with the film without co-sputtering. So, the energy offset between the CBM of Sb2Se3 and the CBM of CdS is larger, suggesting that the photogenerated electron injection from the Sb2Se3 to CdS would be faster. The recombination between CdS and Sb2Se3 would be suppressed. Fig. 7(a) shows the J-V performances of the solar cells with different Se co-sputtering powers, and Table 2 lists the maximum efficiencies of the devices. With the appropriate increase in Se content the short-circuit current density (Jsc), open-circuit voltage (Voc) and fill factor (FF) improved significantly. We achieved an efficiency of up to 3.47% with Jsc, Voc, and FF of 19.2 mA/cm2, 0.350 V, and 51.6% respectively, when the Se power was 10 W. The decline of series resistance (Rs) and the increase of shunt resistance (Rsh) are obvious. The increment of Voc and Jsc with Se incorporation could be caused by the increasement of collection efficiency of photo-generated carriers and the improvement of CdS/Sb2Se3 interface. When the Se target power increased to 15 W, the
Fig. 3. EDX elemental mapping images of the Sb2Se3 films (a) without supplemental Se (b) with 10 W Se. 279
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 4. Cross-sectional SEM images of the solar cells (a) without supplemental Se (b) with the co-sputtering power of Se is 10 W and the EDX mapping based on the cross section of FTO/CdS/Sb2Se3 (c) without supplemental Se (d) with the co-sputtering power of Se is 10 W.
Fig. 5. (a) The absorption spectrum and (b) The (αhυ)2-hυ plot of the Sb2Se3 thin films with different Se co-sputtering power.
4. Conclusion
Se, the films prepared by co-sputtering exhibited better crystallinities and a preferred orientation along the [2 2 1] direction, which provides easier carrier transport. Supplemental Se decreased the number of Se vacancies in the films and improved their optical and electrical properties. We fabricated solar cells with a structure of FTO/CdS/Sb2Se3/Au and achieved an efficiency as high as 3.47%.
Sb2Se3 films were prepared by co-sputtering of Sb2Se3 and Se targets with a substrate temperature of 350 °C and without a subsequent annealing process. The crystal structural properties, surface morphologies, and optical and electrical properties of the films with different Se contents were explored. Compared with the film without supplemental 280
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Fig. 6. (a) and (b) are the UPS spectra of Sb2Se3 films with different Se deposited on CdS film. (c) is the Schematic of the energy alignment.
Fig. 7. (a) J-V curves (b) EQE (c) C−2-V plots and (d) The statistical efficiency distribution chart of the solar cells with different content of Se.
281
Solar Energy 193 (2019) 275–282
C. Ma, et al.
Table 2 Device characteristics of the solar cells with different content of Se. Sample Name
Voc (V)
Jsc (mA/cm2)
Fill Factor (%)
Efficiency (%)
Rs (ohm)
Rsh (ohm)
A
J0 (mA/cm2)
Without Se With 5 W Se With 10 W Se With 15 W Se
0.251 0.311 0.350 0.321
9.1 15.9 19.2 18.1
39.1 41.4 51.6 51.1
0.90 2.04 3.47 2.97
459.7 366.2 222.2 202.2
2224.6 2363.9 3950.0 4048.5
1.68 1.65 1.52 1.56
3.71E-04 2.75E-04 1.12E-04 1.67E-04
Acknowledgement
Lee, D., Cho, J.Y., Heo, J., 2018. Improved efficiency of Sb2Se3/CdS thin-film solar cells: the effect of low-temperature pre-annealing of the absorbers. Sol. Energy 173, 1073–1079. Lei, H.W., Yang, G., Guo, Y.X., Xiong, L.B., Qin, P.L., Dai, X., Zheng, X.L., Ke, W.J., Tao, H., Chen, Z., Li, B.R., Fang, G.J., 2016. Efficient planar Sb2S3 solar cells using a lowtemperature solution-processed tin oxide electron conductor. Phys. Chem. Chem. Phys. 18 (24), 16436–16443. Li, Z., Zhu, H., Guo, Y., Niu, X., Chen, X., Zhang, C., Zhang, W., Liang, X., Zhou, D., Chen, J., 2016. Efficiency enhancement of Sb2Se3 thin-film solar cells by the co-evaporation of Se and Sb2Se3. Appl. Phys Express 9 (5), 052302. Li, Z., Chen, X., Zhu, H., Chen, J., Guo, Y., Zhang, C., Zhang, W., Niu, X., Mai, Y., 2017. Sb2Se3 thin film solar cells in substrate configuration and the back contact selenization. Sol. Energy Mater. Sol. Cells 161, 190–196. Liang, G.X., Zhang, X.H., Ma, H.L., Hu, J.G., Fan, B., Luo, Z.K., Zheng, Z.H., Luo, J.T., Fan, P., 2017. Facile preparation and enhanced photoelectrical performance of Sb2Se3 nano-rods by magnetron sputtering deposition. Sol. Energy Mater. Sol. Cells 160, 257–262. Liang, G.X., Zheng, Z.H., Fan, P., Luo, J.T., Hu, J.G., Zhang, X.H., Ma, H.L., Fan, B., Luo, Z.K., Zhang, D.P., 2018. Thermally induced structural evolution and performance of Sb2Se3 films and nanorods prepared by an easy sputtering method. Sol. Energy Mater. Sol. Cells 174, 263–270. Liu, X., Chen, J., Luo, M., Leng, M., Xia, Z., Zhou, Y., Qin, S., Xue, D.J., Lv, L., Huang, H., Niu, D., Tang, J., 2014. Thermal evaporation and characterization of Sb2Se3 thin film for substrate Sb2Se3/CdS solar cells. ACS Appl. Mater. Interf. 6 (13), 10687–10695. Salomé, P.M.P., Fernandes, P.A., Leitão, J.P., Sousa, M.G., Teixeira, J.P., Cunha, A.F.D., 2014. Secondary crystalline phases identification in Cu2ZnSnSe4 thin films: contributions from Raman scattering and photoluminescence. J. Mater. Sci. 49 (21), 7425–7436. Savory, Christopher N., Scanlon, D.O., 2019. The complex defect chemistry of antimony selenide. J. Mater. Chem. A 7 (17), 10739–10744. Shongalova, A., Correia, M.R., Teixeira, J.P., Leitão, J.P., González, J.C., Ranjbar, S., Garud, S., Vermang, B., Cunha, J.M.V., Salomé, P.M.P., Fernandes, P.A., 2018. Growth of Sb2Se3 thin films by selenization of RF sputtered binary precursors. Sol. Energy Mater. Sol. Cells 187, 219–226. Tao, J., Hu, X., Xue, J., Wang, Y., Weng, G., Chen, S., Zhu, Z., Chu, J., 2019. Investigation of electronic transport mechanisms in Sb2Se3 thin-film solar cells. Sol. Energy Mater. Sol. Cells 197, 1–6. Wang, X., Guo, H., Chen, Z., Ma, C., Fang, X., Jia, X., Yuan, N., Ding, J., 2019. Enhancement of Sb2Se3 thin-film solar cell photoelectric properties by addition of interlayer CeO2. Sol. Energy 188, 218–223. Wen, X., Chen, C., Lu, S., Li, K., Kondrotas, R., Zhao, Y., Chen, W., Gao, L., Wang, C., Zhang, J., Niu, G., Tang, J., 2018. Vapor transport deposition of antimony selenide thin film solar cells with 7.6% efficiency. Nat Commun 9 (1), 2179. Zeng, K., Xue, D.-J., Tang, J., 2016. Antimony selenide thin-film solar cells. Semicond. Sci. Technol. 31 (6). Zhou, Y., Leng, M.Y., Xia, Z., Zhong, J., Song, H.B., Liu, X.S., Yang, B., Zhang, J.P., Chen, J., Zhou, K.H., Han, J.B., Cheng, Y.B., Tang, J., 2014. Solution-processed antimony selenide heterojunction solar cells. Adv. Energy Mater. 4 (8). Zhou, Y., Wang, L., Chen, S.Y., Qin, S.K., Liu, X.S., Chen, J., Xue, D.J., Luo, M., Cao, Y.Z., Cheng, Y.B., Sargent, E.H., Tang, J., 2015. Thin-film Sb2Se3 photovoltaics with oriented one-dimensional ribbons and benign grain boundaries. Nat. Photon. 9 (6), 409.
This work was supported by the National Natural Science Foundation of china (Grant No. 91648109), the Key Research and Development Project of Jiangsu Province (BE2017006-3), the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (18KJD480002), Changzhou science and technology project (CJ20180008) and the Jiangsu Province Cultivation base for State Key Laboratory of Photovoltaic Science and Technology (SKLPST201705). Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.solener.2019.09.046. References Chen, C., Bobela, D.C., Yang, Y., Lu, S., Zeng, K., Ge, C., Yang, B., Gao, L., Zhao, Y., Beard, M.C., 2017a. Characterization of basic physical properties of Sb2Se3 and its relevance for photovoltaics. Front. Optoelectron. 10 (1), 18–30. Chen, Z., Guo, H., Ma, C., Wang, X., Jia, X., Yuan, N., Ding, J., 2019b. Efficiency improvement of Sb2Se3 solar cells based on La-doped SnO2 buffer layer. Sol. Energy 187, 404–410. Chen, S., Hu, X., Tao, J., Xue, J., Weng, G., Jiang, J., Shen, X., Chen, S., 2019a. Effects of substrate temperature on material and photovoltaic properties of magnetron-sputtered Sb2Se3 thin films. Appl. Opt. 58 (11), 2823–2827. Chen, C., Li, W., Zhou, Y., Chen, C., Luo, M., Liu, X., Zeng, K., Yang, B., Zhang, C., Han, J., 2015. Optical properties of amorphous and polycrystalline Sb2Se3 thin films prepared by thermal evaporation. Appl. Phys. Lett. 107 (4), 1301846. Chen, C., Wang, L., Gao, L., Nam, D., Li, D., Li, K., Zhao, Y., Ge, C., Cheong, H., Liu, H., Song, H., Tang, J., 2017b. 6.5% Certified efficiency Sb2Se3 solar cells using PbS colloidal quantum dot film as hole-transporting layer. ACS Energy Lett. 2 (9), 2125–2132. Guo, H., Chen, Z., Wang, X., Cang, Q., Jia, X., Ma, C., Yuan, N., Ding, J., 2019. Enhancement in the efficiency of Sb2Se3 thin-film solar cells by increasing carrier concertation and inducing columnar growth of the grains. Solar RRL 3 (3), 1800224. Hu, X., Tao, J., Weng, G., Jiang, J., Chen, S., Zhu, Z., Chu, J., 2018b. Investigation of electrically-active defects in Sb2Se3 thin-film solar cells with up to 5.91% efficiency via admittance spectroscopy. Sol. Energy Mater. Sol. Cells 186, 324–329. Hu, X., Tao, J., Chen, S., Xue, J., Weng, G., Kaijiang, Hu, Z., Jiang, J., Chen, S., Zhu, Z., Chu, J., 2018a. Improving the efficiency of Sb2Se3 thin-film solar cells by post annealing treatment in vacuum condition. Sol. Energy Mater. Sol. Cells 187, 170–175. Jiang, C., Tang, R., Wang, X., Ju, H., Chen, G., Chen, T., 2019. Alkali metals doping for high-performance planar heterojunction Sb2S3 solar cells. Solar RRL 3 (1). Jin, R., Liu, Z., Yang, L., Liu, J., Yanbin, X.U., Guihua, L.I., 2013. Facile synthesis of sulfur doped Sb2Se3 nanosheets with enhanced electrochemical performance. J. Alloy. Compd. 579 (10), 209–217.
282