Solution-processed efficient CdTe nanocrystal/CBD-CdS hetero-junction solar cells with ZnO interlayer
CdTe nanocrystal (NC)/CdS p–n hetero-junction solar cells with an ITO/ZnO-In/CdS/CdTe/MoOx/Ag-inverted structure were prepared by using a layer-by-layer solution process. The CdS thin films were prepared by chemical bath deposition on top of ITO/ZnO-In and were found to be very compact and pin-hole free in a large area, which insured high quality CdTe NCs thin-film formation upon it. The device performance was strongly related to the CdCl2 annealing temperature and annealing time. Devices exhibited power conversion efficiency (PCE) of 3.08 % following 400 °C CdCl2 annealing for 5 min, which was a good efficiency for solution processed CdTe/CdS NC-inverted solar cells. By carefully designing and optimizing the CdCl2-annealing conditions (370 °C CdCl2 annealing for about 15 min), the PCE of such devices showed a 21 % increase, in comparison to 400 °C CdCl2-annealing conditions, and reached a better PCE of 3.73 % while keeping a relatively high VOC of 0.49 V. This PCE value, to the best of our knowledge, is the highest PCE reported for solution processed CdTe–CdS NC solar cells. Moreover, the inverted solar cell device was very stable when kept under ambient conditions, less than 4 % degradation was observed in PCE after 40 days storage.
KeywordsNanocrystals solar cells CdTe nanocrystals Solution processed Inverted structure Energy conversion
Thin-film solar cells have attracted intense attention in the past several decades due to their high energy conversion efficiency along with lower material consumption and faster deposition rates (Zweibel 1999). Among which, thin-film solar cells based on cadmium telluride (CdTe) and copper indium gallium selenide (CIGS) active layers had exhibited power conversion efficiencies of 16.7 % (Green et al. 2009) and 20.3 % (Jackson et al. 2011), respectively, along with commercial production costs below $1/Watt. However, these inorganic thin-film solar cells are mostly fabricated by vacuum evaporation, magnetron sputtering, or other vacuum-based deposition methods, which still tend to be times consuming and costly. One way to overcome this drawback and further lower the production cost of thin-film solar cells is to use roll-to-roll solution processing of colloidal semiconductor NCs or polymer to deposit the active layers in a device. To date, our research group has reported power conversion efficiency (PCE) of state-of-the art polymer solar cells approaching 10 % in the scientific literature (He et al. 2012). Comparing to polymer solar cells, semiconductor NC solar cells have attracted significant interest due to the benefits of solution processing combined with the potential to harvest the sun’s visible, near-infrared rays. Most important of all, the band gap of NCs can be readily tuned during the time of synthesis, simply by changing the NCs’ size. This merit makes the fabrication of multiple-junction solar cells by a single materials system consisting of different band gaps easier and lower cost (King 2008). NC solar cells, especially PbS or PbSe NC-based solar cells, have seen rapid advances in the past several years. Promising solar power conversion efficiencies up to 5 % (Tang et al. 2012) and ~7 % (Ip et al. 2012; Liu et al. 2011; Tang et al. 2011) have been reported in PbS NC solar cells based on PbS NC homogeneous p–n junction and PbS–TiO2-depleted hetero-junction devices using a post-deposition chemical and organic/inorganic hybrid passivation strategy. On the other hand, in the case of CdTe (Jasieniak et al. 2011), CuInS2 (Weil et al. 2010), and CuInSe2 (Guo et al. 2010) NCs, in order to achieve high efficiency devices, sintering strategies are often used to obtain excellent electronic properties (high mobility, low lattice mismatch, and low defects) of bulk inorganic semiconductors. Gur et al. (2005) first reported all-inorganic CdTe/CdSe hetero-junction NC solar cells with PCEs as high as 2.9 % by employing a solution processing method and thermal treatment steps. More recently, Olson et al. (2010) and our research group (Sun et al. 2012) adapted this approach to fabricate Schottky solar cells based on CdTe nanorods with efficiencies up to 5 %. To decrease the stress developed during the thermal treatment step within a film of NCs, Mulvaney et al. fabricated CdTe–ZnO (Jasieniak et al. 2011) or CdSexTe1−x–ZnO (MacDonald et al. 2012) NC-based solar cells with PCE ~7.0 %, which was believed to be the highest value ever reported for CdTe thin-film solar cells based on solution-processed NC layers. However, CdTe NC devices with a p–n junction or Schottky structure reported to date are seen to be disadvantaged due to a number of factors. For example, as light absorption begins at the ITO ohmic contact side, rather than the junction, many minority carriers (here electrons) must travel the thickness of the entire film before reaching their destination electrode and are therefore more often subjected to unwanted recombination. On the other hand, the open-circuit voltage in a Schottky device is often limited by Fermi-level pinning due to defects at an interface. Investigations have shown that highly efficient polycrystalline CdTe thin-film solar cells are generally fabricated with p–n device configurations for effectively collecting photo-carriers from devices in which the diffusion coefficients are small (Chu et al. 1991). However, n-CdTe layers are still difficult to produce due the self compensation effect during the doping process, so wide-band-gap n-type semiconductors are often used to construct CdTe hetero-junction devices. As CdS has a band gap of 2.42 eV and similar chemical properties as CdTe, it is the most commonly employed n-type semiconductor partner to p-CdTe. In this hetero-junction system, a large fraction of the photo-generated carriers are generated within the depletion layer allowing more efficient collection.
Although high efficiency had been obtained in CdTe/CdS bulk hetero-junction thin-films solar cells fabricated by vacuum deposition techniques, there are still few reports on the solution processed CdTe NC/CdS thin-films solar cells with satisfy PCE (around or up to 3 %) and stability requirements (device could work under ambient conditions for several days/weeks without obvious PCE drop down). Sahoo et al. (Katiyar et al. 2011) reported the solution-processed CdTe–CdS nanocrystal solar cells with ITO/CdS/CdTe structure and very poor PCE was obtained in this case. A key reason for this is the poor interface between the NC and CdS thin film, partially due to a high number of defects between CBD-CdS and CdTe NC layers formed during the sintering process, which will then result in large leakage current to any devices that are subsequently fabricated. To decrease the defects and optimize interfacial properties, we report herein the realization of good efficiency and stability of all-solution processed NC p–n junction solar cells based on CdTe NC active layers and CBD-CdS window layers using ZnO (TiO2 is also a promising candidate) as interlayer. In addition to acting as an electron transport layer, the ZnO layer provides a number of improvements over samples with a CdS layer deposited directly on the ITO; it provides better electrical stability by forming a smooth and pin-hole-free pre-layer on which the smooth CBD-CdS films can be grown, thus eliminating catastrophic shorts from the upper contact directly through CdTe NC layers to the ITO; it improves the Fermi level and blocks the hole transfer to the ITO electrode, thus allowing more efficient photo-generated carrier extraction and transmission, resulting in higher photocurrents and high efficiency. In this paper, CdTe NC–CdS hetero-junction solar cells with an ITO/ZnO-In/CdS/CdTe/MoOx/Ag-inverted structure are fabricated successfully using a layer-by-layer solution process. For comparing, device with ITO/ZnO-In/CdTe/MoOx/Ag and ITO/CdS/CdTe/MoOx/Ag structures were also fabricated. The devices performance versus annealing temperature and annealing time is investigated and will be discussed. It is found that a further CdCl2 annealing under suitable conditions can largely improve device performance, which may be due to the grain size increasing or defect density decreasing in the junction. Under optimized annealing conditions, a PCE as high as 3.73 % coupled with high stability is obtained, which is the highest PCE reported for solution-processed CdTe NC–CdS thin-film solar cells.
CdTe NCs were synthesized according to a previously published method (Sun et al. 2012) and other research group’s report (Peng and Peng 2001; Yu et al. 2003; Nie et al. 2006). The NC product was extracted from the solution by washing three times with methanol/toluene and separated by centrifugation. Then, NC samples were refluxed in pyridine overnight at 110 °C and centrifuged with hexane. The final NC product was then dispersed into a mixture of pyridine and 1-propanol with a volume ratio of 1:1 at 50 mg mL-1.
Indium-doped ZnO thin films of ~60 nm in thickness were deposited on the ITO/glass substrate by magnetron sputtering, as reported before (Lan et al. 2011). As low hole-carrier density and low mobility (Jasieniak et al. 2011) of CdTe NCs, the used of indium-doped ZnO (high electron density and high mobility) is preferred in order to make full depletion of the CdTe NCs/CdS film and increased the carrier-collecting efficiency. The ITO/ZnO-In was then subjected sonication in acetone and isopropyl alcohol in that order and dried in a hotplate to remove any remaining solvent and other impurities. Before the deposition of CdS thin film, the ITO/ZnO-In substrate was dipped into de-ionized water at 60 °C for 1 min in order to eliminate bubble on the surface of ZnO-In film, which will insure pin-hole free of CdS and good adhesion of the CdS film on ZnO. The CdS thin film was formed on the ITO/ZnO-In substrate by a simple CBD method in an aqueous solution-containing cadmium acetate, ammonium acetate, ammonium hydroxide, and thiourea at concentrations of 5 × 10−4, 1 × 10−2, 1 × 10−3, and 0.1 M, respectively, with bath temperature maintained at 90 °C for 30 min (Britt and Ferekides 1993). The ITO/ZnO-In/CdS products were subjected to sonication in deionized water twice and then dried in an oven before use.
CdTe NCs films were deposited using a layer-by-layer spin-coating process under ambient conditions. For each layer, the CdTe NC ink (50 mg mL−1 in pyridine and 1-propanol) was deposited on the ITO/ZnO-In/CdS substrate and spin-cast at 1,000 rpm for 30 s. The substrate was placed on a hot plate at 150 °C for 3 min to remove any solvent and then dipped in a saturated CdCl2 methanol solution for ~3 s, taken out, and rinsed with 1-PrOH then dried under a nitrogen stream. Finally, the sample was sintered on a hot plate at 350 °C for 40 s. This process was repeated several times until a final CdTe NC thin film with thickness of ~500 nm was obtained. To activate the CdTe–CdS junction, a further CdCl2 treatment was carried out. Several drops of saturated CdCl2 methanol solution were deposited on top of the CdTe NC thin film and then spin-cast at 1,000 rpm for 10 s. The substrate was immediately placed on a hot plate at 360–400 °C for different time periods (0–20 min). The ITO/ZnO-In/CdS/CdTe products were subjected to sonication in methanol twice more to remove any remaining CdCl2 and blown dry under a nitrogen stream. The MoOx (~30 nm) and silver (~100 nm) back contact were deposited in sequence via thermal evaporation through a shadow mask. The active area of the solar cells was 0.16 cm2.
PCE of CdTe NC solar cells was measured under an illumination of 1,000 W m−2 with an AM1.5 solar simulator (Oriel model 91192) while the current density–voltage (J–V) curves were measured with a Keithley 240 source measure unit. The external quantum efficiencies (EQE) of the inverted PVCs were measured with a commercial photomodulation spectroscopic setup, and a calibrated Si photodiode was used as a standard.
Photovoltaic performances of solar cells fabricated under different conditions (under irradiation of AM1.5G at 100 mW/cm2)
Annealing temperature (°C)
Annealing time (min)
Summarize devices performance at different light intensity (under irradiation of AM1.5G)
Number of Sun
PCE (one sun) (%)
In conclusion, sputtered ZnO thin film was used as a buffer layer to prepare pin-hole-free CdS thin film and CdTe NC/CdS p–n hetero-junction solar cells with a ITO/ZnO-In/CdS/CdTe/MoOx/Ag-inverted structure were demonstrated. Annealing temperature and annealing time were found to have great effect on the performance of solar cell devices. A PCE of 3.73 % is obtained in the case of CdCl2 treatment at 370 °C, which is the best result ever reported for solution-processed CdTe NC/CdS hetero-junction solar cells. The JSC of devices showed linear behavior upon exposure to light intensity in the range of 0–2 Sun, which indicates no build-up of net space charges even at very high illumination intensity. We also found that solar cell devices have very good stability. When the device was kept under ambient conditions, there was no obvious degradation observed in the Jsc, Voc, and FF, resulting in less than 4 % dropdown in PCE after 40 days. Our results confirm that, if junctions of CdS and CdTe undergo better optimized processing and treatment, the device performance can be definitely improved.
We gratefully acknowledge the financial support of the National Natural Science Foundation of China (Nos. 51073056, 50990065, 51010003, 61274062, and 11204106), National Science Foundation for Distinguished Young Scholars of China (Grant No. 51225301) and SCUT Grant (No. 2013ZZ0016).
- Green MA, Emery K, Hishikawa Y, Warta W (2009) Solar cell efficiency tables (version 33). Prog Photovolt 17:74–85Google Scholar
- He Z, Zhong C, Su S, Xu M, Wu H, Cao Y (2012) Enhanced power-conversion efficiency in polymer solar cells using an inverted device structure. Nat Photonics 6:591–595Google Scholar
- Liu H, Tang J, Kramer I, Debnath JR, Koleilat GI, Wang X, Fisher A, Li R, Brzozowski L, Levina L, Sargent EH (2011) Electron acceptor materials engineering in colloidal quantum dot solar cells. Adv Mater 23:3832–3837Google Scholar