COMMUNICATION © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim (1 of 6) 1501070 wileyonlinelibrary.com Large Efficiency Improvement in Cu 2 ZnSnSe 4 Solar Cells by Introducing a Superficial Ge Nanolayer Sergio Giraldo, Markus Neuschitzer, Thomas Thersleff, Simón López-Marino, Yudania Sánchez, Haibing Xie, Mónica Colina, Marcel Placidi, Paul Pistor, Victor Izquierdo-Roca, Klaus Leifer, Alejandro Pérez-Rodríguez, and Edgardo Saucedo* S. Giraldo, M. Neuschitzer, S. López-Marino, Y. Sánchez, H. Xie, Dr. M. Colina, Dr. M. Placidi, Dr. P. Pistor, Dr. V. Izquierdo-Roca, Prof. A. Pérez-Rodríguez, Dr. E. Saucedo Catalonia Institute for Energy Research, (IREC) Jardins de les Dones de Negre 1 08930, Sant Adrià de Besòs, Barcelona, Spain E-mail: esaucedo@irec.cat Dr. T. Thersleff, Prof. K. Leifer The Ångström Laboratory Department of Engineering Sciences Uppsala University Box 534, 75121, Uppsala, Sweden Prof. A. Pérez-Rodríguez IN2UB Universitat de Barcelona c. Martí i Franquès 1, 08028 Barcelona, Spain DOI: 10.1002/aenm.201501070 Ge has demonstrated rather limited device improvements with respect to the pure Sn kesterite. Furthermore, all reports pub- lished so far are based on the use of large Ge amounts, which may compromise the viability of this technological approach, as Ge has been identified not only as a critical raw material with an earth crust abundance about 1.6 ppm, but also as a scattered material disabling the extraction by mining. [12] For example, Kim et al. reported an increase in the device efficiency from 4.6% to 6.0% using a graded band gap by alloying Cu 2 ZnSnS 4 with Ge, observing a Ge/(Sn+Ge) ratio about 0.50 at the sur- face and an almost pure Ge phase at the back. [7] The efficiency improvement is mainly explained by an increase in the J SC , and the reported values are far from the state-of-the-art efficiencies reported for kesterite solar cells. Bag et al. reported the Ge sub- stitution in CZTSe solar cells too, obtaining a rather small effi- ciency improvement (from 9.07% for the pure Sn kesterite to 9.14% for the Sn–Ge alloyed with 40% of Ge), but observing a remarkable increase in V OC of more than 50 mV. [5] Finally, Hages et al. reported an improved performance of Ge-alloyed Cu 2 Zn(Sn,Ge)(S,Se) 4 solar cells, increasing the efficiency from 8.4% (no Ge) to 9.4% (with 30% of Ge), mainly explained by an increase of 50 mV in V OC . [6] This shows that till now although very promising, attempts of alloying kesterite with large amounts of Ge have demonstrated rather small improvements in the solar cell performance. In this work, we report a breakthrough in kesterite-based technologies. We demonstrate that high-voltage and high-effi- ciency devices can be easily achieved using small quantities of Ge, leading to efficiencies higher than 10%. The Ge-based approach presented here is based on the evaporation of a 10 nm thick Ge layer on top of the Cu/Sn/Cu/Zn metallic precursor stack prior to the selenization step, using a sequential process to synthesize CZTSe absorbers. [13] As will be shown later, this leads to a substitution of less than 1.6% of Sn in the final kes- terite structure, ensuring the sustainability of the developed processes. To analyze the impact of Ge on the optoelectronic proper- ties of the devices, in the following we first present the impact of the superficial Ge nanolayer on the device parameters in a J– V and external quantum efficiency (EQE) comparison. The morphology, elemental composition, and distribution of the resulting absorber layers will be evaluated by scanning electron microscopy (SEM), by time-of-flight secondary ion mass spec- troscopy (TOF-SIMS), and by combining transmission elec- tron microscopy (TEM) with electron energy loss spectroscopy (EELS). Finally, in the Supporting Information, we present a As the knowledge on kesterite photovoltaic absorbers increases, the factors limiting the efficiency of solar cells based on this family of materials become more and more evident. [1] Com- paring the best efficiencies obtained using Cu 2 ZnSn(S,Se) 4 (CZTSSe) as absorber with more mature CdTe and Cu(In,Ga) Se 2 (CIGSe) technologies, it is clear that the voltage deficit is the major challenge that kesterite devices have to face in the near future. [2] Recently, the pure selenide Cu 2 ZnSnSe 4 (CZTSe) compound has demonstrated efficiencies exceeding 11%, with an open circuit voltage ( V OC ) of 423 mV. [3] Commonly, the highest efficiencies were reported for kesterites with Se-rich sulfo-selenide solid solutions, which generally lead to a higher V OC . This highlights the importance to identify and reduce the voltage losses for these materials. [2,4] In this sense, and with the aim to solve the inherent problems of kesterites, the substitu- tion of cations has been explored to some extent. [5–8] In par- ticular, Sn exhibits an intrinsic multicharge character and thus the Sn-site substitution is probably the most interesting. [5,9] This is because multicharge atoms commonly introduce deep defects that increase the nonradiative charge carrier recombina- tion. This has an especially severe impact on the degradation of V OC . [5,9] For this reason, one of the most interesting approaches is to substitute Sn by other cations such as Ge. By alloying CZTSe with Ge, the band gap can be tuned from 1.0 (pure Sn/Zn kes- terite) to 1.35 eV (pure Ge/Zn kesterite). [10,11] Additionally, sev- eral other beneficial effects have been associated with the use of Ge, such as the increase of carrier lifetime and the suppres- sion of Sn +2 state formation. [5,6] Nevertheless till now, the use of Adv. Energy Mater. 2015, 5, 1501070 www.MaterialsViews.com www.advenergymat.de