Project no. 316488 Project Acronym: KESTCELLS Project title: Training for suitable low cost PV technologies: development of kesterite based efficient solar cells. Industry-Academia Partnerships and Pathways Start date of project: 01/09/2012 Duration: 48 months Project coordinator: Dr. Edgardo Saucedo Project coordinator organization name: IREC Project website address: www.kestcells.eu Deliverable 3.3 Kesterite-based solar cell with optimized buffer and window layers and efficiency > 10% Delivery date: Month 36 (July 2016) Dissemination Level PU Public Document details: Workpackage Partners Authors WP3: Implementation of solar cells EMPA / HZB Stefan G. Haass, Yaroslav E. Romanyuk, Ayodhya N. Tiwari Document ID D3.3 Release Date --/--/2016 1
Kesterite-based solar cell with optimized buffer and window layers and efficiency > 10% Introduction: Overcoming the threshold of 10% conversion efficiency is challenging because of several issues concerning the absorber layer synthesis: Avoiding detrimental secondary phases like Zn(S,Se), CuSn(S,Se) 3 and Cu x(s,se) A well crystallized and dense morphology Copper poor and zinc rich stoichiometry Dealing with Sn loss during high temperature annealing Stable and Ohmic back contact in the presence of Mo(S,Se) 2 formation and kesterite phase decomposition This deliverable describes ways to overcome some of the above listed challenges and demonstrates kesterite solar cells with more than 10% conversion efficiency. The CZTSSe absorbers are prepared from precursors obtained from a solution approach with dimethyl sulfoxide (DMSO) as the solvent with subsequent recrystallization by annealing under controlled atmosphere of selenium. Some commonly reported problems of the DMSO-processed kesterite layers are their high porosity, non-uniformity, and numerous grain boundaries that can lead to undesirable recombination. Here, we employ a three-stage annealing process under controlled selenium atmosphere in an SiO x coated graphite box to drastically improve the grain size and morphology of the absorber layer. Importantly, the V oc deficit can be reduced to 0.57 V, which appears to be one of the lowest values reported for kesterite devices. Systematic electrical characterization of absorbers and finished solar cells with time-resolved photoluminescence (TRPL), temperature-dependent current voltage measurements ( JV T ), and admittance spectroscopy (AS) are used to identify the reasons of the improved voltage. Additionally, a series with different cadmium sulfide (CdS) layer thicknesses was processed in order to optimize the buffer layer. Buffer layer: Cadmium sulfide buffer layers still result in the highest device efficiencies, although alternative buffer layers have been successfully applied to kesterite solar cells [2]. Therefore we optimized the CdS buffer layer thickness by variation of the deposition time from 19 minutes to 26 minutes. The CdS buffer layer thickness was measured by SEM cross sections and the finished devices were characterized by current voltage measurements (J-V), external quantum efficiency (EQE) and timeresolved photoluminescence (TRPL). Figure 1 shows the average efficiency, open circuit voltage (V oc), short circuit current (J SC) and fill factor (FF) for all cells. The resulted efficiencies are very similar and no trend is visible. The sample with 24 min deposition time yields a lower FF and therefore also a lower efficiency. The V oc of all samples is very similar, however, the J SC shows a small increase with thicker CdS. In Figure 2 the external quantum efficiency is shown for four representative cells for each 2
deposition time. The absorption in the wavelength region from 400-550 nm indicates a thicker CdS for the longer deposition time as expected and additionally confirmed by SEM cross section images (not shown here). The deposited CdS thickness varies from approximately 45 nm (19 min) to 55nm (22 min), 65 nm (24 min) and finally 75 nm (26min). While thinner CdS gives a slightly higher current in the short wavelength-region, the overall collection from 550nm onwards is higher with thicker CdS, most probably due to a better quality of the p-n junction. The bandgap derived from the derivative of the EQE signal in the long wavelength region is, as expected, not influenced by varying the CdS thickness and remains at 1.05 ev for all samples. Overall it can be concluded that the change in CdS thickness did not have a significant impact on device performance, however, for the champion device, 24 min of CdS deposition was used, as this gave the slightly higher efficiencies in this experiment. Figure 1.: A) efficiency, B) open circuit voltage, C) fill factor and D) short circuit current of solar cells with increasing CdS thickness. Overall there is only very little difference in photovoltaic parameters. There is a small improvement in short circuit current with thicker CdS. 3
Figure 2.: External quantum efficiency of 4 representative devices processed with 19 26 minutes CBD. The change in absorption around 500 nm indicates the different CdS thicknesses. Kesterite solar cell with efficiency > 10%: Figure 3 shows the properties of the champion device. The SEM cross-section image exhibits a largegrain upper crust, a rather narrow small-grain bottom layer and a distinct MoSe 2-layer. The JV-T curve yields a total-area efficiency of 11.2%, whereas 10 best cells had an average efficiency of 10.6 ± 0.3%. The EQE measurement shows a relatively good collection of carriers from the long-wavelength region, which is in line with the long TRPL decay in Figure 3 d). The decay curve is fitted with a single exponential function in the range 10 100 ns since the faster decay during first 10 ns originates from the charge separation in the device [3]. The fitted long minority carrier lifetime of τ 2 = 8.1 ns is responsible for the improved collection in the long-wavelength region, which is manifested by plotting the ratio of reversed bias EQE and zero bias EQE. 4
Figure 3.: Characteristics of the best device. a) SEM cross section exhibiting large grained morphology. b) Dark and illuminated J-V-measurement. c) EQE measured with 0 V and -1 V bias. The integrated EQE yields the current of 36.5 macm -2. The bandgap is estimated from the minimum in the derivative in the long-wavelength range. In the upper part the ratio of EQE(-1V)/EQE(0V) is shown. d) TRPL transient at room temperature yields a minority carrier lifetime of 8.1 ns, and the PL maximum is located at 1260 nm. Further advanced characterization on that sample was conducted using Admittance Spectroscopy (AS) and JV-T measurements. Figure 4 a) shows the JV-T curves in darkness and under illumination. The crossover of illuminated and dark curves is becoming more pronounced at lower temperatures, whereas the increasing rollover of the J-V-curves leads to a complete blocking of the current at the lowest temperature of 123K. Possible explanations for this blocking is a barrier at the interface between absorber and the Mo back contact, which facilitates the minority carrier recombination [4], or an increase in bulk resistivity, due to the lack of shallow acceptor states and therefore a freeze out of deeper acceptor states rendering the device fully depleted and exhibiting high resistivity [5]. The temperature dependence of the V oc extrapolated to T = 0K provides an intercept of 5
E A (Voc-T) = 0.99 ev, representing the activation energy for the dominant recombination mechanism. Since this value is very close to the derived bandgap of 1.05 ev, one can conclude that the dominant recombination paths are located within the bulk of the absorber rather than at the interface [6]. Figure 4.: a) Temperature dependent J-V-measurement (dark curve dotted line, light curve solid line). The inset shows a linear fit of V OC that can be extrapolated to an intersection value of E A = 0.99 ev, which is close to the estimated bandgap of E G = 1.05 ev. b) Temperature dependence of the series resistance R S obtained from the dark J-V curves, which is fitted with a thermal activation energy E A = 182 mev. In Figure 4 b) the temperature dependence of the dark series resistance R s is shown. Using the model for a Schottky barrier at the back contact [7] the barrier height can be calculated from equation (1) R S = R 0 + k e EA qa T kt (1) where A* is the effective Richardson constant, R 0 is the background series resistance and E A the activation energy due to the barrier. The Arrhenius plot in the inset yields an activation energy of E A (R s-t) = 182 mev. AS measurements in Figure 5 show one freeze out in the frequency range from 200 6
Hz to 2 MHz. Assuming a point defect, one can utilize equation (2) in order to extract the activation energy: ω 0 = 2ξ 0 T 2 e E A kt (2) with ω 0 = 2πf the freeze out frequency and ξ 0 the temperature independent thermal emission pre-factor. This results in a thermal emission depth E A (Cf-T) = 108 mev with ξ 0 = 4694 s -1 K -2. Figure 5.: a) Temperature-dependent capacitance frequency measurements in the temperature range from 123 K to 323 K and frequencies from 200 Hz to 2 MHz. b) Arrhenius plot of the freeze out frequency yields an activation energy of E A = 108 mev with ξ 0 = 4694 s -1 K -2. However the freeze out observed in the AS measurement could also be attributed to a Schottky barrier in the device and a corresponding activation energy is derived by a change in the temperature dependence of the pre-factor from T 2 to T 3/2, yielding an activation energy of E A,B (Cf-T) = 115 mev. [8] From the Cf-T measurements the dark series resistance can also be calculated by employing an admittance circuit model of a depletion region in series with the undepleted quasi-neutral region [5]. Comparing these values with the R s derived from JV-T shows one order of magnitude lower values for 7
R s obtained from AS, implying that the activation energies E A (Cf-T) and E A (R s-t) are not attributed to the same effect. Conclusion: In summary, an 11.2%-efficient (total cell area measurement) CZTSSe solar cell is fabricated using the hydrazine-free DMSO solution approach. The best device features the V oc-deficit of only 0.57 V which is amongst the lowest for solution processed CZTSSe devices. The open circuit voltage improvement was possible because of the 3-stage annealing process in a silica-coated closed reactor, which enabled an increased incorporation of selenium and large-grained microstructure. The reduced V oc-deficit is attributed to an increased minority carrier lifetime, low diode saturation current and ideality factor, which are signatures of the semiconductor material with a low concentration of recombination centers. [1] S. Siebentritt, Thin Solid Films 2013, 535, 1. [2] T. Ericson, J. J. Scragg, A. Hultqvist, J. T. Watjen, P. Szaniawski, T. Torndahl, C. Platzer- Björkman, IEEE J. Photovolt. 2014, 4, 465. [3] T. K. Todorov, J. Tang, S. Bag, O. Gunawan, T. Gokmen, Y. Zhu, D. B. Mitzi, Adv. Energy Mater. 2013, 3, 34. [4] R. Scheer, H.-W. Schock, Chalcogenide Photovoltaics: Physics, Technologies, and Thin Film Devices, John Wiley & Sons, 2011. [5] O. Gunawan, T. Gokmen, C. W. Warren, J. D. Cohen, T. K. Todorov, D. A. R. Barkhouse, S. Bag, J. Tang, B. Shin, D. B. Mitzi, Appl. Phys. Lett. 2012, 100, 253905. [6] V. Nadenau, U. Rau, A. Jasenek, H. W. Schock, J. Appl. Phys. 2000, 87, 584. [7] O. Gunawan, T. K. Todorov, D. B. Mitzi, Appl. Phys. Lett. 2010, 97, 233506. [8] T. Eisenbarth, T. Unold, R. Caballero, C. A. Kaufmann, H.-W. Schock, J. Appl. Phys. 2010, 107, 034509. 8