Öffnet in neuem Fenster Opens in a new window Öffnet externe Seite Opens an external site Öffnet externe Seite in neuem Fenster Opens an external site in a new window


CIGS layers grown by multi-source evaporation

While the pursuit of world record efficiencies and future device concepts is a valuable and necessary assessment of a PV technology’s potential, their compatibility with industrial production technologies is often not an issue of concern. Hence, the PVcomB currently establishes a high efficiency reference line, based on thermally coevaporated CIGSe absorbers, with a focus on technology transfer for highly efficient Cu(In,Ga)Se2 (CIGSe) thin film devices and the realization of advanced device concepts.

Our present activities aim at an understanding of the alkali surface treatment and its compatibility with dry buffer layer technologies and – in view of efficient bottom devices in tandem structures – we are looking at concepts for an improved thermal stability of the absorber/buffer interface. 

Recent topics

Heavy Alkali Post Deposition Treatment of CIGS Absorbers

XPS model - enlarged view

Fig. 1: Our current understanding of the working mechanism of a RbF post deposition treatment of a CIGS absorber

An RbF post deposition treatment (PDT) of CIGS absorber layers used in our baseline devices was observed to lead to a strong improvement of the device’s VOC often accompanied by a reduction in FF. There are several mechanisms responsible for these effects [1]: On a bulk level the RbF-PDT enhances the carrier concentration via an exchange mechanism with Na and increases the minority carrier lifetime possibly via defect passivation at the grain boundaries. Furthermore the PDT leads to the formation of a secondary (Rb,Na)-InxSey phase at the surface of the CIGS. The most probable candidate for this phase was identified to be Na-doped RbInSe2 [2]. Dependent on the exact composition, thickness and electronic properties of this layer it may either reduce the interface recombination velocity contributing to the VOC-gain or lead to the formation of a barrier lowering FF. Via adaptation of the CIGS layer properties a more beneficial formation of this surface layer could be achieved and lead to a recovery of the FF maintaining the higher VOC

[1] T. Kodalle, M. D. Heinemann, D. Greiner, H. A. Yetkin, M. Klupsch, C. Li, P. A. van Aken, I. Lauermann, R. Schlatmann, C. A. Kaufmann, Elucidating the Mechanism of an RbF Post Deposition Treatment in CIGS Thin Film Solar Cells, Solar RRL 2 (9), 1800156 (2018).

[2] T. Kodalle, R. Kormath Madam Raghupathy, T. Bertram, N. Maticiuc, H. A. Yetkin, R. Gunder, R. Schlatmann, T. Kühne, C. A. Kaufmann, H. Mirhosseini, Properties of RbInSe2 thin films, Physica Status Solidi: Rapid Research Letters 1800564 (2018)

Building a High Efficiency Reference Line

enlarged view

Fig. 2: Properties of the growing CIGSe thin film as observed via White Light Reflectometry (WLR) in real-time during 3-stage coevaporation: deposition rate from WLR, chalcopyrite phases (158 signifying Cu(In,Ga)5Se8, same for 135 and 112) from ED-XRD, Eg,min from WLR, ESGT from WLR, roughness (not shown) from WLR.[4]

Taking advantage of the relocation of our complete CIGSe coevaporation lab and infrastructure, the vacuum quality of the deposition system was upgraded, such that the effect of the background pressure and composition of the residual gas composition on film growth and resulting device properties could be studied. A clear effect on e.g. the In/Ga interdiffusion during film growth and carrier concentration/collection was demonstrated. [3] Implementation of optical process controls based on the use of either laser light that is scattered or infrared light that is reflected off the growing CIGSe thin film enables precise process control. Lately the use of white light reflectometry (WLR) in combination with in-situ real-time EDXRD revealed the direct correlation of material and optical properties during film formation (see Fig.2).[4] The sensitivity of the WLR signal to the effect of an alkali surface conditioning of the complete CIGSe thin film is currently under investigation. After relocation of the coevaporation lab in August 2014 the complete process chain for device fabrication has been recommissioned, a process that is still ongoing.

Our current in-house top device efficiency is just above 19% (total area 1.1 cm²).

[3] D. Greiner, J. Lauche, M.D. Heinemann et al., Proc. 43rd IEEE PVSC, Portland (2016) 1151-1156, doi:10.1109/PVSC.2016.7749795.

[4] M.D. Heinmann, R. Mainz, F. Österle et al., Scientific Reports 7 (2017) 45463, doi:10.1038/srep45463.

CIGSe Superstrate Devices

Superstrate concepts - enlarged view

Fig. 3 CIGSe device architecture – substrate vs. superstrate [5].

In comparison to the standard substrate device architecture the use of the superstate configuration for CIGSe devices can offer a number of considerable advantages, above all the potential of an improved light harvesting.[5] However, the limited and often unstable power-conversion-efficiencies of superstrate devices, that are reported in the literature, decrease the appeal of the device concept.

When growing Cu(In,Ga)Se2 (CIGSe) by multistep coevaporation directly on an ZnO:Al/i-ZnO double layer TCO front contact an amorphous GaOx layer forms under certain process conditions at the ZnO/CIGSe interface. Suprisingly this layer can act as an efficient buffer layer between CIGS and TCO. However, Cu and/or Na, present as contaminants in the GaOx, are likely to induce acceptor-type defects close to the interface. In our model, this is the main reason for the low efficiencies, continuously observed for CIGSe superstrate devices. Our efforts resulted in a stable 11.4% efficient lab scale device.[6] Based on these results, efforts were made to identify a suitable buffer layer material for a thermally stable pn-junction. In a bold statement we suggest that an amorphous, highly n-doped (In,Ga)Ox thin film could be such a compound.[7]

[5] M. D. Heinemann, F. Ruske, D. Greiner et al., Sol. Eng. Mat. & Sol. Cells 150 (2016) 76-81, doi:10.1016/j.solmat.2016.02.005.

[6] M. D. Heinemann, V. Efimova, R. Klenk et al., Prog. Photovolt: Res. Appl. 23 (2015) 1228-1237), doi:10.1002/pip.2536.

[7] M. D. Heinemann, M. F. A. M. van Hest, M. Contreras er al., Phys. Status Solidi A 214 (2017) 1600870, doi:10.1002/pssa.201600870.

Solar Cells for Space Applications

Space 2 - enlarged view

Fig. 4 In the development of a CIGSe based solar generator for space applications within a German consortium of academic and industrial partners the HZB studied absorber growth at low deposition temperatures on flexible polyimide foil substrates. [8]

From 2007 until 2015 a German consortium of academic and industrial partners, developed an extremely light and flexible Cu(In,Ga)Se2 (CIGSe) thin film solar cell technology for space applications.[8] The combination with a light support structure and an appropriate interconnection technology enables the construction of a solar generator with previously unmatched specific power (W/kg).

While the HZB coordinated the activity, which was funded by the DLR, our scientific contribution focused on the development of a low-temperature CIGSe deposition process on a polyimide foil substrate with the aim to maximize the photo-conversion-efficiency and to perform efficient technology transfer to the device fabricating industry partner. During in-situ real-time growth studies that compared processes with max growth temperatures of 420°C and 530°C, the importance of a Cu-rich growth phase for the lowT process could be demonstrated.[9] A maximum efficiency of up to 17.9% was achieved for lab scale devices (0.5 cm²).[10] At the time, this efficiency was reached without an additional alkali surface treatment.

[8] C. A. Kaufmann, D. Greiner, S. Harndt et al., Proc. 29th EU PVSEC, Amsterdam (2014) 1439-1443, doi:10.4229/EUPVSEC20142014-3AO.5.4.

[9] C. A. Kaufmann, D. Greiner, H. Rodriguez-Alvarez et al., Proc. 39th IEEE PVSC, Tampa (2013) 3058-3061, doi:10.1109/PVSC.2013.6745106.

[10] D. Greiner, J. Lauche, S. Harndt et al., Proc. 42nd IEEE PVSC, New Orleans (2015) 1-6, doi:10.1109/PVSC.2015.7356149.

Connected projects