Recent advances in polymer/silicon heterojunction solar cells

RECENT ADVANCES IN POLYMER/SILICON HETEROJUNCTION SOLAR CELLS
ABSTRACT: In this contribution, we first give a brief historical overview of the recent developments on polymer/silicon heterojunction cells. We then focus on our most recent results concerning: (i) the PEDOT:PSS/c-Si interface design, where the thickness of the native oxide in-between the c-Si surface and the polymer is shown to play a crucial role, (ii) a new type of heterojunction cell, where the hole-selective layer is PEDOT:PSS and the electronselective
layer is well-passivating phosphorus-doped amorphous silicon (n-a-Si:H), (iii) the stability of the
polymer/silicon cells in ambient environment. We demonstrate for the first time that the efficiencies of BackPEDOT cells are long-term stable if a metal foil is used as rear metallization scheme, as the metal foil does not transmit any humidity in contrast to evaporated metals. (iv) We present first results of our attempt to implement a PEDOT:PSS/c- Si junction into an industrial-type large-area (15.6×15.6 cm2) screen-printed c-Si solar cell and reach an efficiency of 19.5% in our first batch of cells.
Keywords: c-Si; Polymer Film; Heterojunction; Hybrid; Passivation; Stability; Screen-Printing

1 BRIEF HISTORICAL OVERVIEW

State-of-the-art junctions in silicon solar cells are realized by technologically advanced methods, such as high-temperature dopant diffusion or plasma-enhanced chemical vapour deposition of doped amorphous silicon layers. A technologically much less demanding approach is the deposition of carrier-selective polymers from a liquid precursor onto the silicon surface. A lot of effort has been devoted in recent years on polymer/silicon heterojunction cells based on the hole-selective polymer poly(3,4-ethylene-dioxythiophene):polystyrenesulfonate
(PEDOT:PSS).

Figure 1: Efficiency evolution of PEDOT:PSS/c-Si front-junction cells. The highest efficiencies reached in the respective year are marked red.

Figure 1 shows the evolution of efficiencies since the first silicon solar cell with PEDOT:PSS on the front was presented by Avasthi et al. [1] in 2011. Although the efficiencies of the front-PEDOT:PSS/c-Si junction devices shown in Fig. 1 increased quite dramatically in recent years [1-8], open-circuit voltages were typically limited in these cells to values Voc < 600 mV, which was assumed to be due to recombination at the PEDOT:PSS/c-Si interface. However, as we were able to show in 2013 by means of contactless lifetime measurements [3], PEDOT:PSS leads to an excellent interface passivation on crystalline silicon. We recently demonstrated [9,10] that Voc values > 700 mV are achievable with the PEDOT:PSS/c-Si junction. It turnedout that most of the cells shown in Fig. 1 were in fact limited by recombination at the full-area metallized rear.

An additional limitation was the parasitic absorption within the PEDOT:PSS at the cell front. We hence proposed to deposit the PEDOT:PSS onto the back of the cell and based on this so-called “BackPEDOT” cell concept [10] we were able to demonstrate a record-high efficiency of 20.6% [9] and Voc values up to 663 mV on device level, clearly proving that BackPEDOT cells have
the potential to reach very high efficiencies.
2 PEDOT:PSS/c-Si INTERFACE DESIGN
One crucial element in the optimization of the PEDOT:PSS/c-Si interface turned out to be the silicon oxide layer in-between the polymer and the silicon substrate [10]. The preparation method and the oxide thickness have a critical impact on the interfacerecombination properties as well as the hole transport properties. We achieve the best passivation properties using a native oxide grown in ambient environment. As the holes have to tunnel through the native SiOx layer, the thickness of the oxide is critical concerning the contact resistance of the polymer/silicon junction. We have optimized the interfacial oxide on test structures and then
verified our results on actual BackPEDOT solar cells
Figure 2 shows the J-V curves of two BackPEDOTsolar cells [11] with conventionally processed, that is, phosphorus-diffused, front surface and evaporated contacts featuring two different pre-treatments of the silicon surface prior to the PEDOT:PSS coating. After finishing the conventionally processed front side of the solar cells all samples were dipped in a 1% HF solution for 1 minute. After this HF dip, some samples were deposited with PEDOT:PSS after 10 minutes storage in air (‘thin’ native SiOx) while other samples were kept in air for 48 hours (‘thick’ native SiOx). Subsequently, the PEDOT:PSS layer was deposited by spin-coating on the
entire rear at 500 revolutions per minute (rpm) for 10 seconds and subsequently 1000 rpm for 30 seconds. Thesamples were then annealed on a hotplate in air at 130°Cfor 15 minutes for drying and to remove residual solvents.