Concerns over climate change are not going to quell anytime soon. Because we rely so much on fossil fuels for our energy needs, the harmful carbon emissions they produce, unfortunately are one major reason for the climate change issues we face currently. Naturally, this has catapulted researchers to look for other ways to meet our great demand for energy worldwide. One of these happens to lean towards harnessing the abundant energy from our own sun using solar cells. As a renewable energy source, there are different types of solar cells but perovskites are tantalizing us with the possibility to achieve higher performance compared to the traditional material used, silicon, which has been the workhorse semiconductor material for the photovoltaics industry for many decades. This is because perovskites are really great optical absorbers and are well tuned to the solar spectrum where absorption can be adjusted easily by changing the ingredients within the crystal structure, which remarkably are made using simple low-cost solution processing rather than the more sophisticated infrastructure needed to fabricate silicon cells. Within the photovoltaics community, perovskite solar cells (PSCs) have sky-rocketed in power conversion efficiency (PCE), going up from just a few percent less than 10 years ago to close to 25.7% currently. But their Achilles heel is their frailty to the environment, making them somewhat unstable. Methyl ammonium lead iodide, MAPbI3, is one of the earliest and most studied formulations for PSCs, but because of its vulnerability to environmental stressors, such as moisture, high temperatures and ultra-violet radiation, all of these issues present major roadblocks for the commercialization of PSCs. The high sensitivity to environmental conditions in the ambient comes from the organic Methyl Ammonium (MA) cations in PSCs, given that MA has a high affinity for water molecules, causing the absorber to break down in the presence of moisture. The photocurrent also degrades after soaking the cells in light for extended periods, because the crystal structure of MAPbI3 is quite unstable. Several approaches have been used to get over the stability issue for the MAPbI3 absorber itself, as well as the contact and interface layers. So really, everything within the solar cell stack needs to function as clockwork, if the cells have to deliver maximum performance as the life of the cell progresses.
The structural stability of the perovskite absorber changes because the composition of the ions that make up the absorber within the crystal structure determines the so-called tolerance factor t. The ideal perovskite structure will come about when t = 1, but a decent range of values for the formability of alkali metal halide perovskites is 0.813 < t < 1.107. Pure perovskite compounds usually fall short, because t is low, arising in thermal or structural instability. So mixing cations and halides has become a good design option in order to make perovskite compounds with greater thermal and structural stability. Although materials engineering approaches have been used to enhance the structural and thermal stability of the intrinsic photoactive absorber material itself through triple cation and mixed halide compositions, other ways in which to affect stability is by considering the entire device structure because material degradation pathways can occur through interfaces causing ion diffusion, which is what we explored here.
In the present work, we have conducted a detailed study of one feature within the solar cell stack, namely the electrode layer, using the more intrinsically stable triple cation, mixed halide absorber, specifically Cs0.05FA0.79MA0.16PbI2.45Br0.55. Besides the Cs0.05FA0.79MA0.16PbI2.45Br0.55 (triple cation) absorber, MAPbI3 was used as a reference to further help us in understanding the degradation pathways involving the metal Anode electrode collector layer. We fabricated triple cation and MAPbI3 based PSCs with both Ag and Au electrodes, as illustrated in Figure 1(a). Our Au-based PSCs showed higher PCE than the Ag-based PSCs, and meanwhile the Ag-based MAPbI3 PSCs were highly unstable. We think this difference comes about due to the rapid generation of AgI in Ag-based PSCs. Iodine ions are liberated when water molecules diffuse through the pin holes in the Spiro-OMeTAD layer reaching the perovskite absorber, where the iodine ions in turn react with Ag to form AgI, compared to the higher temperatures needed to form AuI in Au-based PSCs; so we expect the formation of AuI in our case to be quite minimal. To conduct our material characterization studies and address these questions, we used standard probes, such as X-ray diffraction, scanning electron microscopy and photoluminescence (PL) spectroscopy. Our unique use of PL here allowed us to access key features of the perovskite absorber in situ, while it was buried and in contact with the various layers in the cell stack. Here the laser used for the incoming excitation penetrated through the Spiro-OMeTAD layer, in the lateral vicinity of the opaque metal electrodes, to access the absorber, within the solar cell stack, as shown in Figure 1(a).
The PSCs with Au electrodes showed a prominent PL emission peak for both absorbers with a clear maxima, as shown in Figure 1(b), which provides direct evidence that the Au-based PSCs for both absorbers lead to optically active absorbers while residing within the cell stack, and they are less vulnerable to degradation or the formation of AuI. On the other hand, MAPbI3 contacted with Ag showed a virtually non-existent PL emission peak, as seen in Figure 1(c), while the triple cation-based PSC with Ag showed an optically active peak. The quenching in the PL peak for the Ag-contacted MAPbI3 provided direct evidence of the Ag corrupting the optical properties of the absorber through the formation of AgI and the ease with which the MA ions come apart in the MAPbI3 upon exposure to moisture and the ambient, given its lower t. The ensuing migration of iodine ions (I-) through the pinholes in the Spiro-OMeTAD layer, leads to the rapid formation of AgI, which is corrosive and further affects the stability of the absorber below. These data are explained further in our proposed model explaining the four scenarios with the flux of the I- ions emanating in our two absorbers, in combination with our choice for the two electrodes, illustrated in Figures 1(d)-(g). We believe our work reveals the importance of the electrode material itself to determine the overall stability of the PSCs, and the choice of Au reinforces this notion to enhance device longevity through the device operational stability and environmental storage tests conducted.
Please see our current publication for more details on this work:
Kakaraparthi, M. Parashar, R. Mehta, S. Aryal, M. Temsal, and A. B. Kaul, “Stability and degradation mechanisms in triple cation and methyl ammonium lead iodide-based perovskite solar cells mediated via Au and Ag anode collector interactions,” Nature Scientific Reports 12, 18574 (2022). https://doi.org/10.1038/s41598-022-19541-6
Besides triple cation absorbers, our group has leveraged the extensive prior work we have conducted on van der Waals solids, specifically two-dimensional (2D) layered materials which we have adapted towards perovskites as well. Specifically, we are working on 2D perovskites concurrently, and in the last couple of years, our group has demonstrated one of the first reports on ink-jet printed 2D perovskites for photodetectors formed on flexible substrates . We took this work further toward an all additively manufactured approach to print not only the 2D perovskite absorber, but also the electrical contacts with graphene (Gr) and silver (Ag) inks for efficient charge collection . All ink-jet printed 2D perovskite photodetectors showed a promising photoresponse with additively manufactured electrical contacts on flexible substrates, including through bending and strain tests. In our current activities, we are working toward integrating our 2D perovksites for solar cells, just as our 3D perovskites , for addressing issues of stablility and forming PV sturctures on flexible substrates.
 M. Min,* R. F. Hossain,* N. Adhikari, and A. B. Kaul, “Inkjet printed organo-halide 2D layered perovskites for high-speed photodetectors on flexible polyimide substrates,” ACS Applied Materials and Interfaces 12, 10809 (2020); https://doi.org/10.1021/acsami.9b21053. *These authors contributed equally to the work.
 R. F. Hossain, M. Min, L.C- Ma, S. R. Sakri, and A. B. Kaul, “Carrier photodynamics in 2D perovskites with solution-processed silver and graphene contacts for bendable optoelectronics,” Nature npj 2D Materials and Applications 5, 34 (2021). https://doi.org/10.1038/s41699-021-00214-3
 M. Parashar and A. B. Kaul, “Methylammonium lead tri-iodide perovskite solar cells with varying equimolar concentrations of perovskite precursors,” Applied Sciences (MDPI), special issue of perovskites in opto-electronic application: recent advances and prospects, 11(24), 11689 (2021); https://doi.org/10.3390/app112411689
Figure 1. (a) Our n-i-p device structure where the PL beam (shown in green) was used for in-situ probing of the perovskite absorber layer as it interacted with either Au or Ag electrodes. We purposefully used the more stable triple cation perovskite, where the iodine ions are held more tightly within the perovskite crystal structure because of its higher tolerance factor compared to the lower tolerance factor of MAPbI3 perovskite absorbers, making the former more robust toward environmental stressors (right inset). (b) PL spectra of triple cation and MAPbI3 based PSCs, contacted with Au electrodes, where in both cases an optically active absorber is still present. (c) PL spectra of triple cation and MAPbI3 based PSCs with Ag electrodes, where the Ag-contacted MAPbI3 shows an optically inactive perovskite layer, but the triple cation still yields some optical activity when contacted with Ag. Our proposed model for the (d) Ag-contacted and (e) Au-contacted triple cation absorbers. Both electrodes showed a notable PL emission peak with a clearly visible peak maxima in (b). The smaller flux of iodine ions (shown through the use of thinner arrows representing the movement of iodine ions upwards from the more stable triple cation absorber), the formation of AgI, though present, is not sufficiently severe to cause device failure. On the other hand for (f) Ag-contacted and (g) Au-contacted MAPbI3 absorbers, the flux of iodine ions is larger (shown by the thicker arrows representing the movement of iodine ions upwards) because the MA ions are easily released upon exposure to environmental stressors indicated in the inset of (a). The high flux of iodine ions then results in a rapid reaction with Ag, to yield AgI which corrodes the electrodes and further breaks down the MAPbI3 intrinsically. This results in a negligible PL peak which is corroborated by the data in (c).