as higher PCEs have been achieved due to lower parasitic absorption losses, [4,5] and their low-temperature processing enables a wider choice of bottom cells.The device stability and PCE of silicon/perovskite tandem solar cells are crucially determined by the quality of the contact layer interfaces, [6,7] as they are found to limit especially the open circuit voltage (V OC ) and fill factor. Charge carrier selective contacts and their interfaces toward MHP absorbers are thus a field of recent and ongoing research. [8][9][10] The charge carrier transport across interfaces depends on two main parameters: The energy level alignment of both materials forming the interface, and the density of defect states at the surfaces and interfaces.The fullerene C 60 has become the standard electron transport layer (ETL) for p-i-n MHP solar cells. Similarly, it has also been applied in silicon/perovskite tandem solar cells with a p-i-n structure for the perovskite top cell, yielding the highest-efficiency devices in their class. [10] However, the obtained open circuit voltage for p-i-n MHP solar cells is found to be limited by the electron selective interface due to non-radiative recombination losses, [11,12] which raises interest in thoroughly studying the energetic formation of the perovskite/C 60 interface, including the energy level alignment and density of gap states. The insertion of an ultra-thin (≈1 nm) LiF interlayer between the perovskite and C 60 has been shown to significantly increase the device V OC by reducing the non-radiative recombination while keeping a highThe fullerene C 60 is commonly applied as the electron transport layer in highefficiency metal halide perovskite solar cells and has been found to limit their open circuit voltage. Through ultra-sensitive near-UV photoelectron spectroscopy in constant final state mode (CFSYS), with an unusually high probing depth of 5-10 nm, the perovskite/C 60 interface energetics and defect formation is investigated. It is demonstrated how to consistently determine the energy level alignment by CFSYS and avoid misinterpretations by accounting for the measurement-induced surface photovoltage in photoactive layer stacks. The energetic offset between the perovskite valence band maximum and the C 60 HOMO-edge is directly determined to be 0.55 eV. Furthermore, the voltage enhancement upon the incorporation of a LiF interlayer at the interface can be attributed to originate from a mild dipole effect and probably the presence of fixed charges, both reducing the hole concentration in the vicinity of the perovskite/C 60 interface. This yields a field effect passivation, which overcompensates the observed enhanced defect density in the first monolayers of C 60 .

We report the complex refractive index of methylammonium lead iodide (CH 3 NH 3 PbI 3) perovskite thin films obtained by means of variable angle spectroscopic ellipsometry and transmittance/reflectance spectrophotometry in the wavelength range of 190 nm to 2500 nm. Film thickness and roughness layer thickness are determined by minimizing a global unbiased estimator in the region where the spectrophotometry and ellipsometry spectra overlap. We then determine the optical bandgap and Urbach energy from the absorption coefficient, by means of a fundamental absorption model based on band fluctuations in direct semiconductors. This model merges both the Urbach tail and the absorption edge regions in a single equation. In this way, we increase the fitting region and extend the conventional (α ω) 2-plot method to obtain accurate bandgap values.

We report on the formation of wrinkle-patterned surface morphologies in cesium formamidinium based Cs x FA 1-x Pb(I 1-y Br y) 3 perovskite compositions with x = 0-0.3 and y = 0-0.3 under various spin-coating conditions. By varying the Cs and Br content, perovskite precursor solution concentration, and spin-coating procedure, the occurrence and characteristics of the wrinkleshaped morphology can be tailored systematically. Cs 0.17 FA 0.83 Pb(I 0.83 Br 0.17) 3 perovskite layers were analyzed regarding their surface roughness, microscopic structure, local and overall composition, and optoelectronic properties. Application of these films in p-in perovskite solar cells (PSCs) with ITO/NiO x /perovskite/C 60 /BCP/Cu architecture resulted in up to 15.3% and 17.0% power conversion efficiency for the flat and wrinkled morphology, respectively. Interestingly, we find slightly red-shifted photoluminescence (PL) peaks for wrinkled areas and we are able to directly correlate surface topography with PL peak mapping. This is attributed to differences in local grain size, while there is no indication for compositional de-mixing in the films. We show that perovskite composition, crystallization kinetics, and layer thickness strongly influence the formation of wrinkles which is proposed to be related to the release of compressive strain during perovskite crystallization. Our work helps to better understand film formation and to further improve efficiency of PSCs with widely used mixed perovskite compositions.

We develop a band-fluctuations model which describes the absorption coefficient in the fundamental absorption region for direct and indirect electronic transitions in disordered semiconductor materials. The model accurately describes both the Urbach tail and absorption edge regions observed in such materials near the mobility edge in a single equation with only three fitting parameters. An asymptotic analysis leads to the universally observed exponential tail below the bandgap energy and to the absorption edge model at zero Kelvin above it, for either direct or indirect electronic transitions. The latter feature allows the discrimination between the absorption edge and absorption tails, thus yielding more accurate bandgap values when fitting optical absorption data. We examine the general character of the model using a dimensionless Joint Density of States formalism with a quantitative analysis of a large amount of optical absorption data. Both heavily doped p-type GaAs and nano-crystalline Ga 1−x MnxN, as examples for direct bandgap materials, as well as amorphous Si:Hx, SiC:Hx and SiNx, are modeled successfully with this approach. We contrast our model with previously reported empirical models, showing in our case a suitable absorption coefficient shape capable of describing various distinct materials while also maintaining the universality of the exponential absorption tail and absorption edge.