Insights into the interplay of different recombination mechanisms and their origins (bulk, surface) are provided comparing fresh, light-soaked and aged devices.

Apart from the high power conversion efficiencies (PCEs), [1][2][3] one of the most attractive features of ABX 3 (A = Cs, methylammonium (MA), and formamidinium (FA); B = Pb and Sn; and X = Cl, Br, and I) perovskites is the simplicity of fabrication. Perovskite thin films can be deposited through a variety of different techniques ranging from one-step [4][5][6][7][8] and two-step sequential methods, [9][10][11] vaporassisted solution processing, [12,13] and thermal gas-assisted evaporation. [9,[14][15][16][17][18] However, in a laboratory setting, one-step spin-coating remains the simplest and quickest route for high-quality perovskite layers. To improve film morphology, the spin-coating deposition has been optimized using solvent mixtures (e.g., dimethylformamide (DMF), dimethylsulfoxide (DMSO), γ-butyrolactone (GBL)), [19] and a variety of lead salt precursors. [20][21][22] Importantly, almost all currently reported All current highest efficiency perovskite solar cells (PSCs) use highly toxic, halogenated solvents, such as chlorobenzene (CB) or toluene (TLN), in an antisolvent step or as solvent for the hole transporter material (HTM). A more environmentally friendly antisolvent is highly desirable for decreasing chronic health risk. Here, the efficacy of anisole (ANS), as a greener antisolvent for highest efficiency PSCs, is investigated. The fabrication inside and outside of the glovebox showing high power conversion efficiencies of 19.9% and 15.5%, respectively. Importantly, a fully nonhalogenated solvent system is demonstrated where ANS is used as both the antisolvent and the solvent for the HTM. With this, state-of-the-art efficiencies close to 20.5%, the highest to date without using toxic CB or TLN, are reached. Through scanning electron microscopy, UV-vis, photoluminescence, and X-ray diffraction, it is shown that ANS results in similar mixed-ion perovskite films under glovebox atmosphere as CB and TLN. This underlines that ANS is indeed a viable green solvent system for PSCs and should urgently be adopted by labs and companies to avoid systematic health risks for researchers and employees.

Experimental and theoretical study on the effect of shallow and deep defects on photovoltaic performance, luminescence, surface photovoltage, and density of states.

Perovskite solar cells have attracted considerable interest among the photovoltaics research community, because of their high solar-to-electric power conversion efficiencies and low fabrication costs. [1,2] The employed metal-halide perovskites are solution-processable high-quality semiconductors with exceptional properties such as absorption over a wide spectrum, [3][4][5] a direct band gap, [6] and charge carrier diffusion lengths in the micrometer range. [7] These characteristics have enabled a broad range of possible applications ranging from perovskite solar cells (PSCs), [8] light-emitting diodes, [9] lasers, [10,11] photodetectors, [12,13] X-ray detectors, [14,15] to sensors. [16] In the past few years, organic-inorganic metal halide ABX 3 perovskites (A = Rb, Cs, methylammonium, formamidinium (FA); B = Pb, Sn; X = Cl, Br, I) have rapidly emerged as promising materials for photovoltaic applications. Tuning the film morphology by various deposition techniques and additives is crucial to achieve solar cells with high performance and long-term stability. In this work, carbon nanoparticles (CNPs) containing functional groups are added to the perovskite precursor solution for fabrication of fluorine-doped tin oxide/TiO 2 /perovskite/spiro-OMeTAD/gold devices. With the addition of CNPs, the perovskite films are thermally more stable, contain larger grains, and become more hydrophobic. NMR experiments provide strong evidence that the functional groups of the CNPs interact with FA cations already in the precursor solution. The fabricated solar cells show a power-conversion efficiency of 18% and negligible hysteresis.

Identifying and reducing the dominant recombination processes in perovskite solar cells is one of the major challenges for further device optimization. Here, we show that introducing a thin interlayer of poly(4-vinylpyridine) (PVP) between the perovskite film and the hole transport layer reduces nonradiative recombination. Employing such a PVP interlayer, we reach an open-circuit voltage of 1.20 V for the best devices and a stabilized efficiency of 20.7%. The beneficial effect of the PVP interlayer is proven by statistical analysis of various samples, many of those showing an open-circuit voltage larger than 1.17 V, and a 30 mV increase in average compared to unmodified samples. The reduced nonradiative recombination is proven by enhanced photo- and electroluminescence yields.

When discussing Fig. 9 on page 11, the following (though careful) statement can be found regarding a deep trap state:''The FTPS spectra in Fig. 9a show an additional feature located at 0.9 eV, whose values are close to the resolution limit of our measurement, but observed in all kinds of devices. This feature needs further investigation but could be a deep trap state which is responsible for the SRH recombination leading to n ID 4 1.''Having done further investigations, we are convinced that this feature at 0.9 eV does not result from the sample but is introduced as an artefact of our measurement setup. It turns out that it is a resonant feature appearing for some settings of the preamplifier, giving rise to a response at half of the energy (here around 0.9 eV) of the main signal detected between the absorption onset (1.6 eV) and the cut-off of our low-pass optical filter (ca. 1.8 eV). Therefore, further investigations on the nature of this feature are not required. Nevertheless, all the conclusions in this publication, including recombination through deep defect states, remain valid. Those states are commonly not visible in optical measurements due to their low absorption cross section and density, but can constitute efficient recombination centers. In addition, it is important to note that the FTPS peak observed at 1.35 eV was not affected by this artefact. Also the FTPS data presented in Fig. 9b, showing tail states remains valid.We thank Terry Chien-Jen Yang for collaboration and troubleshooting of the FTPS setup.

Recent advances in heterojunction and interfacial engineering of perovskite solar cells (PSCs) have enabled great progress in developing highly efficient and stable devices. Nevertheless, the effect of halide choice on the formation mechanism, crystallography, and photoelectric properties of the low‐dimensional phase still requires further detailed study. In this work, we present key insights into the significance of halide choice when designing passivation strategies comprising large organic spacer salts, clarifying the effect of anions on the formation of quasi‐2D/3D heterojunctions. To demonstrate the importance of halide influences, we employ novel neo‐pentylammonium halide salts with different halide anions (neoPAX, X=I, Br, or Cl). We find that regardless of halide selection, iodide‐based (neoPA)2(FA)(n‐1)PbnI(3n+1) phases are formed above the perovskite substrate, while the added halide anions diffuse and passivate the perovskite bulk. In addition, we also find the halide choice has an influence on the degree of dimensionality (n). Comparing the three halides, we find that chloride‐based salts exhibit superior crystallographic, enhanced carrier transport, and extraction compared to the iodide and bromide analogs. As a result, we report high power conversion efficiency in quasi‐2D/3D PSCs, which are optimal when using chloride salts, reaching up to 23.35%, and improving long‐term stability.

hole mobilities within the HTL of up to 1.6 × 10 -3 cm 2 V -1 s -1 . [4] However, the use of such highly hygroscopic lithium salts has been shown to dramatically accelerate the degradation of PSCs. [5][6][7][8] Previous works have identified that upon doping with LiTFSI, lithium ions can readily become mobile and diffuse within the perovskite layer, forming hygroscopic LiX salts (where Xis a halide) which in turn rapidly degrade the perovskite. [5,[9][10][11][12][13][14] Wang et al demonstrated that following evaporation of tert-butylpyridine (tBP), a common spiro-OMeTAD additive, LiTFSI begins to form hygroscopic aggregates which hydrate the perovskite/spiro-OMeTAD interface degrading the perovskite over a period of < 1000 hours. [8] Furthermore, it has also been shown that tBP is, by itself, insufficient to prevent the migration of lithium ions. [7] To this extent, Kim et al. recently revealed that the migration of Li + is critical in the degradation of spiro-OMeTAD-based devices and is accelerated at higher temperatures, leading to the rapid degradation of the perovskite. [7] For these combined reasons spiro-OMeTAD-based PSCs consistently fall short of the practical requirements for the commercialization of solar cells. Indeed, the poor stability of LiTFSI doped spiro-OMeTAD as an HTL has become such a problem that many leading studies using the architecture replace the HTL with a more stable, yet lower efficiency, alternative or emit reporting stability entirely. [6,15] Consequently, addressing the impact of lithium on device stability is one of the biggest challenges facing the highest efficiency PSCs.As of writing, two principal strategies have emerged within the literature for improving the stability of n-i-p PSCs, namely (i) searching for new less destructive dopants or (ii) replacing spiro-OMeTAD in favor of alternative stable HTLs such as poly(triarylamine) (PTAA), poly(3-hexathiophene) (P3HT) or a range of novel molecular transporting materials (Table S1, Supporting Information). [5,[16][17][18] While these candidates can yield improvements in the long-term stability, the PCEs of resultant devices remain inferior to the spiro-OMeTAD counterparts. This performance gap has led to a trade-off between efficiency and stability in PSCs. More recently the formation of a blocking layer between the lithium doped spiro-OMeTAD HTL and the perovskite has been suggested using either graphene or flower-like MoS 2 . [11,19] Similar to the previous approaches, devices prepared using a blocking layer have yet to match the high efficiencies achieved using the conventional spiro-OMeTAD architecture, reaching up to 20.18% in devices prepared with MoS 2 . Furthermore, by blocking the Li + at the interface, the mechanism by which hygroscopic agregates are formed is not avoided. This challenge, therefore, requires new strategies that target the Li + ions within the HTL itself, preventing the accumulation of Li + at the interface and subsequent degradation. Very recently, several studies have sought to replace the Li + metal ...