The influence of additives in the stoichiometry of hybrid lead halide perovskites

We investigate the employment of carefully selected solvent additives in the processing of a commercial perovskite precursor ink and analyze their impact on the performance of organometal trihalide perovskite photovoltaic devices. We provide evidence that the use of benzaldehyde can be used as an effective method to preserve the stoichiometry of the perovskite precursors in solution. Benzaldehyde based additive engineering shows to improve perovskite solid state film morphology and device performance of trihalide perovskite based solar cells.

properties such as the energy band gap or the stability of the resulting photoactive layers. 2 Indeed, compositional engineering is one of the main strategies used to boost device performance. [3][4][5] High quality perovskite layers can be obtained by properly controlling the crystal growth. 6,7 However, some of the most commonly used precursors as well as the resulting perovskite layers tend to decompose upon exposure to humidity and high temperature, even under material encapsulation. 8 Unstable perovskite precursors impede a precise control of the stoichiometry to yield high quality films. This is even more severe in the case of mixed halides, where the ratio of the precursors can be critical. 9 On the other hand, solvent engineering and the use of additives have also contributed meaningfully to the field. [10][11][12][13][14] In this work we examine the effect of solvent additives engineering on the morphology, optoelectronic properties and solar cell device performance of solution processed methylammonium lead mixed iodide-chloride perovskites (CH3NH3PbI3-xClx) photovoltaics.
Noteworthy, we use a commercial precursor ink containing methylammounium iodide (MAI), PbCl2 and PbI2 at a molar ratio of 1:1:4 (PbCl2:PbI2:MAI) in anhydrous N,Ndimethylformamide (DMF). Importantly, in order to produce a representative study we employ a set of the following criteria to select the appropriate solvent additives: i) inclusion of both higher and lower polarity solvents compared to the reference host DMF solvent, ii) use of higher boiling point (bp) solvents than DMF (bp 153 °C), while iii) keeping high enough vapour pressure to ensure a complete solvent removal during perovskite film formation at low temperature (80 °C) and iv) different nucleophilic strength and/or reducing character. The selected additives are: 1,2,3,4-tetrahydronaphthalene (aka tetralin), γ-valerolactone, acetophenone, furfural, benzaldehyde, N-methylformamide, dimethyl sulfoxide (DMSO) and benzylamine (see Table I). of benzylamine resulted in the formation of large cuboid-shaped crystals, which could be related to the changes in stoichiometry discussed later in the text. UV-Vis absorbance and relevant Tauc plots indicate that the additive treatment has a minor (maximum shift of 5 meV compared to the reference DMF solvent) and non-systematic influence on the material energy band gap, with the exception of benzylamine where a ~20 meV red-shift is observed (Fig. S1).
This moderate change could be related to lattice structural distortions derived from a modified stoichiometry in the perovskite film formation. 18  In the CH3NH3PbI3−xClx based formulation under investigation we did not observe any clear correlation between the perovskite film morphology nor the device performance with the polarity, boiling point and vapor pressure of the additives tested. Even so, the most significant changes described above were obtained with DMSO, benzylamine and benzaldhyde, three solvents whose boiling points are 25-35 °C above that of the host solvent DMF. Therefore we can not exclude the relevance of this parameter. Moreover, we believe that further studies based on the selection of additives presented here are necessary to fully understand their morphological influence of polarity, boiling point and vapor pressure on alternative perovskite formulations.
Based on the results of our preliminary study, the parameter which appears to have a measurable impact on the film morphology and device performance is the reducing character of the additive. Thus we further explored how the additive reducing character affects the film characteristics and in particular the stoichiometry of the perovskite material. For that reason we characterized the absorbance of the precursor formulations with increasing additive content ( Fig. 3). Firstly, the focus was set on the spectral features at 320 nm and 365 nm, corresponding to absorption signatures of PbI2 and MAI, respectively. 19 It is well known that organic iodides such as MAI are not very stable in air and under light irradiation, mainly due to the oxidation of Iinto I2. By monitoring the PbI2 and MAI features, we aimed to probe potential variations in the stoichiometry of the original precursor ink induced by the additives. The resulting spectra showed no changes when using tetralin, -valerolactone, DMSO or N-methylformamide additives. On the other hand, the PbI2 signal was slightly stronger (weaker) with acetophenone and furfural (benzylamine). Finally, both PbI2 and MAI peaks showed a meaningful increase upon benzaldehyde treatment. We ascribe this effect to the reducing character of benzaldehyde, which appears to prevent the oxidation of the unstable precursors and eventually reduce already properties. Importantly, the fact that this change is observed in solution supports the idea that this is an isolated effect rather than a combined action with the boiling point or the vapor pressure, which one could expect to have an impact on the drying kinetics during film formation.
This valuable finding demonstrates that the limiting instability of the perovskite precursors can be significantly mitigated by the simple addition of small amounts of specific reducing agents such as the herein proposed benzaldehyde. As a result, the modified ink has a higher unreacted precursor content compared to the untreated material, which results in the improved perovskite microstructure. 20 Our results are in good agreement with the work reported by Zhang et al., where the addition of hypophosphorous acid as a reducing agent yielded a significantly enhanced perovskite microstructure. 14 In our case, we believe that the relatively low precursor concentration of the untreated commercial ink is responsible for the lower solar cell efficiencies achieved with such a material compared to the reported state of the art devices. Furthermore, in contrast to the aforementioned work 14  interlayer. 13 We additionally observe that the MAI to PbI2 ratio increases upon benzaldehyde addition. The role of non-stoichiometric precursor ratios is still a controversial topic. 21,22 In our work, the shifted stoichiometry, with a slight excess of MAI compared to the original ink, might also be a contributing factor for the formation of improved quality perovskite films. The addition of this solvent appears therefore as a simple way to control the CH3NH3PbI3−xClx perovskite morphology and stoichiometry and yield better solar cell device performances for the CH3NH3PbI3−xClx precursor ink.
In summary, we have investigated the use of solvent additives in a commercial  The ISOS-D-1 protocol was followed to carry out the stability characterization. Figure S1. a) UV-Vis absorbance spectra and b) Tauc plots of perovskite films prepared from precursor solutions containing different additives.