The role of Pb oxidation state of the precursor in the formation of 2D perovskite microplates

Two-dimensional (2D) lead halide perovskites are an exciting class of materials currently being extensively explored for photovoltaics and other optoelectronic applications. Their ionic nature makes them ideal candidates for solution processing into both thin films and nanostructured crystals. Understanding how 2D lead halide perovskite crystals form is key towards full control over their physical properties, which may enable new physical phenomena and devices. Here, we investigate the effects of the Pb oxidation state of the initial inorganic precursor on the growth of pure-phase (n = 1) – Popper 2D perovskite BA2PbI4 in single-step synthesis. We examine the different crystallisation routes in exposing PbO2 and PbI2 powders to a BAI : IPA organo-halide solution, by combining in situ optical microscopy, UV–VIS spectroscopy and time-resolved high performance liquid chromatography. So far, works using PbO2 to synthesise 3D LHPs introduce a preceding step to reduce PbO2 into either PbO or PbI2. In this work, we find that BA2PbI4 is directly formed when exposing PbO2 to BAI : IPA without the need for an external reducing agent. We explain this phenomenon by the spontaneous reduction/oxidation of PbO2/BAI that occurs under iodine-rich conditions. We observe differences in the final morphology (rectangles vs. octagons) and nanocrystal growth rate, which we explain through the different chemistry and iodoplumbate complexes involved in each case. As such, this work spans the horizon of usable lead precursors and offers a new turning knob to control crystal growth in single-step LHP synthesis.

1 Fabrication of the lead precursor/ITO substrates Lead(IV) oxide (CAS:1309-60-0) and Lead(II) iodide (CAS:10101-63-0 ) was purchased from Sigma-Aldrich. To make different lead precursor films on ITO, corresponding 22mM of lead precursor suspensions were made in anhydrous isopropanol. These precursor suspensions were sonicated to form a stable dispersion. ITO substrates (1.5 * 1.5 cm 2 ) were multi-step cleaned with soap water, acetone, isopropanol, deionized water and dried with a nitrogen gun. Cleaned ITO substrates were placed on a hotplate at 58 degrees and the lead precursor solutions were dropcasted on the substrates. After the solvent evaporated (10 minutes) the substrates were removed from the hotplate. The size of the resulting PbO 2 NPs is estimated to be approximately 200 to 300 nm. The PbI 2 nano-plates are roughly 13-17 nm thick and measure 10-50 µm laterally. Fig. S1 SEM images of lead precursor films dropcasted and sintered on ITO (a) PbO 2 particle cluster, (b) PbI 2 hexagonal plates 2 Preparation of organo-halide solutions n-butylammonium iodide (BAI, CAS Number:36945-08-1) was purchased from Great solar materials.The 0.3M of organohalide solution was prepared by dissolving 0.302g of BAI in 5ml of anhydrous 2-propanol in a nitrogen filled glove-box. The solution was stirred for 10 minutes at room temperature. paper, Grade 2 or KimTech Laboratory wipes) to remove the traces of 2D precursor solution, as shown in figure Figure S2 (b). After the filter paper looks dry, the microplates are picked up from the paper using a clean Polydimethylsiloxane (PDMS) piece. The PDMS containing the microplates can be then used to transfer the microplates onto new substrates (like Silicon, glass) as shown in figure S3 (a). If the used precursor solution with microplates are dropcasted on a clean substrate without intermediate filtering step, the traces can be visible on both optical images and XRD as shown in figure S3 (b, c). However the microplates obtained by the direct drop cast method are much larger in spatial dimensions when compared to the filter/transfer step yield.  The secondary nucleation in the case of PbO 2 is typically promoted by the square BAPI crystals that form in solution which eventually drop to the substrate and start growing quickly along the diagonals similar to the primary nucleation that also exhibit diagonal oriented growth as seen in S5.       HPLC analyses were performed on an Agilent Infinity 1260 HPLC with a Chiralpak IA (250 x 4.6 mm, 5 um) column. Eluent was heptane:IPA (70:30) with a flowrate of 0.7 mL/min. Detection at 220 nm (UV detector). Analyses were run for 12 minutes. Retention times were as follows: 4.8 min (the resulting imine), 5.8 min (benzaldehyde reagent), 8.2 min (R-BCA internal standard). Note: Butylamine has no UV absorption and is thus not directly detected by HPLC-UV. The compound R-BCA is a non-reacting internal standard, fully named (R)-2-(benzylideneamino)-2-(2-chlorophenyl)acetamide, that has a crucially non-overlapping retention time and allows for proper quantitation of imine concentrations.

Derivatization Procedure
The derivatization reaction of a sample containing butylamine to the UV-detectable imine by reaction with benzaldehyde proceeds as follows. A 2 mL HPLC vial is filled with 900 µL isopropanol and then 20 µL of the sample is added. Finally, 80 µL of benzaldehyde solution (500 mM in isopropanol, also containing 1 mg/mL R-BCA as an internal standard) is added and the vial is vortexed thoroughly. After 120 minutes, the vial is submitted to the HPLC for analysis. Reference chromatograms showing the different elution times of the resulting imine and the unreacted benzaldehyde reagent are shown in figure S10 (based on a derivatization of a stock solution of butylamine (10 mmol/g in isopropanol)).

Calibration experiment and butylamine concentration determination
In order to calculate the concentration of butylamine in a sample, a calibration curve was constructed. To this end, stock solutions of butylamine in isopropanol were made with concentrations between 0 and 300 mM. The stock solutions were then analyzed as described under 'Derivatization Procedure' and submitted to the HPLC (injection after exactly 120 minutes). The resulting calibration curve is shown in figure S11. The ratio between the imine peak area and the internal standard peak area (R-BCA) is taken as the response (internal standard procedure); example chromatogram shown as figure S12. The calibration is proper (R2 = 0.995) and abides by the relationship,

Derivatization kinetics validation
To validate that the derivatization reaction is completed within 120 minutes after combining the derivatization reagents, the response (ratio between imine and internal standard peak area) was measured as a function of derivatization time. This experiment was performed by injecting the product resulting from the 'Derivatization Procedure' for a 300 mM butylamine reference sample at several points in time. The results of this experiment are shown in figure S13. Fitting the data with standard chemical reaction kinetics equations (conversion 1˘exp(−k ·t)) allows for estimating that completion of the derivatization is beyond 90% after 120 minutes. For samples with lower concentrations of butylamine than 300 mM, the degree of completion is known, from theory, to increase vastly.

Conversion kinetics for perovskite formation
The conversion reaction is performed in the glovebox. In a 7 mL vial, 10.52 mg of PbO 2 or 20.28 mg of PbI 2 powder was weighed and 2 mL of 300 mM butylammonium iodide solution in IPA was added. At various points in time (1, 2, 5, 15, 30, 60 minutes), a 200 µL sample of the slurry was taken and filtered using a 0.2 µm PTFE-syringe filter. The filtrate was stored in a 2 mL HPLC vial and transported outside the glovebox. This sample was then immediately taken for butylamine concentration analysis using the procedure stated under 'Derivatization Procedure'. It was earlier reported that subjecting a butylammonium iodide solution to the derivatization procedure does not lead to significant imine formation. 1 This statement was confirmed in this study as well, by subjecting the butylamine solution to the derivatization procedure. So, as a control, samples of the initial butylammonium iodide solution and of the aged butylammonium iodide solution were also subjected to analysis to provide a baseline correction.
9 Aliquots of PbI 2 reaction with BAI solution collected at different time intervals 10 Role of water during the conversion of PbO 2 to BAPI The role of water during the formation of perovskites has been a controversial topic. 2 Prior works show both detrimental and beneficial effects of water on 2D perovskites, such as degradation 3,4 , formation of pinholes in spincoated films 5 and defect passivation and stable device performance. 6 These different effects are usually attributed to water (either atmospheric or controlled addition to the precursor solution) that is interacting during the formation of the perovskites. In this work, we perform the conversion experiments of both PbO 2 and PbI 2 under dry N 2 atmosphere, where water only appears as a by-product during the in-situ reduction of PbO 2 . We believe that small quantities of water have a catalytic effect on the BAPI formation, as water reacts with BAI precursor to make more iodide species. These additional iodide species further assist in the formation of highly-valent lead iodide complexes, thus promoting the formation of BAPI microplates. To fully proof this more investigation is needed, which is outside the scope of this work.