Fast A‐Site Cation Cross‐Exchange at Room Temperature: Single‐to Double‐ and Triple‐Cation Halide Perovskite Nanocrystals

Abstract We report here fast A‐site cation cross‐exchange between APbX3 perovskite nanocrystals (NCs) made of different A‐cations (Cs (cesium), FA (formamidinium), and MA (methylammonium)) at room temperature. Surprisingly, the A‐cation cross‐exchange proceeds as fast as the halide (X=Cl, Br, or I) exchange with the help of free A‐oleate complexes present in the freshly prepared colloidal perovskite NC solutions. This enabled the preparation of double (MACs, MAFA, CsFA)‐ and triple (MACsFA)‐cation perovskite NCs with an optical band gap that is finely tunable by their A‐site composition. The optical spectroscopy together with structural analysis using XRD and atomically resolved high‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) and integrated differential phase contrast (iDPC) STEM indicates the homogeneous distribution of different cations in the mixed perovskite NC lattice. Unlike halide ions, the A‐cations do not phase‐segregate under light illumination.

CsPbBr3 and 150 o C for CsPbI3, and then, 400 µL of the pre-heated Cs-OL stock solution was swiftly injected into it under vigorous stirring (1,000 rpm). After 5 s, the vial cooled in an ice-water bath to quench the reaction.
Subsequently, the thus obtained colloidal dispersion was purified by centrifugation at 8000 rpm for 10 min. Then, the supernatant was discarded to remove the unreacted precursors and ligands and the pellet was redispersed in 5 mL of toluene. The colloidal dispersion was centrifuged again at 5,000 rpm for 8 min to remove largest particles in the sediment.

Synthesis of FAPbBr3 and FAPbI3:
In a typical synthesis, 6 mL of the corresponding PbX2 precursor solution in a 20 mL glass vial was heated on a hot-plate until the temperature of the precursor solution reaches 175 ºC in the case of FAPbBr3 and 150 o C for FAPbI3, then, 2.5 mL of the pre-heated FA-OL stock solution was swiftly injected into it under vigorous stirring (1,000 rpm). After 5 s, the vial was removed from the hot-plate and placed it in an ice-water bath to quench the reaction. FAPbBr3The crude solution was purified following the same procedure as described for CsPbBr3 and FAPbBr3 NCs. FAPbI3 crude solution was purified also by centrifugation (8000 rpm, 10 min). Then, the pellet was redispersed in toluene and the colloidal solution was centrifuged again (3500 rpm, 5 min) to remove the largest particles in the sediment.
On the other hand, 10 mL of toluene in a 20 mL vial was heated on a hot-plate until rise 40 ºC under stirring (1,000 rpm), then, 1 mL of the DMF mixture was swiftly injected into it. The solution color turned yellow immediately and it was removed from the hot-plate and placed it in an ice-water bath to quench the reaction.
Before purification, it was necessary to induce aggregation by mixing equal volumes of the MAPbBr3 crude solution and methyl acetate. After that, the MAPbBr3 NCs were purified following the same procedure as described for CsPbX3 NCs.
Synthesis of FAxCs1-xPbI3, FAxCs1-xPbBr3, MAxFAyCs1-x-yPbBr3 NCs by cation exchange: In a typical synthesis, colloidal solutions of CsPbX3, FAPbX3 and MAPbBr3 NCs dispersed in toluene were mixed in different ratios to produce the desired Cs/FA/MA stoichiometry. Before mixing, the absorption spectra of the individual samples were measured, and the concentration was adjusted so that each solution had a similar optical density near the band edge. The mixture was kept stirring (500 rpm) for 5 min at room temperature and the reaction was monitored by photoluminescence spectroscopy.
Kinetics studies: Kinetic studies of different parameters and exchange reactions were carried out in the same way. In a cation or halide exchange reaction, different volumes of colloidal dispersions of the corresponding NCs synthesized previously were mixed under stirring (500 rpm). Before starting the reaction, the absorption spectra of the mono cation and mono halide samples were measured, and the concentration was adjusted so that each solution had a similar optical density neat the band edge. The reaction was monitored by photoluminescence spectroscopy along time. The NCs and measurement parameters employed for each study are described in table 1. For ligands effect study in cation exchange reaction, the colloidal solution in toluene was washed with MeOAc employing proportions 1:1. Then, the mixture was centrifuged (11000 rpm, 15 min) and the sediment was redispersed in toluene. For rich ligand environment conditions, 100 µL of the corresponding ligand-toluene solutions were added after mixing the NCs.  The FAPbBr3 NCs used for the control experiment reported in Figure S9 were prepared by using the above- The lattice parameter analysis was performed using StatSTEM. [3] The location of all atomic columns was determined by fitting Gaussian functions to these columns. Time-resolved photoluminescence spectra were obtained using a FluoroMax-3 (Horiba Jobin Yvon) spectrophotometer. The PL decay traces were measured by exciting the samples at 287 nm using <1.2 ns laser diode. For calibration of the system to extract the absolute number of photons at each point, a two-step process was used for each objective lens used. First, a calibrated white light lamp from Ocean Optics was coupled into an integrating sphere. The objective lens was also coupled into the integrating sphere. Comparing the measured spectrum of the lamp at each point to the calibrated spectrum gives the relative sensitivity of the system both spectrally and spatially. Second, a 657-nm laser was coupled directly to the microscope by an optical fibre. The power of the laser was measured precisely at the output of the fibre using a power meter before coupling to the objective lens. Measuring the laser on the system allows direct conversion between number of counts and photons at this wavelength. Combining this absolute calibration with the relative calibration from the calibrated white light lamp and integrating sphere allows absolute calibration across the spectrum at each point of the sample. For all measurements in this experiment, 100 mW/cm2 power density is used to mimic the power of one sun to investigate the PL stability of the NC films.        Figure S8. Comparison of XRD patterns of CsPbBr3, MAPbBr3, FAPbBr3, CsxFA1-xPbBr3 and CsxMAyFA1-x-yPbBr3 NCs. Figure S9. PL intensity maps of CsMA perovskite NC films at 0 min (a) and 40 min (b) of continuous illumination and their corresponding fitted peak wavelength maps (c,d). Scale bar is 5 μm. The broadening of PL peak is likely caused by the differences in blue-shifting speeds of big or small clusters of NCs on the film. Small clusters of NCs are shifting faster than bigger clusters because each NC received higher photon injections (so cation evaporates faster). With a smoother film, it is highly likely that the overall PL spectrum will be blue shifting uniformly without being broadened.   NCs with respect to their individual PL spectra. The multiple peaks in the PL spectra indicate that the cation cross-exchange didn't complete even after 1440 min, meaning that the DDABr protection on the CsPbBr3 NC surface prevents the cation cross-exchange.