Lead-free double halide perovskites AIMIMIIIX6 (A,M – metal cations, X – halide anions) offer unprecedented variability of structures and compositions, with all four positions variable independently resulting in many thousands of possible compounds for the same set of constituent elements. A detailed screening of the properties of these huge families of compounds requires application of a combination of high-throughput synthetic (HTP) approaches with HTP characterization and HTP testing for possible applications related to light conversion and emission.

We have developed a “green” protocol for a HTP synthesis of brightly luminescent lead-free microcrystalline Cs2AgxNa1-xBiyIn1-yCl6 (CANBIC) perovskite phosphors absorbing strongly in UV range and emitting broadband photoluminescence (PL) in the visible range with PL quantum yields reaching 98±2% for specific compositions with x = 0.35-0.40 and y = 0.01-0.02. The CANBIC perovskites revealed high stability of spectral properties during many months of ambient storage, thermal stability at open air till 200-250 oC and photochemical stability under UV illumination, making these compounds highly promising for applications in luminescent solar light concentrators and down-shifters. Our HTP approach to CANBIC perovskites includes robot-assisted automated synthesis with any desirable x and y steps combined with HTP spectral charcterization which includes absorption, PL, PL excitation, and Raman spectroscopy as well as time-resolved PL characterization, allowing many hundreds of samples to be produced and characterized in a single working session.

This HTP approach was expanded to other CANBIC-like perovskites, produced either by substituting InIII with SbIII, or by substituting BiIII with FeIII. In the first case we developed a HTP protocol for the synthesis of Cs2AgxNa1-xBiySb1-yCl6 (CANBSC) perovskites with x and y fractions varied independently from 0 to 1. The CANBSC perovskites can be converted into Cs2AgxNa1-xBiySb1-yBr6 (CANBSB) and Cs3Bi2ySb2(1-y)I9 (CBSI) perovskites by a single-step anionic exchange with NaBr and NaI, respectively. These compounds reveal a band-bowing effect, the bandgaps of mixed perovskites being typically lower than those of corresponding individual BiIII- and SbIII-based perovskites. For CANBSC and CANBSB compounds the band-bowing effect was found to increase with an increase of the silver fraction. The lowest bandgaps reached were 2.47 eV for CANBSB and 2.08 eV for CBSI both at x = y = 0.50.

We found that BiIII in CANBIC perovskite can be partially or completely substituted by FeIII resulting in the latter case in Cs2AgxNa1-xFeyIn1-yCl6 (CANFIC) perovskites showing strong absorbance in the visible spectral range and a high environmental stability. A special feature of CANFIC compounds is a structure-directing influence of In on the formation of perovskite phase. While for y = 1.00 only a multi-phase mixture can be obtained in our conditions, introduction of mere 1% InIII steers the precipitation exclusively to the formation of cubic perovskite phase with no other phases detectable. The lowest bandgap reached so far is ca. 2.0 eV for x = 1.00 and y = 0.95-0.99. Anionic exchange of CANFIC did not result in stable single-phase bromide and iodide compounds. A similar structure-directing effect was also found for BiIII and SbIII introduced instead of InIII. Our preliminary results show the feasibility of introducing simultaneously three MIII metals, for example, BiIII, InIII, and FeIII, into the chloride perovskite while preserving the single-phase solid-solution character of the final products, opening broad possibilities for the compositional design of the optical properties. The unique variability of the MIII site of these double perovskites will be addressed in future using the combined HTP approach to the synthesis and characterization developed earlier for CANBIC perovskite phosphors.