README FOR DATA RELATED TO PAPER “HIGH THROUGHPUT WORKFLOW FOR SOLID-STATE REACTIONS TO OXIDES RAW DATA CONTENTS PXRD data: CaTiO3 from mixtures of slurries of CaCO3 and TiO2 (Figure 3) BaYxSn1-xO3-x/2 solid solution at three synthesis temperatures: 1200 °C, 1300 °C, 1400 °C (shown in Fig 4) and 1500 °C. Sample code Nominal composition (x) CH402 0.4 CH403 0.5 CH404 0.55 CH405 0.6 CH406 0.65 CH407 0.7 CH408 0.75 CH409 0.8 CH410 0.9 CH430 0 CH445 0.025 CH446 0.05 CH447 0.075 CH448 0.1 CH449 0.2 CH450 0.3 CH451 0.4 CH452 0.55 CH453 0.6 CH454 0.65 CH455 0.7 Nb-Al-P-O phase field at three synthesis temperatures: 1000 °C (shown in Fig 5), 1100 °C, 1200 °C or 1300 °C Sample Nb Al P at% at% at% LA001 81.82 9.09 9.09 LA002 66.67 16.67 16.67 LA003 60.00 20.00 20.00 LA004 50.38 24.81 24.81 LA005 42.86 28.57 28.57 LA006 33.33 33.33 33.33 LA007 25.00 37.50 37.50 LA008 19.76 40.12 40.12 LA009 14.29 42.86 42.86 LA010 11.11 44.44 44.44 LA011 5.26 47.37 47.37 LA012 13.88 77.79 8.33 LA013 27.76 63.91 8.33 LA014 41.64 50.03 8.33 LA015 55.52 36.15 8.33 LA016 69.40 22.27 8.33 LA017 13.33 70.00 16.66 LA018 26.67 56.67 16.66 LA019 40.00 43.34 16.66 LA020 53.33 30.00 16.66 LA021 16.65 58.35 25.00 LA022 33.30 41.70 25.00 LA023 16.67 50.00 33.33 LA024 20.67 35.67 43.65 LA025 33.34 29.34 37.32 LA026 46.01 23.01 30.98 LA027 58.68 16.68 24.64 LA028 71.35 10.35 18.31 LA029 84.02 4.02 11.97 LA030 40.01 20.01 39.98 LA031 50.68 14.68 34.64 LA032 61.35 9.35 29.31 LA033 72.01 4.01 23.97 LA034 55.18 7.18 37.64 LA035 63.18 3.18 33.64 ICP data BaYxSn1-xO3-x/2 solid solution (Figure 4) ADDITIONAL DATA Dispensing of CaCO3 Bowing of resin blocks METHODS Synthesis: Samples were prepared using a high throughput workflow involving dispensing slurries of precursors to mix in desired ratios. The following process steps were used: Wet Milling   Insoluble raw materials (oxides, carbonates, oxalates) were milled in deionised water using zirconia media in a Fritsch Pulverisette 7 planetary mill. An ammonium polyacrylate dispersant was used to reduce suspension viscosity and a water-based acrylic emulsion binder was added to increase the mechanical strength of the discs after drying so that they could be isopressed to increase their density and would remain intact through subsequent calcination and embedding. Each mill generated 15 cm3 of aqueous suspension with a known content of inorganic precursor per unit volume. The solids contents were checked by drying a 1 cm3 sample taken from each mill in an oven at 80 °C overnight and this measured solids content was used to calculate the weight and molarity of inorganic precursor per unit volume of suspension and thus the volumes required in the later dispensing operation.   The suspensions were kept on a disc type sample rotator to prevent sedimentation / separation. A set of three suspensions allows a quaternary oxide system (A-B-C-O) to be addressed. With the current equipment a fourth component could be added and there is no reason in principle why the technique could not be expanded to more complex formulations.  The generation of these suspensions is a conventional process on a small scale which requires significant manual intervention and which is essentially unchanged from normal laboratory practice – hence it is shown as a manual process in Fig. 1a.  2 Wet Mixing  Mixing of the aqueous suspensions was carried out using an Eppendorf epMotion 5075 automated liquid handling station. The starting material suspensions were held on a specially constructed low-profile multi-position magnetic stirrer so that the solids did not settle out during dispensing. The current version of the stirrer has places for four suspensions held in glass vials. The stirrer uses rotating magnetic fields from three sets of stationary coils which are directed to the four stations via soft iron pole pieces. The design minimises the height of the bottom of the vials to increase the volume of material that can be processed, currently to 30 cm3 per suspension. Stirring speed can be varied from 240 – 960 rpm. The conventional PTFE bar stirrers originally used have been replaced with a custom designed 3D printed stirrer with embedded neodymium iron boron magnets. The design has a centre hole which allows the Eppendorf dispensing tips to aspirate from a lower level and reduces the dead volume of liquid which cannot be taken from the vial. The components were dispensed by volume into glass vials (one per formulation) and then the liquid handler mixed each formulation by repeated aspiration and dispensing. The dispensing parameters of the Eppendorf handler were adjusted to allow for the increased density and viscosity of the suspensions compared to water and optimisation of these parameters should allow closer match between nominal and measured compositions.  3 Dispensing  Small aliquots (currently 0.2 cm3) of each mixture were dispensed into specially designed and manufactured vacuum-formed transparent PET trays (Fig. 2a, Fig. S1). The PET is chosen and tested to ensure that no catalyst or filler residues are left when it is subsequently burnt away. Other polymers including biodegradable plant-based materials could be substituted22 provided that they can be vacuum formed and also leave no measurable inorganic residue after combustion (e.g., https://fkur.com/en/applications/trays-from-bioplastics/). Moulds to make the trays were generated by 3D printing using eSun High Temperature resin [https://www.esun3d.com/high-temp-resin-product/] on a Creality LD-002H printer. Tray width is limited by the need to fit them into a laboratory isopress (see below). The initial tray design held 12 wells, each approximately 10 mm in diameter and holding 0.2 cm3 of suspension. Further development is in hand to reduce the well size and thereby increase the number of samples per tray. The trays were held on custom 3D printed holders to locate them in the Eppendorf unit, mimicking part of the layout of a standard 96-well plate. These holders also facilitate subsequent handling. Multiple arrays of compositions can easily be generated for processing under various conditions. For example, in Fig. 2a one set of compositions is being dispensed into four trays for calcination at different temperatures. The same approach can be used to address variations in other process parameters such as compaction pressure, calcination time, calcination atmosphere. From this point on the samples are always handled as sets – there is no one by one handling of individual sample discs.  4 Freeze Drying   The trays, on their holders, were manually transferred to a freezer at -20 °C overnight and then to a Labconco vacuum cabinet attached to a freeze drier. The metal shelves of the cabinet were covered with open cell polymer foam to insulate the trays from the shelves and prevent thawing of the samples during drying. Each liquid aliquot dried to form a porous disc with a flat-bottomed face. This bottom face will subsequently be used for XRD analysis.  5 Isopressing  After drying, the trays of discs were supported in custom made silicone holders (Fig. 2b), covered in 2 mm silicone sheet and vacuum sealed in Nylon bags (Fig. S2) before iso-pressing in an Autoclave Engineers 75 mm diameter wet bag laboratory isopress at 15,000 to 30,000 p.s.i. (105 - 210 MPa) to increase the density and strength of the discs. The silicone holders contained metal inserts to improve the flatness of the bottom faces of the discs. The trays were then inverted onto refractory batts (Fig. 2c) so that the flat faces of the discs were uppermost thus avoiding reaction between this sample surface and the refractory.   6 Calcination  Calcination was carried out in electric box furnaces with a carefully controlled heating cycle which burnt away the PET trays leaving the discs in their original positions on the refractory tray (Fig. 2c). The key element of this cycle was to limit the heating rate to 25 °C/h between 250 °C and 500 °C in order to avoid movement of the discs during the combustion of the PET trays thus preserving the layout and spacing of the samples.  A key element of the process is that the PET trays have carried the samples as sets since they were mixed as suspensions, avoiding any handling of individual samples. As the method develops further and sample numbers increase this becomes ever more important.  7 X-Ray Diffraction  The sets of samples were lifted from the refractories using self-adhesive polymer foam (Technical Foam Services, Corby, Northants, foam grade TFS503 + 4210 SAB adhesive, Fig. 2c) so that the flat faces of the calcined discs were stuck to the foam. They were then surrounded by a 3D printed PLA frame (Fig. S3) which was partially filled with epoxy resin to embed the discs (Fig. 2c). The frame ensures that the top face of the discs will sit parallel to the table in the X ray diffractometer without requiring that the two faces of the cast epoxy be parallel or that the casting is done on a precisely level table (Fig. S4). The one problem we have faced with this system is that the asymmetric filling of the frames can lead to bending as the resin completes its curing. Two solutions to this problem were identified. Curved blocks could be straightened by putting them on a flat metal plate and under a metal weight (> 500 g) in an oven at 125 °C for 1 hour and then cooling with the plate and weight still in place. To avoid having to do this extra step, the hardener to resin ratio was reduced from 0.5:1 to 0.27:1 and the bowing of the frames was reduced to within 0.05 mm of flat (Fig. S5). Further details are given in the Supplementary Information.   The adhesive was released from the cured epoxy using acetone, leaving the samples with their flat surfaces exposed (Fig. 2). Powder X-ray diffraction measurements were carried out in reflection (Bragg-Brentano) mode using a Rigaku SmartLab instrument equipped with a 9 kW rotating anode source providing a parallel beam of Mo radiation K? radiation (0.70930 and 0.71361 Å). The arrays of samples were fixed to an XY table (Fig. 2d) where individual sample positions were selected from an image taken using a camera and scanned in series automatically overnight. For the four samples shown in Figure 4 “measured with LaB6”, samples were removed from the epoxy resin block and remaining resin removed by calcining the samples at 550°C for 4 hours. The powder was then mixed with LaB6 (cubic lattice parameter = 4.156852(2) Å) in an approximate 1: 1 ratio by mass, as an internal standard to confirm the measured lattice parameter. These samples were then measured on a Phillips PANalytical diffractometer with a monochromatic CoK?1 source (? = 1.78901 Å) in Bragg-Brentano geometry. For all diffraction patterns, Pawley refinements were performed using TOPAS academic.23 With the exception of the cubic perovskite phase in BaYxSn1-xO3-x/2 samples, which was allowed to freely refine, lattice parameters for known phases were initially fixed, and all instrumental parameters refined (background, peak shape and zero error), then lattice parameters were released to obtain the refined values.  Chemical analysis of the BaYxSn1-xO3-x/2 samples was performed on a spare set of discs which had not been calcined. Samples from these (~10mg) were digested in Parr vessels at 200 °C for 8 hours using 5 cm3 of 38% HCl. The solutions were diluted to 50 cm3 giving concentrations of Ba, Y and Sn of the order of 10 – 100 ppm which were determined using ICP-OES.