Harris et al. ESI: Ultra-smooth mineral films 1/6 Electronic Supplementary Information Ultra-smooth and space-filling mineral films generated via particle accretion processes Joe Harris, Ingo P. Mey, Corinna F. Böhm, Thi Thanh Huyen Trinh, Simon Leupold, Carsten Prinz, Philip Trippal, Ralf Palmisano and Stephan E. Wolf* * To whom correspondence should be addressed: stephan.e.wolf@fau.de Experimental Details Materials Sodium poly(acrylate) Mw 5100 (Fluka), calcium chloride 1.0 M solution (Carl Roth), magnesium chloride hexa-hydrate (Sigma Aldrich), 2-(N-morpholino) ethanesulfonic acid and fluorescein glycine amide (Thermofisher Scientific), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (Sigma Aldrich), hydrochloric acid (Carl Roth) and ammonium carbonate (Carl Roth) were used as received without further purification. Glassware was cleaned with deionised water, immersed in 1 M HCl for 5 minutes and subsequently rinsed with deionised water before air drying. Film preparation by slow-diffusion mineralization In a typical experiment, film precipitation was performed in a closed 21 litre volume desiccator at 25 °C for 42 hours.[1] A 9 cm diameter petri dish was used as a crystallisation vessel. This contained 25 ml of a 10 mM CaCl2, 0 to 15 mM MgCl2·6H2O, 200 μg ml-1 PAA solution. The petri dish was covered with clingfilm which was punctured with one 1 mm diameter needle hole, this was sealed in a desiccator with two glass vials (10 ml) each containing 3g of freshly crushed ammonium carbonate powder. Each vial was covered with parafilm punctured with a single 1 mm diameter needle hole placed at the bottom of the desiccator. The number of glass vials bearing ammonium carbonate or, alternatively, the number of needle holes in the vial(s) affect the escape rate of carbon dioxide and ammonia released from the decomposing ammonium carbonate powder to the desiccator’s gas phase This provides a very simple but efficient control over the rate by which the composition of the gas phase is saturated with respect to carbon dioxide and ammonia. This, in turn, affects the rate of ammonium carbonate uptake in the reaction solution, eventually controlling the pH and supersaturation profile of the reaction. Alternatively, the desiccator size can be altered; see Harris et al. (2017) for more details [1]. The films were recovered from solution, rinsed with ethanol, and air dried. The defect- free films could be easily extracted from the interface using a microscope coverslip; optical microscopy under cross-polarization, FT-IR, and XRD analyses indicated absence of crystalline polymorphs in case of Mg-doped films. Preparation of fluorescein glycine amide-tagged PAA A 2.5 mg ml -1 sodium poly(acrylate) was prepared in a 100 mM MES buffer at pH 4.9. A 4 ml volume of this solution was used to dissolve 1 mg fluorescein glycine amide. A volume of 0.1 ml of EDC solution (0.19 M) in MES buffer at pH 4.9 was added to the sodium poly(acrylate)/fluorescein glycine amide solution. The reaction was stirred for 4 hours in the dark to yield fluorescein-labelled poly(acrylate). Unreacted EDC and fluorescein glycine amide, and buffer molecules were removed by dialysis (Mw cut off 3500 Da) against deionized water for 48 hours in the dark with regular changes Electronic Supplementary Material (ESI) for Nanoscale Horizons. This journal is © The Royal Society of Chemistry 2019