Molecular catalysis at polarized interfaces created by ferroelectric BaTiO3 † †Electronic supplementary information (ESI) available. See DOI: 10.1039/c6sc05032h Click here for additional data file.

Colloidal suspensions of ferroelectric BaTiO3 nanoparticles act as a dispersible polarized interface that can influence the selectivity of non-faradaic reactions.


Index
Page Experimental methods and additional experimental data S2-S5 Figure S1.
Representative TEM image of bulk BaTiO 3 particles S6 Figure S2.
Representative TEM image of ball milled LiNbO 3 particles S7 Figure S7.
Representative TEM image of unmilled SrTiO 3 particles S8 Figure S10. Representative TEM image of ball milled SrTiO 3 particles S8 Figure S11. Representative TEM image of unmilled TiO 2 particles S8 Figure S12. Representative TEM image of ball milled TiO 2 particles S8 Figure S13. Measurements of displacement current for selected ball milled oxides S9 Figure S14. Sample NMR spectrum showing 1, 2 and 3 S10 Experimental data for

General.
Unless otherwise specified, all solvents were obtained from Fisher Chemical (ACS grade) and used as received. PhCF 3 was purchased from Sigma-Aldrich and used as received.
Preparation of reagents and catalysts. Once ball milling was complete, the grinding bowl was opened and the nanoparticle suspension pipetted out into a glass vial for use in reactions. In order to remove residual oxide, grinding bowls and balls were thoroughly rinsed with hexane and sonicated several times in methanol, then briefly soaked in concentrated nitric acid for about 5 minutes and thoroughly washed with deionized water before being used again.
TEM analysis of bulk and ball milled oxides. TEM analysis was performed using an FEI Tecnai G2 F20 X-TWIN Transmission Electron Microscope operating at 200 kV. Samples were prepared by drop-casting dilute suspensions of the various oxides in methanol (for bulk powders) or heptane (for ball milled nanoparticles) onto copper TEM grids. Particle and nanoparticle sizes were estimated by measuring 100 different particles, from which the average size and standard deviation were calculated. Representative TEM images of the samples are shown in Figures S1-S14. Notably, bulk and ball milled LiNbO 3 particles had a flat, plate-like morphology, unlike the more granular appearance of the other oxides. The entry for LiNbO 3 reflects the particle width and not the thickness, which could not be measured by TEM. Attachment of Rh porphyrin 4b to oxide surfaces. Bulk TiO 2 , CaTiO 3 , SrTiO 3 , PbTiO 3 , or BaTiO 3 powder (~100 mg) was soaked overnight in a solution of 4b in CH 2 Cl 2 (~1 mM). Subsequently, the powder was allowed to settle overnight and the catalyst solution carefully pipetted off. The powder was then allowed to dry at room temperature and then baked at 150 °C under an atmosphere of N 2 . The resulting functionalized bulk powder was rinsed twice with CH 2 Cl 2 and sonicated in CH 2 Cl 2 (4  15 minutes) to remove unattached molecules of 4b. Particle suspensions were centrifuged (500 rpm) between each wash to facilitate this process. The functionalized powder was stored at room temperature, or subjected to ball milling (see above) before use.

Oxide
Spontaneous polarization measurements. Spontaneous polarization of nanoparticles was measured through the electric response of cells filled with nanoparticles suspended in a heptane/oleic acid mixture, 2 an insulating non-polar fluid. An electric field applied across the cell aligns the ferroelectric (i.e. polar) nanoparticles such that their dipole moments become parallel to the field. A periodic ac electric field causes rotation of the polar nanoparticles, which gives rise to an ac displacement current. Integration of the positive (or negative) part of the displacement current density over half a period gives a displacement charge density proportional to the spontaneous polarization of the material, P s , and the volume fraction of the material in the suspension. 3 The cells were made using approximately 2×3 cm 2 glass plates, coated with a transparent conductive layer of indium tin oxide (ITO) on one side of each plate (the side adjacent to the suspension); a space of 8 microns between the ITO layers was achieved using non-polar glass spacer beads. Examples of the displacement current traces under a triangular voltage waveform are given in Figure S15 for the ball milled BaTiO 3 , CaTiO 3 , SrTiO 3 , and TiO 2 nanoparticles. For clarity, the background from the cell filled with just heptane/oleic acid mixture (no nanoparticles) was subtracted from the original data.
The data in Figure S1 show both the result of aggregation as a function of nanoparticle concentration ( Figure S15b) and the differences between various titanium oxide-based materials (Figure S15b,c); Figure S15a shows half of the period of the ac symmetric triangular voltage S4 waveform used to measure the displacement currents. With a low enough concentration, where the nanoparticles do not interact with each other, 4 a single feature is observed. With higher concentrations side band structures (wings) form, which are the result of dynamic aggregation and disaggregation of nanoparticles. 4 When comparing the measured displacement current between BaTiO 3 ( Figure S15b) and CaTiO 3 , SrTiO 3 , and TiO 2 nanoparticles (Figure S15c), one must note the concentrations used in each case. Because of the very weak signal in CaTiO 3 , SrTiO 3 , and TiO 2 , the concentrations used to get a measurable signal were far greater than the case of BaTiO 3 . For example, to get a similar displacement current in CaTiO 3 , 100 times more particles were required with respect to ferroelectric BaTiO 3 . The signals for SrTiO 3 and TiO 2 were noticeably smaller, where the displacement current in TiO 2 was about three orders of magnitude weaker than the measured current for BaTiO 3 . Despite the relatively large concentration of the particles used to obtain data for Figure S15c only a central peak was observed, this lack of aggregation would be expected for suspensions that had limited particle interactions, i.e. little Coulombic interaction, 4 2 and 3 were the only major products observed. Results with other solvents besides CH 2 Cl 2 are summarized in Figure 5 in the main text.
NMR analysis. Nuclear magnetic resonance (NMR) data were obtained on a Varian Inova spectrometer at 600 MHz for 1 H nuclei. After the appropriate reaction time had elapsed, reactions or aliquots were quenched with at least 3 volumes of MeCN (Fisher Scientific) and filtered through a 0.22 µm syringe filter to remove flocculated nanoparticles if necessary. The solvent was evaporated and the residue taken up into CD 3 CN (Cambridge Isotope Laboratories) to produce a sample for NMR analysis. Relative amounts of 1, 2 and 3 were obtained through careful integration of the diagnostic peaks (1: 5.59 ppm, s, 1H; 2: 2.31-2.30 ppm, m, 2H; 3: 6.93-6.90 ppm, dt, 1H) of each substance on the 1 H NMR spectrum (600 MHz, CD 3 CN, nt = 200, at = 4 seconds, d1 = 6 seconds) of the solution, from which conversion and product ratio were calculated. The data was found to be consistent with the results from HPLC analysis (see below). Because of the small quantities of products and long acquisition time required per sample, most samples were analyzed only by HPLC. A sample NMR spectrum of the relevant regions for a reaction of 1 with 4a is shown in Figure S16.
HPLC analysis. Reactions were analyzed on an Agilent 1200 HPLC equipped with a Diode Array Detector and an Agilent Eclipse-XDB-C18 reverse phase column. Analysis of 1, 2 and 3 was carried out using the same method as the literature. 1 If the conversion was less than approximately 5%, two columns in series were used instead with a method of twice the length in order to improve peak separation. Peak areas were calculated after fitting to GaussMod curves in OriginPro 8.5.1, and adjusted for extinction coefficient (see below) to provide conversion and product ratios.
Relative extinction coefficients of 1, 2 and 3. The relative extinction coefficients of 1, 2 and 3 at 210 nm were determined previously. 1 Relative extinction coefficients at 210nm: 1: 1.15 2: 1.62 3: 1.00 S6 Figure S1. Representative TEM image Figure S2. Representative TEM image of bulk BaTiO 3 particles. of ball milled BaTiO 3 nanoparticles.    Experimental data for Figure 2. The observed ratios of 2:3 with various concentrations of 1, catalyst 4a and ball milled BaTiO 3 (or other oxide) in CH 2 Cl 2 are summarized below. This data is presented in Figure 2 in the main text. [