Hydrophobic Porous Liquids with Controlled Cavity Size and Physico‐Chemical Properties

Abstract Developing greener hydrometallurgical processes implies offering alternatives to conventional solvents used for liquid‐liquid extraction (LLE) of metals. In this context, it is proposed to substitute the organic phase by a hydrophobic silica‐based porous liquid (PL). Two different sulfonated hollow silica particles (HSPs) are modified with various polyethoxylated fatty amines (EthAs) forming a canopy that provides both the targeted hydrophobicity and liquefying properties. This study shows that these properties can be tuned by varying the number of ethylene oxide units in the EthA: middle‐range molecular weight EthAs allow obtaining a liquid at room temperature, while too short or too long EthA leads to solid particles. Viscosity is also impacted by the density and size of the silica spheres: less viscous PLs are obtained with small low‐density spheres, while for larger spheres (c.a. 200 nm) the density has a less significant impact on viscosity. According to this approach, hydrophobic PLs are successfully synthesized. When contacted with an aqueous phase, the most hydrophobic PLs obtained allow a subsequent phase separation. Preliminary extraction tests on three rare earth elements have further shown that functionalization of the PL is necessary to observe metal extraction.

For bHSPs, there were two populations of spheres due to the use of a soft template: around 40% had an external diameter of around 149 nm and 60% had an external diameter of around 278 nm in average.obtained with the software SasView over the range [0.00482;0.853]Å -1 .The model used for fitting was a core-shell sphere structure with a sticky hardsphere behavior.The parameters and their value are listed in the Table S1.
Table S2.Parameters for the fit of sHSPs in SasView (see Figure S3).
For bHSPs: The adsorbed volume of gas at P/P0 = 0.87 is 0.78 cm 3 /g in this case (it excludes the interparticle void even though it cannot be totally uncorrelated from the large hollow core of bHSPs).

S3. Characterization of PLs by TEM, SAXS, TGA and DSC experiments
Table S3.Yields of some PLs syntheses at each step.The molar ratio Amine/SiO2 displayed a non-linear behavior with the molecular weight of the amine, with an increase frome 0.32 to 0.37 between PL-sHSPs-5 and PL-sHSPs-11, a further decrease up to 0.22 for PL-sHSPs-50, and a final significant increase for PL-sHSPs-92 with a ratio exceeding 6 (Table S5).The molar ratio SIT/SiO2 increased from 0.31 (PL-sHSPs-5) to 0.57 (PL-sHSPs-22) but decreased to 0.4 for PL-sHSPs-50.It could not be determined for PL-sHSPs-92 (no significant weight loss attributed to the SIT on the TGA trace) but it is expected to be high.

Colum
Additionally, the subequivalent Amine/SIT ratios found based on TGA measurements could mean that the pH-metry was not a sufficient method to follow the real equivalence during the neutralization of the sulfonate groups.
The ratios for PLs with EthA-22 are also given in the Table S6.
Table S5.Weight and molar ratios of Amine, SiO2 and SIT contained in PL-sHSPs with different amines.The FTIR trace of PL-sHSPs-92 could not be normalized since the proportion of amine with respect to the organosilicon was too high to see a well-defined peak at 690 cm -1 .As expected, a non-Newtonian behavior was observed for the liquids along with high viscosities (Figure S9), and the curves could be modelled by a Bingham fitting (with an exception for the viscosity at 20°C that did not follow the same behavior).The values of viscosities obtained with this model (Figure S10) were close to the one obtained at 10 s -1 for each PL (corresponding to a shear rate at the beginning of the zone of Newtonian behavior).

Figure S5 .
Figure S5.Structural properties of PLs with respect to the properties of the silica spheres and of the free amines.a) SiNPs and PL-SiNPs, b) sHSPs and PL-sHSPs, c) bHSPs and PL-bHSPs, and d) pure amines.

Figure
Figure S6.a) Variation of TGA profiles of PLs according to the overall PEG lengths of the EthA involved for each PL-sHSPs, b) Variation of TGA profiles of PLs depending on the nature of the silica core modified with EthA-22.

Figure
Figure S7.a) Chemical composition of PLs-sHSPs with different amine chains lengths, b) chemical composition of PLs grafted with EthA-22 on different silica spheres.

Figure S9 .Figure S10 .
Figure S9.DSC profiles of all PL-sHSPs.The arrows indicate the melting peaks, and the values correspond to the midpoint.For PL-sHSPs-5, no melting was observed, but a glass transition was clearly visible at 85.6°C.

Figure S11 .
Figure S11.Viscosity of PL-sHSPs-22 at different temperatures (log-log scale), recorded at 10 s -1 (black squares) or extracted from a Bingham fitting of the measurements (blue triangles).
Figure S12.a) TGA of PL-sHSPs-11 before (black) and after (red) contact with water.b) TGA of the aqueous phase after contact of PL-sHSPs-22 with water in a w:w 1:20 PL:water ratio.

Table S1 .
Diameters and standard deviations of SiNPs, sHSPs and bHSPs deduced from TEM and SAXS measurements.

Table S4 .
Impact of the global PEGs length of the EthA on the inter-particle distance measured from the centers of the particles.

Table S6 .
Weight and molar ratios of Amine, SiO2 and SIT contained in PLs with EthA-22 and different silica spheres.