Silicon Oxycarbide (SiOC)-Supported Ionic Liquids: Heterogeneous Catalysts for Cyclic Carbonate Formation

Silicon oxycarbides (SiOCs) impregnated with tetrabutylammonium halides (TBAX) were investigated as an alternative to silica-based supported ionic liquid phases for the production of bio-based cyclic carbonates derived from limonene and linseed oil. The support materials and the supported ionic liquid phases (SILPs) were characterized via Fourier transform infrared spectroscopy, thermogravimetric analysis, nitrogen adsorption, X-ray photoelectron spectroscopy, microscopy, and solvent adsorption. The silicon oxycarbide supports were pyrolyzed at 300–900 °C prior to being coated with different tetrabutylammonium halides and further used as heterogeneous catalysts for the formation of cyclic carbonates in batch mode. Excellent selectivities of 97–100% and yields of 53–62% were obtained with tetrabutylammonium chloride supported on the silicon oxycarbides. For comparison, the catalytic performance of commonly employed silica-supported ionic liquids was investigated under the same conditions. The silica-supported species triggered the formation of a diol as a byproduct, leading to a lower selectivity of 87% and a lower yield of 48%. Ultimately, macroporous monolithic SiOC-SILPs with suitable permeability characteristics (k1 = 10–11 m2) were produced via photopolymerization-assisted solidification templating and applied for the selective and continuous production of limonene carbonate with supercritical carbon dioxide as the reagent and sole solvent. Constant product output over 48 h without concurrent catalyst leaching was achieved.


S.1.1 Powdered Silicon Oxycarbide Supports 7a for Batch Experiments
For batch experiments using silicon oxycarbide powder 7a, a preceramic polymer solution 4 with 30 wt.% functionalized polysiloxane and 70 wt.%tert-butyl alcohol 3 was used.The ceramic yield determined after pyrolysis (N=4, heating 1 K min -1 , 1 h dwell time) is in reasonable agreement with the residual mass determined during thermogravimetric analysis, as shown in Figure S1.
Figure S1: Thermogravimetric analysis in argon flow to study the pyrolytic conversion of the green body 6a to silicon oxycarbide 7a and obtained ceramic yields upon pyrolysis at 300, 500, 700, and 900 °C in argon flow.

S.1.2 Silicon Oxycarbide Monoliths 7b for Continuous-Flow Experiments
For continuous experiments using cylindrical, monolithic silicon oxycarbide 7b, a preceramic polymer solution 4 with 20 wt.% functionalized polysiloxane and 80 wt.% tert-butyl alcohol 3 was chosen to combine high porosity and sufficient structural strength.As the pore structure of preliminary samples derived from a preceramic solution 4 using only 10 wt.% functionalized polysiloxane in 90 wt.% tert-butyl alcohol 3 collapsed upon freeze-drying and pyrolytic conversion (Figure S2), a polymer content of 20 wt.% in solution 4 was selected to result in monoliths of maximum porosity achievable via this method.20 wt.% polymer content in the solution 4 was chosen for further experiments to combine high porosity and high permeability while providing enough structural strength.Supporting Information

S4
Aluminum molds with an inner diameter of 12 mm were used to generate cylindrical silicon oxycarbide monoliths 7b with a final diameter of 8.5 mm, thus seamlessly fitting in a column with a 9 mm inner diameter.The linear shrinkage during freeze-drying and pyrolytic conversion was approximately 5 % and 27-29 %, respectively (N=30).The monoliths exhibited a bulk density of 0.4 g cm -3 and an apparent porosity of 81 % (water immersion method, N=4).
Cylindrical samples of a length of 6 cm were obtained, a layer of 1.5 mm from the top and

S.2.2 Permeability Measurement of Silicon Oxycarbide Monoliths
Measurements were performed using filtered compressed air as permeating fluid.Shrinking tubes (RS PRO) were used to seal the cylinders in flow direction, 3D printed rings were used to protect sample edges.The permeating gas flow Q was recorded as function of the pressure drop (p1-p2).Permeated area A and the sample height L were derived from the sample dimensions.The inlet overpressure p1 was varied between 0.2 and 2 bar.Permeability constants (Darcian k1, non-Darcian k2) were determined using Forchheimer's equation for compressible fluids (Formula S1) using least-square fits.The viscosity of air μ was derived from the Sutherland equation (Formula S2) and air density ρ was derived from the ideal gas law. 1

S.2.6 X-Ray Photoelectron Spectroscopy (XPS)
All spectra were charge corrected to adventitious carbon at 284.8 eV according to Biesinger       a Samples were mounted on indium foil and fixed to the stage using double sided carbon tape.Accuracy of XPS measurements falls within 10-20 at% (especially for oxygen and carbon due to adventitious carbon and oxygen) Detection limit in the recorded survey spectra used for quantification lies between 0.1-1 at%.

S.3.1 Limonene Carbonate 15
NMR yields and conversions were determined according to a modification of a protocol 3 previously published by our group.
For the batch reactions running for 5 h, NMR spectra at t = 0 h (before reaction) and at t = 5 h (after reaction) were recorded (Figure S15).Calculations of yields and conversions were performed according to Formulas S4-S9 based on integrals of the protons next to the epoxy and carbonate moiety.Naphthalene was used as an internal standard (δ = 7.82 and 7.45 ppm).
Conversions and, therefore, selectivities (given as ratio of yield and conversion) for flow experiments could not be determined due to the volatility of the starting material and, therefore, partial evaporation of the starting material during the release of CO2 via the backpressure regulator.• 100

S.3.2 Linseed Oil Carbonates 18
For the determination of the conversion of epoxidized linseed oil 17, NMR spectra at t = 0 h (before reaction) and at t = 5 h (after reaction) were recorded (Figure S16).Calculation of the conversions was performed according to Formula S10 based on integrals of the protons next to the epoxy moiety (δ = 3.21 -2.82 ppm).The signal of the α-CH2 of the carbonyl group in the backbone of linseed oil, not participating in the reaction, was used as internal standard (δ = 2.29 ppm).9), other conditions according to footnote a; c reported as sum of cis and trans isomers, detailed information about the determination of yield is summarized in ESI chapter S.3.1 and the experimental part of the main manuscript; Conversions and, therefore, selectivities (ratio of yield and conversion) could not be determined due to partial evaporation of limonene oxide 11 during the release of CO2 via the back-pressure regulator.d limit of detection: 0.1 mg, ≤ 0.1% of total amount of TBAC 8.
bottom part was removed to yield open porosity and parallel bases.The bulk was cut in specimens to stack columns of 220 mm length, requiring a total of 14 specimens.The obtained monolithic specimens had a diameter of 8.4 ± 0.1 mm (N=14).

Figure S15 :
Figure S15: Determination of NMR yields of limonene carbonate 15 via the comparison of the integrals of recorded 1 H-NMR spectra at t = 0 h and t = 5 h.

Figure S16 :
Figure S16: Determination of NMR conversions of epoxidized linseed oil 17 via the comparison of the integrals of recorded 1 H-NMR spectra at t = 0 h and t = 5 h.
5 mm from the bottom and top part of the cylinders (5-6 cm) was removed and the samples were cut (Struers Minitom, diamond cut-off wheel, 150 rpm) to monolith specimen 7b of about 15-16 mm.TBAC 8 (for SILP 1b: 20 wt%, for SILP 2b: 35 wt% dried under high vacuum for 1 d) was dissolved in 100 mL of dry MeOH and monoliths (220 mm as 15 mm pieces, approx.4.9 g, 80 wt%) were added.The suspension was treated in an ultrasonic bath for 2 h at 40 °C.Solvent was removed under vacuo and monolithic SiOC-SILP 1b-2b was further dried under high vacuum for 1 d.Impregnated monoliths were separated from the remaining ionic liquid and the catalyst loading was determined gravimetrically (SILP 1b: 20 wt% of TBAC 8; SILP 2b: 35 wt% of TBAC 8).

Table S2 :
Results of quantification of elements in supports and SILPs via XPS. a

Table S2 :
Results of quantification of supports and SILPs via XPS. a

Table S4 :
Catalyst screening for the formation of limonene carbonate 15 in batch mode a S.

Table S6 :
Catalyst screening for the formation of linseed oil carbonates 18 in batch mode.a