Aqueous Biphasic Dye‐Sensitized Photosynthesis Cells for TEMPO‐Based Oxidation of Glycerol

Abstract This work reports an aqueous dye‐sensitized photoelectrochemical cell (DSPEC) capable of oxidizing glycerol (an archetypical biobased compound) coupled with H2 production. We employed a mesoporous TiO2 photoanode sensitized with the high potential thienopyrroledione‐based dye AP11, encased in an acetonitrile‐based redox‐gel that protects the photoanode from degradation by aqueous electrolytes. The use of the gel creates a biphasic system with an interface at the organic (gel) electrode and aqueous anolyte. Embedded in the acetonitrile gel is 2,2,6,6‐tetramethylpiperidine‐1‐oxyl (TEMPO), acting as both a redox‐mediator and a catalyst for oxidative transformations. Upon oxidation of TEMPO by the photoexcited dye, the in situ generated TEMPO+ shuttles through the gel to the acetonitrile–aqueous interface, where it acts as an oxidant for the selective conversion of glycerol to glyceraldehyde. The introduction of the redox‐gel layer affords a 10‐fold increase in the conversion of glycerol compared to the purely aqueous system. Our redox‐gel protected photoanode yielded a stable photocurrent over 48 hours of continuous operation, demonstrating that this DSPEC is compatible with alkaline aqueous reactions.


Experimental Procedures
Materials and devices Preparation of redox gel based on acetonitrile containing PVDF-HFP polymer. Procedure for gas chromatography samples using a TMS-derivatization method. Standard reaction setup chemical oxidation reactions using TEMPO(BF4) as a chemical oxidant A standard setup for distribution experiments of TEMPO(BF4) and glycerol in a biphasic system Cleaning of the FTO electrodes Preparation of the photoanode FTO|TiO 2 Preparation of the platinum counter electrodes Construction of dye-sensitized solar cells (DSSCs) Illumination setup for of the dye-sensitized photoelectrochemical cells (DSPECs) Quantification of water in redox-gel. TEMPO 0/+ , glycerol, and glyceraldehyde exchange and product conversion between (organic)gel layer and aqueous layer. Tables  Table S1: Determined gas chromatography retention times for all used compounds Figure S1. Gas chromatography calibration curve for glycerol Figure S2. Gas chromatography calibration curve for glyceraldehyde Figure S3. Gas chromatography calibration curve for TEMPO Figure S4. Gas chromatography calibration curve for H 2 in 2 mL air Table S2. Images of setup for chemical oxidation reactions with TEMPO(BF4) in gel biphasic Table S3. Overview chemical oxidation reactions of TEMPO(BF4) and glycerol in gel biphasic systems Table S4. Overview of biphasic exchange studies of TEMPO(BF4) in gel biphasic systems: Table S5. Overview of biphasic exchange studies glycerol in gel biphasic systems Figure S5. Schematic overview of the assembly of a DSSC with Teflon spacer Figure S6. Reference spectrum of used illumination source Figure S7A. Initial chopped light experiments of the various DSPEC systems. Figure S7B. Overview of the photogenerated current density in all DSPEC measurements during 24 hours. Figure S8. Graph of TEMPO present in the aqueous layer in DSPEC experiments obtained over 24 hours Table S6. Increase in water content of gel during 48 hours DSPEC measurements Figure S9A. H 2 measurement in DSPEC experiments over a period of 7 hours. Figure S9B. H 2 and 1/2 electrons produced in DSPEC experiments over a period of 7 hours. Figure S10. Overview of the photogenerated current density in H 2 DSPEC measurements during 7 hours. Figure S11A. GC Chromatogram for the sampling of long-term DSPEC experiments measured over 48 hours. Figure S11B) GC quantification of substrate to product formation corresponding to Figure 11A Figure S12A. Overview of the photogenerated current density produced in DSPECs during 48 hours Figure S12B. Graph of glyceraldehyde and 1/2 electrons produced in DSPECs during 48 hours. Figure S13A. TEMPO in the aqueous layer in two long-term DSPEC experiments measured over 48 hours. Figure S13B. Glycerol and glyceraldehyde in the aqueous layer DSPEC experiments measured over 48 hours Table S7. Overview of biphasic exchange studies and oxidation studies with TEMPO(BF4) of benzyl alcohol Table S8. Overview of biphasic exchange studies and oxidation studies with TEMPO(BF4) of HMF Table S9. Dye leaching experiment with ø 0.5 cm FTO|TiO 2 |AP11 in alkaline or organic anolytes. Table S10. Images of photoanodes used in DSPEC experiments

Standard reaction setup chemical oxidation reactions using TEMPO(BF4) as a chemical oxidant
The standard setup for the biphasic chemical oxidation experiments was a two-layered system of gel/aqueous solution ratios of 1:3, 2:3, and 3:3 (in v/v 100 µL) (images of experimental setup Table S2, . A fully mixed system (400 µL of 0.15 M TEMPO(BF4), 0.1 M glycerol, Table S2, Entry 1) was used as a control. The ratios oxidant and substrate were kept similar, yielding a theoretical maximum conversion of ~80% (Table S3) since the oxidation from glycerol to glyceraldehyde is a two-electron process while TEMPO(BF4) is a one-electron oxidant. As a control, bilayer systems lacking PVDF-HFP were made to analyze the organic layer. The redox-gel contained TEMPO(BF4) (0.15 M) and was obtained as described before. The gel (100 µL, 200 µL, and 300 µL) was added to a vial after which 300 µL glycerol (0.03 M, 0.05 M or 0.1 M in aqueous solution sat. NaCl, NaHCO3 pH 8.3) was slowly added on top of the redox-gel. The equivalent oxidant to the substrate was kept 1.5:1. After 16 hours, aliquots of the aqueous layer and/or the organic layer were obtained and derivatized as described above, prior to GC analysis to study the effect of gel thickness on glycerol oxidation.

A standard setup for distribution experiments of TEMPO(BF4) and glycerol in a biphasic system
The standard setup for distribution experiments was the two-layered system of gel/aqueous solution ratios of 1:3, 2:3, and 3:3 (in v/v 100 µL) (images of an experimental setup v/v 2:3 in Table S2, Entry 2

Cleaning of the FTO electrodes
The Fluorine doped Tin Oxide (FTO) electrodes (Solaronix, 2.2 mm, 15 Ω sq -1 ) were scrubbed with Deconex and rinsed with hot water, followed by wiping with an acetone-soaked tissue and air drying. The FTO electrodes were then rinsed with acetone, toluene, and ethanol and left to air dry in between. The FTO electrodes were then placed inside a glass container in a solution of a teaspoon of Deconex in Milli-Q and sonicated for 30 minutes. The sonication procedure was repeated with Milli-Q, and then ethanol and the electrodes s were left to dry, after which the electrodes were treated with a UV-ozone generator (Ultra-Violet Products, PR-100) for a minimum of 30 minutes.

Preparation of the photoanode FTO|TiO2
A clean FTO plate (Solaronix, 2.2 mm, 15 Ω sq -1 ) (10×5 cm) was added to a 40 mM TiCl 4 solution in Milli-Q (100 mL) and placed in the oven at 70°C for 30 min to create a TiO 2 blocking layer. The FTO plate was consecutively rinsed with Milli-Q, then ethanol and air-dried.

Illumination setup for of the dye-sensitized photoelectrochemical cells (DSPECs)
A three-electrode DSPEC system was composed of two compartments (working and counter) separated by a Nafion-117 membrane However, since we have demonstrated H 2 production in MeCN with 1.0 M AcOH in our previous studies, [10,11] continued the use of a MeCN-based catholyte as improving the photoanode environment is the main focus of this work. Chronoamperometric and choppedlight measurements were performed on a P211 potentiostat (Zahner) without stirring for 23 hours while illuminating the photoanode (illuminated area 0.64 cm 2 ) using a LED white-light source (Zahner, TLS3, 100 mW cm -2 ). A bias potential of 0.1 vs. Ag/AgCl was applied on the WE (P211 potentiostat, Zahner), and determined using preliminary chopped-light experiments, which can be found in Figure S7. To monitor the reaction, aliquots of the aqueous layer were obtained and were TMS-derivatized and analyzed with GC as described before. The integration of half the photocurrent determines the number of electrons to account for two electrons needed per oxidation reaction. The conversion can be enhanced a prolonging the illumination time. However, it is noteworthy that extending the reaction time can cause over-oxidation of the products with sufficient oxidized TEMPO present. Examples of glycerol to ketomalonic acid are known using TEMPO + as an oxidant. [12] There seem to be some photocurrent fluctuations before a steady state of 0.2 mA cm -

Quantification of water in redox-gel.
10 mg of redox-gel unused and used in DSPEC were added to 1 mL of freshly distilled dry toluene. The dry toluene was allowed to take up water from the redox gel over the course of 3 hours. The water in the toluene was then determined by using the Karl Fisher technique on an 831 KF Coulometer (Metrohm) with Hydranal solution.
TEMPO 0/+ , glycerol, and glyceraldehyde exchange and product conversion between (organic)gel layer and aqueous layer.
The desirable separation of the organic and aqueous phase and the gel stability in a mixed system were verified, both needed for longterm photoanode protection. The two-phased system continued to be well separated, and the gel stayed intact over multiple days, as seen in Table S2. The distribution of the compounds throughout the biphasic system is examined next and shown in Table S1. A high affinity of the redox-mediating catalyst TEMPO 0/+ for the redox-gel layer is of great importance since sufficient amounts of the redox mediator near the photoanode is necessary for DSPEC to complete photocycles. In contrast, preference of the product and the substrate for the aqueous layer are desired for product retrievement. TEMPO 0/+ shows a high affinity (~95%) for the acetonitrile-gel throughout different layer thicknesses due to the over-saturated aqueous solution, while both glycerol (~85%) and glyceraldehyde (97%) prefer the aqueous layer. The strong separation of TEMPO + to glycerol in the gel-aqueous system can influence the oxidation reaction considering the two compounds can only react with each other at the interface. Substrate conversion was studied in a biphasic system and compared to a thoroughly mixed reaction mixture. TEMPO(BF4) [2] mimicked the role of in situ photogenerated TEMPO + in the photosynthesis cell and was used as a chemical oxidant for glycerol. The gel consisted of 10%wt PVDF-HFP in acetonitrile and contained the one-electron TEMPO(BF4) (~1.5 eq.) while glycerol (~1 eq.) was added to the aqueous solution (sat. NaCl, NaHCO 3 pH 8.3), which was layered in a 1-to-1 fashion on the acetonitrile-gel. The biphasic system was left to react for 16 hours, after which the two-electron oxidation was monitored, and samples of the organic and aqueous phase were taken at t=16 hours, TMS-derivatized and analyzed with gas chromatography (GC) [4,8] (Table S2- 5). The formation of glyceraldehyde and the distribution of the substrate, product, and TEMPO 0/+ over the gel-aqueous system are quantified, giving the main results in Table 1. A conversion of ~17% of glycerol to glyceraldehyde is seen in the biphasic system and only half of the conversion of 30% in a fully mixed system. [13] The decrease in conversion indicates diffusion limitation plays a role in the oxidation reaction during 16 hours. Considering the presence of the oxidation reaction and the affinity of TEMPO to the gel-layer and the product and substrate to the aqueous layer, we believe that this gel is suitable for a biphasic DSPEC system. Table S1: Determined gas chromatography retention times for all used compounds after the TMS derivatization procedure. Analysis was performed on a Trace GC Ultra machine (Interscience) with an RTX-1 column (30 m, 0.25 mm internal diameter, 0.25 μm film thickness), inlet temperature 70°C hold 2.00 minutes ramp 10°C min -1 to 340°C. The quantification of the pure glyceraldehyde gave two peaks with retention times 11.3 and 11.7 (min). The two peaks were analyzed with GC-MS, after which we surmise the peaks to correspond to the TMS-protected keto-(Glyceraldehyde_A) and enol-(Glyceraldehyde_B) tautomers of glyceraldehyde. A calibration approach involving the sum of the peak areas was used. [8] However, other possibilities such as the incomplete TMS-derivatization, acetals, monomers, and dimers (through hemiacetal and hemiketal bond formation) of glyceraldehyde can also be suitable as the exact species is often hard to determine, as discussed by A.

Tables and Figures
Parodi, E. Diguilio, S. Renzini, I. Magario in Carbohydrate Research 2020, 487, 107885. Since the qualification on the TMS-glyceraldehyde species formed during the aqueous derivatization process is generally very strenuous, we would like to refer the reader to the previous mentioned reference for more information and in-depth discussion on the GC analysis of these compounds. [8] Figure S1,2. Gas chromatography calibration curve for 1) glycerol and 2) glyceraldehyde in aqueous solution (sat. NaCl, NaHCO 3 pH 8.                  Table S9. Dye leaching experiment with ø 0.5 cm FTO|TiO 2 |AP11 in alkaline or organic anolytes. The photoanodes were prepared in the same manner as DSSCs and DSPEC experiments. Plates were soaked overnight in either 1.0 M TEMPO, 1.2 M LiTFSI in acetonitrile (3 mL) or 0.1 M glycerol aqueous solution (sat. NaCl, NaHCO 3 pH 8.3, 3 mL) to simulate reaction conditions. The experiments were performed in the dark to prevent dye decomposition by over-illumination. After 23 hours, the remaining AP11 on the FTO|TiO 2 electrodes was quantified by dipping the electrodes for 24 hours in a 0.01 M TBAOH in DMF solution (1 mL), after which UV-VIS studies were performed using ε= 19000 cm -1 at 364 nm. [1] Entry Environment Dye loading (nmol cm -2 ) An Ag/AgCl RE which was placed close to the WE. The CE compartment was separated by a Nafion-117 membrane and consisted of an FTO|Pt CE and was filled with 1.0 M AcOH in acetonitrile (3 mL). A bias potential of 0.1 vs. Ag/AgCl was applied on the WE (P211 potentiostat, Zahner) and the system was illuminated with a 100 mW cm -2 white LED light source (masked size 0.64 cm 2 ). Differences in control experiments are indicated. Product analysis measured by GC analysis was performed after the TMS derivatization procedure on a Trace GC Ultra machine (Interscience) with an RTX-1 column (30 m, 0.25 mm internal diameter, 0.25 μm film thickness), inlet temperature 70°C hold 2.00 minutes ramp 10°C min -1 to 340°C. * Faradaic efficiency was not obtained for a system with photocurrent near 0 µmol.