Catalytic Furfural/5-Hydroxymethyl Furfural Oxidation to Furoic Acid/Furan-2,5-dicarboxylic Acid with H2 Production Using Alkaline Water as the Formal Oxidant

Furfural and 5-hydroxymethyl furfural (HMF) are abundantly available biomass-derived renewable chemical feedstocks, and their oxidation to furoic acid and furan-2,5-dicarboxylic acid (FDCA), respectively, is a research area with huge prospective applications in food, cosmetics, optics, and renewable polymer industries. Water-based oxidation of furfural/HMF is a lucrative approach for simultaneous generation of H2 and furoic acid/FDCA. However, this process is currently limited to (photo)electrochemical methods that can be challenging to control, improve, and scale up. Herein, we report well-defined ruthenium pincer catalysts for direct homogeneous oxidation of furfural/HMF to furoic acid/FDCA, using alkaline water as the formal oxidant while producing pure H2 as the reaction byproduct. Mechanistic studies indicate that the ruthenium complex not only catalyzes the aqueous oxidation but also actively suppresses background decomposition by facilitating initial Tishchenko coupling of substrates, which is crucial for reaction selectivity. With further improvement, this process can be used in scaled-up facilities for a simultaneous renewable building block and fuel production.

Furfural was purchased from Sigma Aldrich and passed through a silica column to remove brown impurities and stored inside N2 glove box. 5-Hydroxylmethylfurfural (HMF) was purchased from S.L. Moran and used without further purification. Reagent grade sodium hydroxide (NaOH), potassium hydroxide (KOH), lithium hydroxide (LiOH), sodium carbonate (Na2CO3), sodium bicarbonate (NaHCO3), potassium phosphate (K3PO4), 1,8-Diazabicyclo [5.4.0]undec-7-ene (DBU), dimethyl aminopyridine (DMAP) were purchased from commercial sources and used without further purification. All solvents were purified according to standard procedures under an argon atmosphere, bubbled with argon, and stored over 4 Å molecular sieves (MS). Furfural and DBU were bubbled with argon for half an hour prior to their use. 1,3,5-trimethylbenzene (mesitylene) was purchased from commercial sources and used as received. Deionized water was used in the reaction, which was bubbled with argon for half an hour prior to its use. NMR spectra were recorded at room temperature on a Bruker AMX-300 (300 MHz) or AMX-400 (400 MHz) or AMX-500 (500 MHz) spectrometers. Chemical shifts of the NMR spectra are reported relative to residual signals of CDCl3 ( 1 H NMR: δ = 7.26 ppm, 13 C NMR: δ = 77.16 ppm), benzene-D6 ( 1 H NMR: δ = 7.16 ppm, 13 C NMR: δ = 128.06 ppm), or DMSO-d6 ( 1 H NMR: δ = 2.5 ppm, 13 C NMR: δ = 40.00 ppm) or the internal standard mesitylene. 31 P{ 1 H} NMR chemical shifts are reported in ppm downfield from H3PO4 and referenced to an external 85% solution of phosphoric acid in D2O. GC-MS was carried out on HP 6890 (flame ionization detector and thermal conductivity detector) and HP 5973 (MS detector) instruments equipped with a 30 m column (Restek 5MS, 0.32 mm internal diameter) with a 5% phenylmethylsilicone coating (0.25 mm) and helium as carrier gas. IR spectra were recorded on a Nicolet FTIR spectrophotometer (KBr, thin Film). GC was carried out on HP 6890 or Agilent 7890B Series GC System with N2 or Helium as carrier gas.

Synthesis and characterization of complex 8 2.1. Synthesis of the ligand
Procedure. Inside a N2 glove box, 1 mmol (365 mg) of acridine dimethylenebromide was dissolved in 5 mL of hexafluoro-2-propanol, to which 2.2 mmol (410 mg) of diphenylphosphine was added dropwise. The solution was transferred to a pressure tube and was heated at 50 o C for 3 days. Afterwards, it was cooled down to room temperature, and 20 mL DCM (dichloromethane) was added. The solution was quickly* washed with 30 mL of 10% aqueous NaOH. The organic layer was collected, and the aqueous layer was washed once more with 20 mL DCM. The combined DCM solutions were dried with MgSO4, and DCM was removed, affording a crude yellow solid.
The solid was then taken inside glove box and dissolved in 5 mL THF. The solution was filtered through a small Celite pad. THF was then removed, and the resulting yellow solid was washed with pentane, sparing amounts of ether, and methanol to obtain the pure PNP(Ph) acridine ligand in ~70% yield.

Synthesis of the acridine complex RuH(CO)ClAcrPNP(Ph) (8)
Procedure: Inside a N2 glove box, 0.1 mmol of RuHCl(CO)(PPh3)3 (95.2 mg) and 0.12 mmol of the PNP(Ph) acridine ligand (69 mg) were suspended in 5 mL toluene. The solution was transferred to a pressure tube and was heated at 65 o C for 3 hours. Formation of yellow precipitates were observed during the reaction. After 3 hours, the solution was cooled to room temperature and toluene was evaporated in vacuo. The resulting yellow solid was washed with diethyl ether and THF for multiple times, affording complex 8 as a yellow powder in 89% yield as mixtures of possible fac and mer isomers.
The complex is not soluble in less polar solvents such as benzene, toluene, or THF. NMR analysis of a CD2Cl2 solution showed the presence of two major isomers, along with one minor isomer, which was directly used for catalysis. The possibility of different isomers with the ruthenium acridine PNP iPr framework has been discussed before in the context of computational studies. 9 It is thus likely that while for the acridine PNP iPr complex only one isomer is thermodynamically favorable, for the PNP Ph complex, their energies are much more similar, leading to their coobservations. 31 P{ 1 H} NMR (162 MHz, CD2Cl2) δ 53.26 (minor, 0.44 P), 45.40 (major, 0.56 P).

Standard procedure for furfural oxidation to furoic acid using water
In a N2 glove box, 10 µmol of ruthenium catalyst 6 (6 mg) was suspended in 2 mL of dry 1,4-   Figure S8. 1 H NMR spectrum of crude reaction mixture after complete oxidation of furfural to furoic acid salt. Figure S9. 1 H NMR spectrum of the crude reaction mixture after partial oxidation of furfural. Presence of furoic acid salt and furfuryl alcohol is seen in the solution. Figure S10. 1 H NMR spectrum (300 MHz) of isolated furoic acid in DMSO-d6 after acid workup. Figure S11. 13 C NMR spectrum (75 MHz) of isolated furoic acid in DMSO-d6 after acid workup.  Reaction conditions. Furfural (1 mmol), base (1.2 mmol), 6 (1 mol%), water (1 mL), 1,4-dioxane (2 mL), 135 o C, 48 h. Reported yields are that of furoic acid salts prior to acidification. a Yields calculated as in Table S1 b generation of H2/CO2 gas mixture was observed.

Standard procedure for HMF oxidation to FDCA using water
In a N2 glove box, 10 µmol of ruthenium catalyst 6 (6 mg) was suspended in 2 mL of dry 1,4-     Figure S14. 1 H NMR spectrum of isolated FDCA after acid workup in DMSO-d6.

1 H NMR of isolated FDCA
Figure S15. 13 C NMR spectrum of isolated FDCA after acid workup in DMSO-d6.

Procedure for H2 evolution profile experiment
In a N2 glove box, 10 µmol of ruthenium catalyst 6 (6 mg) was suspended in 2 mL of dry 1,4-

Procedure for control experiment with catalyst
In a N2 glove box, 0.01 mmol of complex 6, 1 mmol (96 mg

Procedure for Tishchenko coupling
In a N2 glove box, 5 µmol of complex 9 and 0.5 mmol (48 mg) of furfural were dissolved in 2 mL

Procedure for BHMF oxidation using water
In a N2 glove box, 10 µmol of ruthenium catalyst 6 (6 mg) was suspended in 2 mL of dry 1,4-  6. Mechanistic studies

Synthesis and characterization of the furoate complex 9b
Procedure: Complex 9 (12 mg) was dissolved in 3 mL of THF in a 20 mL vial. To this solution,

Complex 9, in the presence of benzaldehyde, and water, forms the dearomatized benzoate complex 9a at RT
Procedure. To a J Young NMR tube was added a solution of 5 µmol of complex 9 in 0.5 mL THF.
To the solution, 0.025 mmol of benzaldehyde and 0.05 mmol water was added. The tube was then stirred at room temperature for 10 min and the 31 P and 1 H NMR was subsequently recorded.
Formation of the dearomatized benzoate complex 9a was observed by NMR spectroscopy. Figure S35. Stacked 31 P{ 1 H} NMR spectra of reaction between complex 9, benzaldehyde, and water.

Complex 9, in the presence of furfural and water, forms the dearomatized furoate complex 9b upon heating
Procedure. To a J Young NMR tube was added a solution of 5 µmol of complex 9 in 0.5 mL THF.
To the solution, 0.025 mmol of furfural and 0.1 mL water was added. The tube was then stirred at room temperature for 10 min and the 31 P and 1 H NMR spectra were subsequently recorded.
Formation of several unidentifiable species was observed in the 31 P NMR spectrum. The solution was then heated at 130 o C for 10 min and the NMR spectra were measured again, revealing the formation the furoate complex 9b as the major species. Figure S36. Stacked 31 P{ 1 H} NMR spectra of reaction between complex 9, furfural, and water.

The furoate complex 9b can also be accessed from 9 by reaction with furoic acid
Procedure. In a J Young NMR tube was added a solution of 0.01 mmol of complex 9 in 0.5 mL of THF, to which 0.012 mmol of furoic acid was added. The tube was then stirred at room temperature for 10 min and the 31 P and 1 H NMR were subsequently recorded. Formation of the furoate complex 9b was observed as the major species along with the mer furoic acid adduct complex (9e) as the minor species. The NMR was again taken after 1 h of stirring and the furoate complex 9b was observed as the only species. Figure S37. Stacked 31 P{ 1 H} NMR spectra of reaction between complex 9, and furoic acid.

The furoate complex 9b is stable in the presence of water in THF even at elevated temperatures
Procedure. The furoate complex was prepared in situ by mixing 9 with furoic acid in THF. To a J Young tube, 0.5 mL THF solution of 5 µmol complex 9b was added. 0.05 mmol of water was then added and the resulting solution was stirred at room temperature for 10 min and the 31 P and 1 H NMR were subsequently recorded. No reaction was observed by 31 P NMR with only the furoate complex 9b being present in solution. Subsequently, the tube was heated at 130 o C for 10 min and the NMR spectra were recorded again. The furoate complex 9b was observed as the only species. Figure S38. Stacked 31 P{ 1 H} NMR spectra of reaction between complex 9b and water.

The furoate complex 9b forms a new complex upon addition of base, which decomposes when heated to high temperature
Procedure. To the NMR tube from previous experiment, NaOH (0.015 mmol) was added. The resulting solution was stirred for 10 min and the 31 P and 1 H NMR were subsequently recorded.
Formation of a new species, likely the ruthenium hydroxide complex was observed, along with generation of sodium furoate in the 1 H NMR. Subsequently, the tube was heated at 140 o C for 10 min and the NMR were recorded again. Decomposition of the previous complex to multiple unidentified species was observed. Complex 9c is also observed to decompose slowly at RT to other complexes or upon isolation attempts by extraction in benzene, likely signifying the requirement of alkaline conditions for its prolonged stability. Figure S39. Stacked 31 P{ 1 H} NMR spectra of reaction between complex 9b, water and NaOH.

Possible catalytic cycles for alcohol dehydrogenation
The possible catalytic cycles for furfuryl alcohol dehydrogenation to furfural are shown below.
Under the reaction conditions, two alternative catalytic cycles are possible. In the water unassisted pathway (in blue), initial coordination of furfuryl alcohol, followed by H2 evolution can generate the alkoxy complex 9f. Further b hydride elimination from the alkoxy complex can then generate complex 9-fac and furfural. Previous computational studies regarding ethanol dehydrogenation by complex 9 had suggested that the initial H2 evolution step is rate determining step. 13 In the presence of water, an alternative stepwise water assisted dehydrogenation pathway can also be envisioned (in red). 9 In this mechanism, initial water coordination followed by dehydrogenation generates the hydroxy intermediate 9c.
Further alcohol coordination to 9c followed by water elimination generates alkoxide intermediate 9f, in this alternative water assisted stepwise pathway. The generated aldehyde, after subsequent b hydride elimination, converts to furoate salt via dehydrogenative oxidation or disproportionation pathway, as discussed in the main text. Figure S40. Possible catalytic pathways for alcohol dehydrogenation catalyzed by complex 9.

Possible catalytic cycle for Tishchenko coupling
Complex 9 was observed to catalyze the Tishchenko coupling of furfurals to furfuryl furoate ester in 1,4-dioxane ( Figure S19). The tentative mechanism cycle of the process is given below. Initial coordination of aldehyde to 9 followed by hydride transfer generates the alkoxy complex 9f. To 9f, another furfural coordinates in the cis vacant site. Furfuryloxy anion migration to bound furfural can further generate the acetal complex 9g. Finally, b hydride elimination takes place from the acetal complex to generate the ester and complex 9-fac. A relevant computational study for ester formation from aldehydes by complex 9 has been reported recently by our group. 13 Figure S41. Possible catalytic cycle for furfural disproportionative coupling to ester catalyzed by complex 9.

Different ways to access the dearomatized complex 9 from the aromatized complex 6
The dearomatized complex 9 can be accessed from the aromatized complex 6 via several different hydride transfer methods-1) By heating a solution of 6 in H 2 with a base KOH. 6 2) By the treatment of 6 with super-hydride NaBEt 3 H at room temperature. 8 3) By heating a solution of 6 with alcohol/aldehyde substrate in presence of base. 7,9 In this last case, one C-H proton is transferred to the 9 position of the catalyst acridine ring to generate dearomatized complex while oxidizing the hydride donor (aldehyde from alcohol; acid from aldehyde) These different routes are shown in the figure below. However, it should be noted that the dearomatization of the acridine ring is seemingly irreversible, and we have not yet observed the formation of aromatic complexes from the dearomatized complexes under the reaction conditions in this study or the previous studies. Figure S42. Different ways to access the dearomatized complex 9 from 6.

Oxidation of other aldehydes using alkaline water catalyzed by 6
We also explored the possibility of oxidizing other aldehydes by the developed protocol herein.
Under the reaction conditions (similar procedure as in section 3), benzaldehyde, hexanal, and phenylacetaldehyde can also be converted to the corresponding acids in near quantitative yields, demonstrating the versatility of the process. Figure S43. Oxidation of other aldehydes by alkaline water catalyzed by 6

Alternate reaction mechanism involving free acetal dehydrogenation
An alternate mechanistic cycle involving the direct water attack on the aldehyde to form free acetal, followed by acetal dehydrogenation can also be envisioned. In this mechanism, the free acetal can coordinate to complex 9 to generate the acetal adduct complex 9h. H 2 extrusion from 9h will then generate the chelating acetal complex 9d. Subsequently, beta hydride elimination from 9d will generate the acid adduct complex 9f. Further, 9f.can generate the product acid salt and complex 9 in the presence of a base. In this mechanism, the carboxylate complex 9b is not a part of the mechanistic cycle but acts as an off cycle resting intermediate.
Based on the experimental observations, we cannot entirely rule out this alternate mechanism.
Although, given the easy formation of carboxylate complex 9b from 9f, and further equilibrium of the hydroxy complex (9c) with the carboxylate complex (9b) that is observed in presence of NaOH, the mechanism as proposed in the manuscript seems more in line with the experiments. It will be interesting to explore these competing mechanisms by DFT studies in future to find the most energetically favorable reaction pathway. Figure S44. Alternate mechanistic cycle based on free acetal dehydrogenation

Summary of the overall proposed pathways
A summary of the overall ongoing pathways is shown in Figure S45. The five fundamental reactions ongoing under the conditions are-a) [Ru] catalyzed aldehyde oxidation by water to acid salt, b) base catalyzed Cannizzaro disproportionation of the aldehyde to acid and alcohol, c) [Ru] catalyzed Tishchenko coupling of aldehyde to ester, d) base mediated ester hydrolysis to acid and alcohol and e) [Ru] catalyzed alcohol dehydrogenation to the aldehyde. For furfural, all these pathways eventually lead to the formation of furoic acid salt in quantitative yield, whose formation is also irreversible. For HMF oxidation to FDCA, initially the aldehyde group is oxidized to acid fully via the combination of these pathways, and at the next step, the alcohol group is oxidized to acid (which is more challenging to oxidize than the aldehyde group under the reaction conditions), generating the desired product FDCA.