Oxidative Depolymerization of Cellulolytic Enzyme Lignin over Silicotungvanadium Polyoxometalates

The aim of this study was to explore the catalytic performance of the oxidative depolymerization of enzymatic hydrolysis lignin from cellulosic ethanol fermentation residue by different vanadium substituted Keggin-type polyoxometalates (K5[SiVW11O40], K6[SiV2W10O40], and K6H[SiV3W9O40]). Depolymerized products were analyzed by gel permeation chromatography (GPC), gas chromatography–mass spectrometer (GC/MS), and two-dimensional heteronuclear single quantum coherence nuclear magnetic resonance (2D HSQC NMR) analysis. All catalysts showed an effective catalytic activity. The best result, concerning the lignin conversion and lignin oil production, was obtained by K6[SiV2W10O40], and the highest yield of oxidative depolymerization products of 53 wt % was achieved and the main products were monomer aromatic compounds. The HSQC demonstrated that the catalysts were very effective in breaking the β-O-4 structure, the dominant linkage in lignin, and the GPC analysis demonstrated that the molecular of lignin was declined significantly. These results demonstrate the vanadium substituted silicotungstic polyoxometalates were of highly active and stable catalysts for lignin conversion, and this strategy has the potential to be applicable for production of value-added chemicals from biorefinery lignin.


Introduction
Lignin is a heterogeneous renewable biopolymer and the only feedstock in nature with high carbon content and high aromaticity [1]. It is chemically and physically interlaced with cellulose and hemicelluloses in the plant cell walls. Its complex chemical structure and stable chemical properties make most lignin degradation a highly challenging work [2,3]. In a typical biorefinery, the natural lignin is structurally modified under the traditional acid or high temperature fraction or pre-treatment, leading to the irreversible condensation that dramatically affects its further catalytic valorization [2,4,5]. At present, lignin as a raw material for the production of value-added products as an alternative to fossil derived chemicals has attracted ever increasing attention during the past decade, and it is also apparently the most encouraging transformation to high-value utilization of lignin [6]. The processes for conversion lignin can be broadly classified into base or acid depolymerization [7], pyrolysis [8][9][10], hydroprocessing [11,12], oxidation [13][14][15], and other depolymerization processes. Among the above catalytic transformation strategies, oxidative depolymerization is a very promising valorization method and represents various advantages, since it focuses on production of highly functionalized platform aromatic compounds which can be used as starting materials in chemical and pharmaceutical industries [16][17][18].

Preparation and Characterization of the Catalysts
The vanadium substituted silicotungstic polyoxometalates K5[SiVW11O40], K6[SiV2W10O40], and K6H[SiV3W9O40] were prepared according to the literature method [32,33], with some modifications. FTIR spectroscopy was recorded in KBr discs on a Nicolet Magna 560 IR spectrometer. Scanning electron microscope (SEM) was carried out by a FEI-Quanta 200 (Madison, America) equipped with energy dispersive X-ray spectrometer (EDX) analyses at 100 kV, EDX elemental mapping were operated at 200 kV. The redox potential was measured by cyclic voltammetry (CV) on a CS Corrtest CHI760E electrochemical workstation (Shanghai, China) equipped with a carbon paste electrode and saturated calomel as the reference electrode in an electrolyte of 0.1 M sulfuric acid solution.
The FTIR spectra of different vanadium substituted silicotungstic polyoxometalates were investigated. As shown in Figure 1, there were four characteristic peaks at 1008 cm −1 (vas Si-Oa, internal oxygen connecting Si and W), 962 cm −1 (vas W-Od, terminal oxygen bonding to W atom), 896 cm −1 (vas W-Ob, edge-sharing oxygen connecting W), and 755cm −1 (vas W-Oc, corner-sharing oxygen connecting WO6 units) ranged from 600 to 1100 cm −1 , which were attributed to Keggin structural vibrations. Figure 1 also shows the XRD patterns of these three silicotungvanadium polyoxometalates with characteristic diffraction peaks at 8.5°, 10.3°, 25.8° and 34.6°, which were attributed to bodycentered cubic secondary structure of Keggin anion [34].  . These catalysts displayed well-shaped crystalline particles, especially for the one and two vanadium substituted POMs. With the increase of vanadium, the size of these particles gradually decreased, which was due to the ionic radius of tungsten being larger than vanadium. Moreover, EDX was employed to determine the elemental composition of these catalysts. The predominant components were all found to be Si, W, V, and O, in accordance with the proportion of theoretical stoichiometry. CV-potential curves of vanadium-substituted polyoxometalates was shown in Figure 3. These results showed that the introduction of vanadium can alter the redox properties of POMs, and the structural and property performance differences caused by different vanadium substitutions will play an important role in lignin transformation.  These catalysts displayed well-shaped crystalline particles, especially for the one and two vanadium substituted POMs. With the increase of vanadium, the size of these particles gradually decreased, which was due to the ionic radius of tungsten being larger than vanadium. Moreover, EDX was employed to determine the elemental composition of these catalysts. The predominant components were all found to be Si, W, V, and O, in accordance with the proportion of theoretical stoichiometry. CV-potential curves of vanadium-substituted polyoxometalates was shown in Figure 3. These results showed that the introduction of vanadium can alter the redox properties of POMs, and the structural and property performance differences caused by different vanadium substitutions will play an important role in lignin transformation.

Catalytic Decomposition of the CEL
In a typical reaction protocol, 0.25 g of purified CEL, 0.1 mmol POMs, 8 mL CH3OH, and 2 mL H2O were charged into a 10 mL stainless autoclave. The reactor was sealed and purged three times with oxygen, pressurized to 2.0 MPa O2 at room temperature and then heated to the desired temperature with stirring at 500 rpm. After the reaction, the autoclave was cooled to room temperature and depressurized. The reaction mixture was filtered, and the insoluble fraction was washed with methanol, the solution was totally transferred to a round-bottom flask and the methanol was removed under vacuum condition. The obtained mixture was extracted with dichloromethane (DCM) three times (10 mL × 3) and the organic layer was dried with anhydrous MgSO4 then evaporated under reduced pressure. The lignin conversion and lignin oil yields were calculated using the following equations:

Product Analysis
The HSQC NMR spectra was carried out at 25 °C on a Bruker-Advance III 500 MHz spectrometer (Karlsruhe, Germany) equipped with a z-gradient double resonance probe. A 100 mg sample was dissolved in 0.4 mL of DMSO-d6 and 0.1 mL pyridine-d5. The experiment parameters were described previously [35]. The depolymerized aromatic compounds were also analyzed by GC/MS using a Shimadzu GC/MS-QP2010SE (Kyoto, Japan) equipped with an SH-Rxi-5Sil capillary column, and the

Catalytic Decomposition of the CEL
In a typical reaction protocol, 0.25 g of purified CEL, 0.1 mmol POMs, 8 mL CH3OH, and 2 mL H2O were charged into a 10 mL stainless autoclave. The reactor was sealed and purged three times with oxygen, pressurized to 2.0 MPa O2 at room temperature and then heated to the desired temperature with stirring at 500 rpm. After the reaction, the autoclave was cooled to room temperature and depressurized. The reaction mixture was filtered, and the insoluble fraction was washed with methanol, the solution was totally transferred to a round-bottom flask and the methanol was removed under vacuum condition. The obtained mixture was extracted with dichloromethane (DCM) three times (10 mL × 3) and the organic layer was dried with anhydrous MgSO4 then evaporated under reduced pressure. The lignin conversion and lignin oil yields were calculated using the following equations: Lignin oil yield (%) = ( ) ( ) × 100%

Product Analysis
The HSQC NMR spectra was carried out at 25 °C on a Bruker-Advance III 500 MHz spectrometer (Karlsruhe, Germany) equipped with a z-gradient double resonance probe. A 100 mg sample was dissolved in 0.4 mL of DMSO-d6 and 0.1 mL pyridine-d5. The experiment parameters were described previously [35]. The depolymerized aromatic compounds were also analyzed by GC/MS using a Shimadzu GC/MS-QP2010SE (Kyoto, Japan) equipped with an SH-Rxi-5Sil capillary column, and the

Catalytic Decomposition of the CEL
In a typical reaction protocol, 0.25 g of purified CEL, 0.1 mmol POMs, 8 mL CH 3 OH, and 2 mL H 2 O were charged into a 10 mL stainless autoclave. The reactor was sealed and purged three times with oxygen, pressurized to 2.0 MPa O 2 at room temperature and then heated to the desired temperature with stirring at 500 rpm. After the reaction, the autoclave was cooled to room temperature and depressurized. The reaction mixture was filtered, and the insoluble fraction was washed with methanol, the solution was totally transferred to a round-bottom flask and the methanol was removed under vacuum condition. The obtained mixture was extracted with dichloromethane (DCM) three times (10 mL × 3) and the organic layer was dried with anhydrous MgSO 4 then evaporated under reduced pressure. The lignin conversion and lignin oil yields were calculated using the following equations:

Product Analysis
The HSQC NMR spectra was carried out at 25 • C on a Bruker-Advance III 500 MHz spectrometer (Karlsruhe, Germany) equipped with a z-gradient double resonance probe. A 100 mg sample was Polymers 2019, 11, 564 5 of 12 dissolved in 0.4 mL of DMSO-d 6 and 0.1 mL pyridine-d 5 . The experiment parameters were described previously [35]. The depolymerized aromatic compounds were also analyzed by GC/MS using a Shimadzu GC/MS-QP2010SE (Kyoto, Japan) equipped with an SH-Rxi-5Sil capillary column, and the parameters were described previously [36]. Gel permeation chromatography (GPC) analysis was conducted using an Agilent Infinity 1260 HPLC (Palo Alto, CA, USA) equipped with a Refractive Index Detector and an Agilent PLgel MIXED C column at 40 • C using tetrahydrofuran (THF) as the mobile phase and a flow rate of 1 mL/min. Before the GPC experiment, the CEL and lignin oil products were acetylated as shown briefly below: CEL (10 mg) or lignin oil products (10 mg) were treated with a 1:1 mixture of acetic anhydride and pyridine (1.0 mL) at room temperature for 48 h. The acetylated products were extracted by ethyl acetate and the organic layer was washed with brine and dried over anhydrous MgSO 4 . After evaporation under vacuum, the obtained acetylation products were dissolved in THF (2 mg mL −1 ) and filtered over a 0.45 µm syringe filter prior to injection.

Catalytic Oxidation Depolymerization of CEL
The catalytic efficiency of different vanadium substituted silicotungstic polyoxometalates were evaluated by the measurement of lignin conversion and lignin oil yield. As shown in Figure 4, the K 6 [SiV 2 W 10 O 40 ] with two vanadium atoms exhibited both highest lignin conversion at 87% and lignin oil yield at 53%, respectively. As compared to the silicotungvanadium POMs, the control reaction without catalysts and with a vanadium-free POM K 8 [SiW 11 O 39 ] showed both the low lignin conversion and lignin oil yield. This result reflected that the number of vanadium had significant effect on the catalytic transformation of lignin. The use of vanadium-substituted polyoxometalates with strong redox ability enhances the hydrogen liberation and the liberated hydrogen transfer to the reaction intermediates. As a consequence, the lignin conversion and products yields were enhanced. In general, the redox ability of POMs increases with the number of substitutions of vanadium, contributing to the increase of lignin conversion. During these reactions, K 6 [SiV 2 W 10 O 40 ] performed higher efficiency than K 5 [SiVW 11 O 40 ]. This may be attributed to the higher redox potential introduced by vanadium. However, the K 6 H[SiV 3 W 9 O 40 ] with more vanadium atoms did not follow this rule. The possible explanation was that the lignin was over oxidized to carbon dioxide and water, and the lignin fragments were recombined during the reaction. The detail information about the oxidative reaction was characterized by GPC, HSQC and GC/MS as below.
Polymers 2019, 11, x FOR PEER REVIEW 5 of 13 parameters were described previously [36]. Gel permeation chromatography (GPC) analysis was conducted using an Agilent Infinity 1260 HPLC (Palo Alto, USA) equipped with a Refractive Index Detector and an Agilent PLgel MIXED C column at 40 °C using tetrahydrofuran (THF) as the mobile phase and a flow rate of 1 mL/min. Before the GPC experiment, the CEL and lignin oil products were acetylated as shown briefly below: CEL (10 mg) or lignin oil products (10 mg) were treated with a 1:1 mixture of acetic anhydride and pyridine (1.0 mL) at room temperature for 48 h. The acetylated products were extracted by ethyl acetate and the organic layer was washed with brine and dried over anhydrous MgSO4. After evaporation under vacuum, the obtained acetylation products were dissolved in THF (2 mg mL −1 ) and filtered over a 0.45 μm syringe filter prior to injection.

Catalytic Oxidation Depolymerization of CEL
The catalytic efficiency of different vanadium substituted silicotungstic polyoxometalates were evaluated by the measurement of lignin conversion and lignin oil yield. As shown in Figure 4, the K6[SiV2W10O40] with two vanadium atoms exhibited both highest lignin conversion at 87% and lignin oil yield at 53%, respectively. As compared to the silicotungvanadium POMs, the control reaction without catalysts and with a vanadium-free POM K8[SiW11O39] showed both the low lignin conversion and lignin oil yield. This result reflected that the number of vanadium had significant effect on the catalytic transformation of lignin. The use of vanadium-substituted polyoxometalates with strong redox ability enhances the hydrogen liberation and the liberated hydrogen transfer to the reaction intermediates. As a consequence, the lignin conversion and products yields were enhanced. In general, the redox ability of POMs increases with the number of substitutions of vanadium, contributing to the increase of lignin conversion. During these reactions, K6[SiV2W10O40] performed higher efficiency than K5[SiVW11O40]. This may be attributed to the higher redox potential introduced by vanadium. However, the K6H[SiV3W9O40] with more vanadium atoms did not follow this rule. The possible explanation was that the lignin was over oxidized to carbon dioxide and water, and the lignin fragments were recombined during the reaction. The detail information about the oxidative reaction was characterized by GPC, HSQC and GC/MS as below.   To confirm the depolymerized products, the molecular weight distribution was investigated by GPC ( Figure 5). As shown in Figure 5, the parent lignin was essentially insoluble in THF (presumably due to high molecular weight), whereas the post-reaction oils were all significantly soluble. GPC profiles of the CEL and depolymerized products catalyzed by the different number of vanadium substituted POMs suggested the molecular weight was significantly decreased in average molecular weight (M W 1300 g mol −1 ) relative to the raw CEL (M W 4200 g mol −1 ). Further insight into the structural changes were analyzed using two-dimensional heteronuclear single quantum coherence nuclear magnetic resonance (2D HSQC NMR) which was particularly found to be helpful to monitor the changes of lignin structure after oxidative decomposition reaction. The 2D HSQC NMR spectra of CEL mainly separated for three regions, namely the aliphatic region (generally δ C /δ H 0-50/0-2.5 ppm), side chain region (generally δ C /δ H 50-100/2.5-6.5 ppm), and aromatic region (generally δ C /δ H 100-140/6.5-8 ppm) [37,38]. The inter-unit linkages and structural units could be characterized by the side chain region and aromatic region respectively. The aliphatic region was not related with the lignin structure, therefore, it was not discussed here. The detailed information is presented in Figure 6, the assignments of lignin samples were based on recent studies [1,39,40]. By calculating the Cα integral values of the HSQC signals, the relative percentages of the main three inter-unit linkages, namely, β-O-4 (A), β-5(B), and β-β(C) were 80%, 14%, and 6% respectively ( Figure 6). After the reaction, as we can see from Figure 6b,c,d, the signals of A α (δ C /δ H 72.3/5.1 ppm) and A β (δ C /δ H 86.7/4.3 ppm) attributed to the benzylic hydroxyl and secondary alkyl protons, respectively, were vanished completely. This result indicated the silicotungvanadium POMs were effective in the cleavage of C-O-C linkages in CEL under this reaction condition. Except the β-O-4 linkage, the signals for the β-5 and β-β inter-unit linkages disappeared as well. For the aromatic area, we can also observe that the Cα hydroxyl group oxidized syringyl type units generated obviously after the reaction, and the syringyl to guaiacyl ratio after hydrolysis was increased slightly but remained basically the same as the unreacted CEL. These results were in accordance with the analysis detected by GC/MS. To confirm the depolymerized products, the molecular weight distribution was investigated by GPC ( Figure 5). As shown in Figure 5, the parent lignin was essentially insoluble in THF (presumably due to high molecular weight), whereas the post-reaction oils were all significantly soluble. GPC profiles of the CEL and depolymerized products catalyzed by the different number of vanadium substituted POMs suggested the molecular weight was significantly decreased in average molecular weight (MW 1300 g mol −1 ) relative to the raw CEL (MW 4200 g mol −1 ). Further insight into the structural changes were analyzed using two-dimensional heteronuclear single quantum coherence nuclear magnetic resonance (2D HSQC NMR) which was particularly found to be helpful to monitor the changes of lignin structure after oxidative decomposition reaction. The 2D HSQC NMR spectra of CEL mainly separated for three regions, namely the aliphatic region (generally δC/δH 0-50/0-2.5 ppm), side chain region (generally δC/δH 50-100/2.5-6.5 ppm), and aromatic region (generally δC/δH 100-140/6.5-8 ppm) [37,38]. The inter-unit linkages and structural units could be characterized by the side chain region and aromatic region respectively. The aliphatic region was not related with the lignin structure, therefore, it was not discussed here. The detailed information is presented in Figure 6, the

Effect of the Reaction Time and Temperature
The effects of reaction conditions including reaction time and reaction temperature on lignin decomposition are shown in Figure 7. When the reaction time was prolonged from 2 h to 3 h, the lignin conversion yield and oily products yield changed from 74% and 48% to 87% and 53%, respectively. Prolong the reaction time to 6 h, the lignin conversion yield and oily products yield decreased to 83% and 44%, which was probably due to the slight repolymerization reactions of intermediate lignin fragments. These condensation reactions of lignin using POMs as catalysts have been well studied in previous reports [14]. A similar volcano-shaped trend was also observed at different reaction temperatures. Increase in the reaction temperature led to an increase of lignin conversion yield and oily products yield to 150 °C. When temperature was further increased to 190 °C, the lignin oily products yield was decreased from 52.7 wt % at 150 °C to 43.6 wt % at 190 °C. In brief, with the progress of the reaction, higher reaction temperature and longer reaction time would facilitate the depolymerization reaction, but the undesirable condensation reactions occurred in the final stage of reaction, thus inhibiting the efficient degradation of lignin.

Effect of the Reaction Time and Temperature
The effects of reaction conditions including reaction time and reaction temperature on lignin decomposition are shown in Figure 7. When the reaction time was prolonged from 2 h to 3 h, the lignin conversion yield and oily products yield changed from 74% and 48% to 87% and 53%, respectively. Prolong the reaction time to 6 h, the lignin conversion yield and oily products yield decreased to 83% and 44%, which was probably due to the slight repolymerization reactions of intermediate lignin fragments. These condensation reactions of lignin using POMs as catalysts have been well studied in previous reports [14]. A similar volcano-shaped trend was also observed at different reaction temperatures. Increase in the reaction temperature led to an increase of lignin conversion yield and oily products yield to 150 • C. When temperature was further increased to 190 • C, the lignin oily products yield was decreased from 52.7 wt % at 150 • C to 43.6 wt % at 190 • C. In brief, with the progress of the reaction, higher reaction temperature and longer reaction time would facilitate the depolymerization reaction, but the undesirable condensation reactions occurred in the final stage of reaction, thus inhibiting the efficient degradation of lignin.

Depolymerization Products Analysis
Although GPC and HSQC are useful in characterization of molecular weight distribution and structure changes, respectively, they are not powerful at identifying the individual depolymerized products. The purpose of this study is to produce monomer aromatic compounds, so the GC/MS analysis was also employed on the reaction mixture catalysed by K6[SiV2W10O40]. As shown in table 1, the 2-methoxy-4-vinylphenol and vanillin, acetosyringone, and 4-hydroxybenzaldehyde and 4-(methoxycarbonyl)-phenol are the most abundant aromatic compounds in the reaction mixture, derived from the G, S and H units of CEL, respectively. Among these aromatics, the vanillin is a very important commercial compound used in food, cosmetic and pharmaceutical industries [41], and the functionality around the aromatic ring allows it an adaptable platform compounds to synthesize numerous polymers [42]. The other compounds like acetosyringone can also be used in medicine, perfume, pesticide chemistry and organic synthesis industry.

Depolymerization Products Analysis
Although GPC and HSQC are useful in characterization of molecular weight distribution and structure changes, respectively, they are not powerful at identifying the individual depolymerized products. The purpose of this study is to produce monomer aromatic compounds, so the GC/MS analysis was also employed on the reaction mixture catalysed by K 6 [SiV 2 W 10 O 40 ]. As shown in Table 1, the 2-methoxy-4-vinylphenol and vanillin, acetosyringone, and 4-hydroxybenzaldehyde and 4-(methoxycarbonyl)-phenol are the most abundant aromatic compounds in the reaction mixture, derived from the G, S and H units of CEL, respectively. Among these aromatics, the vanillin is a very important commercial compound used in food, cosmetic and pharmaceutical industries [41], and the functionality around the aromatic ring allows it an adaptable platform compounds to synthesize numerous polymers [42]. The other compounds like acetosyringone can also be used in medicine, perfume, pesticide chemistry and organic synthesis industry.

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the aqueous phase of the product extraction step containing all the reacted catalysts was evaporated to

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the aqueous phase of the product extraction step containing all the reacted catalysts was evaporated to

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the aqueous phase of the product extraction step containing all the reacted catalysts was evaporated to

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the aqueous phase of the product extraction step containing all the reacted catalysts was evaporated to dryness by water bath and was then subjected to next reaction. As shown in Figure 8, after five cycles under the optimizing conditions, the lignin oil yield had almost no significant changes with only 6% decrease. The results suggested the vanadium substituted silicotungstic polyoxometalates was stable under such conditions and could be reused without significant loss of its catalytic activity.

Recycling Experiments
Reusability of the catalyst is one of the most important performance requirements in practical applications. During the lignin depolymerization process, the most important properties of POMs are their re-oxidation with oxygen, and the deoxidized POMs can be used for the next catalytic circulation. Considering the silicotungvanadium polyoxometalates are homogeneous catalysts, the aqueous phase of the product extraction step containing all the reacted catalysts was evaporated to dryness by water bath and was then subjected to next reaction. As shown in Figure 8, after five cycles under the optimizing conditions, the lignin oil yield had almost no significant changes with only 6% decrease. The results suggested the vanadium substituted silicotungstic polyoxometalates was stable under such condition activity.

Possible Mechanisms of Lignin Depolymerization
According to the chemical properties of the substrate, there are main three pathways during the oxidative reaction when using POMs as catalysts. Namely, electron-transfer, O-transfer, and radical-transfer [43]. Based on the above discussion on the products and structure changes before and after reaction, a mechanism is proposed for describing the lignin depolymerization over silicotungvanadium polyoxometalates. As shown in Figure 9, the POMs ions undergo repeated cycles of reduction and re-oxidation, which depends on whether the redox potential of POMs is high enough to oxidize the lignin substrate and low enough to be oxidized by oxygen [20,26]. At step 1, the POMs with a redox electrochemical potential oxidized the lignin substrate and the POMs was reduced. During the oxidation, the benzyl hydroxyl was selectively oxidized into the corresponding carbonyl, which could lower the bond energy of the C-O bond and make the β-O-4 linkages easier to fracture [2]. Chemoselective oxidation of the benzylic hydroxyl groups in lignin polymer is of great significance in the lignin oxidative depolymerization, and it represents an applicable way for facilitating the lignin depolymerization [44,45]. After the step 1, the insoluble macromolecular lignin turn into the soluble lignin ox , then the lignin ox undergo C-C and/or C-O cleavage, resulting in the forming of low molecular of aromatic compounds and partial oxidation of the lignin fragments to CO 2 and H 2 O. Finally, the catalytic cycle is completed by the re-oxidation of POM red to POM ox form using O 2 (Step 2). The regenerated POMs could be reused in the next catalytic circulation. significance in the lignin oxidative depolymerization, and it represents an applicable way for facilitating the lignin depolymerization [44,45]. After the step 1, the insoluble macromolecular lignin turn into the soluble ligninox, then the lignin ox undergo C-C and/or C-O cleavage, resulting in the forming of low molecular of aromatic compounds and partial oxidation of the lignin fragments to CO2 and H2O. Finally, the catalytic cycle is completed by the re-oxidation of POMred to POMox form using O2 (Step 2). The regenerated POMs could be reused in the next catalytic circulation.

Conclusions
In summary, we developed an effective degradation strategy for biorefinery CEL in the production of aromatic compounds. Our chemical conversion uses reuseable polyoxometalates to transform CEL into valuable aromatic compounds with an acceptable conversion and yield. The appropriate substituted vanadium number with high catalytic activity could easily oxidise lignin substrate followed by homolytic cleavage of C-C and C-O bonds. The catalyst could be reused five times without obvious loss of catalytic activity. Thus, the strategy of vanadium substituted Keggintype silicotungstic polyoxometalates was evaluated as a potential way for the valorization of biorefinery lignin.

Conflicts of Interest:
The authors declare no conflicts of interest.

Conclusions
In summary, we developed an effective degradation strategy for biorefinery CEL in the production of aromatic compounds. Our chemical conversion uses reuseable polyoxometalates to transform CEL into valuable aromatic compounds with an acceptable conversion and yield. The appropriate substituted vanadium number with high catalytic activity could easily oxidise lignin substrate followed by homolytic cleavage of C-C and C-O bonds. The catalyst could be reused five times without obvious loss of catalytic activity. Thus, the strategy of vanadium substituted Keggin-type silicotungstic polyoxometalates was evaluated as a potential way for the valorization of biorefinery lignin.