Base-Free Selective Oxidation of Glycerol over LDH Hosted Transition Metal Complexes Using 3 % H 2 O 2 as Oxidant

A series of transition metal sulphonato-Schiff base complexes were intercalated into Mg–Al layered-double hydroxides (LDHs). The obtained catalysts were characterized by FTIR, XRD, N2 sorption, SEM and elemental analysis, and then were used in the selective oxidation of glycerol (GLY) using 3% H2O2 as an oxidant. It was found that their catalytic performances were closely related to the loading of active complexes, the Schiff base ligands and the metal centers of the catalysts, as well as the reaction conditions. The optimal conversion of GLY was 85.0%, while the selectivity of 1,3-dihydroxyacetone (DHA) was 56.5%. Moreover, the catalysts could be reused at least 10 times.


Introduction
The use of biorenewable feedstocks to produce commodity chemicals and clean fuels as a substitute for the limited fossil fuel reserves is an essential pathway to sustainable development [1,2].In this context, the biodiesel industry was booming in the past decades.However, as an unavoidable by-product of biodiesel production, 1 mol GLY is formed during the generation of every 3 mol biodiesel methyl esters.So, the effective use of GLY has attracted much attention in academia and industry.Luckily, as a highly functionalized molecule, GLY can produce various valuable compounds by oxidation, dehydration, hydrogenolysis, esterification, transesterificaiton, polymerization, and so on.Among them, the oxidation conversion is of immense current importance for the synthesis of fine chemicals with high added value such as DHA, glyceric acid, glyceraldehyde, hydroxypyruvic acid, mesooxalic acid and tartronic acid.Particularly, DHA is one of the most valuable products due to its great demand in cosmetics [3,4].However, owing to the three hydroxyl groups in GLY, the selective oxidation of GLY to DHA is a crucial dilemma from the point of view of catalyst design.
In the past decades, there were many successful reports on the selective oxidation of GLY using noble metal catalysts [5][6][7][8][9][10][11][12].However, due to the high cost and easy catalyst deactivation of noble metal catalysts, the design of efficient transition metal catalysts was becoming a hot point in the current research area [1,[13][14][15][16][17][18][19][20].Zhou et al. reported that Cu-containing hydrotalcites were active catalysts in the selective oxidation of GLY to glyceric acid, and the highest yield reached 68% [1].Crotti et al. and Shul'pin et al. found that iron complexes and manganese complexes could catalytically oxidize GLY to DHA, respectively (Yield < 15%) [13,18].Although several promising transition metal catalysts have been reported, the activity and selectivity of catalysts still need to be improved.Thus, it was of significance to design highly effective catalysts, especially a low-cost heterogeneous catalyst, for the oxidation reaction of GLY.
LDHs were often considered as catalysts or catalyst supports owing to their adjustable surface basicity, high surface and thermal stability, along with the adjustable variability of their laminate cations, and the exchangeability of their interlayer anions [21][22][23][24].In our previous report, we prepared an LDH-hosted chromium complex, which was an active catalyst for the oxidation reaction of GLY to DHA with 3% H 2 O 2 [17].However, the structure-performance relationship of the catalyst was unclear.In the present investigation, we further prepared a series of LDH-hosted chromium complexes with different Schiff base ligands, metal centers and loadings of active complexes.Their catalytic performance was investigated extensively in the oxidation conversion of GLY using 3% H 2 O 2 as an oxidant.The effect of the structures, the composition of the catalysts, and the oxidation reaction conditions on the catalytic performance of the catalysts was discussed in detail.

Characterization of Catalysts
The results of elemental analysis revealed that the molar ratios of N to the transition metal in LDH-hosted complexes were in conformance with the calculated values according to Scheme 1 (see Table 1).This indicated that the obtained catalysts had the expected elemental composition.
Catalysts 2016, 6, 101 2 of 10 transition metal catalysts have been reported, the activity and selectivity of catalysts still need to be improved.Thus, it was of significance to design highly effective catalysts, especially a low-cost heterogeneous catalyst, for the oxidation reaction of GLY.LDHs were often considered as catalysts or catalyst supports owing to their adjustable surface basicity, high surface and thermal stability, along with the adjustable variability of their laminate cations, and the exchangeability of their interlayer anions [21][22][23][24].In our previous report, we prepared an LDH-hosted chromium complex, which was an active catalyst for the oxidation reaction of GLY to DHA with 3% H2O2 [17].However, the structure-performance relationship of the catalyst was unclear.In the present investigation, we further prepared a series of LDH-hosted chromium complexes with different Schiff base ligands, metal centers and loadings of active complexes.Their catalytic performance was investigated extensively in the oxidation conversion of GLY using 3% H2O2 as an oxidant.The effect of the structures, the composition of the catalysts, and the oxidation reaction conditions on the catalytic performance of the catalysts was discussed in detail.

Characterization of Catalysts
The results of elemental analysis revealed that the molar ratios of N to the transition metal in LDH-hosted complexes were in conformance with the calculated values according to Scheme 1 (see Table 1).This indicated that the obtained catalysts had the expected elemental composition.The FTIR spectra of LDH-[C 6 H 5 COO], Cr(SO 3 -salphen) and LDH-[Cr(SO 3 -salphen)] are typically illustrated in Figure 1, and the important diagnostic bands of the other LDH-hosted complexes were also assigned, as listed in Table 1.Cr(SO 3 -salphen) exhibited υ s (SO 3 ´) and υ as (SO 3 ´) at about 1039 and 1114 cm ´1, respectively.Simultaneously, the bands at about 1624 and 1520 cm ´1 could be due to υ(C=N) and υ(C-O), and the bands at about 605 and 412 cm ´1 were associated with υ(Cr-O) and υ(Cr-N), respectively [25][26][27].This indicated that the homogeneous complex had been prepared successfully.For the LDH-hosted complexes, almost all the above characteristic bands were observed clearly though they turned relatively weaker because of their low concentration, confirming that the transition metal complex was intact during intercalation.Furthermore, compared to LDH-[C 6 H 5 COO], the band of C 6 H 5 COO ´at about 1550 cm -1 was absent in the FTIR spectrum of the LDH-hosted complexes [25], further suggesting that the homogeneous complexes had been intercalated into the LDH interlayer by ion exchange.The FTIR spectra of LDH-[C6H5COO], Cr(SO3-salphen) and LDH-[Cr(SO3-salphen)] are typically illustrated in Figure 1, and the important diagnostic bands of the other LDH-hosted complexes were also assigned, as listed in Table 1.Cr(SO3-salphen) exhibited υs(SO3 − ) and υas(SO3 − ) at about 1039 and 1114 cm −1 , respectively.Simultaneously, the bands at about 1624 and 1520 cm −1 could be due to υ(C=N) and υ(C-O), and the bands at about 605 and 412 cm −1 were associated with υ(Cr-O) and υ(Cr-N), respectively [25][26][27].This indicated that the homogeneous complex had been prepared successfully.For the LDH-hosted complexes, almost all the above characteristic bands were observed clearly though they turned relatively weaker because of their low concentration, confirming that the transition metal complex was intact during intercalation.Furthermore, compared to LDH-[C6H5COO], the band of C6H5COO − at about 1550 cm -1 was absent in the FTIR spectrum of the LDH-hosted complexes [25], further suggesting that the homogeneous complexes had been intercalated into the LDH interlayer by ion exchange.XRD patterns of LDH-[C6H5COO] and LDH-[M(SO3-salphen)] (M: Cr, Mn, Fe, Co, Cu ) are typically illustrated in Figure 2. The samples all showed three sharp characteristic peaks of the (003), (006) and (110) planes, indicating the generation of a hydrotalcite-like structure [25,[28][29][30].Compared to the parent LDH-[C6H5COO], the (003) peaks of LDH-hosted complexes shifted to lower 2θ angles, which could be attributed to the increased interlayer distance originating from the large anion size of Schiff base complexes compared to that of C6H5COO − .XRD patterns of LDH-[C 6 H 5 COO] and LDH-[M(SO 3 -salphen)] (M: Cr, Mn, Fe, Co, Cu ) are typically illustrated in Figure 2. The samples all showed three sharp characteristic peaks of the (003), (006) and (110) planes, indicating the generation of a hydrotalcite-like structure [25,[28][29][30].Compared to the parent LDH-[C 6 H 5 COO], the (003) peaks of LDH-hosted complexes shifted to lower 2θ angles, which could be attributed to the increased interlayer distance originating from the large anion size of Schiff base complexes compared to that of C 6 H 5 COO ´.The N2 sorption isotherms of LDH-hosted complexes in Figure 3 showed type IV isotherms, indicating the formation of mesopores due to the aggregation of particles.Moreover, compared to their parent LDH-[C6H5COO], LDH-hosted complexes exhibited a significantly decreased surface area (see Table 1), which might be related to the intercalation of chromium complexes into the The N 2 sorption isotherms of LDH-hosted complexes in Figure 3 showed type IV isotherms, indicating the formation of mesopores due to the aggregation of particles.Moreover, compared to their parent LDH-[C 6 H 5 COO], LDH-hosted complexes exhibited a significantly decreased surface area (see Table 1), which might be related to the intercalation of chromium complexes into the interlayer of LDH.In addition, typically, the SEM image of LDH-[Cr(SO 3 -salphen)] is shown in Figure 4, and globular-type agglomerated crystals were found due to the introduction of transition metal complexes.The N2 sorption isotherms of LDH-hosted complexes in Figure 3 showed type IV isotherms, indicating the formation of mesopores due to the aggregation of particles.Moreover, compared to their parent LDH-[C6H5COO], LDH-hosted complexes exhibited a significantly decreased surface area (see Table 1), which might be related to the intercalation of chromium complexes into the interlayer of LDH.In addition, typically, the SEM image of LDH-[Cr(SO3-salphen)] is shown in Figure 4, and globular-type agglomerated crystals were found due to the introduction of transition metal complexes.

Catalytic Performance
In the absence of solvent and additive, the as-prepared catalysts were used in the selective oxidation of GLY using 3% H2O2 as an oxidant, and the results are listed in Table 2.It was found that no product was detected without catalyst, while LDH also exhibited a slight catalytic performance (see Entry 1~2).Over homogeneous catalysts, the GLY conversion increased significantly, but the main product was the over-oxidation product of formic acid (see Entry 3~6).Upon the homogeneous complexes being intercalated into LDH, the catalytic performance, especially the selectivity to DHA (a C3 oxygenated product of secondary alcohol) was further improved sharply (see Entry 7~17).Moreover, their catalytic performance increased with the loading of the neat complexes until the chromium content reached 6.30%; however, with the further increase in the loading of the neat complexes, the catalytic performance decreased (see Entry 10~13).It had been accepted that homogeneous complexes tended to deactivate due to the formation of dimers or oligomers originating from their high local concentration of active complexes [31].Thus, such obviously enhanced catalytic performance over the heterogenized catalysts could be attributed to the dispersion effect of the support.
Moreover, the LDH-hosted chromium catalysts with different Schiff base ligands were also

Catalytic Performance
In the absence of solvent and additive, the as-prepared catalysts were used in the selective oxidation of GLY using 3% H 2 O 2 as an oxidant, and the results are listed in Table 2.It was found that no product was detected without catalyst, while LDH also exhibited a slight catalytic performance (see Entry 1~2).Over homogeneous catalysts, the GLY conversion increased significantly, but the main product was the over-oxidation product of formic acid (see Entry 3~6).Upon the homogeneous complexes being intercalated into LDH, the catalytic performance, especially the selectivity to DHA (a C 3 oxygenated product of secondary alcohol) was further improved sharply (see Entry 7~17).Moreover, their catalytic performance increased with the loading of the neat complexes until the chromium content reached 6.30%; however, with the further increase in the loading of the neat complexes, the catalytic performance decreased (see Entry 10~13).It had been accepted that homogeneous complexes tended to deactivate due to the formation of dimers or oligomers originating from their high local concentration of active complexes [31].Thus, such obviously enhanced catalytic performance over the heterogenized catalysts could be attributed to the dispersion effect of the support.
Moreover, the LDH-hosted chromium catalysts with different Schiff base ligands were also found to show significant differences in their catalytic performance (see Entry 7~10 in Table 2).Based on the DHA yield, the catalytic performance of LDH-hosted complexes follows the trend of LDH-[Cr(SO 3 -salphen)] > LDH-[Cr(SO 3 -salen)] > LDH-[Cr(SO 3 -salan)] > LDH-[Cr(SO 3 -sahen)].The relatively higher catalytic performance of LDH-[Cr(SO 3 -salphen)] could be attributed to its bridge groups of phenyl groups, which led to the presence of π-extended coordination structures and the decrease in the system energy.Thus, the reactants and oxidant were easy to access at catalytic active sites.The worst catalytic performance over LDH-[Cr(SO 3 -sahen)] might be related to the chair conformation of the cyclohexyl bridge groups, which induced high steric encumbrance around the active center.The catalytic roles of the metal centers were also investigated.The results in Table 2 (Entry 10~17) and Figure 5 revealed that the type of metal cation remarkably influenced the catalytic performance of the catalysts.The GLY conversion decreased over the catalysts: LDH-[Cr(SO 3 -salphen)] > LDH-[Co(SO 3 -salphen)] > LDH-[Cu(SO 3 -salphen)] > LDH-[Fe(SO 3 -salphen)] > LDH-[Mn(SO 3 -salphen)], which was almost identical to the profile of H 2 O 2 efficiency.This indicated that the catalytic performance of LDH-hosted complexes was probably related to their H 2 O 2 efficiency.In our previous report, we found that O 2 could not oxidize GLY in the presence of an LDH-hosted complex and base-free reaction conditions, and the disproportionation decomposition of H 2 O 2 to O 2 was unproductive for GLY oxidation [17].Thus, the remarkably low catalytic performance of LDH-[Mn(SO 3 -salphen)] and LDH-[Fe(SO 3 -salphen)] could be related to their especially high activities to the disproportionation decomposition of H 2 O 2 .For the other three LDH-hosted complexes, the different catalytic performances were attributed mainly to the increased number of electrons in the three-dimensional (3D) electron orbital from Cr to Cu.Such an increase led to the decrease in the capacity of the metal centers to accept electrons of Schiff base ligands, and then to the decrease in the ability of catalysts to activate H 2 O 2 .Thus, the catalysts with metal centers from Cr to Cu exhibited gradually decreased H 2 O 2 efficiency and GLY conversion.
could be related to their especially high activities to the disproportionation decomposition of H2O2.For the other three LDH-hosted complexes, the different catalytic performances were attributed mainly to the increased number of electrons in the three-dimensional (3D) electron orbital from Cr to Cu.Such an increase led to the decrease in the capacity of the metal centers to accept electrons of Schiff base ligands, and then to the decrease in the ability of catalysts to activate H2O2.Thus, the catalysts with metal centers from Cr to Cu exhibited gradually decreased H2O2 efficiency and GLY conversion.The reaction conditions were further optimized to get the best reaction results with LDH-[Cr(SO3-salphen)] used as the representative catalyst.It was found that the dosage of oxidant and catalyst, the reaction temperature and time all played important roles in the catalytic performance of the obtained LDH-hosted complex catalyst (see Figure 6).With the increase of the reaction time and temperature as well as the oxidant dosage, the glycerol conversion increased continuously, probably because the contact probability of the reactants and oxidant to the active sites of the catalyst increased gradually in the present heterogenized catalytic system.However, The reaction conditions were further optimized to get the best reaction results with LDH-[Cr(SO 3 -salphen)] used as the representative catalyst.It was found that the dosage of oxidant and catalyst, the reaction temperature and time all played important roles in the catalytic performance of the obtained LDH-hosted complex catalyst (see Figure 6).With the increase of the reaction time and temperature as well as the oxidant dosage, the glycerol conversion increased continuously, probably because the contact probability of the reactants and oxidant to the active sites of the catalyst increased gradually in the present heterogenized catalytic system.However, excessive time, temperature and oxidant dosage exhibited an adverse effect on DHA selectivity, indicating the presence of over-oxidation.Interestingly, the GLY conversion and DHA selectivity both increased firstly and then decreased with the increase of the catalyst dosage.This indicated that excessive catalyst was unfavorable to the oxidation reaction due to the increased diffusion resistance and the resultant decreased accessibility of reactants to the active centers of the catalyst.Under the optimal reaction conditions (10 mmol GLY, 0.2 g catalyst, 30 mL 3% H 2 O 2 , 6 h and 60 ˝C), the best GLY conversion and DHA selectivity reached 85.0% and 56.5%, respectively.The present reaction results were better than the best results in the bibliography under the solvent and base-free conditions (GLY conversion of 71.3% and DHA selectivity of 43.5% reported by us previously [17]).
The reusability of heterogenized catalyst is another important performance evaluation index with the exception of its catalytic activity.Here, after the first catalytic run, the representative catalyst of LDH-[Cr(SO 3 -salphen)] was centrifugally separated from the reaction mixture, washed with water, and dried at 100 ˝C overnight.Then, the recovered catalyst was further used in another 10 catalytic runs, and only a slightly decreased catalytic performance was detected (see Figure 7).Elemental analysis showed that the chromium content in the recovered catalyst after 10 runs (Cr wt %: 6.30) was almost the same as the fresh catalyst (see Table 1), suggesting that no significant leaching of active metal was present.Thus, the LDH-hosted complex was a stable heterogenized catalyst for the selective oxidation of GLY.
that excessive catalyst was unfavorable to the oxidation reaction due to the increased diffusion resistance and the resultant decreased accessibility of reactants to the active centers of the catalyst.Under the optimal reaction conditions (10 mmol GLY, 0.2 g catalyst, 30 mL 3% H2O2, 6 h and 60 °C), the best GLY conversion and DHA selectivity reached 85.0% and 56.5%, respectively.The present reaction results were better than the best results in the bibliography under the solvent and base-free conditions (GLY conversion of 71.3% and DHA selectivity of 43.5% reported by us previously [17]).The reusability of heterogenized catalyst is another important performance evaluation index with the exception of its catalytic activity.Here, after the first catalytic run, the representative catalyst of LDH-[Cr(SO3-salphen)] was centrifugally separated from the reaction mixture, washed with water, and dried at 100 °C overnight.Then, the recovered catalyst was further used in another 10 catalytic runs, and only a slightly decreased catalytic performance was detected (see Figure 7).Elemental analysis showed that the chromium content in the recovered catalyst after 10 runs (Cr wt %: 6.30) was almost the same as the fresh catalyst (see Table 1), suggesting that no significant leaching of active metal was present.Thus, the LDH-hosted complex was a stable heterogenized catalyst for the selective oxidation of GLY.

Catalyst Preparation
According to the similar preparation process of LDH-hosted Cr(salen) complexes in our previous reports [17,25], here a series of LDH-hosted transition metal complexes were further synthesized by changing the kinds of ligands and metal centers (see Scheme 1).Basically, the Schiff

Catalyst Preparation
According to the similar preparation process of LDH-hosted Cr(salen) complexes in our previous reports [17,25], here a series of LDH-hosted transition metal complexes were further synthesized by changing the kinds of ligands and metal centers (see Scheme 1).Basically, the Schiff base ligands were prepared firstly by the condensation reaction of salicylaldehyde and different diamines (ethyldiamine, propane diamine, hexamethylene diamine and o-phenylenediamine), and the molar ratio of salicylaldehyde and amines was 2.Then, the obtained Schiff base ligands were mixed with concentrated sulfuric acid to synthesize the sulphonated ligands, and the mass ratio of ligands to sulfuric acid was 5.Moreover, the ion exchange reaction was performed between the sulfonated ligands and the prepared Mg-Al LDH containing C 6 H 5 COO ´anions in advance.Furthermore, the solid products were added into the aqueous solution of metallic salts (MnCl 2 ¨4H 2 O, FeCl 3 ¨6H 2 O, CrCl 3 ¨6H 2 O, NiCl 2 ¨6H 2 O, CuCl 2 ¨6H 2 O) under stirring and N 2 atmosphere to obtain the LDH hosted transition metal complexes.

Characterization
The C, H, O, N and S contents of samples were detected on Vario EL analyzer (Elementar, Hannover, DE-NI, Germany).The contents of Mg, Al and transition metals were determined using PerkinElmer ICP OPTIMA-3000 (Thermo Electron, Waltham, MA, USA).Nitrogen sorption isotherms were measured at ´196 ˝C on Micromeritics ASAP-2000 (Micromeritics, Norcross, GA, USA) by static adsorption procedures, and surface areas were obtained by the BET method.Powder X-ray diffraction (XRD) experiments were tested under ambient condition on Rigaku D Max III VC (Rigaku, Tokyo, Japan) using a Cu target with a Ni filter in a 2θ range of 5 ˝-70 ˝at 30 mA and 50 kV.The Fourier transform infrared (FTIR) spectra were gained on Bruker Tensor-27 FTIR spectrophotometer (Bruker, Ettlingen, DE-BW, Germany).SEM experiments were performed on JEOL JSM-7600F microscope (Hitachi, Akishima, Japan).

Catalytic Test
At atmospheric pressure, the GLY oxidation was carried out in a 100 mL three neck flask equipped with a constant temperature magnetic agitator (DF-II) (Shengwei, Jintan, China).Typically, a certain amount of catalyst was suspended in 25 mL aqueous glycerol solution (0.4 mol L ´1).The above mixture was heated to the required temperature, and then 3% H 2 O 2 was cautiously added, dropwise.After the reaction was performed for a desired time, catalyst was removed by centrifugal separation.The residual aqueous solution was neutralized by adding sulfuric acid, and then was send to a high performance liquid chromatography (Agilent, Santa Clara, CA, USA) with UV-vis detectors and refractive index.H 2 O 2 consumption was determined by iodometry after the reactions.Aminex HPX-87H column (Bio-Rad) (Bio-Rad, Philadelphia, PA, USA) was used for product seperation at 333 K with H 2 SO 4 (0.01 mol/L, 0.5 mL/min) as eluent flowing.A 10 µL injection and a 30 min measure time were need.The calibration curves and retention times were determined using the standard samples with known concentrations.

Conclusions
LDH-hosted complex catalysts were found to be effective catalysts for the GLY oxidation reaction.Moreover, the metal centers, the Schiff base ligands, the loading of complexes on the support and the reaction conditions significantly influenced the performance of the catalysts.In the presence of 3% H 2 O 2 , the main product was C 3 oxygenated products of secondary alcohol, DHA.The GLY conversion and DHA selectivity reached 85.0% and 56.5%, respectively, over LDH-[Cr(SO 3 -salphen)] which exhibited the highest H 2 O 2 efficiency.

Table 1 .
Components, textural parameters and spectroscopic data of catalysts.

Table 2 .
Catalytic performance of catalysts in GLY (glycerol) oxidation with 3% H 2 O 2 a .