recent advances in the catalytic synthesis ... - nsfc.gov.cn

Recent Advances in the Catalytic Synthesis of 2,5-FurandicarboxylicAcid and Its DerivativesZehui Zhang* and Kejian Deng

Key Laboratory of Catalysis and Materials Sciences of the State Ethnic Affairs Commission and Ministry of Education, College ofChemistry and Material Science, South-Central University for Nationalities, Wuhan 430074, P. R. China

ABSTRACT: Catalytic synthesis of value-added chemicalsfrom renewable biomass or biomass-derived platform chem-icals is an important way to reduce current dependence onfossil-fuel resources. In recent years, 2,5-furandicarboxylic acid(FDCA) has received significant attention due to its wideapplication in many fields, particularly as a substitute ofpetrochemical-derived terephthalic acid in the synthesis ofuseful polymers. Therefore, much effort has been devoted tothe catalytic synthesis of FDCA. In this critical review, we willprovide an overview of concise and up-to-date methods for thesynthesis of FDCA from HMF oxidation or directly fromcarbohydrates by one-pot reaction, giving special attention tocatalytic systems, mechanistic insight, reaction pathway, andcatalyst stability. In addition, the one-pot oxidative conversion of carbohydrates into FDCA and the one-pot synthesis of FDCAderivatives are also discussed. It is anticipated that the chemistry detailed in this review will guide researchers to develop effectivecatalysts for the economical and environmentally friendly synthesis of FDCA in large-scale.

KEYWORDS: biomass, carbohydrates, 2,5-furandicarboxylic acid, 5-hydroxymethylfurfural, catalytic oxidation

1. INTRODUCTION

The worldwide consumption of petrochemical products andfossil fuels is increasing rapidly.1 Therefore, there is an urgentneed to produce fuels and chemicals from renewableresources.2 Biomass is one of the most abundant renewableresources with annual production of 170 billion metric tons,and it has been considered as a unique and promisingcandidate.3 In addition, biomass also plays a crucial role incarbon balance, as the generated CO2 output can be offset byCO2 fixation through photosynthesis of plant growth.4

Conversion of biomass into fuels and chemicals, generallycalled “biorefinery technology”, has existed for centuries. Mucheffort is being devoted to develop new methods to convertbiomass into valuable chemicals and fuels using variousmethods.5−8 Catalytic oxidation of biomass or biomass-basedchemicals has received great attention for the production ofvalue-added chemicals in recent years.9−11 5-Hydroxymethyl-furfural (HMF), the dehydration product of C6 carbohydrates,has been considered as a key platform molecule for theproduction of a wide variety of commodity chemicals.12 Forexample, as shown in Figure 1, selective oxidation of HMF cangenerate several kinds of important furanic chemicals such asmaleic anhydride (MA), 2,5-diformylfuran (DFF), 5-hydrox-ymethyl-2-furancarboxylic acid (HFCA), and 2,5-furandicarbox-ylic acid (FDCA).13−15

Among them, FDCA is listed as one of the top-12 value-added chemicals from biomass by the U.S. Department ofenergy.15 FDCA is very stable with a high melting point at 342

°C16 and insoluble in most common solvents. FDCA has beenfound to be useful in many fields. The most importantapplication of FDCA is that it can serve as a polymer buildingblock for the production of biobased polymers such aspolyamides, polyesters, and polyeurethanes.17−19 The mostattractive way is that FDCA can replace the petrochemical-derived terephthalic acid for the synthesis of biobasedpolyesters. Terephthalic acid has been used as a monomer inpolyethylene terephthalate (PET) plastics for a long time.20

PET is usually used for the production of films, fibers, and inparticular bottles for the packaging of soft drinks, water, andfruit juices. One promising biobased polymer to PET ispolyethylene furanoate (PEF), which is the esterificationproduct of ethane-1,2-diol and FDCA.21−23 PEF demonstratessimilar properties as the petroleum-based PET. The Coca Colacompany has collaborated with Avantium, Danone, and ALPLAto develop and commercialize PEF bottles. Their research hasshown that PEF bottles outperform PET bottles in many areas.Besides the main application as monomer for the production ofbiobased polymer, FDCA has also been found to be useful inorganic synthesis, pharmacology, and metal−organic frameworkmaterials.24−26

In view of its wide application, catalytic synthesis of FDCAhas been extensively studied. The focus of this review is to

Received: July 15, 2015Revised: September 9, 2015

Review

pubs.acs.org/acscatalysis

© XXXX American Chemical Society 6529 DOI: 10.1021/acscatal.5b01491ACS Catal. 2015, 5, 6529−6544

pubs.acs.org/acscatalysis

http://dx.doi.org/10.1021/acscatal.5b01491

summarize the most recent findings with a critical discussion ofthe catalytic oxidation of HMF into FDCA. Recent work on theconversion of carbohydrates into FDCA as well as FDCAderivatives are also discussed briefly.

2. FDCA PRODUCTION USING DIFFERENT METHODSIN THE PAST

FDCA was first produced from the dehydration of mucic acidin the presence of strong acid catalyst (Figure 2). In 1876, Fittig

and Heinzelman first reported that the dehydration of mucicacid could produce FDCA using 48% aqueous HBr as thecatalyst and the solvent.27 Later on, other modified methods bythe changing of dehydrating agents were also reported for thesynthesis of FDCA. However, the dehydration of mucic acid toFDCA required severe conditions (highly concentrated acids,high reaction temperature 120 °C, long reaction time >20 h)and all the methods were nonselective with yields <50%.25

More importantly, mucic acid itself is a rare organic acid and itsprice is high. Therefore, this method has never been studied forthe synthesis of FDCA in modern times.The other method for the synthesis of FDCA was via the

oxidation of furfural or HMF by using inorganic oxidants. Asshown in Figure 3, several steps are required to achieve FDCAusing furfural as the starting material.28 Furfural was firstoxidized to 2-furoic acid with nitric acid and the latter was thenconverted to its methyl ester. The ester underwent the

chloromethylation reaction at position C5 to give methyl 5-chloromethylfuroate. The oxidation of 5-chloromethylfuroatewith nitric acid afforded dimethyl 2,5-furandicarboxylate,followed by the hydrolysis to give FDCA.Procedure of the synthesis of FDCA from furfural is complex,

and the total FDCA yield is generally low after so manyreaction steps. The simple method for the synthesis of FDCA isvia the direct oxidation of HMF. In the past decades, theoxidation of HMF into FDCA was performed usingstoichiometric oxidants such as KMnO4.

29 These methodsdemonstrated some distinct drawbacks, such as the lowselectivity, the high cost of the oxidant, and the generation ofhigh toxic waste to the environment.

3. CURRENT METHODS FOR THE OXIDATION OF HMFINTO FDCA

As mentioned in Section 2, those out-of-date methods arecontrary to the concept of green chemistry and show littlesignificance in practice. Thus, many environmentally friendlyand economical methods are currently being developed for theoxidation of HMF into FDCA.

3.1. Electrocatalytic Synthesis of FDCA from HMF.Electrochemical oxidation is driven by the electrochemicalpotential of the electrode. The electrochemical oxidation isrealized through the electron transfer, which eliminates the useof O2 or other chemical oxidants. Thus, electrochemicaloxidation has been considered as a clean synthetic methodand has received great interest.30,31 Electrochemical oxidationcan also offer the advantage of controlling the oxidationpotential and the faradaic current, which can be used tomonitor the thermodynamic driving force, the selectivity of thesurface reaction, and the reaction rate.In 1995, the electrochemical oxidation of HMF into FDCA

was first reported in a H-shaped cell.32 HMF electrochemicaloxidation occurred near an anode (nick oxide/hydroxide as theanode material), affording FDCA in a yield of 71% after 4 h in 1M NaOH solution at the current density of 0.016 A cm−2.However, the electrochemical oxidation of HMF has beenscarcely explored after that work, possibly due to the fact thatthe importance of FDCA has not been well recognized by theresearchers. Until recently, the electrochemical oxidation ofHMF has freshly received attention. There are a few reports onthe electrochemical oxidation of HMF into FDCA. Strasser andco-workers studied the electrochemical oxidation of HMF usinga Pt electrode at pH 10.33 They found that a fraction of HMFwas oxidized into DFF at the current density of 0.44 mA cm−2.However, FDCA was obtained with a negligible yield (less than1%). The authors claimed that water oxidation was the majorcompeting reaction and probably limited the Faradaic efficiency

Figure 1. Schematic illustration of the potential oxidation products from HMF.

Figure 2. Schematic illustration of the dehydration of mucic acid intoFDAC.

Figure 3. Synthesis of FDCA from furfural via multiple reactionsteps.28

ACS Catalysis Review

DOI: 10.1021/acscatal.5b01491ACS Catal. 2015, 5, 6529−6544

6530


for HMF oxidation. Later, Li and co-workers studied theelctrocatalytic oxidation of HMF in alkaline solution overcarbon-black-supported noble-metal catalysts.34 They foundthat the reaction was affected by the potential and theelectrocatalyst (shown in Figure 4). The oxidation of aldehyde

group in HMF was much easier than the oxidation of alcoholgroup over Au/C catalyst, affording HFCA as the intermediate,but high electrode potential was required for further oxidationof the alcohol group in HFCA to FDCA. As far as the oxidationof HMF over Pd/C catalyst, two competitive routes wereobserved for the oxidation of HMF into the intermediateFFCA, which depended on the electrode potential. Oxidationof aldehyde groups occurred much slower on Pd/C than onAu/C at low potentials, but was greatly enhanced at increasedpotentials. It was found that Pd−Au bimetallic catalystsachieved deeply oxidized products (FFCA and FDCA) atlower potentials than monometallic catalysts and the productdistribution depended on the electrode potential and surfacealloy composition. Bimetallic PdAu2/C catalyst significantlyenhanced the efficiency of the electrochemical oxidation ofHMF, affording full HMF conversion and a FDCA yield of 83%at the potential of 0.9 V, much higher than the monometalliccatalyst, which was due to alloy effect. However, FDCA wasobtained with other oxidation intermediates, mainly HFCA,making it difficult to purify the main product from the liquidsolution.Recently, Choi and co-workers made a great process in the

electrochemical oxidation of HMF with high efficiency.35 Asreported by Strasser and co-workers,33 the low efficiency of theelectrochemical oxidation of HMF is the competitive oxidationof water to O2. The authors found that the necessaryoverpotential to initiate HMF oxidation was significantlyreduced using 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO)as the mediator, which inhibited the oxidation of water. Thereaction mechanism is illustrated in Figure 5, and theelectrochemical cell is shown in Figure 6a. TEMPO is oxidizedto its oxidation state of TEMPO+ in the vicinity of the Auelectrode surface, which serves as a mediator and catalyst forHMF oxidation. High FDCA yield (⩾99%) and Faradaicefficiency (⩾93%) were obtained in a pH 9.2 aqueous medium.Kinetic study and cyclic voltammetry indicated that DFF wasthe reaction intermediate.Furthermore, the authors constructed a photoelectrochem-

ical cell (PEC) (Figure 6b) that used TEMPO-mediatedphotoelectrochemical oxidation of HMF as the anode reaction.In this cell, an n-type nanoporous BiVO4 electrode was used asa photoanode that absorbed photons to generate and separate

electron−hole pairs. After separation, electrons were transferredto the Pt counter electrode to reduce water to H2 eq 1),whereas the holes that reached the surface of BiVO4 were usedfor TEMPO-mediated HMF oxidation (eq 2. The overallreaction achieved by this PEC is shown in eq 3. Thephotoelectrochemical method also generated high FDCA yields(⩾99%) and Faradaic efficiency (⩾93%). This method did notrequire an adjustment of the pH during the reaction process asHMF oxidation offset the pH change at the cathode. Thismethod not only afforded a high FDCA yield but alsosimultaneously produced H2 as a clean energy source.

+ → +− −cathode reaction: 6H O 6e 3H 6OH2 2 (1)

+ → + +− −anode reaction: HMF 6OH FDCA 4H O 6e2(2)

+ → +overall: HMF 2H O FDCA 3H2 2 (3)

Although some excellent results on the synthesis of FDCAwere obtained from electrocatalytic method, this method stillremains some problems in the practical synthesis of FDCA.High FDCA yield was obtained at the expense of the additionof high amount of TEMPO (1.5 equiv of TEMPO),35 and thus,the cost of the synthesis of FDCA was high.35 Meanwhile, it isdifficult to separate FDCA from TEMPO or the byproductsand the electrolyte. In addition, the initial HMF concentrationin the reported methods was low in order to achieve highFDCA yield. Therefore, it is very crucial to design the robustelectrocatalysts that will promote the oxidation of HMF withfull conversion at high concentration and 100% selectively ofFDCA.

3.2. Biocatalyst Method for the Synthesis of FDCAfrom HMF. Chemical oxidation reactions are typicallyperformed at high temperature and high pressure. In contrastto chemical processes, biocatalytic transformation is typicallycarried out under relatively mild conditions and usually requiresfewer and less toxic chemicals.36 Despite these evidentadvantages, biocatalytic approaches to FDCA production areless well established.Mara and co-worker reported that DFF could be successfully

oxidized to FDCA by the in situ produced peracids, which wasformed in the presence of lipases as biocatalysts.37 Using lipasesas biocatalysts, alkyl esters as acyl donors, and aqueoussolutions of hydrogen peroxide (30% v/v) added stepwise,peracids were formed in situ, which subsequently oxidized DFFto afford FDCA with high yield (>99%) and excellentselectivity (100%) (Figure 7). However, this method wasinactive for HMF. The use of DFF as a feedstock for the

Figure 4. Proposed reaction pathways of HMF oxidation on Pd/C andAu/C electrocatalysts in alkaline media.34

Figure 5. Scheme of electrocatalytic oxidation of HMF into FDCAwith TEMPO as the mediator.35



6531


synthesis of FDCA requires an additional step of the oxidationof HMF into DFF. Thus, this method was a high cost way toproduce FDCA. A chloroperoxidase from Caldariomyces fumagowas found to have the biocatalytic activity toward the oxidationof HMF into FDCA.38 However, this method could notprovide a complete HMF oxidation, affording FDCA yields of60−75% as well as 25−40% yields of HFCA. This propertyrenders the C. fumago chloroperoxidase a poor biocatalyst forFDCA production, especially when FDCA of very high purity isrequired for specific applications such as for polymermanufacture. Later, a fermentative process using Pseudomonasputida S12 to host the oxidoreductase from Cupriavidusbasilensis HMF14 was studied for the oxidation of HMF intoFDCA.39 In fed-batch experiments using glycerol as the carbonsource, 30.1 g L−1 of FDCA was produced from HMF with ayield of 97%. FDCA was recovered from the culture broth as a99.4% pure dry powder, at 76% recovery using acidprecipitation and subsequent tetrahydrofuran extraction. Thisprocess relies on the activities of both oxidases anddehydrogenases. Most enzymes are restricted to either alcoholor aldehyde oxidations, while the full oxidation of HMF toFDCA requires the enzyme to act on both alcohol and

aldehyde groups. Recently, Fraaije and co-workers identified anFAD-dependent oxidase of the glucose-methanol-cholineoxidoreductase (GMC) family, named HMF oxidase(HMFO), which showed high catalytic activity toward theHMF oxidation.40,41 FDCA yield up to 95% with full HMFconversion was achieved at ambient pressure and temperaturebut requiring a long reaction time of 24 h at a low HMFconcentration of 2 mM. Experiments confirmed that theoxidation of HMF by this method underwent two routes(Figure 8). The reaction rate was controlled by the final step ofthe oxidation of FFCA to FDCA.The conditions of biocatalytic oxidation are mild (room

temperature and ambient atmosphere), but a long reaction timein a low concentration of HMF is required. Constructingenzymes by genetic engineering with high catalytic activity andstability that can oxidize HMF fast at high concentration willmake the biocatalytic method much more competitive in thelarge-scale production of FDCA. It would also be exciting toconstruct a microorganism with multiple enzymatic activitiesthat can use carbohydrates directly to produce FDCA, whichwill greatly decrease the production cost of FDCA.

3.3. Chemical Synthesis of FDCA from HMF byHomogeneous Catalyst. Although some new methodssuch as electrochemical oxidation and biocatalytic oxidationwere reported for the synthesis of FDCA, synthesis of FDCAwas mainly carried out via chemical catalytic method usinghomogeneous catalysts or heterogeneous catalysts. Comparedwith heterogeneous catalysts, there were fewer reports on thesynthesis of FDCA using homogeneous catalysts. In 2001,Partenheimer and Grushin studied the aerobic oxidation ofHMF to FDCA at 125 °C under 70 bar air pressure in aceticacid solvent using Co(OAc)2, Mn(OAc)2, and HBr as thecatalysts, commonly known as the Amoco Mid-Century (MC)

Figure 6. Schematic comparison of the photoelectrochemical and electrochemical cells. a, Photoelectrochemical TEMPO-mediated HMF oxidation.b, Electrochemical TEMPO-mediated HMF oxidation. CB, conduction band; EF, Fermi energy.35

Figure 7. Envisaged lipase-catalyzed peracid formation to perform achemo-enzymatic oxidation of DFF into FDCA.37

Figure 8. Reaction routes of the enzymatic oxidation of HMF into FDCA by HMF oxidase.39



6532


catalyst.42 Similar to the para-xylene oxidation to terephthalicacid over MC catalyst, the oxidation of HMF with the MCcatalyst also proceeded via the formation of peroxyl radical inthe chain propagation step. The peroxyl radicals were formedthrough the abstraction of H atom of HMF by the bromideradical, generated in the catalytic cycle by the oxidation of HBrwith Co(III) or Mn(III), followed by reaction of arylalkylradical with O2. Although both hydroxymethyl and aldehydegroups of HMF could be simultaneously oxidized, the authorsclaimed that hydroxymethyl group might be possiblypreferentially oxidized first. This method afforded FDCA witha yield of 60.9% together with other byproducts. Similarhomogeneous catalysts of Co(OAc)2/Zn(OAc)2/NaBr werelater used for the aerobic oxidation of HMF into FDCA.43 DFFwas observed as the sole oxidation product without an acidadditive, but FDCA was obtained in a yield of 60% withtrifluoroacetic acid as an additive.Besides the use of oxygen as the oxidant, t-BuOOH was also

used as the oxidant for the oxidation of HMF into FDCA.Riisager and co-workers studied the oxidation of HMF intoFDCA in acetonitrile with t-BuOOH as the oxidant and coppersalts as the catalyst.44 The use of CuCl together with LiBr as theadditive afforded FDCA in a low yield of 43% after 48 h atroom temperature and that was 45% using CuCl2 as the catalystwithout additive. The homogeneous reaction systems sufferedfrom two distinct drawbacks in practical applications. On theone hand, FDCA yield was relatively low, accompanied by theformation of some byproducts in the reaction solution. On theother hand, it is difficult to recycle the homogeneous catalystsand also purify FDCA from the metal salts. The use ofheterogeneous catalysts can overcome the drawbacks caused bythe homogeneous catalysts, as heterogeneous catalysts can befacilely separated from the reaction solution.3.4. Catalytic Synthesis of FDCA from HMF by

Supported Noble Metal Catalysts. Catalytic aerobicoxidation of HMF into FDCA has been extensively studiedover various heterogeneous catalysts. The use of molecular

oxygen as the oxidant is cheap and environmentally friendly aswater is the only reduction product. The heterogeneouscatalysts are easily recycled and reused. Thus, the use ofoxygen and the heterogeneous catalysts is in accordance withthe concept of “green and sustainable chemistry”, which are themain method for the synthesis of FDCA from HMF. As oxygenis not easy to be activated, supported Pt, Pd, Au, and Rucatalysts with high activity were the main heterogeneouscatalysts for the oxidation of HMF into FDCA.

3.4.1. Synthesis of FDCA from HMF over Supported PtCatalysts. Compared with Pd, Au, and Pd catalysts, supportedPt catalysts were found to be active toward the aerobicoxidation of HMF into FDCA in the earliest. In 1983,Verdeguer and co-workers studied the oxidation of HMF overPt/C catalyst.45 The authors found that the addition of Pbenhanced the catalytic activity greatly. Under the reactionconditions (1.25 M NaOH solution, 25 °C, with O2 flow rate at2.5 mL/s), FDCA was obtained in a high yield of 99% within 2h over Pt−Pb/C catalyst, whereas the Pt/C catalyst onlyproduced FDCA with a yield of 81% at HMF conversion of100% (Table 1, Entries 1 and 2). HFCA was the reactionintermediate, suggesting the oxidation of formyl group wasmuch easier than the oxidation of hydroxymethyl group overPt−Pb/C catalyst. Besides the additive of Pb, Bi was alsoobserved to show a positive effect on the catalytic performanceof Pt/C catalyst.46 The Pt−Bi/C catalyst with a Pt−Bi molarratio of ca. 0.2 afforded FDCA in a high yield of 98% after 6 hat 100 °C under 40 bar air by the use of 4 equiv of NaHCO3,while that was 69% for Pt/C catalyst (Table 1, Entries 3 and 4).Observed initial intermediates were HFCA and DFF, whichwere rapidly oxidized to FFCA, and the oxidation of FFCA wasthe rate-determining step. In addition, the Pt−Bi/C catalystdemonstrated much higher stability than that of Pt/C catalyst,as the addition of Bi increased the resistance to oxygenpoisoning and prevented the Pt leaching. Recently, the samegroup also observed the superior catalytic activity and stabilityof Pt−Bi/TiO2 to Pt/TiO2 catalyst (Table 1, Entries 5 and

Table 1. Results of the Oxidation of HMF into FDCA over Supported Pt Catalysts

entry catalyst oxidant base T (°C) time (h) HMF concn (%) FDCA yield (%) ref

1 Pt−Pb/C 1 bar O2 1.25 M NaOH 25 2 100 99 452 Pt/C 1 bar O2 1.25 M NaOH 25 2 100 81 453 Pt−Bi/C 40 bar air 2 equiv of Na2CO3 100 6 100 >99 464 Pt/C 40 bar air 2 equiv of Na2CO3 100 6 99 69 465 Pt−Bi/TiO2 40 bar air 2 equiv of Na2CO3 100 6 >99 99 476 Pt/TiO2 40 bar air 2 equiv of Na2CO3 100 6 90 84 477 Pt/RGO 1 atm O2 5 equiv of NaOH 25 24 100 84 488 Pt/C 690 kPa O2 2 equiv of NaOH 22 6 100 79 499 Pt/γ-A12O3 0.2 bar partial O2 pH = 9 60 6 100 99 5010 Pt/γ-A12O3 1 bar O2 1 equiv of Na2CO3 75 12 96 96 52

140 1211 Pt/ZrO2 1 bar O2 1 equiv of Na2CO3 75 12 100 94 52

140 1212 Pt/C 1 bar O2 1 equiv of Na2CO3 75 12 100 89 52

140 1213 Pt/TiO2 1 bar O2 1 equiv of Na2CO3 75 12 96 2 52

140 1214 Pt/CeO2 1 bar O2 1 equiv of Na2CO3 75 12 100 8 52

140 1215 Pt/Ce0.8Bi0.2O2‑δ 1.0 MPa O2 4 equiv of NaOH 23 0.5 100 98 5316 Pt/CeO2 1.0 MPa O2 4 equiv of NaOH 23 0.5 100 20 5317 Pt/PVP 1 bar O2 no base 80 24 100 94 55



6533


6).47 Besides active carbon, reduced graphene oxide (RGO) hasalso been deemed as an excellent support due to its abundantsurface functional groups to anchor the metal nanoparticles.Tsubaki and co-workers studied RGO-supported metal nano-particles for the oxidation of HMF at 25 °C with 5 equiv ofNaOH at an O2 flow rate of 50 mL/min.48 At the reaction timeof 6 h, both Pt/RGO and Pd/RGO afforded 100% HMFconversion, and produced FDCA with yields of 40.6% and30.5%, respectively, while Ru/RGO and Rh/RGO could notafford FDCA. These results suggested that the catalytic activityof Pt catalysts was higher than those of Pd, Ru, and Rhcatalysts. Prolonging the reaction time to 24 h, 84% FDCAyield was obtained over Pt/RGO catalyst (Table 1, Entry 7).HFCA was observed as the only intermediate. The Pt/RGOcatalyst could be reused with full HMF conversion in each run,but with a slight decrease of FDCA yield and a slight increase ofHFCA yield. Similar to the results reported by Tsubaki and co-workers,48 Davis and co-workers also observed that the carbon-supported Pt catalyst (Pt/C) showed a little higher catalyticactivity than carbon-supported Pd catalyst (Pd/C).49 Under thesame reaction conditions (690 kPa O2, 2 equiv of NaOH, 22°C), HMF was completely converted over the two catalysts, butPt/C catalyst produced higher FDCA yield (79%) after 6 h(Table 1, Entry 8) than that of Pd/C catalyst (71%).Besides carbon-material-supported metal catalysts, metal-

oxide-supported Pt catalysts were also studied for the oxidationof HMF into FDCA. In 1990, Vinke et al. found that the Pt/γ-A12O3 catalyst showed high catalytic activity toward theoxidation of HMF into FDCA.50,51 Near quantitative FDCAyield (>99%) was obtained after 6 h at pH = 9, 60 °C, and 0.2bar partial oxygen (Table 1, Entry 9). However, the Pt/A12O3catalyst deactivated by oxygen chemisorptions due to the highactivity of Pt nanoparticles. Recently, Sahu and Dhepecompared the catalytic performance of various metal-oxide-supported Pt catalysts.52 Under the same reaction conditions,Pt/γ-A12O3, Pt/ZrO2, and Pt/C showed high catalytic activitywith FDCA yields being 96%, 94%, and 89%, respectively,whereas Pt/TiO2 and Pt/CeO2 catalyst produced very poorFDCA yields (2 and 8%) even with HMF conversion of ca.100% (Table 1, Entries 10−14). These results suggested thatnonreducible oxides (γ-A12O3, ZrO2, and C) supported Ptcatalysts demonstrated higher catalytic performance thanreducible oxides (TiO2 and CeO2) supported Pt catalysts.The author claimed that the main reason was attributed to thedifferent oxygen storage capacity (OSC) of each support. It isknown that the OSC of the catalysts like Pt/γ-Al2O3, Pt/ZrO2,and Pt/C is quite low, which keeps the active sites in activeform (not covered by oxygen), whereas TiO2 and CeO2 have ahigh OSC because of the presence of Ce4+/Ce3+ or Ti4+/

Ti3+redox couple. Although the high OSC is good for otherreactions such as CO oxidation, it is detrimental in HMFoxidation. Recently, similar to the addition of Bi to Pt for Pt/Ccatalyst,46 Yang and co-workers found the addition of Bi to theCeO2 support could also greatly improved the catalyticperformance of Pt/CeO2.

53 The Pt/Ce0.8Bi0.2O2−δ afforded100% HMF conversion and 98% yield of FDCA, whereas HMFconversion was less than 20% over Pt/CeO2 catalyst (Table 1,Entries 15 and 16). As shown in Figure 9, Pt nanoparticles reactwith hydroxyl group in HMF to form the Pt−alkoxideintermediate, followed by β-H elimination with the help ofhydroxide ions. Bi-containing ceria accelerates the oxygenreduction process because of the presence of a large amount ofoxygen vacancies and the cleavage of the peroxide intermediatepromoted by bismuth. Thus, the surface electrons areconsumed to reduction oxygen, and the catalytic cycle can besmoothly continued. In addition, the Pt/Ce0.8Bi0.2O2−δ catalystcould be reused for five runs without the loss of its catalyticperformance (FDCA yield 98% in the first run vs 97% in thefifth run). In order to produce FDCA in a large scale, Lilga et al.performed the oxidation of HMF into FDCA over Pt/C andPt/Al2O3 catalysts in a continuous reactor, and nearlyquantitative yields of FDCA were obtained over the twocatalysts by the feed of 1 wt % HMF at 100 °C under 7 bar air(LHSV = 4.5 h−1, GHSV = 600 h−1) using stoichiometricaqueous Na2CO3.

54

As listed above, the aerobic oxidation of HMF over Ptcatalysts are generally carried out in the presence of excess base.The disadvantages of basic feeds are that product solutionsmust be neutralized, and inorganic salts must be separated out.More recently, Yan and co-workers found that PVP-stabilizedPt nanoparticles (Pt/PVP) could promote the base-free aerobicoxidation of HMF into FDCA in water. 100% conversion ofHMF and 95% yield of FDCA were obtained at 80 °C after 24h under 1 bar O2. It is observed that PVP/Pt catalyst showed aslight decrease of its catalytic activity in the recycling runs.Although this base-free method is environmental-friendly, ahigh catalyst loading of 5 mol %, a low content of the feedstock(0.29 mmol in each run) and a long reaction time of 24 h wereneeded to achieve 84% yield of FDCA, which also made thecost of the production of FDCA high.

3.4.2. Synthesis of FDCA from HMF over Supported PdCatalysts. Supported Pd catalysts also showed excellentcatalytic performance toward the aerobic oxidation of HMFinto FDCA. As described above, Davis and co-workers foundthe catalytic activity of Pd/C catalyst was comparable to the Pt/C catalyst.49 Under the reaction conditions (690 kPa O2, 2equiv of NaOH, and 23 °C), Pd/C catalyst afforded HMFconversion of 100% and FDCA yield of 71% after 6 h, and Pt/

Figure 9. Proposed reaction mechanism for the oxidation of HMF in alkaline aqueous solution. CeBi* represented the oxygen vacancy accompaniedby the bismuth.53



6534


C catalyst gave HMF conversion of 100% and FDCA yield of79% (Table 2, Entry 1 vs Table 1, Entry 4). Later, DaSiyo andco-workers studied the aerobic oxidation of HMF over PVP-stabilized Pd nanoparticles (Pd/PVP).56 Pd/PVP wereprepared in ethylene glycol with the addition of NaOH, andthe particle size could be controlled by the amount of NaOH. Itwas found that Pd/PVP catalyst with smaller Pd nanoparticlesize afforded FDCA with higher yield. Under the reactionconditions (90 °C, O2 flow rate at 35 mL/min, 1.25 equiv ofNaOH), the maximum FDCA yield of 90% with full HMFconversion was obtained after 6 h with Pd diameter of 1.8 nm(Table 2, Entry 2). FDCA yield decreased to 81% when the Pddiameter was 2.0 nm. The higher catalytic activity of Pd/PVPcatalyst with smaller particle size should be due to a highernumber of surface atoms and a higher amount of coordinatelyunsaturated metal sites. Interestingly, the catalyst activity wasalso found to be dependent on the oxygen flow rate. If oxygenflow rate was far away from the optimum, Pd nanoparticlesdeactivated quickly, probably through blocking of the activesurface sites by byproducts (oxygen flow rate too low) or byinteraction of the Pd surface with oxygen (oxygen flow rate toohigh). The optimal oxygen flow rate was 35 mL/min in theircatalytic system. A similar phenomenon of the effect of oxygenflow rate on the Pt/Al2O3 catalyst was also observed by Sahuand Dhepe.52 The reaction pathway of the oxidation of HMFover Pd/PVP catalyst was affected by the reaction temperature.At a low reaction temperature below 70 °C, the rate of theoxidation of HFCA into FFCA was slower than the rate of theoxidation of FFCA to FDCA, and the main product was HFCA.The rate of the oxidation of HFCA into FFCA was close to therate of the oxidation of FFCA to FDCA at the reactiontemperature of 90 °C. The stability of PVP/Pd in alkalinesolution decreased during the reaction, and PVP/Pd wasdifficult to be recycled and reused. In order to improve thestability of Pd nanoparticles and facilitate the catalyst recycle,DaSiyo and co-workers further deposited Pd/PVP on differentmetal oxides (TiO2, γ-Al2O3, KF/Al2O3, and ZrO2/La2O3) andstudied their catalytic performance toward HMF oxidation. Pd/ZrO2/La2O3 showed the highest catalytic activity with thehighest FDCA yield up to 90% and a relatively stable catalyticperformance than other supported Pd catalysts (Table 2,Entries 4−6). TEM images indicated that there was no obviousaggregation of Pd nanoparticles in the spent Pd/ZrO2/La2O3catalyst, whereas others catalysts showed serious aggregation.XPS confirmed that the majority of Pd was in the metallic stateand that the electronic structure of the Pd nanoparticles wasunchanged in the spent Pd/ZrO2/La2O3 catalyst. Nevertheless,the procedure of the recycling of the heterogeneous catalysts,the tedious recovery procedure via filtration or centrifugation,and the inevitable loss of solid catalysts during the separationprocess still limits their application. Recently, our groups have

made some improvements on the aerobic oxidation of HMFinto FDCA over several kinds of magnetic Pd catalysts.58−60

Followed by our previous work,12 our group prepared themagnetic γ-Fe2O3@HAP-supported Pd catalysts (γ-Fe2O3@HAP-Pd) for the aerobic oxidation of HMF into FDCA.58 Themagnetic core γ-Fe2O3 was coated with a layer of HAP (HAP =hydroxyapatite, Ca10(PO4)6(OH)2), and the Ca2+ in the HAPlayer can be changed with Pd2+, followed by reduction of thePd2+ to Pd(0) nanoparticles. Catalytic oxidation of HMF overγ-Fe2O3@HAP-Pd catalyst afforded 97% conversion of HMFand 92.9% yield of FDCA after 6 h at 110 °C with 0.5 equiv ofK2CO3 under 1 bar O2 (Table 2, Entry 6). The γ-Fe2O3@HAP-Pd catalyst could be facilely separated from the reactionsolution by an external magnet and reused without the loss ofits catalytic activity. TEM images indicated that particle size ofPd nanoparticles did not change in the spent catalyst. Grapheneoxide and carbon have been widely used as supports for theimmobilization of metal nanoparticles due to its high surfacearea and abundant oxygen functional groups. However, therecycling of carbon-supported catalysts is difficult due to itssmall size. Recently, our group successfully prepared magneti-cally separable graphene-oxide-supported Pd catalyst (C−Fe3O4−Pd), in which Fe3O4 nanoparticles and Pd nanoparticleswere simultaneously deposited on graphene oxide by the one-step solvothermal route.59 The C−Fe3O4−Pd catalyst showedexcellent catalytic performance in the aerobic oxidation ofHMF into FDCA, giving high HMF conversion (98.1%) andFDCA yield (91.8%) after 4 h at 80 °C with a K2CO3/HMFmolar ratio of 0.5 (Table 2, Entry 7). The C−Fe3O4−Pdcatalyst could be easily recovered by an external magnet andreused without loss of catalytic activity. Later, with the aim tosuit the sustainability chemistry, our group prepared themagnetic C@Fe3O4-supported Pd nanoparticles (Pd/C@Fe3O4) for the aerobic oxidation of HMF into FDCA undermild conditions.60 The core−shell structure C@Fe3O4 supportwas prepared by the in situ carbonization of glucose on thesurface of the Fe3O4 microspheres. HMF conversion of 100%and FDCA yield of 87.8% were obtained over Pd/C@Fe3O4catalyst after 6 h at 80 °C (Table 2, Entry 8). The Pd/C@Fe3O4 catalyst also showed good stability in the subsequentrecycling experiments, and XPS technology confirmed that themetallic Pd(0) remained in the spent catalyst. All of thecatalytic systems on oxidation of HMF into FDCA overmagnetic Pd catalysts showed the common advantages: (a)These methods did not require a large amount of base; (b)These methods could be performed under atmospheric oxygenpressure, not requiring high oxygen pressure; (c) Thesecatalysts could be easily separated by an external magnet andreused without the loss of catalytic activity.

3.4.3. Synthesis of FDCA from HMF over Supported AuCatalysts. Compared with Pt and Pd catalysts, Au catalysts

Table 2. Results of Oxidation of HMF into FDCA over Supported Pd Catalyst

entry catalyst oxygen pressure base T (°C) time (h) HMF concn (%) FDCA yield (%) ref

1 Pd/C 690 kPa O2 2 equiv of NaOH 23 6 100 71 492 Pd/PVP 1 atm O2 1.25 equiv of NaOH 90 6 >99 90 563 Pd/ZrO2/La2O3 1 atm O2 1.25 equiv of NaOH 90 6 >99 90 574 Pd/Al2O3 1 atm O2 1.25 equiv of NaOH 90 6 >99 78 575 Pd/Ti2O3 1 atm O2 1.25 equiv of NaOH 90 6 >99 53 576 γ-Fe2O3@HAP-Pd 1 atm O2 0.5 equiv of K2CO3 100 6 97 92.9 587 C−Fe3O4−Pd 1 atm O2 0.5 equiv of K2CO3 80 4 98.1 91.8 598 Pd/C@Fe3O4 1 atm O2 0.5 equiv of K2CO3 80 6 98.4 86.7 60



6535


were once considered to be inactive for chemical reactions. Thediscovery in the 1980s that Au could behave as a catalyst hasbeen one of the most stunning breakthroughs in the recentyears.61,62 It has opened up a new field of research that has ledto the discovery of very active catalysts for many potentialapplications.63,64 Recently, supported Au catalyst have shownencouraging catalytic performances for the aerobic oxidation ofHMF to FDCA in water.As discussed above, the catalytic performance of Pt and Pd

catalysts toward HMF oxidation was greatly affected bysupport. Corma and co-workers also observed that the supportshowed a great effect on the catalytic activity of Au catalysts.65

Au/CeO2 and Au/TiO2 catalysts afforded quantitative FDCAyields (>99%) after 8 h under 10 bar O2 with 4 equiv of NaOHand HMF/Au mol ratio = 150, whereas Au/C and Au/Fe2O3

catalysts produced FDCA yields in 44% and 15%, respectively.Further experiments confirmed that Au/CeO2 showed highercatalytic activity and selectivity over Au/TiO2, which was alsoobserved by Albonetti et al.66 At 130 °C and HMF/Au molratio = 640, high FDCA yield of 96% was obtained after 5 hover Au/CeO2 catalyst, whereas that was 84% after 8 h for theAu/TiO2 catalyst (Table 3, Entries 1 and 2). For each case,HMF conversion was 100%. HFCA was determined to be theonly intermediate. As shown in Figure 10, the first step (veryfast) is the oxidation of HMF into HFCA via the formation ofthe intermediate Hemiacetal 1. As no FFCA was determined inthe reaction process, the authors believed that FFCA formedfrom the oxidation of HFCA is rapidly converted into FDCAthrough a second hemiacetal intermediate (Hemiacetal 2).

Substrate degradation was strongly diminished with an increasein catalyst life by performing the reaction in two steps atdifferent reaction temperatures: first, the oxidation of HMF intoFDCA at low reaction temperature of 25 °C and, second, thesubsequent oxidation of HFCA into FDCA at 130 °C.Reductive pretreatment of the Au/CeO2 was shown toefficiently increase the catalytic activity because it increasedthe amount of Ce3+ and oxygen vacancies. The increased Ce3+

and oxygen vacancies were shown to have a great influence intransferring hydride and activating O2 during catalytic oxidationof alcohols in a former report.65 The authors proposed thatCe3+ centers (Lewis acid sites and stoichiometric oxidation sitesof CeO2) and Au+ species of Au/CeO2 could readily accept ahydride from the C−H bond in alcohol or in the correspondingalkoxide to form Ce−H and Au−H, with the formation of acarbonyl compound at the same time. The oxygen vacancies ofceria could activate O2 and form cerium-coordinated super-oxide (Ce−OO) species, which subsequently evolved to ceriumhydroperoxide by hydrogen abstraction from Au−H. Thecerium hydroperoxide then interacted with Ce−H, producingH2O and recovering the Ce3+ centers. Au−H donated H andchanged back to the initial Au+ species. In order to furtherimprove the catalytic activity of Au/CeO2, Yang and co-workersfound that the doping Bi3+ into the nano-CeO2 support couldimprove the O2 activation and hydride transfer of nano-CeO2,which was enhanced by the lone electron pair of Bi3+.67 Underthe same reaction conditions (65 °C, 1.0 MPa, HMF/Au =150) as described by Corma and co-workers,65 both Au/CeO2

and Au/Ce1−xBixO2−δ (0.08≤ x ≤ 0.5) gave 100% HMF

Table 3. Results of the Oxidation of HMF over Supported Au Catalyst

entry catalyst oxidant base T (°C) time (h) HMF concn (%) FDCA yield (%) ref

1 Au/CeO2 10 bar air 4 equiv of NaOH 130 5 100 96 652 Au/TiO2 10 bar air 4 equiv of NaOH 130 8 100 84 653 Au/Ce0.9Bi0.1O2‑δ 1 bar O2 4 equiv of NaOH 65 2 100 >99 674 Au/HY 0.3 atm O2 5 equiv of NaOH 60 6 >99 >99 685 Au/CeO2 0.3 atm O2 5 equiv of NaOH 60 6 >99 73 686 Au/TiO2 0.3 atm O2 5 equiv of NaOH 60 6 >99 85 687 Au/Mg(OH)2 0.3 atm O2 5 equiv of NaOH 60 6 >99 76 688 Au/H-MOR 0.3 atm O2 5 equiv of NaOH 60 6 96 15 689 Au/Na-ZSM-5−25 0.3 atm O2 5 equiv of NaOH 60 6 92 1 6810 Au/TiO2 20 bar O2 20 equiv of NaOH 30 18 100 71 6911 Au−Cu/TiO2 10 bar O2 4 equiv of NaOH 95 4 100 99 7212 Au8−Pd2/C 3 M Pa O2 2 equiv of NaOH 60 2 >99 >99 7413 Au/HT 1 atm O2 no base 95 7 >99 >99 7514 Au−Pd/CNT 0.5 M Pa O2 no base 100 12 100 94 7715 Au−Pd/CNT 1.0 M Pa air no base 100 12 100 96 77

Figure 10. Proposed reaction pathway for aqueous HMF aerobic oxidation over Au/CeO2 catalyst.65



6536


conversion after 1 h, but the FDCA yield largely increased from39% over Au/CeO2 catalyst to about 75% over Au/Ce1−xBixO2−δ catalysts. After 2 h, FDCA was obtained in aquantitative yield (>99%) over Au/Ce0.9Bi0.1O2−δ catalyst(Table 3, Entry 3). More importantly, the Au/Ce0.9Bi0.1O2−δcatalyst showed much higher stability than the Au/CeO2catalyst, albeit a slight decrease of FDCA yield was stillobserved in the recycling experiments. Xu and co-workers alsofound that the catalytic performance of Au catalysts was greatlyaffected by the support.68 HY-zeolite-supported Au catalyst(Au/HY) afforded quantitative FDCA yield (>99%) yield after6 h under mild conditions (60 °C, 0.3 MPa O2, 4 equiv ofNaOH), which was much higher than that of Au supported onTiO2, CeO2, and Mg(OH)2 and channel-type zeolites (ZSM-5and H-MOR) (Table 3, Entries 4−9). Detailed character-izations revealed that Au nanoclusters were well encapsulated inthe HY zeolite supercage, which was considered to restrict andavoid further growing of the Au nanoclusters into largeparticles. The acidic hydroxyl groups of the supercage wereproven to be responsible for the formation and stabilization ofthe gold nanoclusters. TEM results indicated the particle sizewas 1 nm for Au/HT catalyst, while it was 3−20 nm for othercatalysts. Moreover, the interaction between the hydroxylgroups in the supercage and the Au nanoclusters led to theelectronic modification of the Au nanoparticles, which wassupposed to contribute to the high efficiency in the catalyticoxidation of HMF to FDCA. The catalyst could be recycled,but a slight decrease of its catalytic activity was observed.Besides the support, the amount of base also showed a great

effect on the oxidation of HMF over Au catalysts. Riisager andco-worker studied the effect of NaOH amount on the catalyticperformance of Au/TiO2 catalyst toward HMF oxidation.69

With the use of 20 equiv of NaOH, 1 wt % Au/TiO2 catalystwas found to oxidize HMF into FDCA in 71% yield at 30 °Cafter 18 h with 20 bar O2 (Table 3, Entry 10). Lowconcentrations of base (i.e., corresponding to less than 5equiv) afforded relatively more of the intermediate HFCAcompared to FDCA. Without base, the conversion of HMF wasonly 13%, thus suggesting the deactivation of the Au/TiO2catalyst by the initially formed acids. In addition, precipitationof the FDCA onto the catalyst surface as a result of lowsolubility may also have hampered the reaction significantlywithout base. Similar results of the effect of base concentrationon the product selectivity were also observed by Davis, in whichthey used the Au/TiO2 catalyst for the oxidation of HMF.49

Thus, an excess amount of base is generally required to get highFDCA yield over supported Au catalysts.Au-containing bimetallic catalysts have also been used for the

oxidation of HMF into FDCA. Their chemical and physicalproperties of bimetallic catalysts may be easily tuned by varyingthe size, composition, and degree of mixing, and thus, thecatalytic activity of Au-containing bimetallic catalysts was oftenobserved to be higher than that of the monometallic Aucatalysts.70,71 Pasini and co-workers demonstrated that Au−Cu/TiO2 showed higher catalytic activity and stability over Au/TiO2 toward HMF oxidation into FDCA.72 All of the bimetallicAu−Cu/TiO2 catalysts with different Au/Cu mol ratioprepared via a colloidal route displayed an improved activity,by at least factor of 2 with respect to their correspondingmonometallic Au catalysts. The postdeposition of a PVP-stabilized gold sol onto Cu/TiO2 led to a less active catalyst ascompared with the bimetallic sample where a preformed Au−Cu sol was utilized. The postdeposited catalyst not only gave a

lower yield of FDCA but also produced a large number ofbyproducts. Thus, the authors claimed that the homogeneousAu site isolation effect caused by AuCu alloying was the mainreason for the excellent catalytic activity of the Au−Cu/TiO2catalysts prepared via a colloidal route toward HMF oxidationinto FDCA. Under the optimal reaction conditions (10 bar O2,4 equiv of NaOH, 95 °C), HMF conversion of 100%, andFDCA yield of 99% were attained after 4 h (Table 3, Entry 11).A strong synergistic effect was also evident in term of thecatalyst stability and resistance to poisoning. The Au−Cu/TiO2catalyst could be easily recovered and reused without significantleaching and agglomeration of the metal nanoparticles. Thestrong synergistic effect of Au−Cu alloy was also observed byAlbonetti et al.,73,67 in which they used Au−Cu/TiO2 and Au−Cu/CeO2 catalysts for the oxidation of HMF into FDCA,respectively. Besides the strong synergistic effect of Au−Cualloy, Villa and co-workers also observed the synergistic effectof Au−Pd alloy during the aerobic oxidation of HMF over Au−Pd/AC catalyst.74 Under the reaction conditions (3 MPa O2, 2equiv of NaOH and 60 °C), Au/AC or Pd/AC catalystsafforded full HMF conversion, but with the major product ofthe intermediate HFCA after 6 h, while the Au8−Pd2/ACcatalyst produced quantitative FDCA yield (>99%) after 2 h(Table 3, Entry 12). In addition, The Au8−Pd2/AC catalyst alsoshowed higher stability than the monometallic Au/AC catalyst.The Au/AC catalyst showed good product selectivity, but itunderwent deactivation, losing 20% of conversion efficiencyafter the fifth run. No Au leaching from the catalyst wasdetected, and thus, this deactivation was possibly ascribed toirreversible adsorption of intermediates or Au particleagglomeration. However, its stability was increased extraordi-narily by alloying Au with Pd. FDCA yield of 99% was stillobtained even after fifth run over Au8−Pd2/AC catalyst.Generally speaking, the monometallic Au catalysts easily sufferfrom the deactivation by the byproducts or reactionintermediates. The alloying of another metal (e.g., Pd andCu) with Au to form bimetallic alloy catalysts combines theadvantages of different components in the atomic level andenhances the activity and stability.The aerobic oxidation of HMF over Au catalysts is mainly

carried out in the presence of excessive base, which needsadditional acids to neutralize the base after reaction. Thus, base-free oxidation of HMF into FDCA should be an economicaland environmentally benign way. There were also some reportson the oxidation of HMF into FDCA over Au catalysts withoutbase. Ebitani and co-workers developed a base-free process forthe oxidation of HMF into FDCA using hydrotalcite-supportedAu catalyst (Au/HT).75 FDCA was achieved with an excellentyield of 99% after 7 at 95 °C under 1 bar O2 in water (Table 3,Entry 13). Au deposited on neutral supports (Al2O3, C) rarelyshowed activity, and acidic SiO2 had no activity, suggesting thatthe basicity of HT in the oxidation of HMF into FDCA wasimportant. Although MgO is also a strong base, Au/MgOshowed a much poorer catalytic activity with FDCA yield of21% than Au/HT catalyst. TEM images indicated that the Au/HT catalyst had a particle size of 3.2 nm, although the particlesize of Au/MgO was larger (>10 nm). The large size and lowdispersion of Au nanoparticles on MgO should be the reasonfor the lower catalytic activity of Au/MgO. Thus, both thebasicity of solid support and the metal active sites playedimportant roles in HMF oxidation. HFCA was the mainintermediate, which was the same with Au/TiO2 and Au/CeO2catalysts.65 Au/HT catalyst could be reused but with a slight



6537


decrease of its catalytic activity, as complete conversion ofHMF for all cycles were achieved with FDCA yields >99%,92%, and 90% for first, second, and third cycle, respectively.Although Ebitani and co-workers reported that Au/HT catalystcould be reused, Davis and co-workers observed that extensiveleaching of Mg2+ from HT occurred inevitably for the oxidationof HMF over Au/TiO2 catalyst and HT as solid base, whichwas due to the chemical interaction between the basic HT andthe formed FDCA.76 Recently, Wang and co-workersdeveloped an active and stable carbon nanotube (CNT)-supported Au−Pd catalyst for the base-free oxidation ofHMF.77 HMF conversion and FDCA selectivity over theAu−Pd/CNT catalyst could reach 100% and 94% after 12 h at373 K under 0.5 MPa O2, respectively (Table 3, Entry 14). TheAu−Pd/CNT catalyst also showed high catalytic activity usingair as the oxidant, affording HMF conversion of 100% andFDCA selectivity of 96% after 12 h at 373 K under 1.0 MPa Air(Table 3, Entry 15). The authors identified that the carbonyl/quinone and phenol (especially the former) groups on CNTsurfaces facilitated the adsorption of HMF and the intermediate(DFF) but not FDCA, which was believed to be significant forthe high catalytic activity of the Au−Pd/CNT catalyst. Inaddition to the support-enhanced adsorption effect, significantsynergistic effect also existed between Au and Pd in the alloy forthe oxidation of HMF to FDCA. The catalytic performance ofthe Au−Pd bimetallic catalysts was much better than themonometallic Au or Pd catalysts. DFF and FFCA wereidentified as the reaction intermediates (Figure 11a), suggestingthe oxidation of hydroxyl group in HMF was faster than theoxidation of aldehyde group over Au−Pd/CNT catalysts. TheAu/CNT catalyst preferentially catalyzed the oxidation of thealdehyde group in HMF, forming HFCA, which was the sameas the reaction pathway of other supported Au catalysts such asAu/CeO2 and Au/TiO2 in the presence of base.65 However,HFCA mainly underwent ring-opening and degradationreactions to byproducts under base-free conditions (as shownin Figure 11 b). For the Pd/CNT catalyst, the same reactionpathway as the Au−Pd/CNT catalyst was observed. Theaddition of Pd to the Au/CNT catalyst changed the main routefrom HFCA formation to DFF formation by accelerating theoxidation of the hydroxyl group in HMF. In addition, theincorporation of Pd further enhanced the oxidation of FFCA toFDCA, a difficult elementary step over the monometallic Aucatalyst under base-free conditions. More importantly, the Au−Pd/CNT showed a high stability. Although the selectivity ofFDCA decreased slightly in the initial three recycles, both theHMF conversion and FDCA selectivity were sustained in thefurther recycling uses.3.4.4. Synthesis of FDCA from HMF over Supported Ru

Catalysts. The noble-metal Ru is in the same group as Pt, and

it also shows catalytic activity toward HMF oxidation. As far asthe oxidation of HMF, Ru catalysts are mainly used for theoxidation of HMF into DFF in organic solvents,12,78−80 but afew examples were also reported for the oxidation of HMF intoFDCA in water. Riisager and co-workers deposited thecatalytically active Ru(OH)x species onto a series of metaloxides and studied their catalytic performance toward HMFoxidation in water without base.81,82 Under the reactionconditions (2.5 bar O2, 140 °C, 6 h), all of the catalysts werefound to have the catalytic activity toward HMF oxidation withFDCA yields ranging from 20% to 100%, but [Ru(OH)x]deposited on basic carrier materials such as MgO, MgO·La2O3,and HT gave excellent catalytic performance with FDCA yieldsabove 95%. The pH measurement of the postreaction mixtureindicated the pH of the reaction solution with basic-carrier-supported Ru catalysts was higher than those with non-basic-carrier-supported Ru catalysts, and Mg2+ was also determined inthe former reaction solution, suggesting the basic supportsacted as solid base to promote the oxidation of HMF intoFDCA. Through a series of control experiments, the authorsfound that DFF and HFCA were the reaction intermediates atlow temperature and low pressure, but HFCA was the onlyintermediate at high temperature and high oxygen pressure.These results indicated that the oxidation of the aldehyde groupis competitive to the oxidation of alcohol group at lowtemperature and low oxygen pressure, but the oxidation of thealdehyde group was much faster than the oxidation of alcoholgroup at high temperature and high oxygen pressure. Besidesthe reaction in H2O, the same groups also studied the oxidationof HMF into FDCA over supported [Ru(OH)x] catalysts inionic liquids.83 However, this method showed no significance inthe practical application. First, the catalytic results were notsatisfactory, with the maximum FDCA yield of 48%. Second,the cost of ionic liquids is very high, and it is also difficult toseparate FDCA from ionic liquids.According to the results of the aerobic oxidation of HMF

over supported metal catalysts, much progress has already beenachieved for the synthesis of FDCA in recent years. Au catalystsare more stable and selective for the aerobic oxidation of HMFinto FDCA in water than the Pt-, Ru-, and Pd-based catalysts,because the Au catalysts can offer better resistance to water andO2. However, Au catalysts also suffered from the deactivation insome cases by the deposition of the byproducts orintermediates on its active sites. In order to promote theoxidation of HMF smoothly, an excessive base is required toaccelerate the reaction and to maintain the FDCA formed inaqueous solution as the salt of the dicarboxylic acid; thus, thestrong adsorption of the carboxylic acids on the catalyst isavoided. Selecting an appropriate support and the use of alloy

Figure 11. Reaction pathways for the aerobic oxidation of HMF: (a) Pd/CNT and Au−Pd/CNT catalysts, (b) Au/CNT catalyst.77



6538


might make the catalysts show high activity and stability for thebase-free oxidation of HMF into FDCA.3.5. Mechanism of the Oxidation of HMF into FDCA

over Supported Metal Catalysts. Although many research-ers have proposed the reaction pathways of the oxidation ofHMF into FDCA over different supported metal catalysts basedon the concentration change of HMF, the detectableintermediates (such as DFF, HFCA, and FFCA) and FDCA,no single unambiguous reaction mechanism of the HMFoxidation has so far been put forward in previous work.Recently, some researchers tried to understand the reactionmechanism using isotope-labeling technology.Davis and co-workers studied the mechanism of the

oxidation of HMF into FDCA over supported Au and Ptcatalysts by the isotope labeling technology.84,85 The oxidationof HMF catalyzed by Au/TiO2 under 345 kPa O2 at 22 °C with2 equiv of NaOH gave the majority of HFCA (≥98%selectivity) after 6 h. To understand the role of O2, theoxidation of HMF over Au/TiO2 catalyst was carried out using18O2 as the oxidant, but the product HFCA showed noincorporation of 18O atoms. Thus, the oxygen of HFCA shouldcome from the solvent H2O. In order to verify it, HMFoxidation was carried out in labeled water, H2

18O. As expected,mass spectrum analysis revealed two 18O atoms wereincorporated in HFCA and the Na-adduct of HFCA. Thus,O2 was not essential for the oxidation of the aldehyde group inHMF to produce HFCA. The aldehyde side chain is believed toundergo rapid reversible hydration to a geminal diol vianucleophilic addition of a hydroxide ion to the carbonyl andsubsequent proton transfer from water to the alkoxy ionintermediate (Figure 12, step 1). This step accounts for the

incorporation of two 18O atoms in HFCA when the reaction isperformed in H2

18O. The second step is the dehydrogenationof the geminal diol intermediate, facilitated by the hydroxideions adsorbed on the metal surface, to produce the carboxylicacid (Figure 12, step 2). Production of the desired FDCArequires further oxidation of the alcohol side-chain of HFCA.

Without base, no FDCA was formed even in the presence of Auand Pt catalysts, indicating base played a crucial role in thetransformation of HFCA into FDCA. Base is believed todeprotonate the alcohol side-chain to form an alkoxyintermediate, a step that may occur primarily in the solution.86

Hydroxide ions on the catalyst surface then facilitated theactivation of the C−H bond in the alcohol side-chain to formthe aldehyde intermediate (FFCA) (Figure 12, step 3). Thenext two steps (Figure 12, steps 4 and 5) oxidize the aldehydeside-chain of FFCA to form FDCA. These two steps areexpected to proceed analogously to steps 1 and 2 for oxidationof HMF to HFCA. The reversible hydration of the aldehydegroup in step 4 to a geminal diol accounts for two more 18Oatoms incorporated in FDCA when the oxidation is performedin H2

18O. Thus, the sequence in Figure 12 explains theincorporation of all 4 18O atoms in FDCA when the reaction isperformed in H2

18O. Molecular O2 is essential for theproduction of FDCA and plays an indirect role duringoxidation by removing electrons deposited into the supportedmetal particles.The above mechanism involves the participation of base

(OH−) in the oxidation of HMF into FDCA, and thus, thereaction mechanism of the base-free reaction should bedifferent. Yan and co-workers also used the isotope labelingtechnology to study the mechanism of base-free oxidation ofHMF into FDCA over PVP- or poly[bvbim]Cl-stabilized Pdnanoparticles.55 DFF and FFCA were found to be the reactionintermediates, and HFCA was not determined during thereaction process. Other researchers also observed the samereaction pathway when the oxidation of HMF was performedwithout base or under a low pH value.77,85 As shown in Figure13, the release of 2 H+ from the hydroxyl group in HMFgenerated DFF. Then they studied the oxidation of HMF overPVP/Pt catalyst in H2

18O for 4 h. Mass spectrometric analysisof the reaction mixture revealed the oxidation products (FDCAand FFCA) incorporated 18O atoms. Peaks with m/z 163 and161 correspond to 4 and 3 18O atoms incorporated in FDCAand peaks with m/z 145 and 143 correspond to 3 and 2 18Oatoms incorporated in FFCA. The aldehyde group is alsobelieved to undergo rapid reversible hydration to a geminal diolvia nucleophilic addition of water to the carbonyl andsubsequent proton transfer metal nanoparticels, which wassimilar to Figure 12, step 1 in the above mechanism. Lastly,there is a transfer of 2 H to the surface of metal, generating thecarboxylic groups. Molecular oxygen reacts with the surfacehydride to release H2O.Although some information regarding the mechanism of

HMF oxidation of HMF was obtained by the use of isotopelabeling technology, much more effort should be devoted to geta deep understanding of the intrinsic kinetics and mechanismsby other modern technologies. The deep understanding of thereaction mechanism will provide solid insights for the rationaldesign of more efficient and stable catalysts on atomic levels forthe oxidation of HMF into FDCA.

3.6. Catalytic Synthesis of FDCA over Non-NobleMetal Heterogeneous Catalysts. Although noble-metalcatalysts generally show high catalytic performance towardthe oxidation of HMF into FDCA, the industrially large-scaleproduction of FDCA is limited by the use of noble-metalcatalysts due to the high cost of the catalysts. Recently, someinexpensive transition-metal catalysts were also found to beactive toward the oxidation of HMF into FDCA. Saha and co-workers prepared the Fe catalyst by the incorporation of Fe3+ in

Figure 12. Proposed mechanism for the oxidation of HMF in aqueoussolution in the presence of excess base (OH−) and either Pt or Au.84,85



6539


the porphyrin ring center of porphyrin-based porous organicpolymer (Fe−POP), and they studied the catalytic performanceof the Fe−POP catalyst toward aerobic oxidation of HMF inwater.87 HMF conversion of 100% and FDCA yield of 79%were attained after 10 h at 100 °C under 10 bar air. Thiscatalytic system did not require the use of base and wasbelieved to proceed via a radical chain mechanism with theformation of peroxyl radical in the catalytic cycle through EPRspectra analysis. In parallel, our group found that Merrifieldresin-supported Co(II)-meso-tetra(4-pyridyl)-porphyrin (ab-breviated as Merrifield resin-Co-Py) also showed high catalyticactivity toward the oxidation of HMF into FDCA with the useof t-BuOOH as the oxidant.88 High HMF conversion of 95.6%and FDCA yield of 90.4% were attained in acetonitrile at 100°C after 24 h. DFF was the intermediate of the oxidation ofHMF into FDCA over Merrifield resin-Co-Py catalyst. Inaddition, the catalyst could be reused without the significantloss of its catalytic activity. Later, Mugweru and co-workersperformed the oxidation of HMF into FDCA in acetic acid bythe use spinel mixed metal oxide catalyst (Li2CoMn3O8) andNaBr as the additive.89 Full HMF conversion and 80% isolatedyield of FDCA were obtained at 150 °C after 8 h under 800 psiair. The use of acetic acid and NaBr as the additive as well asthe high reaction temperature and high pressure resulted in thiscatalytic system less competitive for the large-scale productionof FDCA. Recently, in order to facilitate the recycling of thecatalysts, our group reported a new method for the oxidation ofHMF into FDCA over a magnetic nano-Fe3O4−CoOx

catalyst.90 High HMF conversion of 97.2% was obtained after12 h at 80 °C, but FDCA yield was obtained in a relative yieldof 68.6%. The catalyst could be facilely recovered by an externalmagnet and reused.Compared with noble-metal catalysts, the use of inexpensive

transition-metal catalysts shows a promising prospect in thepractical synthesis of FDCA, due to the low cost of thesenonprecious transition-metal catalysts. However, the currentlyreported methods were not selective for the production ofFDCA, and the reaction sometimes were carried out in organicsolvents such as acetic acid or the use of t-BuOOH as theoxidant, which are against the “green chemistry”. Therefore,further effort should be continuously devoted to develop newmethods based on nonprecious transition-metal catalysts that

can promote the oxidation of HMF into FDCA in water withhigh catalytic activity and selectivity with O2 as the oxidant.

4. CATALYTIC SYNTHESIS OF FDCA FROMCARBOHYDRATES

Although FDCA could be produced from HMF with nearlyquantitative yields, the cost of HMF is high. Carbohydratessuch as fructose, glucose, and cellulose are much cheaper andmore abundant than HMF. Therefore, it is more attractive tocarry out the oxidative conversion of carbohydrates into FDCAby one-pot reaction over multiple functional catalysts combingacidic and metal sites.The major problem of the direct conversion of carbohydrates

into FDCA is the risk of the simultaneous oxidation ofcarbohydrates. In 2000, Kroger and co-workers realized theone-pot conversion of fructose to FDCA using the strategy oftwo-phase system water/methyl isobutyl ketone (MIBK).91 Asshown in Figure 14, the reaction was carried out in a membrane

reactor divided with a polytetrafluorethylene membrane inorder to prevent fructose from oxidation. Fructose firstdehydrated into HMF in water with a Lewatit SPC (tradename) 108 as the solid acid catalyst. Then HMF was extractedinto MIBK due to the higher solubility of HMF in MIBK,followed by the oxidation into FDCA over metal catalysts,while fructose was insoluble in MIBK, avoiding fructose to beoxidized. This process produced a maximum 25% yield ofFDCA. Interestingly, the authors observed that HMF oxidation

Figure 13. Incorporation of 18O in the reaction steps: 18O (blue) and observed units in dashed ellipses.55

Figure 14. Scheme of the processes in membrane reactor: 1 − HMFformation in water phase, 2 − diffusion of HMF in MIBK phase, and 3− HMF oxidation.91



6540


with this catalyst in a pure MIBK phase, affording DFF as themajor product. The authors claimed that water was needed as acosubstrate for the oxidation of the aldehyde group to formFDCA as the end product. The overall reaction rate wasinfluenced by the diffusion through the membrane. In addition,levulinic acid (25% yield) was also formed as the byproduct.This method produced low FDCA yield, and it was difficult topurify FDCA from the byproducts. Later, Ribeiro andSchuchardt prepared a bifunctional catalyst by the encapsula-tion of Co(acac)3 in sol−gel silica, combing the acidic andredox ability, and studied the one-pot conversion of fructoseinto FDCA, affording fructose conversion of 72% and 99%selectivity of FDCA.92 Compared with the method reported byKroger and co-workers, both FDCA yield and selectivity wasimproved by a factor of several times. However, the reactionwas carried out at 165 °C and high pressure of 20 bar air, whichis difficult to be applied in the practical application.Recently, Zhang and co-workers reported a two-step method

for the conversion of fructose into FDCA.93 The dehydration offructose into HMF was performed in isopropanol catalyzed byHCl. After reaction, isopropanol was collected for the next runby evaporation. Then, HMF was purified with water-extraction,followed by the oxidation over Au/HT catalyst. The overallFDCA yield of 83% was achieved from fructose. The water-extraction step is very important for the whole step. Under thesame conditions, the oxidation of HMF without extraction onlygave 51% yield of FDCA even after a long time of 20 h,although that was 98% after 7 h by water extraction. Theauthors claimed that the impurities such as humins made theAu/HT catalyst deactivation. This method also afforded anoverall FDCA yield of 52% using Jerusalem artichoke tuber(major component of fructose unit) as the feedstock. Later, asimilar work was also reported by the same group, in whichpolybenzylic ammonium chloride resins was used as a solidcatalyst for the dehydration of fructose in the first step,affording FDCA with a yield of 72% from fructose.94 Morerecently, Zhang and co-workers demonstrated a triphasicreactor that can convert sugars into FDCA in a one-potprocess.95 The triphasic system is composed of tetraethylam-monium bromide (TEAB) or water (phase I)−methyl isobutylketone (MIBK) (phase II)−water (phase III). In the designedtriphasic setup, sugars (fructose or glucose) were firstdehydrated to HMF in Phase I. HMF was then extracted,purified, and transferred to Phase III via a bridge (Phase II).Finally, HMF was oxidized to FDCA over Au/HT catalyst inPhase III. Overall, FDCA yields of 78% and 50% were achievedwith fructose and glucose feedstock, respectively. Phase II playsmultiple roles: as a bridge for HMF extraction, transportation,and purification. In order to facilitate the recycling of thecatalysts and reduce the cost of the catalyst, our group recentlydeveloped the one-pot conversion of fructose into FDCA viatwo-step method by the combination of two magneticcatalysts.90 Nano-Fe3O4−CoOx catalyst showed high catalyticactivity toward the oxidation of HMF into FDCA using t-BuOOH as the oxidant. Then, a two-step strategy was appliedfor the synthesis of FDCA from fructose. HMF was first

produced from the dehydration of fructose over the Fe3O4@SiO2−SO3H acid catalyst in DMSO. Then, the magneticFe3O4@SiO2−SO3H was easily separated from the reactionsystem by an external magnet, and HMF in the remainingreaction solution was then oxidized into FDCA with t-BuOOHover nano-Fe3O4−CoOx catalyst. FDCA was obtained in a yieldof 59.8% after 15 h, based on the starting fructose. Ourdeveloped method shows the following two distinct advantages:(1) The use of magnetic catalyst facilitates the catalyst recycle;(2) The use of transition-metal catalyst makes this methodmuch more economical for the practical synthesis of FDCAfrom renewable carbohydrates.The direct conversion of carbohydrates into FDCA is much

more attractive, but the current results are not satisfactory. Thecarbohydrate for FDCA production is focused on fructose.Much more work should be paid to the design of novel catalystswith multiple catalytic sites for the conversion of othercarbohydrates such as glucose or even cellulose into FDCA.In order to avoid the side reactions, the isolation of multiplecatalytic sites is essential to realize the one-pot conversion ofcarbohydrates to FDCA. The acidic sites located in ahydrophilic environment would benefit the adsorption of thecarbohydrates and promote the dehydration of carbohydratesinto HMF and the release of HMF into the reaction solution.The oxidative sites located in a hydrophobic environmentenhance the absorption of the intermediate HMF and promoteits oxidation to FDCA, as well as simultaneously release thehigh polar product FDCA in the reaction solution.

5. CATALYTIC SYNTHESIS OF FDCA DERIVATIVES

The catalytic synthesis of FDCA from HMF can be obtainedwith high yields, but there are few reports on the purification ofFDCA. As the oxidation of HMF into FDCA in most cases iscarried out in the presence of excessive base, FDCA is in theform of its salts together with byproducts and the salts, makingit difficult to purify FDCA. Hence, regarding FDCApurification, there is still a dearth of straightforward and eco-friendly methods. A possible way to overcome this incon-venience is to produce the corresponding ester, 2,5-furandicarboxylicacid dimethyl ester (FDCDM), which can beeasily purified through vacuum distillation and transformed toFDCA through a simple hydrolysis reaction. Moreover, insteadof FDCA, FDCDM can also be used directly to synthesispolymers through transesterification reaction.In 2008, Christensen and co-workers reported an effective

way for the one-pot oxidative esterification of HMF intoFDCDM over Au/TiO2 catalyst. FDCDM could be achieved inan excellent yield of 98% at 130 °C under 4 bar O2 in MeOH inthe presence of 8% MeONa.96 The addition of MeONaaccelerated this reaction remarkably, because the reaction wasincomplete without MeONa even when run for a longerduration. Kinetic study revealed 5-hydroxymethyl methylfur-oate (HMMF) was the reaction intermediate by the fastoxidative esterification of HMF (Figure 15), which was thenslowly converted to FDCDM, suggesting that the oxidation ofthe aldehyde moiety of HMF was faster than the

Figure 15. Proposed oxidation pathway of HMF to FDCDM.96



6541


hydroxymethyl group, which was similar as the oxidation ofHMF into FDCA in water over Au catalysts.65 In contrast toFDCA, FDCDM can be easily purified by sublimation atstandard pressure and 160 °C to afford colorless crystals. Later,Corma and co-workers improved the oxidative esterification ofHMF into FDCDM over Au/CeO2 catalyst.

97 The Au/CeO2catalyst showed high catalytic activity without base, affordinghigh FDCDM yield of 99% after 5 h at 130 °C under 10 bar O2,although 8% MeONa was required to promote the reactionsmoothly over Au/TiO2 catalyst. Therefore, Au/CeO2 showedhigher catalytic activity than Au/TiO2 catalyst, which was dueto the formation of peroxo and superoxo species inmonoelectronic defects of nanocrystalline ceria. Kinetic studiesindicated that the oxidation of alcohol group was the rate-limiting step of the conversion of HMF into FDCDM, whichwas similar to the results reported by Christensen and co-workers using Au/TiO2 catalyst.96 Au/CeO2 could be easilyrecovered and reused with a little loss of activity butmaintaining high selectivity toward FDCDM. Although highFDMCA yields were attained over Au/CeO2 and Au/TiO2catalysts, the use of Au catalysts is high cost. Recently, Fu andco-workers developed an economical way for the oxidativeesterification of HMF into FDCDM using cheap cobalt-basedcatalysts (CoxOy−N@C).98 However, the CoxOy−N@Ccatalyst is not as active as the Au catalysts. Under 1 MPa O2at 100 °C for 12 h, 95% conversion of HMF was obtained, butthe FDCDM yield was only 38%, with other products mainlyincluding 2,5-furandicarboxylic acid monomethyl ester(FDCMM) with a yield of 44% and 5-hydroxymethyl-2-furoicacid methyl ester (HMFM) with a yield of 12%. In order toimprove the oxidation of alcohol group to aldehyde group, thestrong oxidant K-OMS-2 is used to promote this functionalgroup change. Although HMFM yield decreased from 12% to1%, FDCDM was still coproduced with FDCMM, with theyields of 53% and 41%, respectively.Taking the above-described methods into consideration, the

oxidative esterification of HMF into FDMC shows severaladvantages over the oxidation of HMF into FDCA, such as thefacile product purification, and the use of low catalytic amountof base or base-free. It should be a direction to effective andeconomical transformation of carbohydrates derived platformmolecule HMF to 2,5-furandimethylcarboxylate as a buildingblock for the production of biobased polymer. These describedmethods using noble-metal Au catalysts were effective, but thecheap Co catalyst was much less active. Therefore, developingnon-noble-metal catalysts with high catalytic activity andselectively should be paid much more effort.

6. CONCLUSION AND PERSPECTIVE6.1. Conclusions. FDCA is a promising biomass-derived

chemical with wide application in many fields. Moreimportantly, it can serve as a replacement of petrochemical-derived terephthalic acid for the production of useful biobasedpolymer. Thus, catalytic synthesis of FDCA in an effective waywill have huge economic benefits in the future and will also playa vital role in the development of sustainable chemistry.Currently, there are several methods of the catalytic synthesis ofFDCA from HMF or directly from carbohydrates by one-potreaction, including biocatalytic, electrochemical, and chemicalcatalytic approaches.Currently, the chemical catalytic method is the main driving

force for the synthesis of FDCA as it shows high potential inthe large-scale synthesis of FDCA by the use of molecular O2 in

an economical way. Much effort has been devoted to the designof chemical catalytic systems for the aerobic oxidation of HMFinto FDCA, mainly using supported noble-metal catalysts (Au,Pt, and Pd). The activity of the catalyst and the reactionpathway is affected by the catalyst itself (such as the activephase, support, particle size) and reaction conditions (such asoxygen pressure, oxygen flow rate, pH, and temperature).Compared with Pt and Pd catalysts, Au catalysts are morestable and selective, because the Au catalysts can offer betterresistance to O2. However, Au catalysts are also found to bedeactivated by the byproducts or intermediates in some cases.The alloying of another metal (e.g., Pd, Cu, Pt) with Au to formbimetallic alloy catalysts generally show high catalytic activityand stability as compared with the monometallic Au catalysts,due to the alloying effect. HFCA is generally detected as thereaction intermediate in alkaline reaction solution especially forAu catalyst, whereas DFF is the reaction intermediate of base-free aerobic oxidation of HMF, suggesting the oxidation ofaldehyde group in HMF is faster than the oxidation of alcoholgroup at low pH. Despite the designated reaction pathway, theisotope labeling technology indicates that the reactionmechanisms are similar. The oxygen in FDCA is from water,and molecular oxygen plays a key role in the catalytic recycle byremoving electrons deposited into the supported metalparticles. A few examples were also reported for the one-potconversion of carbohydrates into FDCA. However, thefeedstock is mainly focused on fructose, and the two-stepmethod is adopted to avoid the oxidation of carbohydrates.Finally, the diester of FDCA can also be synthesized theoxidative esterification of HMF in one-pot reaction oversupported Au catalysts.

6.2. Perspectives. Although significant achievements havebeen obtained on the catalytic synthesis of FDCA and itsderivatives using various methods, further improvements arestill needed in chemical aerobic oxidation of HMF for realizingthe industrially large-scale and economical production of FDCAand its derivatives. (1) One challenge for the aerobic oxidationof HMF over supported noble-metal catalysts is the instabilityof the catalysts. The stability of the catalyst can be improvedfrom several aspects such as enhancing the interaction of metalnanoparticles and the supports, confining the metal nano-particles in pores with a limited size to avoid their aggregation,improving the catalytic activity with full conversion and 100%selectivity to avoid the catalyst deactivation by the adsorption ofbyproducts, using bimetallic or trimetallic nanoparticles toimprove the ability to prevent their poisoning. (2) Most of thecurrent methods are performed in water in the presence ofexcess base, resulting in the less-green and more expensiveprocess. More importantly, the neutralization of the salt ofFDCA would further increase the operating cost and produceadditional salt byproducts, making the process further less-green and less cost-effective. Therefore, much more effort isneeded to develop more environmentally friendly catalyticsystems that can effectively promote the base-free oxidation ofHMF into FDCA. (3) Compared to noble-metal catalysts, non-noble-metal catalysts are much less studied. From the viewpointof practical application, the development of efficient, low-cost,and stable transition-metal catalysts such as Co, Fe, and Mntoward the aerobic oxidation of HMF into FDCA is thereforevery important. (4) The use of carbohydrates or lignocellulosesas the feedstocks for the one-pot production of FDCA isstrongly desired, as they are cheap and abundant, which willdecrease the production cost. In order to realize this goal,



6542


multifunctional catalysts, combining the metal sites and solidacid/base catalysts, should be carefully designed. It is suggestedthat the acid/base sites used for the dehydration ofcarbohydrates into HMF should be located in hydrophilicenvironments and the metal sites for oxidation reaction shouldbe located in hydrophobic environments, and thus, thecarbohydrates will easily be absorbed on the acid/base sitesand the intermediate HMF will easily desorbe to thehydrophobic metal size to be oxidized. (5) Very few studieshave been performed to examine the kinetics and reactionmechanisms of catalytic synthesis of FDCA either from HMFor direct from carbohydrates. Deep understanding of theintrinsic kinetics and mechanisms will provide insights for therational design of more efficient and stable catalysts on atomiclevels for the oxidation of HMF into FDCA or one-potoxidative conversion of carbohydrates into FDCA. (6) Last butnot least, catalytic synthesis of FDCA in a large scale isextremely important. Many current routes are technicallyfeasible but economically prohibitive. Development of theenergetically and economically viable processes is a long-standing task for us, which involves interdisciplinary problemsof chemistry, material science, process engineering, and so on.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: [email protected]. Tel.: +86-27-67842572.Fax: +86-27-67842572.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by the National Natural ScienceFoundation of China (No. 21203252) and the funding offeredby the China scholarship council (201408420018).

■ REFERENCES(1) Corma, A.; Iborra, S.; Velty, A. Chem. Rev. 2007, 107, 2411−2502.(2) Bozell, J. J. Science 2010, 329, 522−523.(3) Gallezot, P. Chem. Soc. Rev. 2012, 41, 1538−1558.(4) Morais, A. R. C.; da Costa Lopes, A. M.; Bogel-Lukasik, R. Chem.Rev. 2015, 115, 3−27.(5) Liguori, F.; Moreno-Marrodan, C.; Barbaro, P. ACS Catal. 2015,5, 1882−1894.(6) Katryniok, B.; Paul, S.; Dumeignil, F. ACS Catal. 2013, 3, 1819−1834.(7) Beerthuis, R.; Rothenberg, G.; Shiju, N. R. Green Chem. 2015, 17,1341−1361.(8) Amaniampong, P. N.; Jia, X. L.; Wang, B.; Mushrif, S. H.; Borgna,A.; Yang, Y. H. Catal. Sci. Technol. 2015, 5, 2393−2405.(9) Zhang, J. H.; Li, J. K.; Tang, Y. J.; Lin, L.; Long, M. N. Carbohydr.Polym. 2015, 130, 420−428.(10) Hanson, S. K.; Baker, R. T. Acc. Chem. Res. 2015, 48, 2037−2048.(11) Ma, R. S.; Xu, Y.; Zhang, X. ChemSusChem 2015, 8, 24−51.(12) Zakrzewska, M. E.; Bogel-Lukasik, E.; Bogel-Lukasik, R. Chem.Rev. 2011, 111, 397−417.(13) Zhang, Z. H.; Liu, B.; Lv, K. L.; Sun, J.; Deng, K. J. Green Chem.2014, 16, 2762−2770.(14) Lan, J. H.; Lin, J. C.; Chen, Z. Q.; Yin, G. C. ACS Catal. 2015, 5,2035−2041.(15) Werpy, T.; Petersen, G.; Aden, A.; Bozell, J.; Holladay, J.; White,J.; Manheim, A.; Eliot, D.; Lasure, L.; Jones, S.; Top Value AddedChemicals from Biomass. Vol. 1· Results of Screening for PotentialCandidates from Sugars and Synthesis Gas; U. S. Department of Energy

Report. Pacific Northwest National Laboratory/U.S. Department ofEnergy: Oak Ridge, TN, 2004.(16) Haworth, W. N.; Jones, W. G. M.; Wiggins, L. F. J. Chem. Soc.1945, 1−4.(17) Jacquel, N.; Saint-Loup, R.; Pascault, J. P.; Rousseau, A.;Fenouillot, F. Polymer 2015, 59, 234−242.(18) Rajendran, S.; Raghunathan, R.; Hevus, I.; Krishnan, R.;Ugrinov, A.; Sibi, M. P.; Webster, D. C.; Sivaguru, J. Angew. Chem., Int.Ed. 2015, 54, 1159−1163.(19) Wilsens, C. H. R. M.; Wullems, N. J. M.; Gubbels, E.; Yao, Y. F.;Rastogi, S.; Noordover, B. A. Polym. Chem. 2015, 6, 2707−2716.(20) Eerhart, A. J. J. E.; Faaij, A. P. C.; Patel, M. K. Energy Environ.Sci. 2012, 5, 6407−6422.(21) Gandini, A.; Silvestre, A. J. D.; Neto, C. P.; Sousa, A. F.; Gomes,M. J. J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 295−298.(22) Gandini, A. Polym. Chem. 2010, 1, 245−251.(23) Davis, M. E. Top. Catal. 2015, 58, 405−409.(24) Lewkowski, J.; Vaultier, M. Main Group Met. Chem. 2001, 1,17−54.(25) Nagarkar, S. S.; Chaudhari, A. K.; Ghosh, S. K. Cryst. GrowthDes. 2012, 12, 572−576.(26) Rose, M.; Weber, D.; Lotsch, B. V.; Kremer, R. K.; Goddard, R.;Palkovits, R. Microporous Mesoporous Mater. 2013, 181, 217−221.(27) Fittig, R.; Heinzelmann, H. Chem. Ber. 1876, 9, 1198.(28) Gonis, G.; Amstutz, E. D. J. Org. Chem. 1962, 27, 2946−2947.(29) Miura, T.; Kakinuma, H.; Kawano T.; Matsuhisa, H. U.S PatentUS-7411078, 2008.(30) Holade, Y.; Morais, C.; Servat, K.; Napporn, T. W.; Kokoh, K. B.ACS Catal. 2013, 3, 2403−2411.(31) Ciriminna, R.; Palmisano, G.; Pagliaro, M. ChemCatChem 2015,7, 552−558.(32) Grabowski, G.; Lewkowski, J.; Skowron ski, R. Electrochim. Acta1991, 36, 1995−1995.(33) Vuyyuru, K. R.; Strasser, P. Catal. Today 2012, 195, 144−154.(34) Chadderdon, D. J.; Xin, L.; Qi, J.; Qiu, Y.; Krishna, P.; More, K.L.; Li, W.Z. Green Chem. 2014, 16, 3778−3786.(35) Cha, H. G.; Choi, K. S. Nat. Chem. 2015, 7, 328−333.(36) Thomas, S. M.; DiCosimo, R.; Nagarajan, A. Trends Biotechnol.2002, 20, 238−242.(37) Krystof, M.; Perez-Sanchez, M.; de Maria, P. D. ChemSusChem2013, 6, 826−830.(38) Hanke, P. D. U.S. Patent 2009/023174 A2, 2009.(39) Koopman, F.; Wierckx, N.; deWinde, J. H.; Ruijssenaars, H. J.Bioresour. Technol. 2010, 101, 6291−6296.(40) Dijkman, W. P.; Groothuis, D. E.; Fraaije, M. W. Angew. Chem.,Int. Ed. 2014, 53, 6515−6518.(41) Dijkman, W. P.; Binda, C.; Fraaije, M. W.; Mattevi, A. ACSCatal. 2015, 5, 1833−1839.(42) Partenheimer, W.; Grushin, V. V. Adv. Synth. Catal. 2001, 343,102−111.(43) Saha, B.; Dutta, S.; Abu-Omar, M. M. Catal. Sci. Technol. 2012,2, 79−81.(44) Hansen, T. S.; Sadaba, I.; Garcia-Suarez, E. J.; Riisager, A. Appl.Catal., A 2014, 456, 44−50.(45) Verdeguer, P.; Merat, N.; Gaset, A. J. Mol. Catal. 1993, 85, 327−344.(46) Ait Rass, H. A.; Essayem, N.; Besson, M. Green Chem. 2013, 15,2240−2251.(47) Ait Rass, H. A.; Essayem, N.; Besson, M. ChemSusChem 2015, 8,1206−1217.(48) Niu, W. Q.; Wang, D.; Yang, G. H.; Sun, J.; Wu, M. B.;Yoneyama, Y.; Tsubaki, N. Bull. Chem. Soc. Jpn. 2014, 87, 1124−1129.(49) Davis, S. E.; Houk, L. R.; Tamargo, E. C.; Datye, A. K.; Davis, R.J. Catal. Today 2011, 160, 55−60.(50) Vinke, P.; van Dam, H. E.; van Bekkum, H.Studies in SurfaceScience and Catalysis; Elsevier: Amsterdam, 1990; Vol. 55, pp 147−15710.1016/S0167-2991(08)60144-5(51) Vinke, P.; Poel, W. V.; BikkumH. V. Studies in Surface Scienceand Catalysis; Elsevier: Amsterdam, 1991; Vol. 59, pp 385−394.



6543

mailto:[email protected]


(52) Sahu, R.; Dhepe, P. L. React. Kinet., Mech. Catal. 2014, 112,173−187.(53) Miao, Z. Z.; Wu, T. X.; Li, J. W.; Yi, T.; Zhang, Y. B.; Yang, X.G. RSC Adv. 2015, 5, 19823−19829.(54) Lilga, M. A.; Hallen, R. T.; Gray, M. Top. Catal. 2010, 53,1264−1269.(55) Siankevich, S.; Savoglidis, G.; Fei, Z.; Laurenczy, G.; Alexander,D. T. L.; Yan, N.; Dyson, P. J. J. Catal. 2014, 315, 67−74.(56) Siyo, B.; Schneider, M.; Pohl, M. M.; Langer, P.; Steinfeldt, N.Catal. Lett. 2014, 144, 498−506.(57) Siyo, B.; Schneider, M.; Radnik, J.; Pohl, M. M.; Langer, P.;Steinfeldt, N. Appl. Catal., A 2014, 478, 107−116.(58) Zhang, Z. H.; Zhen, J. D.; Liu, B.; Lv, K. L.; Deng, K. J. GreenChem. 2015, 17, 1308−1317.(59) Liu, B.; Ren, Y. S.; Zhang, Z. H. Green Chem. 2015, 17, 1610−1617.(60) Mei, N.; Liu, B.; Zheng, J. D.; Lv, K. L.; Tang, D. G.; Zhang, Z.H. Catal. Sci. Technol. 2015, 5, 3194−3202.(61) Hutchings, G. J. J. Catal. 1985, 96, 292−295.(62) Haruta, M.; Kobayashi, T.; Sano, H.; Yamada, N. Chem. Lett.1987, 16, 405−408.(63) Pan, M.; Brush, A. J.; Pozun, Z. D.; Ham, H. C.; Yu, W. Y.;Henkelman, G.; Hwang, G. S.; Mullins, C. B. Chem. Soc. Rev. 2013, 42,5002−5013.(64) Wittstock, A.; Wichmann, A.; Baumert, M. ACS Catal. 2012, 2,2199−2215.(65) Casanova, O.; Iborra, S.; Corma, A. ChemSusChem 2009, 2,1138−1144.(66) Albonetti, S.; Lolli, A.; Morandi, V.; Migliori, A.; Lucarelli, C.;Cavani, F. Appl. Catal., B 2015, 163, 520−530.(67) Miao, Z. Z.; Zhang, Y. B.; Pan, X. Q.; Wu, T. X.; Zhang, B.; Li, J.W.; Yi, T.; Zhang, Z. D.; Yang, X. G. Catal. Sci. Technol. 2015, 5,1314−1322.(68) Cai, J. Y.; Ma, H.; Zhang, J. J.; Song, Q.; Du, Z. T.; Huang, Y. Z.;Xu, J. Chem. - Eur. J. 2013, 19, 14215−14223.(69) Gorbanev, Y. Y.; Klitgaard, S. K.; Woodley, J. M.; Christensen,C. H.; Riisager, A. ChemSusChem 2009, 2, 672−675.(70) Wu, Y. E.; Wang, D. S.; Zhou, G.; Yu, R.; Chen, C.; Li, Y. D. J.Am. Chem. Soc. 2014, 136, 11594−11597.(71) Rasul, S.; Anjum, D. H.; Jedidi, A.; Minenkov, Y.; Cavallo, L.;Takanabe, K. Angew. Chem., Int. Ed. 2015, 54, 2146−2150.(72) Pasini, T.; Piccinini, M.; Blosi, M.; Bonelli, R.; Albonetti, S.;Dimitratos, N.; Lopez-Sanchez, J. A.; Sankar, M.; He, Q.; Kiely, C. J.;Hutchings, G. J.; Cavani, F. Green Chem. 2011, 13, 2091−2099.(73) Albonetti, S.; Pasini, T.; Lolli, A.; Blosi, M.; Piccinini, M.;Dimitratos, N.; Lopez-Sanchez, J. A.; Morgan, D. J.; Carley, A. F.;Hutchings, G. J.; Cavani, F. Catal. Today 2012, 195, 120−126.(74) Villa, A.; Schiavoni, M.; Campisi, S.; Veith, G. M.; Prati, L.ChemSusChem 2013, 6, 609−612.(75) Gupta, N. K.; Nishimura, S.; Takagaki, A.; Ebitani, K. GreenChem. 2011, 13, 824−827.(76) Zope, B.; Davis, S.; Davis, R. Top. Catal. 2012, 55, 24−32.(77) Wan, X. Y.; Zhou, C. M.; Chen, J. S.; Deng, W. P.; Zhang, Q.H.; Yang, Y. H.; Wang, Y. ACS Catal. 2014, 4, 2175−2185.(78) Artz, J.; Mallmann, S.; Palkovits, R. ChemSusChem 2015, 8,672−679.(79) Nie, J. F.; Xie, J. H.; Liu, H. C. J. Catal. 2013, 301, 83−91.(80) Wang, S. G.; Zhang, Z. H.; Liu, B.; Li, J. L. Ind. Eng. Chem. Res.2014, 53, 5820−5827.(81) Gorbanev, Y. Y.; Kegnaes, S.; Riisager, A. Catal. Lett. 2011, 141,1752−1760.(82) Gorbanev, Y. Y.; Kegnaes, S.; Riisager, A. Top. Catal. 2011, 54,1318−1324.(83) Stahlberg, T.; Eyjolfsdottir, E.; Gorbanev, Y. Y.; Sadaba, I.;Riisager, A. Catal. Lett. 2011, 142, 1089−1097.(84) Davis, S. E.; Zope, B. N.; Davis, R. J. Green Chem. 2012, 14,143−147.

(85) Davis, S. E.; Benavidez, A. D.; Gosselink, R. W.; Bitter, J. H.; deJong, K. P.; Datye, A. K.; Davis, R. J. J. Mol. Catal. A: Chem. 2014, 388-389, 123−132.(86) Zope, B. N.; Hibbitts, D. D.; Neurock, M.; Davis, R. J. Science2010, 330, 74−87.(87) Saha, B.; Gupta, D.; Abu-Omar, M. M.; Modak, A.; Bhaumik, A.J. Catal. 2013, 299, 316−320.(88) Gao, L. C.; Deng, K. J.; Zheng, J. D.; Liu, B.; Zhang, Z. H. Chem.Eng. J. 2015, 270, 444−449.(89) Jain, A.; Jonnalagadda, S. C.; Ramanujachary, K. V.; Mugweru,A. Catal. Commun. 2015, 58, 179−182.(90) Wang, S. G.; Zhang, Z. H.; Liu, B. ACS Sustainable Chem. Eng.2015, 3, 406−412.(91) Kroger, M.; Prusse, U.; Vorlop, K. D. Top. Catal. 2000, 13,237−242.(92) Ribeiro, M. L.; Schuchardt, U. Catal. Commun. 2003, 4, 83−86.(93) Yi, G. S.; Teong, S. P.; Li, X. K.; Zhang, Y. G. ChemSusChem2014, 7, 2131−2135.(94) Teong, S. P.; Yi, G. S.; Cao, X. Q.; Zhang, Y. G. ChemSusChem2014, 7, 2120−2124.(95) Yi, G. S.; Teong, S. P.; Zhang, Y. G. ChemSusChem 2015, 8,1151−1155.(96) Taarning, E.; Nielsen, I. S.; Egeblad, K.; Madsen, R.;Christensen, C. H. ChemSusChem 2008, 1, 75−78.(97) Casanova, O.; Iborra, S.; Corma, A. J. Catal. 2009, 265, 109−116.(98) Deng, J.; Song, H. J.; Cui, M. S.; Du, Y. P.; Fu, Y. ChemSusChem2014, 7, 3334−3340.



6544


recent advances in the catalytic synthesis ... - nsfc.gov.cn

Documents