*2.1. Au-Based Catalysts*

Gold (Au) was generally considered as inactive in the field of catalysis, but in the 1980s, it was a grea<sup>t</sup> achievement in research to discover the catalytic ability of Au, which opened a new gateway for researchers to develop highly active catalysts for many processes [34,35]. Recently, many catalysts are developed using Au that present promising catalytic performance for the aerobic oxidation of HMF to FDCA in aqueous solutions. The performance of di fferent Au-based catalysts for catalytic aerobic oxidation of HMF to FDCA is summarized, and details are also given, in Table 1.


**Table 1.** Summary of reported results for the aerobic oxidation of HMF to FDCA over Au catalysts.

a nNiO = nanosized NiO b La-CaMgAl-LDH = La doped Ca-Mg-Al layered double hydroxide.

The choice of support for Au-based catalysts can also have a grea<sup>t</sup> impact on the catalytic performance in HMF oxidation. When using TiO2 and CeO2 as supports, Au-based catalysts showed nearly quantitative FDCA yields of >99% at 65 ◦C under 10 bar of air after a reaction time of 8 h; in contrast, Au catalysts supported on carbon and Fe2O3 only a fforded FDCA yields of 44% and 15% under the same conditions, separately [36]. According to the reaction mechanism discussed in this study, HMFCA was observed as the only intermediate. As shown in Scheme 3, HMF was first oxidized to HMFCA very fast via the formation of a hemiacetal-1 intermediate. Owing to the fact no FFCA was directly observed, the authors proposed that FFCA was transformed via the oxidation of HMFCA was quickly transformed into FDCA through the production of a second intermediate product, hemiacetal-2. Compared to the one-pot reaction, substrate degradation was strongly diminished and the catalysts life increased by performing the reaction in two steps: first the oxidation of HMF into HMFCA at a low reaction temperature of 25 ◦C and, second, the subsequent oxidation of HMFCA in FDCA at 130 ◦C. Reductive pretreatment of the Au/CeO2 was shown to e fficiently increase the catalytic activity due to increased amount of Ce3+ and oxygen vacancies. The increased Ce3+ species and oxygen vacancies on the support were shown to have a grea<sup>t</sup> e ffect on transferring hydride and activating O2 during the oxidation of the alcohol group. The Lewis acid sites of Ce3+ centers and Au<sup>+</sup> species of Au/CeO2 could easily accept a hydride from the C–H bond in alcohol or in the corresponding alkoxide to form Ce–H and Au–H, with the simultaneous formation of a carbonyl species. The oxygen vacancies of CeO2 could activate O2 and form cerium-coordinated superoxide (Ce–OO) species, which subsequently evolved into cerium hydroperoxide by hydrogen abstraction from Au-H. The cerium hydroperoxide then interacted with Ce–H, producing H2O and recovering the Ce3+ centers. Au–H donated H and changed back to the initial Au<sup>+</sup> species. Further improvement of activity of Au/CeO2 was reported by Lolli et al [37]. An ordered mesoporous CeO2 (m-CeO2) supported Au catalyst was synthesized by nanocasting technique using meso-structured silica SBA-15 as hard template. Au nano-particles immobilized on this high surface area mesoporous CeO2 showed a FDCA yield of 92% with 100% HMF conversion under relatively mild reaction conditions (T = 70 ◦C, PO2 = 10 bar, and t = 4 h).

**Scheme 3.** Reaction mechanisms for aerial oxidation of aqueous HMF over CeO2 supported Au catalysts (reproduced from [36] with permission from Wiley-VCH, copyright 2009).

Formation of undesired humin is a grea<sup>t</sup> issue for HMF transformation, especially in concentrated HMF solutions. Kim et al. reported recently progress in utilizing Au/CeO2 catalyst for achieving 90–95% yield of FDCA via aerobic oxidation of acetal derivatives of HMF [38]. In this approach, protection of aldehyde group of HMF with 1,3-propanediol was proposed to prevent the formation of undesired humin via decomposition and self-polymerization, and to achieve e fficient FDCA yield from the resultant HMF acetal derivative. Even in concentrated solutions of 20% PD-HMF, FDCA could still be obtained in a high yield of 91% at 140 ◦C and 5 bar O2, for 15 h reaction. This example presents a

significant advance over the conventional oxidation of HMF that gives only reasonable FDCA yields in dilute solutions.

Zeolite-supported Au catalysts have also been investigated for the catalytic oxidation of HMF by Xu et al. [40]. The Au/H-Y catalyst showed high yield of FDCA (>99%) with a quantitative HMF conversion under mild reaction conditions (T = 60 ◦C, PO2 = 0.3 bar and t = 6 h), which was much higher than that of Au supported on Mg(OH)2, TiO2, CeO2, H-MOR, and ZSM-5. Further characterization indicated that Au-nanoclusters (approx. 1 nm) are encapsulated inside the supercages of the H-Y-zeolite, and the confinement of the supercage prevented the further agglomeration of Au nanoclusters into large particles (Figure 2). The interaction between the acidic hydroxyl groups in the zeolite supercage and Au clusters has been shown to be responsible for stabilization of the Au species, to which the high catalytic efficiency for the oxidation of HMF to FDCA was ascribed [40].

**Figure 2.** (**i**)-**A** Schematic illustration of synthesis for Au-nanoclusters in HY zeolite supercages (**i**)-**B** Catalytic oxidation process (**ii**)-**a** TEM image of Au/HY (**ii**)-**b** HR-TEM image of Au/HY (reproduced from [40] with permission from Wiley-VCH, copyright 2013).

Various Au-based bimetallic catalysts were also studied for the oxidation of HMF to FDCA. The physical and chemical characteristics of the prepared bimetallic catalysts can be simply tuned by altering catalytic composition, particle size and mixing equality. Pasini et al. reported that the Au-Cu/TiO2 bimetallic catalyst afforded higher catalytic activity and stability over its corresponding mono-metallic Au/TiO2 for HMF oxidation into FDCA [41]. All the bimetallic Au-Cu/TiO2 catalysts with different Au/Cu mole ratio prepared via a colloidal route showed an improved activity, by at least factor of two compared to their corresponding monometallic Au catalysts. Under the optimal reaction conditions (10 bar O2, 4 equiv of NaOH, 95 ◦C), HMF conversion of 100%, and FDCA yield of 99% were attained after 4 h (Table 1). The isolation of Au sites caused by AuCu alloying was the main reason for the excellent catalytic activity of the Au-Cu/TiO2 catalysts for HMF oxidation into FDCA. The Au-Cu/TiO2 catalyst could be easily recovered and reused without significant leaching and agglomeration of the metal nanoparticles. Thus, a strong synergistic effect was also evident in term of the catalyst stability and resistance to poisoning. Similar results in catalytic performance were also demonstrated for the Au-Pd/AC catalyst [42]. Alloying Au with Pd at molar ratio of 8:2 on carbon support significantly increased the catalyst stability and activity than the monometallic counterparts for the aerobic oxidation of HMF to FDCA. The Au/AC catalyst showed good product selectivity, but suffered with catalyst deactivation, with a drop in 20% of HMF conversion after the fifth run. No Au leaching from the catalyst was detected, and the deactivation was mainly attributed to irreversible adsorption of the byproducts or intermediates and the agglomeration of Au particles. In contrast, Au-Pd/AC delivered an excellent stability with a FDCA yield of 99% even after the fifth run. The alloying of a second metal (e.g., Pd or Cu) with Au to form bimetallic alloy catalysts can indeed combine the advantages of different components and thus improve the catalyst performance.

To understand the strategic reaction mechanism and role of base, molecular oxygen and Au catalysts, Davis and co-workers studied the reaction route for oxidation of HMF to FDCA over Au/TiO2 catalyst in alkaline medium (NaOH) using an isotope labeling approach [48,49]. By control experiments, base and a metal catalyst were shown to be important to produce FDCA at 22 ◦C. The aldehyde side-chain of HMF undergoes a rapid reversible hydration to a geminal diol through nucleophilic addition of a hydroxide ion to the carbonyl and subsequent proton transfer from water to the alkoxy ion intermediate (Scheme 4, step 1). This step is due to the incorporation of two 18O atoms in HMFCA when the reaction was performed in H218O. The second step is the dehydrogenation of the geminal diol intermediate, facilitated by the hydroxide ions adsorbed on the metal surface, to form the carboxylic acid HMFCA (Scheme 4, step 2). Further oxidation of the alcoholic group of HMFCA is required to produce FDCA. Base deprotonates the alcoholic group to form an alkoxy intermediate in solution [50]. Hydroxide ions on the catalyst surface then facilitate the activation of the C–H bond in the alcoholic group to form the aldehyde intermediate, 5-formyl-2-furancarboxylic acid (FFCA) (Scheme 4, step 3). The next two steps (Scheme 4, steps 4 and 5) oxidize the aldehyde function of FFCA to form FDCA. These two steps are expected to proceed similarly to steps 1 and 2 for oxidation of HMF to HFCA. The reversible hydration of the aldehyde group in step 4 to a geminal diol accounts for two more 18O atoms incorporated in FDCA when the oxidation is performed in H218O. Overall, complete HMF oxidation to FDCA illustrates that water molecules incorporate all four oxygen atoms in FDCA instead of readily available oxidant (O2). The isotope labeling experiments of 18 O2 and H218O revealed that water was the source of oxygen atoms during the oxidation of HMF to HMFCA and FDCA, probably through direct participation of hydroxide in the catalytic cycle. Molecular oxygen was essential for the production of FDCA and played an indirect role during oxidation by removing electrons deposited into the supported metal particles. Those results provided a fundamental understanding of the roles of added base and molecular oxygen for FDCA production from HMF [49].

Most aerobic oxidations of HMF over Au catalysts are conducted in the presence of excess base, however, considering environmental and economic concerns, base-free HMF to FDCA oxidation systems are more desirable. Thus, some reports on the oxidation of HMF to FDCA over Au-based catalysts without using base were published recently. Gupta et al. reported a hydrotalcite-supported Au catalyst (Au/HT) for the oxidation of HMF into FDCA without using base [43]. An excellent yield of 99% FDCA was demonstrated at 95 ◦C under 1 bar O2 in water after 7 h. Compared to Au deposited on neutral support or acidic SiO2, limited activity was shown, indicating the essential need for basic sites on the catalyst. Although Au/MgO gave a FDCA yield of 21%, it was much lower than that of Au/HT. TEM revealed a much larger size of Au nanoparticles on MgO (>10 nm) than that of Au/HT (3.2 nm), which should be the main reason for the lower catalytic activity of Au/MgO. Although the authors claimed that Au/HT catalyst could be reused, Zope et al. observed a severe leaching of Mg<sup>2</sup>+ from HT over Au/TiO2 catalyst and HT as solid base during the oxidation of HMF, owing to the chemical interaction between the basic HT and the formed FDCA [51]. Further improvement in catalyst stability by modified robust hydrotalcite and activated carbon supported Au catalyst (Au/HT-AC) was demonstrated under base-free conditions [44]. Physical milling of homemade hydrotalcite and commercial activated carbon was applied for catalyst preparation. The Au/HT-AC catalyst showed superior catalytic activity (FDCA Yield = 99.5%) at 100 ◦C, 5 bar O2 pressure after 12 h of oxidation reaction (Table 1) with excellent catalytic stability (6 times). Availability of enough basic sites, large surface area of catalyst, and presence of hydroxyl and carbonyl groups are the reasons for enhanced catalytic performance and improved reusability.

**Scheme 4.** Expanded reaction pathway of HMF oxidation in basic (OH−) media over Au or Pt catalysts (reproduced from [48,49] with permission from Royal Society of Chemistry & Elsevier, copyright 2012 and 2014).

Development of an active and stable bimetallic Au-Pd catalyst also reported by Wan et al [45]. A FDCA yield of 94% could be achieved at 100 ◦C and 0.5 MPa O2 for 10 h, and a FDCA yield of 96% was obtained at 100 ◦C and 1 MPa air for 12 h. The surface carbonyl/quinone and phenol species on CNT was found to facilitate the adsorption of HMF and DFF, rather than FDCA, contributing to the high activity of the Au-Pd/CNT catalyst. In addition, the incorporation of Pd to Au/CNT changed the reaction pathway from HMFCA to DFF route by facilitate the oxidation of the hydroxyl species of HMF, and further enhanced the oxidation of FFCA to FDCA, which is a difficult step for Au catalysts under base-free conditions. Notably, an improved stability with Au-Pd/CNT was also depicted, with marginal loss in activity during the consecutive six runs. Bonincontro et al. further reported an efficient and stable nNiO-supported Au-Pd alloy, with an optimal Au/Pd atomic ratio of 6:4, for base-free oxidation of HMF to FDCA [46]. A nearly quatitative yield of FDCA could be obtained at 90 ◦C after 14 h. NiO was shown to provide basic sites that can promote the reaction and the suitable choice of Au-Pd chemical composition favors the formation of FDCA. Gao et al. reported a highly efficient and stable bimetallic AuPd nanocatalyst over the La-doped Ca-Mg-Al layered double hydroxide (La-CaMgAl-LDH) support for base-free aerobic oxidation of HMF to FDCA in water [47]. A nearly full yield of FDCA could be achieved at 120 ◦C and 0.5 MPa O2 after 6 h. No catalyst deactivation was observed at 100 ◦C and 0.5 MPa O2 after four consecutive runs. The high dispersion of a small amount of La2O3 on the surface of LDH support were attributed to stabilize the support via preventing the deterioration of LDH support by formed carboxylic acid products during reaction, thus resulting in excellent stability and recyclability of AuPd/La-CaMgAl-LDH catalyst.
