Tungsten Promoted Ni/Al2O3 as a Noble-Metal-Free Catalyst for the Conversion of 5-Hydroxymethylfurfural to 1-Hydroxy-2,5-Hexanedione

The conversion of 5-hydroxymethylfurfural (HMF) to 1-hydroxy-2,5-hexanedione (HHD) represented a typical route for high-value utilization of biomass. However, this reaction was often catalyzed by the noble metal catalyst. In this manuscript, W promoted Ni/Al2O3 was prepared as a noble-metal-free catalyst for this transformation. The catalysts were characterized by XRD, XPS, NH3-TPD, TEM, and EDS-mapping to study the influence of the introduction of W. There was an interaction between Ni and W, and strong acid sites were introduced by the addition of W. The W promoted Ni/Al2O3 showed good selectivity to HHD when used as a catalyst for the hydrogenation of HMF in water. The influences of the content of W, temperature, H2 pressure, reaction time, and acetic acid (AcOH) were studied. NiWOx/Al2O3-0.5 (mole ratio of W:Ni = 0.5) was found to be the most suitable catalyst. The high selectivity to HHD was ascribed to the acid sites introduced by W. This was proved by the fact that the selectivity to HHD was increased a lot when AcOH was added just using Ni/Al2O3 as catalysts. 59% yield of HHD was achieved on NiWOx/Al2O3-0.5 at 393 K, 4 MPa H2 reacting for 6 h, which was comparable to the noble metal catalyst, showing the potential application in the production of HHD from HMF.

1-hydroxy-2,5-hexanedione (HHD) was one of the high valueadded compounds obtained from HMF through catalytic hydrogenation (Schiavo et al., 1991;Gupta et al., 2015;Zhu et al., 2019;Yang et al., 2020). The HHD could be used for the preparation of polyols, nitrogen and oxygen heterocycles. Recently, it was reported that the HHD could convert to 2hydroxy-3-methyl-2-cyclopenten-1-one (MCP) through an intramolecular aldol condensation procedure at mild conditions (Duan et al., 2017b;Wozniak et al., 2018). The MCP was a commercialized edible essence produced from petrochemical feedstock by multistep reactions with low yield and severe pollution. The HHD was a potential feedstock candidate for the improvement in the production of MCP.
The transformation of HMF to HHD had been reported by several groups. The reaction was usually conducted in water under H 2 pressure. The formic acid could also be used as the hydrogen source when a homogeneous catalyst was used (Xu et al., 2017). Acid additives such as HCl, Amberlyst-15 or H 3 PO 4 was usually necessary for this conversion (Liu et al., 2014). To avoid acid additives, the introduction of acid sites in the catalyst was a good choice. For example, the reaction could conduct without the addition of acid additives when acid support MIL-101 (Yang et al., 2019b), zeolite (Ramos et al., 2019) or Nb 2 O 5 (Duan et al., 2017b) was used for the preparation of Pd catalyst. The supported Pd was found to be an effective heterogeneous catalyst for this transformation while the Ir complexes were a good homogeneous catalyst candidate (Xu et al., 2017). The Ru (Gupta et al., 2015) complex and supported Au (Ohyama et al., 2014) could also catalyse this reaction. The employment of noble-metal was a barrier for the further application of the conversion of HMF to HHD. To reduce the use of noble-metal, high-performance catalyst with low noble-metal load, high dispersion and activity was designed and applied for this transformation.
The application of noble-metal-free catalyst for the conversion of HMF to HHD without acid additives was a better choice for the improvement. However, there was only one report that HMF in water (~0.3 wt%) could be converted to HHD by Ni 2 P nanoparticles up to now (Fujita et al., 2020). In this manuscript, the tungsten promoted Ni/Al 2 O 3 was simply prepared and used as a noble-metal-free catalyst for the transformation of HMF to HHD. The catalyst showed high activity and stability in the reaction. The introduction of tungsten in the catalyst improved the activity and selectivity to HHD greatly.

Characterization
The X-ray diffraction (XRD) patterns of NiWOx/Al 2 O 3 with different content of W was displayed in Figure 1. Figure 1A were the results of samples after calcined. Except for the diffraction peaks for γ-Al 2 O 3 , Ni/Al 2 O 3 had additional two peaks at 37.  Figure  S1). W species mainly existed as WO 3 after the calcination. However, no diffraction peaks ascribed to WOx could be found for NiWOx/Al 2 O 3 showed that the W had a reasonable degree of dispersion on the support. After the Ni/Al 2 O 3 was reduced in H 2 , the diffraction peaks ascribed to NiO disappeared ( Figure 1B). Two new peaks corresponding to Ni (111) and (200) were immersed at 44.5 o and 51.9 o (PDF#04-0850). This showed that most of the NiO could be reduced to Ni by H 2 reduction, which was consistent with the results of H 2 -Temperature  Figure S2 and Table S1). After the introduction of W, NiWOx/Al 2 O 3 still had the diffraction peaks of Ni. The diffraction peaks moved to a smaller angle with increased W content. This should be caused by the W atoms entering the lattice of Ni.
The surface chemical state was characterized by X-ray photoelectron spectroscopy (XPS) and the results were shown in Figure 2 and Supplementary Figure S3. After deconvolution operation, the Ni/Al 2 O 3 showed two peaks centered at 856.7 and 855.3 eV, which should be ascribed to NiAl 2 O 4 and NiO, respectively ( Figure 2A) (Yang et al., 2016). No peaks (853.0 eV) ascribed to metallic Ni could be found in the XPS spectra. This should be caused by the oxidation of surface Ni when exposed to air. The binding energy shifted to higher energy when W was introduced. The binding energy peaks of NiAl 2 O 4 and NiO were 857.4 and 855.8 eV, respectively for NiWOx/ Al 2 O 3 -0.5. This decrease indicated that the W had interaction with Ni. This interaction between Ni and W should weaken that between Ni and Al. Consequently, the relative content of NiAl 2 O 4 decreased after the introduction of W revealed by the result of XPS ( Figure 2A). The interaction between Ni and W was further proved by the XPS spectra of W ( Figure 2B). The WOx/Al 2 O 3 had peaks at 37.4 and 35.4 eV corresponding to the binding energy of WO 3 of 4f 5/2 and 4f 7/2 (Cao et al., 2014). This value decreased to 37.3 and 35.3 for NiWOx/Al 2 O 3 -0.5. The interaction between Ni and W was also consistent with the results of H 2 -TPR that the reduction temperature peak was changed after the introduction of W (Supplementary Figure S2; Table S1).
The acidic properties of Ni/Al 2 O 3 , NiWOx/Al 2 O 3 -0.5 and WOx/Al 2 O 3 were investigated by the temperatureprogrammed desorption of ammonia (NH 3 -TPD), and the results were shown in Figure 3 and Supplementary Figure  S4. The Ni/Al 2 O 3 showed a sharp peak at 570 K which was ascribed to the weak to medium acid site of Al 2 O 3 . The NH 3 -TPD of NiWOx/Al 2 O 3 -0.5 showed three wide peaks centered at 442, 544, and 700 K corresponding to the weak acid, medium acid, and strong acid sites respectively. The NiWOx/Al 2 O 3 -0.5 posed a relatively strong acid site except for the weak to medium acid site compared to Ni/Al 2 O 3 . This desorption curve was similar to WOx/Al 2 O 3 (Supplementary Figure S4). Hence, the strong acid site was produced due to the introduction of W in NiWOx/ Al 2 O 3 -0.5.
The morphology of Ni/Al 2 O 3 and NiWO x /Al 2 O 3 -0.5 was characterized by transmission electron microscopy (TEM). The TEM images (Figure 4) showed that the Ni was distributed on the surface of the support in granular or rod form for both Ni/Al 2 O 3 , NiWOx/Al 2 O 3 -0.5 in nanoscale. To get a clearer distribution of elements, the energy-dispersive spectrometer (EDS) mapping was taken for both catalysts. The results were shown in Figures 4C,D   shown in TEM images. While the other Ni did not agglomerate and dispersed on the support evenly. This should be ascribed to the unreduced Ni species with high interaction between Al. After the introduction of W, the NiWOx/Al 2 O 3 -0.5 had a similar Ni distribution as Ni/Al 2 O 3 ( Figure 4D). The EDS mappings of W showed that the Ni and W did not have a separated distribution. The W always distributed followed the distribution trend of Ni. This was advantageous for the synergistic effect between Ni and WOx.

Hydrogenation of HMF
The hydrogenation of HMF was conducted in water under a hydrogen atmosphere. We first checked the effect of W for the reaction, and the results were shown in Figure 5. 46% conversion of Frontiers in Chemistry | www.frontiersin.org March 2022 | Volume 10 | Article 857199 4 HMF was obtained when Ni/Al 2 O 3 without W was used as the catalyst. HHD was detected as one of the products, proving that the reaction could be catalyzed by Ni catalyst. However, the selectivity to HHD was only 29% accompanied by 19% selectivity to 3-(hydroxymethyl)cyclopentan-1-one (HCPO) and 12% selectivity to 2,5-Bis(hydroxymethyl) furan (BHMF). Deep hydrogenation products reported in the literature (Yao et al., 2013;Pomeroy et al., 2021;Zhang et al., 2022) for Ni and Ru based catalysts (Ni-Ce/Al 2 O 3 , Ni-Co-Al mixed oxide, and Ru/C) such as 2,5bis(hydroxymethyl) tetrahydrofuran, 1,2,6-hexanetriol and 1,2,5hexanetriol were not found, showing that the catalyst had moderate hydrogenation activity. When a small amount of W was introduced to the Ni/Al 2 O 3 , the conversion of HMF increased to 68% while the selectivity to HHD decreased a little to 25% for NiWOx/ Al 2 O 3 -0.1. A small amount of W had no promoting effect for the selectivity to HHD. However, as the amount of W increased, the selectivity to HHD had a sharp increase to 66% for NiWOx/Al 2 O 3 -0.5 with a little increase in the conversion of HMF. When continued to increase the content of W, the selectivity to HHD maintained around 60% while the conversion of HMF decreased to 59%. An appropriate amount of W was needed for the high performance for hydrogenation of HMF to HHD. When no Ni was used in the catalyst (WOx/Al 2 O 3 ), the HMF was almost unchanged showed that the W did not have the function for hydrogenation. The increased activity for the conversion of HMF could be caused by the interaction between Ni and W which was clued by the XRD, XPS, and H 2 -TPR. Firstly, the electronic state of Ni was changed by the addition of W. Secondly, the interaction between Ni and W weaken that between Ni and Al. So, the content of NiAlO 4 , a species that had no hydrogenation activity, decreased. Therefore, the hydrogenation activity increased after the introduction of W. The high selectivity to HHD should be ascribed to the strong acid site introduced by the addition of W. It was reported that the acids play a key role in the isomerization of furan rings in the conversion of HMF to HHD. The introduced strong acid sites by W were conducive to the rearrangement of furan rings during the reaction. Thus, the selectivity to HHD was enhanced.
The hydrogenation of HMF to HHD was a multistep reaction including hydrogenation and isomerization. The temperature was very important for this multistep reaction. We studied the effect of temperature by conducting the reaction at a temperature between 333 and 453 K and the results were shown in Figure 6. Generally, the conversion should be increased with the rise in temperature. However, the conversion of HMF experienced a process of falling first and then rising. As shown in Figure 6A, the temperature range was divided into three distinct parts. In each temperature range, the conversion of HMF increased smoothly with the rise in temperature. However, there had a sharp descent in the conversion when the temperature increased from 353 to 363 K and a sharp ascend in the conversion when the temperature increased from 403 to 413 K. We first checked the change of catalysts after reacting at different temperature by XRD (Supplementary Figure S9). It could be seen that, both the NiWOx/Al 2 O 3 -0.5 and Al 2 O 3 had no obvious change after reacting at different temperature. This showed the catalysts was stable at the reaction condition. The changes in conversion could be interpreted by the different reaction pathways revealed by the change in selectivity (Figure 6B). At a temperature lower than 363 K, the main product was BHMF which was the hydrogenation of aldehyde in HMF. Both the selectivity to HHD and HCPO, which should be produced by the isomerization of furan rings, was very low. As the reaction temperature raised from 333 to 363 K, the selectivity to BHMF decreased while the selectivity to HHD increased. The maximum increase in the selectivity to HHD was observed when the temperature increased from 353 to 363 K. The hydrogenation active center only required the hydrogenation of aldehyde group for BHMF. However, both the aldehyde group and intermediates needed to be hydrogenated by the hydrogenation active center for the production of HHD. As a result, there had a decline in conversion when the temperature increased from 353 to 363 K. A similar phenomenon also occurred at temperatures increased from 403 to 413 K. The selectivity to HHD decreased sharply when the temperatures increased from 403 to 413 K. At the same time, the selectivity to HCPO began to increase. In a word, the hydrogenation active center should play the role of hydrogenation function in multiple steps that led to the low conversion in the transformation of HMF to HHD. The suitable temperature range for high selectivity to HHD was from 363 to 413 K.
The effect of H 2 pressure and reaction time was studied to optimize the reaction conditions. The conversion of HMF increased with the increase in H 2 pressure (Figure 7). When the H 2 pressure was lower than 4 MPa, the selectivity increased with the increase in pressure. The highest selectivity to HHD was achieved at 4 MPa H 2 . The conversion increased with the extension of time ( Figure 7B). When the reaction was conducted at a shorter time, there had a low selectivity to HHD and BHMF, the intermediate for HHD, was found as the main products. However, as the reaction prolonged to 2 h, the selectivity to HHD increased to 72% while that to BHMF decreased to 3%. The selectivity to HHD kept around 65% when further extension of time. The highest yield of HHD was 59% which was obtained after 6 h reaction.
Frontiers in Chemistry | www.frontiersin.org March 2022 | Volume 10 | Article 857199 6 nickel/alumina was used for the hydrogenation of HMF in water. However, the selectivity to HHD was typically low (less than 15%) when no acid was used. When 10 mg of AcOH was added, there had a little decline in the conversion of HMF. However, the selectivity to HHD increased from 29 to 59% (Table 1, Entry 2). This showed that the acid was conducive to improving the selectivity to HHD rather than the conversion of HMF. Further to increase the amount of acetic acid, there was no significant further improvement in the selectivity to HHD ( Table 1, Entries 3-6). The highest selectivity for HHD was 61% which was acquired with 40 mg of AcOH. These experiments proved the role of acid for the high selectivity to HHD.
Based on the characterization and the hydrogenation of HMF, the reaction pathway and the role of catalyst was proposed as shown in Figure 8. The HMF was firstly hydrogenated to BHMF on Ni. Based on the literature and our previous work (Duan et al., 2017b;Martínez-Vargas et al., 2017;Ramos et al., 2017;Fujita et al., 2020), the BHMF was ready to isomerize to 1-hydroxyhex-3-ene-2,5-dione (HHED) catalyzed by acid. This was the crux for the reaction. In this work, the WOx played the role of acid to catalyze this transformation. At last, the HHED was hydrogenated to HHD on Ni. The suitable acidity of WOx and the moderate hydrogenation activity was the key for this multistep tandem reaction. The adjacent distribution of Ni and WOx accelerated the conversion of intermediates thus avoiding possible polymerization side reactions.

CONCLUSION
In conclusion, the conversion of HMF to HHD was achieved by noble-metal-free W promoted Ni/Al 2 O 3 . The Ni and W uniformly dispersed on the surface of the support. The interaction between W and Ni increased the activity of Ni/ Al 2 O 3 for the hydrogenation of HMF. The introduction of W generated strong acid sites, which were the key for the high selectivity to HHD. The role of acid was proved by the addition of AcOH to unpromoted Ni/Al 2 O 3 . A suitable temperature was needed for the transformation of HMF to HHD smoothly. After the optimization of the conditions, a 59% yield of HHD was acquired at 393 K, 4 MPa H 2 reacted for 6 h on NiWOx/Al 2 O 3 -0.5. This work provided the idea for high selectivity to HHD from HMF by the introduction of suitable acid and improved the feasibility of putting this reaction into practical application by using a non-noble metal catalyst.

Characterization
A Rigaku D/Max 2500/PC powder diffractometer was used to collect the X-ray diffraction (XRD) patterns. Cu Kα radiation at 40 kV was used as the X-ray source. Thermo Escalab 250Xi spectrometer with Al Kα was used to characterize the X-ray photoelectron spectroscopy (XPS) spectra. The sample powder was overspread on a double-faced adhesive tape on aluminum foil. The sample was pressurized to 8 MPa for 30 s and used for measurement. Before measurement, the chamber pressure was vacuumized to <1 × 10 −10 mBar. The binding energy (BE) was adjusted by the binding energy of C1s. The transmission electron microscopy (TEM) images and energy-dispersive spectrometer (EDS) elemental mappings were taken on a JEOL JEM-2100 F field emission transmission electron equipped with An Oxford 80T detector. The temperature-programmed desorption of ammonia (NH 3 -TPD) and H 2 -Temperature programmed reduction (H 2 -TPR) was conducted on the Micromeritics AutoChem II 2920 Instrument. Typically for NH 3 -TPD, 60.0 mg of sample were loaded into the sample tube. The sample was heated to 773 K under He flow (10 mL/min) and kept for 2 h. The tube was cooled to 373 K and 10NH 3 -He (30 mL/min) was introduced for 0.5 h. Then the atmosphere was switched to He (10 mL/min) and kept for 1 h to remove the physical adsorbed NH 3 . After that, the temperature was increased (10 K/min) from 373 to 973 K in an atmosphere of He (10 mL/min). The desorbed NH 3 was detected by the thermal conductivity detector (TCD). For H 2 -TPR, the calcinated catalyst (85.0 mg) was degassed at 573 K under an atmosphere of Ar (10 mL/min) for 2 h. The sample was cooled to 373 K. The temperature was increased (10 K/min) from 373 K to 973 k under the atmosphere of 10H 2 -Ar (30 mL/min). The H 2 consumption was monitored by a TCD detector.

Preparation of Catalysts
In a typical procedure, Ni(NO 3 ) 2 ·6H 2 O (2.50g) and (NH 4 ) 6 H 2 W 12 O 40 ·xH 2 O (1.05 g) were dissolved in water (10.00 g). Then Al 2 O 3 was added to the solution. The mixture was stirred evenly to form a paste and kept standing for 24 h. Then, the paste was dried at 393 k overnight. Followed by calcined at 823 K in the air for 4 h. The obtained solid was ground to pass through 100 mesh sieve and reduced at 773 K in H 2 to afford the NiWOx/Al 2 O 3 -0.5. The 0.5 referred to the mole ratio of W to Ni.

Catalytic Hydrogenation
The hydrogenation reaction was conducted in a 20 mL stainless steel reactor. Typically, the HMF aqueous solution (2.00 g, HMF: 126.0 mg), catalyst (20.0 mg), and magneton were put into a glass lining. The lining was set in the reactor and sealed and purged with H 2 for 4 times to displace the air. Then the reactor was filled with H 2 at a specified pressure and put in an oil bath set at a certain temperature. After the reaction, 0.5 mL ethanol solution of the internal standard (n-decane) was added and the mixture was diluted to 10 mL by ethanol. After centrifugation, the liquid was used for analysis. The qualitative analysis was conducted by GC on a Shimadzu GC-2014 equipped with a SH-Rtx-1701 column (30 m × 0.32 mm × 0.25 µm). The oven temperature was started from 353 K for 2 min and raised to 523 K with 20 K/min heating rate. The oven was kept at 523 K for 1.5 min. The GC-MS was performed on Shimadzu GC/MS-TQ8040 equipped with an SH-Rxi-5Sil MS column (30 m × 0.25 mm × 0.25 µm). The oven temperature was started from 323 K for 1 min and raised to 473 K with 40 K/min heating rate and then raised to 553 K and kept at the temperature for 5 min.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.