Exploiting the Synergetic Behavior of PtPd Bimetallic Catalysts in the Selective Hydrogenation of Glucose and Furfural

Mono and bimetallic catalysts based on Pt and Pd were prepared by a co-precipitation method. They were tested in liquid phase hydrogenation reactions of glucose and furfural at low temperature and pressure. The bimetallic PtPd/TiO2 catalyst proved to be an efficient material in selectively hydrogenating glucose to sorbitol. Moreover, high furfural conversion was attained under relatively soft conditions, and the furfuryl alcohol selectivity was strongly affected by the chemical composition of the catalysts. Furfuryl alcohol (FA) was the major product in most cases, along with side products such as methylfuran (MF), furan, and traces of tetrahydrofuran (THF). These results showed that the PtPd bimetallic sample was more active relative to the monometallic counterparts. A correlation between the catalytic results and the physicochemical properties of the supported nanoparticles identified key factors responsible for the synergetic behavior of the PtPd system. The high activity and selectivity were due to the formation of ultra-small particles, alloy formation, and the Pt-rich surface composition of the bimetallic particles supported on the TiO2 nanowires.


Introduction
In recent decades, the conversion of biomass raw materials to platform molecules has gained interest, as it is one of the most promising routes for the sustainable production of bio-fuels and bio-chemicals [1][2][3][4][5].Furfural and glucose are considered as important biomass-derived compounds among the platform molecules, as they are the source of top most value-added chemicals like sorbitol and furfuryl alcohol, respectively [1,6,7].Furfuryl alcohol is considered as a potential source to produce alkane components in liquid biofuels, and its derivatives have wide applications in the plastics, pharmaceuticals, food, and agrochemical industries [1,[8][9][10].Sorbitol is widely used in the cosmetics, food, and pharmaceutical industries, and is considered as an outstanding building block for commodity chemicals such as isosorbide and sorbitan [4,5,7,11].It has been estimated that more than 60% of the overall furfural produced each year is converted to furfuryl alcohol, whereas the majority of the globally produced sorbitol (~700,000 tons/year) is obtained from glucose [4,7,12].
Due to the aforementioned limitations of the commonly used catalysts in glucose and furfural hydrogenation reactions, noble metal catalysts, mainly based on Pt and Pd, have been studied, as they are known to exhibit high activity, strong deactivation resistance, and regeneration ability in different catalytic reactions [38][39][40].While the Pd catalytic systems have exhibited total conversion or C-C bond cleavage leading to unwanted products, the Pt catalysts have showed higher selectivity at comparatively lower conversions [17,19,[41][42][43].Hydrogenation of furfural performed using bimetallic PdFe and PdCu catalysts proved that the addition of a second metal to a Pd catalyst reduced the C-C cleavage and increased the selectivity to furfuryl alcohol [28,44].Although PtPd bimetallic catalysts have been successfully employed in hydrogenation of various chemicals and are known to possess synergetic behavior [45][46][47][48][49], they have rarely been studied in both glucose and furfural hydrogenation reactions as per our knowledge [42,50].For example, Albilali et al. used a PtPd bimetallic catalyst in furfural hydrogenation to screen for suitable conditions that yield more tetrahydrofurfuryl alcohol, a total hydrogenation product of furfural [50].
In addition to the active phase (noble-metal nanoparticle), the nature of the support of the catalysts is another factor that affects the product distributions in all catalytic systems.Among the commonly known SiO 2 , Al 2 O 3 , TiO 2 , and ZrO 2 supports, TiO 2 -supported catalysts exhibit higher selectivity towards both sorbitol and furfuryl alcohol [51,52].During the hydrogenation of furfural, TiO 2 facilitates the formation of the active furfuryl-oxy intermediate due to charge-transfer interactions between the furfural molecule and an O-vacancy site on the TiO 2 surface, resulting in enhanced furfuryl alcohol selectivity [52].
Based on the aforementioned challenges, we aim in this paper to exploit the synergetic effects of PtPd bimetallic nanocatalysts (relative to their monometallic Pd and Pt counterparts), towards the selective/partial hydrogenations of glucose and furfural to sorbitol and furfuryl alcohol, respectively.Our main focus was to understand the influence of synergetic behavior in well-defined and controlled PtPd bimetallic nanoparticles supported on TiO 2 nanowires as catalysts in selective hydrogenations of glucose and furfural.

Catalyst Characterization
XRD studies were performed in order to confirm the structure of the TiO 2 support and to detect the presence of any large metallic or bimetallic nanoparticles.As shown in Figure 1, no diffraction peaks assigned to metallic Pd or Pt were observed, most probably as a result of the formation of very small metallic nanoparticles (<3 nm NPs) on the TiO 2 nanowire surface.The diffractograms confirmed that the support TiO 2 mainly exists as anatase phase (noted as 'A' on Figure 1) in all catalysts.However, a small contribution from the rutile phase (noted as 'R' on Figure 1) is also confirmed by a peak at 2θ = 28    In order to confirm the presence of Pt and Pd nanoparticles that were not detected by XRD, transmission electron microscopy was performed.The images clearly showed that the TiO 2 support exists as nanowires in all catalysts, as shown in Figure 2A,B.TEM studies confirmed the formation of very small nanoparticles (NPs < 5 nm) of both Pt and Pd metals that are highly and homogeneously dispersed on the TiO 2 surface, as shown in Figure 2C,D.Furthermore, the high-resolution images (HRTEM) confirmed the presence of ultra-small Pt and Pd particles with an average size of about 2 nm, as illustrated in Figure 3.The formation of these highly dispersed ultra-small metallic particles is hence the reason for the absence of metallic Pt and Pd characteristic peaks in XRD.catalyst is necessary to enhance the catalytic activity and durability of metal-based materials [53].
Different methods can be applied [54,55] to obtain homogeneous dispersion of small metallic particles on the support surface, but generally the catalysts with well-dispersed particles are prepared by the traditional impregnation or precipitation methods.In this study, the chemical reduction with NaBH4 in the presence of a stabilizer (polyvinyl alcohol-PVA) during catalyst syntheses permitted us to not only easily obtain zerovalent states in Pt and Pd metals, but also to stabilize particles on the surface of TiO2 supports with low surface area.TEM was also performed for bimetallic Pt0.50Pd0.50catalysts in order to verify if there are any differences in particle size distributions or metal dispersion compared to monometallic catalysts due to the possible formation of alloys in the bimetallic catalyst.This is illustrated in Figure 3.There is no observable differences in particle sizes and dispersion of metallic particles in monometallic and bimetallic catalysts (2.7-3 nm).However, it is difficult to observe the minute differences in the structures of monometallic and bimetallic catalysts due to the presence of very small particle sizes, and so one cannot fully deny the formation of an alloy structure in Pt50Pd50 bimetallic catalysts merely by looking at TEM results.TEM results also confirmed that the reduction-precipitation method used for the syntheses of all catalysts is adequate for the low surface area supports (<50 m 2 g −1 , Table 1) to obtain high dispersion of metallic nanoparticles without any agglomeration.Indeed, the most important problem when working with low surface area oxides is the control of the deposited metallic nanoparticles.It is already well established that a highly dispersed active phase (especially metallic phase) in the catalyst is necessary to enhance the catalytic activity and durability of metal-based materials [53].Different methods can be applied [54,55] to obtain homogeneous dispersion of small metallic particles on the support surface, but generally the catalysts with well-dispersed particles are prepared by the traditional impregnation or precipitation methods.In this study, the chemical reduction with NaBH 4 in the presence of a stabilizer (polyvinyl alcohol-PVA) during catalyst syntheses permitted us to not only easily obtain zerovalent states in Pt and Pd metals, but also to stabilize particles on the surface of TiO 2 supports with low surface area.TEM was also performed for bimetallic Pt 0.50 Pd 0.50 catalysts in order to verify if there are any differences in particle size distributions or metal dispersion compared to monometallic catalysts due to the possible formation of alloys in the bimetallic catalyst.This is illustrated in Figure 3.There is no observable differences in particle sizes and dispersion of metallic particles in monometallic and bimetallic catalysts (2.7-3 nm).However, it is difficult to observe the minute differences in the structures of monometallic and bimetallic catalysts due to the presence of very small particle sizes, and so one cannot fully deny the formation of an alloy structure in Pt 50 Pd 50 bimetallic catalysts merely by looking at TEM results.
Electronic interactions can occur in bimetallic Pt 0.50 Pd 0.50 material, and they can modify the oxidation states of the metals, especially in the case of core-shell structures.For the bimetallic Pt 0.50 Pd 0.50 sample, X-ray photoelectron spectroscopy (XPS) analysis showed the enrichment of Pt on the surface.The trends in the relative percentages of Pd 0 and Pt 0 in the bimetallic sample point out important differences in the interaction between these two metals.The formation of ultra-small Pt-Pd particles with a possible alloy structure was deduced from TEM study, as illustrated in Figure 4. Indeed, the binding energies of Pt 0 4f (4f 7/2 = 72.4eV) and Pd 0 3d (3d 5/2 = 337.0eV) are slightly different from the standard values of bulk Pt 0 (4f 7/2 = 71.30eV) and bulk Pd 0 (3d 5/2 = 334.80eV).These variations in the standard energy peak locations are due to the electronic interactions between the Pt and Pd within the particles.Moreover, the Pt 4f region also showed the existence of Pt 2+ species in the catalysts (about 25%).This is because the surfaces of Pt particles might have been oxidized during the preparation process.The lower intensity of the Pd 3d peak as compared to the Pt 3d peak suggests the existence of local Pt enrichment on the surfaces of nanoparticles in bimetallic catalysts.Electronic interactions can occur in bimetallic Pt0.50Pd0.50material, and they can modify the oxidation states of the metals, especially in the case of core-shell structures.For the bimetallic Pt0.50Pd0.50sample, X-ray photoelectron spectroscopy (XPS) analysis showed the enrichment of Pt on the surface.The trends in the relative percentages of Pd 0 and Pt 0 in the bimetallic sample point out important differences in the interaction between these two metals.The formation of ultra-small Pt-Pd particles with a possible alloy structure was deduced from TEM study, as illustrated in Figure 4.
Indeed, the binding energies of Pt 0 4f (4f7/2 = 72.4eV) and Pd 0 3d (3d5/2 = 337.0eV) are slightly different from the standard values of bulk Pt 0 (4f7/2 = 71.30eV) and bulk Pd 0 (3d5/2 = 334.80eV).These variations in the standard energy peak locations are due to the electronic interactions between the Pt and Pd within the particles.Moreover, the Pt 4f region also showed the existence of Pt 2+ species in the catalysts (about 25%).This is because the surfaces of Pt particles might have been oxidized during the preparation process.The lower intensity of the Pd 3d peak as compared to the Pt 3d peak suggests the existence of local Pt enrichment on the surfaces of nanoparticles in bimetallic catalysts.
Concerning the O 1s XPS region, three different surface oxygen species were identified, namely, lattice oxygen (OL), surface oxygen ions (OS), and oxygen originating from adsorbed species such as water and organic carbon (OW).Binding energies and atomic percentages of all types of oxygen are reported in Table 2. Concerning the O 1s XPS region, three different surface oxygen species were identified, namely, lattice oxygen (O L ), surface oxygen ions (O S ), and oxygen originating from adsorbed species such as water and organic carbon (O W ). Binding energies and atomic percentages of all types of oxygen are reported in Table 2.The distribution of oxygen species does not change significantly within the samples.However, differences in the relative ratio of Me 0 /Me 2+ were observed.The interaction between both metals permits us to obtain a higher value of reduced Pt on the surface (Pt 0 /Pd 0 = 1.23) as compared to the total Pt/Pd ratio of 1.01 (Table 2).

Catalytic Tests
As said above, the activities of all prepared catalysts were tested in liquid phase hydrogenation of both glucose and furfural.

Glucose Hydrogenation
The hydrogenation of glucose to sorbitol is generally performed under a high pressure of H 2 to attain high yields.However, in this work, the glucose hydrogenation tests were conducted at relatively low H 2 pressure (15 bar).All samples showed high activities in the hydrogenation of glucose even at this low pressure.At a glucose:metal ratio of 500, complete glucose conversions were obtained for all three catalysts with different compositions: Pt/TiO 2 , Pt 0.50 Pd 0.50 /TiO 2 , and Pd/TiO 2 .Glucose conversions were reduced to ~50-80% when the glucose:metal ratio increased to 1000.Interesting trends were observed in the sorbitol selectivity results at 80% of glucose iso-conversion, as shown in Figure 5 (glucose:metal ratio of 600).For monometallic catalysts, the sorbitol selectivities varied from 50% to 60%.In contrast, very high sorbitol selectivity of 96% was observed for bimetallic Pt 0.50 Pd 0.50 /TiO 2 catalysts.Zhang et al. reported that larger nanoparticles of Pt (>17nm) and higher H 2 pressure are necessary to obtain higher sorbitol selectivity on Pt/SBA-15 catalysts [19].However, we obtained 80% glucose conversion and 92% sorbitol selectivity even with ultra-small particles (<3 nm) at relatively low H 2 pressure (12 bar) using bimetallic catalysts, which once again confirms the possible positive influence of synergy between Pd and Pt metals on the activity of the bimetallic catalyst, no matter how small the particles and hydrogen pressure are.Furthermore, the influence of support TiO 2 cannot be excluded, as it is known to exhibit higher activity than SiO 2 support due to the presence of the strong metal-support interactions (SMSI) phenomenon and O-vacancy on TiO 2 support [52].
Moreover, the activity was also higher in the case of bimetallic catalysts, compared to their monometallic counterparts.A maximum glucose conversion of 98% was attained with the Pt0.50Pd0.50/TiO2catalyst.Meanwhile, monometallic Pt/TiO2 and Pd/TiO2 catalysts could reach maximum conversions of only 90% and 82%, respectively.These results clearly show that a strong synergistic effect exists between Pt and Pd metals in the bimetallic Pt0.50Pd0.50/TiO2catalyst.This synergy seems to be correlated to the chemical composition of the bimetallic nanoparticles, with a high content of metallic Pt on the surface as observed from XPS studies.

Furfural Hydrogenation Test
Given the very good performances of the Pt0.50Pd0.50/TiO2catalyst in glucose hydrogenation, the efficiency of this material in the reductive conversion of another important biomass-based platform chemical (i.e., furfural) was also tested.It has been well established that the partial hydrogenation of furfural leads to the formation of different products, as shown in Figure 6.All these partial Distribution of products at glucose iso-conversion of 80%.Conditions: 2 mL of aqueous solution of glucose, 10 mg of catalyst, 2 h, glucose:metal (mol) = 600.
Moreover, the activity was also higher in the case of bimetallic catalysts, compared to their monometallic counterparts.A maximum glucose conversion of 98% was attained with the Pt 0.50 Pd 0.50 /TiO 2 catalyst.Meanwhile, monometallic Pt/TiO 2 and Pd/TiO 2 catalysts could reach maximum conversions of only 90% and 82%, respectively.These results clearly show that a strong synergistic effect exists between Pt and Pd metals in the bimetallic Pt 0.50 Pd 0.50 /TiO 2 catalyst.This synergy seems to be correlated to the chemical composition of the bimetallic nanoparticles, with a high content of metallic Pt on the surface as observed from XPS studies.

Furfural Hydrogenation Test
Given the very good performances of the Pt 0.50 Pd 0.50 /TiO 2 catalyst in glucose hydrogenation, the efficiency of this material in the reductive conversion of another important biomass-based platform chemical (i.e., furfural) was also tested.It has been well established that the partial hydrogenation of furfural leads to the formation of different products, as shown in Figure 6.All these partial hydrogenation products are value-added chemicals that can be used as polymer precursors or solvents [3].The formation of tetrahydrofurfuryl alcohol (THFA) involves the reduction of the furan ring, which generally requires relatively high hydrogenation activity of the catalyst [3].The reduction of the aldehyde group of furfural permits the production of furfuryl alcohol (FA), which is used as a green solvent, whereas the hydrogenolysis of the aldehyde group leads to methylfuran (MF), a valuable biofuel [3].Hence, it is necessary to develop a catalyst which is able to selectively and partially hydrogenate the C=O group without reducing the furan ring in order to produce furfuryl alcohol, one of the many aforementioned hydrogenation products.In this section, the ability of the monometallic and bimetallic catalysts to selectively/partially hydrogenate furfural to furfuryl alcohol will be explored.It is generally known that solvents have a great influence on the selective distribution of products during liquid hydrogenation of furfural [29].Hence, the performances of catalysts in furfural hydrogenation were studied in the presence of a protic solvent (isopropanol) and an aprotic solvent (cyclopentyl methyl ether).
The furfural hydrogenation tests were initially performed using isopropanol (protic solvent) as a solvent.The furfural conversions and product distributions obtained on monometallic and bimetallic catalysts are shown in Figure 7.It is generally known that solvents have a great influence on the selective distribution of products during liquid hydrogenation of furfural [29].Hence, the performances of catalysts in furfural hydrogenation were studied in the presence of a protic solvent (isopropanol) and an aprotic solvent (cyclopentyl methyl ether).
The furfural hydrogenation tests were initially performed using isopropanol (protic solvent) as a solvent.The furfural conversions and product distributions obtained on monometallic and bimetallic catalysts are shown in Figure 7.
furfural hydrogenation were studied in the presence of a protic solvent (isopropanol) and an aprotic solvent (cyclopentyl methyl ether).
The furfural hydrogenation tests were initially performed using isopropanol (protic solvent) as a solvent.The furfural conversions and product distributions obtained on monometallic and bimetallic catalysts are shown in Figure 7.The highest furfural conversion was observed for the bimetallic Pt0.50Pd0.50/TiO2catalyst (84%).
The main product for all catalysts was furfuryl alcohol (FA).However, in the case of the bimetallic catalyst, small amounts of methyl furan (MF) and tetrohydrofuran (THF) were also observed.The main inconvenience in using isopropanol as a solvent is the possibility of forming secondary products (low carbon balance) via catalytic transfer of hydrogen from the solvent [54].In this case, the ethyl furfuryl ether was formed via a Meerwein-Ponndorf-Verley reaction.This was confirmed in our case by mass spectroscopy.To overcome this issue, new tests were performed using the aprotic solvent cyclopentyl methyl ether (CPME) in order to eliminate etherification and secondary product The highest furfural conversion was observed for the bimetallic Pt 0.50 Pd 0.50 /TiO 2 catalyst (84%).The main product for all catalysts was furfuryl alcohol (FA).However, in the case of the bimetallic catalyst, small amounts of methyl furan (MF) and tetrohydrofuran (THF) were also observed.The main inconvenience in using isopropanol as a solvent is the possibility of forming secondary products (low carbon balance) via catalytic transfer of hydrogen from the solvent [54].In this case, the ethyl furfuryl ether was formed via a Meerwein-Ponndorf-Verley reaction.This was confirmed in our case by mass spectroscopy.To overcome this issue, new tests were performed using the aprotic solvent cyclopentyl methyl ether (CPME) in order to eliminate etherification and secondary product formation by the transfer of hydrogen.Indeed, when using CPME, the carbon balance reached close to 100%, as can be seen from Figure 8.
Catalysts 2018, 8, x FOR PEER REVIEW 9 of 14 formation by the transfer of hydrogen.Indeed, when using CPME, the carbon balance reached close to 100%, as can be seen from Figure 8.All catalysts showed high selectivities to furfuryl alcohol.In the case of the bimetallic sample, a small quantity of THF was also observed.This indicates that the bimetallic sample could perform hydrogenation of the furan ring, which was not observed for pure monometallic samples.The formation of THF required strong reductive properties, as observed for the bimetallic material.All catalysts showed high selectivities to furfuryl alcohol.In the case of the bimetallic sample, a small quantity of THF was also observed.This indicates that the bimetallic sample could perform hydrogenation of the furan ring, which was not observed for pure monometallic samples.The formation of THF required strong reductive properties, as observed for the bimetallic material.

Catalyst Preparation
The TiO 2 nanowires were obtained by a hydrothermal method as described below.Firstly, 1 g of TiO 2 (P25 Degussa) was dispersed in 75 mL of 10 M NaOH aqueous solution.The solution was stirred for 30 min at room temperature and then transferred into a Teflon-lined stainless steel autoclave.The autoclave was heated and stirred at 250 • C for 72 h, and then the solution was cooled down to room temperature.The obtained precipitate was added into 500 mL of 0.1 M HCl aqueous solution and stirred.The resulting suspension was centrifuged, and the solid was obtained by decantation of the supernatant.This HCl washing was repeated several times.After HCl washing, the solid was also washed with water several times by successive rounds of centrifugation and removal of the supernatant, until the pH was 7. Titanate wires, obtained by this method, were heated in an air atmosphere at 800 • C for 2 h to produce the TiO 2 nanowires.For the synthesis of TiO was mixed and added into the system.After the synthesis, the produced nanomaterials were washed three times with ethanol (15 mL) and water (15 mL) by successive rounds of centrifugation at 6000 rpm for 5 min and by the removal of the supernatant.After washing, the nanomaterials were dried at 80 • C under vacuum.

Catalyst Characterization
The high-resolution transmission electron microscopy (HRTEM) images were obtained with a JEOL JEM2100 microscope (Tokyo, Japan), operated at 200 kV.High-angle annular dark field (HAADF) and energy dispersive x-ray spectroscopy (EDS) images were acquired using a FEI TECNAI G2 F20, operated at 200 kV (Tokyo, Japan).Samples for HRTEM, STEM-HAADF, and STEM-EDX were prepared by drop-casting an aqueous suspension of the nanostructures over a carbon-coated copper grid, followed by drying under ambient conditions.The Pd and Pt atomic percentages were measured by flame atomic absorption spectrometry (FAAS, Tokyo, Japan) with a Shimadzu spectrophotometer (model AA-6300) equipped with an air-acetylene flame.The X-ray diffraction (XRD) data were obtained using a Rigaku-Miniflex equipment (Tokyo, Japan) with CuKα radiation.The diffraction patterns were measured in the range of 0-70 • /2θ, with a 1 • min -1 angular speed scan.X-ray photoelectron spectroscopy (XPS) data of the samples were obtained on a Kratos Axis Ultra electron spectrometer (Shimadzu, Kyoto, Japan), operating with monochromatic AlKα (1486.6 eV) radiation.The pass energy of the hemispherical analyzer was set to 160 eV for the wide scan and 40 eV for the narrow scan.Peak decomposition was performed using curves with 70% Gaussian type and 30% Lorentzian type, and a Shirley nonlinear sigmoid-type baseline.The binding energies (BE) of the Pd 3d, Pt 4f, Ti 2p, and O 1s spectral peaks were referenced to the C 1s peak at 284.8 eV, providing accuracy within ± 0.1 eV.The samples were thermally treated (under 10% hydrogen, 300 • C for 2 h) in an external reactor coupled to the analysis chamber.This permitted quasi in situ analysis of the catalysts before and after thermal treatment.
Textural characteristics of the catalysts were determined from nitrogen adsorption isotherms, recorded at −196 • C in a Quantachrome Nova 1000 (Boynton Beach, FL, USA).The samples (ca. 100 mg) were degassed for 3 h at 150 • C before analysis.Specific surface areas were determined by the Brunauer-Emmett-Teller equation (BET method) from an adsorption isotherm generated in a relative pressure range 0.07 < P/Po < 0.3.The total pore volume was calculated from the amount of N 2 adsorbed at a relative pressure close to unity.The average pore diameter was determined by the Barrett-Joyner-Halenda (BJH) method, from the N 2 desorption isotherms.

Catalytic Tests
Furfural and glucose hydrogenation tests were carried out on the REALCAT platform [22] in a Screening Pressure Reactor system (SPR) from Unchained Labs (London, UK) equipped with 24 parallel batch reactors for high-throughput screening of the catalytic materials.
In the typical glucose hydrogenation reaction, the catalysts (10 mg) were placed in each reactor.Then, 3 mL of an aqueous solution of glucose (1.0 wt.%) were added.The catalytic tests were performed under H 2 (12 bar) for 2 h at 110 • C under non-controlled pH (initial pH = 6).After the reaction, the reactors were cooled down to room temperature.The liquid products were analyzed using High Performance Liquid Chromatography (HPLC, Waters 2410 RJ, Milford, MA, USA) equipped with a Shodex SUGARSH-1011 column (8 × 300 mm) and RI (refractive index) and UV (ultraviolet) detectors.Diluted H 2 SO 4 (5 mM, 1 mL/min) was used as a mobile phase.Commercial standards (glucose and sorbitol) were used for calibration of the HPLC setup.Temperature, initial pressure, hydrogen atmosphere, duration of the reaction, catalyst amount, and initial reactant concentration were set as follows: 110 • C, 12 bar, hydrogen, 2 h, glucose:metal ratio of 500:1 (and 600:1).After the first run, the reactant solution was taken out from the reactor using a syringe.The catalyst was recovered by centrifugation, and then dried at 100 • C overnight.The second run was then performed with the same catalyst amounts and conditions as employed during the first run.
Typical furfural hydrogenation tests were performed at 150 • C under 20 bar of H 2 for 2 h.Before reaction, the catalysts (10 mg) were reduced in situ at 400 • C with pure H 2 for 2 h.Then, the furfural solution (2 mL; 25 wt.% in 2-propanol or CPME) was transferred to the reactors inside a glove box.The reactors were then sealed, charged with H 2 (20 bar), and subsequently heated up to 150 • C. The reaction time was initiated after shaking started (600 rpm).Finally, the system was cooled down to room temperature, and the liquid products were analyzed using a GC (Shimadzu GC 2010 PLUS, Kyoto, Japan) equipped with a ZB-WAX Plus column (30 m × 0.25 mm × 0.25 µm) and a FID (flame ionization detector).GC-MS was used to ensure precision in the identification of the products.
The reactant conversions and the distribution of products were calculated on a mole basis as follows: where mol stands for the number of moles of the reactant or products.

Conclusions
It was shown that the bimetallic Pt 0.50 Pd 0.50 /TiO 2 catalyst is a very efficient material for low-pressure hydrogenation of glucose and furfural reactions.A correlation between the catalytic results obtained and the physicochemical properties of the supported nanoparticles allowed for the identification of key factors responsible for the synergetic behavior of the PtPd system.The high activity and selectivity exhibited by the bimetallic PtPd catalyst was due to the ultra-small particle sizes, alloy formation, and specific Pt-rich composition of the bimetallic particles supported on the TiO 2 nanowires.The catalytic activity of Pt-based materials in the liquid phase hydrogenation reactions of glucose and furfural significantly increased when the reduced Pt was located on the catalyst surface.Besides, the PtPd material with finely dispersed bimetallic particles showed high potential in the reductive transformation of furfural to value-added furfuryl alcohol.This study provides new insights into how the interaction between two metals can be explored to obtain highly active Pt-based NPs that act as excellent catalysts for liquid phase hydrogenation reactions.

Figure 1 .
Figure 1.XRD patterns of the bare TiO2 nanowires support and the TiO2 nanowires support decorated with ultra-small Pd, Pt, and PtPd nanoparticles (NPs).

Figure 2 .
Figure 2. TEM (A,B) and HRTEM (C,D) images of TiO2 nanowires decorated with ultrasmall Pd (A and C) and Pt (B and D) nanoparticles, covering the entire surface of the nanowires.

Figure 1 .
Figure 1.XRD patterns of the bare TiO 2 nanowires support and the TiO 2 nanowires support decorated with ultra-small Pd, Pt, and PtPd nanoparticles (NPs).

Figure 2 .
Figure 2. TEM (A,B) and HRTEM (C,D) images of TiO 2 nanowires decorated with ultrasmall Pd (A and C) and Pt (B and D) nanoparticles, covering the entire surface of the nanowires.

Figure 3 .
Figure 3. TEM particle size distribution of for mono and bimetallic samples.

Figure 3 .
Figure 3. TEM particle size distribution of for mono and bimetallic samples.

Figure 4 .
Figure 4. TEM (A) and STEM high-angle annular dark field (HAADF) (B) images of TiO2 nanowires decorated with ultra-small bimetallic PtPd NPs.The image inset in B shows a STEM-EDX elemental map for a single PtPd nanoparticle at the surface of TiO2 nanowires.

Figure 4 .
Figure 4. TEM (A) and STEM high-angle annular dark field (HAADF) (B) images of TiO 2 nanowires decorated with ultra-small bimetallic PtPd NPs.The image inset in B shows a STEM-EDX elemental map for a single PtPd nanoparticle at the surface of TiO 2 nanowires.
2 nanowires decorated with monometallic Pd or Pt NPs, typically 200 mg of TiO 2 was added to 200 mL of PVP solution (1g/L in EG).The obtained suspension was transferred to a 500 mL round-bottom flask and kept under vigorous stirring at 90 • C for 20 min.Then, 15 mL of 24 mM ascorbic acid aqueous solution and 15 mL of a 24 mM metal precursor solution () were sequentially added to the reaction flask.This mixture was kept under vigorous stirring for another 1 h to produce TiO 2 -Pd or TiO 2 NPs.For the synthesis of TiO 2 nanowires decorated with monometallic PtPd NPs, the same protocol was employed, but 7.5 mL of each 24 mM metal precursor solution (

Table 1 .
Brunauer-Emmett-Teller (BET) surface area and chemical composition of the samples determined by BET and FAAS.

Table 2 .
Binding energies and surface metallic compositions for the TiO 2 nanowires decorated with ultra-small Pt, PtPd, and Pd NPs measured by X-ray photoelectron spectroscopy (XPS) spectroscopy.