Comparative Study of Supported Monometallic Catalysts in the Liquid

Subscriber access provided by TUFTS UNIV

Article

A comparative study of supported monometallic catalysts in the liquid-phase hydrogenation of furfural: batch vs. continuous flow Yantao Wang, Pepijn Prinsen, Konstantinos S. Triantafyllidis, Stamatia A Karakoulia, Pantelis N. Trikalitis, Alfonso Yepez, Christophe Len, and Rafael Luque ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.8b00984 • Publication Date (Web): 19 Jun 2018 Downloaded from http://pubs.acs.org on June 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

A comparative study of supported monometallic catalysts in the liquid-phase hydrogenation of furfural: batch vs. continuous flow Yantao Wang,a Pepijn Prinsen (ORCID iD 0000-0002-5013-7174),b Konstantinos S. Triantafyllidis,c,d Stamatia A. Karakoulia,d Pantelis N. Trikalitis,e A. Yepez,b Christophe Len (0000-0001-8013-0881),a,f and Rafael Luque*,b (ORCID iD 0000-0003-4190-1916) a

Sorbonne Universités, Université de Technologie de Compiègne, Centre de Recherche

Royallieu, CS 60319, F-60203 Compiègne cedex, France. b

Departamento de Química Orgánica, Universidad de Córdoba, Campus de Rabanales, Edificio

Marie Curie (C-3), Ctra Nnal IV, km. 396, E-14014 Cordoba, Spain. c

Department of Chemistry, Aristotle University of Thessaloniki, University Campus, P.O. Box

116, GR-54124 Thessaloniki, Greece. d

Chemical Process & Energy Resources Institute, CERTH, Thermi P.O. Box 60361, GR-57001

Thessaloniki, Greece. e

Department of Chemistry, University of Crete, Voutes, 71003 Iraklio, Greece.

ACS Paragon Plus Environment

1

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

f

Page 2 of 49

PSL Research University, Chimie ParisTech, CNRS, Institut de Recherche de Chimie Paris, 11

rue Pierre et Marie Curie, F-75231 Paris Cedex 05, France.

Correspondence to: Fax: +34957212066. E-mail: [email protected] KEYWORDS Furfural, Hydrogenation, Monometallic carbon supported catalysts, Methylfuran, Batch, Continuous flow

ABSTRACT: This work compares the catalytic performance of well-defined monometallic catalysts supported on commercial micro/mesoporous carbon (Cu/AC, Pd/AC, Pt/AC and Ni/AC) for the liquid phase hydrogenation of furfural, both in batch and in continuous flow. Whereas Ni/AC performed better in batch (in terms of selectivity to 2-methylfuran), better results were obtained with Pt/AC in continuous flow. Re-utilization of the Ni/AC catalyst recovered by simple filtration after batch experiments changed the selectivity from 2methylfuran to furfuryl alcohol. The solvent type and the reaction conditions were essential for the catalytic performance. Metal leaching was the main cause of catalyst deactivation.

INTRODUCTION Important progresses were established in the field of lignocellulose biorefinery for the production of biofuels and biochemicals, including pretreatments1-7 and catalytic conversions.8-10 Starting from primary biomass sources, recent advances have focused mainly on the integration of cellulosic biofuel production with the isolation of lignin fractions for depolymerization to aromatics/phenols,11-13 whereas hemicelluloses are often refined starting from co-generated side streams.7 Hemicelluloses are composed of different carbohydrate


2

Page 3 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


monomers (xylose, arabinose, etc.) and can potentially provide a wide range of different chemicals to replace current ones derived from crude oil. The main disadvantage of their heterogeneity is the fact that C-5 sugars are not converted biochemically as efficient as C-6 sugars (using whole cells or enzymes). Chemocatalytic conversion strategies starting from hemicelluloses are therefore an important research area. Advances were recently reported on the catalytic valorization of hemicelluloses (including their oligomers/monomers in waste or side streams) towards the high value platform chemicals 5-hydroxymethylfurfural14 and furfural,15 produced from glucose and xylose, respectively. The main catalytic routes studied are the conversion to levulinates, lactones and furans.16 Furfural is an important biorefinery platform chemical, having an annual production near 300 kton.17 It can be converted to a wide range of chemicals, either by catalytic oxidation18 or catalytic reduction.15 Other valorization pathways are catalytic amination19 and conversion into aromatics via Diels–Alder cycloaddition with alkenes.20 A wide range of chemicals can be produced in the catalytic hydrogenation of furfural:15 furfuryl alcohol (FA), 2-methylfuran (MF), 2-methyltetrahydrofuran (MTHF), tetrahydrofurfuryl alcohol (THFA), furan, tetrahydrofuran (THF) and pentanediols. Most of the research on catalytic reduction reports high selectivity to FA, which can be applied in the synthesis of renewable materials such as polymers,21 resins22 and carbons.23 But, it can also be further hydrogenated to other compounds, including MF, which shows superior performance as an additive in gasoline fuels15,24 or which can be coupled with cyclopentanone and subsequently hydrotreated for the synthesis of diesel or jet fuel range cycloalkanes.25 The relative low boiling point of MF (63 °C) enables its efficient separation from the other co-products of furfural hydrogenation. Many catalysts have been proposed for the liquid phase hydrogenation of


3


Page 4 of 49

furfural. Most of them consist of supported metal particles: Cu,26-31 Pd,29-35 Ni,26,29,35-38 and Pt,30,39,40 while bimetallic systems,26-28,31,32,35 also with Ru33 and Ir,41 have shown positive effects on the catalyst performance. The reaction conditions, such as solvent type, play also an important role. Whereas protic solvents such as methanol,33,39 ethanol35,39-41 and isopropanol26-28,30,35,37,40 offer high catalytic activity, they often lead to lower product selectivity due to the formation of side products, e.g. alkyl (hemi)acetals and/or alkyl ethers.39,41-44 Other side products include polymeric compounds,33 especially in acidic aqueous media (e.g. humins). Aqueous phase hydrogenation is therefore reported in only few cases.31,36,40 The extent of side product formation remains not well defined as mass balances were often neglected. Aprotic organic solvents such as ethyl acetate,29,34 toluene,33,35,39,40 dichloroethane,33 dioxane,34 n-octane33 and supercritical CO231 have also been reported in liquid phase furfural hydrogenation. Operational parameters such as temperature,26-28,30-32,36-38,40 reaction time,30,31,35,37,38 and (hydrogen) pressure30,37,38 also induce important effects. Most of these studies are based on batch reactor experiments,2628,30,31,33-41

although recently studies on continuous flow hydrogenation also have been

reported.29,30,32 To our knowledge, only Biradar et al. (2014) compared the catalyst performance between batch and continuous flow conditions.30 It is evident from the literature that a high number of variables determine the final product composition, which makes comparison of the catalyst performance rather complex. In this context, the present study aims to compare the performance of well-defined monometallic Cu, Pd, Pt and Ni based catalysts supported on commercial micro-mesoporous carbon with neutral surface properties, allowing to compare the effect of the metal type in a more straightforward way. The experiments were conducted both in batch and continuous flow reactors to study the effect on the catalyst performance.


4

Page 5 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


EXPERIMENTAL SECTION Materials and reagents: Furfural (99 % purity), furfuryl alcohol (FA, 98 % purity), 2methylfuran (MF, 99 % purity), tetrahydrofurfuryl alcohol (THFA, 99 % purity), 2methyltetrahydrofuran (MTHF, 99 % purity), tetrahydrofuran (THF, 99.5 % purity), 1,4pentanediol (PD, 99 % purity) and n-octane (99 % purity) were purchased from Sigma-Aldrich and further used without purification. The metal precursor salts were purchased from Alfa Aesar (H2PtCl6.xH2O), Strem (PdCl2 and Ni(NO3)2.6H2O) and Merck (CuCl2.2H2O). The commercial carbon SX-Plus was purchased from Norit®. Catalyst synthesis and characterization: The 10%Cu/AC, 5% Ni/AC, 10%Ni/AC, 3%Pd/AC and 3%Pt/AC catalysts were prepared by the classic wet impregnation method. First, Norit® SXPlus carbon was activated via thermal treatment at 500 °C under N2 flow for 3 h (activated carbon, AC). In a typical preparation procedure, the required amount of the metal salt precursor was dissolved in 20 mL of deionized water and the solution was added dropwise to a suspension of 10 g AC in 100 mL water under stirring (300 rpm). Stirring was kept for 1 h before the water was removed in a rotary evaporator. The resulting catalysts were dried overnight at 100 °C and were then calcined in air at 500 °C for 3 h under N2 flow (50 mL min-1), followed by treatment at 350 °C (Pt), 400 °C (Pd) and 450 °C (Cu and Ni) for 3 h under H2 flow (50 mL min-1). The catalysts were characterized by N2 porosimetry. Nitrogen adsorption/desorption experiments at 196 °C were performed for the determination of specific surface area (multi-point BET method), total pore volume (at P/Po = 0.99), mircopore area by t-plot analysis and pore size distribution (BJH method using adsorption data or DFT analysis) of the samples which were previously outgassed at 150 °C for 16 h under 5x10-9 Torr vacuum, using an Automatic Volumetric Sorption Analyzer (Autosorb-1MP, Quantachrome). The catalysts were also characterized by X-


5


Page 6 of 49

ray powder diffraction (XRD); experiments were conducted on a Shimadzu X-ray 7000 diffractometer using a CuKα X-ray radiation operating at 45 kV and 100 mA. Counts were accumulated in the range of 5-75º with 0.02º steps (2θ) with counting time 2 s per step. The crystal size of the metal nanoparticles was estimated by the Scherrer equation based on the characteristic peaks (2θ) at 43.3° for Cu(0) and 36.3° for Cu2O, 40.1° for Pd(0), 39.7° for Pt(0), 44.3° for Ni(0) and 43.2° for NiO. 5%Ni/AC catalysts were also analyzed by X-ray Photoelectron Spectrometry (XPS); measurements were carried out with a Phi Quantera Scanning Xray Microprobe (Chevron Energy Technology Company, Richmond, CA, USA) instrument using Al Kα (h·ν=1486.7 eV) radiation at. The instrument was equipped with a hemispherical energy analyzer with multichannel detection and an energy resolution of 1.1 eV. The catalysts were mounted on double-sticky tape (0.8 cm x 0.8 cm). The tape was completely covered by the catalyst powder and the sample surface was smoothed. Spectra were collected for C-1s, O-1s, Ni-2p3 and W-4f photoelectron peaks. Total spectral accumulation times were 100 min per analysis area while irradiating with 100 W of X-ray irradiation. The binding energies (BE) were referenced to the C-1s peak (284.8 eV) to account for charging effects. The XPS spectra were deconvoluted using Gaussian/Lorentzian shaped curves and an iterative least square algorithm provided in Phi Multipak software. The peak areas were computed according to the fitting quality to quantify the Ni and NiO species contents. The metal content of the same set of samples was determined via inductively coupled plasma optical emission mass spectrometry (ICP-MS, Perkin Elmer, NexION 350X) using the corresponding acid digests. Acid digestions were performed with nitric acid in individual N2 pressurized (40 bar) reaction chambers in microwave equipment (Milestone, UltraWave), with the temperature raising during 25 min until reaching 220 °C, which was hold during 15 min. The total digestion time was 50 min and the final HNO3


6

Page 7 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


concentration was 2 %. Transmission electron microscopy (TEM) images were obtained using a JEM-2100 high resolution instrument, equipped with LaB6 filament, operating at 200 kV. Samples were prepared by dipping holey-carbon-coated 200 mesh copper grids, directly into powders. The parent AC material and the corresponding catalysts were also characterized by potentiometric titration measurements with a Mettler Toledo T50 automatic titrator. For the measurement 0.1 g of carbon was added to 50 mL of NaNO3 electrolyte solution (0.01 M) and was stirred overnight. Then, the pH of the solution was adjusted to 10 using NaOH (0.1 M) and was titrated with HCl (0.1 M) under N2 atmosphere. The titrations were carried out for a wide range of pH. The total surface charge, Q (mmol g-1) was calculated using eq. (1).

(1) CA and CB are the acid and base concentrations (M), respectively, [H+] and [OH-] are the equilibrium concentrations of these ions (M) and W is the solid concentration (g L-1). Procedure for batch hydrogenation experiments: The experiments in batch mode were carried out in a stainless-steel autoclave equipped with a thermocouple and magnetic stirring (800 rpm), using 25 mmol furfural or furfuryl alcohol (FA), 12 mmol n-octane (external standard, ES) and 300 mg catalyst in 60 mL solvent. This amount of catalyst (10%Cu, 5%Ni, 10%Ni, 3%Pd and 3%Pt) corresponds with a theoretical content of 1.9 % Cu, 1.0 % Ni, 2.0 % Ni, 0.3 % Pd and 0.2 % Pt, respectively (molar percentages with respect to furfural or FA load). After 3 times purging, 30 bars of H2 were added. The reaction mixture was stirred and heated to 200 °C (ca. 60 bars). The target temperature was reached after ca. 45 min at which time was set to zero. After the reaction, the stirring was stopped and the autoclave was cooled in an ice-batch until reaching a temperature below 35 °C. The residual pressure was read off before releasing the remaining gas


7


Page 8 of 49

and opening the autoclave lit. A schematic illustration of the batch experiments is shown in Figure S1 (see Supporting Information). All results are expressed in molar percentages with respect the substrate load. The furfural conversion and the yield of each hydrogenation product (P) with the corresponding selectivity were calculated with eq. (2), (3) and (4), respectively. The number of moles was calculated from the corresponding concentration (mg mL-1), the corresponding molecular weight and the reaction volume in batch experiments (64 mL). The turnover frequency (TOF) was calculated according to eq. (5), using the metal content (wt%) in monometallic (n = 1) and bimetallic catalysts (n = 2), their corresponding atomic mass (AM, g mol-1), the catalyst load (g) and the reaction time (h). The metal content was based on ICP-MS analysis or similar quantitative methods. When no metal content data was provided, the theoretic metal content was used based on the metal loading used in the synthesis procedure. (2)

(3)

(4)

(5)

Procedure for continuous flow experiments: Continuous flow experiments were carried out in a H-Cube Pro Flow Reactor (ThalesNanoTM, Hungary, see Figure S2), which supplied a continuous H2 flow produced from the electrolysis of water to the central reactor module in which a 30 or 70 mm CatCart was installed, packed with ca. 120 and 280 mg catalyst,


8

Page 9 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


corresponding with reactor volumes of 0.38 and 0.88 mL, respectively. The corresponding reactor volumes packed with catalyst were 0.21 and 0.48 mL, respectively. The corresponding total flow through volumes (including feed lines, reactor and product lines) were 5.0 and 5.5 mL, respectively. The maximum operational temperature allowed for this H-Cube equipment was 150 °C. Full hydrogen mode was used, which corresponds to a hydrogen production capacity reaching 60 mL min-1. A filter was installed in the entrance of the pump to avoid undissolved compounds entering the system. First, pure solvent was pumped through the system before reaching the set temperature (130-150 °C) and pressurization (0-50 bars). Once the reaction conditions were stable, the pure solvent feed was changed to the furfural feedstock. Then, in function of the flow rate (0.1-0.5 mL min-1), the reaction proceeded during a certain time (20-100 min) before collecting the first sample (time zero). Further samples were collected after regular time intervals (20-40 min). The conversion, yield and selectivity were calculated identically as in the batch experiments using concentrations instead of moles (evaporation losses were neglected). TOF was calculated identically as in batch experiments, where MolFurfuralInitial in eq. (4) was calculated by the product of the feed concentration (M) with the flow rate (L h-1) and the reaction time (h). The weight hourly pace velocity (WHSV) on metal basis was calculated according to eq. (6). (6)

Product analysis: The concentrations of furfural and hydrogenation products were determined by GC-FID. First, the response factors (RF) were experimentally determined by measuring calibration curves for furfural, FA, THFA, MF and MTHF were determined in the 0-25 mg mL-1 range, using n-octane (8.1 mg mL-1) as the internal standard. The correlation coefficients of the linear regressions were > 0.99 in all the calibrations. The GC-FID analyses were performed on a


9


Page 10 of 49

AutoSystem XL gas chromatograph (PerkinElmer, Norwalk, CT 06859, United States) coupled with a FID detector equipped with a AT™-1HT GC CAPILLARY COLUMNS (30m × 0.25mm i.d. and 0.1 µm film thickness; Alltech Part No.16368). N2 was used as carrier gas at a rate of 1 mL min-1. The samples were injected with an auto-injector directly onto the column using septum-equipped programmable injector (SPI) system in split mode (20.8:1 ratio). The temperature of the injector was set 350 °C and the oven started at 50 °C, held for 6 min, raised to 130 °C at a rate of 10 °C min-1 and then raised to 250 °C at a rate of 30°C min-1 and held for 2 min. The ionization mode was FID (70 eV, 300 µA, 300 °C). The identification of the compounds was performed by comparison of the retention times with pure standards. The RF of the remaining compounds identified in the product mixtures, 2-(isopropoxy)-methylfuran (iPrOMF) and 2-(isopropoxy)-methyltetrahydrofuran (iPrOMTHF), were calculated by the Effective Carbon Method (ECN)45 relative to MF and MTHF using eq. (7): (7)

RESULTS AND DISCUSSION Catalyst synthesis and characterization: A series of monometallic catalysts supported on commercial activated carbon were prepared via wet impregnation (10%Cu/AC, 5%Ni/AC, 10%Ni/AC, 3%Pd/AC and 3%Pt/AC), followed by thermal treatment in inert and H2 atmosphere. The porosity characteristics (from N2 sorption experiments) of the parent activated carbon (AC) and the AC supported metal catalysts are shown in Table 1. The parent AC is a microporous carbon with a high contribution of meso/macroporosity. About 65 % of its total surface area and 36 % of its total pore volume is attributed to micropores, the rest being attributed to meso/macropores and external surface. The N2 isotherm shapes (Figure S4 in the


10

Page 11 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Supporting Information) exhibit a progressive increase of adsorbed N2 over the whole P/Po range without abrupt changes, indicating the presence of a relatively disordered network of meso- and macropores. Due to this size inhomogeneity, no distinct peaks in the BJH pore size distribution curves were observed, while only a weak peak was identified in the DFT pore size distribution curve with an average diameter of 3.4 nm. The total and micropore surface area, as well as the pore volumes of the metal loaded AC catalysts were essentially unchanged compared to those of the parent AC (decrease of max. 10 %). The catalysts were also characterized using powder XRD analysis and the respective patterns are shown in Figure S5. The 5%Ni/AC catalyst consisted essentially of metallic Ni(0) nanoparticles with average crystal size 6.8 nm. As the Ni content increases to 10%, the Ni(0) crystal size increased to 23.2 nm and a small but XRD visible NiO phase appeared with crystal size 6.1 nm. The crystal sizes of 10%Cu/AC, 3%Pt/AC and 3%Pd/AC were 23.2 nm (a Cu2O phase was also detected with a crystal size of 16.6 nm), 13.6 nm and 16.6 nm, respectively. The dispersion and average size of the metal nanoparticles was also studied by TEM (representative images are shown in Figure S6). An even distribution of relatively uniform, almost spherical Ni nanoparticles with sizes of 5-15 nm (mostly around 10 nm) can be observed for the 5%Ni/AC catalyst. In the case of 3%Pt and 3%Pd/AC catalysts, metal nanoparticles of 10-30 nm (mostly around 20 nm) were observed, being more sparsely distributed on the carbon surface due to their lower loading compared to the 5%Ni/AC. On the other hand, relatively big aggregates of metal nanocrystals with sizes of up to ca. 0.2-0.4 µm were observed in the images of the 10%Cu/AC catalyst due to the high metal loading. The surface heterogeneity in terms of acidic and basic groups, both of the parent AC as well of the corresponding materials after the deposition of metal nanoparticles, was evaluated by potentiometric titration (PT) measurements. The resulting proton binding curves are shown in


11


Page 12 of 49

Figure S7, while the acidic and basic groups estimated from the PT curves are presented in Table S1. From the shape of the curve of the parent AC, a relatively neutral surface can be suggested, resulting from a balance between the acidic and basic sites. This was supported by the pH value of the AC aqueous suspension (7.2). The deposition of the metals, i.e. 5% Ni and 3% Pt and Pd, including their reduction under H2, did not induce any significant alterations in the surface chemistry, since no significant change in the slope/shape of the PT curves was observed as well as no vertical shift of their position with regard to the X axis (Q = 0). Only in the case of 10%Cu/AC, there is an observable alteration in the acidic area, i.e. a change in the slope of the PT curve, indicating increased heterogeneity of the surface groups, leading to higher Q values at lower pH. Both effects can be attributed to the presence of relatively high amounts of Cu(0) and CuO phases on the AC surface. The pH values of the aqueous suspensions of the all the supported metal catalysts were in the range of 7.4 – 8.4, values resulting from the balance of acidic and basic groups on the surface.

((Table 1))

Batch hydrogenation experiments: Table 2 shows the results of experiments conducted in isopropanol (i-PrOH) at 200 °C for 5 h using 30 bars H2. The blank experiment (entry 1) gave almost negligible furfural conversion. High selectivity to FA was seen for the Cu/AC catalyst, but only at low conversion (entry 2). Using Pd/AC catalyst (entry 3), the conversion reached 47 %, with 45 % selectivity to FA, below the values reported in previous works with Pd catalysts.2734

Much higher conversions were achieved with the Pt/AC and Ni/AC catalysts (entries 4 and 5).


12

Page 13 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Whereas Pt/AC was more selective to FA (51 %), high selectivity to MF was achieved with the Ni/AC catalyst (78 %), giving the highest yield reported at temperatures ca. 200 °C in liquid phase.42 The high hydrogenation activity on Ni/AC is also demonstrated by comparing the residual H2 pressure (Figure 1), as it takes more hydrogen to remove the oxygen functionality in furfural to produce MF. In the absence of H2 (entry 6), only 10 % conversion was observed. When using the spent Ni/AC catalyst (entry 9), recovered by simple filtration and drying, the selectivity changed drastically from MF to FA.

((Table 2))

((Figure 1))

In an attempt to compare the catalytic activity with previously reported mono- and bimetallic Cu, Pd, Pt and Ni-based catalysts in liquid phase furfural hydrogenation using H2 as the reductant, the turnover frequency (TOF) was calculated (Table 3). A more accurate way to compare TOF values is to consider the active sites only rather than considering the total metal content, which can be assessed by sorption analysis of probe molecules such as H2. This way effects like metal dispersion are taking into account for the comparison of activity. Unfortunately, these data are not available in the present study, neither in most of the previously reported studies. Another important source of heterogeneity among the data in the literature, apart from the different reaction conditions, is the fact that TOF values highly depend on the


13


Page 14 of 49

reaction coordinate considered for their calculation, at least in batch conditions. For instance, TOF calculated after 0.1 h and after 4 h (Table 3, entries 19 and 20) were 68 and 4 h-1, respectively, for the same catalyst under identical reaction conditions. However, in Table 3 the catalytic activity is compared at near-equilibrium conditions, i.e. the point at which the furfural and product concentrations did not vary significantly anymore over time (after 1.5-7 h). The results show that the activity of the 10%Cu/AC catalyst was relatively low compared to previously reported Cu-based catalysts (entries 1-3). The 3%Pd/AC catalyst exhibited lower activity as well, but the product (entry 9) consisted in great part of ring hydrogenated products (THFA) whereas in the literature (entries 4-8) FA was the main product. The 3%Pt/AC catalyst (entry 14) performed similar to Pt catalysts reported in literature (entries 10-14). The effect of the metal content is most noticeable (entries 12 and 13). This trend is also observed for Ni catalysts; whereas in the literature high Ni contents exhibited relatively low activity (TOF between 1-4 h-1, entries 15-18 and 20), the 5%Ni/AC catalyst in the present study presented high activity (TOF 18 h-1, entry 21). Note that previously reported Ni-based catalysts did not produce any MF, whereas in the present study high selectivity to MF was achieved, in part the effect of the higher reaction temperature.

((Table 3))

Continuous flow hydrogenation experiments: The experiments were conducted using a continuous flow of a liquid furfural feedstock and a continuous H2 gas flow in H-Cube Pro (ThalesNanoTM) equipment. The same solvent (i-PrOH) was used as in the batch experiments.


14

Page 15 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


First, the continuous flow experiments required optimization of the reaction conditions. The effect of temperature, hydrodynamic pressure and flow rate on the conversion and product yields were studied over time (Figure 2). The conversion and product yields increased with higher temperature and pressure and with lower flow rate. At time zero (this is after 20-50 min. at operational pressure and temperature in function of the flow rate), all the conversions were close to 100 %, except at high flow rate (0.5 mL min.-1) or low hydrodynamic pressure (30 bars). When further samples were collected during 2 h on stream, all the conversions decreased gradually. A linear correlation fitted the decrease rate well, showing the catalyst deactivation over time. Only in optimized conditions of 150 °C, 50 bar and 0.2 mL min-1 (Figure 2f), the conversion remained constant (1 % less conversion after 2 h). It was also under these conditions that the MF yield remained the highest (20 %) and quasi constant. Compared to batch hydrogenation, higher amounts of ring hydrogenated products like MTHF and THFA were produced (10-15 %).

((Figure 2, 2 column))

Next, using optimized conditions, different solvents were tested (Figure 3). Whereas the hydrogenation in protic solvents (Figures 3a and 3b) gave full conversion without significant decrease during 2-3 h on stream, the conversions achieved with aprotic organic solvents were substantially lower (50-75 %) and decreased 3-8 % after 2 h. Cyclopentyl methyl ether (CPME) offered the most interesting result (Figures 3e and 3f) in terms of selectivity (mostly FA and THFA while < 3.0 % for MF and MTHF). Higher conversion and THFA yield was obtained


15


Page 16 of 49

using a 0.1 instead of 0.2 M furfural feedstock (Figure 3f). The highest MF yields were produced with protic solvents. But, the mass balance for protic solvents reached only 40-50 %, whereas they were quasi complete (90-100 %) when using aprotic solvents. Batch experiments in MeOH (Table 1, entry 8) gave much higher side product formation compared to i-PrOH (entry 5), leading to lower MF yield. Therefore i-PrOH was a more suitable protic solvent than MeOH. However, the side product formation in i-PrOH in continuous flow was substantial higher than in batch hydrogenation. When looking closer to the chromatogram of the corresponding hydrogenation product (Figure 4a), one can observe two additional peaks at higher retention times in the GC-FID analysis (Figure S6c). Side-product peaks were also present in the chromatograms of the hydrogenation products obtained in MeOH and in EtOH, but at lower retention times (Figure S6a and S6b). These peaks were completely absent in the chromatograms of the products collected using aprotic solvents (Figures S6d-6f). After evaporation under reduced pressure of the hydrogenation product obtained with continuous flow to remove the solvent (i-PrOH), no MTHF and only traces of MF were detected in the residue when analyzed by 1H-NMR (Figure 4b) and 13C-NMR (Figure 4c), as these are the only hydrogenation products with lower boiling point than i-PrOH. Analysis of both spectra revealed the presence of 2-(isopropoxy)-methylfuran (iPrOMF) in agreement with previous works which reported the presence of (hemi-)acetal and/or ether side products when using EtOH or i-PrOH,4044

but which did not provide experimental evidence explicitly. Next to iPrOMF, the

corresponding ring hydrogenated product 2-(isopropoxy)-methyltetrahydrofuran (iPrOMTHF) was also identified in the residue of the hydrogenation product. At first sight, the incorporation of solvent molecules in the hydrogenation product (alcoholysis) is an undesired side reaction and should therefore be suppressed. But, when Chernyak et al. (2016) considered the production of


16

Page 17 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


alkyl furfuryl ethers as a biofuel component precursor,46 they showed that the reaction proceeds spontaneously (but faster with acid catalyst), especially in excess of alcohol and absence of water. The reaction was also susceptible to the formation of some polymerized side products. Gilkey et al. recently demonstrated that when starting from furfuryl alcohol this etherification is reversible.44



Next, using CPME as the solvent (under optimized conditions), different catalysts were tested. The results confirmed that in contrast with protic solvents, mass balances were near to 100 % and kept constant during all experiments (Figure 5). With 10%Cu/AC no significant amounts of products were detected. Whereas with 3%Pd/AC and the 5%Ni/AC the main products were FA and THFA, the 3%Pt/AC yielded mostly FA and MF. The high selectivity to MF with 5%Ni/AC catalyst in batch conditions was not observed in continuous flow. At first sight the different reaction temperature (200 vs. 150 ºC for batch vs. continuous flow) could be the main cause for this change in selectivity, in agreement with the results from Panagiotopoulou and Vlachos (2014), who observed a change from 34 to 59 % selectivity to MF for a Ru/C catalyst when moving from 150 to 195 ºC (both in batch).42 Adsorption effects also may induce this change in selectivity. THFA was produced in all experiments conducted in continuous flow using


17


Page 18 of 49

3%Pd/AC or 5%Ni/AC, but it was only produced in trace amounts in batch conditions. With 3%Pt/AC as catalyst, THFA was quasi absent independent from being conducted in batch or continuous flow mode. The deactivation rates, this is the conversion decrease as seen from the slope of the linear regression fittings (Figure 5), were 12 and 8 % per h for the Pd/AC and Ni/AC catalysts, respectively. The 3%Pt/AC catalyst remained stable during 3 h on stream. When the experiments with 5%Ni/AC and 3%Pt/AC were repeated at lower WHSV, using 280 instead of 120 mg of catalyst at equal flow rate (Figures 5c and 5e), the deactivation rates were minimal (< 1 %), with conversions between 92-100 %. Initially (after 0.5 h on stream, t = 0 h), the selectivity to THFA reached 90 % with the 5%Ni/AC catalyst, but then dropped fast to only 22 % after 2 h. The 3%Pt/AC catalyst in contrast yielded MF with 83 % selectivity and still remained 71 % after 2 h. When fitting the conversion obtained in function of the metal-based weight hourly space velocity (WHSV, h-1), calculated as the furfural mass feed rate (g h-1) divided by the metal content in the catalytic bed (g), a linear correlation was observed (Figure 6), which is a useful guideline for adjusting the flow rate in function of the catalyst deactivation rate. More detailed and extended data sets are required for every metal type to establish more reliable empirical fittings. In a comparison with previously reported catalysts (Table 4), both 3%Pd/AC and 5% Ni/AC catalysts (entries 3 and 7) showed higher catalytic activity, although they were determined at higher WHSV values. In contrast to batch mode, TOF values do not vary over time except for catalyst deactivation effects but do vary in function of the WHSV, as illustrated in Figure S8. The highest activity was achieved with the 3%Pt/AC catalyst (entry 9). Moreover, it´s activity was maintained, at least during the first 2-3 h on stream, whereas the 5%Ni/AC and 3%Pd/C did not remain stable (at high WHSV). Interestingly, the 10%Ni/AC catalyst (entry 8)


18

Page 19 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


showed 4 times lower activity than 5%Ni/AC (entry 7) at similar WHSV, showing the importance of a good metal dispersion.


((Figure 6))

((Table 4))

Furfuryl alcohol hydrogenation: Currently, most of the work published on furfural hydrogenation reported catalysts highly selective to FA27,28,30,31,35,37,39,43 or to THFA,35,37,38 but very few report high selectivity to MF.35,42 Reduction of the carbonyl group is the first step in the furfural hydrogenation pathway whereas the production of MF involves an extra hydrogenolysis step and therefore typically requires higher temperature, which can increase the formation of FA derived side products and cause catalyst deactivation.43 Therefore, further experiments using FA instead of furfural feedstocks were also considered in the present work, under identical conditions (Table 5). In line with the results on furfural batch hydrogenation, low conversion was seen for the 10%Cu/C catalyst (entry 1) while 5%Ni/C and 3%Pt/C gave 97 % conversion (entry 2 and 3). The selectivity in contrast changed drastically, with significant higher yields of THFA and MTHF when using 5%Ni/C. Lower amounts of iPrOMF were detected, which means


19


Page 20 of 49

that the pathway to the ether side products is more efficient when starting from furfural, in line with the kinetic observations reported by Gilkey et al.44 The lack in the mass balance corresponds to other FA derived side products. In continuous flow regime (Figure 7), near to 100 % mass balance was observed during more than 1 h on stream (in contrast with the results when starting from furfural as shown in Figure 3), at 100 % conversion level and with ca. 50 % yield to MTHF for 3%Pd/C and near to 80 % yield to THFA for 5%Ni/C, respectively. The lower reaction temperature (150 °C for continuous flow vs. 200 °C for batch) avoided the production of other FA derived side products. The results also show that the catalysts are more reactive towards FA than furfural. Indeed, the first step in the pathway (reduction of the carbonyl group in furfural) is kinetically constrained limiting all further hydrogenation steps, and hydrogenolysis of FA via ring activation is more likely than direct hydrogenolysis.44 This suggests that ring coordination on the metallic site is more probable than coordination of the carbonyl group in furfural and the hydroxyl group in FA, but for the reaction willing to proceed the carbonyl group needs to reduced first to alcohol. These findings could explain why in the case of FA the product composition is similar when moving from batch to continuous flow hydrogenation, whereas it is not in the case of furfural. Kinetic and thermodynamic differences in adsorption and desorption between furfural and FA also may affect the final product composition. In continuous flow mode, the 3%Pd/AC catalyst produced only FA and THFA when starting from furfural (Figure 5a), whereas the main products were MTHF and THFA when starting from a FA feedstock (Figure 7a). As MTHF was produced from MF (tetrahydro-2furancarbaldehyde was not detected in the present conditions neither in previously reported studies), the results suggest that hydrogenolysis is more likely to occur starting from FA compared to starting from furfural.


20

Page 21 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


((Table 5))


Catalyst deactivation: Despite the promising results described above in terms of conversion and selectivity, significant effects were observed on the catalyst stability (deactivation). The effect of process conditions on the metal dispersion were studied by analyzing the change in mean crystal size, as seen from XRD analysis (Figure 8), using the Scherrer equation. Note that for the fresh Ni/AC, much higher crystal sizes were observed when increasing the Ni load from 5 to 10 wt% during the synthesis, which can explain the worse catalytic performance of the 10%Ni/AC catalyst due to poorer metal dispersion. The crystal sizes in the spent catalysts changed only in small extent. The fresh and spent 5%Ni/AC catalysts from batch and continuous flow were also analyzed with XPS and ICP-MS after acid digestion to study the effect of the process conditions on the extent of metal leaching from the carbon support (Table 6), both on the surface (XPS) and in the bulk catalyst (ICP-MS). The metal loads were slightly lower than their nominal loads used in their synthesis (5 wt%). Based on the ICP-MS analysis, the leaching effect observed in batch conditions (44%) was more pronounced compared to continuous flow (20 %). High stirring rate and high pressure in the batch experiments (200 ºC, 60 bars) may have promoted the metal leaching. Additional experiments at various stirring rates and pressures are required to confirm this. Anyhow, the results play in favor of continuous flow vs. batch hydrogenation. The leaching


21


Page 22 of 49

effect was more pronounced on the carbon surface (XPS) compared to the bulk of the catalyst (ICP-MS).

((Figure 8))

((Table 6))

Metal phases: In contrast to previous works reported on liquid phase furfural hydrogenation, the catalysts in the present work were not activated with H2 just before performing the catalytic hydrogenation experiment itself. Instead, the catalysts were stored under air atmosphere during prolonged time. In such a potential case or application, complete oxygen-free conditions for catalyst preparation, storing and transportation are not industrially feasible. The equilibration under air atmosphere promoted the formation of a NiO phase, at least on the catalyst surface layer as shown by XPS analysis. This might explain the high selectivity of MF with 5%Ni/AC, based on the promoting effect of a metal oxide phase co-existing with the metal phase.42,44 At one hand it was demonstrated that the co-existence of metal oxide with metal phases promotes the hydrogenolysis to MF by binding the cleaved OH groups and via ring activation instead of direct hydrogenolysis, but at the other hand it is known that Lewis acid sites promotes the formation of side products (Meerwein-Ponndorf-Verley reduction). These findings, however, were reported for the catalytic hydrogen transfer from isopropanol (i.e. in the absence of H2).42-44 When using no H2 (Table 1, entry 6), only 1 % MF was formed. At higher temperature (260 ºC)


22

Page 23 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the catalytic hydrogen transfer was much more evident, with 50 % MF yield (entry 7). In continuous flow at 150 ºC and in the absence of H2 (see Table S2), FA yields between 4 and 29 % were observed and in addition some alcoholysis side products, showing that MeerweinPonndorf-Verley reduction on Lewis acid sites, either metal oxides phases or acid groups of the AC support, was also a plausible reaction pathway in the present conditions. It is well known that nickel oxides form easily in Ni-based catalysts. In the present study, however, any XRD signal from a NiO phase was hardly observable (Figure 8), except for the 10%Ni/AC catalyst (Figure S4), which showed that oxide layers existed on the catalyst surface only. The results show that in the present conditions (batch 200 ºC and continuous flow 150 ºC) the partition of NiO as hydrogen transfer catalyst from i-PrOH to furfural for the hydrogenolysis to MF is rather limited, instead metal oxide sites may have acted as Lewis acid sites for the Meerwein-Ponndorf-Verley reduction to FA (and associated with this some alcoholysis side products). Advanced catalyst analysis methods such as in situ or operando XRD and XPS may reveal the evolution of the oxidation state and migration of the catalytic active zero valent metal nanoparticles in the catalyst bulk phase and surface during the course of the hydrogenation reaction. CONCLUSIONS High selectivity (78 %) to 2-methylfuran (MF) at high furfural conversion level (85 %) was demonstrated for a monometallic Ni catalyst supported on commercial micro/mesoporous activated carbon (5%Ni/AC) after 5 hours in batch hydrogenation experiments using 30 bars molecular hydrogen at 200 °C. Re-utilization of the spent catalyst recovered by simple filtration changed the selectivity to 57 % furfuryl alcohol (FA) at 67 % conversion. In continuous flow experiments at 150 °C in contrast, the Ni/AC catalyst was not that selective to MF (20 % at 100 % conversion), with FA and tetrahydrofurfuryl alcohol (THFA) being the other main products.


23


Page 24 of 49

The performance of the catalysts in continuous flow highly depended on the operational conditions, including flow rate, temperature, pressurization and especially the solvent type. When using protic solvents such as isopropanol, ethanol and methanol, important amounts of alkyl methylfuran ether and other side products were formed, as observed in the lacking mass balances and as experimentally evidenced by NMR analysis. Aprotic solvents did not lead to the formation of side products. Under optimized conditions, 3%Pt/AC catalyst showed superior performance with 64/82 % selectivity to MF and 53/92 % conversion at 64 and 27 g g-1 h-1 (metal-based weight hourly space velocity), respectively. Additional experiments with FA instead of furfural as the feedstock showed that the catalysts were more reactive towards FA than furfural and that ring hydrogenated product such as THFA and 2-methyltetrahydrofuran (MTHF) were produced in significant amounts. The product compositions were similar for batch and continuous flow, in strong contrast for the results obtained when starting from furfural. XPS and ICP-MS analysis showed that leaching was the main cause of catalyst deactivation when using Ni supported on high surface area carbon, particularly in batch conditions. ASSOCIATED CONTENT Supporting Information: schematic illustrations of the batch – and continuous flow experiments, nitrogen sorption isotherms and XRD spectra of monometallic catalysts, GC-FID chromatograms of hydrogenation experiments in various solvents and TOF-WHSV graph in continuous flow. ACKNOWLEDGEMENTS The authors gratefully acknowledge the financial support from the European COST Action FP1306 (´Valorization of Lignocellulosic Biomass Side Streams for Sustainable Production of Chemicals, Materials & Fuels Using Low Environmental Impact Technologies´) via Short Term


24

Page 25 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scientific Missions (STSMs) and the financial support from the corresponding partner host institutions: Université de Technologie de Compiègne (France), Universidad de Córdoba (Spain) and Aristotle University of Thessaloniki (Greece). Y. Wang thanks the China Scholarship Council for financial support. The staff of the Central Service for Research Support of the University of Córdoba is acknowledged for its technical support in XPS and ICP-MS measurements.

REFERENCES (1)

Mosier, N.; Wyman, C.; Dale, B.; Elander, R.; Lee, Y. Y.; Holtzapple, M.; Ladisch, M. Features of promising technologies for pretreatment of lignocellulosic biomass. Biores. Technol. 2005, 96, 673-686, DOI 10.1016/j.biortech.2004.06.025.

(2)

Yang, B.; Wyman, C. E. Pretreatment: the key to unlocking low-cost cellulosic ethanol. Biofuels Bioprod. Bioref. 2008, 2, 26-40, DOI 10.1002/bbb.49.

(3)

Brandt, A.; Gräsvik, J.; Hallett, J. P.; Welton, T. Deconstruction of lignocellulosic biomass with ionic liquids. Green Chem. 2013, 15, 550-583, DOI 10.1039/C2GC36364J.

(4)

Barakat, A.; Mayer-Laigle, C.; Solhy, A.; Arancon, R. A. D.; de Vries, H.; Luque, R. Mechanical pretreatments of lignocellulosic biomass: towards facile and environmentally sound technologies for biofuels production. RSC Adv. 2014, 4, 4810948127, DOI 10.1039/C4RA07568D.


25


(5)

Page 26 of 49

Kim, J. S., Lee, Y. Y.; Kim, T. H. A review on alkaline pretreatment technology for bioconversion of lignocellulosic biomass. Biores. Technol. 2016, 199, 42-48, DOI 10.1016/j.biortech.2015.08.085.

(6)

Zhang, K.; Pei, Z.; Wang, D. Organic solvent pretreatment of lignocellulosic biomass for biofuels and biochemicals: a review. Biores. Technol. 2016, 199, 21-33, DOI 10.1016/j.biortech.2015.08.102.

(7)

Steinbach, D.; Kruse, A.; Sauer, J. Pretreatment technologies of lignocellulosic biomass in water in view of furfural and 5-hydroxymethylfurfural production - a review. Biomass Convers. Biorefine. 2017, 7, 247-274, DOI 10.1007/s13399-017-0243-0.

(8)

Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.; Weckhuysen, B. M. The catalytic valorization of lignin for the production of renewable chemicals. Chem. Rev. 2010, 110, 3552-3599, DOI 10.1021/cr900354u.

(9)

Zhou, C.-H.; Xia, X.; Lin, C.-X.; Tong, D.-S.; Beltramini, J. Catalytic conversion of lignocellulosic biomass to fine chemicals and fuels. Chem. Soc. Rev. 2011, 40, 55885617, DOI 10.1039/C1CS15124J.

(10)

Li, C.; Zhao, X.; Wang, A.; Huber, G. W.; Zhang, T. Catalytic transformation of lignin for the production of chemicals and fuels. Chem. Rev. 2015, 115, 11559-11624, DOI 10.1021/acs.chemrev.5b00155.

(11)

Ferrini, P.; Rinaldi, R. Catalytic biorefining of plant biomass to non‐pyrolytic lignin bio‐oil and carbohydrates through hydrogen transfer reactions. Angew. Chem. Int. Ed. 2014, 53, 8634-8639, DOI 10.1002/anie.201403747.


26

Page 27 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(12)

Luterbacher, J. S.; Azarpira, A.; Motagamwala, A. H.; Lu, F.; Ralph, J.; Dumesic, J. A. Lignin monomer production integrated into the γ-valerolactone sugar platform. Energy Environ. Sci. 2015, 8, 2657-2663, DOI 10.1039/C5EE01322D.

(13)

Da Costa Sousa, L.; Foston, M.; Bokade, V.; Azarpira, A.; Lu, F.; Ragauskas, A. J.; Ralph, J.; Dale, B.; Balan, V. Isolation and characterization of new lignin streams derived from extractive-ammonia (EA) pretreatment. Green Chem. 2016, 18, 42054215, DOI 10.1039/C6GC00298F.

(14)

Carrasquillo-Flores, R.; Käldström, M.; Schüth, F.; Dumesic, J. A.; Rinaldi, R. Mechanocatalytic depolymerization of dry (ligno) cellulose as an entry process for highyield production of furfurals. ACS Catal. 2013, 3, 993-997, DOI 10.1021/cs4001333.

(15)

Lange, J.-P.; van der Heide, E.; van Buijtenen, J.; Price, R. Furfural - a promising platform for lignocellulosic biofuels. ChemSusChem 2012, 5, 150-166, DOI 10.1002/cssc.201100648.

(16)

Pileidis, F. D.; Titirici, M.-M. Levulinic acid biorefineries: new challenges for efficient utilization of biomass. ChemSusChem 2016, 9, 562-582, DOI 10.1002/cssc.201501405.

(17)

Mariscal, R.; Maireles-Torres, P.; Ojeda, M.; Sádabaa, I.; López Granados, M. Furfural: a renewable and versatile platform molecule for the synthesis of chemicals and fuels. Energy Environ. Sci. 2016, 9, 1144-1189, DOI 10.1039/C5EE02666K.

(18)

Yi, G.; Teong, S. P.; Zhang, Y. Base-free conversion of 5-hydroxymethylfurfural to 2, 5furandicarboxylic acid over a Ru/C catalyst. Green Chem. 2016, 18, 979-983, DOI 10.1039/C5GC01584G.


27


(19)

Page 28 of 49

Chatterjee, M.; Ishizakaa, T.; Kawanami, H. Reductive amination of furfural to furfurylamine using aqueous ammonia solution and molecular hydrogen: an environmentally friendly approach. Green Chem. 2016, 18, 487-496, DOI 10.1039/C5GC01352F.

(20)

Teixeira, I. F.; Lo, B. T. W.; Kostetskyy, P.; Stamatakis, M.; Ye, L.; Tang, C. C.; Mpourmpakis, G.; Tsang, S. C. E. From biomass‐derived furans to aromatics with ethanol over zeolite. Angew. Chem. Int. Ed. 2016, 55, 13061-13066, DOI 10.1002/anie.201604108.

(21)

Szczurek, A.; Fierro, V.; Thébault, M.; Pizzi, A.; Celzard, A. Structure and properties of poly (furfuryl alcohol)-tannin polyHIPEs. Eur. Polym. J. 2016, 78, 195-212, DOI 10.1016/j.eurpolymj.2016.03.037.

(22)

Mangin, F.; Prinsen, P.; Yepez, A.; Gilani, M. R. H.; Xu, G.; Len, C.; Luque, R. Microwave assisted benzyl alcohol oxidation using iron particles on furfuryl alcohol derived supports. Catal. Commun. 2018, 104, 67–70, DOI 10.1016/j.catcom.2017.10.003.

(23)

Ju, M.; Zeng, C.; Wang, C.; Zhang, L. Preparation of ultrafine carbon spheres by controlled polymerization of furfuryl alcohol in microdroplets. Ind. Eng. Chem. Res. 2014, 53, 3084-3090, DOI 10.1021/ie4029939.

(24)

Wei, H.; Feng, D.; Shu, G.; Pan, M.; Guo, Y.; Gao, D.; Li, W. Experimental investigation on the combustion and emissions characteristics of 2-methylfuran gasoline blend fuel in


28

Page 29 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


spark-ignition engine. Appl. Energ. 2014, 132, 317-324, DOI 10.1016/j.apenergy.2014.07.009. (25)

Li, G.; Li, N.; Wang, X.; Sheng, X.; Li, S.; Wang, A.; Cong, Y., Wang, X.; Zhang, T. Synthesis of diesel or jet fuel range cycloalkanes with 2-methylfuran and cyclopentanone from lignocellulose. Energy Fuels 2014, 28, 5112-5118, DOI 10.1021/ef500676z.

(26)

Srivastava, S.; Jadeja, G. C.; Parikh, J. Synergism studies on alumina-supported coppernickel catalysts towards furfural and 5-hydroxymethylfurfural hydrogenation. J. Mol. Catal. A: Chem. 2017, 426, 244-256, DOI 10.1016/j.molcata.2016.11.023.

(27)

Sharma, R. V.; Das, U.; Sammynaiken, R.; Dalai, A. K. Liquid phase chemo-selective catalytic hydrogenation of furfural to furfuryl alcohol. Appl. Catal., A 2013, 454, 127136, DOI 10.1016/j.apcata.2012.12.010.

(28)

Villaverde, M. M.; Bertero, N. M.; Garetto, T. F.; Marchi, A. J. Selective liquid-phase hydrogenation of furfural to furfuryl alcohol over Cu-based catalysts. Catal. Today 2013, 213, 87-92. DOI 10.1016/j.cattod.2013.02.031.

(29)

Garcia-Olmo, A. J.; Yepez, A.; Balu, A. M.; Romero, A. A.; Li, Y.; Luque, R. Insights into the activity, selectivity and stability of heterogeneous catalysts in the continuous flow hydroconversion of furfural. Catal. Sci. Technol. 2016, 6, 4705-4711, DOI 10.1039/C6CY00249H.

(30)

Biradar, N. S.; Hengne, A. A.; Biradar, S. N.; Swami, R.; Rode, C. V. Tailoring the product distribution with batch and continuous process options in catalytic


29


Page 30 of 49

hydrogenation of furfural. Org. Process Res. Dev. 2014, 18, 1434-1442, DOI 10.1021/op500196x. (31)

Fulajtárova, K., Sotáka, T.; Hronec, M.; Vávra, I.; Dobročka, E.; Omastová, M. Aqueous phase hydrogenation of furfural to furfuryl alcohol over Pd–Cu catalysts. Appl. Catal., A 2015, 502, 78-85, DOI 10.1016/j.apcata.2015.05.031.

(32)

Stevens, J. G.; Bourne, R. A.; Twigg, M. V.; Poliakoff, M. Real‐time product switching using a twin catalyst system for the hydrogenation of furfural in supercritical CO2. Angew. Chem. 2010, 122, 9040-9043, DOI 10.1002/ange.201005092.

(33)

Aldosari, O. F.; Iqbal, S.; Miedziak, P. J.; Brett, G. L.; Jones, D. R.; Liu, X.; Edwards, J. K.; Morgan, D. J.; Knight, D. K.; Hutchings, G. Pd-Ru/TiO2 catalyst - an active and selective catalyst for furfural hydrogenation. J. Catal. Sci. Technol. 2016, 6, 234-242, DOI 10.1039/C5CY01650A.

(34)

Garcia-Olmo, A. J.; Yepez, A.; Balu, A. M., Prinsen, P.; Garcia, A.; Mazière, A.; Len, C.; Luque, R. Activity of continuous flow synthesized Pd-based nanocatalysts in the flow hydroconversion of furfural. Tetrahedr. 2017, 73, 5599-5604, DOI 10.1016/j.tet.2017.02.056.

(35)

Chen, B.; Li, F.; Huang, Z.; Yuan, G. Tuning catalytic selectivity of liquid-phase hydrogenation of furfural via synergistic effects of supported bimetallic catalysts. Appl. Catal., A 2015, 500, 23-29, DOI 10.1016/j.apcata.2015.05.006.


30

Page 31 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(36)

Xinghua, Z.; Tiejun, W.; Longlong, M.; Chuangzhi, W. Aqueous-phase catalytic process for production of pentane from furfural over nickel-based catalysts. Fuel 2010, 89, 2697-2702, DOI 10.1016/j.fuel.2010.05.043.

(37)

Gong, W.; Chen, C.; Zhang, H.; Zhang, Y.; Zhang, Y.; Wang, G.; Zhao, H. Highly selective liquid-phase hydrogenation of furfural over N-doped carbon supported metallic nickel catalyst under mild conditions. Mol. Catal. 2017, 429, 51-59, DOI 10.1016/j.molcata.2016.12.004.

(38)

Chen, X.; Sun, W.; Xiao, N.; Yan, Y.; Liu, S. Experimental study for liquid phase selective hydrogenation of furfuryl alcohol to tetrahydrofurfuryl alcohol on supported Ni catalysts. Chem. Eng. J. 2007, 126, 5-11, DOI 10.1016/j.cej.2006.08.019.

(39)

Taylor, M. J.; Durndell, L. J.; Isaacs, M. A.; Parlett, C. M. A.; Wilson, K.; Lee, A. F.; Kyriakou, G. Highly selective hydrogenation of furfural over supported Pt nanoparticles under mild conditions. Appl. Catal., B 2016, 180, 580-585, DOI 10.1016/j.apcatb.2015.07.006.

(40)

Chen, X.; Zhang, L.; Zhang, B.; Guo, X.; Mu, X. Highly selective hydrogenation of furfural to furfuryl alcohol over Pt nanoparticles supported on g-C3N4 nanosheets catalysts in water. Sci. Rep. 2016, 6, 28558-28571, DOI 10.1038/srep28558.

(41)

Reyes, P.; Salinas, D.; Campos, C.; Oportus, M.; Murcia, J.; Rojas, H.; Borda, G.; García Fierro, J. L. Selective hydrogenation of furfural on Ir/TiO2 catalysts. Quím. Nova 2010, 33(4), 777-780, DOI 10.1590/S0100-40422010000400002.


31


(42)

Page 32 of 49

Panagiotopoulou, P.; Vlachos, D. G. Liquid phase catalytic transfer hydrogenation of furfural over a Ru/C catalyst. Appl. Catal., A 2014, 480, 17-24, DOI 10.1016/j.apcata.2014.04.018.

(43)

Koehle, M.; Lobo, R. F. Lewis acidic zeolite Beta catalyst for the Meerwein–Ponndorf– Verley reduction of furfural. Catal. Sci. Technol. 2016, 6, 3018-3026, DOI 10.1039/C5CY01501D.

(44)

Gilkey, M. J.; Panagiotopoulou, P.; Mironenko, A. V.; Jenness, G. R.; Vlachos, D. G.; Xu, B. Mechanistic insights into metal lewis acid-mediated catalytic transfer hydrogenation of furfural to 2-methylfuran. ACS Catal. 2015, 5, 3988−3994, DOI 10.1021/acscatal.5b00586.

(45)

Grob, R.L.; Barry, E. F. Modern Practice of Gas Chromatography, John Wiley & Sons, Inc., 2004.

(46)

Chernyak, M. Y.; Tarabanko, V. E.; Morosov, A. А.; Kondrasenko, A. A. Preparative synthesis of furfural diethyl acetal through the direct interaction of the alcohol and aldehyde. Journal of Siberian Federal University, Chemistry 2 2016, 9, 146-151, DOI 10.17516/1998-2836-2016-9-2-146-151.


32

Page 33 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Porosity of the monometallic catalysts supported on activated carbon (AC).

Catalyst

Total pore Micropore area Meso/macropore & Total SSA volume g-1) / external area (m2 g-1) (m2 (m2 g-1)a -1 b -1 c (cc g ) volume (cc g ) / volume (cc g-1)d

Activated carbon (AC) 1280 0.95 840 / 0.34 440 / 0.60 10%Cu/AC 1170 0.83 770 / 0.31 400 / 0.52 3%Pd/AC 1340 0.95 890 / 0.36 450 / 0.59 3%Pt/AC 1180 0.85 760 / 0.31 420 / 0.54 5%Ni/AC 1250 0.88 830 / 0.34 420 / 0.54 10%Ni/AC 1250 0.90 810 / 0.33 440 / 0.57 a SSA: specific surface area from N2 sorption at -196 °C (multi-point BET method), b Total pore volume at P/Po = 0.99, c t-plot method, d Meso/macropore & external area = Total SSA Micropore area; Meso/macropore volume = Total pore volume - Micropore volume (t-plot)


33


Page 34 of 49

Table 2. Conversion and product yields in liquid phase furfural batch hydrogenation experiments (FA: furfuryl alcohol, MF: 2-methylfuran, MTHF: 2-methyltetrahydrofuran and iPrOMF: 2(isopropoxy)-methylfuran).a

FA Entry

Catalyst

THFA

Yield (%) MF MTHF

Conversion (%)

iPrOMF

Mass balance (%)

1 2 6 0 0 0 0 104 2 10%Cu/AC 24 22 0 2 1 1 103 3 3%Pd/AC 47 21 1 5 2 5 87 4 3%Pt/AC 93 47 1 24 3 5 87 5 5%Ni/AC 85 6 1 66 2 3 93 b 6 5%Ni/AC 10 10 1 1 0 0 102 7 5%Ni/ACc 95 20 1 50 1 1 78 d 8 5%Ni/AC 87 13 1 9 2 0 38e 9 5% Ni/ACf 67 38 1 17 1 13 103 a b c 200 °C, 5 h, 0.35 M furfural in 60 mL isopropanol, 30 bars H2, 0 bar H2/200 ºC, 0 bar H2/260 ºC, d In methanol, e Unknown compound eluting at 3.8 min in GC analysis (Figure S6a) not included (48 % of total peak area), f Spent catalyst recovered after the experiment in entry 5.


34

Page 35 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 3. Comparison of the catalytic performance of Cu, Pd, Ni and Pt-based catalysts for the liquid phase hydrogenation of furfural in batch mode.

Entry

Catalyst

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Cu:Zn:Cr:Zr (3:2:1:3) CuMgAl 10%Cu/AC 3%Pd/C 5%Pd/C 5%Pd-5% Cu/C 5%Pd-5%Cu/MgO 5%Pd/TiO2 3%Pd/AC 3%Pt/C 1.9%Pt/γ-Al2O3 0.5%Pt/TECN 5.0%Pt/TECN 3%Pt/AC 5%Ni-1.8%Pd/TiO2-ZrO2 10%Ni/SiO2-Al2O3 16%Ni/NAC 16%Ni/AC 10%Ni/Ba-Al2O3 10%Ni/Ba-Al2O3 5%Ni/AC

a

T (ºC) 170 110 200 220 80 110 80 20 200 220 50 100 100 200 130 140 80 80 140 140 200

H2 (bars) 20 10 30 35 6 6 6 3 30 35 1 10 10 30 50 30 40 40 40 40 30

Time (h) 3 4 5 5 1.5 1 1.8 2 5 5 7 5 5 5 5 1 3 3 0.1 4 5

Solvent i-PrOH i-PrOH i-PrOH i-PrOH H 2O H 2O H 2O n-octane i-PrOH i-PrOH MeOH H 2O H 2O i-PrOH EtOH Water i-PrOH i-PrOH Water Water CPME

Selectivity (%) FA MF THFA 96 0 0 100 0 0 92 8 0 16 14 32 86 91 98 29 36 5 33 5 52 74 11 26 99 0 0 99 0 0 99 0 0 51 26 1 40 0 40 0 0 100 22 0 78 0 0 99 0 0 99 7 78 1

TOF (h-1) 2 7 2 69 82 26 74 50 25 90 10 239 74 90 1 2a 1 1 68 4 18

Ref. 27 28

This work 30 31 31 31 33

This work 30 39 40 40

This work 35 36 37 37 47 47

This work

87 % selectivity to C5 alkanes/alcohols, TOF calculated based on assumption of 1 h reaction


35


Page 36 of 49

Table 4. Comparison of the initial catalytic performance (determined after 0.2-0.8 h on stream) of Pd, Ni and Pt-based catalysts for the liquid phase hydrogenation of furfural in continuous flow mode.

5%Pd/AC 1.9%Pd/Al-SBAred

T (ºC) 90 90

H2 (bars)a 50 50

WHSV (h-1) 35 36

3

3%Pd/AC

150

50

4 5

63%Ni/SiAl 92%RaneyNi

90 90

6

5%Ni/AC

7

Entry

Catalyst

1 2

a

Solvent

Selectivity (%) MF THFA 75 0 0 99

C5b 25 0

TOF (h-1) 38 37

EtOAc EtOAc

FA 0 0

64

CPME

25

5

56

0

55

50 50

3 2

EtOAc EtOAc

19 87

9 17

0 0

78 1

1 1

150

50

38

CPME

50

4

42

0

21

5%Ni/AC

150

50

17

CPME

10

0

90

0

12

8

10%Ni/AC

150

50

19

CPME

94

0

1

0

3

9

3%Pt/AC

150

50

64

CPME

29

64

1

0

70

10

3%Pt/AC

150

50

27

CPME

12

82

1

0

51

Ref. 29 35

This work 29 29

This work This work This work This work This work

b

hydrodynamic pressure, C5 alkanes/alcohols

Table 5. Conversion and product yields obtained in furfuryl alcohol batch hydrogenation experiments.a

Entry

Catalyst

1 2 3

10%Cu/AC 3%Pt/AC 5%Ni/AC

Conversion (%) 15 97 97

Yield (%) THFA MF MTHF 1 5 2 18 50 11 48 10 22


iPrOMF 2 1 1

Mass balance (%) 67 83 84

36

Page 37 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a

200 °C, 5 h, 0.35 M furfural in 60 mL i-PrOH, 30 bars H2

Table 6. Comparison of C, O and Ni content as determined by XPS and ICP-MS analysis in fresh and spent 5%Ni/AC catalysts. XPS (wt%)a Entry C O Ni(0) 1 Fresh 86.8 12.3 0.4 2 Spent (batch, 200 °C) 79.1 20.4 0.2 3 Spent (continuous flow, 150 °C) 86.0 13.4 0.2 a based on C (1s), O (1s), Ni(0) (2p3/2) and NiO (2p3/2) signals 5%Ni/AC


NiO 0.5 0.3 0.4

ICP-MS (wt%) Ni(0) 4.1 2.3 3.3

37


Page 38 of 49

Figure 1. Furfural conversion and MF yield vs. H2 consumption observed after 5 h at 200 °C (as pressure drop with respect to initial pressure of 30 bars H2).

Figure 2. Conversion, product yield and mass balance in continuous flow hydrogenation of 0.2 M furfural in i-PrOH with 5%Ni/AC at increasing (a-b) temperature (50 bars, 0.3 mL min-1), (ce) pressurization (150 °C, 0.3 mL min-1) and (f-h) flow rate (150 °C, 50 bars).

Figure 3. Conversion, product yield and mass balance in the continuous flow hydrogenation of 0.2 M furfural with 5%Ni/AC operating at 0.2 mL min-1, 150 °C and 50 bars in (a) isopropanol, (b) ethanol, (c) ethyl acetate, (d) methyl isobutyl ketone, (e) CPME (0.2 M furfural) and (f) CPME (0.1 M furfural).

Figure 4. (a) GC-FID chromatogram of the hydrogenation product obtained after 2 h at 0.2 mL min-1, 150 °C and 50 bars of 0.2 M furfural in i-PrOH, (b) corresponding 1H-NMR spectrum and (c) 13C-NMR spectrum of the hydrogenation product residue after removal of MF, MTHF and iPrOH by rotavaporatization at reduced pressure.

Figure 5. Conversion, product yield and mass balance in the continuous flow hydrogenation of 0.2 M furfural in CPME operating at 0.2 mL min-1, 150 °C and 50 bars using 30 mm packed


38

Page 39 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


beds containing 120 mg of (a) 3%Pd/AC, (b) 3%Pt/AC and (d) 5%Ni/AC, and using 70 mm packed beds containing 280 mg of (c) 3%Pt/C and (e) 5%Ni/AC.

Figure 6. Linear correlation between furfural conversion and metal based WHSV for the continuous flow hydrogenation data obtained with the 3%Pd/AC, 3%Pt/AC and 5%Ni/AC catalysts from Figure 5.

Figure 7. Conversion, product yield and mass balance over 1 h continuous flow hydrogenation at 0.2 mL min-1 of 0.2 M furfuryl alcohol in i-PrOH at 150 ‐ and 50 bars using 120 mg of (a) 3%Pd/AC and (b) 5%Ni/AC.

Figure 8. XRD spectra of fresh and spent catalysts and corresponding crystal sizes based on Scherrer´s equation.

For Table of Contents Use Only


39


Page 40 of 49

Various monometallic catalysts supported on micro/mesoporous carbon showed highly different catalytic performance in the liquid phase hydrogenation of furfural, especially in terms of selectivity, depending on the metal type but also on operational parameters including temperature, solvent type and hydrogenation mode (batch or continuous flow).


40

Page 41 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 84x66mm (300 x 300 DPI)



Figure 2 174x133mm (300 x 300 DPI)


Page 42 of 49

Page 43 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3 173x141mm (300 x 300 DPI)



Figure 4 173x228mm (300 x 300 DPI)


Page 44 of 49

Page 45 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5 148x152mm (300 x 300 DPI)



Figure 6 81x64mm (300 x 300 DPI)


Page 46 of 49

Page 47 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7 173x77mm (300 x 300 DPI)



Figure 8 82x121mm (300 x 300 DPI)


Page 48 of 49

Page 49 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphical abstract 84x47mm (300 x 300 DPI)


Comparative Study of Supported Monometallic Catalysts in the Liquid

Recommend Documents