- Email: [email protected]
Hydrothermal deoxygenation of triglycerides over carbon-supported bimetallic PtRe catalysts without an external hydrogen source
Molecular Catalysis 474 (2019) 110419
Contents lists available at ScienceDirect
Molecular Catalysis journal homepage: www.elsevier.com/locate/mcat
Hydrothermal deoxygenation of triglycerides over carbon-supported bimetallic PtRe catalysts without an external hydrogen source Mingyu Jin, Minkee Choi
T
⁎
Department of Chemical & Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea
ARTICLE INFO
ABSTRACT
Keywords: Hydrothermal deoxygenation Triglyceride PtRe Bimetal Carbon supports
Hydrothermal deoxygenation is a promising route for producing oxygen-free hydrocarbons from triglycerides even without an external H2 source. In this study, PtRe catalysts supported on various carbons were investigated for this reaction. The PtRe catalysts showed enhanced rates of glycerol reforming for in situ H2 production and higher deoxygenation activities, as compared to Pt catalysts. Consequently, the PtRe catalysts showed superior hydrothermal deoxygenation performances. Carbon supports with low oxygen content (e.g., CNT) were beneficial for supporting highly dispersed and well alloyed PtRe particles, which were prerequisites for high catalytic activity. Oxygen-rich carbon surfaces caused the segregation of Re, causing low catalytic activity. PtRe/CNT showed the best performance, yielding 72 wt% n-paraffin (theoretical yield: 79 wt%) from palm oil. The catalyst could be recycled for 5 times without significant loss of activity.
1. Introduction Triglycerides (e.g., microalgal oil, vegetable oil, and waste cooking oil) are a promising feedstock for producing bio-derived fuels. They are esters combining one glycerol and three long-chain fatty acids (generally C14−C22) which contain relatively low oxygen contents and high energy densities compared to other biomass feedstock [1–3]. The synthesis of biodiesel (e.g., fatty acid methyl ester) from triglycerides via transesterification is a major commercially applied process [3,4]. However, the presence of oxygen in biodiesel diminishes its energy density, storage stability, and cold flow properties [5–7]. To overcome the limitations of biodiesel, catalytic upgrading of triglycerides to oxygen-free fuels through hydroprocessing has been developed [8–17]. In these processes, triglycerides are hydrotreated over MoS2-based catalysts [9,10] or supported metal catalysts [11,9–17] under highpressure H2. Under such conditions, triglycerides containing unsaturated fatty acid units are first hydrogenated and cleaved into three saturated fatty acids by hydrogenolysis (Scheme 1a). The resulting fatty acids are then deoxygenated via decarboxylation/decarbonylation (DCO) and hydrodeoxygenation (HDO) pathways. DCO produces nparaffins with one less carbon atom than the initial fatty acids, whereas HDO produces n-paraffins having the same carbon number. With the production of long-chain paraffins, light gaseous hydrocarbons (C1–C3) are also coproduced due to the glycerol hydrogenolysis and the CO
⁎
methanation. These side reactions all consume significant amounts of H2. Although the produced liquid paraffins have significantly improved fuel quality, the large consumption of H2 in these processes is disadvantageous for overall economics and sustainability. It should be noted that H2 is currently primarily produced from natural gas. Hydrothermal deoxygenation of triglycerides in subcritical water (523–653 K) can provide an alternative pathway for producing oxygenfree hydrocarbons [18–23]. Savage and colleagues first reported that the high-temperature hydrothermal treatment of model fatty acids over Pt/C and Pd/C catalysts could produce saturated paraffins even in the absence of an external H2 source [18,19]. The result indicated that hydrogen could be supplied in situ during the reaction. Considering that a significant amount of heavy oligomers were also formed, hydrogen transfer might be involved under those reaction conditions. More recently, other researchers investigated the hydrothermal deoxygenation of natural triglycerides or the mixtures of fatty acids and glycerol [20–23]. The main difference between these studies and the earlier works by Savage is that glycerol in the reactants can be readily transformed to H2 in situ via aqueous phase reforming (APR) (Scheme 1b). In principle, one glycerol can generate seven H2 molecules at maximum through APR [24–32], which is sufficient for the saturation and the deoxygenation of fatty acids. Under this novel reaction scheme, glycerol acts as a source for H2 production rather than a consumer of H2 (the case of conventional hydroprocessing). This can enable essentially net-
Corresponding author. E-mail address: [email protected] (M. Choi).
https://doi.org/10.1016/j.mcat.2019.110419 Received 19 March 2019; Received in revised form 19 May 2019; Accepted 20 May 2019 Available online 30 May 2019 2468-8231/ © 2019 Elsevier B.V. All rights reserved.
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
carried out using a Sigma Probe instrument (Thermo Fischer Scientific VG) equipped with monochromatic Al K radiation (1486.3 eV; anode operating at 15 kV and 7 mA). The pass energy was fixed at 50 eV for all survey scans. Samples were mounted using double-sided adhesive carbon tape, and binding energies were calibrated to the C (1s) binding energy (284.5 eV). N2 adsorption-desorption isotherms of the catalysts were measured using a Tristar volumetric analyzer (Micromeritics) at 77 K. Prior to the measurements, all samples were vacuum-degassed at 473 K for 6 h. H2 and CO chemisorptions were measured using an ASAP 2020c (Micromeritics). Before the isotherm measurement, the Pt and PtRe catalysts were vacuum-degassed at 473 K for 6 h, reduced under H2 at 573 K for 1 h, followed by evacuation at 573 K for 1 h. In the case of Re catalysts, the pre-reduction and evacuation were carried out at 723 K. Adsorption isotherms were measured at 323 K and the chemisorption amount was determined by extrapolating the linear regime (10–30 kPa) of the isotherm to zero pressure. Transmission electron microscopy (TEM) images were taken using a Talos F200X microscope operating at 200 kV. Samples were mounted on a carbon-coated copper grid (300 mesh) using ethanol dispersion. To determine the metal size distributions using TEM, at least 200 metal particles were counted. The surfacearea-weighted mean particle diameter was calculated using the following equation [39]:
Scheme 1. Production of oxygen-free paraffin fuels from triglycerides via (a) hydroprocessing and (b) hydrothermal deoxygenation.
zero H2 consumption in the production of oxygen-free paraffins from triglycerides [21,22]. Pt and Pd catalysts supported on activated carbons have been reported to show activities in hydrothermal deoxygenation [18,19,21]. The choice of noble metal catalysts and carbon supports can be rationalized by their stability even under harsh hydrothermal conditions (563–653 K) that can severely degrade other metals and metal oxides via sintering or phase transition [33,34]. Recent studies by Vardon et al. demonstrated that bimetallic PtRe supported on activated carbons can show superior catalytic activity as compared to a monometallic Pt catalyst [21]. However, these studies mainly focused on the effects of reaction conditions (e.g., temperature and reactant compositions) rather than the effects of catalyst structures. Previous studies showed that carbon-supported catalysts can have a wide range of catalytic properties depending on the carbon structures [35–38]. In particular, the oxygen functional groups on carbon surfaces can significantly affect the alloying state of bimetallic catalysts and thus their catalytic properties [38]. Therefore, in the present work, we carefully investigated the catalytic properties of bimetallic PtRe particles supported on various carbon supports for the hydrothermal deoxygenation of triglycerides (e.g., palm oil).
dTEM (nm) = Σ nidi3 / Σ nidi2
(1)
TPR profiles were collected using a BELCAT II (MicrotracBEL Corp.) equipped with a thermal conductivity detector (TCD). Typically, 50 mg of the catalyst was loaded into a quartz sample tube and degassed in Ar at 473 K for 2 h. TPR profiles were collected while increasing the temperature from 323 K to 1073 K (ramp: 10 K min−1) under 5% H2/Ar flow (30 ml min−1). To remove CO, CO2, and H2O generated by the thermal decomposition of the oxygen functional groups on the carbon surface, a trap containing CuCl-modified activated carbon and 13X molecular sieves was installed before TCD with its temperature set to 273 K using an ice bath [37,40]. DRIFT spectroscopy was carried out using a Bruker Vertex 70 spectrometer equipped with a Pike Diffuse IR cell attachment. Spectra were collected using an MCT detector with a 2 cm−1 resolution. The sample was pre-reduced under H2 at 573 K for 1 h. After cooling to 293 K, background spectra were collected. CO adsorption was carried out by purging the IR cell with 1% CO/He at room temperature for 1 h. After subsequent He purging at 293, 323, 373, 473, and 573 K for 1 h, the spectra were collected. XAS on the Pt and Re L3 edges was carried out in transmission mode at the Pohang Accelerator Laboratory (7D-XAFS beamline). For the measurements, 0.3 g samples were pelletized (13 mm in diameter) and placed in a home-made in situ cell equipped with 0.05 mm-thick aluminum windows. Energy calibration was carried out using Pt and Re foils. Prior to the measurements, catalyst samples were pre-reduced under H2 at 573 K for 1 h. Pt foil, PtO2, Re powder, ReO2, and NH4ReO4 were used as Pt and Re reference samples. The Athena and Artemis programs from the IFEFFIT data package were used for processing the XAS data. Normalization and background removal were carried out using Athena. Normalized EXAFS data were k3-weighted and Fouriertransformed into R space for obtaining the radial distribution functions. The theoretical EXAFS signal was constructed using the FEFF6 code and fitted. For the EXAFS fitting of Re, we used two different Re-O scattering paths at 1.73 and 2.08 Å, as proposed by Bare et al [41].
2. Experimental 2.1. Catalyst preparation CNT (Ctube) and AC (Aldrich) were treated in a 6 M HCl solution at room temperature for 2 h to remove possible metal impurities. To introduce additional oxygen functional groups, 1 g of CNT and AC were oxidized in 100 ml of 3 M HNO3 solution for 2 h at 403 K and 323 K, respectively. After filtration, the carbons were thoroughly washed with deionized water and dried at 373 K. The oxidized CNT and AC samples were denoted as oCNT and oAC. Monometallic Pt (2.5 wt%) and Re (2.5 wt%) catalysts were supported on the carbon supports by incipient wetness impregnation of aqueous solutions of chloroplatinic acid hexahydrate (H2PtCl6·6H2O, Kojima Chemical) and perrhenic acid (HReO4, Aldrich). Bimetallic PtRe catalyst (2.5 wt% Pt/1.0 wt% Re) was supported by co-impregnation of the Pt and Re precursors on the carbon supports. After impregnation, the samples were dried at 373 K overnight, followed by reduction under H2 flow. Pt and PtRe catalysts were reduced at 573 K (ramp: 2 K min−1) for 2 h, while monometallic Re catalysts were reduced at 723 K.
2.3. Analysis of fatty acid composition in palm oil
2.2. Characterization
To analyze the fatty composition of palm oil (Aldrich), transesterification with methanol was carried out. For transesterification, 1 g of palm oil was reacted with 0.25 g of methanol (methanol/triglyceride ratio ˜6) at 338 K for 1 h in the presence of 1 wt% NaOH as a catalyst
Metal concentrations were analyzed using inductively coupled plasma optical emission spectrometer (ICP-OES 720, Agilent). For quantifying the amount of oxygen on carbon surfaces, O 1s XPS was 2
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
[42]. The upper phase (fatty acid methyl esters) were dissolved in chloroform (dodecane was used as an external standard) and then analyzed using a gas chromatograph (GC) equipped with a flame ionization detector (FID) and a HP-5 column (Agilent, 30 m × 0.32 mm).
Table 1 Properties of Pt, Re, and PtRe catalysts supported on CNT.
2.4. Catalytic measurements Hydrothermal deoxygenation of the palm oil was carried out at various temperatures (448, 473, 498, 523, and 558 K) in an autoclave with a 200 ml volume. For a typical reaction, 20 g of palm oil (Aldrich), 80 g of deionized water, and 0.92 g of catalysts, pre-reduced at 573 K, were loaded. After filling with 10 bar N2, the autoclave was heated to the reaction temperature under static conditions. After temperature stabilization, the reaction was initiated by starting magnetic stirring (500 rpm) and continued for 24 h. The deoxygenation of stearic acid in H2 was carried out in a similar manner. For the reaction, 20 g of stearic acid (Aldrich), 80 g of deionized water, and 0.92 g of catalysts were loaded in the autoclave. Instead of N2, the reactor headspace was filled with 10 bar H2 before the reaction. After the reaction, gas compositions in the headspace were analyzed using an online GC equipped with an FID and a GS-GasPro column (Agilent, 30 m × 0.32 mm), as well as a TCD and a Carboxen 1000 column (SUPELCO). Liquid products were mixed with 300 ml of dichloromethane (DCM), followed by the separation of catalysts by filtration. From the filtrate, the organic phase (lower phase) was collected after phase separation. After removing the DCM using a rotary evaporator, the total yield of products in the oil phase was determined by weighing. The detailed compositions of the products were analyzed using a GC equipped with an FID and a DB-FFAP column (Agilent, 30 m × 0.25 mm), after dissolution in chloroform (n-dodecane was used as an external standard). The concentration of products in an aqueous phase was analyzed using high-performance liquid chromatography (HPLC, Ultimate™ 3000, Dionex, USA) with an Aminex HPX-87P column (300 × 7.8 mm, Bio-Rad, USA) and a RefractoMax521 refractive index detector. After the reaction, the amount of coke formation in the catalysts was determined by TGA in air (ramp: 5 K min−1) using a TGA N-1000 (Scinco). For the typical recycling experiments, the used catalysts were collected by filtration and then calcined at 573 K (ramp: 5 K min−1) in air for 2 h prior to the next reaction.
Sample
SBETa (m2 g−1)
Vtotalb (mL g−1)
H2 uptakec (μmol g−1)
CO uptakec (μmol g−1)
Pt/CNT Re/CNT PtRe/CNT
189 184 171
0.387 0.383 0.344
35.0 0.1 26.2
60.7 8.1 55.9
a b c
BET surface areas (SBET) were determined in the P/P0 range of 0.1–0.3. Total pore volumes (Vtotal) were determined at P/P0 = 0.98. H2 and CO uptakes were determined by chemisorption at 323 K.
weighted mean metal particle diameters (dTEM) of Pt/CNT, Re/CNT, and PtRe/CNT were determined as 1.92, 3.42, and 1.79 nm, respectively. Table 1 summarizes the structural properties and H2/CO chemisorption properties of the catalysts. Pt/CNT chemisorbed both H2 and CO, while Re/CNT chemisorbed only a small amount of CO. Such chemisorption trends were consistent with earlier results [24,25]. The PtRe/CNT catalyst exhibited decreased H2 and CO chemisorptions compared to the Pt/CNT, even though the former showed slightly smaller particle size distribution in TEM (Fig. 1). In addition, the decrease in H2 chemisorption was more pronounced than the decrease in CO chemisorption. The results indicated that Re is mainly enriched at the surface of the bimetallic PtRe particles. Hydrothermal deoxygenation of palm oil as a model triglyceride was carried out both with and without the prepared catalysts. According to our analysis, the palm oil was initially composed of mainly palmitic acid (C16:0, 45 mol%) and oleic acid (C18:1, 39 mol%) (Table 2). In Fig. 2, the product distributions in oil, aqueous, and gas phases are plotted as a function of reaction temperature. The result showed that a blank reaction without a catalyst mainly produced fatty acids in the oil phase (Fig. 2a) and glycerol in the aqueous phase (Fig. 2b) due to the hydrolysis of triglycerides. No detectable amount of the deoxygenated products of fatty acids (i.e., n-paraffins) were produced. The fatty acid composition in an oil phase was very similar to that of initial palm oil (Table 2), indicating that hydrogenation of unsaturated fatty acids did not take place in the absence of a catalyst. In the presence of Pt/CNT, the yields of unsaturated fatty acids decreased gradually while those of saturated fatty acids increased up to 523 K (Fig. 2d). Meanwhile, the yield of glycerol decreased and an enhanced formation of smaller oxygenates (e.g., ethylene glycol and acetic acids) was observed (Fig. 2e). This result indicated that the glycerol was decomposed into smaller oxygenates via APR to produce H2 in situ, which was subsequently used in the hydrogenation of unsaturated fatty acids. Above 523 K, n-paraffins (C15–C18) were formed due to the deoxygenation of fatty acids. The produced n-paraffins at 558 K were mostly composed of C15 (43 mol%) and C17 (53 mol%), and the sum of the portions of C16 and C18 was less than 4 mol%. Considering that the initial palm oil was mainly composed of C16 and C18 fatty acids (Table 2), the major formation of C15 and C17 n-paraffins with one less
3. Results and discussion 3.1. Synergetic effects of Pt and Re in hydrothermal deoxygenation We synthesized three different metal catalysts supported on CNT, namely, Pt/CNT (2.5 wt% Pt), Re/CNT (2.5 wt% Re), and PtRe/CNT (2.5 wt% Pt/1.0 wt% Re). TEM images and the metal particle size distributions of the catalysts are shown in Fig. 1. The surface-area-
Fig. 1. High-resolution TEM images of (a) Pt/CNT, (b) Re/CNT, and (c) PtRe/CNT. 3
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
carbon atom indicated that the deoxygenation of fatty acids occurred dominantly via DCO (the removal of oxygen by CO/CO2) rather than HDO (the removal of oxygen by H2O). Even though Pt catalysts are generally known to prefer DCO to HDO [14–16,43–46], such drastic suppression of HDO is uncommon in gas phase reactions [43,44] or liquid phase reactions using organic solvents [14–16,45,46]. The present results can be explained by the fact that the use of water as a solvent suppressed the HDO pathway producing H2O in accordance with Le Chatelier's principle. In the case of the Re/CNT catalyst, a significant amount of unsaturated fatty acids still remained even at elevated temperatures (> 523 K) and the formation of n-paraffins was also insubstantial
Table 2 Fatty acid composition of palm oil. Fatty acid
Composition (mol%) a
Myristic acid (C14:0 ) Palmitic acid (C16:0) Stearic acid (C18:0) Oleic acid (C18:1) Linoleic acid (C18:2) Avg. # of C = C bonds in a fatty acid
2 45 5 39 9 0.57
a In the notation “Cx:y”, x is the carbon length and y is the number of unsaturated C]C double bonds in fatty acids.
Fig. 2. Product distributions in oil, aqueous, and gas phases after hydrothermal deoxygenation at different reaction temperatures for 24 h, without a catalyst (a–c) and with Pt/CNT (d–f), Re/CNT (g–i), and PtRe/CNT (j–l) catalysts. 4
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
(Fig. 2g). In the aqueous phase, the decomposition of glycerol was also much slower compared to that with the Pt/CNT catalyst (Fig. 2h). The result indicated that the monometallic Re catalyst shows very limited activity in the hydrothermal deoxygenation of triglycerides. The bimetallic PtRe/CNT catalyst showed significantly enhanced hydrogenation and deoxygenation activities, when compared to the Pt/ CNT catalyst (Fig. 2j). All the produced fatty acids were fully saturated even at mild temperatures (e.g., 473 K) and most of fatty acids were deoxygenated to n-paraffin at 558 K. The n-paraffin amount in the oil phase at 558 K (88 C%) was significantly greater than that of Pt/CNT (16 C%). As in the case of Pt/CNT catalyst, the produced n-paraffins contained mostly C15 (49 mol%) and C17 (47 mol%), and the sum of the portions of C16 and C18 was less than 2 mol%. This result again indicated that the deoxygenation of fatty acids occurred dominantly via DCO rather than HDO in hydrothermal conditions. In aqueous phase, the PtRe/CNT catalyst showed the fastest decomposition of glycerol with increasing reaction temperature, which revealed the highest APR activity among all catalysts (Fig. 2k). The present results indicated that PtRe/CNT has superior catalytic activity compared to Pt/CNT in the hydrothermal deoxygenation of triglycerides. The faster decomposition of glycerol and the saturation of fatty acids with PtRe/CNT can be explained by its enhanced APR activity. In Fig. 3, the molar amount of H2 produced per mol of glycerol was plotted for all catalysts. The amount of H2 produced was calculated by summing the amount of H2 in gas phase with that used in the hydrogenation of unsaturated fatty acids. We postulated that the deoxygenation step does not require H2, because the major reaction pathway turned out to be DCO rather than HDO. In DCO mechanisms, decarboxylation does not intrinsically require H2, while decarbonylation requires one H2 molecule for producing saturated n-paraffin. However, the CO produced from decarbonylation can be subsequently converted to H2 via a water-gas shift (WGS) reaction, thereby not requiring a net H2 consumption (i.e., the result of decarbonylation combined with WGS is the same as decarboxylation). This can be verified by the fact that CO was not detected in gas phase, while a substantial amount of CO2 was detected (Fig. 2f, i, and l). The H2 production per mol of glycerol increased in the order of Re/CNT < Pt/CNT < PtRe/CNT (Fig. 3). At 558 K, the PtRe/CNT catalyst produced 3.8 mol of H2 per mol of glycerol, while Pt/CNT produced 2.5 mol of H2. In principle, 7 mol of H2 can be produced from 1 mol of glycerol. The H2 production yield lower than the theoretical yield can be attributed to the incomplete conversion of oxygenate intermediates as well as substantial CeO cleavage instead of CeC cleavage [25–32], which could be confirmed by the formation of C1–C3 hydrocarbons in gas phase (Fig. 2). However, considering that
1.71 mol of H2 is required for saturating all the C]C bonds in 1 mol of palm oil (0.57 × 3, Table 2), PtRe/CNT already produced sufficient amounts of the H2 required for hydrothermal deoxygenation. These results confirm that hydrothermal deoxygenation can enable net-zero H2 consumption in the production of paraffins from triglycerides. According to our previous studies [16], the deoxygenation of 1 mol of palm oil via conventional hydroprocessing required the consumption of 5.3 mol of external H2. The reasons for the superior APR activity of PtRe catalysts have been extensively investigated in previous studies [24–32]. A general consensus is that Re weakens CO binding to nearby Pt atoms and reduces the CO coverage on active Pt surfaces [24,25,28,29], which inhibits the binding of other reactants and reaction intermediates. Since cleavage of CeC bond in oxygenated hydrocarbons requires Pt ensembles [30], the coverage of Pt surface even with a trace amount of CO can significantly suppress the APR activities. To investigate the interaction of CO with the Pt/CNT and PtRe/CNT catalysts, diffuse reflectance infrared fourier transform (DRIFT) spectra were collected after CO adsorption (Fig. 4). All DRIFT spectra showed only the stretching band for linearly adsorbed CO (1980–2080 cm−1), while no band for bridging CO (1900–1970 cm−1) was detected. The intensity of the CO stretching band of PtRe/CNT (Fig. 4b) was much smaller than that for Pt/CNT (Fig. 4a), and the former also decreased more rapidly with an increasing desorption temperature. PtRe/CNT also showed a red shift as compared to Pt/CNT. This can be explained by reduced dipole-dipole coupling due to decreased surface CO coverage [47]. These results supported the weakened interaction of CO with PtRe/CNT
Fig. 3. Molar amount of H2 produced per mol of glycerol during the hydrothermal deoxygenation of palm oil for 24 h.
Fig. 4. CO DRIFT spectra of (a) Pt/CNT and (b) PtRe/CNT collected after desorption at different temperatures (293, 323, 373, 473, and 573 K). 5
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
Table 3 Properties of PtRe catalysts supported on various carbon supports. Sample
SBETa (m2 g−1)
Vtotalb (mL g−1)
H2 uptakec (μmol g−1)
CO uptakec (μmol g−1)
O contentd (%)
PtRe/CNT PtRe/oCNT PtRe/AC PtRe/oAC
171 172 977 613
0.344 0.349 0.835 0.607
26.2 23.2 7.5 6.4
55.9 60.3 14.6 13.3
0.8 3.3 10.8 16.2
a b c d
BET surface areas (SBET) were determined in the P/P0 range of 0.1–0.3. Total pore volumes (Vtotal) were determined at P/P0 = 0.98. H2 and CO uptakes were determined by chemisorption at 323 K. O contents were analyzed by O 1s X-ray photoelectron spectroscopy (XPS).
Fig. 5. n-Paraffin yields after the deoxygenation of stearic acid at different reaction temperatures under 10 bar of H2 for 24 h.
as compared to Pt/CNT. It is also known that the addition of Re also increases the hydrogenolysis (more substantially the rate of CeO bond cleavage than that of CeC bond cleavage) [48] as well as WGS activities [31,49–51], which are both beneficial for achieving enhanced APR activities. Several kinetic studies [31,49,50] and DFT calculations [50,51] indicated that Re addition facilitates water dissociation (H2O → OH* + 1/2H2), which is an important step of WGS. The resultant ReeOH* species neighboring Pt can facilitate the formation of carboxyl intermediate (COOH*) and its further decomposition (COOH* + OH* → CO2 + H2O). The enhanced formation of n-paraffins with PtRe/CNT (Fig. 2j) indicated that the bimetallic catalyst had higher deoxygenation activity than the monometallic catalysts. To investigate the deoxygenation activities without the influence of an APR step, the deoxygenation of stearic acid (C18:0) was additionally carried out in water under an H2 atmosphere (Fig. 5). As expected, PtRe/CNT showed much higher activity than Pt/CNT in the deoxygenation of stearic acid. This can be explained by the higher oxygen affinity of Re than that of Pt [49], which facilitates the binding of fatty acids on the catalyst surface via a ReeOeC bond formation.
Fig. 6. High-resolution TEM images of (a) PtRe/CNT, (b) PtRe/oCNT, (c) PtRe/ AC, and (d) PtRe/oAC.
The PtRe catalysts were analyzed using Pt L3 and Re L3 X-ray absorption spectroscopy (XAS). The X-ray absorption near edge structures (XANES) are shown in Fig. 7a. Earlier studies demonstrated a linear correlation between the metal oxidation states and the energy shift of the absorption edge in XANES [25]. Such analysis revealed that Pt is always highly reduced, and its oxidation state is ∼0.15 regardless of the type of carbon support (Fig. S1). On the other hand, the variation of Re oxidation state is more substantial (1.0–1.9), wherein Re becomes more oxidized with increasing oxygen content of a carbon support (Fig. 7b). Extended X-ray absorption fine structure (EXAFS) indicated that Pt-M coordination increased slightly as the oxygen content of a carbon support increased (Table 4). On the other hand, Re-M coordination increased rapidly along with Re-O coordination. These results indicated that Pt always exists as a highly dispersed and reduced species, while Re becomes segregated into large oxidized particles as the oxygen content of a carbon support increases. This means that Pt-Re alloying becomes gradually less efficient with increasing oxygen content of a carbon surface. Temperature programmed reduction (TPR) also supported this conclusion. Pt/CNT showed one reduction peak at 523 K (Fig. S2), while Re/CNT showed a reduction peak at a significantly higher temperature (573 K). The PtRe/CNT (Fig. S2) catalyst showed a very similar TPR profile to that of Pt/CNT, indicating that Pt and Re were well alloyed and hydrogen activated on the surface of Pt helped the reduction of nearby Re atoms. However, as the oxygen content of the carbon support increased (PtRe/oCNT < PtRe/AC <
3.2. Effects of carbon supports on bimetallic PtRe catalysts In previous researches on hydrothermal deoxygenation, only activated carbons (AC) were used for supporting metal catalysts [18–21]. However, carbons can have a wide range of catalytic properties depending on their structures [35–38] and especially the surface oxygen functional groups can substantially affect the alloying state of bimetallic catalysts [38]. In this regard, four different carbons including CNT, oCNT (oxidized CNT), AC, and oAC (oxidized AC) were investigated for supporting PtRe catalysts. Table 3 summarizes the physical properties and H2/CO chemisorption properties of the catalysts. CNT contained much lower oxygen content (0.8%) than AC (10.8%) due to its highly crystalline structure. The oxygen contents of oCNT (3.3%) and oAC (16.2%) were substantially increased compared to those of pristine CNT and AC, indicating that the HNO3 treatment oxidized the carbon surfaces. The H2 and CO chemisorption amounts on the PtRe catalysts generally decreased with increasing oxygen content of the carbon support (i.e., PtRe/CNT > PtRe/oCNT > PtRe/AC > PtRe/oAC). This indicated that the size of PtRe particles became gradually larger as the oxygen content of carbon support increased. TEM investigation (Fig. 6) also confirmed that the metal particle diameter (dTEM) increased in the order PtRe/CNT (1.79 nm) < PtRe/oCNT (2.35 nm) < PtRe/AC (2.80 nm) < PtRe/oAC (8.27 nm). 6
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
Fig. 8. n-Paraffin yield after the hydrothermal deoxygenation of palm oil at 558 K over the PtRe catalysts supported on various carbon supports.
cause selective segregation of Re species and inefficient Pt-Re alloying. Hydrothermal deoxygenation of palm oil over the PtRe catalysts supported on CNT, oCNT, AC, and oAC was carried out at 558 K. The yields of n-paraffin are shown in Fig. 8. Higher n-paraffin yields were achieved with the PtRe catalyst supported on the carbons with less oxygen content (i.e., PtRe/oAC < PtRe/AC < PtRe/oCNT < PtRe/ CNT). Based on the aforementioned structural analyses, it can be concluded that well alloyed and highly dispersed PtRe particles supported on oxygen-deficient carbon supports (e.g., CNT) are more active in the hydrothermal deoxygenation. The n-paraffin yield with respect to the initial weight of the palm oil was 72 wt% over PtRe/CNT, which was quite close to the theoretical n-paraffin yield (79 wt%). The theoretical yield was calculated by assuming that glycerol was fully removed, and all fatty acids were saturated and deoxygenated solely via a DCO pathway.
Fig. 7. (a) XANES of the PtRe catalysts supported on different carbon supports and (b) the oxidation state of Re calculated using the energy shift of the absorption edge.
3.3. Recyclability of PtRe/CNT in the hydrothermal deoxygenation of palm oil
Table 4 EXAFS fitting results of the PtRe catalysts supported on various carbons. Sample
Shell
CN
PtRe/CNT
Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2
7.17 4.36 0.14 0.06 7.91 5.38 0.38 0.33 7.99 6.44 0.88 0.66 8.01 9.12 1.67 0.71
PtRe/oCNT
PtRe/AC
PtRe/oAC
σ2 (Å2)
R (Å) ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.32 0.41 0.05 0.02 0.42 0.78 0.08 0.04 0.62 0.56 0.21 0.14 0.72 1.21 0.21 0.18
2.74 2.59 1.53 1.99 2.74 2.61 1.72 1.93 2.75 2.60 1.83 2.12 2.75 2.64 1.69 2.05
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.02 0.04 0.07 0.02 0.04 0.07 0.02 0.09 0.03 0.06 0.09 0.02 0.04 0.06 0.03 0.09
0.007 0.009 0.005 0.011 0.007 0.012 0.024 0.017 0.007 0.010 0.005 0.004 0.005 0.028 0.007 0.010
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
Catalyst recyclability in the hydrothermal deoxygenation of palm oil was investigated using PtRe/CNT, which showed the best performance among the prepared catalysts. Prior to the recycling experiments, we first analyzed the used catalyst by thermogravimetric analysis (TGA; Fig. 9a). The catalyst collected after the reaction showed the formation of light carbonaceous residue (14 wt%). This can be explained by the polymerization of unsaturated fatty acids via the Diels-Alder reaction and/or a radical-mediated addition at elevated temperatures [45]. Fortunately, due to the high thermal stability of CNT, such polymeric species could be selectively removed by calcination in the temperature range of 473–673 K (Fig. 9a). It is notable that, in the case of AC with low thermal stability, such selective removal of polymeric species is not possible (Fig. S4) because the combustion of AC occurs at a similar temperature regime. Based on the TGA data, we carried out catalyst recycling experiments with and without calcination at 573 K to remove the polymeric species between each reaction cycle (Fig. 9b). Without calcination, the yield of n-paraffin significantly decreased from 72 wt% to 14 wt% after only 3 reaction cycles. With calcination, however, the yield of n-paraffin decreased only slightly from 72 wt% to 66 wt% after 5 cycles. These results showed that the catalyst deactivation is mainly caused by the deposition of polymeric species on the metal surfaces, and thus their removal by mild calcination is necessary for effectively recycling a catalyst. TEM investigation of the PtRe/CNT catalysts after 5 reaction cycles (Fig. S5) indicated a slight increase of metal particle size (dTEM:
0.002 0.001 0.001 0.002 0.001 0.001 0.004 0.002 0.002 0.001 0.002 0.001 0.001 0.004 0.002 0.001
PtRe/oAC), broad peaks were clearly observed above 523 K (Fig. S3). This result supported less efficient Pt-Re alloying on the oxygen-rich carbon surfaces. We believe that the highly oxophilic Re species interact strongly with the oxygen functional groups of the carbon surface, leading to segregation of Re nearby these groups. On the other hand, Pt species interact less strongly with these functional groups, and thus, could be distributed uniformly over the carbon surface. Therefore, the high density of oxygen functional groups on the carbon surface might 7
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
favorable for forming a highly dispersed and well alloyed PtRe catalyst, which is the prerequisite for achieving high hydrothermal deoxygenation activity. An oxygen-rich carbon surface (e.g., oxidized AC) led to the segregation of Re species into bulk oxidized particles, which led to low catalytic activity. The PtRe/CNT showed the best performance, with a paraffin yield of 72 wt% (theoretical yield: 79 wt%) from palm oil, even in the absence of an external H2 source. The PtRe catalyst deactivated mainly via the deposition of polymerized unsaturated fatty acids, and no significant metal sintering/leaching was observed even under the harsh hydrothermal conditions. In the case of PtRe supported on CNT with high thermal stability, the polymerized unsaturated fatty acids could be selectively removed by calcination at a mild temperature (573 K), and the catalyst could be efficiently reused for 5 cycles without significant loss of activity. Because of its high activity and recyclability, PtRe/CNT is a highly promising catalyst for the hydrothermal deoxygenation of triglycerides. Conflicts of interest There are no conflicts to declare. Acknowledgements This work was supported by the Advanced Biomass R&D Center (ABC) of Global Frontier Project funded by the Ministry of Science, ICT, and Future Planning (ABC-2015M3A6A2066121) and the Basic Science Research Program through the National Research Foundation of Korea (NRF-2017R1A2B2002346). The authors thank the Pohang Accelerator Laboratory (PAL) for the use of the beamline. Appendix A. Supplementary data Fig. 9. (a) Thermogravimetric analysis of fresh PtRe/CNT and used PtRe/CNT catalysts in air. (b) Recycling experiments of the PtRe/CNT catalyst with and without calcination (calcination temperature: 573 K) between each reaction cycle.
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.mcat.2019.110419. References
2.17 nm) compared to the fresh sample (dTEM: 1.79 nm). We also performed additional experiments for investigating metal leaching. The PtRe/CNT catalyst was repeatedly treated in pure H2O at 573 K for 2 h and the concentrations of Pt and Re in H2O were analyzed by ICP-OES. The results showed that none of the Pt was leached out during the hydrothermal treatments. On the other hand, 10% of Re in the fresh catalyst was leached out after the first treatment. However, Re leaching from the second treatment was not detectable. This result indicated the hydrothermal stability of PtRe bimetallic particles, which is consistent with an earlier report [29] that unalloyed Re species were readily leached out, but Re species alloyed with Pt were highly stable. The slight sintering of metal catalysts and/or minor leaching of Re species might be the reason for the small decrease of paraffin yield with repeated reaction cycles.
[1] G.W. Huber, A. Corma, Synergies between bio- and oil refineries for the production of fuels from biomass, Angew. Chem. Int. Ed. 46 (2007) 7184–7201, https://doi. org/10.1002/anie.200604504. [2] H. Baumann, M. Bühler, H. Fochem, F. Hirsinger, H. Zoebelein, J. Falbe, Natural fats and oils-renewable raw materials for the chemical industry, Angew. Chem. Int. Ed. Engl. 27 (1988) 41–62, https://doi.org/10.1002/anie.198800411. [3] A. Srivastava, R. Prasad, Triglyceride-based diesel fuels, Renew. Sustain. Energy Rev. 4 (2000) 111–133, https://doi.org/10.1016/S1364-0321(99)00013-1. [4] L.C. Meher, D.V. Sagar, S.N. Naik, Technical aspects of biodiesel production by transesterification-a review, Renew. Sustain. Energy Rev. 10 (2006) 248–268, https://doi.org/10.1016/j.rser.2004.09.002. [5] G. Knothe, Biodiesel and renewable diesel: a comparison, Prog. Energy Combust. Sci. 36 (2010) 364–373, https://doi.org/10.1016/j.pecs.2009.11.004. [6] M. Lapuerta, O. Armas, J. Rodríguez-Fernández, Effect of biodiesel fuels on diesel engine emissions, Prog. Energy Combust. Sci. 34 (2008) 198–223, https://doi.org/ 10.1016/j.pecs.2007.07.001. [7] G. Dwivedi, M.P. Sharma, Impact of cold flow properties of biodiesel on engine performance, Renew. Sustain. Energy Rev. 31 (2014) 650–656, https://doi.org/10. 1016/j.rser.2013.12.035. [8] M. Stumborg, A. Wong, E. Hogan, Hydroprocessed vegetable oils for diesel fuel improvement, Bioresour. Technol. 56 (1996) 13–18, https://doi.org/10.1016/ 0960-8524(95)00181-6. [9] R. Sotelo-Boyás, Y. Liu, T. Minowa, Renewable diesel production from the hydrotreating of rapeseed oil with Pt/Zeolite and NiMo/Al2O3 catalysts, Ind. Eng. Chem. Res. 50 (2011) 2791–2799, https://doi.org/10.1021/ie100824d. [10] B. Veriansyah, J.Y. Han, S.K. Kim, S.A. Hong, Y.J. Kim, J.S. Lim, Y.W. Shu, S.G. Oh, J. Kim, Production of renewable diesel by Hydroprocessing of soybean oil: effect of catalysts, Fuel 94 (2012) 578–585, https://doi.org/10.1016/j.fuel.2011.10.057. [11] B. Peng, X. Yuan, C. Zhao, J.A. Lercher, Stabilizing catalytic pathways via redundancy: selective reduction of microalgae oil to Alkanes, J. Am. Chem. Soc. 134 (2012) 9400–9405, https://doi.org/10.1021/ja302436q. [12] B. Peng, Y. Yao, C. Zhao, J.A. Lercher, Towards quantitative conversion of microalgae oil to diesel-range alkanes with bifunctional catalysts, Angew. Chem. Int. Ed. 51 (2012) 2072–2075, https://doi.org/10.1002/anie.201106243. [13] L. Boda, G. Onyestyák, H. Solt, F. Lónyi, J. Valyon, A. Thernesz, Catalytic hydroconversion of Tricaprylin and caprylic acid as model reaction for biofuel production from triglycerides, Appl. Catal. A Gen. 374 (2010) 158–169, https://doi.org/10.
4. Conclusions We investigated bimetallic PtRe catalysts supported on various carbon supports in the hydrothermal deoxygenation of triglycerides. Bimetallic PtRe catalysts showed a markedly enhanced rate of glycerol reforming for in situ H2 production compared to the Pt catalysts, which can be explained by the weakened interaction of CO with PtRe. The PtRe catalysts also exhibited a significantly enhanced rate for the deoxygenation of fatty acids, which could be attributed to the high oxygen affinity of Re. Consequently, PtRe catalysts showed superior performance in the hydrothermal deoxygenation of triglycerides than monometallic catalysts. Investigations using four different carbon supports (CNT and AC with and without surface oxidation) showed that the carbon support with low surface oxygen content (e.g., CNT) was 8
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi 1016/j.apcata.2009.12.005. [14] A.T. Madsen, E.H. Ahmed, C.H. Christensen, R. Fehrmann, A. Riisager, Hydrodeoxygenation of waste fat for diesel production: study on model feed with Pt/alumina catalyst, Fuel 90 (2011) 3433–3438, https://doi.org/10.1016/j.fuel. 2011.06.005. [15] K. Murata, Y. Liu, M. Inaba, I. Takahara, Production of synthetic diesel by Hydrotreatment of Jatropha oils using Pt-Re/H-ZSM-5 catalyst, Energy Fuels 24 (2010) 2404–2409, https://doi.org/10.1021/ef901607t. [16] M.Y. Kim, J.K. Kim, M.E. Lee, S. Lee, M. Choi, Maximizing biojet fuel production from triglyceride: importance of the hydrocracking catalyst and separate deoxygenation/hydrocracking steps, ACS Catal. 7 (2017) 6256–6267, https://doi.org/ 10.1021/acscatal.7b01326. [17] A. Anumakonda, M.J. McCall, T.A. Brandvold, J.A. Kocal, Controlling Production of Transportation Fuels from Renewable Feedstocks. U. S. Patent 8,058,492, November 15, 2011. [18] J. Fu, X. Lu, P.E. Savage, Catalytic hydrothermal deoxygenation of palmitic acid, Energy Environ. Sci. 3 (2010) 311–317, https://doi.org/10.1039/B923198F. [19] F. Fu, X. Lu, P.E. Savage, Hydrothermal decarboxylation and hydrogenation of fatty acids over Pt/C, ChemSusChem 4 (2011), https://doi.org/10.1002/cssc.201000370 481-186. [20] S.A.W. Hollak, M. Ariëns, K.P. de Jong, D.S. van Es, Hydrothermal deoxygenation of triglycerides over Pd/C aided by in situ hydrogen production from glycerol reforming, ChemSusChem 7 (2014) 1057–1060, https://doi.org/10.1002/cssc. 201301145. [21] D.R. Vardon, B.K. Sharma, H. Jaramillo, D. Kim, J.K. Choe, P.N. Ciesielski, T.J. Strathmann, Hydrothermal catalytic processing of saturated and unsaturated fatty acids to hydrocarbons with glycerol for in situ hydrogen production, Green Chem. 16 (2014) 1507–1520, https://doi.org/10.1039/C3GC41798K. [22] D. Kim, D.R. Vardon, D. Murali, B.K. Sharma, T.J. Strathmann, Valorization of waste lipids through hydrothermal catalytic conversion to liquid hydrocarbon fuels with in situ hydrogen production, ACS Sustain. Chem. Eng. 4 (2016) 1775–1784, https://doi.org/10.1021/acssuschemeng.5b01768. [23] C. Miao, O. Marin-Flores, T. Dong, D. Gao, Y. Wang, M. Garcia-Pérez, S. Chen, Hydrothermal catalytic deoxygenation of fatty acid and bio-oil with in situ H2, ACS Sustain. Chem. Eng. 6 (2018) 4521–4530, https://doi.org/10.1021/acssuschemeng. 7b02226. [24] D.A. Simonetti, E.L. Kunkes, J.A. Dumesic, Gas-phase conversion of glycerol to synthesis gas over carbon-supported platinum and platinum-rhenium catalysts, J. Catal. 247 (2007) 298–306, https://doi.org/10.1016/j.jcat.2007.01.022. [25] E.L. Kunkes, D.A. Simonetti, J.A. Dumesic, W.D. Pyrz, L.E. Murillo, J.G. Chen, D.J. Buttrey, The role of rhenium in the conversion of glycerol to synthesis gas over carbon supported platinum-rhenium catalysts, J. Catal. 260 (2008) 164–177, https://doi.org/10.1016/j.jcat.2008.09.027. [26] D.L. King, L. Zhang, G. Xia, A.M. Karim, D.J. Heldebrant, X. Wang, T. Peterson, Y. Wang, Aqueous phase reforming of glycerol for hydrogen production over Pt-Re supported on carbon, Appl. Catal. B 99 (2010) 206–213, https://doi.org/10.1016/j. apcatb.2010.06.021. [27] L. Zhang, A.M. Karim, M.H. Engelhard, Z. Wei, D.L. King, Y. Wang, Correlation of Pt-Re surface properties with reaction pathways for the aqueous-phase reforming of glycerol, J. Catal. 287 (2012) 37–43, https://doi.org/10.1016/j.jcat.2011.11.015. [28] Z. Wei, A.M. Karim, Y. Li, D.L. King, Y. Wang, Elucidation of the roles of Re in steam reforming of glycerol over Pt-Re/C catalysts, J. Catal. 322 (2015) 49–59, https:// doi.org/10.1016/j.jcat.2014.11.006. [29] Z. Wei, A.M. Karim, Y. Li, Y. Wang, Elucidation of the roles of Re in aqueous-phase reforming of glycerol over Pt-Re/C catalysts, ACS Catal. 5 (2015) 7312–7320, https://doi.org/10.1021/acscatal.5b01770. [30] A. Ciftci, D.A.J.M. Ligthart, E.J.M. Hensen, Influence of Pt particle size and Re addition by catalytic reduction on aqueous phase reforming of glycerol for carbonsupported Pt(Re) catalysts, Appl. Catal. B 174 (2015) 126–135, https://doi.org/10. 1016/j.apcatb.2015.02.027. [31] A. Ciftci, D.A.J.M. Ligthart, A.O. Sen, A.J.F. van Hoof, H. Friedrich, E.J.M. Hensen, Pt-Re synergy in aqueous-phase reforming of glycerol and the water-gas shift reaction, J. Catal. 311 (2014) 88–101, https://doi.org/10.1016/j.jcat.2013.11.011. [32] A. Ciftci, S. Eren, D.A.J.M. Ligthart, E.J.M. Hensen, Platinum-rhenium synergy on
[33]
[34] [35] [36] [37]
[38]
[39] [40] [41]
[42] [43] [44]
[45]
[46] [47] [48] [49] [50]
[51]
9
reducible oxide supports in aqueous-phase glycerol reforming, ChemCatChem 6 (2014) 1260–1269, https://doi.org/10.1002/cctc.201301096. H.N. Pham, A.E. Anderson, R.L. Johnson, K. Schmidt-Rohr, A.K. Datye, Improved Hydrothermal Stability of Mesoporous Oxides for Reactions in the Aqueous Phase, Angew. Chem. Int. Ed. 51 (2012) 13163–13167, https://doi.org/10.1002/anie. 201206675. A. Ciftci, B. Peng, A. Jentys, J.A. Lercher, E.J.M. Hensen, Support effects in the aqueous phase reforming of glycerol over supported platinum catalysts, Appl. Catal. A Gen. 431 (2012) 113–119, https://doi.org/10.1016/j.apcata.2012.04.026. F. Rodríguez-Reinoso, The role of carbon materials in heterogeneous catalysis, Carbon 36 (1998) 159–175, https://doi.org/10.1016/S0008-6223(97)00173-5. D.S. Su, S. Perathoner, G. Centi, Nanocarbons for the development of advanced catalysts, Chem. Rev. 113 (2013) 5782–5816, https://doi.org/10.1021/cr300367d. H.C. Kwon, S. Yook, S. Choi, M. Choi, Comprehensive understanding of the effects of carbon nanostructures on redox catalytic properties and stability in oxidative dehydrogenation, ACS Catal. 7 (2017) 5257–5267, https://doi.org/10.1021/ acscatal.7b01742. S. Yook, H.C. Kwon, Y.G. Kim, W. Choi, M. Choi, Significant roles of carbon pore and surface structure in AuPd/C catalyst for achieving high chemoselectivity in direct hydrogen peroxide synthesis, ACS Sustain. Chem. Eng. 5 (2017) 1208–1216, https://doi.org/10.1021/acssuschemeng.6b02595. M. Choi, Z. Wu, E. Iglesia, Mercaptosilane-assisted synthesis of metal clusters within zeolites and catalytic consequences of encapsulation, J. Am. Chem. Soc. 132 (2010) 9129–9137, https://doi.org/10.1021/ja102778e. H. Hirai, K. Wada, M. Komiyama, Active carbon-supported copper (I) chloride as solid adsorbent for carbon monoxide, Bull. Chem. Soc. Jpn. 59 (1986) 2217–2223, https://doi.org/10.1246/bcsj.59.2217. S.R. Bare, S.D. Kelly, F.D. Vila, E. Boldingh, E. Karapetrova, J. Kas, G.E. Mickelson, F.S. Modica, N. Yang, J.J. Rehr, Experimental (XAS, STEM, TPR, and XPS) and theoretical (DFT) characterization of supported rhenium catalysts, J. Phys. Chem. C 115 (2011) 5740–5755, https://doi.org/10.1021/jp1105218. D. Darnoko, M. Cheryan, Kinetics of palm oil transesterification in a batch reactor, J. Am. Oil Chem. Soc. 77 (2000) 1263–1267, https://doi.org/10.1007/s11746-0000198-y. W. Rachmady, M.A. Vannice, Acetic acid hydrogenation over supported platinum catalysts, J. Catal. 192 (2000) 322–334, https://doi.org/10.1006/jcat.2000.2863. Y.K. Lugo-José, J.R. Monnier, C.T. Williams, Gas-phase, catalytic hydrodeoxygenation of propanoic acid, over supported group VIII noble metals: metal and support effects, Appl. Catal. A Gen. 469 (2014) 410–418, https://doi.org/10. 1016/j.apcata.2013.10.025. T.H. Kim, K. Lee, M.Y. Kim, Y.K. Chang, M. Choi, Effects of fatty acid compositions on heavy oligomer formation and catalyst deactivation during deoxygenation of triglycerides, ACS Sustain. Chem. Eng. 6 (2018) 17168–17177, https://doi.org/10. 1021/acssuschemeng.8b04552. L. Yang, K.L. Tate, J.B. Jasinski, M.A. Carreon, Decarboxylation of oleic acid to Heptadecane over Pt supported on zeolite 5A beads, ACS Catal. 5 (2015) 6497–6502, https://doi.org/10.1021/acscatal.5b01913. J. Raskó, CO-induced surface structural changes of Pt on oxide-supported Pt catalysts studied by DRIFTS, J. Catal. 217 (2003) 478–486, https://doi.org/10.1016/ S0021-9517(03)00147-7. J. Zhang, M.B. Vukmirovic, K. Sasaki, A.U. Nilekar, M. Mavrikakis, R.R. Adzic, Mixed-metal Pt monolayer electrocatalysts for enhanced oxygen reduction kinetics, J. Am. Chem. Soc. 127 (2005) 12480–12481, https://doi.org/10.1021/ja053695i. K.G. Azzam, I.V. Babich, K. Seshan, L. Lefferts, Role of Re in Pt-Re/TiO2 catalyst for water gas shift reaction: a mechanistic and kinetic study, Appl. Catal. B 80 (2008) 129–140, https://doi.org/10.1016/j.apcatb.2007.11.015. R. Carrasquillo-Flores, J.M.R. Gallo, K. Hahn, J.A. Dumesic, M. Mavrikakis, Density functional theory and reaction kinetics studies of the water-gas shift reaction on PtRe catalysts, ChemCatChem 5 (2013) 3690–3699, https://doi.org/10.1002/cctc. 201300365. L.C. Grabow, A.A. Gokhale, S.T. Evans, J.A. Dumesic, M. Mavrikakis, Mechanism of the water gas shift reaction on Pt: first principles, experiments, and microkinetic modeling, J. Phys. Chem. C 112 (2008) 4608–4617, https://doi.org/10.1021/ jp7099702.
Contents lists available at ScienceDirect
Molecular Catalysis journal homepage: www.elsevier.com/locate/mcat
Hydrothermal deoxygenation of triglycerides over carbon-supported bimetallic PtRe catalysts without an external hydrogen source Mingyu Jin, Minkee Choi
T
⁎
Department of Chemical & Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea
ARTICLE INFO
ABSTRACT
Keywords: Hydrothermal deoxygenation Triglyceride PtRe Bimetal Carbon supports
Hydrothermal deoxygenation is a promising route for producing oxygen-free hydrocarbons from triglycerides even without an external H2 source. In this study, PtRe catalysts supported on various carbons were investigated for this reaction. The PtRe catalysts showed enhanced rates of glycerol reforming for in situ H2 production and higher deoxygenation activities, as compared to Pt catalysts. Consequently, the PtRe catalysts showed superior hydrothermal deoxygenation performances. Carbon supports with low oxygen content (e.g., CNT) were beneficial for supporting highly dispersed and well alloyed PtRe particles, which were prerequisites for high catalytic activity. Oxygen-rich carbon surfaces caused the segregation of Re, causing low catalytic activity. PtRe/CNT showed the best performance, yielding 72 wt% n-paraffin (theoretical yield: 79 wt%) from palm oil. The catalyst could be recycled for 5 times without significant loss of activity.
1. Introduction Triglycerides (e.g., microalgal oil, vegetable oil, and waste cooking oil) are a promising feedstock for producing bio-derived fuels. They are esters combining one glycerol and three long-chain fatty acids (generally C14−C22) which contain relatively low oxygen contents and high energy densities compared to other biomass feedstock [1–3]. The synthesis of biodiesel (e.g., fatty acid methyl ester) from triglycerides via transesterification is a major commercially applied process [3,4]. However, the presence of oxygen in biodiesel diminishes its energy density, storage stability, and cold flow properties [5–7]. To overcome the limitations of biodiesel, catalytic upgrading of triglycerides to oxygen-free fuels through hydroprocessing has been developed [8–17]. In these processes, triglycerides are hydrotreated over MoS2-based catalysts [9,10] or supported metal catalysts [11,9–17] under highpressure H2. Under such conditions, triglycerides containing unsaturated fatty acid units are first hydrogenated and cleaved into three saturated fatty acids by hydrogenolysis (Scheme 1a). The resulting fatty acids are then deoxygenated via decarboxylation/decarbonylation (DCO) and hydrodeoxygenation (HDO) pathways. DCO produces nparaffins with one less carbon atom than the initial fatty acids, whereas HDO produces n-paraffins having the same carbon number. With the production of long-chain paraffins, light gaseous hydrocarbons (C1–C3) are also coproduced due to the glycerol hydrogenolysis and the CO
⁎
methanation. These side reactions all consume significant amounts of H2. Although the produced liquid paraffins have significantly improved fuel quality, the large consumption of H2 in these processes is disadvantageous for overall economics and sustainability. It should be noted that H2 is currently primarily produced from natural gas. Hydrothermal deoxygenation of triglycerides in subcritical water (523–653 K) can provide an alternative pathway for producing oxygenfree hydrocarbons [18–23]. Savage and colleagues first reported that the high-temperature hydrothermal treatment of model fatty acids over Pt/C and Pd/C catalysts could produce saturated paraffins even in the absence of an external H2 source [18,19]. The result indicated that hydrogen could be supplied in situ during the reaction. Considering that a significant amount of heavy oligomers were also formed, hydrogen transfer might be involved under those reaction conditions. More recently, other researchers investigated the hydrothermal deoxygenation of natural triglycerides or the mixtures of fatty acids and glycerol [20–23]. The main difference between these studies and the earlier works by Savage is that glycerol in the reactants can be readily transformed to H2 in situ via aqueous phase reforming (APR) (Scheme 1b). In principle, one glycerol can generate seven H2 molecules at maximum through APR [24–32], which is sufficient for the saturation and the deoxygenation of fatty acids. Under this novel reaction scheme, glycerol acts as a source for H2 production rather than a consumer of H2 (the case of conventional hydroprocessing). This can enable essentially net-
Corresponding author. E-mail address: [email protected] (M. Choi).
https://doi.org/10.1016/j.mcat.2019.110419 Received 19 March 2019; Received in revised form 19 May 2019; Accepted 20 May 2019 Available online 30 May 2019 2468-8231/ © 2019 Elsevier B.V. All rights reserved.
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
carried out using a Sigma Probe instrument (Thermo Fischer Scientific VG) equipped with monochromatic Al K radiation (1486.3 eV; anode operating at 15 kV and 7 mA). The pass energy was fixed at 50 eV for all survey scans. Samples were mounted using double-sided adhesive carbon tape, and binding energies were calibrated to the C (1s) binding energy (284.5 eV). N2 adsorption-desorption isotherms of the catalysts were measured using a Tristar volumetric analyzer (Micromeritics) at 77 K. Prior to the measurements, all samples were vacuum-degassed at 473 K for 6 h. H2 and CO chemisorptions were measured using an ASAP 2020c (Micromeritics). Before the isotherm measurement, the Pt and PtRe catalysts were vacuum-degassed at 473 K for 6 h, reduced under H2 at 573 K for 1 h, followed by evacuation at 573 K for 1 h. In the case of Re catalysts, the pre-reduction and evacuation were carried out at 723 K. Adsorption isotherms were measured at 323 K and the chemisorption amount was determined by extrapolating the linear regime (10–30 kPa) of the isotherm to zero pressure. Transmission electron microscopy (TEM) images were taken using a Talos F200X microscope operating at 200 kV. Samples were mounted on a carbon-coated copper grid (300 mesh) using ethanol dispersion. To determine the metal size distributions using TEM, at least 200 metal particles were counted. The surfacearea-weighted mean particle diameter was calculated using the following equation [39]:
Scheme 1. Production of oxygen-free paraffin fuels from triglycerides via (a) hydroprocessing and (b) hydrothermal deoxygenation.
zero H2 consumption in the production of oxygen-free paraffins from triglycerides [21,22]. Pt and Pd catalysts supported on activated carbons have been reported to show activities in hydrothermal deoxygenation [18,19,21]. The choice of noble metal catalysts and carbon supports can be rationalized by their stability even under harsh hydrothermal conditions (563–653 K) that can severely degrade other metals and metal oxides via sintering or phase transition [33,34]. Recent studies by Vardon et al. demonstrated that bimetallic PtRe supported on activated carbons can show superior catalytic activity as compared to a monometallic Pt catalyst [21]. However, these studies mainly focused on the effects of reaction conditions (e.g., temperature and reactant compositions) rather than the effects of catalyst structures. Previous studies showed that carbon-supported catalysts can have a wide range of catalytic properties depending on the carbon structures [35–38]. In particular, the oxygen functional groups on carbon surfaces can significantly affect the alloying state of bimetallic catalysts and thus their catalytic properties [38]. Therefore, in the present work, we carefully investigated the catalytic properties of bimetallic PtRe particles supported on various carbon supports for the hydrothermal deoxygenation of triglycerides (e.g., palm oil).
dTEM (nm) = Σ nidi3 / Σ nidi2
(1)
TPR profiles were collected using a BELCAT II (MicrotracBEL Corp.) equipped with a thermal conductivity detector (TCD). Typically, 50 mg of the catalyst was loaded into a quartz sample tube and degassed in Ar at 473 K for 2 h. TPR profiles were collected while increasing the temperature from 323 K to 1073 K (ramp: 10 K min−1) under 5% H2/Ar flow (30 ml min−1). To remove CO, CO2, and H2O generated by the thermal decomposition of the oxygen functional groups on the carbon surface, a trap containing CuCl-modified activated carbon and 13X molecular sieves was installed before TCD with its temperature set to 273 K using an ice bath [37,40]. DRIFT spectroscopy was carried out using a Bruker Vertex 70 spectrometer equipped with a Pike Diffuse IR cell attachment. Spectra were collected using an MCT detector with a 2 cm−1 resolution. The sample was pre-reduced under H2 at 573 K for 1 h. After cooling to 293 K, background spectra were collected. CO adsorption was carried out by purging the IR cell with 1% CO/He at room temperature for 1 h. After subsequent He purging at 293, 323, 373, 473, and 573 K for 1 h, the spectra were collected. XAS on the Pt and Re L3 edges was carried out in transmission mode at the Pohang Accelerator Laboratory (7D-XAFS beamline). For the measurements, 0.3 g samples were pelletized (13 mm in diameter) and placed in a home-made in situ cell equipped with 0.05 mm-thick aluminum windows. Energy calibration was carried out using Pt and Re foils. Prior to the measurements, catalyst samples were pre-reduced under H2 at 573 K for 1 h. Pt foil, PtO2, Re powder, ReO2, and NH4ReO4 were used as Pt and Re reference samples. The Athena and Artemis programs from the IFEFFIT data package were used for processing the XAS data. Normalization and background removal were carried out using Athena. Normalized EXAFS data were k3-weighted and Fouriertransformed into R space for obtaining the radial distribution functions. The theoretical EXAFS signal was constructed using the FEFF6 code and fitted. For the EXAFS fitting of Re, we used two different Re-O scattering paths at 1.73 and 2.08 Å, as proposed by Bare et al [41].
2. Experimental 2.1. Catalyst preparation CNT (Ctube) and AC (Aldrich) were treated in a 6 M HCl solution at room temperature for 2 h to remove possible metal impurities. To introduce additional oxygen functional groups, 1 g of CNT and AC were oxidized in 100 ml of 3 M HNO3 solution for 2 h at 403 K and 323 K, respectively. After filtration, the carbons were thoroughly washed with deionized water and dried at 373 K. The oxidized CNT and AC samples were denoted as oCNT and oAC. Monometallic Pt (2.5 wt%) and Re (2.5 wt%) catalysts were supported on the carbon supports by incipient wetness impregnation of aqueous solutions of chloroplatinic acid hexahydrate (H2PtCl6·6H2O, Kojima Chemical) and perrhenic acid (HReO4, Aldrich). Bimetallic PtRe catalyst (2.5 wt% Pt/1.0 wt% Re) was supported by co-impregnation of the Pt and Re precursors on the carbon supports. After impregnation, the samples were dried at 373 K overnight, followed by reduction under H2 flow. Pt and PtRe catalysts were reduced at 573 K (ramp: 2 K min−1) for 2 h, while monometallic Re catalysts were reduced at 723 K.
2.3. Analysis of fatty acid composition in palm oil
2.2. Characterization
To analyze the fatty composition of palm oil (Aldrich), transesterification with methanol was carried out. For transesterification, 1 g of palm oil was reacted with 0.25 g of methanol (methanol/triglyceride ratio ˜6) at 338 K for 1 h in the presence of 1 wt% NaOH as a catalyst
Metal concentrations were analyzed using inductively coupled plasma optical emission spectrometer (ICP-OES 720, Agilent). For quantifying the amount of oxygen on carbon surfaces, O 1s XPS was 2
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
[42]. The upper phase (fatty acid methyl esters) were dissolved in chloroform (dodecane was used as an external standard) and then analyzed using a gas chromatograph (GC) equipped with a flame ionization detector (FID) and a HP-5 column (Agilent, 30 m × 0.32 mm).
Table 1 Properties of Pt, Re, and PtRe catalysts supported on CNT.
2.4. Catalytic measurements Hydrothermal deoxygenation of the palm oil was carried out at various temperatures (448, 473, 498, 523, and 558 K) in an autoclave with a 200 ml volume. For a typical reaction, 20 g of palm oil (Aldrich), 80 g of deionized water, and 0.92 g of catalysts, pre-reduced at 573 K, were loaded. After filling with 10 bar N2, the autoclave was heated to the reaction temperature under static conditions. After temperature stabilization, the reaction was initiated by starting magnetic stirring (500 rpm) and continued for 24 h. The deoxygenation of stearic acid in H2 was carried out in a similar manner. For the reaction, 20 g of stearic acid (Aldrich), 80 g of deionized water, and 0.92 g of catalysts were loaded in the autoclave. Instead of N2, the reactor headspace was filled with 10 bar H2 before the reaction. After the reaction, gas compositions in the headspace were analyzed using an online GC equipped with an FID and a GS-GasPro column (Agilent, 30 m × 0.32 mm), as well as a TCD and a Carboxen 1000 column (SUPELCO). Liquid products were mixed with 300 ml of dichloromethane (DCM), followed by the separation of catalysts by filtration. From the filtrate, the organic phase (lower phase) was collected after phase separation. After removing the DCM using a rotary evaporator, the total yield of products in the oil phase was determined by weighing. The detailed compositions of the products were analyzed using a GC equipped with an FID and a DB-FFAP column (Agilent, 30 m × 0.25 mm), after dissolution in chloroform (n-dodecane was used as an external standard). The concentration of products in an aqueous phase was analyzed using high-performance liquid chromatography (HPLC, Ultimate™ 3000, Dionex, USA) with an Aminex HPX-87P column (300 × 7.8 mm, Bio-Rad, USA) and a RefractoMax521 refractive index detector. After the reaction, the amount of coke formation in the catalysts was determined by TGA in air (ramp: 5 K min−1) using a TGA N-1000 (Scinco). For the typical recycling experiments, the used catalysts were collected by filtration and then calcined at 573 K (ramp: 5 K min−1) in air for 2 h prior to the next reaction.
Sample
SBETa (m2 g−1)
Vtotalb (mL g−1)
H2 uptakec (μmol g−1)
CO uptakec (μmol g−1)
Pt/CNT Re/CNT PtRe/CNT
189 184 171
0.387 0.383 0.344
35.0 0.1 26.2
60.7 8.1 55.9
a b c
BET surface areas (SBET) were determined in the P/P0 range of 0.1–0.3. Total pore volumes (Vtotal) were determined at P/P0 = 0.98. H2 and CO uptakes were determined by chemisorption at 323 K.
weighted mean metal particle diameters (dTEM) of Pt/CNT, Re/CNT, and PtRe/CNT were determined as 1.92, 3.42, and 1.79 nm, respectively. Table 1 summarizes the structural properties and H2/CO chemisorption properties of the catalysts. Pt/CNT chemisorbed both H2 and CO, while Re/CNT chemisorbed only a small amount of CO. Such chemisorption trends were consistent with earlier results [24,25]. The PtRe/CNT catalyst exhibited decreased H2 and CO chemisorptions compared to the Pt/CNT, even though the former showed slightly smaller particle size distribution in TEM (Fig. 1). In addition, the decrease in H2 chemisorption was more pronounced than the decrease in CO chemisorption. The results indicated that Re is mainly enriched at the surface of the bimetallic PtRe particles. Hydrothermal deoxygenation of palm oil as a model triglyceride was carried out both with and without the prepared catalysts. According to our analysis, the palm oil was initially composed of mainly palmitic acid (C16:0, 45 mol%) and oleic acid (C18:1, 39 mol%) (Table 2). In Fig. 2, the product distributions in oil, aqueous, and gas phases are plotted as a function of reaction temperature. The result showed that a blank reaction without a catalyst mainly produced fatty acids in the oil phase (Fig. 2a) and glycerol in the aqueous phase (Fig. 2b) due to the hydrolysis of triglycerides. No detectable amount of the deoxygenated products of fatty acids (i.e., n-paraffins) were produced. The fatty acid composition in an oil phase was very similar to that of initial palm oil (Table 2), indicating that hydrogenation of unsaturated fatty acids did not take place in the absence of a catalyst. In the presence of Pt/CNT, the yields of unsaturated fatty acids decreased gradually while those of saturated fatty acids increased up to 523 K (Fig. 2d). Meanwhile, the yield of glycerol decreased and an enhanced formation of smaller oxygenates (e.g., ethylene glycol and acetic acids) was observed (Fig. 2e). This result indicated that the glycerol was decomposed into smaller oxygenates via APR to produce H2 in situ, which was subsequently used in the hydrogenation of unsaturated fatty acids. Above 523 K, n-paraffins (C15–C18) were formed due to the deoxygenation of fatty acids. The produced n-paraffins at 558 K were mostly composed of C15 (43 mol%) and C17 (53 mol%), and the sum of the portions of C16 and C18 was less than 4 mol%. Considering that the initial palm oil was mainly composed of C16 and C18 fatty acids (Table 2), the major formation of C15 and C17 n-paraffins with one less
3. Results and discussion 3.1. Synergetic effects of Pt and Re in hydrothermal deoxygenation We synthesized three different metal catalysts supported on CNT, namely, Pt/CNT (2.5 wt% Pt), Re/CNT (2.5 wt% Re), and PtRe/CNT (2.5 wt% Pt/1.0 wt% Re). TEM images and the metal particle size distributions of the catalysts are shown in Fig. 1. The surface-area-
Fig. 1. High-resolution TEM images of (a) Pt/CNT, (b) Re/CNT, and (c) PtRe/CNT. 3
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
carbon atom indicated that the deoxygenation of fatty acids occurred dominantly via DCO (the removal of oxygen by CO/CO2) rather than HDO (the removal of oxygen by H2O). Even though Pt catalysts are generally known to prefer DCO to HDO [14–16,43–46], such drastic suppression of HDO is uncommon in gas phase reactions [43,44] or liquid phase reactions using organic solvents [14–16,45,46]. The present results can be explained by the fact that the use of water as a solvent suppressed the HDO pathway producing H2O in accordance with Le Chatelier's principle. In the case of the Re/CNT catalyst, a significant amount of unsaturated fatty acids still remained even at elevated temperatures (> 523 K) and the formation of n-paraffins was also insubstantial
Table 2 Fatty acid composition of palm oil. Fatty acid
Composition (mol%) a
Myristic acid (C14:0 ) Palmitic acid (C16:0) Stearic acid (C18:0) Oleic acid (C18:1) Linoleic acid (C18:2) Avg. # of C = C bonds in a fatty acid
2 45 5 39 9 0.57
a In the notation “Cx:y”, x is the carbon length and y is the number of unsaturated C]C double bonds in fatty acids.
Fig. 2. Product distributions in oil, aqueous, and gas phases after hydrothermal deoxygenation at different reaction temperatures for 24 h, without a catalyst (a–c) and with Pt/CNT (d–f), Re/CNT (g–i), and PtRe/CNT (j–l) catalysts. 4
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
(Fig. 2g). In the aqueous phase, the decomposition of glycerol was also much slower compared to that with the Pt/CNT catalyst (Fig. 2h). The result indicated that the monometallic Re catalyst shows very limited activity in the hydrothermal deoxygenation of triglycerides. The bimetallic PtRe/CNT catalyst showed significantly enhanced hydrogenation and deoxygenation activities, when compared to the Pt/ CNT catalyst (Fig. 2j). All the produced fatty acids were fully saturated even at mild temperatures (e.g., 473 K) and most of fatty acids were deoxygenated to n-paraffin at 558 K. The n-paraffin amount in the oil phase at 558 K (88 C%) was significantly greater than that of Pt/CNT (16 C%). As in the case of Pt/CNT catalyst, the produced n-paraffins contained mostly C15 (49 mol%) and C17 (47 mol%), and the sum of the portions of C16 and C18 was less than 2 mol%. This result again indicated that the deoxygenation of fatty acids occurred dominantly via DCO rather than HDO in hydrothermal conditions. In aqueous phase, the PtRe/CNT catalyst showed the fastest decomposition of glycerol with increasing reaction temperature, which revealed the highest APR activity among all catalysts (Fig. 2k). The present results indicated that PtRe/CNT has superior catalytic activity compared to Pt/CNT in the hydrothermal deoxygenation of triglycerides. The faster decomposition of glycerol and the saturation of fatty acids with PtRe/CNT can be explained by its enhanced APR activity. In Fig. 3, the molar amount of H2 produced per mol of glycerol was plotted for all catalysts. The amount of H2 produced was calculated by summing the amount of H2 in gas phase with that used in the hydrogenation of unsaturated fatty acids. We postulated that the deoxygenation step does not require H2, because the major reaction pathway turned out to be DCO rather than HDO. In DCO mechanisms, decarboxylation does not intrinsically require H2, while decarbonylation requires one H2 molecule for producing saturated n-paraffin. However, the CO produced from decarbonylation can be subsequently converted to H2 via a water-gas shift (WGS) reaction, thereby not requiring a net H2 consumption (i.e., the result of decarbonylation combined with WGS is the same as decarboxylation). This can be verified by the fact that CO was not detected in gas phase, while a substantial amount of CO2 was detected (Fig. 2f, i, and l). The H2 production per mol of glycerol increased in the order of Re/CNT < Pt/CNT < PtRe/CNT (Fig. 3). At 558 K, the PtRe/CNT catalyst produced 3.8 mol of H2 per mol of glycerol, while Pt/CNT produced 2.5 mol of H2. In principle, 7 mol of H2 can be produced from 1 mol of glycerol. The H2 production yield lower than the theoretical yield can be attributed to the incomplete conversion of oxygenate intermediates as well as substantial CeO cleavage instead of CeC cleavage [25–32], which could be confirmed by the formation of C1–C3 hydrocarbons in gas phase (Fig. 2). However, considering that
1.71 mol of H2 is required for saturating all the C]C bonds in 1 mol of palm oil (0.57 × 3, Table 2), PtRe/CNT already produced sufficient amounts of the H2 required for hydrothermal deoxygenation. These results confirm that hydrothermal deoxygenation can enable net-zero H2 consumption in the production of paraffins from triglycerides. According to our previous studies [16], the deoxygenation of 1 mol of palm oil via conventional hydroprocessing required the consumption of 5.3 mol of external H2. The reasons for the superior APR activity of PtRe catalysts have been extensively investigated in previous studies [24–32]. A general consensus is that Re weakens CO binding to nearby Pt atoms and reduces the CO coverage on active Pt surfaces [24,25,28,29], which inhibits the binding of other reactants and reaction intermediates. Since cleavage of CeC bond in oxygenated hydrocarbons requires Pt ensembles [30], the coverage of Pt surface even with a trace amount of CO can significantly suppress the APR activities. To investigate the interaction of CO with the Pt/CNT and PtRe/CNT catalysts, diffuse reflectance infrared fourier transform (DRIFT) spectra were collected after CO adsorption (Fig. 4). All DRIFT spectra showed only the stretching band for linearly adsorbed CO (1980–2080 cm−1), while no band for bridging CO (1900–1970 cm−1) was detected. The intensity of the CO stretching band of PtRe/CNT (Fig. 4b) was much smaller than that for Pt/CNT (Fig. 4a), and the former also decreased more rapidly with an increasing desorption temperature. PtRe/CNT also showed a red shift as compared to Pt/CNT. This can be explained by reduced dipole-dipole coupling due to decreased surface CO coverage [47]. These results supported the weakened interaction of CO with PtRe/CNT
Fig. 3. Molar amount of H2 produced per mol of glycerol during the hydrothermal deoxygenation of palm oil for 24 h.
Fig. 4. CO DRIFT spectra of (a) Pt/CNT and (b) PtRe/CNT collected after desorption at different temperatures (293, 323, 373, 473, and 573 K). 5
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
Table 3 Properties of PtRe catalysts supported on various carbon supports. Sample
SBETa (m2 g−1)
Vtotalb (mL g−1)
H2 uptakec (μmol g−1)
CO uptakec (μmol g−1)
O contentd (%)
PtRe/CNT PtRe/oCNT PtRe/AC PtRe/oAC
171 172 977 613
0.344 0.349 0.835 0.607
26.2 23.2 7.5 6.4
55.9 60.3 14.6 13.3
0.8 3.3 10.8 16.2
a b c d
BET surface areas (SBET) were determined in the P/P0 range of 0.1–0.3. Total pore volumes (Vtotal) were determined at P/P0 = 0.98. H2 and CO uptakes were determined by chemisorption at 323 K. O contents were analyzed by O 1s X-ray photoelectron spectroscopy (XPS).
Fig. 5. n-Paraffin yields after the deoxygenation of stearic acid at different reaction temperatures under 10 bar of H2 for 24 h.
as compared to Pt/CNT. It is also known that the addition of Re also increases the hydrogenolysis (more substantially the rate of CeO bond cleavage than that of CeC bond cleavage) [48] as well as WGS activities [31,49–51], which are both beneficial for achieving enhanced APR activities. Several kinetic studies [31,49,50] and DFT calculations [50,51] indicated that Re addition facilitates water dissociation (H2O → OH* + 1/2H2), which is an important step of WGS. The resultant ReeOH* species neighboring Pt can facilitate the formation of carboxyl intermediate (COOH*) and its further decomposition (COOH* + OH* → CO2 + H2O). The enhanced formation of n-paraffins with PtRe/CNT (Fig. 2j) indicated that the bimetallic catalyst had higher deoxygenation activity than the monometallic catalysts. To investigate the deoxygenation activities without the influence of an APR step, the deoxygenation of stearic acid (C18:0) was additionally carried out in water under an H2 atmosphere (Fig. 5). As expected, PtRe/CNT showed much higher activity than Pt/CNT in the deoxygenation of stearic acid. This can be explained by the higher oxygen affinity of Re than that of Pt [49], which facilitates the binding of fatty acids on the catalyst surface via a ReeOeC bond formation.
Fig. 6. High-resolution TEM images of (a) PtRe/CNT, (b) PtRe/oCNT, (c) PtRe/ AC, and (d) PtRe/oAC.
The PtRe catalysts were analyzed using Pt L3 and Re L3 X-ray absorption spectroscopy (XAS). The X-ray absorption near edge structures (XANES) are shown in Fig. 7a. Earlier studies demonstrated a linear correlation between the metal oxidation states and the energy shift of the absorption edge in XANES [25]. Such analysis revealed that Pt is always highly reduced, and its oxidation state is ∼0.15 regardless of the type of carbon support (Fig. S1). On the other hand, the variation of Re oxidation state is more substantial (1.0–1.9), wherein Re becomes more oxidized with increasing oxygen content of a carbon support (Fig. 7b). Extended X-ray absorption fine structure (EXAFS) indicated that Pt-M coordination increased slightly as the oxygen content of a carbon support increased (Table 4). On the other hand, Re-M coordination increased rapidly along with Re-O coordination. These results indicated that Pt always exists as a highly dispersed and reduced species, while Re becomes segregated into large oxidized particles as the oxygen content of a carbon support increases. This means that Pt-Re alloying becomes gradually less efficient with increasing oxygen content of a carbon surface. Temperature programmed reduction (TPR) also supported this conclusion. Pt/CNT showed one reduction peak at 523 K (Fig. S2), while Re/CNT showed a reduction peak at a significantly higher temperature (573 K). The PtRe/CNT (Fig. S2) catalyst showed a very similar TPR profile to that of Pt/CNT, indicating that Pt and Re were well alloyed and hydrogen activated on the surface of Pt helped the reduction of nearby Re atoms. However, as the oxygen content of the carbon support increased (PtRe/oCNT < PtRe/AC <
3.2. Effects of carbon supports on bimetallic PtRe catalysts In previous researches on hydrothermal deoxygenation, only activated carbons (AC) were used for supporting metal catalysts [18–21]. However, carbons can have a wide range of catalytic properties depending on their structures [35–38] and especially the surface oxygen functional groups can substantially affect the alloying state of bimetallic catalysts [38]. In this regard, four different carbons including CNT, oCNT (oxidized CNT), AC, and oAC (oxidized AC) were investigated for supporting PtRe catalysts. Table 3 summarizes the physical properties and H2/CO chemisorption properties of the catalysts. CNT contained much lower oxygen content (0.8%) than AC (10.8%) due to its highly crystalline structure. The oxygen contents of oCNT (3.3%) and oAC (16.2%) were substantially increased compared to those of pristine CNT and AC, indicating that the HNO3 treatment oxidized the carbon surfaces. The H2 and CO chemisorption amounts on the PtRe catalysts generally decreased with increasing oxygen content of the carbon support (i.e., PtRe/CNT > PtRe/oCNT > PtRe/AC > PtRe/oAC). This indicated that the size of PtRe particles became gradually larger as the oxygen content of carbon support increased. TEM investigation (Fig. 6) also confirmed that the metal particle diameter (dTEM) increased in the order PtRe/CNT (1.79 nm) < PtRe/oCNT (2.35 nm) < PtRe/AC (2.80 nm) < PtRe/oAC (8.27 nm). 6
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
Fig. 8. n-Paraffin yield after the hydrothermal deoxygenation of palm oil at 558 K over the PtRe catalysts supported on various carbon supports.
cause selective segregation of Re species and inefficient Pt-Re alloying. Hydrothermal deoxygenation of palm oil over the PtRe catalysts supported on CNT, oCNT, AC, and oAC was carried out at 558 K. The yields of n-paraffin are shown in Fig. 8. Higher n-paraffin yields were achieved with the PtRe catalyst supported on the carbons with less oxygen content (i.e., PtRe/oAC < PtRe/AC < PtRe/oCNT < PtRe/ CNT). Based on the aforementioned structural analyses, it can be concluded that well alloyed and highly dispersed PtRe particles supported on oxygen-deficient carbon supports (e.g., CNT) are more active in the hydrothermal deoxygenation. The n-paraffin yield with respect to the initial weight of the palm oil was 72 wt% over PtRe/CNT, which was quite close to the theoretical n-paraffin yield (79 wt%). The theoretical yield was calculated by assuming that glycerol was fully removed, and all fatty acids were saturated and deoxygenated solely via a DCO pathway.
Fig. 7. (a) XANES of the PtRe catalysts supported on different carbon supports and (b) the oxidation state of Re calculated using the energy shift of the absorption edge.
3.3. Recyclability of PtRe/CNT in the hydrothermal deoxygenation of palm oil
Table 4 EXAFS fitting results of the PtRe catalysts supported on various carbons. Sample
Shell
CN
PtRe/CNT
Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2 Pt-M Re-M Re-O1 Re-O2
7.17 4.36 0.14 0.06 7.91 5.38 0.38 0.33 7.99 6.44 0.88 0.66 8.01 9.12 1.67 0.71
PtRe/oCNT
PtRe/AC
PtRe/oAC
σ2 (Å2)
R (Å) ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.32 0.41 0.05 0.02 0.42 0.78 0.08 0.04 0.62 0.56 0.21 0.14 0.72 1.21 0.21 0.18
2.74 2.59 1.53 1.99 2.74 2.61 1.72 1.93 2.75 2.60 1.83 2.12 2.75 2.64 1.69 2.05
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.02 0.04 0.07 0.02 0.04 0.07 0.02 0.09 0.03 0.06 0.09 0.02 0.04 0.06 0.03 0.09
0.007 0.009 0.005 0.011 0.007 0.012 0.024 0.017 0.007 0.010 0.005 0.004 0.005 0.028 0.007 0.010
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
Catalyst recyclability in the hydrothermal deoxygenation of palm oil was investigated using PtRe/CNT, which showed the best performance among the prepared catalysts. Prior to the recycling experiments, we first analyzed the used catalyst by thermogravimetric analysis (TGA; Fig. 9a). The catalyst collected after the reaction showed the formation of light carbonaceous residue (14 wt%). This can be explained by the polymerization of unsaturated fatty acids via the Diels-Alder reaction and/or a radical-mediated addition at elevated temperatures [45]. Fortunately, due to the high thermal stability of CNT, such polymeric species could be selectively removed by calcination in the temperature range of 473–673 K (Fig. 9a). It is notable that, in the case of AC with low thermal stability, such selective removal of polymeric species is not possible (Fig. S4) because the combustion of AC occurs at a similar temperature regime. Based on the TGA data, we carried out catalyst recycling experiments with and without calcination at 573 K to remove the polymeric species between each reaction cycle (Fig. 9b). Without calcination, the yield of n-paraffin significantly decreased from 72 wt% to 14 wt% after only 3 reaction cycles. With calcination, however, the yield of n-paraffin decreased only slightly from 72 wt% to 66 wt% after 5 cycles. These results showed that the catalyst deactivation is mainly caused by the deposition of polymeric species on the metal surfaces, and thus their removal by mild calcination is necessary for effectively recycling a catalyst. TEM investigation of the PtRe/CNT catalysts after 5 reaction cycles (Fig. S5) indicated a slight increase of metal particle size (dTEM:
0.002 0.001 0.001 0.002 0.001 0.001 0.004 0.002 0.002 0.001 0.002 0.001 0.001 0.004 0.002 0.001
PtRe/oAC), broad peaks were clearly observed above 523 K (Fig. S3). This result supported less efficient Pt-Re alloying on the oxygen-rich carbon surfaces. We believe that the highly oxophilic Re species interact strongly with the oxygen functional groups of the carbon surface, leading to segregation of Re nearby these groups. On the other hand, Pt species interact less strongly with these functional groups, and thus, could be distributed uniformly over the carbon surface. Therefore, the high density of oxygen functional groups on the carbon surface might 7
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi
favorable for forming a highly dispersed and well alloyed PtRe catalyst, which is the prerequisite for achieving high hydrothermal deoxygenation activity. An oxygen-rich carbon surface (e.g., oxidized AC) led to the segregation of Re species into bulk oxidized particles, which led to low catalytic activity. The PtRe/CNT showed the best performance, with a paraffin yield of 72 wt% (theoretical yield: 79 wt%) from palm oil, even in the absence of an external H2 source. The PtRe catalyst deactivated mainly via the deposition of polymerized unsaturated fatty acids, and no significant metal sintering/leaching was observed even under the harsh hydrothermal conditions. In the case of PtRe supported on CNT with high thermal stability, the polymerized unsaturated fatty acids could be selectively removed by calcination at a mild temperature (573 K), and the catalyst could be efficiently reused for 5 cycles without significant loss of activity. Because of its high activity and recyclability, PtRe/CNT is a highly promising catalyst for the hydrothermal deoxygenation of triglycerides. Conflicts of interest There are no conflicts to declare. Acknowledgements This work was supported by the Advanced Biomass R&D Center (ABC) of Global Frontier Project funded by the Ministry of Science, ICT, and Future Planning (ABC-2015M3A6A2066121) and the Basic Science Research Program through the National Research Foundation of Korea (NRF-2017R1A2B2002346). The authors thank the Pohang Accelerator Laboratory (PAL) for the use of the beamline. Appendix A. Supplementary data Fig. 9. (a) Thermogravimetric analysis of fresh PtRe/CNT and used PtRe/CNT catalysts in air. (b) Recycling experiments of the PtRe/CNT catalyst with and without calcination (calcination temperature: 573 K) between each reaction cycle.
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.mcat.2019.110419. References
2.17 nm) compared to the fresh sample (dTEM: 1.79 nm). We also performed additional experiments for investigating metal leaching. The PtRe/CNT catalyst was repeatedly treated in pure H2O at 573 K for 2 h and the concentrations of Pt and Re in H2O were analyzed by ICP-OES. The results showed that none of the Pt was leached out during the hydrothermal treatments. On the other hand, 10% of Re in the fresh catalyst was leached out after the first treatment. However, Re leaching from the second treatment was not detectable. This result indicated the hydrothermal stability of PtRe bimetallic particles, which is consistent with an earlier report [29] that unalloyed Re species were readily leached out, but Re species alloyed with Pt were highly stable. The slight sintering of metal catalysts and/or minor leaching of Re species might be the reason for the small decrease of paraffin yield with repeated reaction cycles.
[1] G.W. Huber, A. Corma, Synergies between bio- and oil refineries for the production of fuels from biomass, Angew. Chem. Int. Ed. 46 (2007) 7184–7201, https://doi. org/10.1002/anie.200604504. [2] H. Baumann, M. Bühler, H. Fochem, F. Hirsinger, H. Zoebelein, J. Falbe, Natural fats and oils-renewable raw materials for the chemical industry, Angew. Chem. Int. Ed. Engl. 27 (1988) 41–62, https://doi.org/10.1002/anie.198800411. [3] A. Srivastava, R. Prasad, Triglyceride-based diesel fuels, Renew. Sustain. Energy Rev. 4 (2000) 111–133, https://doi.org/10.1016/S1364-0321(99)00013-1. [4] L.C. Meher, D.V. Sagar, S.N. Naik, Technical aspects of biodiesel production by transesterification-a review, Renew. Sustain. Energy Rev. 10 (2006) 248–268, https://doi.org/10.1016/j.rser.2004.09.002. [5] G. Knothe, Biodiesel and renewable diesel: a comparison, Prog. Energy Combust. Sci. 36 (2010) 364–373, https://doi.org/10.1016/j.pecs.2009.11.004. [6] M. Lapuerta, O. Armas, J. Rodríguez-Fernández, Effect of biodiesel fuels on diesel engine emissions, Prog. Energy Combust. Sci. 34 (2008) 198–223, https://doi.org/ 10.1016/j.pecs.2007.07.001. [7] G. Dwivedi, M.P. Sharma, Impact of cold flow properties of biodiesel on engine performance, Renew. Sustain. Energy Rev. 31 (2014) 650–656, https://doi.org/10. 1016/j.rser.2013.12.035. [8] M. Stumborg, A. Wong, E. Hogan, Hydroprocessed vegetable oils for diesel fuel improvement, Bioresour. Technol. 56 (1996) 13–18, https://doi.org/10.1016/ 0960-8524(95)00181-6. [9] R. Sotelo-Boyás, Y. Liu, T. Minowa, Renewable diesel production from the hydrotreating of rapeseed oil with Pt/Zeolite and NiMo/Al2O3 catalysts, Ind. Eng. Chem. Res. 50 (2011) 2791–2799, https://doi.org/10.1021/ie100824d. [10] B. Veriansyah, J.Y. Han, S.K. Kim, S.A. Hong, Y.J. Kim, J.S. Lim, Y.W. Shu, S.G. Oh, J. Kim, Production of renewable diesel by Hydroprocessing of soybean oil: effect of catalysts, Fuel 94 (2012) 578–585, https://doi.org/10.1016/j.fuel.2011.10.057. [11] B. Peng, X. Yuan, C. Zhao, J.A. Lercher, Stabilizing catalytic pathways via redundancy: selective reduction of microalgae oil to Alkanes, J. Am. Chem. Soc. 134 (2012) 9400–9405, https://doi.org/10.1021/ja302436q. [12] B. Peng, Y. Yao, C. Zhao, J.A. Lercher, Towards quantitative conversion of microalgae oil to diesel-range alkanes with bifunctional catalysts, Angew. Chem. Int. Ed. 51 (2012) 2072–2075, https://doi.org/10.1002/anie.201106243. [13] L. Boda, G. Onyestyák, H. Solt, F. Lónyi, J. Valyon, A. Thernesz, Catalytic hydroconversion of Tricaprylin and caprylic acid as model reaction for biofuel production from triglycerides, Appl. Catal. A Gen. 374 (2010) 158–169, https://doi.org/10.
4. Conclusions We investigated bimetallic PtRe catalysts supported on various carbon supports in the hydrothermal deoxygenation of triglycerides. Bimetallic PtRe catalysts showed a markedly enhanced rate of glycerol reforming for in situ H2 production compared to the Pt catalysts, which can be explained by the weakened interaction of CO with PtRe. The PtRe catalysts also exhibited a significantly enhanced rate for the deoxygenation of fatty acids, which could be attributed to the high oxygen affinity of Re. Consequently, PtRe catalysts showed superior performance in the hydrothermal deoxygenation of triglycerides than monometallic catalysts. Investigations using four different carbon supports (CNT and AC with and without surface oxidation) showed that the carbon support with low surface oxygen content (e.g., CNT) was 8
Molecular Catalysis 474 (2019) 110419
M. Jin and M. Choi 1016/j.apcata.2009.12.005. [14] A.T. Madsen, E.H. Ahmed, C.H. Christensen, R. Fehrmann, A. Riisager, Hydrodeoxygenation of waste fat for diesel production: study on model feed with Pt/alumina catalyst, Fuel 90 (2011) 3433–3438, https://doi.org/10.1016/j.fuel. 2011.06.005. [15] K. Murata, Y. Liu, M. Inaba, I. Takahara, Production of synthetic diesel by Hydrotreatment of Jatropha oils using Pt-Re/H-ZSM-5 catalyst, Energy Fuels 24 (2010) 2404–2409, https://doi.org/10.1021/ef901607t. [16] M.Y. Kim, J.K. Kim, M.E. Lee, S. Lee, M. Choi, Maximizing biojet fuel production from triglyceride: importance of the hydrocracking catalyst and separate deoxygenation/hydrocracking steps, ACS Catal. 7 (2017) 6256–6267, https://doi.org/ 10.1021/acscatal.7b01326. [17] A. Anumakonda, M.J. McCall, T.A. Brandvold, J.A. Kocal, Controlling Production of Transportation Fuels from Renewable Feedstocks. U. S. Patent 8,058,492, November 15, 2011. [18] J. Fu, X. Lu, P.E. Savage, Catalytic hydrothermal deoxygenation of palmitic acid, Energy Environ. Sci. 3 (2010) 311–317, https://doi.org/10.1039/B923198F. [19] F. Fu, X. Lu, P.E. Savage, Hydrothermal decarboxylation and hydrogenation of fatty acids over Pt/C, ChemSusChem 4 (2011), https://doi.org/10.1002/cssc.201000370 481-186. [20] S.A.W. Hollak, M. Ariëns, K.P. de Jong, D.S. van Es, Hydrothermal deoxygenation of triglycerides over Pd/C aided by in situ hydrogen production from glycerol reforming, ChemSusChem 7 (2014) 1057–1060, https://doi.org/10.1002/cssc. 201301145. [21] D.R. Vardon, B.K. Sharma, H. Jaramillo, D. Kim, J.K. Choe, P.N. Ciesielski, T.J. Strathmann, Hydrothermal catalytic processing of saturated and unsaturated fatty acids to hydrocarbons with glycerol for in situ hydrogen production, Green Chem. 16 (2014) 1507–1520, https://doi.org/10.1039/C3GC41798K. [22] D. Kim, D.R. Vardon, D. Murali, B.K. Sharma, T.J. Strathmann, Valorization of waste lipids through hydrothermal catalytic conversion to liquid hydrocarbon fuels with in situ hydrogen production, ACS Sustain. Chem. Eng. 4 (2016) 1775–1784, https://doi.org/10.1021/acssuschemeng.5b01768. [23] C. Miao, O. Marin-Flores, T. Dong, D. Gao, Y. Wang, M. Garcia-Pérez, S. Chen, Hydrothermal catalytic deoxygenation of fatty acid and bio-oil with in situ H2, ACS Sustain. Chem. Eng. 6 (2018) 4521–4530, https://doi.org/10.1021/acssuschemeng. 7b02226. [24] D.A. Simonetti, E.L. Kunkes, J.A. Dumesic, Gas-phase conversion of glycerol to synthesis gas over carbon-supported platinum and platinum-rhenium catalysts, J. Catal. 247 (2007) 298–306, https://doi.org/10.1016/j.jcat.2007.01.022. [25] E.L. Kunkes, D.A. Simonetti, J.A. Dumesic, W.D. Pyrz, L.E. Murillo, J.G. Chen, D.J. Buttrey, The role of rhenium in the conversion of glycerol to synthesis gas over carbon supported platinum-rhenium catalysts, J. Catal. 260 (2008) 164–177, https://doi.org/10.1016/j.jcat.2008.09.027. [26] D.L. King, L. Zhang, G. Xia, A.M. Karim, D.J. Heldebrant, X. Wang, T. Peterson, Y. Wang, Aqueous phase reforming of glycerol for hydrogen production over Pt-Re supported on carbon, Appl. Catal. B 99 (2010) 206–213, https://doi.org/10.1016/j. apcatb.2010.06.021. [27] L. Zhang, A.M. Karim, M.H. Engelhard, Z. Wei, D.L. King, Y. Wang, Correlation of Pt-Re surface properties with reaction pathways for the aqueous-phase reforming of glycerol, J. Catal. 287 (2012) 37–43, https://doi.org/10.1016/j.jcat.2011.11.015. [28] Z. Wei, A.M. Karim, Y. Li, D.L. King, Y. Wang, Elucidation of the roles of Re in steam reforming of glycerol over Pt-Re/C catalysts, J. Catal. 322 (2015) 49–59, https:// doi.org/10.1016/j.jcat.2014.11.006. [29] Z. Wei, A.M. Karim, Y. Li, Y. Wang, Elucidation of the roles of Re in aqueous-phase reforming of glycerol over Pt-Re/C catalysts, ACS Catal. 5 (2015) 7312–7320, https://doi.org/10.1021/acscatal.5b01770. [30] A. Ciftci, D.A.J.M. Ligthart, E.J.M. Hensen, Influence of Pt particle size and Re addition by catalytic reduction on aqueous phase reforming of glycerol for carbonsupported Pt(Re) catalysts, Appl. Catal. B 174 (2015) 126–135, https://doi.org/10. 1016/j.apcatb.2015.02.027. [31] A. Ciftci, D.A.J.M. Ligthart, A.O. Sen, A.J.F. van Hoof, H. Friedrich, E.J.M. Hensen, Pt-Re synergy in aqueous-phase reforming of glycerol and the water-gas shift reaction, J. Catal. 311 (2014) 88–101, https://doi.org/10.1016/j.jcat.2013.11.011. [32] A. Ciftci, S. Eren, D.A.J.M. Ligthart, E.J.M. Hensen, Platinum-rhenium synergy on
[33]
[34] [35] [36] [37]
[38]
[39] [40] [41]
[42] [43] [44]
[45]
[46] [47] [48] [49] [50]
[51]
9
reducible oxide supports in aqueous-phase glycerol reforming, ChemCatChem 6 (2014) 1260–1269, https://doi.org/10.1002/cctc.201301096. H.N. Pham, A.E. Anderson, R.L. Johnson, K. Schmidt-Rohr, A.K. Datye, Improved Hydrothermal Stability of Mesoporous Oxides for Reactions in the Aqueous Phase, Angew. Chem. Int. Ed. 51 (2012) 13163–13167, https://doi.org/10.1002/anie. 201206675. A. Ciftci, B. Peng, A. Jentys, J.A. Lercher, E.J.M. Hensen, Support effects in the aqueous phase reforming of glycerol over supported platinum catalysts, Appl. Catal. A Gen. 431 (2012) 113–119, https://doi.org/10.1016/j.apcata.2012.04.026. F. Rodríguez-Reinoso, The role of carbon materials in heterogeneous catalysis, Carbon 36 (1998) 159–175, https://doi.org/10.1016/S0008-6223(97)00173-5. D.S. Su, S. Perathoner, G. Centi, Nanocarbons for the development of advanced catalysts, Chem. Rev. 113 (2013) 5782–5816, https://doi.org/10.1021/cr300367d. H.C. Kwon, S. Yook, S. Choi, M. Choi, Comprehensive understanding of the effects of carbon nanostructures on redox catalytic properties and stability in oxidative dehydrogenation, ACS Catal. 7 (2017) 5257–5267, https://doi.org/10.1021/ acscatal.7b01742. S. Yook, H.C. Kwon, Y.G. Kim, W. Choi, M. Choi, Significant roles of carbon pore and surface structure in AuPd/C catalyst for achieving high chemoselectivity in direct hydrogen peroxide synthesis, ACS Sustain. Chem. Eng. 5 (2017) 1208–1216, https://doi.org/10.1021/acssuschemeng.6b02595. M. Choi, Z. Wu, E. Iglesia, Mercaptosilane-assisted synthesis of metal clusters within zeolites and catalytic consequences of encapsulation, J. Am. Chem. Soc. 132 (2010) 9129–9137, https://doi.org/10.1021/ja102778e. H. Hirai, K. Wada, M. Komiyama, Active carbon-supported copper (I) chloride as solid adsorbent for carbon monoxide, Bull. Chem. Soc. Jpn. 59 (1986) 2217–2223, https://doi.org/10.1246/bcsj.59.2217. S.R. Bare, S.D. Kelly, F.D. Vila, E. Boldingh, E. Karapetrova, J. Kas, G.E. Mickelson, F.S. Modica, N. Yang, J.J. Rehr, Experimental (XAS, STEM, TPR, and XPS) and theoretical (DFT) characterization of supported rhenium catalysts, J. Phys. Chem. C 115 (2011) 5740–5755, https://doi.org/10.1021/jp1105218. D. Darnoko, M. Cheryan, Kinetics of palm oil transesterification in a batch reactor, J. Am. Oil Chem. Soc. 77 (2000) 1263–1267, https://doi.org/10.1007/s11746-0000198-y. W. Rachmady, M.A. Vannice, Acetic acid hydrogenation over supported platinum catalysts, J. Catal. 192 (2000) 322–334, https://doi.org/10.1006/jcat.2000.2863. Y.K. Lugo-José, J.R. Monnier, C.T. Williams, Gas-phase, catalytic hydrodeoxygenation of propanoic acid, over supported group VIII noble metals: metal and support effects, Appl. Catal. A Gen. 469 (2014) 410–418, https://doi.org/10. 1016/j.apcata.2013.10.025. T.H. Kim, K. Lee, M.Y. Kim, Y.K. Chang, M. Choi, Effects of fatty acid compositions on heavy oligomer formation and catalyst deactivation during deoxygenation of triglycerides, ACS Sustain. Chem. Eng. 6 (2018) 17168–17177, https://doi.org/10. 1021/acssuschemeng.8b04552. L. Yang, K.L. Tate, J.B. Jasinski, M.A. Carreon, Decarboxylation of oleic acid to Heptadecane over Pt supported on zeolite 5A beads, ACS Catal. 5 (2015) 6497–6502, https://doi.org/10.1021/acscatal.5b01913. J. Raskó, CO-induced surface structural changes of Pt on oxide-supported Pt catalysts studied by DRIFTS, J. Catal. 217 (2003) 478–486, https://doi.org/10.1016/ S0021-9517(03)00147-7. J. Zhang, M.B. Vukmirovic, K. Sasaki, A.U. Nilekar, M. Mavrikakis, R.R. Adzic, Mixed-metal Pt monolayer electrocatalysts for enhanced oxygen reduction kinetics, J. Am. Chem. Soc. 127 (2005) 12480–12481, https://doi.org/10.1021/ja053695i. K.G. Azzam, I.V. Babich, K. Seshan, L. Lefferts, Role of Re in Pt-Re/TiO2 catalyst for water gas shift reaction: a mechanistic and kinetic study, Appl. Catal. B 80 (2008) 129–140, https://doi.org/10.1016/j.apcatb.2007.11.015. R. Carrasquillo-Flores, J.M.R. Gallo, K. Hahn, J.A. Dumesic, M. Mavrikakis, Density functional theory and reaction kinetics studies of the water-gas shift reaction on PtRe catalysts, ChemCatChem 5 (2013) 3690–3699, https://doi.org/10.1002/cctc. 201300365. L.C. Grabow, A.A. Gokhale, S.T. Evans, J.A. Dumesic, M. Mavrikakis, Mechanism of the water gas shift reaction on Pt: first principles, experiments, and microkinetic modeling, J. Phys. Chem. C 112 (2008) 4608–4617, https://doi.org/10.1021/ jp7099702.