Research Article pubs.acs.org/acscatalysis
Methane Upgrading of Acetic Acid as a Model Compound for a Biomass-Derived Liquid over a Modified Zeolite Catalyst Aiguo Wang,† Danielle Austin,† Abhoy Karmakar,‡ Guy M. Bernard,‡ Vladimir K. Michaelis,‡ Matthew M. Yung,§ Hongbo Zeng,∥ and Hua Song*,† †
Department of Chemical and Petroleum Engineering, University of Calgary, 2500 University Drive, NW, Calgary, Alberta T2N 1N4, Canada ‡ Department of Chemistry, University of Alberta, 11227 Saskatchewan Drive, Edmonton, Alberta T6G 2G2, Canada § National Bioenergy Center, National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, Colorado 80401, United States ∥ Department of Chemical and Materials Engineering, University of Alberta, 9211-116 Street NW, Edmonton, Alberta T6G 1H9, Canada S Supporting Information *
ABSTRACT: The technical feasibility of coaromatization of acetic acid derived from biomass and methane was investigated under mild reaction conditions (400 °C and 30 bar) over silver-, zinc-, and/or gallium-modified zeolite catalysts. On the basis of GC-MS, Micro-GC, and TGA analysis, more light aromatic hydrocarbons, less phenol formation, lower coke production, and higher methane conversion are observed over 5%Zn-1%Ga/ ZSM-5 catalyst in comparison with catalytic performance over the other catalysts. Direct evidence of methane incorporation into aromatics over 5%Zn-1%Ga/ZSM-5 catalyst is witnessed in 1 H, 2H, and 13C NMR spectra, revealing that the carbon from methane prefers to occupy the phenyl carbon sites and the benzylic carbon sites, and the hydrogen of methane favors the aromatic and benzylic substitutions of product molecules. In combination with the 13C NMR results for isotopically labeled acetic acid (13CH3COOH and CH313COOH), it can be seen that the methyl and carbonyl carbons of acetic acid are equally involved in the formation of ortho, meta and para carbons of the aromatics, whereas the phenyl carbons directly bonded with alkyl substituent groups and benzylic carbons are derived mainly from the carboxyl carbon of acetic acid. After various catalyst characterizations by using TEM, XRD, DRIFT, NH3-TPD, and XPS, the excellent catalytic performance might be closely related to the highly dispersed zinc and gallium species on the zeolite support, moderate surface acidity, and an appropriate ratio of weak acidic sites to strong acidic sites as well as the fairly stable oxidation state during acetic acid conversion under a methane environment. Two mechanisms of the coaromatization of acetic acid and methane have also been proposed after consulting all the collected data in this study. The results reported in this paper could potentially lead to more cost-effective utilization of abundant natural gas and biomass. KEYWORDS: methane, acetic acid, catalyst, aromatization, ZSM-5
1. INTRODUCTION Today’s society is highly dependent on energy from fossil fuels. In 2015 alone, the world consumed 4.362 billion tons of oil. Comparatively, just 0.365 billion tons of the oil equivalent of renewable energy was consumed.1 This heavy reliance on fossil fuels raises concerns about resources and economic stability as well as about the negative environmental impact. Due to these concerns, Canada, in addition to many other countries, is trying to reduce carbon emissions and dependence on fossil fuels. Therefore, it is desirable to find a renewable energy source that is cost competitive with traditional fuels. The upgrading of biomass is an attractive option because it is low cost and consumes waste while generating liquid fuels as well as valuable © 2017 American Chemical Society
chemicals. Biomass is the general term referring to organic matter which can include wood, plant matter, and organic municipal solid waste (MSW). One widely used method of converting biomass to bio-oil is fast pyrolysis. Fast pyrolysis is the decomposition of biomass at high temperatures that occurs in a few seconds in the absence of oxygen, producing gas, biooil, and char.2 Due to the high oxygen content in biomass, the bio-oil product is rich in oxygen-containing compounds, such as carboxylic acids, resulting in a low-quality fuel and incompatReceived: January 27, 2017 Revised: March 17, 2017 Published: April 19, 2017 3681
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis ibility with industrially used pipelines and storage units.3,4 Hydrodeoxygenation is the predominant method used to upgrade bio-oil.5,6 Due to the high operating pressure (usually 10−14 MPa) and substantial consumption of hydrogen, there is a high cost associated with hydroprocessing.78 Methane, as the main component in natural gas, is a lower cost, abundant alternative to the use of hydrogen for upgrading bio-oil. According to the U.S. Department of Energy, in July 2016 natural gas had an average cost of $2.05 per gasoline gallon equivalents (GGE) while hydrogen had an average price of $13.68/GGE.9 In a process called methanolysis, biomass undergoes fast pyrolysis and upgrading simultaneously, using methane as a proton donor at low pressures as opposed to hydrogen, which could result in the production of biofuels that are cost competitive with oil. In addition to upgrading the biooil quality, the produced aromatics have a wide range of uses, including pharmaceuticals, chemical applications (especially benzene and toluene), and dye production.10 Methane is not currently heavily utilized due to its low heating value and high stability, with a C−H bond activation energy of 413 kJ/mol.11 Much research has been done on the activation of CH4, and it has been shown that methane can be effectively activated through the nonoxidative pathway at low temperatures and atmospheric pressure over zeolite-supported catalysts.11−13 It has been demonstrated that methane can be more efficiently converted to aromatics in the presence of higher hydrocarbons or oxygenated hydrocarbons under mild conditions.13−17 It has also been shown that methane may be incorporated into the liquid aromatic product.17,18 Due to the complexity of bio-oil, it is necessary to utilize model compounds for a better understanding of the interaction between reagents and the catalyst. Acetic acid has been identified as a large component in bio-oil, reportedly making up 15−18% of crude bio-oil by dry weight depending on the feedstock and pyrolysis conditions. 19−21 As such, the elimination of acetic acid would greatly enhance fuel quality by reducing corrosiveness to infrastructure, improving the heating value, and increasing stability.2,22 Acetic acid in the biooil could be upgraded directly or extracted from the bio-oil and upgraded separately. There are many ways to extract acetic acid from bio-oil, one of which requires neutralization with calcium oxide to form precipitates followed by filtration and regeneration of the acid using sulfuric acid.23 Reactive liquid− liquid extraction can also be used to remove carboxylic acids from bio-oil. This method uses aliphatic tertiary amines as the extractant and is followed by hot water treatment for recovery of the acid.24 Nanofiltration has also been shown to be effective in the separation of acetic acid from bio-oil.25 These methods are neither economically nor environmentally friendly due to the production of large amounts of waste requiring subsequent treatment. Acetic acid has previously been utilized as a model compound for bio-oil in multiple studies, primarily for hydrodeoxygenation and steam re-forming.26−29 Chang et al. have previously demonstrated a synergistic effect of methanol on acetic acid aromatization over a zeolite support, providing the groundwork for this study.30 The ease of access to methane drives the motivation to determine if the same synergistic effect can be found from methane on the aromatization of acetic acid. In the present work, the performances of five catalysts, selected from their promising performance in our previous studies,18,31 were evaluated for the aromatization of acetic acid in a methane environment under relatively mild conditions.
The understanding gained from using acetic acid as a model compound can lead to the development of a novel catalyst for the upgrading of biomass that can produce a high-quality liquid product. Once this can be done economically, it could change the landscape of how we obtain chemicals and fuels, reducing dependence on fossil fuels and providing an environmentally superior option.
2. EXPERIMENTAL SECTION 2.1. Catalyst Preparation. The ammonium ZSM-5 zeolite with Si/Al = 23 and specific surface area of 425 m2 g−1 was purchased from Zeolyst and calcined at 600 °C for 5 h in air to attain the H-type ZSM-5 for further use. 5%Zn-1%Ga/ZSM-5 was prepared by incipient wetness impregnation of HZSM-5 with a Zn(NO3)2·6H2O (99%, Alfa Aesar) and Ga(NO3)3 (99.9%, Alfa Aesar) solution, dried in the oven at 88 °C overnight, followed by calcining at 600 °C for 3 h in ambient air. In a similar manner, 1%Ag-1%Ga/ZSM-5 was synthesized by using a 0.1 mol/L AgNO3 (99.9+%, Alfa Aesar) and 0.15 mol/L Ga(NO3)3 (99.9%, Alfa Aesar) solution. 5%Zn/ZSM-5 and 1%Ag/ZSM-5 were also prepared for the experiments. 2.2. Performance Evaluation. The acetic acid aromatization reaction was conducted using a 300 mL Parr reactor under batch mode. In a typical run, ∼1.0 g of catalyst and a glass vial filled with ∼3.0 g of acetic acid were loaded into the reactor. After the reactor passed a leak test, it was pressurized to 5 bar with reactive gas (i.e., methane) or nitrogen after the air inside the reactor was purged out. The reactor temperature was then ramped up with a rate of 20 °C/min to the target temperature (400 °C) and held for 40 min. Upon reaction completion, the reactor was cooled to room temperature before product collection. The formed liquid product embedded into the charged solid catalyst was extracted using 10 mL of CS2 (GC grade, EMD Chemicals) as the solvent. In a similar manner, isotopically labeled reactions between acetic acid and 13CH4/ CD4 (99.9% 13C and 99% 2H, respectively, Cambridge Isotope Laboratories Inc.) and between CH4 and 13CH3COOH/ CH313COOH (99% 13C, Sigma-Aldrich) were conducted in the same reactor. Following the same initial procedure, several experiments were also conducted with varying reaction times of 1, 3, 6, 9, and 20 min. The reactor was submerged in cold water to quench the reaction at the desired completion time, and the product was extracted as before. The liquid yields, liquid product selectivity, and gas product selectivity reported in this paper are given by the following equations: liquid yield =
weight of liquid collected in the reactor weight of acetic acid × 100%
liquid product selectivity weight of collected liquid product in the reactor = wieght of converted acetic acid × 100%
gas product selectivity weight of gas product = × 100% weight of converted acetic acid 3682
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis 1
All selectivity is calculated on the basis of acetic acid. Methane incorporation in the reaction will lead to a total selectivity of greater than 100%. 2.3. Sample Characterization. The adsorption and desorption of nitrogen on each catalyst was measured using a Quadrasorb SI instrument from Quantachrome Instruments. Samples were outgassed overnight under vacuum at 350 °C and then brought to 77 K via immersion in a liquid nitrogen bath. The total surface area was calculated using a multipoint Brunauer−Emmett−Teller (BET) analysis. Pore surface area and pore volume were calculated using a Barrett−Joyner− Halenda (BJH) analysis. The t-plot method was used with the DeBoer model for the calculation of the statistical thickness to distinguish the contribution of micropores ( alkyl. 3.2. Characterization. To improve our understanding of this catalytic system and the correlation between the physical properties of the catalysts and their catalytic performances during the aromatization of acetic acid, versatile characterizations including N2 physisorption, XRD, TEM, XPS, DRIFTS, NH3-TPD, and TGA have been performed. The results of N2 physisorption in Table S2 in the Supporting Information indicate that after metal loading the ratio of microporous surface area to total surface area remained constant, showing that during synthesis there was no pore blockage that would affect catalytic performance. The XRD patterns in Figure S1 in the Supporting Information indicate that these metal species are highly dispersed on the support surface or in the channel of the zeolite. Additionally, the crystal structure of the five studied catalysts and the zeolite framework are well maintained after the reaction. The morphology and particle size and metal distribution of the samples were further investigated using the TEM technique combined with electron energy-loss spectroscopy (EELS). As shown in Figures S2 and S3 in the Supporting Information, the dispersion of Ga species remained in both methane and nitrogen environments, while the Zn species were found to have agglomerated after reaction in a nitrogen environment. The features of surface acidity over various catalysts were studied using pyridine as the probe and NH3-TPD. The DRIFT spectra upon pyridine absorption depicted in Figure S4 in the Supporting Information and NH3TPD profiles presented in Figure S5 in the Supporting Information imply that the abundant surface acid sites and appropriate ratio of weak surface acidic sites to strong surface acidic sites may contribute to the superior performance of the
(Figure 8a,c) for insight into hydrogen incorporation during acetic acid aromatization. The peaks at 7.24 ppm in the 1H NMR spectra are due to the solvent CDCl3, the intensities of which are similar in Figure 8c,d. The peaks located between 7.1 and 7.6 ppm are due to the H atoms on the phenyl rings, while those in the 2.4−2.8 ppm region are assigned to benzylic hydrogen. The peaks appearing below 2 ppm are attributed to H sites of the alkyl groups that are not directly bonded to the phenyl rings. The peak area ratios with respect to that of CDCl3 are given in Table 2 as reference. The peak signals as a result of Table 2. 1H Liquid NMR Peak Area Ratio with Respect to CDCl3 of the Products from Acetic Acid Aromatization under CH4 and CD4 Environment type of proton
chem shift (ppm)
acetic acid + CH4
acetic acid + CD4
aromatic benzylic alkyl
7.15−7.55 2.44−2.58 1.5
11.52 19.30 2.76
9.22 15.21 2.57
aromatic and benzylic H sites decrease greatly for the isotopically labeled products, while the peak signals due to alkyl H change little, implying that the hydrogens of methane prefer to be incorporated into the aromatic and benzylic sites of product molecules. This is highly consistent with the appearance of two clusters of large peaks (Figure 8b) in the aromatic (7.0−7.6 ppm) and benzylic (2.4−2.8 ppm) hydrogen regions as well as a small peak at the alkyl (1.5 ppm) hydrogen region of the 2H NMR spectra. This coincides with the observation in the 13C NMR spectra that methane prefers to be incorporated into the benzene ring and benzylic carbon site. On the basis of the aforementioned 13C and 2H NMR analysis, it is easily concluded that, during the aromatization of acetic acid
Figure 9. XPS spectra of 5%Zn-1%Ga/ZSM-5 before and after reactions under different environments in the (a) Ga 2p, (b) Zn 2p, (c) O 1s, and (d) C 1s regions. 3687
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis
the catalyst and a masking effect of the surface O signal from SiO2. Finally, there are no visible additional O 1s shoulders corresponding to the presence of oxygen-containing organic compounds witnessed around the single peak over the spent catalysts, implying that there might not be residual carbon oxygenates left on the catalyst external surface after the reaction or the concentration of these surface species might be too low to be detectable. The C 1s signal was also monitored over the catalysts collected from various runs to trace the tendency for carbon deposition which has been identified as the main contributor to catalyst deactivation. As demonstrated in Figure 9d, the carbon signal is observed to some degree from all of the samples tested. The residual carbon on the surface of the fresh catalyst might originate from the adsorption of CO2 and/or other carboncontaining compounds during sample storage. Coke formation is clearly witnessed from the stronger peak intensity in the spectra collected over the catalyst after reaction in both N2 and CH4 environments. Nevertheless, a lower C 1s signal at a binding energy of 284.8 eV is clearly noticed in the spectra from the methane run relative to that from its N2 counterpart, which is consistent with the coke yield results as shown in Figure 1. In addition, it is also worth noting that there is an additional small peak at 289.5 eV present in the spectra collected over the spent catalysts, which might be due to the existence of residual hydrocarbons left over from the reaction on the catalyst surface.47 On the basis of the above analyses along with the performance results released in Figures 1−4, the possible reaction sites on the catalyst’s surface for coaromatization of acetic acid and methane might be inferred as such: acetic acid becomes converted into intermediates on the external surface, and these active intermediates diffuse into the catalyst’s porous structure to undergo aromatization triggered by the Zn particles; on the other hand, methane becomes activated with the facilitation of Ga, which is abundant on the external surface, and the activated moieties formed then migrate to the inner pores for the following aromatization reaction with the reaction intermediate derived from acetic acid. This initial reaction pathway speculation on the catalyst surface is further explored in section 3.3. 3.3. Mechanism Investigation. 3.3.1. Reaction Pathway Investigation. In order to probe the reaction pathway for coaromatization of acetic acid and methane, the liquid and gas products were collected after various reaction times at 400 °C under a CH4 environment and further analyzed by GC-MS and micro-GC, the results of which are shown in Figures 10 and 11, respectively. It is clearly observed that acetic acid is consumed quickly as the reaction progresses with the trend of an increase in CO2 products. This is because the initial step of the reaction sequence involves oxygen elimination of acetic acid through ketonization to form acetone and CO2.30 It is also noticed that there is a large initial concentration of acetone that is gradually converted. This is attributed to two parallel reactions of the intermediate acetone. A portion of acetone undergoes condensation and dehydration to form the intermediate mesityl oxide, which is observed in the liquid products collected from the 1 min trial as shown in Figure S7 in the Supporting Information. Simultaneously, some of the acetone molecules are consumed to create isobutene by means of aldol condensation followed by β-scission,30 which is detected as a C4 hydrocarbon by micro-GC in the gas products. As the reaction proceeds, more aromatics are steadily formed,
5%Zn-1%Ga/ZSM-5 catalyst. Furthermore, the analysis of coke formation on various catalysts by TGA (Figure S6 and Table S3 in the Supporting Information) indicates that the addition of 5%Zn and/or 1%Ga on the support can enhance the coke resistance. More details of the N2 physisorption, XRD, TEM, DRIFTS, NH3-TPD, and TGA analysis can be found in the Supporting Information. To obtain a further understanding of the distribution of each element loaded on the 5%Zn-1%Ga/ZSM-5 catalyst surface and their corresponding chemical state, XPS was employed for conducting specific scans at Ga 2p, Zn 2p, O 1s, and C 1s regions, respectively, and the results are disseminated in Figure 9. As shown in Figure 9a, the Ga 2p region has significantly split spin−orbit components located at 1145.81 and 1118.85 eV corresponding to Ga 2p1/2 and Ga 2p3/2 of Ga native oxide (Ga2O3), respectively.41,42 The binding energy difference (Δ) between them is 26.96 eV, which is consistent with the reported data (Δ = 26.9 eV).43 All of the above data suggest that a major part of Ga species on the catalysts is in the highest oxidation state, Ga3+. The insignificant peak shift and lack of an additional peak imply that the Ga valence state is relatively stable in CH4 or N2. Furthermore, there is no significant change in Ga content after reactions in different environments, indicating that the gallium distribution is well maintained in acetic acid aromatization, which is further evidenced by the absence of large bright spots (or disappearance of bright spots) in images of Ga-L EELS mapping, as seen in Figure S3d−f in the Supporting Information. This relative stability of dispersion and high oxidation state of Ga on the surface might be associated with its better catalytic performance. As displayed in Figure 9b, it is also observed that Zn 2p has two fitting peaks found at about 1023.09 and 1046.16 eV corresponding to Zn 2p3/2, and Zn 2p1/2, respectively.44−46 The binding energy separation between Zn 2p3/2 and Zn 2p1/2 is 23.07 eV, which is in agreement with reported data (Δ = 23 eV).45 These observations suggest that the Zn species in the catalysts are in the 2+ oxidation state, indicating the presence of ZnO. When the spectra collected from the acetic acid and CH4 run are compared with those from a freshly synthesized sample, it is clearly noticed that the intensities of the ZnO peaks are significantly reduced, which can be explained on the basis of the possibility that Zn species might migrate into the inner pores of the zeolite support and become dispersed better on the surface in the presence of methane. In contrast, the surface concentration of Zn is increased after reaction in an N2 environment, probably as a result of the agglomeration leading to subsequently intense distribution. These observations are highly consistent with the observations in images of TEM and Zn-M EELS mapping shown in Figures S2c and S3a−c in the Supporting Information. In a complementary matter, O surface concentration is reduced after reaction in an N2 environment, as shown in Figure 9c. Nonetheless, the surface O signal is increased after the methane run in comparison to that from the fresh sample, indicating that more oxygen of the zeolite framework is exposed to the catalyst’s surface probably due to Zn diffusion toward internal channels. In addition, the binding energy of O at 532.55 eV matches the characteristic energy from that in SiO2 abundant in the ZSM-5 support. It is interesting to note that the presence of surface oxygen coming from zinc and gallium oxides is not observed in all three spectra included in Figure 9c, which is probably due to their significantly low concentration in 3688
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis
along with activated methane diffuse from the outer surface to the catalyst’s inner pores to provide the building blocks for the formation of C6−C8 naphthenes via oligomerization and cyclization.48 These naphthenes are immediately converted into aromatics through hydrogen transfer; meanwhile, C2−C4 olefins become saturated to form C2−C4 paraffins.35 The diffusion of hydrocarbons into the channel of the catalyst might be the limiting reaction step, causing the aromatics to be generated more slowly in the course of coaromatization. In addition, the phenols initially increase and gradually decrease in the liquid products as the reaction proceeds, as shown in Figure 10. This is as a result of the intermediate mesityl oxide originating from acetone, which then continues to be converted into isophorone,33,48 which might further react with activated methane to ultimately produce xylenol by virtue of demethylation from the quaternary carbon upon dehydrogenation.33 The generated phenolics can be partially catalyzed to produce aromatics via hydrodeoxygenation under the facilitation of gallium species that mainly deposit on the outer surface of the catalyst.39 On the basis of these observations as well as the aforementioned TEM, NMR, and XPS analysis results in this study, a hypothetical reaction pathway for coaromatization of acetic acid and methane over 5%Zn-1%Ga/ZSM5 is proposed in Figure 12, which requires further investigation. 3.3.2. Methane Incorporation Mechanism. Two different isotope-labeled molecules of acetic acid (13CH3COOH and CH313COOH) were employed to gain more insight into the reaction mechanism of the coaromatization between acetic acid and methane over 5%Zn-1%Ga/ZSM-5 at 400 °C. The results of the 13C NMR experiments are displayed in Figure 13a,c. The enlarged views of the 120−140 ppm region corresponding to the aromatic products are also included in Figure 13b,d, and the ratios of major peak areas with respect to CDCl3 within that region are reported in Table 3. The significant increase in peak intensity in the 13C NMR spectrum of the liquid products obtained from the isotope-labeled runs are clearly observed in Figure 13a,c, in comparison to that from their non-isotopelabeled counterpart. In conjunction with the quantitative analysis presented in Table 3, it is concluded that the carbon atoms of acetic acid strongly contribute to the phenyl ring and benzylic carbon. It is interestingly found that the carboxyl carbon shows an increased preference over the methyl carbon for the phenyl carbon site directly bonded with alkyl substituent groups and the benzylic site, which is evidenced by the
Figure 10. Composition of liquid products collected after different reaction times at 400 °C under a CH4 environment with an initial pressure of 5 bar.
Figure 11. Concentration of CO2 and hydrocarbons in the gas phase collected after various reaction times at 400 °C under CH4 environments with an initial pressure of 5 bar.
accompanied by a decrease in C4 and increase in C2 and C3 gas products. One reasonable explanation for this phenomenon is that two isobutene molecules continue to condense into intermediate C8 olefins and subsequently trigger the cleavage reaction to form C2 and C3 hydrocarbons as gas byproducts on the external surface of the catalyst. The formed hydrocarbons
Figure 12. Hypothetical reaction pathway for coaromatization of acetic acid and methane over 5%Zn-1%Ga/ZSM5 at 400 °C. 3689
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis
Figure 13. 13C liquid NMR spectra (a, c) and an enlarged view of the 120−140 ppm chemical shift region (b, d) of the liquid products collected from the reactions between 13CH3COOH/CH313COOH and CH4. The triplet peak centered at 77.23 ppm belongs to CDCl3, which serves as an internal reference for more accurate comparison.
Table 3. 13C Liquid NMR Peak Area Ratio with Respect to That for CDCl3 of the Products from CH313COOH and 13 CH3COOH Aromatization under a CH4 Environment 13
chem shift (ppm)
peak assigned
137.55, 134.30 130.12, 129.8, 129.17 128.38 126.26, 125.56 39.23, 21.5, 19.85 15.93
phenyl carbon directly bonded with alkyl substituent groups ortho and meta positions of the alkyl-substituted phenyl ring meta positions of the alkyl-substituted phenyl ring para positions of the alkyl-substituted phenyl ring benzylic carbon alkyl carbon sites not directly attached to the phenyl ring
significant increase in 13C NMR peak intensity at these two sites when CH313COOH is used. The high preference of carboxyl carbon for a benzylic carbon site might be due to the carboxyl carbon first occupying the phenyl carbon site and then transferring from the phenyl carbon site to the benzylic carbon site by a ring-expansion/-contraction mechanism (Scheme 2).16
CH3COOH + CH4 0.17 5.37 6.62 6.01 3.91 0.31
CH313COOH + CH4
CH3COOH + CH4
1.63 5.13 6.24 6.06 6.47 0.71
0.095 0.57 0.29 0.29 0.48 0.11
In combination with the 13C NMR results of the labeled methane (13CH4) trial in Table 1, on the basis of how pronounced the increase in 13C NMR peak intensity is, it can be inferred that methane contributes a small portion of carbons to all substitution carbon sites of aromatics and the methyl and carbonyl carbons of acetic acid are equally involved in the formation of ortho, meta, and para carbons of the aromatics; whereas the phenyl carbons directly bonded with alkyl substituent groups and benzylic carbons are derived mainly from the carboxyl carbon of acetic acid, the minority is donated by the methyl carbon of acetic acid. Figure S8 in the Supporting Information illustrates 13C CP MAS NMR spectra of the surface species on the 5%Zn-1%Ga/ ZSM-5 catalyst collected after various reaction times at 400 °C under methane with enriched 13C. The peak at a chemical shift of −7 ppm is assigned to 13CH4 absorbed on the catalyst’s surface.49,50 The signals appearing at 21 and 30 ppm are due to the benzylic site on the substitute group of aromatics,51 including the methyl group.49,52,53 The resonances at 129 and
Scheme 2. Mechanism of Carboxyl Carbon Transfer from Acetic Acid to the Benzylic Carbon Site of the Formed Aromatics during Methane and Acetic Acid Co-conversion over 5%Zn-1%Ga/ZSM-5
3690
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis
4. CONCLUSION The present work demonstrates the technical feasibility of coaromatization between acetic acid and methane over various transition-metal-modified zeolite catalysts under mild conditions. Among them, 5%Zn-1%Ga/ZSM-5 exhibits the most promising catalytic performance. The loaded metal species not only promote methane activation but also appropriately modulate the surface acidity, leading to the enhancement of aromatics formation and methane conversion in a methane environment. Furthermore, methane participation into acetic acid aromatization has been demonstrated by using NMR spectroscopy, the results of which imply that the carbon and hydrogen of methane preferentially tend to be incorporated in the phenyl rings and benzylic sites of the product molecules in the reaction. In combination with isotope-tracing experiments with 13C-labeled acetic acid (13CH3COOH/CH313COOH), this demonstrates the equal involvement of the two carbons of acetic acid into the ortho, meta, and para carbons of the benzene ring, while the carboxyl carbon of acetic acid has higher preference in comparison to the methyl carbon for the phenyl carbons as well as the benzylic carbons in the aromatic compounds. Two mechanisms of the coaromatization have been proposed. Extensive catalyst characterizations suggest that the excellent catalytic performance of 5%Zn-1%Ga/ZSM-5 might be attributed to a better metal species dispersion on the surface and its appropriate distribution between the internal and external surface, the relatively stable chemical state, the moderate surface acidity, and the suitable ratio of weakly acidic sites to strongly acidic sites. However, the modification of the catalyst to increase the yield of liquid products and further catalytic reaction mechanistic investigations are critical future work.
131 ppm belong to the ortho, meta, and para sites of the phenyl ring. The peaks centered at 141 and 153 ppm are attributed to the phenyl carbon atom that is directly bonded to alkyl and hydroxyl groups, respectively.52,54 There are also small resonances centered at 183 ppm due to the carbonyl carbon of acetic acid.52 The 13C CP MAS NMR spectra are consistent with the spectra obtained for the sample in solution (Figure 7). More importantly, it is noted that the peak intensity of acetic acid (at 183 ppm) decreased with the reaction time, and the peak at 141 ppm became more and more clear, indicating that more aromatics are produced. These observations are in reasonable agreement with the GC-MS results presented in Figure 10. 13C CP MAS NMR spectra of a catalyst sample exposed to natural-abundance methane (CH4) at 400 °C for 9 min was also acquired (Figure S9 in the Supporting Information); in comparison to that of the sample exposed to 13 C-enriched methane (13CH4), the noticeable difference in the intensity of the peaks at 153, 141, 131, and 21 ppm demonstrates that methane is not only incorporated into the ortho, meta, and para sites of phenyl ring but also scrambles the phenyl carbon directly bonded to the alkyl or hydroxyl group while on the catalyst surface. According to the 13C liquid and solid-state NMR analyses as well as the proposed reaction pathway above, two mechanisms for the embedding of methane carbon into the aromatics ring are hypothesized. Methane together with hydrocarbons (C2− C4) interact with active sites (such as Zn species) in the channel of the catalyst to form the intermediate naphthene species, which are further converted to aromatics by dehydroaromatization with the assistance of the ZSM-5 acidic support (Scheme 3).55
■
Scheme 3. Mechanism of the 13C Label of Methane Embedding into Aromatics on 5%Zn-1%Ga/ZSM-5
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.7b00296. Liquid product GC-MS data, gas product composition results and catalyst characterization results from N2 physisorption, XRD, TEM, TGA, NH3 TPD, SS NMR, and DRIFTS upon pyridine adsorption (PDF)
■
The other approach for the conversion of methane into aromatic compounds involves the alkylation of aromatic molecules originating exclusively from acetic acid by methane and the intramolecular scrambling of carbon sites in methylbenzenes by a ring-expansion/-contraction mechanism (Scheme 4).16,55,56 This hypothesis is quite reasonable, since it is observed that the methane-13C goes to the phenyl carbon directly bonded with an OH group in phenolics or alkyl substituent groups in Figure 7b and Figure S9 in the Supporting Information.
AUTHOR INFORMATION
Corresponding Author
*H.S.: fax, +1 (403) 284-4852; tel, +1 (403) 220-3792; e-mail,
[email protected]. ORCID
Vladimir K. Michaelis: 0000-0002-6708-7660 Hongbo Zeng: 0000-0002-1432-5979 Hua Song: 0000-0002-2791-1723
Scheme 4. Mechanism of the Transfer from 13CH4 into Aromatic Compounds during Methane and Acetic Acid Co-conversion on 5%Zn-1%Ga/ZSM-5
3691
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
Research Article
ACS Catalysis Notes
(32) Hasan, M.; Zaki, M.; Pasupulety, L. Appl. Catal., A 2003, 243, 81−92. (33) Salvapati, G.; Ramanamurty, K.; Janardanarao, M. J. Mol. Catal. 1989, 54, 9−30. (34) Linstead, R. P.; Thomas, S. L. S. J. Chem. Soc. 1940, 1127−1134. (35) Viswanadham, N.; Saxena, S. Fuel 2013, 105, 490−495. (36) Zhao, C. Y.; Krall, A.; Zhao, H.; Zhang, Q.; Li, Y. Int. J. Hydrogen Energy 2012, 37, 9967−9976. (37) Yu, C.; Xu, H.; Ge, Q.; Li, W. J. Mol. Catal. A: Chem. 2007, 266, 80−87. (38) Majhi, S.; Pant, K. J. Ind. Eng. Chem. 2014, 20, 2364−2369. (39) Ausavasukhi, A.; Huang, Y.; Anh, T.; Sooknoi, T.; Resasco, D. J. Catal. 2012, 290, 90−100. (40) Pretsch, E.; Buhlmann, P.; Badertscher, M. Structure Determination of Organic Compounds Tables of Spectral Data, 4th ed.; Springer: Zurich, Switzerland, 2009; p 93. (41) He, P.; Lou, Y.; Song, H. Fuel 2016, 182, 577−587. (42) Cossu, G.; Ingo, G. M.; Mattogno, G.; Padeletti, G.; Proietti, G. Appl. Surf. Sci. 1992, 56-58, 81−88. (43) Thermo Scientific XPS. Gallium; http://xpssimplified.com/ elements/gallium.php (accessed Oct 16, 2016). (44) Lupan, O.; Emelchenko, G.; Ursaki, V.; Chai, G.; Redkin, A.; Gruzintsev, A. Mater. Res. Bull. 2010, 45, 1026−1032. (45) Al-Gaashani, R.; Radiman, S.; Daud, A.; Tabet, N.; Al-Douri, Y. Ceram. Int. 2013, 39, 2283−2292. (46) Deerattrakul, V.; Dittanet, P.; Sawangphruk, M.; Kongkachuichay, P. J. CO2 Util. 2016, 16, 104−113. (47) He, P.; Zhao, L.; Song, H. Appl. Catal., B 2017, 201, 438−450. (48) Setiadi, S.; Nasikin, M.; Tsutsui, T.; Kojima, T. Jurnal Teknologi 2008, 4, 328−336. (49) Luzgin, M. V.; Rogov, V.; Arzumanov, S.; Toktarev, A. V.; Stepanov, A.; Parmon, V. Catal. Today 2009, 144, 265−272. (50) Gabrienko, A.; Arzumanov, S.; Moroz, I.; Toktarev, A. V.; Wang, W.; Stepanov, A. J. Phys. Chem. C 2013, 117, 7690−7702. (51) Liu, Y.; Li, D.; Wang, T.; Liu, Y.; Xu, T.; Zhang, Y. ACS Catal. 2016, 6, 5366−5370. (52) Wang, X.; Xu, J.; Qi, G.; Wang, C.; Wang, Q.; Deng, F. Chem. Commun. 2014, 50, 11382. (53) Luzgin, M. V.; Rogov, V.; Arzumanov, S.; Toktarev, A. V.; Stepanov, A.; Parmon, V. Angew. Chem., Int. Ed. 2008, 47, 4559−4562. (54) Hu, J. Z.; Solum, M. S.; Taylor, C. M. V.; Pugmire, R. J.; Grant, D. M. Energy Fuels 2001, 15, 14−22. (55) Guo, J.; Lou, H.; Zheng, X. J. Nat. Gas Chem. 2009, 18, 260− 272. (56) Luzgin, M. V.; Rogov, V. A.; Arzumanov, S. S.; Toktarev, A. V.; Stepanov, A. G.; Parmon, V. N. Angew. Chem., Int. Ed. 2008, 47, 4559− 4562.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We gratefully acknowledge financial support from the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant Program (RGPIN/04385-2014 and 2016-05447).
■
REFERENCES
(1) British Petroleum. BP Statistical Review of World Energy; BP p.l.c.: London, 2016. (2) Bridgwater, A. Biomass Bioenergy 2012, 38, 68−94. (3) Mettler, M.; Vlachos, D.; Dauenhauer, P. Energy Environ. Sci. 2012, 5, 7797−7809. (4) French, R.; Czernik, S. Fuel Process. Technol. 2010, 91, 25−32. (5) Bridgwater, A. Catal. Today 1996, 29, 285−295. (6) Thangalazhy-Gopakumar, S.; Adhikari, S.; Gupta, R.; Tu, M.; Taylor, S. Bioresour. Technol. 2011, 102, 6742−6749. (7) Wright, M.; Daugaard, D.; Satrio, J.; Brown, R. Fuel 2010, 89, S2−S10. (8) Brown, T.; Thilakaratne, R.; Brown, R.; Hu, G. Fuel 2013, 106, 463−469. (9) U.S. Department of Energy. Clean Cities Alternative Fuel Price Report; U.S. Department of Energy: Washington, DC, 2016. (10) Yang, F.; Wang, S.; Xia, Y.; Bi, W.; Wang, W. Pet. Sci. Technol. 2011, 29, 1675−1684. (11) Tang, P.; Zhu, Q.; Wu, Z.; Ma, D. Energy Environ. Sci. 2014, 7, 2580−2591. (12) Gabrienko, A.; Arzumanov, S.; Moroz, I.; Toktarev, A.; Wang, W.; Stepanov, A. J. Phys. Chem. C 2013, 117, 7690−7702. (13) Anunziata, O.; Mercado, G.; Pierella, L. Catal. Commun. 2004, 5, 401−405. (14) Choudhary, V.; Kinage, A.; Choudhary, T. Science 1997, 275, 1286−1288. (15) Anunziata, O.; Mercado, G.; Pierella, L. Catal. Lett. 2003, 87, 167−171. (16) Luzgin, M.; Rogov, V.; Arzumanov, S.; Toktarev, A.; Stepanov, A.; Parmon, V. Catal. Today 2009, 144, 265−272. (17) Choudhary, V.; Mondal, K.; Mulla, S. Angew. Chem., Int. Ed. 2005, 44, 4381−4385. (18) Wang, A.; He, P.; Yung, M.; Zeng, H.; Qian, H.; Song, H. Appl. Catal., B 2016, 198, 480−492. (19) Demirbas, F. Energy Sources, Part A 2006, 28, 1181−1188. (20) Valle, B.; Gayubo, A.; Aguayo, A.; Olazar, M.; Bilbao, J. Energy Fuels 2010, 24, 2060−2070. (21) Peng, J.; Chen, P.; Lou, H.; Zheng, X. Bioresour. Technol. 2009, 100, 3415−3418. (22) Su-Ping, Z.; Yong-Jie, Y.; Zhengwei, R.; Tingchen, L. Energy Sources 2003, 25, 57−65. (23) Sukhbaatar, B.; Steele, P.; Ingram, L.; Kim, M. Bioresources 2009, 4, 1319−1329. (24) Mahfud, F.; van Geel, F.; Vanderbosch, R.; Heeres, H. Sep. Sci. Technol. 2008, 43, 3056−3074. (25) Teella, A.; Huber, G.; Ford, D. J. Membr. Sci. 2011, 378, 495− 502. (26) Nogueira, F.; Assaf, P.; Carvalho, H.; Assaf, E. Appl. Catal., B 2014, 160−161, 188−199. (27) Bimbela, F.; Oliva, M.; Ruiz, J.; Garcia, L.; Arauzo, J. J. Anal. Appl. Pyrolysis 2007, 79, 112−120. (28) Wang, S.; Cai, Q.; Zhang, F.; Li, X.; Zhang, L.; Luo, Z. Int. J. Hydrogen Energy 2014, 39, 18675−18687. (29) Rioche, C.; Kulkarni, S.; Meunier, F.; Breen, J.; Burch, R. Appl. Catal., B 2005, 61, 130−139. (30) Chang, C. D.; Chen, N. Y.; Koenig, L. R.; Walsh, D. E. 185th American Chemical Society National Meeting, Seattle, WA, 1983. (31) He, P.; Song, H. Ind. Eng. Chem. Res. 2014, 53, 15862−15870. 3692
DOI: 10.1021/acscatal.7b00296 ACS Catal. 2017, 7, 3681−3692
