Article pubs.acs.org/EF
Hydropyrolysis of Lignin Using Pd/HZSM‑5 Oliver Jan,† Ryan Marchand,‡ Luiz C. A. Anjos,§ Gabriel V. S. Seufitelli,† Eranda Nikolla,‡ and Fernando L. P. Resende*,† †
School of Environmental and Forest Sciences, University of Washington, Seattle, Washington 98195, United States Department of Chemical Engineering and Materials Science, Wayne State University, Detroit, Michigan 48202, United States § Department of Chemical Engineering and Chemistry, Federal Institute of Education, Science and Technology, Recife, PE 50740-540, Brazil ‡
S Supporting Information *
ABSTRACT: The aim of this work was to study the formation of cycloalkanes from hydropyrolysis of lignin with HZSM-5 and Pd/HZSM-5 catalysts. Cycloalkanes are high-octane-rating molecules and are a major component of jet fuels. We observed that palladium supported on HZSM-5 catalyzed hydrogenation and deoxygenation reactions that converted phenolic compounds into aromatic hydrocarbons and cycloalkanes. This in situ study analyzed the effect of the catalyst-to-lignin ratio, H2 partial pressure, and temperature on the yields of hydrocarbons with HZSM-5 and 1 wt % Pd/HZSM-5. Pd/HZSM-5 produced 44% more aromatic hydrocarbons than HZSM-5 at a catalyst-to-lignin ratio of 20:1, 650 °C, and a constant H2 partial pressure of 1.7 MPa. The presence of palladium led to a significant difference in yields only at 1.7 MPa H2 partial pressure. In addition, the hydropyrolysis temperature played a substantial role in the equilibrium conversion of hydrogenation reactions that led to cycloalkanes directly from lignin. In an attempt to bypass the thermal limitations of the in situ process, we performed ex situ catalytic upgrading experiments and also observed formation of cycloalkanes at a hydropyrolysis temperature of 650 °C, packed bed temperature of 300 °C, and H2 partial pressure of 1.7 MPa.
1. INTRODUCTION The lignin component of biomass is a potential feedstock for the production of fuels and chemicals. Lignin comprises nearly 15− 30 wt % of lignocellulosic biomass and is the second most abundant source of renewable carbon on earth, behind only cellulose.1−3 It is an abundant waste product, with approximately 50 million tons of lignin isolated annually in pulp and paper mills around the world.4,5 Advancements in cellulosic ethanol production in biorefineries will also lead to a surplus of lignin, which must be separated from cellulose and hemicellulose for effective hydrolysis and fermentation processes.2,6,7 Lignin accounts for approximately 40% of the total energy density of lignocellulosic biomass, making it an attractive option for fuel production. 3 Despite this potential, lignin is currently implemented as a low-grade fuel for steam and electricity production.8 If efficient routes for lignin conversion were developed, it could become a sustainable and economical source of fuels and chemicals. A major impediment for lignin transformation to fuels and chemicals is the formation of char/coke that occurs during thermal or chemical depolymerization of the bulk aromatic structure. Lignin is comprised of a macromolecular assembly of phenylpropanoid aromatic rings connected through carbon− carbon bonds and ether linkages, with the β-O-4 aryl ether bonds being the most dominant.9 As the structure is largely aromatic, commodity chemicals such as toluene and p-xylene could theoretically be produced from lignin. However, etherified linkages in lignin readily decompose into char/coke upon thermal or chemical treatment by radical-induced polymerization of aromatic rings.9,10 These solid agglomerates impede molecular diffusion and adsorption on the active sites of catalysts, © XXXX American Chemical Society
eventually leading to catalyst deactivation and clogging in reactors. Thermochemical processes, such as hydropyrolysis, have been relatively unexplored for the production of biofuels from lignin. Hydropyrolysis is a thermal degradation process in which biomass is pyrolyzed into monomeric and oligomeric units in the presence of H2. The H2 environment plays an important role in the production of hydrocarbons from lignin. H2 has been proposed as a suitable reaction medium for reducing the amount of coke on catalytic active sites, which makes hydropyrolysis of lignin a suitable thermochemical process to minimize coke formation on catalytic surfaces.11 Researchers from GTI have shown potential in using catalytic hydropyrolysis to convert woody biomass into a fungible fuel, producing deoxygenated liquid oil rich in hydrocarbons.12 In addition, RTI International have also developed bench-scale hydropyrolysis reactors to catalytically improve the physical and chemical properties of biooil.13 However, the design of proper catalysts for lignin hydropyrolysis is critical, as lignin also contains oxygen functional groups that must be removed to produce a fuel with a heating value similar to liquid fuel produced from crude oil and petroleum.1,3 Therefore, a catalyst that could facilitate the removal of oxygen heteroatoms and the addition of hydrogen atoms would be ideal for converting lignin into liquid fuels for transportation. Bifunctional catalysts, comprised of supported metals on acidic zeolite catalysts, have been proposed to facilitate hydrogenation Received: December 15, 2014 Revised: February 2, 2015
A
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels and deoxygenation reactions.14−18 This reaction mechanism, known as hydrodeoxygenation (HDO), is an important reaction pathway for lignin valorization into molecules such as cycloalkanes and branched aliphatic compounds commonly found in conventional liquid fuels. Cycloalkanes are high-octane-rating molecules and a major component of jet fuels, comprising 52.8 vol % of JP-5.19 The majority of present research has attempted HDO of model lignin components, such as guaiacol and anisole, with bifunctional catalysts consisting of platinum and zeolite supports.20,21 However, incomplete HDO was observed in these reports, as the product distribution consisted primarily of aromatic hydrocarbons and phenolic compounds. In addition, very little work has been performed on hydropyrolysis of raw lignin in general, suggesting the need for additional studies. In the present work, we performed hydropyrolysis of raw, natural lignin with and without catalysts using a pyroprobe interfaced with a gas chromatograph/mass spectrometer (PyGC/MS). For the catalytic experiments, lignin was mixed with the catalyst HZSM-5 or Pd/HZSM-5. Palladium was used in these studies since it is a highly active hydrogenation catalyst and the combination of palladium with acidic zeolite catalysts has not been previously tested under H2 conditions. We varied the catalyst-to-lignin ratio, H2 partial pressure, and hydropyrolysis temperature to observe the relationship of each parameter to the yields of aromatic hydrocarbons and cycloalkanes. To the best of our knowledge, this is the first hydropyrolysis study on natural lignin to yield cycloalkanes and the first to demonstrate the bifunctional behavior of Pd/HZSM-5 for HDO reactions.
The pressure in the pyroprobe unit was controlled using a highpressure module. A backpressure regulator maintained the pressure within the system, and 6.0-grade helium gas (99.9999%) acted as the carrier/sweeping gas. The valve oven and transfer line were both set constant at 300 °C throughout all experiments, and 5.0-grade H2 reactant gas (99.999%) was introduced during the interface temperature ramp. The interface had a ballistic heating curve from 100 to 300 °C upon software initialization. Once the interface temperature was stable at 300 °C, the platinum coil rapidly heated to the set point at a ramp of 1000 °C/s, and the set point temperature was maintained for 60 s. Vapors produced from hydropyrolysis were then absorbed onto a downstream TENAX trap at 50 °C to allow for pyrolysate collection. However, permanent gases could not be collected by this trapping setup and were sent through the purge vent. Upon completion, hydropyrolysis products absorbed onto the trap were then thermally desorbed at 300 °C for 5 min and swept by helium into a gas chromatography/mass spectrometer system for downstream analysis (QP2010 Ultra, Shimadzu). Gas chromatography/mass spectrometry (GC/MS) was used to qualitatively and quantitatively determine the volatile hydropyrolysis products. The injector was kept at a constant 300 °C. Chromatographic separation of the hydropyrolysis products was achieved using a ZB-5 ms capillary column (30 m length × 0.25 mm inner diameter × 0.25 μm film thickness) (Phenomenex). The helium flow rate was set to 206.5 mL/ min at an initial pressure of 113.3 kPa and a split ratio of 125. The temperature of the GC oven started at 35 °C (held for 2 min), followed by a ramp to 200 °C at 5 °C/min (held at 200 °C for 5 min), and ramped to 300 °C at 10 °C/min (held for 5 min). The mass spectrometer was operated at 0.88 kV in SCAN mode. Target masses were analyzed from m/z of 45.00 to 500.00 at a scan speed of 1666 amu/s. NIST MS libraries were used to identify chromatographic peaks of compounds with a similarity index of 90 or greater. We performed screening experiments by varying the hydrogen partial pressure to determine a suitable operating pressure. We observed that 1.72 MPa was an optimal hydrogen partial pressure in the range tested, and all catalytic experiments reported were performed at this condition. In this work, we report only the volatile products at 300 °C. There are also permanent gases (CO, CO2, CH4) and nonvolatile compounds that could not be measured with the pyroprobe setup. In addition, the mass of the resulting solid (catalyst + coke) was too small to be measured accurately and is not reported. 2.4. Data Analysis. Five-point calibration curves of major hydropyrolysis products (similarity index >90) were constructed using analytical standards purchased from Sigma-Aldrich and Fisher through autosampler injection in HPLC-grade methanol. A total of 50 compounds were quantified by TIC using 19 external standards: 7 aromatic hydrocarbons and 12 phenolics. Response factors for the remaining compounds were approximated by those of standards with similar chemical structures. All trials were performed in triplicate for reproducibility. Error bars reported represent the standard deviations of the replicated trials. 2.4.1. Yield Calculations. Throughout this work, yield is defined as
2. MATERIALS AND METHODS 2.1. Lignin. Lignin used throughout this study was prepared via sulfur-catalyzed steam explosion of hybrid poplar (195 °C, 3% SO2, 5 min). Klason acid hydrolysis (121 °C, 72% H2SO4, 60 min) was performed to further purify the poplar lignin of any residues and undissolved sugars. An elemental analyzer (EA 2400, PerkinElmer) was used to determine the weight percentages of carbon, hydrogen, and nitrogen in the lignin feedstock. Inductively coupled plasma atomic emission spectroscopy (ICP/OES) allowed for the determination of sulfur. Ash content was determined using the procedure outlined in ASTM E1534-93.22 Oxygen content was calculated by difference of the carbon, hydrogen, nitrogen, sulfur, and ash content. 2.2. Catalyst Preparation. ZSM-5 (CBV-2314, Zeolyst International) was purchased in the nominal ammonium cation form. The acidic form, HZSM-5, was prepared by calcination in air for 5 h at 600 °C. The specified surface area was 425 m2/g with a SiO2/Al2O3 ratio of 23. Pd/HZSM-5 catalysts were prepared through a wet impregnation technique. Appropriate amounts of Pd(NO3)·6H2O and HZSM-5 were mixed to obtain 1 wt % loading of Pd with respect to the total catalyst at 700 rpm in deionized water at room temperature for 24 h. The water was then removed via evaporation. The resulting dried catalyst was calcined at 600 °C for 2 h under an oxygen atmosphere. Finally, the catalyst was reduced using 10% H2 balance in argon gas at 550 °C for 3 h. 2.3. Hydropyrolysis−Gas Chromatography/Mass Spectrometry. Microscale hydropyrolysis experiments were conducted using a Pyroprobe 5200 model with high pressure and reactant gas options (Chemical Data Systems Analytical). The pyroprobe system was equipped with a platinum coil that served simultaneously as a sample holder and resistive heating element. Hydropyrolysis samples were prepared in quartz tubes with up to 0.5 mg of lignin and/or catalyst surrounded by two layers of quartz wool, as described previously.4 An analytical balance with a resolution of 0.1 μg was used for accurate weighing of hydropyrolysis samples (AD 6000 Ultra Microbalance, PerkinElmer). Catalysts were mixed with lignin in ratios of 1:1, 10:1, and 20:1. Fresh catalysts were used for all trials.
% yieldi =
mass of compound i in product × 100 mass of lignin
% selectivityi =
mass of compound i in product × 100 mass of total observed products
on the basis of previous reports.4,22 2.4.2. Lignin Characterization. The elemental composition of the steam-exploded poplar lignin was determined by elemental analysis and ICP/OES. The results are shown in Table 1. The empirical formula of the lignin was found to be C3H2O. Nitrogen and sulfur were excluded from the formula, as they constitute less than 0.5 wt % of the lignin. The ash content was found to be negligible.
3. RESULTS AND DISCUSSION 3.1. Noncatalytic Hydropyrolysis of Lignin. Poplar lignin was first pyrolyzed in H2 in the absence of catalysts to determine B
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
mentioned previously, making it more susceptible to hydrogenolysis by H2 radicals. 3.2. Catalytic Hydropyrolysis of Lignin. Catalytic hydropyrolysis of lignin was studied using HZSM-5 or Pd/HZSM-5 catalysts mixed with lignin in the pyroprobe. The catalyst-tolignin ratio and hydropyrolysis temperature were varied in this study to observe their effect on the yields of aromatic hydrocarbons and cycloalkanes. 3.2.1. Effect of Catalyst-to-Lignin Ratio. This study aimed to determine, within the range of conditions studied, the optimal amount of catalyst necessary to generate a high aromatic hydrocarbon yield and low phenolic yield for HZSM-5 and Pd/ HZSM-5 catalysts. For each catalyst, three catalyst-to-lignin ratios were used: 1:1, 10:1, and 20:1. It has been previously observed that higher ratios of catalyst to biomass have a positive effect on the hydrocarbon yield.24 We performed screening experiments in the pressure range of 0.00−2.58 MPa at 650 °C to test the effect of the catalyst-to-lignin ratio on the hydrocarbon yield. Within the range studied, the optimal experimental conditions for aromatic hydrocarbon generation and deoxygenation occurred at a H2 partial pressure of 1.72 MPa. Only this situation is reported in Figures 2 and 3. At the lowest catalyst-to-lignin
Table 1. Elemental Analysis Results of Lignin Derived from Hybrid Poplar
a
C (%)
H (%)
N (%)
O (%)
S (%)a
61.7
3.21
0.267
34.5
0.24
Tested by ICP/OES, in a single run.
the base case product distribution, as shown in Figure 1. Phenols (phenol, m/p-cresol, catechol, methylcatechol), guaiacols
Figure 1. Major phenolic and aromatic hydrocarbons produced from noncatalytic hydropyrolysis of lignin at 650 °C.
(guaiacol, creosol), and syringol were among the most prevalent phenolic compounds observed. Trace amounts of aromatic hydrocarbons were also detected during this initial study. We conducted pressurized experiments with only helium and observed no significant differences, as no aromatic hydrocarbons were produced without hydrogen. We increased the H2 partial pressure (gauge) to determine its effect on noncatalytic hydropyrolysis of lignin. For major compounds such as guaiacol and m/p-cresol, the overall effect of H2 partial pressure was small from 0.00 to 1.72 MPa. However, smaller yields of larger oxygenated compounds (catechol, methylcatechol, isoeugenol, vanillin) were observed at higher pressures. Conversely, as the H2 partial pressure increased, there was an apparent increase in the yields of the aforementioned aromatic hydrocarbons (benzene, toluene, xylenes, methylnaphthalene, azulene). Interestingly, phenol yields increased from about 2 wt % to approximately 3.5 wt % as the H2 partial pressure increased from 0.00 to 0.85 MPa, which suggests deoxygenation of larger phenolic compounds into phenol. Phenol then deoxygenated further to produce aromatic hydrocarbons. Overall, these results suggest that H2, in the absence of catalysts, has only a minor effect on the yields. H2, particularly at high temperatures, may induce hydrogenolysis mechanisms where H2 radicals promote demethoxylation, decarbonylation, and dehydration reactions. Interestingly, the yields of all aromatic and phenolic compounds decreased as the H2 partial pressure was further increased to 2.58 MPa. This suggests that a pressurized H2 environment may induce severe hydrogenolysis of the lignin structure and increased production of noncondensable gases, i.e., CO, CO2, CH4, C2H4, and C2H6. The highly acidic pretreatment process may have resulted in a structurally weaker lignin, as
Figure 2. Effect of palladium on total aromatic hydrocarbon yield. Reaction conditions: Tpyr = 650 °C, PH = 1.72 MPa.
ratio (1:1), the Pd/HZSM-5 catalyst behaves similarly to HZSM5 as both catalysts produce similar yields of aromatic hydro-
Figure 3. Effect of palladium on total phenolic yield. Reaction conditions: Tpyr = 650 °C, PH = 1.72 MPa. C
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 4. Aromatic hydrocarbon distribution over HZSM-5 and Pd/HZSM-5 at 650 °C, PH = 1.72 MPa, and 20:1 catalyst-to-lignin ratio.
between Pd and HZSM-5 is substantital during hydropyrolysis of lignin, with the Pd assisting HZSM-5 in deoxygenating the phenolic products. 3.2.2. Effect of Pd on Product Distribution. Aromatic compounds, from C6 to polyaromatic hydrocarbons, are listed in Figure 4 at the optimal experimental conditions (650 °C, 1.72 MPa, and 20:1 catalyst-to-lignin ratio) with their respective yields from HZSM-5 and Pd/HZSM-5 catalysts. Both catalysts, at the specified conditions, exhibited enhanced reactivity at 1.72 MPa. Pd/HZSM-5, however, produced higher yields of C6, C7, C8, C9, C10, and polyaromatic hydrocarbons than HZSM-5. This result is consistent with Figure 2, as Pd/ HZSM-5 produced 44% more aromatic hydrocarbons than HZSM-5, with overall yields of approximately 40 and 26 wt %, respectively. In particular, the toluene yield with Pd/HZSM-5 was 10 wt % (27% selectivity), a 27% increase compared to that of HZSM-5, which produced 8.4 wt % (33% selectivity). The pxylene yield increased 28% with Pd/HZSM-5, with yields increasing from 5 wt % (21% selectivity) to 7 wt % (17% selectivity). This result is critical as both toluene and p-xylene are commercially valuable products in both the fuel and chemical industries. Naphthalenes and methylnaphthalenes, grouped in “Polyarom”, were major products formed from lignin over Pd/HZSM5. In fact, the addition of Pd more than doubled the polyaromatic hydrocarbon yield (15% selectivity) compared to that of HZSM5 (21% selectivity). As polyaromatic hydrocarbons are generally known as precursors to char formation, this result may suggest that the addition of Pd metal impedes char formation, as these smaller ordered polyaromatic compounds were analyzed by a downstream GC/MS. For HZSM-5, it is possible that the polyaromatic compounds that were produced further reacted to form highly condensed species, otherwise known as char. We hypothesize that Pd/HZSM-5, in H2 conditions, minimizes char agglomeration through in situ interactions of aryl radicals with the Pd metal and H2 atmosphere. Rapid polymerization of aryl radicals among neighboring aryl radicals may be inhibited by hydrogen atoms that are facilitated by the surface chemistry of the Pd metal, nullifying the formation of higher ordered polyaromatic rings and increasing the yield of smaller polyaromatics. 3.2.3. Effect of Temperature. Typical hydropyrolysis processes occur in a range from 400 to 600 °C. This study focused on the effect of temperature (400, 550, and 650 °C) and the product distribution from lignin hydropyrolysis over Pd/ HZSM-5. Primarily, this study was performed to investigate the role of temperature in the formation of cycloalkane compounds.
carbons. This behavior is also observed in Figure 3, as both catalysts produced a similar amount of phenolic compounds with no major statistical differences in their deoxygenation ability. However, as the catalyst-to-lignin ratio increased from 1:1 to 10:1 to 20:1, Pd/HZSM-5 surpassed HZSM-5 at each ratio in aromatic hydrocarbon yield and phenolic deoxygenation. The aromatic hydrocarbon yields observed in the present work are higher compared to similar work performed in the literature, which suggests that the effect of the catalyst and the reaction conditions can significantly impact the formation of aromatic hydrocarbons.10,23,25 At a 20:1 catalyst-to-lignin ratio, Pd/HZSM-5 produced 44% more aromatic hydrocarbons than HZSM-5 alone, and virtually no phenolic compounds were detected through mass spectrometry with Pd/HZSM-5. These results suggest that the Pd metal is capable of accelerating deoxygenation reactions that are normally facilitated by HZSM-5, resulting in higher yields of aromatic hydrocarbons. However, throughout these experiments, no cycloalkanes were observed. This indicates that hydrogenation is not a major reaction pathway under these conditions. Similarly, Thangalazhy-Gopakumar et al.26 observed no cycloalkane products using pinewood biomass with metal−supported zeolites at a hydropyrolysis temperature of 650 °C. The aromatic hydrocarbon yields observed are higher compared to similar work performed in the literature, which suggests that the effect of the catalyst and the reaction conditions can significantly impact the formation of aromatic hydrocarbons.10,23,25 No cycloalkane formation was observed with Ni/HZSM-5, Co/HZSM-5, Mo/ HZSM-5, and Pt/HZSM-5, with minimal variation in aromatic hydrocarbon yield over a range of operating pressures among the four catalysts.26 To explain the effects of adding a metal to a support, various research groups have alluded to a synergistic effect that leads to an improvement in the overall catalytic performance.21,27−33 Resasco’s group looked at the effect of using Pt/HBeta with lignin model compound anisole. It was observed that Pt/HBeta was an effective catalyst in the hydrodeoxygenation of anisole to aromatic compounds. They hypothesized that the Pt metal accelerated alkyl transfer reactions while adjacent Brønsted acid sites were responsible for carrying out deoxygenation reactions.21 Ga/HZSM-5 was investigated by Cheng et al. for improving the rate of aromatic production from lignocellulosic biomass. Cheng et al. hypothesized a catalytic bifunctionality in which Ga improved the rates of decarbonylation and olefin aromatization while the zeolite was responsible for oligomerization reactions, thereby leading to an improvement in aromatic yield over that of monofunctional zeolite.32 We believe that the synergistic effect D
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels Catalyst ratio and pressure were held constant for all experiments at 20:1 and 1.72 MPa, respectively, as these conditions displayed improved aromatization from the previous results. Figure 5 depicts the yields of aromatic hydrocarbons and cycloalkane compounds at 400, 550, and 650 °C. Five distinct
Figure 6. Hydrodeoxygenation reaction network of p-cresol to methylcyclohexane.
Although the analysis of specific reaction mechanisms inherent with raw lignin hydropyrolysis is complex, for our results we suggest that the Pd metal is responsible for HYD reactions and also assists HZSM-5 in DDO reactions. As the hydropyrolysis temperature increased from 400 to 650 °C, cycloalkanes were no longer observed. The complementary hydrogenation reactions appeared to be severely affected by the increase in temperature. This may be due to thermodynamic limitations, as hydrogenation reactions are exothermic. As a result, the aromatic yield increases dramatically from 8.7 wt % at 400 °C to 40.3 wt % at 650 °C. Higher temperatures have been observed to improve the overall yields in the hydropyrolysis of lignin, allowing for an increase in the amount of volatile vapor produced, which leads to increased overall yields. Thangalazhy-Gopakumar et al. performed similar studies but with an experimental design that kept the hydropyrolysis temperature constant at 650 °C. Their results showed no formation of cycloalkane compounds at any H2 partial pressure (0−2.76 MPa) for any of the four supported catalysts that they used.26 This is consistent with the result that we observed. This phenomenon is likely due to the low equilibrium constants of hydrogenation reactions at increasing temperatures, as depicted in Figure 7. All equilibrium constants, except those at 550 and 650 °C, were reported literature values, with corresponding reaction enthalpy values for benzene, toluene, and p-xylene hydrogenation.37 We assumed that the enthalpy of reaction remained constant over this temperature range. Equilibrium constants at 550 and 650 °C were calculated using the following equation, with 200 °C as the reference temperature.38
Figure 5. Aromatic and cycloalkane yields with Pd/HZSM-5 at 400, 550, and 650 °C at PH = 1.72 MPa.
cycloalkane species (methylcyclopentane, dimethylcyclopentane, methylcyclohexane, dimethylcyclohexane, and ethylcyclohexane) appeared at 400 °C. Although the total cycloalkane yield is 3 wt % (25% selectivity), this result indicates that Pd/HZSM-5 successfully facilitated HDO reactions of phenolic compounds or aromatic hydrocarbons into cycloalkanes. In addition, no phenolic compounds were detected in the product distribution. As far as we know, this is the first time that cycloalkanes are produced from natural lignin. The presence of cycloalkanes also demonstrates the bifunctionality of the Pd/HZSM-5 in hydropyrolysis conditions, since both the deoxygenation and hydrogenation reactions are needed to produce these molecules. The mechanism of formation of cycloalkanes from phenolic compounds has been reported previously.13,30−32 Several studies on lignin model compounds have indicated that noble metals supported with an acidic support proceed via an initial hydrogenation (HYD) stage to remove the aromatization of the ring structure. A subsequent deoxygenation (DDO) step of the remaining C−O bond yields the resulting cycloalkane.15 Hong et al. analyzed the HDO of phenol using a Pt/HY bifunctionalized catalyst in a fixed bed reactor under elevated H2 pressure (4 MPa) with moderate reaction temperatures. The authors proposed that the successful HDO of phenol to cyclohexane occurred through a series of competing HYD/ DDO reactions to yield cyclohexanol, followed by dehydration to eventually form cyclohexane.34 The two competing mechanisms are represented in Figure 6, as adapted from Wang et al.15 Zhao and Lercher experimented with Ni/HZSM-5 catalysts to upgrade phenolic-rich bio-oil in a semibatch reactor at 250 °C. They observed that, with the addition of H2 (5 MPa), Ni/ HZSM-5 facilitated complete hydrodeoxygenation of phenolic monomers and dimers into cycloalkane molecules.35 Wan et al. observed that the orientation of the adsorbed molecule on the active site determined whether or not the reaction proceeded first via HYD or DDO. Cresol that was vertically adsorbed to the metal surface was more susceptible to DDO, resulting in toluene. HYD was proposed to be the major reaction pathway when cresol was adsorbed coplanar to the metal surface, resulting in methylcyclohexanol and subsequently methylcyclohexane.36
⎡ ΔH o ⎛ 1 1 ⎞⎤ Rx Kc(T ) = Kc(T1) exp⎢ ⎜ − ⎟⎥ ⎢⎣ R ⎝ T1 T ⎠⎥⎦
(1)
The exothermicity of the hydrogenation reaction leads to a decrease in product moles as the reaction temperature increases; high H2 partial pressures and low temperatures would therefore favor optimal equilibrium conversion.11 The decreasing equilibrium constant indicates that very few cycloalkane molecules will be produced at higher reaction temperatures.37 The equilibrium constant for benzene hydrogenation to cyclohexane at 300 °C is more than 21 000 times that of the constant at 650 °C, meaning that infinitesimal amounts of cyclohexane will be present, if any at all, at this higher temperature. We observed no formation of cycloalkanes at 650 °C, which is consistent with the findings of Thangalazhy-Gopakumar et al.26 However, by decreasing the temperature to 400 °C and E
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 7. Equilibrium constants for hydrogenation of various aromatic hydrocarbons.37
Figure 8. Ex situ upgrading of lignin into cycloalkane and aromatic compounds with Pd/HZSM-5 at varying packed bed reactor temperatures. Hydrogen partial pressure = 1.72 MPa, and approximately 20:1 catalyst-to-lignin ratio was held constant. Probe temperature = 650 °C.
Figure 9. Product distribution of cycloalkanes produced using Pd/HZSM-5. Conditions: secondary reactor T = 300 °C, 20:1 catalyst-to-lignin ratio, hydrogen partial pressure = 1.72 MPa.
maintaining a moderate H2 partial pressure of 1.72 MPa, the hydrogenation reactions favor cycloalkane formation. 3.3. Ex Situ Hydropyrolysis of Lignin. To mitigate the thermodynamic limitations between high-temperature hydro-
pyrolysis and hydrogenation, we decoupled the lignin and catalyst into two distinct zones for pyrolysis and catalytic upgrading, respectively. This catalytic configuration is referred to as ex situ hydropyrolysis and it was performed using the F
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
with inexpensive catalysts and zeolitic supports may elucidate other effective bifunctional catalysts for HDO of natural lignin.
pyroprobe and a downstream packed bed reactor. Approximately 0.5 mg of lignin was loaded in a quartz tube in the pyroprobe. A packed bed reactor downstream from the pyroprobe was packed with quartz wool and loaded with approximately 10 mg of catalyst. Fresh catalyst was used for each run. The hydrogen partial pressure and hydropyrolysis temperature were held constant at 1.72 MPa and 650 °C, respectively. The temperature of the packed bed reactor was tested at 300, 400, 500, and 600 °C; these temperatures were measured at the outside wall of the reactor, where the heater and thermocouple are located. In the range of temperatures tested in Figure 8, a secondary reactor temperature of 300 °C produced the highest yield of cycloalkanes at 0.85 wt %, a 71% decrease compared to the cycloalkane yield observed in the in situ upgrading results discussed earlier. As the temperature increased to 600 °C, we observed a noticeable inverse relationship between the cycloalkane and aromatic yields. This trend is consistent with the in situ results shown in Figure 5. The aromatic hydrocarbon yield at 600 °C is also much lower compared to that of in situ hydropyrolysis at 650 °C. It is possible that secondary reactions are occurring after the pyrolysis reactor, leading to cracking and agglomeration among the predominantly phenolic species. Our future work in quantifying the permanent and nonvolatile gases will help elucidate the differences in vapor yields between in situ and ex situ upgrading. The observable yield of cycloalkanes is low at a packed bed reactor temperature of 300 °C, but this might result from the combination of favorable thermodynamics with unfavorable kinetics. The product distribution in Figure 9 shows a wide range of cycloparaffins from methylcyclopentane to bi/tricyclic species, which indicates complete HDO of the lignin hydropyrolysis products flowing through the packed bed of Pd/HZSM-5. The high-temperature hydropyrolysis (650 °C) is effective for breaking down the lignin into phenolic compounds observed in Figure 1, while the packed bed reactor is held at a lower temperature for catalytic upgrading. The effect of lowering the reaction temperature results in effective HDO in the packed bed, but the low temperature results in a low reaction rate that negatively affects the product yield. The main reason for the low yield of cycloalkanes in the ex situ configuration may have been the low yields from noncatalytic hydropyrolysis in the probe (those yields are shown in Figure 1).
■
ASSOCIATED CONTENT
S Supporting Information *
Tables summarizing the hydropyrolysis products of lignin with and without catalysts and the equilibrium constants for hydrogenation of aromatics to cyclohexane. This material is available free of charge via the Internet at http://pubs.acs.org/
■
AUTHOR INFORMATION
Corresponding Author
*Tel: +1 206 221 1580. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors would like to acknowledge Weyerhauser for supporting this study. The authors would also like to acknowledge funding from CAPES (Coordination for the Improvement of Higher Education Personnel) from Brazil. Additionally, the authors thank Dr. Renata Bura, Dr. Shannon Ewanick, and Rodrigo Morales for the poplar lignin feedstock, Dr. Layan Sivithra for his help with preparing the catalysts, and Tatiane Dos Santos Lisboa for her help with operating the pyroprobe GC/MS system.
■
REFERENCES
(1) Huber, G. W.; Iborra, S.; Avelino, C. Synthesis of transportation fuels from biomass: Chemistry, catalysts, and engineering. Chem. Rev. 2006, 106, 4044−4098. (2) Azadi, P.; Inderwildi, O. R.; Farnood, R.; King, D. A. Liquid fuels, hydrogen and chemicals from lignin: A critical review. Renewable Sustainable Energy Rev. 2013, 21, 506−523. (3) Alonso, D. M.; Wettstein, S. G.; Dumesic, J. A. Bimetallic catalysts for upgrading of biomass to fuels and chemicals. Chem. Soc. Rev. 2012, 41, 8075−98. (4) Zhang, M.; Resende, F. L. P.; Moutsoglou, A. Catalytic fast pyrolysis of aspen lignin via Py-GC/MS. Fuel 2014, 116, 358−369. (5) Roy, M. J. Hydrodeoxygenation of lignin model compounds via thermal catalytic reactions. M.S. Thesis, Georgia Institute of Technology, Atlanta, GA, 2012. (6) Kleinert, M.; Barth, T. Towards a lignocellulosic biorefinery: Direct one-step conversion of lignin to hydrogen-enriched biofuel. Energy Fuels 2008, 46, 1371−1379. (7) Patwardhan, P. R.; Brown, R. C.; Shanks, B. H. Understanding the fast pyrolysis of lignin. ChemSusChem 2011, 4, 1629−1636. (8) Nimmanwudipong, T.; Runnebaum, R. C.; Ebeler, S. E.; Block, D. E.; Gates, B. C. Upgrading of lignin-derived compounds: Reactions of eugenol catalyzed by HY zeolite and by Pt/γ-Al2O3. Catal. Lett. 2011, 142, 151−160. (9) Laskar, D. D.; Tucker, M. P.; Chen, X.; Helms, G. L.; Yang, B. Noble-metal catalyzed hydrodeoxygenation of biomass-derived lignin to aromatic hydrocarbons. Green Chem. 2014, 16, 897. (10) Ma, Z.; Troussard, E.; van Bokhoven, J. a. Controlling the selectivity to chemicals from lignin via catalytic fast pyrolysis. Appl. Catal. A Gen. 2012, 423−424, 130−136. (11) Gates, B. C. Catalytic Chemistry; Wiley Series in Chemical Engineering; Wiley: New York, 1992. (12) Roberts, M.; Marker, T. Biomass to Gasoline and Diesel Using Integrated Hydropyrolysis and Hydroconversion; Technical Report; Gas Technology Institute: Des Plaines, IL, 2012. (13) Dayton, D. C.; Carpenter, J.; Farmer, J.; Turk, B.; Gupta, R. Biomass hydropyrolysis in a pressurized fluidized bed reactor. Energy Fuels 2013, 27, 3778−3785.
4. CONCLUSION The hydropyrolysis of lignin was performed with HZSM-5 and Pd/HZSM-5 to assess the hydrodeoxygenation of lignin-derived phenolics to aromatics and cycloalkanes. One of the key findings was that Pd/HZSM-5 produced 44% more aromatic hydrocarbons than HZSM-5 at a catalyst-to-lignin ratio of 20:1, H2 partial pressure of 1.72 MPa, and hydropyrolysis temperature set to 650 °C. The most abundant aromatic compounds were toluene, xylene, and polyaromatic compounds such as naphthalene and methylnaphthalene. The hydropyrolysis temperature was shown to affect the yields of cycloalkanes over Pd/HZSM-5 in H2 through thermodynamic limitations. Although the yield of cycloalkanes is low, this result provides evidence that the bifunctional catalyst was successful in facilitating hydrodeoxygenation reactions of phenolics into cycloalkanes. The ex situ results also represent successful conversion of phenolic compounds into cycloalkanes. However, the low cycloalkane yields can be attributed to the packed bed temperature of 300 °C, which may result in low reaction rates that severely impact the product yield. Further experimentation G
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels (14) Chorkendorff, I.; Niemantsverdriet, J. W. Concepts of Modern Catalysis and Kinetics; Wiley-VCH: New York, 2005. (15) Wang, H.; Male, J.; Wang, Y. Recent advances in hydrotreating of pyrolysis bio-oil and its oxygen-containing model compounds. ACS Catal. 2013, 3, 1047−1070. (16) Barbaro, P.; Liguori, F.; Linares, N.; Marrodan, C. M. Heterogeneous bifunctional metal/acid catalysts for selective chemical processes. Eur. J. Inorg. Chem. 2012, 2012, 3807−3823. (17) Bejblová, M.; Zámostný, P.; Č ervený, L.; Č ejka, J. Hydrodeoxygenation of benzophenone on Pd catalysts. Appl. Catal. A Gen. 2005, 296, 169−175. (18) Bu, Q.; et al. A review of catalytic hydrodeoxygenation of ligninderived phenols from biomass pyrolysis. Bioresour. Technol. 2012, 124, 470−477. (19) Jet Fuels JP-5 and JP-8; Agency for Toxic Substances & Disease Registry: Atlanta, GA. (20) Nimmanwudipong, T.; et al. Selective hydrodeoxygenation of guaiacol catalyzed by platinum supported on magnesium oxide. Catal. Lett. 2012, 142, 1190−1196. (21) Zhu, X.; Lobban, L. L.; Mallinson, R. G.; Resasco, D. E. Bifunctional transalkylation and hydrodeoxygenation of anisole over a Pt/HBeta catalyst. J. Catal. 2011, 281, 21−29. (22) ASTM E1534-93. Standard Test Method for Determination of Ash Content of Particulate Wood Fuels; ASTM International: West Conshohocken, PA, 2013. (23) Mullen, C. A.; Boateng, A. A. Catalytic pyrolysis−GC/MS of lignin from several sources. Fuel Process. Technol. 2010, 91, 1446−1458. (24) Carlson, T.; Tompsett, G.; Conner, W.; Huber, G. Aromatic production from catalytic fast pyrolysis of biomass-derived feedstocks. Top. Catal. 2009, 52, 241−252. (25) Wang, K.; Kim, K. H.; Brown, R. C. Catalytic pyrolysis of individual components of lignocellulosic biomass. Green Chem. 2014, 16, 727. (26) Thangalazhy-Gopakumar, S.; Adhikari, S.; Gupta, R. B. Catalytic pyrolysis of biomass over H + ZSM-5 under hydrogen pressure. Energy Fuels 2012, 26, 5300−5306. (27) Yu, W.; et al. One-step hydrogenation−esterification of furfural and acetic acid over bifunctional Pd catalysts for bio-oil upgrading. Bioresour. Technol. 2011, 102, 8241−6. (28) Zahmakıran, M.; Ozkar, S. Metal nanoparticles in liquid phase catalysis; From recent advances to future goals. Nanoscale 2011, 3, 3462−81. (29) Zahmakıran, M.; Román-Leshkov, Y.; Zhang, Y. Rhodium(0) nanoparticles supported on nanocrystalline hydroxyapatite: Highly effective catalytic system for the solvent-free hydrogenation of aromatics at room temperature. Langmuir 2012, 28, 60−64. (30) Nikolaev, S. a.; et al. Novel gold catalysts for the direct conversion of ethanol into C3+ hydrocarbons. J. Catal. 2013, 297, 296−305. (31) Treesukol, P.; Srisuk, K.; Limtrakul, J.; Truong, T. N. Nature of the metal−support interaction in bifunctional catalytic Pt/H-ZSM-5 zeolite. J. Phys. Chem. B 2005, 109, 11940−5. (32) Cheng, Y.-T.; Jae, J.; Shi, J.; Fan, W.; Huber, G. W. Production of renewable aromatic compounds by catalytic fast pyrolysis of lignocellulosic biomass with bifunctional Ga/ZSM-5 catalysts. Angew. Chem., Int. Ed. Engl. 2012, 51, 1387−90. (33) Zhao, C.; Lercher, J. a. Upgrading pyrolysis oil over Ni/HZSM-5 by cascade reactions. Angew. Chem., Int. Ed. 2012, 51, 5935−40. (34) Hong, D.-Y.; Miller, S. J.; Agrawal, P. K.; Jones, C. W. Hydrodeoxygenation and coupling of aqueous phenolics over bifunctional zeolite-supported metal catalysts. Chem. Commun. (Cambridge, U. K.) 2010, 46, 1038−40. (35) Zhao, C.; Lercher, J. a. Selective hydrodeoxygenation of ligninderived phenolic monomers and dimers to cycloalkanes on Pd/C and HZSM-5 catalysts. ChemCatChem 2012, 4, 64−68. (36) Wan, H.; Chaudhari, R. V.; Subramaniam, B. catalytic hydroprocessing of p-cresol: Metal, solvent and mass-transfer effects. Top. Catal. 2012, 55, 129−139. (37) Ancheyta, J.; Speight, J. G. Hydroprocessing of Heavy Oils and Residua; CRC Press: Boca Raton, FL, 2007.
(38) Fogler, H. S. Elements of Chemical Reaction Engineering; PrentissHall: Upper Saddle River, NJ, 2006.
H
DOI: 10.1021/ef502779s Energy Fuels XXXX, XXX, XXX−XXX
