Modeling of Supercritical Water Gasification of Xylose to Hydrogen

ARTICLE pubs.acs.org/IECR

Modeling of Supercritical Water Gasification of Xylose to Hydrogen-Rich Gas in a Hastelloy Microchannel Reactor Aaron K. Goodwin*,† and Gregory L. Rorrer‡ † ‡

Bend Research Inc., Bend, Oregon 97701, United States Department of Chemical, Biological, and Environmental Engineering, Oregon State University, Corvallis, Oregon 97331, United States ABSTRACT: Microchannel reactors offer high rates of heat transfer that intensify biomass gasification in supercritical water by sustaining the reaction temperature in the presence of endothermic reforming reactions and providing rapid fluid heating. Furthermore, the large ratio of surface area to volume in the microchannels enhances “unintentional” catalytic activity from the reactor wall for reactors comprised of nickel alloys such as Hastelloy. In this study, a parallel-channel Hastelloy C-276 microreactor was used to gasify xylose, a hemicellulose model compound, at 650 °C and 250 bar. The reactor consisted of 14 parallel microchannels, each measuring 127 μm 1000 μm, integrated into a contiguous reactor block using scalable microfabrication techniques. The channels were configured in a serpentine design, which resulted in temperature gradients within the channels during fluid heating isolated from those in subsequent channel passes. Complete conversion of a 4.0 wt % aqueous xylose solution to hydrogen-rich gas was achieved with an average fluid residence time of 1.4 s. Computational fluid dynamics (CFD) modeling was used to simulate xylose gasification in the microchannel reactor and investigate temperature gradients produced by the heat of reaction for xylose gasification. Additional CFD simulations were used to show the effect of short residence times (less than 1.0 s) on the reacting fluid temperature while in the reactor. The results from this study suggest that parallel-channel Hastelloy microreactors potentially offer a unique way to improve biomass gasification by supercritical water.

’ INTRODUCTION Declining fossil fuel reserves and concerns about increasing atmospheric carbon dioxide (CO2) concentrations from the combustion of fossil fuels have motivated a considerable body of research in the field of alternative energy. Biomass is a renewable, CO2-neutral, readily available feedstock that can be thermochemically converted to fuels and chemicals.1 Candidate thermochemical conversion technologies include gasification, combustion, pyrolysis, and liquefaction. Unlike combustion, which typically is focused on heat generation from burning biomass in the presence of oxygen, and pyrolysis, which thermally decomposes biomass in the absence of oxygen to produce biooil, charcoal, and gas, gasification is focused on conversion of biomass to combustible gases by partial combustion with a controlled amount of oxidant. The product gas can be directly combusted for heat generation, fed to a hydrogen fuel cell to generate electricity, or used as a feedstock for the FischerTropsch synthesis for the production of fuels and chemicals. Supercritical water is an excellent platform for biomass gasification because of its high reactivity and ability to solubilize non polar compounds.2 Since water plays the role of solvent and reactant, high-moisture-content biomass can be directly processed in supercritical water without energy-intensive dewatering or drying pretreatment steps needed for other thermochemical conversion technologies.1 Furthermore, biomass gasification in supercritical water generates additional hydrogen (H2) through reforming and produces a compressed product gas is potentially low in carbon monoxide (CO). Recently, several reviews have been published on biomass processing in supercritical water.315 Two strategies for biomass r 2011 American Chemical Society

gasification by supercritical water have emerged: (1) a lowtemperature (350500 °C) heterogeneous catalytic route that produces methane (CH4) as its major gas product and (2) a hightemperature (500750 °C) route that produces H2 as its major gas product. Although catalysts designed for low-temperature gasification using supercritical water have shown the potential to lower the activation energy for biomass gasification reactions and increase the selectivity in the gas products, the stability of these catalyst systems must be addressed for this technology to become feasible.1621 With regard to the high-temperature route, biomass gasification by supercritical water can be substantially intensified by increasing the rate of heat transfer to the reacting fluid.2226 Previous kinetic studies for glucose and xylose—biomass model compounds for cellulose and hemicelluloses—suggest that these biomass constituents rapidly decompose via two competing pathways in near-critical and supercritical water.2735 Below the critical temperature of water, cellulose and hemicellulose sugars dehydrate to furan-type compounds, including 5-(hydroxymethyl)furfural and furfural, because of an ionicdominated reaction environment. However, once above the critical temperature of water, cellulose and hemicellulose react via a retro-aldol condensation to organic acids because of a change from an ionic to a free-radical reaction environment.36,37 Given that furfurals are more difficult to gasify than organic acids, a rapid fluid heating period, which can be produced by high Received: December 15, 2010 Accepted: April 18, 2011 Revised: April 8, 2011 Published: April 18, 2011 7172

dx.doi.org/10.1021/ie102482y | Ind. Eng. Chem. Res. 2011, 50, 7172–7182

Industrial & Engineering Chemistry Research

ARTICLE

Figure 1. View of the Hastelloy C-276 parallel-channel microreactor. The arrows indicate the direction of flow. The microchannel reactor components are (1) the inlet distribution header, (2) thermowells, (3) alignment pinholes used during fabrication, and (4) the outlet header.

rates of heat transfer, will decrease the furfural concentration and, therefore, the time necessary to gasify biomass feedstocks.26 Furthermore, rapid heat transfer is needed to sustain the reaction temperature in the presence of highly endothermic biomass reforming reactions. For example, when xylose is reformed to CO2 and H2, the enthalpy of the reaction at 25 °C and 1.01 bar (based on water in the gas phase) is 287 kJ mol1. In addition to rapid heat transfer, the chemical composition of the reactor material may significantly influence gasification rates and product selectivity. Reactor walls comprised of a high nickel alloy, such as Hastelloy and Inconel, have been shown to catalyze biomass gasification reactions, suppress coke formation, and generate additional H2 through reforming.10,3842 Therefore, a high-surface-area, high-heat-flux reactor fabricated from a nickel alloy has the potential to substantially improve biomass gasification yields and increase H2 product selectivity for biomass gasification in supercritical water. Microchannel reactors are ideal for supercritical water gasification of biomass because of the high rates of heat transfer and high ratios of surface area to volume that are characteristic of micrometer-sized reactor passages. In our previous work, we integrated a series of parallel microchannels into a stainless steel microchannel reactor and reported a significant enhancement for the gasification of glucose.23 In addition, we have used Hastelloy and stainless steel tubular reactors with micrometer-diameter channels to investigate the effect of intensified heat transfer on the cogasification of xylose and phenol and to estimate kinetics for the gasification of xylose under intrinsic reaction conditions.22 Specifically, we were able to completely gasify a 4.0 wt % aqueous solution of xylose to H2 and CO2 with a residence time of less than 1.0 s.27 Although the Hastelloy microtubular reactor demonstrated the greatest enhancement and H2 selectivity for xylose gasification in supercritical water, the reactor is not scalable and is

therefore not feasible for continuous biomass gasification on a larger scale. Biomass gasification in the stainless steel microchannel reactor benefitted from intensified heat transfer to the reacting fluid but lacked the gasification enhancement from a high concentration of nickel in the reactor wall. This study describes the gasification of xylose by supercritical water in a Hastelloy C-276 microchannel reactor at 650 °C and 250 bar. The microreactor has a parallel array of 14 microchannels (127 μm 1000 μm) integrated into a single device. A twodimensional computational fluid dynamics (CFD) model describing heat and fluid transport in a single channel of the microreactor was solved with a simplified kinetic model for xylose gasification by supercritical water. On the basis of a comparison of the CFD modeling results and the experimental results, the Hastelloy microchannel reactor offers a novel way to intensify supercritical water gasification of biomass constituents.

’ MICROCHANNEL REACTOR CONFIGURATION AND TEST LOOP The reactor investigated is a Hastelloy C-276 parallel-channel microreactor designed for continuous gasification of biomass by supercritical water. A transparent three-dimensional schematic of the reactor is presented in Figure 1. The reactor components and dimensions are summarized in Table 1. Briefly, the reactor consists of 14 parallel microchannels that serpentine 15 times vertically through the reactor. Each channel pass is 3.0 cm long, with the exception of the first channel pass, which is 4.0 cm long to minimize a drop in the fluid temperature in subsequent channel passes (because of a temperature gradient caused by the fluid heating period in the first channel pass). A crosssectional view of a single channel is presented in Figure 2. The reacting fluid is distributed to each of the 14 channels by an inlet header and recombined at the exit of the reactor by an outlet 7173

dx.doi.org/10.1021/ie102482y |Ind. Eng. Chem. Res. 2011, 50, 7172–7182


ARTICLE

header. The average surface roughness (Ra) of the reactor walls is less than 100 nm. The continuous-flow microreactor test loop is shown in Figure 3. All wetted parts, including thermocouples and pressure gauges, were constructed of 316 stainless steel. The microchannel reactor described in Figure 1 is sandwiched between two 375 W flat-plate ceramic heaters. The heater block assembly is insulated by a 3.8-cm-thick Fibercraft board (Thermcraft Inc., Winston-Salem, NC). Conduction from the ceramic heating plates to the upper and lower plates of the microchannel reactor was the primary mode of heat transfer. The reactor temperature was controlled by a proportionalintegralderivative controller with a type J thermocouple imbedded into the bottom plate of the microchannel reactor. A feed solution containing 4.0 wt % (41.7 g L1) R-D-xylose (Sigma-Aldrich X1500, >99% purity, CAS 58-86-6, molecular weight 150.13) was fed to the reactor at 25 °C and 250 bar with Table 1. Dimensions of the Components of a Hastelloy C-276 Parallel-Channel Microreactor Reactor Dimensions reactor width

5.0 cm

reactor length

5.0 cm

reactor height reactor volume

1.7 cm 0.9 cm3 Microhannel Dimensions

microchannel width

1000 μm

microchannel height

127 μm

microchannel hydraulic diameter

225 μm

total length of each channel

46.2 cm

number of microchannels

14

number serpentine layers single microchannel volume

15 0.06

Header Dimensions header channel length

4.5 cm

header channel width

0.100 cm

header channel height

0.075 cm

total header volume

0.034 cm3

an Isco Teledyne 260D syringe pump (266-mL capacity) operating at constant flow. All feed solutions were degassed by sparging with helium before use. The liquid feed flow rate to the reactor was varied from 0.5 to 2.5 mL min1. The reactor residence time (τ) was estimated by the continuity equation, τ = VRF(T,P)/v0F0, where VR is the reactor volume, v0 is the volumetric flow rate of the liquid feed solution at the reactor at inlet temperature T0 and system pressure P, F0 is the density of the feed solution at T0 and P, and F(T,P) is the fluid density at the set-point reactor at T and P. The average fluid residence time does not account for changes in the fluid density during heating and quenching of the reacting fluid or any change in the product density from the generation of gas. The hot effluent fluid exiting the reactor was cooled to 25 °C with a shell-and-tube heat exchanger using water as the coolant. The volume in the tube (127 μm diameter) between the reactor outlet and the shell-andtube heat exchanger was approximately 0.003 cm3 and did not significantly contribute to any further reaction outside of the reactor. An adjustable precision back-pressure regulator (stainless steel, KHB1WOA6C2P6000, Swagelok Inc., Solon, OH) stepped down the pressure from 250 to 1.01 bar. The liquid products were collected in the gasliquid separator. The gas products were dried inline, metered with a mass flowmeter [FMA 1800 series, Omega Engineering Inc., Stamford, CT; 0200 standard cubic centimeters per minute (sccm) and 01000 sccm, aluminum/brass body], and then collected into a 2.0 L Tedlar gas collection bag. The volumetric flow rate reading from the mass flowmeter was corrected for the gas composition.

’ ANALYTICAL PROCEDURES The procedures for analysis of the condensed liquid-phase products by high-performance liquid chromatography were described previously.27 The gas products were quantitatively analyzed by an online multiple gas analyzer #1 gas chromatography (GC) instrument (SRI Instruments, Torrance, CA) equipped with a thermal conductivity detector for H2 analysis and a flame ionization detector with a methanizer for CO, CH4, CO2, C2H2, C2H4, and C2H6 analysis. The gas mixture was separated on two columns: a 2 m molecular sieve 13X column and a 2 m silica gel column. The GC oven temperature was held at 40 °C for 3 min, then ramped to 135 °C at a rate of 16 °C min1,

Figure 2. Cross-sectional view of a single channel in the Hastelloy C-276 parallel-channel microreactor. The arrows indicate the direction of flow. The microchannel reactor components are (1) the inlet header, (2) the outlet header, (3) the integrated preheater, and (4) the microchannel segment, which serpentines 15 times vertically. 7174



ARTICLE

Figure 3. Microchannel-reactor test loop (GC = gas chromatography; HPLC = high-performance liquid chromatography).

and held at 135 °C for 2.67 min. A standard gas injection volume (250 μL) was injected three times for each gas sample. The concentration of the gas species was reported as an average of three injections. Gas products were identified by the retention time and quantified by external calibration against a standard gas mixture (Gas Standard No. 19792, Alltech Associates Inc., Bannockburn, IL). Calibration was performed with three 100 μL standard gas injections.

’ CFD MODELING AND SIMULATION CFD modeling was used to simulate xylose gasification by supercritical water in a single channel of the Hastelloy microchannel reactor. Finite-volume CFD simulations were performed using Fluent, version 6.3.26, software operated in double precision mode and interfaced with Gambit, version 2.4.6, software. Model simulations were run on a HP XW4300 workstation using a Linux operating system equipped with a Pentium 4 processor and 8 GB of random access memory (RAM). A representative threedimensional CFD model of the microchannel reactor was too computationally expensive to be feasible; therefore, a twodimensional cross section of a single microchannel was used and is presented in Figure 2. The two-dimensional model geometry is a good approximation because of the high aspect ratio of the channels (i.e., 7.9). The thermophysical properties of water at 250 bar and a function of temperature were used to approximate the properties of the working fluid. However, the density of the reacting fluid was held constant and approximated by the density of water at 650 °C and 250 bar (i.e., the model did not account for density changes due to changes in the reacting fluid temperature or density changes as a result of gas generation). A generic binary diffusion coefficient of 1 108 m2 s1 was used for each species in the reacting mixture. The fluid flow was modeled by the NavierStokes equations for laminar flow. The mass flow rate at the channel inlet boundary ranged from 2.2 to 10.8 kg h1 (Fluent assumes a 1.0 m channel width for twodimensional models). A prescribed pressure of 250 bar was used for the reactor outlet boundary condition, and a no-slip boundary condition was used to describe the velocity at the channel walls. For the general heat equation, a prescribed temperature of 650 °C was used as a boundary condition at the top and bottom

Figure 4. Reaction mechanism for xylose gasification by supercritical water (WSHS = water-soluble humic substance).

of the reactor, where the heater contacts the reactor wall and a zero heat flux boundary condition was used on the sides of the reactor. A prescribed fluid temperature of 25 °C was used at the microchannel entrance, and a zero heat flux boundary condition was used at the microchannel outlet. The kinetic model reaction mechanism for xylose gasification by supercritical water is presented in Figure 4. The kinetic model was previously developed to predict gas yields for supercritical water gasification of xylose at reaction conditions where gasification is dominant and is appropriate for the conditions considered by this study.27 All of the liquid-phase decomposition reactions are pseudo-first-order, and the temperature dependence of the rate constants is described by the Arrhenius equation. Specifically, in the kinetic model, xylose is dehydrated to furfural or reacted to water-soluble humic substances (WSHS), a general term that encompasses all of the liquid decomposition products from xylose and other liquid intermediates. In the kinetic model, furfural is further decomposed to WSHS, which is stoichiometrically gasified to yield CO and H2. The watergas shift and methanation reactions are assumed to be in thermodynamic equilibrium at the temperature of the reacting fluid in the reactor. The temperature dependence on the equilibrium constant for the watergas shift and methanation reactions 7175



ARTICLE

was calculated by the van’t Hoff equation. The fugacity of the gasphase species was used to account for nonideal behavior at 250 bar, was based on the PengRobinson equation of state for the pure gas species, and did not account for mixing interactions in the reacting fluid. Standard heats of reaction were calculated by Hess’s law at the reacting fluid temperature and 1.01 bar. The thermodynamic properties of WSHS were calculated based on the products of the following equation: C5 H10 O5 ðXyÞ f C2 H4 O2 þ C3 H4 O2 þ H2 O

ð1.1Þ

Although WSHS consists of many more liquid intermediate compounds, major liquid decomposition products were acetic acid and acrylic acid27 and are a good approximation of the thermodynamic properties of the bulk term. A structured quadrahedral mesh was used to discretize the area inside the microchannel, and an unstructured quadrahedral mesh was used to discretize the reactor block. A mesh containing 930 000 nodes was found adequate for a mesh-independent solution. This conclusion was reached by refining the mesh based on gradients of pressure and temperature, resolving the problem, and comparing solutions. The Fluent pressure-based solver was used to solve the fluid flow, energy equation, and species transport equations in the microchannel reactor. The SIMPLE algorithm was used for pressure velocity coupling, and a second-order upwind discretization scheme was used for species equations, momentum, and energy. The standard discretization scheme was used for the pressure, and the Green Gauss Node based gradient was chosen for the gradient option. The default under relaxation factors was sufficient for all variables. Simulations were solved by initially solving the steady-state fluid flow and energy equations, commonly referred to as the “cold-flow solution”. Once a steady-state cold-flow solution was reached, the species transport was solved for using the laminar finite-rate model with the stiff chemistry option and the unsteady-state solver with a fixed time step of 1 105 s. The concentrations of H2 and WSHS were monitored at the microchannel outlet, and the simulation was considered converged when the concentration of these species reached an asymptote and no longer changed with each time step. The difference in the mass flux at the inlet and outlet was less than 1 1013 kg s1 for each simulation.

’ RESULTS Effect of the Residence Time on the Gas Yield. Supercritical

water gasification of xylose in the microchannel reactor produced H2-rich gas and a clear aqueous liquid phase. No coking or char was observed for any of the reaction conditions tested. The effect of the residence time on the total gas yield and the percentage of recovered carbon in the gas, also referred to as the carbon gasification efficiency (CGE), are compared with the results from CFD simulations and presented in Figure 5A. The experimental and predicted total gas yields based on mass and CGE were independent of residence time between 1.4 and 7.1 s; average values are reported in Table 2. With respect to the CGE and H2 yield, defined as the moles of H2 generated per mole of xylose fed, CFD simulation results were in good agreement with the experimental values and indicated near-complete gasification of the feed substrate. Experimental and model results for the overall gas yield, on a mass basis, were greater than 100% for residence times greater than 1.0 s due to xylose reforming. Higher overall gas yields from CFD simulations are due to the presence

Figure 5. Composition and yield of gas and liquid products versus residence time for xylose gasification at 650 °C and 250 bar in the microchannel reactor: (A) gas yield, CGE, and H2 yield; (B) liquid intermediate concentration; and (C) gas composition. The filled markers are predictions from CFD simulations, and the outline markers are experimental results.

of CH4 in the experimental gas composition not predicted by the model. Although experimental residence times did not go below 1.4 s, CFD simulations predicted a sharp decrease in the overall gas yield and CGE for residence times less than 1.4 s because of incomplete gasification. Measured and predicted liquid intermediate concentrations versus residence times are presented in Figure 5B. Although no liquid intermediates were detected for xylose gasification experiments, CFD simulations predicted a sharp increase in the furfural and WSHS concentrations below a 1.4 s residence time. The effect of the residence time on the H2 yield, for xylose gasification is presented in Figure 5A. The theoretical H2 yield for 7176



ARTICLE

Table 2. Average Values for Carbon Recovery in the Gas (CGE), Gas Yield, and H2 Yield for Gasification of Xylose and Xylan at 650 °C and 250 bar

xylose xylose (model)

residence

CGE

gas yield

H2

CO2

CH4

CO

C2H6

time (s)

(%)

(g of gas per g of xylose fed; %)

(mol %)

(mol %)

(mol %)

(mol %)

(mol %)

96 ( 1.6 99 ( 1.6

140 ( 1.6 156 ( 3.0

65.0 ( 1.1 66.1 ( 0.1

30.0 ( 0.7 32.4 ( 0.2

3.9 ( 0.6 0.1 ( 0.0

0.8 ( 0.1 1.4 ( 0.3

0.3 ( 0.1

1.47.1 1.47.1

Figure 6. Sample streamline plot through the microchannel reactor for a 1.4 s residence time, a reactor temperature of 650 °C, and a pressure of 250 bar. The streamline represents the path taken by a massless particle released from the center of the reactor inlet. Plots of the following process variables for this residence time are plotted on this streamline: (A) entire channel (not to scale); (B) dashed rectangular section in Figure 2 (to scale).

xylose gasification is 5 mol based on the concentration of H2 in the feed stock and 10 mol based on reforming. The additional 5 mol of H2 generated from reforming are a result of the water gas shift reaction. The experimental H2 yield was independent of the residence time and averaged 8.9 ( 1.0 mol of H2 generated per 1 mol of xylose fed. For the range of experimental residence times tested, the H2 yield based on CFD simulations was 9.7 ( 0.2 mol of H2 generated per 1 mol of xylose reacted. For residence times less than 1.0 s, CFD simulations predicted a substantial decrease in the H2 yield due to incomplete gasification of the feed substrate, given that the gas-phase reactions are in thermodynamic equilibrium. Effect of the Residence Time on the Gas Composition. Major gas products for the gasification of xylose in the microchannel reactor consisted of H2 and CO2, and minor gas products included CO, CH4, and C2H6. In addition to the major and minor gas products, C2H4 and C2H2 were tested for but not identified in the product gas. The effect of the residence time on the gas composition for xylose gasification is presented in Figure 5C. The experimental and predicted gas compositions were generally in good agreement and independent of the residence time, which ranged from 1.4 to 7.1 s. One notable difference between CFD predictions and experimental values was the concentration of CH4 in the product gas. On the basis of the equilibrium values for the watergas shift and methanation reactions, an average CH4 concentration of 0.1 ( 0.0 was predicted, whereas the average CH4 concentration measured over the range of residence times tested was 3.9 ( 0.6. For residence times of less than 1.0 s, CFD simulations predicted gas compositions similar to those with longer residence times with the exception of a sharp increase in the CO concentration, up to 5.3% for a 0.7 s residence time. Although C2H6 was measured in the product gas, the kinetic model did not include a pathway for C2H6 formation, so it was not reported in Table 2. CFD Simulation: Microchannel Temperature Profile during Xylose Gasification. Because biomass gasification reaction rates and equilibrium gas compositions are sensitive to the

reaction temperature,22,4348 CFD simulations were used to gain insight into temperature gradients within the reactor. The temperature profile through the reactor is dependent on the fluid Reynolds number and the feed concentration because of the endothermic nature of reforming reactions. To address the effect from heat of reaction on the temperature of the reacting fluid, the temperature profile through a single microchannel was simulated with and without reactions at 1.4 and 7.1 s residence times. The temperature profile through the microchannel was calculated on a streamline that was released at the center of the entrance of the channel. A sample plot showing the microchannel geometry and the streamline through the microchannel for a 1.4 s residence time is presented in Figure 6. The streamline was located at the center of the channel or the part of the flow with the highest velocity. A plot showing the fluid velocity on the streamline for 1.4 and 7.1 s residence times is presented in Figure 7. The vertical lines in the figure represent the 15 serpentine channel passes in each microchannel. In both cases, a decrease in the fluid velocity is observed when the fluid goes around the 180° turn between serpentine channel passes, and the effect becomes more pronounced at shorter residence times. A velocity vector plot also presented in Figure 7C shows the flow field for xylose gasification for a 1.4 s residence time. From the velocity vector plot, it is apparent that an eddy and a small vortex form in the transition area between channel passes. The Reynolds number has a considerable effect on the flow field between channel passes and therefore affects the fluid residence time distribution. A residence time distribution study and CFD model could be used to better understand how this phenomenon will affect hydrodynamics in the microchannel and the ultimately product distribution; however, such an investigation is beyond the scope of the current study. The nonreacting- or cold-flow streamline temperature profile through the microchannel is compared with that of the reacting flow and presented in Figure 8. For a 7.1 s residence time, the temperature of the reacting fluid rapidly increases from 25 °C 7177



ARTICLE

Figure 7. Streamline velocity profile through the reactor for xylose gasification at 650 °C and 250 bar, showing (A) the velocity through the entire microchannel, (B) the velocity profile around the 180° turn between serpentine layers, and (C) the velocity vector plot for the region between serpentine layers for a 1.4 s residence time. The velocity vector plot is represented at two different velocity scales for the same residence time.

Figure 8. Streamline temperature profile through the reactor at a reactor temperature of 650 °C and a pressure of 250 bar for (A) a 7.1 s residence time and (B) a 1.4 s residence time. The temperature prediction for xylose gasification was based on a 4.0 wt % feed solution. The area between vertical dotted lines corresponds to the 15 serpentine channel passes in a single microchannel.

past the critical temperature of water within the first 200 μm of the first channel pass. Once past the critical temperature of water, xylose dehydration to furfural is no longer favored because of a change in the reaction environment from ionic to free radical. At the end of the first channel pass, the temperature of the fluid

fluctuates about 10 °C between subsequent channel passes. Generally, a decrease in the reaction temperature is predicted when the fluid is flowing in the direction of the microchannel entrance because of a temperature gradient as a result of the fluid heating period. The microchannel temperature profiles are 7178


Industrial & Engineering Chemistry Research similar for the cold and reacting flows, but a slight drop in the fluid temperature, less than 1.2 °C, is observed in the first seven channel passes due to the heat of reaction from WSHS being gasified to CO and H2. After the first seven passes, the temperatures of the cold and reacting flows are similar because of complete gasification of the feed substrate. For a 1.4 s residence time, the fluid temperature rapidly increases to a maximum of 646 °C by the end of the first channel but decreases to 618 °C by the end of the second channel. The lowest average temperature was predicted for channel passes located at the middle of the reactor because these channels are farthest from the reactor heaters. Through a comparison of the cold- and reactingflow streamline temperatures for a 1.4 s residence time, it is evident that endothermic xylose gasification reactions have a greater effect on the temperature of the reacting fluid than longer residence times. However, a maximum temperature drop in the reacting fluid of only 9.6 °C was predicted compared to the temperature drop in the coldflow solution. In addition, the average streamline temperature through the microchannel was 632 °C for the cold flow and 626 °C for the reacting flow; therefore, it is unlikely that the temperature difference will substantially affect the gas yield or H2 selectivity for xylose gasification. CFD Simulation: Liquid Intermediate Formation. To investigate the effects of the residence time and reaction temperature on liquid and gas intermediate formation in the microchannel reactor, streamline concentrations of WSHS, furfural, CO2, and H2 were calculated for 1.4 and 7.1 s residence times. The results are presented in Figure 9. Given that xylose is completely reacted in the first 3 mm of the first channel for a 1.4 s residence time, the concentrations of WSHS and furfural were used to represent liquid intermediates in the reacting flow. The maximum concentrations for WSHS and furfural were similar for both residence times. However, for the 7.1 s residence time, furfural was completely reacted to WSHS by the first 10 cm (three channel passes) in the microchannel reactor and WSHS was completely gasified to CO and H2 in 20 cm (seven channel passes) of the microchannel reactor. In comparison, for a 1.4 s residence time, a small concentration of WSHS was predicted at the reactor outlet. As shown in Figure 5B, this is contrary to our experimental results, in which no liquid intermediates were measured at the outlet. It should be noted that the average predicted concentration of WSHS at the outlet will be lower than the streamline value, given that the streamline velocity is near the maximum velocity of the fluid and not an average velocity. A streamline with a lower velocity will have a longer residence time, and therefore the concentration of WSHS will be less. For an average residence time of 7.1 s, the streamline concentrations of CO2 and H2 increased for the first 15 cm of the microchannel until they reached an asymptote, indicating that all of the liquid intermediates had been gasified. Small fluctuations in the H2 and CO2 concentrations on the asymptote are due to changes in the equilibrium of the watergas shift and methanation reactions caused by the fluctuating reaction temperature. At an average fluid residence time of 1.4 s, the concentrations of H2 and CO2 increase throughout the entire length of the microchannel and approach the species concentration asymptotes predicted for a 7.1 s residence time. CFD Simulation of Nonreacting Flow for Residence Times of Less Than 1.0 s. Considering CFD simulations within the experimental average range for fluid residence times in this study, it is evident that the fluid velocity has a greater effect on the

ARTICLE

Figure 9. Streamline concentration profiles for (A) liquid decomposition products and (B) gas products from CFD simulations for gasification of 4.0 wt % xylose at 650 °C and 250 bar at 1.4 and 7.1 s residence times.

reacting fluid temperature than do endothermic xylose reforming reactions. Therefore, CFD simulations were used to determine the effect of residence times of less than 1 s on the streamline temperature of the cold flow. The results are presented in Figure 9 for residence times ranging from 0.2 to 1.0 s. Temperature fluctuations between channel passes were similar for residence times between 0.5 and 1.0 s, but the effect of decreasing residence time was a lower overall average streamline temperature through the length of the channel. The lowest average streamline temperature, 474 °C, was predicted for a 0.2 s residence time.

’ DISCUSSION This paper describes the continuous rapid gasification of xylose by supercritical water in a novel Hastelloy parallel-channel microreactor at conditions at which gasification is dominant. The reactor contains 14 parallel microchannels integrated into a single device constructed from Hastelloy C-276. CFD modeling was used to simulate the fluid-, heat-, and mass-transport properties in a single microchannel. This modeling aided in understanding the effect of the residence time on product formation and temperature gradients within the reactor. The microchannel reactor proved beneficial in reforming xylose to H2-rich gas because of the high rates of heat transfer to the reacting fluid. It is well established that increasing the rate of heat transfer during biomass gasification with supercritical water increases the gasification efficiency.25,26,49 In our previous work, we exploited this concept by gasifying xylose in a Hastelloy microtubular reactor.22,27 This reactor configuration provided the heat transfer needed for isothermal gasification of xylose at 7179


Industrial & Engineering Chemistry Research fast residence times and resulted in complete gasification of a 4.0 wt % xylose solution to H2-rich gas at a residence time of less than 1.0 s. In this study, a Hastelloy microchannel reactor was used instead of a microtubular reactor. The microchannel reactor offers two significant advantages over the microtubular reactor. The first advantage is scalable microfabrication. Depending on the desired reactor volume, complete channels and channel layers can be added to the existing device to increase the reactor volume. Because, theoretically, the physics in individual microchannels will remain identical, the device performance will not be affected appreciably by scaling up the reactor volume. However, two technical issues are associated with scale-up of this particular microreactor design: (1) the formation of temperature gradients in the vertical direction of the reactor with the addition of channel passes, as seen in Figure 8B, and (2) an equal microchannel flow distribution with the addition of complete microchannels. Temperature gradients in the vertical direction from an increase in the number of shims (and, therefore, channel passes) can be addressed by integrating periodic internal heaters into the shim stack section of the reactor. Alternatively, a portion of the product gas could be internally combusted to sustain an isothermal reaction temperature. Combustion microchannels have previously been successfully integrated into microreactors for catalytic steam reforming of methane to self-sustain the reaction temperature.50,51 The issue of microchannel flow uniformity upon scaling-up the number of microchannels can be addressed by the use of CFD to modify the header design.52,53 The second advantage the microchannel reactor has over the microtubular reactor is enhanced heat transfer as a result of a higher surface area to volume ratio. For example, the heattransfer coefficient, h, for supercritical water at 650 °C and 250 bar in the microchannel reactor was 2691 W m2 K1 at a Reynolds number of 87, which corresponded to an average fluid residence time of 2.4 s. On the basis of the same residence time and reaction conditions, the heat-transfer coefficient in a 508μm-diameter Hastelloy microtubular reactor used in previous kinetic studies was 528 W m2 K1 at a Reynolds number of 1130, a 5-fold higher heat-transfer coefficient.22,27 Although CFD simulations predict small temperature fluctuations between microchannel passes (as shown in Figure 8) due to a temperature gradient caused by the fluid heating period, the rate of heat transfer to fluid is sufficient to drive endothermic xylose reforming reactions. This is clearly shown in Figure 8 by comparing the temperature profile for the reacting and nonreacting flows. Because increasing the feed concentration will increase the gasification efficiency but intensify the endothermic heat of reaction, increasingly high rates of heat transfer are needed to sustain the reaction temperature for high feed concentrations. In addition to intensified heat transfer, xylose gasification in the microchannel reactor benefits from enhanced catalytic activity from a high surface area to volume ratio due to the presence of iron and nickel in the reactor walls. The microchannel reactor was fabricated from Hastelloy C-276, a nickel alloy that contains nominally 5060% nickel, 15% chromium, 16% molybdenum, and 47% iron. When exposed to supercritical water, nickel alloys such as Hastelloy C-276 develop an outer oxide layer at the surface consisting of nickel oxide and an inner layer rich in chromium, oxygen, and nickel clusters.54,55 It has been shown that the exposed nickel on the surface of the reactor wall substantially increases gas yields for supercritical water gasification of cellulose and lignin by promoting the

ARTICLE

watergas shift reaction, methanation reaction, and liquid intermediate decomposition reactions.10,3841 In addition, increasing the ratio of the catalyst surface area to biomass weight has been shown to improve gasification yields and H2 selectivity for cellulose gasification.42 Therefore, the low concentration of CO in the product gas, less than 1.0%, and an H2 yield approaching the theoretical limit for reforming xylose were attributed to the enhanced promotion of the watergas shift and methanation reactions from a high concentration of nickel exposed on the surface of the reactor walls and the high surface area to volume ratio. Although previous studies have shown that nickel alloy reactors such as Inconel and Hastelloy may be susceptible to corrosion, which has been reported to influence reaction kinetics,40 no such effects were observed in the current study. The result from xylose gasification by supercritical water in the microchannel reactor was complete gasification of a 4.0 wt % xylose solution within a 1.4 s residence time with no observable coke formation. Additionally, given that no intermediates were measured in the liquid products and the gas composition and yield were independent of the residence time, it is likely that the reaction was approaching equilibrium. This is further substantiated by the close agreement between the experimental and model gas compositions and yields because the watergas shift and methanation reactions are in thermodynamic equilibrium in the kinetic model. However, one significant difference between the experimental and model results is the concentration of CH4 in the product gas. The model prediction of 0.1% CH4 is substantially less than the average experimentally measured value of 4.0 wt %. The higher experimental CH4 concentration is likely a direct decomposition product from WSHS (Figure 4), which was not accounted for by the kinetic model.27,56 A consequence of the higher CH4 concentration in the product gas was a decrease in the H2 yield from a theoretical value of 10 to 8.9, given that each 1 mol of reformed CH4 produces 3 mol of H2. Aside from our previous studies with the microtubular reactor, there do not appear to be any reported investigations that noncatalytically gasify xylose or xylan by supercritical water at conditions that promote gasification. Previous catalytic studies in sealed batch reactors have reported results of H2 yields of less than 1.0 and carbon recoveries of less than 70% for xylan gasification just above the critical temperature of water for a fixed reaction time of about 20 min.57,58 Although these studies were mechanistic in nature and did not necessarily focus on gasification optimization, the current work clearly shows that microchannel reactors offer a unique alternative to catalytic biomass gasification for process intensification. Furthermore, on the basis of model kinetic data, an increase in the reaction time from 1.0 s to 1.0 min would reduce the reaction temperature to approximately 500 °C. Complete gasification may be achievable at shorter fluid residence times; however, CFD simulations of the nonreacting flow, corresponding to residence times of less than 1.0 s (seen in Figure 10), indicate a substantial drop in the fluid temperature due to a temperature gradient from heating of the reacting fluid to 650 °C from 25 °C at the entrance of the reactor. Although the fluid temperature at the entrance of the microchannel reactor was set to 25 °C in the CFD model simulations, the actual fluid temperature at the channel entrance will be higher because of the fluid residence time in the reactor header, which makes up approximately 3.7% of the total residence time in the microchannel reactor. In addition, the channel entrance temperature 7180



Figure 10. Comparison of cold-flow streamline temperatures through the microchannel at 650 °C and 250 bar at residence times below 1.0 s.

will be affected by the fluid velocity and position of the channel down the length of the header. Because the reactor is a contiguous block and there is no way to measure the fluid temperature at a specific channel inlet, a 25 °C boundary condition was used.

’ CONCLUSION In conclusion, xylose was successfully gasified to H2-rich gas by supercritical water in a novel Hastelloy microchannel reactor. The reactor’s large ratio of surface area to volume provided high rates of heat transfer and the high concentration of nickel in the reactor wall intensified the catalytic activity, resulting in complete gasification of a 4.0 wt % aqueous solution of xylose in a 1.4 s residence time. Major gas products consisted of CO2 and H2, and the average H2 yield was 8.9 mol of H2 generated per 1 mol of xylose fed. Results from CFD simulations suggest that endothermic xylose gasification reactions have little on the reaction temperature. However, for this specific reactor configuration, the reacting fluid temperature is sensitive to the average fluid residence time because of the temperature gradient from the fluid heating period. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This research was supported by Bend Research Inc., Bend, OR, and by the U.S. Army under the Tactical Energy Systems program administered through the Oregon Nanoscience and Microtechnologies Institute (ONAMI). The CFD modeling software and workstation was provided by ONAMI. ’ REFERENCES (1) Zhang, L.; Xu, C.; Champagne, P. Overview of Recent Advances in Thermo Chemical Conversion of Biomass. Energy Convers. Manage. 2010, 51, 969–982. (2) Loppinet-Serani, A.; Aymonier, C.; Cansell, F. Supercritical Water for Environmental Technologies. J. Chem. Technol. Biotechnol. 2010, 85, 583–589. (3) Demirbas, A. Hydrogen Production from Biomass via Supercritical Water Gasification. Energy Sources, Part A 2010, 32, 1342–1354. (4) Matsumura, Y.; Minowa, T.; Potic, B.; Kersten, S.; Prins, W.; van Swaaij, W.; Beld, B.; Elliott, D.; Neuenschwander, G.; Kruse, A.; Antal,

ARTICLE

M. J., Jr. Biomass Gasification in Near- and Supercritical Water: Status and Prospects. Biomass Bioenergy 2005, 29, 269–292. (5) Calzavara, Y.; Joussot-Dubien, C.; Boissonnet, G.; Sarrade, S. Evaluation of biomass Gasification in Supercritical Water Process for Hydrogen Production. Energy Convers. Manage. 2005, 46, 615–631. (6) Loppinet-Serani, A.; Aymonier, C.; Cansell, F. Current and Foreseeable Applications of Supercritical Water for Energy and the Environment. ChemSusChem 2008, 1, 486–503. (7) Navarro, R. M.; Sanchez-Sanchez, M. C.; Alvarez-Galvan, M. C.; del Valle, F.; Fierro, J. L. G. Hydrogen Production from Renewable Sources: Biomass and Photocatalytic Opportunities. Energy Environ. Sci. 2009, 2, 35–54. (8) Peterson, A. A.; Vogel, F.; Lachance, R. P.; Fr€oling, M.; Antal, M. J., Jr.; Tester, J. Thermochemical Biofuel Production in Hydrothermal Media: A Review of Sub- and Supercritical Water Technologies. Energy Environ. Sci. 2008, 1, 32–65. (9) Kruse, A. Supercritical Water Gasification. Biofuels, Bioprod. Biorefin. 2008, 2, 415–437. (10) Guo, Y.; Wang, S. Z.; Xu, D. H.; Gong, Y. M.; Ma, H. H.; Tang, X. Y. Review of Catalytic Supercritical Water Gasification for Hydrogen Production from Biomass. Renewable Sustainable Energy Rev. 2010, 14, 334–343. (11) Kruse, A.; Vogel, H. Heterogeneous Catalysis in Supercritical Media: 2. Near-Critical and Supercritical Water. Chem. Eng. Technol. 2008, 31, 1241–1245. (12) Savage, P. E. A Perspective on Catalysis in Sub- and Supercritical Water. J. Supercrit. Fluid 2009, 47, 407–414. (13) Vogel, F. In Handbook of Green Chemistry: Green Catalysis; Anastas, P. T., Crabtree, R. H., Eds.; Wiley-VCH Verlag GmbH & Co KGaA: Berlin, Germany, 2009; Vol. 2; pp 281324. (14) Ding, Z. Y.; Frisch, M. A.; Li, L.; Gloyna, E. F. Catalytic Oxidation in Supercritical Water. Ind. Eng. Chem. Res. 1996, 35, 3257–3279. (15) Savage, P. E. Heterogeneous catalysis in supercritical water. Catal. Today 2000, 62, 167–173. (16) Byrd, A. J.; Gupta, R. B. Stability of Cerium-Modified γAlumina Catalyst Support in Supercritical Water. Appl. Catal., A 2010, 381, 177–182. (17) Furusawa, T.; Sato, T.; Saito, M.; Ishiyama, Y.; Sato, M.; Itoh, N.; Suzuki, N. The Evaluation of the Stability of Ni/MgO Catalysts for the Gasification of Lignin in Supercritical Water. Appl. Catal., A 2007, 327, 300–310. (18) Osada, M.; Sato, O.; Arai, K.; Shirai, M. Stability of Supported Ruthenium Catalysts for Lignin Gasification in Supercritical Water. Energy Fuels 2006, 20, 2337–2343. (19) Tomita, K.; Oshima, Y. Stability of Manganese Oxide in Catalytic Supercritical Water Oxidation of Phenol. Ind. Eng. Chem. Res. 2004, 43, 7740–7743. (20) Yu, J.; Savage, P. E. Catalyst Activity, Stability, and Transformations during Oxidation in Supercritical Water. Appl. Catal., B 2001, 31, 123–132. (21) Aki, S. N. V. K.; Ding, Z.; Abraham, M. A. Catalytic Supercritical Water Oxidation: Stability of Cr2O3 Catalyst. AIChE J. 1994, 2, 1995–2004. (22) Goodwin, A. K.; Rorrer, G. L. Conversion of Xylose and XylosePhenol Mixtures to Hydrogen-Rich Gas by Supercritical Water in an Isothermal Microtube Flow Reactor. Energy Fuels 2009, 23, 3818–3825. (23) Goodwin, A. K; Rorrer, G. L. Conversion of Glucose to Hydrogen-Rich Gas by Supercritical Water in a Microchannel Reactor. Ind. Eng. Chem. 2008, 47, 4106–4114. (24) Lu, Y. J.; Guo, L. J.; Ji, C.; Zhang, X.; Hao, X.; Yan, Q. Hydrogen Production by Biomass Gasification in Supercritical Water: A Parametric Study. Int. J. Hydrogen Energy 2006, 31, 822–31. (25) Watanabe, M.; Aizawa, Y.; Iida, T.; Levy, C.; Aida, T. M.; Inomata, H. Glucose Reactions within the Heating Period and the Effect of Heating Rate on the Reactions in Hot Compressed Water. Carbohydr. Res. 2005, 340, 1931–39. 7181


Industrial & Engineering Chemistry Research (26) Sinag, A.; Kruse, A.; Rathert, J. Influence of the Heating Rate and the Type of Catalyst on the Formation of Key Intermediates and on the Generation of Gasses During Hydropyrolysis of Glucose in Supercritical Water in a Batch Reactor. Ind. Eng. Chem. Res. 2004, 43 502–508. (27) Goodwin, A. K.; Rorrer, G. L. Reaction Rates for Supercritical Water Gasification of Xylose in a Microtubular Reactor. Chem. Eng. J. 2010, 163, 10–21. (28) Matsumura, Y.; Yanachi, S.; Yoshida, T. Glucose Decomposition Kinetics in Water at 25 MPa in the Temperature Range of 448673 K. Ind. Eng. Chem. Res. 2006, 45, 1875–1879. (29) Lee, I.; Kim, M.; Ihm, S. Gasification of Glucose in Supercritical Water. Ind. Eng. Chem. Res. 2002, 41, 1182–1188. (30) Kabyemela, B. M.; Adschiri, T.; Malaluan, R. M.; Arai, K. Glucose and Fructose Decomposition in Subcritical and Supercritical Water: Detailed Reaction Pathway, Mechanisms and Kinetics. Ind. Eng. Chem. Res. 1999, 38, 2888–2895. (31) Kabyemela, B. M.; Adschiri, T.; Malaluan, R. M.; Arai, K. Kinetics of Glucose Epimerization and Decomposition in Subcritical and Supercritical Water. Ind. Eng. Chem. Res. 1997, 36, 1552–1558. (32) Qi, J.; Xiuyang, L. Kinetics of Non-Catalyzed Decomposition of D-xylose in High Temperature Liquid Water. Chin. J. Chem. Eng. 2007, 15, 666–669. (33) Antal, M. J.; Leesomboom, T.; Mok, W. S.; Richards, G. N. Mechanism of formation of 2-furaldehyde from D-xylose. Carbohydr. Res. 1991, 217, 71–85. (34) Antal, M. J.; Mok, W. S.; Richards, G. N. Mechanism of formation of 5-(hydroxymethyl)-2-furaldehyde from D-fructose and sucrose. Carbohydr. Res. 1990, 91–109. (35) Antal, M. J.; Mok, W. S.; Richards, G. N. Four-carbon model compounds for the reactions of sugars in water at high temperature. Carbohydr. Res. 1990, 111–115. (36) Watanabe, M.; Aizawa, Y.; Iida, T.; Levy, C.; Aida, T. M.; Inomata, H. . Glucose Reactions within the Heating Period and the Effect of Heating Rate on the Reactions in Hot Compressed Water. Carbohydr. Res. 2005, 340, 1931–1939. (37) Sasaki, M.; Hayakawa, T.; Arai, K.; Adschiri, T. Hydrothermal Reactions and Techniques; The Proceedings of the Seventh International Symposium on Hydrothermal Reactions; World Scientific Publishing Co. Pte. Ltd.: River Edge, NJ, 2003; pp 169176. (38) Kersten, S. R. A.; Potic, B.; Prins, W.; Van Swaaij, W. P. M. Gasification of Model Compounds and Wood in Hot Compressed Water. Ind. Eng. Chem. Res. 2006, 45, 4169–4177. (39) Antal, J. A. J.; Allen, S. G.; Schulman, D.; Xu, X. Biomass Gasification in Supercritical Water. Ind. Eng. Chem. Res. 2000, 39, 4040–4053. (40) Yu, D.; Aihara, M.; Antal, J. A. J. Hydrogen Production by Steam Reforming Glucose in Supercritical Water. Energy Fuels 1993, 7, 574–577. (41) DiLeo, G. J.; Neff, M. E.; Savage, P. E. Gasification of Guaiacol and Phenol in Supercritical Water. Energy Fuels 2007, 21, 2340–2345. (42) Resende, F. L. P.; Savage, P. E. Effect of Metals on Supercritical Water Gasification of Cellulose and Lignin. Ind. Eng. Chem. Res. 2010, 49, 2694–2700. (43) Resende, F. L. P.; Fraley, S.; Berger, M.; Savage, P. E. Previous Noncatalytic Gasification of Lignin in Supercritical Water. Energy Fuels 2008, 22, 1328–1334. (44) Resende, F. L. P.; Savage, P. E. Expanded and Updated Results for Supercritical Water Gasification of Cellulose and Lignin in MetalFree Reactors. Energy Fuels 2009, 23, 6213–6221. (45) Resende, F. L. P.; Neff, M. E.; Savage, P. E. Noncatalytic Gasification of Cellulose in Supercritical Water. Energy Fuels 2007, 21, 3637–3643. (46) Lu, Y.; Guo, L.; Zhang, X.; Yan, Q. Thermodynamic modeling and analysis of biomass gasification for hydrogen production in supercritical water. Chem. Eng. J. 2007, 131, 233–244. (47) Tang, H.; Kitagawa, K. Supercritical Water Gasification of Biomass: Thermodynamic Analysis with Direct Gibbs Free Energy Minimization. Chem. Eng. J. 2005, 106, 261–267.

ARTICLE

(48) Yan, Q.; Guo, L.; Lu, Y. Thermodynamic Analysis of Hydrogen Production from Biomass Gasification in Supercritical Water. Energy Convers. Manage. 2006, 47, 1515–1528. (49) Matsumura, Y.; Harada, M.; Nagata, K.; Kikuchi, Y. Effect of Heating Rate of Biomass Feedstock on Carbon Gasification Efficiency in Supercritical Water Gasification. Chem. Eng. Commun. 2006, 193, 649–659. (50) Cremmers, C.; Pelz, A.; Stimming, U.; Haas-Santo, K.; Goerke, O.; Pfeifer, P.; Schubert, K. Microstructured Methane Steam Reformer with Integrated Catalytic Combustor. Fuel Cells 2007, 7, 91–98. (51) Stefanidis, G. D.; Vlachos, D. G.; Kaisare, N. S.; Kaisare, N. S.; Maestri, M. Methane Steam Reforming at Microscales: Operation Strategies for Variable Power Output at Millisecond Contact Times. AIChE J. 2009, 55, 180–191. (52) Mettler, M. S.; Stefanidis, G. D.; Vlachos, D. G. Scale-out of Microreactor Stacks for Portable and Distributed Processing: Coupling of Exothermic and Endothermic Processes for Syngas Production. Ind. Eng. Chem. Res. 2010, 49, 10942–10955. (53) Saber, M.; Commenge, J. M.; Falk, L. Microreactor Numbering-up in Multi-scale Networks for Industrial-Scale Applications: Impact of Flow Maldistribution on the Reactor Performances. Chem. Eng. Sci. 2010, 65, 372–379. (54) Zhang, Q.; Tang, R.; Yin, K.; Luo, X.; Zhang, L. Corrosion Behavior of Hastelloy C-276 in Supercritical Water. Corros. Sci. 2009, 51, 2092–2097. (55) Boukis, N.; Habicht, W.; Franz, G.; Dinjus, E. Behavior of Nibase Alloy 625 in MethanolSupercritical Water Systems. Mater. Corros. 2003, 54, 326–330. (56) Resende, F. L. P.; Savage, P. E. Kinetic Model for Noncatalytic Supercritical Water Gasification of Cellulose and Lignin. AIChE J. 2010, 56, 2412–2420. (57) Yoshida, T.; Matsumura, Y. Gasification of Cellulose, Xylan, and Lignin Mixtures in Supercritical Water. Ind. Eng. Chem. Res. 2001, 40, 5469–5474. (58) Guo, L. J.; Lu, Y. J.; Zhang, X. M.; Ji, C. M.; Guan, Y.; Pei, A. X. Hydrogen Production by Biomass Gasification in Supercritical Water: A Systematic Experimental and Analytical Study. Catal. Today 2007, 129, 275–286.

7182


Modeling of Supercritical Water Gasification of Xylose to Hydrogen

Recommend Documents