Steam

Article pubs.acs.org/EF

Hydrogen Production from High Temperature Pyrolysis/Steam Reforming of Waste Biomass: Rice Husk, Sugar Cane Bagasse, and Wheat Straw Qari M. K. Waheed and Paul T. Williams* Energy Research Institute, University of Leeds, Leeds, LS2 9JT, United Kingdom ABSTRACT: Hydrogen production from pyrolysis, steam reforming, and catalytic steam reforming of sugar cane bagasse, wheat straw, and rice husk were investigated using a two stage pyrolysis-reforming system. Biomass samples were pyrolyzed in the first stage, and the volatiles and liquids were reformed in the second stage in the presence of steam. During all experiments, a high temperature of 950 °C was chosen for both the pyrolysis and reforming stages. As compared to low temperatures, pyrolysis/ reforming carried out at higher temperature showed higher gas yields, particularly hydrogen gas yield. In addition, dolomite and 10 wt % Ni-dolomite were used to investigate the catalytic steam reforming of the biomass. In terms of hydrogen production, steam reforming using 10 wt % Ni-dolomite was the most effective, producing 25.44, 25.41, and 24.47 mmol of hydrogen per gram for rice husk, sugar cane bagasse, and wheat straw, respectively. The amount of deposited carbon on the reacted catalyst was from 1.31 wt % to 10.13 wt % and was in the form of amorphous and graphitic carbon. Relatively lower carbon deposits were found on the 10 wt % Ni-dolomite as compared to the calcined dolomite. XRD analysis of the reacted catalyst showed the presence of Ni, NiO, and NiMgO2 phases for the 10 wt % Ni-dolomite. The highest hydrogen yield of 25.44 mmol g−1 was obtained from rice husk, and the highest hydrogen concentration in the gas mixture was found to be 59.14 vol % from rice husk using 10 wt % Ni-dolomite.

1. INTRODUCTION If produced from sustainable and renewable sources such as biomass, hydrogen could be a possible solution for future energy sustainability and a reduction of the impacts due to climate change. Hydrogen has many advantages as an energy carrier, having three times higher energy density per unit mass as compared to gasoline and producing only water vapor as a product of combustion.1,2 According to the U.S. Department of Energy, more than 50% of global hydrogen production comes from the steam reforming of natural gas.3 However, the production of hydrogen from sustainable sources such as biomass is of current interest, due to its net carbon neutrality.4 There is growing interest in the application of thermochemical processes such as pyrolysis and gasification for the recovery of energy from biomass waste. Pyrolysis has attained much attention, as the process can be optimized to obtain liquid oils and char while gasification is more favorable for the production of gaseous products at higher temperatures. Various factors affecting the product yield including process operating parameters, the addition of a gasifying agent, a catalyst, and the reactor design have been investigated.5−9 Two stage combined pyrolysis-gasification has been found to be effective for thermochemical biomass conversion producing enhanced gas yield, low tar contents, and high biomass conversion rates.10−14 In the two stage process, biomass is pyrolyzed in the first stage, and then the derived pyrolysis gases, volatiles, and tar are gasified/reformed in the second stage already at higher temperatures (T > 800 °C). The addition of different catalysts together with the steam in the second stage has also shown a positive effect on biomass to gas conversion and tar reduction. Examples of two-stage systems have been reported; for example, Mun et al.5 investigated the gasification of sewage © 2013 American Chemical Society

sludge in a two-stage gasifier; gasification of bagasse was investigated by De Filippis et al.;15 Wu et al.16 investigated the effect of two different catalysts on pyrolysis and gasification of biomass components, cellulose, xylan, and lignin. In this study, pyrolysis-steam reforming and pyrolysiscatalytic steam reforming of rice husk, sugar cane bagasse, and wheat straw was performed in a two-stage reactor system. Calcined dolomite and 10 wt % Ni-dolomite were used as the catalysts in the second stage to enhance the hydrogen production. The aim of this study was to investigate the comparative effectiveness of different thermochemical processes such as pyrolysis, steam reforming, and catalytic steam reforming in terms of hydrogen production at a very high temperature of 950 °C. It was also interesting to compare the hydrogen production from three different biomass samples for a given process using the two-stage pyrolysis/reforming system.

2. MATERIALS AND METHODS 2.1. Biomass Samples. Sugar cane (Saccharum officinarum) bagasse (BG), rice (Oryza sativa) husk (RH), and wheat (Triticum aestivum) straw (WS) samples were sourced from Pakistan. Being an agricultural based economy, Pakistan is one of the main producers of wheat, rice, and sugar cane. In Pakistan, it is estimated that 16.3 million tonnes of bagasse is available in the form of residues from the sugar industry; 1.4 million tonnes of rice husk and 35.8 million tonnes of wheat straw are available annually.17 The biomass samples were ground and sieved to a particle size of 1.4−2.8 mm and kept in Received: June 19, 2013 Revised: October 4, 2013 Published: October 7, 2013 6695

dx.doi.org/10.1021/ef401145w | Energy Fuels 2013, 27, 6695−6704

Energy & Fuels

Article

airtight containers to ensure consistent composition until final usage. 2.2. Catalysts. Naturally occurring dolomite and 10 wt % Ni-dolomite were used in this study. Dolomite was calcined in an air atmosphere at 1000 °C for 3 h. The 10 wt % Ni-dolomite was synthesized in the laboratory using a wet impregnation method.18 A known quantity of Ni(NO3)2·6H2O was dissolved in 25 mL of deionized water; 10 g of dolomite was then dissolved in the solution and heated up to 105 °C with constant stirring. The catalyst was dried overnight at 105 °C and then calcined at 900 °C for 3 h in an air atmosphere. All the catalysts were later ground and sieved to achieve a final particle size between 0.050 and 0.212 mm. 2.3. Materials Characterization. A Stanton Redcroft 1000 thermogravimetric analyzer (TGA) operating at atmospheric pressure was used to carry out the proximate analysis of the biomass samples, while a Thermoquest CE Flash EA 2000 series instrument was employed for C, H, N, and S analysis of the biomass samples. Proximate and ultimate analysis results are shown in Table 1. Thermogravimetric analysis of all three

Figure 1. TGA and DTG thermograms of rice husk, sugar cane bagasse, and wheat straw at a heating rate of 20 °C min−1.

Emmett and Teller)), pore volume and pore size distribution of fresh catalysts (shown in Table 2) were measured using a Nova-

Table 1. Proximate and Ultimate Analysis of Feedstock

Table 2. Surface Properties of the Freshly Prepared Catalysts

Proximate Analysis feed stock bagasse rice husk wheat straw

volatile matter (wt %)

fixed carbon (wt %)

moisture (wt %)

ash (wt %)

81.55 63.71 73

10.91 12.98 17.47

5.92 6.1 2.63

1.62 17.21 6.9

catalyst dolomite noncalcined dolomite calcined at 1000 °C 10% Ni-dolomite calcined at 900 °C

Ultimate Analysis

a

feed stock

C (wt %)

bagasse rice husk wheat straw

45.5 39.82 32.96

H (wt %) N (wt %) 5.63 5.4 4.33

0.8 1.27 1.18

S (wt %)a

O (wt %)b

nd nd 0.12

48.07 53.51 61.41

BET surface area, m2 g−1

BJH pore volume, cm3 g−1

Average pore size, nm

2.0 4.7

0.0083 0.0110

2.22 2.84

5.6

0.0308

2.21

2200e surface area and pore size analyzer from Quantachrome Instruments USA. Adsorption and desorption curves were obtained by increasing the relative pressure from 0 to 1 at 77 K using liquid nitrogen. A nonlocal density function theory (NLDFT) equilibrium model was used to calculate the pore volume and pore size distribution. The deposition of carbon on the used catalysts after reaction was investigated by temperature-programmed oxidation (TPO) using a Stanton Redcroft 1000 thermogravimetric analyzer. The used catalyst after reaction was heated in an air atmosphere, from ambient temperature to 800 °C at a 15 °C min−1 heating rate, with a dwell time of 10 min. A Field Emission Gun Scanning Electron Microscope (FEGSEM) LEO 1530 equipped with 80 mm XMax SDD detector was used to analyze the microscopic structure of fresh and reacted catalysts. Each specimen was coated with a 10−15 nm platinum layer. High resolution electron microscopy of 100 000× magnification was carried out under vacuum conditions at a 2.5−3 mm working distance with a supply voltage of 3 kV. X-ray diffraction analysis of the catalysts was carried out using a D8 Focus from Bruker Corporation to examine the crystal structure using Cu α1 radiation. The angle 2θ between the X-ray source and detector was varied from 10 to 80°. Eva software along with the ICDD PDF2 (International Centre for Diffraction Data Powder Diffraction Files) database was used for phase identification. A Phillips CM200 Field Emission Gun Transmission Electron Microscope (FEG-TEM) coupled with an energy dispersive Xray spectrometer (EDX) was also used to further investigate the morphology of the fresh and reacted 10 wt % Ni-dolomite catalysts.

nd - not detected. bCalculated by difference.

biomass samples (15 mg) was also investigated at a 20 °C min−1 heating rate to a final temperature of 900 °C with a dwell time of 10 min, using nitrogen as a carrier gas. With the rise in temperature, sample weight loss was continuously recorded. This raw data was used to calculate thermogravimetric (TG) and differential thermogravimetric (DTG) curves, shown in Figure 1. The main components of biomass are hemicelluloses, cellulose, and lignin. Williams and Besler19 showed that hemicelluloses begin to thermally decompose at 250 °C, and the main weight loss occurs between 250 and 350 °C, after which there is a slower weight loss as the temperature is raised further. Cellulose, after a small loss of volatile matter, begins to decompose at about 325 °C, and the main weight loss occurs between 325 and 400 °C. Lignin produces a gradual loss of weight over a temperature range of 200−700 °C. The temperature ranges depend on the heating rate, with a shift to higher temperature ranges at higher heating rates. Figure 1 shows some distinct peaks which reflect the presence of the three main components of biomass. Bagasse is mainly composed of hemicellulose and cellulose, as also reported by Varhegyi and Antal.20 Rice husks are mainly composed of hemicellulose and cellulose with some lignin.21 Wheat straw is mainly composed of cellulose with lower fractions of hemicelluloses and lignin.22 The freshly prepared and used catalysts were characterized using a range of techniques. The surface area (BET (Brunauer, 6696


Energy & Fuels

Article

2.4. Experimental Reactor. Pyrolysis of biomass samples (4 g) and steam reforming of the evolved gas components was carried out in a two stage fixed bed reactor system, 60 cm in length with a 2.5 cm inner diameter and constructed of Inconel (Figure 2). The reactor system consisted of two stages, a

collection time was 90 min. For pyrolysis/reforming experiments, the reforming stage was heated first from ambient temperature to 950 °C in ∼50 min. After the catalyst bed had reached 950 °C, the pyrolysis stage was heated to 950 °C in ∼50 min. The gases were collected for another 40 min to ensure good material balance. The total gas collection time for reforming experiments was 140 min. Repeatability experiments with and without steam were performed on the reactor system to ensure the reliability and suitability of the system for the research work. 2.5. Gas Analysis. Two different gas chromatographs (GCs) were used to analyze the gas samples collected in the gas sample bag. The gases were analyzed for hydrocarbons (C1− C4) using a Varian CP-3380 gas chromatograph with a column packed with an 80−100 mesh Hysep with a flame ionization detector (GC/FID) and using nitrogen as a carrier gas. Permanent gases (H2, CO, N2, O2, CO2) were analyzed using a second Varian CP-3380 chromatograph comprised of two columns with two thermal conductivity detectors (GC/TCD). One column packed with a 60−80 mesh molecular sieve was used to separate hydrogen, carbon monoxide, nitrogen, and oxygen, and the other column packed with 80−100 mesh Hysep was used to analyze carbon dioxide. The carrier gas used was argon. In order to calculate the amount of gas in grams, the area obtained for each gas from gas chromatography was compared with the area of calibration gas, and the percentage of each gas in the mixture was calculated. As a known quantity of nitrogen was introduced into the reactor during the experiment, the total gas volume could be calculated. Using the volume percentage of each gas and the total gas volume, the number of moles for each gas was calculated. The number of moles of each gas was converted into grams using their molecular weights, and the total mass of the gas mixture was obtained from the sum of all individual gases.

Figure 2. Schematic diagram of the two-stage fixed-bed high temperature reactor.

pyrolysis stage and a reforming stage, each separately heated by electrical furnaces. Nitrogen was used as purge gas. The flow rate of nitrogen was 100 mL min−1 for all experiments. The first pyrolysis reactor stage produces pyrolysis gases which are then passed directly to the second stage containing the catalyst, and the pyrolysis volatiles and gases are reformed in the presence of steam and the catalyst. For steam reforming and catalytic steam reforming experiments, the second stage catalytic reactor was heated first to 950 °C. The catalyst (2 g) was placed on perforated mesh placed in the reforming reactor. Once the second stage reactor reached the required temperature of 950 °C, heating up of the pyrolysis stage began along with the steam injection in the second stage. The pyrolysis stage was heated at 20 °C min−1 to 950 °C. Steam was introduced into the reforming reactor via water injected from a syringe pump which was immediately converted into steam at a high temperature and swept through the reactor by the nitrogen carrier gas. The steam to biomass ratio was 1.37 for reforming experiments as steam was introduced at 0.1 g min−1 for 55 min (5.5 g) for 4 g of biomass. A sand bed was used for steam reforming experiments. Product condensable gases and excess steam were condensed in a series of condensers. The total evolved noncondensable gases were collected using a Tedlar gas sample bag and were analyzed offline using gas chromatography. For pyrolysis experiments in the absence of a catalyst, the biomass sample was placed in the pyrolysis reactor stage, and both stages were heated simultaneously at 20 °C min−1 for ∼50 min to achieve the final temperature of 950 °C. The gases were collected for another 40 min to ensure good material balance. For pyrolysis experiments, the total gas

3. RESULTS AND DISCUSSION 3.1. Characterization of Fresh Catalysts. The surface properties of the dolomite and 10 wt % Ni-dolomite are shown in Table 2. It is evident from the results that calcination of naturally occurring dolomite resulted in an increase in BET surface area from 2.0 m2 g−1 to 4.7 m2 g−1. Sasaki et al.23 reported a strong influence of calcination temperature on pore size distribution of naturally occurring dolomite. They described calcination as a two-step process: the formation of MgO between 600 and 700 °C in the first step and formation of CaO from 700 to 900 °C in the second step. A similar mechanism was also proposed by Yoosuk et al.24 The addition of Ni into dolomite slightly increased the BET surface area to 5.6 m2 g−1. Scanning electron microscope (SEM) characterization of the fresh catalysts is shown in Figure 3. Calcination of dolomite (Figure 3b) resulted in a granular morphology due to the thermal degradation of the larger grains (Figure 3a). SEM of the 10 wt % Ni-dolomite shown in Figure 3c indicates the presence of a granular hexagonal plate-like morphology. From the X-ray diffraction results of the fresh catalysts shown in Figure 4a, it can be inferred that the raw dolomite primarily consisted of MgCa(CO3)2; however the presence of CaCO3 in dolomite has also been reported.23 Calcination of dolomite resulted in a complete breakdown of MgCa(CO3)2 into CaO and MgO, as shown in Figure 4b.25 The addition of Ni into the dolomite resulted in the formation of NiO and NiMgO2 along with CaO and MgO. Srinakruang et al.26 have suggested that 6697


Energy & Fuels

Article

Figure 3. SEM of fresh catalysts (a) fresh dolomite noncalcined, (b) dolomite calcined at 1000 °C, (c) 10 wt % Ni-dolomite calcined at 900 °C.

Table 3. Product and Hydrogen Yield from the Pyrolysis of Different Biomass Samples at 20 °C min−1 to a Final Temperature of 950 °C gas/biomass (wt %) solid/biomass (wt %) oil/biomass (wt %) mass balance (wt %) H2 (mmol g−1 biomass) H2/CO CO/CO2 H2/CO2 H2/CH4 CH4/CO CH4/CO2

rice husk

sugar cane bagasse

wheat straw

22.29 30.00 42.25 94.54 2.12 0.94 0.74 0.70 2.46 0.38 0.28

22.53 20.25 54.25 97.03 2.07 0.99 0.65 0.64 2.32 0.43 0.27

24.19 24.75 50.00 96.44 2.22 0.96 0.67 0.64 2.68 0.36 0.24

of 2.12 mmol g−1, 2.07 mmol g−1, and 2.22 mmol g−1 were obtained from the three biomass samples. A relatively larger oil fraction was recovered from bagasse and wheat straw compared to rice husk. The rice husk sample exhibited the lowest gas and oil yield resulting in the lowest feedstock to volatile conversion. This lowest conversion and highest solid yield from rice husk was primarily due to the higher ash contents in rice husk (Table 1).27 It is also interesting to note that the relatively higher oil yield was obtained from sugar cane bagasse and wheat straw. This was in agreement with the TGA and DTG results shown in Figure 1, indicating the presence of a larger quantity of volatile matter in these biomass samples. Burhenne et al.28 performed the pyrolysis of wheat straw in a fixed bed pyrolysis reactor at a much lower pyrolysis temperature of 500 °C; they

Figure 4. XRD results of fresh catalysts (a) fresh dolomite noncalcined, (b) fresh dolomite calcined at 1000 °C, (c) fresh 10 wt % Ni-dolomite.

the NiO phase was only present at a calcination temperature of 500 °C. According to their results, a more stable form, NiMgO2, was observed at high temperature, due to the strong interaction between Ni species and dolomite. 3.2. Pyrolysis of Biomass. Table 3 shows the results for the pyrolysis of the three biomass samples heated at 20 °C min−1 to a final temperature of 950 °C. A gas yield of 22.29, 22.53, and 24.19 wt % was attained from rice husk, bagasse, and wheat straw, respectively. As shown in Table 3, hydrogen yields 6698


Energy & Fuels

Article

Neves et al.33 explained the pyrolysis of biomass as a three step process. In the first step, as-received biomass releases moisture and is converted into dry biomass. In the second step, this dry biomass undergoes primary pyrolysis, in which feedstock is converted into char with the evolution of volatiles, permanent gases, moisture, and tar. During the last step, the complex interaction between these species results in reforming, cracking, oxidation, polymerization, and other gasification reactions. It has been reported34 that during the pyrolysis of cellulose and hemicellulose components present in biomass, CO and CO2 were the major gas components from the decarboxylation of these biomass components at lower temperatures (900 °C) two-stage system over other configurations.40 3.4. Dolomite Catalytic Steam Reforming of Biomass. The product and hydrogen yields derived from the pyrolysis/ reforming of the three biomass samples using dolomite are shown in Table 5. From these results, it is shown that in

The gas composition derived from the pyrolysis and steam reforming of rice husk (RH), sugar cane bagasse (BG), and wheat straw (WS) are shown in Figure 5. In contrast to the pyrolysis results, the presence of steam in the second stage reactor resulted in a substantial change in gas composition. Carbon monoxide and CH4 concentrations were somewhat reduced, and only trace amounts of C2−C4 hydrocarbons were found in the gas mixture during steam reforming. However, the hydrogen concentration in the product gas mixture was enhanced from around 25 vol % during pyrolysis to 55 vol % during two stage steam reforming. Hydrogen concentrations of 55.62 vol % for rice husk, 54.12 vol % for bagasse, and 55.94 vol % for wheat straw were achieved. Chang et al.36 investigated the steam gasification of bagasse in a fluidized bed reactor at a lower temperature of 700 °C, where less than 20 vol % hydrogen was observed at an Equivalence Ratio (ER) of 0.29 and a steam/biomass ratio of 0.5. Similar findings were reported by Skoulou et al.,37 who investigated the gasification of olive kernels in a fluidized bed reactor at 750 °C. The highest hydrogen concentration of 23.98 vol % was obtained for an ER of 0.2. It is evident from these findings that the higher temperature used in this study effectively enhanced the hydrogen gas yield. The steam reforming results (Table 4), when compared with the pyrolysis results (Table 3), showed that the CO/CO2 ratio was increased significantly for the pyrolysis/steam reforming of all three biomasses. For example, the CO/CO2 ratio was increased from 0.67 to 1.46 for wheat straw. Also, the concentration of CO2 reduced to less than half from 38 vol % during pyrolysis to 17 vol % during steam reforming. This decrease in CO2 concentration at a higher temperature indicates the increase in the forward Boudouard reaction (eq 2). As reported by Yang et al.,38 this reaction is favored at higher temperatures. The endothermic reaction between CO2 and CH4 (eq 4) was also supported at high temperatures contributing toward the decrease in CO2 and CH4 concentrations. The high H2 production during steam gasification/reforming can be attributed to the following chemical reactions. C + H 2O → CO + H 2 ΔH = 131.3 kJ/mol

(1)

C + CO2 → 2CO ΔH = 172 kJ/mol

(2)

CO + H 2O → H 2 + CO2 ΔH = − 41.2 kJ/mol

(3)

CH4 + CO2 → 2CO + 2H 2 ΔH = 247 kJ/mol

(4)

CH4 + H 2O → CO + 3H 2 ΔH = 206.4 kJ/mol

(5)

CmHn + nH 2O → nCO + [n + (m /2)]H 2

(6)

Table 5. Product and Hydrogen Yield from the Dolomite Catalytic Steam Reforming of Different Biomass Samples at 950 °C

gas/(biomass + water) (wt %) solid/(biomass + water) (wt %) mass balance (wt %) gas/biomass (wt %) solid/biomass (wt %) H2 (mmol g−1 biomass) H2/CO CO/CO2 H2/CO2 H2/CH4 CH4/CO CH4/CO2

rice husk

sugar cane bagasse

wheat straw

24.63 12.45 94.69 62.31 31.50 22.55 2.87 1.02 2.93 18.98 0.15 0.15

25.55 9.87 92.78 60.82 23.50 22.30 2.49 1.34 3.34 18.77 0.13 0.18

25.63 11.16 95.71 61.46 26.75 21.97 2.51 1.29 3.25 12.94 0.19 0.25

relation to biomass (corrected for no input water), more than 60 wt % of the biomass sample was converted into gas; however the gas yield in relation to biomass + water was found to be 24.63 wt % for rice husk, 25.55 wt % for bagasse, and 25.63 wt % for wheat straw, respectively. The solid yield was similar for these experiments as was found for pyrolysis since conditions in the pyrolysis stage were identical. Hydrogen yield did not increase significantly, being 22.55 mmol g−1 for rice husk, 22.30 mmol g−1 for bagasse, and 21.97 mmol g−1 for wheat straw. It has been reported that the use of calcined dolomite (CaOMgO) minimized tar production in the product gas mixture.41 The better activity of calcined dolomite as compared to noncalcined dolomite was due to the higher surface area and higher CaO and MgO contents.42 Simell et al.43 suggested that the CaO was more reactive than dolomite. González et al.11 investigated the two-stage gasification of olive cake using dolomite. They reported an improvement in hydrocarbon and tar cracking reactions. Wang et al.44 used a two-stage gasification and catalytic system to enhance the hydrogen yield from pig compost. The gasification and catalytic stages were kept at constant temperatures of 800 and 900 °C, respectively. It was reported that the presence of calcined modified dolomite enhanced the hydrogen yield, but this effect was more evident at a lower temperature of 800 °C. Their results showed that hydrogen yield was increased from 10.62 mmol g−1 of a sample to 18.76 mmol g−1 of a sample. Although the presence of dolomite did not improve hydrogen yield significantly, it was reported that the addition of Ni to dolomite was a promising option for enhanced hydrogen yield. Gas composition results from the catalytic two stage pyrolysis reforming of the biomass samples using calcined

The injected steam also increased the hydrogen concentration by reacting with CH4, C, and CO. The carbon steam gasification reaction (eq 1) and steam methane reforming reaction (eq 5) were favored at high temperatures due to their endothermic nature since the CH4/CO ratio was decreased from 0.38 (for pyrolysis) to 0.23 (for steam reforming) for rice husk. The water gas shift reaction (eq 3), on the other hand, is slightly exothermic, but the equilibrium can be shifted toward the products at higher steam to biomass ratios. The presence of steam enhanced the H2/CO ratio from 0.94 (for pyrolysis) to 2.58 (for steam reforming) for rice husk. Herguido et al.39 reported an increase in hydrogen concentration (up to 60 vol %) with a corresponding decrease in CO concentration when 6700


Energy & Fuels

Article

dolomite are presented in Figure 6. Compared to the steam reforming results shown in Figure 5, a slight increase in

Table 6. Product and Hydrogen Yield from the 10 wt % NiDolomite Catalytic Steam Reforming of Different Biomass Samples at 950 °C

gas/(biomass + water) (wt %) solid/(biomass + water) (wt %) mass balance (wt %) gas/biomass (wt %) solid/biomass (wt %) H2 (mmol g−1 biomass) H2/CO CO/CO2 H2/CO2 H2/CH4 CH4/CO CH4/CO2

Figure 6. Gas composition from the dolomite catalytic steam reforming and 10 wt % Ni-dolomite catalytic steam reforming of rice husk (RH), sugar cane bagasse (BG), and wheat straw (WS).

rice husk

sugar cane bagasse

wheat straw

26.30 13.02 96.65 63.64 31.50 25.44 2.59 1.41 3.64 33.24 0.08 0.11

28.04 8.98 96.80 70.24 22.50 25.41 2.39 1.37 3.28 32.69 0.07 0.10

26.20 10.80 97.23 64.90 26.75 24.47 2.63 1.26 3.31 28.48 0.09 0.12

different compositions of Ni-dolomite were calcined at 900 °C in an argon atmosphere. With the introduction of Ni-dolomite, a significant increase in gas yield along with a decrease in tar and char formation was reported. Various researchers have reported the effectiveness of Ni for tar removal and hydrogen production from gasification.50−53 Because of the two stage configuration and high temperature employed in this study, very little to no carbon was observed on the catalyst after reaction. Hydrogen production results in this study are compared with some literature data in Table 7.36,37,54−64 It is evident that the combination of high temperature steam gasification/reforming and the tar cracking

hydrogen concentration was observed, with a corresponding decrease in CO concentration and a slight increase in CO2 concentration and decrease in CH4 concentration. However, no C2−C4 hydrocarbons were detected during the catalytic steam reforming of all three biomass samples. This suggests that the higher temperature steam reaction was more effective in reforming and cracking of hydrocarbons and tar components, thereby enhancing gas and hydrogen yield. André et al.45 reported an increase in hydrogen concentration due to the presence of dolomite. Similar results have been reported by González et al.,11 suggesting that the presence of dolomite improved the water gas shift reaction as was evidenced by a higher hydrogen concentration with the reduction in CO concentration in the gas mixture. Other authors46−48 also reported the effectiveness of dolomite as a gasification/ reforming catalyst. Dolomite was found to be very effective for the reduction of tar compounds. As reported by Olivares et al.,47 the tar cracking capability of dolomite was mainly due to the steam reforming and dry reforming reactions. 3.5. 10 wt % Ni-Dolomite Catalytic Steam Reforming of Biomass. Product yield and hydrogen production from the two stage pyrolysis reforming of the biomass samples using 10 wt % Ni-dolomite are presented in Table 6. Introducing 10 wt % Ni into the calcined catalyst increased the gas yield as well as hydrogen yield. For example, compared to the dolomite results, hydrogen yield was enhanced from 22.55 mmol g−1 to 25.44 mmol g−1 for rice husk, from 22.30 mmol g−1 to 25.41 mmol g−1 for sugar cane bagasse, and from 21.79 mmol g−1 to 24.47 mmol g−1 for wheat straw. Enhanced gas yield and hydrogen yield was also reported by Wang et al.44 They used a two-stage gasification system to enhance the hydrogen yield from pig compost. Introduction of Ni into the dolomite catalyst increased the gas yield from 0.97 N m3 kg−1 to 1.33 N m3 kg−1, and hydrogen production was also enhanced from 18.76 mmol g−1 to 32.45 mmol g−1 of the sample. Ni-dolomite was also reported to be very effective in tar reduction, only 0.24 g Nm−3 of tar was found with the use of Ni-dolomite catalyst. Corujo et al.49 investigated the influence of calcined dolomite and Ni-dolomite on the product yield from the steam gasification of forestry residue at 900 °C. Dolomite and

Table 7. Comparison of Hydrogen Results with the Literature sample

atmosphere

temperature (°C)

hydrogen (vol %)

ref

rice husk

FixB

steam

950

59.13

bagasse

FixB

steam

950

57.02

wheat straw

FixB

steam

950

58.23

empty fruit bunch wood pellets bagasse agrol

FB

air

850

26.7

this study this study this study 54

FB FB FB

766 700 800−820

37.5 18 24

55 36 56

willow

FB

800−820

28

56

dry distillers grain saw dust mixture wood pellets coir pith rice husk oliver kernels oliver kernels animal waste pine waste wood chips

FB

800−820

20

56

FB

air + steam air + steam steam + oxygen steam + oxygen steam + oxygen air

739

16.36

57

FixB FB FB FB FixB FB FB FixB

air air steam air steam air + steam steam air

600 773 750 750 1050 660 800 700

15.73 27.76 47.81 23.98 42 48 30 15

58 59 60 37 61 62 63 64

a

6701

reactora

FB = fluidized bed; FixB = fixed bed. dx.doi.org/10.1021/ef401145w | Energy Fuels 2013, 27, 6695−6704

Energy & Fuels

Article

capabilities of 10 wt % Ni-dolomite is a promising option for recovering higher hydrogen yield from biomass. The gas composition derived from the pyrolysis reforming of the biomass samples using 10 wt % Ni-dolomite is compared with the dolomite reforming results in Figure 6. For rice husk, the CO concentration slightly increased along with an increase in hydrogen concentration. CO2 and CH4 concentrations were reduced. Similar trends were observed for bagasse and wheat straw. The increase in hydrogen concentration can be partly attributed to the further cracking of tar components in the gas stream by the 10 wt % Ni-dolomite catalyst. Other authors49 have also suggested that the increase in hydrogen concentration was due to the increase in overall gas yield due to thermal cracking of tar. Ther increase in CO concentration along with the decrease in CO2 concentration might be due to the reverse water gas shift reaction. At high temperatures, the equilibrium of the water gas shift reaction changes toward the reactants. The reduction in CH4 concentration can be explained due to the enhanced steam reforming reaction. Wang et al.44 also reported a positive effect of the addition of Ni onto modified dolomite, reporting that hydrogen concentration was enhanced from 36.60 vol % for no catalyst to 43.32 vol % for a modified calcined dolomite and finally to 54.49 vol % for 10 wt % Ni impregnated onto calcined modified dolomite. 3.6. Characterization of Reacted Catalysts. Different characterization techniques including temperature-programmed oxidation (TPO) and transmission electron microscopy (TEM) were employed to characterize the reacted calcined dolomite and 10 wt % Ni-dolomite for carbon deposition and other morphological changes. The amount of carbon deposited on the catalyst was calculated using eq 7. w=

(w1 − w2) × 100 (wt%) w1

Figure 7. TGA−TPO and DTG−TPO results of reacted dolomite (Dol) and reacted 10 wt % Ni-dolomite (Ni-Dol) catalysts during the catalytic steam reforming of rice husk (RH), bagasse (BG), and wheat straw (WS) at 950 °C.

(7)

where w is the amount of deposited carbon on the catalyst in wt %, w1 is the initial catalyst weight after moisture loss, and w2 is the final catalyst weight after oxidation. For rice husk, the amount of carbon deposited on the catalyst decreased from 5.66 wt % to 1.33 wt % when the catalyst was changed from dolomite to 10 wt % Ni-dolomite. For bagasse, 10.13 wt % carbon deposits were found on dolomite compared to 5.55 wt % carbon found on 10 wt % Ni-dolomite. Similarly for wheat straw, 9.74 and 5.84 wt % carbon deposits were found on dolomite and 10 wt % Ni-dolomite, respectively. It is evident from these results that the addition of Ni to the dolomite enhanced the catalytic activity at high temperatures with lower carbon deposition. TGA-TPO and DTG-TPO thermograms for reacted dolomite and 10 wt % Ni-dolomite are shown in Figure 7. Two distinct peaks were observed from the TGA−TPO profiles of both catalysts. The first peak observed at a temperature of ∼425 °C can be assigned to amorphous carbon,53 while the second peak found at ∼650 °C was most likely due to the presence of graphitic carbon.65 The presence of graphitic carbon on the reacted 10 wt % Nidolomite catalysts were evident from the TEM image shown in Figure 8. Similar graphitic carbons deposited on nickel catalysts were reported by Sehested66 and Wang et al.67 Various researchers68−70 have investigated the formation and morphology of the carbon deposited on nickel based catalysts. It has been reported71 that the deposition of carbon on the catalyst surface initiates with the dissociation of hydrocarbons derived from the pyrolysis of biomass, leading to the formation

Figure 8. TEM image of reacted 10 wt % Ni-dolomite (Ni-Dol) catalyst.

of highly reactive monatomic carbon. This highly reactive carbon if not converted into CO can react with the Ni phase (produced from the in situ reduction of the NiO phase) to form carbides, which results in the formation of carbon whiskers by further dissolving and diffusing of the reactive layered carbon into the Ni particles.72 This process of deposition of layered carbon at the rear of the Ni particle results in the formation of filamentous carbon.70 Further investigations on the formation of filamentous carbons was performed by Wang et al.67 It was suggested that filamentous carbon consisted of graphite sheets piled up in the shape of hollow cones. Similar findings were reported by Natesakhawat et al.68 and Kep̨ iński et al.70 showing the presence of Ni particles on the tip of the filamentous carbon.

4. CONCLUSIONS Experiments of high temperature pyrolysis, steam reforming, and catalytic steam reforming of rice husk, bagasse, and wheat straw were performed in a two stage, fixed-bed reactor system to obtain a syngas with a high hydrogen content. The following may be concluded: 6702


Energy & Fuels

Article

Hydrogen yield increased dramatically, for example, from ∼2 mmol g−1 biomass for pyrolysis to ∼21 mmol g−1 of biomass during the two-stage pyrolysis/reforming of biomass in the presence of steam. This suggests that the high temperature (950 °C) employed in this study is a promising process for hydrogen production as compared to the conventional pyrolysis and gasification/reforming performed at lower temperatures. The use of calcined dolomite and 10 wt % Ni-dolomite catalysts in the second stage further increased syngas yield and hydrogen yield. The highest hydrogen yield of 25.44 mmol g−1 of biomass was obtained from the pyrolysis/reforming of rice husk using 10 wt % Ni-dolomite. The highest hydrogen concentration in the gas mixture was found to be 59.14 vol %. The higher gas yield obtained in this study was primarily due to the higher temperature of the second stage steam reformer, leading to thermal cracking of tar components and effective reforming of methane and other hydrocarbons produced during the pyrolysis stage of the process, thereby enhancing the hydrogen yield. Significantly lower carbon deposits were found on the reacted 10 wt % Ni dolomite, suggesting that the high temperature of the process has the potential to maintain the initial higher catalytic activity by suppressing carbon formation and deposition on the catalyst.

■

(15) De Filippis, P.; Borgianni, C.; Polucci, M.; Pochetti, F. Biomass Bioenergy 2004, 27, 247−252. (16) Wu, C.; Wang, Z.; Dupont, V.; Huang, J.; Williams, P. T. J. Anal. Appl. Pyrol. 2013, 99, 143−148. (17) Bhutto, A. W.; Bazmi, A. A.; Zahedi, G. Renewable Sustainable Energy Rev. 2011, 15, 3207−3219. (18) Wu, C.; Williams, P. T. Int. J. Hydrogen Energy 2009, 34, 6242− 6252. (19) Williams, P. T.; Besler, S. Renew. Energy 1996, 7, 233−250. (20) Varhegyi, G.; Antal, J R. Energy Fuels 1989, 3, 329−335. (21) Williams, P. T.; Besler, S. Fuel 1993, 72, 151−159. (22) Thomsen, M. H. H.; Henrik, N. J. Ind. Microbiol. Biotechnol. 2008, 35, 9−16. (23) Sasaki, K.; Qiu, X.; Hosomomi, Y.; Moriyama, S.; Hirajima, T. Microporous Mesoporous Mater. 2013, 171, 1−8. (24) Yoosuk, B.; Udomsap, P.; Puttasawat, B. Appl. Catal., A 2011, 395, 87−94. (25) Kristóf-Makó, É.; Juhász, A. Z. Thermochim. Acta 1999, 342, 105−114. (26) Srinakruang, J.; Sato, K.; Vitidsant, T.; Fujimoto, K. Fuel 2006, 85, 2419−2426. (27) Abu Bakar, M. S.; Titiloye, J. O. J. Anal. Appl. Pyrolysis 2013, 101. (28) Burhenne, L.; Messmer, J.; Aicher, T.; Laborie, M. P. J. Anal. Appl. Pyrolysis 2013, 101, 177−184. (29) Kırtay, E. Energy Convers. Manage. 2011, 52, 1778−1789. (30) Onay, O.; Koçkar, O. M. Biomass Bioenergy 2004, 26, 289−299. (31) Onay, O.; Kockar, O. M. Renewable Energy 2003, 28, 2417− 2433. (32) Garcìa-Pèrez, M.; Chaala, A.; Roy, C. Fuel 2002, 81, 893−907. (33) Neves, D.; Thunman, H.; Matos, A.; Taarelho, L.; Gomez-Barea, A. Prog. Energy Combust. Sci. 2011, 37, 611−630. (34) Chen, Y.; Yang, H.; Wang, X.; Zhang, S.; Chen, H. Bioresour. Technol. 2012, 107, 411−418. (35) Becidan, M.; Skreiberg, O.; Hustad, J. E. J. Anal. Appl. Pyrolysis 2007, 78, 207−213. (36) Chang, A. C. C.; Chang, H. F.; Lin, F. J.; Lin, K. H.; Chen, C. H. Int. J. Hydrogen Energy 2011, 36, 14252−14260. (37) Skoulou, V.; Koufodimos, G.; Samaras, Z.; Zabaniotou, A. Int. J. Hydrogen Energy 2008, 33, 6515−6524. (38) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Liang, D. T.; Zheng, C. Fuel Process. Technol. 2006, 87, 935−942. (39) Herguido, J.; Corella, J.; Gonzalez-Saiz, J. Ind. Eng. Chem. Res. 1992, 31, 1274−1282. (40) Devi, L.; Ptasinski, K. J.; Janssen, F. J. J. G. Biomass Bioenergy 2003, 24, 125−140. (41) Xu, C.; Donald, J.; Byambajav, E.; Ohtsuka, Y. Fuel 2010, 89, 1784−1795. (42) Hu, G.; Xu, S.; Li, S.; Xiao, C.; Liu, S. Fuel Process. Technol. 2006, 87, 375−382. (43) Simell, P.; Kurkela, E.; Stfthlberg, P.; Hepola, J. Catal. Today 1996, 27, 55−62. (44) Wang, J.; Xiao, B.; Liu, S.; Hu, Z.; He, P.; Guo, D.; Hu, M.; Qi, F.; Luo, S. Bioresour. Technol. 2013, 133, 127−133. (45) André, R. N.; Pinto, F.; Franco, C.; Dias, M.; Gulyurtlu, I.; Matos, M. A. A.; Cabrita, I. Fuel 2005, 84, 1635−1644. (46) Corella, J.; Herguido, J.; Gonzalez-Saiz, J.; Alday, F. J. In Research in Thermochemical Biomass Conversion; Bridgwater, A. V., Kuester, J. L.,Eds.; Springer: The Netherlands, 1988; pp 754−765. (47) Olivares, A.; Aznar, M. P.; Caballero, M. A.; Gil, J.; Frances, E.; Corella, J. Ind. Eng. Chem. Res. 1997, 36, 5220−5226. (48) Rapagnà, S.; Foscolo, P. U. Int. J. Hydrogen Energy 1998, 23, 551−557. (49) Corujo, A.; Yerman, L.; Arizaga, B.; Brusoni, M.; Castiglioni, J. Biomass Bioenergy 2010, 34, 1695−1702. (50) Blanco, P. H.; Wu, C.; Onwudili, J. A.; Williams, P. T. Appl. Catal., B 2013, 134, 238−250. (51) Wu, C.; Wang, Z.; Huang, J.; Williams, P. T. Fuel 2013, 106, 697−706.

AUTHOR INFORMATION

Corresponding Author

*Tel.: #44 1133432504. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The financial support provided by the Government of Pakistan for one of us (Q.M.K.W.) is gratefully acknowledged. REFERENCES

(1) Housecroft, C. E.; Sharpe, A. G. Inorganic Chemistry, 2nd ed.; Pearson Education Limited: Upper Saddle River, NJ, 2005; p 888. (2) Tanksale, A.; Beltramini, J. N.; Lu, G.M. Renew. Sus. Energ. Rev. 2010, 14, 166−182. (3) U.S. Department of Energy. Todays Hydrogen Production Industry. http://www.fossil.energy.gov/programs/fu els/h ydrogen/ currenttechnology.html (accessed Mar 2, 2013). (4) Lee, D. H.; Hung, C. P. Int. J. Hydrogen Energy 2012, 37, 15753− 15765. (5) Mun, T.-Y.; Kim, J. W.; Kim, J. S. Int. J. Hydrogen Energy 2013, 38, 5226−5234. (6) Mun, T.-Y.; Kim, J. W.; Kim, J. S. Fuel 2012, 102, 326−332. (7) Van de steene, L.; Tagutchou, J. P.; Mermoud, F.; Martin, E.; Salvador, S. Fuel 2010, 89, 3320−3329. (8) Efika, C. E.; Wu, C.; Williams, P. T. J. Anal. Appl. Pyrol. 2012, 95, 87−94. (9) Zhao, X.; Wang, M.; Liu, H.; Zhao, C.; Ma, C.; Song, Z. J. Anal. Appl. Pyrol. 2013, 100, 39−55. (10) Fassinou, W. F.; Van den Steene, L.; Toure, S.; Volle, G.; Girard, P. Fuel Process. Technol. 2009, 90, 75−90. (11) González, J. F.; Roman, S.; Engo, G.; Encinar, J. M.; Martinez, G. Biomass Bioenergy 2011, 35, 4324−4330. (12) Šulc, J.; Stojdl, J.; Richter, M.; Popelka, J.; Svoboda, K.; Smetana, J.; Vacek, J.; Skoblja, S.; Buryan, P. Waste Manage. 2012, 32, 692−700. (13) Wu, C.; Wang, L. Z.; Williams, P. T.; Shi, J.; Huang, J. Appl. Catal., B 2011, 108, 6−13. (14) Xiao, X.; Meng, X.; Dung, D.; Takarada, T. Bioresour. Technol. 2011, 102, 1975−1981. 6703


Energy & Fuels

Article

(52) Wu, C.; Williams, P. T. Fuel 2010, 89, 1435−1441. (53) Wang, S. Ind. Eng. Chem. Res. 1999, 38, 2615−2625. (54) Mohammed, M. A. A.; Salmiaton, A.; Azlina, W. A. K. G.; Amran, M. S.; Razi, F. R. Energy Convers. Manage. 2011, 52, 1555− 1561. (55) Campoy, M.; Gomez-Barea, A.; Vidal, F. B.; Ollero, P. Fuel Process. Technol. 2009, 90, 677−685. (56) Meng, X.; De Jong, W.; Fu, N.; Verkooijen, A. H. M. Biomass Bioenergy 2011, 35, 2910−2924. (57) Li, X. T.; Grace, J. R.; Lim, C. J.; Watkinson, A. P.; Chen, H. P.; Kim, J. R. Biomass Bioenergy 2004, 26, 171−193. (58) Plis, P.; Wilk, R. K. Energy 2011, 36, 3838−3845. (59) Subramanian, P.; Sampathrajan, A.; Venkatachalam, P. Bioresour. Technol. 2011, 102, 1914−1920. (60) Karmakar, M. K.; Datta, A. B. Bioresour. Technol. 2011, 102, 1907−1913. (61) Skoulou, V.; Swiderski, A.; Yang, W.; Zabaniotou, A. Bioresour. Technol. 2009, 100, 2444−2451. (62) Xiao, X.; Dung, D.; Morishita, K.; Zhang, S.; Li, L.; Takarada, T. Fuel Process. Technol. 2010, 91, 895−902. (63) Franco, C.; Pinto, F.; Gulyurtlu, I.; Cabrita, I. Fuel 2003, 82, 835−842. (64) Son, Y. I.; Yoon, S. J.; Kim, Y. K.; Lee, J. G. Biomass Bioenergy 2011, 35, 4215−4220. (65) Goula, M. A.; Lemonidou, A. A.; Efstathiou, A. M. J. Catal. 1996, 161, 626−640. (66) Sehested, J. Catal. Today 2006, 111, 103−110. (67) Wang, P.; Tanabe, E.; Ito, K.; Jiab, J.; Morioka, H.; Shishido, T.; Takehira, K. Appl. Catal., A 2002, 231, 35−44. (68) Natesakhawat, S.; Watson, R. B.; Wang, X.; Ozkan, U. S. J. Catal. 2005, 234, 496−508. (69) Kroll, V. C. H.; Swaan, H. M.; Mirodatos, C. J. Catal. 1996, 161, 409−422. (70) Kępiński, L.; Stasińska, B.; Borowiecki, T. Carbon 2000, 38, 1845−1856. (71) Trimm, D. L. Catal. Today 1999, 49, 3−10. (72) Wu, C.; Williams, P. T. Appl. Catal., B 2010, 96, 198−207.

6704


Steam

Recommend Documents