Catalytic decomposition of biomass tar compound by calcined coal gangue: A kinetic study

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Catalytic decomposition of biomass tar compound by calcined coal gangue: A kinetic study Feiqiang Guo a, Yuping Dong b,c,*, Pengfei Fan b, Zhaochuan Lv b, Shuai Yang b, Lei Dong c a

School of Electric Power Engineering, China University of Mining and Technology, 221116 Xuzhou, PR China School of Mechanical Engineering, Shandong University, 250061 Jinan, PR China c Shandong Baichuan Tongchuang Energy Company Ltd., 250101 Jinan, PR China b

article info

abstract

Article history:

The kinetics of thermal decomposition of biomass tar was studied using phenol as a model

Received 24 February 2016

compound under isothermal conditions by a two stage micro fluidized bed reactor.

Received in revised form

Calcined coal gangue and g-Al2O3 were employed as catalyst, particularly gangue repre-

26 April 2016

senting one type of catalyst for tar decomposition that prepared by the thermal conversion

Accepted 15 May 2016

of inferior coal. The Friedman method and integral method were employed for the deter-

Available online 2 June 2016

mination of kinetic parameters for forming H2 and CO by phenol decomposition. The results demonstrated that the activation energy for generating H2 was lower compared with

Keywords:

CO, indicating that the reactions for forming H2 was easier in phenol decomposition.

Phenol

Higher activation energies for H2 and CO generating were obtained using gangue as cata-

Decomposition

lyst, implying that the catalytic performance gangue was slightly lower toward phenol

Two-stage micro fluidized bed

decomposition process due to the high content of SiO2 in gangue.

Reaction kinetics

© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Isothermal

Introduction With regard to global issues of sustainable energy and environmental protection, biomass thermal conversion is considered to be a promising technology in reaching targets for renewable energy and greenhouse gases emissions reduction [1]. However, a serious problem in the utilization of biomass gas from thermal conversion, such as pyrolysis or gasification, is the trouble with tar contained in the gas [2e4]. Tars are formed during the thermal decomposition of biomass in series of complex reactions. Tar can be defined as a complex mixture of condensable fraction of the organic products, which are largely aromatic hydrocarbons, including single

ring to 5-ring aromatic compounds along with other oxygencontaining hydrocarbons and complex polycyclic aromatic hydrocarbon (PAH) hydrocarbons [5,6]. Tars may condense on cooler surfaces downstream, leading to blockage and fouling problem in some equipments. In addition, operational problems may be caused as a result of the possible formation of aerosols and soot formation [7,8]. Thus, researching on tar elimination from the product gas is becoming increasingly attractive. Tar reduction by using a hot gas cleaning catalyst has been widely considered as a potential method to convert tar into clean gases at lower temperature [9,10]. Among all the catalysts, char from the pyrolysis of coal or biomass has become of interest because char production and tar reduction can be

* Corresponding author. School of Mechanical Engineering, Shandong University, 250061 Jinan, PR China. Tel.: þ86 516 83592000. E-mail address: [email protected] (Y. Dong). http://dx.doi.org/10.1016/j.ijhydene.2016.05.126 0360-3199/© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

simultaneously implemented inside the reactor by controlling the parameters and configuration [11,12]. Coal gangue is the unexpected product of coal mining with average production of 10% during the raw coal production in China [13]. Coal gangue has low carbon content and high oxides content (both metallic oxides and nonmetallic oxides) [14,15], implying that the ash of gangue has catalytic performance toward biomass tar. Thus, this material can be chosen as a catalyst for the tar decomposition. The catalytic decomposition characteristics of biomass tar have been studied using char and char-supported catalysts from biomass or coal. Phenol, benzene and naphthalene are usually chosen as model compounds. The conversion of naphthalene and phenol was carried out by El-Rub et al. [16] to compare the catalytic performance of biomass chars with other catalysts, founding that biomass chars gave the good phenol and naphthalene conversion. Zhang et al. [17] studied the in situ tar reforming of coal tar using brown coal char, showing that the major components of tar cracked over the hot char bed. Coal chars were employed as well by Wang et al. [18] to catalyze the tar reforming reactions during the pyrolysis and gasification of brown coal, the results showing that chars derived from brown coal showed excellent catalytic activity for tar reforming and the formation of large aromatics in tar was enhanced. The catalytic property of coal and biomass char depends upon the metallic oxides in the ash. Thus, there are reasons to believe that the coal gangue has the catalytic performance toward tar due to its high metallic oxides content. A significant fraction of biomass tars originates from pyrolysis and gasification products mainly composed of phenolic compounds [19,20]. Phenol and phenolic compounds are identified as typical representatives by Morf et al. [21] for the secondary tar component class. Phenol appeared in all tar classes as precursor or intermediate and belonged to the problematic producer gas impurities for the practical application of biomass gasification. The simplest phenolic model compound with a phenyl and a hydroxyl, which is phenol, has been chosen as a surrogate for primary tar, as a first step toward developing an advanced elementary kinetic model [22]. A thorough understanding of pyrolysis kinetics of phenol is a key component in the efficient design of its decomposition processes. However, there have been very few works on detailed kinetic mechanism for phenol decomposition against well-controlled thermal conditions with analysis of products. Thermogravimetric analysis (TGA) has been generally used to analyze kinetics of gasesolid reactions based on monitoring the mass variation of a spot sample under a specified heating program. For isothermal conditions, the fluidized bed reactor can be used to obtain kinetic parameters based on measuring the releasing of major gas components in the thermal degradation of the tested sample [23e25]. As suggested by Yu et al. [26,27], compared with TGA, the micro fluidized bed reactor has the advantage in the minimized diffusion inhibition, online feed of reactant sample, quick heating for isothermal conditions, and the test at arbitrary temperatures and in various gaseous atmospheres. Toward liquid materials like biomass tar in this work, the micro fluidized bed reactor was therefore employed to calculate the kinetics and analyze the

13381

reaction mechanism based on the measured time-series of product gas composition. Since a large amount of coal gangue is produced during coal mining activities and its potential toward biomass tar decomposition. The decomposition characteristics of biomass tar using gangue as catalyst has both theoretical research meaning and practical applicative values. The aim of this work is to deeply understand the phenol conversion process using calcined coal gangue as catalyst and g-Al2O3 were employed as control. The kinetics of phenol catalytic decomposition was investigated experimentally at intermediate temperatures (750e850  C) to predict the formation of light gases (H2 and CO) under isothermal conditions using a twostage fluidized bed reactor.

Experimental methods Sampling The ultimate and proximate analyses of coal gangue are shown in Table 1. In order to avoid the influence of volatiles and fixed carbon of gangue on the gas generation of phone decomposition, the gangue was calcined for about 30 min at 850  C under air atmosphere until it turned into creamy-white before the test. The chemical composition of the gangue after calcination was analyzed as well and listed in Table 1. It can be seen that the ash of gangue after calcination mainly consists of SiO2 which is inert toward tar decomposition. The most abundant of the metallic oxides is Al2O3, which accounts for 31.5% of the ash. Thus, g-Al2O3 was chosen as a catalyst in comparison to study the catalytic characteristics of gangue for phenol decomposition. The particle size of calcined gangue and g-Al2O3 was in the range of 100e150 mm.

Table 1 e Chemical composition of coal gangue. Sample Proximate analysis (wt.%, db) Moisture Volatile Fixed Carbon Ash Ultimate analysis (wt.%, daf) Carbon Hydrogen Oxygen (diff) Nitrogen Sulfur Elemental ash analysis (wt %) SiO2 Al2O3 Fe2O3 CaO K2O Na2O MgO P2O5 SO3

Gangue 1.27 10.21 33.74 54.78 74.38 4.73 18.92 1.58 0.39 57.2 31.5 3.13 0.64 0.46 0.85 0.93 0.03 0.01

Notes: db: dry basis; daf: dry ash free basis; diff: calculated by difference.

13382

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

Apparatus and operational conditions Fig. 1 shows a schematic of the experimental apparatus that was used in the experiments discussed in this paper. The main components are: a two-stage micro fluidized bed reactor of 20 mm in diameter, a liquid injector system, a temperature and pressure sensor and a mass spectrometer (AMETEK, American). The experimental apparatus is capable of rapidly heating the feedstock to the desired temperatures to make the phenol decompose at constant conditions. The two-stage micro fluidized bed was made of quartz and consists of two porous plates with total height of 150 mm. The zone below the lower plate is designed as a chamber to realize uniform distribution of the fluidized gas. The zone between the two porous plates (the first stage for the reaction) of 40 mm in height is the zone where decomposition reactions of phenol occur. The zone above the upper porous plate (the second stage for the reaction) of 60 mm in height is the zone where the remaining phenol escaping from the lower stage are caught to realize further catalytic decomposition. There are two branches at the middle of the reactor, one for phenol sample injection, the other one for temperature and pressure sensors installing. In order to feed micro liquid sample quickly and accurately, a special liquid injector system was designed, as shown in Fig. 1. A liquid syringe and two cylinders were employed in the injector system. After the phenol sample was loaded in the syringe, the syringe and cylinder 1 were conveyed ahead to ensure the syringe needle reached the reaction zone of the fluidized bed. At that time, a certain amount of phenol was injected into the reactor from the syringe pushed by cylinder 1. After that, the syringe and

cylinder 1 were pushed back immediately by cylinder 2, and one stroke of phenol feeding was finished. The total time spent during the liquid sample feeding was less than 1 s by this injector system. The temperature of the furnace, carrier gas flow rate and actions of sample injector system are all controlled by a computer. Meantime, the temperature inside the reactor, pressures at the reactor inlet and outlet and the date of the product gas from the mass spectrometer are logged into the computer. Argon gas was used as carrier gas with flow rate of 300 NmL/min. Quartz sand with diameter in 100e150 mm was used as the fluidization medium in the first stage. Before each test, 3 g of quartz sand was put into the lower layer and 2 g of catalyst (calcined gangue or g-Al2O3) was put into the upper layer. The reactor was heated by the furnace in fluidization state to the desired temperature (750e850  C). After that, 0.02 ml phenol was injected into the reactor and the product gas (CO and H2) was measured by the mass spectrometer continuously. To assure the reliability of the test results, each test was repeated for three times.

Isothermal kinetic method During the experiments, the variation of the concentrations of the product gases with time under different reaction temperatures was measured by the mass spectrometer, and then the degree of conversion of the product gases (CO and H2) is expressed as: Z mt ¼

t

12 

4i  qv dt

0

(1)

22:4

Exhaust Flow sensor

Needle vavle

Pressure sensor

Cylinder 1 Cylinder 2

Mass spectrometer Gas filter

Syringe Electric furnace Temperature and pressure sensor

Fixation apparatus Quartz tubular reactor Gas valve

Mass flowmeter

Compressed gas

Fig. 1 e Schematic diagram of the experimental setup.

Display interface

13383

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

Z mt0 ¼

Zx

4i  qv dt

0

(2)

22:4 Z

x¼

t0

12 

t

mt t ¼ Z 0te mt0

GðxÞ ¼ 0

dx ¼ kðTÞt f ðxÞ

lnðkðTÞÞ ¼  4i  qv dt (3) 4i  qv dt

t0

Where, mt and mt0 denote the weight of the gas component i (CO and H2) at time t and the reaction end (t0) respectively, g; 4i represents the concentration of gas specie i, vol. %; and qv denotes the flow rate of the product gas from the reactor, L/ min; 12 is the molar mass of carbon and 22.4 is the molar volume of gas at standard condition; x is the conversion degree of product gas. dx ¼ kðTÞf ðxÞ dt   E kðTÞ ¼ A exp  RT

(7)

E þ lnðAÞ RT

(8)

where G(x) is the integral reaction model. Based on Eq. (7), the points of G(x) versus t at different reaction temperatures can be fitted to a straight line. The slope and intercept of the line corresponds to E/R and ln(A). Toward the thermochemical reactions between gas and solid, nineteen reaction models have been widely employed for the kinetic study of solid-state reactions, as listed in Table 2. In this paper, the form of G(x) which gives a straight line with the highest correlation coefficient will be considered to be the function of the model that best represents the kinetic gas releasing from phenol decomposition.

(4) (5)

Where, f(x) is a function, the type of which depends on the reaction mechanism; k(T) is the reaction rate constant defined by the Arrhenius equation and it is a constant in isothermal process, which means it can be separated from f(x). When kinetics follow a single expression such as Eq. (4), various methods can be applied to find the kinetic parameters. Friedman method [28] is widely used without considering the mechanism function f(x), based on the following equation:   dx E ¼ þ ln A þ lnðf ðxÞÞ ln dt RT

(6)

Where, E is the activation energy, kJ/mol; A denotes to the preexponential factor, 1/s; T refers to the temperature, K; R represents the gas constant, 8.314 J/(mol K). Unlike the Friedman method, an integration function considering the mechanism function is shown as below:

Results and discussion Temperature profiles In order to ensure that the decomposition reaction of phenol sample occurs at the desired temperature, the temperature in the main reaction zone (Fig. 1) of the reactor was measured every 1 s to monitor the temperature variation during each test. Fig. 2 presents the temperature profiles during the phenol decomposition under 750e850  C using gangue and g-Al2O3 as catalyst. It is apparent that the temperature profile maintains in a steady state with the variation of measuring data within around ±1  C, implying that the decomposition process of phenol can be ensured to occur at the desired temperature.

The releasing characteristics of CO and H2 The tests of phenol decomposition were carried out at a given temperatures varying in 750e850  C using coal gangue and g-

Table 2 e Typical mechanism functions using in gasesolid reactions. Symbol G(1) G(2) G(3) G(4) G(5) G(6) G(7) G(8) G(9) G(10) G(11) G(12) G(13) G(14) G(15) G(16) G(17) G(18) G(19)

Model Diffusional

Nucleation and growth

Autocatalytic reaction Mampel Power law

Order based

Phase interfacial reaction

f(x) 1-D diffusion 2-D diffusion 3-D diffusion (Jander) 3-D diffusion (GeB) 3-D diffusion (AeJ) n ¼ 2/3 n ¼ 1/2 n ¼ 1/3 n ¼ 1/4 n ¼ 1/2 n ¼ 1/3 n ¼ 1/4 Third order n ¼ 3 Second order n ¼ 2 First order n ¼ 1 Mampel Power law Contraction sphere Contraction cylinder

1/2$x [ln(1x)]1 [1.5(1x)2/3]/[1(1x)1/3] 1.5[1(1x)1/3]1 1.5(1x)2/3[(1 þ x)1/31]1 1.5(1x) [ln(1x)]1/3 2(1x) [ln(1x)]1/2 3(1x) [ln(1x)]2/3 4(1x) [ln(1x)]1/3 x (1x) 2x1/2 3x2/3 4x3/4 (1x)3 (1x)2 1x 1 3(1x)2/3 2(1x)1/2

G(x) 2

x xþ(1x) ln(1x) [1(1x)1/3]2 12x/3(1x)2/3 [(1 þ x)1/31]2 [ln(1x)]2/3 [ln(1x)]1/2 [ln(1x)]1/3 [ln(1x)]1/4 ln[x (1x)] x1/2 x1/3 x1/4 [(1x)21]/2 (1x)11 ln(1x) x 1(1x)1/3 1(1x)1/2

13384

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

826

850

Temperature ( C)

825

o

824 950

1000

1050

1100

1150

800

Gangue as catalyst Al2O3 as catalyst

750 0

1000

2000

3000

4000

Time (s) Fig. 2 e The temperature profiles inside the reactor.

little earlier in each test, indicating that the releasing of H2 may be easier compared with CO during phenol decomposition. It is well known that the activation energy represents essentially the difficulty to start a reaction. Therefore, the values of activation energy for releasing H2 should be smaller in comparison with CO using gangue and Al2O3 as catalyst. It can also be seen in Fig. 3 that the releasing pattern of the gas components differed with the variation of catalyst and temperature. The intensity of CO was stronger than H2 using gangue as catalyst, indicating that using gangue as catalyst benefit the generation of CO. The intensity of H2 was much stronger using g-Al2O3 as catalyst, particularly at T ¼ 750  C, also implying that the generation of H2 may be enhanced as a result of the catalytic effect of g-Al2O3. However, the intensity of H2 and CO became close when the temperature increased to 850  C.

Kinetic parameters by the Friedman method Al2O3 as catalyst. Since CO and H2 are the mainly target gas components from phenol decomposition, the releasing characteristics of them were measured by the mass spectrometer and four typical experiments are plotted in Fig. 3. It was apparent that the time to start or end the gas release was different. There is a law in common that H2 was released a

Setting the time when phenol samples were injected as the beginning of reaction, the phenol decomposition reaction is analyzed in terms of the conversion, x, defined against the volume of the produced gas components according to Eqs. (1)e(3). Here, the largest yield of H2 and CO at the end of the

-8

2.5x10

-8

-7

H2

Gangue as catalyst

CO

o

T=750 C

2.0x10

2.0x10

H2

Gangue as catalyst

CO

o

1.5x10

-7

1.0x10

-7

5.0x10

-8

T=850 C

-8

Intensity

Intensity

1.5x10

-8

1.0x10

-9

5.0x10

0.0

0.0 130

140

150

130

160

140

150

160

Time (s)

Time (s)

-7

2.5x10

3.5x10

Al2O3 as catalyst

H2

o

3.0x10

CO

T=750 C

Al2O3 as catalyst

-7

2.0x10

H2 CO

o

T=850 C

-7

2.5x10

-7

Intensity

Intensity

1.5x10

1.0x10

2.0x10

-7

1.5x10

-7

1.0x10 5.0x10

-8

5.0x10 0.0 130

140

150

160

170

0.0 130

140

Time (s)

Fig. 3 e The releasing characteristics of CO and H2.

150

Time (s)

160

170

13385

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

reaction corresponds to the conversion of 100%. The calculated results of conversion of CO and H2 versus time at different reaction temperatures using gangue and g-Al2O3 as catalyst are shown in Fig. 4. Temperature is a key factor influencing the tar cracking progress [29], and it also affect the time for complete releasing of the gas component. The time for complete releasing of H2 and CO decreased with elevating temperature, indicating that higher temperature promotes the cracking reactions. The slope of the conversion curves represents the reaction rate at a certain condition, and thus the results in Fig. 4 showing that higher temperature benefit quick gas generation during phenol decomposition. The degree of increment in the reaction rate was higher for CO in comparison with H2. Besides, the curves varied using different catalyst, representing that the catalytic property of gangue and g-Al2O3 was different as well toward phenol decomposition. It also can be seen that the difference of the time for complete releasing of CO and H2 was not significant between gangue and g-Al2O3, and therefore the catalytic performance of them may be similar. Based on the data sets achieved of conversion of CO and H2 in Fig. 4, the correlation of dx/dt and x for individual component was obtained under 750e850  C, as illustrated in Fig. 5. It was apparent that all the curves presented a single peak,

corresponding to the maximum releasing rate of CO and H2 under different conditions. The reaction rate increased sharply at the beginning and reached the maximum values before x ¼ 0.4. As a stable temperature was ensured using the micro fluidized bed, the quick heating was achieved soon after the phenol was injected into the reactor to realize its decomposition under isothermal condition. In Fig. 5, it also can be seen that the calculated reaction rate increased obviously with the increasing temperature, referring to a result from increasing both the chemical kinetics and physical diffusion [26]. This is also consistent with the fact that higher temperature benefit the cracking process of biomass tar. The data of Fig. 5 were re-correlated according to the Friedman method, and the correlation of lnðdx=dtÞ versus 1000/T was obtained, resulting a family of activation energy for the conversion x. The reaction was influenced by the reactant heating during the first and the last period, leading to apparently variation of the activation energy. The intrinsic activation energy of the reaction can thus be determined by averaging the values at x ¼ 0.2e0.8. The calculated activation energies together with the square of correlation coefficient (R2) are listed in Table 3. An average activation energy of 79.1 kJ/mol was obtained for generating CO using coal gangue as catalyst. The

1.0

1.0

CO

H2 0.8 o

0.6

750 C o 775 C o 800 C o 825 C o 850 C

0.4

0.2

0.0

Conversion x(-)

Conversion x(-)

0.8

0.6

0.4

0.2

Gangue as catalyst

Gangue as catalyst

0.0

0

5

10

15

20

o

750 C o 775 C o 800 C o 825 C o 850 C

25

0

5

10

Time(s)

1.0

25

H2 0.8

o

750 C o 775 C o 800 C o 825 C o 850 C

0.6

0.4

Conversion x (-)

Conversion x (-)

20

1.0

CO

0.8

Al2O3 as catalyst

o

750 C o 775 C o 800 C o 825 C o 850 C

0.6

0.4

Al2O3 as catalyst

0.2

0.2

0.0

15

Time(s)

0.0

0

5

10

15

20

25

0

5

Time (s) Fig. 4 e Relative conversion of CO and H2 versus time.

10

15 Time (s)

20

25

13386

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

0.15

o

750 C o 775 C o 800 C o 825 C o 850 C

CO

0.10

H2 0.15

Gangue as catalyst

0.10

Gangue as catalyst

dx/dt

dx/dt

o

750 C o 775 C o 800 C o 825 C o 850 C

0.05

0.05

0.00

0.00 0.0

0.2

0.4

0.6

0.8

0.0

1.0

0.2

0.4

o

CO

750 C o 775 C o 800 C o 825 C o 850 C

0.10

dx/dt

Al2O3 as catalyst

H2

0.10

0.05

0.00

0.00

0.6

0.8

o

0.15

0.05

0.4

1.0

750 C o 775 C o 800 C o 825 C o 850 C Al2O3 as catalyst

dx/dt

0.15

0.2

0.8

Conversion x(-)

Conversion x(-)

0.0

0.6

0.0

1.0

0.2

0.4

0.6

0.8

1.0

Conversion x(-)

Conversion x(-)

Fig. 5 e dx/dt vs x for CO and H2 at different temperatures.

Table 3 e Activation energies by the Friedman method. Conversion x (e) CO

0.2 0.3 0.4 0.5 0.6 0.7 0.8 Equal value

g-Al2O3 as catalyst

Gangue as catalyst CO

H2

H2

E/(kJ/mol)

R2

E/(kJ/mol)

R2

E/(kJ/mol)

R2

E/(kJ/mol)

R2

41.2 61.7 79.6 92.0 90.5 94.6 96.9 79.1

0.99 0.99 0.99 1.00 1.00 0.99 0.93

23.0 26.9 29.9 48.6 74.9 99.2 88.7 55.9

0.94 0.99 0.99 0.99 0.97 0.96 0.94

48.2 48.7 49.3 49.6 53.3 58.2 78.2 54.3

1.00 0.99 0.99 0.99 0.99 0.99 0.97

29.9 34.3 36.8 44.5 51.6 65.0 70.5 47.5

0.95 0.99 1.00 0.99 0.97 0.95 0.96

corresponding activation energy for generating H2 was only 55.9 kJ/mol, which was much lower compared with CO. When g-Al2O3 was used, the average activation energies for generating CO and H2 dropped to 54.3 kJ/mol and 47.5 kJ/mol respectively. It was apparent that the average activation energies using coal gangue was lower for generating both CO and H2, implying that catalytic activity of g-Al2O3 was better than coal gangue in the decomposition process of phenol. It can be deduced that although many kinds of metallic oxide in the

coal gangue, its catalytic performance was reduced due to the high SiO2 content. In addition, from the values of activation energy for generating CO and H2, it was apparent that the values for generating H2 was lower which meant that H2 formation was easier during phenol decomposition process. As suggested by Horn et al. [30] and Emdee et al. [31], the formation of CO was relevant to the breaking of CeC bond and H2 was the product of H radical evolving, the activation energy values of CO and

13387

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

H2 can also represent the difficult of breaking CeO and CeH bonds. Thus, the reactions for generating CO was more different to start compared with the reactions for H2 generating, which corresponded to the higher activation energy for generating CO in this study.

Table 4 e Kinetic parameters by the universal integral method. Catalyst

Gas

G(x)

E/(kJ/mol)

A/(1/s)

x

Gangue

CO

G(3) G(4) G(16) G(3) G(4) G(16) G(3) G(4) G(16) G(3) G(4) G(16)

48.8 53.3 78.6 45.1 45.9 54.7 44.7 54.9 48.4 47.7 46.7 47.4

19.8 24.8 319.3 25.0 6.9 16.7 15.0 53.7 194.5 24.5 12.3 166.3

0e0.9 0e0.9 0e0.9 0e0.85 0e0.85 0e0.85 0.1e0.9 0.1e0.85 0.1e0.9 0e0.95 0e0.95 0e0.95

Kinetic parameters by the integral method For all the experiments under isothermal conditions, the integral method can be used to deduce the activation energy and pre-exponential factor based on the mechanism function, G(x). In this work, the most probable mechanism functions G(x) for phenol decomposition are determined by fitting of G(x) with the reaction time according to Eqs. (7)e(8). G(3), G(4) and G(16) appeared the best fitting accuracy among the nineteen reaction models. Fig. 6 illustrates the resulting fittings for the three mechanism models that enabled the better description for the isothermal testing data of the micro fluidized bed under different conditions. The three models used enabled the linear fitting in a range wider than 0.1e0.85, indicating that the phenol decomposition behavior is reasonably described by these three mechanism functions. Table 4 shows the activation energies calculated by the mechanism function models G(3), G(4), and G(16). The most probable reaction model can be determined from the three probable models based on the closeness of the activation

H2

Al2O3

CO

H2

energies obtained from the universal integral method and Friedman method, as suggested by Yu et al. [26]. Compared with the average values in Table 4, G(16) and G(4) could be used to describe the evolution profiles of CO using coal gangue and g-Al2O3 as catalyst respectively. The corresponding preexponential factor was 319.3 and 53.7 s1 respectively. The activation energies obtained by G(16) for both the two catalysts were closer to the average values of H2 in Table 3, corresponding to the pre-exponential factors of 16.7 and 166.3 s1respectively.

Fig. 6 e The fitting of three typical mechanism functions with time under different conditions.

13388

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

The thermal decomposition of phenol in the micro fluidized bed using sand in the upper stage was studied in our previous study [32], showing that the activation energies for generating CO and H2 were 137.5 kJ/mol and 85.1 kJ/mol respectively. Horn et al. [30] determined the value of activation energy for phenol decomposition into C5H6 and CO as 254 kJ/mol with A ¼ 1.02  1012 s1 using a shock-tube reactor. The further decomposition of the C5H6O into CO and C5H5 was studied by Frank et al. [33], and the obtained activation energy was 184 kJ/mol. It is apparent that the activation energies lowered significantly using gangue and Al2O3 as catalyst, particularly toward CO formation. Since activation energy represents the minimum energy required to break the chemical bonds between atoms during a chemical reaction process, higher value of the activation energy leads to slower reaction. The activation energy value obtained using gangue as catalyst was apparently lowered compared with sand, indicating that the gangue has the catalytic performance toward biomass tar. Thus, it can be concluded that the existence of coal gangue promotes the forming of CO and H2, which in turn accelerates the phenol decomposition process. However, The activation energy values obtained using gangue were a little higher compared with g-Al2O3 for both CO and H2 formation. It can be concluded that g-Al2O3 has better catalytic performance toward phenol decomposition. It has been reported that Al2O3 was quite effective for eliminating tar in the biomass pyrolysis process [34,35]. Calcined gangue in this paper represents one type of catalyst for tar decomposition that prepared by the thermal conversion of inferior coal or biomass. Although the gangue contains many kinds of metallic oxides, SiO2 accounts for more than half of the gangue ash. Thus, it can be speculated that the catalytic performance of gangue was weakened by the high content of SiO2. Considering the large production of gangue during coal mining activities, it will be meaningful to use gangue as catalyst for the tar decomposition in biomass thermal conversion. Adopting this approach, biomass and gangue can be utilized together to produce clean combustible gases.

Conclusion In this study, using phenol as a surrogate for biomass tars, the catalytic effect of calcined coal gangue and g-Al2O3 on the phenol decomposition was investigated in a micro twostage fluidized bed reactor under isothermal conditions. The results demonstrated that the catalytic performance of coal gangue and g-Al2O3 can explained by the activation energy obtained for phenol decomposition. g-Al2O3 provided a better catalytic performance toward phenol decomposition process, in terms of having lower activation energies for H2 and CO generating. The values of activation energy for generating H2 varied significant with that of CO, indicative of the different mechanisms involved in forming these two gas species. The activation energy for generating H2 was lower using these two catalysts, implying the formation of H2 is easier compared with CO in the decomposition process of phenol.

Acknowledgment This work was financially supported by the National Natural Science Foundation of China (51406226), National key foundation for exploring scientific instrument (2011YQ12003906) and China Postdoctoral Science Foundation (2014M551693). We gratefully acknowledge the valuable cooperation with Prof. Xu Guangwen and the members of his laboratory.

references

[1] Kalinci Y, Hepbasli A, Dincer I. Biomass-based hydrogen production: a review and analysis. Int J Hydrogen Energy 2009;34:8799e817. [2] Hosseini SE, Wahid MA, Ganjehkaviri A. An overview of renewable hydrogen production from thermochemical process of oil palm solid waste in Malaysia. Energy Convers Manag 2015;94:415e29. [3] Guan GQ, Kaewpanha M, Hao XG, Abudula A. Catalytic steam reforming of biomass tar: prospects and challenges. Renew Sustain Energy Rev 2016;58:450e61. [4] Li D, Koike M, Chen J, Nakagawa Y, Tomishige K. Preparation of NieCu/Mg/Al catalysts from hydrotalcite-like compounds for hydrogen production by steam reforming of biomass tar. Int J Hydrogen Energy 2014;39:10959e70. [5] Li C, Suzuki K. Tar property, analysis, reforming mechanism and model for biomass gasificationean overview. Renew Sustain Energy Rev 2009;13:594e604. [6] Tursun Y, Xu SP, Wang GY, Wang C, Xiao YH. Tar formation during co-gasification of biomass and coal under different gasification condition. J Anal Appl Pyrolysis 2015;111:191e9. [7] Palma CF. Modelling of tar formation and evolution for biomass gasification: a review. Appl Energy 2013;111:129e41. [8] Di Carlo A, Borello D, Sisinni M, Savuto E, Venturini P, Bocci E, et al. Reforming of tar contained in a raw fuel gas from biomass gasification using nickel-mayenite catalyst. Int J Hydrogen Energy 2015;40:9088e95. [9] Zhang S, Asadullah M, Dong I, Tay HL, Li CZ. An advanced biomass gasification technology with integrated catalytic hot gas cleaning. Part II: tar reforming using char as a catalyst or as a catalyst support. Fuel 2013;112:646e53. [10] Guo FQ, Dong YP, Dong L, Guo C. Effect of design and operating parameters on the gasification process of biomass in a downdraft fixed bed: an experimental study. Int J Hydrogen Energy 2014;39:5625e33. [11] Krerkkaiwan S, Mueangta S, Thammarat P, Jaisat L, Kuchonthara P. Catalytic biomass-derived tar decomposition using char from the co-pyrolysis of coal and giant leucaena wood biomass. Energy Fuels 2015;29:3119e26. [12] Kaewpanha M, Guan G, Ma Y, Hao X, Zhang Z, Reubroychareon P, et al. Hydrogen production by steam reforming of biomass tar over biomass char supported molybdenum carbide catalyst. Int J Hydrogen Energy 2015;40:7974e82. [13] Liu H, Liu Z. Recycling utilization patterns of coal mining waste in China. Resour Conserv Recycl 2010;54:1331e40. [14] Zhou CC, Liu GJ, Wu SC, Lam PK. The environmental characteristics of usage of coal gangue in bricking-making: a case study at Huainan, China. Chemosphere 2014;95:274e80. [15] Xiao H, Ma X, Liu K. Co-combustion kinetics of sewage sludge with coal and coal gangue under different atmospheres. Energy Convers Manag 2010;51:1976e80. [16] El-Rub ZA, Bramera EA, Brem G. Experimental comparison of biomass chars with other catalysts for tar reduction. Fuel 2008;87:2243e52.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 1 3 3 8 0 e1 3 3 8 9

[17] Zhang LX, Matsuhara T, Kudo S, Hayashi J, Norinaga K. Rapid pyrolysis of brown coal in a drop-tube reactor with cofeeding of char as a promoter of in situ tar reforming. Fuel 2013;112:681e6. [18] Wang FJ, Zhang S, Chen ZD, Liu C, Wang YG. Tar reforming using char as catalyst during pyrolysis and gasification of Shengli brown coal. J Anal Appl Pyrolysis 2014;105:269e75. [19] Polychronopoulou K, Bakandritsos A, Tzitzios V, Fierro JLG, Efstathiou AM. Absorption-enhanced reforming of phenol by steam over supported Fe catalysts. J Catal 2006;241:132e48.  J, Farriol X, Montane  D. Steam reforming [20] Coll R, Salvado model compounds of biomass gasification tars: conversion at different operating conditions and tendency towards coke formation. Fuel Process Technol 2001;74:19e31. [21] Morf P, Hasler P, Nussbaumer T. Mechanisms and kinetics of homogeneous secondary reactions of tar from continuous pyrolysis of wood chips. Fuel 2002;81:843e53. [22] Huelsman CM, Savage PE. Reaction pathways and kinetic modeling for phenol gasification in supercritical water. J Supercrit Fluids 2013;81:200e9. [23] Guo FQ, Dong YP, Lv ZC, Fan PF, Yang S, Dong L. Pyrolysis kinetics of biomass (herb residue) under isothermal condition in a micro fluidized bed. Energy Convers Manag 2015;93:367e76. [24] Gai C, Dong YP, Lv ZC, Zhang ZL, Liang JC, Liu YT. Pyrolysis behavior and kinetic study of phenol as tar model compound in micro fluidized bed reactor. Int J Hydrogen Energy 2015:7956e64. [25] Guo FQ, Dong YP, Lv ZC, Fan PF, Yang S, Dong L. Kinetic behavior of biomass under oxidative atmosphere using a micro-fluidized bed reactor. Energy Convers Manag 2016;108:210e8.

13389

[26] Yu J, Zeng X, Zhang JW, Zhong M, Zhang GY, Wang Y, et al. Isothermal differential characteristics of gasesolid reaction in micro-fluidized bed reactor. Fuel 2013;103:29e36. [27] Yu J, Liu Z, Zhu JH, Dong L, Xu GW. Kinetics and mechanism of solid reactions in a micro fluidized bed reactor. AIChE J 2010;56:2905e12. [28] Friedman HL. Kinetics of thermal degradation of charforming plastics from thermogravimetry. Application to a phenolic plastic. J Polym Sci Part C Polym Symp 1964;6:183e95. [29] Guo FQ, Dong YP, Dong L, Jing YZ. An innovative example of herb residues recycling by gasification in a fluidized bed. Waste Manag 2013;33:825e32. [30] Horn C, Roy K, Frank P, Just T. Shock-tube study on the hightemperature pyrolysis of phenol. In: Symposium (International) on combustion, vol. 27; 1998. p. 321e8. [31] Emdee JL, Brezinsky K, Glassman I. A kinetic model for the oxidation of toluene near 1200 K. J Phys Chem 1992;96:2151e61. [32] Guo FQ, Dong YP, Fan PF, Lv ZC, Yang S, Dong L. Detailed kinetic study of phenol decomposition under isothermal conditions to understand tar catalytic cracking process. J Anal Appl Pyrolysis 2016;118:155e63. [33] Frank P, Herzler J, Just T, Wahl C. High-temperature reactions of phenyl oxidation. In: Symposium (International) on combustion, vol. 25; 1994. p. 833e40. [34] Aysu T, Durak H. Catalytic pyrolysis of liquorice (Glycyrrhiza glabra L.) in a fixed-bed reactor: effects of pyrolysis parameters on product yields and character. J Anal Appl Pyrolysis 2015;111:156e72. [35] Tao J, Lu Q, Dong CQ, Du XZ, Dahlquist E. Effects of electric current upon catalytic steam reforming of biomass gasification tar model compounds to syngas. Energy Convers Manag 2015;100:56e63.