Thermogravimetric study on the pyrolysis kinetics of apple pomace as waste biomass

ww.sciencedirect.com

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 en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 7

Available online at w

ScienceDirect

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

Thermogravimetric study on the pyrolysis kineticsof apple pomace as waste biomass

M. R. Baray Guerrero a, M. Marques da Silva Paula b,M. Mel�endez Zaragoza a, J. Salinas Guti�errez a, V. Guzm�an Velderrain a,A. L�opez Ortiz a, V. Collins-Martınez a,*

a Departamento de Materiales Nanoestructurados, Centro de Investigaci�on en Materiales Avanzados, S.C.,

Miguel de Cervantes 120, Chihuahua, Chih. 31109, Mexicob Laboratory of Synthesis of Multifunctional Complexes, PPGCS, Universidade do Extremo Sul Catarinense,

88806-000 Criciuma, SC, Brazil

a r t i c l e i n f o

Article history:

Available online 23 July 2014

Keywords:

Apple pomace

Waste pyrolysis

TGA

Kinetics

* Corresponding author. Tel.: þ52 614439112E-mail address: [email protected]

http://dx.doi.org/10.1016/j.ijhydene.2014.06.00360-3199/Copyright © 2014, Hydrogen Ener

a b s t r a c t

Biomass waste-to-energy is an attractive alternative to fossil feedstocks because of

essentially zero net CO2 impact. A viable option consists in an integrated process, in which

biomass is partly used to produce valuable chemicals with residual fractions employed for

hydrogen production. One example of a biomass waste is the apple pomace, which is the

residue generated in the process of extraction of apple juice. In this research, a kinetic

study of the pyrolysis of apple pomace biomass (APB) was performed by TGA aiming its

liquid and gaseous products be utilized for the production of valuable chemicals and

hydrogen. Characterization of APB consisted in calorific value, compositional, proximal

and elemental analyzes. Kinetics were evaluated using three iso-conversional TGA models

at 5, 10, 15 and 20 �C/min. Activation energy values of 213.0 and 201.7 kJ/mol were within

the range for hemicellulose and cellulose, respectively, which are the main components of

biomass.

Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

Introduction

The current global economic development is based on the

trade and processing of oil, however, depletion is expected

during the first quarter of this century, which poses both

economic and energy supply problems because energy de-

mand is met mainly from fossil fuels [1]. Biomass as a

renewable source not only allows to partially replace fossil

fuels, but also to reduce concentrations of gaseous pollutants

9.u.mx (V. Collins-Martıne12gy Publications, LLC. Publ

(carbon oxides) emitted into the atmosphere [2]. Agro-

industrial residues represent a renewable source of energy,

as obtained in large quantities as a result of industrial pro-

cessing of fruits and vegetables and are a cheap raw material

for conversion to biofuels [3].

Moreover the use of renewable energy technologies such as

wind, geothermal, hydro, solar, hydrogen and those obtained

frombiomassarealternatives in themediumand long-termfor

the replacement of fossil fuels [4]. Todayhydrogen is generated

mostly from fossil fuels with a consequent, release of CO2

z).

ished by Elsevier Ltd. All rights reserved.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 716620

during its production stage.While, biomasswaste-to-energy is

an attractive alternative to fossil feedstocks because of

essentially zero net CO2 impact. A viable economical option

consists in an integrated process, in which biomass is partly

used to produce valuable materials or chemicals with residual

fractions employed for hydrogen production. Therefore, the

transformation of waste biomass into energy valued com-

pounds (i. e. H2) is a research field that is considered of great

importance in the present due to the current energy crisis and

environmental pollution issues [2]. Furthermore, biofuels

produced fromvarious lignocellulosicmaterials such aswood,

agricultural or forest residues, have the potential to be a valu-

able substitute (or supplement to gasoline biofuels) to liquid or

gaseous fuels for the transport sector [4].

One specific example of a biomass waste is the apple

pomace, which is the residue generated in the process of

extraction of apple juice. This apple pomace, is formed by a

complex mixture of shell, seed kernel, calyx, stem and soft

tissue, which is representative of the pomace, and this con-

tains mainly cellulose, hemicellulose, lignin and pectin.

The processing of biomass is grouped into three major

groups: biochemical, thermochemical and physicochemical.

Basically three types of thermochemical processes are

distinguished: pyrolysis, gasification and combustion. The

term pyrolysis refers to the incomplete thermal degradation

which leads to the production of coal tars and condensable

liquids and gases. In its strictest sense, pyrolysis must be

performed in complete absence of oxygen, however, this term

is nowused in a broader connotation, to describe the chemical

changes caused by the action of heat [5].

Moreover, pyrolysis is typically studied based on hypotheti-

cal models [6], where it is considered that the overall perfor-

mance of pyrolysis is the combination of the behavior of each

individual component [6,7]. Therefore, the determination of the

kinetic parameters provides key information of the processes

that take place, as well as the structure and composition of its

constituents [6]. Furthermore, the analysis of the thermal

degradation volatile products, identifies the gaseous species

emitted by the biomass, and thus provides insights to the pro-

cesses through which such decomposition occurs. The deter-

mination of the decomposition kinetics of lignocellulosic

biomass involves the knowledge of the reaction mechanisms.

However, the number of reactions occurring simultaneously in

the simplest pyrolysis process is so great that prevents to

develop a kinetic model that takes into account all these re-

actions. A kinetic study aims to reveal how the thermal

decomposition takes place (whether one ormore processes and

what range of conversions occur) through the characteristic ki-

netic constants provided by the kinetic models. This last is

critical to thedesign,constructionandoperationona large-scale

reactor for thepyrolysisofapplepomacesubject to study, for the

use of valued chemicals that may be generated or for the pro-

duction of hydrogen from gaseous products or simply to get rid

of certain wastes in a clean way [6]. The kinetic analysis of the

thermaldecompositionofbiomass isgenerally basedonthe rate

equation of decomposition of solids [8].

The present research is aimed to perform a basic charac-

terization of apple pomace (from the region of Cuauhtemoc,

Chihuahua, Mexico), the determination of the kinetic param-

eters (activation energy and pre-exponential factor) of the

pyrolysis reaction under a nitrogen atmosphere using differ-

ential and integral non-isothermal iso-conversional models.

Furthermore, the models employed in the present research

were: the differential Friedman and two integral

FlynneWalleOzawa (FWO) and KissingereAkahiraeSunose

(KAS)models onTGAdata for the apple pomacebiomass (APB).

Experimental

Sample characterization

Apple pomace sampleswere collected from the northern state

of Chihuahua (Cuauht�emoc, Chihuahua) and subjected to a

drying process, crushed, grounded and sieved to achieve a

particle size of 150 mm.

The elemental and proximal analyzes for the apple pomace

sample were performed using a Carlo Erba EA-1110 elemental

analyzer and an atomic emission spectrometer coupled with

ICP (ICP Thermo Jarrell Ash IRIS/AP DUO), calorific power was

determined through an adiabatic bomb calorimeter (Parr-1341

Oxygen Bomb Calorimeter) following the standard test

method ASTM D-2015-96.

Lignin, cellulose, hemicellulose and pectin content from

the pomace were determined using gravimetric techniques,

described in ASTM (E 1756-95, D1106-95) andASTM (D1103-60).

Moisture, volatiles and ash content was determined according

to the procedure described in ASTM E (871-82), ASTM (872-82)

and ASTM (1755e1795), respectively. In order to determine the

particle size (dp), samples were analyzed with dimensions:

dp < 150 mm (150 mm), 150 < dp < 180 mm (180 mm),

180 < dp < 250 mm (250 mm), and 250 < dp < 450 mm (450 mm)

under 100 cm3/min N2 flow and heating from room tempera-

ture to 800 �C at a rate of 10 �C/min. To verify the effect of the

heating rate on the generation of volatiles and to obtain the

kinetic parameters, the pomace sample was used with the

same particle size, which was subjected to different heating

rates of 5, 10, 15 and 20 �C/min.

Thermogravimetric analysis

TGA tests were carried out under an inert atmosphere (N2)

using a TGA-Q-500, TA Instruments equipment. Heating rates

(b) were controlled at 5, 10, 15 and 20 �C/min. Experiments

were performed under a nitrogen atmosphere with a flowrate

of 100 cm3/min and by duplicate. In all TGA tests between 20

and 30 mg of apple pomace biomass (APB) sample with a

specific particle size were deposited on the crucible of the

thermo balance. Then this sample was the subjected to a

specific heating rate from room temperature to 800 �C.

Kinetic models

During the pyrolysis primary reactions occur, so that the ki-

netic study of these are of paramount importance with TGA

being a very powerful tool. The determination of decomposi-

tion kinetics of lignocellulosic materials involves the knowl-

edge of the reaction mechanisms. However, the number of

reactions occurring simultaneously during a simple pyrolysis

process is so great that prevents the development of a kinetic

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 en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 7 16621

model that takes into account all these reactions. Thus, py-

rolysis is typically studied through hypothetical [5] models, in

which the overall pyrolysis behavior is considered as the

combination of each individual component [6,7].

Moreover, the kinetic study attempts to unravel how the

thermal decomposition takes place (whether one or more

processes and what range of conversions occur) through the

characteristic kinetic constants provided by the kinetic

models. This is critical to the design, build and operation of a

large-scale reactor for pyrolysis of the apple pomace biomass,

subject of the present study.

Kinetic analysis of the thermal decomposition of biomass

is generally based on the rate equation of decomposition of

solids [8]:

dadt

¼ A exp

��Ea

RT

�fðaÞ (1)

where t denotes time, a indicates the fraction of sample that

has reacted and the degree of conversion, da/dt is the rate of

the process, A and Ea are the pre-exponential factor and the

activation energy, respectively, from the Arrhenius equation

f(a) is a conversion function that represents the reaction

model used and depends on the controlling mechanism.

In this study, the degree of conversion, a, is defined as:

a ¼ W0 �WW0 �Wf

(2)

where w0, wf and w are the sample masses at the beginning,

end, and at a specific time t, respectively.

The unknown terms in equation (1) are the formal kinetic

parameters (A, Ea and f(a)) which are used to characterize

biomass pyrolysis reactions [9].

For non-isothermal reactions, where the heating rate,

b¼ dT/dt is constant, the above equationmay be expressed as:

dadt

¼ bdadT

¼ A exp

��Ea

RT

�fðaÞ (3)

The techniques developed for the evaluation of the kinetic

parameters for non-isothermal thermogravimetric analysis can

be divided into fitting models and free models. With the free

model is not necessary to assume a kinetic reaction, while ki-

netic parameters are obtained as a function of conversion or

temperature.Within suchmodels there are the iso-conversional

methods, which assume a constant degree of conversion, and

therefore the reaction rate depends only on the temperature.

Thus, these methods allow the estimation of the activation en-

ergy, Ea as a function of conversion, a, and independently of the

reactionmodel, f(a).

TheTGAdataprocessing of iso-conversionalmethodscanbe

either differential or integral. This paper presents results froma

differential (Friedman) and two integral; FlynneWalleOzawa

(FWO) and KissingereAkahiraeSunose (KAS) methods.

The Friedman iso-conversional method is a differential

technique, which involves taking natural logarithms of both

sides of equation (3) [10],:

ln

�dadt

�¼

�bdadT

�¼ ln½AfðaÞ� ¼

��Ea

RT

�(4)

It is assumed that the conversion function f(a) remains

constant, which means that the biomass degradation is

temperature independent and depends exclusively on the rate

of mass loss. A plot of ln(da/dt) versus 1/T for the same degree

of conversion of data taken at various heating rates, will result

in a series of lines with slopes equal to �Ea/R for each value of

conversion, a, at different heating rates.

The FlynneWalleOzawa method (FWO) is an integral iso-

conversional technique where regrouping the terms of equa-

tion (4), and integrating these with respect to a and T variables

and using the approximation of Doyle the following expres-

sion is obtained:

ln bylog

�A

Ea

RgðaÞ�� 2:315� 0:4567

�Ea

RT

�(5)

Thus, in the FWO method the plot of log(AEa/Rg(a)) vs 1/T

or ln(b) vs 1/T for different heating rates allows to obtain

parallel lines for a fixed degree of conversion. The slope

(�0.4567Ea/R) of these lines is proportional to the apparent

activation energy. If equal Ea values are obtained for different

values of a, it can be assumed with certainty that there is a

single reaction step. By contrast, a change in Ea with an in-

crease in the conversion degree is indicative of a complex

reaction mechanism [11].

Another widely used integral iso-conversional technique is

the KissingereAkahiraeSunose (KAS) method, obtained from

the CoatseRedfern approximation and based on the following

equation [12,13]:

ln

�b

T2

�¼ ln

�AR

EagðaÞ���Ea

RT

�(6)

Assuming that a has a fixed value, the activation energy

(Ea) can be determined from the slope of the straight line ob-

tained by plotting ln(b/T2) vs 1/T.

Results and Discussion

Chemical analyses

Table 1 presents results from the proximal, elemental and

compositional analyses for apple pomace.

From the results of Table 1 it can be seen that the apple

pomace has a small amount of N (0.78%), while for the case of

S, this could not be detected, which is advantageous because

it minimizes the corrosion problems associated with the

formation of acids in the process equipment [4]. It is also

evident that the greatest elemental amount corresponds to

carbon with a 47.98% followed by oxygen with 37.44%.

Moreover, there is a low ash (3.4%) and a high volatile

(81.32%) contents, characteristic of lignocellulosic materials,

which makes this biomass very attractive for thermal

degradation processes [9]. The lignocellulosic composition is

typically of biomass, although it is important to notice its

high cellulose content (47.49%). With respect to the mineral

composition of the ashes the major contributions come from

barium (37.75%) and titanium (17.51%).

Thermogravimetric analysis

Experimental TGA curves obtained for the apple pomace

under different heating rates are presented in Fig. 1. The

Table 1 e Results from calorific power, proximal,elemental and compositional analyses for the APB.

Parameter Unit Magnitude

Elemental Analysis

C % 47.98

H % 6.65

N % 0.78

O % 37.44

S % N.D

Compositional Analysis

Cellulose % 47.49

Hemicellulose % 27.77

Lignin % 24.72

Proximate Analysis

Moisture % 8.87

Fixed Carbon % 6.41

Volatile matter % 81.32

Ash % 3.40

Composition of ash

Al % 0.51

B % 11.43

Ba % 37.75

Ca % 2.01

Cr % 0.80

Cu % 8.50

Fe % 0.35

K % 7.73

Mg % 1.54

Mn % 10.24

Na % 0.23

Ti % 17.51

V % 0.56

Zn % 0.84

Removable

Ethanol % 2.89

Physical Properties

Density kg/m3 1103

Calorific Power kJ/kg 22,420

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 716622

thermal decomposition of the apple pomace reveals two

main regions attributed to the decomposition of cellulose and

hemicellulose. The first decomposition mass loss that occurs

at low temperatures can be associated to the process of py-

rolysis of hemicellulose and at higher temperatures the

0 100 200 300 400 500 600 700 800 900

0

5

10

15

20

25

30

35

Wei

ght,

%

Temperature, °C

5 °C/min 10°C /min 15 °C/min 20 °C/min

Fig. 1 e TGA curves for the pyrolysis of apple pomace under

N2 at different heating rates.

weight loss is associated with decomposition of cellulose.

Moreover, mass losses for the decomposition of lignin are not

observed in this temperature range. Understanding of the

volatilization of apple pomace is important because pyrolysis

is the first step in a process of gasification or combustion [8].

Furthermore, in this TGA plot it is observed that the greatest

amount of volatile material is produced at a heating rate of

5 �C/min, which is the curve that ends its decomposition at a

lower temperature (~540 �C). When heating rate increases,

the required time to reach a certain temperature value in-

creases, enabling dehydration, depolymerization, carbonyla-

tion, carboxylation and transglycosylation reactions. As a

consequence, the amount of devolatilized matter is

increased. The obtained curves at different heating rates,

after a certain decomposition stage at high heating rate,

reach a common value typical of the mass solid residue [14].

The largest sample weight loss is located from a temperature

range of 200e450 �C, as can be observed in Fig. 1, and this can

be attributed to the devolatilization process. Analyzing this

behavior, according to the literature the weight loss that oc-

curs at 200 �C is related to the beginning of the lignin and

hemicellulose pyrolysis contained in the apple pomace [15].

From 250 and up to approximately 350 �C the high decom-

position rate arises and in this region the maximum devo-

latilization of hemicellulose, cellulose and lignin is achieved.

Remaining molecules of these compounds generate the next

weight loss that corresponds to the temperature (up to

550 �C) in which the reaction ends.

Fig. 2 presents a TGA plot of the apple pomace biomass

(APB) subjected to different particle size fractions at a heating

rate of 10 �C/min. In this Figure it can be observed that particle

sizes smaller or equal than 150 mm are the ones that allow the

generation of the greater amount of volatile matter. The dif-

ference in volatiles production with respect to the size of the

particles is attributed to a main reason: particles greater than

425 mm, during the devolatilization process, some problems

arise related to the heat and mass transfer [16], as reported by

Lou and Stanmore [17].

Fig. 2 e TGA curves for the pyrolysis of the apple pomace

under N2 at a heating rate of 10 �C/min and different

particle sizes.

Fig. 3 e DTGA curves for the pyrolysis of apple pomace

under N2 at different heating rates.

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 en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 7 16623

The derivative of weight vs temperature plots for the

different heating rates (b) of 5, 10, 15, 20 �C/min for the apple

pomace are shown in Fig. 3. In TGA and derivative

thermogravimetric analysis (DTGA) from the pyrolysis of

lignocellulosic materials typically, at least three peaks are

observed, which can be associated to the cellulose,

200 300 400 500 600 700 800 900 1000 11000.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion,

Temperature, K

5 °C/min

200 300 400 500 600 700 800 900 1000 11000.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion,

Temperature, K

15 °C/min

Fig. 4 e Isothermal residence time effect for

hemicellulose and lignin. Thus, indicating that although

there appear interactions between fractions they usually

overlap in their decomposition, while their identity is

maintained [5]. Specifically, at temperatures below 200 �Cthere is a small change in conversion of the sample and is

this usually attributed to moisture removal, which is bonded

to the surface of the sample. The apple pomace decompo-

sition started around 250 �C as shown in Figs. 1 and 2. While,

in Fig. 3 a displacement of the curves to the right with the

increase in heating rate is observed. This shift happened

due to a greater reaction times that occurred at higher

temperatures, In addition, the maximum decomposition

rate tends to increase at higher heating rates, because a

greater thermal energy is provided that facilitates the heat

transfer around and within the samples [18]. Furthermore,

TGA curves show that the major decomposition occurs be-

tween 220 and 600 �C. Given the fact that the biomass con-

tains mainly cellulose, hemicellulose, lignin and pectin, it

has been found that cellulose decomposes between 277 and

427 �C, hemicellulose around 197 and 327 �C and lignin be-

tween 277 and 527 �C [19]. Also, it can be observed that the

decomposition of the pomace after 400 �C proceeds at a

slower rate because of the characteristic lignin decomposi-

tion rate [20].

200 300 400 500 600 700 800 900 1000 11000.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion,

10 °C/min

Temperature, K

200 300 400 500 600 700 800 900 1000 11000.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion,

Temperature, K

20 °C/min

the apple pomace (from 300 to 1070 K).

Fig. 5 e Friedman differential model for the calculation of the activation energies of APB.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 716624

Kinetic analysis

The thermal behavior of the apple pomace was studied

through TGA in the temperature range of 300e1070 K. In Fig. 4

the change in conversionwith temperature for all four heating

Fig. 6 e FWO integral model for the calcula

rates; 5, 10, 15 and 20 �C/min in a nitrogen environment can be

seen. These plots were used for the kinetic analysis based on

the three kinetic models above described.

According to the Friedman differential model, activation

energy, Ea, based on equation (4) can be determined from a

tion of the activation energies of APB.

Fig. 7 e KAS integral model for the calculation of the activation energies of APB.

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 en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 7 16625

plot of ln[da/dt] versus 1/T and results are presented in Fig. 5.

Here, the slopes of the iso-conversional lines give (�Ea/R) at

progressing conversion degrees a. Activation energy has also

been calculated using the integral FWO method using equa-

tion (5) and Fig. 6 presents the iso-conversional linear plots of

ln(b) versus 1/T where slopes give �0.453 Ea/R at progressing

conversion degrees for this method. Finally, Fig. 7 shows re-

sults for the integral KASmethod using the equation (6) where

linear plots of ln(ß/T2) versus 1/T provide slopes to determine

(�Ea/R) at progressing conversion degrees a.

The calculated activation energies for the Friedman, OFW

and KAS methods are presented in Table 2. Because of low

correlation values at conversion degrees below 0.2 and above

0.8 these values are not included [21]. The mean activation

energies calculated from Friedman, OFW and KAS methods

were 197.7, 213.0 and 201.7 kJ/mol, respectively. Also, excel-

lent linear correlation coefficients were obtained with a R2

very close to 1 with 0.973, 0.996 and 0.978, for the Friedman,

Table 2 e Activation energy results as a function of (a) for the

Conversion (a) E, model KAS,kJ/mol

R2 E, modekJ/m

0.1 137 0.9984 14

0.2 189 0.9967 19

0.3 228 0.9954 23

0.4 237 0.9963 24

0.5 280 0.9732 29

0.6 170 0.9589 18

0.7 134 0.9661 18

0.8 174 0.9596 14

0.9 105 0.9501 11

OFW and KASmethods respectively. Results obtained from all

models were in a good agreement with a deviation below 8%.

The small deviations from the highest activation energy

(OFW) with respect to the Friedman and KAS methods were

7.1 and 5.6%, respectively, which validate the reliability of

calculations and confirmed the predictive power of KAS and

OFW methods [22].

Kinetic analysis results showed that activation energy is

highly depended on conversion which means that the apple

pomace pyrolysis is a complex process consisting of several

reactions. Fig. 8 shows the change in activation energy with

respect to progressing conversions. For the calculated values

from Friedman, FWO and KASmodels Ea increases from 0.2 to

0.5 conversions. While, all the models follow almost the same

increasing Ea trend with respect to conversion, and they peak

at a conversion around a ¼ 0.49. Different reaction mecha-

nisms are responsible for the change in Ea values as conver-

sion increases. Due to the fact that activation energy is the

Friedman, FWO and KAS kinetic models.

l FWO,ol

R2 E, model Friedmann,kJ/mol

R2

5 0.9969 148 0.9966

8 0.9957 213 0.9963

8 0.9959 235 0.9863

7 0.9958 249 0.9994

0 0.9916 267 0.9764

2 0.9987 175 0.9575

9 0.9975 145 0.9467

7 0.9994 100 0.9469

9 0.9825 78 0.904

Fig. 8 e Activation energy as a function of progressing conversions (a) for the Friedman, FWO and KAS kinetic models.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 716626

minimum energy required for a reaction to begin, the higher

the Ea values the slower reactions will proceed. Furthermore

Ea is also employed in the calculation of the reactivity of a fuel

[23]. Gai et al. reported the kineticmechanism of rice husk and

corn cob and calculated Ea values of 79 and 129 kJ/mol,

respectively. Damartzis et al. [21] studied the pyrolysis ki-

netics of cardoon and found an Ea ¼ 224.1 kJ/mol for cardoon

stems and 350.07 kJ/mol for cardoon leaves. LopezeVelazquez

et al. [22], reported Ea values for pyrolysis of orange waste

between 120 and 250 kJ/mol. Amutio et al. [24] studied pine-

wood waste pyrolysis kinetics and found that Ea changed be-

tween 62 and 206 kJ/mol. In general, calculated Ea values for

apple pomace was similar to those reported in previous waste

biomass studies.

Moreover, the values of activation energy obtained in the

present research for the models applied are within the range

of values of activation energy of hemicellulose (67e105 kJ/

mol), cellulose (210e240 kJ/mol) and lignin (65e67 kJ/mol) [25].

Huang et al. [25] have studied the evolution of the activation

energy values as a function of the degree of conversion,

finding lower Ea values for small conversions, a growth of

Table 3 e Pre-exponential calculation results as a function of (

Conversion (a) b ¼ 5 �C/min b ¼ 10 �C

0.1 0.000292856 0.000585

0.2 0.000574277 0.001148

0.3 0.000858529 0.001717

0.4 0.001174112 0.002348

0.5 0.001518704 0.003037

0.6 0.001783433 0.003566

0.7 0.002189155 0.004378

0.8 0.002781244 0.005562

0.9 0.003803871 0.007607

these at intermediate conversions to return to low values to-

wards the end. These authors have related this behavior to the

decomposition of the hemicellulose, cellulose and lignin

fractions. Since, Ea values obtained in these conversion ranges

are close to the tabulated values of the pure compounds.

However, it should be noted that the activation energy values

that are determined for any conversion value should not be

considered as the actual values of a particular reaction step,

but as an apparent value that represents the contributions of

numerous parallel and competing reactions, which contribute

to the overall reaction rate. For such a complex biomass

devolatilization process, the contributions will vary with the

temperature and the conversion and very often overlap one

another [26,27].

In order to validate the kinetic parameters, the pre-

exponential factors as a function of conversion were deter-

mined using CoatseRedfern method [28,29]. Since the KAS

method ismore reliable, activationenergiesobtained from this

model was used in the CoatseRedfern equations for calcula-

tion of the pre-exponential factor. Calculated pre-exponential

factor values are shown in Table 3. It is important to notice

a) and (b) for the KAS kinetic model.

/min b ¼ 15 �C/min b ¼ 20 �C/min

713 0.000878569 0.00087857

554 0.001722831 0.00172283

057 0.002575586 0.00257559

223 0.003522335 0.00352234

408 0.004556113 0.00455611

865 0.005350298 0.0053503

311 0.006567466 0.00656747

487 0.008343731 0.00834373

743 0.011411614 0.01141161

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 en e r g y 3 9 ( 2 0 1 4 ) 1 6 6 1 9e1 6 6 2 7 16627

that in the CoatseRedfern model the pseudo-order n, has no

physical meaning and therefore this parameter was not

calculated in the present research. Finally, it is important to

notice that thepresent study is thefirst report onapplepomace

biomass (APB) pyrolysis kinetics and the obtained mid-

activation energy values for APB makes it an attractive ener-

getic waste biomass for a waste-to-energy potential fuel.

Conclusion

In this study the pyrolysis of waste apple pomace biomass

(APB) has been investigated for the first time by means of

thermogravimetric analysis. The low moisture and ash con-

tent and high volatile matter makes apple pomace a high

potential candidate for production of bio-chemicals and with

further processing for hydrogen production. Apple pomace

pyrolysis kinetics using data obtained from TGA analysis

showed good agreement with experimental data. This kinetic

data will be an important tool to model, design and develop a

thermochemical system for apple pomace in the near future.

The results of this study are crucial as they provide many

options for future application of APB as a waste-to-energy

resource for energy and chemicals.

Acknowledgments

This researchwas supported by Consejo Nacional de Ciencia y

Tecnologıa (CONACYT-M�exico)

r e f e r e n c e s

[1] Bassam EN. Renewable energy for rural communities. RenewEnergy 2001;24(3):401e8.

[2] Tang L, Huang H. Biomass gasification using capacitivelycoupled RF plasma technology. Fuel 2005;16:2055e63.

[3] Hong Y, Yuan Z, Junguo L. Land and water requirements ofbiofuel and implications for food supply and theenvironment in China. Energy Policy 2009;37(5):1876e85.

[4] I.D.A.E.e Instituto para laDiversificaci�onyelAhorrodeEnergıa.Biomasa. Available at: http://www.idae.es/index.php/idpag.233/relcategoria.1037/relmenu.48/mod.pags/mem.detalle

[5] Conesa JA. Curso B�asico de An�alisis T�ermico.Termogravimetrıa, cin�etica de reacciones y an�alisis t�ermicodiferencial. Mayo: Universidad de Alicante; 2000.

[6] Conesa JA, Domene A. Biomasses pyrolysis and combustionkinetics through n-th order parallel reactions. ThermochimActa 2011;523:176e81.

[7] Garcıa Barneto A, Ariza Carmona J, Martın Alfonso JE, DıazBlanco J. Kinetic models based in biomass components forthe combustion and pyrolysis of sewage sludge and itscompost. J Anal Appl Pyrolysis 2009;86:108e14.

[8] Min F, Zhang M, Chen Q. Non-isothermal kinetics ofpyrolysis of three kinds of fresh biomass. J China Univ MinTechnol 2007;17(1):105e11.

[9] White JE, Catallo WJ, Legendre BL. Biomass pyrolysis kinetics:a comparative critical review with relevant agriculturalresidue case studies. J Anal Appl Pyrolysis 2011;91:1e33.

[10] Friedman H. Kinetics of thermal degradation of char-formingplastics from thermogravimetry. Application to a phenolicplastic. J Polym Sci Part C Polym Lett 1964;6(1):183e95.

[11] Ozawa T. A new method of analyzing thermogravimetricdata. Bull Chem Soc Jpn 1965;38:1881e6.

[12] Kissinger HE. Reaction kinetics in differential thermalanalysis. Anal Chem 1957;29:1702e6.

[13] Akahira T, Sunose T. Trans. Joint convention of fourelectrical institutes; 1969. p. 246.

[14] Zabaniotou AA, Kantarelis EK, Theodoropoulos DC.Sunflower shells utilization for energetic purposes in anintegrated approach of energy crops: laboratory studypyrolysis and kinetics. Bioresour Technol2008;99:3174e81.

[15] Munir S, Daood SS, Nimmo W, Cunliffe AM, Gibbs BM.Thermal analysis and devolatilization kinetics ofcotton stalk, sugar cane bagasse and shea meal undernitrogen and air atmospheres. Bioresour Technol2009;100:1413e8.

[16] Antony Raja S, Robert Kennedy Z, Pillai BC, Lindon RobertLee C. Flash pyrolysis of jatropha oil cake in electricallyheated fluidized bed reactor. Energy 2010;35:2819e23.

[17] Luo M, Stanmore B. The combustion characteristics of charfrom pulverized bagasse. Fuel 1992;71:1074e6.

[18] Alexander G, Wolfgang K, Sonia R. Tesis de Doctorado,Universi-dad Nacional de Colombia, Colombia y Universidadde Kassel; 2008. Alemania.

[19] Du ZY, Sarofim AF, Longwell JP. Activation-energydistribution in temperature-programmed desorption e

modeling and application to the soot e oxygen system.Energy Fuels 1990;4:296e302.

[20] Xiaodong Z, Min X, Rongfeng S, Sun L. Study on biomasspyrolysis kinetics. J Eng Gas Turbines 2006;128:493e6.

[21] Damartzis T, Vamvuka D, Sfakiotakis S, Zabaniotou A.Thermal degradation studies and kinetic modeling ofcardoon (Cynara cardunculus) pyrolysis usingthermogravimetric analysis (TGA). Bioresour Technol2011;102:6230e8.

[22] Lopez-Velazquez MA, Santes V, Balmaseda J, Torres-Garcia E.Pyrolysis of orange waste: a thermokinetic study. J Anal ApplPyrolysis 2013;99:170e7.

[23] Gai C, Dong Y, Zhang T. The kinetic analysis of the pyrolysisof agricultural residue under non-isothermal conditions.Bioresour Technol 2013;127:298e305.

[24] Amutio M, Lopez G, Aguado R, Artetxe M, Bilbao J, Olazar M.Kinetic study of lignocellulosic biomass oxidative pyrolysis.Fuel 2012;95:305e11.

[25] Huang YF, Kuan WH, Chiueh PT, Lo SL. Pyrolysis of biomassby thermal analysis-mass spectrometry (TA-MS). BioresourTechnol 2011;102(3):3527e34.

[26] Vyazovkin S. Model-free kinetics. J Therm Anal Calorim2006;83:45e51.

[27] Akinwale OA, Thomas JH, Marion C, Edson L, Meyer RS,Knoetze JH, et al. Non-isothermal kinetic analysis of thedevolatilization of corn cobs and sugar cane bagasse in aninert atmosphere. Thermochim Acta 2011;517:81e9.

[28] Suuberg E. Approximate solution technique for non-isothermal, Gaussian distributed activation energy models.Combust Flame 1983;50:243e5.

[29] Teng H, Hsieh CT. Activation energy for oxygenchemisorption on carbon at low temperature. Ind Eng ChemRes 1999;38:292e7.