Combustion of pistachio shell:

physicochemical characterization and

evaluation of kinetic parameters

Jean Constantino Gomes da Silva, José

Luiz Francisco Alves, Wendell Venicio
de Araujo Galdino, Regina de Fátima
Peralta Muniz Moreira, et al.
Environmental Science and Pollution
Research

ISSN 0944-1344

Environ Sci Pollut Res

DOI 10.1007/s11356-017-8945-1

1 23
Your article is protected by copyright and
all rights are held exclusively by Springer-
Verlag Berlin Heidelberg. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.

1 23
 Author's personal copy
Environ Sci Pollut Res
DOI 10.1007/s11356-017-8945-1

ADVANCES IN ENVIRONMENTAL BIOTECHNOLOGY AND ENGINEERING 2016

Combustion of pistachio shell: physicochemical characterization

and evaluation of kinetic parameters
Jean Constantino Gomes da Silva 1 & José Luiz Francisco Alves 2 &
Wendell Venicio de Araujo Galdino 1 & Regina de Fátima Peralta Muniz Moreira 2 &
Humberto Jorge José 2 & Rennio Felix de Sena 1 & Silvia Layara Floriani Andersen 1

Received: 22 November 2016 / Accepted: 27 March 2017

# Springer-Verlag Berlin Heidelberg 2017

Abstract The study of different renewable energy sources Introduction

has been intensifying due to the current climate changes;
therefore, the present work had the objective to characterize The socio-economic and energetic development led to an ex-
physicochemically the pistachio shell waste and evaluate ki- ponential increase in the use of fossil fuels for decades, where
netic parameters of its combustion. The pistachio shell was it was intensified during important periods in history such as
characterized through proximate analysis, ultimate analysis, the Industrial Revolution and the petroleum industry growth
SEM, and FTIR. The thermal and kinetic behaviors were eval- (Huber 2009). However, this increased consumption brought
uated by a thermogravimetric analyzer under oxidant atmo- environmental problems on a global scale, mainly due to the
sphere between room temperature and 1000 °C, in which the greenhouse gases emission (CO2, NOx, and SOx) in many
process was performed in three different heating rates (20, 30, different ways, from extraction to final use. There is a lot of
and 40 °C min−1). The combustion of the pistachio shell pre- discussion about these problems; thus, several nations’ leaders
sented two regions in the derivative thermogravimetric curves, have been meeting to discuss the limitation of greenhouse
where the first represents the devolatilization of volatile matter gases emissions and evaluate renewable energy alternatives,
compounds and the second one is associated to the biochar in order to decentralize the main energy supply of countries.
oxidation. These zones were considered for the evaluation of Among the several ways to obtain energy from renewable
the kinetic parameters Ea, A, and f(α) by the modified method resources, the biomass is presented as one of the main ones.
of Coats-Redfern, compensation effect, and master plot, re- Biomass is a renewable resource, defined as a non-fossil or-
spectively. The kinetic parameters for zone 1 were ganic and biodegradable material originated from plants, ani-
E a1 = 84.11 kJ mol −1 , A 1 = 6.39 × 10 6 min −1 , and mals, and microorganisms, chemically composed of hydrocar-
f(α)1 = 3(1 − α)2/3, while for zone 2, the kinetic parameters bons, which may have different forms. However, the
were E a2 = 37.47 kJ mol −1 , A 2 = 57.14 min −1 , and agroindustrial waste has been distinguished between different
f(α)2 = 2(1 − α)1/2. forms of biomass because of its low cost and it does not com-
pete with arable lands for food production (Ferreira-Leitão
et al. 2010). In this context, the residues generated by the
Keywords Combustion . Isoconversional method . Kinetic pistachio crop can present a considerable potential for energy
parameters . Pistachio shell application.
The pistachio (Pistacia vera L.) is widely cultivated in the
Responsible editor: Philippe Garrigues Mediterranean region being Iran and Turkey its main pro-
ducers. According to FAO (2015), the pistachio crop has ex-
* Jean Constantino Gomes da Silva hibited a constant growth in the last decades, which presented
[email protected]
an average annual growth rate of 8.5% between the years of
2000 and 2013. In 2014, 638,000 tons of pistachio (in shell)
1
Federal University of Paraíba, João Pessoa 58051-900, Brazil was produced in the world, which represents a growth of 37%
2
Federal University of Santa Catarina, Florianópolis 88040-900, compared to 2013 (Taghizadeh-Alisaraei et al. 2017). This
Brazil annual growth rate in production is accompanied by an
 Author's personal copy
Environ Sci Pollut Res

increase in the generation of residues from the pistachio shell, Several mathematical methods can be found in order to
which represents on average between 35 and 45% of the total obtain the kinetic parameters from the weight loss data as the
weight. This fact motivates the study for its exploitation, es- fitting model or isoconversional model. In the isoconversional
pecially when talking about renewable energy source. or free model, the activation energy can be determined from
Biomass can be converted into solid, liquid, or gas by dif- the different curves of weight loss data without the knowledge
ferent conversion routes, where the product of these processes of the pre-exponential factor and reaction model, while in the
can be used to generate electricity and heat, or used in chem- fitting model, the activation energy is determined from the
ical industries as a feedstock to convert into other products. reaction model. However, this anticipated knowledge for reac-
The conversion routes can be classified as thermochemical tion model leads to an evaluation of several models to get a
(combustion, gasification, and pyrolysis) and biochemical better approximation of the data, making the isoconversional a
(fermentation and digestion). Combustion is one of the most better choice to obtain the kinetic parameters. The kinetic pa-
used conversion processes of biomass; it is widely used in rameters are usually obtained through methods of Flynn-Wall-
several applications, from generating heat to producing elec- Ozawa, Friedman, Coats-Redfern, and Vyazovkin (Vyazovkin
tricity through turbines, which corresponds to 97% of the pro- et al. 2011).
duced bioenergy (Nanda et al. 2014). Concerning about the increase in generation of residues
Biomass consists mainly of cellulose, hemicellulose, and from pistachio crops, associated to the low number of studies
lignin, where these polymers are commonly called lignocellu- reported on its application in thermochemical processes, this
lose (Kan et al. 2016). The efficiency of thermal conversion study aims to evaluate the combustion process of the pistachio
processes can be provided by the mass of those compounds in shell (P. vera L.) characterizing it physicochemically as well
the biomass; therefore, compounds that present high concen- as to evaluate the thermal behavior under oxidant and inert
tration of hydrocarbons consequently have a high energy po- atmospheres to determine the kinetic parameters. The thermo-
tential. Besides, in the lignocellulose biomasses, there are or- gravimetric data were obtained from a thermogravimetric an-
ganic extractives and inorganic compounds (ashes); however, alyzer (TGA), where the combustion data were used to obtain
these compounds are in smaller concentrations (Anca-Couce activation energy, pre-exponential factor, and reaction model
and Obernberger 2016). The ashes present in the biomass can by means of the modified Coats-Redfern method, compensa-
be considered as a factor that affects the conversion process; tion effect, and master plot, respectively. In addition, the phys-
this is due to the presence of some inorganic compounds which icochemical characteristics of the pistachio shell were evalu-
have the possibility of melting at low temperatures, interfering ated through elemental analysis, proximate analysis, calorific
in heat and mass transfer process during combustion (Jenkins value, Fourier transform infrared spectroscopy (FTIR), and
et al. 1998). Few studies on the energetic application of the scanning electron microscopy (SEM).
pistachio shell by combustion process can be found in the
literature; however, the pyrolysis process has many studies.
The influence of the physical parameters on the products of Materials and methods
the pyrolysis in fixed bed reactor was evaluated by Açıkalın
et al. (2012), which observed the increase of the gas yield with Preparation and characterization
the elevation of the reaction temperature. Lua and Yang (2009)
presented the applicability of the SO2 adsorption by activated The samples of pistachio shell (P. vera L.) were initially dried
carbon produced from the pistachio shell, while Komnitsas in an oven at 60 °C for 24 h and subsequently grounded in a
et al. (2015) presented the potential applicability of biochar ball mill to obtain smaller particle sizes at a range between
for adsorption of heavy metals. 105 μm (140 mesh) and 125 μm (120 mesh). Pistachio shell
For development and optimization of combustion systems, was characterized by proximate analysis, ultimate analysis,
it is necessary the initial knowledge of physicochemical char- calorific value, FTIR, and SEM.
acteristics, thermal decomposition behavior, and kinetic pa- The proximate analysis was accomplished according to
rameters involving the decomposition reactions (activation ASTM E-1131 (ASTM 2003) for the determination of moisture,
energy, pre-exponential factor, and reaction model) ash, and volatile matter, while the fixed carbon was calculated
(Vyazovkin et al. 2011). Thermogravimetric analysis is a sim- from the difference in dry basis. This data was obtained through
ple technique commonly used to obtain this data, where dur- a thermogravimetric analyzer DTG-60 Shimadzu. The ultimate
ing analysis, one or more properties are monitored and curves analysis of pistachio shell was performed through a Series II
representing thermal events as a function of time or tempera- CHNS/O Analyzer Perkin Elmer 2400, where the composition,
ture are obtained. From these data, it is possible to evaluate the in weight, of carbon (C), hydrogen (H), nitrogen (N), and sulfur
stage of biomass decomposition and oxidation; moreover, it is (S) was obtained experimentally, while the oxygen (O) content
possible to obtain the values of the kinetic parameters of the was calculated using differences in C, H, N, S, and ash. The
decomposition reactions (Brown et al. 2000). heating value was calculated using Channiwala and Parikh
 Author's personal copy
Environ Sci Pollut Res

equation (Eq. 1), which is measured in megajoules per kilogram. equation which relates the influence of temperature over reac-
This correlation takes into consideration the weight of C, H, N, tion rate, where A is the pre-exponential factor (min−1), Ea is
S, and O obtained of ultimate analysis, and biomass ashes of the activation energy (J mol −1 ), R is the gas constant
proximate analysis, showing a good approximation of the results (8.314 J K−1 mol−1), and T is the temperature (K). In thermal
(Channiwala and Parikh 2002). This correlation can be applied analysis field, the pressure dependence is rarely mentioned in
to materials with C, H, O, N, S, and ash as 0.00–92.25, 0.43– kinetic studies using computational methods; for this, a high
25.15, 0.00–50.00, 0.00–5.60, 0.00–94.08, and 0.00–71.4%, re- flow rate of a purge gas in gas-solid reactions is necessary to
spectively, with an absolute error of 1.45%. effectively remove reactive gaseous products; thus, the condi-
tion h(P) became a constant (Vyazovkin et al. 2011). Another

 issue about Eq. 2 is the relationship between conversion and
HHV MJ kg−1 ¼ 0:3491C þ 1:1783H þ 0:1005S ð1Þ time. Generally, in thermogravimetric data, the conversion
−0:1034O−0:0151N−0:0211Ash processes are evaluated as a function of temperature, where
temperature increases linearly with time (dT = βdt), with β
FTIR is useful to obtain information about the chemical being the heating rate. Thus, admitting the process as
composition in terms of functional groups present in sample. h(P) = constant and adding β to Eq. 2, Eq. 3 is obtained:
The analysis was performed using a Shimadzu IRPrestige-21
using the KBr method, in a proportion of 1:100 (sample/KBr)
in weight, and pressed in pellet shape, through scanning be- dα A −Ea
¼ e RT f ðαÞ ð3Þ
tween 4000 and 400 cm−1. The biomass surface morphology dT β
was obtained by SEM on a JEOL JSM-6390LV scanning
electron microscope device. Equation 3 does not have an analytical solution; therefore,
The thermal behavior evaluation was performed using ap- there is a need to use several methods to calculate kinetic
proximately 10 mg of biomass with a particle size of about parameters. The isoconversional methods are used to deter-
105 μm, under oxidant (synthetic air, 79.997% N 2 and mine Ea without prior knowledge of A and f(α) and they are
20.000% O2) and inert (99.995% of N2) atmospheres at a frequently called Bmodel-free^ methods. The modified Coats-
heating rate of 20 °C min − 1 , with a gas flow of Redfern method (Eq. 4) is an isoconversional method widely
50 mL min−1. All analyses were performed in duplicate. used, and with a satisfactory approximation (Yao et al. 2008;
Magdziarz and Wilk 2013; Braga et al. 2014).
Kinetic analysis of combustion
    
β Aα R 2RT α Ea;α 1
ln ¼ ln − 1− − ð4Þ
The kinetic study of combustion was performed through three Tα 2 E a;α gðαÞ E a;α R Tα
different heating rates (20, 30, and 40 °C min−1) with an initial
mass sample of about 10 mg and particle size of about
The temperature dependence with conversion in
105 μm, in order to minimize the diffusion effects (Branca
isoconversional method is obtained by means of a series of
and Di Blasi 2013). The samples were heated from room
runs with different temperature programs, and then through
temperature to 1000 °C under oxidant atmosphere (synthetic
line plotting in terms of ln(β/Tα2) against 1/Tα. At Eq. 4, it is
air, 79.997% N 2 and 20.000% O 2 ) with a gas flow of
possible to obtain activation energy for each conversion value.
50 mL min−1. The runs were accomplished on a thermogravi-
Due to the impossibility to obtain the value of the pre-
metric analyzer DTG-60 Shimadzu with a resolution of 0.1 μg
exponential factor by isoconversional method, the pre-
and precision of temperature measurement of ±0.25 °C. The
exponential factor was obtained using the compensation meth-
precision data were reached by duplicate analysis of each
od (Eq. 5), which uses the activation energy value obtained
heating rate run (20, 30, and 40 °C min−1).
previously (Vyazovkin et al. 2011).
The kinetic study of the oxidative decomposition of the
solid state was based on a single-step reaction described in
Eq. 2: lnðAα Þ ¼ aEa;α þ b ð5Þ

The compensation effect consists in using the model-fitting

dα −E a
¼ k ðT Þf ðαÞhðPÞ ¼ Ae RT f ðαÞhðPÞ ð2Þ method for a single heating by different models fj(α) to deter-
dt
mine several lnAj e Eaj values, obtaining the compensation
where k(T) is the rate constant, f(α) is the reaction model, and parameters a and b, where the term i indicates each reaction
h(P) is a pressure-dependent term. Equation 2 represents the model showed in Table 1 (Khawam and Flanagan 2006). The
conversion rate (dα/dt) of the solid state to gas state as a reaction model can be obtained through the curve g(α) using
function of time. The rate constant is based on the Arrhenius the parameters Ea,α and Aα calculated.
 Author's personal copy
Environ Sci Pollut Res

The reaction model was obtained using the master plot The second decomposition stage is represented by the oxida-
based on the integral form of the kinetic data, calculated by tion of the solid carbonaceous residue through reactions at the
Eq. 6. solid surface between the carbon bonds and the oxygen
(Branca and Di Blasi 2013). Thus, the high content of volatile
matter and fixed carbon observed in Table 2 can be directly
g ðαÞ pð x Þ
¼ ð6Þ related to the amount of fuel material; moreover, these contents
gð0:5Þ pð0:5Þ are proportional to the reactivity of sample (García et al. 2012).
where g(α) represents the integral function result of f(α) The decomposition process of biomass occurs through the
shown in Table 1, while g(0.5) represents the result of the degradation of lignocellulose compounds; in this way, the
integral equation when the conversion reaches 50%. The ex- knowledge of chemical bonds and identification of these com-
pression p(x) can be calculated by the Senum-Yang approxi- pounds become important (Asadieraghi and Wan Daud 2014;
mation (Senum and Yang 1977). These reference theoretical Braga et al. 2015). Cellulose, hemicellulose, and lignin can be
curves depend on the kinetic model but generally they are identified through infrared spectroscopy. Figure 1 shows the
independent of the kinetic parameters of the process (Gotor FTIR spectrum for the pistachio shell. It is noted that the band
et al. 2000). at 3412 cm−1 corresponds to vibrations of hydroxyl groups
OH present in the lignocellulose in which this band can be
associated with water, acid groups, and alcohol groups. The
band at 2916 cm−1 is attributed to the asymmetric C–H
Results and discussion stretching of methylene groups in lignocellulose, however
with high intensity in the cellulose and hemicellulose. The
Characterization band at 2139 cm−1 is attributed to alkyne group stretching
vibration of C≡C. The aliphatic aldehyde C=O stretching is
Table 2 shows the results of the characterization of pistachio observed in the sample at wavelength 1735 cm−1, which is
shell waste through proximate analysis, ultimate analysis, and associated with hemicellulose. The band at 1647 cm−1 is as-
calculated heating value. sociated to C=C stretching in alkene groups at hemicellulose.
The results of proximate analysis show that the biomass has At the bands 1593 and 1506 cm−1 occur the C=C vibrations in
a favorable applicability in the combustion process mainly due aromatic rings present in the lignin. The band at 1460 cm−1
to its low moisture content and low ash content, which are corresponds to the C–H methylene scissoring at all lignocel-
important parameters in combustion process. A low moisture lulosic compounds while the band at 1379 cm−1 shows the
content can result in a higher heating value, because it does not symmetrical methyl C–H bending. The bands at 1427 and
need a previous evaporation process, while a low ash content 1246 cm−1 are attributed to C–O–H in-plane bending and
reduces the chance of problems caused by fouling and slagging C–O stretching, respectively, for carboxylic acid groups in
in equipment. The results of the moisture and ash contents, lignin. The band at 1329 cm−1, which has the first overtone
together with the elemental composition of the biomass, can of the C–H bending vibration, appears as a weak, broad band
be associated to the calorific value through Eq. 1. The high in the 1370–1220 cm −1 region. The band at 1165 and
oxygen/carbon ratio may cause a negative effect during the 1047 cm−1 can be correlated with C–O stretching vibrations
first stage of decomposition, which results in a reduction in in alcohols and phenols of lignocellulose which produce a
the higher heating value (HHV) (Jenkins et al. 1998). Similar strong band in the 1260–1000 cm−1 region of the spectrum.
HHV was reported by Açıkalın et al. (2012), where the pista- The band at 899 cm−1 can be attributed to the glycosidic link-
chio shell studied by the author presented a calorific value of age in the cellulose, while the band at 604 cm−1 can be related
~17.39 MJ kg−1. When comparing the HHV of the pistachio with =C–H bending in alkynes (700–600 cm−1) (Xu et al.
shell to other biomass used in combustion systems, such as 2013; Asadieraghi and Wan Daud 2014).
sugarcane bagasse (~18 MJ kg −1 ), pinewood sawdust The microstructures of the pistachio shell, obtained by
(15.90 MJ kg−1), and coconut shell (17.25 MJ kg−1), it can SEM, are shown in Fig. 2. It is noted in Fig. 2a, b the surface
be noted the potential of pistachio shell for energetic applica- and cross section of pistachio shell as a dense surface with few
tion (Liu et al. 2014; Varma and Mondal 2017). cracks; these characteristics were also observed in other stud-
The volatile matter represents the fraction of gases released ies (Yang and Lua 2003; Lua and Yang 2004).
in the first stage of oxidative decomposition. The first stage is
characterized by the pyrolysis process of the biomass which Thermogravimetric analysis
has the release of combustible gases (CxHy, CO, and H2) and
non-combustible (CO2, SO2, NOx, and H2O). The combustible For the kinetic study, it is necessary to know the thermal be-
gases released in the first stage are burned when in contact with havior of the sample; thereby, Fig. 3 shows TG/DTG and DTA
the oxidizing atmosphere by means of exothermic reactions. curves for pistachio shell decomposition under oxidant and
 Author's personal copy
Environ Sci Pollut Res

Table 1 List of reaction models

used in solid-state reaction Model Mechanism g(α) f(α)
kinetics
P2 Power law α1/2 2α1/2
P3 Power law α1/3 3α2/3
P4 Power law α1/4 4α3/4
P2/3 Power law α3/2 2/3α−1/2
F1 First-order reaction − ln(1 − α) 1−α
R2 Contracting cylinder (contracting area) 1 − (1 − α)1/2 2(1 − α)1/2
R3 Contracting sphere (contracting volume) 1 − (1 − α)1/3 3(1 − α)2/3
D1 One-dimensional diffusion α2 1/2α−1
D2 Two-dimensional diffusion (1 − α) ln(1 − α) + α [− ln(1 − α)]−1
D3 Three-dimensional diffusion [1 − (1 − α)1/3]2 3/2(1 − α)2/3[1 − (1 − α)1/3]−1
A2 Avrami-Erofeev [− ln(1 − α)]1/2 2(1 − α)[− ln(1 − α)]1/2
A3 Avrami-Erofeev [− ln(1 − α)]1/3 3(1 − α)[− ln(1 − α)]2/3
A4 Avrami-Erofeev [− ln(1 − α)]1/4 4(1 − α)[− ln(1 − α)]3/4

inert atmospheres, which can be observed in different decom- onset of lignin also occurs in a slow rate (Brebu et al. 2013).
position stages. The second stage of decomposition occurs between the tem-
In Fig. 3a, it is possible to observe two well-defined de- peratures of 325 and 400 °C and it can be characterized by the
composition stages after the event of water volatilization un- majority decomposition of cellulose, with weight loss of 27%.
der inert atmosphere (pyrolysis). The first stage of decompo- Due to the intramolecular and intermolecular chemical inter-
sition occurs between 200 and 325 °C and can be character- action forces and high degree of polymerization, cellulose
ized by the major devolatilization of hemicellulose, which decomposition occurs at higher temperatures, when compared
presented a weight loss of 40%. Werner et al. (2014) evaluated to hemicellulose. The quick decomposition of cellulose is due
the thermal behavior under inert atmosphere of seven different to the depolymerization process which occurs the breakdown
hemicelluloses (β-glucan, arabinogalactan, arabinoxylan, of the glycosidic linkage with formation of levoglucosan be-
galactomannan, glucomannan, xyloglucan, and xylan) and tween temperatures of 300 and 390 °C (Collard and Blin
noticed that the majority of the polymers show decomposition 2014). According to Brebu et al. (2013), in this stage occurs
onset at low temperatures when compared to cellulose, also an increase in the decomposition rate of lignin which results in
because it presents a chemical structure with lower molecular the highest generation of volatile compounds. Lignin is a com-
weight and lower degree of polymerization (Collard and Blin plex structure formed of phenylpropane units, justifying the
2014; Braga et al. 2015). In this first stage, the decomposition difficulty to break down chemical bonds; thus, lignin is de-
graded under the whole experimental range and it is the main
responsible for the formation of the solid residue at the end of
Table 2 Physicochemic-
al characterization of Proximate analysis (wt%) the pyrolysis process. The residual solid of the pyrolysis pro-
pistachio shell Moisturea 3.59 ± 0.21 cess is called biochar; this carbonaceous solid residue has a
Volatile matterb 82.97 ± 1.82 high heating value.
Fixed carbonb,c 16.67 ± 1.46 In oxidant atmosphere, two main zones can be identified,
Ashb 0.36 ± 0.15 however in different temperature ranges when compared to
Ultimate analysis (wt%) thermal decomposition under inert atmosphere. The first zone
Carbonb 44.89 between the temperatures of 200 and 400 °C is represented by
Hydrogenb 5.66 the oxidative degradation of three main compounds (cellulose,
Nitrogenb 0.53 hemicellulose, and lignin). Unlike the inert atmosphere, in this
Sulfurb 0.97 region, there are two near peaks which represent the overlap of
Oxygenb,c 47.95 the curves of hemicellulose, cellulose, and lignin decomposi-
Heating value (MJ/kg) tion. The presence of a single peak is due to the oxidation of
HHV 17.47 the combustible gases that burn near the surface. This exother-
mic reaction causes the increase of the temperature of the
a
Wet basis samples to one above that programmed in the thermogravi-
b
Dry basis metric analyzer (Branca and Di Blasi 2013); this can be seen
c
Calculated by difference in Fig. 3. The second region between 400 and 520 °C is
 Author's personal copy
Environ Sci Pollut Res

Fig. 1 Fourier transform infrared

spectroscopy spectrum of
pistachio shell

characterized by the oxidation of the solid residue formed by of E a and presented a correlation coefficient of
pyrolysis. The solid residue oxidation occurs quickly, show- R2 > 0.9899. Moreover, in Fig. 4, it can be seen that the
ing a high reactivity of the solid at higher temperatures under fitted lines are nearly parallel, indicating a small variation
oxidant atmosphere. of Ea values in function of conversion; in addition, there is
a possibility that the reaction occurs in a single-step mech-
Kinetic study anism (Vyazovkin 2015). The oxidation reactions occur
through parallel and complex reactions; however, Fig. 4
The isoconversional methods were applied to a single-stage shows that the overall kinetic of decomposition can be
reaction; however, it was noted that the oxidation process considered as a single stage. This behavior of Ea values
occurs through two main regions of weight loss (Fig. 3b), is noted in Fig. 5. Ea for the two regions presented a small
which results in two different kinetic stages. In this way, the decline with conversion increase, presenting values be-
evaluation of the kinetic parameters of the combustion of tween 69.7 and 103.6 kJ mol−1 and 19.8 and
the pistachio shell was performed using two different re- 57.4 kJ mol−1 for zone 1 and zone 2, respectively. The
gions of weight loss. The first region (zone 1) was associ- decline of the activation energy value can be associated
ated to the biomass devolatilization (200–400 °C) and the with exothermic reactions of the volatile matter near the
second region (zone 2) represents the oxidation of the solid particle surface. These exothermic reactions cause the in-
residue (biochar) (400–640 °C). The activation energy was crease of the temperature above from the one programmed
evaluated by the Coats-Redfern modified isoconversional in particle, making the decomposition process of lignocel-
method (Eq. 4) in the conversion range between lulose quicker (Sullivan and Ball 2012). The low Ea value
0.05 ≤ α ≤ 0.95 from the thermogravimetric data as shown obtained for the combustion region of the biochar (zone 2)
in Fig. 4. can be associated with the exothermic reactions of the
The activation energy values obtained through ln(β/T2) solid produced by the initial stage of decomposition as
vs. 1000/T from the modified Coats-Redfern method pre- observed in Fig. 3b. According to Vyazovkin (2015), the
sented a satisfactory applicability in obtaining the values reduction of the energy required for conversion of a

Fig. 2 Scanning electron

micrographs at a magnification
×1000 (a) of pistachio shell
surface and magnification ×3000
(b) of cross section
 Author's personal copy
Environ Sci Pollut Res

Fig. 3 a TG/DTG and b DTA

curves from (black line) pyrolysis
and (red line) combustion process
at 20 °C min−1

material is associated with a high exothermicity, which adequate to obtain other parameters as pre-exponential fac-
favors the breaking of the carbon-carbon bonds of the tor and reaction model (Vyazovkin 2015). The evaluation of
biochar. The reduction of Ea value in the combustion of the pre-exponential factor and reaction model by compen-
the biochar was also observed by López-González et al. sation effect and master plot, respectively, is presented in
(2015), which associated with the possible catalytic reac- Fig. 6.
tion of the inorganic compounds present in the ashes. The linear fitting, in Fig. 6a, generated by compensation
The average value of activation energy (Ea,av) for each effect (Eq. 5) for each zone presented an excellent result with
zone was 85.11 and 37.47 kJ mol−1 for zone 1 and zone 2, regression coefficients (R2) of 0.9978 and 0.9946 for zone 1
respectively. In a thermochemical process, the activation and zone 2, respectively. The compensation parameters ob-
energy represents the main parameter to be evaluated from tained were a1 = −1.576 and b1 = 0.2027 for zone 1, while
the thermogravimetric analysis; this is because the solid-gas for zone 2, the values obtained were a2 = −2.2271 and
conversion of the biomass requires heat for the breakdown b2 = 0.1633. The average values of ln(A) obtained through
of the chemical bonds of the lignocellulose in combustible the compensation effect for zones 1 and 2 are 15.67 and
gas; in other words, the activation energy may provide the 4.05, respectively, which are equal to 6.39 × 10 6 and
information of energy needed for to start the chemical de- 57.14 min−1.
composition reactions (Yao et al. 2008; Vyazovkin 2015). From the results of activation energy and the pre-
Based on these conditions, comparing Ea,av values of other exponential factor, it became possible, through the master
biomass, under the same atmosphere conditions, the pista- plot, to estimate the reaction model of each zone as shown
chio shell presents a lower thermal stability under air atmo- in Fig. 6b. It is noted that the oxidation process of pistachio
sphere, requiring less energy for the breakdown of the shell behaves as a reaction controlled by the phase boundary,
chemical bonds of lignocellulose (Magdziarz and Wilk where in this process, the degradation rate is controlled by the
2013; Kocabaş-Ataklı et al. 2015). resulting reaction interface progress toward the center of the
The isoconversional methods are useful to obtain the val- particle (Khawam and Flanagan 2006). The reaction models
ue of the activation energy; however, these methods are not R3 and R2 were the more adequate models for zone 1 and

Fig. 4 Linear fit of ln(β/T2) vs.

1000/T for a zone 1 and b zone 2
 Author's personal copy
Environ Sci Pollut Res

Comparing the experimental and simulated results in

Fig. 7, it is noted that when the heating rate decreases, in
experimental analysis, the DTG peak is more evidenced,
where it is not observed in the simulated curve. This may be
associated with multiple reaction mechanisms occurring with
the reduction of the heating rate, because in Eq. 7, it was
considered a unified reaction model for R2 reaction. In addi-
tion, the negligence of the partial pressure of O2 in Eq. 2 may
have influenced to the difference between simulated and ex-
perimental results, mainly in the second region (Labaki and
Jeguirim 2016). However, the simulated results from Eq. 7
show a satisfactory approximation, although the use of aver-
age values of Ea and A. In this way, the results presented in
Fig. 7 demonstrated the possibility to perform a graphical
Fig. 5 Activation energy results calculated by modified Coats-Redfern modeling with a simple numerical method through the kinetic
method data parameters obtained initially by isoconversional
methods.
zone 2, respectively. Note that the reaction model obtained is
not controlled by diffusion; this is due to the quick volatiliza-
tion and small amount of mass and particle size used for the Conclusion
study.
From the kinetic parameters (Ea,av, Aav, and f(α)) calculated The application of the solid waste from the pistachio crop
for oxidation process of the pistachio shell samples, it was (pistachio shell) can represent a favorable energy resource
possible to obtain Eq. 7 which describes this process. in thermochemical processes such as combustion. Its ap-
plicability is mainly due to its low content of moisture and
8
dα1 6:39  106 −85110=RT h i ash, which favors a high HHV despite the high oxygen/
>
> 2

dα < ω 1 ¼ e 3 ð 1−αÞ 3 carbon ratio. SEM analysis showed that the surface of the
dT β ð7Þ
dα2 57:14 −37470=RT h i pistachio shell shows the dense surface with few cracks,
dT >
> 1
: ω2 ¼ e 2ð1−αÞ 2 while the infrared spectrum (FTIR) showed the functional
dT β
groups present in the lignocellulose. From the thermogra-
vimetric analyses under inert and oxidant atmospheres, it
In Eq. 7, the parameter ω represents the weight loss per- was possible to identify the main regions of weight loss,
centage of each sample region on a dry basis. Thus, using a which are associated with the devolatilization of lignocel-
numerical method such as Runge-Kutta 4th-order (RK4), the lulose and the formation of biochar. The presence of well-
simulated curves can be obtained and, consequently, com- defined two regions was observed when under inert atmo-
pared with the experimental analyses. The simulated DTG sphere, where the first one appears between the tempera-
curves are shown in Fig. 7. ture range of 200–350 °C and the second one between the

Fig. 6 Graph result for a

compensation effect and b master
plots of theoretical g(α)/g(0.5) vs.
α for each zone
 Author's personal copy
Environ Sci Pollut Res

Fig. 7 Comparison of
experimental and simulated
curves for the pistachio shell
samples for the heating rate of 20,
30, and 40 °C min−1

temperatures of 325 and 400 °C. These regions were as- isoconversional method data, presenting a satisfactory ap-
sociated with the major decomposition of hemicellulose proximation by means of simple numerical methods.
and cellulose, respectively. In the oxidant atmosphere,
the presence of two regions of weight loss was noticed.
The first region between temperatures of 200 and 400 °C Acknowledgements The authors gratefully acknowledge financial
was associated with overlapping of the decomposition of support from the National Council for Scientific and Technological
Development (CNPq) and the Coordination for the Improvement of
the lignocellulose components, while the second region Higher Education Personnel (CAPES).
between the temperature range of 400–520 °C was asso-
ciated with oxidation of the residual solid (biochar).
From the knowledge of the thermal behavior obtained
from the TG/DTG, it became possible to obtain the kinet- References
ic parameters for each stage (zone 1 and zone 2). The
methods used to estimate the kinetic parameters (Ea, A, Açıkalın K, Karaca F, Bolat E (2012) Pyrolysis of pistachio shell: effects
of pyrolysis conditions and analysis of products. Fuel 95:169–177.
and f(α)) present satisfactory results, which was observed doi:10.1016/j.fuel.2011.09.037
a good proximity to experimental results. Besides, the Anca-Couce A, Obernberger I (2016) Application of a detailed biomass
method used allowed to obtain a better understanding of pyrolysis kinetic scheme to hardwood and softwood torrefaction.
the combustion process of pistachio shell. The kinetic Fuel 167:158–167. doi:10.1016/j.fuel.2015.11.062
parameters for zone 1 were E a = 84.11 kJ mol − 1 , Asadieraghi M, Wan Daud WMA (2014) Characterization of lignocellu-
losic biomass thermal degradation and physiochemical structure:
A = 6.39 × 106 min−1, and f(α) = 3(1 − α)2/3, while for effects of demineralization by diverse acid solutions. Energy
zone 2, the kinetic parameters were Ea = 37.47 kJ mol−1, Convers Manag 82:71–82. doi:10.1016/j.enconman.2014.03.007
A = 57.14 min−1, and f(α) = 2(1 − α)1/2. The low Ea value ASTM (2003) Standard test method for compositional analysis by
observed in zone 2 can be associated with the highly thermogravimetry. ASTM International, West Conshohocken, pp
1–5
exothermic oxidation reaction in this stage, which reduces
Braga RM, Costa TR, Freitas JCO et al (2014) Pyrolysis kinetics of
the activation energy. From the calculated data, it became elephant grass pretreated biomasses. J Therm Anal Calorim 117:
possible to simulate the DTG curves and to compare with 1341–1348. doi:10.1007/s10973-014-3884-2
the experimental results, which was possible to obtain a Braga RM, Queiroga TS, Calixto GQ et al (2015) The energetic charac-
satisfactory adjustment of the results. terization of pineapple crown leaves. Environ Sci Pollut Res 22:
18987–18993. doi:10.1007/s11356-015-5082-6
In conclusion, this piece of work demonstrated the appli- Branca C, Di Blasi C (2013) A unified mechanism of the combustion
cability of the pistachio shell in combustion processes and the reactions of lignocellulosic fuels. Thermochim Acta 565:58–64. doi:
possibility to simulate decomposition curves through 10.1016/j.tca.2013.04.014
 Author's personal copy
Environ Sci Pollut Res

Brebu M, Tamminen T, Spiridon I (2013) Thermal degradation of various corresponding hydrochars. Appl Energy 113:1315–1322. doi:10.
lignins by TG-MS/FTIR and Py-GC-MS. J Anal Appl Pyrolysis 1016/j.apenergy.2013.08.087
104:531–539. doi:10.1016/j.jaap.2013.05.016 López-González D, Avalos-Ramirez A, Giroir-Fendler A et al (2015)
Brown ME, Maciejewski M, Vyazovkin S et al (2000) Computational Combustion kinetic study of woody and herbaceous crops by ther-
aspects of kinetic analysis: part A: the ICTAC kinetics project-data, mal analysis coupled to mass spectrometry. Energy 90:1626–1635.
methods and results. Thermochim Acta 355:125–143 doi:10.1016/j.energy.2015.06.134
Channiwala SA, Parikh PP (2002) A unified correlation for estimating Lua AC, Yang T (2004) Effects of vacuum pyrolysis conditions on the
HHVof solid, liquid and gaseous fuels. Fuel 81:1051–1063. doi:10. characteristics of activated carbons derived from pistachio-nut
1016/S0016-2361(01)00131-4 shells. J Colloid Interface Sci 276:364–372. doi:10.1016/j.jcis.
Collard F-X, Blin J (2014) A review on pyrolysis of biomass constituents: 2004.03.071
mechanisms and composition of the products obtained from the Lua AC, Yang T (2009) Theoretical and experimental SO2 adsorption
conversion of cellulose, hemicelluloses and lignin. Renew Sust onto pistachio-nut-shell activated carbon for a fixed-bed column.
Energ Rev 38:594–608. doi:10.1016/j.rser.2014.06.013 Chem Eng J 155:175–183. doi:10.1016/j.cej.2009.07.031
FAO (2015) FAO statistical pocketbook: world food and agriculture Magdziarz A, Wilk M (2013) Thermal characteristics of the combustion
2015. FAO, Rome process of biomass and sewage sludge. J Therm Anal Calorim 114:
Ferreira-Leitão V, Gottschalk LMF, Ferrara MA et al (2010) Biomass 519–529. doi:10.1007/s10973-012-2933-y
residues in Brazil: availability and potential uses. Waste and Nanda S, Mohammad J, Reddy SN et al (2014) Pathways of lignocellu-
Biomass Valorization 1:65–76. doi:10.1007/s12649-010-9008-8 losic biomass conversion to renewable fuels. Biomass Convers
García R, Pizarro C, Lavín AG, Bueno JL (2012) Characterization of Biorefinery 4:157–191. doi:10.1007/s13399-013-0097-z
Spanish biomass wastes for energy use. Bioresour Technol 103: Senum GI, Yang RT (1977) Rational approximations of the integral of the
249–258. doi:10.1016/j.biortech.2011.10.004 Arrhenius function. J Therm Anal 11:445–447
Gotor FJ, Criado JM, Malek J, Koga N (2000) Kinetic analysis of solid-
Sullivan AL, Ball R (2012) Thermal decomposition and combustion
state reactions: the universality of master plots for analyzing isother-
chemistry of cellulosic biomass. Atmos Environ 47:133–141. doi:
mal and nonisothermal experiments. J Phys Chem A 104:10777–
10.1016/j.atmosenv.2011.11.022
10782. doi:10.1021/jp0022205
Taghizadeh-Alisaraei A, Assar HA, Ghobadian B, Motevali A (2017)
Huber MT (2009) Energizing historical materialism: fossil fuels, space
Potential of biofuel production from pistachio waste in Iran.
and the capitalist mode of production. Geoforum 40:105–115. doi:
Renew Sust Energ Rev 72:510–522. doi:10.1016/j.rser.2017.01.111
10.1016/j.geoforum.2008.08.004
Jenkins BM, Baxter LL, Miles TR, Miles TR (1998) Combustion prop- Varma AK, Mondal P (2017) Pyrolysis of sugarcane bagasse in semi
erties of biomass. Fuel Process Technol 54:17–46. doi:10.1016/ batch reactor: effects of process parameters on product yields and
S0378-3820(97)00059-3 characterization of products. Ind Crop Prod 95:704–717. doi:10.
Kan T, Strezov V, Evans TJ (2016) Lignocellulosic biomass pyrolysis: a 1016/j.indcrop.2016.11.039
review of product properties and effects of pyrolysis parameters. Vyazovkin S (2015) Isoconversional kinetics of thermally stimulated pro-
Renew Sust Energ Rev 57:1126–1140. doi:10.1016/j.rser.2015.12. cesses, 1st edn. Springer International Publishing, Cham
185 Vyazovkin S, Burnham AK, Criado JM et al (2011) Thermochimica Acta
Khawam A, Flanagan DR (2006) Solid-state kinetic models: basics and ICTAC Kinetics Committee recommendations for performing kinet-
mathematical fundamentals. J Phys Chem B 110:17315–17328. doi: ic computations on thermal analysis data. Thermochim Acta 520:1–
10.1021/jp062746a 19. doi:10.1016/j.tca.2011.03.034
Kocabaş-Ataklı ZÖ, Okyay-Öner F, Yürüm Y (2015) Combustion char- Werner K, Pommer L, Broström M (2014) Thermal decomposition of
acteristics of Turkish hazelnut shell biomass, lignite coal, and their hemicelluloses. J Anal Appl Pyrolysis 110:130–137. doi:10.1016/
respective blends via thermogravimetric analysis. J Therm Anal j.jaap.2014.08.013
Calorim 119:1723–1729. doi:10.1007/s10973-014-4348-4 Xu F, Yu J, Tesso T et al (2013) Qualitative and quantitative analysis of
Komnitsas K, Zaharaki D, Pyliotis I et al (2015) Assessment of pistachio lignocellulosic biomass using infrared techniques: a mini-review.
shell biochar quality and its potential for adsorption of heavy metals. Appl Energy 104:801–809. doi:10.1016/j.apenergy.2012.12.019
Waste and Biomass Valorization 6:805–816. doi:10.1007/s12649- Yang T, Lua AC (2003) Characteristics of activated carbons prepared
015-9364-5 from pistachio-nut shells by physical activation. J Colloid Interface
Labaki M, Jeguirim M (2016) Thermochemical conversion of waste Sci 267:408–417. doi:10.1016/S0021-9797(03)00689-1
tyres—a review. Environ Sci Pollut Res. doi:10.1007/s11356-016- Yao F, Wu Q, Lei Y et al (2008) Thermal decomposition kinetics of
7780-0 natural fibers: activation energy with dynamic thermogravimetric
Liu Z, Quek A, Balasubramanian R (2014) Preparation and characteriza- analysis. Polym Degrad Stab 93:90–98. doi:10.1016/j.
tion of fuel pellets from woody biomass, agro-residues and their polymdegradstab.2007.10.012