Biochar potential evaluation of palm oil wastes through slow pyrolysis: Thermochemical characterization and pyrolytic kinetic studies

Bioresource Technology 236 (2017) 155–163

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Biochar potential evaluation of palm oil wastes through slow pyrolysis: Thermochemical characterization and pyrolytic kinetic studies Xin Jiat Lee a, Lai Yee Lee a,⇑, Suyin Gan a, Suchithra Thangalazhy-Gopakumar a, Hoon Kiat Ng b a b

Department of Chemical and Environmental Engineering, The University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor, Malaysia Department of Mechanical, Materials and Manufacturing Engineering, The University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor, Malaysia

h i g h l i g h t s
PKS-char and EFB-char have high carbon contents of 59.92 and 53.78 wt%. 1


HHV of PKS-char (27.50 MJ kg

) and EFB-char (26.18 MJ kg1) comparable with coal.


High yields of PKS-char (37.07 wt%) and EFB-char (35.14 wt%) achieved by pyrolysis.
Multi-step pyrolysis kinetics demonstrated by isoconversional KAS and FWO methods.

a r t i c l e

i n f o

Article history: Received 18 January 2017 Received in revised form 16 March 2017 Accepted 17 March 2017 Available online 22 March 2017 Keywords: Biochar Empty fruit bunch Palm kernel shell Palm oil sludge Slow pyrolysis

a b s t r a c t This research investigated the potential of palm kernel shell (PKS), empty fruit bunch (EFB) and palm oil sludge (POS), abundantly available agricultural wastes, as feedstock for biochar production by slow pyrolysis (50 mL min1 N2 at 500 °C). Various characterization tests were performed to establish the thermochemical properties of the feedstocks and obtained biochars. PKS and EFB had higher lignin, volatiles, carbon and HHV, and lower ash than POS. The thermochemical conversion had enhanced the biofuel quality of PKS-char and EFB-char exhibiting increased HHV (26.18–27.50 MJ kg1) and fixed carbon (53.78–59.92%), and decreased moisture (1.03–2.26%). The kinetics of pyrolysis were evaluated by thermogravimetry at different heating rates (10–40 °C). The activation energies determined by KissingerAkahira-Sunose and Flynn-Wall-Ozawa models were similar, and comparable with literature data. The findings implied that PKS and EFB are very promising sources for biochars synthesis, and the obtained chars possessed significant biofuel potential. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction The increase in fossil fuel usage due to rapid industrialisation and population growth has resulted in detrimental impacts on the environment such as air pollution, global warming and climate change. This, in addition to diminishing fossil fuel reserves, poses a serious concern worldwide for governments and researchers alike. In response to this concern, many researchers have focussed on developing alternative routes for sustainable and clean energy production, especially biofuel from lignocellulosic wastes (Jeguirim et al., 2014a). Biofuel from lignocellulosic materials is considered renewable, abundant and low cost since the feedstock is readily available worldwide in the forms of agricultural waste, forestry refuse, industrial biomass by-products, municipal sewage sludge, aquatic ⇑ Corresponding author. E-mail address: [email protected] (L.Y. Lee). http://dx.doi.org/10.1016/j.biortech.2017.03.105 0960-8524/Ó 2017 Elsevier Ltd. All rights reserved.

algae and plants (Chen et al., 2015; Jeguirim et al., 2014b; Liang et al., 2008). These solid wastes can be transformed into various biofuels such as syngas, bio-oil and biochar through gasification and pyrolysis (Jeguirim et al., 2014b). Comparing with standard fossil fuels, lignocellulosic wastes have lower nitrogen and sulphur composition hence harvesting biofuel from them would generate lesser toxic gases including nitrogen oxides and sulphur dioxide. Furthermore, production of biofuel would lead to comparatively lower carbon footprint since lignocellulosic wastes are part of nature’s carbon cycle. Thus, the production of bioenergy from lignocellulosic wastes is more environmentally friendly (Lee et al., 2013). Oil palm (Elaeis guineensis) is grown extensively in Malaysia with a total planted area of 5.74 million hectares in 2016. The production of edible oil from the palm fruits has reached 17.32 million tonnes per year (MPOB, 2016a). Apart from palm oil production, the industry also generates considerable amount of agricultural wastes such as palm kernel shell (PKS), empty fruit bunch (EFB)

156

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

and palm oil sludge (POS). PKS and EFB are solid residues from fresh fruit bunch (FFB) after oil extraction, accounting for 4.5 and 22 wt% per tonne of FFB, respectively (Garcia-Nunez et al., 2016). Based on the FFB yield of 15.91 tonnes per hectare in 2016, the quantities of PKS and EFB generated were significant (MPOB, 2016a). Palm oil sludge (POS) is the semi-solid residue resulting from the treatment of palm oil mill effluent (POME) by acidification, anaerobic and aerobic methods. POS is generated in significant quantity in mills located throughout Malaysia due to the large amount of POME (800 dm3 POME per tonne of FFB) resulting from high consumption of water in oil extraction and cleaning processes (Garcia-Nunez et al., 2016). The disposal of untreated POS, PKS and EFB could lead to adverse consequences on the environment. It is thus necessary to control the wastes discharge from palm oil industry. This research involved re-utilizing the three major palm oil mill residues as feedstock for biochar thereby providing an attractive option for managing the wastes and imparting economic value. Biochar is a carbon rich and porous solid, often produced by slow pyrolysis of waste biomass without or with partial presence of oxygen (Creamer et al., 2014; Islam et al., 2016). The pyrolytic reaction releases moisture and volatiles in the biomass, leaving behind a porous structure whilst retaining the aromatic compounds and chemical functional groups (Tan et al., 2017). These desirable attributes enable biochar to be used in various applications such as soil amendment, energy production and pollution control. Biochar when added to soil can improve its fertility by enhancing nutrients and water retention (Beesley et al., 2011; Wang et al., 2014; Zhang et al., 2013). It has also been utilized as an effective pollutant adsorbent in environmental remediation (Creamer et al., 2014; Hodgson et al., 2016). A number of agricultural wastes such as pelletized grape vine and sunflower husks (Colantoni et al., 2016), paper mill sludge (Devi and Saroha, 2015), olive solid waste, date palm trunks, pine sawdust, Posidonia oceanica balls (Jeguirim et al., 2014b), rice husk and elm sawdust (Wang et al., 2014) have been pyrolyzed to biochars indicating the process versatility with regard to feedstock type. The objectives of this research were to investigate the potential of PKS, EFB and POS as the precursor for biochar, and to evaluate the biofuel potential of the obtained chars. To date, a number of investigations have been reported on pyrolysis of palm oil residues by thermogravimetric (TG) analysis whereby the obtained data were analyzed using different kinetic models and/or model-free isoconversional approaches. For instance, Jeguirim et al. (2014a) examined the thermochemical conversion of PKS and palm mesocarp fibre (PMF) under nitrogen (N2) atmosphere. PMF was determined to be the most suitable feedstock exhibiting the highest heating values. The pyrolysis kinetic was represented by devolatilization followed by char formation. Nyakuma et al. (2015) who assessed the potential of palm EFB as the pyrolysis feedstock reported that the pyrolytic kinetics were well represented by a model-free isoconversional method. Luangkiattikhun et al. (2008) studied the non-isothermal decomposition of palm shell, fibre and kernel, and determined that the experimental data were best fitted with the two step-parallel reactions model. Other similar works on palm oil residues reported in the literature include that by Yang et al. (2004) and Ma et al. (2015). This research aims to compare the pyrolytic behaviour of PKS, EFB and POS sourced from the same palm oil mill. There are 453 FFB mills spread across Malaysia, and each mill operates with different parameters to process FFB from various plantations hence generating by-products of varying properties (MPOB, 2016b). In this work, the rationale behind the approach of using agricultural residues acquired from the same mill is to eliminate the possible influence of variation in milling and plantation conditions on the research findings (Luangkiattikhun et al., 2008; Ma et al., 2015;

Yang et al., 2004). The physicochemical properties of the three biomasses and derived biochars were determined by Fourier transform infrared spectroscopy, bomb calorimetry, scanning electron microscopy, energy dispersive X-ray and proximate analysis. The pyrolysis kinetics were evaluated by TG analyzer, and the data obtained were correlated with the Kissinger-Akahira-Sunose and Flynn-Wall-Ozawa models. 2. Materials and methods 2.1. Preparation of biochars PKS, EFB and POS were collected from Seri Ulu Langat Palm Oil Mill Sdn. Bhd., Dengkil, Selangor, Malaysia. The waste materials were washed repeatedly with distilled water to remove impurities and surface oil, and dried in an oven (Memmert, Germany) at 80 °C, for 72 h. The dried biomass were pulverized in a miller (Retsch SM 100, Germany) and then sieved to 0.5–2 mm particles using a vibratory sieve shaker (Retsch AS 200, Germany). The biomass particles was then pyrolyzed in a stainless steel bed reactor with internal diameter 0.04 m and length 1.22 m, placed in a horizontal tubular furnace (Carbolite CTF12, UK). The sample was heated to 500 °C at 10 °C min1 under a N2 flow of 50 mL min1. It was held up at this temperature for 1 h until complete degradation of hemicellulose and cellulose was achieved (Lee et al., 2013). The pyrolyzed products were referred to as PKS-char, EFB-char and POSchar. The char yield was determined based on the sample weight difference before and after pyrolysis. 2.2. Scanning electron microscopy The morphological and elemental characteristics of PKS, EFB, POS and their corresponding biochars were examined by scanning electron microscope (SEM, Quanta 400F, USA) equipped with an energy dispersive X-ray (EDX) spectrometer and X-max detector (Oxford-Instruments INCA 400, UK). The sample was mounted on the SEM stub using double-side carbon tape and inserted into the chamber operated at approximately 1.4  103 Pa, between 10– 20 kV accelerating voltages and 2500–10,000 times magnification. 2.3. Lignocellulose content determination The extractives, hemicellulose, cellulose and lignin contents of the raw biomass were evaluated based on procedures described by Ayeni et al. (2013) and Tan et al. (2011). The extractives content was evaluated using a Soxhlet extractor in which 1 g of biomass was refluxed with 60 mL acetone at 60 °C for 6 h. The sample was then dried in the oven at 110 °C until a constant weight was achieved. The difference in the initial sample weight and final dried sample weight represented the extractives weight. The hemicellulose content was determined by sodium hydroxide (NaOH) treatment. Approximately 1 g of extractive-free biomass was poured into an Erlenmeyer flask containing 150 mL NaOH of concentration 0.5 mol L1. The mixture was agitated in a waterbath shaker (Protech, Malaysia) at 150 rev min1 and 80 °C for 3.5 h. The contents were filtered through a filter paper (Sartorius, Grade 293) and rinsed with excess distilled water until the pH of rinsing approached 7. The residue was then dried at 110 °C until its weight was constant. The sample weight difference before and after this treatment was taken as the hemicellulose weight. The lignin content of biomass was determined by a two-step acid hydrolysis process. Approximately 300 mg of extractive-free biomass was treated with 3 mL sulphuric acid of concentration 14 mol L1. The mixture was agitated in the waterbath shaker at 150 rev min1 and 30 °C for 2 h. Thereafter, 84 mL of distilled

157

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

water was added to the mixture to dilute the acid. The mixture was then autoclaved (Labtech, UK) at 121 °C (2 bar)1 for 1 h. Upon weak acid hydrolysis, the hydrolyzed contents were cooled, filtered and washed with distilled water until the pH of solution was neutral. The solid residue was dried at 110 °C to a constant weight representing the lignin weight. By assuming the extractive-free biomass consists of hemicellulose, cellulose and lignin, the cellulose content can be determined as follows:

%W cellulose ¼ 100  ð%W hemicellulose þ %W lignin Þ

ð1Þ

where %W is the weight percentage. 2.4. Proximate analysis The proximate analysis including moisture content, volatile matter, fixed carbon and ash content, was evaluated by a TG analyzer (Mettler Toledo, USA). The analysis was programmed using STARe (Mettler Toledo, Version 13.00, USA) based on ASTM D7582-15 method (ASTM International, 2015). 15 mg of sample was heated at 5 °C min1, from room temperature to 110 °C under N2 purge (50 mL min1) and held up for 10 min. The sample was next heated to 800 °C and kept for 7 min. Thereafter, oxygen (O2) (50 mL min1) was supplied to the TG chamber for oxidation. The sample was finally heated up to 900 °C and retained for 30 min. The weight of sample was recorded until a constant value was achieved. The ash content was determined by deducting the total of moisture, volatile matter and fixed carbon from 100%. 2.5. Calorimetry analysis The higher heating value (HHV) of samples was determined by a bomb calorimeter (Parr 6100, USA) according to ASTM D5865-07 method (ASTM International, 2013). A pellet of 1 g created by a press (Parr 2811, USA) was placed in the combustion vessel which was next filled with O2 at 30 bar. The vessel was immersed in 2 L of water contained in a bucket. Upon combustion, the temperature rise of the water was recorded for determination of HHV. 2.6. Fourier transform infrared analysis The surface chemical properties of PKS, EFB, POS and biochars were determined by Fourier transform infrared (FTIR) spectroscopy (Perkin Elmer, USA). The sample was mixed with potassium bromide (KBr, Sigma-Aldrich, IR grade) which had been dried at 110 °C for 24 h, at a ratio of 1 mg sample to 10 mg KBr. The mixture was ground to ensure homogeneity and then pressed into a thin disc. The FTIR spectrum of the disc was recorded at wavenumber within the range of 400–4000 cm1 with a resolution of 1 cm1.


 da Ea ¼ kðTÞf ðaÞ ¼ A exp  f ðaÞ dt RT

ð2Þ

where da/dt (mg mg1 min1) is the conversion rate, t (min) is the reaction time, f(a) is the differential reaction model, k(T) is the temperature-dependent rate constant, Ea (kJ mol1) is the apparent activation energy, A (min1) is the pre-exponential factor, R (J mol1 K1) is the universal gas constant and T (K) is the absolute temperature. The degree of conversion, a (mg mg1), is defined as follows:

a¼

m0  mt m0  mf

ð3Þ

where m0 (mg), mt (mg), and mf (mg) are the initial mass, mass at time t and final mass of sample, respectively. For non-isothermal pyrolysis, the constant heating rate applied to the process is bK (K min1) = dT/dt. Hence Eq. (2) can be written as follows:


 da A Ea dT exp ¼ f ðaÞ bK RT

ð4Þ

Upon integration, Eq. (4) becomes:

gðaÞ ¼

Z

a

0

Z

da A ¼ f ðaÞ bK

Ta

exp 0


 Ea AEa dT ¼ PðxÞ RT bK R

ð5Þ

where g(a) is the integral form of conversion rate, Ta (K) is the temperature at conversion, P(x) is the temperature integral without an exact analytical solution and x is Ea/RT. Eq. (5) can be solved by using approximation methods such as Kissinger-Akahira-Sunose (KAS) and Flynn-Wall-Ozawa (FWO). The models considered that the reaction rate is a function of temperature only for fixed conversion degree. The KAS and FWO are isoconversional models used commonly to determine the activation energy for a given conversion without specification of reaction mechanism (Ceylan and Topcu, 2014; Damartzis et al., 2011). The KAS model is based on the following empirical approximation (Kissinger, 1957):

log PðxÞ ffi

ex ; for 20 6 x 6 50 x2

ð6Þ

By substituting into Eq. (5) and rearranging, the KAS model as shown in Table 1 was obtained. According to this model, the activation energy can be determined from the slope of the linear plot of ln (bK/T2) against 1/T at progressing conversion. The FWO model utilizes the Doyle’s approach for heterogeneous reaction (Doyle, 1961), given by:

log PðxÞ ffi 2:315  0:4567x; for 20 6 x 6 60

ð7Þ

2.7. Pyrolysis kinetic evaluation The pyrolysis kinetic of PKS, EFB and POS was evaluated using the TG analyzer (Mettler Toledo, USA). 15 mg of sample was heated from 30 to 900 °C at 10 °C min1 in N2 atmosphere (50 cm3 min1). The maximum weight loss point was evaluated from the derivative of TG (DTG) curves. The procedures were repeated using heating rates of 20 and 40 °C min1 (Ma et al., 2015; Mu et al., 2015). All experiments were repeated at least three times to ensure data reproducibility and the average results are presented in this paper along with the standard errors. 2.8. Background of pyrolysis kinetics The reaction rate of thermochemical decomposition of biomass can be expressed by the following (Mu et al., 2015; White et al., 2011):

Table 1 Expressions for KAS and FWO models. Model KAS

FWO

Expression     AEa Ea ln bTK2 ¼ ln Rgð aÞ  RT

log bK ¼ log





AEa RgðaÞ

E  a  2:315  0:4567 RT

Plot

Reference

ln (bK/T2) versus 1/T Slope: Ea/R Intercept: ln[AEa/Rg (a)] log(bK) versus 1/T Slope: 0.4567 (Ea/R) Intercept: log[AEa/Rg (a)]

Kissinger (1957), Akahira and Sunose (1971)

Flynn and Wall (1966), Ozawa (1965), Doyle (1961)

158

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

The model obtained by substituting Eq. (7) into Eq. (5) is shown in Table 1. From the plot of log bK against 1/T, the activation energy can be evaluated.

contained typical lignocellulosic contents (Abnisa et al., 2013; Ma et al., 2015). The slight deviations might be attributed to the different biomass sources and methods of analysis.

3. Results and discussion

3.3. Proximate analysis

3.1. Surface morphology

The proximate analysis was evaluated according to ASTM D1762-84 standard procedure and the results are displayed in Table 2 in terms of moisture, volatile matter, fixed carbon and ash contents. The volatile matter in PKS (63.01%) and EFB (72.88%) were relatively high indicating that both materials consisted of high amount of thermally degradable compounds. This result implied that PKS and EFB were promising feedstocks for biochar formation (Ceylan and Topcu, 2014). Table 2 also shows that PKS (7.54%) and EFB (11.10%) have relatively low ash content. Generally, a feedstock with high ash content is undesirable as it can produce char with high ash content which lowers its heating value (Ceylan and Topcu, 2014). Compared to PKS and EFB, POS contained more ash and less volatile matter, making it the least favourable feedstock for biochar production. Additionally, the results show that the fixed carbon content of PKS and EFB increased after pyrolysis due to the decrease in volatile matter. The high fixed carbon content of PKS-char (59.92%) and EFB-char (53.78%) further supported PKS and EFB as promising feedstocks for biochar production. The moisture contents of the three raw biomass varied from 3.47 to 5.67% are well below than the range feasible for pyrolysis, i.e. <10% (Mehmood et al., 2017). PKS and EFB experienced a reduction in retained moisture upon pyrolysis. The ash content of the three biomasses was increased due probably to the mineral matter turning into ash during the thermal treatment (Wang et al., 2014). The ash content of POS-char was consistent with that reported by Thangalazhy-Gopakumar et al. (2015). The results of proximate analysis revealed the suitability of PKS and EFB for pyrolysis into biochars.

The SEM images for PKS, EFB, POS and the biochars were obtained at magnifications ranging from 2500–10,000 times. The surface of PKS contained protuberances and irregular patches, however after pyrolysis, some of the lumps were eliminated resulting in cavities. The SEM image of EFB consisted of circular structure of crystalline agglomerates. The crystalline agglomerates were silica structures commonly found in EFB. Similar observations have been reported by Shamsudin et al. (2012) and Bouraoui et al. (2015). Upon pyrolysis of EFB, some of the agglomerates were removed forming loop holes. Meanwhile, POS surface contained irregular pores which have not changed significantly upon pyrolysis. The honeycomb-like structure in POS and POS-char may be favourable active sites for certain applications such as sequestration of pollutants which is worthy of further investigation. 3.2. Lignocellulose content The results on lignocellulose contents of PKS, EFB and POS are listed in Table 2. It can be seen that the three agricultural wastes have different contents of hemicellulose, cellulose, lignin and extractives. EFB has the highest hemicellulose content (38.46%) whereas PKS has the highest lignin content (58.30%). The highest content of extractives (15.66%) was found in EFB which might be lipids, proteins, moisture, hydrocarbons and other minerals. Among the three biomasses, PKS not only had the highest lignin content, it also had the lowest hemicellulose which is favourable for biochar synthesis. Generally, cellulose and hemicellulose are strongly bound to lignin with hemicellulose acting as the linking agent. During pyrolytic reaction, hemicellulose would be decomposed first followed by cellulose and finally lignin. On comparison with literature data for PKS and EFB, the biomass assayed

3.4. Biochar yield The yields of char from PKS, EFB and POS are summarized in Table 2. The yield of POS was the highest (79.16%) followed by

Table 2 Physicochemical characteristics of PKS, EFB, POS and respective biochars. EFB

POS

PKS-char

EFB-char

POS-char

Lignocellulose analysis (wt%): Hemicellulose 14.20 ± 0.44 Cellulose 27.51 ± 0.39 Lignin 58.30 ± 0.5 Extractives 7.50 ± 0.05

PKS

38.46 ± 0.83 24.23 ± 1.28 37.32 ± 1.14 15.66 ± 0.07

34.70 ± 2.77 35.27 ± 3.03 30.04 ± 0.58 5.57 ± 2.25

– – – –

– – – –

– – – –

Proximate analysis (wt%): Moisture content Volatile matter Fixed carbon Ash content(a)

4.53 ± 0.60 63.01 ± 3.80 24.91 ± 2.89 7.54 ± 0.86

3.47 ± 0.57 72.88 ± 1.82 12.56 ± 1.89 11.10 ± 0.50

5.67 ± 0.07 28.16 ± 0.92 23.21 ± 2.49 42.95 ± 1.50

2.26 ± 0.50 30.26 ± 1.56 59.92 ± 0.23 7.56 ± 1.83

1.03 ± 0.21 27.46 ± 4.67 53.78 ± 2.50 17.73 ± 1.97

5.86 ± 0.17 24.07 ± 1.05 21.37 ± 0.69 48.70 ± 1.57

Elemental analysis (wt%): C O Mg Al Si P S Cl K Ca Fe

55.89 ± 0.08 41.15 ± 0.11 – 2.37 ± 0.18 0.61 ± 0.07 – – – – – –

50.89 ± 0.41 42.44 ± 0.28 – 0.20 ± 0.02 6.43 ± 0.16 – – – – – –

53.21 ± 1.92 33.02 ± 0.63 0.63 ± 0.12 2.58 ± 1.53 5.10 ± 2.24 0.69 ± 0.15 0.64 ± 0.24 0.46 ± 0.22 1.95 ± 0.67 0.89 ± 0.13 0.95 ± 0.44

73.11 ± 0.76 25.88 ± 0.71 – 0.43 ± 0.05 0.59 ± 0.1 – – – – – –

54.5 ± 0.14 32.94 ± 0.17 – 0.19 ± 0.02 12.13 ± 0.25 – – – – 0.26 ± 0.04 –

72.43 ± 1.06 20.91 ± 0.81 0.51 ± 0.04 0.78 ± 0.06 2.52 ± 0.22 0.46 ± 0.09 0.41 ± 0.1 – 0.88 ± 0.07 0.71 ± 0.06 0.42 ± 0.04

Yield (wt%): HHV (MJ kg1):

– 19.72 ± 0.15

– 19.65 ± 0.05

– 10.69 ± 0.56

37.07 ± 1.19 27.50 ± 0.16

35.14 ± 0.35 26.18 ± 0.05

79.16 ± 0.52 9.80 ± 0.48

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

159

PKS (37.07%) and finally, EFB (35.14%). This could be explained by the ratio of volatile matter to fixed carbon in the raw biomass. The calculated ratios for POS, PKS and EFB were 1.21, 2.53 and 5.80, respectively which were inversely proportional to the yield. The observed correlation between the biochar yields and volatile matter to fixed carbon ratios are in agreement with the literature (Lee et al., 2013; Liu et al., 2014). Despite its high yield, POS-char would be unsuitable as a biofuel due to its high ash content as discussed previously. The PKS-char (37.07%) and EFB-char (35.14%) yields compared well with the char yields of camel grass (30.46%) (Mehmood et al., 2017), cocopeat (38.7%) and paddy straw (41%) (Lee et al., 2013). The results of this study reflected again that PKS and EFB have high biochar potential.

This suggests that bond cleavage of alkane groups most probably in hemicellulose, cellulose and lignin had occurred (Leng et al., 2011). Additionally, the oxygen-containing functional groups (peaks at 1052, 1054, 1722 and 1726 cm1) in the palm oil wastes were removed after pyrolysis. This is in agreement with the elemental analysis results (Table 2) whereby O wt% of the raw biomass was reduced after conversion to biochars. Pyrolysis had also introduced aromatic compounds in PKS-char (884 and 1593 cm1) and carboxylic acids in EFB-char (1581 cm1) and POS-char (1579 cm1). Overall, the FTIR analysis demonstrated that pyrolysis of the three agricultural wastes was accompanied by structural changes in various chemical groups such as alkanes, alcohols, alky halides, carboxylic and aromatic compounds.

3.5. Heating value

3.8. Effect of heating rate on TG and DTG curves

The HHV of the prepared biochars are shown in Table 2. As can be seen, the HHV of PKS-char (27.50 MJ kg1) and EFB-char (26.18 MJ kg1) are higher than those of their respective raw forms. The obtained HHV are comparable with that of coal which typically ranged from 25 to 30 MJ kg1 (Ilyushechkin et al., 2014). The HHV for POS-char (9.80 MJ kg1) was the lowest likely due to its relatively high ash and low volatiles contents (Table 2). The results suggested that PKS and EFB could be used for biochar synthesis.

The thermal degradation of biomass involved sets of complex chemical reactions, and to evaluate their bioenergy potential, it is necessary to obtain the reaction kinetics. These can be achieved from TG curve which represents mass loss at increasing temperature and a fixed heating rate (Damartzis et al., 2011; El may et al., 2012). The effect of heating rate on the TG behaviours of PKS, EFB and POS was investigated at 10, 20 and 40 °C min1 in N2 atmosphere, and the results are depicted in Fig. 1(a)–(c). For all heating rates assayed, the mass loss profiles for the three palm oil wastes were identical, but skewed towards higher temperatures at higher rate. This trend could be caused by poorer heat transfer efficiency as the heating rate was increased. At high heating rates, heating of biomass might occur too rapidly causing poorer heat flow to the interior of the solid (Ceylan and Topcu, 2014). There were three distinct stages in the TG curves: 1) an initial mass reduction from ambient temperature to about 200 °C attributed to evaporation of moisture and light volatiles; 2) a significant mass loss from about 200 to 500 °C at which heavy organic compounds such as hemicellulose and cellulose were devolatilized; 3) a slow mass loss above 500 °C caused by degradation of lignin and other compounds with stronger chemical bonds. As displayed in Fig. 1(d)–(f), the DTG plots of PKS, EFB and POS contain three peaks labelled as I, II and III corresponding to the temperatures at which the maximum mass loss rate occurred. For PKS, the three peaks occurred at 120, 282 and 356 °C; for EFB, they existed at 66, 295 and 356 °C; and for POS they occurred at 56, 131 and 330 °C. The three peaks for PKS were shifted to 150, 300 and 375 °C; for EFB they were moved to 75, 320 and 375 °C; and for POS they were shifted to 95, 160 and 347 °C when the heating rate was increased to 40 °C min1. The temperatures for maximum decomposition rates were different for the three biomass as their hemicellulose, cellulose and lignin contents were different (Chen et al., 2015). The DTG curves also indicated that 500 °C was sufficient for thermal decomposition of the hemicellulose and cellulose in PKS, EFB and POS as there was no apparent peak beyond this temperature. The slow mass loss occurring between 500 and 900 °C corresponded to the final decomposition involving aromatization of lignin compound (Nyakuma et al., 2015).

3.6. Elemental contents The EDX analysis (Table 2) shows the presence of C, O, Al and Si in PKS, EFB and POS. Other elements such as Mg, P, S, Cl, K, Ca and Fe were also detected in POS which might be linked to its high ash content. The presence of these minerals in POS could be due to the chemicals usage during POME treatment in the palm oil mill. After pyrolysis, the charred products contained higher C wt%, but lower O wt%. The thermal process might have eliminated the oxygenated chemical groups in the biomass. The results indicated that pyrolysis had different impacts on the elemental compositions of the three palm oil wastes. Furthermore, no harmful metal elements were detected in the biomass and biochars indicating their potential use as non-toxic biofuel and adsorbents for wastewater treatment. 3.7. Surface chemical groups The FTIR spectra of PKS, EFB, POS and their derived biochars contained a number of peaks which indicate that the structure of the materials was composed of various chemical groups. The peak wavenumbers on the spectra had been identified and the possible chemical groups causing each peak are listed in Table 3 together with literature data. The FTIR spectra of PKS and EFB are similar in terms of overall shape and intensity, but that of POS showed relatively weak intensities. This might be attributed to the lower volatiles in POS as compared with PKS and EFB, and this result is in good agreement with the proximate analysis (Table 2). The FTIR spectra of the three biochars demonstrated weaker peak intensities relatively to those of raw biomass, reflecting that volatilization of chemical groups had occurred during pyrolysis. This finding is also consistent with the decrease in volatiles of the biomass after pyrolysis (Table 2). The results shown in Table 3 revealed that pyrolysis of the palm oil wastes had greatly influenced their chemical compositions. The devolatilization effect is evident from the FTIR spectra as there are changes in some peaks before and after pyrolysis. For instance, some of the peaks between 2855 and 2926, 1438 and 1443 and at 721 cm1 attributed to C–H stretching, CH2 and CH3 deformation, and C–H rocking of alkane groups (Coats, 2000), respectively, were shifted and some were eliminated after pyrolysis (Table 3).

3.9. Pyrolysis kinetic The pyrolytic reaction of the biomass was non-isothermal and heterogeneous which was affected by the activation energy (Ea) (Chen et al., 2015). The Ea of PKS, EFB and POS pyrolysis were determined from the linear plots of isoconversional KAS and FWO models as displayed in Fig. 2. Table 4 shows that the Ea values estimated by KAS model for PKS, EFB and POS varied within 130.55–235.59, 130.55–235.59, and 84–445.05 kJ mol1, respectively, while those estimated by FWO ranged between 133.04– 234.28, 133.04–234.28, and 87.81–433.81 kJ mol1, respectively.

160

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

Table 3 Chemical functional groups in PKS, EFB, POS and respective biochars. Class

Functional group

Wavenumber range (cm1)b

PKS

PKS-char

Wavevnumber range (cm

a b c

Alcohols & Phenols Alkanes

O–H stretch(s)a C–H stretch(s)

3600–3400 3000–2850

Aldehydes and Ketones Alkenes Amides Aromatics Carboxylic acids Alkanes Alkyl halides

C@O stretch(s) C@C stretch(v) N–H out of plane C–C stretch(m) C–O stretch(m) CH2, CH3 deformation C–F stretch(s)

1730–1720 1645–1637 1640–1600 1600–1585 1610–1550 1470–1350 1350–1000

Alcohols and phenols Alkenes Aromatics Aromatics Alkenes

C–O stretch(s) @C–H out of plane(s) C–H out of plane(m) C–H out of plane(m) @C–H out of plane(s)

1320–1000 915–905 885–870 865–810 840–790

Alkyl halides

C–Cl stretch(s)

850–750

Alkanes Alkenes Alky halides

C–H rock(m) @C–H out of plane(m) C–Br stretch(s)

725–720 725–675 680–500

Aryl disulfides

S-S stretch

500–430

3406 2926 2855 1726 – 1618 – – 1443 1247 – 1166 1052 – – 853 – – – 771 – – – 606 – –

– – 2866 – – – 1593 – – – 1214 – – – 884 – 829 – – – 759 – – – – –

EFB

EFB-char

POS

POS-char

3423 2930 2866 – – – – 1581 – – 1217 – – – – – 829 – – – – – – – – –

3440 2928 – – 1644 – – – – – – 1034 – 914 – – – – 796 – – – 695 – 539 467

3411 2936 2866 – – – – 1579 – – – 1033 – – – – – – 797 – – – – – – 468

1 c

)

3411 2926 2855 1722 1638 – – – 1438 1245 – 1164 1054 – – – – 8138 – 771 – 721 – 609 – –

s-Strong, m-medium and v-varied. Based on (Coats, 2000). Based on Fig. S2.

Fig. 1. TG curves for PKS (a), EFB (b), POS (c), and DTG curves for PKS (d), EFB (e), POS (f) at different heating rates.

The results calculated by KAS model are very close with those calculated by FWO model, with deviations ranging between 0.0851 and 4.543%. The good agreement between the two result sets verified the reliability and applicability of KAS and FWO models in the

data prediction (Islam et al., 2016). The average Ea evaluated for PKS, EFB and POS by KAS model were 205.70, 169.36 and 261.26 kJ mol1, respectively. Meanwhile, using FWO model, the values were 205.34, 169.76 and 258.77 kJ mol1 for PKS, EFB and

161

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

Fig. 2. Kinetic plots by KAS model for PKS (a), EFB (b) and POS (c), and by FWO model for PKS (d), EFB (e) and POS (f).

Table 4 Activation energies of EFB, PKS and POS determined by KAS and FWO models. Conversion

KAS Model

FWO Model

Deviation (%)

Ea (kJ mol1)

R2

Ea (kJ mol1)

R2

PKS 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Average

130.55 146.46 235.51 235.12 235.35 235.59 221.31 205.70

0.8835 0.9878 0.9639 0.9639 0.9639 0.9639 0.8982

133.04 148.30 233.35 233.41 233.92 234.28 221.12 205.34

0.8971 0.9892 0.9667 0.9668 0.9669 0.9669 0.9068

1.9073 1.2542 0.9181 0.7280 0.6100 0.5533 0.0851

EFB 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Average

130.55 133.88 137.20 155.02 235.19 235.59 158.09 169.36

0.8835 0.8874 0.8909 0.9932 0.9639 0.9639 0.9164

133.04 136.49 139.94 157.16 233.62 234.28 153.75 169.76

0.8971 0.9007 0.9039 0.9940 0.9669 0.9669 0.9144

1.9073 1.9526 1.9959 1.3774 0.6682 0.5533 2.7453

POS 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Average

83.996 97.737 202.45 210.16 445.05 410.07 379.33 261.26

0.9100 0.9366 0.8238 0.8569 0.9815 0.9964 0.9994

87.812 101.88 202.04 209.80 433.81 401.56 374.48 258.77

0.9241 0.9465 0.8374 0.8683 0.9824 0.9966 0.9995

4.5433 4.2423 0.2017 0.1702 2.5236 2.0750 1.2770

POS, respectively. These values are in close agreement with the Ea values for other biomasses determined by similar kinetic approaches, as shown in Table 5. Generally, Ea represents the minimum energy required to initiate a chemical reaction. Therefore, it is economically more feasible to pyrolyze a biomass with lower Ea to biofuel products due to the lower energy requirement. As the

average Ea values for PKS and EFB were determined to be lower than that for POS, they are better candidates for biochar production. Table 4 also shows that Ea of PKS, EFB and POS were influenced by the conversion degree. The Ea determined by KAS and FWO models increased when conversion progressed from 0.2 to 0.7 for

162

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163

Table 5 Comparison of activation energies of PKS, EFB and POS with literature data. Sample

Ea (kJ mol1)

Reference

PKS EFB POS Pophyra yezoensis Plocamium telfairiae Harv Corallina pilulifera Cardoon stems Cardoon leaves Hazelnut husk Rice straw Soybean straw Cymbopogon schoenanthus Urochloa mutica Elephant grass

130.04–235.59 130.55–235.59 84.0–445.05 118.7–176.1 153.0–320.80 191.90–291.20 177.36–338.77 314.44–382.64 103.04–162.06 139.04–282.46 46.32–179.31 84.59–193.17 103.87–241.62 176.34–248.87

Present work Present work Present work Li et al. (2011) Li et al. (2011) Li et al. (2011) Damartzis et al. (2011) Damartzis et al. (2011) Ceylan and Topcu (2014) Mishra and Bhaskar (2014) Huang et al. (2016) Mehmood et al. (2017) Ahmad et al. (2017) Collazzo et al. (2017)

PKS and EFB, and from 0.2 to 0.6 for POS, and then decreased with further progression in conversion. The change in Ea with progressing conversion implied that pyrolysis of PKS, EFB and POS involved different reaction mechanisms. The multi-step kinetics could be due to degradation of hemicellulose and cellulose as well as the slow thermal cracking of lignin. The observations are in accordance with pyrolysis of hazelnut husk (Ceylan and Topcu, 2014) and cardoon (Damartzis et al., 2011). 4. Conclusions The physicochemical properties and pyrolysis kinetics of PKS, EFB, POS and their corresponding biochars were determined in this research. Analysis of TGA data using KAS and FWO model-free isoconversional methods suggested that pyrolysis of the three biomasses involved multi-step kinetics. The estimated average Ea were 205.70, 169.36 and 261.26 kJ mol1 (KAS) and 205.34, 169.76 and 258.77 kJ mol1 (FWO) for PKS, EFB and POS, respectively. The higher HHV (26.18–27.50 MJ kg1) and char yield (35.14–37.07%), and lower average Ea (169.36–205.70 kJ mol1) and moisture content (1.03–2.26%) demonstrated that PKS and EFB have higher biochar potential than POS. Acknowledgements The authors acknowledge the financial support received from the Faculty of Engineering, University of Nottingham Campus Malaysia Campus. The authors are thankful to the management of Seri Ulu Langat Palm Oil Mill Sdn. Bhd., Dengkil, Selangor, Malaysia for supplying the research materials. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2017.03. 105. References Abnisa, F., Arami-Niya, A., Daud, W.M.A.W., Sahu, J.N., 2013. Characterization of biooil and bio-char from pyrolysis of palm oil wastes. BioEnergy Res. 6 (2), 830– 840. Ahmad, M.S., Mehmood, M.A., Al Ayed, O.S., Ye, G., Luo, H., Ibrahim, M., Rashid, U., Arbi Nehdi, I., Qadir, G., 2017. Kinetic analyses and pyrolytic behavior of Para grass (Urochloa mutica) for its bioenergy potential. Bioresour. Technol. 224, 708–713. Akahira, T., Sunose, T., 1971. Joint convention of four electrical institutes. Sci. Technol. 16, 22–31. ASTM International, 2013. ASTM D5865–13, Standard Test Method for Gross Calorific Value of Coal and Coke. ASTM International, West Conshohocken, PA.

ASTM International, 2015. ASTM D7582–15, Standard Test Methods for Proximate Analysis of Coal and Coke by Marco Thermogravimetric Analysis. ASTM International, West Conshohocken, PA. Ayeni, A.O., Hymore, F.K., Mudliar, S.N., Deshmukh, S.C., Satpute, D.B., Omoleye, J.A., Pandey, R.A., 2013. Hydrogen peroxide and lime based oxidative pretreatment of wood waste to enhance enzymatic hydrolysis for a biorefinery: process parameters optimization using response surface methodology. Fuel 106, 187– 194. Beesley, L., Moreno-Jimenez, E., Gomez-Eyles, J.L., Harris, E., Robinson, B., Sizmur, T., 2011. A review of biochars’ potential role in the remediation, revegetation and restoration of contaminated soils. Environ. Pollut. 159 (12), 3269–3282. Bouraoui, Z., Jeguirim, M., Guizani, C., Limousy, L., Dupont, C., Gadiou, R., 2015. Thermogravimetric study on the influence of structural, textural and chemical properties of biomass chars on CO2 gasification reactivity. Energy 88, 703–710. Ceylan, S., Topcu, Y., 2014. Pyrolysis kinetics of hazelnut husk using thermogravimetric analysis. Bioresour. Technol. 156, 182–188. Chen, Z., Hu, M., Zhu, X., Guo, D., Liu, S., Hu, Z., Xiao, B., Wang, J., Laghari, M., 2015. Characteristics and kinetic study on pyrolysis of five lignocellulosic biomass via thermogravimetric analysis. Bioresour. Technol. 192, 441–450. Coats, J., 2000. Interpretation of infrared spectra, a practical approach. In: Meyers, R. A. (Ed.), Encyclopedia of Analytical Chemistry. Joyhn Wiley & Sons Ltd., Chichester, pp. 10815–10837. Colantoni, A., Evic, N., Lord, R., Retschitzegger, S., Proto, A.R., Gallucci, F., Monarca, D., 2016. Characterization of biochars produced from pyrolysis of pelletized agricultural residues. Renew. Sust. Energy Rev. 64, 187–194. Collazzo, G.C., Broetto, C.C., Perondi, D., Junges, J., Dettmer, A., Dornelles Filho, A.A., Foletto, E.L., Godinho, M., 2017. A detailed non-isothermal kinetic study of elephant grass pyrolysis from different models. Appl. Therm. Eng. 110, 1200– 1211. Creamer, A.E., Gao, B., Zhang, M., 2014. Carbon dioxide capture using biochar produced from sugarcane bagasse and hickory wood. Chem. Eng. J. 249, 174– 179. Damartzis, T., Vamvuka, D., Sfakiotakis, S., Zabaniotou, A., 2011. Thermal degradation studies and kinetic modeling of cardoon (Cynara cardunculus) pyrolysis using thermogravimetric analysis (TGA). Bioresour. Technol. 102 (10), 6230–6238. Devi, P., Saroha, A.K., 2015. Effect of pyrolysis temperature on polycyclic aromatic hydrocarbons toxicity and sorption behaviour of biochars prepared by pyrolysis of paper mill effluent treatment plant sludge. Bioresour. Technol. 192, 312–320. Doyle, C.D., 1961. Kinetic analysis of thermogravimetric data. J. Appl. Polym. Sci. 5 (15), 285–292. El may, Y., Jeguirim, M., Dorge, S., Trouvé, G., Said, R., 2012. Study on the thermal behavior of different date palm residues: characterization and devolatilization kinetics under inert and oxidative atmospheres. Energy 44 (1), 702–709. Flynn, J.H., Wall, L.A., 1966. A quick, direct method for the determination of activation energy from thermogravimetric data. J. Polym. Sci. B: Polym. 4, 323– 328. Garcia-Nunez, J.A., Rodriguez, D.T., Fontanilla, C.A., Ramirez, N.E., Silva Lora, E.E., Frear, C.S., Stockle, C., Amonette, J., Garcia-Perez, M., 2016. Evaluation of alternatives for the evolution of palm oil mills into biorefineries. Biomass Bioenergy 95, 310–329. Hodgson, E., Lewys-James, A., Rao Ravella, S., Thomas-Jones, S., Perkins, W., Gallagher, J., 2016. Optimisation of slow-pyrolysis process conditions to maximise char yield and heavy metal adsorption of biochar produced from different feedstocks. Bioresour. Technol. 214, 574–581. Huang, X., Cao, J.-P., Zhao, X.-Y., Wang, J.-X., Fan, X., Zhao, Y.-P., Wei, X.-Y., 2016. Pyrolysis kinetics of soybean straw using thermogravimetric analysis. Fuel 169, 93–98. Ilyushechkin, A.Y., Roberts, D.G., Harris, D.J., 2014. Characteristics of solid byproducts from entrained flow gasification of Australian coals. Fuel Process. Technol. 118, 98–109. Islam, M.A., Auta, M., Kabir, G., Hameed, B.H., 2016. A thermogravimetric analysis of the combustion kinetics of karanja (Pongamia pinnata) fruit hulls char. Bioresour. Technol. 200, 335–341. Jeguirim, M., Bikai, J., Elmay, Y., Limousy, L., Njeugna, E., 2014a. Thermal characterization and pyrolysis kinetics of tropical biomass feedstocks for energy recovery. Energy Sustainable Dev. 23, 188–193. Jeguirim, M., Elmay, Y., Limousy, L., Lajili, M., Said, R., 2014b. Devolatilization behavior and pyrolysis kinetics of potential Tunisian biomass fuels. Environ. Prog. Sustainable Energy 33 (4), 1452–1458. Kissinger, H.E., 1957. Reaction kinetics in differential thermal analysis. Anal. Chem. 29 (11), 1702–1706. Lee, Y., Park, J., Ryu, C., Gang, K.S., Yang, W., Park, Y.K., Jung, J., Hyun, S., 2013. Comparison of biochar properties from biomass residues produced by slow pyrolysis at 500 °C. Bioresour. Technol. 148, 196–201. Leng, L.Y., Husni, M.H., Samsuri, A.W., 2011. Comparison of the carbon-sequestering abilities of pineapple leaf residue chars produced by controlled combustion and by field burning. Bioresour. Technol. 102 (22), 10759–10762. Li, D., Chen, L., Zhang, X., Ye, N., Xing, F., 2011. Pyrolytic characteristics and kinetic studies of three kinds of red algae. Biomass Bioenergy 35, 1765–1772. Liang, L., Sun, R., Fei, J., Wu, S., Liu, X., Dai, K., Yao, N., 2008. Experimental study on effects of moisture content on combustion characteristics of simulated municipal solid wastes in a fixed bed. Bioresour. Technol. 99 (15), 7238–7246. Liu, X., Zhang, Y., Li, Z., Feng, R., Zhang, Y., 2014. Characterization of corncob-derived biochar and pyrolysis kinetics in comparison with corn stalk and sawdust. Bioresour. Technol. 170, 76–82.

X.J. Lee et al. / Bioresource Technology 236 (2017) 155–163 Luangkiattikhun, P., Tangsathitkulchai, C., Tangsathitkulchai, M., 2008. Nonisothermal thermogravimetric analysis of oil-palm solid wastes. Bioresour. Technol. 99 (5), 986–997. Ma, Z., Chen, D., Gu, J., Bao, B., Zhang, Q., 2015. Determination of pyrolysis characteristics and kinetics of palm kernel shell using TGA–FTIR and model-free integral methods. Energ. Convers. Manage. 89, 251–259. Mehmood, M.A., Ye, G., Luo, H., Liu, C., Malik, S., Afzal, I., Xu, J., Ahmad, M.S., 2017. Pyrolysis and kinetic analyses of Camel grass (Cymbopogon schoenanthus) for bioenergy. Bioresour. Technol. 228, 18–24. Mishra, G., Bhaskar, T., 2014. Non isothermal model free kinetics for pyrolysis of rice straw. Bioresour. Technol. 169, 614–621. MPOB, 2016b. Malaysian Palm Oil Board (MPOB). Number and capacities of palm oil sectors December 2016 (tonnes/year) http://bepi.mpob.gov.my/index.php/en/ statistics/sectoral-status/170-sectoral-status-2016/758-number-a-capacitiesof-palm-oil-sectors-2016.html (accessed 26.2.17). MPOB, 2016a. Malaysian Palm Oil Board (MPOB). Overview of the Malaysian Oil Palm Industry 2016 http://bepi.mpob.gov.my/images/overview/Overview_of_ Industry_2016.pdf (accessed 26.2.17). Mu, L., Chen, J., Yin, H., Song, X., Li, A., Chi, X., 2015. Pyrolysis behaviors and kinetics of refining and chemicals wastewater, lignite and their blends through TGA. Bioresour. Technol. 180, 22–31. Nyakuma, B.B., Ahmad, A., Johari, A., Abdullah, T.A.T., Oladokun, O., Aminu, D.Y., 2015. Non-isothermal kinetic analysis of oil palm empty fruit bunch pellets by thermogravimetric analysis. Chem. Eng. Trans. 45, 1327–1332. Ozawa, T., 1965. A new method of analysing thermogravimetric data. Bull. Chem. Soc. Jpn. 38, 1881–1886.

163

Shamsudin, S., Md Shah, U.K., Zainudin, H., Abd-Aziz, S., Mustapa Kamal, S.M., Shirai, Y., Hassan, M.A., 2012. Effect of steam pretreatment on oil palm empty fruit bunch for the production of sugars. Biomass Bioenergy 36, 280–288. Tan, H.T., Lee, K.T., Mohamed, A.R., 2011. Pretreatment of lignocellulosic palm biomass using a solvent-ionic liquid [BMIM]Cl for glucose recovery: An optimisation study using response surface methodology. Carbohydr. Polym. 83 (4), 1862–1868. Tan, X.-F., Liu, S.-B., Liu, Y.-G., Gu, Y.-L., Zeng, G.-M., Hu, X.-J., Wang, X., Liu, S.-H., Jiang, L.-H., 2017. Biochar as potential sustainable precursors for activated carbon production: Multiple applications in environmental protection and energy storage. Bioresour. Technol. 227, 359–372. Thangalazhy-Gopakumar, S., Al-Nadheri, W.M.A., Jegarajan, D., Sahu, J.N., Mubarak, N.M., Nizamuddin, S., 2015. Utilization of palm oil sludge through pyrolysis for bio-oil and bio-char production. Bioresour. Technol. 178, 65–69. Wang, Y., Yin, R., Liu, R., 2014. Characterization of biochar from fast pyrolysis and its effect on chemical properties of the tea garden soil. J. Anal. Appl. Pyrolysis 110, 375–381. White, J.E., Catallo, W.J., Legendre, B.L., 2011. Biomass pyrolysis kinetics: A comparative critical review with relevant agricultural residue case studies. J. Anal. Appl. Pyrolysis 91 (1), 1–33. Yang, H., Yan, R., Chin, T., Liang, D.T., Chen, H., Zheng, C., 2004. Thermogravimetric analysisfourier transform infrared analysis of palm oil waste pyrolysis. Energy Fuels 18 (6), 1814–1821. Zhang, X., Wang, H., He, L., Lu, K., Sarmah, A., Li, J., Bolan, N.S., Pei, J., Huang, H., 2013. Using biochar for remediation of soils contaminated with heavy metals and organic pollutants. Environ. Sci. Pollut. Res. Int. 20 (12), 8472–8483.