Journal of Environmental Management 146 (2014) 189e197

Contents lists available at ScienceDirect

Journal of Environmental Management

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

Characteristics of biochars from crop residues: Potential for carbon

sequestration and soil amendment
Jayne H. Windeatt a, *, Andrew B. Ross a, Paul T. Williams a, Piers M. Forster b,
Mohamad A. Nahil a, Surjit Singh a
a
Energy Research Institute, Faculty of Engineering, University of Leeds, Leeds LS2 9JT, UK
b
School of Earth and Environment, Faculty of Environment, University of Leeds, Leeds LS2 9JT, UK

a r t i c l e i n f o a b s t r a c t

Article history: Biochar has potential to sequester carbon in soils and simultaneously improve soil quality and plant
Received 17 June 2014 growth. More understanding of biochar variation is needed to optimise these potential benefits. Slow
Received in revised form pyrolysis at 600
C was undertaken to determine how yields and characteristics of biochars differ when
1 August 2014
produced from eight different agricultural residues. Biochar properties such as carbon content, surface
Accepted 7 August 2014
Available online
area, pH, ultimate and proximate analysis, nutrient and metal content and the R50 recalcitrance index
were determined. Significant variations seen in biochar characteristics were attributed to feedstock
variation since pyrolysis conditions were constant. Biochar yields varied from 28% to 39%. Average carbon
Keywords:
Biochar
content was 51%. Ash content of both feedstocks and biochars were correlated with biochar carbon
Crop residue content. Macronutrients were concentrated during pyrolysis, but biochar macronutrient content was low
Pyrolysis in comparison to biochars produced from more nutrient rich feedstocks. Most biochars were slightly
Carbon alkaline, ranging from pH 6.1 to pH 11.6. pH was correlated with biochar K content. Aromaticity was
Sequestration increased with pyrolysis, shown by a reduction in biochar H/C and O/C ratios relative to feedstock values.
Recalcitrance The R50 recalcitrance index showed biochars to be either class 2 or class 3. Biochar carbon sequestration
potential was 21.3%e32.5%. The R50 recalcitrance index is influenced by the presence of alkali metals in
the biochar which may lead to an under-estimation of biochar stability. The residues assessed here, at
current global availability, could produce 373 Mt of biochar. This quantity of biochar has the potential to
sequester 0.55 Pg CO2 yr1 in soils over long time periods.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction achieved using a number of processes including (slow, intermedi-

This study characterises biochars produced from eight agricul-
tural residues and assesses the global potential of these residues to
sequester carbon in soils. Biochar, a recalcitrant form of carbon
made by the thermo-chemical conversion of biomass (Manya,
2012), is increasingly discussed as a potential tool in research
fields such as negative emissions technologies, climate change
mitigation, soil quality and food security. The production and use of
biochar, including its addition to soils, animal feed and building
products, may offer an opportunity to tackle a number of these
issues simultaneously (Xu et al., 2012; Zhao et al., 2013). Production
of biochar can also yield oil and gas products which could be used
as a renewable energy source (Manya, 2012). The thermal decom-
position of biomass into biochar, oil and gaseous products can be
* Corresponding author. Tel.: þ44 113 343 2556.
E-mail address: ee06jhw@leeds.ac.uk (J.H. Windeatt).
ate or fast) pyrolysis, gasification, hydrothermal carbonisation
(HTC), torrefaction and traditional carbonisation methods
(Bridgwater, 2003; van der Stelt et al., 2011). Slow pyrolysis pro-
duces higher biochar yields relative to the other processes, and is
considered further here. During pyrolysis the biomass is heated in
the absence of oxygen so that full combustion does not occur.
Volatiles and semi-volatiles are released from the biomass as oil
and gas products and chemical and physical changes leave a bio-
char product. Depending on vapour residence time, a secondary
char forming process may occur with vapours re-condensing onto
the char material, further increasing biochar yields (Antal, 2003).
It is reported within the literature that the characteristics of the
biochar product are influenced by many variables such as the
biomass feedstock, the temperature of pyrolysis, the maximum
temperature of pyrolysis, the hold time at the maximum temper-
ature, and the pyrolysis atmosphere. For example, biochar charac-
teristics such as fixed carbon, surface area, ash content, nutrient
content, pH and cation exchange capacity may vary between

http://dx.doi.org/10.1016/j.jenvman.2014.08.003
0301-4797/© 2014 Elsevier Ltd. All rights reserved.
190 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
biochars due to both feedstock properties and process conditions
(Manya, 2012). The source(s) of variation in biochar characteristics
require further research including the effects of initial feedstock
characteristics on those of the biochar. Studies which include
detailed characterisation of the biochar feedstock are limited in
number. Studies which explore the effects of feedstock on biochar
characteristics, detailing the characteristics of biochars produced
from different feedstocks under uniform process conditions, are
also limited. The variety of process conditions used in studies
throughout the literature often makes it difficult to compare results
regarding the effect of feedstock properties on biochar character-
istics. Various potential biomass feedstocks exist for biochar pro-
duction including agricultural and forestry residues, municipal
wastes, animal manures and purpose grown biomass. A number of
factors should be considered, however, when determining feed-
stock suitability, such as the desired biochar characteristics, sus-
tainability requirements, possible toxicity of the biochar and end-
use. A prominent research focus for the application of the biochar
is its use in soil amendments. This may enable removal of carbon
from the atmosphere resulting in long term storage of carbon, with
potential co-benefits such as soil improvement and increased plant
growth (Biederman and Harpole, 2013). Upon addition of biochar to
soil, the variation seen in biochar characteristics can lead to vari-
ation in the effects on soil processes and plant growth. The
important agronomical properties of biochars when used in soil
amendment include porosity, pH, water holding capacity, nutrient
content and cation exchange capacity. It is also essential that the
biochars contain low levels of contaminants. Improving knowledge
of the properties of biochars produced from different crop residues
using a uniform pyrolysis process will enable further understand-
ing of how feedstocks affect biochar characteristics. It will also
enable discussion of their potential effects on soils, and of the po-
tential of biochars to sequester carbon. This study documents eight
crop residues and their resulting biochars, exploring the effects of
feedstock on biochar yield and characteristics. This primary data is
combined with a modelling study, providing projections of the
carbon sequestration potential of the residues assessed. This brings
together a number of methodologies and different areas of biochar
research.
2. Materials and methods
2.1. Feedstock selection
Agricultural crop residues are an often under-utilised resource
which, if large scale conversion to biochar and co-products was
achieved, could offer both agronomic and carbon sequestration
benefits. Eight agricultural crop residues were selected here for
analysis. Crop production data for the years 2004e2006, from the
Food and Agriculture Organisation's (FAO) statistical database;
FAOSTAT, was used to determine regionally predominant crop types
(FAO, 2014). Residue to product ratios (RPRs) from the literature
were applied to the production quantities of the predominant crop
types, determining those crop types that would yield the highest
residue quantities. The RPR values were calculated as averages from
the literature (Supplementary Material, Table S1). Crops with high
productivity and high RPR were chosen for analysis due to their
potential availability, in large quantities, for biochar production.
The feedstocks used for biochar production were: coconut husk,
coconut shell, cotton stalk, olive pomace, palm shell, rice husk,
sugarcane bagasse, and wheat straw.
Wheat straw and rice husk samples were sourced from fields in
the Faisalabad District, Punjab province, Pakistan (31 21 N, 72 59 E),
Sugarcane bagasse was sourced from Samundri, Pakistan (30 48 N,
71 52 E). Samples were transported in plastic bags and, on receipt,
ground and sieved to 1.4e2.8 mm particle size. Coconut husk and
shell, and palm kernel shells were sourced from the waste stream of
coconut and palm kernel oil processing in the western region of
Ghana. The coconut and palm kernel shells, as received, ranged
from 3.35 to 10 mm particles. Cotton stalks were sourced from
Northern Syria. Olive pomace, sourced from the waste stream of an
olive processing plant in the Southern Mediterranean region, was
received in powdered form with particle size <2 mm. All samples
were stored in air tight containers after grinding, prior to pyrolysis.
2.2. Pyrolysis reactor
A laboratory scale fixed-bed slow pyrolysis reactor was devel-
oped to pyrolyse the biomass feedstock. The reactor was 250 mm in
length by 30 mm internal diameter and was externally heated by a
1.2 kW tube furnace (Supplementary Material, Figure S2). A stain-
less steel sample crucible was used to hold 6 g of each crop residue
sample. The eight biomass types were pyrolysed separately with a
heating rate of 5
C min1 and peak temperature of 600
C which
was held for one hour. The intermediate peak temperature of
600
C was used here as lower pyrolysis temperatures may result in
incomplete charring of the feedstock material and higher pyrolysis
temperatures may reduce the yield of biochar. Williams and Besler
(1996) produced biochar at yields of between 16.2% and 60.8% from
the pyrolysis of wood with temperatures ranging from 300
C to
720
C and heating rates of between 5
C min1 and 80
C min1.
Biochar yields were higher at lower temperatures and lower
heating rates and, whilst the highest biochar yields were achieved
at 300
C, the higher yields may indicate incomplete conversion of
biomass to biochar. Although it is well documented that increased
temperature corresponds to a decreased yield of char, it has also
been documented that the quality of char may increase with
increasing temperature. Other studies have found an increase in
carbon content and calorific value of the biochars with increasing
temperature (Peters, 2011). Biochar recalcitrance is also seen to
increase with increasing temperature (Harvey et al., 2012). Pyrol-
ysis at 600
C was used to balance these factors. Pyrolysis was
undertaken in a nitrogen (N2) atmosphere with a flow rate of
200 ml min1 at 20
C and 1 bar. The N2 carrier gas was introduced
10 min before heating commenced to purge oxygen from the sys-
tem. Biochar was collected and weighed after cooling. The oil
product was collected using a three condenser system, with glass
wool used to remove uncondensed semi-volatiles from the gas
stream. Product gases were collected in a gas tedlar bag and ana-
lysed off-line by gas chromatography.
2.3. Analytical methodology
Proximate analysis of the raw feedstocks and biochars was
conducted using a standardised method. Samples, in a ceramic
crucible with lid, were oven dried at 105
C for 2 h then heated to
550
C and held for 4 h. The lids were removed and the samples
held at 550
C for 1 h to combust the fixed carbon, leaving the
residual ash. Moisture, volatiles and ash were calculated by direct
weight loss and fixed carbon content calculated by difference.
Carbon, hydrogen, nitrogen, sulphur and oxygen content of the
feedstocks and biochars was determined by ultimate analysis using
a Flash 2000 organic element analyser with thermal conductivity
detector. 3 mg of each sample was analysed, in duplicate, with va-
nadium pentoxide (V2O5) added to aid combustion. Oxygen was
calculated by difference. Carbon yield, defined here as % carbon
retained from the original biomass carbon, was calculated from the
ultimate analysis results. Macronutrient content of feedstocks and
biochars was analysed using inductively coupled plasma mass
spectrometry (ICP-MS). 0.2 g of sample was digested in 8 ml of
 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
nitric acid (HNO3), with microwave or sand bath heating to aid
sample digestion. Temperature programmed oxidation (TPO) of the
biochars was conducted by thermal gravimetric analysis (TGA),
using a Mettler Toledo TGA/DSC1 analyser with alumina crucible
and aluminium lid. 5 ug of each biochar sample was heated, in air,
from 35
C to 900
C at 10
C min1. Thermograms were corrected
for moisture and ash following the method described by Harvey
et al. (2012). R50 values, determining the recalcitrance potential of
the biochars, were calculated from the TPO data using the method
described by Harvey et al. (2012) and described in equation (1):
T50 x
R50 x ¼
T50 graphite
where T50 x and T50 graphite are the temperatures at which 50% of the
biochar and graphite were oxidised respectively. The value of
886
C used here for T50 graphite was taken from Harvey et al. (2012).
It was hypothesised that high concentrations of alkali metals such
as potassium (K) within biochars may influence the oxidation
temperature during TPO (Jiang et al., 2013), resulting in a conser-
vative R50 value. Potassium is likely to be leached from the biochar
very rapidly upon addition to the soil (Lindstrom, 1986). To simu-
late this, a sample of the wheat straw biochar was washed in water
at 80
C for 2 h. TPO was performed as described above, in dupli-
cate, on the washed and unwashed wheat straw biochars for
comparison. The carbon sequestration (CS) potential of the bio-
chars, giving a measure of the amount of the original feedstock
carbon that would be retained in biochar for long time periods
upon addition to soil, was determined following the method
described by Zhao et al. (2013) and is described by equation (2).

ðM  Ch  CCh  R50 Þ
CS % ¼
M  CF
where, M is the mass of feedstock (g), Ch is the yield of biochar (%),
CCh is the carbon content of the biochar, R50 is the recalcitrance
index and CF is the carbon content of the feedstock. Biochar surface
area and pore volume were determined using a Quantasorb
continuous flow gas adsorption unit with N2 adsorbate. Vacuum
outgassing was conducted at 200
C for 3 h. Surface area was
determined using the BET method of isotherm analysis. To deter-
mine biochar pH, a solution of 1:20 biochar to deionised water was
shaken for 30 min and left to stand for 10 min before testing with a
standard pH probe. Typical cellulose, hemicellulose and lignin
content were estimated for each feedstock using averages of values
available in the literature (Table 1 and Supplementary Material,
Table S3).
Biochar calorific value was determined using a PARR 6200 bomb
calorimeter. Regression analysis determined significant relation-
ships between characteristics of the raw feedstock and biochars,
and also between the different biochar characteristics. A 95% con-
fidence level (p ¼ 0.05) was used.
2.4. Global carbon sequestration potential
A preliminary assessment of the long term carbon sequestration
potential of biochars produced from the feedstocks detailed in this
Table 1
Typical cellulose, hemicellulose and lignin content of raw feedstocks (db).
Palm shell Sugarcane bagasse Rice husk
Cellulose (%) 30 39 38
Hemicellulose (%) 18 26 18
Lignin (%) 53 24 22
191
paper is presented here. The assessment aims to give an upper limit
estimate of the biochar potential, and examines the potential of
these feedstocks at current global production levels. The scenarios
of carbon sequestration assume that conversion of feedstocks to
biochar is undertaken by slow pyrolysis, and results in properties
and yields of biochar similar to those in this study which are within
the limits of those within the wider literature. Assumptions are
made about crop residue factors and residue availability to deter-
mine an upper limit for biochar potential. The maximum avail-
ability of residues is discussed, however scenarios of reduced
availability are also detailed within the results to show the impact
of more conservative estimates of residue availability on carbon
(1)
sequestration potential. The maximum availability scenario is, from
this point, termed Scenario A. Conservative scenarios detailing 50%
and 10% of the maximum availability are termed Scenario B and
Scenario C respectively. This work, including the assumptions
made, will be approached in more detail in following publications.
The average annual global production of each commodity was
calculated using data for 2005e2012 from the FAOs FAOSTAT (FAO,
2014). Crop specific RPRs were applied to the production values for
each commodity, giving a quantity of residue. The RPR values used
were an average for each crop type, determined from the literature
(Supplementary Material, Table S1). The potential availability of
residues, considering competing uses, was calculated to estimate
the contribution that each feedstock could make to biochar pro-
duction and subsequent carbon sequestration. Palm kernel shell,
coconut shell and husk, olive pomace, rice husk and sugarcane
bagasse are all residues from the processing streams of commercial
agricultural commodities. A number of these residues are currently
used as a fuel source in industry or domestically. Koopmans and
Koppejan (1997) discuss that between 50% and 70% of rice husk is
(2) currently used as a fuel source by the processing industry, and the
remaining 30%e50% is currently a waste product. They also discuss
that the main use of sugarcane bagasse is for energy production by
the sugarcane processing industry. In Scenario A residues currently
used for fuel purposes are assumed to be utilised in a dual biochar/
bioenergy system, such as slow pyrolysis, providing both energy
and biochar products. The slow pyrolysis process has a lower en-
ergy output per unit of feedstock when compared to combustion or
gasification processes. This is justified here by the utilisation within
the pyrolysis system of the residues not currently used, and the co-
benefits of energy production and carbon sequestration. This could
be further incentivised by the inclusion of biochar for carbon
sequestration into carbon credit schemes or other regulatory
mechanisms. This conversion of energy systems is an idealised
scenario to highlight the maximum potential, which in reality may
carry economic impacts which are beyond the scope of this study. It
is therefore assumed for scenario A that all rice husk residues and
sugarcane bagasse will be available for conversion to biochar and
bioenergy. Around 20% of coconut shells and 30% of coconut husks
are used for purposes such as copra drying, with the remainder
mainly used for fuel sources. Around 10% of oil palm shells may
currently be used for purposes such as road construction
(Koopmans & Koppejan, 1997). Scenario A therefore assumes that
80% of coconut shell, 70% of coconut husk and 90% of oil palm shell
residues are available for pyrolysis. Olive stones are currently used
for energy production and for activated carbon production (Guinda,
Coconut shell Wheat straw Cotton stalk Olive pomace Coconut fibre
20 35 35 34 46
49 25 39 15 15
30 19 21 20 33
192 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197

Table 2
Characteristics of raw feedstocks (top) and biochars (bottom).

Raw materials Units Palm shell Sugarcane bagasse Rice husk Coconut shell Wheat straw Cotton stalk Olive pomace Coconut fibre

Ultimate analysis
Ca % 53.1 45.9 42.5 52.6 48.1 46.0 49.2 44.7
Ha % 7.1 6.7 6.5 6.2 6.8 7.6 6.8 7.5
Na % 0.7 0.9 1.3 2.0 1.8 5.6 2.0 0.8
a
S % 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0
O (by diff) % 46.8 59.2 46.0 53.1 49.0 54.5 45.8 61.8
H/C 1.6 1.7 1.8 1.4 1.7 2.0 1.6 2.0
O/C 0.7 1.0 0.8 0.8 0.8 0.9 0.7 1.0
Proximate analysis
Moisture % 3.0 5.8 5.7 5.7 5.6 6.1 5.7 7.5
Volatilesb % 74.1 85.3 80.9 77.2 85.9 93.1 80.5 85.3
b
Fixed Carbon % 25.9 14.7 19.1 22.8 14.1 6.9 19.5 14.7
c
Ash % 2.0 4.4 19.6 0.6 7.9 4.2 4.5 5.3
HHV MJ kg1 19.9 14.6 15.5 17.2 17.3 16.8 18.2 14.9

Chars Units Palm shell Sugarcane bagasse Rice husk Coconut shell Wheat straw Cotton stalk Olive pomace Coconut fibre

Product yields
char % 31.8 27.7 39.0 28.2 30.3 28.0 30.5 30.8
oil % 50.3 50.3 33.5 43.7 50.0 53.6 44.8 47.8
gas % 17.9 23.6 21.8 28.1 17.6 18.4 29.2 25.1
Ultimate analysis
Ca % 90.6 88.6 54.5 93.9 75.3 83.2 71.8 82.6
Ha % 2.8 2.8 2.1 3.0 2.6 3.2 2.8 2.7
Na % 0.9 1.3 1.1 0.4 1.0 4.8 1.9 2.4
a
S % 0.0 0.1 0.0 0.0 0.2 0.1 0.0 0.0
O (by diff) % 7.9 13.7 5.4 2.6 4.5 14.2 11.6 12.8
H/C 0.4 0.4 0.5 0.4 0.4 0.5 0.5 0.4
O/C 0.1 0.1 0.1 0.02 0.04 0.1 0.1 0.1
Proximate analysis
Moisture % 2.2 3.7 5.7 7.1 8.1 8.5 10.0 10.4
Volatilesb % 11.5 30.1 13.9 8.1 21.2 28.8 20.9 25.1
b
Fixed carbon % 88.5 69.9 86.1 91.9 78.8 71.2 79.1 74.9
c
Ash % 6.7 13.0 47.0 4.1 23.4 9.5 18.1 13.5
Other characteristics
HHV MJ kg1 33.6 30.1 19.3 33.7 26.5 31.4 24.1 26.6
Total pore volume Cc g1 0.16 0.18 0.10 0.15 0.01 0.05 0.00 0.04
Surface area M2 g1 220.0 149.1 114.9 222.5 6.3 121.2 1.2 23.2
pH 6.1 8.6 9.9 8.5 11.6 10.3 10.5 9.6
Stored carbonc % 54.3 53.5 50.0 50.3 47.4 50.6 44.5 56.9
R50 0.61 0.53 0.54 0.59 0.46 0.50 0.58 0.49
d
CS % 32.5 27.3 26 28.7 21.3 23.8 24.5 26.8
a
As received basis.
b
Dry ash free basis.
c
Dry basis.
d
CS ¼ carbon sequestration potential.

2006). Scenario A assumes that all olive residues would be available
for pyrolysis as activated carbons can often be either substituted
with, or produced from, biochars (Azargohar and Dalai, 2006).
Activated carbon is often used as an adsorbent, for example in
remediation treatments, and may therefore need careful consid-
eration before addition to soils (Ghaedri et al., 2014). Wheat straw
and cotton stalk residues are normally separated from the com-
modity in-field. As discussed by Lindstrom (1986), a conservative
estimate of 70% residue is required to remain in situ for soil health
and nutrient recycling and has been applied here for these in-field
residues. In reality much of these residues may be burnt in the field
or used locally as fuel. It has been assumed here that all residues
burnt in the field (on average 20% and 5% of residues in the
developing and developed nations respectively (Yevich and Logan,
2003)) are available instead for collection for biochar production.
Ballesteros et al. (2006) state that wheat straw is used for animal
bedding, with the remainder burnt for fuel or left in situ. Following
the methodology of Woolf et al. (2010a) Scenario A assumes that all
straw used for animal bedding will eventually become available for
pyrolysis. Straw currently burnt for energy is also assumed to be
available for pyrolysis. The majority of cotton stalk is currently
burnt in the field, with the remainder used for fuel purposes
(Koopmans and Koppejan, 1997) therefore Scenario A assumes all
cotton stalk residue to be available for biochar production.
Following the calculation of available feedstock for the three
scenarios, the experimental data published here regarding biochar
yields, biochar carbon content and recalcitrance were used to
determine the potential biochar yield, biochar carbon content, and
the long-term carbon sequestration potential of each residue type.
Following the method of Zhao et al. (2013) for the carbon
sequestration (CS) potential of biochars, as detailed in equation
(2), the amount of feedstock carbon expected to remain in soils
long-term, through conversion to biochar and additions to soil,
was calculated. There are a number of caveats involved in this
projection including assumptions made about RPR values,
competing uses of residues, and rates of biochar degradation. In
reality, these values may vary spatially and temporally. Yields and
characteristics of biochars are, in reality, dependent on a number
of factors including feedstock properties which can vary within
strain and with growing conditions, and with biochar production
process and conditions. Other residues, such as rice straw and
sugarcane trash would also be produced alongside those residues
examined here, but are not included in this assessment of biochar
potential.
 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
3. Results and discussion
The proximate and ultimate analysis characterisations of the
feedstocks are listed in Table 2 together with proximate and ulti-
mate analysis, pH, CV, surface area, pore volume, calorific value
(CV), stored carbon, R50 values and carbon sequestration potential
(CS) of the biochars.
3.1. Feedstock composition
The typical lignocellulosic biomass components of the feedstock
are listed in Table 1 and the proximate and ultimate analysis are
listed in Table 2. Carbon content, on an as received basis, was
highest in the palm shell feedstock (531 g kg1) and lowest in the
rice husk (425 g kg1). This corresponds with the typical lignocel-
lulosic composition and the ash content, with the rice husk feed-
stock having correspondingly high ash content at 196 g kg1 and
low lignin content, the palm shell having low ash content
(20 g kg1) and the highest lignin content. The concentrations of
macronutrients (potassium (K), phosphorus (P), calcium (Ca),
magnesium (Mg)) in the feedstocks are shown in Fig. 1(a).
3.2. Biochars
Biochar yields, as a percentage of the original biomass material,
were found to range from 28 wt % for sugarcane bagasse biochar to
39 wt % for rice husk biochar. Examples from the literature of
biochar yield from slow pyrolysis are 25e62 wt % using different
types of charcoal kiln in research by Antal (2003), and 26e63 wt %
by Peng et al. (2011) with yields being dependent on temperature
and residence time. Higher yields of biochar, above 40%, have been
reported from the pyrolysis of cellulose, hemicelluloses and woody
materials under high pressure (Antal and Varhegyi, 1995).
Demirbas (2004) discussed that biomass with higher lignin content
produces higher yields of biochar. Here biochar carbon content
varied from 55% for rice husk biochar to 94% for coconut shell
Fig. 1. Macronutrient content (g kg1) (dry basis) of (a) raw feedstocks and (b) biochars. N.B. Scales vary between nutrient species.
193
biochar. The balance of carbon stored in the biochar, as the mass
fraction of the carbon remaining in biochar from the original
feedstock carbon, was between 45% for olive pomace and 57% for
coconut fibre. The average value of carbon stored from the original
feedstock carbon was 51%. This compares well with the values of
49% used by Woolf et al. (2010b) and 50% seen by Xu et al. (2012).
The volatile content of the biochars ranged from 8% to 30%. Ash
content was found to range from 4.1% in coconut shell biochar to
47% in rice husk biochar. In each case, the biochar contained higher
levels of ash than its corresponding feedstock although some
feedstocks increased more than others, suggesting the ash
composition may be influencing the biochar yields. A positive
correlation was seen between biochar yield and both feedstock and
biochar ash content (r2 ¼ 0.76, p ¼ 0.005; r2 ¼ 0.74, p ¼ 0.006
respectively). In general, the higher the ash content, the lower the
carbon content of the biochar. The lowest carbon content was seen
in rice husk biochar which has the highest ash content, conversely
the highest carbon content was seen in the coconut shell biochar
which has the lowest ash content. This aligns well with Enders et al.
(2012) suggestion that biomass feedstocks with relatively high ash
contents produced relatively low fixed carbon biochars, which was
attributed to the high ash content inhibiting the formation of aro-
matic carbon forms. Despite this, for feedstocks such as olive stone
and wheat straw, the increase in ash content is higher than for the
other feedstocks. There also appears to be a link between ligno-
cellulosic composition and fixed carbon content; the feedstock with
highest lignin content (palm shell) produced a high carbon biochar,
and the feedstock with highest cellulose content (sugarcane
bagasse) produced a biochar with the lowest carbon content.
Clearly there is a shift from high volatile, low fixed carbon
composition for the raw biomass to low volatile, high fixed carbon
composition for the product biochars. The results here confirm the
discussion of Enders et al. (2012) that biochars have low volatile
content and high fixed carbon contents, when compared with that
of raw biomass, for a range of feedstocks. Kim et al. (2012) also
showed, via a Van Krevelen diagram, that H/C and O/C ratios in
194 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
biochar are considerably reduced compared to the original biomass
feedstock. They reported H/C and O/C ratios, for biochar produced
from the fast pyrolysis of pitch pine at 500
C, which were similar to
those reported in this work for the slow pyrolysis of biochars
produced at 600
C. This same trend is seen in H/C and O/C ratios
for the feedstocks and biochars examined here. The decrease in H/C
and O/C ratios shows the loss of volatile hydrocarbon and
oxygenated hydrocarbon species and the increasing carbon content
of the biomass feedstock during pyrolysis (Supplementary Material,
S4). Krull et al. (2009) reported that a decreasing ratio of hydrogen
to carbon in biochars indicates an increasing aromatic structure in
the biochar. This also corresponds with an increase in the higher
heating value (HHV) of the biochar, meaning the energy density of
the product is increased (Basu, 2013). The lower H/C ratios of the
biochars here tend to correspond with higher HHV.
3.2.1. Agronomical properties of biochars
The surface areas of the biochars ranged from 1.2 m2 g1 (olive
pomace) to 223 m2 g1 (coconut shell). Surface area can be influ-
enced by both feedstock properties and process conditions. A low
H/C ratio in the feedstock may contribute to lower surface area in
the biochar due to less graphitisation of the structure during py-
rolysis (Uchimiya et al., 2011). In this study, olive pomace and
wheat straw result in low surface area biochars. For olive pomace
this may be due to the presence of residual oils in the biomass,
which has been observed in other studies. Pagnanelli et al. (2008)
attributed the low surface area of olive pomace biochar to parti-
cle conglomeration and the blocking of pores by the decomposition
and deposition of tars during the pyrolysis process. The low surface
area of wheat straw biochar may be related to the lower lignin
content of the feedstock, meaning less structural support during
the charring process (Uchimiya et al., 2011). The other six biochars
exhibited higher surface areas, with palm shell and coconut shell
biochars having the highest surface areas. Previous studies have
shown the influence of temperature on surface area of biochars.
Maximum biochar surface areas were identified by Uchimiya et al.
(2011) at production temperatures between 500
C and 900
C. The
BET surface areas detailed in this study, produced within the op-
timum temperature range for maximum surface area, are indicative
of the upper levels of surface areas achievable for the feedstocks
assessed. Further increases in surface area can be achieved by
including a further activation process, for example steam activation,
resulting in an activated carbon. At higher temperatures (~900
C)
pore structure may begin to break down resulting in reduced sur-
face area. Total pore volume ranged from negligible for olive
pomace biochar to 0.18 cc g1 for sugarcane bagasse biochar. A
significant correlation was seen between surface area and total
pore volume (r2 ¼ 0.8, p ¼ 0.003). In general, high surface area is
regarded as a positive attribute in biochars used for soil amend-
ment. The physical properties of soil, such as structure and water
holding capacity, and the biological properties, such as microbial
communities and earthworm presence, may be altered by biochar
application (Biederman and Harpole, 2013). The high surface areas
and porosities observed for some of the biochars in this work, for
example coconut shell, indicate good water retention properties
and potential habitat for microbial communities (Atkinson et al.,
2010).
Biochar pH varied from pH 6.1 (palm shell biochar) to pH 11.6
(wheat straw biochar) with the majority of biochars being slightly
alkaline. In general higher levels of alkali and alkali earth metals
corresponded to higher pH of the biochar. Correlation was seen
between biochar pH and both feedstock and biochar potassium (K)
concentration (r2 ¼ 0.74, p ¼ 0.01; r2 ¼ 0.70, p ¼ 0.02 respectively).
Palm shell produced slightly acidic biochar and was also found to be
low in alkali metals and high in lignin content. Previous literature
reports biochar pH values (without further processing of the chars)
of between pH 4 and pH 12, with typical values being above pH 7.
Zhao et al. (2013) found biochar pH levels between 8.8 and 10.8,
depending on biomass feedstock type. They also reported that the
pH of biochars increases with increasing pyrolysis temperature due
to the enrichment of ash as the temperature increases. All of the
biochars produced here were within the pH range found in the
literature. Alkaline biochars have the potential to buffer excess soil
acidity. Application of biochar to soils has been found to have an
overall liming effect, increasing soil pH (Biederman and Harpole,
2013), although knowledge of the initial pH of both biochar and
soil is necessary to amend soil pH using biochar. Increasing the pH
of acidic soils has been seen to increase microbial activity,
increasing soil organic matter mineralisation and increasing
nutrient availability to plants. This may in some circumstances,
however, cause a priming effect resulting in the increased emission
of CO2 from soils and may also have only a short term effect on
microbial activity (Sohi et al., 2009).
Both direct and indirect effects of biochar addition to soil may
occur due to influence from different biochar characteristics,
although nutrient content and the ability of biochar to retain
nutrients are important. The nutrient species and concentrations
required for optimal plant growth are dependent on factors such
as soil type, initial soil nutrient concentrations and plant specific
requirements (Havlin et al., 1999). In some soils, an excess of
nutrient may already exist and the biochar could potentially
adsorb nutrients. Knowledge of the nutrient content of biochars
from different feedstocks will enable biochar production to
compliment efforts to achieve the optimum nutrient concentra-
tions in soils, with potential for reduction in fertiliser requirement
(Roberts et al., 2010). The potential for biochars to trap nutrients
via cation exchange capacity is, however, also an important
consideration. The phosphorus (P), potassium (K), calcium (Ca)
and magnesium (Mg) content of the raw biomasses and the bio-
chars are shown in Fig. 1.
Nutrient concentrations vary between both the different raw
feedstock and the different biochars. For most feedstocks inves-
tigated, the total nutrient content was concentrated within the
biochar during the pyrolysis process. Examination of the litera-
ture indicated that total concentrations of N, P and K are often
under-reported, however a large range of N, P and K concen-
trations exist within those reported. For example, Chan and Xu
(2009), in a meta-analysis of the current literature, reported
concentrations of total N of 1.8 g kg1 to 56.4 g kg1, total P of
2.7 g kg1 to 480 g kg1 and total K of 1.0 g kg1 to 58 g kg1. For
the biochars produced here, total N ranged from 4.3 g kg1
(coconut shell biochar) to 47.8 g kg1 (cotton stalk biochar). Total
P content of the biochars was between <0.1 g kg1 (palm shell
and coconut shell biochars) to 4.2 g kg1 (olive pomace biochar).
Total K ranged from 0.6 g kg1 (palm shell biochar) to 60 g kg1
(wheat straw biochar). It is important, however, to note that total
levels of nutrients do not imply bio-availability. Often, concen-
trations of mineral N are low even where total N is high, as in
biochars produced from poultry manures (Chan et al., 2007).
Conversely, due to the high solubility of K, a higher proportion of
the total amount is typically bioavailable. Chan and Xu (2009)
analysed biochar content of both total and available P, finding a
range of 0.2e73 g kg1 total P and 15 to 11,600 mg kg1 available
P. The biochar with 0.2 g kg1 total P had 15 mg kg1 available P,
and the biochar with 25.2 g kg1 total P had 11,600 mg kg 1,
indicating that available P increases with total P, however the
limited reporting of either total P and/or available P for all bio-
chars makes it difficult to determine trends. For total N and P, the
values seen in this study are low compared to the higher values
of the range found in the literature.
 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197 195

3.2.2. Biochar stability and sequestration potential

3.2.2.1. Recalcitrance and carbon content. Although biochar carbon
often has a higher recalcitrance than the carbon in the raw biomass,
due in large to increased aromatic structure, the lifetime and sta-
bility of the carbon in soils is still uncertain and variable between
biochars, soil types and environmental conditions, with estimates
ranging from tens to thousands of years. One method of estimating
the recalcitrance of biochars, in relation to that of graphite, is the
R50 recalcitrance index (Harvey et al., 2012). Comparison of R50
values with carbon mineralisation rates led Harvey et al. (2012) to
classify biochars, by their degradation potential, into three classes:
R50  0.7, most recalcitrant; 0.5  R50 < 0.7, minimal degradation;
R50 < 0.5, more degradable (Fig. 2).
Using the R50 recalcitrance classification system, three of the
biochars studied here were rated as class 3 biochars (most
degradable) and 6 of the biochars as class 2 (slightly degradable).
None of the biochars examined were class 1 (highly resistant to Fig. 3. Temperature programmed oxidation of washed and unwashed wheat straw
degradation). Using the R50 values the biochars, from most to least biochar. 50% oxidation of the biochars is shown by the horizontal dotted line and T50
recalcitrant, are: palm shell, coconut shell, olive pomace, rice husk, values are indicated for the washed and unwashed biochars.
sugarcane bagasse, cotton stalk, coconut fibre and wheat straw.
Palm shell biochar (R50 ¼ 0.60) would be the most resistant to value for the unwashed biochar relative to that of the washed
degradation and wheat straw biochar (R50 ¼ 0.45) would be the biochar (Fig. 3). The R50 value of wheat straw was increased to 0.55
least resistant. This corresponds to the biochars with the highest with washing prior to TPO. The addition of biochar to soils would
typical lignin content (palm shell biochar) and the lowest typical likely see rapid leaching of the alkali metals by water, therefore the
lignin content (wheat straw biochar). Zhao et al. (2013) found that R50 values, determined by TPO, of biochars high in alkali metals
R50 values increased with pyrolysis temperature, with wheat straw may be lower than the recalcitrance expected in soils. The R50
biochars produced at 500
C and 650
C having R50 values of 0.54 values for the biochars investigated here which were high in alkali
and 0.71 respectively. Increasing pyrolysis temperatures could, metals, for example wheat straw and olive pomace, may therefore
therefore, potentially increase the recalcitrance of the biochars be conservative values. These biochars, therefore, have a potentially
produced here. The carbon sequestration (CS) potentials of the longer lifetime in soil than indicated by their R50 values.
biochars, calculated using the method described by Zhao et al.
(2013), indicate that between 21.3% (wheat straw) and 32.5% 3.2.2.2. Global carbon sequestration potential. Assessment of the
(palm shell) of the original feedstock carbon would be retained long global potential for long term carbon storage using biochars
term in soils. These are within the range, but towards the lower showed that some feedstocks assessed here have higher potential
end, of CS potentials determined by Zhao et al. (2013) who inves- than others (Table 3).
tigated the production of biochars from 12 different biomass In Scenario A 373 Mt yr1 of biochar could be produced, con-
feedstocks and reported CS potentials from 21.1% to 47.1%. Corre- taining 283 Mt carbon. This equates to 1.04 Pg CO2 yr1. Scenarios B
lation was seen between the typical lignin content of feedstocks and C see the CO2 equivalent content of the biochar carbon reduced
and biochar CS potential (r2 ¼ 0.78, p ¼ 0.004). We investigated the to 0.5 Pg CO2 yr1 and 0.1 Pg CO2 yr1 respectively. Some of the
potential influence of alkali metals on the R50 recalcitrance method. biochar carbon would, however, be released back to the
As hypothesised, the thermal degradation of biochar was catalysed
by high alkali metal content, leading to thermal degradation at
Table 3
lower temperatures. Theoretical potential for biochar production, carbon content of biochar, equivalent
Potassium (K) content was reduced by 92% with washing, from CO2 of the carbon in biochar, and long term carbon storage potential of biochars
2173 mg kg1 to 171 mg kg1. A catalytic effect of the high K content from the residues detailed in this study. Scenarios A, B and C represent the
maximum residue availability scenario, 50% and 10% of the maximum availability
of the unwashed wheat straw biochar was seen, with a lower T50
respectively.

Biochar Biochar CO2 eq of C remaining CO2 eq of C stored

yield carbon biochar long-term long-term (Mt)
(Mt) (Mt) carbon (Mt)
(Mt)

Scenario A
coconut husk 2.5 1.9 7.0 0.8 2.9
coconut shell 4.9 4.4 16.1 2.7 9.9
olive pomace 1.7 1.3 4.8 0.7 2.6
palm kernel 2.9 2.6 9.5 1.6 5.9
shell
rice husk 79.7 68.6 251.8 23.4 85.9
cotton stalk 16.0 11.3 41.5 6.7 24.6
sugarcane 182.1 127.3 467.2 85.5 313.8
bagasse
wheat straw 83.2 65.6 240.8 28.8 105.7
Total 373.0 283.0 1038.6 150.2 551.2

Scenario B 186.5 141.5 519.3 75.1 275.6

Fig. 2. Temperature programmed oxidation of the biochars, corrected for moisture and
ash. Central dashed line indicates 50% oxidation of the biochar. Thermograms have
Scenario C 37.3 28.3 103.9 15.0 55.1
been corrected for moisture and ash content.
196 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
atmosphere as CO2 as the biochar degrades. The long term carbon
storage potential is estimated as 0.55 Pg CO2 yr1, 0.28 Pg CO2 yr1,
and 0.06 Pg CO2 yr1 for Scenarios A, B and C respectively. In an
assessment of the global potential of biochar for climate change
mitigation, Woolf et al. (2010a) estimated emissions reductions of
1.8 Pg CO2eCe yr1. The estimate includes a number of other resi-
dues, including manures and purpose grown biomass, making
direct comparison difficult. Rice husk, sugarcane bagasse and
wheat straw residues combined currently have 93% of the total
biochar production potential of the feedstocks assessed. Biochars
from these three feedstocks would contain 92% of the total carbon
content of all biochars produced in Scenario A.
Global CO2 emissions from fossil fuel burning were around 36
billion tonnes (or Petagrams (Pg)) in 2013. Using the CS potentials
of the biochars produced here, Scenario A would therefore be able
to sequester the carbon equivalent of 1.5% of annual carbon dioxide
emissions from fossil fuel burning in 2013. Scenario B would
sequester 0.8%, and Scenario C 0.2% of carbon dioxide emissions
from fossil fuel burning in 2013. Sugarcane bagasse demonstrates
the largest potential for long-term carbon storage, at 57% of the
total, and sugarcane bagasse, wheat straw and rice husk residues
combined contribute 92% of the total long term carbon storage
potential. Although rice husk residues have the second highest
biochar production potential, it falls into third place behind wheat
straw residues for long-term carbon storage potential due to the
carbon content of the respective biochars. Sugarcane bagasse, rice
husk and wheat straw residues have a comparatively high biochar
production potential, biochar carbon content and overall long-term
carbon storage potential when compared with the other residues
assessed. This assessment highlights that factors such as residue
availability and biochar recalcitrance must be considered alongside
the yield and other biochar properties when determining the
suitability of a feedstock. Further assessment of these factors
affecting the net potential of biochars to sequester carbon long-
term will be approached in related publications.
4. Conclusions
Biochars from eight crop residues exhibited variation in a num-
ber of characteristics. Biochar yields were 28 wt % to 39 wt %, with
relationships seen between biochar yield and ash content of both
feedstock and biochar. Overall, the nutrient contents of the feed-
stocks and biochars were low, with the majority of species being
concentrated during pyrolysis. Biochar pH range (pH 6.1 to pH 11.6)
showed the majority of biochars were slightly alkaline, with re-
lationships seen between K content and pH. Average carbon storage
was 51%. High lignin feedstocks produced high carbon biochars with
high recalcitrance. Feedstock and biochar H/C and O/C ratios indicate
increasing aromaticity upon charring. The biochars were either
moderately or more highly degradable. None were of the most
recalcitrant classification. Recalcitrance values were conservative
for biochars with high alkali metal content due to catalysis during
TPO. Between 21.3% and 32.5% of the feedstock carbon would be
stored long-term in biochars. Currently 150 Mt yr1 of carbon, which
equates to 0.55 Pg CO2 yr1, could be stored long-term in soils as
biochar produced from the feedstocks assessed here.
Acknowledgements
The award of an Engineering & Physical Sciences Research
Council (EPSRC) Scholarship and funding through the EPSRC
Doctoral Training Centre in Low Carbon Technologies to J. Windeatt
(Grant No EP/G036608/01, the European Commission for funding of
the Fertipus FP7 project (Proposal N
289853) supporting A.B Ross
and S. Singh, and the Royal Society Wolfson Merit Award to P.M.
Forster are gratefully acknowledged.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jenvman.2014.08.003.
References
Antal, M.J., Varhegyi, G., 1995. Cellulose pyrolysis kinetics: the current state of
knowledge. Ind. Eng. Chem. Res. 34, 703e717.
Antal, M.J., 2003. The art, science, and technology of charcoal production. Ind. Eng.
Chem. Res. 42, 1619e1640.
Atkinson, C.J., Fitzgeraki, J.D., Hipps, N.A., 2010. Potential mechanisms for achieving
agricultural benefits from biochar application to temperate soils: a review. Plant
Soil 337, 1e18.
Azargohar, R., Dalai, A.K., 2006. Biochar as a precursor of activated carbon. Appl.
Biochem. Biotech. 129e132, 762e773.
Ballesteros, I., Jose-Negro, M., Olvia, J.M., Cabanas, A., Manzanares, P., Ballesteros, M.,
2006. Ethanol production from steam-explosion pretreated wheat straw. Appl.
Biochem. Biotech. 129e132, 496e507.
Basu, P., 2013. Biomass characteristics. In: Biomass Gasification, Pyrolysis and Tor-
refaction: Practical Design and Theory, second ed. Academic Press, London.
Biederman, L.A., Harpole, S., 2013. Biochar and its effects on plant productivity and
nutrient cycling: a meta-analysis. Glob. Change Biol. 5, 202e214.
Bridgwater, A.V., 2003. Renewable fuels and chemicals by thermal processing of
biomass. Chem. Eng. J. 91, 87e102.
Chan, K.Y., Van Zweiten, L., Meszaros, I., Downie, A., Joseph, S., 2007. Assessing the
agronomic values of contrasting char materials on Australian hardsetting soil.
Proceedings of the Conference of the International Agrichar Initiative (Terrigal,
NSW, Australia).
Chan, K.Y., Xu, Z., 2009. Biochar: nutrient properties and their enhancement. In:
Lehmann, J., Joseph, S. (Eds.), Biochar for Environmental Management: Science
and Technology. Easthscan, London, pp. 67e84.
Demirbas, A., 2004. Effects of temperature and particle size on bio-char yield from
pyrolysis of agricultural residues. J. Anal. Appl. Pyrol. 72, 243e248.
Enders, A., Hanley, K., Whitman, T., Joseph, S., Lehmann, J., 2012. Characterisation of
biochars to evaluate recalitrance and agronomic performance. Bioresour.
Technol. 114, 644e653.
FAO, 2014. FAOSTAT: Food and Agriculture Organisation of the United Nations:
Statistics Division. Online: http://faostat3.fao.org/faostat-gateway/go/to/home/
E (accessed 31.07.14.).
Ghaedri, M., Ghaedri, A.M., Abdi, F., Roosta, M., Sahraei, R., Daneshfar, A., 2014.
Principal component analysis-artificial neural network and genetic algorithm
optimization for removal of reactive orange 12 by copper sulfide nanoparticls-
activated carbon. J. Ind. Eng. Chem. 20, 787e795.
Guinda, A., 2006. Uses of solid residue from the oil industry. Grasas Aceites 57,
107e115.
Harvey, O., Kuo, L., Zimmerman, A., Louchouarn, P., Amonette, J., Herbert, B., 2012.
An index-based approach to assessing recalcitrance and soil carbon seques-
tration potential of engineered black carbons (biochars). Environ. Sci. Technol.
46, 1415e1421.
Havlin, J.L., Beaton, J.D., Tisdale, S.L., Nelson, W.R., 1999. Soil Fertility and Fertilizers,
an Introduction to Nutrient Management, sixth ed. Prentice Hall, New Jersey.
Jiang, L., Hu, S., Sun, L., Su, S., ZXu, K., He, L., Xiang, J., 2013. Influence of different
demineralization treatments on physicochemical structure and thermal
degradation of biomass. Bioresour. Technol. 146, 254e260.
Kim, K.H., Kim, J.Y., Cho, T.S., Choi, J.W., 2012. Influence of pyrolysis temperature on
pyhsicochemical properties of biochar obtained from the fast pyrolysis of pitch-
pine (Pinus rigida). Bioresour. Technol. 118, 158e162.
Koopmans, A., Koppejan, J., 1997. Agricultural and forestry residues e generation,
utilization and availability. Paper Presented at the Regional Consultation on
Modern Applications of Biomass Energy (Kuala Lumpur, Malaysia).
Krull, E.S., Baldock, J.A., Skjemstad, J.O., Smernik, R.J., 2009. Characteristics of bio-
char: organo-chemical properties. In: Lehmann, J., Joseph, S. (Eds.), Biochar for
Environmental Management: Science and Technology. Earthscan, Sterling, USA
and London, UK, pp. 53e66.
Lindstrom, M.J., 1986. Effects of residue harvesting on water runoff, soil erosion and
nutrient loss. Agr. Ecosyst. Environ. 16, 103e112.
Manya, J.J., 2012. Pyrolysis for biochar purposes: a review to establish current
knowledge gaps and research needs. Environ. Sci. Technol. 46, 7939e7954.
Pagnanelli, F., Mainelli, S., Toro, L., 2008. New biosorbent materials for heavy metal
removal: product development guided by active site characterisation. Water
Res. 42, 2953e2962.
Peng, X., Ye, L.L., Wang, C.H., Zhou, H., Sun, B., 2011. Temperature- and duration-
dependent rice straw-derived biochar: characteristics and its effects on soil
properties of an Ultisol in southern China. Soil Till. Res. 112, 159e166.
Peters, B., 2011. Prediction of pyrolysis of pistachio shells based on its components
hemicellulose, cellulose and lignin. Fuel Process. Technol. 92, 1993e1998.
 J.H. Windeatt et al. / Journal of Environmental Management 146 (2014) 189e197
Roberts, K.G., Gloy, B.A., Joseph, S., Scott, N.R., Lehmann, J., 2010. Life cycle assess-
ment of biochar systems: estimating the energetic, economic, and climate
change potential. Environ. Sci. Technol. 44, 827e833.
Sohi, S., Loez-Capel, E., Krull, E., Bol, R., 2009. Biochar's roles in soil and climate
change: a review of research needs. CSIRO Land Water Sci. Rep. 05/09, 64.
Uchimiya, M., Wartelle, L.H., Klasson, K.T., Fortier, C.A., Lima, I.M., 2011. Influence of
pyrolysis temperature on biochar property and function as a heavy metal sor-
bent in soil. J. Agr. Food Chem. 59, 2501e2510.
van der Stelt, M.J.C., Gerhauser, H., Kiel, J.H.A., Ptasinski, K.J., 2011. Biomass
upgrading by torrefaction for the production of biofuels: a review. Biomass
Bioenerg. 35, 3748e3762.
Williams, P.T., Besler, S., 1996. The influence of temperature and heating rate on the
slow pyrolysis of biomass. Renew. Energy 7, 233e250.
197
Woolf, D., Amonette, J.E., Street-Perrott, F.A., Lehmann, J., Joseph, S., 2010a. Sus-
tainable biochar to mitigate global climate change. Nat. Commun. 1, 1e9.
Woolf, D., Amonette, J.E., Street-Perrott, F.A., Lehmann, J., Joseph, S., 2010b. Sus-
tainable biochar to mitigate global climate change: supplementary information.
Nat. Commun. 1, 1e9.
Xu, G., Yingchun, L., Junne, S., Hongbo, S., Linlin, W., 2012. Recent advances in
biochar applications in agricultural soils: benefits and environmental implica-
tions. Clean 40, 1093e1098.
Yevich, R., Logan, J.A., 2003. An assessment of biofuel use and burning of agricul-
tural waste in the developing world. Glob. Biogeochem. Cy. 17, 1e42.
Zhao, L., Cao, X., Masek, O., Zimmerman, A., 2013. Heterogeneity of biochar prop-
erties as a function of feedstock sources and production temperatures.
J. Hazard. Mater. 256, 1e9.