Energy Conversion and Management 151 (2017) 190–198

Contents lists available at ScienceDirect

Energy Conversion and Management

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

A study on co-pyrolysis of bagasse and sewage sludge using TG-FTIR and Py- MARK
GC/MS
⁎
Yan Lin, Yanfen Liao , Zhaosheng Yu, Shiwen Fang, Xiaoqian Ma
School of Electric Power, South China University of Technology, 510640 Guangzhou, China
Guangdong Province Key Laboratory of Efficient and Clean Energy Utilization, 510640 Guangzhou, China

A R T I C L E I N F O A B S T R A C T

Keywords: In this paper, the pyrolysis process of mixed samples, bagasse with the addition of sludge, were investigated by
Co-pyrolysis TG-FTIR, and Py-GC/MS. For the mixed samples, the proportions of sludge were 10%, 30% and 50%. With the
TG-FTIR sludge ratio increased, the initial temperature and the terminated temperature increased. Eleven chemical
Py-GC/MS structure groups were resolved from both of bagasse and sludge by KBr-Tabletting based FTIR analysis. Aromatic
KBr-Tabletting based FTIR
compounds, CO2, C]O, CeO, NH3 and HCN were the major gaseous products surveyed from the TG-FTIR
Sewage sludge
inspection. The percentage of pyrolysis products (acetic acid, C7#, furfural, benzene derivatives, C7P) were
Bagasse
figured out from Py-GC/MS analysis. The addition of sludge was beneficial for the production of group C7P,
whose carbon atoms exceeded 7. This research provided useful background knowledge for co-pyrolysis of ba-
gasse and sewage sludge.

1. Introduction
A great number of sewage sludge is produced with the dramatic
growing of municipal wastewater treatment plant. As reported in
China, at least 11.2 M tons of dry sewage sludge is generated by
thousands of wastewater treatment plant, and the growth rate is 10%
a year [1]. Some researches had found that approximate 50% weight of
sewage sludge was organic content, most of which were hemicelluloses,
cellulose, lignin, lipids and protein [2,3]. It is consequently feasible to
achieve the resourceful usage of sewage sludge. Based on the existing
sugar milling industry situation of the world, the bagasse, a residue of
sugar milling, is needed to dispose. Bagasse is a major solid residue
from sugar milling industry, and is also an ideal feedstock for renewable
industry [4]. There were more than 70 million tons of sugarcane pro-
duction annually, which produced about 7 million tons of sugarcane
bagasse after sugar milling [5].
As a convenient route for the reduction of waste, incineration was
broadly implemented worldwide. Approximately 55% of generated
sewage sludge in Japan and 25% in Denmark were disposed by the
incineration technology [6]. Although incineration had been proven
as an efficient way for the resourceful usage of waste [7], the pre-
treatment and pollution control made a huge cost in the incineration
[2]. Pyrolysis, the thermal destruction lacking of oxygen, could achieve
huge reduction in waste volume, the stabilization of organic matter, as
well as the production of fuels. Compared to incineration, pyrolysis
converted the biomass energy into different types of bio-fuel, which
provided a more flexible pattern for the usage of biomass energy.
Hence, pyrolysis of solid waste attracted lots of attentions of researchers
[8–11]. Alvarez et al. found that the catalytic activity of the sludge ash
promoted the bio-oil formation [12]. Lin et al. had observed that the
thermochemical reactivity was improved at low temperature on the co-
pyrolysis of oil-palm solid wastes and paper sludge [13]. They attrib-
uted into the hydrogenation and thermal catalytic effects. Fang et al.
had investigated the catalytic co-pyrolysis process with metal oxide,
such as MgO, Al2O3 and ZnO [14]. With the addition of metal oxide, the
pyrolysis reaction began at lower temperature, and the apparent acti-
vation energy was significantly reduced. In our previous work [11], we
had studied the co-pyrolysis interaction between oil shale and sewage
sludge using TG-FTIR system. Interestingly, low proportion of sludge
blending with oil shale could improve the production of methane. The
pyrolysis behavior of mixed materials was distinctly improved because
of the presence of sludge. Nevertheless, the trial on bagasse pyrolysis
with adding sludge was seldom reported.
The current researches on pyrolysis of bagasse focused on the im-
provement of bio-oil and bio-char [15–19], which provided useful in-
formation for the subsequent usage of pyrolysis products. However, the
further investigation on reaction pathway was rarely performed, espe-
cially the co-pyrolysis reaction pathway. In this paper, KBr Pellets -
by Infrared Spectroscopy, TG-FTIR and Py-GC/MS system were used in
the co-pyrolysis process of bagasse and sewage sludge. The knowledge

⁎
Corresponding author at: School of Electric Power, South China University of Technology, 510640 Guangzhou, China.
E-mail address: yfliao@scut.edu.cn (Y. Liao).

http://dx.doi.org/10.1016/j.enconman.2017.08.062
Received 11 July 2017; Received in revised form 21 August 2017; Accepted 22 August 2017
Available online 07 September 2017
0196-8904/ © 2017 Elsevier Ltd. All rights reserved.
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Table 1 Eexp (x ) = ∑ Zi (x ) (1)

The ultimate analyses and proximate analyses of bagasse and sludge on dry basis.
Each of characteristic peak was simulated by Gauss-Lorentz
Samples Ultimate analysis (wt.%) Proximate analysis (wt.%)
Amplitude function:
C H O N S Volatile Fixed carbon Ash Zi (x ) = a3 Gi (x ) + (1−a3) Li (x ) (2)
Bagasse 44.2 6.517 42.78 0.67 0.196 67.18 27.15 5.63 Specifically, Gauss and Lorentz Amplitude were expressed as below:
Sludge 22.21 4.013 17.51 3.42 0.464 33.38 13.97 52.39
2
(x −a1i) ⎤
Gi (x ) = a0i exp ⎡
⎢− 2a 2 ⎥
obtained from these methods presented the basic information for co- ⎣ 2i ⎦ (3)
pyrolysis process of bagasse and sewage sludge, which was significant a 0i
for the development of co-utilization technique. Li (x ) =
( x − a1i ) 2
1+⎡ ⎤
⎣ a2i
⎦ (4)

2. Materials and methods Software Peakfit 4.0 was used to separate the characteristic peaks of
functional groups. The FTIR spectra was recorded from 4000 to
2.1. Materials 900 cm−1.

The sewage sludge used in this study was obtained from Liede
Wastewater Treatment Plant of Guangzhou, China. The bagasse was
collected from Zhanjiang, Guangdong province of China. Samples were
put in the loft drier, holding at 105 °C for 48 h air-drying. Then, dried
samples were crushed and sieved. Finally, the particles less than
178 μm was obtained. The ultimate and proximate analysis results were
shown in Table 1. The test of ash was performed on wavelength dis-
persive X-ray fluorescence Axio Pw4400 (Holland, PANalytical B.V.),
and the results were presented in Table 2. Sludge was accounted for the
weight proportion of 10%, 30% and 50% in the blends, which were
named as 0.1sludge, 0.3sludge and 0.5sludge, respectively.
2.2. Experimental facility and methods
2.2.1. KBr-Tabletting based FTIR
Nicolet™iS™ 10 FT-IR spectrometer was used to obtain the KBr-
Tabletting based spectrum of bagasse and sludge. FTIR analysis was
used as a qualitative technique for analyzing the functional groups of
materials. This method was based on the characteristic peak separations
of FTIR spectra, which has been employed widely in previous re-
searches [20,21]. The separation theory was that: the experimental
spectrum Eexp(x) was assumed a superposition of groups of single
characteristic peaks Zi(x) (i = 1, 2, … , n), and matching functions were
used to search for these peaks.
Table 2
The result of X-ray Fluorescence on ash (XRF) of sewage sludge
on dry basis.
Ash composition (XRF) wt.%
2.2.2. TG-FTIR
To investigate the pyrolysis behavior of bagasse, sewage sludge and
their blends, TG-FTIR experiments were carried out by using thermo-
gravimetric analyzer (METTLER TOLEDO TGA/DSC1) coupled with
Fourier transform infrared measurements (Nicolet™iS™ 10 FT-IR spec-
trometer). The temperature precision of TGA is ± 0.5 °C and sensitivity
of microbalance is lower than ± 0.1 μg. Pure N2 was used as the agent
gas at a flowing rate of 80 ml/min. All of samples were kept at the
weight of 10 ± 0.5 mg for reducing heat transfer limitation. It was
heated from 50 °C to 1100 °C with the heating rate of 20 °C/min. In
addition, the volatiles released during co-pyrolysis process were de-
tected online by FTIR spectrometer, in which spectra was recorded from
4000 to 900 cm−1. FTIR analysis was carried out using a resolution of
4 cm−1 and 8 scan per sampling, and it started when the sample tem-
perature reached at 100 °C. The transfer line and detector were heated
at a constant temperature of 225 °C for preventing condensation. To
eliminate the background signal, blank experiments were carried out
before loading samples.
TG(t) was defined as Eq. (5):
mt
TG (t ) = × 100%
m0 (5)
where m0 and mt represented the initial mass and the mass at tem-
perature t, °C.
And DTG(t),%/min, was defined as Eq. (6):
d [TG (t )]
DTG (t ) =
dT (6)
where T, min, represented the time, and there was a relationship,
t = 20 T, according to the heating rate.

SiO2 52.77 2.2.3. Py-GC/MS

Al2O3 22.32
P2O5 9.65
Py-GC/MS experiments were designed for studying more details
Fe2O3 6.22 about the distribution of organic products from co-pyrolysis process. In
K2O 3.21 this study, CDS Pyroprobe 5200 HP pyrolyser (Chemical Data systems)
CaO 2.77 connected with a gas chromatography/mass spectrometry (7890B GC
MgO 1.19
System/5977A MSD) was used to perform these experiments. All of
TIO2 0.7
SO3 0.43 samples were kept at 1 ± 0.1 mg. Pyrolysis was performed at 700 °C
MnO 0.21 for 20 s using a heating rate of 20 °C/ms. The temperatures of oven and
Na2O 0.13 transfer line were both 300 °C. The GC system equipped with a HP-5MS
Cr2O3 0.09 capillary column (30 m × 0.25 mm × 0.25 μm), and its interface tem-
ZnO 0.08
BaO 0.06
perature was held on 300 °C. The GC oven started at 50 °C and was kept
NiO 0.05 for 2.25 min. After that, the heating program was performed with a
CuO 0.04 heating rate of 10 °C/min, and it finished when the oven temperature
SrO 0.03 reached 300 °C. The GC oven was held on 300 °C for 5 min, then the GC
ZrO2 0.02
program came into the end. Pure helium was used as agent gas of GC/
Cl 0.02
As2O3 0.01 MS with a flowing rate of 1 ml/min and a 1:50 split ratio. Mass spectra
detecting covered a range of 35–300 amu. The compounds were

191
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Table 3
The pyrolysis characteristic parameters of bagasse, sludge and their blends.

Samples Tia (°C) Tfb (°C) Mfc (%) DTG1d T1e (°C) DTG2d T2e (°C)
(%/min) (%/min)

Bagasse 176.7 563 27.57 −8.15 228.7 −13.34 346.0

0.1sludge 177.5 561 31.49 −7.39 228.8 −11.93 346.2
0.3sludge 180.8 559 37.90 −5.94 230.5 −9.98 346.6
0.5sludge 186.0 564 43.64 −4.72 230.8 −8.32 347.0
Sludge 196.7 965 56.70 −3.19 291.9 −0.59 930.4

a
Ti, the initial temperature.
b
Tf, the terminated temperature.
c
Mf, the residue mass.
d
DTG1, DTG2, the mass loss rate of the first peak and the second peak.
e
T1, T2, the temperature where the first peak and the second peak appeared.

Fig. 1a. The TG profiles of bagasse, sewage sludge and their blends under heating rate
20 °C/min.
determined by referring to the spectral data with those in the NIST
library.
All of the experiments mentioned above, the ultimate analysis, the
proximate analysis, the test of ash contents, TG-FTIR analysis and Py-
GC/MS analysis, were performed at least twice for the repeatability of
these results.
3. Results and discussions
3.1. Thermogravimetric analysis
The TG and DTG profiles of bagasse, sludge and their blends under
heating rate 20 °C/min were displayed in Fig. 1. From DTG curves, the
shape and trend of co-pyrolysis profiles of blends were similar to ba-
gasse. The pyrolysis process of the blends and bagasse could be divided
into four individual stages, including moisture removal
(start ∼ 160 °C), devolatilization (160–250 °C), stabilization
(250–560 °C) and the inorganic decomposing (560–1000 °C). The pyr-
olysis process of sludge was divided into three stages, and had an ob-
vious mass-loss peak appeared at a range of 950–990 °C due to the large
ash content in sludge [11]. During the devolatilization stage, hemi-
celluloses was first to decompose due to its low polymerization degree.
At the same time, some primary depolymerizing reaction started in
cellulose [22,23]. Therefore, a shoulder was observed between the first
Fig. 1b. The DTG profiles of bagasse, sewage sludge and their blends under heating rate
20 °C/min.
and the second peak. Beyond 320 °C, the reaction of cellulose became
more intensified. The maximum peak of DTG appeared in 346 °C, which
located in the major stage of cellulose decomposing. Lignin began to
decompose and had a shoulder at the range of 400–560 °C [24].
Through increasing the addition of sludge, two mass-loss peaks de-
creased from −8.15 to −4.72%/min (occurred at about 229 °C) and
from −13.34 to −8.32%/min (occurred at about 346 °C), respectively.
However, the shoulder (400–560 °C) was slightly strengthened, which
might corresponded to the protein decomposing in sludge [12].
The pyrolysis characteristic parameters of samples were presented
in Table 3. The intersection of the TG curve tangent and the horizontal
curve is defined as the initial temperature. The terminated temperature
is obtained when the mass lose accounts for 98% of the total mass loss.
With the increasing of the sludge ratios, the initial temperature (from
176.7 °C to 196.7 °C), the terminated temperature (from 563 °C to
965 °C) and the residue mass (from 27.57% to 56.70%) of samples in-
creased. While, the DTG1 and DTG2 decreased. The second peak of in-
dividual sludge was 930.4 °C, which was higher than other samples.
This phenomenon might attributed to the slight decomposition of ash
content. According to the mentioned above, sludge could reduce the
intensification and delay the initial and terminated temperature of
pyrolysis reaction.
3.2. Chemical structures in KBr-Tabletting based FTIR
As a non-food agricultural residue, bagasse was totally different
from sewage sludge whatever elementary compositions or chemical
components (Table 1). Higher proportion of hydrogen and carbon, also
extremely less ash content, made the greater high heat value of bagasse
than sewage sludge.
A summary of functional groups were listed in Table 4 [25–27]. The
separation was performed in the FTIR spectra to distinguish the func-
tional groups. Each characteristic peak had an individual fitting func-
tion, and these data were put on the supporting file (FTIR Fitting
Results). The coefficient of correlation R2 between fitting curves and
the experimental data was over 0.99. Bagasse was a typical biomass,
containing cellulose, hemicelluloses and lignin. The FTIR fitting ana-
lysis using Gauss-Lorentz Amplitude function of bagasse was shown in
Fig. 2a. The aromatic-nucleus C]C and aromatic CeH belonged to the
structure units of lignin, involving sinapinic alcohol, coniferyl alcohol
and p-coumaryl alcohol [28]. The CeO bond was included in the units
of cellulose and hemicellulose as well as methoxyl of lignin unit. Ap-
parently, the C]O bond was not common to observe in cellulose and
lignin. But, some of hemicellulose unit carried the branches of carbonyl
groups, like xylan [29]. Besides, the linkage of lignin and hemi-
celluloses might consist of C]O bond, which had been found by Yuan
et al. [30]. Heteroatom-connected CeH stretching vibration appeared
in 2761 cm−1, which was related to the methoxy group carried by
syringyl lignin and guaiacyl lignin [24]. The hydroxyl of cellulose and
hemicellulose caused the strong absorbance happened at the range of

192
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Table 4
The chemical structures in KBr-Tabletting based FTIR of bagasse and sludge on dry basis.

Functional groups Peaks Bagasse Sludge

(cm−1)

OeH bending vibration 913 / √

925 √ /
CeO stretching vibration (alcohol, ether, ester, 990 √ √
carboxyl) 1051 √ √
1110 √ √
1131 √ √
1161 √ √
CeH flexible vibration (eCH3, eCH2) 1210 √ √
1260 √ √
1321 √ √
1381 √ √
C]C stretching vibration (aromatic nucleus) 1431 √ √
1460 √ √
1510 √ √
1600 √ √
C]O stretching vibration (ketone, aldehyde, 1633 √ √
carboxyl, ester) 1660 √ √
Fig. 2b. The FTIR resolving of sewage sludge.
1731 √ √
XeY]Z or X^Y stretching vibration (X, Y and Z 2253 √ /
represents one of the C, N, S and O, 2577 √ √
bond and peptide bond had a C]O linkage, so the absorbance of C]O
respectively)
CeH stretching vibration (heteroatom- 2761 √ √
bond was higher in sludge than bagasse. However, benzo-fused deri-
connected) vatives did not exist as much as that of bagasse. It was determined by
CeH stretching vibration (eCH3, eCH2) 2848 √ √ the characteristics of protein and lipid, which were not as benzene-rich
2884 √ √ as biomass-derived components. As a result, the absorbance of aro-
2921 √ √
matic-nucleus C]C bond in sludge became weaker. Biomass compo-
2970 √ √
CeH stretching vibration (aromatic nucleus) 3070 √ √ nents had a large amount of hydroxyl groups in their branches, while
OeH stretching vibration (hydrogen-bonding 3241 √ √ sludge was slightly comprised of them [8]. Thus, the absorbance of
association among molecules) sludge happened at a range of 3200–3600 cm−1 was weaker. NeH
eNH2 and eNH stretching vibration √ √
3329
bonds were observed in 3619 and 3693 cm−1, which did not appear in
(dissociated)
OeH stretching vibration (hydrogen-bonding 3414 √ √
bagasse spectra. According to the result of ultimate analysis (Table 1),
association inside molecule) the nitrogen content of sludge was over five times than bagasse, which
OeH stretching vibration (dissociated) 3544 √ √ was mostly carried by protein [11]. The emission of nitrogen pollution
NeH stretching vibration (oxime, amine) 3619 / √ should be concerned during the co-pyrolysis process.
3693 / √
The chemical structures determined the generated products during
pyrolysis. According to the results of FTIR analysis, CeO, C]O and
OeH abounded in raw material. In the generated bio-energy, it was
common to found a certain content of carboxylic acid, aldehyde and
ketone. These compounds decreased the total energy density of the
generated bio-fuel due to its high ratio of O/C and low carbon atoms.
Besides, The low carbon chain length weakened the restriction to hy-
drogen atom of carboxyl. The hydrogen ionization was improved, and
consequently caused the acidity of bio-fuel. Reducing the acid content
was needed for the equipment protection.

3.3. Gaseous products and functional groups by TG-FTIR

TG-FTIR was used for online detection of the gas-phase products

during co-pyrolysis process. Six groups of IR profiles [10] were in-
vestigated and were showed in Table 5. The absorbance via time (min)
profiles of their compounds were presented in Fig. 3.
From Fig. 3a, the first peak of aromatic compounds of blends and
bagasse occurred at 12.35 min, and the corresponding temperature was
about 347 °C. The aromatization of cellulose unit which was common to
Fig. 2a. The FTIR resolving of bagasse.
Table 5
The IR absorbance for gaseous products and functional groups.
−1
3200–3600 cm .
Sewage sludge was mostly comprised of lipid and protein [31]. As Groups Wavenumber (cm−1)

presented in Fig. 2b, the CeO and C]O of sludge were more con- Aromatic compounds 3015
spicuous than that of bagasse, while the absorbance of aromatic-nucleus CO2 2358
C]C was opposite. Aliphatic acid was one of the major components in C]O 1768
lipids, and it was linked with glycerinum through esterification [32]. CeO 1121
NH3 980
On another respect, amino acid was connected with another one by
HCN 719
peptide bond, and formed the structure unit of protein. Both of ester

193
Y. Lin et al.
Fig. 3a. The aromatic compounds versus time of samples.
Fig. 3b. The CO2 absorbance versus time of samples.
see in the later stage of cellulose decomposing, as well as the primary
cracking of lignin, contributed to the first generation of aromatic
compounds. When the monitoring came into 20.67 min (513 °C), a
highly huge and extremely strong absorption was observed in these
profiles. That was the signal that the lignin became activated and came
into its major disintegrating stage. At this stage, the release of aromatic
compounds were strengthened. For sludge, the first peak occurred at
19.95 min (499 °C). The amino acids, the protein unit, began to fall
apart and generated compounds carrying benzene-ring, like phenols,
indoles and pyrroles [33]. The curve showed a small peak at nearly
38.18 min (863.6 °C), where was usually the temperature for the in-
organic decomposing. That might attributed to lignin, a tiny amount in
sludge, which were extremely stable and hardly reacted below 800 °C
[34]. Fig. 3b depicted the CO2 absorbance versus time of samples. The
break and reforming of thermolabile functional groups, like eCOOH,
C]O, and R-O-R, contributed to the release of CO2 during the pyrolysis
process [35]. When temperature exceeded 800 °C, the CO2 peak ap-
peared in the spectrum again. It might be contributed by the conversion
of CaCO3 to CO2:
C aCO3 → C aO + CO2
Adding 10% sludge slightly improved CO2 absorbance peak at
11.87 min (337.4 °C). In addition, according to previous researches
[36], the pyrolysis reaction of biomass was intensified when co-pyr-
olysis with another high ash content feedstock. The extra metallic mi-
neral in the ash contributed to the positive interaction during pyrolysis.
Energy Conversion and Management 151 (2017) 190–198
Fig. 3c. The C]O absorbance versus time of samples.
Fig. 3d. The CeO absorbance versus time of samples.
Figs. 3c and 3d showed the curves of functional groups C]O and CeO.
These two functional groups related to the products derived from cel-
lulose, hemicelluloses and lignin. With growing proportion of sludge,
C]O and CeO presented a downtrend in their peak. It indicated that
the release of ketones, carboxylic acid, aldehydes, esters, alcohols might
be reduced.
A lot of researches showed that NH3 and HCN were the precursors of
nitrogen oxides [37]. The monitoring of these N-containing products
was useful for the pollution control during the co-pyrolysis process. The
nitrogen of product mainly presented as HCN in bagasse and NH3 in
sludge respectively. Both of HCN and NH3 significantly fell to a low
level during the co-pyrolysis process. Some interactions between ba-
gasse and sludge certainly reduced the (NH3 + HCN)-N yield. Yuan
et al. had discovered something familiar in their research [38]. They
believed that (NH3 + HCN)-N yield, from fast co-pyrolysis process of
biomass/coal, decreased and converted into other volatile-N yield. The
production of (NH3 + HCN)-N, which occurred at 5–20 min
(200–500 °C) and showed in Figs. 3e and 3f, almost covered the same
range of temperature related to the devolatilization. The maximum
(7) absorbance was observed in 11.72 min (334.4 °C) in NH3 as well as
HCN profiles. It seemed to be a good advantage of co-pyrolysis in the
reduction of (NH3 + HCN)-N yields. Guo et al. found that calcium-
based compounds was helpful for the transformation of N2 from fuel N
[39]. The reaction followed as below:
C aO + C harN → C aC x+ CO + N2 (8)
194
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Fig. 3e. The NH3 absorbance versus time of samples.

Fig. 3g. The integral value of each compound of the samples.

The benzene derivatives were characterized by the benzene ring:

Based on the quantity of carbon atom (≤C7 or > C7), the other
products were divided into two groups. The number of carbon atom less
than or equal to 7 was named as C7#, and the others were named as
C7P. The identified fraction accounted for over 71% in these samples,
and the specific detail of these products were presented in the sup-
porting file (Py-GC/MS data). The experiments were performed at least
twice, and the results were repeatable as Figs. 4a and 4b showed. The
classification and peak-area percentage of bagasse and sewage sludge
were showed in Table 6.
Fig. 3f. The HCN absorbance versus time of samples. Table 7 showed the classification and peak-area percentage of
samples. The largest part of bagasse products were benzene and its
derivatives, which mostly came from the decomposition of lignin, and
Therefore, the reduction of (NH3 + HCN)-N yields might be caused
also partly from the aromatization of cellulose and hemicelluloses.
by the calcium-based compounds, which has been detected in the XRF
Especially, over 30% of benzene derivatives were benzofuran. The
test of sludge ash (Table 2).
phenylbenzofuran structure was ordinary to be found in lignin [24].
Based on the Beer-Lambert’s law, the concentrations of gaseous
Lou et al. found that the secondary pyrolysis of intermediate phe-
products presented a linear relationship with the FTIR absorbance.
nylbenzofuran played an important role in the benzofuran forming
Hence, the integral of FTIR profiles could be used to evaluate the re-
lative yields of the gaseous products [10]. The integral values of dif-
ferent sample were depicted in Fig. 3g. The yields of aromatic com-
pounds and CO2 reached their maximum when 10% sludge was added.
The reduction of C]O and CeO indicated that the yields of oxygenated
compounds declined. The yields of nitrogen pollutants (NH3 and HCN)
were also decreased in varying degrees with different percentage of
sludge. This phenomenon suggested that the interaction between ba-
gasse and sludge had an inhibitive effect on the evolution of nitrogen
pollutants.

3.4. Py-GC/MS analysis

According to the thermogravimetric analysis in Section 3.1, the

devolatilization finished before 700 °C, which was selected as the
temperature settled for Py-GC/MS experiments. Five groups were es-
tablished for classifying various of products (acetic acid, furfural,
benzene derivatives, C7# and C7P). The characteristic of furfural group
was the furfural based structure:
Fig. 4a. The Py-GC/MS results of bagasse.

195
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Table 7
The classification and peak-area percentage of samples from Py-GC/MS.

Compounds Bagasse 0.1sludge 0.3sludge 0.5sludge Sludge

Acetic acid 11.37 10.25 10.15 9.67 4.36

C7# 16.84 16.43 14.87 13.42 7.00
Furfural 11.39 9.76 8.09 5.93 1.44
Benzene derivatives 32.71 33.42 32.38 28.35 20.22
C7P 2.95 4.27 7.45 14.62 38.46
Identified fraction 75.26 74.12 72.94 72.00 71.48
Unidentified 24.74 25.88 27.06 28.00 28.52

Fig. 4b. The Py-GC/MS results of sewage sludge.

Table 6
The classification and peak-area percentage of bagasse and sewage sludge from Py-GC/MS.

No. Bagasse Group Sludge Group

RT/min Area Pct/ Compound RT/min Area Pct/ Compound

% %

1 1.56 4.56 Methyl glyoxal C7# 1.71 2.59 Propanal, 2-methyl- C7#
2 1.80 1.69 2,3-Butanedione C7# 1.95 4.36 Acetic acid Acetic acid
3 2.04 11.37 Acetic acid Acetic acid 2.15 2.00 Butanal, 3-methyl- C7#
4 2.23 3.89 2-Propanone, 1-hydroxy- C7# 3.03 0.91 Pyridine BDa
5 2.58 0.96 Furan, 2,5-dimethyl- Furfural 3.15 1.03 N-Pyrrylcarbinol C7#
6 4.28 2.52 Furfural Furfural 3.30 4.75 Toluene BD
7 5.54 0.38 2-Cyclopenten-1-one, 2-methyl- C7# 4.32 1.44 Furfural Furfural
8 5.73 0.69 2(5H)-Furanone C7# 4.42 1.38 1H-Pyrrole, 3-methyl- C7#
9 5.93 2.58 1,2-Cyclopentanedione C7# 4.77 1.57 Pyridine, 3-methyl- BD
10 6.56 1.41 2-Furancarboxaldehyde, 5-methyl- Furfural 4.89 0.96 p-Xylene BD
11 6.95 1.49 Phenol BDa 5.26 2.37 Styrene BD
12 7.70 1.02 1,2-Cyclopentanedione, 3-methyl- C7# 6.97 0.45 1-Decene C7P
13 8.16 0.92 Phenol, 2-methyl- BD 7.01 1.18 Phenol BD
14 8.52 1.05 p-Cresol BD 7.63 0.84 D-Limonene C7P
15 8.68 1.56 Phenol, 2-methoxy- BD 7.72 0.94 1-Butanamine, 2-methyl-N-(2- C7P
methylbutylidene)-
16 9.19 0.36 2-Cyclopenten-1-one, 3-ethyl-2-hydroxy- C7# 8.56 2.96 p-Cresol BD
17 9.35 0.63 2H-Pyran-3(4H)-one, dihydro-6-methyl- C7# 9.49 0.61 Benzyl nitrile BD
18 9.71 1.04 4H-Pyran-4-one, 2,3-dihydro-3,5-dihydroxy-6- C7# 9.73 0.71 Phenol, 2,6-dimethyl- BD
methyl-
19 9.94 0.89 Phenol, 4-ethyl- BD 10.04 1.03 Phenol, 4-ethyl- BD
20 10.30 0.44 Creosol BD 10.19 0.68 1-Dodecene C7P
21 10.71 13.10 Benzofuran, 2,3-dihydro- BD 11.64 0.32 1-Tridecene C7P
22 11.07 6.50 5-Hydroxymethylfurfural Furfural 11.83 2.15 Indole BD
23 11.46 0.44 1,2-Benzenediol, 3-methoxy- BD 12.99 0.93 1-Tetradecene C7P
24 11.55 0.48 Phenol, 4-ethyl-2-methoxy- BD 13.09 1.01 Indole, 3-methyl- BD
25 12.06 3.56 2-Methoxy-4-vinylphenol BD 14.27 0.46 1-Pentadecene C7P
26 12.57 1.99 Phenol, 2,6-dimethoxy- BD 15.25 0.58 Dodecanoic acid C7P
27 12.64 0.28 Eugenol BD 15.48 0.66 Cetene C7P
28 13.32 0.48 Vanillin BD 15.61 0.45 2,2,4-Trimethyl-1,3-pentanediol C7P
diisobutyrate
29 13.81 0.47 3,5-Dimethoxy-4-hydroxytoluene BD 17.04 0.63 7-Tetradecene, (Z)- C7P
30 13.86 0.78 trans-Isoeugenol BD 17.17 0.47 Cyclododecane C7P
31 14.93 0.55 2-Propanone, 1-(4-hydroxy-3-methoxyphenyl)- BD 17.46 2.01 Tetradecanoic acid C7P
32 15.62 0.18 2,2,4-Trimethyl-1,3-pentanediol diisobutyrate C7P 17.70 0.95 9-Octadecene, (E)- C7P
33 15.71 0.33 Phenol, 2,6-dimethoxy-4-(2-propenyl)- BD 18.12 0.68 Pentadecanoic acid C7P
34 16.28 0.27 (E)-2,6-Dimethoxy-4-(prop-1-en-1-yl)phenol BD 18.62 0.96 Cyclopentadecane C7P
35 16.84 1.38 (E)-2,6-Dimethoxy-4-(prop-1-en-1-yl)phenol BD 18.85 0.66 Hexadecanenitrile C7P
36 17.27 0.29 Ethanone, 1-(4-hydroxy-3,5-dimethoxyphenyl)- BD 19.56 12.52 n-Hexadecanoic acid C7P
37 17.36 1.41 4-((1E)-3-Hydroxy-1-propenyl)-2-methoxyphenol BD 21.21 4.47 9-Octadecenoic acid, (E)- C7P
38 19.46 1.09 n-Hexadecanoic acid C7P 21.37 3.13 Octadecanoic acid C7P
39 19.94 0.55 trans-Sinapyl alcohol BD 21.55 1.57 Hexadecanamide C7P
40 21.13 0.41 Oleic acid C7P 23.09 2.48 9-Octadecenamide, (Z)- C7P
41 21.32 0.45 Octadecanoic acid C7P 23.14 0.57 9-Octadecenamide, (Z)- C7P
42 23.09 0.43 9-Octadecenamide, (Z)- C7P 23.28 0.72 Octadecanamide C7P
43 26.29 0.39 1-Nonadecene C7P 24.10 0.36 1-Nonadecene C7P

a
BD, Benzene and its derivatives.

196
Y. Lin et al. Energy Conversion and Management 151 (2017) 190–198

Fig. 4c. The area percent of products of bagasse, sewage sludge and their blends. Fig. 4d. The area percent of furfural of bagasse, sewage sludge and their blends.

[40]: compounds linked by CeC bond, and resulted in the growth on the area
percent of group C7P.
The pathway of the yield of group C7P during co-pyrolysis was de-
picted as follow:

The area percent of benzene and its derivatives was slightly higher
in 0.1sludge than that of bagasse (Fig. 4c). With increasing proportion
of sludge, the source of the benzene and its derivatives declined in the
mixture, and the results hence behaved a slide form the area percent.
On another respect, the yield of group C7P was larger in sludge, and the
growth of sludge content provided larger amount of source for the
production of group C7P.
It was common to see ketones, carboxylic acid, aldehydes, esters,
alcohols and benzene derivatives presented in the products of bagasse
[41]. The oxygenated compounds were formed from the breakages of The addition of sludge significantly improved the yield of group
CeOeC and eCOOH. Most of these oxygen-containing compounds had C7P, and three possible pathways for the sludge based catalysis con-
a low boiling point and usually existed as a small molecule, whose tributed to the growth on group C7P.
carbon atom (C7#) was less than or equal to 7. As Fig. 4c showed, the
area percent of group C7# and acetic acid were decreased from 11.37% 3.5. Future work
to 9.67% and from 16.84% to 13.42% respectively, and the group C7P
remarkably increased from 2.95% to 14.62%. It was noteworthy that As growing demand of solid waste disposal, converting solid waste
acetic acid and most of group C7# compounds carried unsaturated into valuable products using pyrolysis technique is regarded as an ef-
bond. With the catalysis of sludge ash, these compounds might combine ficient route [45]. This paper provided useful background information
for the carbon chain extension, and the group C7P presented a uptrend for co-pyrolysis of bagasse and sewage sludge. Although the framework
from its area percent when the addition of sludge increased. of pyrolysis process was elementally established, the co-pyrolysis pro-
Furfurals were regarded as the products from dehydrating cycliza- cess was complex, and more effects were needed for the reaction con-
tion of cellulose unit, and they possessed a stable structure [42]. The trol. The gaseous effluents were needed to be attentive during the
ratio of furfural products was reduced from 11.39% to 5.93% for in- pyrolysis of solid waste. The downstream utilization of co-pyrolysis
creasing proportion of sludge. The area percent of 5-Hydro- products should be considered. Hence, specific investigation of co-
xymethylfurfural remarkably decreased, and resulted in the reduction pyrolysis reactions, systematic investigation on gaseous effluents and
of total furfural products (Fig. 4d). George W. Huber et al. had found downstream utilization of pyrolysis products were the interests in our
that acid catalyst (e.g., SiO2eAl2O3) had a selectivity on the formation subsequent work.
of liquid alkanes with the number of carbon atom ranging from C7 to
C15, and this process was followed by aldol condensation over solid
4. Conclusions
base catalysts [43]. SiO2 and Al2O3 were common to see in the sludge
ash [44] and accounted for over 40% of the sludge ash. In the XRF test
The characteristics through TGA, the FTIR resolving through KBr-
of sludge ash (Table 2), the weight proportion of SiO2 and Al2O3 in
Tabletting based FTIR, the gaseous products and functional groups
sludge ash were 52.77% and 22.32% respectively. With the catalysis of
through TG-FTIR, the types of products and their proportions through
SiO2 and Al2O3, the furfural products were converted to long-chain
Py-GC/MS of co-pyrolysis of bagasse, sewage sludge and their blends

197
Y. Lin et al.
were studied in this paper. The bagasse and blends could be divided
into four decomposition phases, while sludge had three stages. With the
sludge ratio increased, the initial temperature, terminated temperature
and the residue mass increased, while DTG1 and DTG2 decreased. OeH,
CeO, CeH, C]C, C]O, XeY]Z, X^Y, OeH & NeH these chemical
structures groups were resolved from bagasse and sludge. Gauss-
Lorentz Amplitude function was adopted to the FTIR fitting analysis. In
addition, the contents of CeO, C]O and OeH were abundant in both of
bagasse and sewage sludge. Aromatic compounds, CO2, C]O, CeO,
NH3 and HCN were the major gaseous products and functional groups
surveyed from the FTIR spectrums. The calcium-based content were
benefit for the N-pollutant control in co-pyrolysis process. The per-
centage of pyrolysis products (acetic acid, C7# (the number of carbon
atom ≦7), furfural, benzene derivatives, C7P (the number of carbon
atom > 7)) was classified. With the adding of sludge, the contents of
acetic acid, C7# and furfural were decreased and the content of C7P
increased gradually. The SiO2 and Al2O3 of sludge ash catalyzed the
opening of furfural products, and promoted the combination of C7# and
acetic acid. Both of them contributed to the growth on group C7P. This
researches investigated the products during co-pyrolysis, and provided
useful information for the background of co-pyrolysis process.
Acknowledgements
This work was supported by National Basic Research Program of
China (973 Program) (2013CB228100); the National Natural Science
Foundation of China (51676074); General Administration of Quality
Supervision, Inspection of Public Projects (20140159); Guangzhou
Science and Technology Plan (201604020022); the Guangdong Natural
Science Foundation (2015A030313227); the Key Laboratory of Efficient
and Clean Energy Utilization of Guangdong Higher Education
Institutes, South China University of Technology (KLB10004).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.enconman.2017.08.062.
References
[1] Zhang M, Yang C, Jing Y, Li J. Effect of energy grass on methane production and
heavy metal fractionation during anaerobic digestion of sewage sludge. Waste
Manage 2016;58:316–23.
[2] Manara P, Zabaniotou A. Towards sewage sludge based biofuels via thermochemical
conversion – a review. Renew Sustain Energy Rev 2012;16:2566–82.
[3] Huang Y-F, Shih C-H, Chiueh P-T, Lo S-L. Microwave co-pyrolysis of sewage sludge
and rice straw. Energy 2015;87:638–44.
[4] Alves M, Ponce GHSF, Silva MA, Ensinas AV. Surplus electricity production in su-
garcane mills using residual bagasse and straw as fuel. Energy 2015;91:751–7.
[5] Wang Yunpu LJ, Yuhuan Liu, Wenwen Zeng, Liu Yang, Ruan Rongsheng LC, Yiqin
Wan. Comprehensive utilization of bagasse: state of the art. Chin Agric Sci Bull
2010;26:370–5.
[6] Wang Zhiguo LM, Xiaolin Liu. Research of sewage sludge incineration technology.
Environ Protect Eng 2016;34:148–9.
[7] Anukam A, Mamphweli S, Reddy P, Meyer E, Okoh O. Pre-processing of sugarcane
bagasse for gasification in a downdraft biomass gasifier system: a comprehensive
review. Renew Sustain Energy Rev 2016;66:775–801.
[8] Martínez JD, Veses A, Mastral AM, Murillo R, Navarro MV, Puy N, et al. Co-pyrolysis
of biomass with waste tyres: upgrading of liquid bio-fuel. Fuel Process Technol
2014;119:263–71.
[9] Fang S, Yu Z, Lin Y, Hu S, Liao Y, Ma X. Thermogravimetric analysis of the co-
pyrolysis of paper sludge and municipal solid waste. Energy Convers Manage
2015;101:626–31.
[10] Fang S, Yu Z, Ma X, Lin Y, Lin Y, Chen L, et al. Co-pyrolysis characters between
combustible solid waste and paper mill sludge by TG-FTIR and Py-GC/MS. Energy
Convers Manage 2017;144:114–22.
[11] Lin Y, Liao Y, Yu Z, Fang S, Lin Y, Fan Y, et al. Co-pyrolysis kinetics of sewage
sludge and oil shale thermal decomposition using TGA–FTIR analysis. Energy
Convers Manage 2016;118:345–52.
[12] Alvarez J, Amutio M, Lopez G, Bilbao J, Olazar M. Fast co-pyrolysis of sewage
sludge and lignocellulosic biomass in a conical spouted bed reactor. Fuel
2015;159:810–8.
198
Energy Conversion and Management 151 (2017) 190–198
[13] Lin Y, Ma X, Yu Z, Cao Y. Investigation on thermochemical behavior of co-pyrolysis
between oil-palm solid wastes and paper sludge. Biores Technol 2014;166:444–50.
[14] Fang S, Yu Z, Lin Y, Lin Y, Fan Y, Liao Y, et al. Effects of additives on the co-
pyrolysis of municipal solid waste and paper sludge by using thermogravimetric
analysis. Bioresour Technol 2016;209:265–72.
[15] David GF, Perez VH, Rodriguez Justo O, Garcia-Perez M. Effect of acid additives on
sugarcane bagasse pyrolysis: production of high yields of sugars. Bioresour Technol
2016;223:74–83.
[16] Dewangan A, Pradhan D, Singh RK. Co-pyrolysis of sugarcane bagasse and low-
density polyethylene: influence of plastic on pyrolysis product yield. Fuel
2016;185:508–16.
[17] Henkel C, Muley PD, Abdollahi KK, Marculescu C, Boldor D. Pyrolysis of energy
cane bagasse and invasive Chinese tallow tree (Triadica sebifera L.) biomass in an
inductively heated reactor. Energy Convers Manage 2016;109:175–83.
[18] Hua M-Y, Li B-X. Co-pyrolysis characteristics of the sugarcane bagasse and
Enteromorpha prolifera. Energy Convers Manage 2016;120:238–46.
[19] Mendes FL, Ximenes VL, de Almeida MBB, Azevedo DA, Tessarolo NS, de Rezende
Pinho A. Catalytic pyrolysis of sugarcane bagasse and pinewood in a pilot scale unit.
J Anal Appl Pyrol 2016;122:395–404.
[20] Nytoft HPJAJB-KH. Coal-generated oil: source rock evaluation and petroleum
geochemistry of the Lulita oil field, Danish North Sea. J Pet Geol 2008;23:55–90.
[21] Li Qingzhao LB, Changsui ZHAO, Weifang WU. Chemical structure analysis of coal
char surface based on fourier-transform infrared spectrometer. Proc CSEE
2011;31:46–52.
[22] Yang J, Chen H, Zhao W, Zhou J. TG–FTIR-MS study of pyrolysis products evolving
from peat. J Anal Appl Pyrol 2016;117:296–309.
[23] Fang S, Yu Z, Lin Y, Lin Y, Fan Y, Liao Y, et al. A study on experimental char-
acteristic of co-pyrolysis of municipal solid waste and paper mill sludge with ad-
ditives. Appl Therm Eng 2017;111:292–300.
[24] Cheng Hui YJ, Meiqin Yao, Guangwen Xu. Mechanism analysis of lignin slow
pyrolysis. CIESC J 2013;64:1757–65.
[25] Ke Yikan DH. Chemical analysis manual. In: Ren Huimin TH, editor. Chemical
analysis manual (spectroanalysis). Beijing: Chemical Industry Press; 1998.
[26] Tian L, Shen B, Xu H, Li F, Wang Y, Singh S. Thermal behavior of waste tea pyrolysis
by TG-FTIR analysis. Energy 2016;103:533–42.
[27] Oluwadayo TH, Sonibare O, Foley Stephen F. Structural characterization of
Nigerian coals by X-ray diffraction, Raman and FTIR spectroscopy. Energy
2010;35:5347–53.
[28] Shewa WA, Lalman JA, Chaganti SR, Heath DD. Electricity production from lignin
photocatalytic degradation byproducts. Energy 2016;111:774–84.
[29] Chen W-H, Kuo P-C. Isothermal torrefaction kinetics of hemicellulose, cellulose,
lignin and xylan using thermogravimetric analysis. Energy 2011;36:6451–60.
[30] Yuan TQ, Sun SN, Xu F, Sun RC. Characterization of lignin structures and lignin-
carbohydrate complex (LCC) linkages by quantitative 13C and 2D HSQC NMR
spectroscopy. J Agric Food Chem 2011;59:10604–14.
[31] Wang X, Zhao B, Yang X. Co-pyrolysis of microalgae and sewage sludge: biocrude
assessment and char yield prediction. Energy Convers Manage 2016;117:326–34.
[32] Zhu L, Cheung CS, Huang Z. Impact of chemical structure of individual fatty acid
esters on combustion and emission characteristics of diesel engine. Energy
2016;107:305–20.
[33] Lin-ling Li RZ, Tong Dong-mei, Hu∗ Chang-wei. Fractional pyrolysis of algae and
model compounds. Chin J Chem Phys 2015;28:525–32.
[34] Zhang J, Chen T, Wu J, Wu J. Multi-Gaussian-DAEM-reaction model for thermal
decompositions of cellulose, hemicellulose and lignin: comparison of N(2) and CO
(2) atmosphere. Biores Technol 2014;166:87–95.
[35] Wang S, Wang K, Liu Q, Gu Y, Luo Z, Cen K, et al. Comparison of the pyrolysis
behavior of lignins from different tree species. Biotechnol Adv 2009;27:562–7.
[36] Peng X, Ma X, Lin Y, Guo Z, Hu S, Ning X, et al. Co-pyrolysis between microalgae
and textile dyeing sludge by TG–FTIR: kinetics and products. Energy Convers
Manage 2015;100:391–402.
[37] Shoji M, Yamamoto T, Tanno S, Aoki H, Miura T. Modeling study of homogeneous
NO and N2O formation from oxidation of HCN in a flow reactor. Energy
2005;30:337–45.
[38] Yuan S, Zhou Z-J, Li J, Chen X-L, Wang F-C. HCN and NH3 (NOx precursors) re-
leased under rapid pyrolysis of biomass/coal blends. J Anal Appl Pyrol
2011;92:463–9.
[39] Guo X, Zhang W, Wang L, Hao J. Comparative study of nitrogen migration among
the products from catalytic pyrolysis and gasification of waste rigid polyurethane
foam. J Anal Appl Pyrol 2016;120:144–53.
[40] Lou Rui Ws-b, Hao-liang Dong, Gao-jin Lu. Fast pyrolysis of enzymatic/mild acid-
olysis lignin from moso bamboo. J Fuel Chem Technol 2015;43:42–7.
[41] Pattiya A, Sukkasi S, Goodwin V. Fast pyrolysis of sugarcane and cassava residues in
a free-fall reactor. Energy 2012;44:1067–77.
[42] Shen DK, Gu S. The mechanism for thermal decomposition of cellulose and its main
products. Biores Technol 2009;100:6496–504.
[43] George JNC, Huber W, Barrett Christopher J, Dumesic∗ James A. Production of
liquid alkanes by aqueous-phase processing of biomass-derived carbohydrates.
Science 2005;308:1446–9.
[44] Alvarez J, Amutio M, Lopez G, Barbarias I, Bilbao J, Olazar M. Sewage sludge va-
lorization by flash pyrolysis in a conical spouted bed reactor. Chem Eng J
2015;273:173–83.
[45] Miandad R, Barakat MA, Aburiazaiza AS, Rehan M, Ismail IMI, Nizami AS. Effect of
plastic waste types on pyrolysis liquid oil. Int Biodeterior Biodegrad
2017;119:239–52.