Bioresource Technology 221 (2016) 534–540

Contents lists available at ScienceDirect

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

Kinetics of coffee industrial residue pyrolysis using distributed

activation energy model and components separation of bio-oil by
sequencing temperature-raising pyrolysis
Nanwei Chen a,b, Jie Ren a, Ziwei Ye a, Qizhi Xu a, Jingyong Liu a, Shuiyu Sun a,b,⇑
a
School of Environmental Science and Engineering, Guangdong University of Technology, Guangzhou 510006, PR China
b
Guangdong Polytechnic of Environmental Protection Engineering, Foshan 528216, PR China

h i g h l i g h t s

DAEM is used to study the kinetics of coffee industrial residue.

Sequencing temperature-raising pyrolysis (new method) is used.

A solution of components separation of bio-oil was put forward.

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

Article history: This study was carried out to investigate the kinetics of coffee industrial residue (CIR) pyrolysis, the effect
Received 11 August 2016 of pyrolysis factors on yield of bio-oil component and components separation of bio-oil. The kinetics of
Received in revised form 10 September CIR pyrolysis was analyzed using distributed activation energy model (DAEM), based on the experiments
2016
in thermogravimetric analyzer (TGA), and it indicated that the average of activation energy (E) is
Accepted 13 September 2016
Available online 15 September 2016
187.86 kJmol1. The bio-oils were prepared from CIR pyrolysis in vacuum tube furnace, and its compo-
nents were determined by gas chromatography/mass spectrometry (GC–MS). Among pyrolysis factors,
pyrolysis temperature is the most influential factor on components yield of bio-oil, directly concerned
Keywords:
Coffee industrial residue
with the volatilization and yield of components (palmitic acid, linoleic acid, oleic acid, octadecanoic acid
Pyrolysis and caffeine). Furthermore, a new method (sequencing temperature-raising pyrolysis) was put forward
Kinetics and applied to the components separation of bio-oil. Based on experiments, a solution of components
Bio-oil separation of bio-oil was come out.
Components separation Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction average annual production of 8 million tons in the world (ICO,

Biomass is a clean and renewable energy source leading to envi-
ronmental, technical and economical benefits. With the concern of
environmental protection, the energetic recovery of biomass via
pyrolysis, gasification or combustion has attracted increasing
worldwide interest. The pyrolysis is a thermal decomposition pro-
cess producing char with ash content, volatiles and low molecular
weight gases. Without char and volatiles oxidation or less, it’s the
better way to recycle energy from biomass (e.g. microalgae, wood,
straw and coffee waste) (White et al., 2011; Bridgwater, 2012;
Collard and Blin, 2014).
Coffee is considered as the second largest traded commodity in
the world. Coffee plants are cultivated in over 70 countries with an
⇑ Corresponding author at: School of Environmental Science and Engineering,
Guangdong University of Technology, Guangzhou 510006, PR China.
E-mail address: sysun@gdut.edu.cn (S. Sun).
2014). Coffee industrial residue (CIR) is the by-product from
instant coffee manufacturing process, left behind at the factory.
By the statistics, instant coffee production consumes around 20%
of the global production of green coffee beans (Gonzalez et al.,
2013) and for each kilogram of instant coffee produced 0.91 kg of
CIR (Silva et al., 1998). In addition, the CIR contains great amounts
of organic compounds (i.e. fatty acids, lignin, cellulose, hemi-
cellulose and other polysaccharides) (Pujol et al., 2013), while its
comprehensive utilization has become a new study in recent years.
For example, CIR has been investigated for biodiesel production
(Gonzalez et al., 2013; Kwon et al., 2013), as production of acti-
vated carbon (Tsai et al., 2012; Ma and Ouyang, 2013; Tehrani
et al., 2015), as source of sugars (Passos et al., 2014) or raw mate-
rial of fatty acids (Cruz et al., 2014).
Thermogravimetric analysis is one of the most common tech-
niques used to investigate the thermal behavior of small samples

http://dx.doi.org/10.1016/j.biortech.2016.09.062
0960-8524/Ó 2016 Elsevier Ltd. All rights reserved.
 N. Chen et al. / Bioresource Technology 221 (2016) 534–540 535

(Zhao et al., 2012). Thermogravimetry of CIR pyrolysis has been determined according to DD CEN/TS 14775: 2004, which heats
studied by several researchers (Skreiberg et al., 2011; Bok et al., solid samples to 550 ± 10 °C in muffle furnace with temperature
2012; Li et al., 2014), but the kinetics of CIR pyrolysis has not been programming for determination of ash content. The volatile con-
investigated. Kinetics is a theoretical analysis to obtain reaction tent was analyzed according to DD CEN/TS 15148: 2005, which
laws (e.g. activation energy and frequency factor), based on ther- heats solid samples at 900 ± 10 °C for 7 min in muffle furnace for
mogravimetric analysis. Numerous kinetic models have been determination of volatile content. The fixed carbon was calculated
developed, such as single-step global reaction model, multiple- using Eq. (1).
step model, semi-global model, and distributed activation energy
FC ¼ 100%  V  A  M ð1Þ
model (DAEM) (Soria-Verdugo et al., 2013). Among these models,
DAEM is considered as an accurate and versatile approach to model where FC, V, A and M are the content (wt.%) of fixed carbon, volatile,
the pyrolysis process (Cai et al., 2014). Therefore, DAEM was used ash and moisture, respectively.
to study the kinetics of CIR pyrolysis.
Bio-oil is one of products from biomass pyrolysis, and its yield 2.2.2. Ultimate analysis
and components are influenced by pyrolysis condition The C, H, N and S contents in the sample were measured using
(Bridgwater, 2012). The conventional separation technology of an elemental analyzer (Perkin Elmer Series II 2400, USA). The O
bio-oil components includes solvent extraction, column chro- content was calculated using Eq. (2).
matography and distillation. Solvent extraction is a method which
uses specific solvent to dissolve target components of a mixture of O ¼ 100%  C  H  N  S  A  M ð2Þ
components. Column chromatography commonly utilizes silica gel
where O, C, H, N, S, A and M are the content (wt.%) of O, C, H, N, S, ash
and aluminum oxide as the stationary phase, where the mobile
and moisture, respectively.
phase is selected according to the polarity of bio-oil components.
Distillation separates bio-oil components according to their differ-
ent volatilities (Kim, 2015). In this study, a new method of bio-oil 2.2.3. Heating value
components separation (sequencing temperature-raising pyroly- The higher heating value (HHV) was determined by an
sis) is put forward and used. isothermal-jacket bomb calorimeter (SHR-15, China).
This paper presents a study about kinetics of CIR pyrolysis and
components separation of bio-oil. First of all, the pyrolysis of 2.3. Thermogravimetric analysis
grounded CIR was carried out in thermogravimetric analyzer at
nitrogen atmosphere, with different heating rates. The thermo- The experiments were carried out in a thermalgravimetric ana-
gravimetric analysis and kinetic analysis using DAEM of CIR pyrol- lyzer (Henven HCT-3, China). Pyrolysis was performed under a car-
ysis were studied. Moreover, the bio-oils were prepared from CIR rier gas flow (nitrogen) of 30 mL min1, with a sample mass of
pyrolysis in vacuum tube furnace on different pyrolysis conditions. approximately 10 mg, which was heated up to 800 °C at heating
The effects of pyrolysis factor (pyrolysis temperature, time, pres- rates of 5, 10, 15, 20, 25 and 30 °Cmin1.
sure and heating rate) on yield of bio-oil component were
researched by gas chromatography/mass spectrometry (GC–MS). 2.4. Kinetic analysis using DAEM
Finally, the bio-oil components were separated by sequencing
temperature-raising pyrolysis, and the absolute yield of compo- The distributed activation energy model (DAEM) is originally
nents in separation liquid were estimated. Those results would developed by Vand (1943). It has been widely applied to analyze
make a contribution on the understanding of CIR pyrolysis, espe- the complex reactions in the pyrolysis of biomass (Wang et al.,
cially the utilization of bio-oil. 2008; Shen et al., 2011), which assumes that a number of parallel,
irreversible and first-order reactions with different energies occur
simultaneously.
2. Materials and methods
When the model is used to analyze biomass, the DAEM is
expressed as (Sonobe and Worasuwannarak, 2008):
2.1. Materials
Z 1  Z t 
V E
1 ¼ exp k0 eRT dt f ðEÞdE ð3Þ
The coffee industrial residue (CIR) used for this study was col- V 0 0
lected from a coffee processing plant in Dongguan, Guangdong pro-
vince, China. The original sample was oven-dried at 105 °C for 24 h where V is the total volatile content at temperature T, V⁄ is the effec-
and then became the sample of experiments in vacuum tube fur- tive volatile content of the biomass, t is against time, E is the activa-
nace. Dried CIR was grounded with a micromill (Yili QE-300, China) tion energy, f (E) is the normalized distribution curve of the
to pass through a 150 lm sieve, taken as the sample for character- activation energy of many reactions, and k0 is the frequency factor
ization and thermalgravimetric analyzer. According to proximate corresponding to the E value, R is gas constant (8.314 J mol1 K1).
analysis, CIR (dry basis) includes 6.98 wt.% of moisture, 64.94 wt. Eq. (3) was simplified to Eq. (4) (Miura, 1995; Miura and Maki,
% of volatile, 21.03 wt.% of fixed carbon and 7.05 wt.% of ash. The 1998).
result of ultimate analysis showed that CIR (dry basis) includes Z 1 Z ES
V
46.62 wt.% of C, 8.87 wt.% of H, 23.90 wt.% of O, 5.64 wt.% of N ¼1 f ðEÞdE ¼ f ðEÞdE ð4Þ
V ES 0
and 0.94 wt.% of S. And the higher heating value of CIR (dry basis)
is 21.3 MJ kg1. In this simplified model, the Arrhenius equation can be
described as follows (Andon et al., 2002; Munir et al., 2009):
   
2.2. Characterization of materials a k0 R E 1
ln ¼ ln þ 0:6075   ð5Þ
T2 E R T
2.2.1. Proximate analysis
The moisture analysis was conducted according to DD CEN/TS where a is the heating rate of pyrolysis, k0 is the frequency factor.
14774-3: 2004, which heats solid samples at 150 ± 2 °C in air dry Using Eq. (5), we can estimate both E and k0 from the Arrhenius plot
oven for determination of moisture content. The ash content was of ln(a/T2) vs. 1/T at the selected V/V⁄ values for different a values.
536 N. Chen et al. / Bioresource Technology 221 (2016) 534–540
We can determine the activation energies from the Arrhenius plots
at different levels of V/V⁄.
2.5. Preparation of bio-oil from CIR pyrolysis in vacuum tube furnace
The experimental setup is shown in Fig. 1. The tube furnace (SK-
G06123K, China) consists of a 1 m long quartz tube (diameter
60 mm), heated by well-insulated, controllable heating elements.
The quartz tube is connected to condensation pipes, bottles and a
circulating water vacuum pump. The circulating water vacuum
pump removed the organic vapors and other gas products from
the reactor through the condensation pipes. The condensable
vapors were then condensed in the vacuum traps and recovered
as liquid (bio-oil), but non-condensable gas was absorbed by alka-
line (NaOH) liquor in the third bottle or escaped. A control program
was used to control the heating rate, final pyrolysis temperature
and pyrolysis time, and the valve before circulating water vacuum
pump was used to control pyrolysis pressure.
The experiments were conducted in batch runs with variable
temperatures (300–500 °C), heating rates (5–15 °Cmin1), pyroly-
sis times (20–60 min) and pressures (30–50 kPaabs). Once the
quartz tube was loaded (quartz boats with 40 g of dried CIR), heat-
ing started after the quartz tube was sealed and evacuated to the
desired operating pressure. The quartz tube was natural cooling
after control program finished. The setup was stopped cooling
when the control program was over, but under vacuum until the
sample temperature reached low temperature (approximately
100 °C). Then the bottle with bio-oil were removed and weighed.
As the empty quartz boats and bottle had been weighed, the
weights of char and bio-oil were obtained. At last, the yields of
char, bio-oil and non-condensable gas were estimated using Eqs.
(6)–(8).
mS
YS ¼  100% ð6Þ
m0
mL
YL ¼  100% ð7Þ
m0
mG m0  mS  mL
YG ¼  100% ¼  100% ð8Þ
m0 m0
where YS, YL and YG are the yield (wt.%) of char, bio-oil and non-
condensable gas, respectively; m0, mS, mL and mG are the weight
(g) of sample, char, bio-oil and non-condensable gas, respectively.
Fig. 1. Experimental setup.
2.6. Characterization of bio-oil
Moisture of bio-oil was measured using infrared moisture
meter (Guanya SFY-20A, China). Components in the bio-oil were
analyzed by gas chromatography/mass spectrometry (GC/MS:
Agilent 7890A/5975C, USA). A HP-5MS capillary column
(30 m  250 lm  0.25 lm) was employed to separate the
components. Bio-oil without moisture was diluted in acetone and
passed through the organic filter membrane (aperture size
0.22 lm) using an injector, then 0.2 lL sample solution was
injected into column at 250 °C, with a split ratio of 10:1. Helium
was used as carrier gas at flow rate of 20 mL min1. Oven temper-
ature was programmed from 100 to 240 °C at 10 °Cmin1, held at
initial and final temperatures for 2 min. The detector and ion
source temperature were 150 °C and 230 °C, respectively. Absolute
yield of component in bio-oil was estimated using Eq. (9).
Y A ¼ 100  Y R  ðY B  M B Þ ð9Þ
where YA is the absolute yield (g/100 g CIR) of component from CIR
pyrolysis, YR is the relative yield (wt.%) of component in bio-oil
without moisture, YB is the relative yield (wt.%) of bio-oil and MB
is the moisture (wt.%) of bio-oil.
3. Results and discussion
3.1. Thermogravimetric analysis
Fig. 2 shows the thermogravimetry (TG) and derivative thermo-
gravimetry (DTG) curves of CIR pyrolysis at different heating rates.
With heating rate increasing, TG curve shifted to higher tempera-
ture and peak of DTG curve became bigger. Below 150 °C, there
was a small weight loss that represents the evaporation of mois-
ture, corresponding to a peak in DTG curves. The devolatilization
of CIR starts at above 150 °C and it is completed by 600 °C. For
example, at the volatile content conversion rate (V/V⁄) of 0.5, the
pyrolysis temperatures of CIR at the heating rate of 5, 10, 15, 20,
25 and 30 °Cmin1 were 329, 342, 349, 353, 358 and 362 °C,
respectively. When V/V⁄ changed from 0.8 to 0.2, the temperature
of CIR pyrolysis at heating rate of 5–30 °Cmin1 increased from
288 to 420 °C. It was characterized by a major mass loss (approx-
imately 65 wt.% of total weight) and only one syncretic peak was
observed in DTG curves. After 600 °C, the loss of weight was very
slow, while the carbonization of pyrolysis residues was underway.
 N. Chen et al. / Bioresource Technology 221 (2016) 534–540
Fig. 2. TG and DTG curves of CIR pyrolysis under 30 mLmin1 nitrogen flow: (a)
TG; (b) DTG.
3.2. Kinetic analysis
The kinetic analysis using DAEM was to study the main stage
(150–600 °C) of CIR pyrolysis. Because DAEM gives an excellent
fit to the pyrolytic mass loss curves, it is plausible that the thermal
decomposition of CIR under inert atmosphere probably undergoes
a set of single (first-order) reactions. It also needs to be noted that
DAEM reflects the E distribution of the whole pyrolysis process in
terms of consecutive mass conversion in spite of heating rate (Li
et al., 2013). The necessary parameters in Eq. (5) were calculated,
based on the data of TG curve at different heating rates (5, 10,
15, 20, 25 and 30 °Cmin1). The Arrhenius plot of ln(a/T2) vs. 1/T
was performed in Fig. 3, which was to obtain E at different levels
of V/V⁄ (0.2–0.8), and the results are shown in Table 1.
As Table 1 showed, all plots had fairly high linear correlation
coefficients (R2) greater than 0.97, which meant that the linear fit-
ting equation more accorded with rule of weight losing during CIR
pyrolysis. The values of k0 changed greatly (from 2.93  1025 to
2.39  1029 min1) with E values at different mass conversion.
And, E values changed from 163.42 to 239.49 kJmol1, which
average was 187.86 kJmol1. Sonobe and Worasuwannarak
(2008) have used DAEM to study several agricultural residues
(rice straw, rice husk, corncob and cellulose) pyrolysis, and found
E spread from 120 to 250 kJmol1. That is to say, the energy
requirement for CIR pyrolysis is closed to other agricultural resi-
due pyrolysis.
537
Fig. 3. Arrhenius plot of ln(a/T2) vs. 1/T at selected V/V⁄ values for CIR.
3.3. Effect of pyrolysis factors on yield of bio-oil component
3.3.1. Effect of pyrolysis temperature on yield of bio-oil component
The experiments in vacuum tube furnace were carried out at
different pyrolysis temperatures (300, 400 and 500 °C), with the
pyrolysis pressure at 10 kPaabs. The sample (40 g) was heated at
10 °Cmin1. Once the desired temperature was reached, the sam-
ple was kept at this pyrolysis temperature for 60 min.
Table 2 presents the absolute yield of component of bio-oils from
CIR pyrolysis at different temperatures. All those bio-oils contained
palmitic acid, linoleic acid, oleic acid, octadecanamide, hexadeca-
nenitrile, methyl palmitate and caffeine. As temperature increased,
octadecanoic acid, pentadecane, tetradecane and heptadecane grad-
ually came out, the number of palmitic acid, linoleic acid, oleic acid
and octadecanoic acid increased. However, the bio-oil at 300 °C did
not have octadecanoic acid, because the temperature was not high
enough for making it volatilized. The kind and yield of alkane (pen-
tadecane, tetradecane and heptadecane) in bio-oil turned more with
temperature increasing, while pyrolysis reaction was enhanced. Caf-
feine is the most special component in bio-oil, as it turned less with
temperature increasing. This might be because the temperature was
not low enough for its condensing. In short, pyrolysis temperature is
greatly concerned with the volatilization and yields of bio-oil com-
ponent, and can be used to the separation of components.
3.3.2. Effect of pyrolysis time on yield of bio-oil component
In experiments using vacuum tube furnace, the pyrolysis pres-
sure was hold at 10 kPaabs, and the heating rate was 10 °Cmin1.
Samples (40 g) were kept at 400 °C for 20, 40 and 60 min,
respectively.
Table 2 shows the absolute yield of component of bio-oils from
CIR pyrolysis for different pyrolysis times. The kind of component
of bio-oils from CIR pyrolysis for different pyrolysis times were the
same. With pyrolysis time extended, the number of palmitic acid,
linoleic acid, oleic acid and octadecanoic acid increased. For exam-
ple, the number of palmitic acid increased from 15.61 to
17.69 g/100 g CIR, when pyrolysis time extended from 20 to
60 min. The longer pyrolysis time is, the more heat and mass trans-
fers happen, such as evaporation of water and escape of volatiles,
then water and volatiles became bio-oil after cooling. As a result,
pyrolysis time has less influence on the yield of bio-oil compo-
nents, but should be long enough.
3.3.3. Effect of heating rate on yield of bio-oil component
The experiments were carried out by a vacuum tube furnace.
Dried CIR (40 g) were heated from normal temperature to 400 °C
for 60 min, with various heating rates (5, 10 and 15 °Cmin1),
under a pressure of 10 kPaabs.
As shown in Table 2, the kind of component of bio-oils from CIR
pyrolysis with different heating rates were the same, but the bio-
oil with heating rate of 5 °Cmin1 was quite different on yield of
538 N. Chen et al. / Bioresource Technology 221 (2016) 534–540

Table 1
Activation energies of CIR pyrolysis at different V/V⁄ values based on different heating rates (5, 10, 15, 20, 25 and 30 Kmin1).

V/V⁄ Linear fitting equation R2 k0 (min1) E (kJmol1)

29
0.2 Y = 28806X + 57.985 0.9816 2.39  10 239.49
0.3 Y = 22603X + 50.267 0.9933 8.34  1025 187.92
0.4 Y = 22053X + 50.374 0.9961 9.05  1025 183.35
0.5 Y = 22213X + 51.291 0.9965 2.28  1026 184.68
0.6 Y = 21806X + 51.264 0.9956 2.18  1026 181.30
0.7 Y = 21029X + 50.714 0.9889 1.21  1026 174.84
0.8 Y = 19656X + 49.360 0.9762 2.93  1025 163.42

In the linear fitting equation, Y represents ln(a/T2), X represents 1/T.

Table 2
Absolute yield of component of bio-oils on different pyrolysis conditions.

No. Chemical compound Absolute yield (g/100 g CIR)

300 °C 400 °C 500 °C 50 kPaabs 40 kPaabs 30 kPaabs 5 °Cmin1 10 °Cmin1 15 °Cmin1 20 min 40 min 60 min
1 1,2-Benzenediol 0.95 0.97 0.88 1.05 1.09 0.94
2 Tetradecane 0.72
3 Pentadecane 1.22 1.58 0.66 0.84 0.43 1.30 0.86 0.89 0.89 1.18 0.79
4 Heptadecane 0.55
5 Caffeine 2.85 1.84 1.65 0.55 1.60 1.29 1.50 1.74 1.73 1.25 1.60 1.45
6 Hexadecanenitrile 0.19 0.98 1.27 0.95 0.74 0.70 1.42 0.91 0.99 0.93 1.15 0.82
7 Methyl palmitate 0.87 0.90 0.89 0.67 0.66 0.55 1.39 0.83 0.91 0.89 1.03 0.75
8 Palmitic acid 4.83 16.91 16.26 12.51 15.46 17.18 15.90 16.58 16.56 15.61 15.84 17.69
9 Linoleic acid 0.65 4.04 3.91 3.08 3.48 4.03 3.80 3.96 3.95 3.73 3.78 4.23
10 Oleic acid 1.39 6.21 6.30 4.80 6.43 8.03 5.84 6.09 6.08 5.74 5.82 6.50
11 Octadecanoic acid 2.71 2.76 2.13 2.28 3.10 2.55 2.66 2.66 2.51 2.54 2.84
12 Octadecanamide 0.45 0.79 0.92 0.55 1.60 1.29

The other pyrolysis factors of 300 °C, 400 °C and 500 °C: ‘‘pressure = 10 kPaabs, pyrolysis time = 60 min and heating rate = 10 °Cmin1”; the other pyrolysis factors of 50 kPaabs,
40 kPaabs and 30 kPaabs: ‘‘pyrolysis temperature = 400 °C, pyrolysis time = 60 min and heating rate = 10 °Cmin1”; the other pyrolysis factors of 5 °Cmin1, 10 °Cmin1 and
15 °Cmin1: ‘‘pyrolysis temperature = 400 °C, pressure = 10 kPaabs and pyrolysis time = 60 min”; the other pyrolysis factors of 20 min, 40 min and 60 min: ‘‘pyrolysis tem-
perature = 400 °C, pressure = 10 kPaabs and heating rate = 10 °Cmin1”.

components. The reason might be that the temperature-rising per-
iod of 5 °Cmin1 was quite long, slowing down volatilization
speed of components and intensifying secondary reactions during
pyrolysis, then increased the amount of pentadecane, hexadeca-
nenitrile and methyl palmitate. Therefore, if the secondary reac-
tions were to be decreased, heating rate should be controlled
more than 10 °Cmin1.
3.3.4. Effect of pyrolysis pressure on yield of bio-oil component
The experiments were carried out in the vacuum tube furnace.
Dried CIR (40 g) was heated, with 10 °Cmin1 heating rate, from
normal temperature to 400 °C for 60 min, under various pyrolysis
pressure (30, 40 and 50 kPaabs).
Table 2 illustrates the absolute yield of component of bio-oils
from CIR pyrolysis as a function of pressure. As pyrolysis pressure
reduced from 50 kPaabs to 30 kPaabs, the absolute yield of palmitic
acid, linoleic acid, oleic acid and octadecanoic acid increased, hex-
adecanenitrile and methyl palmitate decreased. This is an intuitive
result, since it is commonly accepted that at higher pressures,
there will be more recondensation reactions which would lead to
more of the vapors becoming trapped on the char. At the lower
pressure, volatiles extract from tube faster, and there are fewer
reactions of volatiles bringing about more bio-oils (de Jongh
et al., 2011). For this reason, pyrolysis pressure should be lower
in order to increase the yield of main components and reduce
the yield of minor components caused by the reaction of volatiles.
3.4. Components separation of bio-oil by sequencing temperature-
raising pyrolysis
Illuminated by the results of Section 3.3.1, a new separation
method (sequencing temperature-raising pyrolysis) was put for-
ward and applied to components separation of bio-oil. This experi-
ment was also carried out in the vacuum tube furnace. A sample
(40 g CIR) was heated up to 150, 200, 250, 300, 350 and 400 °C, suc-
cessively and respectively. The heating rate was 10 °Cmin1, the
pressure was 30 kPaabs and each temperature was held 30 min.
After each pyrolysis, the separation liquid was transferred and a
new bottle was to be loaded. After the sequencing temperature-
raising pyrolysis, the bio-oil was divided into six parts of liquids
in different temperature ranges, including 0–150, 150–200,
200–250, 250–300,300–350 and 350–400 °C. The separation
liquids were weighted using electronic balance and determined
by GC–MS.
Table 3 shows the absolute yield of component of separation
liquids in different temperature ranges. The absolute yield of sep-
aration liquids were ordered: 250–300 °C separation liquid
(15.05 g/100 g CIR) > 300–350 °C separation liquid (9.63 g/100 g
CIR) > 200–250 °C separation liquid (6.35 g/100 g CIR) > 0–150 °C
separation liquid (4.23 g/100 g CIR) > 350–400 °C separation liquid
(4.10 g/100 g CIR) > 150–200 °C separation liquid (0.68 g/100 g
CIR). There are several features of CIR sequencing temperature-
raising pyrolysis as follows:
(a) Each component has a devolatilization temperature range.
For instance, palmitic acid is devolatilized at 200–400 °C,
octadecanoic acid is devolatilized at 300–350 °C, linoleic
acid and oleic acid are devolatilized at 200–350 °C.
(b) Caffeine is the most special component in separation liquid.
The separation process of caffeine can be divided into two
parts: one is separated out with moisture below 200 °C,
the other is separated out with fatty acid (eg. palmitic acid)
at 200–250 °C.
(c) Four main components (palmitic acid, linoleic acid, oleic
acid, octadecanoic acid) have similar properties. The
devolatilization temperature ranges of main components
are superposed so that it seems impossible to separate them
one by one.
 N. Chen et al. / Bioresource Technology 221 (2016) 534–540 539

Table 3
Absolute yield of component of separation liquids in different temperature ranges.

No. Chemical compound Absolute yield (g/100 g CIR)

0–150 150–200 200–250 250–300 300–350 350–400
°C °C °C °C °C °C
1 Caffeine 0.34 0.68 0.58 2.07
2 Palmitic acid 3.97 7.42 6.85 1.02
3 Linoleic acid 1.35 0.36
4 Octadecanoic acid 0.52
5 Oleic acid 0.37 4.55 0.89
6 Methyl palmitate 0.04 0.39 0.50 0.23
7 1,2-Benzenediol 0.04 0.62
8 Hexadecanenitrile 0.51 1.58
9 Pentadecane 1.27
10 Water 3.89
Subtotal 4.23 0.68 6.35 15.05 9.63 4.10

Fig. 4. Flow diagram of separation solution.

(d) Minor components (1,2-Benzenediol, methyl palmitate, hex- is accomplished at 200 °C for 30 min, 5 g aqueous solution of caf-
adecanenitrile, pentadecane) are gradually separated out as feine (1.0 g caffeine + 4.0 g water) and 3 g non-condensable gas
temperature increased from 200 to 400 °C. can be separated out. And then with vacuum pyrolysis of the front
pyrolysis char is accomplished at 400 °C for 30 min, 35 g mixed
3.5. A solution of components separation of bio-oil solution of fatty acid (2.7 g caffeine + 18.3 g palmitic acid + 2.7 g
linoleic acid + 2.6 g octadecanoic acid + 6.8 g oleic acid + 0.7 g
According to the results of Section 3.4, a solution of components methyl palmitate + 0.2 g 1,2-Benzenediol + 0.7 g hexadecaneni-
separation of bio-oil was put forward and implemented. The solu- trile + 0.3 g pentadecane) and 22 g non-condensable gas will be
tion is shown in Fig. 4. The experiment was also carried out in the separated out. At last, there are 35 g pyrolysis char left.
vacuum tube furnace. A sample (40 g CIR) was heated up to 200
and 400 °C, successively and respectively. The heating rate was
4. Conclusions
10 °Cmin1, the pressure was 30 kPaabs and each temperature
was held 30 min. The separation liquids were analyzed by GC/
In this study, pyrolysis of CIR was carried out in thermogravi-
MS, and the absolute yield of each component were obtained
metric analyzer and vacuum tube furnace, respectively. The kinet-
(Table 4). The results show that as vacuum pyrolysis of 100 g CIR
ics of CIR pyrolysis was analyzed using DAEM, and it indicated that
E values changed from 163.42 to 239.49 kJmol1 (average
Table 4 187.86 kJmol1). It tends to promote pyrolysis as temperature
Absolute yield of component of separation liquids in separation solution. increased, pyrolysis pressure decreased or pyrolysis time extended,
No. Chemical Absolute yield (g/100 g CIR)
while main components (palmitic acid, linoleic acid, oleic acid and
compound octadecanoic acid) increased. The sequencing temperature-raising
Aqueous solution of Mixed solution of fatty
pyrolysis is one of the effective ways to separate components of
caffeine (0–200 °C) acid (200–400 °C)
bio-oil, and a fraction of caffeine in CIR can be separated out.
1 Caffeine 1.01 2.65
2 Palmitic acid 18.26
3 Linoleic acid 2.71 Acknowledgements
4 Octadecanoic acid 2.52
5 Oleic acid 6.81
6 Methyl palmitate 0.66 This study was financially supported by the Science and Tech-
7 1,2-Benzenediol 0.16 nology Planning Project of Guangdong Province (No.
8 Hexadecanenitrile 0.63 2014B020216009), the Special Fund of Department of Environ-
9 Pentadecane 0.27 mental Protection of Guangdong Province (2016), the Special Fund
10 Water 3.99
of University Discipline Construction of Department of Education
Subtotal 5.00 34.67
of Guangdong Province (No. 2014KTSP022), the Major Achieve-
540 N. Chen et al. / Bioresource Technology 221 (2016) 534–540
ments Cultivating Foundation of Guangdong University of Technol-
ogy (No. 201206) and the National Natural Science Foundation of
China (No. 51308132).
References
Andon, N.L., Hollingworth, S., Koller, A., Greenland, A.J., Yates, J.R., Haynes, P.A.,
2002. Proteomic characterization of wheat amyloplasts using identification of
proteins by tandem mass spectrometry. Proteomics 2, 1156–1168.
Bok, J.P., Choi, H.S., Choi, Y.S., Park, H.C., Kim, S.J., 2012. Fast pyrolysis of coffee
grounds: characteristics of product yields and biocrude oil quality. Energy 47,
17–24.
Bridgwater, A.V., 2012. Review of fast pyrolysis of biomass and product upgrading.
Biomass Bioenergy 38, 68–94.
Cai, J., Wu, W., Liu, R., 2014. An overview of distributed activation energy model and
its application in the pyrolysis of lignocellulosic biomass. Renew. Sustain.
Energy Rev. 36, 236–246.
Collard, F., Blin, J., 2014. A review on pyrolysis of biomass constituents: mechanisms
and composition of the products obtained from the conversion of cellulose,
hemicelluloses and lignin. Renew. Sustain. Energy Rev. 38, 594–608.
Cruz, M.V., Paiva, A., Lisboa, P., Freitas, F., Alves, V.D., Simões, P., Barreiros, S., Reis,
M.A.M., 2014. Production of polyhydroxyalkanoates from coffee industrial
residues oil obtained by supercritical fluid extraction technology. Bioresour.
Technol. 157, 360–363.
de Jongh, W.A., Carrier, M., Knoetze, J.H.H., 2011. Vacuum pyrolysis of intruder plant
biomasses. J. Anal. Appl. Pyrol. 92, 184–193.
Gonzalez, A.S., Plaza, M.G., Pis, J.J., Rubiera, F., Pevida, C., 2013. Post-combustion CO2
capture adsorbents from spent coffee grounds. Energy Procedia 37, 134–141.
ICO, 2014. Annual Review Available from: http://www.ico.org/news/annual-
review-2012-13-e.pdf.
Kim, J., 2015. Production, separation and applications of phenolic-rich bio-oil – a
review. Bioresour. Technol. 178, 90–98.
Kwon, E.E., Yi, H., Jeon, Y.J., 2013. Sequential co-production of biodiesel and
bioethanol with spent coffee grounds. Bioresour. Technol. 136, 475–480.
Li, L., Wang, G., Wang, S., Qin, S., 2013. Thermogravimetric and kinetic analysis of
energy crop Jerusalem artichoke using the distributed activation energy model.
J. Therm. Anal. Calorim. 114, 1183–1189.
Li, X., Strezov, V., Kan, T., 2014. Energy recovery potential analysis of spent coffee
grounds pyrolysis products. J. Anal. Appl. Pyrol. 110, 79–87.
Ma, X., Ouyang, F., 2013. Adsorption properties of biomass-based activated carbon
prepared with spent coffee grounds and pomelo skin by phosphoric acid
activation. Appl. Surf. Sci. 268, 566–570.
Miura, K., 1995. A new and simple method to estimate f(E) and k0(E) in the
distributed activation-energy model from three sets of experimental data.
Energy Fuels 9, 302–307.
Miura, K., Maki, T., 1998. A simple method for estimating f(E) and k0(E) in the
distributed activation energy model. Energy Fuels 12, 864–869.
Munir, S., Daood, S.S., Nimmo, W., Cunliffe, A.M., Gibbs, B.M., 2009. Thermal analysis
and devolatilization kinetics of cotton stalk, sugar cane bagasse and shea meal
under nitrogen and air atmospheres. Bioresour. Technol. 100, 1413–1418.
Passos, C.P., Moreira, A.S.P., Domingues, M.R.M., Evtuguin, D.V., Coimbra, M.A., 2014.
Sequential microwave superheated water extraction of mannans from spent
coffee grounds. Carbohydr. Polym. 103, 333–338.
Pujol, D., Liu, C., Gominho, J., Olivella, M.A., Fiol, N., Villaescusa, I., Pereira, H., 2013.
The chemical composition of exhausted coffee waste. Ind. Crops Prod. 50, 423–
429.
Shen, D.K., Gu, S., Jin, B., Fang, M.X., 2011. Thermal degradation mechanisms of
wood under inert and oxidative environments using DAEM methods. Bioresour.
Technol. 102, 2047–2052.
Silva, M.A., Nebra, S.A., Silva, M., Sanchez, C.G., 1998. The use of biomass residues in
the Brazilian soluble coffee industry. Biomass Bioenergy 14, 457–467.
Skreiberg, A., Skreiberg, O., Sandquist, J., Sorum, L., 2011. TGA and macro-TGA
characterisation of biomass fuels and fuel mixtures. Fuel 90, 2182–2197.
Sonobe, T., Worasuwannarak, N., 2008. Kinetic analyses of biomass pyrolysis using
the distributed activation energy model. Fuel 87, 414–421.
Soria-Verdugo, A., Garcia-Hernando, N., Garcia-Gutierrez, L.M., Ruiz-Rivas, U., 2013.
Analysis of biomass and sewage sludge devolatilization using the distributed
activation energy model. Energy Convers. Manage. 65, 239–244.
Tehrani, N.F., Aznar, J.S., Kiros, Y., 2015. Coffee extract residue for production of
ethanol and activated carbons. J. Clean. Prod. 91, 64–70.
Tsai, W., Liu, S., Hsieh, C., 2012. Preparation and fuel properties of biochars from the
pyrolysis of exhausted coffee residue. J. Anal. Appl. Pyrol. 93, 63–67.
Vand, V., 1943. A theory of the irreversible electrical resistance changes of metallic
films evaporated in vacuum. Proc. Phys. Soc. 55, 222–246.
Wang, G., Li, W., Li, B., Chen, H., 2008. TG study on pyrolysis of biomass and its three
components under syngas. Fuel 87, 552–558.
White, J.E., Catallo, W.J., Legendre, B.L., 2011. Biomass pyrolysis kinetics: a
comparative critical review with relevant agricultural residue case studies. J.
Anal. Appl. Pyrol. 91, 1–33.
Zhao, H., Cao, Y., Sit, S.P., Lineberry, Q., Pan, W., 2012. Thermal characteristics of
bitumen pyrolysis. J. Therm. Anal. Calorim. 107, 541–547.