Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

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

Catalytic pyrolysis of lignin using a two-stage fixed bed reactor

comprised of in-situ natural zeolite and ex-situ HZSM-5
Hyung Won Lee a , Young-Min Kim a,b , Jungho Jae c,d , Bong Hyun Sung e , Sang-Chul Jung f ,
Sang Chai Kim g , Jong-Ki Jeon h , Young-Kwon Park a,∗
a
School of Environmental Engineering, University of Seoul, Seoul 02504, Republic of Korea
b
Frontier Laboratories Ltd., 1-8-14, Saikon, Koriyama, Fukushima 963-8862, Japan
c
Clean Energy Research Center, Korea Institute of Science and Technology, Seoul 02792, Republic of Korea
d
Department of Clean Energy and Chemical Engineering, Korea University of Science and Technology, Daejeon 34113, Republic of Korea
e
Bionergy & Biochemical Research Center, Korea Research Institute of Bioscience & Biotechnology, Daejeon 34141, Republic of Korea
f
Department of Environmental Engineering, Sunchon National University, Suncheon 57922, Republic of Korea
g
Department of Environmental Education, Mokpo National University, Muan 58554, Republic of Korea
h
Department of Chemical Engineering, Kongju National University, Cheonan 31080, Republic of Korea

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

Article history: The two-stage catalytic pyrolysis of lignin over in-situ natural zeolite (NZ) and ex-situ HZSM-5 was exam-
Received 28 March 2016 ined using a tandem fixed bed reactor. The physicochemical properties of the catalysts, HZSM-5 and NZ,
Received in revised form 20 July 2016 were characterized by N2 adsorption-desorption and temperature programed desorption of ammonia.
Accepted 1 September 2016
The overall performance for the catalytic pyrolysis of lignin was evaluated by the comparing lignin con-
Available online 5 October 2016
version, aromatic formation, and amount of coke deposited from the two-stage catalytic pyrolysis with
those from a single-stage catalytic pyrolysis with ex-situ HZSM-5. Compared to the single-stage catalytic
Keywords:
pyrolysis, the two-stage catalytic pyrolysis produced a larger amount of aromatics with a smaller amount
Lignin
Catalytic pyrolysis
of coke due to the pre-catalytic effect of NZ. These positive effects caused by the use of the two-stage
Two-stage reactor catalyst were maximized by increasing the amounts of in-situ natural zeolite and the temperature of the
Natural zeolite ex-situ HZSM-5 catalyst bed to 600 ◦ C.
HZSM-5 © 2016 Elsevier B.V. All rights reserved.

1. Introduction biomass pyrolysis process, its actual commercialization is yet to be

Due to the intensive use of fossil fuels, energy shortages and
climate change have become urgent issues that need to be solved.
In this respect, biomass is being considered the only sustainable
source of carbon to replace fossil fuels. Biomass can be converted
to valuable energy, fuels, or chemicals via biological, thermal, or
chemical conversion techniques. Recent advances in biomass con-
version technologies have made it possible to produce fuels and
chemicals, which have an identical chemical structure to petroleum
products [1–5].
The pyrolysis of biomass, which is one of the thermal conver-
sion technologies for biomass conversion, has been investigated
widely and is usually aimed to maximize the yield of bio-oil because
bio-oil can be applied easily to the current petroleum infrastruc-
ture. Although many researchers have attempted to optimize the
∗ Corresponding author.
E-mail address: catalica@uos.ac.kr (Y.-K. Park).
achieved because of the low quality of bio-oil caused by its high oxy-
gen content and thermal instability. Therefore, many researchers
have examined the catalytic pyrolysis of biomass to upgrade the
bio-oil quality [6–10].
Lignocellulosic biomass consists of three main components of
hemicellulose, cellulose, and lignin and their distribution varies
according to the type of biomass. The lignocellulosic components
have a significant effect on the quality of bio-oil. Among the three
major components of biomass, lignin is the most difficult compo-
nent to convert to bio-oil, even via catalytic pyrolysis. Compared
to hemicellulose and cellulose, lignin produces a large amount of
char because of its complicated three dimensional structure. Nev-
ertheless, the importance of lignin utilization as an energy source
is being emphasized due to not only the higher portion of lignin in
biomass occupying 30% of carbon but also large amount of lignin
production as the byproduct of paper industry [11–13].
Research on the pyrolysis of lignin has increased steadily to
produce phenolic compounds from various lignins [14–18]. Espe-
cially, many researchers have attempted to upgrade the bio-oil

http://dx.doi.org/10.1016/j.jaap.2016.09.015
0165-2370/© 2016 Elsevier B.V. All rights reserved.
 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288
quality or produce aromatic chemicals via a cost-effective cat-
alytic pyrolysis process. Many researchers tested various kinds of
catalysts, such as zeolites [19,20] solid phosphoric acid, hydrotreat-
ing catalyst (Co/Mo/Al2 O3 ) [21], NaCl, permutite [22], amorphous
silica-alumina, silicalite [23], natural clinoptilolite [24], meso-
porous Y zeolite, and mesoporous MFI zeolite [25,26], for the
catalytic pyrolysis of lignin. These studies concluded that the struc-
ture and acidity of the catalyst are important factors for the efficient
deoxygenation reaction together with the operation conditions,
such as temperature, reaction time, etc.
In addition to the type of catalysts and process conditions
employed in catalyst pyrolysis, another important issue to consider
for successful catalytic pyrolysis is to minimize the level of catalyst
deactivation, which influences the lifetime of the catalyst and the
operation cost of the process. The catalyst deactivation during the
catalytic pyrolysis of lignin is more severe than those of hemicellu-
lose and cellulose due to the high thermal resistance of lignin and
low reactivity of lignin-derived phenolic compounds [27]. Some
studies examined the deactivation and regeneration of the zeo-
lite catalysts during the catalytic pyrolysis of lignin [28,29]. They
reported that the zeolite catalysts undergo the irreversible deacti-
vation due to structural changes during the reaction-regeneration
process. Therefore, more studies on finding an alternative method
to increase the catalyst lifetime will be required before the actual
catalytic pyrolysis of lignin can be commercialized.
Some researchers also attempted to find a reason for catalyst
deactivation during the catalytic pyrolysis of lignin. To et al. [30]
reported that the reaction of phenolic compounds over HZSM-
5. To maintain the catalytic activity without severe deactivation,
they suggested increasing the amount of catalyst because a large
amount of catalyst can provide sufficient acid sites for the conver-
sion of phenolic compounds. On the other hand, the use of a large
amount of catalyst causes an increase in the total cost of the cat-
alytic pyrolysis process. Therefore, a more cost-effective method to
reduce catalyst deactivation during the catalytic pyrolysis of lignin
needs to be developed. The use of a low-price additive catalyst
together with the expensive high- performance catalyst, such as
HZSM-5, can be the good option to decrease the total cost of the
catalytic pyrolysis process but it has few applications. One of the
possible catalysts is natural zeolite (NZ) because it is abundant and
inexpensive.
In this study, the two-stage catalytic pyrolysis of lignin was
investigated using a tandem fixed bed reactor system, in which
lignin is first pyrolyzed in the presence of low cost NZ catalyst in the
first reactor and the pyrolysis vapors are then upgraded to valuable
aromatic chemicals with HZSM-5 in a second catalytic reactor. The
synergy effects using natural zeolite as an additive catalyst on the
catalytic pyrolysis of lignin with HZSM-5 were evaluated based on
the aromatic formation efficiency and catalyst deactivation behav-
ior.
2. Materials and methods
2.1. Materials
A powder form of Kraft lignin, which was purchased from Sigma
Aldrich, was used as a sample. Table 1 lists the results for the
proximate and ultimate analysis of Kraft lignin. The ultimate and
proximate analyses were performed by using an elemental analyzer
(Flash EA 1113 series, CE Instruments) and a thermogravimetry
analyzer (Pyris 1, Perkin Elmer) according to the following method.
The lignin was heated from room temperature to 110 ◦ C under an
N2 flow and then this temperature was maintained for 0.5 h to
determine the moisture followed by heating to 900 ◦ C to determine
the volatile matter. After changing to an oxygen flow, the tem-
283
Fig. 1. Schematic diagram of fixed bed catalytic pyrolysis reactor.
perature was maintained at 800 ◦ C until the sample mass reached
constant value. The total mass change observed at this point was
regarded as the fixed carbon and the rest was regarded as the ash.
HZSM-5 (SiO2 /Al2 O3 : 50) and NZ (SiO2 /Al2 O3 : 5.4–7.3) were pur-
chased from Zeolyst and Rota Mining Corporation, respectively.
The catalysts were characterized using various techniques. N2 -
physisorption measurements of the catalysts were performed in
a BELSORP-mini II (BEL Japan, Inc.) to determine the pore volume
and the surface area. NH3 -temperature programmed desorption
(NH3 -TPD) was performed using a BELCAT-M (BEL Japan) instru-
ment equipped with a TCD (thermal conductivity detector) to
determine the total acidity, according to the procedure reported
in the previous literature [31]. Pyridine FT-IR was also performed
to characterize the distributions of Brønsted and Lewis acid sites
using Spectrum GX (Perkinelmer). The Brønsted and Lewis acid
sites of the catalysts were assigned to the peaks at 1545 cm−1 and
1439 cm−1 , respectively [32,33]. The microporous surface area and
volume of HZSM-5 and NZ were listed in Table S1. As shown in
Table S1, HZSM-5 showed much higher microporous surface and
volume than NZ. All the catalysts were calcined at 550 ◦ C for 3 h.
HZSM-5 was pelletized and sieved to make a uniform particle size
between 1.0 and 1.7 mm.
2.2. Catalytic pyrolysis using a fixed bed
The catalytic pyrolysis of lignin was performed using a two stage
fixed bed reactor system with a U-type quartz reaction tube (Fig. 1).
The system consisted of two individual furnaces with a reaction gas
controller, temperature controller, bio-oil condenser, and gas sam-
pling parts. Two furnaces were used for the pyrolysis of lignin (3 g)
at the 1st reactor and the catalytic upgrading of pyrolysis vapor
over a HZSM-5 catalyst (1 g) at the 2nd reactor (ex-situ), respec-
tively. The additive catalyst, NZ (3 g), was added to the 1st reactor
together with lignin (in-situ). To prevent the movement of HZSM-5
catalyst, both sides of the HZSM-5 catalyst were blocked by quartz
wool in the 2nd reactor. After loading the sample and catalysts into
the quartz tube, the system was purged with 50 mL/min of nitrogen
gas to eliminate the residual air in the system. After nitrogen purg-
ing, the temperature of each furnace was set to 500 ◦ C for the 1st
furnace and to 500 or 600 ◦ C for the 2nd furnace. During the sample
loading and furnace preheating time, the furnace was positioned on
the left side of the quartz tube to prevent any reaction during the
standby time. Once the furnace temperature and carrier gas flow
rate were stabilized, the in-situ catalytic pyrolysis of lignin over
NZ was initiated by sliding the 1st furnace (500 ◦ C) to the sample
position. The pyrolysis vapor emitted from the 1st reactor was also
284 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288

Table 1
Physicochemcial characteristics of kraft lignin.

wt% Ultimate analysis Proximate analysis

C H Oa N S Moisture Volatile matter Fixed carbon Ash

Kraft lignin 62.45 5.68 30.61 0.56 0.7 1.5 56.95 39.19 2.36
a
By difference.

upgraded by the catalytic reaction over the ex-situ HZSM-5 catalyst 60

Solid residue
located in the 2nd reactor. The final products were moved into the Oil
liquid condensing part. To collect the liquid products, a condenser 50 Gas
at −20 ◦ C was used and the gas products were collected by a Teflon
gas bag connected to the outlet of the liquid collection flask.
The yields of the liquid and solid products were determined by 40

Yield (wt%)
actual weight measurements after the reaction. The gas yield was
calculated by subtracting the summed yield of the liquid and solid 30
products from 100%.

20
2.3. Analysis of gas, oil, and coke

The oil dissolved in acetone and gas products obtained from 10

the catalytic pyrolysis of lignin were analyzed by gas chromatogra-
phy/mass spectrometry (GC/MS, Agilent 7890A/5975 inert XL) and 0
GC/TCD/FID (Young Lin ACME 6000), respectively. Detailed analysis nin (1g
)
(3 g)
Lig - NZ -NZ
conditions of GC/MS and GC/TCD/FID were shown in Table S2 and nin nin
Lig Lig
Table S3.
The amount and property of the deposited coke on the HZSM-5 Fig. 2. Yields of lignin and additive mixture pyrolysis products.
catalyst was measured by the non-isothermal calcination of coke in
air by thermogravimetric analysis (TGA) and evolved gas analysis- 50
mass spectrometry (EGA-MS), respectively. For TG analysis, 5 mg of Lignin
the coke deposited catalyst was heated from ambient temperature Lignin-NZ(1g)
to 600 ◦ C at a heating rate of 10 ◦ C/min under air atmosphere, and 40
Lignin-NZ(3g)

the coke amount was estimated by the weight difference between

Distribution (area%)

before and after TG analysis. The thermal property and relative

amount of coke was identified by comparing the temperature and 30
MS intensity of CO2 emitted from the catalyst during the non-
isothermal oxidative EGA-MS analysis of coke deposited catalyst
[34]. For this, 2 mg of the coke deposited catalyst was heated from 20

ambient temperature to 600 ◦ C at a heating rate of 10 ◦ C/min in the

pyrolyzer coupled with MS under air atmosphere, and the mass
spectrum of product vapors was monitored in real time. After the 10

experiment, thermal profile of CO2 was obtained by extracting the

ion thermogram of CO2 (m/z: 44) from total ion thermograms. All
0
the experiments and product analyses were performed more than ls s
ls
ho col es ols es
twice to confirm the reproducibility of the data, and the RSD values e no tec aia zen gen zat
ph Gu en Eu oly
Alk
yl
yroc a lk ylb y pyr
were within 3%. P
o xy
a
hea
v
eth er
Dim Oth

3. Results and discussion Fig. 3. Product distribution of lignin and additive mixture pyrolysis.

3.1. Effect of the additive catalyst in the 1st stage reactor on the
pyrolysis of lignin (in-situ)
Fig. 2 shows the yields of gas, oil, and solid residue obtained from
the thermal and in-situ catalytic pyrolysis of lignin (3 g) over NZ (1 g
and 3 g) at 500 ◦ C. The non-catalytic pyrolysis of lignin produced a
large amount of char (51.7 wt.%) together with oil (33.6 wt.%) and
gas (14.7 wt.%). The large amount of char formation in the lignin
pyrolysis was attributed to the thermal stability of lignin [27].
Lignin pyrolysis char has a cross-linked structure and is produced
by the stepwise process of a reaction of primary char with pyroly-
sis vapors [35]. In the presence of NZ catalyst, the yields of gas, oil
and solid residue were changed only slightly. Compared to the non-
catalytic pyrolysis, the solid residue yield decreased, while the gas
and oil yields increased. A decrease in the solid residue yield with
the NZ catalyst can be explained by the catalytic effect of the NZ
catalyst, which converts the initial pyrolysis intermediates to liq-
uid or gas rather than to solid residue. Ma et al. [23] also reported
that the use of a catalyst during the pyrolysis of lignin increases the
oil yield and decreases the char yield because the catalyst converts
the active pyrolysis intermediates to a liquid, instead of solid char.
The chemical composition of the liquid product (Fig. 3) was
changed using the NZ catalyst, indicating the acidic properties
of NZ as a catalyst. As shown in detailed chemical distribu-
tion of Table S4, the quantities of methoxylated chemicals,
such as 4-ethyl-2-methoxy-phenol, 2-methoxy-4-vinylphenol,
3,4-dimethoxytoluene, and eugenol, decreased, whereas alkylphe-
nols and pyrocatechols were increased using 1 g of NZ. This
means that the demethoxylation and dealkylation of methoxy-
 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288
HZSM-5
NZ
Intensity (a.u.)
100 200 300 400
o nin
Temperature ( C)
Fig. 4. NH3 -TPD curves of catalysts.
lated chemicals are the main reactions catalyzed by the NZ catalyst
during the catalytic pyrolysis of lignin. The amount of heavy
lignin pyrolyzates, such as 4-hydroxy-3-methoxy-benzeneacetic
acid, was also decreased using NZ.
The catalytic effect of NZ on the catalytic pyrolysis of lignin
was improved greatly by increasing the amount of NZ from 1 g
to 3 g. Using 3 g of NZ, the amounts of guaiacols and heavy lignin
pyrolyzates were decreased greatly, while the amount of alkylphe-
nols were increased significantly due to the larger number of acid
sites available for the catalytic reactions. On the other hand, aro-
matic hydrocarbon formation could not be achieved, even with
large NZ catalyst loadings. This indicates that the addition of NZ
catalyst has a limited catalytic effect due to the relatively weak
acidity of the NZ catalyst compared to that of commercial zeolites,
such as HZSM-5. As shown in Fig. 4, NZ showed only one desorp-
tion peak of NH3 at a relatively low temperatures of approximately
200 ◦ C compared to two peaks at 200 ◦ C and 400 ◦ C for HZSM-5,
indicating that it has only weak acid sites. This result corresponds
well with pyridine FT-IR spectra (Fig. S1). Compared to NZ, HZSM-5
showed much higher Brønsted acidity.
Gas composition was also changed using NZ. Compared to non-
catalytic pyrolysis, increased amounts of CO, CO2 and olefins, such
as ethene and propene (Table S5) were observed, indicating that
catalytic reactions, such as decarbonylation, decarboxylation and
dealkylation were activated using the NZ catalyst during the cat-
alytic pyrolysis of lignin.
3.2. Two-stage catalytic pyrolysis using a combination of in-situ
NZ and ex-situ HZSM-5
Fig. 5 presents the yields of gas, oil, solid residue and coke
obtained from catalytic pyrolysis with single-stage in-situ NZ, with
single-stage ex-situ HZSM-5 and with two-stage NZ-HZSM-5 cata-
lysts. Compared to catalytic pyrolysis with single-stage in-situ NZ,
the catalytic pyrolysis over single-stage ex-situ HZSM-5 and two-
stage NZ-HZSM-5 produced smaller amounts of oil. Interestingly,
the amount of coke deposited on HZSM-5 decreased when the NZ
catalyst was applied from 2.7% over HZSM-5 to 1.9% over NZ-HZSM-
5. The smaller amount of coke formation on HZSM-5 with the co-use
of NZ can be explained by two possible behaviors of coke forma-
tion intermediates. First, potential coking compounds on ex-situ
HZSM-5 can be reduced because some of them are deposited on
the 1st stage in-situ NZ catalyst before being transferred to the
2nd stage ex-situ HZSM-5 catalyst. Second, lignin pyrolyzates pro-
duced in the 1st stage in-situ NZ catalyst consisted of the proper
285
60
Solid residue
Oil
50 Gas
Coke
40
Yield (wt%)
30
20
10
0
500 ) -5 -5
(3g SM SM
-NZ +HZ Z
nin (3 g)-H
Lig Lig -NZ
nin
Lig
Fig. 5. Product yields from catalytic pyrolysis of lignin with additive.
molecules that do not cause significant coke formation. In these
single-stage pyrolysis results (Fig. 2), the amount of solid residue
(char plus coke in NZ) was not increased using the NZ catalyst com-
pared to the non-catalytic pyrolysis of lignin. This means that the
reduction of coke in the two-stage reaction system is likely to be
due to the second reason. As shown in Fig. 3, smaller molecular
pyrolyzates are produced during the catalytic pyrolysis of lignin
over in-situ NZ, compared to the non-catalytic pyrolysis and will
be transferred to ex-situ HZSM-5 in the two-stage catalytic pyrol-
ysis system. Therefore, smaller lignin pyrolyzates, demethoxylated
phenols, and pyrocatechols formed by the pre-catalytic reactions
over the in-situ NZ catalyst may not act as the precursors of coke
on the ex-situ HZSM-5 catalyst.
Fig. 6 presents the extracted ion thermogram of CO2 obtained
from the non-isothermal oxidative EGA-MS of the coke deposited
HZSM-5 catalyst. The properties and relative amount of coke can
be understood by comparing the thermal profiles and total area of
the extracted ion thermogram for CO2 (m/z: 44) [34]. Compared to
single-stage ex-situ HZSM-5, two-stage NZ (3 g)-HZSM-5 showed a
smaller CO2 peak. A combination of the results in Figs. 5 and 6 sug-
gests that the smaller molecular size of lignin pyrolyzates produced
by the NZ catalyst may result in a lesser extent of coke forma-
tion on the HZSM-5 catalyst. Therefore, in the two-stage reaction
system, the NZ catalyst acts as a guard catalyst to increase the cat-
alytic activity and lifetime of the HZSM-5 catalyst. The increased
lifetime of the HZSM-5 catalyst can reduce the total cost of the cat-
alytic pyrolysis reaction system made by the relatively high cost of
HZSM-5.
The chemical distributions of the product oil obtained from
the catalytic pyrolysis over the two- stage NZ-HZSM-5 catalyst
were also different from those over the single-stage ex-situ HZSM-
5 (Fig. 7). Compared to catalytic pyrolysis over the single-stage
HZSM-5 catalyst, catalytic pyrolysis over the two-stage NZ-HZSM-
5 catalyst produced larger amounts of alkylphenols with smaller
amounts of guaiacols and heavy pyrolyzates. This suggests that the
use of an in-situ NZ catalyst greatly influences the chemical com-
position of the final product oil by altering the reaction pathways
over the 2nd stage ex-situ HZSM-5. Although Mullen et al. [36]
reported similar results in that phenols and cresols are generated by
the partial deoxygenation of guaiacols during the catalytic pyrol-
ysis of Asian, Lignol, Bagasse, and ETEK lignins over HZSM-5, the
additional deoxygenation using an additive catalyst has not been
reported. In this aspect, the finding that the use of a low cost NZ
catalyst and HZSM-5 in tandem can provide more efficient deoxy-
286 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288

Fig. 6. Ion thermograms for CO2 evolved during EGA-MS coke analysis.

genation and lower coke formation than the single use of ex-situ 70
HZSM-5 is quite valuable. The smaller molecular dimensions of the Lignin-NZ(3g)
pyrolyzates emitted from the 1st reactor in the presence of NZ 60 Lignin+HZSM-5
Lignin-NZ(3g)+HZSM-5
can be a better feeding chemical to produce a larger amount of
deoxygenated alkylphenols over the 2nd stage HZSM-5 catalyst. 50
Distribution (area%)

The additional use of additive NZ on the catalytic pyrolysis over

ex-situ HZSM-5 also affected the aromatic formation efficiency. 40
5.65% (MS peak area%) of BTEX (benzene, toluene, ethylbenzene,
xylenes) obtained from the catalytic pyrolysis of lignin over single- 30
stage HZSM-5 was increased by up to 8.69% when 3 g of NZ was
applied as the additive catalyst (Fig. 7). Meanwhile, product distri-
20
butions for the other mono aromatics (OMAs) were similar. Based
on these results, the use of NZ catalyst enhances BTEX formation for
10
the catalytic pyrolysis of lignin over HZSM-5. The possible reaction
mechanism for the catalytic pyrolysis of lignin over HZSM-5 and
0
NZ can be summarized as below. s
BT
EX tics ols ols col ene no
ls
ate
s
Over NZ: ma hen ech aia enz Euge lyz
o aro lk ylp ro cat Gu lk ylb p yro
mo
n A py oxy
a vy
Lignin → Guaiacols, Other heavy pyrolyzates er e th he a
Oth er
Dim Oth
→ Alkylphenols, Pyrocatechols
Fig. 7. Product distribution of catalytic pyrolysis of lignin with additive.
Over HZSM-5:
Lignin → Guaiacols, Other heavy pyrolyzates from lignin pyrolyzates is believed to occur through these two
mechanisms.
→ Alkylphenols, Aromatics
In the present study, the use of NZ catalyst also increased the
Over NZ and HZSM-5: formation of olefins presumably by the enhanced dealkylation
Lignin → Alkylphenols, Pyrocatechols activity of guaiacols, eugenols, or heavy pyrolyzates. In addition,
the selectivity to light phenolics, such as alkylphenols increased
→ Alkylphenols, Aromatics dramatically from 39.0% to 58.8% due to the enhanced decarbonly-
The mechanism of aromatic formation during the catalytic ation and demethoxylation. Therefore, the enhanced formation of
pyrolysis of lignin can be explained by a hydrocarbon pool mecha- olefins and light phenolics with the NZ catalyst may be the main
nism [23] and phenolic pool mechanism [30]. Aromatic formation reason for the enhanced production of aromatics over HZSM-5.
by the hydrocarbon pool mechanism can be explained by the
oligomerization of light olefins, such as ethene and propene, which 3.3. Effect of catalyst bed temperature of ex-situ HZSM-5 catalyst
can be produced by the dealkylation of lignin pyrolyzates, such as
eugenols [28]. In addition, the phenolic pool mechanism is related To examine the effect of the catalyst bed temperature on the
to the condensation of phenolic molecules. To et al. [30] suggested two-stage catalytic pyrolysis of lignin, the temperature of the ex-
that the conversion of phenolics to aromatics occurs via cracking situ HZSM-5 catalyst bed was varied from 500 to 600 ◦ C. Fig. 8
and H-transfer reaction of the surface pool formed by phenolics presents the yields for gas, oil, coke, and solid residue obtained at
condensation on zeolite. Therefore, the production of aromatics different HZSM-5 catalyst bed temperatures. Compared to 500 ◦ C,
 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288 287

60 70
Solid residue
Lignin-NZ(3g)+HZSM-5_500oC
Oil
Gas 60 Lignin-NZ(3g)+HZSM-5_600oC
50
Coke

50

Distribution (area%)
40
Yield (wt%)

40
30

30

20
20

10
10

0 o o 0
00C 00C EX s ls ls ls ne ls es
-5_5 -5_6 BT atic eno techo aco nze eno zat
-H ZSM
- HZ SM a rom y l ph c a G u ai y lbe E ug y roly
3g) 3g) no Alk o a lk p
NZ( NZ( mo pyr oxy vy
er eth h ea
O th er
Dim Oth
Fig. 8. Product yields from catalytic pyrolysis of lignin as a function of catalyst bed
temperature. Fig. 10. Product distribution of catalytic pyrolysis of lignin as a function of catalyst
bed temperature.

a higher gas yield and lower oil yield were obtained at 600 ◦ C. In
600 ◦ C. In addition, the selectivity to alkylphenols and guaiacols
addition, the coke yield at 600 ◦ C was lower than that of 500 ◦ C,
decreased considerably with increasing temperature from 500 ◦ C
as shown in EGA-MS coke analysis (Fig. 9). The smaller amount of
to 600 ◦ C (Fig. 10 and Table S6). The higher selectivity to aromatic
coke at higher HZSM-5 catalyst bed temperatures can be attributed
hydrocarbons at 600 ◦ C can be explained by the pore expansion of
to the elimination of thermal coke at elevated temperatures. The
HZSM-5 at higher temperatures through which the pyrolyzates can
thermal coke produced by the large molecular phenolics can be
diffuse easily into the small pores of the HZSM-5 [37]. This efficient
decomposed by the thermal cleavage reactions at elevated tem-
diffusion of pyrolyzates enhances the cracking and aromatization
peratures [36]. To et al. [30] also suggested that coke intermediates
activity over the HZSM-5 catalyst. In addition, the increased for-
formed from the condensation of phenolics were converted to aro-
mation of olefins by the dealkylation of guaiacols and eugenols can
matics at high temperatures, instead of forming coke. Therefore,
be another reason for the increased aromatic formation at higher
coke intermediates are believed to be released from HZSM-5 at
temperatures of the HZSM-5 catalyst bed.
600 ◦ C, resulting in less coke formation.
Fig. 10 shows the distribution of the product oil obtained from
the catalytic pyrolysis of lignin at different HZSM-5 catalyst bed 4. Conclusions
temperatures. The selectivity to BTEX and other monoaromatics
increased from 8.69% to 14.56% and from 4.69% to 13.35%, respec- The two-stage catalytic pyrolysis of lignin using a tandem fixed
tively, with increasing temperature from 500 ◦ C to 600 ◦ C. The bed reactor system was investigated using in-situ NZ and ex-situ
maximum yield (1.45 wt.%) of monoaromatics was obtained from HZSM-5 as the additive and main catalyst, respectively. Although
the catalytic pyrolysis of lignin over NZ (3 g) and HZSM-5 (1 g) at the use of NZ additive as an in-situ catalyst did not have a signif-

Fig. 9. Ion thermograms for CO2 evolved during EGA-MS coke analysis (effect of catalyst bed temperature).
288 H.W. Lee et al. / Journal of Analytical and Applied Pyrolysis 122 (2016) 282–288
icant effect on the formation of aromatic hydrocarbons due to the
weak acidity of the NZ catalyst, it was very helpful in increasing
the amount of alkylphenols and pyrocatechols with a decrease in
the amounts of guaiacols and eugenols. In addition, the NZ catalyst
decreased the amounts of heavy lignin pyrolyzates. The pyrolyzates
produced in the presence of the NZ catalyst had a beneficial effect
on the downstream ex-situ HZSM-5 upgrading process. Compared
to the single catalytic pyrolysis with HZSM-5, two-stage NZ-HZSM-
5 catalytic pyrolysis produced a smaller amount of coke and a
higher quality of product oil, having larger amounts of BTEXs and
alkylphenols. An increase in the temperature of the ex-situ HZSM-5
catalyst bed resulted in an increase in aromatic formation through
the hydrocarbon pool and phenolic pool mechanism. The two-stage
catalytic pyrolysis process using a series of NZ and HZSM-5 cata-
lysts developed in this study will be very helpful for making the
catalytic pyrolysis process of lignin more economical by extending
the lifetime of the expensive HZSM-5 catalyst.
Acknowledgement
This work was supported by the New & Renewable Energy Core
Technology Program of the Korea Institute of Energy Technology
Evaluation and Planning (KETEP), granted financial resource from
the Ministry of Trade, Industry & Energy, Republic of Korea. (No.
20153030101580). Also, this work was supported by the National
Research Council of Science & Technology (NST) grant by the Korea
government (MSIP) (No. CAP-11-04-KIST).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jaap.2016.09.015.
References
[1] E. Cséfalvay, G.R. Akien, L. Qi, I.T. Horváth, Catal. Today 239 (2015) 50.
[2] J.S. Cha, S.H. Park, S.C. Jung, C. Ryu, J.K. Jeon, M.C. Shin, Y.K. Park, J. Ind. Eng.
Chem. 40 (2016) 1.
[3] W. Liu, P. Yin, X. Liu, S. Zhang, R. Qu, J. Ind. Eng. Chem. 21 (2015) 893.
[4] M.D.N. Meinita, B. Marhaeni, T. Winanto, D. Setyaningsih, Y.K. Hong, J. Ind.
Eng. Chem. 27 (2015) 108.
[5] S.H. Park, H.J. Cho, C. Ryu, Y.K. Park, J. Ind. Eng. Chem. 36 (2016) 314.
[6] W. Mu, H. Ben, A. Ragauskas, Y. Deng, Bioenergy Res. 6 (2013) 1183.
[7] D.J. Nowakowski, A.V. Bridgwater, D.C. Elliott, D. Meier, P. de Wild, J. Anal.
Appl. Pyrolysis 88 (2010) 53.
[8] M. Milina, S. Mitchell, J. Pérez-Ramírez, Catal. Today 235 (2014) 176.
[9] D.E. Resasco, S.P. Crossley, Catal. Today 15 (2015) 185.
[10] E.H. Lee, R.S. Park, H. Kim, S.H. Park, S.C. Jung, J.K. Jeon, S.C. Kim, Y.K. Park, J.
Ind. Eng. Chem. 37 (2016) 18.
[11] S. Farag, J. Chaouki, Bioresour. Technol. 175 (2015) 254.
[12] H. Ben, A.J. Ragauskas, RSC Adv. 2 (2012) 12892.
[13] N. Smolarski, High-value opportunities for lignin: Unlocking its potential,
Frost & Sullivan, 2012.
[14] C. Amen-Chen, H. Pakdel, C. Roy, Bioresour. Technol. 79 (2001) 277.
[15] S. Zhou, M. Garcia-Perez, B. Pecha, S.R.A. Kersten, A.G. McDonald, R.J.M.
Westerhof, Energy Fuels 27 (2013) 5867.
[16] T.C. Drage, C.H. Vane, G.D. Abbott, Org. Geochem. 33 (2002) 1523.
[17] S. Chu, A.V. Subrahmanyam, G.W. Huber, Green Chem. 15 (2013) 125.
[18] C.H. Ko, S.H. Park, J.K. Jeon, D.J. Suh, K.E. Jeong, Y.K. Park, Korean J. Chem. Eng.
29 (2012) 1657.
[19] H. Ben, A.J. Ragauskas, Energy Fuels 25 (2011) 4662.
[20] Y. Zhao, L. Deng, B. Liao, Y. Fu, Q.X. Guo, Energy Fuels 24 (2010) 5735.
[21] M.A. Jackson, D.L. Compton, A.A. Boateng, J. Anal. Appl. Pyrolysis 85 (2009)
226.
[22] R. Lou, S.B. Wu, G.J. Lv, J. Anal. Appl. Pyrolysis 89 (2010) 191.
[23] Z. Ma, E. Troussard, J.A.V. Bokhoven, Appl. Catal. A: Gen. 423–424 (2012) 130.
[24] N. Rajić, N.Z. Logar, A. Rečnik, M. El-Roz, F. Thibault-Starzyk, P. Sprenger, L.
Hannevold, A. Andersen, M. Stöcker, Microporous Mesoporous Mat. 176
(2013) 162.
[25] H.W. Lee, T.H. Kim, S.H. Park, J.K. Jeon, D.J. Suh, Y.K. Park, J. Nanosci.
Nanotechnol. 13 (2013) 2640.
[26] S.S. Kim, H.W. Lee, R. Ryoo, W. Kim, S.H. Park, J.K. Jeon, Y.K. Park, J. Nanosci.
Nanotechnol. 14 (2014) 2414.
[27] K. Wang, K.H. Kim, R.C. Brown, Green Chem. 16 (2014) 727.
[28] Z. Ma, J.A.V. Bokhoven, ChemCatChem 4 (2012) 2036.
[29] P. Djinović, T. Tomše, J. Grdadolnik, Š. Božič, B. Erjavec, M. Zabilskiy, A. Pintar,
Catal. Today 258 (2015) 648.
[30] A.T. To, D.E. Resasco, Appl. Catal. A: Gen. 487 (2014) 62.
[31] H.W. Lee, S.H. Park, J.K. Jeon, R. Ryoo, W. Kim, D.J. Suh, Y.K. Park, Catal. Today
232 (2014) 119.
[32] R. Yadav, A.K. Singh, A. Sakth, Catal. Today 245 (2015) 155.
[33] H.W. Lee, I.G. Lee, S.H. Park, J.K. Jeon, D.J. Suh, J. Jung, Y.K. Park, J. Nanosci.
Nanotechnol. 14 (2014) 2990.
[34] B.S. Kim, Y.M. Kim, J. Jae, C. Watanabe, S. Kim, S.C. Jung, S.C. Kim, Y.K. Park,
Bioresour. Technol. 194 (2014) 312.
[35] L. Jia, Y. Le-Brech, B. Shrestha, M.B. Frowein, S. Ehlert, G. Mauviel, R.
Zimmermann, A. Dufour, Energy Fuels 29 (2015) 7364.
[36] C.A. Mullen, A.A. Boateng, Fuel Process. Technol. 91 (2010) 1446.
[37] J.Y. Kim, J.H. Kim, J. Park, J.K. Kim, D. An, I.K. Song, J.W. Choi, J. Anal. Appl.
Pyrolysis 114 (2015) 273.