Enhancing the production of renewable petrochemicals by co-feeding of biomass with plastics in catalytic fast pyrolysis with ZSM-5 zeolites

Article abstract of DOI:10.1016/j.apcata.2014.05.015

This study investigated catalytic fast pyrolysis (CFP) of a series of biomass (cellulose, lignin, and pine wood), plastics (low-density polyethylene (LDPE), polyethylene (PP), and polystyrene (PS)), and their mixtures with ZSM-5 zeolite. Co-feeding of cellulose with LDPE (mixing ratios of 4-1) produced much higher petrochemical (aromatics and olefins) yields (52.1-55.6 C%) and lower solid (coke/char) yields (22.6-10.9 C%) than those expected if there were no chemical interactions between the two feedstocks in co-feed CFP (37.4-39.2 C% and 25.0-15.9 C% for petrochemicals and solid, respectively, calculated by linear addition of the corresponding yields determined in CFP of cellulose and LDPE individually). This result indicates that cellulose and LDPE have a significant synergy that enhances the production of valuable petrochemicals and decreases the undesired coke in CFP. Similar synergy was also observed in co-feed CFP of pine wood and LDPE mixtures (mixing ratio of 2), which produced 49.5 C% petrochemicals and 19.5 C% solid residue. In comparison, CFP of pine wood and LDPE individually produced only 31.6 C% and 41.0 C% petrochemicals and 46.5 C% and 6.74 C% solid, respectively. This synergy, however, was less pronounced for the other combinations of biomass and plastics (cellulose/PP, cellulose/PS, and lignin/LDPE) tested in this study. The results suggest that the interactions between the primary pyrolysis products of cellulose and LDPE, especially Diels-Alder reactions of cellulose-derived furans with LDPE-derived linear α-olefins, play an important role in the synergy for petrochemical production and coke reduction in co-feed CFP. Co-feeding LDPE thus has great potential in improving the performance of CFP of natural lignocellulosic biomass, which usually contains a significant fraction (40-50 wt.%) of cellulose component.

Full text of DOI:10.1016/j.apcata.2014.05.015

Applied Catalysis A: General 481 (2014) 173–182
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Enhancing the production of renewable petrochemicals by co-feeding
of biomass with plastics in catalytic fast pyrolysis with ZSM-5 zeolites
Xiangyu Lia,b, Jian Li , Guoqiang Zhou , Yu Feng , Yujue Wanga,b,∗, Gang Yua,
a
a
a
a
a
a
Shubo Deng , Jun Huang , Bin Wang
School of Environment, State Key Joint Laboratory of Environmental Simulation and Pollution Control, Tsinghua University, Beijing 100084, China
Shanghai Tobacco (Group) Corporation, Shanghai 200082, China
a
b
a r t i c l e i n f o
a b s t r a c t
Article history:
This study investigated catalytic fast pyrolysis (CFP) of a series of biomass (cellulose, lignin, and pine
wood), plastics (low-density polyethylene (LDPE), polyethylene (PP), and polystyrene (PS)), and their
mixtures with ZSM-5 zeolite. Co-feeding of cellulose with LDPE (mixing ratios of 4–1) produced much
higher petrochemical (aromatics and olefins) yields (52.1–55.6 C%) and lower solid (coke/char) yields
Received 26 January 2014
Received in revised form 29 April 2014
Accepted 17 May 2014
Available online 27 May 2014
(22.6–10.9 C%) than those expected if there were no chemical interactions between the two feedstocks in
co-feed CFP (37.4–39.2 C% and 25.0–15.9 C% for petrochemicals and solid, respectively, calculated by lin-
ear addition of the corresponding yields determined in CFP of cellulose and LDPE individually). This result
indicates that cellulose and LDPE have a significant synergy that enhances the production of valuable pet-
rochemicals and decreases the undesired coke in CFP. Similar synergy was also observed in co-feed CFP
of pine wood and LDPE mixtures (mixing ratio of 2), which produced 49.5 C% petrochemicals and 19.5 C%
solid residue. In comparison, CFP of pine wood and LDPE individually produced only 31.6 C% and 41.0 C%
petrochemicals and 46.5 C% and 6.74 C% solid, respectively. This synergy, however, was less pronounced
for the other combinations of biomass and plastics (cellulose/PP, cellulose/PS, and lignin/LDPE) tested in
this study. The results suggest that the interactions between the primary pyrolysis products of cellulose
and LDPE, especially Diels–Alder reactions of cellulose-derived furans with LDPE-derived linear ␣-olefins,
play an important role in the synergy for petrochemical production and coke reduction in co-feed CFP.
Co-feeding LDPE thus has great potential in improving the performance of CFP of natural lignocellulosic
biomass, which usually contains a significant fraction (40–50 wt.%) of cellulose component.
Keywords:
Catalysis
Pyrolysis
Biomass
Plastics
Aromatic hydrocarbons
Olefins
Zeolite
©
2014 Elsevier B.V. All rights reserved.
1
. Introduction
Catalytic fast pyrolysis (CFP) of lignocellulose with zeolite cat-
alysts (e.g., ZSM-5) is a promising technology that can convert
lignocellulosic feedstocks directly into valuable aromatics and
olefins [4,5]. CFP involves fast heating biomass feedstocks in the
Petrochemicals derived from petroleum are important feed-
stocks for manufacturing a wide range of useful products such
as plastics, synthetic fibers, solvents, and pharmaceuticals. The
two main classes of petrochemicals are olefins and aromatics. Par-
ticularly, ethylene, propylene, butadiene, benzene, toluene, and
xylenes (BTX) are the six most important primary petrochemicals
from which numerous petrochemical intermediates and commer-
cial products can be made. As the petroleum reserves on the
earth are depleted at a rapid rate, producing petrochemicals from
renewable lignocellulosic biomass has attracted increasing atten-
tion [1–3].
◦
presence of zeolite catalysts at about 400–700 C. Volatile interme-
diate species (e.g., anhydrosugars, furans, and phenolics) formed in
the initial thermal decomposition of biomass are then further cat-
alytically converted to aromatics and olefins via a series of reactions
(e.g., cracking, deoxygenation, oligomerization, and aromatization)
over the zeolite catalysts. The whole conversion process can be
completed in a single reactor at short reaction times. The major
aromatic products include benzene, toluene, xylenes, and naph-
thalenes, while the prominent olefin products are ethylene and
propylene [4]. Thus, of the six most important primary petroche-
micals, five can be produced from CFP of lignocellulosic biomass
[
4–6].
Although CFP of lignocellulose can produce important petroche-
∗ Corresponding author at: School of Environment, Tsinghua University, Beijing
00084, China. Tel.: +86 10 62772914; fax: +86 10 62785687.
E-mail address: wangyujue@tsinghua.edu.cn (Y. Wang).
micals, it also produces large amounts of solid residue, including
both char and coke (carbon yields usually above 30 C%) [7–9].
1
http://dx.doi.org/10.1016/j.apcata.2014.05.015
926-860X/© 2014 Elsevier B.V. All rights reserved.
0

1
74
X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
Char is defined herein as the primary solid product formed in non-
catalytic pyrolysis of carbonaceous materials [8,10], whereas coke
is defined as the carbon deposits formed on catalysts via hetero-
geneous reactions of primary pyrolysis vapors (e.g., dehydration
of biomass-derived oxygenates) at the catalytic sites of catalysts
Nankai University (Tianjin, China). All biomass, plastics, and zeolite
catalyst samples were (crushed and) sieved through a 140 mesh
(0.105 mm) sieve, and then stored in a desiccator before use. The
content of cellulose, hemicellulose, and lignin in the pine wood was
analyzed using the protocol of Goering and Van Soest [19]. Carbon,
hydrogen, and nitrogen contents of the biomass and plastics were
analyzed with an elemental analyzer (CE-440, Exeter Analytical,
Inc., North Chelmsford, MA). For co-pyrolysis tests, a particular
biomass sample (cellulose, lignin, or pine wood) was mixed with
a plastic sample (LDPE, PP, or PS) in a biomass-to-plastic ratio
between 4 and 1. To prepare the samples for CFP tests, ZSM-5
zeolite was thoroughly mixed with a particular reactant (biomass,
plastic, and their co-feed mixture) in a catalyst-to-reactant ratio of
15. It is noted that due to the small sample sizes used in pyroprobe
pyrolysis tests, high catalyst-to-reactant ratios (e.g., 10–20) and
complete mixing are usually required to ensure that the primary
pyrolysis products of biomass/plastics can be effectively converted
within the framework of catalysts [13,20,21].
[
8,10,11]. The formation of char/coke decreases the carbon effi-
ciency of biomass to petrochemical products. Moreover, the coke
deposits would cause catalyst deactivation, leading to significant
loss in the conversion efficiency of biomass with increasing opera-
tion times in CFP processes [8,12]. It is thus crucial to decrease the
formation of char/coke (especially coke) in CFP of biomass.
High solid yield in CFP of lignocellulose is due to its oxygen rich
(
usually >40 wt.%) and hydrogen deficient (∼6–7 wt.%) nature [7,8].
The initial thermal decomposition of lignocellulose thus produces
many highly oxygenated primary products (e.g., hydroxyacetalde-
hyde, acetic acid, and furfural) [8,13]. These oxygenates would
produce large amounts of catalytically derived coke when they
are converted over zeolites [8,12,14]. The similar coking problems
have been observed in previous studies on zeolite upgrading of
bio-oils and their model compounds (e.g., acetic acid and acetone)
2.2. Fast pyrolysis analysis
[
12,15,16]. However, researchers have found that coke formation
can be largely decreased through co-processing of bio-oils with
hydrogen-rich reactants (e.g., methanol) during zeolite upgrading
Fast pyrolysis tests were conducted with a Pyroprobe 5200 ana-
lytical pyrolyzer (CDS Analytical, Inc.) using a protocol described
[
15,17]. This result suggests that by co-feeding hydrogen-deficient
elsewhere [9,22]. About 1 mg (for non-CFP) or 4 mg (for CFP) of
biomass with hydrogen-rich feedstocks in CFP, it might be possible
to decrease coke formation from CFP of biomass.
sample was rapidly heated to a preset temperature (450–750 ◦C)
at a heating rate of 20 ◦C/ms, and then held for 60 s under high-
We have therefore proposed to co-feed biomass with plastic
wastes in CFP to decrease coke formation [18]. Plastic wastes are
generated in huge quantities each year worldwide. They are usu-
ally rich in hydrogen (e.g., ∼14–15 wt.% for polyethylene (PE) and
polypropylene (PP)), and may thus be used as a cheap and abundant
hydrogen source to decrease coke formation in CFP of biomass. To
examine this hypothesis, we have preliminarily tested co-feeding
purity helium flux during fast pyrolysis. The volatiles emitted via
this fast heating were carried by helium through a heated trans-
porting tube (300 ◦C) to a gas chromatograph (Agilent 7890A) that
was equipped with a mass spectrometer (5975C MSD), a flame ion-
ization detector (FID), and a thermal conductivity detector (TCD).
The condensable pyrolysis products were separated with an HP-5
MS column (30 m × 0.25 mm i.d. × 0.25 ␮m film thickness, Agilent
Technologies, Inc.).
The non-condensable gases were separated with a fused sil-
ica capillary column (HP-Plot/Q, 30 m × 0.32 mm i.d. × 20 ␮m film
thickness, Agilent Technologies, Inc.). For tentative identification
of pyrolysis products, all mass spectra were compared to the
NIST mass spectrum library. The yields of hydrocarbons (alkanes,
of cellulose with low-density polyethylene (LDPE) (mixing ratio of
◦
1
:1) in CFP at 650 C. The result is very promising, showing that cel-
lulose and LDPE have a significant synergy that not only decreases
coke formation, but also considerably enhances the production of
valuable aromatic products (e.g., BTX compounds) in the co-feed
CFP [18].
Based on the promising result of our preliminary work, this
study aimed to further investigate co-feed CFP of biomass and
plastics systematically. A series of biomass (cellulose, lignin, pine
wood), plastics (LDPE, PP, and polystyrene (PS)), and their mixtures
were catalytically pyrolyzed with ZSM-5 zeolites. The effects of
important operation factors (reaction temperature and biomass-
to-plastic ratio) on the synergy between cellulose and LDPE were
evaluated. The product distribution in CFP of the different biomass,
plastics, and their mixtures were analyzed to determine if simi-
lar synergy for petrochemical production and coke reduction exists
between the different biomass and plastics. The results of this study
may provide the necessary information required to optimize the
CFP process and select proper feedstocks to maximize the synergy
between biomass and plastics in co-feed CFP.
olefins, and aromatics) and carbon oxides (CO and CO ) were quan-
2
tified with FID and TCD, respectively. Triplicate CFP tests were
performed for the condensable and non-condensable product anal-
yses, respectively. The carbon contents in the spent catalysts were
measured in triplicate using the elemental analyzer to determine
the solid (char/coke) yield in CFP tests. All yields are reported in
terms of carbon yield (moles of carbon atoms in the products rela-
tive to those in the reactants, Eq. (1)). Aromatic and olefin selectivity
are calculated according to Eqs. (2) and (3), respectively.
moles of carbon in a product
moles of carbon in reactant
Carbon yield (%) =
× 100
(1)
(2)
(3)
moles of carbon in an aromatic product
moles of carbon in all aromatic products
Aromatic selectivity (%) =
× 100
moles of carbon in an olefinic product
moles of carbon in all olefinic products
Olefin selectivity (%) =
× 100
2
. Experimental
3
. Results and discussion
.1. Elemental composition analysis
The elemental composition of the biomass and plastic sam-
2.1. Materials
3
Microgranular cellulose was purchased from Sigma–Aldrich.
Lignin was kindly provided by a company in Beijing that produced
bioethanol via hydrolysis of corn stover. Pine wood sawdust
was acquired from a furniture factory in Beijing. Plastic samples
ples are listed in Table 1. The effective hydrogen index (H/Ceff),
which is a measure of the relative abundance of hydrogen in a
feedstock, is calculated according to Eq. (4) [15]. Previous stud-
ies have suggested that feedstocks with higher H/Ceff are generally
easier to convert to hydrocarbons and produce less coke in CFP,
(
LDPE, PP, and PS (d < 0.105 mm)) were purchased from Li Yang
Technology Corporation (Shanghai, China). ZSM-5 zeolite with
SiO /Al O ratio of 25 was acquired from the Catalyst Plant of
2
2
3

X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
175
Table 1
plastics still contained significant fractions of biomass-derived oxy-
genates (see Fig. S3 in SD), which may cause aging and corrosion
during storage and combustion [26,32]. The liquid products pro-
duced in non-CFP of biomass and plastic mixtures thus cannot be
directly used in conventional engines and boilers, and need further
upgrading [5].
Elemental analysis of the biomass and plastic samples used in this study.
C
H
N
O
H/Ceff
Cellulose
Lignin
Pine wood
LDPE
PP
PS
43.3
65.1
51.9
85.5
85.3
92.2
6.20
5.72
6.36
14.5
14.7
7.80
0.45
0.97
0.51
0
0
0
50.05
28.20
41.23
0
0
0
0.0
0.4
0.3
2.0
2.1
1.0
3.3. CFP of individual biomass and plastic samples
Although the biomass and plastics produced significantly dif-
ferent products in non-CFP, they produced similar final products
when catalytically pyrolyzed with ZSM-5 zeolites (i.e., CFP). Fur-
thermore, co-feeding of the biomass with plastics in CFP also
produced the same products (see Fig. S4 in SD for the TIC of the
pyrolysis products). For all samples, the liquid products were com-
posed predominantly of aromatic hydrocarbons (see Table S7 in
whereas feedstocks with lower H/Ceff (e.g., <1) may produce large
quantities of coke during zeolite upgrading, leading to rapid cata-
lyst deactivation [8,11]. As reported in Table 1, the lignocellulosic
biomass samples had a H/Ceff of 0–0.4, whereas the plastics had
a H/Ceff of 1.0–2.1. This result indicates that the biomass samples
are hydrogen-deficient feedstocks for CFP conversion, whereas the
plastics are much more hydrogen-rich.
SD). In addition, they all produced C –C5 hydrocarbons (alkanes
1
and olefins) as the major gaseous products (biomass also produced
H
H − 2O
considerable amounts of CO and CO ) and considerable amounts of
2
=
(4)
Ceff
C
solid residues (char/coke).
Many researchers have investigated the reaction pathways in
catalytic conversion of biomass (e.g., cellulose [5,7], lignin [7,10,13],
hemicellulose [7], and their model compounds [12,16,20,33]) and
plastics (e.g., PE [34–36], PP [34,37], and PS [27,31]) over ZSM-5
zeolites. The results suggest that although the starting feedstocks
may have very different molecular structures and chemical compo-
sitions, they may undergo several common reaction steps to form
similar products (e.g., aromatics and olefins) during ZSM-5 cat-
alyzed conversion processes: (1) cracking (and deoxygenation) of
3
.2. Non-catalytic fast pyrolysis (non-CFP)
The biomass and plastics produced significantly different prod-
ucts when they were fast pyrolyzed in the absence of catalysts
i.e., non-CFP) (see Supplementary Data (SD) for the total ion chro-
(
matograms (TIC) and lists of the pyrolysis products). In general, the
biomass samples produced predominantly oxygenated products,
whereas the plastics produce hydrocarbons. The major products
identified in this study for these biomass and plastic samples were
essentially the same as those reported in literature (e.g., cellu-
lose [7,9,23], lignin [7,10,24], pine wood [25,26], PE [27–29], PP
polymer structures to small olefins (e.g., C –C5); (2) oligomeriza-
2
tion of the small olefins to C –C olefins, which then transform
6
10
to C –C dienes via hydride transfer reactions; and (3) cyclization
6
10
[
28–30], and PS [27,31]). For example, the most prominent products
and aromatization of the dienes to form aromatics [12,16,34,35]
(this is with the exception of PS [31,38], see SD for more discus-
sion on the reaction pathways). As a result, CFP of the biomass and
plastics with ZSM-5 zeolites produced similar final products.
The major products in CFP of the individual biomass or plas-
tic samples have been quantified and summarized in Table 2. The
product distributions reported herein were generally comparable
to those reported in similar studies (e.g., CFP of cellulose [5,7,18],
lignin [7,13], pine wood [4], LDPE [18], and PS [38]). In general, CFP
of the biomass feedstocks produced less petrochemicals (aromat-
ics and olefins) and considerably more solid residues than CFP of
the plastic samples (Table 2). The highest petrochemical yield (82.1
C%) was obtained in CFP of PS. In contrast, CFP of lignin produced
the lowest petrochemical yield (9.55 C%) and the highest solid yield
(64.7 C%).
were levoglucosan and furfural for cellulose (Fig. S2(a)), and phe-
nols, guaiacols, and syringols for lignin (Fig. S2(b)). These cellulose-
and lignin-derived products were also the major products from
non-CFP of pine wood sawdust (Fig. S2(c)), which contained con-
siderable fractions of cellulose (46.1 wt%) and lignin (24.7 wt%).
In addition, non-CFP of the pine wood also produced consider-
able amounts of hemicellulose-derived products (e.g., acetic acid)
because the pine wood contained 15.8 wt.% hemicellulose compo-
nent. The major pyrolysis products of LDPE consisted of a series of
linear alkanes, alkenes, and dienes (Fig. S2(d)), which are formed
from random scission of the original polymer chains of LDPE during
thermal decomposition [28]. In comparison, PP produced predomi-
nantly branched alkanes and alkenes (Fig. S2(e)) due to its branched
polymer structure [29]. The predominant products in non-CFP of
polystyrene were styrene and its oligomers (Fig. S2(f)), similar to
those reported by other researchers [27,31].
As expected, the biomass samples produced much higher yields
of solid residues (char/coke) in CFP than the plastics. This can
be mainly attributed to the fact that the biomass had consid-
erably lower hydrogen contents (H/Ceff = 0–0.4) than the plastics
(H/Ceff = 1–2.1) (see Table 1), and would thus produce much more
catalytically derived coke when they are converted in CFP with
ZSM-5 zeolite than the plastics [7,8,11]. In addition, lignin and pine
wood produced considerable amounts of thermally derived char
(33.8 wt.% and 13.8 wt.%, respectively, determined by thermogravi-
metric analysis, see Fig. S1 in SD) during pyrolysis, whereas the
plastics produced negligible char (0.33–1.42 wt.%). To decrease the
coke formation in CFP of biomass, we have proposed to co-feed
hydrogen-deficient biomass with hydrogen-rich plastic wastes to
increase the H/Ceff of the feedstocks in CFP [18]. Our preliminary
study has tested co-feed CFP of cellulose and LDPE mixture with
Similar to what has been observed in non-CFP of cellulose and
LDPE mixture [18], minimal interactions between the biomass (cel-
lulose, lignin, and pinewood) and plastics (LDPE, PP, and PS) were
observed when they were co-pyrolyzed in the absence of ZSM-
5
catalysts. For example, non-CFP of the mixture of pine wood
and LDPE just produced a “combined” chromatogram that resem-
bles the superposition of the two individual components (Fig. S3
in SD). All identified major products in non-CFP of the mixture
have previously been observed in pyrolysis of the two individual
components. No new products or significant change in the relative
yields of any products from either the biomass or the plastic were
observed. Similarly, Bhattacharya et al. [25] found that negligible
cross-over products of biomass and plastics were formed in co-
pyrolysis of pine wood with PE, PP, and PS. These results indicate
that co-pyrolysis of biomass with plastics without using catalysts
◦
a mixing ratio of 1:1 at 650 C. The results show that cellulose
and LDPE have a significant synergy that decreases coke formation
and enhances the production of valuable aromatics (e.g., benzene,
toluene, and xylenes) in co-feed CFP. To optimize the synergy
(
i.e., non-CFP) has little effect on improving product distributions.
The liquid products produced from co-pyrolysis of biomass and

1
76
X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
Table 2
with the reaction temperature. In contrast, the yields of less valu-
able alkanes and undesired coke decreased as the temperature was
increased from 450 to 750 C. Increasing the reaction temperature
thus enhanced the conversion of cellulose and LDPE to desired pet-
rochemicals (but consumes more energy).
Product distribution in CFP of individual biomass and plastic samples (reaction
temperature 550 C; catalyst-to-reactant ratio of 15).
◦
◦
Cellulose Lignin Pine wood LDPE PP
PS
Overall yield (C%)
Aromatics
Olefins
Alkane
CO
32.9
2.68
2.06
23.0
5.96
7.84 28.5
28.3
12.7
55.1
0.00
0.00
6.74
35.4
13.9
48.9
0.00
0.00
80.2
1.94
2.68
0.00
0.00
The reaction temperature had a pronounced effect on the distri-
bution of aromatic and olefin products (Fig. 1). As the temperature
1.71
2.65
3.09
2.49
◦
was increased from 450 to 750 C, the selectivity for benzene
2.05 15.1
2.05 3.16
64.7 46.5
9.55 31.6
increased considerably, while the selectivity for xylenes and higher
alkylbenzenes decreased. This trend was observed in CFP of both
cellulose and LDPE, and their mixture (Fig. 1(a–c)), and can be
attributed to the fact that dealkylation reactions are enhanced at
higher temperatures [39]. On the other hand, the selectivity for
toluene, indenes, and polyaromatics were not significantly affected
CO2
Solid (coke and char) 34.0
Petrochemicals 35.6
6.38 10.2
49.3
41.0
82.1
Aromatic selectivity (%)
Benzene
12.6
8.6
21.6
21.0
5.2
10.1
23.8
21.0
6.6
14.2
35.4
24.6
13.0
0.0
15.2
35.7
22.0
12.7
0.0
44.2
11.6
1.8
8.6
0.0
Toluene
25.1
15.6
6.2
Xylenesa
◦
by the reaction temperature within the test range (450–750 C).
b
Other alkylbenzenes
c
For olefin products, the selectivity for C2–C3 olefins (ethylene and
propylene) increased with the reaction temperature, while the
Phenols
Indenes
0.0
3.9
0.0
d
5.1
3.6
5.4
2.1
2.1
6.9
26.9
Polyaromaticse
35.5
36.1
33.2
10.7
12.4
selectivity for C –C5 olefins decreased (Fig. 1(d–f)). This result is
4
consistent with the general observation that increasing reaction
temperature enhances cracking reactions to produce small olefins
Olefin selectivity (%)
Ethylene
Propylene
C4 olefins
C5 olefins
54.5
35.5
9.9
59.1
31.0
7.1
61.4
30.7
5.7
18.4
36.9
25.3
19.4
31.7
39.6
19.1
9.7
46.2
37.6
11.9
4.3
[
34].
Notably, co-feeding of cellulose with LDPE in CFP produced
0.0
2.8
2.1
considerably higher aromatic yields (36.8–42.3 C%) than CFP of
the two components individually (25.6–36.1 C% for cellulose and
a
Xylenes include m-, p-, and o-xylene.
Other alkylbenzenes include ethylbenzene, trimethylbenzenes, tetramethyl-
b
2
6.3–32.7 C% for LDPE) at all tested temperatures (Table 3). This
benzenes, propenylbenzene, and butenylbenzene.
c
result indicates that cellulose and LDPE have a synergy for aro-
matic production in the co-feed CFP, which has been attributed
mainly to the interactions between cellulose- and LDPE-derived
primary pyrolysis products [18]. It has been suggested that in the
presence of ZSM-5 zeolites, cellulose-derived oxygenates may react
with LDPE-derived hydrocarbons in many ways [18]. For exam-
ple, cellulose-derived furans (e.g., furan and furfural) can react
with LDPE-derived olefins (e.g., ethylene and propylene) to form
aromatics (e.g., toluene and xylenes) via Diels–Alder reactions (in
which furans and olefins act as the diene and dienophile compo-
nents, respectively) and dehydration reactions (Fig. 2) [39–41]. This
can enhance the aromatic production and decrease coke forma-
tion from polymerization of furans [6,18]. In addition, LDPE-derived
hydrocarbons (e.g., alkanes and olefins) may provide hydrogen
for cellulose-derived oxygenates, which can decrease coke for-
mation from zeolite-catalyzed conversion of hydrogen-deficient
oxygenates [8,15,42].
To evaluate how the reaction temperature influenced the syn-
ergy between cellulose and LDPE in co-feed CFP, we calculated
additive carbon yields in the co-feed CFP for the major products
using the corresponding yields determined in CFP of cellulose
and LDPE individually (Eq. (5)). These additive yields would
have equaled the yields experimentally determined in CFP of the
cellulose and LDPE mixture if there had been no interactions
between the two feedstocks. However, remarkable differences
were observed between the calculated additive and experi-
mentally determined yields (Fig. 3), indicating that significant
Phenols include phenol, cresols, and guaiacol.
Indenes include indane, indene, 2,3-dihydro-2,2-dimethyl-1H-indene,
d
2
2
,3-dihydro-4-methyl-1H-indene,
2,3-dihydro-1,3-dimethyl-1H-indene,
-methylindene, and 1,3-dimethyl-1H-indene.
e
Polyaromatics include naphthalene, methylnaphthalenes, dimethylnaph-
thalenes, and polycyclic aromatic hydrocarbons (benzene rings >2).
between cellulose and LDPE, this study further evaluated the effects
of reaction temperature and cellulose-to-LDPE ratio on the product
distributions in co-feed CFP.
3
.4. Co-feeding of cellulose with LDPE in CFP
The product distribution in CFP of cellulose, LDPE, and their
◦
mixture (1:1) were at an array of temperatures between 450 C
◦
and 750 C are summarized in Table 3. For cellulose, increasing
the reaction temperature increased the yields of gaseous (carbon
oxides, alkanes, and olefins) and liquid (aromatics) products, and
decreased the yield of solid residue in CFP (Table 3). This trend
is consistent with the general observation that increasing tem-
perature enhances the production of gaseous and liquid products
during CFP of biomass [4,7]. As a result, the yield of petrochemi-
cals (aromatics plus olefins) increased considerably from 27.1 C%
◦
to 39.7 C% as the reaction temperature was increased from 450 C
◦
to 750 C (Table 3). Similar trends were observed for the product
distributions in CFP of LDPE alone and the mixture of cellulose and
LDPE (Table 3). The yields of both aromatics and olefins increased
Table 3
Effects of reaction temperature on the product distribution in CFP of cellulose, LDPE, and their mixture (cellulose-to-LDPE ratio of 1).
Cellulose
450
LDPE
450
Mixture (1:1)
◦
Temperature ( C)
550
650
750
550
650
750
450
550
650
750
Overall yield (C%)
Aromatics
Olefins
Alkanes
CO
25.6
1.54
1.05
19.8
5.37
47.9
27.1
32.9
34.8
36.1
26.3
10.8
56.5
0
28.3
12.7
55.1
0
29.4
15.8
51.6
0
32.7
18.2
44.6
0
36.8
42.1
10.9
26.6
4.71
0.93
10.9
53.0
42.1
14.2
24.7
6.11
1.08
10.3
56.3
42.3
15.8
23.0
6.59
1.09
8.62
2.68
2.06
23.0
5.96
34.0
35.6
3.45
2.23
23.9
5.65
28.0
38.3
3.59
2.35
25.4
5.65
21.9
39.7
8.69
28.1
4.37
0.90
18.0
45.5
CO2
0
0
0
0
Solid (coke and char)
Petrochemicals
6.99
37.1
6.74
41.0
6.65
45.2
5.80
50.9
58.1

X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
177
Fig. 1. Effects of reaction temperature on the distributions of aromatic and olefin products in CFP of cellulose, LDPE, and their mixture (1:1).
chemical interactions occurred between cellulose and LDPE in co-
feed CFP.
The fact that the experimental yield was higher than the additive
yield for olefins was somewhat unexpected because LDPE-derived
olefins are consumed in the Diels–Alder reactions with cellulose-
derived furans in co-feed CFP. However, a close examination of the
reaction pathways suggests that the presence of cellulose-derived
oxygenates may help to preserve olefins from being hydrogenated
to alkanes. In the presence of ZSM-5 zeolites, olefins can undergo
zeolite-catalyzed oligomerization, hydrogen transfer, and cycliza-
tion to form C6–C10 cyclic diolefins, which then release hydrogen
to form aromatics (see Fig. S6(c) in SD) [34,35]. Without the exist-
ence of biomass-derived oxygenates (e.g., in CFP of LDPE alone),
the hydrogen released during the transformation of hydrogen-
rich olefins to hydrogen-deficient hydrocarbons (dienes, cyclo
diolefins, and aromatics) would be transferred to other olefins,
converting them to alkanes [34,35,43]. However, in co-feed CFP,
biomass-derived oxygenates may act as hydrogen acceptors to
accept the hydrogen released during aromatization of olefins. This
would protect other olefins from being hydrogenated to alkanes.
This inference is corroborated by the trend observed for alkanes
mYbioCbio% + nYpCp%
Yadditive, (m:n) (%) =
× 100
(5)
(
m + n)Cmix
%
where m and n are the mass ratio of biomass and plastic in co-feed
mixture, respectively; Ybio and Yp are the average carbon yields
determined in CFP of biomass and plastic, respectively; Cbio, Cp,
and Cmix% are the carbon contents in the biomass, plastic and their
mixture (biomass-to-plastic ratio of m:n).
As shown in Fig. 3, the experimental yields were considerably
higher than the additive yields for aromatics and olefins. In con-
trast, the experimental yields were lower than the additive yields
for alkanes, carbon oxides, and coke. This result indicates that
co-feeding of cellulose with LDPE has a beneficial effect on improv-
ing the production of valuable petrochemicals (i.e., aromatics and
olefins) and decrease the formation of less valuable alkanes and
undesired coke in CFP.
Fig. 2. Aromatic formation from Diels–Alder reactions between cellulose-derived furans and LDPE-derived olefins in co-feed CFP.

1
78
X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
Fig. 3. Comparison between the experimental yields (᭹) and the calculated additive yields (ꢀ) in catalytic fast pyrolysis of the cellulose and LDPE mixture (1:1).
(
Fig. 3(c)), which shows that the experimental yields in co-feed
mixing ratio of 20–50 wt.%) (see Table S9 in SD for the product dis-
tribution). In addition, to evaluate how the feed ratio influence the
synergy between the two feedstocks, the experimental and addi-
tive yields of aromatics, olefins, alkanes, carbon oxides, and solid
residues in these co-CFP tests are compared in Fig. 4. It is noted that
the yields are plotted against the H/Ceff of the feedstocks, which
increased from 0.7 to 1.4 with the increasing mass ratio of LDPE in
the mixture from 20 wt.% to 50 wt.%.
As the H/Ceff of the feedstock was increased (i.e., the feed
ratio of LDPE was increased), the yields of olefins and alkanes
increased monotonically, while the yields of coke and carbon oxides
decreased (see Table S9 in SD or the experimental yield curves in
Fig. 4). However, the aromatic yield went through a maximum at
cellulose-to-LDPE ratio of 2 (H/Ceff = 1.0). Consequently, the high-
est yield (55.6 C%) of petrochemicals (aromatics plus olefins) was
obtained at a cellulose-to-LDPE ratio of 2.
As Fig. 4 shows, the calculated additive yields changed linearly
with the H/Ceff of the feedstocks. However, the experimental yields
deviated considerably from these linear trends, indicating that sig-
nificant chemical interactions occur between cellulose and LDPE.
At all three co-feeding ratios, the experimental yields were con-
siderably higher than the additive yields for aromatics (Fig. 4(a)),
but lower for alkanes, carbon oxides, and coke (Figs. 4(c–e)). This
result is consistent with the previous result that co-feeding of cellu-
lose with LDPE enhances the production of valuable aromatics and
decreases the formation of less valuable (alkanes) and undesired
(coke) products (Fig. 3).
CFP were much lower than the additive yields. In addition, the
experimental yields of carbon oxides (CO and CO ) were also lower
2
than the additive yields (Fig. 3(d)), indicating that the formation
of carbon oxides is decreased in co-feed CFP. Previous studies
have indicated that the oxygen in biomass feedstocks are removed
mainly as CO, CO , and H O in CFP [20,39,44]. Considering the oxy-
gen balance, the trend observed for carbon oxides suggests that the
oxygen removal is shifted to a dehydration route instead of decar-
bonylation and decarboxylation in co-feed CFP, agreeing with the
above inference that cellulose-derived oxygenates may accept the
2
2
hydrogen released during olefin aromatization to form H O.
2
It was noticed that the aromatic yield increased monotoni-
◦
cally as the reaction temperature was increased from 450 C to
7
a plateau at 550 C in co-feed CFP (see Table 3). Coincidentally,
Cheng and Huber [39] reported that the highest aromatic yield
from Diels–Alder reactions of furan with propylene over ZSM-5
was obtained at 550 C within the tested temperature range of
3
tions, which convert cycloadducts back to starting materials (diene
and dienophile), become more common [45,46]. Aromatic produc-
tion via the Diels–Alder reaction pathway may thus become less
◦
50 C in CFP of cellulose and LDPE individually, whereas it reached
◦
◦
◦
00–600 C [39]. At higher temperatures, reverse Diels–Alder reac-
◦
favored as the reaction temperature is increased above 600 C. This
may explain why the aromatic yield in co-CFP did not increase as
the reaction temperature was increased above 550 C (Fig. 3(a)).
◦
The above results show that co-feeding of cellulose with LDPE
(
1:1) in CFP enhances petrochemical production and decreases
Notably, the aromatic yield went through a maximum at
cellulose-to-LDPE ratio of 2 (H/Ceff = 1.0). This trend indicates that
there exists an optimal cellulose-to-LDPE ratio for aromatic pro-
duction in co-feed CFP. It is well-known that Diels–Alder reactions
occur only when the diene component is in the s-cis conforma-
tion [47]. Furans are permanently in the s-cis conformation, and
are therefore a good source of diene for Diels–Alder reactions with
olefins. The existence of cellulose-derived furans can thus promote
coke formation considerably (see Fig. 3). Because biomass is much
more abundant and cheaper than waste plastics, it is desirable to
increase the mass ratio of biomass in co-feed mixtures if this will
not decrease petrochemical production significantly. To evaluate
how the cellulose-to-LDPE ratio influences the product distribu-
tions in co-feed CFP, we investigated CFP of cellulose and LDPE
mixtures with different mixing ratios between 4 and 1 (i.e., LDPE
Fig. 4. Comparison between (᭹) the experimental yields and (ꢀ) the calculated additive yields of the major products in catalytic fast pyrolysis of cellulose, LDPE, and their
◦
mixture with different mixing ratios (reaction temperatures: 550 C).

X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
179
◦
Fig. 5. (a) Aromatic and (b) olefin selectivity in catalytic fast pyrolysis of cellulose, LDPE, and their mixture with different mixing ratios (reaction temperature: 550 C).
the conversion of LDPE-derived olefins to aromatics in co-feed CFP,
and vice versa. However, when cellulose or LDPE is fed in excess rel-
ative to its counterpart, the conversion of excess furans or olefins to
aromatics requires multiple reaction steps and is thus less efficient
cellulose-to-LDPE ratio was decreased in co-feed CFP (Fig. 5(b)).
This trend can be easily rationalized because LDPE produced con-
siderable fractions of higher olefins in CFP (e.g., C4 and C5 olefin
selectivity of 25.3% and 19.4%, respectively).
(
see SD Fig. S6 for the reaction pathways of aromatic formation in
The results presented above indicate that co-feeding of cellulose
with LDPE can greatly enhance the production of valuable petro-
chemicals (especially aromatics) and decrease the coke formation
in CFP (see Figs. 3 and 4). In addition, aromatic distribution was
also improved by co-feeding, which increased the selectivity for
more valuable monoaromatics (Fig. 5(a)). Increasing the co-feed
ratio of LDPE (i.e., decreasing the cellulose-to-LDPE ratio) decreases
the coke yield steadily (Fig. 4(e)), and can thus decrease the rate of
catalyst deactivation in CFP. However, the petrochemical yield (aro-
matics plus olefins) was similar at the cellulose-to-LDPE ratios of
4–1 (52.1, 55.6, and 53.0 C% for the ratio of 4, 2, and 1, respectively,
see Table S9 in SD). Thus, from the perspective of petrochemical
production, only small quantities of LDPE (e.g., 20 wt.%) are needed
to co-feed with cellulose. Adding more LDPE will not increase pet-
rochemical yield considerably, and thus be a waste of LDPE since it
is less available than biomass.
CFP of cellulose and LDPE). For example, when LDPE is fed in excess
relative to cellulose, the excess olefins need to undergo a series of
reactions such as oligomerization, hydrogen transfer, cyclization,
and aromatization to form aromatics [35]. Meanwhile, some olefins
would be converted to alkanes in the hydrogen transfer reactions
(
similar to what occur in CFP of individual LDPE) [35]. The results
presented in Fig. 4 suggest that LDPE is probably overly fed rela-
tive to cellulose at the cellulose-to-LDPE ratio of 1, which produced
less aromatics (42.1 C%) but considerably more alkanes (26.6 C%)
than those (48.4 C% for aromatics and 14.5 C% for alkanes) at the
cellulose-to-LDPE ratio of 2.
For olefins, the experimental yields were slightly lower than the
additive yields at cellulose-to-LDPE ratios of 4 and 2, but higher
than the additive yields at cellulose-to-LDPE ratio of 1 (Fig. 4(b)).
This result suggests that olefin production is sensitive to the co-
feed ratio of cellulose and LDPE in CFP. As discussed previously,
LDPE-derived olefins would then be consumed in Diels–Alder reac-
tions with cellulose-derived furans to aromatics in co-feed CFP.
However, the existence of cellulose-derived oxygenates can also
protect olefins from being hydrogenated to alkanes. The trend
observed for olefins suggests that the first mechanism dominates at
high cellulose-to-LDPE ratios (e.g., 4 and 2), whereby LDPE-derived
olefins would be largely consumes in Diels–Alder reactions with
cellulose-derived furans. However, the latter mechanism becomes
more important at low cellulose-to-LDPE ratios (e.g., 1), whereby
LDPE produces excess olefins relative to cellulose-derived furans
for the Diels–Alder reactions.
3.5. Effects of types of biomass and plastics on co-feed CFP
The above results indicate that LDPE and cellulose have a
significant synergy that improve the production of valuable pet-
rochemicals and decrease coke formation in CFP. To determine
whether the similar synergistic effects exist between other plastics
and lignocellulosic biomass, we further investigated co-feeding of
cellulose with PP or PS (another two major components of waste
plastics), as well as co-feeding of LDPE with lignin or pine wood
sawdust in CFP. The biomass-to-plastic mass ratio was 2 for all
co-feed tests discussed below.
The co-feed ratio of cellulose and LDPE influenced the distribu-
tion of aromatics and olefins considerably. As the cellulose-to-LDPE
ratio was decreased (i.e., more LDPE in the co-feed mixture),
the selectivity for most monoaromatics (toluene, xylenes, and
other alkylbenzenes) increased considerably, while the selectiv-
ity for polyaromatics declined steadily (Fig. 5(a)). Previous studies
have indicated that in CFP of biomass, polyaromatics are formed
mainly from the reactions of monoaromatics with small oxy-
genates [20,39]. Co-feeding of cellulose with LDPE in CFP can thus
protect monoaromatics from being converted to polyaromatics
because LDPE-derived olefins would compete with monoaromatics
for reacting with cellulose-derived oxygenates. This result indicates
that co-feeding of cellulose with LDPE in CFP can improve the dis-
tribution of aromatic products because monoaromatics (especially
BTX compounds) are more valuable products than polyaromat-
ics. The olefin selectivity was shifted toward higher olefins as the
The major products in these co-feed tests are summarized in
Table 4. Co-feeding of the different combinations of biomass and
plastics produced 23.3–56.6 C% petrochemicals. The highest pet-
rochemical yield was obtained for co-feeding of cellulose with PS.
By contrast, co-feeding of lignin with LDPE produced the least pet-
rochemicals and the highest yield of solid residues (40.5 C%). To
evaluate how the interactions between the biomass and plastics
influenced the product distribution in co-feed CFP, the experimen-
tal and additive yields of major products in the co-feed tests are
compared in Fig. 6.
Compared with cellulose/LDPE (Fig. 6(a)), co-feeding of cellu-
lose with PP in CFP achieved a much smaller synergy for aromatic
production (the experimental and additive yields were 37.5 C% and
3
4.1 C%, respectively) (Fig. 6(b)). Consequently, although CFP of PP
alone produced more aromatics (35.4 C%) than CFP of LDPE alone
28.3 C%) (see Table 2), co-CFP of PP with cellulose produced much
(

1
80
X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
Table 4
Product distribution in co-feed CFP of the different combinations of biomass and plastics (reaction temperature: 550 C; biomass-to-plastic: 2).
◦
Feedstock
H/Ceff
Cellulose/LDPE
1.0
Cellulose/PP
1.0
Cellulose/PS
0.5
Lignin/LDPE
1.1
Pine/LDPE
1.1
Overall yield (C%)
Aromatics
Olefins
Alkanes
CO
48.4
7.21
14.5
9.15
2.00
18.1
55.6
37.5
9.56
21.9
11.1
2.74
18.9
47.1
54.4
2.16
1.90
10.9
2.54
23.4
56.6
14.1
9.17
16.1
1.13
1.17
40.5
23.3
39.9
9.58
18.4
6.20
1.20
19.5
49.5
CO2
Solid (coke and char)
Petrochemicals
Aromatic selectivity (%)
Benzene
Toluene
11.4
30.6
23.2
11.0
0.0
13.3
31.4
20.7
9.5
0.0
3.2
34.4
15.0
6.0
6.4
0.0
12.8
33.6
19.8
5.0
0.2
2.5
11.4
31.9
26.5
8.1
0.0
3.8
Xylenesa
b
Other alkylbenzenes
c
Phenols
Indenesd
4.0
19.9
7.2
30.9
Polyaromaticse
21.9
26.2
18.4
Olefin selectivity (%)
Ethylene
Propylene
C4 olefins
C5 olefins
30.5
42.8
17.1
9.7
35.2
43.9
15.8
5.2
52.3
35.6
8.8
38.1
44.1
13.0
4.8
32.6
41.0
17.3
9.1
3.4
a
Xylenes include m-, p-, and o-xylene.
Other alkylbenzenes include ethylbenzene, trimethylbenzenes, tetramethylbenzenes, propenylbenzene, and butenylbenzene.
Phenols include phenol, cresols, and guaiacol.
b
c
d
Indenes include indane, indene, 2,3-dihydro-2,2-dimethyl-1H-indene, 2,3-dihydro-4-methyl-1H-indene, 2,3-dihydro-1,3-dimethyl-1H-indene, 2-methylindene, and 1,3-
dimethyl-1H-indene.
e
Polyaromatics include naphthalene, methylnaphthalenes, dimethylnaphthalenes, and polycyclic aromatic hydrocarbons (benzene rings >2).
less aromatics (37.5 C%) than co-CFP of LDPE with cellulose (48.4
C%). As shown in Fig. S2 in SD, LDPE produced linear ␣-olefins
as the major primary pyrolysis products, whereas PP produced
predominantly branched olefins (e.g., 2,4-dimethyl-1-heptene, (Z)-
and aldehydes [7]) can act as hydrogen acceptors to receive the
hydrogen released during aromatization of LDPE-derived olefins,
which prevents other olefins from being hydrogenated to alkanes.
However, no significant differences were observed for the other
major products (aromatics, carbon oxides, and coke). This result
suggests that lignin and LDPE have only weak interactions in the
co-feed CFP. As shown in Fig. S2 in SD, the primary pyrolysis prod-
ucts of lignin were predominantly phenolics, which have a high
tendency for adsorption onto the acid sites of ZSM-5 leading to
coke formation [10]. They may thus not react significantly with
LDPE-derived products in the co-feed CFP.
4
-methyl-2-pentene, and 4-methyl-2-undecene) (see Fig. S2 in SD).
It has been suggested that branched olefins are less reactive than
linear ␣-olefins in Diels–Alder reactions due to steric hindrance
and electronic effects of their branchings [48]. This may explain
why cellulose and PP have much smaller synergy for aromatic pro-
duction (Fig. 6(b)) than cellulose and LDPE (Fig. 6(a)), although
both LDPE and PP produces significant amounts of hydrogen-rich
olefins as the primary pyrolysis products. This comparison also sug-
gests that Diels–Alder reactions between cellulose-derived furans
and LDPE-derived ␣-olefins play an important role in enhancing
aromatic production in the co-feed CFP of cellulose/LDPE.
A significant synergy for petrochemical production and coke
reduction was observed when pine wood sawdust and LDPE were
co-fed in CFP (Fig. 6(e)). As natural biomass, the pine wood
contained 46.1 wt.% cellulose, 24.7 wt.% lignin, and 15.8 wt.% hemi-
cellulose. The interactions between pine wood and LDPE in co-feed
CFP are therefore expected to be more complex than those in
co-feeding of individual biomass components (e.g., cellulose and
lignin) with LDPE. For example, the primary pyrolysis products
of cellulose (e.g., furans) can react with LDPE-derived olefins
to form aromatics. In addition, thermal decomposition of hemi-
cellulose produced significantly amounts of high-oxygen-content
oxygenates such as acetic acid and hydroxyacetaldehyde [7]. These
oxygenates have a low H/Ceff (e.g., zero for acetic acid) and produce
large quantities of catalytically derived coke during zeolite upgrad-
ing [8,12]. However, coke formation can be decreased significantly
when these oxygenates are co-fed with hydrogen-rich compounds
(e.g., methanol, which has a H/Ceff = 2) [6,15,17]. This decrease
is mainly because hydrogen-rich compounds can provide hydro-
gen to biomass-derived oxygenates during their zeolite-catalyzed
conversion processes [6,42]. The result presented herein suggests
that similar hydrogen-transfer reactions may occur between LDPE-
derived hydrocarbons and pine wood-derived oxygenates when
pine wood and LDPE are co-fed in CFP. Consequently, co-feeding
of pine wood with LDPE in CFP decreased the coke formation con-
siderably.
Although co-feeding of PP with cellulose did not significantly
enhance the aromatic production, it improved the aromatic dis-
tribution toward more valuable monoaromatics (see Table 4). The
selectivity for toluene and xylenes was respectively increased from
2
5.1% and 15.6% for CFP of cellulose alone to 31.4% and 20.7% for the
co-feed CFP (see Tables 2 and 4). Conversely, the selectivity for less
valuable polyaromatics (naphthalenes) was decreased from 35.5%
to 21.9% by the co-feeding. Co-feeding of cellulose with PP thus has
a beneficial effect on improving product distribution in CFP.
When cellulose and PS were co-fed in CFP, insignificant dif-
ferences were observed between the experimental and additive
yields for all major products (Fig. 6(c)). This result suggests that
cellulose and PS have little interaction in the co-feed CFP. Unlike
LDPE and PP, PS produces aromatics (predominantly styrene) as
the major primary pyrolysis products [31]. Styrene is a notoriously
poor dienophile component for Diels–Alder reactions [49], and thus
may not react actively with cellulose-derived furans. Cellulose and
PS thus did not interact much in the co-feed CFP.
Fig. 6(d) shows that olefin production was enhanced, but alkane
production was decreased when lignin was co-fed with LDPE in CFP.
This is possibly because some lignin-derived oxygenates (e.g., acids

X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
181
The above results indicate that the interactions between
biomass and plastics in co-feed CFP are strongly dependent on the
types of feedstocks. When LDPE is co-fed with cellulose or pine
wood, they have a significant synergy that enhances petrochem-
ical production and decreases coke formation in CFP. However,
the other combinations of biomass and plastics (cellulose/PP, cel-
lulose/PS, and lignin/LDPE) tested in this study exhibited a much
smaller or no synergy for petrochemical production and coke
reduction in co-feed CFP. This difference can be mainly attributed
to the fact that LDPE-derived linear ␣-olefins can react readily
with cellulose-derived furans to form aromatics via Diels–Alder
reactions, whereas the major primary pyrolysis products of PP
(branched olefins) and PS (styrene) are not reactive with furans.
In addition, the primary pyrolysis products of lignin are mainly
phenolic compounds. They have high tendency to form coke over
ZSM-5 [10], and may thus not react actively with LDPE-derived
products in co-feed CFP. Consequently, co-feed CFP of cellulose/PP,
cellulose/PS, and lignin/LDPE did not result in a significant synergy
for petrochemical production as have been observed for CFP of cel-
lulose/LDPE and pine wood/LDPE. These results indicate that proper
selection of biomass and plastic feedstocks is crucial to optimize the
synergy between the two feedstocks in co-feed CFP.
4. Conclusions
The results of this study demonstrate that CFP is a robust con-
version process that can convert the different biomass (cellulose,
lignin, and pine wood), plastic (LDPE, PP, and PS), and their mixtures
to similar petrochemical products. Furthermore, when cellulose or
pine wood is co-fed with LDPE in CFP, they have a significant syn-
ergy that enhances petrochemical (especially aromatic) production
and decreases coke formation. This synergy has been attributed
mainly to the Diels–Alder reactions of cellulose-derived furans with
LDPE-derived olefins, which produce aromatics in the presence
of ZSM-5 zeolites. In addition, reactions such as hydrogen trans-
fer between plastic-derived hydrocarbons and biomass-derived
oxygenates may help to decrease coke formation during zeolite-
catalyzed conversion of hydrogen-deficient oxygenates. Because
cellulose is usually the most dominant component (40–50 wt.%) of
natural lignocellulosic biomass, similar synergy for petrochemical
production and coke reduction is expected to occur when natural
biomass and LDPE are co-fed in CFP (as has been shown for the co-
feed CFP of pine wood with LDPE). Co-feeding LDPE thus has a great
potential in improving the performance of CFP of natural biomass.
Acknowledgements
This research is supported by the National Key Technology
R&D Program (2010BAC66B03), the Funding of Shanghai Tobacco
(
Group) Corporation (2011-1-010), and the special fund of State Key
Joint Laboratory of Environment Simulation and Pollution Control
13Y01ESPCT).
(
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.apcata.
2
014.05.015.
References
[
1] E. Taarning, C.M. Osmundsen, X. Yang, B. Voss, S.I. Andersen, C.H. Christensen,
Energy Environ. Sci. 4 (2011) 793–804.
Fig. 6. Comparison between the experimental yields and the calculated additive
[2] P. Gallezot, Chem. Soc. Rev. 41 (2012) 1538–1558.
yields of the major products in co-feed CFP of (a) cellulose/LDPE, (b) cellulose/PP, (c)
cellulose/PS, (d) lignin/LDPE, and (e) pine wood/LDPE (reaction temperature: 550 C;
biomass-to-plastic: 2).
[3] D.M. Alonso, J.Q. Bond, J.A. Dumesic, Green Chem. 12 (2010) 1493–1513.
[4] T.R. Carlson, Y.-T. Cheng, J. Jae, G.W. Huber, Energy Environ. Sci. 4 (2011)
◦
145–161.

1
82
X. Li et al. / Applied Catalysis A: General 481 (2014) 173–182
[
5] T.R. Carlson, T.R. Vispute, G.W. Huber, ChemSusChem 1 (2008) 397–400.
6] H. Zhang, T.R. Carlson, R. Xiao, G.W. Huber, Green Chem. 14 (2012) 98–110.
7] K. Wang, K.H. Kim, R.C. Brown, Green Chem. 16 (2014) 727–735.
8] S. Czernik, R. French, Fuel Process. Technol. 91 (2010) 25–32.
[27] N. Miskolczi, L. Bartha, G. Deák, Polym. Degrad. Stab. 91 (2006) 517–526.
[28] I. Cit, A. Sinag, T. Yumak, S. Ucar, Z. Misirlioglu, M. Canel, Polym. Bull. 64 (2010)
817–834.
[29] L. Sojak, R. Kubinec, H. Jurdakova, E. Hajekova, M. Bajus, J. Anal. Appl. Pyrolysis
78 (2007) 387–399.
[
[
[
[
9] J. Li, X. Li, G. Zhou, W. Wang, C. Wang, S. Komarneni, Y. Wang, Appl. Catal. A 470
(
2014) 115–122.
[30] P.F. Greenwood, Org. Geochem. 37 (2006) 755–771.
[31] D.P. Serrano, J. Aguado, J.M. Escola, Appl. Catal. B 25 (2000) 181–189.
[32] D. Mohan, C.U. Pittman, P.H. Steele, Energy Fuels 20 (2006) 848–889.
[33] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Ind. Eng. Chem. Res.
43 (2004) 2610–2618.
[34] J. Aguado, D.P. Serrano, J.L. Sotelo, R. Van Grieken, J.M. Escola, Ind. Eng. Chem.
Res. 40 (2001) 5696–5704.
[35] D.B. Lukyanov, N.S. Gnep, M.R. Guisnet, Ind. Eng. Chem. Res. 33 (1994) 223–234.
[36] D.P. Serrano, J. Aguado, J.M. Escola, J.M. Rodriguez, G. San Miguel, J. Anal. Appl.
Pyrolysis 74 (2005) 370–378.
[
[
10] C.A. Mullen, A.A. Boateng, Fuel Process. Technol. 91 (2010) 1446–1458.
11] H. Zhang, Y.-T. Cheng, T.P. Vispute, R. Xiao, G.W. Huber, Energy Environ. Sci. 4
(
2011) 2297–2307.
[
[
[
12] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, M. Olazar, J. Bilbao, Ind. Eng.
Chem. Res. 43 (2004) 2619–2626.
13] X.Y. Li, L. Su, Y.J. Wang, Y.Q. Yu, C.W. Wang, X.L. Li, Z.H. Wang, Front. Environ.
Sci. Eng. 6 (2012) 295–303.
14] K.K. Ramasamy, M.A. Gerber, M. Flake, H. Zhang, Y. Wang, Green Chem. 16
(
2013) 748–760.
[
[
[
15] N.Y. Chen, T.F. Degnan, L.R. Koenig, Chemtech 16 (1986) 506–511.
16] J.D. Adjaye, N.N. Bakhshi, Biomass Bioenergy 8 (1995) 131–149.
17] A.G. Gayubo, B. Valle, A.s.T. Aguayo, M. Olazar, J. Bilbao, Ind. Eng. Chem. Res. 49
[37] R.C. Mordi, J. Dwyer, R. Fields, Polym. Degrad. Stab. 46 (1994) 57–62.
[38] R. Lin, R.L. White, J. Appl. Polym. Sci. 63 (1997) 1287–1298.
[39] Y.-T. Cheng, G.W. Huber, Green Chem. 14 (2012) 3114–3125.
[40] C.L. Williams, C.-C. Chang, P. Do, N. Nikbin, S. Caratzoulas, D.G. Vlachos, R.F.
Lobo, W. Fan, P.J. Dauenhauer, ACS Catal. 2 (2012) 935–939.
[41] N. Nikbin, P.T. Do, S. Caratzoulas, R.F. Lobo, P.J. Dauenhauer, D.G. Vlachos, J.
Catal. 297 (2013) 35–43.
[42] H.Y. Zhang, J. Zheng, R. Xiao, D.K. Shen, B.S. Jin, G.M. Xiao, R. Chen, RSC Adv. 3
(2013) 5769–5774.
[43] F. Angelescu, G. Pop, G. Pogonaru, R. Stanescu, G. Musca, V. Pirvulescu, Rev.
Roum. Chim. 38 (1993) 411–423.
[44] A.J. Foster, J. Jae, Y.T. Cheng, G.W. Huber, R.F. Lobo, Appl. Catal. A 423 (2012)
154–161.
[45] W.G. Dauben, H.O. Krabbenhoft, J. Am. Chem. Soc. 98 (1976) 1992–1993.
[46] M. Jones, S.A. Fleming, Organic Chemistry, WW Norton & Company Incorpo-
rated, 2010.
[47] F. Fringuelli, A. Taticchi, Dienes in the Diels–Alder Reaction, Wiley, New York,
1990.
[48] L.H. Slaugh, L.A. Fenouil, H.L.-H. Fong, Process for Separating Linear Alpha
Olefins from 2-Branched and/or 3-Branched Alpha Olefins, 1999.
[49] L.E. Overman, R.L. Freerks, C.B. Petty, L.A. Clizbe, R.K. Ono, G.F. Taylor, P.J. Jessup,
J. Am. Chem. Soc. 103 (1981) 2816–2822.
(
2009) 123–131.
18] X. Li, H. Zhang, J. Li, L. Su, J. Zuo, S. Komarneni, Y. Wang, Appl. Catal. A 455 (2013)
14–121.
[
[
1
19] H. Goering, P.J. Van Soest, Forage Fiber Analyses, Apparatus, Reagents, Proce-
dures, and Some Applications, US Agricultural Research Service, Washington,
1
970.
20] T.R. Carlson, J. Jae, Y.-C. Lin, G.A. Tompsett, G.W. Huber, J. Catal. 270 (2010)
10–124.
[
[
[
[
[
[
[
1
21] C. Torri, M. Reinikainen, C. Lindfors, D. Fabbri, A. Oasmaa, E. Kuoppala, J. Anal.
Appl. Pyrolysis 88 (2010) 7–13.
22] Y. Yu, Y. Zeng, J. Zuo, F. Ma, X. Yang, X. Zhang, Y. Wang, Bioresour. Technol. 134
(
2013) 198–203.
23] Y.C. Lin, J. Cho, G.A. Tompsett, P.R. Westmoreland, G.W. Huber, J. Phys. Chem. C
13 (2009) 20097–20107.
24] Y.Q. Yu, X.Y. Li, L. Su, Y. Zhang, Y.J. Wang, H.Z. Zhang, Appl. Catal. A 447 (2012)
15–123.
1
1
25] P. Bhattacharya, P.H. Steele, E.B.M. Hassan, B. Mitchell, L. Ingram, C.U. Pittman,
Fuel 88 (2009) 1251–1260.
26] M. Brebu, S. Ucar, C. Vasile, J. Yanik, Fuel 89 (2010) 1911–1918.