Reaction routes in catalytic reforming of poly(3-hydroxybutyrate) into renewable hydrocarbon oil

Article abstract of DOI:10.1039/c5ra03195h

Poly(3-hydroxybutyrate) or PHB is an energy storage material of microbial organisms and can be reformed into hydrocarbon oils rich with aromatic compounds. This work investigated the main reaction routes from PHB to the key intermediates and final hydrocarbons. The main sequential reactions under catalysis of phosphoric acid at moderate temperatures (200-230 °C) consist of: (1) decomposition of PHB into crotonic acid, a major monomeric intermediate, (2) deoxygenation of crotonic acid, and (3) combination of the deoxygenated molecules. The oxygen in PHB is removed as CO₂ and H₂O in stage (2), involving decarboxylation and ketonization of crotonic acid. The main aromatic compounds are formed in stage (3) from propylene and 2,3-dimethyl-2-cyclopenten-1-one as two key intermediates, the former from decarboxylation and the latter from ketonization of crotonic acid. The reaction routes reveal that the formation of aromatics is affected to a great extent by the concentrations of phosphoric acid and water in the reaction, which can be used to control the composition of hydrocarbon oil.

Full text of DOI:10.1039/c5ra03195h

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Reaction routes in catalytic reforming of poly(3-
hydroxybutyrate) into renewable hydrocarbon oil†
Cite this: RSC Adv., 2015, 5, 30005
*
Shimin Kang and Jian Yu
Poly(3-hydroxybutyrate) or PHB is an energy storage material of microbial organisms and can be reformed
into hydrocarbon oils rich with aromatic compounds. This work investigated the main reaction routes from
PHB to the key intermediates and final hydrocarbons. The main sequential reactions under catalysis of
phosphoric acid at moderate temperatures (200–230 ꢀC) consist of: (1) decomposition of PHB into
crotonic acid, a major monomeric intermediate, (2) deoxygenation of crotonic acid, and (3) combination
of the deoxygenated molecules. The oxygen in PHB is removed as CO2 and H2O in stage (2), involving
decarboxylation and ketonization of crotonic acid. The main aromatic compounds are formed in stage
(3) from propylene and 2,3-dimethyl-2-cyclopenten-1-one as two key intermediates, the former from
decarboxylation and the latter from ketonization of crotonic acid. The reaction routes reveal that the
formation of aromatics is affected to a great extent by the concentrations of phosphoric acid and water
in the reaction, which can be used to control the composition of hydrocarbon oil.
Received 19th February 2015
Accepted 12th March 2015
DOI: 10.1039/c5ra03195h
www.rsc.org/advances
dependent on the composition of biomass feedstock which
varies with season and location.
1 Introduction
Poly(3-hydroxybutyrate), PHB (C4H6O2)n (n ¼ 6000–12 000)
Transportation fuels derived from biomass could reduce fossil
fuel consumption and CO2 emission since the renewable feed-
stock originates from CO2, water and sunlight. Because of the
high oxygen content of biomass, however, bio-based fuels such
as ethanol (C2H6O) contain a substantial amount of oxygen
which not only reduces the fuel's energy content but also
requires advanced engine technologies.1 Bio-based drop-in
gasoline that has the same chemical composition of the petro-
chemical counterpart could solve the problem, but presents
technical challenges to reneries including: deoxygenation of
biofuel to a desired low oxygen content and reforming of the
chemical structure of hydrocarbons for high antiknock quality
(high octane number). In general, straight-chain paraffins are
more prone to knocking than branched-chain paraffins, olens
and cyclic hydrocarbons. Likewise, cycloparaffins are more
knock-prone than aromatics. To satisfy modern high perfor-
mance automotive engines, for example, unleaded gasoline
is an energy storage material of many microbial species and
can be produced through fermentation of renewable carbo-
hydrates or CO2/H2.4 Recently, we reported a simple one-pot
conversion of PHB into hydrocarbon oil from which a
gasoline-like drop-in fuel was obtained through conventional
distillation. In the reaction catalyzed with H3PO4, PHB was
completely converted and deoxygenated to a great extent with
formation of CO2 and propylene (C3H6), and more interest-
ingly with formation of substantial amount of aromatic
compounds.5 We found that crotonic acid (CA, trans-2 bute-
noic acid, C4H6O2), was a key monomeric intermediate in the
early stage of PHB conversion. Under typical reaction condi-
tions (100% H3PO4 solution, 220 ꢀC and 3 hours), CA was
completely converted and 90.9% of the original carbon was
recovered from the gas and liquid products.5 We, however,
had little knowledge on how CA was deoxygenated and how
aromatic rings were formed under the reaction conditions.
Compared to the conventional biomass feedstock, PHB is a
simple consistent raw material and can be conveniently con-
verted into drop-in fuels or aromatics that can be blended
with conventional bio-fuels such as ethanol and fatty acids
methyl ester to improve the quality of biofuels. The reaction
mechanism of PHB reforming, especially the decarboxylation
of CA and formation of aromatic compounds, would help us
understand and control the new reaction system. In this work,
we explored the reaction routes from PHB to the key inter-
mediates and major nal hydrocarbons.
contains
a substantial amount (maximum 35 vol%) of
aromatics that replace lead-containing additives and sustain a
high octane number.2 Conventional thermochemical process-
ing of biomass may produce a drop-in fuel, but needs multiple
catalytic reactions and consumes a large amount of hydrogen.3
Besides, the thermochemical process conditions are very much
Hawaii Natural Energy Institute, University of Hawaii at Manoa, Honolulu, Hawaii
96816, USA. E-mail: jianyu@hawaii.edu
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c5ra03195h
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 30005–30013 | 30005

View Article Online
RSC Advances
Paper
estimated from the ratio of peak areas of isocrotonic acid to CA
multiplied by the yield of CA.
2 Experimental section
2.1 Materials
Crotonic acid (CA, 98%), 2,3-dimethyl-2-cyclopenten-1-one
(99%), 3-methyl-2-cyclohexenone (>98%), 3-butenoic acid
(97%), 3-hydroxybutyric acid (3HB, 98%), toluene (99.8%) and
polyphosphoric acid (115% H3PO4 basis) were purchased from
Sigma-Aldrich (St Louis, MO, USA). Butyric acid (99%) was
obtained from Alfa Aesar (MA, USA). Solutions of phosphoric
acid (50% to 100% H3PO4) were prepared from polyphosphoric
acid and distilled de-ionized water. Poly(3-hydroxybutyrate)
(PHB, 98%) was obtained from Bio-on (Bologna, Italy). Its
weight-average molecular weight (138 000 Da) and number-
average molecular weight (53 100 Da) were measured with gel
permeation chromatography (GPC) and calibrated with poly-
styrene standards. The polystyrene standards with narrow
molecular weight distribution were purchased from Sigma-
Aldrich.
2.3 Gas formation and analysis
In a typical experiment for gas products and analysis, 3.6 g of
PHB or crotonic acid and 72 mL of H3PO4 solution were added
into a 180 mL pyro-beaker, which was put in a 600 mL autoclave
(Parr Instrument, IL, USA). The autoclave was purged with
nitrogen at least ten times of the reactor volume to remove air.
The reactor was then sealed and heated to a pre-determined
temperature. When the temperature reached the setting value
in about half an hour, the reaction time was set as zero, and
thereaer recorded. Aer reaction, the reactor was cooled down
in ambient conditions. The quantitative determination of CO2
and propylene was performed by using a gas chromatograph
equipped with a thermal conductivity detector (GC-TCD, Bruker
450-GC, FL, USA) and a Carboxen-1006 Plot column. Both CO2
and propylene were calibrated with pure gases against helium.
2.4 Reactions of propylene
2.2 Liquid products formation and analysis
The reactions of propylene as a sole or co-reagent were con-
ducted in the 600 mL autoclave in which a 180 mL pyro-beaker
contained a phosphoric acid solution and co-reagents. Aer
purge with nitrogen, propylene was introduced into the reactor
till an initial pressure of 760 kPa. The reaction was conducted at
230 ꢀC for 3 hours. For propylene as the sole reagent, 50 mL of
100% H3PO4 was put in the pyro-beaker. For the reactions of
propylene with 2,3-dimethyl-2-cyclopenten-1-one (or 3-methyl-
2-cyclohexenone), 1 g of 2,3-dimethyl-2-cyclopenten-1-one (or
3-methyl-2-cyclohexenone) and 30 mL of 100% H3PO4 were put
in the beaker. For the reaction of propylene with toluene, 8 g of
toluene and 50 mL of 100% H3PO4 were put in the pyro-beaker.
Aer the reactions, the reactor was cooled down in ambient
conditions. The liquid products were recovered with methylene
chloride for GCMS analysis.
In oil-producing experiments, a pre-determined amount (0.5–
1 g) of reagents (PHB, CA, 3HB) and 10 mL of H3PO4 solutions
were put into a 20 mL polytetrauoroethylene (PTFE) reactor,
and the reactor was purged with N2 for about 10 minutes. The
PTFE reactor was then sealed and le for a predetermined
reaction time in a thermostat oven that was maintained at a
desired temperature (165 to 240 ꢀC). Aer reaction, the PTFE
reactor was quickly cooled down under tap water. The reaction
ꢀ
under the conditions of 100% H3PO4 solution at 230 C for 3
hours is dened as a typical reaction thereaer. In controlled
reactions with butyric acid (or 2,3-dimethyl-2-cyclopenten-1-one
or 3-methyl-2-cyclohexenone), 1 g of butyric acid (or 2,3-
dimethyl-2-cyclopenten-1-one, or 3-methyl-2-cyclohexenone)
and 10 mL of 100% H3PO4 solution were put into a 20 mL
polytetrauoroethylene (PTFE) reactor, and the reaction was
ꢀ
maintained at 230 C for 3 hours. For the 2-propanol reaction
3 Results and discussion
3.1 Monomeric intermediates
with 2,3-dimethyl-2-cyclopenten-1-one, 0.5 g 2-propanol and
0.3 g 2,3-dimethyl-2-cyclopenten-1-one were put in 10 mL 100%
ꢀ
H3PO4 at 230 C for 3 hours. In order to check acylation reac- CA and 3HB are two possible monomeric intermediates of PHB
tion, toluene (2 g) and CA (0.5 g) were reacted in a solution of degradation and their predominance depends on reaction
ꢀ
100% H3PO4 (0.5 g) at 150 C for 3 hours.
conditions. Under pyrolysis conditions, PHB is primarily
The liquid products were recovered with methylene chloride decomposed into CA, via a b-elimination six-member ring
and analyzed with a gas chromatograph-mass spectrometer structure (Scheme 1).6,7 Under acidic hydrolysis conditions,
(GC-MS, Bruker 436-GC, USA) equipped with an Rtx-1 capillary both 3HB and CA are formed.8 The molar ratio of 3HB to CA
column and EI ion source (70 eV). The temperature of the reects the relative extent of hydrolysis versus pyrolysis of PHB.
ꢀ
injector was set at 250 C and the temperature program of the As shown in Table 1, the molar ratio of CA to 3HB in the reaction
ꢀ
ꢀ
column was started at 30 C for 10 min, increased to 230 C at solutions (50–70% H3PO4) at 220 ꢀC is much lower than that in
8 ꢀC minꢁ1 anꢁd1held for 2 min, and further increased to 260 ꢀC the pyrolysis of PHB. It indicates that the acidic hydrolysis,
at 10 ꢀC min and held for 10 min. The compounds were instead of the b-elimination six-member ring structure could be
identied by using a NIST11 mass spectral data library. The the main route of PHB degradation in the H3PO4 solutions. The
concentrations of organic acids (CA, isocrotonic acid, 3-bute- double bond of CA may be formed from dehydration of 3HB via
noic acid, and 3-hydroxybutyric acid) in H3PO4 solutions were b-elimination.6 Table 1 also shows that 3HB dehydration was
determined with a high performance liquid chromatograph promoted with increase of H3PO4 concentration and tempera-
(HPLC, Shimadzu, Japan), aer the reaction solution was ture. In a solution of 100% H3PO4, CA was the dominant
appropriately diluted. The yield of isocrotonic acid was monomeric product even at relative low temperatures (165–
30006 | RSC Adv., 2015, 5, 30005–30013
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Scheme 1 CA formation via six-membered ring state of PHB in pyrolysis condition.6,7
Table 1 The yields of monomeric products from PHB degradation under different reaction conditions
H3PO4 (%)
Temperature (ꢀC)
Time (h)
3HB (wt%)
CA (wt%)
CA/3HB (mol/mol)
100
100
100
100
85
85
70
50
165
180
220
220
220
210
220
220
3
3
3
1
3
3
3
3
2.9
0.2
nga
ng
28.8
3.4
ng
1.3
ng
0.2
3.7
12.0
12.0
20.6
—
—
—
0.8
0.4
0.2
ng
0.3
11.2
72.6
a
ng ¼ negligible.
ꢀ
180 C). In a reaction under the conditions of 100% H3PO4 at identied with GCMS analysis as shown in Fig. 1A. They are
220 ꢀC for 3 hours, both CA and 3HB were completely converted 1,2,3,4-tetramethyl-5-(1-methylethyl)-benzene (compound
1
into gas and liquid products. In the following experiments, the thereaer), 1,2,3,4-tetramethylbenzene (compound 2 there-
ꢀ
typical reaction conditions (100% H3PO4, 230 C and 3 hours) aer), 1,2,3,5-tetramethylbenzene (compound 3 thereaer), and
were used for analysis of nal products. We also used mild 1-indanone,3,3,4,5,7-pentamethyl-benzene (compound 4 there-
reaction conditions (e.g., 85% H3PO4, 230 ꢀC and 1 hour) to nd aer). In the typical reaction, compound 1 is the predominant
the key intermediates that might appear in the early stage of the compound of hydrocarbon oil and has a similar chemical
reaction. Fig. 1 reveals an interesting fact with GCMS analysis structure of compounds 2 and 3, probably involving the same
that PHB, CA and 3HB, when they were used as the initial reaction intermediate and mechanism.
reagent, were converted into the same nal products under the
typical reaction conditions. Fig. 1C1 and C2 reveal that CA was
3.3 Crotonic acid decarboxylation and propylene formation
the main intermediate of 3HB conversion and observed only in
Deoxygenation of PHB occurs via CA decarboxylation because
CO2 and propylene are the predominant gaseous products
under different reaction conditions as shown in Table 2. The
mole and weight yields of CO2 and propylene from PHB are
provided for comparison. Since little CO was detected in the gas
phase, decarbonylation of PHB or CA was negligible and not
considered in this study. According to the stoichiometry of CA
decarboxylation (C4H6O2 / C3H6 + CO2), equal molar concen-
trations of propylene and CO2 were expected, but the actual
molar ratio of CO2 to C3H6 ranged from 5.5 to 17.1, depending
on the reaction conditions. It might be attributed to further
conversion of propylene into nal products, but also to other
the early stage under the mild reaction conditions. Under the
typical reaction conditions, some 3HB might be formed
initially, but quickly converted into CA and further to nal
hydrocarbons. Our previous work also revealed that the same
gas products were formed from PHB and CA.5 Based on these
facts, it was concluded that CA was the key monomeric inter-
mediate in PHB reforming. In the following experiments, either
PHB or CA was used as the reagent to nd the intermediates and
mechanisms of the reactions following PHB degradation.
3.2 Final gas and liquid products
We rst check the nal gas and liquid products under different reaction routes in which CO2 is formed with little propylene
reaction conditions. It was found that PHB and CA were to a being formed.
great extent deoxygenated with CO2 and propylene (C3H6) being
Early studies have pointed out that decarboxylation of CA
the major products in the gas phase (Table 2) and aromatic occurs primarily via a six-member ring transition state (Scheme 2)
compounds (alkylbenzenes and naphthalenes) being the major in which CA was rst converted to 3-butenoic acid.9,10 It has also
compounds in the liquid oil (Table 3). Specically, the chemical been reported that isocrotonic acid is an essential intermediate
composition of liquid oil changed depending on the reaction in formation of 3-butenoic acid, a precursor to the six-member
conditions such as reaction temperature, time, and the ring structure intermediate.9 Very interestingly, both iso-
concentration of H3PO4 (Table 3). Four major compounds were crotonic acid and 3-butenoic acid were detected, especially in
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 30005–30013 | 30007

View Article Online
RSC Advances
Paper
Fig. 1 GCMS spectra of hydrocarbon oils and the chemical structures of compounds 1–4. (A) Spectrum of hydrocarbon oil from PHB in the
typical reaction conditions (100% H3PO4, 230 ꢀC, 3 hours); (B) spectrum of hydrocarbon oil from CA in the typical reaction conditions; (C)
spectrum of hydrocarbon oil from 3HB in the typical reaction conditions; (C1) spectrum of pure 3HB; (C2) spectrum of the reaction products in
the initial stage of 3HB conversion (85% H3PO4, 230 ꢀC, 1 hour).
Table 2 The yields of CO2 and propylene from PHB under different reaction conditions
CO2/propylene
(mol/mol)
H3PO4 (%)
Temperature (ꢀC)
Time (h)
CO2 (wt%)
aCO2 (mol%)
Propylene (wt%)
aPropylene (mol%)
70%
100%
115%
100
100
100
220
220
220
220
220
165
195
240
1
1
1
3
6
3
3
3
25.9
33.2
19.9
35.3
38.5
12.6
32.4
38.2
50.7
65.0
38.9
69.1
75.3
24.7
63.4
74.7
4.5
2.4
1.4
2.5
2.6
1.1
1.8
2.9
9.2
4.9
2.9
5.1
5.3
2.3
3.7
5.9
5.5
13.3
13.4
13.5
14.2
10.7
17.1
12.7
100
100
a
The molar yields of CO2 and propylene are based on the equivalent moles of CA in PHB.
the early stage of the reaction (Table 4). Based on the infor-
mation above, a reaction route is suggested for PHB decarbox-
ylation via CA (Scheme 1 and 2). It should be pointed out that
the reactions in Scheme 2 can also occur at high temperatures
3.4 Propylene oligomerization and alkylation
In contrast to inert CO2, propylene is an active reagent under
the reaction conditions and participates in formation of other
chemicals in hydrocarbon oil. In a controlled experiment with
propylene being the sole reagent, propylene was oligomerized to
C6–C12 alkenes determined with GCMS analysis (Fig. 2B), but no
aromatic compounds were formed. Interestingly, these alkenes
from propylene are quite similar to the alkenes obtained from
PHB reaction (Fig. 2A). This fact implies that the high content of
(e.g., 320 C) in the absence of H3PO4,9 but proceed quickly at
ꢀ
ꢀ
much lower temperatures (200–240 C) with the acid catalysis.
Most likely, the isomerization of CA to 3-butenoic acid in this
work is promoted by protons, because isomerization involves
electron transfer. In the absence of H3PO4 solution, 3-butenoic
acid was not found in the same reaction conditions (Table 4).
30008 | RSC Adv., 2015, 5, 30005–30013
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Table 3 Influence of reaction conditions on chemical composition of liquid oil from CA reactiona
Reaction conditions
O-containing compounds
Hydrocarbon compounds
Benzenes
H3PO4
(%)
Temperature
Time
(h)
Ketones
(%)
Othersb
(%)
Alkenes
(%)
Naphthalenes
(%)
(ꢀC)
Cyclicc
(%)
85
230
230
230
210
180
3
0.5
3
3
3
12.6
15.5
5.5
16.0
22.5
ngd
39.7
1.0
0.4
0.3
51.1
ng
1.6
1.4
1.0
5.3
4.8
3.8
3.4
5.1
30.6
40.0
78.8
72.4
70.6
ng
ng
9.3
6.4
0.6
100
100
100
100
a
The percentage composition is based on the relative peak area in the GCMS analysis. b The others include organic acids and esters, etc. c The cyclic
d
hydrocarbons did not include the aromatics. ng ¼ negligible.
Scheme 2 The proposed decarboxylation mechanism for propylene production.9,10
Table 4 Yields of intermediates in the early stage of CA decarboxylation in H3PO4 solutions at 230 ꢀC for 30 min
3-Butenoic acid
(3BA) (mol%)
Isocrotonic acid
(IA) (mol%)
Remaining CA
(mol%)
3BA : IA : CA
(mole ratio)
Solution
100% H3PO4
85% H3PO4
Nonea
0.7
1.7
2.1
3.8
2.9
34.7
47.4
94.5
1 : 2.3 : 49.6
1 : 2.2 : 27.9
—
Not detected
a
Pyrolysis of CA in the absence of phosphoric acid solution.
Fig. 2 GCMS spectra of hydrocarbon oils from PHB and propylene reaction. (A) Spectrum of hydrocarbon oil from PHB reaction in 85% H3PO4 at
230 ꢀC for 3 hours; (B) spectrum of hydrocarbon oil from propylene reaction in 100% H3PO4 at 230 ꢀC for 3 hours.
alkenes formed in low acid solution (85% H3PO4, in Table 3) to formation of major aromatic compounds in the hydrocarbon
may be attributed to propylene oligomerization. When oil.
propylene and toluene were the co-reagents under the same
reaction conditions, propylene participated in alkylation with
toluene to form several substituted aromatic compounds
3.5 Ketonization of crotonic acid
(Fig. S1†). Obviously, propylene alone cannot form the major
aromatic compounds (1–4) in the liquid oil (Fig. 1A). In another
word, decarboxylation of CA is not the only reaction route of
PHB deoxygenation. There must be other main routes that lead
As shown in Table 3, ketones were formed as the major O-
containing compounds from CA under the reaction condi-
tions. Based on our earlier observation that water, in addition to
CO2, was also formed from CA deoxygenation,5 ketonization of
CA is a possible reaction, in which two molecules of carboxylic
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 30005–30013 | 30009

View Article Online
RSC Advances
Paper
Fig. 3 GCMS spectra of two initial intermediates and their further reaction products. (A) Spectrum of the reaction products in an initial stage of
PHB conversion (85% H3PO4, 230 ꢀC, 1 hour); (A1) spectrum of 2,3-dimethyl-2-cyclopenten-1-one (5) and its chemical structure; (A2) spectrum
of 3-methyl-2-cyclohexen-1-one (6) and its chemical structure; (B) spectrum of oil products from compound 5 reaction in 100% H3PO4 at 230
ꢀC for 3 hours; (C) spectrum of oil products from compound 5 reaction with propylene in 100% H3PO4 at 230 ꢀC for 3 hours; (C0) spectrum of oil
products from compound 5 reaction with 2-propanol in 100% H3PO4 at 230 ꢀC for 3 hours; (D) spectrum of oil products from compound 6
reaction in 100% H3PO4 at 230 ꢀC for 3 hours; (E) spectrum of oil products from compound 6 reaction with propylene in 100% H3PO4 at 230 ꢀC
for 3 hours.
acids are converted into one ketone, one CO2, and one water.11 acids does occur in the reaction solution. According to the
In an controlled reaction with butyric acid as the reagent, the unsaturation degrees of reagents and products, two CAs with
saturated C4 organic acid remained largely intact under the total four unsaturation degrees may be ketonized into one
typical reaction conditions, except formation of a very small C7H10O compound (three unsaturation degrees), one CO2 (one
amount of 4-heptanone detected with GCMS (Fig. S2†). The unsaturation degree) and one H2O (zero unsaturation degree).
mass spectrum of 4-heptanone is compared with a reference We checked the products distribution in the early stage of CA and
spectrum in Fig. S3.† The butyric acid reaction reveals two PHB reactions in H3PO4 solution, and nd two unsaturated
important facts: (1) that the double bond of CA (an unsaturated seven-carbon compounds, 2,3-dimethyl-2-cyclopenten-1-one
C4 acid) plays a determinant role in its conversion under the (C7H10O, compound 5 thereaer) and 3-methyl-2-cyclohexenone
typical reaction conditions, and (2) that ketonization of organic (C7H10O, compound 6 thereaer). They were detected with
30010 | RSC Adv., 2015, 5, 30005–30013
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Scheme 3 The proposed formation and conversion of compounds 5 and 6.
Scheme 4 A proposed alkylation mechanism for the formation of compound 1.
GCMS and conrmed with pure chemicals obtained from Sigma- compound 5 very well in the H3PO4 solution, and could be
Aldrich. Their chemical structures are shown in Fig. 3A and mass quickly converted to propylene in acid catalytic conditions.12
spectrum in Fig. S3.† Detected only in the early time of CA and Fig. 3C0 is the GCMS spectrum of the products formed in the
PHB reactions, compounds 5 and 6 were active intermediates and reaction of 2-propane with compound 5 under the typical
further converted to aromatic compounds (benzenes and naph- reaction conditions. Very interestingly, naphthalenes were
thalenes) which is described in the next subsection. Although the signicantly reduced while compounds 1–3 were increased.
detailed mechanism of compounds 5 and 6 from CA is not clear Importantly, the aromatics distribution in Fig. 3C0 is quite
yet, the presence of these two compounds conrms ketonization similar to those of aromatics formed from PHB and CA (Fig. 1).
of CA (Scheme 3). Importantly, the ketonization reaction provides We also noticed that the peaks of alkenes in Fig. 3C0 were
an alternative route to CO2 formation with no propylene being much higher than those in Fig. 1, which was attributed to the
formed. It also explains the high molar ratio of CO2 to propylene excessive 2-popanol used in the reaction. These results indi-
mentioned above.
cate that both compound 5 and propylene were the key inter-
mediates for the formation of main alkyl benzenes
(compounds 1–3). Compounds 2 and 3 might be formed via an
acid-catalyzed Diels–Alder reaction13 from propylene and
3.6 Formation of main aromatic compounds
Compound 5, one of the detected unsaturated ketones, was found compound 5, followed by acid catalytic dehydration. Similar to
a key intermediate in the formation of main aromatics from CA the alkylation of toluene with propylene (Fig. S1†) in the
and PHB. When compound 5 was used as an initial reagent under typical reaction conditions, compound 2 could react with
the typical reaction conditions, the products were much depen- propylene through a 2-propyl carbocation mechanism to form
dent on propylene. In the absence of propylene, naphthalenes compound 1, the major aromatic component of liquid oil
(compounds a–d in Fig. 3B) were the main products rather than (Scheme 4).
the main aromatic compounds 1–4 from CA or PHB. In the
Compound 6, another unsaturated ketone, could also be
presence of propylene, however, the main compounds 1–3 were converted into aromatics in the typical reaction conditions with
formed as shown in Fig. 3C. Interestingly, some naphthalenes or without propylene (Fig. 3D and E). In the presence of
(compounds a–d in Fig. 3B) were also detected in the reaction of propylene, compound 6 was converted to several substituted
compound 5 with propylene. This might be attributed to the fact benzenes (compounds h–j in Fig. 3E), which was also detected
that compound 5 was dissolved in the H3PO4 solution while in the nal hydrocarbon oil from CA and PHB.5 However, none
propylene was in the gas phase. The separation of two reagents of the main compounds 1–4 were formed from compound 6
favored the formation of naphthalenes. To check this hypothesis, even in the presence of propylene. This fact indicates that
we replaced propylene with 2-propanol in the reaction with compound 6 is an intermediate, but not the key intermediate of
compound 5. 2-Propanol was dissolved and mixed with the main aromatics (compound 1–4) from PHB.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 30005–30013 | 30011

View Article Online
RSC Advances
Paper
Scheme 5 A proposed acylation reaction of compound 3 with CA for the formation of 1-indanone,3,3,4,5,7-pentamethyl-(compound 4).
Based on the information above, it can be concluded that possible that compound 3 was acylated with CA to form
both decarboxylation and ketonization of CA occurred under compound 4 as suggested in Scheme 5.
the typical reaction conditions, the former provided propylene
and the latter provided compound 5. The decarboxylation and
ketonization are competitive for CA and the competition is
affected by the concentration of H3PO4 solution. As shown in
4 Summary
Based on the results above, the reaction routes from PHB to
hydrocarbons are summarized in Scheme 6. The main
sequential reactions can be divided into three stages: (1)
degradation of PHB to monomeric compounds, (2) deoxygen-
ation of crotonic acid, and (3) combination of the deoxygenated
small molecules into nal hydrocarbons.
In the rst stage, H3PO4 is an effective catalyst in PHB
hydrolysis and degradation (steps 1–3) and CA is the primary
intermediate. Deoxygenation mainly occurred in the second
stage, including decarboxylation and ketonization of CA to
propylene and unsaturated ketones such as compound 5 (steps
Table 3, the acyclic alkenes from propylene oligomerization are
the major products in the 85% H3PO4 solution, while the
aromatics from compounds 5 and 6 via ketonization become
the major ones in the 100% solution. This result indicates that
high concentration of H3PO4 is benecial to the ketonization
while low concentration of H3PO4 to decarboxylation. The
composition of hydrocarbon oil can therefore be controlled by
simply controlling the acid concentration.
3.7 Oxygen-containing aromatics
As shown in Fig. 1, compound 4 is an oxygen-containing 4 and 5). The reactions are competitive for CA and are affected
aromatic compound in the hydrocarbon oil derived from by H3PO4 concentration. In the third stage, multiple reactions
PHB. It was not found in the controlled reactions of compound occur including oligomerization of propylene (step 6), aromatic
5 or 6 with propylene (Fig. 3). Mostly likely, CA was involved in cyclization of propylene with compound 5 to form compounds 2
the formation of compound 4 via acylation of the aromatic and 3 (steps 7and 8), alkylation of propylene with compound 2
compounds in the reaction conditions.14 In a controlled reac- to form compound 1 (step 9), and acylation of compound 3 with
tion of toluene and CA under the typical conditions, toluene was CA to form compound 4 (step 10). In addition, deoxygenation of
acylated with CA to form H-inden-1-one,2,3-dihydro-3-3- the O-containing ketones (compounds 5 and 6) leads to naph-
dimethyl-as a main product (Fig. S4†). It is therefore highly thalenes (step 11). In this overall reaction pathway, CA,
Scheme 6 Proposed reaction routes for the main compounds of hydrocarbon oil derived from PHB.
30012 | RSC Adv., 2015, 5, 30005–30013
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
propylene and compound 5 are the key intermediates which
determine the nal composition of hydrocarbon oil.
6 C. Chen, B. Fei, S. Peng, Y. Zhuang, L. Dong and Z. Feng, J.
Appl. Polym. Sci., 2002, 84, 89.
7 (a) H. Nishida, H. Ariffin, Y. Shirai and M. A. Hassan, in
Biopolymers, ed. M. Elnashar, InTech, 2010, pp. 369–386;
(b) H. Ariffin, H. Nishida, Y. Shirai and M. A. Hassan,
Polym. Degrad. Stab., 2008, 93, 1433; (c) A. Gonzalez,
L. Irusta, M. J. Fernandez-Berridi, M. Iriarte and J. J. Iruin,
Polym. Degrad. Stab., 2005, 87, 347; (d) G. E. Yu and
R. H. Marchessault, Polymer, 2000, 41, 1087.
8 J. Yu, D. Plackett and L. X. L. Chen, Polym. Degrad. Stab.,
2005, 89, 289.
9 D. B. Bigley and M. J. Clarke, J. Chem. Soc., Perkin Trans.,
1982, 2, 1.
10 (a) R. T. Arnold, O. C. Elmer and R. M. Dodson, J. Am. Chem.
Soc., 1950, 72, 4359; (b) D. B. Bigley and R. W. May, J. Chem.
Soc. B, 1967, 557; (c) A. Al-Borno and D. B. Bigley, J. Chem.
Soc., Perkin Trans., 1982, 2, 15; (d) D. B. Bigley,
R. H. Weatherhead and R. W. May, J. Chem. Soc., Perkin
Trans., 1977, 2, 745.
Acknowledgements
The work is supported by Hawaii Natural Energy Institute and
the Office of Naval Research.
Notes and references
1 (a) B. K. Bailey, in handbook on bioethanol: production and
utilization, ed. C. E. Wyman, Taylor & Francis, 1996, pp.
37–60; (b) A. C. Hansen, Q. Zhang and P. W. Lyne,
Bioresour. Technol., 2005, 96, 277; (c) A. K. Agarwal, Prog.
Energy Combust. Sci., 2007, 33, 233.
2 (a) P. T. M. Do, S. Crossley, M. Santikunaporn and
D. E. Resasco, Catalysis, 2007, 20, 33; (b) G. Marsh, N. Hill,
J. Sully, in consultation on the need to reduce the sulphur
content of petrol and diesel fuels below 50 ppm: A Policy
Maker's
Summary.
European
Commission,
DG
11 (a) T. N. Pham, T. Sooknoi, S. P. Crossley and D. E. Resasco,
ACS Catal., 2013, 3, 2456; (b) R. Martinez, M. C. Huff and
M. A. Barteau, J. Catal., 2004, 222, 404; (c) S. Rajadurai,
Catal. Rev.: Sci. Eng., 1994, 36, 385; (d) J. A. Martens,
M. Wydoodt, P. Espeel and P. A. Jcobs, Stud. Surf. Sci.
Catal., 1993, 78, 527.
12 (a) M. J. Antal Jr, M. Carlsson, X. Xu and D. M. Anderson, Ind.
Eng. Chem. Res., 1998, 37, 3820; (b) J. Bedia, J. M. Rosas,
J. Marquez, J. Rodriguez-Mirasol and T. Cordero, Carbon,
2009, 47, 286; (c) J. Bedia, R. Ruiz-Rosas, J. Rodriguez-
Mirasol and T. Cordero, J. Catal., 2010, 271, 33.
13 (a) H. B. Kagan and O. Riant, Chem. Rev., 1992, 92, 1007; (b)
E. J. Corey, Angew. Chem., Int. Ed., 2002, 41, 1650.
14 (a) C. De Castro, J. Primo and A. Corma, J. Mol. Catal. A:
Chem., 1998, 134, 215; (b) C. Castro, A. Corma and
J. Primo, J. Mol. Catal. A: Chem., 2002, 177, 273.
Environment, AEAT/ENV/R/0372, 2000; (c) L. Gibbs,
B. Anderson, K. Barnes, et al., Motor gasolines technical
review (FTR-1), Chevron Company, 2009; (d) A. A. Gureev,
T. N. Mitusova, V. V. Sokolov, V. E. Emel'yanov and
N. P. Spirkina, Chem. Technol. Fuels Oils, 1993, 29, 525.
3 (a) D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem.,
2010, 12, 1493; (b) C. H. Zhou, X. Xia, C. X. Lin, D. S. Tong
and J. Beltramini, Chem. Soc. Rev., 2011, 40, 5588.
4 (a) J. Yu, A. Dow and S. Pingali, Int. J. Hydrogen Energy, 2013,
38, 8683; (b) J. M. B. T. Cavalheiro, M. De Almeida,
C. Grandls and M. M. R. Da Fonseca, Process Biochem.,
2009, 44, 509; (c) R. A. J. Verlinden, D. J. Hill,
M. A. Kenward, C. D. Williams and I. Radecka, J. Appl.
Microbiol., 2007, 102, 1437; (d) N. Tanadchangsaeng and
J. Yu, Biotechnol. Bioeng., 2012, 109, 2808.
5 S. Kang and J. Yu, RSC Adv., 2014, 4, 14320.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 30005–30013 | 30013