On the Diels-Alder approach to solely biomass-derived polyethylene terephthalate (PET): Conversion of 2,5-dimethylfuran and acrolein into p-xylene

Article abstract of DOI:10.1002/chem.201101580

Polyethylene terephthalate (PET) is a polymeric material with high global demand. Conventionally, PET is produced from fossil-fuel-based materials. Herein, we explored the feasibility of a sustainable method for PET production by using solely bio-renewable resources. Specifically, 2,5-dimethylfuran (derived from lignocellulosic biomass through 5-(hydroxymethyl)furfural) and acrolein (produced from glycerol, a side product of biodiesel production) were converted into the key intermediate p-xylene (a precursor of terephthalic acid). This synthesis consists of a sequential Diels-Alder reaction, oxidation, dehydration, and decarboxylation. In particular, the pivotal first step, the Diels-Alder reaction, was studied in detail to provide useful kinetic and thermodynamic data. Although it was found that this reaction requires low temperature to proceed efficiently, which presents a limitation on economic feasibility on an industrial scale, the concept was realized and bio-derived p-xylene was obtained in 34 % overall yield over four steps. Making PET as green as grass: Polyethylene terephthalate (PET) can be prepared from solely bio-renewable sources by converting 2,5-dimethylfuran and acrolein into the key intermediate p-xylene (see scheme). This atom-economic route consists of a sequential Diels-Alder reaction, oxidation, dehydrative aromatization and decarboxylation. We examined the feasibility of this process with an emphasis on the Diels-Alder reaction step. Copyright

Full text of DOI:10.1002/chem.201101580

DOI: 10.1002/chem.201101580
On the Diels–Alder Approach to Solely Biomass-Derived Polyethylene
Terephthalate (PET): Conversion of 2,5-Dimethylfuran and Acrolein into
p-Xylene
Mika Shiramizu and F. Dean Toste*[a]
Abstract: Polyethylene terephthalate
(PET) is a polymeric material with
high global demand. Conventionally,
PET is produced from fossil-fuel-based
materials. Herein, we explored the fea-
sibility of a sustainable method for
PET production by using solely bio-re-
newable resources. Specifically, 2,5-di-
methylfuran (derived from lignocellu-
losic biomass through 5-(hydroxyme-
thyl)furfural) and acrolein (produced
from glycerol, a side product of biodie-
sel production) were converted into
the key intermediate p-xylene (a pre-
cursor of terephthalic acid). This syn-
thesis consists of a sequential Diels–
Alder reaction, oxidation, dehydration,
and decarboxylation. In particular, the
pivotal first step, the Diels–Alder reac-
tion, was studied in detail to provide
useful kinetic and thermodynamic data.
Although it was found that this reac-
tion requires low temperature to pro-
ceed efficiently, which presents a limi-
tation on economic feasibility on an in-
dustrial scale, the concept was realized
and bio-derived p-xylene was obtained
in 34% overall yield over four steps.
Keywords: biomass · Diels–Alder
reactions · polyethylene terephtha-
late · sustainable chemistry
Introduction
sumption in the US: 5.01 million metric tons in 2007).[7] Al-
though the direct substitution of FDCA for TA in polymer
synthesis has been known since the 1970s,[8] it has not been
successfully commercialized to date. This can be attributed
not only to the difference in chemical properties between
FDCA and TA, but also to the difficulty of introducing new
chemicals into an existing market.[9]
Based on the aforementioned background, it is desirable
to develop a novel route to convert HMF into an already es-
tablished commodity chemical, such as TA. The envisioned
transformation of HMF into TA requires the addition of a
building block of at least two carbon atoms, which would
ideally also come from renewable sources. The most rele-
vant work in this context is described by two individually
published patents.[10] They both claim methods to convert
2,5-dimethylfuran (DMF), the hydrogenated form of HMF,
and ethylene gas into p-xylene through a one-pot Diels–
Alder reaction and thermal dehydrative aromatization. Al-
though this simple method gave p-xylene in good yields (up
to 92%), the process requires a high pressure of ethylene, as
well as a high reaction temperature. We thus considered
other biomass-derived dienophiles and chose acrolein as the
most attractive target molecule.[11]
In response to the increasingly urgent issue of sustainability,
lignocellulosic biomass has attracted attention as a renewa-
ble and carbon-neutral energy source. However, the selec-
tive acidic hydrolysis of cellulose to glucose, the feedstock
for fermentative bioethanol production, is difficult due to
the recalcitrant crystalline structure of cellulose[1] and the in-
stability of the product sugars.[2] Instead, 5-(hydroxymethyl)-
furfural (HMF), a dehydrated aromatic form of glucose or
fructose, is typically obtained as a major byproduct under
harsh conditions[3] or even as the predominant product in
the presence of chromium catalysts.[4]
The utility of HMF warrants further exploration. Due to
its low energy density and instability, HMF is not a promis-
ing candidate as a liquid fuel, but it is a potential intermedi-
ate for renewable chemicals. One often-proposed approach
is the oxidation of HMF to 2,5-furandicarboxylic acid
(FDCA).[5] FDCA has been suggested as a potential re-
placement for terephthalic acid (TA).[6] TA is an important
raw material for the production of polyethylene terephtha-
late (PET), a widely used plastic with applications ranging
from beverage containers to synthetic fibers (total PET con-
Acrolein can be efficiently prepared from the dehydration
of glycerol, which is currently overproduced as a side-prod-
uct of biodiesel production.[6,12] An efficient method for the
conversion of acrolein into a bulk chemical would enhance
the value of bio-refinery processes while decreasing waste
output. Acrolein also has the advantage of high atom econo-
my, almost comparable to that of ethylene, as it only con-
tains one excess carbon, which can be easily removed as CO
or CO2. It is known that ethylene glycol, the ester condensa-
[a] M. Shiramizu, Prof. F. D. Toste
Department of Chemistry, University of California
Berkeley, Berkeley, CA 94720 (USA)
Fax : (+1)510-643-9480
E-mail: fdtoste@berkeley.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101580.
12452
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12452 – 12457

FULL PAPER
tion partner in the final polyester synthesis step, can be pro-
duced from sugar-based feedstocks[13] or from cellulose by
using a nickel–tungsten carbide catalyst.[14] Therefore, this
approach could provide fully renewable, solely biomass-de-
rived PET.
placement with bio-derived p-xylene would be easily imple-
mented because few changes would be required in the
downstream infrastructure and operational methods. As step
one is precedented,[15] this work focuses on the conversion
of DMF and acrolein to p-xylene (steps 2 and 3).
To realize this concept, our strategy consists of the follow-
ing five steps (Scheme 1): 1) hydrogenation of HMF to give
DMF; 2) Diels–Alder reaction with acrolein to construct
Results and Discussion
7-oxabicycloACHTUNGTRENNUNG[2,2,1]hept-2-ene core structure 1; 3) oxidation
of the aldehyde, aromatization, and decarboxylation to
obtain p-xylene, 4) oxidation of p-xylene to TA; and finally
5) condensation of TA and ethylene glycol to form PET.
Currently, PET is mass-produced by steps four and five by
using p-xylene obtained from crude oil. Therefore, its re-
To the best of our knowledge, there are only two reports on
the Diels–Alder reaction between DMF and acrolein.[16,17]
Burrell et al. used very high pressure (15 kbar) whereas
Laszlo et al. used an FeIII-doped K-10 bentonite clay cata-
lyst. Due to the rather moderate yields (ꢀ40%) in both
methods, we decided to seek a more efficient simple Lewis
acid catalyst capable of mediating this transformation. The
difficulty soon became apparent; our initial screening of
common Lewis acids, such as AlCl3, Et2AlCl, TiCl4, SnCl2,
BF3·Et2O, ZnCl2, and ZnI2 (room temperature, 08C or
À788C, DMF (1–3 equiv)) showed no promise, mainly due
to the instability of acrolein. In most cases, we observed
either no reaction or immediate and complete decomposi-
tion of acrolein. However, when we specifically focused on
group three and four metal species (Sc, Y, Zr, and Hf), in-
spired by a reported HfCl4-catalyzed Diels–Alder reaction
of DMF and benzyl acrylate,[18] we observed a trace amount
1
of 1 in the H NMR spectrum by using Sc
A
À408C, DMF (2 equiv), CDCl3; TfO=trifluoro methanesul-
fonate).[19]
Considering the negative reaction entropy (DS8) of the
Diels–Alder reaction, we thought that decreasing the tem-
perature and increasing the concentration would be benefi-
cial to the reactivity. Indeed, both parameters had dramatic
effects. In general, lowering the temperature gave better
yields and reduced starting-material decomposition. Higher
concentrations were also favorable; however, the neat reac-
tion mixture froze at about À608C (DMF m.p. À628C). We
found that a ꢀ1:1 (v/v) mixture of DMF and chloroform
(m.p. À648C) does not freeze even at À788C due to cryo-
scopy, and thus this mixture was employed as the standard
conditions to investigate the lower temperature reactions.
To further suppress acrolein decomposition and improve the
selectivity for 1, we also decreased the catalyst loading while
prolonging the reaction time. We concluded that 0.1 mol%
was the optimal amount of the catalyst (–608C, CDCl3; see
Table S1 in the Supporting Information). When more than
0.5 mol% of ScACHTNUTRGNEUNG(OTf)3 was employed, the decomposition of
acrolein was significant. When the catalyst loading was de-
creased to 0.05 mol%, the selectivity for 1 was good but the
reaction proceeded much more slowly. The addition of mo-
lecular sieves (MS; 4 ꢁ) or other drying agents was also
found to be essential for the progression of the reaction at a
reasonable rate. This is likely due to the deactivation of Sc-
AHCTUNGRTEG(NNUN OTf)3 by water. By using the optimized conditions, we were
finally able to obtain the desired adduct 1 in good yield as a
mixture of the endo and exo diastereomers (Scheme 2). En-
visioning the formation of the same product from both ste-
Scheme 1. The proposed PET synthesis by using biomass-derived carbon
feedstocks.
Chem. Eur. J. 2011, 17, 12452 – 12457
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12453

F. D. Toste and M. Shiramizu
Scheme 2. Optimized conditions for the Diels–Alder reaction of DMF
and acrolein.
reoisomers after the aromatization step (Scheme 1), we
were not concerned by the low diastereoselectivity.
The significant temperature dependence of the yield ob-
served in the optimization process implied that this Diels–
Alder reaction was under thermodynamic rather than kinet-
ic control. We thought that a more thorough understanding
of the reaction mechanism and thermodynamics could offer
insights that might lead to further improvements in the yield
and efficiency of the reaction. To investigate this point, the
reaction profiles were compared at different temperatures
by using the conditions in Scheme 2 (Figure 1). At higher
temperatures, the reaction proceeded faster but the yield
reached a plateau at a lower level (e.g., 84% in 3 days at
À608C vs. 75% in less than 24 h at À558C). When the equi-
Figure 2. The linear relationship between Keq and 1/T for 1a (top; y=
3225.4xÀ15.33; R2 =0.996) and 1b (bottom; y=3217.5xÀ15.47; R2 =
0.998).
librium constant Keq =[1]/
by using the converged values at each temperature (À608C
to RT) and ln(Keq) was plotted against 1/T (T in K), a clear
ACHUTGTNREN(UNG [DMF]CAHTUNGTNER[NUGN acrolein]) was calculated
ACHTUNGTRENNUNG
linear relationship was discovered (Figure 2). Thus, it was
found that this reaction is indeed under thermodynamic con-
trol, and the standard enthalpy (DH8) and entropy (DS8)
were experimentally determined to be DH8=À6.41 kcal
molÀ1 and DS8=À0.0304 kcalKÀ1 molÀ1 for 1a, and DH8=
À6.39 kcalmolÀ1 and DS8=À0.0307 kcalKÀ1 molÀ1 for 1b.
Since cooling is a cost- and energy-intensive operation, it
is desirable that this Diels–Alder reaction be performed at
ambient temperature. Unfortunately, the thermodynamic
values show that the low temperature is an inherent require-
ment regardless of the catalytic conditions. The only other
method to increase the equilibrium yield is to increase the
concentration of the starting materials. From the DH8 and
DS8 values, we estimated how many equivalents of DMF are
theoretically required to reach a particular yield at a given
temperature. For example, at À558C, 6 equivalents of DMF
would give ꢀ84% of 1a+1b on a reasonable time scale.
However, more than 40 equivalents of DMF would be nec-
essary to obtain a 90% yield (see Table S2 in the Supporting
Information). Due to the smaller Keq at room temperature,
even more DMF is needed to reach similar yields. Because
this is impractical, improving the yield by increasing the
amount of DMF was not pursued further.
Although the low temperature required for the Diels–
Alder reaction step presents a severe economic challenge
for scaling up the reaction, it is still valuable to achieve the
synthesis of p-xylene solely from biomass waste products.
Given the good yield and product selectivity achieved, we
decided to isolate 1 and investigate its downstream conver-
sion into p-xylene. To obtain a good yield of 1, it was essen-
tial to quench the catalyst at low temperature after comple-
tion of the reaction. As expected from the measured equilib-
rium constants, 1 readily undergoes a retro-Diels–Alder re-
action if allowed to warm in the presence of ScACTHNUGTRENUNG(OTf)3. In
Figure 1. The temperature dependence of the DMF–acrolein Diels–Alder
fact, equilibration was extremely rapid above 08C; if the
À608C reaction mixture containing catalyst and an 83%
~
&
^
*
reaction (ꢂ: À788C; : À608C; : À558C; : À408C; : 08C; : 258C
*
(RT)).
12454
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12452 – 12457

Biomass-Derived PET
FULL PAPER
yield of 1 was brought to 08C or RT, the same equilibrium
states as those obtained from the forward reaction (Figure 1,
20% yield at 08C and 8% yield at RT) were reached in less
than 10 min.
Moreover, at room temperature, the retro-Diels–Alder re-
action of 1 was fast even after removal of the catalyst.[20]
This raised a fundamental technical problem because the
common isolation/purification techniques could not be ap-
plied to 1 at room temperature. Although we clearly could
not relieve the inherent thermodynamic driving force, we
anticipated that a reduction in workup temperature would
render the reverse reaction slow enough to allow the preser-
vation of 1. To address this issue and determine which tem-
perature is sufficient for this purpose, we sought more infor-
mation regarding the kinetics of the thermal retro-Diels–
Alder reaction.
Thus, the first-order decay of 1 in the absence of the cata-
1
lyst was monitored by H NMR spectroscopy at room tem-
perature (Figure 3), and the rate constant (k), Gibbs free
energy of activation (DG°) and half-life (T1/2) at this temper-
ature were determined in several different solvents
(Table 1). Because the transition state represents a state be-
tween two molecules and one molecule, DS° is likely to be
positive. Therefore, DG° at lower temperatures should be
larger than DG°RT for this retro-Diels-Alder reaction. If the
DG°RT values in Table 1 are used, the half-lives (T1/2) for 1a
and 1b at 08C, in chloroform are calculated to be 85 and
95 h, respectively, suggesting that 1 is stable towards a retro-
Diels–Alder reaction at ꢁ08C. Indeed, when NMR samples
of 1 were kept at 08C instead of RT, essentially no reaction
was observed over a period of 10 h in CDCl3, CD2Cl2,
CD3CN, or [D8]toluene.
Based on the knowledge garnered from this mechanistic
investigation, the best approach to isolate the Diels–Alder
product appeared to be a three-step sequence consisting of
a Diels–Alder reaction at À60 to À558C, quenching the cat-
alyst at the same low temperature, and derivatization of 1 to
a more stable compound at ꢁ08C. Regarding the choice of
derivatization method, we considered that the oxidation of
aldehyde 1 to more stable carboxylic acid 2 could solve the
problem of the retro-Diels–Alder reaction while also ad-
vancing towards the desired end product, p-xylene
(Scheme 1).[21] In particular, we chose the Pinnick oxidation
by using H2O2 as a mild, green, and economical method[22]
and successfully devised an operationally straightforward,
one-pot Diels–Alder/Pinnick oxidation protocol (Scheme 3).
After confirming ꢀ75% conversion of acrolein into 1 by
1H NMR spectroscopy, the catalyst was quenched by adding
an aqueous NaH2PO4/CH3CN mixture at À558C. The reac-
tion mixture was allowed to warm to 08C, and H2O2 and
NaClO2 were added. The oxidation was complete in 5 h,
giving 2 in good yield (>99% for oxidation step, 77% from
DMF and acrolein). Notably, the excess reagents and by-
products can all be removed by simple aqueous workup and
evaporation under reduced pressure, leaving 2 in nearly
pure form without further purification. Carboxylic acid 2
was much less prone to the retro-Diels–Alder reaction than
Figure 3. The kinetics of the thermal retro-Diels–Alder reaction at room
&
temperature for 1a (top) and 1b (bottom). Top: : CDCl3, y=0.2795x+
1.0905, R2 =0.9994;
:
CD2Cl2, y=0.1919x+1.1088, R2 =0.9976;
:
~
^
CD3CN, y=0.1364x+1.0701, R2 =0.9987; ꢂ: [D8]toluene, y=0.0694x+
2
1.1676, R2 =0.9792. Bottom: : CDCl3, y=0.3103x+1.2256, R =0.9989;
&
CD2Cl2, y=0.2129x+1.2329, R2 =0.992;
: CD3CN, y=0.1946x+
~
:
^
1.1948, R2 =0.9995; ꢂ: [D8]toluene, y=0.0801x+1.2764, R2 =0.9827.
Table 1. The thermal retro-Diels–Alder reaction at room temperature.
1a
1b
Solvent
kRT
[ꢂ10À5
DG°RT
T1/2,RT kRT
DG°RT
T1/2,RT
[h]
G
]
[kcalmolÀ1
]
[h]
A
]
[kcalmolÀ1
]
CDCl3
CD2Cl2
CD3CN
7.76
5.33
3.79
23.04
23.27
23.47
23.87
2.48
3.61
5.08
9.99
8.62
5.91
5.41
2.22
22.98
23.21
23.26
23.78
2.23
3.26
3.56
8.65
[D8]toluene 1.93
1 and could be stored for a few hours at room temperature
or for several months at À208C.[23]
With isolated 2 in hand, we studied its conversion into
p-xylene. The reported aromatization reactions of the simi-
Chem. Eur. J. 2011, 17, 12452 – 12457
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12455

F. D. Toste and M. Shiramizu
Experimental Section
Diels–Alder reaction of DMF and
acrolein: A vial (20 mL) fitted with a
PTFE septum screw cap was charged
with a magnetic stirring bar, ScACHTUNGTRENNUNG(OTf)3
(3.0 mg, 0.006 mmol), and activated
molecular sieves (4 ꢁ powder, 300 mg)
and then purged with nitrogen by
using a vacuum/N2 cycle (ꢂ4). Tetrae-
Scheme 3. The one-pot Diels–Alder/Pinnick sequence.
thylsilane
(115 mL,
90.9 mg,
0.63 mmol) and CDCl3 (1.8 mL) were
added and the mixture was cooled to À558C. DMF (1.93 mL, 19 mmol)
and acrolein (400 mL, 5.9 mmol) were subsequently added over a few mi-
nutes. The reaction progress was monitored by 1H NMR spectroscopy as
follows: A small aliquot (ꢀ10 mL) of the reaction mixture was removed
by use of a syringe and diluted in pre-cooled (À558C) CDCl3. The
sample was immediately filtered through a Pasteur pipette filled with
Na2SO4 into a NMR tube and the 1H NMR spectrum was collected at
RT. The sample was kept frozen at À788C (acetone/dry ice bath) if im-
mediate 1H NMR analysis was not possible. The yield and conversion
were determined by using the peak of tetraethylsilane as an internal stan-
dard. After stirring at À558C for 24 h, a 75% 1H NMR yield of 1 (1a=
2.4 mmol, 1b=2.0 mmol, 1a/1b=1.2) was obtained (remaining DMF=
13 mmol, acrolein=1.4 mmol). The signals used for the yield/conversion
calculation: tetraethylsilane d=0.51 ppm (q, J=8.0 Hz, 8H); DMF d=
5.86 ppm (s, 2H); acrolein d=6.52 ppm (dd, J=9.5, 1.0 Hz, 1H); 1a d=
6.12 ppm (d, J=5.6 Hz, 2H); 1b d=6.16 ppm (d, J=5.6 Hz, 2H). The
relative stereochemistry of 1a and 1b was determined by using d=2.84–
2.88 (m, 1H) and 2.39–2.34 ppm (m, 1H). In analogy to the exo hydrogen
in 2a, which has a more downfield resonance than the endo hydrogen in
2b (see the Supporting Information), the former was assigned to be 1a
(endo, major) and the latter was assigned to be 1b (exo, minor).
lar 7-oxabicycloACHTUNGTRENNUNG[2,2,1]hept-2-ene structure mainly use
strongly acidic conditions for dehydration.[24] When we treat-
ed the crude 2a/2b mixture with concentrated H2SO4, com-
pound 3 was obtained in 48% yield (Scheme 4). Base-cata-
lyzed dehydration by using potassium hexamethyldisilazide
(KHMDS) or other reagents[25] was unsuccessful, presuma-
bly due to the presence of an unprotected COOH group.
Direct pyrolysis[26] was also inapplicable because 2 under-
goes
a retro-Diels–Alder reaction and decomposes if
heated. We are currently seeking more efficient aromatiza-
tion methods in our laboratory. The Cu2O-catalyzed proto-
decarboxylation of aromatic carboxylic acids was recently
reported by Gooben et al.[27] When 3 was subjected to this
system, p-xylene was obtained in 91% yield without optimi-
zation. Combined with the two steps described in Scheme 3,
this route from DMF and acrolein to p-xylene gives a 34%
overall yield over four steps, thereby realizing the concept
of biomass-derived PET synthesis (Scheme 1).
One-pot Pinnick oxidation of
1 to
form 2: Following the procedure de-
scribed above, a Diels–Alder reaction
of DMF and acrolein was conducted in
a septum-capped flask (100 mL) on a
14.3 mmol scale (À558C, 25 h). A pre-
cooled (08C) mixture of CH3CN
(14 mL),
NaH2PO4·H2O
(1.11 g,
8.0 mmol), and H2O (6 mL) was slowly
added to this mixture at À558C. The
Scheme 4. Conversion of 2 into p-xylene.
mixture was kept at À558C for 10 min
and then aqueous H2O2 (34%, 7 mL,
70 mmol, 08C) was added. The mix-
Conclusion
ture was allowed to warm to 08C and NaClO2 (80%, 1.77 g, 20 mmol) in
H2O (20 mL, 08C) was added in small portions over 3 h. The reaction
progress was monitored by H NMR spectroscopy (by using the sampling
We have developed a route to convert DMF and acrolein
into p-xylene for bio-renewable PET production, with the
aim to expand the range of major commodity chemicals that
can be synthesized from biomass. Both raw materials are de-
rived from waste products (HMF and glycerol) of biofuel
production. Our method consisting of a Diels–Alder reac-
tion, oxidation, dehydrative aromatization, and decarboxyla-
tion is designed to maximize atom economy and avoid toxic
byproducts. Unfortunately, the process revealed in this study
would certainly not be immediately practical due to the low-
temperature conditions required in the Diels–Alder reaction
step and the moderate yield of the aromatization step.
Nonetheless, this solely bio-renewable PET synthesis serves
as a valuable demonstration of sustainable chemistry in the
field of biomass utilization.
1
procedure described above) and TLC. Upon complete consumption of 1
in an additional 5 h, the molecular sieves were filtered off and the organ-
ic solvents were removed in vacuo without heating. The mixture was ex-
tracted with CH2Cl2 (3ꢂ40 mL), dried over MgSO4, filtered, and concen-
trated in vacuo to give crude 1. The aqueous layer was then acidified to
pH 3 with aqueous HCl (0.1n), extracted with CH2Cl2 (3ꢂ40 mL), dried
over MgSO4, filtered, and concentrated in vacuo to give crude 2. Both
crude products gave a sufficiently pure 2a/2b mixture. The diastereose-
lectivity (endo/exo ratio) was determined by 1H NMR analysis: d=
3.00 ppm (dd, J=9.0, 3.5 Hz, 1H; 2a, endo major), d=2.62 ppm (dd, J=
8.0, 3.5 Hz; 1H, 2b, exo minor). Crude 1: thick pale-yellow oil, 1.62 g
(9.7 mmol, 2a/2b=1.2); crude 2: thick colorless oil, 0.238 g (1.4 mmol,
2a/2b=1.5), Overall: 1.86 g (11.1 mmol, 2a/2b=1.2), 77% yield over the
two steps from DMF and acrolein. LCMS found two peaks corresponding
to diastereomers 2a and 2b: both m/z calcd for [C9H11O3]À: 167.1; found:
167.1. Retention times matched those obtained from a Diels–Alder reac-
tion of DMF and 2,2,2-trifluoroethyl acrylate followed by hydrolysis (see
the Supporting Information).
12456
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12452 – 12457

Biomass-Derived PET
FULL PAPER
[14] N. Ji, T. Zhang, M. Zheng, A. Wang, H. Wang, X. Wang, J. G. Chen,
Angew. Chem. 2008, 120, 8638–8641; Angew. Chem. Int. Ed. 2008,
47, 8510–8513.
[15] a) Y. Romꢄn-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic, Nature
2007, 447, 982–985; b) M. Chidambaram, A. T. Bell, Green Chem.
2010, 12, 1253–1262; c) G. C. A. Luijkx, N. P. M. Huck, F. van Rant-
wijk, L. Maat, H. van Bekkum, Heterocycles 2009, 77, 1037–1044.
[16] a) S. W. G. Dauben, H. O. Krabbenhoft J. Am. Chem. Soc. 1976, 98,
1992–1993; b) P. Laszlo, J. Lucchetti, Tetrahedron Lett. 1984, 25,
4387–4388.
[17] For aromatization of DMF through a Diels–Alder reaction with
other dienophiles, see: a) A. de La Hoz, A. Diaz-Ortiz, J. M. Fraile,
M. V. Gomez, J. A. Mayoral, A. Moreno, A. Saiz, E. Vazquez, Syn-
lett 2001, 0753–0756; b) J. Fraile, J. Garca, M. A. Gmez, A. de La
Hoz, J. Mayoral, A. Moreno, P. Prieto, L. Salvatella, E. Vzquez, Eur.
J. Org. Chem. 2001, 2891–2899.
[18] Y. Hayashi, M. Nakamura, S. Nakao, T. Inoue, M. Shoji, Angew.
Chem. 2002, 114, 4253–4256; Angew. Chem. Int. Ed. 2002, 41, 4079–
4082.
Acknowledgements
This work was supported by British Petroleum through the Energy Bio-
sciences Institute. M.S. is grateful for an Ito Foundation for International
Education Exchange Graduate Fellowship and The Dow Sustainability
Innovation Student Challenge Award. The authors also thank Dr. J. B.
Binder for helpful discussions.
[1] R. Rinaldi, F. Schuth, ChemSusChem 2009, 2, 1096–1107.
[2] Q. Xiang, Y. Y. Lee, R. W. Torget, Appl. Biochem. Biotechnol. 2004,
115, 1127–1138.
[3] a) G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–
4098; b) J. N. Chheda, Y. Roman-Leshkov, J. A. Dumesic, Green
Chem. 2007, 9, 342–350.
[4] a) J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–
1985; b) C. Li, Z. Zhang, Z. K. Zhao, Tetrahedron Lett. 2009, 50,
5403–5405.
[5] a) P. Verdeguer, N. Merat, A. Gaset, J. Molec. Catal. 1993, 85, 327–
344; b) Y. Y. Gorbanev, S. K. Klitgaard, J. M. Woodley, C. H. Chris-
tensen, A. Riisager, ChemSusChem 2009, 2, 672–675; c) O. Casano-
va, S. Iborra, A. Corma, ChemSusChem 2009, 2, 1138–1144.
[6] a) Top Value Added Chemicals From Biomass, US Department of
Energy, http://www.eere.energy.gov/biomass/pdfs/35523.pdf, 2004;
b) J. J. Bozell, G. R. Petersen, Green Chem. 2010, 12, 539–554.
[7] B. Kuczenski, R. Geyer, Resour. Conserv. Recycl. 2010, 54, 1161–
1169.
[19] HfCl4 and ScCl3 were less active in this reaction. Hf
ACHTUNGRTEN(NUNG OTf)4 was later
found to be as effective as Sc(OTf)3, but optimization and other
AHCTUNGTRENNUNG
studies were not conducted with this species.
[20] At room temperature, the equilibrium yield (7–8%) was also
reached in the forward direction with no catalyst; see Table S1,
entry 7 in the Supporting Information.
[21] Hydrogenation of the double bond on Pd/C and reduction of the al-
dehyde to the alcohol by NaBH4 or LiAlH4 were also attempted in
one pot at À558C and the desired compounds were obtained. Rh-
catalyzed decarbonylation and base- or acid-catalyzed aromatization
were not successful on 1.
[22] a) A. Raach, O. Reiser, J. Prakt. Chem. 2000, 342, 605–608; b) E.
Dalcanale, F. Montanari, J. Org. Chem. 1986, 51, 567–569; c) F. E.
Ziegler, Y. Wang, J. Org. Chem. 1998, 63, 7920–7930.
[23] Acrylic acid and decomposition products were observed when left at
RT for a few days.
[24] a) T. Chan, T. C. W. Mak, C. Poon, H. N. C. Wong, J. H. Jia, L. L.
Wang, Tetrahedron 1986, 42, 655–661; b) G. P. Jana, B. K. Ghorai,
Tetrahedron 2007, 63, 12015–12025; c) F. I. Zubkov, E. V. Boltukhi-
na, K. F. Turchin, A. V. Varlamov, Tetrahedron 2004, 60, 8455–8463;
d) A. D. Mance, B. Borovicka, B. Karaman, K. Jakopcic, J. Hetero-
cycl. Chem. 1999, 36, 1337–1341.
[25] a) F. Brion, Tetrahedron Lett. 1982, 23, 5299–5302; b) M. E. Bun-
nage, T. Ganesh, I. B. Masesane, D. Orton, P. G. Steel, Org. Lett.
2003, 5, 239–242.
[8] A. Gandini, Adv. Polym. Sci. 1977, 25, 47–96.
[9] A. Gandini, A. J. D. Silvestre, C. P. Neto, A. F. Sousa, M. Gomes, J.
Polym. Sci. Part A: Polym. Chem. 2009, 47, 295–298.
[10] a) K. Takanishi, S. Sone (Toray Industries, Tokyo, Japan), WO
2009110402A1, 2009, ; b) T. Brandvold (UOP LLC, Illinois, USA),
WO 2010151346A1, 2010.
[11] In the course of our investigation, we also studied maleic anhydride
and 2,2,2-trifluoroethyl acetate as alternative raw materials. p-
Xylene was obtained in both cases in moderate overall yields. Acryl-
ic acid is an oxidized form of acrolein and thus can be obtained
from glycerol. Maleic acid can also be produced from bio-derived
fumaric acid. See the Supporting Information for details.
[12] B. Katryniok, S. Paul, V. Belliere-Baca, P. Rey, F. Dumeignil, Green
Chem. 2010, 12, 2079–2098.
[13] Ethylene glycol is conventionally produced from another petro-
chemical ethylene. However, its replacement with biomass-based
ethylene glycol has already been successfully commercialized in
some products, such as The Coca-Cola Companyꢃs PlantBottle; for
methods of bio-derived ethylene glycol production, see: a) T. P.
Binder, P. D. Bloom, G. B. Poppe (Kirkpatrick&Lockhart Preston
Gates Ellis LLP, Pittsburgh, PA), USA 20080103340Al, 2008; b) S. P.
Chopade, D. J. Miller, J. E. Jackson, T. A. Werpy, J. G. Frye, A. H.
Zacher (Michigan State University, East Lansing, MI; Battelle Me-
morial Institute, Richland, WA), US-A 6291725B1, 2001.
[26] W. A. Yarnall, E. S. Wallis, J. Org. Chem. 1939, 4, 270–283.
[27] L. J. Gooßen, W. R. Thiel, N. Rodriguez, C. Linder, B. Melzer, Adv.
Synth. Catal. 2007, 349, 2241–2246.
Received: May 23, 2011
Revised: July 23, 2011
Published online: September 16, 2011
Chem. Eur. J. 2011, 17, 12452 – 12457
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12457