A facile catalytic synthesis of trimethylene carbonate from trimethylene oxide and carbon dioxide

Article abstract of DOI:10.1039/c0gc00136h

The coupling of oxetane (trimethylene oxide) and carbon dioxide catalyzed by VO(acac)₂ in the presence of an onium salt was studied. The process was found to be highly selective and quantitative for the production of the six-membered cyclic carbonate, trimethylene carbonate, under very mild reaction conditions of 60°C and 1.7 MPa. Other derivatives of trimethylene oxide were shown to similarly selectively afford the corresponding cyclic carbonates upon reaction with CO₂.

Full text of DOI:10.1039/c0gc00136h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/greenchem | Green Chemistry
A facile catalytic synthesis of trimethylene carbonate from trimethylene
oxide and carbon dioxide
Donald J. Darensbourg,* Adolfo Horn Jr† and Adriana I. Moncada
Received 21st May 2010, Accepted 29th June 2010
First published as an Advance Article on the web 19th July 2010
DOI: 10.1039/c0gc00136h
The coupling of oxetane (trimethylene oxide) and carbon
dioxide catalyzed by VO(acac) in the presence of an onium
the six-membered cyclic carbonate is thermodynamically less
stable than its polymer, and thus retains CO
the ring-opening polymerization process (eqn (4)).
2
when it undergoes
2
6
salt was studied. The process was found to be highly
selective and quantitative for the production of the six-
membered cyclic carbonate, trimethylene carbonate, under
The proposed pathway for cyclic carbonate production from
cyclic ethers and carbon dioxide involves a backbiting reaction
◦
from the initial ring-opening process with subsequent CO2
very mild reaction conditions of 60 C and 1.7 MPa. Other
insertion or later on from the growing polymer chain. This is
derivatives of trimethylene oxide were shown to similarly
selectively afford the corresponding cyclic carbonates upon
illustrated below for the coupling of oxetane and CO utilizing
2
a metal catalyst and anion initiator (eqn (5)) Although epoxides
are easier to ring-open than oxetane because of higher ring-
strain, oxetane is a much better ligand for binding to metals. The
reaction with CO .
2
Carbon dioxide is an abundant, inexpensive, and nontoxic
source of chemical carbon. Hence, its utilization as a feedstock
for providing value-added products has drawn much attention
pK of oxetane is 3.13 vs. 6.94 for propylene oxide. It should be
b
noted here as well that the process depicted in eqn (5) is enhanced
1
during the last two decades. Concomitantly, it is a greenhouse
in the initial steps if X is a good leaving group. Recently,
we have employed this feature of the reaction mechanism for
tuning the reaction pathway for poly(trimethylene carbonate)
2
gas and a direct effect of its utilization by chemical means plays
a role in curtailing CO emissions at power-generation plants
2
or industrial processes. Alternatively, an indirect effect would
result from waste reduction and the health and safety benefits
derived from toxic chemicals substitution. The preparation of
aliphatic polycarbonates and/or cyclic carbonates via the metal-
formation from oxetane and CO to proceed via preformed
7
trimethylene carbonate by way of ring-opening polymerization.
This was accomplished by means of utilizing a less electrophilic
cobalt(II) Schiff base complex which prevents polymer chain
catalyzed coupling of epoxides with CO
2
represents one of the
propagation or using (salen)CrCl in the presence of n-Bu NBr
and lowering the reaction temperature where six-membered
cyclic carbonate production is enhanced over polymer chain
growth. It is worthwhile noting that copolymer formation by
4
most promising methodologies for achieving these goals (eqn
2
(
1) and (2)). Similarly, we have demonstrated that these same
catalyst systems, e.g., (salen)CrCl in the presence of onium salts,
are capable of affording polycarbonates from oxetane and CO
3
2
.
way of CO and oxetane enchainment generally results in a
2
copolymer with some ether linkages, i.e., not a completely
alternating copolymer.
(
1)
2)
(
(
3)
4)
(
A major difference between the coupling of CO and three-
2
and four-membered cyclic ethers is that the resulting cyclic car-
bonates have opposite thermal stabilities relative to their respec-
4
tive copolymers. That is, the five-membered cyclic carbonate
Trimethylene carbonate is traditionally synthesized from 1,3-
propanediol with phosgene or its derivatives (di- and triphos-
gene) which are highly poisonous gases. Alternatively, it can be
produced in a greener more sustainable manner by transesteri-
fication of 1,3-propanediol and dialkylcarbonate. This is partic-
ularly true since both 1,3-propanediol and dialkylcarbonate are
afforded from epoxides is more stable than its ring-opened
polymer, and hence its ring-opening polymerization reaction
5
occurs with significant loss of CO
2
(eqn (3)). On the other hand,
Department of Chemistry, Texas A&M University, 3255 TAMU, College
Station, TX, USA. E-mail: djdarens@mail.chem.tamu.edu; Fax: +1 979
available from renewable resources, namely, glycerol or glucose,
8
45 0158; Tel: +1 979 845 5417
8
and propylene oxide/CO
2
, respectively. On a related note,
†
On sabbatical leave from Laborat o´ rio de Ci eˆ ncias Qu ´ı micas, Uni-
numerous reports have demonstrated binary systems composed
of a simple metal complex and an onium salt, such as ZnCl
versidade Estadual do Norte Fluminense, 28013-602, Campos dos
Goytacazes, RJ, Brazil.
2
1
376 | Green Chem., 2010, 12, 1376–1379
This journal is © The Royal Society of Chemistry 2010

View Article Online
a
or VCl
3
and n-Bu
4
NOAc or n-Bu
4
NI to be effective catalysts
2
Table 1 Results of coupling reaction of oxetane and CO .
for the coupling of epoxides and carbon dioxide to provide
f
Selectivity (%)
TMC
2,9
cyclic carbonates in good yield. Trimethylene carbonate has
been shown to undergo ring-opening polymerization to yield
poly(trimethylene carbonate) with complete retention of its
Reaction Pressure Conversion
b
f
Entry time (h) (MPa)
(%)
Poly(TMC)
c
1
2
3
4
5
6
7
8
9
2(4)
2
4
4
4
4
2
4
8
3.5
3.5
3.5
1.7
0.6
3.5
3.5
3.5
3.5
46 (57)
56
76
80
34
56
59
84
95
97.8 (98.2)
96.4
94.7
91.3
82.4
100.0
100.0
100.0
100.0
2.2 (1.8)
3.6
5.3
8.7
17.6
—
—
—
—
CO
catalysts.
major components of biomaterials for the preparation of various
2
content by a variety of benign metal- as well as organo-
4,10
This is of importance since these polymers are
11
medical devices.
d
e
e
e
(
5)
a
All reactions were carried out with 5.0 mol% VO(acac) /n-Bu NBr
b
2
4
◦
loading at 60 C, except where noted. Values in parenthesis represent
4 h runs. Catalyst loading 2.5 mol%. 50 C. With added 10 mL of
toluene. Determined by H NMR spectroscopy.
c
d
◦
e
Herein, we describe a catalytic system for the efficient syn-
f
1
thesis of trimethylene carbonate (TMC) from oxetane and CO
involving a readily available complex derived from a biocompati-
ble metal, VO(acac) (acac = acetylacetonato or 2,4-pentadiono)
which serves as a Lewis acid for the activation of oxetane,
and n-Bu NBr which provides the nucleophile for ring-opening
oxetane. In a typical experiment a green solution of VO(acac)
2
2
well as the presence of an added solvent. Previous studies
involving a closely related process have shown that n-Bu
4
NBr
7
4
is the best cocatalyst for this reaction. Table 1 lists reaction
conditions, along with the corresponding product yields and
selectivities. As readily seen in Table 1 all reactions are highly
selective for cyclic carbonate (TMC) formation.
2
(
2.5–5.0 mol%) in oxetane (4.0 g or 0.069 mol) containing one
equivalent of n-Bu NBr, and in some instances with added
4
toluene (10 mL), was introduced into a previously dried at 100
◦
As noted in entry 1 for 2 and 4 h runs at 60 C and 3.5 MPa
◦
12
C under vacuum stainless steel Parr autoclave. The reactor
with a catalyst loading of 2.5 mol% the conversion of oxetane
to products was 46 and 57%, respectively, with a selectivity for
TMC of 98%. Alternatively, an increase in catalyst loading to 5.0
mol% for a 2 h run (entry 2) also provided a 56% conversion with
similar selectivity of cyclic carbonate formation. Entries 3, 4, and
◦
was pressurized with carbon dioxide and heated to 60 C for
the described time. Upon cooling the autoclave in an ice bath
and venting the excess CO in a fume hood, the contents of the
2
reactor were extracted with dichloromethane and the conversion
1
to cyclic carbonate was measured by H NMR spectroscopy.
5
demonstrate the effect of varying the CO
at 3.5 and 1.7 MPa yielding similar reactivity. However, upon
decreasing the CO pressure to 0.6 MPa (entry 5) there was a
significant drop in both reactivity and selectivity. The effect of
2
pressure, with runs
The product was isolated as a green-brown solid (6.56 g) which
contained the VO(acac)
2
/n-Bu NBr catalyst and trimethylene
4
2
carbonate. The cyclic carbonate product was purified by elution
from a silica-gel column by diethyl ether, gradually increasing
the solvent’s polarity to 10% dichloromethane to provide a
white solid in a 60% yield. The melting point, determined by
◦
◦
lowering the reaction temperature from 60 C to 50 C while
keeping the other reaction parameters constant is best seen in
comparing entries 3 and 6 where the % conversion decreased
from 76 to 56% concomitantly with a slight increase in selectivity
◦
◦
13
DSC, was found to be 47.1 C (literature value 45–47 C).
Anal. Calcd. for C : C, 47.06; H, 5.92; Found: C, 47.22;
H, 6.10. H NMR in CDCl d (ppm): 4.45(t, 4H, J = 6 Hz,
CH O), 2.14 (quintet, 2H, J = 6 Hz, CH
CDCl d (ppm): 148.4 (C O), 67.8 (CH O), 21.5 (CH
in CH (cm ): 675(s), 756(vs), 779(m), 1064(m), 1110(vs),
140(s), 1186(vs), 1249(s), 1417(s), 1463(m), 1726(s), 1744(m),
H
6
O
3
◦
4
from 95 to 100%. Increasing the reaction temperature to 70 C
1
3
and higher greatly decreases the selectivity for TMC formation.
That is, subsequent ring-opening polymerization of preformed
TMC leads to poly(TMC) formation. For the purposes of
preparing synthetic quantities of trimethylene carbonate, where
maximum conversion is desired in a batch reaction, using a non-
coordinating cosolvent, toluene, and longer reaction time led
to approximately 100% yield and selectivity of the product (see
entries 7–9). The oxetane derivatives, 3-ethyl-3-oxetanemethanol
1
3
2
2
). C NMR in
). IR
3
2
2
-
1
2
Cl
2
1
2
914(vw), 2927(vw), 2960(vw), 2983(vw) and 3018(vw).
Early reports on the formation of trimethylene carbonate
from oxetane and CO
or organotin iodide/n-Bu
harsher reaction conditions.
2
have utilized tetraphenylstibonium iodide
3
PO catalysts which operate under
(
1) and 3,3-dimethyloxetane (2), provided 100% selectivity for
cyclic carbonate production upon coupling with CO under sim-
14,15
Furthermore, these processes
2
were not well-developed as synthetic methodologies for the
preparation of TMC. So far there have been no communications
ilar reaction conditions. Nevertheless, presumably due to steric
inhibition these reactions occur at a much slower rate than those
for catalyzing the selective coupling of oxetane and CO to
2
involving oxetane and CO
2
. More detailed studies of substituted
provide TMC utilizing transition metal coordination complexes,
especially those derived from biocompatible metals such as
vanadium. The process described here is much greener, i.e.,
produces less waste and utilizes less energy, than the synthesis
oxetanes and CO coupling reactions are currently underway.
2
8b
of TMC from 1,3-propanediol and diethylcarbonate.
In order to find the best yield and selectivity for TMC forma-
tion vs. poly(TMC) we have evaluated variation of the reaction
In an effort to examine this CO coupling reaction in more
2
conditions of temperature, catalyst loading, CO
2
pressure, as
quantitative detail, we have monitored the progress of the
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1376–1379 | 1377

View Article Online
1
6
process by in situ infrared spectroscopy. This was accomplished
employing the reaction conditions (entry 9) which provided a
nearly quantitative yield of trimethylene carbonate. Fig. 1 con-
tains the infrared traces for trimethylene carbonate formation
as a function of time, along with the reaction profile for TMC
In an attempt to gain evidence in support of the suggested
reaction pathway outlined in eqn (5), we have examined the
competitive binding ability of oxetane vs. the anion provided by
the cocatalyst n-Bu NX. In order to accomplish this we have
4
employed the anion X = azide because of its strong, isolated
band in the infrared. Upon addition of one equivalent of n-
production as monitored by the infrared absorption in the nCO
region at 1770–1750 cm . Since the nCO bands in TMC and
poly(TMC) are grossly overlapped at 1770 and 1750 cm in
toluene solution, it is important to point out here that the
infrared band monitored in Fig. 1 is exclusively that of TMC
3
-
1
Bu
4
NN
3
to VO(acac)
2
in CH
2
Cl
2
two n
N
infrared absorptions
3
3
-
1
-1
are observed at 2062 and 2005 cm , where the former is due
to VO(acac)
-
3
2
N
3
and the later to free azide ion. Subsequent
addition of 100 equivalents of oxetane resulted in an increase
in the quantity of free azide with a concomitant decrease in
the quantity of vanadium bound azide. That is, azide is a
1
as verified by H NMR spectroscopy. Fig. 2 illustrates the linear
plot obtained from the data in Fig. 1 of ln(A
•
- A
t
), where A is
the absorbance of TMC, vs. time. This provides a pseudo-first
much better ligand for VO(acac) than oxetane. It would be
anticipated that oxetane would be more competitive with the
anion bromide used in our synthetic procedure. Nevertheless as
2
-
4
-1
order rate constant of about 10
s
for a reaction involving a
◦
catalyst loading of 5 mol% VO(acac)
2
at 60 C and 3.5 MPa CO
2
pressure. Further detailed studies of this kinetically well-behaved
reaction will be forthcoming from our laboratories.
we have observed, the use of excessive quantities of n-Bu NBr
greatly inhibits the reaction’s progress. At this point in our
studies we do not have a solid state structure of the adduct
4
formed between VO(acac)
are structures in the literature between VO(acac)
base such as DMAP (4-dimethylaminopyridine) which exhibit
2
and azide or oxetane. However, there
2
and a good
17
a cis-coordination geometry. Endeavours to obtain crystals
of an oxetane derivative of VO(acac) have led to isolation of
X-ray quality crystals of VO(acac) with no oxetane bound to
vanadium or found anywhere in the crystal lattice.
2
2
18,19
Future investigations from our laboratory will be directed
at more comprehensive mechanistic studies of the selective
coupling reaction of various oxetanes and CO
2
to provide six-
membered cyclic carbonates employing a variety of vanadium
catalysts, e.g., in addition to VO(acac) , VO(salen) was found to
2
catalyze this coupling process.
We gratefully acknowledge the financial support from the
National Science Foundation (CHE 05–43133) and the Robert
A. Welch Foundation (A-0923). Professor Adolfo Horn also
gratefully acknowledges funds for a sabbatical leave from
CAPES (Brazil).
Fig. 1 (A) Three-dimensional stack plot of IR spectra collected every
3
2
min during the coupling reaction of oxetane and CO . (B) Reaction
Notes and references
profile indicating trimethylene carbonate formation with time. Reaction
carried out at 60 C in toluene in the presence of 5 mol% of VO(acac)
and 1 equiv. of n-Bu NBr at 3.5 MPa of CO pressure.
4 2
◦
2
1
Carbon Dioxide as Chemical Feedstock, ed. M. Aresta, Wiley-VCH,
Weinheim, 2010.
2
D. J. Darensbourg and M. W. Holtcamp, Coord. Chem. Rev., 1996,
1
2
2
53, 155; G. W. Coates and D. R. Moore, Angew. Chem., Int. Ed.,
004, 43, 6618; M. H. Chisholm and Z. Zhou, J. Mater. Chem.,
004, 14, 3081; H. Sugimoto and S. Inoue, J. Polym. Sci., Part A:
Polym. Chem., 2004, 42, 5561; D. J. Darensbourg, R. M. Mackiewicz,
A. L. Phelps and D. R. Billodeaux, Acc. Chem. Res., 2004, 37,
8
36; B. Ochiai and T. Endo, Prog. Polym. Sci., 2005, 30, 183; D. J.
Darensbourg, Chem. Rev., 2007, 107, 2388.
3
D. J. Darensbourg, P. Ganguly and W. Choi, Inorg. Chem., 2006,
4
5, 3831; D. J. Darensbourg, A. I. Moncada, W. Choi and J. H.
Reibenspies, J. Am. Chem. Soc., 2008, 130, 6523; D. J. Darensbourg
and A. I. Moncada, Inorg. Chem., 2008, 47, 10000.
4
5
J. H. Clements, Ind. Eng. Chem. Res., 2003, 42, 663.
K. Soga, S. Hosoda, Y. Tazuke and S. Ikeda, J. Polym. Sci., Polym.
Lett. Ed., 1976, 14, 161; K. Soga, Y. Tazuke, S. Hosoda and S. Ikeda,
J. Polym. Sci.: Polym. Chem., 1977, 15, 219; R. F. Harris, J. Appl.
Polym. Sci., 1989, 37, 183; R. F. Harris and L. A. McDonald,
J. Appl. Polym. Sci., 1989, 37, 1491; L. Vogdanis and W. Heitz,
Makromol. Chem. Rapid Commun., 1986, 7, 543; R. F. Storey and
D. C. Hoffman, Macromolecules, 1992, 25, 5369; L. Vogdanis, B.
Martens, H. Uchtmann, F. Hensel and W. Heitz, Makromol. Chem.,
1990, 191, 465; J. C. Lee and M. H. Litt, Macromolecules, 2000, 33,
Fig. 2 Plot of ln(A
absorbances for trimethylene carbonate (1765 cm ) at time = infinitive
and time = t, respectively. Slope = 1.0 ¥ 10
•
- A
t
) vs. time, where A
•
and A
t
are the infrared
-1
-4
-1
2
s
with an R value of
0
.9985.
1
378 | Green Chem., 2010, 12, 1376–1379
This journal is © The Royal Society of Chemistry 2010

View Article Online
1
2
618; S. K e´ ki, J. T o¨ r o¨ k, G. D e´ ak and M. Zsuga, Macromolecules,
001, 34, 6850.
11 M. Penco, R. Donetti, R. Mendichi and P. Ferruti, Macromol. Chem.
Phys., 1998, 199, 1737; A. Smith and I. M. Hunneyball, Int. J. Pharm.,
1986, 30, 215; A. P. P eˆ go, B. Siebum, M. J. A. Van Luyn, X. J. Gallego
y Van Seijen, A. A. Poot, D. W. Grijpma and J. Feijen, Tissue Eng.,
2003, 9, 981; J. J. Marler, J. Upton, R. Langer and J. P. Vacanti, Adv.
Drug Delivery Rev., 1998, 33, 165; A. P. P eˆ go, M. J. A. Van Luyn,
L. A. Brouwer, P. B. van Wachem, A. A. Poot, D. W. Grijpma and J.
Feijen, J. Biomed. Mater. Res., 2003, 67A, 1044.
6
7
8
H. R. Kricheldorf and B. Weegen-Schulz, Macromolecules, 1993, 26,
5
4
991; H. R. Kricheldorf and B. Weegan-Schulz, Polymer, 1995, 36,
997.
D. J. Darensbourg and A. I. Moncada, Macromolecules, 2009, 42,
063; D. J. Darensbourg and A. I. Moncada, Macromolecules, 2010,
3, DOI: 10.1021/ma100896x.
4
4
D. J. Darensbourg, A. I. Moncada and S. J. Wilson, in Green Poly-
merization Methods, ed. Robert T. Mathers and Michael A. R. Meier,
Wiley-VCH, Weinheim, 2010, in press; B. Hu, R. Zhuo and C. Fan,
Polym. Adv. Technol., 1998, 9, 145.
2
12 VO(acac) was purchased from Strem Chemicals and used without
further purification.
13 H. R. Kricheldorf and J. Jenssen, J. Macromol. Sci., Part A: Pure
Appl. Chem., 1989, 26, 631.
9
T. Bok, E. K. Noh and B. Y. Lee, Bull. Korean Chem. Soc., 2006, 27,
14 A. Baba, H. Kashiwagi and H. Matsuda, Tetrahedron Lett., 1985, 26,
1323.
1
2
4
171; J. Mel e´ ndez, M. North and R. Pasquale, Eur. J. Inorg. Chem.,
007, 3323; M. North and R. Pascuale, Angew. Chem., Int. Ed., 2009,
8, 2946; S. W. Chen, R. B. Kawthekar and G. J. Kim, Tetrahedron
15 A. Baba, H. Kashiwagi and H. Matsuda, Organometallics, 1987, 6,
137.
Lett., 2007, 48, 297.
0 G. Rokicki, Prog. Polym. Sci., 2000, 25, 259; D. J. Darensbourg,
W. Choi, P. Ganguly and C. P. Richers, Macromolecules, 2006, 39,
16 D. J. Darensbourg, J. L. Rodgers, R. M. Mackiewicz and A. L. Phelps,
Catal. Today, 2004, 98, 485.
1
17 S. P. Anthony, L. Srikanth and T. P. Radhakrishnan, Mol. Cryst. Liq.
Cryst., 2002, 381, 133.
4
2
374; D. J. Darensbourg, W. Choi and C. P. Richers, Macromolecules,
007, 40, 3521; D. J. Darensbourg, W. Choi, O. Karroonnirun and N.
18 Crystals isolated from an oxetane solution of VO(acac)
hexanes were shown to be pure VO(acac) as identified by X-ray
crystallography (cell parameters a = 7.3191(3) A, b = 8.1328(3) A, c
2
layered with
Bhuvanesh, Macromolecules, 2008, 41, 3493; F. Nederberg, B. G. G.
Lohmeijer, F. Leibfarth, R. C. Pratt, J. Choi, A. P. Dove, R. M.
Waymouth and J. L. Hedrick, Biomacromolecules, 2007, 8, 153; N. E.
Kamber, W. Jeong, R. M. Waymouth, R. Pratt, B. G. G. Lohmeijer
and J. L. Hedrick, Chem. Rev., 2007, 107, 5813; J. Mindemark, J.
Hilborn and T. Bowden, Macromolecules, 2007, 40, 3515.
2
˚
˚
˚
◦
◦
◦
= 11.1863(4) A and a = 72.86 , b = 72.26 , g = 66.97 ).
19 R. P. Dodge, D. H. Templeton and A. Zalkin, J. Chem. Phys., 1961,
35, 55; M. Hoshino, A. Sekine, H. Uekusa and Y. Ohashi, Chem.
Lett., 2005, 34, 1228.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1376–1379 | 1379