Highly Efficient Organocatalyzed Conversion of Oxiranes and CO2 into Organic Carbonates

Article abstract of DOI:10.1002/cssc.201500710

A binary catalyst system based on tannic acid/NBu₄X (X=Br, I) is presented as a highly efficient organocatalyst at very low catalyst loading for the coupling of carbon dioxide and functional oxiranes to afford organic carbonates in good yields. The presence of multiple polyphenol fragments within the tannic acid structure is considered to be beneficial for synergistic effects that lead to higher stabilization of the catalyst structure during catalysis. The observed turnover frequencies (TOFs) exceed 200 h^-1 and are among the highest reported to date for organocatalysts in this area of CO₂ conversion. This organocatalyst system presents a useful, readily available, inexpensive, and, above all, reactive alternative for most of the metal-based catalyst systems reported to date.

Full text of DOI:10.1002/cssc.201500710

DOI: 10.1002/cssc.201500710
Full Papers
Highly Efficient Organocatalyzed Conversion of Oxiranes
and CO into Organic Carbonates
Sergio SopeÇa, Giulia Fiorani, Carmen Martín, and Arjan W. Kleij*
2
[a]
[a]
[a]
[a, b]
A binary catalyst system based on tannic acid/NBu X (X=Br, I)
ture during catalysis. The observed turnover frequencies (TOFs)
4
À1
is presented as a highly efficient organocatalyst at very low
catalyst loading for the coupling of carbon dioxide and func-
tional oxiranes to afford organic carbonates in good yields.
The presence of multiple polyphenol fragments within the
tannic acid structure is considered to be beneficial for synergis-
tic effects that lead to higher stabilization of the catalyst struc-
exceed 200 h and are among the highest reported to date
for organocatalysts in this area of CO conversion. This organo-
2
catalyst system presents a useful, readily available, inexpensive,
and, above all, reactive alternative for most of the metal-based
catalyst systems reported to date.
Introduction
Current (catalytic) research centered on the use of CO as an
the generally observed lower reactivity of organocatalysts still
poses a major challenge to their ability to compete with
metal-based catalysts and eventually provide more-sustainable
catalysis solutions.
2
inexpensive and renewable source of carbon focuses on its in-
corporation into other organic scaffolds to enable the partial
[1–9]
replacement of fossil-fuel-based chemistry.
Considerable
progress in this area has led to a wide variety of organic struc-
In the course of our studies devoted to the development of
a new and more efficient organocatalytic methodology for or-
ganic carbonate formation, we and others have shown that
the activation of oxiranes through hydrogen bonding with pol-
[10–14]
tures from CO as a molecular synthon.
Among these or-
have a prominent
2
[7,8,15–17]
ganic products, organic carbonates
position as they are useful in a number of applications, includ-
ing their use as fuel additives, aprotic solvents, and monomer
yphenolic compounds, fluorinated alcohols, and silanediols is
[18]
[45–48]
intermediates for polymeric structures. In the last decade,
highly efficient catalytic methods for their preparation have
emerged, and those incorporating Lewis acidic metal ions are
very attractive.
In particular, pyrogallol (PG, i.e., 1,2,3-trihy-
droxybenzene) is an example of a highly efficient organocata-
lyst system that can mediate the coupling between terminal
[15,19–22]
among the most-active reported to date.
Nevertheless,
epoxides and CO under extremely mild reaction conditions
2
the use of metal-based catalysts in industrial settings is not
always desired, and the presence of trace amounts of (toxic)
metals in the final products is subject to an increasingly lower
limit accepted by end-users in commercial settings.
(25–458C, 0.2–1 MPa CO pressure, 2 mol% catalyst). Of partic-
2
ular importance is the cooperative nature of the adjacent
phenol groups that allows the formation of an extended hy-
drogen-bond network upon activation of the oxirane, which
lowers the activation barrier for its ring-opening by an external
Therefore, from this viewpoint, one would ideally like to use
organocatalysis for the formation of organic carbonates, and
progress in this respect is characterized by the use of various
[
31,42]
nucleophile (Scheme 1b).
We thought that an even higher
local concentration of phenolic sites would be beneficial for
catalytic turnover and, therefore, considered tannic acid (TA,
Scheme 1a) as a catalyst additive, as it is a naturally occurring
plant polyphenol that is commercially available and inexpen-
sive. The presence of multiple phenol fragments in TA should
facilitate the highly efficient activation of oxiranes through sim-
ilar synergistic effects to those noted for PG.
[
23–25]
types of catalysts based on ionic liquids;
(poly)alcohols
[
26–31]
and Brønsted acids;
um, or imidazolium salts and their derivatives;
(supported) phosphonium, ammoni-
[32–38]
and
However, organocatalysts are usually much less ef-
[
39–44]
others.
fective in the activation of organic substrates and require
longer reaction times, much higher reaction temperatures,
[
4]
and/or (much) higher loadings for effective turnover. Thus,
Herein, we report on the use of binary catalyst systems de-
rived from TA and suitable nucleophiles for the coupling of
CO2 and various epoxides; these new catalyst systems are
[
a] S. SopeÇa, Dr. G. Fiorani, Dr. C. Martín, Prof. Dr. A. W. Kleij
Institute of Chemical Research of Catalonia (ICIQ)
Av. Pa ¨ı sos Catalans 16, 43007 Tarragona (Spain)
E-mail: akleij@iciq.es
[
44]
among the most active organocatalysts reported to date,
and their appreciable turnover frequencies (TOFs) compete
with those reported for many known metal-based catalysts.
The observation of synergistic effects for improved catalyst sta-
bility and lifetime creates the potential for new organocatalyst
[
b] Prof. Dr. A. W. Kleij
Catalan Institute for Research and Advanced Studies (ICREA)
Pg. Lluis Companys 23, 08010 Barcelona (Spain)
design in this area of CO catalysis.
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201500710.
2
ChemSusChem 2015, 8, 3248 – 3254
3248
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 1. Screening of conditions and (co)catalyst loadings using various
[a]
nucleophiles, TA, and using 1,2-epoxyhexane (1a) as the substrate.
[b]
Entry
Loading [mol%]
NBu
Solvent
T
[8C]
Yield
[%]
TA
4
X
[
c]
1
2
3
4
5
6
7
8
9
0.50
2.0
I (5.0)
I (5.0)
I (5.0)
I (5.0)
I (2.0)
0
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
MEK
ACE
MEK
MEK
MEK
MEK
MEK
45
80
80
80
80
80
80
80
80
80
80
80
80
70
60
80
80
80
80
80
80
80
80
80
16
[
c]
93
>99
47
41
0
>99
>99
>99
24
82
79
47
76
53
>99
>99
70
>99
>99
62
[d]
0.05
0
0
0.05
0.05
0.05
0.05
0.05
0.05
0.03
0.01
0.05
0.05
0.05
0.05
0.05
0.05
I (4.0)
I (3.0)
I (2.0)
I (2.0)
I (1.0)
I (2.0)
I (2.0)
I (5.0)
I (5.0)
I (5.0)
Br (2.0)
Cl (2.0)
I (2.0)
I (2.0)
I (2.0)
I (2.0)
I (2.0)
I (2.0)
[
e]
10
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
3
4
5
6
7
8
9
0
1
2
3
4
[f]
[
[
[
g]
h]
i,j]
0.025
0.15
0.03
0.15
0.30
35
33
55
Scheme 1. (a) Chemical structures of TA and PG. (b) Stabilization of inter-
[h,i]
[h,i]
mediates in organic carbonate synthesis through hydrogen bonding using
PG.
0
[
a] General conditions: 1,2-epoxyhexane (8.3 mmol), pCO2 =1 MPa, cocata-
lyst NBu X (amount indicated), 18 h, 30 mL autoclave as reactor; MEK or
ACE (acetone; 5 mL). [b] Yield determined by H NMR spectroscopy
CDCl ) using mesitylene as an internal standard; selectivity for the cyclic
4
1
Results and Discussion
(
3
carbonate was >99%. [c] Not fully homogeneous. [d] Isolated yield 99%.
e] Conditions: MEK (2.5 mL), t=2 h; TON=472, TOF=236 h ; the reac-
tion in the absence of TA afforded only 8% of 1b. [f] pCO2 =0.6 MPa.
g] Only 2.5 mL solvent. [h] PG as catalyst. [i] Reaction time 6 h. [j] Average
À1
[
First, we combined TA with NBu I to obtain a binary catalyst
4
0
system capable of mediating the coupling of 1,2-epoxyhexane
[
(
1a, a benchmark substrate) and CO . Our previous study
À1
2
TA TOF=195 h .
using PG/NBu I as a binary catalyst system [5 mol% of each
4
with respect to the same substrate, 2-butanone (methyl ethyl
[49]
ketone, MEK) as solvent] gave rise to a yield of 100% (93%
isolated) after 18 h at 458C. Thus, we first attempted to use
similar conditions (Table 1, entry 1) using a slightly lower
amount of hydrogen-bond donor, that is, TA (0.50 mol%).
However, under these conditions, we observed that the TA was
not fully soluble; therefore, the epoxide turnover became
more complicated. We found that an increase of the reaction
temperature to 808C resulted in a homogeneous catalyst solu-
tion if the catalyst loading was kept low enough, that is, at
À1
high (236 h ; Table 1, entry 10; t=2 h); in the absence of TA,
only an 8% yield of 1b was noted after 2 h. The catalyst load-
ing, [TA], could be further reduced to approximately
0.03 mol% (Table 1, entry 12) to yield a substantially higher
conversion of 1a compared to the turnover facilitated by the
cocatalyst NBu I alone (cf., Table 1, entry 5). The reaction tem-
4
perature had a significant effect on the conversion rate (see
Table 1, entries 14 and 15 vs. 3), and 808C seems to be optimal
for this catalytic system.
0
.05 mol% of TA (Table 1, entry 3), and led to the quantitative
formation of the cyclic carbonate 1b with high selectivity
>99%). The use of the nucleophilic reagent alone (see
At a lower initial pressure of 0.6 MPa, the yield of 1b re-
mained quantitative (Table 1, entry 16). Upon changing the
nature of the nucleophile (Table 1, entries 17 and 18), a lower
yield of 1b was apparent if the chloride was used whereas the
bromide-based binary catalyst gave rise to a similar result, that
is, quantitative conversion of epoxide 1a could be attained
(cf., Table 1, entry 9). The use of an alternative solvent (ace-
tone; Table 1, entry 19) also resulted in productive catalysis in
line with our previous results on Zn-catalyzed carbonate for-
(
Table 1, entries 4 and 5) led to a much lower yield of cyclic car-
bonate 1b, which emphasizes the imperative role of the TA to
mediate this conversion. In the presence of TA and absence of
cocatalyst (NBu I), no conversion was noted (Table 1, entry 6).
4
[
50]
We then varied the cocatalyst loading (Table 1, entries 7–11)
and kept [TA] at 0.05 mol%; quantitative conversion of 1a to
1
b could still be achieved at a NBu I loading of 2 mol%. Inter-
4
[
51]
estingly, the initial TOF under these conditions was relatively
mation. A further decrease of the TA loading to 0.025 mol%
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3249
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
in only 2.5 mL of MEK still resulted in quantitative conversion
conversion and activity of the TA-based catalyst (Table 2,
entry 1) and those consisting of 1, 5, or 10 equivalents of the
polyphenols PG, CC, or PGA (Table 2, entries 3–11). According
to the data in Table 2, the TA-based systems show favorable
comparative reactivity behavior with high molecular turnover
numbers (TONs) and TOFs. It should be mentioned that is diffi-
cult to use a correct reference system for TA as PG and CC are
electronically different from the pseudo-PG units within the TA
structure, and PGA probably represents a better electronic
match. Furthermore, TA contains a significant amount of water
[
52]
(
Table 1, entry 20).
Finally, a comparison was made between the TA-based
binary catalyst TA/NBu I and our previously reported binary
4
couple PG/NBu I (Table 1, entries 21–24). As the TA structure
4
[53]
contains five (substituted) PG units,
the comparison was
made with the synthesis of carbonate 1b mediated by 5 equiv-
alents of PG. After 18 h, a difference between the yields of 1b
promoted by TA (Table 1, entry 12: 79%) and PG (5 equiv.,
Table 1, entry 21: 62%) was observed; for a reduced reaction
time of 6 h, a much smaller difference in the yield of 1b was
noted (cf., Table 1, entries 22 and 23; 35 vs. 32% yield). Under
these conditions, the TA-derived binary catalyst system (i.e.,
[
53]
(12% weight loss upon drying), and the reported TON and
TOF values in Table 2 are uncorrected. The reactive polyphenol
units within TA are nonrandomly distributed during catalysis,
which likely reduces their accessibility relative to the other in-
vestigated catalyst systems. We hypothesize that intramolecu-
lar hydrogen bonding controls the accessibility of the polyphe-
nol units, a phenomenon that cannot be (fully) counterbal-
anced by the use of a moderately polar solvent such as MEK.
As the reactions in MEK needed an increased reaction temper-
ature (808C) for the full dissolution of both catalyst compo-
nents, it seems plausible to assume that this solvent is not
able to break up intra- and intermolecular hydrogen bonding
between the separate polyphenol units at lower temperatures.
Despite these features, at very low TA loading (0.03 mol%) the
relative reactivity seems to indicate that the high local concen-
tration of phenol groups provides some degree of synergy
that leads to efficient catalysis behavior. Thus, one should con-
sider the overall catalytic effect rather than attempting to cor-
relate quantitatively the findings in Table 2.
TA/NBu I) still displayed an appreciably high average TOF of
4
À1
1
95 h . Next, we decided to make a further comparison be-
tween TA and various polyphenol-based structures including
PG, catechol (CC), and propyl gallate (PGA, see Table 2).
Table 2. Screening of various polyphenols in the synthesis of carbonate
[a]
1
b.
[
c]
[d]
Entry
Phenol
Amount
mol%]
Conversion
[%]
Yield
[%]
TON
TOF
[
b]
[
Remarkably, upon comparing the reactivity of PG, CC, and
PGA as catalyst additives (cf., Table 2, entries 3–11), one can
note the lower efficiency of PG among the polyphenols stud-
ied. This result contrasts with our previous findings in which
the catalytic efficiency of PG was markedly better than that ob-
1
2
3
4
5
6
7
8
9
TA
TA
0.03
0.15
0.03
0.15
0.30
0.03
0.15
0.30
0.03
0.15
0.30
0
21
47
10
32
43
15
42
51
11
46
65
8
20
44
9
634
295
303
200
138
461
272
158
344
290
218
–
159
74
76
50
34
115
68
40
86
72
54
–
PG
PG
PG
CC
CC
CC
PGA
PGA
PGA
–
30
41
14
41
48
10
44
64
7
[45]
served for CC at 458C. Intrigued by this discrepancy, we de-
cided to investigate the long-term temperature effect on the
catalytic performance of the polyphenol additives in more
detail by measuring the conversion of 1a to carbonate 1b at
1
0
1
1
8
08C at various time intervals (full data in the Supporting Infor-
[e]
1
2
mation, Tables S1–S3). First, we compared the kinetic profiles
of TA, PG, CC, and PGA during the first 6 h using equimolar
amounts of polyphenol (0.03 mol%; see Figure 1). Interestingly,
both triphenolic derivatives PG and PGA show inferior catalytic
behavior as the conversion already seems to reach a plateau
after 4 h at this catalyst loading whereas TA and CC retain
good activity. These results seem to indicate some catalyst
degradation for the PG- and PGA-based binary catalysts under
the operating conditions.
0
[
(
a] General conditions: 1,2-epoxyhexane (4.15 mmol), pCO2 =1 MPa, NBu
4
I
2.0 mol%), 4 h, 808C, MEK (2.5 mL), AMTEC reactor. [b] Yield determined
by H NMR spectroscopy (CDCl ) using mesitylene as an internal standard;
3
selectivity for the cyclic carbonate was >99%. [c] Total TON per molecule
of catalyst based on reported yields. [d] Average TOF per molecule of cat-
1
4
alyst based on reported yields. [e] Only 2.0 mol% NBu I used.
For the comparative studies, we used a high-throughput ex-
perimentation platform (AMTEC reactor, see the Supporting In-
formation) and estimated the reactivity of the polyphenols
under similar reaction conditions (Table 2, entries 1–11; reac-
tion time 4 h). Moreover, for completion, the conversion ob-
tained in the absence of the polyphenol additive was also ex-
amined (Table 2, entry 12). In the latter case, very low conver-
sion was noted (8%); thus, the production of carbonate 1b
under these conditions is caused by the binary catalysts con-
taining the polyphenols. Comparisons were made between the
To make the comparison more realistic, we also compared
the catalytic performance of TA (0.03 mol%) and PG
(0.15 mol%) over 18 h (Figure 2). After approximately 5 h, the
reactivity of the PG-based binary catalyst decreased drastically
whereas the TA-based system still showed appreciable activity.
This further supports the view that TA is a more stable catalyst
under these conditions and has a longer lifetime than the PG-
based catalyst. The more effective catalytic behavior of TA is
probably the result of a higher local concentration of phenol
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3250
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
4
Scheme 2. Recycling studies with 2a (6 mmol) as the substrate and TA/NBu I
(0.25–0.50 and 2.5–5.0 mol%, respectively) in MEK (15 mL) at pCO2 =1 MPA.
Figure 1. Comparison of the catalytic performance of polyphenol-based
binary catalysts in the conversion of 1a to 1b. For reactions conditions, see
Table 2.
were extracted. Two solid catalyst fractions (FR1 and FR2,
Scheme 2) could be separated and reused in a second catalyst
cycle. The catalyst recycled at 808C showed a significant drop
in conversion (89!27%), whereas a similar though slightly re-
duced conversion drop (54!24%) was noted at 458C. After
the second cycle, the catalyst was separated from the product/
1
substrate phase and subjected to H NMR spectroscopy, IR
spectroscopy, and thermogravimetric analysis (TGA). The com-
bined analytic data showed a clear loss of reactive phenol
units, likely caused by competing reactions between the poly-
phenol and the substrate/nucleophile (for analytical data see
the Supporting Information). This afforded a halohydrin and
NBu -based TA salt byproducts, which were identified by
4
1
H NMR spectroscopy, as also reported previously for the deg-
[
46]
radation of PG under forcing conditions. The regeneration of
the TA structure was probed by acid treatment of the isolated
solid fraction from the recycling experiments. The treatment of
the recycled material with concentrated HCl (37%) regenerated
Figure 2. Comparison of the catalytic performance of TA- (0.03 mol%) and
CC-based (0.15 mol%) binary catalysts in the conversion of 1a to 1b. For re-
actions conditions, see Table 2.
1
the TA species, as evidenced by H NMR and IR spectroscopy
analysis; the spectra showed very high similarity to the spectra
of the commercial product. Thus, the acid treatment indicates
the possibility of catalyst regeneration (see the Supporting In-
formation). The recycled binary catalyst was reused in the syn-
thesis of cyclic carbonate 2b (Scheme 2), and an improved
yield of 51% (vs. 24% for the untreated recovered TA) was de-
termined at 808C.
The mass balance for both TA (FR1, Scheme 2) and NBu4I
(FR2, Scheme 2) was then checked carefully. Although a virtually
complete isolation of the original TA amount (24.6 vs. 25.7 mg;
96%) was noted for FR1, a clear loss of the cocatalytic NBu4I
(FR2, 35.2 vs. 56.8 mg; 62%) was apparent. Finally, we checked
separately the activity of a regenerated TA sample using
groups in the catalyst structure, which likely does not induce
[
45]
strong intermolecular effects. As we reported previously, for
PG (Scheme 1b), potential catalyst degradation may occur
through irreversible proton transfer from PG to the substrate
with the nucleophilic additive also involved. This eventually
translates into lower catalytic efficiencies, and higher reaction
temperatures in combination with very low catalyst loadings
(
0.03–0.15 mol%) may lead quickly to unproductive catalysis
behavior and incomplete substrate conversion. Therefore, it
seems that TA holds promise as a catalytic additive under
dilute conditions whereas much higher concentrations are re-
quired to maintain similarly effective turnovers for the other
polyphenols. To probe the hypothesis that the disappearance
of phenol sites may be responsible for the loss of catalytic ac-
tivity, we decided to investigate the recyclability of the binary
a fresh amount of NBu I in the synthesis of cyclic carbonate
4
1b and compared the conversion with the original data for
0.03 mol% TA at 808C for 18 h (79% conversion; see Table 1,
entry 12). Fortunately, the comparable conversion (75%) found
for the regenerated TA catalyst supports the view that it can
be readily recycled upon acid treatment.
catalyst TA/NBu I in the conversion of 1,2-epoxydodecane (2a,
4
Scheme 2) into carbonate 2b at two different reaction temper-
atures (45 and 808C).
We then further investigated the application of the binary
At both reaction temperatures (see the Supporting Informa-
tion for details), the catalyst was separated readily from the
product-containing hexane phase. The hexane solution was
then concentrated and showed virtually pure carbonate prod-
uct 2b, indicating that no significant catalyst components
TA/NBu I catalyst in the synthesis of carbonate 1b using rela-
tively low amounts of TA (Table 3, 0.01–0.05 mol%) combined
4
with 10 equivalents of NBu I (with respect to TA) as the nucleo-
4
phile. At 0.05 mol% TA (Table 3, entry 1), carbonate 1b was
produced in 86% yield after 24 h with a high TON of 1721 and
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3251
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 3. TA-mediated synthesis of carbonate 1b under various
[a]
conditions.
[
b]
[c]
[d]
Entry
Loading [mol%]
t
Conversion Yield
TON
TOF
TA
NBu
4
I
[h] [%]
[%]
1
2
3
4
5
6
7
8
0.05
–
0.05
–
0.01
–
0.01
–
0.5
0.5
0.5
0.5
0.1
0.1
0.1
0.1
24
24
66 100
66
24
24
66
66
96
36
86
32
100
76
24
16
1721
–
1985
–
2220
–
4458
–
72
–
30
–
92
–
68
–
76
26
18
53
55
45
48
0
[
a] General conditions: 1,2-epoxyhexane (10 mmol), pCO2 =1 MPa, 808C,
1
MEK (5 mL), AMTEC reactor. [b] Yield determined by H NMR spectroscopy
CDCl ) with mesitylene as an internal standard; selectivity for the cyclic
(
3
carbonate was >99%. [c] Total TON per TA equivalent based on reported
yields. [d] Average TOF per TA equivalent based on reported yields.
À1
an average TOF of 72 h . Notably, in the absence of TA
(
Table 3, entry 2), carbonate 1b was produced in only 32%
yield. An increased total reaction time of 66 h (Table 3, en-
tries 3 and 4) reduced this difference as expected, which may
also be an effect of partial catalyst degradation. When reduc-
ing the TA loading further to 0.01 mol% (Table 3, entries 5 and
7
) the difference between the binary couple and the catalyst
system comprising the iodide nucleophile alone became less
significant despite the higher and valuable TONs observed
(
2220 after 24 h, 4458 after 66 h). Clearly, the long-term stabili-
ty plays a key role in the attainment of high average activity.
Next, the substrate scope was investigated (see Scheme 3)
under conditions that would enable high conversion of the ep-
oxide substrates 1a–16a at the reported temperature (808C)
and pressure (1 MPa). All carbonate products 1b–16b were
produced in a high-throughput reactor system at a 2 mmol
substrate scale. Most of the products were obtained at high
conversion and selectivity with good-to-excellent isolated
yields of up to 94% (except for 6b; 57%). Under these reac-
tion conditions, the synthesis of 1b was also probed in the ab-
sence of TA and only a low yield of 22% was achieved. The
Scheme 3. Substrate scope for the TA-mediated synthesis of organic carbo-
nates 1b–16b. General conditions: epoxide (2 mmol), TA (0.5 mol%), NBu4I
(5 mol%), 808C, pCO20=1 MPa, 18 h, MEK (5 mL), AMTEC reactor; under
these conditions, the yield obtained for 1b was only 22% in the absence of
TA.
Conclusions
binary catalyst TA/NBu I tolerates several functional groups in-
We present here a new binary catalyst system based on a natu-
rally occurring and fairly inexpensive polyphenol, that is, tannic
acid (TA), which shows excellent catalytic reactivity at excep-
tionally low loadings. Therefore, this system is an attractive
and sustainable organocatalytic alternative. Comparative catal-
ysis studies have indicated that some degree of synergy be-
tween the various polyphenol units within the TA structure
may help to increase the catalyst lifetime; this provides a con-
ceptually interesting approach for the further improvement of
4
cluding alcohol, alkyl halide, heterocyclic ring systems, carb-
amate, ether, alkene, and alkyne groups. Furthermore, the ster-
ically more hindered substrate 9a could also be converted
(
69%) and provided 40% isolated yield of carbonate 9b
whereas the internal epoxide 10a resulted in much poorer re-
sults, as expected, in line with the more challenging nature of
[43]
this conversion for organocatalytic catalyst systems.
The conversion of internal epoxides was computed to be
more energetically demanding (cf., higher kinetic barriers) than
polyphenol-based organocatalysis in the area of CO conver-
2
[54]
the conversion of terminal epoxides. Nonetheless, the forma-
tion of all carbonates was mediated by only 0.5 mol% TA; this
is an attractive feature within the context of providing a sus-
tainable and reactive alternative for metal-based carbonate-for-
mation reactions.
sion. Our future work will be focused on merging these con-
cepts with the design of organocatalyst systems with im-
proved reactivity and stability for similar CO conversions and
2
the preferable use of hydrogen-bonding as a substrate-activa-
tion strategy.
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3252
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
0
.5 MPa) were performed before final stabilization of the pressure
Experimental Section
at 1 MPa. The reactor was then heated to the required tempera-
ture, and the mixture was stirred for an additional 18 h. Then, the
reactor was cooled and depressurized, and the reaction mixture
was separated from the precipitate (Solid 1; FR1) and moved to
a flask. The solvent was removed under vacuum, and hexane
General
Methyl ethyl ketone and carbon dioxide (purchased from PRAXAIR)
were used as received without further purification or drying. All
phenolic compounds were purchased from Sigma–Aldrich and
used without further purification; the TA was reagent grade, see
also Ref. [53]. The NMR spectra were recorded using Bruker AV-400
and AV-500 spectrometers and referenced to the residual deuterat-
ed solvent signals. The FTIR measurements were performed with
a Bruker Optics FTIR Alpha spectrometer equipped with a deuterat-
(
80 mL) was added. Then, the hexane solution was cooled to
À308C. After 3 h, the flask was warmed to room temperature, the
solution was filtered, and the collected precipitate (Solid 2; FR2)
was washed with hexane (80 mL). The combined organic phases
were then concentrated under vacuum to afford the pure cyclic
carbonate. For a new reaction cycle, Solid 1 (FR1) was put into
a stainless-steel reactor, and Solid 2 (FR2) was dissolved in MEK
ed triglycine sulfate (DTGS) detector and a KBr beam splitter at
À1
4
cm resolution.
(
15 mL) and added to the same reactor. A new portion of 1,2-epox-
ydodecane (6.00 mmol, 1.25 g) was added. Finally, the procedure
continued as described above. To regenerate the catalyst, FR1 was
treated with concentrated HCl for 18 h. Then, the mixture was fil-
tered and washed with diethyl ether and hexane. The obtained
precipitate was then dried under vacuum and combined with FR2
in a new catalytic cycle in MEK (15 mL) and transferred to a pres-
sure reactor, and 1,2-epoxydodecane (6.0 mmol, 1.25 g) was added.
Finally, the procedure continued as described above.
Standard autoclave screening experiments
All reactions were performed in a 30 mL stainless-steel reactor. In
a typical experiment, a solution of TA (0.050 mol%, 7.05 mg), NBu4I
(
2.00 mol%, 61.3 mg), 1,2-epoxyhexane (8.30 mmol, 831 mg), and
mesitylene (1.00 mL, 7.18 mmol) in MEK (5 mL) was added to
a stainless-steel reactor. Three cycles of pressurization and depres-
surization of the reactor (pCO2 =0.5 MPa) were performed, and the
pressure was stabilized at 1 MPa. The reactor was heated to the re-
quired temperature, and the mixture was stirred for a further 18 h. Acknowledgements
Then, the reactor was cooled and depressurized, and an aliquot of
1
the solution was analyzed by means of H NMR spectroscopy using
C.M. gratefully thanks the Marie Curie COFUND Action from the
European Commission for cofinancing a postdoctoral fellowship.
G.F. kindly acknowledges financial support from the European
Community through a Marie Curie Intra-European Fellowship
CDCl as the solvent. The yield was determined using mesitylene
as the internal standard. In all cases, the selectivity for the cyclic
carbonate product was determined to be >99%.
3
(
FP7-PEOPLE-2013-IEF, project RENOVACARB, Grant Agreement
no. 622587). Further acknowledgements go to ICIQ, ICREA, and
the Spanish Ministerio de Economía y Competitividad (MINECO)
through project CTQ2014-60419-R, and support through Severo
Ochoa Excellence Accreditation 2014–2018 (SEV-2013-0319). Dr.
Marta GimØnez and Cristina Rivero are thanked for their help re-
garding the high-pressure experiments.
Substrate scope experiments
All reactions were performed in an SPR16 Slurry Phase Reactor
(
(
Amtec GmbH). First, TA (0.500 mol%, 17.0 mg) and NBu4I
5.00 mol%, 36.9 mg) were put into the reactors. Then, the reaction
vessels were tested for leaks using N (1.5 MPa) to a final reduced
2
pressure of 0.2 MPa. After the injection into the reactors of the
chosen epoxide (e.g., 2.00 mmol, 200 mg for 1,2-epoxyhexane) in
MEK (5 mL) and mesitylene (10.0 mol%, 24.0 mg) as an internal
standard (IS), the vessels were heated to the desired reaction tem-
perature (T=808C). Once the operating temperature was reached,
Keywords: carbon dioxide
·
homogeneous catalysis
·
hydrogen bonding · oxygen heterocycles · ring expansion
the CO pressure was increased to 1 MPa and the reaction mixture
2
[
[
[
1] Carbon Dioxide as Chemical Feedstock (Ed.: M. Aresta), Wiley-VCH, Wein-
heim, 2010.
2] M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann, F. E. Kühn, Angew.
Chem. Int. Ed. 2011, 50, 8510; Angew. Chem. 2011, 123, 8662.
3] M. Peters, B. Kçhler, W. Kuckshinrichs, W. Leitner, P. Markewitz, T. E.
Müller, ChemSusChem 2011, 4, 1216.
was stirred at the appropriate temperature for 18 h. At the end of
the reaction, the analysis of the crude product was performed as
reported above for the screening phase. The crude products were
obtained by removing the solvent and unreacted substrate under
vacuum (50 Pa). The residue was then dissolved in dichlorome-
thane (DCM, except for the conversion of substrate 4a, for which
the solvent was ethyl acetate) and filtered through silica; after the
removal of the solvent, the pure cyclic carbonate products were
obtained. The identity of each of the carbonate products was con-
firmed by comparison to previously reported literature data, and
full tabular data and copies of all spectra are provided in the Sup-
porting Information.
[4] G. Fiorani, W. Guo, A. W. Kleij, Green Chem. 2015, 17, 1375.
[5] M. Mikkelsen, M. Jørgensen, F. C. Krebs, Energy Environ. Sci. 2010, 3, 43.
[6] D. R. Darensbourg, Chem. Rev. 2007, 107, 2388.
[
[
[
7] M. North, R. Pasquale, C. Young, Green Chem. 2010, 12, 1514.
8] P. P. Pescarmona, M. Taherimehr, Catal. Sci. Technol. 2012, 2, 2169.
9] N. Kielland, C. J. Whiteoak, A. W. Kleij, Adv. Synth. Catal. 2013, 355, 2115.
[
10] O. Jacquet, X. Frogneux, C. Das Neves Gomes, T. Cantat, Chem. Sci.
2013, 4, 2127.
[
[
[
[
11] H. Mizuno, J. Takaya, N. Iwasawa, J. Am. Chem. Soc. 2011, 133, 1251.
12] M. D. Greenhalgh, S. P. Thomas, J. Am. Chem. Soc. 2012, 134, 11900.
13] T. León, A. Correa, R. Martin, J. Am. Chem. Soc. 2013, 135, 1221.
14] Y. Li, I. Sorribes, T. Yan, K. Junge, M. Beller, Angew. Chem. Int. Ed. 2013,
Recycling experiments
All reactions were performed in a 45 mL stainless-steel reactor. In
a typical experiment, a solution of TA (0.250 mol%, 25.5 mg),
5
2, 12156; Angew. Chem. 2013, 125, 12378.
[
15] C. Martín, G. Fiorani, A. W. Kleij, ACS Catal. 2015, 5, 1353.
nBu NI (2.50 mol%, 55.4 mg), and 1,2-epoxydodecane (6.00 mmol,
4
[16] B. Schäffner, F. Schäffner, S. P. Verevkin, A. Bçrner, Chem. Rev. 2010, 110,
1
.25 g) in MEK (15 mL) was added to a stainless-steel reactor. Three
4554.
cycles of pressurization and depressurization of the reactor (pCO2
=
[17] J. H. Clements, Ind. Eng. Chem. Res. 2003, 42, 663.
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3253
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
[
[
18] See, for example: D. J. Darensbourg, A. I. Moncada, W. Choi, J. H. Rei-
benspies, J. Am. Chem. Soc. 2008, 130, 6523.
19] C. J. Whiteoak, N. Kielland, V. Laserna, E. C. Escudero-Adµn, E. Martin,
A. W. Kleij, J. Am. Chem. Soc. 2013, 135, 1228.
[43] B. Chatelet, L. Joucla, J.-P. Dutasta, A. Martinez, V. Dufaud, Chem. Eur. J.
2014, 20, 8571.
[44] M. E. Wilhelm, M. H. Anthofer, M. Cokoja, I. I. E. Markovits, W. A. Herr-
mann, F. E. Kühn, ChemSusChem 2014, 7, 1357.
[45] C. J. Whiteoak, A. Nova, F. Maseras, A. W. Kleij, ChemSusChem 2012, 5,
2032.
20] T. Ema, Y. Miyazaki, S. Koyama, Y. Yano, T. Sakai, Chem. Commun. 2012,
4
8, 4489.
[
[
21] J. MelØndez, M. North, P. Villuendas, Chem. Commun. 2009, 2577.
22] J. Song, Z. Zhang, S. Hu, T. Wu, T. Jiang, B. Han, Green Chem. 2009, 11,
[46] C. J. Whiteoak, A. H. Henseler, C. Ayats, A. W. Kleij, M. A. Pericàs, Green
Chem. 2014, 16, 1552.
1
031.
[47] Quite recently, Mattson et al. reported interesting and highly active sila-
nediol catalysts for organic carbonate formation, see: A. M. Hardman-
Baldwin, A. E. Mattson, ChemSusChem 2014, 7, 3275.
[48] S. Gennen, M. Alves, R. MØreau, T. Tassaing, B. Gilbert, C. Detrembleur,
C. Jerome, B. Grignard, ChemSusChem 2015, 8, 1845.
[49] MEK was previously used by us and others as it is a good medium for
[
[
23] J. Sun, S. Zhang, W. Cheng, J. Ren, Tetrahedron Lett. 2008, 49, 3588.
24] Z.-Z. Yang, Y.-N. Zhao, L.-N. He, J. Gao, Z.-S. Yin, Green Chem. 2012, 14,
5
19.
25] Z. Z. Yang, L.-N. He, C.-X. Miao, S. Chanfreau, Adv. Synth. Catal. 2010,
52, 2233.
26] J. L. Song, Z. F. Zhang, B. X. Han, S. Q. Hu, W. J. Li, Y. Xie, Green Chem.
008, 10, 1337.
[
[
[
3
2
CO dissolution: A. Decortes, M. Martínez Belmonte, J. Benet-Buchholz,
2
A. W. Kleij, Chem. Commun. 2010, 46, 4580.
[50] Compared with the reaction listed in Table 1, entry 5, the background
27] S. G. Liang, H. Z. Liu, T. Jiang, J. L. Song, G. Y. Yang, B. X. Han, Chem.
Commun. 2011, 47, 2131.
reactions in the presence of 2 mol% of NBu
4
Cl (44%) or 2 mol%
[
[
28] C. Qi, J. Ye, W. Zeng, H. Jiang, Adv. Synth. Catal. 2010, 352, 1925.
29] J. Sun, W. G. Cheng, Z. F. Yang, J. Q. Wang, T. T. Xu, J. Y. Xin, S. J. Zhang,
Green Chem. 2014, 16, 3071.
30] J. Sun, L. J. Han, W. G. Cheng, J. Q. Wang, X. P. Zhang, S. J. Zhang, Chem-
SusChem 2011, 4, 502.
31] J. Q. Wang, J. Sun, W. G. Cheng, K. Dong, X. P. Zhang, S. J. Zhang, Phys.
Chem. Chem. Phys. 2012, 14, 11021.
32] J. Q. Wang, D. L. Kong, J. Y. Chen, F. Cai, L.-N. He, J. Mol. Catal. A 2006,
NBu Br (60%) show similar or improved conversion rates. However, in
4
the presence of TA, the chloride-based binary catalyst (cf., Table 1,
entry 18) is somewhat less effective whereas the bromide- and iodide-
based ones (Table 1, entries 9 and 17) give quantitative conversions. As
[
[
[
[
[
4
our previous studies were performed using NBu I as the nucleophilic
additive, we continued using this cocatalytic mediator for comparative
reasons.
51] A. Decortes, A. W. Kleij, ChemCatChem 2011, 3, 831.
[
[
52] Possible product inhibition was also probed using a 1:1 mixture of the
cyclic carbonate 1b and its precursor 1a, taking the reaction in the ab-
sence of the product as a reference. In both cases, a similar conversion
was noted at the end of the reaction; therefore, no (significant) product
inhibition occurs (see the Supporting Information for details).
53] Commercially available TA is usually a combination of polygalloyl gluco-
ses or polygalloyl quinic acid esters, with the number of galloyl moiet-
ies per molecule depending on the plant source. Here, we have used
the commercial product from Aldrich (ACS reagent, 500 g=77.50 E),
which loses approximately 12% of its weight upon drying. Further, as
the general structure of TA contains five pseudo-PG units, we have
compared the catalytic performance of TA (0.03 mol%; 1 equiv.) with
that of PG (0.15 mol%; 5 equiv.).
2
49, 143.
33] Q.-W. Song, L.-N. He, J.-Q. Wang, H. Yasuda, T. Sakakura, Green Chem.
013, 15, 110.
2
34] M. H. Anthofer, M. E. Wilhelm, M. Cokoja, M. Drees, W. A. Herrmann, F. E.
Kühn, ChemCatChem 2015, 7, 94.
35] V. Caló, A. Nacci, A. Monopoli, A. Fanizzi, Org. Lett. 2002, 4, 2561.
36] X. Chen, J. Sun, J. Q. Wang, W. G. Cheng, Tetrahedron Lett. 2012, 53,
[
[
[
2
684.
[
[
[
[
37] T. Werner, H. Buettner, ChemSusChem 2014, 7, 3268. See also: C. Kohrt,
T. Werner, ChemSusChem 2015, 8, 2031.
38] T. Werner, N. Tenhumberg, H. Buettner, ChemCatChem 2014, 6, 3493–
3
500.
39] Y. Tsutsumi, K. Yamakawa, M. Yoshida, T. Ema, T. Sakai, Org. Lett. 2010,
2, 5728.
[
54] F. Castro-Gómez, G. Salassa, A. W. Kleij, C. Bo, Chem. Eur. J. 2013, 19,
1
6289.
40] R. A. Watile, K. M. Deshmukh, K. P. Dhake, B. M. Bhanage, Catal. Sci. Tech-
nol. 2012, 2, 1051.
[
[
41] Y. M. Shen, W. L. Duan, M. Shi, Adv. Synth. Catal. 2003, 345, 337.
42] B. Chatelet, L. Joucla, J.-P. Dutasta, A. Martinez, K. C. Szeto, V. Dufaud, J.
Am. Chem. Soc. 2013, 135, 5348.
Received: May 26, 2015
Revised: July 20, 2015
Published online on September 2, 2015
ChemSusChem 2015, 8, 3248 – 3254
www.chemsuschem.org
3254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim