Catalytic Coupling of Carbon Dioxide with Terpene Scaffolds: Access to Challenging Bio-Based Organic Carbonates

Article abstract of DOI:10.1002/cssc.201600238

The challenging coupling of highly substituted terpene oxides and carbon dioxide into bio-based cyclic organic carbonates catalyzed by Al(aminotriphenolate) complexes is reported. Both acyclic as well as cyclic terpene oxides were used as coupling partners, showing distinct reactivity/selectivity behavior. Whereas cyclic terpene oxides showed excellent chemoselectivity towards the organic carbonate product, acyclic substrates exhibited poorer selectivities owing to concomitant epoxide rearrangement reactions and the formation of undesired oligo/polyether side products. Considering the challenging nature of these coupling reactions, the isolated yields of the targeted bio-carbonates are reasonable and in most cases in the range 5060 %. The first crystal structures of tri-substituted terpene based cyclic carbonates are reported and their stereoconnectivity suggests that their formation proceeds through a double inversion pathway.

Full text of DOI:10.1002/cssc.201600238

DOI: 10.1002/cssc.201600238
Full Papers
Catalytic Coupling of Carbon Dioxide with Terpene
Scaffolds: Access to Challenging Bio-Based Organic
Carbonates
Giulia Fiorani,*[a] Moritz Stuck,[a] Carmen Martín,[a] Marta Martínez Belmonte,[a] Eddy Martin,[a]
Eduardo C. Escudero-Adµn,[a] and Arjan W. Kleij*[a, b]
The challenging coupling of highly substituted terpene oxides
and carbon dioxide into bio-based cyclic organic carbonates
catalyzed by Al(aminotriphenolate) complexes is reported.
Both acyclic as well as cyclic terpene oxides were used as cou-
pling partners, showing distinct reactivity/selectivity behavior.
Whereas cyclic terpene oxides showed excellent chemoselec-
tivity towards the organic carbonate product, acyclic substrates
exhibited poorer selectivities owing to concomitant epoxide
rearrangement reactions and the formation of undesired oligo/
polyether side products. Considering the challenging nature of
these coupling reactions, the isolated yields of the targeted
bio-carbonates are reasonable and in most cases in the range
50–60%. The first crystal structures of tri-substituted terpene
based cyclic carbonates are reported and their stereoconnec-
tivity suggests that their formation proceeds through a double
inversion pathway.
Introduction
The conversion of carbon dioxide (CO2) into useful organic
products is a prominent strategy towards the valorization of
this renewable-carbon-based feed stock.[1] Despite the exciting
progress in the area of catalytic conversion of CO2 into various
products such as polymers,[2] synthetic fuels[3] and fine chemi-
cals,[4] a remaining challenge in the area of CO2 conversion is
represented by the limited number of chemical functionalities
that can be derived from this C1 building block. So far, most ef-
forts have been devoted to CO2 conversion through reductive
and non-reductive coupling chemistry. Within the latter cate-
gory, CO2 is incorporated into an organic substrate without af-
fecting the oxidation state of the carbon center. The most
prominent among this type of transformation is the formation
of organic carbonates using epoxides as reaction partners.[5]
The catalytic coupling of epoxides and CO2 is an atom-
efficient reaction that has been well-studied and developed
over the years, with important contributions to solve key chal-
lenges associated with this process, including enhanced reac-
tivity/selectivity,[6] improved overall sustainability,[7] and the de-
velopment of commercially attractive, recyclable catalysts.[8]
However, there are still important open challenges in this spe-
cific area of research, including the limited number of sub-
strates that can be used owing to the lack of efficient catalytic
strategies for their coupling. Highly substituted epoxides that
upon coupling with CO2 would give 1,3-dioxolan-2-ones, in-
cluding those with a 4,5-disubstitution[9] and 4,4’,5-trisubstitu-
tion,[10] remain challenging reaction partners despite the pres-
ence of such patterns in many naturally occurring plant-based
organic carbonates (Figure 1).[11]
We have recently become interested in the use of renewable
feed stocks such as terpenes as substrates for CO2 conver-
[a] Dr. G. Fiorani, M. Stuck, Dr. C. Martín, Dr. M. M. Belmonte, Dr. E. Martin,
E. C. Escudero-Adµn, Prof. Dr. A. W. Kleij
Institute of Chemical Research of Catalonia (ICIQ)
The Barcelona Institute of Science and Technology
Av. Pa¨ısos Catalans 16, 43007 Tarragona (Spain)
E-mail: akleij@iciq.es
gfiorani@iciq.es
[b] Prof. Dr. A. W. Kleij
Catalan Institute of Research and Advanced Studies (ICREA)
Pg. Lluís Companys 23, 08010 Barcelona (Spain)
Figure 1. Top: naturally occurring organic carbonates based on complex
scaffolds; the carbonate unit is marked in blue. Bottom: some representative
terpene scaffolds with similar di- and tri-substituted alkene units highlighted
in yellow.
Supporting Information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cssc.201600238.
ChemSusChem 2016, 9, 1304 – 1311
1304
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
sion.[12] Terpenes, including limonene (Figure 1),[10,13a] have in-
creasingly found applications in polymer science and sustaina-
ble materials.[13b] The formation of cyclic organic carbonates
from terpene scaffolds ultimately results in 100% bio-derived
compounds with relatively unexplored potential,[14] and we
thus set out to develop an efficient catalytic process for their
formation. Here we describe the AlIII catalyzed coupling be-
tween CO2 and both cyclic and acyclic terpene oxides affording
new bio-based carbonates. The first crystal structures of ter-
pene derived carbonates are also presented and the observed
stereoconnectivity is linked to a double inversion pathway in
the coupling between the oxirane precursors and CO2.
Table 1. Screening of the reaction conditions in the coupling of 1-meth-
ylcyclohexene oxide 1a and CO2 to afford cyclic carbonate 1b.[a]
Entry
Catalyst
PPNCl [mol%]
t [h]
Yield 1b[b] [%]
1
2
3
4
5
6
7
8
9
10
11
12
13
–
–
–
3.0
3.0
3.0
–
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
1.0
24
48
66
24
24
24
24
24
48
66
66
66
66
18
19
31
0
26
7
22
17
40
49[c]
1[d]
34
14
A
A
B
C
D
A
A
A
B
A
Results and Discussion
Terpene Carbonates with bicyclic structures
We first devoted our attention to the conversion of 1-methyl-
cyclohexene oxide 1a to the corresponding cyclic carbonate
product 1b by CO2 insertion. We were unable to detect any
formation of cyclic carbonate 1b at temperatures below 708C.
This is in line with a previously reported density functional
theory analysis of the coupling between limonene oxide (LO,
a structurally related trisubstituted oxirane) and CO2, showing
that the kinetic barrier towards cyclic carbonate formation
(through backbiting) is comparatively higher than the one as-
sociated with chain propagation.
[a] Reaction conditions: 1-methylcyclohexene oxide 1a (5 mmol), MEK
(methyl ethyl ketone, 1.0 mL), catalyst (1.0 mol%, when present), p(CO2)=
1.0 MPa (10 bar), T=858C. [b] Isolated yield, selectivity for 1b was >99%
1
in all cases determined by H NMR. [c] The NMR yield was 75%, in the ab-
sence of complex A the NMR and isolated yield were both 31% (see
entry 3). [d] No MEK used.
Indeed, the only product observed during the coupling of
LO and CO2 at lower temperatures (40–708C) was the poly(li-
monene)carbonate.[12a] Hence, a higher reaction temperature
(858C) was used to direct the chemoselectivity towards the
cyclic carbonate product. The substrate 1-methylcyclohexene
oxide 1a was taken as a representative terpene model sub-
strate (Table 1). AlIII aminotriphenolate complexes A–D were
used as catalysts together with bis(triphenylphosphine)imini-
um chloride (PPNCl) as the nucleophilic additive. Although in
the presence of the nucleophile only (entries 1–3, Table 1) con-
version to 1b is noted, the addition of a Lewis acid (A–D) sub-
stantially improves the overall kinetics, most noteworthy when
complex A is used (entries 5, 9 and 10). As expected, in the ab-
sence of PPNCl (entry 4) no conversion was noted. A lower
loading of PPNCl (entry 13, 1 mol%) reduced the yield of 1b to
only 14%, indicating the requirement for a higher loading of
the nucleophile. In the absence of solvent [methyl ethyl
ketone (MEK)] virtually no conversion was obtained (see
entry 11), underlining the critical role of this solvent to accom-
modate the appropriate mixing of all the reactants. The struc-
ture of 1b was confirmed by X-ray analysis and is depicted in
Figure 2.
Figure 2. X-ray molecular structure for 1b. Selected bond lengths (Š) and
angles (8) for 1b: C(1)–O(1)=1.1988(18), C(1)–O(2)=1.3442(19), C(1)–
O(3)=1.3487(18); O(2)–C(1)–O(3)=111.47(12), O(1)–C(1)–O(2)=124.23(14).
examined LO as a substrate using the best conditions deter-
mined for the synthesis of 1b. Whereas pure trans-LO was con-
verted into limonene carbonate 2b in appreciable conversion
(73%) and yield (57%), the use of the pure cis isomer only led
to a 4% yield of cis-3b (by NMR) of its respective carbonate.
The extremely sluggish outcome observed for CO2 coupling
with the above mentioned cis isomer was confirmed using
commercially available LO (a 60:40 trans/cis mixture) resulting
in the almost exclusive formation of trans-4b from trans-LO in
43% isolated yield. The proposed trans-connectivity pattern in
2b was confirmed by NMR analysis and further substantiated
by X-ray diffraction (see Figure 4). As far as we know, Figure 4
represents the first crystallographically characterized terpene
based carbonates. The trans nature of carbonate 2b enforces
With these optimal conditions in hand we decided to inves-
tigate a wider scope of trisubstituted carbonates derived from
cyclic epoxides based on terpene scaffolds (see Figure 3).
Owing to the challenging nature of these coupling reactions,
conversion of the respective terpene epoxides proceeds slowly
and reaction times of 66 h are typically required (Figure 3).
Longer reaction times at 858C did not improve the conversion
levels and neither did higher reaction temperatures.[15] We first
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1305
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 3. Investigated scope for the formation of trisubstituted bicyclic terpene carbonates. Reaction conditions (unless stated otherwise): 1.0 mol% AltBu
,
3.0 mol% PPNCl, 1.0 mL MEK, 66 h, 858C, p(CO2)=1.0 MPa. The diastereoisomeric ratios (dr values, major isomer indicated) and selectivity towards the carbon-
1
ate products were also determined by H NMR; the dr relates to the relative position of both substituents on the cyclohexyl ring. [a] Conversion relates to the
1
total amount of epoxide groups converted to cyclic carbonate ones determined by H NMR (CDCl3) integration; the NMR yield of 6b was 74%. [b] The NMR
yield of 7b was 27% with 24% of 6b present along with polyether byproduct. [c] Reaction performed at 1208C. [d] The major stereoisomer assignment was
not possible by 1D/2D NMR owing to a combination of complex patterns.
cyclic carbonate formation from LO is in agreement with the
higher computed kinetic barriers compared with the corre-
sponding copolymerization process.[12a]
Coupling between CO2 and other (bi)cyclic terpene oxides
(including carvone oxide, limonene diepoxide and menthene
oxide) exhibited a similar reactivity giving the respective ter-
pene (mono)carbonates 5b, 6b and 8b in 45–52% yield and
high chemoselectivity.[16] We also attempted to optimize the
synthesis of limonene bis-carbonate 7b (27%) using more forc-
ing reaction conditions (1208C) but at the cost of the chemo-
selectivity. This is likely because of the observed increase of
the viscosity of the reaction mixture during the reaction of li-
monene diepoxide, limiting the diffusion of CO2 towards the
reactive center; the presence of significant amounts of oligo/
1
polyether product, as determined by H NMR, is in line with
this assumption (see Supporting Information for details). The
molecular structures of terpene (bis)carbonates 2b, 5b, 6b
and 7b were determined by X-ray crystallography and con-
firmed their proposed formulations. Notably, as far as we
know, these X-ray analyses represent the first ones reported
for this family of bio-based carbonates, and bis-carbonate 7b
was isolated for the first time and fully characterized.[10c]
Figure 4. X-ray molecular structures for terpene carbonates trans-2b, cis-5b,
6b and trans-7b with the adopted numbering scheme. Selected bond
lengths (Š) and angles (8) for 2b: C(2)–O(2)=1.221(8), C(2)–O(1)=1.326(8),
C(2)–O(3)=1.344(7); O(2)–C(2)–O(3)=123.1(7), O(1)–C(2)–O(3)=112.8(5); se-
lected data for 5b: C(1)–O(1)=1.200(2), C(1)–O(2)=1.350(2), C(1)–
O(3)=1.338(2); O(2)–C(1)–O(3)=111.73(14), O(1)–C(1)-O(2)=123.53(17); se-
lected data for 6b: C(1)–O(1)=1.330(7), C(1)–O(2)=1.197(7), C(1)–
O(3)=1.333(8), C(6)–O(4)=1.448(9); O(1)–C(1)–O(3)=112.6(5), O(1)–C(1)–
O(2)=124.5(7); selected data for 7b: C(1)–O(1)=1.336(7), C(1)–
O(2)=1.474(6), C(1)–O(3)=1.354(7), C(11)–O(4)=1.362(7), C(11)–
O(5)=1.182(8), C(11)–O(6)=1.332(7); O(1)–C(1)–O(3)=111.3(5), O(1)–C(1)–
O(2)=124.6(5), O(4)–C(11)–O(6)=110.7(6), O(4)–C(11)–O(5)=124.0(6).
Formation of Acyclic Terpene Carbonates
We next turned our focus on the formation of terpene carbo-
nates based on acyclic scaffolds. Our investigation started with
the epoxide derived from citronellyl acetate 9a (Table 2) using
the optimal conditions developed for the synthesis of most of
the bicyclic terpene carbonates (i.e., 858C, 1.0 MPa, 66 h;
3.0 mol% PPNCl) but using a chloride-substituted Al(aminotri-
phenolate) complex (AlCl, 1 mol%) as the Lewis acid.[17] Where-
as the conversion of 9a already proceeds at 708C (entry 1)
with a 31% NMR yield for the corresponding cyclic carbonate
9b, at 858C (entry 3) we unexpectedly observed a poorer out-
come in terms of chemoselectivity. Increasing the scale of the
reaction (entry 4) did result in further improvement, but a com-
bination of increasing the scale and applying a higher CO2
the idea that these conversions occur with complete retention
of configuration, as would be expected from a double inver-
sion pathway.[9b,12c] Interestingly, we previously observed that
cis-LO reacts faster than its trans isomer in copolymerization re-
actions, resulting in an almost exclusive formation of a trans
configured copolymer (i.e., a formal inversion occurred). Con-
versely, at higher temperatures (858C) cyclic carbonate forma-
tion starting from the trans isomer seems to be the favored
process (formation of trans-2b and trans-4b, Figure 3). The ob-
served requisite for higher reaction temperatures in the case of
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1306
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
nolate-assisted rearrangement of the epoxide 9a;
such conversions have precedence though are typi-
cally performed with strong bases.[20] Additionally,
Lewis acid induced Meinwald-type rearrangement
of epoxides to ketones has been well-
established.[21]
Table 2. Screening and optimization of the reaction conditions in the coupling of acy-
clic terpene-based epoxides 9a–14a and CO2 to afford their cyclic carbonate products
9b–14b.[a] The commercial olefinic, terpene precursors are drawn together with their
epoxidized derivatives.
To rule out the possibility that the terpene scaffold
itself induces the aforementioned side-product for-
mation, we performed the synthesis of the non-
functionalized, tri-substituted cyclic carbonate 14b
derived from trimethyloxirane 14a; in essence the
same product distribution was observed (Table 2, en-
tries 10 and 11, Figure 5), confirming that side-prod-
uct formation does not depend on the nature of the
terpene scaffold. Furthermore, the less polar and
more volatile nature of 14a seems to influence the
course of the coupling reaction as for this substrate
the use of the tert-butyl-substituted Al(aminotriphe-
nolate) complex (AltBu) gave better results (Table 2,
entries 10 and 11).
Entry Substrate Catalyst p(CO2)
T
t
Epoxide[b] AA[b] K[b] CC[b] PE[b]
[MPa] [8C] [h] [%]
[%]
[%] [%]
[%]
1
2
3
4[c]
5
6
7
8
9[c]
10
11
12[c]
13[c]
14[c]
15[d]
9a
9a
9a
9a
9a
9a
9a
9a
9a
14a
14a
10a
11 a
12a
13a
AlCl
–
AlCl
AlCl
–
AlCl
AlCl
–
AlCl
AltBu
AlCl
AlCl
AlCl
AlCl
AlCl
1.0
1.0
1.0
1.0
1.0
3.0
4.0
4.0
4.0
1.0
1.0
4.0
4.0
4.0
1.0
70
70
85
85
85
70
70
70
85
85
85
85
85
85
85
66
66
66
7
57
22
6
2
5
12
19
<1
<1
2
7
2
2
2
9
31
21
8
35
24
2
22
36
23
28
13
16
35
16
9
<1 17
20
14 52
19
3
66 <1
66
66
66
66
66
24
24
66
66
66
17
16
23
27
3
23
40
8
7
Having determined the optimized conditions for
the preparation of acyclic terpene carbonates
(Table 2, entry 9) we then extended the scope of
these transformations to terpene oxides 10a–13a,
which were converted into the respective cyclic car-
bonate products 10b–13b (entries 12–15). In general,
apart from the desired carbonate products, similar
distributions of side-products (allylic alcohol, ketone
and oligo/polyether) were observed, together with
small amounts of the starting material. The desired
cyclic carbonate products were conveniently isolated
by column purification of the crude mixtures in yields
of up to 61% (for 11 b and 12b; Figure 5). Notably,
using a myrcene based scaffold (13b) led to the for-
mation of a complex reaction mixture as other side
reactions involving the conjugated double bonds oc-
curred. Indeed, we were not able to isolate a pure
sample of 13b using chromatographic separation as
further degradation occurred during the purification
step. It is well-known that isoprenes show good poly-
<1 45
<1 36
<1 15
7
6
9
8
7
3
56
51
31
47
61
71
5
3
5
4
66 <1
15
16 19
30
[a] Reactions conditions: substrates 9a–14a (2.5 mmol), MEK (methyl ethyl ketone,
0.5 mL), catalyst (1.0 mol%, when present), PPNCl (3.0 mol%), pressure/time/tempera-
ture as indicated. [b] NMR yields from 1H NMR analysis using mesitylene as internal
standard; for the use of 9a: allylic alcohol (AA) peak at d=4.93 ppm (septet) and
4.84 ppm (quintet), ketone (K) peak at d=2.61 ppm (quintet), epoxide at d=
2.68 ppm (triplet), cyclic carbonate (CC) peak at d=4.20 ppm (multiplet) and oligo/
polyether peak in the region d=3.3–3.8 ppm. [c] Substrate amount was 5.0 mmol.
[d] The 1H NMR spectrum of the crude mixture showed apart from the CC product
13b a number of peaks in the region 5.0–5.5 ppm attributed to unknown olefin-con-
taining byproducts.
pressure (entries 6, 7 and 9; 3.0–4.0 MPa) gave slightly more
satisfactory results with the formation of the corresponding
cyclic carbonate in 56% NMR yield (51% isolated, Figure 5)
and further suppressed byproduct formation. Notably, the re-
action performed with only the nucleophilic additive (PPNCl)
proceeded with significantly lower efficiency (entries 2 and 8).
It should be noted that the overall chemoselectivity towards
the carbonate 9b was much poorer than the ones observed
starting from the aforementioned bicyclic terpene oxides.[18]
Beside the formation of substantial amount of oligo/polyether
(Table 2, entry 9, 28%; 9e),[19] two other side-products was
noted. These side-products were isolated and fully character-
ized for the citronellyl acetate case (see Experimental Section
and Supporting Information for details), and were unambigu-
ously identified as an allylic alcohol (9c) and a ketone (9d),
both arising from the rearrangement of the starting epoxide
9a (see Figure 6). The allylic alcohol 9c may arise from a phe-
merization potential in the presence of Lewis acids[22] and this
subunit is also present in 13b (highlighted in green). Thus, it
seems reasonable to assume that further impurities to the
ones reported in Table 2 (entry 15) may also be ascribed to
oligo-myrcene species (see Supporting Information for details)
resulting from diene activation catalyzed by complex B.
The significant byproduct formation generally observed for
the synthesis of acyclic terpene carbonates 9b–14b as op-
posed to the chemoselective formation of bicyclic carbonate
structures 1b–5b and 8b can be explained by taking into ac-
count the different transformations involving the metal-
alkoxide intermediate formed after nucleophilic ring-opening
of the epoxide group.[23] In particular, it should be noted that
polyether formation remains a competitive process even for
CO2 coupling reactions with acyclic substrates performed at
higher pressure (4.0 MPa, entry 4 and 9, Table 2). This observa-
tion suggests that after the initial coordination of the epoxide
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1307
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
(Table 2). Whereas the use of AlCl gives rise to virtually full con-
version of the starting epoxide 9a (97%), significant formation
of oligo/polyether 9e (28%) is observed, with a calculated che-
moselectivity towards the carbonate 9b of 57% (56% NMR
yield). Notably, the presence of AltBu results in lower conversion
of epoxide 9a (71%) alongside with formation of a much
lower amount of oligo/polyether (8%) with an overall chemo-
selectivity of 67% for carbonate 9b (46% NMR yield).
In contrast, the use of AltBu with bicyclic terpene oxides led
to the chemoselective formation of the respective terpene car-
bonates (Figure 3). This is in agreement with our former notion
that this Lewis acidic complex is primarily involved at the initial
stage of the coupling between the epoxide and CO2, as previ-
ously proposed for oxetane/CO2 coupling reactions promoted
by AltBu [24]
Upon formation of a metal hemi-carbonate inter-
.
Figure 5. Schematic structures of the terpene derived carbonates 9b–14b,
their isolated yields after chromatographic purification and diastereoisomeric
ratios (dr) of 9b and 11 b. For terpene carbonates 10b and 12b the amount
of (E) and (Z) isomers, respectively, was>98%.
mediate, the metal-mediated formation of the cyclic carbonate
is sterically disfavored. This causes the release of the hemi-car-
bonate from the coordination sphere of the metal following
fast outer-sphere cyclization. The observed differences in che-
moselectivity underline the importance of matching the reac-
tivity and product selectivity features in the formation of ter-
pene based cyclic carbonates.
to the Lewis acid catalyst and subsequent nucleophile-assisted
ring opening, acyclic substrates favor the insertion of an addi-
tional epoxide molecule rather than CO2 insertion. This behav-
ior can be ascribed to the nature of the Lewis acid catalyst em-
ployed: in particular, AlCl was used with acyclic substrates 9a–
14a. Previous studies have shown that this AlCl complex com-
bined with NBu4I is a highly active binary catalyst for terminal
epoxide/CO2 conversion into cyclic carbonates. Detailed com-
putational and kinetic studies supported the view that CO2 in-
sertion into the initially formed Al–alkoxide bond is the rate-
limiting step.[6d] Thus, the higher Lewis acidity of AlCl relative to
AltBu results in a relatively more stable Al–alkoxide intermediate
with a lower nucleophilic character, thus making the CO2 inser-
tion step slower. This observation seems to corroborate with
the modest chemoselectivities attained in the synthesis of acy-
clic terpene carbonates 9b and 11 b–14b. As a result, substan-
tial oligo/polyether formation is noted.
Conclusions
We have developed the first general and effective route to-
wards bio-based, cyclic carbonates derived from various ter-
pene scaffolds. A binary catalyst comprising a Lewis acidic AlIII
complex and a nucleophilic additive (PPNCl) is able to mediate
these challenging conversions with different degrees of che-
moselectivity; bicyclic terpene oxides were converted with
high selectivity whereas the acyclic substrates showed inferior
chemoselectivity under comparable temperature/pressure con-
ditions. The highly challenging nature of these conversions
limit to some extent the isolated yields of the terpene carbo-
nates (typically between 50–60%), although the results shown
herein offer a reference point for catalyst development to fur-
ther optimize the reactivity/selectivity features of such process-
es. Our future activities will aim to improve these aspects by
carefully designing suitable (aminophenolate) ligand scaffolds
and selecting appropriate Lewis acid metals favoring CO2 inser-
To qualitatively assess the influence of the structure of the
Al-complex on the overall selectivity, we decided to prepare
terpene carbonate 9b (based on citronellyl acetate) using the
AltBu complex under the same conditions as reported in entry 9
Figure 6. 1H NMR spectrum (CDCl3, region 2.0–5.1 ppm) of the crude product obtained in the conversion of terpene oxide 9a and the assigned peaks. The R
group in the schematic figures represents a –CH2CH2CH(Me)CH2CH2OAc side chain. For a full NMR spectrum, see the Supporting Information.
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1308
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
around 10% mol with respect to the epoxide substrate) was
added. A quantitative 1H NMR determination was performed on
samples consisting of 100 mL of this solution diluted with 0.50 mL
of CDCl3. The cyclic carbonate product, 5-(5,5-dimethyl-2-oxo-1,3-
dioxolan-4-yl)-3-methylpentyl acetate 9b, was isolated in 51%
yield by purification using column chromatography (support: SiO2,
eluent: hexane/diethyl ether 1:1 v/v, Rf =0.16). The byproducts 6-
hydroxy-3,7-dimethyloct-7-en-1-yl acetate 9c (Rf =0.26) and 3,7-di-
methyl-6-oxooctyl acetate 9d (Rf =0.18) were also isolated and
fully characterized; the yields were 1 and 5%, respectively.
tion into the metal–alkoxide bond after nucleophilic activation
of the epoxide substrate.
Experimental Section
General information
1H, 13C{1H}, and 2D NMR spectra were recorded on a Bruker AV-
300, AV-400 or AV-500 spectrometer using the residual solvent
peak for calibration. Mass spectrometric analyses and XRD studies
were performed by the Research Support Group at the ICIQ (Tarra-
gona). FTIR measurements were performed on a Bruker Optics FTIR
Alpha spectrometer. Carbon dioxide was purchased from PRAXAIR
and used without further purification. Solvents used in the synthe-
sis of the complexes were dried using an Innovative Technology
PURE SOLV solvent purification system.
Data for 5-(5,5-dimethyl-2-oxo-1,3-dioxolan-4-yl)-3-methylpentyl
acetate (9b): mixture of diastereoisomers with a diastereomeric
ratio (dr) of 51:49: 1H NMR (300 MHz, 298 K, CDCl3): d=4.28–4.17
(m, 1H), 4.17–4.00 (m, 2H), 2.04 (s, 3H), 1.81–1.48 (m, 7H), 1.49
(3H), 1.48–1.38 (m, 1H), 1.37 (s, 3H), 1.32–1.06 (m, 1H), 0.94 ppm
3
(d, J=6.3 Hz, 3H); 13C{1H} NMR (101 MHz, 298 K, CDCl3): d=171.26,
154.19, 85.95, 85.78, 84.10, 62.67, 35.39, 33.30, 33.18, 29.81, 29.79,
26.83, 26.22, 21.34, 21.13, 19.30 ppm; FTIR (neat): 1786, 1733 cmÀ1
(C=O); HRMS (ESI+, CH3OH): m/z: calcd for C13H22O5 +Na+:
281.1359 (M+Na)+; found: 281.1362.
General procedure for preparation of bicyclic terpene carbo-
nates
All reactions were performed in a 30 mL stainless steel reactor. As
a typical experiment, 1-methylcyclohexene oxide (1a, 560.8 mg,
5.0 mmol) was introduced in the inner Teflon vessel (30 mL) of an
autoclave along with AltBu (34.7 mg 0.05 mmol), of PPNCl (86.1 mg,
0.15 mmol), and methyl ethyl ketone (MEK, 1.0 mL). Three cycles of
pressurization and depressurization of the reactor (pCO2 =0.5 MPa)
were performed, and the pressure was finally stabilized at 1.0 MPa.
The reactor was heated to the required temperature (T=858C),
and the mixture was then stirred for another 66 h. The autoclave
was allowed to cool down by immersion in an ice bath for 1 h
before venting. The reaction mixture was diluted in CDCl3 (2.0 mL)
and a weighed quantity of mesitylene (typically around 10 mol%
with respect to the epoxide substrate) was added. A quantitative
1H NMR determination was performed on samples consisting of
100 mL of this solution diluted with 0.5 mL of CDCl3. After concen-
tration of the crude reaction mixture in vacuo, the cyclic carbonate
product 1b (3a-methylhexahydrobenzo[d][1,3]dioxol-2-one) was
isolated upon purification through a short silica path using CH2Cl2
as eluent.
Data for 6-hydroxy-3,7-dimethyloct-7-en-1-yl acetate (9c), mixture
of diastereoisomers with a dr of 1:1. 1H NMR (500 MHz, 298 K,
CDCl3): d=4.93–4.87 (m, 1H), 4.81 (s, 1H), 4.13–4.02 (m, 2H), 4.00
3
(t, J=6.5 Hz, 1H), 2.02 (m, 3H), 1.82 (s, 1H), 1.69 (s, 3H), 1.68–1.18
(m, 6H), 1.13–1.01 (m, 1H), 0.94–0.84 ppm (m, 3H); 13C NMR
(126 MHz, 298 K, CDCl3): d=171.33, 171.32, 147.66, 147.56, 111.29,
111.09, 76.28, 76.12, 63.03, 63.01, 35.55, 35.53, 32.69, 32.65, 32.26,
32.21, 29.90, 29.89, 21.11, 19.53, 17.60, 17.44 ppm; FTIR (neat): n˜ =
1737 cmÀ1 (C=O); HRMS (ESI+, CH3OH): m/z: calcd for C12H22O3 +
Na+ 237.1461 (M+Na)+; found: 237.1457.
Data for 3,7-dimethyl-6-oxooctyl acetate (9d): 1H NMR (300 MHz,
298 K, CDCl3): d=4.09 (m, 2H), 2.61 (m, 1H), 2.46 (m, 2H), 2.04 (s,
3
3
3H), 1.74–1.33 (m, 5H), 1.09 (d, J=6.9 Hz, 6H), 0.91 ppm (d, J=
6.2 Hz, 3H); 13C{1H} NMR (101 MHz, 298 K, CDCl3): d=215.2, 171.6,
63.2, 41.3, 38.2, 35.8, 31.0, 30.0, 21.5, 19.7, 18.8, 18.7 ppm; FTIR
(neat): n˜ =1737, 1710 cmÀ1 (C=O); HRMS (ESI+, CH3OH): m/z: calcd
for C12H22O3 +Na+: 237.1461 (M+Na)+; found: 237.1459.
Data for 3a-methylhexahydrobenzo[d][1,3]dioxol-2-one (1b):
Crystallographic Studies
3
1H NMR (500 MHz, 298 K, CDCl3): d=4.35 (t, J=4.1 Hz, 1H), 2.09–
The measured crystals were stable under atmospheric conditions;
nevertheless, they were treated under inert conditions and im-
mersed in perfluoropolyether as protecting oil for manipulation.
Data Collection: measurements were made on a Bruker-Nonius dif-
fractometer equipped with an APPEX II 4 K CCD area detector,
a FR591 rotating anode with MoKa radiation, Montel mirrors and
a Kryoflex low temperature device (T=À1738C). Full-sphere data
collection was used with w and f scans. Programs used: Data col-
lection Apex2V2011.3 (Bruker-Nonius 2008), data reduction Saint+
Version 7.60 A (Bruker AXS 2008) and absorption correction
SADABS V. 2008–1 (2008). Structure Solution: SHELXTL Version 6.10
(Sheldrick, 2000) was used.[25] Structure Refinement: SHELXTL-97-
UNIX VERSION. CCDC 1451771, 1451772, 1451773, 1451774, and
1451775 contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre.
2.03 (m, 1H), 1.85–1.80 (m, 2H), 1.76–1.65 (m, 2H), 1.56–1.49 (m,
2H), 1.48 (s, 3H), 1.38–1.27 ppm (m, 1H); 13C{1H} NMR (126 MHz,
298 K, CDCl3): d=154.95, 82.92, 81.19, 33.80, 26.01, 23.42, 20.52,
18.80 ppm; FTIR (neat): n˜ =1771 cmÀ1 (C=O); HRMS (ESI+, CH3OH):
m/z: calcd for C8H12O3 +Na+: 179.0679 (M+Na)+; found: 179.0678.
General procedure for preparation of acyclic terpene carbo-
nates
All reactions were performed in a 30 mL stainless-steel reactor. In
a
typical experiment, 5-(3,3-dimethyloxiran-2-yl)-3-methylpentyl
acetate (9a, 1.07 g, 5.0 mmol) was introduced in the inner Teflon
vessel of a 30 mL autoclave along with AlCl (34.8 mg, 0.05 mmol),
PPNCl (86.1 mg, 0.15 mmol), and MEK (1.0 mL). Three cycles of
pressurization and depressurization of the reactor (pCO2 =0.5 MPa)
were performed, after which the CO2 pressure was finally stabilized
at 4.0 MPa. The reactor was heated to the required temperature
(T=858C), and the mixture was stirred for a further 66 h. The auto-
clave was allowed to cool completely by immersion in an ice bath
for 1 h before venting. The reaction mixture was diluted with
2.0 mL of CDCl3 and a weighed quantity of mesitylene (typically
Crystallographic Data for 1b
C8H12O3, Mr =156.18, monoclinic, P2(1), a=7.6622(8) Š, b=
6.4129(6) Š, c=7.7635(8) Š, a=908, b=98.300(3)8, g=908, V=
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1309
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
377.48(8) Š3, Z=2, 1=1.374 mgMÀ3
,
m=0.104 mmÀ1
,
l=
Acknowledgements
0.71073 Š, T=100(2) K, F(000)=168, crystal size=0.40”0.20”
0.20 mm, q(min)=2.6518, q(max)=32.208, 3344 reflections collect-
ed, 2037 reflections unique (Rint =0.0282), GoF=1.063, R1 =0.0345
and wR2 =0.0868 [I>2s(I)], R1 =0.0358 and wR2 =0.0880 (all indi-
ces), min/max residual density=À0.319/0.379 [e·ŠÀ3]. Complete-
ness to q(32.208)=90.4%. CCDC 1451773.
We thank ICIQ, ICREA, and the Spanish Ministerio de Economía y
Competitividad (MINECO) through project CTQ-2014-60419-R and
the Severo Ochoa Excellence Accreditation 2014–2018 through
project SEV-2013-0319. Dr. Noemí Cabello, Sofía Arnal and Vanes-
sa Martínez are acknowledged for the mass analyses. G.F. kindly
acknowledges financial support from the European Community
through FP7-PEOPLE-2013-IEF project RENOVACARB (grant agree-
ment no. 622587), C.M. is grateful to the Marie Curie COFUND
action from the European Commission for co-financing a postdoc-
toral fellowship and M.S. thanks the Erasmus program of the EU
for support.
Crystallographic Data for 2b
C11H16O3, Mr =196.24, monoclinic, P2(1), a=6.048(4) Š, b=
15.614(9) Š, c=6.338(5) Š, a=908, b=117.776(16)8, g=908, V=
529.5(6) Š3, Z=2, 1=1.231 mgMÀ3, m=0.088 mmÀ1, l=0.71073 Š,
T=100(2) K, F(000)=212, crystal size=0.20”0.20”0.20 mm,
q(min)=2.618, q(max)=25.898, 3756 reflections collected, 1719 re-
flections unique (Rint =0.1029), GoF=1.014, R1 =0.0873 and wR2 =
0.2090 [I>2s(I)], R1 =0.1027 and wR2 =0.2162 (all indices), min/
max residual density=À0.379/0.540 [e·ŠÀ3], flack parameter x=
0.7(10). Completeness to q(25.898)=98.8%. CCDC 1451771.
Keywords: aluminium · biocarbonates · carbon dioxide ·
homogeneous catalysis · terpenes
[1] a) M. Aresta, A. Dibenedetto, A. Angelini, Chem. Rev. 2014, 114, 1709–
1742; b) M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann, F. E. Kühn,
Angew. Chem. Int. Ed. 2011, 50, 8510–8537; Angew. Chem. 2011, 123,
8662–8690; c) Q. Liu, L. Wu, R. Jackstell, M. Beller, Nat. Commun. 2015,
6, 5933; d) N. Kielland, C. J. Whiteoak, A. W. Kleij, Adv. Synth. Catal. 2013,
355, 2115–2138; e) T. Sakakura, K. Kohno, Chem. Commun. 2009, 1312–
1330; f) M. Peters, B. Kçhler, W. Kuckshinrichs, W. Leitner, P. Markewitz,
T. E. Müller, ChemSusChem 2011, 4, 1216–1240; g) D. R. Darensbourg,
Chem. Rev. 2007, 107, 2388–2410.
[2] a) D. J. Darensbourg, Inorg. Chem. 2010, 49, 10765–10780; b) S. Klaus,
M. W. Lehenmeier, C. E. Anderson, B. Rieger, Coord. Chem. Rev. 2011,
255, 1460–1479; c) M. R. Kember, A. Buchard, C. K. Williams, Chem.
Commun. 2011, 47, 141–163; d) X.-B. Lu, D. J. Darensbourg, Chem. Soc.
Rev. 2012, 41, 1462–1484; e) G. W. Coates, D. R. Moore, Angew. Chem.
Int. Ed. 2004, 43, 6618–6639; Angew. Chem. 2004, 116, 6784–6806.
[3] a) G. A. Olah, A. Goeppert, G. K. S. Prakash, J. Org. Chem. 2008, 74, 487–
498; b) A. Bansode, A. Urakawa, J. Catal. 2014, 309, 66–70; c) W. Wang,
S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 2011, 40, 3703–3727; d) P. G.
Jessop, T. Ikariya, R. Noyori, Chem. Rev. 1995, 95, 259–270.
Crystallographic Data for 5b
C11H14O4, Mr =210.22, orthorhombic, P2(1)2(1)2(1), a=6.8787(9) Š,
b=7.5688(9) Š, c=20.495(3) Š, a=908, b=908, g=908, V=
1067.1(2) Š3, Z=4, 1=1.309 mgMÀ3
,
m=0.099 mmÀ1
,
l=
0.71073 Š, T=100(2) K, F(000)=448, crystal size=0.40”0.20”
0.20 mm, q(min)=1.998, q(max)=30.608, 11222 reflections collect-
ed, 3225 reflections unique (Rint =0.0271), GoF=1.078, R1 =0.0370
and wR2 =0.0928 [I>2s(I)], R1 =0.0418 and wR2 =0.1018 (all indi-
ces), min/max residual density=À0.204/0.314 [e·ŠÀ3], flack param-
eter x=À0.1(4). Completeness to q(30.608)=98.7%. CCDC
1451774.
[4] For a selection: a) A. Tlili, X. Frogneux, E. Blondiaux, T. Cantat, Angew.
Chem. Int. Ed. 2014, 53, 2543–2545; Angew. Chem. 2014, 126, 2577–
2579; b) T. G. Ostapowicz, M. Schmitz, M. Krystof, J. Klankermayer, W.
Leitner, Angew. Chem. Int. Ed. 2013, 52, 12119–12123; Angew. Chem.
2013, 125, 12341–12345; c) T. Mita, J. Chen, Y. Sato, Org. Lett. 2014, 16,
2200–2203; d) B. Yu, L.-N. He, ChemSusChem 2015, 8, 52–62; e) Y. Li, T.
Yan, K. Junge, M. Beller, Angew. Chem. Int. Ed. 2014, 53, 10476–10480;
Angew. Chem. 2014, 126, 10644–10648; f) D. Yu, S. P. Teong, Y. Zhang,
Coord. Chem. Rev. 2015, 293–294, 279–291.
[5] For recent reviews: a) J. W. Comerford, I. D. V. Ingram, M. North, X. Wu,
Green Chem. 2015, 17, 1966–1987; b) C. Martín, G. Fiorani, A. W. Kleij,
ACS Catal. 2015, 5, 1353–1370; c) G. Fiorani, W. Guo, A. W. Kleij, Green
Chem. 2015, 17, 1375–1389; d) P. P. Pescarmona, M. Taherimehr, Catal.
Sci. Technol. 2012, 2, 2169–2187.
[6] High reactivity systems: a) Y. Qin, H. Guo, X. Sheng, X. Wang, F. Wang,
Green Chem. 2015, 17, 2853–2858; b) T. Ema, Y. Miyazaki, S. Koyama, Y.
Yano, T. Sakai, Chem. Commun. 2012, 48, 4489–4491; c) C. J. Whiteoak,
N. Kielland, V. Laserna, E. C. Escudero-Adµn, E. Martin, A. W. Kleij, J. Am.
Chem. Soc. 2013, 135, 1228–1231; d) C. J. Whiteoak, N. Kielland, V. La-
serna, F. Castro-Gómez, E. Martin, E. C. Escudero-Adµn, C. Bo, A. W. Kleij,
Chem. Eur. J. 2014, 20, 2264–2275; e) J. Qin, P. Wang, Q. Li, Y. Zhang, D.
Yuan, Y. Yao, Chem. Commun. 2014, 50, 10952–10955.
[7] a) M. Cokoja, M. E. Wilhelm, M. H. Anthofer, W. A. Herrmann, F. E. Kühn,
ChemSusChem 2015, 8, 2436–2454; b) Q. He, J. W. O’Brien, K. A. Kitsel-
man, L. E. Tompkins, G. C. T. Curtis, F. M. Kerton, Catal. Sci. Technol. 2014,
4, 1513–1528; c) A. Tlili, E. Blondiaux, X. Frogneux, T. Cantat, Green
Chem. 2015, 17, 157–168.
Crystallographic Data for 6b
C11H16O4, Mr =212.24, triclinic, P-1, a=6.4255(11) Š, b=
9.2997(16) Š, c=9.3869(15) Š, a=103.713(5)8, b=94.235(5)8, g=
103.972(5)8, V=523.67(15) Š3, Z=2, 1=1.346 mgMÀ3
,
m=
0.102 mmÀ1, l=0.71073 Š, T=100(2) K, F(000)=228, crystal size=
0.06”0.04”0.005 mm, q(min)=2.268, q(max)=25.448, 11714 re-
flections collected, 1810 reflections unique (Rint =0.0813), GoF=
2.226, R1 =0.1174 and wR2 =0.3344 [I>2s(I)], R1 =0.1607 and
wR2 =0.3434 (all indices), min/max residual density=À0.475/0.421
[e·ŠÀ3]. Completeness to q(25.448)=94.5%. CCDC 1451772. The
very small size of the crystal resulted in poor quality crystal data;
despite this, the atom connectivity patterns as proposed in
Figure 4 could be confirmed.
Crystallographic Data for 7b
C12H16O6, Mr =256.25, orthorhombic, P2(1)2(1)2(1), a=8.3285(17) Š,
b=11.495(2) Š, c=12.522(3) Š, a=908, b=908, g=908, V=
1198.8(4) Š3, Z=4, 1=1.420 mgMÀ3, m=0.114 mmÀ1, l=0.71073 Š,
T=100(2) K, F(000)=544, crystal size=0.30”0.10”0.10 mm,
q(min)=2.418, q(max)=24.708, 4780 reflections collected, 2006 re-
flections unique (Rint =0.0751), GoF=1.030, R1 =0.0625 and wR2 =
0.1430 [I>2s(I)], R1 =0.0922 and wR2 =0.1624 (all indices), min/
max residual density=À0.245/0.316 [e·ŠÀ3]. Completeness to
q(24.708)=99.2%. CCDC 1451775.
[8] Efficient recycling under mild reaction conditions: a) C. J. Whiteoak,
A. H. Henseler, C. Ayats, A. W. Kleij, M. A. Pericàs, Green Chem. 2014, 16,
1552–1559; b) W. Desens, C. Kohrt, M. Frank, T. Werner, ChemSusChem
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1310
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
2015, 8, 3815–3822; c) M. North, B. Wang, C. Young, Energy Environ. Sci.
2011, 4, 4163–4170; d) S. Liang, H. Liu, T. Jiang, J. Song, G. Yang, B.
Han, Chem. Commun. 2011, 47, 2131–2133.
MEK was used as solvent since the acyclic terpene oxides are viscous
oils and a relatively high loading of AlCl and the nucleophilic additive
are required. The neat reaction between methyl cyclohexene oxide and
CO2 (Table 1, entry 11; only 1% conversion) is further reason for consid-
ering the use of a solvent in these reactions.
[9] Successful systems reporting ample substrate scope: a) V. Laserna, G.
Fiorani, C. J. Whiteoak, E. Martin, E. Escudero-Adµn, A. W. Kleij, Angew.
Chem. Int. Ed. 2014, 53, 10416–10419; Angew. Chem. 2014, 126, 10584–
10587; b) C. J. Whiteoak, E. Martin, E. Escudero-Adµn, A. W. Kleij, Adv.
Synth. Catal. 2013, 355, 2233–2239; c) C. Beattie, M. North, P. Villuendas,
C. Young, J. Org. Chem. 2013, 78, 419–426; d) C. J. Whiteoak, E. Martin,
M. Martínez Belmonte, J. Benet-Buchholz, A. W. Kleij, Adv. Synth. Catal.
2012, 354, 469–476; for a seminal contribution: e) W. J. Kruper, D. V.
Dellar, J. Org. Chem. 1995, 60, 725–727.
[10] As far as we know, the only examples reporting the formation of limo-
nene-based carbonates, see: a) C. M. Byrne, S. D. Allen, E. B. Lobkovsky,
G. W. Coates, J. Am. Chem. Soc. 2004, 126, 11404–11405; b) O. Hauen-
stein, M. Reiter, S. Agarwal, B. Rieger, A. Greiner, Green Chem. 2016, 18,
760–770; c) M. Bähr, A. Bitto, R. Mülhaupt, Green Chem. 2012, 14,
1447–1454.
[11] H. Zhang, H. B. Liu, J. M. Yue, Chem. Rev. 2014, 114, 883–898.
[12] For examples including either terpene or fatty acid based (poly)-
carbonates: a) L. PeÇa Carrodeguas, J. Gonzµlez-Fabra, F. Castro-Gómez,
C. Bo, A. W. Kleij, Chem. Eur. J. 2015, 21, 6115–6122; for examples using
renewable fatty-acid-based epoxides, see: b) M. Alves, B. Grignard, S.
Gennen, C. Detrembleur, C. Jerome, T. Tassaing, RSC Adv. 2015, 5,
53629–53636; c) J. Langanke, L. Greiner, W. Leitner, Green Chem. 2013,
15, 1173–1182; d) P. G. Parzuchowski, M. Jurczyk-Kowalska, J. Ryszkow-
ska, G. Rokicki, J. Appl. Polym. Sci. 2006, 102, 2904–2914; e) M. Bähr, R.
Mülhaupt, Green Chem. 2012, 14, 483–489; f) Y.-Y. Zhang, X.-H. Zhang,
R.-J. Wei, B.-Y. Du, Z.-Q. Fan, G.-R. Qi, RSC Adv. 2014, 4, 36183–36188.
[13] a) R. Ciriminna, M. Lomeli-Rodriguez, P. Demma Carà, J. A. Lopez-San-
chez, M. Pagliaro, Chem. Commun. 2014, 50, 15288–15296; b) N. J.
van Zee, G. W. Coates, Angew. Chem. Int. Ed. 2015, 54, 2665–2668;
Angew. Chem. 2015, 127, 2703–2706.
[14] Recently, Schäffner et al. reported the synthesis of bio-sourced cyclic
carbonates and their use as alternative plasticizers, see: B. Schäffner, M.
Blug, D. Kruse, M. Polyakov, A. Kçckritz, A. Martin, P. Rajagopalan, U.
Bentrup, A. Brückner, S. Jung, D. Agar, B. Rüngeler, A. Pfennig, K. Müller,
W. Arlt, B. Woldt, M. Graß, S. Buchholz, ChemSusChem 2014, 7, 1133–
1139.
[15] Higher reaction temperatures tended to create different phases in the
reactor, in which an ideal mixing of the substrate, CO2 and catalyst
components was not possible leading to lower conversion levels.
[16] A larger scale reaction using 10 g (65.8 mmol) of cis/trans limonene
oxide and CO2 (1.0 MPa) in a 45 mL stainless steel autoclave at 858C for
66 h was also performed. Somewhat lower conversion (35%) and isolat-
ed yield (3.7 g, 29%) was obtained compared to the formation of 4b in
Figure 3 (43% isolated yield). We ascribe this to a lower CO2-to-sub-
strate ratio inside the reactor; however, the result suggests that larger-
scale synthesis of 4b is potentially feasible. See Supporting Information
for more details.
[18] In some of the reactions we noted an imperfect mass balance by com-
bining the molar amounts of cyclic carbonate, starting epoxide and by-
products, as reported in Table 2. We observed that upon cooling the re-
action mixture that under certain conditions up to 10–15% of epoxide
remained in other sections of the reactor, that is, not in the Teflon
insert containing the reaction mixture. In a separate experiment we de-
termined this amount by washing these sections following 1H NMR
analysis. This showed unambiguously the presence of the epoxide sub-
strate.
[19] The formation of oligo/polyether side products is rationalized by the
sluggish nature of these conversions as the viscosity of the reaction
mixture increases. As such, the transport of CO2 towards the reactive
center is hampered and becomes a limiting factor, and side reactions
such as oligo/polyether formation become more significant.
[20] For allylic alcohol formation from epoxides using strong bases: a) R. P.
Thummel, B. Rickborn, J. Org. Chem. 1971, 36, 1365–1368; b) M. J. Sç-
dergren, S. K. Bertilsson, P. G. Andersson, J. Am. Chem. Soc. 2000, 122,
6610–6618. The triphenolate ligand of the Al complex may act as
a basic residue promoting relatively slow allylic alcohol formation.
[21] For Lewis acid catalyzed Meinwald rearrangement of epoxides see:
a) J. R. Lamb, Y. Jung, G. W. Coates, Org. Chem. Front. 2015, 2, 346–349;
A recent contribution showed the use of Al-catalysis as a means to rear-
range internal epoxides to ketones, see: b) J. R. Lamb, M. Mulzer, A. M.
LaPointe, G. W. Coates, J. Am. Chem. Soc. 2015, 137, 15049–15054.
[22] a) M. L. Senyek, Isoprene Polymers in Encyclopedia of Polymer Science
and Technology, Wiley, 2002. Please note that Lewis acid catalyzed
Diels–Alder chemistry of myrcene has also been reported, refer to:
b) D. Yin, C. Li, B. Li, L. Tao, D. Yin, Adv. Synth. Catal. 2005, 347, 137–
142.
[23] For a detailed mechanistic description of the key steps (epoxide ring-
opening, CO2 insertion and ring-closure) involved in organic carbonate
formation using binary catalysts comprising a Lewis acid and a nucleo-
philic additive: a) F. Castro-Gómez, G. Salassa, A. W. Kleij, C. Bo, Chem.
Eur. J. 2013, 19, 6289–6298; b) M. North, R. Pasquale, Angew. Chem. Int.
Ed. 2009, 48, 2946–2948; Angew. Chem. 2009, 121, 2990–2992; Very re-
cently, the formation of in situ prepared organic carbonates was mech-
anistically linked to diastereoselective conversions with a key role of
the Lewis acid catalyst, see: c) W. Guo, V. Laserna, E. Martin, E. C. Escu-
dero-Adµn, A. W. Kleij, Chem. Eur. J. 2016, 22, 1722–1727.
[24] J. Rintjema, W. Guo, E. Martin, E. C. Escudero-Adµn, A. W. Kleij, Chem.
Eur. J. 2015, 21, 10754–10762.
[25] G. M. Sheldrick, SHELXTL Crystallographic System, version 6.10; Bruker
AXS, Inc.: Madison, WI, 2000.
[17] For the acyclic terpene oxide substrates we found that the use of the
AlCl complex provided the best combination of conversion and selectiv-
ity and therefore we proceeded with this complex. Note that again
Received: February 22, 2016
Published online on May 9, 2016
ChemSusChem 2016, 9, 1304 – 1311
www.chemsuschem.org
1311
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim