Highly active aluminium catalysts for the formation of organic carbonates from CO2 and Oxiranes

Article abstract of DOI:10.1002/chem.201302536

Al^III complexes of amino-tris(phenolate) ligand scaffolds have been prepared to attain highly Lewis acidic catalysts. Combination of the aforementioned systems with ammonium halides provides highly active catalysts for the synthesis of organic carbonates through addition of carbon dioxide to oxiranes with initial turnover frequencies among the highest reported to date within the context of cyclic carbonate formation. Density functional theory (DFT) studies combined with kinetic data provides a rational for the relative high activity found for these Al^III complexes, and the data are consistent with a monometallic mechanism. The activity and versatility of these Al^III complexes has also been evaluated against some state-of-the-art catalysts and the combined results compare favorably in terms of catalyst construction, stability, activity, and applicability. Copyright

Full text of DOI:10.1002/chem.201302536

DOI: 10.1002/chem.201302536
Full Paper
&
Lewis Acids
Highly Active Aluminium Catalysts for the Formation of Organic
Carbonates from CO2 and Oxiranes
Christopher J. Whiteoak,[a] Nicola Kielland,[a] Victor Laserna,[a] Fernando Castro-Gꢀmez,[a]
Eddy Martin,[a] Eduardo C. Escudero-Adꢁn,[a] Carles Bo,[a, c] and Arjan W. Kleij*[a, b]
Abstract: AlIII complexes of amino-tris(phenolate) ligand
scaffolds have been prepared to attain highly Lewis acidic
catalysts. Combination of the aforementioned systems with
ammonium halides provides highly active catalysts for the
synthesis of organic carbonates through addition of carbon
dioxide to oxiranes with initial turnover frequencies among
the highest reported to date within the context of cyclic car-
bonate formation. Density functional theory (DFT) studies
combined with kinetic data provides a rational for the rela-
tive high activity found for these AlIII complexes, and the
data are consistent with a monometallic mechanism. The ac-
tivity and versatility of these AlIII complexes has also been
evaluated against some state-of-the-art catalysts and the
combined results compare favorably in terms of catalyst
construction, stability, activity, and applicability.
tion[7] of CO2 to epoxides is a well-documented area of re-
search in which a number of different catalytic solutions have
been proposed using ionic liquids,[8] binary,[9] or bifunctional
(metallosalen) complexes,[10] and simple organic motifs such as
“onium salts”.[11] However, only in a limited amount of cases
have these catalytic efforts resulted in systems with (very) high
activities expressed in high turnover frequencies (TOFs) and
high turnover numbers (TONs).
Introduction
The use of carbon dioxide (CO2) as a renewable carbon feed-
stock is of both academic and industrial relevance.[1] However,
there are only a limited number of chemical processes that
have been successfully developed in which CO2 is used as a re-
agent[2] and even less that classify as truly sustainable from the
point view of energy use and carbon recycling.[3] One method
of increasing the sustainability of a chemical process is
through the use of catalysis and therefore use of such may
enable the realization of truly sustainable CO2 conversions.[4]
The largest challenge in CO2 conversion is its relative high ki-
netic and thermodynamic stability; whereas catalysis can be
used to effectively lower the barriers associated with the kinet-
ic stability, the choice for higher-energy substrates can help to
overcome the thermodynamic hurdle. With this latter aspect in
mind, the use of epoxides (oxiranes) has the advantage that
the ring strain is released upon ring opening and its conver-
sion in the presence of CO2 providing either organic poly-[5] or
cyclic carbonates[6] (Scheme 1) is indeed exothermic. This addi-
In addition, when taking into account possible application
of catalysts in an industrial setting[12] long-term stability, recy-
cling features, accessibility, cost-effectiveness, toxicity, and cat-
alyst modularity are features that are of crucial importance.
Recently, Ema, Sakai, and co-workers reported on bifunction-
al metalloporphyrin catalysts (M=Mg, Zn) that were able to
produce cyclic carbonates based on terminal epoxides with
high TOFs of up to 12000 hꢀ1 and TONs up to 103000.[10b] This
work inspired us to design and construct catalyst systems with
similar or even improved potential in this area with a focus on
high reactivity and wide application potential (substrate
scope). We became interested in using amino-tris(phenolate)
scaffolds[9b,13a] as ligand alternatives for the widely used salen
and/or salans,[14] because these provide different molecular
conformations upon complexation of suitable metal ions.[13]
We envisioned that the presence of fewer donor atoms in the
plane of the metal would be beneficial for those cases in
which the steric requirement of the coordinating epoxide sub-
strate would limit catalytic turnover. Because in the majority of
cases binary catalyst systems are employed (i.e., a metal com-
plex combined with a suitable nucleophile), the nucleophile
needed in the ring-opening step of the coordinated epoxide
would have difficulty approaching the substrate and entering
the coordination sphere of the metal. We have recently report-
ed that FeIII-based amino-tris(phenolate) complexes are effi-
cient catalysts for the formation of a wide series of cyclic car-
[a] Dr. C. J. Whiteoak, Dr. N. Kielland, V. Laserna, F. Castro-Gꢀmez,
Dr. E. Martin, E. C. Escudero-Adꢁn, Prof. Dr. C. Bo, Prof. Dr. A. W. Kleij
Institute of Chemical Research of Catalonia (ICIQ)
Av. Pa¨ısos Catalans 16, 43007 Tarragona (Spain)
Fax: (+34)977-920-828
E-mail: akleij@iciq.es
[b] Prof. Dr. A. W. Kleij
Catalan Institute of Research and Advanced Studies (ICREA)
Pg. Lluꢂs Companys 23, 08010 Barcelona (Spain)
[c] Prof. Dr. C. Bo
Departament de Quꢂmica Fꢂsica i Inorgꢁnica
Universitat Rovira i Virgili
Marcel·lꢂ Domingo s/n 43007 Tarragona (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201302536.
Chem. Eur. J. 2014, 20, 2264 – 2275
2264
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Results and Discussion
Synthesis and analysis of com-
plexes
The amino-tris(phenol) ligands
1
and 2 were prepared by
a known Mannich condensation
reaction,[18] whereas 3 was pre-
pared in stepwise manner.[19] The
ligand bearing chloride substitu-
ents (ligand 4) has been report-
ed by Kol and co-workers[20] but
the preparation was achieved by
using a different work-up proce-
dure to that reported; after heat-
ing 2,4-dichlorophenol and hex-
amethyltetramine (HMTA) to-
gether at reflux at 1108C for
2.5 h, an (unexpected) adduct of
HMTA and the desired ligand,
4·HMTA, was isolated (X-ray
structure, see the Supporting In-
formation). The HMTA could be
removed by an acid/base work-
up to yield the desired ligand 4
Scheme 1. Top: addition of CO2 to epoxides yielding cyclic- and polycarbonates and below some reported suc-
cessful catalyst systems (A–C) based on aluminium for this reaction. Note that catalyst systems A and C are binary
systems and are combined with NBu4X (X=Br or I) whereas catalyst B is a one-component, bifunctional system.
bonates under virtually ambient conditions (258C, P(CO2)=
2 bar) and also effectively mediate the conversion of internal
epoxides.[9b] Further to that, computational analyses[15] on the
Zn(salphen)-mediated formation of cyclic carbonates[4c,9e] have
indicated that metallosalen catalyst systems may have limita-
tions concerning the more sterically demanding substrates.
The fact that amino-tris(phenolate) complexes show better cat-
alytic potential is in line with these theoretical studies.
in good purity although low yield. This ligand is difficult to pre-
pare selectively as a result of the poor solubility of both the
corresponding amino-bis(phenol) and the amino-tris(phenol) in
the reaction medium. Upon formation of the amino-tris-
(phenol) ligands, a degradation may occur towards the amino-
bis(phenol) as reported by Solomon and co-workers for ligand
1.[21] In the case of 4 this degradation reaction appears to be
rapid as the poor solubility of the amino-bis(phenol) drives the
equilibrium towards this undesired product as soon as the
amino-tris(phenol) is obtained, thereby limiting the yield of the
desired ligand.
In an attempt to create even more powerful catalyst sys-
tems, we have turned our focus to aluminium systems, be-
cause recent literature has demonstrated that complexes
based on Al can be attractive within the context of organic
carbonate formation (Scheme 1; see systems A–C).[16] Our pre-
liminary results using an [Al{amino-tris(phenolate)}] complex
(Scheme 1, C) demonstrated unprecedented activity (initial
TOFs up to 36000 hꢀ1 and TONs exceeding 118000) and sub-
strate scope;[17] here we will demonstrate that these kind of
complexes based on this earth-abundant metal can be easily
fine-tuned giving rise to catalyst systems with activities among
the highest reported to date. Further theoretical, kinetic, stabil-
ity, and structural studies are also presented that clearly show
a combination of attractive properties for these powerful cata-
lyst systems during CO2 conversion into value-added commod-
ities such as cyclic carbonates. Computational investigations
have shed important insight on reasons for the observation of
high activities of these new Al-based catalysts.
The [Al{amino-tris(phenolate)}] complexes were obtained
through the reaction of the ligand precursors 1–4 with trime-
thylaluminium ([AlMe3]) in THF (Scheme 2). The characteriza-
tion data for complexes 5, 6, and 8 are consistent with mono-
meric complexes, in which a THF ligand occupies the apical
Scheme 2. Synthesis of [Al{(amino-tris(phenolate)}] complexes 5–8 from pre-
cursor ligands 1–4.
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2265
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. Displacement ellipsoid plot of the dimeric structure of 12 together
with a partial numbering scheme. Co-crystallized solvent molecules and ro-
tational disorder are omitted for clarity. The bond lengths/angles around
both Al centers are rather similar. Selected lengths [ꢃ] and angles [8]: Al(1)ꢀ
O(1)=1.740(3), Al(1)ꢀO(2)=1.734(3), Al(1)ꢀO(3)=1.846(2), Al(1)ꢀ
O(8)=1.948(2), Al(1)ꢀN(1)=2.003(3); O(1)-Al(1)-O(2)=124.28(13), O(1)-Al(1)-
O(3)=115.08(12), O(1)-Al(1)-O(8)=91.15(11), O(1)-Al(1)-N(1)=93.50(12), O(2)-
Al(1)-O(8)=91.14(11), O(3)-Al(1)-O(8)=76.99(10), Al(1)-O(8)-Al(2)=39.94(8),
Al(1)-O(8)-Al(2)-O(3)=176.71(18).
Figure 1. Displacement ellipsoid plot of the molecular structure of 8·THF to-
gether with a partial numbering scheme. Co-crystallized solvent molecules
and rotational disorder in the THF ligand is omitted for clarity. Selected
bond lengths [ꢃ] and angles [8]: Al(1)ꢀO(1)=1.7604(14), Al(1)ꢀ
O(2)=1.7527(15), Al(1)ꢀO(3)=1.7689(15), Al(1)ꢀO(4)=1.9216(15), Al(1)ꢀ
N(1)=2.0715(17); O(4)-Al(1)-N(1)=178.11(7), O(1)-Al(1)-O(2)=120.94(7), O(2)-
Al(1)-O(3)=117.02(7), O(1)-Al(1)-O(4)=87.31(7), O(2)-Al(1)-O(4)=89.50(7),
N(1)-Al(1)-O(1)=90.94(7), N(1)-Al(1)-O(3)=91.43(7).
position of trigonal-bipyrimidal coordination geome-
try. Complex 7 is presumed to exist as a dimeric spe-
cies (as is its FeIII analogue)[13] considering its lower
solubility in most solvents. The molecular structure
for 8·THF (Figure 1) was determined by X-ray crystal-
lography from crystals obtained from THF/hexane.
The molecular structures of 5·THF and 6·THF were re-
ported previously.[22]
Further fine-tuning of the amino-tris(phenol) scaf-
fold can also be accomplished through the synthesis
of nonsymmetrical ligands. In this respect, an amino-
bis(phenol) (i.e., a secondary amine) may serve as
a suitable precursor, since simple alkylation of the
amine with an appropriate reagent directly affords
nonsymmetrical systems. This was probed for the
secondary amine 9,[23] which was alkylated with the
Scheme 3. Synthesis of nonsymmetrical [Al{(amino-tris(phenolate)}] complex 12.
commercially available alkyl bromide 10; the desired
nonsymmetrical amino-tris(phenol) ligand 11 was iso-
lated in 68% yield. Metalation, as described for Al
complexes 5–8, was achieved similarly to furnish 12 in 78%
yield. The identity of 12 was first established by H NMR spec-
(THF, pyridine) or substrate (epoxide)[9b,13] providing monomer-
ic species which is a prerequisite for application in catalysis
(see below).
1
troscopy (details in the Supporting Information) and further af-
firmed by using X-ray crystallography (Figure 2). The structure
reported for 12 represents, to our knowledge, the first crystal-
lographically characterized nonsymmetrical amino-tris(pheno-
late) complex (Scheme 3).
Catalysis screening with Al complexes 5–8 and 12
The catalytic potential of the Al complexes was first evaluated
under mild conditions (T=308C, P(CO2)=1.0 MPa=10 bar)
using 1,2-epoxyhexane (13) as a benchmark substrate and
NBu4I as the co-catalyst (Table 1, entries 1–14). In general, at
catalyst loadings of 0.05 mol% low conversions were noted,
whereas increasing this by at least tenfold (0.5–0.7 mol%) re-
Complex 12 crystallized as a dimeric structure comprising
two m2-phenoxo bridges. Such dimerization potential has been
well-documented in the literature and recently we have shown
that dimers of [Fe{amino-tris(phenolate)}] complexes can be
easily disrupted by addition of a suitable, competing ligand
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2266
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
plex 12. It is known that these complexes have a higher ten-
dency to form dimeric structures, which can compete with the
required substrate coordination. Temperature is an important
factor controlling the monomer–dimer equilibrium, with higher
temperatures favoring higher concentrations of the more reac-
tive monomeric species.[13] Therefore, to compare and better
evaluate the catalytic potential of Al complexes 5–8 and 12,
we then focused on higher reaction temperatures to further in-
crease the turnover frequencies (Table 2).
Table 1. Initial catalysis results obtained for the synthesis of the cyclic
carbonate derived from 1,2-epoxyhexane at 308C using NBu4I as co-cata-
lyst.[a,b]
Table 2. Catalysis results obtained for the synthesis of the cyclic carbon-
ate derived from 1,2-epoxyhexane at various temperatures using NBu4I as
the co-catalyst.[a,b]
Entry
Cat.
Cat.
[mol%]
Co-cat.
[mol%]
Conv.
[%][c]
TOF
[hꢀ1
]
[d]
1
2
3
4
5
6
7
8
5
5
5
5
6
6
6
7
7
8
8
8
12
12
–
0.5
0.05
0.5
0.7
0.05
0.5
0.7
0.05
0.5
0.5
0.05
0.5
0.05
0.5
–
0
0
6
0
0.25
2.5
2.5
0.25
2.5
2.5
0.25
2.5
0
0.25
2.5
0.25
2.5
0.25
2.5
2.5
2.5
2.5[e]
60
74
67
50
65
59
20
18
0
100
91
40
27
–
74
94
5
65
83
2
18
0
10
91
4
27
0
0
46
6
50
Entry
Cat.
[mol%]
Co-cat.
[mol%]
T
[8C]
Conv.
[%][c]
TOF
[hꢀ1
]
[d]
1
2
3
4
5
6
7
8
5
5
5
5
5
6
6
6
6
6
7
7
7
7
8
8
8
8
8
8
–
–
–
–
0.05
0.05
0.05
0.05
0.005
0.05
0.05
0.05
0.01
0.005
0.05
0.05
0.05
0.005
0.05
0.05
0.05
0.05
0.01
0.005
–
0.25
0.25
0
50
70
90
90
110
50
70
90
90
110
50
70
90
110
50
70
90
90
90
110
50
70
90
110
41
60
0
410
600
0
9
10
11
12
13
14
15
16
17
18
19
0.25
0.025
0.25
0.25
0.25
0.05
0.025
0.25
0.25
0.25
0.025
0.25
0.25
0
0.25
0.05
0.025
0.25
0.25
0.25
0.025
89
22
20
52
74
23
17
15
58
78
21
41
82
0
96
42
26
2
15
17
22
890
2200
200
520
740
1150
1700
150
580
780
2100
410
820
0
960
2100
2600
–
–
–
9
–
–
–
46
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
D
A
A
0.5
0.5
0.5
50
[a] Reaction conditions: 1.0 g 1,2-epoxyhexane (9.98 mmol), 1.0 MPa CO2
initial pressure, 2 h, 308C. [b] Catalyst concentration calculated on a per-
Al-atom basis and amount related to the substrate. [c] Determined by
1H NMR spectroscopy of the crude reaction mixture; selectivity in all
cases >99%. Conversions were found to be reproducible within ꢁ2% in
four independent runs for selected experiments. [d] Average turnover fre-
quency (TOFhꢀ1) during the first 2 h. [e] By using NBu4Br as co-catalyst.
–
–
–
–
sulted in high conversion levels of up to 94% (Table 1, entry 4).
The Al complexes themselves proved to be inactive (Table 1,
entries 1 and 10). The best result, in terms of initial TOF, was
obtained for Al complex 8 (Table 1, entry 11) which gave an
average TOF of 100 hꢀ1. It should be noted that the co-catalyst
alone (NBu4I) under similar conditions reported for the Al com-
plexes (either at 0.05 or 0.5 mol% catalyst loading) proved to
be ineffective and no conversion into carbonate was detected
(Table 1, entries 15 and 16). Further to that, upon comparison
with two benchmark catalyst systems (North’s bimetallic [Al-
(salen)] A combined with NBu4I (Table 1, entry 18) or NBu4Br
(entry 19) and our previously reported [Zn(salphen)] catalyst D)
we found that these latter systems were less effective and
showed about half the activity (TOFs are 50 and 46 hꢀ1, respec-
tively; Table 1, entries 17 and 19) compared with Al complex 8
at the same metal loading (entry 12; TOF=91 hꢀ1).[24] The
lowest activities (expressed in initial TOFs) were found for the
nonsubstituted Al complex 7 and the nonsymmetrical Al com-
[a] Reaction conditions: 1,2-epoxyhexane (1.0 g, 9.98 mmol), 1.0 MPa CO2
initial pressure, 2 h. [b] Catalyst concentration calculated on a per-Al-
atom basis and amount related to the substrate. [c] Determined by
1H NMR spectroscopy of the crude reaction mixture; selectivity in all
cases >99%. Conversions were found to be reproducible within ꢁ2% in
four independent runs for selected experiments. [d] Average turnover fre-
quency (TOFhꢀ1) during the first 2 h.
For this second stage of the catalysis screening, tempera-
tures up to 1108C were considered and Al complexes 5–8
were utilized because their substitution pattern allows for
a direct comparison of activities within this series. As expected,
at higher temperatures complexes 5–8 display higher reactivity
and an increase in the observed average TOFs for all com-
plexes, with the complexes in the absence of NBu4I being inac-
tive (Table 2, entries 3 and 17; 908C). For instance, for the Al
complex 5 at 508C an increase in the TOF is noted to 410 hꢀ1,
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2267
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
which reaches 2200 hꢀ1 at 1108C at a catalyst loading of only
0.005 mol% (Table 2, entries 1–5). For the other catalysts 6–8
similar features were noted, that is, an increase in the TOF with
the highest reactivity was noted at 1108C. The Al complex 7,
which is likely to form (at least in part) a dimeric structure in
solution shows, as expected, higher reactivity at higher tem-
peratures as a result of a more favorable monomer–dimer
equilibrium.[13] Nonetheless, clearly for Al complex 8 at all tem-
peratures the highest TOFs were observed suggesting that this
system has the highest catalytic potential with a notable TOF
of up to 2600 hꢀ1 (Table 2, entry 20).
Table 3. Optimization studies towards the activity of Al catalyst 8 in the
synthesis of the cyclic carbonates derived from epoxides 13–16 at 908C
using NBu4I as co-catalyst and comparison with reactions using only the
co-catalyst.[a,b]
At increased reaction temperatures there are two crucial fea-
tures to consider: first, at elevated temperatures the efficacy of
the co-catalyst alone (i.e., the ammonium salt)[11] is needed for
comparison.[25] Critical assessment of the conversion data at
110 8C shows that the absence/presence of any Al complex
(Table 2, entries 5, 10, 14, and 20 versus 24) gives rise to
almost the same amount of product formation and the effect
of the presence of the Lewis acidic Al complex is thus minimal.
At 908C however, a significant difference in conversion is still
noted (compare, for instance, Table 2, entries 18 and 23; 96 vs.
17% conversion) in line with the hypothesis that the Al com-
plex under these conditions plays an imperative role. Second,
whereas up to a 908C reaction temperature the selectivity for
the cyclic carbonate is maintained (ꢂ99% as determined by
1H NMR spectroscopy), at 1108C in some cases we observed
trace amounts of side products that probably arise from ther-
mal decomposition of the carbonate target. Therefore, because
of the above-mentioned reasons further optimization/investi-
gations of the TOF were carried out with Al complex 8 at
a maximum temperature of 908C.
Entry
Cat. 8
[mol%]
0.05
–
0.01
0.01[e]
0.0025
–
0.0010
0.0005
0.0005
0.0005
0.0005
0.0005
–
Co-cat.
[mol%]
0.25
0.25
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05[f]
0.05[f]
0.05[g]
0.05[g]
Substrate
Conv.
[%][c]
96
17
42
11
38
13
33
24
27
24
17
29
21
36
19
TOF
[hꢀ1
960
–
]
[d]
1
2
3
4
5
6
7
8
13
13
13
13
13
13
13
13
14
15
16
13
13
13
13
2100
550
7600
–
16500
24000
27000
24000
17000
29000
–
9
10
11
12
13
14
15
0.0005
–
36000
–
[a] Reaction conditions: epoxide substrate (10 mmol), 1.0 MPa CO2 initial
pressure, 2 h, 908C. [b] Catalyst concentration calculated on a per-Al-
atom basis and amount related to the substrate. [c] Determined by
1H NMR spectroscopy of the crude reaction mixture; selectivity in all
cases >99%. Conversions were found to be reproducible within ꢁ2% in
four independent runs for selected experiments. [d] Average turnover fre-
quency (TOFhꢀ1) during the first 2 h. [e] Using the North catalyst system
A from Scheme 1. [f] Using PPN-I as co-catalyst. [g] Using PPN-Br as co-
catalyst.
Optimization of the performance of Al complex 8 and stabil-
ity tests
To assess the catalytic potential of Al complex 8 in the forma-
tion of cyclic carbonates, initial TOFs were then determined at
various concentrations of 8 (Table 3). In each case, a compari-
son was made between the efficiency of the binary catalyst
couple (i.e., Al complex 8 and NBu4I) versus the co-catalyst
alone to ensure that the Al complex shows a significant contri-
bution in the observed activity. At a loading of 0.05 mol%
(Table 3, entry 1, 96%) a TOF of 960 hꢀ1 is attained, whereas
the co-catalyst alone (entry 2) shows much lower conversion
(17%). Lowering the loading of Al complex 8 to 0.01 mol%
(Table 3, entry 3) increases the TOF value to 2100 hꢀ1; it is im-
portant to note here that the actual potential of a catalyst
should be measured at the initial stage of the reaction (i.e., at
low conversion levels) when substrate concentration limitation
on the observed activity is minimal. On decreasing the amount
of 8 while maintaining a co-catalyst amount of 0.05 mol%,
very high initial TOFs of 7600 (Table 3, entry 5; 0.0025 mol% of
8), 16500 (entry 7; 0.0010 mol% of 8), and 24000 hꢀ1 (entry 8;
0.0005 mol% of 8) were observed whereas the co-catalyst
alone provided a significantly lower conversion (entry 6; 13%).
One of the benchmark systems ([Al(salen)] dimer A, Scheme 1)
from the literature was again compared with our best-perform-
ing complex 8. At a total Al loading of 0.01 mol% (entries 3
and 4 in Table 3), catalyst 8 gave an initial TOF of 2100 hꢀ1,
which is four times higher than observed for dimeric [Al(salen)]
complex A (550 hꢀ1). Other substrates (14–16) were then
tested with Al complex 8 to see whether such high initial TOFs
could also be achieved as observed for 1,2-epoxyhexane 13
(Table 3, entries 9–11). With these alternative epoxides 14–16,
initial TOFs amounted to 27000 (14), 24000 (15), and
17000 hꢀ1 (16), respectively, demonstrating the wider applica-
bility of catalyst 8 and a very high activity in general. As a final
part of the optimization of the activity of the binary catalyst
couple (Al complex 8 and a suitable co-catalyst) we then fo-
cused on the use of bis(triphenylphosphoranylidene)ammoni-
um (PPN) salts of bromide and iodide because these have
been shown to give excellent results in the context of (cyclic/
poly) carbonate formation.[26] Whereas the use of Al complex
8/PPN-I (Table 3, entry 12) showed a modest increase in initial
TOF to 29000 hꢀ1 (cf., 24000 hꢀ1 in the case in which NBu4I
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2268
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
was used; entry 8), the use of PPN-I alone (entry 13) showed
slightly lower conversion of the substrate. The combination of
Al complex 8 and PPN-Br, however, showed significantly higher
activity than the co-catalyst alone (Table 3, compare entries 14
and 15) and a remarkably high initial TOF of 36000 hꢀ1 was
achieved.
paring the conversion/yield for three different substrates with
those reported previously.[17]
Interestingly, the results from entries 9–14 in Table 4 show
that neither the presence of air nor water has a dramatic long-
term effect (cf., TON) on the conversion of the epoxide nor the
yield of the isolated carbonate product with typical yields
under these conditions being high (up to 98%). Therefore, by
combining the results from Tables 1–4 it may be concluded
that Al complex 8 is both a highly active as well as robust
system that upon combination with a suitable nucleophile
(NBu4X, PPNꢀX; X=Br, I) gives rise to a binary catalyst couple
with initial activities amongst the highest reported to date.
To establish the robustness of Al complex 8 several further
experiments were conducted. First, at
a
loading of
0.0005 mol% of 8 using the conditions reported in entry 8
(Table 3) the TON was determined after 18 h, being a notable
112000. This shows that the catalyst is stable enough in the
course of time at 908C providing further turnover when com-
pared with the initial stage of the reaction (Table 3, entry 8;
2 h).
Following the average TOF in time (Table 4, entries 1–8) the
TOFhꢀ1 stays stable within the first two hours of the reaction
Mechanistic considerations
As the mechanism for dimeric [Al(salen)] complexes has been
recently investigated by North and co-workers,[9d] we have fur-
ther focused on some of the features of the operative mecha-
nism[27] when using Al complexes based on amino-tris(pheno-
late) scaffolds. Previously, the first step in the [Zn(salphen)]-cat-
alyzed formation of cyclic carbonates was proposed to be the
coordination of the epoxide to the metal center.[9e,28] We were
able to grow crystals of a similar structure based on Al com-
plex 8 and oxetane; its structure is presented in Figure 3.
The structure for 8·oxetane simply serves to demonstrate
a likely starting point in the conversion of any substrate con-
verted by the [Al{amino-tris(phenolate)}] complexes reported
in this work. Kinetic studies (see the Experimental Section for
more details) were performed with the objective to examine
the order in the binary catalyst. These kinetic studies were per-
Table 4. Reactivity in time and stability studies using Al-catalyst 8 in the
synthesis of cyclic carbonates derived from epoxides 13 and 15–17 at
70–908C using NBu4I as co-catalyst and comparison with reactions using
only the co-catalyst.[a,b]
Entry
Cat. 8
[mol%]
Co-cat.
[mol%]
Substrate
t
[h]
Conv.
[%][c]
TOF
[hꢀ1
]
[d]
1
2
3
4
5
6
7
8
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.25
0.25
0.25
0.25
0.25
0.25
13
13
13
13
13
13
13
13
15
15
16
16
17
17
0.5
1
1.5
2
3
4
6
13
18
24
26
30
34
38
99(98)[f]
84(79)[f]
99(92)[f]
99(93)[f]
99(93)[f]
91(86)[f]
24000
26000
24000
24000
17333
15000
13600
12666
–
–
–
–
–
5
6
9
18[e]
18[g]
18[e]
18[g]
18[e]
18[g]
10
11
12
13
14
0.05
–
[a] Reaction conditions: Epoxide substrate (10 mmol), 1.0 MPa CO2 initial
pressure, 2 h, 908C. [b] Catalyst concentration calculated on a per-Al-
atom basis and amount related to the substrate. [c] Determined by
1H NMR spectroscopy of the crude reaction mixture; selectivity in all
cases >99%. Conversions were found to be reproducible within ꢁ2% in
four independent runs for selected experiments. [d] Average turnover fre-
quency (TOFhꢀ1) during the first 2 h. [e] At P(CO2)=1 MPa together with
P(Air)=0.2 MPa at 708C. [f] The yield of the isolated product is in paren-
theses. [g] Additionally added 10 mL (5 mol%) of water, reaction per-
formed at 708C.
Figure 3. Displacement ellipsoid plot of the structure of 8·oxetane together
with a partial numbering scheme. Co-crystallized solvent molecules are
omitted for clarity, and only one of the crystallographically independent
molecules in the unit cell is shown. Selected bond lengths [ꢃ] and angles [8]:
Al(1B)ꢀO(1B)=1.742(2), Al(1B)ꢀO(2B)=1.759(2), Al(1B)ꢀO(3B)=1.768(2),
Al(1B)ꢀO(4B)=1.907(3), Al(1B)ꢀN(1B)=2.073(3); O(1B)-Al(1B)-
O(2B)=124.84(12), O(1B)-Al(1B)-O(3B)=115.79(11), O(2B)-Al(1B)-
O(4B)=84.56(10), O(1B)-Al(1B)-O(4B)=88.67(11), O(1B)-Al(1B)-
N(1B)=92.11(12), O(4B)-Al(1B)-N(1B)=175.76(10).
and amounts to 12666 hꢀ1 after 6 h; this lower value may be
expected when the kinetics of a process follows a non-zero-
order dependency on the substrate(s) concentration. Then, the
stability of the catalyst system was further probed by investi-
gating the turnover under more competitive reaction condi-
tions, that is, in the presence of air or water (Table 4) and com-
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2269
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
e
formed with Al complex 8 in the range 0.063–0.25 mol% (the
range wherein most of the experiments were carried out with
complexes 5–8 and 12) with respect to the substrate 1,2-
epoxyhexane 13 at 308C and an initial CO2 pressure of 1 MPa
(10 bar) with an iodide loading of 0.5 mol%. A reaction tem-
perature of 308C was chosen to isolate the activity of the
binary catalyst system from that of the background conversion
related to the co-catalyst, which occurs at higher tempera-
tures; the iodide nucleophile itself (at 0.5 mol%) does not lead
to any observable conversion at this temperature (separately
checked; 0% conversion). Activity is only observed at 308C
when both the Al complex 8 and the iodide nucleophile are
present; the absence of either component results in a non-ob-
servable formation of product. The results of the kinetic stud-
ies are highlighted in Figure 4. The conversion of the substrate
Rate ¼ kobs ꢃ ½Catꢄ with ½Catꢄ ¼ binary catalyst system
ð3Þ
Both the Al complex 8 as well as the iodide nucleophile are
considered to be equally important for catalytic turnover; they
do not act separately and as a result one can assume that
c=d and e=c+d. Therefore Equation (2) can be simplified in
the form shown in Equation (3). By taking the natural loga-
rithm of Equation (3), Equation (4) is obtained:
lnðRateÞ ¼ lnðkobsÞ þ eln½Catꢄ
ð4Þ
The kinetic data from the double logarithmic plot revealed
that e is close to 2.0 suggesting a (pseudo) second order de-
pendence on the binary catalyst system. As the rate-determin-
ing step likely involves the Al complex 8 as well as the iodide,
this suggests that only one metal center is actively involved in
the mechanistic cycle in contrast with the cycle previously re-
ported for dinuclear [Al(salen)] complexes.[9d] Also, the compu-
tational investigations appear to confirm the requirement of
a single Al complex for effective catalytic turnover as will be
discussed below in further detail.
To obtain further details about the individual reaction steps
and their energy requirements, computational studies were
performed by using DFT-based methods and considering
a monometallic mechanism. We recently reported on the com-
putational analysis of the full reaction mechanism based on
the binary catalyst [Zn(salphen)] D/NBu4I examining various ep-
oxide substrates.[15] A number of the previously examined sub-
strates[17] converted by Al complex 8 required methyl ethyl
ketone (MEK) as co-solvent, whereas the former DFT studies
(BP86) were carried out with the [Zn(salphen)]-based catalyst
by using CH2Cl2 as medium. Therefore, we have now recalcu-
lated the free energies of all transition states (TSs) and inter-
mediates that are involved in the [Zn(salphen)]/NBu4I-catalyzed
reaction using MEK as medium, and have compared these
computational results with those obtained for the binary cata-
lyst system 8/NBu4I using propylene oxide at 258C at the
B3LYP/6-311G**/LANL2DZ level of theory (see Figure 5 includ-
ing schematic representations of the involved intermediates
and transition states).
Figure 4. Double logarithmic plot of the order determination in binary cata-
lyst under pseudo-steady-state conditions using 1,2-epoxyhexane/CO2 as
substrates and different concentrations of the Al complex 8 at 308C.
1
was determined by H NMR spectroscopic analysis after 2 h in
all experiments by analyzing the crude reaction mixture ali-
quots. Interestingly, these kinetic studies appear to reveal
a pseudo-second-order dependence (2.1; Figure 4) on the
(binary) catalyst system. The rate law for this process may be
written as [Equation (1)]:
Considering a similar type of monometallic mechanism for
both types of catalysts, the reaction starts with coordination of
the epoxide (propylene oxide) forming an initial complex IC.
Then, the iodide mediates the ring opening of the coordinated
epoxide at the b position; this step involves the breaking of
the CbꢀO bond and the simultaneous formation of a CbꢀI
bond through TS1 leading to the metal-alkoxide intermediate
Int-1. It can be observed that this process is energetically more
favorable by 21.6 kcalmolꢀ1 for the Al complex 8 compared
with the same step involving the [Zn(salphen)] complex. The
next step is the insertion of the CO2 molecule into the metal–
oxygen bond of intermediate Int-1, leading to the formation
of linear carbonate Int-2 through transition state TS2.
a
b
c
d
ð1Þ
Rate ¼ k ꢃ ½CO2ꢄ ꢃ ½epoxideꢄ ꢃ ½Alꢄ ꢃ ½Iꢄ
in which, k represents the rate constant, and [Al] and [I] the
concentration of complex 8 and the iodide, respectively. As-
suming steady state conditions at the beginning of the reac-
tion (and thus CO2 and epoxide concentrations being more or
less constant) Equation (1) can be rewritten as Equation (2)
and (3)
Unlike for the [Zn(salphen)] catalyst system, for the Al-cata-
lyzed reaction this process involves two possible pathways
(i.e., path a and path b, see Figure 5). Thus, two TS2 structures
could be obtained depending on the relative orientation (cf.,
c
d
a
b
ð2Þ
2270
Rate ¼ kobs ꢃ ½Alꢄ ꢃ ½Iꢄ with kobs ¼ k ꢃ ½CO2ꢄ ꢃ ½epoxideꢄ
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. Free-energy profiles for the ring-expansion addition of CO2 to propylene oxide catalyzed by [Zn(salphen)]/NBu4I (in blue) and Al complex 8/NBu4I (in
red) considering the b pathway (i.e., initial ring opening occurs at the non-substituted carbon center). As a solvent, MEK was considered at 258C. IC stands
for initial complex formation; FC stands for the intermediate complex having the carbonate product coordinated.
axial or equatorial coordination position of O1 in Figure 6) of
the incoming CO2 molecule with respect the metal center as
shown in Figure 6. Interestingly, in these transition states,
a hexacoordinated Al center is observed, which helps to de-
crease the energy requirement of the CO2 insertion step. Al-
though TS2-a and TS2-b are lower in energy than the corre-
sponding transition state for the [Zn(salphen)] system by 11.3
and 20.2 kcalmolꢀ1, respectively, TS2-b would be the most
favorable route due to the lower barrier required to afford
Int2-b. However, since axial-to-equatorial interconversion be-
tween intermediates Int2-a and Int2-b should be fairly easy, it
is proposed that these species can undergo isomerization. In-
termediate Int2-a was found more stable than Int2-b by 7 kcal
molꢀ1, thus suggesting that the reaction proceeds through
path a. Following the reaction coordinate, the subsequent step
is ring closing, which occurs through the non-coordinating
O-atom (O2) of the carbonate fragment (Figure 6) through TS3
giving the final intermediate FC. Herein, the transition states
TS3-a and TS3-b also could afford intermediates FC-a and
FC-b, respectively. Thus, path a still continues to be the most
likely pathway because of the higher stability of TS3-a over
TS3-b by 14.1 kcalmolꢀ1, and also because the lower absolute
value of most other barriers. In addition, path a has an activa-
tion energy of 34.2 kcalmolꢀ1, with the CO2 insertion being the
rate-limiting step; whereas for path b, a higher activation
energy was found (39.3 kcalmolꢀ1), and the rate-limiting step
corresponds to the ring-closing step. Once the coordinated
cyclic carbonate is formed, it is released from the complex al-
lowing for further epoxide turnover. The overall reaction is ex-
ergonic with a release of 2.4 kcalmolꢀ1 in the case of both cat-
alyst systems.
The competing mechanism was also examined by using the
energetic span (dE) model.[29] Computed free-energy spans for
path a and path b are fairly high (dE=34.2 and 39.3 kcalmolꢀ1
,
respectively); these are identical to the activation barrier of
each pathway. IC was found to be the TOF-determining inter-
mediate (TDI) for both catalytic cycles, and TS2-a and TS3-
b correspond to the TOF-determining TS (TDTS) of the alterna-
tive cycles. On the other hand, by mixing both pathways (con-
sidering the most stable intermediates and TS), a smaller ener-
getic span of 25.3 kcalmolꢀ1 was obtained, in which IC is the
TDI. However, the exact position of the TDTS is difficult to dis-
tinguish because of the small difference between TS2-
b (15.4 kcalmolꢀ1) and TS3-a (15.3 kcalmolꢀ1). Consequently,
a similar degree of TOF control of 0.54 for TS2-b and 0.46 for
TS3-a is observed.
The energetic span of the catalytic cycle involving the [Zn-
(salphen)] complex D was found to be higher (dE=35.6 kcal
molꢀ1) compared with the energetic span of the binary Al-con-
taining catalyst system. The latter, together with the fact that
for Al complex 8 all the involved TSs and intermediates are
significantly less energy-demanding, is in line with the observa-
tion that complex 8 is a much more efficient catalyst than [Zn-
(salphen)] D (see also Table 1 for experimental evidence). Of
particular note is the fact that the initial coordination of the
epoxide in the case of the [Zn(salphen)] complex is endother-
mic (+5.6 kcalmolꢀ1), whereas the coordination of propylene
oxide to Al complex 8 is exothermic by 9.9 kcalmolꢀ1. This
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2271
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
complex 8 to form a less energy-demanding hexacoordinated
transition state is a unique feature (cf., DFT studies) and to-
gether with its associated lower energetic span these are fea-
tures that help to explain the higher catalytic potential of this
system versus the catalyst based on [Zn(salphen)] D. The
simple access to Al complex 8 and its high potential as
a strong Lewis acid mediator make this system a highly potent
system for other types of Lewis acid-mediated transformations
such as the formation of polycarbonates under appropriate
conditions. Studies towards this objective are currently ongo-
ing in our laboratory.
Experimental Section
General Procedures
All epoxide substrates and reagents were commercially available
and were used as received. Carbon dioxide was purchased from
PRAXAIR and used without further purification. Solvents used in
the synthesis of the complexes were dried using an Innovative
Technology PURE SOLV solvent purification system. The ligands
1,[18] 2,[18] 3,[19] and 4[20] were prepared as described previously al-
though 4 was isolated by a modified procedure. The synthesis of
[Al{amino-tris(phenolate)}] complexes was carried out using stan-
dard vacuum line and Schlenk or cannular techniques and once
1
synthesized the products were stored in a vial kept in air. H NMR
spectra were recorded on a Bruker AV-400 or AV-500 spectrometer
and referenced to the residual NMR solvent signals. Elemental anal-
ysis was performed by the Unidꢁd de Anꢁlisis Elemental at the Uni-
versidad de Santiago de Compostela. Mass spectrometric analysis
and X-ray diffraction studies were performed by the Research Sup-
port Group at the ICIQ.
Syntheses
Ligand precursor 4: 2,4-dichlorophenol (4.0 g, 24.5 mmol) and
hexamethylene tetramine (HMTA; 1.2 g, 8.5 mmol) were heated at
110 8C for 2.5 h. Cold MeOH (20 mL) was then added to the yellow
solution before being kept at 08C for 30 min. The solution was so-
nicated and the precipitate which subsequently formed was fil-
tered off and dried under vacuum. The white solid was found to
be a co-crystal with the formula 4·HMTA (790 mg, 14%). Crystals
suitable for single-crystal X-ray diffraction studies of this co-crystal-
lized adduct were obtained by storing the MeOH solution at
ꢀ308C for a further 3 days. 1H NMR (400 MHz, 298 K, CDCl3): d=
Figure 6. Optimized structures of the transition states TS1–TS3 for the Al
complex 8/NBu4I catalyzed reaction, together with the most relevant calcu-
lated distances (in ꢃ) and the values of the negative (imaginary) vibrational
frequencies.
result also demonstrates that 8 is a much more powerful Lewis
acid mediator than [Zn(salphen)] complex D.
3
3
7.23 (d, JHH =2.3 Hz, 3H, ArH), 7.01 (d, JHH =2.3 Hz, 3H, ArH), 6.92
(brs, 3H, ArOH), 4.80 (s, 12H, NCH2N), 3.69 ppm (s, 6H, ArCH2N);
13C{1H} NMR (125 MHz, 298 K, CDCl3): d=150.50, 129.03, 128.63,
125.22, 124.19, 121.16 (all aromatic C), 74.34 (NCH2N), 55.59 ppm
(ArCH2N); elemental analysis calcd (%) for C28H29N5O3Cl6: C 47.53, H
3.99, N 10.27; found: C 47.33, H 3.86, N 10.39. The resulting
4·HMTA compound was then suspended in water (20 mL) and
acidified to pH 2 with HCl (37% in water) with stirring before
being basified to pH 11, at which point the compound was fully
dissolved. The solution was neutralized using HCl, during which
the product precipitated. It was then filtered and dried under
vacuum. The resulting yellowish solid was dissolved in a small
amount of CH2Cl2 and hexane added to precipitate the product,
which was finally filtered and dried to yield a white powder
Conclusion
We have reported on a new family of powerful [Al{amino-tris-
(phenolate)}] catalysts for the addition of CO2 to epoxides with
initial activities among the highest reported to date in this
area and turnover numbers exceeding 100000. Furthermore,
the most active catalyst based on Al complex 8 is a highly
robust system as shown by various stability and activity stud-
ies. Kinetic data and DFT calculations appear to support that
a monometallic mechanism is likely operative, with the CO2 in-
sertion step being rate-limiting and the initial coordination of
the epoxide to the Lewis acidic center more favorable by
around 15 kcalmolꢀ1 compared with our previously reported
[Zn(salphen)]/NBu4I binary catalyst system. The potential of Al
1
3
(680 mg, 37%). H NMR (400 MHz, 298 K, CDCl3): d=7.24 (d, JHH
=
2.4 Hz, 3H, ArH), 7.04 (d, 3JHH =2.4 Hz, 3H, ArH), 5.53 (brs, 3H,
ArOH), 3.73 ppm (s, 6H, ArCH2N); 13C{1H} NMR (125 MHz, 298 K,
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2272
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
CDCl3): d=149.83, 129.10, 128.54, 124.84, 124.72, 121.09 (all aro-
matic C), 55.35 ppm (ArCH2N); MS (ESI, MeOH): m/z calcd: 540.9
[M]+; found: 541. elemental analysis calcd (%) for C21H15NO3Cl6: C
46.53, H 2.79, N 2.58; found: C 46.73, H 2.73, N 2.57.
due was purified by column chromatography (EtOAc/hexane,
20:80 v/v) to yield a yellow powder (721 mg, 68%). 1H NMR
3
4
(400 MHz, 298 K, CDCl3): d=8.04 (dd, JHH =8.9 Hz, JHH =2.8 Hz 1H,
4
4
ArH), 7.96 (d, JHH =2.8 Hz, 1H, ArH), 7.24 (d, JHH =2.4 Hz, 2H, ArH),
4
3
Al complex 5[22a]: [AlMe3] (2m in heptane, 1.19 mL, 2.38 mmol) was
slowly added to a solution of ligand precursor 1 (1.0 g, 2.38 mmol)
in THF (20 mL). The solution was stirred at RT for 2 h and then con-
centrated. Hexane was added to the concentrate resulting in pre-
cipitation of the complex, which was isolated by filtration and fur-
ther dried under vacuum to yield a white powder (963 mg, 78%).
6.98 (d, JHH =2.4 Hz, 2H, ArH), 6.86 (d, JHH =8.9 Hz, 1H, ArH), 5.91
(brs, 3H, ArOH), 3.85 (s, 2H, ArCH2N), 3.78 (s, 4H, ArCH2N), 1.41 (s,
18H, ArC(CH3)3), 1.28 ppm (s, 18H, ArC(CH3)3); 13C{1H} NMR
(125 MHz, 298 K, CDCl3): d=162.40, 151.31, 142.59, 140.41, 135.92,
126.45, 125.69, 125.62, 124.20, 123.19, 121.68, 116.73 (all aromatic
C), 57.68, 56.94 (both ArCH2N), 34.60, 34.19 (both ArC(CH3)3), 31.55,
29.88 ppm (both ArC(CH3)3); MS (ESIꢀ, MeOH): m/z calcd: 603.4
[MꢀH]ꢀ; found: 603; elemental analysis calcd (%) for
3
1H NMR (400 MHz, 298 K, CDCl3): d=6.90 (d, JHH =1.4 Hz, 3H, ArH),
3
2
6.63 (d, JHH =1.4 Hz, 3H, ArH), 4.59 (br, 4H, THF), 4.28 (br d, JHH
=
C37H52N2O5·1= H2O: C 72.40, H 8.70, N 4.56; found: C 72.57, H 8.63,
13.5 Hz, 3H, ArCH2N), 2.86 (br d, 2JHH =13.5 Hz, 3H, ArCH2N), 2.22
(s, 9H, ArCH3), 2.21 (s, 9H, ArCH3), 2.18 ppm (br, 4H, THF); 13C{1H}
NMR (125 MHz, 298 K, CDCl3): d=154.29, 131.02, 126.97, 126.89,
125.83, 120.63 (all aromatic C), 71.31 (THF), 58.67 (ArCH2N), 25.60
(THF), 20.40 (ArCH3), 16.98 ppm (ArCH3); MS (MALDI+, dctb): m/z
calcd: 443.2 [MꢀTHF]+; found: 443.
2
N 4.47.
Al complex 12: This compound was prepared in a similar manner
to 5 by treatment of 11 (500 mg, 0.82 mmol) with [AlMe3] (2m in
heptane, 0.41 mL, 0.82 mmol) to yield a white powder (422 mg,
78%). Crystals suitable for single-crystal X-ray diffraction studies
were obtained by layering of a concentrated THF solution of the
complex with hexane. 1H NMR (400 MHz, 298 K, [D5]pyridine): d=
Al complex 6[22b]: This compound was prepared in a similar
manner to 5 by treatment of ligand precursor 2 (1.0 g, 1.49 mmol)
with [AlMe3] (2m in heptane, 0.75 mL, 1.49 mmol) to yield a white
powder (869 mg, 76%). 1H NMR (400 MHz, 298 K, CDCl3): d=7.25
3
4
4
8.27 (dd, JHH =8.9, JHH =2.8 Hz, 1H, ArH), 8.13 (d, JHH =2.8 Hz, 1H,
ArH), 7.51 (d, 4JHH =2.3 Hz, 2H, ArH), 7.12 (brs, 2H, ArH), 7.03 (d,
3JHH =8.9 Hz, 1H, ArH), 4.15 (brs, 3H, ArCH2N), 3.25 (brs, 3H,
ArCH2N), 1.43 (s, 18H, ArC(CH3)3), 1.34 ppm (s, 18H, ArC(CH3)3);
13C{1H} NMR (125 MHz, 298 K, [D5]pyridine): d=165.44, 155.13,
139.77, 139.07, 137.48, 135.53, 125.82, 125.75, 124.30, 123.70,
123.53, 119.59 (all aromatic C), 58.83, 56.90 (both ArCH2N), 34.61,
34.11 (both ArC(CH3)3), 31.71, 29.52 ppm (both ArC(CH3)3); m/z
calcd: 628.35 [M]+; found: 628.4; elemental analysis calcd (%) for
3
3
(d, JHH =2.5 Hz, 3H, ArH), 6.87 (d, JHH =2.5 Hz, 3H, ArH), 4.36–4.27
(m, 7H; overlapping ArCH2N and THF), 2.95 (brs, 3H, ArCH2N), 2.06
(m, 4H, THF), 1.43 (s, 27H, ArC(CH3)3), 1.29 ppm (s, 27H, ArC(CH3)3);
13C{1H} NMR (125 MHz, 298 K, CDCl3): d=154.82, 139.37, 137.11,
123.78, 123.60, 121.43 (all aromatic C), 70.26 (THF), 58.83 (ArCH2N),
34.88 (ArC(CH3)3), 34.03 (ArC(CH3)3), 31.71 (ArC(CH3)3), 29.70
(ArC(CH3)3), 25.42 ppm (THF); MS (MALDI+, dctb): m/z calcd: 695.5
[MꢀTHF]+; found: 695.
C37H49N2O5Al·H2O·1= (C6H14): C 69.64, H 8.47, N 3.91; found: C 69.59,
2
H 8.78, N 3.80.
Al complex 7: This compound was prepared in a similar manner to
5 by treatment of ligand precursor 3 (350 mg, 1.04 mmol) with
[AlMe3] (2m in heptane, 0.51 mL, 1.04 mmol) to yield a white
1
Typical catalysis procedure
powder (286 mg, 76%). H NMR (400 MHz, 298 K, [D5]pyridine): d=
3
4
3
7.37 (ddd, JHH =7.6 and 7.7, JHH =1.5 Hz, 3H, ArH), 7.18 (dd, JHH
=
1,2-epoxyhexane (1.0 g), the aluminium complex (0.005 mol%),
and Bu4NI (0.025 mol%) were charged into a stainless steel auto-
clave (30 mL). The autoclave was then subjected to three cycles of
pressurization and depressurization with CO2 (0.5 MPa, 5 bar),
before final stabilization of the pressure to 1.0 MPa (10 bar). The
autoclave was sealed and heated to the required temperature and
left stirring for 2 h. An aliquot of the resulting mixture was taken
7.6, 4JHH =1.5 Hz, 3H, ArH), 7.06 (dd, 3JHH =7.7, 4JHH =0.9 Hz, 3H,
3
4
ArH), 6.92 (ddd, JHH =7.6 and 7.7, JHH =0.9 Hz, 3H, ArH), 4.49 (brs,
3H, ArCH2N), 3.14 ppm (brs, 3H, ArCH2N); 13C{1H} NMR (125 MHz,
298 K, [D5]pyridine): d=160.21, 130.77, 130.52, 123.50, 120.69,
119.09 (all aromatic C), 59.46 ppm (ArCH2N); MS (MALDI+, dctb):
m/z calcd: 359.11 [M]+; found: 359.4; elemental analysis calcd (%)
for C21H18NO3Al·H2O·1= (C6H14): C 67.74, H 5.94, N 3.51; found: C
4
1
and conversion analyzed by means of H NMR spectroscopy using
67.41, H 5.95, N 3.25.
CDCl3 as the solvent. The identity of the cyclic carbonate product
was confirmed by comparison to previously reported literature
values. For a photo of the reactor please refer to the Supporting
Information.
Al complex 8: This compound was prepared in a similar manner
to 5 by treatment of ligand precursor 4 (500 mg, 0.92 mmol) with
[AlMe3] (2m in heptane, 0.46 mL, 0.92 mmol) to yield a white
powder (350 mg, 60%). Crystals suitable for single-crystal X-ray dif-
fraction studies were obtained by layering of a concentrated THF
solution of the complex with hexane. 1H NMR (400 MHz, 298 K,
Kinetic experiments
3
3
CDCl3): d=7.33 (d, JHH =2.6 Hz, 3H, ArH), 6.90 (d, JHH =2.6 Hz, 3H,
2
ArH), 4.69 (m, 4H, THF), 4.18 (br d, JHH =13.3 Hz, 3H, ArCH2N), 2.95
The AMTEC vessels were charged with NBu4I and connected to the
hardware. A leak test was first performed with 1.5 MPa (15 bar) of
CO2, with the pressure finally reduced to 0.2 MPa (2 bar). The reac-
tors were filled with 0.4 MPa (4 bar) of CO2, and degassed to 2 bar,
and this cycle was repeated three times. A solution of the Al cata-
lyst 8 in 1,2-epoxyhexane 13 (1.20 mL) was injected into the reac-
tion vessel, the pressure was raised to 1.5 MPa (15 bar), and the
temperature was set to 308C. The reactions were stirred for 2 h at
600 rpm. The stirrer was then stopped and degassed. Hereafter, an
aliquot of the resultant reaction mixture was taken and the conver-
sion was determined by means of 1H NMR spectroscopy using
CDCl3 as the solvent. For a photo of the reactor system please
refer to the Supporting Information.
(br d, 2JHH =13.3 Hz, 3H, ArCH2N), 2.23 ppm (m, 4H, THF); 13C{1H}
NMR (125 MHz, 298 K, CDCl3): d=152.46, 129.75, 127.13, 124.68,
123.08, 122.16 (all aromatic C), 72.38 (THF), 58.25 (ArCH2N),
25.48 ppm (THF); MS (MALDI+, dctb): m/z calcd: 564.9 [MꢀTHF]+;
found: 565; elemental analysis calcd (%) for C25H20AlCl6NO4: C
47.05, H 3.16, N 2.19; found: C 46.75, H 3.22, N 2.12.
Ligand precursor 11: NEt3 (123 mL, 1.76 mmol) was added with
stirring to a solution of 6,6’-(azanediyl-bis(methylene))-bis(2,4-di-
tert-butylphenol) (0.80 g, 1.76 mmol) and 2-(bromomethyl)-4-nitro-
phenol (204 mg, 1.76 mmol) in THF (20 mL). The reaction mixture
was stirred at reflux temperature for 18 h and filtered after cooling.
The volatiles were removed from the filtrate and the obtained resi-
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2273
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The structure for 4·HMTA can be found in the Supporting Informa-
tion; CCDC-938871 contains the supplementary crystallographic
data for this structure. This data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif
X-ray diffraction studies
The measured crystals were stable under atmospheric conditions;
nevertheless they were treated under inert conditions immersed in
perfluoropolyether as protecting oil for manipulation. Data Collec-
tion: Measurements were made on a Bruker-Nonius diffractometer
equipped with an APPEX II 4K CCD area detector, a FR591 rotating
anode with MoKa radiation, Montel mirrors and a Kryoflex low tem-
perature device (T=ꢀ1738C). Full-sphere data collection was used
with w and f scans. Programs used: Data collection Apex2V2011.3
(Bruker-Nonius 2008), data reduction Saint+Version 7.60A (Bruker
AXS 2008) and absorption correction SADABS V. 2008–1 (2008).
Structure Solution: SHELXTL Version 6.10 (Sheldrick, 2000) was
used. Structure Refinement: SHELXTL-97-UNIX VERSION.
Computational details
All calculations were performed by using the Gaussian 09 pack-
age[31] with the hybrid B3LYP functional.[32] The standard
6–311G(d,p) basis set was used to describe the H, C, N, and O
atoms. The relativistic effective core pseudo potential LANL2DZ
was used, together with its associated basis set, for Al, Zn, Cl, and I
atoms. Full geometry optimizations were performed without con-
strains. The nature of the stationary points encountered was char-
acterized either as minima or transition states by means of har-
monic vibrational frequencies analysis. The zero-point, thermal,
and entropy corrections were evaluated to compute enthalpies
and Gibbs free energies (T=298 K, p=1 bar). Solvent effects for
MEK have been introduced by using the Polarizable Continuum
Model (PCM).
Crystallographic details for Al-complex 8·THF: Formula:
C25H20Cl6NO4Al; Mw =638.10; crystal size 0.30ꢄ0.15ꢄ0.05 mm3; or-
thorhombic; space group Pbca; a=17.2360(19), b=15.7823(17),
c=19.047(2) ꢃ; a=b=g=90.08; V=5181.2(10)8; Z=8; 1calcd
=
1.636 mgmꢀ13; m(MoKa)=0.733 mmꢀ1; T=100(2) K; q(min/max)=
2.05/26.25; F(000)=2592; 93683 reflections collected; 5209 unique
reflections (Rint =0.0535); absorption correction empirical; refine-
ment method: full-matrix least-squares on F2; data/restraints/pa-
rameters: 5209/86/364; GoF on F2 =1.074; R1 =0.0287 and wR2 =
0.0663 [I>2s(I) ]; R1 =0.0404 and wR2 =0.0728 (all data); largest
diff. peak and hole: 0.333 and ꢀ0.305 e3 ꢃꢀ3. CCDC-902502 con-
tains the supplementary crystallographic data for this structure.
This data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
Acknowledgements
We thank ICIQ, ICREA, the Spanish Ministry of Science and In-
novation (MINECO, projects CTQ-2011–27385 and CTQ2011–
29054-C02–02 and FPI to FCG) and Generalitat de Catalunya
(project 2009-SGR-259) for support. We acknowledge Dr.
Noemꢅ Cabello, Vanessa Martꢅnez, and Sofꢅa Arnal for perform-
ing the MS studies.
Crystallographic details for Al complex 12: Formula:
C95H122N4O10Al2; Mw =1533.93; crystal size 0.15ꢄ0.05ꢄ0.05 mm3;
monoclinic; space group P21/c; a=29.417(3), b=10.9419(12), c=
30.189(3) ꢃ; a=g=90, b=118.896(4); V=8507.2(17)8; Z=4;
1calcd =1.198 mgmꢀ13; m(MoKa)=0.095 mmꢀ1; T=100(2) K; q(min/
max)=0.79/24.80; F(000)=3304; 79528 reflections collected;
14140 unique reflections (Rint =0.1388); absorption correction em-
pirical; refinement method: full-matrix least-squares on F2; data/re-
straints/parameters: 14140/441/1127; GoF on F2 =1.015; R1 =
0.0643 and wR2 =0.1280 [I>2s(I) ]; R1 =0.1523 and wR2 =0.1658
(all data); largest diff. peak and hole: 0.546 and ꢀ0.554 e3 ꢃꢀ3. This
structure is a toluene solvate and has a tert-butyl group disordered.
CCDC-938872 contains the supplementary crystallographic data for
this structure. This data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Keywords: aluminium
· carbon dioxide · carbonates ·
homogeneous catalysis · ligands
[1] a) Carbon Dioxide as Chemical Feedstock (Ed.: M. Aresta), Wiley-VCH,
Weinheim, 2010; b) M. Peters, B. Kçhler, W. Kuckshinrichs, W. Leitner, P.
Markewitz, T. E. Mꢆller, ChemSusChem 2011, 4, 1216–1240; c) M. Cokoja,
C. Bruckmeier, B. Rieger, W. A. Herrmann, F. E. Kꢆhn, Angew. Chem. 2011,
123, 8662–8690; Angew. Chem. Int. Ed. 2011, 50, 8510–8537; d) T. Saka-
kura, J. C. Choi, H. Yasuda, Chem. Rev. 2007, 107, 2365–2387; e) R.
Martꢅn, A. W. Kleij, ChemSusChem 2011, 4, 1259–1263.
[2] See for instance: a) G. Pagani, Hydrocarb. Proc. 1982, 61, 87–91; b) refer
also to www.novomer.com for a company focusing on green polymer
synthesis using CO2 as a feedstock.
[3] a) E. A. Quadrelli, J.-L. Duplan, G. Centi, S. Perathoner, ChemSusChem
2011, 4, 1194–1215; b) G. Centi, G. Iaquaniello, S. Perathoner, ChemSu-
sChem 2011, 4, 1265–1273.
Crystallographic details for Al-complex 8·oxetane: Formula:
C25H19.75NO4.25Cl6.75Al; Mw =668.44; crystal size 0.50ꢄ0.05ꢄ
0.05 mm3; triclinic; space group P-1; a=13.702(4), b=14.048(4),
c=14.784(3) ꢃ; a=94.689(7), b=93.252(7), g=100.374(10)8; V=
2782.5(12)8; Z=4; 1calcd =1.596 mgmꢀ13; m(MoKa)=0.757 mmꢀ1; T=
100(2) K; q(min/max)=1.39/28.23; F(000)=1354; 21763 reflections
collected; 21763 unique reflections (Rint =0.000); absorption cor-
rection empirical; refinement method: full-matrix least-squares on
F2; data/restraints/parameters: 21763/72/758; GoF on F2 =1.021;
R1 =0.0570 and wR2 =0.1528 [I>2s(I) ]; R1 =0.0928 and wR2 =
0.1898 (all data); largest diff. peak and hole: 0.881 and ꢀ0.940
e3 ꢃꢀ3. The disorder in co-crystallized solvent molecules and the co-
ordinated oxetane was done with TWINABS.[30] CCDC-938873 con-
tains the supplementary crystallographic data for this structure. .
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
[4] For some original recent examples: a) C. D. Neves Gomes, O. Jacquet, C.
Villiers, P. Thuꢇry, M. Ephritikhine, T. Cantat, Angew. Chem. 2012, 124,
191–194; Angew. Chem. Int. Ed. 2012, 51, 187–190; b) T. Kimura, K.
Kamata, N. Mizuno, Angew. Chem. 2012, 124, 6804–6807; Angew. Chem.
Int. Ed. 2012, 51, 6700–6703; c) A. Decortes, A. W. Kleij, ChemCatChem
2011, 3, 831–834; d) I. I. F. Boogaerts, S. P. Nolan, J. Am. Chem. Soc.
2010, 132, 8858–8859; e) D. R. Dalton, T. Rovis, Nat. Chem. 2010, 2,
710–711; f) A. Correa, R. Martꢅn, J. Am. Chem. Soc. 2009, 131, 15974–
15975; g) T. Fujihara, K. Nogi, T. Xu, J. Terao, Y. Tsuji, J. Am. Chem. Soc.
2012, 134, 9106–9109; h) M. R. Kember, C. K. Williams, J. Am. Chem. Soc.
2012, 134, 15676–15679; i) S. Wesselbaum, U. Hintermair, W. Leitner,
Angew. Chem. 2012, 124, 8713–8716; Angew. Chem. Int. Ed. 2012, 51,
8585–8588; j) A. Monassier, V. D’Elia, M. Cokoja, J. Pelletier, J.-M. Basset,
F. E. Kꢆhn, ChemCatChem 2013, 5, 1321–1324.
[5] a) D. J. Darensbourg, Inorg. Chem. 2010, 49, 10765–10780; b) D. J. Dare-
nsbourg, Chem. Rev. 2007, 107, 2388–2410; c) M. R. Kember, A. Buchard,
C. K. Williams, Chem. Commun. 2011, 47, 141–163; d) G. W. Coates, D. R.
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2274
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Moore, Angew. Chem. 2004, 116, 6784–6806; Angew. Chem. Int. Ed.
2004, 43, 6618–6639; e) X.-B- Lu, D. J. Darensbourg, Chem. Soc. Rev.
2012, 41, 1462–1484; f) G.-P. Wu, W.-M. Ren, Y. Luo, B. Li, W.-Z. Zhang,
X.-B. Lu, J. Am. Chem. Soc. 2012, 134, 5682–5688.
[21] X. Zhang, D. H. Solomon, Polymer 1998, 39, 405–412.
[22] a) A. L. Johnson, M. G. Davidson, Y. Pꢇrez, M. D. Jones, N. Merle, P. R.
Raithby, S. P. Richards, Dalton Trans. 2009, 5551–5558; b) J. Zhang, A.
Liu, X. Pan, L. Yao, L. Wang, J. Fang, J. Wu, Inorg. Chem. 2011, 50, 9564–
9570. For some catalytic applications of Lewis acidic [Al{amino-tris(phe-
nolates)}] see: c) S. M. Raders, J. G. Verkade, J. Org. Chem. 2009, 74,
5417–5428; d) S. M. Raders, J. G. Verkade, Tetrahedron Lett. 2009, 50,
5317–5321.
[6] a) M. North, R. Pasquale, C. Young, Green Chem. 2010, 12, 1514–1539;
b) A. Decortes, A. M. Castilla, A. W. Kleij, Angew. Chem. 2010, 122,
10016–10032; Angew. Chem. Int. Ed. 2010, 49, 9822–9837; c) P. P. Pes-
carmona, M. Taherimehr, Catal. Sci. Technol. 2012, 2, 2169–2187.
[7] The often used term “cyclo” addition reaction is in our view incorrect as
it formally refers to a pericyclic chemical conversion between one or
more unsaturated molecules. This combination should then give
a cyclic product with a net reduction of the bond multiplicity.
[8] a) J. Sun, S.-I. Fujita, M. Arai, J. Organomet. Chem. 2005, 690, 3490–
3497; b) Y. Zhao, C. Yao, G. Chen, Q. Yuan, Green Chem. 2013, 15, 446–
452; c) Z.-Z. Yang, L.-N. He, C.-X. Miao, S. Chanfreau, Adv. Synth. Catal.
2010, 352, 2233–2240.
[9] a) R. L. Paddock, S. T. Nguyen, Chem. Commun. 2004, 1622; b) C. White-
oak, E. Martin, M. Martꢅnez Belmonte, J. Benet-Buchholz, A. W. Kleij, Adv.
Synth. Catal. 2012, 354, 469–476; c) R. M. Haak, A. Decortes, E. C. Escu-
dero-Adꢁn, M. Martꢅnez Belmonte, E. Martin, J. Benet-Buchholz, A. W.
Kleij, Inorg. Chem. 2011, 50, 7934–7936; d) M. North, R. Pasquale,
Angew. Chem. 2009, 121, 2990–2992; Angew. Chem. Int. Ed. 2009, 48,
2946–2948; e) A. Decortes, M. Martꢅnez Belmonte, J. Benet-Buchholz,
A. W. Kleij, Chem. Commun. 2010, 46, 4580–4582; f) J. Langanke, L.
Greiner, W. Leitner, Green Chem. 2013, 15, 1173–1182.
[10] a) A. Buchard, M. R. Kember, K. G. Sandeman, C. K. Williams, Chem.
Commun. 2011, 47, 212–214; b) T. Ema, Y. Miyazaki, S. Koyama, Y. Yano,
T. Sakai, Chem. Commun. 2012, 48, 4489–4491; c) M. V. Escꢁrcega-Boba-
dilla, M. Martꢅnez Belmonte, E. Martin, E. C. Escudero-Adꢁn, A. W. Kleij,
Chem. Eur. J. 2013, 19, 2641–2648; d) J. Melꢇndez, M. North, P. Villuen-
das, C. Young, Dalton Trans. 2011, 40, 3885–3902; e) M. A. Fuchs, T. A.
Zevaco, E. Ember, O. Walter, I. Held, E. Dinjus, M. Dçring, Dalton Trans.
2013, 42, 5322–5329.
[11] V. Calꢀ, A. Nacci, A. Monopoli, A. Fanizzi, Org. Lett. 2002, 4, 2561–2563.
[12] For the synthesis of cyclic carbonates by using a flue gas-based feed-
stock: M. North, B. Wang, C. Young, Energy Environ. Sci. 2011, 4, 4163–
4170.
[23] I. S. Belostotskaya, N. L. Komissarova, T. I. Prokof’eva, L. N. Kurkovskaya,
V. B. Vol’eva, Russian J. Org. Chem. 2005, 41, 703–706.
[24] Note that a comparison between catalysts at higher conversion levels
does not necessarily imply the best evaluation of the relative catalyst
potential. This should be done at the early stage of the reactions (i.e.,
at low conversions), in which initial TOFs give a better description of
the potential of a catalyst.
[25] C. Beattie, M. North, P. Villuendas, C. Young, J. Org. Chem. 2013, 78,
419–426.
[26] For illustrative examples: a) W.-M. Ren, G.-P. Wu, F. Lin, J.-Y. Jiang, C. Liu,
Y. Luo, X.-B. Lu, Chem. Sci. 2012, 3, 2094–2102; b) A. Berkessel, M. Bran-
denburg, Org. Lett. 2006, 8, 4401–4404; c) D. J. Darensbourg, S. J.
Wilson, J. Am. Chem. Soc. 2011, 133, 18610–18613.
[27] For an early discussion around fundamental mechanistic features in
CO2 chemistry refer to: W. Leitner, Coord. Chem. Rev. 1996, 153, 257–
284.
[28] M. Taherimehr, A. Decortes, S. M. Al-Amsyar, W. Lueangchaichaweng,
C. J. Whiteoak, A. W. Kleij, P. P. Pescarmona, Catal. Sci. Technol. 2012, 2,
2231–2237.
[29] a) S. Kozuch, S. Shaik, Acc. Chem. Res. 2011, 44, 101–110; b) A. Uhe, S.
Kozuch, S. Shaik, Comput. Chem. 2011, 32, 978–985; c) S. Kozuch, WIREs
Comput. Mol. Sci. 2012, 2, 795–815; See also: d) A. Uhe, M. Hçlscher, W.
Leitner, Chem. Eur. J. 2012, 18, 170–177 for the use of the energetic
span model in CO2 chemistry.
[30] TWINABS, version 2008/4, Bruker AXS; R. H. Blessing, Acta Crystallogr.
Sect. A 1995, 51, 33–38.
[31] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢈ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[13] a) C. J. Whiteoak, B. Gjoka, E. Martin, M. Martꢅnez Belmonte, E. C. Escu-
dero-Adꢁn, C. Zonta, G. Licini, A. W. Kleij, Inorg. Chem. 2012, 51, 10639–
10649; b) G. Licini, M. Mba, C. Zonta, Dalton Trans. 2009, 5265–5277.
[14] For some recent reviews on salen chemistry and systems see, for exam-
ple: a) C. J. Whiteoak, G. Salassa, A. W. Kleij, Chem. Soc. Rev. 2012, 41,
622–631; b) A. W. Kleij, Chem. Eur. J. 2008, 14, 10520–10529.
[15] F. Castro-Gꢀmez, G. Salassa, A. W. Kleij, C. Bo, Chem. Eur. J. 2013, 19,
6289–6298.
[16] a) W. Clegg, R. W. Harrington, M. North, R. Pasquale, Chem. Eur. J. 2010,
16, 6828–6843; b) D. Tian, B. Liu, Q. Gan, H. Li, D. J. Darensbourg, ACS
Catal. 2012, 2, 2029–2035.
[17] C. J. Whiteoak, N. Kielland, V. Laserna, E. C. Escudero-Adꢁn, E. Martin,
A. W. Kleij, J. Am. Chem. Soc. 2013, 135, 1228–1231.
[32] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. Lee, R. G. Parr,
W. Yang, Phys. Rev. B 1988, 37, 785–789.
[18] a) A. Chandrasekaran, R. O. Day, R. R. Holmes, J. Am. Chem. Soc. 2000,
122, 1066–1072; b) M. Kol, M. Shamis, I. Goldberg, Z. Goldschmidt, S.
Alfi, E. Hayut-Salant, Inorg. Chem. Commun. 2001, 4, 177–179.
´
[19] L. J. Prins, M. Mba, A. Kolarovic, G. Licini, Tetrahedron Lett. 2006, 47,
Received: July 1, 2013
2735–2738.
Revised: November 15, 2013
Published online on January 20, 2014
[20] S. Groysman, I. Goldberg, M. Kol, E. Genizi, Z. Goldschmidt, Adv. Synth.
Catal. 2005, 347, 409.
Chem. Eur. J. 2014, 20, 2264 – 2275
www.chemeurj.org
2275
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim