A recyclable trinuclear bifunctional catalyst derived from a tetraoxo bis-Zn(salphen) metalloligand

Article abstract of DOI:10.1002/chem.201204132

A tetraoxo bis-Zn(salphen) supramolecular host can bind various divalent metal salts, thereby providing access to trinuclear bifunctional systems that incorporate both Lewis acid sites and dynamically bound nucleophilic anions. The formation of these trinuclear species was investigated and their stability features were also determined. The application of these trinuclear complexes as bifunctional catalysts was evaluated in the formation of cyclic organic carbonates from epoxides and CO₂. The catalytic data, in combination with control experiments, clearly demonstrate that these trinuclear compounds show much higher recycling potential compared to various control compounds and they can be used in up to five cycles without an observable loss in activity. Furthermore, this new recyclable catalytic system does not require any additives and can be applied under solvent-free conditions.

Full text of DOI:10.1002/chem.201204132

FULL PAPER
DOI: 10.1002/chem.201204132
A Recyclable Trinuclear Bifunctional Catalyst Derived from a Tetraoxo Bis-
ZnACTHNUTRGNEUG(N salphen) Metalloligand
Martha V. Escꢀrcega-Bobadilla,[a] Marta Martꢁnez Belmonte,[a] Eddy Martin,[a]
Eduardo C. Escudero-Adꢀn,[a] and Arjan W. Kleij*[a, b]
Abstract: A tetraoxo bis-Zn
(salphen)
lysts was evaluated in the formation of
cyclic organic carbonates from ep-
AHCTUNGTERNNoGUN xides and CO2. The catalytic data, in
combination with control experiments,
clearly demonstrate that these trinu-
clear compounds show much higher re-
cycling potential compared to various
control compounds and they can be
used in up to five cycles without an ob-
servable loss in activity. Furthermore,
this new recyclable catalytic system
does not require any additives and can
be applied under solvent-free condi-
tions.
supramolecular host can bind various
divalent metal salts, thereby providing
access to trinuclear bifunctional sys-
tems that incorporate both Lewis acid
sites and dynamically bound nucleo-
philic anions. The formation of these
trinuclear species was investigated and
their stability features were also deter-
mined. The application of these trinu-
clear complexes as bifunctional cata-
Keywords: bifunctional catalysts ·
carbon dioxide fixation · metallolig-
AHCTUNGTRENNUNG
Introduction
Inspired by the work of Nabeshima and co-workers,[5] we
set out to explore a new design of an atom-economical syn-
thetic route towards recyclable bifunctional supramolecular
catalysts. The Nabeshima group has reported on acyclic
Metalloligands are structures that allow the construction of
more-complex (supramolecular) metal-containing architec-
tures.[1] From an applications point of view, the resultant su-
perstructures can display new properties, including guest-
recognition and catalytic ability.[2] Supramolecular cataly-
sis,[3] one of the fields for the application of these com-
pounds, has taken up a prominent position in chemical sci-
ence with a strong focus on the discovery and development
of new reactivity patterns. Unfortunately, very little atten-
tion has been paid to systems that incorporate features that
allow solvent-free processes and easy recycling of the cata-
lyst ensemble.[4] For their possible implementation in indus-
trial processes, it is crucial that active catalysts can be effi-
ciently recycled to maximize turnover efficiencies and lower
overall costs. Hence, new approaches need to be developed
to address these latter issues, such that supramolecular cata-
lysts may become viable alternatives with respect to co-
oligoACTHNURTGNEU“GN salamo”-based bis-metal complexes that, upon forma-
tion, create a second coordination sphere that is comprised
of multiple O donors, which resemble (in part) the well-
known crown-ether framework. The O-rich cavities can be
exploited for selective and cooperative metal-cation binding
and their use in various applications, such as the creation of
chirality[6] and in selective macrocyclic synthesis by using
templated ring-closing metathesis (RCM) chemistry,[7] have
been reported. Thus, this type of metallohost (Figure 1a)
ACHTUNGTRENNUNGvalent-type homogeneous catalysts.
Figure 1. Schematic representation of metalloACTHNUGTRENNUGliACHTUGTNRENNUGgand systems that have
been used for cation binding, as reported by a) Nabeshima and co-work-
ers and b) others.
[a] Dr. M. V. Escꢀrcega-Bobadilla, Dr. M. Martꢁnez Belmonte,
E. Martin, E. C. Escudero-Adꢀn, Prof. Dr. A. W. Kleij
Institute of Chemical Research of Catalonia (ICIQ)
Av. Paꢂsos Catalans 16, 43007–Tarragona (Spain)
Fax: (+34)977920828
E-mail: akleij@iciq.es
holds great promise for the binding of Lewis acidic metal
cations, thus providing stable and functional trinuclear su-
pramolecular structures.[8]
[b] Prof. Dr. A. W. Kleij
Catalan Institute for Research and Advanced Studies (ICREA)
Pg. Lluis Companys 23, 08010–Barcelona (Spain)
Other trimetallic complexes (Figure 1b) that were derived
from mono-salen complexes have been used as catalysts for
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201204132.
Chem. Eur. J. 2013, 19, 2641 – 2648
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2641

different transformations, such as the halogenation of aro-
Results and Discussion
matic rings mediated by Cu3 bis
(salen) complexes[9] and the
ꢀ
ꢀ
use of Ni2Mn bis-salen complexes as analogues of catechol
oxidase.[10] The principal difference between these latter sys-
tems and those reported by Nabeshima and co-workers is
the fact that the central, third metal cation is bound to two
metallosalen complexes through a bidentate coordination
mode that is comprised of two phenolic O atoms in each re-
spective salen unit. In the Nabeshima system, the third
metal cation is surrounded by a polydentate O-coordinating
As discussed above, mono-salen complexes may afford tri-
nuclear bis-salen complexes (Figure 1b); however, herein,
we considered the use of a metalloACHTUNTRGENNlUG iACHUTNGTRENgUNG and that contained a
polydentate O-coordination pocket, which is crucial for af-
fording trinuclear complexes with increased stability. Previ-
ꢀ
ously, we reported on a Zn3Cl2 bis
N
was derived from compound 1 by treatment with ZnCl2; this
complex was successfully used as an achiral host system for
effective chirality transfer (chirogenesis) with a variety of
mono- and ditopic chiral guests.[14a] By comparison with the
systems reported by Nabeshima and co-workers (for exam-
ple, see Figure 1a), we realized that compound 1 could pro-
vide access to a variety of trinuclear complexes that may be
of use in catalysis by taking advantage of the presence of
both Lewis acid centers and anionic nucleophiles (Scheme 1,
compounds 2–4). The incorporation of these nucleophilic
sites in the same structure is advantageous in cases in which
recycling of the catalyst is desirable, because binary versions
of Lewis acids and halides are often not easily recycled.
metalloACHTUNGTRENNUNGliACHTUNGTRENNUNGgand, which likely affords higher stabilization and a
better recycling potential (Figure 1) and this is the major
focus of this work.
The catalytic formation of organic carbonates through the
cycloaddition of carbon dioxide to oxiranes has become a vi-
brant field in homogeneous catalysis.[11] The most widely
used strategy is the combination of a Lewis acid (LA) with
a nucleophilic co-catalyst (usually a halide), in which the
LA site activates the epoxide for ring opening by the incom-
ing nucleophile. This approach has been successful in, for in-
stance, the case of simple metal salts, as previously demon-
strated by Fujita and co-workers[12] and by Aresta and
Therefore, bis-ZnACTHNUTRGNEUNG(salphen) complex 1 was used as starting
Diben
G
N
point in the synthesis of trinuclear complexes 2–4. Then, the
central Zn center would represent a Lewis acidic center for
epoxide activation, whereas, upon dissociation from the
outer Zn ions, the anions could act as reactive nucleophiles
for epoxide ring-opening. Previous work from our group
demonstrated the dynamic behavior of anion coordination
binary catalysts, respectively. Herein, we report our efforts
to combine these types of simple catalysts with the advanta-
geous chelating properties of metalloligands to create new,
stable supramolecular catalysts that have excellent recycling
potential. As a starting point, we used bis-Zn
ACHTUNGTREN(NGNU salphen) struc-
ture 1 (Scheme 1)[14a] and the incorporation of various metal
to ZnACHTNUGRTNEUNG
(salphen) complexes[16] and, thus, the potential of the
anions in complexes 2–4 to act as nucleophilic co-catalysts.
Reversible, supramolecular interactions between halide
anions and the Zn ions is a prerequisite for catalytic activity,
because the halide is involved in nucleophilic ring-opening
ꢀ
of the epoxide substrate and Zn X (X=halide) dissociation
also allows for coordination of the epoxide to the Lewis
acidic Zn center. After the reaction, the halide anion re-co-
ordinates to the Zn center and the bifunctional system can
be recycled if stable enough.
Trinuclear complexes 2–4 were readily synthesized in
almost-quantitative yield by the addition of MX2 (M=Zn,
Mg; X=I, OAc) salts to metalloACTHNURTGENNlGU iACHUTGTNRENgUNG and complex 1 in THF
(Scheme 1 and the Experimental Section). For complexes 2–
4, X-ray diffraction confirmed the proposed stoichiometries
and connectivity patterns; only the structure for complex 2
is discussed herein (for more details, see Figure 2 and the
Supporting Information).[17]
The central Zn atom, Zn1, is surrounded by a tetraden-
tate O4-coordinating metallohost with the four oxygen
atoms in the equatorial plane of the octahedral geometry.
The axial positions are occupied by coordinating DMF li-
gands (not shown). From this structure, it can be inferred
that the introduction of the third metal ion causes the other
Zn ions to displace from the basal plane that is defined by
the N2O2 coordination pockets, as evident from the observed
Scheme 1. Synthesis of trinuclear complexes 2–4 from bis-Zn
metalloligand 1.
ACHTUNGTRENN(UNG salphen)
A
A
cations and different nucleophiles was investigated. The new
trinuclear complexes were applied as bifunctional catalysts
in the cycloaddition of CO2 to epoxides, thereby exploiting
the presence of a Lewis acid and a nucleophile (halide)
within the same molecule. This feature obviates the use of
co-catalytic amounts of additives in the form of commonly
used tetrabutyl ammonium halide salts, which could decom-
pose at high temperatures.[15] The best catalyst reported
herein can be recycled up to five times without a significant-
ly loss of activity, thus demonstrating its high stability and
superior recycling potential compared with a typical binary
system used as a control.
ꢀ
bond lengths and angles around Zn2. For instance, the Zn2
ꢀ
O1 and Zn2 O2 bond lengths are markedly different
(1.9561 versus 2.0970 ꢄ) when compared to other Zn-
2642
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2641 – 2648

Recyclable Trinuclear Bifunctional Catalyst
FULL PAPER
addition of one equivalent of ZnI2 is sufficient to induce
these observed changes. The final UV/Vis spectrum was
identical to that measured for isolated complex 2 and the
addition of excess of ZnI2 induced no further changes.
Different control experiments were then performed with
complexes 5b–7 (Figure 4) by using UV/Vis spectroscopy.
First, the titration of mono-ZnACTHNUGTRNEUNG(salphen) complex 5b with
ZnI2 in THF was performed to examine whether trinuclear
assemblies (Figure 1b) could be obtained. However, no
changes were observed, even after the addition of excess
ZnI2, thus suggesting that, in solution, the stabilization
effect of mono-ZnACTHNUTRGNEUNG(salphen) is not high enough to allow for
the formation of binuclear and/or trinuclear assemblies.
ꢀ
Figure 2. X-ray molecular structure of Zn3I2 bisACTHNUGRTENUNG(salphen) complex
2·(DMF)2. Coordinating DMF molecules (to atom Zn1), co-crystallized
solvent molecules, and hydrogen atoms are omitted for clarity. Partial
atom numbering is provided. Selected bond lengths [ꢄ] and angles [8]
ꢀ
with estimated standard deviations in parentheses: Zn2 O1 1.9561(11),
ꢀ
ꢀ
ꢀ
ꢀ
Zn2 O2 2.0970(11), Zn2 N1 2.2278(13), Zn2 N2 2.0419(13), Zn2 I1
ꢀ
ꢀ
ꢀ
2.5854(3), Zn1 O1 2.0162(11), Zn1 O2 2.1291(11), Zn1 O3 2.1192(11),
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Zn1 O4 2.0223(11); O2 Zn2 N1 147.63(5), N2 Zn2 O1 126.04(5), O1
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Zn2 I1 119.043, N2 Zn2 I1 114.69(4), O1 Zn1 O3 170.16(4), O2 Zn1
O4 174.31(4).
ACHTUNGTRENNUNG(salphen) structures that contain axially coordinated
anions.[16] This difference is likely an effect of the require-
ments that are imposed by the introduction of a third metal
center. Zn ions Zn2 and Zn3 from the metalloACTHUNRGTENNUGligAHCUTNGTRENaNGUN nd both
Figure 4. Salphen complexes 5b–7 that were used for the UV/Vis control
have one iodide anion that is coordinated in a mutual trans
mode. Whereas complex 4 is
comprised of two potentially bi-
dentate OAc anions, monoden-
tate coordination is observed in
the molecular structure, with
DMF molecules occupying the
axial ligation sites of the central
Zn metal center (as for com-
pounds 2 and 3, see the Sup-
porting Information).
experiments.
With the aim of studying the
formation of these new trinu-
clear complexes in more detail,
UV/Vis titration of metallo
ACHTUNGTNERlNUNG ig-
ACHTUNGTRENNUNGand 1 with ZnI2 was performed.
Upon titration in THF, a signifi-
cant change in the absorption
behavior was noted and four
clear isobestic points, together
with
trace, was noted in the region
l=200–500 nm (Figure 3).
a
blue-shifted UV/Vis
Moreover, complex 1 is a red
solid, whereas trinuclear com-
pound 2 is yellow. From the ti-
Figure 3. UV/Vis titration of complex 1 with ZnI2, which shows the formation of complex 2. The dashed line
corresponds to the UV/Vis spectrum of metallohost 1; the bold line corresponds to the UV/Vis spectrum of
ꢀ
complex 2. Inset: Titration plot, which shows that only one equivalent of ZnI2 is necessary to form the Zn3I2
tration plot, it is clear that the bisACTHNUTRGNEUNG(salphen) complex (2).
Chem. Eur. J. 2013, 19, 2641 – 2648
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2643

A. W. Kleij et al.
In the second control experiment, bis-Ni
(salphen) com-
other products (such as the mono- and/or dimetalated spe-
cies, 1) were observed. By using relative signal integration,
the amount of complex 2 that formed was determined to be
30% (expected: 33%). After the addition of two equiva-
lents of ZnI2, the amount of complex 2 increased to 64%
(expected: 66%) and, as anticipated, the addition of three
equivalents of ZnI2 resulted in the complete conversion of
L-H4 into complex 2. These results indicate a cooperative
formation of compound 2, similar to some of the multinu-
clear complexes reported by Nabeshima and co-workers,
and again points to the high stability of this trinuclear spe-
cies.[5] Notably, ligand precursor L-H4 can exist in two tauto-
meric forms (cf., keto–enol equilibrium), thus giving rise to
two separate NMR patterns; this feature has been observed
previously for similar Schiff base compounds.[19]
plex 6 was evaluated for its propensity to form a trinuclear
complex upon combination with ZnI2. Complex 6 is com-
prised of a similar “O4” cavity, which can potentially bind
the Zn cation but lacks the ability to interact with halides
after initial complexation because the Ni centers in complex
6 are coordinatively saturated. Again, after the addition of
excess ZnI2, no visible changes could be observed, thus indi-
cating the necessity of having Lewis acidic Zn ions present
for binding the halide anions.
Finally, the use of mono-ZnACTHNUTRGNEUGN(salphen) complex 7, which
also incorporated an O4 cavity, was also titrated with ZnI2
and the reaction was followed by UV/Vis spectroscopy.
Upon titration in THF, a clear change in the absorption be-
havior was noted and two clear isosbestic points, together
with a red-shifted UV/Vis trace, were noted in the region
l=200–500 nm. From the titration plot, it is clear that the
addition of one equivalent of ZnI2 is sufficient to induce the
observed changes, presumably owing to the formation of a
bis-Zn complex (Figure 5) that is reminiscent of a binuclear
species that we isolated previously by combining compound
A possible formation mechanism for these trinuclear com-
plexes is shown in Scheme 2. First, the metal center in the
7 with ZnACHTUNGTRENNUNG
(OAc)2.[18] Thus, the results from these UV/Vis
studies clearly point to two requirements for the formation
of stable multinuclear complexes: 1) The presence of a poly-
dentate O-rich cavity, and 2) the presence of Zn ions in the
metalloACHTUNGTRENNUNGliACHTUNGTRENNUNGgand.
The formation of trinuclear complex 2 was further studied
1
by H NMR spectroscopy in [D7]DMF. Starting with ligand
precursor L-H4 (Figure 6),[14a] increasing amounts of ZnI2
were added in a stepwise manner and the products that
arose from these additions were identified. After the addi-
tion of one equivalent of ZnI2, the reaction seemed to be se-
lective for the formation of trinuclear complex 2 and no
ꢀ
Scheme 2. Proposed mechanism for the formation of Zn2MX2 bis-
AHCTUNGERTG(NNUN salphen) complexes 2–4.
MX2 salt is coordinated by two
phenolic oxygen atoms of one
of the salphen moieties on the
metallohost (1). In the second
step, rotation of the biphenyl
bridge allows the iodide atoms
to move closer to the outer-rim
Zn ions and the central Zn
center is concomitantly coordi-
nated by the other salphen unit
through two additional O-coor-
dination bonds, thereby rigidify-
ing the structure. Finally, the in-
teractions of the iodides with
the outer Zn atoms of the met-
allohost allows “splitting” of
the MX2 salt by forming trime-
tallic supramolecular complexes
2–4.
Figure 5. UV/Vis titration of ZnACTHUNTRGNEU(GN salphen) 7 with ZnI2, which shows the formation of a new binuclear species.
The stability of these trinu-
clear complexes is extremely
important for eventual catalytic
The dashed line corresponds to the UV/Vis spectrum of compound 7; the bold line corresponds to the UV/Vis
spectrum of the new binuclear complex that is formed in situ. Inset: Titration plot, which shows that only
one equivalent of ZnI2 is necessary to form the Zn2(salphen) complex that is derived from compound 7.
2644
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2641 – 2648

Recyclable Trinuclear Bifunctional Catalyst
FULL PAPER
(A) was used as the substrate. When compound 2 was used
in the first run, high conversion (84%) was achieved
(Table 1, entry 7). Trinuclear complex 2 could be easily recy-
cled by precipitation and extraction of the product (for de-
tails, see the Experimental Section). Both recycling experi-
ments with the residue (Table 1, entries 8 and 9) showed
much lower conversion. Control experiments with ZnI2
(Table 1, entry 2), ZnCl2 (Table 1, entry 3), and bimetallic
complex 1 also revealed low conversions and we tentatively
ascribe the drop in activity of complex 2 to possible halide
exchange between the substrate and the catalyst structures
because chloride is known to be a much poorer nucleophile
for the ring-opening of an epoxide than iodide.[15,20] In an at-
tempt to improve the recycling features of these trinuclear
systems, the reaction of substrate B with MI2 salts (M=Zn,
Mg; Table 1, entries 4 and 5) revealed high conversions into
the desired product. However, the recycling of the simple
salts was impossible, owing to gel formation (see the Sup-
porting Information) after cooling the reaction mixtures to
ambient temperature. The presence of iodide as a nucleo-
phile remains an important requirement because the reac-
tion with ZnACTHUNRGTNEUNG(OAc)2 (Table 1, entry 6) did not lead to any
product formation.
After supporting the salts (Table 1, entries 4–6) by using
trinuclear complexes 2–4, the resulting structures were ex-
amined for their catalytic and recycling properties (Table 1,
entries 10–18) by using substrate B. In the case of complex
2, excellent conversion, yield, and selectivity for the car
ACHTUNGTRENNUNGbon-
AHCTUNGTRENNUNG
up to five consecutive runs without any significant loss of ac-
tivity. In contrast, when complex 3 (i.e., the heterotrimetallic
Zn2Mg complex) was applied, the conversion dropped to
81% in the third run, thus showing somewhat inferior recy-
cling potential. As expected, when complex 4 was used as
the catalyst under similar conditions, only poor conversion
into the carbonate product was achieved, in line with the
lower nucleophilicity of the OAc anion. As a first control
experiment, the reaction of mononuclear complex 7 with
one equivalent of ZnI2 was also evaluated (Table 1,
entry 19); this reaction showed high conversion, but recy-
cling of the catalyst was not possible owing to the formation
of a gel after the reaction, which did not allow separation of
the product and the catalyst. Then, binuclear complex 1 was
combined with two equivalents of NBu4I (i.e., a typical
binary catalyst) to investigate its activity and recycling; al-
though complete conversion was achieved (Table 1,
entry 20), the catalyst was inseparable, because two liquid
phases were observed during isolation, which hampered its
recycling (also see the Supporting Information). Therefore,
trinuclear bifunctional complex 2 is the best catalyst in this
series because it can advantageously combine high and se-
lective substrate conversion with excellent recycling poten-
tial.
Figure 6. 1H NMR formation studies of compound
2
in [D7]DMF:
ZnI2; c) L-
a) Ligand
L-H4;
b) L-H4+one equivalent
of
H4+two equivalents of ZnI2; d) L-H4+three equivalents of ZnI2; e) an
ꢀ
authentic sample of Zn3I2 bis
peaks that are related to ligand L-H4, whereas the red diamonds indicate
peaks that are related to trinuclear complex 2.
G
applications and recycling studies. Therefore, the stability of
trinuclear complex
2 was studied by high-temperature
1H NMR spectroscopy ([D7]DMF, 25!125!258C). Subse-
quently, the solution was kept at 858C for 18 h; in both ex-
periments (see the Supporting Information) no changes
could be observed, thus confirming the high thermal stabili-
ty of compound 2 under these conditions.
Because previous work from ourselves and others re-
vealed that ZnACHTUNGTRENNUNG(salen) complexes have good potential for the
cycloaddition reactions of CO2 with oxiranes,[20] trinuclear
complexes 2–4 were subsequently evaluated in the catalytic
formation of cyclic organic carbonates under neat condi-
tions. The presence of different anions with distinct nucleo-
philicities and different metal cations in complexes 2–4
allows for an optimization of the catalytic performance. In
the first series of experiments (Table 1), epichlorohydrin
The proposed catalytic cycle is shown in Scheme 3: First,
dissociation of one or two iodide anions occurs (this dissoci-
ation proceeds more readily at higher temperatures), after
which the epoxide coordinates to the central Zn center.
Chem. Eur. J. 2013, 19, 2641 – 2648
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2645

A. W. Kleij et al.
Table 1. Catalytic cycloaddition of CO2 to epoxides A and B.[a]
Then, the iodide nucleophile attacks the coordi-
ꢀ
nated substrate, thereby resulting in a Zn alkox-
ide intermediate that, after CO2 insertion, gives a
linear carbonate structure. In the last step, the
product is formed through ring-closure, by virtue
of an intramolecular nucleophilic attack and, final-
ly, the product dissociates from the metal center.
Entry
Cat.
Substrate
Cycle
Conv.
[%][b]
Yield[c]
[%]
Selectivity
[%][d]
ACHTUNGERTN(NUNG Cat.)r
[%][e]
1
2
3
4
5
6
7
8
1
ZnI2
ZnCl2
ZnI2
MgI2
ZnOAc2
2
2
2
2
2
2
2
2
3
3
A
A
A
B
B
B
A
A
A
B
B
B
B
B
B
B
B
B
B
B
–
–
–
–
–
–
1
2
3
1
2
3
4
5
1
2
3
–
–
–
0
11
0
84
100
0
80
47
47
100
100
100
100
97
100
100
81
–
–
–
–
–
–
>99
–
>99
>99
–
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
–
–
–
–
–
Conclusion
New supramolecular trinuclear complexes have
been obtained by using metalloACTHNUTRGENNUGliACHUTGTNRENgUNG and 1 and vari-
–
–
65
44
44
96
97
96
95
93
98
96
77
–
97
91
89
95
98
96
99
97
97
96
96
–
ous salts of type MX2, thereby providing highly
stable compounds that combine the binding of a
Lewis acid cation and reversibly bind nucleophilic
anions within the same structure. Trinuclear com-
plexes 2–4 were applied as bifunctional catalysts
in the cycloaddition of CO2 to epoxides A and B,
thus affording cyclic carbonates under conditions
that allow for halide dissociation and epoxide acti-
vation. Catalysis data have clearly shown that
compound 2 can be recovered and reused up to
five times without any significant drop in the con-
version of the epoxide substrate or the yield of
the isolated carbonate product. This supramolecu-
lar approach towards catalyst immobilization and
recycling is highly promising, as demonstrated in a
typical reaction that requires both Lewis acid and
nucleophilic activation; the recycling potential of
compound 2 is superior to the commonly used
binary-type catalysts that comprise both (but sepa-
rate) components.
9
10
11
12
13
14
15
16
17
18
19
20
3
4
13
100
100
7[f]
1[g]
–
–
–
–
[h]
[a] Neat conditions, 18 h, 858C, pACTHNUTRGNE(NUG CO2)=10 bar, catalyst loading: 2.5 mol% of active Zn
sites. [b] Determined by 1H NMR spectroscopy (CDCl3). [c] Yield of the isolated prod-
uct. [d] For the cyclic carbonate product, as determined by 1H NMR spectroscopy
(CDCl3). [e] Mass percentage of the recovered catalyst after each cycle. [f] Used in com-
bination with one equivalent of ZnI2. [g] Used in combination with two equivalents of
NBu4I. [h] Recycling was not achieved, owing to the formation of two liquid phases.
Experimental Section
General: All of the starting materials were purchased from commercial
sources and used without further purification. Bis-Zn
(salphen) 1,[14a] Zn-
were prepared according to
AHCTUNGTRENNUNG
[14a]
AHCTUNGTRENNUNG
(salphen) 7,[18] and ligand precursor L-H4
literature procedures. Elemental analysis was performed at the Unidad
de Anꢀlisis Elemental of the University of Santiago de Compostela
(Spain). NMR spectroscopy was performed on a Bruker-400 MHz spec-
trometer at ambient temperature unless otherwise stated; chemicals
shifts are given in ppm versus TMS. MS and X-ray crystallography were
performed by the Research Support Unit of the ICIQ. UV/Vis spectra
were recorded on a Shimadzu UV1800 spectrophotometer.
ꢀ
Synthesis of Zn3I2 bis
(salphen) complex 2: Bis-Zn
N
(500 mg, 0.55 mmol) and ZnI2 (177 mg, 0.55 mmol) were dissolved in
THF (20 mL) and the solution was stirred for 24 h. The yellow precipi-
tate was filtered off and washed with MeOH to afford complex 2 as a
yellow solid (669 mg, 99% yield). Single crystals suitable for X-ray dif-
fraction were obtained from recrystallization in DMF. 1H NMR
(300 MHz, [D7]DMF): d=6.17 (m, 4H; CHAr), 6.95 (q, J=7.1 Hz, 4H;
CHAr), 7.07 (m, 6H; CHAr), 7.33 (m, 4H; CHAr), 7.43–7.62 (m, 10H;
CHAr), 7.79 (d, J=6.5 Hz, 2H; CHAr), 7.88 (d, J=8.1 Hz, 2H; CHAr),
9.29 ppm (s, 2H; CHimine); the compound was not soluble enough for
13C {1H} NMR analysis; MS (MALDI+, 2-[2E-3-(4-tert-butylphenyl)-2-
methylprop-2-enylidene]malononitrile (DCTB)): m/z calcd: 1101 [MꢀI]+;
found: 1101; elemental analysis calcd (%) for C52H34I2N4O4Zn3·8H2O:
C 45.49, H 3.67, N 4.08; found: C 45.74, H 3.40, N 3.79.
Scheme 3. Proposed catalytic cycle with trinuclear bifunctional complex
2.
2646
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2641 – 2648

Recyclable Trinuclear Bifunctional Catalyst
FULL PAPER
ꢀ
Synthesis of Zn2MgI2 bis
G
G
found: 894; elemental analysis calcd (%) for C52H34N4Ni2O4·2H2O:
C 67.19, H 4.17, N 5.91; found: C 66.99, H 4.11, N 6.01.
1 (50 mg, 0.055 mmol) and MgI2 (15.3 mg, 0.055 mmol) were dissolved in
THF (5 mL) and the solution was stirred for 24 h at RT. The yellow pre-
cipitate was filtered off and washed with MeOH to afford complex 3 as a
yellow solid (63 mg, 97% yield). Single crystals suitable for X-ray diffrac-
tion were obtained from recrystallization in DMF. 1H NMR (500 MHz,
[D7]DMF): d=6.15 (m, 4H; CHAr), 6.90 (t, J=7.4 Hz, 2H; CHAr), 6.97
(m, 2H; CHAr), 7.08 (m, 4H; CHAr), 7.29 (m, 4H; CHAr), 7.48 (m, 2H;
CHAr), 7.57 (m, 8H; CHAr), 7.76 (d, J=7.7 Hz, 2H; CHAr), 7.87 (d, J=
8.3 Hz, 2H; CHAr), 9.25 ppm (s, 2H; CHimine); the compound was not
soluble enough for 13C {1H} NMR analysis; MS (MALDI+, pyrene): m/z
calcd: 1061 [MꢀI]+; found: 1061; elemental analysis calcd (%) for
C52H34I2N4O4Zn2Mg·8H2O: C 46.89, H 3.78, N 4.21; found: C 47.07,
H 3.75, N 3.34.
UV/Vis titration procedure: A 1.0ꢅ10ꢀ5 m solution of the corresponding
salphen complex in THF was prepared and the UV/Vis spectrum was re-
corded by using a quartz cell (path length: 1 cm). A 1.0ꢅ10ꢀ4 m solution
of ZnI2 in THF was prepared that contained the corresponding salphen
complex (1.0ꢅ10ꢀ5 m) and the solution was added as small aliquots
(10 mL) to the Zn(salphen) complex in THF having the concentration in-
dicated above (2.0 mL). After each addition, a spectrum was recorded.
Formation studies of Zn3I2 bisACHTNUGRTNEUNG
(salphen) complex 2 by using 1H NMR
ꢀ
spectroscopy in [D7]DMF: The bis-salphen ligand precursor to com-
pound 1 (5 mg, 6.4ꢅ10ꢀ3 mmol) was dissolved in [D7]DMF (0.7 mL) and
a
1H NMR spectrum was acquired. In a second tube, the bis-salphen
ligand (5 mg, 6.4ꢅ10ꢀ3 mmol) and one equivalent of ZnI2 was dissolved
in [D7]DMF (0.7 mL) and a 1H NMR spectrum was acquired; then, the
tube was heated at 608C for 18 h and another H NMR spectrum was ac-
ꢀ
Synthesis of Zn
(OAc)2 bis
1
plex 1 (50 mg, 0.055 mmol) and ZnI2 (12 mg, 0.055 mmol) were dissolved
in THF (3 mL) and the solution was stirred for 24 h. The yellow precipi-
tate was filtered off and washed with MeOH to afford complex 4 as a
yellow solid (57 mg, 95% yield). Single crystals suitable for X-ray diffrac-
tion were obtained from recrystallization in DMF. 1H NMR (400 MHz,
[D7]DMF): d=1.58 (s, 6H; CH3), 6.13 (m, 2H; CHAr), 6.88–7.05 (m,
10H; CHAr), 7.23–7.53 (m, 16H; CHAr), 7.75 (m, 4H; CHAr), 9.27 ppm (s,
2H; CHimine); the compound was not soluble enough for 13C {1H} NMR
analysis; MS (MALDI+, DCTB): m/z calcd: 1033 [MꢀOAc]+; found:
1033; elemental analysis calcd (%) for C56H40N4O8Zn3·4H2O: C 57.73,
H 4.15, N 4.81; found: C 57.41, H 3.89, N 4.62.
quired. In a third tube, the bis-salphen ligand (5 mg, 6.4ꢅ10ꢀ3 mmol) and
two equivalents of ZnI2 were dissolved in [D7]DMF (0.7 mL) and a
1H NMR spectrum was acquired; then, the tube was heated at 608C for
18 h and another 1H NMR spectrum was acquired. In a fourth tube, the
bis-salphen ligand (5 mg, 6.4ꢅ10ꢀ3 mmol) and three equivalents of ZnI2
were dissolved in [D7]DMF (0.7 mL) and a 1H NMR spectrum was ac-
quired.
ꢀ
Stability studies of Zn3I2 bis
(salphen) complex 2 by using VT-NMR
1
spectroscopy: Complex 2 (5 mg) was dissolved in [D7]DMF and H NMR
spectra were acquired over the range 25–1258C at 208C intervals by first
heating to 1258C and then stepwise cooling to 258C. Then, the NMR
tube was kept at 858C for 18 h and a 1H NMR spectrum was acquired
and compared with the sample before heating.
Synthesis of ligand precursor to compound 5b (compound 5a): (E)-2-
{[(2-aminophenyl)imino]ACHTUNGTRENNUNG(phenyl)methyl}phenol (200 mg, 0.7 mmol) and
2-hydroxybenzaldehyde (85 mg, 0.7 mmol) were dissolved in MeOH and
the solution was stirred for 18 h. The yellow precipitate was filtered off,
washed with MeOH, and dried in air to afford compound 5a as a yellow
solid (270 mg, 98% yield). 1H NMR (400 MHz, CDCl3): d=6.72 (dt, 3J=
8.3 Hz, J=1.27 Hz, 1H; CHAr), 6.80 (m, 1H; CHAr), 6.91 (dt, J=7.5 Hz,
4J=1.2 Hz, 1H; CHAr), 6.97 (d, J=8.4 Hz, 1H; CHAr), 7.04–7.11 (m, 8H;
CHAr), 7.20 (m, 2H; CHAr), 7.31–7.39 (m, 3H; CHAr), 8.39 (s, 1H; CHim),
12.98 (s, 1H; OH), 14.08 ppm (s, 1H; OH); 13C {1H} NMR (100 MHz,
CDCl3): d=117.60, 118.10, 118.29, 118.61, 119.04, 119.47, 119.73, 123.31,
125.63, 127.06, 128.01, 128.55, 129.05, 132.31, 132.50, 133.32, 133.57,
134.74, 140.13, 141.91, 161.36, 162.64, 162.77, 174.77 ppm; HRMS
(MALDI+): m/z calcd for C26H19N2O2: 391.1447 [MꢀH]+; found:
391.1465.
Catalytic experiments: Typical procedure: A mixture of (pre)catalyst 2
(330 mg, 2.5 mol% of active Zn sites) and glycidyl methyl ether (1.00 g,
11.35 mmol) was added to a stainless steel reactor. Three cycles of pres-
surization and depressurization of the reactor (with CO2, 5 bar) were per-
formed before the pressure was finally stabilized at 10 bar. Then, the re-
actor was heated to 858C and the mixture stirred for a further 18 h.
Then, an aliquot of the solution was analyzed by 1H NMR spectroscopy
(CDCl3) and the conversion was determined. Next, the (pre)catalyst and
the cyclic carbonate product were isolated (see below).
4
Catalyst recovery and recycling: The reaction mixture was dissolved in a
minimum amount of CH2Cl2 and n-pentane was added to precipitate the
catalyst. Then, the solid (catalyst) was filtered off, dried, and reused.
After catalyst separation, the solvent was evaporated from the organic
phase and the epoxide/cyclic-carbonate mixture was kept under reduced
pressure for 3 h. Then, the organic carbonate was obtained as a pure
compound and the yield was determined.
Synthesis
hydroxybenzylidene)amino)phenyl)imino)
200 mg, 0.51 mmol) and Zn(OAc)2·2H2O (34 mg, 0.61 mmol) were dis-
of
ZnN
solved in MeOH (5 mL) and the solution was stirred for 18 h. Then, the
solvent was evaporated and the yellow solid that remained was washed
with MeOH and dried in air to afford complex 5b as a yellow solid
(230 mg, 99% yield). 1H NMR (400 MHz, [D6]DMSO): d=6.28 (dt, 3J=
7.5 Hz, 4J=1.3 Hz, 1H), 6.49 (m, 2H), 6.66 (d, J=8.0 Hz, 1H), 6.76 (d,
3J=8.5 Hz, 4J=1.1 Hz, 1H), 6.81 (m, 1H), 6.86 (d, 3J=8.3 Hz, 4J=
1.8 Hz, 1H), 7.09 (m, 1H), 7.16 (dt, 3J=7.6, 4J=1.8 Hz, 1H), 7.23 (m,
3H), 7.38 (m, 4H), 7.56 (m, 1H), 8.85 ppm (s, 1H); 13C {1H} NMR
(125 MHz, [D6]DMSO): d=112.16, 112.86, 116.79, 119.41, 120.32, 123.00,
123.14, 124.19, 125.50, 125.69, 128.21, 128.47, 128.76, 133.23, 134.30,
134.31, 135.98, 136.77, 139.62, 139.82, 162.47, 172.23, 172.56, 173.55 ppm;
MS (MALDI+, pyrene): m/z calcd: 454 [M]+; found: 454; elemental
analysis calcd (%) for C26H18N2O2Zn·4H2O: C 59.16, H 4.96, N 5.31;
found: C 59.56, H 3.76, N 4.99.
Acknowledgements
This work was supported by the ICIQ, the ICREA, and the Spanish Min-
isterio de Economꢁa y Competitividad (MINECO, CTQ2011-27385). We
thank Dr. Noemꢁ Cabello, Vanessa Martꢁnez, and Sofꢁa Arnal for per-
forming the MS studies.
[1] S. J. Garibay, J. R. Stork, S. Cohen, Progr. Inorg. Chem. 2009, 56,
335.
Synthesis of bis-Ni
ACHTUNGTRENNUNG
[2] For some general reviews on metal-containing hosts/systems and
their potential applications, see: a) R. Chakrabarty, P. S. Mukherjee,
P. J. Stang, Chem. Rev. 2011, 111, 6810; b) M. Yoshizawa, J. K. Klos-
terman, M. Fujita, Angew. Chem. 2009, 121, 3470; Angew. Chem.
Int. Ed. 2009, 48, 3418; c) M. J. Wiester, P. A. Ulmann, C. A. Mirkin,
Angew. Chem. 2011, 123, 118; Angew. Chem. Int. Ed. 2011, 50, 114;
d) D. Fiedler, D. H. Leung, R. G. Bergman, K. N. Raymond, Acc.
Chem. Res. 2005, 38, 349; e) T. S. Koblenz, J. Wassenaar, J. N. H.
Reek, Chem. Soc. Rev. 2008, 37, 247; f) A. W. Kleij, J. N. H. Reek,
(50 mg, 0.06 mmol) and NiAHCTUNTGRENNUNG
(20 mL) was stirred for 18 h. Then, the solvent was evaporated and the
solid brown residue was washed with MeOH (10 mL) to afford complex
1
6 as a brown solid (53 mg, 98% yield). H NMR (300 MHz, [D6]DMSO):
d=5.94 (d, J=8.4 Hz, 2H), 6.14 (m, 2H), 6.52–6.75 (m, 12H), 6.90 (d,
J=7.1 Hz, 4H), 7.04 (t, J=7.6 Hz, 4H), 7.40–7.55 (m, 10H), 7.98 (d, J=
8.2 Hz, 2H), 8.76 ppm (s, 2H); the compound was not soluble enough for
13C{1H} NMR analysis; MS (MALDI+, pyrene): m/z calcd: 894 [M]+;
Chem. Eur. J. 2013, 19, 2641 – 2648
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2647

A. W. Kleij et al.
Chem. Eur. J. 2006, 12, 4218; g) J. Meeuwissen, J. N. H. Reek, Nat.
Chem. 2010, 2, 615.
[3] Supramolecular Catalysis, (Ed.: P. W. N. M. van Leeuwen), Wiley-
VCH, Weinheim, Germany 2008.
[4] For an example, see: D. de Groot, B. F. M. van der Waal, J. N. H.
Reek, A. P. H. J. Schenning, P. J. C. Kamer, E. W. Meijer, P. W. N. M.
van Leeuwen, J. Am. Chem. Soc. 2001, 123, 8453.
[5] For selected reviews, see: a) S. Akine, T. Nabeshima, Dalton Trans.
2009, 10395; b) S. Akine, J. Inclusion Phenom. Macrocyclic Chem.
2012, 72, 25.
[13] M. Aresta, A. Dibenedetto, J. Mol. Catal. A 2002, 182, 399.
[14] a) M. V. Escꢀrcega-Bobadilla, G. Salassa, M. Martꢁnez-Belmonte,
E. C. Escudero-Adꢀn, A. W. Kleij, Chem. Eur. J. 2012, 18, 6805; see
also: b) S. J. Wezenberg, G. Salassa, E. C. Escudero-Adꢀn, J. Benet-
Buchholz, A. W. Kleij, Angew. Chem. 2011, 123, 739; Angew. Chem.
Int. Ed. 2011, 50, 713.
[15] W. Clegg, R. W. Harrington, M. North, R. Pasquale, Chem. Eur. J.
2010, 16, 6828.
[16] S. J. Wezenberg, E. C. Escudero-Adꢀn, J. Benet-Buchholz, A. W.
Kleij, Chem. Eur. J. 2009, 15, 5695.
[6] a) S. Akine, T. Taniguchi, T. Nabeshima, J. Am. Chem. Soc. 2006,
128, 15765; b) S. Akine, T. Taniguchi, T. Nabeshima, Inorg. Chem.
2008, 47, 3255.
[7] S. Akine, S. Kagiyama, T. Nabeshima, Inorg. Chem. 2007, 46, 9525.
[8] For recent contributions that exploit such O-donor-rich cavities to
create larger supramolecular structures, see: a) S. Akine, S. Sunaga,
T. Taniguchi, H. Miyazaki, T. Nabeshima, Inorg. Chem. 2007, 46,
2959; b) P. D. Frischmann, A. J. Gallant, J. H. Chong, M. J. Mac-
[17] Crystallographic details for complex 2: C67H69N9I2O9Zn3; Mr =
¯
1594.22; triclinic; space group P1; a=10.2948(8), b=17.9546(14),
c=18.8896(14); a=85.143(2), b=77.857(2), g=75.857(2)8; V=
3307.9(4) ꢄ3; Z=2; 1=1.601 mgmꢀ3; m=2.075 mmꢀ1; l=0.71073 ꢄ;
T=100(2) K; F
ACHTUNGTREN(NUNG 000)=1604; crystal size=0.08ꢅ0.08ꢅ0.06 mm; q-
A
ACHTUNGTRENNUNG
unique reflections (Rint =0.0207); GoF=1.020; R1 =0.0213 and
wR2 =0.0542 [I>2s(I)], R1 =0.0248 and wR2 =0.0563 (all indices),
min/max residual density: ꢀ0.518/1.013 [eꢄꢀ3], completeness to
ACHTUNGTRENNUNGLachACHTUNGTRENNUNGlan, Inorg. Chem. 2008, 47, 101; c) T. Nabeshima, H. Miyazaki,
A. Iwasaki, S. Akine, T. Saiki, C. Ikeda, Tetrahedron 2007, 63, 3328;
q (30.04)=89.4%. CCDC-894525 (2), CCDC-894526 (3), and
d) P. D. Frischmann, G. A. Facey, P. Y. Ghi, A. J. Gallant, D. L.
Bryce, F. Lelj, M. J. MacACTHNUTRGENNLGU achACHUTNGTRENlNNGU an, J. Am. Chem. Soc. 2010, 132, 3893.
[9] I. Gamba, P. Gamez, E. Monzani, L. Casella, I. Mutikainen, J. Reed-
ijk, Eur. J. Inorg. Chem. 2011, 4360.
CCDC-894527 (4) contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[10] P. Seth, L. K. Das, M. G. B. Drew, A. Gosh, Eur. J. Inorg. Chem.
2012, 2232.
[18] L. San Felices, E. C. Escudero-Adꢀn, J. Benet-Buchholz, A. W.
Kleij, Inorg. Chem. 2009, 48, 846.
[11] a) D. J. Darensbourg, Chem. Rev. 2007, 107, 2388; b) A. Decortes,
A. M. Castilla, A. W. Kleij, Angew. Chem. 2010, 122, 10016; Angew.
Chem. Int. Ed. 2010, 49, 9822; c) M. R. Kember, A. Buchard, C. K.
Williams, Chem. Commun. 2011, 47, 141; d) M. North, R. Pasquale,
C. Young, Green Chem. 2010, 12, 1514; e) T. Sakakura, K. Kohno,
Chem. Commun. 2009, 1312; f) T. Sakakura, J.-C. Choi, H. Yasuda,
Chem. Rev. 2007, 107, 2365; g) T. Ema, Y. Miyazaki, S. Koyama, Y.
Yano, T. Sakai, Chem. Commun. 2012, 48, 4489.
[19] A. Filarowski, J. Phys. Org. Chem. 2005, 18, 686–698.
[20] a) A. Decortes, A. W. Kleij, ChemCatChem 2011, 3, 831; b) A. De-
cortes, M. Martꢁnez Belmonte, J. Benet Buchholz, A. W. Kleij,
Chem. Commun. 2010, 46, 4580. Also see: c) Y.-M. Shen, W.-L.
Duan, M. Shi, J. Org. Chem. 2003, 68, 1559; d) J.-W. Huang, M. Shi,
J. Org. Chem. 2003, 68, 6705.
Received: November 19, 2012
[12] J. Sun, S.-I. Fujita, F. Zhao, M. Arai, Appl. Catal. A 2005, 287, 221.
Published online: January 25, 2013
2648
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2641 – 2648