Synthesis and electrochemical study of an original copper(II)-capped salen-cyclodextrin complex

Article abstract of DOI:10.1002/ejic.200901208

A new metallocapped cyclodextrin (CD) was synthesized by the regioselective debenzylation, induced by diisobutylaluminium hydride (DIBAL-H), of perbenzylated cyclodextrins. This reaction allowed for the efficient preparation of an unprecedented CD-salen type copper(II) complex. The electrochemical behavior of both the bound and unbound CD-salen compounds was investigated by cyclic voltammetry. Notably, it was shown that the presence of tert-butyl groups at the ortho- and para-positions of the salen aromatic rings stabilized the copper(II) phenoxyl radical species that was generated upon the one-electron oxidation of the starting compound. Importantly, this stabilization remained effective when the salen-type ligand was covalently attached to the CD. This allowed for investigations of the reactivity of the copper(II) phenoxyl radical complex towards a primary alcohol to be performed by cyclic voltammetry. This reaction can be considered as mimicking the behavior of galactose oxidase. However, under these conditions, no reactivity was observed in the presence of benzyl alcohol. This may be due to distortion, either of the initially square planar salen ligand after its grafting to the CD primary face, and/or of the CD itself. On the other hand, the electrochemical reduction of the un-grafted copper(II) salen-type ligand led to a transient anionic species that exhibited significant stability on the time-scale of the slow cyclic voltammetry measurement in the absence of the CD, but was unstable in the presence of the CD. In the latter case, it was demonstrated that the anionic species was protonated by the CD. Importantly, this protonation was not fast enough to prevent catalytic activation of iodomethane by the electro-generated copper(I)-capped salen CD complex. A new metallocapped cyclodextrin was synthesized by a regioselective debenzylation reaction that was induced by diisobutylaluminium hydride (DIBAL-H). The electrochemical investigation of this original complex demonstrated that grafting of a copper salen-type complex to a cyclodextrin leads to changes in the reactivity of the intermediary species.

Full text of DOI:10.1002/ejic.200901208

FULL PAPER
DOI: 10.1002/ejic.200901208
Synthesis and Electrochemical Study of an Original Copper(II)-Capped
Salen–Cyclodextrin Complex
Elise Deunf,[a] Elena Zaborova,[b] Samuel Guieu,[b] Yves Blériot,[b] Jean-Noël Verpeaux,[a]
Olivier Buriez,*[a] Matthieu Sollogoub,*[b] and Christian Amatore*[a]
Keywords: Copper / Metalloenzymes / Cyclodextrins / Electrochemistry
A new metallocapped cyclodextrin (CD) was synthesized by
the regioselective debenzylation, induced by diisobutylalu-
minium hydride (DIBAL-H), of perbenzylated cyclodextrins.
This reaction allowed for the efficient preparation of an un-
precedented CD–salen type copper(II) complex. The electro-
chemical behavior of both the bound and unbound CD–salen
compounds was investigated by cyclic voltammetry. Notably,
it was shown that the presence of tert-butyl groups at the
ortho- and para-positions of the salen aromatic rings stabi-
lized the copper(II) phenoxyl radical species that was gener-
ated upon the one-electron oxidation of the starting com-
pound. Importantly, this stabilization remained effective
when the salen-type ligand was covalently attached to the
CD. This allowed for investigations of the reactivity of the
copper(II) phenoxyl radical complex towards a primary
alcohol to be performed by cyclic voltammetry. This reaction
can be considered as mimicking the behavior of galactose
oxidase. However, under these conditions, no reactivity was
observed in the presence of benzyl alcohol. This may be due
to distortion, either of the initially square planar salen ligand
after its grafting to the CD primary face, and/or of the CD
itself. On the other hand, the electrochemical reduction of
the un-grafted copper(II) salen-type ligand led to a transient
anionic species that exhibited significant stability on the
time-scale of the slow cyclic voltammetry measurement in
the absence of the CD, but was unstable in the presence of
the CD. In the latter case, it was demonstrated that the an-
ionic species was protonated by the CD. Importantly, this
protonation was not fast enough to prevent catalytic acti-
vation of iodomethane by the electro-generated copper(I)-
capped salen CD complex.
The second approach involves covalently attaching one
or more ligands of the transition metal complex to the pri-
mary or secondary face of the CD. In such a configuration,
the metal center can be kept close to the CD cavity regard-
less of the solvent polarity. We recently used this approach
to design a new cobalt(II)–CD complex, and to investigate
the stability/reactivity duality of the corresponding low-val-
ent cobalt(I) species that was electrochemically generated.[3]
This cobalt-CD complex was prepared from CoX2 (X =
Br and BF4) in the presence of 1 equiv. of 6-deoxy-N-(2-
methyliminopyridine)-β-CD ligand. Under these condi-
tions, we demonstrated for the first time, that the electro-
generated cobalt(I) species could be kinetically and thermo-
dynamically stabilized by the presence of the CD. Import-
antly, this did not prevent the cobalt(I) species from remain-
ing sufficiently reactive to enable it to activate aryl halides.
We also used this approach to generate a Pd-capped CD,
and demonstrated for the first time the inherent chirality of
regioisomerically functionalized CDs.[1a]
Introduction
Combining cyclodextrins (CDs) with transition metal
salts or complexes has proven to be a very attractive way of
designing new catalysts, enzyme mimics, sensors, and mo-
lecular wires.[1] Two main approaches are generally used to
immobilize transition metals in the vicinity of CDs . In the
first one, the CD acts as a second-sphere ligand binding
noncovalently to the first ligand sphere of the metal center.
This supramolecular approach, which relies on the ability of
the CDs to include part of a metal complex in their internal
hydrophobic cavities, is strongly dependent on the nature of
the solvent. Indeed, the driving force for the complexation
between a CD and a substrate involves several factors such
as van der Waals forces, hydrophobic interactions, elec-
tronic effects, and steric factors, all of which may be too
weak to enforce a strong association between the adduct
partners.[2] Hence, when the metal ligands are hydrophobic,
dissociation is favored in organic solvents and water ap-
pears to be the most suitable solvent for complexation.
Copper is an important transition metal in catalysis.
Copper is used in organic synthesis,[4] but is also found in
metalloenzymes such as galactose oxidase (GO) that cataly-
ses the oxidation of primary alcohols to their corresponding
aldehydes with concomitant reduction of molecular di-
oxygen to hydrogen peroxide.[5] Interestingly, this two-elec-
[a] Ecole Normale Supérieure, Département de Chimie,
UMR CNRS-ENS-UPMC 8640 “PASTEUR”,
24 rue Lhomond, 75231 Paris cedex 05, France
[b] UPMC Univ. Paris 06, Institut Parisien de Chimie Moléculaire
(UMR CNRS 7201), C.181,
4 place Jussieu, 75005 Paris, France
View this journal online at
wileyonlinelibrary.com
4720
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4720–4727

Copper(II)-Capped Salen–Cyclodextrin Complex
tron oxidation process is promoted by a single copper atom with the cavity, hence forming a so-called metallocapped
acting in synergy with a tyrosyl radical from the protein. CD (see 3 in Scheme 1).[11] The presence of the CD cavity
The fact that GO possesses a N2O2 coordination sphere has right under the metal center could enhance the selectivity
led to the preparation of numerous compounds that model of the catalyst towards oxidation reactions, and allow reac-
its active site.[5] A wide number of copper(II) complexes of tions to be performed in aqueous media. Moreover, combi-
salen-type ligands have been prepared and investigated, but nation of the rigid salen ligand with a less rigid cyclodextrin
only a limited number of reports of functional models that ligand could result in ligands that exhibit interesting chemi-
exhibit activity towards the oxidation of primary alcohols cal reactivity. Since tetrahedral distortion of the salen moi-
are available. A functional model of GO was described in ety was found to be of importance in aiding the oxidation
1996 by Stack et al. involving the one-electron oxidized cop- of primary alcohols, 6A,6D-dideoxy-6A,6D-bis[(tert-butyl-N-
per(II) phenolate complex of a salen-type ligand.[6] This salicylidene)imino]-α-cyclodextrin (3) has been designed to
model compound was able to catalyze the aerobic oxidation modify the initial geometry of the salen moiety, and most
of benzyl alcohol to benzaldehyde with high turnover val- likely the CD itself.
ues. To achieve such reactivity, the phenolate moieties were
Electrochemical investigations of both this new cop-
ortho- and para-substituted with electron donating groups per(II)–CD–salen complex and its parent complexes
that had the effect of stabilizing the phenoxyl radical, (Scheme 1) were useful, not only for probing the redox
furthermore a significant tetrahedral distortion was in- properties of this complex, but also for delineating the
duced around the copper atom by the binaphthyl linker. structure/reactivity relationship of the electro-generated
However, oxidation of the initial copper(II)-phenolate com- transient species.
plex to the catalytically active copper(II)-phenoxyl radical
species required the use of a strong chemical oxidant. Alter-
natively, it was shown that the copper(II) complex of N,NЈ-
bis(2-hydroxy-3,5-di-tert-butylbenzyl)-1,2-ethylenediamine
Results and Discussion
was readily electrochemically oxidized, and allowed for the
a) Synthesis of Complex (3)
oxidation of primary alcohols to aldehydes but could not
oxidize secondary alcohols under the same conditions.[7]
In order to synthesize complex 3 (Scheme 2) two diamet-
Other copper(II) complexes of salen-type ligands have been rically opposed primary hydroxyl groups of the CD ligand
prepared and tested, and have been found to oxidize pri- had to be selectively modified. Our strategy was based on
mary alcohols with reasonable success.[8]
the perbenzylation-debenzylation sequence developed by
In this context, and taking advantage of our recent works some of us[12] to regioselectively access a wide variety of
on CDs,[1a,3,9] we decided to prepare and examine the reac- functional CDs.[13] This gives access, in very high yield, to
tivity of a complex in which the salen-type ligand would be α-cyclodextrin 4, which is selectively protected at all but the
covalently attached to the primary face of the CD. To the 6A and 6D positions. Amino groups were then introduced
best of our knowledge, there is only one report dedicated by mesylation, followed by substitution with an azido group
to discussing the challenge of building of a salen-type li- to yield CD 5. CD 5 was then reduced by LiAlH4 to give
gand onto a cyclodextrin rim.[10] This report detailed the CD 6. Hydrogenolysis of the benzyl groups on CD 6 af-
synthesis of a CD–salen-type ligand in which the two salen forded the diamino-α-cyclodextrin 7 in 52% overall yield.
moieties were covalently attached to the A and B positions It is worth noting that the azido groups had to be reduced
of the CD. This ligand was used to prepare the correspond- prior to hydrogenolysis of the benzyl groups to avoid poi-
ing manganese(III) complex, a superoxide dismutase mimic. soning the catalyst. This route is also an improvement on a
Our aim here is to observe the influence of the CD on previously described synthesis method for 7 that gave only
the electrochemical behavior of a copper(II) complex; we a 3.7% yield.[14] Copper salen complex 3 was then synthe-
therefore designed a CD-based salen ligand that would sized according to a classical procedure.[15] The condensa-
bring the metal center of the complex into close proximity tion reaction involving 3,5-di-tert-butyl-2-hydroxybenzalde-
Scheme 1. The copper(II) complexes of salen-type ligands investigated in this work.
Eur. J. Inorg. Chem. 2010, 4720–4727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4721

O. Buriez, M. Sollogoub, C. Amatore et al.
FULL PAPER
Scheme 2. Synthesis route for complex 3.
hyde was performed in ethanol and DMF for solubility
The oxidation of 1 was poorly resolved at a glassy carbon
reasons. The final complex was obtained using CuClO4 as electrode surface. Therefore, a platinum electrode was used
a copper source, with an overall yield of 40% based on the to follow the oxidation process of 1 (Figure 1, B). Under
native α-cyclodextrin. (Scheme 2).
these conditions an irreversible two-electrons oxidation
wave was exhibited at +1.03 V, showing the low stability of
the electro-generated copper phenoxyl radical species with
respect to the time-scale of the slow cyclic voltammetry ex-
periment.
b) Electrochemical Behavior of Complexes 1, 2, and 3
Cyclic voltammograms obtained during reduction or oxi-
Importantly, when the same experiment was run with 2
dation scans of the copper(II) salen complex 1 are shown in instead of 1 the oxidation process gained some reversibility
Figure 1. As already reported, the reduction of 1 proceeded (E° = +0.88 V), and tended towards a one-electron process.
through two distinct electron-transfer processes located at This confirmed that a sterically bulky and electron-donat-
–1.22 V (E° = –1.18 V) and –2.49 V (Figure 1, A).[16] The ing substituent at the ortho- and para-positions of the two
first wave corresponded to the electrochemically reversible aromatic rings thermodynamically and kinetically stabilized
CuII/CuI transition, whereas the second one certainly fea- the phenoxyl radical (Figure 1, D). A similar effect involv-
tured a reduction of CuI to the corresponding Cu0 complex, ing an ethylenediamine linker substituting for an ethylenedi-
followed by removal of the metal center from the ligand imine linker has already been evidenced based on cyclic vol-
(Scheme 3). Although no Cu0 deposit could be observed in tammetry and EPR studies.[7a] The reversible oxidation
the anodic range, which was explored, the reduction of the wave OЈ2 may therefore be attributed to the oxidation of
resulting free dianionic ligand could be observed at a more the µ-phenolato moiety to a µ-phenoxyl radical. In other
negative potential value (E = –2.95 V) (not shown in Fig- words, the oxidation process involves the ligand rather than
ure 1).
the metal (Scheme 3).
4722
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4720–4727

Copper(II)-Capped Salen–Cyclodextrin Complex
Figure 1. Voltammetry of 1 (2 m) in DMF with TBABF4 (0.1 ) as the supporting electrolyte, recorded at a scan rate of 200 mVs–1.
Reductive (A) and oxidative (B) directions for the initial scans were recorded at a glassy carbon electrode (1 mm diameter) and at a
platinum electrode (0.5 mm diameter), respectively. Voltammetry of 2 in DMF with TBABF4 (0.1 ) as the supporting electrolyte, recorded
at a glassy carbon electrode (1 mm diameter), at a scan rate of 100 mVs–1. Reductive (C) and oxidative (D) scans. The concentration of
2 was 2 m for scan C and 1 m for D.
Scheme 3. Oxidation and reduction processes of complexes 1 and 2.
On the other hand, the presence of the four tert-butyl shown in Figure 1, C) once again leading to a rapid expul-
groups in complex 2 did not lead to a change in the re- sion of Cu0.
duction process when compared to 1. The reduction still
The electrochemical behavior of complex 3 in which the
occurred at E° = –1.18 V and remained reversible. A second salen-type ligand is covalently attached to the CD primary
reduction wave was also observed at higher potentials (not face, was significantly different from that of 1 and 2 (Fig-
Eur. J. Inorg. Chem. 2010, 4720–4727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4723

O. Buriez, M. Sollogoub, C. Amatore et al.
FULL PAPER
Figure 2. Voltammetry of 3 (1 m) in DMF with TBABF4 (0.1 ) as the supporting electrolyte, recorded at a scan rate of 200 mVs–1 (A,
B, D) and 30 Vs–1 (C). Reductive (A, B, C) and oxidative (D) scans recorded at a glassy carbon electrode (1 mm diameter) in the absence
and the presence of iodomethane (B, C).
ure 2). Although the reduction of 3 proceeded through an the electro-generated copper(I)-capped salen–cyclodextrin
irreversible wave RЈЈ1 (–1.71 V) at slow scan rates, the pro- and the alkyl halide. This process is faster than the proton-
cess became reversible above 30 Vs–1 which is evidence for ation reaction.
a rather slow follow up reaction involving the initial CuI
Interestingly, and as observed with the nongrafted com-
species in the presence of the cyclodextrin (Figure 2, A and plex 2, the oxidation of 3 appeared to be reversible at
C). A second reduction wave RЈЈ2 could be also observed 0.2 Vs–1, confirming the stabilization of the electro-gener-
at –2.40 V at slow scan rates (Figure 2, A), but again, and ated phenoxyl radical in the presence of the tert-butyl
contrary to previous reported works,[16,17] no Cu0 deposit groups despite the presence of the cyclodextrin (Figure 2,
could be observed despite the irreversibility of RЈЈ1. Since D). This prompted us to investigate, by cyclic voltammetry,
CDs are known to be good proton donors, the irreversibil- the reactivity of the phenoxyl radical towards a primary
ity of the reduction wave may involve protonation of the alcohol to examine its potential as a galactose oxidase
anionic species produced in RЈЈ1. To verify this point one mimic. Thus, increasing amounts of benzyl alcohol were
molar equivalent of native α-CD was added to a solution added to a solution of 3 and voltammetric data were col-
containing complex 2, and under these conditions the ini- lected at slow scan rates. However, no changes were ob-
tially reversible reduction wave RЈ1 became irreversible. On served in the voltammograms when compared to the initial
the other hand, RЈ1 remained reversible after the addition cyclic voltammogram of 3 showing that no reaction oc-
of one molar equivalent of a per-benzylated α-CD. These curred over the voltammetric time-scale (t Ͻ 1 s). This lack
results clearly established the feasibility of a reaction be- of reactivity may be due to a distortion of the square planar
tween the phenate moiety and a proton donated by a CD salen-type ligand as previously observed when a binaphthyl
moiety, even under less favorable conditions than in the re- linker was attached to the complex.[6] Indeed, the flexibility
action incorporating 2 (bimolecular reaction with 1:1 con- and degree of distortion required to shift from a square
centration ratio). Such a reaction may then lead to the con- planar towards a tetrahedral geometry is certainly modu-
version of the initial CuI to the corresponding protonated lated by the nature of the linker between the salicylidene
form that is prone to being reduced in RЈЈ2.
moieties.[5] Yet, other effects may also explain such lack of
Importantly, over a short time-scale (v Ն 30 Vs–1) the reactivity, such as a possible distortion of the CD itself due
addition of increasing amounts of iodomethane led to the to its natural flexibility, or because of steric hindrance
disappearance of the oxidation wave OЈЈ1 and a slight in- caused by the presence of the CD.
crease in the peak current intensity of RЈЈ1. This current
peak increase was more pronounced when the same experi-
Conclusions
ment was performed over a longer time-scale (v =
100 mVs–1; compare Figure 2 , B and C). These results
For the first time, a copper(II) salen-type complex was
clearly showed a redox catalysis process occurring between synthesized by covalent attachment of a salen ligand to the
4724
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4720–4727

Copper(II)-Capped Salen–Cyclodextrin Complex
trode cell with an Autolab potentiostat (PGSTAT 20). The refer-
ence electrode was a SCE (Tacussel), which was separated from
the solution by a bridge compartment filled with the same solvent/
supporting electrolyte solution that was used in the cell. The
counter electrode was a gold wire, 1 cm in length. The glassy
carbon (1 and 3 mm diameter; Goodfellow) and platinum (0.5 mm
diameter; Goodfellow) working electrodes were prepared in-house.
primary face of an α-cyclodextrin at the A–D position. The
preparation of this new complex was possible due to the
remarkable ability of diisobutylaluminium hydride to re-
gioselectively de-O-benzylate perbenzylated cyclodex-
trins.[12,13] The electrochemical behaviors of the grafted and
nongrafted copper(II) salen-type ligands were investigated
by cyclic voltammetry. It was shown that the presence of
tert-butyl groups at the ortho- and para-positions of the
salen aromatic rings stabilized the copper phenoxyl radical
obtained by oxidation of the starting compound. A similar
stabilization was obtained when the salen-type ligand was
covalently attached to the CD. Similarly, the electrochemi-
cal reduction of the nongrafted copper(II) salen-type ligand
led to an anionic species that appeared to be a little bit
less stable in the presence of an added cyclodextrin, most
Syntheses and Chemical Characterizations: Reactions were moni-
tored by thin-layer chromatography (TLC) on a precoated plate of
silica gel 60 F254 (layer thickness 0.2 mm; E. Merck, Darmstadt,
Germany) with detection permitted by charring with sulfuric acid.
1H NMR [13C NMR] spectra were recorded at room temperature
with a 400 MHz [100 MHz] Bruker AVANCE 400 spectrometer.
Chemical shifts are given in ppm, referenced to the residual reso-
nances of the solvents (δ = 7.26 or 77.16 ppm, for CDCl3). Cou-
pling constants (J) are given in Hertz [Hz]. Assignments were aided
probably because it was protonated by the CD. The reactiv- by COSY, J-mod technique and HMQC. Optical rotations were
determined with a Perkin–Elmer 343 polarimeter (c = g/100 mL,
CHCl3). Mass spectrometry experiments were carried out on a
Micromass ZABSpecTOF or a Bruker microTOF.
ity of the electro-generated copper(I)-capped salen CD
complex and of the corresponding copper phenoxyl radical
were investigated towards iodomethane and benzyl alcohol,
respectively. Iodomethane could be activated through a re-
dox catalytic process which occurred faster than the proton-
α-CD–Disalen–Cu Complex 3: An ethanol solution of NaOH
(0.24 mL, 5 g/L, 2 equiv.) and CuCl2 (2.0 mg, 15 µmol, 1 equiv.)
ation of the copper(I)-capped salen CD. On the other hand, was added to a solution of disalen α-CD 8 (21 mg, 15 µmol,
1 equiv.) in ethanol (1.0 mL) at room temp. The reaction mixture
was stirred under nitrogen for 12 h. Then the solvent was removed
by evaporation under vacuum. The residue was washed with diethyl
ether (2ϫ3 mL) to give CD 3 (21 mg, 96%) as a green amorphous
powder. [α]2D0 = +104 (c = 0.05, methanol). HRMS (ESI) [M +
Na]+: calcd. for C66H100N2O30CuNa 1486.5549; found 1486.5096.
no reactivity was observed between the copper bound phen-
oxyl radical and the primary alcohol. This lack of reactivity
may be due to a possible distortion either of the initially
square planar salen ligand after its grafting to the CD pri-
mary face, and/or of the CD itself.
These results clearly showed that grafting of a copper
salen-type complex on to a CD led to intermediary species
with new and interesting reactivity properties. This may
lead to new methods for the design of new metal-based cat-
alysts and enzyme mimics. A systematic study is in progress
that involves grafting the salen-type ligand onto other CD
positions with the purpose of tailoring the molecular distor-
tions, and to test the catalytic ability of these new com-
plexes towards the oxidation of primary alcohols.
6A,6D-Dideoxy-6A,6D-diazido-2A–F,3A–F,6B,6C,6E,6F-hexadeca-O-benz-
yl-α-cyclodextrin (5): NaN3 (1.6 g, 25.2 mmol, 8 equiv.) was added
to
a solution of a
known dimesylated α-CD[13a,13e] (8.1 g,
3.15 mmol, 1 equiv.) in DMF (45 mL) at room temp. under nitro-
gen. The reaction mixture was heated at 80 °C for 12 h, then cooled
to room temp., and the DMF removed by evaporation. The residue
was dissolved in EtOAc (20 mL) and washed with water (20 mL).
The aqueous layer was extracted with EtOAc (3ϫ20 mL). The
combined organic layers were dried with MgSO4, filtered, and con-
centrated under vacuum. Silica gel flash chromatography of the
residue (cyclohexane/EtOAc: 5:1Ǟ3:1) gave 5 (7.1 g, 92%) as an
amorphous powder. [α]2D0 = +52.0 (c = 1.0, CHCl3); Rf = 0.50 (cy-
clohexane/AcOEt, 3:1). 1H NMR (400 MHz, CDCl3): δ = 7.36–
Experimental Section
Chemicals: Tetrabutylammonium tetrafluoroborate (TBABF4),
which was dissolved in dimethylformamide (DMF) and used as
the supporting electrolyte for electrochemical measurements, was
prepared from NaBF4 (Acros), and nBu4NHSO4 (Acros). The
TBABF4 was recrystallized from ethyl acetate/hexane (both sol-
vents were purchased from Acros), and dried at 60 °C.
2
7.15 (m, 80 H, H arom.), 5.31 (d, J = 10.7 Hz, 2 H, 2ϫ CHPh),
3
2
5.23 (d, J1,2 = 3.5 Hz, 2 H, 2ϫ 1-H), 5.19 (d, J = 11.0 Hz, 2 H,
2
3
2ϫ CHPh), 5.05 (d, J = 11.4 Hz, 2 H, 2ϫ CHPh), 5.00 (d, J1,2
3
= 3.1 Hz, 2 H, 2ϫ 1-H), 4.99 (d, J1,2 = 3.1 Hz, 2 H, 2ϫ 1-H),
4.92 (d, 2J = 11.0 Hz, 2 H, 2ϫ CHPh), 4.89 (d, 2J = 11.4 Hz, 2 H,
2
2
2ϫ CHPh), 4.86 (d, J = 10.7 Hz, 2 H, 2ϫ CHPh), 4.65 (d, J =
2
12.3 Hz, 2 H, 2ϫ CHPh), 4.56 (d, J = 11.8 Hz, 4 H, 4ϫ CHPh),
Solvents were freshly distilled from Na/benzophenone (THF, tolu-
ene) or P2O5 (CH2Cl2). Reactions were carried out under Ar. Flash
column chromatography was performed on silica gel 60 (230–
400 mesh, E. Merck). Diisobutylaluminium was purchased from
Aldrich as a 1.5  solution in toluene. All other chemicals were
purchased.
4.53 (d, 2J = 12.7 Hz, 2 H, 2ϫ CHPh), 4.50 (d, 2J = 12.4 Hz, 2 H,
2
2ϫ CHPh), 4.47 (d, J = 13.6 Hz, 4 H, 4ϫ CHPh), 4.45 (s, 4 H,
2ϫ CH2Ph), 4.43 (d, 2J = 11.2 Hz, 2 H, 2ϫ CHPh), 4.18 (dd, 3J2,3
3
= 8.3, J3,4 = 9.7 Hz, 2 H, 2ϫ 3-H), 4.14–4.09 (m, 6 H, 4ϫ 3-H,
2ϫ 6-H), 4.05–3.99 (m, 8 H, 4ϫ 4-H, 2ϫ 5-H, 2ϫ 6-H), 3.95 (br.
3
3
3
d, J4,5 = 9.6 Hz, 2 H, 2ϫ 5-H), 3.89 (br. dd, J4,5 = 9.4, J5,6
=
Complexes 1 and 2 were synthesized as described elsewhere by
combining stoichiometric amounts of the appropriate salen ligand
and copper(II) acetate.[16,18] These complexes were not prepared in-
situ due to the possibility of the acetate ions reacting with the final
Cu(salen) complex.[16]
2.9 Hz, 2 H, 2ϫ 5-H), 3.77 (br. d, 2J = 11.0 Hz, 2 H, 2ϫ 6-H),
3.75 (dd, J3,4 = 8.6, J4,5 = 8.6 Hz, 2 H, 2ϫ 4-H), 3.65 (br. d, J
= 10.3 Hz, 2 H, 2ϫ 6-H), 3.59 (dd, J5,6 = 2.0, J = 10.5 Hz, 2 H,
2ϫ 6-H), 3.56 (dd, 3J1,2 = 3.7, 3J2,3 = 9.9 Hz, 2 H, 2ϫ 2-H), 3.53–
3.49 (m, 4 H, 2ϫ 2-H, 2ϫ 6-H), 3.45 (dd, 3J1,2 = 3.3, 3J2,3 = 9.9 Hz,
2 H, 2ϫ 2-H) ppm. 13C NMR (100 MHz, CDCl3): δ = 140.25 (2
C), 140.22 (2 C), 140.20 (2 C), 139.38 (2 C), 139.20 (2 C), 139.12
3
3
2
3
2
Instrumentation: Cyclic voltammetry experiments were performed
at room temperature under an argon atmosphere in a three-elec-
Eur. J. Inorg. Chem. 2010, 4720–4727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4725

O. Buriez, M. Sollogoub, C. Amatore et al.
FULL PAPER
(2 C), 139.01 (2 C), 138.99 (2 C, 16ϫ C arom. quat.), 128.38– 82.97 (2 C, 6ϫ C-4), 75.02 (2 C), 74.89 (2 C), 74.61 (2 C, 6ϫ C-
126.88 (m, 80ϫ C arom. tert.), 98.96 (2 C), 98.79 (2 C), 98.25 (2 3), 74.06 (2 C), 73.63 (2 C), 73.59 (2 C), 73.53 (2 C), 73.52 (2 C,
C, 6 ϫ C-1), 80.87 (2 C), 80.81 (2 C), 80.68 (2 C, 6ϫ C-3), 80.42
6ϫ C-2, 4ϫ C-5), 69.70 (2ϫ C-5), 61.95 (2 C), 61.88 (2 C), 41.96 (2
(2 C), 80.03 (2 C), 79.61 (2 C, 6ϫ C-4), 79.32 (2 C), 78.99 (2 C), C, 6 ϫ C-6) ppm. HRMS (ESI) [M + H]+: calcd. for C36H63N2O28
78.44 (2 C, 6ϫ C-2), 75.86 (2 C), 75.74 (2 C), 75.04 (2 C), 73.48 (2 971.35619; found 971.35553 (–0.7 ppm).
C), 73.46 (2 C), 73.12 (2 C), 72.90 (2 C), 72.60 (2 C, 16ϫ CH2Ph),
α-CD–Disalen 8: 3,5-Di-tert-butyl-2-hydroxybenzaldehyde (33 mg,
71.86 (2 C), 71.64 (2 C), 70.78 (2 C, 6ϫ C-5), 69.47 (2 C), 68.91 (2
140 µmol, 3 equiv.) was added to a solution of diamine α-CD 7
C), 52.34 (2 C, 6ϫ C-6) ppm. HRMS (ESI) [M + 2Na]++: calcd.
(40 mg, 610 µmol, 1 equiv.) in ethanol (5.0 mL) and in DMF
for C148H154N6O28Na2 1254.52978; found 1254.53318 (+2.8 ppm).
(1.0 mL) under argon at room temp. The reaction mixture was
6A,6D-Dideoxy-6A,6D-diamino-2A–F,3A–F,6B,6C,6E,6F-hexadeca-O-benz-
heated at 90 °C for 12h, cooled to room temp., and the solvents
removed by evaporation. The residue was washed with acetone to
give 8 (47 mg, 80%) as a yellow amorphous powder. [α]2D0 = +89.0
(c = 1.0, methanol). 1H NMR (400 MHz, MeOD): δ = 8.46 (s, 2
yl-α-cyclodextrin (6): LiAlH4 (92 mg, 2.44 mmol, 6 equiv.) was
added to a solution of diazido α-CD 5 (1.0 g, 0.406 mmol, 1 equiv.)
in dry THF (10 mL) at room temp. under nitrogen. The reaction
mixture was heated at 60 °C for 2 h, then water was added, and
THF was removed by evaporation. The residue was dissolved in
EtOAc (15 mL) and washed with water (10 mL) and diluted HCl
(1 mol·L–1 in water, 10 mL). Then the layers were separated and
the aqueous layer was extracted with EtOAc (3ϫ15 mL). The com-
bined organic layers were washed with NaOH (1 mol. L–1 in water,
10 mL), dried with MgSO4, filtered, and concentrated under vac-
uum. Silica gel flash chromatography of the residue (CH2Cl2/
MeOH, 40:1) gave 6 (807 mg, 82%) as an amorphous powder.
[α]2D0 = +36.6 (c = 1.0, CHCl3); Rf = 0.30 (CH2Cl2/MeOH, 95:5).
1H NMR (400 MHz, CDCl3): δ = 7.35–6.85 (m, 80 H, H arom.),
4
H, 2ϫ N=C–H), 7.38 (d, J = 2.1 Hz, 2 H, 2ϫ H arom.), 7.19 (d,
3
4J = 2.1 Hz, 2 H, 2ϫ H arom.), 4.98 (d, J1,2 = 2.8 Hz, 2 H, 2ϫ
3
3
1-H), 4.95 (d, J1,2 = 2.7 Hz, 4 H, 4ϫ 1-H), 4.17 (br. dd, J = 9.2,
3J = 4.3 Hz, 2 H, 2ϫ 5-H), 4.07–3.92 (m, 12 H, 4ϫ 6-H, 2ϫ 5-H,
6ϫ 3-H), 3.88–3.72 (m, 10 H, 8ϫ 6-H, 2ϫ 5-H), 3.59–3.44 (m, 12
H, 6ϫ 2-H, 6ϫ 4-H), 1.41 (s, 18 H, tBu), 1.31 (s, 18 H, tBu) ppm.
13C NMR (100 MHz, MeOD): δ = 170.06 (2ϫ C=N), 159.45 (2
C), 141.29 (2 C), 137.55 (2 C), 119.54 (2 C, 8ϫ C arom. quat.),
127.97 (2 C), 127.56 (2 C, 4ϫ C arom. tert.), 104.01 (2 C), 103.84
(2 C), 103.57 (2 C, 6ϫ C-1), 85.24 (2 C), 83.26 (2 C), 83.00 (2 C,
6ϫ C-4), 75.27 (2 C), 75.21 (2 C), 75.07 (2 C, 6ϫ C-3), 73.90 (8
C, 6 ϫ C-2, 2ϫ C-5), 73.76 (2ϫ C-5), 71.98 (2ϫ C-5), 61.91 (2 C),
61.73 (2 C), 60.23 (2 C, 6ϫ C-6), 35.91 (2 C), 35.01 (2 C, 4ϫ C
tBu quat.), 31.93 (2 C), 30.03 (2 C, 4ϫ CH3) ppm. HRMS (ESI) [M
+ Na]+: calcd. for C66H102N2O30Na 1425.64096; found 1425.64119
(+0.2 ppm).
3
2
5.50 (d, J1,2 = 3.6 Hz, 2 H, 2ϫ 1-H), 5.31 (d, J = 10.5 Hz, 2 H,
2
2ϫ CHPh), 5.09 (d, J = 10.8 Hz, 2 H, 2ϫ CHPh), 4.85–4.60 (m,
18 H, 14ϫ CHPh, 4ϫ 1-H), 4.50–4.20 (m, 14 H, 10ϫ CHPh, 4ϫ
3
5-H), 4.11 (t, J2,3 = 9.4 Hz, 2 H, 2ϫ 3-H), 4.04–3.84 (m, 12 H,
4ϫ 3-H, 4ϫ 4-H, 4ϫ 6-H), 3.82–3.65 (m, 8 H, 4ϫ CHPh, 2ϫ 4-
3
3
H, 2ϫ 5-H), 3.62 (d, J5,6 = 10.4 Hz, 2 H, 2ϫ 6-H), 3.54 (d, J5,6
3
3
= 10.6 Hz, 2 H, 2ϫ 6-H), 3.46 (dd, J1,2 = 3.8, J2,3 = 10.0 Hz, 2
Acknowledgments
3
3
H, 2ϫ 2-H), 3.35 (dd, J1,2 = 3.4, J2,3 = 9.8 Hz, 2 H, 2ϫ 2-H),
3.23 (dd, 3J1,2 = 3.0, 3J2,3 = 9.6 Hz, 2 H, 2ϫ 2-H), 2.78 (br. d, 3J5,6
This work was supported in part by the Centre National de la Re-
cherche Scientifique (CNRS) (UMR CNRS-ENS-UPMC 8640)
and the Ministère de la Recherche (Ecole Normale Supérieure).
The French Ministry of Research (MESR) is also thanked for sup-
porting the PhD grant of E. D.
3
= 13.4 Hz, 2 H, 2ϫ 6-H), 2.77 (br. d, J5,6 = 13.8 Hz, 2 H, 2ϫ 6-
H) ppm. 13C NMR (100 MHz, CDCl3): δ = 139.35 (2 C), 139.32
(2 C), 139.31 (2 C), 138.57 (2 C), 138.36 (2 C), 138.21 (2 C), 138.10
(2 C), 138.02 (2 C, 16ϫ C arom. quat.), 128.40–128.10, 128.05–
127.85, 127.75–127.50, 127.17–126.93, 126.77, 126.39 (m, 80ϫ C
arom. tert.), 98.36 (2 C), 98.27 (2 C), 98.02 (2 C, 6ϫ C-1), 81.38
(2 C), 80.98 (2 C), 80.92 (4 C), 80.76 (2 C), 80.66 (2 C, 6ϫ C-3,
6ϫ C-4), 79.89 (2 C), 79.10 (2 C), 77.99 (2 C, 6ϫ C-2), 76.24 (2
C), 75.95 (2 C), 74.13 (2 C), 73.48 (2 C), 73.41 (2 C), 73.34 (2 C),
72.99 (2 C), 72.26 (2 C, 16ϫ CH2Ph), 71.98 (2 C), 71.90 (2 C),
71.06 (2 C, 6ϫ C-5), 69.69 (2 C), 69.10 (2 C), 42.64 (2 C, 6ϫ C-
6) ppm. HRMS (ESI) [M + H]+: calcd. for C148H159N2O28:
2412.1079; found 2412.1084 (+0.2 ppm).
6A,6D-Dideoxy-6A,6D-diamino-α-cyclodextrin (7): The diamine α-
CD 6 (100 mg, 0.041 mmol, 1 equiv.) was dissolved in a mixture of
H2O (2 mL) and THF (6 mL) under nitrogen, then TFA (25 µL)
and Pd/C 10% (50 mg) were added to the solution. The reaction
mixture was stirred under an H2 atmosphere for 12 h, filtered
through a Celite pad, and concentrated under vacuum. The residue
was washed with mixture of H2O/MeOH (5 mL) and the solvents
were removed by evaporation under vacuum. The residue was
washed with CHCl3 (5 mL) to give 7 (39 mg, 99%) as an amorph-
ous powder. [α]2D0 = +76.2 (c = 0.5, methanol). 1H NMR (400 MHz,
[1] For a recent example of a catalyst developed by the authors of
this article, see: a) S. Guieu, E. Zaborova, Y. Blériot, G. Poli,
A. Jutand, D. Madec, G. Prestat, M. Sollogoub, Angew. Chem.
Int. Ed. 2010, 49, 2314; review articles: b) F. Hapiot, S. Tilloy,
E. Monflier, Chem. Rev. 2006, 106, 767; c) E. Engeldinger, D.
Armspach, D. Matt, Chem. Rev. 2003, 103, 4147; d) J. M.
Haider, Z. Pikramenou, Chem. Soc. Rev. 2005, 34, 120; e) E.
Rizzarelli, G. Vecchio, Coord. Chem. Rev. 1999, 188, 343.
[2] a) J. Szejtli, Cyclodextrin and Their Inclusion Complexes, Akad-
emiai Kiado, Budapest, 1982; b) W. Saenger, Angew. Chem. Int.
Ed. Engl. 1980, 19, 344; c) S. Li, W. C. Purdy, Chem. Rev. 1992,
92, 1457; d) J. Szejtli, Chem. Rev. 1998, 98, 1743.
[3] E. Deunf, O. Buriez, E. Labbé, J. N. Verpeaux, C. Amatore,
Electrochem. Commun. 2009, 11, 114.
[4] G. Evano, N. Blanchard, M. Toumi, Chem. Rev. 2008, 108,
3054.
[5] F. Thomas, Eur. J. Inorg. Chem. 2007, 2379, and references
cited therein.
[6] a) Y. Wang, T. D. P. Stack, J. Am. Chem. Soc. 1996, 118, 13097;
b) Y. Wang, J. L. Dubois, B. Hedman, K. O. Hodgson, T. D. P.
Stack, Science 1998, 279, 537.
[7] a) E. Saint-Aman, S. Ménage, J. L. Pierre, E. Defrancq, G. Gel-
lon, New J. Chem. 1998, 22, 393; b) F. Thomas, O. Jarjayes, C.
Duboc, C. Philouze, E. Saint-Aman, J. L. Pierre, Dalton Trans.
2004, 2662.
[8] a) M. Vaidyanathan, M. Palaniandavar, R. S. Gopalan, Ind. J.
Chem. 2003, 42A, 2210; b) R. C. Pratt, T. D. P. Stack, J. Am.
3
3
MeOD): δ = 5.00 (d, J1,2 = 3.1 Hz, 2 H, 2ϫ 1-H), 4.97 (d, J1,2
=
3
3.3 Hz, 2 H, 2ϫ 1-H), 4.93 (d, J1,2 = 3.1 Hz, 2 H, 2ϫ 1-H), 4.17
3
3
(br. dd, J = 8.0, J = 8.8 Hz, 2 H, 2ϫ 5-H), 4.02–3.68 (m, 18 H,
6ϫ 3-H, 4ϫ 5-H, 8ϫ 6-H), 3.67–3.43 (m, 12 H, 6ϫ 2-H, 6ϫ 4-
H), 3.17 (br. d, 3J = 6.5 Hz, 2 H, 2ϫ 6-H), 3.14 (br. d, 3J = 7.9 Hz,
2 H, 2ϫ 6-H) ppm. 13C NMR (100 MHz, MeOD): δ = 103.80 (2
C), 103.70 (2 C), 103.20 (2 C, 6ϫ C-1), 85.20 (2 C), 83.29 (2 C),
4726
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4720–4727

Copper(II)-Capped Salen–Cyclodextrin Complex
Chem. Soc. 2003, 125, 8716; c) R. C. Pratt, T. D. P. Stack, In-
org. Chem. 2005, 44, 2367.
goub, Angew. Chem. Int. Ed. 2008, 47, 7060; M. Sollogoub,
Eur. J. Org. Chem. 2009, 1295; h) S. Xiao, M. Yang, P. Sinaÿ,
Y. B l ériot, M. Sollogoub, Y. Zhang, Eur. J. Org. Chem. 2010,
1510; i) E. Zaborova, Y. Blériot, M. Sollogoub, Tetrahedron
Lett. 2010, 51, 1254.
D.-Q. Yuan, C. Yang, T. Fukuda, K. Fujita, Tetrahedron Lett.
2003, 44, 565.
A. Puglisi, G. Tabbì, G. Vecchio, J. Inorg. Biochem. 2004, 98,
969.
M. J. Samide, D. G. Peters, J. Electroanal. Chem. 1998, 443, 95.
[9] a) C. Amatore, O. Buriez, E. Labbé, J.-N. Verpeaux, J. Elec-
troanal. Chem. 2008, 621, 134; b) O. Buriez, J. M. Heldt, E.
Labbé, A. Vessières, G. Jaouen, C. Amatore, Chem. Eur. J.
2008, 14, 8195.
[14]
[15]
[16]
[10] A. Puglisi, G. Tabbi, G. Vecchio, J. Inorg. Biochem. 2004, 98,
969.
[11] D. Armspach, D. Matt, Chem. Commun. 1999, 1073.
[12] T. Lecourt, A. Hérault, A. J. Pearce, M. Sollogoub, P. Sinaÿ,
Chem. Eur. J. 2004, 10, 2960.
[13] a) O. Bistri, P. Sinaÿ, M. Sollogoub, Tetrahedron Lett. 2005,
46, 7757; b) O. Bistri, P. Sinaÿ, M. Sollogoub, Chem. Commun.
2006, 1112; c) O. Bistri, P. Sinaÿ, M. Sollogoub, Chem. Lett.
2006, 534; d) O. Bistri, P. Sinaÿ, M. Sollogoub, Tetrahedron
Lett. 2006, 47, 4137; e) O. Bistri, P. Sinaÿ, J. Jiménez Barbero,
M. Sollogoub, Chem. Eur. J. 2007, 13, 9757; f) S. Guieu, M.
Sollogoub, J. Org. Chem. 2008, 73, 2819; g) S. Guieu, M. Sollo-
[17] a) S. Seka, O. Buriez, J. Périchon, Chem. Eur. J. 2003, 9, 3597;
b) S. Seka, O. Buriez, J.-Y. Nédélec, J. Périchon, Chem. Eur. J.
2002, 8, 2534.
[18] R. Sanzenbacher, M. Huber, J. Glerup, M. Neuburger, J.
Springborg, Inorg. Chem. 1996, 35, 7493.
Received: December 14, 2009
Published Online: August 27, 2010
Eur. J. Inorg. Chem. 2010, 4720–4727
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4727