Limits of the Inversion Phenomenon in Triazolyl-Substituted β-Cyclodextrin Dimers

Article abstract of DOI:10.1002/ejoc.201301681

Six different β-cyclodextrin (β-CD) dimers have been synthesized by copper-catalyzed azide alkyne cycloaddition. The nature of the spacer connecting the two β-CDs has been varied in length and hydrophilicity. For each dimer, a detailed NMR study was carried out in D₂O. It was found that, whereas β-CD dimers having short spacer exhibited only one conformation, β-CD dimers having long and/or hydrophobic spacers showed two different conformations. Indeed, the latter underwent an inversion process leading to self-inclusion of the spacer in one of the two CD cavities. The behavior of the six β-CD dimers in water has thus been rationalized and this allowed the proportion of "free" cavities actually available for molecular recognition to be determined. Six β-cyclodextrin (β-CD) dimers have been synthesized by CuAAC. Depending on the nature of the spacer connecting the CDs, one or two conformations could be detected by NMR spectroscopic analysis. An inversion process was observed for dimers with a long and/or hydrophilic spacer. The proportion of "free" β-CD cavities actually available for molecular recognition could be determined.

Full text of DOI:10.1002/ejoc.201301681

FULL PAPER
DOI: 10.1002/ejoc.201301681
Limits of the Inversion Phenomenon in Triazolyl-Substituted β-Cyclodextrin
Dimers
Jonathan Potier,[a,b] Stéphane Menuel,[a,b] Nathalie Azaroual,[a,c] Eric Monflier,[a,b] and
Frédéric Hapiot*[a,b]
Keywords: Host-guest systems / Cyclodextrins / Macromolecules / Conformation analysis / Structure elucidation
Six different β-cyclodextrin (β-CD) dimers have been synthe-
sized by copper-catalyzed azide alkyne cycloaddition. The
nature of the spacer connecting the two β-CDs has been var-
ied in length and hydrophilicity. For each dimer, a detailed
NMR study was carried out in D2O. It was found that,
whereas β-CD dimers having short spacer exhibited only one
conformation, β-CD dimers having long and/or hydrophobic
spacers showed two different conformations. Indeed, the lat-
ter underwent an inversion process leading to self-inclusion
of the spacer in one of the two CD cavities. The behavior of
the six β-CD dimers in water has thus been rationalized and
this allowed the proportion of “free” cavities actually avail-
able for molecular recognition to be determined.
alteration is not without consequence. At best, the water-
solubility of the CD host is enhanced and the association
constant outmatches that measured with the native counter-
part. In other cases, addition of substituents at random on
the CD structure may disrupt the intramolecular hydrogen
bond network, resulting in an enhanced solubility in water
without affecting the molecular recognition properties. For
example, the water-solubility of the randomly methylated β-
CD (RAME-β-CD, substitution degree: 1.8 methyl group
per glucopyranose unit) is Ͼ 500 gL–1 at 25 °C,[10] which is
far higher than that of native β-CD (18.8 gL at 25 °C).[11]
Yet, native β-CD and RAME-β-CD display similar associa-
tion constants regarding guest molecules.[12] However,
grafting a substituent on the CD structure does not always
lead to better intrinsic properties. Two key phenomena may
hinder the inclusion of a guest within the cavity of substi-
tuted CDs. First, if the hydrophobic character of the sub-
stituent is high and its structure can fit the CD cavity, then
a self-inclusion process may occur. As such, we recently
showed that the recognition ability of chemically modified
CDs can be dramatically altered by flexible hydrophobic
substituents that are capable of entering the CD cavity.[13]
In this case, the molecular recognition of guests remains
possible but is significantly hampered. Once the CD has
been substituted, another phenomenon can also alter the
recognition process. In 2004, Kano et al. revealed that
double full rotation of the substituted glucopyranose unit
can occur, leading to a CD conformation in which the cav-
ity is fully occupied by the substituent,[14] thus demonstrat-
ing that the CD cavity is not as rigid as first anticipated. In
line with this observation, Liu et al. reported in 2008 that
the flipping phenomenon can also take place with tetrakis-
(permethyl-β-CD)-modified zinc(II) porphyrins.[15] Simi-
larly, in 2010, Harada et al. revealed that CD dimers were
Introduction
Despite the huge amount of work already dedicated to
cyclodextrins (CDs),[1] these truncated cone-shaped mo-
lecules continue to fascinate researchers in many different
areas such as drug release,[2] cosmetics,[2] the food indus-
try,[3] sensing,[4] biomimetics,[5] chiral enantiomer separa-
tion,[6] and catalysis.[7] The recognition ability of these α-d-
glucopyranose-based cyclic molecules towards organic mo-
lecules mainly explains this fascination. From linear alkyl
chains to more bulky aromatic moieties, α-, β- and γ-CDs
are able to host numerous substrates in water, provided that
the guests display hydrophobic fragments that are sterically
compatible with the CD cavity. For decades the idea that
native (unmodified) CDs are rigid structures has fueled re-
search into their molecular recognition properties.[8] The
case of the native β-CD is especially illustrative. Its recogni-
tion ability is essentially rationalized by the greater rigidity
conferred by the participating intramolecular hydrogen
bonds. Thus, many different substrates can be included
within the native β-CD cavity. To widen the scope of CDs
as molecular hosts, chemical modifications of their struc-
ture has been developed. A wide range of substituents has
thus been covalently grafted on either the CD primary or
secondary face.[9] Nevertheless, the resulting structural
[a] Université Lille Nord de France,
59000 Lille, France
[b] UArtois, Unité de Catalyse et de Chimie du Solide (UCCS),
Faculté des Sciences Jean Perrin,
Rue Jean Souvraz, SP 18, 62307 Lens Cedex, France
E-mail: frederic.hapiot@univ-artois.fr
http://www.uccs.univ-artois.fr/frederichapiot.htm
[c] UDSL, EA 4481, Faculté de Pharmacie, Université de Lille 2,
BP 83, 59006 Lille, France
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201301681.
Eur. J. Org. Chem. 2014, 1547–1556
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1547

J. Potier, S. Menuel, N. Azaroual, E. Monflier, F. Hapiot
FULL PAPER
also involved. In their study, an alkyl altro-α-CD dimer was
converted into the pseudo[1]rotaxane dimer through tum-
bling of the altropyranose unit of altro-α-CD in D2O.[16]
They also showed that the tumbling process originated from
the breakage of the hydrogen bond network, which consti-
tuted the main energetic barrier for the conformational
change of the alkyl altro-α-CD dimer. Their study was ex-
tended to [2]rotaxane and [3]rotaxane.[17] Whereas [3]rotax-
ane resisted reorientation of its altro-α-CD stopper group
due to insufficient space, the altro-α-CD stopper of [2]-
rotaxane tumbled, resulting in conversion into pseudo[2]-
rotaxane in D2O. Recently, we demonstrated that triazolyl-
substituted CD dimers were also subjected to inversion phe-
nomena. More precisely, the recognition ability of water-
soluble β-CD-based ditopic receptors linked together by a
bis(triazolyl)dimethoxybenzene spacer strongly depended
upon the location of the substituents on the phenyl ring.[18]
Because of the inversion phenomenon, the para-isomer
proved to be inappropriate for the recognition of long alkyl-
chain substrates because only one out of two CD cavities
was available. On the other hand, the bulkier meta- and
ortho-isomers disfavored the inversion phenomenon, re-
sulting in an enhanced recognition ability. Similar observa-
tions on the inversion process were made very recently by
Désiré et al. on β-CD glycerol dimers.[19]
Considering the above observations, we wanted to ratio-
nalize the behavior of triazolyl-substituted CDs when incor-
porated in a dimer structure. To this end, we investigated a
series of β-CD-triazolyl-containing dimers to evaluate
(i) the role of the triazolyl group, (ii) the impact of the bulk
of the spacer, (iii) the role of spacer flexibility, and (iv) the
influence of spacer hydrophobicity on the inversion phe-
nomenon. To do so, six different β-CD dimers having link-
ers of different nature were synthesized and their behavior
in water was observed by using 1D and 2D NMR spectro-
scopic analyses. Note that the mechanism of inversion re-
garding the CD is beyond the scope of this study.
Scheme 1. Synthesis of CD dimer 1.
Dimer 2 was synthesized by following the same experi-
mental procedure from meta-diethynylbenzene and mono-
azido β-CD (Scheme 2). After workup, 2 was isolated as a
white powder in 80–85% yield. This compound consisted
of two β-CDs linked together by a rigid meta-bis(triazolyl)-
phenyl spacer. By following the same experimental pro-
cedure, dimer 3 was easily attainable from para-diethyn-
ylbenzene as a reactant. Changing from meta to para geom-
etry on the aromatic phenyl ring altered neither the reaction
time nor the yield.
Results and Discussion
Synthesis
Various synthetic approaches have been developed to ac-
cess six β-CD dimers in which the β-CD residues are
bridged by linkers of different length, hydrophobicity and
flexibility. Synthesis of 1 was realized in dimethyl sulfoxide
(DMSO) at 30 °C by copper-catalyzed azide alkyne cyclo-
Scheme 2. Synthesis of CD dimers 2 and 3.
Unsymmetrical dimers required a multistep strategy. The
addition (CuAAC) using mono-azido β-CD and dipropynyl linker had to be prepared first before the coupling reaction
ether as reactants (Scheme 1) and the CuSO4·5H2O/Na-as- with the β-CD residues was possible. To do so, 4-iodo-
corbate couple as a catalyst. After stirring for 15 h, the phenol and trimethylsilylethyne were reacted in pyrrolidine
crude product was precipitated in acetone. The Cu-catalyst at 60 °C for 5 h under nitrogen in the presence of CuI and
was removed by complexation with ammonia and subse- [PdCl2(PPh3)2] (Scheme 3). Liquid-liquid extraction with
quent chromatography on a silica column. After evapora- NaHCO3-saturated water and ethyl acetate gave a crude
tion of the water–ammonia mixture, the product was sub- product that was subsequently purified by chromatography
mitted to an additional chromatography column using a on silica column using CH2Cl2 as eluent. 4-(Trimethylsilyl-
CH3CN/H2O mixture (8:2 v/v) to purify the β-CD dimer ethynyl)phenol (4) was isolated as a white powder (96%
(1). Dimer 1 was obtained in 80–85% yield as a white pow- yield), which was then reacted with propargyl bromide in
der.
CH3CN under reflux overnight.
1548
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 1547–1556

Inversion Phenomenon in β-Cyclodextrin Dimers
Scheme 4. Synthesis of CD dimer 8.
Scheme 3. Synthesis of CD dimer 6.
After liquid-liquid extraction, the crude product was
purified by chromatography to give the para-disubstituted
benzene derivative 5 as a white powder in 75% yield. Note
that during the course of the reaction the alkyne fragment
was completely deprotected, thus avoiding the need for an
additional deprotection step. Compound 5 was then sub-
jected to CuAAC with mono-azido-β-CD under the experi-
mental conditions described above, leading to dimer 6 in
80–85% yield.
More flexible β-CD dimers were prepared according to
the following procedure. Pentaethylene glycol and propargyl
bromide were mixed in CH3CN in the presence of Cs2CO3
and heated to reflux overnight (Scheme 4). Dialkynyl com-
pound 7 was obtained in poor yield (14%) after liquid-li-
quid extraction and chromatography on the silica column
(CH2Cl2/EtOAc, 7:3 v/v). Subsequent reaction of 7 with
2 equiv. mono-azido-β-CD under CuAAC conditions pro-
duced dimer 8 in 80–85% yield.
Scheme 5. Synthesis of CD dimer 10.
Similarly, 1-chloro-2-[2-(2-chloroethoxy)ethoxy]ethane
reacted overnight with 4 under basic conditions at reflux in
CH3CN (Scheme 5). The crude product was first purified
by liquid-liquid extraction with NaHCO3-saturated water
and ethyl acetate.
NMR Measurements
We sought to clarify the key parameters that control the
inversion phenomenon by a careful examination of the in-
Subsequent chromatography on a silica column (CH2Cl2/ fluence of the structure of the linker on the dimer confor-
heptane, 50:50 to 100:0 v/v) gave dialkynyl derivative 9 as a mation. In this context, the synthesized β-CDs dimers were
pale-yellow powder in 37% yield. As noted above for the submitted to 1D and 2D NMR experiments in D2O at
synthesis of 5, the alkyne was completely deprotected dur- 25 °C. Three different groups of protons served as probes
ing the course of the reaction. Compound 9 and mono- to evaluate the structural modification of the β-CD dimers
azido-β-CD reacted in DMSO under CuAAC conditions to in D2O, namely the triazolyl protons, the linker protons,
1
give dimer 10 (80–85%) after purification.
and the inner CD protons (H-3 and H-5). The H NMR
Eur. J. Org. Chem. 2014, 1547–1556
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1549

J. Potier, S. Menuel, N. Azaroual, E. Monflier, F. Hapiot
FULL PAPER
1
Figure 1. H NMR spectrum of dimer 1 in D2O at 25 °C.
spectrum of 1 is displayed in Figure 1. Only one singlet pyranoside methylene group (H-6 protons) attached to the
could be detected in the triazolyl proton region (ca. 8 ppm) triazolyl cycle. Accordingly, 100% of dimer 1 was in a sym-
indicative of a single dimer structure. However, the numer- metrical conformation.
ous signals observed between 4.5 and 5.0 ppm were clearly
The 1H NMR spectrum of 2 is displayed in Figure 2.
illustrative of the magnetic inequivalence of the H-6 protons Among the several signals detected in the 7–8.5 ppm region,
and suggested that the CD was not as rigid as first antici- only one corresponded to triazolyl protons (δ = 8.30 ppm),
pated. It seemed that the CD attempted to rotate but the with the others being attributed to aromatic protons. From
shortness of the linker and the hindrance of the bulky CD the triazolyl singlet, we deduced that all the triazolyl pro-
structure prevented the inversion phenomenon in this case. tons were in the same chemical environment. To determine
A 2D NOESY experiment confirmed the absence of inter- whether the two CD cavities were unaffected or reversed,
action between the linker and the CD cavities. Indeed, no a 2D NOESY experiment was performed. The absence of
cross-peaks could be detected between the triazolyl protons correlations clearly indicated that the β-CD cavities of 2
and the inner CD protons. Additionally, no nuclear Over- remained untouched. Thus, the cavities contained neither
hauser effect (NOE) could be observed between the triaz- the triazolyl moieties nor the phenyl ring. Accordingly, as
olyl protons H-8 and either the β-CD protons or the gluco- observed for 1, dimer 2 adopted a symmetrical conforma-
1
Figure 2. H NMR spectrum of dimer 2 in D2O at 25 °C.
1550
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 1547–1556

Inversion Phenomenon in β-Cyclodextrin Dimers
1
tion. However, here again, the inequivalent resonances de-
The H NMR spectrum of the unsymmetrical dimer 6
tected for the CH2–N pattern in the 4.50–5.00 ppm range displayed more intricate resonance patterns (Figure 4). In
were indicative of local disorder around the substituted the 8.2–6.8 ppm region, both singlets and doublets were de-
glucopyranoside unit.
tected. Using 1H-13C HSQC and HMBC, the most de-
Under the same conditions, 3 behaved in a similar way shielded singlets [δ = 8.18 (H-8) and 8.04 (H-12) ppm] were
to 2. In addition to the singlet at δ = 8.30 ppm characteristic attributed to triazolyl protons in a dimer conformation for
of the triazolyl protons (Figure 3), the singlet at δ = which both CD cavities were unaffected by the inversion
7.83 ppm reinforced the idea of a symmetrical structure be- phenomenon. The existence of a nonreversed conformation
cause the four aromatic protons were equivalent. Here was confirmed by the singlet at δ = 5.21 ppm, indicative of
again, no cross-peak between aromatic and CD protons two magnetically equivalent H-11NR protons (as observed
could be detected in the 2D NOESY spectrum of 3 in D2O. for H-9 in dimer 1). An NOE revealed dipolar interactions
The inequivalent resonances detected for the methylene between H-8NR and both H-6NR and H-9NR (henceforth,
protons linking the β-CD and the triazolyl cycle still sug- the following subscript abbreviations will be used: NR: non-
gested that the substituted glucopyranose unit was rather reversed; R: reversed; RI: reversed and included). Con-
flexible and possessed some degrees of freedom.
versely, the singlets at δ = 7.90 and 7.24 ppm were attrib-
1
Figure 3. H NMR spectrum of dimer 3 in D2O at 25 °C.
1
Figure 4. H NMR spectrum of dimer 6 in D2O at 25 °C (NR: nonreversed; R: reversed; RI: reversed and included).
Eur. J. Org. Chem. 2014, 1547–1556
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1551

J. Potier, S. Menuel, N. Azaroual, E. Monflier, F. Hapiot
FULL PAPER
1
uted by H-13C HSQC and HMBC to the triazolyl protons H-8NR and both H-6NR and H-9NR. The existence of two
H-8R and H-12RI, respectively, in a reversed conformation. additional peaks unambiguously demonstrated that one
Additionally, cross-peaks could be detected in the NOESY CD in two was able to undergo the inversion phenomenon
spectrum between H-12RI and the inner CD protons H-3 (Figure 5). Only one additional peak would have been ob-
and H-5 (see the Supporting Information). Aromatic pro- served for 100% symmetrical conformations with both their
tons were also illustrative of two different populations of CD reversed. The existence of a fully reversed stable confor-
dimers because two sets of doublets were detected in the mation of dimer 8 in D2O could be ruled out at this stage
7.7–6.8 ppm region. Whereas the doublets at 7.69 (H-9NR
)
because of the equal value of integrations of the two reso-
and 7.11 (H-10NR) ppm did not show any correlations with nances at δ = 7.93 and 7.87 ppm, indicative of two different
the CD protons in the NOESY spectrum, the doublets at triazolyl protons in an unsymmetrical conformation for
7.58 (H-9RI) and 6.88 (H-10RI) ppm exhibited intense cross- which one CD was reversed and the second nonreversed. If
peaks with the inner CD protons H-3 and H-5, indicative present, the fully reversed conformation existed in D2O in
of the inclusion of the phenyl moiety within the CD cavity. a very small proportion that was below the detection limits
Dipolar contacts between H-10RI and H-12RI were illustra- of the NMR measurements. The absence of NOE between
tive of the distorted structure of the spacer within the CD H-8NR or H-8R and the CD protons clearly indicated that
cavity. Moreover, the two doublets at δ = 5.27 and 5.38 ppm the ethoxylated fragment was included in the reversed CD
were characteristic of two magnetically inequivalent methyl- cavity. The ratio of symmetrical and unsymmetrical confor-
ene protons H-11R in a dimer structure in which one CD mations was given by integration of the triazolyl signals as
out of two was reversed. Protons H-11R became inequiva- 62% symmetric vs. 38% unsymmetrical conformations.
1
lent, probably due to the chiral environment exhibited by
The H NMR spectrum of dimer 10 was interpreted by
the close CD cavity and the torsional stress of the molecule detailed analysis of the aromatic and triazolyl resonances
required to bring the aromatic cycle into the CD cavity. (Figure 6). Three distinct signals were identified for the tria-
1
Note that the significant difference in chemical shifts Δδ zolyl protons by H-13C HSQC and HMBC (see the Sup-
between H-5R or H-5NR and H-5R indirectly confirmed the porting Information). Interestingly, the proportion of non-
presence of the inversed conformation. The relative pro- reversed conformations steadily decreased with time. The
portions of symmetrical and unsymmetrical conformations peak at δ = 7.98 ppm was assigned to two triazolyl protons
were calculated from integration of the two singlets at δ = in a nonreversed, symmetrical structure (H-8NR). The sing-
8.18 and 7.90 ppm; thus, 23% symmetrical (nonreversed) let at δ = 7.76 ppm was identified as a triazolyl proton (H-
and 77% unsymmetrical (partly reversed) dimer were iden-
tified.
8
RI) included in a CD cavity (NOE with H-6RI and inner
CD protons H-3 and H-5). The third singlet at δ =
Dimer 8 also displayed two different conformations in 7.69 ppm was assigned to a triazolyl proton (H-8R) that was
D2O. In fact, three triazolyl protons were detected in the not included in the CD cavity (NOE with H-6R). The exis-
8.0–7.8 ppm region (Figure 5). The most deshielded signal tence of two different populations of dimers was corrobo-
at δ = 7.97 ppm was identified as triazolyl protons H-8NR rated by the six doublets observed between 7.5 and
in a symmetrical conformation. As observed above for 6, 6.60 ppm, two of which were overlapped (δ = 7.40 ppm).
NOE experiments revealed dipolar interactions between The doublets at δ = 7.44 and 6.84 ppm were attributed to
1
Figure 5. H NMR spectrum of dimer 8 in D2O at 25 °C (NR: nonreversed; R: reversed).
1552
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 1547–1556

Inversion Phenomenon in β-Cyclodextrin Dimers
1
Figure 6. H NMR spectrum of dimer 10 in D2O at 25 °C (NR: nonreversed; R: reversed; RI: reversed and included).
aromatic protons (H-9NR and H-10NR) in a nonreversed conformation could not be obtained by inclusion of the
architecture (NOE with H-8NR and H-9NR, respectively). spacer within the CD cavity. Even though the spacer could
One of the doublets at δ = 7.39 ppm and the doublet at δ enter the CD cavity by the secondary face, the resulting
= 7.08 ppm were attributed to aromatic protons (H-9R and close proximity of the two CDs precluded the existence of
H-10R) not included in the CD cavity of a dimer for which a stable reversed conformation. Thus, one of the CDs at-
at least one of the CD cavity tumbled. Finally, the remain- tempted to tumble but the short spacer and the bulk of the
ing doublets (7.40 and 6.67 ppm) were assigned to aromatic CD prevented it from doing so. Although less congested
protons (H-9RI and H-10RI) included in the CD cavity. In than 2, dimer 3 also yielded 100% symmetrical conforma-
this case, 43% symmetrical vs. 57% unsymmetrical confor- tion. However, the local disorder detected around the sub-
mations were identified (Figure 6). Dipolar interactions stituted glucopyranoside unit of these dimers (multiple res-
were detected between H-8RI, H-9RI and H-10RI and the onances detected for the CH2–N pattern) clearly indicated
CD protons H-3 and H-5. In contrast to H-11R and H- that a partial tumbling of the CD remained possible. Dimer
11NR (singlets), the two H-11RI protons were magnetically 6 showed that the limit of the inversion process could be
inequivalent due to their proximity to the CD cavity, as easily overcome because an additional flexible group (CH2–
clearly shown in the HSQC spectrum (see the Supporting O) between the phenyl ring and one of the triazolyl cycles
Information). Accordingly, two different populations have allowed for the inversion process to take place. The distance
been detected for dimer 10, with 43% symmetrical and 57% between the nitrogen atoms of the N–C6 fragments con-
unsymmetrical conformations (Figure 6).
nected to the CDs was estimated to be 18 Å in this case,
which is far greater than the depth of the CD cavity
(ca. 8 Å). Symmetrical and unsymmetrical conformations
were then observed, with the unsymmetrical conformation
Discussion
Based on the above analysis, several conclusions could predominating (77 vs. 23%). Upon the 360° rotation, H-
be drawn. First, the analysis of 1, 2 and 3 highlighted one RI, H-10RI, H-11RI and H-12RI were included within the
9
of the main features guiding the inversion phenomenon in CD cavity. The higher stability of the latter regarding the
CD dimers, namely the spacer length. Each of the com- “free cavities” dimer probably resulted from a conjunction
pounds displayed 100% symmetrical conformations, thus of two parameters. First, the hydrophobic spacer had a nat-
demonstrating that the inversion process was greatly disfa- ural tendency to enter the hydrophobic CD cavity (hydro-
vored. The higher flexibility of the spacer of 1 did not lead phobic effect). Second, hydrogen bonds between the pri-
to higher percentage of reversed dimer, suggesting that the mary face of the nonreversed CD could interact with the
inversion process was not correlated to the spacer rigidity secondary face of the reversed CD.
but rather to the distance between the CDs. A complete
The analysis of dimer 8 brought additional information
tumbling was inconceivable because of the shortness of the on the stability of CD dimers in water. In this case, the
triazolyl–methyloxymethyl–triazolyl or triazolyl–phenyl–tri- symmetrical “free cavities” dimer predominated (62 vs.
azolyl spacer. Actually, the distance between the nitrogen 38%). Although the polyethoxylated spacer was more flexi-
atoms of the N–C6 fragments connected to the CDs has ble and longer than the spacer of dimers 1, 2, 3 and 6, its
been estimated to be 8 Å for 1 and 9 Å for 2. Knowing more balanced hydrophobic/hydrophilic character disfa-
that the β-CD cavity is approximately 8 Å deep, a stable vored the inversion process. Thus, as expected, the chemical
Eur. J. Org. Chem. 2014, 1547–1556
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1553

J. Potier, S. Menuel, N. Azaroual, E. Monflier, F. Hapiot
FULL PAPER
acid in methanol. NMR spectra were recorded either with a Bruker
nature of the spacer was also of importance for the control
of the inversion process. Dimer 10 confirmed the NMR
spectroscopy results obtained with 6, although the relative
proportions of symmetrical and unsymmetrical conforma-
tions were much more balanced for 10 (43 vs. 57%, respec-
tively). One of the CD groups underwent the inversion phe-
nomenon, whereas the cavity of the second remained
“free”. Among the three different groups constituting the
spacer (triazolyl, phenyl, and ethoxylated chain), the CD
1
DRX300 spectrometer operating at 300 MHz for H nuclei and at
75 MHz for 13C nuclei, or with a Bruker Avance spectrometer op-
1
erating at 500 MHz for H and at 126 MHz for 13C nuclei with a
TXI probe at 295 K with a mixing time of 500 ms. CDCl3 (99.50%
isotopic purity), [D6]DMSO (99.80% isotopic purity) and D2O
(99.92% isotopic purity) were purchased from Euriso-Top. Mass
spectra were recorded with a MALDI-TOF/TOF Bruker Daltonics
Ultraflex II spectrometer in positive reflectron mode with 2,5-DHB
as the matrix. The distances between CDs within the CD-dimers
cavity encapsulated the more hydrophobic phenyl moiety, were estimated by using Avogadro, version 1.1.0 (Force field:
MMFF94s, Geometry Optimization by 500 steps with steepest de-
scent algorithm).
as clearly shown by NOESY experiments.
From the above considerations, it was possible to estab-
lish the essential criteria to be respected to avoid the inver-
sion phenomenon occurring in CD dimers. First, the in-
clusion of a spacer within a CD cavity depended more upon
its length than upon its rigidity. Second, the length of the
spacer should be as short as possible. Third, high spacer
hydrophilicity was required to favor the displacement of the
equilibrium towards conformations in which the CDs have
“free” cavities. Note that although the triazolyl group might
initiate the inversion process, it did not especially stabilize
the reversed conformation once the inversion process had
taken place.
(4-Hydroxyphenyl)(trimethylsilyl)ethyne (4): A solution of 4-iodo-
phenol (1.52 g, 6.9 mmol) and copper iodide (66 mg, 0.35 mmol,
0.05 equiv.) in pyrrolidine (100 mL) was treated with bis(triphenyl-
phosphine)palladium(II) dichloride (500 mg, 0.69 mmol, 0.1 equiv.)
The solution was degassed under nitrogen, (trimethylsilyl)ethyne
(1.01 g, 10.3 mmol, 1.5 equiv.) was added, and the mixture was
stirred under nitrogen for 5 h at 60 °C. The crude product was ex-
tracted with ethyl acetate and washed with a saturated aqueous
solution of sodium hydrogen carbonate. After evaporation of ethyl
acetate, the product was purified by chromatography (CH2Cl2) to
give 4 (1.26 g, 96%) as a white powder. Rf = 0.48 (CH2Cl2). 1H
NMR (300 MHz, [D6]DMSO, 25 °C): δ = 9.93 (s, 1 H, OH), 7.26
[d, J = 8.7 Hz, 2 H, CH(5)], 6.73 [d, J = 8.7 Hz, 2 H, CH(6)], 0.19
[s, 9 H, CH3(1)] ppm. 13C NMR (75 MHz, [D6]DMSO, 25 °C): δ =
158.2 (C4), 133.3 (C5), 115.6 (C6), 112.4 (C7), 106.08 (C3), 91.5
(C2), 0.1 (C1) ppm.
Conclusions
The present study sought to address the scope of the in-
version phenomenon on triazolyl-substituted β-CDs. It was
shown that the proportion of CDs undergoing a 360° rota-
tion mainly depended upon the nature of the spacer. Large
proportions of dimers for which only one out of two CD
cavities was available have been observed for β-CDs linked
together by long and/or hydrophobic spacers, thus limiting
the application of such CD dimers in aqueous media. Con-
1-Ethynyl-4-(prop-2-yn-1-yloxy)benzene (5): Compound 4 (700 mg,
3.6 mmol) and Cs2CO3 (1.76 g, 5.4 mmol, 1.5 equiv.) were dissolved
in acetonitrile (50 mL) and the solution was degassed under nitro-
gen. Propargyl bromide (80 wt.-% in toluene, 523 mg, 4.4 mmol)
was added and the resulting mixture was stirred overnight under
nitrogen at 85 °C. The crude product was extracted with ethyl acet-
ate and washed with a saturated aqueous solution of sodium hydro-
gen carbonate. After evaporation of ethyl acetate, the product was
versely, short spacers proved to be appropriate to ensure the purified by chromatography (heptane/CH2Cl2, 5:5 to 0:5) to give 5
(420 mg, 75%) as a white solid. Rf = 0.52 (CH2Cl2). 1H NMR
(300 MHz, [D6]DMSO, 25 °C): δ = 7.41 [d, J = 8.7 Hz, 2 H,
CH(6)], 6.96 [d, J = 8.7 Hz, 2 H, CH(5)], 4.81 [d, J = 1.8 Hz, 2 H,
existence in water of CD dimers having “free” cavities. As
such, we unambiguously demonstrated that CD dimers
composed of two β-CDs linked together by a short triazo-
lyl–methyloxymethyl–triazolyl or triazolyl–phenyl–triazolyl
CH2(3)], 4.02 [s, 1 H, CH(9)], 3.58 [t, J = 1.8 Hz, 1 H, CH(1)] ppm.
13C NMR (75 MHz, [D6]DMSO, 25 °C): δ = 157.9 (C4), 133.6
spacer displayed 100% symmetrical conformation in water.
(C6), 115.6 (C5), 114.9 (C7), 83.8 (C8), 79.9 (C9), 79.4 (C2), 79.0
This is of crucial importance for catalytic systems in which
(C1), 55.9 (C3) ppm.
multivalency and cooperativity are crucial.[20] More gen-
4,7,10,13,16,19-Hexaoxodocosa-1,21-diyne (7): Pentaethylene glycol
erally, any system based on β-CD dimers for which molecu-
(300 mg, 1.3 mmol) and Cs2CO3 (1.26 g, 3.9 mmol) were dissolved
lar recognition processes are involved can benefit from the
results of this study. Our current investigations are focused
on the use of these “free cavities” CD dimers in aqueous
catalysis.
in acetonitrile (30 mL) and degassed under nitrogen. Propargyl
bromide (80 wt.-% in toluene, 345 mg, 2.9 mmol) was added and
the resulting mixture was stirred overnight under nitrogen at 85 °C.
The crude product was extracted with ethyl acetate and washed
with saturated aqueous sodium hydrogen carbonate. After evapora-
tion of ethyl acetate, the product was purified by chromatography
(ethyl acetate/CH2Cl2, 7:3) to give 7 (59 mg, 14%) as a white pow-
Experimental Section
1
der. Rf = 0.53 (EtOAc/CH2Cl2, 7:3). H NMR (300 MHz, CDCl3,
General Remarks: All chemicals were purchased from Acros or
Aldrich Chemicals in their highest purity. All solvents were used as
supplied without further purification. Distilled water was used in
all experiments. Column chromatography was carried out by using
the “flash” method with 40 μm silica. Analytical thin-layer
chromatography (TLC) was performed on E. Merck aluminum-
backed silica gel (Silica Gel F254). Compounds were identified with
UV light (254 nm) and/or by staining with a solution of sulfuric
25 °C): δ = 4.20 [d, J = 2.4 Hz, 4 H, CH2(3)], 3.65–3.69 [m, 20 H,
CH2(4 and 5)], 2.42 [t, J = 2.4 Hz, 2 H, CH(1)] ppm. 13C NMR
(75 MHz, CDCl3, 25 °C): δ = 58.5 (C3), 69.2 (C2), 70.6–70.5 (C4
and C5), 74.6 (C1), 83.7 (C2), 114.3 (C6), 114.6 (C5), 133.6 (C4),
159.2 (C3) ppm.
Bis[2-(4-ethynylphenoxy)ethoxy]ethane (9): Bis(2-chloroethyl) ether
(984 mg, 5.25 mmol), 4 (2.0 g, 10.5 mmol, 2 equiv.) and cesium
1554
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 1547–1556

Inversion Phenomenon in β-Cyclodextrin Dimers
carbonate (6.8 g, 21 mmol, 4 equiv.) were dissolved in acetonitrile
(100 mL) and the resulting mixture was stirred overnight under ni-
trogen at 85 °C. The crude product was extracted with ethyl acetate
and washed with saturated aqueous sodium hydrogen carbonate.
After evaporation of ethyl acetate, the product was purified by
chromatography (heptane/dichloromethane, 5:5 to 0:5) to give 9
(950 mg, 37%) as a pale-yellow powder. Rf = 0.52 (CH2Cl2). 1H
NMR (300 MHz, CDCl3, 25 °C): δ = 7.40 [d, J = 8.7 Hz, 2 H,
CH(4)], 6.83 [d, J = 8.7 Hz, 2 H, CH(5)], 4.11 [t, J = 4.5 Hz, 4 H,
NMR (500 MHz, D2O, 25 °C): δ = 7.91 [s, 1 H, CH(8NR)], 8.05 [s,
1 H, CH(12NR)], 7.91 [s, 1 H, CH(8R)], 7.69 [d, J = 8 Hz, 2 H,
CH(9NR)], 7.58 [d, J = 8 Hz, 2 H, CH(9R)], 7.25 [s, 1 H, CH(12RI)],
7.11 [d, J = 8 Hz, 2 H, CH(10NR)], 6.88 [d, J = 8 Hz, 2 H,
CH(10R)], 5.40 [d, J = 14 Hz, 1 H, 1/2 CH2(11R)], 5.29 [d, J =
14 Hz, 1 H, 1/2 CH2(11R)], 5.23 [s, 14 H, CH2(11NR)], 5.11–4.81
(m, 14 H, H-1), 4.68–5.06 (m, 2 H, H6R), 4.95–4.54 (m, 4 H,
H6NR), 4.67–4.39 (m, 2 H, H6RI), 4.15 (t, J = 9 Hz, 1 H, H5R
substituted subunit), 4.11–2.61 (m, 159 H, H-2, H-3, H-4, H-5, H-
CH2(7)], 3.85 [t, J = 4.5 Hz, 4 H, CH2(8)], 2.99 [s, 2 H, CH(1)], 6) ppm. 13C NMR (126 MHz, D2O, 25 °C): δ = 128–125 (C9),
3.74 [s, 4 H, CH2(9)] ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ
= 159.2 (C3), 133.6 (C4), 114.6 (C5), 114.3 (C6), 83.7 (C2), 75.9
(C1), 71.0 (C9), 69.6 (C8), 67.4 (C7) ppm.
125.0 (C12), 122.5 (C8), 118.6 (C10), 115.9 (C10), 104–100 (C1),
84–80 (C4, C5), 75–70 (C2, C3), 62.6 (C11), 61–58 (C6 nonsubsti-
tuted subunit and C6 substituted subunit), 52.1 (C6 substituted
subunit) ppm. HRMS: calcd. for [C95H146N6O69 + Na]+ 2498.80;
found 2497.98.
General Procedure for Copper-Catalyzed Azide Alkyne Cycload-
dition (CuAAC): Mono-6-azido-β-CD (2 equiv.) and hydrated cop-
per sulfate (4 equiv.) were added to a solution of dialkynyl deriva-
tive (1 equiv.) in DMSO (50 mL). After subsequent dropwise ad-
dition of a freshly prepared solution of sodium ascorbate (8 equiv.)
dissolved in H2O (10 mL), the solution was stirred at room temp.
for 24 h. The crude product was then precipitated with acetone to
give a green solid. After filtration, the residue was purified by col-
umn chromatography (water/ammoniac, 80:20) to remove the cop-
per salts. After solvent evaporation, the product was precipitation
in a water/acetone mixture. The solid was filtered and purified by
chromatography (CH3CN/H2O, 7:3) to give the pure dimer (80–
85% yield) as a white powder.
1,18-Bis[1-(6A-deoxy-β-
D-cyclodextrin)-1H-1,2,3-triazol-4-
yl]-2,5,8,11,14,17-hexaoxaoctadecane (8): 1H NMR (500 MHz,
D2O, 25 °C): δ = 7.97 [s, 2 H, CH(8NR)], 7.93 and 7.88 [s, 2 H,
CH(8R)], 5.11–4.85 (m, 14 H, H-1), 4.96–4.62 (m, 2 H, H6RI), 4.52–
4.97 (m, 2 H, H6R), 4.54–4.93 (m, 4 H, H6NR), 4.61 [m, 8 H,
CH2(9)], 4.11 (t, J = 9 Hz, 2 H, H5NR), between 3.98–3.22 (m, 155
H, H-2, H-3, H-4, H-5, H-6), 3.06 (d, J = 12 Hz, 2 H, H-6), 2.71
(d, J = 12 Hz, 2 H, H-6) ppm. 13C NMR (126 MHz, D2O, 25 °C):
δ = 126–123 (C8), 104–100 (C1), 84–80 (C4, C5), 75–70 (C2, C3),
72–68 (C10, C11, C12, C13, C14), 63.7 (C9), 61–58 (C6 nonsubsti-
tuted subunit and C6 substituted subunit), 52.4 (C6 substituted
subunit) ppm. HRMS: calcd. for [C100H164N6O74 + Na]+ 2656.92;
found 2656.16.
Bis[1-(6A-deoxy-β-
D-cyclodextrin)-1H-1,2,3-triazol-4-ylmethyl]
Ether (1): 1H NMR (500 MHz, D2O, 25 °C): δ = 7.99 [s, 2 H,
CH(8)], between 4.85 and 5.10 (m, 14 H, H-1), 4.54–4.93 (m, 4 H,
H-6 substituted subunit), 4.65 [s, 4 H, CH2(9)], 4.11 (t, J = 9 Hz,
2 H, H-5 substituted subunit), 3.97–3.35 (m, 74 H, H-2, H-3, H-4,
H-5, H-6), 3.07 (d, J = 12 Hz, 2 H, H-6), 2.73 (d, J = 12 Hz, 2 H,
H-6) ppm. 13C NMR (126 MHz, D2O, 25 °C): δ = 127.3 (C8), 104–
100 (C1), 84–80 (C4, C5), 75–70 (C2, C3), 64.6 (C9), 63–60 (C6
nonsubstituted subunit and C6 substituted subunit), 51.7 (C6 sub-
stituted subunit) ppm. HRMS: calcd. for [C90H144N6O69 + Na]+
2435.78; found 2435.62.
1,2-Bis(2-{4-[1-(6A-deoxy-β-
D-cyclodextrin)-1H-1,2,3-triazol-4-
1
yl]phenoxyl}ethoxy)ethane (10): H NMR (500 MHz, D2O, 25 °C):
δ = 7.98 [s, 2 H, CH(8NR)], 7.75 [s, 1 H, CH(8RI)], 7.69 [s, 1 H,
CH(8R)], 7.44 [d, J = 8 Hz, 4 H, CH(9NR)], 7.40 [d, J = 8 Hz, 2 H,
CH(9RI)], 7.39 [d, J = 8 Hz, 2 H, CH(9R)], 7.08 [d, J = 8 Hz, 2 H,
CH(10R)], 6.84 [d, J = 8 Hz, 4 H, CH(10NR)], 6.67 [d, J = 8 Hz, 2
H, CH(10RI)], 5.14–4.78 (m, 14 H, H-1), 5.02–4.45 (m, 2 H, H6R),
4.9 (m, 2 H, H6RI), 4.88–4.39 (m, 4 H, H6NR), 4.3 [br. s, 2 H,
CH2(11R)], 4.15 (m, 1 H, H5RI), 4.02 [br. s, 4 H, CH2(11NR)], be-
1,3-Bis[1-(6A-deoxy-β-
D-cyclodextrin)-1H-1,2,3-triazol-4-yl]benz-
tween 2.58 and 4.00 (m, 181 H, H-2, H-3, H-4, H-5, H-6, H-11RI
,
H-12, H-13) ppm. 13C NMR (126 MHz, D2O, 25 °C): δ = 128–126
(C9), 122–118 (C8), 117–113 (C10), 104–101 (C1), 85–80 (C4, C5),
75–70 (C2, C3), 70–67 (C11, C12), 61–58 (C6 nonsubstituted sub-
unit and C6 substituted subunit), 53–50 (C6 substituted sub-
unit) ppm. HRMS: calcd. for [C106H160N6O72 + Na]+ 2692.90;
found 2692.21.
ene (2): 1H NMR (500 MHz, D2O, 25 °C): δ = 8.32 [s, 2 H, CH(8)],
8.14 [s, 1 H, CH(13)], 7.73 [d, J = 8 Hz, 2 H, CH(11)], 7.53 [t, J =
8 Hz, 1 H, CH(12)], 5.13–4.81 (m, 14 H, H-1), 5.02–4.56 (m, 4 H,
H-6 substituted subunit), 4.08 (t, J = 9 Hz, 2 H, H-5 substituted
subunit), 4.00–3.20 (m, 74 H, H-2, H-3, H-4, H-5, H-6), 3.00 (d, J
= 12 Hz, 2 H, H-6), 2.79 (d, J = 12 Hz, 2 H, H-6) ppm. 13C NMR
(126 MHz, D2O, 25 °C): δ = 130.5 (C12), 124.3 (C11), 123.1 (C8),
121.9 (C13), 102–98 (C1), 83–79 (C4, C5), 75–70 (C2, C3), 61–58
(C6 nonsubstituted subunit and C6 substituted subunit), 51.6 (C6
Supporting Information (see footnote on the first page of this arti-
cle): NMR spectra.
substituted subunit) ppm. HRMS: calcd. for [C94H144N6O68
+
Na]+ 2468.79; found 2468.52.
Acknowledgments
1,4-Bis[1-(6A-deoxy-β-
D-cyclodextrin)-1H-1,2,3-triazol-4-yl]benz-
J. P. is grateful to the Région Nord-Pas-Calais and the Centre
National de la Recherche Scientifique (CNRS) for financial sup-
port (2010–2013). The authors are grateful to Roquette Frères
(Lestrem, France) for a gift of β-cyclodextrin.
ene (3): 1H NMR (500 MHz, D2O, 25 °C): δ = 8.30 [s, 2 H, CH(8)],
7.83 [s, 4 H, CH(9)], 5.13–4.81 (m, 14 H, H-1), 5.00–4.55 (m, 4 H,
H-6 substituted subunit), 4.08 (t, J = 9 Hz, 2 H, H-5 substituted
subunit), 4.02–3.20 (m, 74 H, H-2, H-3, H-4, H-5, H-6), 3.00 (d, J
= 12 Hz, 2 H, H-6), 2.74 (d, J = 12 Hz, 2 H, H-6) ppm. 13C NMR
(126 MHz, D2O, 25 °C): δ = 126.4 (C Ar), 123.2 (C8), 105–100
(C1), 84–80 (C4, C5), 75–70 (C2, C3), 62–58 (C6 nonsubstituted
subunit and C6 substituted subunit), 51.4 (C6 substituted sub-
unit) ppm. HRMS: calcd. for [C94H144N6O68 + Na]+ 2468.79;
found 2468.79.
[1] a) J. Szejtli, T. Osa, in: Comprehensive Supramolecular Chemis-
try: Cyclodextrins (Eds.: J. L. Atwood, J. E. D. Davies, D. D.
MacNicol, F. Vögtle), Pergamon, 1999, vol. 3; b) H. Dodziuk,
in: Cyclodextrins and Their Complexes, Wiley-VCH, Weinheim,
Germany, 2006.
[2] a) E. Bilensoy, in: Cyclodextrins in Pharmaceutics, Cosmetics
and Biomedecine: Current and Future Industrial Applications,
John Wiley & Sons, Hoboken, New Jersey, 2011; b) A. L. Laza-
1-(6A-Deoxy-β-
D D-cyclodex-
-cyclodextrin)-4-(4-{[1-(6A-deoxy-β-
1
trin)-1H-1,2,3-triazol-4-yl]methoxy}phenyl)-1H-1,2,3-triazol (6): H
Eur. J. Org. Chem. 2014, 1547–1556
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1555

J. Potier, S. Menuel, N. Azaroual, E. Monflier, F. Hapiot
FULL PAPER
Knoerr, R. Gref, P. Couvreur, J. Drug Targeting 2010, 18, 645–
656.
[3] a) G. Cravotto, A. Binello, E. Baranelli, P. Carraro, F. Trotta,
Curr. Nutr. Food Sci. 2006, 2, 343–350; b) G. Astray, C. Gonza-
lez-Barreiro, J. C. Mejuto, R. Rial-Otero, J. Simal-Gándara,
Food Hydrocolloids 2009, 23, 1631–1640.
[4] V. M. Mirsky, A. Yatsimirsky, in: Artificial Receptors for Chem-
ical Sensors, Wiley-VCH, Weinheim, Germany, 2011.
[5] a) R. Breslow, in: Artificial Enzymes, Wiley-VCH, Weinheim,
Germany, 2005; b) R. Breslow, S. D. Dong, Chem. Rev. 1998,
98, 1997–2011.
[6] O. A. Shpigun, I. A. Ananieva, N. Y. Budanova, E. N. Shapov-
alova, Russ. Chem. Rev. 2003, 72, 1035–1054.
[7] a) F. Hapiot, A. Ponchel, S. Tilloy, E. Monflier, C. R. Chim.
2011, 14, 149–166; b) H. Bricout, F. Hapiot, A. Ponchel, S.
Tilloy, E. Monflier, Curr. Org. Chem. 2010, 14, 1296–1307; c)
H. Bricout, F. Hapiot, A. Ponchel, S. Tilloy, E. Monflier, Sus-
tainability 2009, 1, 924–945.
[8] J. Szejtli, Chem. Rev. 1998, 98, 1743–1753.
[9] A. R. Khan, P. Forgo, K. J. Stine, V. T. D’Souza, Chem. Rev.
1998, 98, 1977–1996.
[11]
[12]
[13]
W. Cai, T. Sun, X. Shao, C. Chipot, Phys. Chem. Chem. Phys.
2008, 10, 3236–3243.
F. Hapiot, L. Leclercq, N. Azaroual, S. Fourmentin, S. Tilloy,
E. Monflier, Curr. Org. Synth. 2008, 5, 162–172.
F. Hapiot, H. Bricout, S. Tilloy, E. Monflier, Eur. J. Inorg.
Chem. 2012, 1571–1578.
R. Nishiyabu, K. Kano, Eur. J. Org. Chem. 2004, 4985–4988.
Y. Liu, C.-F. Ke, H.-Y. Zhang, J. Cui, F. Ding, J. Am. Chem.
Soc. 2008, 130, 600–605.
[14]
[15]
[16] K. Yamauchi, A. Miyawaki, Y. Takashima, H. Yamaguchi, A.
Harada, Org. Lett. 2010, 12, 1284–1286.
[17] K. Yamauchi, A. Miyawaki, Y. Takashima, H. Yamaguchi, A.
Harada, J. Org. Chem. 2010, 75, 1040–1046.
[18] S. Menuel, N. Azaroual, D. Landy, N. Six, F. Hapiot, E.
Monflier, Chem. Eur. J. 2011, 17, 3949–3955.
[19] V. Legros, C. Vanhaverbeke, F. Souard, C. Len, J. Désiré, Eur.
J. Org. Chem. 2013, 2583–2590.
[20] J. Potier, S. Menuel, D. Fournier, S. Fourmentin, P. Woisel, E.
Monflier, F. Hapiot, ACS Catal. 2012, 2, 1417–1420.
Received: November 8, 2013
[10] Y. Cui, C. Wang, J. Mao, Y. Yu, J. Chem. Technol. Biotechnol.
2010, 85, 248–251.
Published Online: December 19, 2013
1556
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 1547–1556