Synthesis of monofacially functionalized cyclodextrins bearing amino pendent groups

Article abstract of DOI:10.1021/jo9711549

Derivatives of the cyclodextrins, αCD, βCD, and ΥCD, in which all primary hydroxyls are substituted by amine pendant groups, may be synthesized efficiently from the per-6-bromo-6-deoxy-CD derivatives by direct reaction with amines. These ACD derivatives, which bear six, seven, or eight amine pendent groups, represent interesting biomimetic receptors and catalysts. The synthetic strategy relies on quantitative transformation and efficient purification as is demonstrated by preparation of 11 homogeneous ACD derivatives. The limitations of the synthesis and potential adaptations are illustrated by the synthesis of several more ACD derivatives to >95% purity. A synthetic route to a CD persubstituted with primary amine functionalities at the primary face, per-6-bromo-6-deoxy-CD, yields an alternative reagent to the simple per-6-bromo-6-deoxy-CD, which is more suitable for further synthetic transformations. The synthetic strategy is further adapted to preparation of a prototypical (6 + 1)-ACD derivative in which one primary position is substituted with a sulfide group and the remaining six primary face positions are substituted with amine pendent groups.

Full text of DOI:10.1021/jo9711549

8760
J . Org. Chem. 1997, 62, 8760-8766
Syn th esis of Mon ofa cia lly F u n ction a lized Cyclod extr in s Bea r in g
Am in o P en d en t Gr ou p s
Dragos Vizitiu, Caroline S. Walkinshaw, Borio I. Gorin, and Gregory R. J . Thatcher*
Department of Chemistry, Queen’s University, Kingston, Ontario K7L 3N6, Canada
Received J une 25, 1997X
Derivatives of the cyclodextrins, RCD, âCD, and γCD, in which all primary hydroxyls are substituted
by amine pendant groups, may be synthesized efficiently from the per-6-bromo-6-deoxy-CD
derivatives by direct reaction with amines. These ACD derivatives, which bear six, seven, or eight
amine pendent groups, represent interesting biomimetic receptors and catalysts. The synthetic
strategy relies on quantitative transformation and efficient purification as is demonstrated by
preparation of 11 homogeneous ACD derivatives. The limitations of the synthesis and potential
adaptations are illustrated by the synthesis of several more ACD derivatives to >95% purity. A
synthetic route to a CD persubstituted with primary amine functionalities at the primary face,
per-6-(aminomethyl)-6-deoxy-CD, yields an alternative reagent to the simple per-6-amino-6-deoxy-
CD, which is more suitable for further synthetic transformations. The synthetic strategy is further
adapted to preparation of a prototypical (6 + 1)-ACD derivative in which one primary position is
substituted with a sulfide group and the remaining six primary face positions are substituted with
amine pendent groups.
In tr od u ction
derivatives have been enduring features of the field of
enzyme mimicry.7 CDs remain attractive for study as
enzyme mimics because of the incorporation of substrate
binding and product release in aqueous solution within
the catalytic mechanism.8
The cyclodextrins RCD, âCD, and γCD are naturally
occurring cyclomaltooligosaccharides containing six, seven,
and eight R-(1f4)-D-glucopyranosyl rings, respectively.1
The use of synthetic CD derivatives as molecular scaf-
folds is of interest since the primary (6′) and secondary
(2′ and 3′) hydroxyl groups may be used as points of
functionalization.2 However, the hydrophobic interior of
these torus-shaped molecules has provided the focus of
much of the chemistry and applications of cyclodextrins.
The cavity binds hydrophobic, organic molecules of ap-
propriate size, yielding inclusion complexes.3 The po-
tential utility of this inclusion phenomenon includes
solubilization, encapsulation, and transport of small
hydrophobic molecules including toxins and drugs.4 The
facile ionization of the hydroxyl groups on the secondary
face annulus provides a nucleophilic site that has been
seen as analogous to that of the serine proteases.5 Thus,
early studies proposed that substrate binding within the
cavity, combined with the reactivity of the secondary
annulus, made cyclodextrins ideal candidates for enzyme
models.6 In numerous cases, cyclodextrins provide ac-
celeration of reaction, in contrast to true catalysis;
however, both natural cyclodextrins and synthetic CD
Derivatization at one or two of the 6′ carbons yields a
mono- or difunctionalized primary face, respectively. This
procedure is facile, although overall yields are often
unsatisfactory.9 Monofunctionalization of the secondary
face is less facile and has been limited to a much smaller
selection of pendent groups.10 Many of these synthetic
CD derivatives have been reported as enzyme models,
utilizing in some cases the secondary face hydroxyls, and
in other examples, the pendent, prosthetic group as the
catalytic or reactive functionality.6 Procedures for mono-
facial derivatization have been developed for both the
primary face (all 6′ positions) and the secondary face (all
2′ and 3′ positions) of CD. Procedures for CD “perderiva-
tization” yield both homogeneous and heterogeneous,
modified cyclodextrins.11 In the heterogeneous examples,
the CD derivative is described by the degree of substitu-
tion. In these polysubstituted CD derivatives, the pen-
dent groups have been utilized primarily to modify the
bulk properties of the CD: to extend the hydrophobic
cavity in order to influence binding and encapsulation
and, most importantly, to increase the poor aqueous
solubility of natural CD itself.
* To whom correspondence should be addressed. Fax: (613) 545-
6669. E-mail: thatcher@chem.queensu.ca.
X Abstract published in Advance ACS Abstracts, November 1, 1997.
(1) Szetli, J . Cyclodextrins and their inclusion complexes; Akademiai
Kiado: Budapest, 1982. Wenz, G. Angew. Chem., Int. Ed. Engl. 1994,
33, 803.
Monofacially functionalized CD derivatives persubsti-
tuted on the primary face with six, seven, or eight
(2) (a) Hanessian, S.; Benalil, A.; Laferriere, C. J . Org. Chem. 1995,
60, 4786. (b) Ashton, P. R.; Koniger, R.; Stoddardt, J . F.; Alker, D.;
Harding, V. D. J . Org. Chem. 1996, 61, 903. (c) Yoon, J .; Hong, S.;
Martin, K. A.; Czarnik, A. W. J . Org. Chem. 1995, 60, 2792.
(3) For example: Rudiger, V.; Eliseev, A.; Simova, S.; Schneider,
H. J .; Blandamer, M. J .; Cullis, P. M.; Meyer, A. J . J . Chem. Soc.,
Perkin Trans. 2 1996, 2119.
(4) Komiyama, M. In Comprehensive Supramolecular Chemistry;
Lehn, J . M., Ed.; Pergamon: New York, 1996; Vol. 3 and references
therein. Eastburn, S. D.; Tao, B. Y. Biotech. Adv. 1994, 12, 325. Sjetli,
J . Pharm Technol. Int. 1991, 3, 15. Schipper, N. G. M.; Hermens, W.
A. J . J .; Romeijn, S. G.; Verhoef, J .; Merkus, F. W. H. M. Int. J . Pharm.
1990, 64, 61.
(6) Breslow, R. Acc. Chem. Res. 1995, 28, 146. Tee, O. S.; Breslow,
R.; Trainor, G.; Ueno, A. J . Am. Chem. Soc. 1983, 105, 2739. Bozzi,
M.; Hoeven, J . J .; Gadosy, T. A. J . Am. Chem. Soc. 1993, 115, 8990.
Tee, O. S. Adv. Phys. Org. Chem. 1994, 29, 1.
(7) Breslow, R.; Duggan, P. J .; Light, J . P. J . Am. Chem. Soc. 1992,
114, 3982. Anslyn, E.; Breslow, R. J . Am. Chem. Soc. 1989, 111, 893.
Breslow, R.; Czarnik, A. W. J . Am. Chem. Soc. 1983, 105, 1390.
(8) Kirby, A. J . Adv. Phys. Org. Chem. 1980, 17, 183.
(9) Tabushi, I.; Shimokawa, K.; Shimizu, N.; Shirikata, H.; Fujita,
J . Am. Chem. Soc. 1976, 98, 7855. Tabushi, I.; Shimokawa, K.; Fujita,
K. Tetrahedron Lett. 1977, 18, 1527. Yoon, J .; Hong, S.; Martin, K. A.;
Czarnik, A. W. J . Org. Chem. 1995, 60, 2792. Breslow, R.; Doherty, J .
B.; Guillot, G.; Lipsey, C. J . Am. Chem. Soc. 1978, 100, 3227.
(10) Murakami, T.; Harata, K.; Morimoto, S. Tetrahedron Lett. 1986.
Ueno, A.; Breslow, R. Tetrahedron Lett. 1982, 23, 3451. Rong, D.;
D’Souza, V. T. Tetrahedron Lett. 1990, 31, 4275.
(5) D’Souza, V. T.; Bender, M. L. Acc. Chem. Res. 1987, 20, 146.
Palmer, D. R. J .; Buncel, E.; Thatcher, G. R. J . J . Org. Chem. 1994,
59, 5286.
S0022-3263(97)01154-7 CCC: $14.00 © 1997 American Chemical Society

Synthesis of Monofacially Functionalized Cyclohextrins
J . Org. Chem., Vol. 62, No. 25, 1997 8761
problems involve the peculiar difficulties in separation
and purification, in particular using chromatography,
which are amplified in charged CD derivatives. By using
an efficient synthesis of the per-6-bromo-6-deoxy deriva-
tives of RCD, âCD, and γCD, the primary CD face may
be per-substituted with a variety of pendent groups.17 The
syntheses of a number of ACD derivatives are reported
herein, together with synthesis of a prototypical ACD in
which one 6′ position is selectively modified with an
alternative pendent group.
F igu r e 1. 6-Aminocyclodextrins.
pendent groups represent potentially interesting recep-
tors and biomimetics (Figure 1). First, since the depth
of the CD cavity is only one glucose residue (∼7.9 Å), the
corona formed from the pendent groups of a monofacially
substituted CD may represent a significant alternate
receptor to the cavity itself. Second, synergistic effects
may be seen between the pendent groups, not seen in
simple mono- or disubstituted CD derivatives.
Resu lts a n d Discu ssion
The most significant problems in regioselective CD
synthesis are associated with difficulties in separation
and purification, in particular using chromatography,
which can be considerable in the case of charged CD
derivatives. Low yields, often reported in the synthesis
of CD derivatives exacerbate these problems. For ex-
ample, in the synthesis of primary face functionalized,
per-6-substituted 6-deoxy-âCD derivatives, incomplete or
over-reaction leaves the problem of separating the fully
heptasubstituted from hexa- and/or octasubstituted âCD
derivatives, the physical properties of which are not
distinctly different. A further peculiar complication is
tight binding of some reaction byproducts within the CD
cavity. Thus, quantitative conversions are required with
suitable purification techniques that do not reduce the
isolated yield excessively.
The predominant strategy for ACD synthesis has
utilized halogenation at the 6′ position with in situ or
subsequent substitution with azide, followed by efficient
conversion to amine via the Staudinger reaction.,18,19
Lehn’s early strategies yielded isomeric products and
required secondary face functionalization for purifica-
tion.19 Subsequently, several modifications have been
reported leading to Defaye and co-workers’ more recent
perfacial bromination procedure using Br2/PPh3 in DMF.16
However, Stoddart and co-workers have reported dif-
ficulties with the workup presented in this and a further
paper. In its place, Stoddart’s group employ selective
precipitation and exhaustive Soxhlet extraction to remove
impurities and byproducts, in particular, residual tri-
phenylphosphine oxide.2b We have been using an alter-
native procedure in which the Vilsmeier-Haack reagent
[(CH3)2NCHBr]+Br- 20 is prepared and washed prior to
reaction with unprotected CD.17,21 In this way, a signifi-
cant portion of the Ph3PO can be removed before reaction
with CD, hence reducing the subsequent effort required
to remove this persistent byproduct. We have reported
this procedure for synthesis of the per-6-deoxy-6-bromo
derivatives of RCD, âCD, and γCD and subsequent
efficient conversion to a variety of monofacially substi-
tuted analogues including perazido, peramino, percyano,
permercapto, and perthioureido CD derivatives.17
Several attempts have been reported to couple CD to
amino acids to form peptidomimetics, including mono-
facially substituted cyclodextrins.12 An alternative strat-
egy to the use of amino acids is to provide pendent groups
that mimic the important peptide side chains. Amino-
CD (ACD) derivatives with varied amine pendent groups
provide a binding site rich in amino groups atop the
hydrophobic binding site of the cavity (Figure 1). There
are numerous examples of proteins with receptor sites
rich in the basic residues Arg, Lys, and His, many atop
hydrophobic pockets. Some of these are the functional
receptor sites of proteins, such as the glycosaminoglycan
sulfate binding site in, for example, antithrombin III.13
Others are functional active sites of enzymes, for ex-
ample, phosphatidylinositol-specific phospholipase C from
B. cereus.14 Further proteins, notably hemoglobin and
serum albumin, possess hydrophobic, Lys-rich, binding
sites that are capable of catalyzing unnatural reactions.15
Water-soluble ACD derivatives combine these hydropho-
bic and electrostatic binding sites.
Despite recent advances, several obstacles remain to
regioselective CD derivatization.2,16 The most significant
(11) Ashton, P. R.; Uwood, P. E.; Staton, I.; Stoddardt, J . F. J . Org.
Chem. 1991, 56, 7274. Folkman, J .; Weisz, P. B.; J oullie, M. M.; Li,
W. W.; Ewing, W. R. Science 1989, 243, 1490. Sakairi, N.; Kuzuhara,
H.; Okamoto, T.; Yajima, M. Bioorg. Med. Chem. 1996, 4, 2187. Moriya,
T.; Kurita, H.; Matsumoto, K.; Otake, T.; Mori, H.; Morimoto, M.; Ueba,
N.; Kunita, N. J . Med. Chem. 1991, 34, 2301. Takeo, K.; Mitoh, H.;
Uemura, K. Carbohydr. Res. 1989, 187, 203. Pregel, M. J .; Buncel, E.
Can. J . Chem. 1991, 69, 130. Fujedi, P. Carbohydr. Res. 1989, 192,
366. Coleman, A. W.; Zhang, P.; Parrot-Lopez, H.; Ling, C. C.; Miocque,
M.; Mascrier, L. Tetrahedron Lett. 1991, 32, 3997. Zhang, P.; Ling, C.
C.; Coleman, A. W.; Parrot-Lopez, H.; Galons, H. Tetrahedron Lett.
1991, 32, 2769. Khan, A. R.; Barton, L.; D′Souza, V. T. J . Chem. Soc.,
Chem. Commun. 1992, 1112. Takeo, K.; Mitoh, H.; Uemura, K.
Carbohydr. Res. 1989, 187, 203. Boger, J .; Brenner, D. G.; Knowles,
J . R. J . Am. Chem. Soc. 1979, 101, 7630. Fugedi, P. Carbohydr. Res.
1989, 192, 366.
(12) Parrot-Lopez, H.; Djedaini, F.; Perly, B.; Coleman, A. W.;
Galons, H.; Miocque, M. Tetrahedron Lett. 1990, 1999. Cucinotta, V.;
D’Allessandro, F.; Impellizzeri, G.; Pappalardo, G.; Rizzarelli, E.;
Vecchio, G. J . Chem. Soc. Chem. Commun. 1991, 293. Akerfeldt, K.
S.; DeGrado, W. F. Tetrahedron Lett. 1994, 4489.
(13) J ackson, R. L.; Busch, S. J .; Cardin, A. D. Physiol. Rev. 1991,
71, 481.
Preparation of per-6-amino-per-6-deoxy-CD derivatives
has been reported.22,23 Synthesis via the per-6-azido-per-
6-deoxy-CD derivative permits efficient Staudinger con-
(14) Vizitiu, D.; Campbell, A. S.; Thatcher, G. R. J . J . Mol. Recog.
1996, 9, 197. Heinz, D. W.; Ryan, M.; Smith, M. P.; Weaver, L. H;
Keana, J . F.; Griffith, O. H. Biochemistry 1996, 35, 9496.
(15) Thorn, S. N.; Daniels, R. G.; Auditor, M. M.; Hilvert, D. Nature
1995, 373, 228. Kikuchi, K.; Thorn, S. N.; Hilvert, D. Nature 1996,
118, 8184. Hollfelder, F.; Kirby, A. J .; Tawfik, D. S. Nature 1996, 383,
60. Elbaum, D.; Wiedenmann, B.; Nagel, R. J . Biol. Chem. 1982, 257,
8454.
(17) Gorin, B. I.; Riopelle, R. J .; Thatcher, G. R. J . Tetahedron. Lett.
1996, 37, 4647.
(18) Parrot-Lopez, H.; Ling, C-C.; Zhang, P.; Baszkin, A.; Albrecht,
G.; de Rango, C.; Coleman, A. W. J . Am. Chem. Soc. 1992, 114, 5479.
(19) Boger, J .; Corcoran, R. J .; Lehn, J .-M., Helv. Chim. Acta 1978,
61, 2190.
(20) Hepburn, D. R.; Hudson, H. R. J . Chem. Soc., Perkin Trans. 1
1976, 754.
(21) Borrajo, A. M. P.; Gorin, B. I.; Dostaler, S. M.; Riopelle, R. J .;
Thatcher, G. R. J . Bioorg. Med. Chem. Lett. 1997, 7, 1185.
(16) (a) Gadelle, A.; Defaye, J . Angew. Chem., Int. Ed. Engl. 1991,
30, 78. (b) Baer, H. H.; Berenguel, A. V.; Shu Y. Y.; Defaye J .; Gadelle,
A.; Gonzalez, F. S. Carbohydr. Res. 1992, 228, 307.

8762 J . Org. Chem., Vol. 62, No. 25, 1997
Vizitiu et al.
version to the per-6-amino-per-6-deoxy-CD and has the
advantage that direct secondary face modification of the
azide is possible without recourse to standard techniques
of primary face protection. However, this route exacer-
bates the problem with residual Ph3PO, and furthermore,
the product per-6-amino-per-6-deoxy-CD has poor solu-
bility in all common solvents in the free base form (e.g.,
DMSO, DMF, H2O). Nevertheless, coupling of amino
acids with per-6-amino-per-6-deoxy-CD has been re-
ported.2b Reaction of per-6-aminoper-6-deoxy-CD with
pyridyl carboxaldehydes and subsequent reduction is also
a feasible route to extended pyridyl-ACD derivatives.
However, the alternative route to ACD derivatives via
direct amination of the per-6-bromo-per-6-deoxy-CD (1)
is more facile, providing a pathway to a number of ACD
derivatives.
Route I: the amine excess was removed as the HBr salt.
Acidification of the crude ACD product with HBr followed
by exhaustive washings with hot abs. ethanol removed
all the excess amine. Trace ethanol was removed by
methanol washing, followed by thorough drying. Route
II requires dissolution of the crude ACD product in a
minimum amount of refluxing methanol followed by
precipitation in acetone, washing and subsequent titra-
tion with aqueous HBr. The ACD products obtained by
these methods were characterized as the per-HBr salts,
as confirmed by elemental analysis. In some cases, the
ACD derivatives were treated with anion-exchange resins
in order to obtain the free amine forms. Per-6-[(2-
methoxyhydroxyethylamino)]-6-deoxy-âCD was synthe-
sized in a similar fashion. Model reactions were also
carried out with R- and γCD. The reactions between 1R
and 1γ and either a primary amine, ethanolamine, or
secondary amine, N-methylethanolamine, were success-
fully performed under conditions similar to those for
reaction of 1â. All ACD compounds were highly soluble
in water, not only as HBr salts, but also as free bases,
and consequently, all the NMR spectra were recorded in
D2O, displaying the high-symmetry expected from quan-
titative persubstitution of the primary face (Figure 1).
It is clear that the use of amine as reagent and solvent
contributes to the quantitative conversion. Reaction with
tris(hydroxymethyl)aminomethane (TRIS), which must
be performed in a minimum amount of DMSO or DMF
as solvent, does not proceed to completion. Further,
when simple diethylamine was used, no reaction occurred
because of the insolubility of 1 even at reflux (55 °C).
P er -6-[(2-a m in oa lk yl)a m in o]-6-d eoxy-CD. Reac-
tion of 1 with primary and secondary amine sites of
2-hydroxyethylamines is clearly facile. To examine
competitive reaction of different amine sites and further
extend the family of ACD derivatives, reactions with N,N-
dimethylethylenediamine and N,N,N′-trimethylethylene-
diamine were examined. Reaction with N,N-dimethyl-
ethylenediamine, followed by acetone precipitation, yielded
a solid product that by 13C NMR showed complete
substitution of the bromide with a pattern corresponding
to displacement by primary amine alone. Purification
to remove the excess amine was achieved after treatment
with an anion-exchange resin, followed by exhaustive
washing of the neutral ACD product with acetone. The
reaction of 1 with N,N,N′-trimethylethylenediamine was
carried out under the same conditions, but the result was
different. In this case, a very small, but observable
competitive substitution by the tertiary amine nucleo-
phile was observed by 13C NMR. Given the finding that
a tertiary amine was nucleophilic toward 1, the reaction
with N,N,N′,N′-tetramethylethylenediamine was exam-
ined. Unfortunately, DMF was required as solvent,
leading again to incomplete substitution. The ACD
derivatives synthesized by these methods are highly
soluble in water in salt and free base forms. Further-
more, there is no evidence from NMR, MS or combustion
analysis of intramolecular or oligomerization reactions
(Figure 2).
P er -6-[(2-h yd r oxya lk yl)a m in o]-6-d eoxy-CD. Heat-
ing of 1 at 75-80 °C, in excess of amine, without solvent,
resulted in quantitative conversion to the corresponding
ACD. The reaction took 1-2 days for completion, de-
pending on the amine used as starting material. With
liquid amines, reaction was complete in 24 h, whereas
with the molten amine, serinol, up to 48 h was required
for completion. The amine excess was calculated as 10
equiv of amine/Br, serving both as reagent and solvent.
The workup procedure requires removal of excess free
amine by high-vacuum distillation, followed by precipita-
tion in acetone and washing of the precipitate. Where
distillation proved inefficient, direct acetone precipitation
of the reaction mixture was performed. Our experience
in handling cyclodextrins shows that temperatures higher
than 80 °C lead to CD ring cleavage products, which
cannot be removed cleanly at any stage of the workup.
Accordingly, in both reactions and distillations, the
temperature never exceeded 80 °C.
The crude solid resulting from acetone precipitation
showed total conversion of 1 by 13C NMR, but in every
case, starting amine was present presumably as the HBr
salt. Further purification could follow one of two routes.
ACD Der ivatives Bear in g Am in o Acid Side Ch ain s.
Direct imidazole substitution of 1 was attempted in order
to obtain a per-6-imidazolyl-6-deoxy-CD. The reaction
proceeded well in a minimum amount of DMF, at 75-80
°C, in 24 h. After removal of DMF, the remaining residue
was precipitated in water. The poor aqueous solubility
of the imidazolyl-substituted CD assisted in separation
and purification, but raises concerns over the utility of
(22) Tagaki, W.; Yano, K.; Yamanaka, K.; Yamamoto, H.; Miyasaki,
T. Tetrahedron Lett. 1990, 3897. Matsui, Y.; Tanemura, E.; Nonomura,
T. J . Am. Chem. Soc. 1993, 66, 2827.
(23) Synthesis of per-6-(methylamino)-6-deoxy-CD from direct con-
densation of amine with 1â has been reported previously to proceed
with high but incomplete substitution at the 6′ position (Breslow, R.;
Czarniecki, M. F.; Emert, J .; Hamaguchi, H. J . Am. Chem. Soc. 1980,
102, 762. Eliseev, A. V.; Schneider, H. J . J . Am. Chem. Soc. 1994, 116,
6081).

Synthesis of Monofacially Functionalized Cyclohextrins
J . Org. Chem., Vol. 62, No. 25, 1997 8763
Elemental analysis confirmed quantitative substitution
of the primary face, but the 13C NMR of the product
revealed competitive N and O substitution. Although
O-protection would be required to obtain the homoge-
neous N-substituted ACD, this result clearly indicates
the reactivity of 1 toward phenolate nucleophiles.
P er -6-(a m in om eth yl)-6-d eoxy-CD. The per-6-cy-
ano-6-deoxy-CD derivative (16) has clear potential as an
ACD synthon, for example, in amino acid coupling, if
effective reduction of the nitrile could be achieved.
Nitrile reduction to the corresponding primary amine
would provide an alternative ACD scaffold. However, it
is well-known that catalytic hydrogenation of nitriles may
give rise to a number of products, which include primary,
secondary, and tertiary amines, imines, hydrocarbons,
aldehydes, amides, and alcohols. The major product,
derived as a result of these competing reactions, depends
primarily on the catalyst, substrate, and reaction condi-
tions. The potential for intramolecular reactions at the
primary CD face suggests that the opportunity for side
reactions may be enhanced in this case.
The nitrile 16 is obtained readily by reaction of 1 with
KCN in dry DMF with heating at 75-80 °C for 24 h and
subsequent purification. Nitrile hydrogenation was car-
ried out in a Parr-pressure hydrogenator under a 50-55
psi H2 pressure, at room temperature. Water was used
as the reaction medium, although the nitrile was in-
soluble in water. PtO2 was used as a catalyst, but no
prehydrogenation of the catalyst was performed.
A
relatively large amount of catalyst (approximately 125
mg of PtO2/1 mmol of CN) was selected in order to
increase the rate of the main reaction. To minimize
secondary amine formation an acidic medium was used.
Various literature reports suggest that strong acids are
required in order to prevent the addition reaction be-
tween primary amine and imine, but in our case the use
of strong acid is severely limited by the stability of the
cyclodextrin ring. A previous attempt at CD-nitrile
hydrogenation reported a mixture of products.16b We
employed an equimolar amount of HCl in a dilute
aqueous solution to prevent the serious degradation
observed when excess HCl was employed. Use of a
milder acid, such as formic acid, did not yield the desired
product, but 13C NMR suggested formation of secondary
and tertiary amines. In the absence of acid catalyst, no
reaction occurred. The optimum reaction time was
selected as 3 h to minimize undesired side reactions and
since the nitrile was observed to be entirely consumed
within this period. The subsequent workup procedure
is extremely simple, involving only the filtration of the
catalyst through a prewashed Celite bed followed by
concentration of the clear filtrate. The 13C NMR of the
final product clearly indicates the seven major peaks of
the desired product, in addition to a limited number of
smaller side-product peaks (Figure 4). The product,
obtained as the HCl salt, may be converted to free base
form using anion-exchange resin. Surprisingly, in con-
trast to per-6-amino-6-deoxy-CD, per-6-(aminomethyl)-
6-deoxy-CD is freely water soluble in base form. The
purity of the per-6-(aminomethyl)-6-deoxy-CD obtained
in this way was estimated at g98% on the basis of 13C
NMR. The synthetic utility of 6-amino-6-deoxy-CD itself
is severely limited by the poor solubility of the free amino
derivatives and Ph3PO contamination. The route to per-
6-(aminomethyl)-6-deoxy-CD described herein provides
an alternative, improved, nucleophilic ACD synthon.
F igu r e 2. 100 MHz J -MOD 13C NMR spectra for 2, 6â, 4â,
and 7â.
F igu r e 3. 100 MHz J -MOD 13C NMR spectra for perimid-
azolide 9â.
this ACD derivative. Excess imidazole was removed by
purification route I (vide supra). Curiously, the product
displayed an anomalous reversed phase for the imida-
zolyl-C2′ at δ ∼136.24 in the standard J -MOD 13C NMR
spectrum (Figure 3). The low intensity and the reversed
phase of this C2′ signal is caused by the large J CH coupling
constant (220 Hz), and the assignment is confirmed by
an heteronuclear shift correlation experiment. The im-
idazoyl-CD product was >95% pure by 13C NMR but
clearly showed evidence of small amounts of other
products (Figure 3). Coupling of two CD units at the
primary face by an imidazole linker has been reported.
The ease of reaction of 1 with imidazole disallows
synthesis of an histidyl-CD mimic by reaction of 1 with
histamine.
To obtain a tyrosyl-CD mimic, reaction of tyramine
with 1 was examined. We had shown that benzylamine
reacted cleanly with 1 without cosolvent, although the
product had limited aqueous solubility. For reaction with
tyramine, DMF was required as solvent. After 24 h, the
reaction mixture was worked up using route I, yielding
an off-white solid product with limited water solubility.

8764 J . Org. Chem., Vol. 62, No. 25, 1997
Vizitiu et al.
the simple per-6-amino-6-deoxy-CD, which is soluble in
a range of polar solvents. Finally, the utility of the
synthetic strategy to synthesis of (n + 1)-ACD derivatives
is demonstrated by preparation of an homogeneous (6 +
1)-âACD. ACD derivatives possess catalytic activity at
physiological pH,25 since reaction is not dependent upon
ionization of a secondary hydroxyl: thus, (n + 1)-ACD
derivatives in which one 6′ position is functionalized with
an alternative receptor or catalytic group have significant
scope as biomimetics.
Exp er im en ta l Section
Gen er a l Meth od s. Unless otherwise stated, chemicals
were used as received. DMF was dried over CaH2 prior to use
and stored over the same. Cyclodextrins were recrystallized
from water and dried under vacuum in a drying pistol over
P2O5 at 100 °C prior to use. Reactions were performed under
inert atmosphere (Ar or N2) unless water was a component or
1
byproduct of reaction. Routine H and 13C NMR spectra were
recorded on a Bruker AC/F-200 or Bruker AM-400 spectrom-
eter (at 50, 100, 200 or 400 MHz). High-field assigments are
reported. Mass spectra were recorded on a VG mass spec-
trometer, using fast atom/ion bombardment (FAB, ion source)
in a glycerol or thioglycerol matrix. Carbon, hydrogen, nitro-
gen, and oxygen analyses were carried out in a gas chroma-
tography unit (Carlo Erba 1104) in which the sample is burned
and the amount of the resulting gases are measured by
thermal conductivity. CHN analyses involve combustion of
the sample in oxygen; the resulting carbon dioxide, water, and
nitrogen are measured by thermal conductivity and compared
directly with known standards, most commonly acetanilide and
nitroaniline. Oxygen is analyzed in a similar manner, but
carbon monoxide is the gas formed and measured by thermal
conductivity.
P er -6-br om o-6-d eoxy-â-cyclod extr in s (1). The bromi-
nation of cyclodextrin was carried out using the Vilsmeier-
Haack reagent [(CH3)2NCHBr]Br, which was prepared by
dropwise addition of Br2 (2.75 mL, 53.3 mmole) to a solution
of triphenylphosphine (14 g, 53.3 mmole) in 60 mL of dry DMF
with vigorous stirring, under Ar. During the addition, the
reaction mixture was not allowed to warm above 60 °C. The
resulting precipitate was further stirred at rt for 30 min. After
cooling to 0 °C, filtration of the reaction mixture under Ar
afforded the iminium salt as a yellow crystalline solid. Suc-
cessive washing on the filter with cold dry DMF, under Ar,
resulted in the white crystalline Vilsmeier-Haack reagent,
almost clean of the triphenylphosphine oxide side product. The
yield was calculated as approximately 80% by isolation of the
salt. However, in all subsequent usage, this reagent was used
in situ without isolation and a value of 80% was assumed in
calculations of molar equivalents. To the Vilsmeier-Haack
reagent (12 equiv for RCD, 14 equiv for âCD, and 16 equiv for
γCD) were added 100 mL of dry DMF and recrystallized
â-cyclodextrin (freshly dried, 3.4 g, 3 mmol), and the reaction
mixture was heated at 75-80 °C, under Ar. It is extremely
important not to exceed 80 °C, in which case the solution turns
dark-brown and the yield is severely diminished. After 10-
12 h, the DMF was removed under vacuum and the remaining
oil was poured into 50 mL of 3 M sodium methoxide solution
in methanol, previously cooled to 0 °C. After 1 h of stirring at
rt, the mixture was precipitated in a large amount of cooled
water and filtered followed by successive washings with
methanol to afford 1 as a white solid, in 95% isolated yield:
13C NMR (DMSO-d6) δ 102.10 (C1), 84.63 (C4), 72.29 (C3), 72.05
(C2), 71.01 (C5), 34.42 (C6); MS (+FAB, Cs+ source) M + H+
1576.
F igu r e 4. 100 MHz J -MOD 13C NMR spectra for 1â, 16â, and
17â.
6 + 1 Su bstitu ted âCD. Monotosylation of âCD
proceeds cleanly although in low yield after rigorous
purification from water.24 Displacement of the tosyl
group by benzylthiolate proceeded to completion in 12 h,
at 65-70 °C, in DMF. After evaporation of DMF, a crude
solid remained, which was washed several times with
water and then acetone. 13C NMR of the resulting
product indicated the presence of the benzyl disulfide,
which could not be rigorously removed from the CD
product, possibly because of tight-binding with the CD
derivative, involving the hydrophobic cavity and p-
stacking interactions. The crude mono-6-(benzylthio)-6-
deoxy-âCD was brominated using the Vilsmeier-Haack
reagent in an identical fashion to that used for simple
âCD, yielding a product contaminated with dibenzyl
disulfide. To obtain an example of (6 + 1)-âACD,
N-methylethanolamine was used to displace bromide.
Reaction with the amine as reagent and solvent, at 75-
80 °C, for 24 h afforded the (6 + 1)-âACD, which was
readily purified by evaporation of excess amine, precipi-
tation in acetone, and acidification in water to pH 3 with
1 M HBr. The resulting emulsion was washed with
chloroform, concentrated, and then washed with hot
absolute ethanol. Precipitation from water allowed for
the complete removal of the benzyl disulfide contaminant
after filtration through Celite. The 13C NMR spectrum
of the product obtained from the filtrate showed the
desired product 6 + 1 substitution pattern, confirmed by
MS and elemental analysis.
Su m m a r y
An efficient synthetic route to ACD derivatives has
been described and demonstrated by synthesis of 11
homogeneous ACD derivatives. The limitations of this
route and potential adaptations are illustrated by the
synthesis of several ACD derivatives to >95% purity. A
new synthetic route to a CD persubstituted with primary
amine functionalities at the primary face, per-6-(ami-
nomethyl)-6-deoxy-CD, yields an alternative synthon to
Gen er a l P r oced u r e: P er -6-(a lk yla m in o)-6-d eoxycyclo-
d extr in s (2-6). Per-6-(alkylamino)cyclodextrins were syn-
thesized by treating 1 with an excess of alkylamine (10 molar
equiv of amine per Br equivalent) at 80 °C for 24-48 h. The
(24) Zarzycki, R.; Hoehn, P.; Duong, B. Minutes Int. Symp. Cyclo-
dextrins, 6th 1992, 663.
(25) McCracken, P. G.; Vizitiu, D.; Walkinshaw, C. S.; Gorin, B. I.;
Wang, Y.; Thatcher, G. R. J . Manuscript submitted.

Synthesis of Monofacially Functionalized Cyclohextrins
J . Org. Chem., Vol. 62, No. 25, 1997 8765
reagent/solvent was removed under vacuum, and the residue
was precipitated in acetone. After filtration, and washing with
acetone, the solid was taken up in water and the pH was
carefully brought down to 4 with 1 M HBr. The aqueous
solution was evaporated under vacuum and further dried on
the vacuum line. The resulting solid was washed three times
with hot absolute ethanol in order to remove all of the
unreacted starting amine. The last filtrate was perfectly clear.
After drying on the vacuum line, persubstitution of the
primary face was confirmed by 13C NMR, MS, and elemental
analysis.
P er -6-[(2-h yd r oxyet h yl)a m in o]-6-d eoxy-r-cyclod ex-
tr in (2a ): 13C NMR (D2O) δ 101.11 (C1), 82.29 (C4), 72.30 (C3),
71.20 (C2), 67.59 (C5), 56.47 (CH2OH), 49.87 (NCH2), 48.02 (C6);
MS (+FAB) M + H+ 1231.3, M + H2Br+ 1313.2. Anal. Calcd
for C48H90N6O30‚6HBr‚4H2O: C, 32.23; H, 5.86; N, 4.70.
Found: C, 32.04; H, 5.59; N, 4.55.
P e r -6-[(2-h yd r oxye t h yl)a m in o]-6-d e oxy-â-cyclod e x-
tr in (2b): 13C NMR (D2O) δ 101.07 (C1), 81.91 (C4), 72.00 (C3),
71.60 (C2), 67.47 (C5), 56.65 (CH2OH), 49.83 (NCH2), 48.24 (C6);
MS (+FAB) M + H+ 1435.7. A sample for elemental analysis
was passed through a column with basic exchange resin
(Dowex 1-X8) to afford the free base form of the compound.
Anal. Calcd for C56H105N7O35‚H2O: C, 46.24; H, 7.41; N, 6.74;
O, 39.6. Found: C, 46.53; H, 7.39; N, 6.63; O, 39.96.
P e r -6-[(2-h yd r oxye t h yl)a m in o]-6-d e oxy-γ-cyclod e x-
tr in (2g): 13C NMR (D2O) δ 99.38 (C1), 79.73 (C4), 71.88 (C3),
71.43 (C2), 67.02 (C5), 56.55 (CH2OH), 49.79 (NCH2), 48.29 (C6);
MS (+FAB) M + H+ 1641.7. Anal. Calcd for C64H120N8-
O40‚8HBr‚4H2O: C, 32.56; H, 5.81; N, 4.75. Found: C, 32.70;
H, 5.56; N, 4.41.
P er -6-(N-m eth yl-N-(2-h yd r oxyeth yl)a m in o)-6-d eoxy-r-
cyclod extr in (3a ): 13C NMR (D2O) δ 99.21 (C1), 79.67 (C4),
71.89 (C3), 71.55 (C2), 66.37 (C5), 56.83 (C6), 55.35 (CH2OH,
CH2N), 41.55 (CH3N); MS (+FAB) M + H+ 1315.4. Anal.
Calcd for C54H112N6O30‚6HBr‚2H2O: C, 35.31; H, 6.15; N, 4.58.
Found: C, 35.26; H, 6.12; N, 4.67.
P er -6-[N-m eth yl-N-(2-h yd r oxyeth yl)a m in o]-6-d eoxy-â-
cyclod extr in (3b): 13C NMR (D2O) δ 99.80 (C1), 80.65 (C4),
71.70 (C3, C2), 66.57 (C5), 56.85 (C6), 55.25 (CH2OH, CH2N),
41.74 (CH3N); MS (+FAB) M + H+ 1534.7, M + H2Br+ 1616.5,
M + H3Br2+ 1696.5. Anal. Calcd for C63H119N7O35‚7HBr‚4H2O:
C, 34.82; H, 6.21; N, 4.51; O, 28.71. Found: C, 34.42; H, 5.98;
N, 4.42; O, 27.47.
P er -6-[N-m eth yl-N-(2-h yd r oxyeth yl)a m in o]-6-d eoxy-γ-
cyclod extr in (3g): 13C NMR (D2O) δ 100.80 (C1), 82.27 (C4),
72.14 (C3), 71.48 (C2), 67.18 (C5), 58.07 (C6), 56.86 (CH2N),
55.39 (CH2OH), 42.36 (CH3N); MS (+FAB) M + H+ 1753.5.
Anal. Calcd for C72H136N8O40‚8HBr‚5H2O: C, 34.71; H, 6.23;
N, 4.50. Found: C, 34.51; H, 5.91; N, 4.51.
P er -6-[(1,3-d ih yd r oxy-2-p r op yl)a m in o]-6-d eoxy-â-cy-
clod extr in (4b): 13C NMR (D2O) δ 100.94 (C1), 81.36 (C4),
71.98 (C3), 71.52 (C2), 67.40 (C5), 60.91 (C2′), 57.96 (C1′), 57.22
(C3′), 45.84 (C6); MS (+FAB) M + H+ 1645.8 found. Anal.
Calcd for C63H119N7O42‚7HBr‚3H2O: C, 33.38; H, 5.87; N, 4.32.
Found: C, 33.39; H, 5.67; N, 4.17.
P er -6-[b is(h yd r oxyet h yl)a m in o]-6-d eoxy-â-cyclod ex-
tr in (5b): 13C NMR (D2O) δ 99.35 (C1), 79.81 (C4), 72.59 (C2),
71.77 (C3), 69.89 (C5), 58.19 (C2′), 56.53 (C1′), 55.77 (C6); MS
(+FAB) M + H+ 1744.1. Anal. Calcd for C70H133N7O42‚4H2O:
C, 46.27; H, 7.82; N, 5.40. Found: C, 46.20; H, 7.34; N, 4.96.
Anal. Calcd for C70H133N7O42‚7HBr‚4H2O: C, 35.58; H, 5.46;
N, 4.15. Found: C, 35.74; H, 5.98; N, 3.84.
acetone, the solid was taken up in water and treated with
Amberlite IRA-410 ion-exchange resin in the hydroxide form.
The initial chloride commercial form of the resin was changed
into the hydroxide form prior to use by washing with 20
volumes of 1 M NaOH followed by rinsing with 4 volumes of
water. The resin load was calculated based on a ratio of 3.4
mequiv of anion/g wet resin, and the amount used was twice
the calculated one. The exchange was performed by stirring
the resin beads with the diluted aqueous solution of crude CD
for 2 h at rt. Filtration and washing of the resin with water
afforded a clear filtrate that was concentrated and dried. The
solid residue was thoroughly washed with acetone to yield the
pure CD-(ethylenediamino) derivatives in their free base form.
P er -6-[(N,N-d im eth yl-2-a m in oeth yl)a m in o]-6-d eoxy-r-
cyclod extr in (7b):
73.07 (C3), 72.02 (C2), 70.45 (C5), 57.61 (C2′), 49.19 (C1′), 46.74
(C6), 44.53 (2‚CH3); MS (+FAB) M + H+ 1625. Anal. Calcd
for C70H140N14O28‚3H2O: C, 50.05; H, 8.76; N, 11.67. Found:
C, 49.99; H, 8.11; N, 11.50.
P er -6-[(N,N,N′-tr im eth yl-2-am in oeth yl)am in o]-6-deoxy-
â-cyclod extr in (8b): 13C NMR (D2O) δ 100.66 (C1), 81.97
(C4), 73.13 (C3), 71.98 (C2), 69.59 (C5), 57.59 (C2′), 55.60 (C2′),
55.41 (C6), 44.60 (CH3), 42.47 (2‚CH3); MS (+FAB) M + H+
1723.4. Anal. Calcd for C77H154N14O28‚7H2O: C, 49.98; H,
9.15; N, 10.60. Found: C, 50.50; H, 8.22; N, 9.81.
P er -6-im id a zolyl-6-d eoxy-â-cyclod extr in (9b). A mix-
ture of 1b (0.3152 g, 0.0002 mol) and imidazole (0.272 g, 0.004
mol) in a minimum amount of anhydrous DMF (0.3 mL) was
stirred at 75-80 °C for 24 h. The reaction mixture was poured
into 30 mL of H2O with vigorous stirring. The resulting white
precipitate was filtered and then taken in 20 mL of H2O. HBr
(1 M) was added slowly to the suspension until the pH of the
resulting solution dropped down to 3. The aqueous solution
was rotary evaporated and then further dried on the vacuum
line. The resulting solid was washed three times with hot
absolute ethanol in order to remove all of the unreacted
imidazole. Removal of the ethanol traces from the final
compound was achieved by succesive additions and evapora-
tions of MeOH under vacuum, followed by vacuum drying to
yield the 8b‚7HBr salt (0.3 g, 73%) as a white powder: 13C
NMR (D2O) δ 136.24 (C2′), 123.2 (C4′), 120.65 (C5′), 101.97 (C1),
82.36 (C4), 72.14 (C3), 71.68 (C2), 69.49 (C5), 49.61 (C6); MS
(+FAB) M + H+ 1485.1.
P er -6-tyr a m in o-6-d eoxy-â-cyclod extr in (10b). Over a
suspension of 3.84 g of tyramine (0.028 mol) in 12 mL of dry
DMF at 70 °C was added in portions 1b (3.152 g, 0.002 mol).
One hour after the addition was completed, dossolution
occurred and the reaction mixture darkened. The solution was
stirred overnight at 75-80 °C to allow completion of the
substitution. Evaporation of DMF, followed by precipitation
in acetone, afforded a light-brown powder. The aqueous
solution of the crude product was titrated with 1 M HBr to
pH 3 and then rotavapped. The remaining residue was
refluxed in absolute ethanol and filtered successively until the
filtrate was clear. The resulting solid was concentrated from
methanol in order to remove the ethanol traces to yield
10b‚7HBr as a white powder: 13C NMR (D2O) δ 154.54 (ipso
COH), 129.88 (arom CH), 127.38 (ipso-C), 115.57 (arom CH),
101.26 (C1), 82.40 (C4), 72.12 (C3), 71.59 (C2), 67.67 (C5), 49.78
(CH2N), 48.73 (C6), 31.00 (CH2 phenol); MS (+FAB) M + H+
1967.5. Anal. Calcd for C98H133N7O35‚7HBr‚10H2O: C, 43.34;
H, 5.97; N, 3.61. Found: C, 43.01; H, 5.62; N, 3.80.
13C NMR (D2O) δ 101.66 (C1), 82.66 (C4),
P er -6-(ben zylam in o)-6-deoxy-â-cyclodextr in (11b). See
th e gen er a l p r oced u r e for p er -6-(a lk yla m in o)-6-d eoxy-
cyclod extr in : 13C NMR (D2O) δ 130.49 (ipso-C), 130.34 (2
arom CH), 129.64 (1 arom CH), 129.01 (2 arom CH), 101.38
(C1), 82.70 (C4), 72.00 (C3), 71.61 (C2), 67.51 (C5), 51.83 (benzyl
CH2), 48.77 (C6); MS (+FAB) M + H+ 1757.9. Anal. Calcd
For C91H119N9O28‚7HBr‚6H2O: C, 44.92 H, 5.72; N, 4.03.
Found: C, 44.51; H, 5.20; N, 4.52.
Mon otosyl-â-cyclod extr in (12). âCD (17.22 g, 0.015 mol)
was dissolved at rt in 200 mL of 1% NaOH solution. To the
stirring colorless solution was added 2.9 g of tosyl chloride
(0.015 mol) dissolved in 11 mL of CH3CN dropwise over 80
min. A white fine suspension was visible immediately after
the start of the addition, and a bulk precipitation ended the
P er -6-[(2-m et h oxyet h yl)a m in o]-6-d eoxy-â-cyclod ex-
tr in (6b): 13C NMR (D2O) δ 100.98 (C1), 81.84 (C4), 72.00 (C3),
71.58 (C2), 67.35 (C5), 66.83 (CH2O), 58.67 (OCH3), 48.57
(NCH2), 47.79 (C6); MS (+FAB) M + H+ 1534.6, M + H2Br+
1616.5. Anal. Calcd for C70H133N7O42‚7HBr‚4H2O: C, 34.82;
H, 6.22; N, 4.51. Found: C, 34.92; H, 6.12; N, 4.51.
Gen er a l P r oced u r e. P er -6-[(N,N-d im eth yl-2-a m in o-
eth yl)a m in o]-6-d eoxy-cyclod extr in s (7 a n d 8). Per-6-
(N,N-dimethylethylenediamino)-â-cyclodextrins were synthe-
sized by treating 1 with an excess of ethylenediamine (10 molar
equiv amine per Br equivalent) at 80 °C for 12 h. The reagent/
solvent was removed under vacuum, and the residue was
precipitated in acetone. After filtration and washing with

8766 J . Org. Chem., Vol. 62, No. 25, 1997
Vizitiu et al.
addition. The reaction mixture was stirred another 2 h at rt
and then filtered. The filtrate was acidified to about pH 2-3
with 1 M HCl and the product allowed to precipitate at 2 °C
overnight. Filtration afforded a white solid that combined with
the previous precipitate yielded 2.4 g of crude product. Re-
crystallization from water (130 mL) afforded 1.22 g (0.000 95
mol, 6.3%) of material, homogeneous by NMR: 1H NMR
(DMSO-d6) δ 7.75 (d, J ) 7.5 Hz, 2H, o-Ar), 7.43 (d, J ) 7.5
Hz, 2H, m-Ar), 5.72 (br, 14 H, s, OH), 4.84 (s, 6H, H1), 4.77 (s,
1H, H1′), 4.48 (br, 6H, pr OH), 4.34 (d, J ) 8 Hz, 1H), 4.20 (m,
1H), 3.90-3.00 (m, 40H), 2.43 (s, 3H, Me); 13C NMR (DMSO-
d6) δ 144.82 (ipso-Ar), 132.67 (p-Ar), 129.90 (o-Ar), 127.59 (m-
Ar), 102.25, 101.94, 101.86 (C1), 101.29 (C1′), 81.67, 81.52,
81.44, 81.18 (C4), 80.78 (C4′), 73.08, 72.97, 72.74, 72.45, 72.38,
72.19, 72.05, 71.89 (C2,3,5), 69.73 (C6′), 68.93 (C5′), 59.93, 59.54,
59.28 (C6), 21.23 (CH3).
HBr to pH 3 afforded an emulsion that after extraction twice
with chloroform resulted in a clear aqueous solution. The
aqueous layer was further evaporated down and the remaining
residue washed thoroughly with hot absolute ethanol. The last
traces of benzyl disulfide were removed by filtration through
a Celite bed of the suspension resulting from water precipita-
tion. Evaporation of the clear solution afforded 15‚6HBr (110
mg, 53.1%) as a glassy solid: 13C NMR (D2O) δ 138.41 (ipso-
Ar), 129.21 (o-Ar), 129.00 (m-Ar), 127.66 (p-Ar), 100.83, 100.32,
100.08, 99.64 (C1), 82.71 (C1′), 81.34, 80.97, 99.64 (C4), 72.93,
72.02, 71.85, 71.76, 71.65, 71.44, 71.36 (C2,3), 67.64, 67.12,
66.77, 66.54, 66.41, 66.27, 66.02 (C5), 56.47 (C6), 55.39, 55.23
(CH2), 50.54 (C6′), 42.26, 41.73 (CH3), 37.00 (CH2-benzyl); MS
(+FAB) M + H+ 1583.6. Anal. Calcd for C67H118N6O34S‚
6HBr‚6H2O: C, 36.96; H, 6.30; N, 3.86. Found: C, 36.96; H,
6.00; N, 4.12.
Mon o-6-(ben zylth io)-6-d eoxy-â-cyclod extr in (13). NaH
(60% dispersion in oil, 0.4 g, 0.001 mol) was washed twice with
dry toluene, under Ar. After removal of toluene, the remaining
suspension was taken in 6 mL of dry DMF, and 0.117 mL of
benzylmercaptan (0.001 mol) was added dropwise. The mix-
ture was stirred at rt for 1 h to allow quantitative formation
of the anion. Monotosyl-â-cyclodextrin (0.45 g, 0.000 35 mol)
was added in one portion, and the reaction mixture was heated
to 65-70 °C. Above 50 °C, the initial suspension became a
clear solution. After 12 h, the DMF was rotavapped and the
solid residue was taken in water to afford a white suspension.
The precipitate was filtered and successively washed with
water, acetone, and diethyl ether until no mercaptan smell
could be detected: 1H NMR (DMSO-d6) δ 7.29 (br, 5H, Ar),
5.71 (br, 14H, sec OH), 4.83 (s, 6H, H1), 4.53 (s, 1H, H1′), 4.45
(br, 6H, pr OH), 3.9-3.2 (m, 42H), 2.96 (s, 2H, benzyl); 13C
NMR (DMSO-d6) δ 138.92 (ipso-Ar), 129.40 (o-Ar), 128.98 (m-
Ar), 127.36 (p-Ar), 102.35, 101.97, 101.64 (C1), 85.04 (C1′),
81.71, 81.56, 81.45 (C4), 81.22 (C4′), 73.07, 72.87, 72.45, 72.30,
72.20, 72.07 (C2,3,5), 71.42 (C5′), 59.88 (C6), 41.62 (C6′), 36.55
(C-benzyl).
Mon o-6-(ben zylth io)h exa -6-br om o-6-d eoxy-â-cyclod ex-
tr in (14). The bromination of mono-6-(benzylthio)-â-cyclo-
dextrin was carried out following the same procedure described
above, using the Vilsmeier-Haack reagent [(CH3)2NCHBr]Br.
Br2 (0.193 mL, 3.75 mmol) was added dropwise to a solution
of triphenylphosphine (0.9825 g, 3.75 mmol) in 4 mL of dry
DMF with vigorous stirring, under Ar. The washed Vils-
meier-Haack reagent was combined with 6-(benzylthio)-â-
cyclodextrin (0.2482 g, 0.2 mmol) in 3 mL of dry DMF and
stirred overnight at 75-80 °C. After the known workup, 0.262
g of 14 was obtained as a white powder in 81% isolated yield:
1H NMR (DMSO-d6) δ 7.30 (br, 5H, Ar), 5.90 (Br, 14H, sec
OH), 4.96 (s, 6H, H1), 4.60 (s, 1H, H1′), 4.2-3.1 (m, 42H), 3.05
(s, 2H, benzyl); 13C NMR (DMSO-d6) δ 138.78 (ipso-Ar), 129.35
(o-Ar), 128.82 (m-Ar), 127.29 (p-Ar), 101.96 (C1), 85.12 (C4′),
84.60, 84.23 (C4), 72.31, 72.05, 71.08, 70.77, 70.37 (C2,3,5), 41.61
(C6′), 37.06 (C-benzyl), 34.35, 34.22 (C6).
Gen er a l P r oced u r e for P er -6-cya n o-6-d eoxy-CD (16a ,-
h ,g). The per-6-cyanoCD was formed from reaction of per-
bromoCD (1) with KCN (1.3n equiv; n ) 6, 7, 8) in DMF with
heating at 80 °C for 24 h. After evaporation of solvent, water
was added, yielding an off-white precipitate that was filtered.
Thorough washing with water and methanol was responsible
for isolation of the nitrile product in 65-70% yield.
P er -6-cya n o-6-d eoxy-r-cyclod ext r in (16a ): 13C NMR
(DMSO-d6) δ 118.35 (CN), 101.94 (C1), 85.84 (C4), 72.44 (C3),
71.65 (C2), 66.85 (C5), 20.93 (C6).
P er -6-cya n o-6-d eoxy-â-cyclod extr in (16b): 13C NMR
(DMSO-d6) δ 118.08 (CN), 102.11 (C1), 85.32 (C4), 72.16 (C3),
71.84 (C2), 67.01 (C5), 20.64 (C6).
P er -6-cya n o-6-d eoxy-γ-cyclod extr in (16g): 13C NMR
(DMSO-d6) δ 118.11 (CN), 102.18 (C1), 84.85 (C4), 72.11 (C3),
72.04 (C2), 67.08 (C5), 20.58 (C6).
Gen er a l P r oced u r e for P er -6-(a m in om eth yl)-6-d eoxy-
CD (17). The per-6-cyano-6-deoxy-CD (0.2 mmol) was hydro-
genated in a Parr-pressure hydrogenator under 50-55 psi H2
pressure, at room temperature. The nitrile was taken in water
(60 mL), and then the catalyst (n, 0.2, 125 mg PtO2; n ) 6, 7,
8) and the 1 M HCl (n, 0.2 mmols; n ) 6, 7, 8) were added to
the suspension. PtO2 was used without prehydrogenation.
After only 3 h of hydrogenation the reaction mixture was clear,
and the reduction was stopped. Filtration of the catalyst
suspension through a water prewashed Celite bed followed by
evaporation of the aqueous solution afforded 17‚7HCl as a
white solid.
P er -6-(a m in om eth yl)-6-d eoxy-r-cyclod extr in (17a ): 13C
NMR (D2O) δ 101.30 (C1), 84.00 (C4), 73.11 (C3), 71.63 (C2),
69.77 (C5), 36.05 (C7), 27.67 (C6); MS (+FAB) M + H+ 1051.2.
P er -6-(a m in om eth yl)-6-d eoxy-â-cyclod extr in (17b): 13C
NMR (D2O) δ 101.33 (C1), 83.51 (C4), 72.70 (C3), 71.90 (C2),
69.83 (C5), 36.20 (C7), 27.72 (C6); MS (+FAB) M + H+ 1226.4.
P er -6-(a m in om eth yl)-6-d eoxy-γ-cyclod extr in (17g): 13C
NMR (D2O) δ 100.83 (C1), 82.62 (C4), 72.54 (C3), 72.12 (C2),
69.82 (C5), 36.40 (C7), 27.89 (C6); MS (+FAB) M + H+ 1401.7.
Mon o-6-(b en zylt h io)h exa -6-[N-m et h yl-N-(2-h yd r oxy-
eth yl)a m in o]-6-d eoxy-â-cyclod extr in (15). 14 (0.162 g, 0.1
mmol) was stirred in 2 mL of N-methylethanolamine at 75-
80 °C for 24 h. After evaporation of the excess amine, the
residue was precipitated in acetone. Protonation with 1 M
Ack n ow led gm en t. NSERC Canada is thanked for
financial support.
J O9711549