Regioselectivity in the formation of di- and tri-6-O-mesitylenesulfonates of α-cyclodextrin

Article abstract of DOI:10.1016/j.carres.2014.10.027

The quantitative analysis of the reaction products for α-cyclodextrin (α-CD) with mesitylenesulfonyl chloride (MessCl) showed that di- and tri-mesitylenesulfonylation of the primary hydroxy groups of α-CD is regioselective. The reaction of mono-6-O-mesity

Full text of DOI:10.1016/j.carres.2014.10.027

Carbohydrate Research 401 (2015) 58–63
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Regioselectivity in the formation of di- and tri-6-O-
mesitylenesulfonates of a-cyclodextrin
⇑
Keisuke Yoshikiyo , Misaki Shinjo, Yoshihisa Matsui, Tatsuyuki Yamamoto
Faculty of Life and Environmental Science, Shimane University, 1060 Nishikawatsu, Matsue, Shimane 690-8504, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 10 July 2014
Received in revised form 23 October 2014
Accepted 27 October 2014
Available online 8 November 2014
The quantitative analysis of the reaction products for
a-cyclodextrin (a-CD) with mesitylenesulfonyl
chloride (MessCl) showed that di- and tri-mesitylenesulfonylation of the primary hydroxy groups of
-CD is regioselective. The reaction of mono-6-O-mesitylenesulfonyl- -CD with MessCl in pyridine gave
less 6A,6C-di-O-mesitylenesulfonyl- -CD than 6A,6B-di-O-mesitylenesulfonyl-
-CD. The reaction of
6A,6D-di-O-mesitylenesulfonyl- -CD with MessCl gave less 6A,6B,6E-tri-O-mesitylenesulfonyl-
-CD than
6A,6B,6D-tri-O-mesitylenesulfonyl-
-CD. These results indicate that the mesitylenesulfonyl group
attached to glucopyranose-A (Glc-A) retards further mesitylenesulfonylation of the primary hydroxy
group of Glc-C. The 1H NMR spectra of these modified
-CDs showed that the signal for the primary
a
a
a
a
a
a
a
Keywords:
a-Cyclodextrin
a
Mesitylenesulfonylation
Regioselectivity
Regioisomers
hydroxy and anomeric protons of Glc-C are significantly shifted upfield by the mesitylenesulfonyl group
of Glc-A.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
mono- and poly-6-O-sulfonylated derivatives.9 As illustrated in
Scheme 1, di-6-O-mesitylenesulfonylated
three regioisomers such as 6A,6B-, 6A,6C-, and 6A,6D-di-O-mes-
itylenesulfonyl- -CDs (AB-, AC-, and AD-isomers, respectively),
and tri-6-O-mesitylenesulfonylated -CD, -CD(Mess)3, has four
regioisomers such as 6A,6B,6C-, 6A,6B,6D-, 6A,6B,6E-, and 6A,6C,6E-
tri-O-mesitylenesulfonyl- -CDs (ABC-, ABD-, ABE-, and ACE-iso-
mers, respectively). These regioisomers have successfully been
separated and identified.10–13 However, not much has been
reported on the quantitative analysis on the formation ratio of
a-CD, a-CD(Mess)2, has
Cyclodextrins (CDs) are cyclic oligosaccharides composed of 6–
8 glucopyranoses (Glc) linked via
a
-1,4 glycosidic bonds, having
a
hydrophobic cavities at the center of the molecules, in which
guests, molecules or ions, are selectively included to form inclusion
complexes. Owing to the guest selectivity, CDs have long been
studied as model compounds for biological receptors1,2 or artificial
molecular recognition tools.3–5 Especially, the chemically modified
CDs have attracted considerable attention, as the modifications of
functional groups on CDs often result in the expression of unprec-
edented functions such as a stronger guest binding comparable to
the biological systems, or enhanced chiral recognition ability,
etc.6,7 Most of the CD derivatives are synthesized via activated
compounds in which CD hydroxy groups are substituted with good
leaving groups. Among them, the sulfonyl group is one of the most
excellent leaving groups, with high reactivity to nucleophiles. The
sulfonylation of hydroxy groups of CDs with p-toluenesulfonyl
chloride is frequently used for this purpose. However, the reaction
is nonselective and often produces a mixture of the toluenesulf-
onylated products both at the primary as well as the secondary
side, together with di- and tri-substituted derivatives.8 The sulf-
onylation of CD using mesitylenesulfonyl chloride (MessCl) was
reported to be an efficient method for a selective sulfonylation of
the primary hydroxy groups, C(6)-OHs, and gives a mixture of
a
a
a
regioisomers of
works by Fujita et al. which reported yield of three regioisomers
of
-CD(Mess)2.10,11
From the results of our recent experiments, we also found that
AC-isomer, one of the three regioisomers of -CD(Mess)2, was
formed in about half the yield of AB-isomer in the reaction of
-CD with MessCl in pyridine. If this mesitylenesulfonylation reac-
a-CD(Mess)2 and a-CD(Mess)3, except for the
a
a
a
tion has no regioselectivity, AB- and AC-isomers would be formed
in the same yield. The apparent difference in yield between these
isomers indicates that the reaction is regioselective. This finding
had led us to investigate these reactions in detail.
2. Results and discussions
Dried a-CD was allowed to react with MessCl in pyridine at 2 °C,
and aliquots were withdrawn at hourly intervals. The concentra-
tions of the reaction products in the aliquots were determined by
means of UFLC (ultra-fast liquid chromatography) as described in
⇑
Corresponding author.
E-mail address: yoshikiyo@life.shimane-u.ac.jp (K. Yoshikiyo).
http://dx.doi.org/10.1016/j.carres.2014.10.027
0008-6215/Ó 2014 Elsevier Ltd. All rights reserved.

K. Yoshikiyo et al. / Carbohydrate Research 401 (2015) 58–63
59
Scheme 1. Mesitylenesulfonates of
a-CD examined in the present study and formation pathways. Letters A–F represent the Glc residues when viewed from the primary
hydroxy side.
the Experimental. The concentrations of mono-6-O-mes-
itylenesulfonyl- -CD, -CD(Mess)1, and three regioisomers of
CD(Mess)2 formed are plotted versus the reaction time in Figure 1.
-CD(Mess)1 was smoothly produced after the reaction started
has the same reactivity with MessCl, then AB- and AC-isomers
would be produced in an equal concentration, and in twice of
AD-isomer; (AB/AC/AD = 1:1:0.5). However, the molar ratio of the
products was about 1.00:0.55:0.45 (AB/AC/AD) in every aliquot,
showing that the AC-isomer was produced considerably less than
AB-isomer. A similar result has been reported, in which the molar
ratio of the product was 1.00:0.67:0.54 (AB/AC/AD).10,11 The results
a
a
a-
a
(left scale), together with three regioisomers of
a-CD(Mess)2 with
lower concentrations (right scale) than the
a
-CD(Mess)1. The
concentrations of
reaction and gradually increased with time, suggesting that
-CD(Mess)2 is formed by the sulfonylation of the formerly pro-
duced -CD(Mess)1. If the C(6)-OH of each Glc of the -CD(Mess)1
a-CD(Mess)2 were low at the beginning of the
clearly show that the second Mess group is introduced into the
CD(Mess)1 with regioselectivity.
a-
a
a
a
In order to investigate the formation of
detail, -CD(Mess)1 was directly mesitylenesulfonylated in pyri-
dine to form -CD(Mess)2. The time-dependent changes in the con-
centration of three regioisomers of -CD(Mess)2 are shown in
a-CD(Mess)2 in more
a
a
a
Figure 2. The result showed that the molar ratio of AB/AC/
AD = 1.00:0.57:0.44, virtually the same as the direct reaction of
a
a
-CD as
-CD(Mess)3 in the crude product was also determined to be less
-CD(Mess)2. This shows that further sulfonylation of
-CD(Mess)2 to -CD(Mess)3 did not practically affect the molar
-CD(Mess)2. AC-isomer is formed from -CD(Mess)1 when
a reactant. The total amount of four isomers of
than 4% of
a
ratio of
a
a
a
a
C(6)-OH of either Glc-C or -E is sulfonylated. Glc-C is the second
adjacent Glc in a counter-clockwise direction from Glc-A, and
Glc-E is the second adjacent Glc in a clockwise direction. The small
molar ratio of AC-isomer to AB-isomer shows that C(6)-OHs of Glc-
C and/or Glc-E in a-CD(Mess)1 were less reactive than the others.
Then, we carried out experiments using AD-isomer as a reactant
to investigate which of the C(6)-OH of Glc-C or -E is more restricted
in the sulfonylation. If the Mess groups of Glc-A and -D of the
AD-isomer retard the sulfonylation of the C(6)-OH of the second
adjacent Glc in a counter-clockwise direction (Glc-C and -F),
ABE-isomer would be a minor product, and ABD-isomer, a major
one. On the contrary, if they retard the sulfonylation of the C(6)-
OH of the second adjacent Glc in a clockwise direction (Glc-E
and -B), ABD-isomer would be a minor, and ABE-isomer, a major
product. The quantitative analysis of the product in this reaction
Figure 1. Plots of the concentration of
AC-(N), and AD-isomers (j), right axis, versus the reaction time for the reaction of
-CD with MessCl in pyridine at 2 °C. Average molar ratio of AB/AC/AD was
1.00:0.55 0.01:0.45 0.01.
a-CD(Mess)1 (}), left axis, and AB-(ꢀ),
a

60
K. Yoshikiyo et al. / Carbohydrate Research 401 (2015) 58–63
the molar ratio of the ACE-isomer, that is produced by the
substitution of the C(6)-OH on Glc-E (the second adjacent Glc in
a counter-clockwise direction from Glc-C), was the smallest.
The reaction of MessCl with AB-isomer, having Mess groups on
two adjacent Glcs, produces ABC-isomer, if the substitution occurs
on either Glc-C or -F. On the other hand, ABD- and ABE-isomers are
formed, if the substitution occurs on Glc-D and -E of the AB-isomer,
respectively. The molar ratio of the product in this reaction would
be 1.0:0.5:0.5 (ABC/ABD/ABE), if the substitution occurs without
regioselectivity. The analysis of the products in the reaction
showed the ratio to be 1.00:0.53:0.52 as shown in Figure 5, which
is almost consistent with the theoretical estimation on the basis of
no regioselectivity in the reaction. However, this result is consid-
ered to be a specific one, because all other results in the present
study suggest the regioselectivity for the reaction. It seems that
interactions between two Mess groups on adjacent Glcs bring
about some change in the orientation of Mess groups and, conse-
quently, disappearance of the regioselectivity.
From the quantitative analysis on the products in the reaction
of four reactants, namely,
with MessCl in pyridine, it is strongly suggested that the Mess
group introduced to C(6)-O of -CD gives steric hindrance effect
on the substitution of the C(6)-OH of the second Glc in a coun-
ter-clockwise direction with the another Mess group.
a-CD, a-CD(Mess)1, AD-, and AC-isomers
Figure 2. Plots of the concentration of AB-(ꢀ), AC-(N), and AD-isomers (j) versus
a
the reaction time for the reaction of
a-CD(Mess)1 with MessCl in pyridine at 2 °C.
Average molar ratio of AB/AC/AD was 1.00:0.57 0.01:0.44 0.01.
In order to confirm this suggestion, we have checked the
orientation of Mess groups of AD-isomer, together with that of
a
(Fig. 6). The spectrum for AD-isomer was fully assigned by means
of 1D HOHAHA, 2D COSY, and 2D ROESY (Table 1). The signals for
O(6)-H and C(1)-H of Glc-C and -F were fairly shifted upfield
revealed that ABE-isomer was formed in smaller amount than
ABD-isomer, with the molar ratio of 0.80:1.00 (Fig. 3). This result
suggests that the Mess group of Glc-A retards the sulfonylation
of the C(6)-OH of the second adjacent Glc in a counter-clockwise
direction (Glc-C).
-CD(Mess)1, in DMSO-d6 by means of 1H NMR spectroscopy
(D
d = ꢀ0.19 and ꢀ0.25 ppm, respectively) compared with those
for -CD. On the other hand, the signal for O(6)-H of Glc-B and
-E gave almost the same chemical shift value as that of -CD
Next, AC-isomer was used as a reactant in place of the AD-iso-
a
mer, and the molar ratio of the regioisomers of a-CD(Mess)3 in the
product was investigated. As Glc-B, -D, -E, and -F in the AC-isomer
a
(
D
d = ꢀ0.05 ppm). The upfield shift of the signals is attributable
are chemically unequivalent to one another, the substitution on
to the ring current of the benzene rings in Mess groups. The ROESY
spectrum of AD-isomer gave a cross peak between o-Me and C(1)-
H of Glc-C and -F (Fig. 7). Therefore, it is supposed that the Mess
group on Glc-A is oriented adjacent to the C(6)-OH and C(1)-H of
Glc-C, and that on Glc-D to the C(6)-OH and C(1)-H of Glc-F. For
each of these Glcs gives four regioisomers of
a-CD(Mess)3. If the
substitution reaction occurs without regioselectivity, then these
regioisomers should be produced in an equal yield. Interestingly,
quantitative analysis revealed the molar ratio of these regioisom-
ers to be 0.56:0.35:1.00:0.25 (ABC/ABD/ABE/ACE), being substan-
tially different from each other (Fig. 4). This result again suggests
the regioselectivity for the reaction. Among the four regioisomers,
a
-CD(Mess)1, a full assignment of the spectrum was not accom-
plished because of the complexity of the spectrum (Fig. 6b). The
Figure 4. Plots of the concentration of ABC-(4), ABD-(h), ABE-(s), and ACE-
isomers (d) versus the reaction time for the reaction of AC-isomer with MessCl in
pyridine at 2 °C. Average mole ratio of ABC/ABD/ABE/ACE was 0.56 0.01:0.35
0.01:1.00:0.25 0.01.
Figure 3. Plots of the concentration of ABD-(h), and ABE-isomers (s) versus the
reaction time for the reaction of AD-isomer with MessCl in pyridine at 2 °C. Average
molar ratio of ABD/ABE was 1.00:0.80 0.01.

K. Yoshikiyo et al. / Carbohydrate Research 401 (2015) 58–63
61
counter-clockwise direction and retards the sulfonylation of it,
resulting in regioselectivity.
3. Experimental
3.1. Materials
a
-CD was supplied by Ensuiko Sugar Refining Co., Ltd, and dried
overnight in vacuo at 110 °C before use. MessCl (Wako Pure Chem-
ical Industries Ltd) was used without further purification. Reagent-
grade pyridine was dried over CaH2 and distilled in the presence of
CaH2 before use. Acetonitrile for high performance liquid chroma-
tography was purchased from Wako Pure Chemical Industries Ltd.
Dimethyl-d6 sulfoxide (DMSO-d6) containing 0.05% v/v tetrameth-
ylsilane (Cambridge Isotope Laboratories, Inc., 99.9 atm % D) was
used for 1H NMR measurement. 6-O-Mesitylenesulfonates of
a
-CD were prepared and separated from one another according
to the methods previously described.10,11
3.2. Apparatus
Figure 5. Plots of the concentration of ABC-(4), ABD-(h), and ABE-isomers (s),
versus the reaction time for the reaction of AB-isomer with MessCl in pyridine at
2 °C. Average molar ratio of ABC/ABD/ABE was 1.00:0.53 0.01:0.52 0.01.
The mesitylenesulfonates of
a-CD in a reaction mixture were
quantitatively determined with a Shimadzu prominence UFLC sys-
tem, which was composed of a pair of pumps (LC-20 AD), an auto-
sampler (SIL-20 ACHT), a column oven (CTO-20AC), a UV detector
(SPD-M20A), and a system controller (CBM-20A). Shim-pack XR-
ODS (100 mm ꢁ 3.0 mm i.d.) was used as a column, through which
acetonitrile/water was passed as an eluent. The flow rate of eluent
was 0.5 mL/min, and the eluate was detected at 275 nm. Column
temperature was maintained at 40 °C. Linear standard curves were
prepared between the peak areas of the sulfonates and their concen-
signal appearing at d = 4.13 ppm was supposed to be due to the
O(6)-H of Glc-C, judged from the similarity of the spectrum for
AD-isomer. The ROESY spectrum of
peak between o-Me and C(1)-H of Glc-C. The O(6)-H and C(1)-H
assigned to Glc-C of -CD(Mess)1 were up-field shifted to a greater
a-CD(Mess)1 also gave a cross
a
extent than the O(6)-H and C(1)-H of Glc-C and -F of AD-isomer,
indicating the greater steric hindrance effect of the Mess group
trations. The 1H NMR spectra of the sulfonylated
a-CDs in DMSO-d6
on
a
-CD(Mess)1 than those on AD-isomer. It is to be noted that
were recorded on a JEOL Model JNM-A400 FT NMR spectrometer
no cross peak between o-Me and C(1)-H of any Glc is observed in
the ROESY spectrum of AB-isomer, suggesting the less steric hin-
drance effect of the Mess group in this isomer would be one of
the reasons for the no selectivity in the reaction with MessCl.
(400 MHz) with a sample tube of 5.0 mm diameter at 323 K.
3.3. Reaction of a-CD with MessCl and product analysis
In conclusion, the Mess group attached to C(6)-O of
a-CD is
a
-CD (1.01 g, 1.04 mmol) was dissolved in pyridine (50 mL).
oriented to the C(6)-OH of the second adjacent Glc in
a
The solution was boiled to remove a trace amount of water as
Figure 6. 1H NMR spectra of
a-CD (a), a-CD(Mess)1 (b), and AD-isomer (c) in DMSO-d6 at 323 K. Letters in parentheses represent the Glc residues.

62
K. Yoshikiyo et al. / Carbohydrate Research 401 (2015) 58–63
Table 1
Assignments of 1H NMR spectra and their chemical shifts (ppm) of AD-isomer and native
a-CD
Glc residues of AD-isomer
a-CD
A,D
B,E
C,F
C(1)-H
C(2)-H
C(3)-H
C(4)-H
C(5)-H
C(6)-H(g)
C(6)-H(t)
O(2)-H
O(3)-H
O(6)-H
m-H
4.68
3.10
3.73
3.21
3.81
4.27
4.20
5.40
5.35
(d, J = 3.4)
(m)
(m)
(t, J = 9.0)
(dd, J = 2.4, 9.8)
(d, J = 10.7)
(dd, J = 4.6, 11.0)
(d, J = 6.3)
(d, J = 2.9)
4.80
3.27
3.75
3.40
3.56
3.50
3.50
5.43
5.33
4.30
(d, J = 3.4)
(m)
(m)
(t, J = 9.0)
(m)
(m)
4.65
3.24
3.75
3.47
3.63
3.52
3.52
5.41
5.39
4.15
(d, J = 3.4)
(m)
(m)
(t, J = 8.8)
(m)
(m)
4.80
3.28
3.78
3.39
3.59
3.66
3.66
5.40
5.36
4.34
(d, J = 3.4)
(m)
(td, J = 2.7, 9.0)
(t, J = 9.0)
(dd, J = 1.7, 9.5)
(m)
(m)
(m)
(m)
(d, J = 6.8)
(d, J = 2.9)
(t, J = 5.9)
(d, J = 5.4)
(d, J = 2.9)
(t, J = 5.6)
(d, J = 7.3)
(d, J = 2.9)
(t, J = 6.6)
7.07
2.53
2.29
(s)
(s)
(s)
o-Me-H
p-Me-H
PPM
4.8
4.7
4.6
C1(C,F)
O-Me
Figure 7. ROESY spectrum of AD-isomer in DMSO-d6 at 323 K. A cross peak between o-Me and C(1)-H of Glc-C and -F is observed.
an azeotropic mixture, and the resulting solution (36.5 mL) was
stirred in an ice-water bath (2 °C). MessCl (0.561 g, 2.57 mmol)
was added to the stirred solution. At hourly intervals, 3.0 mL
aliquots were withdrawn, added to 0.3 mL water to stop the
reaction, and analyzed by UFLC with acetonitrile/water (24/76
and 35/65, v/v, for the separation of mono- and di-sulfonates,
acetonitrile/water (35/65, v/v) as mobile phase for the separation
of di-sulfonates formed.
3.5. Reaction of AD-, AC-, or AB-isomer of
MessCl and product analysis
a-CD(Mess)2 with
respectively) as mobile phase. Retention times of
a-CD(Mess)1,
AB-isomer (103 mg, 0.077 mmol), AC-isomer (102 mg,
0.076 mmol), or AD-isomer (102 mg, 0.076 mmol) of -CD(Mess)2
and AD-, AC-, and AB-isomers of
5.1, and 6.9 min, respectively.
a-CD(Mess)2, were ca. 2.6, 3.6,
a
was dissolved in pyridine (25 mL) and the solution was stirred in
an ice-water bath (2 °C). MessCl (39 mg, 0.18 mmol) was added
to the stirred solution. At hourly intervals, 2.0 mL aliquots were
withdrawn, added to 0.2 mL water to stop the reaction, and ana-
lyzed by UFLC with acetonitrile/water (43/57, v/v) as mobile phase
for the separation of tri-sulfonates formed. Retention times of ABE-,
3.4. Reaction of
a-CD(Mess)1 with MessCl and product analysis
a
-CD(Mess)1 (105 mg, 0.091 mmol) was dissolved in pyridine
(25 mL), and the solution was stirred in an ice-water bath (2 °C).
MessCl (38.6 mg, 0.177 mmol) was added to the stirred solution
At hourly intervals, 2.0 mL aliquots were withdrawn, added to
0.2 mL water to stop the reaction, and analyzed by UFLC with
ABD-, ABC-, and ACE-isomers of
15.5, and 17.8 min, respectively.
a-CD(Mess)3 were ca. 12.6, 13.3,

K. Yoshikiyo et al. / Carbohydrate Research 401 (2015) 58–63
63
4. Watanabe, K.; Kitagishi, H.; Kano, K. Angew. Chem., Int. Ed. 2013, 52, 6894–6897.
5. Yamamura, H.; Suzuki, K.; Uchibori, K.; Miyagawa, A.; Kawai, M.; Ohmizo, C.;
Katsu, T. Chem. Commun. 2012, 892–894.
6. Yuan, D. Q.; Izuka, A.; Fukudome, M.; Rekharsky, M. V.; Inoue, Y.; Fujita, K.
Tetrahedron Lett. 2007, 48, 3479–3483.
Supplementary data
Supplementary data (chromatograms, 1D 1H NMR, 2D COSY, 2D
ROESY, HOHAHA spectra) associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.carres.2014.
10.027.
7. Lurie, I. S.; Klein, R. F. X.; Dalcason, T. A.; Lebelle, M. J.; Brenneisen, R.;
Weinberger, R. E. Anal. Chem. 1994, 66, 4019–4026.
8. Khan, A. R.; Forgo, P.; Stine, K. J.; D’Souza, V. T. Chem. Rev. 1998, 98, 1977–1996.
9. Tsujihara, T.; Kurita, H.; Kawazu, M. Bull. Chem. Soc. Jpn. 1977, 50, 1567–1571.
10. Fujita, K.; Matsunaga, A.; Imoto, T. J. Am. Chem. Soc. 1984, 106, 5740–5741.
11. Fujita, K.; Yamamura, H.; Matsunaga, A.; Imoto, T.; Mihashi, K.; Fujioka, T. J. Am.
Chem. Soc. 1986, 108, 4509–4513.
12. Matsui, Y.; Fujie, M.; Sakate, H. Bull. Chem. Soc. Jpn. 1988, 61, 3409–3415.
13. Yamamura, H.; Nagaoka, H.; Saito, K.; Kawai, M.; Butsugan, Y.; Nakajima, T.;
Fujita, K. J. Org. Chem. 1993, 58, 2936–2937.
References
1. Rizzarelli, E.; Vecchio, G. Coord. Chem. Rev. 1999, 188, 343–364.
2. Cotner, E. S.; Smith, P. J. J. Org. Chem. 1998, 63, 1737–1739.
3. Menuel, S.; Duval, R. E.; Cuc, D.; Mutzenhardt, P.; Marsura, A. New J. Chem.
2007, 31, 995–1000.