Formation of 1:1 and 2:1 host-guest inclusion complexes of α-cyclodextrin with cycloalkanols: A 1H and 13C NMR spectroscopic study

Article abstract of DOI:10.1016/j.molstruc.2014.05.051

Binding constants (K_a's) for the formation of inclusion complexes of α-cyclodextrin (α-CD) with cycloalkanols (c-CnOH; n = 4-8) were determined by means of ¹H and ¹³C NMR titration, under two different conditions: (i) only 1:1 host-guest inclusion complexes are formed when the guest is in excess; (ii) the formation of 2:1 inclusion complexes occurs only after that of 1:1 inclusion complexes, when the host is in excess. The results of this work showed that α-CD can include c-C4OH or c-C5OH only when the molar ratio is 1:1; larger ring-sized cycloalkanols such as c-C6OH, c-C7OH or c-C8OH can be included only when the molar ratio is 2:1. These findings, together with those obtained for the four derivatives of α-CD, per-6-O-methyl-α-CD, per-2-O-methyl-α-CD, per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD, suggested that α-CD forms 2:1 inclusion complexes with c-C6OH, c-C7OH or c-C8OH in a tail-to-tail manner, in which the secondary hydroxy sides of the two CD molecules face each other. Two-dimensional ROESY measurements confirmed our results.

Full text of DOI:10.1016/j.molstruc.2014.05.051

Accepted Manuscript
Formation of 1:1 and 2:1 host-guest inclusion complexes of α-cyclodextrin with
cycloalkanols: A 1H and 13C NMR spectroscopic study
Tomoki Akita, Keisuke Yoshikiyo, Tatsuyuki Yamamoto
PII:
S0022-2860(14)00565-1
DOI:
http://dx.doi.org/10.1016/j.molstruc.2014.05.051
MOLSTR 20650
Reference:
Journal of Molecular Structure
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 April 2014
19 May 2014
22 May 2014
Please cite this article as: T. Akita, K. Yoshikiyo, T. Yamamoto, Formation of 1:1 and 2:1 host-guest inclusion
1
13
complexes of α-cyclodextrin with cycloalkanols: A H and C NMR spectroscopic study, Journal of Molecular
Structure (2014), doi: http://dx.doi.org/10.1016/j.molstruc.2014.05.051
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Formation of 1:1 and 2:1 host-guest inclusion complexes of α-cyclodextrin with cycloalkanols: A
1H and 13C NMR spectroscopic study
Tomoki Akita, Keisuke Yoshikiyo, and Tatsuyuki Yamamoto*
Faculty of Life and Environmental Science, Shimane University, 1060, Nishikawatsu-cho, Matsue 690-8504, Japan
( Received
)
Tel+Fax: +81-(852)-32-6551
E-mail: tyamamot@life.shimane-u.ac.jp
Keywords α-Cyclodextrin, Cycloalkanol, Binding constant, 1:1 and 2:1 Inclusion complexes
Abstract
Binding constants (Ka’s) for the formation of inclusion complexes of α-cyclodextrin (α-CD) with
1
cycloalkanols (c-CnOH; n=4 ~ 8) were determined by means of H and 13C NMR titration, under two different
conditions: i) only 1:1 host-guest inclusion complexes are formed when the guest is in excess; ii) the formation of
2:1 inclusion complexes occurs only after that of 1:1 inclusion complexes, when the host is in excess. The results of
this work showed that α-CD can include c-C4OH or c-C5OH only when the molar ratio is 1:1; larger ring-sized
cycloalkanols such as c-C6OH, c-C7OH or c-C8OH can be included only when the molar ratio is 2:1. These
1

findings, together with those obtained for the four derivatives of α-CD, per-6-O-methyl-α-CD, per-2-O-methyl-α-CD,
per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD, suggested that α-CD forms 2:1 inclusion complexes with c-
C6OH, c-C7OH or c-C8OH in a tail-to-tail manner, in which the secondary hydroxy sides of the two CD molecules
face each other. Two-dimensional ROESY measurements confirmed our results.
Introduction
The stoichiometry of complexation of cyclodextrins (CDs) employed as hosts strongly depends on
specific chemical properties, including the shape and hydrophobicity of the guest as well as the size of the internal
cavity and/or the substituents of the host [1]. At concentrations of CDs significantly higher than those of the guests,
and for guests larger than the cavities of CDs, the guest may be included by several CD molecules at multiple
points. Our recent study on the formation of inclusion complexes with 6-O-α-D-glucopyranosyl-β-CD (G1-β-CD) and
D- or L-tryptophan clearly showed that the binding constants (Ka’s) for the 1:1 and 2:1 host-guest inclusion
complexes could be obtained by NMR titration [2]. For a kinetic study on the inclusion systems of α- or G1-β-CD
with dimethyl(nitrophenyl) phosphates or thiophosphates, the assumption that a 2:1 inclusion complex occurs was
needed for the curve-fitting analysis of the system [3]. In systems in which a dimerized α-CD (which has two
independent hydrophobic cavities per molecule) was used as a host, with an exceeding concentration of the guest,
the dimer included two alkanol molecules at a time in different cavities, forming 1:2 inclusion complexes [4]. 13C
NMR data obtained for the formation of inclusion complexes of α-CD with 1-alkanols or 1-alkanoates, which are
characterized by relatively long chains, suggested that the formation of 2:1 inclusion complexes occurred together
with that of 1:1 inclusion complexes [5]. These results indicated that the existence of ternary complexes should be
2

taken into consideration when investigating the interaction of a host with a relatively large guest.
We have recently demonstrated that some cycloalkanols, namely cyclohexanol, cycloheptanol, and
cyclooctanol, form precipitates in aqueous solutions with α-CD, but not with β- or γ-CD. Analysis of the precipitates
revealed the stoichiometry of these complexes to be 2:1; that is, 2:1 inclusion complexes could possibly be formed.
Ka’s for the complexation of cycloalkanols with CDs determined by isothermal calorimetry [6, 7] and indirect
visible-light absorbance titration were already reported [8]. In our opinion, however, the authors of these studies did
not investigate in detail the possibility of the formation of 2:1 inclusion complexes. To the best of our knowledge,
Ka’s for the 2:1 complexation of α-CD with cycloalkanols have not yet been reported; this relevant point needs to be
investigated with the use of appropriate techniques.
NMR spectroscopy is a powerful technique to determine Ka’s of inclusion complexes of CD with
various guests [9]. In a typical NMR titration experiment, the concentration of one of two components of the
complex (either the host or the guest) is fixed, while the changes of the concentration of the other component are
monitored. The chemical shift (δ) of an NMR signal of the component with the constant concentration increases
gradually while the inclusion complex is formed by the addition of its counterpart. The chemical shift change (Δδ) of
the NMR signal plotted against the molar concentration of its counterpart produces the NMR titration plot. If the
stoichiometry of the inclusion complex is known, then Ka’s can be estimated from the most fitted curve of the NMR
plot. Because of the number of atoms in the host-guest systems, plural NMR signals should be used. This approach
allows an independent and simultaneous estimation of a number of Ka’s from one single NMR titration experiment.
The closer to one another the obtained Ka’s are, the better the curve-fitting analysis is for the assumed
3

stoichiometry. If the existence of 2:1 inclusion complexes is assumed, the Ka value for 1:1 inclusion complexes
should be determined at those conditions for which only 1:1 inclusion complexes are formed. This can be
accomplished with an NMR titration experiment carried out at a low and constant concentration of the host, while
the guest is in excess. Conversely, the Ka’s for 2:1 inclusion complexes can be determined by performing NMR
titration experiments at a low and constant concentration of the guest, while the host is in excesses. If the obtained
Ka’s are not similar to one another, participation of more than two host molecules to the formation of the complex
has to be taken into consideration.
In the present study, we determined the Ka’s of inclusion complexes of α-CD with cycloalkanols by
1
means of H and 13C NMR titration experiments. In particular, the Ka’s for the formation of 1:1 inclusion
1
complexes were determined by H NMR titration, as this approach is faster than that based on 13C NMR; Ka’s for
2:1 inclusion complexes were obtained with 13C NMR titration, because 13C NMR spectra can provide sharply
separated signals for these systems, due to the presence of carbon atoms of cycloalkanols [5]. In addition, the
molecular orientation of the two host molecules in 2:1 inclusion complexes was estimated by two-dimensional
NMR measurements. Moreover, the importance of secondary hydroxy groups for the formation of 2:1 inclusion
complexes is discussed based on the results of the titration experiments using derivatives of α-CD, such as per-6-O-
methyl-α-CD, per-2-O-methyl-α-CD, per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD.
Experimental
Sample preparation
α-CDs were obtained from Ensuiko Sugar Refining Co., Ltd., and were used after being cleaned with
4

activated charcoal followed by freeze-drying. The per-6-O-methyl-α-CD, per-2-O-methyl-α-CD, and per-3-O-
methyl-α-CD compounds were synthesized as described in previous studies [10-13]; per-2,6-O-dimethyl-α-CD was
purchased from Wako Chemical Co. Cyclobutanol (c-C4OH), cyclopentanol (c-C5OH), cyclohexanol (c-C6OH),
cycloheptanol (c-C7OH) and cyclooctanol (c-C8OH) were purchased from Wako Chemical Co., and were used
after purification by distillation. All other chemical reagents used for this study were of reagent grade.
D2O solutions of α-CD and/or its derivatives as well as cycloalkanols were prepared and used for 1H and
13C NMR measurements. The molar concentration of CDs was kept constant and that of cycloalkanols was changed
to produce NMR titration plots (Experiment 1). Ka’s for the formation of 1:1 inclusion complexes determined in
this way were then used to obtain those of 2:1 inclusion complexes. For this purpose, the molar concentration of a
cycloalkanol was kept constant and that of the α-CD was changed (Experiment 2). The NMR titration plots were
then curve-fitted with the assumption of the formation of both 1:1 and 2:1 inclusion complexes.
1H NMR titration experiments with cycloalkanols in excess (Experiment 1)
The concentration of α-CD was fixed at 2 mmol dm-3; the concentration of the cycloalkanol was varied
from 0 to 40 mmol dm-3 and from 0 to 30 mmol dm-3 for c-C4OH, c-C5OH or c-C6OH, and for c-C7OH and c-
C8OH, respectively. The exact concentration of α-CD was 2.103, 2.097, 2.103, 2.171 and 2.309 mmol dm-3 for c-
C4OH, c-C5OH, c-C6OH c-C7OH and c-C8OH, respectively. Tetramethylammmonium chloride (TMA) at 10.1
mmol dm-3 was added to be used as an internal reference for 13C NMR signals. These solutions were inserted into
NMR-sample tubes with a diameter of 5.0 mm and shaken for a minute with a test tube mixer (Shibata Scientific
5

1
Technology Ltd., TTM-1). The H NMR spectra were recorded as described in the section entitled “1H and 13C
1
NMR spectral measurements”. The chemical shift change (Δδ) of the H NMR signal of H3 of α-CD was plotted
against the concentration of cycloalkanols. To estimate the Ka’s, these plots were curve-fitted with the assumption
that the formation of 1:1 inclusion complexes occurred [14]. Similar experiments were performed for the four α-CD
derivatives, e.g., per-6-O-methyl-α-CD, per-2-O-methyl-α-CD, per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-
CD, using c-C6OH or c-C7OH as guests.
13C NMR titration experiments with α-CD in excess (Experiment 2)
The concentration of cycloalkanol was fixed at 5 mmol dm-3, while that of α-CD was varied from 0 to 50
mmol dm-3. The exact concentrations of c-C4OH, c-C5OH, c-C6OH c-C7OH and c-C8OH were 5.354, 6.270,
6.190, 5.360 and 5.429 mmol dm-3, respectively. TMA (10.1 mmol dm-3) was added as an internal reference for 13
C
NMR. D2O solutions of c-C6OH, c-C7OH or c-C8OH with α-CD in excess produced white precipitates at room
temperature; these were dissolved at 323 K and re-cooled to room temperature. The solutions were then inserted
into NMR-sample tubes with a diameter of 5.0 mm, and mixed for 1 min with a test tube mixer. In order to avoid
the formation of white precipitates, the NMR tubes were heated at 323 K before the 13C NMR measurements. The Δδ
of the 13C NMR signals of the guest was plotted against the concentration of the host. These plots were curve-fitted
with the assumption that either only the formation of 1:1 or of both 1:1 and 2:1 inclusion complexes occurred [15].
Similar experiments were performed for the four α-CD derivatives, per-6-O-methyl-α-CD, per-2-O-methyl-α-CD,
per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD, using c-C6OH or c-C7OH as guests.
6

1H and 13C NMR measurements
1
One-dimensional H and 13C NMR measurements were performed with a 400 MHz JEOL JNM-A400
FT-NMR spectrophotometer at 298 K. To produce 1H NMR spectra, eight scans were recorded and averaged within
several minutes; 1370 scans were recorded and averaged to produce the 13C NMR spectra in about 50 min. This
record time was chosen to avoid reformation (estimated to occur in 1 h) of white precipitates in the solutions of c-
1
C6OH, c-C7OH or c-C8OH. For H NMR spectra, δwas determined using that of the HDO (4.650 ppm) signal as
an internal reference. Δδ for the H3 of α-CD, which gives a signal at 3.853 ppm in the absence of cycloalkanols, was
calculated. For 13C NMR spectra, the δvalue was determined using the 13C NMR signal of TMA (57.952 ppm) as an
internal reference. It was proven that the interaction of CD with TMA is very weak [14], i.e., the obtained δvalue is
reliable. δ values for the 13C NMR signals of the guest in the absence of α-CD were determined as follows: c-C4OH,
C1 68.89, C2 35.55, C3 14.42 ppm; c-C5OH, C1 76.18, C2 37.45, C3 25.86 ppm; c-C6OH, C1 73.04, C2 36.98, C3
26.52, C4 27.65 ppm; c-C7OH, C1 75.55, C2 38.95, C3 24.90, C4 30.42 ppm; c-C8OH, C1 74.99, C2 35.97, C3
24.96, C4 29.59, C5 27.45 ppm. Numbering for the hydrogen and carbon atoms of α-CD and those of cycloalkanols
is shown in Figure 1.
ROESY spectral measurements
The two-dimensional rotating-frame nuclear Overhauser effect spectroscopy (ROESY) was employed.
ROESY spectra were acquired with the pulse routine named “roesy” in the spectrophotometer software. Relevant
7

parameters were set as follows: mixing time as 250 ms, relaxation delay as 0.9359 s, number of scans as 1024,
spectral width as 8,000 Hz, and numbers of points as 512 for t2 and as 256 points for t1, This was followed by zero-
filling. The ROESY measurements were performed at two different concentrations of α-CD, e.g., 20 and 50 mmol
dm-3; the concentration of cycloalkanols was fixed at 20 mmol dm-3. Thus, the relative host-guest concentrations
obtained were 1:1 and 2.5:1. Sample solutions were heated up to 323 K and re-cooled to room temperature before
measurements to avoid the formation of white precipitates.
Results and Discussion
Determination of Ka’s
1
The H NMR signal of H3 in α-CD shifted as a consequence of the increase of the cycloalkanols
concentration (see Figure 2). The calculated Δδwas plotted against the molar concentration of c-C7OH to produce
the 1H NMR titration plot as shown in Figure 2. The Ka value for the formation of the 1:1 inclusion complexes was
determined using a method previously described [14]. The 13C NMR signals of the carbon atoms of cycloalkanols
shifted with an increase of the α-CD concentration (Figure 3). The calculated Δδwas plotted against the concentration
of α-CD to produce the 13C NMR titration plots for four carbon atoms of c-C7OH is also shown in Figure 3. For
each 13C NMR titration plot of the carbons of cycloalkanols, we initially curve-fitted the data under the assumption
that only 1:1 inclusion complexes were formed. The Ka’s determined by 1H NMR titration (Experiment 1) and 13
C
NMR titration (Experiment 2) together with the correlation coefficients are listed in Tables 1, 2, and 3. The
average values and standard deviations are also given for those determined with 13C NMR titration.
8

The calculated Ka value for c-C4OH was determined to be 35 mol-1dm3 (Experiment 1), and those of C1,
C2 and C3 were determined to be 36, 34, and 35 mol-1dm3 (Experiment 2), respectively (Tables 1 and 2). Thus, our
data strongly suggested that c-C4OH exclusively forms 1:1 inclusion complexes with α-CD in D2O. The correlation
coefficient (near to 1) confirmed this point. However, a worse correlation was found for C1 as compared to C2 and
C3; this may be due to the chemical shift change in the 13C NMR signal of C1, which is significantly smaller than
that of C2 or C3. The calculated Ka’s for c-C5OH were determined to be 34 mol-1dm3; those of C1, C2 and C3 were
determined to be 39, 35 and 35 mol-1dm3, respectively (Tables 1 and 2). These data indicated that when c-C5OH is
employed, only 1:1 inclusion complexes are formed. However, these findings are inconsistent with the results of
previous calorimetric experiments on inclusion complexes of α-CD with cyclopentane, which suggested the
formation of a 2:1 system [16]. The discrepancy between our results and those previously reported may be
attributed to the presence in our systems of the hydroxy group. In particular, when a cyclopentane molecule is
included in the hydrophobic internal cavity of α-CD, the counter side of the included guest is exposed to the
aqueous solution. This allows the access of another α-CD molecule to include the exposed moiety of the guest,
which is still hydrophobic. In contrast, in the case of c-C5OH, the hydroxy group is exposed to the aqueous
medium, and the exposed moiety is not hydrophobic enough to be included by another α-CD molecule. As a result,
only a 1:1 inclusion complex can be formed with c-C5OH and α-CD.
Assuming the exclusive formation of 1:1 inclusion complexes, and considering experimental conditions
in which the host was in excess, c-C6OH was found to produce scattered Ka’s. These were determined with 13C
NMR titration of C1, C2, C3 and C4 to be 77, 26, 57 and 27, respectively (data not shown), the average and the
9

standard deviation being 47 and 24 mol-1dm3, respectively. Although the correlation coefficients for C1, C2, C3
and C4 were found to be satisfactory ( 0.9973), the large standard deviation suggested that inclusion complexes
other than 1:1 are formed. Similarly, c-C7OH and c-C8OH produced scattered Ka’s.
Considering the large standard deviation of the Ka’s for the complexation of c-C6OH, c-C7OH and c-
C8OH, the assumption of an exclusive formation of 1:1 inclusion complexes may be not appropriated. Thus, white
precipitates were prepared by adding these three cycloalkanols to D2O with an excess of α-CD; the precipitates were
then collected and dried up to be dissolved again in D2O in the absence of α-CD; 1H NMR spectra were recorded to
establish whether the precipitate contained α-CD and cycloalkanols. The area intensity of the 1H NMR signal of H1
atom of α-CD was compared to the total area intensity of all hydrogen atoms of a cycloalkanol molecule, except for
that attached to C1 carbon of the cycloalkanol. The results showed that the molar ratio of the white precipitates for
c-C6OH, c-C7OH or c-C8OH with α-CD is 2:1, indicating that using an excess of α-CD results in the formation of
2:1 inclusion complexes.
Therefore, we re-fitted the 13C NMR titration plots with the new assumption that c-C6OH, c-C7OH or
c-C8OH may also form 2:1 inclusion complexes; in this new set of calculations, Ka of 1:1 and of 2:1 inclusion
complexes were defined as K1 and K2, respectively; by employing a curve-fitting procedure, K1 and K2 values were
determined for c-C6OH, c-C7OH or c-C8OH (Table 3) [15]. The Ka value for the formation of the 1:1 inclusion
complex, determined by 1H NMR titration (Experiment 1) with the guest in excess, was used as the initial value for
the curve-fitting calculation. As shown in Table 3, the standard deviations for K1 and K2 are smaller than those
obtained with the previous approach, indicating that c-C6OH, c-C7OH or c-C8OH can also form 2:1 inclusion
10

complexes in presence of an excess of α-CD.
Molecular orientation of the two α-CD molecules in 2:1 inclusion complexes
The formation of white precipitates in 2:1 inclusion complexes of c-C6OH, c-C7OH or c-C8OH with α-
CD suggested that a strong hydrogen bonding occurs between the hydroxy groups of two different α-CD molecules
to decrease the water solubility of the inclusion complex. Based on this finding, we explored the possibility that c-
C6OH, c-C7OH or c-C8OH may form white precipitates with α-CD derivatives, in which one of the three hydroxy
groups is substituted with a methoxy group in every glucopyranosyl unit; four α-CD derivatives were employed,
namely per-6-O-methyl-α-CD, per-2-O-methyl-α-CD, per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD. In
particular, per-6-O-methyl-α-CD lacks a hydroxy group in the primary rim of the α-CD in every glucopyranosyl
unit; per-2-O-methyl-α-CD and per-3-O-methyl-α-CD lack a hydroxy group on C2 and C3, respectively, in the
secondary hydroxy side of the α-CD in every glucopyranosyl unit; per-2,6-di-O-methyl-α-CD lacks two hydroxy
groups attached on carbons C6 and C2, in both of the primary and secondary hydroxy side of the α-CD in every
glucopyranosyl unit. If the two α-CD molecules interact with each other in a head-to-head manner (e.g., the primary
hydroxy sides face each other), the 2:1 inclusion complex cannot be formed, due to the a lack of hydroxy groups on
the primary hydroxy side. This in turn suggests that the calculated Ka’s, determined under the assumption of an
exclusive formation of 1:1 inclusion complex, should not scatter. Conversely, if the two α-CD molecules interact in
a tail-to-tail manner (e.g., the secondary hydroxy sides face each other), per-2-O-methyl-α-CD, per-3-O-methyl-α-
CD, and per-2,6-di-O-methyl-α-CD should not form 2:1 inclusion complexes. In such a scenario, Ka’s, determined
11

under the assumption of an exclusive formation of a 1:1 inclusion complex, are expected to be very similar to one
another.
1
To explore this hypothesis, H and 13C NMR titration experiments were performed for the four α-CD
derivatives. The Ka’s along with the correlation coefficients are listed in Table 4, 5, and 6. Only c-C6OH and c-
C7OH were investigated, as the results of these experiments are, in our opinion, sufficient to establish which one
between the primary and the secondary hydroxy group is the most important for complexation.
In the presence of an excess of the host, c-C6OH and per-6-O-methyl-α-CD produced a precipitate
1
similar to that of the native α-CD. The Ka’s were estimated with H NMR titration to be 77 mol-1dm3 (Table 4);
Ka’s determined using 13C NMR titration, under the assumption of an exclusive formation of 1:1 inclusion
complexes, were estimated to be 75, 37, 57, and 38 mol-1dm3 for C1, C2, C3, and C4, respectively (data not shown),
the average and the standard deviation being 52 and 18 mol-1dm3, respectively. These data confirmed that in these
conditions inclusion complexes other than 1:1 are formed. Thus, the 13C NMR titration plots were fitted again,
assuming the formation of both 1:1 and 2:1 inclusion complexes. The averaged values of K1 and of K2 for c-C6OH
with per-6-O-methyl-α-CD were calculated to be 81 3 and 85 6 mol-1dm3, respectively (Table 5). Similarly, the
combination of c-C7OH and per-6-O-methyl-α-CD produced scattered Ka’s, under the assumption of an exclusive
formation of a 1:1 inclusion complex, the averaged Ka value and the standard deviation being 24 and 29 mol-1dm3,
respectively. When the formation of 2:1 inclusion complexes was also taken into account, the averaged values of K1
and K2 were calculated as 97 14 and 130 16 mol-1dm3, respectively (Table 5). Thus, these results strongly
indicated that the formation of 2:1 inclusion complexes of c-C6OH or c-C7OH with per-6-O-methyl-α-CD is not
12

inhibited by the methoxy groups at the primary hydroxy side (Figure 4)
Interestingly, neither c-C6OH nor c-C7OH formed white precipitates with per-2-O-methyl-α-CD, per-3-
O-methyl-α-CD , or per-2,6-di-O-methyl-α-CD. The calculated Ka’s for these systems were found to be very similar
to one another, as shown in Tables 4 and 6. In particular, the average Ka’s and standard deviations determined
by13C NMR titration were calculated to be 69 5, 49 10, and 125 3 mol-1dm3 when c-C6OH was combined with
per-2-O-methyl-α-CD, per-3-O-methyl-α-CD, and per-2,6-di-O-methyl-α-CD, respectively. The Ka’s for c-C7OH
with per-2-O-methyl-α-CD, per-3-methyl-α-CD, and per-2,6-di-O-methyl-α-CD were estimated as 96 12, 32 3,
and 156 13 mol-1dm3, respectively (Table 6). These data clearly showed that only 1:1 inclusion complexes are
formed when these compounds are combined, and that per-O-methylation of the secondary hydroxy groups of the α-
CD can prevent the formation of 2:1 inclusion complexes when c-C6OH or c-C7OH are used (Figure 4).
Considering these findings together with those obtained for the native α-CD with cycloalkanols, we claim that c-
C4OH or c-C5OH can exclusively form 1:1 inclusion complexes with α-CD, while larger ring-sized cycloalkanols,
such as c-C6OH, c-C7OH or c-C8OH, form 2:1 inclusion complexes. The two α-CD molecules of 2:1 inclusion
complexes are thus expected to interact with each other in a tail-to-tail manner, via the formation of hydrogen
bonds between the secondary hydroxy groups.
To confirm these results, two-dimensional ROESY NMR spectra were collected. The cross peak
intensity in two-dimensional ROESY spectra, between 1H NMR signals of the hydrogen atoms, which belong to the
internal cavity of the CD and to the guest, can provide crucial information about the orientation of the host and/or
1
the guest in an inclusion complex. Thus, in this work, the cross peaks between H NMR signals of H3 or H5 of α-
13

CD and those of the cycloalkanols were examined. These hydrogen atoms are located in the internal cavity of the α-
CD in the vicinity of the secondary (hydrogen H3) and primary (hydrogen H5) hydroxy sides. As shown in Figure
1
5, the intensity of the cross peaks between the H NMR signals of H3 or H5 and that of c-C5OH hardly change
when the concentration of α-CD was increased from 20 to 50 mmol dm-3 (Figure 5a, b). This result suggested that
the c-C5OH molecule is included deeply enough in the internal cavity of an α-CD molecule to produce strong cross
peaks between both these hydrogen atoms. The two-dimensional ROESY spectra for c-C4OH showed virtually the
1
same results, however, the intensity of the cross peak between the H NMR signals of H5 and that of c-C7OH
significantly decreased when the α-CD concentration was increased from 20 to 50 mmol dm-3. This finding
indicated that the distance between H5 and c-C7OH increases, unlike that between H3 and c-C7OH (Figure 5c, d).
This can be well explained by the formation of a 2:1 inclusion complex in a tail-to-tail manner (Figure 4a). When α-
CD and c-C7OH have the same concentrations, the c-C7OH molecule cab be included in an α-CD molecule deeply
enough to produce cross peaks with H3 and H5 in a two-dimensional ROESY spectrum. In contrast, when the
concentrations of α-CD and c-C7OH are different enough (such as 50 and 20 mmol dm-3 respectively, in our study),
the c-C7OH molecule moves to the middle of the two α-CD molecules, to form a 2:1 inclusion complex in a tail-to-
tail manner, and consequently is stepped away from H5, although still crosses to H3. Based on the data reported in
this contribution, we suggest that this rearrangement is the cause for the decrease in the intensity of the cross peak
1
between H NMR signals of H5 and that of c-C7OH. The two-dimensional ROESY spectra of c-C6OH and c-
C8OH produced virtually the same results of c-C7OH, which support our conclusion that c-C6OH, c-C7OH and c-
C8OH form 2:1 inclusion complexes with α-CD in a tail-to-tail manner.
14

Conclusion
The Ka’s for the formation of 1:1 inclusion complexes of α-CD with cycloalkanols were determined by
1
means of H and 13C NMR titration experiments. The analysis of the Ka’s showed that α-CD forms 1:1 inclusion
complexes with c-C4OH or c-C5OH; in contrast, with larger ring-sized cycloalkanols, such as c-C6OH, c-C7OH or
c-C8OH, our data suggested that α-CD form 2:1 inclusion complexes. The Ka’s for the formation of both 1:1 and 2:1
inclusion complexes, K1 and K2, were also calculated, using the Ka obtained by 1H NMR titration as the initial value.
This was carried out under the assumption that the formation of 1:1 and 2:1 inclusion complexes occur
simultaneously and with an excess concentration of the host. These data together with those obtained with ROESY
measurements confirmed that the two α-CD molecules of 2:1 inclusion complexes interact with each other in a tail-
to-tail manner via hydrogen bonds between the secondary hydroxy groups,
Acknowledgments
We thank Professor Yoshihisa Matsui of the Shimane University for his kind support for the NMR
experiments and for the calculations of the binding constants. This study has been financially supported partly by
the Grand-In-Aid for Scientific Research of Japan Society for the promotion of Science for T. Y. and by the budget
for Project research of Shimane University.
15

References
[1] Funasaki, N., Yodo, H., Hada, S., Neya, S., Stoichiometries and equilibrium constants of cyclodextrin-
surfactant complexations., Bull. Chem. Soc. Jpn. 65 (1992) 1323-1330
[2] Akita, T., Matsui, Y., Yamamoto, T., A 1H NMR titration study on the binding constants for D- and L-
tryptophan inclusion complexes with 6-O-α-D-glucosyl-β-cyclodextrin. Formation of 1:1 and 2:1 (host : guest)
complexes., J. Mol. Struct. 1060 (2014) 138-141
[3] Nagata, T., Yamamoto, K., Yoshikiyo, K., Matsui, Y., Yamamoto, T., Kinetic study on the interactions of
cyclodextrins with organic phosphates and thiophosphates., Bull. Chem. Soc. Jpn. 82 (2009) 76-80
[4] Yoshikiyo, K., Ohta, H., Matsui, Y., Yamamoto, T., Okabe, Y., Complexation of a disulfide-linked α-
cyclodextrin dimer with 1-alkanols., J. Mol. Struct. 891 (2008) 420-422
[5] Ohtsuki, H., Kamei, K., Nagata. T., Yamamoto, T., Matsui, Y., 13C NMR spectroscopy on the complexation of
α-cyclodextrin with 1-alkanols and 1-alkanoate ions., J. Incl. Phenom. Macrocycl. Chem. 50 (2004) 25-30.
[6] Castronuovo, G., Elia, V., Niccoli, M., Velleca, F., Viscardi, G., Role of the functional group in the formation
of the complexes between α-cyclodextrin and alkanols or monocarboxylic acids in aqueous solutions. A
calorimetric study at 25℃., Carbohyd. Res. 306 (1998) 147-155
[7] Castronuovo, G., Elia, V., Iannone, A., Niccoli, M., Velleca, F., Factors determining the formation of
complexes between α-cyclodextrin and alkylated substances in aqueous solutions: a calorimetric study at 25℃.,
Carbohyd. Res. 325 (2000) 278-286
16

[8] Matsui, Y., Mochida, K., Binding forces contributing to the association of cyclodextrin with alcohol in an
aqueous solution., Bull. Chem. Soc. Jpn. 52 (1979) 2808-2814
[9] Schneider, H.-J., Hacket, F., Ruediger, V., Ikeda, H., NMR studies of cyclodextrins and cyclodextrin
complexes., Chem. Rev. 98 (1998) 1755-1785
[10] Takeo, K., Uemura, K., Mitoh, H., Derivatives of α-cyclodextrin and the synthesys of 6-O-α-D-gulcopyranosyl-
α-cyclodextrin., J. Carbohyd. Chem. 7 (1988) 293-308
[11] Nagata, T., Yoshikiyo, K., Matsui, Y., Yamamoto, T., Binding and catalytic properties of 2-O- and 3-O-
permethylated cyclodextrins., Bull. Chem. Soc. Jpn. 82 (2009) 196-201
[12] Ashton, P. R., Boyd, S. E., Gattuso, G., Hartwell, E. Y., Königer, R., Spencer, N., Stoddart, J. F., A novel
approach to the synthesis of some chemically-modified cyclodextrins., J. Org. Chem. 60 (1995) 3898-3903
[13] Bergeron, R. J., Meeley, M. P., Machida, Y., Selective alkylation of cycloheptaamylose., Bioorg. Chem. 5
(1976) 121-126
[14] Matsui, Y., Tokunaga, S., Internal reference compounds available for the determination of binding constants
for cyclodextrin complexes by 1H NMR spectrometry., Bull. Chem. Soc. Jpn. 69 (1996) 2477-2480
[15] Fielding, L., Determination of association constants (Ka) from solution NMR data, Tetrahedron 56 (2000)
6151–6170.
[16] Osajima, T., Deguchi, T., Sanemasa I., Association of cycloalkanes with cyclodextrins in aqueous medium.,
Bull. Chem. Soc. Jpn. 64 (1991) 2705-2709
17

Table 1. Ka’s (mol-1dm3) for the formation of the 1:1 inclusion complexes with α-CD and cycloalkanols in D2O at
1
298 K; these were determined by H NMR titration, with an excess concentration of the guest and under the
assumption that only the formation of 1:1 inclusion complexes occurred; r is the correlation coefficient.
Guest
c-C4OHa
35
c-C5OHa
34
c-C6OHa
c-C7OHb
54
c-C8OHb
44
Ka
r
53
0.9994
0.9997
0.9991
0.9994
0.9991
a : The concentrations of host and guest are 2 and 0-40 mmol dm-3.
b : The concentrations of host and guest are 2 and 0-30 mmol dm-3.
18

Table 2. Ka’s (mol-1dm3) for the formation of the 1:1 inclusion complex with α-CD and cycloalkanols in D2O at 298
K. These were determined by 13C NMR titration method, with an excess concentration of the host and under the
assumption that only the formation of 1:1 inclusion complexes occurred; r is correlation coefficient.
guesta
Numbering of
c-C4OH
Ka
c-C5OH
Ka
carbons
r
r
1
2
3
36
34
35
0.9687
0.9997
0.9997
39
35
35
0.9998
0.9998
0.9997
Ave. SD
35
1
36
2
a : The concentrations of host and guest are 0-50 and 5 mmol dm-3.
19

Table 3. Ka’s (mol-1dm3) for the formation of both 1:1 and 2:1 inclusion complexes with α-CD and cycloalkanols in
D2O at 298 K. These were determined by 13C NMR titration method, with an excess concentration of the host and
under the assumption that the formation of both 1:1 and 2:1 inclusion complexes occurred; r is correlation
coefficient.
guesta
Numbering
of carbons
c-C6OH
K2
c-C7OH
K2
c-C8OH
K1
51
58
53
57
r
K1
55
55
56
57
r
K1
K2
r
b
b
b
1
34
0.9991
0.9995
0.9995
0.9995
39
0.9978
0.9995
0.9987
0.9992
-
-
-
2
32
39
38
38
39
38
50
49
50
50
0.9994
0.9997
0.9997
0.9999
3
36
40
4
5
35
39
Ave. SD
55
3
34
2
56
1
39
1
38
1
50
1
a : The concentrations of host and guest are 0-50 and 5 mmol dm-3.
b : Data are not shown because of the weak the intensity of the signals.
20

Table 4. Ka’s (mol-1dm3) for the formation of the 1:1 inclusion complexes with α-CD derivatives and cycloalkanols
1
in D2O at 298 K. These were determined by H NMR titration method, with an excess concentration of the guest
and under the assumption that only the formation of 1:1 inclusion complexes occurred; r is correlation coefficient.
Guest
c-C6OHa
c-C7OHb
host
Ka
r
Ka
r
6-O-methyl-α-CD
2-O-methyl-α-CD
77
80
29
0.9987
0.9994
0.9984
87
88
48
0.9989
0.9993
0.9977
3-O-methyl-α-CD
2,6-di-O-methyl-α-CD
136 0.9997
125 0.9995
a : The concentrations of host and guest are 2 and 0-40 mmol dm-3.
b : The concentrations of host and guest are 2 and 0-30 mmol dm-3.
21

Table 5. Ka’s (mol-1dm3) for the formation of both 1:1 and 2:1 inclusion complexes with 6-O-methyl-α-CD and
cycloalkanols in D2O at 298 K. These were determined by 13C NMR titration method, with an excess concentration
of the host under the assumption that the formation of both 1:1 and 2:1 inclusion complexes occurred; r is
correlation coefficient.
guest
Numbering of
carbons
c-C6OHa
c-C7OHb
K1
82
83
84
77
K2
84
81
81
93
r
K1
116
96
K2
r
1
2
3
4
0.9971
0.9992
0.9996
0.9972
125
143
110
142
0.9945
0.9982
0.9984
0.9976
90
85
Ave. SD
81
3
85
6
97 14
130 16
a : The concentrations of host and guest are 0-30 and 3 mmol dm-3.
b : The concentrations of host and guest are 0-15 and 5 mmol dm-3.
22

Table 6. Ka’s (mol-1dm3) for the formation of the 1:1 inclusion complex with α-CD derivatives and cycloalkanols in
D2O at 298 K. These were determined by 13C NMR titration method, with an excess concentration of the host and
under the assumption that only the formation of 1:1 inclusion complexes occurred; r is correlation coefficient.
Guest
Host
Numbering of carbons
c-C6OH
Ka
c-C7OH
Ka
r
r
1
2
3
4
62
74
71
67
0.9983
0.9998
0.9997
0.9996
114
88
0.9990
0.9999
0.999
0.9999
2-O-methyl-α-CDa
3-O-methyl-α-CDb
89
93
Ave. SD
1
69
5
96 12
35
60
42
53
39
0.9995
0.9998
0.9993
0.9997
0.9996
0.9996
0.9998
0.9998
2
3
4
28
34
30
Ave. SD
1
49 10
32
174
3
d
d
-
-
0.9992
0.9998
0.9993
0.9991
2
3
4
127
128
122
0.9997
0.9997
0.9997
155
143
151
2,6-di-O-methyl-α-CDc
Ave. SD
125
3
156 13
a : The concentrations of host and guest are 0-40 and 5 mmol dm-3, respectively.
b : The concentrations of host and guest are 0-30 and 3 mmol dm-3 when c-C6OH was used; 0-40 and 5 mmol dm-3
when c-C7OH was used.
c : The concentrations of host and guest are 0-80 and 5 mmol dm-3 when c-C6OH was used; 0-40 and 5 mmol dm-3
when c-C7OH was used.
d : These data were not used for calculations because the Δδof the signal was too small.
23

Figure captions
Figure 1. Structures for the native α-CD, its four derivatives (top) and the five cycloalkanols (bottom). For the sake
of clarity, all hydrogen atoms are omitted, except for H3, which was used for the 1H NMR titration. The numbering
of the carbon atoms used for the 13C NMR titration is also shown.
Figure 2. Left: The 1H NMR spectra in the region of H3 signals of α-CD (2 mmol dm-3) in D2O at 298 K; the molar
concentration of c-C7OH is 0, 15, and 30 mmol dm-3 in a), b), and c), respectively. The arrow shows the position of
1
H3 signal. Right: The Δδof the H NMR signal for H3 of α-CD, plotted against the concentration of c-C7OH. The
solid line shows the calculated value.
Figure 3. Left: The 13C NMR spectra in the region of 13C NMR signals of C2, C3 and C4 carbon atoms of c-C7OH
(5 mmol dm-3) in D2O at 298 K; the molar concentration of α-CD is 0, 15, and 30 mmol dm-3 in a), b), and c),
respectively. The arrows show the positions of 13C NMR signals. The 13C NMR signal for C1, which appears
around 76 ppm, is not shown. Right: The Δδof the 13C NMR signals for C1, C2, C3 and C4 carbon atoms of c-
C7OH plotted against the concentration of α-CD. The carbon atom numbers correspond to those indicated in Figure
1.
Figure 4. a) Schematic drawing of the 2:1 inclusion complex of 6-O-methyl-α-CD with a larger ring-sized
cycloalkanol; the methoxy groups at the primary hydroxy side of the CDs are not incorporated in the formation of
24

the 2:1 inclusion complex, while the secondary hydroxy groups of the CDs are. Hydrogen bonds are formed
between the secondary hydroxy groups of the two α-CD derivatives, which interact in a tail-to-tail manner.
b) The stable 1:1 inclusion complex does not allow the access of another α-CD derivative, 2-O-methyl-α-CD, 3-O-
methyl-α-CD, or 2,6-di-O-methyl-α-CD. Because of the methoxy groups at the secondary hydroxy side, the
hydrogen bonding between the secondary hydroxy groups of the two CDs are not strong enough to stabilize the 2:1
inclusion complex.
Figure 5. Two-dimensional ROESY spectra for c-C5OH or c-C7OH in the presence of α-CD; [α-CD]:[c-C5OH] =
a) 20:20, b) 50:20, and [α-CD]:[c-C7OH] = c) 20:20, d) 50:20 mmol dm-3. In the one-dimensional spectra, the
region of 1H NMR signals for cycloalkanols is displayed on the horizontal axis, those for α-CD are displayed on the
vertical axis.
25

Figure 1. Akita et al.
26

Figure 2. Akita et al.
27

Figure 3. Akita et al.
28

Figure 4. Akita et al.
29