Diverse associations in the ternary systems of β-cyclodextrin, simple carbohydrates and phenyl derivatives of inorganic oxoacids

Article abstract of DOI:10.1016/j.carres.2010.12.014

Complex formation reactions of phenylboronic, phenylphosphonic, phenylarsonic and 4-aminophenyl arsonic acids with β-cyclodextrin (cycloheptaamylose, β-CD) and some simple carbohydrates (mannitol, sorbitol, glucose) have been studied using spectrophotomet

Full text of DOI:10.1016/j.carres.2010.12.014

Carbohydrate Research 346 (2011) 833–838
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Diverse associations in the ternary systems of b-cyclodextrin, simple
carbohydrates and phenyl derivatives of inorganic oxoacids
Zita Sebestyén a, , Katalin Máthé a, Ágnes Buvári-Barcza a, , Elemér Vass a, Ferenc Ruff a, Julianna Szemán b,
⇑
Lajos Barcza a
a Institute of Chemistry, Loránd Eötvös University, Pázmány sétány 1/A, H-1117 Budapest (H-1518 Budapest, PO Box 32), Hungary
b Cyclolab Research and Development Laboratory Ltd, Illatos ut 7, H-1097 Budapest (H-1525 Budapest, PO Box 435), Hungary
a r t i c l e i n f o
a b s t r a c t
Article history:
Complex formation reactions of phenylboronic, phenylphosphonic, phenylarsonic and 4-aminophenyl
arsonic acids with b-cyclodextrin (cycloheptaamylose, b-CD) and some simple carbohydrates (mannitol,
sorbitol, glucose) have been studied using spectrophotometric, potentiometric methods and solubility
measurements, supplemented with HPLC and IR analyses of the solid samples. Equilibrium constants
have been determined at ionic strength of 0.2 M (NaCl) and 25 °C. b-CD forms the most stable complexes
with the neutral, undissociated forms of the acids, the stability constants are as follows: phenylboronic
acid: 320 36, phenylphosphonic acid: 108 25, phenylarsonic acid: 97 4 and 4-aminophenyl arsonic
acid: 107 10. The stability constants for the b–CD-complexes of the ionic forms are much lower. Ternary
complexes of low stability could be detected in the case of phenylphosphonic acid and sorbitol with the
undissociated form and with glucose and the dianion. In more concentrated solutions phenylboronic acid
forms insoluble complexes with mannitol, sorbitol and b-CD. The solid phases obtained in the ternary
systems are predominantly mixtures of ester type 3:1 complexes with the carbohydrate and 1:1 inclusion
complex with the b-CD. No significant interaction has been found with glucose. The phenomena can be
explained by the differences in the structures of the components and by the changes in the H-bonding
network of b-CD on the complex formation.
Received 14 September 2010
Received in revised form 26 November 2010
Accepted 14 December 2010
Available online 26 January 2011
Keywords:
b-Cyclodextrin
Phenylboronic acid
Phenylphosphonic acid
Glucose
Sorbitol
Ternary complexes
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
can be attached to the complex containing an appropriate guest.6,7
This can be the explanation of the observation that in the case of
The recognition and separation of carbohydrates and related
compounds is important both from practical and theoretical points
of view. Work in this field is mainly done on empirical basis, rela-
tively few reliable data on the equilibria are available.
some insoluble drugs much higher increase of the solubility could
be achieved when CD-derivatives and hydroxy acids (tartaric acid,
citric acid) were applied together, than with the cyclodextrins by
themselves.8
Cyclodextrins (CD-s) are known to form stable inclusion com-
plexes with a large variety of guests bearing sufficiently hydropho-
bic moieties of appropriate size.1 Strongly hydrophilic compounds,
however, cannot be included, so no significant interaction could be
detected between b-cyclodextrin (cycloheptaamylose, b-CD) and
sugars, for example, glucose.2–4 On the other hand, the primary
and secondary alcoholic OH groups on the outer rims of the CD-
ring can take part in H-bonds, contributing either to the stabiliza-
tion of the inclusion complex,5 or to the interaction with further
molecules, thus resulting in the formation of ternary complexes.
Owing to the modification of the hydrogen bonding system, smal-
ler, hydrophilic molecules like acetic acid or even perchlorate ion
Carbohydrates themselves do not form complexes of significant
stability with cyclodextrins, as it has been proved by several
authors.2–4 The interaction of boric acid or its derivatives with car-
bohydrates, however, is a well known and thoroughly studied
fact.9–11 It has been shown that the reaction is preferred in alkaline
solutions, and results in an ester-like product containing C–O–B
bonds and tetrahedrally coordinated boron.11–13 Sugar alcohols
like mannitol or sorbitol and phenylboronic acid may form poorly
soluble 1:3 associates,11 while cyclic sugars can not bind more than
one or two borate or boronate ions because of the unfavourable
steric position of the OH-groups.13 The topic has gained a renewed
interest in recent years as the basis of sugar sensors,13–18 neverthe-
less, the association constants given for the same system spread
over a wide range (e.g., for phenylboronic acid–D-glucose between
⇑
K = 200 and K = 11 dm3 molꢀ1).11,14–16
Corresponding author.
E-mail address: buvari@chem.elte.hu (Ágnes Buvári-Barcza).
Present address: Institute of Pharmacy, Semmelweis University, 1092 Budapest,
Högyes E. u. 7, Hungary.
The interaction of boric acid or borate ion itself with cyclodex-
trins is also negligible,2 but the phenyl derivative [phenylboronic
0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2010.12.014

834
Z. Sebestyén et al. / Carbohydrate Research 346 (2011) 833–838
acid (PhBA) or phenylboronate ion] is expected to give a complex
via inclusion of the phenyl group. Thus the phenylboronate–
cyclodextrin and phenylboronate–carbohydrate interactions may
mutually influence each other and result in the formation of
ternary complexes, as it is proved—though with low stability—in
the case of benzoic acid.6
As phenylboronic acid takes part in the interaction with polyols
mainly as a tetrahedral boronate anion, this structure can be re-
lated to those of phenylphosphonic and phenylarsonic acids.
Phenylboronic acid is shown to have bacteriostatic activity11
and both phenylboronic and phenylarsonic acids are subjects of
pharmacological research as inhibitors of carbonic anhydrase en-
zymes.19,20 A further derivative, 4-aminophenyl-arsonic acid may
be used in the synthesis of antileukemic drugs.21
2.3. Potentiometric measurements
Aliquots of 5 ꢁ 10ꢀ3–2 ꢁ 10ꢀ2 M solutions of the acids were ti-
trated with 0.2 M carbonate free NaOH solution under stirring with
nitrogen, in the absence and in the presence of 5 ꢁ 10ꢀ3
–
1.4 ꢁ 10ꢀ2
M b-CD or/and 0.1–0.2 M carbohydrate. The ionic
strength was adjusted to 0.2 M by the addition of NaCl. The e.m.f.
was recorded with a Radelkis OP 208/1 type pH-meter fitted with
a Radelkis OP 0808 P combined glass electrode and a Schott-Gerdte
T80/20 automatic burette. The system was calibrated by the titra-
tion of 20.00 mL 0.02 M HCl (0.18 M in NaCl) solution with 0.2 M
NaOH.
The temperature was kept constant at 25 0.5 °C in all types of
measurements.
In the present paper investigations on the interactions of the
above mentioned acids with b-cyclodextrin and some simple
carbohydrates are reported—with special regard to the possible
formation of ternary complexes. Two open chain sugar alcohols:
2.4. Solubility measurements and analysis of the solid products
Solubility measurements were applied for phenylboronic acid
(PhBA) as it is a too weak acid and is the most sparingly soluble
among the acids investigated. Solid PhBA in a small excess with re-
spect to the expected solubility was equilibrated with pure water
and with solutions of b-CD or/and the chosen carbohydrates. The
concentration of the b-CD varied from 0 to 1.3 ꢁ 10ꢀ2 M (near
the solubility limit) and those of the carbohydrates from 0 to
0.25 M. The solubility of PhBA has been found higher (0.145 M)
than the values given in the manuals, merely the dissolution pro-
cess is extremely slow, because the molecules in the solid phase
are connected not only by hydrogen bridges but by the stacking
of the aromatic rings, which hinder hydration. Sonication of the
system can not be recommended since an unfilterable colloidal
solution is produced. Equilibration at 25 0.1 °C lasted at least
2–3 days. The concentration of the dissolved PhBA was determined
spectrophotometrically after appropriate dilution at k = 261 nm. (It
was checked in separate experiments that the absorbance was not
influenced by the other components present in the actual concen-
trations.) As in the presence of sorbitol and mannitol and of higher
concentrations of P-CD a decrease in the solubility and perceptibly
the formation of a new phase was observed, the solid phase was fil-
tered, washed quickly with a small amount of water and dried at
105 °C to constant mass. The composition was analysed by HPLC,
and for PhBA spectrophotometrically, too, after dissolution.
Infrared spectra were recorded with a Bruker PM-37 instrument
in nujol, with optical elements made of barium-fluoride. In spite of
the high absorption of nujol in the frequency range of 2800–
3000 cmꢀ1, and at 1460 cmꢀ1 and 1380 cmꢀ1, the KBr pellet
technique could not be applied because of the definite interaction
between bromide and CD.8
D-mannitol and
D-sorbitol and the cyclic sugar D-glucose have been
chosen as model carbohydrates.
The complex formation equilibria were studied using pH-
potentiometric titrations, spectrophotometric method based on
the competing reaction with an appropriate indicator (phenol-
phthalein in alkaline and methyl orange in acidic medium),22,23
and solubility measurements. The solid phases were analysed by
means of infrared spectroscopy and HPLC.
2. Experimental
2.1. Materials
b-Cyclodextrin was obtained from Cyclolab Ltd (Hungary) as a
gift within the frames of research collaboration, and was used after
recrystallization from hot water.
Phenylboronic acid, phenylphosphonic acid and 4-aminophenyl-
arsonic acid (Sigma–Aldrich) and phenylarsonic acid (Merck) were
of analytical grade and used without further purification. The con-
centrations of the solutions were checked by the potentiometric
titration.
D-Sorbitol and D-glucose were of pharmacopoeial grade (Ph. Hg.
VII), all the other materials were commercial products of analytical
grade (Reanal, Hungary) and used without purification, except
phenolphthalein which was recrystallized twice from an etha-
nol–water solution.
Carbonate free NaOH-solution was prepared by the Sorensen
method. All the solutions were made in double distilled water.
TheHPLC analysis was performed in Cyclolab Ltd, on an Agilent
1100 HPLC system equipped with an Alltech 3300 Evaporative
Light Scattering Detector (temperature: 70 °C, gas flow rate:
1.3 L/min). The further experimental conditions were as follows:
2.2. Spectrophotometry
Spectrophotometry measurements were carried out with Spec-
tromom 195D and Camspec M330 instruments, in 10 mm cells.
In alkaline medium phenolphthalein was used as indicator in a
concentration of 3 ꢁ 10ꢀ3 M. The pH was adjusted with 0.02 M
Na2CO3. The concentration of b-CD varied stepwise from 0 to
2.4 ꢁ 10ꢀ4 M and that of the salts of the acids (prepared in solution
from weighed amounts of the acids by the addition of calculated
amounts of NaOH) from 1 ꢁ 10ꢀ2 M to 6 ꢁ 10ꢀ2 M. The absor-
bances were measured at k = 550 nm.
column: CD-Screen, 5
l
m, 4.0 ꢁ 250 nm (ChiroQuest Ltd, Budapest,
Hungary), eluents: A: water, B: acetonitrile, gradient program: 0–
5 min: 0% B, 5–20 min: B increasing linearly 0–50%, flow rate:
1 mL/min, temperature of the column: 30 °C. Sample preparation:
10 mg of the samples was dissolved in 10 mL of 20% acetonitrile–
water mixture and 20 lL was injected at once.
3. Results and discussion
3.1. Evaluation of the experimental data
The complex formation of the acidic forms was studied in 0.1 M
HCl solutions with methyl orange as indicatior, at wavelengths of
506 nm and 319 nm. The concentrations were as follows: methyl
The simplest equilibria for a divalent acid (supposing 1:1 com-
plexes only with the b-CD) can be symbolized as in Scheme 1.
where H2A stands for the undissociated acid, and b-CD and the car-
bohydrate are symbolized by D and C, respectively.
orange: 2 ꢁ 10ꢀ5 M, b-CD: 0–9 ꢁ 10ꢀ3
M and the acids: 0–
2 ꢁ 10ꢀ2 M. Further details of the spectrophotometric methods
and the method of the evaluation can be found in Refs. [24,25].

Z. Sebestyén et al. / Carbohydrate Research 346 (2011) 833–838
835
A similar equation can be given for cC if a third component is also
present, and
X
p
q
r
s
cH
¼
ðp þ 2qÞbpqrs½Hþꢂ ½H2Aꢂ ½Dꢂ ½Cꢂ
ð8Þ
The values of b0100 b0010 etc. are 1.0 by definition, and the corre-
sponding products give the equilibrium concentrations of the free
components.
Since [H+], cA and cD (and cC) are known, the values of bpqrs can
be computed using Eqs. 3–8 by an iterative computer program
with assumed starting values, searching for the best fit between
experimental data and the calulated values. (A more detailed
explanation of the method is described in Ref. 23.) In our case
the calculations justified the assumption applied in Scheme 1, that
is, q = r = s = 1.
As it is seen in Scheme 1, the equilibria are not independent of
each other, and the effects observed in the pH-potentiometric titra-
tions depend largely on the ratio of the equilibrium constants for
two consecutive processes. Therefore potentiometry itself can not
give definite values for the individual constants, it is advisable to
determine some of them by some independent method, too.
Scheme 1. Dissociation and complex formation equilibria of a divalent acid in
solutions contaning b-CD (D) and a simple carbohydrate (C).
In the case of phenylboronic acid the scheme is simplified, as
PhBA can behave as a monovalent acid only, but the acidic dissoci-
ation must be described as
H2A þ H2O () H2AðOHÞꢀ þ Hþ
rather than that in Scheme 1. Concerning the further interactions
either with the carbohydrate or with the CD, pH-potentiometric
measurements can not distinguish between the simple association
like that symbolized in Scheme 1 and the formation of an ester type
complex by the loss of H2O.
Under the conditions of the spectrophotometric measurements
only one species of acids is present predominantly, so the equilib-
ria are simplified to the association of that species with the CD and
the carbohydrate (and the competing reaction between CD and the
indicator). The method of the evaluation is described in details in
Refs. 24,25.
3.2. Stability constants from spectrophotometry and
potentiometry
The stability constants obtained from the spectrophotometric
measurements for some of the complex formation processes are
summarized in Table 1.
These data were used as starting values in the evaluation of the
potentiometric measurements. Dissociation constants of the acids
were taken from the literature26 and refined in the evaluation pro-
cess for the titration of the pure acids, to fit the applied conditions.
The results are summarized in Table 2.
In most general, the equilibria can be characterised as follows
(charges omitted):
pHþ þ qA þ r D þ s C () HpAqDrCs
ð1Þ
and the equilibrium constant:
It must be noted that the uncertainty of the equilibrium con-
stants for the maximally protonated forms of phenylphosphonic
acid (Ka1 or bꢀ1100) is unusually high because of the uncertainty
of potentiometric pH-measurements in the appropriate range
(1 < pH < 2), and this involves similar uncertainty in the corre-
sponding complex stability constants.. However, the equilibrium
½HpAqDrCsꢂ
s
bpqrs
¼
ð2Þ
p
q
r
½Hþꢂ ½Aꢂ ½Dꢂ ½Cꢂ
where square brackets denote equilibrium concentrations in
mol dmꢀ3 (M) units.
When H2A is chosen as the basic species, q means the number of
H2A units in the given particle and p is the number of additional
protons. E.g. for the A2ꢀ anion p = ꢀ2, and the formation of the
A2ꢀ. D complex can be described as:
constant for the second dissociation step (Ka2 = bꢀ2100/b
)
ꢀ1100
proved to be fairly constant in the calculations. On the other hand,
because of the successive complex formation steps, the effect of
the change in one equilibrium constant may be compensated
partly by the change in that of the next step. So the uncertainty
of the successive complex formation constants (K) is often lower
than those of the directly calculated overall values.
Binary interactions between the acids (except PhBA) and the
small carbohydrate molecules proved to be negligible: the calcula-
tions were insensitive to the supposed values and the correspond-
ing successive stability constants were <1.
H2A ꢀ 2Hþ þ D () A2ꢀ ꢃ D
2
½A2ꢀ ꢃ Dꢂ½Hþꢂ
bꢀ2110
¼
:
ð3Þ
½H2Aꢂ½Dꢂ
For the dissociation of the acid:
H2A () A2ꢀ þ 2Hþ
The stepwise equilibrium constants (K) derived from the overall
values of Table 2 for the formation of the different binary and ter-
nary associates are collected in Table 3.
2
½A2ꢀꢂ½Hþꢂ
bꢀ2100
¼
;
ð4Þ
½H2Aꢂ
consequently the stepwise formation constant of the A2ꢀꢃD complex
Table 1
can be obtained as:
Stability constants (K/dm3 molꢀ1
) of the b-CD-complexes obtained from the
2ꢀ þ D () A2ꢀ ꢃ D
½A2ꢀ ꢃ Dꢂ
spectrophotometric measurements
A
H3A+
H2A
HAꢀ
A2ꢀ
(H2AOHꢀ)
bꢀ2110
bꢀ2100
K
¼
¼
:
ð5Þ
ꢀ2110
½A2ꢀꢂ ꢃ ½Dꢂ
Phenylboronic acid
Phenylphosphonic acid
Phenylarsonic acid
4-Aminophenyl arsonic
acid
320 36 23.7 2.7
58 5a
29
5
110 30
12.5 1.6
16
The equations of the mass balances:
36 7b
5
X
p
q
r
s
cA
cD
¼
¼
qbpqrs½Hþꢂ ½H2Aꢂ ½Dꢂ ½Cꢂ
ð6Þ
ð7Þ
X
a
cca 79% H2A and 21% HAꢀ.
cca 90% H3A+ and 10% H2A.
p
q
r
s
rbpqrs½Hþꢂ ½H2Aꢂ ½Dꢂ ½Cꢂ
b

836
Z. Sebestyén et al. / Carbohydrate Research 346 (2011) 833–838
Table 2
Overall stability constants [b, for the definition see Eqs. 2–4 and footnote] obtained from the potentiometric measurements for the different possible associates in the acid-b-CD-
carbohydrate systems
Acid (H2A)
X
Y
11XYa
01XYa
ꢀ11XYa
ꢀ21XYa
Phenyl phosphonic acid
0
0
0
—
—
—
—
—
—
—
(1.0)
108 25
nd
250 50
nd
(2.6 0.5) ꢁ 10ꢀ2
(2.85 0.6) ꢁ 10ꢀ9
(8.5 1.0) ꢁ 10ꢀ8
(5.8 1.2) ꢁ 10ꢀ7
(2.3 0.5) ꢁ 10ꢀ8
nd
b-CD
b-CD
b-CD
b-CD
0
b-CD
0
b-CD
1.06 0.2
4.6 1.0
1.7 0.5
nd
Glucose
Sorbitol
Mannitol
0
0
0
0
Phenylarsonic acid
(3.95 0.04) ꢁ 10ꢀ4
(2.5 0.6) ꢁ 10ꢀ3
(1.11 0.04) ꢁ 10ꢀ4
(1.8 0.3) ꢁ 10ꢀ3
(1.63 0.02) ꢁ 10ꢀ12
(5.4 0.4) ꢁ 10ꢀ11
(2.08 0.14) ꢁ 10ꢀ13
(1.8 0.4) ꢁ 10ꢀ12
97
4
4-Amino-phenylarsonic acid
86
6
(2.7 0.6) ꢁ 103
107 10
nd: not detectable.
a
The equilibrium processes corresponding to the different items: 11XY: H++H2A+X+Y,H3A+ꢃX.Y, for example, when X = 0 and Y = 0, H+ + H2A,H3A+, when X = b-CD and
Y = 0, H+ + H2A + D,H3A+ꢃD 01XY: H2A + X + Y ,H2AꢃXꢃY, for example, when X = b-CD and Y = 0, H2A + D,H2AꢃD, when X = b-CD and Y = carbohydrate, H2A + D + C,H2AꢃDꢃC
ꢀ21XY: H2A + X + Y , A2ꢀꢃꢃY + 2H+, for example, when X = 0 and Y = 0, it means the overall dissociation constant of the acid, when X = b-CD and Y = carbohydrate,
H2A + D + C , A2ꢀꢃDꢃC + 2 H+.
From the results it is clear that the stability constants of the bin-
ary complexes with b-CD are always the highest with the neutral,
undissociated forms, and the values are similar for the three acids.
This fact shows that the complex formation is mainly governed by
the inclusion of the aromatic ring, while it is hindered by the
strong hydration of polar groups. The comparison with the data
for 4-aminophenyl-arsonic acid suggests that for the additional
H-bonding interactions with the primary and secondary alcoholic
OH-groups of the CD, the presence of some proton donating
group(s) is more favourable than the completely unprotonated
form which could function as proton acceptor only, while the
hydration of the amino group causes further hindrance against
the inclusion.
The stability constant for the complex of the much weaker
phenylboronic acid (PhBA, see Table 1) is substantially higher rel-
ative to the other three acids and is closer to that of other weak
aromatic acids, for example, benzoic acid.27 This can be explained
again by the inhibiting role of the increased hydration of the stron-
ger acids, in agreement with the observation that the stability con-
stants of b-CD-complexes are often antiparallel to the solubility of
the guest.28
Concerning the ternary interactions with phenylphosphonic
acid and carbohydrates it is very interesting that no ternary com-
plex could be proved with mannitol, while an opposite trend was
found in the interactions with differently protonated forms of the
primary guest in the case of sorbitol and glucose. This can be
understood considering that glucose is more acidic,16,26 so it can
take part in H-bonds as proton donor, while sorbitol is rather a pro-
ton acceptor.
In the case of PhBA pH-potentiometry was not used because of
the very low acidity and because spectrophotometry was appropri-
ate for the investigation of the two possible forms. Concerning the
ternary complexes, some qualitative indications in spectrophotom-
etry experiments with methyl red showed a definite increase of the
stability of the PhBA–b-CD complex in the presence of sorbitol
(suggesting the formation of ternary complexes), and to a lesser
extent with mannitol, while no effect was caused by glucose. Sim-
ilar conclusion can be drawn from the absorption spectra of PhBA
recorded in pure water and in the presence of b-CD (6 ꢁ 10ꢀ3 M)
or/and the carbohydrates (0.1 M). b-CD causes a small increase in
the absorbances and a small bathochromic shift near 272 and
266 nm, and the fine structure of the spectrum becomes more ex-
pressed, which can be attributed to the less polar environment in
the CD-cavity. Glucose has no influence on the spectra at all, nei-
ther in itself nor in the presence of b-CD. Mannitol and sorbitol
causes changes in the absorbances similar to the effect of b-CD
(mannitol h sorbitol) but, in contrast to that, accompanied by a
slight coalescence of the fine structure. When sorbitol and b-CD
are present together, the increase in the absorbances is higher than
the sum of the individual effects. However, these changes are
small, and the evaluation of spectrophotometric experiments for
ternary complexes is very complicated and uncertain because of
the large number of interdependent equilibria. Therefore, solubility
measurements were chosen to get further information on the
phenylboronic acid–b-CD–carbohydrate systems.
3.3. Solubility experiments, composition of the precipitates
The results of the solubility experiments could not be evaluated
to obtain complex stability constants. Glucose had no influence on
the solubility of PhBA at all. In the presence of mannitol and sorbi-
tol, an insoluble new phase appeared at rather low concentrations
already (5 ꢁ 10ꢀ2 M and 1 ꢁ 10ꢀ3 M, respectively) and, following a
short increasing range (cCD < 1 ꢁ 10ꢀ3 M), in the presence of b-CD,
too. In ternary systems the effects of b-CD and the carbohydrate
seemed to be summarized.
Table 3
Stepwise formation constants (K/dm3 molꢀ1) of the differently protonated acid-b-CD-carbohydrate associates [derived from the data of Table 2, according to equations like Eq.
(5)]
Association process
H3A+
H2A
HAꢀ
A2ꢀ
(H3A+ + D , H3A+ꢃD)
(H2A + D , H2AꢃD)
(HAꢀ + D , HAꢀꢃD)
(A2ꢀ + D , A2ꢀꢃD)
4-Aminophenyl-arsonic acid + b-CD
30
5
107 10
17
4
9
2
Phenylarsonic acid + b-CD
97
4
6.3 1.5
33 2.4
Phenylphosphonic acid + b-CD
108 25
40 10
30
8
Phenylphosphonic acid ꢀ b-CD complex + carbohydrate
(H3A+ꢃD + C , H3A+ꢃDꢃC)
(H2AꢃD + C , H2AꢃDꢃC)
(HAꢀꢃD + C , HAꢀꢃDꢃC)
(A2ꢀꢃD + C , A2ꢀꢃDꢃC)
Mannitol
Sorbitol
Glucose
nd
nd
1.6 0.5
4.3 1.0
nd
<1
6.8 1.0
2.3 0.4
nd
nd: not detectable.

Z. Sebestyén et al. / Carbohydrate Research 346 (2011) 833–838
837
The composition of the precipitates was determined by HPLC
and spectrophotometry. The PhBA-content from the HPLC-results
was in good agreement with those of the photometric determina-
tion. The sum of the total amounts of the components [calculated
with C6H5B(OH)2] was always >100%, but the excess correlates
fairly well with the PhBA-content, so it can be attributed to the loss
of H2O during the ester type complex formation (see IR spectra la-
ter). The results in molar units are collected in Table 4.
The molar ratios nPhBA/nb-CD = 1.18 0.02 for both the pure
PhBA–b–CD and the PhBA–b-CD–glucose systems (no glucose
could be detected in the latter sample). It might mean the presence
of 1:1 complex dominantly, accompanied by some undissolved
free acid, but the reproducibility of the value >1 suggests rather
the formation of some (probably outer sphere type, H-bonded or
ester-like) 2:1 complex, too, as it was proved earlier with benzoic
acid.27
The further data of Table 4 show that in the ternary systems the
molar ratios of any two of the components do not have any stoi-
chiometrical meaning, the precipitates can not be regarded as stoi-
chiometrically uniform ternary complexes. Supposing the presence
of 1:1 PhBA–b-CD complex and subtracting the corresponding
amount of the acid from the total, however, the rest gives
3.13 0.08 for the nPhBA/ncarbohydrate ratio with sorbitol and
3.22 0.08 with mannitol. (Though the homogeneity of the latter
sample was poorer, the amounts of the acid and mannitol changed
parallel to each other and antiparallel with the CD-content.)
From the reproducibility of the (nA ꢀ nD)/nC ratio and its good
agreement with the nA/nC obtained for the binary PhBA-sorbitol
system it can be concluded that the solid phases obtained in the
ternary systems with mannitol or sorbitol are mainly the mixtures
of 1:1 PhBA–b-CD and 3:1 PhBA–carbohydrate complexes. The sol-
ubility of the mannitol-complex is higher than that of the sorbitol-
complex, therefore, the ratio of the b-CD complex is higher in the
former system. The excess PhBA may be present as a 2:1 PhBA–
b-CD complex or interact with two monomeric units forming
bridges between them, attached by stacking of the aromatic rings
to the carbohydrate complex or simply accommodated in intersti-
tial positions.
Fig. 1. Infrared spectra recorded in nujol: (a) phenylboronic acid, (b) sorbitol, (c)
the precipitate obtained on the dissolution of phenylboronic acid in 0.1 M solution
of sorbitol.
The spectrum of the PhBA–b-CD system (Fig. 2b) is dominated
by b-CD. The stretching vibrations of the C–H bonds of the aro-
matic ring (3100–3000 cmꢀ1) are very suppressed (even if the
relatively smaller PhBA-content of the sample is taken into
account), while the C–H deformation and the aromatic ring vibra-
tions are somewhat shifted (from 703 cmꢀ1 to below 700 cmꢀ1
and 1608 cmꢀ1?1601 cmꢀ1, respectively). The OH-band in the
region >3000 cmꢀ1 remains almost unchanged, though it becomes
somewhat narrower, that may be an indication of the modifica-
tion in the H-bond system of the b-CD. The same effect (in
addition to the decreased moisture content of the complex) can
be the source of some minor shifts and significant decreases in
the absorptions at 1648 cmꢀ1 and in the region <900 cmꢀ1. (b-
CD samples, even if dried, retake water very fast, while in the
case of inclusion complexes, water is excluded from the cavity.)
Summarizing the observations, the differences compared to the
spectra of the pure components indicate the formation of an
inclusion complex, without any ester formation. The decreased
solubility of the complex relative to that of the pure b-CD can
Glucose does not form any insoluble complex either with PhBA
itself or in the ternary system with b-CD.
3.4. Infrared spectra
In the IR spectrum of the precipitate obtained in the reaction of
PhBA and sorbitol (Fig. 1c), the OH-bands completely disappeared
[not only the characteristic, broad stretching band between 3400
and 3180 cmꢀ1 but the deformation vibrations (1000–700 cmꢀ1
)
as well], and significant changes can be observed in the region
of the C–O stretching (1200–1000 cmꢀ1), proving the formation
of ester-like C–O–B bonds. Accordingly, the strong absorption
bands of PhBA at 1348 cmꢀ1 and 1306 cmꢀ1 (B–O vibrations) are
shifted to 1342 cmꢀ1 and 1317 cmꢀ1, respectively, while the mC–B
(1442 cmꢀ1) and the vibration characteristic for the aromatic
C–C bonds (ꢄ1608 cmꢀ1) remain practically unchanged.
Table 4
Compositions of the precipitates obtained on the dissolution of phenylboronic acid in the solutions of other components (n, mol/100 g dry sample)
nA ꢀnD
Solution (concn)
Phenyl-boronic acid (A)
b-CD(D)
Carbohydrate (C)
nA/nD
nA/nC
nC/nD
nC
1. b-CD(6.4 ꢁ 10ꢀ3 M)
2. Sorbitol (0.1 M)
3. b-CD + sorbitol
4. b-CD + mannitol
5. b-CD + glucose
0.100 0.002
0.627 0.031
0.458 0.004
0.29 0.06
0.0848 0.0006
1.18 0.02
0.209 0.029
0.138 0.004
0.072 0.023
<LODa
3.0 0.6
3.3 0.1
4.1 0.5
0.0245 0.0010
0.055 0.009
0.0831 0.002
18.7 0.9
5.4 1.9
1.19 0.02
5.6 0.4
1.4 0.6
3.13 0.08
3.22 0.08
0.0985 0.0005
a
LOD: limit of detection.

838
Z. Sebestyén et al. / Carbohydrate Research 346 (2011) 833–838
The cyclic saccharide glucose does not form any insoluble prod-
uct, and has no effect on the association equilibria of the other
components. This fundamental difference can be understood con-
sidering that the tetrahedral form of PhBA16 and the alpha-fura-
nose form of glucose13 are suitable for the interaction, and the
equilibrium concentrations of both are very low in neutral aqueous
solution.
Nevertheless, the stoichiometry of the associates dominating in
solution and in solid phase is not necessarily the same, owing to
the different solubilities, so these findings do not contradict the re-
sults obtained by spectrophotometry and potentiometry in more
diluted solutions where some low stability ternary complexes
could be detected.
In spite of formal similarities, the behaviour of phenylphos-
phonic acid and phenylboronic acid in the interactions is substan-
tially different. The much stronger acidity and real dissociation of
the former prevents direct association with the small carbohy-
drate molecules, while the inclusion complex with b-CD can take
part in further interactions to form ternary complexes of low
stability. On the other hand, the structural differences of the small
carbohydrate molecules lead to significantly different complex
forming behaviour.
Acknowledgement
We are grateful to Cyclolab Ltd for supplying the b-CD and to
the Hungarian Research Foundation (OTKA T-32470) for the partial
financial support.
References
Fig. 2. Infrared spectra recorded in nujol: (a) b-CD, (b) b-CD-phenylboronic acid
complex, (c) the precipitate obtained in the b-CD-phenylboronic acid-sorbitol
system.
1. Cyclodextrin Technology; Szejtli, J., Ed.; Kluwer Academic Publishers.:
Dordrecht/Boston/London, 1988; pp 79–84.
2. Paal, T. L.; Szejtli, J. Acta Chim. Acad. Sci. Hung 1981, 106, 9–15.
3. Hacket, F.; Coteron, J. M.; Schneider, H.-J.; Kazachenko, V. P. Can. J. Chem. 1997,
75, 5254.
4. Tarnai, M.; Buvári-Barcza, Á.; Barcza, L. J. Inclusion Phenom. Macrocycl. Chem.
1999, 34, 311–319.
5. Buvári, Á.; Barcza, L. Acta Chim. Acad. Sci. Hung 1989, 126, 455–462.
6. Buvári-Barcza, Á.; Csampai, A.; Barcza, L. J. Inclusion Phenom. Macrocycl. Chem.
2002, 42, 209–212.
7. Buvári, Á.; Barcza, L. J. Inclusion Phenom. Mol. Recognit. 1989, 7, 379–389.
8. Fenyvesi, É.; Vikmon, M.; Szeman, J., et al J. Inclusion Phenom. Macrocycl. Chem.
2000, 36, 355–370.
9. Kuivala, H. G.; Keough, A. H.; Soboczenski, E. J. J. Org. Chem. 1954, 19, 780–783.
10. Torsell, K. Arkiv för Kemi 1957, 10, 541–547.
be attributed to the modified hydrogen bonding ability caused by
the b-CD–PhBA interaction.
In the IR spectrum of the PhBA–sorbitol–b-CD ternary system
(Fig. 2c) the characteristic features of both the PhBA-sorbitol and
PhBA–b-CD systems can be observed, as if a somewhat decreased
PhBA-sorbitol spectrum were superimposed on
a decreased
PhBA–b-CD spectrum.
Similar observations can be made in comparison of the IR
spectra of the PhBA–mannitol–b-CD system and those of the pure
components and binary systems: the characteristic absorptions of
PhBA appear somewhat shifted, with relatively low intensity, the
broad OH-band shows the shape of that of the b-CD-complex and
the OH-bands of mannitol disappeared.
11. Lorand, J. P.; Edwards, J. O. J. Org. Chem. 1959, 24, 769–774.
12. Paál, T. L. Acta Chim. Acad. Sci. Hung 1981, 106, 71–81.
13. Norrild, J. C.; Eggert, H. J. Am. Chem. Soc. 1995, 117, 1479–1484.
14. Arimori, S.; Ward, C. J.; James, T. D. Tetrahedron Lett. 2002, 43, 303–305.
15. Springsteen, G.; Wang, B. Tetrahedron 2002, 58, 5291–5300.
16. Bosch, L. J.; Files, T. M.; James, T. D. Tetrahedron 2004, 60, 11175–11190.
17. Nicholls, M. P.; Paul, P. K. C. Org. Biomol. Chem. 2004, 2, 1434–1441.
18. Mader, H. S.; Wolfbeis, O. S. Microchim. Acta 2008, 162, 1–34.
19. Innocenti, A.; Firnges, M. A.; Antel, J., et al Bioorg. Med. Chem. Lett. 2004, 14,
5769–5773.
4. Conclusions
20. Dagger, P. S.; Smith, K. S.; Iverson, T. M.; Ferry, J. G.; Rees, D. C. J. Biol. Chem.
2001, 276, 10299–10305.
Summarizing the results, the IR spectra are in complete agree-
ment with the conclusions drawn from the results of the HPLC
analysis: the precipitates formed on the dissolution of PhBA in
solutions of open chain carbohydrates and b-CD consist mainly of
binary associates: ester type 3:1 complexes with the carbohydrate
and 1:1 inclusion complex with the b-CD. The sequence of the sol-
ubilities is: PhBA-sorbitol < PhBA-mannitol < PhBA–b-CD. The
presence of some minor amounts of outer sphere type (hydrogen
bonded) ternary associates can not be excluded.
21. Liu, Y. P.; Narla, R. K.; Uekun, F. M. Bioorg. Med. Chem. Lett. 2003, 13, 581–583.
22. Buvári, Á.; Barcza, L.; Kajtar, M. J. Chem. Soc., Perkin Trans. 2 1988, 1687–1690.
23. Csernák, O.; Buvari-Barcza, Á.; Samu, J.; Barcza, L. J. Inclusion Phenom.
Macrocycl. Chem. 2005, 51, 59–63.
24. Buvári, Á.; Barcza, L. J. Inclusion Phenom. Mol. Recognit. 1989, 7, 313–320.
25. Buvári, Á.; Bodnar-Gyarmathy, D.; Barcza, L. J. Inclusion Phenom. Mol. Recognit.
1994, 18, 301–306.
26. Dissociation Constants of Organic Acids in Aqueous Solution; Kortüm, G., Vogel,
W., Andrussow, K., Eds.; Butterworth: London, 1961.
27. Buvári, Á.; Szejtli, J.; Barcza, L. Acta Chim. Acad. Sci. Hung 1982, 110, 51–57.
28. Schlenk, H.; Sand, D. M. J. Am. Chem. Soc. 1961, 83, 2313–2320.