Potentiometric saccharide detection based on the pKa changes of poly(aniline boronic acid)

Article abstract of DOI:10.1021/ja0267371

A novel approach for the potentiometric detection of saccharides using poly(aniline boronic acid) is presented. A model is described in which the electrochemical potential is sensitive to the change in the pK_a of the conducting polymer as a result of boronic acid-diol complexation. In this system, boronic acid complexation is the mode of transduction and it is manifested as changes in the electrochemical potential of the polymer with remarkable selectivity. Characteristics of both transient and steady-state response associated with the complexation are discussed. The presence of Nafion and fluoride during the electrochemical polymerization of 3-aminophenylboronic acid are shown to impact the sensitivity and the stability of the electrode response. The sensor sensitivity is improved significantly by increasing the concentration of sodium fluoride during the polymerization. Finally, the nature of the selectivity of the boronic acid-diol reaction under these conditions is explored by using molecular orbital calculations.

Full text of DOI:10.1021/ja0267371

Published on Web 09/28/2002
Potentiometric Saccharide Detection Based on the pKa
Changes of Poly(aniline boronic acid)
Eiichi Shoji1a and Michael S. Freund*,1b
Contribution from the Molecular Material Research Center, Beckman Institute, Mail Code
139-74, California Institute of Technology, Pasadena, California 91125
Received April 30, 2002
Abstract: A novel approach for the potentiometric detection of saccharides using poly(aniline boronic acid)
is presented. A model is described in which the electrochemical potential is sensitive to the change in the
pKa of the conducting polymer as a result of boronic acid-diol complexation. In this system, boronic acid
complexation is the mode of transduction and it is manifested as changes in the electrochemical potential
of the polymer with remarkable selectivity. Characteristics of both transient and steady-state response
associated with the complexation are discussed. The presence of Nafion and fluoride during the
electrochemical polymerization of 3-aminophenylboronic acid are shown to impact the sensitivity and the
stability of the electrode response. The sensor sensitivity is improved significantly by increasing the
concentration of sodium fluoride during the polymerization. Finally, the nature of the selectivity of the boronic
acid-diol reaction under these conditions is explored by using molecular orbital calculations.
of the enzyme.9-11 However, many difficulties remain4 and there
is considerable interest in developing alternative sensing ap-
Introduction
One of the most successful and historically important
electrochemical approaches for detecting saccharides uses
enzyme-based electrodes. These have been under development
since the 1960s.2 To date, the in situ detection of saccharides,
including D-glucose and D-fructose, has been the focus of
considerable research due to the importance of saccharide
monitoring in such diverse areas as medical diagnostics3,4 and
bioprocessing.5 The use of enzyme-based monitoring approaches
in these areas has been hampered due to the fact that enzyme-
based sensors typically require mediators, their steady-state
response is a function of the mass transport of the analyte, and
they consume the analytes of interest. In systems using
oxidoreductases, oxygen/peroxide is widely used as the media-
tor;4 however, in this configuration a large excess of oxygen is
required for the limiting current to be proportional to the
concentration of the analyte, and fluctuations in oxygen
concentrations can result in variations in sensitivity. In recent
years these issues have been addressed through the use of
mediators other than oxygen6-8 and through the direct oxidation
proaches for the continuous monitoring of saccharides.
Sensors based on the complexation of boron compounds with
saccharides are an attractive alternative to enzymatic approaches.
The first reports of the complexation of saccharides with
boric12,13 and boronic13 acids appeared in the 1950s, and this
pioneering work has since attracted considerable attention. To
date, boronic acid-based detection schemes have proven im-
portant for saccharide detection.14-18 Boronic acid chemistry
has been employed in a number of sensing strategies including
direct pH measurements,19 fluorescence,20-22 UV-vis,23,24
circular dichroism,14,25 near-IR,26 surface plasmon resonance
(9) Kulys, J. J.; Samalius, A. S.; Svirmickas, G. J. S. FEBS Lett. 1980, 114, 7.
(10) Genas, N. K.; Kulys, J. J. Bioelectrochem. Bioenerg. 1980, 8, 103.
(11) Albery, W. J.; Bartlett, P. N.; Craston, D. H. J. Electroanal. Chem. 1985,
194, 223-235.
(12) Boeseken, J. AdV. Carbohydr. Chem. 1949, 4, 189-210.
(13) Kuivila, H. G.; Keough, A. H. J. Org. Chem. 1954, 19, 780.
(14) Tsukagoshi, K.; S., S. J. Org. Chem. 1991, 56, 4089-4091.
(15) Shiomi, Y.; Saisho, M.; Tsukagosi, K.; Shinkai, S. J. Chem. Soc., Perkin
Trans. 1 1993.
(16) Sandanayake, K. R. A. S.; Shinkai, S. J. Chem. Soc., Chem. Commun. 1994,
1083-1084.
(17) Dusemund, C.; Sandanayake, K. R. A. S.; Shinkai, S. J. Chem. Soc., Chem.
Commun. 1995, 333-334.
* Address correspondence to this author. E-mail: michael_freund@
umanitoba.ca.
(1) Current addresses: (a) Department of Human and Artificial Intelligent
Systems, Fukui University, 3-9-1 Bunkyo, Fukui 910-8507, Japan. (b)
Department of Chemistry, University of Manitoba, Winnipeg, MB R3T
2N2, Canada.
(2) Clark, L., Jr.; Lyons, C. Ann. N.Y. Acad. Sci. 1962, 102, 29.
(3) Gough, D. A.; Armour, J. C. Diabetes 1995, 44, 1005-1009.
(4) Wilson, G. S.; Hu, Y. Chem. ReV. 2000, 100, 2693-2704.
(5) Vaidyanathan, S.; Macaloney, G.; Vaughan, J.; McNeil, B.; Harvey, L. M.
Crit. ReV. Biotechnol. 1999, 19, 277-316.
(6) Lanniello, R. M.; Lindsay, T. J.; Yacynych, A. M. Anal. Chem. 1982, 54,
1980.
(18) James, T. D.; Sandanayake, K. R. A. S.; Shinkai, S. Nature 1995, 374,
345-347.
(19) Wu, W. H.; Greene, C. Clin. Chem. 1986, 32, 1193-1193.
(20) The boronic acid group can act as a quencher for the fluorescence of dyes
upon the change in hybridization from sp2 to sp3 (i.e., boronate anion),
which occurs as a result of saccharide complexation. This behavior has
been exploited in the development of many fluorescence approaches for
saccharide detection.
(21) Yoon, J.-Y.; Czarnik, A. W. J. Am. Chem. Soc. 1992, 114, 5874-5875.
(22) James, T. D.; Linnane, P.; Shinkai, S. Chem. Commun. 1996, 281-288.
(23) Shinmori, H.; Takeuchi, M.; Shinkai, S. Tetrahedron 1995, 51, 1893-
1902.
(24) Yamamoto, H.; Ori, A.; Ueda, K.; Dusemund, C.; Shinkai, S. Chem.
Commun. 1996, 407-408.
(25) Takeuchi, M.; Imada, T.; Shinkai, S. J. Am. Chem. Soc. 1996, 118, 10 658-
10 659.
(7) Cass, A. E. G.; Davis, G.; Francis, G. D.; Hill, H. A. O.; Higgins, I. J.;
Plotkin, E. V.; Scott, L. D. L.; Turner, A. P. F. Anal. Chem. 1984, 56, 677.
(8) Degani, Y.; Heller, A. J. Phys. Chem. 1987, 91, 1285-1289.
9
12486
J. AM. CHEM. SOC. 2002, 124, 12486-12493
10.1021/ja0267371 CCC: $22.00 © 2002 American Chemical Society

Potentiometric Saccharide Detection
A R T I C L E S
Scheme 1
aromatic system is part of a conducting polymer. The oxidation
state of a conducting polymer is readily varied and hence may
be used to tune the electronic properties of the polymer’s
backbone, thereby influencing the properties of the boronic acid
moiety and hence its complexation with saccharides. In addition,
since the complexation changes the electronic properties of the
boronic acid substituent, we can monitor the binding event by
measuring changes in the electrochemical potential of the
polymer.
Conducting polymers41 have indeed attracted considerable
attention for use in sensor applications.42 Polyaniline in particular
has been the focus of considerable research interest due to its
stability, high conductivity, ease of production, and pH depend-
ent behavior,43-45 as well as its mediation/catalytic abilities46-48
Although there have been numerous accounts in the literature
over the years on the preparation of functionalized polyanilines,
only recently has boronic acid-substituted polyaniline been
reported.49,50 In these reports, it has been shown that amino-
phenylboronic acid can be polymerized as a copolymer with
aniline or as a homopolymer through a fluoride-catalyzed
reaction.49,50 In the latter case, it was shown that complexation
of fluoride with the boronic acid moiety substantially reduced
the oxidation potential required for polymerization, thereby
eliminating deleterious side reactions that occur at more positive
potentials. This breakthrough has opened up the possibility of
developing new electrochemical approaches to saccharide
detection using poly(aniline boronic acid) (PABA).30
In this work we provide a detailed investigation into the nature
of the boronic acid-saccharide complexation in PABA and its
influence on the electrochemical potential of the polymer.
Specifically, we explore the effect of the changing inductive
and resonance effects that occur upon complexation on the
electrochemical potential. Our findings demonstrate that in
addition to transient fluctuations associated with short-lived pH
changes, complexation also results in a steady-state change in
the electrochemical potential that is dominated by a change in
the pKa of the polymer. Furthermore, we explore the nature of
the selectivity of the response in terms of the distribution of
saccharide isomers and the preferential complexation between
boronic acids and cis-cyclopentanediols. Molecular orbital
calculations are used to reveal the important parameters that
predict sensitivity in this system. Finally, the role of Nafion
and fluoride used in the polymerization on the sensitivity and
stability are explored.
spectroscopy,27 potentiometry,28-30 conductance measure-
ments,12,27,31 and quartz crystal microbalance measurements.27
These approaches overcome many of the issues that plague
enzymatic systems as described above. Most importantly, these
approaches are “reagentless” and their sensitivity is not de-
pendent on the mass transport of analyte.
The complexation of saccharides (as well as alkyl and
aromatic diols) with aromatic boronic acids produces a stable
boronate anion12 and a proton (1:1 stoichiometry) in the pH
range 6-10 (see Scheme 1). Since the complexation is revers-
ible,32,33 saccharides bind to aromatic boronic acids as a function
of the saccharide concentration in the physiological pH range
6.8-7.5. The percentage of complex produced with phenyl-
boronic acid at pH 7.4 is ca. 30% for D-glucose and ca. 80%
for D-fructose (boronic acid:saccharides ) 1:2 (molar ratio)).33
While the degree of complexation is significant at high molar
concentrations, for analytical applications in which the saccha-
ride concentration is relatively small, increasing the percentage
of boronate complex near neutral pH remains an important goal.
Since complexation involves a tetrahedral boronate anion as an
intermediate,34 binding constants are increased at higher pH.
Two approaches have been used successfully to increase binding
constants of phenylboronic acid with saccharides at neutral
pH: (1) the addition of neighboring amine groups35 or fluoride36
to form a stable covalent bond with the boron atom, thereby
driving the boron from sp2 to sp3 hybridization, and (2) the
addition of an electron-withdrawing group to the phenyl ring
to shift the pKa to a lower value.33 The latter approach illustrates
the interdependence of the complexation reaction and the
electronic structure of the adjoining aromatic system. This
relationship is further demonstrate by recent work where
complexation involving boronic acid-substituted ferrocene
resulted in a corresponding change in redox potential.29,37 This
shift in potential can be understood in terms of the formation
of the boronate anion, which can result in a change in the
inductive38 and resonance properties of the boron-containing
substituent.39,40
Given the above knowledge that complexation and the nature
of the aromatic system are interdependent, we have undertaken
a study of the complexation reactions occurring when the
Experimental Section
(26) Pringsheim, E.; Terpetschnig, E.; Piletsky, S. A.; Wolfbeis, O. S. AdV.
Mater. 1999, 11, 865-868.
Materials. 3-Aminophenylboronic acid hydrochloric salt, D-glucose,
D-fructose, R-methyl-D-glucoside, cis-1,2-cyclopentanediol, trans-1,2-
cyclopentanediol, cis-1,2-cyclohexanediol, trans-1,2-cyclohexanediol,
(27) Gabai, R.; Sallacan, N.; Chegel, V.; Bourenko, T.; Katz, E.; Willner, I. J.
Phys. Chem. B 2001, 105, 8196-8202.
(28) Kikuchi, A.; Suzuki, K.; Okabayashi, O.; Hoshino, H.; Kataoka, K.; Sakurai,
Y.; Okano, T. Anal. Chem. 1996, 68, 823-828.
(29) Moore, A. N. J.; Wayner, D. D. M. Can. J. Chem. 1999, 77, 681-686.
(30) Shoji, E.; Freund, M. S. J. Am. Chem. Soc. 2001, 123, 3383-3384.
(31) Arnold, F. H.; Zheng, W. G.; S., M. A. J. Membr. Sci. 2000, 167, 227-
239.
(41) Reddinger, J. L.; Reynolds, J. R. AdV. Polym. Sci. 1999, 145, 57-122.
(42) McQuade, D. T.; Pullen, A. E.; Swager, T. M. Chem. ReV. 2000, 2537-
2574.
(43) Focke, W. W.; Wnek, G. E.; Wei, Y. J. Phys. Chem. 1987, 91, 5813-
(32) Consden, R.; Stanier, W. M. Nature 1952, 169, 783-785.
(33) Barker, S. A.; Chopra, A. K.; Hatt, B. W.; Somers, P. J. Carbohydr. Res.
1973, 26, 33-40.
5818.
(44) Genies, E. M.; Lapkowski, M.; Tsintavis, C. New J. Chem. 1988, 15, 373-
377.
(34) London, R. E.; Gabel, S. A. J. Am. Chem. Soc. 1994, 116, 2562-2569.
(35) Wulff, G. Pure Appl. Chem. 1982, 54, 2093-2102.
(36) Westmark, P. R.; Valencia, L. S.; Smith, B. D. J. Chromatogr. A 1994,
664, 123-128.
(45) MacDiarmid, A. G.; Epstein, A. J. Faraday Discuss. Chem. Soc. 1989, 88,
317-332.
(46) Shouji, E.; Buttry, D. A. J. Phys. Chem. B 1999, 103, 2239-2247.
(47) Tatsuma, T.; Matsui, H.; Shouji, E.; Oyama, N. J. Phys. Chem. 1996, 100,
(37) Ori, A.; Shinkai, S. J. Chem. Soc. Chem. Commun. 1995, 1771-1772.
(38) Hine, J. Structural Effects on Equilibria in Organic Chemistry; Wiley: New
York, 1975.
14016-14021.
(48) Ohsaka, T.; Chiba, K.; Oyama, N. Nippon Kagaku Kaishi 1986, 3, 457-
464.
(39) Muetterties, E. L. The Chemistry of Boron and its Compounds; Wiley:
New York, 1967; pp 495-500.
(40) Matteson, D. S. In Progress in Boron Chemistry; Brotherton, R. J. S., H.,
Ed.; Pergamon Press: New York, 1970; Vol. 3; pp 117-176.
(49) Pringsheim, E.; Zimin, D.; Wolfbeis, O. S. AdV. Mater. 2001, 13, 819-
822.
(50) Nicolas, M.; Fabre, B.; Marchand, G.; Simonet, J. Eur. J. Org. Chem. 2000,
9, 1703-1710.
9
J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002 12487

A R T I C L E S
Shoji and Freund
and Nafion solution (perfluorinated ion-exchange resin, 5 wt % lower
aliphatic alcohol and 45% water) were purchased from Aldrich
Chemical Inc. Bulk distilled water was first filtered and ion exchanged
to yield 18.3 MΩ quality water using an EasyPure RF, Barnstead
Thermolyne model 7031. Phosphate buffer saline (PBS) stock solution
(pH 7.4) was purchased from EM Science. Unless otherwise noted,
purchased chemicals were used as received.
Electrochemical Measurements. Cyclic voltammetry (CV) was
performed at 25 °C using a EG&G 362 potentiostat with an EG&G
175 universal programmer and a Hewlett-Packard 7046B X-Y recorder.
Measurements were taken in a three-electrode cell configuration by
using a glassy carbon disk (BAS, 3.0-mm diameter) as the working
electrode, a Pt-coil auxiliary electrode, and a Ag/AgCl reference
electrode. Unless otherwise noted, all CV experiments were conducted
in a 0.5 M HCl aqueous solution and at a scan rate of 100 mV s-1
.
Electrochemical Production of PABA. PABA was produced
electrochemically on a glassy carbon electrode.50 The reaction procedure
and conditions were as follows: 3-aminophenylboronic acid hydro-
chloride salt (87 mg, 40 mM) and sodium fluoride (21 mg, 40 mM)
were dissolved in 0.2 M HCl solution 12.5 mL. Nafion solution (2
mL) was added and the solution mixture was vigorously stirred. The
potential was scanned between 0.0 and +1.1 V in the unstirred solution
until the charge in the cathodic scan reached 10 mC cm-2. The deep
bluish-green film that was obtained was washed with water and placed
in a 0.1 M HCl solution to verify its redox behavior with CV. The
potential window for CV measurement was 0.0-0.8 V. After verifying
that the redox response was stable, the potential was scanned to +0.8
V and held at that potential for 10 s. The electrode was then rinsed
with water, followed with PBS solution, and then soaked in PBS
solution for 24 h to allow the electrochemical potential to stabilize. At
this point, the polymer film was a deep blue color. PABA produced in
the presence of higher sodium fluoride concentrations (105 mg, 200
mM) was prepared in an identical fashion with the exception of the
absence of Nafion and the use of 0.5 M HCl.
Figure 1. Cyclic voltammogram of a GC electrode in (a) 0.5 M HCl blank,
(b) 40 mM phenylboronic acid plus 200 mM NaF in 0.5 M HCl, and (c) 40
mM aminophenylboronic acid plus 200 mM NaF in 0.5 M HCl and (d)
subsequent scans of (c). Scan rate: 100 mV s-1
.
electrode was used repeatedly (at least 5 times) before being rinsed
with water and equilibrated in PBS. No problems with adhesion were
observed.
Open-Circuit Measurements. Open-circuit potential measurements
(Eoc) were performed by using a CH Instrument, CHI-660 workstation
controlled by PC. Measurements were taken in a three-electrode cell
configuration with a polymer-modified (PABA) glassy carbon disk
(BAS, 3.0-mm diameter) as the working electrode, a Pt-coil (0.5-mm
diameter, 50-cm length) auxiliary electrode, and a Ag/AgCl reference
electrode. Unless otherwise noted, all open circuit experiments were
conducted in a stirred PBS (pH 7.4) solution. The electrode potential
was allowed to settle in PBS solution prior to performing all open-
circuit measurements.
Flow Cell and Pump Controlled Experiments. To examine the
performance of a PABA electrode under dynamic concentrations, an
HPLC quat pump (Hewlett-Packard, G1311A) with degasser (Hewlett-
Packard, G1322A) was used with a specially designed flow cell
fabricated in house. PABA was produced on the glassy carbon electrode
at the higher NaF concentration, as described above. Electrical isolation
between the pump and flow cell was required to reduce noise and to
obtain a stable and reliable response. To achieve this, two isolation
points were placed in the setup. Separate bottles containing PBS and
100 mM of D-fructose in PBS were connected, and the concentration
of the analyte was programmed to achieve 0, 5, 0, 10, 0, 20, 0, 10, 0,
5, 0 mM at a delivery rate of 1 mL min-1. Each concentration was
held constant for 10 min and changes in concentrations were achieved
within a 30-s time period.
Results and Discussion
Production of PABA. The impact of fluoride complexation
with aromatic boronic acids is illustrated by the significant
increase in the solubility of phenylboroninc acid in both neutral
and acidic solutions upon the addition of fluoride. The increase
in solubility is a result of the formation of the fluorinated
boronate anion.51 Figure 1a and b demonstrate that the fluori-
nated boronate anion species is electrochemically inert within
the potential window between -0.2 and +1.1 V. In contrast,
3-aminophenylboronic acid is readily oxidized in the presence
of fluoride at approximately +1.0 V (see Figure 1c). Continued
cycling of the potential results in the growth of PABA on the
electrode surface in agreement with previous reports.50
It has been found that the stability and pH range of poly-
aniline films can be enhanced by the addition of polyanioninc
materials such as Nafion.52 The electrochemical polymerization
of 3-aminophenylboronic acid was achieved both in a Nafion
suspension and at a Nafion-coated electrode, and there were no
significant differences in the polymerization behavior or the
redox characteristics of the films.
Open-Circuit Potential Measurements. Polyaniline 1 (see
Scheme 2) consists of benzenoid diamine and quinone diimine
groups.53-55 The distribution of these groups is a function of
Computational Calculations. Geometry optimization for each
possible quasi-stable structure was performed using MOPAC (AM1
and PM3) and MM2 in Chem3D software version 4.0 (CambridgeSoft
Corp.). RMS: 0.001 was set for the calculation of MM2, AM1, and
PM3 with all standard parameters.
Responses of Analytes. After the PABA electrode was allowed to
settle, the change in Eoc was monitored before and after the injection
of an aliquot of analyte stock solution by using a microsyringe. The
(51) Valeur, B.; Pouget, J.; Bourson, J.; Kaschke, M.; Ernsting, N. P. J. Phys.
Chem. 1992, 96, 6545-6549.
(52) Bartlett, P. N.; Wang, J. H. J. Chem. Soc., Faraday Trans. 1996, 92, 4137-
4143.
(53) Quillard, S.; Louarn, G.; Lefrant, S.; MacDiarmid, A. G. Phys. ReV. B:
Condens. Matter 1994, 50, 12498-12508.
9
12488 J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002

Potentiometric Saccharide Detection
A R T I C L E S
Scheme 2
the oxidation state of the polymer and the degree of protonation
of the polymer is a function of the pH (the pKa of protonated
amine and imine are 2.5 and 5.5, respectively).56,57 Since the
redox chemistry of 1 involves both electron and proton-transfer
processes, the electrochemical potential is sensitive to changes
in pH. More precisely, the dependence of the electrochemical
potential on pH can be described by a double-square scheme.58
Under conditions where the system appears to undergo a
thermodynamically reversible 2e-2H+ redox reaction, the
Nernstian expression simplifies to
Figure 2. (A) Response curve of a poly(anilineboronic acid) electrode as
a function of time upon addition of 6.8 mM: (a) R-Methyl-D-glucoside;
(b) D-glucose; (c) D-fructose in pH 7.4 PBS. (B) Response of poly-
(vinylphenylboronic acid) coated polyaniline upon two additions of 3.4 mM
D-fructose in pH 7.4 PBS.
Changes in the properties of a substituent group will therefore
influence the electrochemical potential of a proton coupled redox
reaction in two ways: (1) altering the formal potential of the
2+/QH2
redox couple E°′QH
and (2) altering the acid dissociation
2
constant, Ka.59 In the case of the conversion of the boronic acid
to the boronate complex by cis diol (Scheme 1), it is expected
that there will be changes in both the inductive and resonance
properties of the boron moiety.39 Specifically, formation of the
boronate complex leads to an increase in its inductive electron-
donating ability, while eliminating the mesometrically electron-
withdrawing nature associated with the vacant p-orbital on the
boron.40 Under conditions where dE/dpH ) 0 (e.g., both
oxidized and reduced forms are completely dissociated), an
increase in electron-donating ability of a substituent will
decrease the formal potential. This behavior has in fact been
observed for ferrocene boronic acid, where a -140 mV change
in formal potential is reported between the uncomplexed and
complexed species.29 On the other hand, increasing the electron-
donating ability of a substituent is expected to stabilize the acid
form (i.e., QH22+). As such, converting boronic acid into the
boronate anion complex is also expected to reduce the formal
potential as well as to reduce the Ka of the protonated quinone
diimine group. It becomes apparent that the two processes offset
one another as indicated in eq 7, although the net effect
(discussed below) will depend on the relative magnitude of the
two influences and is ultimately expected to be a function of
the binding constant and concentration of the analyte.
E ) E°′Q/QH + (RT/2F) ln[Q][H+]2/[QH2]
(1)
2
Alternatively, it can be shown that the electrochemical potential
will be a function of the Ka of the protonated quinone diimine
QH22+. Assuming a concerted 2e- 2H+ reaction, the acid-base
reaction associated with the quinone diimine group in eq 1 can
be written as
Q + 2H+ a QH2
(2)
2+
with the corresponding acid dissociation constant:
Ka ) [Q][H+]2/[QH2
]
(3)
2+
The reduction of the protonated quinone diimine structure is
given by
QH22+ + 2e- a QH2
(4)
with the corresponding Nernst expression:
E ) E°′QH
+ (RT/2F) ln[QH22+]/[QH2]
(5)
2+
/QH2
2
The influence of both the transient pH change and the net
thermodynamic effect of complexation on Eoc are shown in
Figure 2 where saccharides with varying binding constants (D-
fructose > D-glucose > R-methyl-D-glucoside)60 are added. The
The net proton coupled redox reaction (combining reactions 2
and 4) is then
Q + 2H+ + 2e- a QH2
(6)
E
oc spikes generated upon addition of a saccharide are attributed
to transient pH changes that occur upon complexation (see
Scheme 1 and eq 7). The rapid decrease in the Eoc is likely due
to diffusion of protons from the film into the bulk solution. In
fact, similar transient profiles are observed (see Figure 2b) using
a composite electrode consisting of poly(vinylphenylboroinic
acid)-coated polyaniline/GC electrode, where the polyaniline-
coated electrode acts as a pH-sensitive electrode.
Substituting the expression for [QH22+] from eq 3 into eq 5
yields the corresponding Nernst expression for the net reaction,
which is a function of proton concentration as well as Ka:
E ) E°′QH
+ (RT/2F) ln[Q][H+]2/Ka[QH2] (7)
2+
/QH2
2
The gradual change in Eoc observed in Figure 2a is due to
the net effect of complexation on the electrochemical potential.
The increase in potential suggests that Ka change plays a greater
(54) Boyer, M. I.; Quillard, S.; Rebourt, E.; Louarn, G.; Buisson, J. P.; Monkman,
A.; Lefrant, S. J. Phys. Chem. B 1998, 102, 7382-7392.
(55) Hatchett, D. W.; Josowicz, M.; Janata, J. J. Phys. Chem. B 1999, 103,
10 992-10 998.
(56) Menardo, C.; Nechtschein, M.; Rousseau, A.; Travers, J. P. Synth. Met.
1988, 25, 311-322.
(59) Zuman, P. Effects of substituents in quinonoid compounds. In Substituent
effects in organic polarography; Plenum Press: New York, 1967; pp 273-
308.
(60) Oshima, K.; Toi, H.; Aoyama, Y. Carbohydr. Lett. 1995, 1, 223-230.
(57) Pringsheim, E.; Terpetschnig, E.; Wolfbeis, O. S. Anal. Chim. Acta 1997,
357, 247-252.
(58) Meunierprest, R.; Laviron, E. J. Electroanal. Chem. 1992, 328, 33-46.
9
J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002 12489

A R T I C L E S
Shoji and Freund
Figure 3. Sensitivity of open circuit response for 10 mM addition of
D-glucose as a function of film thickness as indicated by the total charge
measured during the reduction of the deposited film. Polymerization
conditions: 40 mM 3-aminophenylboronic acid, 40 mM NaF electropoly-
merized through a Nafion thin film (2 µL of Nafion alcohol solution was
cast onto a 3-mm diameter glassy carbon electrode and dried under room
temperature for 2 h).
Figure 4. Saccharide response curve of a PABA electrode in pH 7.4 PBS
as a function of time. Additions resulted in the following series of
concentrations: (1) 3.4; (2) 6.8; (3) 10.2; (4) 13.6; (5) 17.0; (6) 20.4; (7)
23.8; (8) 40.8 mM. (a) D-Glucose and (b) D-fructose response curves of a
poly(anilineboronic acid) electrode in pH 7.4 PBS as a function of time.
(c) Calibration curves for D-glucose (b) and D-fructose (O) using a PABA
electrode and of D-glucose (2) for polyaniline coated electrode in pH 7.4
PBS.
role than the change in the formal potential under these
conditions. This is likely due to the increase in electron density
ortho to the boronate complex, directly impacting the Ka of the
protonated amine.39 Furthermore, Figure 2a shows that the
equilibrated Eoc is a function of the binding constant of the
saccharide.
occurs preferentially with the cis-hydroxyls and the anomeric
hydroxyl under neutral conditions, and that the geometry of the
five-membered ring is critical. Thus, it can be concluded that
the apparent sensitivity of a boronic acid-based sensor will also
be a function of the binding constants of the various isomers
and their relative distributions in solution (Vide infra).
Sensor Characteristics. Sensor stability and sensitivity were
influenced by both the presence of Nafion and the polymeri-
zation conditions. Initial polymerization conditions made use
of a 1:1 concentration of fluoride to monomer as reported
previously.50 It was found that incorporation of Nafion under
these conditions improves stability of the films, presumably due
to electrostatic interaction between PABA and Nafion. Delami-
nation of the films was observed in the absence of Nafion during
stirring of the solution, while the films grown in the presence
of Nafion adhered well. Composite sensors made in this fashion
demonstrated good selectivity and sensitivity consistent with
reported binding constants (i.e., D-fructose > D-glucose).
Sensitivity increased with the amount of polymer deposited over
a limited range (see Figure 3) and increased to a limiting value,
suggesting that the transition is likely associated with complete
coverage of the electrode surface.
The selectivity of the boronic acid complexation for various
saccharides is influenced in part by their isomeric composition
in aqueous solution.61 Specifically, when a saccharide is
dissolved in water, mutarotation, which can occur on a time
scale of minutes, leads to an equilibrium mixture of four
forms: R-pyranose, â-pyranose, R-furanose, and â-furanose. In
the case of D-glucose the distribution at equilibrium under
neutral conditions is 39.4, 60.2, 0.2, and 0.21%, respectively,62
whereas the distribution for D-fructose is 2.0, 68, 6.0, and 23%,
respectively.63 13C NMR studies have revealed a five-membered
ring containing 1,2-cis-hydroxyls is involved in the complexation
of D-glucose and D-fructose, corresponding to the â-furanose
form of D-fructose (23%)63,64 and the R-furanose form of
D-glucose (2%).65 These findings suggest that complexation
Figure 4 shows the calibration curves for D-fructose and
D-glucose. Although the selectivity and sensitivity of the boronic
acid complexation is clearly retained in the electrode response,
sensitivity is modest (ca. 1-4 mV mM-1). To investigate the
effect of fluoride on the polymerization and sensitivity of the
resulting electrodes, the concentration of sodium fluoride was
increased from 40 to 200 mM.66 Since the complexation of
fluoride with boronic acid leads to a negatively charged
species,67 PABA should have a self-doped structure.50 Further,
since the complexation shifts the formal potential of an outer-
sphere redox couple (ca. 59 mV pF-1) over a large concentration
range (from 1 mM to 5 M),17 fluoride concentration should have
a significant effect on the polymerization rate and in turn the
morphology68 and behavior of the polymer.69 Indeed, increasing
the concentration of sodium fluoride resulted in significant
negative shifts in the oxidation potential, thereby enhancing
polymerization rates, and is consistent with the formation of a
negatively charged substituent17 and an increase in pH due to
the formation of HF (pKa ) 3.1).
In addition to enhancing the electropolymerization process,
higher fluoride concentrations significantly increased the sen-
sitivity. As shown in Figure 5a, a 3-fold enhancement in
sensitivity is seen compared to the addition of the same quantity
(10 mM) of analyte to a polymer grown under low fluoride
conditions (e.g., Figure 4, addition 3). Reversibility of the
complexation (see Figure 5b) is also maintained as indicated
by a return to baseline. The excellent selectivity of electrodes
grown in the presence of higher sodium fluoride concentrations
(65) Norrild, J. C.; Eggert, H. J. Am. Chem. Soc. 1995, 117, 1479-1484.
(66) The solution was saturated at this concentration.
(67) Cooper, C. R.; Spencer, N.; James, T. D. Chem. Commun. 1998, 1365-
1366.
(68) Kim, Y. T.; Yang, H. G.; Bard, A. J. J. Electrochem. Soc. 1991, 138,
L71-L74.
(69) Desilvestro, J.; Scheifele, W. J. Mater. Chem. 1993, 3, 263-272.
(61) Angyal, S. J. Angew. Chem., Int. Ed. 1969, 8, 157-166.
(62) Maple, S. R.; Allerhand, A. J. Am. Chem. Soc. 1987, 109, 3168-3169.
(63) Norrild, J. C.; Eggert, H. J. Chem. Soc., Perkin Trans. 2 1996, 2583-
2588.
(64) Shull, B. K.; Spielvogel, D. E.; Head, C.; Gopalaswamy, R.; Sankar, S.;
Devito, K. J. Pharm. Sci. 2000, 89, 215-222.
9
12490 J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002

Potentiometric Saccharide Detection
A R T I C L E S
as well as structural changes in the polymer. However, it was
possible to estimate an effective diffusion coefficient from the
initial rate of sorption70 by using
Mt
1/2
4
π1/2
Dt
)
(8)
2
( )
M∞
l
where Mt is mass uptake as a function of time t, D is the
diffusion coefficient, and l is the film thickness. Assuming that
the measured electrochemical potential is proportional to mass
uptake and determining the slope of the linear portion of a plot
of Mt/M∞ versus t1/2 (Mt/M∞ < 0.5) and a film thickness of
approximately 0.5 µm (the film thickness was estimated on the
basis of total charge passed during polymerization71), an
effective diffusion coefficient of 2 × 10-12 cm2 s-1 was
estimated for D-fructose. This value is about 2 orders of
magnitude less than that reported for methanol diffusion through
polyaniline,72 which is consistent with both the larger molecular
size of fructose and the coupled complexation process. These
results suggest that the response is largely diffusion controlled
and that any factor that accelerates mass transport through the
film will shorten the response time.
It is interesting to note that the time required to reach steady
state is noticeably longer for D-glucose than for D-fructose (e.g.,
Figure 5B). This difference is likely due to a displacement of
the equilibrium among the different isomeric forms of D-
glucose.61 Complexation with boric acid is known to displace
the relative concentrations of various isomers of D-glucose Via
mutarotation.33,73,74 Since the percentage of the complexing form
of D-glucose (R-furanose, 0.2%) is small relative to that of
D-fructose (â-furanose, 23%) and since the rate of mutarotation
of D-glucose is significantly slower than that of D-fructose, it is
expected that tautomerization-induced response changes that
occur slowly as a function of time will be more significant in
the case of D-glucose relative to D-fructose.
Figure 5. PABA (film was made from 40 mM 3-aminophenylboronic acid,
200 mM NaF), (a) 10 mM D-fructose injection, (b) soaked in the PBS
solution (no fructose added). (c) Open-circuit potential measurement for
(i) ethyl alcohol, (ii) ethyleneglycol, (iii) R-methyl-D-glucoside, (iv)
D-glucose, and (v) D-fructose. Each concentration was 3.4 mM.
In addition to increased sensitivity, polymer films prepared
with higher concentrations of fluoride exhibited better stability
and adhesion properties. Removing Nafion from the system had
the effect of eliminating the potential spikes associated with
transient pH changes observed in Figure 2. This behavior is
likely due to the permselective nature of Nafion, in which the
negatively charged sulfonate groups attached to the polymer
backbone repel solution phase anions.75 By excluding phosphate
anions, the buffer capacity inside the film will be reduced,76
making the local pH changes more dramatic. This hypothesis
is supported by the return of potential spikes (similar to those
seen in Figure 2) upon dilution of the PBS solution by an order
of magnitude, thereby reducing its buffering capacity.
Figure 6. Sensitivity of PABA to D-fructose (b), D-glucose (O), cis-1,2-
cyclopentanediol (0), trans-1,2-cyclopentanediol (4), cis-1,2-cyclohexandiol
(9), and trans-1,2-cyclohexanediol (2) and in PBS.
Selectivity Based on Boronic Acid-Diol Interactions. To
determine the structural parameters that determine the sensitivity
is shown in both Figure 5B and Figure 6. Figure 5B demon-
strates that the sensor is relatively insensitive to linear alcohols
and 1,2-alkanediols. Similarly, Figure 6 shows that the sensor
response is insensitive to cyclic 1,2-alkanediols with the
exception of cis-1,2-cyclopentanediol.
The time dependence of the electrochemical potential upon
the addition of complexing diols did not follow a simple model
for non-steady-state diffusion into a thin film on top of an
impermeable substrate.70 This behavior is likely due to com-
plications associated with diffusion coupled with complexation
(70) Crank, J. In The Mathematics of Diffusion; Oxford University Press: Oxford,
1975; pp 47-49.
(71) Dinh, H. N.; Vanysek, P.; Birss, V. I. J. Electrochem. Soc. 1999, 146,
3324-3334.
(72) Ball, I. J.; Huang, S. C.; Wolf, R. A.; Shimano, J. Y.; Kaner, R. B. J.
Membr. Sci. 2000, 174, 161-176.
(73) Mazurek, M.; Perlin, A. S. Can. J. Chem. 1963, 41, 2403-2411.
(74) Acree, T. E. The chemistry of sugars in boric acid solutions. In
Carbohydrates in Solution; Gould, R. F., Ed.; American Chemical Society:
Washington, DC, 1971; Vol. 117; pp 208-219.
(75) Shu, C. F.; Anson, F. C. J. Am. Chem. Soc. 1990, 112, 9227-9232.
(76) Soldatkin, A. P.; Elskaya, A. V.; Shulga, A. A.; Jdanova, A. S.; Dzyadevich,
S. V.; Jaffrezicrenault, N.; Martelet, C.; Clechet, P. Anal. Chim. Acta 1994,
288, 197-203.
9
J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002 12491

A R T I C L E S
Shoji and Freund
Table 1. Calculated Dihedral Angles (θ)a and Interatomic
Distances (d)b
Figure 7. Dihedral angle for two hydroxy groups of cis-1,2-cyclopen-
tanediol and cis-1,2-cyclohexanediol.
of our approach, we performed molecular orbital calculations
on cyclic alkanediols. Phenylboronic acid is known to form
stable complexes with cis-1,2-cyclopentanediols but not with
trans-cyclopentanediols. However, complexation with 1,2-cyclo-
hexane is fairly weak for both cis- and trans-cyclohexanediols.12
As seen in Figure 6, the results obtained with our system are in
good agreement of these observations, i.e., that there is a
dramatic difference in sensitivity between cis- and trans-1,2-
cyclopentanediol and virtually no response for cis- and trans-
cyclohexanediol. This behavior has been attributed to the steric
constraints imposed by the different structures and this assertion
has been used to explain the remarkable selectivity of this
complexation for other diols and sugars.12,73
a Dihedral angle of O-C-C-O. b Interatomic distance between two
oxygen atom of hydroxy groups. c Oshima, K.; Toi, H.; Aoyama, Y.
Carbohydr. Lett. 1995, 1, 223-230.
More detailed studies have shown that the selectivity of 1,2-
diols decreases in the order five-membered-cis > six-membered-
cis > six-membered-trans isomers.60 This study revealed several
interesting points in addition to the cis/trans stereoselectivity
of five- or six-membered rings. For example, there is a
remarkable ring-size selectivity of cis-diols, as indicated by the
ratio of binding constants (ca. 200) for cis-cyclopentandiol/cis-
cyclohexanediol.60 The affinities of cyclic 1,2-diols are cor-
related with the smaller dihedral angles between the adjacent
C-OH groups, and earlier work using molecular mechanic
calculation (MM2) by Oshima et al.60 suggests that the five-
membered cis isomer has a more planar conformation, thereby
facilitating the complexation with boronic acid groups. To
explore our experimental results that indicate that cis-1,2-
cyclopentanediol has a more favorable complexation in com-
parison to other diols, AM1 and PM3 (MOPAC) molecular
orbital as well as MM2 molecular mechanics calculations were
performed to estimate the dihedral angles in addition to the
calculation of oxygen-oxygen interatomic distances (see Figure
7).
As seen in Table 1, there is no significant difference between
the results obtained by either computational method. However,
for cis-1,2-cyclopentanediol, the calculated dihedral angles have
a strong dependence on the Hamiltonians used, and all calculated
angles for cis-1,2-cyclopentanediol were significantly smaller
than for the other three diols. The PM3 calculated dihedral
angles for cis-1,2-cyclopentandiol and cis-1,2-cyclohexandiol
are approximately 37° and 55°, respectively. These results
suggest that cis-cyclopentandiol should bind more strongly due
to its more planer conformation. Our experimental results are
in good agreement with this finding. The differences in
interatomic distances of oxygen atoms of hydroxy groups for
cis-1,2-cyclopentandiol and cis-1,2-cyclohexandiol were small.
Figure 8. Open circuit potential measurement using flow cell driven by a
pump. (A) The experimental setup and (B) the response of D-fructose, 0, 5,
0, 10, 0, 20, 0, 10, 0, 5, and 0 mM at 1 mL min-1 in PBS. Each interval
was for 10 min and the time required to change the concentration was 10
s. PABA was produced on the GC electrode in the presence of 200 mM
NaF.
Thus, ring-size effects do not provide a noticeable change in
this distance, and it is not a major factor in observed effects.
Performance under Dynamic Conditions. Since this ap-
proach is “reagentless”, it is ideally suited to continuous
monitoring applications. To demonstrate this capacity, flow
injection experiments where performed. D-Fructose was chosen
as an analyte due to its higher response and stability. Figure 8
demonstrates the performance of the sensor upon repetitive
exposures under various concentrations of D-fructose. The
9
12492 J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002

Potentiometric Saccharide Detection
A R T I C L E S
sensitivity in this configuration was ca. 1 mV mM-1 for
D-fructose with a background drift of ca. 0.2 mV h-1, corre-
sponding to 0.2 mM h-1. The small contribution of the
background drift permitted reliable monitoring of concentrations
over 10 h.77 These preliminary results indicate that the response
is reversible and stable and could be applied to monitor other
saccharides continuously.
geometries indicate that the O-C-C-O dihedral angle and not
oxygen atom separation reconciles with known complexation
trends. The response profiles observed in this work indicate that
the responses are diffusion controlled, although in the case of
D-glucose, secondary effects are observed associated with
changes in the equilibrium concentrations of the isomers
diffusing into the film. It is therefore expected that strategies
to enhance mass transport within the polymer films, such as
the creation of more porous or composite thin films, will
enhance the sensor’s response time. Efforts are currently
underway to address these possibilities.
The sensitivity and mechanical stability of PABA is enhanced
significantly by increasing the fluoride concentration during the
electrochemical polymerization. This suggests that eliminating
detrimental side reactions that occur at higher potentials is
crucial to increasing the sensitivity of the system. Further efforts
to increase sensitivity and the dynamic range of the sensor are
also under way. Finally, the utility of this sensing strategy was
demonstrated in a dynamic environment, indicating that it is
suited for the continuous monitoring of saccharides or other
diols.
Conclusions
In summary, the complexation of boronic acid group in PABA
produces a sufficient change in the pKa of PABA to result in a
significant increase in the electrochemical potential. This
behavior provides a selective means for measuring the solution
concentration of various sugars and vicinal diols. A Nernstien
expression was derived to explain the role of changes in the
formal potential and the pKa on the measured electrochemical
potential. These results demonstrate that the boronic acid
complexation is well behaved and also suggest that this strategy
may be extended to any reversible binding mechanism that
changes the pKa of a redox polymer.
The selectivity for different saccharides can be understood
from the relative concentrations of various isomers of particular
sugars and the relative binding affinity of boronic acid toward
these isomers. In the case of 1,2-cycloalkanediols, calculated
Acknowledgment. This work was supported by the Arnold
and Mabel Beckman Foundation.
(77) In this demonstration, the maximum concentration was arbitrarily limited
to 20 mM. Larger concentration ranges are possible.
JA0267371
9
J. AM. CHEM. SOC. VOL. 124, NO. 42, 2002 12493