11863
2008, 112, 11863–11867 Published on Web 09/03/2008
Perturbation of Second and Farther Hydration Shells of Alkali Cations and Bromide in Concentrated Aqueous Protein as a Water-Shortage Medium Takumi Ohki, Makoto Harada, and Tetsuo Okada* Department of Chemistry, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8551, Japan ReceiVed: March 31, 2008; ReVised Manuscript ReceiVed: August 17, 2008
The Gibbs free energies of transfer of selected ions from water to concentrated aqueous ovalbumin and albumin (∆W′ W G°j) have been determined by ion-transfer voltammetry. Negative values for the tetrabutylammonium ion suggest its direct binding to ovalbumin. In contrast, for alkali cations and bromide, the ∆W′ W G°j values are positive and increase with increasing ovalbumin concentration. Positive values are confirmed for concentrated aqueous albumin and poly(styrenesulfonate) as well. The largest value (ca. 10 kJ mol-1) is found for the transfer of K+ from water to 30 wt % ovalbumin. To reveal the solvation structure of these ions in ovalbumin solutions, X-ray absorption fine structure (XAFS) measurements have been performed at the K, Rb, and Br K-edges. Interestingly, the spectra obtained in 30 wt % ovalbumin solutions are identical to those for the corresponding hydrated ions. This strongly suggests that the first coordination shell structures of these ions are not affected by a large concentration of ovalbumin. The detected positive free energy of transfer is slightly lower than the hydrogen bonding energy of a water molecule and should thus come from the perturbation of the second and farther hydration shells of the ions under a water-shortage condition caused by a high concentration of ovalbumin. Introduction The interactions of proteins with ions play important roles in a number of biological systems, such as ion channels, enzyme reactivity, and molecular recognitions by receptors, and have been studied from various viewpoints. The protein-ion interactions involve a wide variety of molecular mechanisms, including the coordination bond formation between a metal ion and ligand groups of a protein, the ion pair formation between an ion and an oppositely charged group, hydrogen bonding, etc. Calorimetry, NMR, and potentiometry with ion-selective electrodes have been often utilized to reveal such interactions.1-12 One of the key issues is whether an ion keeps its hydration structure or is dehydrated to some extent when it is bound by a protein. Unfortunately, most of the methods employed so far give no direct evidence for its hydration shell structure of an interacting ion. Concentrated protein solutions found in biological systems such as serum and cell sap can be regarded as water-shortage media. The hydration shell of an ion should undergo severe perturbation therein, because sufficient water molecules are not available for the formation of the complete hydration shell of the ion. It should be of importance to see whether such perturbation reaches the first hydration shell or the first shell remains unaffected while the hydration sphere outside the first shell is influenced. Although a number of experimental and theoretical studies of the outer hydration shells have been reported, no consensus has been reached on their structures and thermodynamics because of the intrinsically weak interaction ofsolventmoleculeswithionsandthelargethermalperturbation.13-17 It has been, for example, reported that the rotational dynamics * Phone and Fax: +81-3-5734-2612. E-mail:
[email protected].
10.1021/jp802734a CCC: $40.75
of water molecules in the second shell is the same as those in bulk water.17 Thus, it is very difficult to draw a clear-cut molecular picture of the second and farther hydration shells of an ion, albeit its importance is widely recognized. In previous papers, we measured the Gibbs free energy of transfer of an ion from water to concentrated solution of a hydrophilic polymer, and the solvation structures of some ions were revealed by X-ray absorption fine structure (XAFS).18,19 In concentrated poly(ethylene glycol) solutions, alkali cations and bromide are directly coordinated by the polymers, which replace some water molecules in the first coordination shells of the ions. Our attention is focused on the solvation of ions in aqueous protein solutions in the present work. As stated above, the first interest is whether proteins invade the first coordination shells of ions or not. The second one is to relate the solvation energy of an ion in a concentrated protein solution to its solvation structure therein. We will discuss the contribution from the second and farther hydration shells through the present approach. Experimental Section Chemicals. Reagents were of the highest grade available. Some salts were purified by recrystallization. Ovalbumin was purchased from Wako Chemicales and bovine serum albumin (BSA) was purchased from Nacalai Tesque. These proteins were used as received. Poly(4-styrenesulfonate) sodium salt (PSS, average molecular weight is 70000) was purchased from Aldrich. Water was purified with a MilliQ system. Sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaBAr4), bis(triphenylphosphine)iminium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate ([Ph3PdNdPPh3]BAr4), [Ph3PdNdPPh3]ClO4, Pn4NBAr4, 2008 American Chemical Society
11864 J. Phys. Chem. B, Vol. 112, No. 38, 2008 Me4NBPh4, KBPh4, RbBPh4, CsBPh4 and Et4NBPh4 were synthesized as previously reported.19 Voltammetric Measurements. Ion transfer through a microinterface between an aqueous solution (prepared in water, W, or in aqueous albumin, W′) and nitrobenzene (NB) was voltammetrically measured. The zero current-transfer potential of a given ion was measured between W and NB and between W′ and NB. A microinterface was formed using a polyester film with a 50 µm hole according to the literature.20,21 The concentration of ovalbumin in W′ was varied from 5 to 30 wt %. The pH of these solutions was almost neutral; pH ) 7.22, 7.14, 7.08, and 7.05 for 5 wt %, 10 wt %, 20 wt %, and 30 wt % ovalbumin. Also, 30 wt % of BSA solution was used as W′, which was neutralized with Li2CO3. The electrochemical cell used in this work is summarized as Ag/AgCl | 10 mM LiCl(W) | 5 mM Li2SO4(W) | W′ (or W) phase |* NB phase | 1 mM [Ph3PdNdPPh3][BAr4](NB) | 10 mM [Pn4N][BAr4] (NB) | 10 mM Pn4NCl(W) | AgCl/Ag. The W′ (or W) phase contained 1 mM of an analyte anion (as a Li+ salt) or 1 mM analyte cation (as a SO42- or Cl- salt), and NB phase contained 1 mM of a tetrahexylammonium or [Ph3PdNdPPh3]+ salt for an anionic analyte or a [BAr4]- salt for a cationic analyte. A supporting electrolyte in W or W′ was 5 mM Li2SO4, and that in NB was 1 mM [Ph3PdNdPPh3][BAr4]. Voltamograms at the W′ (or W)| NB interfaces were recorded by scanning the potential difference, E, between the Ag/AgCl electrodes. The potential applied to the Ag/AgCl electrode in W′ (or W) vs that in NB was defined as E. The potential scanning rate was 0.1 mV s-1. A triangular wave generated by a function generator Model HB-III (Hokuto Denko Corp.) was fed to a potentiostat Model HA1010 mM1A (Hokuto Denko Corp.). Voltammograms were recorded by an XY recorder Model F-35CA (Riken Denshi Co., Ltd.), and the digital data were acquired by a personal computer via the recorder. XAFS Measurements. Transmission XAFS measurements at the Br K-edge, the K K-edge, and the Rb K-edge, were performed at the room temperature (25 °C) at the beamlines BL-7C and BL-9A of Photon Factory of High Energy Accelerator Research Organization (KEK) in Tsukuba, Japan. These beamlines equipped a Si(111) double-crystal monochromator. Higher harmonic X-ray was removed by a Si mirror. Incident and transmitted X-ray intensities were determined with ionization chambers filled with appropriate gases. A sample solution was sealed in a polyethylene pouch and set on a light path of X-ray. The thickness of a sample was adjusted so that an appropriate signal intensity was obtained. The normalized XAFS interference function in the k-space, χ(k), is defined as
µ(k) - µb(k) - µ0(k) χ(k) ) µ0(k) k)
2m (E - E0) p2
where m is the electron mass, E is the incident X-ray energy, E0 is the threshold energy, and µ(k), µ0(k), and µb(k) are the total absorption coefficient, the absorption due only to the K shell excitation of a priori isolated ion, and the background absorption depending on the circumstances of the absorbing atom, such as absorption from the other shells and long-range solvation effects, respectively. The energy giving a half of the
Letters edge jump was chosen as E0. The background absorption, µb(k), was estimated with the Victreen’s formula, aE-3 - bE-4 - c. The XAFS spectra in k space, χ(k), were analyzed by curvefitting with the following equation.
χ(k) )
∑
SjNjFj(kj) krj2
j
kj )
k2 -
exp(-2σj2kj2) sin[2kjrj + φj(kj)]
2m ∆E0j p2
where j is the coordination shell number, rj is the distance between the atom under study and scattering atoms, SjNj is the amplitude factor (where Sj is the amplitude reduction factor), σj is the Debye-Waller factor, E0j is the absorption edge shift, and Fj(kj) is the backscattering amplitude. The scattering amplitudes and phase shifts for model systems were calculated with FEFF8.02. The models used for the analyses of the spectra at the Br K-edge and K K-edge were Br-H-O (rBr-H ) 2.258 Å and rBr-O ) 3.20 Å) and K-O (rK-O) 2.80 Å). Experimental XAFS spectra, χ(k), were Fourier-transformed with a Hanning window over the range k ) 0-6.8 Å-1 at the Br K-edge, k ) 0-10 Å-1 at the K K-edge. The fitting parameters were the number of coordinating atoms, the energy shift, the distance between the absorbing and scattering atoms, and the Debye-Waller factor. These parameters were determined by curve-fitting over the range k ) 1.7-6.8 Å-1 for the Br K-edge and over the range k ) 1.8-9 Å-1 for the K K-edge. Results and Discussion The determination of the Gibbs free energy of transfer of an ion from water (W) to aqueous ovalbumin or albumin (W′) is briefly described here. Because W and W′ phases are miscible, the zero current potential (EI)0) of a given ion is first measured between W and NB. W EI)0 ) ∆NB φ ° + const +
γW RT ln zF γNB
(1)
W where ∆NB φ° is the standard liquid junction potential between W and NB, γW and γNB are the activity coefficients in W and NB, respectively, and const is the sum of interfacial potentials but the liquid junction potential under study. EI)0 values measured for monovalent ions should contain the identical const term, as long as the same experimental system is employed. Thus, the following relation should hold when EI)0 values are measured for two ions, j and k.
∆EI)0 ≡ EI)0,j - EI)0,k
( (
W ) ∆NB φ°j + const +
W ∆NB φ°k + const +
γk(W) RT ln zkF γk(NB)
W W ) ∆NB φ°j - ∆NB φ°k +
(
) )
γj(W) RT ln zjF γj(NB)
)
γj(W) γk(W) RT RT (2) ln ln zjF γj(NB) zkF γk(NB)
A similar relation can be derived for the ion transfer between an aqueous albumin (W′) and the NB phases.
Letters
J. Phys. Chem. B, Vol. 112, No. 38, 2008 11865
W W ∆E′I)0 ) ∆NB φ°j - ∆NB φ°k +
(
γk(W′) RT γj(W′) RT ln ln zjF γj(NB) zkF γk(NB)
)
(3)
TABLE 1: ∆W′ W G°j for Selected Ions from Water to Aqueous Ovalbumin Solutions W′ ∆W G°j/kJ mol-1
ref ion
Br- Bu4N+ ClO4- Et4N+ Me4N+
Cs+
Rb+
K+
Me4N+ Et4N+ ClO4-
1.3 1.7 1.8
0.6 0.2 0.1
W′ ) 5% ovalbumin -0.5 0.4 0 -0.2 0 -0.4 0 -0.2 -0.5
0.4 -0.0 -0.1
2.2 1.8 1.7
4.7 4.2 4.2
Me4N+ Et4N+ ClO4-
4.5 4.3 5.3
-1.8 -1.5 -2.6
W′ ) 20% ovalbumin -0.8 -0.1 0 -1.0 0 0.1 0 -1.0 -0.8
2.9 3.1 2.2
6.5 6.6 5.6
9.9 10.0 9.1
Me4N+ Et4N+ ClO4-
6.0 5.6 6.3
-3.0 -2.6 -3.3
W′ ) 30% ovalbumin -0.3 -0.3 0 -0.7 0 0.3 0 -0.7 -0.3
3.7 4.0 3.3
7.0 7.3 6.6
10.7 10.9 10.3
A difference between ∆EI)0 and ∆E′I)0 gives
∆∆EI)0 ≡ ∆EI)0 - ∆E′I)0 W W W W ) {(∆NB φ°j - ∆NB φ°k) - (∆NB φ°j - ∆NB φ°k)}
+
{( (
) )}
γj(W) γk(W) RT RT + ln ln zjF γj(NB) zkF γk(NB)
γk(W′) RT γj(W′) RT ln ln zjF γj(NB) zkF γk(NB)
W ) (∆W W φ°k - ∆W φ°j) -
(
)
RT γj(W′) RT γk(W′) (4) ln ln zjF γj(W) zkF γk(W)
Our previous work indicated that the activity coefficient terms are negligible (i.e., γj(W)) γj(W′) and γk(W)) γk(W′)) for concentrated PEG solutions.18 Even though the activity coefficients of a given ion are not identical in W and W′, the activity coefficient terms in (eq 4) can reasonably be canceled for ions with the same charge. Thus, W ∆∆EI)0 ) ∆W W φ°k - ∆W φ°j
Finally, we obtain
(
∆W W G°j ) zj F · ∆∆EI)0 +
∆W W G°k zk
(5)
)
(6)
W′ ∆∆EI)0 is directly related to the ∆W′ W G°j if ∆W G°k is known. W′ We showed that the ∆W G° values for ClO4 , Et4N+, or Me4N+ are not affected by the concentration of hydrophilic polymers added in aqueous solution.18,19 This suggests that these appreciably bulky ions can be utilized as references for the calculation of ∆W′ W G°j of a single ionic species. Similar trends for these ions were confirmed in concentrated salt solutions.22 ∆W′ W G°j can thus be derived by
∆W W G°j ) zjF∆∆EI)0
(7)
The voltammetric baseline was flat for any W′ phase tested in the polarizable potential range, which indicates that the transfer of ionic impurities does not occur in the potential window. Major ionic impurities in 1 wt % ovalbumin were 1.4 × 10-4 M Ca2+ and 5.8 × 10-3 M Na+. These ions are so hydrophilic that their transfers to the NB phase were not observed. Table 1 summarizes ∆W′ W G°j values (W′ ) 5 wt %, 20 wt %, and 30 wt % albumin) calculated on the basis of (eq 7) with reference to k ) ClO4-, Et4N+, or Me4N+. In general, ∆∆EI)0 involves an ambiguity of 2-3 mV, which produces an W′ W′ error of 0.2-0.3 kJmol-1 in ∆W G°j. The ∆W G°j values for + + ClO4 , Et4N , and Me4N are almost zero even if any of them are taken as a reference ion. In addition, the ∆W′ W G°j value for a given ion is almost constant irrespective of a reference ion. Thus, these strongly support the consideration that these reference ions do not undergo solvation energy changes when transferred from W to the concentrated ovalbumin phase. The averages of three ∆W′ W G°j values obtained with different reference ions are plotted against the concentration of ovalbumin in W′ + in Figure 1. The ∆W′ W G°j value for Bu4N becomes more negative as the ovalbumin concentration increases, whereas those for Br-, Cs+, Rb+, and K+ become more positive.
As stated above, most of the related studies on ion binding by protein assumed direct interaction between ions and ionic groups of protein molecules.1-12 If direct ion associates are formed, the ∆W′ W G°j values for the involved ions must be negative because the interactions are energetically favorable. This situation holds for Bu4N+. The pI value for ovalbumin is 4.8-5,23 and this protein is negatively charged at natural pH (pH 7.1-7.2 in this study). Although the electrostatically induced ion-pair formation can occur between Bu4N+ and ovalbumin, the extent is not very large because of the hydrophilic nature of ovalbumin. Ion-pair formation is in general accompanied by dehydration from the ions involved in the reaction and is often driven by an entropic gain coming from dehydration processes. Bu4N+ has been used as an ion-pair reagent for chromatography and solvent extraction of various compounds.24-27 However, successful application of the ion-pair formation of Bu4N+ with amino acids and peptide has not been reported. This is related to small negative ∆W′ W G°j values, which imply the weak ion-pair formation of Bu4N+ with the anionic carboxylic groups contained in ovalbumin. The ∆W′ W G°j values of cations appear to correlate with their hydration energies, albeit the direct comparison is difficult for tetraalkylammonium ions because of the lack of reliable relevant data. Although the direct complexation of cations must occur for transition metal ions, such direct binding should not occur for alkali cations studied judging from the positive ∆W′ W G°j values as stated above. Therefore, the positive ∆W′ W G°j values should
W′ Figure 1. Changes in ∆W G°j with ovalbumin concentration in W′. W′ Reference ions, ClO4-, Et4N+, or Me4N+. ∆W G°j values determined in terms of different reference ions are averaged.
11866 J. Phys. Chem. B, Vol. 112, No. 38, 2008
Letters TABLE 2: Comparison of XAFS Parameters Determined for K+ and Br+ in Water with Those in 30 wt % Ovalbumin Solutiona N
σ/Å
K K-edge, k ) 2.0-7.0 Å K+ in water 2.76 K+ in 30 wt% ovalbumin 2.72
4.04 4.07
0.16 0.16
Br K-edge, k ) 2.0-6.8 Å-1 Br- in water Br- in 30 wt% ovalbumin
6a 5.77
0.18 0.18
r/Å -1
Figure 2. XAFS spectra of Br- in water (black solid) and in aqueous 30 wt % ovalbumin solution (red broken). Sample, 0.1 M RbBr.
Figure 3. XAFS spectra of K+ in water (black solid) and in aqueous 30 wt % ovalbumin solution (red broken). Sample, 0.1 M KBr.
originate not from the direct interaction of ions with the protein but from other mechanisms that are not accompanied by the perturbation of the first coordination shells. Interestingly, Br- shows positive ∆W′ W G°j values similar to alkali cations. In the present approach, halide ions except for Br- were not studied because of methodological limitations; F- and Cl- cannot be transferred into the NB phase within an appropriate potential range, and I- is easily oxidized in the NB phase. For these reasons, systematic measurements were unfortunately not possible for halide ions. The bindings of halide ions to a protein have been studied with various methods. Ionselective electrodes were, for example, employed for the determination of binding constants between BSA and halide ions. It has been reported that binding constants are ∼103 and decrease in the order F-, Br-, and I-.3 In contrast, entirely opposite trends are reported for the albumin-halide ion bindings in the study of protein errors of pH indicator dyes, the binding ability decreases in the order I- > Br- > Cl-.10 In these methods based on solution equilibria, there are several common problems: that is, we do not know what is measured by an employed method; no structural information is obtained from such studies. XAFS measurements were performed at the Br, Rb, and K K-edges to get further insight into the structures of the first hydration shells of Br-, Rb+, and K+ in ovalbumin solutions. XAFS χk3 spectra are shown in Figures 2 and 3. In these figures, the spectra obtained for hydrated Br- and K+ are depicted for comparison (the spectra for Rb+ are given as Supporting Information). Interestingly, the spectrum for a hydrated ion is identical to that obtained in a 30 wt % ovalbumin solution. The spectra for K+ and Br- were analyzed assuming water is the scattering path. The spectra for Rb+ were not analyzed because the k range utilizable for analyses is very restricted due to the disturbance from the [1s3d] multielectron excitation at k ) ca.
3.25 3.24
a Samples, 0.1 M KCl or RbBr. The hydration number of Br- in water was assumed to be six.30 The N values for K+ determined by curve-fitting well agree with the literature values.30
6.1 Å-1.28,29 The results are summarized in Table 2. As intuitively inferred from the spectra, the XAFS parameters, including the coordination distance (r), the number of scattering groups (N), and Debye-Waller factor (σ), are the same for both media. This strongly suggests that the structures of the first hydration shells of these ions are maintained even in such a concentrated ovalbumin solution. Thus, though these ions may interact with ovalbumin, the structure of the first hydration shell is not affected by the protein-ion binding. This is consistent with the above discussion based on ∆W′ W G°j. A special mention should be made for the positive ∆W′ W G°j values for some ions taking the results of XAFS into account. Because the first coordination shell is not perturbed in the presence of a large amount of ovalbumin, the weaker solvation in concentrated aqueous ovalbumin should come from the second or farther hydration shells of the ions. Also, more positive values are determined for more hydrophilic or less polarizable ions with smaller crystalline radii. In concentrated ovalbumin, sufficient water molecules are not available for the formation of the complete hydration shell of an ion. In particular, the second or farther hydration shells, which have poorly organized structures, are more largely influenced than the first W′ shell. Therefore, the positive ∆W G°j values found for alkali cations and bromide should reflect the perturbation in the second or farther hydration shell when the ion is transferred from water to a concentrated ovalbumin solution. The essential question is whether the present discussion is applicable to other similar media or not. The ∆W′ W G°j values from W to concentrated BSA or PSS were determined for selected ions to asses this point. For 30 wt % BSA, the ∆W′ W G°j values for K+ and Br- were 2.2 kJ mol-1 (vs Et4N+ reference) and 5.5 kJ mol-1 (vs ClO4- reference), respectively. The value for K+ is smaller than the corresponding one for ovalbumin reported in Table 1. There may be two reasons. One is the specific binding of K+ with BSA. The other is a difference in the extent of the interaction of the proteins with water. The strong interaction of ovalbumin, which is a glycoprotein, with water induces the significant reduction of water activity and results in more severe water-shortage circumstance. To confirm this consideration, a ∆W′ W G°j value (W′ ) 40 wt % PSS) was also determined. PSS as well as coexistent Na+ as a countercation is highly hydrated, and, in turn, a very limited number of free water molecules are available for hydration of a coexistent ion under study. Actually, the ∆W′ W G°j value for Br was determined -1 to be 13.9 kJ mol (vs ClO4 reference). The measurement for K+ was not conducted because its electrostatic interaction with the polymer chain was expected. Thus, the positive ∆W′ W G°j value is obviously related to the extent of a water-shortage in a medium.
Letters A theoretical calculation has suggested that the outer hydrogen bonding between water molecules contributes to the total hydration energy of an ion to some extent.16 However, a recent neutron diffraction study has indicated that weak orientational correlations occur in the range outside the first hydration shell of halide anions though being like bulk water.13 Thus, the contribution from the second hydration shell to the entire hydration energy of an ion is not very large. The energy of a hydrogen bond is estimated to be 17.5-32.2 kJ mol-1.31,32 The second hydration shell formation basically relies on the hydrogen bonding between the first shell and outer water molecules. Thus, it is reasonable to conclude that the positive ∆W′ W G°j values come from the perturbation of the second and farther hydration shells. Heavy alkali cations as well as Br- are often called “structure breakers”. However, even such ions, the hydration shell outside the first hydration sphere is involved in their complete hydration shell formation, albeit the energetic contribution is lower than 5% of the entire hydration energy. Acknowledgment. This work was in part supported by a Grant-in-Aid for Scientific Research from the Japan Society for Promotion of Science. Supporting Information Available: XAFS spectra and parameters of Rb+ in water and ovalbumin. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Chaturvedi, U. C.; Shrivastava, R. FEMS Immunol. Med. Microbiol. 2005, 43, 105. (2) Ayranci, E.; Duman, O. Protein Peptide Lett. 2004, 11, 331. (3) Ayranci, E.; Duman, O. Food Chem. 2003, 84, 539. (4) Luehrs, D. C.; Johnson, W. C. Fluoride 1986, 19, 86. (5) Qu, S.-S.; Liu, Y.; Wang, T.-Z.; Gao, W.-Y. Chemosphere 2002, 46, 1211. (6) Aime, S.; Canton, S.; Crich, S. G.; Terreno, E. Magn. Reson. Chem. 2002, 40, 41.
J. Phys. Chem. B, Vol. 112, No. 38, 2008 11867 (7) Sadler, P. J.; Viles, J. H. Inorg. Chem. 1996, 35, 4490. (8) Goux, W. J.; Venkatasubramanian, P. N. Biochemistry 1986, 25, 84. (9) Purcell, M.; Neault, J. F.; Malonga, H.; Arakawa, H.; Tajmir-Riahi, H. A. Can. J. Chem. 2001, 79, 1415. (10) Suzuki, Y. Anal. Sci. 2006, 22, 269. (11) Suzuki, Y. Anal. Sci. 2006, 22, 907. (12) Barone, J. R.; Dangaran, K. L.; Schmidt, W. F. J. Appl. Polym. Sci. 2007, 106, 1518. (13) Soper, A. K.; Weckstroem, K. Biophys.Chem. 2006, 124, 180. (14) Guardia, E.; Marti, J.; Garcia-Tarres, L.; Laria, D. J. Mol. Liq. 2005, 117, 63. (15) Bock, C. W.; Markham, G. D.; Katz, A. K.; Glusker, J. P. Theor. Chem. Acc. 2006, 115, 100. (16) Tunell, I.; Lim, C. Inorg. Chem. 2006, 45, 4811. (17) Omta, A. W.; Kropman, M. F.; Woutersen, S. J. Chem. Phys. 2003, 119, 12457. (18) Ohki, T.; Harada, M.; Okada, T. J. Phys. Chem. B 2006, 110, 15486. (19) Ohki, T.; Harada, M.; Okada, T. J. Phys. Chem. B 2007, 111, 7245. (20) Ohde, H.; Uehara, A.; Yoshida, Y.; Maeda, K.; Kihara, S. J. Electroanal. Chem. 2001, 496, 110. (21) Beriet, C.; Girault, H. H. J. Electroanal. Chem. 1998, 444, 219. (22) Ogura, K.; Kihara, S.; Suzuki, M.; Matsui, M. J. Electroanal. Chem. 1993, 352, 131. (23) Ahamed, T.; Nfor, B. K.; Verhaert, P. D. E. M.; van Dedem, G. W. K.; van der Wielen, L. A. M.; Eppink, M. H. M.; van de Sandt, E. J. A. X.; Ottens, M. J. Chromatogr. A 2007, 1164, 181. (24) Takayanagi, T.; Motomizu, S. Bull. Chem. Soc. Jpn. 2007, 80, 183. (25) Botsoglou, N. A.; Fletouris, D. J.; Psomas, I. E.; Mantis, A. I. Anal. Chim. Acta 1997, 354, 115. (26) Okumura, M.; Honda, S.; Fujinaga, K.; Seike, Y. Anal. Sci. 2005, 21, 1137. (27) Takahasi, T.; Kaneko, E.; Yotsuyanagi, T. Anal. Sci. 2006, 22, 1585. (28) De Panfilis, S.; Di Cicco, A.; Filipponi, A.; Comez, L.; Borowski, M. J. Synchrotron Rad. 2001, 8, 764–766. (29) Kodre, A.; Arcon, I.; Gomilsek, J. P.; Preseren, R.; Frahm, R. J. Phys. B 2002, 35, 3497–3513. (30) Ohtaki, H.; Radnai, T. Chem. ReV. 1993, 93, 1157. (31) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press: London, 1985. (32) Ohtaki, H. Yoekino Kagaku (Chemistry of Solution); Dainihon Tosho: Tokyo, 1987.
JP802734A
