•Ion-transfer voltammetry was performed for amine drugs at the DCE | water interface.•The standard ion-transfer potential is useful for predicting drug permeability.•This index would be more available than log P in the octanol–water system.Ion transfer voltammetry with the polarized 1,2-dichloroethane (DCE)|water (W) interface was carried out to determine the standard ion-transfer potential (∆OWϕ∘) and the distribution coefficient (logKD) for amine drugs, including desipramine, imipramine, labetalol, etc. The values of ∆OWϕ∘ and logKD are the hydrophobicity scales for the ionized and neutral forms of a drug, respectively, and were found to show a good mutual correlation. However, these hydrophobicity scales did not show a good correlation to the 1-octanol–water partition coefficient (logPoct) being conventionally used. In contrast, the permeability coefficient (logPpampa) in the parallel artificial membrane permeation assay showed a clear and characteristic dependence on ∆OWϕ∘ (or logKD). These results suggested that the solvation environment in DCE should be more similar to that in the hydrocarbon region of lipid bilayers than in 1-octanol. Thus, it was shown that ∆OWϕ∘, which can be easily determined by ion-transfer voltammetry, is a promising scale for predicting drug permeability through lipid membranes.

•The redox of ubiquinone-10 and vitamin K1 incorporated in SAMs was studied.•The redox was significantly slowed by the presence of the SAM|solution interface.•The adsorption of hydrophobic cations on the SAM caused the structural disorder.•The adsorption of cations could be confirmed by capacity measurements.A quinone derivative XQ (= ubiquinone-10 or vitamin K1) was incorporated into alkanethiol self-assembled monolayers (SAMs) formed on a gold electrode and the effect of the SAM|solution interface on the electron-transfer kinetics of XQ was investigated by cyclic voltammetry. It was confirmed that when a longer alkanethiol (e.g., 1-octadecylmercaptan) was used to form the SAM, the redox reactions of XQ were slowed significantly not only by the separation of XQ from the Au|SAM interface but also by the inhibition of the electron-transfer-coupled proton transfer across the SAM|solution interface. When a strongly hydrophobic supporting-electrolyte cation such as tetrabutylammonium or tetrapentylammonium was added to the aqueous solution, however, the redox reaction of XQ at the SAM modified electrode became faster. In order to explain this reason, we have proposed a model in which the adsorption of strongly hydrophobic cations on the SAM|solution interface induces a structural disorder of the monolayer and then facilitates the supply of protons from the solution to XQ in the monolayer. An evidence of the adsorption of the strongly hydrophobic cations on the SAM|solution interface was obtained from the measurements of the double-layer capacitance of the SAM modified electrode by alternating-current voltammetry.

Various organic anions (sulfonates (RSO3–), carboxylates (RCO2–), and phenolates (RO–)) and ammonium cations (RNH3+, R2NH2+, and R3NH+) were distributed in the nitrobenzene (NB)–water system by using Crystal Violet and dipicrylaminate, respectively. The number of water molecules (n) being coextracted into NB with an ion was then determined by the Karl Fischer method. The n values determined and those reported previously showed the variation from 0.51 to 3.4, depending on not only the charged groups but also the noncharged R-groups. In this study, we focused our attention to the strong electric field on the charged group and its facilitation effect for binding water molecules in NB. The local electric field (Ei) on the surface of an organic ion was evaluated by using Gaussian09 program with a subprogram developed in our recent study. It was found that the n values showed a clear dependence on the average value of Ei on oxygen or hydrogen atoms, respectively, of an anionic or cationic group.

The non-Bornian solvation model was applied for evaluation of the standard Gibbs energy (ΔGtr°,W→O) of transfer of organic ions from water (W) to organic solvent (O = nitrobenzene). The solvation energy of an ion in either W or O is basically formulated as the energy required for the formation of a nanosized ion–solvent interface around the ion; however, many organic ions with strongly charged groups (e.g., −SO3–, −CO2–, −NH3+) are preferentially hydrated in O. Here we divided the surface of an ion into “hydrated” and “non-hydrated” surfaces and then carried out regression analyses with experimental values of ΔGtr°,W→O. In the analyses, the local electric field on the surface of an organic ion was evaluated through density functional theory calculation. Good regression results were then obtained with the mean absolute error of 1.9 and 2.4 kJ mol–1 for 34 anions and 63 cations, respectively. These errors correspond to the error of ∼20 mV in the standard ion-transfer potential (ΔOWϕ°), being only two times larger than the typical experimental error (∼10 mV) in the voltammetric measurement. This non-Bornian model is promising for theoretical prediction of ΔGtr°,W→O (or ΔOWϕ°) for organic ions and possibly of the biomembrane permeability for ionic drugs.

Recently, a non-Bornian model was successfully applied to evaluate the Gibbs energy of hydration (ΔG°hyd) for spherical ions (mainly inorganic ions). In this model, the long-range, Born-type electrostatic ion–solvent interaction is not explicitly included in the calculation of ΔG°hyd, since its contribution is small, whereas the short-range interaction, including Coulomb, polarization, and charge-transfer interactions, is considered as the dominant factor that determines the ΔG°hyd of ions. The ΔG°hyd scaled by the surface area of an ion can be given by a quadratic function of the surface field strength (E) of the ion. In this study, the non-Bornian model was further applied to organic ions with charged groups. Using the Gaussian 09 program package, the geometries of ions in vacuum were optimized at the B3LYP/6-311++G(2d,p) level, and the partial atomic charges were computed in the Mulliken, Merz-Kollman (MK), natural population analysis (NPA), Hirshfeld, and ChelpG methods. Introducing a new subprogram, we could estimate local electric fields on the ion surface (van der Waals surface or solvent-accessible surface (SAS)). This enabled us to perform regression analyses based on the non-Bornian model, by using the experimental values of ΔG°hyd for 109 ions. When the NPA-SAS combination was chosen, the best regression result was obtained, giving the mean absolute error of 4.3 kcal mol−1. The non-Bornian model would provide a simple and relatively accurate way of determining ΔG°hyd of ions.

The mixed potential (MP) theory was successfully utilized to design an ionophore-based polyvinyl chloride (PVC) membrane K+ ion-selective electrode (ISE). Prior to the application of the MP theory, the transfer of K+ and interfering ions (Na+, Li+, and H+) facilitated by bis(benzo-15-crown-5) (BB15C5) or dibenzo-18-crown-6 (DB18C6) at a micro PVC membrane/water interface was studied by ion-transfer voltammetry (ITV). The reversible half-wave potentials were then obtained for the facilitated transfer of the ions. Using such voltammetric data and the literature data about diffusion coefficients of ions, we could well-predict the potential responses of the BB15C5- or DB18C6-based K+ ISE, as the function of the concentrations of primary and interfering ions, and also of the counterion for K+ [e.g., tetrakis(4-chlorophenyl)borate] added to the membrane. Thus, the MP theory has been proven to be useful to optimize the membrane composition for a higher ion selectivity and a lower detection limit. It has also been found that the leaching of ions from an inner solution is too small to affect the detection limit, at least for the designed PVC membrane ISE.

•The non-Bornian theory for the ion-transfer energy has been revised.•The transfer energy is formulated based on short-range ion–solvent interactions.•The Uhlig formula is not used for evaluating the cavity formation energy.•The transfer energy is a simple function of the surface field strength of the ion.Previously, Osakai and Ebina (1998) proposed a non-Bornian theory for the Gibbs energy (ΔGtr°,W→O) of ion transfer at the organic solvent/water interface. In a conventional manner, however, ΔGtr°,W→O was divided into ionic charge-dependent and -independent terms. The former was formulated based on non-Bornian, short-range interactions of an ion with primary solvent molecules, while the latter was evaluated as the energy of cavity formation by using the Uhlig formula. In this study, we have successfully shown that the above two terms can be brought together in the context of short-range ion–solvent interactions, and that the ΔGtr°,W→O scaled by the ionic surface area can be given by a quadratic function of the surface field strength of the ion. The obtained semiempirical equation of ΔGtr°,W→O is very simple, and would be useful for the prediction of ΔGtr°,W→O.Graphical abstract

Previously, an excellent theory was proposed for understanding the potential response of liquid membrane ion-selective electrodes (ISEs), which was based on a concept of zero-current potential or “mixed potential (MP)” of the interface between the liquid membrane and a sample solution containing interfering ions. Unfortunately, however, the MP theory has not been utilized explicitly in the development of the most popular, polyvinyl chloride (PVC) membrane ISEs, because the theory has not been verified by using the PVC membrane/water interface. In this study, we have successfully employed a micro PVC membrane/water interface to perform voltammetric measurements for the interfacial transfer of ions, including tetramethylammonium (TMA+), tetraethylammonium (TEA+), and tetrapropylammonium (TPrA+) ions. Using the thus obtained voltammetric data (i.e., the standard ion-transfer potentials), the potential response of a PVC membrane TEA+-ISE in the presence of TMA+ or TPrA+ as the interfering ion could be well elucidated by the MP theory. This theory would be available for the sophisticated design of high performance ISEs, which is not determined by conventional “trial-and-error” procedures.Graphical abstractHighlights► The mixed-potential theory has been applied to a PVC membrane TEA+-ISE. ► The transfer of TEA+ was studied at a micro PVC membrane/water interface. ► The standard ion-transfer potentials of TEA+ and interfering ions were determined. ► The selectivity coefficients of the ISE could be estimated theoretically.

The electron transfer (ET) between cytochrome c (Cyt c) in water (W) and 1,1′-dimethylferrocene (DiMFc) in 1,2-dichloroethane (DCE) was studied. The cyclic voltammograms obtained for the interfacial ET under various conditions could be well reproduced by digital simulation based on the ion-transfer (IT) mechanism, in which the ET process occurs not at the DCE/W interface but in the W phase nearest the interface. In this mechanism, the current signal is due to the IT of DiMFc+ as the reaction product. On the other hand, the measurement of the double-layer capacity showed that Cyt c is adsorbed at the DCE/W interface. However, the contribution from the adsorbed proteins to the overall ET is considered to be small because of the thicker reaction layer in the IT mechanism. These findings would offer a useful suggestion for the behaviors of Cyt c in vivo.

The catalytic activity of a membrane-bound enzyme, d-fructose dehydrogenase (FDH), at the polarized oil/water (O/W) interface was studied. Multisweep cyclic voltammetry and ac voltammetry were carried out to show the irreversible adsorption of FDH at the interface. Using the thusly prepared FDH-adsorbed O/W interface, clear steady-state catalytic current was observed in amperometry and cyclic voltammetry, where 1,1′-dimethylferrocenium ion (DiMFc+, electron acceptor) and d-fructose (substrate) were added to the O and W phases, respectively. The observed catalytic current was then analyzed by using two mechanisms. In mechanism (A), the heme c site of FDH, where DiMFc+ is reduced, was assumed to be located in the O-phase side of the interface. The intramolecular electron transfer in FDH should be affected by the Galvani potential difference of the interface (ΔOWϕ). However, the theoretical equations derived for the catalytic current could not reproduce the experimental data. In mechanism (B), the heme c   site was assumed to be in the W-phase side. In this case, ΔOWϕ should affect the interfacial distribution of DiMFc+. This mechanism could reproduce well the observed potential dependence of the catalytic current.

Voltammetric behaviors of various globular proteins, including cytochrome c, ribonuclease A, lysozyme, albumin, myoglobin, and α-lactalbumin, were studied at the polarized 1,2-dichloroethane/water (DCE/W) interface in the presence of four different anionic surfactants, that is, dinonylnaphthalenesulfonate (DNNS), bis(2-ethylhexyl)sulfosuccinate (Aerosol-OT; AOT), bis(2,2,3,3,4,4,5,5,6,6,7,7-dodecafluoroheptyl)sulfosuccinate (BDFHS), and bis(2-ethylhexyl)phosphate (BEHP). When the W phase was acidic (pH = ∼3.4), the surfactants (except for BEHP) added to DCE facilitated the adsorption of the above proteins to the DCE/W interface and gave a well-developed voltammetric wave due to the adsorption/desorption of the proteins. This voltammetric wave, which we here call “protein wave”, is promising for direct label-free electrochemical detection of proteins. The current for the adsorption of a protein to the interface showed a linear dependence on the protein concentration in the presence of excess surfactant. The foot potential at which the protein wave appeared in cyclic voltammetry showed different values depending on the natures of the protein and surfactant. Multivariate analysis for the foot potentials determined for different proteins with different surfactants revealed that the protein selectivity should depend on the charged, polar, and nonpolar surface areas of a protein molecule. On the basis of these voltammetric studies, it was shown in principle that online electrochemical separation/determination of proteins could be performed using a two-step oil/water-type flow-cell system.

Two interfacial electron transfer (ET) systems, i.e  ., the decamethylferrocene (DcMFc)–Fe(CN)63- system and the MTPP+–Fe(CN)64- system (M = Zn or Cd; TPP = 5,10,15,20-tetraphenylporphyrin) at oil/water interfaces were studied by means of scanning electrochemical microscopy. The second-order rate constants (k) for the ET systems could be accurately determined at different “standard” Gibbs energies (ΔG°) by changing the nature and concentration ratio of a common ion added to both phases. The driving force dependence, i.e., the log k vs. ΔG° plot did not show a simple upward parabola, with the k values being limited in a certain range of ΔG  °. This clearly suggested that there should be a bimolecular-reaction effect, as predicted previously [T. Osakai, H. Hotta, T. Sugihara, K. Nakatani, J. Electroanal. Chem. 571 (2004) 201–206]. However, the observed diffusion-controlled rate constants are typically one-order smaller than the theoretical ones, and have shown only small dependence on the viscosity of organic solvents for the DcMFc–Fe(CN)63- system. These results were unexpected from the previous model based on “microscopic diffusion” of a redox species in the vicinity of the interface, and then suggested an alternative model, in which the rate-determining step is “interfacial diffusion” of a redox species across the oil/water interface.

The spectrofluorometric behavior of a membrane potential-sensitive dye, 1-(3-sulfonatopropyl)-4-[β-{2-(di-N-butylamino)-6-naphthyl}vinyl]pyridinium betaine (di-4-ANEPPS), at the polarized 1,2-dichloroethane/water interface was studied by means of potential-modulated fluorescence (PMF) spectroscopy. The results, combined with those from cyclic and alternating current voltammetry, clearly suggested that the dye adsorbed at the interface underwent a reorientation with increasing the interfacial potential, giving a well-developed PMF response as well as a voltammetric response. In addition to the PMF response, another PMF response was observed by addition of dilauroyl phosphatidylcholine (DLPC). This additional response was well explained in terms of a reorientation of di-4-ANEPPS at the interface, which would be induced by the potential-dependent desorption of DLPC from the interface. Thus, the present study supported the reorientation/solvatochromic mechanism for the membrane potential-sensitive dye rather than the electrochromic mechanism.

Glucose oxidase (GOD)-catalyzed electron transfers between some oxidants in nitrobenzene (NB) and glucose in water (W) were studied by cyclic voltammetry. When an electrically neutral compound, chloranil (CQ), was employed as the oxidant in NB, the enzymatic reaction could not be regulated because of the spontaneous transfer of CQ from NB to W. In this case, the voltammetric wave observed for the enzyme-catalyzed electron transfer was increased depending on the standing time until the voltage scan was started. However, when an ionic oxidant, dimethylferricenium ion (DiMFc+), was employed as the oxidant, the electrochemical control of the enzymatic reaction was achieved by controlling the interfacial transfer of DiMFc+, so that well-reproducible voltammograms could be obtained for different concentrations of DiMFc+ and for different scan rates. The voltammetric behaviors were successfully explained by a digital simulation based on the ion-transfer mechanism, which involves the interfacial transfer of DiMFc+ and the succeeding GOD-catalyzed electron transfer which occurs not heterogeneously at the interface, but homogeneously in the W phase.

In contrast to conventional electrode reactions, an electron transfer (ET) at an oil (O)|water (W) interface is a bimolecular reaction, so that the “microscopic” diffusion of a redox species in the immediate vicinity of an O|W interface should be not a linear one, but like a hemispherical diffusion. Accordingly, the second-order rate constant obtained from usual kinetic measurements involves such a bimolecular-reaction effect, having a certain upper limit determined by the microscopic diffusion process. In this study, the diffusion-controlled rate constant of ET at an O|W interface has been calculated in the analogy of the Smoluchowski–Debye theory for a bimolecular reaction in a homogeneous medium. It has been shown that when the heterogeneous ET process is very fast, the overall or observed rate constant may be restricted by the diffusion-controlled rate constant.

In the nitrobenzene (NB)/water system, primary to tertiary ammonium ions with Me, Et and n-Bu groups were distributed with dipicrylaminate (DPA−) and tetraphenylborate (TPB−), and the number of water molecules being co-extracted to NB by an alkylammonium ion (RmNH4−m+) was measured by the Karl Fischer method. In the use of TPB− as the counter ion, the numbers of co-extracted water molecules, mainly for primary and secondary alkylammonium ions, were somewhat smaller than those in the use of DPA−, showing a possible influence of ion-pair formation between RmNH4−m+ and TPB− in NB. Conductivity measurements were then performed to determined the association constants of the RmNH4−m+ ions with TPB− ions as well as DPA− in ‘‘water-saturated ’’ NB, and the concentrations of free RmNH4−m+ ions were evaluated. By assuming that the ion pairs of RmNH4−m+ would have no water molecules in NB, the ‘‘ true’’ hydration numbers (nh) of the RmNH4−m+ ions in NB were successfully obtained. The nh values of the RmNH4−m+ ions, being little affected by the alkyl chain length, were found to decrease with heightening the class of the RmNH4−m+ ion: nh=1.64, 1.04 and 0.66, respectively, for the primary, secondary and tertiary ammonium ions. It was suggested that water molecules interact directly with hydrogen atom(s) bound to the central nitrogen atom.