COMMUNICATION Hofmeister Phenomena in Bioelectrochemistry: The Supporting Electrolyte Affects the Response of Glucose Electrodes Cristina Carucci, [a,b] Philipp Haltenort, [a] Marcela Salazar, [a] Andrea Salis* [b,c] and Edmond Magner* [a] Dedicated to the memory of Professor John Albery FRS. Abstract: The faradaic response of ferrocene-methanol when used as a mediator for the catalytic oxidation of glucose by glucose oxidase is examined in a range of electrolytes. The response is modulated by the nature of the salt used. A two-fold difference was observed with the sodium salts of fluoride and thiocyanate, with a 24 % difference between NaCl and KCl. The changes in the response can be explained by the Hofmeister effect, with specific ion effects arising between the mediator and co-substrate binding pocket of the enzyme. Such differences can significantly affect the response of electrochemically mediated glucose biosensors and biofuel cells and emphasize the importance of carefully considering the solution conditions when evaluating the properties of glucose oxidase based biosensors and biofuel cells. Understanding the properties of redox proteins and enzymes is of significant importance in the development of biosensors [1-5] and biofuel cells. [6, 7] The commercial development of biosensors has been largely driven by glucose biosensors, the success of which is evident by annual sales that are estimated at a level of €10 billion per annum. [6] Electrochemical glucose biosensors utilise an enzyme such as glucose oxidase (GOx) to provide the selectivity necessary to detect glucose. GOx is a dimeric glycoprotein composed of two identical subunits each containing a flavin adenine dinucleotide (FAD) cofactor that is reduced to FADH2 on oxidation of glucose [1, 4] (Scheme 1). The reduced cofactor, normally oxidized by oxygen, is directly oxidized by a mediator such as a ferrocene (Fc) or osmium complex, as the direct electrochemical oxidation of the reduced, enzymatic bound cofactor at a bare electrode is not feasible. [1-5] The reduced mediator is oxidized at the electrode surface producing a current that is directly proportional to the concentration of glucose. In an analogous manner, biofuel cells utilize enzymes such as glucose oxidase to oxidize a fuel (glucose). Oxygen reducing enzymes such as laccase or bilirubin oxidase are used at the cathode to complete the electrical circuit. [8] Scheme 1. Reaction scheme for mediated electron transfer. Provided the conductivity of the solution is adequate, the nature and concentration of the electrolyte in solution is not normally considered to affect the faradaic properties of a mediator, [9] nor to affect the response of a biosensor or a biofuel cell. However, it is well established that the nature of the anion or cation can affect the properties of a protein such as aggregation, [10, 11] or enzymatic activity. [12-17] We have recently shown that the faradaic response of the model redox protein, cytochrome c, can be modulated by the nature of the ions present in solution. [18] Such specific ion effects follow the Hofmeister series: [19, 20] Anions: SO4 2- , > F - > Cl - > NO3 - > Br - > I - > ClO4 - > SCN - . Cations: Cs + > NH4 + > K + > Na + > Li + > Mg 2+ . Specific ion effects that follow these series have been widely reported. [19, 21, 22] In the case of proteins, the nature of the ions influences surface charge, [23, 24] electrophoretic mobility, [25] or adsorption on mesoporous materials. [26] In addition, ion specificity affects the aggregation of polymers, [27, 28] pH measurements, [29, 30] , etc.. A number of theoretical approaches have been developed to explain the Hofmeister effect. [31, 32] Collins’ “law of matching water affinities” (LMWA) is based on electrostatic and hydration forces between ions and water or between ions and charged groups at the surfaces of biomacromolecules. [33, 34] Ninham pointed out that conventional theories of electrolytes (i.e. Debye Hückel) explain ion interactions in terms of electrostatics only. [35] If ion dispersion forces are also considered, both the specific structuring of water molecules adjacent to ions and the specific interaction of ions with interfaces can be explained. [36, 37] While the two approaches, LMWA and ion dispersion forces, have been regarded as separate, recent theoretical developments have suggested an integrated approach that combines both approaches. [32, 36, 38] Here we report the measurements of specific ion effects on the faradaic response of ferrocene-methanol when used as a mediator for the catalytic oxidation of glucose by GOx. We demonstrate that the nature of the ions influences the catalytic response in a manner than can be explained by the Hofmeister effect. The specific ion effects likely arise from interactions between the ions and the co-substrate binding site of GOx and can affect the response of glucose oxidase based biosensors and biofuel cells. Cyclic voltammograms of Fc-MeOH alone and with the addition of GOx were recorded in the presence of different salts (200 mM) to examine specific anion (Figure 1A) and cation (Figure 1B) effects. The voltammograms are representative of a one electron process with the exception of NaSCN where values of 42 mV and 1.77 were obtained for Ep and ia/ic, respectively (Table 1, Figure 1). On addition of GOx no significant changes were observed in the response (Table S1). While the salts do affect both the cathodic and anodic peak currents and the redox potential of Fc-MeOH in the presence and absence of GOx, no discernible ion specific trend was observed at salt concentrations of 200 mM (Figures S1 and S2). On addition of glucose, voltammograms typical of an EC catalytic process (equations 1–3) were obtained (Figure 2). In contrast to the mediator alone (or with GOx), the nature of supporting electrolyte strongly affects the electrochemical response with a clear ion specific trend. [a] Prof. E. Magner, C. Carucci, P. Haltenort, M. Salazar Department of Chemical and Environmental Sciences & Materials and Surface Science Institute, University of Limerick, Limerick, Ireland. E-mail: edmond.magner@ul.ie [b] Dr. A. Salis, C. Carucci Department of Chemical and Geological Sciences, University of Cagliari, Cittadella Universitaria, SS 554 bivio Sestu, 09042 Monserrato (CA), Italy E-mail: asalis@unica.it [c] Dr. A. Salis Department of Applied Mathematics, Research School of Physics and Engineering, The Australian National University, 0200 ACT, Australia. Supporting information for this article is available on WWW under http://