CONTROL OF LIQUID JUNCTIONS - THE SYSTEM HCL-KCL

Extending a recently published investigation [3], a new theoretical method has been used to evaluate emfs of the cell: Ag(s)\AgCl(s)\KCl(aq,c2)\l.j.\HCl(aq,c1)\AgCl(s)\Ag(s) where \l.j.\ denote different liquid junctions of HCl-KCl solutions. The method uses operationally defined variables and opens for experimental control of assumptions frequently made in calculations of liquid junctions. Distinction is made between emf contributions due to: a) electrolyte concentration variations, b) neutral electrolyte activity changes, and c) differences in mobility ratios. Three shapes of concentration profiles between the boundary values c2 and c1 in the cell have been studied: 1) linear concentration profiles, 2) profiles obtained from the Nernst-Planck flux equations, and 3) those from the Miller LNI approximation. The contributions to emf from all profiles have been calculated within +/- 0.1 mV. The emfs of seven cells with linear concentration profiles have been measured within +/- 0.2 mV. Calculated and measured results agree within error limits of the empirical data set. The results show that contributions to emf from activity coefficient variations are almost always significant; under the conditions applied here they amount up to more than 10 mV. The contributions from mobility differences attain the same order of magnitude; the error made by using constant mobilities in the calculations is limited within +/- 1 mV. The concentration dependencies of the activity coefficients are most important for a good calculation of the emf. The method gives a quantitative base for discussion and application of salt bridges in analytical chemistry. Particularly it is demonstrated in detail that Nernstian behaviour cannot be expected from electrochemical cells having a perfect ion-selective electrode and a salt bridge.