ABSTRACT
The recognition of the importance of pH and Eh as “master variables” has been a great contribution to environmental science, particularly to aquatic chemistry. An accurate determination of these two parameters would enable the prediction of concentrations and chemical forms of the major and trace elements in water and aquatic sediments. Each combination of Eh and pH values uniquely adjusts the proportions of all dissolved species in solution. The appreciation of this potential has led investigators of biogeochemical processes to study and refine methods for measuring Eh and pH in the sediments. However, the natural aquatic environment is not a thermodynamic system at perfect equilibrium. Chemical reactions proceed at variable rates, and instruments suffer biases. To be useful, the measured values require careful interpretation. The following questions usually arise and need to be considered during the interpretation of the results obtained by measuring pH and Eh. Is the system at equilibrium, and if not, can we assume a predictable offset from the theory? What are the important reactions influencing the instrument(s) used in the measurements? Are these reactions relevant to the study? Is the system really homogeneous, or are we just assuming that it is? These and similar questions exemplify the problems associated with instruments based on rigorous theoretical requirements that are often not maintained during the measurement. In a typical field operation, Eh and pH electrodes are first calibrated with standard solutions for the purpose of adjusting the instrument(s) under some well-defined equilibrium conditions. The instrument is then used in the real environmental conditions (i.e., measurements of Eh and pH of water and sediments), and the values of the measured 48parameters are recorded. The belief that the measurements actually indicate the real environmental conditions is based on two implicit points. First, we assume that somehow the real conditions can be extrapolated from those existing in the standard solutions. Second, we also assume that the instrument response is similar and predictable under both conditions. In the end, the reality of the measurements largely depends on the extent to which these two assumptions are respected.
