Measurement and analysis of rate data: The rate of reaction of ferric iron with pyrite

Abstract

Geochemists use several types of experiments to measure the rates of geochemical processes and analyze the results of these experiments by a variety of methods. In order to assess the efficacy of these experiments and methods of analysis, we measured the rate of reaction of ferric iron with pyrite using batch, mixed flow, and plug flow reactors and analyzed the results using differential, integration, and hybrid methods. The reaction of ferric iron with pyrite occurs very far from equilibrium, so there is no significant back reaction; and the rate can be fit to a general differential rate law of the form r = km n . This study showed that rates derived from ideal mixed flow, batch, and plug flow reactor experiments give comparable rates when the rates were determined from the batch and plug flow reactor data using the initial rate method. When these data were fit to the general rate law, we found the best estimate of k = 3.0 (±2.0) × 10 −5 mol m −2 sec −1 and n t = 0.62 (±0.10). Rates calculated from the general rate law using these parameters agree within a factor of two with the results of other studies. Integration and hybrid methods of analysis of these same data failed completely because there is an inhibitor among the products of this reaction. Fits of concentration vs. time data to the integrated rate law gives n, (the reaction order with respect to time), which is as much as five times higher than n c (the reaction order with respect to concentration). Furthermore, the rate constants from these fits range over tens of orders of magnitude from experiment to experiment. Analysis of the data using Wilkinson's method (a hybrid method) gave similar results. Thus, determining rate laws using integration or hybrid methods can be quite treacherous.