The linearization of the callbratlon curve of the automated methylthymol blue (MTB) procedure for the determination of sulfates is described. Thls is accomplished by altering the barium-to-MTB molar ratio of the barium-MTB reagent used. Because of the low purlty of the commerclal dye, bariumto-MTB molar ratios of 0.9:l or lower are usually needed for obtaining linear callbratlon curves. A llnear relatlon between absorbance and sulfate concentration with a correlatlon coefficient of 0.9995 or better is obtained. The relative standard deviation among repllcates, in the range of 0 to 100 pg/ml, is 1.4 to 0.4%. The sensitivity for this range is 4 pg/ml SO4*-.Routine determination of sulfate in a large number of water and air samples is usually performed by a variety of automated wet chemical methods. The automated methylthymol blue method developed by Lazrus et al. (I) is gradually displacing all the other methods, and currently most of the Federal and State water and air pollution agencies have adopted this method for the determination of sulfates.The method involves displacement of methylthymol blue (MTB) from a barium-MTB complex by sulfate and colorimetric measurement of the freed MTB in highly basic solution.The absorption maximum of the barium-MTB complex is at 610 nm and that of the free MTB is a t 460 nm. Therefore, the free MTB can be measured a t 460 nm without major interference from the barium-MTB complex. As Equation 2 shows, the concentration of MTB is directly related to the concentration of sulfate, which permits determination of the sulfate by measuring the concentration of the displaced MTB.One of the most unwanted characteristics of this method has been the nonlinear relationship between the output of the instrument (absorbance) and the sulfate concentration. The plot of absorbance vs. concentration appears to be a parabolic function which becomes linear above a certain concentration of sulfate. Thus, in most cases, a calibration curve may be assumed as consisting of two linear portions of different slopes. The first linear portion corresponds to low sulfate concentrations (about one fifth of the total range), and the second to higher sulfate concentrations. The slope of the first linear portion of the calibration curve is lower by at least a factor of two from the slope of the second portion. Currently, two different approaches are used for the utilization of the nonlinear calibration curves for determination of sulfates. In the first approach, the output signals corresponding to each point of the calibration curve are linearized electronically with respect to the standard sulfate concentration, and thus the entire range can be utilized. In the second approach, only the linear portion of the calibration curve corresponding to higher sulfate concentrations is used for the analysis of samples. Both approaches have several disadvantages. For the electronic linearization, additional equipment and some extra effort to set the calibration are required. In the other approach, the sensitivity of the metho...