Dynamics of disinfection of selected preservatives against Escherichia coli.

Mathematical models were determined relating preservative concentration and D values (decimal reduction times at 25 degrees C; pH 6.9-7.1) against Escherichia coli in aqueous medium. Preservatives investigated were 2-bromo-2-nitro-1,3-propanediol (Bronopol), N-(hydroxymethyl)-N-(1,3-dihydroxymethyl-2, 5-dioxo-4-imidazolidinyl)-N'-(hydroxymethyl)urea (Germall II), phenethyl alcohol, and benzyl alcohol. Linear regression was used to determine D values [i.e., the time required for a particular concentration of preservative at a specified pH, temperature, and medium to cause a 90% reduction of viable organisms (E. coli)] from a number of concentrations of each preservative. Linear regression of the log D values versus the log of the concentration (a minimum of 4 concentrations per preservative) were used to derive power curves. Concentration exponents, eta values (the logarithmic values relating changes in rates of kill for specified changes in concentrations) and A values (extrapolated D values at 1% concentration), were determined. Correlation coefficients for these power fits ranged from -0.987 to -0.999. Plots depicting the closeness of fit of the models to the actual data are shown.