Optimization of E. coli Inactivation by Benzalkonium Chloride Reveals the Importance of Quantifying the Inoculum Effect on Chemical Disinfection

Abstract

Optimal disinfection protocols are fundamental to minimize bacterial resistance to the compound applied, or cross-resistance to other antimicrobials such as antibiotics. The objective is twofold: guarantee safe levels of pathogens and minimize the excess of disinfectant after a treatment. In this work, the disinfectant dose is optimized based on a mathematical model. The model explains and predicts the interplay between disinfectant and pathogen at different initial microbial densities (inocula) and dose concentrations. The study focuses on the disinfection of Escherichia coli with benzalkonium chloride, the most common quaternary ammonium compound. Interestingly, the specific benzalkonium chloride uptake (mean uptake per cell) decreases exponentially when the inoculum concentration increases. As a consequence, the optimal disinfectant dose increases exponentially with the initial bacterial concentration.

Keywords: Escherichia coli; benzalkonium chloride (alkyldimethylbenzylammonium chloride); disinfection; inactivation; inoculum effect; kinetic modeling.

Figure 1

Classical models under disinfectant demand-free conditions assume constant disinfectant concentration and are a special case of the Generalized Differential Rate Law (Gyürék and Finch, ; Hunt and Mariñas, 1999). The Integrated form of most of these models can be seen in Gyürék and Finch (1998). Common models using disinfectant demand conditions assume decay independent on the microorganism density: first order decay (Lambert and Johnston, 2000) and second-order rate (Hunt and Mariñas, 1999). Only Fernando (2009) considers that disinfectant decay depends on the microorganism density.

Figure 2

Evolution of E. coli viable counts and extracellular BAC concentration with contact time under demanding conditions at different inoculum and dose concentrations. Left and right columns show E. coli inactivation and BAC decay, respectively, while each row corresponds to a different dose concentration (100, 200, and 300 ppm). Each panel shows the dynamics with high and low inoculum concentration in, respectively, blue and red. Replicates are represented with asterisks for viable counts, and lines go through their mean values. Detection limit (2 logs) for viable counts is represented with ≤ 2 and gathers all results below this limit. (A,C,D) show E. coli inactivation by 100, 200 and 300 ppm of initial BAC concentration respectively. (B–D) despite BAC decay for the same dose concentration (100, 200 and 300 ppm).

Figure 3

Dependence of BAC uptake and specific BAC uptake (first and second rows of panels respectively) with inoculum size and disinfectant dosage. (A) shows the dependence of BAC uptake with dose concentrations at two different inoculum concentrations for experiments in Figure 2. (B) depicts the dependence of BAC uptake with the inoculum concentration for a set of new experiments. (C) shows the isotherms of uptake for experiments in Figure 2. (D) The dependence of specific BAC uptake with dose concentrations at two different inoculum concentrations for experiments in Figure 2. (E) depicts the correlation between inoculum and specific BAC uptake for the new experiments at different inocula. Finally (F) shows the proposed correlation to explain specific BAC uptake at different inoculum and dose concentrations.

Figure 4

Best fits for model (6) assuming that specific BAC uptake is constant (dashed line) or depends on inoculum and dose concentrations (continuous line). Experimental data marked with asterisks correspond with the results shown in Figure 2. High and low inoculum concentration are shown in blue and red, respectively. The model with specific BAC uptake of the form $\alpha =10^{-a}{(C_0/N_0)}^b$ fits the data considerably better than the model assuming α = 10^−a = cte with b = 0. (A,C,D) show model simulations and data of E. coli inactivation by 100, 200 and 300 ppm of initial BAC concentration, respectively. (B–D) model simulations and data of BAC decay for the same dose concentration (100, 200 and 300 ppm).

Figure 5

New data prediction (first row) and optimal BAC dose concentration (second row).