"Bacteria Disinfection with Hydrogen Peroxide"

M. LABAS, C. ZALAZAR, R. BRANDI, AND A. CASSANO

Clausthal-Zellerfeld (Alemania)

Conferencia; 4th Conference on Oxidation Technologies for Water and Wastewater Treatment; 2006

INTRODUCTIONThe antimicrobial properties of H2O2 have been known for many years now because of its efficacy, versatility and reasonable manipulation safety. Although the exact mechanism by which H2O2 produces lethal consequences for many microorganisms has not been clearly elucidated, it is well known that due to its strong oxidative properties it can produce damage to nucleic acids, proteins and lipids.    In this work we studied the disinfection kinetics of E. coli using hydrogen peroxide. With the employed approach the obtained kinetic parameters are useful for scaling-up purposes because they are independent of the reactor shape, size and operating conditions.METHODS The employed reactor is a well-stirred batch reactor (1000 cm3) with a jacket to keep the system temperature constant at 20ºC. E. coli strain ATCC 8739 was used throughout this work. The working solution was prepared from a culture with nutrient broth afterwards was brought to a 1/1000 dilution with physiological saline to simulate transparent waters. The prepared culture is mixed with different concentrations of H2O2 (40 - 300 ppm). Each sample was subjected to the following measurements: absorbance a 350 nm and CFU counting using specific Pretrifilm™ plates .Proposed kinetic methodsA significant proportion of the published literature on inactivation of microorganisms reports the existence of a shoulder in a semi logarithmic plot of survivors vs. time. This behavior has been attributed to an initial resistance to the disinfectant by all or a part of the involved species as well as other more specific aspects of the chemical attack. Two mechanistic types of kinetic models have been proposed: Series-Event ModelIt is based on the idea that an event is assumed to be a unit of microorganism damage. Events occur in a stepwise fashion and each step is considered a separate event. The model is thought of as a series of consecutive “damaging reactions”. Combining the kinetic expression with the mass balance, after integration one gets:     Multi-target ModelIt is assumed that a microorganism contains a finite number, nc, of discrete critical targets, each of which must be hit prior to reach the desired full inactivation of the living particle. If uniform clumping is presumed, rather than the internal resistance of a single organisms, a different rate equation must be devised which includes the decrease in probability of attaining lethal hits on viable organisms as the number of viable organisms is depleted.Inserting the kinetic model in the mass balance, after integration:     RESULTSThe kinetic constants of the model were obtained using a non-linear, multiparameter optimization program.Events in Series model: The events number is function of the H2O2 concentration: n = 1 for   100 ppm, n = 2 for  < 100 ppm. The kinetic constants obtained are:     and  .Multitarget model:The targets number is function of the H2O2 concentration: nc = 1 for  100 ppm, nc = 2 for  < 100 ppm. The kinetic constants obtained are:      and        Events in Series modelBacteria inactivation. Comparison of model predictions and experimental data. Solid lines: model predictions    Multitarget modelBacteria inactivation. Comparison of model predictions and experimental data. Solid lines: model predictionsCONCLUSIONSThe described results permit to summarize the following conclusions. The inactivation rate presents a direct dependence on the H2O2 concentration. The initial time lag for bacteria inactivation is clearly reduced when the hydrogen peroxide concentration is increased. The dependence of the inactivation rate with the H2O2 concentration is not of first order Both models represent reasonably well the experimental data. However, for concentrations of hydrogen peroxide below 100 ppm, the series event model seems to provide a better agreement with the data. Conversely, for oxidant concentrations above 100 ppm, the multitarget model shows a better adjustment.