Quantifying bacterial thermal resistance for drying and roasting processes using a novel low-humidity control method

Accurate mathematical models are vital for validating the microbial safety of low-moisture food processes like high temperature drying and roasting. However, model development is hindered by a lack of thermal resistance data for pathogens such as Salmonella spp. under relevant high-temperature, low-humidity conditions. This study aimed to develop and validate a method using conditioned desiccants to control humidity at temperatures above 90 °C. We evaluated three desiccants, silica gel, activated alumina, and molecular sieve 3A, for their ability to control relative humidity (RH) in a thermal a_w cell (TAC) at temperatures from 80 to 140 °C. System equilibration was characterized, and the inactivation kinetics for a three-strain Salmonella cocktail and its surrogate Enterococcus faecium were determined at 120 °C under various low-RH conditions. Results showed that molecular sieve 3A was most effective at creating extremely dry conditions (approaching 0% RH), due to its high monolayer moisture binding capacity. Temperature and humidity equilibration within the TDT cell headspace was rapid, occurring in under 3 min. Microbial inactivation for both Salmonella and E. faecium followed first-order kinetics. D-values at 120 °C showed no significant difference when using the three unconditioned (or raw) desiccants (minimum RH) but was significantly dependent on RH in the 0-40% range, confirming the extreme protective effect of desiccation at elevated temperatures. D-values at 120 °C for the Salmonella cocktail, for example, plummeted 27-fold from 10.8 ± 4.2 min at ∼0.5% RH to 0.43 ± 0.14 min at 29.7% RH, demonstrating a critical dependence on humidity. This work successfully demonstrates a reliable method for generating critical thermal inactivation data in previously difficult-to-study, high-temperature, low-humidity environments. The resulting kinetic parameters can be used to strengthen predictive models, enabling more robust process validation for the food industry.