Milk powder is a versatile food ingredient with several benefits. It is highly nutritious and has a long shelf life which facilitates its storage and distribution (Zhong et al., 2017). Milk powder can be used as reconstituted milk or as an additive in various food products, such as puddings, sauces, and baked goods to enhance texture and flavour (Wei et al., 2021). Due to its low water activity (aw), it is commonly perceived as “safe”. The low aw creates an adverse environment that inhibits bacterial growth. However, several microorganisms can survive for prolonged periods of time, creating challenges with respect to food safety upon rehydration of the product (Dag et al., 2022).

Over the years, several outbreaks have been traced back to contaminated milk powder (ECDPC and EFSA, 2019; Michael et al., 2014). One of the most prevalent foodborne pathogens implicated in these outbreaks is Salmonella. Salmonella is known to survive the spray drying process which is the last critical control point for eliminating microbial contamination (Wei et al., 2021). Listeria monocytogenes is another microorganism that can withstand the spray drying process and survive in low moisture foods for an extended period of time, even if it hasn't been linked to any known milk powder outbreaks thus far (Ballom et al., 2020; Ly et al., 2019).

The reduction of microbial contamination in food products is typically addressed through conventional thermal processing technologies (Zhang et al., 2021a). Such technologies have certain limitations regarding the heating rate and uniformity, especially when it comes to the treatment of low moisture food products like milk powder. Radio Frequency (RF) heating is an innovative technology which utilises electromagnetic radiation and generates thermal energy within the product (Altemimi et al., 2019). Numerous studies are available on the RF treatment of various low moisture food products, with a particular interest on milk powder (Dag et al., 2020). The milk powder's composition exerts a big influence on the RF treatments (Dag et al., 2019), but the influence on the bacterial survival in milk powders has not been studied yet. Furthermore, RF treatments have been mainly focused on Salmonella spp., despite the fact that L. monocytogenes, if present, can survive storage up to a year (Ballom et al., 2020). In most of the RF inactivation studies, a common approach involves the initial use of RF to elevate temperatures to a specific target, followed by isothermal holding using (i) hot air for a set duration at the target temperature (Michael et al., 2014; Wang et al., 2016; Wei et al., 2021) or (ii) placing products in insulation boxes (Ballom et al., 2021). Oven holding involving hot air is the most commonly applied strategy, since small temperature decreases can often occur during product transfer to insulation boxes (Ballom et al., 2021). However, the stage of isothermal holding needed to allow time for internal heat to mitigate non-uniform RF heating (i.e., to ensure bacteria in the cold zones receive adequate treatments) can result in overtreating (i.e., too long time at high temperature) of the hot zones, consequently leading to a decline in quality (Wang et al., 2016; Dag et al., 2022). RF-only treatments consisting of a slow RF heating phase and no conventional oven holding phase could be used to better preserve product quality, according to a Low Temperature Long Time (LTLT) strategy (Dominguez-Hernandez et al., 2018; Liu et al., 2022a). However, this strategy requires more research prior to industrial application, since (i) RF has not been used as a sole technology aiming at bacterial inactivation thus far, and (ii) the potential of RF-only dynamic heating has not been fully explored yet. In contrast to typical RF processes involving an oven heating phase, RF heating is more significant in the overall process, and hence, the interaction between RF, the powder properties (e.g. composition) and the specific pathogenic microorganisms (e.g., influence of the bacterial gram properties), becomes a critical aspect to consider. The effect of these influences has not been studied so far, not for RF processing in general, and especially not in cases involving an RF-only dynamic heating process.

The main objective of this study was to elucidate the influence of milk powder composition and bacterial properties on RF inactivation kinetics during a RF-only dynamic decontamination process. The specific subobjectives were to (i) compare the RF inactivation kinetics of two microorganisms with different Gram-types (i.e., S. Typhimurium and L. monocytogenes), (ii) unravel the effect of the powder's composition on the microbial survival (i.e., using skimmed and whole fat milk powder), and (iii) validate the use of the non-pathogenic microorganism Enterococcus faecium as a surrogate for the aforementioned microorganisms, food products and RF process. In order to study the influence of these factors solely during RF exposure, a novel, RF-only (non-isothermal) temperature profile was used, not including the typical conventionally heated temperature holding stage. A log-linear model including a Bigelow-type temperature dependency was fitted to the acquired experimental data in order to systematically quantify the effect of the aforementioned factors.