Corona discharge plasma jet for inactivation of Escherichia coli O157:H7 and Listeria monocytogenes on inoculated pork and its impact on meat quality attributes

Abstract

Corona discharge plasma jet (CDPJ) was used for inactivation of Escherichia coli O157:H7 and Listeria monocytogenes inoculated experimentally onto the surfaces of fresh and frozen pork. CDPJ was generated with an output voltage of 20 kV direct current and a frequency of 58 kHz. Optimal inactivation was found with plasma created at a current of 1.5 A, and at a span length of 25 mm between the plasma electrode tip and the sample. Following CDPJ treatment (0–120 s) of pork samples, reductions in E. coli O157:H7 and L. monocytogenes were 1.5 log and >1.0 log units, respectively. The inactivation pattern fitted well to the Singh-Heldman model or pseudo-first-order kinetics. Compared to untreated pork, with the exception of color and appearance, no statistically significant (P > 0.05) changes were observed in volatile basic nitrogen, peroxide value, or 2-thiobarbituric acid reactive substances of CDPJ-treated pork. Furthermore, CDPJ treatment had no significant impact on the sensory characteristics of frozen pork although there was some effect on unfrozen pork.

Introduction

In meat and meat products, contamination by various pathogenic microorganisms such as Salmonella Typhimurium, Escherichia coli O157:H7, Campylobacter jejuni, Listeria monocytogenes, Clostridium spp., etc., can cause severe foodborne illness (Jayasena et al. 2015). Although conventional thermal treatments are effective in inactivation of foodborne pathogens, they can have a negative impact on the nutrient value and sensory qualities of food (Jayasena et al. 2015). In recent years, however, several novel non-thermal technologies, such as beta and gamma irradiation, power ultrasound, high hydrostatic pressure, pulsed light, UV treatment, treatments using ozone and chlorine dioxide, non-thermal or cold plasmas and so on, are being recognized increasingly for food decontamination purposes (Khadre et al. 2001; Gomez-Lopez et al. 2007, 2009; Misra et al. 2011).

Although plasmas have been known for a long time, cold plasma as bio-decontaminant has drawn substantial attention from food scientists and researchers recently (Basaran et al. 2008; Selcuk et al. 2008; Lee et al. 2015). Non-thermal plasma (NTP)-based techniques are becoming attractive substitutes for conventional chemical-based methods of bio-decontamination as chemicals used in such treatments are known for their intrinsic toxicity and for leaving toxic residues on surfaces (Ehlbeck et al. 2011; Morent and De Geyter 2011; Mok et al. 2015).

NTP can be generated at atmospheric pressure, which makes it more widely applicable (Afshari and Hosseini 2014). At atmospheric pressure, classical approaches for plasma generation include the corona discharge, dielectric barrier discharges (DBD), radio frequency plasma (RFP) and the gliding arc discharge (Misra et al. 2011). Barrier glow discharge-based NTP systems are employed widely for food applications, i.e., for microbial decontamination. The effectiveness of cold atmospheric pressure plasma (APP) for the inactivation of Listeria monocytogenes on the surface of agar plates and slices of cooked chicken and ham (Lee et al. 2011), for inactivation of L. monocytogenes, E. coli and Salmonella Typhimurium on sliced bacon (Kim et al. 2011), and for Listeria innocua decontamination on a sliced ready-to-eat (RTE) meat products (Rod et al. 2012) has been demonstrated. A recent study showed that radio-frequency APP discharges were effective in Staphylococcus aureus reduction, by 3–4 log colony forming units (CFU) after a 10-min treatment time, on beef jerky (Kim et al. 2014).

Corona discharge in air is one of the simplest and best-reproducible sources of NTP. Even the initial expenses and operating costs are lower compared to other kinds of discharges, which generally require complex systems (Scholtz et al. 2010). The chemical composition of plasma generated by atmospheric pressure corona discharges in air is complex and not fully understood (Timoshkin et al. 2012). These discharges are known to produce chemically active species: oxygen ions and other charged species such as N⁺, NO⁺, NO⁻, hydroxyl and hydroperoxyl radicals (OH and HO₂), hydrogen peroxide (H₂O₂), nitrous oxide (N₂O) and other nitrogen oxide species (NO, NO₂, N₂O₄ and N₂O₅), and neutral species in their ground and excited states, including atomic oxygen, ozone and oxygen molecules in the singlet state (Pontiga et al. 2002; Joshi et al. 2011), which act as very strong oxidizers (Deng et al. 2007).

Corona discharges in air have a strong bactericidal effect. Attempts have previously been made to study the inactivation effect of different corona discharges on microbes (Julak et al. 2006; Fletcher et al. 2007; Korachi et al. 2009; Machala et al. 2010; Kim et al. 2015). Cold atmospheric air direct current (DC) discharges (streamer corona) can be used efficiently for bio-decontamination of water and surfaces contaminated by bacteria (Salmonella Typhimurium) (Machala et al. 2010). Corona discharge plasma jet (CDPJ) has been used for surface decontamination of dried laver without adversely affecting its physical as well as functional properties (Kim et al. 2015). Corona discharges have been used even for disinfection of water contaminated with different microorganisms, including bacteria (E. coli, Staphylococcus aureus, Pseudomonas aeruginosa, Streptococcus mutans and Bacillus subtilis), yeasts (Candida albicans), fungi (Aspergillus niger) and green algae (Korachi et al. 2009).

There are no previous studies dealing with the use of CDPJ for foodborne pathogen inactivation on meat and meat products. In this study, pork was used as the meat model substrate. Both fresh and frozen pork were used in the present experiment, because, although freezing is a common practice to preserve meat and to prevent microbial growth, it is not a reliable procedure for eliminating bacterial pathogens in foods (ICMSF 1996; Farkas 1997). E. coli can survive under frozen conditions for long periods. Doyle and Schoeni (1984) demonstrated that E. coli O157:H7 strain 932 can survive for 9 months in frozen ground beef patties with little decline in population. A recent study has shown that enterohemorrhagic E. coli (EHEC) can survive in frozen meat products without a significant decline in the population, even after 180 days of frozen storage (Ro et al. 2015). Therefore, in this study, the effect of CDPJ on inactivation of experimentally inoculated foodborne bacterial pathogens, E. coli O157:H7 and L. monocytogenes, onto pork was investigated. In addition, the possible biochemical changes in pork due to the plasma treatment were evaluated.

Materials and methods

Preparation of samples

Different batches of packaged pork slices (20 × 10 cm), unfrozen (fresh raw meat) and frozen (stored at −20 °C), were purchased from a local market in Seongnam, Korea. The frozen pork slices were thawed in a refrigerator at 4 °C for a brief period of 10 min. The slices were chopped into small pieces measuring ca. 20 × 15 × 10 mm in the sterile environment of a laminar flow cabinet.

Microorganisms and inoculation onto pork

Pure cultures of E. coli O157:H7 (ATCC 43894) and L. monocytogenes (KCTC 3569) were procured from the Korean Culture Center of Microorganisms (KCCM). Both were cultivated individually using tryptic soy broth (TSB) containing 0.6 % yeast extract (BD Company, Le Pont de Claix, France) at 37 °C for 24 h. The pork pieces were ultraviolet light (UV)-sterilized under a biosafety hood for 15 min, and then 100 μL culture from each strain was transferred individually onto the surface of each pork piece. After spreading, the pork samples were maintained at room temperature (25 °C) for 10 min under sterile conditions, enabling the attachment of microorganisms to the pieces. The final cell load in pork samples was approximately 7–8 log CFU/g. Uninoculated UV-sterilized pork pieces (control) also were monitored for sterility throughout the experiment.

CDPJ generation

The corona discharge plasma generating system designed by Mok and Lee (2013) was used in the present study. As shown in Fig. 1, the corona discharge plasma generating electrode was connected to an air blower and power supply. A glass plate was placed on top of the sample stand. For plasma plume protrusion or plasma jet creation, a centrifugal air blower (Ventur Tekniska, Goteborg, Sweden) was operated at a constant rotational speed of 3312 rpm. Thus, the feed gas for the plasma emitter was dried, filtered air. A single-phase 220 V AC power source with an output voltage of 20 kV DC and a frequency of 58 kHz was used for plasma generation. the plasma emission slit in the electrode measures 6 mm × 35 mm. The intensity of plasma can be controlled by varying the strength of the current and span length between electrode tip and sample surface. Initially, we tested the inactivation effect of CDPJ generated in current strength range of 1.0–1.5 A, and span length range of 15–35 mm (data not shown). Optimal inactivation was observed with CDPJ generated at a current strength of 1.50 A and treatment at a span length of 25 mm.

Fig. 1

Schematic diagram of corona discharge plasma jet (CDPJ) treatment system

Treatment with CDPJ

The pork samples were treated with CDPJ (span length: 25 mm, current: 1.50 A) for 0, 30, 60, 90 and 120 s. The sample size for each treatment was n = 3, and each sample consisted of three sub-samples taken from different batches. Each sample was taken aseptically in a glass petri dish and exposed to CDPJ for a predetermined amount of time. Immediately after plasma treatment, samples were subjected to microbial analysis, instrumental color measurement and visual appearance determination. Biochemical analysis was conducted within 48 h post-treatment.

Microbiological analysis

For detection and enumeration of viable inoculated pathogens from pork after treatment, each piece (3 g) was incubated with shaking in sample bag containing 0.85 % sterile saline solution (30 mL) at room temperature (25 °C) for 5 min. Thereafter, the samples were serially diluted using sterile saline solution. A 100-μL aliquot of each dilution was spread on plates containing selective enrichment media (Difco, Becton Dickinson and Co., Sparks, MD): PALCAM agar for L. monocytogenes, eosin-methylene blue agar for E. coli. The plates were incubated at 37 °C for 48 h, and the colonies developed were counted.

Modeling of inactivation

Microbial inactivation is known to follow first-order kinetics as follows (Eq. 1):

$$ \log \frac{No}{N}=\frac{k}{2.303}\cdot t $$

(1)

where N ₀ is initial microbial population, N is microbial population at time t, t is exposure time (min), and k is inactivation rate constant. This equation is valid only for linear inactivation curves. However, as many microbial inactivation curves are non-linear, there have been some reports on the suitability of pseudo-first-order kinetics or Singh and Heldman (2009) for the non-linear survival curves (Mok and Lee 2012; Lee et al. 2015).

The Singh and Heldman model is described by Eq. (2), and was modified as shown in Eq. (3). Using log (t) vs log [log (N ₀/N)], the intercept and slope of the regression line were obtained and the D-like value (D′-value) was calculated.

$$ \log \frac{No}{N}={\left[\frac{t}{D^{\prime }}\right]}^n $$

(2)

$$ \log \left[ \log \left(\frac{No}{N}\right)\right]=n \log (t)-n \log \left({D}^{\prime}\right) $$

(3)

where D′ is D-like value (min) similar to the decimal reduction time, and n is curve shape factor.

Physicochemical properties

Instrumental color measurement

Color characteristics of untreated and CDPJ-treated pork pieces were determined using a Hunter Lab colorimeter (Color difference meter, CR-200, Konica-Minolta, Tokyo, Japan) with an aperture size of 8.0 mm and illuminant D₆₅; values are expressed in terms of the L* (lightness), a* (redness), and b* (yellowness). Also, the total color change (ΔE) was calculated. Prior to analysis, the instrument was calibrated with a standard white plate (L* = 96.77, a* = −0.02, b* = 1.99).

Volatile basic nitrogen

Volatile basic nitrogen (VBN) content was determined using the Conway’s microdiffusion technique as modified by Miwa and Iida (1973).

Peroxide value

The peroxide value (POV) of pork was determined according to the AOAC method 965.33 (AOAC 2005) and expressed as meq O₂/kg lipid.

Thiobarbituric acid reactive substances

Lipid oxidation was measured in terms of 2-thiobarbituric acid reactive substances (TBARS) content using a spectrophotometer (Model T60U, PG Instruments) as described by Buege and Aust (1978). Results were expressed as milligrams of malondialdehyde (MDA)/kilogram muscle.

Sensory evaluation

The sensory attributes evaluated were appearance, color, off flavor and overall acceptability of the product using a 5-point hedonic scale; where 5 was very good; 4 good; 3 common; 2 poor; and 1 very poor. The meat was tasted by a ten-member panel consisting of staff members from our school who were previously experienced and familiar with meat characteristics.

Statistical analysis

Data are presented as the means ± standard error of the mean (SEM) from at least three samples per each treatment condition. All statistical analyses were carried out using the SPSS software package (SAS software, version 9.2; SAS Institute, Cary, NC). From microbial data (plate counting), values of log₁₀ CFU were calculated. The statistical significance (P < 0.05) of the physicochemical and sensory evaluation data was analyzed by a one-way ANOVA test (Duncan’s test). Plasma treatment conditions were fixed and the treatment time was varied.

Results and discussion

Natural microflora of pork

Both fresh and frozen pork naturally contain a considerable number of aerobic bacteria and molds (Table 1). Counts of these two types of microbes were relatively higher in frozen compared to fresh pork; however, other food-borne microorganisms such as E. coli and Listeria were not detected in either sample.

Table 1 Natural microflora of pork pieces. Values are log CFU/g. CFU Colony forming units, SEM standard error of the mean

Inactivation of inoculated foodborne pathogens

Time-dependent inactivation of inoculated E. coli O157:H7 and L. monocytogenes was observed in pork by CDPJ treatment (Fig. 2). E. coli seems to be more susceptible to corona discharge plasma, which were reduced by about 1.5 log units (96.84 %) in 120 s treatment, than the L. monocytogenes (above 1.0 log or 90 %). There was a significant difference in the susceptibility of E. coli to the plasma between unfrozen and frozen pork. Higher E. coli susceptibility was noted in unfrozen pork. In unfrozen pork, more than 1.0 log reduction was observed within 60-s treatment time. However, no significant difference in the susceptibility of L. monocytogenes to the plasma was noted between frozen and unfrozen pork; and 1.0 log reduction was observed at 90–120 s treatment. The higher rate of E. coli inactivation in fresh pork was probably due to species-specific susceptibility under increased moisture and temperature conditions. Freezing, frozen storage and thawing all contribute to a decrease in the water-holding capacity of meat (Vieira et al. 2009; Leygonie et al. 2012).

Fig. 2

Inactivation pattern of a Escherichia coli O157:H7, and b Listeria monocytogenes on pork by CDPJ treatment (span length: 25 mm, current: 1.50 A)

Different atmospheric pressure plasmas have been used successfully for microbial inactivation in meat and related products. Cold atmospheric gas plasmas have the potential to decontaminate chicken muscle; 3 log reductions in L. innocua were achieved for muscle after a 4-min treatment (Noriega et al. 2011). APP jets were found effective for the inactivation of L. monocytogenes on sliced meats and for prolonging the shelf-life of such foods (Lee et al. 2011). In the latter study, after treatment for 2 min with APP jets of He, He + O₂, N₂, or N₂ + O₂, L. monocytogenes inoculated onto sliced chicken breast and ham were reduced by 1.37 to 4.73 and 1.94 to 6.52 log units, respectively. Recently, inactivation of pathogens on sliced cheddar cheese by using thin-layer dielectric barrier discharge (DBD) plasma was demonstrated by Yong et al. (2015), who showed that study, in response to a 10-min plasma treatment, E. coli O157:H7, L. monocytogenes, and Salmonella Typhimurium populations decreased significantly by 3.2, 2.1 and 5.8 log CFU/g, respectively. Indirect plasma treatment was able to prolong the shelf life of porcine musculus longissimus dorsi with regard to microbiological contamination. A plasma exposure time of 2 × 2.5 min was sufficient to hold the microbial load at the detection limit of 10² CFU/g over a storage period of 20 days at 5 °C (Frohling et al. 2012).

In an earlier study, atmospheric corona discharges obtained from both DC and AC high voltage power supplies were able to sterilize water efficiently from the majority of microorganisms tested (Korachi et al. 2009). The predominant reactive species in corona plasma was reported to be ozone (Gaunt et al. 2006). Positive steady-state corona discharges in air normally generate less ozone as compared with negative corona at the same energization potential (Timoshkin et al. 2012). The energy of the reactive species can be transferred effectively to other molecules through inelastic collisions (Zou et al. 2006). The mechanism of corona plasma inactivation is not yet fully understood. The predominant mechanism of inactivation is believed to occur through oxidative damage from reactive oxygen species (ROS), according to earlier reports (Machala et al. 2010; Dobrynin et al. 2011). ROS species are able to induce damage to biomolecules, including DNA and proteins. They can induce alterations in the functions of biological membranes by interacting with lipids (Timoshkin et al. 2012).

Modeling of inactivation pattern

Kinetic modeling of inactivation provides information for quantitative risk assessment. A scatter plot shows the linear relationship between the two variables, log inactivation and time of plasma exposure. At the initial stages of plasma exposure, 0–60 s, rate of inactivation was increased at an exponential rate (data not shown). However, in the latter half, the slope becomes gentler, which implies that the inactivation effect was reduced. Using a first-order kinetic inactivation model, no satisfactory explanation can be given for the inactivation pattern over the entire test period. Compared to first-order kinetics for inactivation, Singh-Heldman model (pseudo-first-order) fitted better to the non-linear inactivation curves of E. coli and L. monocytogenes, as shown in Table 2. The coefficients of determination (R ²) were closer to 1, implying that inactivation patterns are better explained using pseudo-first-order kinetics. Also, the high coefficient of determination confirms a stronger inactivation effect of CDPJ on the microorganisms tested.

Table 2 Comparison of regression analysis for first-order and pseudo-first-order inactivation kinetics of Escherichia coli and Listeria monocytogenes in unfrozen and frozen pork

To date, the experimental work on the germicidal effects of cold, atmospheric pressure plasmas has shown that survivor curves take different shapes depending on the type of microorganism, the type of medium supporting the microorganism, and the method of exposure, “direct exposure” or “remote exposure” (Laroussi 2005).

D′-values

The D′-value is a measure of the heat resistance of a microorganism. D′-value is similar to, if n = 1, decimal reduction time (D-value) in case of heat sterilization, which is the time required at a certain temperature to kill 90 % of the organisms being studied. E. coli cells on unfrozen pork were susceptible to CDPJ-induced inactivation relatively quickly compared to those on frozen pork, as lower decimal reduction values were noted on unfrozen pork (Table 3). In the case of L. monocytogenes, the inoculated cells were almost equally susceptible to inactivation effect regardless of meat type, i.e., frozen or unfrozen. Several factors other than plasma intensity and exposure time influence the initial survival rate of bacteria during plasma exposure (Lee et al. 2015). It has been shown that the inactivation effects of APP on L. monocytogenes are strongly dependent on the type of food (Song et al. 2009). In that study, after 120 s APP treatments at 75, 100, and 125 W, the viable cells of the cocktail of L. monocytogenes were reduced by 1.70, 2.78, and 5.82 log in sliced cheese, respectively (Song et al. 2009). More than 8 log reductions can be achieved in 120 s at 150 W. In contrast, reductions after 120 s ranged from 0.25 to 1.73 log CFU/g in sliced ham.

Table 3 Inactivation decimal reduction values (D′-value) of Escherichia coli and Listeria monocytogenes in pork by CDPJ treatment (span length: 25 mm, current 1.50 A). Values (mins) are given as mean ± standard error of the mean (SEM; n = 3)

Physicochemical properties

Color and visual appearance

As shown in Table 4, there was a significant (P < 0.05) change in brightness (L*-value) and yellowness (b*-value) of unfrozen pork upon CDPJ exposure. Both brightness and yellowness increased markedly; however, the changes in redness (a*-value) were insignificant. Higher b* values could be due to a high concentration of metmyoglobin in the pork. Metmyoglobin is formed by oxidation of deoxymyoglobin or oxymyoglobin (Mancini and Hunt 2005) and the radicals formed in the plasma may lead to oxidation to metmyoglobin, resulting in higher b*-value (Frohling et al. 2012). In the case of frozen pork, values of the three color parameters (brightness, redness and yellowness) were altered profoundly due to the plasma treatment. Brightness was decreased and redness as well as yellowness were increased significantly (P < 0.05), compared to untreated pork. The decreased brightness could be due to accelerated lipid oxidation in the frozen pork (Frohling et al. 2012). Overall, the total color change (ΔE) differed significantly between untreated and CDPJ-treated pork. According to Francis and Clydesdale (1975), when ΔE > 3, color differences are obvious to the human eye. Therefore, as the ΔEfor unfrozen pork treated for 60, 90 and 120 s was > 4, color differences were visually noticeable as shown in Fig. 3. Figure 3 also depicts the enhanced brightness and redness of the CDPJ-treated unfrozen pork, compared to the untreated pork. In addition, the decreased brightness in frozen pork upon plasma exposure is just noticeable.

Table 4 Changes in color characteristics of CDPJ-treated unfrozen and frozen pork. Values are given as mean ± SEM (n = 3). Distinct lower case letters within the same column indicate significant differences (P < 0.05). L* Lightness, a* redness, b* yellowness

Fig. 3

Visual appearance of pork pieces upon treatment by CDPJ (span length: 25 mm, current: 1.50 A)

Compared to the plasma-untreated fresh pork, frozen pork (untreated) exhibited a higher brightness value and a low redness value. Myoglobin accounts, in part, for the change in the color stability of meat after freezing and thawing (Anon and Cavelo 1980). At some stage during freezing, frozen storage and thawing, denaturation of the globin moiety of the myoglobin molecule occurs, as reported by Calvelo (1981). The denaturation leads to an increased susceptibility of myoglobin to autoxidation and subsequent loss of optimum color presentation (Leygonie et al. 2012).

Levels of VBN, POV and TBARS

Total VBN content is an important reference index in evaluating pork freshness (Huang et al. 2014). In the present study, there was no significant change (P > 0.05) in VBN content of either unfrozen or frozen pork upon plasma treatment (Table 5). The initial VBN content was about 6.16 mg VBN/100 g pork. During the CDPJ treatment, the total VBN levels of pork pieces increased insignificantly even up to 120 s. When the VBN <15 mg/100 g, the pork is considered fresh; when the value is between 15 to 30 mg/100 g, the pork is stale meat; when VBN > 30 mg/100 g, the pork is rotting (Xiao et al. 2014). Therefore, it can be concluded that pork freshness is unaffected by plasma treatment.

Table 5 Changes in volatile basic nitrogen (VBN) content, peroxide value (POV) and 2-thiobarbituric acid reactive substances (TBARS) of pork upon CDPJ treatment (span length: 25 mm, current 1.50 A). Values are given as mean ± SEM (n = 3). Distinct letters within the same column indicate significant differences (P < 0.05). MDA Malondialdehyde

During lipid oxidation, although lipid hydroperoxides are the primary products, they eventually convert to peroxides (Kim et al. 2013a). Therefore, to study the extent of oxidation in meat, it seemed reasonable to determine the concentration of peroxides in the pork samples. Peroxide value (POV) is commonly expressed as milliequivalents (meq) of active oxygen per kilogram of lipid. As shown in Table 5, POV of unfrozen pork was not changed significantly (P > 0.05) during CDPJ treatment. The POVs of unfrozen pork were in the range 10.33–10.67 meq/kg, after treatment at different intervals. On the other hand, compared to the control, POV significantly (P < 0.05) increased in frozen pork upon CDPJ treatment. Mean POVs of pork after 60, 90 and 120 s treatment durations were 12.33, 12.67 and 13.33 meq/kg, respectively. The increase of POV in frozen pork during plasma treatment may result from catalysis of intracellular contents that oozed out due to the destruction of cell structure by freezing. The peroxidation of lipids may also be facilitated by oxygen during frozen storage. Increased POV during storage of processed pork has been reported (Juntachote et al. 2006).

TBARS are formed as a byproduct of lipid peroxidation. Compared to untreated pork, there were no significant changes (P > 0.05) in TBARS levels of CDPJ-treated pork. TBARS levels were 0.55–0.56 mg MDA/kg for unfrozen pork, and 0.54–0.55 mg MDA/kg for frozen pork. According to previous studies, the organoleptic detectable threshold of off-flavor in meat is observed around TBARS values of 0.5–2.0 mg MDA/kg (Chang et al. 1961; Gray et al. 1996).

Sensory characteristics

Appearance, color, off-flavor and overall acceptance of unfrozen pork were significantly (P < 0.05) altered upon plasma exposure compared to untreated control pork; however, these changes were not statistically significant in the case of frozen pork (Table 6). The average scores for pork appearance were in the range of 4.17–2.92 (unfrozen) and 3.92–3.75 (frozen), on a 5-point hedonic scale. Other sensory parameter scores include pork color: 4.17–2.33 for unfrozen, 4.42–4.17 for frozen; off-flavor: 3.08–2.25 for unfrozen, 2.67–2.17 for frozen; and overall acceptance: 3.67–2.25 for unfrozen, 4.17–3.42 for frozen. Overall, compared to untreated pork, sensory scores of CDPJ-treated unfrozen pork deteriorated significantly over the 120 s treatment time. An earlier study reported similar results; DBD-plasma treatment substantially lowered the scores measured for appearance, color, odor, and overall acceptability of raw pork loin (Kim et al. 2013b). It has been reported that free radicals generated during plasma treatment trigger lipid and/or protein oxidation and generate secondary oxidation products such as alkanes, alkenes, aldehydes, alcohols, ketones, and acids (Kim et al. 2013b). These molecules produce fishy, metallic, rancid, and oxidized flavors (Kochhar 1996). The temperature difference between unfrozen and frozen pork could also contribute to the enhanced off-flavors in the unfrozen, plasma-treated pork.

Table 6 Changes in sensory properties of CDPJ-treated pork. Values are given as mean ± standard error of the mean (SEM; n = 3). Distinct lower case letters within the same row indicate significant differences (P < 0.05)

The deterioration of sensory characteristics of CDPJ-treated unfrozen pork should be considered and minimized. Addition of natural antioxidants to meat and meat products has been recommended as a possible strategy to minimize lipid oxidation (Jayasena and Jo 2014). Therefore, such interventions would improve the sensory quality of CDPJ-treated unfrozen pork.

Conclusions

In conclusion, although CDPJ showed somewhat low inactivation effect on L. monocytogenes compared to E. coli, it still holds promise as the plasma exposure time was minimal. The inactivation kinetics of the pathogens tested can be better explained using Singh-Heldman model. It is clear from these results that the microbial safety of pork can be enhanced using CDPJ treatment. Regarding physicochemical properties, there were small but significant changes in a few instrumental color parameters of pork upon CDPJ treatment. VBN content was unchanged due to treatment; total VBN levels in the treated samples were well below the recommended limit for pork freshness. POV was relatively high in plasma-treated frozen pork, but not to abnormal levels; and TBARS content remains unaffected during the plasma treatment, indicating that lipid oxidation was minimal. Sensory scores of the plasma-treated frozen pork were similar to that of untreated frozen pork although changes were noted in unfrozen samples. Such deterioration in sensory quality could be minimized by using antioxidants. Overall, it can be concluded that CDPJ is suitable for E. coli and L. monocytogenes decontamination on pork without affecting its physicochemical and sensory characteristics.