Inactivation of internalized Salmonella Typhimurium in lettuce and green onion using ultraviolet C irradiation and chemical sanitizers



Jiyoung Lee, 406 Cunz Hall, 1841 Neil Avenue, Columbus, OH 43210, USA. E-mail:



The internalized human pathogens in fresh produce are not effectively removed during conventional washing, and therefore, it may cause foodborne illness when the produce is consumed raw. Thus, effective nonthermal processes are needed to prevent this risk.

Methods and Results

Green fluorescence protein-tagged Salmonella Typhimurium was either sprayed on the surface of iceberg lettuce or injected into the bottom part (bulb) of green onions to induce bacterial internalization. The contaminated vegetables were collected after 2 days and subjected to surface disinfection. Different fluencies of UV-C radiation (75–900 mJ cm−2) and two fluencies of UV-C (450, 900 mJ cm−2) combined with chlorine and peracetic acid (PAA) were applied to the produce to examine the inactivation efficiency of internalized bacteria. A range of 1·96–2·52 log reduction in the internalized Salmonella was achieved when the lettuce was treated with higher UV-C fluency (150, 450, 900 mJ cm−2) or UV-C combined with chemical disinfectants. Significant reduction (1·00–1·49 log CFU g−1) in internalized Salmonella was observed in green onion treated with UV-C with the fluency of 150 or 900 mJ cm−2 or UV-C-chlorine/PAA. No significant reduction was observed in either lettuce or green onion treatments when chlorine or PAA was used alone. The food quality measured with firmness was not changed during any treatments. However, a slight colour change was observed in lettuce only when UV-C was used at 900 mJ cm−2.


High fluency UV-C can significantly inactivate the internalized Salmonella in lettuce and green onion while maintaining the food quality.

Significance and Impact of the Study

This research provides applicable research outcomes for developing nonthermal methods to inactivate internalized pathogens in fresh produce.


According to the estimation of the United States Centres for Disease Control and Prevention (CDC), foodborne pathogens caused illness in one of six Americans (48 million people), 128 000 hospitalizations and 3000 deaths each year (Centers for Disease Control and Prevention 2012). During the past two decades, fresh produce-linked foodborne diseases increased dramatically due to its increased consumption (Heaton and Jones 2008). The fresh produce consumption is encouraged because of its health benefits; however, fresh produce can be contaminated with human pathogens in any stage of the food supply chain, which increases the risk of foodborne outbreaks (Lynch et al. 2009).

Many bacterial types have been reported to become internalized in fresh vegetables and fruits, including Salmonella and Escherichia coli (Delaquis et al. 2007; Kroupitski et al. 2011; Deering et al. 2012). To improve food safety and maintain the freshness, the products are subject to washing with chemical sanitizers, and one of the most commonly used disinfectants in food industry is chlorine (50–200 ppm). However, it is reported that chlorine is not effective to inactivate internalized pathogens due to its limited penetration into the internal areas of the plants (Sapers 2001). In addition, previous research showed that chlorine reacted with organic matter and can generate disinfection by-products (DBPs), such as chloroform and bromodichloromethane, which are known as carcinogenic (Chu et al. 1982; Reckhow et al. 1990; Environmental Protection Agency 1998). Therefore, investigation of other disinfection methods is warranted while maintaining the food safety and quality. For instance, to minimize toxic DBPs formation, PAA has been used as a food disinfectant for decades to treat fresh produce, seeds and equipments (Greenspan and Margulies 1950; Hei 2000). The potential DBPs from PAA decompose spontaneously into acetic acid, water and oxygen, which are not a threat to human health (Dell'Erba et al. 2007).

Other than chemical-based disinfectants, one of the most frequently studied fresh produce disinfection methods is ionizing radiation, especially gamma radiation, due to its high penetration ability (Niemira 2008; Gomes et al. 2009). Despite its desired reduction in internalized pathogens, the use of gamma rays in food processing has not been well perceived by consumers. In addition, the cost, equipment maintenance and process safety are of concern for full-scale industrial applications (Durante 2002; Mahapatra et al. 2005; Stefanova et al. 2010). UV-C (254 nm), as a nonionizing radiation, demonstrates powerful germicidal activity and has been used in food, water and solid surface disinfection (Guerrero-Beltrán and Barbosa-Cánovas 2004; Yaun et al. 2004; Cutler and Zimmerman 2011). Several UV disinfection systems have been created and patented (Brandt and Klebaum 2000; Uchino et al. 2008; Yuan 2009). SAMRO Ltd., a Swiss company, developed the Ventafresh technologies ready for marketing, which integrated gaseous ozone and UV radiation primarily for root vegetable and fruit disinfection (Steffen et al. 2010). Hadjok et al. (2008) reported that the combination of UV-C radiation and hydrogen peroxide resulted in a 2·84 log reduction in the internalized Salmonella Montevideo in lettuce, which was the only study investigating the efficacy of the internalized pathogens inactivation using UV-C radiation. Before this study, UV-C was commonly considered having low penetration into the internal region of the food matrix (Morgan 1989). There is a research need for better understanding the effectiveness of UV-C and other disinfectants in inactivating internalized human pathogens in fresh produce, especially their combined treatments and synergistic effects for broad application of this approach for industrial practice.

In this study, the effectiveness of different fluencies of UV-C irradiation was tested in inactivating internalized Salmonella Typhimurium in iceberg lettuce and green onion, which are often consumed raw. In addition, the combined treatments with UV-C and chemical disinfectants, chorine and peracetic acid (PAA) were examined. The food quality was determined after the treatments by measuring colour and texture of the fresh produce. The outcome of this research can provide a platform for future studies about inactivation of the internalized pathogens with nonthermal processing and practical application of UV-C radiation in fresh produce industry.

Materials and methods

Plant growth

Iceberg lettuce (Lactuca sativa var. capitata) and green onion (Allium fistulosum) seeds were purchased from Burpee and Co. (Warmister, PA, USA). All the plants germinated and grew in a greenhouse for 4 weeks at the Department of Horticulture and Crop Science, The Ohio State University (Columbus, OH, USA). The cultivation conditions in the greenhouse were followed as described in our previous study (Ge et al. 2012). The 4-week-aged lettuce and green onion were transported to a laboratory growing system for Salmonella inoculation. The growing conditions were: 23–27°C, 30–45% of relative humidity, irrigated with 250 ml of tap water per pot (each pot was filled with 360 g sterile soil (Metro-Mix 360; SunGro Horticulture, Vancouver, BC, Canada) and 16 h of light (Hydrofarm Fluorescent Grow Light System, Fairless Hills, PA, USA) per day.

Salmonella Typhimurium inoculation

S. Typhimurium (ATCC 19585) was labelled with green fluorescence protein (GFP) using the method in our previous study (Ge et al. 2012). The GFP-labelled S. Typhimurium was cultured in Luria-Bertani (LB) broth supplemented with 100 μg ml−1 of ampicillin for 18 h at 37°C in a shaking incubator (200 rpm). The broth was centrifuged at 10 000 g for 10 min, and the bacterial pellet was resuspended in deionized (DI) water (Milli-Q; Millipore, Billerica, MA, USA). The concentration of S. Typhimurium was measured using a cell density meter (WPA Biowave; Biochrom, Cambridge, UK) and was later confirmed with a plate count method. The bacteria suspension was diluted to c. 108 colony forming unit (CFU) ml−1 using DI water and then was sprayed onto the adaxial surface (upper side) of the lettuce leaf. In the preliminary experiments, the internalization incidence of different batches of green onion showed very inconsistent results when using surface spraying. Therefore, an injection method was applied to effectively generate Salmonella internalization in the green onion. Each head of green onion was injected with 10 μl of S. Typhimurium (108 CFU ml−1) at the location of 5 mm above the soil. The contaminated plants were grown in the laboratory growth system for 2 more days as described previously to simulate the internalization of human pathogens during the preharvest. One group of lettuce/green onion without S. Typhimurium inoculation was used as a control group.

Plant harvest and inactivation of bacteria on surface

The lettuce leaves (edible part) were cut at 5 mm above the soil, and the green onions were collected after removing the roots. Then, all the plants were disinfected with 80% ethanol for 10 s, 1% AgNO3 for 5 min, flushed with tap water and DI water sequentially to inactivate and remove the bacteria on the plant surface (Franz et al. 2007). To verify the surface disinfection efficiency, lettuce and green onion were sprayed with S. Typhimurium (108 CFU ml−1) on their surfaces while making sure there was no dripping. All the inoculum droplets on the plants were dried in a laminar flow biosafety cabinet (Thermo Scientific 1300 Series A2 Class II, Asheville, NC, USA). Then, the plants were subjected to the surface disinfection as described earlier, and the samples were processed to check their bacterial concentrations as described below. A total of 10 samples of each plant were tested.

Disinfection of internalized Salmonella with UV-C irradiation, chlorine and PAA

A UV-C disinfection system was constructed in our laboratory. The whole enclosure was made of plywood (60 cm × 60 cm × 80 cm) with a door in the front. Two UV-C lamps (254 nm, 15 W; Philips TUV15 W G15T8 Long L, Eindhoven, the Netherlands) were positioned on the top of the enclosure. The intensity of UV was determined using a UV-C meter (UV512C, Digital UV-C Meter; General Tools, New York, NY, USA). The UV-C fluency was adjusted by changing the distance between the UV-C lamps and food samples (changing intensity) and also by changing treatment time. UV fluency (mJ cm−2) was calculated as intensity × time. Chemical disinfectants, chlorine (Clorox, Oakland, CA, USA) and PAA (35%, FMC Corporation, Philadelphia, PA, USA), and the combinations of these disinfectants with UV-C was tested as well to assess their effectiveness in inactivation of the internalized S. Typhimurium in the plants. The entire disinfection treatments used in this study are summarized in Table 1. All the lettuce was placed on the aluminium alloy process trays with the adaxial side up towards the UV-C bulbs for irradiation (no overlap between the leaves). About 25 g lettuce leaves was spread on a tray to be processed at a time (Treatments 9–16), and 100 g lettuce was processed using each treatment method. A total of 10 heads of green onion (c. 10 g) were loaded on the process tray for UV-C irradiation at a time, and the samples were turned every 2 min using sterile tweezers to facilitate the entire surface of green onion was exposed to the UV-C light during the treatment (UV light was turned off while rolling the green onion). Prior to the UV irradiation, 250 ml of DI water was poured onto the vegetables to prevent water loss or wilting under a high fluency of UV-C radiation (observed in the preliminary study). In the treatments using chemical disinfectants (Treatment 1–8), the lettuce leaves or the green onion samples were submerged into the disinfectant solutions (100 or 200 ppm of chlorine; 40 or 80 ppm of PAA) for 5 or 10 min and then rinsed with DI water. For the combined treatments (Treatments 13–16), the plants were irradiated with UV-C light first and sequentially treated with chlorine or PAA solution. The plants were also rinsed with DI water to eliminate the residual chemicals on the surface. All the 16 treatments were completed at the room temperature (c. 25°C) and repeated three times.

Table 1. Each treatment of contaminated fresh produce using UV-C, chorine, peracetic acid (PAA) and combined methods
Treatment numberTreatment methoda
  1. a

    Treatment parameters included UV-C intensity and irradiation fluency, concentrations of chlorine and PAA, contact time and C × t value.

1Chlorine (100 mg l−1 for 5 min, C × t value = 500 mg min l−1)
2Chlorine (100 mg l−1 for 10 min, C × t value = 1000 mg min l−1)
3Chlorine (200 mg l−1 for 5 min, C × t value = 1000 mg min l−1)
4Chlorine (200 mg l−1 for 10 min, C × t value = 2000 mg min l−1)
5PAA (40 mg l−1 for 5 min, C × t value = 200 mg min l−1)
6PAA (40 mg l−1 for 10 min, C × t value = 400 mg min l−1)
7PAA (80 mg l−1 for 5 min, C × t value = 400 mg min l−1)
8PAA (80 mg l−1 for 10 min, C × t value = 800 mg min l−1)
9UV-C (0·25 mW cm−2 for 5 min, irradiation fluency = 75 mJ cm−2)
10UV-C (0·25 mW cm−2 for 10 min, irradiation fluency = 150 mJ cm−2)
11UV-C (1·5 mW cm−2 for 5 min, irradiation fluency = 450 mJ cm−2)
12UV-C (1·5 mW cm−2 for 10 min, irradiation fluency = 900 mJ cm−2)
13UV-C (1·5 mW cm−2) for 5 min (irradiation fluency = 450 mJ cm−2) + chlorine (200 mg l−1) for 5 min (C × t value = 1000 mg min l−1)
14UV-C (1·5 mW cm−2) for 10 min (irradiation fluency = 900 mJ cm−2) + chlorine (200 mg l−1) for 10 min (C × t value = 2000 mg min l−1)
15UV-C (1·5 mW cm−2) for 5 min (irradiation fluency = 450 mJ cm−2) + PAA (80 mg l−1) for 5 min (C × t value = 400 mg min l−1)
16UV-C (1·5 mW cm−2, irradiation fluency = 900 mJ cm−2) + PAA (80 mg l−1) for 10 min (C × t value = 800 mg min l−1)

Microbial test: quantification of the internalized Salmonella Typhimurium

Each batch of processed plants was cut into pieces using sterilized scissors and mixed to minimize the variability in the microbial enumeration. Ten grams of lettuce leaves and 1 g of green onion were mixed with 30 and 10 ml peptone water (0·1%), respectively, and the mixture was sealed in a sterile Whirl-pak© bag (NASCO, Fort Atkinson, WI, USA) for 2 min-homogenization using a stomacher (Stomacher® 80, Seward, West Sussex, UK). The blended plant suspension (100 μl) was plated onto the LB agar supplemented with 100 μg ml−1 of ampicillin in duplicate. The plates were incubated at 37°C for 18 h. Only the colonies that showed green fluorescence under UV light (365 nm) were counted.

Food quality test: colour

The plant colour was measured using a portable tristimulus colorimeter with an 8 mm diameter measuring area in the L*a*b* mode (CR 300 series Chroma Meters; Konica Minolta, Ramsey, NJ, USA). The values of L* (lightness, white-black), a* (green-red) and b* (blue-yellow) at three selected areas on the green onion surface (samples were sliced to form a single layer to cover the whole measurement area) and the adaxidal side of lettuce leaf surface were measured. The hue angle, which was used as an important indicator of discoloration, can be calculated using the following equation; Hue angle = Tan−1 (b*/a*) (Castaner et al. 1996; Hosoda et al. 2000). For both colour test and texture analyses, the plant samples that went through the exact same treatments as listed in Table 1 were used, but they were not contaminated with S. Typhimurium because any samples containing pathogens were not allowed to put in the colorimeter and texture analyser in our research setting. Ten samples from each treatment were randomly selected for the colour measurement.

Food quality test: texture

The texture of the leaves was determined using a Stable Micro Systems TA.XT Plus texture analyser (Texture Technologies Corporation, Scarsdale, NY, USA) with a TA-91 Kramer Shear 5 blade probe. The plants were cut and tailored to c. 8 cm, and the sample size for one measurement of lettuce and green onion were 10 and 2 g, respectively. The detection parameters were set as follows: pretest speed at 6·0 mm s−1, test speed at 2·5 mm s−1, post-test speed at 10·0 mm s−1 and return distance of 2 cm. Six replicates were measured in each treatment. The peak force (kg) was recorded using Texture Expert software Version 1·22 (Texture Technologies Corp.) to express the firmness for evaluating the leaf texture.

Statistical analysis

The experimental data, including the microbial reduction test for determining the disinfection efficiency, colour and texture tests for the food quality after treatments, were analysed using SPSS 17.0 statistical software (SPSS Inc., Chicago, IL, USA). Analyses of variance (anova) were performed to determine the differences between the means of the data in each independent experiment. Results were considered significant at < 0·05.


Reduction in internalized Salmonella Typhimurium in fresh produce

To focus on internalized bacteria only, the efficiency of surface disinfection steps was tested and confirmed by following Franz et al. (2007). There were no remaining bacteria on the plant surfaces after going through the surface disinfection (80% ethanol and 1% AgNo3 treatment) in any plant samples that were tested. This demonstrated that the microbial counts obtained from all the experimental settings reflect only the internalized S. Typhimurium in the fresh produce. The reduction in the internalized S. Typhimurium in the lettuce and green onion after all of the 16 treatments is summarized in Tables 2 and 3, respectively. The number of samples that contained remaining internalized Salmonella was also counted and is presented in the Tables 2 and 3. Among the untreated lettuce samples, 28 of 30 lettuce samples and all of the green onion samples showed positive result of Salmonella internalization. The initial concentrations of internalized Salmonella in the lettuce and green onion were 2·76 ± 1·11 and 4·10 ± 1·12 log CFU g−1, respectively. The chlorine or PAA treatments (Treatment 1–8) did not significantly reduce the Salmonella in lettuce. The reduction efficiency ranged from 0·43 to 0·99 log CFU g−1 and the internalized viable Salmonella remained in 66·7–86·7% of the lettuce samples. When the lettuce leaves were treated with higher fluency of UV-C (Treatment 10–12 with 150, 450 and 900 mJ cm−2), the internalized S. Typhimurium was significantly lower than the untreated sample (< 0·05) and more than 2-log reduction was achieved using these UV-C fluencies. Even though the Treatment 9 (UV-C fluency of 75 mJ cm−2) caused 1·40 log CFU g−1 of Salmonella inactivation, the remaining viable internalized Salmonella was not significantly different from the untreated group. The combined treatments (Treatment 13–16) also resulted in a significant decrease in the internalized Salmonella, and among those, the combination of UV-C (900 mJ cm−2) and PAA (800 mg min l−1) achieved the highest Salmonella reduction (2·52 log CFU g−1). However, the reductions in internalized Salmonella that resulted from the combined treatments (Treatment 13–16) and the high fluencies of UV-C (Treatment 10–12) were not significantly different (> 0·05). The concentrations of the remaining internalized S. Typhimurium in all of the chlorine or PAA-treated lettuce were significantly higher than those in the lettuce treated with high fluency of UV-C (Treatment 10–12) or the combined treatments (Treatment 13–16) (< 0·05).

Table 2. Efficiency of UV-C, chlorine and peracetic acid (PAA) in inactivating internalized S. Typhimurium in lettuce
TreatmentaLog reduction (log CFU g−1)bNumber of sample with internalized SalmonellacSurviving Salmonella (log CFU g−1)
  1. a

    Treatment was numbered as shown in Table 1.

  2. b

    Log reduction in each treatment was equal to the internalized Salmonella in the control group minus the remaining internalized Salmonella in the leaves after each treatment.

  3. c

    The total sample number in each treatment was 30, which is in the parentheses.

  4. d

    Significant reduction in the internalized S. Typhimurium was achieved (< 0·05).

  5. The number of log reduction is in bold format.

UntreatedN/A28 (30)2·76 ± 1·11
Chlorine treatment
10·43 ± 1·0826 (30)2·23 ± 1·08
20·88 ± 1·2421 (30)1·88 ± 1·24
30·92 ± 1·0021 (30)1·84 ± 1·00
40·99 ± 1·3820 (30)1·77 ± 1·38
PAA treatment
50·99 ± 1·3623 (30)2·29 ± 1·36
60·89 ± 1·4420 (30)1·88 ± 1·44
70·45 ± 1·3524 (30)2·32 ± 1·35
80·95 ± 1·4720 (30)1·81 ± 1·47
UV treatment
91·40 ± 1·0913 (30)1·37 ± 1·09
10 2·16 ± 0·97 d 9 (30)0·60 ± 0·97
11 2·28 ± 1·02 d 7 (30)0·48 ± 1·02
12 2·29 ± 1·08 d 5 (30)0·47 ± 1·08
Combined treatment
13 2·17 ± 1·03 d 7 (30)0·50 ± 1·03
14 2·40 ± 0·83 d 5 (30)0·36 ± 0·83
15 1·96 ± 1·18 d 9 (30)0·81 ± 1·18
16 2·52 ± 0·64 d 4 (30)0·24 ± 0·64
Table 3. Efficiency of UV-C, chlorine and peracetic acid (PAA) in inactivating internalized S. Typhimurium in green onion
TreatmentaLog reduction (log CFU g−1)bNumber of sample with internalized SalmonellacRemaining Salmonella (log CFU g−1)
  1. a

    Treatment was numbered as shown in Table 1.

  2. b

    Log reduction in each treatment was equal to the internalized Salmonella in the control group minus the remaining internalized Salmonella in the green onion after treatment.

  3. c

    The total sample number in each treatment was 30 which is in the parentheses.

  4. d

    Significant reduction in the internalized S. Typhimurium was achieved (< 0·05).

  5. The number of log reduction is in bold format.

UntreatedN/A30 (30)4·10 ± 1·12
Chlorine treatment
10·14 ± 0·4330 (30)3·96 ± 0·43
20·06 ± 0·4730 (30)4·04 ± 0·47
30·07 ± 0·8730 (30)4·03 ± 0·87
40·17 ± 0·9829 (30)3·93 ± 0·98
PAA treatment
50·15 ± 0·5530 (30)3·95 ± 0·55
60·22 ± 0·4330 (30)3·88 ± 0·43
70·54 ± 0·8930 (30)3·56 ± 0·89
80·63 ± 0·8829 (30)3·47 ± 0·88
UV treatment
90·05 ± 1·2628 (30)4·05 ± 1·26
10 1·45 ± 1·99 d 20 (30)2·65 ± 1·99
110·23 ± 1·2228 (30)3·87 ± 1·22
12 1·44 ± 1·83 d 14 (30)2·66 ± 1·83
Combined method treatment
13 1·48 ± 1·93 d 26 (30)2·62 ± 1·93
14 1·18 ± 1·72 d 23 (30)2·92 ± 1·72
15 1·00 ± 1·23 d 27 (30)3·10 ± 1·23
16 1·49 ± 1·69 d 22 (30)2·61 ± 1·69

In most green onions, the concentration of remaining internalized S. Typhimurium was still high (> 2·5 log CFU g−1) even after the treatments (Table 3). Chlorine treatments (Treatment 1–4) alone achieved 0·06 to 0·17 log reduction in the internalized Salmonella, and the remaining bacteria concentration was c. 4 log CFU g−1 of lettuce. In general, PAA treatments (Treatment 5–8) seem to achieve slightly higher reduction in the internalized Salmonella than chlorine, but their log reductions were not significantly different (> 0·05) and the differences of the remaining viable Salmonella were not significant between the two chemical treatment groups (> 0·05). Ninety-seven per cent of the green onions (29 of 30 samples) still contained the internalized Salmonella after the chlorine or PAA treatments. Two UV-C treatments (Treatment 10 and 12) significantly decreased the internalized bacteria (c. 1·5 log reduction) (< 0·05), while the other two (Treatment 9 and 11) did not. The four UV-C-chlorine/PAA combined treatments (Treatment 13–16) achieved 1–1·5 log reduction, and significantly decreased the internalized Salmonella in the green onion compared with either the untreated or the chlorine/PAA-treated groups (< 0·05). The remaining internalized S. Typhimurium after Treatment 13 (UV-C 450 mJ cm−2+ chlorine 1000 mg min l−1) and 15 (UV-C 450 mJcm−2 + PAA 400 mg min l−1) were significantly lower than the Salmonella levels after Treatment 11 (UV-C 450 mJ cm−2) (< 0·05).

Evaluation of food quality of the processed fresh produce

The colour parameters (L*, a*, b* and hue angle) of the lettuce and green onion before and after treatments are presented in Tables 4 and 5, respectively. The lightness of lettuce was 52·42 ± 2·43 before the treatments and ranged from 50·80 to 54·13 after disinfection. The hue angle of the lettuce was 121·73° before the treatments and varied from 120 to 123° after the treatments. No treatment was found to affect the lettuce lightness or hue angles significantly (> 0·05). Both of the chlorine and PAA treatments did not change the green colour (−a*) or the yellow colour (b*) of the lettuce (> 0·05). When the lettuce was processed with UV-C alone, the highest fluency (900 mJ cm−2) decreased green and yellow colour (< 0·05), while the other three UV-C treatments did not alter the colours (> 0·05). The minor colour change was also observed when using two of the combined treatments (Treatment 14 and 16). The green colour was decreased by 0·58 unit (< 0·05) when the lettuce was treated with UV-C (900 mJ cm−2) and chlorine (2000 mg min l−1) (Treatment 14), and both green and yellow colours were reduced when being treated with UV-C (900 mJ cm−2) and PAA (800 mg min l−1) (Treatment 16).

Table 4. Color measurements (L*a*b* and hue angle) of lettuce after disinfection treatments with UV-C, chlorine and PAA
LaaabaHue angle (°)
  1. a

    Treatment was numbered as shown in Table 1, and the sample size was 10 in each treatment.

Untreated52·42 ± 2·43−23·35 ± 0·4137·85 ± 1·00121·73 ± 0·99
Chlorine treatment
151·31 ± 1·61−22·66 ± 0·5236·45 ± 1·95122·69 ± 0·36
252·49 ± 2·19−23·26 ± 0·6937·97 ± 1·87121·55 ± 1·01
353·74 ± 1·20−23·16 ± 0·7037·81 ± 1·96121·55 ± 1·36
453·65 ± 1·26−23·05 ± 1·0738·23 ± 2·09121·14 ± 1·50
PAA treatment
554·13 ± 0·97−22·91 ± 0·8837·55 ± 1·93121·45 ± 1·10
653·42 ± 1·44−22·95 ± 2·3837·57 ± 1·57121·48 ± 1·09
753·79 ± 1·37−23·01 ± 0·5837·38 ± 0·67121·68 ± 0·33
852·78 ± 2·00−23·04 ± 0·8937·55 ± 2·66121·59 ± 1·16
UV treatment
953·72 ± 0·97−22·91 ± 0·9937·47 ± 2·22121·50 ± 0·95
1053·29 ± 2·05−22·93 ± 0·7437·11 ± 2·16121·77 ± 1·21
1152·73 ± 2·30−23·10 ± 0·7237·19 ± 2·36121·91 ± 1·11
1251·32 ± 1·61−22·66 ± 0·5236·45 ± 1·95121·93 ± 1·23
Combined method treatment
1353·59 ± 2·42−22·92 ± 0·8337·80 ± 1·97121·29 ± 1·32
1450·80 ± 1·98−22·77 ± 0·5136·81 ± 2·55121·80 ± 1·34
1553·70 ± 2·05−23·18 ± 0·6339·00 ± 2·59120·79 ± 0·75
1651·96 ± 2·46−22·72 ± 0·7235·93 ± 1·70122·37 ± 0·92
Table 5. Color parameters (L*a*b* and hue angle) of green onion after disinfection treatments with UV-C, chlorine and PAA
LaaabaHue angle (°)
  1. a

    Treatment was numbered as shown in Table 1, and the sample size was 10 in each treatment.

Untreated32·87 ± 1·09−12·01 ± 0·9911·70 ± 2·10135·81 ± 2·10
Chlorine treatment
132·74 ± 2·21−12·22 ± 1·0213·25 ± 0·96132·75 ± 1·97
232·42 ± 2·64−12·41 ± 1·5713·05 ± 1·97133·63 ± 1·84
331·49 ± 1·95−12·02 ± 1·6912·97 ± 2·51132·89 ± 3·65
432·09 ± 1·64−12·82 ± 1·3512·53 ± 1·39135·72 ± 4·11
PAA treatment
532·48 ± 2·08−12·09 ± 1·0212·48 ± 2·04134·16 ± 2·93
631·02 ± 2·09−12·04 ± 1·2712·82 ± 2·51133·27 ± 3·37
732·79 ± 1·29−12·15 ± 2·3113·32 ± 2·83132·43 ± 2·03
832·30 ± 2·20−12·34 ± 1·8712·56 ± 1·50134·56 ± 2·33
UV treatment
934·02 ± 2·29−13·29 ± 1·4013·94 ± 1·92133·70 ± 3·65
1031·65 ± 2·00−12·37 ± 1·9012·38 ± 1·52135·04 ± 2·40
1133·66 ± 1·99−13·37 ± 0·1014·40 ± 1·57132·94 ± 1·55
1232·50 ± 2·22−11·78 ± 1·4411·55 ± 2·37135·63 ± 2·56
Combined method treatment
1333·87 ± 1·85−12·86 ± 1·0013·14 ± 1·63134·45 ± 1·62
1432·21 ± 1·29−12·97 ± 1·2113·84 ± 2·33133·21 ± 2·37
1531·83 ± 1·48−12·09 ± 1·2012·19 ± 2·14134·83 ± 2·34
1632·24 ± 1·67−12·05 ± 0·9711·81 ± 1·90135·65 ± 2·46

The lightness of green onion was 32·87 before disinfection, and the measurements varied from 31 to 34 after being treated with different methods. Similar to the lettuce disinfection, none of the treatment significantly affected the green onion lightness (> 0·05). The green colour and hue angle were stable during the treatments including the UV-C radiations (> 0·05). The yellow colour of most green onion was not affected by the treatments either, and the only yellow colour change (2·14 unit increase, < 0·05) was found in the green onions which were processed with UV-C (900 mJ cm−2) combined with chlorine (2000 mg min l−1) (Treatment 14).

The firmness of the lettuce and green onion was 13·49 and 4·37 kg, respectively. As shown in Table 6, the texture (firmness) of the fresh produce was not significantly deteriorated by the UV-C radiation, chlorine/PAA, or the combined methods.

Table 6. Firmness of lettuce and green onion before and after treatment with UV-C, chlorine and peracetic acid (PAA)
TreatmentaFirmness (kg)
LettuceGreen onion
  1. a

    Treatment was numbered as shown in Table 1 and the sample size was 6 in each treatment.

Untreated13·49 ± 1·874·73 ± 1·18
Chlorine treatment
114·40 ± 0·974·67 ± 0·87
212·81 ± 1·474·19 ± 0·59
313·50 ± 1·454·84 ± 1·80
414·34 ± 1·564·87 ± 1·27
PAA treatment
513·43 ± 2·404·38 ± 0·73
612·73 ± 0·794·55 ± 0·79
713·69 ± 1·274·57 ± 0·63
813·97 ± 1·424·67 ± 0·29
UV treatment
913·31 ± 1·254·84 ± 1·80
1013·49 ± 2·214·87 ± 1·27
1113·52 ± 1·174·76 ± 0·72
1213·13 ± 1·264·61 ± 0·27
Combined method treatment
1314·44 ± 5·114·85 ± 0·61
1412·90 ± 0·864·41 ± 0·91
1513·90 ± 1·094·60 ± 0·58
1613·04 ± 0·704·80 ± 0·75


Ionizing radiation, such as gamma ray and electric beam, has been frequently examined for its potential of inactivating the internalized pathogens in fresh produce (Niemira 2008; Gomes et al. 2009; Chimbombi et al. 2011) and it was shown as an effective physical disinfectant to inactivate the human pathogens with low fluency in food without generating toxic hazards or making food radioactive, which has been confirmed by World Health Organization (Urbain 1986; Doyle 1999). Nevertheless, the public perception of the gamma ray-irradiated food has not been improved (Mahapatra et al. 2005). In addition, the cost of irradiation equipments and their maintenance, as well as the safety concerns during processing, make the ionizing radiation less competitive in its application in food industry (Durante 2002).

Compared with the ionizing radiation methods, the application of ultraviolet light has grown steadily in various fields such as water, wastewater, aerosol, surface and food as its germicidal effect was firstly described by Ward (1892). The gaining popularity of UV is attributed to its simplicity (the essential component of the UV processing is only the UV lamp) and cost-effectiveness. Most of the previous studies about UV-C application in food are associated with beverages and solid food surface using various UV fluency, food types and bacterial species (Guerrero-Beltrán and Barbosa-Cánovas 2004). Yaun et al. (2004) used UV-C radiation (1·5–24 mW cm−2) to inactivate Salmonella and E. coli O157:H7 on the surfaces of apple, tomato and lettuce. They concluded that 2–3 log reduction was achieved, and even the low UV-C fluency worked effectively in killing bacteria on the plant surfaces. After being treated with 410 mJ cm−2 of UV-C, a 3-log microbial reduction was obtained in the fresh-cut watermelon, and the count was c. 1 log lower in the UV-C irradiated plant compared with the control group, which exhibited better microbial inactivation efficiency than hydrogen peroxide (2%) or chlorine (c. 1200 ppm) (Fonseca and Rushing 2006).

Combination of UV-C with chemical sanitizers also has been investigated for food surface disinfection. UV-C (15 W lamp, fluency was not indicated) and ozone (1·6 ppm) washing inactivated over 3-log and 5-log of mesophilic bacteria on the surface of onion and escarole, respectively, with 20-min exposure time (Selma et al. 2008). Synergistic inactivation with sequential use of UV-C (90–150 mJ cm−2) and gaseous ozone (1·46 g m−2 for 1 min) was found to decrease 4·6 log of Salmonella Enteritidis on the surface of shell eggs (Rodriguez-Romo and Yousef 2005). A photocatalytic reaction of titanium dioxide (TiO2) under UV-C light showed higher bactericidal efficiency than using the UV-C alone. Surrounded by TiO2-coated quartz (0·7–0·9 μm thickness), UV-C (25 μW cm−2) removed 2·6 log CFU g−1 of E. coli, 2·5 log CFU g−1 of Listeria monocytogenes, 2·3 log CFU g−1 Staphylococcus aureus and 2·8 log CFU g−1 of S. Typhimurium on the surface of iceberg lettuce, respectively, in 20 min (Kim et al. 2009). Another study showed that the TiO2/UV-C treatment (16 mW cm−2 for 20 min) inactivated 2·1, 2·3, and 1·8 log CFU g−1 of E. coli, S. Typhimurium, and Bacillus cereus, respectively, on carrots (Cho et al. 2007).

It is known that UV-C light cannot penetrate deeply into the fresh produce (Morgan 1989). To our knowledge, only one study reported the use of UV-C to inactivate the internalized human pathogens in vegetables conducted by Hadjok et al. (2008) using UV-C with 50°C 1·5% H2O2 (continuous spraying or 10–30 s contacting time during the UV-C illumination at 25·2–56·7 mJ cm−2). Their results showed that UV-C (37·8 mJ cm−2) combined with 1·5% H2O2 continuous spraying at 50°C achieved a 2·84 log reduction in the internalized S. Montevideo in the iceberg lettuce, which was more efficient than using the UV-C or H2O2 alone. In this study, various UV treatments (higher UV-C fluencies, 75–900 mJ cm−2) were examined, UV-C alone or sequential treatment with chlorine/PAA, to test its disinfection efficiency of the internalized S. Typhimurium in iceberg lettuce and green onion at room temperature. In the lettuce processing, over 2-log reduction in the internalized S. Typhimurium was obtained when the UV-C fluency was 150 mJ cm−2 or higher, which was significantly better than using the chemicals alone. However, the synergistic effect was not observed in this study when compared the combined methods to using UV-C alone (Table 2). Unlike the lettuce, the radiation time in treating the green onion seemed more influential to the microbial reduction other than the UV intensity or the total irradiation fluency, because Treatment 10 (0·25 mW cm−2 for 10 min, irradiation fluency = 150 mJ cm−2) exhibited more effective than Treatment 11 (1·5 mW cm−2 for 5 min, irradiation fluency = 450 mJ cm−2) (Table 3). Generally, two UV-C treatments (Treatment 10 and 12) and all of the combined methods (Treatment 13–16) achieved 1–1·5 log Salmonella reduction, which was lower than the bacterial reduction achieved in the lettuce treatments, although the internalized Salmonella was significantly decreased in the treated green onion compared with the control group. None of the chlorine or PAA treatments were effective to remove the internalized bacteria. Considering the bacteria were injected into the green onion, more S. Typhimurium could have colonized deeper inside the green onion, which could have reduced the chance of bacterial exposure to the sanitizers, and therefore, the internalized bacteria were possibly less affected by either the UV-C light, chlorine or PAA. Adding chlorine/PAA washing after the UV-C treatments did not significantly enhance the inactivation efficiency when the radiation time was 10 min, but a synergistic effect was found when the chemical disinfection was used together with the UV treatment for 5 min (Treatment 11 vs Treatment 13 or 15).

The highest UV-C fluency (900 mJ cm−2) induced a slight loss of the green colour (Treatment 12, 14 and 16) and the yellow colour (Treatment 12 and 14) in the lettuce, and only one group of green onion showed a colour change (increase in the yellow colour after Treatment 14). Nevertheless, the colour was not apparently altered after UV-C irradiation, and the discoloration cannot be differentiated without using the professional equipments such as colorimeter (pictures are not shown). The firmness was also determined, and no texture loss was detected in any treatment. The UV-C treatments have little influence on food quality and did not cause unacceptable quality loss, which has been supported by other studies. For example, after being treated with UV-C (three fluencies from 454 to 1135 mJ cm−2), the chlorophyll content of the prepackaging spinach was not affected, and only a slight lightness change was found during a 13-day-shelf life (Artés-Hernández et al. 2009). No discoloration was observed on the UV-C (37·8 mJ cm−2)-H2O2 (1·5%) treated lettuce even after 6-day storage at 4°C or 25°C (Hadjok et al. 2008).

In conclusion, washing with chlorine or PAA alone was not effective to inactivate the internalized Salmonella in the lettuce and green onion. UV-C irradiation with higher fluencies (150, 450, 900 mJ cm−2) can significantly reduce the internalized S. Typhimurium in the iceberg lettuce. However, the mechanism of UV-C against internalized bacteria in the plant has not been clearly revealed and needs further exploration. Additional treatment of chlorine or PAA washing did not further enhance the reduction in the internalized Salmonella in the lettuce after the UV-C irradiation, and thus, it seems there is no additional benefit from adding the chemical sanitizers. However, for the green onion, combination of UV-C and chemical sanitizers was more effective in reducing the internalized Salmonella when the irradiation time was short (5 min). The colour of most treated green onions was not affected, and lettuce colour was slightly changed when processed with the highest intensity of UV-C (900 mJ cm−2), but unacceptable discoloration was not observed. The texture of all the fresh produce samples was well maintained during the processing.


This study was partially supported by funds provided by the Public Health Preparedness for Infectious Diseases (PHPID) programme ( of The Ohio State University.