Microbubbles Remove Listeria monocytogenes from the Surface of Stainless Steel, Cucumber, and Avocado

Fresh produce may be contaminated by bacterial pathogens, including Listeria monocytogenes, during harvesting, packaging, or transporting. A low-intensity cavitation process with air being injected into water was studied to determine the microbubbles’ efficiency when detaching L. monocytogenes from stainless steel and the surface of fresh cucumber and avocado. Stainless steel coupons (1″ × 2″), cucumber, and avocado surfaces were inoculated with L. monocytogenes (LCDC strain). After 1, 24 or 48 h, loosely attached cells were washed off, and inoculated areas were targeted by microbubbles (~0.1–0.5 mm dia.) through a bubble diffuser (1.0 L air/min) for 1, 2, 5, or 10 min. For steel, L. monocytogenes (48 h drying) detachment peaked at 2.95 mean log reduction after 10 min of microbubbles when compared to a no-bubble treatment. After 48 h pathogen drying, cucumbers treated for 10 min showed a 1.78 mean log reduction of L. monocytogenes. For avocados, L. monocytogenes (24 h drying) detachment peaked at 1.65 log reduction after 10 min of microbubbles. Microbubble applications may be an effective, economical, and environmentally friendly way to remove L. monocytogenes, and possibly other bacterial pathogens, from food contact surfaces and the surfaces of whole, intact fresh produce.

L. monocytogenes (strain LCDC, serotype 4b) was transferred from frozen (−80 • C) storage to Tryptic Soy Broth (TSB) and incubated at 35 ± 2 • C for 24 h. Then, the culture was transferred to fresh TSB and incubated at 35 ± 2 • C for 24 h. This culture was subsequently used to inoculate stainless steel, cucumbers and avocados. Culture identity and purity was confirmed by colony appearance on Rapid L'mono agar (Bio-Rad Laboratories, Richmond, CA, USA) and with the Microgen Listeria-ID biochemical identification kit (Microbiology International, Frederick, MD, USA). Additionally, this culture was confirmed to be able to adhere to the test materials and could be recovered 24 and 48 h after inoculation at higher concentrations than three other test strains of L. monocytogenes.
Stainless-steel sheet (type 304, #2 Finish; Speedy Metals, Milwaukee, WI, USA) was cut into 1 × 2-inch coupons. Slicing cucumbers (Cucumis sativas) and avocados (Persea americana cultivar Hass) were purchased from local supermarkets. The degree of maturity for all cucumbers and all avocados was likely similar, since growers and packers had determined they were suitable for distribution. These markets generally sell these produce items within three days after they are made available to the consumer. The Hass avocado is the most common cultivar of commercial avocados. Before the bacterial inoculation onto cucumbers, wax was removed from the cucumber surfaces by washing them with peptone water and drying them gently with paper wipes. Additionally, cucumbers and avocados were sprayed with 70% ethanol and then wiped and air-dried to reduce background microflora.
Stainless-steel coupons were spot inoculated with 0.1 mL of a L. monocytogenes culture (~2 × 10 8 CFU/mL). The spots of inoculation were marked before inoculation. Inoculums were allowed to dry for 1, 24, or 48 h at room temperature (22 • C) in a sealed box with two small beakers of a sodium chloride solution to maintain the humidity from 50% to 60%. Cucumbers and avocados were also spot-inoculated with 0.1 mL of a L. monocytogenes culture (~2 × 10 8 CFU/mL) and stored at 10 • C. Inoculums were allowed to dry for 1, 24, or 48 h.

Bubble Treatments for Stainless Steel
An air compressor (Campbell Hausfeld, Cincinnati, OH, USA, 120 V, 60 Hz, 2 A, 1 gallon tank, maximum rated 110 psi) was used to deliver air through a bubble diffuser (Model MBD75 (Diffuser Area 6-1/8" × 1-1/8", Range up to 0.75 LPM, Flow Rate @ 50 PSI 2.2 LPM, Gas Inlet Connection 1/4" Hose Barb); Pentair, Golden Hills, MN, USA) submerged in the bottom of a water tank. The manufacturer states that this diffuser produces bubbles with an approximately 100-500 micron diameter in water, hereafter referred to as microbubbles. A Masterflex flow meter (Cole-Parmer, Vernon Hills, IL, USA) was used to regulate air flow to 1.0 L/min.
After stainless-steel coupons were dried (1, 24, or 48 h), but before the application of bubble treatments, the inoculated steel coupons were rinsed with 10 mL peptone water to remove loosely attached cells. These rinsates were collected and further analyzed as described in Section 2.4. Stainless steel coupons were individually placed into a 13 L plastic tank containing 4.5 L distilled water. Plastic forceps were used to hold the rectangle stainless-steel coupon when applying bubbles to the inoculated surface.
Stainless-steel coupons were treated for 0, 1, 2, 5 or 10 min using the bubble diffuser with a 1.0 L/min air flow rate. Three steel samples were tested for each combination of inoculum drying time and bubble treatment time. After treatment with a stream of heavy bubbles, stainless-steel coupons were collected in a sterile bag with peptone water and sonicated (Aquasonic Ultrasonic Cleaner, Volts: 117/120, 50/60 Hz) for 2 min at room temperature. Solutions were then diluted with peptone water and plated onto Oxford agar (Neogen, Lansing, MI, USA) plates. Plates were incubated at 35 ± 2 • C for 48 h. Three replications of the above process were conducted for a total of 108 samples.

Bubble Treatment of Cucumber and Avocado Surfaces
After inoculated cucumbers and avocados were dried (1, 24, or 48 h), but before the application of bubble treatments, they were rinsed with 10 mL peptone water to remove loosely attached cells. These rinses were collected and further analyzed as described in Section 2.4. Rinsed cucumbers and avocados were individually placed into a 13 L plastic tank containing 4.5 L distilled water. Two elastic rubber wires on the top the opposite edges of the tank were used to hold the cucumbers in place over the bubble stream. A compressed air supply was used to deliver 1.0 L/min of air through a bubble diffuser as described in Section 2.2. A stream of bubbles (~0.1-0.5 mm diameter) was applied for either 0, 1, 2, 5 or 10 min. Three cucumbers and three avocados were tested for each combination of inoculum drying time (3) and bubble treatment time (5) for each of the three replications.

Recovery of L. monocytogenes from Inoculated Surfaces
Before the application of bubble treatments, inoculated steel coupons, cucumbers and avocados were rinsed with 10 mL distilled water to remove loosely attached cells. The rinses were collected, diluted with peptone water and plated onto Oxford agar plates. Plates were incubated at 35 ± 2 • C for 48 h and enumerated. Then, the rinsed test materials were treated with a bubble application for 0, 1, 2, 5, or 10 min. Firmly attached cells were removed from the steel coupons, cucumbers and avocados by placing the sample in a sterile bag or cup with added peptone water. Containers were then sonicated for 2 min at room temperature (22 • C). Solutions were then diluted and plated onto Oxford agar plates and incubated at 35 ± 2 • C for 48 h. Finally, the cell counts of the solutions after sonication of bubble-treated samples were compared to the cell counts of the solutions after sonication where samples were not treated with bubbles (0 min treatment).

Cross-Contamination from Inoculated to Un-Inoculated Produce
To simulate the cross-contamination of produce with L. monocytogenes, one inoculated cucumber and one un-inoculated cucumber were added to a water tank. All cucumbers were wiped to remove wax and reduce background microflora, as described previously. Inoculated cucumbers received 0.1 mL of a L. monocytogenes (LCDC) culture and allowed to dry for 1, 24 or 48 h before they were placed into a water tank with one un-inoculated cucumber. A heavy bubble stream from the bubble diffuser was applied for 2 or 10 min with an airflow rate of 1.0 L/min. As a control, pairs of inoculated and uninoculated cucumbers were placed in the water tank for 1 min, but without a bubble application. Then, each sample cucumber was placed into separate sterile bags, and L. monocytogenes were recovered from the sample surface, as described previously. Oxford agar plates were incubated at 35 ± 2 • C for 48 h. A similar inoculation, treatment and recovery protocol was carried out for avocados. Three sample pairs of cucumbers and avocados were tested for each combination of inoculum drying time and bubble treatment time. These experiments were replicated three times for both cucumbers and avocados. The proportion of cells that were transferred to the uninoculated samples was determined.

Microbial Quality Evaluation of Treated Cucumbers and Avocados after Refrigerated Storage (10 • C)
Fresh cucumbers and avocados were stored in a refrigerator at 10 • C for up to 7 d. After storage for 0, 1, 4, or 7 d, three cucumbers and three avocados were individually placed in a sterile bag, diluted with peptone water, and solutions were plated onto Tryptic Soy Agar plates. Additional cucumbers and avocados were treated to a heavy bubble stream, as described previously, for 2 or 10 min. These samples were stored for either 0, 1, 4 or 7 d at 10 • C. After storage, three cucumbers and three avocados (from each storage time) were individually placed in a sterile bag, diluted with peptone water, and solutions were plated onto Tryptic Soy Agar plates. All plates were incubated at 35 ± 2 • C for 24 h, then enumerated. The above process was repeated for 3 replications.

Data Analysis
L. monocytogenes cell counts recovered from fresh produce and stainless-steel coupons were compared to the cell counts of original inoculation on produce and stainless steel to determine the pathogen detachment from the surface of fresh produce and stainless steel after microbubble treatment. A one-way analysis of variance (ANOVA) was used to determine significant differences between means for each variable tested at a statistical significance of α = 0.05. All calculations were performed with JMP 16.1 Statistical Software (SAS Institute, Inc., Cary, NC, USA).

Bubble Treatments of Stainless-Steel Coupons
After rinsing off loosely attached cells, approximately 6.65, 6.30, and 6.08 log mean CFU of L. monocytogenes per stainless steel coupon could be recovered after 1-, 24-, or 48-h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the total mean CFU of L. monocytogenes recovered from stainless steel decreased for each set of coupons with different drying times (1, 24, or 48 h). Thus, an extended bubble treatment time appears to remove more bacteria, since less could be recovered from the steel coupons. Figure 1 shows the mean log CFU L. monocytogenes recovered per steel coupon after different bubble treatment times, ranging from 0 to 10 min. After inoculated steel was dried for 48 h, the log-CFU-recovered difference (from the no-bubble treatment) was highest after 10 min treatments (2.95 log CFU), and 1.55 log CFU was highest after a 1 min bubble treatment. Additionally, the mean log CFU difference recovered from stainless steel with 1 h drying time was highest after 10 min treatments (2.08 log CFU) and lowest after 1 min treatments (0.31 log CFU). The effect of bubble treatment time was significant for each of the three steel inoculum drying times (p < 0.01).

Cavitation Treatment of Cucumber Surfaces
After rinsing off loosely attached cells, approximately 7.15, 6.48, and 6.79 log mean CFU of L. monocytogenes per cucumber could be recovered after a 1, 24, or 48 h drying time, respectively. Figure 2 shows the mean log CFU L. monocytogenes recovered per cucumber after rinsing off loosely attached cells, followed by a bubble treatment from 0 to 10 min. The difference in Listeria recovery between no-bubble treatments (0 min) and 10 min treatments was 1.09, 1.39 or 1.78 mean log CFU per cucumber for cucumbers dried

Cavitation Treatment of Cucumber Surfaces
After rinsing off loosely attached cells, approximately 7.15, 6.48, and 6.79 log mean CFU of L. monocytogenes per cucumber could be recovered after a 1, 24, or 48 h drying time, respectively. Figure 2 shows the mean log CFU L. monocytogenes recovered per cucumber after rinsing off loosely attached cells, followed by a bubble treatment from 0 to 10 min. The difference in Listeria recovery between no-bubble treatments (0 min) and 10 min treatments was 1.09, 1.39 or 1.78 mean log CFU per cucumber for cucumbers dried

Cavitation Treatment of Cucumber Surfaces
After rinsing off loosely attached cells, approximately 7.15, 6.48, and 6.79 log mean CFU of L. monocytogenes per cucumber could be recovered after a 1, 24, or 48 h drying time, respectively. Figure 2 shows the mean log CFU L. monocytogenes recovered per cucumber after rinsing off loosely attached cells, followed by a bubble treatment from 0 to 10 min. The difference in Listeria recovery between no-bubble treatments (0 min) and 10 min treatments was 1.09, 1.39 or 1.78 mean log CFU per cucumber for cucumbers dried

Cavitation Treatment of Cucumber Surfaces
After rinsing off loosely attached cells, approximately 7.15, 6.48, and 6.79 log mean CFU of L. monocytogenes per cucumber could be recovered after a 1, 24, or 48 h drying time, respectively. Figure 2 shows the mean log CFU L. monocytogenes recovered per cucumber after rinsing off loosely attached cells, followed by a bubble treatment from 0 to 10 min. The difference in Listeria recovery between no-bubble treatments (0 min) and 10 min treatments was 1.09, 1.39 or 1.78 mean log CFU per cucumber for cucumbers dried , and different bubble treatment times from 0 to 10 min (Mean of: 3 replications × 3 samples each). A,B,C,D Denote significant differences in recovery (p < 0.01) between treatment times for each of the inoculum drying times.

Bubble Treatment of Cucumber Surfaces
After rinsing off loosely attached cells, approximately 7.15, 6.48, and 6.79 log mean CFU of L. monocytogenes per cucumber could be recovered after a 1, 24, or 48 h drying time, respectively. Figure 2 shows the mean log CFU L. monocytogenes recovered per cucumber after rinsing off loosely attached cells, followed by a bubble treatment from 0 to 10 min. The difference in Listeria recovery between no-bubble treatments (0 min) and 10 min treatments was 1.09, 1.39 or 1.78 mean log CFU per cucumber for cucumbers dried for 1, 24 or 48 h, respectively. For each of the drying times, extending the bubble treatment times by up to 10 min appears to remove more bacteria, since bacterial recovery from treated cucumbers was reduced. The mean difference in recovery with a 1 min bubble treatment was 1.02 log CFU for cucumbers, where inocula was dried for 48 h, but only 0.24 log CFU when inoculated cucumbers were dried for 24 h. The effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When cucumbers were dried for 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different (lowered) between each of the bubble treatment times (p < 0.01).

Cavitation Treatment of Avocado Surfaces
After rinsing off loosely attached cells, approximately 8.13, 7.52, and 7.79 log mean CFU of L. monocytogenes per avocado could be recovered after 1, 24, or 48 h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the mean CFU of L. monocytogenes recovered from avocados decreased for each set of fruit with different drying times (1, 24, or 48 h). Therefore, extended bubble treatment times may lead to greater bacterial removal or detachment. Figure 3 shows the mean log CFU L. monocytogenes recovered per avocado after different bubble treatment times, ranging from 0 to 10 min. After a 1 min treatment, the mean difference (from 0 min treatment) log CFU recovered from avocado with a 24 h drying time (1.08 log CFU) was greater than that for the avocados that were dried for 1 or 48 h. The maximum difference from a no-bubble treatment occurred with a 10 min treatment for avocadoes where the inoculum was dried for 24 h (mean 1.65 log CFU per avocado). Similar to cucumbers, the effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When avocados were dried for 1, 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different between most of the bubble

Cavitation Treatment of Avocado Surfaces
After rinsing off loosely attached cells, approximately 8.13, 7.52, and 7.79 log mean CFU of L. monocytogenes per avocado could be recovered after 1, 24, or 48 h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the mean CFU of L. monocytogenes recovered from avocados decreased for each set of fruit with different drying times (1, 24, or 48 h). Therefore, extended bubble treatment times may lead to greater bacterial removal or detachment. Figure 3 shows the mean log CFU L. monocytogenes recovered per avocado after different bubble treatment times, ranging from 0 to 10 min. After a 1 min treatment, the mean difference (from 0 min treatment) log CFU recovered from avocado with a 24 h drying time (1.08 log CFU) was greater than that for the avocados that were dried for 1 or 48 h. The maximum difference from a no-bubble treatment occurred with a 10 min treatment for avocadoes where the inoculum was dried for 24 h (mean 1.65 log CFU per avocado). Similar to cucumbers, the effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When avocados were dried for 1, 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different between most of the bubble

Cavitation Treatment of Avocado Surfaces
After rinsing off loosely attached cells, approximately 8.13, 7.52, and 7.79 log mean CFU of L. monocytogenes per avocado could be recovered after 1, 24, or 48 h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the mean CFU of L. monocytogenes recovered from avocados decreased for each set of fruit with different drying times (1, 24, or 48 h). Therefore, extended bubble treatment times may lead to greater bacterial removal or detachment. Figure 3 shows the mean log CFU L. monocytogenes recovered per avocado after different bubble treatment times, ranging from 0 to 10 min. After a 1 min treatment, the mean difference (from 0 min treatment) log CFU recovered from avocado with a 24 h drying time (1.08 log CFU) was greater than that for the avocados that were dried for 1 or 48 h. The maximum difference from a no-bubble treatment occurred with a 10 min treatment for avocadoes where the inoculum was dried for 24 h (mean 1.65 log CFU per avocado). Similar to cucumbers, the effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When avocados were dried for 1, 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different between most of the bubble

Cavitation Treatment of Avocado Surfaces
After rinsing off loosely attached cells, approximately 8.13, 7.52, and 7.79 log mean CFU of L. monocytogenes per avocado could be recovered after 1, 24, or 48 h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the mean CFU of L. monocytogenes recovered from avocados decreased for each set of fruit with different drying times (1, 24, or 48 h). Therefore, extended bubble treatment times may lead to greater bacterial removal or detachment. Figure 3 shows the mean log CFU L. monocytogenes recovered per avocado after different bubble treatment times, ranging from 0 to 10 min. After a 1 min treatment, the mean difference (from 0 min treatment) log CFU recovered from avocado with a 24 h drying time (1.08 log CFU) was greater than that for the avocados that were dried for 1 or 48 h. The maximum difference from a no-bubble treatment occurred with a 10 min treatment for avocadoes where the inoculum was dried for 24 h (mean 1.65 log CFU per avocado). Similar to cucumbers, the effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When avocados were dried for 1, 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different between most of the bubble , and different bubble treatment times from 0 to 10 min (Mean of 3 replications × 3 samples each). A,B,C,D,E Denote significant differences in recovery (p < 0.01) between treatment times for each of the inoculum drying times.

Bubble Treatment of Avocado Surfaces
After rinsing off loosely attached cells, approximately 8.13, 7.52, and 7.79 log mean CFU of L. monocytogenes per avocado could be recovered after 1, 24, or 48 h drying time, respectively. As the bubble treatment time increased from 0 to 1, 2, 5, or 10 min, the mean CFU of L. monocytogenes recovered from avocados decreased for each set of fruit with different drying times (1, 24, or 48 h). Therefore, extended bubble treatment times may lead to greater bacterial removal or detachment. Figure 3 shows the mean log CFU L. monocytogenes recovered per avocado after different bubble treatment times, ranging from 0 to 10 min. After a 1 min treatment, the mean difference (from 0 min treatment) log CFU recovered from avocado with a 24 h drying time (1.08 log CFU) was greater than that for the avocados that were dried for 1 or 48 h. The maximum difference from a no-bubble treatment occurred with a 10 min treatment for avocadoes where the inoculum was dried for 24 h (mean 1.65 log CFU per avocado). Similar to cucumbers, the effect of bubble treatment time was significant for each of the cucumber drying times (p < 0.01). When avocados were dried for 1, 24 or 48 h, the bacterial recovery of firmly attached cells was significantly different between most of the bubble treatment times (p < 0.01).  Table 1 shows the percent transfer of viable L. monocytogenes cells from cross-contaminated cucumbers with different bubble treatment times and the three different inocula drying times. The percent transfer from contaminated cucumbers to uninoculated cucumbers increased when bubble treatment time increased. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes (0.203% and 0.548%) between cucumbers was the highest from inoculated cucumbers dried for 24 h compared to those dried for 1 or 48 h. When pairs of cucumbers were placed in water for 1 min and without a bubble treatment, no more than 0.002% of the recovered L. monocytogenes was transferred to uninoculated cucumbers. Table 1 also shows that the percent transfer from cross-contaminated avocados increases when bubble treatment time increases, but at a lower level than seen with cucumbers. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes between avocados (0.017% and 0.115%, respectively) was highest from inoculated avocados dried for 24 h than for 1 or 48 h. When pairs of avocados were placed in water for 1 min without a bubble stream, no more than 0.001% of the recovered L. monocytogenes was transferred to uninoculated avocados.  Table 1 shows the percent transfer of viable L. monocytogenes cells from cross-contaminated cucumbers with different bubble treatment times and the three different inocula drying times. The percent transfer from contaminated cucumbers to uninoculated cucumbers increased when bubble treatment time increased. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes (0.203% and 0.548%) between cucumbers was the highest from inoculated cucumbers dried for 24 h compared to those dried for 1 or 48 h. When pairs of cucumbers were placed in water for 1 min and without a bubble treatment, no more than 0.002% of the recovered L. monocytogenes was transferred to uninoculated cucumbers. Table 1 also shows that the percent transfer from cross-contaminated avocados increases when bubble treatment time increases, but at a lower level than seen with cucumbers. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes between avocados (0.017% and 0.115%, respectively) was highest from inoculated avocados dried for 24 h than for 1 or 48 h. When pairs of avocados were placed in water for 1 min without a bubble stream, no more than 0.001% of the recovered L. monocytogenes was transferred to uninoculated avocados.  Table 1 shows the percent transfer of viable L. monocytogenes cells from cross-contaminated cucumbers with different bubble treatment times and the three different inocula drying times. The percent transfer from contaminated cucumbers to uninoculated cucumbers increased when bubble treatment time increased. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes (0.203% and 0.548%) between cucumbers was the highest from inoculated cucumbers dried for 24 h compared to those dried for 1 or 48 h. When pairs of cucumbers were placed in water for 1 min and without a bubble treatment, no more than 0.002% of the recovered L. monocytogenes was transferred to uninoculated cucumbers. Table 1 also shows that the percent transfer from cross-contaminated avocados increases when bubble treatment time increases, but at a lower level than seen with cucumbers. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes between avocados (0.017% and 0.115%, respectively) was highest from inoculated avocados dried for 24 h than for 1 or 48 h. When pairs of avocados were placed in water for 1 min without a bubble stream, no more than 0.001% of the recovered L. monocytogenes was transferred to uninoculated avocados.  Table 1 shows the percent transfer of viable L. monocytogenes cells from cross-contaminated cucumbers with different bubble treatment times and the three different inocula drying times. The percent transfer from contaminated cucumbers to uninoculated cucumbers increased when bubble treatment time increased. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes (0.203% and 0.548%) between cucumbers was the highest from inoculated cucumbers dried for 24 h compared to those dried for 1 or 48 h. When pairs of cucumbers were placed in water for 1 min and without a bubble treatment, no more than 0.002% of the recovered L. monocytogenes was transferred to uninoculated cucumbers. Table 1 also shows that the percent transfer from cross-contaminated avocados increases when bubble treatment time increases, but at a lower level than seen with cucumbers. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes between avocados (0.017% and 0.115%, respectively) was highest from inoculated avocados dried for 24 h than for 1 or 48 h. When pairs of avocados were placed in water for 1 min without a bubble stream, no more than 0.001% of the recovered L. monocytogenes was transferred to uninoculated avocados.

Avocado
, and different bubble treatment times from 0 to 10 min (mean of 3 replications × 3 samples each). A,B,C,D,E Denote significant differences in recovery (p < 0.01) between treatment times for each of the inoculum drying times. Table 1 shows the percent transfer of viable L. monocytogenes cells from cross-contaminated cucumbers with different bubble treatment times and the three different inocula drying times. The percent transfer from contaminated cucumbers to uninoculated cucumbers increased when bubble treatment time increased. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes (0.203% and 0.548%) between cucumbers was the highest from inoculated cucumbers dried for 24 h compared to those dried for 1 or 48 h. When pairs of cucumbers were placed in water for 1 min and without a bubble treatment, no more than 0.002% of the recovered L. monocytogenes was transferred to uninoculated cucumbers. Table 1 also shows that the percent transfer from cross-contaminated avocados increases when bubble treatment time increases, but at a lower level than seen with cucumbers. For 2 and 10 min bubble treatments, the percent transfer of L. monocytogenes between avocados (0.017% and 0.115%, respectively) was highest from inoculated avocados dried for 24 h than for 1 or 48 h. When pairs of avocados were placed in water for 1 min without a bubble stream, no more than 0.001% of the recovered L. monocytogenes was transferred to uninoculated avocados.  Figure 4 displays the log mean CFU per cucumber recovered after different storage times after different bubble treatment times. When no bubbles were applied (0 min) and for a 2 min bubble treatment time, the log mean CFU per cucumber increased from day 0 to day 7. For 10 min bubble treatment, log mean CFU per cucumber increased from day 0 to storage day 4, then decreased slightly by storage day 7. Overall, the log mean CFU per cucumber after 0, 4 and 7 d storage was the lowest when bubbles were applied for 10 min.

Quality Evaluation of Produce during Refrigerated Storage
3.5.1. Cucumber Figure 4 displays the log mean CFU per cucumber recovered after different storage times after different bubble treatment times. When no bubbles were applied (0 min) and for a 2 min bubble treatment time, the log mean CFU per cucumber increased from day 0 to day 7. For 10 min bubble treatment, log mean CFU per cucumber increased from day 0 to storage day 4, then decreased slightly by storage day 7. Overall, the log mean CFU per cucumber after 0, 4 and 7 d storage was the lowest when bubbles were applied for 10 min.

Avocado
As seen in Figure 5, the log means CFU per avocado increased during storage from day 0 to day 7 when avocados received no bubble treatment or when they received a 10min bubble treatment. For the 2 min' bubble treatment, the log mean CFU per avocado decreased by approximately 1 log after 7 d storage, yet almost no decrease was observed after 4 d storage. The mean log CFU recovered immediately after a 2 min bubble treatment   Figure 4 displays the log mean CFU per cucumber recovered after different storage times after different bubble treatment times. When no bubbles were applied (0 min) and for a 2 min bubble treatment time, the log mean CFU per cucumber increased from day 0 to day 7. For 10 min bubble treatment, log mean CFU per cucumber increased from day 0 to storage day 4, then decreased slightly by storage day 7. Overall, the log mean CFU per cucumber after 0, 4 and 7 d storage was the lowest when bubbles were applied for 10 min.

Avocado
As seen in Figure 5, the log means CFU per avocado increased during storage from day 0 to day 7 when avocados received no bubble treatment or when they received a 10min bubble treatment. For the 2 min' bubble treatment, the log mean CFU per avocado decreased by approximately 1 log after 7 d storage, yet almost no decrease was observed after 4 d storage. The mean log CFU recovered immediately after a 2 min bubble treatment   Figure 4 displays the log mean CFU per cucumber recovered after different storage times after different bubble treatment times. When no bubbles were applied (0 min) and for a 2 min bubble treatment time, the log mean CFU per cucumber increased from day 0 to day 7. For 10 min bubble treatment, log mean CFU per cucumber increased from day 0 to storage day 4, then decreased slightly by storage day 7. Overall, the log mean CFU per cucumber after 0, 4 and 7 d storage was the lowest when bubbles were applied for 10 min.

Avocado
As seen in Figure 5, the log means CFU per avocado increased during storage from day 0 to day 7 when avocados received no bubble treatment or when they received a 10min bubble treatment. For the 2 min' bubble treatment, the log mean CFU per avocado decreased by approximately 1 log after 7 d storage, yet almost no decrease was observed after 4 d storage. The mean log CFU recovered immediately after a 2 min bubble treatment   Figure 4 displays the log mean CFU per cucumber recovered after different storage times after different bubble treatment times. When no bubbles were applied (0 min) and for a 2 min bubble treatment time, the log mean CFU per cucumber increased from day 0 to day 7. For 10 min bubble treatment, log mean CFU per cucumber increased from day 0 to storage day 4, then decreased slightly by storage day 7. Overall, the log mean CFU per cucumber after 0, 4 and 7 d storage was the lowest when bubbles were applied for 10 min.

Avocado
As seen in Figure 5, the log means CFU per avocado increased during storage from day 0 to day 7 when avocados received no bubble treatment or when they received a 10min bubble treatment. For the 2 min' bubble treatment, the log mean CFU per avocado decreased by approximately 1 log after 7 d storage, yet almost no decrease was observed after 4 d storage. The mean log CFU recovered immediately after a 2 min bubble treatment bubble stream. (mean of 3 replications × 3 samples each). A,B,C Denote significant differences in recovery (p < 0.01) between storage times for each of the bubble treatment times.

Avocado
As seen in Figure 5, the log means CFU per avocado increased during storage from day 0 to day 7 when avocados received no bubble treatment or when they received a 10-min bubble treatment. For the 2 min' bubble treatment, the log mean CFU per avocado decreased by approximately 1 log after 7 d storage, yet almost no decrease was observed after 4 d storage. The mean log CFU recovered immediately after a 2 min bubble treatment (log 4.96 for day 0), which was higher than the initial count recovered for the no-bubble treatment or 10 min treatment (4.45 or 3.85 log CFU per avocado, respectively).
(log 4.96 for day 0), which was higher than the initial count recovered for the no-bubble treatment or 10 min treatment (4.45 or 3.85 log CFU per avocado, respectively).

Discussion
Based on this study, a stream of targeted by microbubbles (<0.5 mm dia.) from a bubble diffuser (1.0 L air/min) can detach cells of L. monocytogenes from steel and produce surfaces. Extended bubble treatment times (up to 10 min) promoted greater removal of L. monocytogenes from stainless steel coupons, since significantly fewer L. monocytogenes were recovered from stainless steel after 10 min bubble treatments compared with no bubble treatment. The bubble application used in this study may reduce L. monocytogenes on stainless by nearly three logs (2.95 log CFU per sample) after an inoculum was dried for 48 h and loosely attached cells were removed. For cucumber and avocado, a 5 min application of the same heavy bubble stream removed at least 1.0 log more of the firmly attached bacteria than submerging the produce in water without cavitation bubbles. Lee et al. (2018) applied cavitating bubbles to inoculated fresh Roma tomatoes and cantaloupes using larger bubble sizes (~1-3 mm diameter) and higher air flows (3.5-14 L/min) than were used in this study [12]. Those authors reported that L. monocytogenes significantly decreased on Roma tomatoes (~1.0 log CFU after 30 sec and ~1.2 log CFU after 60 sec treatment) with a 14 L/min airflow rate compared to no cavitating bubbles. L. monocytogenes significantly decreased on cantaloupes (0.65 log CFU for 30 sec and ~0.74 log CFU for 60 sec treatment) when the maximum air-flow rate was used. In the study reported here, the log reduction in L. monocytogenes on cucumber and avocado surfaces also increased with increasing bubble treatment time, even though the airflow rate was lower, the treatment times longer, and the bubble diameters smaller. An important difference in these studies is that Lee et al. (2018) did not remove loosely attached cells prior to any bubble treatments. Here, in our study with cucumbers, avocados and stainless-steel coupons, the number of loosely attached cells recovered and enumerated after 1, 24 or 48 h drying was always higher than the quantity that could be recovered from the samples that were not treated with bubbles (treatment time = 0 min).
Chlorine is commonly applied to wash and clean fresh produce due to its bactericidal capacity and economical cost. When chlorine makes contact with wash water during washing, it yields an oxidizer, HOCI, which can inactivate pathogens effectively [16]. The (log 4.96 for day 0), which was higher than the initial count recovered for the no-bubble treatment or 10 min treatment (4.45 or 3.85 log CFU per avocado, respectively).

Discussion
Based on this study, a stream of targeted by microbubbles (<0.5 mm dia.) from a bubble diffuser (1.0 L air/min) can detach cells of L. monocytogenes from steel and produce surfaces. Extended bubble treatment times (up to 10 min) promoted greater removal of L. monocytogenes from stainless steel coupons, since significantly fewer L. monocytogenes were recovered from stainless steel after 10 min bubble treatments compared with no bubble treatment. The bubble application used in this study may reduce L. monocytogenes on stainless by nearly three logs (2.95 log CFU per sample) after an inoculum was dried for 48 h and loosely attached cells were removed. For cucumber and avocado, a 5 min application of the same heavy bubble stream removed at least 1.0 log more of the firmly attached bacteria than submerging the produce in water without cavitation bubbles. Lee et al. (2018) applied cavitating bubbles to inoculated fresh Roma tomatoes and cantaloupes using larger bubble sizes (~1-3 mm diameter) and higher air flows (3.5-14 L/min) than were used in this study [12]. Those authors reported that L. monocytogenes significantly decreased on Roma tomatoes (~1.0 log CFU after 30 sec and ~1.2 log CFU after 60 sec treatment) with a 14 L/min airflow rate compared to no cavitating bubbles. L. monocytogenes significantly decreased on cantaloupes (0.65 log CFU for 30 sec and ~0.74 log CFU for 60 sec treatment) when the maximum air-flow rate was used. In the study reported here, the log reduction in L. monocytogenes on cucumber and avocado surfaces also increased with increasing bubble treatment time, even though the airflow rate was lower, the treatment times longer, and the bubble diameters smaller. An important difference in these studies is that Lee et al. (2018) did not remove loosely attached cells prior to any bubble treatments. Here, in our study with cucumbers, avocados and stainless-steel coupons, the number of loosely attached cells recovered and enumerated after 1, 24 or 48 h drying was always higher than the quantity that could be recovered from the samples that were not treated with bubbles (treatment time = 0 min).
Chlorine is commonly applied to wash and clean fresh produce due to its bactericidal capacity and economical cost. When chlorine makes contact with wash water during washing, it yields an oxidizer, HOCI, which can inactivate pathogens effectively [16]. The (log 4.96 for day 0), which was higher than the initial count recovered for the no-bubble treatment or 10 min treatment (4.45 or 3.85 log CFU per avocado, respectively).

Discussion
Based on this study, a stream of targeted by microbubbles (<0.5 mm dia.) from a bubble diffuser (1.0 L air/min) can detach cells of L. monocytogenes from steel and produce surfaces. Extended bubble treatment times (up to 10 min) promoted greater removal of L. monocytogenes from stainless steel coupons, since significantly fewer L. monocytogenes were recovered from stainless steel after 10 min bubble treatments compared with no bubble treatment. The bubble application used in this study may reduce L. monocytogenes on stainless by nearly three logs (2.95 log CFU per sample) after an inoculum was dried for 48 h and loosely attached cells were removed. For cucumber and avocado, a 5 min application of the same heavy bubble stream removed at least 1.0 log more of the firmly attached bacteria than submerging the produce in water without cavitation bubbles. Lee et al. (2018) applied cavitating bubbles to inoculated fresh Roma tomatoes and cantaloupes using larger bubble sizes (~1-3 mm diameter) and higher air flows (3.5-14 L/min) than were used in this study [12]. Those authors reported that L. monocytogenes significantly decreased on Roma tomatoes (~1.0 log CFU after 30 sec and ~1.2 log CFU after 60 sec treatment) with a 14 L/min airflow rate compared to no cavitating bubbles. L. monocytogenes significantly decreased on cantaloupes (0.65 log CFU for 30 sec and ~0.74 log CFU for 60 sec treatment) when the maximum air-flow rate was used. In the study reported here, the log reduction in L. monocytogenes on cucumber and avocado surfaces also increased with increasing bubble treatment time, even though the airflow rate was lower, the treatment times longer, and the bubble diameters smaller. An important difference in these studies is that Lee et al. (2018) did not remove loosely attached cells prior to any bubble treatments. Here, in our study with cucumbers, avocados and stainless-steel coupons, the number of loosely attached cells recovered and enumerated after 1, 24 or 48 h drying was always higher than the quantity that could be recovered from the samples that were not treated with bubbles (treatment time = 0 min).
Chlorine is commonly applied to wash and clean fresh produce due to its bactericidal capacity and economical cost. When chlorine makes contact with wash water during washing, it yields an oxidizer, HOCI, which can inactivate pathogens effectively [16]. The (log 4.96 for day 0), which was higher than the initial count recovered for the no-bubble treatment or 10 min treatment (4.45 or 3.85 log CFU per avocado, respectively).

Discussion
Based on this study, a stream of targeted by microbubbles (<0.5 mm dia.) from a bubble diffuser (1.0 L air/min) can detach cells of L. monocytogenes from steel and produce surfaces. Extended bubble treatment times (up to 10 min) promoted greater removal of L. monocytogenes from stainless steel coupons, since significantly fewer L. monocytogenes were recovered from stainless steel after 10 min bubble treatments compared with no bubble treatment. The bubble application used in this study may reduce L. monocytogenes on stainless by nearly three logs (2.95 log CFU per sample) after an inoculum was dried for 48 h and loosely attached cells were removed. For cucumber and avocado, a 5 min application of the same heavy bubble stream removed at least 1.0 log more of the firmly attached bacteria than submerging the produce in water without cavitation bubbles. Lee et al. (2018) applied cavitating bubbles to inoculated fresh Roma tomatoes and cantaloupes using larger bubble sizes (~1-3 mm diameter) and higher air flows (3.5-14 L/min) than were used in this study [12]. Those authors reported that L. monocytogenes significantly decreased on Roma tomatoes (~1.0 log CFU after 30 sec and ~1.2 log CFU after 60 sec treatment) with a 14 L/min airflow rate compared to no cavitating bubbles. L. monocytogenes significantly decreased on cantaloupes (0.65 log CFU for 30 sec and ~0.74 log CFU for 60 sec treatment) when the maximum air-flow rate was used. In the study reported here, the log reduction in L. monocytogenes on cucumber and avocado surfaces also increased with increasing bubble treatment time, even though the airflow rate was lower, the treatment times longer, and the bubble diameters smaller. An important difference in these studies is that Lee et al. (2018) did not remove loosely attached cells prior to any bubble treatments. Here, in our study with cucumbers, avocados and stainless-steel coupons, the number of loosely attached cells recovered and enumerated after 1, 24 or 48 h drying was always higher than the quantity that could be recovered from the samples that were not treated with bubbles (treatment time = 0 min).
Chlorine is commonly applied to wash and clean fresh produce due to its bactericidal capacity and economical cost. When chlorine makes contact with wash water during washing, it yields an oxidizer, HOCI, which can inactivate pathogens effectively [16]. The bubble stream (mean of 3 replications × 3 samples). A,B,C Denote significant differences in recovery (p < 0.01) between storage times for each of the bubble treatment times.

Discussion
Based on this study, a stream of targeted by microbubbles (<0.5 mm dia.) from a bubble diffuser (1.0 L air/min) can detach cells of L. monocytogenes from steel and produce surfaces. Extended bubble treatment times (up to 10 min) promoted greater removal of L. monocytogenes from stainless steel coupons, since significantly fewer L. monocytogenes were recovered from stainless steel after 10 min bubble treatments compared with no bubble treatment. The bubble application used in this study may reduce L. monocytogenes on stainless by nearly three logs (2.95 log CFU per sample) after an inoculum was dried for 48 h and loosely attached cells were removed. For cucumber and avocado, a 5 min application of the same heavy bubble stream removed at least 1.0 log more of the firmly attached bacteria than submerging the produce in water without bubbles. Lee et al. (2018) applied cavitating bubbles to inoculated fresh Roma tomatoes and cantaloupes using larger bubble sizes (~1-3 mm diameter) and higher air flows (3.5-14 L/min) than were used in this study [12]. Those authors reported that L. monocytogenes significantly decreased on Roma tomatoes (~1.0 log CFU after 30 s and~1.2 log CFU after 60 s treatment) with a 14 L/min airflow rate compared to no cavitating bubbles. L. monocytogenes significantly decreased on cantaloupes (0.65 log CFU for 30 s and~0.74 log CFU for 60 s treatment) when the maximum air-flow rate was used. In the study reported here, the log reduction in L. monocytogenes on cucumber and avocado surfaces also increased with increasing bubble treatment time, even though the airflow rate was lower, the treatment times longer, and the bubble diameters smaller. An important difference in these studies is that Lee et al. (2018) did not remove loosely attached cells prior to any bubble treatments. Here, in our study with cucumbers, avocados and stainless-steel coupons, the number of loosely attached cells recovered and enumerated after 1, 24 or 48 h drying was always higher than the quantity that could be recovered from the samples that were not treated with bubbles (treatment time = 0 min).
Chlorine is commonly applied to wash and clean fresh produce due to its bactericidal capacity and economical cost. When chlorine makes contact with wash water during washing, it yields an oxidizer, HOCI, which can inactivate pathogens effectively [16]. The common dosage of commercial chlorine in industry ranges from around 50 to 200 mg/L, but 100 mg/L is the most common dosage used in industry [17]. The application of this dosage of chlorine requires a short impact time, of about 1 to 2 min, and a preferred pH of from 6.0 to 7.5 is used to maintain the stabilization of HOCI in order to prevent chemicals from corroding the processing equipment. Banach et al. (2015) showed that a five-log CFU of E. coli on fresh spinach can be completely removed with wash water containing a level of 7 mg/L free chlorine [16]. Although some chlorine solutions have been shown to inactivate pathogens at a higher level than reported in the present study using streams of microbubbles, some regulators, consumers, and food industries believe that an overreliance on chlorine rinses by food industries can lead to the excessive formation of chlorate residues that can harm human health.
Other researchers have compared the addition of several antimicrobial chemicals to the treatment water of fresh produce regarding their ability to inactivate or reduce pathogens, including Listeria monocytogenes. Chlorine, peracetic acid (PAA), chlorine dioxide (CIO 2 ), ozone, and electrolyzed oxidizing water (EOW) are some of the most common antimicrobial agents [18]. Rodgers et al. (2004) showed that both E. coli O157:H7 and L. monocytogenes can be inactivated over five logs when exposed to aqueous chemical sanitizers, including ozone, ClO 2 , chlorine, and PAA, for a 2-5 min exposure time [19]. Additionally, ozonation can be an excellent alternative for chlorine, as an economical and environmentally friendly disinfectant for this industry [20]. Additionally, ultraviolet light energy could be an effective measure by which industry could clean fresh produce. Yaun et al. (2004) reported using ultraviolet energy (24 mW/cm 2 ) to remove Salmonella spp. or Escherichia coli on the surface of fresh produce, including red apples, leaf lettuce, and tomatoes. These researchers reported that both Salmonella (2.65 log reduction) and E. coli (2.79 log reduction) can be inactivated to a similar level [21].
The present study was designed to evaluate microbubbles for removal or detachment of L. monocytogenes on stainless steel, avocado and cucumber surfaces. Even though microbubbles may reduce the concentration of L. monocytogenes on steel and some produce surfaces, they may or may not enhance microbial inactivation. Nevertheless, microbubbles can be an environmentally friendly alternative to chlorine for reducing the microbial surface contamination of fresh produce, which will be odorless and possibly less corrosive to equipment.
Based on the results, the cross-contamination of L. monocytogenes from inoculated to clean cucumbers and avocados increased with extended bubble treatment time (2 vs. 10 min). The transfer rate to cucumbers and avocados was also higher when inoculum was dried for 24 h versus 1 or 48 h. This result could be explained by differences in the total number of firmly or loosely attached cells that occurred with each drying time. The variable numbers of attached cells on inoculated cucumbers and avocados influence the number of cells that detach and adhere to uninoculated produce. The percentage of cells that were transferred between cucumbers was always higher than the percentage transfer between avocados, even though the initial inoculums were similar. Compared with cucumbers, the crosscontamination of avocados by L. monocytogenes was less affected by bubble treatments. In Lee et al. (2018), the cross-contamination of L. monocytogenes on Roma tomatoes was not significantly different with bubble treatment time (30 or 60 s) in terms of the transfer between inoculated to uninoculated Roma tomatoes, but significantly different between airflow rates (0, 7 or 14 SLPM) [12].
The application of a bubble stream similar to that used in this study could be used in a produce-packing house prior to the packaging and shipment of some produce to retail markets. This technique is not likely to be employed by consumers due to the need to purchase equipment that will be only occasionally used. After the consumer's retail purchase, the shelf-life of fresh, whole cucumbers is approximately 7 d when refrigerated (1-10 • C), and the shelf-life of fresh avocados is~3-4 d at room temperature (20-25 • C) and 7-10 d when refrigerated [22,23]. In this study, extending refrigerated storage from 4 to 7 d resulted in variable increases or decreases in surface bacteria growth and changes in quality for cucumbers and avocados. While an extended bubble treatment time appears to have higher effectiveness for removing bacteria from cucumbers, the surface aerobic plate count concentrations did not decrease over time. Similarly, for avocados, surface aerobic plate counts increased with longer storage times, and surface concentrations could