Control of pathogenic bacteria on the surface of rolling fruits by an atmospheric pressure air dielectric barrier discharge system

The schematic diagram of the plasma device and the corresponding experimental setup are shown in figure 1. The plasma device consisted of a copper mesh as H.V. electrode, a quartz tube (inner diameter: 4.6 cm, outer diameter: 5 cm, 40 cm, length: 40 cm) as dielectric barrier and a grounded copper foil (length: 40 cm, width: 1.0 cm). The fruit is placed on copper foil as a ground electrode. As shown in figure 1(a), the copper mesh is tightly attached to the outer wall of the quartz tube, and connected to DC nanosecond pulsed H.V. power supply. The copper mesh is 38.0 cm in length, 1.0 cm away from each end of the quartz tube, and its width is 12.4 cm, covering 85% of the outer circumference of the tube. Two electrical tapes with 1.0 cm width are affixed to both ends of the quartz tube to avoid redundant discharges. Inside the quartz tube, the copper foil is tightly attached to the inner wall.

Zoom In Zoom Out Reset image size

The plasma device is driven by the DC nanosecond pulsed H.V. power supply (Xi'an Smart Maple Electronic Technology Co., Ltd HVP-20P). In this study, the values of applied voltage, frequency, and pulse width of the voltage are kept constant at 17 kV, 1 kHz, and 3 µs, respectively. The rising and falling slopes of the positive square wave H.V. pulse have a voltage slope of approximately 340 V ns⁻¹. The development of plasma is captured by an ICCD camera (Andor iStar DH712). The detailed trigger method of the H.V. power supply and the camera has been described in our previous study [23]. The applied voltage and the discharge current are measured by an H.V. probe (Tektronix P6015A) and a current monitor (Pearson 4100), respectively. The waveforms are recorded by an oscilloscope (Tektronix MDO3104). In order to better study the development of plasma on the surface of the sphere and ensure the repeatability of discharge, the same size steel ball is used instead of the apricot for the diagnosis and analysis of the plasma development.

In order to better understand the development of plasma, the finite element method is used to calculate the distribution of applied Laplacian electric field distribution. The above calculations are implemented using COMSOL version 5.3 software [24], and the detail method has been described in our previous study [25]. The structure parameters of the model were consistent with those in the experimental configuration.

Staphylococcal food poisoning is the most common bacterial food poisoning, which may lead to sudden severe nausea and vomiting, abdominal cramps, diarrhea and fever [26]. 20 µl of S. aureus solution with the concentration of 10⁹ CFU ml⁻¹ is inoculated on aseptic gauze with an area of 1 cm². The distribution of the gauze pieces on the apricot surface is shown in figure 2. The five specific locations, named A, B, C, D and E from right to left (front view) are evenly distributed on the upper half of the fruit. The concentration of S. aureus inoculated on the gauze is calculated by following equation:

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle {{C}_{1}}=\frac{V\times {{C}_{2}}}{S}\nonumber \end{align}$$

where C₁ and C₂ are the concentration of S. aureus inoculated on the gauze and the concentration of S. aureus solution, V represents the volume of inoculated S. aureus solution. In this study, the value of C₁ was 2  ×  10⁷ CFU cm⁻².

Zoom In Zoom Out Reset image size

Apricots with good surface and no mechanical damage were purchased in Dalian local mall for microbial purification experiment. Five pieces of inoculated gauze were placed at A, B, C, D and E on the apricots surface, respectively. In this study, the apricot was treated by plasma in rolling and static states, respectively. After 15 s plasma treatment, the gauzes were vibrated in stroke-physiological saline solution for 5 min, and then 100 µl of liquid were absorbed and inoculated on LB medium. The inoculated LB medium was cultured for 24 h at 38 °C. The colonies were counted after the culture period. The gauze inoculated with S. aureus was placed on the surface of apricot as control and rolled for 15 s without the plasma device running, and then the number of S. aureus on the gauze was evaluated by the above method. The number of S. aureus in the control sample after rolling is about 1.5  ±  0.31 $\times$ 10⁷ CFU cm⁻². All data were obtained from three replicate experiments with a completely randomized design. Moreover, the surface temperature of the fruit under plasma treatment is measured by a thermal imager (Fluke Ti450).

The steel ball with a diameter of 45 mm was used as the ground electrode. The typical waveforms of the applied voltage and current are shown in figure 3(a). It should be point out that the measured current contains the displacement current and discharge current. The discharge power was calculated to be 48.4 W based on the data of applied voltage and discharge current given in figure 3(a).

Zoom In Zoom Out Reset image size

By comparison with the ICCD images, the discharge period in the rising and falling slopes during one H.V. pulse was marked in figure 3(b). In the rising slope, the plasma was generated between 80–140 ns, during which the applied voltage increases from 8.3 kV to 18.5 kV. In the falling slope, the plasma was generated between 3145–3210 ns.

The development of plasma was studied by using a steel ball with a diameter of 45 mm as the ground electrode. ICCD camera was used to capture the development of plasma from front and side views. The spatial-temporal evolution of discharge from the front view is given in figure 4. It is obvious from the ICCD images that the discharge at the rising slope was intense than that at the falling slope. However, no obvious plasma filaments were observed between the quartz tube and the surface of the steel ball from the ICCD images. From the front view, the uniform plasma was directly generated over the surface of the steel ball. The discharge was ignited on both sides of the lower part. With the increase of the applied voltage, the discharge developed from side to top along the electrode gap. The breakdown of the air gap at atmospheric pressure follows Paschen's law. The minimum value ((Pd)_m) of the product of air pressure (P) and the gap distance between electrodes (d) is 75.6  ×  10⁻² Pa·m and the P value at atmospheric pressure is 101.3 kPa. Therefore, the minimum gap with the lowest breakdown voltage (330 V) is 7.5 µm [27]. With the increase of the applied voltage, the breakdown gradually developed from the small to the large air gap region, which was consistent with the measurement in figure 4.

Zoom In Zoom Out Reset image size

Figure 5 shows the spatial-temporal evolution of the discharge captured by the ICCD from the side view during the rising and falling slopes. The arrows indicated the development direction of the discharge. During the rising slope of the applied voltage, the discharge was ignited at the bottom of the device (at the small gap region) at ~80 ns and then developed upwards and sidewards (to the large gap region), which exhibited a uniformly 'scanning' pattern. Based on the ICCD measurements, the average development velocity of the discharge was calculated to be ~10⁵ m s⁻¹. The discharge was sustained about 60 ns and extinguished at 140 ns. During the falling slope of the applied voltage, the discharge was ignited at the bottom of the device at 3145 ns, the discharge pattern and the sustaining time were consistent with the discharge during the rising slope. It should be noted that there is a significant difference in the intensity of the discharge during its development. The difference could be explained as follows: with increasing the gap distance, strong electron avalanches were spawned, resulting in increased discharge intensity [28]. When the discharge developed to the top of the steel ball, the emission intensities of the discharge were integrated to the ICCD from the side view (90–130 ns in the rising slope, 3160 to 3210 ns in the falling slope). Different from the randomness of discharge channels generated by plate-to-plate DBD, the development of discharge channel was artificially controllable. Discharge channel was ignited in the place where the air gap was small. With the increase of the applied voltage, the discharge channels developed regularly from the small air gap to the large air gap.

Zoom In Zoom Out Reset image size

An electric field simulation with the same structural parameters as the experiment was used to verify that the development of the discharge. The simulation showed that the development of the discharge was related to the distance of the electrode gap. The potential of the H.V. electrode was set to a constant value of 20.0 kV. It should be noted that, the calculation was carried out without plasma. In figure 6(a), a 3D distribution of the electric field in the space is given. The solid line represents the outer surface of the quartz tube and the dotted line represents the inner surface of the quartz tube. The electric field induced by the applied voltage showed an uneven distribution in the whole space, and the norm value of the electric field showed a rapid diminishing trend from the axis of the steel ball to the both sides of the space. In order to observe the distribution of the electric field over the surface of steel ball, the distribution of the electric field over the surface of steel ball from the side view is given in figure 6(b). The electric field over the surface of the steel ball in the axial direction gradually decreases from the bottom of the steel ball to the top of the steel ball, and the electric field was gradually reduced from the axis to both sides of the steel ball. As the distance of the air gap increased from 0 to 1 mm, the electric field significantly decreased since the gap distance greatly increases. The norm value of the surface electric field on the bottom of the steel ball was 7.7 log₁₀ V m⁻¹, which was 0.7 log₁₀ V m⁻¹ higher than that of the surface electric field on the top of the steel ball. From the side view, the minimum value of the surface electric field on both sides of the steel ball was 6.2 log₁₀ V m⁻¹, which was 1.5 log₁₀ V m⁻¹ lower than that of the maximum value. The discharge was firstly ignited in the small gap region where the electric field was relatively intense. With the increase of the applied voltage, the breakdown electric field was achieved and the discharge developed to the large gap region.

Zoom In Zoom Out Reset image size

The generation and extinction mechanism of discharge during rising and falling slopes is schematically shown in figure 7. In our case, when the applied voltage increased to 8.3 kV, the discharge was ignited in the small gap region (figure 7(a)). The electrons on the cathode surface drifted along the opposite direction of the E₀ in the air gap. The electron avalanche multiplication occurred when electrons gained enough energy to ionize or excite neutral atoms or molecules in the air gap [24]. When the electron avalanche developed from the cathode to the surface of the quartz tube, a discharge channel was formed, and the space charge was transported to the quartz tube and accumulated on the surface of the tube. Due to the surface charge accumulation, an electric field was built in the opposite direction of the applied electric field (Region1, figure 7(b)). The establishment of the reverse electric field led to a rapid decrease of the electric field strength in the discharge channel. The discharge was extinguished when the superposed field cannot maintain the discharge. In the same position, a second breakdown occurred in the original position only when the voltage in the air gap rose again to the initial air gap breakdown voltage. In this study, the minimum voltage for igniting discharge during the rising slope was 8.3 kV, however, no secondary discharge was observed with increasing the applied voltage to 18.5 kV (figure 3(b)). The charge accumulated on the surface of quartz tube per unit area is defined as Q(s,t):

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle Q\left(s,t \right)={{C}_{d}}\left[ {{U}_{a}}\left(t \right)-{{U}_{g}}\left(s,t \right) \right]\nonumber \end{align}$$

where C_d is the capacitance value of quartz tube, Q(s,t) is the amount of charge accumulated on the surface of quartz tube per unit area, U_a(t) is the applied voltage, U_g(s,t) is the voltage across the air gap. Q(s,t), U_a(t) and U_g(s,t) are functions of time.

Zoom In Zoom Out Reset image size

For the current design, U_g(s,t) was determined by the distance of air gap. After charges were accumulated, the capacitance of the quartz tube remained unchanged. $Q\left(s,t \right)$ increased with increasing ${{U}_{a}}\left(t \right)$. At a given ${{U}_{a}}\left(t \right)$, the larger the air gap, the larger the ${{U}_{g}}\left(s,t \right)$, resulting in a smaller $Q\left(s,t \right)$. Therefore, the smaller the air gap, the greater the amount of charge accumulated on the surface of quartz tube. The accumulation of surface charge on the surface of quartz tube was uneven. With the increase of air gap, the less charge accumulated on the surface of quartz tube per unit area (figure 7(b)). The accumulated charges which may flow along the surface would generate a strong reverse electric field. The reverse electric field would prevent the occurrence of secondary breakdown.

The micro-discharge of a unipolar high voltage pulse at falling slope was generally recognized as being caused by a reverse electric field generated by the charge accumulated on the surface of the dielectric [29, 30]. The former anode was 'converted' to be the cathode due to the negative surface charge. When the voltage provided by the E in the air gap reaches the breakdown voltage of the air gap, the discharge was ignited. Electrons accumulated on the surface of quartz tube were transferred to the anode under the action of E (figure 7(c)), and neutralized with the positive charge in space during development, which eventually led to the decrease of the number of negative charge accumulated on the surface of quartz tube, the reducing of the E intensity generated in the air gap, and finally the extinction of discharge (Region3, figure 7(d)). For unipolar positive pulse discharge, the length of pulse width has a significant effect on the generation of discharge channel at the falling slope, and the detailed work has been reported [29].

As shown in figure 2, S. aureus was inoculated at five specific locations on the apricot surface. The inoculation concentration of S. aureus at each special location on the apricot surface was ~7 log₁₀ CFU cm⁻². The plasma treatment time of the fruit in static state and rolling state (Fruit rolls in one direction at a constant speed in the device) was 15 s. Due to the different discharge intensity on five specific locations, the bactericidal effects of the apricot induced by ACP may result in different performances. When the apricot was treated by plasma in static state, the survival numbers of bacterial on A, B, C, D and E decreased from 7 to 3.28, 2.11, 1.82, 2.83 and 3.19 log₁₀ CFU cm⁻², respectively (figure 8). When the apricot was treated by plasma in rolling state, the survival number of bacteria at different locations on the apricot surface was significantly decreased. The distance of air gap increased from location A (E), B (D) to C. The larger the air gap was, the longer the development of the electron avalanche, leading to the high the concentration of active species generated by plasma, such as singlet oxygen, super oxygen anion etc. Besides, the plasma treatment time was inconsistent for different locations of the apricot surface. As shown in figure 5, discharge was rarely generated in the middle position of the steel ball along the axial direction (A/E location). During the rising slope of the H.V. pulse, the ignition times of discharge at the location of A (E), B (D) and C were 5, 10 and 35 ns, respectively (figure 5). The treatment time play an important role in the plasma sterilization, the bactericidal efficiency of plasma increased with the increase of treatment time [8, 9, 31].

Zoom In Zoom Out Reset image size

In the rolling state, the bacterial survival numbers on different locations of A, B, C, D and E decreased from 7 to 1.71, 1.2, 1.1, 1.15 and 1.86 log₁₀ CFU cm⁻², respectively. When the apricot is treated by plasma in static state, the development of the plasma is limited by the distance and hence some areas on the gauze cannot be directly exposed to the plasma. With the apricots rolling in the device, uniform plasma could be generated over each part of the gauze, and the bactericidal efficiency was greatly improved.

Figure 9 presents the ICCD images captured while the steel ball with a diameter of 45 mm or 38 mm was rolling in the device. The relative sizes of the steel ball in the device are illustrated in figures 9(a) and (b). The ICCD images were taken in single shot mode with a gate width of 1.0 s. Figure 9(c) shows the corresponding discharge image with the steel ball of diameter of 45 mm. With the ball rolling in the device, plasma could be uniformly generated over the ball. The discharge image with the steel ball of diameter of 35 mm is shown in figure 9(d). The discharge was only generated at the bottom of the plasma device, and the discharge intensity was obviously decreased. By contrast, it was apparent that when large steel ball rolling in the plasma device, intense and uniform plasma could be generated. The size of the steel ball had a significant effect on the uniformity of the plasma in such a device. The more suitable the size of the steel ball, the smaller the air gap will be. According to Paschen's law, the smaller the distance of the air gap, the easier the breakdown occurred, thus the uniform plasma in the space could be obtained.

Zoom In Zoom Out Reset image size

Figure 10 shows the survival numbers of bacteria on apricot surface with different size treated by plasma in rolling state at five specific locations. When the height of the apricot is 4.4 cm and the width is 4 cm (larger one), the bacterial survival numbers on A, B, C, D and E decreased from 7 to 2.39, 1.79, 1.34, 1.86 and 2.47 log₁₀ CFU cm⁻², respectively. When it came to the apricot of 3.1 cm tall and 2.9 cm wide (smaller one), the bacterial survival numbers on A, B, C, D and E decreased from 7 log₁₀ to 4.37, 3.31, 2.18, 3.46 and 4.14 log₁₀ CFU cm⁻², respectively. At the same location on the surface of apricot with different sizes, the survival numbers of bacteria were significantly different. For the same inoculated area, the survival number of bacteria on the surface of large apricot is significantly lower than that of small apricot. In the practical industrial application of the device, it is very important to select the appropriate inner diameter of quartz tube according to the size of apricot.

Zoom In Zoom Out Reset image size

Figure 11 shows the infrared thermal images of the apricot before, during and after the plasma treatment for 15 s in rolling state. The temperature of the apricot surface increased from 27.1 °C to 27.9 °C after the plasma treatment, showing that the plasma treatment had no vital impact on the apricot surface temperature. In figure 11(b), the discharge between the ground electrode and the surrounding high voltage electrode resulted in the temperature of 79.1 °C around the ground electrode. However, the change in the apricot surface temperature was not significant as the apricot kept rolling in the device within 15 s.

Zoom In Zoom Out Reset image size