Effects of plasma activated species produced by a surface micro-discharge device on growth inhibition of cyanobacteria

The increase of nutrient supply is accelerating primary productivity and eutrophication. An obvious and problematic symptom of eutrophication is the rapid growth of phytoplankton, causing discoloration in the affected waters [1]. These events are termed blooms. Numerous freshwater genera including diverse phytoplankton are capable of forming blooms, among which the cyanobacteria is the most notorious one. cyanobacteria has a strong ecological competitive advantage after long-term evolution. It can produce a range of organic compounds, including those which are toxic to high-ranker consumers, from zooplankton to those further up the food chain. Harmful blooms negatively affect aquatic ecosystems, fisheries, and human health [2]. The 2006 toxic algae blooms on Ultah Lake led to lake closure due to the associated risks, and the nuisance algae blooms that invaded Lake Erie in August 2014 left over 400,000 residents in Ohio without drinking water [3]. Taihu Lake, the third largest fresh-water lake in China, suffers from much more serious cyanobacteria blooms comparing with Ultah Lake and Lake Erie, in pace with the rapid economic growth in China, and the highest level of cyanobacteria accounts for more than 91% of the total algae species, among which microcystis aeruginosa is the main species [4]. Therefore, how to kill algae in polluted water without bringing new pollution sources becomes an important research hotspot in environmental protection.

Great progresses have been achieved in sterilization basing on low temperature plasma in the last decade [3, 5–7]. However, the growth environment of cyanobacteria is different from that of bacteria. The target solution to be treated has a relatively low concentration of cyanobacteria, and the liquid environment represents an additional barrier which will impede direct plasma-microorganism interaction. Consequently, plasma-induced effects are mediated by the liquid phase. Reactive species generated in the plasma have to penetrate or diffuse into the liquid to interact with microorganism cells. Moreover, plasma-induced changing of the liquid environment may have an impact on ecosystems, which has to be considered. For example, the eruption of cyanobacteria is usually accompanied by the pH rising of the lake water. In summer, the pH can increase over 8 during the outbreak of cyanobacteria in Taihu Lake. However, the active species such as reactive nitrogen species (RNS) generated by plasma discharge can turn the solution acidic [8, 9]. If the pH decreases to 3 ∼ 4, the antibacterial effectiveness is known to drop significantly [10], which is unacceptable to the aquatic ecosystem. Therefore, it is necessary to study the effects of the acidity and alkalinity on the growth inhibition of cyanobacteria.

In this article, we present a kind of simple surface micro-discharge (SMD) device, which is designed to inhibit cyanobacteria growth. The structure of this device and the experimental setup is described in section 2. The experimental results and physical understanding are presented in section 3 and in section 4 a summary.

Figure 1(a) shows the SMD device used as the plasma producer in the experiments. Figure 1(b) is the left view photo of discharge. High voltage (at the order of tens of kilovolts, 17.75 kHz) is applied between the stainless steel spiral wound and the brass tube, which work as two electrodes. The diameter of the spiral wound is 18 mm, with a pitch of 3.2 mm and the total length of 142 mm. The brass tube has an inner radius of r = 10 mm, an outer radius of 12 mm and a length of l = 139 mm. A 1 mm-thick quartz glass tube is sandwiched between the electrodes as the dielectric barrier. All of these components are wrapped up in apolytetrafluroethene (PTFE). Air flow is pumped into the tube from one end and carries RONS out of the tube from the other end. Figure 2(a) shows the photo of this SMD device and figure 2(b) shows the discharging in dark environment. The typical power is at the order of tens of Watts, according to the applied voltages. It is to be noted here in figure 2(a) the stainless steel spiral wound and the quartz glass tube are pulled out of the PTFE just to show these inner components clearly. The sizes, together with the electrode shape, is carefully designed before fabrication based on the principle of simplicity. The main physical sizes, such as the thickness of the quartz glass d and the inner radius of the brass tube r, determine the plasma property of SMD. The thicker the quartz glass, the lower the SMD input energy [11] without changing the applied voltage, resulting in the decreasing of RONS. Here we define the per unit-area yield of RONS as M. Then the average concentration C of RONS in this tube follows

$$\begin{eqnarray*}&&C=\displaystyle \frac{\left(2\pi r\right)\cdot l}{\pi {r}^{2}\cdot l}\cdot M=\displaystyle \frac{2M}{r}.\end{eqnarray*}$$

It is clear that the concentration C is inversely proportional to the radius of the brass tube r. But for a much small radius r, it is difficult to introduce the air flow. As a compromise, the radius of r = 10 mm is finally chosen.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A high voltage probe (P6015A,Tekronic), together with an oscilloscope (TDS2004C,tektronix) is used to measure the voltage parameters. Figure 1(c) shows the sketch of ozone concentration measurement. The confined gas absorption cell is made of quartz glass with a length of 150 mm, a radius of 25 mm and a quartz thickness of 2 mm. Air plasma is blown into the quartz cell with a flow rate of 4 L min⁻¹ (changeable) in most experiments presented in this article, except for the gas flow experiments.

Ozone generated by SMD is measured in this cell by ultraviolet (UV) absorption spectroscopy. An argon/mercury lamp (HG-1, Ocean optics) works as a UV source. Two iris diaphragms with a 1mm-diameter aperture each are set at the center of both ends of the cell. UV light passing through the two diaphragms is collected by an optical fiber, which is linked to a UV spectrometer (STS-UV, Ocean optics). The intensity of UV light with a wavelength of 254 nm is recorded to calculate ozone concentration based on Beer's law. The photon absorption cross section is 1.15 × 10⁻²¹ m². The length of the optical path/measurement line is 150 mm. The temporal resolution and detection limit of measurements in this article are 1 s and 1ppm, respectively.

Microcystisaeruginosa (FACHB-912, Freshwater Algae Culture Collection at the Institute of Hydrobiology, FACHB-collection), the dominant algae specie among cyanobacteria species, is chosen to be investigated in the experiments. The cyanobacteria is cultured in light incubator (MGC-300,Shanghai Sanfa scientific instruments Co., Ltd) using BG11 medium(Blue-Green Medium, pH 7.5–9.0) with a temperature of 25 °C, an illumination intensity of 2500Lx, and a light-dark cycle of 12/12 h. Cyanobacteria cells are numbered under the microscope (SK2009, Saikedigital) using platelet counting method, and five squares are randomly selected according to the biometric method and averaged.

The inhibitory effect of plasma on the growth of cyanobacteria is quantified by the log reduction, i.e. log(N₀/N), where N₀ stands for the number of cyanobacteria cells which are not treated and N stands for the number of cyanobacteria cells treated by air plasma. The treated cyanobacteria cells in solution are counted after 4 days of culture.

The pH values are measured by a pH meter (PHSJ-3F, Shanghai REX Instrument Factory).

It is known that RONS are main products in indirect plasma, including O₃, NO₂, N₂O, N₂O₅, HNO₂, HNO₃ and H₂O₂ [12]. Different RONS have different effects on cells and tissues, especially in contact with water. NO₂, N₂O and N₂O₅ will react with water to form nitric acid and nitrite, resulting in a pH decrease of the solution [9]. Under acidic condition, H₂O₂ and some other species can achieve long-term bacteriostatic effect [13]. However, the percentage of H₂O₂ in RONS is always very small, and what's more, acid would do harm to the ecological environment. Ozone, a kind of strong oxidant, which can be produced by SMD discharge, offers an option to effectively treat cyanobacteria in the water without changing pH value.

Figure 3 shows time evolutions of ozone concentrations under different applied discharging voltages. We can roughly divide these discharges into three modes, according to applied voltages and the resulting ozone production.

Mode 1: The first mode includes discharges with an applied voltage of lower than 20 kV. When the discharge ignites, the concentration of ozone increases and after seconds it keeps in a steady level. The steady ozone concentration increases along with the increase of applied voltage.

Mode 2: When the voltage exceeds 20 kV but is lower than 30 kV, the discharge changes to a complete different mode. When the discharge ignites, the ozone concentration sharply increases to a high peak level, and then begins to decrease very slowly. The peak value continues to increase along with the increase of voltage.

Mode 3: When the applied voltage exceeds 30 kV, although the ozone concentration increases to its peak value in seconds, it drops so sharply that after tens of seconds (or even shorter, according to the applied voltage) it is below the detection limit (∼1 ppm) in the experiments.

Zoom In Zoom Out Reset image size

These mode behaviors are physically and chemically understandable. Generally speaking, ozone production is related to power deposition [14]. However, shimizu [15] found that the energy input per molecules is not the only global parameter to comprehensively describe the complex, transient and non-linear behavior of air plasma chemistry. The main cause is the reaction between vibrationaly excited nitrogen molecules (N₂(v)) and O atoms. As the applied voltage increases, the fraction of vibrationally excited nitrogen increases, thereby greatly increasing the rate of creation of NO, the major quencher of O₃, because NO reacts with O₃ to O₂ & NO₂ or other nitrogen oxides. N≡N bond has a strong bond energy (942 kJ mol⁻¹), while O=O bond energy is much lower (498 kJ mol⁻¹). This chemical characteristic makes nitrogen discharging much more difficult than oxygen or some other discharging gases, such as argon. The empirical discharging voltage threshold for nitrogen is from teen kV to over 20 kV, according to devices and discharging conditions. Figure 3 confirms the key role of applied discharging voltage in SMD. For Mode 1, the applied voltage is lower than nitrogen discharging threshold of this SMD device, resulting in a pure oxygen discharging. The ozone concentration increases along with the applied voltage, or in other words, the power deposition, and keeps at a saturated level when the production rate and decomposition rate are equal. In this mode, pure oxygen discharging without any nitrogenous species is experimentally confirmed by the undetectable pH change of the solution after being passed by air plasma flow. For Mode 3, the applied voltage is far higher than the nitrogen discharging threshold. A great deal of nitrogenous species are produced, resulting in a quick quenching of ozone. pH value checking experiments are also performed (see figure 6) and the solution turns acidic very quickly (in seconds), confirming the existence of nitrogenous species. For Mode 2, the applied voltage is higher than the threshold but not very high. The ozone concentration increases along with the increase of applied voltage (power deposition). The concentration curve reaches its peak in seconds, indicating the fact that ozone generation still dominates the discharge in the first seconds, and then falls slowly because of the chemical reaction with nitrogenous species. This physical image is confirmed by experimental results as are shown in figure 7(a), in which pH value decreases much more slowly (in minutes) comparing with that in figure 5, implying a limited quantity of nitrogenous species. So the underlying physics and chemistry of these three modes are 'competition' between ozone and nitrogenous species and these three are ozone dominating, transition and nitrogen oxides dominating modes, respectively. It is to be noted here that in Mode 3 discharge, the saturated electrode working temperature is experimentally proven to be over 40 °C, which is much higher than those in Mode 1 and Mode 2 because of the higher power deposition. This high temperature will further expedite the ozone decomposition.

Figure 4 shows the evolutions of ozone concentrations under different gas flows with an applied voltage of 28 kV. From the figure, we can see that ozone concentration increases along with the increase of gas flow. This is reasonable because the gas flow carries ozone. This increasing saturates when the gas flow reaches its threshold of 12 L min⁻¹. This saturation comes from the limited ozone production rate of this SMD.

Zoom In Zoom Out Reset image size

The inhibition of ozone on the growth and reproduction of cyanobacteria is experimentally investigated. A 17 kV voltage is applied onto the SMD plasma source (Mode 1), and in this kind of discharge configuration the main reactive component is ozone. Two types of treatments are performed. Type 1: Firstly, the air which has already been treated by SMD discharging is blown into the culture fluid (BG11) for 1 ∼ 13 min, and then cyanobacteria is inoculated into the culture. Type 2: We firstly inoculate cyanobacteria into the culture and then we use the air plasma flow to deal with cyanobacteria solution for 1 ∼ 13 min.

After cultured in light incubator for four days, the cyanobacteria cells are counted. In fact, cyanobacteria cells increase 6 times in the control experiment, or in other words, a log reduction value of larger than 0.8 indicates basically a good inhibition of the growth of cyanobacteria.

Figure 5(a) shows the inhibitions of cyanobacteria growth in Type 1 (grey squares) and Type 2 (red dots) experiments. The red fold line shows a quasi-linear characteristic, while the grey line shows a very weak and un-stable treatment. After a Type 2 treatment for longer than 8 min, the log reduction value would be larger than 0.8, indicating an effective inhibition of cyanobacteria growth. The grey fold line shows a un-stable character and the log reduction value hovers around 0.2, indicating a weak effectiveness of ozone in Type 1 experiments. The underlying physics for this is the limited solubility of ozone. Ozone will saturate in the culture fluid soon and can not promote the inhibition ability, no matter how long the culture fluid is treated by the SMD plasma. But for Type 2 experiments, ozone keeps taking effect on cyanobacteria, resulting in a continuous increase of inhibiting effect. The effect of the increase of ozone concentration on reduction is experimentally checked. Mode 1 (Type 2) is chosen as the discharging mode to avoid cyanobacteria reduction caused by nitrogenous species. The decrease or increase of ozone concentration is achieved by means of changing the applied voltage without changing the discharging mode. As is shown in figure 5(b), the log reduction increases apparently along with the ozone production, confirming the efficiency of ozone on cyanobacteria growth inhibition.

Zoom In Zoom Out Reset image size

Another set of experiments are performed with an applied voltage of 34 kV (Mode 3). After 1-min discharge, ozone is thoroughly quenched. Then the air plasma is blown into cyanobacteria solution (Type 2). The pH values and log reduction are drawn in figure 6 by little black squares and red inverted triangles, respectively. One can clearly seen the gradual pH decreasing as time goes by. The log reduction shows an un-steady slow increase along with the decrease of pH, but it explosively increases when pH reaches a threshold of 6.8. After a 200-second treatment, the pH drops to about 6.7, and all the cyanobacteria cells cultured in this acidic environment have died. The upper layer of the solution is cyan, and there are white small particles deposited at the bottom of the solution. It is to be noted here that, in the experiments microcystis aeruginosa, which is suitable to live in alkaline environment, is chosen as the main study object. The best pH for it is 9, higher than the general algae and microorganisms [16]. That is why Mode 3 SMD plasma can exterminate cyanobacteria in the solution. Unfortunately, in this discharging mode the dropping in pH means the increase of the nitrate ion density in the solution (nitrite ions will convert to nitrate), which may aggravate the eutrophication of water bodies. Further more, some algae can fix nitrogen and increase nitrogen content in solution even when nitrogen content is low during blooms. A comprehensive assessment of the impact of nitrogen on water quality and a compromise between acidic treatment and ozone treatment are valuable and worthy of deep researching.

Zoom In Zoom Out Reset image size

Inhibition effects of ozone and pH decrease on the growth and reproduction of cyanobacteria are investigated in Mode 2 SMD discharges with an applied voltage of 21 kV. In this mode, the production of nitrogen oxides leads to a decrease of ozone concentration. Blowing the air plasma into the solution will turn the solution to acidic. The acidity and ozone work together to kill cyanobacteria. pH changes in Type 1 and Type 2 experiments are shown in figure 7(a) by black and red, respectively. pH in each type drops quasi-linearly by about 0.05 per minute. Log reduction in each type is shown in figure 7(b). In Type 2 experiments, the log reduction steadily increases following a quasi-linearity, and reaches the joyful value of 0.8 after the treatment for a time period of about 7.5 min, characterizing the inhibition ability of ozone. It is to be noted here that this finding is of great difference to Type 2 experiments in Mode 3 discharges as was shown in figure 6. In Type 1 experiments, the log reduction keeps very low, which is similar to that in Mode 1 Type 1 discharges, until the sharp rocketing when pH falls lower than 7, characterizing the inhibition ability of acid. These Type 1 experiments confirm the acidity threshold of pH 6.8 what is the conclusion drawn in Mode 3 Type 2 experiments.

Zoom In Zoom Out Reset image size

As a supplement, one can experience the Mode 2 Type 2 experiments visually in figure 8. The conical beakers are located from left to right according to the treatment time period, i.e. 0, 1, 4, 7, 10 and 13 min, respectively. A 7-min treatment can effectively inhibit the growth of cyanobacteria.

Zoom In Zoom Out Reset image size

Therefore, in general, when pH is higher than 6.8, the ozone inhibition is dominant, and when below 6.8, the effect of acidity exceeds the effect of ozone. From above experiments, we find an effective and optimal operating regions for growth and reproduction inhibition of cyanobacteria by using SMD plasma, at least for SMD used in this article. That is an applied voltage range of 20 to 30 kV (Mode 2) , Type 2 treatment and a suitable treatment time period. This conclusion tells us that the treatment time period is not the longer the better, and for our case, a period of 8 to 11 min is optimal.

A kind of SMD air plasma flow device has been developed to produce the reactive oxygen/nitrogen species (RONS). It has not a complicate structure, but a high productivity. By means of easily changing the applied voltage of this SMD, the yield of ozone and nitrogen species can be controlled. This characteristic makes it a convenient device to deal with plasma treatment in different circumstances, such as cyanobacteria inhibition presented in this article. The size is designed to reach a compromise between the productivity of ozone and the air flow rate. Cylindrical electrodes promise an even surface productivity of ozone and a steady gas flow for convenient Type 2 treatments. As is shown in figure 4, the ozone concentration maximizes at a gas flow of 12 L min⁻¹, which implies a saturation. The size and the long cylindrical shape make it easy to group SMDs to enhance the treatment ability. Further more, another main attraction of this kind of SMD is economy because of its simple structure, which is a key issue for huge-volume water treatment in practical applications. All of these make it an effective and acceptable device not only for environmental but also for hygienic or industrial usages.

Three discharging modes have been found in the experiments performed in this article, defined and differentiated from each other by the applied voltages. In Mode 1 with an applied voltage of lower than 20 kV, the ozone concentration saturates at a steady level and no nitrogenous species are produced. In Mode 3 with an applied voltage of over 30 kV, the ozone concentration reaches its peak in seconds and then drops sharply to almost zero. In this mode, a great deal of nitrogenous compounds are confirmed by the great pH decreasing of the solution in seconds, resulting in the quenching of ozone production. In Mode 2 with an applied voltage of between 20 kV and 30 kV, the ozone concentration reaches its peak in seconds and then decreases slowly. This mode is a kind of moderate discharging, in which ozone concentration keeps much higher than the other two and is proven to be a desired discharging mode without changing the pH significantly. These differences among modes come from the interruption of N≡N chemical bonds in the working air.

The inhibition of growth and reproduction of cyanobacteria is experimentally investigated by using SMD air plasma flow. When pH is over 6.8, ozone plays a major role. After about 7-min treatment, the growth of cyanobacteria is effectively inhibited. When pH drops below 6.8, the pH decrease/acidity begins to take effect in this inhibition. These results indicate that plasma-generated active substances are effective in inhibiting the growth of cyanobacteria. A suitable discharging voltage and a suitable treatment time period are key parameters determining the treatment effect. The air flow with RONS should be blown directly into cyanobacteria solution.

It is to be noted here that the algae species treated in the experiments are microcystis aeruginosa. For other algae species which have larger volumes or stronger vitality, the effect of plasma active substance treatment maybe not so obvious, which requires further experimental researches.

This work is supported by the National Magnetic Confinement Fusion Energy Research Project (Grant No.2015GB120002) and the National Natural Science Foundation of China (Grant Nos. 11535013, 11875124). The authors also acknowledge the technical support from Nanjing Kepu Medical Technology, Co., Ltd.