## Abstract

The disinfection of recirculating nutrient solution is essential to avoid disasters due to the dispersal of pathogens in closed soilless culture systems. This work presents the development of a new technique to perform disinfection, based on pulsed electric field (PEF) treatment. Hoagland solution inoculated with Fusarium oxysporum was selected as the specimen to verify the effectiveness of PEF treatment. The results indicate that PEF deactivated most Fusarium oxysporum in nutrient solution within a few seconds with the maximum disinfection efficiency achieved being 99.84%. The disinfection efficiency became higher with the extending of treatment time or the increase of PEF strength, while the initial microbial density was proven to play no role. Temperature rise indicates that PEF treatment played the dominant role in the process of disinfection. The energy efficiency depends mainly on treatment time and electric field intensity. The optimal operating conditions were found to be: treatment time less than 10 seconds and electric field strength around 5 kV/cm. This investigation implies that, in the future, PEF treatment has the potential to be applied to disinfect nutrient solution.

## INTRODUCTION

A closed soilless culture system with recirculating nutrient solution features advantages over conventional drain-to-waste or open systems, such as reducing the cost of water and fertilizer, and avoiding environmental pollution (Ehret et al. 2001). However, the recycling use of nutrient solution could result in dispersal of pathogens in a growing system (van Os 1999). A few disinfection techniques are currently being utilized to avoid an outbreak of root-borne diseases due to contaminated nutrient solution: for example, ozone treatment (Kobayashi et al. 2011), chlorine or chlorine dioxide processing (Scarlett et al. 2016; López-Gálvez et al. 2017), UV radiation (Sutton et al. 2000), heat sterilizing (Runia & Amsing 2001a), and removing pathogens with slow filtration or constructed wetlands (Wohanka et al. 1999; Gruyer et al. 2013).

Each method has its noticeable merits, but also a few inherent weaknesses that also need to be considered. Ozone may react with specific iron chelates, manganese and organic matter in solution, hence, the effect of ozone treatment could be reduced. Moreover, the low dissolvability of ozone in solution implies potential health and environmental hazards (Stewart-Wade 2011). Interacting with some compounds (ammonium, organic nitrogen, etc.), chlorine forms chloramines (Raudales et al. 2014), which could hinder the growth of lettuce (Date et al. 2005). UV radiation most likely leads to photo-degradation of iron chelates (Albano & Miller 2001) and also the transparency of nutrient solution needs to be at least 60% to maintain the efficacy of UV (Yiasoumi et al. 2014). Heat sterilization not only consumes a large amount of energy but also may have undesirable effects on some slightly soluble chemical compounds (Runia 1994). Tu & Harwood (2005) reported that slow filtration needs a large infrastructure and sediment clogging requires frequent maintenance.

The pulsed electric field (PEF) method has been widely investigated as a technology for the disinfection of liquid food and beverages (Sharma et al. 2014). The commonly accepted disinfection mechanism of PEF is the effect of electroporation and it has been demonstrated that this microorganism inactivation technology is both environmentally friendly and non-thermal (Barbosa-Cánovas et al. 1999). These advantages of the treatment could overcome some of the drawbacks of the presently existing methods. However, to the authors' best knowledge, the treatment of recirculating nutrient solution using PEF technology has not yet been reported.

In this work, to assess the PEF technique, a novel PEF-based disinfection system was developed to investigate various factors for disinfection efficiency. The study ends with positive conclusions.

## MATERIALS AND METHODS

### The preparation of the inoculated nutrient solution

Fusarium oxysporum is a serious pathogen in greenhouses, which may invade seedling and fruit production (Song et al. 2004; Elmer 2015), leading to a significant decrease in yield and a contamination of recirculating nutrient solution. Therefore, this study selected F. oxysporum as a typical pathogen. Since Hoagland solution is widely used in plant-growing (Hothem et al. 2003), it was chosen as the living environment of the pathogen.

The activated F. oxysporum, which was isolated from infected cucumber plants, was inoculated on potato dextrose agar medium (PDA, 50 g potato, 5 g dextrose, 5 g agar) in petri dishes, and then had 72 hours of cultivation at 28 °C in an incubator. To expand the cultivation, a 1-cm diameter colonized medium was cut from a petri dish and transferred into a 250 ml conical flask, which contained 80 ml potato dextrose broth (PDB, 50 g potato, 5 g dextrose). After that, the conical flask was put on an orbital shaker (SPH-200B, SHIPING, Shanghai, China) at a speed of 150 rotations/min at an ambient temperature of 28 °C. The suspension in the conical flask was cultivated with shaking for 150 hours. In the last step, the diluted Hoagland solution was mixed with the suspension of F. oxysporum at a composition rate of 495 ml/5 ml, after which the inoculated nutrient solution for the disinfection experiments was ready (Huang et al. 2018).

### Experimental setup

Figure 1 presents the scheme of the PEF treatment system, which was basically made from two distinctive parts: a high-voltage pulsed power supply and a series of treatment chambers.

Figure 1

Schematic diagram of the PEF disinfection system.

Figure 1

Schematic diagram of the PEF disinfection system.

The pulsed power supply contained a high-frequency transformer, a rectifier unit, a pulse-forming network (PFN) and a rotating spark gap switch (SGS). The PFN was made of a nine-stage L-C ladder with a characteristic impedance (ZC = (L0/C0)1/2) close to 33 Ω, where L0 = 18.7 µH and C0 = 18.1 nF. The repetition rate of the rotating SGS was 4.5 pulses per second. The pulsed peak voltage (Vp) could be adjusted from 4 kV to 8 kV.

The waveform of the pulsed voltage impulse was measured using a high-voltage probe (1000:1, P6015A, Tektronix, USA) connected to a digital oscilloscope (TDS2012C, Tektronix, USA). The pulse width is determined by 2n (L0C0)1/2, where the number of stages of the L-C ladder was n = 9. The same pulse width of 10 µs was applied in all the tests performed.

The PEF treatment chamber was designed as shown in Figure 1. The nutrient solution was filled into the space between the two circular electrodes, which were made of stainless steel with a diameter of 30 mm and a thickness of 5 mm. The distance (d) between the electrodes was 10 mm. The body of the treatment chamber was made of acrylic glass. In all tests, the resistance of each fully filled chamber was kept at (60 ± 1) Ω.

### Evaluation of disinfection efficiency

The total plate count (TPC) method (Kobayashi et al. 2011) was adopted to detect the microbial density of the suspension. The disinfection efficiency δ (%) was calculated by:
(1)
where N0 and N are the colony numbers of F. oxysporum before and after PEF treatment, respectively. All experiments were carried out in triplicate.

### Experimental design

#### Experiment of disinfection efficiency versus treatment time

The initial microbial density of the inoculated nutrient solution treated in the experiments was 1.04 × 106 colony-forming units per millilitre (CFU/ml). The 24 chambers were randomly divided into eight equal groups for different treatment times (0 s, 5 s, 10 s, 15 s, 20 s, 25 s, 30 s, and 60 s). The applied pulsed voltage impulse for all the groups was the same: 4.6 kV, 10 µs, 4.5 pulses per second.

#### Experiment of disinfection efficiency versus PEF strength

To investigate the effect of PEF strength on the disinfection efficiency, 18 treatment chambers were randomly divided into six equal groups for different PEF strengths (E: kV/cm). Each treatment time was 10 seconds. The initial microbial density of nutrient solution for each experiment was 4.15 × 105 CFU/ml.

#### Experiment of disinfection efficiency versus initial microbial density

The experiment was to investigate the relationship between the initial microbial density and disinfection efficiency. In each experiment, the treatment time was 15 seconds, and the PEF strength was 6 kV/cm. The four initial microbial densities used for the experiments were 3.5 × 103, 3.5 × 104, 4.0 × 105 and 4.0 × 106 CFU/ml.

#### Experiment of temperature rise in the nutrient solution due to PEF treatment

The experiment was to confirm whether PEF treatment represents the dominant factor of disinfection. An infrared thermal imager (U5855A, Agilent Technologies, USA) was used to measure the temperature of the nutrient solution. The thermal imager captured temperature every 10 seconds during the 60 seconds of treatment. The PEF strength ranged from 4 kV/cm to 7 kV/cm. The measuring point was focused on the surface of the nutrient solution.

## RESULTS AND DISCUSSION

### Effect of treatment time on disinfection efficiency

Figure 2 shows that the residual microbial density decreased rapidly in the first few seconds. Meanwhile, the disinfection efficiency reached 95.38% under a 5-second treatment. The results demonstrate that the F. oxysporum in the nutrient solution was mostly deactivated within a few seconds. This result is very important and implies that, in the near future, a continuous-flow PEF treatment system can be developed.

Figure 2

The variation of the residual microbial density and of the disinfection efficiency with PEF treatment time.

Figure 2

The variation of the residual microbial density and of the disinfection efficiency with PEF treatment time.

The maximum disinfection efficiency of 99.84%, acquired during a 60-second treatment, is equivalent to a 2.80-order reduction. Disinfection efficiency increased with the increase of treatment time, however, the growth rate of the efficiency tended to saturate after 15 seconds. It follows that a treatment time of less than 10 seconds is preferable.

### Effect of pulsed electric field strength on disinfection efficiency

Figure 3 shows that, as the PEF strength rose from 4 kV/cm to 8 kV/cm, the disinfection efficiency increased from 97.2% to 99.47%. The PEF across cells might lead to the generation of irreversible pores in the membrane, resulting in the leakage of intracellular compounds and even in the death of the cells. With the increase of PEF strength, the probability of electroporation increases correspondingly. The cells of F. oxysporum could tend to be inactivated due to the high occurrence probability of irreversible pores. In practice, for an electric field of 5 kV/cm the resulting average disinfection efficiency was 98.28%.

Figure 3

The variation of the residual microbial density and of the disinfection efficiency with PEF strength.

Figure 3

The variation of the residual microbial density and of the disinfection efficiency with PEF strength.

### Influence of initial microbial density

As shown in Figure 4, disinfection efficiency stayed around an average percentage of 97.5% with a fluctuation of less than 1%. This indicates the initial pathogen population has little influence on the disinfection efficiency of the PEF treatment. Since the density of pathogens in actual soilless culture systems varies over a large range, this result implies that the PEF treatment could work well practically in most conditions.

Figure 4

Disinfection efficiency for different initial microbial densities.

Figure 4

Disinfection efficiency for different initial microbial densities.

### Temperature rise of nutrient solution caused by the PEF treatment

It can be observed from Table 1 that the temperature almost linearly increases with the treatment time. As expected, an increase in PEF strength increases the variation of the temperature rise. If no cooling systems are implemented, the temperature would keep rising with the treatment time.

Table 1

Temperature rise (°C) with PEF strength and time

Treatment TimePEF strength
5 kV/cm6 kV/cm7 kV/cm
0s 28 28 28
10s 32.2 33.4 35.2
20s 33.2 34.4 38.4
40s 41 41.4 45.2
60s 45.1 46 53.3
Treatment TimePEF strength
5 kV/cm6 kV/cm7 kV/cm
0s 28 28 28
10s 32.2 33.4 35.2
20s 33.2 34.4 38.4
40s 41 41.4 45.2
60s 45.1 46 53.3
The Joule's loss (Q, unit: J) and the temperature rise generated by a pulsed voltage can be calculated by:
(2)

(3)
where U (unit: V) is the peak of pulsed voltage, R (unit: Ω) is the resistance of the treatment chamber filled with nutrient solution, t (unit: s) represents the equivalent treatment time obtained by multiplying the pulse-number with the pulse-width, C (unit: J kg−1°C−1) is the heat capacity of the nutrient solution, m (unit: kg) is the mass of the treated liquid and finally ΔT (unit: °C) represents the temperature variation.

For a typical experiment: U= 6,000 V, R= 60 Ω, t= 4.5 × 10−4 s (45 pulses of 10 µs within 10 s), C = 4,185.5 J·kg−1· °C−1 (using the data of water), 7.1 ml nutrient solution in the chamber. It follows that Q = 270 J and ΔT = 8 °C. The corresponding temperature rise measured during the experiment (as shown in Table 1) was equal to 5.4 °C, close to the calculated value. The discrepancy is probably because the calculation did not consider the dissipation of heat. A more accurate calculation will require models based on finite element analysis (FEA).

In Table 1, for a peak pulsed voltage of 7 kV and a treatment duration of 60 seconds, the maximum temperature was 53 °C, corresponding to a temperature rise of about 25 °C. When the treatment time was less than 25 seconds the temperature did not exceed 40 °C, which is below 54 °C, the lethal temperature for F. oxysporum (Runia & Amsing 2001b). However, even under these conditions the disinfection efficiency was still high. Clearly this indicates that the PEF played the dominant role in the present disinfection experiments. This means that PEF treatment can be utilized as a non-thermal method for the disinfection of nutrient solution. However, cooling facilities will probably be required to be implemented in future practical PEF treatment systems, to control the temperature.

### Energy consumption of PEF treatment method

For a practical PEF technique, energy efficiency and energy density are also two important factors to be considered. Since kilowatt-hour (kWh) is commonly used as a billing unit (1 kWh = 3.6 MJ, $0.072/kWh in China), the energy efficiency is defined as the volume (unit: L) of treated nutrient solution per kilowatt-hour (L/kWh). In addition, the definition of energy density adopted here is the electric energy of PEF per millilitre of nutrient solution (J/ml). Table 2 shows the calculated injected electric energy, energy density, and energy efficiency for different parameters of the PEF treatment. The data in Table 2 indicate that energy consumption very much depends on the treatment conditions. The maximum energy efficiency in Table 2 is 322 L/kWh, for treatment parameters 4.6 kV/cm and 5 seconds. The design of future PEF systems need to consider the balance between two factors: the disinfection efficiency and the energy efficiency. On the condition that the disinfection efficiency is already high, it is reasonable to decrease both the treatment time and the PEF strength. Table 2 Calculated injected electric energy, energy density, energy efficiency and cost per unit for different parameters such as treatment time and applied voltage Treatment time (s)Applied voltage (kV)Injected energy (J)Energy density (J/ml)Energy efficiency (L/kWh)Cost per unit ($/m3)Disinfection efficiency (%)
4.6 79.35 11.17 322 0.22 95.38
10 120 16.90 213 0.34 97.2
10 4.6 158.7 22.35 161 0.45 97.5
10 270 38.03 94.67 0.77 98.4
10 367.5 51.76 69.55 1.04 99.05
10 480 67.61 53.25 1.36 99.47
15 4.6 238.05 33.53 107.37 0.67 98.27
15 405 57.04 63.11 1.15 98.24
Treatment time (s)Applied voltage (kV)Injected energy (J)Energy density (J/ml)Energy efficiency (L/kWh)Cost per unit (\$/m3)Disinfection efficiency (%)
4.6 79.35 11.17 322 0.22 95.38
10 120 16.90 213 0.34 97.2
10 4.6 158.7 22.35 161 0.45 97.5
10 270 38.03 94.67 0.77 98.4
10 367.5 51.76 69.55 1.04 99.05
10 480 67.61 53.25 1.36 99.47
15 4.6 238.05 33.53 107.37 0.67 98.27
15 405 57.04 63.11 1.15 98.24

## CONCLUSION

In this work, a PEF treatment method to disinfect recirculating nutrient solution of closed soilless culture systems has been investigated. A Hoagland solution inoculated with F. oxysporum was treated by the novel PEF treatment system. The results indicate that the PEF technique is effective in deactivating most of the F. oxysporum in the nutrient solution. The increase of treatment time and electric field strength enhanced the disinfection efficiency, respectively. The maximum disinfection efficiency of 99.84% was obtained under the treatment condition: 4.7 kV/cm, 60 s, 10 µs. Initial microbial density seems not to have an obvious effect on the disinfection efficiency. PEF played the dominant role in the deactivation of F. oxysporum. From a practical perspective, the recommended operating conditions for PEF treatment should be short treatment time (less than 10 seconds) and low electric field strength (around 5 kV/cm). The results of this study demonstrate that PEF treatment implies a future technological alternative for the disinfection of nutrient solutions.

## ACKNOWLEDGEMENTS

The study was funded by the National College Students Innovation and Training Program (The Ministry of Education of China, Grant No. 201610019065). Dr Jiankang Cao (College of Food Science & Nutritional Engineering, China Agricultural University) also provided much useful information about microbiological detection. We also thank Professor Bucur Novac (Loughborough University, UK), Dr David Herbert (Canada) and Ms Irene (Canada) for their helpful advice during the writing of the present paper.

## REFERENCES

REFERENCES
Barbosa-Cánovas
G. V.
,
Góngora-Nieto
M. M.
,
Pothakamury
U. R.
,
Swanson
B. G.
1999
Preservation of Foods with Pulsed Electric Fields.
,
London, UK
, pp.
1
155
.
Date
S.
,
Terabayashi
S.
,
Kobayashi
Y.
,
Fujime
Y.
2005
Effects of chloramines concentration in nutrient solution and exposure time on plant growth in hydroponically cultured lettuce
.
Sci. Hortic.
103
,
257
265
.
Ehret
D. L.
,
Alsanius
B.
,
Wohanka
W.
,
Menzies
J. G.
,
Utkhede
R.
2001
Disinfestation of recirculating nutrient solutions in greenhouse horticulture
.
Agronomie
21
,
323
339
.
Elmer
W. H.
2015
Management of Fusarium crown and root rot of asparagus
.
Crop Prot.
73
,
2
6
.
Gruyer
N.
,
Dorais
M.
,
Zagury
G. J.
,
Alsanius
B. W.
2013
Removal of plant pathogens from recycled greenhouse wastewater using constructed wetlands
.
Agric. Water Manag.
117
,
153
158
.
Hothem
S. D.
,
Marley
K. A.
,
Larson
R. A.
2003
Photochemistry in Hoagland's nutrient solution
.
J. Plant Nutr.
26
,
845
854
.
Kobayashi
F.
,
Ikeura
H.
,
Ohsato
S.
,
Goto
T.
,
Tamaki
M.
2011
Disinfection using ozone microbubbles to inactivate Fusarium oxysporum f. sp. melonis and Pectobacterium carotovorum subsp. carotovorum
.
Crop Prot.
30
(
11
),
1514
1518
.
López-Gálvez
F.
,
Sampers
I.
,
Gil
M. I.
,
Allende
A.
2017
Modelling of E. coli inactivation by chlorine dioxide in irrigation water
.
Agric. Water Manag.
192
,
98
102
.
Raudales
R. E.
,
Parke
J. L.
,
Guy
C. L.
,
Fisher
P. R.
2014
Control of waterborne microbes in irrigation: a review
.
Agric. Water Manag.
143
,
9
28
.
Runia
W. T.
,
Amsing
J. J.
2001a
Disinfection of recirculation water from closed cultivation systems by heat treatment
.
Acta Hortic.
548
,
215
222
.
Scarlett
K.
,
Collins
D.
,
Tesoriero
L.
,
Jewell
L.
,
van Ogtrop
F.
,
Daniel
R.
2016
Efficacy of chlorine, chlorine dioxide and ultraviolet radiation as disinfectants against plant pathogens in irrigation water
.
Eur. J. Plant Pathol.
145
,
27
38
.
Sharma
P.
,
Bremer
P.
,
Oey
I.
,
Everett
D. W.
2014
Bacterial inactivation in whole milk using pulsed electric field processing
.
Int. Dairy J.
35
,
49
56
.
Song
W.
,
Zhou
L.
,
Yang
C.
,
Cao
X.
,
Zhang
L.
,
Liu
X.
2004
Tomato Fusarium wilt and its chemical control strategies in a hydroponic system
.
Crop Prot.
23
,
243
247
.
Sutton
J. C.
,
Yu
H.
,
Grodzinski
B.
,
Johnstone
M.
2000
Relationships of ultraviolet radiation dose and inactivation of pathogen propagules in water and hydroponic nutrient solutions
.
Can. J. Plant Pathol.
22
,
300
309
.
Wohanka
W.
,
Luedtke
H.
,
Ahlers
H.
,
Luebke
M.
1999
Optimization of slow filtration as a means for disinfecting nutrient solutions
.
Acta Hortic.
481
,
539
544
.
Yiasoumi
W.
,
Evans
L.
,
Rogers
L.
2014
Farm Water Quality and Treatment
.
NSW Department of Primary Industries
,
Orange, NSW, Australia
.

## Author notes

These authors contributed equally to the work.