Salt removal of greenhouse soils using electrokinetic technology

Excess nutrients are easily accumulated in greenhouse soils due to the interception of rainfall by plastic film and repeated over-application of compost and fertilizers. This study was conducted to evaluate the application of electrokinetic (E/K) technology for salt removal from soils with high electrical conductivity (EC) in greenhouses. Three types of soil in plastic film greenhouses were used: artificial soil (Site A), poorly drained soil (Site B), and well-drained soil (Site C). The salt-enriched surface soils were used to fill 37-cm-long-box-type E/K cells, and a constant voltage gradient was applied at a rate of 1 V cm−1 for 30 days. The decrease in soil EC was achieved with water content of greater than 30% for silt loam (Sites A and B) and 20% for sandy loam (Site C). The E/K technology decreased soil EC by more than 80%, with a greater reduction ratio for sandy loam than for silt loam. After 30 days of the E/K treatment, 98–99% of NO3-N and 95–99% of sodium were removed in all three sites under saturated condition, implying that nitrate and sodium ions had higher mobility than the other ions during the E/K process. The results obtained from the study suggested, therefore, that the E/K technology is highly efficient for wet and nitrate-enriched sandy loam soils, and the technology can be a feasible and environmentally sound practice for the removal of excessive nutrients in greenhouse soils without water pollution by nutrients such as nitrate as can be caused by flooding and repeated washing with water.


Introduction
Farmers can increase crop productivity and economic profit using plastic film greenhouses and through multiple cropping in a year, as well as by regulation of growing conditions such as soil moisture, temperature, and air humidity [1,2]. In terms of soil environment, the interception of rainfall and over-application of compost and fertilizers result in excess nutrients remaining in the greenhouse soil after crop harvest, consequently causing salt accumulation and high electrical conductivity (EC). The average content of available phosphate (P 2 O 5 ) and exchangeable potassium (K) of 1334 greenhouse soils in Korea in 2012 were 1049 mg kg -1 and 1.58 cmol ? kg -1 , respectively, higher than the optimum levels of 300-550 mg kg -1 and 0.5-0.8 cmol ? kg -1 [1]. As a result, soil EC was 3.1 dS m -1 , higher than the optimum level of less than 2 dS m -1 . The excess nutrients in soil eventually cause reductions in crop productivity, as well as quality deterioration through water uptake depression due to osmosis, imbalanced nutrition [2,3], and accumulation of NO 3 -N in crops [4,5].
To reduce excess nutrient levels in greenhouse soils, numerous agricultural practices have been developed and disseminated to farmers, such as the cultivation of crops with a high nutrient absorption ability, the addition of low-fertility soil, rotation cropping, flooding, and high-ridge establishment [2,[6][7][8][9][10][11]. However, adoption of a rotation cropping system or the cultivation of crops with a high nutrient absorption ability, such as maize, may reduce profitability, compared with cash crops such as vegetables. Application of low-fertility soil, or the removal of surface soil with high EC, may be a temporary solution and may also require labor and transportation. In addition, flooding and repeated washing with water can involve the consumption of large volumes of water and can cause contamination of water resources with nutrients such as nitrate and phosphorus. Therefore, an environment-friendly technology needs to be developed to reduce salt concentration and soil EC in greenhouse soil, without causing environment pollution.
Electrokinetic (E/K) technology has been widely applied to remediation of contaminated soils since Puri and Anand [12] used it to remove exchangeable bases from alkaline soils. It has been used on soils contaminated with heavy metals, nutrients, or organic pollutants, and also to purify water of contaminants such as radioactive materials, oils, organic pollutants, and heavy metals [13][14][15][16][17][18][19][20][21][22][23][24][25]. When a low-intensity direct current is applied to soil between anode and cathode, charged species produced by electrolysis [14,26] move toward cathode or anode through E/K processes such as electroosmosis (water transport from anode to cathode [27][28][29]), electromigration (ion transport to the opposite electrode [14,[30][31][32][33]), and electrophoresis [34,35]. The E/K technology can be used to effectively remove pollutants even in soils with low hydraulic conductivity [36,37]. Thevanayagam and Wang [38] showed that cations including heavy metals around the cathode can be removed through deposition, with OHproduced at cathode. The objective of this study was to evaluate the application of E/K technology in the removal of salts (excess nutrients) from greenhouse soils in Korea.

Characteristics of greenhouse soils
Three types of plastic film greenhouses were selected for the study: artificial soil (Site A), poorly drained soil (Site B), and well-drained soil (Site C). Site A was located in Jinjeop-up, Namyangju-si, Gyeonggi-do, Korea; Site B in Daejeo2-dong, Gangseo-gu, Busan-si, Korea; and Site C in Gangdong-dong, Gangseo-gu, Busan-si, Korea. The soil series of Site B was Gimhae, and Site C was Nagdong. The soil series for indigenous soil of Site A was Deogcheon. The soil textures of Sites A and B were silt loam, and that at Site C was sandy loam. Soil EC and soluble ions in soil from surface to 100 cm depth were measured for soil samples taking from the three sites. The distribution of soil EC, soluble cations, and soluble anions with depth is shown in Fig. 1 for Site A, Fig. 2 for Site B, and Fig. 3 for Site C. Soil EC was high at 0-10 cm depth for the three sites, and decreased with soil depth except at Site A where another EC peak at occurred at 70 cm, the boundary of artificial and indigenous soils. Site C, which had sandy textured soil and no compacted zone, had the thinnest upper salt accumulated zone and lowest EC values below 10 cm depth of soil. The main soluble ions were Ca 2? and SO 4 2for Site A, Na ? and Clfor Site B, and Ca 2? and NO 3 for Site C. Soil chemical properties were determined using the soil analysis method devised by the National Institute of Agricultural Science and Technology [39]. Briefly, soil pH and EC were measured after mixing soil with H 2 O (1:5). After EC was measured, soluble ions were analyzed with inductively coupled plasma spectrophotometer (CINTRA6, GBC Scientific Equipment, Braeside, Victoria, Australia) and ion chromatography (DX-600, Dionex, Themo Fisher Scientific, Waltham, MA, USA) using the water extracts. Soil bulk density was determined using a core sampler.

Electrokinetic (E/K) experiments
The salt-enriched soils from upper part of each site were used for establishing the methodology of E/K technology for salt removal. The 37-cm-long-box-type cells were filled with surface soil (5 cm depth for Site A, and 10 cm for Sites B and C) from the three sites (Fig. 4). The soil water content in each cell of the E/K apparatus was set to be different, ranging from the initial state to a saturated state. A constant voltage gradient was applied at a rate of 1 V cm -1 for 30 days. Soil EC and the concentration of soluble ions were determined for soils taken from five spots of each cell.
Column experiments Column experiments were performed under saturated conditions, using soils from Sites B and C. Columns with a radius of 3.63 cm and height of 3.63 cm were filled with 360 g of dried soil. The bulk density of the soil was 1.20 Mg m -3 , and the saturated hydraulic conductivity was 0.06 cm h -1 for Site B and 1.49 cm h -1 for Site C. The concentrations of soluble ions and EC were measured for each effluent fraction of total five pore volume. The mobile ion ratio was obtained by comparing the total amount of each ion in effluents with the initial total amount of the soluble ion.

Results and discussion
The physical and chemical characteristics of the soils used in E/K experiments are shown in Tables 1 and 2, respectively. The initial soil moisture content was 17.3% for Site A, 21.1% for Site B, and 6.3% for Site C. After applying an electrokinetic current, the soil pH decreased in the anode compartments to almost pH 3, whereas it increased in cathode compartments up to almost pH 12, especially under high water content conditions, when compared with the initial pH (Fig. 5). H ? is produced through water oxidation at the anode, whereas OHis produced through water reduction at the cathode [16,18,19,21,23,24,40,41]. The soil EC for Site A decreased, especially for water contents greater than 30%, compared with an initial EC of 3.6 dS m -1 (Fig. 6). For Site B, the soil EC decreased, except for a water content of 21%, compared with the initial EC of 15.0 dS m -1 (Fig. 7). For Site C, the soil EC decreased, especially for water contents of 30 and 49%, compared with the initial EC of 11.2 dS m -1 (Fig. 8). The concentrations of soluble ions for the three sites are listed in Tables 3, 4, and 5. The concentrations after 30 days of E/K treatment decreased, probably through electromigration, diffusion, and electroosmosis.
The reduction of soluble ions ranged from 74 to 99% in high soil water content, based on the contents of soluble ions at spots 2-4 except around the anode and cathode ( Table 6). The highest reduction was observed for NO 3 After 30 days of the E/K treatment, 98-99% of the initial NO 3 -N was eliminated for the three sites under saturated water condition. Cho et al. [40] reported that 86% of NO 3 -N was removed after only 8 days of E/K treatment of a sulfate-accumulated saline greenhouse soil, whereas it took 16 days to remove   70% of sulfates, due to their high affinity for soil organic matter, and their tendency to precipitate with Ca 2? . Lee et al. [24] also reported that nitrate removal efficiency was greater than 90% from a greenhouse soil with a dimension of 5 by 6 m through in situ E/K process. Manokararajah and Ranjan [25] suggested that the main mechanisms for nitrate removal are electromigration toward the anode and denitrification to nitrogen gas. Choi et al. [16] conducted in situ E/K process for a greenhouse soil with a 3 m by 2 m plot and reported that removal efficiency of nitrate and chloride was greater than 80% after 30 days of the treatment, due to the easier desorption from soil particles and the higher solubility in the pore water than for sulfate. In addition, sodium showed the highest removal efficiency among the cations because of its low adsorption strength, although sodium has the lowest value of diffusion and ionic valence. Calcium and magnesium are relatively easy to adsorb to soil particles, because of their high adsorption strength and their ability to precipitate with anions such as sulfate and hydroxide [16]. Cho et al. [40] reported potassium removal of 46-68%, and magnesium removal of up to 70%, after 16 days of E/K treatment. Lee et al. [24] reported that an in situ E/K process removed 93% of     N  S   21  1  2497  522  1164  1650  334  759   2  1339  422  801  1958  266  879   3  1322  445  832  2016  244  935   4  1289  451  831  2023  245  894   5  969  436  624  2065  217  927   30  1  921  475  761  39  485  429   2  1608  317  410  33  312  912   3  1263  206  266  37  16  1347   4  1495  190  395  58  14  1672   5  132  424  14  2178  4  potassium after 26 days. Site C, which had sandy loam soil, showed higher reductions than the sites with silt loam, especially for calcium and magnesium, which are more resilient to the E/K treatment. Among the silt loam soils, a higher salt reduction was shown at saturated water content than for unsaturated conditions. For the column experiment, Fig. 9 shows changes in EC and ions in effluents as a function of pore volume. Concentrations of ions in effluents decreased to almost 0 with a pore volume of 3 for Site B with silt loam, and 1.5 for Site C with sandy loam. Nitrate and chloride showed relatively fast movement before one pore volume.
Mobile ion ratios, calculated by the total amount of each ion in effluents divided by the initial total amount of the soluble ion for column experiment, were compared with reduction percentages for the E/K experiment (Table 7). Mobile ion ratios for Site B (silt loam) ranged from 106 to 167%, greater than the reduction ratios of 85-99%. For Site C (sandy loam), E/K reduction ratios ranged from 96 to 98%, greater than the mobile ion ratios of 62-91%. Mobile ion ratios for Site B were greater than those for Site C, possibly because of the site's low hydraulic conductivity and long reaction time. Please note that saturated hydraulic conductivity for Site C was about 25 times greater than that for Site B. The results of this study indicated that the efficiency of the E/K technology is greater than that of flooding and repeated washing method for sandy loam soil.
However, electrical energy consumption can be one of the main barriers for the application of E/K process in agriculture, because it accounts for 25% of the operating cost of the technology [37]. Jo et al. [21] reported that a pulsed E/K process reduced the electrical energy consumption by 42% compared with conventional processes for the restoration of greenhouse soil. Moreover, the technology can be integrated with other remediation processes such as bioremediation, permeable reactive barriers, phosphate amendment, and phytoremediation [15,17,42,43].
This study demonstrated the applicability of E/K technology to reduce excess nutrients in greenhouse soils in an environmentally friendly manner. The E/K technology decreased soil EC by more than 80% for silt loam with a water content of greater than 30% (Sites A and B), and 20% for sandy loam (Site C). After 30 days of the E/K treatment, 98-99% of NO 3 -N and 95-99% of sodium were   removed. Therefore, E/K technology can provide a feasible and environmentally sound practice for the removal of excess nutrients from greenhouse soils. This is achieved without causing water pollution by nutrients such as nitrates, as occurs when flooding and repeated washing with water is used, especially for wet and nitrate-enriched sandy loam soils. It should be noticed that the E/K technology may enhance emissions of nitrous oxide, one of the greenhouse gases, during denitrification under wet soil conditions. Further studies are required to investigate ways to increase the salt removal efficiency and to decrease the cost and the treatment period. Nonetheless, E/K technology is currently already feasible as a method for the reduction of EC in greenhouse soils.