Phytoremediation of arsenic- and molybdenum-contaminated alkaline wastewater by Eleocharis acicularis in winter in Japan

Phytoremediation using aquatic plants is a sustainable, low-cost measure for remediating water contaminated by toxic heavy metals. In this study, we conducted a channel experiment using Eleocharis acicularis in heavy metal-contaminated mildly alkaline wastewater under unfavorable plant habitat conditions in winter in northeastern Japan. The wastewater from an embankment consisting of Neogene marine sediments had a temperature of 10-15 °C and a pH of about 9, and it contained ∼0.02 mg/L of As and ∼ 0.23 mg/L of Mo. About 16 kg (fresh weight) of E. acicularis was laid in a plastic channel measuring 30 cm in width by 20 m in length, and the channel was enclosed in a tunnel greenhouse. The experiment was conducted for the 3 months from November 2015 at an average flow rate of 0.3 L/min and an air temperature of -4 to 19 °C. No reductions in As or Mo concentrations in the outflow were detected. However, at 3 months, the E. acicularis showed accumulations of ∼7 mg/kg As and ∼18 mg/kg Mo as dry weight, indicating that this remediation method is workable in an unfavorable low-temperature, mildly alkaline environment.


Introduction
Toxic heavy metal elements, such as arsenic (As) and cadmium (Cd), in drinking water are a key issue of concern to human and animal health [1][2]. In light of this, a sustainable, low-cost remediation method is required for the treatment of contaminated wastewater that is drained from soil and rock excavated from building sites, tunnels and mines. Phytoremediation is method of remediating contaminated soil and water by using plants that are able to hyperaccumulate organic and inorganic contaminants, including heavy metal elements [3]. Phytoremediation is considered to be more sustainable than methods that use chemical media or physical removal and sealing of the excavation site [4]. Phytoremediation using aquatic plants has advantages in that some species can be grown in sized sheets in free-floating, submerged or emergent conditions, and can take up elements directly through their roots and leaves from flowing or pooled water [5]. The sized sheets are easy to install, replace and dispose of, making them convenient in terms of treatment site design and management. Some aquatic macrophytes can be grown in or on water without soil. This enables complex factors such as surface adsorption and desorption of heavy metal elements in a soil medium to be excluded, which is advantageous in simplifying the evaluation of element uptake efficiency from wastewater to aquatic macrophytes. However, a disadvantage is that the range of growth conditions favorable for aquatic plants in wastewater (e.g., chemistry, temperature and flow rate of wastewater) is more limited than in a soil environment and there is no buffering of pH and nutrients by soil material. Therefore, many case studies under various conditions of water and weather are needed before phytoremediation can be used for the various types of contaminated wastewater. This paper focuses on the phytoremediation of wastewater contaminated with heavy metals by using the aquatic macrophyte Eleocharis acicularis. This species is known as a hyperaccumulator of elemental heavy metals such as As, cupper (Cu), zinc (Zn), lead (Pb), indium (In), silver (Ag) and Cd [6][7][8][9][10] and it is also good at accumulating antimony (Sb) [11]. E. acicularis grows as an autochthon worldwide, including in northern Japan, which has snowfall and subzero air temperatures. It can grow in submerged, emergent and soil conditions in nature, and in sheets it can grow in a water pool with freely flowing water. Many studies using aquatic plants have demonstrated its effectiveness at accumulating toxic heavy metal elements in the field and in the laboratory [5]. To our knowledge, no studies have addressed the phytoremediation of actual wastewater flow in a channel with aquatic macrophytes under winter weather conditions, where the low temperatures and shorter day length would be unfavorable for plant growth and element uptake. The main objectives of the present study are to understand a phytoremediation efficiency in a 90-day experiment conducted in the winter from November 2015 to February 2016 using E. acicularis in mildly alkaline wastewater contaminated with As and molybdenum (Mo) in a flow-controlled channel at a wastewater treatment site in Japan at about 38° North latitude.

Site description and experimental design
In this study, an experiment channel was set up at a wastewater treatment site about 150 meters above sea level in Sendai City, northeastern Japan. At this site, an embankment mainly consisting of about 800,000 m 3 of Neogene (late Miocene to early Pliocene) marine sedimentary rocks steadily drains mildly alkaline wastewater contaminated with As at a flow rate of about 20 L/min. The wastewater from the embankment is first collected in a drainage tank before an As treatment plant.

Experimental design
The experiment channel, with a total length of 20 m, is set adjacent to the drainage tank. The experimental channel was constructed from two small channels that were 10 m in length (figure 1). The channels are made of 1 cm thick PVC resin plates with 30 cm in width and 10 cm in sidewall  height. The two channels were connected by hosepipe and placed with slope angle of 0.5° (figure 1b).
To protect plants in the channel from low temperatures and snowfall in winter, the channel was covered with a tunnel greenhouse of 0.2-mm-thick clear plastic sheeting and the sidewalls and bottom were covered with 20-mm-thick urethane foam (figure 1). The rate of inflow of the raw wastewater could be roughly controlled by faucet handle at the tank and was measured by an integrating flowmeter (NW05-NTN of Aich Tokei Denki Co., Ltd.) at the inflow point.

Sample preparation
Sheets of E. acicularis with a plant height of 10-15 cm, a width of 30 cm and a length of 60 cm were weighed for fresh weight after being hung out for 1-2 hours to dry off the water that was added for shipping. The fresh weight of each sheet ranged from 261 to 906 g, and the total fresh weight in the channel was 16.02 kg. After weighing, the sheets were laid in the channel so as to leave 50 cm free of plants at the uppermost and lowermost parts of each channel, in order to keep clearance for water flow. The area density of the E. acicularis was 29.74 kg/m 2 as fresh weight.

Sampling and measurement
During the channel experiment, 7 inflow and 17 outflow samples were collected using a PP-bottle that was rinsed with ultrapure water after being soaked in 1M nitric acid for at least a week. At sampling, the water temperature and the pH of inflow and outflow were measured by portable tester (HI98121 Hanna Inc.). As pre-treatment, 0.6 ml of nitric acid (69% HNO 3 for ultratrace analysis from Wako Pure Chemical Industries, Ltd.) was added to a 50-ml sample after filtration using a syringe filter with a pore diameter of 0.45 μm in the laboratory. Several pinches of E. acicularis were sampled from around 11.5 m from the upstream end of the channel 13 times during the experiment (table 3). The sample was dried at 50°C in a drying oven for at least 24 hours and then was measured for dry weight. The sample was powdered as whole plant using sample mixer (IKA A11 basic) with stainless steel blade. Acid digestion of E. acicularis was performed, according to the following steps. 0.1 g of dried powder was mixed with 5 ml ultrapuregrade HNO 3 and 2 ml H 2 O 2 in a closed, 100-ml PTFE vessel, and then the mixture was digested with the assistance of a microwave (Titan MPS PerkinElmer Co., Ltd.), with the sequence of 5 minutes at 150°C, 15 minutes at 190°C and 10 minutes at 50°C. The analytical solution was prepared by dilution up to 50 ml with ultrapure water.
The chemical components in the wastewater and the E. acicularis samples were analyzed using one of four methods: Inductively-coupled plasma mass spectrometry (ICP-MS), Inductively-coupled plasma atomic emission spectrometry (ICP-AES), atomic adsorption spectroscopy (AAS) and adsorption spectroscopy using molybdenum blue. Table 2 lists the method that was used for each element and sample type. The methods for water samples are subdivided into M1 and M2. The

Experimental conditions and plant growth
The change in air temperature outside of the channel from day 12 to day 90 is graphed in Figure 2a.
The outside temperature ranged from -8.7 to 14.1°C. On successive days between day 44 and day 51 and after day 56, the daily minimum air temperature was below zero at midnight and the daily mean was below 5°C. The air temperature inside the tunnel greenhouse in the upper channel ranged from -3.5 to 32.1°C from the start to day 43 (figure 2b). The tunnel greenhouse's contribution to the increase in inside air temperature can be estimated as +1.8 to +17.2°C from the daytime temperature difference between the inside and outside of the channel (figure 2b). The water temperature at the point of inflow during the 90 days was 11.6°C on average and ranged from 5.4-19.1°C. At the point of outflow, the range of water temperature was greater than at inflow. Between day 60 and day 69, the water temperature of outflow was almost 0°C. The total volume of flow into the experiment channel during the 90 days was 40,098 L. The flow rate decreased from about 2 L/min at day 1 to 0.5 L/min at day 43. The average flow rate during the experiment was 0.31 L/min (table 1 and figure 2d). Throughout the experiment period, the inflow showed no turbidity and the pH was roughly stable, at the mildly alkaline values of 9.0-9.3 (table 1). However, the treated wastewater showed pale green turbidity during the first several hours and turned transparent after that. The pH of the treated wastewater gradually changed from 6.11 at the start to 7.89 at day 43 (table 1). The recorded photos of E. acicularis growth at 16 m from the inflow point represent a lowering of water depth at day 61 and a further lowering at days 65-66, which is recorded as a decrease in flow rate between days 40 and 90   Mg, 86.2±4.1 mg/L for Na, 252±11 mg/L for Ca and 11.8±0.5 mg/L for K. The elements Al, Mn, Fe and Zn were almost below their detection thresholds (table 3). The treated wastewater immediately after the start (~day 0.3) shows notably higher concentrations of Cd, Zn, Mn, Fe, Pb, Sb, P and Al than the concentrations in the raw wastewater (table 3 and figure  4). These enrichments could be explained by mixing of groundwater and fertilizer attached to the E. acicularis, which were used for growing it in sheet form before the experiment. From day 1 to day 89, there were no noticeable differences in the element concentrations between the raw and treated wastewaters, except for Cd and for temporary decreases in As, Mo and Sb concentrations at day 12 and day 16. The concentration of Cd showed a rapid decrease from day 0.05 to day 0.3, and the decrease continued gently to day 8.

Chemical compositions of E. acicularis
Concentrations of As, Mo and Mn in E. acicularis during the experiment all show a remarkable linear increase from day 0 to day 63 and then a slight decrease from day 89 (table 3 and figure 4). From day 0 to day 63, As increased from 0.51 to 8.79 mg/kg dry wt., Mo from 4.82 to 26.62 mg/kg dry wt. and Mn from 85.9 to 884.9 mg/kg dry wt. Although Na also tended to increase from day 0 to day 63, it did not show a linear trend. The rate of increase gradually slowed toward day 63. The elements Sr and Ca increased slightly from 29.5 to 35.5 mg/kg dry wt. and from 4357 to 8058 mg/kg dry wt., respectively. The concentrations of Cd, Zn, Se and Sb decreased with time from 11.1 to 1.8 mg/kg dry wt., 150 to 110 mg/kg dry wt., 0.38 to 0.22 mg/kg dry wt. and 5.47 to 0.73 mg/kg in dry wt., respectively. The concentrations of K and P in plants are responsive to the addition or depletion of fertilizer in the growth environment. These elements increased slightly until day 8 and then decreased linearly. The variations of Fe and Al concentrations did not show any systematic tendencies.

Discussion
The nonsystematic decreases in As, Mo, Mn and Na concentrations observed in the sample of E. acicularis at day 89 suggests that the lowering of water level and flow rate at day 65 with low temperature from day 60 to day 66 caused to the plant shoots to wither in the channel. Therefore, this study evaluates the phytoremediation capability of E. acicularis for As, Mo and Mn, whose concentrations showed linear increases from the start to day 63 in the winter experiment. The bio-concentration factor (BCF), a factor for estimating phytoremediation potential, is defined as the ratio of the concentration of a target element in plant tissue to that in the growth solution or soil [12]. In the most environments, the BCF for solution is higher than those for soil due to the small portion of the element responsible for uptake from soil materials and pore water to plant root. Thus, using BCFs for the comparison of phytoremediation potentials between soil and solution environments should be carefully considered. The translocation factor, which is defined as the ratio of an element's concentration in the shoot to its concentration in the root, is also a key factor in evaluating of phytoremediation ability [13]. In this study, E. acicularis was analyzed in terms of the whole plant, including the roots. Thus, the BCF value in this study was calculated by using the concentration in the whole plant (BCF whole ). The preliminary measurement for a sample with shoots separated from roots shows a translocation factor of 0.74 for As, 1.18 for Mo and 1.77 for Mn. These translocation factors are near 1. Thus, BFC whole did not greatly differ from BCF for shoots and is comparable to the BCF obtained in a previous study.

Arsenic (As)
In our experiment from day 0 to 63, the As concentrations of E. acicularis increased from 0.51 to 8.79 mg/kg dry wt. in the wastewater flow with 0.026 mg/L and flow velocity about 0.31 L/min. The average uptake volume per day for As from the wastewater is calculated as 0.143 mg/kg dry wt./day at day 63. The BCF whole for As in this study is ranges from 18-85 up to day 42 and then increases to 346 at day 69. A previous channel experiment in air temperature of 24.6-31.9°C in summer using E. acicularis in the same type of sloped channel with a length of 50 m and three 100-L inflows, with temperature of 26.2-34.5°C, pH 7.5, flow volume of 0.4-0.6 L/min and As concentration of 0.234 mg/L, found an increase in As concentration of E. acicularis from 0.14 to 4.16 mg/kg dry wt. [14]. The BCF whole for solution were calculated from 0.37 to 17.8, for 2 days. The uptake rate for this 50 m channel experiment was calculated about 2 mg/kg dry wt./day, which is over 10 times higher than 0.145 mg/kg dry wt./day of this study. This corresponds to the increase in BCF whole with time in our 63 days study. The lower As uptake efficiency in this study is considered as the difference of growth condition In other case of soil environment, the two month cultivation of E. acicularis in the mine tailing site from September in Japan has been reported As concentration of 29.2-872 mg/kg dry wt. in shoots. These E. acicularis were grown in the contaminated sediment site with 188-1470 mg/kg dry wt. of As along a small river with flow velocity of 10 cm/s and As concentration of 0.039 mg/L. The BCF are estimated as 0.21-7.39 for the contaminated sediments and 1003-29966 for the river water. The BCF whole for As in our flow-controlled channel experiment was lesser than 10-100 times as high as the BCF for the river water in the soil environment. This suggests that As taken up from the As and Mo contaminated wastewater in the adverse growth season is accumulated less efficiently than the As take up from the solutions in favorable growth media and season. Previous study for three other As appear not to be caused by plant growth but by basal metabolism with stable uptake rate at equilibrium with water temperature, flow rate and element concentration of solution, because of the almost linearly increasing of As and Mo with time and As and Mo concentrations in plant tend to decrease after shoots change to non-green colored. As a future experimental design, the estimation of element uptake efficiency of E.acicularis should be investigated by the comparison with the water and air temperature and growth rate, with or without several types of fertilizer.

Conclusions
The 90-day monitoring results of weather, water temperature, flow volume and chemical compositions in water and E. Acicularis showed a relationship between the temperature of the wastewater flow and the uptake of heavy metal elements. While the temperatures of the inside air and the water flow in the channel were continuously kept above 0°C by the tunnel greenhouse in the channel, the bioconcentration factors of As and Mo were found exceed 1. This demonstrates that phytoremediation using E. acicularis is workable for the elements As and Mo in moderately contaminated wastewater flow, even in low-temperature conditions that are inhospitable to aquatic macrophytes.