Heavy Metals Removal from Simulated Wastewater using Horizontal Subsurface Constructed Wetland

This study aimed to assess the efficiency of Nerium oleander in removing three different metals (Cd, Cu, and Ni) from simulated wastewater using horizontal subsurface flow constructed wetland (HSSF-CW) system. The HSSFCW pilot scale was operated at two hydraulic retention times (HRTs) of 4 and 7 days, filled with a substrate layer of sand and gravel. The results indicated that the HSSF-CW had high removal efficiency of Cd and Cu. A higher HRT (7 days) resulted in greater removal efficiency reaching up to (99.3% Cd, 99.5% Cu, 86.3% Ni) compared to 4 days. The substrate played a significant role in removal of metals due to adsorption and precipitation. The N. oleander plant also showed a good tolerance to the uptake of Cd, Cu, and Ni ions from water. The highest removal of the heavy metals indicated that the HSSF-CW would be a promising technology for heavy metal contaminated wastewater as well as in electroplating and manufacturing industries.


INTRODUCTION
The environmental contamination with heavy metals is causing an increasing concern nowadays due to the industrial growth, agricultural practices, and transportation means. These metals include lead (Pb), chromium (Cr), cadmium (Cd), mercury (Hg), copper (Cu), nickel (Ni) and zinc (Zn). Cadmium, copper and nickel are among the major contaminants of water sources with heavy metals [Amarioarei et al., 2017]. Improper disposal of heavy metals into water bodies, even at low concentrations has become a major challenge due to their noxious impact on all creatures, persistence in nature and difficulty in being transformed by microorganisms; therefore, they accumulate in the ecosystem Constructed wetlands (CWs) are an alternative treatment method with special merits of being sustained, efficient, economically and environmentally friendly treatment that mimics the natural processes [Stefanakis, 2019; Said et al., 2020]. Researchers used diff erent aquatic and terrestrial plants in their pilot or fi eld studies to treat diff erent contaminants by various mechanisms such as containment or contaminants removal into the upper plants parts or turn them into less harmful forms . CWs are comprised of chemical, physical, and biological processes that infl uence each other, resulting in properly treated water. Many plants have high potential for successful uptake of contaminants. They have the ability to absorb and accumulate metals which are essential for their growth (Cu, Ni, Fe, Mn, Zn, and Mg) as well as non-essential metals (Cd, Cr, Pb, Co, Ag, Se, Hg).
In this study, a pilot scale horizontal subsurface fl ow constructed wetlands (HSSF-CWs) were investigated to remove Cd 2+ , Cu 2+ , and Ni 2+ ions from simulated wastewater. A Nerium oleander plant was used in this study due to its potential to remove heavy metals. Oleander is a native ornamental plant that is found in almost diff erent Iraqi regions. It is fast-growing, evergreen, has high biomass, withstands extreme temperature and is easily transplanted into most soils, although it grows well in poor, resistant to salt stress [Kumar et al., 2017]. This is an ever-green, non-edible, terrestrial plant which can endure the Iraqi climate. The main objectives of the study were to examine the performance of the HSSF-CW system in removing Cd 2+ , Cu +2 , and Ni 2+ from simulated water under the climatic condition of Iraq and to determine the eff ect of substrate in heavy metals retention.

Simulation of heavy metal contaminated wastewater
Cadmium nitrate Cd(NO 3 ) 2 .4H 2 O (Tetenal photowerk GmbH, Germany), copper nitrate Cu(NO 3 ) 2 .3H 2 O (BDH chemicals Ltd, UK), and nickel chloride Ni(Cl) 2 .6H 2 O (Tetenal photowerk GmbH, Germany), all with purities above 98% were to prepare simulated heavy metal contaminated wastewater by dissolving them with tap water. The concentrations of Cd 2+ , Cu 2+ , and Ni 2+ ions were calculated and adjusted at pH value of 5 to maintain the soluble cation forms.

Pilot-Scale Horizontal Sub-Surface Flow -Constructed Wetland
In this study, four planted glass basins were constructed outdoor and operated continuously with a dimension of 70×30×30 cm (LxWxD) and two diff erent hydraulic retention times (HRTs) of 4 and 7 days were examined. A slope of (1%) was used to maintain the hydraulic gradient and aspect ratio of 2.3:1. Figures 1 and 2 show the photo and schematic diagram of HSSF-CW, respectively. Four storage tanks (each with 30 L capacity) were connected to four control tanks (each with 8 L capacity) with an opening to drain the excess simulated water that was fed from storage tanks to obtain constant water elevation throughout the whole study. The outlet valve of each control tank was attached to a stopcock to control the water fl ow into each basin. The fl ow rates were measured using a stopwatch and measuring cylinders. Each HSSF-CW Figure 1. Photo of HSSF-CW was connected to a PVC pipe which controls water level within each bed as well as for periodic collection of samples from the effl uent, and a valve to empty the basins. Coarse gravel with a size of > 4 mm was added at 5 cm length from the inlet and outlet zones with depth of 20 cm in each basin in order to obtain a uniform distributed fl ow. Then, the remaining area of the basins was fi lled with 7 cm sand, 7 cm fi ne gravel and 6 cm coarse gravel. The oleanders used in this study had an average height of 25 cm and a root length ranging from 10-12 cm.
Hydraulic retention time (HRT) is the actual volume of wastewater in the bed divided by the average design fl ow. Many studies found that HRT is not suffi cient to achieve the required contaminants removal [EPA, 1993]. Many researchers suggested that it is more convenient to apply The total quantity of simulated wastewater used during each experiment was 45 L and was fed manually into each storage tank. Each basin was treating a volume of 5 and 9 L/d for 4-and 7-days HRT, respectively. The three CW were irrigated with Cd 2+ , Cu 2+ and Ni 2+ , respectively, while the fourth basin was irrigated with tap water as a control. Treated heavy metal contaminated wastewater was sampled and analyzed.

Heavy Metals Measurement in Simulated Wastewater
The wastewater samples were analyzed daily. The samples were fi ltered using (Whatman No. 45) fi lter paper for the estimation of heavy metals, collected in plastic test tube and then analyzed using atomic absorption spectrometry (ASS) (Sen-sAA GBC Scientifi c Equipment, Malaysia). The percentage of Cd 2+ , Cu 2+ and Ni 2+ removal on each sampling day was determined with Eq. (2): where: C o -concentration of heavy metal entering the HSSF-CW (mg/l); C i -concentration of heavy metal fl owing out of the HSSF-CW (mg/l).

Heavy Metals Measurement in Sand Samples
The sand was sampled at the beginning and at the end of each experiment during the fi rst and last day of the experiments. These samples were taken at diff erent points of 5 cm depths and mixed to ensure representative sampling. After that, the samples were air-dried and digested. The total (Cd 2+ , Cu 2+ , Ni 2+ ) concentrations in sand was extracted using the hot aqua regia digestion procedure (ISO standard 11466). Then, the concentration of heavy metals was analyzed using ASS.

Heavy Metals Measurement in Plant Tissues
The plant tissues were analyzed on the fi rst and last days of the experiments. Plant samples

Heavy Metals Removal from Water
The results of the Cd 2+ concentration during the 4 and 7 days HRT are shown in Figure 3 (a) and (b), respectively. At 4 days HRT, the Cd 2+ concentration in the effl uent ranged from 0.213 to 0.626 mg/l, which exceeds the permissible limits for local river conservation criteria (< 0.01 mg/l). The average removal effi ciency was 96.5%. At 7 days HRT experiment, the Cd 2+ concentrations in the effl uent were 1.59 mg/l at day-1, 0.17 at day-5 sampling, and decreased to meet the permissible limits starting from day-7 sampling to be 0.008 mg/l. The average removal effi ciency was 99.3% which was higher than the removal effi ciency at 4 days HRT.
The results of the Cu 2+ concentration at 4 and 7 days HRT in the effl uent are shown in Figure 4 (a) and (b), respectively. During the 4 days HRT, the Cu 2+ concentration at day 1 sampling was 3.63 mg/l, ending by 0.661 mg/l, which exceeds the permissible limits for local river conservation criteria (< 0.2 mg/l). The average removal effi ciency were taken from the beginning and the end of each basin at each sampling time. The dried plant samples were ground, and about 1 g of each part (root, stem and leaves) was treated with 10 mL of concentrated HNO 3 . The solution was placed on a hot plate for 30-45 min and left to cool. After that, 4 mL of 20% hydrogen peroxide (H 2 O 2 ) was added. The solution was reheated on a hot plate until it became semi-dried and clear. Then, the suspension was allowed to cool and was fi ltered into a 50 mL volumetric fl ask and diluted with distilled water [Rashid et al., 2016].

Chemical Parameters
The parameters of pH, dissolved oxygen (DO), and total dissolved solids (TDS) were measured to observe the potential changes in water. The pH in the water samples was measured by means of a HANNA pH meter model (HI 98103), TDS was measured using a HAN-NA conductivity tester model (HI 98301), and DO was measured using a HACH meter (HQ430d, USA).  was 96.3%. At 7 days HRT, the Cu 2+ concentrations were 1.01 mg/l at day-1, 0.3 mg/l at day-5, and decreased to meet the permissible limits starting from day-6 with 0.088 mg/l. The average removal efficiency was 99.5%.
The results of Ni 2+ concentration at 4 and 7 days HRT in the effluent are shown in Figure 5 (a) and (b), respectively. At 4 days HRT, the concentrations of Ni 2+ exceeded the permissible limits for local river conservation criteria (< 0.2 mg/l) and ranged from 15.1-20.5 mg/l. The average removal efficiency was 71%. At 7 days HRT, the Ni concentration ranged of 7.6 to 9.41 mg/l, and also exceeded the permissible limits with an average removal efficiency of 86.3%.

Regression Models for Cd 2+ , Cu 2+ , and Ni 2+ Removal in HSS-CW
The results of the regression analysis between the time (t) and effluent concentration (C) of the three metals are shown in Table 1. A square polynomial was used to calculate the concentrations of Cd 2+ , Cu 2+ , and Ni 2+ ions in the effluent water from the CW. The analysis showed a significant relationship between the variables t and C (R 2 > 0.9), and the quadratic equation can be used to predict the Cd 2+ , Cu 2+ , and Ni 2+ concentration in the effluents of simulated wastewater by time.

Heavy Metals in Sand
The results of sand analyses are summarized in Table 2. The heavy metals were high on the last day for both experiments compared to the first day. This could be attributed to the gypsiferous and calcareous characteristics of the sand, in addition to the neutral pH, the presence of clay content and -to a lesser extent -organic matter [Pappalardo et al., 2010]. Moreover, it was clearly shown that higher HRT resulted in higher heavy metals accumulation in soil [Gola et al., 2016].

Heavy Metals Concentration in Plant Tissues
The results of the plant tissues analyses are summarized in Table 3. The oleanders planted near  the HSSF-CW inlet showed higher concentrations accumulated in their tissues than those planted near the outlet. This could be attributed to the metal ions absorbed by the plant along with their flow through the HSSF-CW, in addition to the metal sorption capacity of the sand. At 7 days HRT, the oleander plant accumulated higher Cd 2+ , Cu 2+ , and Ni 2+ than at 4 days HRT. Moreover, it was obvious that oleander accumulated more Cd 2+ , Cu 2+ , and Ni 2+ at the last day of the experiments than the first day.

Monitoring of pH, DO and TDS
The pH, DO and TDS were recorded for each metal throughout the study and the profiles are shown in Figures 6, 7, and 8, respectively.
The DO ranged between 5.1 to 7.2 mg/l for Cd 2+ , 4.3 to 6.7 mg/l for Cu 2+ , 3.6 to 6.4 mg/l for Ni 2+ , and 6.1 to 7.4 mg/l for control during the 4 days HRT. At 7 days HRT, the DO ranged between 3.4 to 6.2 mg/l for Cd 2+ , 3.3 to 5.4 mg/l for Cu 2+ , 4.1 to 5.8 mg/l for Ni 2+ , and 4.3 to 7.2 mg/l for control. According to Podedworna and Zubrowska-Sudol [2010], these

CONCLUSIONS
Higher removal efficiencies from simulated wastewater were observed in 7 days HRT with an average removal efficiency of 99.3, 99.5, and 86.3% for Cd 2+ , Cu 2+ , and Ni 2+ , respectively. At 4 days HRT, the removal efficiencies were only 96.5 (Cd 2+ ), 96.3 (Cu 2+ ), and 71% (Ni 2+ ). This indicated the essential role of HRT on the treatment efficiency of heavy metals in HSSF-CWs. The HSSF-CW was able to reduce the Cd 2+ and Cu 2+ concentrations to meet the permissible limits at 7 days HRT, while increasing HRT has a slight contribution on the removal of Ni 2+ ions and it is obvious that the metal concentrations were higher than the permissible limits at both HRTs.