3.1 Control Column
Before conducting main reactor runs, several blank experiments were conducted in order to determine the reduction route of MTBE through the mechanisms other than ZVI chemical reactions. The control runs were performed with sand alone column in the absence of ZVI with a constant concentration of MTBE at neutral pH. The data in Fig. 2 presents that there is little adsorption of MTBE at different HLR of 0.23, 0.57 and 1.14 m3/m2.d (equal to flow rate of 0.2, 0.5 and 1 ml/min) in the column led to 16.97, 8.12 and 3.83 removal percent, respectively.
3.2 Batch Experiments
Figure 3 shows the effect of retention time in MBTE removal in a batch system. The amount of MBTE removal in time periods of up to 60 min by weight values of ZVI in the range of 2 to 40 g/L was investigated. At retention time of 30 min the maximum removal of 58% was obtained. Also, with increasing retention time to 60 min, MTBE removal found a constant trend. While, the findings of other study have presented that with increasing of retention time, removal percent of halogenated aliphatic hydrocarbons are increasing. Hence, at a retention time of about 100 h with ZVI powder in the batch system, the residual concentration of organic matter fell nearly to zero (Gillham and O'Hannesin 1994). Another study by using ZVI powder showed that 45% of MTBE removed within 24 h in a batch system (Russo et al. 2015). The batch experiments demonstrate that the removal of MTBE is affected significantly by pH. As shown in Fig. 4, the fastest rate of MTBE degraded occurs within 30 min under acidic pH values (degree of conversion obtained 66.34% for pH = 5 but 27.29% for pH = 9). Although little decreasing was observed in neutral pH value. The results suggest that with increasing pH, the removal percent decreases dramatically. This provides further support to earlier work that MTBE in the presence of ZVI can be significantly removed in the acidic pH in columns filled with modified natural zeolites (Russo et al. 2015).
3.3 Column Experiments
3.3.1 Effect of Initial MTBE Concentrations
The disappearance of MTBE at different initial concentrations (Co = 0.5-5 mg/L) and neutral pH are compared in Fig. 5. The best results were obtained at low initial concentration, where that 80.76% of MTBE removed in less than 22.4 h reaction time. The removal of MTBE decreased with increasing Co under the conditions studied; however, the total amount of MTBE degraded actually increased in the PRB system. In general, the ZVI can remove the MTBE by reduction via iron or electron relocation at the superficial of corrosion intermediates, adsorption, or precipitation (Obiri-Nyarko et al. 2014, Liu et al. 2015).
Similarly, the results of other study for MTBE removal with using zeolite granule in batch and column systems showed that in less initial concentration, the removal percent of MTBE increases, because in high concentrations larger pore is needed for obtaining more removal efficiency (Abu-Lail et al. 2010). This result is in agreement with those reported by other researchers (Levchuk et al. 2014). The application of a PRB system filled with natural pyrite to treat a Cr (VI)-contaminated groundwater have shown the same results as other researchers (Noubactep 2008). Similar results presented with heavy metals removal in ZVI/pumice permeable reactive barriers. Increasing initial concentration of Ni and Cu led to decrease efficiency of PRB system. As well as increasing of ratio ZVI/Pumice, increased removal of Ni and Cu (Chen et al. 2010).
3.3.2 Effect of HLR
It is clearly shown in Fig. 6 that the performance of PRB system is dependent of HLR and significant effect was observed at this operating condition. The MTBE removals in HLR of 0.23, 0.57 and 1.14 m3/m2.d were 90.32%, 77.31% and 69.47%, respectively. This is due to the fact that the motion of MTBE is speed up with an increase in the flow rate, which could cause inadequate retention time of MTBE in the PRB column (Zhang et al. 2019). The obtained result showed that with increasing the amount of inflow, system efficiency decreases. This provides further support to earlier work that perchlorate ethylene (PCE) removal from water by PRB with ZVI media can be obtained by reducing the amount of inflow to the column and with a flow of 2 ml/min PCE removal of 93% has achieved (Eslami et al. 2015). Similarly, the findings of other study for MTBE removal using of zeolite in batch and column systems concluded that removal amount of MTBE increases in low flow rates. With considering the flow rate of 5.2 ml/min, removal of 95% has been obtained (Abu-Lail et al. 2010).
3.3.3 Effect of pH
The pH of the solution prevails the degradation process due to the strongly pH dependent of the properties such as ZVI surface charge condition and dissociation of the solution. As shown in Fig. 7, the most rate of MTBE vanished take places within 22.4 h under different pH values (degree of removal obtains 93.4% for pH 5 but 90.32% for pH 7 and 61.35% for pH 9). The removal rate increase in lower pH value due to ZVI stability is less disturbed. The main reason to explain this behavior may be the surface reaction led to the corrosion of ZVI, which generates hydrogen gas and hydroxyl ion according to following equations:
Fe0 → Fe2+ + 2e− (1)
2H2O + 2e− →2OH− + H2 (2)
These equations recommend that iron corrosion could have a destructive effect on the removal of MTBE as water competes with MTBE for the electrons from ZVI. At higher pH, the ferrous and hydroxyl ions form ferrous hydroxide and precipitate. The deposition of Fe(III) on ZVI could impede the movement of the MTBE and clog the reactive spots on ZVI, and hence reduce the overall performance. As the solubility of Fe(III) is strongly dependent on pH, content of Fe(III) on the surface of ZVI is controlled by the pH. Therefore, pH could affect the rate of iron corrosion and the amount of precipitation on ZVI exterior. Similarly, the results of a study on the effects of pH in the range of 1.7 to 10 on dechlorination of TCE by ZVI declared that pH 4.9 had the most efficiency. No observed any removal at pH 9 and 10 and with increasing pH, the removal efficiency had dramatic reduction (Chen et al. 2001). This result is in agreement with those reported by other works (Noubactep 2008, Levchuk et al. 2014). Also, long-term efficiency of ZVI barriers during clean-up of copper containing solutions has illustrated that increasing of pH values resulted to decrease in copper solution removal efficiency (Komnitsas et al. 2007). This gives verification of the amended efficiency of ZVI/sand integrations.
3.3.4 Effect of Nitrate in the Durability of PRB System
As shown in Fig. 8, more than 90% MTBE removal could be obtained at 22.4 h in the absence of nitrate. The initial MTBE concentration was kept constant at 1 mg/L, while the HLR was 0.23 m3/m2.d. The being of high nitrate concentrations had a substantial effect on the removal rate. The MTBE removal in nitrate concentration of 30, 60 and 180 mg/L were 64.15%, 52.75% and 44.61%, respectively. These results proved that the presence of nitrate as disturbing factor, decreases MTBE removal. This provides further backing to the earlier supposition that one of the problems associated with ZVI particles is the enhancement of reactivity of these particles with reduction of selectivity of pollutant removal from water that caused ZVI reaction with non-target materials (disturbing ions) and consequently caused to reduce efficiency of target pollutant removal in water (Elliott and Zhang 2001). To elucidate this decline in target pollutant removal, it was described that the active sites of nZVI surface have a positive charge state originally; thus, a contest between various negatively charged anions such as nitrate, phosphate and sulfate ions involves with in overlaying these surface active sites, resulting in a drop in removal efficiency (Khalil et al. 2018).
3.3.5 Effect of Hardness in the Durability of PRB System
Figure 9 exhibits the efficiency of the PRB column in MTBE removal in the presence of calcium hardness as a nuisance factor. All of operating conditions were the same as the nitrate. The MTBE removal with hardness concentrations of 100, 120 and 200 mg/L were 66.66%, 57.75% and 51.4%, respectively. The results show that hardness as an interfering factor reduces MTBE removal. Similarly, the findings of another study illustrated that the presence of hardness ions caused reduction of Cr removal from water to 33% by ZVI filled column. They have proven that the existence of calcium, magnesium and carbonate ions in groundwater have vast impact on the ZVI by development of inactive deposits, such as CaCO3 and Mg(OH)2, on the ZVI surface resulting in a decreased durability of the ZVI by preventing electron conduction (Lo et al. 2006). This result is in agreement with those reported by other works (Khalil et al. 2018). Similarly, the other study has investigated the dechlorination of TCE under different quality of groundwater and anaerobic environments. For the two water types tested, dechlorination of TCE was a little superior for the soft and low alkalinity groundwater than for the hard and high alkalinity groundwater. The results recommend that the mixture of groundwater is presumably to strongly change the capability of ZVI to reduce TCE (Cundy et al. 2008).
3.3.6 Reaction Kinetics
The main factor in the column kinetic study is the retention time defined as the time needed for water and the filling media inside the PRB to be in contiguity to get treatment aims. The retention time tR, of the pseudo-first-order reaction, can be calculated using Eq. (3) (Obiri-Nyarko et al. 2014).
tR=[-ln(Ct/C0)/k] (3)
where Ct is the MTBE concentration down-gradient of the PRB; C0 is the concentration of the MTBE entering the PRB; and k is the rate of reaction. Hence, since the flow rate and the initial porosity of the reactive material are known, the distances through the column can be easily converted to retention times. The reduction of MTBE as a function of residence time in the PRB system is presented in Fig. 10. In order to evaluate of MTBE removal from the PRB system in the presence and absence of interfering factors (i.e. nitrate and calcium hardness), the reaction rate constants were compared.
As can be seen in Fig. 10, the results revealed that the both of disturbing factors have a negative effect on MTBE removal rates. As predicted, lack of interfering factors had a considerable effect on the rate of MTBE degradation with rate constant equal to 0.0741 1/min that is approximately 2.65 and 4.11 times higher than in the presence of calcium hardness and nitrate, respectively. This is consistent with the findings of other studies (Zhang et al. 2010, Zhang et al. 2017). Reduction of nitrate by ZVI has fitted well the pseudo-first order rate constant depending on nitrate concentration (Wang et al. 2018). The k of reductive denitrification of nitrate was reported to be 0.086 1/min with nanoscale iron dosage of 1.0 g/L and pH in 6.7 (Zhang et al. 2010). Denitrification by ZVI is primarily depends on pH, nitrate concentration and, surface-mediated process (Zhang et al. 2017).