Heavy metal contamination in underground water commonly occurs in industrial areas in Taiwan. Wine-processing waste sludge (WPWS) can adsorb and remove several toxic metals from aqueous solutions. In this study, WPWS particles were used to construct a permeable reactive barrier (PRB) for the remediation of a contaminant plume comprising HCrO4, Cu2+, Zn2+, Ni2+, Cd2+, and AsO33− in a simulated aquifer. This PRB effectively prevented the dispersals of Cu2+, Zn2+, and HCrO4, and their concentrations in the pore water behind the barrier declined below the control standard levels. However, the PRB failed to prevent the diffusion of Ni2+, Cd2+, and AsO33−, and their concentrations were occasionally higher than the control standard levels. However, 18% to 45% of As, 84% to 93% of Cd, and 16% to 77% of Ni were removed by the barrier. Ni ions showed less adsorption on the fine sand layer because of the layer's ineffectiveness in multiple competitive adsorptions. Therefore, the ions infiltrated the barrier at a high concentration, which increased the loading for the barrier blocking. The blocking efficiency was related to the degree of adsorption of heavy metals in the sand layer and the results of their competitive adsorption.

  • A wine-processing waste sludge (WPWS) permeable reactive barrier (PRB) successfully prevented the spread of HCrO4, Cu2+, and Zn2+.

  • WPWS PRB had an extremely high affinity to HCrO4.

  • WPWS PRB had a low affinity to AsO33−.

  • Multiple competitive adsorption results showed WPWS PRB as a poor block to Ni2+.

  • Adsorption from sand layer to metals also affected the block efficiency.

The Environmental Protection Administration, Executive Yuan, discovers more than ten toxic-metal contamination sites each year through routine investigations of groundwater in all large-scale industrial districts of Taiwan. Spontaneous seepage of chemicals and illegal disposal of industrial wastewater into the soil are considered to be the major causes for these pollution incidents. Factories and manufacturing units commonly utilize large metallic tanks to store chemicals. Complicated pipeline systems have been constructed to transfer such chemicals. Once cathodic corrosion occurs (Okyere 2019), the chemicals, containing high concentration of heavy metals, can seep from the underground tanks or pipelines and probably reach the aquifer within a short time period, as long as they flow along the soil cracks (Hillel 1999). However, negatively charged soil minerals can strongly adsorb the transition elements; that is, Cr3+, Cu2+, Zn2+, Ni2+, and Cd2+, and inhibit their mobilization (Logan 1999). Heavy metal anions, such as HCrO4 and AsO33−, are more easily leached into the aquifer, thereby causing greater harm to the environment (Ahuja 2008; Komárek et al. 2013; Chen et al. 2019). Contamination of groundwater with heavy metals is usually caused by numerous industrial applications, such as the production of alloys and stainless steel (Yuan et al. 2014), electroplating (Wei et al. 2013), leather tanning (Nur-E-Alam et al. 2020), pigmentation, and wood preservation (Rent et al. 2017). Additionally, in the United States, more than 90% of the total arsenic consumption is for agricultural purposes. This includes the production of wood preservatives (74% of total), herbicides, insecticides, fungicides, desiccants, antiparasitic medications, and growth stimulants for plants and animals (Agency for Toxic Substances and Disease Registry 2007). In Taiwan, arsenic pollution in groundwater was only caused by an illegal dumping and burial from a specific pesticide manufacturing plant.

Toxicity of heavy metals is attributed to their good solubility in water, long biological half-lives, and propensity towards bioaccumulation. Accumulation of heavy metals in different body tissues has the potential to affect vital organs such as kidneys, bones, and liver, and can cause serious health hazards (Wu 2020). The threshold levels for Cd2+, HCrO4 and AsO33− in standards for drinking water are very low compared to some other elements because these compounds are highly toxic to living organisms. They have been classified as a Group A human carcinogen by the U.S. Environmental Protection Agency (Dhal et al. 2013; Kim et al. 2015).

In summary, heavy metal ions are hazardous to human health and the environment, even at low concentrations. Thus, effective remediation technologies to reduce the risk of spreading contaminated water plumes throughout aquifers and to decrease negative effects on the ecosystem are in high demand (Kumarasinghe et al. 2018). The conventional pump-and-treat method combining groundwater extraction and ex-situ treatment is widely used in remediating contaminated groundwater, but this option is expensive and often ineffective in meeting water quality standards (Maamoun et al. 2020).

Permeable reactive barriers (PRBs) have received considerable attention as a practical and cost-effective methodology for the in-situ treatment of contaminated groundwater in recent years (Obiri-Nyarko et al. 2014). Using natural water flow, the installation of a PRB downstream of the contaminant plume enables the trapping of pollutants and minimizes the spread of pollutants. The major advantage of PRB technology is that it utilizes natural groundwater flow to transport pollutants toward the reactive materials without requirements for an energy input or above-ground facilities (Gavaskar 1999; Snape et al. 2001). An important consideration in PRB design is the selection of the reactive material. Various adsorbent media, including zero-valent iron (Henderson & Demond 2007; Chen et al. 2011; Singh & Singh 2018; Hu et al.,2019b), zero-valent aluminum (Han et al. 2016), zeolite (Statham et al. 2016), fly ash (Czurda & Haus 2002), biochar (Hu et al. 2019a), activated carbon (Liu et al. 2012b; Singh & Singh 2018), waste green sand, peat (Guerin et al. 2002; Erto et al. 2011), and even mixed adsorbents (Kumarasinghe et al. 2018; Mittal et al. 2021) are widely used in groundwater remediation. PRB is considered a practical approach to treat heavy metal ions that contaminate subsurface water owing to its low cost, high efficiency, and environmental friendliness (Dong et al. 2009). However, treating co-existing heavy metal ions is difficult, especially when heavy metal ions coexist in the form of cations and anions (Han et al. 2016).

Hualien Winery & Distillery produces approximately 300 tons of waste sludge a month from the final clarifier and settling basin of the wastewater treatment process. The waste removal fee is a considerable operating cost. However, the abundant organic material and high binding affinities of wine-processing waste sludge (WPWS) make it a useful adsorbent for the removal of cationic pollutants from aqueous solutions (Liu et al. 2005; Liu et al. 2006; Liu et al. 2009). The C/N ratio of WPWS ranged from 5:1 to 10:1, indicating that it is a mature and stable material that does not stink. WPWS has previously been used to build a new type of PRB, termed WPWS PRB, to block a gasoline plume in a simulated aquifer, and an extraordinary removal rate was achieved (Liu et al. 2020). However, few studies have simultaneously remediated disparate contaminants using PRB. The purpose of this study was to assess the effectiveness of utilizing the WPWS PRB to treat a mixed metal plume comprising Cu2+, Zn2+, Ni2+, Cd2+, HCrO4, and AsO33−, in a simulated aquifer. The interception efficiency of the WPWS PRB was evaluated, and the distribution of the tested metals in the PRB was revealed.

Preparation of wine-processing waste sludge from winery

The utilized WPWS was collected from a wastewater treatment plant in Hualien Winery & Distillery. It was dried at room temperature, ground, and sieved through 16- and 35-mesh sieves. Then, WPWS particles were mixed evenly using a machine and stored in cans. Finer particles were not considered for PRB because the sludge particles undergo a restricted swelling when they soak in water for several days, which would likely cause the barrier to block itself. An analysis of the chemical properties of WPWS was conducted to determine the major elements, pH, cation exchange capacity (CEC), and organic matter content. The total C, H, N, S, and O contents were evaluated using an auto elemental analyzer (vario EL cube), and K, Ca, Mg, Fe, Al, Mn, and P contents were determined via energy-dispersive spectrometry (Kevex level 4). The pH was measured at a 1:5 ratio of WPWS to water using a pH meter. The organic matter content was obtained by multiplying 1.724 by the carbon content from the wet oxidation method (Walkley & Black 1934). The cation exchange capacity of WPWS was determined using a conventional ammonium-sodium exchange method (Avom et al. 1997). All the experiments about the characteristics of WPWS were performed in duplicate.

Experiment for mixed plume of heavy metals blocked by WPWS PRB

A transparent acrylic water tank (external size of 106 cm × 50 cm × 100 cm) was used to simulate the heavy-metal plume blockage in the aquifer of the WPWS PRB. The distributions of the first sand zone, reactive barrier, second sand zone, and inlet/outlet water chambers in the water tank are presented in Figure 1. The sand layer, which was divided into two sections by the PRB, was located in the middle of the water tank and had a total length of 135 cm and a filling height of 80 cm. The WPWS PRB was located downstream of the plume, traversing the entire tank with a height of 80 cm, the same as the sand layer, and a thickness of 10 cm. Its filling mass was 49.9 kg, and the bulk density was 1.25 g cm−3. The depth of the aquifer layer was a constant 70 cm. To assess the effectiveness of the WPWS PRB in intercepting heavy metals, a sampling point was set at the center point 3 cm downstream from the PRB. Three glass tubes of various lengths (inner diameter R = 6.0 mm, outer diameter R = 6.3 mm) were buried at the sampling point to collect pore water: one near the top of the aquifer layer (1 cm below the water table), one at the middle of the aquifer layer (35 cm from the tank bottom), and one near the bottom of the aquifer layer (1 cm from the tank bottom). In contrast, another sampling point was established 3 cm upstream from the WPWS PRB to monitor the residual amounts of heavy metals in the sand layer. Three glass tubes with different lengths were also set for the three sampled water depths, similar to the other point. This was useful for determining the end time of the experiment. The sand layer porosity was approximately 0.45. A heavy metal mixture solution was prepared and stored in a huge plastic tank with a pH of 5.0. Their initial concentrations were 0.25, 5, 25, 0.5, 0.025, and 0.25 mg/L for Cr (HCrO4), Cu2+, Zn2+, Ni2+, Cd2+, and As (AsO33−), respectively; these concentrations represent five times the control standard levels for groundwater quality. The water tank was initially filled with running water. During the experiment, two peristaltic pumps were utilized to adjust the flow rate of the system, maintained at 100 cm day−1, with one to pump the metallic mixture solution from the reservoir into the tank, and the other to pump water out of the tank. Sampling downstream of the PRB was conducted every 8 h using a fine silicone tube to extend into the end of each sampling tube to draw 50 mL of pore water using a peristaltic pump. However, pore water in the first sand zone was sampled every 24 h. When the ninth sampling was completed, the running water began to replace the solution mixture pumped into the inlet chamber until the experiment was completed. This action was designed to simulate a situation in which the plume was about to pass through. The heavy metal mixture solution in the first sand zone was gradually diluted over time in this stage. Sampled water was filtered through 0.45 μm membranes, and the Cr, Cu, Zn, Ni, and Cd concentrations in the filtrates were then analyzed using a flame atomic absorption spectrometer (AAS, Hitachi Z2300, Japan). The As concentrations were determined using the same AAS equipped with a hydride generator (Hitachi HFS-3). The experiment was terminated at 176 h, and all the metal concentrations in the second sand zone decreased below the control standard levels. Average block efficiency for a pollutant was calculated by the following formulae:
formula
formula
formula
formula
  • : pollutant block rate at a specific sampling time

  • i: sampling order

  • a: sampling site before PRB

  • b: sampling site after PRB

  • : an average of pollutant concentration in the aquifer before PRB

  • : an average of pollutant concentration in the aquifer after PRB

  • Symbols S, M, and L represent sampling depth:

  • S (surface): below the water table of 1 cm

  • M (middle): half depth of the aquifer

  • L (low): 1 cm above the tank bottom

Figure 1

Components used in the water tank WPWS PRB blocking investigation and the sampling zones in the WPWS PRB.

Figure 1

Components used in the water tank WPWS PRB blocking investigation and the sampling zones in the WPWS PRB.

Distribution of heavy metals in the WPWS barrier

When the pore water flowed in the horizontal direction through the simulated aquifer, the heavy metals in the plume gradually moved downward along the water flow owing to its high density. By investigating the distribution of these metals in the WPWS PRB after the reaction's completion, the behavior of each pollutant penetrating the barrier and their migration in the aquifer were revealed. The WPWS PRB was divided into ten intervals of the same size for sample collection, from the bottom to the highest water level of the aquifer (Figure 2). Ten grams of the WPWS sample collected from the barrier was added to a 500 ml flask, and then a 35% H2O2 solution was used to decompose the sample's organic component at 50 °C until it was completely removed. Afterwards, the residual fraction was treated with an aqua regia solution followed by filtration. The concentrations of heavy metals in the filtrate were determined using an AAS system. The metal distribution and their penetration mechanisms for WPWS PRBs were revealed. All the experiments were conducted in duplicate.

Figure 2

Sampling zones within the WPWS PRB (F: first half area, S: second half area, ⇒flow direction).

Figure 2

Sampling zones within the WPWS PRB (F: first half area, S: second half area, ⇒flow direction).

Characterization of WPWS

Wine-processing waste sludge is neutral (pH 6.8). The high organic matter content (52.6%) and CEC (169 cmolc kg−1) revealed that the sludge had a large number of adsorption sites. The dominant elements in WPWS were N (3.1%), P (0.9%), S (3.2%), Ca (4.7%), Mg (0.4%), Na (0.25%), Fe (6.5%), Al (1.2%), Mn (0.1%), and Si (5.4%). The high Ca content originated from the lime that was employed to enhance the dehydration of the raw sludge. The Fe and Al may have originated from the coagulation reagents of the WPWS.

Investigation of heavy metals penetrating the WPWS PRB

As shown in Table 1, more than half of the sampling sites did not contain any Cr, and the Cr concentrations in the residual sites were smaller than the control standard level. A previous study revealed that HCrO4 transforms to Cr(III) once it contacts the organic component of WPWS (Liu et al. 2006). However, Cr has the highest charge compared to the other tested metals; therefore, the WPWS barrier showed the best adsorption of Cr. The interception efficiency of Cu by the WPWS PRB was second only to that of HCrO4. Minimal Cu was detected at all sampling sites during the entire reaction, and no components exceeded the control standard level (Table 2). Copper showed the highest adsorption by organic matter compared to other divalent transition elements. This is attributed to its electronic configuration and adjustable electron orbit (McBride 1989; Liu et al. 2011). The highest Cu concentration in the first sand zone was only 1.775 mg/L. This was still higher than the control standard but far below its concentration in the reservoir (15 mg/L), indicating that the sand layer had a significant adsorption of Cu2+. As shown in Table 3, WPWS PRB also effectively diminished the dispersal of Zn2+. All the Zn concentrations declined under the control standard, but most of the Zn sank or moved around the middle and bottom layers. Although previous batch experiments showed that the WPWS was poor in Zn2+ adsorption under a high concentration (>100 mg Zn2+ L−1), in this study, the WPWS PRB exhibited a high interception ratio of Zn2+ owing to adsorption under a low concentration (<15 mg Zn2+ L−1) (Liu et al. 2009). The WPWS PRB absorbed 96% of the Zn imported into the system.

Table 1

Dynamics of HCrO4 in the aquifer from the metal mixture plume (mg L−1)

 
 
Table 2

Dynamics of Cu2+ in the aquifer from the metal mixture plume (mg L−1)

 
 
Table 3

Dynamics of Zn2+ in the aquifer from the metal mixture plume (mg L−1)

 
 

Ni had the lowest removal rate from the WPWS PRB (Table 4). The nickel concentration in the second sand zone reached its maximum after 8 h of processing. The high Ni concentration behind the PRB persisted until the experiment was completed. Other studies have revealed that the adsorption of Ni is far less than those of Cr and Cu under multiple competitive adsorption conditions (Liu et al. 2006; Liu et al. 2009; Liu et al. 2012a). At the sampling point of the first sand zone, the Ni concentrations exceeded the control standard by 4 to 6 times during most of the reaction time. We speculated that Ni adsorption was poor in the sand layer, which would increase the loading for the WPWS PRB intercepting the Ni. As shown in Table 5, Cd migration seemed to dominate in both the middle and upper sand layers. All the Cd concentrations in the first sand zone were below 0.1 mg/L; however, they were higher than 4.8 to 18.6 times the control standard level (0.005 mg/L). When the high-concentration Cd2+ mass flow arrived at the frontier of the PRB on the third day, excess Cd2+ allowed the PRB to fail in the block so that their concentrations after PRB exceeded the control standard for 80 h. However, only 7–16% of Cd remained in the pore water when the flow passed the WPWS PRB; that is, 84–93% of Cd was removed from the aquifer stream. As expected, AsO33− penetrated the WPWS PRB more easily than the other metals. Considerable amounts of As were observed behind the WPWS PRB at any time, thereby indicating that it is difficult for the WPWS PRB to prevent the spread of As (Table 6). Negatively charged carboxylic groups in WPWS expelled AsO33− ions. The AsO33− mass flow seemed to advance within both the middle and bottom layers, which increased the adsorption loading at the lower part of the PRB. Thus, when the PRB encountered an invasion from large amounts of AsO33−, the As concentrations behind the PRB did not meet the control standard for 80 h. Interestingly, the WPWS PRB continued to intercept high amounts of AsO33−, with only 18 to 45% of As remaining in the flow. In summary, the average block efficiencies in this study for Cr, Cu, Zn, Ni, Cd, and As were calculated to be 96, 87, 91, 51, 94, and 62%, respectively. This indicates that WPWS showed least intercept to Ni2+ and AsO33−.

Table 4

Dynamics of Ni2+ in the aquifer from the metal mixture plume (mg L−1)

 
 
Table 5

Dynamics of Cd2+ in the aquifer from the metal mixture plume (mg L−1)

 
 
Table 6

Dynamics of AsO33− in the aquifer from the metal mixture plume (mg L−1)

 
 

Distribution of heavy metals in WPWS PRB

The residual contents of various metals in each section of the WPWS PRB are listed in Table 7. The heavy metal amounts in the second half of the PRB were clearly higher than the expected levels. A high-efficiency PRB intercepted most of the pollutants in its first half. In general, if the first half cannot capture excess pollutants, residual pollutants will accumulate in the second half and accelerate their breakthrough to the barrier. In summary, the adsorbed metal distribution in the WPWS PRB was consistent with their migration dynamics in the first sand zone. Chromium continued to show the highest amount in WPWS PRB, except for Zn. This indicates that HCrO4 ions reacted with the organic component in PRB and rapidly transformed to Cr(III). WPWS steadily maintained Cr(III) levels, which resulted in a scarcity of HCrO4 behind the barrier. However, the porosity of the WPWS PRB was near 0.5, so the quantities of HCrO4 ions were fixed in the second half. This study proves that the WPWS PRB is useful in remediating Cr(VI) contamination in groundwater. Zinc had the highest accumulation amount in the WPWS PRB owing to its extremely high concentration in the original mixture solution. Compared to other metals, Zn has the least toxicity in humans. The Zn adsorption distribution was fairly uniform throughout the WPWS PRB. The PRB also intercepted high amounts of Cu, and the first half adsorbed more Cu than the other parts (1:0.85). The distribution of Ni in the WPWS PRB was non-uniform. Its content in the first half was lower than that in the second half (0.93:1). We suspected that Ni2+ encountered multiple competitive adsorptions from other species of cations in this area; therefore, the expelled Ni2+ was forced to move further and then was trapped. More Cd was observed in the first half of the experiment, as its content was approximately 1.25 times higher than that in the other part. Arsenite was the least frequently intercepted pollutant in the WPWS PRB. The residual average of As was only 2 ppm, and this amount was only up to 0.6% of HCrO4 adsorption. This indicates that the affinity between AsO33− and WPWS was quite small. Therefore, AsO33− easily passed through the barrier. However, the distribution of As in the WPWS PRB was relatively uniform.

Table 7

Distribution of adsorbed metal ions in the WPWS PRB

Zones in barrierCrCuZnNiCdAs
ppm
F5, S5 296 ± 4, 347 ± 15 172 ± 9, 191 ± 13 1,625 ± 26, 1,630 ± 30 132 ± 10, 112 ± 8 16.0 ± 0.9, 12.0 ± 0.2 2.1 ± 0.3, 1.4 ± 0.1 
F4, S4 218 ± 7, 432 ± 18 310 ± 20, 211 ± 14 1,600 ± 29, 1,615 ± 21 137 ± 6, 216 ± 7 19.5 ± 0.8, 15.0 ± 0.6 2.2 ± 0.1, 2.1 ± 0.1 
F3, S3 448 ± 21, 321 ± 16 261 ± 11, 148 ± 8 1,600 ± 25, 1,630 ± 18 235 ± 13, 168 ± 11 17.5 ± 0.9, 10.0 ± 0.7 1.1 ± 0.1, 2.1 ± 0.3 
F2, S2 349 ± 16, 395 ± 21 192 ± 13, 181 ± 13 1,630 ± 17, 1,610 ± 22 132 ± 5, 199 ± 13 12.5 ± 0.3, 13.5 ± 0.4 2.3 ± 0.2, 2.3 ± 0.3 
F1, S1 337 ± 10, 424 ± 15 117 ± 16, 167 ± 10 1,615 ± 14, 1,495 ± 17 186 ± 10, 186 ± 11 14.0 ± 0.6, 13.0 ± 0.5 2.3 ± 0.4, 2.2 ± 0.3 
Average 330 364 210 179 1,614 1,596 164 176 15.9 12.7 2.0 2.0 
             
Zones in barrierCrCuZnNiCdAs
ppm
F5, S5 296 ± 4, 347 ± 15 172 ± 9, 191 ± 13 1,625 ± 26, 1,630 ± 30 132 ± 10, 112 ± 8 16.0 ± 0.9, 12.0 ± 0.2 2.1 ± 0.3, 1.4 ± 0.1 
F4, S4 218 ± 7, 432 ± 18 310 ± 20, 211 ± 14 1,600 ± 29, 1,615 ± 21 137 ± 6, 216 ± 7 19.5 ± 0.8, 15.0 ± 0.6 2.2 ± 0.1, 2.1 ± 0.1 
F3, S3 448 ± 21, 321 ± 16 261 ± 11, 148 ± 8 1,600 ± 25, 1,630 ± 18 235 ± 13, 168 ± 11 17.5 ± 0.9, 10.0 ± 0.7 1.1 ± 0.1, 2.1 ± 0.3 
F2, S2 349 ± 16, 395 ± 21 192 ± 13, 181 ± 13 1,630 ± 17, 1,610 ± 22 132 ± 5, 199 ± 13 12.5 ± 0.3, 13.5 ± 0.4 2.3 ± 0.2, 2.3 ± 0.3 
F1, S1 337 ± 10, 424 ± 15 117 ± 16, 167 ± 10 1,615 ± 14, 1,495 ± 17 186 ± 10, 186 ± 11 14.0 ± 0.6, 13.0 ± 0.5 2.3 ± 0.4, 2.2 ± 0.3 
Average 330 364 210 179 1,614 1,596 164 176 15.9 12.7 2.0 2.0 
             

F, First half of WPWS barrier; S, Second half of WPWS barrier.

Table 8 presents the comparison of block capacity of WPWS from metallic concentration in the aquifer or in PRB for HCrO4, Cu2+, Zn2+, Ni2+, Cd2+, and AsO33− with that of various materials reported in literatures (Song et al. 2021). Clearly, the zero-valent iron exhibits more block potential for most metals. Other composite materials with a higher intercept capacity were also prepared going through a complicated process, this increasing the cost of operation. Our RPB was built to a simple structure, but it enabled relative high intercept to be shown for HCrO4, Cu2+, Zn2+, and Cd2+, approaching those proposed by the literatures. However, using WPWS for the PRB remediation is feasible and more economic.

Table 8

A comparison of intercept capacity of experimented metal for various materials under similar operational condition

PollutantPRB materialTimeCondition of removalInitial conc.Intercept rateAdsorption (mg/g)Literature
HCrO4 Bioactive sand 50 min Length: 30 cm; diameter: 3.75 cm; 480 g silica sand; pore volume: 104 mL; 12 mL/min 0.52 mg/L 100%  Han et al. (2016)  
 cellulomonas sp. Strain ES6 120 d Length: 17 cm; diameter: 2.5 cm; 1.32 mL/h 2 mg/L 100%  Viamajala et al. (2008)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.25 mg/L 96% 0.347 This work 
Cu2+ Mixture of municipal compost and calcite 16 h Length: 50 cm; diameter: 5 cm; porosity: 0.5; 0.5 mL/min 16 mg/L >99%  Gibert et al. (2005)  
 EGDE–CS–NZVI beads 6 h Length: 50 cm; diameter: 1.5 cm ; 35.6 cm adsorbent filling; 60 mL/h; pH: 6.4 10 mg/L >96% 67.2 Liu et al. (2013)  
 Mixture of limestone, vegetal compost and ZVI cutting 36 mon Thickness:140 cm, parallel to groundwater flow; width: 30 m long perpendicular to groundwater flow; average 6.0 m deep; 0.5–1 m/d 1.2 mg/L 76%  Gibert et al. (2011)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 5 mg/L 87% 0.195 This work 
Zn2+ Acid-washed ZVI/ZVAl mixtures 20 d Length: 45 cm; diameter: 5 cm; height of filled sorbent: 5 cm; pH 5.4; 1.0 mL/min; acid-washed ZVI/ZVAl: 80 g/40 g 20 mg/L 99.5%  Han et al. (2016)  
 Mixture of limestone and vegetal compost 36 mon Thickness: 140 cm, parallel to groundwater flow; width: 30 m, perpendicular to groundwater flow; average 6.0 m deep; 0.5–1 m/d 20 mg/L 47%  Gibert et al. (2011)  
 Mixture NZVI and pumice 17 d Length: 50 cm; diameter: 5 cm; 0.5 mL/min 23 mg/L 94.2% 13.6 Bilardi et al. (2013)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 25 mg/L 91% 1.605 This work 
Ni2+ ZVI/zeolite/activated carbon mixture 92.7 h Weight ratio of ZVI, zeolite and activated carbon: 5:1:4; porosity: 43.2%; 0.5 mL/min 0.73 mg/L 70.7%  Zhou et al. (2014)  
 Pervious concrete 1 d Length: 50 cm; diameter: 10 cm; 0.35 mL/min 1.3 mg/L 69.2%  Shabalala et al. (2017)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.5 mg/L 51% 0.17 This work 
Cd2+ Corn straw, fly ash, zeolite, and Fe-Mn nodule mixture 1 h Length: 20 cm; diameter: 5 cm; quartz sand: corn straw: synthesized zeolites: fly ash: iron-manganese nodule: quartz sand (in sequence) = 1: 2: 2: 2: 2: 2 (volume); 6 mL/min 0.5 mg/L 97%  Fan et al. (2018)  
 Compost, ZVI, limestone gravel, and granite pea gravel mixture 30 mon Thickness:7.9 m, depth: 4.1 m, width: 1.8 m; 30% (v/v) leaf/yard compost, 20% (v/v) ZVI, 5% (v/v) limestone gravel, 45% (v/v) granite pea gravel; average hydraulic gradient: 0.002 1.44 mg/L 99.9%  Ludwig et al. (2009)  
 EGDE–CS–NZVI beads 6 h Length: 50 cm, diameter: 1.5 cm; 35.6 cm adsorbent filling; 60 mL/h; pH 6.4 10 mg/L >96%  Liu et al. (2013)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.025 g/L 94% 0.0143 This work 
AsO43− 6% leaves, 9% compost, 3% Fe(0), 30% Si sand, 30% perlite, 22% limestone mixture (w/w %) 18 mon Thickness: 80 cm, diameter: 20 cm; 0.5 mL/min 2 mg/L 99% 517 Viggi et al. (2010)  
AsO33− ZVI bound with aluminosilicate 3 yr 0.3–6 L/min 0.21 mg/L 98.9%  Morrison et al. (2002)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.25 mg/L 62% 0.002 This work 
PollutantPRB materialTimeCondition of removalInitial conc.Intercept rateAdsorption (mg/g)Literature
HCrO4 Bioactive sand 50 min Length: 30 cm; diameter: 3.75 cm; 480 g silica sand; pore volume: 104 mL; 12 mL/min 0.52 mg/L 100%  Han et al. (2016)  
 cellulomonas sp. Strain ES6 120 d Length: 17 cm; diameter: 2.5 cm; 1.32 mL/h 2 mg/L 100%  Viamajala et al. (2008)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.25 mg/L 96% 0.347 This work 
Cu2+ Mixture of municipal compost and calcite 16 h Length: 50 cm; diameter: 5 cm; porosity: 0.5; 0.5 mL/min 16 mg/L >99%  Gibert et al. (2005)  
 EGDE–CS–NZVI beads 6 h Length: 50 cm; diameter: 1.5 cm ; 35.6 cm adsorbent filling; 60 mL/h; pH: 6.4 10 mg/L >96% 67.2 Liu et al. (2013)  
 Mixture of limestone, vegetal compost and ZVI cutting 36 mon Thickness:140 cm, parallel to groundwater flow; width: 30 m long perpendicular to groundwater flow; average 6.0 m deep; 0.5–1 m/d 1.2 mg/L 76%  Gibert et al. (2011)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 5 mg/L 87% 0.195 This work 
Zn2+ Acid-washed ZVI/ZVAl mixtures 20 d Length: 45 cm; diameter: 5 cm; height of filled sorbent: 5 cm; pH 5.4; 1.0 mL/min; acid-washed ZVI/ZVAl: 80 g/40 g 20 mg/L 99.5%  Han et al. (2016)  
 Mixture of limestone and vegetal compost 36 mon Thickness: 140 cm, parallel to groundwater flow; width: 30 m, perpendicular to groundwater flow; average 6.0 m deep; 0.5–1 m/d 20 mg/L 47%  Gibert et al. (2011)  
 Mixture NZVI and pumice 17 d Length: 50 cm; diameter: 5 cm; 0.5 mL/min 23 mg/L 94.2% 13.6 Bilardi et al. (2013)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 25 mg/L 91% 1.605 This work 
Ni2+ ZVI/zeolite/activated carbon mixture 92.7 h Weight ratio of ZVI, zeolite and activated carbon: 5:1:4; porosity: 43.2%; 0.5 mL/min 0.73 mg/L 70.7%  Zhou et al. (2014)  
 Pervious concrete 1 d Length: 50 cm; diameter: 10 cm; 0.35 mL/min 1.3 mg/L 69.2%  Shabalala et al. (2017)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.5 mg/L 51% 0.17 This work 
Cd2+ Corn straw, fly ash, zeolite, and Fe-Mn nodule mixture 1 h Length: 20 cm; diameter: 5 cm; quartz sand: corn straw: synthesized zeolites: fly ash: iron-manganese nodule: quartz sand (in sequence) = 1: 2: 2: 2: 2: 2 (volume); 6 mL/min 0.5 mg/L 97%  Fan et al. (2018)  
 Compost, ZVI, limestone gravel, and granite pea gravel mixture 30 mon Thickness:7.9 m, depth: 4.1 m, width: 1.8 m; 30% (v/v) leaf/yard compost, 20% (v/v) ZVI, 5% (v/v) limestone gravel, 45% (v/v) granite pea gravel; average hydraulic gradient: 0.002 1.44 mg/L 99.9%  Ludwig et al. (2009)  
 EGDE–CS–NZVI beads 6 h Length: 50 cm, diameter: 1.5 cm; 35.6 cm adsorbent filling; 60 mL/h; pH 6.4 10 mg/L >96%  Liu et al. (2013)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.025 g/L 94% 0.0143 This work 
AsO43− 6% leaves, 9% compost, 3% Fe(0), 30% Si sand, 30% perlite, 22% limestone mixture (w/w %) 18 mon Thickness: 80 cm, diameter: 20 cm; 0.5 mL/min 2 mg/L 99% 517 Viggi et al. (2010)  
AsO33− ZVI bound with aluminosilicate 3 yr 0.3–6 L/min 0.21 mg/L 98.9%  Morrison et al. (2002)  
 WPWS 176 h Thickness: 10 cm; height: 70 cm; width: 50 cm; 243 mL/min 0.25 mg/L 62% 0.002 This work 

ZVI, zero-valent iron; ZVAl, zero-valent aluminum; NZVI, nano-ZVI; EGDE, ethylene glycol diglycidyl ether.

A WPWS PRB may weaken the hazards from HCrO4, Cu2+, and Zn2+ in a simulated aquifer, but it was unable to prevent the dispersal of AsO33−, Ni2+, and Cd2+. These results may be attributed to the slight hydrophobicity of the WPWS, which caused a decrease in the water permeability of the WPWS PRB. However, the sand layer showed significant adsorption of positively charged heavy metals, large amounts of metal cations were fixed in the zone. Two kinds of competitive adsorption are speculated to have occurred in this system, including adsorptive competition between heavy metals and sands, as well as adsorptive competition between heavy metals and WPWS. The effect of multiple competitive adsorption on Ni2+ enhanced the formation of the high concentration Ni2+ mass flow in the sand layer, which eventually led to a failure in the block.

A WPWS can feasibly be utilized to create a PRB for remediation in a brown field, although it did not block all the metals in this study. There are three reasons for this: (1) the effect of competitive adsorption, (2) conducting the experiment under a super high concentrations, and (3) the experiment utilized a PRB only 10 cm thick. Brown field plumes rarely contain more than three species of metals, and metal concentrations are usually not as high as those tested in this study. When remediation in a brown field concludes, we suggest digging up the WPWS PRB and sending its debris to a brick kiln factory or a cement factory for final disposal. The Fe, Al, Ca, and Si in the WPWS may serve as raw materials for producing bricks or cement. Toxic metals trapped in WPWS particles would then be sintered and fixed into the structure of brick or cement, and their toxicity would hence be removed.

This work was financially supported by the Environmental Protection Administration Executive Yuan, Republic of China, under Project 109C003913.

All relevant data are included in the paper or its Supplementary Information.

Ahuja
S.
2008
Arsenic Contamination of Groundwater: Mechanism, Analysis, and Remediation
.
John Wiley & Sons
Hoboken, NJ
.
ATSDR, Agency for Toxic Substances and Disease Registry
2007
Toxicological Profile for Arsenic
.
PHS (August 2007)
.
U.S. Department of Health and Human Services
,
Washington, DC
.
Avom
J.
,
Mbadcam
J. K.
,
Noubactep
C.
&
Germain
P.
1997
Adsorption of methylene blue from an aqueous solution onto activated carbon from palm-tree cobs
.
Carbon
35
(
3
),
365
369
.
Bilardi
S.
,
Calabro
P. S.
,
Care
S.
,
Moraci
N.
&
Noubactep
C.
2013
Improving the sustainability of granular iron/pumice systems for water treatment
.
Journal of Environmental Management
121
,
133
141
.
Chen
L.
,
Liu
F. Y.
,
Liu
L.
,
Dong
H. Z.
&
Colberg
P. J. S.
2011
Benzene and toluene biodegradation down gradient of a zero-valent iron permeable reactive barrier
.
Journal of Hazardous Materials
188
,
110
115
.
Chen
K. Y.
,
Tzou
Y. M.
,
Chan
Y. T.
,
Wu
J. J.
,
Teah
H. Y.
&
Liu
Y. T.
2019
Removal and simultaneous reduction of Cr(VI) by organo-Fe(III) composites produced during coprecipitation and coagulation processes
.
Journal of Hazardous Materials
376
,
12
20
.
Dhal
B.
,
Thatoi
H. N.
,
Das
N. N.
&
Pandey
B. D.
2013
Chemical and microbial remediation of hexavalent chromium from contaminated soil and mining/metallurgical solid waste: a review
.
Journal of Hazardous Materials
250–251
,
272
291
.
Dong
J.
,
Zhao
Y. S.
,
Zhang
W. H.
&
Hong
M.
2009
Laboratory study on sequenced permeable reactive barrier remediation for landfill leachate-contaminated groundwater
.
Journal of Hazardous Materials
161
,
224
230
.
Erto
A.
,
Lancia
A.
,
Bortone
I.
,
Di Nardo
A.
,
Di Natale
M.
&
Musmarra
D.
2011
A procedure to design a permeable adsorptive barrier (PAB) for contaminated groundwater remediation
.
Journal of Environmental Management
92
,
23
30
.
Gibert
O.
,
Pablo
J. D.
,
Cortina
J. L.
&
Ayora
C.
2005
Municipal compost-based mixture for acid mine drainage bioremediation: metal retention mechanisms
.
Applied Geochemistry
20
(
9
),
1648
1657
.
Gibert
O.
,
Rötting
T.
,
Cortina
J. L.
,
Pablo
J. D.
,
Ayora
C.
,
Carrera
J.
&
Bolzicco
J.
2011
In-situ remediation of acid mine drainage using a permeable reactive barrier in Aznalcóllar (Sw Spain)
.
Journal of Hazardous Materials
191
(
1–3
),
287
295
.
Henderson
A. D.
&
Demond
A. H.
2007
Long-term performance of zero-valent iron permeable reactive barriers: a critical review
.
Environmental Engineering Science
24
,
401
423
.
Hillel
D.
1999
Soil and hydrologic cycle
. In:
The Nature and Properties of Soils
, 12th edn (
Brady
N. C.
&
Weil
R. R.
, eds).
Prentice-Hall, Inc.
,
Upper Saddle River, New Jersey
, pp.
213
264
.
Hu
B.
,
Song
Y.
,
Wu
S.
,
Zhu
Y.
&
Sheng
G.
2019a
Slow released nutrient-immobilized biochar: a novel permeable reactive barrier filler for Cr (VI) removal
.
Journal of Molecular Liquids
286
,
110876
.
Kim
H. S.
,
Kim
Y. J.
&
Seo
Y. R.
2015
An overview of carcinogenic heavy metal: molecular toxicity mechanism and prevention
.
Journal of Cancer Prevention
20
(
4
),
232
240
.
Kumarasinghe
U.
,
Kawamoto
K.
,
Saito
T.
,
Sakamoto
Y.
&
Mowjood
M. I. M.
2018
Evaluation of applicability of filling materials in permeable reactive barrier (PRB) system to remediate groundwater contaminated with Cd and Pb at open solid waste dump sites
.
Process Safety and Environmental Protection
120
,
118
127
.
Liu
C. C.
,
Wang
M. K.
&
Li
Y. S.
2005
Removal of nickel from aqueous solution using wine processing waste sludge
.
Industrial & Engineering Chemistry Research
44
,
1438
1445
.
Liu
C. C.
,
Wang
M. K.
,
Li
Y. S.
,
Chiou
C. S.
,
Lin
Y. A.
&
Huang
S. S.
2006
Chromium removal and sorption mechanism from aqueous solutions by wine processing waste sludge
.
Industrial & Engineering Chemistry Research
45
,
8891
8899
.
Liu
C. C.
,
Wang
M. K.
,
Chiou
C. S.
,
Li
Y. S.
&
Lin
Y. A.
2009
Biosorption of chromium, copper and zinc by wine-processing waste sludge: single and multi-component system study
.
Journal of Hazardous Materials
171
,
386
392
.
Liu
C. C.
,
Li
Y. S.
,
Chen
Y. M.
,
Wang
M. K.
,
Chiou
C. S.
,
Yang
C. Y.
&
Lin
Y. A.
2011
Biosorption of chromium, copper and zinc on rice wine processing waste sludge in fixed bed
.
Desalination
267
,
20
24
.
Liu
C. C.
,
Chen
Y. M.
,
Wang
M. K.
&
Lin
Y. A.
2012a
Adsorption of Cu(II) from aqueous solution by wine processing waste sludge
.
Water Environment Research
84
(
9
),
733
743
.
Logan
W. B.
1999
Soil colloid their nature and practical significance
. In:
The Nature and Properties of Soils
, 12th edn (
Brady
N. C.
&
Weil
R. R.
, eds)
Prentice-Hall, Inc.
,
Upper Saddle River, New Jersey
, pp.
307
342
.
Ludwig
R. D.
,
Smyth
D. J. A.
,
Blowes
D. W.
,
Spink
L. E.
,
Wilkin
R. T.
&
Jewett
D. G.
2009
Treatment of arsenic, heavy metals, and acidity using a mixed ZVI-compost PRB
.
Environmental Science & Technology
43
(
6
),
1970
1976
.
Maamoun
I.
,
Eljamal
O.
,
Falyouna
O.
,
Eljamal
R.
&
Sugihara
Y.
2020
Multi-objective optimization of permeable reactive barrier design for Cr(VI) removal from groundwater
.
Ecotoxicology and Environmental Safety
200
,
110773
.
McBride
M.
1989
Reactions controlling heavy metal solubility in soils
.
Advances in Soil Science
10
,
1
56
.
Morrison
S. J.
,
Metzler
D. R.
&
Dwyer
B. P.
2002
Removal of As, Mn, Mo, Se, U, V and Zn from groundwater by zero-valent iron in a passive treatment cell: reaction progress modeling
.
Journal of Contaminant Hydrology
56
(
1/2
),
99
116
.
Nur-E-Alam
M.
,
Mia
M. A. S.
,
Ahmad
F.
&
Rahman
M. M.
2020
An overview of chromium removal techniques from tannery effluent
.
Applied Water Science
10
,
205
.
Obiri-Nyarko
F.
,
Grajales-Mesa
S. J.
&
Malina
G.
2014
An overview of permeable reactive barriers for in situ sustainable groundwater remediation
.
Chemosphere
111
,
243
259
.
Okyere
M. S.
2019
Corrosion Protection for the Oil and Gas Industry: Pipelines, Subsea Equipment, and Structures
.
CRC Press
,
Boca Raton, FL
.
Rent
G.
,
Wang
X.
,
Huang
P.
,
Zhong
B.
,
Zhang
Z.
,
Yang
L.
&
Yang
X.
2017
Chromium (VI) adsorption from wastewater using porous magnetite nanoparticles prepared from titanium residue by a novel solid-phase reduction method
.
Science of the Total Environment
607–608
,
900
910
.
Shabalala
A. N.
,
Ekolu
S. O.
,
Diop
S.
&
Solomon
F.
2017
Previous concrete reactive barrier for removal of heavy metals from acid mine drainage-column study
.
Journal of Hazardous Materials
323
,
641
653
.
Snape
I.
,
Morris
C. E.
&
Cole
C. M.
2001
The use of permeable reactive barriers to control contaminant dispersal during site remediation in Antarctica
.
Cold Regions Science and Technology
32
(
2–3
),
157
174
.
Viamajala
S.
,
Peyton
B. M.
,
Gerlach
R.
,
Sivaswamy
V.
&
Petersen
J. N.
2008
Permeable reactive biobarriers for in situ Cr(VI) reduction: bench scale tests using cellulomonas sp. strain ES6
.
Biotechnology & Bioengineering
101
(
6
),
1150
1162
.
Viggi
C. C.
,
Pagnanelli
F.
,
Cibati
A.
,
Uccelletti
D.
,
Palleschi
C.
&
Toro
L.
2010
Biotreatment and bioassessment of heavy metal removal by sulphate reducing bacteria in fixed bed reactors
.
Water Research
44
,
151
158
.
Wei
X.
,
Kong
X.
,
Wang
S.
,
Xiang
H.
,
Wang
J.
&
Chen
J.
2013
Removal of heavy metals from electroplating wastewater by thin-film composite nanofiltration hollow-fiber membranes
.
Industrial & Engineering Chemistry Research
52
(
49
),
17583
17590
.
Wu
B.
2020
Chapter 6 - Human health hazards of wastewater
. In: (
Ren
H.
&
Zhang
W.
, eds),
High-Risk Pollutants in Wastewater. Elsevier
,
Amsterdam, The Netherlands
, pp.
125
139
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).