1 Introduction

Heavy metal pollution from both natural and anthropogenic sources is toxic and persistent in the environment. Although some metals are essential trace elements naturally found in food and water in small amounts, they can be harmful above certain limits specified by the World Health Organization (WHO 2011; J. J. Zhao et al. 2019). Bioaccumulation, high toxicity, and carcinogenic effects of heavy metals pose a threat to human health (Järup 2003). Some health implications associated with some notable heavy metals are high risks of cancer, renal and neurological effects, developmental delay, hypertension, and mental retardation (Shi et al. 2019). These health risks have raised global concerns over the last 100 years about regulating most harmful metals such as Pb, Zn, Cu, Mn, Cd, Hg, As, and Cr in domestic products (European Commission DG ENV. 2002). On the other hand, mine waste is another medium through which heavy metals find their way into waterways and finally into the food chain (Gyamfi et al. 2019; Hudson et al. 2018).

Frongoch mine near Pont-rhyd-y-groes, central Wales, was a lead and zinc mine site operated in the 1700s–1900s covering over 11 hectares of land area, which left vast waste dumps at the end of its operation (Fig. 1a). Heavy metals are discharged from the underground workings via the abandoned Frongoch Adit study site situated 52° 21′ 04.211″ N, 3° 53′ 28.612″ W. Based on the 2015 classification, Frongoch mine metal discharge majorly impacted the ecological and chemical state of Frongoch stream, Nant Cell, Nant Cwmnewydion, and the River Magwr downstream, and consequently failed to achieve the EU Water Framework Directive standards (Edwards and Williams 2016; Hudson et al. 2018). The heavy metal pollution also affected economic and recreational activities and could be poisonous to aquatic species. Among the 4908 non-coal abandoned mines on the priority list in Wales and England (Fig. 1c), Jarvis and Mayes (2012) noted that Pb, Cd, and Zn were major metal pollutants of national concern from Frongoch mine. Prior to remediation, approximately 23 and 1.5 tons of zinc and lead metals was discharged annually, which ranked Frongoch as Wales’ second most polluted mine and a top priority among other mine sites (Edwards et al. 2016).

Fig. 1
figure 1

Abandoned Frongoch metal mine site and its remediated state. a Abandoned mine ruins in 2020 (Paul Edwards 2010). b Impermeable capping of mine waste and creating a wetland area by NRW in 2015. c Heavy metal-polluted water bodies impacted by abandoned non-coal mines in England and Wales (Jarvis and Mayes 2012)

Full size image

Several remediation approaches have been undertaken by the Natural Resources Wales (NRW) to minimize heavy metal flow downstream, such as stream diversion in 2011, surface drain around the mine waste and flood attenuation pond in 2013, low-permeability capping of mine waste and creation of a wetland area in 2015, and further capping in 2018 (Fig. 1b). Though these remediation strategies have yielded considerable success in immobilizing polluting metals from the mine waste and diverting watercourse, their impact is still a concern. In the monitoring data for water quality, Edwards et al. (2016) noted an 80% water flow reduction, and consequently, zinc, lead, and cadmium by 44%, 63%, and 58%, respectively. Besides the cost-intensive nature of these projects, it is imperative to clearly understand the relationship between point and diffused pollutant sources, considering the long-term metal deposits downstream along pathways into river catchments (Environment Agency 2012). However, lead and zinc metal tailings are still lingering along the watercourse path through the resuspended metal-rich sediments. Hence, there is a need for further studies on appropriate non-invasive remediation approach with minimal costs and disturbances on the catchment pollution sources (Crane et al. 2017).

Biochar is a carbon-rich, porous, and stable solid formed when pyrolyzed under limited oxygen. It is considered a promising eco-friendly and cost-effective adsorbent of heavy metals using pyrolyzed waste biomass materials. These include animal waste (J. Liu et al. 2020; W. Zhang et al. 2020), waste pine cone biomass (Biswas et al. 2020), waste crop residues (R. Xiao et al. 2019), and industrial waste materials of biosolids (Qiu and Duan 2019). The unique adsorbent properties of biochar are induced by pyrolysis heat transformation that forms a carbon-ring without oxygen or hydrogen (Hu et al. 2020; Kim et al. 2019).

Although biochar properties and adsorption efficiency of Pb2+ and Zn2+ metal have been investigated and reviewed over time in many studies (Betts et al. 2013; Li et al. 2017; L. Wang et al. 2019a), no previous on-site field study using biochar has been published on this watercourse. Unique properties of biochar vital for metal adsorption are surface chemistry (Song et al. 2020), mineral ions content (Cheung et al. 2000), pH level (Y. Xiao et al. 2017), surface area, and functional group (Inyang et al. 2016). Different biochar amendments have improved their properties to acquire optimal adsorption capability (Godwin et al. 2019). Nevertheless, Wu et al. (2019) noted that most modifications are uneconomical for broad field-scale implementation and could be another secondary pollution source. Commercial activated carbon has recorded unique adsorptive properties for heavy Pb2+ and Zn2+ metals in wastewater treatment studies attributed to having a high surface area (Loganathan et al. 2018). Compared to other commercial metal adsorbents such as activated carbon, cactus, alumina, and silica gel (Amari et al. 2019; Rabia et al. 2018; Renu et al. 2017), biochar has enormous environmental benefits. These include but are not limited to a reduction in global warming (Ma et al. 2019), biomass-waste valorization (Foong et al. 2020), minimized agricultural waste (L. Liu et al. 2019), and economically estimated at 200% lower than activated carbon (Thompson et al. 2016). Hence, it is considered most suitable for a large-scale field application.

In a review, L. Wang et al. (2019a) noted that there is limited field application of conventional biochar on water treatment due to a relatively low metal adsorption efficiencies metals such as Cr6+, As5+, and Pb2+ (Tan et al. 2016), while other studies noted that its efficiency could vary by the nature of biomass and production processes (Boeykens et al. 2019; Kwak et al. 2019). This study carried out a field-scale application on Pb2+ and Zn2+ metal adsorption along Frongoch mine adit’s watercourse. We hypothesize that pyrolyzed waste biomass of pine feedstock (leylandii) and activated carbon could be promising adsorbents for Pb2+ and Zn2+ heavy metal pollutants in Frongoch Adit. Hence, the objectives of this study are (1) to determine the adsorption capability of pyrolyzed waste leylandii feedstock in its untreated form for a possible large-scale and low-cost application and (2) to investigate the effect of post-amendments on biochar at different saturation ratios of biochar per ash supplement (kg/kg).

2 Materials and Methods

2.1 Pyrolysis and Ash Amendment

Dried mixed chips of pinewood waste (leylandii) feedstocks were pyrolyzed at 700 °C in a burning chamber by Sion Brackenbury pyrolysis company. After pyrolysis, a known proportion of the pyrolyzed biochar was treated with ash in a mixer and periodically sprayed with water to facilitate the binding of ash to the biochar surface. The treatment was scaled by a mass ratio (kg/kg) of supplement at saturation ratios of 1:10 (BB1), 1:35 (BB2), 1:50 (BB3), and 1:100 (BB4) of biophos ash to biochar, respectively. Another portion of biochar was amended with cockleshell ash and biophos ash at a mass ratio (kg/kg/kg) of 1:1:100 (CBB). Commercial activated carbon (AC) and a portion of the biochar were unamended (BC), while the contaminated water was used as the study control. The samples’ surface area, pH, and metal ions were determined via ASTM Brunauer-Emmett-Teller (BET) analysis, black carbon’s ASTM standards (ASTM D6851 2011), and Rigaku benchtop XRF spectrometer, respectively.

Leylandii was chosen for this study because it is one of the fastest-growing native coniferous trees common in Wales and England. If a suitable metal adsorbent, it could serve as a strategy to improve its waste management. Cockleshells are wastes from harvested marine bivalve mollusks common in Wales. They contain 98% calcium carbonate (CaCO3) (Moideen et al. 2020) and have been recorded as a cost-effective adsorbent of Pb2+ in a previous study (Budin et al. 2014). Biophos ashes are waste from Swansea power station, which was provided by Sion Brackenbury of Commons Vision Ltd. Biophos ash was tested for its adsorbent capabilities similar to other kinds of ashes reported in earlier studies such as fly ash (Novais et al. 2016), biomass ash (Xu et al. 2018), and coal ash (Shaobin Wang et al. 2006). Additionally, it contains significant Na+ and Mn+, notable for Pb2+ and Zn2+ metal adsorption (Esrafili et al. 2019; Holguera et al. 2018). These supplements were used to alter the biochar surface and mineral ion content.

2.2 Field Experiment

The study site is located at Frongoch Adit, situated 52° 21′ 04.211″ N, 3° 53′ 28.612″ W in central Wales (near Pontrhydygroes). Fifty grams of BB1, BB2, BB3, and BB4, cockleshell-biophos supplement CBB, unamended BC, and AC samples were weighed out. Using a 1-mm sieve, the samples were filtered to remove unbounded excess ashes from the biochar and packaged in long socks of 100-μm nylon mesh filter bag to a total of 63 sample balls: 9 balls in each role of 7 socks. The sample balls were compartmentalized with plastic cable ties 2 × 140 mm, one cable tie at each sample’s top and bottom, leaving 1-m space between ties. The partitioned sample socks were tied on a 10-m nylon rope attached to a wooden frame and allowed to float in the pool of metal-contaminated water of an estimated average volume of 45 L for 46 days at sampling intervals of 72, 240, 432, 624, and 1104 h (Fig. 2). Each collection was done by cutting off the last sample in each sock and taken to Swansea University in a sealed polyethylene bag for further analysis.

Fig. 2
figure 2

a Schematic diagram illustrating biochar sample preparation and the mechanism of metal removal. b Site installation

Full size image

2.3 Sample Collection and Statistical Analysis

The collected samples were dried in labeled crucibles at 80 °C for 24 h. The dried samples were pounded with a ceramic mortar and pestle and sieved at <63 μm, as recommended for benchtop XRF analysis machine. Two grams of each sample is placed in a 24-mm XRF-labeled circumference sampling plastics cups, enclosed with 4-μm prolene film at its bottom. Each sample in three replicates was compressed with a Pana press vice at 350-in. pounds’ scale and 7-mm height before enclosing with the cover and transferred to the benchtop XRF spectrometer for analysis. The water samples were collected and analyzed by NRW via inductively coupled plasma mass spectrometer (ICP-MS) and industrial and chemical measurement (ICM-QES).

The sample size consists of only three replicates per sample, suggesting a non-parametric test due to sample size. Statistical analysis was performed using a Kruskal-Wallis test in R-stat software version 3.6.1 to determine samples’ statistical significance at a 95% confidence interval. Regression analysis was used to examine the correlation between sample mineral ions and metal adsorption using IBM SPSS statistics software version 25 and Microsoft Excel for graphical adsorption analysis.

3 Results and discussion

3.1 Pre-experimental Characteristics of Samples and Water Quality Analysis

The water quality result shows that zinc concentration (15.8 mg/L) was 25 times higher than lead (0.62 mg/L), while cadmium (0.02 mg/L) was considered the least among the three metals of major concern. Other parameters include ion content of sodium (Na+): 6.9 mg/L, magnesium (Mg2+): 4.6 mg/L, potassium (K+): 1.0 mg/L, and calcium (Ca2+): 16.1 mg/L; pH: 7.2; conductivity; and flow: 0.014 m3/s (Table 1).

Table 1 Parameters of Frongoch adit contaminated water quality
Full size table

The samples’ physicochemical properties showed variations in metal content, mineral ions, and BET surface area. BC and amended BB1–4 had a relatively equal amount of Pb2+, Mn+, and Zn2+ higher than AC. The Fe2+ content in AC at 958 ppm was higher than other samples by 124% than BC, 139% average than BB1–4, and 46% than CBB. The mineral ion contents of K+, Ca2+, Mg2+, and Na+ suggest a variation by nature of samples and amendment supplement. No significant difference in K+ content was shown between BC and AC but slightly differed among treated samples. CBB had more Ca2+ by 108% among treated samples than BB1–4, but AC had the highest Ca2+ content. Initial Pb2+ and Zn2+ contents in treated BB1–4 samples were low but slightly higher in Na+ than untreated BC. The Brunauer-Emmett-Teller (BET) surface area analysis revealed that AC at 987 m2/g was 2.7 times higher than BC and 8.3 times higher than the average of BB1–4. Comparatively, the ash amendment supplement caused a decrease in BB1–4 samples’ surface area by 65% and CBB sample by 76% relative to the unamended BC (Table 2).

Table 2 Adsorbent samples physicochemical properties
Full size table

Similarly, a larger surface area of AC was observed by H. Wang et al. (2020), who noted that AC at 1658 m2 g−1 was 96.2 times larger than the non-activated BC at 17.9 m2 g−1. Biophos ash supplement had little or no significant effect on the samples’ pH, which is similar to Bentley and Summers (2020) findings, who noted an insignificant effect of ash on pine biochar pH and slightly lower than AC. Other mineral ions such as Fe2+, Mg2+, Ca2+, and K+ were slightly affected by increasing the biophos amendment ratio (kg/kg), while Mn+ and Na+ were increased. Cockleshell and biophos supplement significantly increased by 1.6 g/kg and 0.38 g/kg of Ca2+ and Fe2+ content respectively in the CBB sample, which could be attributed to the 98% calcium similarly observed in cockleshell (Moideen et al. 2020), but significantly decreased its surface area.

Notable differences in mineral ions between BC and AC were a higher Ca2+ content in BC, while Fe2+ and Na+ were higher in AC. These variations can be attributed to the differences in the feedstock and production processes between the commercially produced and conventional biochar (Yargicoglu and Reddy 2014). However, it is unclear how these properties could affect the samples’ adsorbent capabilities, but cation exchange and complexation are two critical mechanisms determined by the nature of feedstock, mineral content, and pyrolysis (Lehmann and Joseph 2009; Shengsen Wang et al. 2015; J. J. Zhao et al. 2019).

3.2 Adsorption Correlation with Adsorbent Physicochemical Features

The net adsorption at 72, 240, 432, 624, and 1104 h of total sampling intervals revealed that all samples were good absorbent of Pb2+ and Zn2+ at different degrees (Table 3). Mineral ions such as Na+, K+, Ca2+, and Mg2+ are assayed for their role in Pb2+ and Zn2+ sorption through cation exchange capacity (CEC) interaction (Li et al. 2017; Y. Wang and Liu 2017; J. J. Zhao et al. 2019), which varies by the nature of biomass and pyrolysis conditions (Mahdi et al. 2018). Other physicochemical properties which include but are not limited to pH and BET surface area are also correlated by regression analysis (Li et al. 2017; Mahdi et al. 2018).

Table 3 Metal adsorption in all adsorbent samples
Full size table

3.2.1 Pb2+ Adsorption

Adsorption per unit adsorbent mass shows that BC by 2 g adsorbed 1335 ppm in 1104 h, equivalent to 1.3 g/kg, and better than AC (1.0 g/kg) in a concentration of 0.62 mg/L (Fig. 3). The total net Pb2+ adsorbed by 50 g of BC and AC in 1104 h field setup was 33 g/kg and 26 g/kg, respectively, from 0.66 mg/L concentration at a flow rate of 0.014 m3/s. The optimum Pb2+ adsorption was between the first sampling interval of 0–240 h at the rate of 1.9 ppm/h and reduced by 63% between the third and fifth intervals. Pb2+ adsorption positively correlated with K+ and Mg2+ at p < 0.05, reflecting the high K+ content notably in BC, AC, and BB4 samples, while Mg2+ was higher in only BC and AC samples (Fig. 4a–f). The general knowledge of heavy metal replacement of alkaline metal or earth metals such as K+, Ca2+, Mg2+, and Na+ on adsorbent surfaces and within is observable in biochar (Jiang et al. 2015; Yu et al. 2019). The Pb2+ adsorption difference between BC and AC can be attributed to the higher Ca2+ content in BC (Ca2+ 571 ppm), and greater open pores than the ash amended samples (BB1–4). A similar effect of initial Ca2+ content on Pb2+ adsorption was observed by Zhao et al. (2019) in a comparative adsorption potential using different biomass feedstocks. The underlying Ca2+ mechanism on Pb2+ adsorption in biochar was explained by Wu et al. (2019), who noted that CaCO3 and Ca3(PO4)2 release their inorganic anions such as CO32−, SO42−, and OH− in biochar to take up Pb2+ by precipitation. This inorganic anion precipitation mechanism has been earlier observed using XRD analysis (Chi et al. 2017; T. Zhang et al. 2017). Contrarily, among treated samples, Ca2+ did not positively correlate with Pb2+ or Zn2+ adsorption due to the highly reduced surface area that provides sorption sites. The hindered surface area’s effect is most reflected in CBB (760 ppm) with the least BET and sorbed Pb2+, despite its high Ca2+ content. Adsorbent’s BET surface area is known to greatly determine the pore structure, volume, and diameter, which provides the needed sorption sites for cationic exchange capacity (CEC) (Fahmi et al. 2018). Besides other factors, ash content has been reported to significantly disrupt biochar surface area, porosity, and functional groups (Fahmi et al. 2018).

Fig. 3
figure 3

a Pb2+ net absorption in 1104 h of sampling time and b comparative box-and-whiskers plot showing net adsorption differences between untreated biochar (BC), activated carbon (AC), and treated BB1–4 and CBB samples

Full size image
Fig. 4
figure 4

Pb2+ adsorption correlations with adsorbent physicochemical properties. a K+ ion. b Ca2+ ion. c Mg2+ ion. d Na+ ion, e BET surface area. f pH level. Pearson’s R2 coefficients and significance (p) values indicated at 95% confidence level

Full size image

Generally, treated samples’ adsorption exponentially increased by their increasing surface area (BB1 > BB2 > BB3 > BB4 > CBB), while the most treated BB1 and CBB samples were the least adsorbents of Pb2+. However, there was no statistically significant difference in overall net adsorption (Kruskal-Wallis chi-squared = 2.6818, df = 6, p value = 0.8476) and between treated samples (Kruskal-Wallis chi-squared = 0.42771, df = 4, p value = 0.9801). This result suggests that besides Ca2+ content, the surface area is vital for Pb2+ adsorption.

In a laboratory batch experiment, Z. Liu and Zhang (2009) recorded a total of 4.25 mg/g Pb2+ adsorption using untreated pinewood biochar, while J. Zhang et al. (2019) used KOH-activated sludge biochar to sorb 57.48 mg/g. The higher adsorption capability is attributed to increased BET surface area. Using a Fe2+-coated bamboo biochar, Zhang et al. (2013) achieved 298.7 m2/g BET, which is lower than that of the Leylandii pinewood in this study. However, chemical modification of biochar is relatively not feasible in a field application due to cost-intensiveness. Also, Li et al. (2017), in a study by Godwin et al. (2019), noted that there were few reported field applications of biochar to remove heavy metals from contaminated discharged water, which has shown to be a promising adsorbent in this study.

Adsorption rate with contact time in all samples BC, AC, BB1–4, and CBB shows that BC adsorbed 80% of its net total Pb2+ in 26 days compared to AC with 76% (1036 ppm). Other lower adsorbent of treated samples BB1–4 and CBB adsorbed a relatively similar average of 62% net. The adsorption rate in all samples suggests that BC is a faster and better adsorbent material for Pb2+. This result is contrary to Ramola et al. (2019), who observed that unmodified biochar, pyrolyzed at 700 °C, adsorbed less Pb2+ though with a bentonite calcite supplement amendment. However, there was no significant difference in Pb2+ sorption between BC, AC, treated BB1–4, and CBB samples (Kruskal-Wallis chi-squared = 1.3036, df = 2, p value = 0.5211).

3.2.2 Zn2+ Adsorption

The total net Zn2+ adsorption revealed that all the samples are suitable adsorbents but differed in sorption rates and capacity. At 1104-h final sampling interval, AC adsorbed 1221 ppm, BC 666 ppm, and lower sorption in other treated BB1–4 and CBB samples (Fig. 5). Zn2+ adsorption in AC was the most rapid compared to other adsorbents, which occurred at the first 72-h interval at 1.4 g/kg but rapidly underwent desorption afterward. Though AC was 3.2 folds higher than BC at 0.43 g/kg in the first 72-h sampling time, it was most unstable among other adsorbents. But there was no statistically significant difference in Zn2+ adsorption among all samples (Kruskal-Wallis chi-squared = 11.038, df = 6, p value = 0.08721).

Fig. 5
figure 5

a Zn2+ net absorption in 1104 h of sampling time and b comparative box-and-whiskers plot shows net adsorption differences between untreated biochar (BC), activated carbon (AC), treated BB1–4, and CBB samples

Full size image

Comparatively, Zn2+ adsorption between BC, AC, treated BB1–4, and CBB revealed that untreated AC was a better adsorbent with a statistically significant difference (Kruskal-Wallis chi-squared = 6.8, df = 2, p value = 0.03337). BB4, the least treated sample, was a better adsorbent than other treated samples by 28%. Among all samples, CBB was the least adsorbent (367 ppm) and had the lowest surface area but showed no statistical difference when compared with other treated samples (Kruskal-Wallis chi-squared = 2.3546, df = 4, p value = 0.6709). The total net Zn2+ adsorbed in this field setup by 50 g of each adsorbent in the 15.8 mg/L zinc concentration in 1104 h was 30 g/kg and 18 g/kg by AC and BC, respectively, in 0.014 m3/s flow rate. In previous studies, different adsorbent materials have varied in their Zn2+ sorption capabilities. Biswas et al. (2019) recorded a maximum of 40.6 mg/g adsorption using Ca-alginate-biochar composite (CABC) adsorbent in a constant 50 mg/L concentration, while 7.48 mg/g and 4.98 mg/g were adsorbed using hardwood and corn straw, respectively, from 100 mM Cu(II) and Zn(II) stock solution (Chen et al. 2011).

In a batch experiment of Cu and Zn adsorption from a mine-contaminated water, Rodríguez-Vila et al. (2018) observed lesser Zn2+ adsorption in circumneutral mine discharged than the acid mine drain (AMD) water. This adsorption was attributed to a higher Zn2+ concentration of 14.16 mg/L than 8.98 mg/L in the circumneutral mine, which is slightly lower than Frongoch discharge at 15.8 mg/L concentration and 7.2 pH. Notwithstanding, these studies were carried out in laboratory conditions under a constant metal concentration and not flowing water like Frongoch watercourse. Hence, the need for a field trial where adsorption conditions are not regulated is suggested.

The adsorptive correlation analysis suggests that Fe2+ content, Na+ content, and surface area are possible mechanisms attributable to increased Zn2+ adsorption (Fig. 6a–f). The Fe2+ and Na+ contents in AC were 4.4 and 3.0 folds, respectively, greater than BC and suggestive of the 0.6 g/kg higher Zn2+ adsorption in AC. A previous study by Jiang et al. (2015) observed significant differences in Zn2+ adsorption with changes in Na2SO4 concentrations, which revealed that Na+ ion content enhanced Zn2+ adsorption. A similar attribute was observed in water hyacinth-activated biochar, where FeCl2-activated biochar exhibited a robust magnetic property and adsorption capability observed through FTIR spectra (Nyamunda et al. 2019). The BET surface area was positively correlated with Pb2+ and Zn2+ adsorption, but most statistically (p < 0.001) correlated with Zn2+ adsorption due to a 90.5% higher surface area of AC than other samples. The high porosity and large surface area of AC suggest to have increased adsorbent surface exposure for ion exchange and binding sites for heavy metals, which has been reported earlier (Chen et al. 2011; Jeyakumar and Chandrasekaran 2014; J. J. Zhao et al. 2019). In conformity with previous studies, Cibati et al. (2017) noted that commercial Norit AC was 1.9 times better than non-activated BC from Miscanthus biomass due to its 5.8 m2/g greater BET surface area. These notable factors suggest that with a higher BET surface area and more significant Fe2+ binding sites, AC had adsorbed more Zn2+ than BC. In a batch and column test of mixed metals, Ding et al. (2016) achieved 1.83 mg/g Zn2+ removal in a mixed 100 mg/L metal solution of Zn2+, Pb2+, Cd2+, Cu2+, and Ni2+ in 140 min using an alkaline-modified biochar surface area (873 m2/g). Xue et al. (2020) noted that in a medium where there is a competitive adsorption, the effect of adsorbed Pb2+ on biochar surface could also alter the sorption sites for Zn2+ on BC, while pH positively correlates with adsorption of both metals but not statistically significant, which is similar to this study.

Fig. 6
figure 6

Zn2+ adsorption correlations with adsorbents’ physicochemical properties. a K+ ion. b Ca2+ ion. c Mg2+ ion. d Na+ ion. e BET surface area. f pH level. Pearson’s R2 coefficients and significance (p) values indicated at 95% confidence level

Full size image

3.3 Adsorption Equilibrium Over Time and Possible Adsorbent Modification to Optimize Sorption Capabilities

Comparative adsorption equilibrium between Pb2+ and Zn2+ at different intervals is based on the adsorption curve shown in Figs. 3 and 5. Pb2+ adsorption declined from 3.3 to 0.6 ppm/h in BC, while Zn2+ from 6.0 to 0.3 ppm/h between the first and last sampling intervals. These adsorption differences show that BC has more potential to sorb more Pb2+ than Zn2+. The gradual sorption decrease towards equilibrium suggests an effect of surface coverage, which is described as a chemisorption surface factor by a rapid initial surface reaction followed by slow secondary accumulation (Betts et al. 2013; Xue et al. 2020). Moreover, Zn2+ adsorption was relatively inconsistent due to metal desorption intervals and occurred in all samples, but notable in the most treated sample with the least surface area. Conversely, AC had the highest surface area but most desorbed metal after the initial sampling interval and never regained its optimal sorbed Zn2+. A regression analysis to decipher the adsorption relationship between Pb2+ and Zn2+ showed a positive correlation with treated and untreated biochar samples, while AC was inversely or negatively correlated (Fig. S1). Hence, it suggests a competitive surface adsorption effect amidst multiple metal ions (Pb, Zn, Al, Cd, Fe, Ni, Co, Cu, and Cr) present in the contaminated water. Thus, Pb2+ over Zn2+ on biochar surface than activated carbon was favored, as reported in a previous study (Bearcock et al. 2010).

Generally, biochar with large surface area and pore volumes can provide sufficient binding sites for heavy metals, otherwise subjected to competitive and inconsistent adsorption (J. H. Park et al. 2016). Among several studies of competitive metal adsorption in aqueous solution using various adsorbents (Ni et al. 2019; J. H. Park et al. 2015a, b; Sellaoui et al. 2019). J. H. Park et al. (2016) observed higher adsorption of Pb2+ in a multimetal aqueous solution containing Cu, Cr, Zn, and C, irrespective of the metals’ electronegativity. This affinity was attributed to the smaller hydrated radius of Pb2+ (4.01 Å) compared to Zn2+ (4.30 Å), and more significant affinity to functional groups in the organic matter, which include phenolic and carboxylic groups. Other reported attributes include its higher atomic weight and radius, hydrolysis constant, and Misono softness value (J. H. Park et al. 2016). Sellaoui et al. (2019) also observed an antagonistic reduction in Zn2+ adsorption by 66% as well as 10% and 9.8% of Hg2+ and Pb2+ adsorption in Hg2+ and Pb2+ multi-metallic solution. The endothermic and competitive adsorption was explained by a density functional theory (DTF), which noted that atomic binding energies of metals contribute to adsorbent performance, depending on the ions present. Thus, by this study, Pb2+ sorption application using BC adsorbent in the Frongoch watercourse is recommended at an optimal sampling interval of 1104 h at 0.6 ppm/h and Zn2+ at 432 h (0.9 ppm/h), at a concentration of 0.62 mg/L and 0.014 m3/s flow. Similarly, using AC adsorbent is recommended at 1104-h interval at 0.5 ppm/h for Pb2+ adsorption while Zn2+ at 72 h (19.0 ppm/h).

3.3.1 Recommended Amendments for Optimal Sorption

The adsorbents’ physicochemical properties and correlation with metal sorption show that ash amendment after pyrolysis altered the surface area necessary for metal binding. Therefore, it is suggestive of further trial applications for improvement. Firstly, by amendment process of biomass co-pyrolysis with ash. This co-amendment has recorded a successfully increased BET surface area of adsorbent biomass in previous studies (Z. Wang et al. 2019b; Ye et al. 2016). Furthermore, hydroxyl group modifications such as NaOH, KOH, NH4OH, and HNO3 before or after pyrolysis have been observed to improve porosity and surface area of biochar (Li et al. 2017; J.-H. Park et al. 2015a, b; Senthilkumar and Prasad 2020; X. Yang et al. 2019b). In a previous study, Ding et al. (2016) modified hickory wood biochar using alkaline (20 mL of 5 M NaOH) solution at 70 °C for 4 h to increase biochar surface area from 256 to 873 m2/g and CEC from 45.7 to 124.5 cmol/kg. Indications of the surface functional group show that alkaline modification decreased the surface C from 86.7 to 57.9% while increased surface O from 12.0 to 33.3%, to yield an increased atomic ratio of surface O/C from 0.14 to 0.58. The alkaline (NaOH) modification increased the specific surface areas and binding energy of target heavy metal ions (Pb2+, Cd2+, Cu2+, Zn2+, and Ni2+), which has also been recorded in most alkaline-modified studies (Cazetta et al. 2011; Uchimiya et al. 2011). Also, OH modification; Fe impregnation using alpha-FeOOH, alpha- and gamma-Fe2O3, CuFe2O3, FeCl2, and FeCl3; and saturation have been recorded in previous studies (Godlewska et al. 2020; Nyamunda et al. 2019; L. Yang et al. 2019a; T. Zhao et al. 2020). In a previous study, Nyamunda et al. (2019) saturated 200 g of dry water hyacinth biomass in a 1000 mL (2 M) mixture of FeCl2 and FeCl3 with an additional 5 M NaOH to increase its pH to 10 under rigorous magnetic stirring for 30 min. The modification increased its surface area from 610.88 to 1038 m2 g−1 at 450 °C pyrolysis temperature. Nevertheless, due consideration to biomass type is recommended. A reconstruction of the installation pool site might be necessary to channel all contaminated mine water through the designed adsorbent biochar column for optimal sorption efficiency.

4 Conclusion

This study provides the opportunity to utilize low-cost adsorbent materials through pyrolysis of waste pinewood feedstock (leylandii) to limit the exposure pathway of Pb2+ and Zn2+ contamination from abandoned mines. The untreated BC and purchased activated biochar AC are better promising absorbents for minimizing Pb2+ and Zn2+ pollution. In 1104 h (46 days), a total net of 33 g/kg and 25 g/kg of Pb2+ was adsorbed using 50 g of BC and AC, respectively, while a total net of 30 g/kg and 18 g/kg of Zn2+ was adsorbed by AC and BC respectively.

Analysis of adsorption results revealed that adsorbent physicochemical properties such as K+, Ca2+, Mg2+, Na+, BET, and pH are primary mechanisms attributed to the differences in Pb2+ and Zn2+ adsorption. The BC samples contained 570.80 ppm Ca2+ and 640.72 ppm Mg2+ mineral ions which enhanced its surface for metal binding sites. More also, Fe2+ and Na+ at 958.09 ppm and 1635.01 ppm in AC correlatively increased Zn2+ adsorption. High BET surface area enhanced metal adsorption mostly in untreated BC and AC samples, while post-ash amendment supplement significantly altered treated adsorbents’ properties BB1–4 and CBB, notably the surface area, which provides binding sites for ion exchange.

Zn2+ metal desorption suggests the possibility of ion exchange by competitive adsorption, which slightly favored Pb2+ uptake in the multimetal-contaminated water by displacement interaction. Previous studies attributed similar ion exchange effect to electronegativity, a smaller hydrated radius of Pb2+, and a more significant affinity to functional groups such as phenolic and carboxylic groups. Thus, it should be a factor to be considered in further studies.

Recommended modifications to improve adsorbent physicochemical properties include hydroxyl group modifications such as NaOH, KOH, NH4OH, and HNO3, before or after pyrolysis, and also OH modification and Fe impregnation using alpha-FeOOH, alpha- and gamma-Fe2O3, CuFe2O3, FeCl2, and FeCl3, as recorded in previous studies.