Interactive Relationship between Cementitious Materials and Acid Mine Drainage: Their Effects on Chromium Cr(VI) Removal

Abstract

Elevated hexavalent chromium (Cr(VI)) levels in pervious concrete may undermine its successful application in water treatment. Portland cement CEM I 52.5R (CEM I), coal fly ash (FA), natural zeolite and ground granulated blast-furnace slag (GGBS) were evaluated as adsorbents for removal of Cr(VI) from acid mine drainage (AMD). Adsorption experiments were conducted at dosages of 6, 10, 30 and 60 g of adsorbent in 200 mL of AMD, while the mixing contact time was varied from 15 to 300 min. It was found that the use of CEM1 and FA adsorbents strongly increased the Cr(VI) concentration in AMD. Conversely, zeolite and GGBS removed up to 76% and 100% of Cr(VI) from AMD, respectively, upon their use at dosages of at least 10 g of the adsorbent. Freundlich isotherm was found better fitted with a high correlation coefficient (R² = 0.998 for zeolite and 0.973 for GGBS) than to the Langmuir model (R² = 0.965 for zeolite and 0.955 for GGBS). Adsorption and ion exchange seem to be active mechanisms for the Cr(VI) removal. These results suggest that zeolite and GGBS can be considered as partial cement replacement materials for effective reduction or removal of Cr(VI) from the treated water.

1. Introduction

Chromium exists in various oxidation states ranging from Cr(II) to Cr(VI). Among these states, Cr(III) and Cr(VI) are the most common and most stable species [1,2]. Different species of chromium originate from its industrial applications, as well as incineration facilities, cement, contaminated landfill, asbestos lining erosion, tobacco smoke, topsoil and rocks [3]. The resulting industrial waste materials containing Cr(VI) are often disposed of into the environment, leading to contamination of natural water resources and soils, amongst others. The Cr(VI) varieties found in water are Cr₂O₇²⁻, CrO₄²⁻, H₂CrO₄, and HCrO₄⁻ [4]. Of these, the variety that becomes prevalent in aqueous solution is determined by pH of the solution, chromium concentration, the presence of oxidising and reducing compounds, and the redox potential [3]. Cr₂O₇²⁻ and HCrO₄⁻ are the predominant varieties at the pH range of 2.0 to 6.0. At pH above 6.0, the dominant species is CrO₄²⁻ [3,4]. Equations (1) and (3) show the dissolved forms of Cr(VI) typically present in solution, as determined by the solution pH level.

H₂CrO₄→H⁺ + HCrO₄⁻                (pH 2.0 to 6.0)

(1)

2HCrO₄⁻→Cr₂O₇²⁻ + H₂O             (pH 2.0 to 6.0)

(2)

HCrO₄⁻→H⁺ + CrO₄²⁻                (pH > 6)

(3)

Fly ash (FA), ground granulated blast-furnace slag (GGBS) and natural zeolite, all exhibit pozzolanic properties and are conventionally used in concrete as partial cement replacement materials for improvement of durability, among other beneficial effects [5,6,7,8,9]. Various researchers have demonstrated the potential of FA as an effective adsorbent for wastewater treatment [10,11,12]. Partial replacement of pervious concrete with FA has been shown to improve the workability of fresh concrete and the mechanical strength and durability of hardened concrete [13]. Fly ash can be utilised for a cleaner production of pervious concrete possessing compatible hydrological property and pollution control potential, compared to the ordinary pervious concrete [14]. Zeolites are low-cost ion exchangers that have been used as adsorbents for removal of heavy metals including chromium, etc., [15,16] from polluted water. Due to their net negative charge, zeolites have a strong affinity for transition metal cations. Zeolites are often abundantly available in natural deposits; they also exhibit high chemical stability [17,18,19]. A recent investigation by Bae et al. [20] found that the replacement of Portland cement with a small amount of only 5 wt% GGBS, significantly reduced the dissolved Cr(VI) in aqueous solution. There was complete sorption and reduction of Cr(VI) to Cr(III) using GGBS in a Ca(OH)₂ solution of pH > 12.5.

Cr(VI) is an extremely toxic chromium species that is known to cause severe environmental and health problems [21] to human, animal, and aquatic life systems. Such adverse health effects of Cr(VI) include contact dermatitis, lung carcinoma, diarrhoea, ulcers, kidney failure, liver damage, and other diseases of the gastrointestinal organs [3,22,23,24,25]. Not only is Cr(VI) highly carcinogenic and mutagenic, it is also known to cause birth defects [3]. Inhalation of Cr(VI) can cause asthma, bronchitis, pneumonitis, inflammation of the larynx and liver, perforation of the nasal septum, and increased risks of cancer of the respiratory system [21,22,23,26]. Cr(VI) is classified as a Group 1 human carcinogen by the International Agency for Research on Cancer [27] and was identified by the United States Environmental Protection Agency (USEPA) as one of the 17 chemicals that pose a threat to human life [28]. The World Health Organization specifies 0.05 mg/L as the maximum permissible limit of Cr(VI) in water for domestic use [29]. Given the severe health hazards of Cr(VI), it is a crucial necessity to ensure its removal from any sources that may lead to its direct contact with life systems.

The conventional approach for Cr(VI) removal from wastewater involves its reduction from the hexavalent to the trivalent state, followed by its precipitation as Cr(III) hydroxide, by means of physical, chemical, or bioremediation methods. Physical methods employ the physico-chemical properties of materials to achieve chromium remediation through various mechanisms including adsorption, electrodialysis, membrane filtration, photocatalysis, amongst others. Chemical remediation utilises chemicals such as sulphur dioxide, sodium metabisulfite, ferrous sulfate, sodium sulfite, barium sulfite, lime and limestone for the reduction of Cr(VI) to Cr(III). Bioremediation refers to the use of living organisms including bacteria, fungi, yeast, algae, and plants to remove pollutants [3]. The problems associated with use of physico-chemical methods for Cr(IV) removal include the high operating costs of the treatment, high energy consumption, excessive use of chemicals, generation of toxic sludge and air pollution resulting from the use of sulphur-based reducing agents [22,30]. These reductants generate SO₂ or H₂S under strong acidic conditions and emit an unpleasant odour, both of which seriously affect the health of workers [10]. Although bioremediation for chromium removal from wastewater may show good performance under laboratory testing, it is not often suitable for large-scale field systems, since the living organisms utilised for the treatments are not naturally available in abundance and have to be grown or cultured [1].

Adsorption offers the benefits of a clean, cost-effective, easily controlled, efficient process [4]. Natural bio-adsorbents, such as sawdust, have been successfully used for adsorption of Cr(VI). However, low adsorption capacities, high chemical, and biological oxygen demand, as well as leaching of organic components remain the main drawbacks of applying biosorbents [31]. Natural polymer-based adsorbents offer the benefits of low-cost, high effectiveness and regeneration potential. Examples of natural polymer-based adsorbents are chitin [32], cellulose [33] and chitosan [34]. Chitosan contains chemical functional groups, such as NH₂ and OH, while cellulose is mostly dominated by OH⁻ and CHO⁻ groups that can aid in trace metal sequestration and complexation. [31]. Carbon-based adsorbents, such as activated carbon [35] and biochar [36], possess a vast network of internal pores by which they have high surface areas and become excellent adsorbents [37]. Other adsorbents that have shown varying success in Cr(VI) adsorption include bentonite clay, zero-valent iron, bauxite and iron oxide nanoparticles [38,39,40,41]. However, most of the adsorbents show limitations in adsorption capacity and are often not re-usable [4].

Due to its particle retention capacity during filtration, concrete products can function as pollution sink [42,43]. Contaminants, such as heavy metals and organic compounds, are absorbed into the internal concrete body and are eliminated from the runoff. Other potential benefits of using this type of pavement include recharging of groundwater, saving of water by recycling, and prevention of pollution [44]. Reductions in suspended solids, biochemical oxygen demand, chemical oxygen demand and ammonia levels in surface runoffs demonstrate the high treatment efficiency of pervious pavement systems [45]. It can also lead to a reduction in oil, grease, and petroleum products, from the water effluent drained through pervious concrete [43]. Several recent studies [43,44,45,46,47,48,49] have shown pervious concrete to be a promising alternative reactive material for the treatment of acid mine drainage (AMD). However, elevated Cr(VI) levels in the treated water arising from the use of cement and FA in pervious concrete, may potentially undermine concrete’s successful application in water treatment. Chromium in cement is found primarily in the form of Cr(III), while Cr(IV), Cr(V) and Cr(VI) may exist in smaller quantities [50,51]. The oxidising environment in kilns under the typically high operating temperatures of up to 1450 °C, transforms the Cr(III) found in the materials into hexavalent state [21]. Cr(VI) in cement emanates from different sources including some refractory bricks, raw materials, fuels, and the chromium alloys used in grinding mills. The release of Cr(VI) into the environment when cement or concrete products are exposed to water, soil and air, may result in a threat to environmental safety and human health.

In the present study, adsorption experiments were conducted on Portland cement CEM I 52.5R (CEM I), FA, natural zeolite and GGBS, to assess their effects as adsorbents for effective removal of Cr(VI) from AMD. The present study aimed at identifying effective adsorbent(s) that may be incorporated into pervious concrete for removal of Cr(VI) from the concrete-treated AMD. Batch adsorption experiments were implemented under different criterions, including adsorbent dosage, contact time, and pH. Fourier-transform infrared spectroscopy (FTIR), X-ray powder diffraction (XRD), and scanning electron microscopy-energy-dispersive X-ray spectroscopy (SEM-EDS), were used to investigate the physico-chemical properties of the adsorbents.

2. Materials and Methods

Naturally occurring AMD was obtained from an abandoned coal mine for use in the experiment. The AMD was collected from its discharge source using high-density polyethylene containers and transported to the laboratory for use in the experiments. The pH of the raw AMD was 3.01, while its composition [47] showed high concentrations of Ca (470 mg/L), Mg (214 mg/L), Na (3061 mg/L), Fe (9 mg/L), Al (6 mg/L) and SO₄ (2870 mg/L). The concentration of Cr(VI) in the raw AMD was 0.042 mg/L.

Adsorption tests were conducted on CEM I; Class F, FA; natural zeolite and GGBS as adsorbents for Cr(VI) removal from AMD. The CEM I, FA, and zeolite materials used, were obtained from AfriSam SA (Pty) Ltd. (Roodepoort, South Africa), Ash Resources SA (Pty) Ltd. (Edenvale, South Africa), and Serina Trading (Pty) Ltd. (Heidelberg, South Africa), respectively. GGBS was also supplied by AfriSam SA (Pty) Ltd. Phase identification of the unreacted and reacted adsorbents was carried out using X-ray diffraction (XRD) (Malvern Panalytical Ltd., Malvern, UK) at a scan rate of 0.02°, 2-theta per minute using a Pan Analytical X-ray X’pert PRO diffractometer (Malvern Panalytical (Pty) Ltd., Malvern, UK). The XRD instrument consisted of a PW3830 X-ray generator operated (Malvern Panalytical (Pty) Ltd., Malvern, UK) at 40 kV, 40 mA and a copper X-ray tube. Morphological characterisation of the adsorbents was conducted by scan electron microscopy–energy dispersive spectrometer (SEM–EDS) (TESCAN VEGA 3 SEM, (Tescan Orsay Holding, Brno, Czech Republic). The SEM samples were cross-section cut and dried under vacuum. Then, the samples were fixed on the aluminium stub and sputter-coated with gold particle. An AZtec energy dispersive spectrometer (EDS, Oxford instruments, Abingdon, UK) was employed to determine elemental composition of phases in the samples. Functional groups were determined using BX-II PerkinElmer FTIR (Perkin Elmer, Waltham, MA, USA) equipped with the universal attenuated total reflectance (ATR) diamond crystal. Then, 0.5 g of the adsorbent is placed into contact with the ATR sampling crystal and subjected to IR radiation. Some of this radiation is absorbed by the sample and is measured as an infrared spectrum. The ATR technique gave a spectral range from 4000 to 400 cm⁻¹.

The concentration of Cr(VI) in AMD was determined using a WLAB/046/Discrete Chromium Analyser (Skalar Analytical B.V., Breda, The Netherlands) based on the American Public Health Association, Standard Method 3500 Cr B Colorimetric Method [52]. The AMD sample was reacted with diphenylcarbazide under acid conditions (pH of 2 ± 0.5). Cr(VI) produces a red-violet colour which is measured photometrically at wavelength 540 nm. The limit of detection (LOD) was 0.0048 mg/L. The amount of Cr(VI) contaminant adsorbed (q_e) in mg/g and the removal efficiency (RE) levels were calculated using Equations (4) and (5), respectively [53].

$${\mathrm{q}}_{\mathrm{e}} = {(\mathrm{C}}_{\mathrm{o}} -{ \mathrm{C}}_{\mathrm{e}}) \times \frac{\mathrm{V}}{\mathrm{m}}$$

(4)

$$\mathrm{RE} (\%) = \frac{{\mathrm{C}}_{\mathrm{o}} -{ \mathrm{C}}_{\mathrm{e}}}{{\mathrm{C}}_{\mathrm{o}}} \times 100$$

(5)

where C_o is the initial concentration of the contaminant in raw AMD (mg/L), C_e is the equilibrium concentration of the contaminant (mg/L), V is the volume of AMD (litres), m is the mass of the adsorbent (g).

Batch adsorption experiments were carried out in 500 mL graduated borosilicate glass beakers. The experiments were performed using 200 mL of AMD containing 0.042 mg L⁻¹ of Cr(VI) at 25 °C and 150 rpm for 300 min. Adsorption efficiency of Cr(VI) by the adsorbents was studied by varying the adsorbent dosage. The following dosages were used for the individual adsorbents: 6.0, 10.0, 30.0 and 60.0 g. A high precision electrical balance (Ohaus Scout Pro Portable Electronic Balance) was used for weighing. The mixture was filtered through a 0.45 μm Whatman^® PTFE membrane filter. Treated AMD samples were transferred into 200 mL plastic vials. The pH, Electrical Conductivity (EC) and Total Dissolved Solids (TDS) were measured using a digital pH meter (MP-103 microprocessor-based pH/mV/Temp tester (Gondo electronic Co. Ltd., Taipei, Taiwan). The pH electrode used was calibrated using standard NIST-traceable pH 2.0, 4.0, 7.0 and 10.0 buffers. Treated AMD samples were stored in a refrigerator set at a constant temperature of 3 °C before the determination of Cr(VI) concentration.

Column tests were conducted to determine the leaching behaviour of Cr(VI) from CEM I, FA and GGBS. Pervious concrete was made, using 6.7 mm granite aggregate and Portland cement consisting of CEM I 52.5R (CEM I) alone or CEM I with 30% fly ash or CEM I with 50% GGBS. A mixture of 0.27 water/cementitious ratio was used. The mixture was placed into standard 100 mm cube moulds and then compacted using a vibrating table. The fresh concrete cubes were de-moulded after about 24 h and placed in a water curing bath until 28 days of age [47]. Four concrete cubes of CEM I, 30%FA and 50% GGBS were placed in separate columns. The columns used in this study were 500 mm in height, and had an internal diameter of 100 mm. An AMD sample was pumped at a flow rate of 0.35 mL/min. Leachate samples were collected daily for the first three months, then once every third day, and thereafter once a week for a total of 320 days. The concentrations of Cr(VI) were determined using WLAB/046/Discrete Chromium Analyser.

Langmuir and Freundlich isotherms are the most commonly used models for adsorption studies. Equation (6) gives the Langmuir model which can also be expressed in the linearised form given in Equation (7). By plotting the graphs of adsorption data as 1/q_e versus 1/C_e, a trendline is obtained whose slope and intercept is used to directly determine the values of $q_m$ and $K$, based on Equation (7) [26,54].

$${\mathrm{q}}_{\mathrm{e}} = \frac{{\mathrm{q}}_{\mathrm{m}}{\mathrm{KC}}_{\mathrm{e}}}{1 +{ \mathrm{KC}}_{\mathrm{e}}}$$

(6)

$$\frac{1}{{\mathrm{q}}_{\mathrm{e}}} = \frac{1}{{\mathrm{Kq}}_{\mathrm{m}}} \times \frac{1}{{\mathrm{C}}_{\mathrm{e}}} + \frac{1}{{\mathrm{q}}_{\mathrm{m}}}$$

(7)

where q_e (mg/g) is the amount of contaminant adsorbed (adsorbate) at the equilibrium concentration, C_e (mg/L) is the equilibrium concentration of adsorbate, q_m (mg/g) represents the maximum monolayer adsorption capacity, and K (l/mg) is the Langmuir constant related to energy of adsorption and the affinity of the binding sites. The essential characteristics of the Langmuir isotherm can be expressed in terms of a dimensionless constant called the separation factor, R_L, as shown in Equation (8) [33]:

$${\mathrm{R}}_{\mathrm{L}} = \frac{1}{1 +{ \mathrm{K}}_{\mathrm{L}}{\mathrm{C}}_0}$$

(8)

where C₀ is the maximum initial concentration of the adsorbate (heavy metal ion) and K_L(L/mg) is a constant related to the affinity of the binding sites. When the values of ${\mathrm{R}}_{\mathrm{L}}>1$, it is an indicator that adsorption is unfavourable, ${\mathrm{R}}_{\mathrm{L}} = 1$ indicates linear adsorption, favourable when 0 < ${\mathrm{R}}_{\mathrm{L}}<1$ and irreversible when ${\mathrm{R}}_{\mathrm{L}} = 0$.

The Freundlich isotherm of Equation (9) accounts for multilayer physico-chemical adsorption on heterogeneous surfaces and can be expressed in a linearised form as given in Equation (10).

$${\mathrm{q}}_{\mathrm{e}}={ \mathrm{K}}_{\mathrm{F}}·{\mathrm{C}}_{\mathrm{e}}{}^{\frac{1}{\mathrm{n}}}$$

(9)

where C_e is the concentration of the adsorbate at equilibrium (mg/L), K_F is the Freundlich capacity factor, and 1/n is the intensity. The latter two parameters are obtained by fitting Equation (9) to graphs of adsorption data plotted as $\log {\mathrm{q}}_{\mathrm{e}}$ versus $\log { \mathrm{C}}_{\mathrm{e}}$.

$$\log {\mathrm{q}}_{\mathrm{e}} = \frac{1}{\mathrm{n}}\log {\mathrm{C}}_{\mathrm{e}} +{ \mathrm{logK}}_{\mathrm{F}}$$

(10)

3. Results

3.1. Characterisation of Adsorbents

Chemical analyses of the adsorbents were performed using X-ray fluorescence (XRF, Bruker, Karlsruhe, Germany), giving the chemical compositions shown in

Table 1. Chemical compositions of Portland cement (CEM I), fly ash (FA), natural zeolite and ground granulated blast-furnace slag (GGBS). LOI = Loss on Ignition.

Adsorbent	SiO2	Al2O3	CaO	Fe2O3	MgO	TiO2	Mn2O3	SO3	Na2O3	K2O	LOI
CEM I (%)	21.9	4.75	65.44	3.68	2.17	0.49	0.40	1.92	0.17	0.25	1.57
FA (%)	50.32	24.57	7.31	5.91	1.83	1.53	0.05	0.16	0.16	0.76	5.59
Zeolite (%)	58.12	11.44	1.01	1.57	1.30	0.12		0.18	2.03	1.44	22.36
GGBS (%)	37.03	13.39	36.62	0.61	8.00	0.63	0.87	2.31	0.23	1.11	0.19

. The high CaO and SiO₂ contents of CEM I are responsible for strength development in concretes. FA and zeolite are pozzolanic materials composed of mainly amorphous aluminosilicate elements, while GGBS is a latent hydraulic cement exhibiting both pozzolanic and cement properties. It can be seen in Table 1 that GGBS had intermediate levels of all the three oxides comprising CaO, SiO₂ and Al₂O₃. The typical concentrations of total chromium Cr₂O₃ concentrations in South African CEM I, FA, GGBS are 102, 200 and 193 ppm, respectively [51,55,56]. Potgieter et al. [51] also reported that 30 to 80% of the total chromium in cement is Cr(VI) of which 8 to 26% is leachable or water soluble. A study by Eštoková et al. [57] stated that the average concentrations of the total chromium in cements vary from 178.5 to 257.3 mg per kg of cement while the average concentrations of hexavalent chromium ranged from 0.5 to 2.46 mg/kg.

3.2. Cr(VI) Adsorption Tests

3.2.1. Effect of Adsorbent Dosage

It can be seen in Figure 1a that CEM I showed an increase in Cr(VI) concentration with increase in dosage of the adsorbent, attaining a maximum concentration of 4.27 mg/L at the dosage of 60 g of the adsorbent in 200 mL of AMD. As mentioned in Section 1, chromium is present in Portland cement as an impurity. In recent studies [47,48,49], it was reported that mine water treated with pervious concrete exhibited high levels of Cr(VI) owing to leaching from the cement and fly ash materials used in PERVC mixtures. FA also increased the concentration of Cr(VI) in AMD, but to a lesser extent than CEM I. Cr(VI) increased from 0.042 mg/L in raw AMD to 0.211 mg/L and 0.535 mg/L when 6 g and 60 g of FA were mixed with 200 mL of AMD, respectively. Clearly, both CEM I and FA increased the Cr(VI) concentration from 0.042 mg/L in raw AMD to levels exceeding the maximum permissible limits of 0.1 mg/L and 0.05 mg/L, in the treated AMD, specified in Environmental Protection Agency (EPA) [58] and World Health Organization (WHO) [29] or National Water Act (NWA) [59], respectively. The observed elevation of Cr(VI) in the AMD following its treatment using CEM I or FA, is attributed to leaching of the chromium present in the powder materials, into the polluted mine water [51]. Interestingly, Cr(VI) in AMD was removed or significantly reduced when the zeolite and GGBS adsorbents were used, as seen in Figure 1b. The Cr(VI) concentration in AMD reduced with an increase in the GGBS dosage, achieving complete removal at the dosage of 30 g of the adsorbent in 200 mL of AMD. Bae et al. [20] suggested that chromium removal by GGBS, could be attributed to the dissolved anions from slag, which may serve as reducing agents to transform Cr(VI) to Cr(III). Additionally, in the present study, a small dosage of only 6 g of zeolite in 200 mL of AMD, was sufficient for effective reduction of the Cr(VI) concentration level.

3.2.2. Effect of Contact Time

In the experiment, contact time between the adsorbent and AMD was varied from 15 to 300 min. The CEM I and FA adsorbents strongly increased the Cr(VI) concentration in AMD, more so with longer contact time, as seen in Figure 2a. Notably for FA, there was no significant change in the concentration of Cr(VI) after 15 min of contact time. After 300 min of contact, the Cr(VI) concentration had risen from 0.042 mg/L in raw AMD to equilibrium levels of 2.05 mg/L for CEM I and 0.558 mg/L for FA, both of which exceed the maximum limits of 0.10 and 0.05 mg/L specified in EPA [58] and NWA [59] or WHO [29], respectively. It can be seen in Figure 2b that zeolite and GGBS showed rapid removal of Cr(VI) at the early ages, with near complete removal of the contaminant occurring within 15 min of contact between the adsorbent and AMD. The Cr(VI) concentration was reduced from 0.042 mg/L in raw AMD to equilibrium concentrations 0.0107 mg/L and 0.002 mg/L after the water treatment using zeolite and GGBS, respectively. There was no significant change in concentration of Cr(VI) after 15 min of contact, indicating that the adsorption phase had reached equilibrium. Evidently, a minimum contact time of 15 min was required for removal or reduction of Cr(VI) by zeolite or GGBS to low concentration levels that meet the limits specified in national water standards [58,59].

3.2.3. Change in the pH of AMD as a Function of Adsorbent Dosage and Its Influence on Cr(VI) Adsorption

Adsorption of Cr(VI) varies as a function of the pH. In dissolved form, Cr(VI) can be present in any of the different forms H₂CrO₄, HCrO₄⁻, CrO₄²⁻ or Cr₂O₇²⁻ [4]. At low pH, the H⁺ ions tend to protonate at the surface of adsorbent, resulting in a positively charged surface, which in turn has strong affinity for the negatively charged HCrO₄⁻ [3]. At higher pH, a decrease in proton concentration leads to a negatively charged adsorbent surface which has less affinity for the adsorption of HCrO₄⁻. At higher pH, interferences from OH⁻ ions may lead to lower removal efficiency of Cr(VI). Increased amounts of OH⁻ increases the competition between Cr(VI) and OH^- for occupying exchange sites in absorbent pore [60]. A similar pH dependent trend was also observed by a number of researchers for the Cr(VI) removal by various adsorbents [1,3,34,61,62]. At the adsorbent dosage of 10 g in 200 mL of AMD, the maximum pH levels attained by CEM I, FA, zeolite and GGBS were 12.0, 8.4, 4.8 and 9.2, respectively, as seen in Figure 3. At the higher adsorbent dosage of 60 g in 200 mL of AMD, the respective pH levels attained were significantly higher, giving 13.2, 10.1, 6.7 and 10.6 for CEM I, FA, zeolite and GGBS. In all cases, however, the pH rise was rapid and complete within 15 min of contact between AMD and the adsorbents, which also coincided with the maximum Cr(VI) removal levels achieved by zeolite and GGBS. In the AMD that was treated using CEM I and FA, the concentration of Cr(VI) strongly increased as pH of the AMD increased—a concern attributed to the overriding effect of the leaching of chromium from the absorbents into AMD. Cr(VI) in cement emanates from sources such as refractory bricks, fuels, and the chromium alloys; Chromium is present in coal mainly as trivalent chromium (Cr(III)). During the combustion of coal, Cr is usually oxidized from Cr(III) to Cr(VI). When exposed to water, Cr(VI) may leach out from the adsorbents [63]. Maximum Cr(VI) concentration levels of 4.27 mg/L and 0.558 mg/L were attained at pH levels of 13.6/8.77 of the treated mine water, under the dosage of 60 g of CEM I/FA in 200 mL of AMD, respectively. In a study of Cr(VI) removal from aqueous systems, Kantar et al. [64] reported that Cr(VI) reduction efficiency decreased with increasing solution pH. In contrast, Cr(VI) reduction efficiency increased with increasing solution pH when GGBS was used as an adsorbent. This is because GGBS’s activation is strongly dependent on pH values. Enhanced Cr(VI) reduction by GGBS is achieved at pH values higher than 11.5 [20]. The effects of increased alkalinity on efficient removal of Cr(VI) by GGBS was confirmed by the XRD/SEM results in Section 3.4.

3.2.4. Removal Efficiencies

The RE levels of Cr(VI) removal were determined for the various adsorbents, as summarised in

Table 2. Removal Efficiency (RE) levels for Cr(VI) removal by Portland cement, fly ash, natural zeolite and ground granulated blast-furnace slag.

Adsorbent	Parameter	Dosage of Adsorbent in 200 mL of AMD
		6 g of Adsorbent	10 g of Adsorbent	30 g of Adsorbent	60 g of Adsorbent
CEM I	RE (%)	−7519	−4781	−5305	−10,067
FA	RE (%)	−402	−1229	−990	−1174
Zeolite	RE (%)	76	71	66	75
GGBS	RE (%)	48	93	100	95

. For each adsorbent dosage, the average equilibrium concentration of Cr(VI) was calculated for the contact durations of 15 to 300 min. The average Ce was then used in the calculation of RE as per Equation (5). It can be seen in Table 2 that the GGBS adsorbent gave the highest RE, achieving 100% removal of Cr(VI) followed by 76% removal by zeolite. CEM I or FA did not remove or reduce Cr(VI) concentration, but these adsorbents instead increased the chromium levels in the treated AMD, thus giving negative RE values which apparently were also very high. At the dosage of 10 g of adsorbent in 200 mL of AMD, CEM I and FA gave the RE values of −4781% and −1229%, respectively. The high Cr(VI) levels in the CEM I- and FA-treated AMD are a result of undesirable leaching behaviours of the two adsorbents.

3.2.5. Isotherm Models

The Langmuir and Freundlich isotherm models were used to describe the equilibrium relationship between Cr(VI) and the adsorbents (zeolite and ground granulated blast-furnace slag). Correlation coefficient values are presented in

Table 3. Langmuir and Freundlich adsorption isotherm models determined for the adsorbents.

Adsorbent	Langmuir		Freundlich
Zeolite	qm (mg/g)	0.0262	n	0.596
	KL (L/mg)	−31.58	K	125.81
	RL	0.999	R2	0.998
	R2	0.965
GGBS	qm (mg/g)	0.0144	n	1.187
	KL (L/mg)	4.729	K	1.311 × 10−4
	RL	0.782	R2	0.973
	R2	0.955

.

Langmuir Isotherm

The correlation coefficients R² for the linear plot were >0.99 indicating that the experimental data of Cr(VI) adsorption using both adsorbents fitted well with the Langmuir isotherm model, as can be seen in Figure 4. Langmuir isotherm assumes monolayer adsorption on a homogenous surface [36]. The Langmuir dimensionless constant separation factor RL, was used to predict favourability of the isotherm. If RL < 1, then adsorption is favourable, while RL > 1 represents unfavourable adsorption, and RL = 0 irreversible process. According to the obtained results, RL was found to be 0.782 for GGBS, showing favourable adsorption of Cr(VI) by GGBS under the study conditions. RL value for zeolite was 0.999, indicating linear adsorption of Cr(VI) by zeolite.

Freundlich Isotherm

The linear form of Freundlich model equation is reported in Figure 5 for zeolite and GGBS, respectively. The R² values were 0.998 and 0.973 for zeolite and GGBS, respectively, indicating good correlation between the experimental data and Freundlich adsorption models for Cr(VI). Both the zeolite and GGBS possess high adsorption capacity for Cr(VI). The Freundlich equation can also be used to determine the adsorption intensity, 1/n, where n is an indicator of the change of intensity of the adsorption process—a value of n above 1 (n > 1) indicates favourable adsorption. A value of n below 1 (n < 1) indicates poor adsorption characteristics [65]. Our results showed n = 0.596 for zeolite, indicating poor adsorption of Cr(VI) by zeolite. The calculated n value for adsorption of Cr(VI) by GGBS was 1.187, showing good efficiency for Cr(VI) adsorption by GGBS adsorbent. The Freundlich model was found to fit the Cr(VI) adsorption better than the Langmuir model for both adsorbents, suggesting a multilayer adsorption of Cr(VI) on a heterogeneous surface. Other studies also reported that the Freundlich model described a much better fit than the Langmuir model in relation to Cr(VI) adsorption [37,38].

Comparison of Different Adsorbents for Cr(VI) Ions Removal

The maximum adsorption capacities of various adsorbents are presented in

Table 4. Comparison of maximum removal capacity of Cr(VI) by various adsorbents.

Adsorbent	Qmax (mg/g)	Reference
Zeolite	0.0262 (76% removal)	Present study
GGBS	0.0144 (100% removal)	Present study
Activated carbon	72.46	[35]
Dolachar	0.904	[65]
Zeolite modified ZVI	2.49	[66]
Illite carbon nanocomposite	0.57	[67]
Activated Akadama clay	2.17	[68]

. By comparison, it was found that zeolite and GGBS were also able to efficiently remove Cr(VI) ions from aqueous media. Most adsorbents presented on Table 4 have been chemically modified to achieve high adsorption capacities. In the current study, maximum adsorption capacities were obtained without modifying the adsorbents and the adsorption conditions.

3.3. Cr(VI) Leaching Tests

Column tests were conducted to determine the leaching behaviour of Cr(VI) from CEM I, FA and GGBS. After 320 days of the experiment, the concentration of Cr(VI) for CEM1, FA and GGBS concrete samples was 0.184, 0.069 and 0.059 mg/L, respectively (Figure 6). CEM I and FA showed higher leachability of Cr(VI) that exceeded the recommended maximum concentration specified by the World Health Organisation (WHO) of 0.05 mg/L, and therefore poses an environmental risk. This is in agreement with adsorption experiments that showed elevated Cr(VI) levels in the AMD following its treatment using CEM I or FA. Hartwich and Vollpracht [69] reported that higher chromium content in cement results in higher pore solution concentrations and a higher release during leaching.

3.4. Characterisation and FTIR Analysis

The XRD pattern of raw zeolite, as given in Figure 7a, identified the dominant minerals present to be the low-crystallised clinoptilolite ((Na,K,Ca)_2–3Al₃(Al,Si)₂Si₁₃O₃₆·12H₂O), quartz (SiO₂), and mordenite (Al₂Si₁₀O₂₄·7H₂O). In the reacted zeolite (Figure 7a), mordenite disappeared while a new phase muscovite appeared, indicating the transformation of the former to the latter phase. The co-existence of mordenite with the clinoptilolite in zeolites, was also observed by Diale et al. [70]. The high SiO₂ and Al₂O₃ in the zeolite adsorbent (Table 1) is responsible for its hydrophilic nature, an important property for adsorption of contaminants [71]. The cations Na, K, Ca and Mg found in zeolites generally, give the adsorbent its cation exchange capacity [16]. Figure 7b gives the XRD patterns of GGBS before and after its exposure to AMD. The wide hump between the 25° and 35° two-theta angles, shows that the GGBS was characteristically amorphous. The diffraction peak observed at the 30° two-theta angle of unreacted GGBS was identified as calcium carbonate (CaCO₃), most likely from the small amounts of limestone typically present in GGBS. It may be recalled that GGBS is a by-product of pig iron production from a mixture of limestone and forsterite or dolomite in some cases [72]. It can also be seen that the reaction between GGBS and AMD resulted in a significant formation of gypsum. GGBS is mainly composed of SiO₂, Al₂O₃ and CaO, as shown in Table 1. Again, the dissolution of CaO releases Ca ions, which combine with the SO₄²⁻ in the AMD solution and/or sulphuric acid from pyrite oxidation, leading to the formation of gypsum [47].

Figure 8a is an SEM micrograph of unreacted zeolite, showing clustered crystals of clinoptilolite plates. These structures provide a large surface area for fast and efficient adsorption of contaminants by the zeolite through ion exchange between cations of the adsorbent and contaminants in the mine water, including chromium [16,73]. The exchangeable ions Ca, K, Na found in unreacted zeolite as seen in the EDS of Figure 8b, could no longer be detected in the reacted zeolite, as evident in Figure 8c,d, indicating the occurrence of ion exchange between cations in zeolite and the positively charged ions in the mine water [73]. Jorfi et al. [60] reported that in aqueous solutions, chromium form oxyanions. At low pH values, these oxyanions are adsorbed onto the positively charged functional groups of zeolite. Therefore, the negatively charged Cr(VI) ions reduce the cationic exchange capacity and favour adsorption onto zeolite. The SEM micrograph of unreacted GGBS given in Figure 9a shows predominantly irregular particles of the slag. Figure 9b gives the EDS spectra for unreacted GGBS showing high content of C, Ca, Mg, Al and Si which are typical constituents of GGBS. When the CaO and MgO in GGBS (Table 1) react with water to form Ca(OH)₂ and Mg(OH)₂, respectively, an alkaline environment develops. With the resulting increase in pH, cement hydration products such as Ca(OH)₂, calcium-silicate-hydrate (C-S-H) and ettringite are formed [47]. These hydration products have a large specific surface area that provides physical sorption and co-precipitation sites for heavy metal ions such as Cr(VI) [20]. Cr(VI) may have also co-precipitated with gypsum, which was identified as the dominant product formed from the reaction of GGBS with AMD. GGBS particles also had a large surface area and high porosity levels, both of which promote metal sorption. Figure 9c gives the SEM micrograph of reacted GGBS showing an accumulation of gypsum inside an existing pore. The EDS for the reacted GGBS, as given in Figure 9d, shows a remarkable decrease in intensity levels of Ca²⁺, Al³⁺ and Si⁴⁺ suggesting their consumption from the slag during AMD treatment.

FTIR spectra of the reacted and unreacted natural zeolite are given in Figure 10a, indicating the complex formations between the contaminants in AMD and different functional groups of the adsorbent. It can be seen that adsorption of contaminants was particularly associated with the silicon oxide (Si–O), the carboxylate anion (C=O) and hydroxyl (–OH) groups [74]. The spectrum of the unreacted zeolite indicates the presence of Si–O symmetric bands of quartz with a double peak at the wavenumbers 778 to 797 cm⁻¹, which majorly shifted to a single peak at 797 cm⁻¹ of the reacted zeolite. The –OH group in the unreacted zeolite shows a signal at 2988 cm⁻¹ which also shifted to form a strong broad band between 3264 cm⁻¹ and 3395 cm⁻¹, denoting strong stretching vibrations in the reacted zeolite. Both the Si–O and –OH groups show the shifting of bands to higher numbers, indicating the adsorption of relatively lighter contaminant elements by the functional groups. The –OH group in the reacted zeolite also gave a strong increase and broadening of transmittance. The carboxylate C=O stretching at 1634 cm⁻¹ showed no major shift but gave a strong increase and broadening of its transmittance peak. The observed broadening of the peaks belonging to the –OH and C=O groups in reacted zeolite, indicates an increase in contaminant concentrations and possible formation of crystalline phases, which points to chemisorption through cation exchange. In Figure 10b, the FTIR spectra of unreacted GGBS exhibit a band between wavenumbers 876 cm⁻¹ and 963 cm⁻¹, representing stretching vibrations of the AlO₄⁻¹ group along with the related antisymmetric stretching vibrations of (Al)–O, respectively. The weak band at wavenumber 714 cm⁻¹ is ascribed to Si–O–Si(Al) bridges, linked or related to the SiO₄ tetrahedral [75,76]. The transmittance wavenumbers (876 to 963, 714 cm⁻¹) observed in the unreacted slag were strongly reduced in the reacted GGBS, which may indicate the role of Al and Si towards effective contaminant removal by slag [76]. In a study of chromite and slag interactions by Lee and Nassaralla [77], it was reported that a replacement of Al₂O₃ in slag by SiO₂, reduced Cr(VI) formation. It was further shown that use of calcium silicate slags led to the formation of much smaller levels of Cr(VI) from chromite, while calcium aluminate slags gave very high chromium levels. The strong band at 1477 cm⁻¹ indicates the presence of C–O stretching mode of calcite, with no role in contaminant adsorption as there was no major shift or change in its transmittance peak.

4. Conclusions

An investigation was conducted on Portland cement, fly ash (FA), natural zeolite and ground granulated blast-furnace slag (GGBS), as adsorbents for Cr(IV) removal from acid mine drainage (AMD). The following findings have been drawn.

• Effective removal of Cr(VI) from AMD was achieved using at least 6 g of zeolite or 30 g of GGBS in 200 mL of AMD, giving the removal efficiency levels of 76% and 100%, respectively. In both cases, effective Cr(VI) removal was attained within 15 min of contact between the adsorbents and AMD.
• In the converse, CEM I and FA strongly increased the Cr(VI) concentrations in AMD to levels far exceeding the permissible limits prescribed in national water standards, a concern attributed to leaching of the chromium present in the adsorbents into AMD.
• The Freundlich model correlated best with Cr(VI) adsorption data, for both zeolite and GGBS, suggesting a multilayer adsorption of Cr(VI) on the surface of the adsorbents.
• XRD and SEM analysis revealed high SiO₂ and Al₂O₃ in zeolite, which was responsible for its hydrophilic nature. Core-shell spheres and clustered crystals observed, confirmed the large surface area of zeolite that could lead to a fast and efficient adsorption of Cr(VI).
• Cr(VI) removal, when using GGBS, can be attributed to its adsorption onto the surface of GGBS. Hydration products formed provided physical sorption and co-precipitation sites for Cr(VI). Cr(VI) may have also co-precipitated with gypsum, which was identified as the dominant product formed from the reaction of GGBS with AMD.
• FTIR showed changes in O–H and C–O functional groups of zeolite and GGBS after CR(VI) adsorption, which is an indication of decationised adsorbents.
• Zeolite and GGBS were found to be effective adsorbents for Cr(VI) removal and may be used as partial cement replacement materials in systems that employ pervious concrete technology for AMD treatment.