Application of Magnesium Oxide for Metal Removal in Mine Water Treatment

Abstract

In this study low-grade magnesium oxide (MgO) produced by calcinations of natural magnesite was used in mine water treatment using a laboratory-column device. The treatment of mine water from the abandoned Osor mine (NE Spain) with MgO showed the removal of metals from both mine water and tailing leachates. The PHREEQC numerical code and the Geochemist’s Workbench code (GWB) were used to evaluate the speciation of dissolved constituents and calculate the saturation state of the effluents. The analysis of the treated mine water showed the removal of As (from 1.59 to 0.31 μg/L), Cd (from 1.98 to <0.05 μg/L), Co (from 19.1 to <0.03 μg/L), F (from 2730 to 200 μg/L), Mn (from 841 to 0.6 μg/L), Ni (from 17.9 to <2 μg/L, U (from 9.16 to 0.08 μg/L), and Zn (from 2900 to 68.5 μg/L). Pb was also removed (from 98 to 35.2 μg/L) in the treatment of contaminated leachates from the mine waste. The mixing of MgO and water at room temperature may promote the formation of a stabilizing agent composed of hydroxides, carbonates, and magnesium-silicate-hydrates (MSH), which may remove Cd, Zn, and similar metals by sorption on MSH, substitution on the MSH lattice, and precipitation or co-precipitation with some of the hydrated phases.

1. Introduction

Although mine water treatment is often focused on acid drainage, metals such as arsenic, cobalt, molybdenum, nickel, antimony, and zinc are soluble at near-neutral pH and can contaminate soil, surface waters, and groundwater even under non-acidic conditions [1,2,3,4,5]. Neutral mine drainage occurs when sufficient neutralization potential of rocks and/or tailings reacts and consumes acid in the waters that have been in contact with the sulfides or when sulfide oxidation is weak due to the low sulfide content in ore and/or mine waste [6,7,8,9]. The main hydrochemical reactions for circum-neutral mine water (CNMW) occurrence are shown in Equations (1)–(6) and are associated with sulfide oxidation and carbonate dissolution [5,10,11]:

(1) Pyrite and sphalerite oxidation.

FeS₂ + 7/2 O₂ + H₂O → Fe²⁺ + 2SO₄²⁻ + 2 H⁺

(1)

ZnS + 2 O₂ → Zn²⁺ + SO₄²⁻

(2)

(2) Calcite dissolution.

CaCO₃ + H⁺ = Ca²⁺ + HCO₃⁻ (pH > 6.4)

(3)

CaCO₃ + 2H⁺ = Ca²⁺ + H₂CO₃ (pH < 6.4)

(4)

The presence of some cations, such as Na, Mg, and K, in high concentrations in mine water may be associated with the weathering of some silicates (albite and muscovite) and may also explain the detection of high pH. The progression of the feldspar alteration reactions changes the activities of the dissolved solutes: Ca²⁺, Na⁺, K⁺, H⁺, SiO₂(aq); therefore. albite and other aluminosilicates detected are not stable in these waters and should dissolve:

(3) Silicate (albite) dissolution.

2NaAlSi₃O₈ + 2H⁺ + 9H₂O → 2Na⁺ + Al₂Si₂O₅(OH)₄ + 4H₄SiO₄

(5)

NaAlSi₃O₈ + H⁺ + H₂O → Na⁺ + Al(OH)₃ + 3H₄SiO₄

(6)

Mine drainage from abandoned surface and underground mines may be treated in several ways, including active technologies (requiring dedicated professional teams and control systems) and passive remediation systems [12,13,14,15,16]. Most mine water treatment technologies are associated with the removal of metals from “acid mine water” (AMD) using active, passive or semi-passive systems [12,13,14,15,16,17,18]. However, there are relatively few studies of CNMW treatment [19,20].

The use of magnesium oxide as an alternative to lime and caustic soda for hydroxide precipitation has been proposed in some studies. However, its cost is a disadvantage since it is approximately three times the cost of hydrated lime [21]. Nonetheless, the use of caustic magnesia (by-product of MgO production through the calcination of MgCO₃) may be useful in the attenuation of metals such as Cu, Mn, and Zn [17]. Significant reduction in metal concentration were obtained after using MgO to treat effluents with high concentrations of Zn (50 mg/L), Cu (10 mg/L), Al (20 mg/L), Fe (360 mg/L), and SO₄²⁻ (960 mg/L) that were associated with brucite dissolution [17]. Metal removal may be linked to co-precipitation reactions of metals with amorphous Fe and Al oxy-hydroxides. Thus, iron in the ferric state as well as other metals should readily precipitate as oxyhydroxide compounds in mine water with high Fe concentrations [18]:

Fe³⁺ + 2OH⁻ → FeOOH + H⁺

(7)

Additionally, aluminum removal may be associated with the precipitation of Al(OH)₃. However, this species can redissolve when pH increases to 8.5 [18]:

Al³⁺ + 3OH⁻ → Al(OH)₃

(8)

Al(OH)₃ + OH⁻ → [Al(OH)₄]⁻

(9)

Likewise, MgO from caustic magnesia has been used to remove metals from contaminated groundwater and water contaminated with metals and Cr (VI) [22,23,24,25]. The removal of Ni (II) from wastewater was found by adsorption using globular magnesium oxide [26]. Moreover, AMD treatment with magnesite and MgO showed the decrease of As, Cd, Cr, Zn, Ni, and Mn in the treated mine water [27,28]. Tailings amended with magnesium oxide and other binders were more highly buffered and resistant against anthropogenic re-acidification than other tailings amended with conventional reactive materials [29,30]. In the treatment of contaminated soils by solidification/stabilization techniques, the use of MgO may reduce the leachability of some metals, specially Pb and Zn [31,32]. Additionally, MgO mixed with cementitious materials or Si-rich minerals has been used as a binder in the treatment of contaminated sediments [33,34].

The PHREEQC geochemical model uses solution equilibrium models to calculate aqueous speciation and saturation state of mineral phases, among other possibilities, and has been applied to evaluate the removal of heavy metals from lab-scale experiments with MgO [22,23]. Additionally, PHREEQC code is used in the study of AMD treatment by membrane technology and Zn and Ni removal from mine water using limestone treatment [35,36,37,38]. In the treatment of AMD and CNMW, PHREEQC is used in order to predict the mineral phases involved in sulfate removal [39]. Likewise, PHREEQC showed the removal of metals from mine water treated with magnesite which precipitated and formed oxy-hydroxysulfates, hydroxides, gypsum, and dolomite [40]. In addition, PHREEQC is used to understand the geochemical processes in mine water treatment plants and dynamic modeling using the chemical reaction module [41,42,43] and to study the use of thermal activated non-cryptocrystalline magnesite on the treatment of AMD [44].

The overall objectives of this study were to investigate the potential use of MgO in passive mine water treatment, especially in the CNMW treatment, where the laboratory-column studies may be a novelty contribution in the mine water treatment. Additionally, the study evaluated the possible mechanisms of metal removal using the PHREEQC geochemical model. The PHREEQC code calculated the speciation and saturation index of the possible mineral phases which may control the mass transfer processes. Geochemical modeling may be a cost effective way for assessing the appropriate treatment processes for particular mine water.

2. Materials and Methods

2.1. Characterization of MgO

The low-grade MgO used in this study was produced by calcination in a rotary kiln of natural magnesite (Figure 1). The temperature of the calcination ranged from 500 °C to 1300 °C. The rotary kiln was 81 m in length and the magnesite remained inside of it for approximately 6 h. The MgO obtained under these conditions is termed “hard-burned” and shows a narrow range of reactivity. This grade is typically used when slow degradation or chemical reactivity is required, as is the case with animal feeds and fertilizers [22].

The MgO used in the experiments is a solid waste by-product projected to be reused as a metal adsorbent in treatment of mine water drainage or tailings remediation leachates. Additionally, the following physical characteristics of MgO were evaluated: pore size distribution, porosity, field capacity, and bulk density. Particle size distribution was determined by sieve analysis (RETSCH AS 200) using the following aperture ranges: 4, 2, 1, 0.5, 0.25, 0.125, 0.063, and 0.045 mm.

2.2. Geochemistry of MgO and Mine Wastes

The geochemical composition was analyzed using instrumental neutron activation analysis (INAA) and ICP-AES (inductively coupled plasma atomic emission spectrometry) in Actlabs (Ontario, Canada). The following elements were quantified by INAA: Au, Ag, As, Ba, Br, Ca, Co, Cr, Cs, Fe, Hf, Hg, Ir, Mo, Na, Ni, Rb, Sb, Sc, Se, Sn, Sr, Ta, Th, U, W, Zn, La, Ce, Nd, Sm, Eu, Tb, Yb, and Lu. In addition, the concentrations of the following elements were determined by acid digestion (employing HF, HClO₄, HNO₃ and HCl) and subsequent analysis by ICP-AES: Ag, Cd, Cu, Mn, Mo, Ni, Pb, Zn, Al, Be, Bi, Ca, K, Mg, P, Sr, Ti, V, Y, and S (

Table 1. Composition of Osor tailings (TAL), ore sample, and MgO. CAL (*): The Catalonian soil intervention values (industrial use). ND: not determined. Data for Osor tailings and ore from [5].

Element	Al	Ag	As	Au	Ba	Be	Bi	Ca	Cd	Co	Cr	Cu	Fe	Hg	K	Mg	Mn	Mo	Na
Unit	%	ppm	ppm	ppb	ppm	ppm	ppm	%	ppm	ppm	ppm	ppm	%	ppm	%	%	ppm	ppm	%
Detection limit	0.01	0.3	0.01	2	50	1	2	0.01	0.3	0.1	0.3	1	0.01	0.05	0.01	0.01	1	0.05	0.01
Osor tailings (TAL)	4.46	0.6	12.5	<2	5110	2	0.3	7.33	7.6	14	39	47	1.78	<1	1.86	0.49	684	1	1.0
Osor ore	0.24	66.2	164	32	2190	1	<2	21.4	68.5	15	119	88	0.96	5	0.43	0.08	109	1	0.14
MgO	0.42	<0.3	17.5	<2	<50	<1	<2	7.07	<0.3	47	95	28	2.16	<1	0.30	32.6	965	<1	0.08
CAL *	---	---	30	---	1000	90	---	---	55	90	1000	1000	---	30	---	---	---	70	-----
Element	Ni	P	Pb	Rb	S	Sb	Se	Sr	Ta	Ti	Th	U	V	W	Zn	La	Ce
Unit	ppm	%	ppm	ppm	%	ppm	ppm	ppm	ppm	%	ppm	ppm	ppm	ppm	ppm	ppm	ppm
Detection limit	1	0.001	1	1	0.01	0.005	0.1	1	0.5	0.01	0.1	0.01	2	0.05	1	0.01	0.1
Osor tailings (TAL)	18	0.041	940	103	0.23	1	<0.1	105	<0.5	0.24	5.2	2.4	45	<1	2370	27.8	55
Osor ore	31	0.008	>5000	<15	5.78	208	<3	74	<0.5	0.03	1.5	<0.5	15	<1	34000	9	12
MgO	27	0.030	<3	<15	0.04	1.4	<3	83	<0.5	0.02	0.8	1.5	47	<1	6	<0.5	13
CAL *	1000	--	550	--	--	30	70	--	45	--	--	--	1000	--	1000	--	--

). The grain size distribution of mine wastes was determined by sieving (RETSCH AS 200), the porosity by water displacement in a test tube due to their sandy character [23], and the field capacity by numerical methods.

2.3. Mine Water, Mine Wastes, and MgO

Mine water was collected from the Coral adit, which drained the abandoned mine of Osor (Girona, NE Spain). The abandoned Osor mining area lies approximately 35 km SE of Girona, on the La Selva basin and Montseny–Guilleries massif, which belongs to the Catalonian Coastal Ranges (CCR) in the NE of the Iberian Peninsula. The main mine drainage was Coral adit, which drains the Osor vein system and had an estimated discharge in the Osor river varying between 800 and 1500 m³/day of near-neutral contaminated mine water. Annually, mine water discharge was higher in the winter or spring, and lower, generally, in the late summer and early fall [5,45].

Mine waste at the study area consisted of flotation tailings and waste rock generated from the exploitation of the Osor vein deposit. Mineral deposits are located 4 km SE of the Anglés town, and they include several fluorite-barite-sphalerite-galena veins, exploited until 300 m of depth [5]. Gangue minerals included quartz, barite, calcite, pyrite, and silicates (mainly muscovite, albite, and biotite).

2.4. Column Leaching Experiments

The column leaching method was a modified PrEN 14,405 procedure, the standard European percolation test [46]. A column with the following dimensions was used: 750 mm long, 150 mm diameter, and 5 mm thick (Figure 1C). A 222 mm long plastic funnel with an internal diameter of 186 mm was attached to the bottom of the column. Inside the funnel, a fiberglass plate with holes in it acted as a support column. Methacrylate was coated onto a mesh that acted as a filter and retained the porous medium. The entire device was mounted on a metal structure to hold the column in a vertical position at a regular height above the surface. The contaminated mine water (GC sample) and the contaminated lixiviate (LT sample) used in the experiments were added in the column using a rain simulator connected to a metering pump, which provided a maximum flow of 10 L/h and could be adjusted from 1% to 100% of low-mineralized water (Figure 1C).

Contaminated mine water (GC sample) and tailings leachate (LT sample) percolated each column obtaining 6 samples of treated mine water: samples G-1 to G-6 (

Table 2. Composition of circum-neutral mine water and treated mine water samples. GC: mine water sample used in the column experiment. G-1 to G-6: treated mine water.

Analyte Symbol	Time	pH	Eh	EC	Mn	Fe	Co	Ni	Zn	Pb	As	Se	U	Mo	Cd	Sb	Cu
Unit	min	pH unit	mV	μS/cm	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L
Detection Limit					0.1	10	0.005	0.3	0.5	0.01	0.03	0.2	0.001	0.1	0.01	0.01	0.2
GC	---	7.46	98.6	678	841	120	19.1	17.9	2900	2.11	1.59	4.4	9.16	1.0	1.98	0.17	7.3
G-1	0	10.7	−750	7200	50.4	50	1.66	2.3	329	118	1.54	2.6	0.957	6.4	1.04	5.87	29.8
G-2	30	12.4	−725	4530	0.4	370	<0.005	0.5	8.2	1970	0.36	2.2	0.017	9.9	0.04	5.64	2.8
G-3	60	12.6	−776	6080	3.2	580	<0.03	<2	73.5	2470	0.34	1.4	0.104	7.1	0.06	13	4.9
G-4	90	12.6	−820	6470	0.9	550	<0.03	<2	76.5	1870	0.48	2.2	0.093	6.3	<0.05	12.9	6.1
G-5	120	12.7	−810	6730	2	630	<0.03	<2	69.5	1350	0.31	2	0.098	5.7	0.05	11.6	5.5
G-6	150	12.7	−750	6780	0.6	710	<0.03	<2	68.5	1000	0.4	1.8	0.081	5.4	<0.05	11.3	6

and

Table 3. Main anion contents of mine water (GC) and first and last treated mine water samples.

Analyte Symbol	Time	F	Cl	NO3 (as N)	SO4	HCO3−
Unit Symbol	min	mg/L	mg/L	mg/L	mg/L	mg/L
Detection Limit	---	0.01	0.03	0.01	0.03	1
GC	---	2.73	22.4	0.4	210.7	408
G-1	0	0.08	50.4	0.3	254.8	1222
G-6	150	0.2	30.1	0.1	123.3	3013

) and six treated leachates (samples T-1 to T-6 in

Table 4. Composition of untreated leached tailings (LT sample) and treated leachates (T-1 to T-6 samples).

Analyte Symbol	Time	pH	Eh	EC	Mn	Fe	Co	Ni	Zn	Pb	As	Se	U	Mo	Cd	Sb	Cu
Unit	min	pH unit	mV	uS/cm	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L	µg/L
Detection Limit					0.1	10	0.005	0.3	0.5	0.01	0.03	0.2	0.001	0.1	0.01	0.01	0.2
LT	0	7.77	122	2420	80.9	830	1.85	3.8	2380	98.1	2.25	1.5	1.17	6.4	8.73	1.59	10.4
T-1	0	12.3	−700	4570	117	130	2.55	1.4	60.9	0.89	1.6	2.1	0.225	5.4	0.19	6.42	5
T-2	30	12.6	−678	5870	2.7	170	<0.03	<2	105	85.4	0.6	2.2	0.118	3.8	<0.05	12.1	3.4
T-3	60	12.6	−800	6400	<0.5	120	<0.03	<2	63.5	52.7	0.31	2.3	0.026	3.5	<0.05	12.8	1.9
T-4	90	12.6	−600	6700	<0.5	170	<0.03	<2	64.5	44.2	0.33	2	0.088	3.1	<0.05	12.9	2.8
T-5	120	12.7	−710	6860	<0.5	230	<0.03	<2	59.5	37.8	0.3	2.2	0.07	2.9	<0.05	13	2.7
T-6	150	12.7	−560	7030	<0.5	480	<0.03	<2	54	35.2	0.26	1.9	0.06	2.5	<0.05	14.4	2.6
MCL *	----	----	----	----	----	----	----	120	1200	150	60	40	----	200	20	100	600

(*) MCL: maximum contaminated level of solid waste leachates (percolation test) in order to deposit in landfills (inert solid waste).

and

Table 5. Main anion contents of untreated leached tailings (LT sample) and treated leachates (T-1 and T-6 samples).

Analyte Symbol	Time	F	Cl	NO2 (as N)	Br	NO3 (as N)	PO4 (as P)	SO4	HCO3−
Unit Symbol	min	mg/L	mg/L	mg/L	mg/L	mg/L	mg/L	mg/L	mg/L
Detection Limit	---	0.01	0.03	0.01	0.03	0.01	0.02	0.03	1
LT	0	2.3	36.4	0.33	<0.6	34.3	<0.4	1790	72
T-1	0	0.25	78.8	---	---	0.5	---	119.7	344
T-6	150	0.38	8.4	---	---	0.1	---	87.4	2950
MCL *	---	40	8500	---	---	---	---	7000	---

(*) MCL: maximum contaminated level of solid waste leachates (percolation test) in order to deposit in landfills (inert solid waste).

). The samples were collected at the bottom of the column during 150 min at 0, 30, 60, 90, 120, and 150 min of the first effluent. The pH, redox potential (Eh; mV), water temperature, and electrical conductivity (EC; μS/cm) measurements were calibrated using standard solutions and measured in situ with portable devices (HACH model sensION TM378). The samples were filtered using a cellulose nitrate membrane with a pore size of 0.45 um. Samples for cation analysis were later acidified to pH < 2.0 by adding ultra-pure HNO₃. The samples were collected in 110 mL high-density polypropylene bottles, sealed with a double cap, and stored in a refrigerator until analysis.

Metal concentrations were measured by inductively coupled plasma mass spectrometry (ICP-MS) at Actlabs (Ancaster, ON, Canada). Chloride, nitrate, and sulfate concentrations were analyzed in unfiltered samples by ion chromatography (IC). Water alkalinity was analyzed by titration. Standard reference material NIST 1640 (ICP-MS) was used to confirm accuracy. In order to preserve the carbonate equilibrium and CO₂ (g) levels in water samples, the samples collected for main anion determination were quickly closed and rapidly taken to the laboratory after sampling.

2.5. Geochemical Modeling

Hydrogeochemical analyses of leachates were performed using the PHREEQC [47] numerical code (version 3.0.6-7757) to evaluate the speciation of dissolved constituents and calculate the saturation state of the effluents. The MINTEQ thermodynamic database was used for chemical equilibrium calculations.

3. Results and Discussion

3.1. Mine Wastes and MgO

Chemical analysis of mine tailings revealed that Pb and Zn concentrations were above Catalonia’s generic reference levels for human health protection applicable to soils under industrial use (550 and 1000 mg/kg, respectively) (Table 1) [48]. Results from this analysis suggest that waste deposited in an open area is equivalent to contaminated soil, is inadequate for industrial use, and is susceptible to leaching, thus allowing the release of pollutants to surface water and groundwater. The bulk density and porosity of tailings were 1.291 and 0.22 g/cm³, respectively, while the equivalent diameter obtained from grain size analysis was 1.5 mm.

The bulk composition of MgO is shown in Table 1. The calcium content was due to the presence of small amounts of dolomite, MgCa(CO₃)₂, in the natural magnesite. The presence of iron, aluminum, and silica did not interfere with the treatment of mine water and mine tailings, which remained inert in the reactive material. The elevated content of Cr and Ni may be associated to the grinding process of the natural magnesite. Porosity and equivalent diameter of MgO were 0.57 and 1.55 mm, respectively. The LG-MgO large particle size was necessary to establish a porosity that ensured the permeability of the medium.

3.2. Mine Water Treatment

The mine water sample used in the treatment (GC sample in Table 2 and Table 3) showed high concentrations of Ca (167 mg/L), HCO₃ (408 mg/L), and SO₄ (210.7 mg/L) and a pH of 7.46, which might indicate calcite dissolution in the Coral adit. Additionally, the elevated concentrations of Mn (0.84 mg/L) and Zn (2.9 mg/L) suggest the dissolution of sphalerite and Mn oxides. Moreover, the mine water showed Na, Mg, and K concentrations above surface water and springs concentrations. The weathering of some detected silicates (albite and muscovite) might also explain the high pH detected [5].

A sharp increase in pH (from 7.46 to 12.75) can be observed when comparing samples collected from the column test (Table 2 and Table 3) with the sample of mine drainage water. This could indicate that pH was controlled by the solubility of portlandite (Ca(OH)₂), formed by the rapid hydration reaction of the calcium dioxide present in the mixture of MgO [22].

In addition to pH, conductivity and redox potential markedly varied in the first post-treatment sample collected (G1) relative to the Coral gallery sample (GC). The initial EC in the contaminated water (less than 1000 μS/cm, GC sample) increased to over 7000 μS/cm, thus indicating a sharp rise in water mineralization, and decreased to 4530 μS/cm in sample G2, suggesting the progressive dissolution of some elements from the MgO substrate. The redox potential in the untreated Coral gallery sample was positive while the test samples presented values quite negative, which means that reduction reactions were favored over oxidation.

The main metal and metalloid concentrations in MgO-treated elutriates are listed in Table 2 and Table 3. MgO was very effective in the removal of Cd (from 1.98 to <0.05 μg/L), Co (from 14.1 to <0.03 μg/L), F (from 2.73 to 0.2 mg/L), Mn (from 841 to 0.6 μg/L), Ni (from 17.9 to 0.5 μg/L), Si (from 9.3 to 0.3 mg/L), Ti (6.2 to <0.5 μg/L), U (from 9.16 to 0.08 μg/L), and Zn (from 2900 to 68.5 μg/L). Removal rates varied between 80% and 99%.

Other elements such as (from 1.59 to 0.31 μg/L) and Se (from 4.4 to 1.8 μg/L) presented a lower reduction in the last sample of the test compared to Coral gallery water. This indicates a reduced effectiveness of the treatment for these metals and metalloids when compared with others previously mentioned. Column tests using heavy metal contaminated waste showed that the degree of metal leaching was highly dependent on pH [49]. Increasing pH reduced the solubility of metals and metalloids. The presence of hydroxyl (OH⁻) ions in aqueous solution and the negative Eh favored reduction reactions, which generated sparingly soluble metal compounds that precipitated in the solid matrix in the form of hydroxides, oxy-hydroxides, and carbonates [22].

These processes may explain removal rates observed for some of the metals analyzed after treatment with the MgO layer. Figure 2 shows the evolution of the concentration of major pollutants—Mn, Zn, Co, Ni, and U—during the column experiment. There was a sharp reduction in these five metals concentrations in the first collected sample (G1) compared with the untreated sample (GC). Concentrations continued to decrease in the following 30 min (G2) increased at the 60 min mark (G3) and remained stable until the last sample was collected. The lower concentration changes of the sample relative to the G3-G2 evolution may be related to the decrease in EC (from 7200 to 4530 μS/cm) and the increase in pH (10.76 to 12.46) during the test as precipitation and other reactions occurred in the column experiment.

The observed increase of Pb in the solution after the interaction with the MgO substrate may be explained by the change of solubility of lead hydroxide (Pb(OH)₂) as a function of pH. At a pH of about 9.5, the solubility of this compound is low. However, it increases significantly at pH values above 11 [49]. Studies using cement to stabilize Pb-rich wastes also detected the increased release of Pb due to favorable conditions for generating the high alkalinity of the cement [50].

3.3. Treatment of Mine Tailings

Table 4 and Table 5 and Figure 3 show the physicochemical parameters obtained in the leachates of the column experiment with flotation tailings and MgO (samples T-1 to T-6). The samples showed that pH varied between 12.3 and 12.7, Eh was always negative and ranged from −560 to −800 mV, and EC progressively increased throughout the experiment. The continuous decrease in concentrations during steady state flow indicates that metal contents in the leachate were controlled more by availability than by solubility of pure mineral phases [51]. The evolution of the main metals during the experiment showed a continuous decrease in concentration (Table 4). Results from leaching experiments also showed that Zn was the metal most released, with concentrations of 2.3 mg/L in tailings matching with the most abundant metal in waters. The concentration of Zn in pre-MgO treatment tailings (Figure 3) was above the maximum contaminant level for inert solid waste (1200 μg/L). Additionally, during MgO treatment there was a continuous decrease in Cd concentration (Table 5), with values below the maximum contaminant level (20 μg/L).

There was a continuous drop in Mn concentrations from 117.9 to less than 0.5 μg/L, while Fe increased from 130 to 480 μg/L, possibly because of the pH-Eh conditions of the experiment (Table 4, Figure 3). The leaching of Pb and Zn also showed an initial increase in concentrations in the sample collected at 30 min from the beginning of leaching (Figure 3). This phenomenon may be associated to an equilibrium process in column effluent during percolation and/or flow irregularities, which may produce increasing concentrations [51,52]. When comparing post-treatment concentration with metal and metalloid concentrations in untreated tailings (sample LT in Table 4 and Table 5), we can see that the layer of MgO was very effective in removing As (from 2.25 to 0.26 μg/L), Cd (from 8.73 to <0.05 μg/L), Co (from 1.85 to <0.03 μg/L), F (from 2.3 to 0.38 mg/L), Mn (from 80.9 to <0.5 μg/L), Pb (from 98 to 35.2 μg/L), U (from 1.17 to 0.06 μg/L), and Zn (from 2380 to 54 μg/L). The reduction in metals and metalloids due to treatment was above 78%. Fe (from 830–480 μg/L). Mo (from 6.4 to 2.5 μg/L) and Cu (from 10.4 to 2.6 μg/L) showed a range of reduction between 50 and 78%, which indicates the efficacy of such a reactive layer.

3.4. Geochemical Modeling

Hydrogeochemical analyses of leachates were conducted to evaluate the speciation of dissolved constituents and to calculate the saturation state of the effluents (

Table 6. Distribution of species for leachates. Data calculated using PHREEQC and database Minteq. Values in molality.

Species	G-1	G-6	T-1	T-6
CdCO3	8.16 × 10−9	4.47 × 10−11	2.01 × 10−11	6.41 × 10−12
Cd(OH)5	1.66 × 10−10	2.29 × 10−9	1.26 × 10−9	2.42 × 10−10
CdOH+	6.92 × 10−11	1.10 × 10−11	1.28 × 10−11	1.18 × 10−12
Cd2+	2.48 × 10−11	4.59 × 10−14	1.15 × 10−13	6.04 × 10−15
MgCO3	3.53 × 10−3	1.47 × 10−6	4.65 × 10−4	3.83 × 10−7
Mg2+	2.42 × 10−3	3.29 × 10−7	6.53 × 10−4	8.16 × 10−8
MgSO4	3.13 × 10−4	9.68 × 10−9	3.44 × 10−5	1.75 × 10−9
MgHCO3	1.88 × 10−5	1.61 × 10−6 *	1.51 × 10−3 *	3.58 × 10−7 *
Mn2+	4.83 × 10−7	7.40 × 10−12	1.50 × 10−8	8.53 × 10−12
MnOH+	4.3 × 10−7	5.37 × 10−10	5.36 × 10−7	5.55 × 10−10
MnSO4	6.26 × 10−8	2.05 × 10−13	7.87 × 10−10	1.74 × 10−13
MnHCO3+	5.79 × 10−9	1.04 × 10−8 **	1.58 × 10−6 **	8.57 × 10−9 **
Zn(CO3)2−2	2.81 × 10−6	5.94 × 10−11	1.68 × 10−11	7.68 × 10−11
Zn(OH)2	1.83 × 10−6	1.52 × 10−8	6.72 × 10−8	1.46 × 10−8
Zn(OH)3−	3.52 × 10−7	3.57 × 10−7	5.51 × 10−7	3.04 × 10−7
ZnCO3	2.79 × 10−8	6.79 × 10−7 ***	5.13 × 10−7 ***	5.10 × 10−7 ***

*: MgOH⁺; **: Mn(OH)₃⁻; ***: Zn(OH)₄⁻².

and

Table 7. Calculated saturation index for leachates. Saturation indices calculated using PHREEQC and database MINTEQ.

Species	GC	G-1	G-6	T-1	T-6
Artinite	−2.68	3.43	0.24	5.33	−1.05
Brucite	−2.70	1.80	1.85	4.45	1.15
Cd(OH)2	−5.63	−3.16	−2.01	−2.27	−2.99
Chrysotile	2.44	13.69	5.80	16.20	3.89
Nesquehonite	−1.60	0.06	−3.18	−0.69	−3.77
Otavite	0.53	0.30	−1.98	−2.35	−2.84
Hydrozincite	39.79	35.12	17.90	20.87	18.04
Smithsonite	−0.55	−2.85	−8.35	−7.82	−8.25
Talc	3.79	12.95	−3.02	10.01	−4.72
ZnCO3:H2O	−0.21	−2.64	−8.09	−7.56	−7.99
Portlandite	−8.06	−4.89	−0.09	−0.86	−0.25

). In addition, some pH-Eh diagrams were drawn using GWB [53]. The study of aqueous Cd speciation showed that the predominant species in treated leachates were Cd(OH)₂ and CdCO₃ (Table 6), which may indicate the mobilization of Cd if these species are stable in the pH-Eh conditions of the experiments. Mg speciation showed that MgCO₃ and Mg²⁺ are the predominant species, whereas Mn²⁺ and Mn(OH)³⁻ were the most stable in Mn speciation. The geochemical speciation of Zn leachates indicates that Zn may be retained in the treated samples, although this metal may be mobilized as Zn (CO₃)₂²⁻ and Zn(OH)₄²⁻ in the MgO-treated samples.

The PHREEQC calculations indicated that all leachates were saturated with respect to artinite (except leachate T-6), brucite, chrysotile, and hydrozincite (Table 7). Additionally, some leachates were saturated with respect to nesquehonite and talc (Table 7). According to the pH-Eh diagram (Figure 4), Mg first should precipitate as carbonate and/or brucite. Thus, metal removal in the treatment experiments could be associated first with the precipitation of brucite and metal hydroxides:

MgO + H₂O = 2OH⁻ + Mg²⁺

(10)

2Mg²⁺ + 2OH⁻ = 2Mg(OH)₂

(11)

M²⁺ + Mg(OH)₂ = M(OH)²⁺ Mg²⁺

(12)

However, the pH-Eh conditions suggested that brucite may be an ephemeral phase and drive quickly to carbonation reactions producing the formation of magnesite and/or artinite and lansfordite/nesquehonite [33,54]. The carbonate species are formed as CO₂ dissolves in water [33,54]. Thus the main carbonation reactions may be:

Mg(OH)₂ + H₂O + H₂CO₃ → MgCO₃·3H₂O (nesquehonite)

(13)

Finally, with the pH increase (Figure 4), the hydration products: Talc (Mg₃Si₄O₁₀(OH)₂) and chrysotile (MgSi₂O₅(OH)₄) may be present [33] and might be partially responsible for the metal immobilization (Figure 4). Additionally, in similar MgO-based studies the presence of talc was detected by XRD analysis [33].

According to the pH-Eh conditions, Zn should be a low mobile element, especially under high pH (pH > 12) and slightly reducing conditions such as those of the MgO tests. Thus, Zn should be less mobile in alkaline and low redox environments due to the possible precipitation of zinc oxides and hydroxides. Additionally, geochemical modeling of leachates indicated that the most possible solid phase sink for Zn could be hydrozincite (Zn₅(OH)₆(CO₃)₂) since all the leachates showed super-saturation with respect to this mineral.

Divalent metals such as Mn, Cu, and Zn may be removed by adsorption and/or catalytic oxidation [55,56,57,58,59]. These metal removal mechanisms together with hydration products may be the main chemical processes related with the application of magnesium oxide in mine water treatment. Especially in the case of As, due to MgO, the adsorbent may exhibit a higher environmental stability [59].

4. Conclusions

MgO was tested as a reactive material in the removal of metals from mine water and mine waste leachate. Treatment performance was evaluated using column experiments and the PHREEQC numerical code. The use of MgO layer in the treatment of mine water was very effective in the removal of As (from 1.59 to 0.31 μg/L), Cd (from 1.98 to <0.05 μg/L), Co (from 14.1 to <0.03 μg/L), F (from 2.73 to 0.2 mg/L), Mn (from 841 to 0.6 μg/L), Ni (from 17.9 to 0.5 μg/L), U (from 9.16 to 0.08 μg/L), and Zn (from 2900 to 68.5 μg/L).

The treatment of leachates from Osor mining waste showed that the use of MgO was very effective in removing As (from 2.25 to 0.26 μg/L), Cd (from 8.73 to <0.05 μg/L), Co (from 1.85 to <0.03 μg/L), F (from 2.3 to 0.38 mg/L), Mn (from 80.9 to <0.5 μg/L), Pb (from 98 to 35.2 μg/L), U (from 1.17 to 0.06 μg/L), and Zn (from 2380 to 54 μg/L).

Geochemical modeling showed that the mixing of MgO and contaminated water at environmental temperature may promote the formation of a stabilizing agent composed of carbonates and possible brucite. However, the pH-Eh conditions suggested that brucite may be an ephemeral phase and drive to carbonation reactions producing the formation of magnesite and/or artinite and lansfordite/nesquehonite. The carbonate species are formed as CO₂ dissolves in water.

Later, with the pH increase magnesium-silicate-hydrates (MSH) may remove Cd, Zn, and similar metals by sorption on MSH, substitution on the MSH lattice, and precipitation or co-precipitation with some of the hydrated phases. The pH-Eh experimental conditions suggested the possible formation of talc (Mg₃Si₄O₁₀(OH)₂).

Geochemical modeling of leachates indicated that the most possible solid phase sink for Zn would be hydrozincite (Zn₅(OH)₆(CO₃)₂) given that all the leachates showed super-saturation with respect to this mineral.

Divalent metals such as Mn, Cu, and Zn may be removed by adsorption and/or catalytic oxidation. These metal removal mechanisms, together with hydration products, may be the main chemical processes related to the application of magnesium oxide in mine water treatment.