Removal of boron from aqueous solution using cryptocrystalline magnesite

The present study aimed to evaluate the ef ﬁ ciency of using cryptocrystalline magnesite to remove boron ions from aqueous systems. Batch experimental protocols were used to evaluate the adsorption capacity of magnesite for boron. Parameters optimized included: time, dosage, chemical species concentration and pH. Optimum conditions were observed to be 30 min of agitation, 1 g dosage of magnesite per 100 mL of aqueous solution and 20 mg/L initial boron concentration. Removal of boron from aqueous solution was observed to be independent of initial pH of the aqueous solution. The adsorption of boron onto magnesite was observed to ﬁ t better to pseudo-second-order kinetics than pseudo- ﬁ rst-order kinetics hence proving chemisorption. The intra-particle diffusion model revealed that the adsorption of boron from aqueous system occurs through multiple reaction phenomena. Adsorption isotherms proved that the removal of boron by magnesite ﬁ tted well to both Langmuir and Freundlich adsorption isotherms hence proving that both mono- and multi-site adsorption processes are taking place. Under optimized conditions, magnesite was able to attenuate the boron concentration to < 0.01 mg/L which is below levels stipulated in World Health Organization guidelines. It was concluded that this comparative study will be helpful for further application of magnesite in remediation of boron-contaminated aqueous systems.


INTRODUCTION
Pollution of water by boron has caused serious environmental problems that require intensive environmental mitigation (Masindi et al. a). Boron can enter the environment through natural processes and anthropogenic activities. Naturally, boron is found in association with coal and gold seams, and other mineral seams.
After mining, it is exposed to weathering, hence boron leaches to surface and underground water resources.
Anthropogenically, it enters the environment through the use of fertilizers, insecticides, corrosion inhibitors, antifreeze, cooling systems, pharmaceuticals, dyes, bleaching agents, detergents and nuclear reactors (Türker et al. ). These industrial uses generate boron-rich wastewaters that are discharged to aquatic ecosystems. In aqueous environments, boron exists as boric acid MT which can be mined for the coming 20 years (Masindi et al. c). Thus, the use of cryptocrystalline magnesite is expected to be an economically viable way of removing boron from wastewater.

Sampling of the adsorbent
Raw magnesite rocks were collected before any processing at the mine from the Folovhodwe Magnesite Mine. Magnesite samples were milled to a fine powder using a Retsch RS 200 miller (Haan, The Netherlands) and passed through a 32-μm particle size sieve. After sieving, the samples were tightly kept in zip lock plastic bags until use.

Effect of shaking time
To evaluate the effect of equilibration time on the removal of boron, shaking time was varied from 1 to 360 min. Fixed parameters were: initial boron concentration 10 mg/L; magnesite in aqueous solution 1 g/100 mL; initial pH <6; 250 rpm shaking; and room temperature.

Effect of magnesite dosage
To evaluate the effect of magnesite dosage, the dosage was varied from 0.1 to 8 g. Fixed parameters were boron 10 mg/L; reaction time 30 min; initial pH <6; 250 rpm shaking; and room temperature.

Effect of chemical species concentration
To evaluate the effect of initial boron concentration, the initial concentrations were varied from 0.1 to 60 mg/L.
Fixed parameters were 1 g magnesite per 100 mL of solution; reaction time 30 min; initial pH <6; 250 rpm shaking; and room temperature.

Effect of pH
To evaluate the effect of initial pH, the initial pH was varied from 2 to 12. The initial pH was adjusted by using 0.1 M of NaOH and nitric acid. Fixed parameters were boron 10 mg/L; reaction time 30 min; 1 g magnesite per 100 mL of solution; 250 rpm shaking; and room temperature.
Optimized conditions obtained from the batch experiments were used in experiments to treat boron rich wastewater.

Calculation of boron removal and adsorption capacity
Computation of % removal and adsorption capacity was done using Equations (1) and (2).
Adsorption capacity (q e ) ¼ where: C i ¼ initial concentration, C e ¼ equilibrium ion concentration, V ¼ volume of solution and m ¼ mass of magnesite.

Adsorption kinetics
Adsorption kinetics was done using pseudo-first-order kinetic, pseudo-second-order kinetic and intra-particle diffusion models (Masindi et al. b).

Adsorption isotherms
Adsorption isotherms were determined using Langmuir and Freundlich adsorption models (Masindi et al. b).

Mineralogical composition by X-ray diffraction
The mineralogical composition of cryptocrystalline magnesite is presented in Figure 1.
The results revealed that cryptocrystalline magnesite consists of magnesite, dolomite, periclase, brucite, forsterite and quartz as the main constituents. The wide broadening of X-ray diffraction patterns and some sharp peaks indicates that the material is cryptocrystalline in nature. This corroborates SEM-EDS results and a study by Nasedkin et al. ().
Scanning electron microscopy-elemental dispersive spectrometry analysis The spot elemental composition and morphology of magnesite is shown in Figure 2.
The SEM image indicates that magnesite is composed mostly of aggregates with particle sizes generally below 1 μm. The elemental content of magnesite was determined using EDS at four spots. The analysis showed that the mineral is composed of Mg, C and O with weight percentage (wt %) of 43.9, 9.4 and 44.4, respectively. The impurities Si and Ca were also observed on cryptocrystalline magnesite surfaces. Again, the results correspond to arithmetic averages of four data points obtained from randomly selected locations on the surface of magnesite hence validating the EDS on each point.

Surface area by BET
The surface area of South African magnesite from Folovhodwe is shown in Table 1.
BET revealed that magnesite has a surface area of 14.6 m 2 /g, which is the sum of micropore area and external surface area. It has mean pore diameter of 223 Å. The large pore diameter of magnesite allows boron ions to diffuse into the pore channel of the mesoporous material due to smaller radius of boron species. Therefore, boron can be adsorbed onto the surfaces of magnesite through outer and inner sphere complexes.

Optimization experiments
Optimum boron removal conditions by cryptocrystalline magnesite were established by evaluating the effects of    from 0.5 g/100 mL to 1 g/100 mL, the adsorption of boron by magnesite was very rapid and above 1 g/100 mL no more notable change was observed denoting that the adsorption has reached equilibrium. The flattening of the adsorption curve above 1 g/100 mL shows that boron has been depleted from the aqueous system. As such, 1 g/100 mL was taken as the optimum dosage for the subsequent experiments.

Adsorption kinetics
The effect of contact time on removal of boron from aqueous solution was evaluated using different kinetic models to reveal the nature of the adsorption process and rate-limiting processes. A Lagergren pseudo-first-order kinetic model is a well-known model that is used to describe mechanisms of species adsorption by an adsorbent. It can be written as follows (Masindi et al. b): ln (q e À q t ) ¼ ln q e À k 1 t where k 1 (min À1 ) is the pseudo-first-order adsorption rate coefficient and q e and q t are the values of the amount adsorbed per unit mass at equilibrium and at time t, respectively. The experimental data were fitted by using the pseudofirst-order kinetic model by plotting ln(q e À q t ) vs t, and the results are shown in Table 2. The pseudo-first-order was applied and it was found to converge poorly with the experimental data; the correlation coefficient was <0.95.
Moreover, the calculated amounts of boron ions adsorbed by magnesite [q e , calc (mg g À1 )] were less than the experimental values [q e, exp (mg g À1 )] ( Table 2). The finding indicated that the Lagergren pseudo-first-order kinetic model is inappropriate to describe the adsorption of boron ions from aqueous system by magnesite.
The pseudo-second-order kinetic model is another kinetic model that is widely used to describe the adsorption process from an aqueous solution. The linearized form of the pseudo-second-order rate equation is given as follows: where k 2 [g (mg min À1 )] is the pseudo-second-order adsorption rate constant and q e and q t are the values of the amount adsorbed per unit mass at equilibrium and at time t, respect- ively. An application of the pseudo-second-order rate equation for adsorption of chemical species to magnesite showed a better fit with experimental data ( Table 2). The results obtained confirm that pseudo-second-order model is the more suitable kinetic model to describe adsorption of boron ions by magnesite from aqueous systems. Moreover, this also confirms that the mechanism of boron removal from aqueous solution is chemisorption. Different  Table 2. Note the theoretical adsorption capacity is close to the experimental adsorption capacity further confirming that this model describes the adsorption data. The overall kinetics of the adsorption from solutions may be governed by the diffusional processes as well as by the kinetics of the surface chemical reaction as proposed by Weber and Morris (Masindi et al. b). To determine whether film diffusion or intra-particle diffusion is the rate limiting step, the intra-particle diffusion model parameters were calculated ( Table 2). The model suggested that if the sorption mechanisms are via intraparticle diffusion then a plot of q t vs t 1=2 will be linear and intraparticle diffusion is the only rate limiting step when such plot passes through the origin. When the sorption process is controlled by more than one mechanism, then the plot of q t vs t 1=2 will be multi-linear. Intra-particle diffusion is computed using the following expression: where k id (mg g À1 min À1/2 ) is the intra-particle diffusion coefficient (slope of the plot of q t vs t 1=2 ) and C i is the intra-particle diffusion rate constant.
The C value (Table 2) indicate the thickness of the boundary layer which was observed to be very small for magnesite thus suggesting that surface diffusion plays a lesser role as the rate-limiting step in the overall sorption process.

Adsorption isotherms
The relationship between the amounts of ions adsorbed and the ion concentration remaining in solution is described by an isotherm. The two most common isotherm models for describing this type of system are the Langmuir and Freundlich adsorption isotherms (Masindi et al. b). These models describe adsorption processes on a homogenous (monolayer) or heterogeneous (multilayer) surface, respectively. The most important model of monolayer adsorption came from Langmuir (Masindi et al. b). This isotherm is given as: The constants Q 0 and b are characteristics of the Langmuir equation and can be determined from a linearized form of Equation (6). The Langmuir isotherm is valid for monolayer adsorption onto a surface with a finite number of identical sites and can be expressed in the following linear form: where, C e ¼ equilibrium concentration (mg/L), Q e ¼ amount adsorbed at equilibrium (mg/g), Q m ¼ Langmuir constant related to adsorption capacity (mg/g) and b ¼ Langmuir constant related to energy of adsorption (L/mg).
A plot of C e =Q e vs C e should be linear if the data conform to the Langmuir isotherm. The value of Q m is determined from the slope and the intercept of the plot. It is used to derive the maximum adsorption capacity and b is determined from the original equation and represents the degree of adsorption.
The Freundlich adsorption isotherm describes the heterogeneous surface energy by multilayer adsorption and is expressed as: The equation may be linearized by taking the logarithm of both sides of the equation and can be expressed in linear form as: where C e ¼ equilibrium concentration (mg/L), q e ¼ amount adsorbed at equilibrium (mg g À1 ), K f ¼ partition coefficient (mg/g) and n ¼ degree of adsorption.
A linear plot of log q e vs log C e indicates whether the data conform to the Freundlich isotherm.

CONCLUSION
In this study, the feasibility of using cryptocrystalline magne- The adsorption of boron onto magnesite was observed to fit better to pseudo-second-order kinetic than pseudo-firstorder-kinetic hence proving chemisorption. The intra-particle diffusion model revealed that the adsorption of boron from aqueous system is through multiple reaction phenomena. Adsorption isotherms proved that the removal of boron by magnesite fitted well to both Langmuir and Freundlich adsorption isotherm hence proving that both monoand multisite adsorption are taking place. This study proved that cryptocrystalline magnesite can be used as low-cost material for removal of boron from contaminated water bodies.