Valuation of Unmodified Rice Husk Waste as an Eco-Friendly Sorbent to Remove Mercury: a Study Using Environmental Realistic Concentrations

Abstract

The present work explores the sorption capacity of an inexpensive and highly available agricultural waste, rice husk, to remove mercury using realistic concentrations of this metal. The efficiency of the process was evaluated for two initial Hg(II) concentrations, one representing the maximum value for Hg discharges from industrial sectors (0.05 mg L⁻¹), and the other ten times higher. A very small amount of rice husk (0.25 and 0.50 g L⁻¹) was able to reduce the Hg(II) levels in more than 80 % for an initial concentration of 0.05 mg L⁻¹ and in more than 90 % for 0.50 mg L⁻¹, corresponding to residual concentrations of Hg(II) of 0.048 and 0.009 mg L⁻¹, respectively. The biosorvent was reused in further cleaning treatments, maintaining the efficiency and high performance. The sorption kinetics of the Hg–rice husk system is well fitted by the Elovich model and the diffusion models suggested that, depending on the initial Hg(II) concentrations, the sorption process can be controlled by intraparticle diffusion or by both film and intraparticle diffusion. The equilibrium data are well described by the linear isotherm and the distribution coefficient found was 36.1 L g⁻¹.

APPENDIX A

1.1 Sorption Kinetic Models

Sorption kinetic models can be divided in two main categories: reaction-based models and diffusion-based models (Ho et al. 2000; Al-Degs et al. 2006). While the reaction-based models presuppose the presence of chemical groups in the sorbent which can undergo a chemical reaction and form a chemical bond (e.g., ion-exchange or chelation), the diffusion based-models presuppose an inert sorbent surface that may only provide physical sites for diffusional controlled bond formation (Ho et al. 2000).

1.1.1 Reaction-Based Models

The kinetic reaction models most commonly used in batch experiments are the pseudo-first-order model, the pseudo-second-order model, and the Elovich model (Ho et al. 2000).

The pseudo-first-order model was firstly applied by Lagergren and is mathematically expressed by:

$$ \frac{d{q}_t}{ dt}={k}_1\left({q}_e-{q}_t\right) $$

(A.1)

where k ₁ (h⁻¹) is the rate constant of pseudo-first-order and q _e (mg g⁻¹) is the amount of solute sorbed per gram of sorbent at equilibrium. After integration and application of the boundary condition q _t = 0 at t = 0, Eq. (A.1) can be expressed by the following non-linear equation

$$ {q}_t={q}_e\left(1-{e}^{-{k}_1t}\right) $$

(A.2)

However, it is well known that Lagergren model may not represent the sorption evolution along full time range (Ho et al. 2000).

The pseudo-second-order was firstly described by Ho, and in contrast with the previous model, usually correlates the behavior over the whole range of sorption (Ho et al. 2000). The kinetic rate equation is expressed as:

$$ \frac{d{q}_t}{ dt}={k}_2{\left({q}_e-{q}_t\right)}^2 $$

(A.3)

where k ₂ (g mg⁻¹ h⁻¹) is pseudo-second-order rate constant. By applying the boundary conditions t = 0 to t = t and q _t = 0 to q _t  = q _e , the integrated form of Eq. (A.3) is:

$$ {q}_t=\frac{q_e^2{k}_2t}{1+{q}_e{k}_2t} $$

(A.4)

The initial sorption rate (h) can be obtained as q/t approaches zero and is expressed by means of the following equation:

$$ h={k}_2{q}_e{}^2 $$

(A.5)

The Elovich model was established by Zeldowitsch and describes the kinetic law of chemisorption (Ho 2006) and it is described by the following equation:

$$ \frac{d{q}_t}{ dt}=\beta \exp \left(-\alpha {q}_t\right) $$

(A.6)

where q _t is the quantity of gas adsorbed during the time t, α (g mg⁻¹ h⁻¹) is the initial adsorption rate and β (g mg⁻¹) is the desorption constant. Integrating equation (A.6), using the initial condition q(t = 0) = 0, the Elovich’s equation is defined by:

$$ {q}_t=\frac{1}{\beta } \ln \left(1+\alpha \beta \right) $$

(A.7)

The Elovich’s equation has been widely used to describe the sorption of gases onto solid systems, but in recent years as has also been used to describe the sorption of pollutants from aqueous solutions (Ho 2006).

1.1.2 Diffusion-Based Models

The rate-controlling step involved in the sorption process can be determined by means of the diffusion-based models. The sorbate transport from the solution phase surface of the sorbent particles can be controlled either by one step, e.g., film or external diffusion, pore diffusion, surface diffusion, sorption on the pore surface, or by combination of more than one step (Ho et al. 2000). The models most commonly used to study the mechanism of the sorption process are Boyd’s film diffusion and Webber’s intraparticle diffusion. By using these models it is possible to distinguish whether if film (external) diffusion or intraparticle (pore) diffusion is the rate-controlling steps in the sorption process and in the case of a multi-linear behavior, if both mechanisms are involved (Malash and El-Khaiary 2010; El-Khaiary and Malash 2011).

The intraparticle diffusion (internal surface and pore diffusion) model can be expressed by the Weber and Morris equation (Malash and El-Khaiary 2010; El-Khaiary and Malash 2011; Ho et al. 2000):

$$ {q}_t={k}_{\mathrm{id}}\;{t}^{1/2} $$

(A.8)

where k _id is the internal diffusion rate constant (mg g⁻¹ h^−1/2). In a situation where the q _t vs t ^1/2 plot is linear, with a slope that equals k _id and with an intercept that passes through the origin, then the rate-limiting process is controlled only by intraparticle diffusion. If not, some other mechanism along with intraparticle diffusion must also be involved. Weber and Morris plots often present several linear segments and in this case several diffusion rate constant k _id can be calculated, each of them representative of the corresponding linear segment.

The film diffusion model of Boyd is based on the assumption that the main resistance to diffusion is the boundary layer surrounding the sorbent particles and is expressed as (Hameed and El-Khaiary 2008; Malash and El-Khaiary 2010; El-Khaiary and Malash 2011):

$$ F=1-\frac{6}{\pi^2}{\displaystyle \sum {{}_n^{\infty}}_{=1}\frac{1}{n^2} \exp}\left(-{n}^2 Bt\right) $$

(A.9)

where F is the fractional attainment of equilibrium at different times t and Bt is a function of F:

$$ F=\frac{q_t}{q_e} $$

(A.10)

The term Bt is calculated by the following Reichenberg equations:

$$ \mathrm{For}\;F\;\mathrm{values}>0.85\kern0.5em Bt=-0.4977- \ln \left(1-F\right) $$

(A.11)

$$ \mathrm{For}\;F\;\mathrm{values}>0.85\kern0.5em Bt={\left(\sqrt{\pi }-\sqrt{\pi -\frac{\pi^2F}{3}}\right)}^2 $$

(A.12)

Calculating Bt for each F value and plotting Bt vs t (Boyd’s plot), if the plot is non-linear or linear with a slope equal to B and with an intercept that do not passes through the origin, then it can be concluded that film diffusion or chemical reaction is the rate-controlling step. If the plot is linear and passes through the origin the intraparticle diffusion controls the rate of mass transference. The diffusion coefficient, (cm² h⁻¹) can be calculated from eq. (A.9):

$$ B=\frac{\pi^2{D}_i}{r^2} $$

(A.13)

where r is the radius of the sorbent particles assuming spherical shape. Boyd plots can present several linear segments and under these circumstances, each step must be analyzed separately to obtain the corresponding diffusion coefficient D _i .

1.2 Sorption Isotherms

The equilibrium data provides fundamental information on the affinity or capacity of a certain sorbent, in a particular sorption process. Sorption equilibrium is established when the concentration of sorbate in the bulk solution is in dynamic balance with that of the interface (Ho et al. 2000). The relationship between the sorbent and the sorbate is described by an isotherm, which is usually the relation between the amount of solute sorbed by unit of solid sorbent and the amount of solute that remains in solution, at a fixed temperature at equilibrium (Ho et al. 2000; Feng et al. 2004; Kumar and Bandyopadhyay 2006). The isotherms play an important role in the predictive modeling procedures for analysis and design of sorption systems (Kumar and Bandyopadhyay 2006).

In the simplest form it is assumed that the sorption isotherm for the sorbate is linear. However, the linearity is limited to a certain low concentration range and at higher sorbate levels, the sorption isotherms becomes non-linear and often have a convex (downward) shape. The Langmuir and Freundlich isotherms are the most frequent models used to represent the equilibrium data of a given sorption process (Feng et al. 2004; Kumar and Bandyopadhyay 2006). Both models have shown to be suitable for describing short-term and monocomponent sorption of metal ions by different materials (Ho et al. 2002).

1.2.1 Linear Isotherm

The linear isotherm indicates a partitioning process of the sorbate onto the sorbent:

$$ {q}_e={K}_l{C}_e $$

(A.14)

where K _l (L g⁻¹) is the distribution coefficient (Voudrias et al. 2002).

1.2.2 Langmuir Isotherm

Langmuir model assumes the presence of a finite number of binding sites, homogeneously distributed over the sorbent surface, presenting the same affinity for sorption of a single layer, and with no interaction between sorbed species. The non-linear form of Langmuir equation is represented by:

$$ {q}_e=\frac{q_m{K}_L}{1+{K}_L{C}_e} $$

(A.15)

where q _m (mg g⁻¹) represents the maximum sorption capacity of the sorbent, i.e., a complete monolayer coverage, and K _L (L mg⁻¹) is a constant related to the affinity of the binding sites (Feng et al. 2004; Kumar and Bandyopadhyay 2006; El-Said 2010).

1.2.3 Freundlich Isotherm

The Freundlich expression is an equation based on heterogeneous surfaces suggesting that binding sites are not equivalent and/or independent. The non-linear form of the Freundlich isotherm is represented by the following equation:

$$ {q}_e={K}_f{C}_e^{1/n} $$

(A.16)

where K _f (mg^1−1/n L^1/n g⁻¹) and n are the Freundlich constants characteristics of the system. K _f is related with the sorption capacity and n is associated to the sorption intensity (Ho et al. 2002).