3.1. Synthesis and characterization of iron oxide nanoparticles.
The morphology of magnetite nanoparticles was determined by High Resolution Transmission Electron Microscopy (HRTEM). The obtained images of synthesized nanoparticles at different magnification scales are shown in Figure 1(a-d). The fabricated nanoparticles were arranged in a beaded chain like structures with spherical morphology. The average size of the magnetite nanoparticles revealed by HRTEM was 42.90 nm. The polycrystalline nature of magnetite nanoparticles was confirmed by Selected Area Electron Diffraction Pattern (SAED) as shown in Figure 1(e). The Energy Dispersive X-Ray (EDX) peaks showed the constituents of the fabricated nanoparticles before and after Cr6+ adsorption (Figure 2a, b). The EDX composition includes the iron, oxygen, copper, and carbon peaks. The copper and carbon peaks were due to sample mounting on carbon coated copper grid. The X-Ray Diffraction (XRD) of the powdered magnetite nanoparticles give the structural information. The XRD pattern is shown in the Figure 3. The Fe3O4 nanoparticles possess orthorhombic structure with lattice constant (a) value of 2.799 A⸰. The XRD spectra showed multiple peaks at angle 2θ = 34.08, 38.47, 58.35, 83.99, 85.46 with hkl values of 023, 112, 006, 028, 173 (COD database, 96-900-2027) [26]. By using different peaks, crystallite size was determined by the following Scherer equation [27]:
$$D=\frac{k\lambda }{b{cos}\theta }$$
3
Where D is the “crystallite size (nm)”, λ is the “wavelength (Cu Kα = 0.154)”, b is the “full width at half maxima” (rad) for diffraction peak, k is the constant having value equal to 0.9 and θ is the “diffraction angle in degrees”.
3.2 Batch experimental studies
3.2.1. Variation in pH
The solution pH gives an idea about the effect of functional groups of adsorbents, the solubility of metal ions, adsorbate’s degree of ionization and the counter ion concentration during the reaction [28]. Fe3O4 contains Fe2+ therefore, hydrolysis products such as FeOH+, Fe(OH)20, and Fe(OH)3− varies with fluctuations in pH [29, 30, 31]. Different chromium species such as CrO42−, Cr2O72−, HCr2O7−, HCrO4− commonly exists in aqueous solution. The Cr6+ removal decreases with increase in pH (as shown in figure 4). The reason behind such behaviour may be the presence of more H+ ions or HCrO4− as major species at lower pH as OH− groups are neutralised by H+ ions and thus facilitates adsorption. At high pH, due to the presence of CrO42−, Cr2O72−, abundance of OH− species hinders the diffusion of dichromate ions and decreases the percent removal with increasing pH [32]. Hence, maximum adsorption was observed at pH 2 and minimum at pH 9.
3.2.2. Variation in initial ion concentration
Variation of initial ion concentration on Cr6+ removal was studied by adsorption on magnetite nanoparticles. The initial Cr6+ ion concentration was studied in the range of 10-60 mg L-1. The results obtained were plotted in figure 5 and the maximum (88.3%) removal percentage was observed at 10 mg L-1. The removal percent decreases with increase in concentration (also reported earlier by [33, 32]). However, according to the surface chemistry related theories, with increasing ionic strength an electric double layer decreases resulting in reduced adsorption of heavy metals [34].
3.2.3. Variation in adsorbent dose
The removal efficiency increases with increase in adsorbent dose with maximum at 60 mg and then decreases on increasing dose as shown in figure 6. However, no remarkable increase was observed in adsorption efficiency on further addition of dose upto 90 mg. The reason behind can be the aggregation of clustering nanoparticles on increasing NPs dose and hence, the availability of adsorption sites got reduced. As the same amount of Cr6+ ions got adsorbed on the larger surface area resulted in reduction of adsorption capacity [35].
3.2.4. Variation in contact time
To study the effect of contact time on adsorption, a range of 15-60 minutes was selected, and graph was plotted (Figure 7). On increasing agitation time, the Cr6+ removal also got increased. A maximum (75.3%) was observed at 50 minutes and removal percent decrease with increase in contact time. However, by increasing contact time, adsorption capacity got decreased.
3.2.5. Effect of temperature
Cr6+ adsorption was found to be decreased with increased temperature as represented by figure 8. The adsorption rate was increased suddenly from 15˚C to 20˚C and then declines with increasing temperature up to 40˚C. maximum Cr6+ removal was achieved at 20˚C and considered as optimum temperature. The decreased removal with increased temperature was possibly due to the tendency of Cr6+ ions to remain in the aqueous phase [36].
3.3. Isothermal Studies
Adsorption isotherms are “represented as the relation between number of adsorbate molecules per unit mass of adsorbent and the residual adsorbate concentration in bulk solution at a constant temperature”. In this work, Langmuir, Freundlich, Temkin and D-R (Table 1) were evaluated by adsorption of Cr6+ on magnetite nanoparticles [37, 38, 39, 40, 41].
Langmuir isotherm describes the equilibrium between adsorbent and adsorbate where, adsorption process is limited to monolayer. It also describes the homogenous surface of adsorbent assuming that there are no sideways interactions between the nearby adsorbed molecules, where every molecule occupies a single adsorption site [42]. The Langmuir equation used is given below:
Ce/qe = 1 / (KL × qm) + Ce/qm (4)
Where, qm is the Langmuir capacity and a plot between Ce/qe vs Ce was used to calculate the value of constants as shown in Figure 9a. The Langmuir constant (KL) and maximum adsorption capacity (qm) can be derived from the above equation 4 (Table 1). The value of qm derived from Langmuir modelling is 0.154 mg g−1. Separation factor (RL) is another parameter which evaluates the feasibility of Langmuir isotherm and can be derived from the following equation:
RL = 1 / (1 + K L × Co) (5)
The RL indicates the nature of isotherm, when RL ˃ 1, then the adsorption is considered as unfavourable whereas when RL = 1, there is a linear adsorption when 0 < RL <1, there is favourable adsorption while RL = 0, adsorption is irreversible
The RL value for magnetite nanoparticles was found to be 0.767, which is less than one and showed potential interaction among Fe3O4 nanoparticles and Cr6+ ions.
The adsorption data was further evaluated to fit Freundlich isotherm equation. This model facilitates the heterogenous surface adsorption forming multilayers on the surface [43]. The Freundlich model equation is given below:
lnqe = lnKf + 1/n ln Ce (6)
where, “Kf and n are the Freundlich constants”. The graph plot between lnqe vs lnCe (Figure 9b) gives the value of Kf (0.459) and n (18.28) (Table 1). The lower the value of 1/n (0.054) and higher value of n denotes efficient interactions among magnetite nanoparticles and chromium.
Temkin model was also evaluated for description of equilibrium mechanism by considering adsorbent-adsorbate interactions. The Temkin equation given below:
qe = (RT/BT) × ln (AT × Ce) (7)
gives the value of equilibrium binding constant (AT) and Temkin constant (BT). A curve was plotted qe vs lnCe (Figure 9c) to get the linear equation and find the values of constant. The value of BT is 24.33 kJ mol−1 which is the heat of adsorption, further the value of BT greater than 20 kJ mol−1 is the feature of physisorption.
D-R model is based upon multilayer adsorption model involving Van der Waal forces. It expresses adsorption mechanism with gaussian energy distribution onto heterogenous surfaces [44]. The equation to derive D-R parameters is given below:
lnqe = lnqm – KЄ2 (8)
Є = RT × ln [1+ (1/Ce)] (9)
$${E}_{m}=\frac{1}{\sqrt{2K}}$$
10
The curve was plotted against lnqe vs Є2 (Figure 10d). From D-R model, the value of qm was 4.89 mg g−1, this shows the maximum monolayer value of chromium adsorption per unit weight of adsorbent. The “value of mean free energy (Em)” is 1.46 kJ mol−1 at 293 K. In D-R isotherm, adsorption can be one of these three conditions i.e. if Em < 8.0 kJ mol−1, it is physisorption, Em = 8.0-16.0 kJ mol−1 then it is ion exchange while if Em ˃ 16-400 kJ mol−1, it is chemisorption. Hence, it is assumed that “process and concentrations of adsorbate and adsorbent are involved in rate determining step and the adsorption is physisorption” according to this model.
Based on correlation value of the experimental data the best fit model were compared and R2 order of the adsorption is Temkin˃ Langmuir˃ D-R˃ Freundlich.
3.4. Adsorption kinetics
The adsorption data was analysed for kinetics study of chromium removal by Pseudo first order, second order kinetics and intraparticle diffusion (Table 2). The adsorption kinetic model estimates the rate and mechanism of adsorption reactions.
First order kinetic equation describes “the solute adsorption on the adsorbent surface”. The linear form of first order kinetics is given as follows:
ln(qe-qt) =lnqe-k1t (11)
A curve was plotted, log(qe-qt) against t to find the values of k1 and qe (Figure 10a). A linearized form of pseudo second order equation is:
t/qt = 1/(k2qe2) + (1/qe) t (12)
The value of rate constant k2 (0.985) was derived by plotting a curve between t/qt vs t (Figure 10b) and are summarized in Table 3. The correlation coefficient (R2) of Lagergren’s first order and second order was found to be 0.68 and 0.98 respectively. Hence, pseudo second order kinetic model was the best fit to the experimental data.
Intraparticle diffusion model describes the adsorbate intraparticle uptake and pore diffusion during adsorption process. The linear equation for intraparticle model is:
qt = ki √t + xi (13)
A curve for intraparticle diffusion model (Figure 10c) gives the value of Xi (0.076) at 293 K for 10 mg L−1 Cr6+ solution. The boundary layer effect is directly proportional to the value of Xi i.e. higher the Xi value, higher will be the boundary layer effect.
The adsorption of Cr6+ on magnetite nanoparticles is a three-step process which involves the transportation of Cr6+ ions to the surface of Fe3O4 nanoparticles from bulk solution. Secondly, Cr6+ diffusion to the interior of the Fe3O4 pores. And ultimately adsorption of Cr6+ onto active sites of Fe3O4 nanoparticles surface [45].
3.4.4. Thermodynamics of adsorption
To assess the adsorption feasibility and spontaneity, thermodynamic parameters have wide role to play as they give basic information to design adsorption process. Feasibility and spontaneity are governed by parameters such as entropy (∆S˚), heat of enthalpy (∆H˚) and Gibbs free energy (∆G˚). various equations were used to calculate the thermodynamic parameters of this adsorption process which were given as follows [46, 47, 48]:
Distribution coefficient can be calculated by the equation,
lnKd = ∆S/R – ∆H/RT (14)
where Kd = qe/Ce (15)
Equilibrium chromium concentration is given by, lnCe = -lnK0 + ∆H/RT (16)
∆G = ∆H-T∆S (17)
By plotting a curve between lnKd vs 1/T and lnCe vs 1/T (Figure 11a and 11b respectively) values of thermodynamic parameters were calculated (Table 3) from slope and intercept of obtained equation.
The adsorption of chromium on magnetite nanoparticles was observed to favour lower temperature. With increase in temperature, the adsorption rate decreases due to weak adsorbate-adsorbent interactions. The negative value of ∆G˚ is evident to feasibility and spontaneity of the adsorption process at a given temperature. The physisorption mechanism of adsorption was confirmed by the negative value of ∆H˚ which lie in the range of -20 to 40 kJ mol−1 [49]. However, the positive ∆S˚ value suggests spontaneous adsorption.