Flower Buds Like MgO Nanoparticles: From Characterisation to Indigo Carmine Elimination

2.1 MgO Nanoparticles Synthesis

The MgO samples were fabricated by pyrolysis of magnesium carbonate (MgCO₃₎ (Sigma Aldrich) without the addition of any other chemical. In a typical preparation, 4.5 g of MgCO₃ was placed in a ceramic boat and calcinated in a tubular oven (Nabertherm, Germany) under ambient conditions at 700 °C and 5 °C⋅min⁻¹ for different time intervals namely 1, 3, 6, 9 and 12 h. The pyrolysis temperature was chosen to be considerably lower than the melting point of MgCO₃ (851 °C) to ensure the formation of the targeted product.

2.2 Characterisation

The MgO samples were studied by means of the powder XRD at room temperature, using D8 Advance Diffractometer (Bruker, Germany) with Cu-Kα radiation (λ = 0.15406 nm) with an accelerating voltage of 40 kV and an emission current of 30 mA. The morphology and elemental composition of the synthesised nanopowder were performed using SEM (PhenomXL, the Netherlands). The specific surface area and pores distribution of the nanopowder were conducted using a Micrometrics ASAP 2020 apparatus (Degas temperature: ambient to 200 °C for 20 min with pressure range from 0 to 950 mmHg). The chemical bonding of MgO was recorded using an FTIR (Model: Nicolet 6700) in the range 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹.

2.3 Adsorption Experiment Details

It is well documented that Indigo Carmine pigment is widely utilised in a vast range of industries such as textile, food and cosmetics; however, it is extremely toxic and can induce health risks to human, biota and environment. Hence, environmental remediation of Indigo Carmine by the fabricated MgO was conducted. The equilibrium adsorption experiment was carried out in 50 ml conical flasks containing 10 mg of MgO. After 1 h of heating time, 20 ml of dye with concentrations 5, 15, 25, 35, 60 and 80 ppm were added under a magnetic stirrer for 80 min. Following the adsorption equilibrium, the suspension nanomaterials were isolated from the solution via filtration, and the residual dye concentration was determined using the spectrophotometer (Labomed, UVS-2800) at a maximum wavelength of 610 nm.

For adsorption kinetic study, the volume and initial concentration of the dye were 220 ml and 60 ppm, respectively, and the mass of the MgO was 110 mg. The experiment was conducted in the dark under vigorous magnetic stirring. Later, 4 ml of the suspension was withdrawn at predetermined time intervals, and centrifuged for measuring the remaining dye concentration. The quantity of the dye adsorbed per gram of the catalyst at time t (min) can be calculated by the following equation [14]:

where q_t (in mg g⁻¹) is the mass of the dye adsorbed by a unit mass of nanopowder m (g) at time t (min), V is the solution volume (l), C₀ is the concentration of the metal ion initially present and C_t is that concentration at time t in mg l⁻¹. At equilibrium, an analogous formula is employed to find out the mass adsorbed q_e:

Adsorption equilibrium data are generally modelled by adsorption isotherms. In particular Langmuir and Freundlich representations are widely used to delineate the established equilibrium. They are symbolised in linear forms to make their graphical illustrations pertinent.

Here q_m is the quantity of the adsorbed solid-phase that accomplishes a monolayer coverage of adsorption sites and K_L is a parameter associated with the adsorption free energy [15]. q_m and K_L represent the slope and intercept of C_e/q_e versus C_e plotting. While k and n of Freundlich formula are obtained from the ln q_e versus ln C_e diagram, K_F and n of the Freundlich model are correlated with the adsorptive bond strength and distribution, respectively [15].

Either a pseudo-first- or pseudo-second-order law typically presents adsorption kinetics data. Equation (5) describes a pseudo-first-order model [16] having a rate constant k₁ (min⁻¹). From the slope and intercept of ln(q_e – q_t) versus t plot, k₁ and q_e can, respectively, be determined.

A pseudo-second-order is modelled by (6), where k₂ denotes the rate constant [g (mg⋅min)⁻¹] [17]. Use of the t/q_t against graph, q_e and k₂ are correspondingly obtained from the slope and intercept.

Figure 1:

X-ray diffraction patterns of magnesium oxide (MgO) nanoparticles heated at different time intervals.

3.1 Crystal Structure of MgO

The XRD patterns of the MgO nanoparticles after annealing at different heating times in the 2θ range of 10°–80° are shown in Figure 1. All the diffraction peaks in these patterns are indexed as cubic MgO JCPDS card No: 01-1235, space group:Fm-3m [18]. Existence of sharp diffraction peaks at orientation (200) in the pattern is an influential indication of good crystallinity of the nanoparticles. The diffraction peaks illustrated in the XRD pattern can also be used to obtain information regarding the particle size of nanomaterials at various heating rates. The average crystallite size (D) of the nanomaterials was calculated from the XRD results using the Scherrer equation [19] and the obtained values are displayed in Table 1.

where λ, β and θ are the crystallite size, wavelength of the X-ray source (Cu K_α), full width at half maximum and Braggs’ diffraction angle, respectively. In addition, the crystallite size (D) and the strain (ε) were estimated using the Williamson-Hall equation [20].

The D values obtained from both Sherrer’s and Hall’s formulae exhibit monotonous increments as the heating time gets longer (Fig. 2a,b, Tab. 1). This observation is not an unexpected one as nanomaterials generally aggregate as more heating is done according to the Ostwald ripening phenomenon [21]. One of the explanations given is that larger particles are thermodynamically having less energy as their surface area to volume is small [22]. The particle enlargement with extended heating is manifested by the decrease in the strain calculated by both methods.

Figure 2:

(a) Williamson-Hall graph for MgO samples and (b) crystallite size from Scherrer and William-Hall calculations.

3.2 MgO Morphology

Scanning electron microscopy is one of the best techniques for the topographic study of nanoparticles and it provides important information regarding the growth mechanism, shape and size of the nanoscale materials. The surface morphology of the MgO nanopowder at different time intervals and magnifications are photographed in Figure 3a–f. The SEM images reveal that MgO nanopowder has the non-uniform pattern, whereas some of them look like the flower buds. Obviously the SEM images have clearly shown that the product is in the nanoscale form. The energy dispersive X-ray analysis data (Fig. 3) reveal the formation of pure MgO as given by the listed elemental ratios.

Figure 3:

(a)–(f) Scanning electron microscopy, energy dispersive X-ray analysis (EDX) of MgO samples and (f) EDX of MgO at 6 h.

3.3 Nitrogen Adsorption Study

Nitrogen adsorption-desorption isotherms (Fig. 4a) and the pore size distribution of the MgO nanopowder annealed at 700 °C and heating time in the range of 1–12 h are displayed in Figure 4b. The isotherms demonstrate a type IV with H2 hysteresis loops, which is a feature of mesoporous materials. The specific surface area values obtained from the BET method are 27.6, 12.0 and 6.7 m²g⁻¹ for MgO samples heated for 1, 6 and 12 h, respectively. The pore size distribution with centres about 25, 27 and 12.5 nm for MgO heated at different time intervals are observed. The BET surface area and the pore volume of the as-prepared MgO nanoparticles were 0.26, 0.12 and 0.03 cm³ g⁻¹, respectively (Tab. 2).

Figure 4:

(a) Nitrogen adsorption-desorption isotherms and (b) pore size distribution of the MgO nanopowder.

3.4 FTIR Spectra of MgO Nanoparticles

As displayed in Figure 5, the OH⁻ of the physically adsorbed water stretching vibration is indicated by the broad vibration band at 3448 cm⁻¹ [23]. The strong and broad band at 411 cm⁻¹ is associated with the Mg–O bond [23], whereas the band at 1107 and 1500 cm⁻¹ are the distinctive symmetric and anti-symmetric stretching vibrations of CO₃²⁻, respectively [11]. The weak FTIR peak at 3726 cm⁻¹ may be assigned to the A_2u(OH) lattice vibration of a minor amount Mg(OH)₂ [24] that may have resulted from the MgO hydrolysis by the atmospheric humidity.

Figure 5:

Fourier transformation infrared spectroscopy of MgO samples.

3.5 Adsorption Kinetics Study

Figure 6a exhibits the dye absorbance at different time intervals in batch experiment carried at 610 nm wavelengths. The graph indicates a gradual decrease in the dye concentration as the time was increased.

The progress of the sorption data of the dye at different time intervals is demonstrated by Figure 6b for the nanoparticles at ambient temperature. It is evident that initially the adsorption capacity (q_t) of the nanomaterials has increased sharply reaching a steady state after 50 min. In accordance, 50 min was taken as a suitable time to accomplish the equilibrium. A graph of t/q_t versus t representing the pseudo-second-order kinetics is displayed in Figure 7a. The plot shows linearity and a good fitting of the data with regression coefficient r² = 0.999. Concomitantly, the computed and the experimentally obtained values of q_m are in agreement (Tab. 3) in contrast to the value corresponding to the pseudo-second-order that deviates considerably. These results accentuate excellent fitting to the pseudo-second-order kinetic model for the dye adsorption on MgO nanoparticles. Moreover, the compliance of the dye adsorption performed in this work is concurrent with a considerable number of investigations [25], [26], [27], [28] that reported similar findings.

Figure 6:

(a) Dye sample absorbance at different wavelengths and (b) plot of the equilibrium adsorption of the dye on MgO nanopowder as a function of time.

Figure 7:

(a) Langmuir isotherm, (b) pseudo-second-order kinetic model of dye adsorption on MgO, (c) q_t versus t^1/2 plot for the intra-particle diffusion, and (d) Boyd plot for the dye adsorption on MgO nanoparticles.

3.6 Equilibrium Study

Adsorption isotherms provide explanations for the adsorption process at equilibrium conditions. By knowing the adsorbent capacity we can realise the appropriate adsorption mechanism. The most extensively investigated adsorption isotherms to mock-up the experimental data are the Freundlich and Langmuir models. For equilibrium studies, 120 min was taken as a suitable time to accomplish the equilibrium. Figure 7b shows the linearised Langmuir adsorption isotherm for the dyes on MgO at room temperature where the batch experiment data were fitted using the least squares method. The calculated isotherm parameters (q_m, K_L, K_F and n) in addition to the regression coefficient (r²) are listed in Table 4. The high value of the correlation coefficient, close to the unit, is an apparent clue for the better fitting of the Langmuir model for the dye adsorption on MgO equilibrium data. This finding is consistent with the data from the previous investigations for the removal of ciprofloxacin [28], reactive and vat dyes [27], fast orange and bromo-phenol blue dyes [26]. From the Langmuir model, the amount of dye removed (q_m) was found to be 159 mg⋅g⁻¹ which is higher than some previously reported values for dyes removal by MgO nanostructures. For instance the value reported in this study is higher than the 81 mg ⋅ g⁻¹ of Orange G adsorbed by commercial MgO, 156 mg ⋅ g⁻¹, by flower-like mesoporous MgO microspheres [26], 59.17 mg⋅g⁻¹ methylene blue dye [29] and 21.5 mg⋅g⁻¹ for RR195 and OG dyes [25]. The competence of MgO nanoparticles for the Indigo Carmine elimination showed a maximum adsorption capacity of 158.0 mg⋅g⁻¹ which is superior to the mesoporous Mg/Fe [30], Co(OH)₂ nanoparticles [31] and chitosan [32] that achieved 62.8, 62.5 and 72.6 mg⋅g⁻¹, respectively. Thus the adsorbent produced by the pyrolysis of MgCO₃ manifested better adsorption efficiency than other nanoadsorbents indicating that MgO is a promising adsorbent of Indigo Carmine.

An essential trait of the Langmuir isotherm is a dimensionless equilibrium constant R_L termed as the separation factor or equilibrium parameter [16].

where C₀ is the initial concentration and K_L is the constant related to the energy of adsorption (Langmuir Constant). The R_L value classifies the adsorption mode as irreversible if R_L = 0, favourable if 0 < R_L < 1, linear if R_L = 1 and unfavourable if R_L > 1. In addition to the r² ≈ 0.9879, the tabulated R_L value (0.9360) provides an additional support for the appropriateness of the Langmuir model to the experimental data obtained in this work.

3.7 Mechanism of Adsorption

The uptake of dyes from the solution can be influenced by a mass transfer process at the boundary film of liquid or via an intra-particular mass transfer. An external mass transport coefficient, β_L (m⋅s⁻¹) of a dye at the boundary film is assessed using the equation: [33], [34].

where C_t and C₀ (both in mg ⋅ l⁻¹) are the respective dye concentrations at time t and zero, K_a(l⋅g⁻¹) is related to the Langmuir constants: K_a = Q₀⋅b; m(g) stands for the adsorbent mass, and S_S represents the adsorbent surface area (m²⋅g⁻¹). The liability of the model is achieved by ln[(C_t/C₀ − 1/(1 + m ⋅ K_a)] versus t linear relationship. Adsorbed species may alternatively be conveyed from the bulk of solution to the solid phase through intra-particle diffusion/transport process. Intra-particular diffusion is a limiting step in a number of adsorption processes. The likelihood of intra-particular diffusion is investigated using the Weber and Morris diffusion approach [35].

where C is the intercept and k_dif is the intra-particle diffusion rate constant. The k_dif values for the tested adsorbent are calculated from the slopes of the plots (Fig. 7c) and reported in Table 5 two-stage or multi-linear sorption [36]. It is clearly displayed in the intra-particle diffusion graph of q_t versus t^1/2 (Fig. 7c). The first sharp stage may be ascribed to the dye diffusion through the solution to the external surface of nanoparticles and then crossing the boundary layer to reach the surface. In the meantime, the second stage may signify a final equilibrium where the intra-particle diffusion starts to decline due to the lower concentration gradient of the dye.

The tabulated data (Tab. 5) show a decrease in the diffusion rate as contact time was lengthened.

This can be rationalised by the fact that as contact time increases less pores will be accessible for diffusion as the dye had previously diffused into the inner structure of nanoparticles in the first stage. This is supported by the less value of rate constant k_dif2 relative to k_dif1. Moreover, the parameter (C) that indicates the boundary layer thickness is greater in the second-stage manifesting enhanced boundary layer effect [37]. The plots do not pass the origin and reveal a two-stage linearity, and this designates that the rate determining step of the dye adsorption on MgO adsorbent is not solely controlled by an intra-particle diffusion mechanism. Other mechanisms such as bulk diffusion and film diffusion may have contributed to the diffusion process [38].

Boyd model is generally a useful indicator on kinetic data to validate the contribution of film diffusion. The model is formulated as [39]:

where F is the fraction of the material adsorbed at any time t (min) and β_t is an arithmetical function of F. As the linear plots plot of β_t versus t (Fig. 7d) does not pass through the origin, it can be derived that there is a film diffusion control of the dye adsorption process. Assuming the MgO to be an spherical adsorbent nanoparticles, the overall rate constant of the adsorption process can be related to the pore and film diffusion coefficients as given by [40]:

The terms r₀, δ, c¯ and c stand for adsorbent radius, film thickness (10⁻³ cm) [41], amount adsorbed and initial concentration, respectively. Substituting the relevant values, D_p and D_f will be, respectively, 4.30 × 10⁻¹² and 9.02 × 10⁻⁷ cm²s⁻¹. For film diffusion controlled diffusion, the D_f value should range between 10⁶ and 10⁻⁸ cm² s⁻¹. Whereas for pore diffusion controlled process the D_p value can range between 10⁻¹¹ and 10⁻¹³ cm²s⁻¹. It is apparent that the diffusion of the dye is mutually controlled by the pore and film diffusion processes [42]. This finding is consistent with the two-stage sorption process shown in Figure 7c.