Adsorption of phenol over bio-based silica/calcium carbonate (CS-SiO₂/CaCO₃) nanocomposite synthesized from waste eggshells and rice husks

Chemical and reagents

The waste eggshells (ES) and rice husks (RH) used in this work were obtained from local fast-food restaurants in Mubi, Adamawa state Nigeria. The Low-molecular-weight Chitosan (CS) (75–85% degree of deacetylation) was purchased from Sigma-Aldrich (St Louis, MO, USA). Ethanol (C₂H₅OH, 99.7%), hydrochloric acid (HCl, 37%), sulfuric acid (H₂SO₄, 98.5%), hydrogen peroxide (H₂O₂), sodium hydroxide (NaOH, 97%), acetone (C₃H₆O, 99%), acetic acid (CH₃COOH, 99.85%) and phenol (C₆H₆O, 99%) were purchased from BDH Chemical Ltd England.

Preparation and synthesis of the adsorbents

The preparation of the adsorbent follows sequential steps. Briefly, a thoroughly washed RH was subjected to acid pretreatment using HCl: H₂SO₄ (10:30 wt%) and further treatment with H₂O₂ (30% v/v) at 70 °C for 60 min (Lu & Hsieh, 2012; Thuc & Thuc, 2013). The formed slurry was rinsed with distilled water, oven-dried at 600 °C for 4 h. The sieved RH is then dispersed in an aqueous 0.5M NaOH solution under vigorous stirring for 1 h to form the RH-derived sodium silicate (RH-Na₂SiO₃). The RH-silicate (RHSiO₂) is obtained by subjecting RH-Na₂SiO₃, in 12 wt% H₂SO₄ under vigorous stirring for 15 min. The formed RHSiO₂ is washed in water and aged in ethanol at 60 °C for 60 min. The final material is further rinsed in water and dried under ambient temperature (Lu & Hsieh, 2012; Le et al., 2013; Ghorbani, Sanati & Maleki, 2015). The proteins in the Chicken eggshells are deactivated/denatured by boiling after subjecting the same to washing using acetone/alcohol (1:1 ratio), dried, and made into powder. The powdered samples were further dispersed in distilled water under sonication, collected and heated in an oven at 600 °C for 3 h, and finally sieved using a 40–75 µm to form the ES-calcium carbonate (ESCaCO₃) particle (Minakshi et al., 2019).

An aqueous solution of ESCaCO₃ (5 wt%) was prepared under vigorous stirring at 3,000 rpm for 30 min. To the formed precursor, 1 g of RHSiO₂ in 20 ml of distilled water was added gradually under heating at 80 °C and stirred for 60 min to form the composite mixture of RHSiO₂/ESCaCO₃ particles (Morsy, El-Sheikh & Barhoum, 2019). The Bio-based CS-SiO₂/CaCO₃ nanocomposites were prepared by the dropwise addition of aqueous RHSiO₂/ESCaCO₃ (1 g/20 mL) solution into chitosan (CS) suspensions (0.1 g in 10 mL of 1% acetic acid) under stirring for 30 min in other to stabilize the particles. The formed precursor was incubated at 60 °C for 3 h, collected by centrifugation, and further allowed to age in 2.0M NaOH solution for 2 h. Then wash using distilled water and filtered to obtain the CS-SiO₂/CaCO₃ nanocomposites.

Instrumental analysis

In this study, different instruments were used at various stages of the analysis to characterize the prepared samples. The structure and phase identification of the prepared RHSiO₂, ESCaCO₃, and the CS-SiO₂/CaCO₃ nanocomposites were established using an X-ray diffraction (XRD) instrument (Malvern PAAnalytical Empyrean, Nederland) equipped with Cu anode (λCuKα radiation source) X-ray tube. The identification of the functional groups was determined using Fourier Transform Infrared (FTIR) spectroscopy using PerkinElmer (Waltham, MA, USA, N3895) with a range covering 4,000–500 cm⁻¹. The elemental composition and morphology were studied using Field Emission Scanning Electron Microscopy and Energy Dispersive X-ray (FESEM/EDX) on JEOL JSM-7600F (JEOL, Tokyo, Japan).

Adsorption experiment

The adsorbents were brought into contact with aqueous phenol with a concentration ranging from 10 to 30 mg/L and agitated at a speed of 120 rpm at room temperature. The percentage of phenol in each of the aliquot samples taken were determined after 30, 45, 60, 70, 80, and 90 min by measuring the difference in absorbance using UV–Vis Spectrophotometer (PerkinElmer Lambda 35 spectrometer). The study was conducted using changes in pH (4, 5.4, 6, 7.4, and 8) and adsorbent dose (0.05, 0.1, 0.15, 0.2, and 0.25 g). The adsorption kinetics were determined using the Pseudo-first-order and Pseudo-second-order kinetic model. The Pseudo-first-order kinetic model was determined using Eq. (1): (1)

Characterization of adsorbent

FTIR analysis

The fingerprints or functional groups involved in the various stages of the synthesis identified by FTIR are shown in Fig. 1. In the figure, the broadband in the spectra of ESCaCO₃ centered around 3,104–3,630 cm⁻¹ is an O–H bond from water, while the peak at 1,651 cm⁻¹ is the adsorption band of H–O–H bending vibrations from adsorbed water molecules (Ihli et al., 2014; Habte et al., 2019). The three peaks at 1,413 cm⁻¹, 873 cm⁻¹, and 712 cm⁻¹ are absorption peaks due to the CO⁻²₃ group in CaCO₃ (Nyquist & Kagel, 2012; Morsy, El-Sheikh & Barhoum, 2019; Du & Amstad, 2020). The peaks at 3,285–3,629 cm⁻¹, 1,651 cm⁻¹, 1,056 cm⁻¹, and 796 cm⁻¹ in the spectra of RHSiO2 are from the silanol OH groups of water, the H–O–H bending vibration, siloxane asymmetric bending vibrations bands (Si–O–Si), and the symmetric bending mode of silanol (Si–O) (Kamath & Proctor, 1998; Ghorbani, Sanati & Maleki, 2015). Chitosan’s spectra show stretching vibration of OH group around 3,372 cm⁻¹ and a –CH₂ symmetric vibration at 2,878 cm⁻¹. The peaks at 1,645 cm⁻¹, 1,378 cm⁻¹, and the prominent peak at 1,031 cm⁻¹ are C=O vibration mode of amides, C–N stretching, and C–O stretching vibrations, respectively (Saifuddin & Dinara, 2012; Dehaghi et al., 2014).

The broadband around 3,312–3,618 cm⁻¹ and the peak centered at 1,643 cm⁻¹ are fingerprints from O–H bond and H–O–H bending vibrations of ESCaCO₃, and RHSiO₂, in the CS-SiO₂/CaCO₃ nanocomposite. The absorption peaks centered at 1,412 cm⁻¹ and 873 cm⁻¹ are from the CO⁻²₃ group in ESCaCO₃, while the peaks at 1,065 cm⁻¹, 795 cm⁻¹, and 713 cm⁻¹ are from the Si–OH stretching, Si–O–Si asymmetric, and Si–O–Si symmetric stretching vibrations from the RHSiO₂ respectively (Nyquist & Kagel, 2012; Morsy, El-Sheikh & Barhoum, 2019). The Si–O–Si asymmetric vibrational mode from the spectra of RHSiO₂ centered at 1,056 cm⁻¹ slightly shifted to 1,065 cm⁻¹ in the nanocomposites, suggesting a possible interaction with the C–O stretching vibrations of CS (1,031 cm⁻¹).

XRD analysis

Figure 2 represents the XRD patterns of (a) ESCaCO₃, (b) RHSiO₂, (c) CS and, (d) CS–SiO₂/CaCO₃ nanocomposite. Figure 2A shows the XRD patterns for ESCaCO₃ having all the diffraction peaks corresponding to the standard pattern for calcite (JCPDS card no. 005-0586), indicating 81% CaCO₃ (Hassan, Rangari & Jeelani, 2014; Minakshi et al., 2019). The maximum peaks at 2θ = 29.46° reflect the calcite phase of CaCO₃. This result agrees with the findings reported by Chraibi et al. (2016), showing the peak at 2θ = 29.48° for CaCO₃ derived from the eggshell. Figure 2B shows the XRD patterns of silica prepared from rice husk (RHSiO₂). The peak corresponding to the semi-crystalline phase of Si appeared at a broad peak centered at 2θ = 22.08°. The analysis of the particles indicates the presence of silica (96%). The result agrees with the study by Morsy, El-Sheikh & Barhoum (2019) showing 82% silica and a broad peak centered at 2θ = 22.5° corresponding to silica nanoparticles’ semi-crystalline phase. A strong peak centered at ~20° was observed in the spectra of pure chitosan (Fig. 2C). This high degree of crystallinity was observed to transcend to a very weak peak that is not noticeable in the spectra of CS–SiO₂/CaCO₃ nanocomposite, arising from possible disarray in the chain alignment of chitosan by the diffraction peaks of calcite and the semi-crystalline phase of Si (Dehaghi et al., 2014). Furthermore, Fig. 2D shows the XRD patterns of the CS–SiO₂/CaCO₃ nanocomposites. The results show the diffraction peaks corresponding to the calcite phase of CaCO₃ at 2θ = 29.51° and a semi-crystalline phase related to silica at 2θ = 22.79°. The result further shows that CS–SiO₂/CaCO₃ nanocomposite predominantly consists of 9% silica and 52% CaCO₃.

FESEM/EDX analysis

Figure 3 shows FESEM/EDX micrographs with the corresponding particle size distribution of ESCaCO₃, RHSiO₂, and CS-SiO₂/CaCO₃ nanocomposites. The micrograph of ESCaCO₃, as shown in Fig. 3A, reveals an irregular surface structure with a random assembly of aggregated grains that form a densely packed structure and size distribution up to 100 nm (Fig. 3B). Similar morphological features were reported by Minakshi et al. (2019) and Ahmad et al. (2020), showing the structure of eggshells-CaCO₃ having an irregular surface structure with a size distribution ~200 nm and 89 nm, respectively. The morphology of the RHSiO₂, as shown in Fig. 3C, shows nearly spherical densely packed grains with an average particle size of 19.07 ± 4.63 nm (Fig. 3D). The images are similar to those reported by Le et al. (2013), Morsy, El-Sheikh & Barhoum (2019), and Phoohinkong & Kitthawee (2014). The result in Fig. 3E shows that the CS-SiO₂/CaCO₃ nanocomposite have better and good monodispersed spherical morphology than those of the single EsCaCO₃ particles. The dispersion of the SiO₂/CaCO₃ into CS suspension increases the uniformity of the grain size of CS–SiO₂/CaCO₃, with an average size distribution of 32.15 ± 6.20 nm (Fig. 3F). The result falls near to the findings reported by Morsy, El-Sheikh & Barhoum (2019), showing SiO₂/CaCO₃ nanocomposites to be nearly spherical with particle sizes ranging 70–110 nm. The corresponding EDX results, as presented in Figs. 3A, 3C, and 3E, showed representative peaks for Ca, Si, and O for all the samples. The EDX of ESCaCO₃ consists mainly of C (36%), O (52%), Ca (10%), while the spectra of RHSiO₂ show majorly the peaks of Si (30%) and O (69). The CS-SiO₂/CaCO₃ nanocomposite show similar peaks corresponding to the spectra of SiO2 and CaCO3, which indicate a successful formation of CS-SiO₂/CaCO₃ nanocomposite.

Factors affecting adsorption behaviors

The Phenol removal efficiency of the adsorbent, as shown in Fig. 4A, shows the adsorption capacity increases in the order of CS-SiO₂/CaCO₃ > RHSiO₂ > ESCaCO₃ with an increase in the adsorbent dose. The synthesized CS-SiO₂/CaCO₃ nanocomposite due to additional binding sites from SiO₂, CaCO_3, and CS adsorbed more phenol than the RHSiO₂ and ESCaCO₃ particles_. The study was conducted at a varying initial dosage of 0.05–0.25 g, pH 5.4, and initial phenol concentration of 25 mg/L at 120 rpm for 60 min. The highest removal efficiency of 80% was achieved at a dose of 0.15 g of CS-SiO₂/CaCO₃ nanocomposite and remain more or less the same up to 0.25 g. Suggesting that the removal efficiency at the onset of the adsorption processes was higher due to the more available surface area, additional adsorption sites, and then remained the same due to saturating absorbent sites following an increase in the dose (Mandal, Mukhopadhyay & Das, 2019). The RHSiO₂ and ESCaCO₃ at varying initial dosage show an increase in phenol removal efficiency. The highest removal efficiency of phenol was observed at a dose of 0.25 g of RHSiO₂ (75%) and ESCaCO₃ (76%), respectively. An increase in phenol removal efficiency with an increase in adsorbent dosage was also observed and reported by Sarker & Fakhruddin (2017) and Asgharnia et al. (2019) using rice straw as adsorbent. Similar observations were reported using CaCO₃ derived from the eggshell. The study reported an optimum dosage of eggshell powder adsorbent at 4 g (Kashi, 2017).

Figure 4B described the adsorption efficiency of the adsorbent at varying pH environments. The study was carried out at varying pH 4, 5.4, 6, 7.4, and 8, with an initial phenol concentration of 25 mg/L and an adsorbent dose of 0.15 g at 120 rpm for 60 min. The plots show higher removal efficiency occurring at acidic pH range (4–6) and decreasing at pH of 7.4 and 8. At these pH ranges, the acidic environment substantially increases the electrostatic attraction with the phenolates’ negative ions, increasing phenol adsorption in the process (Khare & Kumar, 2012; Daraei et al., 2013; Ouallal et al., 2019). The adsorption processes following the pH variation were observed to follow the trend CS–SiO₂/CaCO₃ > RHSiO₂ > ESCaCO₃. The highest percentage removal of 83% was observed in the final nanocomposite (CS–SiO₂/CaCO₃) at a pH of 5.4. The presence of available exchangeable ions (Kashi, 2017), multiple adsorption sites on the adsorbent at this pH range leads to the observed removal efficiency in the CS–SiO₂/CaCO₃ nanocomposite compared to the other adsorbents. Asgharnia et al. (2019), while investigating the effect of pH on phenol removal efficiency, observed higher efficiency for rice husk activated carbon (RHAC) at pH 6 and rice husk carbon (RHC) at pH 5. The result further agrees with the findings reported by Kashi (2017) and Daraei et al. (2013) using powdered eggshell.

Varying the initial concentration of phenol from 5–30 mg/L, as shown in Fig. 4C, shows the removal efficiency increases as the concentration increases from 5–25 mg/L and remains the same at 30 mg/L. The CS-SiO₂/CaCO₃ nanocomposite gives the highest removal efficiency (87%) at a concentration range of 25 mg/L and 30 mg/L. The removal efficiency follows the trend of CS-SiO₂/CaCO₃ > RHSiO₂ > ESCaCO₃, at a fixed adsorbent dose of 0.15 g, pH 5.4 at 120 rpm for 60 min. An increase in concentration leads to a rise in phenol adsorption on the surface of the sorbent. But above 25 mg/L, the removal efficiency remains the same, which suggests a saturation of the sorption sites available for sorption at this concentration of the sorbate molecules (Dehaghi et al., 2014; Sarker & Fakhruddin, 2017).

Figure 4D shows the effect of contact time (30, 45, 60, 70, 80, and 90) on the percentage removal of phenol using 0.15 g of adsorbent in 20 ml of 25 mg/L phenol, pH 5.4 at 120 rpm. The result shows an increase in phenol removal efficiency after 60 min of contact with CS–SiO₂/CaCO₃ nanocomposite and RHSiO₂ particles. The CS-SiO₂/CaCO₃ nanocomposite shows removal efficiency of 84% after 60 min and 89% at 90 min of contact time. The result agrees with findings reported by Asgharnia et al. (2019) and Sarker & Fakhruddin (2017) using RH as an adsorbent.

Kinetic adsorption study

To further understand phenol adsorption behavior and mechanism onto CS–SiO₂/CaCO₃, RHSiO₂, and ESCaCO3, the study was subjected to kinetic analysis using pseudo-first-order and pseudo-second-order models. The models help to predict possible interaction mechanisms between the adsorbent and the adsorbate. The result is shown in Fig. 5 and further elaborated in Table 1. The plot from the graph of t/qt vs t fit well with pseudo-second-order with the highest correlation coefficient (R²) of 0.99 and qm values of 3.80 mg/g observed in CS–SiO₂/CaCO₃ nanocomposite (Fig. 5A). The result further shows phenol adsorption over RHSiO₂ (R² = 0.98) and ESCaCO₃ (R² = 0.95) fitting well with the pseudo-second-order. The outcome suggests that the adsorption process is chemisorption and the adsorbent particles are heterogeneous (Mandal, Mukhopadhyay & Das, 2019). On the other hand, the pseudo-first-order graph of log (q_e−q_t) against t, as shown in Fig. 5B, poorly described the absorption processes of ESCaCO₃ (R² = 0.72), but a bit linear for CS-SiO₂/CaCO₃ nanocomposite (R² = 0.91) and RHSiO₂ (R² = 0.90) respectively. Pseudo-first-order suggests that the adsorption process is physisorption and the adsorbent particles are homogeneous (Husein, Hassanien & Al-Hakkani, 2019). This study’s result agrees with the findings reported by Mandal, Mukhopadhyay & Das (2019). Reporting pseudo-second-order supporting the adsorption processes of phenol using rice husk as adsorbent. Though pseudo-second-order best described the adsorption processes, the R² derived from the pseudo-first-order model of CS–SiO₂/CaCO₃ and RHSiO₂, though poorly described the adsorption process further characterize the homogeneity or the heterogeneity of the adsorption sites (Ouallal et al., 2019). These observations also suggest some semblance of pseudo-first-order mechanism influencing the adsorption process of CS–SiO₂/CaCO₃ nanocomposite and RHSiO₂ particles. Which further aggress with isotherm result from this study, showing the adsorption processes to possess good monolayer coverage on the surface and heterogeneous surface with different binding sites for CS–SiO₂/CaCO₃ nanocomposite (Song, Johnson & Elimelech, 1994). The higher values of K₂, compared to K₁ (Table 1), further suggest a pseudo-first-order reaction’s applicability (Ali, AL-Othman & Alwarthan, 2016).

Adsorption isotherms study

As presented in Fig. 6, the result showed the equilibrium data for all the adsorbents at an initial concentration of 5–30 mg/L, demonstrating a better matching to the Langmuir isotherm model. Observed to follow the trend CS–SiO₂/CaCO₃ > ESCaCO₃ > RHSiO₂. Chraibi et al. (2016) and Kashi (2017), in a separate study, show Langmuir isotherm describing the adsorption of phenol onto a calcined eggshell. Giraldo & Moreno-Piraján (2014) observed similar findings over activated carbons derived from eggshell particles. Asgharnia et al. (2019) also reported a better matching of phenol adsorption with the Langmuir isotherm model using RHS and RHAC as adsorbents.

Except for CS–SiO₂/CaCO₃ nanocomposite (Fig. 6A), the adsorption data for ESCaCO₃ (Fig. 6B) and RHSiO₂ (Fig. 6C) base on their R² values, fitted poorly to the Freundlich isotherm model. However, the magnitude of their K_F (4.60 and 3.3) showed the tendency of phenol adsorption over the adsorbents (Khare & Kumar, 2012). The Langmuir equation and Freundlich model in this study show a higher correlation coefficient (R² = 0.9912 and 0.9905) for phenol adsorption onto CS–SiO₂/CaCO₃ nanocomposite (Fig. 6A). Suggesting good monolayer coverage on the surface of the adsorbent (Langmuir) and equally possess heterogeneous surfaces with available binding sites (Freundlich) (Kermani et al., 2012). The K_F (3.4) and 1/n (2.5) values calculated from the Freundlich isotherm for CS–SiO₂/CaCO₃ nanocomposite further suggest that the sorption process is complimentary and physicochemical (Song, Johnson & Elimelech, 1994). The magnitude of 1/n indicates that phenol was favorably adsorbed by CS–SiO₂/CaCO₃ nanocomposite compared to the other adsorbent (Ouallal et al., 2019). From the data, the magnitude of qm, which defines the amount of phenol per unit weight of sorbent, was higher for CS–SiO₂/CaCO₃ nanocomposite in comparison to RHSiO₂ and ESCaCO₃ particles. Similar results were also reported by Kermani et al. (2012) using rice husk ash as an adsorbent. In the study, both the Langmuir and Freundlich isotherm model adequately described the adsorption process. Djebbar et al. (2012) reported similar findings for phenol’s adsorption using natural clay.

The R_L estimation at various C_i further shows the adsorption to be more favored by the Langmuir isotherm model, having values of <1 for all the adsorbents. The results indicate favorable adsorption in all cases. Similar observations have been reported for phenol’s sorption on rice husk and chitin (Milhome et al., 2009; Chraibi et al., 2016). Table 2 shows the correlation coefficients (R²), the Langmuir constants (qm, and K_L), the Freundlich parameters (K_F, 1/n), and the separation factors (R_L) estimated from this study.

Some authors reported materials derived from rice husk/straw to show an adsorption capacity of 13.9 (Mandal, Mukhopadhyay & Das, 2019). Sarker & Fakhruddin (2017) reported an adsorption capacity of 3.8 and 5.8 mg/g from a similar study using rice straw. Nadavala et al. (2009) and Du et al. (2020) reported an adsorption capacity of 1.26 and 8.55 mg/g using chitosan as adsorbents. About 30–58 mg/g were also reported using calcine eggshells by Chraibi et al. (2016) and 191 mg/g from activated carbon derived from eggshells (Giraldo & Moreno-Piraján, 2014). The studies reported and presented in Table 3 show phenol adsorption capacity that is either higher or lower than the values recorded in this study. However, research shows that the adsorption capacity of an adsorbent is a function of so many parameters used in the synthesis and adsorption studies. Parameters influencing adsorption processes include the following; the initial phenol concentration, the adsorbent’s characteristics, the adsorbent dose, the particle size of the adsorbent, temperature, pH, and contact time (Mandal, Mukhopadhyay & Das, 2019).

Adsorption and removal mechanism of phenol over CS–SiO2/CaCO3 nanocomposite

The starting materials used in the adsorbents’ synthesis (ESCaCO₃.RHSiO₂ and, CS-SiO₂/CaCO₃) are low-cost, readily available, and naturally derived from agricultural waste. Phenol adsorbed on the adsorbent can be decomposed upon re-calcination or incubation in an alkaline medium (Tabana et al., 2020). The removal of phenol was carried out by incubating the spent CS-SiO₂/CaCO₃ nanocomposite in 0.1 M NaOH solution for 60 min at 45 °C followed by washing using DI water (Dehaghi et al., 2014; Tural, Ertaş & Tural, 2016). Fig. 7 shows the FTIR result for CS–SiO₂/CaCO₃ nanocomposite. The study shows broadband at 3,500–3,000 cm^−1, and the small peak at 1,643 cm⁻¹ disappeared in CS–SiO₂/CaCO₃ nanocomposite spectra after phenol adsorption. Suggests electrostatic interaction of the phenolate with the adsorbent’s hydroxide layers (Ouallal et al., 2019). The CS–SiO₂/CaCO₃ nanocomposite on contact with phenol further shows a change in peak intensity of CO²⁻₃ at 1,410 cm⁻¹ from 1,412 cm⁻¹ before the adsorption. The phenol adsorption influences the Si–OH stretching vibration activity at 1,065 cm⁻¹ to 1,045 cm⁻¹. Similarly, a change in the asymmetric and symmetric stretching mode at 795 cm^–1 associated with Si–O–Si was observed on interaction with phenol to 790 cm⁻¹.

Furthermore, the spectra on the phenol’s desorption from the CS–SiO₂/CaCO₃ nanocomposite show a broad contour’s appearance, suggesting to be from moisture-induced crystallization (Ouallal et al., 2019). Moisture induced crystallization could arise from a humid environment (Singer et al., 2012) or at least in parts related to moisture contents directly bound to the atoms (Ihli et al., 2014; Jensen et al., 2018) or structurally trapped in the crystals or due to fracture of hydrogen bonds formed by H₂O…CO²⁻₃ hydrogen bonds. (Cheng, Sun & Wu, 2019; Du & Amstad, 2020). Certain soluble additives also facilitate moisture-induced crystallization formation (Cheng, Sun & Wu, 2019; Du & Amstad, 2020). The CS–SiO₂/CaCO₃ nanocomposite on phenol adsorption shows a shift in the peak at 1,410 cm⁻¹, 1,045 cm⁻¹, and 874 cm⁻¹, which corresponds to the peaks at 1,427 cm⁻¹, 1,074 cm⁻¹, and 873 cm⁻¹ after the desorption.