CeO2/Fe3O4/g-C3N4 nanohybrid for adsorptive removal of Rose Bengal from aqueous stream

ABSTRACT Rose Bengal (RB) is a halogen-containing water-soluble dye that is frequently used for therapeutic purposes. The dye, on the other hand, is poisonous and may cause inflammation, stinging and other adverse reactions in eyes and skin of living beings as it comes in water resources with medical waste. As a reason, it is thought to be desirable to propose a systematic approach for removing or minimising the effect of RB dye via adsorption. Here in this work, a novel CeO2/Fe3O4/g-C3N4 (CFC) nanocomposite has been synthesised by the co-precipitation-assisted hydrothermal method. The morphological and structural properties were studied by powder HRTEM, XRD, EDAX, FTIR and XPS, and also, the surface area of the CFC was measured by BET. UV-DRS was employed for the band gap calculation. This nanocomposite showed excellent adsorption properties for Rose Bengal (RB). The synthesised nanostructure exhibited a maximum adsorption capacity of 83.05 mgg−1 for RB dye. Furthermore, the nanocomposite has shown adsorption efficiency close to 99% after being regenerated and reused up to five times. It followed pseudo-second-order kinetics and an isotherm best fit with the Langmuir model. Based on the adsorption studies and FTIR studies, it is confirmed that the hydrogen bonding and electrostatic interactions are the predominating factors for the adsorption process.


Introduction
Rapid industrialisation, population growth, urbanisation, extensive use of non-sustainable resources and uncontrolled exploitation of natural resources are causing significant, irreversible and severe environmental degradation [1].Our natural resources are being jeopardised by effluents and discharges from various industries and anthropogenic activities [2].The aquatic bodies, such as lakes, ponds, rivers, and oceans, are the most vulnerable.Everyday, millions of gallons of wastewater are discharged by textiles, paper, pulp, chemicals, fertilisers, pesticides, metal plating, batteries, food processing, refineries and pharmaceutical industries, polluting our neighbouring terrestrial lands and water bodies [3,4].The majority of the pollutants listed above are resistant to decomposition and can harm human health and living organisms, causing cancer, congenital impairments and immune and reproductive system dysfunction and jeopardising baby and child growth, among other things [5].The major harmful pollutants include dyes, pesticides, medicinal components, phenolic compounds, fertilisers, hydrocarbons, plasticisers and so on [6][7][8][9].Organic dyes are complex compounds that impart colour to fabrics and materials and are usually resistant to detergents and heat.Dyes are used in a wide range of products and are an essential part of the manufacturing process.There are over 100,000 different types of dyes available on the market, with an annual production of about 7 × 10 5 tonnes [10].The majority of dyes are water-soluble, have little biodegradability and can be challenging to detect at low concentrations [11].The release of organic dyes into the hydrosphere produces an unwanted colour and reduces sunlight penetration, jeopardising aquatic life's photochemical and biological processes.Many dyes are known to be hazardous and have been classified as mutagenic, carcinogenic, or teratogenic [12].The use of dye-contaminated water poses a health risk because it can harm human organs such as the kidney, liver, brain and central nervous system and cause skin allergies [13,14].
Rose Bengal (RB) (Figure 1), commonly known as Acid Red 94, is a fluorescein dye that contains halogens.It is a pink-coloured water-soluble dye.It has a variety of medical diagnostic uses, including liver function tests, staining of necrotic tissue in the eyes and devitalised corneal cells in conjunctivitis [15].Rose Bengal is also used to treat eczema, and psoriasis with a low concentration-dose Rose Bengal has strong toxic effects on the human corneal epithelium, in addition to its many clinical applications [16,17].When the dye comes into contact with skin, it causes irritation, itching, scaling, reddening and sometimes blistering.Inflammation, redness, watering and itching in the eyes are other symptoms.When inhaled, it destroys mucous membranes, causing respiratory discomfort in people [18].
Dye-contaminated water serves no purpose in the home or in agriculture, but instead pollutes the environment, resulting in water shortages.In various parts of the globe, water shortage is a threat, and many people are unable to meet their basic necessities [19].The removal of various types of organic contaminants like dyes from wastewater using economically viable and environmentally safe methods garners a lot of attention.In this framework, the study of the interaction between organic pollutants in water and the materials utilised to remove or degrade them efficiently and economically is becoming increasingly relevant for technological implementation.The adsorption technique is one of the most effective methods for removing contaminants from water.The adsorption technique's ability to remove and recover bulkier organic molecules without breakdown has also made it more popular than other current analytical, electrochemical and photochemical approaches [20][21][22][23].According to reports, the bulky dye Rose Bengal's breakdown products are far more hazardous than the original molecule.Adsorption is widely recognised as one of the most effective techniques for removal of expensive dye Rose Bengal from water, with numerous benefits over other physico-chemical approaches such as electrochemical treatment [24], photochemical treatment [25], biodegradation [26], froth flotation [27] and reverse osmosis [28].
Numerous absorbents have already been developed for the adsorption of RB dye from wastewater; however, no adsorbent has been shown to be capable of fully removing the dye with high efficiency.As a result, the design of adsorbents with improved RB dye adsorption efficiency is of significant importance.Moreover, some adsorbents with good adsorption capacity for RB dye removal have been reported, e.g.chitosan-TiO 2 nanocomposite [29], bottom ash [18], Fe(III)-Montmorillonite [30], carbon nanosphere [31], Cu 2 O nanoparticles [32], biosorbents [33] and hierarchical multiparous cellulose beads [34].However, these sorbents are suffering from limitations, such as high-temperature synthetic approaches and aggregation of particles, leading to decreased adsorption capacity [35].Hence, it is necessary to design a new low-cost material that provides a simple and environmentally benign method for adsorption of RB dye completely.In view of the above, our research has cantered on the employment of metal oxides in cooperation with appropriate nanosheet composites for the adsorption of RB dye.
Nanosized metal oxides (NMOs), such as nanosized ferric oxides, manganese oxides, titanium oxides, magnesium oxides, aluminium oxides and cerium oxides, are the most interesting adsorbents for removing heavy metals [36][37][38] and for dye removal [39][40][41][42] from aqueous systems amongst the existing adsorbents.This is due in part to their vast surface areas and high activity levels, which are a result of the size-quantisation effect [43,44].Among all these, although CeO 2 is very less reported for adsorption purposes still cerium oxide nanoparticles have gained a lot of interest because of their unusual properties including UV absorption, thermal stability and electrical properties.It has been utilised in a variety of large applications, including solar cells, photocatalytic photodegradation under solar light and organic material extraction from wastewater [45].According to recent research, several NMOs have a high capacity and sensitivity for sorption of contaminants, resulting in significant removal of hazardous metals to fulfil more stringent requirements [46].However, as metal oxides shrink in size from micrometres to nanometres, the increasing surface energy causes them to lose stability.As a result of van der Waals forces or other interactions, NMOs are prone to agglomeration and their high capacity and sensitivity are significantly lowered or completely destroyed [47].To improve its applicability and sensitivity in practically treating wastewater, NMOs were subsequently incorporated onto porous supports of high size to create composite adsorbents.Activated carbon, carbon nanosheets, natural materials, synthetic polymeric hosts, and others are examples of porous supports that are usually employed [48].Due to its low cost, remarkable physicochemical characteristics and good chemical and thermal stabilities, graphitic carbon nitride (g-C 3 N 4 ), a structurally graphite-like material, has recently been intensively researched in photocatalytic hydrogen generation [49] and water contaminant removal [50].However, due to its low surface area, restricted functional groups and few binding sites, studies on g-C 3 N 4 as an adsorbent to adsorb dyes were few.To adsorb MB, Peng and colleagues produced a variety of mesoporous carbon nitrides and the adsorbent showed a notable effect [51].Keeping noteworthy properties of g-C 3 N 4 , it could be employed as a support material for designing nanocomposites.Magnetic nanosized metal oxides, in addition to conventional nano-sized metal oxides, are gaining popularity as they can be readily isolated from water after completion of the adsorption process using a magnetic rod [52].Magnetic NMO-based composite adsorbents also allowed for simple water solution separation for recycling or regeneration [53].This ease of separation is an important and essential step for increasing operational efficiency and lowering costs during wastewater treatment.
In this work, we have designed a nanocomposite, CeO 2 , as nano-sized metal oxide with g-C 3 N 4 sheet as a support material, and magnified it with Fe 3 O 4 for easy recovery.To the best of the knowledge of the authors, a very smaller number of CeO 2 -based nanocomposites are reported for the adsorption of dye from wastewater.Therefore, in this work, a new g-C 3 N 4 based CeO 2 nanocomposites was designed for the effective removal of Rose Bengal from water.The structural, morphological and surface studies of the synthesised CeO 2 /Fe 3 O 4 /g-C 3 N 4 (CFC) nano-adsorbents were evaluated.Further, adsorption isotherms and kinetics were observed and investigated via adsorption of Rose Bengal dye.Maximum adsorption capacity and adsorption efficiency were calculated, which showed that the prepared nanocomposites are excellent photocatalytic adsorbent.A plausible adsorption mechanism has been provided based on its adsorption properties.Hence, the designed nano-adsorbent could be utilised as an emerging material for the removal of Rose Bengal dye from wastewater.

Synthesis of pristine samples of CeO 2 , Fe 3 O 4, and g-C 3 N 4
Hydrothermal synthesis was carried out for pristine CeO 2 and Fe 3 O 4 synthesis.25 mmol of CeCl 3 .7H 2 O was transferred in deionised water and kept for vigorous stirring.After 20 min of stirring, solution of NaOH was added in a dropwise manner and the mixture was keptunder stirring for 5 min after completion of NaOH addition.Then, the whole combination was transferred into the Teflon autoclave, further, which was placed in an oven at 180 ̊C for 20 h.A precipitate of white colour was obtained, which was then centrifuged (2000 rpm, 5 min), washed with ethyl alcohol multiple times, collected and allowed to dry overnight.The dried sample was then placed in a muffle furnace for calcination at 220 ̊C for three hours to get pristine CeO 2.
Furthermore, the pristine Fe 3 O 4 was synthesised using the same technique.38.5 mmol of FeCl 3 and 25 mmol of FeSO 4 .7H 2 O were added to deionised water placed in a magnetic stirrer, and then dropwise addition of NaOH was proceeded.The whole combination was kept in the same vigorous stirring for 20 h, and thereafter, the mixture was shifted to a Teflon autoclave and was placed in an oven at 180°C for 20 h.The brown colour precipitate was obtained, centrifuged (2000rpm, 5 min.),washed with deionised water and ethyl alcohol multiple times, collected and allowed to dry for 24 h at room temperature.The material obtained after drying was kept in a muffle furnace at 200 ̊ o C for 3 h to get pristine Fe 3 O 4 nanoparticles.The pristine g-C 3 N 4 powder was synthesised simply by heating the urea at 555°C for four hours in an air-tight condition.A yellowish-white residue is obtained and then ground in a mortar and pestle to get a pristine g-C 3 N 4 powder.

Synthesis of CeO 2 /Fe 3 O 4
38.5 mmol of FeCl 3 and 25 mmol of FeSO 4 .7H 2 O were transferred in deionised water and kept for mechanical agitation.After 20 minutes of passing, a dropwise addition of NaOH was proceeded.The combination was kept under the stirring condition for 20 h, and after that, 25 mmol of CeCl 3 .7H 2 O was transferred into it followed by a dropwise addition of NaOH solution.The resulting combination was then shifted into an autoclave, which was kept in an oven at 180°C for 20 h.A brownish precipitate was obtained which was further passed through multiple washing with ethyl alcohol, collected and dried.The dried sample was placed in a muffle furnace for calcination at 200°C for 3 h.So, the final obtained brown colour sample was considered as CeO 2 /Fe 3 O 4 and labelled as CF.

Synthesis of CeO 2 /Fe 3 O 4 /g-C 3 N 4
As freshly prepared, each CeO 2 /Fe 3 O 4 and g-C 3 N 4 powder were taken in 40 ml of deionised water in separate beakers, and both the mixtures were passed through the ultra-sonication for 10 min (two rounds of 5 min each).After ultra-sonication, both the mixtures were mixed.The resulting mix containing CeO 2 /Fe 3 O 4 and g-C 3 N 4 powder was shifted to ultra-sonication for 20 min (two rounds of 10 min each) to get the final nanocomposite, CeO 2 /Fe 3 O 4 /g-C 3 N 4 .The final product was collected, dried and labelled as CFC.The schematic diagram for the synthesis of CFC is shown in Figure S1.

Characterisation
The surface morphology, crystallinity, and composition of the synthesised nanocomposites were examined through numerous characterisation methods.PXRD patterns of CeO 2 /Fe 3 O 4 /g-C 3 N 4 , CeO 2 , Fe 3 O 4 and g-C 3 N 4 were recorded on a Phillips XPERT powder X-ray diffractometer using CuKa radiations (ʎ = 1.5418Å) within the 2Ɵ range of 20 ̊ to 80 ̊ at a voltage of 10Kv.The FTIR of the materials was carried out in KBr pellets on a Perkin-Elmer spectrum-2 spectrophotometer.A JEOL JEM 2100 (accelerating voltage 200 kV) instrument was used to obtain high-resolution transmission electron microscopy (HRTEM) and SAED images of CeO 2 /Fe 3 O 4 /g-C 3 N 4 .Energy-dispersive X-ray analysis (EDS) and elemental mapping of CeO 2 /Fe 3 O 4 /g-C 3 N 4 were performed using Bruker XFlash 6130.Brunauer-Emmett-Teller (BET) analysis of CeO 2 /Fe 3 O 4 /g-C 3 N 4 was performed on a Quanta Chrome Nova 2200 gas adsorption analyser (before performing BET, CeO 2 /Fe 3 O 4 /g-C 3 N 4 was degassed at 150 ̊C for four hours).To ascertain the surface area of CeO 2 /Fe 3 O 4 /g-C 3 N 4 , N 2 sorption-desorption isotherms were mapped at a bath temperature of 77.3 K.The XPS spectrum of CeO 2 /Fe 3 O 4 /g-C 3 N 4 was obtained using a PHI 5000 Versa Prob II spectrometer.All the absorption spectra of materials were obtained employing a GENESYS 10S UV-Visible spectrophotometer.

Evaluation of adsorption activity
Initially, the adsorption capacity and adsorption efficiency of the synthesised CeO 2 /Fe 3 O 4 /g-C 3 N 4 (CFC) nanoadsorbent with rose Bengal (RB) were determined.Typically, 10 mg of CeO 2 /Fe 3 O 4 /g-C 3 N 4 nanoadsorbent was transferred to 30 mL of each solution (concentration 5 to 50 mg/L) at neutral pH.Then, the mixture was transferred into a 100 mL beaker with a medium and constant stirring speed of 150-200 rpm at room temperature.Then, requisite quantity of suspension was collected from the mixture over a significant period of time, and the supernatant was collected using a centrifugal separator.A UV-Vis spectrophotometer was used to determine the final concentration of the adsorbates based on the wavelength of maximum absorption.Equations ( 1) and ( 2) may be used to calculate the adsorption capacity (q e ) and adsorption efficiency of adsorbent, respectively, where Co and Ce denote the initial and equilibrium concentrations of adsorbates, M (in g) represents the weight of the adsorbent, and V (in L) is the volume of the solution used in the experiment.The mass balance of the adsorbent in the system was used to determine its adsorption capacity.

Time, concentration and pH dependent analysis
All the adsorption experiments were carried out in batch approach.A 100-ppm stock solution was made using an adequate quantity of Rose Bengal dye in deionised water.The quantity of adsorbent was optimised using 50 mL of stock solutions.For kinetic studies, 30 mL of stock solution was added separately with adsorbent (0.010 g) at neutral pH and shaken 150 rpm.For concentration-dependent experiments, RB dye concentrations 2, 5, 10, 15, 20, 30, 40 and 50 ppm were used.Further studies were carried out at three different temperatures to investigate the influence of temperature on the adsorption process.
The pH of RB dye solution was modified to 2, 3, 4, 5, 6, 7, 8, 9 and 10 using a suitable proportion of either a NaOH or HCl solutions (0.1 M) to understand the influence of pH over removal efficiency of adsorbent.In 50 mL of 10 ppm RB dye solution, CFC (0.01 g) was added and shaken.After 24 hours, part of samples (20 mL) were collected, filtered and evaluated.

Desorption and recyclability of CFC
RB dye on the CFC surface was desorbed utilising NaOH (10 mL, 1 M) with the application of an orbital shaker, by shaking the sample mixture for 24 h at 100 rpm.The portion of sample mixture (20 mL) was collected, filtered and further utilised for the analysis to calculate the adsorption efficiency.The same process was repeated until the five cycles of utilisation of adsorbent.

XRD studies
The X-ray diffraction pattern (Figure 2a) of the nanocomposite revealed characteristic peaks of CeO 2 in the 2Ɵ range of 25° to 32° superimposed on the prominent peaks of    063) planes (JCPDS 89-6466).A characteristic peak of pristine g-C 3 N 4 (Figure 3d) could be seen at 26.50 ̊ (JCPDS 87-1526) corresponding to the (002) plane with some other peaks at 34.45֩ , 37.91֩ , 44.06֩ , 46.19֩ , 46.85֩ and 61.18֩ corresponding to (102), ( 110), ( 200), ( 201), ( 112) and ( 211) panes (JCPDS 87-1526).The presence of graphitic carbon nitride sheets in the final compound was further confirmed by TEM and SAED images.The presence of a significant amount of C and N in EDAX further bore evidence of graphitic carbon nitride sheets in the final nanocomposite.Scherrer's formula gave the average crystallite size of 6.1 nm for CFC.The crystallite size for pristine CeO 2 and Fe 3 O 4 lies in the range of 7.9 nm and 3.4 nm, respectively.

TEM and SAED analysis
The final nanocomposite (CFC) morphology was determined by Transmission Electron Spectroscopy (TEM).TEM micrographs (Figure 3a-c and 3e) display images of CeO 2 nanorods and Fe 3 O 4 nanoparticles distributed over the g-C 3 N 4 sheet.The shades, shapes and lattice fringes of the CeO 2 nanorods clearly separated them from the Fe 3 O 4 nanoparticles.The HRTEM picture (Figure 3d) indicates a clear heterojunction generated by the components g-C 3 N 4 , CeO 2 and Fe 3 O 4 .Additionally, HRTEM micrographs indicated that the average length and breadth of the CeO 2 nanorods are 51.97 nm and 7.98 nm, respectively.The aspect ratio of these nanorods was determined to be 6.5.The average diameter of the Fe 3 O 4 nanostructures was determined to be 11.32 nm.The spacing between the lattice fringes of CeO 2 nanorods was determined to be 0.320 nm, which corresponds to (111) planes.The gap of 0.332 nm between the bright fringes of g-C 3 N 4 may be attributed to the CeO 2 nanorod's (0 0 2) facet.The 0.260 nm interplanar spacing may also be related to the (023) plane of Fe 3 O 4 nanoparticles.SAED patterns (Figure 3f) exhibited concentric rings, indicating the polycrystalline structure of the sample.Numerous lattice planes corresponding to CeO 2 (2 2 0), (1 1 1) and Fe 3 O 4 (0 2 3), (0 4 1) and (1 4 2) planes were identified, as shown by a pre-recorded XRD pattern (JCPDS 81-0792 for CeO 2 and JCPDS 89-6466 for Fe 3 O 4 ).The inner and outermost rings might be related to the (0 0 2) and (2 1 1) planes of g-C 3 N 4 , which correspond to JCPDS 87-1526, respectively.

EDAX analysis
The elemental composition of the CeO 2 /Fe 3 O 4 /g-C 3 N 4 nanostructure was determined using EDAX analysis.In the obtained spectra, there was a distinct appearance of signals belonging to Ce, Fe, C, N and O (Figure 4).Around 6.5 keV, 0.2 keV, 0.3 keV and 0.4 keV, respectively, the peaks could be correlated with Fe, C, N and O, all of which match the K-series.At around 4.8 keV, a strong signal associated with the L-series of Ce can be observed.A supplementary table with the spectra illustrates the relative proportions of the various components.Thus, the EDAX data provided further evidence that the desired nanostructure was successfully synthesised.

XPS analysis
The XPS spectra of CeO 2 /Fe 3 O 4 /g-C 3 N 4 nanostructures were investigated further to understand each element's chemical interfacial bonding state, as shown in Figure 5.The XPS survey spectra (Figure 5a) indicated that elements C, O, Fe and Ce existed in the CeO 2 /Fe 3 O 4 /g-C 3 N 4 nanostructure.Figure 5(b) shows multiple overlapping peaks corresponding to cerium in the range of 880 eV to920 eV responsible for Ce +4 with some concentration of Ce +3 [54].Fine structured O1s spectra indicated different chemical shifts due to the multiple oxygen environments (Figure 5c).Broad O1s spectra have been deconvoluted with the Gaussian XPS peak fit leading to a superposition of three peaks located at 529.3 eV, 530.7 eV and 534.6 eV.The peak at 529.3 eV is characteristic of O 2− ions of the surface lattice oxygen in the α-Fe 2 O 3 matrix [55].Higher binding energy at the 530.7 eV peak attributed to oxygen defects in the matrix of metal oxides, related to oxygen vacancies [56,57].
Anionic vacancies change the net electronic charge density; this non-lattice oxygen peak has been attributed to surface O− ions with lower electron density.The peak at 534.6 eV could be associated with a chemisorbed oxygen peak [58].As shown in Figure 5 (d), peaks corresponding to 709.7 eV and 711.29 eV were identified as the 2p3/2 peaks for Fe3+ and Fe2+ ions, respectively.A satellite peak at 718.12 eV could be associated with a Fe 3+ ion.The broad peak at 724.06 eV corresponded to the superposed 2p1/2 peaks of Fe 3+ and Fe 2+ ions [59].The C 1s spectrum (Figure 5e) indicated that three main peaks located at 284.67 eV, 286 eV and 288.20 eV in the spectrum of the final composite could

BET analysis
Nitrogen adsorption-desorption isotherms were analysed to determine the surface area and pore structures of the synthesised nano-heterojunctions.The isotherm constitutes a typical type IV curve with H2 hysteresis loops (Figure 6a), characteristic of typical mesoporous materials with irregular pore structures.The synthesised materials' BET surface area and BJH pore size (Figure 6b) were calculated and found to be 50.74m 2 g −1 and 3.50 nm, respectively.The high surface area and mesoporous morphology are likely to benefit the adsorption performance.

Adsorption kinetic studies
The data were examined using both the pseudo-first and pseudo-second-order kinetic models to better understand the kinetics of RB adsorption onto CFC nanocomposite surfaces.The sorption kinetics were performed in batch mode with initial concentrations of 10 ppm RB.Equations ( 3) and ( 4) can be used to express the pseudo-first-order and pseudo-second-order kinetic models, In the above equations, q t (mg g −1 ) is the quantity of adsorbate (RB) adsorbed on the sorbent (CFC) surface at any time t, k 1 and k 2 are the first-order rate constant (min −1 ) and second-order rate constant (g mg −1 min −1 ), respectively, and t is the time (min) [61][62][63].Figure 7(a and b) show time-dependent tests of RB dye on CFC nanocomposite surfaces, which indicate that at a concentration of 10 ppm, 99% of the dye is removed within 120 min.The adsorption process followed the pseudo-second order kinetics with R 2 values near to unity.The actual and estimated values of maximum quantity adsorbed (q e ), as well as the rate constants k 1 and k 2 , were determined from the plots (Figure 7c,d) and are shown in Table 1.

Adsorption isotherms studies
Adsorption isotherms reveal adsorption capacity and adsorbate-adsorbent interaction.The adsorption process was described using two commonly used isotherm models, the Langmuir and Freundlich isotherms (Figure S2 (a and b).The Langmuir isotherm describes monolayer adsorption on homogeneous surfaces, whereas the Freundlich isotherm is frequently employed to explain multilayer adsorption on heterogeneous surfaces.Equations ( 5) and ( 6) define the linear form of the Langmuir and Freundlich isotherms, K L and q m denote the Langmuir adsorption constant (L mg −1 ) and maximum monolayer adsorption capacity (mg g −1 ).K F and n, are the Freundlich adsorption capacity and adsorption intensity [64][65][66][67].To display the adsorption isotherm plots (Figure S2 (a & b) relying on the observed data, various parameters were computed and are given in Table 2.Both adsorption models agree well for RB adsorption over the CFC surface, but the Langmuir model better fitted to the adsorption, as the model displays all the points aligning on the straight line with the higher R 2 value, which further suggests monolayer RB adsorption [68].The separation factor (R L ) was computed using the formula R L = 1/ (1+ K L ×C 0 ), where C 0 is the maximum starting concentration, and it was found to be 0< R L >1, indicating that the Langmuir adsorption process is favourable [69].Furthermore, the 1/n value of the Freundlich constant is very less than 1, suggesting good interaction between adsorbate and Table 1.Various kinetic models with their parameters for the RB adsorption over the CFC surface.

Thermodynamic parameters evaluation for the adsorption process of RB
The thermodynamic parameters, such as Gibb's free energy, entropy and enthalpy, are used to determine the spontaneity of a sorption process.The standard change in Gibb's free energy, which must be negative for considerable adsorption to take place, predicts the degree of spontaneity [72].Mathematically, Gibb's free energy is represented as in Equation (7), where R, T and K ad are gas constant, temperature during reaction and adsorption equilibrium constant respectively.Furthermore, K ad = C ad/ C e , with C ad (mg L −1 ) and C e (mg L −1 ) indicating the adsorbed solute concentration and concentration of solute in solution at equilibrium, respectively [73].However, to evaluate the thermodynamic parameters of adsorption, the Langmuir adsorption equilibrium constant (K L ) is utilised instead of the adsorption equilibrium constant (K ad ) [74].Moreover, the following is the connection amongst ΔG 0 , ΔH 0 and ΔS 0 of adsorption expressed in Equation ( 8), Correlating Equations ( 7) and ( 8), we get Equation (9), which is known as integrated Van't Hoff equation, The adsorption thermodynamic parameters were calculated by plotting lnK ad against 1/T (Figure S3) of Equation ( 9), and the slope and intercept of the linear plot were used to obtain the values of ΔH 0 and ΔS 0 .
The Van't Hoff plot for RB adsorption is shown in Figure 9, and the computed thermodynamic parameters are presented in Table 3.The process of adsorption is endothermic, as seen by the positive ΔH 0 value, and the numeric value below 40 KJmol −1 indicates physisorption [75].A positive value of ΔS 0 denotes greater randomisation of the system over the solid-liquid interface [76], whereas negative values of ΔG 0 at all temperatures denote the spontaneity and feasibility of the adsorption process [77].

pH effect and recyclability
The pH of zero-point charge (pH ZPC ) of CFC nanocomposites was evaluated and observed to be about 7, indicating that CFC possess a positively charged surface at a pH lower than that of pH ZPC .The pH above pH ZPC CFC surface is negatively charged [78].It is observed from Figure 8(a) that there was no any significant dip in the adsorption efficiency of CFC in the pH range of (2-6) due to better interaction between the positively charged surface of CFC and anionic Rose Bengal dye.However, there is a sudden fall in efficiency after pH 7, and it dipped up to 50% at higher pH, which could be due to repulsion between the negatively charged surface of CFC and anionic dye [79].Further adsorption experiments were conducted at pH 6 to justify the real ground water conditions.The adsorption-desorption of RB was examined in order to determine the CFC's recyclability.After adsorption, the dye was desorbed with 5 mL of ethanol, and the adsorbent was recovered and treated with ethanol and water until the dye colour was removed.Subsequently, the adsorbent was vacuum dried and re-used in the next experiment.CFC was recyclable up to 4 times and has an adsorption efficiency of more than 94% on the surface (Figure 8b).

Adsorption mechanism
There are several factors, like surface area of adsorbent, exchange of ions, dye and metal coordination, hydrogen bonding, among others, which could be major contributor for the dye adsorption mechanism.As suggested from Figure 9(a), the CFC surface is positively charged at pH less than 7.This positively charged surface of CFC may trigger electrostatic interactions and play a key role among other factors in anionic RB dye adsorption [80,81].A good surface area value (50.74 m 2 g −1 ) of CFC is also an important factor governing the significant effect over the adsorption mechanism.CFC shows high removal efficiency and good capacity for anionic RB dye, as RB contains -O, which may be accountable for hydrogen bonding with surface-capped water (confirmed by FTIR in Figure 10) over the CFC surface.Besides above factors, RB contains halogens capable of forming the coordinate bond with metals of CFC (Figure 10).
FTIR analyses of CFC before and after dye adsorption with pristine constituents of CFC were performed for the confirmation of between CFC and RB dye (Figure 10).It has been observed that the broad absorption band of pure g-C 3 N 4 from 3300 to 3000 cm −1 is attributed to the stretching vibration modes of N H bonds, resulting from the incomplete condensation of amino groups.The absorption peaks at 805 and 890 cm −1 are ascribed to the typical breathing mode of tri-s-triazine units and the deformation mode of N H bonds, respectively.A series of peaks observed in the range of 1650-1200 cm −1 indicates the presence of typical stretching modes of vibrations of heterocyclic CN [82].All the characteristic absorption peaks of g-C 3 N 4 appear in the CFC composite spectrum.
Therefore, the structure of g-C 3 N 4 remains intact after the growth of CeO 2 and Fe 3 O 4 nanoparticles on the g-C 3 N 4 surface.
Moreover, the FTIR of the Fe 3 O 4 spectrum shows water molecule stretching vibration bands at 3410 and 1629 cm −1 and -OH vibration bands at 893 cm −1 .The asymmetrical solid band at 578 cm −1 characteristics of Fe-O stretching vibrations includes the Fe-OH vibration band located at ~630 cm −1 [83].The absorption band at about 555 cm −1 present in both CeO 2 and CFC is representative of Ce-O stretching vibration [84].Thus, FTIR spectra confirm the grafting of Fe 3 O 4 and CeO 2 on the g-C 3 N 4 surface and perfect growth of CFC.A weak broad peak was observed after RB's adsorption on the CFC surface, suggesting the amount of H 2 O on the CFC surface.A peak shift was observed corresponding to the stretching frequency of the -NH 2 group and -OH group in the broad peak region after RB adsorption, suggesting the existence of H-bonding interactions between -OH and -NH groups on the nanocomposites with halo groups and -COONa functional groups of dyes.Peak shifts were observed from 1587 to 1581 cm −1 for RB corresponding to stretching frequency aromatic C = C groups, which suggest π-π interactions.RB shows high adsorption capacity because of halogen groups responsible for hydrogen bonding formation, which enhances the active sites.The other IR bands after RB adsorption show weak peaks less than that of pure CFC, which may be due to partial coverage of active adsorption sites of CFC.

Conclusion
In conclusion, we have designed an efficient and novel CFC nanocomposite via an easy set of hydrothermal process, which has shown its specific characteristics of working as an adsorbent and as a photocatalyst as well.The structural and the morphological properties of CFC have been thoroughly characterised employing various characterisation techniques.The maximum adsorption capacity of the CFC was calculated to be 83.05mg g −1 .CFC followed the pseudo-second-order kinetics and Langmuir isotherm model.All the thermodynamic parameters were studied and concluded that the process of adsorption is spontaneous and feasible.The adsorption mechanism of the process is established with the help of pH and FTIR analysis and suggested that metal coordination, hydrogen bonding, π-π interaction and electrostatic interactions are the major responsible factors for the adsorption phenomenon.The recyclability and the reusability of the CFC were also observed, and it was found that CFC has shown up to 94% of the adsorption efficiency for the fourth cycle.Overall, the ease of the synthetic process, considerable adsorption capacity and high adsorption and reusability provide extraordinary opportunities for the researchers working in the field of environmental remediation.

Figure 6 .
Figure 6.(a) The nitrogen adsorption-desorption isotherm curve of CeO 2 /Fe 3 O 4 /g-C 3 N 4 nanocomposite and (b) particle size distribution derived from desorption curve according to BJH model.

Figure 7 .
Figure 7. (a) display UV-vis absorbance spectra suggesting decrease in absorbance of RB dye with the increase of time; (b) Time profile of RB adsorption at neutral pH and 293 K temperature; (c) Pseudo first and, (d) Pseudo second order kinetics plots for RB dye adsorption.

Figure 8 .
Figure 8.(a) pH effect over adsorption efficiency of CFC for RB removal (b) CFC recyclability for RB dye.

Figure 9 .
Figure 9. Plausible adsorption mechanism for adsorption of rose Bengal dye on the surface of CeO 2 /Fe 3 O 4 /g-C 3 N 4 adsorbent.

Figure 10 .
Figure 10.FTIR spectra of CFC before and after RB adsorption with their pristine constituents.

Table 2 .
Various isotherm parameters of the models, Langmuir and Freundlich at different increasing temperaturea with respect to room temperature for adsorption of RB dye on the CFC nanoadsorbent surface.Langmuir maximum adsorption capacity of the adsorbent was found to be 83.05mgg −1 at 333 K.It is worth mentioning that the adsorption of RB was carried out with a 10 ppm solution at pH 6 to replicate actual polluted waste water circumstances, and the findings demonstrated a high removal efficiency close to 99% of the initial concentration, which is unmatched by other known adsorbents.

Table 3 .
Various thermodynamic parameters for RB adsorption at varying temperatures.