Abstract

In this article, the Methylene Blue dye adsorption is studied using NiFe₂O₄-Graphene Oxide composite as adsorbent. The NiFe₂O₄-graphene oxide (NiFe₂O₄-GO) nano-composite made by single step solvothermal approach. The characterization study revealed the abundance of functional group and nanomaterial features in prepared material. The MB adsorption is increased with rising adsorbent doses, pH, temperature and initial MB solution concentrations. Pseudo second order kinetic model got fitted by adsorption kinetics. As compared to BET, Temkin, Freundlich and Dubinin-Radushkevich model, Langmuir model is suitable for adsorption isotherm. Thermodynamic studies suggest adsorption process’s endothermic nature and spontaneity. The adsorption advances through π-π interaction, H-bonding and electrostatic attraction. Reusability study reveals the prepared adsorbent is a promising as well as cost effective sorbent for high efficiency and excellent renewability.

1. Introduction

The basic dye, Methylene Blue (MB) (tetra-methylthionine chloride) has been utilized in microbiologic as well as histologic staining. Its clinical use is done in methemoglobinemia treatment ¹. However, the toxic effect of MB had also been reported in a neonate ² and results in acute renal failure, hyperbilirubinemia, hemolytic anemia, parathyroid adenoma ³, Alzheimer’s disease ⁴ etc.

Many chemicals are defiant to processing of predictable treatment, like simple biological, chemical and physical water refinement strategies. Such issues draw the attention of researchers to develop novel carbonaceous nano adsorbent for advance water treatment ^{5, 6, 7} via developing and implementation of various approaches for the contaminants removal from waste water.

Vast applications for the graphene-family nanomaterials were reported. It has the similar graphite structure that consists of (a) few-layer graphene (b) reduced GO(rGO) with a few oxygen groups (c) graphene oxide(GO) containing hydroxyl, carbonyl, epoxide, as well as carboxyl functional groups on the surface and/or on the edges ⁸. The nanomaterials of graphene-family exhibit excellent properties like mechanical strength, electrical conductivity and for organic/inorganic contaminant’s high adsorption capacity ⁹. The usage of nanomaterials of graphene family for adsorption required a great energy deal from aqueous solutions for recovering or reusing them ¹⁰. To overcome such problems, spinel ferrites (MFe₂O₄, where M=Mn, Ni, Zn, Cu, Co, etc.) will be combined as interior materials in nanomaterials of graphene-family. The ferrites fascinated a great interest because of their extraordinary electrical, catalytic, and magnetic properties which have potential usefulness for varied practical applications ¹¹.

The ferrites in the interior of the graphene nano-sheets resist the agglomeration, whereas toxic substance leaching are restricted by the graphene and adsorption properties are enhanced through a particular chemical stability and large surface area. The graphene nano-sheets associated with magnetic ferrites (GNSF) provide additional advantages of easy recovery of the adsorbent after removing and reusing contaminants by utilizing an external magnetic field. This study critically highlighted the application of GNSFs as specific nano adsorbents for removal of MB from aqueous solution.

2. Materials and Methods

Graphite powder, H₃PO₄, H₂SO₄, ethylene glycol(EG), NaOH, Ni(NO₃)₂.6H₂O, Fe(NO₃)₃.9H₂O, EtOH, Methylene Blue (Molecular formula: C₁₆H₁₈Cl N₃S, CI Classification Number: 52015, Figure S1)(Mark chemical laboratory reagent Co. Ltd., Mumbai) chemicals of analytical grade were used without purification.

The main precursor GO was prepared by using modified Hummer process ¹², 1.5g Graphite flakes and 9.0g KMnO₄ added successfully into a mixture of conc. H₃PO₄ and conc. H₂SO₄ of ratio 1:9 by volume under unremitting stirring for 12hrs at 323K. The mixture was cooled and poured slowly below dynamic stirring into ice chilled beaker containing H₂O₂ (30%), the mixture was washed with 30% ethanol and HCl at 350°C.

Synthesized GO (0.9g) is exfoliated in 80mL of EG by ultrasonication for more than 3hrs. 1.5g of NaOH, 0.95g of Ni(NO₃)₂.6H₂O and 1.72g of Fe(NO₃)₃.9H₂O are dissolved at ambient temperature in solution of GO-EG. After careful stirring for 30mins, the solution is shifted to 100mL stainless-steel teflon-lined autoclave to preserved at 473K for 6hrs and then the mixture is cooled to normal temperature. A black precipitate obtained, washed, centrifuged many times using EtOH. Lastly, the materials extracted is dried at 333K in a vacuum oven ¹³.

Batch adsorption methodology followed for the dye removal experiment. The predetermined adsorbent amount is mixed with 50mL dye solution in 100mL conical flask and allowed to shaken in thermostated water bath shaker (ISO 9001:2008; SUPERIOR SCIENTIFIC INDUSTRIES) for specific period of time. At presets interval of time, centrifugation used for separation of solution from the mixture. The Visible spectrometer is used to measure adsorbed dye’s concentration in supernatant solution (Elico SL 177). Furthermore, mass balanced equation is used for calculating adsorption extent (%) and dye adsorbed per unit adsorbent mass(q).

(1)

(2)

where, q(mgg^-1) is the adsorbate amount adsorbed per mass(g) of adsorbent. C₀(mgL^-1) and C_e(mgL^-1) are dye’s initial and equilibrium concentration. V(L) is solution’s volume of experiment and m(g) is adsorbent’s mass.

The process of adsorption presented under solution temperature, pH, interaction time, adsorbent load, and initial dye concentrations at diverse experimental condition (Table 1). The adsorbent’s reusability investigation carried out subsequent to loaded adsorbent’s desorption for the adsorption capacity.

The investigation of powdered XRD pattern carried out in Bruker AXS(Germany) “X-ray powder diffractometer Model D8”, which focus on Cu Kα monochromatized radiation which has 0.02°(2θ) step size as well as 0.15418nm wavelength. The determination of NFO-GO composite’s pore volume, pore diameter and particular surface area were executed by Brunauer-Emmet-Teller (BET) N₂ gas approaches by utilizing automated gas sorption analyser (Quantachrome® ASiQwin™ Instrument, NOVA-1000 version 3.70). The analysis of morphology (FESEM images) and elemental study (EDX) is achieved using Carl Zeiss (Germany) SUPRA 55VP, Gemini Column with air lock system model. Laser Raman microscope (Labram HR Evolution, Horiba) is used to get the Raman spectra. The X-ray photoelectron spectroscopy (XPS) measurement is carried out in X-ray photoelectron spectrometer Multilab2000. The Zeta sizer, Malvern is used to measure zeta potentials and the charges nature on surface. The Differential scanning calorimetry (DSC) analysis and Thermogravimetric (TG) analysis are studied” out in DSC1 Star System Mettler Toledo model to study the thermal stability.

3. Result and Discussion

XRD patterns of NiFe₂O₄ (NFO) and NiFe₂O₄-Graphene oxide composite (NFO-GO) was published elsewhere ¹³ (Figure S2). NFO-GO represented wide peak at 25.1° (2θ) with the inter planar spacing of 0.36nm, might be due to (002) plane for GO. Crystals sizes of 11.48nm, 5.72nm, 6.99nm, 7.02nm and 12.84nm were obtained from Scherrer equation (D_p=0.94λ/(β_1/2Cosθ, where β_1/2= width at X-ray line broadening reflection’s half maximum, θ =diffraction angle, λ= Cu K_α radiation’s wavelength ¹⁴.

FESEM images (Figure 1A) clearly indicated the association of NFO particles in the GO sheet preventing GO sheet from agglomeration and make easy to peel off by ultrasonication. The figure indicated that GO sheets are separated largely by the introduction of NFO particles.

The EDX (Figure 1B) analysis confirmed the presence of carbon, oxygen, iron and nickel elements in NFO-GO nano-composites ¹⁵.

Raman spectra of NFO-GO (Figure 2) displayed noticeable peaks at 1598.55cm^-1 and 1341.08cm^-1 for two (G and D) bands. Assignment of G band is preferred for E_2g mode which observes sp²-carbon domain while structural disorders and defects are connected through D band which results in breaking selection rule as well as symmetry ¹⁶. The intensity ratio(I_D/I_G) utilized to measure the disorder ¹⁷ and found to be 1.002 showed an enhanced value indicating localized sp³ defects within the sp² carbon network indicated the successful oxidation of graphene. The Raman spectra also exhibits five Raman bands ¹⁸ typically for the inverse spinel structure of NiFe₂O₄. The group theoretical calculation shows five Raman band A_1g+E_g+3T_2g for NiFe₂O₄ crystal of space group Fd-3m ¹⁹.

The literature of XPS analysis revealed that GO consists of two main peak, hydrophobic π- conjugated sp² domain and hydrophilic oxygen containing functional groups with sp³ domain ²⁰. The full screen XPS spectrum (Figure 3) exhibited the presence of C, O, Fe and Ni elements at the binding energy 285.42eV, 532.07eV, 725.95eV and 856.78eV respectively. The Carbon/Oxygen content implied that a greater number of oxygen containing group successfully introduced into the prepared composites. A computational1multi peak resolution methods of C1s band deconvoluted into four peaks at 284.85eV, 285.75eV, 287.05eV and 289.01eV, resulting C=C/C-C, C-O, C=O and COOH groups respectively ²¹. FT-IR spectra (Figure 4) supported the report obtained in XPS analysis.

At high relative pressure, BET plots indicate type-I for wider hysteresis loop of isotherm by IUPAC classification ¹³ (Figure S3, published elsewhere), the isotherm’s feature discovered that with narrow slit pores high adsorption performance is presented by NFO-GO composite.

BET specific surface area of 88.027m²g^-1 with pore diameter 3.48nm and pore volume 0.131ccg^-1 indicate an impressive result as adsorbent whereas the reported GO’s specific surface area is 31.4m²g^{-1 22}.

At 5.0 to 9.0pH range Zeta potential’s measurement estimate no zero point charge of NFO-GO indicating the adsorbent’s negative surface charge (Figure 5).

Thermogravimetric analysis of NFO-GO presented that approximately 10% weight loss observed to 373K, might be due to water vapor’s loss. In 373K-800K temperature range 16% weight loss occurs that might be because of carbon-oxidation ²³. After oxidations, nearly ~74% weight is related to NFO-GO weight. In DSC curve an endothermic peak was observed at nearly 385K attributing to glass transition temperature (Figure 6).

The contact time effect for Methylene Blue adsorption on NFO-GO is achieved at various temperatures (303- 333K).

MB adsorption (303K) per unit mass of adsorbent is enhanced from 39.28mgg^-1 (10min) to 48.15mgg^-1 (240min) steadily (Figure S4). However, for NFO, the maximum value of q_e(303K) is 1.4mgg^-1(240min). At beginning, maximum adsorption has taken place and equilibrium attained by slowing down within 240 min. At initial periods high adsorption capacity might be because of enormous active sites that encourage the dye molecule in proficient adsorbent attachment. Adsorption equilibriums was attained due to saturation of binding site ^{24, 25}.

Negatively charged surface of adsorbent may compete for proton along with the dye molecules thus the adsorption process increases with increasing pH of the solution (Figure S5). The electrostatic attraction between cationic MB with NFO-GO composite increases with pH ²⁶. Therefore, the adsorbent’s negative surface attracts the dye molecule very easily that supported by Zeta potential values.

The adsorption of MB per unit adsorbent’s mass enhanced from 48.15mgg^-1 to 97.37mgg^-1 at 303K with the concentration of dye from 10mgL^-1 to 30mgL^-1 (Figure S6). The dye’s mass transfer resistance between the solid phase and the aqueous phase is overcome by providing higher driving force via enhancing the dye concentration which results more collision among adsorbent’s solid phase and dye molecule. At the higher dye concentration, adsorbent’s mass is exposed to thousands dye species and appropriate binding sites are gradually filled by the dye’s progressive high numbers. It results in q_e enhancement, however, there is declination in the net adsorption ²⁷.

There exists a reduction in adsorption capacity (q_e) when adsorbent amount is increased from 0.08gL^-1 to 0.36gL^-1. However, the extent of adsorption presented a gradual increase as adsorbents load increases (Figure S7).

The enhancement of extent of adsorption may be because of surface negative charge enhancement as well as electrostatic potential reduction near the solid surfaces that favours adsorbent-adsobate interaction. While, the greater adsorbent amount efficiently decreased the unsaturations of adsorption sites due to overcrowding and correspondingly such sites per unit mass decreases which results in reasonably low adsorption capacity(q_e) at higher dosage of adsorbent ^{28, 29}. Moreover, adsorption sites are easily accessed by the dye species in less adsorbent amount.

MB adsorption capacity is increased with rise in temperature from 303K to 333K indicating endothermic nature of adsorbent-dye interaction (Figure 7). Increase in temperature might result porosity swelling that enabled the adsorbate molecule to quick diffusion into NFO-GO composites pores as well as into the external boundary layer ^{30, 31}. The increase in adsorption capacity of MB is preferred by temperature rise and an activation energy provided more driving force in overcoming such energy barrier to get surface attachment ³².

Pseudo first order kinetic ³³ and pseudo second order kinetic ³⁴ is widely used in adsorption phenomenon as

(3)

(4)

where, q_e and q_t is MB adsorbed(mgg^-1) at equilibrium and at time t(min) respectively. The linearity of the curve1log (q_e - q_t) Vs t and t/q_t Vs t represents pseudo first1order and pseudo second1order kinetic models (Figure 8)1respectively. K₁(min^-1) and K₂ (gmg^-1min^-1) is1the first order and second order rate constant respectively.

The value of adsorption1capacity calculated (q_e,cal.) for pseudo first order model differed from experimental data (q_e,exp.) with a deviation of -69.94%. However, the deviation for pseudo second order model is relatively small (1.29%), suggesting the closeness of experimental value to theoretical value. The minute deviations observed in pseudo second order model might be because of experimental error. Table 2 summarizes the information attained at various temperatures from pseudo second order model.

The Elovich kinetic model ³⁵ applied to the adsorption phenomenon considering that (i) The adsorbent surface is energetically heterogeneous and (ii) At the low surface coverage adsorbed dyes neither interacted nor desorpted that can significantly affect the adsorption kinetic. The Elovich equation is

(5)

The parameter α and β are the initial rate and Elovich constant respectively. The calculated plot values for co-efficient correlation, α(adsorption co-efficient) and β(desorption co-efficient) at different experimental temperature are tabulated in Table 2.

In the process of solid/liquid sorption the solute1transfer may consist of either1the mass transfer step (film diffusion) or diffusion of intra-particle or both. The diffusion mechanism of dye removal from aqueous solution by adsorption is multi-step process ³⁶. For porous adsorbent, dye molecule’s diffusion into the adsorbent’s pores cannot be denied. Therefore, this model is also considered in finding out the reasonable adsorption kinetic model ³⁷. Weber and Morris simplified the intra-particle diffusion model ³⁸ and can be expressed as-

(6)

k_i (mgg^-1min^0.5) represented the rate of intra-particle diffusions. The data attained for rate constant (k_i) is 2.67mgg^-1min^0.5 at i-stage, The intercept(C) suggested the boundary layer thickness. The non-linearity of the q_t v_s t^0.5 curve indicated that diffusion of intra-particle may not have substantial part in the overall process of dye adsorption.

The q_t V_s t^0.5 plots (Figure 9) are multilinear consisting three linear segments. The 1^st segments with great slope due to MB transport to NFO-GO composite’s external surface from solution. The 2^ndsegment explains the steady adsorption stage corresponds to MB molecule’s diffusion into the adsorbent’s pore from the external surface (intra-particle diffusion). The 3^rd segment with small1slope describes the stage of final1equilibrium in which diffusion of intra-particle initiates to decrease.

The non-zero value of the intercept(C) for every linear segment (Table 2) indicating that1the intra-particle diffusion is not1the only controlling rate step in every stage in diffusion process ³⁹. Once the solute particles are loaded, the sorption process1is controlled by intra-particle1diffusion.

Boyd’s model is used for analyzing the experimental data ⁴⁰ to get insight into adsorption process’s actual rate controlling step.

(7)

Where, F is the equilibrium’s fractional attainment at various time (t) and B_t represented F’s mathematical function.

(8)

Thus, B_t value could be measured from equation (7). The linearity of the plot indicated consistent information that the adsorption rate is controlled by transfer of external mass (film diffusion) phenomenon. At initial stage, the plot of calculated B_t Vs t did not pass through the origin indicating adsorption rate controlled by film diffusion and subsequently switch to other mechanism like intra-particle diffusion.

Chemical species accumulation is allowed by the adsorption into the adsorbent’s solid phase as well as into their interfaces. Empirical isotherm models provide information related to experimental findings. BET, Dubinin-Radushkevich, Temkin, Freundlich, and Langmuir isotherm model are used to get insight the adsorption process.

The linear form of Freundlich ⁴¹ and Langmuir ⁴² equation can be expressed as-

(9)

(10)

Where, c_e(mgL^-1) =equilibrium concentration of MB solution, q_e(mgg^-1)=equilibrium adsorption capacity, q_m(mgg^-1)= Langmuir maximum adsorption capacity corresponds to total coverage of monolayer, b = Langmuir constant to binding site’s affinity as well as adsorption energy, k_f(mg^1-1/n L^1/ng^-1) =Freundlich constant related to capacity of adsorption, 1/n =factor of heterogeneity indicating adsorption process’s feasibility.

The linear relation between c_e/q_e Vs c_e plots (Figure 10) indicated the adsorption process obey Langmuir isotherm model. Table 3 listed the correlation coefficients and isotherm values. q_m and b increased with enhancement of temperature due to the endothermic nature of adsorption process ⁴³. Other important parameter R_L ⁴⁴, a dimensionless separation constant can be represented as-

(11)

Where, c_o represented maximum initial concentration of MB (mg/L). The separation constant specified that the isotherm is favourable for R_L<1, not favourable for R_L>1, linear for R_L=1 and irreversible for R_L=0 ^{45, 46}. The R_L value in between 0.007 to 0.02 indicating the adsorption process is favourable.

The examination of linear fitting curve of logq_e Vs logc_e for Freundlich isotherm (equation 9) showed that the correlation co-efficient (R²) value is much less than that of Langmuir isotherm curve (Table 3), although it increased with increasing temperature.

As compared to Freundlich isotherm model, Langmuir isotherm model fitted better with experimental value. The monolayer Langmuir adsorption capacity is measured as 97.94mgg^-1 for MB.

It assumed that heat of the molecules reduces with coverage due to adsorbent-adsorbate interaction. Temkin Isotherm model explained in linear form ⁴⁷ in adsorption as-

(12)

where, T= solution temperature(K), R(JK^-1mol^-1)= gas constant, A_T(g^-1)= equilibrium binding constant consistent to maximum binding energy, RT/b_T=B (Jmol^-1) represented Temkin constant corresponding to sorption. The linear fitting of q_eVs logc_e curve for MB monolayer adsorption on NFO-GO revealed the R²= 0.929.

The Dubunin-Rudushkevich (D-R) Isotherm play significant part for the porous adsorbent, characteristics with wide variety of pore shape and sizes ⁴⁸. The linear form of D-R isotherm and desorption energy equation is given by ⁴⁹-

(13)

(14)

Where, E(kJmol¹)= adsorbate’s energy per molecule for eliminating a molecule to infinity from its sorption site’s location, ε=Polanyi potential [ε=RTln(1+1/C_e)], K_dr(mol²/J²) = D-R isotherm constant linked with free energy, q_s(mgg^-1) = D-R theoretical saturation capacity. The parameter values (Table 3) from lnq_e Vs ε² linear fitting curve for D-R Isotherm(Figure S8) is q_s= 85.86mgg^-1 and E= 8.3kJmol^-1. It recommended the MB’s physical adsorption on adsorbent.

BET adsorption model is dependent on the statement that there are chances that onto the adsorbent surface adsorbate will be adsorbed causing multi layer formation in adsorbed dye’s random distribution ⁵⁰. It has been assumed that the condensation energy as well as adsorption energy are the reason for the successive layer’s adsorption as well as for the first monolayer respectively. The BET equation’s linear form ⁵¹ is-

(15)

Where, K_b= BET constant, q_m= adsorbate amount that forms total monolayer (mgg^-1), q_e represents adsorbate amount adsorbed onto adsorbent(mgg^-1), c_o represents adsorbate’s saturation concentration (mgL^-1), c_e represents adsorbate’s equilibrium concentration(mgL^-1) in solution. The plots of Vs c_e/c_o at various temperature are plotted (Figure S9) and calculated monolayer capacity are listed in Table 3. The parameters indicated the multilayer formation of MB-NFO-GO interaction.

The thermodynamics parameters, namely, Gibb’s free energy (ΔG), entropy(ΔS) and enthalpy (ΔH) for dye adsorption are measured by following equation ⁵².

(16)

(17)

where, R= ideal gas constant(8.314 JK^-1mol^-1), T= temperature (K), K_d(q_e/c_e)= distribution coefficient.

In order to calculate the adsorption’s activation energy Arrhenius equation applied as-

(18)

where, A= Arrhenius factor, E_a(kJmol^-1)= Arrhenius activation energy, and K₂(gmg^-1min^-1)= pseudo second order rate constant.

The slope and intercept from linear plot of lnK_dVs 1/T is used for calculating ΔH and ΔS values (Figure 11, Table 4). The positive value of ΔH recommended the endothermic adsorbate-adsorbent interaction. The rise of randomness is indicated by positive value of ΔS which might be due to the structural changes of adsorbent in adsorption process.

Gibb’s negative free energy (ΔG) decreased from -13.17kJmol^-1 to -21.11kJmol^-1 with rise in temperature (303K-333K). MB adsorption process’s spontaneity is more positive at high temperature ⁵³. Depending on ΔG value, MB’s adsorption onto NFO-GO thought to be involving physisorption ⁵⁰.

The Arrhenius parameter was determined by linear fitting curve of logK₂ Vs 1/T (equation 18) offered the -E_a/2.303R slope and logA intercept (Figure 12). For activation energy result attained is 33.458 kJmol^-1 which indicates the process of adsorption may be controlled physically ⁵⁴.

The pH influence for MB adsorption suggested that the electrostatic attraction could be the prime adsorption force. Although the process of adsorption observed to occur through a complicated adsorption mechanism where various interaction types might be involve in adsorption of MB onto NFO-GO composites. Investigation of FT-IR spectrums (Figure 4) of MB loaded NFO-GO and NFO-GO are carried out for verifying the adsorption process mechanism. The FT-IR spectra of MB loaded NFO-GO showed various changes. Adsorption band at 1696cm^-1 and 1735cm^-1 in NFO-GO composite conforming C=O and COOH slightly shifted to 1695cm^-1 and 1733cm^-1 respectively, indicating that the carboxylic group has significant role in the adsorption process. It might be because of the ionization of carboxylic group that made NFO-GO composite surfaces negatively charged. The adsorption peak at 1650cm^-1 and 3427cm^-1 belong to C=C and O-H shifted to 1677cm^-1 and 3415cm^-1 respectively, suggest that hydrogen bond and π=π interaction could contribute MB adsorption ^{55, 56}. The characteristic adsorption peaks for aromatic rings system at 834cm^-1, 790cm^-1 and 1547cm^-1 confirmed π-π interaction of the carbon atom of NFO-GO with B molecules which contain C=C double bond as well as benzene ring ⁵⁷. The C-N stretching frequency at 1350cm^-1 and 1321cm^-1 confirms the hydrogen bond existence between MB’s nitrogen and NFO- GO’s hydroxyl ⁵⁸.

The regeneration and reusability study is a significant factor for the adsorbent’s applicability and economy. The recyclability of NFO-GO composites is investigated performing five adsorption/desorption cycles (Figure 13). The desorption of adsorbed MB dye is done at every cycle ⁵⁹ from acetic/ethanolic solution after stirring for 12hrs at room temperature by magnetic stirrer. The adsorbent is reused for successive adsorption processes. It is noticed that adsorbent’s adsorption capacity reduced in each consecutive cycle from 97.9% to 80.4% after the 5^th cycle of the original adsorption capacity. After first cycle, the uptake capacity reduced down to 53% without desorption, that furthermore after the 5^th cycle reduced to original value’s 20.3%. The adsorption efficiency recovered effectively by ethanol/acetic acid solution desorption. The result indicated that NFO-GO composites can be a potential, efficient and cost effective adsorbent for MB removal because of attractive regeneration performance.

4. Conclusion

The Methylene Blue dye adsorption is studied using NiFe₂O₄-Graphene Oxide composite. The study summarized as follows -

• Successful preparation of adsorbent NFO-GO was done as well as it has been suggested by the characterization study that it was nano range composite.

• Langmuir isotherm obeyed during the dye adsorption process with 97.94mgg^-1 monolayer adsorption capacities.

• Second order kinetic model followed in the adsorption process.

• Nature of dye adsorptions is endothermic.

• There exists spontaneous interaction of NFO-GO-dye as well as controlled by Gibbs free energy reduction.

• pH sensitive dye adsorptions is observed.

Acknowledgements

We are thankful to NEHU, shilling, IITG Guwahati and USIC Gauhati University, IIAST for measuring TEM, XRD and Zeta potential respectively. We are also grateful to the department of chemistry, B N College, Dhubri for allowing to carry out the experimental work.

Supplementary File (Figure)

References