Synthesis of surfactant-coated cobalt ferrite nanoparticles for adsorptive removal of acid blue 45 dye

Nanotechnology has introduced various fascinating applications of nanoscale materials that have been used in bulk form during last several decades. Among these materials, nanoferrites have gained a great deal of intention due to their utilization in ferrofluids as adsorbent of toxic dyes for water purification, high density magnetic storage devices, targeted drug delivery, tumor treatment, contrast agents in magnetic resonance imaging, magnetic separation of biological molecules like proteins and enzymes, tissue repairing and cell separation [1–5]. Cobalt ferrite nanoparticles (CFNPs) are distinguished hard magnetic nanomaterials owing to their large anisotropic properties as compared to other ferrites [6]. The cobalt ferrites (CoFe₂O₄) have inverse spinel structure with cubic closed-packed oxygen lattice in which trivalent ions (Fe⁺³) stay at A (called tetrahedral) and B (called octahedral) sites, while divalent ions (Co⁺²) occupy only B sites. The electrical and magnetic properties of CFNPs at atomic level strongly depend on the size, purity, morphology and applied synthesis techniques [7, 8]. The CFNPs have wide range of applications owing to their high monodispersity and homogeneity. Several preparation techniques like combustion [9–13], microwave sintering [14], hydrothermal [15, 16], spray-spin heating coating [17], sonochemical [18], and modified citrate-gel [19] methods have been used to assemble the CFNPs with desired properties.

The decrease in particle size of magnetic CFNPs improves their superparamagnetism and fosters their applications in cell labelling and separation, drug delivery and cell thawing agents [20–23]. To control the surface properties, particle size and stability of CFNPs, various surfactants have been used. From the reported literature, it is supposed that the addition of surfactant is a best way to enhance monodispersity and control the size of NPs. The effect of polyethylene glycol as surfactant on crystallinity and magnetic characteristics of CFNPs has been studied in details [24]. The CFNPs with an average size of ∼8 nm using sodium citrate as surfactant has also been synthesized [25]. The cubic shape CFNPs under the effect of Bis-(2-hydroxy-1-naph- thaldehyde)-butanediamine Schiff-base ligand (L) as capping agent have also been synthesized. However, the spherical geometry of nanoparticles was non-uniform [26]. The oleic acid assisted synthesis of CFNPs to make stable colloidal suspensions in non-polar organic medium has also been reported [27]. The surfactants like ricinoleic acid and oleic acid induce the changes on surface of CoFe₂O₄ to impede the ferrite NPs growth [28]. It has been evidenced that the clustering of ferrite nanoparticles results in loss of their unique features and poor control over the inverse spinel composition of ferrites.

The T-1x, a polyglycol ether, is a non-ionic surfactant and can be used for extensive biomedical applications [29]. Whereas, the DDAB is a cationic water-soluble surfactant and increases the lipids structural order [30, 31]. In current work, we studied the inhibition effect of T-1x and DDAB as capping agents on secondary growth of CoFe₂O₄ nanoparticles to control the coalescence behaviour of magnetic nanoparticles. The CFNPs/T-1x and CFNPs/DDAB show good adsorptive removal of acid blue 45 dye as compared to uncoated CFNPs. The acid blue 45 dye is extensively used in carpet dyeing industry in various countries and it causes many hazardous effects to the environment. Consequently, the assembled materials can find significant applications in removal of dyes and toxic metals from several media.

Iron (III) chloride hexahydrate (99%), Cobalt (II) chloride hexahydrate (99%), sodium hydroxide (99%), T-1x (99%) and DDAB (99%) were used as received from Sigma Aldrich. Acid blue 45 dye was purchased from Clariant (PVT) Ltd, Lahore, Pakistan. Magnetic CoFe₂O₄ nanoparticles were synthesized using a modified capping agents-assisted coprecipitation method in which the chlorides of Fe (III) and Co (II) were used as reactants. In brief, 30 ml of 0.4 M ferric chloride (FeCl₃.6H₂O) and 30 ml of 0.2 M cobalt chloride (CoCl₂.6H₂O) were added into an inert atmosphere of three neck flask and stirred continuously on magnetic hot plate at room temperature. The 3 M potassium hydroxide (KOH) solution and precipitating agent were added to solution till pH of ∼11 was achieved. To observe the effect of capping agents, 3 drops of 0.05 M T-1x or DDAB were added to the solution and temperature was increased and maintained at 90 °C for 80 min. The resultant black residue was washed multiple times with deionized water and ethanol to make it free from chloride and sodium ions. The resultant sample was dried in oven at 60 °C for overnight. The synthesized products were named as CFNPs (without capping agents), CFNPs/T-1x (coated with T-1x) and CFNPs/DDAB (coated with DDAB).

The adsorption tests for acid blue 45 dye were performed by mixing 0.1 g of adsorbent in 50 ml (50 mg L⁻¹) of dye solution. The effect of agitation time (10–80 min), pH (2–12), temperature (10–80 °C), adsorbent dose (0.02–0.16 g) and dye concentration (10–100 mg L⁻¹) on removal of dye was investigated. The samples were subjected to filter to remove the adsorbent. The dye adsorption was confirmed spectrophotometrically using spectronic-20 with wavelength for maximum absorbance (λ_max). The percentage adsorption of dye was determined with the help of following relation [32]:

$$\begin{eqnarray} &&\% \,{\rm{age}}\,{\rm{adsorption}}=\displaystyle \frac{Co-Ce}{Co}\times 100\end{eqnarray} \tag{ 1 }$$

Whereas, Co is initial dye concentration (mg/L) and Ce is dye concentration at equilibrium.

To confirm the inverse spinel structure of synthesized CFNPs, the x-ray diffraction analysis was performed using Philips analytical x-ray diffractometer with CuKα radiation (λ = 0.154 nm) and Ni as filter. The diffraction images were recorded in terms of 2θ between 10 and 80°. The average crystallite size was determined from high intensity diffraction plane (311) with the help of Debye Sherrer formula (D = kλ/βcosθ, where k is the shape factor (0.9), β is full width at half maximum of the (311) reflection plane in radian, λ is the x-ray wavelength and θ is the Bragg's angle) [33]. The IR frequencies of the CFNPs for M-O bond identification were measured using Shimadzu ATR-FTIR-8400 s spectrophotometer in wavenumber range 400–4000 cm⁻¹. The magnetic response of all the prepared samples (CFNPs, CFNPs/T-1x, and CFNPs/DDAB) was evaluated using Lakeshore (model 7404) vibrating sample magnetometer (USA) at 300 K under maximum applied variable magnetic field 10 kOe. The microscopic morphology, hydrodynamic diameter and polydispersity were investigated by scanning electron microscopy (Hitachi, TM-1000) and atomic force microscopy (Park, XE-15). Double beam spectrophotometer (UV-6100) was used to determine the % age removal of acid blue 45 dye using calibration curve.

To analyze the characteristic bonds in the prepared samples (CFNPs, CFNPs/T-1x and CFNPs/DDAB), attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy was carried out in wavenumber range 400–4000 cm⁻¹. The absorption bands in wavenumber region 400–1000 cm⁻¹ are associated with lattice vibrations of metal ions in solids especially in ferrites as reported in literature [34]. Figure 1 shows the two characteristics absorption bands at ∼603 cm⁻¹ and ∼495 cm⁻¹ which are the intrinsic stretching vibrations of the metal oxygen (M-O) bond at tetrahedral (T_h) and octahedral (O_h) sites. The change in area and position of the bands are attributed to the addition of surfactants (T-1x and DDAB) [35, 36]. No absorption peak is observed in wavenumber range 1000–4000 cm⁻¹ which evidences the high-level purity of the assembled materials.

Zoom In Zoom Out Reset image size

X-ray diffraction patterns of CFNPs CFNPs/T-1x and CFNPs/DDAB are given in figure 2. The strong diffraction peaks (111), (220), (311), (222), (400), (422), (511) and (440) are consistent with the JCPDS Card No. 22–1086 which suggests the crystalline and inverse spinel phase structure of native CFNPs [37, 38]. The average crystallite size was calculated from diffraction plane (311) and its values for CFNPs, CFNPs/T-1x and CFNPs/DDAB were 37 nm, 21 nm and 13 nm, respectively. As the capping agents decrease the agglomeration of nanoparticles, consequently, these are responsible for decrease in grain size [39].

Zoom In Zoom Out Reset image size

To examine the influence of capping agents on microscopic morphology and topography of CFNPs, CFNPs/T-1x and CFNPs/DDAB, atomic force microscopy (AFM) and scanning electron microscopy (SEM) were employed. Figures 3(a) and (d) show the elongated and oval shape morphology of the CFNPs with a slight coalescence which is attributed to the prominence of dipolar-dipolar magnetic interactions and van der Waals forces between the magnetic nanoparticles [40, 41]. However, the agglomeration was decreased for CFNPs/T-1x and CFNPs/DDAB which is ascribed to the addition of capping agents. The average particle size of CFNPs, CFNPs/T-1x and CFNPs/DDAB measured from AFM and SEM images was ∼43 nm, ∼26 nm and ∼15 nm, respectively.

Zoom In Zoom Out Reset image size

The AFM and SEM results shown in figure 3 clearly assert the addition of capping agents in CFNPs. It is obvious that due to addition of capping agents, the distribution of NPs is enhanced. The chemical nature of capping agents has profound influence on changing the particle size. The nucleation growth model attributes this particle distribution enhancement to stabilization properties and steric hindrance of the capping agents. During the coprecipitation synthesis at pH ∼11–12 (alkaline), the precipitation of CFNPs is achieved which is acknowledged as Ostwald's ripening (aggregation and coarsening) [42]. After the precipitation, the NPs own highly negative charge on the surface owing to extensive hydroxyl groups at alkaline pH. Consequently, the cationic DDAB is adsorbed on the surface of NPs and resist their aggregation. The neutral T-1x with hydroxyl group and ether linkage, also tends to adsorb on the surface of NPs which further results in production of steric effects and enhancement of particle distribution.

The magnetic character of CFNPs, CFNPs/DDAB and CFNPs/T-1x was studied by vibrating sample magnetometry (VSM) analysis and results have been displayed in figure 4. The characteristic hysteresis loop formation is persistent in all the prepared samples which reflects the typical ferromagnetic response of the samples. The saturation magnetization (M_s) and intrinsic coercivity (H_c) results have been tabulated in table 1. The values of M_s and H_c measured at room temperature under an applied magnetic field of ±10 kOe are lower for CFNPs/DDAB and CFNPs/T-1x as compared to naked CFNPs. The decrease in M_s and H_c can be ascribed to magnetically dead layer of surfactants on the surface of CFNPs. The surfactant layers influence the interaction between core and surface spins of particle-particle association and constrain the magnetic moment alignment under the applied magnetic field [43]. Thus, the addition of capping agents in CFNPs has pronounced effect on particle size, surface anisotropy and exchange coupling energies of CFNPs.

Zoom In Zoom Out Reset image size

The adsorptive properties of CFNPs, CFNPs/T-1x and CFNPs/DDAB for acid blue 45 dye were investigated. The influence of adsorbent dose on adsorption of acid blue 45 dye was studied at 25 °C by changing the dose from 0.02 g to 0.16 g; the initial dye concentration was 50 ppm and volume of the dye solution was 100 ml. The adsorption of dye was increased with increasing concentration of adsorbents (CFNPs, CFNPs/T-1x and CFNPs/DDAB). The removal of dye was 78.95%, 89.22% and 95.52% for adsorbents CFNPs, CFNPs/T-1x and CFNPs/DDAB (each with dose 0.16 g), respectively, as given in the figure 5(a). The increasing adsorption is attributed to the introduction of frequent binding sites provided by adsorbent, which is consistent with the reported literature [44, 45].

Zoom In Zoom Out Reset image size

The effect of dye concentration was also studied for constant dose of adsorbent (0.1 g). As the dye concentration is increased from 10 to 100 mg L⁻¹, the dye removal is decreased from 81.18% to 72.44%, 93.22% to 84.72% and 95.71% to 89.41% for CFNPs, CFNPs/T-1x and CFNPs/DDAB, respectively, as shown in figure 5(b). The dye removal efficiency greatly depends on initial dye concentration. At low concentration of dye, there are extensive vacant binding positions on the surface of adsorbent. However, when initial concentration of dye is increased, the binding sites for the adsorption of dye molecules are reduced which results in decrease in dye removal efficiency. The increase in initial concentration of dye causes an increase in the loading capacity of the adsorbent owing to higher driving force for mass transfer [46–50]. The greater adsorption of dye by CFNPs/DDAB is due to cationic layer of DDAB over ferrite NPs which provides maximum binding sites for anionic acid blue 45 dye. Whereas, the adsorption of dye over CFNPs/T-1x is owing to porous surface and van der Waals interactions. The naked CFNPs have less affinity for the removal of acid blue 45 dye.

The dye removal is also increased with increasing agitation time, as shown in figure 5(c). In first ∼30 min, the rate of dye adsorption was very fast due to maximum available vacant sites. However, subsequently, the rate was decreased gradually and maximum adsorption of 71.33%, 82.26% and 98.15% was achieved in 60 min for CFNPs, CFNPs/T-1x and CFNPs/DDAB, respectively. During adsorption, the dye molecules get adsorbed on outer surface of CFNPs and may transfer to the pores and internal structures of the nanoparticles. As adsorbent has limited capacity of adsorption, the rate of adsorption becomes constant and further did not depend on agitation time. Therefore, the rate of dye removal from aqueous medium depends on rate of transfer of the adsorbate from outer to inner sites of the adsorbent [51].

The effect of pH on removal of dye was also examined and an aqueous solution of dye (50 mg L⁻¹) was treated with 0.1 g dose of adsorbent with pH from 1 to 10. The pH of the solution was adjusted with the help of 0.1 M HCl and 0.1 M NaOH solution. As synthesized CFNPs showed a maximum of 86.14% dye adsorption at pH 1 which was gradually decreased to 59.43% at pH 10, as shown in figure 5(d). In case of CFNPs/T-1x, a maximum dye adsorption of 94.20% was achieved at pH 1 and it was decreased to 79.89% at basic pH 10. Whereas for CFNPs/DDAB, a maximum dye adsorption of 99.78% was achieved at pH 1 and it was decreased to 88.52% at pH 6. However, a slight increase in dye adsorption was observed in pH range 6 to 8. Afterward, the removal of dye was further decreased from 90.07% to 80.96% with decrease in pH from 8 to 10.

The results show that the low pH is an effective factor for the adsorption of acid blue 45 dye. It plays a significant role especially for the adsorption of coloured compounds. It controls the quantity of electrostatic charges imparted by ionized dye compounds [52]. Generally, the rate of dye adsorption decreases for cationic dye and increases for anionic dye. Consequently, high pH is better for cationic dye adsorption and low pH for anionic dye adsorption [49]. At high pH, the positive charge is decreased at the solution interface and adsorbent surface becomes negatively charged [53]. As a result, the cationic dye adsorption is increased and anionic dye adsorption is decreased. While at low pH, the positive charge is increased on the solution interface and adsorbent surface becomes positively charged. Consequently, the cationic dye adsorption is decreased and anionic dye adsorption is increased [49]. Similar results have been reported in biosorption studies. At high acidic pH, negative charge is appeared on surface of adsorbing materials. Apart from this, a lower pH effects the physiochemical process and hydrolysis of the dyes [54].

The temperature has a significant influence on rate of dye adsorption. The rate of dye adsorption is increased with increase in temperature. A maximum dye adsorption of 74.27%, 88.73% and 94.30% was obtained at 60 °C for CFNPs, CFNPs/T-1x and CFNPs/DDAB, respectively, as shown in figure 5(e). The increase in temperature enhances the flow of dye ions and creates the swelling effect in the inner structuer of the adsorbent during the removal of dye, which further increases the movement of dye molecules. Consequently, the rate of dye removal is also associated with adssorbate-adsorbent interactions [55–57].

The state of equilibrium during adsorption is the measure of dye distribution between liquid phase and adsorbent. It determines the capacity of adsorbent to adsorb dye from the aqueous solution and can be expressed by common adsorption theories like Freundlich and Langmuir adsorption models [58]. Langmuir adsorption isotherm supports the monolayer adsorption of adsorbate molecules at the surface of adsorbent. Langmuir model is a quantitative form of adsorption. The equilibrium data received for varying and constant adsorbate concentrations was added to Langmuir equation and the outcomes have been displayed in figure 6(a) and table S1 of supplementary material which is available online at stacks.iop.org/MRX/5/035058/mmedia. The Langmuir isotherm equation is expressed as follows [59]:

$$\begin{eqnarray}&&\displaystyle \frac{1}{{{\boldsymbol{q}}}_{{\boldsymbol{e}}}}=\displaystyle \frac{1}{b}\times {{\boldsymbol{Q}}}_{{\bf{\max }}}\times \displaystyle \frac{1}{{\boldsymbol{Ce}}}+\displaystyle \frac{1}{{Q}_{\max }}\end{eqnarray} \tag{ 2 }$$

Zoom In Zoom Out Reset image size

Whereas, b (is equal to '(1/slope) × Q_max') is empirical constant which indicates the affinity of sorbent towards the sorbate, Q_max (is equal to '1/intercept') is the maximum amount of dye that can be adsorbed per unit dry weight of sorbent, C_e is the equilibrium concentration of dye (ppm), and q_e is the amount of dye adsorbed (mg/g) at equilibrium. Following expressions were used to measure the value of Langmuir's equilibrium constant (K_L and R_L) [60]:

$$\begin{eqnarray}&&{K}_{L}={Q}_{\max }.\,b\end{eqnarray} \tag{ 3 }$$

$$\begin{eqnarray}&&{R}_{L}=\displaystyle \frac{1}{b}+{K}_{L}.{C}_{o}\end{eqnarray} \tag{ 4 }$$

Where K_L is Langmuir's equilibrium constant and it informs about affinity of binding sites. The value of R_L is calculated as 0.0079 for CFNPs, 0.0081 for CFNPs/T-1x and 0.0085 for CFNPs/DDAB. The literature suggests that the adsorption will be favourable if R_L has value between 0 and 1 [58].

The equilibrium data received for varying and constant concentrations of adsorbent was applied to the following Freundlich equation [59]:

$$\begin{eqnarray}&&\mathrm{log}\,{q}_{e}=\,\mathrm{log}\,{K}_{f}+\displaystyle \frac{1}{n}\,\mathrm{log}\,{C}_{e}\end{eqnarray} \tag{ 5 }$$

Where C_e is the equilibrium concentration of dye, m is the mass of adsorbent used, and K_f and n are the constants for adsorption capacity and intensity of adsorption, respectively. The values of K_f and n are received from the intercept and slope of the graph plotted between logq_e versus logC_e, as shown in figures 6(d)–(f). The values of n for CFNPs, CFNPs/T-1x and CFNPs/DDAB lie between 0 and 1 which shows that the adsorption is favourable as per Freundlich isotherm model. The results are consistent with the previous reports [59].

To study the nature of adsorption that either it is chemisorption or physiosorption, following Dubinin-Radkohvisch (DR) equation was used [61, 62]:

$$\begin{eqnarray}&&\mathrm{ln}\,q=\,\mathrm{ln}\,{q}_{\max }\mbox{--}\beta {\varepsilon }^{2}\end{eqnarray} \tag{ 6 }$$

Where q is maximum sorption capacity, q_max is theoretical saturation capacity, β is adsorption per mole of adsorbate, and ε is polanyi potential which is equal to 'RT(1 + 1/C_e)'. For sorption energy, following equation was used [62]:

$$\begin{eqnarray}&&E=-\displaystyle \frac{1}{{(2\beta )}^{\tfrac{1}{2}}}\end{eqnarray} \tag{ 7 }$$

The value of sorption energy (E) was 2.208 KJ/mole for CFNPs, 2.219 KJ/mole for CFNPs/T-1x and 3.162 KJ/mole for CFNPs/DDAB. Such a small amount of released energy shows the less adsorption capacity of pure CFNPs. All values are less than 8 kJ/mole which indicates that the nature of adsorption was physiosorption and there was just a physical interaction between adsorbate and adsorbent [63].

Cationic (DDAB) and neutral (T-1x) capping agents have pronounced effect on particle size, morphology and monodispersity of CFNPs. The cationic surfactants are more feasible to reduce the particle size (CFNPs/DDAB, ∼13 nm) as compared to neutral one (CFNPsT-1x, ∼21 nm). It is fascinating that the addition of capping agents has no prominent influence on ferromagnetic behaviour of CFNPs at room temperature. However, a small decrease in saturation magnetization and coercivity with addition of capping agents is observed which is ascribed to decrease in particle size of surfactant coated CFNPs. The CFNPs, CFNPs/T-1x and CFNPs/DDAB materials show efficient adsorption of acid blue 45 dye. The applied adsorption models fit well and indicate the physical and endothermic nature of adsorption.

This research work was funded by Higher Education Commission (HEC), Islamabad, Pakistan.