Highly efficient simultaneous ultrasonic-assisted adsorption of methylene blue and rhodamine B onto metal organic framework MIL-68(Al): central composite design optimization

Abstract

In the present work, metal organic framework MIL-68 (MIL = Material of Institute Lavoisier), with metal sites of aluminum ions (namely MIL-68(Al)), was synthesized by a simple, fast and low-cost process for simultaneous removal of Methylene Blue (MB) and Rhodamine B (RhB), regarded to be toxic and even carcinogenic, from aqueous solution. The adsorbent was characterized by FTIR, SEM, XRD, BET and TG analysis. Optimization of significant variables such as pH, adsorbent dosage, initial MB concentration, initial RhB concentration and adsorption time affecting both single and binary dye removal was developed by Central Composite Design (CCD) under response surface methodology (RSM), and the optimized values for these parameters in binary solutions were found to be 6.45, 0.014 g, 60 mg L⁻¹, 15 mg L⁻¹ and 9.9 min, respectively. The kinetics and isotherm of the adsorption were investigated in detail. Maximum sorption capacities of 1666, 1111, 227 and 29 mg g⁻¹ were obtained from the Langmuir isotherms for MB and RhB in single and binary solutions, respectively. The kinetics study results suggested that the sorption of the studied dyes onto MIL-68(Al) follows the pseudo-second order model. According to the reusability test of the adsorbent, dye uploaded MIL-68(Al) can be regenerated using methanol. The regenerated sorbent was used for three cycles use with high performance. The effect of ionic strength on the removal efficiency of the sorbent was also tested with different kinds of salts. The applicability of the sorbent was examined for real samples spiked with specific quantities of both dyes in water from tap water, sea water, spring water and laundry wastewater. The short time required for dye uptake makes this MOF a promising sorbent for simultaneous and rapid removal of dyes even in the real polluted environment.

---

Introduction

Organic dyes and pigments are considered as one of the most hazardous contaminants that are widely used in many industries such as textiles, cosmetics, food additives, dyeing, printing, electroplating, plastic, leather and related industries.^1–3

But, it is necessary to note that the contamination of drinking water by dyes even at low concentrations can make it harmful for human consumption.⁴ Methylene Blue (MB) as a cationic dye and also a redox indicator is used in coloring paper, cottons, rubber, wools and various tests of analytical chemistry.⁵ Rhodamine B (RhB) is also a basic red dye of the xanthene class. Both of mentioned dyes are very resistant to biodegradability and their elimination is difficult by conventional methods.⁶ The aromatic rings in most dyes particularly in azo dyes make them carcinogenic and mutagenic.⁷ Consequently, the removal of colored effluents from wastewaters is challenging subject to preserve human health and produce a safe environment.⁸ Wastewater can be treated using various physical, biological and chemical methods like electrochemical techniques, precipitation, membrane filtration, flocculation, coagulation, ozonation, enzymatic techniques and adsorption with their own advantages and disadvantages.^6,9,10 Adsorption process which is based on the transfer of pollutants from the solution to the solid phase is generally used as powerful and widely acceptable technique in wastewater treatment due to its unique advantages including high efficiency, simplicity and large scalability.¹¹

Recently, numerous porous materials have been investigated for adsorptive removal of organic compounds from wastewaters.^12–15 Among them, metal–organic frameworks (MOFs) known as porous coordination polymers, and also soft analogues of zeolites, are a new class of crystalline porous materials designing a structure with a particular pore size and shape from different transition metal ions (or clusters) and multifunctional ligands as organic linkers.¹⁶ Due to their diverse, porous structures, easy tenability of pore size and shape, high thermal and mechanical stability, high surface areas (up to 3000 m² g⁻¹) and ease of preparation, MOFs have been widely studied in the field of catalysis, gas adsorption/storage, drug delivery, separation, molecular transport, adsorption of organic and inorganic molecules, luminescence and electrode materials.^17–23 The adsorption abilities of MOFs have been largely unexplored although their adsorptive removal of hazardous materials from polluted aqueous environment is one of the most important areas of MOFs.²⁴ MIL-n materials, a class of MOFs, are composed of trivalent metal cations such as Al³⁺, Cr³⁺, V³⁺, In³⁺ or Ga³⁺ and carboxylic acid groups.²⁵ The three-dimensional networks of MIL-68(Al) contain two kinds of channels with an opening diameter of 6.0–6.4 Å and 16–17 Å exhibiting high surface area, suitable selectivity to trap a particular species and sufficient thermal stability.²⁶

Todays, ultrasound irradiation is regarded as motivation force in chemical process to accelerate mass transfer. This phenomenon is described by the acoustic cavitation with its growth and collapse of micrometrical bubbles. The increase in the rate of mass transfer is due to the acoustic streaming induced by sonic waves creates microscopic turbulence through the solid particles.²⁷

The conventional and classical one factor at a time approach is time consuming and non-feasible to give the true optimum conditions due to requiring high number of experiments and the lack of interactions among the factors. Statistical experimental design is proposed in order to reduce the process development time, overall costs and also decrease the number of experiments by consideration the interaction between variables on the basis of mathematical modeling.^28,29 Nowadays, central composite design (CCD) under response surface methodology (RSM) is widely employed as a statistical tool to evaluate the sole and combination interactions between the variables. A useful predicative model is then obtained for first or second order polynomial equations to the experimental responses in the experimental design followed by a variance analysis (ANOVA) of the model. In this way the real optimal conditions can be found with respect to the tridimensional graphs. Significance of process variable was analyzed by p-value and F-value parameters.²⁹

In this study, the as-prepared MOF material MIL-68(Al) was synthesized by facile solvothermal method and characterized using several instrumental techniques such as X-ray diffraction (XRD), scanning electron microscopy (SEM), Brunauer–Emmett–Teller (BET) surface area measurements and Fourier transform infrared (FT-IR) spectroscopy. In the next step, the adsorption behavior of MIL-68(Al) for removal of MB and RhB following the optimization of the effective parameters including pH, amount of adsorbent, sonication time and initial dye concentration on the removal efficiency of MB or RhB (R%) were investigated by central composite design (CCD) combined with response surface methodology (RSM). Langmuir, Freundlich and Temkin isotherm models were applied to experimental data for evaluation the adsorption capacity of MIL-68(Al). The adsorption rates were investigated by fitting the experimental data to conventional kinetic models such as Elovich, pseudo-first and pseudo-second order models. Moreover the reusability of the loaded sorbent was tested by various chemical reagents. Finally, the validity of the proposed method studied for different real samples.

Experimental

Materials

All chemicals including terephthalic acid, AlCl₃·6H₂O, H₃PO₄, HCl, NaCl, KCl, methanol, MB and RhB with the analytical grade are purchased from Merck (Darmstadt, Germany). The most important properties of methylene blue and Rhodamine B are described in Table 1. Regarding to adsorption spectrum curve for the mix solution of MB and RhB (Table 1) it seems that there is no severe or significant overlap between the spectra of the components especially near the maximum wavelength of each dye. Therefore simultaneous analysis of these dyes mixture is possible without using the analytical techniques for resolving the overlapping spectra of binary solutions. An accurately weighted amount of MB or RhB was dissolved in deionized water (1000 mg L⁻¹) as stock solution and the working concentrations was prepared by suitable diluting this solution. The pH of sample solution was adjusted by addition of phosphate buffer solution (0.010 M) which was prepared from phosphoric acid/sodium hydroxide solutions.

Table 1 Properties of the dyes

Properties	Methylene blue	Rhodamine B
Color index number	52 015	45 170
CAS number	61-73-4	81-88-9
Chemical formula	C16H18N3SCl	C28H31ClN2O3
Molecular weight (g mol^−1)	319.85	479.02
Maximum wavelength (nm)	664	554
Adsorption spectrum
Type of dye	Basic blue	Basic red
Use	Redox indicator, coloring paper, cottons and hair colorant	Coloring paper, wools and a tracer

---

Instrumentation

Dye concentration was calculated using T80+ UV-Vis spectrometer at maximum wavelength of each dye (666 nm for MB and 554 nm for RhB) based on linear calibration curve obtained by plotting absorbance toward MB or RhB concentration over desired concentration range. A Metrohm pH meter with a combined double junction glass electrode was used for pH measurements. An ultrasonic bath with heating system (Elmasonic model) was used for the ultrasound-assisted removal procedure. The morphology of the adsorbent was observed by SEM (scanning electron microscopy; Tescan Vega II) under an acceleration voltage of 15 kV. XRD (X-ray diffraction) pattern was recorded with an automated Philips X'Pert X-ray diffractometer with Cu K. radiation (40 kV and 30 mA) for 2θ values over 5–40°. In order to determine the existence of active functional groups on MIL-68(Al), FTIR spectrometer (Shimadzu-8400S, Japan) was used at room temperature. The spectral range was used from 4000 to 400 cm⁻¹. The BET surface areas of the adsorbent were measured using ASAP 2020 (Micromeritics Instrument Corporation) surface area analyzer where N₂ gas was used as adsorbate. Prior to measurement, the adsorbent was degassed at 423 K for 12 h. The pore size distribution and mesopore volume were calculated from the adsorption branch of the isotherm using BJH (Barrett–Joyner–Halenda) method. The micropore volume was also obtained by t-plot micropore analysis. In order to investigate the thermal stability of the sorbent, thermogravimetric analysis (TGA) was performed using STA 504 instrument. Sample was heated at rate of 5 degrees per minute to 873 K under air flow and the approximate weight of sample was 10 mg. Design-Expert, a statistical software package 7.0 was used for experimental design analysis and subsequent regression analysis.

Synthesis of MIL-68(Al)

MIL-68(Al) was synthesized from AlCl₃·6H₂O, 1,4-benzene dicarboxylic acid (terephthalic acid or H₂BDC) and N,N-dimethylformamide (DMF) following the reported method with a little modification.²⁶ Briefly, terephthalic acid (1.67 g; 10 mmol) and AlCl₃·6H₂O (1.63 g; 6.74 mmol) were dissolved separately in 50 mL of DMF (94.4 g; 1292 mmol). The mixture was placed in a round bottom flask equipped with a condenser and was kept stirring and heated for 18 hours at 398 K under air. The mixture was then slowly returned to room temperature. The yellow solid was recovered by filtration. In order to remove the free acid which can remain in the pores, the as synthesized product was dispersed in 3 × 25 mL of DMF under ultrasonic radiation (at 35 °C). To further remove the DMF out of the pores, the same procedure was repeated four times using 4 × 25 mL of MeOH instead of DMF. The as-synthesized MIL-68(Al), after filtration, was dried overnight in 200 °C under vacuum and stored in a desiccator.

Batch adsorption experiments

The influence of variables including pH, initial dye concentration, adsorbent dosage and ultrasonic time, on dye removal was examined in batch mode. All experiments were carried out in triplicate. Specified amounts of each dye solution (25 mL) at a desired concentration and pH (2–10 being adjusted using buffer solution) with a known amount of adsorbent (2–18 mg) were loaded into the flask. Thereafter, the flask was maintained in an ultrasonic bath for specific time intervals (6–14 min) at room temperature. The mixtures were then centrifuged at 6000 rpm for 5 min. Finally, the sample was taken for UV-Vis analysis (using T80+ UV-Vis spectrometer) at maximum wavelength of each dye (666 nm for MB and 554 nm for RhB). The concentration of each dye was determined using a linear calibration curve from 0.3 to 5 mg L⁻¹ and 0.5 to 5 mg L⁻¹ for MB and RhB dyes respectively. The efficiency of dye removal was determined in different experimental conditions according to CCD method. The isotherm and kinetics of the adsorption were determined by measuring the adsorptive uptake of the dye in various initial dye concentrations and at different time intervals under optimum conditions respectively.

The dye removal percentage was calculated using the following relationship:³⁰

The equilibrium adsorption capacity of dyes was calculated according to following equation:³¹

where q_t is the amount of dye adsorbed per unit weight of adsorbent at any time t (mg g⁻¹); C₀ and C_t are the initial and liquid-phase concentrations of the dye solution at any time t (mg L⁻¹), respectively; V is the volume of the dye test solution (L); and w is the amount of the sorbent used (g). When t is equal to the equilibrium agitation time (i.e., C_t = C_e, q_t = q_e), then the amount of dye adsorbed at equilibrium, q_e, (mg g⁻¹) is calculated using eqn (2).

Central composite design (CCD)²⁷

The study of removal efficiency of each dye in individual and joint condition was performed using five level CCD with five factors for binary solution (pH, initial MB concentration, initial RhB concentration, sonication time and adsorbent dosage) and four factors for single dye solution (pH, initial MB or RhB concentration, sonication time and adsorbent dosage). All tests were designed using Design-Expert statistical software package 7.0 leading to 30 and 32 runs for single and binary dye solutions respectively (Tables S1–S3, ESI†). In binary solutions, half-fractional CCD design was applied while using full-CCD for single solutions. The center points were used in order to study the experimental error and the reproducibility of the data.²⁹ The independent variables were coded based on (−1, +1) interval indicating the low and high levels, respectively. The axial points were located at a distance of α from the center and make the design rotatable. The mathematical relationship between five independent variables can be written by the second order polynomial model:³²where, y is the predicted removal percentage (R%); X_i is showing the independent variables (initial dye concentration, pH, sonication time and adsorbent dosage) determined for each experimental run. The parameter β₀ is the model constant; β_i is the linear coefficient; β_ii are the quadratic coefficients and β_ij are the cross-product coefficients.³²

Results and discussion

Characterization of MIL-68(Al)

FTIR spectrum of the as-synthesized MIL-68(Al) presents several peaks in the 1400–1700 cm⁻¹ region (Fig. S1, ESI†). The peak that appears at 1656 cm⁻¹ corresponds to the DMF which indicates there is still the presence of solvent in the material. However the spectrum is in accordance with previous reports.²⁶

According to the Fig. 1, SEM photographs of the as-synthesized MIL-68(Al) show the needlelike crystals of MOF with different lengths. The average thickness of each rod is about 20–30 nm.

Fig. 1 SEM images of the as-prepared MIL-68(Al).

Based on the XRD pattern (Fig. S2, ESI†), the crystalline structure of MIL-68 which is orthorhombic with a double cell volume was confirmed by comparison with literature.^33,34

BET analysis and some textural properties of the prepared MIL-68(Al) are given in Table 2. The BET and Langmuir surface area was measured as 976 and 1484 m² g⁻¹ of synthesized adsorbent respectively. This large surface area and total pore volume favors the suitability of the MOF to remove large amount of dyes very fast using small amount of the adsorbent. The total pore volume (V_Total) evaluated by the liquid N₂ volume at relative pressure of 0.983 was about 0.70 cm³ g⁻¹.

Table 2 Textural properties of synthesized MIL-68(Al)

Sample	SBET (m^2 g^−1)	SINT (m^2 g^−1)	SEXT (m^2 g^−1)	VTotal (cm^3 g^−1)	VMICRO (cm^3 g^−1)
MIL-68(Al)	976	776	200	0.70	0.44

---

The synthesized MOF shows major nitrogen sorption at relative pressures less than 0.25 (Fig. S3, ESI†). Therefore the sample is highly microporous. It is necessary to note that the presence of a small hysteresis loop at high relative pressures indicates the presence of mesopores. Pore size distribution can be more clearly observed in Fig. S3 inset (for details, see ESI†).

Thermal stability and decomposition of the tagged MOF were monitored by TGA analysis (Fig. S4, ESI†). The curve indicates similar weight loss behaviors with previous reports except a little difference.^26,33 In this study, three successive events was occurred instead of two events. The first weight loss is attributed to the departure of guest molecules (H₂O, MeOH and DMF) from room temperature up to 350 °C. The second one between 350 and 500 °C is assigned to the removal of unreacted ligand species which was not seen in other reports. This may be due to the presence of traces of free dicarboxylic acid in the powder or to insufficient purification. The last weight loss around 525 °C occurs with the decomposition of the terephthalate ligand following the collapse of the structure to produce dense metal oxide, Al₂O₃.²⁶

Central composite design (CCD)

As mentioned above, CCD is an efficient five-level design for fitting second-order response surfaces. In this step the independent factors (contact time, pH, adsorbent dosage and dye concentration) were prescribed into three levels (low, basal and high) with coded value (−1, 0, +1) and the star points of +2 and −2 for +α and −α respectively for each set of experiments. The corresponding design matrix used by CCD design model consisted of 30 and 32 experiments including six center points in each set for single and binary dye solution respectively (for details, see Tables S1–S3, ESI†).

Analysis of variance (ANOVA) calculated by Design-Expert 7.0.0 determined the adequacy of the model besides the significance and magnitude of the effect of main and interaction of factors (see Tables S4–S7, ESI†). A p-value less than 0.05 in the ANOVA table demonstrates the statistical significance of an effect at 95% confidence level.

Moreover, F-test was applied to evaluate the statistical significance of all terms in the polynomial equation within 95% confidence interval.²⁹ The Model F-value of 218.68, 107.90, 286.73 and 48.37 for single solution of MB, single solution of RhB, MB in binary solution and RhB in binary solution respectively, implies that the model is significant. The values ≥0.050 illustrate the model terms that are not significant. For example in the case of MB uptake in binary solutions, pH (B), adsorbent dosage (C), MB concentration (D) and interactions BC, CD, B², C² and D² are significant model terms. The lack of fit (LOF) is the variation of the data around the fitted model. The non-concurrence between the proposed model and experimental data is not favorable and makes the LOF significant.³⁵ According to our results (Tables S4–S7, ESI†), the LOF of p-value is 0.0823, 0.1666, 0.1352 and 0.1235 for single solution of MB, single solution of RhB, MB in binary solution and RhB in binary solution respectively. These data clearly show that LOF is not significant relative to the pure error indicating suitability of the model for well-fitting the experimental data.

According to Fig. 2, the points of all predicted and actual responses obtained by experiments confirm good fit. Based on the achieved statistical results, it can be concluded that central composite design was adequate to predict the removal percentage of the considered dyes within the range of variables determined. The final predicted model in terms of significant actual factors for evaluation the removal of MB in single solution (R₁), RhB in single solution (R₂), MB in binary solution (R₃) and RhB in binary solution (R₄) were specified by Design-Expert software and were expressed as the eqn (S1)–(S4) (for details, see ESI†), respectively.

Fig. 2 The experimental data versus predicted data for (a) MB in single solution, (b) RhB in single solution, (c) MB in binary solution and (d) RhB in binary solution.

For MB dye in single solution, RhB in single solution, MB in binary solution and RhB in binary solution the “Predicted R²” (0.967, 0.945, 0.9598 and 0.7625 respectively) is in reasonable agreement with the “Adjusted R²” (0.991, 0.982 0.994 and 0.968 respectively) and it can be inferred that the model regression coefficient (R²: 0.996, 0.992, 0.998 and 0.988 respectively) is reasonable (see Tables S4–S7, ESI†).

Response surface methodology

In this step, after finding the critical factors representing the most impact on final results, response surface methodology (RSM) was utilized to optimize these vital parameters. In order to ease of study, the response surface plots of removal efficiency (R₁% and R₂% for MB and RhB respectively) versus significant factors are depicted in Fig. 3 only for binary solutions. These plots were resulted for a given pair of factors at fixed and optimal values of other variables. It is obvious that in the case of MB removal from binary solution of MB and RhB dyes, adsorbent dosage, pH and initial concentration of MB had the most effect on adsorption yield (Fig. 3a–d). Not only from the eqn (S3)† but also from the Fig. 3 it can be concluded that dosage and pH had significant positive effect on removal, if pH or adsorbent dosage increased then removal efficiency may be increased. In fact the surface charge of the sorbent plays an important role in electrostatic interaction between the cationic dye and active sites of the surface. Therefore, at higher pH value, negatively charged surface of the MOF resulted into higher removal of cationic dye through the electrostatic attraction. At low pH values, a competition will occur between high amounts of H⁺ and cationic dyes to occupy the active sites of the adsorbent which leads to decrease in dye adsorption efficiency. This phenomenon is also amplified by enhancement the repulsive force between cationic dye and adsorbent as a result of the protonation of analyte and adsorbent in acidic condition.³⁶

Fig. 3 3D response surface graphs for MB removal in binary solution of dyes: (a) dose vs. pH (b) MB concentration vs. dose. (c) MB concentration vs. pH, (d) RhB concentration vs. time and for RhB removal in binary solution of dyes: (e) MB concentration vs. dose, (f) RhB concentration vs. dose.

The removal percentage of both dyes increased by addition the adsorbent dosage which is related to the enhancement of surface area as well as the availability of more adsorption sites and lower ratio of dye molecule to vacant sites of the adsorbent.^37,38 On the other hand an increase in MB concentration causes the saturation of adsorption sites leading to decrease in the removal percentage. It has been also observed from Fig. 3d the surface plot of MB removal (R%) vs. time and RhB concentration have no significant influence on the response. This phenomenon confirms that the active sites on the sorbent are first occupied by MB component and RhB can't be considered as a competitor for MB uptake in mix solutions of these dyes. Furthermore, very rapid adsorption rate of MIL-68(Al) was verified by negligible effect of time factor (in the range of 6 to 14 min) on both of dyes removal process.

As is displayed in eqn (S4)† and Fig. 3e and f for RhB uptake, time and pH had the minimum effects among the other main factors but adsorbent dose, MB concentration and RhB concentration posed remarkable impact on final results. According to our conclusions, the presence of MB especially at high concentration leads to substantial decrease in RhB removal compared to the dye uptake percentage in single solution of RhB. Therefore the competition to occupy the adsorption sites on the surface lead to considerable reduction in removal of RhB component. However according to our responses obtained by experimental design, it will be possible to eliminate more than 80% of both dyes from aqueous solutions by regulating the conditions (see Table S1, Run11, ESI†). Dyes are generally small molecules. Therefore pores with dimensions greater than 1 nm are suitable for their internalization and adsorption by p–p interaction.³⁹ The greater affinity of MIL-68(Al) for methylene blue rather than Rhodamine B may be due to the presence of sulfur atom in MB's structure which make it to be more attracted by sorbent through acid–base interaction of sulfur from analyte and metal cation of the MIL-68(Al).⁴⁰

The optimized values obtained from RSM values for the factors such as adsorbent dosage, pH, MB concentration, RhB concentration, sonication time and the predicted removal percentage for each dye are displayed in Table 3. At this conditions the other studies including isotherm, kinetics and reusability of the sorbent was done.

Table 3 Optimum conditions derived by RSM design for dyes removal in single and binary solutions

Optimal conditions					Predicted removal%
Adsorbent dosage (g)	pH	MB concentration (mg L^−1)	RhB concentration (mg L^−1)	Sonication time (min)	R1	R2	R3	R4
0.01185	7.96	67.00	—	9.60	102.2	—	—	—
0.01375	5.00	—	19.5	10.00	—	97.2	—	—
0.01400	6.50	60.00	15.00	9.91	—		103.11	71.25

---

Adsorption isotherms of dyes

In this section, several adsorption isotherm models have been used in order to attain useful information about the adsorption mechanism, potential adsorption capacity of the used MOF and evaluate tendency of the adsorbent toward each dye mathematically.^35,41,42 Under optimum conditions of each dye (see Table 3), the Langmuir, Freundlich and Temkin models as the most frequently employed models were studied to explain the type of interaction between the adsorbate (dye) and adsorbent (MIL-68) at equilibrium. Adsorption isotherms were undertaken over batch system at room temperature with constant adsorbent dosage (0.01185, 0.01375 and 0.01400 g for single MB solution, RhB solution and both dyes in a binary solution respectively) and pH (7.96, 5.0, 6.5 for single MB solution, single RhB solution and both dyes in a binary solution respectively) with 25 mL of dye solution at various initial concentrations. The conventional isotherm equations followed by their constant parameters and the correlation coefficient (R²) for each model are summarized in Table 4.

Table 4 Isotherm parameters for the adsorption of dyes onto MIL-68(Al) adsorbent

Isotherm	Equation	Plot	Parameters	Values of parameters
				In single component system		In multi component system
				MB	RhB	MB	RhB
Langmuir		The values of qm and KL were calculated respectively from the slope and intercept of the plot of Ce/qe versus Ce	qm (mg g^−1)	1666.6	1111.1	227.27	29.32
			KL (L mg^−1)	0.0199	0.0157	1.467	0.457
			R^2	0.9745	0.9851	0.9900	0.9507
Freundlich		The values of KF and n were calculated respectively from the intercept and slope of the plot of ln qe versus ln Ce	KF (L mg^−1)	141.99	44.99	130.02	11.654
			n	2.508	1.9139	6.32	3.30
			R^2	0.9714	0.9929	0.9370	0.9817
Temkin	qe = B1 ln KT + B2 ln Ce	The values of B1 and KT were calculated from the plot of qe against ln Ce	KT (L mg^−1)	1.371	0.679	139.60	4.53
			B1	256.84	150.9	26.715	6.2741
			R^2	0.9222	0.9006	0.9762	0.9676

---

According to Langmuir isotherm model (see Table 4), the values of K_L (the Langmuir adsorption constant, L mg⁻¹) and q_m (theoretical maximum monolayer adsorption capacity, mg g⁻¹) were calculated from the intercept and slope of the plot of C_e/q_e vs. C_e, respectively.⁴³ From Freundlich isotherm model K_F (the Freundlich coefficient corresponding to adsorption capacity, (mg g⁻¹) (mg L⁻¹)^−1/n) and n (an exponential coefficient indicative of the adsorption intensity) were obtained from the intercept and slope of the linear plot of ln q_e vs. ln C_e, respectively. The heat of the adsorption (B₁ = RT/b, J mol⁻¹) and K_T (the equilibrium binding constant corresponding to the maximum binding energy, L mol⁻¹) were also achieved by Temkin isotherm model. In this equation, T is the absolute temperature (K), R is the universal gas constant (8.314 J mol⁻¹ K⁻¹) and K_T is the equilibrium binding constant (L mg⁻¹).^28,42,44

As shown in Table 4, for MB dye in both single and binary solution, the high correlation coefficient of Langmuir model confirms the applicability of this model for interpretation of the experimental data over the whole concentration range. Furthermore, the higher R² value for RhB dye in both single (0.9929) and binary solutions (0.9817) and also the value of n represents the suitability of this model for fitting the experimental data over the whole concentration range of RhB. It is important to imply that MIL-68(Al) showed an enormous potential to adsorb both analytes in single component solution of each dye. As seen in Table 4, maximum adsorption capacities (q_m) for MB and RhB in single solutions are about 1666 and 1111 mg g⁻¹, respectively indicating the considerable adsorption capability of the used MOF for MB and RhB dyes removal. Although this parameter (q_m) had a significant decrease in binary solution, it is still comparable with the materials reported so far.

Adsorption kinetic study

In general, the adsorption rate is considered as one of the most important factors for choosing an ideal adsorbent especially in industries. For this purpose the adsorption capability of MB and RhB dyes by MIL-68(Al) was analyzed at various sonication times (with constant adsorbent dosage, dye concentration and pH) and the kinetic data was evaluated by pseudo first-order, pseudo second-order and Elovich models (Table 5).

Table 5 Kinetic parameters for the adsorption of dyes onto MIL-68(Al) adsorbent

Model	Equation	Plot	Parameters	Values of parameters
				In single component system		In multi component system
				MB	RhB	MB	RhB
First-order		The values of k1 and qe1 were calculated from the slope and intercept of the plot of log(qe − qt) versus t, respectively	qe (mg g^−1)	1.23	11.168	2.694	16.62
			k1 (L min^−1)	2.98	0.8136	0.356	0.540
			R^2	0.8902	0.9996	0.8931	0.8505
Second-order		The values of k2 and qe2, were calculated from the intercept and slope of the plot of t/qt versus t, respectively	k2	0.1183	4.006	0.288	0.0339
			qe2	153.84	34.84	107.52	21.98
			R^2	1.0000	1.0000	1.0000	0.9941
Elovich	qt = l/β ln(αβ) + 1/β ln(t)	The value of β were calculated from the slope of the plot of qt versus ln(t)	β	2.129	0.530	1.850	0.206
			R^2	0.9575	0.9348	0.8985	0.8953
			qe(exp)	152.69	34.348	106.75	19.48

---

According to our results the sorption process under optimum condition is surprisingly fast. For MB uptake in single component solution, a contact time of 30 seconds was enough to reach the equilibrium. This parameter for single RhB dye solution was about 300 seconds which is relatively too short. In the case of binary solution, MB removal is still very fast but a little more time (360 seconds) is needed for RhB dye to achieve the equilibrium. The estimation of three models was performed by plotting the values of log(q_e − q_t) vs. t for pseudo-first-order model, t/q_t vs. t for pseudo-second-order model and q_t vs. ln t for Elovich model.^38,45–47 The rate constants k₁ and k₂ and also B₁ were obtained from the slopes of corresponding linear plots (see Table 5).

Regarding to the high regression coefficient of the pseudo-second-order, the adsorption of dyes on MIL-68(Al) are best evaluated by this kinetic model (R² = 0.99–1.00). Moreover, the closeness of calculated q_e values for pseudo-second-order kinetic model and experimental q_e values indicates the applicability of this model to describe the adsorption process. It should be noted that, the general mechanism of adsorption process can be explained by the following three steps: (a) mass transfer; (b) intraparticle or pore diffusion and (c) sorption onto interior active sites or chemical reaction between analyte and adsorption. In other words, adsorption kinetics is controlled by either liquid phase mass transport rate or intra-particle mass transport rate. If pore diffusion limits the sorption process, the relationship between the analyte concentration and adsorption rate will be linear in and the adsorption kinetic will agree with pseudo first-order kinetic model. Furthermore, the overall rate of dyes sorption processes will be governed by the chemical process if the adsorption data show a good compliance with pseudo second-order reaction mechanism. According to our results, as the pseudo second-order model presented high correlation coefficient, chemical reaction appears to be rate controlling stage.⁴⁸

Regeneration study

Regenerability of the dye-loaded adsorbent for further use is one of the determining factors for evaluation the performance of adsorbents.⁴⁹ So, in the current study some low cost chemicals were examined to retrieve the adsorbent which was loaded previously by single or binary solutions of dyes under optimum conditions. Among the different chemicals and solvents such as HCl (0.01 M), NaCl (0.01 M) and methanol, the latter was found the best. After the first dye-loading, the adsorbent was washed several times by methanol under ultrasonication and centrifugation and then dried in vacuum at 190 °C for 10 h afterward the adsorbent reused for the subsequent run under optimum condition and the concentration of each dye was determined using UV-Vis spectrophotometry. As shown in Fig. 4 the adsorption performance can be regained by this solvent-washing method and after two cycles of adsorption/desorption, the removal efficiency of MB and RhB still reach 91.37 and 60.14% respectively.

Fig. 4 Removal percentage for toward sorption/desorption cycles for simultaneous removal of MB and RhB dyes using MIL-68(Al).

Application to real samples

Regarding to the practical applicability of the proposed nanopore adsorbent for dye removal of real samples, it was also tested for water samples from North Sea water, tap water, spring water (Cheshmeh-Ali, Shahr-e-Rey, Iran) and laundry wastewater. All samples were spiked with certain concentrations of MB (60, 70 and 80 mg L⁻¹) and RhB (15, 20 and 25 mg L⁻¹) dyes under optimum condition and after treating by adsorbent, the remaining dye concentration was measured via UV-Vis spectrophotometry. The results presented in Table 6 show good adsorption efficiency especially for MB dye supporting the validity of the proposed method for removal of these dyes from real samples.

Table 6 Removal percentage of MB and RhB from their binary system from different real samples using MIL-68(Al) (for 3 replicates)

Sample	MB (mg L^−1)	RhB (mg L^−1)	% removal MB	% removal RhB
Sea water	60	15	99.16 ± 2.25	70.48 ± 2.39
	70	20	98.05 ± 1.83	63.99 ± 2.01
	80	25	96.34 ± 1.76	58.22 ± 3.07
Tap water	60	15	98.59 ± 1.21	35.26 ± 3.17
	70	20	97.72 ± 2.73	30.22 ± 2.73
	80	25	94.78 ± 2.25	27.15 ± 2.48
Spring water	60	15	98.51 ± 2.33	72.62 ± 1.64
	70	20	92.08 ± 1.89	68.96 ± 2.15
	80	25	90.76 ± 2.13	55.37 ± 3.26
Laundry wastewater	60	15	91.12 ± 1.63	69.69 ± 2.18
	70	20	88.40 ± 2.22	63.07 ± 3.73
	80	25	82.23 ± 1.55	50.47 ± 2.09

---

An investigation through the effect of ionic strength

To evaluate the effect of salt on simultaneous removal efficiency of the dyes, NaCl and KCl solutions with different concentrations of 0.01, 0.05 and 0.1 mol L⁻¹ were used separately to adjust the solution salinity under optimum condition. Our experiments confirmed that the removal percentage of both dyes was not significantly influenced by ionic strength. However by addition of these salts, an ignorable increase in removal efficiency occurred which may be due to the decrease in solubility of the dyes in water and improvement the hydrophobic interaction between dye molecule and the adsorbent in the presence of salt leading to enhance dye removal percentage.²¹

Comparison with other methods

The removals of MB and RhB have been studied by various adsorbents and dye adsorption capacities have been reported in literatures. Table 7 displays the performance of the proposed method comparing with different adsorbents in terms of adsorption capacities and uptake condition. From these data, it can be seen that the adsorption capacities of the synthesized adsorbent for both dyes are significantly higher in most cases previously reported. In the other words, MIL-68(Al) has great potential ability for rapid dye removal with high efficiency under nearly mild condition.

Table 7 Comparison for the removal of dyes by different methods and adsorbents

Adsorbent	Dye	tct^a (min)	qm/mg g^−1	Ref.
a Contact time.b Maximum sorption capacity in single dye solution.c Maximum sorption capacity in binary dyes solution.d Not reported.
ZnS:Cu/AC	MB	2.2	106.9^b	32
			51.7^c
Sodium montmorillonite	RhB	1440	42.19	50
Fe3O4 nanoparticles	MB	2	45.4^c	51
ZnS/AC	MB	15	100	52
Wood millet carbon	MB	18	4.93	53
NiS/AC	MB	5.46	52	54
Porous GO/hydrogel	MB	NR^d	769.23	4
Ni doped FeO(OH)–NWs–AC	RhB	2	210.17	36
MIL-125(Ti)	RhB	180	59.92	8
MCM-22	RhB	14 400	0.054	55
MIL-68(Al)	MB	8	1666.67^b	This study
		9.9	227.27^c
MIL-68(Al)	RhB	10	1111.11^b	This study
		9.9	29.32^c

---

Conclusion

In summary, MIL-68(Al) as a kind of MOF was successfully synthesized through a simple one-pot refluxing route reported previously. The potential of the prepared material for MB and RhB dyes removal in both single-component and multi-component solutions was examined by experimental design methodology under various condition of pH, adsorbent dosage, initial dye concentration and ultrasonic time. The results of this research confirmed that adsorbent dosage and pH of the solution had significant effect on the removal efficiency of dyes.

In our study, MIL-68(Al) presented great capability to adsorb MB and RhB especially in separate aqueous solutions however by applying a specific condition both dyes can be removed simultaneously with high efficiency. According to the equilibrium studies, MB uptake in single and binary solutions of dyes fitted well with Langmuir model while the Freundlich isotherm was the best model for describing the multilayer adsorption of RhB from both single and binary dye solutions.

According to our results the sorption process is surprisingly fast under optimum condition and using the ultrasonic waves. For MB uptake in single component solution, a contact time of 30 seconds was enough to reach the equilibrium. This parameter for single RhB dye solution was about 300 seconds which is relatively too short. In the case of binary solution, MB removal is still very fast but a little more time (360 seconds) is needed for RhB dye to achieve the equilibrium. The maximal adsorption capacities of MB and RhB in single and binary solutions evaluated from Langmuir isotherm were 1666, 1111, 227 and 29 mg g⁻¹ respectively. Adsorption kinetics of MB and RhB in single and binary solutions onto the sorbent followed the pseudo-second-order kinetic model. In addition, the adsorbent can be easily regenerated by methanol and reused for subsequent runs. The results from investigation on real samples indicated that the as-prepared MOF could be considered from a practical point of view as a potential adsorbent for simultaneous removal of MB and RhB from water.

Acknowledgements

The financial support of this study by Iran University of Science and Technology and Nano Technology Initiative Council are gratefully acknowledged. The author acknowledges financial support from the Iran National Science Foundation (INSF).

References

1. P. K. S. Malik and S. Malik, RSC Adv., 2013, 3, 1902–1915.
2. M. Ghaedi, E. Nazari, R. Sahraie and M. K. Purkait, Desalin. Water Treat., 2014, 52, 5504–5512.
3. A. Pourjavadi, M. Nazari, B. Kabiri, S. H. Hosseini and C. Bennett, RSC Adv., 2016, 6, 10430–10437.
4. M. Shariati-Rad, M. Irandoust, S. Amri, M. Feyzi and F. Jafari, Int. Nano Lett., 2014, 4, 91–101.
5. R. Ansari, M. S. Tehrani and A. Mohammad-khah, J. Wood Chem. Technol., 2012, 32, 198–209.
6. L. Qu, T. Han, Z. Luo, C. Liu, Y. Mei and T. Zhu, J. Phys. Chem. Solids, 2015, 78, 20–27.
7. A. Ahmad, S. H. Mohd-Setapar, S. C. Chuo, A. Khatoon, W. A. Wani, R. Kumar and M. Rafatullah, RSC Adv., 2015, 5, 30801–30818.
8. H. Guo, F. Lin, J. Chen, F. Li and W. Weng, Appl. Organomet. Chem., 2015, 29, 12–19.
9. L. Cui, X. Guo, Q. Wei, Y. Wang, L. Gao, L. Yan, T. Yan and B. Du, J. Colloid Interface Sci., 2015, 439, 112–120.
10. R. Ansari, M. S. Tehrani and M. B. Keivani, J. Wood Chem. Technol., 2013, 33, 19–32.
11. P. Sharma and M. R. Das, J. Chem. Eng. Data, 2013, 58, 151–158.
12. M. S. Sajab, C. H. Chia, C. H. Chan, S. Zakaria, H. Kaco, S. W. Chook, S. X. China and A. M. Noor, RSC Adv., 2016, 6, 19819–19825.
13. A. Shahat, M. R. Awual, M. A. Khaleque, M. Z. Alam, M. Naushad and A. M. S. Chowdhury, Chem. Eng. J., 2015, 273, 286–295.
14. D. Zheng, M. Zhang, L. Ding, Y. Zhang, J. Zheng and J. Xu, RSC Adv., 2016, 6, 11973–11979.
15. L. Xie, D. Liu, H. Huang, Q. Yang and C. Zhong, Chem. Eng. J., 2014, 246, 142–149.
16. X. Lian and B. Yan, RSC Adv., 2016, 6, 11570–11576.
17. P. Falcaro, R. Ricco, C. M. Doherty, K. Liang, A. J. Hill and M. J. Styles, Chem. Soc. Rev., 2014, 43, 5513–5560.
18. E. Haque, N. Abedin Khan, C. M. Kim and S. H. Jhung, Cryst. Growth Des., 2011, 11, 4413–4421.
19. S. Y. Jia, Y. F. Zhang, Y. Liu, F. X. Qin, H. T. Ren and S. H. Wu, J. Hazard. Mater., 2013, 262, 589–597.
20. S. Lin, Z. Song, G. Che, A. Ren, P. Li, C. Liu and J. Zhang, Microporous Mesoporous Mater., 2014, 193, 27–34.
21. S. Huo and X. Yan, Analyst, 2012, 137, 3445–3451.
22. F. Ke, L. G. Qiu, Y. P. Yuan, X. Jiang and J. F. Zhub, J. Mater. Chem., 2012, 22, 9497–9500.
23. A. R. Chowdhuri, D. Bhattacharya and S. K. Sahu, Dalton Trans., 2016, 45, 2963–2973.
24. S. H. Huo and X. P. Yan, J. Mater. Chem., 2012, 22, 7449–7455.
25. L. N. Jin, X. Y. Qian, J. G. Wanga, H. Aslan and M. Dong, J. Colloid Interface Sci., 2015, 453, 270–275.
26. Q. Yang, S. Vaesen, M. Vishnuvarthan, F. Ragon, C. Serre, A. Vimont, M. Daturi, G. D. Weireld and G. Maurin, J. Mater. Chem., 2012, 22, 10210–10220.
27. M. Saghanejhad Tehrani and R. Zare-Dorabei, Spectrochim. Acta, Part A, 2016, 160, 8–18.
28. S. Honary, P. Ebrahimi, H. Asgari-rad and F. Mohamadpour, Int. J. Biomed. Nanosci. Nanotechnol., 2014, 10, 257–261.
29. M. Roosta, M. Ghaedi and M. Mohammadi, Powder Technol., 2014, 267, 134–144.
30. S. Wang, J. Wei, S. Lv, Z. Guo and F. Jiang, Clean: Soil, Air, Water, 2013, 41, 992–1001.
31. R. Liu, Y. Liu, X. Zhou, Z. Zhang, J. Zhang and F. Dang, Bioresour. Technol., 2014, 154, 138–147.
32. A. Asfaram, M. Ghaedi, S. Hajati, M. Rezaeinejad, A. Goudarzi and M. K. Purkait, J. Taiwan Inst. Chem. Eng., 2015, 53, 80–91.
33. B. Seoane, V. Sebastián, C. Téllez and J. Coronas, CrystEngComm, 2013, 15, 9483–9490.
34. T. Han, Y. Xiao, M. Tong, H. Huang, D. Liu, L. Wang and C. Zhong, Chem. Eng. J., 2015, 275, 134–141.
35. M. A. Islam, V. Sakkas and T. A. Albanis, J. Hazard. Mater., 2009, 170, 230–238.
36. F. N. Azad, M. Ghaedi, A. Asfaram, A. Jamshidi, G. Hassani, A. Goudarzi, M. H. A. Azqhandi and A. Ghaedi, RSC Adv., 2016, 6, 19768–19779.
37. M. K. Sahu, U. K. Sahu and R. K. Pate, RSC Adv., 2015, 5, 42294–42304.
38. M. Ravanan, M. Ghaedi, A. Ansari, F. Taghizadeh and D. Elhamifar, Spectrochim. Acta, Part A, 2014, 123, 467–472.
39. A. Banerjee, R. Gokhale, S. Bhatnagar, J. Jog, M. Bhardwaj, B. Lefez, B. Hannoyer and S. Ogale, J. Mater. Chem., 2012, 22, 19694–19699.
40. G. Blanco-Brieva, J. M. Campos-Martin, S. M. AL-Zahrani and J. L. G. Fierro, Global NEST J., 2010, 12, 296–304.
41. D. Kavak, J. Hazard. Mater., 2009, 163, 308–314.
42. M. Roosta, M. Ghaedi, A. Daneshfar and R. Sahraei, Spectrochim. Acta, Part A, 2014, 122, 223–231.
43. B. K. Jung, Z. Hasan and S. H. Jhung, Chem. Eng. J., 2013, 234, 99–105.
44. F. H. Malayeri, M. R. Sohrabi and H. Ghourchian, Int. J. Biomed. Nanosci. Nanotechnol., 2012, 8, 79–86.
45. B. Fang, Z. Bao, L. Lu, L. Zhao and H. Wang, RSC Adv., 2016, 6, 61–64.
46. V. K. Gupta, D. Pathania, N. C. Kothiyal and G. Sharma, J. Mol. Liq., 2014, 190, 139–145.
47. X. Zhu, Y. Liu, F. Qian, C. Zhou, S. Zhang and J. Chen, Bioresour. Technol., 2014, 154, 209–214.
48. M. Özacar and I. A. Sengil, Process Biochem., 2005, 40, 565–572.
49. M. H. Mashhadizadeh and M. Amoli-Diva, J. Anal. At. Spectrom., 2013, 28, 251–258.
50. P. P. Selvam, S. Preethi, P. Basakaralingam, N. Thinakaran, A. Sivasamy and S. Sivanesan, J. Hazard. Mater., 2008, 155, 39–44.
51. M. Ghaedi, S. Hajjati, Z. Mahmudi, I. Tyagi, S. Agarwal, A. Maity and V. K. Gupta, Chem. Eng. J., 2015, 268, 28–37.
52. M. Ghaedi, H. A. Larki, S. N. Kokhdan, F. Marahel, R. Sahraei, A. Daneshfar and M. K. Purkait, Environ. Prog. Sustainable Energy, 2012, 32, 535–542.
53. M. Ghaedi and S. N. Kokhdan, Spectrochim. Acta, Part A, 2015, 136, 141–148.
54. M. Ghaedi, M. Pakniat, Z. Mahmoudi, S. Hajati, R. Sahraei and A. Daneshfar, Spectrochim. Acta, Part A, 2014, 123, 402–409.
55. S. Wang, H. Li and L. Xu, J. Colloid Interface Sci., 2006, 295, 71–78.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra28052d

---