Solid-phase extraction of organic dyes on mixed-ligand Zr(IV) metal-organic framework

Effective removal of organic dyes from aqueous solution is an important task for targeted wastewater treatment. In this work, a mixed-ligand Zr(IV) metal organic framework based on terephthalic acid and 1,10-phenanthroline was used for the solid-phase extraction of dyes. The maximum adsorption capacity of the sorbent for Congo red and methylene blue dyes reached 40 mg/g. The adsorption behavior was studied, including the effect of time, temperature, adsorbent dosage, pH and coexisting ions. The reliability of the sorbent for removing artificial dyes from food samples and in artificial seawater was tested. The possibility of using the sorbent as a filler for column chromatography for dye separation is shown.


Introduction
Zirconium-based metal-organic frameworks (Zr-MOF) have quickly gained a lot of attention in analytical chemistry due to their promising properties, including good selectivity, high productivity, low cost and regenerability, and strong interaction of Zr-OH groups with specific functionality of pollutants (Cavka et al. 2008;Butova et al. 2016;Tsivadze et al. 2019;Solovtsova et al. 2020;Pribylov et al. 2021;Uflyand et al. 2021;Bai et al. 2016).This is explained by high chemical and mechanical stability of Zr-MOFs due to the formation of strong bonds of zirconium with carboxyl groups (Cavka et al. 2008;Wu et al. 2013;Howarth et al. 2016;Schaate et al. 2011).As a kind of stable MOFs, Zr-MOFs demonstrate high efficiency in removing dyes (Molavi et al. 2018), antibiotics (Chen et al. 2017), and heavy metal ions (He et al. 2018).The disadvantages of Zr-MOFs are low adsorption values and long equilibrium time.Among the methods for increasing the adsorption capacity, the following should be noted.Thus, the use of crystallization modulators in the synthesis makes it possible to increase the size of Zr-MOF crystallites up to four orders of magnitude (from 10 to 100 μm) (Valenzano et al. 2011;Wißmann et al. 2012;Gutov et al. 2016;Marshall et al. 2016;Shearer et al. 2016).Typical modulators are monocarboxylates to coordinate with metal clusters (coordination modulators) and thus to compete with the linker (Valenzano et al. 2011;Wißmann et al. 2012;Gutov et al. 2016;Marshall et al. 2016) or acidic species to reduce linker deprotonation (protonation modulators) (Katz et al. 2013).The stability of these solids is well suited for further studies of the effect of functionalization on their properties, and several experimental studies have already determined the possibility of additional introduction of new functional groups into the structure of these compounds (Garibay and Cohen 2010).Among the accepted approaches to synthesis, including stepwise, bottom-up, etc., the strategy of mixed ligands plays an important role in the construction of MOFs (Dzhardimalieva and Uflyand 2017;Yang et al. 2020).Mixed ligands mean that two types of organic ligands with similar geometric structures or completely different configurations can be introduced to synthesize MOF.Regarding ligand acceptance, the carboxyl ligand can satisfy the coordinated configuration of the metal ion due to its flexible coordination mode, and the polypyridine organic ligand can terminate or chelate with the metal ion.As an example, we note the use of the mixed ligand strategy to obtain Zr-MOF, including, in addition to carboxylate ions, various polypyridine ligands such as 2,2'-bipyridine and 1,10-phenanthroline (Feng et al. 2021;Uflyand et al. 2020;Qiu et al. 2009;Dun et al. 2020;Yang et al. 2003;Yang et al. 2019;Xu et al. 2004;Go et al. 2004;Go et al. 2005).
The aim of our work was to use a ligand-mixed Zr-MOF based on terephthalic acid and 1,10-phenanthroline in solid-phase extraction (SPE) of organic dyes.
The adsorbates were dyes methylene blue (MB), Congo red (CR), and malachite green (MG) with molecular formulas C16H18N3SCl, C32H22N6Na2O6S2, and C23H25ClN2, water solubility of 50, 10, and 110 g/L (20 °C), as well as molecular weights of 319.85, 696.66, and 301.453 g/mol, respectively (Fig. 1).They were purchased from Sigma-Aldrich.The original dye solution with a concentration of 200 mg/L was obtained by dissolving an accurate weighed portion of the dye in distilled water.For the preparation of experimental solutions, the dye stock solution was diluted in precise proportions to the required initial concentrations.

Synthesis of sorbent
The synthesis of the sorbent was carried out according to described procedure (Kharissova et al. 2022).

Experimental methods
Elemental analyzes were performed on a CHNOS vario EL cube analyzer (Elementar Analysensysteme GmbH, Germany).Zirconium was determined on an energy dispersive X-ray fluorescence spectrometer «X-Art M» (Comita, Russia) or atomic absorption spectrometer «MGA-915» (Lumex, Russia).IR spectra were recorded on a Varian Excalibrum 3100 FTIR spectrometer using KBr pellets and Softspectra data analysis software.X-ray diffraction (XRD) analysis was carried out on a Phywe XR 4.0 diffractometer with CuKα radiation (λCu = 1.54184Å) in the range of 2θ = 5-80° angles 2θ at a scanning rate of 5°/min and a temperature of 25 °C.The radiation intensity was 35 kV.Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX) was obtained on a Zeiss LEO SUPRA 25 scanning electron microscope.The nitrogen adsorption/desorption isotherms were obtained at 77 K (liquid N2) using the AUTOSORB-1 system (Quantachrome, Boynton Beach, FL, USA) by the static volumetric method; before analysis the samples were degassed by heating at 150 °C for 12 h in vacuum.The Brunauer-Emmett-Teller surface area was obtained from the amount of N2 physically sorbed at various relative pressures (P/P0), based on the linear part of the 6-point adsorption data at P/P0 = 0.02-0.10.For assess the adsorption capacity, ultrahigh purity gases (99.995%) were used.The equilibrium values of the adsorbed gas volume were calculated as a function of the relative pressure in the system considering the weight of the sample and the volumes of various parts of the instrument.

Solid-phase extraction procedures
In this work, serial adsorption experiments were carried out to remove MB, CR, and MG as basic dyes from aqueous solutions using the synthesized sorbent.
Solutions of the corresponding dyes with a volume of 200 mL were placed in a thermostated 300 mL beaker at 10, 18, and 35 °C.When the set temperature is reached, a sorbent (0.1 g) was added and every 5, 10, 15, 30, 45, 60 min, 10 mL of the sorbent suspension was taken from the dye solution and centrifuged for 5 min at 4500 rpm.A UV-visible spectrophotometer (Varian, Cary 50) was used to determine the concentration of the residual dye in the fugate for CR at λmax equal to 492 nm (Abo-State et al. 2017), and for MB at λmax equal to 664 nm (Hu and Tong, 2007).
Removal efficiency R (%) was calculated using the equation ( 1 where C0 is the initial concentration; Сt is the concentration of the dye at time t.The equilibrium adsorption qe (mg/g) was calculated using the equation ( 2): where V is the adsorbate solution volume (L), m is the adsorbent weight (g).
The pseudo-first order adsorption model can be expressed by equation (3) (Simonin 2016): (3) where k1 (min -1 ) is the rate constant of the pseudo-first order model.
After a definite integration, equation (3) takes the form: ln(qeqt) = lnqe -k1t (4) The constant k1 can be determined experimentally by the slope of the linear graphs ln(qeqt) versus t.
The effect of pH on the adsorption of the dye was studied in solutions with a pH of 3-11.The prepared suspensions were shaken for 120 min to reach equilibrium at a temperature of 18 °C, filtered through membrane filters with a pore size of 0.2 μm, and the concentration of the residual dye was determined.
Adsorption isotherms were calculated according to the linearized equations of Langmuir ( 5) and Freundlich (6).
The thermodynamic parameters of adsorption ∆G 0 , ∆H 0 and ∆S 0 are calculated either graphically from the dependence of lnKD on 1/T, where KD is the adsorbate distribution coefficient calculated by the equation:

Synthesis and characterization of the sorbent
In this work, the sorbent was synthesized by the solvothermal method by the interaction of ZrCl4, terephthalic acid, and 1,10-phenanthroline in DMF.Based on the literature data and the results obtained, the following structure of the synthesized compound can be proposed (Fig. 2).

Fig. 2 Plausible structure of the sorbent
The obtained adsorption-desorption isotherms of N2 at 77 K and the pore size distribution for the synthesized MOF are shown in Fig. 3. Analysis of nitrogen physiosorption isotherms shows that MOF has constant porosity and a mesoporous structure with a large surface area corresponding to adsorption type II.These characteristics are typical of high-quality MOF material with little pore collapse and no residual reactant.A high nitrogen adsorption of 561 cm 3 /g at 77 K was determined for the MOF sample with a pore size of 9.6 Å.

Solid-phase extraction of organic dyes
We have studied the adsorption activity of the sorbent with respect to aqueous solutions of CR and MB according to the method described in (Dzhardimalieva et al. 2020).The effect of pH of the sample solution, amount of adsorbent, adsorption time on the removal efficiency was studied.The results are shown in Fig. 4. The adsorption capacity of the sorbent rapidly increases in the initial period, and then significantly decreases with an increase in the adsorption time.First, diffusion proceeds to the outer surface, which is accompanied by diffusion into the pores of the sorbent because of which equilibrium is quickly achieved.The limiting adsorption according to experimental data is 40 mg/g.

Fig. 4 Dependence of adsorption on time
The temperature dependence of the removal efficiency at pH = 7 is shown in Fig. 5.It turned out that an increase in temperature promotes a more rapid achievement of adsorption equilibrium and the limiting degree of removal is reached within 20 min.Based on the results obtained, it can be concluded that the sorbent effectively extracts the CR dye from an aqueous solution at a temperature of 35 °C, however, at a temperature of 18 °C, the complex proved to be insufficiently effective.On the contrary, we observe the situation when the MB food coloring sorbent is extracted in an aqueous solution.More efficient extraction is observed at a temperature of 18 °C with a short-term increase in the first 10 min, apparently, there is a process of activation of the substance, since after 15 min the concentration of the MB dye sharply decreased in 15 min from 25.8 mg/L to 0.4 mg/L.That is, by 64.5 times in 15 min, and in 60 min, the concentration of the dye decreased by 7500 times, which indicates a high efficiency of the sorbent when extracting the MB dye at room temperature.
The moment of activation is also present during the extraction of the dye at a temperature of 35 °C, but it is shifted and appears after 30 min.At the time point from 30 min to 45 min, the concentration of the food coloring MB in an aqueous solution decreases from 16.8 mg/L to 4.4 mg/L, that is, almost 4 times.This indicates the effectiveness of the use of the sorbent, although not as effective as at room temperature.
The effect of acidity was studied in the pH range 3-11.We found that the degree of MB removal is practically independent of the pH of the medium in this range, while for CR this dependence is quite significant (Fig. 7, 8).(3), 4 -initial solution The adsorption of dyes using the synthesized sorbent is satisfactorily described by a pseudo-first order kinetic model, as shown by fitting the experimental data to the model using least squares regression analysis (Fig. 9).The pseudo-first order adsorption rate constants were calculated graphically and are presented in Table 1.The Langmuir and Freundlich adsorption isotherms of dyes using a sorbent are shown in Fig. 10, 11, and Table 1 shows the coefficients of these isotherms.Most of the R 2 values exceed 0.9 for all isotherm models.The RL values for Langmuir adsorption of the dye on the sorbent were between 0 and 1, indicating favorable adsorption.The KL values (0.93, 2.16, and 3.12 L mg -1 at 283, 291, and 308 K, respectively) indicate an increase of the adsorption of the dye on the sorbent with temperature.The KF and 1/n values increase and decrease with an increase in temperature, respectively.The 1/n values were between 0 and 1, indicating favorable adsorption.The graphs of lnKD versus 1/T (Fig. 12) were used to calcite the thermodynamic parameters of the adsorption of dyes on the sorbent.As you can see from the Table 2, the values of ΔG 0 at temperatures of 283, 291, and 308 K are negative, indicating that the adsorption process was spontaneous.A decrease in ΔG 0 with an increase in temperature testifies about more efficient adsorption at a higher temperature.In addition, a positive ΔS 0 value indicates that the degrees of freedom increase at the solid-liquid interface during the adsorption of dyes on the sorbent and reflect the affinity of the sorbent for dye ions in aqueous solutions and may indicate some structural changes in adsorbents.

Solid-phase extraction of dyes in real samples
While dyes improve the appearance of food products, they have a significant impact on food safety.Therefore, an important problem in the use of dyes in food is the threat to human health due to the accumulation of dyes in food (Yang et al. 2018;Wang et al. 2019).To test the reliability of a sorbent for removing artificial dyes from a food sample, such as the low-alcohol drink "Cherry", experiments were carried out to study the kinetics and isotherm of adsorption.The drink contains the dye carmosine (azorubin) E-122 (Fig. 13) and natural cherry juice.

Fig. 13 Formula of carmosine
We have taken the absorption UV-vis spectrum to identify the dyes that make up the drink.In Fig. 14, an absorption band characteristic of carmosine is noted in the region of 516 nm, and its intensity decreases depending on the time of contact with the sorbent (Wood et al. 2004;McCune et al. 2010).

Solid-phase extraction of dyes in artificial seawater
Today the marine environment is becoming more and more polluted due to the pollution of the marine environment with hazardous dyes.Therefore, we investigated the adsorption characteristics of the sorbent in artificial seawater.The method for preparing artificial seawater belongs to the Khan scheme (Hanaa and Abdalsamad 2020).
The Lyman and Fleming formula (Table 3) was used to prepare an artificial seawater solution by dissolving analytical grade reagents in bidistilled water.The method for studying the adsorption in synthetic seawater is like the above-described method for studying in ideal water.The adsorption characteristics of the sorbent in artificial seawater slightly deteriorate (Fig. 17).Various inorganic salts in artificial seawater have less effect on the adsorption of a low concentration dye solution compared to a high concentration dye solution.The decrease in the adsorption capacity of the sorbent in artificial seawater in comparison with the case of an ideal aqueous solution is only 3-6%.A decrease in the adsorption capacity is due to a decrease in the solubility of dyes after the addition of various inorganic salts to the solution (Asgher and Bhatti 2017) and the competition of SO4 2-, Cl -, F -, etc. ions with dye molecules for unsaturated sites on adsorbents (O'mahony et al. 2002;Qin et al. 2009).

The sorbent as filler for column chromatography for dye separation
Based on the excellent characteristics of adsorption of dyes on the sorbent and considering the previously obtained results (Liu et al. 2019a;Liu et al. 2019b), we used it as the stationary phase of the chromatographic column.The chromatographic column consists of a glass tube, which was packed with a sorbent as a stationary phase.A mixed solution of MB, malachite green and CR at the same concentration was injected into the chromatographic column, respectively.As shown in Fig. 18, CR was adsorbed to the stationary phase for a long time along with the eluent flow, while MB and malachite green passed through the column.In the upper part of the photo there is a concentrated mixture of malachite green and MB, CR sorbed on the sorbent is in the lower part.The results obtained indicate the possibility of dye separation by passing through a chromatographic column, which is easy to see with the naked eye.These experiments highlight the potential of the sorbent as a filler for column chromatography for dye separation.In summary, a ligand-mixed Zr(VI)-based MOF based on terephthalic acid and 1,10-phenanthroline is a promising sorption material for the solid-phase extraction of organic pollutants (synthetic dyes), demonstrating a high efficiency of removal of the target pollutant.The adsorption behavior was investigated considering several important parameters, including adsorption kinetics, isotherms, and initial pH.The results showed that the maximum adsorption capacity of MOF reached 40 mg/g.Moreover, it has been applied in practice in food samples with good characteristics.The excellent performance of this sorbent in artificial seawater makes it a promising material for seawater treatment.It can potentially serve as a column chromatography filler for the separation of dye molecules.

Fig. 1
Fig. 1 Formulas of Congo red (CR), methylene blue (MB) and malachite green (MG) adsorption-desorption isotherms (A) and the pore size distribution curve (B) for MOF

Fig. 5
Fig. 5 Dependence of removal efficiency on time at different temperatures: A -CR, B -MBThe product of solid-phase extraction of CR and the synthesized sorbent luminesces in blue when irradiated with UV light with a wavelength of 385 nm, which may indicate that during the extraction, a π-π interaction occurs between the dye molecule and the adsorbent (Fig.6).

Fig. 7
Fig. 7 Effect of pH on the removal efficiency of CR by the sorbent

Fig. 17
Fig. 17 Removal efficiency of CR and MB from distilled water and artificial seawater at 293 K at a concentration of 20 mg L -1 (A, C) and 40 mg L -1 (B, D)

Fig. 18
Fig. 18 Selective separation of a mixture of dyes (methylene blue, malachite green and Congo red) on a column filled with the studied sorbent

Table 1
Best fit parameters for the dye adsorption onto the sorbent by the Langmuir and Freundlich models