Removal of 2,4-Dichlorophenoxyacetic Acid from Aqueous Solutions Using Al₂O₃/Graphene Oxide Granules Prepared by Spray-Drying Method

Abstract

Within this study, aluminum oxide granules with 0.25%vol. of graphene oxide were prepared by a spray-drying method to make an adsorbent for the 2,4-Dichlorophenoxyacetic acid (2,4-D) herbicide removal from aqueous solutions. The obtained adsorbent was studied using infrared spectroscopy, scanning electron microscopy and Raman spectroscopy. The presence of graphene in the spray-dried powder was confirmed. The adsorption removal of 2,4-D using the obtained material was performed at an ambient temperature by varying the process parameters such as pH and adsorption time. The adsorption of 2,4-D was a monolayer chemisorption according to the Langmuir isotherm pattern and a pseudo-second-order kinetic model. The maximum Langmuir adsorption capacity of the monolayer was 35.181 mg/g. The results show that the Al₂O₃-0.25%vol. GO powder obtained by spray drying is suitable for the production of adsorbents for toxic herbicides.

1. Introduction

Today, natural water reservoirs contain large amounts of pollutants as a result of human economic activities. Xenobiotic substances including pesticides and herbicides used in agriculture, e.g., 2,4-dichlorophenoxyacetic acid (2,4-D), contribute the most to that pollution. Being widespread and highly soluble, 2,4-D and its transformation products have polluted natural water and soil resources [1]. Moreover, 2,4-D was reported to stand behind the cancer-disease development in mammals, as well as endocrine and central nervous system disorders [2,3,4].

The method of adsorption extraction for pollutants from aqueous solutions is widely used for wastewater treatment thanks to the ease of the adsorber design, low cost, mild process conditions and the absence of harmful secondary products [5]. A large list of sorbents was already studied for the 2,4-D removal from aqueous solutions, including magnetically activated carbons [6], a metal-organic framework [7], polymer compounds [8] and graphene oxide aerogel [9].

Sorbents based on aluminum oxide (Al₂O₃) have been widely used in various fields of industry and medicine for a long time due to their beneficial properties [10]. The Al₂O₃ compound combines high mechanical strength, hardness, a developed surface and chemical resistance, while its cost is reasonably low [11,12]. Currently, special attention is paid to adsorbents with alumina, which exhibit very good adsorption properties towards heavy metals [13,14]. In addition, being water-resistant, aluminum oxide is often used as an adsorbent for drying and processing media containing condensed moisture, while the possibility of repeated temperature regeneration by burning ensures the long-term operation of the adsorbent [15].

In recent years, much attention has been paid in the literature to graphene oxide (GO) as a promising adsorbent for organic pollutants due to the presence of oxygen-containing functional groups on its surface, which can act as adsorption centers [16]. However, two-dimensional GO has poor structural stability in aqueous environments, making it difficult to recover and reuse [9,17].

Adsorbents containing a single component often require a functionalization of their surface, such as, for example, increasing the specific surface area, creating additional micro and mesopores, and applying functional groups to improve the adsorption capacity in relation to target pollutants. Such functionalization can be achieved through the additional activation of the raw material, doping the sorbent with various substances [18], and also creating composites.

To enhance aluminum oxide, it was suggested to combine it with graphene oxide by spray drying to obtain a granulated Al₂O₃-0.25%vol. GO adsorbent. The technique provides a spherical shape to the particles that contributes to the adsorption process. Spray drying makes it possible to obtain granules ranging in size from several to tens of microns from highly dispersed and ultra-disperse ceramic powders from solutions or suspensions by drying [19]. Compared to other methods for obtaining carbon nanomaterials such as milling, sintering and crushing, the spray granulation method avoids the agglomeration of GO nanosheets and reduces damage to their structure [20].

Thus, the purpose of this work was to obtain a new sorbent from aluminum oxide with the addition of 0.25% vol of GO and to study its adsorption ability for 2,4-D from aqueous solutions.

2. Materials and Methods

2.1. Materials

For the research, the following raw materials were obtained: 2,4-Dichlorophenoxyacetic acid (Merck, Darmstadt, Germany), NaOH (≥99%), NaCl (≥99%), HCl (37%) (Khimprom, Kemerovo, Russia), α-Al₂O₃ powder with purity 98.9–99.9% and average particle size of 500 nm (Plasmotherm Ltd., Moscow, Russia), polyvinyl alcohol (Ruskhim Ltd., Moscow, Russia) and aqueous suspension of GO with a concentration of 50 mg/mL (Graphenox, Moscow, Russia).

2.2. Adsorbent Preparation

The Al₂O₃-0.25GO sorbent was produced by spray drying from a prepared suspension for granulation. To obtain a suspension, at the first stage, graphene oxide was dispersed in distilled water with a concentration of 0.6 mg/L by the IKA T-18 (IKA-Werke GmbH & Co. KG, Staufen, Germany) ultrasound homogenizer. Next, aluminum oxide and a dispersant (0.01 wt.% Al₂O₃) were added to the suspension to prevent the sticking of Al₂O₃ particles and the formation of large agglomerates. Then, the suspension was placed on a MR Hei-Tec (Heidolph Instruments GmbH & Co. KG, Schwabach, Germany) magnetic stirrer, and a solution of polyvinyl alcohol (0.2% wt.) was added as a binder. The composition was mixed for 6 h at a speed of 400 rpm/min. To demonstrate the effect of graphene oxide on the adsorption properties of the composite, aluminum oxide powder was obtained in a similar way without the addition of graphene oxide (Al₂O₃-0GO).

2.3. Characterization of Adsorbent

Microstructural analysis was performed using scanning electron microscope Vega 3 LMH (SEM, Tescan, Brno, Czech Republic). In order to avoid electron charging and improved imaging, samples were coated by electrically conductive thin gold film using Q150T Emscope (Quorum Technologies Ltd., Newhaven, UK) sputter coater. Raman spectroscopy of the produced samples was carried out using DXR ^TM2 microscope (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a 780 nm laser with a power of 15 mW. The laser beam was to pass through 50× optical zoom lens projecting into 0.8 μm spot on the examined area with an integration time of about 6 s for each spectrum. Fourier transform infrared spectroscopy (FTIR) spectra were recorded using Vertex 70 (Bruker Optik GmbH, Ettlingen, Germany) spectrometer. To calculate the specific surface area S_BET, m²/g for the Al₂O₃-0.25GO specimen, a Brunauer–Emmett–Teller (BET) analysis was performed using Autosorb iQ Station 2 gas sorption analyzer (Quantachrome Instruments, Boynton Beach, FL, USA). The point of zero charge (pHzpc) of the adsorbent resulted from the pH drift method, described in [21], using S 210 pH meter (Mettler-Toledo, Columbus, OH, USA).

2.4. Adsorption

The adsorption of 2,4-D from aqueous solutions onto Al₂O₃-0.25GO and Al₂O₃ was carried out in thermostatic cells at a temperature of 25 °C with continuous mixing. To study the kinetics of adsorption, equal portions of the adsorbent weighing 25 mg were added to 100 mL of 2,4-D solution with concentrations of 10–200 mg/L. During the experiment, samples were taken from prepared adsorbate solutions with added adsorbent at specified time intervals. Adsorbate concentration in samples was determined using UV spectroscopy at 283 nm on a Hitachi U-1900 instrument (Tokyo, Japan). The number of ions of 2,4-D adsorbed per 1 g of the adsorbent (mg/g) was calculated according to Equation (1) below:

$$q_e=\frac{\left(C_0-C_e\right)·V}{m},$$

(1)

where $C_0$ and $C_e$—initial and equilibrium concentration of each adsorbate ion, respectively; mg/L; $V$—volume of the solution, l; $m$—the mass of the adsorbent, g.

Preliminary experiments showed an insignificant adsorption ability of Al₂O₃-0GO. Thus, the calculation of the isotherms and kinetics of the adsorption process on aluminum oxide powder without the addition of graphene oxide was not considered in this work. All adsorption experiments were performed in triplicate, with the average values used for calculations.

3. Results and Discussion

3.1. Characterization of the Adsorbent

The calculated specific surface area of the sorbent was 6.852 m²/g, the total pore volume—0.043 cm³/g and the average pore diameter—25.32 nm. The microstructure of the sorbent is shown in Figure 1. The particles were almost spherical, with sizes ranging from 1 μm to 20 μm (Figure 1A). A microporous structure and a developed surface were also observed (Figure 1B).

Since the low content of GO in granules makes it hard to detect using SEM, it was identified using Raman spectroscopy. The Raman spectrum of the sorbent is given in Figure 2.

The results show that the spectrum has pronounced peaks of D and G bands with wave numbers of 1312 cm⁻¹ and 1592 cm⁻¹, respectively, which is typical for GO. This pattern is also common for carbon-containing materials. The G band is attributed to a perfect graphite structure while the D band shows defects in the microstructure. Therefore, after the spray drying, the sorbent retained the GO structure, which is consistent with the literature [22,23,24].

Figure 3 shows the FTIR spectra of the Al₂O₃-0GO and the Al₂O₃-0.25GO specimens. Strong absorption bands at 600 cm⁻¹ and 447 cm⁻¹ can be attributed to stretching vibrations of the Al-O bond [25]. The Al₂O₃-0.25GO specimen demonstrates characteristic bands 1215 cm⁻¹ and 1280 cm⁻¹, confirming the presence of C-O-C bonds and the C-O stretching of an epoxy group [26]. The 1660 cm⁻¹ peak can be identified as the in-plane vibrations of the skeletal C=C band of the hexagonal aromatic ring from the graphene sheet [25]. The peak at 3406 cm⁻¹ is connected to the OH stretching of carboxylic groups. The characteristic band at 1760 cm⁻¹ can be associated with the C-O stretching of the carboxylic group [26].

3.2. Adsorption Studies

3.2.1. Effect of pH

The pH of the solution is one of the most important parameters influencing the process of interaction of the adsorbate with the adsorbent. When the pH changed in the range from 2 to 10, the interaction between 2,4-D and the adsorbent also changed. The pH_PZC value was determined to characterize the surface charge of the adsorbent. It describes the electrically neutral environment of the sorbent surface at a certain pH of the solution. For an adsorbent, pH_PZC was found to be 7.6, which means that the total surface charge of the adsorbent will be positive at pH > 7.6 and negative at pH < 7.6. 2,4-D is a weak acid and is present in solution mostly in its anionic form. Based on pH_PZC data and the resulting dependence presented in Figure 4, it is shown that the pH region most effective for adsorption is in the range of values from 2 to 4.

The obtained correlation and the pH_PZC concept suggest that electrostatic interactions are one of the main adsorption mechanisms of 2,4-D on that sorbent. However, a small adsorption capacity of Al₂O₃-0.25GO to 2,4-D was observed in an environment with a pH higher than the pH_PZC value, which may be due to interactions such as π-π stacking between the aromatic rings of 2,4-D and GO or hydrogen interactions with oxygen-containing groups on the surface of the sorbent.

3.2.2. Adsorption Kinetics

Adsorption kinetics were evaluated by applying nonlinear pseudo-first-order (PFO) [27], pseudo-second-order (PSO) [28] and Elovich [29] kinetic models. The models are usually taken to evaluate rate constants and adsorption rates. The kinetic parameters of the nonlinear PFO and PSO models were calculated using Equations (2) and (3), respectively:

$$q_t=q_e·\left(1-ex{p}^{-k_{1·}t}\right),$$

(2)

$$q_t=\frac{q_e^2·k_2·t}{1+q_e^2·k_2·t},$$

(3)

where $k_1$(min⁻¹) and $k_2$ (g/(mg∙min)) are the rate constants for PFO and PSO, respectively, and $q_e$ and $q_t$ (mg/g) are the number of adsorbed ions of 2,4-D at equilibrium and at time $t$, respectively.

The Elovich model is b is expressed in the following way (4):

$$q_t=\frac{1}{\beta }\ln \left(\alpha \beta t+1\right),$$

(4)

where α (mg/g min)—initial adsorption rate; β (g/mg)—desorption constant during each experiment.

Figure 5 shows the results of the nonlinear models of a pseudo-first-order, pseudo-second-order and the Elovich kinetic model for the adsorption of 2,4-D on Al₂O₃-0.25GO, and

Table 1. Kinetic models for the removal of 2,4-D.

C0, mg/L	Pseudo-First-Order			Pseudo-Second-Order			Elovich
	k1	qe1, mg/g	R2	k2	qe2, mg/g	R2	α, mg/(g⋅min)	β, g/mg	R2
25	0.023	25.19	0.989	0.001	28.914	0.998	1.84	0.15	0.988
50	0.026	28.189	0.993	0.001	32.173	0.999	2.393	0.15	0.982
75	0.027	30.092	0.993	0.001	34.21	0.999	2.787	0.15	0.982
100	0.036	30.622	0.984	0.002	33.888	0.999	6.89	0.18	0.988

contains the variables of the kinetic models.

The best fit to the kinetic model of 2,4-D adsorption on Al₂O₃-0.25GO was shown using the PSO model (R² > 0.99). This indicates that the adsorption-limiting step involves inter- and intramolecular interactions between 2,4-D and the adsorbent. In this case, the rate of filling the adsorption centers on the Al₂O₃-0.25GO surface is proportional to the number of unoccupied centers [30]. Although the Elovich kinetic model correlated less with experimental 2,4-D adsorption data, the β variable could be used to confirm the occurrence of inter or intramolecular interactions between the adsorbate and the adsorbent. The Elovich model assumes that real solid surfaces are energetically inhomogeneous, while the desorption process and the interaction between the adsorbed particles do not have a significant effect on the adsorption kinetics at low surface coverage. The α and β coefficients of the model represent the initial adsorption rate (mg/(g·min)) and the desorption coefficient (g/mg), respectively. The α coefficient was >1 mg/(g·min) over the entire concentration range, while the β coefficient varied from 0.15 to 0.18 g/mg. This indicates a low desorption rate due to the effective interaction of 2,4-D and Al₂O₃-0.25GO [31].

3.2.3. Adsorption Isotherms

The three widely used models of Langmuir [32], Freundlich [33] and Temkin [34] were used to describe the adsorption isotherms of 2,4-D on Al₂O₃-0.25GO. The simulation of adsorption isotherms provides insight into the interaction of the solute with the adsorbent and the nature of the adsorption process. These models show the dispersion behavior of the adsorbed fragments in the aqueous and solid phases when adsorption reaches equilibrium.

The nonlinear form of the Langmuir isotherm could be represented by Formula (5):

$$q_e=q_m{b}_l{c}_e/\left(1+b_L{c}_e\right),$$

(5)

where $q_m$—maximum capacity of the monolayer (mg/g), and $b_l$—the adsorption coefficient (L/g). The Langmuir model assumes a single-layer coating of 2,4-D on the surface of the adsorbent.

According to Freundlich’s theory, at a constant temperature, the amount of adsorbed solute per unit mass of the adsorbent is proportional to the equilibrium concentration of the substance adsorbed from the solution, raised to a certain power that is less than their unity. Mathematically, the Freundlich isotherm can be presented as in Equation (6):

$$q_e=K_F{c_e}^{1/n},$$

(6)

where $K_F$—the coefficient of distribution or adsorption coefficient (L/g).

The Temkin isotherm model takes into account the adsorbate–adsorbent interaction and assumes that the heat of the adsorption of all the molecules in the layer will decrease linearly as adsorbed molecules accumulate on the surface of the adsorbent [35]. The isotherm was calculated according to Equation (7).

$$q_e=\left(RT/b_T\right)\ln \left(A{c}_e\right),$$

(7)

where $b_T$—the adsorption coefficient (J/mole); R—the universal gas constant of 8.314 J/(mol∙K); A—the constant, L/g; T—the absolute temperature (K).

Figure 6 shows the results of applying the isotherm models to the experimental data of the adsorption 2,4-D on Al₂O₃-0.25GO.

Comparing the obtained R² values of the applied models, the Langmuir and Freundlich equations describe the experimental data the best. This indicates that during the adsorption of 2,4-D, both chemical and physical interaction processes occur between the adsorbate and the adsorbent [35]. It is clear that the Langmuir isotherm fits the isotherm data best, with the highest R² value of 0.996. Thus, based on the data presented in

Table 2. Isotherm parameters for 2,4-D.

Model Parameters		Calculated Values
Langmuir	qm, mg/g	35.181
	bl, L/g	0.109
	R2	0.996
Freundlich	KF, L/g	10.467
	1/n	0.256
	R2	0.891
Temkin	bT, J/mol	373.848
	A, L/g	1.601
	R2	0.984

, it can be concluded that the adsorption of 2,4-D on the Al₂O₃-GO sample followed the Langmuir isotherm and was monolayered.

Table 3. Comparison of capacity values for 2,4-D ions absorbed by different adsorbents.

Adsorbent	qe, mg/g	Equilibrium Time, Min	Ref.
[Co–Al–Cl] layered double hydroxide	20.51	60	[36]
Glutaraldehyde-crosslinked chitosan	252.55	-	[37]
Reduced graphene oxide	270.1	1440	[38]
Modified tiger-nut residue	90.2	120	[39]
Organophilic clay	6.45	6	[40]
Al2O3-0.25GO	35.18	250	This work

shows comparative data on adsorption capacity, specific surface area and time to establish the adsorption equilibrium with respect to 2,4-D for different sorbents described in the literature. It shows that the suggested Al₂O₃-0.25GO composite demonstrates a satisfactory adsorption capacity for the target component.

4. Conclusions

In the present study, graphene oxide alumina granules were shown to have good adsorption capacity towards the pollutant herbicide 2,4-D. Experimental adsorption data were consistent with the Langmuir isotherm and the PSO model. Solution pH played a key role in adsorption efficiency, which may imply the predominance of electrostatic interactions in the adsorption process. 2,4-D was adsorbed most effectively at pH 2 to 4. However, adsorption was also observed at higher pH values, which may indicate that several mechanisms are involved in adsorption, such as, for example, π-π stacking or hydrogen bonds.