Antibiotics are among the most significant discoveries of humanity, having saved countless lives and improved the quality of life for many people by significantly changing the way we treat a wide range of infectious diseases. However, the widespread use of antibiotics also leads to environmental pollution, including surface and groundwater, bottom sediments, and soils, where they act as toxins on native microorganisms [1].

Antibiotics used for medical purposes are not always completely absorbed by a human organism, and their active ingredients, as well as their metabolites, end up in wastewater and, after passing through treatment plants, can enter aquatic ecosystems [2]. Medical institutions and pharmaceutical companies are major sources of antibiotic pollution in wastewater [3]. Using wastewater for irrigation and sewage sludge as fertilizer can lead to the contamination of agricultural soil with antibiotics. Soil pollution is also caused by the extensive use of antibiotics in livestock farming and the use of animal waste (urine and manure) as fertilizers [4,5]. It has been estimated that the amount of antibiotics accumulated in agricultural soils can reach one kilogram per hectare or more, comparable to the concentrations of pesticides [6]. The migration of these antibiotics through food chains can lead to their accumulation in the bodies of both farm animals and humans.

The main classification schemes for antibiotics are based on their molecular structure, character of action, and activity spectrum. Antibiotics within a structural group generally exhibit similar patterns of action, toxicity, and potential side effects. The following main groups of antibiotics are distinguished based on their chemical or molecular structure: beta-lactam antibiotics, tetracyclines, macrolides, quinolones, aminoglycosides, sulfonamides, glycopeptides, and oxazolidinones [7].

Within the framework of the concept of sustainable development, an important issue is the development of materials and technologies that can minimize the risk of water and soil contamination with antibiotics and reclaim already contaminated sites [8]. Industrial techniques for cleaning water based on physical and chemical processes, such as oxidation and UV irradiation, show promising results in removing organic pollutants. However, these methods can be expensive and have side effects that are difficult to predict. Additionally, techniques such as sedimentation, coagulation–flocculation, ion exchange, solvent extraction, membrane filtration, reverse osmosis, and ozonation can be used to purify water from organic pollutants [9,10,11,12].

Adsorption methods based on porous materials are a simple and efficient way to remove organic pollutants from water [13]. Common sorbents such as zeolites, activated carbon, graphene oxide, hydrogels, and metal–organic frameworks can be used to adsorb antibiotics from water [14,15,16,17,18]. However, these sorbents have some disadvantages, including not always being highly efficient and specific in their adsorption, having high costs, and being difficult to regenerate or dispose of.

Sorption materials based on clay minerals, on the other hand, are economical, abundant, have a large specific surface area (SSA), and a high cation exchange capacity (CEC) [19,20,21,22]. They are also physically and chemically stable, non-toxic, and have the ability to desorb pollutants after use, making them effective sorbents for purifying water from antibiotics [23].

Although clay minerals are important in many environmental processes, especially in controlling the migration of molecules and ions in soil, they are also used as materials to create various substances that solve other environmental problems [24,25]. Modification of the properties of these minerals to increase their adsorption capacity for organic pollutants is often achieved by chemical methods [20].

Currently, active research is being conducted into the composition, properties, and adsorption capacity of natural clay minerals as potential sorbents of various pollutants. This research focuses on their ability to bind antibiotics, as demonstrated by a large number of publications on Sciencedirect, using the keywords “Antibiotics adsorption by clay minerals” (Figure 1).

This review attempts to summarize the main patterns of antibiotic adsorption by various natural clay minerals based on the structures and characteristics of both adsorbents and adsorbates. The main properties of the adsorption solution that affect the interaction between antibiotics and natural clay minerals are described.

Clay minerals are secondary hydrous silicates, aluminosilicates, and ferrosilicates with a standard structure consisting of layers comprising a tetrahedral silicate sheet bonded to an octahedral aluminum hydroxide sheet. In each unit of the tetrahedral sheet, one Si atom is connected to four O atoms or OH groups in a tetrahedral configuration.

The arrangement of tetrahedral units in the sheet results in the formation of a hexagonal network characterized by the chemical composition Si₂O₆(OH)₄. The octahedral sheet is composed of Al or Mg and Fe atoms surrounded by a closely packed six OH groups or O atoms in an octahedral configuration (Figure 2). In the presence of Al atoms, the sheet is characterized by the chemical composition Al₂(OH)₆. When an Al³⁺ ion with a charge of 3+ is present in an octahedral sheet, only 2/3 of the possible positions are filled to balance the charges, and the mineral is called dioctahedral. In the presence of a Mg²⁺ ion with a charge of 2+, all three positions are filled to balance the structure, and the mineral is called trioctahedral.

A Si tetrahedral sheet and an Al octahedral sheet are connected together by sharing an apical O or OH to form a layer of a clay mineral. There are two layer types. Layers consisting of one tetrahedral sheet and one octahedral sheet form a 1:1 (or tetrahedrooctahedral, TO) layer type. The second type of layer includes two tetrahedral sheets that are on either side of an octahedral sheet—this is the 2:1 (or TOT) layer type. Each clay mineral layer has two basal surfaces. In 2:1 layer-type clay minerals, these surfaces are siloxane surfaces (i.e., planes of basal O atoms) (Figure 2). In TO-type clay minerals, one basal surface is a siloxane surface, and the other is a plane of protonated octahedral O atoms. The ratio of tetrahedral and octahedral sheets in the structural layer of the mineral is the basis of the most common classification of phyllosilicates (Table 1).

By isomorphic substitution, the structural cations in the tetrahedral and octahedral sheets with low-valence metal ions, the surfaces of clay minerals, especially the 2:1 layer type, become negatively charged. In the minerals with dioctahedral structure, the most frequently encountered substitutions are of Si⁴⁺ by Al³⁺ in the tetrahedral sheets and of Al³⁺ by Mg²⁺, Fe²⁺, or Fe³⁺ in the octahedral sheets. In addition, the presence of vacant octahedral sites in trioctahedral minerals, the presence of trioctahedral units in dioctahedral minerals, and partial dehydroxylation of the octahedral sheets as a result of the oxidation/reduction of structural octahedral iron can affect the charge of the mineral layer [27]. The layer charge is not necessarily distributed spatially uniformly in the layer, since the localization of isomorphic substitutions can be ordered, clustered, or randomly distributed [28]. In addition, the pH of the solutions in which clay suspensions are formed can alter the charges at the edges of clay particles [29]. It is believed that the basal oxygen atoms on the siloxane surface cannot be protonated in the range of water pH, although experiments by Gupta and Miller (2010) may suggest otherwise [30]. Conversely, the Al₂-OH functional groups on the octahedral basal surface of kaolinite have a well-established ability to gain or lose protons when exposed to liquid water that contributes to the observed pH-dependent surface charge of kaolinite [31,32]. The total negative charge of the layer is balanced by the presence of exchangeable, mainly alkaline (Na⁺, K⁺) and alkaline earth metals (Ca²⁺, Mg²⁺) on the basal surface of the minerals. Clay minerals of the 1:1 layer type (TO), including kaolinite, have a layer charge close to zero.

The stacking of up to tens of clay mineral layers forms a clay mineral particle. From crystallographic data, it can be determined that the distance between the planes of oxygen atoms on opposite surfaces of the layers is 6.54 Å for montmorillonite and 4.5 Å for kaolinite. Clay minerals generally have high aspect ratios with variable morphologies: kaolinite and well-crystallized illite tend to have hexagonal and elongated hexagonal morphologies, respectively, while montmorillonite and less well-crystallized illite have mostly irregular platy or lath-shaped morphologies [33]. When packed together, clay mineral particles form aggregates and their external surfaces delineate interparticle spaces. Assemblies of aggregates delineate interaggregate spaces [34].

Due to their unique structure, clay minerals have the highest specific surface area among known minerals in nature [33]. The total specific surface area of smectite minerals can be 450–850 m² g⁻¹, while the outer specific surface area of smectite group minerals reaches 40–70 m² g⁻¹ [35,36]. Figure 3 shows the structures of various clay minerals, which reflect the different arrangements of atoms and ions within the crystal due to chemical changes or thermal fluctuations that occur during weathering.

The surface properties of the minerals consist of the surface properties of the basal surfaces and the surface properties of the edges. As already noted, the negative charge of the layer, which arises mainly as a result of isomorphic substitutions, is compensated primarily by cations located in the interlayer space or on the outer basal surfaces. Unless these cations are in the closed interlayer spaces of non-swelling clay minerals, they exchange fairly easily with other cations in solution. In swelling clay minerals, cation exchange occurs simultaneously on both the internal and external basal surfaces. The size of the clay particle and the value of the interlayer distance can influence the ion exchange process [38].

In non-swelling clay minerals, the properties of the layers that form the outer basal surfaces of the clay particle may differ from those of the layers in the bulk of the particle [33]. However, this heterogeneity has little or no effect on the specific surface area but may strongly influence the surface charge and hence the surface reactivity of the mineral particles.

The negative charge of the basal surfaces of clay minerals is compensated by the adsorption of exchangeable cations and the exclusion of anions near the mineral surface in the electrical double layer (EDL). Measurements of anion exclusion and electrophoretic mobility in aqueous suspensions of clay particles indicate that the EDL is in the order of a few nanometers in thickness and is strongly dependent on the ionic strength [39]. The EDL can be theoretically divided into a Stern layer containing inner- and outer-sphere surface complexes and a diffuse layer containing ions interacting with the surface via long-range electrostatic forces [40,41]. The composition and structure of the diffuse layer have been intensively studied since they influence many macroscopically observable phenomena, including swelling, osmosis, and particle aggregation [42].

The edge structure of clay minerals has not been fully established and is apparently quite diverse. Traditionally, the edges have been assumed to have a TO or TOT structure similar to the bulk structure, with the surface elements (Si, Al, and their substitution atoms) having the same coordination as in the internal structure and with the outer oxygen atoms being undercoordinated [43]. However, a number of experimental and modeling results suggest that the octahedral and tetrahedral elements at the edges of clay layers may adopt a different stoichiometry and coordination than in the bulk mineral structure, at least under some conditions [44]. The edges of clay minerals carry a pH-dependent surface charge that arises from the acid–base properties of the functional groups present at the edges of the clay layers. The edge surfaces of clay minerals can bind inorganic or organic cations, anions, and molecules through short-range interactions [45,46]. Non-specific interactions that compensate for the charge of the edge surface can also be observed [39,47].

The overall negative surface charge on clay minerals, despite the presence of a number of pH-dependent positive charges at edges and cleavages, leads to preferential adsorption of hydrophilic, positively charged particles on clay minerals. The anionic adsorbing capacity of clay minerals is low: less than 5 cmol kg⁻¹ for smectites [48] and not more than 2 cmol kg⁻¹ for kaolinites [49]. Thus, a key factor in the adsorption of antibiotics by clay is the hydrophilicity of organic molecules, which can react with the negatively charged surface of the mineral.

The distribution coefficient in the octanol–water system (K_ow or P) is an indicator of the hydrophilicity and hydrophobicity of a substance. It is defined as the ratio of the concentration of the substance in n-octanol, a lipophilic solvent, to its concentration in water, a hydrophilic solvent, under equilibrium conditions. Higher K_ow values indicate a more hydrophobic substance, meaning it is more soluble in lipophilic media and less soluble in water.

For example, for tetracyclines, the log K_ow values are −2.2 to −1.3 [50], and the solid/water distribution coefficients (K_d) are 300–2000 L kg⁻¹ [51]. The low K_ow values of tetracyclines suggested that the antibiotic molecule is hydrophilic with a high water solubility. A high affinity of tetracyclines for clay minerals was shown in many studies [52,53,54]. Kulshrestha et al. [53] reported a high adsorption capacity of oxytetracycline on montmorillonite of 800 mg g⁻¹ at pH 5. Parolo et al. [54] obtained a tetracycline adsorption capacity of 0.9 mmol g⁻¹, corresponding to 400 mg g⁻¹, at pH 3 on a bentonite. Our own experiments show that the hydrophilic basal surfaces of clay layers of montmorillonite and kaolinite interact weakly with hydrophobic molecules of the antibiotic amoxicillin (log K_ow = 0.87) and more effectively adsorb hydrophilic oxytetracycline molecules (log K_ow = −0.90). Therefore, in aqueous solutions, molecules with negative K_ow will have strong interaction with clay minerals, resulting in their adsorption on the surface and intercalation in the interlayers of the swelling minerals.

In the presence of a negative charge on the mineral surface, two important factors that affect the adsorption of antibiotics are the specific surface area and the cation-exchange capacity of the clay mineral. These values may depend on the deposit of an individual mineral, the characteristics of sample preparation (grinding), and the method of determining the parameter (using various swelling-inducing agents or adsorbents that do not cause swelling).

The specific surface area of certain minerals determined by the N₂ gas adsorption method (BET) is shown in Table 2.

SSA data obtained using other methods may have different absolute values, but they are generally related to each other. In this study, the specific surface areas of illite and montmorillonite minerals were determined using atomic force microscopy (AFM) and compared to SSA values obtained by N₂ gas adsorption (BET) and liquid adsorption using ethylene glycol monomethyl ether (EGME).

For illite, the SSA was estimated to be 41 ± 3 m²·g⁻¹ by BET and 83 ± 5 m² g⁻¹ by AFM. For montmorillonite, BET estimated a SSA of 61 ± 2 m² g⁻¹, while the analysis of AFM images yielded a much higher mean SSA of 346 ± 37 m² g⁻¹. The authors suggest that the sample preparation for AFM imaging may have caused delamination of clay mineral particles. The specific surface area estimated by EGME for illite was 112 m² g⁻¹ and 475 m² g⁻¹ for montmorillonite, which is about 30%–40% greater than the AFM values [56].

There is a significant difference in the cation exchange capacity (CEC) of various clay minerals. For example, the CEC of montmorillonite typically ranges from 70 to 130 meq 100 g⁻¹ of soil, while that of illite ranges from about 20 to 40 meq 100 g⁻¹, and that of kaolinite ranges from 3 to 15 meq 100 g⁻¹. The CEC values for halloysite, hectorite, palygorskite, sepiolite, and vermiculite are 5–10 meq, 80–130 meq, 30–40 meq, and 30–40 meq 100 g⁻¹, respectively [57]. Chemical composition of the minerals, specific surface area, and acidic and alkaline environments are the main factors affecting CEC of clay minerals. Three-layer swelling minerals have the highest specific surface area and cation-exchange capacity. Below is a table of adsorption models for some antibiotics on clay minerals, both modified and native .