Factors influencing the adsorption of antibiotics onto activated carbon in aqueous media

For a long time, antibiotics have significantly contributed to the health and longevity of mankind, domestic, and aquatic animals (Mishra et al. 2012; Ben et al. 2019; Cycon et al. 2019). Several infectious diseases have been cured using this group of medicines. However, the use of antibiotics beyond the need and improper disposal practices thereof has resulted in various repercussions. The development of resistance in microorganisms towards antibiotics is one of the outcomes (Diwan et al. 2009). Increasing usage of antibiotics among humans/animals and their inappropriate disposal have resulted in increased presence of antibiotics in the environment through excreta, hospital wastewater discharge, solid waste etc. (Diwan et al. 2009, 2010; Duan et al. 2019; Bilal et al. 2020; Parashar et al. 2022). Over time, the consistent presence of small amounts of antibiotics in aqueous media has resulted in the development of resistance in microorganisms towards antibiotics. Considering the public health aspects, this phenomenon is challenging to deal with as once the microbes become resistant, these become difficult to treat. Therefore, suitable technologies are needed in order to remove antibiotics from environmental matrices.

Pore size and specific surface area of the activated carbon (adsorbent) play an important role in determining the adsorption of antibiotics. The size of the adsorbent's pores may be in the range of micropores (<2 nm), mesopores (2–50 nm), or macropores (>50 nm) (IUPAC 1972). Moreover, pore size and specific surface area are the interrelated terms, as it has been seen that microporous substances generally have more surface area compared to the mesoporous or macroporous substances (Ji et al. 2010). Further, the higher the surface area, the more would be the adsorption and vice-versa. Adsorption patterns were studied for sulfamethoxazole, tetracycline, and tylosin over activated carbons, microporous and mesoporous carbons, and graphite. The lowest antibiotic adsorption was reported over graphite owing to the smallest surface area (Ji et al. 2010). Nevertheless, the pore-filling mechanism is also important to consider, as the highest adsorption is accomplished when the pore size of the adsorbent (activated carbon) and molecular size of the adsorbate (antibiotic) matches well (Ismadji & Bhatia 2001).

Activated carbon is synthesized using various activating agents, which may be acidic, alkaline, or neutral in nature. These agents affect the carbon yield (Table 1) as well as physico-chemical properties of the activated carbon significantly, which in turn affect the adsorption characteristics (Gao et al. 2020). For example, activated carbon synthesized from mango kernel resulted in 47.3, 45.7, and 8.4% carbon yield, upon using the H₃PO₄, ZnCl₂, and KOH as the activating agents, respectively (Andas & Wazil 2019). Thus, the same precursor may result in varied carbon yields depending upon the activating agent chosen (Table 1). Activating agents also affect the pore size, specific surface area, and pore distribution in the synthesized activated carbon. It has been found that alkaline activating agent, such as KOH, results in high specific surface area, which renders excellent adsorption characteristics (Gao et al. 2020). Further, KOH activation resulted in microporous activated carbon while H₃PO₄ activation resulted in mesoporous carbon (Heidari et al. 2014). Nevertheless, morphological features and chemical properties of the activated carbon are also influenced by the activating agents. Chiu & Lin (2019) found that acidic activating agents usually result in granular activated carbon while alkaline agents result in powdered activated carbon. This happens due to the different activation mechanisms of the agents applied. Where alkaline activating agents undergo dehydration reaction, water gas reaction, and formation of potassium containing constituents; activating mechanisms of the acidic agents can be attributed to the accelerated carbonization, swelling role, dehydration and elimination, oxidation, aromatic condensation etc. Pore forming mechanisms of the neutral activating agents, on the other hand, include oxidation, template, and gasification (Gao et al. 2020). Similarly, surface chemical composition and elemental distribution also get affected by the activating agents (Singh et al. 2019). While acidic activating agents result in acidic surface functional groups, alkaline agents result in basic surface groups. These properties of the synthesized activated carbon do affect the adsorption characteristics and therefore pose significant effects on the antibiotics’ adsorption as well.

The type and amount of functional groups present over the activated carbon surface influence the adsorption behavior for adsorbate. In an adsorption study, where activated carbon was used for the adsorption of bisphenol A (BPA), it was found that phenolic groups resulted in better adsorption capacity compared to carboxyl group present over the surface of activated carbon (Liu et al. 2017a). It was so because the carboxyl group reduced the ability of carbon to act as π-donor which led to the reduction in adsorption. Though BPA is not an antibiotic, similar effects may be seen in the case of antibiotics as well. The role of surface functional groups over the adsorption of triclosan antibiotic was studied by Fang et al. (2009). It was found that the presence of oxygen groups on the activated carbon reduced the adsorption, while nitrogen functional groups resulted in enhancement. However, the possible reason for this could not be elucidated. Influence of functional groups present on the surface of activated carbon was also studied for the adsorption of amoxicillin antibiotic, where the standard activated carbon was compared with the NH₄Cl induced activated carbon. The NH₄Cl induction resulted in higher adsorption of amoxicillin in terms of rate as well as capacity (Moussavi et al. 2013). The possible reason for the better performance of NH₄Cl induced activated carbon was attributed to its high monolayer adsorption capacity due to the increased interactions between adsorbate (amoxicillin) and adsorbent (activated carbon). The increased interactions were due to the presence of more surface functional groups and higher density of charge in the NH₄Cl induced activated carbon (Moussavi et al. 2013). Similarly, Yu et al. (2020) studied the adsorption of amoxicillin and cephalexin on to the activated carbon derived from Phragmites australis. The un-doped activated carbon obtained from Phragmites australis showed carboxyl, phenolic, and carbonyl groups. However, upon doping it with powdery puffed waterfowl feather (a common protein waste), nitrogen-containing functional groups were introduced, rendering better adsorption owing to the presence of more surface functional groups and more microporous structure (Yu et al. 2020). Adsorption of sulfamethazine was also compared over natural and modified activated carbon (Liu et al. 2017b). Experiments demonstrated that Fe³⁺ treatment of the activated carbon increased the number of surface oxygenic functional groups and zeta potential which resulted in a significant increase in the adsorption of sulfamethazine. The possible adsorption mechanisms included the hydrogen bonding, π-π electron donor-acceptor, and coordination interactions (Liu et al. 2017b).

Aging is one of the important parameters which affect the adsorption of compounds over activated carbon. Adsorption of organic vapors over the aged activated carbon was found to be less compared to the un-aged carbon (Amitay-Rosen et al. 2015). Effect of aging was also studied by Lompe et al. (2018). It was demonstrated that micropollutants, such as sulfamethaxazole, resulted in 10 times reduced adsorption on aged activated carbon (90 days), compared to the virgin carbon. Generally, aging enhances the affinity of water vapor to the carbon, i.e. hydrophilicity. Aging of the activated carbon thus affects the hydrogen-bonding capacity and interaction between water and adsorption sites on the activated carbon surface, which alters the adsorption properties and ultimately reduces the adsorption of adsorbate (Amitay-Rosen et al. 2015). Aging also results in blocking of pores/active sites on to the activated carbon surface which leads to reduced adsorption (Brown et al. 1989; Beck et al. 2002).

Moisture content in the activated carbon may also lead to the reduction in its adsorption capacity (Jonas et al. 1985; Werner 1985; Abiko et al. 2010). As of now, there are no reports about the effects of moisture content over the antibiotics’ adsorption on to the activated carbon; however, this effect has been seen for many other compounds. In the case of organic vapors adsorption, it was found that moisture content reduces the adsorption capacity of coconut shell derived activated carbon (Abiko et al. 2010). Similarly, adsorption of chloroform onto the carbon was also seen to be decreasing when the relative moisture content was above 40% (Jonas et al. 1985). This may have happened due to its insolubility/low solubility in water which does not allow it to hydrolyze, thus reducing adsorption. It has been seen that hydrophobic adsorbates are affected more due to moisture content compared to the hydrophilic compounds (Werner 1985). A decreasing trend due to moisture content was also observed in the case of vapor phase adsorption of trichloroethylene on to the activated carbon (Werner 1985). There are certain factors which influence the adsorption phenomena with respect to moisture, viz. preconditioning of the activated carbon, adsorbate concentration, type of adsorbate compound, and type of activated carbon (Werner 1985). Therefore, it is necessary to consider the effect of moisture while designing and operating the adsorption process for the removal of antibiotics. Further, as the available studies of the effect of moisture are mostly about the vapor phase adsorption, there is ample scope to investigate the impact of moisture on the solid/liquid phase (antibiotics) adsorption.

Surface properties of the antibiotics, such as charge, polarity, hydrophilicity/hydrophobicity etc., may have implications on its adsorption properties. For example, aminoglycoside antibiotics are basic, strongly polar, and hydrophilic. Similarly, tetracyclines are polar and water soluble. Sulfonamides are weakly basic and acidic polar molecules with high water solubility (Ozumchelouei et al. 2020). All these properties do affect the interaction with the activated carbon which itself may be acidic/alkaline/neutral in nature. According to a study carried out by Ridder et al. (2011), when the negatively charged granular activated carbon was employed for the adsorption of pharmaceuticals, the positively charged pharmaceuticals were adsorbed to a much greater extent (32–98%) rather the negatively charged pharmaceuticals (0–58%). It was also inferred in the study that charge effects are more distinct in pure water compared to natural water owing to absence of any ion in pure water to shield this effect.

Solubility of the antibiotics in aqueous solutions is one of the important requisites for effective adsorption. It has been seen that solubility is dependent on the properties of solute (antibiotics), properties of the solvent (water or any other solvent), and solute-solvent interactions. Varanda et al. (2006) and Caco et al. (2008) studied the solubility aspects of antibiotics in various solvents, viz. water, ethanol, 2-propanol, and acetone. The studied hydrochloride forms of the antibiotics, viz. tetracycline hydrochloride, moxifloxacin hydrochloride, and ciprofloxacin hydrochloride, were found to follow the solubility order as water > ethanol > 2-propanol > acetone (Varanda et al. 2006). Another important finding was reported as solubility increased upon increasing the temperature for all the solvents studied. However, when the same antibiotics were studied in their non-hydrochloride forms (or in the base forms), the solubility decreased by two orders of magnitude (Caco et al. 2008). In the basic forms, both ciprofloxacin and tetracycline showed poor solubility in water. The possible cause of this phenomenon was elucidated to be the presence of hydrochloride group (•HCl) which renders the formation of ionic species in water and protonation of amine groups. The increased ions and protonation in the solvent enhances the ion-dipole interaction, thus increasing the solubility (Caco et al. 2008).

Saturation of the adsorption process is determined by the initial concentrations of adsorbate and adsorbent in the medium. Usually, high initial concentration of adsorbate tends to lower the removal efficiency, while high initial concentration of adsorbent results in high removal efficiency. This depends upon the time taken in saturation of adsorption sites (Ahmed et al. 2015). High concentration of adsorbate will result in faster saturation and vice-versa. It has been seen that saturation time reduced from 37.4 to 25.6 min for sulfamethoxazole and from 51.8 to 25.5 min for sulfapyridine, upon increasing the initial concentration of antibiotics (adsorbate) four times (Tian et al. 2013). The effect of adsorbent concentration has also been studied in the case of ciprofloxacin and norfloxacin. It was found that antibiotic removal increases when the dose of adsorbent (activated carbon) is increased up to 0.75 g/L and 0.5 g/L for ciprofloxacin and norfloxacin, respectively (Ahmed & Theydan 2014). The increased removal rates were supposed to be due to the enhancement in surface area and pore volume upon increasing the concentration, which resulted in more active sites and functional groups to act upon (Jiang et al. 2012). Therefore, proper optimization of initial concentration of adsorbent and adsorbate is significant for enhancing the adsorption and thereby removal efficiency.

Similar to the initial concentration of adsorbate and adsorbent, the contact time between the two is also important for the adsorption process. Contact time between adsorbate and adsorbent in any medium is the minimum time in which the highest removal of adsorbate from the solution could be achieved. Contact time is usually reflected from the kinetics of the adsorption process. Pseudo-second-order kinetics was found to be appropriate for adsorption of various antibiotics on the activated carbon, as reflected from their correlation coefficients (Pouretedal & Sadegh 2014). It depicts that chemisorption is the main mechanism of the adsorption for antibiotics, and in such a case, the adsorption rate is relatively more influenced by the concentration of the adsorbent. In the case of imidazoles, the contact time was reported to be 8 days for significant removal using 0.1 mg/L of activated carbon (Rivera-Utrilla et al. 2009). For tetracycline removal, the contact time at different temperatures was approximately 17 hours on the 0.05 g of powdered activated carbon when 500 mL of tetracycline solution (100 mg/L) was used (Zhang et al. 2015). In this case, the pseudo-second-order model was found to be appropriate for the adsorption process. Similarly, penicillin adsorption also followed the pseudo-second-order model where it took 48–72 hours to reach equilibrium irrespective of the type of carbon adsorbent (Ania et al. 2011).

Adsorption phenomenon is always associated with the change in temperature. Therefore, it is important to understand the effect of temperature of the medium onto the adsorption mechanism of antibiotics over activated carbon. Zhang et al. (2015) studied the effect of temperature on the adsorption process of tetracycline over powdered activated carbon. It was found that the adsorption capacity was more at high temperature compared to low temperature. This is because the adsorption is an endothermic process, and therefore, higher temperature is more favorable for the adsorption. In the study, three different temperatures were chosen, viz. 303, 313, and 323 K. The highest adsorption of tetracycline was found at 323 K, viz. 1121.5 mg/g of activated carbon. Adsorption isotherms further revealed that the adsorption of tetracycline on activated carbon is the heterogeneous surface adsorption process (Zhang et al. 2015).

pH affects the adsorption process of antibiotics onto the activated carbon up to a significant extent by affecting the solubility of adsorbate and surface charge characteristics of adsorbent. Various classes of antibiotics solubilize to a different extent in the medium owing to their existing state (cations, zwitter ions, anions) at a particular pH. Varied ionization states of antibiotics will therefore be different in dissimilar media and the same would be affected by the pH of the solution. For example, the adsorption of tetracycline is promoted at acidic pH while it decreases gradually when the pH is increased. This is due to the fact that high pH causes electrostatic repulsion between the antibiotic species (having basic structure) and the negative charge of the adsorbent surface (Pouretedal & Sadegh 2014; Zhang et al. 2015). Upon increasing the pH, alkaline conditions prevail which render the active sites of adsorbent to convert into hydroxide and phenoxide ions, thereby causing repulsion. Nevertheless, it has also been reported that strong acidic conditions further reduce the adsorption rate owing to the damage of carbon structure and interference caused by H⁺ ions (Choi et al. 2008). Fluoroquinolones, being weak organic acids, behave differently in the medium owing to their pH dependent speciation. It was found that ciprofloxacin and norfloxacin have optimum removal efficiency at pH 9 and 5, respectively (Ahmed & Theydan 2014).

Ionic strength details about the concentration of ions in the solution/medium. Therefore, it is one of the important factors affecting the adsorption process. It has been seen that upon increase in the salt ions concentration in the medium, adsorption of antibiotics on to the activated carbon decreases. Further, ionic salts affect the configuration and electrostatic nature of the antibiotics resulting in decreased solubility which further affects the adsorption capacity of antibiotics (Xiang et al. 2019). Zhang et al. (2015) studied the effect of salts over the adsorption of tetracycline on petroleum-coke derived activated carbon. It was seen that upon increasing the salt concentration from 0.01 to 0.1 M, the adsorption decreased significantly. As the tetracycline molecules are positively charged, the cations of the solution result in competition with tetracycline molecules, resulting in decreased adsorption. It has also been shown that divalent cations cause more inhibition than monovalent cations (Zhang et al. 2015).

Similar to the effect of ionic strength, natural organic materials present in the medium also affect the adsorption. It has been seen that activated carbon acts efficiently for the removal of natural organic matter/humic substances/low molecular weight organics by adsorbing them (Dastgheib et al. 2004; Velten et al. 2011; Tafvizi et al. 2021). Thus, these substances will compete with the antibiotics for active sites on the activated carbon surface, thereby interfering with the adsorption of the latter. Influence of natural organic matter over the adsorption of various pharmaceuticals onto the activated carbon was studied by Ridder et al. (2011). It was found that organic matter reduced the adsorption and it was more pronounced in the case of natural water, rather than in ultrapure water owing to higher hydrophobicity of organic matter in natural water (Ridder et al. 2011). Bajracharya et al. (2019) studied the effect of natural organic matter over the adsorption of microcystin-LR on powdered activated carbon. It was found that upon increasing the concentration of organic matter from 5 to 10 mg/L, the adsorption rate as well as the extent of microcystin-LR adsorption was reduced considerably. The mechanisms involved here were postulated to be the competition of the organic molecules with natural organic matter and pore blockage. The size of the natural organic matter does play a significant role here. Where small molecular weight fraction of natural organic matter (<1000 Da) competes for the adsorption sites on activated carbon; the large molecular weight fraction creates pore blockage or pore constriction, thus interfering with the adsorption of targeted organics in both the cases (Li et al. 2003).

Antibiotics removal from the aqueous media through adsorption over activated carbon is an efficient method; however, it often faces various challenges. One of the challenges is the amount of waste generated during the adsorbent preparation, which renders this process unsuitable to scale up at the industrial level. Moreover, proper optimization of various process parameters is yet another challenge to deal with. In this review, adsorption of antibiotics on to the activated carbon was discussed in light of various physico-chemical properties of activated carbon as well as antibiotics. The effect of process parameters was also reviewed which play an important role in modulating the overall reaction kinetics. It was discussed that initial concentration of adsorbate/adsorbent, pH, ionic strength, contact time, and presence of organic matter in the reaction media have a significant effect on the adsorption mechanism. Thus, all these factors must be taken into consideration for effective removal of antibiotics through activated carbon adsorption in aqueous media.

The authors are thankful to the Swedish Research Council, Sweden for the financial support (project grant number 2021-00889).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.