Adsorption of Amoxicillin using Oxidized Carbon Nanotubes of Aqueous Solution

Introduction

The widespread use of antibiotics has become a major concern due to their negative effects on both health and the environment ^{1, 2}. Studies have shown the detection of antibiotics in various aquatic environments and water supplies. In most cases, these antibiotics exist as mixtures rather than individual compounds ^{3, 4}. Even at low levels, the persistence of antibiotics in water bodies can promote the emergence of antibiotic-resistant genes and harm aquatic organisms, leading to ecosystem disruption and potential threats to human health ⁵. Therefore, the development of efficient methods for the removal of antibiotics from aquatic environments is imperative ^{6, 7}.

A variety of techniques have recently been explored for wastewater treatment, including biodegradation, adsorption, ion exchange, membrane separation methods, and catalytic oxidation, all aimed at degrading and eliminating antibiotics ^{8, 9}. Among these techniques, adsorption processes offer a promising approach for eliminating antibiotics due to their efficiency, simplicity, flexibility, and cost-effectiveness ¹⁰. Effective water purification and antibiotic removal require adsorbents with high adsorption capacity and specific selectivity, attributes typically found in porous adsorbents with large pore diameters and abundant functional groups such as -OH and -COOH ^{11, 12}. However, the adsorption of large-molecule antibiotics from aqueous solutions by adsorbents with smaller pore sizes is hindered by the size exclusion effect. Additionally, the presence of numerous functional groups on material surfaces enhances antibiotic removal from wastewater ^{13, 14}. Mesoporous carbon, with its and abundant functional groups, high surface area, and large pore volume, emerges as a promising candidate for adsorbing large-molecule contaminants ^{15, 16}.

The limited biodegradability of antibiotics renders conventional biological treatment processes relatively ineffective in treating antibiotic manufacturing wastewater ^{17, 18}. Consequently, physical or chemical treatments are often favored. Nonetheless, these methods can be costly and may not adequately address the broad spectrum of antibiotic wastewaters ^{19, 20}. AC serves as a popular adsorbent for antibiotic removal through adsorption ²¹. Nevertheless, the expensive nature of AC has driven the exploration and adoption of numerous alternative adsorbents for removing antibiotics from aqueous solutions ^{22, 23}.

Ever since their discovery, CNTs have been extensively studied across various disciplines ²⁴. The vast potential for future engineering applications is indicated by the exceptional mechanical properties, remarkable electrical conductivity, small size, and large surface area of these materials ^{25, 26}. Within the spectrum of CNTs, SWCNTs and MWCNTs are delineated based on the number of layers they exhibit ²⁷. The production approaches for MWCNTs are relatively straightforward and can be easily scaled up to accommodate large-scale production.

Carbon nanotubes (CNTs) are cylindrical structures composed of one or more rolled-up sheets of graphene. Oxidation refers to the process of introducing oxygen-containing functional groups (like carboxylic acid, hydroxyl, and carbonyl) onto the surface of CNTs. This chemical modification significantly alters their physical and chemical properties ²⁰.

Pristine CNTs tend to aggregate due to strong van der Waals forces. Oxidation introduces polar groups, enhancing their solubility in polar solvents and improving dispersion, making them easier to process and integrate into composites.  Functional groups introduced through oxidation increase CNT reactivity, enabling better adhesion to other materials and facilitating chemical modification for specific applications ²¹. Oxidized CNTs exhibit increased wettability, making them suitable for applications involving contact with liquids, such as filtration and sensors. In some cases, oxidation can enhance the electrical conductivity of CNTs by creating defects that act as charge carriers ²². In this study, we used oxidized carbon nanotubes to remove amoxicillin, which has not been used in other studies.

Materials and Methods

The main chemical employed in our study, as the pollutant, was AMX prepared by Merck (Darmstadt, Germany). Fig. 1 illustrates the chemical structure of AMX. To adjust the pH of the solution, either 0.1 M H₂SO₄ or NaOH was used. Experiments were accomplished in conical flasks (250 mL), in batch mode.

Figure 1: Chemical structure of the used AMX Click here to View Figure

For the production of MWCNTs, chemical vapor deposition (CVD) of acetylene in a hydrogen atmosphere at 760 ◦C was employed, with Ni–Fe nanoparticles serving as catalysts. Fe(NO₃)₂ and Ni(NO₃)₂ underwent a sol–gel process and subsequent calcination to obtain FeO and NiO, respectively. Fe and Ni nanoparticles were generated through deoxidation of the mentioned compounds using H₂. The ends of the tubes were uncapped by placing the raw MWCNTs in a liquid consisting of 3M HNO₃. We did this so we can use the MWCNTs for something else. We mixed 3 grams of MWCNTs with 400 mL of 3 M HNO₃ and used sound waves to stir it for 24 hours. After that, we filtered the mixture and rinsed it with clean water until the liquid had a pH of about 6. Then, we dried the mixture at 80 degrees Celsius. Following this, the oxidized MWCNTs underwent a calcination process at 450 ◦C for 24 hours to eliminate any remaining amorphous carbon, after which they were utilized in the subsequent experiments. The experiments used chemicals that were of analytical purity and were directly employed without any additional purification steps. Additionally, Milli-Q water was utilized to prepare the solutions.

The experiments were fulfilled at ambient conditions and T = 28 ± 2 ◦C; for this purpose, polyethylene tubes, as batch system, were employed. To conduct the measurements, a specific quantity of OMWCNT was mixed with AMX solution (this was done in 100 mL flasks) and shaken at a fixed rate of 200 rpm for a certain period. Once the adsorption experiments were finished, an external magnet was used to separate the OMWCNT. UV-vis spectroscopy was used to track changes in absorbance of AMX solution samples at specific time intervals. The maximum wavelengths (λ_max) of 622 nm were utilized for the determination of the changes. The removal ratio (R%) of AMX was computed by comparing its initial and equilibrium concentrations. All the experimental results were obtained from duplicate determinations and had a relative error of roughly 5%.

The thermodynamic parameters, including the free energy change (ΔG⁰), enthalpy change (ΔH⁰), and entropy change (ΔS⁰), were estimated using the equations presented below ²⁰.

Kc is the symbol used to represent the sorption equilibrium constant, R is the symbol used to represent the gas constant, and T is the symbol used to indicate the absolute temperature in Kelvin.

Results and Discussion

Analysis using the N2-BET method revealed the specific surface area of the OMWCNTs to be 214 m²/g. Subsequent examination of the OMWCNT morphologies was conducted via scanning electron microscopy (SEM) utilizing a JEOL JSM-6700F instrument (see Fig. 2). The SEM illustrates OMWCNTs characterized by exceptionally smooth surfaces, with diameters varying from 10 to 30 nm and lengths measuring several micrometers.

Figure 2: SEM image of OMWCNTs sample Click here to View Figure

Effect of parameters on antibiotic adsorption

The design of cost-effective wastewater treatment systems relies heavily on determining the equilibrium time, which is a critical parameter. In order to study the adsorption kinetics, five distinct AMX quantities were selected. Fig. 3 illustrates the relationship between the adsorption capacity (q_t) of AMX onto OMWCNTs and contact time for different initial concentrations of AMX. The qt value showed a steep rise from 98.8 mg/g to 829.4 mg/g as the concentration rose from 10 mg/L to 100 mg/L, as seen in Fig. 3.

The impact of the initial concentration of AMX on the adsorption process was studied by varying it from 10 to 100 mg/L. The pH was kept constant at 7, and the amount of adsorbent was fixed at 0.2 g/L. The contact time was set at 10-150 min. According to figure 3, the efficiency of elimination reduced with the rise in the initial AMX concentration. The minimum efficiency was observed at an initial AMX concentration of 100 mg/L.

Adsorption process was also deliberated at four different temperatures to analyze its temperature dependence. The initial AMX concentrations were varied over a range, and the adsorption isotherms were used to analyze how changes in temperature affect the studied process. Fig. 4 illustrates the variation in adsorption isotherms with temperature. The experimental results indicated an enhancement in the equilibrium uptake by rising the equilibrium AMX concentrations within the experimental concentration range.

To investigate the kinetic behavior of AMX adsorption onto OMWCNTs, three commonly used kinetic models were employed, namely the PFO, PSO, IPD model. In PSO model, the rate-limiting step is the surface adsorption that involves chemisorption, where the removal from a solution is due to physicochemical interactions between the two phases ²⁷. The model that best fit the data was designated based on the correlation coefficient values (R²) obtained through linear regression. Equation 3 represents the PFO model ^{28, 29}.

The quantities of AMX adsorbed at equilibrium and time t are represented by q_e and q_t (mg/g), respectively. The rate constant of the PFO is denoted by k₁ (min^-1). The slope and intercept of the log(q_e-q_t) versus t plots can be employed to compute the values of k₁ and q_e. These calculations are presented in Fig 5a. The R² gotten for all considered concentrations are in the range of 0.75-0.85 and are low. Additionally, the q_e,exp (mg/g) experimental values are significantly different from the calculated q_{e, cal} (mg/g). As a result, the PFO does not effectively describe the adsorption process.

Equation 4 expresses a PSO in a linear form ^{30, 31}.

The PSO rate constant, k₂ (g/mg min), can be derived from the rate equation. The values of q_e and k₂ can be obtained by plotting t/qt against t (Fig. 5b). The high R² (>0.999) of the linear plots at different concentrations indicate that the adsorption of AMX onto OMWCNTs follows the PSO predominantly. Besides, the q_e,cal are in good agreement with the q_e, exp. Table 1 summarizes the equations related to kinetic models and the results obtained from their application.

The two aforementioned models are unable to determine the diffusion mechanism that occurs during the studied process. Therefore, the experimental data is examined using the IPD model (equation 5) ^{32, 33}.

In above equation, k_d (mg/g min^1/2) signifies the IPD rate constant, and C is the intercept. The adsorption mechanism of a solute from a solution onto porous adsorbents progresses through three distinct phases. Following this, the process transitions into a stage of gradual adsorption (step II), where the rate-limiting factor becomes IPD. Finally, equilibrium is achieved, with IPD slowing down caused by the minimal concentrations of adsorbate left in the solutions ³⁴. The regulation of the adsorption rate is influenced by one or more of these three stages.

In the graph depicted in Figure 5c, where qt is plotted against t^1/2, a dual linear trend is observed. Initially, it is posited that AMX quickly migrates to the outer surface of OMWCNTs via film diffusion within a short duration. The initial linear phase likely reflects the penetration of AMX molecules into the OMWCNTs particle through IPD. Subsequently, the subsequent linear portion signifies the establishment of final equilibrium ³⁵.

Over the past few years, the identification of human medicines and pharmaceutical drugs in sewage, wastewater, and domestic water systems has brought attention to their potential ecotoxicological effects on human health in numerous countries ³⁶. The detrimental impacts of pharmaceuticals in water resources pose serious risks to human health, as well as to animals and plants, with far-reaching consequences ³⁷.

The data presented in Fig. 3 demonstrate a direct relationship between equilibrium uptake and increasing concentrations of AMX within the experimental concentration range. The explanation for this could lie in the enhanced driving force resulting from the concentration gradient. For   higher concentration of AMX, more AMX ions surrounded the active sites of OMWCNTs, facilitating adsorption and consequently leading to an increase in q_e values ³⁸. Additionally, it was noted that the q_e rose with increasing temperature, suggesting the endothermic nature of AMX adsorption onto OMWCNTs ^{39, 40}. It’s feasible that this arises from the strengthening of physical bonding between AMX molecules and the active sites of the adsorbent at higher temperatures, coupled with the increased solubility of AMX, which enhances interaction forces between the solvent and the solute ⁴¹.

The concentration of the pollutant (AMX) at the start of the study is a critical factor in adsorption and requires assessment. Previous research has demonstrated significant effects of this factor up to a certain limit. Additionally, the adsorption capacity of the adsorbent for AMX molecules depends on the presence of sorption sites on its surface. Once the optimal value is exceeded, the adsorption efficiency decreases due to the saturation of active sites on the adsorbent’s surface ³⁵. Moreover, there is minimal available space between the layers of OMWCNTs, and the adsorption process involves replacing interlayer anions of the adsorbent with contaminants. This space can accommodate the ideal values of OMWCNTs and enable efficient adsorption ⁴⁰.  Consequently, the saturation of active sites in the interlayer of OMWCNTs prevents further diffusion of contaminants onto the adsorbent.

The concentration of AMX during initial stages significantly impacts its adsorption capacity on OMWCNTs adsorbent. The observation reveals that as the initial AMX concentration increases, the adsorption capacity of AMX also increases. A heightened driving force of the concentration gradient with the increase in the initial concentration could be the reason behind this phenomenon. During the initial stages of adsorption (contact time<45 min), the process is rapid, whereas it slows down later on. The presence of many vacant surface sites at the outset may contribute to this, making it easier for AMX molecules to adsorb. However, as the surface sites get occupied, the unoccupied sites become challenging to fill due to repulsive forces between the AMX molecules on the OMWCNTs and the bulk phase ³⁴. The curves observed during the process are single, smooth, and continuous towards saturation, signifying that monolayer coverage of AMX molecules on OMWCNTs surface has been achieved ³⁵.

The thermodynamic parameters provide valuable information such as the strength of adsorbate attachment to the adsorbent surface, whether the process is chemisorption or physisorption, whether it is exothermic or endothermic, and the entropic changes during adsorption. Physisorption occurs when the ∆𝐺° value is approximately 20 kJ/mol, while chemisorption occurs when the value ranges from 80-400 kJ/mol ³³. According to the results of the adsorption model, the adsorption of the drug on the OMWCNTs adsorbent was determined to be physisorption. A negative ∆𝐺° value designates that the adsorption is endothermic and spontaneous. It was observed that the change in free energy decreases with increasing temperature, but the nature of adsorption is non-normal. It is possible that the reason for this is the strengthening of the adsorption process caused by rising temperature ³⁵. The ∆𝐺° for AMX adsorption was found to be within the range of -2.2 to -8.9 kJ/mol, indicating that the adsorption was primarily physical adsorption ³⁶. A ∆S°>0 specifies a developed randomness between solid-solution interfaces during the adsorption of AMX on OMWCNTs ³⁸.

The graph in Figure 5c illustrates the plot between qt and t^0.5 for AMX adsorption and desorption onto OMWCNTs. It is evident from the graph that both adsorption and desorption exhibit a multi-linear behavior, with two distinct steps visible in the plots representing the initial and second stages ³⁶. The initial stage is likely associated with the boundary layer diffusion effect, while the next stage may be a result of IPD effects. The current study revealed that the rate constants (k_d) governing both adsorption and desorption exhibit higher magnitudes, showing an upward trend with increasing initial AMX concentration. Consequently, the linear section at the upper portion of the graph can be interpreted as a rate parameter (k_d), indicative of the adsorption and desorption rates within the domain where IPD was identified as the rate-controlling step ³⁴.

Table 2 shows a comparison of the adsorption capacity (q_e) of different materials reported in the articles as adsorbent for AMX adsorption from aqueous solution under different experimental conditions.

The mechanism of adsorption of AMX to the OMWCNTs material involves a multi-step process where molecules move into the material.  This process is seen in the data as a steep part (bulk diffusion) followed by a less steep part (film and pore diffusion) (35).  

Additionally, the OMWCNTs surface has different chemical features like -OH and -C=O groups that can bind to AMX in various ways, including strong (covalent) and weak (hydrogen bonding, ionic interaction, etc.) connections. The -NH₂ groups can also bond with AMX through strong (imine) and weaker (amide) connections, and other interactions (Fig 6).

Figure 3: Effect of contact time (temperature: 303 K, adsorbent dose: 0.1 g/L) Click here to View Figure

Figure 4: Effect of temperature (C0:100 mg/L, adsorbent dose: dose: 0.1 g/L, time: 75 min). Click here to View Figure

Figure 5: PFO (a); PSO (b); IPD (C) kinetic plots for AMX adsorption onto OMWCNTs. Click here to View Figure

Table 1: Values of kinetic parameters for the adsorption of AMX

AMX (mg/L)	(qe) exp	IPD model			PFO			PSO
		Kd	I	R2	(qe) cal	K1	R2	(qe) cal	K2	R2
10	41.55	2.11	11.62	0.752	11.6	0.451	0.895	21.2	0.004	0.997
25	81.34	3.65	25.47	0.787	19.47	0.274	0.912	65.2	0.001	0.996
50	119.1	7.94	29.76	0.884	28.94	0.011	0.904	161.2	0.0009	0.994
75	155.1	10.25	38.48	0.816	59.73	0.046	0.876	214.2	0.0004	0.998
100	196.1	14.2	44.3	0.793	67.2	0.074	0.871	251.7	0.0007	0.992

 

Figure 6: Adsorption mechanism of AMX onto OMWCNTs Click here to View Figure

Table 2: Comparison of the maximum adsorption of various adsorbent for AMX

Adsorbents	Qe (mg/g)	Ref
AC- Jujube nuts	56.2	13
MWCNT	465.1	14
AC- cashew of Para	56.95	15
AC-Date Pits	84.93	18
AC-olive stone	84.1	21
Synthesized NiO	94.3	26
Magnetic graphene oxide	93.2	34
OMWCNTs	926.2	This study

 

Conclusion

 In the present investigation, it was found that the adsorbent demonstrated a maximum uptake capacity of 829.4 mg/g and achieved a total percentage removal of 98.71% of AMX from the aqueous solution. Optimal physicochemical parameters were determined to be a dosage of 0.1 g, a contact time of 75 minutes, and a temperature of 40 °C, respectively. Various adsorption kinetic models, including the PFO, PSO, and IPD models, were employed to assess the kinetics of maltose sorption by OMWCNTs. The results indicate that the sorption of AMX onto OMWCNTs conforms well to the PSO.

Acknowledgement

The authors are grateful to Zahedan University of Medical Sciences, because of supporting this research (IR.ZAUMS.REC.1401.147).

Funding Sources

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Conflict of Interest

The author(s) do not have any conflict of interest.

Data Availability Statement

This statement does not apply to this article.

Ethics Statement

This research did not involve human participants, animal subjects, or any material that requires ethical approval.

References

1. Fei Y, Li Y, Han S, Ma J, J. Colloid Interface Sci. 2016; 484, 196–204.
   CrossRef
2. Balarak D, Mahvi AH, Shim MJ, Lee SM. Desalin Water Treat. 2021; 212, 390-400.
   CrossRef
3. Ahmadi S, Banach A, Mostafapour FK. Desalin Water Treat. 2017; 89: 297-303.
   CrossRef
4. Yilmaz M, Al-Musawi TJ, Saloot MK.  Biomass Conv. Bioref.  2022. https://doi.org/ 10.1007/s13399-021-02279-y
5. Al-Musawi TJ, Mengelizadeh N, Taghavi M. Biomass Conv. Bioref. 2021. https://doi.org/10.1007/s13399-021-01962.
6. Wang H, Fang C, Wang Q, Chu Y, Song Y, Chen Y, Xue X. RSC Adv. 2018; 8:16260–16268.
   CrossRef
7. Wang J, Zhang M, Zhou R, Li J, Zhao W, Zhou J. Water Sci Technol. 2020; 82:242–254.
   CrossRef
8. Wu S, Hu YH. Chem Eng J. 2021; 409:127739.
   CrossRef
9. Mahvi AH, Mostafapour FK. Fresenius Environ Bull. 2018; 27(8):5759-67.
10. Balarak D, Taheri Z, Shim MJ, Lee SM, Jeon C. Desalin Water Treat. 2021; 215; 183-193.
    CrossRef
11. Azarpira H, Mahdavi Y, Khaleghi O. Pharm Lett. 2016; 8(11):107-13.
12. Shikuku VO, Zanella, Kowenje CO.   Appl Water Sci. 2018; 8, 175. https://doi.org/ 10.1007/s13201-018-0825-4.
    CrossRef
13. Belaissa, Y., Saib, F. & Trari, M. Reac Kinet Mech Cat; 2022; 135, 1011–1030.
    CrossRef
14. Balarak D, Mostafapour FK, Bazrafshan E, Saleh T. A.  Water Sci Technol. 2017; 75; 1599-1606.
    CrossRef
15. Lima, DR., Lima, EC., Umpierres, C.S.  Environ Sci Pollut Res. 2019; 26, 16396–16408. https://doi.org/10.1007/s11356-019-04994-6.
    CrossRef
16. Leonard JM. Githinji, MKM., Ramble O. Ankumah. Water Air Soil Pollutant, 2011. 2019: 191-201.
    CrossRef
17. Ibáneza M, Bijlsmaa L, Moralesb E, Pastorb L, Hernándeza F. J. Hazard. Mater.  2013. 260: 389-398.
18. Belhachemi, M., Djelaila, S. Environ. Process. 2017; 4, 549–561.
    CrossRef
19. Boukhelkhal, A., Benkortbi, O., Hamadeche, M. Water Air Soil Pollut. 2015; 226, 323. https://doi.org/10.1007/s11270-015-2587-z.
    CrossRef
20. Moarefian, A., Golestani, H.A. & Bahmanpour, H. J Environ Health Sci Eng. 2014; 12, 127. https://doi.org/10.1186/s40201-014-0127-1.
    CrossRef
21. Limousy, L., Ghouma, I., Ouederni, A. Environ Sci Pollut Res. 2017;  24, 9993–10004. https://doi.org/10.1007/s11356-016-7404-8.
    CrossRef
22. Gao H, Zhao S, Cheng X, Wang X, Zheng L. Chem. Eng. J. 2013. 223; 84–90.
    CrossRef
23. Balarak D, Mostafapour FK, Joghataei A. Pharm. Chem. 2016, 8(8):138-145.
24. Balarak D, Mahdavi Y, Bazrafshan E, Mahvi AH, Esfandyari Y. Fluoride. 2016; 49(1)71-83.  
25. Naghsh N, Chandrika K, Balarak D. Int. J. Pharm. Investigation. 2024;14(2):365-70.
26. Eltaweil AS, Abd El-Monaem EM., El-Subruiti GM.  J Porous Mater. 2023; 30, 607–618.
    CrossRef
27. Balarak D, Chandrika K. Int J Pharm Investig. 2019;9(3):117-21. 
    CrossRef
28. Rout, DR, Jena, HM. Environ Sci Pollut Res. 2022; 29, 32105–32119. https://doi.org/10.1007/s11356-021-17944-y.
    CrossRef
29. Mostafapour FK, Bazi M, Siddiqui SH, Bagheri H. Int J Pharm Investig. 2021;11(4):384-8. 
    CrossRef
30. El Rouby WMA, Farghali AA, Sadek MA.  J Inorg Organomet Polym. 2018; 28, 2336–2349. https://doi.org/10.1007/s10904-018-0885-9.
    CrossRef
31. Li R, Wang Z, Zhao X.  Environ Sci Pollut Res. 2018; 25, 31136–31148. https://doi.org/10.1007/s11356-018-3064-1.
    CrossRef
32. Balarak D, Ganji F, Chandrika K, Haseeb S. Int J Pharm Investig. 2020;10(2):122-6.
    CrossRef
33. Mostafapour FK, Haseeb S, Balarak D, Moein H, Sajadi AA, Jalalzaei Z. Int J Pharm Investig. 2021;11(1):41-5.
    CrossRef
34. Kong X, Liu Y, Pi J.  Environ Sci Pollut Res. 2017; 24, 6679–6687.
    CrossRef
35. Liang, H., Zhu, C., Ji, S. Biochar. 2022;  4, 3. https://doi.org/10.1007/ s42773-021-00130-1.
    CrossRef
36. Shan DN, Deng SB, Zhao TN, Wang B, Wang YJ, Huang J, Yu G, Winglee J, Wiesner MR. J Hazard Mater. 2016; 305:156–163.
37. Zisti F, Chandrika K. Int. J. Pharm. Investigation. 2023;13(4):778-83.
    CrossRef
38. Azadfar M,  Tahermansouri H,   Qomi M.  J. Chin. Chem. Soc, 2021;68:2103–2117.
    CrossRef
39. Gholizadeh, H, Ghorbani-HasanSaraei A, Tahermansouri H.  Carbon Lett. 2019; 29, 273–279.
    CrossRef
40. Gholizadeh, H,Ghorbani-HasanSaraei A, Tahermansouri H.  Shahidi SA. J. Chin. Chem. Soc, 2020;67: 546–557.
    CrossRef

---

This work is licensed under a Creative Commons Attribution 4.0 International License.