Tetracycline removal from an aqueous solution using bimetallic-impregnated activated carbon synthesized via one-step dry-pyrolysis of hawthorn kernels Authors: Hakimeh Sharififard, Behniya Hafezi Published: 2025 ABSTRACT A novel bimetallic Fe-Mn-impregnated activated carbon (Fe: Mn: AC) composite was synthesized via the one-step dry pyrolysis method and is applied to tetracycline (TC) removal from aqueous solution. Characterization results revealed a mesoporous structure with various functional groups on the Fe: Mn: AC adsorbent. The impacts of solution pH, adsorption time (kinetics), and initial TC concentration on equilibrium in batch adsorption were studied. The highest TC removal efficiency using Fe: Mn: AC occurred at a pH of 6. The kinetic analysis using various models revealed that the data best fit the Elovich model, highlighting the significance of external diffusion (film diffusion) during the adsorption process. The equilibrium data aligned well with the Langmuir isotherm, which predicted a maximum adsorption capacity of 463.040 mg/g for TC adsorption at 25 °C. The primary mechanisms of adsorption included Hydrogen bonding, surface complexation with metal functional groups (M-π), Lewis acid-base interactions, π-π interactions, and pore filling. Continuous mode experiments in five cycles confirm the high adsorption capacity of the Fe: Mn: AC adsorbent for effectively removing pharmaceutical contaminants from aqueous media (27.5% reduction in adsorption ability).Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-31638-2. Introduction Pharmaceutical compounds are widely used in the treatment of diseases. Among these substances, antibiotics are widely used to prevent and treat microbial infections in humans and animals1,2. Tetracycline is one of the most widely used antibiotics by humans and animals (due to better antibacterial activity, lower cost, and lower side effects). Tetracycline, consumed by humans and animals, is not fully absorbed and metabolized, and it is transported to the environment through their feces and urine1,2. The presence of antibiotics in the environment (water and soil resources), even in very small amounts, is very harmful to both ecosystems and human health3. The presence of antibiotics, such as tetracycline, in the environment leads to antibiotic resistance in water bacteria, which causes drug resistance in humans and animals, posing a serious threat to public health1. Therefore, it is crucial to remove the antibiotics from water. Several separation methods including biodegradation4, electrochemical method5, photocatalytic processes6, membrane technique7, and adsorption8, have been suggested for treating wastewater containing tetracycline. Among the new methods that have attracted the attention of many researchers in recent years is the adsorption method, which is simple, low-cost (with low energy consumption due to being carried out at ambient temperature and pressure), flexible, and highly efficient9–17. Considering the role of the solid adsorbent in the adsorption process, various synthetic and natural adsorbents have been developed and utilized to adsorb a range of contaminants. Since adsorption is carried out on the surface of the adsorbent particles, adsorbents with a high specific surface area lead to high adsorption capacity and also high adsorption rate. Among the most popular adsorbents is activated carbon. Activated carbon is a special adsorbent that can adsorb various adsorbates from liquids and gases because of its porous surface, high chemical stability, and high mechanical strength13,15,17. Today, activated carbon receives considerable attention due to the availability and affordability (cheapness) of raw materials for synthesis, which are often prepared from agricultural or industrial waste. The process of producing activated carbon from these wastes involves two carbonization and activation stages. In the carbonization stage, by heating the raw material (at temperatures between 400 and 800 and sometimes 1000 degrees Celsius) and in the presence of an inert gas, non-carbon materials such as oxygen and hydrogen are destroyed. This is done to achieve a constant mass of carbon and also to create a primary and porous structure. To increase the porosity of the carbon material obtained from the carbonization stage, the activation stage is performed. Activation is performed by two methods: physical and chemical activation processes, or a combination of the two processes. One effective method to enhance the adsorption capacity of activated carbon adsorbents is to deposit metal nanoparticles on their surfaces. This deposition not only increases adsorption capacity but also imparts magnetic properties to the adsorbent, enabling its separation from the solution using an external magnetic force18. Most research on depositing metallic nanostructures onto activated carbon surfaces employs the wet-impregnation method following the synthesis of activated carbon18–22. In contrast, one-step dry-pyrolysis is a less energy-intensive method for producing carbon adsorbents that are impregnated with metal nanoparticles23. The dry-mixed pyrolysis method can anchor metallic nanostructures onto the activated surface. In this technique, the metal salts (acting as the activating agent) are mixed with the raw material in a dry state, and the pyrolysis process is carried out on this dry mixture. This approach is more feasible for industrial-scale applications compared to the traditional wet-impregnation method, which mixes the raw material with a solution of activating agent (metal salts) before pyrolysis24,25. Despite the advantages of the dry mixed pyrolysis method—such as being achievable in one step, which saves time and energy—limited studies have been conducted in this area23. There are many Hawthorn (Crataegus) trees in the Kohgiluyeh and Boyer-Ahmad Province. Therefore, a huge amount of hawthorn kernels is available and can be used as agricultural by-products to produce activated carbon. In this research, first, Fe3O4/Mn2O4/activated carbon (Fe: Mn: AC) was synthesized from hawthorn kernel, a biodegradable and inexpensive adsorbent, using a one-step dry-mixed pyrolysis method. This material was then applied to remove tetracycline (TC) from aqueous solutions via both batch and continuous adsorption processes. Materials and methods Synthesis of fe: mn: AC via dry-mixed pyrolysis After washing with distilled water, the hawthorn kernel were dried at room temperature for 72 h, then ground and placed in a sieve shaker to prepare a 35- to 40-mesh particle-size powder. 5 g of biomass powder, 2 g NaOH (≥ 99%, Merck), 3 g FeSO4.7H2O (≥ 99.5%, Merck), and 1 g KMnO4 (≥ 99%, Merck) were mixed homogeneously (using a household mixer). The mixture was put into a stainless steel container. The N2 gas was purged into the container for 1 h to remove the air (gas flow rate: 500 mL/min). The pyrolysis process was performed in two steps. In the first step, the container was heated at 350 °C using a muffle furnace (M11L-1100 model) for 60 min (heating rate: 10 °C/min). This step is performed to ensure that the reaction of biomass with NaOH occurs in the molten state23. In the second step, the mixture was heated at 600 °C for 150 min (heating rate: 10 °C/min). After cooling to room temperature, the product was washed with distilled water several times until reaching a constant pH of the washing water. Additionally, the concentrations of iron (Fe) and manganese (Mn) in the washing water were detected, and washing continued until neither iron nor manganese was detected in the washing water (analysis performed via AAS). This product was named Fe: Mn: AC. Fe: AC (Fe-impregnated activated carbon) adsorbent was synthesized in the same process (without KMnO4) to study the impact of two metallic elements on the properties and the adsorption capacity. The hawthorn kernel was pyrolyzed with NaOH in the same process, and the product was named AC (activated carbon). The production yields (%) for synthesizing AC, Fe: AC, and Fe: Mn: AC carbonaceous adsorbents were 39.78%, 44.67%, and 45.31%, respectively24. Characterization of prepared adsorbent Scanning electron microscopy (SEM) technique using TESCAN, TESCAN-Vega3 equipment equipped with energy-dispersive X-ray analysis (EDX) and element mapping was applied to analyze the purity and morphology of prepared adsorbents. The crystal structures of the metallic elements (Fe and Mn) coated onto the Fe: Mn: AC adsorbent surface were analyzed using XRD pattern (X-ray powder diffraction) with Rigaku Ultima IV (Japan) equipment. Zetasizer equipment (UK, Malvern) was applied to measure the zeta potential of prepared adsorbents. The specific surface area and porosity properties of the prepared adsorbent were determined using BELSORP mini II (Japan) equipment. Fourier transform infrared spectroscopy (FTIR) analysis was applied to detect the functional groups of prepared adsorbents using Perkin-Elmer Spectrum GX equipment. A VSM (Vibrating Sample Magnetometer) analyzer (MDK, VSM, Iran) was applied to detect the magnetic properties of prepared Fe: Mn: AC adsorbent. TC removal via batch and continuous adsorption process The batch process In the first stage of this study, the effect of pH on the adsorption performance of the synthesized adsorbents was investigated, covering a range from 1 to 12. During this stage, the capabilities of the prepared adsorbents in removing TC were compared. The percentage of TC (C22H24N2O8. HCl, ≥ 95%, Sigma-Aldrich) removal was calculated according to Eq. (1). C0 (mg L−1) and Cf (mg L−1) denote the initial concentration and the final concentration of TC in the samples, respectively (measured using a Jasco730, Japan UV spectrophotometer at λmax = 357 nm)1. In the second stage, the effect of contact time (adsorption kinetics) and the effect of initial concentration of TC at equilibrium time (equilibrium experiments) were investigated. The operating conditions of the experiments performed are stated in Table S1 (presented in the Supporting Information file). The amount of TC adsorbed per gram of the solid adsorbent (q (mg/g)) in kinetic and equilibrium experiments was calculated using Eq. (2) and then used to analyze the data. In the batch process, all experiments have been done twice, and the mean value was applied as the final data. Due to the magnetic property of Fe: Mn: AC adsorbent, the separation of Fe: Mn: AC adsorbent from the solution after the adsorption process was performed using an external magnetic force (magnet).1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R\:\left(\mathrm{\%}\right)=\frac{{C}_{0}-{C}_{f}}{{C}_{0}}\times\:100$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:q=\frac{{(C}_{0}-{C}_{f})V}{m}$$\end{document} Continuous adsorption process One key aspect of the adsorption process is the potential to use the synthesized adsorbent in a continuous (industrial) system. To assess this capability, adsorption experiments are performed continuously in the laboratory. In this study, the adsorption capacity of the Fe: Mn: AC adsorbent in a continuous setup was tested using a glass column that is 50 cm long and 0.9 cm in diameter. A peristaltic pump was utilized to deliver the feed solution to the top of the column. More details about this experiment are available in the Supporting Information file. The continuous process of TC adsorption was conducted in five adsorption/desorption cycles. After adsorption, adsorbed TC molecules were removed from the saturated Fe: Mn: AC surface by contacting it with a 0.1 M HCl (37%, Merck) solution for 12 h at a flow rate of 5 mL/min1. Then, distilled water was passed through the column for 24 h before starting the next adsorption cycle. More details are presented in the Supporting Information file. The operational conditions for the adsorption process during regeneration experiments: Feed flow rate: 5 mL/min, adsorbent mass: 2.5 g, pH = 6, C0 = 200 mg L− 1. Results and discussion Characterization of adsorbent In the FTIR spectrum of the synthesized AC (presented in Fig. 1a), peaks at 3426 cm− 1 and 2981 cm− 1 indicate O-H bond vibrations, and the stretching of the C-H bond. The presence of carbonyl groups (C = O) in the structure of AC is evident with peaks in the range of 1600–1750 cm− 1. The carbonyl group is a characteristic functional group of carboxylic acids, ketones, and aldehydes. The peaks appearing in the range of 1000–1380 cm− 1 correspond to the vibration of the C-O band, which is present in phenol and alkoxy functional groups. In the FTIR spectrum of Fe: AC adsorbent (Fig. 1a), peaks corresponding to the hydroxyl, carboxylic acid, phenolic, aldehyde, and ketone functional groups in the activated carbon structure are present. However, the intensity of these peaks has changed, and they have also shifted slightly, indicating the interactions between the surface functional groups of activated carbon with iron. Additionally, new peaks have been observed at 705 cm− 1 and 579 cm− 1, indicating the F-O bond vibrations21,23. In the FTIR spectrum of Fe: Mn: AC adsorbent (Fig. 1a), new peaks appear at 723 cm− 1, 573, and 450, indicating vibrations of Fe-O and Mn-O bonds21,23. The appearance of these new peaks, along with the hydroxyl, carboxylic acid, aldehyde, phenol, and ketone functional groups on the surface of the activated carbon, indicates the formation of new and additional functional groups on the modified adsorbent surface. The XRD pattern of the Fe: Mn: AC nanocomposite, shown in Fig. 1b, displays characteristic peaks at 2θ values of 35.8°, 44.8°, and 57.2°. These peaks indicate the face-centered cubic (fcc) structure of Fe3O4. Also, a broad peak at 2θ of 20°−30° corresponds to the amorphous phase of activated carbon. The peak at 2θ = 35.9° and 62.6° is also attributed to the tetragonal crystal structure of manganes oxide (Mn2O4). Figure 1c presents the VSM loop of Fe: Mn: AC adsorbent. This loop demonstrates the magnetic properties of this adsorbent, due to the presence of Fe3O4 and Mn2O4 metallic structures on its surface, as confirmed by XRD and EDX analyses. The saturation magnetization of Fe: Mn: AC adsorbent is about 27 emu/g. Therefore, the Fe: Mn: AC adsorbent can be separated from the solution using an external magnetic force. The nitrogen adsorption-desorption isotherms for the carbon-based adsorbents, along with the pore size distribution diagram, are presented in Fig. 2. The porosity parameters are detailed in Table 1. The BET (Brunauer–Emmett–Teller) method was applied to measure the total pore volume (Vtotal pore) and the total specific surface area (SBET). Additionally, the t-plot method was used to determine the micropore volume (Vmicro(cm3/g STP))26,27. The isotherm diagrams (Fig. 2a) are of the IV type, indicating the meso-microporous structure of the synthesized adsorbents. Comparing the specific surface area values shows a decrease in surface area with the presence of nanoparticles, which is caused by the closure of metal nanoparticles in the pores of the activated carbon surface and a reduction in its surface area23,28. The data present in Table 1 indicate a decrease in the total pore volume of activated carbon by embedding nanoparticles in its structure. However, the volume percent of the adsorbents’ mesopores has increased by embedding nanoparticles on their surface. By depositing nanoparticles on the surface of activated carbon, the volume percentage of mesopores increased from 34% for AC to 53% for Fe: AC and 98% for Fe: Mn: AC (Fig. 2b). Since adsorbate molecules/ions move more easily in mesopores than micropores, this enhances their diffusion rate in the adsorption onto Fe: AC and Fe: Mn: AC adsorbents9,29,30. In addition to increasing the volume percentage of mesopores, the average pore diameter increased from 6.2817 nm to 10.601 nm for Fe: AC and 16.049 nm for Fe: Mn: AC by placing metal nanoparticles on the surface of the activated carbon. Fig. 1(a): FTIR spectra of AC, Fe: AC, Fe: Mn: AC adsorbents; (b): XRD patern of Fe: Mn: AC adsorbent, (c): The VSM loop of Fe: Mn: AC adsorbent. (a): FTIR spectra of AC, Fe: AC, Fe: Mn: AC adsorbents; (b): XRD patern of Fe: Mn: AC adsorbent, (c): The VSM loop of Fe: Mn: AC adsorbent. Fig. 2(a): The adsorption/desorption isotherms; (b): BJH (Barrett-Joyner-Halenda) plot (pore size distribution) for AC, Fe: AC, Fe: Mn: AC adsorbents. (a): The adsorption/desorption isotherms; (b): BJH (Barrett-Joyner-Halenda) plot (pore size distribution) for AC, Fe: AC, Fe: Mn: AC adsorbents. Table 1Porosity characteristics of synthesized carbon adsorbents.AdsorbentSBET (m2/g)Vtotal pore (cm3/g)Vmeso (cm3/g)Vmicro (cm3/g)Mean Pore diameter Dp (nm)AC10401.23410.41950.81456.2817Fe: AC5880.740.39330.347810.601Fe: Mn: AC3910.65040.64010.010516.049 Table 1Porosity characteristics of synthesized carbon adsorbents.AdsorbentSBET (m2/g)Vtotal pore (cm3/g)Vmeso (cm3/g)Vmicro (cm3/g)Mean Pore diameter Dp (nm)AC10401.23410.41950.81456.2817Fe: AC5880.740.39330.347810.601Fe: Mn: AC3910.65040.64010.010516.049 Porosity characteristics of synthesized carbon adsorbents. The increase in the average pore diameter facilitates and increases the movement rate of adsorbate molecules/ions into the inner channels of the adsorbent structure, allowing for faster diffusion. In general, although the existence of nanoparticles on the activated carbon surface reduces SBET of the adsorbent, it optimizes the pore size and pore size distribution, ultimately improving the movement and diffusion rate of adsorbate molecules/ions. Figure 3 shows the SEM analysis images of the synthesized adsorbents. The SEM image of the synthesized AC surface reveals the highly porous surface of the synthesized AC, which is also confirmed by the BET results. Surface morphology images of Fe: Mn: AC and Fe: AC adsorbents indicate a decrease in adsorbent porosity due to the presence of metal nanoparticles. Furthermore, the EDX spectrum demonstrates that the synthesized AC has a high level of purity. The EDX spectrum of the Fe: AC and Fe: Mn: AC adsorbents confirms the existence of iron (Fe) and manganese (Mn) atoms in the structure of these adsorbents. Also, element mapping images reveal that the Fe and Mn elements are evenly distributed on the AC surface. Fig. 3SEM images, EDX espectra and element mapping of AC, Fe: AC, Fe: Mn: AC adsorbents. SEM images, EDX espectra and element mapping of AC, Fe: AC, Fe: Mn: AC adsorbents. TC removal Batch process The pH parameter is a highly influential factor in adsorption from an aqueous medium, and its impact on TC removal through the adsorption process is very significant. Besides impacting the charge of the adsorbent surface, the pH of the solution also affects the ionic state of the organic substances such as TC in the wastewater. The pHZPC (zero point charge) is a key property of the adsorbent that defines its surface charge type (positive or negative) at various pH levels. According to the zeta potential analysis (presented in Fig. 4a), the pHZPC value for AC, Fe: AC, and Fe: Mn: AC adsorbents is 8.4, 7, and 6.6, respectively. The adsorbent surface is positive at pH below pHZPC (due to protonation of the functional groups) and has a negative charge at pH above pHZPC due to deprotonation of the functional groups). Fig. 4(a): Zeta potential diagram of AC, Fe: AC, Fe: Mn: AC adsorbents, (b): The effect of pH solution on TC removal ability of prepared adsorbents. (a): Zeta potential diagram of AC, Fe: AC, Fe: Mn: AC adsorbents, (b): The effect of pH solution on TC removal ability of prepared adsorbents. According to the pKa values of tetracycline in an aqueous environment, tetracycline exists as a cation H4TC+ at pH < 3.4, as an uncharged molecule H3TC0 at pH between 3.4 and 7.7, as an anion H2TC− 1 at pH between 7.7 and 9.7, and as an anion H1TC− 2 at pH above 9.7. In an acidic environment with a pH of less than 3.4, the adsorption efficiency is low due to the repulsion between the protonated functional groups (-OH2+, -COOH2+, M-OH2+) on the positively charged adsorbent surface and the H4TC+ 1 cation. As the pH increases (pH < 3.4 or pH > 7.7), the protonation of the functional groups decreases, reducing the repulsion. This allows tetracycline neutral molecules (H3TC0) to interact more effectively with the functional groups of the adsorbent surface, resulting in an increased adsorption efficiency and TC removal25,31,32. The maximum removal of tetracycline using AC, Fe: AC, and Fe: Mn: AC adsorbents occurs at pH values of 7, 7, and 6, respectively (Fig. 4b). As the pH increases (above the pHzpc of the adsorbents), the adsorption amount decreases due to negative charge (because of the deprotonation) of the adsorbent surface and the presence of tetracycline in its anionic form. Time is the critical parameter impacting the economics of the process. When the process is completed in less time, it becomes more economically viable. Figure 5a shows the amount of TC adsorbed per gram of AC, Fe: AC, Fe: Mn: AC adsorbents (q (mg/g)) over different time intervals. The Fe: Mn: AC adsorbent can remove over 90% of TC within 50 min of contact. The data show that the adsorption process reaches equilibrium in approximately 420 min, 360 min, and 100 min for TC adsorption using AC, Fe: AC, AND Fe: Mn: AC, respectively. At this equilibrium time, the equilibrium adsorption capacities of AC, Fe: AC, and Fe: Mn: AC adsorbents are 161.0 mg/g, 177.60 mg/g, and 196.28 mg/g, respectively. The data indicate a significant reduction in the time required to reach equilibrium in the TC adsorption process by depositing metal nanoparticles on the surface of the activated carbon adsorbent. This improvement can be due to the change in the porous structure of the activated carbon adsorbent and the change in its surface functional groups following the deposition of metal nanoparticles. By depositing iron and manganese metal nanoparticles on the AC surface, the proportion of mesopores increases significantly- by 53% for the adsorbent modified with iron nanoparticles (Fe: AC) and by 98% for the adsorbent modified with two metals (Fe: Mn: AC). The presence of mesopores facilitates the movement of TC molecules and reduces the time to reach equilibrium. The mesopores reduce internal mass-transfer resistance, which facilitates faster adsorption kinetics and desorption processes. This enhances the overall rate of adsorption on mesoporous adsorbents compared to purely microporous adsorbents. The diffusion of adsorbate molecules into the mesopores occurs through surface and Knudsen diffusion mechanisms, which these diffusion mechanisms depend on the adsorbate-surface interactions and pore size. These diffusions in mesopores complement adsorption on the pore walls, contributing to efficient mass transport33–36. Furthermore, the creation of new functional groups, such as F-O, Mn-O, creates more adsorption sites on the surface of activated carbon. This enhances the adsorption of TC molecules through complex formation and Lewis acid-base interactions37. Experimental kinetic data were analyzed with pseudo-first-order (PFO), pseudo-second-order (PSO), and Elovich models to determine the dominant type of interaction. Additionally, to assess the significance of the intra-particle diffusion mass transfer step relative to film diffusion, the experimental data were fitted to the intra-particle diffusion model. Suppose the experimental data fit the intra-particle diffusion model (a regression coefficient near 1 and a plot that passes through the origin). In that case, this suggests that intra-particle diffusion (internal diffusion) plays a more dominant and controlling role compared to film diffusion (external diffusion). The relationships of these kinetic models are presented in Table S2. To fit the experimental data with these models, nonlinear regression was used, taking into account two statistical parameters, the regression coefficient (R2) and the chi-square (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document}), whose relationships are also presented in Table S2. The constants and statistical parameters for the PFO, PSO, Elovich, and intra-particle diffusion kinetic models are presented in Table 2 for TC adsorption using AC, Fe: AC, Fe: Mn: AC adsorbents. Comparison of the values of R2 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} parameters indicates that the kinetics of TC adsorption on AC adsorbent with a micro-mesopore structure is in good agreement with the PFO (pseudo-first-order) and the Elovich models (the qe value predicted by the PFO model is very close to the experimental value). This suggests that pore diffusion and physical adsorption (filling of pores), in addition to chemical adsorption (matching the Elovich model), play important roles in TC adsorption using AC adsorbent38–43. The comparison of statistical parameters of the PFO, PSO, and Elovich models for TC adsorption using Fe: AC and Fe: Mn: AC shows a very good fit of the kinetics experimental data with the Elovich model, as evidenced by the highest R² value and the lowest Chi-square. The strong alignment of experimental data with the Elovich model means that the chemisorption of TC molecules on the heterogeneous surface of Fe: AC and Fe: Mn: AC adsorbents is dominant41,43. The discrepancy between the kinetic data and the intra-particle diffusion model highlights the significance of both external (film) diffusion and internal diffusion stages in the process of TC adsorption through adsorption on the synthesized adsorbents. The distance from the origin value emphasizes the role of the liquid film thickness surrounding the adsorbent particle. Table 2Kinetics parameters for TC removal via adsorption onto AC, fe: AC, fe: mn: AC.Kinetics modelParameterACFe: ACFe: Mn: ACqe, exp (mg/g) = 161.0qe, exp (mg/g) = 177.60qe, exp (mg/g) = 196.28Pseudo-first-order (PFO)k1(min− 1)0.010370.019400.0265qe (mg/g)158.9051167.8052186.9136 R 2 0.97470.89460.8256 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 3.485026.185027.2385Pseudo-second-order (PSO)k2 (g mg− 1min− 1)0.000060.0001530.000244qe (mg/g)192.247187.4045201.0124 R 2 0.93840.93640.8961 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 15.99337.4468019.3457Intra-Particle diffusionkid (mg g− 1 min− 1/2)8.55774.39008.5577Ɵ (mg/g)49.811064.758949.8110 R 2 0.57820.80620.6534 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 113.077131.652098.1613Elovichβ (g/mg)0.02330.030980.0388α (mg.g− 1.s− 1)4.3319.50211 R 2 0.96950.97660.9818 ꭓ 2 3.76172.96205.006 Table 2Kinetics parameters for TC removal via adsorption onto AC, fe: AC, fe: mn: AC.Kinetics modelParameterACFe: ACFe: Mn: ACqe, exp (mg/g) = 161.0qe, exp (mg/g) = 177.60qe, exp (mg/g) = 196.28Pseudo-first-order (PFO)k1(min− 1)0.010370.019400.0265qe (mg/g)158.9051167.8052186.9136 R 2 0.97470.89460.8256 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 3.485026.185027.2385Pseudo-second-order (PSO)k2 (g mg− 1min− 1)0.000060.0001530.000244qe (mg/g)192.247187.4045201.0124 R 2 0.93840.93640.8961 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 15.99337.4468019.3457Intra-Particle diffusionkid (mg g− 1 min− 1/2)8.55774.39008.5577Ɵ (mg/g)49.811064.758949.8110 R 2 0.57820.80620.6534 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 113.077131.652098.1613Elovichβ (g/mg)0.02330.030980.0388α (mg.g− 1.s− 1)4.3319.50211 R 2 0.96950.97660.9818 ꭓ 2 3.76172.96205.006 Kinetics parameters for TC removal via adsorption onto AC, fe: AC, fe: mn: AC. The α (mg.g− 1.s− 1) parameter in the Elovich kinetic model represents the initial rate of adsorption. A comparison of this parameter for AC, Fe: AC, and Fe: Mn: AC adsorbents indicates a significant increase in the initial rate of TC adsorption with the Fe: Mn: AC adsorbent compared to the Fe: AC and AC adsorbents. The enhancement of the initial adsorption rate for TC using Fe: Mn: AC is primarily due to the significant increase in mesopores, which provide larger pathways for TC molecules to move. Examining the adsorption equilibrium and identifying a suitable equilibrium model to describe the equilibrium behavior of the adsorption process is a critical aspect of the adsorption process. To achieve this, the qe (mg/g) values obtained from the equilibrium experiments (investigating the effect of the initial concentration of TC at equilibrium time) are analyzed with the well-known equilibrium models. Figure 5b presents the equilibrium experimental q (mg/g) for TC adsorption using Fe: Mn: AC at three temperatures. The Langmuir, Freundlich, and R-D models were applied to analyze the data, and the relations of these equilibrium models are presented in Table S2. To fit the experimental data with these models, nonlinear regression was used, taking into account two statistical parameters, the regression coefficient (R2) and the chi-square (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document}). The constant parameters and statistical parameters for the equilibrium models are presented in Table 3. Fig. 5(a): The effect of adsorption time on the adsorption ability of AC, Fe: AC, and Fe: Mn: AC adsorbents for TC removals, (b): The equilibrium isotherms of TC adsorption at 25 °C, 35 °C, 45 °C using Fe: Mn: AC adsorbent, (c): Ln (KL) versus 1/T for TC adsorption using Fe: Mn: AC adsorbent. (a): The effect of adsorption time on the adsorption ability of AC, Fe: AC, and Fe: Mn: AC adsorbents for TC removals, (b): The equilibrium isotherms of TC adsorption at 25 °C, 35 °C, 45 °C using Fe: Mn: AC adsorbent, (c): Ln (KL) versus 1/T for TC adsorption using Fe: Mn: AC adsorbent. Comparing the statistical parameters in Table 3 and observing Fig. 5b show that the equilibrium data of TC adsorption align well with the Langmuir model. Considering the comparison of equilibrium data of TC adsorption using Fe: Mn: AC with the Langmuir model and the assumptions of this model, it can be concluded that the adsorption occurs in a single layer. The Fe: Mn: AC adsorbent’s maximum adsorption capacities (qmax) for TC adsorption, predicted by the Langmuir model, was 463.040 mg/g. Table 4 presents the maximum adsorption capacities of some adsorbents applied to TC adsorption1,2,19,31,32,37,44–46. Table 3Constant parameters of equilibrium models for TC adsorption using fe: mn: AC.IsothermParameters25 °C35 °C45 °C Langmuir KL (L/mg)0.18830.30940.5251qmax (mg/g)463.040482.7358496.5481R20.97430.97020.9694 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 10.742013.498015.7993 Freundlich KF ((mg/g)(L/mg)1/n)1.45771.50491.8044n0.10740.0780.072R20.49140.60690.5963 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 513.8850236.7401404.9120 D-R E (kJ/mol)9.62208.8669.622qmax (mmol/g)395.063404.5202422.2752R20.85430.81460.8566 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 80.7820107.986768.9390 Table 3Constant parameters of equilibrium models for TC adsorption using fe: mn: AC.IsothermParameters25 °C35 °C45 °C Langmuir KL (L/mg)0.18830.30940.5251qmax (mg/g)463.040482.7358496.5481R20.97430.97020.9694 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 10.742013.498015.7993 Freundlich KF ((mg/g)(L/mg)1/n)1.45771.50491.8044n0.10740.0780.072R20.49140.60690.5963 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 513.8850236.7401404.9120 D-R E (kJ/mol)9.62208.8669.622qmax (mmol/g)395.063404.5202422.2752R20.85430.81460.8566 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\chi^2$$\end{document} 80.7820107.986768.9390 Constant parameters of equilibrium models for TC adsorption using fe: mn: AC. Table 4Some adsorbents used for TC adsorption1,2,19,31,32,3744,45,46.AdsorbentOperation conditionqmax (mg/g)ReferencesMagnetic biocharC01: 100–800 mg L− 1; Dosage2: 0.5 g/L; Temperature: 45 °C, pH = 41152 1 Fe-loaded porous hydrothermal biocharsC0 = 30–300 mg L− 1; Dosage: 0.2 g/L; Temperature: 25 °C, pH = 5606.52 2 Magnetic activated carbonC0=--- mg L− 1; Dosage: --g/L; Temperature: 25 °C, pH = 414 19 Activated carbon/Fe3O4/ZIF-8C0 = 20–100 mg L− 1; Dosage: 1 g/L; Temperature: 25 °C, pH = 757.47 31 Halloysite nanotubesC0=--- mg L− 1; Dosage: 1 g/L; Temperature: 55 °C, pH = 790 32 Ce-based tubular boron carbon nitrideC0 = 50–300 mg L− 1; Dosage: 4 g/L; Temperature: 25 °C, pH = 3294.1 37 Modified waste printed circuit boardsC0 = 20–200 mg L− 1; Dosage: 2 g/L; Temperature: ---, pH = 794.22 44 Reed biocharC0=--- mg L−1; Dosage: 0.25 g/L; Temperature: 25 °C, pH = 5776.21 45 Carbonaceous materials derived from Brassica oleracea plantC0 = 10–200 mg L− 1; Dosage: --g/L; Temperature: 25 °C, pH = 3163.93 46 Fe: Mn: ACC0 = 25–300 mg L− 1; Dosage: 0.5 g/L; Temperature: 25 °C, pH = 6463.040This work1Initial concentration.2Adsorbent dosage. Table 4Some adsorbents used for TC adsorption1,2,19,31,32,3744,45,46.AdsorbentOperation conditionqmax (mg/g)ReferencesMagnetic biocharC01: 100–800 mg L− 1; Dosage2: 0.5 g/L; Temperature: 45 °C, pH = 41152 1 Fe-loaded porous hydrothermal biocharsC0 = 30–300 mg L− 1; Dosage: 0.2 g/L; Temperature: 25 °C, pH = 5606.52 2 Magnetic activated carbonC0=--- mg L− 1; Dosage: --g/L; Temperature: 25 °C, pH = 414 19 Activated carbon/Fe3O4/ZIF-8C0 = 20–100 mg L− 1; Dosage: 1 g/L; Temperature: 25 °C, pH = 757.47 31 Halloysite nanotubesC0=--- mg L− 1; Dosage: 1 g/L; Temperature: 55 °C, pH = 790 32 Ce-based tubular boron carbon nitrideC0 = 50–300 mg L− 1; Dosage: 4 g/L; Temperature: 25 °C, pH = 3294.1 37 Modified waste printed circuit boardsC0 = 20–200 mg L− 1; Dosage: 2 g/L; Temperature: ---, pH = 794.22 44 Reed biocharC0=--- mg L−1; Dosage: 0.25 g/L; Temperature: 25 °C, pH = 5776.21 45 Carbonaceous materials derived from Brassica oleracea plantC0 = 10–200 mg L− 1; Dosage: --g/L; Temperature: 25 °C, pH = 3163.93 46 Fe: Mn: ACC0 = 25–300 mg L− 1; Dosage: 0.5 g/L; Temperature: 25 °C, pH = 6463.040This work1Initial concentration.2Adsorbent dosage. Some adsorbents used for TC adsorption1,2,19,31,32,3744,45,46. 1Initial concentration. 2Adsorbent dosage. The data indicate the acceptable ability of Fe: Mn: AC adsorbent for TC removal, and it can be considered for use on an industrial scale for removing pharmaceutical contaminants from aqueous media. Generally, increasing the initial concentration of a solution leads to a higher adsorption capacity of the adsorbent. Additionally, reducing the amount of adsorbent per unit volume of solution also enhances the adsorption capacity. To determine the thermodynamic nature of the adsorption process, the equilibrium experiments were performed at three temperatures (as shown in Fig. 5b). KL (Langmuir constant parameter) of each temperature was used to find the Gibbs free energy changes (ΔG° (kJ/mol), Eq. (3)) as well as the enthalpy (ΔH° (kJ/mol)) and entropy (ΔS° (kJ/(mol.K))) changes (Eq. (4) and Fig. 5c). The thermodynamic properties of adsorption are listed in Table 5. The results show that the TC adsorption processes using Fe: Mn: AC adsorbent are spontaneous (negative Gibbs free energy changes) and endothermic (positive enthalpy changes).3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{G}^{^\circ\:}=-RTln{K}_{L}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{ln}{K}_{L}=\frac{\varDelta\:{S}^{^\circ\:}}{R}-\frac{\varDelta\:{H}^{^\circ\:}}{RT}\:\:$$\end{document} Table 5Thermodynamic parameters for TC adsorption using fe: mn: AC.Temperature (K)KL (L/mg)KLΔG◦ (kJ/mol)ΔH◦ (kJ/mol)ΔS◦ (kJ/mol. K) R 2 298.150.18894−11.262040.46750.17340.9987308.150.3094154.7−12.9160318.150.525262.50−14.7338 Table 5Thermodynamic parameters for TC adsorption using fe: mn: AC.Temperature (K)KL (L/mg)KLΔG◦ (kJ/mol)ΔH◦ (kJ/mol)ΔS◦ (kJ/mol. K) R 2 298.150.18894−11.262040.46750.17340.9987308.150.3094154.7−12.9160318.150.525262.50−14.7338 Thermodynamic parameters for TC adsorption using fe: mn: AC. Regeneration of fe: mn: AC adsorbent through continuous (Column) process The regeneration and reuse of the adsorbent (Fe: Mn: AC) in a continuous process was investigated in five cycles. Figure 6 displays the breakthrough curves resulting from the continuous tests using Fe: Mn: AC bed. The breakthrough time t0.05 (the time at C/C0 = 0.05) and the total amount of adsorbed TC (qtot, Eq. (5)) for five cycles are listed in Table 6. Also, the regeneration efficiency percentages of the adsorption cycle were determined according to Eq. (6), and presented in Table 6.5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{q}_{tot}=\frac{Q}{1000}{\int\:}_{t=0}^{{t}_{tot}}{C}_{ads}dt=\frac{Q}{1000}{\int\:}_{0}^{{t}_{tot}}\left({C}_{0}-C\right)dt$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Regeneration\:efficiency\:percentage\:\left(REc\:\mathrm{\%}\right)\:for\:cycle\:i=\frac{{q}_{tot,i}}{{q}_{tot,1}}$$\end{document} The data show a 20.3% decrease in breakthrough time and a 27.5% decrease in the amount of adsorbed TC. Additionally, the regeneration efficiency percentage decreases from 89% to 72% from cycle 2 to cycle 5. These declines in capacity and regeneration efficiency can be attributed to several factors, including the unavailability of active adsorption sites and porosity caused by incomplete desorption of contaminants from the adsorbent surface, as well as changes in the morphology and porosity of the adsorbent after multiple adsorption cycles, as shown in Fig. 7a. Fig. 6The breakthrough curves of five cycles for TC adsorption using Fe: Mn: AC adsorbent. The breakthrough curves of five cycles for TC adsorption using Fe: Mn: AC adsorbent. Figure 7b, c shows the adsorption/desorption isotherms and BJH diagram of Fe: Mn: AC after five cycles, indicating a reduction in specific surface area from 391 m²/g to 281.2 m²/g and a decrease in total pore volume from 0.6504 cm3g to 0.5411 cm3/g. According to the BJH diagram, the mesoporous structure of Fe: Mn: AC adsorbent has changed to meso-macro porous after five cycles. Table 6The performance parameters of fe: mn: AC bed in five cycles.Operation condition:t0.05 (minutes)qtot (mg)REc (%)Cycle 18101350.006-Cycle 27701214.9260.899941185Cycle 37331101.3840.815836374Cycle 46901013.4600.750707775Cycle 5645978.7540.724999741 Table 6The performance parameters of fe: mn: AC bed in five cycles.Operation condition:t0.05 (minutes)qtot (mg)REc (%)Cycle 18101350.006-Cycle 27701214.9260.899941185Cycle 37331101.3840.815836374Cycle 46901013.4600.750707775Cycle 5645978.7540.724999741 The performance parameters of fe: mn: AC bed in five cycles. Fig. 7(a): SEM image of surface morphology, (b): The adsorption/desorption isotherms of N2, (c): BJH diagram for Fe: Mn: AC after 5 cycles of TC adsorption. (a): SEM image of surface morphology, (b): The adsorption/desorption isotherms of N2, (c): BJH diagram for Fe: Mn: AC after 5 cycles of TC adsorption. Adsorption mechanism Figure 8a displays the FTIR spectra of Fe: Mn: AC adsorbent before and after TC adsorption. The changes are obvious. The peak intensity and position have shifted considerably, and new peak has appeared. These changes in the peak strength and location of surface functional groups (including -COOH, -OH, Fe-O, Mn-O) reveal the chemical interactions between the surface functional groups of Fe: Mn: AC adsorbent and the molecules of tetracycline. The functional groups on the surface of activated carbon (-OH and -COOH) attract TC molecules by forming hydrogen bonds with the -OH and -NH2 groups in the TC structure47. Variations in the position and intensity of peaks associated with -OH, C-O, and C = O functional groups in the FTIR spectrum of Fe: Mn: AC following TC adsorption can indicate this chemical interaction (confirmed by the Elovich model). The aromatic (graphite-like) rings in the activated carbon structure attract TC molecules by creating a π-π interaction with the benzene ring in the TC structure30,47. Metal functional groups (Fe-O, Mn-O) can also form surface complexes with organic compounds (surface complexation) such as TC molecules1,2. Also, these metal centers (Fe and Mn) on the adsorbent surface act as Lewis acids and adsorb TC molecules by creating Lewis acid-base interactions with the oxygen of the -C = O, -OH groups (as Lewis bases) in the structure of tetracycline molecules37. The shift and change in the intensity of the peaks of these metal functional groups in the Fe: Mn: AC FTIR spectrum after adsorption confirm these interactions (confirmed by the Elovich model). Table 7 displays the changes in the characteristic peaks. Another mechanism of TC adsorption with the mesoporous Fe: Mn: AC adsorbent is pore filling (physical adsorption). To investigate the importance of pore diffusion (pore filling), a BET analysis was performed on the Fe: Mn: AC adsorbent after TC adsorption. Reports indicate that the characteristic peak of TC is found at a wavelength of 1452 cm− 1, which is also evident in the FTIR spectrum of Fe: Mn: AC after TC adsorption1. Figure 8b presents the nitrogen adsorption-desorption isotherm diagram following the adsorption of TC. When compared to the adsorption-desorption diagrams of the adsorbent prior to TC adsorption, there is a noticeable decrease in surface porosity and specific surface area, with values dropping from 391 m²/g to 125 m²/g. Additionally, the total pore volume decreased from 0.65 cm³/g to 0.3415 cm³/g. This decrease in specific surface area and total pore volume indicates the diffusion of TC molecules into the adsorbent mesoporous structure and occupation of the pores. Figure 8c presents the main adsorption mechanisms for TC molecules adsorption using Fe: Mn: AC adsorbent. Table 7The peaks shifts in the FTIR spectra.Functional groupPeak before TC adsorption (cm− 1)Peak after TC adsorption (cm− 1)-OH34263368-C = O16261607-C-O1361, 10721301, 1037-Fe-O, -Mn-O450, 723, 573765, 705, 565 Table 7The peaks shifts in the FTIR spectra.Functional groupPeak before TC adsorption (cm− 1)Peak after TC adsorption (cm− 1)-OH34263368-C = O16261607-C-O1361, 10721301, 1037-Fe-O, -Mn-O450, 723, 573765, 705, 565 The peaks shifts in the FTIR spectra. Fig. 8(a): FTIR spectra of Fe: Mn: AC before and after adsorption process, (b): N2 Adsorption/desorption isotherms for Fe: Mn: AC before and after TC adsorption, (c): Possible adsorption mechanisms. (a): FTIR spectra of Fe: Mn: AC before and after adsorption process, (b): N2 Adsorption/desorption isotherms for Fe: Mn: AC before and after TC adsorption, (c): Possible adsorption mechanisms. Cost analysis To analyze the costs, the focus was placed on direct operating costs48. These costs include consumables such as biomass, chemical supplies, water for washing, and electricity for the furnace, drying, and stirring operations involved in synthesizing the adsorbents. In the adsorption process section, the operating costs encompass water, electricity, and chemicals used during the process. This includes hydrochloric acid and sodium hydroxide for pH adjustment, as well as hydrochloric acid for the desorption process. Detailed information about these items is provided in Table 8. The cost of consumable chemicals was obtained from the Merck Company website (November 2025), while water and electricity costs were calculated using industrial unit rates in Iran for 2024–2025. Table 8Direct operating costs for TC removal in this research work (basedon the laboratory scale).ProcessConsumable materialAmount usedPrice per unit (USD)Total cost (USD)Adsorbent SynthesisBiomass0.5 kG0.014 USD/kG0.007NaOH0.016 kG68.27 USD/kG1.092FeSO4.7H2O0.021 kG133.07 USD/kG2.794KMnO40.006 kG146.95 USD/kG0.8817Water (for washing)0.003 m³0.0057 USD/m30.000017Energy (electricity)14 KWh0.0143 USD/KWh0.2002Adsorption processWater0.012 m30.0057 USD/m30.0000684Energy (electricity)2.1 KWh0.0143 USD/KWh0.0300HCl (37%)20 mL45.59 USD/Lit0.9118NaOH0.010 kG68.27 USD/kG0.6827 Table 8Direct operating costs for TC removal in this research work (basedon the laboratory scale).ProcessConsumable materialAmount usedPrice per unit (USD)Total cost (USD)Adsorbent SynthesisBiomass0.5 kG0.014 USD/kG0.007NaOH0.016 kG68.27 USD/kG1.092FeSO4.7H2O0.021 kG133.07 USD/kG2.794KMnO40.006 kG146.95 USD/kG0.8817Water (for washing)0.003 m³0.0057 USD/m30.000017Energy (electricity)14 KWh0.0143 USD/KWh0.2002Adsorption processWater0.012 m30.0057 USD/m30.0000684Energy (electricity)2.1 KWh0.0143 USD/KWh0.0300HCl (37%)20 mL45.59 USD/Lit0.9118NaOH0.010 kG68.27 USD/kG0.6827 Direct operating costs for TC removal in this research work (basedon the laboratory scale). Conclusion In this study, hawthorn kernels were transformed into a porous mesoporous adsorbent with multiple functional groups through a single-step dry pyrolysis process. This process utilized iron and manganese metal salts along with sodium hydroxide. The resulting bimetallic Fe-Mn-impregnated activated carbon (Fe: Mn: AC) was tested for its ability to remove tetracycline (TC) from aqueous solutions. The kinetic data fitting using the Elovich model indicated that chemical adsorption played a significant role. Furthermore, the equilibrium data aligned well with the Langmuir model. Various chemical interactions, including surface complexation, hydrogen bonding, Lewis acid-base interactions, π-π interactions, and physical adsorption, contributed to the adsorption of TC molecules by the Fe: Mn: AC adsorbent. Results from continuous mode experiments over five cycles demonstrated the high adsorption capacity of the Fe: Mn: AC adsorbent for effectively removing pharmaceutical contaminants from water, with only a 27.5% reduction in adsorption capability observed. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1