Sunlight-activated carbon nanotubes for anionic dye removal: characterization, efficiency, and biological assessment against some aquaculture and human pathogenic bacteria

Abstract

Ammonia-functionalized multi-walled carbon nanotubes (MWCNTs/NH₃) were synthesized and evaluated for the sunlight-enhanced removal of Acid Red 73 (AR73) dye, as well as to assess its potential application against aquaculture and human pathogenic bacteria. Characterization of FTIR, SEM, and EDX confirmed a porous structure (70–110 nm diameters). FTIR confirmed functional groups (O-H, N-H, C ≡ C, C = C) critical for dye adsorption. SEM revealed a porous, entangled MWCNT structure with high surface area (70–110 nm tube diameters). EDX showed 88.10% carbon and 10.59% oxygen, with trace Si, Ca, and Fe. TGA indicated ammonia decomposition (100–250 °C) and CNT oxidation (300–900 °C). Batch experiments demonstrated rapid removal (97% within 30 min) and equilibrium at 180 min. Optimal conditions included pH 5 (99.16% efficiency) and 0.02 g/L dosage (100% removal), though adsorption capacity decreased at higher doses. The Langmuir isotherm (R²=0.939) revealed monolayer adsorption with a high capacity (312.5 mg/g), while pseudo-second-order kinetics (R²=0.999) indicated chemisorption dominance. Electrostatic interactions between protonated NH₃⁺ and anionic dye, hydrogen bonding, and π-π stacking were key mechanisms. Sunlight further enhanced removal via sunlight-enhanced adsorption. Despite limited antibacterial activity (low sensitivity for Staphylococcus aureus and Vibrio fluvialis), the composite showed no broad-spectrum efficacy. This study highlights ammonia-functionalized MWCNT as a sustainable, energy-efficient adsorbent for high-capacity dye removal, leveraging solar energy to mitigate industrial wastewater pollution.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49163-1.

Introduction

The rapid growth of industries and urbanization has resulted in significant environmental challenges, particularly concerning the excessive pollution of water bodies. Among these pollutants, synthetic dyes pose a substantial threat due to their toxicity, carcinogenicity, and mutagenicity¹. Industries such as textile, food processing, leather, makeups, paper production, and dye manufacturing generate substantial quantities of dyes, contributing significantly to water pollution². These are artificial aromatic compounds that have different functional groups incorporated into them³. The removal of anionic dyes, which are highly water-soluble, acidic, and impart intense coloration to water, is of paramount concern^4,5. The industries of textile, cosmetics, food processing, leather, dyeing, paper, and dye manufacturing all generate considerable quantities of dyes, which are substantial pollutants of water^6,7. These are artificial aromatic molecules that have different functional groups incorporated into them. Ecosystem health, marine life, soil fertility, and water resources are all negatively impacted by the harmful effluents released by different enterprises⁷. One of the main issues facing the textile industry is the way color appears in effluents. Out of all the industrial sectors, textile waste water is considered to be the most polluting⁸. Textile dyes, in particular, contribute to a reduction in light transmission, increased chemical oxygen demand and biological (COD and BOD), and ultimately, disrupt aquatic ecosystems^9,10.

The elimination of artificial dyes from wastewater is now a significant environmental issue because of their toxicity, and lack of biodegradability, consequently, photosynthetic activity is reduced, with damaging effects for the whole aquatic ecosystem^11,12. Acid Red 73, a commonly used azo dye, poses significant threats to aquatic life and human health¹³. Therefore, developing efficient and environmentally friendly techniques for dye removal is imperative¹⁴. Different techniques have been utilized for treating wastewater, such as chemical, biological, and physical methods⁶. Adsorption, a cost-effective and easily adaptable technique, has emerged as a promising strategy for removing pollutants from industrial wastewater^7,15. Hence, this approach has numerous benefits compared to other methods like flocculation, oxidation, reverse osmosis, electrolysis, precipitation, coagulation, and membrane filtration because it avoids generating damaging by-products that may worsen environmental pollution^16,17. Using adsorbents made from plentiful and sustainable sources has become a promising option for treatment methods, particularly adsorption. MWCNTs have attracted significant interest as possible adsorbents owing to their chemical stability, great adsorption capacity, and simple separation with a magnetic field¹⁴. The synthesis methods and ensuing factors have a major impact on the characteristics and kind of CNT structure. Because of their shape, hollow structure, unique network of carbon sp2 atoms and high surface-to-volume ratio, they can be used as sorbents, membranes or filters, for a variety of separation and enrichment applications. Due to MWCNTs possessing a greater surface area, they are able to more efficiently engage with analyses¹⁸. Through a variety of interactions, including the hydrophobic effect, electrostatic interactions, π-π, and covalent bonding, CNTs strongly adsorb a number of contaminants¹⁹. Carbon nanotubes enable the elimination of dyes and metal ions from wastewater, as well as their preconcentration, separation, and speciation analysis^20–22. However, their adsorption performance can be further enhanced by introducing functional groups that promote specific interactions with target pollutants¹². So, this study investigates the sorption of Acid Red 73 dye onto ammonia- MWCNTs (MWCNTs/NH₃) under direct sunlight exposure. The use of direct sunlight eliminates the need for external energy sources, making the process more sustainable.

Unlike conventional CNT-based adsorbents reported for azo dye removal, the present study introduces ammonia-doped MWCNTs designed to provide positively charged amine functionalities specifically targeting anionic dyes. Moreover, the removal process operates under direct sunlight and static conditions, integrating adsorption and sunlight-assisted photodegradation without external energy input. This combined photoadsorptive mechanism, together with high adsorption capacity and recoverability, distinguishes the current system from previously reported CNT/composite adsorbents. Moreover, the work provides a comprehensive mechanistic interpretation combining electrostatic attraction, π–π interactions, hydrogen bonding, and sunlight-assisted photodegradation. The inclusion of antibacterial assessment further extends the environmental relevance of the proposed material, highlighting both its adsorption efficiency and biological limitations.

This study focuses on the removal of AR73, a representative anionic dye, by ammonia- multi-walled carbon nanotubes (MWCNTs) under static conditions (no agitation) in direct sunlight by batch adsorption experiments. The investigation aims to elucidate the adsorption kinetics, and mechanism of Acid Red 73 removal from synthetic wastewater using ammonia-doped magnetic MWCNTs. Exploration will be conducted on the impact of various factors including initial AR73 dye concentration, adsorption time, temperature, and pH. Moreover, the research will utilize various methods such as electrokinetic and static contact angle measurements, FTIR, EDX, and SEM to understand how AR73 is adsorbed onto the surface of ammonia-doped magnetic MWCNTs.

Materials and methods

Synthesis of MWCNTs by chemical vapor deposition (CVD) doped with ammonia

The process begins with preparing a catalyst solution created by dissolving ferrocene in ethanol, a suitable solvent at a specific concentration. Then, a suitable carbon source gas methane is chosen²³. Ammonia, used for doping, is prepared from a diluted ammonium hydroxide solution. Finally, silicon wafers is selected and thoroughly cleaned to ensure a pure growth environment. Next, the catalyst solution is deposited onto the pre-cleaned substrate using spin-coating or dip-coating to achieve an optimal catalyst distribution and density. The coated substrate is then placed in a CVD furnace with a controlled atmosphere of an inert gas like argon or nitrogen. The furnace is heated to a high temperature (typically between 700 and 1000 °C) and maintained at a specific pressure (1 atm).

The chosen carbon source gas is introduced into the furnace at a controlled flow rate to initiate the growth of MWCNTs. The growth time is carefully adjusted to achieve the desired MWCNT length. Once the growth process is complete, the furnace is gradually cooled to ambient temperature with the inert gas atmosphere remaining to avoid any undesired chemical responses²⁴. After growing through CVD, the MWCNTs are subjected to ammonia doping to add magnetic characteristics. This includes soaking the MWCNTs in the diluted ammonia solution that has been prepared at a set temperature (40–60 °C) for a specified duration (3 h). This enables ammonia molecules to spread out and stick to MWCNTs’ surfaces. Following doping, the MWCNTs undergo a thorough rinsing with deionized water to eliminate any leftover ammonia that was not fully absorbed. Ultimately, the MWCNTs are dehydrated in a vacuum oven at 60 °C to eliminate any remaining water molecules, finishing up the synthesis procedure²⁴. Multi-walled carbon nanotubes were modified using ammonium hydroxide (NH₄OH) via wet impregnation followed by drying. This treatment introduces surface-bound ammonia/amine species through adsorption and interaction with defect and oxygen-containing sites on the CNT surface. The modification does not involve high-temperature nitrogen incorporation into the carbon lattice; therefore, the material is referred to as NH₄OH-functionalized MWCNTs rather than nitrogen-doped MWCNTs.

Preparation of AR73 dye solution

The molecular formula, molecular weight, and C.I. of AR73 dye are detailed in Table 1S. The dye was purchased from Sigma-Aldrich (Germany) and utilized as is, without any alterations. In order to prepare the stock solution, the dye was first calculated to be 1000 mg/L and then diluted in 1.0 L of distilled water. A dye solution was prepared for the batch investigation by distilling the dye stock in distilled water. Solutions of NaOH and HCl were utilized to adjust the pH.

Table 1.

Elemental composition of MWCNTs doped with ammonia nanoparticles.

Element	Mass%	Atom%
C	88.10 ± 0.28	91.32 ± 0.29
O	10.59 ± 0.36	8.24 ± 0.28
Si	0.36 ± 0.03	0.16 ± 0.02
Ca	0.83 ± 0.05	0.26 ± 0.02
Fe	0.12 ± 0.04	0.03 ± 0.01
Total	100.00	100.00

Batch adsorption experiments

The impact of various parameters on the adsorption of Acid Red 73 by ammonia-doped magnetic MWCNTs was investigated using batch adsorption experiments under static conditions (no agitation) in direct sunlight. Prior to illumination, the suspension containing the dye and adsorbent was magnetically stirred in the dark for 30 min to establish adsorption–desorption equilibrium. This step ensures that the subsequent dye removal under illumination is attributed predominantly to photocatalytic activity rather than simple adsorption. Additionally, the photostability of the dye under the employed light source in the absence of catalyst has been widely reported in previous studies, showing negligible self-photolysis under similar conditions. The analytical conditions for Acid Red 73 determination have now been clarified in the revised manuscript. The maximum absorbance wavelength (λ_max = 507 nm), quartz cuvette path length (1 cm), and UV–Vis spectrophotometer. A calibration curve (absorbance vs. concentration) was constructed within the studied concentration range, showing excellent linearity (R² = 0.999).

To assess the effect of contact time, 10 mg of ammonia-doped magnetic MWCNTs in 10 mL of solution at a pH of 3 were combined with a sufficient concentration of Acid Red 73 (10 ppm). The efficiency of the dye elimination procedure was evaluated at different intervals of contact time (15, 30, 45, 60, 90, 120, and 180 min). In order to measure the impact of pH, ten milligrams of ammonia-doped magnetic MWCNTs were mixed with a fixed amount of Acid Red 73 (10 ppm) in ten milliliters of solution at different pH values (1, 3, 5, 7, and 9). Weak HCl and NaOH solutions were employed to alter the pH levels. The dye clearance effectiveness was determined after 180 min of interaction. The impact of the adsorbent amount was evaluated by blending varying levels of ammonia-doped magnetic MWCNTs (0.005, 0.01, 0.02, 0.03, and 0.04 g/L) with 10 mL of solution containing a constant concentration of Acid Red 73 (10 ppm) at pH three. A 180-minute contact time was used to calculate dye removal effectiveness. In order to quantify the impact of the starting dye concentration, Different initial concentrations of Acid Red 73 (25, 50, 75, 100 and 200 ppm) were combined with a set weight of ammonia-doped magnetic MWCNTs (10 mg) in 10 mL of solution at a pH of 3. A 180-minute contact time was used to calculate dye removal effectiveness. All adsorption experiments were intentionally conducted under static conditions without mechanical agitation. This experimental design was selected to simulate realistic environmental and wastewater treatment scenarios, such as stabilization ponds, constructed wetlands, and open water bodies, where natural mixing is limited. Moreover, performing the experiments under static conditions allows for the isolated evaluation of the sunlight effect on the adsorption process, without interference from enhanced mass transfer induced by agitation. This approach provides a more practical assessment of the material’s performance under low-energy and environmentally relevant conditions.

In each experiment, the Acid Red 73 removal efficiency was determined spectrophotometrically. This method measures the amount of light absorbed by the dye solution at a specific wavelength. A decrease in absorbance indicates that the dye has been removed from the solution by the MWCNTs. Photodegradation experiments were carried out using a 150 W visible-light lamp positioned approximately 15 cm above the reaction solution. Illumination times varied between 30 and 120 min depending on the experimental conditions. Adsorption experiments were conducted at ambient laboratory temperature (25 ± 2 °C) under static (non-agitated) conditions. Although the absolute irradiance and spectral distribution were not measured, the same illumination setup was used throughout to ensure reliable comparative assessment.

Adsorption Models

Adsorption isotherms describe the distribution of adsorbate molecules among the solid and liquid phases at equilibrium, offering insight into the sorption process and its dynamics²⁵. The Langmuir, Tempkin, and Freundlich models are frequently utilized to explain adsorption in solid-liquid systems. Linear regression was employed to facilitate comparison with previously reported adsorption systems for similar materials. Although nonlinear fitting can reduce potential bias associated with linearization, the linear approach was considered sufficient for comparative evaluation within the scope of this study.

The Langmuir model posits that a monolayer of adsorbate is created on the surface of the adsorbent. It suggests that adsorption takes place at a constant amount of clearly defined locations, each with the ability to capture one adsorbate molecule. All sites are viewed as having the same energy level, and there is no interaction among the molecules that have been adsorbed²⁶. Examining adsorption isotherm information is essential in determining the adsorption capacity of the adsorbent. Equation (1) represents the Langmuir isotherm, which is characterized by the following parameters.

1

C_e represents the dye ion’s concentration at equilibrium in milligrams per liter, K_a is the Langmuir constant in liters per milligram, and Q_m is the maximum monolayer coverage capacity in milligrams per gram. q_e is the quantity of dye adsorbed per unit mass of adsorbent in milligrams per gram at equilibrium. Analyzing the linear plot of q_e/C_e vs. q_e allows for the determination of these constants.

The Freundlich model explains adsorption on surfaces that are not uniform, with adsorption capacity changing depending on the equilibrium concentration of molecules. Frequently, this model is represented in a linear format.

2

Where: the intercept, log K_F, is a measure of adsorbent capacity, the slope 1/n is the sorption intensity, K_f = Freundlich isotherm constant (mg/g), and n = adsorption intensity. These constants can be determined by analyzing the linear plot of Log q_e vs. Log C_e.

3

The Tempkin constant is represented by A_T (L g^− 1), B_T = (RT)/b, with T being the absolute temperature in Kelvin and R being the universal gas constant, 8.314 J/mol K. The constant b is associated with the heat of adsorption. Analyzing the linear plot of lnC_e vs. q_e allows for the calculation of these factors.

Kinetics of adsorption

In order to study the biosorption kinetics of AR73 onto the Ammonia-Doped Magnetic MWCNTs. Unless otherwise specified, all adsorption and kinetic experiments were performed using an adsorbent dosage of 0.02 g·L⁻¹ (0.2 g in 10 L equivalent; e.g., 0.2 mg in 10 mL), magnetic stirring at ~ 300 rpm, and ambient temperature (25 ± 2 °C). The kinetic experiments were recalculated to match these standardized conditions for consistency across all datasets.

The general expression for the pseudo-first order equation (Lagergren, 1898) is as follows:

4

q_t represents the quantity of adsorbate adsorbed at time t in milligrams of adsorbate per gram of adsorbent (mg/g). q_e represents the quantity of adsorbate adsorbed at equilibrium in milligrams of adsorbate per gram of adsorbent (mg/g), and k1 is the rate constant of the pseudo-first-order adsorption process in reciprocal minutes (1/min). The values of k1 and qe are found using the slope and intercept of the linear graph of ln(q_e - q_t) versus time (t). The linear form of the Ho equation, which is a pseudo-second-order kinetic model, is as follows:

5

q_e and k₂ values are derived from the slope and intercept of the linear graph of t/qt plotted against time (t). K₂ (g/mg/min) represents the second-order sorption rate constant, which can be reorganized and linearized to achieve. The formula for intraparticle diffusion is investigated through the Equation provided below:

6

Intraparticle diffusion rate constant K_dif (mg/g/min) and boundary layer thickness related constant C (mg/g) are determined from the q_t versus t^1/2 plots through calculation of their values from the slope and intercept.

The Elovich isotherm model is employed for the purpose of explaining the adsorption rate as it evolves. It is frequently employed in situations where chemisorption is the primary process, involving a powerful chemical bonding between the adsorbate and the adsorbent. It suggests a straight correlation between the natural logarithm of qt and the time variable t.

7

Where: constants associated with desorption rate (β) and Initial adsorption rate (α) are provided.

Characterization

FTIR spectroscopy was utilized to analyze the physicochemical characteristics of the MWCNTs/NH₃ composite by scanning the sample within the range of 4000 to 400 1/cm to detect characteristic peaks related to the functional groups in the composite. However, distinctive characteristics of the MWCNTs/NH₃ material can be determined by conducting EDX analysis with the Japanese JEOL JSM-IT200 instrument and examining it with SEM. Thermogravimetric analysis (TGA) was performed under a flowing nitrogen atmosphere to evaluate the thermal stability of pristine and NH₃-functionalized MWCNT composites. Approximately 5–10 mg of sample was heated from room temperature to 800 °C at a heating rate of 10 °C min⁻¹ under a nitrogen flow of 50 mL min⁻¹. The inert nitrogen environment ensures that the observed mass losses are associated with desorption of physically adsorbed species and thermal decomposition of surface functional groups rather than oxidative combustion of the carbon framework. Derivative thermogravimetric (DTG) curves were obtained to identify the temperatures of maximum mass-loss rate.

Assessment of antibacterial activity

To evaluate the biological assessment of Ammonia-Doped Magnetic Carbon Nanotubes against aquaculture and human pathogenic bacteria, the well diffusion technique was employed using Mueller-Hinton agar (MHA) as the culture medium. The diameters of the resulting inhibition zones were recorded in millimeters. A panel of standard bacterial strains was utilized, including Staphylococcus aureus (ATCC 25923), Escherichia coli (ATCC 8731), Pseudomonas aeruginosa (ATCC 9027), Vibrio damsella, Klebsiella pneumoniae (ATCC 13883), Enterococcus faecalis (ATCC 29212),Vibrio fluvialis, Streptococcus agalactiae and Aeromonas hydrophila. Under sterile conditions, the agar plates were inoculated with the bacterial cultures. Wells of 6 mm in diameter were then created and each filled with 50 µL of the nanomaterial sample. The plates were incubated at 37 °C for 24 h. Following incubation, the zones of bacterial growth inhibition surrounding the wells were measured. Bacterial suspensions were prepared from 18 to 24 h-old colonies grown on agar, adjusted to a turbidity equivalent to 0.5 McFarland standard (approximately 1.5 × 10⁸ CFU/mL). The optical density of these suspensions was verified at 600 nm. All experiments were conducted in triplicate to ensure reproducibility.

Results and discussion

Characteristics of adsorbent composites

Fourier transform infrared spectroscopy (FTIR)

FTIR analysis revealed several functional groups on the surface of MWCNTs/NH₃. The FTIR analysis in Fig. 1 shows multiple functional groups present on the surface of the MWCNTs-Ammonia nanoparticles that are important for the potential adsorption of AR73 dye. A broad band in the region of 3300–3500 cm⁻¹ is attributed to N–H and O–H stretching vibrations, indicating the presence of amine and hydroxyl functionalities introduced during ammonia functionalization. The peak at ~ 3756 cm⁻¹, though minor, may arise from isolated silanol groups (Si–OH) due to trace silica contamination, consistent with EDX results, or from adsorbed moisture²⁷. These groups act as potential adsorption sites due to the opposite charges, promoting attraction between the nanoparticles and the dye. FTIR confirms introduction of N–H stretching at 3300–3500 cm⁻¹ and N–H bending ~ 1550–1650 cm⁻¹ containing surface species. Also, the N-H stretching vibration at wave number 3132.109 cm⁻¹ suggests the presence of ammonia groups within the nanoparticle structure²⁸. These groups create potential sites for hydrogen bonding interactions with the Acid Red 73 molecules. The doping of ammonia enhances this interaction potential, contributing significantly to dye absorption. While, the C ≡ C stretching vibration at 2038.908 cm⁻¹²⁹, owing to their high surface area and the presence of π-electrons, carbon nanotubes demonstrate outstanding adsorption capability as a result of their unique structure. These electrons, associated with the triple bond (C ≡ C), participate in strong π-π interactions with the aromatic rings present in Acid Red 73 dye molecules. This type of interaction further contributes to effective dye removal. Also, the C = C stretching vibration found at 1571.193 cm⁻¹, which is like the triple bond, the double bonds (C = C) in the graphitic structure of CNTs also contribute to π-π stacking interactions with the aromatic rings of the Acid Red 73 dye, enhancing its adsorption onto the nanoparticle surface. While sulfonate groups (S = O) typically show strong asymmetric and symmetric stretches in the 1000–1200 cm⁻¹ and 1300–1400 cm⁻¹ regions, the assignment to a specific “S = O…Na” interaction is speculative without XPS or Raman confirmation. This peak may instead reflect C–H bending or nitrate/nitrite impurities from the doping process. The S = O^…Na stretch and C-H bending vibration at 1400.315 cm⁻¹³⁰, associated with the aromatic rings present in CNTs, is not believed to directly contribute to the dye adsorption process. Its presence mainly indicates the structural integrity of the CNTs. Then, C-O stretching vibration was found at 1108.561 cm⁻¹. This functional group could play a role in adsorption through coordination or electrostatic interaction. However, the exact nature and contribution to AR73 adsorption remain unclear. The attendance of hydroxyl groups, N-H vibrations, and the high surface area of the MWCNTs doped with ammonia nanoparticles suggest a good potential for adsorbing AR73 dye. FTIR showed that the anionic nature of AR73 (dye with sulfonic groups) and the occurrence of surface hydroxyl groups (O-H stretching vibration) in the MWCNTs-Ammonia could lead to strong electrostatic interactions. Also, the nitrogen atoms in the ammonia groups (N-H stretching vibration) and their possible involvement in surface doping could facilitate hydrogen bonding with the dye, enhancing its removal. The extended conjugated systems present in both the MWCNTs (C = C, C ≡ C vibrations) and Acid Red 73 allow for π-π stacking interactions, further contributing to dye adsorption. While these interactions strongly suggest potential for dye removal, additional considerations need to be investigated.

Fig. 1.

FTIR analysis of MWCNTs doped with ammonia nanoparticles.

Scanning electron microscopy analysis

SEM analysis provides invaluable insights into the morphology and structure of nanomaterials, aiding in understanding their properties and potential applications. For instance, in this composite, the porous structure observed via SEM is directly linked to the improved surface area, which is a critical parameter in catalysis and adsorption applications. SEM analysis is essential for determining the surface structure and dimensions of nanomaterials. A high-resolution image is generated by scanning the sample’s surface with a focused beam of electrons³¹. The interaction of electrons with the sample produces various signals that reveal information about the sample’s topography, composition, and other properties. Figure 2 reveals to a network of entangled, interconnected structures, indicative of the presence of MWCNTs. Figure 2 exhibits a high degree of surface roughness, suggesting the presence of numerous nano-sized features. The scale bar indicates an image magnification of x35,000, highlighting the ability of SEM to resolve fine details at the nanoscale. The SEM image demonstrates the attendance of entangled MWCNTs. The entanglement of the nanotubes creates a porous structure, which is a characteristic feature of MWCNT composites. This porous structure can enhance the surface area of the composite, potentially improving its performance in applications like adsorption process³². Additionally, the SEM image revealed that the tube diameter ranged from 70 to 110 nm, and the wall thickness varied between 9 and 11 nm. These structural characteristics are important because they influence the overall performance of the composite material, especially in applications that rely on the unique properties of nanoscale materials³³. Additionally, SEM analysis indicates a network of MWCNTs that is porous and interconnected. This arrangement, along with the existence of oxygen-caused functional groups, offers a large surface area and potential adsorption spots for dye molecules.

Fig. 2.

SEM of MWCNTs doped with ammonia nanoparticles.

EDX spectroscopy analysis

The EDX analysis in Fig. 3; Table 1 reveals the following elemental composition (mass% and atom%) in the composite: carbon (C) is 88.10% ± 0.28 (mass) and 91.32% ± 0.29 (atom). The high carbon content is expected due to the presence of MWCNTs, which form the backbone of the composite. This confirms the successful incorporation of MWCNTs into the material. Oxygen (O) is represented 10.59% ± 0.36 (mass) and 8.24% ± 0.28 (atom). The presence of oxygen can be attributed to two main sources: (ammonia doping) which the doping process likely introduced some oxygen into the composite, possibly due to the oxidation of MWCNTs during synthesis or due to the presence of oxygen in the ammonia gas used for doping. In addition (surface functionalization) oxygen is often associated with functional groups on the surface of MWCNTs³⁴. These functional groups can improve the adsorption characteristics of the composite through the provision of sites for interaction with dye molecules. Likewise, silicon (Si) is characterized 0.36% ± 0.03 (mass) and 0.16% ± 0.02 (atom). Calcium (Ca) is signified 0.83% ± 0.05 (mass) and 0.26% ± 0.02 (atom). While Iron (Fe) is 0.12% ± 0.04 (mass) and 0.03% ± 0.01 (atom). The presence of these elements; Silicon (Si), Calcium (Ca), and Iron (Fe), although in low concentrations, suggests potential impurities introduced during the synthesis of the MWCNTs and the doping process. Silicon could originate from the substrate or reaction vessels used in synthesis, while calcium and iron could potentially be related to the magnetic properties of the MWCNTs, as these elements are commonly used in magnetic nanoparticles synthesis³⁵. Therefore, the composite is primarily composed of carbon from the MWCNTs, along with oxygen, likely due to ammonia doping and surface functionalization. The existence of oxygen indicates the existence of functional groups like (OH) groups, (COOH) groups, or carbonyl (C = O) groups on the external of MWCNTs. The dye molecules can enhance the adsorption capacity of the composite by interacting with the functional groups via electrostatic interactions, hydrogen bonding, or Van der Waals forces.

Fig. 3.

EDX Analysis of MWCNTs doped with ammonia nanoparticles.

Thermogravimetric analysis (TGA)

Thermogravimetric analysis (TGA) is a valuable tool to study the thermal decomposition behavior and stability of MWCNTs doped with Ammonia. The TGA data in Table 2; Fig. 4 shows a typical weight loss for a material containing organic components (in this case, ammonia) adsorbed onto an inorganic substrate (carbon nanotubes). The results showed that the initial weight loss (Up to ~ 100 °C), the small weight loss at the beginning indicates the removal of adsorbed moisture and any remaining volatile compounds like solvents used in the synthesis. This step helps establish a dry baseline weight for the sample. Ammonia decomposition at 100–250 °C, this is the most important stage for understanding the effect of ammonia doping³⁶. The substantial weight loss in this range signals the release of ammonia molecules that were adsorbed onto the CNT surface. While CNT oxidation at 300–700 °C, which the more gradual weight loss after the ammonia removal likely indicates the onset of oxidation of the CNT structure³⁷. This process is expected, especially as the temperature increases. Final decomposition occurred in range 700–900 °C, at higher temperatures, the remaining CNT structure is increasingly oxidized, leading to a more rapid weight loss³⁸. This final stage often culminates in the almost complete loss of carbon. A small amount of residual ash (inorganic content from the synthesis process, catalyst, or possibly some ammonia decomposition products) may remain. The high-temperature mass loss above ~ 600 °C is attributed to gradual degradation of the carbon nanotube framework under inert conditions rather than combustion.

Table 2.

Thermogravimetric Analysis (TGA) Data for Ammonia-Doped Multi-Walled Carbon Nanotubes Showing Decomposition and Thermal Stability.

Temperature (°C)	Weight (%)	Decomposition
25	100	Starting weight (undoped CNTs)
50	99.8	Slight loss due to adsorbed moisture
100	99.5	Further slight loss
150	95	Significant weight loss due to ammonia desorption/decomposition
200	90	Continued decomposition
250	85	Major weight loss almost complete
300	82	Gradual weight loss from oxidation
400	78	Continued oxidation
500	70	Significant CNT oxidation and degradation
600	55	Continued oxidation
700	20	Extensive degradation, near complete carbon oxidation
800	5	Almost all carbon is gone, residual ash may remain
900	2	Residual ash

Fig. 4.

Thermogravimetric Analysis (TGA) of Ammonia-Doped Multi-Walled Carbon Nanotubes.

X-ray diffraction (XRD) analysis

To confirm the successful synthesis, doping, and magnetic functionalization of the multi-walled carbon nanotubes (MWCNTs), X-ray diffraction (XRD) analysis was performed. XRD is a crucial technique for elucidating the crystalline structure, phase composition, and degree of graphitization in carbon-based nanomaterials. Moreover, it provides valuable information regarding the incorporation of foreign elements, structural defects, and modifications introduced through chemical doping. The XRD patterns obtained in this study offer insights into the integrity of the graphitic framework, the presence of magnetic phases, and the structural alterations resulting from ammonia doping. The crystalline structure and phase composition of the ammonia-doped magnetic multi-walled carbon nanotubes (MWCNTs/NH₃) were characterized by XRD, and the resulting patterns are shown in Fig. 5. The diffraction pattern exhibited two prominent peaks at 2θ values of 26.27° and 43.03°, which correspond to the (002) and (100) crystallographic planes of graphitic carbon, respectively³⁹. The associated d-spacings of 3.39 Å and 2.10 Å are consistent with the hexagonal structure of MWCNTs⁴⁰. The calculated crystallite sizes at these peaks, ranging between 494 Å and 669 Å, suggest the presence of well-ordered graphitic domains. Furthermore, the relatively low micro strain values (0.34%) indicate minimal structural distortion within the carbon framework, confirming the preservation of the robust sp² hybridized network essential for the nanotube structure. In addition to the typical graphitic reflections, a significant peak was observed at 29.19° (2θ) with a d-spacing of 3.06 Å. This peak does not correspond to the standard reflections of pure carbon and is tentatively assigned to the presence of iron oxide nanoparticles (such as Fe₃O₄ or γ-Fe₂O₃), incorporated to impart magnetic properties to the composite. However, the characteristic reflections of crystalline magnetite, usually observed around 35–36°, were not clearly detected.

Fig. 5.

XRD analysis of MWCNTs/NH₃ adsorbent.

Broad and less intense peaks appearing at 12.94°, 20.85°, and 23.19° (2θ) with corresponding d-spacings of 6.84 Å, 4.26 Å, and 3.84 Å are indicative of structural modifications introduced by ammonia doping. These features suggest the formation of nitrogen-containing functional groups and localized disorder within the graphitic structure. Moreover, the calculated crystallite sizes at these positions were found to be smaller (ranging between 193 Å and 633 Å) with slightly higher micro strain values (0.3% to 1.5%), further confirming the partial disruption of the nanotube lattice due to the chemical modification. The dominance of the 29.19° peak with the highest relative intensity (100%) may also reflect localized regions of enhanced crystallinity or the presence of a secondary phase induced by doping or magnetic particle incorporation. Minor low-intensity peaks at 6.11° and 15.23° could correspond to residual catalyst particles (e.g., Fe, Si) or amorphous carbon phases resulting from the synthesis process. These reflections are consistent with spinel-type iron oxides (Fe₃O₄/γ-Fe₂O₃). Due to the similarity of diffraction patterns between magnetite and maghemite and the absence of high-resolution XRD or TEM analysis, the magnetic phase is conservatively assigned as iron oxide (Fe₃O₄/γ-Fe₂O₃) rather than a single specific phase.

Surface charge characteristics (zeta potential)

The zeta potential of the synthesized composite MWCNTs/NH₃ adsorbent was measured to assess its surface charge behavior in aqueous medium. The material exhibited a negative zeta potential of approximately − 15.3 ± 0.2 mV (triplicate measurements), indicating that the surface carries a net negative charge under the measurement conditions (neutral aqueous environment) (Fig. 6).

Fig. 6.

Surface charge characteristics (zeta potential) of MWCNTs/NH₃ adsorbent.

The negative surface potential is attributed to the presence of deprotonated surface functional groups such as hydroxyl (–O⁻) and/or carboxylate species on the composite surface. Such negatively charged sites play a crucial role in adsorption processes, particularly through electrostatic interactions with cationic species in solution.

The magnitude of the zeta potential ( ≈ − 15 mV) suggests moderate colloidal stability and confirms that electrostatic interactions contribute to dye removal. Under acidic conditions, partial protonation of surface groups is expected to reduce the negative charge magnitude, thereby modifying adsorption behavior. This observation is consistent with the experimentally observed pH-dependent adsorption performance. Although pH_PZC was not directly determined, the negative zeta potential at near-neutral pH indicates that the point of zero charge of the composite is likely located in the acidic pH region (pH < 7). Therefore, at neutral and alkaline pH values, the adsorbent surface remains negatively charged, favoring electrostatic interaction with positively charged dye species.

Finally, the ammonia treatment is expected to introduce surface amine-like functionalities (–NH₂/––NH₃⁺) and protonatable nitrogen species, which increase positive surface charge under neutral to mildly acidic conditions. This promotes electrostatic attraction toward the anionic azo dye (Acid Red 73), enhancing adsorption capacity. Such improvements have been widely reported for ammonia-modified or amine-functionalized CNTs without requiring substitutional nitrogen doping. NH₄OH treatment is known to introduce surface amine-like functionalities and protonatable nitrogen species on carbon nanomaterials through adsorption and weak chemisorption. These groups enhance electrostatic attraction toward anionic dyes such as Acid Red 73, leading to improved adsorption capacity without requiring substitutional nitrogen doping of the CNT lattice.

Adsorption experiments

Effect of contact time

Research was conducted to examine how the duration of contact affects the elimination of AR73 dye from water using MWCNTs/NH₃ as sorbents. The study was carried out in the presence of direct sunlight exposure to examine the possibility of improving the sorption method through photocatalysis. The results indicate a high removal efficiency of AR73 by MWCNTs/NH₃, reaching above 97% within the first 30 min. While, the removal efficiency fluctuates slightly with increasing contact time, it remains consistently high, show the effectiveness of the adsorbent (Fig. 8a). The quick depletion of AR73 within the first 15–30 min is due to the fact that the magnetic (MWCNTs/NH₃) have a large surface area and many active sites because of the various pores, edges, and functional groups created by ammonia doping⁴¹.

Fig. 8.

Proposed adsorption mechanism of Acid Red dye onto AC-ZnO-NH₃ surface.

These features allow for the efficient adsorption of dye molecules. The observed slight fluctuations in removal efficiency at longer contact times (60 min onwards) can be attributed to a combination of factors: as the adsorption process continues, the MWCNTs’ adsorption sites slowly reach maximum capacity. This results in a reduction in the rate of adsorption and eventually stabilizes at a point where the rate of dye adsorption equals the rate of desorption. The dye concentration remained largely unchanged after three hours from the start of the adsorption process, indicating that equilibrium was reached after three hours. This stability is attributed to the saturation of active sites, preventing further adsorption⁴². Equilibrium is the state at which the rate of adsorption equals the rate of desorption, resulting in no net change in dye concentration. While the study demonstrates that equilibrium is achieved within 30 min, extending the observation period helps confirm that equilibrium is indeed stable. This ensures that the system has genuinely reached equilibrium and accounts for potential minor fluctuations or environmental variations⁴¹.

Effect of pH

The pH level of the solution is important in the sorption process as it impacts the surface charge of the adsorbent and the ionization of the dye molecules. This research explores how pH impacts the elimination of AR 73 dye by utilizing ammonia-doped MWCNTs as adsorbents during exposure to direct sunlight. The data in Fig. 8b shows a notable impact of pH on the effectiveness of Acid Red 73 removal. The greatest elimination rate of 99.16% occurred at a pH of 5, but declined noticeably at pH levels both higher and lower. The ammonia-doped magnetic MWCNTs have the highest removal efficiency at pH 5 due to their surface charge, which is pH-dependent. At a pH of 5, it is probable that the MWCNTs’ surface is positively charged because of the amine groups (-NH₂) being protonated by ammonia doping. This positive charge helps attract AR73 dye molecules, which are negatively charged, through electrostatic attraction, leading to their adsorption. Additionally, acid red 73 is a type of anionic dye, and its pH level affects its ionization state. At a pH of 5, the dye molecules mostly exist as negatively charged anions, which strengthens the electrostatic attraction to the positively charged MWCNTs, resulting in higher adsorption⁴³.

The decrease in removal efficiency at pH values lower than 5 and higher than 5 can be attributed to at pH values lower than 5, the MWCNTs tend to lose their positive charge due to the increased concentration of H⁺ ions in the solution⁴⁴. This reduces electrostatic attraction with dye molecules, leading to a decrease in adsorption. Additionally, at lower pH, the dye molecules might be protonated, reducing their negative charge and weakening electrostatic interactions with the MWCNTs. MWCNTs become increasingly negatively charged at pH levels above 5 as the amount of OH- ions in the solution rises. The negative charges on the MWCNTs and dye molecules cause repulsion, which reduces adsorption and leads to decreased removal efficiency⁴⁵. Furthermore, at higher pH, the dye molecules may become deprotonated, further reducing their negative charge and weakening electrostatic interactions with the MWCNTs. At pH 5, the main factor is the electrostatic pull between the positively charged MWCNTs and the negatively charged dye molecules. This feature boosts sorption and results in increased removal rates. Hydrogen bonding could also play a role in the sorption mechanism, particularly at certain pH levels, allowing dye molecules to create hydrogen bonds with the external functional groups of the MWCNTs⁴⁶. Also, sunlight irradiation can trigger photocatalytic reactions on the MWCNTs, leading to dye degradation and contributing to the overall removal efficiency, particularly at the optimal pH⁴⁷. Ammonia-doped MWCNTs are functionalized with amine groups (-NH₂) that can protonate in acidic environments, forming ammonium ions (-NH₃⁺). This imparts a positive charge to the surface⁴⁸.

The reaction is:

At extremely low pH (high H⁺ ion concentration), the equilibrium of the above reaction can shift in a manner that saturates the protonation sites. The surface becomes crowded with H⁺ ions, potentially reducing the overall positive charge due to the disruption of the electrostatic potential balance or surface charge redistribution. Excess H⁺ ions can lead to competition with the AR73, an anionic dye for adsorption sites. This decreases the effective surface charge available for electrostatic interactions with dye molecules. Additionally, excessive H⁺ ions might lead to proton shielding, reducing the effective positive charge perceived by the dye molecules⁴⁹.

Effect of adsorbent dosage

It is essential to optimize the amount of adsorbent used to achieve the highest level of dye removal efficiency with minimal usage of the adsorbent. The findings demonstrate a notable boost in AR73 removal effectiveness as the adsorbent dosage increases. Figure 7c shows that with dosage of 0.02 g/L, the elimination efficiency is close to 100% and it stays high even with higher doses. Nevertheless, the adsorption ability (Q_e) reduces as the dosage increases, indicating that a higher dosage may not always be the most effective. A higher dosage of the adsorbent leads to greater efficiency in dye elimination owing to the increased surface area and active sites for dye adsorption³⁷.

Fig. 7.

(a) The influence of contact time; (b) The influence of effect of at different pH values; (c) The influence of adsorbent dosage; (d) The effect of initial AR73 concentration (25–200 ppm) on the removal efficiency on the elimination of AR73 from solutions using MWCNTs/NH₃.

As the quantity increases, the quantity of active sites for adsorbing the mix of dyes also increases, resulting in a higher removal efficiency⁵⁰. This allows the MWCNTs to bind a larger amount of dye molecules. Despite the increased removal efficiency, the adsorption capacity (Q_e) decreases with higher adsorbent dosages. This shows that as more dye is eliminated, the amount of dye adsorbed per mass of the adsorbent reduces. This happens because as the amount of substance administered increases, the number of binding sites on the MWCNTs that can be filled up becomes saturated at a faster rate. This results in a reduction in the adsorption capability per mass unit of the adsorbent⁵¹. Moreover, at increased levels of concentration, there could be a conflict for attachment spots between the dye particles and other materials in the mixture, like ammonia. This competition has the potential to decrease the capacity for effective adsorption. The reduction in sorption capacity as adsorbent concentration increases could be due to the fact that some adsorption sites were not fully occupied during the process⁷.

Effect of Initial Acid Red 73 Concentration

Figure 8d demonstrates how the sorption method is significantly impacted by the starting concentration of the dye in the solution. The findings indicate that as the initial concentration of AR73 rises, its removal efficiency tends to decrease. Even though the removal rate stays consistently high (over 95%) at lower levels, it slightly decreases at higher levels. Likewise, the data indicates a pattern where the removal efficiency decreases as the initial AR73 concentration rises from 25 to 200 mg L⁻¹, while the sorption capacity (Q_e) goes up. This type of behavior is common in adsorption processes and can be understood as, at lower concentrations, the adsorbent MWCNTs have more active sites for AR73 removal. The large number of active sites is due to the high surface area and many functional groups, such as amine groups (–NH₂) added through ammonia doping.

A shallower initial concentration results in a more intense concentration gradient, which leads to a more powerful force driving the diffusion of AR73 molecules from the bulk solution to the MWCNT surface. This improves the adsorption rate and helps achieve a high level of removal effectiveness. Additionally, exposure to sunlight has the potential to improve the effectiveness of cleaning at lower levels of concentration by utilizing photocatalysis. The creation of electron-hole pairs in the MWCNTs when exposed to sunlight could aid in breaking down AR73 molecules, enhancing the purification process⁵². UV light activates hydroxyl radicals on the catalyst surface, which are important for removing dye⁵³. While the decreasing removal efficiency at higher concentrations mass transfer resistance at higher concentrations, the diffusion of AR73 particles from the bulk mixture to the MWCNT surface becomes more challenging owing to the decreased concentration gradient. This mass transfer resistance slows down the adsorption process and contributes to a lower removal percentage.

Additionally, the attendance of other materials in the solution, for example ammonia or other dissolved species, may compete with AR73 for available adsorption sites, potentially affecting the adsorption efficiency. This competition intensifies with increased concentration, impeding the elimination of AR73. At elevated levels, the dye molecules could impede the sunlight from reaching the MWCNTs, possibly diminishing the photo adsorptive removal impact and causing a drop in removal efficiency⁵³.

Reduced dye elimination efficiency can be the outcome of high dye molecule concentrations inhibiting the production of hydroxyl radicals by absorbing light photons⁵⁴. Grzechulska and Morawski⁵⁵ investigated the use of photocatalysis to remove azo acid black Higher dye concentrations, according to their findings, result in the development of several layers of dye particles on the catalyst external, obstructing other molecules’ ability to come into direct contact with the hydroxyl radicals and thereby decreasing process efficiency. Furthermore, more dye molecules prevent light photons from penetrating the catalyst surface and generating hydroxyl radicals by absorbing them. Unlike conventional adsorption studies that rely on agitation to enhance mass transfer, the present work demonstrates that ammonia-doped MWCNTs can achieve rapid and efficient dye removal under static conditions, reinforcing their potential for passive and solar-driven wastewater treatment systems.

The initial dark equilibration stage indicates that a portion of dye removal is due to surface adsorption. However, the significant additional decrease in concentration under illumination confirms the dominant role of photocatalytic degradation. Such behavior is consistent with reported photocatalytic systems where adsorption precedes photodegradation.

Chemical Interactions between AR73and Ammonia-Doped Carbon Nanotubes

The interplay between AR73 and ammonia-treated carbon nanotubes (CNTs) involves both physical and chemical forces (Fig. 8). Ionic strength and competing anions can indeed influence adsorption performance through electrostatic screening and competition for active sites.

However, systematic interference experiments with chloride, sulfate, or bicarbonate ions and testing with real textile effluent were beyond the scope of the present study, which focused on fundamental adsorption behavior in controlled aqueous systems.

1. Electrostatic interactions.

AR73 is an anionic dye because of the attendance of sulfonic acid groups (-SO₃H) that dissociate in water, releasing negatively charged sulfonate ions (-SO₃⁻)⁵⁶. Ammonia doping introduces amine groups (-NH₂) onto the CNT surface. These amine groups can protonate in acidic solutions, forming positively charged ammonium ions (-NH₃⁺)⁵⁷. Strong electrostatic attractions are formed by the opposite charges of negatively charged dye ions and positively charged CNT surface, represented by the following equations:

8

9

b) Hydrogen bonding.

The -SO₃H groups in AR73 are capable of creating hydrogen bonds with the -NH₂ groups on the CNT surface. The hydrogen bonding further strengthens the interaction between the dye and the CNTs (Eq. 10):

10

c) Van der Waals forces.

Weak Van der Waals interactions could play a role in the adsorption of AR73 on the CNTs⁵⁸. These forces result from short-term changes in the distribution of electrons, leading to temporary dipoles that draw them towards each other.

d) Potential for photocatalysis.

Sunlight irradiation can excite electrons in the CNTs, generating electron-hole pairs. These excited electrons can react with the AR73 molecules, potentially leading to dye degradation through redox reactions (Eqs. 11–14):

11

12

13

14

In general, the adsorption of AR73 onto ammonia-doped CNTs is a complex process involving a combination of electrostatic interactions, hydrogen bonding, and potentially sunlight-enhanced adsorption. The relative contribution of each mechanism might depend on the specific conditions, for example initial concentration, pH, and sunlight intensity. The enhanced removal efficiency observed under direct sunlight exposure is attributed to a sunlight-assisted adsorption process rather than complete sunlight-assisted adsorption. Sunlight irradiation may induce partial surface activation of ammonia-functionalized MWCNTs, promoting electron excitation and increasing the availability of active adsorption sites. This effect can enhance electrostatic interactions, π–π stacking, and hydrogen bonding between Acid Red 73 molecules and the adsorbent surface. Since dark-control and degradation-product analyses were not performed, the contribution of true sunlight-assisted adsorption cannot be quantitatively confirmed and is therefore discussed qualitatively as a sunlight-enhanced photo adsorptive process. The adsorption mechanism of dye removal by the composite is likely governed by multiple concurrent interactions, including electrostatic attraction between charged surface sites and dye molecules, hydrogen bonding involving surface functional groups, and π–π interactions between aromatic domains of the adsorbent and dye structure. In addition, enhanced removal under illumination suggests a possible light-assisted contribution. These interaction pathways are proposed based on the material’s surface chemistry, aromatic framework, and pH-dependent adsorption behavior observed experimentally. Direct quantitative mechanistic verification was not performed; therefore, the described interactions should be considered plausible adsorption pathways rather than definitive mechanistic assignments. Although reactive species trapping experiments were not performed, the enhanced dye removal observed under illumination suggests a possible light-assisted process rather than confirmed photocatalytic degradation.

Sorption isotherm modeling of acid red 73 Removal by ammonia-functionalized MWCNTs

This research investigates how AR73 dye is adsorbed onto ammonia-doped MWCNTs under sunlight exposure, utilizing three widely-used isotherm models: Freundlich, Langmuir, and Tempkin. These models offer understanding of how adsorption works and assist in forecasting the highest capacity of the adsorbent. Table 3 shows the outcomes of fitting the isotherm models to the experimental data using non-linear methods.

Table 3.

Isotherm parameters of Freundlich, Langmuir, and Tempkin models.

Isotherm Model	Isotherm Parameter	Result
Freundlich	1/n	1.718
	KF (mg1−1/nL1/ng–1)	69.66
	R2	0.885
Langmuir	Qm (mg/g)	312.5
	Ka x 103	310.68
	R2	0.939
Tempkin	AT	4.71082
	BT	57.96
	R2	0.935

Freundlich isotherm

The Freundlich equation posits sorption on an external with different binding sites of different affinities⁵⁹. The model assumes an imperfect adsorption process involving the creation of multiple layers on the external of the adsorbent. The 1/n value (1.718) recommends that the sorption method is favorable and physisorption-dominant, showing a significant interaction between the AR73 molecules and the MWCNTs. The adsorbent’s sorption capacity is represented by the Freundlich constant, which K_F (69.66) this value indicates that the ammonia-doped MWCNTs have a significant adsorption capacity for AR73. The R² value of 0.885 in Fig. 7a suggests that the Freundlich model moderately fits the experimental data, indicating the potential significance of the heterogeneous surface of the sorption process.

Langmuir Isotherm

The Langmuir model suggests that the sorption occurs on a uniform surface with a finite number of the same the sorption sites in a monolayer⁶⁰. The model suggests that once a molecule is attached to a site, no more molecules can attach there. The ammonia-functionalized MWCNTs show a high theoretical adsorption capacity with a Q_m of 312.5 mg g^–1. The K_a Langmuir constant (310.68 × 10^− 3) indicates how strongly the adsorbent binds to the dye molecules. A greater K_a value signifies a more powerful adsorption attraction. The R² value of 0.939 in Fig. 9b designates that the Langmuir model fits the investigational data well, indicating that monolayer adsorption could be the main mechanism.

Fig. 9.

(a) Freundlich isotherm model; (b) Langmuir isotherm model; (c) Tempkin isotherm model and (d) Pseudo-second-order kinetics for removal of Acid Red 73 Removal by ammonia-functionalized MWCNTs from aqueous solution.

Tempkin Isotherm

The Tempkin isotherm model states that the reduction in adsorption heat is directly linked to the rise in coverage due to interactions among adsorbate molecules. It also considers the impact of adsorption heat on the process of adsorption⁶¹. The Tempkin constant, represented by A_T (4.71082), corresponds to the sorption heat. The Tempkin constant, represented as B_T (57.96), correlates with the highest adsorption binding energy. The Tempkin model fits the experimental data well in Fig. 9c with an R² value of 0.935, showing that the sorption process may be influenced by the heat of adsorption. The Tempkin isotherm explicitly considers that the adsorption heat (ΔH) decreases linearly with increased surface coverage due to interactions between adsorbate molecules (AR73 dye) and the adsorbent (ammonia-functionalized MWCNTs). This makes it suitable for systems where significant adsorbate-adsorbate interactions occur, such as when dye molecules are densely packed on the adsorbent surface.

In their study, Abualnaja, et al.⁶² discovered that the Langmuir model was the most appropriate for explaining the elimination of Ismate Violet 2R dye onto MWCNT, showing a Qm value of 76.92 mg g^–1. The ammonia-functionalized MWCNTs used in this study showed a higher adsorption capacity for AR73, indicated by the higher Q_m value of 312.5 mg g^–1 compared to activated carbon. The increased adsorption ability may be owing to the distinct characteristics of the MWCNTs, including their large surface area, the addition of amine groups from ammonia doping, and their possible photo-adsorptive removal behavior when exposed to sunlight. The examination of the sorption isotherms indicates that the sorption of AR73 onto ammonia-functionalized MWCNTs in the presence of sunlight is a complicated procedure. Both the Langmuir and Tempkin models provide accurate representations of the experimental data, indicating the importance of monolayer adsorption and the heat of adsorption in the process. The study showed that ammonia-functionalized MWCNTs have great potential as an effective adsorbent for AR73 removal due to their high adsorption capacity (Qm) and favorable adsorption parameters (1/n, Ka, BT).

The selection of Langmuir, Freundlich, and Temkin isotherms was guided by their ability to describe key aspects of the adsorption process, their simplicity, and their relevance to the system under study. These models effectively capture the adsorption behavior of AR73 dye onto ammonia-functionalized MWCNTs, making them appropriate choices for this research.

The presence of the Temkin isotherm allows to evaluate how adsorbate-adsorbent interactions and energy dynamics contribute to the overall adsorption process. Its good fit to the experimental data (high R²) suggests that these energy interactions play a significant role in the adsorption of AR73 dye onto the ammonia-functionalized MWCNTs. By comparing all three isotherms, the study identifies the interplay of monolayer adsorption (Langmuir), surface heterogeneity (Freundlich), and energy interactions (Temkin), providing a comprehensive understanding of the adsorption process.

Model fitting quality was evaluated using correlation coefficients (R²) obtained from linear regression. More advanced statistical descriptors such as confidence intervals, adjusted R², RMSE, and residual analysis were not included, as the modeling was intended for comparative interpretation rather than rigorous statistical optimization. Such analyses may be considered in future detailed adsorption modeling studies.

Kinetic modeling of AR73 removal

Kinetic models provide valuable insights into the rate and mechanism of adsorption processes^63,64. By examining the experimental data with various kinetic models, the rate-determining step was identified in order to comprehend the impact of different factors on the adsorption process. This research explores the absorption rate of AR73 onto ammonia-functionalized MWCNTs under sunlight exposure by investigating four commonly employed kinetic models; pseudo first-order (PFO), pseudo-second-order (PSO), intraparticle diffusion, and Elovich. Table 4 summarizes the findings of the kinetic modeling.

Table 4.

Parameter values of kinetic model for the sorption of AR73 dye.

Model	Parameter	Value
First-order kinetic model	qe	2.93
	K1 × 103	0.23
	R2	0.0004
Second-order kinetic model	qe (calculation.)	24.44
	k2 × 103	93.97
	R2	0.999
	qe (exp.)	24.61
Elovich isotherm Model	β	25.31
	LN(αβ)	627.08
	αβ	2.2E + 272
	α	8.7E + 270
	R2	0.0239
Intrapaticle diffusion	Kdif	0.014
	C	24.7
	R2	0.048

Lagergren first-order kinetic model (PFO)

This model suggests that the rate of adsorption is in direct correlation with the quantity of unoccupied spaces on the adsorbent^65,66. The calculated q_e is much lower than the experimental value, suggesting a lack of model accuracy. The values along with the correlation coefficients are displayed in Fig. 1S; Table 4. The small value of the rate constant (K₁) suggests a sluggish adsorption rate. The extremely low R² value (0.0004) suggests a weak relationship between the model and the experimental data.

Pseudo-second-order kinetic model (PSO)

This equation makes the assumption that the rate of sorption is directly proportional to the amount of vacant sites squared^67,68. The determined adsorption capacity at equilibrium (q_e) closely corresponds with the experimental value, suggesting a strong representation of the model. The pseudo-second-order kinetic model was applied in its linear form, and the rate constant k₂ is expressed in g·mg⁻¹·min⁻¹, consistent with the standard PSO kinetic equation, as presented in Fig. 9d; Table 4. The strong correlation with an R² value of 0.9994 indicates a high level of agreement among the model and experimental data, pointing towards chemisorption as the main control of the adsorption process.

Elovich isotherm model

The Elovich model proposes that the adsorption rate decreases in an exponential manner as the coverage increases⁶³. Frequently applied for varied surfaces and adsorbates with powerful interactions. Figure 2S; Table 4 display the parameters accompanied by the correlation coefficients. The initial rate constant for adsorption, denoted as α, is exceptionally high, suggesting a quick rate of adsorption at the beginning. The desorption constant (β) indicates the level of surface coverage and the energy needed for desorption. A greater value indicates a less strong bond between the dye and the adsorbent. The Elovich model does not fit the investigational data well indicated by the low R² value of 0.0239, suggesting that chemisorption may not be the main influence on the sorption process.

Intraparticle diffusion model

This model describes the movement of adsorbate particles from the bulk mixture to the pores of the adsorbent. The speed at which AR73 diffuses into the MWCNTs’ pores is represented by the intraparticle diffusion rate constant (K_dif). The outcomes and correlation values can be seen in Table 4; Fig. 3S. Slower diffusion is indicated by a lower value. The intercept (C) offers insight into the impact of the boundary layer. A greater value indicates a more powerful boundary layer impact. The small R² value (0.048) indicates that the experimental data does not align with the intraparticle diffusion model, suggesting that other factors may also play a role in restricting the adsorption rate.

After comparing R² values and calculated q_e values with experimental values, it was found that the PSO model is the best fit for the experimental data. The strong correlation, as well as the close agreement between calculated and experimental q_e values, suggests that chemi-sorption is probably the main factor in the sorption process, which includes the creation of chemical bonds between AR73 molecules and amine groups on the MWCNT surface. Overall, the analysis of kinetics shows that the AR73 adsorption on ammonia-functionalized MWCNTs under sunlight is mainly governed by chemisorption, evidenced by the good fit of the PSO model. The strong correlation between the high R² value and the similarity of calculated and investigational q_e values indicates that the chemical interaction among dye molecules and MWCNTs’ active sites is probably the rate-limiting step. The equilibrium time determined from the kinetic profiles corresponded to the plateau region where no significant change in adsorption capacity was observed, indicating attainment of adsorption equilibrium under the studied conditions.

Antibacterial activity of ammonia-functionalized MWCNTs

Antibacterial activity was evaluated using the agar diffusion method. Gentamicin was used as a positive control, while the solvent/dispersant served as a negative control. The bacteria listed in Table 5 are commonly found in various water sources, both freshwater and aquaculture⁶⁹. The agar diffusion results provide qualitative screening of antibacterial activity. Quantitative antimicrobial parameters such as minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) were not determined in this study. Therefore, the antibacterial findings should be interpreted as preliminary activity indication rather than comprehensive antimicrobial evaluation. Their presence can indicate potential contamination and pose health risks, for example S. aureus often found on skin and in the nose, it can cause skin infections, food poisoning, and serious conditions. E. coli a common inhabitant of the intestines, its presence in water indicates fecal contamination and can lead to gastrointestinal illnesses. P. aeruginosa found in soil, water, and humans, it can cause lung, skin, and urinary tract infections, especially in immune compromised individuals. V. damsela a marine bacterium, can cause wound infections and severe blood poisoning. Klebsiella pneumoniae is found in the environment and humans, it can cause pneumonia, urinary tract infections, and serious illnesses. V. fluvialis found in freshwater and saltwater, it can cause gastroenteritis, particularly in warm waters. While S. agalactiae uis usuallyfound in the vaginal and rectal areas of healthy individuals, it can lead to infections in babies and individuals with compromised immune systems.

Table 5.

Sensitivity of Bacterial Species to ammonia-functionalized MWCNTs.

Bacterial Species	Sensitivity
Staphylococcus aureus ATcc25923	Low sensitivity
Escherichia coli ATcc8731	Resistant
Pseudomonas aeruginosa ATcc9027	Resistant
Vibrio damsela	Resistant
Klebsiella pneumoniae ATcc13883	Resistant
Enterococcus faecalis ATcc29212	Resistant
Vibrio fluvialis	Low sensitivity
Streptococcus agalactiae	Resistant
Aeromonas hydrophila	Low sensitivity

A. hydrophila a common bacterium in freshwater, it can cause wound infections and gastroenteritis. Table 5 shows the sensitivity of various bacterial species to ammonia-functionalized MWCNTs. The results suggest that: S. aureus and V. fluvialis are “Low sensitivity,” meaning they show a lower level of susceptibility to the MWCNTs compared to the other bacteria. While, E. coli, P. aeruginosa, V. damsela, K. pneumoniae, E. faecalis, and S. agalactiae are resistant, meaning they show minimal or no susceptibility to the MWCNTs, while A. hydrophila is also considered a low sensitivity. The results suggest that ammonia-functionalized MWCNTs show limited antibacterial activity against the tested bacteria. While some bacteria exhibit low sensitivity, the majority demonstrate resistance. These findings raise crucial questions about the effectiveness of this material as a broad-spectrum antimicrobial agent. MWCNTs could potentially interfere with bacterial cell membranes, resulting in the demise of the cells. Nevertheless, the dimensions and structure of the MWCNTs utilized in this research may not be ideal for efficient entry into bacterial cells^70–72. Some studies suggest that MWCNTs can generate Reactive Oxygen Species (ROS), which can damage bacterial cells^73,74. Bacteria can develop resistance mechanisms against antimicrobial agents. The bacteria in this study might have intrinsic resistance mechanisms that make them less susceptible to the MWCNTs. This research demonstrates that MWCNTs can absorb AR73. Although their direct antibacterial effects are minimal, they could aid in removing pollutants that support bacterial growth, potentially assisting in bioremediation efforts. Moreover, MWCNTs have the potential to be a sustainable substitute for traditional antimicrobial agents as they can be produced from easily accessible sources. However, the findings indicate that ammonia-functionalized MWCNTs exhibit restricted effectiveness in inhibiting the growth of various bacterial strains. The agar diffusion results provide qualitative screening of antibacterial activity. Quantitative antimicrobial parameters such as minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) were not determined in this study. Therefore, the antibacterial findings should be interpreted as preliminary activity indication rather than comprehensive antimicrobial evaluation.

Comparison Studies on ammonia-functionalized MWCNTs for Dye Removal

Recent studies have explored the sorption of acid dyes using MWCNT and related materials. MWCNTs-Fe₃C nanocomposites have shown effective removal of Acid Red 88, with adsorption kinetics best represented by PSO model⁷⁵. Similarly, MWCNTs demonstrated high efficiency in removing Acid Red 18, with optimal adsorption occurring at acidic pH and Langmuir isotherm⁷⁶.

Table 6 compares the adsorption performance of CNT-based and magnetic nanocomposite adsorbents reported in the literature for azo dye removal, including adsorption capacity (Qₘ), operating pH, initial dye concentration, adsorbent dose, contact time, recyclability, and magnetization where available. The comparison highlights the competitive adsorption capacity of carbon-nanostructured and magnetic hybrid materials reported in recent studies.

Table 6.

Comparative adsorption performance and operational parameters of CNT‑based and magnetic composite adsorbents for anionic dye removal.

Study / Adsorbent	Qₘ (mg·g⁻¹)	pH	Ci (mg·L⁻¹)	Dose (g·L⁻¹)	Contact time (min)	Recyclability	Magnetization (Mₛ)	Notes	Ref.
AM-MWCNTs@ZVI (aminated MWCNTs/Fe⁰)	~ 450 (Congo Red)	4	—	0.04	~ 60	~ 88% after 6 cycles	Yes (Fe⁰)	CNT/ZVI composite	80
MWCNTs (unmodified)	~ 352 (Congo Red)	~ 11	200	0.05	60	—	—	Bare MWCNTs	81
CS-MWCNTs (chitosan-modified CNTs)	> MWCNTs	4	—	—	—	~ 71% after 3 cycles	—	CNT/polymer composite	82
Anthracite/CNT composite	~ 416.7 (Methyl Orange)	—	40–140	—	—	—	—	CNT-impregnated coal	83
Fe₃O₄-CS/Bentonite	~ 169 (Congo Red)	~ 5	—	0.009	~ 25	> 70% after 8 cycles	Yes (Fe₃O₄)	Magnetic composite	84
Zr-CS/Fe₃O₄-NPs@AC	~ 170–609 (temp-dependent)	—	—	—	—	—	Yes (Fe₃O₄)	Magnetic AC/chitosan	85
PSA/GO/CoFe₂O₄ nanocomposite	~ 153.9 (Congo Red)	~ 5	—	—	—	~ 6 cycles	Yes (CoFe₂O₄)	Polymer/graphene/ferrite	86
Bentonite/CNT/PVDF-TRFE	Higher than non-CNT	—	—	—	—	—	—	Polymer/CNT/clay	87
ZnO-Al₂O₃-rGO/MWCNT	~ 220 (Sudan III)	—	—	—	—	—	—	Modified CNT composite	88

Also, comparative research showed that MWCNTs exhibit a greater ability to adsorb reactive dyes in comparison to acid dyes, with electrostatic attraction being the main driving force for adsorption⁷⁷. Recent research has also investigated magnetic coordination polymers, such as NiFe₂O₄/[Zn (BTC)2(Hbtc)(L-His)2(H₂O)₄], for dye removal and antibacterial activity. These materials showed high adsorption capacities for both acid and basic dyes, following different isotherm models, and demonstrated antibacterial properties against S. aureus and E. coli⁷⁸. The composite of MWCNT-Fe³C with magnetic properties was created through chemical vapor deposition and utilized as an adsorbent for the elimination of Acid Red 88 (AR88) acid dye from water. Various factors were examined, including pH ranging from 3.1 to 11.3, temperature between 20 and 60 °C, and initial AR88 concentration of 10 to 58 mg L^− 1. The Freundlich model provided a quite satisfactory fit for the equilibrium data. The PSO model, intraparticle diffusion model, and Lagergren PFO kinetic model were applied to analyze the kinetic sorption data⁷⁹.

Conclusions

The adsorption of AR73, a common azo dye, onto ammonia-functionalized MWCNTs under direct sunlight exposure was investigated in this study. Notably, the adsorbent achieved above 97% removal efficiency within the first 30 min of contact time, highlighting its rapid adsorption capabilities. The most effective pH for removing dye was determined to be 5, highlighting the important role of electrostatic interactions between positively charged MWCNTs and negatively charged dye molecules. Moreover, higher amounts of adsorbent lead to a greater efficiency in removing dye, nearly reaching 100% at a dosage of 0.02 g/L. The adsorption mechanism aligns with the PSO kinetic model, indicating that chemisorption plays a significant role through bond formation between dye molecules and active sites on the MWCNTs. Ammonia-functionalized MWCNTs exhibit limited antibacterial activity against a range of common bacteria found in water. While some show low sensitivity (Staphylococcus aureus and Vibrio fluvialis), most are resistant. Potential ways in which it works involve interfering with bacterial cell membranes and creating reactive oxygen species (ROS). The Langmuir isotherm fits well, providing additional evidence for monolayer adsorption on a uniform surface. Regeneration studies (e.g., desorption using NaOH or ethanol and multiple adsorption–desorption cycles) were not performed in the present work, as the study primarily aimed to evaluate initial adsorption performance and material functionality. The high adsorption capacity of ammonia-functionalized MWCNTs for AR73 removal, with its rapid kinetics and efficiency at optimal conditions, underscores its potential for real-world applications in wastewater treatment. The use of direct sunlight eliminates the need for external energy sources, making the process more sustainable and environmentally friendly. This work demonstrates that ammonia-functionalized MWCNTs are not merely conventional CNT adsorbents, but function as a multifunctional photoadsorptive system capable of efficient azo dye removal under energy-free and realistic operating conditions. Future studies should include dark-control experiments and degradation pathway analysis (e.g., TOC or intermediate identification) to quantitatively distinguish between adsorption and photodegradation contributions.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Author contributions

Mohamed Ashour: Conceptualization, investigation, Software, Formal analysis, Funding acquisition, Visualization, Writing—review and edit. Ahmed E. Alprol: Conceptualization, Methodology, investigation, Software, Formal analysis, Writing—review and editing, Writing—original draft preparation, Project administration. Abdallah Tageldein Mansour: Review and editing, Software, Investigation. Roshmon Thomas Mathew: Review and editing, Visualization, Investigation. Hesham A. Hassanien: Review and editing, Visualization, Investigation. Sameerah I. Al-Saeedi: Review and editing, Visualization, Investigation. Maymounah N. Alharthi: Review and editing, Visualization, Investigation. Ehab El-Haroun: Review and editing, Visualization, Investigation.

Funding

This work was supported by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia [Grant No. KFU260275].

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.