Recent progress in nanomaterial-based adsorption technologies for dye removal from wastewater

Abstract

Dye-contaminated wastewater from textile, printing, leather, paper, cosmetic, and related industries poses serious environmental and health risks due to the high stability, low biodegradability, and potential toxicity of synthetic dyes. Adsorption has emerged as a practical and efficient treatment method because of its simplicity, flexibility, high removal efficiency, and suitability for different dye classes under mild conditions. This review highlights recent advances in dye removal using advanced adsorbents, particularly nanoparticle-based materials, carbon-based adsorbents, polymers, biopolymers, metal–organic frameworks, and zeolitic imidazolate frameworks. Adsorption efficiency is mainly influenced by surface area, pore structure, surface charge, functional groups, adsorbent dosage, initial dye concentration, contact time, temperature, and solution pH. The main removal mechanisms include electrostatic attraction, hydrogen bonding, π–π interactions, ion exchange, surface complexation, pore filling, and coordination interactions. Most reported systems follow pseudo-second-order kinetics, while Langmuir and Freundlich isotherms are commonly used to describe equilibrium behavior. Despite promising laboratory results, practical application remains limited by high production cost, difficult separation, poor regeneration, reduced efficiency in real wastewater, and possible secondary pollution. Future studies should focus on low-cost, reusable, stable, eco-friendly, and scalable adsorbents, with greater attention to real effluents, toxicity assessment, regeneration, economic analysis, and pilot-scale validation.

Dye-contaminated wastewater from textile, printing, leather, paper, cosmetic, and related industries poses significant environmental and public health risks due to the high stability, low biodegradability, and potential toxicity of synthetic dyes.

1. Introduction

The rapid growth of textile, leather, paper, printing, cosmetic, pharmaceutical, and food industries has increased the discharge of dye-contaminated wastewater into the environment.^1–3 Synthetic dyes are widely used because of their strong color, chemical stability, and resistance to fading; however, these properties also make them persistent and difficult to remove from wastewater.^4–6 Even at low concentrations, dyes can cause visible coloration, reduce light penetration, inhibit photosynthesis, decrease dissolved oxygen levels, and negatively affect aquatic organisms. In addition, many dyes and their degradation products may be toxic, mutagenic, carcinogenic, or resistant to biodegradation, making their removal before discharge an important environmental and health requirement.^4,7

Different treatment methods have been applied for dye removal, including coagulation–flocculation, membrane filtration, biological treatment, chemical oxidation, photocatalysis, and electrochemical processes.^8,9 Although these methods can be effective, they often suffer from limitations such as high cost, sludge production, membrane fouling, high energy demand, incomplete degradation, or formation of secondary pollutants.¹⁰ Therefore, adsorption has attracted significant attention as a simple, efficient, flexible, and cost-effective method for removing dyes from aqueous solutions.¹¹ It can be applied under mild operating conditions and used either as a main treatment process or as a polishing step for industrial effluents.

The adsorption efficiency depends on several factors, including solution pH, initial dye concentration, adsorbent dose, contact time, temperature, ionic strength, and the presence of competing pollutants.^12,13 The properties of the adsorbent, such as surface area, pore structure, surface charge, morphology, and functional groups, also play an important role in dye uptake. Dye adsorption usually occurs through electrostatic attraction, hydrogen bonding, π–π interactions, ion exchange, surface complexation, coordination interaction, and pore filling. Kinetic studies are commonly described using pseudo-first order and pseudo-second-order models, while Langmuir and the Freundlich isotherms are widely used to explain monolayer or heterogeneous adsorption behavior.^14–16

Recently, nanomaterial-based adsorbents have received increasing attention due to their high specific surface area, abundant active sites, tunable surface chemistry, and enhanced adsorption capacity.^17,18 Various materials have been investigated for dye adsorption, including metal oxide nanoparticles, metallic nanoparticles, spinel oxide nanoparticles, metal sulfide nanoparticles, carbon-based materials such as CNTs, GO, rGO, activated carbon, and biochar, as well as polymeric and biopolymer composites such as chitosan-based adsorbents.^17–19 In addition, MOF- and ZIF-based materials have shown promising performance because of their highly porous structures, adjustable pore sizes, and tunable functional groups.^20,21

Despite these advantages, several challenges still limit the practical application of adsorption technology. Many adsorbents show high efficiency in synthetic dye solutions but lower performance in real industrial wastewater due to salts, organic matter, surfactants, heavy metals, and mixed dyes.^22,23 Other limitations include high preparation costs, difficulty in separating nanosized adsorbents, poor regeneration, loss of capacity after reuse, and possible secondary pollution from spent adsorbents.^24,25 Therefore, future research should focus on developing low-cost, reusable, stable, environmentally safe, and scalable adsorbents, with greater attention to real wastewater treatment, regeneration, toxicity assessment, cost analysis, and pilot-scale applications.²⁶

This review discusses recent progress in dye adsorption using different classes of advanced adsorbents, including metal oxides, metallic nanoparticles, spinel oxides, metal sulfides, carbon-based materials, polymeric/biopolymer composites, MOFs, and ZIFs. It also highlights adsorption mechanisms, influencing factors, kinetic and isotherm behavior, limitations, and future perspectives for sustainable treatment of dye-contaminated wastewater.

2. Classification of dyes

Dyes are classified based on their chemical structure, charge, and application method. These classifications are essential for understanding the challenges in treating dye-laden wastewater, particularly when using nanoparticles for dye removal. This section covers various dye categories, their characteristics, and the types of nanoparticles used for their removal, drawing from recent literature to compare results and highlight the effectiveness of different treatment strategies (see Fig. 1).

Fig. 1. Classification of dyes.

Dyes can be classified according to their chemical structure, ionic charge, water solubility, and origin. Structurally, major classes include azo, anthraquinone, triphenylmethane, phthalocyanine, indigoid, and xanthene dyes. Based on ionic character, dyes are categorized as cationic, anionic, non-ionic, or zwitterionic, which strongly influences their solubility and adsorption behavior. According to water solubility, dyes may be water-soluble or insoluble, determining their application and treatment requirements. Finally, dyes can be classified by origin as natural, derived from plants, animals, or minerals, or synthetic, which are widely used because of their high color intensity, stability, and versatility.

3. Removal mechanisms of dyes

The removal of synthetic dyes from wastewater using nanoparticle-based materials occurs through several complementary mechanisms, including adsorption, electrostatic interaction, photocatalytic degradation, and Fenton-like oxidation. These mechanisms are strongly influenced by the physicochemical properties of both the dye molecules and the nanomaterials, such as surface charge, particle size, surface area, porosity, functional groups, crystallinity, and catalytic activity. Since dye-contaminated wastewater usually contains different classes of dyes with complex molecular structures, no single mechanism is sufficient in all cases. Therefore, efficient dye removal often results from the combined action of surface adsorption and chemical degradation processes. The main mechanism involved in dye removal is described in Fig. 2. Comparison between different approaches used for dye removal is listed in Table 1.

Fig. 2. The main mechanisms involved in dye removal.

Table 1. Comparison between different approaches used for dye removals.

Removal mechanism	Literature example	Target dye(s)	Material/system used	Main reported role/mechanism	Ref.
Adsorption	Chitosan–graphene oxide nanocomposite membrane	Methylene blue and Congo red	Chitosan–GO/electrospun nanofiber membrane	Surface adsorption through functional groups of chitosan and GO; suitable for simultaneous cationic/anionic dye removal	27
	Chitosan-modified hybrid nanocomposite	Congo red and methylene blue	Chitosan-modified hybrid adsorbent	Adsorption controlled by pH, dye charge, and available active sites	28
	Magnetic Fe3O4/polypyrrole/carbon black nanocomposite	Methylene blue and Congo red	Fe3O4/PPy/C magnetic nanocomposite	High uptake of basic and acidic dyes; magnetic component supports recovery	29
	Reduced graphene oxide adsorbent	Methylene blue	Eco-friendly reduced graphene oxide	MB adsorption attributed to large surface area, π–π interactions, and oxygen-containing groups	30
	Cobalt ferrite–chitosan composite	Congo red and methyl orange	CoFe2O4 and CoFe2O4–chitosan composite	Adsorption of anionic dyes from single and binary solutions; chitosan improves surface functionality	31
Electrostatic interaction	Delaminated Ti3C2 MXene	Six organic dyes	Ti3C2 MXene nanosheets	Study directly reported that electrostatic interactions between ionizable MXene groups and dye molecules strongly controlled removal efficiency	32
	MXene/PEI sodium alginate aerogel	Congo red	MXene/polyethyleneimine-functionalized sodium alginate aerogel	Very high Congo red uptake was attributed to strong electrostatic attraction plus surface/intercalation adsorption	33
	GO/PAMAM nanocomposite	Congo red	Graphene oxide/poly(amidoamine) dendrimer nanocomposite	Removal of anionic Congo red was enhanced by amine-rich surface groups and charge-based attraction	34
	Nanoceria/H2O2 system	Methylene blue	CeO2 nanoparticles with H2O2	The study reported different adsorption–degradation behavior, including electrostatic attraction for methylene blue and stronger Lewis acid–base interaction for Congo red	35
Photocatalytic degradation	TiO2 thin films	Methyl orange and methylene blue	Anatase TiO2 thin films prepared by MOCVD	UV/sunlight photocatalysis degraded dyes; activity depended on deposition temperature and surface properties	36
	ZnO nanoparticles	Methylene blue	ZnO nanoparticle thin films	MB degradation under UV/laser irradiation was used to evaluate photocatalytic activity	37
	ZnO/TiO2 composite	Methyl orange	ZnO/TiO2 photocatalyst	Composite showed enhanced degradation of methyl orange due to synergistic ZnO–TiO2 effects	38
	CuO/ZnO nanocomposite	Methylene blue	CuO nanoplates/ZnO nanoparticles	ZnO–CuO heterostructure improved MB photodegradation by enhancing charge separation	39
	NiO/Ag/TiO2 nanocomposite	Methylene blue	NiO/Ag/TiO2 photocatalyst	Visible-light photocatalytic removal of MB; improved reaction rate compared with less modified TiO2 systems	40
Fenton-like oxidation	Fe3O4/ZnO/graphene composite	Methylene blue and Congo red	Fe3O4/ZnO/graphene	Used as heterogeneous Fenton, photo-Fenton, sono-Fenton, and sonophoto-Fenton catalyst for cationic and anionic dyes	41
	Nanoceria/H2O2 system	Congo red	CeO2 nanoparticles with H2O2	Fenton-like adsorption–degradation system; surface adsorption influenced subsequent oxidative degradation	35
	Fe3O4/montmorillonite composite	Methylene blue	Fe3O4/MMT with H2O2	Heterogeneous Fenton oxidation achieved near-complete MB color removal after optimization of H2O2 and temperature	42
	α-Fe2O3/SiO2 composite	Dye pollutants	Mesoporous α-Fe2O3/SiO2	Reported as a heterogeneous Fenton-like catalyst with both adsorption capacity and catalytic degradation ability	43
	Fe3O4@rGO@TiO2 system	Methylene blue	Magnetic Fe3O4@rGO@TiO2	Rapid photo-Fenton degradation of MB; magnetic core also supports catalyst recovery	44

3.1. Adsorption

Adsorption is one of the most common mechanisms involved in dye removal. In this process, dye molecules accumulate on the surface of an adsorbent through physical or chemical interactions. Nanoparticle-based materials are highly effective adsorbents because they possess a large surface area, abundant active sites, and tunable surface chemistry. Metal oxide nanoparticles, carbon-based materials, biopolymers, and hybrid porous materials can all adsorb dye molecules from aqueous solutions.

Physical adsorption usually occurs through weak forces such as van der Waals interactions, hydrogen bonding, and π–π interactions. For example, carbon-based materials such as graphene oxide, graphene, and carbon nanotubes can adsorb aromatic dye molecules through π–π stacking between the aromatic rings of the dye and the carbon surface. This mechanism is particularly important for dyes such as methylene blue, Congo red, and methyl orange, which contain aromatic structures. In contrast, chemical adsorption involves stronger interactions, such as coordination bonding or surface complexation, between functional groups on the dye molecules and active sites on the nanomaterial surface.

The adsorption efficiency depends on several factors, including solution pH, initial dye concentration, contact time, adsorbent dosage, temperature, and ionic strength. The pH of the solution is especially important because it affects both the ionization state of the dye and the surface charge of the adsorbent. For instance, cationic dyes such as methylene blue are more readily adsorbed onto negatively charged surfaces, while anionic dyes such as Congo red and methyl orange are more strongly attracted to positively charged surfaces.

3.2. Electrostatic interactions

Electrostatic interaction is another major mechanism in dye removal, particularly for charged dyes. Many synthetic dyes exist as ionic species in water. Cationic dyes carry a positive charge, while anionic dyes carry a negative charge. The surface charge of the nanomaterial determines the strength of attraction or repulsion between the dye and the adsorbent.

For example, negatively charged nanomaterials can attract cationic dyes such as methylene blue, crystal violet, and rhodamine B. On the other hand, positively charged materials can interact strongly with anionic dyes such as Congo red, methyl orange, and Reactive Red 120. This interaction can significantly enhance dye adsorption and improve removal efficiency. Surface modification is often used to improve electrostatic attraction. For example, coating nanoparticles with functional groups such as amine, hydroxyl, or carboxyl can alter their surface charge and enhance their affinity for specific dye molecules.

The pH plays a crucial role in electrostatic dye removal. At low pH, many adsorbent surfaces become protonated and positively charged, which favors the adsorption of anionic dyes. At high pH, the surface may become negatively charged, improving the removal of cationic dyes. Therefore, controlling solution pH can optimize the interaction between the dye and the nanomaterial surface.

3.3. Photocatalytic degradation

Photocatalytic degradation is an advanced removal mechanism in which dye molecules are degraded into smaller and less harmful compounds under light irradiation. Semiconductor nanoparticles, such as titanium dioxide and zinc oxide, are widely used photocatalysts because they generate reactive oxygen species upon exposure to ultraviolet or visible light.

During photocatalysis, light energy excites electrons from the valence band to the conduction band of the semiconductor, creating electron–hole pairs. These electrons and holes react with water and dissolved oxygen to produce highly reactive species such as hydroxyl radicals and superoxide radicals. These reactive species attack the chromophoric groups and aromatic rings of dye molecules, leading to decolorization and degradation.

For example, dyes such as methylene blue, methyl orange, rhodamine B, and Congo red can be degraded through photocatalytic oxidation using TiO₂, ZnO, or modified semiconductor nanomaterials. The degradation process usually begins with the breakdown of the chromophore responsible for color, followed by the further oxidation of the resulting intermediate products. In ideal conditions, complete mineralization can convert dye molecules into carbon dioxide, water, and inorganic ions.

However, photocatalytic efficiency can be limited by rapid recombination of electron–hole pairs and poor visible-light absorption. To overcome these limitations, photocatalysts are often modified by doping, coupling with carbon-based materials, forming heterojunctions, or combining with magnetic nanoparticles for easier recovery.

3.4. Fenton-like oxidation

Fenton-like oxidation is another important mechanism for dye degradation. This process involves the generation of hydroxyl radicals from hydrogen peroxide in the presence of metal-based catalysts. Iron-containing nanoparticles, especially magnetite nanoparticles, are commonly used in Fenton-like systems because they can activate hydrogen peroxide and produce highly reactive radicals.

In this mechanism, iron species on the nanoparticle surface react with hydrogen peroxide to generate hydroxyl radicals. These radicals are powerful oxidizing agents that can attack dye molecules and break down their complex structures. For example, azo dyes such as Congo red and methyl orange can be degraded by cleavage of the azo bond, while other dyes, such as methylene blue and rhodamine B, can undergo oxidative destruction of their chromophoric groups.

A major advantage of Fenton-like systems is their strong oxidative ability and effectiveness toward a wide range of dye pollutants. Magnetic nanoparticles also allow easy separation of the catalyst from treated water using an external magnetic field. However, the process may be affected by pH, hydrogen peroxide concentration, catalyst dosage, and the presence of interfering ions in real wastewater. Excess hydrogen peroxide can also reduce efficiency by scavenging hydroxyl radicals, so reaction conditions must be carefully optimized.

3.5. Combined removal pathways

In many nanoparticle-based systems, dye removal does not occur through only one mechanism. Instead, several mechanisms operate simultaneously. For example, a magnetic graphene oxide composite may remove dyes through adsorption, π–π interaction, electrostatic attraction, and magnetic separation. Similarly, TiO₂-based nanocomposites may first adsorb dye molecules onto the catalyst surface and then degrade them photocatalytically under light irradiation. Hybrid porous materials, such as metal–organic frameworks and zeolitic imidazolate frameworks, can combine high adsorption capacity with catalytic activity due to their large pore volumes, tunable structures, and active metal centers.

This combined mechanism is important because adsorption alone transfers dye molecules from water to a solid phase, but it does not necessarily destroy the pollutant. In contrast, degradation mechanisms such as photocatalysis and Fenton-like oxidation can break down dye molecules into smaller compounds. Therefore, coupling adsorption with catalytic degradation offers a more comprehensive and sustainable approach to treating dye-contaminated wastewater.

3.6. Factors affecting dye removal mechanisms

The effectiveness of dye removal depends on several operational and material-related factors. Solution pH controls both the dye's ionization and the adsorbent surface charge. Contact time determines whether the system reaches adsorption equilibrium or complete degradation. Initial dye concentration affects the number of available active sites and the reaction rate. Nanomaterial dosage influences the total surface area and number of catalytic sites. Temperature may affect diffusion and reaction kinetics, while light intensity is especially important in photocatalytic processes.

The dye's structure also plays an important role. Dyes with large molecular size may diffuse more slowly into porous materials, while dyes with multiple sulfonate or amine groups may interact strongly with charged surfaces. Similarly, dyes containing azo, anthraquinone, or triphenylmethane structures may show different degradation pathways depending on their chemical stability and functional groups.

4. Experimental parameters affecting dye removal

The removal efficiency of dye from aqueous solutions is strongly influenced by several experimental parameters, including contact time, initial dye concentration, adsorbent dosage, solution pH, temperature, and agitation speed (see Fig. 3). Understanding the effect of these parameters is essential for optimizing the treatment process and improving the overall removal performance.^45,46

Fig. 3. Schematic illustration of the main experimental parameters affecting dye removal, including contact time, initial dye concentration, adsorbent dosage, solution pH, temperature, and agitation speed. These parameters collectively influence adsorption behavior, dye uptake efficiency, equilibrium attainment, and the overall performance of the dye removal process.

Several operational parameters, including contact time, initial dye concentration, adsorbent dosage, solution pH, temperature, and agitation speed, significantly influence dye adsorption performance. Contact time determines the duration available for dye molecules to interact with the adsorbent. Adsorption is generally rapid at the initial stage because numerous vacant and accessible active sites are available. As adsorption proceeds, these sites become progressively occupied, resulting in a gradual decrease in the adsorption rate until equilibrium is reached, beyond which further increases in contact time produce little improvement.

The initial dye concentration affects both the removal efficiency and adsorption capacity. At low concentrations, the available adsorption sites are generally sufficient to capture most dye molecules, resulting in high removal percentages. Increasing the dye concentration increases the concentration gradient and provides a stronger driving force for mass transfer, which can enhance adsorption capacity. However, when the number of dye molecules exceeds the available active sites, the percentage removal generally decreases. Similarly, increasing the adsorbent dosage usually enhances dye removal by providing a greater surface area and more active sites. Nevertheless, beyond an optimum dosage, further increases may produce limited improvement because of particle aggregation, overlapping of adsorption sites, or reduced effective surface area. Therefore, an appropriate adsorbent dosage is required to achieve efficient removal while minimizing unnecessary material consumption (Fig. 4).

Fig. 4. Types of catalyst for dye adsorption.

The solution pH is another critical parameter because it affects the surface charge of the adsorbent and the ionization state of dye molecules. Adsorption of cationic dyes is often favored under alkaline conditions, where the adsorbent surface becomes more negatively charged and electrostatic attraction is enhanced. In contrast, acidic conditions may favor the adsorption of anionic dyes by increasing the positive charge of the adsorbent surface. However, the optimum pH depends on the chemical characteristics of both the dye and adsorbent.

Temperature can influence adsorption by affecting molecular mobility, diffusion, pore accessibility, and the interactions between dye molecules and the adsorbent surface. Increasing temperature may enhance adsorption when the process is endothermic by facilitating molecular diffusion and penetration into the adsorbent pores. Conversely, adsorption may decrease with increasing temperature for exothermic processes because of weakening of the interactions between the dye and adsorbent. Finally, agitation speed improves mixing and facilitates mass transfer between the dye solution and adsorbent surface. Increasing agitation generally reduces boundary-layer resistance and enhances dye diffusion; however, excessively high speeds may provide little additional benefit and may cause particle fragmentation or desorption. Thus, optimization of these operational parameters is essential for maximizing dye removal efficiency and adsorption capacity.

5. Thermodynamic insights into dye–adsorbent interactions

Thermodynamic parameters provide valuable insight into the driving forces governing dye adsorption at the solid–liquid interface and should not be interpreted solely in terms of endothermic or exothermic behavior. The spontaneity of adsorption is determined by the Gibbs free energy change (ΔG°), which is related to enthalpic and entropic contributions according to:

ΔG° = ΔH° − TΔS° 1

where ΔH° is the standard enthalpy change, ΔS° is the standard entropy change, and T is the absolute temperature. Negative ΔG° values indicate that dye adsorption is thermodynamically favorable, while increasingly negative values at higher temperatures suggest enhanced adsorption affinity. The Gibbs free energy is also related to the equilibrium constant (K) through:

ΔG° = −RTln K 2

where R is the universal gas constant (8.314 J mol⁻¹ K⁻¹). Thus, larger equilibrium constants correspond to more favorable adsorption processes.

The enthalpy change (ΔH°) reflects the net energy associated with adsorbate–adsorbent interactions, including electrostatic attraction, hydrogen bonding, π–π interactions, van der Waals forces, and possible coordination interactions with surface functional groups. Positive ΔH° values indicate that adsorption requires an input of energy, often due to desolvation and surface rearrangement processes, whereas negative ΔH° values imply that the formation of surface interactions releases energy. Importantly, the entropy change (ΔS°) provides information about the degree of organization within the entire adsorption system rather than the adsorbed molecules alone. In aqueous media, dye molecules and adsorbent surfaces are surrounded by structured hydration shells formed through hydrogen bonding and electrostatic interactions. During adsorption, partial displacement and reorganization of these hydration layers may occur, releasing constrained water molecules into the bulk solution where they possess greater translational and rotational freedom. This desolvation process can generate positive entropy changes, making the term (TΔS°) a significant driving force for adsorption. Therefore, positive ΔS° values are consistent with increased disorder resulting from interfacial water release and restructuring at the solid–liquid interface. Overall, dye adsorption is governed by a balance between energetic contributions (ΔH°) and entropic gains (TΔS°), with the resulting ΔG° determining the thermodynamic favorability of the process.

6. Adsorption of dyes using nanoparticles for dye removal applications

Nanoparticles have received increasing attention as efficient adsorbents for dye removal from aqueous solutions due to their nanoscale size, high specific surface area, abundant active sites, and tunable surface properties. Among the different types of nanoadsorbents, metal oxide nanoparticles are widely studied because they possess reactive surface hydroxyl groups, variable surface charge, chemical stability, and strong affinity toward different classes of dye molecules. Comparison between different metal-based surfaces for dye adsorption represented in Table 2. Dye adsorption onto metal oxide nanoparticles can occur through several mechanisms, including electrostatic attraction, hydrogen bonding, surface complexation, ion exchange, and pore-filling, depending on the physicochemical properties of both the adsorbent and the dye. The adsorption efficiency is strongly affected by operational conditions such as solution pH, initial dye concentration, adsorbent dosage, contact time, temperature, and ionic strength. In addition, surface modification and incorporation of metal oxide nanoparticles into composite materials can improve adsorption capacity, dispersion stability, regeneration ability, and separation efficiency. Therefore, metal oxide nanoparticles represent a promising class of nanomaterials for the treatment of dye-contaminated wastewater, although further studies are still required to improve their reusability, environmental safety, and large-scale applicability (see Fig. 5).

Table 2. Comparison table for transition metal-based adsorbent.

Material/catalyst	Dye removed	Reported efficiency (%)	q max (mg g−1)	Main mechanism	Year	Ref.
Flower-like MoS2 nanomaterials	Methylene blue (MB)	>93.47	200	Langmuir monolayer adsorption; pseudo-second-order kinetics; electrostatic attraction	2022	101
α-MoO3 nanoparticles	Methylene blue (MB)	100	152	pH-assisted adsorption; pseudo-second-order behavior	2018	102
Magnetic NiFe2O4 nanoparticles	Remazol Red	96.5 ± 0.4; up to 99 at optimum pH	169.5 ± 0.8	Electrostatic attraction; favorable Langmuir adsorption; pH-dependent surface charge	2025	103
MgFe2O4 nanoparticles	Methylene blue (MB)	≈99 reported high removal	537.63	Electrostatic attraction and H-bonding between MB and MgFe2O4 surface hydroxyls	2023	79
Fe/Al di-metal nanostructured composite	Congo red (CR)	≥99	411	Diffusion-limited adsorption; high surface area and pores; heterogeneous active sites	2022	104
Iron-impregnated nanoclay	Methylene blue (MB)	99.9	79.68	Physisorption with complexation/precipitation contribution; possible chromophore disruption	2021	72
Fe3O4@C nanoparticles	Methylene blue (MB)	≈99 with H2O2	18.52	Adsorption plus Fenton-like oxidation; carbon shell adsorption and Fe3O4 activation of H2O2	2021	49
Teff-straw-assisted ZnO nanoparticles	Methylene blue (MB)	97.8; 92.5 in textile wastewater	62.5	Surface adsorption on ZnO; pH-dependent electrostatic interaction	2025	56
Biogenic CuO nanoparticles	Congo red (CR)	78% retained after reuse; lower CR removal than MG	12.06	Langmuir adsorption; pseudo-second-order kinetics; electrostatic interaction	2024	105
Malva sylvestris-derived CuO nanoparticles	Congo red (CR)	≈80 rapid initial removal	6.99	Surface adsorption on CuO active sites; Langmuir monolayer adsorption	2025	59
Biogenic TiO2 nanoparticles	Congo red (CR)	90	N/A	Photocatalytic/adsorptive removal; surface active sites under optimized dose	2025	106
nZVI nanoparticles	Acid Black 1	96.3	N/A	Reductive degradation/adsorption on zero-valent iron; heterogeneous multilayer adsorption	2025	107
P/Al2O3 biogenic nanoparticles	Congo red (CR)	N/R	27.78	Langmuir adsorption; surface active sites	2025	61
Cu,N-doped TiO2/PANI nanocomposite	Methyl orange (MO)	High removal reported	N/A	Photocatalytic degradation and adsorption synergy; doped TiO2/PANI charge separation	2021	108
TiO2 nanoparticles-modified smectite	Rhodamine B (RhB)	High removal reported	N/A	Adsorption on TiO2/smectite surface; kinetic/isotherm-controlled uptake	2021	109
Ag–Cu–CeO2 nanocomposite	Congo red (CR)	96.83	4.71	Freundlich adsorption; pseudo-second-order kinetics; spontaneous exothermic removal	2024	64
ZnO@EPS nanoparticles	Methyl orange (MO)	High photodegradation under optimum conditions	49.1	Photocatalytic degradation; Freundlich-like surface uptake; pseudo-second-order model	2025	57
ZnO–rGO nanocomposite	Rhodamine B (RhB)	99	N/A	Adsorption enhanced by rGO support; π–π interactions and surface charge effects	2024	58
Fe–Sn–Ni ternary metal oxide nanocomposite	Congo red (CR)	High adsorption reported	N/A	Porous metal oxide adsorption; abundant surface sites; likely electrostatic/H-bonding	2025	71
CuO/biochar nanocomposite	Reactive Red 120 (RR120)	99.2	1399	Adsorption on CuO-modified biochar; electrostatic attraction, pore filling, and surface complexation; Langmuir-type uptake	2020	110
Fe3O4/black cumin carbon nanobiosorbent	Methylene blue (MB)	High optimized removal	62.1	Magnetic nanobiosorption; electrostatic interaction and porous carbon-supported Fe3O4 active sites	2025	111
Ocimum basilicum-waste-derived ZnO nanoparticles	Methylene blue (MB)	Rapid high removal	N/A	Green ZnO adsorption; pH-dependent surface charge; pseudo-second-order/Freundlich-type behavior reported	2025	65
Magnetite Fe3O4 nanoparticles	Methylene blue (MB)	95.11 by adsorption; 98.5 by AOP	N/A	Adsorption and advanced oxidation/Fenton-like pathway with H2O2; magnetic separation enables reuse	2024	112
Chitosan-templated NiFe2O4 nanoparticles	Methylene blue (MB)	96.8	N/A	Ferrite nanoparticle adsorption; pH-dependent electrostatic attraction and surface interactions	2025	113
Green-synthesized NiO/ZnO nanocomposite	Methylene blue (MB)	99	66.7	Langmuir adsorption and pseudo-second-order kinetics; electrostatic attraction and surface complexation	2025	66
Green-synthesized NiO/ZnO nanocomposite	Crystal violet (CV)	99	52.9	Langmuir adsorption and pseudo-second-order kinetics; cationic dye interaction with oxide surface sites	2025	66
ZnO@activated carbon/phosphate composite	Celestine blue (CEB)	High optimized removal	172.41	Hybrid nanoparticle/carbon adsorption; electrostatic attraction and surface active sites	2025	114

Fig. 5. Effect of adsorption parameters on methylene blue (MB) removal using Fe₃O₄@C nanoparticles: (a) adsorbent dosage, (b) initial MB concentration, (c) temperature, and (d) initial pH. Reproduced from ref. 49 (Xiang et al., 2021) under a Creative Commons Attribution 4.0 International License (CC BY 4.0).

6.1. Metal oxide nanoparticles

Metal oxide nanoparticles have attracted considerable attention as adsorbents for dye removal because their high specific surface area, abundant surface hydroxyl groups, chemical stability, and pH-responsive surface charge collectively enhance interactions with dye molecules. Their adsorption performance is primarily governed by the availability of active surface sites, electrostatic interactions, and the possibility of coupling adsorption with catalytic degradation. Among these materials, Fe₃O₄-based nanostructures offer an additional advantage because their magnetic properties facilitate rapid recovery and reuse, while their redox-active iron species enable Fenton-like reactions in the presence of H₂O₂. Consequently, dye removal is achieved through the combined effects of adsorption and oxidative degradation, often resulting in substantially higher removal efficiencies than adsorption alone.

The development of sustainable biosorbents has recently been advanced through the integration of agricultural waste with metal oxide nanoparticles. A recent study developed ZnO-impregnated Musa acuminata peel as a sustainable biosorbent for Congo red (CR) removal from aqueous solutions. The ZnO modification enhanced the adsorption performance compared with untreated banana peel, with the process being evaluated under different pH, dosage, contact time, temperature, and initial dye concentration conditions. The adsorption followed pseudo-second-order kinetics, while the Langmuir model provided the best fit for the ZnO-modified material. Thermodynamic analysis revealed negative values of ΔG°, ΔH°, and ΔS°, indicating that CR adsorption was spontaneous and exothermic, with a decrease in randomness at the solid–solution interface. The negative enthalpy value (ΔH° = −20.63 kJ mol⁻¹) further suggested that the adsorption was predominantly physical in nature. These findings demonstrate the potential of combining agricultural waste with ZnO nanoparticles to develop sustainable and efficient adsorbents for dye-contaminated wastewater.⁴⁷

Another recent sustainable approach involved the fabrication of a magnetic bio-hybrid adsorbent by functionalizing avian eggshell membrane with Fe₃O₄ nanoparticles (Fe₃O₄@ESM) for Congo red removal. The eggshell membrane provided abundant amino, carboxyl, and hydroxyl functional groups for dye binding, whereas the Fe₃O₄ nanoparticles enabled convenient magnetic separation and recovery of the adsorbent after treatment. The composite achieved more than 96% CR removal with an adsorption capacity of 45.04 mg g⁻¹ under optimized conditions, and the adsorption behavior followed the Langmuir isotherm and pseudo-second-order kinetic model. Thermodynamic analysis showed negative ΔG° and ΔH° values together with a positive ΔS° value, confirming that the adsorption was spontaneous and exothermic and was accompanied by increased randomness at the adsorbent–solution interface. The calculated values (ΔG° = −16.02 to −16.71 kJ mol⁻¹, ΔH° = −5.68 kJ mol⁻¹, and ΔS° = 34.12 J mol⁻¹ K⁻¹) further support the favorable nature of CR uptake.⁴⁸

The adsorption behavior of Fe₃O₄@C nanoparticles illustrates the strong dependence of dye removal on operating conditions (Fig. 5). Increasing the adsorbent dosage enhances dye removal by providing a larger number of accessible adsorption sites; however, the adsorption capacity per unit mass decreases because many sites remain unsaturated at higher dosages (Fig. 5a). In contrast, increasing the initial dye concentration strengthens the concentration gradient and increases the adsorption capacity, although the overall removal efficiency declines as the finite number of active sites approaches saturation (Fig. 5b). Temperature also plays a decisive role, with reduced adsorption at elevated temperatures indicating that methylene blue adsorption is predominantly an exothermic process (Fig. 5c). Furthermore, alkaline conditions improve both adsorption capacity and removal efficiency because the reduced competition from H⁺ ions increases the availability of negatively charged surface sites for electrostatic attraction of the cationic methylene blue molecules (Fig. 5d).

These observations highlight that adsorption efficiency depends not only on the intrinsic properties of the adsorbent but also on the interplay between surface chemistry and operational parameters. Although Fe₃O₄@C nanoparticles achieved an adsorption capacity of 18.52 mg g⁻¹ and nearly complete methylene blue removal (≈99%) in the presence of H₂O₂, the exceptional performance results from the synergistic coupling of adsorption with Fenton-like oxidation rather than adsorption alone. This distinction is important when comparing Fe₃O₄-based materials with conventional adsorbents, as adsorption capacities should be interpreted alongside catalytic contributions to avoid overestimating the intrinsic adsorption capability of the material.⁴⁹

The adsorption performance of ZnO- and TiO₂-based nanomaterials is strongly influenced by their surface chemistry, morphology, and the availability of active adsorption sites. Rather than relying solely on high surface area, these materials remove dyes through a combination of electrostatic attraction, hydrogen bonding, surface complexation, and, in some cases, photocatalytic oxidation. Consequently, their adsorption efficiency depends not only on the intrinsic properties of the oxide nanoparticles but also on operating parameters such as pH, dye concentration, and contact time.

Surface charge plays a particularly important role in the adsorption of cationic dyes such as methylene blue (MB) and methyl green. Under neutral or alkaline conditions, oxide surfaces become increasingly negatively charged, strengthening electrostatic interactions with positively charged dye molecules while minimizing competition from H⁺ ions. This behavior explains the consistently high removal efficiencies reported for ZnO-based adsorbents under near-neutral or alkaline conditions. For example, ZnO nanoparticles achieved almost complete methyl green removal (99.96%) at pH 7, while biosynthesized ZnO nanoparticles exhibited enhanced MB adsorption with increasing contact time but reduced efficiency at higher dye concentrations because of progressive saturation of adsorption sites.^50,51 Likewise, adsorption studies performed over a broad pH range confirmed that alkaline conditions favor MB uptake through the combined effects of electrostatic attraction and hydrogen bonding.⁵²

Adsorption kinetics reported for ZnO- and TiO₂-based systems consistently indicate that surface interactions, rather than simple physical diffusion, dominate dye uptake. Most studies are better described by pseudo-second-order or general-order kinetic models, suggesting that adsorption proceeds through strong interactions between dye molecules and surface functional groups. Similarly, Freundlich isotherm behavior is frequently observed, implying heterogeneous surfaces with multilayer adsorption rather than uniform monolayer coverage.^50,51,53 Although these kinetic and equilibrium models are widely used, they provide only indirect evidence of adsorption mechanisms, emphasizing the need for complementary spectroscopic analyses to verify the nature of dye–surface interactions.

Engineering the nanoparticle surface through doping or composite formation further enhances adsorption by modifying the electronic structure, increasing the density of active sites, and improving structural stability. For instance, La–Na co-doped TiO₂ nanoparticles achieved approximately 99% MB removal within only 15 min, demonstrating that lattice modification can accelerate adsorption kinetics while maintaining high removal efficiency.⁵³ Similarly, immobilizing ZnO nanoparticles on halloysite nanoclay or cellulose nanocrystals improves particle dispersion, reduces agglomeration, and exposes more accessible adsorption sites, thereby maintaining high dye removal efficiencies over a wide concentration range.^54,55 These findings demonstrate that support materials contribute not only mechanical stability but also additional functional groups that facilitate dye adsorption.

Green synthesis has emerged as an effective strategy for improving both adsorption performance and environmental sustainability. Plant extracts and other biological resources introduce oxygen-containing functional groups that serve as natural capping agents, improving nanoparticle dispersion and creating additional adsorption sites. For example, teff straw-assisted ZnO nanoparticles exhibited nearly complete MB removal under optimized conditions and retained high efficiency (92.5%) during treatment of real textile wastewater, highlighting the practical advantages of bio-derived surface functionalities (Fig. 6a).⁵⁶ The proposed mechanism involves electrostatic attraction between positively charged dye molecules and negatively charged ZnO surfaces, while biomolecules derived from teff straw provide additional binding sites that strengthen dye adsorption.

Fig. 6. (a) Proposed adsorption mechanism of methylene blue dye onto teff straw-assisted ZnO nanoparticles. (b) Schematic mechanism of methyl orange dye degradation using ZnO and ZnO@EPS photocatalysts under UV irradiation, (c) visual observation of rhodamine B dye adsorption using ZnO–rGO nanocomposite at different contact times from 5 to 40 min. Reproduced from ref. 56–58, (Rorissa et al., 2025; Ali et al., 2025; Gbawoquiya et al., 2024) under a Creative Commons Attribution 4.0 International License (CC BY 4.0).

Beyond adsorption, several ZnO-based nanomaterials integrate photocatalytic activity, enabling simultaneous dye adsorption and oxidative degradation. This dual-function mechanism enhances overall treatment efficiency because adsorbed dye molecules are subsequently decomposed by reactive oxygen species generated under light irradiation. Exopolysaccharide-modified ZnO nanoparticles illustrate this synergistic effect, where surface modification improves light absorption, inhibits electron–hole recombination, and enhances the production of hydroxyl and superoxide radicals responsible for methyl orange degradation (Fig. 6b).⁵⁷ Likewise, coupling ZnO with reduced graphene oxide significantly improves charge separation and electron transport while increasing the available adsorption surface, resulting in superior rhodamine B removal compared with pristine ZnO (Fig. 6c).⁵⁸

Copper oxide nanoparticles have also been applied for dye adsorption. Green synthesis of CuO nanoparticles using Malva sylvestris leaf extract produced a rapidly acting adsorbent for Congo red (CR), reaching adsorption equilibrium within approximately 30 min. Despite the short equilibration time, the maximum adsorption capacity was relatively low (6.99 mg g⁻¹) compared with many MOF-, graphene-, and chitosan-based adsorbents. This contrast highlights an important trade-off between adsorption kinetics and capacity: rapid uptake does not necessarily translate into high contaminant loading. The use of a plant extract as both a reducing and stabilizing agent provides an environmentally favorable synthesis route, while the reported reusability supports the potential for repeated application. Nevertheless, improving the density and accessibility of active sites would be necessary to overcome the relatively low adsorption capacity and enhance the practical competitiveness of the CuO nanoparticles.⁵⁹

Fig. 7 shows the effect of different adsorption parameters on Congo red removal using CuO nanoparticles. Fig. 7A shows that removal efficiency decreased as the initial dye concentration increased because active adsorption sites became saturated. Fig. 7B and C indicate that pH strongly affected adsorption, with better removal around neutral pH; the pHpzc was about 9.37, meaning the CuO surface is positively charged below this value and can attract anionic Congo red molecules. Fig. 7D shows that removal decreased with increasing temperature, suggesting an exothermic adsorption process. Fig. 7E shows that the optimum adsorbent dose was 0.02 g, while Fig. 7F and G confirm rapid adsorption within the first 30 min. Fig. 7H and I demonstrate the reusability of CuO nanoparticles, with good removal maintained for several cycles before declining.

Fig. 7. Influence of various parameters on the absorbance and removal of CR dye using CuO NPs: (A) initial CR dye concentration; (B) pH; (C) point of zero charge (pHpzc); (D) temperature; (E) catalyst dosage; (F) contact time; (G) absorbance of CR dye at different contact times; (H) recyclability of CuO NPs for CR dye removal; and (I) absorbance of CR dye after different recycling cycles. Reproduced from ref. 59 (Ameen et al., 2025) under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).

In addition, ben Mosbah et al. investigated CuO nanoparticles for methyl green adsorption over a pH range of 6–8. The study showed that dye adsorption was strongly affected by solution pH, adsorbent dosage, and contact time, with a rapid adsorption stage occurring within the first 60 min.⁶⁰

Biogenic Al₂O₃ nanoparticles derived from waste pharmaceutical packaging exhibited high efficiency for Congo red (CR) removal, achieving approximately 97.81% removal under optimized conditions. The use of Padina pavonica extract as a reducing and stabilizing agent, together with waste-derived aluminum precursors, provides a potentially sustainable synthesis route that combines waste valorization with dye remediation. The adsorption equilibrium was described by the Langmuir model, suggesting predominant monolayer uptake, while the pseudo-second-order model provided the best kinetic fit. However, the high removal percentage should be interpreted in conjunction with adsorption capacity and operating conditions, since percentage removal alone can be strongly influenced by adsorbent dosage and initial dye concentration. Thus, although the material demonstrates promising removal efficiency, its practical advantage over high-capacity adsorbents cannot be established without comparing q_max, regeneration performance, and adsorption kinetics under comparable conditions.⁶¹

Fig. 8 shows the effect of different adsorption parameters on Congo red removal using P/Al₂O₃-NPs. Fig. 8A shows that adsorption was strongly affected by pH, with high removal near the native/near-neutral pH and lower removal under alkaline conditions due to electrostatic repulsion between negatively charged dye molecules and the negatively charged nanoparticle surface. Fig. 8B shows that increasing the adsorbent dose improved removal up to the optimum dose because more active sites became available, while further increase slightly reduced efficiency due to particle aggregation. Fig. 8C indicates rapid adsorption within the first 30 min, followed by a plateau as active sites became saturated. Fig. 8D shows that removal increased with dye concentration up to 30 mg L⁻¹, then slightly decreased due to limited active sites. Fig. 8E shows that increasing NaCl concentration reduced adsorption because chloride ions competed with Congo red molecules for active sites. Fig. 8F shows that adsorption decreased with increasing temperature, confirming the exothermic nature of the process.

Fig. 8. Effect of various factors on the removal efficiency and adsorption capacity of biogenic P/Al₂O₃-NPs derived from P. pavonica using waste pharmaceutical packages as precursors for CR dye removal: (A) pH; (B) adsorbent dosage; (C) contact time; (D) initial CR dye concentration; (E) ionic strength; and (F) temperature. Reproduced from ref. 61 (Al-Ahmari et al., 2025) under a Creative Commons Attribution 4.0 International License (CC BY 4.0).

Other metal oxide systems have also been reported. Bekele et al. synthesized γ-Al₂O₃ nanoparticles for methylene blue adsorption and optimized the process using response surface methodology, showing that the adsorption efficiency depended on experimental parameters such as pH, adsorbent dose, and contact time.⁶² MgO nanoparticles exhibited an exceptionally high adsorption capacity toward Congo red (CR), reaching 1429 mg g⁻¹, placing this material among the high-capacity adsorbents reported for anionic dye removal. The high uptake can be associated with the abundance of reactive surface sites on MgO and favorable interactions with the anionic dye. The adsorption kinetics were described by a second-order model, indicating that the adsorption rate depends on the interaction between the adsorbate and available surface sites. However, the kinetic model alone should not be interpreted as definitive evidence of strong chemical bonding or chemisorption. Given the unusually high reported capacity, comparison under identical experimental conditions—including initial dye concentration, adsorbent dosage, pH, and equilibrium time—is particularly important before concluding that MgO is intrinsically superior to other high-capacity adsorbents.⁶³ Ag–Cu–CeO₂ nanocomposite exhibited heterogeneous adsorption behavior toward Congo red (CR), with the equilibrium data being better described by the Freundlich isotherm, consistent with adsorption on a heterogeneous surface and the possibility of multilayer uptake. The pseudo-second-order model provided the best kinetic fit, indicating that the adsorption rate is associated with the availability of adsorption sites and adsorbate–surface interactions. The combination of Ag, Cu, and CeO₂ can provide chemically diverse surface sites, potentially contributing to the heterogeneous adsorption behavior. However, fitting the Freundlich and pseudo-second-order models alone does not conclusively establish multilayer adsorption or a specific chemical mechanism; complementary surface characterization and mechanistic analysis are required to distinguish the contributions of electrostatic attraction, surface complexation, and other interactions.⁶⁴

Green-synthesized ZnO nanoparticles derived from Ocimum basilicum L. waste exhibited exceptionally rapid methylene blue (MB) removal, exceeding 90% within only 4 min. This rapid uptake represents a clear kinetic advantage over many conventional adsorbents that require substantially longer contact times. However, the maximum adsorption capacity of 21.41 mg g⁻¹ was relatively modest, indicating a trade-off between adsorption rate and overall uptake capacity. The Freundlich isotherm suggests adsorption on a heterogeneous surface, while the pseudo-second-order model provided the best kinetic description. Importantly, these model fits should not be interpreted alone as evidence for multilayer adsorption or chemisorption. The use of plant waste as a precursor further strengthens the sustainability aspect of the material, although its relatively low capacity may limit its competitiveness in applications requiring high contaminant loading.⁶⁵

Fig. 9 illustrates the proposed synthesis and adsorption mechanism of ObL-ZnONp for methylene blue removal. The phytochemicals present in Ocimum basilicum extract, such as phenolics, flavonoids, tannins, alkaloids, ketones, ethers, and reducing sugars, act as reducing, stabilizing, and capping agents during ZnO nanoparticle formation. Zinc nitrate provides Zn²⁺ ions, which are reduced and nucleated into Zn atoms, then stabilized to form ObL-ZnONp. In the adsorption stage, methylene blue molecules interact with the surface of ObL-ZnONp through available functional groups and active sites, leading to efficient dye removal from water. Therefore, Fig. 9 supports the dual role of plant waste extract in nanoparticle synthesis and the role of ObL-ZnONp as an effective adsorbent for MB dye removal.

Fig. 9. Proposed mechanism for green synthesis of ObL-ZnONp using Ocimum basilicum L. waste extract and methylene blue dye removal by adsorption. Reproduced from ref. 65 (Turna et al., 2025) under a Creative Commons Attribution 4.0 International License (CC BY 4.0).

Green-synthesized NiO/ZnO nanocomposites exhibited strong photocatalytic activity toward methylene blue (MB), achieving 97.64% removal under natural sunlight and 94.02% degradation under UV irradiation. The higher performance under natural sunlight is particularly noteworthy, as it suggests that the heterojunction can utilize a broader portion of the solar spectrum and offers potential for lower-energy treatment compared with UV-dependent systems. The degradation process followed pseudo-first-order kinetics, providing a useful description of the reaction rate. However, the reported removal should be distinguished from adsorption-based uptake, since photocatalytic degradation transforms the dye rather than simply transferring it to the catalyst surface. Thus, the key advantage of the NiO/ZnO system lies in its ability to couple semiconductor photocatalysis with green synthesis, although assessment of mineralization, degradation intermediates, catalyst stability, and reusability would be necessary to establish its practical treatment potential.⁶⁶

Fig. 10 shows the effect of JA-NiO/ZnO catalyst dosage on methylene blue removal under natural sunlight. Fig. 10a shows that increasing the catalyst dose from 0.2 to 0.8 mg mL⁻¹ enhanced dye removal because more active sites and photocatalytic surfaces became available. However, further increasing the dose to 1.0 and 1.2 mg mL⁻¹ reduced the removal efficiency, probably due to catalyst agglomeration and reduced light penetration. Fig. 10b confirms the pseudo-first-order kinetic behavior through the linear relationship of ln(C₀/C_t) with irradiation time. Fig. 10c shows the gradual decrease in MB absorbance with time at the optimum dose, confirming dye degradation. Fig. 10d shows that 0.8 mg mL⁻¹ was the optimum catalyst dose, achieving 97.64% MB removal within 90 min with a rate constant of 0.04072 min⁻¹.

Fig. 10. Effect of JA-NiO/ZnO dosage on the removal of MB under natural sunlight: (a) percentage removal as a function of irradiation time; (b) ln(C₀/C_t) as a function of irradiation time; (c) absorbance spectra of the dye solution; and (d) degradation kinetics as a function of irradiation time at the optimized JA-NiO/ZnO dosage (0.80 mg mL⁻¹). Reproduced from ref. 66 (Singh et al., 2024), Elsevier, license no. 6272710349146.

6.2. Metallic nanoparticles

Metallic nanoparticles have also been used for dye adsorption because of their high surface reactivity, electron-transfer ability, and strong interaction with dye molecules. Batool et al. synthesized biosilver nanoparticles for methylene blue adsorption and reported a maximum adsorption capacity of 121.04 mg g⁻¹. The adsorption followed the Langmuir isotherm and pseudo-second-order kinetic model, indicating monolayer adsorption controlled by strong surface interactions.⁶⁷ Green-synthesized silver nanoparticles using Urena lobata extract were evaluated for methylene blue adsorption across pH 2–12 and contact times of 5–120 min. Dye uptake was strongly influenced by both parameters, highlighting the importance of surface charge and contact time in controlling adsorption. The study demonstrates the potential of plant-mediated Ag nanoparticles as adsorbents, although adsorption capacity and regeneration performance are needed to assess their practical competitiveness.⁶⁸

Several studies have also focused on silver nanoparticles for rhodamine B removal. Cobweb-assisted silver nanoparticles showed effective rhodamine B adsorption, reaching a maximum capacity of 59.85 mg g⁻¹ at an optimum pH of 3.4. The Langmuir model provided an excellent fit (R² = 0.9901), while the pseudo-second-order model described the adsorption kinetics. These results indicate favorable uptake on relatively uniform adsorption sites, although the kinetic model alone does not confirm chemisorption. The use of cobwebs as a supporting material also offers an unconventional low-cost approach to nanoparticle-based adsorption.⁶⁹ Silver nanoparticles synthesized using Acacia ehrenbergiana extract were effective for rhodamine B adsorption, with optimum uptake at pH 3 and a contact time of 75 min. The adsorption data were well described by the Langmuir isotherm and pseudo-second-order kinetic model, suggesting favorable adsorption on relatively uniform sites. The plant-mediated synthesis provides a potentially sustainable route to Ag-based adsorbents, although adsorption capacity and regeneration performance are needed to determine their practical advantage over other nanoparticle systems.⁷⁰

Fe–Sn–Ni ternary metal oxide nanocomposite (FSN-MONC) demonstrated effective Congo red (CR) removal, achieving 87.92% removal and a maximum adsorption capacity of 76.87 mg g⁻¹. The relatively high removal efficiency, together with the moderate adsorption capacity, indicates that the material can provide effective dye uptake under the reported conditions, although its performance is not exceptional compared with several high-capacity adsorbents. The Langmuir model described the equilibrium behavior, suggesting predominant monolayer adsorption on a comparatively defined set of adsorption sites, while the pseudo-second-order model provided the best kinetic fit. However, as with other systems, the kinetic model should not be considered sufficient evidence for chemisorption without supporting mechanistic characterization. Overall, the ternary oxide illustrates how combining different metal centers can generate multifunctional adsorption surfaces, but direct comparison with other metal-oxide nanocomposites under identical conditions is needed to determine whether the multicomponent composition provides a meaningful performance advantage.⁷¹

Fig. 11 shows the optimization of adsorption conditions for Congo red removal using FSN-MONC. Fig. 11a presents the point of zero charge (POZC = 8.12), indicating that the adsorbent surface is positively charged below this pH, which favors attraction with anionic Congo red molecules. Fig. 11b shows that adsorption was high in acidic to near-neutral conditions and decreased at higher pH due to electrostatic repulsion. Fig. 11c shows that increasing adsorbent dose improved removal until the optimum dose, after which no major improvement occurred. Fig. 11d indicates that adsorption increased with contact time and reached equilibrium at about 60 min. Fig. 11e shows that higher initial dye concentration increased adsorption capacity, while Fig. 11f demonstrates that adsorption decreased with increasing temperature, confirming the exothermic nature of the process.

Fig. 11. (a) Point of zero charge (pHpzc) analysis of FSN-MONC; optimization of batch adsorption parameters, including (b) pH, (c) adsorbent dosage, (d) contact time, (e) initial CR dye concentration, and (f) temperature, for the adsorption of CR dye onto FSN-MONC. Reproduced from ref. 71 (Rehman et al., 2026) under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0).

Nano zero-valent iron has received particular attention because it can remove dyes through adsorption, reduction, and surface complexation. A comparative assessment of nZVI, nanoclay, and iron-impregnated nanoclay demonstrated that iron incorporation substantially enhanced methylene blue (MB) removal, with the iron-impregnated nanoclay achieving approximately 99.9% removal. The superior performance relative to the individual components indicates that combining iron species with the nanoclay support can create a more favorable adsorption environment, potentially by increasing the availability of reactive sites and modifying surface charge. The pseudo-second-order model provided the best kinetic description, although this fit should not be interpreted alone as evidence of chemisorption. More importantly, the direct comparison among three materials strengthens the conclusion that support modification can be more effective than using either nZVI or nanoclay independently. However, adsorption capacity, regeneration behavior, and iron leaching should also be considered before translating the high removal efficiency into a practical-performance advantage.⁷² nZVI-coated corncob biochar was developed for simultaneous methylene blue and methyl orange removal in a continuous-flow column. The composite achieved 89.64% removal with a maximum adsorption capacity of 118.06 mg g⁻¹, while the Thomas model accurately described the column adsorption behavior (R² = 0.9886). The use of a column system strengthens the practical relevance of the material for continuous wastewater treatment compared with conventional batch adsorption studies.⁷³ Green-synthesized zero-valent iron nanoparticles derived from sweet lime pulp achieved 98.9% methylene blue (MB) removal under optimized conditions (10 mg L⁻¹ initial concentration, 1.2 g L⁻¹ adsorbent dose, and 25 °C). The Freundlich and Sips models provided good fits (R² ≈ 0.98), indicating heterogeneous adsorption behavior. The use of fruit waste as a precursor adds a sustainability advantage, although the relatively high adsorbent dosage should be considered when assessing the material's practical and economic feasibility.⁷⁴ Green-synthesized copper nanoparticles achieved up to 97% removal of Reactive Yellow S3R at around pH 6, with adsorption increasing with contact time and adsorbent dosage. The material showed a high maximum adsorption capacity of 136.99 mg g⁻¹, while the Langmuir (R² = 0.9992) and pseudo-second-order (R² = 0.9993) models provided excellent fits. These results indicate favorable adsorption on relatively homogeneous sites, although the kinetic fit alone does not confirm chemisorption. The combination of high capacity, green synthesis, and effective dye removal highlights the potential of Cu nanoparticles for treating reactive dyes.⁷⁵ A humic acid/gold nanocomposite showed effective methylene blue (MB) adsorption, with optimum performance at pH 7 and equilibrium reached within approximately 90 min at 30 ppm MB. The Langmuir model provided a better equilibrium fit than Freundlich (R² = 0.992), suggesting predominant monolayer adsorption on relatively uniform sites, while the pseudo-first-order model best described the kinetics. The neutral optimum pH and relatively moderate equilibrium time indicate that surface chemistry plays an important role in MB uptake. However, without a reported adsorption capacity, the material cannot be directly compared with high-capacity adsorbents, limiting assessment of its practical advantage.⁷⁶ Wu et al. developed Fe⁰-modified tea-residual biochar for methylene blue adsorption. The material exhibited a high adsorption capacity of 452.5 mg g⁻¹ under optimized conditions, including pH 7.0, 30 °C, 60 min contact time, 0.33 g L⁻¹ adsorbent dose, and 150 mg per L MB concentration. The high capacity was attributed to the combined effect of zero-valent iron and biochar surface functionality, which enhanced adsorption through electrostatic interaction, pore filling, and surface complexation.⁷⁷

6.3. Spinel oxide nanoparticles

Spinel oxide nanoparticles, especially ferrites, are attractive adsorbents because they combine high adsorption ability with magnetic separation. Simonescu et al. synthesized CoFe₂O₄ nanoparticles for Congo red and methyl orange adsorption and reported maximum adsorption capacities of 162.68 mg g⁻¹ for Congo red and 94.46 mg g⁻¹ for methyl orange. The adsorption was pH-dependent and followed the Langmuir isotherm, indicating monolayer adsorption on the ferrite surface.³¹ In another study, surfactant-assisted CoFe₂O₄ nanoparticles achieved adsorption capacities of 182.7 mg g⁻¹ for Congo red and 256.3 mg g⁻¹ for methyl orange. The enhanced uptake was attributed to surfactant-induced changes in surface properties, textural characteristics, and the accessibility of active sites. The higher capacity toward methyl orange suggests that dye–surface interactions are influenced by the specific molecular structure and charge distribution of the adsorbate. Overall, surfactant modification demonstrates that tailoring ferrite surface chemistry can substantially improve anionic dye adsorption, although the stability and potential leaching of the surfactant should be considered when assessing practical applicability.⁷⁸ Magnesium ferrite nanoparticles have also shown excellent adsorption capacity. Ultrasonication-assisted MgFe₂O₄ nanoparticles achieved a high methylene blue (MB) adsorption capacity of 537.63 mg g⁻¹, with rapid uptake under sonication. The Langmuir isotherm and pseudo-second-order kinetic model provided good fits (R² = 0.9922), indicating favorable adsorption on the MgFe₂O₄ surface. The exceptionally high capacity was attributed to enhanced mass transfer under ultrasonication and electrostatic interactions between MB and surface sites. However, the contribution of ultrasonication should be distinguished from the intrinsic adsorption capacity of MgFe₂O₄, since improved mixing and mass transfer can substantially enhance apparent uptake. This highlights ultrasonication as an effective process-intensification strategy, but its additional energy demand should be considered for practical wastewater treatment.⁷⁹ Nickel ferrite-based adsorbents have also been reported for dye removal. Sodium docusate modification substantially enhanced crystal violet adsorption by NiFe₂O₄, increasing the adsorption capacity from 237.53 to 384.62 mg g⁻¹ at the optimum pH of 10. The modified material achieved 91.83% removal within 50 min, while the Langmuir and pseudo-second-order models best described the equilibrium and kinetic data, respectively. The marked improvement after surface modification demonstrates that tailoring ferrite surface chemistry can be more effective than relying on the pristine material alone. However, the specific contribution of sodium docusate to surface charge, active-site availability, and dye–surface interactions should be further clarified to establish the underlying mechanism.⁸⁰ NiFe₂O₄-loaded nanocomposite adsorbents showed enhanced removal of Direct Red 31 and Direct Blue 78 under acidic conditions, with pH 2 favoring adsorption of both anionic dyes. The improved uptake at low pH can be attributed to protonation of surface sites, increasing electrostatic attraction toward the negatively charged dye molecules. This pH-dependent behavior highlights the importance of surface charge in ferrite-based adsorption; however, the need for strongly acidic conditions may limit practical application because pH adjustment would increase treatment requirements and operating costs.⁸¹ Hybrid spinel oxide systems can further improve adsorption due to higher surface area and additional active sites. Ag–Zn–Ni spinel ferrite@mSiO₂ nanoparticles combined a mesoporous silica shell with a magnetic ferrite core for removing methyl green, indigo, and carmine. The mSiO₂ coating increased the accessible surface area and adsorption sites, while the ferrite core enabled magnetic separation after treatment. The Langmuir and pseudo-second-order models provided the best fits for equilibrium and kinetic data, respectively. This core–shell design therefore addresses both adsorption performance and adsorbent recovery, although quantitative comparison with uncoated ferrite and mSiO₂ alone is needed to confirm the extent of the synergistic improvement.⁸² A comparative evaluation of MFe₂O₄ ferrites, including CoFe₂O₄ and ZnFe₂O₄, demonstrated that alkaline conditions generally favored methylene blue (MB) uptake, consistent with enhanced electrostatic attraction between the cationic dye and negatively charged ferrite surfaces. More importantly, the variation in equilibrium behavior among the different ferrite compositions—described by either Langmuir or Freundlich models—indicates that metal substitution can alter the surface heterogeneity and distribution of adsorption sites. The use of both pseudo-first-order and pseudo-second-order models further reflects differences in adsorption kinetics among the ferrites. Rather than identifying a single universally superior ferrite, these results suggest that the chemical composition of the ferrite governs both surface heterogeneity and dye–surface interactions, making metal-ion selection an important design parameter for ferrite-based adsorbents.⁸³ Supporting NiFe₂O₄ on biochar substantially improved methylene blue (MB) adsorption, increasing the maximum adsorption capacity from 35.71 mg g⁻¹ for bare biochar to 68.79 mg g⁻¹ for the NiFe₂O₄/biochar composite. This nearly twofold enhancement indicates that heterostructure formation can modify the surface chemistry and increase the accessibility or density of adsorption-active sites. Importantly, the comparison with the unmodified biochar provides stronger evidence for the benefit of NiFe₂O₄ incorporation than a capacity value reported for the composite alone. Nevertheless, attributing the improvement solely to an increased number of active sites may be an oversimplification, because changes in surface charge, pore structure, and MB–surface interactions may also contribute. The study therefore supports heterostructure engineering as an effective strategy for improving biochar-based adsorption, while highlighting the need to distinguish the individual contributions of surface area and chemical functionality.⁸⁴ A magnetic activated carbon/ZnFe₂O₄ nanocomposite demonstrated effective adsorption of crystal violet (CV), with the optimum uptake occurring at pH 7.2. The Langmuir adsorption capacity increased from 208.29 to 246.19 mg g⁻¹ as the temperature increased from 23 to 45 °C, indicating a favorable temperature-dependent enhancement in dye uptake and suggesting that higher temperature facilitates access to adsorption sites. The pseudo-second-order model provided the best kinetic fit, while diffusion analysis indicated that both external film diffusion and intraparticle diffusion contributed to the overall adsorption process. This is important because it shows that adsorption was not controlled by surface interactions alone; mass transfer through the boundary layer and into the porous structure also influenced the uptake rate. The combination of activated carbon with magnetic ZnFe₂O₄ therefore provides both high adsorption capacity and the potential for magnetic separation, although the increasing capacity with temperature may have implications for the energy requirements of practical treatment systems.⁸⁵ Incorporating hydroxyapatite into CoFe₂O₄ substantially increased the BET surface area from approximately 15 to 34 m² g⁻¹, demonstrating how heterostructure formation can improve the accessibility of adsorption sites. The resulting CoFe₂O₄@hydroxyapatite composite achieved a Congo red (CR) adsorption capacity of approximately 15.25 mg g⁻¹, with adsorption occurring rapidly during the initial stage; about 71% of the maximum uptake was reached within only 5 min. Although the pseudo-second-order model provided an excellent kinetic fit (R² = 0.9984), this correlation should not be taken as conclusive evidence of strong chemical interaction or chemisorption. More importantly, the relatively low adsorption capacity despite the increased surface area suggests that surface area alone does not determine CR uptake, and that the chemical nature and accessibility of surface sites are also critical. Thus, the study highlights the importance of optimizing both pore structure and surface chemistry rather than relying solely on increasing BET surface area.⁸⁶ Co_0.5Mn_0.5Fe₂O₄ nanoparticles provided a magnetic platform for Congo red (CR) adsorption, combining nanoscale dimensions (≈31.5 nm) with a saturation magnetization of 25.25 emu g⁻¹ that enabled magnetic recovery after treatment. The Freundlich isotherm described the equilibrium behavior better than the Langmuir model, indicating a heterogeneous adsorption environment, while the pseudo-second-order model provided the best kinetic fit. However, these model fits alone do not establish multilayer adsorption or chemisorption; the thermodynamic results provide additional evidence that the process was exothermic and involved favorable adsorbate–surface interactions. The combination of adsorption functionality and magnetic recoverability is particularly relevant for practical applications because it can simplify post-treatment separation. Nevertheless, the relatively moderate magnetic saturation highlights a potential trade-off between maximizing adsorption-active surface chemistry and maintaining sufficient magnetic responsiveness for rapid recovery.⁸⁷ Mesoporous MgFe₂O₄ nanoparticles provided a magnetic platform for the removal of indigo and carmine, with adsorption performance optimized through pH, adsorbent dosage, initial dye concentration, and contact time. The mesoporous architecture is particularly relevant because it can increase the accessibility of active sites and facilitate dye transport into the porous structure, while the magnetic character enables straightforward separation of the adsorbent after treatment. The inclusion of kinetic and isotherm analyses further supports evaluation of the adsorption process beyond simple removal efficiency. However, the practical value of MgFe₂O₄ depends not only on its adsorption performance but also on its regeneration stability and capacity retention over repeated cycles. Thus, the combination of mesoporosity and magnetic recoverability represents the principal design advantage of this material for anionic dye treatment, although quantitative comparisons with non-magnetic mesoporous adsorbents are needed to establish the extent of this advantage.⁸⁸ A magnetic SnFe₂O₄@activated carbon nanocomposite combined the high adsorption potential of activated carbon with the magnetic separability provided by SnFe₂O₄ for crystal violet (CV) removal. Incorporation of activated carbon increased the availability of surface area and adsorption sites, while the magnetic ferrite component addressed a major practical limitation of conventional powdered adsorbents by facilitating solid–liquid separation after treatment. Adsorption performance was strongly dependent on pH, initial dye concentration, adsorbent dosage, and contact time, emphasizing the importance of operating conditions in determining removal efficiency. Thus, the main contribution of this composite is its integration of adsorption capacity with magnetic recoverability, rather than simply increasing the number of adsorption sites. However, quantitative comparison of adsorption capacity and regeneration stability would be necessary to establish whether the added magnetic component provides a meaningful overall advantage over activated carbon alone.⁸⁹

6.4. Metal sulfide nanoparticles

Metal sulfide nanoparticles have also been explored for dye removal due to their reactive sulfide surfaces, tunable morphologies, and potential synergy between adsorption and photocatalysis. Ouni et al. synthesized ZnS nanoparticles using an Artemisia herba-alba extract for the removal of methylene blue. Adsorption was carried out at pH 7, and the maximum adsorption capacity was 30.92 mg g⁻¹. The study suggested that ZnS nanoparticles can contribute to both adsorption and photocatalytic dye removal, and that the adsorption process was described as spontaneous and exothermic.⁹⁰ ZnS nanoparticles were evaluated as adsorbents for Acid Green 25 using response surface methodology (RSM), allowing the combined effects of pH, adsorbent dosage, initial dye concentration, and contact time to be optimized rather than assessed independently. This multivariable approach provides a more realistic assessment of adsorption performance because these parameters can interact and collectively determine dye uptake. The inclusion of kinetic and equilibrium analyses further supports the interpretation of the adsorption process. However, without reporting the adsorption capacity and the specific kinetic and isotherm models that best described the data, the study provides limited evidence for comparing the intrinsic adsorption performance of ZnS with other nanomaterials. Thus, its main contribution lies in demonstrating the value of process optimization rather than establishing ZnS as a superior high-capacity adsorbent.⁹¹ Lead sulfide nanoparticles have been functionalized to improve dye adsorption. Rosly et al. synthesized calix[6]arene-modified PbS nanoparticles for methylene blue adsorption and found that the optimum conditions were pH 6, 20 mg L⁻¹ dye concentration, 44 mg adsorbent dose, and 31 °C. The maximum adsorption capacity was 5.495 mg g⁻¹, and the adsorption followed the Langmuir isotherm and pseudo-second-order kinetic model, indicating monolayer adsorption controlled by strong interaction between dye molecules and active sites.⁹² Molybdenum sulfide-based materials have also shown strong adsorption ability. Schneider et al. immobilized MoS₂ nanoflakes on PLA fibers for methylene blue adsorption and reported a maximum adsorption capacity of 111.2 mg g⁻¹. The immobilization of MoS₂ improved material handling and allowed the adsorbent to be used in a more practical form.⁹³ Xiong et al. prepared a MoS₂–reduced graphene oxide composite for Congo red adsorption and reported rapid adsorption with high selectivity toward Congo red. The improved adsorption was attributed to the combined effect of MoS₂ active sites and the large surface area of reduced graphene oxide.⁹⁴ MoS₂/MXene composite exhibited exceptionally rapid removal of rhodamine B (RhB), achieving approximately 99% removal within only 11 min. This rapid uptake can be associated with the highly accessible surface of the layered composite, which facilitates contact between RhB molecules and active adsorption sites. However, the relatively low adsorption capacity of 12.82 mg g⁻¹ indicates a clear distinction between removal efficiency and adsorption capacity: the material removes a high fraction of dye under the reported experimental conditions but cannot accommodate a particularly large dye mass per unit adsorbent. Therefore, its principal advantage appears to be rapid treatment rather than high-capacity adsorption. This trade-off highlights the need to evaluate both kinetics and adsorption capacity when comparing layered nanomaterials for practical dye removal.⁹⁵ Other metal sulfide systems have also been used for the adsorption of organic dyes. Khan et al. synthesized cobalt sulfide, Co₄S₃, for the adsorption of organic dyes and found that the adsorption followed the pseudo-second-order kinetic model, indicating that the dye removal was mainly controlled by surface interaction.⁹⁶ Flower-like MoS₂ nanostructures demonstrated preferential adsorption toward cationic dyes over anionic dyes, indicating that surface charge and the layered sulfide architecture strongly influence dye–adsorbent interactions. The enhanced affinity for cationic species was attributed primarily to electrostatic attraction, while the high surface accessibility associated with the flower-like morphology may further facilitate dye uptake. The pseudo-second-order model described the adsorption kinetics, although this kinetic fit alone should not be interpreted as definitive evidence of chemisorption. Importantly, the ability to regenerate and reuse the MoS₂ adsorbent distinguishes this system from materials whose high initial adsorption capacities are accompanied by poor cycling stability. Thus, the key advantage of the flower-like MoS₂ structure lies not only in its dye affinity but also in its combination of selective adsorption and regenerability, both of which are important for practical wastewater treatment.⁹⁷ The MoS₂/WO₃ composite exhibited heterogeneous adsorption behavior toward methylene blue (MB), with the equilibrium data better described by the Freundlich model. This behavior suggests that the composite provides a heterogeneous distribution of adsorption environments and potentially allows multilayer uptake, although the Freundlich fit alone does not conclusively establish multilayer adsorption. The pseudo-second-order model provided the best kinetic description, while intraparticle-diffusion analysis indicated that diffusion within the porous structure contributed to the overall adsorption process. Notably, adsorption was favored at lower temperatures, indicating an exothermic uptake process and suggesting that elevated temperatures may reduce the affinity of MB for the composite surface. The combined effects of surface heterogeneity and pore diffusion therefore appear to be important in controlling MB uptake.⁹⁸

WS₂ nanosheets exhibited adsorption behavior toward methylene blue (MB) that was strongly influenced by solution pH, initial dye concentration, contact time, and adsorbent dosage, indicating that both surface chemistry and mass-transfer conditions govern dye uptake. The pseudo-first-order model provided a better description of the adsorption kinetics, suggesting that the adsorption rate may be associated with the availability of surface sites and concentration gradients. Although this behavior is consistent with a contribution from physical adsorption, kinetic-model fitting alone cannot conclusively establish physisorption as the dominant mechanism. The use of two-dimensional WS₂ nanosheets is nevertheless attractive because their high exposed surface area can provide accessible sites for dye interaction. Further mechanistic characterization would be required to distinguish physical adsorption from possible electrostatic, van der Waals, or other surface interactions.⁹⁹ Binary Bi₂S₃/ZnS and ZnS/Bi₂S₃ nanocomposites illustrate the potential of integrating adsorption with photocatalytic degradation for methylene blue (MB) removal. Under optimized conditions, a contact time of approximately 120 min and a catalyst dosage of 0.05 g L⁻¹ were identified, while the pseudo-first-order model provided the best kinetic description. However, the kinetic fit alone does not establish the nature of the adsorption mechanism or distinguish surface uptake from photocatalytic degradation. The principal advantage of this hybrid strategy is that adsorption can initially concentrate MB at the catalyst surface, where subsequent photocatalytic reactions may degrade the dye rather than simply transfer it to a solid phase. Nevertheless, the overall removal efficiency should be deconvoluted into adsorption and photodegradation contributions using appropriate dark-adsorption controls and degradation analyses. Without such separation, it remains difficult to determine whether the Bi₂S₃/ZnS heterostructure provides a genuine synergistic effect or whether the observed removal is dominated by one process. Thus, the study highlights the promise of adsorption–photocatalysis coupling, while also emphasizing the need for mechanistic quantification when evaluating multifunctional nanocomposites.¹⁰⁰

6.5. MOF and ZIF-based materials for dye adsorption

MOF- and ZIF-based adsorbents have strong potential for dye removal due to their high porosity, large surface area, tunable functional groups, and adjustable pore structures. Comparison between different MOF/ZIF surfaces for dye adsorption is represented in Table 3. ZIF-8, ZIF-67, UiO-66, MIL-101, and MOF-5-based composites can remove both cationic and anionic dyes via electrostatic attraction, pore-filling, hydrogen bonding, π–π stacking, coordination interactions, and host–guest effects. Amine functionalization substantially enhanced the adsorption performance of MIL-101(Fe) toward methylene blue (MB), increasing the maximum adsorption capacity from 149.25 mg g⁻¹ for the unmodified framework to 312.5 mg g⁻¹ after functionalization. This more than twofold increase demonstrates the importance of surface chemistry in addition to the intrinsic porosity of the MOF, as the introduced amine groups can provide additional interaction sites and alter the surface affinity toward the cationic dye. Adsorption was relatively rapid during the first hour, suggesting efficient accessibility of the modified pore structure. The proposed uptake mechanism involves pore filling together with electrostatic interactions, hydrogen bonding, and π–π stacking, indicating that multiple interactions contribute to MB retention. However, the simultaneous assignment of several mechanisms requires careful experimental validation, since adsorption-capacity enhancement alone cannot distinguish their individual contributions.¹¹⁵ Loading ZIF-8 onto UiO-66-NH₂ substantially enhanced the selective uptake of methylene blue (MB) from mixed-dye solutions containing MB, methyl orange, and rhodamine B. The maximum adsorption capacity increased from 55.04 mg g⁻¹ for the parent UiO-66-NH₂ to 173.01 mg g⁻¹ for the ZIF-8@UiO-66-NH₂ composite, demonstrating that heterostructure formation can significantly improve both adsorption capacity and dye selectivity. The enhanced preference for MB suggests that the composite provides a more favorable combination of pore accessibility, surface chemistry, and adsorbate–framework interactions than either parent MOF. The pseudo-second-order model described the kinetic data very well (R² ≈ 0.999), although this statistical fit should not be interpreted alone as definitive evidence of strong surface bonding or chemisorption. Importantly, the use of a mixed-dye system provides stronger evidence of practical selectivity than single-dye experiments, where competition between different contaminants is not considered.¹¹⁶ Hierarchical pore engineering substantially enhanced the adsorption performance of MIL-101(Cr) toward methyl orange (MO), increasing the maximum adsorption capacity from 229.4 mg g⁻¹ for pristine MIL-101(Cr) to 369.8 mg g⁻¹. This approximately 1.52-fold improvement demonstrates that introducing interconnected pore structures can increase the accessibility of adsorption sites and facilitate mass transfer within the MOF. Adsorption was most favorable at pH 7 and occurred rapidly during the first 20 min, indicating efficient transport of MO molecules to the available adsorption sites. Although the pseudo-second-order model provided an excellent fit (R² = 0.999), this kinetic correlation should be interpreted cautiously and does not, by itself, establish a specific adsorption mechanism. Overall, the comparison with pristine MIL-101(Cr) provides stronger evidence for the beneficial role of hierarchical porosity than a simple report of adsorption capacity, highlighting pore architecture as an important performance bottleneck in MOF-based dye adsorption.¹¹⁷ UiO-66-NO₂ exhibited markedly different adsorption affinities toward methyl orange (MO) and methylene blue (MB), with a substantially higher capacity for the anionic MO (142.9 mg g⁻¹) than for the cationic MB (41.7 mg g⁻¹). The more than threefold difference highlights the importance of dye charge and adsorbate–framework interactions in determining adsorption performance. The stronger affinity toward MO can plausibly be associated with favorable interactions between the electron-deficient functionalized framework and the anionic dye, although the specific contribution of electrostatic, hydrogen-bonding, and pore-confined interactions requires further clarification. The pseudo-second-order model provided a better description of the adsorption kinetics than the pseudo-first-order model, while both Langmuir and Freundlich models were considered for the equilibrium behavior. Overall, the contrasting uptake of MO and MB demonstrates that MOF adsorption performance is highly dye-specific and cannot be predicted solely from surface area or porosity.¹¹⁸ Incorporation of MoS₂ into ZIF-8 enhanced the adsorption of methylene blue (MB), increasing the maximum adsorption capacity from 68.02 mg g⁻¹ for pristine ZIF-8 to 85.52 mg g⁻¹ for the MoS₂/ZIF-8 composite. This improvement indicates that MoS₂ introduces additional adsorption sites and complementary surface interactions, although the relatively modest increase in capacity suggests that the benefit of MoS₂ incorporation is limited compared with some other ZIF-based hybrid systems. The adsorption equilibrium was described by the Langmuir model, consistent with predominant uptake on relatively uniform adsorption sites, while the pseudo-second-order model provided the best kinetic fit. However, the kinetic model alone should not be interpreted as definitive evidence of chemisorption. The proposed contribution of electrostatic attraction, chemisorption, and van der Waals interactions therefore requires support from complementary spectroscopic or surface-chemical evidence to distinguish their relative roles in MB uptake.¹¹⁹

Table 3. MOF-ZIF-based adsorbent materials.

Material/catalyst	Dye removed	Reported efficiency (%)	q max (mg g−1)	Main mechanism	Year	Reference
Fe3O4@MIL-53(Fe,Cu)	Methyl orange (MO)	97	142	Electrostatic attraction, H-bonding, π–π stacking, mesopore filling	2025	132
UiO-66-NH2/TOCNF aerogel	Congo red (CR)	98.7	549.45	Chemisorption; electrostatic attraction; H-bonding; surface complexation; π–π stacking	2025	133
UiO-66-NH2/TOCNF aerogel	Orange II (ORII)	97.7	171.23	Chemisorption; electrostatic attraction; H-bonding; surface complexation; π–π stacking	2025	133
H-MMT/MOF-5 composite	Methylene blue (MB)	97.52	192.3	Electrostatic attraction, π–π interactions, H-bonding; Langmuir adsorption	2023	134
H-MMT/MOF-199 composite	Methylene blue (MB)	98.73	238.09	Electrostatic attraction, π–π interactions, H-bonding; Langmuir adsorption	2023	134
MIL-101(Fe)	Methylene blue (MB)	94	58.82	Adsorption on Fe-MOF active sites; pH/dose dependent uptake	2020	124
ZIF-8	Malachite green (MG)	95	224.14	Freundlich/pseudo-second-order adsorption; pore/surface interactions	2019	135
rGO/ZIF-8 aerogel	Malachite green (MG)	High; pH 3–10	608.8	3D aerogel adsorption; π–π stacking/electrostatic interactions; large surface area	2024	136
Pullulan gel/ZIF-8@PDA hydrogel	Malachite green (MG)	>88 after 5 cycles	615.4	pH-responsive surface charge conversion; electrostatic and π–π interactions	2024	126
Pullulan gel/ZIF-8@PDA hydrogel	Congo red (CR)	>88 after 5 cycles	525.8	pH-responsive surface charge conversion; π–π interactions for CR; adsorption	2024	126
ATTM@ZIF-8 modified pullulan hydrogel	Malachite green (MG)	High in real water systems	1645.28	Electrostatic attraction, π–π stacking; pseudo-second-order adsorption	2024	137
ATTM@ZIF-8 modified pullulan hydrogel	Congo red (CR)	High in real water systems	680.33	π–π stacking and surface interactions; pseudo-second-order adsorption	2024	137
CA-β-CD-MOF	Congo red (CR)	≈90 retention after 4 cycles	900.9	Electrostatic attraction between protonated CD-MOF surface and anionic CR	2023	125
Zn-MOF	Aniline blue (AB)	N/R	55.34	Pseudo-second-order adsorption; surface interaction on Zn-MOF	2023	138
ZnCuCr-based TpIm-functionalized MOF	Congo red (CR)	High optimized removal	N/A	Functionalized MOF adsorption; likely electrostatic and pore filling	2024	139
Ce-BTC@microcrystalline cellulose	Congo red (CR)	98 at low pH	N/A	pH-dependent electrostatic attraction; MOF/cellulose composite adsorption	2025	140
Cu-MOF with novel linker	Tropaeolin OO/Congo red	89–96	N/A	Multilayer adsorption; spontaneous dye removal	2025	141
MIL-101(Fe)@MnFe2O4	Methylene blue (MB)	98.6	N/A	Magnetic MOF composite adsorption; electrostatic/pore interactions	2026	142
Fe-UiO-66-NH2-PEI	Organic dyes	>98 after 5 cycles	N/A	Chemisorption accelerated by Fe; PEI amine groups; electrostatic interaction	2025	120
Mesoporous zinc-based MOF-5	Congo red (CR)	99.73	351.86	Selective adsorption through weak hydrogen bonding and π–π stacking; Langmuir/PSO behavior	2025	143
[Ni2F2(4,4′-bipy)2(H2O)2](VO3)2·8H2O MOF	Congo red (CR)	Optimized high removal	242.1	Sonochemically synthesized MOF; electrostatic adsorption and pore filling; BBD optimization	2020	144
ZIF-8@GO composite	Methyl orange (MO)	High removal at acidic pH	82.78	GO-assisted ZIF-8 adsorption; electrostatic attraction at pH below pHpzc and pore diffusion	2025	127
ZIF-8@GO composite	Methylene blue (MB)	High removal under optimized conditions	12.36	Adsorption by GO/ZIF-8 active sites; electrostatic interaction and surface complexation	2025	127
Cationic Cu-organic framework with 3D cages	Congo red (CR)	High targeted adsorption	N/A	Charge-selective adsorption in cationic cages; electrostatic attraction and π–π interaction	2024	123
AgCl-grafted NH2-MIL-101(Fe)	Congo red (CR)	Efficient removal reported	N/A	MOF adsorption coupled with AgCl-assisted photocatalytic/charge-transfer pathway	2023	145
GO/ZIF-8/HKUST-1 composite	Congo red (CR)	Enhanced removal	N/A	Synergistic MOF/GO adsorption and photocatalytic degradation; π–π and electrostatic interactions	2023	122
UiO-66@β-cyclodextrin@Fe3O4 hybrid	Organic dyes	High-performance dye removal	N/A	Magnetic MOF/cyclodextrin adsorption; host–guest inclusion, H-bonding, and electrostatic interaction	2025	146

Polyethyleneimine (PEI) grafting substantially enhanced the ability of Fe-UiO-66-NH₂ to remove anionic dyes, with the resulting composite achieving approximately 98% Eosin B removal and a maximum adsorption capacity of 1.58 mmol g⁻¹. The high affinity toward the anionic dye can be attributed to the positively charged amine-rich PEI layer, which provides abundant sites for electrostatic attraction while increasing the density of accessible functional groups. Importantly, the composite maintained more than 98% removal efficiency after five adsorption–desorption cycles, demonstrating that surface functionalization can improve not only adsorption performance but also regeneration stability. This behavior contrasts with systems that exhibit high initial capacities but substantial performance losses during reuse, highlighting the importance of combining adsorption capacity with long-term cycling stability when assessing practical adsorbents.¹²⁰

Fig. 12a illustrates the proposed adsorption mechanism of Eosin B onto Fe-UiO-66-NH₂-PEI. The mechanism involves three main interactions: electrostatic attraction between the positively charged protonated amine groups of PEI and the negatively charged groups of Eosin B, coordination interactions between the dye functional groups and the Fe/Zr metal sites in the MOF structure, and π–π interactions between the aromatic rings of Eosin B and the aromatic/triazine structures in the adsorbent. Therefore, Fig. 12a confirms that the high adsorption efficiency of Fe-UiO-66-NH₂-PEI results from a synergistic mechanism rather than a single adsorption pathway.

Fig. 12. (a) Proposed adsorption mechanism of Eosin B onto Fe-UiO-66-NH₂-PEI through electrostatic attraction, metal coordination, and π–π interactions. (b) Proposed adsorption mechanism of methylene blue on sulfur-incorporated orthorhombic bismuth ferrite prepared using DMSO solvent. Reproduced from ref. 120 and 121 (Qian et al., 2025; Singh et al., 2023) with permission from Elsevier, license no. 6272700539067 and 6272700987999, respectively.

Singh et al. synthesized mixed-phase bismuth ferrite (BFO) particles using a non-aqueous sol–gel method in different solvents, namely THF, DMF, and DMSO, and evaluated their methylene blue adsorption performance. The DMSO-derived sample showed the highest adsorption efficiency of 96.54%, which was attributed to the formation of a more open, less crystalline orthorhombic phase with sulfur incorporation and a needle-like morphology that enhanced dye adsorption.¹²¹

Fig. 12b illustrates the proposed structural explanation for the enhanced adsorption of methylene blue on the BFO-DMS sample. The figure shows the orthorhombic Bi₂Fe₄O₉ phase, where Fe–O octahedra form columnar chains along the c-axis and are linked by Fe–O tetrahedral units. In the DMSO-prepared sample, sulfur from the solvent may replace oxygen sites in the orthorhombic structure, creating additional interaction sites for methylene blue molecules. This sulfur incorporation, together with the open columnar structure, promotes interaction between the π-orbitals of methylene blue and the lone-pair electrons on sulfur, which helps explain the superior adsorption efficiency of BFO-DMS compared with the THF- and DMF-prepared samples.

Integrating graphene oxide (GO) with ZIF-8/HKUST-1 enabled the composite to combine adsorption with photocatalytic degradation of Congo red (CR), with GO(10)/ZIF-8/HKUST-1 delivering the highest performance and achieving 91.81% degradation within 60 min under UV-LED irradiation. The enhanced photocatalytic activity can be attributed to the role of GO as a conductive component that facilitates charge separation and suppresses electron–hole recombination, thereby increasing the availability of reactive charge carriers. This hybrid adsorption–photocatalysis strategy offers an important advantage over conventional adsorbents because the contaminant can potentially be degraded rather than merely transferred from water to a solid phase. However, the optimum GO content also indicates that excessive incorporation of GO may not necessarily improve photocatalytic performance, emphasizing the need to balance interfacial charge transfer with the availability of photoactive sites.¹²²

Fig. 13 shows the effects of pH and scavengers on Congo red degradation using GO(10)/ZIF-8/HKUST-1. Fig. 13a indicates that acidic conditions improved the degradation efficiency, with the highest removal at pH 5, because the photocatalyst surface becomes more positively charged and strongly attracts the anionic Congo red dye. In contrast, increasing the pH reduced degradation efficiency due to electrostatic repulsion between the negatively charged surface and Congo red molecules. Fig. 13b shows that the addition of IPA and ascorbic acid strongly inhibited degradation, confirming that hydroxyl radicals and superoxide radicals are the main active species responsible for Congo red degradation. The weaker inhibition caused by DMSO and methanol indicates that electrons and holes have a less dominant role in the degradation process.

Fig. 13. (a) Effect of solution pH and (b) radical scavengers on Congo red degradation using GO(10)/ZIF-8/HKUST-1 photocatalyst. Reproduced from ref. 122 (Ediati et al., 2022) under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0).

A cationic Cu-MOF with a three-dimensional cage structure exhibited selective adsorption toward Congo red (CR), achieving a maximum adsorption capacity of 119.76 mg g⁻¹. The selective uptake highlights the importance of electrostatic and pore-confined interactions between the cationic framework and the anionic dye. More importantly, the nearly complete recovery of adsorbed CR using a NaCl–DMF solution and the sustained performance over repeated adsorption–desorption cycles demonstrate that the material is designed not only for dye removal but also for dye recovery and adsorbent regeneration. However, its adsorption capacity remains moderate compared with several high-capacity MOF and graphene-based composites. Therefore, its principal advantage lies in selectivity and recoverability rather than maximum adsorption capacity, illustrating that practical adsorbent performance should be evaluated using multiple criteria rather than q_max alone.¹²³

Fig. 14 demonstrates the selective adsorption behavior of Cu-MOF toward Congo red compared with other dyes. Fig. 14a shows that after treatment with Cu-MOF, only the CR solution became colorless, while the other anionic, cationic, and neutral dye solutions remained almost unchanged, confirming the high selectivity of Cu-MOF for CR. Fig. 14b supports this observation by showing a significant decrease in CR absorbance, whereas the other dyes showed only slight changes in absorbance. Fig. 14c shows the gradual decrease in the UV-vis absorption peak of CR with increasing contact time, indicating rapid and efficient CR adsorption. Fig. 14d shows that the PXRD pattern of Cu-MOF after CR adsorption remained similar to the original material, confirming that the Cu-MOF framework maintained good structural stability during adsorption.

Fig. 14. (a) Illustration of the color changes of dye solutions during the dye adsorption experiments and the corresponding molecular structures; (b) change in absorbance before and after dye adsorption; (c) UV-vis absorption spectra of CR during adsorption at different contact times; and (d) PXRD patterns of Cu-MOF before and after CR adsorption. Reproduced from ref. 123 (Wang et al., 2024), with permission from Elsevier, license no. 6272700731680.

MIL-101(Fe) exhibited favorable adsorption toward the cationic methylene blue (MB), reaching a maximum adsorption capacity of 58.82 mg g⁻¹ at pH 9. The enhanced uptake under alkaline conditions suggests that pH-dependent surface charge plays an important role in promoting interactions between the MOF and the cationic dye. The Langmuir model provided the best fit to the equilibrium data, indicating predominant monolayer adsorption on relatively homogeneous sites, while the pseudo-second-order model described the adsorption kinetics. However, the relatively low adsorption capacity compared with several graphene-, chitosan-, and MOF-based composites reported in the literature indicates that the intrinsic adsorption performance of MIL-101(Fe) may be limited without further surface modification. This comparison highlights the potential benefit of functionalizing MOFs to increase the density and accessibility of active adsorption sites rather than relying solely on their porous structure.¹²⁴

Fig. 15 presents the selectivity and reusability performance of MIL-101(Fe). Fig. 15a compares the adsorption of cationic MB and anionic methyl orange (MO) dyes at different concentrations, showing that MIL-101(Fe) adsorbed MB much more effectively than MO. At 200 mg L⁻¹, the adsorption capacity reached 58.82 mg g⁻¹ for MB but only 11.24 mg g⁻¹ for MO, confirming the higher affinity of MIL-101(Fe) toward cationic dyes. This behavior is mainly attributed to the interaction between the Lewis basic –N(CH₃)₂ groups of MB and the Lewis acidic Fe sites of MIL-101(Fe), in addition to possible π–π interactions. Fig. 15b shows that MIL-101(Fe) maintained good reusability, with MB removal efficiency remaining above 70% after ten adsorption–desorption cycles, indicating its stability and practical potential as a reusable adsorbent.

Fig. 15. (a) Adsorption performance of MIL-101(Fe) toward the cationic MB and anionic MO dyes; and (b) reusability of MIL-101(Fe) for MB dye removal. Reproduced from ref. 124 (Eltaweil et al., 2020) with permission from Elsevier, license no. 6272700881953.

Citric acid modification transformed β-cyclodextrin metal–organic frameworks (CA-β-CD-MOFs) into a water-stable and recyclable platform for the simultaneous removal of Congo red (CR) and Cu(ii). The optimized CA-β-CD-MOF-1 exhibited a remarkably high adsorption capacity of 900.9 mg g⁻¹ for CR, compared with 287.4 mg g⁻¹ for Cu(ii), highlighting the strong dependence of adsorption performance on the nature and charge characteristics of the target species. The Langmuir model provided the best description of the equilibrium data, suggesting predominant monolayer adsorption on relatively homogeneous sites, while the pseudo-second-order model adequately described the adsorption kinetics. The substantially higher affinity toward CR suggests that the modified framework provides particularly favorable interactions with the large anionic dye molecules. Importantly, the reported water stability and recyclability address a major limitation of conventional MOFs, although direct comparison with other high-capacity adsorbents should account for differences in solution chemistry, adsorbent dosage, and equilibrium conditions.¹²⁵

Fig. 16 shows the effects of citric acid modification, adsorbent dosage, and solution pH on the adsorption of CR and Cu(ii). Fig. 16a and d show that citric acid modification improved the removal efficiency compared with β-CD-MOF because the introduced –COOH groups provided additional adsorption sites; however, excessive citric acid reduced efficiency due to electrostatic repulsion for CR and possible water-cluster formation that hindered Cu(ii) complexation. Fig. 16b and e show that increasing the adsorbent dosage improved CR and Cu(ii) removal until the optimum dosage, after which the increase became limited due to competition and saturation of active sites. Fig. 16c shows that CR removal was highest at pH 4 and decreased at higher pH values, suggesting electrostatic interaction between CR and CA-β-CD-MOF. Fig. 16f shows that Cu(ii) removal increased with pH because fewer H⁺ ions competed with Cu(ii) for adsorption sites, allowing better complexation with carboxyl groups.

Fig. 16. Effect of various parameters on the adsorption of CR and Cu(ii) onto CA-β-CD-MOF-1: (a and d) effect of sample type on the adsorption capacities of CR and Cu(ii), respectively; (b and e) effect of adsorbent type on the adsorption of CR and Cu(ii), respectively; and (c and f) effect of solution pH on the adsorption of CR and Cu(ii), respectively, onto CA-β-CD-MOF-1. Reproduced from ref. 125 (Zhang et al., 2023) with permission from Elsevier, license no. 627270111022.

PDA-modified ZIF-8 encapsulated within a pullulan-based nanocomposite hydrogel (ZPP10) provided a high-capacity platform for the simultaneous treatment of structurally different organic dyes. The optimized hydrogel achieved adsorption capacities of 615.4 mg g⁻¹ for malachite green (MG) and 525.8 mg g⁻¹ for Congo red (CR), demonstrating that the hybrid structure can accommodate both cationic and anionic dyes. The high uptake can be associated with the combination of the porous ZIF-8 framework, PDA-derived functional groups, and the hydrophilic pullulan matrix, which collectively provide diverse interaction sites. Importantly, the hydrogel maintained its adsorption performance and structural stability during repeated use, addressing a common limitation of powdered MOF-based adsorbents. Nevertheless, the difference in capacity between MG and CR indicates that dye-specific interactions remain important, and further mechanistic analysis would be needed to determine how electrostatic interactions, hydrogen bonding, π–π interactions, and pore accessibility contribute to the adsorption of each dye.¹²⁶

Fig. 17 presents the proposed adsorption mechanism, recyclability, and biocompatibility of ZPP10. Fig. 17a shows that MG and CR adsorption occurs through several interactions, including hydrogen bonding, electrostatic attraction, van der Waals forces, and π–π stacking between dye molecules and the ZPP10 hydrogel network. The PDA coating and ZIF-8@PDA particles provide additional active sites, while the pullulan hydrogel structure improves separability and prevents nanoparticle aggregation. Fig. 17b confirms the good reusability of ZPP10, as the adsorption capacity remained high after five cycles, retaining about 88% for MG and 93% for CR. Fig. 17c shows the live/dead cell assay, indicating good cytocompatibility and supporting the safe potential application of ZPP10 in wastewater treatment.

Fig. 17. (a) Proposed adsorption mechanisms for the removal of CR and MG using the ZPP10 composite hydrogel; (b) recyclability of ZPP10 for CR and MG adsorption; and (c) cytotoxicity evaluation of ZPP10 using a live/dead assay. Reproduced from ref. 126 (Su et al., 2024) with permission from Elsevier, license no. 6272701243619.

Room-temperature fabrication of ZIF-8/GO composites produced an adsorbent capable of efficiently removing both cationic methylene blue (MB) and anionic methyl orange (MO), demonstrating that GO incorporation can broaden the applicability of ZIF-8 toward dyes with different charges. The ZIF-8@0.5GO composite achieved 95.2% MB removal and 94.6% MO removal within 60 min, with corresponding adsorption capacities of 87.39 and 82.78 mg g⁻¹, respectively. The comparable performance toward oppositely charged dyes suggests that adsorption cannot be attributed solely to electrostatic attraction and may instead arise from the combined contributions of the porous ZIF-8 structure and the diverse surface functionalities introduced by GO. Nevertheless, the moderate adsorption capacities relative to several GO-based hybrid adsorbents indicate that high removal efficiency should not be interpreted as evidence of exceptionally high adsorption capacity. A direct comparison under identical initial concentrations, adsorbent dosages, and equilibrium conditions would be required to establish the relative advantage of the composite.¹²⁷

Fig. 18 shows the effect of temperature and thermodynamic behavior of MB and MO adsorption onto GO, ZIF-8, ZIF-8@0.1GO, and ZIF-8@0.5GO. Fig. 18a and b show that increasing temperature from 20 to 80 °C enhanced the removal efficiency of both MB and MO, indicating that the adsorption process is endothermic and favored at higher temperature. Fig. 18c and d present the van't Hoff plots used to calculate the thermodynamic parameters. The negative ΔG° values confirm that dye adsorption is spontaneous and thermodynamically favorable, while the positive ΔH° values indicate an endothermic process. In addition, the positive ΔS° values suggest increased randomness at the solid–liquid interface during adsorption. These results support the strong interaction between dye molecules and the ZIF-8@GO composites, especially due to the synergistic effect between porous ZIF-8 and GO sheets.

Fig. 18. Effect of various parameters on the degradation of MB and MO dyes: (a and b) effect of temperature on the removal of MB and MO dyes, respectively (adsorbent dosage = 2 g L⁻¹, C₀ = 50 mg L⁻¹, and pH = 8 for MB and 5.5 for MO); and (c and d) van’t Hoff plots for the adsorption of MB and MO dyes, respectively, onto GO, ZIF-8, ZIF-8@0.1GO, and ZIF-8@0.5GO. Reproduced from ref. 127 (Younis et al., 2025) under the terms of the Creative Commons Attribution-Non-Commercial 4.0 International License (CC BY-NC 4.0).

Zn/Co-ZIF@CNT composites demonstrated effective removal of both methyl orange (MO) and methylene blue (MB), with removal efficiencies exceeding 85% under optimized conditions. The composite architecture combines the high surface area and porous structure of ZIFs with the additional adsorption sites provided by CNTs, offering a complementary surface for interactions with structurally different dyes. The distinct equilibrium behavior of the two dyes is particularly informative: MO adsorption followed the Langmuir model, suggesting relatively uniform adsorption sites and predominant monolayer uptake, whereas MB was better described by the Freundlich model, indicating greater surface heterogeneity and possible multilayer adsorption. Both systems followed pseudo-second-order kinetics, highlighting the contribution of surface interactions to the overall adsorption rate. However, the comparison of removal percentages should be interpreted alongside the adsorption capacity and adsorbent dosage, since removal efficiency is strongly dependent on the experimental conditions, including the use of 30 mg of adsorbent for 30 mL of dye solution.¹²⁸ ZIF-67@bentonite nanocomposite exhibited adsorption capacities ranging from 38 to 187 mg g⁻¹ for methyl orange, with the substantial variation reflecting the strong influence of the operating conditions on dye uptake. The Avrami kinetic model provided the best fit to the experimental data (R² = 0.9949), indicating that the adsorption process may involve multiple kinetic pathways or heterogeneous adsorption behavior rather than a single well-defined mechanism. FTIR analysis suggested that electrostatic attraction, hydrogen bonding, and coordination interactions with Co²⁺ and Al³⁺ sites contribute collectively to methyl orange uptake. The coexistence of these interactions highlights the multifunctional nature of the ZIF-67/bentonite surface; however, spectroscopic evidence alone does not establish the relative contribution of each mechanism. Further comparison of adsorption capacities under identical experimental conditions would therefore be necessary to determine the actual advantage of this hybrid adsorbent over other materials.¹²⁹ Malachite green adsorption by MOF-5 was strongly pH-dependent, with the removal efficiency increasing from approximately 32% at pH 4 to 67% at pH 12. The improved uptake under alkaline conditions suggests that surface charge and the ionization of functional groups play important roles in the interaction with the cationic dye. Although aminated corn-stover-derived reduced graphene oxide exhibited superior adsorption performance, the response of MOF-5 to pH variation demonstrates its potential as an adsorbent. However, the comparison also exposes a critical limitation of MOF-based materials: structural stability in aqueous media. Thus, high adsorption performance alone may not be sufficient to establish practical applicability, particularly when prolonged exposure to water can compromise the framework structure and potentially affect its adsorption capacity and reusability.¹³⁰ Chitosan decoration of the bimetallic Fe/MOF-5 framework produced a multifunctional composite film for Congo red adsorption, with a Langmuir maximum adsorption capacity of 219.78 mg g⁻¹. The enhanced adsorption can be associated with the coexistence of multiple interaction sites provided by the MOF and chitosan components, including electrostatic attraction, host–guest interactions, π–π interactions, and coordination bonding. The adsorption kinetics were well described by the pseudo-second-order model, indicating that surface interactions contributed substantially to the adsorption process. Despite the relatively high initial adsorption capacity, the material retained only about 18.54% of its original adsorption capacity after 10 adsorption–desorption cycles, indicating a pronounced decline in performance upon repeated use. This loss in capacity represents an important limitation for practical application and suggests that the stability of the composite and the effectiveness of the regeneration process require further improvement.¹³¹

6.6. Carbon-based nanomaterials

Carbon-based and chitosan-containing nanoadsorbents show strong potential for dye adsorption because they combine high surface area, porous structure, π-conjugated carbon frameworks, and abundant functional groups such as –OH, –COOH, and –NH₂. CNTs and rGO mainly enhance adsorption through π–π stacking, pore filling, and electrostatic attraction, while GO and chitosan-based composites provide additional oxygen- and nitrogen-containing binding sites. Overall, these materials frequently follow pseudo-second-order kinetics, while their equilibrium behavior is commonly described by Langmuir or Freundlich isotherms, depending on whether adsorption occurs as a monolayer or a heterogeneous multilayer. Comparison between different carbon based surfaces for dye adsorption is represented in Table 4.

Table 4. Carbon-based adsorbent materials.

Material/catalyst	Dye removed	Reported efficiency (%)	q max (mg g−1)	Main mechanism	Year	Ref.
GO–alginate microbeads	Methylene blue (MB)	>96 after 8 cycles	120.37	Adsorption on GO oxygen groups; spontaneous endothermic adsorption; electrostatic/π–π interactions	2024	163
PAN@PRGO nanocomposite	Indigo carmine (IC)	>99 in water matrices; 100 at ≤300 mg L−1	N/A	Electron-sharing interactions; π–π stacking; pseudo-second-order adsorption	2025	164
Spathodea campanulata activated carbon	Methylene blue (MB)	99.95	N/A	Multilayer adsorption on porous activated carbon; Freundlich best fit	2024	151
Graphene oxide/montmorillonite composite	Methylene blue (MB)	High dye uptake	315	Surface adsorption; GO oxygen groups; electrostatic and π–π interactions	2024	152
Mn–Prussian blue–CNT composite (MnPBCNT)	Methylene blue (MB)	91	31.2	Redox-active coordination sites plus CNT adsorption; electrostatic interaction	2026	153
PKSPMA@MWCNT polymer-brush carbon nanotubes	Methylene blue (MB)	98	411.62	Electrostatic attraction between dye and polymer brush/CNT surface	2024	165
PMETAC@MWCNT polymer-brush carbon nanotubes	Methylene blue (MB)	High	129.22	Electrostatic attraction; cationic/anionic selectivity via polymer brush	2024	165
GO/ZnTiO3/TiO2 composite	Methylene blue (MB)	High optimized removal	N/A	Photocatalysis/adsorption synergy; GO charge transfer reduces recombination	2024	166
MnO2@rGO nanocomposite	Methylene blue (MB)	98–99	156	H-bonding, π–π interaction, electrostatic interaction, chemisorption	2023	167
ZnO–rGO nanocomposite	Rhodamine B (RhB)	99	N/A	Adsorption; π–π interactions on rGO; ZnO–rGO surface sites	2024	58
GO adsorbent	Rhodamine B (RhB)	>84.51 in environmental water	N/A	Adsorption on GO oxygenated groups; sonication/pH-assisted removal	2022	168
Activated carbon–chitosan composite	Methylene blue (MB)	High	22.52	Langmuir adsorption; pseudo-second-order kinetics	2025	169
GO–isatin–PVP composite	Model organic pollutant/dye-like adsorbate	High	119.76	Chemisorption; pseudo-second-order kinetic model	2025	170
GO/silica nanosheets with multistage pores	Congo red (CR)	High	N/R	Hierarchical pore adsorption; electrostatic/π–π interactions	2020	171
Graphene oxide-based composite granule	Methylene blue (MB)	Maintained after six cycles	951.35	Spontaneous exothermic adsorption; pseudo-second-order kinetics; dye condensation on GO-composite surface	2024	157
Rumex abyssinicus-derived activated carbon	Methylene blue (MB)	82.2–99.9	N/A	Porous activated carbon adsorption; Langmuir/kinetic modeling under optimized textile wastewater conditions	2023	172
H3PO4-activated chicken bone carbon	Methylene blue (MB)	92.29	N/A	Carbon/bone-char adsorption; pH-dependent electrostatic interaction and pore filling	2025	154
rGO–biochar@ZrO2 composite	Methylene blue (MB)	Good reuse for 5 cycles	N/A	Multilayer adsorption fitted by Freundlich model; rGO/biochar/ZrO2 surface interactions	2025	173
Sulfonated carbon derived from fishbone	Methylene blue (MB)	High removal	N/A	Sulfonated carbon adsorption; electrostatic attraction and functional-group binding	2021	174
Sulfonated carbon derived from fishbone	Congo red (CR)	High removal	N/A	Sulfonated carbon adsorption; anionic dye interaction with carbon functional groups	2021	174
Solvothermally reduced graphene oxide (rGO-160)	Methylene blue (MB)	87.39 adsorption; 98.57 photodegradation	N/A	Adsorptive removal plus UV-C photocatalytic degradation; π–π/electrostatic interactions	2019	175
GO/rGO-modified natural clinoptilolite	Methylene blue (MB)	About 85 at pH > 10; pH dependent	N/A	Carbon oxidation-state effect; negatively charged sites interact with MB cations	2026	176
Carbon-derived black soldier fly larvae adsorbent	Methylene blue (MB)	High removal reported	N/A	Porous biomass-derived carbon adsorption; electrostatic attraction and surface functional groups	2025	177
Polypyrrole–polyethyleneimine (PPy–PEI) nano-adsorbent	Methylene blue (MB)	100	183.3	N-rich active sites; Langmuir monolayer adsorption; pseudo-second-order kinetics	2022	178
Cyclodextrin polymer magnetic adsorbent	Direct Red 83:1	High removal at optimum pH 5	32.0	Heterogeneous surface adsorption; pseudo-second-order/intraparticle diffusion	2022	179
Pristine chitosan powder	Methyl orange (MO)	98.15 first cycle; 81.43 by fifth	7.843	Chitosan amine/hydroxyl adsorption; electrostatic and H-bonding	2026	180
Pristine chitosan powder	Methylene blue (MB)	93.76 first cycle; 77.28 by fifth	7.605	Chitosan functional groups; electrostatic and H-bonding	2026	180
Acrylic-acid-grafted guggul gum hydrogel	Methylene blue (MB)	High	N/A	Hydrogel swelling; electrostatic interaction and diffusion into polymer network	2025	181
Cyanoguanidine-modified chitosan	Congo red (CR)	High	N/A	Electrostatic attraction between protonated amines and sulfonate groups	2021	182
Hydrogel nanocomposite adsorbent	Astrazon red	High	313.3	Adsorption on nanocomposite hydrogel surface; spontaneous process	2023	183

In addition to nanoparticle-based biosorbents, recent work has demonstrated the potential of lignin valorization for sustainable pollutant removal. A functionalized lignin adsorbent material was prepared through an in situ one-step strategy for simultaneous lignin recovery and functionalization from industrial black liquor using a sustainable organic acid–choline chloride system. Methylene blue (MB) was employed as a model cationic compound to investigate the adsorption behavior, and the functionalized lignin achieved a high adsorption capacity of 725.98 mg g⁻¹ across a broad pH range, substantially outperforming commercial kraft lignin. The adsorption data were well described by the Freundlich isotherm, indicating heterogeneous multilayer adsorption, while the study also demonstrated the practical application of the material for ammonia recovery from animal feedlot wastewater. Unlike the two Congo red studies, the main emphasis of this work was on sustainable material preparation, adsorption performance, and circular resource recovery rather than a detailed thermodynamic evaluation of dye adsorption. The ammonia-enriched lignin was subsequently reused as a slow-release fertilizer, demonstrating an integrated circular-economy strategy linking wastewater treatment, waste valorization, nutrient recovery, and agricultural reuse.¹⁴⁷

The adsorption of methylene blue varies considerably among carbonaceous materials, reflecting differences in surface area, surface chemistry, and molecular interactions. Activated carbon, graphene oxide, and carbon nanotubes exhibited distinct adsorption behaviors, with oxygen-containing functional groups and π–π interactions playing important roles in the interaction between the aromatic dye molecules and carbon surfaces. The comparison highlights that adsorption performance cannot be attributed solely to surface area; rather, the chemical nature and accessibility of surface sites also strongly influence dye uptake. This distinction is particularly important when comparing conventional activated carbon with nanoscale carbon materials, where enhanced surface accessibility and tunable surface chemistry can provide additional adsorption mechanisms.¹⁴⁸ Multi-walled carbon nanotubes (MWCNTs) exhibited a relatively high adsorption capacity toward methylene blue, reaching 95.30 mg g⁻¹ at 323 K. The favorable adsorption performance can be associated with the accessible carbon surface and the interactions between the aromatic dye molecules and the nanotube surface. Equilibrium data followed the Langmuir isotherm, suggesting predominant monolayer adsorption, while the pseudo-second-order kinetic model provided the best description of the adsorption rate, indicating a strong contribution from surface interactions. The increase in adsorption capacity at elevated temperature further suggests a temperature-dependent adsorption process; however, distinguishing the contributions of surface area, π–π interactions, and surface functional groups would be necessary to establish the dominant adsorption mechanism.¹⁴⁹

Modification of eucalyptus bark (EB) with CuFe₂O₄ significantly improved its adsorption performance toward basic fuchsin, with CuFe₂O₄/EB achieving the highest adsorption capacity (≈9.64 mg g⁻¹) among the investigated adsorbents. The superior performance of the magnetic composite suggests that CuFe₂O₄ incorporation enhanced the adsorption characteristics of the EB-based material. The adsorption data were best described by the Langmuir isotherm and pseudo-second-order kinetic models, indicating monolayer adsorption on relatively homogeneous sites and a substantial contribution of surface interactions to the adsorption process. However, the relatively modest adsorption capacity underscores the need to further optimize the surface chemistry and composite structure to enhance dye uptake.¹⁵⁰

Fig. 19 shows the effect of different operating parameters on basic fuchsin adsorption using EB, CuFe₂O₄/EB, and PPY/GO/EB. Fig. 19A and B show that pH strongly affected dye adsorption; the best removal occurred at pH 5 for EB, pH 9 for CuFe₂O₄/EB, and pH 10 for PPY/GO/EB, which is related to the pHpzc values and electrostatic attraction between the positively charged basic fuchsin dye and negatively charged adsorbent surfaces. Fig. 19C shows that adsorption capacity decreased with increasing adsorbent dose due to active site overlap and underutilization. Fig. 19D and E indicate that adsorption increased with dye concentration and contact time until equilibrium was reached at about 90 min. Fig. 19F shows that adsorption decreased with increasing temperature, confirming an exothermic process. Fig. 19G and H show that 120 rpm agitation speed and smaller particle size gave better adsorption, while Fig. 19I and J show that salts and surfactants reduced adsorption because of competition and shielding effects.

Fig. 19. Effect of various parameters on the degradation of BF by EB, CuFe₂O₄/EB (MAG/EB), and PPY/GO/EB: (a) pH; (b) point of zero charge (pHpzc); (c) adsorbent dosage; (d) initial dye concentration; (e) contact time; (f) temperature; (g) agitation speed; (h) adsorbent particle size; (i) salts; and (j) surfactants. Experimental conditions: initial BF concentration, 10 mg L⁻¹; adsorbent dosage, 0.05 g L⁻¹; agitation speed, 120 rpm; contact time, 90 min; and temperature, 303 K. Reproduced from ref. 150 (Rashid et al., 2025) under the terms of the Creative Commons Attribution-Non-Commercial 4.0 International License (CC BY-NC 4.0).

Phosphoric-acid activation of Spathodea campanulata stem produced a high-surface-area activated carbon (1054 m² g⁻¹) capable of removing 99.95% of methylene blue (MB) under optimized conditions. The high surface area provides extensive accessible adsorption sites and likely contributes substantially to the strong removal performance. The Freundlich model provided the best description of the equilibrium data, suggesting adsorption on a heterogeneous surface, while the pseudo-second-order model adequately described the kinetics. However, the exceptionally high removal percentage should be interpreted together with the experimental adsorbent dosage and initial dye concentration, since these parameters can strongly influence percentage removal. Importantly, the conversion of agricultural biomass into a high-surface-area adsorbent demonstrates a potentially attractive route for combining waste valorization with wastewater treatment.¹⁵¹

Fig. 20 illustrates the preparation steps of Spathodea campanulata-based activated carbon. The raw plant stem was first size-reduced to increase the contact area and make the material suitable for chemical treatment. It was then further ground and impregnated with phosphoric acid, which helps develop porosity and activate the carbon structure. After thermal activation by pyrolysis, the final black activated carbon was obtained. This figure clearly supports the conversion of low-cost plant biomass into a porous carbon adsorbent suitable for methylene blue adsorption.

Fig. 20. Preparation of activated carbon from Spathodea campanulata stem: (a) size reduction, (b) chemical impregnation, (c) final activated carbon after pyrolysis. Reproduced from ref. 151 (Dimbo et al., 2024) under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0).

Integrating graphene oxide (GO) with montmorillonite markedly enhanced the adsorption of malachite green oxalate from aqueous solution. The GO–montmorillonite composite achieved an adsorption capacity of approximately 675 mg g⁻¹, more than twice that of pristine GO (315 mg g⁻¹). This enhancement can be attributed to the introduction of additional surface functional groups and the greater availability of accessible adsorption sites provided by the clay component. The pronounced improvement demonstrates that hybridization can overcome some limitations of individual carbon-based adsorbents by combining complementary surface properties. However, the relative contributions of increased surface area, additional functional groups, and possible electrostatic or π–π interactions should be distinguished to establish the dominant mechanism responsible for the enhanced dye uptake.¹⁵²

Fig. 21a illustrates the proposed adsorption mechanism of malachite green oxalate dye onto GO and GO-based montmorillonite composites. The adsorption mainly occurs through electrostatic attraction between the positively charged dye molecules and the negatively charged surface of the adsorbents, especially under near-neutral to basic conditions. In addition, hydrogen bonding and π–π interactions also contribute to dye adsorption, which explains why adsorption can still occur even under acidic conditions.

Fig. 21. (a) Proposed mechanism for malachite green oxalate adsorption onto GO and GO-based montmorillonite composites through electrostatic interaction, hydrogen bonding, and π–π interactions, (b) proposed adsorption mechanism of methylene blue onto Mn/PB/CNT composite through electrostatic attraction, intra-particle diffusion, π–π stacking, and hydrogen bonding, (c) preparation procedure of H₃PO₄-activated chicken bone adsorbent for methylene blue dye removal. Reproduced from ref. 152 (Bhandari et al., 2024) under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0); reproduced from ref. 153 (Alsowayigh et al., 2026) with permission from the copyright holder, license no. 6282381498659; reproduced from ref. 154 (Muteeb et al., 2025) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0).

A manganese–Prussian blue–carbon nanotube (Mn/PB/CNT) nanocomposite combined several complementary adsorption features for methylene blue (MB) removal, achieving 91% removal under optimized conditions with a maximum adsorption capacity of 31.2 mg g⁻¹. The composite architecture integrates the porous structure of CNTs with the chemical stability of Prussian blue and the redox/coordination functionality of manganese, providing a broader range of potential interaction sites than the individual components. Nevertheless, the relatively modest adsorption capacity indicates that the principal benefit of the ternary composite may lie in its rapid and efficient removal rather than exceptionally high dye loading. The study therefore illustrates an important design principle: combining materials with different physicochemical functions can improve overall removal performance, but increasing compositional complexity does not necessarily translate into a proportionally higher adsorption capacity.¹⁵³

The study reported a maximum MB removal efficiency of 92.29%, with adsorption behavior best described by the Temkin isotherm and pseudo-second-order kinetic model, indicating a spontaneous and exothermic adsorption process.¹⁵⁴Fig. 21b illustrates the proposed adsorption mechanism of MB onto the Mn/PB/CNT composite. The mechanism involves several interactions, including electrostatic attraction between the negatively charged composite surface and cationic MB molecules, π–π stacking between the aromatic rings of MB and the graphitic CNT structure, and hydrogen bonding with surface functional groups. In addition, the porous CNT network supports intra-particle diffusion, while Mn sites enhance redox-based interaction and surface complexation.

Fig. 21c illustrates the preparation procedure of the activated chicken bone adsorbent. The process starts with collecting and cleaning chicken bone waste, followed by drying, pulverizing, and thermal carbonization in a muffle furnace to obtain bone charcoal. The material is then chemically activated using H₃PO₄, neutralized, ground, and sieved to obtain the final CBAAC adsorbent. This figure clearly supports the sustainable conversion of low-cost chicken bone waste into an activated adsorbent with improved surface properties for MB dye removal.

Magnetic alginate/γ-Fe₂O₃/functionalized CNT composite beads exhibited strongly pH-dependent methylene blue (MB) adsorption, with the uptake increasing from 37.1 to 82.7 mg g⁻¹ as the solution pH increased from 3 to 10. This pronounced pH dependence indicates that surface charge and the ionization state of surface functional groups play an important role in MB uptake. Although equilibrium was reached after 48 h, the composite achieved a high adsorption capacity, with the Sips model predicting 320.5 mg g⁻¹, which is in close agreement with the experimental value of 307.0 mg g⁻¹. The good agreement between the modeled and experimental capacities supports the suitability of the Sips model for describing the heterogeneous adsorption behavior of the composite. The pseudo-second-order kinetic fit (k₂ = 0.030 g mg⁻¹ min⁻¹) further suggests that surface interactions contribute substantially to the adsorption process. However, the relatively long equilibrium time may represent a practical limitation for wastewater treatment, emphasizing the need to balance high adsorption capacity with adsorption kinetics.¹⁵⁵ Functionalization of multi-walled carbon nanotubes (MWCNTs) with iron oxide nanoparticles and Citrus reticulata peel extract produced a highly efficient and rapidly acting adsorbent for cationic dyes. Methyl blue removal reached 93% within only 10 min, whereas methyl green removal reached 84% within 25 min, demonstrating faster uptake than many adsorption systems that require extended equilibration periods. The material also retained approximately 90% removal efficiency after seven reuse cycles, highlighting its potential for repeated wastewater treatment. The enhanced performance was attributed to electrostatic attraction between negatively charged surface groups and cationic dye molecules, complemented by π–π interactions between the aromatic dye rings and the MWCNT framework. Nevertheless, removal efficiency alone cannot establish the superiority of this system over other adsorbents; adsorption capacity, adsorbent dosage, initial dye concentration, and regeneration efficiency should also be considered when assessing its practical performance.¹⁵⁶ Eco-friendly reduced graphene oxide (rGO) exhibited a substantial adsorption capacity toward methylene blue (MB), reaching 121.95 mg g⁻¹, with adsorption increasing as the solution pH rose to approximately 9. This pH dependence highlights the importance of surface charge and functional-group ionization in governing the interaction between rGO and the cationic dye. The adsorption equilibrium was better described by the Langmuir model, suggesting predominant monolayer uptake, while the pseudo-second-order model provided the best kinetic fit, indicating a significant contribution of surface interactions to the adsorption rate. The enhanced MB uptake was attributed to the combined effects of hydrogen bonding, electrostatic attraction, and π–π interactions with the rGO surface. Nevertheless, the moderate adsorption capacity compared with some graphene-based composites suggests that optimizing surface functionality and accessible adsorption sites could further improve the performance of rGO for dye removal.³⁰ Graphene oxide–chitosan composite granules demonstrated exceptionally high adsorption performance toward methylene blue (MB), reaching a maximum capacity of 951.35 mg g⁻¹, substantially exceeding the capacities reported for many conventional adsorbents. The markedly enhanced uptake can be associated with the complementary surface characteristics of graphene oxide and chitosan, which provide abundant and chemically diverse adsorption sites. The Freundlich isotherm provided the best equilibrium description, indicating heterogeneous surface interactions and multilayer adsorption, while the pseudo-second-order model described the adsorption kinetics. Intraparticle diffusion played an important role during the first 4 h, suggesting that mass transfer within the porous structure contributes significantly to the overall adsorption process. Importantly, the composite retained its adsorption performance over six adsorption–desorption cycles, demonstrating promising regenerability. However, the relatively long diffusion period and the need to assess performance under realistic wastewater conditions should be considered when evaluating its practical applicability.¹⁵⁷ Chitosan/polyacrylamide/graphene oxide (GO) composite membranes exhibited their optimum methylene blue (MB) adsorption performance at a GO loading of 20 wt%, achieving a Langmuir-derived monolayer capacity of 510.2 mg g⁻¹ at 20 °C. The enhanced uptake can be attributed to the combined contribution of chitosan and GO, which provides chemically active sites while the polyacrylamide component improves the structural integrity of the composite membrane. Despite the relatively high adsorption capacity, equilibrium required approximately 2800 min, indicating slow adsorption kinetics that may limit practical application in wastewater treatment. This contrast between high adsorption capacity and prolonged equilibration highlights the importance of considering both adsorption efficiency and mass-transfer kinetics when evaluating GO-based composite adsorbents.¹⁵⁸ Crosslinked O-carboxymethyl chitosan hydrogel exhibited enhanced methylene blue (MB) adsorption with increasing pH and temperature, highlighting the importance of surface ionization and temperature-dependent interactions in governing dye uptake. The adsorption kinetics were exceptionally well described by the pseudo-second-order model (R² = 0.998–1.000), while the Langmuir model showed strong applicability to the equilibrium data. The close agreement between the calculated and experimental adsorption capacities (294.1 and 290.6 mg g⁻¹, respectively) further supports the reliability of the equilibrium analysis. However, the interpretation of pseudo-second-order kinetics should not, by itself, be taken as definitive evidence of chemisorption, as kinetic model fitting alone cannot establish the underlying adsorption mechanism.¹⁵⁹ Magnetic chitosan/graphene oxide/copper ferrite nanocomposite exhibited dye-dependent adsorption behavior, with the optimum pH shifting from 3 for indigo carmine to 7 for safranin O. The corresponding removal efficiencies reached 87.86% and 88.13%, while the Langmuir model yielded maximum adsorption capacities of 110.42 and 66.5 mg g⁻¹, respectively. The marked difference in adsorption capacity between the two dyes despite their comparable removal efficiencies highlights the strong influence of dye structure and surface–adsorbate interactions on adsorption performance. Electrostatic attraction, hydrogen bonding, and π–π interactions collectively contributed to dye uptake, while the pseudo-second-order model provided a good kinetic fit (R² ≈ 0.99). Nevertheless, the relatively moderate capacities compared with some graphene/chitosan-based systems suggest that optimizing surface functionality and charge characteristics could be important for improving adsorption toward different dye classes.¹⁶⁰

A dual-network chitosan-based hydrogel (CMAPP) demonstrated broad-spectrum removal of several cationic dyes, achieving removal efficiencies of 97.62% for methylene blue, 96.67% for malachite green, 98.12% for crystal violet, and 99.32% for basic fuchsin. The consistently high removal across structurally different cationic dyes suggests that the hydrogel provides a versatile adsorption environment rather than a highly dye-specific interaction. The proposed mechanism involves electrostatic attraction, hydrogen bonding, and n–π interactions, reflecting the contribution of multiple functional groups within the chitosan-based network. Importantly, the material retained its adsorption performance over five adsorption–desorption cycles, indicating useful regeneration stability. Nevertheless, the reported removal percentages should be considered alongside adsorption capacities and adsorbent dosages, because high percentage removal alone does not necessarily indicate high adsorption capacity.¹⁶¹

Fig. 22 discusses the adsorption kinetics, interference resistance, recyclability, and selectivity of CMAPP. Fig. 22a and b show that the adsorption process was better fitted by the pseudo-second-order kinetic model, indicating that chemisorption plays an important role in dye adsorption. Fig. 22c presents the intraparticle diffusion model, showing that adsorption occurs through several stages, including surface adsorption, diffusion into hydrogel pores, and equilibrium. Fig. 22d shows that interfering ions reduced dye adsorption because they competed with dye molecules for active sites, while Fig. 22e confirms that CMAPP maintained good removal efficiency after five cycles, proving its reusability. Fig. 22f shows weak adsorption toward anionic dyes, confirming that CMAPP is more suitable for cationic dye removal.

Fig. 22. (a) PFO, (b) PSO, and (c) IPD kinetic models for CMAPP; (d) effect of interfering ions on the adsorption performance of CMAPP; (e) recycling performance of CMAPP; and (f) adsorption performance of CMAPP toward anions. Reproduced from ref. 161 (Miao et al., 2024) with permission from Elsevier, license no. 6272710493198.

A separable pullulan-based nanocomposite hydrogel incorporating PDA-modified ZIF-8 (ZPPs) provided a high-capacity platform for removing both cationic malachite green (MG) and anionic Congo red (CR), with adsorption capacities of 615.4 and 525.8 mg g⁻¹, respectively. The comparable high uptake toward dyes with opposite charges suggests that the composite benefits from multiple adsorption pathways rather than relying exclusively on electrostatic attraction. The combination of PDA functional groups, the porous ZIF-8 framework, and the hydrophilic pullulan network likely creates a high density of accessible sites and facilitates mass transfer. More importantly, the separable hydrogel architecture and stable performance over five reuse cycles address two practical limitations of powdered MOFs: difficult solid–liquid separation and loss of adsorption efficiency during regeneration. Thus, the main advantage of this system is not simply its high q_max, but the integration of high adsorption capacity, broad dye compatibility, separability, and recyclability within a single composite.¹²⁶

Fig. 23 summarizes the structural and adsorption-related properties of ZIF-8, ZIF-8@PDA, and the pullulan composite hydrogels. Fig. 23a and b show that PDA modification improved the dye adsorption capacity of ZIF-8, especially for MG and CR, due to enhanced surface area and additional interaction sites. Fig. 23c confirms that ZIF-8@PDA has a higher BET surface area and pore volume than pristine ZIF-8, supporting better dye uptake. Fig. 23d shows that the characteristic XRD peaks of ZIF-8 remained after PDA coating and hydrogel incorporation, confirming structural stability. Fig. 23e and f indicate that adding ZIF-8@PDA reduced swelling and water retention because the gel network became more rigid. Fig. 23g and h show that ZPP10 had the best dye adsorption performance, while Fig. 23i confirms improved mechanical strength compared with pure pullulan hydrogel.

Fig. 23. (a) Dye adsorption capacity of ZIF-8; (b) dye adsorption capacity and (c) N₂ adsorption–desorption isotherms of ZIF-8 and ZIF-8@PDA; (d) XRD patterns of ZPP0, ZPP10, ZIF-8, and ZIF-8@PDA; (e) swelling ratio, (f) water retention, (g) dye adsorption capacity, and (i) stress–strain curves of ZPP0 and ZPP10; and (h) dye adsorption capacity of ZIF-8@PDA incorporated into the hydrogels. Reproduced from ref. 126 (Su et al., 2024) with permission from Elsevier, license no. 6272710729891.

Encapsulating polyaniline (PANI) on quartz sand produced a low-cost QS@PANI composite with improved Orange G adsorption compared with either quartz sand or PANI used independently. The enhanced performance indicates a synergistic effect between the inert mineral support and the functional polymer layer, with electrostatic attraction, π–π interactions, and hydrogen bonding contributing to dye uptake. The use of quartz sand is particularly relevant from a practical perspective because it provides an inexpensive and readily available support while potentially improving the handling of the polymeric adsorbent. In addition, the reported recyclability over repeated cycles strengthens the practical potential of the composite. However, the absence of a reported adsorption capacity in the available results limits direct comparison with high-capacity adsorbents, so its main advantage should be considered low cost, composite synergy, and reusability rather than demonstrated superior adsorption capacity.¹⁶²

Fig. 24 shows the effects of adsorbent dose and solution pH on Orange G adsorption. In Fig. 24a and b, increasing the adsorbent dose enhanced the removal efficiency because more active sites became available; however, the adsorption capacity decreased at higher doses due to overlapping and overcrowding of adsorption sites. QS@PANI showed the best performance compared with QS and PANI, reaching about 94.14% removal and 78.45 mg g⁻¹ adsorption capacity at 1.2 g L⁻¹. Fig. 24c shows that QS@PANI maintained high Orange G removal from acidic to weakly alkaline pH, while removal decreased at strong alkaline pH because OH⁻ ions competed with dye molecules. Fig. 24d shows that the pHpzc of QS@PANI is 6.6, meaning the surface is positively charged below this value and can strongly attract anionic Orange G molecules through electrostatic interaction.

Fig. 24. (a and b) Effects of adsorbent dose, (c) Effect of solution pH on the adsorption of OG, (d) pH of point zero charge. Reproduced from ref. 162 (Amjlef et al., 2023) with permission from Elsevier, license no. 6272710623382.

7. Dye adsorption for wastewater treatment

Dye adsorption is a highly important approach for treating wastewater and industrial effluents, particularly from textile, leather, printing, cosmetic, and dye-manufacturing industries. Large quantities of synthetic dyes are discharged with industrial wastewater, and many of these dyes are chemically stable, resistant to biodegradation, and difficult to remove using conventional biological treatment alone. Untreated dye-containing effluents reduce light penetration into water bodies, inhibit photosynthetic activity, increase chemical and biochemical oxygen demand, and negatively affect aquatic organisms. In addition, several dyes and their degradation products may exhibit toxic, mutagenic, teratogenic, allergenic, or carcinogenic effects, posing serious environmental and human health concerns. Therefore, efficient removal of dyes before wastewater discharge is essential. Among different treatment methods, adsorption is considered one of the most effective and practical techniques because it is simple, cost-effective, highly efficient, flexible, and applicable to a wide range of dye pollutants. It also offers advantages such as low sludge formation, potential for adsorbent regeneration, and suitability as a polishing step in industrial wastewater treatment. The adsorption process depends on several factors, including adsorbent type, surface area, pore structure, functional groups, solution pH, adsorbent dose, initial dye concentration, contact time, and temperature. Therefore, the development of low-cost, reusable, and environmentally safe adsorbents is essential to improve dye removal efficiency and support sustainable wastewater management.

8. Economic feasibility and industrial scale-up of nanomaterial-based adsorbents

The economic feasibility of adsorption-based wastewater treatment should be evaluated beyond adsorption capacity and removal efficiency, particularly when advanced or nanostructured adsorbents are considered for large-scale implementation. Recent techno-economic studies demonstrate that both capital and operating expenditures can substantially influence the overall feasibility of advanced treatment systems. For example, Saad et al. performed a techno-economic assessment of an electrocoagulation–nanofiber membrane system theoretically scaled to treat 1000 m³ d⁻¹ of dye wastewater, incorporating equipment, installation, utilities, membrane requirements, energy consumption, electrode consumption, labor, and maintenance into the CAPEX and OPEX calculations.¹⁸⁴ The integrated system achieved 96% dye removal with an estimated CAPEX of approximately USD 437 025 and an annual OPEX of USD 42 857, whereas the conventional photocatalytic process required approximately USD 399 075 in CAPEX and USD 94 265 in annual OPEX. Over a 15 year plant lifetime, the standalone electrocoagulation process was the least expensive option, with an estimated total cost of USD 325 820, highlighting that improved treatment performance does not necessarily translate into superior economic feasibility when additional membrane or nanomaterial-related components are introduced. Similarly, Dadebo et al. demonstrated that techno-economic feasibility should account for material consumption, energy requirements, maintenance, labor, waste management, and potential revenue from resource recovery. For a treatment capacity of 50 m³ d⁻¹, their integrated bio-digestion/phytoremediation/co-pyrolysis system had a total operating expenditure of approximately USD 14 551 per year (USD 0.83 per m³), while valorization of biochar, energy recovery, and carbon credits generated revenues of up to USD 29 931 per year (USD 1.71 per m³), resulting in a net annual profit of USD 15 380, an NPV of USD 70 358, and a payback period of 5.43 years.¹⁸⁵ These findings emphasize that claims of cost-effectiveness for emerging adsorbents should consider not only the intrinsic adsorption performance but also raw-material costs, synthesis and processing energy, reactor and separation requirements, regeneration or replacement frequency, maintenance, and opportunities for material recovery or reuse. Importantly, the transition of exotic nanomaterials from milligram- or laboratory-scale synthesis to continuous-flow treatment requires validation of production yield, process reproducibility, material recovery, hydraulic performance, and long-term stability at substantially larger throughput. Thus, future studies should integrate adsorption performance with CAPEX, OPEX, cost per unit mass of pollutant removed, regeneration cycles, energy demand, and continuous-flow scale-up metrics to provide a more realistic assessment of the commercial potential of advanced nanomaterial-based adsorption technologies.

9. Future perspectives and a research roadmap for dye adsorption

The next phase of dye-adsorption research should move beyond demonstrating high adsorption capacities under idealized laboratory conditions and instead focus on establishing whether proposed adsorbents can function reliably, economically, and sustainably in real wastewater-treatment systems. The field has generated a large number of promising materials, yet the gap between laboratory performance and practical deployment remains substantial. Future research should therefore prioritize realistic wastewater conditions, continuous-flow operation, regeneration, environmental safety, scalability, techno-economic feasibility, and life-cycle sustainability, rather than the continued synthesis of increasingly complex materials with only marginal improvements in adsorption capacity.

The field should move away from single-dye batch experiments as the primary evidence of adsorbent performance. Batch experiments remain useful for understanding fundamental adsorption mechanisms, but future studies should prioritize dynamic fixed-bed columns and continuous-flow systems, reporting breakthrough curves, bed capacity, empty-bed contact time, pressure drop, mass-transfer behavior, and long-term stability. Adsorbents should also be tested in chemically complex wastewater containing multiple dyes, salts, surfactants, organic matter, and heavy metals. Researchers should quantify the loss of adsorption capacity caused by competitive adsorption and high ionic strength to distinguish materials that perform well in idealized solutions from those that remain effective under realistic conditions.

Regeneration and adsorbent lifetime should also be treated as core performance indicators. Future studies should report repeated adsorption–desorption cycles, capacity retention, regeneration efficiency, structural stability, regenerant consumption, and the fate of concentrated regeneration waste. In parallel, the environmental safety of nanomaterials should be evaluated through leaching, particle release, degradation, and ecotoxicity assessments, particularly after repeated use. An adsorbent should not be considered sustainable solely because it achieves high dye removal; its potential for secondary contamination must also be considered.

Techno-economic analysis (TEA) and life-cycle assessment (LCA) should become integral components of the development of advanced nanoadsorbents. TEA should consider precursor and reagent costs, energy and water consumption, synthesis complexity, activation or functionalization, separation, regeneration, transportation, and end-of-life management. Recent work on biomass-derived materials demonstrates the value of integrating TEA with process-scale assessment to evaluate CAPEX, OPEX, profitability, and the transition from laboratory production to industrial-scale operation. Similarly, recent LCA research has shown that the environmental benefits of bio-derived materials cannot simply be assumed from the use of agricultural waste; solvent selection, extraction, purification, solvent recovery, and energy consumption can represent important environmental hotspots. Therefore, future nanoadsorbents should increasingly be designed from renewable or waste-derived feedstocks using green solvents, simplified synthesis routes, low-energy processing, and recyclable or biodegradable components, with cradle-to-grave LCA used to substantiate claims of sustainability.^186,187

Future material design should consequently prioritize simplicity, scalability, reproducibility, and environmental compatibility over unnecessary structural complexity. Researchers should report material yield, mass balance, energy demand, chemical consumption, and batch-to-batch reproducibility. Promising materials should then progress through a clear technology-readiness pathway from laboratory studies to continuous-flow, pilot-scale, and demonstration-scale systems. Standardized benchmarking against commercial or established adsorbents under identical conditions is also needed to enable meaningful comparison among materials.

Finally, adsorption should increasingly be considered as one component of an integrated treatment train rather than a stand-alone technology. Hybrid systems combining adsorption with photocatalysis, advanced oxidation, membrane separation, biological treatment, or electrochemical processes should be evaluated based on overall pollutant removal, mineralization, energy demand, regeneration requirements, environmental impacts, and treatment cost. Such a systems-level approach will be essential for converting high-performing nanoadsorbents from laboratory-scale materials into genuinely sustainable and industrially deployable technologies for dye-contaminated wastewater treatment.

10. Artificial intelligence and machine learning in adsorbent design

Artificial intelligence (AI) and machine learning (ML) are emerging as valuable tools for accelerating the design and optimization of advanced adsorbents by establishing predictive relationships between material properties, operating conditions, and adsorption performance. Recent studies have demonstrated that ML algorithms can capture complex and nonlinear structure–property–performance relationships that are difficult to resolve using conventional experimental approaches. For example, ANN, random forest, and gradient boosting models have been used to predict adsorption capacity from physicochemical descriptors and identify the parameters that most strongly control pollutant uptake. In magnetic clay–Fe₃O₄ composites, ML models achieved high predictive accuracy (R² ≈ 0.91), while explainability analyses identified zeta potential, isoelectric point, hydrodynamic particle size, and adsorption kinetics as important predictors, highlighting that surface chemistry may be more influential than surface area alone.¹⁸⁸ Similarly, the integration of design of experiments with ML has been shown to optimize adsorption/separation conditions while maximizing removal efficiency or minimizing adsorbent dosage.¹⁸⁹ These developments indicate that AI/ML can reduce experimental trial-and-error, guide the rational selection of material properties, and support faster screening of candidate nanoadsorbents. However, their broader application remains dependent on the availability of high-quality, sufficiently large, and standardized datasets, as well as transparent model validation and interpretability. Future research should therefore integrate experimental adsorption databases with explainable AI, multi-objective optimization, and mechanistic understanding to enable reliable prediction of adsorbent performance and accelerate the development of scalable nanomaterials for real wastewater treatment.

11. Limitations and drawbacks of nanomaterials for dye adsorption

Although nanomaterials have attracted considerable attention as highly efficient adsorbents for dye removal because of their large specific surface area, abundant active sites, tunable surface chemistry, and high adsorption capacity, their practical application is associated with several limitations and potential environmental concerns. Importantly, the excellent adsorption performance of many nanoadsorbents reported under controlled laboratory conditions does not necessarily translate to real industrial wastewater, where competing ions, salts, surfactants, natural organic matter, heavy metals, and mixtures of different dyes can interfere with adsorption and reduce the availability of active sites. In addition, the synthesis of advanced nanomaterials may involve expensive precursors, toxic solvents or reagents, sophisticated preparation procedures, and considerable energy consumption, which can increase their overall environmental footprint and limit their economic feasibility at large scale.

Another important challenge is the recovery and separation of nanomaterials after wastewater treatment. Owing to their very small particle size, powdered nanoadsorbents can be difficult to separate completely from treated water, raising concerns regarding nanoparticle release into aquatic environments. The potential toxicity, persistence, mobility, and bioaccumulation of certain nanomaterials should therefore be carefully considered, particularly when materials contain potentially hazardous metals or when nanoparticles can undergo dissolution, transformation, or leaching during treatment. Mechanical and chemical stresses, repeated regeneration cycles, changes in pH and ionic strength, and prolonged contact with water may promote nanoparticle detachment or the release of potentially toxic metal ions, thereby creating a possible secondary source of aquatic contamination. Furthermore, nanomaterials may undergo aggregation, surface fouling, structural changes, or loss of active sites during repeated adsorption–desorption cycles, resulting in reduced adsorption capacity and limited long-term reusability.

Regeneration is another major limitation. Although regeneration can reduce the amount of solid waste generated and improve the economic viability of nanoadsorbents, the process may require substantial amounts of chemical regenerants, solvents, or energy and can generate secondary wastewater containing concentrated dye molecules. Moreover, adsorption itself generally transfers dye molecules from the aqueous phase to the nanomaterial surface rather than destroying or mineralizing them. Consequently, improper disposal or inadequate regeneration of spent nanoadsorbents may result in secondary contamination. The environmental fate and long-term ecotoxicological effects of spent nanoadsorbents also require greater attention, as accumulated dyes, released nanoparticles, or dissolved metal ions may interact with aquatic organisms after disposal or accidental release. The performance of nanoadsorbents is also strongly dependent on operational parameters, including pH, temperature, contact time, adsorbent dosage, initial dye concentration, and ionic strength, making optimization necessary for different wastewater matrices.

Therefore, the assessment of nanomaterials for dye adsorption should extend beyond adsorption capacity and removal percentage. Their synthesis cost, energy and resource requirements, environmental footprint, toxicity, stability, recovery efficiency, regeneration performance, recyclability, and behavior in real wastewater should also be evaluated. Future research should adopt safe-by-design principles by considering material composition, nanoparticle immobilization, structural stability, resistance to leaching, and end-of-life management during the development of nanoadsorbents. In parallel, systematic studies should evaluate nanoparticle and metal-ion leaching under realistic mechanical and chemical conditions, as well as the long-term ecotoxicity of both fresh and spent adsorbents toward representative aquatic organisms. Such assessments, together with life-cycle and techno-economic analyses, are essential to determine whether the laboratory-scale advantages of nanoadsorbents can be translated into environmentally safe, cost-effective, and scalable industrial applications.

12. Conclusion

Dye-contaminated wastewater is a major environmental concern due to stability, low biodegradability, and potential toxicity of synthetic dyes. Adsorption is an effective and practical treatment method because it is simple, efficient, and suitable for different dye classes. This review highlights the use of various adsorbents, including metal oxide nanoparticles, metallic nanoparticles, spinel oxides, metal sulfides, carbon-based materials such as CNTs, GO, rGO, activated carbon, and biochar, as well as polymeric and biopolymer-based materials such as chitosan composites. These materials remove dyes mainly through electrostatic attraction, hydrogen bonding, π–π interactions, ion exchange, surface complexation, and pore filling. Their performance depends on pH, dye concentration, adsorbent dose, contact time, temperature, surface area, pore structure, and functional groups. However, challenges such as high cost, difficult separation, poor regeneration, reduced efficiency in real wastewater, and secondary pollution still limit large-scale applications. Future research should focus on developing low-cost, reusable, stable, eco-friendly, and scalable adsorbents, with more attention to real wastewater testing, regeneration, toxicity assessment, cost analysis, and pilot-scale studies. Overall, adsorption remains a promising strategy for the sustainable treatment of dye-contaminated wastewater.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.