Introduction

Clean water is needed all over the world for the health and well-being of living things (Tortajada and Biswas 2018). Clean water scarcity has become a serious problem in most developing, as well as some developed countries (Falkenmark 2022). This is due to rapid technological growth, industrial effluent pollution, a fast increase in human population, and irregular rainfall, affecting water quality (Bello et al. 2018). Regular use of modern facilities and products in our day-to-day activities, such as petrol in cars, plastics, textile clothing, food products, and drugs, just to mention a few, have generated large amounts of waste products, which have significantly contaminated pristine water bodies (Ahmad and Danish 2018; Ogbu et al. 2019). The use of polluted waters has resulted in severe adverse effects on plants, humans, and living organisms in general (Ezemonye et al. 2019; Lellis et al. 2019). Groundwater which is the most reliable source of clean water for use by the local population in developing countries is rapidly being polluted from the agricultural application of pesticides and fertilizers, as well as from organic and inorganic effluent discharge from industries (Edokpayi et al. 2018; Kiwaan et al. 2020; Asiwaju-Bello et al. 2020). In recent years, toxic pollutants were found to be present in unacceptable quantities in water in some developing countries, which poses a serious health threat to the local population (Ahmad and Danish 2018; Imran et al. 2020). Previously, when pollutant discharge to the environment was still low, contaminated waters were easily remediated by the adsorption process of soil, allowing clean water to penetrate, filter through, and get accumulated in underground water sources. Recently, due to the sharp increase in the release and concentration of pollutants, the soil uptake capacities have become saturated, thus also allowing the penetration of pollutants to underground sources.

Heavy metals are one of the most dangerous contaminants damaging the environment (Vardhan et al. 2019). Metals are natural elements that have been mined from the soil and used for human industry and products for millennia, making them a prominent source of worldwide distributed pollution. The massive growth in heavy metals use over the last few decades has undoubtedly resulted in an increase in metallic substance flux in the aquatic and terrestrial environment (Gautam et al. 2016). Due to their toxicity, longevity in the environment, non-biodegradability, and accumulation in the food chain, heavy metal contamination in aqueous media and industrial effluents is a serious ecological hazard (Chukwuemeka-Okorie et al. 2018; Mitra et al. 2022). At certain amounts, most heavy metals are carcinogenic. Zinc, copper, nickel, mercury, cadmium, lead, cobalt, arsenic, and chromium are toxic heavy metals of particular concern in the environment (Uddin 2017; Chakraborty et al. 2022). Similarly, the rising usage of radioactive materials in nuclear power plants, nuclear medicine, research, industry, and agriculture has raised the likelihood of radioactive water pollution (Yu et al. 2015). There have been various nuclear incidents in the past that have resulted in direct radioactive poisoning of water. These include the Fukushima Daiichi nuclear power plant accident in Japan in 2011, the Chernobyl disaster in Russia in 1986, the Three Mile Island nuclear power plant accident in the United States in 1979, the Kyshtym nuclear power plant explosion in Russia in 1975, and the K-19 nuclear power plant explosion in the North Atlantic in 1961 (Wang and Zhuang 2019; Kumar et al. 2020). The pollution of environmental waters with radionuclides such as uranium, thorium, selenium, lanthanum, cerium, ruthenium, and vanadium can also be associated with some toxic effects at a certain concentration (Gendy et al. 2021; Hassan et al. 2022; Akl 2022; Niu et al. 2022; Ji and Zhang 2022). The sources, harmful effects, and the United States Environmental Protection Agency’s (USEPA) permissible limits of heavy metals and radionuclides in drinking water are presented in Table 1.

Table 1 The US Environmental Protection Agency’s maximum permissible limit, sources, and toxicity of heavy metals and radionuclides
Full size table

As a result of the hazardous effects associated with heavy metal and radionuclide pollution of environmental waters, numerous strategies for treating contaminated waters have been developed. Electrochemical treatment, forward osmosis, flotation, reverse osmosis, lime softening, ion exchange, filtration, adsorption, solvent extraction, electrodialysis, photocatalysis, precipitation, and coagulation/flocculation are just a few of the processes available (Ibeji et al. 2020; Qasem et al. 2021; Karimi et al. 2022; Xu et al. 2022b). Most of these systems, however, cannot be deployed on a large scale because of high capital costs, hazardous intermediate production, and the inability to regenerate and self-clean (Crini and Lichtfouse 2019; Ghamry and Abdelmonem 2022). Adsorption, on the other hand, is an efficient and cost-effective method for removing heavy metals and radionuclides from water at very low concentrations due to the low risk of secondary contamination, cheap cost, simplicity, ease of operation, and simple adsorbent regeneration (Lee and Shin 2021; Liu et al. 2022; Xu et al. 2022a; Ankrah et al. 2022). As a result, adsorbents like agricultural waste, biochar, activated carbon, clay, polymer, nanoparticles, metal–organic framework, graphene, chitosan, and zeolite have been used to treat heavy metal and radionuclide-contaminated water (Yu et al. 2015; Renu et al. 2017; Gupta et al. 2021; Xu et al. 2022b).

Nanomaterials have recently acquired popularity as potential adsorbents for pollutant removal in wastewater treatment. This is because nanostructured adsorbents have substantially greater efficiencies and faster adsorption rates in water treatment than traditional materials, owing to their high surface area (Sadegh et al. 2017; Tee et al. 2022). Zinc oxide nanoparticles (ZnONPs) in particular have attracted a lot of attention among nanoparticle adsorbents because of their biocompatibility, affordable price, long-term stability, surface characteristics, photocatalytic activity, nontoxicity, and powerful antibacterial activity against microbes often found in water (Akbar et al. 2019; Gu et al. 2020; Akpomie et al. 2021). As a result, various studies on the adsorption of heavy metals and radionuclides onto ZnONPs have been conducted (Kumar et al. 2013; Kaynar et al. 2014; Azizi et al. 2017; Lagashetty et al. 2020; Gu et al. 2020; Alqahtany and Khalil 2021; Davarnejad and Nikandam 2022). Due to the importance of ZnONPs in water treatment, a review of their capability as a water decontaminating agent via adsorption and photocatalysis was written (Bharti et al. 2022). Another review recorded the synthetic parameters influencing the characteristics of ZnONPs and their use in wastewater treatment (Shaba et al. 2021). Likewise, the synthesis and characterization of ZnONPs were also documented (Agarwal et al. 2017; Rl et al. 2019). However, the existing reviews lacked information on the isotherms, kinetics, and thermodynamics of adsorption onto ZnONPs, which is essential for a thorough knowledge of any adsorption process. Only a basic sectional description of heavy metal polluted water treatment was provided in the reviews. This review addresses this shortcoming by offering valuable insight into these model interpretations as they pertain to heavy metal and radioactive adsorption onto ZnO nanoparticles. The equilibrium adsorption capacities obtained for the adsorption of heavy metals and radionuclides under different experimental conditions were examined. The isotherms, kinetics and thermodynamics were evaluated in addition to the regeneration and reuse of ZnONPs. Moreover, the mechanism of adsorption of the heavy metals and radionuclides onto ZnONPs was also considered.

Adsorption capacity of ZnONPs

The adsorption of heavy metals and radionuclides on ZnONPs can be expressed in terms of the adsorption capacity. Unlike the percentage removal, which is a representation of the adsorbate (heavy metals and radionuclides) removed from the solution, the adsorption capacity is a characteristic of the adsorbent (ZnONPs). The percentage removal only expresses the amount of adsorbate removed from the solution at equilibrium but does not give an adequate representation of the amount of adsorbate present on the adsorbent (Gu et al. 2020; Rezaei-Aghdam et al. 2021). Thus, the affinity of an adsorbent for different adsorbates is effectively compared by considering the adsorption capacity of the adsorbent for the pollutants in solution. Moreover, certain experimental factors such as solution pH, contact time, adsorbent dose, temperature, and adsorbate concentration could influence the adsorption capacity of an adsorbent material (Hegazy et al. 2021; Yadav and Dasgupta 2022). Therefore, such experimental conditions must also be provided alongside the adsorption capacity for a holistic and reliable comparison. Several studies on the adsorption of heavy metals and radionuclides onto ZnONPs presented a wide range of adsorption capacities for the various pollutants. For example, in 2015, the adsorption of thorium (IV) onto ZnONPs was performed and a high adsorption capacity of 1500 mg/g was obtained, which indicated the potency of ZnONPs for the decontamination of thorium-contaminated water (Kaynar et al. 2015). In 2017, the green synthesis of ZnONPs using zerumbone was conducted and applied for the adsorption of lead (II) ions as shown in Fig. 1. The zerumbone-mediated green synthesized ZnONPs were found to be efficient in the uptake of Pb(II) with an adsorption capacity of 15.65 mg/g at 300 K and pH 5.0 (Azizi et al. 2017). The following year, a much higher adsorption capacity of 434.8 mg/g was obtained for commercially ZnONPs for Pb(II) ions at pH 6.5 and a temperature of 298 K (Yin et al. 2018). This indicates that the method of preparation or source of the nanoparticle as well as the experimental conditions could significantly influence the adsorption capacity for a particular pollutant. In another report, solvothermal synthesized ZnONPs exhibited adsorption capacities of 5.084 mg/g, 2.248 mg,/g, and 1.761 mg/g for Cu(II), Cd(II) and Ni(II) ions, respectively, after appropriate conversions from mmol/g (Wang et al. 2018). The enhanced adsorption ability of ZnONPs toward Cu(II) revealed a relatively stronger selective adsorption. The binding force between heavy metal ions and ZnONPs accounts for such selectivity. The adsorption of heavy metal ions is based on the electronegativity of the metal ions and the surface hydroxyl groups of ZnONPs (Wang et al. 2013). Cu(II), Ni(II), and Cd(II) have electronegativities of 2.00, 1.91, and 1.69, respectively, resulting in the strongest copper-OH bond among these three heavy metals (Wang et al. 2018). The hydrolyzable characteristics of heavy metals also alter the affinity between the heavy metals and the surface of ZnONPs. Metal ions that are easily hydrolyzed have a strong affinity for the surface (Mustafa et al. 2002). The obtained adsorption capacities followed the order Cu(II) > Ni(II) > Cd(II), which corresponds to the electronegativity order. The induced selectivity of the hydroxyl group of ZnONPs for the metal ions was verified by the attenuated total reflection-Fourier transform infrared (ATR-FTIR) before and after adsorption as shown in Fig. 2. The ATR-FTIR hydroxyl shifts at 3400 cm−1 and the increased metal–oxygen absorptions supported the higher uptake of Cu(II) ions onto ZnONPs. Recently, hydrothermal synthesized ZnONPs were prepared and characterized with the X-ray diffraction (XRD) as shown in Fig. 3. The hydrothermal synthesized ZnONPs revealed crystalline phases corresponding to the wurtzite hexagonal form of ZnONPs with an adsorption capacity of 64.6 mg/g for Ba(II) ions (Abdulkhair et al. 2021). The result showed the potency of ZnONPs in the treatment of barium-polluted water via adsorption technique.

Fig. 1
figure 1

The schematic representation of the green synthesis of zinc oxide nanoparticles using zerumbone and its application in the adsorption of Pb(II) ions from solution (Azizi et al. 2017). Zerumbone crystals were dissolved in 100 mL ethanol at room temperature with gentle stirring. After complete dissolution, 2.19 g of zinc acetate dihydrate was added to the zerumbone solution to react for 2 h at 70 °C with constant magnetic stirring. The white solid was recovered by centrifugation at 8000 rpm for 15 min, washed with ethanol to remove excess zerumbone, and dried for 2 h at 100 °C. The obtained ZnONPs were used for Pb(II) ions adsorption and showed the occurrence of a chelating mechanism via the zerumbone moiety

Full size image
Fig. 2
figure 2

The attenuated total reflection-Fourier transform infrared spectra of a ZnONPs before adsorption and after adsorption of b Ni(II) c Cd(II) and d Cu(II) ions from solution. Reproduced from (Wang et al. 2018) with permission from Elsevier. The peaks of hydroxyl groups (3400 cm−1) migrated to lower wavenumbers after the adsorption of Ni, Cd, and Cu ions, showing that the adsorption might be attributed to the interaction between the metal ions and hydroxyl groups. Meanwhile, when compared to pure ZnONPs, the absorption peaks of metal-O located around 546 cm−1 of the Zn–O stretching enhanced after adsorption, suggesting that metal-O bonding was created after engaging with hydroxyl groups, particularly significant for the Cu–O bond. The hydroxyl group shift of ZnO (30.2 cm−1) following Cu ion adsorption is larger than that of Ni (8.9 cm−1) and Cd (2.2 cm−1), indicating that ZnO and Cu have the strongest binding contact

Full size image
Fig. 3
figure 3

The X-ray diffraction of ZnO nanoparticles prepared by hydrothermal synthesis. Reproduced from (Abdulkhair et al. 2021) with permission from Elsevier. The peaks at 31.71°, 34.38°, 36.30°, 47.52°, 56.56°, 62.91°, and 67.93° correspond to the (10 0), (0 02), (101), (102), (110), (103), and (112) planes of zinc oxide nanoparticles, respectively. The diffractions are consistent with the wurtzite hexagonal structure of zinc oxide nanoparticles (JCPDS 36-1451)

Full size image

A summary of the adsorption capacity of ZnONPs for heavy metals and radionuclides uptake from solution is presented in Table 2. It is observed that the ZnONPs used so far in the adsorption of heavy metals and radionuclides apart from the commercial ones are prepared by the green, sol–gel, precipitation, solvothermal, co-precipitation, chemical reduction, hydrothermal, and the combustion methods. Moreover, the surface areas presented by the ZnONPs were in the range of 3.93–58.0 m2/g, with the highest value obtained from the hydrothermal synthesis. The surface area was much lower than other potent adsorbents with high surface areas such as activated carbon (200–2640 m2/g) (Pui et al. 2019) and metal–organic frameworks (1000–10,000 m2/g)(Li et al. 2019). Despite the comparably low surface area presented by the ZnONPs, they exhibited significantly high adsorption capacities of 380–1500 mg/g for Pb(II), Cd(II), Hg(II), V(V), Th(IV), and U(VI) ions. This indicates that a low surface area of an adsorbent does not imply a low adsorption capacity of the material (Akpomie and Conradie 2020a) and that the adsorption efficiency of an adsorbent is not solely dependent on the surface area. This deduction was corroborated by the high adsorption capacity of 1111 mg/g obtained in the adsorption of U(VI) onto ZnONPs with low surface areas of 3.93–8.72 m2/g (Kaynar et al. 2014). In general, the adsorption capacities in the range of 0.30–1500 mg/g were obtained for the adsorption of heavy metals and radionuclides onto ZnONPs except for Cu(II) adsorption reported (Leiva et al. 2021), where a negative adsorption capacity was presented. It is rare to obtain a negative adsorption capacity for an adsorbent as other researchers obtained a positive adsorption capacity (5.084–137.5 mg/g) of ZnONPs for Cu(II) ions (Mahdavi et al. 2012; Wang et al. 2018; Primo et al. 2020; Ali and Hassan 2022). Moreover, from Table 2, several studies have been conducted on the adsorption of heavy metals on ZnONPs (except for Mn(II) ions) with only a few reports on radionuclides adsorption. Therefore, future research on adsorption onto ZnONPs should focus on the treatment of water contaminated with manganese and radionuclides such as radium, ruthenium, and radon.

Table 2 Adsorption of heavy metals and radionuclides onto zinc oxide nanoparticles
Full size table

Adsorption isotherm modeling

An adsorption isotherm describes the equilibrium performance of adsorbents at a constant temperature. The adsorbent, adsorbate species, and other physical parameters of the solution, such as temperature, ionic strength, and pH, all influence the equilibrium isotherm (Yan et al. 2017). Adsorption isotherms are established when the adsorbent and the adsorbate come into contact at a time when the interface concentration is in dynamic balance with the adsorbate concentration in the bulk solution. Adsorption isotherms are commonly utilized in the design of commercial adsorption processes as well as material characterization. The equilibrium isotherm provides the most crucial piece of information for a comprehensive understanding of an adsorption process (Al-Ghouti and Da’ana 2020). Furthermore, adsorption isotherm models describe the mechanistic interactions between contaminants and adsorbent materials by taking into account both adsorption parameters and equilibrium data.

The Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich isotherm models are the most used isotherm models applied in the adsorption of pollutants from solution onto various adsorbents. The Langmuir isotherm is based on a monolayer surface coverage with identical and equivalent definite localized sites for adsorption. There should be no steric hindrance or lateral contact between the adsorbed molecules, even on nearby sites. The Langmuir isotherm model implies that adsorption is homogeneous, with each molecule having the same constant enthalpies and sorption activation energy. The isotherm involves no adsorbate transmigration in the surface plane, and all sites should have the same affinity for the adsorbate (Gupta et al. 2021; Hamidon et al. 2022). The Langmuir model equation is written in linear form as (Umeh et al. 2021):

$$\frac{{C}_{e}}{{q}_{e}}=\frac{1}{{q}_{L}{K}_{L}}+\frac{{C}_{e}}{{q}_{L}}$$
(1)

The appropriateness of the Langmuir isotherm to the adsorption process is indicated by a straight line obtained from plotting Ce/qe against Ce. On the other hand, a reversible and non-ideal adsorption process is described by the Freundlich isotherm model. The Freundlich model, unlike the Langmuir isotherm model, is not bound to monolayer formation and can be applied to multilayer adsorption. Adsorption heat and affinities do not need to be evenly distributed across the heterogeneous surface in this model. The surface heterogeneity, as well as the exponential distribution of active sites and their energies, is defined by the Freundlich isotherm model (Al-Ghouti and Da’ana 2020). The linear expression of the Freundlich isotherm model is written as (David et al. 2020):

$$\mathrm{log}{q}_{e}= \mathrm{log}{K}_{F}+ \left(\frac{1}{n}\right)\mathrm{log}{C}_{e}$$
(2)

A straight line obtained from the plot of log qe versus log Ce confirms the applicability of the Freundlich isotherm. Furthermore, by neglecting extremely low and high concentrations, the Temkin model assumes that the heat of adsorption of all molecules in the adsorbent layer decreases linearly rather than logarithmically with coverage. It is characterized by a uniform distribution of binding energies until maximum binding energy. The Temkin equation is good for forecasting gas-phase equilibrium, but it rarely fits complex adsorption systems involving liquid-phase adsorption (Foo and Hameed 2010). The Temkin isotherm equation is written in its linearized form as:

$${q}_{e}=B\mathrm{ln}A+B\mathrm{ln}{C}_{e}$$
(3)

A straight line formed by plotting qe versus ln Ce indicates the Temkin isotherm’s suitability for the adsorption process. Moreover, the Dubinin–Radushkevich isotherm model does not presuppose a homogeneous surface or a constant adsorption potential of the Langmuir model. The distribution of Gaussian energy onto heterogeneous surfaces is generally connected to this adsorption isotherm. This model, unlike the Langmuir and Freundlich isotherms, is a semiempirical equation using the pore-filling mechanism. Multilayer adsorption with Van der Waal's forces is also assumed in this model (Al-Ghouti and Da’ana 2020). The Dubinin–Radushkevich isotherm equation is written in linear form as (Dawodu and Akpomie 2014):

$$\ln q_{e} = \ln q_{m} + \beta \varepsilon^{2}$$
(4)

Applicability of the Dubinin–Radushkevich isotherm to the adsorption is verified by a linear plot of ln qe versus ε2.

Table 3 shows the applied isotherm and best fit models for the adsorption of heavy metals and radionuclides onto ZnONPs. It is observed that the Freundlich. Langmuir, Temkin, Dubinin–Radushkevich, Sips, and Halsey isotherm models have been applied so far in adsorption. Moreover, the Langmuir model was found to give the best fit to the adsorption of Pb(II), Hg(II), Co(II), As(III), Cr(III), Ba(II), Zn(II), Se(IV) and V(V) with coefficients of determination (R2) in the range 0.988–1.000. However, in one of the studies, the Sips isotherm also provided the best fit alongside the Langmuir model in the adsorption of Pb(II) ions (Radhakrishnan et al. 2016). The implication of the good fit of the Langmuir model to the adsorption of these metal ions on ZnONPs is that process is restricted to monolayer adsorption onto a homogenous surface. On the other hand, the Freundlich model presented the best fit to the radionuclide adsorption of Pd(II), La(III),and U(VI) on ZnONPs attributed to a multilayer heterogeneous uptake. However, more than one isotherm model was found to be applicable in the adsorption of Cd(II), Cu(II), Ni(II), Cr(VI,) and Ce(III) indicating complex adsorption involving multiple mechanisms occurring simultaneously. Chemisorption usually includes the formation of a monolayer (Langmuir isotherm), whereas physisorption entails the formation of a multilayer (Freundlich isotherm) (Al-Ghouti and Da’ana 2020). However, it is not recommended to conclude on the chemical or physical nature of adsorption based on the good fit of the Langmuir or Freundlich model alone, rather a reliable conclusion would involve a holistic consideration of the isotherm, kinetics thermodynamics, desorption, and mechanistic interpretations. Furthermore, the favorability of the adsorption process or efficient interaction between the metal ions in solution and ZnONPs can be deduced from the Freundlich n value in the range of 1–10 (Chukwuemeka-Okorie et al. 2021). As shown in Table 3, it is observed that the values of n obtained for heavy metals and radionuclides adsorption were all in the favorable range except for a few studies involving the adsorption of Cu(II), Pb(II) and Pd(II) (Somu and Paul 2018; Primo et al. 2020; Leiva et al. 2021; Davarnejad and Nikandam 2022). This shows that metal contaminants in water and ZnONPs have a good affinity for efficient water decontamination. Besides, efficient interaction between the metal ions and ZnONPs (favorable adsorption) can also be obtained from the Langmuir separation factor (RL = 1/(1 + KLCo). The separation factor indicates whether the adsorption is linear (RL = 1), irreversible (RL = 0), unfavorable (RL > 1), or favorable (0 < RL < 1) (David et al. 2020). Again, the calculated RL values (not shown) in all the studies were in the favorable range which confirms the suitability of ZnONPs in the decontamination of wastewater polluted with heavy metals and radionuclides. Other isotherms, such as Hills, Flory–Huggins, Scatchard, Redlich–Peterson, Toth, and Jovanovich models, should be included in future studies on adsorption onto ZnONPs to provide more insight into the adsorption process.

Table 3 The isotherm modeling of heavy metals and radionuclides adsorption onto ZnO nanoparticles
Full size table

Kinetics of adsorption

Kinetic model assessments are crucial because they aid in the design of adsorption systems by assisting with retention times and reactor dimensions. They also provide crucial information on pollutant adsorption mechanisms involving diffusion and adsorption on active sites (Akpomie and Conradie 2020a). Adsorption is a complicated process that often involves a combination of surface adsorption and diffusion into the pores (Benjelloun et al. 2021). In the adsorption process, there are three basic steps. External mass transfer of the adsorbate from the bulk solution to the adsorbent's external surface comes first, followed by internal diffusion of the adsorbate to the sorption sites, and ultimately sorption. Some models assume that the rate-limiting stage in the adsorption process is sorption, whereas others assume that the rate-limiting step is diffusion. As a result of the fitting to the kinetic models, the adsorption mechanism may be deduced (Largitte and Pasquier 2016). The pseudo-first-order, pseudo-second-order, and intraparticle diffusion equations are three kinetic models mostly applied in the adsorption of pollutants from solution onto various adsorbents. According to the pseudo-first-order or Lagergren model, the rate of adsorption site occupancy is proportional to the number of vacant sites (Blázquez et al. 2011). The pseudo-first-order equation is written in linear form as (Eze et al. 2021):

$$\mathrm{log}\left({q}_{e}-{q}_{t}\right)=\mathrm{log}{q}_{e}-\frac{{K}_{1}}{2.303}t$$
(5)

A linear fit obtained from the plot of log (qe–qt) versus t indicates the suitability of the pseudo-first-order model. On the other hand, the pseudo-second-order model assumes that sorption site occupation is proportional to the square of the number of empty sites. The pseudo-second-order equation is written in the linear form as (Pang et al. 2022):

$$\frac{t}{{q}_{t}}=\frac{1}{{K}_{2}{{q}_{e}}^{2}}+\frac{t}{{q}_{e}}$$
(6)

A straight line obtained from the plot of t/qt against t confirms the applicability of the pseudo-second-order model. The pseudo-first-order and pseudo-second-order models do not provide information on the diffusion mechanism, and thus, information on the mechanism of diffusion can be obtained from the intraparticle diffusion model expressed as (Umeh et al. 2021; Mogale et al. 2022):

$$q_{t} = K_{d} t^{{1}/{2}}+ C$$
(7)

If the plot of qt vs t1/2 is linear and passes through the origin (C = 0), intraparticle diffusion is the only rate-determining step. However, adsorption is regulated by both film and intraparticle diffusion mechanisms when C is not equal to 0. The bigger the value of C and the greater the plot's divergence from linearity, the more substantial the film diffusion (boundary layer diffusion) effect (An et al. 2022).

The kinetic models applied to the adsorption of heavy metals and radionuclides onto ZnONPs are presented in Table 4. Information on the best-fitted kinetic model, coefficient of determination (R2), and model constants for the best-fitted model and diffusion mechanism is also presented. It is evident that the pseudo-second-order model presented the best for the adsorption of all the heavy metals and radionuclides onto ZnONPs with R2 values in the range 0.986–1.000. Many researchers over the years have attributed the best fit of the pseudo-second-order model to the chemisorption mechanism (Liu et al. 2022). Such a conclusion based on the best fit of the pseudo-second-order model alone is not recommended as this model always presents a good fit to this kinetic adsorption data irrespective of the nature of adsorption or the rate-controlling mechanism (Simonin 2016; Akpomie et al. 2017). Rather, the pseudo-second-order model's good fit shows that the rate of adsorption is controlled by both the unoccupied active sites in the adsorbent and the concentration of metal ions in the solution. It also implies that valence forces may be involved in electron exchange and sharing between the functional groups of the adsorbent and the adsorbate species (Vishan et al. 2019). This was corroborated by the previous discussion in Fig. 2, where a potent interaction between the hydroxyl groups of ZnONPs and the metals resulted in the formation of a metal–oxygen bond (Wang et al. 2018). Moreover, the pseudo-second-order rate constant (k2) for the adsorption of different heavy metals and radionuclides on ZnONPs was in the range of 1 × 10–5–4.42 g/mgmin. The differences in the hydrated ionic radii of metal ions influence their rate of adsorption onto adsorbents, as metals with smaller ionic radii tend to diffuse faster resulting in a faster adsorption rate. In addition, the differences in other properties of the metals, such as electronegativity, acidity strength, and the pKOH values of the metal hydroxides in solution, could also influence the rate of adsorption (Barka et al. 2013). Pertaining to the diffusion mechanism of adsorption, it is evident from Table 4 that a good number of researchers did not consider this. However, the investigations available demonstrated that both film and intraparticle diffusion play a role in the overall adsorption of heavy metals and radionuclides on ZnONPs, with varying degrees of contribution. This suggests that the adsorption of the pollutants on ZnONPs is a complex process involving several mechanisms. However, in order to arrive at a more elaborate conclusion, future research on the adsorption of heavy metals and radionuclides should also take into consideration the diffusion kinetics. Moreover, additional kinetics models, such as the Elovich, Bangham, Crank, Boyd, and film diffusion, could also be used to gain a better understanding of the kinetics of adsorption onto ZnONPs (Qiu et al. 2009; Largitte and Pasquier 2016).

Table 4 The kinetic modeling of heavy metals and radionuclides adsorption onto ZnO nanoparticles
Full size table

Thermodynamics of adsorption

The thermodynamics of adsorption is significant because it reveals how temperature influences the adsorption process. It also gives useful information about adsorption's feasibility or spontaneity, the process's exothermic or endothermic nature, the system's disorderliness or randomness, and the physical or chemical nature of adsorption (Mogale et al. 2022). The calculation of thermodynamic parameters such as Gibbs free energy changes (ΔG°), entropy changes (ΔS°), and enthalpy changes (ΔH°) yields this crucial thermodynamic information. The three thermodynamics parameters are calculated from the Gibbs free energy and Van’t Hoff’s equations expressed as (Ezekoye et al. 2020):

$$\Delta G^\circ = - {\text{RT}} \ln K_{c}$$
(8)
$$\ln K_{C} = - \left( {\frac{\Delta H^\circ }{{{\text{RT}}}}} \right) + \left( {\frac{\Delta S^\circ }{R}} \right)$$
(9)

A linear plot of ln Kc versus 1/T allows for the determination of ΔH° and ΔS° from the slope and intercept of the plot, respectively. Generally, negative and positive values of ΔG° correspond to a spontaneous and non-spontaneous adsorption process, respectively (Eze et al. 2021). Similarly, positive and negative values of ΔS° are ascribed to an increase and decrease in randomness at the adsorbent/adsorbate interface, respectively (Singh et al. 2022). Furthermore, positive ΔH° is indicative of endothermic adsorption, while negative ΔH° values correspond to an exothermic process. Moreover, the magnitude of ΔH° in the range of 21.0–418.4 kJ/mol or less than 21.0 kJ/mol indicates the dominance of chemisorption or physisorption in the overall adsorption process, respectively (Nanthamathee and Dechatiwongse 2021). The thermodynamic parameters obtained for the adsorption of heavy metals and radionuclides onto ZnONPs are shown in Table 5. It is evident that spontaneous adsorption of the pollutants on ZnONPs was achieved in many studies due to the negative ΔHo values obtained. This suggests favorable interaction between the heavy metals and radionuclides in solution and the ZnONPs adsorbent. However, a non-spontaneous removal was obtained for the adsorption of Ba(II) at a higher concentration (100 mg/L) (Abdulkhair et al. 2021), which indicates that the higher concentration was not favorable for the adsorption. This was corroborated by the decrease in the magnitude of ΔH° from 79.22 to 17.06 kJ/mol with an increase in concentration from 10 to 100 mg/L showing a shift from stronger chemical interactions to weaker physical bonding. In addition, non-spontaneous adsorption was also obtained at a higher temperature for Cd(II), Hg(II), Zn(II), and Ba(II) in some investigations (Sheela et al. 2012; Abdulkhair et al. 2021). This indicates that higher temperature was not supportive of the adsorption process as corroborated by the negative ΔH° obtained in these studies attributed to exothermic adsorption. For the ΔS°, both increase and decrease in the randomness at the ZnONPs/metal interfaces were observed in the studies, which was strongly influenced by the exothermic or exothermic nature of the process. Moreover, the uptake of some of the metals from the solution was endothermic, while others were exothermic, suggesting the suitability of ZnONPs in the adsorption of heavy metals and radionuclides at various temperatures under tropical and temperate conditions (Akpomie and Conradie 2020a). According to the magnitude of ΔHo, the adsorption of Hg(II), Zn(II), Th(IV), and U(VI) is dominated by chemical forces, while the uptake of Co(II), As(III), Cu(II) and Ni(II) corresponds to a physical adsorption process. Recall from the kinetic analysis that the pseudo-second-order equation presented the best fit for all the metal ions and radionuclides. Therefore, as stated earlier, it would be misleading to conclude that an adsorption process is a chemisorption based on the good fit of the pseudo-second-order model alone. Therefore, a holistic consideration of the isotherm, kinetic, thermodynamic, and mechanism would be beneficial to arrive at a reliable classification of a process as physisorption or chemisorption.

Table 5 Thermodynamic investigations on the adsorption of heavy metals and radionuclides onto ZnO nanoparticles
Full size table

Reusability of ZnO nanoparticles

When considering the costs of resources, adsorbent preparation, and secondary waste management, the performance of an adsorbent in reuse trials is critical for life-cycle assessment. As a result, adsorbents' chemical and physical stability must be maintained over a long period for them to be used on a large scale (Maia et al. 2021). Besides, desorption and reusability studies can also be used to determine the nature of the adsorbent–adsorbate interaction (physical or chemical adsorption) throughout the adsorption process. The adsorbate molecules are weakly bound to the adsorbent surface (physisorption), as evidenced by desorption through a neutral pH solution or water. If the desorption was carried out using a strong solvent, very acidic or basic solution, it is reasonable to assume that adsorbate molecules occupied the adsorbent surface via an ion-exchange mechanism or chemisorption (Akpomie et al. 2015). Although, this is not conclusive (Iwuozor et al. 2022). The usage of organic acid as a desorption solvent implies that a chemisorption mechanism was involved. A successful desorption process is entirely reliant on solvents that have been carefully chosen. The solvent for the desorption process should be chosen based on the adsorption mechanism. However, over the years, researchers have used desorbing solvents at random, without considering these factors (Ahmad and Danish 2022). Furthermore, an adsorbent must not only have a high adsorption capacity and quick removal kinetics, but it must also be able to be regenerated and reused over time to be designated as efficient for commercial application. Despite the importance of the reuse of adsorbents, only a few of the publications on the adsorption of heavy metals and radionuclides on ZnONPs conducted desorption and reusability experiments as shown in Table 6. So far, hydrochloric acid, sodium hydroxide, nitric acid, methanol, water, calcium chloride, and ethyl acetate have been utilized as eluents, with hydrochloric acid solution predominating. The predominance of acid solution in the desorption of metal ions could be due to the reduced adsorption of these species at low pH values due to the increased competition or replacement with hydrogen ions on the surface of the adsorbent. It is also observed that the ZnONPs were successfully regenerated and reused from 2 to 10 adsorption–desorption cycles. Moreover, the regeneration performance showed that ZnONPs were efficient in the uptake of Pb(II), Cd(II), Co(II), Pd(II), Ba(II), and Se(IV) ions from the solution. This proves the viability of ZnONPs for practical applications in the treatment of wastewater contaminated with these metals. However, poor regeneration and reuse were obtained in the adsorption of Cu(II) and Cr(VI) ions probably due to the poor performance of the eluent used. Therefore, other desorbing agents should be considered to achieve efficient desorption and reuse, which should be selected based on the mechanism of adsorption. Furthermore, because there are few data on the regeneration and reuse of ZnONPs in the adsorption of heavy metals and radionuclides, more research is needed.

Table 6 The reusability of ZnO nanoparticles in the adsorption of heavy metals and radionuclides
Full size table

Mechanism of adsorption onto ZnO nanoparticle

The mechanism of adsorption of heavy metals on adsorbents is usually via electrostatic interaction, hydrophobic interaction, chelation, ion exchange, hydrogen bonding, precipitation, reduction, complexation, π-π interaction, or weak Van der Waals interaction (Singh et al. 2020; Akpomie and Conradie 2020b). The adsorption process proceeds via one or a combination of two or more of these interactions. Several factors influence the mechanism, including the pH of the solution, the textural qualities of the adsorbent, and the chemical structure of the target molecules. Although pinpointing the specific interactions at work is difficult, many researchers find the Fourier transform infrared spectroscopy to be a useful technique for investigating solute–adsorbent interactions (Qureshi et al. 2020). The process might be predicted arbitrarily based on FTIR and surface charge since the adsorption mechanism is heavily influenced by surface functional groups (Ahmad and Danish 2022). The hydroxyl, amino, and carboxyl moieties are the main functional groups that interact with metals (Rajapaksha et al. 2016). However, the functional groups of ZnONPs are mainly the hydroxyl groups, and hence, we expect a limited number of mechanisms in the removal of metals since metals do not possess adequate functionalities as organic contaminants. Most investigations on the adsorption of heavy metals and radionuclides did not consider the adsorption mechanism. The few studies available only presented a cursory interpretation of the formation of bonds between the functional groups of ZnONPs and the metals (Kumar et al. 2013; Wang et al. 2018; Yuvaraja et al. 2018). Some studies proposed the complexation mechanism between the ZnONPs and metals, which was highly influenced by the solution pH (Khezami et al. 2017a; Azizi et al. 2017; Primo et al. 2020). Yin et al. reported that the surface of ZnONPs contains hydroxyl groups due to the adsorption of water and partial dissociation of water molecules (Yin et al. 2018). The abundance of hydroxyl groups affords sites for metal adsorption. The percentage of hydroxyl groups in ZnONPs dropped with the formation of metal–oxygen bonds, according to their FTIR and XPS studies, indicating that the hydroxyl groups were complexing with V(V) and Pb(II) via a complexation mechanism (Yin et al. 2018). Another study reported that the dominant mechanism in the adsorption of Cu(II) ions onto ZnONPs was complexation and precipitation based on the pH study (Mahdavi et al. 2012). An ion-exchange mechanism was observed in another work by a drop in pH of the solution following adsorption of Cd(II) and Pb(II) due to the release of hydrogen ions (Radhakrishnan et al. 2016). Furthermore, electrostatic interaction was proposed in the adsorption of Cr(VI) as a result of the drop in the zeta potential of ZnONPs from 32.6 to 19.0 mV after adsorption at pH 4.0 due to the uptake of negatively charged chromium species (Zhao and Qi 2012). Therefore, based on these reports, the complexation, precipitation, ion exchange, and electrostatic interactions are the probable mechanisms in the adsorption of heavy metals and radionuclides on ZnONPs with a predominance of complexation. However, due to the limited and superficial data currently available, future studies should focus on and undertake an in-depth analysis of the mechanism of adsorption of heavy metals and radionuclides onto ZnONPs.

Conclusions and future recommendations

The poisoning of ambient waters with heavy metals and radionuclides is on the rise as a result of rapid technological improvement. Humans and the ecosystem are both at risk from these toxins. Adsorption onto zinc oxide nanoparticles (ZnONPs) has been proven to be an efficient and low-cost method for treating contaminated wastewater. As a result, various investigations on the adsorption of heavy metals and radionuclides onto ZnONPs have been done. In this review, we looked at the isotherm, kinetics, thermodynamics, and mechanism of adsorption of the contaminants on ZnONPs to gain a better understanding of the adsorption process. The ZnONPs produced using various techniques had surface areas ranging from 3.93 to 58.0 m2/g and adsorption capacities ranging from 0.30 to 1500 mg/g. In the isotherm analysis, the Langmuir model was shown to be the best fit for the adsorption of Pb(II), Hg(II), Co(II), As(III), Cr(III), Ba(II), Zn(II), Se(IV), and V(V), whereas the Freundlich model was adequate for Pd(II), La(III), and U(VI). The Langmuir separation factor showed favorable adsorption onto ZnONPs in all cases. The kinetic evaluation revealed that the pseudo-second-order model presented the best fit in all reports with the occurrence of both film and intraparticle diffusion mechanisms. The pseudo-second-order rate constant (k2) for the adsorption on ZnONPs was in the range of 1 × 10–5–4.42 g/mgmin. Thermodynamics revealed spontaneous adsorption of the heavy metals and radionuclides on ZnONPs (not in all cases). Moreover, both endothermic and exothermic processes were observed. Desorption investigations showed that hydrochloric acid, sodium hydroxide, nitric acid, methanol, water, calcium chloride, and ethyl acetate have been utilized as eluents with the predominance of hydrochloric acid. ZnONPs exhibited potent reusability in the uptake of Pb(II), Cd(II), Co(II), Pd(II), Ba(II), and Se(IV) ions but displayed poor performance for Cu(II) and Cr(VI). The complexation, precipitation, ion exchange and electrostatic interactions were the probable mechanisms in the adsorption of heavy metals and radionuclides on ZnONPs with a predominance of complexation. The overview demonstrated the potential of ZnONPs as an efficient adsorbent in the decontamination of heavy metal and radionuclide-contaminated wastewater.

The investigations that are now available have some flaws or information gaps. Therefore, these flaws must be considered to broaden the scope of the application of ZnONPs for the adsorption of heavy metals and radionuclides. Since industrial effluents contain various metal ions and radionuclides, more research into the competitive adsorption of these pollutants from multipollutant systems is needed. There are currently no studies on the adsorption of manganese and radionuclides such as radium, ruthenium and radon onto ZnONPs, these aspects should also be considered. In the isotherm analysis, only the Langmuir, Freundlich, Temkin, Dubinin–Radushkevich, Halsey and Sips isotherms have been analyzed. Therefore, other isotherms, such as Hills, Flory–Huggins, Scatchard, Redlich–Peterson, Toth and Jovanovich models, should be incorporated to gain a better understanding of the adsorption process. Few studies evaluated the diffusion mechanism of adsorption; therefore, future research should also consider this. Moreover, additional kinetics models such as Bangham, Crank, Boyd and film diffusion could be applied to gain a better understanding of the kinetics of adsorption onto ZnONPs. Future investigations should consider holistically the isotherm, kinetic, thermodynamic and mechanism of adsorption when determining whether a process is a chemisorption or physisorption. This is to prevent drawing erroneous conclusions about a chemisorption mechanism based solely on the good fit of the pseudo-second-order model. The reusability of ZnONPs in the adsorption of heavy metals and radionuclides is currently limited in the investigation. Future research should take this into account by using highly efficient solvents, selected based on the adsorption mechanism to obtain optimum reusability. There are currently no comprehensive investigations on the mechanism of metal and radionuclide adsorption on ZnONPs, which should be examined to fully comprehend the treatment process. Considering these aspects will help provide significant insights into the design of ZnONPs adsorption systems for the treatment of real wastewater polluted with heavy metals and radionuclides.