The Darvill Wastewater Treatment Plant (WWTP) (29.60125°S, 30.42881°E) in Pietermaritzburg, KwaZulu-Natal, South Africa, was selected as the study site. This WWTP is a major facility designed to treat up to 100,000 m³ of wastewater daily (Umgeni. Annual Report, 2022/2023, 2023). Since its establishment in 1992, the plant has played a crucial role in managing the city’s domestic and industrial wastewater (Sikhakhane, 2001). The plant utilizes a combination of biological filtration and activated sludge process technologies, both of which are key components of secondary treatment, to effectively reduce organic matter and suspended pollutants before discharging the treated wastewater. The treated effluent from Darvill WWTP is released into the Msunduzi River, a significant tributary of the Umgeni River. The Umgeni River is a vital water resource in the region, serving as the primary drinking water source for rural communities in the Valley of a Thousand Hills and eventually feeding into Inanda Dam. This dam is part of the Umgeni-Uthukela catchment and supplies potable water to the Durban metropolitan area, supporting millions of residents and businesses.

The water samples were collected from the effluent point using high-quality plastic containers (PE, Lasec, South Africa) to prevent sample contamination and adsorption on the walls of the container. The containers were then sealed airtight to prevent the heavy metals from oxidizing and forming insoluble compounds in the presence of air (Abd-El-Nabey et al., 2022). A total of five 2-L samples were collected at the sampling point, ensuring sufficient volume for the experiments conducted in this study. Samples were temporarily stored in a cooler box with ice during transportation and kept in a fridge at about 4°C at the laboratory.

Parameters such as pH, conductivity, temperature, total dissolved solids (TDS), salinity, and dissolved oxygen (DO) were measured using a portable multiparameter meter (Bante Instruments 900P, China). Before experiments, the wastewater effluent samples underwent filtration using a 55 mm CHM shift filtration paper (F1001 grade, Separations, South Africa) to eliminate large particles.

The uMgeni-uThukela Water (Pietermaritzburg, South Africa) provided the Chlorella sp. strain (FSB41.2) used in this study, with cultivation routinely done using tris-acetate phosphate (TAP) medium, as outlined by Harris (Harris, 1989). An axenic culture was established by sequential subculturing and purification of single colonies of Chlorella sp. using an acetate-free medium under autotrophic growth conditions with regular culture-based (Nutrient agar) and microscopic (phase contrast) monitoring as reported previously (Adedoyin and Schmidt, 2023). Subsequently, biomass production occurred in a controlled growth environment (approximately 23°C with a 20-hour light-dark cycle) using 1,000 mL Erlenmeyer flasks, each filled with 600 mL of the TP culture medium following autoclaving. The growth of cultures was monitored over time by absorbance measurements at 680 nm using a UV/Visible Spectrophotometer (JENWAY 7310, United Kingdom).

The Chlorella sp. biomass was extracted from the TP culture medium after reaching the late exponential growth phase by centrifugation (3354 g for 15 min) and washed 3 times with distilled water to remove the remaining medium. Following centrifugation, the resulting Chlorella sp. cell pellets were subjected to overnight oven drying at 80°C, since very low or exceedingly high temperatures negatively affect the cell structure (Tattibayeva et al., 2022). Subsequently, an electric grinder (MRC SM-450, Material Research Company, Israel) was used to grind the dried pellet, followed by sieving (KINGTEST sieve shaker VB200/300, Kingtest, China) and storing the dried biomass in a sealed container to prevent rehydration.

Dried Chlorella sp. biomass (12.5 ± 0.1 mg) obtained from axenic cultures (see Section 2.5 for biomass preparation details) was added to conical flasks containing 30 mL of deionized water. The water was spiked with a mixture of Cu, Pb, and Zn ions, each applied at the same initial concentration within the range of 0.25–16 mg/L. The initial experiments were conducted to optimize adsorption conditions, including biomass dosage, metal ion concentration, pH, and contact time.

The optimal conditions were then applied to real wastewater to assess adsorption performance in a realistic scenario and evaluate the removal efficiency of contaminants in actual water systems. Adsorption behavior was evaluated using Freundlich, Langmuir, and Temkin isotherms, while kinetic studies were assessed using intraparticle diffusion models, pseudo-first-order, and pseudo-second-order kinetics. For isotherm and kinetic experiments, the samples were incubated in a rotary shaker at 150 rpm and 25°C for 60 min. The initial concentrations of heavy metals and the contact time of 60 min were chosen based on the optimization results, which identified conditions that achieved optimal adsorption performance. The temperature and shaker speed were maintained constant throughout the study, as these parameters are widely supported in the literature for ensuring consistency and reliability in adsorption experiments (Li et al., 2020; Pfeifer and Skerget, 2020; Cherono et al., 2021; Tattibayeva et al., 2022; Wang and Guo, 2023). The samples were centrifuged at 3354 g for 15 min to separate the Chlorella cells from the liquid. The liquid samples were then filtered with a 0.22 μm pore size filter (Merck Mille, Germany) and a 30 mm sterile syringe to remove residual algal cells before analysis with ICP OES. The impact of temperature on adsorption equilibrium was explored by altering the temperature within the 298–333 K range (Tang et al., 2017; Shi et al., 2023; Soto-Ramírez et al., 2023).

The adsorption process was quantified using Equation 2: q e = C 0 − C e   m   x   V (2)

Where qe is the adsorption capacity (mg/g), C0 is the initial concentration of the adsorbate (mg/L), Ce is the equilibrium concentration of the adsorbate (mg/L), V is the volume of the solution (L), m is the mass of the adsorbent (g).

The degree of removal (R) of heavy metal ions was calculated by using Equation 3: %   R e c o v e r y = C 0 − C e   C 0 x   100 (3)

Where: RE is the removal efficiency in percentage, C0 is the initial concentration before adsorption, and Ce is the concentration at equilibrium.

Scanning electron microscopy (SEM) is a powerful tool for visualizing the surface features and morphology of microorganisms such as Chlorella sp. (Nasir et al., 2023). As expected, the cells of Chlorella sp. used even in the dried state, showed an almost intact surface area, enabling the binding of metal ions via interaction with exposed functional groups (Figure 1a) (Spain et al., 2021; Tattibayeva et al., 2022). An image taken after adsorption indicated morphological changes resembling cell shrinkage and increased surface irregularities, suggesting an interaction of the cell surface with metal ions (Figure 1b).

The results of the analysis of the elemental composition of Chlorella sp. biomass, which comprises carbon (C), nitrogen (N), oxygen (O), phosphorus (P), and sulfur (S) are shown in Figure 2a. The obtained elemental composition in this study closely aligns with the literature values reported for Chlorella sp. by Mandalam and Palsson, with the exception of slightly higher levels of nitrogen (N) and sulfur (S) in this study (Mandalam and Palsson, 1998). The detected elements are associated with microalgal biomass, such as Chlorella sp., and are key elements present in microalgal proteins, nucleic acids, lipids, carbohydrates, and organic acids (Ma et al., 2020; Yaakob et al., 2021). Organic functional groups, such as carboxyl (-COO) and hydroxyl (-OH), are abundant on the cell surface of Chlorella sp. and are crucial in the adsorption of heavy metals by providing binding sites (Yang et al., 2019). Similarly, nitrogen-containing functional groups such as amide (C-N) associated with cell surface proteins form various interactions with metal ions (Giarikos et al., 2021), which applies as well to phosphate and sulfur-containing groups.

Analytical methods such as Fourier-transform infrared spectroscopy (FTIR) are vital in identifying functional groups present in microbial biomass. For Chlorella sp., the functional groups present in biomass samples analysed included hydroxyl (-OH), methyl (-CH₃), carbonyl (C=O), alkene (C=C), carboxyl (-COO), amide (C-N), and ether (C-O) (Figure 2b). These functional groups increase the overall binding capacity and reactivity of Chlorella biomass. This study results agree with the findings of Ahmad et al. (2019), El-Naggar et al. (2020), Sultana et al. (2020), Shi et al. (2023), who reported the occurrence of similar functional groups that play a significant role in heavy metal adsorption. Notably, a shift from 3,270 to 3,281 cm^-1 after adsorption was observed in the peak associated with the hydroxyl group, which enables hydrogen bonding interactions (Wohlert et al., 2022; Wang and Guo, 2023). This observed change in peak intensity and broadness confirmed the interactions of hydroxyl groups with metal ions. The observed shift in C-H stretching vibrations from 2,925 to 2,968 cm^-1 provided evidence of their involvement in hydrophobic interactions (Barchi and Strain, 2023). The increased intensity, as well as shifts in C=O, C=C, and C-N stretching vibrations from 1,618 to 1,638 cm^-1, 1,538 to 1,528 cm^-1, and 1,234 to 1,250 cm^-1, respectively, suggests the interaction of carbonyl, alkene, and amide groups with metal ions during the adsorption process through metal coordination (Spain et al., 2021; Li H. et al., 2022; Barchi and Strain, 2023). Other noticeable shifts in the carboxyl group (1,407–1,398 cm^-1) and ether group (1,036–1,028 cm^-1) were observed. The carboxyl groups promote electrostatic interactions through negatively charged sites, and ether groups play a crucial role in the overall surface polarity (Barczak, 2019; Zhang et al., 2019). These results demonstrated a diversity of crucial functional groups present on the dried Chlorella sp. biomass, enabling the complexity of interactions and ultimately impacting the adsorption capacity of the Chlorella biomass for target heavy metals.

Finding optimum conditions for critical parameters such as biomass dosage is crucial in Chlorella-biomass-based adsorption studies as it ensures cost-effectiveness and enhances the efficiency and environmental sustainability of the process. The results depicting the effect of biomass dosage are illustrated in Figure 4a. The ANOVA results (p < 0.05) showed that biomass dosage significantly affects metal removal efficiencies (Supplementary Table S5).

Copper removal efficiency decreased as the biomass dosage increased (p < 0.05), with the highest removal being 56% at 5 mg. This could be ascribed to competitive adsorption as the heavy metals were present in a mixture, with the three metals potentially competing for the same available binding sites. These findings align with the observations made by Ni et al. (2019), who documented competitive adsorption behaviour between cadmium and lead, where cadmium showed a decreased adsorption in the presence of lead. This was ascribed to lead’s higher electronegativity and binding affinity, larger hydrolysis constant, atomic weight, ionic radius, and a larger Misono softness value, increasing its binding favourability (Al Hamouz and Ali, 2013; Ni et al., 2019; Musumba et al., 2020). In this study, the Chlorella sp. biomass consistently removed lead and zinc with high removal efficiencies, with maximum values of 86% and 95% recorded at 25 mg, respectively. Tukey’s post hoc test exhibited that adsorption at 25 mg and 50 mg was not significantly different for lead (p = 0.6539) and zinc (p = 0.5255), indicating a plateau in adsorption efficiency at these dosages (Supplementary Table S5). However, the 100 mg dosage led to a significant reduction in zinc removal (p < 0.05) (Supplementary Table S8), likely due to biomass aggregation, which reduces the surface area exposed to the solution, severely impacting adsorption capacity (Pfeifer and Skerget, 2020). A biomass dosage of 12.5 mg was selected for the later experiments. This selection was justified by the nearly identical removal efficiencies observed at 12.5 mg and 25 mg for lead and zinc (above 80%), as well as notably better removal for copper.

Figure 4b illustrates the effect of a varying pH at constant biomass dosage, contact time, temperature, and metal concentration upon the metal removal efficacy. The low removal efficiencies observed at acidic conditions (pH 3-5), particularly for copper and zinc, could be attributed to the protonation of metal binding functional groups (e.g., -NH₂, -OH, -SH) in Chlorella biomass, causing them to behave like positively charged species that repel metal cations (Cherono et al., 2021; Tattibayeva et al., 2022). Similarly, several studies reported that low pH promoted competition between metal ions and protons for binding sites, resulting in reduced adsorption of their analytes (Kiruba et al., 2014; Gonte and Balasubramanian, 2016; Shi et al., 2023). Additionally, Almomani and Bhosale documented that the solubility of heavy metals in water increases in acidic conditions, resulting in decreased adsorption rates (Almomani and Bhosale, 2021). The removal efficiencies for copper, lead, and zinc remained relatively stable between pH 7 and 9, with values of 56.35% and 55.75% for copper, 89.68% and 89.86% for lead, and 88.98% and 89.40% for zinc, respectively. Tukey’s post hoc test confirmed that the differences in removal efficiencies between pH 7 and 9 were not statistically significant for copper (p = 0.3207), lead (p = 0.9989), and zinc (p = 0.5604), suggesting an equilibrium point for metal binding within this pH range (Supplementary Tables S1-S4). However, when the pH becomes too basic (pH 11), a decrease in removal efficiencies was observed, particularly for copper and, to a lesser degree, for lead and copper, which can be attributed to the formation of insoluble metal hydroxides, hindering metal adsorption onto binding sites (Al-Saadi et al., 2013; Cherono et al., 2021). These findings match studies by Zhang and Depci et al. (Depci et al., 2012; Zhang, 2011) that assessed the removal of heavy metals employing dairy manure compost and activated carbon derived from apple pulp. The significant drop in adsorption at pH 11, as indicated by Tukey’s post hoc test (p < 0.001), supports that extreme pH conditions negatively impact metal removal efficiency. Thus, pH 7 was chosen as the optimum condition as it helps to prevent the disruption of pH-sensitive biological processes and the accumulation of toxic substances in receiving water bodies, contributing to the overall health and sustainability of the environment.

The effect of contact time on removing selected heavy metals was investigated (Figure 4c). An increase in lead and zinc removal efficiencies from 15 to 30 min was observed, attributable to lead and zinc having a larger ionic radius than copper, resulting in stronger interactions with binding sites on the Chlorella sp. cell surface, leading to higher and faster adsorption than small copper ions (Ni et al., 2019). Tukey’s post hoc test confirmed significant differences between 15 and 30 min for lead (p < 0.05) and zinc (p < 0.05), reinforcing the impact of contact time on metal removal efficiency (Supplementary Tables S11, 12). As the contact time increased (45–115 min), the removal of these 2 compounds did not increase substantially, indicating saturated binding sites and the establishment of an equilibrium between metal ions and the binding sites available on the surface of Chlorella biomass. Tukey’s post hoc further confirmed this trend as the results for lead showed that p-values between 45 and 75 min (0.9996), 45 and 90 min (0.0616), and 75 and 90 min (0.1226) were not statistically different (Supplementary Table S11). Similarly, for zinc, the p-values between 45 and 75 min (0.6205), 45 and 90 min (0.9170), and 75 and 90 min (0.1406) were not statistically different.

The copper removal efficiency increased from 30 to 60 min, stipulating different copper binding kinetics for the Chlorella biomass (p < 0.05). However, at ≥75–115 min, slightly lower removal efficacy indicated partial saturation or changes in binding site accessibility. The decrease observed could also indicate desorption, where previously adsorbed copper ions are released from the binding sites back into the solution (Chatterjee and Abraham, 2019). Tukey’s post hoc results for copper showed no significant differences in adsorption between 75 and 90 min (p = 0.9999), 75 and 115 min (p = 0.7975), and 90 and 115 min (p = 9,166), indicating a plateau in adsorption efficiency beyond 60 minutes (Supplementary Table S10). It is important to note that at 15 min, copper removal was not apparent, possibly due to the microalgal biomass not having enough time to adsorb copper ions effectively. Thus, the optimum time was established as 60 min, with maximum removals of 70% observed for copper, 95% for lead, and 100% for zinc.

This study explored the effect of metal concentration on the adsorption by Chlorella sp. biomass to understand the dynamic relationship between initial metal ion concentrations and their adsorption behaviour. For this analysis, the other four variables (e.g., biomass dosage, pH, temperature, and contact time were kept constant (Figure 4d). The ANOVA results (p < 0.05) showed that metal concentration significantly affected the adsorption (Supplementary Table S13). A notable increase in copper removal efficiency was observed, from 37% at 0.25 mg/L to 79% at 1.0 mg/L. This could be due to the increased availability of copper ions at higher concentrations (0.5 and 1.0 mg/L), providing more opportunities for adsorption onto the Chlorella biomass. Tukey’s post hoc test confirmed this trend, revealing a significant increase in adsorption between 0.25 mg/L and 0.5 mg/L (p < 0.05), as well as between 0.25 mg/L and 1.0 mg/L (p < 0.05) (Supplementary Table S14). Al-Homaidan et al. (2015) reported a similar trend when assessing the removal of cadmium ions by dried biomass of Spirulina (Arthrospira) platensis. However, the lead and zinc removal efficiencies from 0.25 mg/L to 1.0 mg/L did not change notably. This can also be ascribed to the abundant availability of the binding sites on the biomass surface at low metal concentrations. Tukey’s post hoc test showed no significant difference in lead and zinc removal between 0.25 mg/L, 0.5 mg/L, and 1.0 mg/L (p > 0.05), supporting the observation that their adsorption remained relatively constant at lower concentrations (Supplementary Tables S15, S16). Above metal concentrations of 1 mg/L, a decrease in copper and zinc removal efficiencies was observed, but to a lesser degree for lead. This is statistically supported by Tukey’s post hoc test, which confirmed significant reductions in copper and zinc removal efficiencies at concentrations above 1 mg/L, with all comparisons yielding p-values < 0.05 (Supplementary Tables S15, S16). The increased concentration of heavy metal ions in the solution may overwhelm the active sites, resulting in heightened competition among available ions for limited binding sites (Musumba et al., 2020). Overall, copper showed lower adsorption, contributing to it being less efficiently removed than the other two metals. Kiruba et al. also reported reduced removal as cadmium concentrations increased due to the saturation of the active sites of the surface-modified Eucalyptus seeds system (Kiruba et al., 2014).

For EDTA (Figure 5a), there is a rapid decline in removal efficiencies, particularly by Cycle 2, suggesting that the employed EDTA concentration has limited effectiveness in desorbing metals from the biomass after the first cycle. This is likely because lead (Pb) and zinc (Zn) exhibit strong binding affinities to functional groups, such as carboxyl, hydroxyl, and amino groups, on the Chlorella sp. biomass surface (Al Hamouz and Ali, 2013; Ni et al., 2019; Musumba et al., 2020). These metals form strong ionic bonds or stable complexes that require more energy to disrupt. EDTA at the tested concentration may have acted as a weaker chelating agent, failing to fully break these bonds, leaving residual metals that progressively reduce the biomass’s adsorption capacity over cycles. Thus, increasing the EDTA concentration could potentially improve desorption efficiency by providing a stronger chelating effect. Supporting this, Kumar et al. (2018) reported that desorption capacity increased with higher EDTA concentrations when desorbing cadmium. Other studies have similarly shown that the desorption efficiency of HCl improves with higher concentrations (Wu et al., 2015; Zhang et al., 2021b).

Unlike EDTA, the use of 0.05 M HCl (Figure 5b) resulted in a slightly slower decline in removal efficiencies, suggesting that HCl is more effective at regenerating the biomass for reuse. The acidic environment created by HCl protonates the biomass’s functional groups by donating protons (H⁺ ions) to surface groups like carboxyl and hydroxyl on the Chlorella sp. biomass (Chatterjee and Abraham, 2019; Zhang et al., 2021b; Yadav et al., 2022). This protonation reduces electrostatic interactions between the biomass and adsorbed metal ions, facilitating metal desorption by weakening the strong bonds with lead and zinc (Martínez et al., 2023). Additionally, HCl promotes ion exchange between hydrogen ions (H⁺) and the adsorbed metal ions, releasing metals like lead, zinc, and copper more effectively (Qasem et al., 2021). Although HCl demonstrated slightly better performance, after repeated cycles, 0.05 M HCl may still be insufficient to fully disrupt all metal-biomass interactions, particularly for Pb and Zn, which may require a stronger acid or longer desorption time. Moreover, the observed decline in adsorption in later cycles may also be attributed to alterations in the biomass’s surface chemistry, such as damaged physical structure, degraded functional groups, reduced surface area, and decreased porosity, diminishing the biomass’s adsorption capacity in future cycles. Additionally, the desorption process lasted only 2 h, which may not have been sufficient to fully desorb all metal ions, especially Pb and Zn, which form stronger bonds with the biomass. Extending the desorption time could potentially enhance removal efficiency, particularly for metals that bind tightly. The results further indicated that 0.05 M HCl maintains higher removal efficiencies for copper across cycles compared to EDTA. The protonation effect weakens copper’s coordination bonds with the biomass surface (Chatterjee and Abraham, 2019), and HCl also supplies chloride ions, which can form soluble CuCl₂ complexes, facilitating copper desorption. This enables the biomass to retain its adsorption efficiency for copper across cycles, resulting in better overall performance with HCl as the desorption agent.

The results of this study showed that while Chlorella sp. offers promise for heavy metal removal, its reusability is limited by a decline in adsorption efficiency after multiple cycles. The slower decline observed with HCl suggests that it may be a more effective desorption agent compared to EDTA. Future studies should focus on optimizing desorption processes, including adjusting desorption times and concentrations, to improve long-term effectiveness. Additionally, exploring ways to maintain or enhance the biomass’s structural integrity and surface properties could help retain its adsorption capacity in subsequent cycles.