Purple Ipe Leaf as a Sustainable Biosorbent for the Removal of Co(II) and Cd(II) Ions from Aqueous Samples

Abstract

The increasing contamination of water resources by wastewater has stimulated extensive research into advanced methods for effluent analysis, monitoring, and treatment. Heavy metals are among the most concerning pollutants due to their toxicity, persistence, and potential for bioaccumulation and biomagnification in living organisms. This study investigates the use of purple ipe (Handroanthus impetiginosus) leaves as a biosorbent for the removal of Co(II) and Cd(II) ions from aqueous solutions. The biosorbent was characterized using FTIR, NMR, EDX, SEM, and elemental analysis, revealing a porous and heterogeneous surface with functional groups suitable for metal adsorption. The point of zero charge (pH_PZC) was 5.8, and the zeta potential was −14.7 mV, indicating a negatively charged surface at higher pH values. Maximum removal efficiency was observed in the pH range of 5–6. Kinetic data showed the best fit to a pseudo-second order model, while adsorption equilibrium was most accurately described by the Langmuir isotherm, suggesting a monolayer adsorption process. The maximum adsorption capacities were 0.823 mmol g⁻¹ for Co(II) and 0.270 mmol g⁻¹ for Cd(II). The results demonstrate that purple ipe leaves are a sustainable, efficient, and low-cost biosorbent for wastewater treatment, showing great potential for mitigating environmental impacts associated with heavy metal pollution.

1. Introduction

The contamination of aquatic environments with toxic heavy metals has emerged as a major global environmental concern in recent decades, driven by rapid industrialization, urban expansion, and inadequate wastewater management practices. Unlike organic pollutants, heavy metals are non-biodegradable, bioaccumulative, and capable of inducing toxic effects at trace concentrations, which makes their presence in water bodies particularly alarming. Among these contaminants, cadmium (Cd), chromium (Cr), nickel (Ni), cobalt (Co), copper (Cu), and zinc (Zn) are the most frequently reported, originating from processes such as electroplating, metallurgical operations, mining, leather tanning, and fertilizer production [1]. Chronic exposure to such elements is associated with neurological, renal, cardiovascular, and carcinogenic disorders, posing significant risks to both ecosystems and human health [2].

Within this broader context, cadmium (Cd) and cobalt (Co) were selected in this study due to their environmental relevance, toxicity, and representativeness in real contamination scenarios. Cadmium is well known for its high toxicity, carcinogenic nature, and strong bioaccumulative behavior, becoming more available in acidic soils and entering the food chain through aquatic plants and organisms. Cobalt, although an essential micronutrient in trace amounts, is released by both natural processes, such as volcanic activity and forest fires, and anthropogenic sources including fossil fuel combustion, mining, and fertilizer use. Once deposited, Co tends to adsorb strongly to soils, especially under acidic conditions, and can cause gastrointestinal, respiratory, and systemic toxic effects. Together, these characteristics make Cd and Co particularly relevant contaminants for adsorption studies [3,4,5].

To mitigate these hazards, various conventional treatment methods have been extensively employed, including ion exchange, reverse osmosis, chemical precipitation, and coagulation/flocculation. Although often effective, these approaches present critical limitations: they require high operational costs, sophisticated infrastructure, and substantial chemical and energy inputs; furthermore, they generate large volumes of toxic sludge that demand secondary treatment and safe disposal [6]. Consequently, recent research has increasingly focused on adsorption-based processes, which stand out for their operational simplicity, versatility, cost-effectiveness, and potential for adsorbent regeneration [7].

Within this context, biosorption has gained prominence as a sustainable and eco-friendly alternative. Biosorbents—materials derived from agricultural, forestry, or agro-industrial residues—combine high adsorption efficiency with environmental compatibility and low production cost. These materials exhibit diverse functional groups such as hydroxyl, carboxyl, carbonyl, and amino moieties, which serve as active binding sites for metal ions [8]. Furthermore, biosorption aligns with the principles of the circular economy by promoting waste valorization, thereby reducing environmental burdens associated with both waste disposal and water pollution [9]. A variety of biomasses have been investigated as biosorbent feedstocks, including fruit shells, seed husks, sawdust, leaves, crop residues, and pruning by-products. Recent studies have shown that many of these biosorbents achieve high removal efficiencies under controlled laboratory conditions and exhibit sorption behaviors well described by Langmuir or Freundlich isotherm models [10]. However, despite substantial progress in the characterization and optimization of biosorption processes, several research gaps persist. First, most reported studies have been conducted under single-metal systems, while real industrial effluents typically contain multiple metal ions and other competing solutes. The simultaneous presence of various metals often results in competitive adsorption phenomena, significantly affecting removal efficiency and selectivity [11]. Consequently, results obtained from single-metal systems may not reflect the actual performance of biosorbents under real wastewater conditions. Second, many studies lack comprehensive characterization of biosorbent materials, including surface area, pore size distribution, functional group density, and surface charge (pH_PZC). Such data are fundamental for correlating the physicochemical properties of the biosorbent with its adsorption performance and for establishing predictive models [12]. Third, the regeneration and reusability of biosorbents remain underexplored in many cases, despite their importance for the economic and practical feasibility of large-scale applications [13].

Another relevant gap lies in the limited exploration of native plant species as potential biosorbent sources. Most of the biomasses investigated so far are derived from globally available agricultural wastes (such as rice husk, banana peel, or coconut shell), while region-specific and naturally abundant species remain underutilized. In this sense, the purple ipe (Handroanthus impetiginosus)—a tree native to South and Central America, particularly abundant in Brazil—presents a promising alternative [14]. This deciduous species sheds its leaves naturally during the dry season, resulting in a large seasonal biomass yield that can be collected without destructive harvesting. The leaves of the material are rich in lignocellulosic compounds and display reactive functional groups capable of coordinating metal ions, suggesting a high potential for adsorption applications [15].

Preliminary studies on the use of Handroanthus impetiginosus biomass have indicated favorable adsorption capacities toward certain heavy metals, as well as potential for chemical modification to enhance performance. Moreover, physiological studies have demonstrated that the species responds sensitively to heavy metal stress, such as mercury exposure, which implies the presence of active binding mechanisms involving phenolic and carboxylic moieties. Despite these promising findings, systematic investigations on the adsorption behavior of the material in multi-metal systems, including regeneration and stability assessments, are still scarce in the literature.

Thus, this study aims to evaluate the potential of Handroanthus impetiginosus leaves, an underutilized residue from one of Brazil’s most economically valuable tree species, as a low-cost and sustainable biosorbent for the removal of cobalt or cadmium ions from aqueous solutions under monometallic conditions. The work focuses on the physicochemical characterization of the biosorbent, the investigation of the factors influencing the interaction mechanisms between the material and the target metal ions, and the evaluation of its adsorption and preconcentration performance in batch and continuous-flow systems. The underlying hypothesis is that the lignocellulosic matrix of the material, rich in oxygen- and nitrogen-containing functional groups, can effectively bind these ions, offering a low-impact alternative to conventional adsorbents.

2. Materials and Methods

2.1. Preparation of Purple Ipe Leaf

Purple ipe (Handroanthus impetiginosus) leaves were collected from trees near Botucatu, São Paulo, Brazil (coordinates: −22.87916, −48.44484). After collecting, the leaves were washed with running water to remove superficial particles and non-adherent residues, then dried in a ventilated oven at 50 °C (323.15 K) for 24 h. The dried material was ground and sieved using standardized laboratory sieves and a mechanical sieve shaker to obtain a powder with a particle size range of 106–250 µm, which was used in subsequent batch adsorption experiments.

Following sieving, the Soxhlet extraction technique was employed for the efficient removal of persistent soluble organic contaminants. The biosorbent powder was enclosed in quantitative filter paper thimbles and placed into the extractor, which was coupled to a 500 mL round-bottom flask containing 300 mL of deionized water. The flask was heated on a heating mantle, and the condenser was connected to a recirculating chiller system. The system was maintained under continuous reflux at 100 °C (373.15 K) for 48 h, with extraction completeness confirmed by the absence of visible coloration in the solvent. Subsequently, the sample was redried in the oven at 50 °C (323.15 K) for 48 h to achieve constant mass and was stored in a desiccator under anhydrous conditions until use.

2.2. Chemicals and Reagents

Cobalt and cadmium standard solutions were prepared from cobalt(II) nitrate hexahydrate (Co(NO₃)₂·6H₂O, 98%, P.A., VETEC, Speyer, Germany) and cadmium(II) nitrate tetrahydrate (Cd(NO₃)₂·4H₂O, ≥99%, P.A., Sigma-Aldrich, Taufkirchen, Germany), respectively. Each salt was dissolved in ultrapure water (Direct-Q^® 3 UV, Millipore, Guyancourt, France). Working standard solutions for Atomic Absorption Spectrometry (AAS) were obtained by serial dilution of a certified 1000 mg L⁻¹ single-element stock solution (Specsol, São Paulo, Brazil). Solutions of hydrochloric acid (Carlo Erba, Milan, Italy) and sodium hydroxide (Merck, Darmstadt, Germany) were also prepared using ultrapure water. Leaves of purple ipe (Handroanthus impetiginosus) were acquired in their natural state and were washed using a Soxhlet apparatus prior to further use to remove soluble impurities and surface contaminants, thus ensuring greater reproducibility and comparability of the characterization and adsorption experiments.

2.3. Apparatus

Fourier-transform infrared (FTIR) spectroscopy was performed on a Nicolet iS5 spectrometer (Thermo Scientific, Waltham, MA, USA). Spectra were acquired in transmittance mode across the 4000–400 cm⁻¹ spectral range at a resolution of 4 cm⁻¹, co-adding 200 scans per spectrum. Samples were prepared by homogenizing 95 mg of material with potassium bromide (KBr) at a 2% (m/m) ratio and subsequently compressing the mixture into pellets. This methodology enabled the identification of characteristic absorption bands for the functional groups present.

Solid-state ¹³C nuclear magnetic resonance (NMR) experiments were conducted on an Agilent DD2 spectrometer operating at 250 MHz for ¹H. The purple ipe leaf sample was packed in a 4 mm zirconia rotor and analyzed using the combined cross-polarization (CP) ¹H→¹³C and magic angle spinning (MAS) technique at 5 kHz to enhance sensitivity and spectroscopic resolution. Specific acquisition parameters included a Hartmann–Hahn contact time of 1.0 ms. Spectra were accumulated from approximately 20,000 transients using a 5 s recycle delay between pulses, resulting in a total acquisition time of approximately 28 h. Chemical shifts were referenced indirectly to tetramethylsilane (TMS) by setting the CH₂ peak of adamantane to 38.6 ppm as a secondary standard.

Scanning electron microscopy (SEM) imaging of the purple ipe leaf particles was performed using an FEI Quanta 200 microscope equipped with an Everhart–Thornley secondary electron detector (FEI Company, Hillsboro, OR, USA). Elemental identification was carried out by energy-dispersive X-ray spectroscopy (EDS) using an Oxford Instruments X-Max 510XMX1119 detector (Oxford Instruments plc, Bucks, UK). The sample was prepared by compressing the material at 2 ton cm⁻² using a hydraulic press to ensure surface uniformity. The specimen was then mounted on an aluminum stub with carbon fiber adhesive tape and sputter-coated with a thin gold layer (~30 nm) to improve electrical conductivity.

Total nitrogen content was determined via the Kjeldahl method, which involved digestion with a H₂SO₄/K₂SO₄/CuSO₄ mixture, distillation with excess NaOH, and titration with 0.1 mol L⁻¹ HCl. Sulfur content was quantified by turbidimetry following an acid digestion with 65% HNO₃ (12 h) and 70% HClO₄. The liberated SO₄²⁻ ions were precipitated with 5% BaCl₂, and the turbidity of the resulting BaSO₄ suspension was measured spectrophotometrically at 420 nm after a 5 min reaction period [16]. Analytical accuracy (RSD < 5%) was ensured using certified standards and calibration curves.

The concentrations of cadmium (Cd) and cobalt (Co) ions were quantified by flame atomic absorption spectrometry (FAAS) using a Shimadzu AA-7000 spectrometer (Shimadzu Scientific Instruments, Inc., Columbia, MD, USA) equipped with hollow cathode lamps. The monochromator was set to the most intense analytical absorption lines for each element (228.80 nm for Cd and 240.73 nm for Co). External calibration curves were prepared using five concentration levels for each metal: Co (1, 2, 4, 8, and 12 mg L⁻¹) and Cd (0.1, 0.2, 0.3, 0.4, and 0.5 mg L⁻¹), showing linearity with coefficients of determination (R²) of 0.9993 and 0.9996, respectively. Reagent blanks consisting of ultrapure water acidified with 1 mol L⁻¹ HNO₃ were analyzed in triplicate, and limits of detection (LOD) and quantification (LOQ) were calculated as 3σ/slope and 10σ/slope, respectively. The obtained LOD/LOQ values were 0.18/0.60 mg L⁻¹ for Co and 0.01/0.03 mg L⁻¹ for Cd, and all sample measurements were within the validated linear range of the method.

2.4. Determination of Point of Zero Charge (pH_PZC) and Zeta Potential Distribution

The point of zero charge (PZC) was determined according to a previously established methodology. Briefly, 20 mg of the biosorbent were weighed into Falcon tubes containing 20 mL of a NaCl solution (0.01 mol L⁻¹) to maintain constant ionic strength, with initial pH values adjusted between 1 and 10. pH adjustments were performed using dilute 0.1 mol L⁻¹ HCl and NaOH solutions, with measurements taken using a pH meter (Metrohm, model 827, São Paulo, Brazil).

The samples were agitated for 24 h at room temperature and then centrifuged to separate the solid phase. The pH of the supernatant (pHf) was measured, and the difference from the initial pH (pHi) was calculated (ΔpH = pHf − pHi). The PZC was determined from a plot of pHf (y-axis) versus pHi (x-axis), corresponding to the point where the curve intersects the line pHf = pHi, indicating zero net surface charge.

Zeta potential measurements of the purple ipe leaf aqueous dispersions were conducted on a Zetasizer Nano ZS90 (Malvern Instruments Ltd., Malvern, UK). Dispersions were prepared in ultrapure water (η = 0.8872 cP, dielectric constant ε = 78.5) and measured at 25 °C (298.15 K). Measurements consisted of 12 consecutive runs per sample. The Smoluchowski model was applied for zeta potential calculation, with the medium conductivity maintained at 0.0279 mS/cm. The measured count rate was 183.3 kcps. Data were processed using the Zetasizer Software Suite v7.10 (Malvern Panalytical, Malvern, UK), and results are expressed as mean ± standard deviation.

2.5. Batch Adsorption Experiments

The adsorption capacity of purple ipe (Handroanthus impetiginosus) leaf powder for metal ion removal from aqueous solutions was investigated through a series of batch experiments. This methodology is standard for determining fundamental adsorption parameters, including equilibrium isotherms, kinetics, and thermodynamics.

Batch kinetic and pH edge experiments were conducted using solutions of Co(II) and Cd(II) ions. To assess the adsorption capacity, 20 mg of biosorbent was added to a 50 mL centrifuge tube containing 20 mL of a metal ion solution at an initial concentration of 50 mg L⁻¹. These specific solid/liquid ratio and concentration values were selected based on preliminary tests and prior literature to optimize the process.

The mixture was agitated using an axial shaker at 150 rpm for 1 h and maintained at a constant temperature of 26 ± 1 °C (299.15 K). Subsequently, the samples were centrifuged at 3000 rpm for 5 min (Centrifuge: Excelsa 2 Model 206 BL, Fanem, São Paulo, Brazil) to ensure efficient solid phase. The supernatant was then filtered through a quantitative ashless filter paper (Whatman Grade 41, 20–25 µm), which is widely used in analytical procedures involving metal ions due to its very low ash content (<0.007%) and minimal release or retention of inorganic species, ensuring that the filtration step does not interfere with subsequent metal quantification. The clarified supernatant was stored for analysis. Residual metal concentrations were determined by Flame Atomic Absorption Spectrometry (FAAS), with absorbance readings referenced against pre-established calibration curves.

The adsorption capacity at equilibrium, $q_e$ (mmol g⁻¹), was calculated according to Equation (1):

$$q_e={\displaystyle \frac{\left(C_0- C_e\right) \times V}{m M}}$$

(1)

where $C_0$ and $C_e$ are the initial and equilibrium concentration of the adsorbate in the solution (mg L⁻¹), respectively, $V$ is the volume of the solution (L), $M$ is the molar mass of the adsorbate (mg mmol⁻¹) and $m$ is the mass of the adsorbent (g).

The study systematically evaluated three critical adsorption parameters: contact time, initial pH, and maximum uptake capacity. Kinetic studies were performed by varying the contact time from 1 to 240 min. The influence of pH was examined across a range of 1.0 to 6.0 over a 1 h contact period. Adsorption isotherms were constructed using initial metal concentrations ranging from 1 to 400 mg L⁻¹, with a contact time of 1 h at the predetermined optimal pH. The optimal conditions for adsorption were identified by maximizing the equilibrium uptake capacity ($q_e$).

2.6. Continuous-Flow Column Adsorption

Continuous-flow adsorption experiments were performed using a recirculating system equipped with a peristaltic pump (Watson Marlow, São Paulo, Brazil) and Tygon tubing (Ø = 2.057 mm). A metal ion solution (50 mL, 0.2 mg L⁻¹) was percolated through the column at a flow rate of 1 mL min⁻¹ under constant agitation. After adsorption, the retained metal ions were eluted with 5 mL of eluent, and aliquots were collected for analysis.

The adsorption performance in continuous-flow experiments was evaluated based on enrichment factor and metal recovery, calculated according to Equations (2) and (3), respectively.

$$EF={\displaystyle \frac{C_e}{C_0}}$$

(2)

$$Recovery \left(\%\right)={\displaystyle \frac{C_e\times V_e}{C_0\times V_0}}\times 100$$

(3)

where $C_0$ is the influent metal concentration (mg L⁻¹), $V_0$ is the influent volume (mL), $C_e$ is the metal concentration measured in the eluate (mg L⁻¹) and $V_e$ is the eluent volume (mL).

2.7. Adsorption Models

Various kinetic models are employed to analyze experimental adsorption data, providing critical insight into the governing mechanisms of the process. As adsorption is a mass transfer operation, its rate and kinetics can be quantitatively determined. Among the most prevalent models are the pseudo-first order (PFO) model, originally developed by Lagergren XIX [17], which analyzes kinetics based on the solid’s retention capacity and suggests a physisorption mechanism, and the pseudo-second order (PSO) model, proposed by Ho and McKay [18], which models the adsorption rate based on adsorbate uptake at the adsorbent surface over time and implies a chemisorption mechanism. Equations (4) and (5) represent the non-linearized forms of the PFO and PSO models, respectively.

$$q_t=q_e(1-e^{-k1t})$$

(4)

$$q_t={\displaystyle \frac{k_2{q}_e^{2}t}{1+k_2{q}_{e}t}}$$

(5)

where $q_t$ (mmol g⁻¹) is the amount of adsorbate adsorbed at time $t$ (min), and $q_e$ (mmol g⁻¹) is the amount adsorbed at equilibrium. For Equation (2), $k_1$ (min⁻¹) is the PFO rate constant and in Equation (3), $k_2$ (g mmol⁻¹ min⁻¹) is the PSO rate constant.

The equilibrium adsorption data for both metals were analyzed using the Langmuir and Freundlich isotherm models to determine the maximum adsorption capacity. The Langmuir model postulates a homogeneous surface with identical, energetically equivalent adsorption sites, resulting in monolayer coverage. This theory assumes that adsorption occurs at specific sites without interaction between adsorbed species and with no possibility of multilayer formation [19]. The model is represented by its non-linearized form in Equation (6).

$$q_e={\displaystyle \frac{q_{max}{K}_L{C}_e}{1+K_L{C}_e}}$$

(6)

where $q_e$ is the amount adsorbed at equilibrium (mmol g⁻¹), $q_{max}$ is the maximum monolayer adsorption capacity (mmol g⁻¹), $C_e$ is the equilibrium concentration of adsorbate in solution (mg L⁻¹ or mmol L⁻¹) and $K_L$ is the Langmuir equilibrium constant (L mmol⁻¹).

The Freundlich model is an empirical equation describing adsorption on heterogeneous surfaces featuring sites with different adsorption energies, which facilitates multilayer adsorption [20]. It is expressed in non-linearized form in Equation (7).

$$q_e=K_F{C}_e^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$n$}\right.}$$

(7)

where $q_e$ is the amount adsorbed at equilibrium (mmol g⁻¹), $C_e$ is the equilibrium concentration in solution (mg L⁻¹ or mmol L⁻¹) and $K_F$ is the Freundlich constant related to adsorption capacity and $n$ is the heterogeneity factor (dimensionless). An $n$ value between 1 and 10 indicates a favorable adsorption process.

The kinetic and equilibrium adsorption data were fitted using non-linear regression. The goodness of fit of the models was evaluated using the root mean square error (RMSE), while model comparison was further supported by the Akaike information criterion (AIC) Equation (8) and the Bayesian information criterion (BIC) Equation (9) [21,22]. Lower values of RMSE, AIC, and BIC were considered indicative of better model performance.

$$AIC=-2\mathrm{l}\mathrm{n}\left(\widehat{L}\right)+2k$$

(8)

$$BIC=-2\mathrm{l}\mathrm{n}\left(\widehat{L}\right)+k\mathrm{l}\mathrm{n}\left(n\right)$$

(9)

where $\widehat{L}$ is the maximum value of the model’s likelihood function, $k$ is the number of fitted parameters and $n$ is the number of experimental data points.

The 95% confidence intervals (95% CI) of the fitted parameters were obtained from non-linear regression based on Student’s t distribution, providing an estimate of parameter uncertainty.

Thermodynamic parameters were assessed by calculating the standard Gibbs free energy change (ΔG°, J mol⁻¹) according to Equation (10) [23]:

$$\mathsf{\Delta }{G}^{°}=-RT\mathrm{l}\mathrm{n}K$$

(10)

where $R$ is the universal gas constant (8.314 J mol⁻¹ K⁻¹), $T$ is the absolute temperature (K), and $K$ is the equilibrium constant derived from the adsorption isotherms.

The Dubinin–Radushkevich (D–R) isotherm model was applied to describe the adsorption mechanism and estimate the mean adsorption energy [24]. The non-linear form of the D–R equation is expressed as Equation (11):

$$q_e=q_m\mathrm{e}\mathrm{x}\mathrm{p}(-\beta {\epsilon }^2)$$

(11)

where $q_e$ is the equilibrium adsorption capacity (mmol g⁻¹), $q_m$ is the theoretical saturation capacity (mmol g⁻¹), $C_e$ is the equilibrium concentration of adsorbate in solution (mg L⁻¹ or mmol L⁻¹), $\beta$ is a constant related to the adsorption energy (mol² kJ⁻²), and $\epsilon$ is the Polanyi potential (kJ mol⁻¹), defined by Equation (12):

$$\epsilon =RT\mathrm{l}\mathrm{n}\left(1+{\displaystyle \frac{1}{C_e}}\right)$$

(12)

3. Results

3.1. Characterization

Fourier-transform infrared (FTIR) spectroscopy was applied to identify characteristic bands linked to the functional groups existing in the structure of the purple ipe leaf. This analysis permitted inference of the presence of potential active sites acting as Lewis bases, which are favorable for the adsorption process. The interpretation of the spectra was based on tables of wavenumbers and their correspondences with functional groups described in existing literature. This allowed a comparison between our experimental values and data previously reported for biosorbents of a similar nature (Figure 1). A broad and intense band is observed at 3356 cm⁻¹, attributed to the vibrational stretching of the hydroxyl group (–OH) present in the cellulose’s structure. The band at 2916 cm⁻¹ is corresponding to the symmetric vibrational stretching of C–H bonds. The band found at 1646 cm⁻¹ indicates carbonyl groups (C=O), which are characteristic of aldehydes, ketones, and carboxylic acids. It is also associated with primary amides present in proteins, possibly due to the presence of peptide bonds. The band at 1315 cm⁻¹ is related to the axial deformation of –CN groups, attributed to amines and amides. Finally, the band at 1060 cm⁻¹ is assigned to the C–O stretching of alcohols present in cellulose and lignin, as well as to the elongation of the C–O–C bond, which is typical for hemicellulose and cellulose structures [25].

Nuclear Magnetic Resonance (NMR) spectroscopy was next applied to assist in the determination of the functional groups constituting the sample’s chemical architecture (Figure 2). The obtained ¹H–¹³C CP/MAS NMR spectra for the extracts of purple ipe (Handroanthus impetiginosus) leaves presented distinct chemical signatures across several spectral regions. Resonances identified within the 100 to 70 ppm range are typically ascribed to carbohydrate-derived carbon atoms, notably those characteristics of cellulose and hemicellulose structures [26]. The grouping of signals observed between 50 and 80 ppm is furthermore indicative of sp³-hybridized carbons engaged in a covalent bond with oxygen. Finally, the appearance of signals in spectral regions close to 30 ppm suggests the presence of characteristic CH₂ and CH moieties. This specific chemical shift allows for several interpretation possibilities, with potential assignments including, but not being limited to, carbon atoms involved in S–C and N–C bonds.

A comprehensive characterization of the biosorbent material is a prerequisite for the interpretation of its metal adsorption performance. Consequently, the surface morphology and physical properties of the purple ipe leaf biomass were investigated using a suite of analytical techniques.

Analysis by Scanning Electron Microscopy (SEM) provided micrographs at various levels of magnification (Figure 3). These images reveal a material possessing a highly irregular surface topography, composed of particles with a broad size distribution. This observed heterogeneity indicates a clear absence of morphological uniformity. The analysis further confirmed that, within the selected 106–250 μm fraction, the average particle size was 200 μm and, importantly, the material exhibited a notably porous structure throughout.

Energy Dispersive X-ray Spectroscopy (EDS) analysis was employed to complement and corroborate the findings obtained through other methodologies described in this study. This technique can identify elements present in amounts above about 2 wt%. Quantification was performed by acquiring spectra from three distinct regions on the sample surface, from which the arithmetic means of the atomic concentrations of carbon and oxygen were calculated. It is recognized that EDS is not the optimal technique for the precise quantification of low atomic number elements such as carbon and oxygen [27]. Their ubiquitous presence as atmospheric contaminants, coupled with the inherent limitations of the spectrometer’s vacuum level, can introduce measurable uncertainty into the analysis. The resulting spectrum exhibited characteristic peaks for carbon and oxygen. The peak at approximately 300 eV is assigned to carbon, originating from the emission of electrons from the K-shell (1 s transition). Similarly, the peak observed at 530 eV is attributed to oxygen, also resulting from K-shell electron emission (1 s transition) [28], as summarized in

Table 1. Values referring to the mass percentage (% m/m) of the elements present in the purple ipe leaf at a 20 kV spectrum and 500× magnification.

Chemical Elements (K 1 s Layer)	Mass Concentration (%)
C	54.58
O	33.40
Al	0.13
Si	0.21
Ca	2.87
Co	0.07
Au	8.75

and Figure 4.

However, the analysis of nitrogen and sulfur, elements which are frequently indicative of functional groups serving as active adsorption sites, could not be performed by EDS. This inability is likely a consequence of their low concentration in the sample, potentially falling below the method’s detection threshold. To address this, specific elemental analysis for nitrogen and sulfur was conducted. The resulting data, presented in

Table 2. Elemental analysis of nitrogen and sulfur in purple ipe leaves.

Nitrogen (g kg−1)	Sulfur (g kg−1)
19	0.8

, correspond to the total bulk content of these elements within the material. It is important to note that these values represent the entire quantity present, and not specifically the fraction participating in adsorption processes.

The point of zero charge (pH_PZC) of the purple ipe leaf biosorbent was found to be 5.8 (Figure 5). This parameter corresponds to the pH at which the biosorbent surface has no net electrical charge, and it provides insight into how the surface charge varies with changes in solution pH. It is important to note, however, that the pH_PZC represents an average surface condition and does not necessarily reflect the availability of specific active sites or the aqueous speciation of the adsorbate. Even so, knowledge of the pH_PZC is essential for interpreting and optimizing electrostatic interactions between the biosorbent surface and the adsorbate species.

The red line and the dot in the figure indicate the pH at the point of zero charge (pH_PZC), which separates the pH regions where the surface charge changes. At pH values above the pH_PZC, the surface becomes predominantly negatively charged, favoring the adsorption of cationic species through electrostatic interactions. At pH values below the pH_PZC, the surface is positively charged, and adsorption is hindered due to competition between metal cations and protons (H⁺), originating from hydronium ions (H₃O⁺), for the active sites. The presence of H⁺ in the figure represents this proton competition, which decreases as pH increases. The optimal adsorption observed at pH 5–6 results from a balance between reduced proton competition and favorable metal speciation, prior to hydrolysis or complexation at higher pH values [29].

Figure 6 and

Table 3. Electrokinetic parameters of purple ipe leaves.

Parameter	Value
Zeta potential (mV)	−14.7 ± 8.89
Dominant peak	−14.7 (100% area)
Conductivity (mS/cm)	0.0279 ± 0.0005
Count rate	183.3

present the zeta potential distribution for the purple ipe leaf extract, which was characterized by a monomodal distribution with a mean value of −14.7 ± 8.89 mV. The analysis indicated a predominance of negative surface charges on the particles. The high count rate observed (183.3 kcps), coupled with a low standard deviation, suggests a satisfactory homogeneity of the sample in suspension.

3.2. Batch Adsorption Experiments

Batch experiments were conducted in triplicate to investigate the influence of pH and adsorption kinetics. For these studies, 20 mg of purple ipe leaf biomass and 20 mL of each metal ion solution at a fixed concentration of 50 mg L⁻¹ were employed, as these parameters were established as optimal based on existing literature. The effect of pH on adsorption efficiency was evaluated over a pH range from 1 to 6. The pH was adjusted using 0.1 mol L⁻¹ and 2 mol L⁻¹ HNO₃ (Labsynth, Brazil) and 0.1 mol L⁻¹ and 1 mol L⁻¹ NaOH (Merck, Germany), and measurements were performed using a pH meter (Metrohm, model 827).

As shown in Figure 7, the adsorption capacity ($q_e$) of both Co²⁺ and Cd²⁺ ions increased with increasing pH, reaching maximum values in the pH range between 5 and 6. For Cd²⁺, the adsorption capacity observed at pH 4 was also close to the maximum value, indicating an earlier stabilization of adsorption compared to Co²⁺. The experimental equilibrium adsorption capacities as a function of pH are summarized in

Table 4. Experimental equilibrium adsorption capacity of Co(II) and Cd (II) as a function of pH. Values are reported as mean ± standard deviation (n = 3).

pH	Co(II)	Cd(II)
	${\mathit{q}}_{\mathit{e}}$ (mmol g−1)	${\mathit{q}}_{\mathit{e}}$ (mmol g−1)
1	0.04627 ± 0.00312	0.00404 ± 0.00024
2	0.26556 ± 0.01850	0.04587 ± 0.03480
3	0.28191 ± 0.02240	0.10158 ± 0.00894
4	0.47029 ± 0.03580	0.20445 ± 0.01630
5	0.50130 ± 0.04020	0.20808 ± 0.01810
6	0.50415 ± 0.04260	0.20706 ± 0.01940

, while the global statistical metrics obtained from the sigmoidal model fitting are presented in

Table 5. Global statistical metrics for the sigmoidal model fitting.

Metrics	Co(II)	Cd(II)
RMSE	0.0416	0.01035
AIC	−32.15	−48.85
BIC	−32.78	−49.48

.

The experimental $q_e$ values as a function of pH were further fitted using a sigmoidal model, which adequately describes the observed trend. The quality of the model fitting was evaluated using the root mean square error (RMSE) and the Akaike (AIC) and Bayesian (BIC) information criteria, confirming good agreement between experimental and theoretical values for both metal ions.

Kinetic studies were conducted a fixed pH of 6, utilizing 20 mg of the biosorbent and 20 mL of a metal ion solution at a concentration of 50 mg L⁻¹. The contact time varied from 1 to 240 min. The results are presented in Figure 8 and

Table 6. Adsorption capacity of Co(II) and Cd(II) at different contact times. Values are reported as mean ± standard deviation (n = 3).

Time (min)	Co(II)	Cd(II)
	${\mathbf{q}}_{\mathbf{t}}$ (mmol g−1)	${\mathbf{q}}_{\mathbf{t}}$ (mmol g−1)
1	0.304 ± 0.111	0.149 ± 0.010
5	0.348 ± 0.028	0.192 ± 0.012
15	0.378 ± 0.016	0.206 ± 0.015
30	0.527 ± 0.002	0.216 ± 0.022
45	0.530 ± 0.002	0.213 ± 0.010
60	0.539 ± 0.001	0.213 ± 0.009
90	0.535 ± 0.031	0.214 ± 0.010
120	0.527 ± 0.015	0.212 ± 0.004
180	0.528 ± 0.017	0.207 ± 0.019
240	0.528 ± 0.015	0.206 ± 0.006

.

As shown in Figure 8, the purple ipe leaf biomaterial exhibited rapid adsorption kinetics during the initial minutes of the experiment. The highest adsorption rate was observed within the first 5 min. The minimum contact time required to achieve effective adsorption was 30 min for both metal ions studied, indicating similar interaction kinetics between them.

Following the construction of the adsorption capacity plots, the adsorption kinetics were evaluated using the pseudo-first order (PFO) and pseudo-second order (PSO) models, and the corresponding kinetic parameters are summarized in

Table 7. Parameters obtained from PFO and PSO kinetic models for the purple ipe leaf for Co(II) and Cd(II).

Model	Specie	R2	${\mathbf{q}}_{\mathbf{e}}$ (mmol g−1)	95% CI $({\mathbf{q}}_{\mathbf{e}}$)	k	95% CI (k)	RMSE	AIC	BIC
Pseudo-First Order	Co(II)	0.443	0.496	0.448–0.545	0.82 min−1	0.157–1.48	0.065	−54.6	−54.0
	Cd(II)	0.886	0.209	0.198–0.219	1.25 min−1	0.843–1.651	0.006	−101.1	−100.5
Pseudo-Second Order	Co(II)	0.695	0.520	0.479–0.561	1.65 g mmol−1 min−1	0.338–2.958	0.048	−60.6	−60.0
	Cd(II)	0.961	0.212	0.204–0.220	11.02 g mmol−1 min−1	7.86–14.18	0.004	−111.9	−111.3

. Among the evaluated models, the PSO model provided a better statistical description of the experimental data, as indicated by higher non-linear R² values, lower RMSE, and reduced AIC and BIC values compared to the PFO model. In addition, the equilibrium adsorption capacities predicted by the PSO model were in closer agreement with the experimental values. These results suggest that the PSO model is more consistent with the observed kinetic behavior under the studied conditions, without implying a definitive adsorption mechanism.

The maximum biosorption capacity was determined by conducting a study as a function of the initial concentration of Co²⁺ and Cd²⁺ ions. The experimental conditions, namely a minimum contact time of 30 min and a pH of 6, were selected based on previous studies to optimize the adsorptive process efficiency. A concentration range of 1–400 mg L⁻¹ was established, with the upper limit selected to prevent analyte precipitation and thus ensure the occurrence of adsorption. The experimental data are presented in Figure 9, while the equilibrium adsorption capacities of Co(II) e Cd(II) at different equilibrium concentrations are summarized in

Table 8. Equilibrium adsorption capacity of Co(II) and Cd(II) at different equilibrium concentrations ($C_e$). Values are reported as mean ± standard deviation (n = 3).

${\mathit{C}}_{\mathit{e}}$ (mg L−1)	Co(II)	Cd(II)
	${\mathit{q}}_{\mathit{e}}$ (mmol g−1)	${\mathit{q}}_{\mathit{e}}$ (mmol g−1)
1	0.016 ± 0.00023	0.00552 ± 0.00009
5	0.057 ± 0.00119	0.0565 ± 0.00054
10	0.099 ± 0.00347	0.091 ± 0.00066
25	0.223 ± 0.0128	0.152 ± 0.00302
50	0.360 ± 0.0207	0.238 ± 0.0162
75	0.526 ± 0.0516	0.274 ± 0.0292
100	0.649 ± 0.0737	0.225 ± 0.0292
200	0.756 ± 0.0859	0.264 ± 0.0520
300	0.763 ± 0.0856	0.261 ± 0.0400
400	0.711 ± 0.0892	0.275 ± 0.0334

.

The adsorption curves for both metal ions exhibit similar behavior in the initial phase, demonstrating a direct proportional relationship between the concentration of Co²⁺/Cd²⁺ and the adsorption capacity until a saturation point is reached. At this equilibrium plateau, the maximum adsorption capacities for the purple ipe biosorbent were determined to be 0.823 mmol g⁻¹ and 0.270 mmol g⁻¹ for Co(II) and Cd(II), respectively.

Subsequently, the experimental data for the maximum adsorption capacity, together with the kinetic data, were analyzed using the Langmuir and Freundlich isotherm models, and the corresponding parameters are summarized in

Table 9. Parameters obtained through the Langmuir and Freundlich equations for the purple ipe leaf with the Co(II) ion.

Model	q Theoretical (mmol g−1)	IC 95% (q)	R2	K	IC 95% (K)	n	IC 95% (n)	RMSE	AIC	BIC
Langmuir	$q_{max}$ = 0.823	0.791–0.855	0.997	$K_L$ = 3.20 L mmol−1	2.66–3.73	—	—	0.0170	−77.40	−76.8
Freundlich	$q_{exp}$ = 0.422	—	0.926	$K_F$ = 0.511 mmol g−1(L mmol−1)1/n	0.433–0.590	3.38	2.22–4.54	0.080	−46.60	−46.0

and

Table 10. Parameters obtained through the Langmuir and Freundlich equations for the purple ipe leaf with the Cd(II) ion.

Model	q Theoretical (mmol g−1)	IC 95% (q)	R2	K	IC 95% (K)	n	IC 95% (n)	RMSE	AIC	BIC
Langmuir	$q_{max}$ = 0.270	0.256–0.284	0.980	$K_L$ = 50.9 L mmol−1	32.9–68.9	—	—	0.0138	−55.31	−52.71
Freundlich	$q_{exp}$ = 0.276	—	0.910	$K_F$ = 0.055 mmol g−1(L mmol−1)1/n	0.04–0.07	5.31	3.6–7.0	0.0292	−38.32	−37.72

. The Langmuir model assumes monolayer adsorption on a homogeneous surface with energetically equivalent sites and no lateral interactions between adsorbed species, whereas the Freundlich model is an empirical equation commonly applied to heterogeneous surfaces with a non-uniform distribution of adsorption energies. It should be emphasized, however, that a direct association of the Langmuir model with chemisorption and the Freundlich model with physisorption is an oversimplification, as these models do not mechanistically discriminate between adsorption types. Instead, the quality of fit to these models should be interpreted as being consistent with specific surface and adsorption characteristics, rather than as definitive evidence of the underlying adsorption mechanism. Therefore, the isotherm models are used primarily as mathematical tools to describe adsorption behavior under the studied conditions [30,31,32].

The comparative analysis of the parameters presented in Table 9 and Table 10 highlights differences in the fitting performance of the Langmuir and Freundlich models for the adsorption of the evaluated metal ions. For both Co(II) and Cd(II), the Langmuir model exhibited higher maximum adsorption capacities ($q_{max}$), with relatively narrow confidence intervals, as well as higher Langmuir equilibrium constants ($K_L$), indicating a better statistical description of the experimental data within the studied concentration range. The corresponding Gibbs free energy changes (ΔG°,

Table 11. Gibbs free energy change (ΔG°) for the biosorption of Co(II) and Cd(II) onto purple ipe leaf at 299.15 K *.

Metal Ion	${\mathit{K}}_{\mathit{L}}$	ΔG°
	(L mmol−1)	(kJ mol−1)
Co(II)	3.20	−20.0
Cd(II)	50.9	−27.0

* ΔG° was estimated using the Langmuir equilibrium constant at a single temperature (299.15 K); therefore, ΔH° and ΔS° could not be determined.

), estimated from the Langmuir equilibrium constants, were negative for both ions and became more negative in the order Co(II) < Cd(II). This trend, derived from monometallic systems, suggests that the adsorption process is thermodynamically favorable and hints at a potentially higher intrinsic affinity for Cd(II) over Co(II). Although the Freundlich model exhibited n values greater than 1, which are commonly interpreted as indicative of favorable adsorption behavior, it was also characterized by higher error values, as reflected by increased RMSE, along with less negative AIC and BIC values. The coefficients of determination were likewise higher for the Langmuir model; however, these metrics were considered together with the other statistical indicators, which collectively suggest a better overall representation of the experimental data by this model under the studied conditions.

In order to further investigate the adsorption mechanism, the Dubinin–Radushkevich (D–R) isotherm model was applied to the equilibrium adsorption data of Co(II) and Cd(II). This model allows the estimation of the mean free energy of adsorption (E), which provides insight into the nature of the adsorption process. In addition, the standard Gibbs free energy change was evaluated using the equilibrium constants obtained from the Langmuir isotherm, enabling assessment of the spontaneity of the adsorption process. The calculated thermodynamic and D–R parameters are summarized in Table 11 and

Table 12. Dubinin–Radushkevich isotherm parameters for Co(II) and Cd(II) adsorption.

Metal Ion	${\mathit{q}}_{\mathit{m}}$	β	ε
	(mmol g−1)	(mol2 kJ−2)	(kJ mol−1)
Co(II)	0.730	207.0	0.049
Cd(II)	0.254	22.25	0.150

.

To complement the batch adsorption and thermodynamic analyses, the applicability of Handroanthus impetiginosus leaves was examined in a continuous-flow column system. As shown in Figure 10 and

Table 13. Metal concentration in the eluate, recovery, and enrichment factor obtained during continuous-flow column experiments using purple ipe leaf.

Column	Eluate Volume (mL)	${\mathit{C}}_{\mathit{e}}$ (mg L−1)	Enrichment Factor	Recovery (%)
1	5	1.3697	6.85	68.5
2	5	1.2770	6.39	63.9
3	5	1.1450	5.73	57.3

, effective metal preconcentration was achieved, with eluate concentrations ranging from 1.145 to 1.3697 mg L⁻¹ over three consecutive column runs, indicating consistent retention and elution behavior. The corresponding recovery values varied from 57.3 to 68.5%, while enrichment factors between 5.73 and 6.85 were obtained. A slight decrease in concentration, recovery, and enrichment factor was observed along the runs, suggesting a gradual reduction in adsorption efficiency during consecutive column operation.

4. Discussion

4.1. Material Characterization

Energy-dispersive X-ray spectroscopy (EDS) and Fourier-transform infrared spectroscopy (FTIR) were employed to complement the sample characterization. As indicated in Table 1, the sample exhibited predominantly high values for carbon and oxygen, a result consistent with the material’s chemical composition, which includes cellulose and hemicellulose, molecules rich in these elements. Furthermore, Figure 1 and Figure 2 confirm the presence of functional groups that are important for the adsorption process, particularly those with the potential to act as active sites, such as oxygen- and nitrogen-containing groups [33,34]. These data reinforce the chemical characterization of the sample, confirming the expected presence of these constituents, and are in agreement with the literature, which highlights the relevance of such groups in the adsorption of metal ions.

Considering that purple ipe leaf was used for solid-phase extraction in the batch-test method, factors such as particle size and surface area are fundamental to the process. In this context, scanning electron microscopy (SEM) served as an essential tool for the detailed investigation of the sample’s structure, texture, and roughness. The micrographs obtained at different magnifications (Figure 3) complement the understanding of the material’s properties. The material was observed to possess a heterogeneous surface with a considerable quantity of pores, which are regularly identifiable in fibrous plant materials [35,36].

To further characterize the surface composition of the biosorbent, energy-dispersive X-ray spectroscopy (EDS) analysis was performed in conjunction with SEM (Figure 4). The EDS spectrum confirmed the presence of the main inorganic elements typically associated with plant-based materials. The gold (Au) signal observed arises from the deposition of a thin gold layer during sample preparation for SEM analysis, a necessary step to prevent surface charging in non-conductive samples. In addition, cobalt (Co) was detected at a very low concentration, which is consistent with its natural occurrence as a trace micronutrient in plant tissues.

Regarding the elemental analysis, which was conducted to complement data not acquired by EDS, the presence of nitrogen (N) and sulfur (S) was verified. These are fundamental elements for adsorption, as they are frequently associated with functional groups that act as active sites. These sites, such as amino (–NH₂) and thiol (–SH) groups, exhibit a high capacity for coordinating metal ions [37], forming stable complexes through electron pair donation or electrostatic interactions, which explains the material’s efficiency in metal uptake. The results indicated a significantly higher concentration of nitrogen compared to sulfur. However, as presented in

Table 14. Comparative analysis between different adsorbent surfaces with respect to elemental analysis.

Adsorbent	Nitrogen (g kg−1)	Sulfur (g kg−1)	Reference
Brachyura shell	19.0	0.9	[38]
TABOA	13.0	0.8	[38]
Activated carbon functionalized with melamine	24.8	—	[39]
Cassava root husks powder	10.2	6.0	[40]
Purple ipe leaf	19.0	0.8	*

* Present work.

, the purple ipe leaf demonstrates a balanced distribution of these elements compared to other adsorbents reported in the literature.

The determination of the point of zero charge (${\mathrm{p}\mathrm{H}}_{\mathrm{P}\mathrm{Z}\mathrm{C}}$) under controlled ionic strength conditions resulted in a value of 5.8, as shown in Figure 5, corresponding to the pH at which the biosorbent surface exhibits a zero average net charge. The adsorption experiments indicated that the highest metal uptake occurred in the pH range of 5–6, within the investigated pH interval. These results demonstrate a clear dependence of adsorption performance on solution pH and confirm the effectiveness of the purple ipe leaf biosorbent for metal retention in aqueous systems. Additionally, the adsorption efficiency observed within the pH 5–6 range supports the applicability of this material for real wastewater treatment, whose pH commonly falls within this interval.

Regarding the obtained average zeta potential value of −14.7 mV, it is situated within the range of moderate electrostatic stability (−10 to −20 mV) [41]. This indicates that the particles exhibit a certain degree of mutual repulsion, which contributes to dispersion stability. However, this level of stability may be insufficient for long-term applications, and the adoption of strategies such as pH adjustment or the addition of surfactants is recommended. The predominance of negative surface charge is likely attributable to the presence of phenolic or carboxylic groups on the sample surface, which is characteristic of plant extracts rich in bioactive compounds [42,43,44]. The high homogeneity of the dispersion, evidenced by low data variability and a high count rate, reinforces the quality of the obtained colloidal system.

4.2. Experimental Tests

The effect of pH on the adsorption of Co²⁺ and Cd²⁺ ions onto purple ipe leaf biomass is shown in Figure 7. For both metal ions, the equilibrium adsorption capacity (q_e) increased with increasing pH, indicating a strong dependence of the adsorption process on solution acidity. Low $q_e$ values were observed under acidic conditions (pH 1–2), while a marked increase in adsorption occurred between pH 3 and 5, followed by a plateau at higher pH values. The experimental $q_e$ data as a function of pH were fitted using a sigmoidal model, which adequately described the observed trend for both Co²⁺ and Cd²⁺ ions. The quality of the model fit was confirmed by low RMSE values and negative AIC and BIC values. The 95% confidence intervals for the experimental q_e values, calculated from the triplicate measurements, are provided in the Supplementary Material (Tables S1 and S2).

The kinetic profiles show a more pronounced adsorption during the initial stages of the process, which is commonly attributed to the greater availability of accessible surface sites on the biosorbent. Raw concentration data for Co(II) and Cd(II) obtained during the batch kinetic experiments, measured in triplicate at each contact time, are provided in the Supplementary Material (Tables S3 and S4). Kinetic modeling suggested that the pseudo-second order (PSO) model provided a better statistical description of the experimental data, particularly for Cd²⁺, as reflected by higher R² values, lower RMSE, AIC, and BIC values, and a closer agreement between estimated and experimental $q_e$ values (Table 7). These results indicate differences in the kinetic representation of the data between the two ions under monometallic conditions, without implying preferential adsorption or selectivity. The robustness of the equilibrium adsorption capacity ($q_e$) estimates was further evaluated using a bootstrap approach, and the corresponding 95% confidence intervals (percentile method, 20.000 resamples) for Co(II) and Cd(II) are reported in the Supplementary Material (Tables S5 and S6). Consistent trends were also observed in the adsorption isotherm analysis, in which the Langmuir model provided a more consistent overall fit to the experimental data for both ions, as indicated by lower RMSE, AIC, and BIC values and well-defined fitted parameters within their confidence intervals (Table 9 and Table 10). Raw concentration and adsorption capacity data obtained from the batch isotherm experiments, measured in triplicate at each equilibrium concentration, are available in the Supplementary Material (Tables S7 and S8). Overall, these results suggest adsorption behavior consistent with the predominance of energetically similar surface sites and limited multilayer contributions within the investigated concentration range. However, these interpretations are based on model fitting and therefore do not constitute definitive evidence of a specific adsorption mechanism or of selectivity between Co²⁺ and Cd²⁺.

Although kinetic and spectroscopic analyses suggest the involvement of chemisorption, the Dubinin–Radushkevich (D–R) isotherm (Table 12) revealed low mean free adsorption energies for both Co(II) and Cd(II), which is often associated with physisorption [45]. These observations are consistent with a process where physical interactions may dominate the overall equilibrium, while potential chemical interactions such as surface complexation and ion exchange could occur at specific active sites. This interpretation of multimodal behavior is typical of biosorbents and reflects the heterogeneous nature of their surface functional groups. In line with the D–R results, the magnitude of the Gibbs free energy change also supports the presence of spontaneous but energetically moderate interactions. The ΔG° value in Table 11 for Co(II) (−20.0 kJ mol⁻¹) indicates a favorable and spontaneous adsorption process. Its magnitude falls within a range often consistent with a contribution from both physical and ion-exchange interactions in biosorption systems. In contrast, the more negative ΔG° value observed for Cd(II) (−27.0 kJ mol⁻¹) suggests a stronger apparent affinity toward the purple ipe leaf biomass under the studied monometallic conditions, in agreement with its higher Langmuir constant [46]. This difference may be attributed to variations in ionic radius, hydration energy, and binding preference toward oxygen-containing functional groups. Overall, the thermodynamic results corroborate the isotherm analysis and support the interpretation of a spontaneous and favorable biosorption of both metal ions.

The differences observed in the $q_e$ values between Co²⁺ and Cd²⁺ ions can be partially attributed to distinctions in their structural characteristics, particularly their ionic radii. Ions with larger radii, such as Cd²⁺, tend to exhibit adsorptive behavior different from that of smaller ions, such as Co²⁺, especially during the initial stages of the process. Additionally, the greater number of electron shells associated with Cd²⁺, resulting from its higher atomic number, may influence its affinity for the biosorbent surface [47].

Differences in the apparent adsorbate–adsorbent interactions between Co²⁺ and Cd²⁺ may be further discussed in the context of Pearson’s hard and soft acids and bases (HSAB) concept. Within this framework, Co²⁺ is generally regarded as an intermediate Lewis acid, whereas Cd²⁺ is classified as a soft acid, which may influence their interactions with surface functional groups. Under the monometallic conditions investigated, the higher equilibrium adsorption capacities observed for Co²⁺ could be related to its classification as an intermediate acid, which may favor a higher saturation of the available amine sites on the biosorbent, whose nitrogen atoms act as Lewis bases and are often reported to interact more favorably with intermediate acids than with soft acids [48]. However, these differences reflect variations in adsorption capacity and apparent affinity under single-metal systems and should not be interpreted as selectivity, which would require evaluation under competitive adsorption systems.

The adsorption of Co²⁺ and Cd²⁺ onto the purple ipe leaf biosorbent can also be discussed in terms of possible contributions from interactions between the metal ions and surface functional groups. The presence of oxygen- and nitrogen-containing groups identified by FTIR suggests the potential involvement of coordination-type interactions, while the negative surface charge indicated by the zeta potential likely facilitates the initial electrostatic attraction of the cations. In addition, the porous morphology observed by SEM provides accessible sites that may support adsorption. The rapid initial uptake, together with the improved statistical description obtained using the pseudo-second order kinetic model and the Langmuir isotherm, is supportive of an adsorption behavior predominantly influenced by surface-site interactions and limited multilayer contributions within the investigated concentration range. Nevertheless, these interpretations are based on model fitting and indirect physicochemical evidence and should therefore be regarded as suggestive rather than confirmatory of a complexation-driven chemisorption mechanism. The arrows in Figure 11 represent a schematic illustration of the interactions between metal ions and the available electron-donating sites on the biosorbent surface. This figure is intended solely as a conceptual representation of the proposed adsorption pathway and does not depict a detailed mechanistic process.

In summary, the purple ipe leaf biosorbent exhibited a higher adsorption capacity for Co(II), reaching 0.823 mmol g⁻¹, compared to 0.270 mmol g⁻¹ for Cd(II) under the investigated monometallic conditions. These values are considered significant when compared with those reported for other biosorbents in the literature (

Table 15. Capacity of different adsorbents and biosorbents in retaining the metal ions Co(II) and Cd(II).

	Adsorption Capacity (mmol g−1)
Adsorbent	Co(II)	Cd(II)	Reference
Polyurethane foam	0.011	–	[49]
Blackberry leaf	0.195	–	[50]
Indian Chrysanthemum Flower	0.251	–	[51]
Serratia marcescens	0.337	–	[52]
Lemon peel	0.373	–	[53]
Biomass from rose waste	0.460	–	[54]
Ulva fasciata	0.678	–	[55]
Orchid Tree	–	0.113	[56]
Cassava root husks powder	–	0.140	[57]
Modified carbon foam	–	0.197	[58]
Freshwater mussel shell	–	0.230	[59]
Activated Carbon with Melamine	–	0.279	[39]
PVC-AHTT	–	0.320	[60]
Brachyura carapace	–	0.430	[38]
Purple ipe leaf	0.823	0.270	*

* Present work.

).

As observed in Table 15, the purple ipe leaf demonstrated high efficiency in the removal of Co(II), with performance comparable to that of marine algae such as Ulva fasciata and Colpomenia spp. [61]. Although algae-based biosorbents are known for their adsorption capacity, their large-scale application may be impacted by high production costs, the need for controlled cultivation, and logistical challenges [62,63]. In contrast, the purple ipe leaf may represent a potentially accessible alternative because it can be obtained from pruning residues, reducing the need for dedicated cultivation.

However, it is important to note that the results discussed in this study were obtained under controlled laboratory conditions and focused on monometallic systems. The influence of competing ions, increased matrix complexity, and long-term performance under repeated adsorption–desorption cycles was not investigated and may affect the applicability of the biosorbent in more complex scenarios. Within these limitations, the adsorption behavior observed in batch experiments and the enrichment and recovery achieved under continuous-flow conditions are consistent with the physicochemical characteristics of the material. In particular, the ability to concentrate metal ions from dilute solutions into a small eluent volume under dynamic conditions supports the potential application of the purple ipe leaf biomass not only as an adsorbent but also as a preconcentration material. These findings reinforce the relevance of exploring underutilized biomaterials as low-cost and potentially lower-impact alternatives for environmental applications.

4.3. Limitations and Suggestions for Future Research Tests

This study is subject to certain limitations. Specifically, the availability of Handroanthus impetiginosus leaves is seasonal due to natural abscission, a phenomenon typical of tropical and subtropical climates, which may restrict their continuous use. Nevertheless, despite this limitation, the leaves demonstrated promising potential as a natural biosorbent, particularly for complementary or targeted treatment systems. Future studies focused on chemical modification and process optimization may help mitigate these constraints.

Beyond the techniques employed in the present study, the investigation of optimized desorption–elution conditions may further improve metal recovery and contribute to the assessment of biosorbent regeneration. In addition, the application of flow-through systems under a broader range of operational conditions could provide further insight into the performance of the material under dynamic regimes relevant to practical scenarios. The implementation of such approaches may strengthen the evaluation of the biosorbent’s applicability in environmental processes.

5. Conclusions

The results suggest that purple ipe (Handroanthus impetiginosus) leaves exhibit physicochemical characteristics favorable for metal adsorption, including the presence of oxygen- and nitrogen-containing functional groups, a porous surface, and adequate electrostatic stability. The biosorbent showed effective adsorption performance for Co(II) and Cd(II) under the investigated monometallic conditions, which is consistent with the possible contribution of Lewis acid–base interactions involving functional groups within the lignocellulosic matrix, without implying a definitive adsorption mechanism. Under the experimental conditions evaluated, the adsorption and preconcentration performance of the material was comparable to that reported for several conventional adsorbents described in the literature.

Although the material offers advantages related to availability and sustainability, additional studies addressing regeneration, long-term performance, and behavior in more complex aqueous matrices are required to better assess its practical applicability. Within these limitations, purple ipe leaves may be considered a promising and eco-friendly biosorbent for the adsorption and preconcentration of metal ions from aqueous systems.