Functionalized Nano-Hydroxyapatite with Chitosan and Glutamic Acid: A Versatile Nano-Adsorbent for Cd(II) Removal from Aqueous Media and a Selective Antifungal Agent

INTRODUCTION
Water pollution from heavy metals remains one of the more stubborn problems facing public health, particularly in regions where industrial growth has outpaced wastewater regulation [1–4]. Among the elements typically flagged as hazardous, cadmium (Cd) sits in a difficult position: it is highly toxic even at sub-ppm levels, accumulates in tissues over time and reaches humans through drinking water and the food chain [5–9]. The World Health Organization has set the permissible limit for Cd(II) in drinking water at 0.005 mg L⁻¹, a value that demands rather efficient removal strategies [10]. Several techniques are commonly used – chemical precipitation, ion exchange, membrane filtration, electrodeposition and adsorption – yet adsorption is generally favoured for being inexpensive, scalable and relatively clean [11,12]. A wide variety of solid sorbents (zeolites, clays, biomass, activated carbon and polymers) have been explored, but capacity and selectivity are often limited [13–17].
Hydroxyapatite, Ca₁₀(PO₄)₆(OH)₂, is attractive because of its low solubility, chemical stability and high affinity for di- and trivalent cations [18–19]. Reducing it to the nano-scale further improves its sinterability and reactive surface area [20]. Chitosan, in turn, brings a dense population of –NH₂ and –OH groups that chelate metal ions and disrupt microbial membranes [21–23], while acidic amino acids such as glutamic and aspartic acid alter the crystallinity of HAp through electrostatic and stereochemical effects [24–27]. Combining the three should, in principle, yield a low-crystallinity, highly reactive nano-composite. Building on the n-HAp–Cs–Glu material we previously reported [28], the current work examines its performance as a Cd(II) nano-adsorbent across a range of operational conditions and assesses its antimicrobial behaviour against two bacterial and two fungal isolates.

MATERIALS AND METHODS
Materials
The n-HAp–Cs–Glu nano-composite powder was the same batch we synthesised and partially characterised in our earlier work [29]. Cadmium chloride hemipentahydrate (CdCl₂·2.5H₂O) was supplied by Hangzhou Hyper Chemicals (China). Hydrochloric acid (35–38 %) was obtained from Alpha Chemika (India) and sodium hydroxide from Fluka. All chemicals were of analytical grade and were used as received. Solutions were prepared with deionised water throughout.

Synthesis of the n-HAp–Cs–Glu nano-composite
The synthesis is summarised in Fig. 1 and is briefly described here for clarity. A CaCl₂ solution was first mixed with chitosan (2 %) and the pH adjusted to 10 with NH₄OH. In parallel, glutamic acid (0.2 M, prepared in 1.3 M ammonium acetate) was heated to 80 °C and brought to pH 10. A Na₂HPO₄ solution at the same pH was then added drop-wise to the glutamic mixture under magnetic stirring. The CaCl₂/chitosan stream was subsequently introduced drop-wise into the reactor; pH was checked and re-adjusted if needed. Ageing was carried out at 80 °C and pH 10 for 5 h. The white slurry was washed three times with deionised water followed by absolute ethanol, separated by centrifugation, dried at 37°C and kept in a sealed vial. A pristine n-HAp control sample was prepared by repeating the protocol without chitosan or glutamic acid [30].

Additional characterisation
In addition to the techniques reported earlier [31], three complementary methods were applied here. AFM imaging was carried out on a NaioAFM (Nanosurf AG, Switzerland) to obtain 2D and 3D surface views and the mean-diameter distribution of the particles. Elemental composition and elemental mapping were obtained on a Thermo Scientific Axia ChemiSEM diffractometer (EDS). Thermal stability was probed on an HZ2329 TG analyser, with simultaneous TGA, DTG and DSC traces recorded under an air flow at 20 °C min⁻¹.

Adsorption procedure
Stock Cd(II) solutions were prepared by dissolving CdCl₂·2.5H₂O in deionised water and freshly diluted before each run. Batch experiments were performed in 250 mL glass beakers in which 50 mL of the Cd(II) solution was contacted with the n-HAp–Cs–Glu powder under magnetic stirring at 200 rpm. After equilibrium, the suspension was centrifuged at 4000 rpm for 5 min and the residual Cd(II) in the supernatant measured by flame atomic absorption spectroscopy (PG Instruments, Japan; λ = 228.80 nm). The variables tested were the adsorbent dose (0.1–0.2 g), contact time (15–90 min), initial Cd(II) concentration (20–100 mg L⁻¹), pH (3–9; adjusted with 0.1 M HCl or NaOH) and temperature (25–45 °C). Removal efficiency (%) and equilibrium uptake qe (mg g⁻¹) were calculated from Eqs. 1 and 2:

R % = (C₀ − Ce) × 100 / C₀                                        (1)

qe = (C₀ − Ce) V / m                                                  (2)

where C₀ and Ce are the initial and equilibrium Cd(II) concentrations (mg L⁻¹), V the solution volume (L) and m the mass of adsorbent (g) [32].

Antimicrobial assays
Antibacterial and antifungal screening was performed by the agar well diffusion method on Mueller-Hinton (bacteria) and Sabouraud-dextrose (fungi) agar [33]. Each microbial inoculum was prepared from a single fresh colony suspended in sterile saline. The lawn was spread with a sterile swab and 6 mm wells were punched in the agar. A 100 µL aliquot of n-HAp–Cs–Glu suspension at 250, 500, 750 or 1000 µg mL⁻¹ (in DMSO) was loaded into each well. DMSO alone served as the negative control. Bacterial plates were incubated at 37 °C for 18–24 h and fungal plates for 24–48 h. Inhibition-zone diameters were measured in millimetres.

RESULTS AND DISCUSSION
Characterisation of the nano-composite
XRD patterns of n-HAp and n-HAp–Cs–Glu (Fig. 2a) show the characteristic apatite reflections, the most intense at 2θ ≈ 31.7°. After functionalization the peaks broaden and lose intensity, a clear sign of reduced crystallinity. Applying the Scherrer equation to the (211) reflection gave mean crystallite sizes of 21.91 nm for n-HAp and 12.52 nm for n-HAp–Cs–Glu, confirming that chitosan and glutamic acid act as growth-inhibitors during precipitation [34,35]. FESEM micrographs (Fig. 2b and c) show a rough surface with loose aggregates for n-HAp and a clearly more porous, plate-like network for the functionalized composite – useful for adsorption since it raises the available surface area. TEM imaging (Fig. 2d and e) backs this up: the parent n-HAp displays uniform, needle-shaped crystallites whereas n-HAp–Cs–Glu forms flower-like aggregates with a more complex morphology. This kind of morphological evolution is typical when amino acids are present during apatite crystallization [36,37].
AFM imaging supported these observations. The 2D and 3D scans (Fig. 3a and b) revealed a granular, hilly surface with a mean particle diameter of 60.45 nm and a maximum height of 16.48 nm; the size distribution histogram (Fig. 3c) is heavily skewed toward the small-to-medium fraction. EDS analysis (Fig. 3d) detected C, N, O, P and Ca only – an internally clean spectrum without spurious peaks. The corresponding atomic percentages were 28.9 % C, 7.4 % N, 46.5 % O, 5.8 % P and 11.4 % Ca. The Ca/P ratio sat above 1.67, consistent with stoichiometric apatite carrying additional surface-bound Ca²⁺ coordinated by carboxyl and amine groups from glutamic acid and chitosan, in line with prior reports [38]. Thermal analysis (Fig. 3e) showed four mass-loss steps between 280 and 550 °C totalling roughly 14 wt %, with the main DTG/DSC peak near 292 °C corresponding to depolymerization of chitosan and decomposition of the glutamic acid moiety. Above 600 °C the residue was effectively constant, confirming the high thermal stability of the apatite framework [39,40].

Adsorption studies
Effect of contact time and kinetic modelling
Fig. 4a shows that Cd(II) uptake rose sharply during the first 60 min and reached a plateau of 42.93 mg g⁻¹ at 75 min; no further change was seen at 90 min, so 75 min was used as the equilibrium contact time in subsequent runs. Such fast initial uptake is expected: a freshly contacted surface has many vacant binding sites, and these saturate as the adsorbed layer grows [41,42].
To probe the underlying kinetics, both pseudo-first-order (Eq. 3) and pseudo-second-order (Eq. 4) models were fitted [43,44]:

ln(qe − qt) = ln qe − k₁ t                                            (3)

t / qt = 1/(k₂ qe²) + (1/qe) t                                     (4)

The pseudo-first-order plot (Fig. 4b) was clearly poorer (R² = 0.84) and yielded a calculated qe of only 15.25 mg g⁻¹ – very different from the experimental value, which is itself a strong indication that the model does not apply here. The pseudo-second-order plot (Fig. 4c), in contrast, was almost perfectly linear (R² = 0.9998), with a calculated qe of 43.86 mg g⁻¹ that essentially matches the experimental plateau. Together these results imply that chemisorption – involving the sharing or exchange of electrons between Cd²⁺ and surface functional groups – is the rate-controlling step (Table 1).

Effect of adsorbent dose and pH
Increasing the n-HAp–Cs–Glu mass from 0.1 to 0.2 g raised the Cd(II) removal modestly from 85.9 % to 89.0 %, while the equilibrium uptake fell from 42.9 to 22.3 mg g⁻¹ (Fig. 5a). The trend is the usual one: adding more sorbent introduces more binding sites and pushes removal up, but at the same time many of those sites stay unoccupied because the available Cd(II) is fixed, so the per-gram uptake drops [47-45].
The solution pH had a stronger influence (Fig. 5b). At pH 3, qe was only 34.7 mg g⁻¹; raising pH to 9 increased it to 49.0 mg g⁻¹. At low pH, H⁺ ions compete with Cd²⁺ for adsorption sites and protonate the chitosan amine groups, both of which weaken Cd(II) binding [48,49]. As pH rises, this competition fades and the surface acquires a net negative character that favours cation uptake [50]. The catch is that at pH > 8, Cd(OH)₂ begins to precipitate, and a fraction of the apparent removal is no longer adsorption. To stay within a regime where adsorption dominates, pH 7 was selected for all subsequent experiments [51].

Effect of initial Cd(II) concentration and temperature
With m = 0.1 g, t = 75 min and pH 7, the Cd(II) concentration was varied between 20 and 100 mg L⁻¹ at three temperatures (Fig. 6). At 25 °C the removal efficiency dropped from 98.9 % at 20 mg L⁻¹ to 87.9 % at 100 mg L⁻¹ (Fig. 6a). At low concentrations there are easily enough vacant sites to capture most of the metal, whereas at high concentrations the fixed amount of sorbent saturates and the residual fraction in solution grows [52]. The opposite trend was seen for qe, which climbed from 9.89 to 43.94 mg g⁻¹ (Fig. 6b). A higher driving force across the solution-solid boundary explains this behaviour[53].
Heating the system from 25 to 45 °C lifted both removal efficiency and qe across the entire concentration range. At C₀ = 100 mg L⁻¹, qe rose from 43.94 to 45.33 mg g⁻¹, indicating an endothermic process [54,55]. The effect was most pronounced at the highest concentrations, where saturation effects were already starting to bite at 25 °C; warming gave the system enough thermal energy to populate additional, slightly less favourable sites.

Adsorption isotherms
Equilibrium data at 25, 35 and 45 °C were modelled using the Langmuir (Eq. 5) and Freundlich (Eq. 7) equations [56–58,60,61]:

Ce/qe = 1/(qm kL) + Ce/qm                                    (5)

RL = 1 / (1 + kL C₀)                                                    (6)

ln qe = (1/n) ln Ce + ln KF                                        (7)

Fig. 7a and b show the linearised Langmuir and Freundlich plots respectively. Across the three temperatures the Langmuir fit was consistently the better of the two, with R² values in the 0.974–0.994 range against 0.959–0.971 for Freundlich (Table 2). qm rose modestly from 46.30 to 48.54 mg g⁻¹ between 25 and 45 °C, again pointing to an endothermic interaction. The separation factor RL stayed below 0.04 in all cases (0 < RL < 1), confirming that the adsorption is favourable [56]. The Freundlich exponent n was always between 2.6 and 2.9, also indicating a favourable process [51]; the slight decrease in n with temperature suggests a mild increase in surface heterogeneity, perhaps as previously inaccessible sites become available at higher T.
Fig. 7c presents the van’t Hoff plots used to extract the thermodynamic parameters from Eqs. 8–10 [53]:

K = CAe / Ce                                                               (8)

ΔG° = −RT ln K                                                           (9)

ln K = −ΔH°/(RT) + ΔS°/R                                       (10)

Across all initial concentrations and temperatures, ΔG° was negative and grew more negative with T – the adsorption is spontaneous and grows more spontaneous on heating . ΔH° was positive in every case, confirming the endothermic nature suggested by the equilibrium runs, while positive ΔS° values point to increased randomness at the solid–solution interface, possibly because hydration shells around Cd²⁺ are partially shed as the cation binds to the surface [53]. The values for the 60 mg L⁻¹ run sit somewhat outside the trend of the other concentrations (ΔH° ≈ 31.8 kJ mol⁻¹), and the corresponding van’t Hoff R² is lower (0.917). With only three temperature points per concentration, this kind of scatter is not unusual and does not change the qualitative picture, but a fourth temperature would be helpful in future work.

Antimicrobial activity
The well-diffusion test produced an interesting and somewhat split picture. At every concentration tested (250–1000 µg mL⁻¹), the n-HAp–Cs–Glu nano-composite gave no measurable inhibition zone against either S. aureus (Gram-positive) or E. coli (Gram-negative); see Fig. 8a and b. The same suspensions, however, did produce clear zones around the wells loaded onto plates inoculated with C. tropicalis and C. albicans (Fig. 8c and d). The largest zone, 8.5 mm, was recorded for C. albicans at 1000 µg mL⁻¹, while C. tropicalis reached 6 mm at the same dose (Table 4).
This selective response is consistent with what has been reported for other chitosan-functionalised inorganic carriers [21,22,62]. Bacterial envelopes – the thick peptidoglycan of Gram-positive cells and the lipopolysaccharide outer membrane of Gram-negative cells – act as steric and electrostatic barriers that limit the penetration of macromolecules. Fungal cell walls, dominated by β-glucan and chitin, are more permissive; the chitosan moiety on the composite surface is also chemically similar to chitin, which probably aids adsorption onto the wall and the subsequent disruption of membrane integrity [63,21]. We note that the dose–response is not strictly monotonic for either yeast (e.g. C. tropicalis gave 4 mm at both 750 and 250 µg mL⁻¹), which suggests that experimental variability between wells and the well-diffusion technique itself contribute meaningfully to the spread; this is a known limitation of the method and a future study using broth microdilution would put more accurate MIC values on these numbers[59,64].

CONCLUSION
A nano-hydroxyapatite functionalised with chitosan and glutamic acid was prepared by wet precipitation and tested for cadmium removal and antimicrobial activity. AFM, EDS and TGA confirmed the nano-scale character (mean particle diameter ≈ 60 nm), the elemental integrity (Ca/P > 1.67) and a thermally robust apatite framework. Cd(II) adsorption fitted the pseudo-second-order kinetic model (R² = 0.9998) and the Langmuir isotherm (qm = 48.54 mg g⁻¹ at 45 °C), while RL values below unity confirmed favourable uptake. Thermodynamics were spontaneous (ΔG° < 0), endothermic (ΔH° > 0) and entropy-driven (ΔS° > 0). On the biological side, the same nano-composite acted as a selective antifungal agent against C. albicans and C. tropicalis but produced no detec activity against S. aureus or E. coli. Taken together, the results identify n-HAp–Cs–Glu as an inexpensive, environmentally benign nano-adsorbent with a built-in antifungal property – a useful combination for water-treatment applications where biofouling by yeasts is also a concern.

CONFLICT OF INTEREST
The authors declare that there is no conflict of interests regarding the publication of this manuscript.