Hyperconfined bio-inspired Polymers in Integrative Flow-Through Systems for Highly Selective Removal of Heavy Metal Ions

Abstract

Access to clean water, hygiene, and sanitation is becoming an increasingly pressing global demand, particularly owing to rapid population growth and urbanization. Phytoremediation utilizes a highly conserved phytochelatin in plants, which captures hazardous heavy metal ions from aquatic environments and sequesters them in vacuoles. Herein, we report the design of phytochelatin-inspired copolymers containing carboxylate and thiolate moieties. Titration calorimetry results indicate that the coexistence of both moieties is essential for the excellent Cd²⁺ ion-capturing capacity of the copolymers. The obtained dissociation constant, K_D ~ 1 nM for Cd²⁺ ion, is four-to-five orders of magnitude higher than that for peptides mimicking the sequence of endogenous phytochelatin. Furthermore, infrared and nuclear magnetic resonance spectroscopy results unravel the mechanism underlying complex formation at the molecular level. The grafting of 0.1 g bio-inspired copolymers onto silica microparticles and cellulose membranes helps concentrate the copolymer-coated microparticles in ≈3 mL volume to remove Cd²⁺ ions from 0.3 L of water within 1 h to the drinking water level (<0.03 µM). The obtained results suggest that hyperconfinement of bio-inspired polymers in flow-through systems can be applied for the highly selective removal of harmful contaminants from the environmental water.

Fig. 1. Design of copolymer (pAA–Cys5) inspired by plant phytochelatin.

a Schematic of the process of phytoremediation occurring in plant cells. The phytochelatin forms complexes with heavy metal ions such as Cd²⁺ to sequester the ions in vacuoles. b Chemical structure of the designed phytochelatin-inspired synthetic polymers (pAA–Cys5).

Fig. 2. Selective capture of Cd²⁺ ions by bio-inspired pAA-Cys5.

a Plots of additional thermal power (dQ/dt) and enthalpy (ΔH) versus the molar ratio of the cysteine side chain by titrating pAA–Cys5 with CdCl₂ and CaCl₂, respectively. The best-fit curves for ITC data using a two-site model are shown (solid lines). b ATR-IR spectra obtained by subtracting the spectrum of a buffer from that for pAA–Cys5 (4 w/v%) in the absence (black) and presence of 8 mM Ca²⁺ (blue) or Cd²⁺ (red). Inset: magnified view of the peak top for ν_COO– (antisymmetric) band. c ATR-IR spectra obtained by subtracting the spectrum of pAA–Cys5 (4 w/v%) from that recorded for pAA–Cys5 (4 w/v%) in the presence of Ca²⁺ or Cd²⁺ (2, 4, 6, and 8 mM). d Peak intensity plotted against the concentration of Ca²⁺ or Cd²⁺ for (c). The data obtained in the presence of Cd²⁺ and Ca²⁺ are indicated in red and blue, respectively. Error bars corresponds to the instrumental noise determined from the baseline. e ¹H NMR spectra of NAC (1 mM) in the absence and presence of 1 mM of Ca²⁺ or Cd²⁺. f Possible structures of the Cd²⁺–pAA–Cys5 complex deduced from the NMR and FTIR spectra.

Fig. 3. pAA-Cys5 is selective adsorbent for Cd²⁺ ions with high loading capacity.

a Schematic for the experimental procedure. Only the ions that did not bind to pAA–Cys5 can pass through the filter because the ions bound to pAA–Cys5 polymer (M_w ≈ 1.7 ×  10⁴Da) cannot pass the filter. b [Cd²⁺] in the flow-through in the absence (light gray) and presence (red) of pAA–Cys5 (0.1 mg mL^–1). The corresponding data for pHPMA–Cys5 (gray) and pPEGMA–Cys5 (gray) are shown for comparison. The initial concentration of Cd²⁺ ([Cd]_initial) was 100 µM. Error bars indicate the calibration accuracy of Spectroquant® colorimetric assay. c Table showing the initial concentrations of pAA–Cys5 and Cd²⁺ ([pAA–Cys5] and [Cd]_initial, respectively) and Cd²⁺ in the flow-through ([Cd]_flow-through). These values were used to calculate the amount of Cd²⁺ bound to 1 g of pAA–Cys5, whose maximum is 5.2 mmol g^–1. (d) Concentrations of Cd²⁺ in the flow-through after mixing simulated wastewater ([Cd²⁺]_initial = 10 µM, [Na⁺]_initial = 1 mM, [K⁺]_initial = 0.2 mM, [Mg²⁺]_initial = 0.5 mM, [Ca²⁺]_initial = 0.5 mM) with 10⁻⁴, 10⁻³, 10⁻², or 10^–1 mg mL^–1 of pAA–Cys5, confirming an outstanding Cd²⁺ selectivity. Error bars indicate the calibration accuracy of Spectroquant® colorimetric assay. Note that [Cd²⁺] for the WHO’s drinking water standard (0.03 µM) corresponds to 0.003 ppm.

Fig. 4. Hyperconfinement of pAA-Cys5 into column-based microreactor prototype.

a Functionalization of silica microparticles (diameter: 10 µm). The homogeneity of the surface coating was confirmed by the confocal fluorescence image of silica microparticles coated with fluorescently labeled pAA–Cys5–biotin (Fluor–pAA–Cys5–biotin, Supplementary Fig. 21). b The flow-through microreactor prototype based on an FPLC column that enables the confinement of silica microparticles (surface area of 1.1 m²) functionalized with 0.6 µmol of pAA–Cys5 in a volume of 1.8 mL. c [Cd²⁺] of the eluent plotted as a function of the fraction number. Inset: magnified plots near the onset. Error bars for x- and y-axes indicate the fraction volume and the calibration accuracy of Spectroquant® colorimetric assay, respectively. d Total amount of removed Cd²⁺ plotted as a function of elution volume. [Cd²⁺] satisfied the concentration criteria recommended by WHO for drinking water.

Fig. 5. High-throughput, integrative water treatment system based on pAA–Cys5–functionalized membrane and particles.

Combination of (a) silica microparticles (diameter: 1.2 µm) functionalized with pAA–Cys5 and (b) cellulose membrane coated with chitosan functionalized with pAA–Cys5. The surface of silanized silica microparticles was functionalized with a lipid monolayer incorporating 2 mol% of pAA–Cys5–DOPE. c Photograph of the device consisting of the microparticles and a membrane functionalized with pAA–Cys5. d [Cd²⁺] in the eluent plotted as a function of elution volume. The unmodified cellulose membrane and silica microparticles without polymers showed a significant leakage from the very first fraction (light gray). The cellulose/chitosan–g–pAA–Cys5 membrane showed a leakage of Cd²⁺ already at V = 60 mL (gray). In contrast, the combination of the membrane and the pAA–Cys5-functionalized microparticles (black) could remove 10 µM Cd²⁺ (typical industrial wastewater level) from 0.3 L in 1 h down to the level approved for drinking water. Error bars for x- and y-axes indicate the fraction volume and the calibration accuracy of Spectroquant® colorimetric assay, respectively.