Fine-Tuning Porous Structure of Zirconium-Based Metal-Organic Frameworks for Efficient Separation and Purification of Astaxanthin by Defect Engineering

Abstract

Efficient separation of bioactive compounds from nature source, particularly that of astaxanthin (AXT), remains challenging due to their low content in complicated matrix and readily degradable structure. Herein, a modulator-induced defect engineering is presented on the stable zirconium-based metal-organic frameworks (Zr-MOFs) to optimize pore size and pore chemistry for the efficient separation and purification of AXT for the first time. High adsorption capacity of 26.21 mg g^-1 is achieved on the best-performing defect Zr-MOF (d-UiO-67-4), superior over the other reported adsorbent for AXT. Meanwhile, d-UiO-67-4 exhibits the selective adsorption of AXT over other carotenoids analogues with similar structure and properties. This is attributed to the preferential non-covalent interactions between defect framework and AXT revealed by the spectroscopy analysis and density functional theory (DFT) calculations. High purity of AXT with 89.0% ± 2.3% extraction efficiency can be realized after the purification of AXT by d-UiO-67-4. The practical separation performance of d-UiO-67-4 for AXT extracted from Haematococcus pluvialis is demonstrated by fixed-bed column-based dynamic adsorption and desorption experiments. This work broadens the preparation methods for thermosensitive active substances and provided new research ideas for the controlled adsorption of functional food factors.

Keywords: astaxanthin; metal–organic frameworks; purification; separation.

Figure 1

a) Schematic illustration of defective‐free UiO‐67 and d‐UiO‐67‐X; PXRD patterns b) and FT‐IR spectra c) of UiO‐67 and d‐UiO‐67‐X.

Figure 2

¹H NMR spectra of UiO‐67 and d‐UiO‐67‐X.

Figure 3

TGA curves of UiO‐67 and d‐UiO‐67‐X a–d); N₂ sorption isotherms at 77 K e) and Pore size distribution f) of d‐UiO‐67‐X.

Figure 4

a) The time‐dependent adsorption capacity curves, b) Pseudo‐first‐order model fitting, c) intra‐particle diffusion model fitting, d) AXT concentration dependent adsorption capacity curves, and e) Freundlich isotherm fitting of UiO‐67 and d‐UiO‐67‐X; f) Adsorption selectivity of d‐UiO‐67‐4.

Figure 5

DFT calculated binding sites and IGMH isosurface maps of AXT a,b), fucoxanthin c,d) and β‐carotene e,f) on optimized defective UiO‐67 cluster model.

Figure 6

a) Desorption rate of d‐UiO‐67‐X by different eluent solvents. b) Desorption rate of d‐UiO‐67‐X by different ratio of ethanol‐water mixture. c) HPLC spectra of AXT standard, AXT purification by d‐UiO‐67‐X, crude AXT after saponification and solvent extracted crude AXT with the same concentration (4 mg L⁻¹).

Figure 7

a) Breakthrough curves of fixed‐bed adsorption and b) desorption for d‐UiO‐67‐X and c) the correspond schematic diagram of fixed‐bed column apparatus.