Mechanistic study of pH effect on organic solvent nanofiltration using carboxylated covalent organic framework as a modeling and experimental platform
Summary¶
This wiki entry documents the same DOI and science as [[2022duong-separation-a-mechanistic-study]] but points at an alternate corpus PDF (papers/Phuoc_Sep_Purif_Tech_COF_2022.pdf). The publication investigates pH-dependent organic solvent nanofiltration through a carboxylated COF selective layer using methanol feeds with HCl, neutral, and NaOH conditions. ReaxFF MD supplies molecular interpretations for methanol self-diffusion, ion pairing, pore metrics, and electrostatic shifts of carboxylate groups, while bench experiments report permeance and dye rejection for Alcian Blue and related probes.
Methods¶
MD application (ReaxFF)¶
- Engine / code: ReaxFF molecular dynamics on periodic C-COF slab supercells in methanol electrolytes with HCl / NaOH concentration sweeps (Section 2.3 in
pdf_path). - System & composition: C-COF + methanol + electrolyte ions with HCl/NaOH pH sweeps; full atom counts in
pdf_path. - PBC / boundaries: Periodic in-plane PBC slab supercells as in
pdf_path. - Ensemble, timestep, duration, thermostat, barostat: 0.25 fs integration; NPT equilibration at 300 K and 0 atm reference pressure with Berendsen thermostat and barostat (100 fs / 1000 fs damping per the protocol); NVT production RMD 200 ps in the documented protocol. Self-diffusion and pore/dye solvation analyses:
pdf_path. - Shear, electric field, MSST, umbrella: N/A unless the PDF/SI states otherwise. ReaxFF Coulomb and QEq schedules:
pdf_path.
Experiments (OSN, rejection, permeance)¶
Membranes pair a carboxylated COF selective layer with an anodized aluminum oxide support (~800 nm selective thickness, ~20 nm support pores) following prior synthesis routes cited in the paper. Dead-end filtration at 1 bar transmembrane pressure and 25 °C on 2.5 cm discs, discarding initial permeate volumes (10–15 mL) before sampling. Permeance per the article; rejection via UV–visible dye assays. For figure-stable protocol locators, prefer [[2022duong-separation-a-mechanistic-study]] (same DOI, alternate PDF path on this page).
Force-field training in this work¶
- N/A for a de novo ReaxFF training paper: the work applies a published ReaxFF for the COF/methanol/ion chemistry, as cited in
pdf_path.
Findings¶
Methanol permeance decreases when either acid or base is added relative to less perturbed feeds, while Alcian Blue rejection increases from about 23% at pH 2.2 to about 98% at pH 10.1. Atomistic models show only modest changes in average pore dimensions, implicating solvated ion structures and membrane charge as dominant knobs: larger methanol–ion clusters retard diffusion, and deprotonated carboxylates strengthen dye–framework interactions that boost selectivity. For stable locators and figure alignment, prefer the primary curation on [[2022duong-separation-a-mechanistic-study]] unless this PDF path is required for manifest provenance. Side-by-side comparison of the two PDFs is recommended after major corpus updates because OCR or compression differences can shift pagination even when scientific content matches. When citing transport coefficients derived from MD, note that finite system sizes and short production windows affect error bars on diffusivity, so the paper’s experimental permeance remains the ground-truth check for engineering predictions.
Limitations¶
ReaxFF charges and sampling length constrain quantitative pH behavior in organic media; experiments use simplified dye cocktails versus industrial feeds.
Relevance to group¶
Adri C. T. van Duin is co-author; ReaxFF MD supplies molecular interpretation for COF OSN pH effects.