Reticular and Network Design
Enumerate compatible monomers, connectivity patterns, topologies, linkage families, interpenetration states, and post-synthetic handles while enforcing valence, geometry, and precursor constraints.
Turn linker chemistry, network topology, pore environment, and operating conditions into a synthesis-aware shortlist for adsorption, separation, and capture.
Online InquiryPorous organic polymers (POPs) span crystalline covalent organic frameworks (COFs), porous aromatic frameworks (PAFs), conjugated microporous polymers (CMPs), hypercrosslinked polymers, and polymers of intrinsic microporosity. Their performance depends on more than nominal pore size: monomer geometry, linkage chemistry, network interpenetration, defects, framework flexibility, accessible functional groups, residual solvent, and water stability all reshape the adsorption landscape.
Within our AI for Porous Adsorbent Materials Screening and Design platform, we combine synthesis-aware structure generation, pore characterization, quantum-informed host–guest interactions, molecular simulation, interpretable machine learning, and process metrics. Candidates are ranked for a defined feed, pressure, temperature, humidity, cycle, and regeneration strategy—not by a single idealized uptake value.

Enumerate compatible monomers, connectivity patterns, topologies, linkage families, interpenetration states, and post-synthetic handles while enforcing valence, geometry, and precursor constraints.
Clean experimental or hypothetical structures; resolve disorder, missing hydrogen atoms, residual guests, cell consistency, accessible pore volume, and duplicate networks with auditable provenance.
Quantify pore-limiting diameter, largest cavity, accessible area and volume, dimensionality, void connectivity, functional-group density, electrostatic environment, and likely transport bottlenecks.
Use Henry-regime calculations, GCMC, mixture simulation, selectivity, working capacity, regenerability, heat of adsorption, and humidity competition at project-specific conditions.
Evaluate guest mobility, residence times, constrictions, framework flexibility, size exclusion, and diffusion selectivity using molecular dynamics or enhanced sampling where equilibrium metrics are insufficient.
Train applicability-domain-aware surrogate models, interpret linker and pore descriptors, quantify uncertainty, and use active learning to select high-information calculations or experiments.
| Stage | Key Activities | Decision Output |
|---|---|---|
| 1. Separation and process scoping | Define feed composition, impurities, water activity, temperature, pressure, cyclic operation, product specification, regeneration energy, shaping, and safety constraints. | Target metrics and boundary conditions |
| 2. Evidence and data audit | Review structures, isotherms, synthesis records, activation history, PXRD/BET data, force fields, assay conditions, missing metadata, and train–test leakage risks. | Fit-for-purpose data package |
| 3. Candidate generation and curation | Build or curate COF, PAF, CMP, HCP, or PIM candidates; enforce bonding and topology rules; remove duplicates and inaccessible or chemically implausible structures. | Traceable computation-ready library |
| 4. Multiscale property evaluation | Calculate pore descriptors, charges and interaction sites; run Henry, GCMC, MD, or DFT-informed calculations at relevant mixture and humidity conditions. | Adsorption, transport, and stability evidence |
| 5. AI ranking and synthesis triage | Build interpretable surrogate models, estimate uncertainty and applicability domain, balance performance with precursor access, reaction robustness, activation, and shaping constraints. | Pareto-ranked and synthesis-aware shortlist |
| 6. Validation and model update | Specify synthesis or procurement route, activation protocol, isotherm and breakthrough tests, humidity/stability checks, acceptance criteria, and active-learning updates. | Experimental validation plan |
Versioned structures, monomer and linkage annotations, topology, provenance, quality flags, duplicate handling, and simulation readiness.
Geometric descriptors, accessible channels, functional-group maps, electrostatic features, and structure–property interpretation.
Uptake, selectivity, working capacity, regenerability, heat of adsorption, and diffusion metrics across agreed feed conditions.
Pareto fronts, score definitions, uncertainty intervals, applicability-domain flags, sensitivity analysis, and alternative candidates.
Suggested precursor routes, linkage conditions, solvent and catalyst considerations, activation risks, shaping constraints, and characterization checkpoints.
Recommended isotherm, mixture breakthrough, cycling, moisture, stability, and transport experiments with data templates for model updates.
CO₂/N₂ and CO₂/H₂ screening across flue gas, pre-combustion, and direct-air-capture-relevant conditions, including water competition and regeneration.
CO₂/CH₄, H₂S/CH₄, water, and trace-contaminant removal with working capacity, product recovery, and cyclic stability in view.
Adsorbent or membrane candidate evaluation for H₂/CO₂, H₂/CH₄, and related mixtures using adsorption and diffusion evidence.
Functional-group and pore-size design for aromatic, chlorinated, polar, or low-concentration organic vapors under realistic humidity.
Hydrophobic, ionic, hydrogen-bonding, and size-selective pore environments for target uptake, selectivity, regeneration, and leaching-risk assessment.
CH₄ or H₂ deliverable capacity assessed between charge and discharge conditions, with density, thermal effects, and framework stability considered.
A published COFInformatics workflow integrates experimental and hypothetical COF libraries, calculated features, molecular simulation, and machine-learning models to screen natural-gas purification candidates. The study illustrates why candidate selection should combine adsorption metrics with structure descriptors and process conditions.1

1 Aksu, G. O.; Keskin, S. Rapid and Accurate Screening of the COF Space for Natural Gas Purification: COFInformatics. ACS Applied Materials & Interfaces 2024, 16, 19806–19818. https://doi.org/10.1021/acsami.4c01641. Distributed under Open Access license CC BY 4.0, with modification.
Every recommendation keeps the source structure, chemistry assumptions, simulation conditions, model version, uncertainty, and synthesis constraints visible. Share your target molecules, operating window, candidate structures, experimental isotherms, or monomer library through Contact Us or the Online Inquiry below, and we can define a staged screening and validation program.
Submit your project details below, and our team will respond within 24 hours.
Talk to our technical team about your project!
I Want To Talk