Capture Target Definition
Translate the industrial stream into a computational specification.
- Feed, humidity, and trace components
- Purity, recovery, and removal targets
- PSA, VSA, TSA, or equilibrium context
AI for Carbon Capture and Molecular Separation
AI for Carbon Capture Materials
Move beyond dry-gas uptake rankings. CD ComputaBio screens porous and functional adsorbents against the composition, pressure swing, humidity, regeneration limits, and product targets that define your real carbon-capture decision.
Stream specificMaterial broadProcess aware
Service coverage
Carbon Capture Services
Material Library Curation
Build a computation-ready set with chemistry, pore geometry, and structure-quality filters.
- Experimental and hypothetical structures
- Charge and force-field readiness
- Diversity-preserving down-selection
Adsorption Screening
Estimate where and how strongly CO2 binds across the operating window.
- Henry coefficients and heats of adsorption
- GCMC isotherms and working capacity
- Binding-site and pore analysis
Mixture and Moisture Analysis
Challenge selectivity with competing gases and water before expensive validation.
- CO2/N2 and CO2/CH4 mixtures
- Humidity and impurity scenarios
- Competitive adsorption risk
Process-Aware Ranking
Combine molecular outputs with regeneration, density, kinetics, stability, and form-factor constraints.
- Energy–productivity trade-offs
- Pellet and bed assumptions
- Decision-ready candidate dossier
Candidate landscape
Materials We Screen
Candidate families are selected for the target stream and development pathway, rather than forced into a universal database.
Metal–Organic Frameworks
Tunable pores, open metal sites, and functional linkers.
Covalent Organic Frameworks
Lightweight networks with modular topology and chemistry.
Zeolites
Robust inorganic frameworks for demanding cyclic conditions.
Porous Polymers
Functional networks and amine sorbents for dilute capture.
Carbons and Hybrids
Activated carbons, supported amines, and composites.
Decision model
What We Rank
No single adsorption number defines a carbon-capture material. The ranking model can be adjusted for capture cost, product purity, recovery, cycle productivity, moisture tolerance, or regeneration energy. We rank performance across the loading–release cycle and expose the trade-offs that control the next decision.
The console is illustrative and does not represent measured material results.
Ranking dimensionsAdjustable model
Working capacity
CYCLE
Mixture selectivity
FEED
Water tolerance
RISK
Regeneration load
ENERGY
Transport
RATE
Form-factor fit
BED
Project workflow
From Stream to Shortlist
Each stage produces a reviewable decision gate before more expensive calculations or experiments are commissioned.
Frame the Stream
Define feed composition, humidity, pressure, temperature, product target, cycle concept, and material exclusions.
Curate Candidates
Standardize structures, remove inaccessible duplicates, calculate descriptors, and preserve chemical diversity.
Simulate and Learn
Combine adsorption simulation with AI triage, uncertainty review, and structure–performance insight.
Select for Process
Re-rank with working capacity, regeneration, kinetics, stability, and process assumptions; define validation.
Project package
Inputs and Deliverables
Recommended Inputs
- Feed composition, impurities, and humidity range
- Adsorption/desorption pressure or temperature window
- Target CO2 purity, recovery, or removal specification
- Preferred families, exclusions, and available experimental data
Quality Controls
- Structure provenance and readiness checks
- Force-field and charge-method documentation
- Scenario-specific ranking and sensitivity analysis
- Uncertainty, missing-property, and stability flags
01 / DATACurated material library
02 / RESULTSAdsorption and mixture dataset
03 / DECISIONProcess-aware shortlist
04 / NEXT STEPValidation roadmap
From prediction to measurement
Experimental Validation Support
Computational screening is most valuable when its predictions are translated into focused measurements. CD ComputaBio helps define the experimental evidence needed to confirm adsorption performance, challenge material stability, and refine the candidate ranking.
Validation plans are matched to the uncertainty that remains after screening, rather than applying the same experimental package to every material.
Plan Your Validation Strategy01
Confirm equilibrium behavior
Adsorption Isotherms and Selectivity
Define pressure, temperature, and gas-composition conditions for measuring CO2 uptake, working capacity, heat of adsorption, and competitive adsorption.
- Single-component and mixture isotherms
- Low-pressure or high-pressure measurement windows
- IAST comparison where the assumptions are appropriate
02
Challenge the real feed
Breakthrough and Humidity Testing
Translate predicted competition effects into dynamic tests that examine separation under realistic flow, humidity, and impurity conditions.
- Binary or multicomponent breakthrough design
- Dry-versus-humid performance comparison
- Trace-gas and feed-variation scenarios
03
Verify material robustness
Structure and Stability Characterization
Identify whether adsorption, moisture, heating, or repeated cycling changes the framework, pore environment, functional groups, or accessible capacity.
- PXRD, TGA, spectroscopy, and porosity measurements
- Water, thermal, and chemical stability assessment
- Pre- and post-cycling material comparison
04
Connect powder to process
Cycling, Shaping, and Process-Relevant Tests
Evaluate whether a promising intrinsic material retains useful performance after regeneration cycles, pelletization, binder addition, and bed-level implementation.
- Repeated adsorption–desorption cycling
- Pellet, binder, and bulk-density effects
- Pressure drop, kinetics, and productivity implications
05
Learn from measured results
Model Updating and Candidate Re-Ranking
Integrate measured isotherms, breakthrough curves, cycling data, and stability observations to update assumptions, recalibrate predictive models, and refine the shortlist.
- Prediction-versus-experiment comparison
- Uncertainty reduction and sensitivity review
- Updated ranking and next-experiment recommendation
Published data
Evidence Behind the Workflow
Data mining identified water-tolerant CO2 binding motifs
1Screen the landscapeMine more than 300,000 MOF structures.
2Extract binding motifsIdentify adsorbaphores linked to selective CO2 binding.
3Challenge with waterPrioritize hydrophobic motifs and water-stable frameworks.
Boyd and colleagues connected high-throughput screening, motif discovery, and experimental synthesis, highlighting how water competition changes the credible shortlist.[1]
View publicationPellet and cycle variables changed material performance
1Generate adsorption dataCalculate CO2 and N2 isotherms with GCMC.
2Model VSA operationConnect molecular properties to a process model.
3Optimize the form factorEvaluate pellets, cycle variables, and Pareto trade-offs.
Farmahini and colleagues showed that intrinsic adsorption properties alone do not fix a ranking; pellet structure and process configuration can change predicted energy and productivity.[2]
View publicationPractical guidance
Common Questions
Scope, fidelity, and simulation cost should follow the stream and the decision.
Can you screen a proprietary material library?
Yes. Client structures, compositions, and measured adsorption data can support a project-specific library under the agreed data-handling and confidentiality framework.
Why is dry CO2/N2 selectivity not enough?
Water, trace gases, pressure-dependent loading, desorption, and cycle conditions can change usable working capacity and selectivity.
Which simulation methods are used?
Projects may combine geometric descriptors, machine learning, GCMC, molecular dynamics, selected quantum calculations, IAST where appropriate, and process models.
Can measured data be included?
Yes. Isotherms, breakthrough curves, cycling, and stability data can calibrate predictions and identify the next most valuable measurement.
What is included in the final shortlist?
Ranked candidates, scoring logic, scenario sensitivity, property gaps, adsorption outputs, process implications, and recommended validation steps.
Start with the Stream
Share the feed composition, capture target, operating window, and available material data. We will define a focused screening strategy and the evidence needed to select a credible adsorbent.
Scientific References
- Boyd, P. G., Chidambaram, A., García-Díez, E., et al. Data-driven design of metal–organic frameworks for wet flue gas CO2 capture. Nature 576, 253–256 (2019). https://doi.org/10.1038/s41586-019-1798-7
- Farmahini, A. H., Friedrich, D., Brandani, S., and Sarkisov, L. Exploring new sources of efficiency in process-driven materials screening for post-combustion carbon capture. Energy & Environmental Science 13, 1018–1037 (2020). https://doi.org/10.1039/C9EE03977E
