Feed and Product Specification
Define what must be removed, retained, or stored before screening begins.
- H2 fraction and impurity envelope
- Product purity and recovery target
- PSA, TSA, membrane, or storage context
AI for Carbon Capture and Molecular Separation
AI for Hydrogen Purification and Storage
Turn feed composition, hydrogen purity, pressure swing, temperature, and storage targets into a defensible adsorbent or membrane shortlist. CD ComputaBio combines molecular simulation, machine learning, transport analysis, and process-aware ranking for hydrogen-rich streams.
Syngas specificPurity awareStorage ready
Service coverage
Hydrogen Purification and Storage Services
Impurity-Selective Adsorbent Screening
Prioritize materials that retain CO2, CH4, N2, CO, or water while preserving hydrogen recovery.
- GCMC mixture adsorption
- Working capacity and regenerability
- Humidity and competitive-adsorption risk
Hydrogen-Selective Membrane Screening
Assess whether permeability and diffusion selectivity support the required hydrogen-rich product.
- Molecular dynamics and transport descriptors
- Adsorption–diffusion trade-offs
- Polymer, MOF, and hybrid membrane context
Adsorptive Hydrogen Storage Screening
Rank porous materials by usable hydrogen delivery across the actual charge–discharge window.
- Gravimetric and volumetric capacity
- Deliverable rather than maximum uptake
- Pressure and temperature sensitivity
Process-Aware Candidate Ranking
Connect molecular performance with cycle recovery, energy, stability, density, and form factor.
- PSA or membrane assumptions
- Pellet and packing effects
- Decision-ready validation plan
Hydrogen material landscape
Materials for Clean Hydrogen
Purification and storage favor different pore chemistries. The library is therefore organized by the job each material must perform: retain impurities, transport hydrogen, or deliver it across a pressure swing.
MOFs for Impurity Capture
Functional pores can retain CO2, CH4, CO, or water while supporting cyclic release.
COFs for Lightweight Storage
Low-density ordered networks offer adjustable binding environments for hydrogen delivery.
Zeolites for Cyclic Purification
Rigid inorganic pores provide thermal robustness for repeated impurity-removal cycles.
Hydrogen-Selective Membranes
Thin selective layers are evaluated through permeability, diffusion, and mixed-gas recovery.
Porous Carbons and Shaped Hybrids
Dense, formable materials connect adsorption capacity with packing and pressure-drop needs.
Hydrogen decision model
How We Prioritize Hydrogen Materials
Hydrogen decisions are not defined by a single uptake or ideal selectivity. We rank candidates against product purity, hydrogen recovery, impurity working capacity, deliverable storage, transport rate, regeneration duty, stability, and the operating window that will be used.
Illustrative weighting only; the bars are not measured material results.
Purification & storage criteriaScenario weighted
H2 recovery
PRODUCT
Impurity working capacity
CYCLE
Product purity
SPEC
Deliverable H2
STORAGE
Regeneration duty
ENERGY
Stability and form
RISK
Hydrogen project path
From Hydrogen Stream to Shortlist
Follow the hydrogen stream from impurity definition to a material and operating window that can be validated.
Frame the Hydrogen Stream
Define pressure, temperature, impurities, humidity, product specification, cycle concept, and storage window.
Curate Candidate Materials
Standardize structures, calculate pore descriptors, assess readiness, and preserve chemistry and topology diversity.
Simulate Adsorption and Transport
Combine GCMC, molecular dynamics, AI triage, mixture analysis, and uncertainty review.
Rank for Delivery
Compare purity, recovery, deliverable capacity, regeneration, stability, and validation needs.
Project package
Hydrogen Data and Decision Outputs
Hydrogen Project Inputs
- Hydrogen-rich feed composition, pressure, temperature, impurities, and humidity
- Required H2 purity, recovery, flow, or storage delivery target
- Adsorption/desorption or membrane pressure window and cycle concept
- Preferred material families, exclusions, and available experimental data
Model and Delivery Checks
- Structure provenance and simulation-readiness checks
- Force-field, charge, and quantum-correction documentation
- Dry, wet, and multicomponent sensitivity analysis
- Density, form-factor, stability, and missing-property flags
01 / DATACurated hydrogen-material library
02 / RESULTSMixture adsorption or transport dataset
03 / DECISIONPurification/storage shortlist
04 / NEXT STEPValidation and process roadmap
Prove purity and delivery
Validate Hydrogen Performance
Computational rankings become useful when they are tested against the pressure, temperature, impurities, cycling, and form factor of the intended hydrogen process. CD ComputaBio helps define measurements that challenge the highest-value assumptions first.
Purification candidates are challenged with realistic impurities; storage candidates are tested for usable delivery across the intended charge–discharge window.
Plan Your Validation Strategy01
Confirm equilibrium behavior
Hydrogen and Impurity Isotherms
Measure H2 and priority impurity uptake over the charging, adsorption, and regeneration windows.
- Single- and multicomponent isotherms
- Low- and high-pressure measurement
- Heat of adsorption and working capacity
02
Challenge the real feed
Breakthrough and Humidity Testing
Test whether CO2, CH4, N2, CO, water, or trace contaminants change hydrogen recovery.
- Dynamic breakthrough curves
- Dry versus humid comparison
- Trace-impurity sensitivity
03
Verify hydrogen delivery
Charge–Discharge Storage Testing
Quantify usable rather than maximum hydrogen capacity under the selected pressure and temperature swing.
- Gravimetric and volumetric delivery
- Packing-density effects
- Thermal management observations
04
Verify transport and robustness
Permeation, Cycling, and Stability
Evaluate membrane transport or adsorbent cycling alongside chemical, thermal, and mechanical stability.
- Mixed-gas permeation where relevant
- Repeated adsorption–desorption cycles
- Shaping and pressure-drop implications
05
Learn from measured results
Model Updating and Re-Ranking
Use measured adsorption, transport, and cycling data to update assumptions and refine the material shortlist.
- Prediction-versus-experiment review
- Uncertainty reduction
- Next-experiment recommendation
Published data
Research Behind Hydrogen Screening
High-throughput simulation connected adsorption and membrane routes
1Screen structuresEvaluate thousands of experimental MOFs.
2Simulate mixturesCombine GCMC adsorption with MD transport.
3Rank by processSeparate PSA, VSA, and membrane candidates.
Avci and colleagues screened 3,857 MOFs for CO2/H2 separation and showed that pore characteristics favoring adsorbents can differ from those favoring hydrogen-selective membranes.[1]
View publicationAI accelerated screening for deliverable hydrogen capacity
1Build the libraryPrepare computation-ready porous structures.
2Simulate deliveryEvaluate pressure-swing working capacity.
3Train and prioritizeUse graph learning to accelerate candidate triage.
Wang and colleagues combined automated simulation with a modified crystal graph neural network to screen about 11,600 MOFs for room-temperature hydrogen working capacity, illustrating how AI can focus more expensive calculations.[2]
View publication
Hydrogen project guidance
Hydrogen Purification and Storage Questions
Choose model detail according to the impurity risk, purity target, storage window, and decision cost.
Can purification and storage be assessed in one project?
Yes. When both decisions matter, the project can compare impurity removal, hydrogen recovery, and usable storage capacity within one material and process framework.
Which hydrogen impurities can be modeled?
Projects may include CO2, CH4, N2, CO, water, and other stream-specific species when suitable interaction models and data are available.
Do you evaluate membranes as well as adsorbents?
Yes. Membrane projects can include adsorption, diffusion, permeability, selectivity, and mixed-gas risk; adsorbent projects focus on cyclic working capacity, recovery, and regeneration.
Why is deliverable storage capacity more useful than maximum uptake?
A storage material must release hydrogen across the selected pressure and temperature window. Maximum uptake alone does not show how much hydrogen can actually be delivered.
Can measured isotherms or permeation data be included?
Yes. Client data can calibrate simulations, test assumptions, and reduce uncertainty in the final ranking.
Define the Hydrogen Decision
Share the feed, impurity envelope, purity or storage target, pressure–temperature window, and available material data. We will define a focused computational and validation strategy.
Scientific References
- Avci, G., Velioglu, S., and Keskin, S. High-Throughput Screening of MOF Adsorbents and Membranes for H2 Purification and CO2 Capture. ACS Applied Materials & Interfaces 10, 33693–33706 (2018). https://doi.org/10.1021/acsami.8b12746
- Wang, L., Feng, S., Zhang, C., et al. Artificial Intelligence and High-Throughput Computational Workflows Empowering the Fast Screening of Metal–Organic Frameworks for Hydrogen Storage. ACS Applied Materials & Interfaces 16, 36444–36452 (2024). https://doi.org/10.1021/acsami.4c06416
