atmospheric-co2-capture

Atmospheric CO₂ capture

Process Mars-native isru
TRL Mars
Energy intensity
Required by
0
Requires
2

Extracts and concentrates CO₂ from the Mars atmosphere as Sabatier feedstock. Mars CO₂ is abundant (95.3 % of atmosphere) but dilute (600 Pa total pressure → ~570 Pa CO₂ partial pressure). Cryogenic freezing during the −90 °C Mars night, pressure-swing adsorption on zeolite, and direct mechanical compression are the three competing architectures. Energy-per-kilogram differs by an order of magnitude between the cheapest (radiative) and most flight-mature (compression).

Last reviewed: 2026-06-08

Governing equations

Mars atmospheric CO₂ partial pressure at typical surface conditions. Bookmark the number: this is what the inlet sees. [1]

Ideal isothermal compression work per mole of gas. Mars surface inlet → 30 bar Sabatier feed is a 5170× ratio; ideal work is ~21 kJ/mol = 480 kJ/kg CO₂. [2]

Direct deposition of CO₂ to solid (dry ice) — bypasses the liquid phase entirely below the triple point (5.18 bar, 216.55 K). On Mars, deposition occurs at ~150 K at typical low pressures. [3]

Carnot coefficient of performance for active cryocooling. Pumping heat from 150 K to a 300 K Mars-day sink gives COP ≈ 1.0; using a passively-radiated 200 K cold plate gives COP ≈ 2.0 — substantial energy savings. [4]

Key constants & quantities

Symbol Value Units Conditions Description
x_CO₂ 95.32 ±0.5% vol% Gale Crater, multi-season average CO₂ mole fraction in Mars atmosphere from Curiosity SAM measurements. Remainder: N₂ 2.7 %, Ar 1.6 %, O₂ 0.13 %, CO 0.08 %, water 0.03 %.[5]
p_Mars 610 ±20 % seasonal Pa Reference Mars surface pressure. Varies seasonally by ~25 % from CO₂ polar cap condensation/sublimation cycle.[1]
T_deposition 150 K (at Mars ambient pressure) Temperature at which gaseous CO₂ deposits directly to solid at ~600 Pa. The polar caps reach this in winter — natural cryotrap.[3]
ΔH_sublimation 25.2 kJ/mol CO₂ sublimation enthalpy at 195 K. Energy required to convert solid CO₂ back to gas during the capture-cycle desorb phase.[3]
W_compression,real 800–2000 ±25 % kJ / kg CO₂ Real multi-stage compression energy from 600 Pa to 30 bar Sabatier feed, including intercooling losses, blower inefficiency, motor losses. ~3× ideal isothermal.[6]
W_cryo,radiative 50 ±30 % kJ / kg CO₂ Energy for cryogenic capture via radiative cooling during Mars night. Most of the heat is rejected for free; only blower work + sublimation energy charged.[7]

Operating envelope

ParameterRangeUnitsSource
Inlet pressure (Mars ambient) 400 – 900 Pa [1]
Outlet pressure (Sabatier feed) 1 – 30 bar [6]
Cryotrap cold-side T 120 – 180 K [7]
Compression ratio 10 – 5000 × [2]
Throughput (per crew, propellant production) 3 – 6 kg CO₂ / sol / crew [7]

Mass balance

Basis: 1 kg CO₂ delivered to Sabatier at 30 bar

Inputs

Mars atmosphere processed 22 kg (95% CO₂ + 5% inerts) [5]
Electrical energy (cryogenic radiative) 0.014 kWh [7]
Electrical energy (mechanical compression) 0.45 kWh [6]
  • Mars atmosphere processed: Inert N₂, Ar, O₂, CO removed by selective adsorption or cryogenic separation; vented back to atmosphere.
  • Electrical energy (cryogenic radiative): Blower work to move 22 kg of atmosphere through cold trap during night cycle.
  • Electrical energy (mechanical compression): Multi-stage isentropic compression — order of magnitude more energy-intensive than radiative.

Outputs

Pure CO₂ at delivery pressure 1 kg [6]
Inert gas vent (N₂, Ar, O₂) 1.05 kg [5]
  • Pure CO₂ at delivery pressure: > 99.5 % purity after one separation stage; > 99.99 % after two.
  • Inert gas vent (N₂, Ar, O₂): Vented back to atmosphere. Could be captured separately for ECLSS buffer gas (N₂) or breathing oxidant.
TRL · Earth
7/ 9
TRL · Mars
5/ 9
Mechanical compression of Mars atmosphere demonstrated at lab scale (Hartvigsen 2017 NASA prototype). MOXIE-on-Mars (2021–) demonstrates the front-end intake + filtration for ISRU CO₂ — though MOXIE itself converts to O₂ rather than passing CO₂ through. Cryogenic radiative capture demonstrated in analog (Mars-chamber tests, Mars Society analog facilities). Mars TRL 5 reflects analog demonstration without integrated end-to-end flight.[8]
Energy budget
0.45 kWhe / kg CO₂ at 30 bar (mechanical compression baseline) [6]

Radiative cryotrap reduces this to ~0.014 kWh/kg by exploiting Mars-night cold-side temperature. For a 4-crew base producing 20 kg CH₄/sol, the compression route demands ~25 kWh/sol — significant but tractable on a 10 kWe nuclear baseload.

Variants & trade-offs

Mechanical compression (Hartvigsen-class)

[6]

Multi-stage scroll or piston compressor pulls CO₂ from Mars atmosphere through dust filters and chemical drying, delivers directly to Sabatier at 1–30 bar. The MOXIE / industrial-prototype approach.

Compression stages
2–5
Inlet flow
10–100 kg/sol atmosphere processed
Stack lifetime
10000–30000 h
Materials: Scroll or oil-free piston compressor · HEPA + nano-filter dust pre-cleaner · Molecular sieve dryer · Pressure vessel + intercoolers
  • Highest TRL — direct flight heritage from MOXIE front-end
  • Continuous operation, day or night
  • Tight control over outlet purity and pressure
  • Highest energy demand of the three variants
  • Compressor wear is the primary failure mode
  • Vibration sensitivity — careful mounting needed

Cryogenic radiative freeze (Zubrin Mars Direct architecture)

[7]

During Mars night (−90 °C ambient), a radiator panel passively cools a cold-plate to < 150 K. CO₂ deposits as dry ice on the plate, leaving N₂ and Ar gaseous. By dawn, the cold-plate is heated to 200 K, dry ice sublimes, and CO₂ exits at 6 bar — high enough for Sabatier without further compression.

Cold-plate T (capture)
120–150 K
Cycle period (one Mars sol)
12–14 h (capture phase)
Stack lifetime
50000–100000 h
Materials: Anti-dust radiator panel (aluminized polyimide) · Thermal switch (heat pipe with shutoff valve) · Pressure vessel for dry-ice accumulation · Sublimation heater
  • Lowest energy demand by an order of magnitude — radiative cooling is free
  • Mechanically simple — minimal moving parts
  • Self-purifies: N₂, Ar, O₂ remain gaseous at capture T
  • Mars Direct flagship architecture; Zubrin's mass budget rests on it
  • Diurnal duty cycle — only collects during cold half of sol
  • Dust storms suppress radiative cooling for weeks
  • Cold-plate fouling by N₂/Ar deposition at very low temperatures
  • Lower flight TRL than mechanical

Pressure-swing adsorption (PSA, zeolite)

[9]

Zeolite molecular sieve selectively adsorbs CO₂ at low pressure, releases at lower pressure (vacuum desorption) or higher temperature. Two-bed alternating architecture provides continuous output, similar to CDRA design.

Adsorption T
200–280 K
Desorption pressure
10–100 Pa (vacuum swing)
Stack lifetime
40000–80000 h
Materials: Zeolite 13X or proprietary CO₂-selective sieve · Stainless bed canisters · Vacuum pump for desorption
  • Continuous operation possible with bed alternation
  • Direct heritage from ISS life-support hardware
  • High purity output (> 99.9 %)
  • Vacuum desorption needs a vacuum pump on Mars (atmosphere is the "vacuum" but only ~600 Pa — borderline)
  • Bed thermal swing demands sustained heater power
  • Throughput per sorbent mass lower than mechanical compression

Failure modes

Mode Cause Detection Mitigation
Compressor seal wear (mechanical variant)[6] Mars dust ingress to compressor cylinders abrades seals; thousands of hours of duty cycling degrades elastomers. Outlet pressure droop at constant flow; vibration signature change. HEPA + electrostatic dust pre-cleaner; oil-free compressor design (avoids dust-oil agglomeration); programmed seal replacement every 8000 h.
Dust storm radiator fouling (cryotrap variant)[7] Sustained dust storms deposit fines on radiator panels, reducing emissivity and lifting cold-side T above the CO₂ deposition threshold. Cold-plate T fails to reach setpoint during night cycle; sublimation rate drops. Vertical-facing radiator orientation; electrostatic dust shedding; mechanical wiper; buffer storage to bridge multi-sol storms.
Zeolite water co-adsorption (PSA variant)[9] Trace water vapor in Mars atmosphere (~0.03 %) co-adsorbs onto CO₂-selective sites; reduces sorbent capacity over thousands of cycles. CO₂ throughput declines at fixed cycle period; outlet humidity climbs. Upstream silica-gel pre-dryer; periodic high-T bake-out (200 °C); programmed sorbent replacement.
Inert gas accumulation in cryotrap[7] At cold-plate T < 130 K, N₂ and Ar begin to deposit alongside CO₂, contaminating the product stream. Outlet GC shows rising N₂/Ar fraction. Tight cold-plate T control (130–145 K window); short collection cycle to limit inert build-up; vent flush between cycles.
Cold-end heater failure (cryotrap)[7] Resistance heater that drives sublimation phase fails open-circuit. Output flow drops to zero during desorb phase; cold-plate T stays cold. Redundant heater elements; back-up sublimation via natural daytime warming with vent valve closed.
Intake clogging by dust (all variants)[6] Mars dust storms (regional or global) accumulate fine particulates on intake filters faster than nominal. Pressure drop across intake filter climbs; throughput falls. Multi-stage dust filter with field-replaceable cassettes; self-cleaning intake with reverse-flow purge; redundant intake paths.

Mars adjustments

Atmospheric pressure 0.6 % of Earth[7]

Impact: The Sabatier reactor needs 1–30 bar feed; Mars atmosphere is 600 Pa. Compression ratio of 50–5000× sets the energy floor for any non-cryogenic architecture.

Mitigation: Cryogenic radiative capture sidesteps compression entirely — CO₂ deposits as solid at ~150 K, sublimes to 6 bar at 220 K. The compression "ratio" is replaced by a T cycle that radiative cooling delivers for free.

Night-side cold trap availability[7]

Impact: Mars night-side surface drops to −90 °C (183 K) typically, −125 °C (148 K) during polar winter. A radiator pointing at the cold sky reaches even lower temperatures — natural cryotrap.

Mitigation: Cryogenic capture only works during the cold half of each sol; system sized to collect a full sol's production during ~12 hours. Storage buffer between capture and Sabatier feed bridges the gap.

Dust contamination at the intake[8]

Impact: Mars dust is fine (1–3 µm modal), perchlorate-rich, and abrasive. Atmospheric intake filters at ground level see continuous dust loading that ramps under storms.

Mitigation: Two-stage filtration (electrostatic precipitator + HEPA), elevated intake snorkel (1+ m above grade), reverse-flow self-cleaning cycles. MOXIE intake provides flight-validated design.

Seasonal CO₂ availability[1]

Impact: Mars atmospheric pressure swings ±25 % seasonally as CO₂ condenses/sublimes from polar caps. Capture throughput tracks this — winter production drops below summer at non-polar sites.

Mitigation: Size for winter throughput; buffer-storage methane production during summer to even out propellant output for transit windows.

Alternatives & substitutes

Direct CO₂ electrolysis (MOXIE)[8]

  • Eliminates the Sabatier step entirely — produces O₂ directly from atmospheric CO₂
  • MOXIE flight-proven on Perseverance since 2021
  • No H₂ feedstock required
  • Output is O₂ + CO, not propellant
  • CO must be vented or used as Fischer-Tropsch feed
  • Doesn't produce CH₄ — separate pathway needed for fuel

When preferred: When O₂ alone is the goal (ECLSS, EVA, oxidizer-only architectures).

In-situ atmospheric collection via airlock cycling[7]

  • No dedicated infrastructure — uses existing airlock pumps
  • Zero new mass on launch manifest
  • Throughput orders of magnitude too low for propellant production
  • Contaminates internal systems with Mars dust
  • Only useful as a buffer or backup

Requires

References

  1. Haberle, R. M., Clancy, R. T., Forget, F., Smith, M. D., & Zurek, R. W. (Eds.) (2017). The Atmosphere and Climate of Mars. Cambridge University Press. ISBN 978-1-107-01618-7. — Reference handbook for Mars atmospheric pressure, temperature, dust climatology.
  2. Green, D. W., & Southard, M. Z. (2019). Perry's Chemical Engineers' Handbook, 9th Edition. McGraw-Hill Education. ISBN 978-0-07-183408-3. — Canonical chemical-engineering reference: thermodynamic calculations, equipment sizing, unit operations.
  3. Linstrom, P. J., & Mallard, W. G. (Eds.) (2024). NIST Chemistry WebBook, NIST Standard Reference Database Number 69. National Institute of Standards and Technology. doi:10.18434/T4D303 — Thermodynamic properties of H₂O, H₂, O₂. ΔH°, ΔG°, S° at standard state.
  4. Barron, R. F. (1999). Cryogenic Heat Transfer. Taylor & Francis. ISBN 978-1-56032-551-7. — Classic cryogenic engineering reference — heat-leak calculation, vacuum-jacketed vessel design, stratification.
  5. Franz, H. B., Trainer, M. G., Malespin, C. A., Mahaffy, P. R., et al. (2020). Initial SAM calibration gas experiments on Mars: Quadrupole mass spectrometer results and implications. Planetary and Space Science, 138, 44-54. doi:10.1016/j.pss.2017.01.014 — Mars atmospheric composition from Curiosity SAM — CO₂ 95.32 %, N₂ 2.7 %, Ar 1.6 %, O₂ 0.13 %.
  6. Hartvigsen, J. J., Elangovan, S., Frost, L., Larsen, D., Elwell, J., Bayless, A., & Stoots, C. (2017). Carbon Dioxide Electrolysis for Mars ISRU. ECS Transactions, 78(1), 2953-2966. doi:10.1149/07801.2953ecst — MOXIE precursor work — solid-oxide CO₂ electrolysis at Mars conditions.
  7. Zubrin, R., & Wagner, R. (1996). The Case for Mars: The Plan to Settle the Red Planet and Why We Must. Free Press, New York. — Mars Direct mission architecture, in-situ propellant production, water electrolysis context.
  8. Hecht, M. H., Hoffman, J. A., Rapp, D., McClean, J. B., et al. (2021). Mars Oxygen ISRU Experiment (MOXIE). Space Science Reviews, 217(1), 9. doi:10.1007/s11214-020-00782-8 — MOXIE flight instrument — first ISRU demonstration on Mars (2021-).
  9. Knox, J. C. (2005). International Space Station Carbon Dioxide Removal Assembly Testing. 35th International Conference on Environmental Systems, SAE 2005-01-2864. doi:10.4271/2005-01-2864 — CDRA architecture, zeolite 13X/5A duty cycle, ISS performance history.