Carbon Nanofiber Filters: Direct Air Capture Inside Building HVAC

Imagine your office building's HVAC system filtering PM2.5 — and also pulling CO₂ out of the air while it does so.

Not science fiction. Science Advances, Oct 2025 demonstrated it.

How Does the Filter Work?

Carbon nanofiber (CNF) substrate coated with polyethyleneimine (PEI) polymer as the CO₂ sorbent.

Sorption → Reaction → Regeneration

Step 1: sorption Building ventilation air naturally flows over the filter. Ambient CO₂ (~420 ppm) contacts the PEI coating.

Step 2: chemical reaction Amine groups (-NH₂) on PEI react with CO₂ to form ammonium carbamate:

R-NH₂ + CO₂ → R-NH-COO⁻

This is chemisorption (not physical) — binding energy is 10×+ van der Waals, and CO₂ is locked in place.

Step 3: regeneration When the filter saturates, heat to 80–100 °C. The carbamate decomposes and releases pure CO₂. The released CO₂ can:

• ▸Be sequestered in geological formations
• ▸Feed synthetic fuels or plastics
• ▸Supply food-grade CO₂ industries

The regenerated filter continues capturing. Cycle repeats.

The Key Numbers: How Much Carbon Does This Actually Remove?

A full Life Cycle Assessment (LCA) was conducted. Key findings:

92.1 % net carbon removal efficiency

For every 1 ton of CO₂ captured, the total system (manufacturing + transport + regeneration energy) emits 0.079 ton.

Context: one tree absorbs about 21 kg CO₂/year — 50 tree-years offset one passenger car's annual emissions. A DAC filter integrated into building HVAC operates 24/7 at a small footprint.

USD 209–668 per ton

Higher than industrial centralized DAC (e.g., Switzerland's Climeworks at ~USD 100–300/ton).

But the advantage: distributed DAC needs no dedicated land, facility, or standalone power. It rides on existing buildings and existing airflow. Cost reduction comes from shared infrastructure.

596 Mt per year global potential

Upper-bound estimate for wide deployment across global buildings. Context: global CO₂ emissions are about 37 Gt/year. 596 Mt equals ~1.6 % of annual global emissions.

Sounds modest? That's from one technology. Climate targets need many such 1.6 %'s stacked together.

Fundamental Difference From Traditional Filters

The filter role shifts from "passive" to "active"

Traditional filter: particle hits media → particle held. Pure physical process, filter itself doesn't change.

DAC filter: CO₂ molecule → chemically reacts with PEI → becomes new compound → released on regeneration. The filter actively participates in a chemical cycle — not just a filter, also a chemical reactor.

The filter goes from "consumable" to "equipment"

Traditional: saturate → replace → old filter to landfill.

DAC: saturate → heat regenerate → continue. The filter itself may run 5–10 years rather than 5–10 months.

This changes procurement logic: no longer "X units per year consumable" but "one set every few years of capital equipment."

Honest Assessment: How Far From Commercial Deployment?

The paper demonstrates principle feasibility and reasonable LCA — but a gap remains to universal deployment:

Technical

• ▸Long-term stability still needs validation (5–10 year cycles)
• ▸Efficiency in high-humidity or with co-pollutants
• ▸Performance across climate zones

Economic

• ▸Room for cost reduction below 209–668 USD/ton?
• ▸Why would building owners install it? (Carbon pricing, regulation)
• ▸Where does regeneration energy come from? (Fossil-fueled regen cuts the benefit)

Policy

• ▸What happens to captured CO₂? Sequestration needs infrastructure
• ▸Can carbon-credit markets certify distributed DAC contribution?

What This Means for the Filter Industry

First, the definition of "filter" is expanding. Historically filter = filter. Future filter may = multi-function chemical reactor.

Second, "green building" purchasing may include DAC capability. Imagine LEED or WELL certifications adding "ventilation system integrates DAC filter" as a credit.

Third, the supply chain will change. Carbon nanofiber, PEI, photocatalysts, antimicrobial metals — materials with no prior connection to air filtration will enter the supply chain.

The real significance isn't "a filter that catches CO₂" — it's that filter media, as a material, can evolve from passive medium to active functional material. Expect similar lines over the next few years: formaldehyde-capture filters, virus-decomposing filters, ozone-generating filters (one medium, many functions).

This Science Advances paper may just be the beginning.