A direct air capture (DAC) plant under construction in Ector County, Texas, is designed to remove 500,000 tonnes of CO₂ per year using potassium hydroxide (KOH) absorption. The efficiency of that scale-up now has a research instrument behind it.

Researchers at the University of Colorado Boulder have built a custom laboratory flow cell that images, in real time, the chemical reactions happening at the gas-liquid interface where alkaline DAC systems capture carbon. The work, published 13 May 2026 in ACS Energy Letters (DOI: 10.1021/acsenergylett.5c04139), replaces indirect input-output measurements with spatial maps of the reactive zone at sub-millimetre resolution.

What did the imaging reveal?

The maps contradicted simple intuition. "We saw that the equilibrium reaction is in effect going backwards near the surface," said lead researcher Jason Pfeilsticker, a graduate student in the group of RASEI Fellow Wilson Smith.

When fresh KOH first contacts CO₂, hydroxide ions rapidly convert the gas to carbonate near the porous membrane. The reaction then depletes local hydroxide supply. Because the channel runs in laminar flow with no turbulent mixing, a thin bicarbonate layer forms immediately next to the membrane. The pattern intensifies further along the flow channel.

The team also documented two operating variables that alter the reactive zone:

• Higher liquid flow rates changed the shape and extent of the reactive layer.
• Doubling the KOH concentration shifted the product balance and appeared to reduce hydroxide depletion near the membrane.

How does the flow cell work?

The instrument, which the team describes as the only one of its kind, brings CO₂ gas into contact with flowing KOH through a porous membrane, mimicking a real capture contactor. Confocal Raman spectroscopy scans the interface in a grid pattern, reading scattered laser light to identify and quantify carbonates and bicarbonates as they form.

Pfeilsticker said the device was the product of 60 to 70 iterations of prototyping. The team initially considered professional machining, but each iteration would have cost thousands of dollars. They instead identified a chemically compatible resin and a low-cost resin 3D printer, allowing proof-of-principle work for less than a dollar per print.

Three engineering problems drove the design iterations:

• Sealing — mechanisms borrowed from drumhead design
• Bubble suppression — reactor geometry angles to keep the laser path clear
• Laminar flow — shaped inlets, outlets, and a custom flow dampener

What changes for DAC system designers?

The combination of the flow cell and a computational model validated against the experimental data gives researchers a screening tool for novel capture liquids. Until now, alkaline DAC development has largely measured inputs and outputs, leaving interface chemistry to inference. The new platform tests that chemistry directly.

Pfeilsticker framed the work as a shift comparable to medical imaging: "This really is a case of if you want to know about something, just look at, really carefully."

The authors note the device could extend beyond DAC into electrochemical CO₂ conversion and critical mineral separation, wherever reaction and transport interact at an interface.

What happens next?

The CU Boulder group is opening the validated model and flow cell maps to the DAC research community as an initial screening step before bench-scale system builds. On a 500,000 tpy facility, even fractional gains in reaction efficiency translate into material operating cost — a margin the sector is under pressure to widen as alkaline DAC moves from first-of-a-kind plants to multi-asset portfolios.