Scientists at ETH Zurich have turned proteins extracted from dairy whey and tofu wastewater into biodegradable microbeads that pull carbon dioxide straight out of ambient air. The beads, treated with potassium hydroxide, captured up to 2.20 mmol of CO2 per gram under real-world humidity and temperature conditions, and they can be refreshed in roughly ten to twelve minutes using only dilute acid and base at room temperature. The work, published in PNAS on June 8, 2026, offers a potential path toward direct-air capture systems that skip the energy-hungry heating and vacuum steps used by most commercial approaches today.
Why food-waste sorbents change the direct-air-capture calculus
Most direct-air capture (DAC) plants rely on temperature-vacuum swing adsorption, a process that demands substantial heat or reduced pressure to release trapped CO2 from solid or liquid sorbents. A peer-reviewed life-cycle assessment of that mainstream approach, published in Nature Energy, documented the significant energy and emissions burdens that come with industrial-scale thermal regeneration. The ETH Zurich team sidesteps that bottleneck entirely. Their amyloid fibril beads release CO2 through alternating washes of dilute acid and alkali at ambient temperature, completing each cycle in about ten to twelve minutes with no furnace, no compressor, and no vacuum pump.
That speed matters because regeneration time directly limits how many capture-release cycles a sorbent can complete per day, which in turn sets the economics of any DAC installation. A sorbent that refreshes in minutes rather than hours can process far more air through the same hardware. The hypothesis that pairing these beads with on-site electrodialysis regeneration could push net energy use below 1 GJ per tonne of CO2 while keeping capture rates above 1.5 mmol/g is technically plausible given the reported numbers, but no primary energy-balance data or life-cycle inventory for the KOH-treated beads has been published yet. Testing that idea would require a pilot unit of at least 100 kg of sorbent operating under variable weather, a step the current lab-scale work has not attempted.
From whey protein to 2.20 mmol/g: how the amyloid beads perform
The Mezzenga Lab at ETH Zurich detailed the full fabrication and performance data in a technical preprint. Researchers extracted proteins from two common food-processing waste streams, dairy whey and soy (tofu) wastewater, then coaxed those proteins into amyloid fibrils, thread-like protein assemblies with high surface area. The fibrils were shaped into microbeads and activated with KOH, which introduces the chemical sites that grab CO2 molecules from air containing only about 420 parts per million of the gas.
The resulting sorbent achieved up to 2.20 mmol of CO2 per gram under ambient conditions. That figure sits well above the performance floor that many analysts consider necessary for economically viable direct-air capture. Earlier work from the same research group had already shown that amyloid-based materials can bind CO2 selectively. A foundational study in protein nanofibers demonstrated selective CO2 binding by designed amyloid fibers, establishing the scientific basis that the 2026 beads build on. A separate peer-reviewed paper in ChemSusChem extended that line of research by showing how amine-modified aerogels can be engineered for CO2 adsorption, providing a direct comparison point for the newer KOH-based approach.
The biodegradability of the beads adds a second environmental benefit. Unlike synthetic amine sorbents or metal-organic frameworks, which can persist in landfills or require specialized disposal, protein-based beads are designed to break down after their useful life. The institutional release from ETH Zurich framed the approach as closing a loop between food waste and atmospheric carbon removal. In principle, the same dairy and soy processing facilities that generate the waste streams could host small DAC units that capture CO2 for local use in greenhouses, beverages, or mineralization processes.
Durability, cost, and scale questions the data does not yet answer
Several gaps separate the lab results from a working climate technology. First, the preprint does not report long-term cycling stability. A sorbent that delivers 2.20 mmol/g on its first pass but degrades after a few hundred cycles would be impractical for continuous outdoor operation. Real-world DAC systems aim for tens of thousands of cycles over years of service. Without data on mechanical abrasion, chemical fouling, or loss of active sites over time, it is impossible to estimate replacement rates or maintenance costs.
Field performance under variable humidity, temperature swings, and airborne contaminants such as sulfur dioxide or particulate matter also remains untested. The laboratory experiments exposed the beads to controlled humidity and clean air, conditions that differ sharply from dusty, polluted, or coastal environments where large DAC arrays might be sited. Interactions between the KOH-treated surfaces and trace gases could either reduce capacity or require periodic cleaning steps that the current process flow does not include.
Second, no detailed cost model or supply-chain analysis accompanies the research. Dairy whey is abundant in countries with large dairy industries, and tofu wastewater is plentiful across East and Southeast Asia, but the volumes needed for a meaningful DAC installation have not been mapped against actual waste availability. Seasonal variation in protein content and regional differences in waste-stream composition could affect sorbent quality. Standardizing bead properties at industrial scale would likely require pre-treatment and quality-control steps that add cost and complexity.
Third, the 2.20 mmol/g figure has not been independently replicated under blinded conditions. The preprint is not yet peer-reviewed in its current form, and small differences in bead drying, KOH loading, or measurement protocols can significantly shift apparent capacity. Independent labs will need to verify both uptake and regeneration performance using standardized testing methods before investors or policymakers can treat the numbers as bankable.
There are also open questions about the fate of the captured CO2. The current work focuses on the sorbent itself rather than on downstream handling, but any practical system must compress, transport, or convert the gas. Because the beads regenerate at ambient temperature using mild solutions, they could, in principle, be integrated with low-pressure CO2 utilization processes such as biological fixation or mineral carbonation. However, those couplings remain speculative until process-level studies quantify flow rates, concentrations, and compatibility with existing equipment.
What comes next for protein-based direct-air capture
Turning the ETH Zurich concept into a deployable technology will require a sequence of engineering steps. Pilot-scale bead production is the first hurdle: moving from gram-scale batches to hundreds of kilograms while maintaining uniform size, porosity, and KOH content. Parallel work will need to design contactors-devices that bring air into contact with the beads-that minimize pressure drop while maximizing exposure time, likely through packed beds or fluidized columns.
Researchers will also have to optimize the regeneration loop. The current acid-base washes are simple but generate mixed salt streams that must be managed or recycled. Integrating electrodialysis or other membrane processes could close that loop, reducing chemical consumption and waste, but would add capital costs and operational complexity. Detailed techno-economic analyses will be essential to determine whether the low-temperature advantage of the beads offsets the extra handling steps.
Regulators and standards bodies may eventually play a role as well. Because the beads are biodegradable and derived from food-industry byproducts, they could fit comfortably within emerging sustainability frameworks for carbon removal, provided that life-cycle assessments confirm low upstream emissions and minimal land-use impacts. Certification schemes for durable carbon removal will expect clear accounting of how long the captured CO2 remains out of the atmosphere, whether in geological formations, long-lived products, or stable mineral forms.
For now, the amyloid microbeads represent a promising laboratory demonstration rather than a turnkey climate solution. They show that high-capacity CO2 capture is possible using inexpensive, renewable feedstocks and mild operating conditions, challenging the assumption that direct-air capture must rely on exotic materials and high-grade heat. The next few years of peer review, replication, and pilot testing will determine whether this bio-based approach can move from benchtop curiosity to a practical tool in the broader portfolio of carbon removal technologies.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
