Direct air capture companies are in a race to prove they can reduce the cost of removing carbon from the atmosphere down below $100 per ton. Now, one is closing in on the prize with a first-of-its-kind deal.
On Tuesday, Google announced it will pay the startup Holocene $10 million to remove 100,000 tons of carbon from the atmosphere, to be delivered “by the early 2030s.” The tech giant said the price point was made possible by the federal tax credit for carbon sequestration, and its own willingness to cough up the bulk of the funds upfront.
There’s no question the deal is risky on both sides. Today, most estimates place the cost of direct air capture at upwards of $600 per ton. Bringing the cost down is essential if the tech is ever going to play a meaningful role in tackling climate change. But even the companies that are farthest along, like the Swiss pioneer Climeworks, aren’t sure they will be able to offer a price of $100 per ton by 2030. Holocene has yet to build a commercial plant, so its ability to remove carbon for $100 per ton is pure projection at this point.
But for Google, the goal is more to catalyze a potentially important climate solution than to clean up its carbon footprint.
“The point of our program is to help Google reach net zero in whatever way most helps the world reach net zero,” Randy Spock, the company’s carbon credits and removals lead, told me in an email. “So this deal is an example of us identifying what the planet needs (long-term cost reduction for Direct Air Capture) and then doing what we can to help it take a step in that direction.”
Though Holocene is relatively new to the direct air capture market, it was started by veterans. Co-founders Anca Timofte and Tobias Rüesh spent roughly six years working in research and development at Climeworks back in its early days, when the company was building its first prototypes. Timofte left in 2020 to get an MBA at Stanford, and while there, came across some exciting research out of Oak Ridge National Laboratory that described a new approach to removing carbon from the ambient air — one that seemed to have distinct advantages. Seeing the potential, Timofte decided to start Holocene with Rüesh and another Stanford classmate and, in 2023, licensed the Oak Ridge technology.
“The chemistry from Oak Ridge is special,” Timofte told me. “It's different than all other chemistries, we think, in direct air capture.”
Most direct air capture systems fall into one of two categories, liquid or solid, and each approach has trade-offs. Liquid systems typically have simpler engineering and can capture CO2 continuously, but require more heat, and therefore more energy. Solid systems have lower heat requirements, but work sort of like cartridges that get “charged” with CO2 and have to be “discharged,” and therefore capture CO2 in batches rather than in perpetuity.
Timofte described Holocene’s process as the “best of both worlds.” It captures CO2 in water and operates in a continuous loop, but requires relatively low heat — between 70 to 100 degrees Celsius (158 to 212 degrees Fahrenheit) — which could potentially come from a source of waste heat like a data center. The enabling discovery was the use of two chemicals — an amino acid and a compound called guanidine — that attract CO2 and then further concentrate it within the water, making it easier and less energy-intensive to isolate so that it can be stored securely underground.
After licensing the tech, Holocene moved quickly. Within a year, the team had built a small pilot plant in Knoxville, Tennessee that’s capable of capturing about 10 tons of CO2 annually. That’s, of course, a totally insignificant amount, but it’s enough for the team to demonstrate its approach to potential funders and to keep testing variations on the basic chemistry to refine the system, Timofte told me.
Timofte said the company has made it this far with just over $6 million in grants and prizes from the Department of Energy, Bill Gates’ Breakthrough Energy, and Frontier Climate, a coalition of carbon removal buyers that includes Google in addition to other tech companies. The $500,000 that Holocene got from Frontier was technically a pre-purchase of 332 tons of removal, which would put the current cost per ton at roughly $1,500.
Frontier’s pre-purchases are not a precise indicator of price as they are meant to “pressure-test the viability of novel CDR solutions,” and are granted with the expectation that some ventures will fail. Still, even with a fresh influx of cash from Google and the prospect of a $180 per ton tax credit from the federal government, the company has a steep climb ahead. Timofte told me the team is beginning to fundraise to build their next project — a 2,000- to 5,000-ton per year demonstration plant. When asked about how it reached the $100 per ton deal with Google, she stressed that having a delivery date past 2030 was crucial to the deal.
The industry’s fixation on achieving $100 per ton is somewhat arbitrary. A 2019 National Academies of Sciences report found that estimates of the cost of capturing CO2 via direct air capture spanned “an order of magnitude, from $100 to $1,000” per ton. In 2021, the Biden administration’s Department of Energy set a goal to bring the cost of all kinds of carbon removal below $100 per ton, which seemed to solidify the goal across the field. In 2022, the nonprofit CarbonPlan surveyed carbon removal buyers, suppliers, and brokers, and found that $100 per ton was a common benchmark. “If cost were $100/ton, demand would be practically unlimited,” one supplier said. “Bringing down cost to $100/ton for CDR would be the sweet spot,” said a buyer. CarbonPlan pointed out, however, that the responses weren’t consistent on whether $100 per ton was the desired break-even point for carbon removal companies or the desired price for buyers.
“I think we focus too much on the cost of DAC,” Erin Burns, the executive director of the nonprofit Carbon180 told me when I asked her if $100 per ton was a meaningful goal. “Sure, DAC should and will get cheaper. But we need to also be thinking, right now, about things like renewable energy availability, infrastructure, and reducing emissions as quickly as possible.”
Finding clean sources of power for direct air capture is becoming more of an issue as companies try to scale. At the end of August, a startup called CarbonCapture Inc. announced it would try to relocate a commercial-scale project it had planned to build in Wyoming because it was struggling to procure enough clean energy to power the plant due to competition with data centers and cryptocurrency miners.
Timofte agreed that “clean electrons are hard to come by,” but added that Holocene’s potential to use waste heat might make it a little easier for the company.
“I don't want to dismiss the challenge. I think this is the challenge that everyone faces. We each have to solve it, and the solutions are going to be individual.”
The 2026 Ioniq 5 Limited. Image courtesy of Hyundai
