Machines

When Acidic Gas Recovery Stops Waiting for a Refill

A carbon-capture patent arrives under a retired corporate name, exposing the costly, stubborn work behind cleaner exhaust.

Clay-diorama editorial illustration related to Acidic Gas Recovery Apparatus And Acidic Gas Recovery Method.

What changes in this design

A reader might expect that if material used to absorb acidic gas leaks, more must be added before gas recovery can continue.

Here, absorption and regeneration units are connected so an acidic-gas absorbent circulates between them. The system is intended to recover acidic gas continuously from exhaust gas or air without additional replenishment, even when the absorption component leaks.

The central point is continuity: recovery is meant to continue through that leak condition without stopping for an added refill. How much gas it recovers and how it performs are not stated.

Ten kilograms at Senju

The number I keep returning to is 10 kilograms a day. In January 2024, Toshiba installed a truck-portable carbon-capture unit at Tokyo Gas's Senju Techno Station, measuring about 3.1 by 2 by 2.8 metres and weighing 2.3 tonnes. Its stated capacity was modest: up to 10 kilograms of CO2 a day.

That size matters to me because carbon capture is often discussed through giant plants and distant targets. Here was a machine intended for manufacturers and local governments, not a power station. It puts a practical question in view: once a company has taught a liquid to pull CO2 from gas, how long can that liquid keep doing its job before it becomes another costly material to buy, treat, and discard?

April 1, 2026

That question sits beneath a new US patent publication, US20260257174A1, released on September 3, 2026. Its front page names both Toshiba Corporation and Toshiba Energy Systems & Solutions Corporation as assignees, although the latter had been integrated into Toshiba Corporation on April 1, 2026 under a wider group restructuring plan.

The apparent oddity caught my attention. The application was filed on July 1, 2025 and claims priority to a Japanese application filed July 5, 2024, so the two-company attribution reflects work set in motion before the merger, not necessarily an error in the later publication. I read it as a small legal trace of a long engineering programme continuing while the company structure around it changes.

The salt left behind

The filing concerns an apparatus that connects an absorption unit and a regeneration unit, circulating an acidic-gas absorbent between them. Its central claim is continuous recovery of acidic gas from exhaust gas or air without replenishing the absorbent, even if an absorption component leaks.

I found the family history more revealing than the wording alone. A same-titled US patent granted in 2018 used electrodialysis with bipolar and anion-exchange membranes to remove heat-stable amine salts from CO2-absorbing liquid. The newer filing extends that basic effort: keep the working liquid useful rather than letting losses and contaminants turn it into a recurring supply problem.

This is not a promise that capture becomes easy. Amine gas treating has been used since 1930, and common amines can degrade when exposed to CO2 and oxygen. In moist systems they can also contribute to corrosion. The difficult part is often not getting a gas into the liquid once, but preserving the liquid through thousands of cycles.

Five hundred tons a day

Toshiba says it has developed its amine solvent chemistry since 2007, and that its commercial and demonstration plants have captured more than 600 tonnes of CO2 a day. In 2020, its facility at the Mikawa biomass power plant in Fukuoka began operating with stated capture above 500 tonnes daily, or more than half of the plant's daily emissions.

The programme stretches in both directions. There is the 10-kilogram portable unit at Senju, and there is also a company roadmap for CO2 electrolysis that described a basketball-court-sized electrode area processing about 100,000 tonnes annually. That roadmap targeted commercialization in 2026 after a 2023 prototype. Such targets are not proof of deployment, but they show why solvent upkeep matters: it is one unglamorous constraint shared by very different scales of machinery.

The harder half of the problem

The figure I could not shake was the estimate that regeneration can consume up to 70% of an amine system's operating costs. Heating, stripping, cleaning up degradation products, and replacing lost material all compete with the value of the captured gas. A membrane-based way to protect the solvent may therefore be more consequential than its quiet patent language suggests.

Still, my view is that this filing should be read as maintenance work, not a declaration that carbon capture's economics are settled. A carbon-capture advocacy group listed 45Q incentives, in an August 2025 update, of $85 per metric tonne for industrial or power capture and $180 for direct-air capture. A separate cost analysis argues that the familiar $100-per-tonne direct-air-capture ambition rests on idealised assumptions, citing a 2024 estimate of $230 to $540 per tonne by 2050.

Those figures do not measure this apparatus directly, and I would not use either to pronounce the technology viable or doomed. They do explain why a patent about leaks, salts, and solvent circulation deserves attention. Capturing carbon is the visible half of the task. Keeping the capture system affordable enough to run, day after day, is the harder half.

Sources

Reporting for this piece drew on the published filing and the references below.

The record, briefly. US20260257174A1, “Acidic Gas Recovery Apparatus And Acidic Gas Recovery Method” An acidic gas recovery apparatus and method recover acidic gas from exhaust gas or air continuously without additional replenishment even when an acidic gas absorption component leaks, using connected absorption and regeneration units that circulate an acidic gas absorbent.

More angles on this

Timing at the Kilogram-Scale Mushroom HandoffAnother reading from the same B01D classification family, separation.
Degassing a bottle: a chamber, a pump, and the judgment left to whoever holds itAnother reading from the same B01D classification family, separation.
CPC B01D: SeparationThe official classification family this record was filed into, published in the USPTO Cooperative Patent Classification scheme.
When Shared Pipes Need a PlanAnother reading from the neighbouring B01J classification family, chemical or physical processes, e.g. catalysis or colloid chemistry; their relevant apparatus.
A Lanyard System That Waits for EvidenceAnother reading from the wider B66F classification family, hoisting.

Read around this one:Public patent record USPTO published figure