The second article says "the team has proven the system can withstand at least 7,000 charging-discharging cycles, with a 30 percent loss in efficiency over that time. The researchers estimate that they can readily improve that to 20,000 to 50,000 cycles."
That's without consumables, just a stack of electrodes coated with an organic compound and carbon nanotubes.
Your examples are illustrative of the point. These are the equivalent of "in mice" articles in medicine. Showing DAC in a lab is the easy part.
Just to offset current emissions, we need to pull 40 billion tons of CO2 out of the air each year. This must work within the constraints of supply chains, upstream chemical feedstock, and available planetary mineral reserves. People are not comprehending the scale: this would be the largest industrial chemistry process in history by orders of magnitude, larger than all other industrial chemistry combined.
Scaling either of the above DAC processes to 40 billion tons per year is infeasible. You don't just have to scale the process, you also have to scale the upstream processes and resource extraction activities that enable the process to be built. At 40 billion tons, any chemical engineer can quickly pencil out that these processes have upstream dependencies that have no ability to be scaled sufficient to support the target production process, not even close. In computer terms, it is like trying to train the latest LLMs on a computer from the 1990s.
I see a lot of DAC proposals that, at scale, imply a supply chain that instantly consumes the global reserves of critical agricultural minerals like potassium or phosphorus, or which would literally produce e.g. concentrated acid waste at the scale of billions of tons, with no plan for how to keep people from starving or disposing of the chemical waste. Consequently, these processes are not real solutions.
Yes I understand the argument, but you're still not posting actual numbers for any specific materials required by these particular technologies. I could just as well mention the immense materials required for grid-scale batteries, but that doesn't mean they're necessarily infeasible; it all depends on which materials we use.
Before you say "industrial chemistry is different" I'll note that the MIT tech is actually quite similar to a battery. Aside from the CO2 it doesn't have a flow of chemicals that turn into waste; I've seen proposals like that too and this is different.
The Princeton method is new in the lab but the MIT project started a company that has been working on commercializing for five years now. So far they've won an XPrize, gotten investment from the Gates fund, and started collaborating with an aluminum company.
https://engineering.princeton.edu/news/2024/03/14/engineers-...
https://news.mit.edu/2019/mit-engineers-develop-new-way-remo...
The second article says "the team has proven the system can withstand at least 7,000 charging-discharging cycles, with a 30 percent loss in efficiency over that time. The researchers estimate that they can readily improve that to 20,000 to 50,000 cycles."
That's without consumables, just a stack of electrodes coated with an organic compound and carbon nanotubes.