Inside Arizona State University in Tempe’s lab sit the parts for an artificial tree, capable of extracting carbon dioxide 1,000 times faster than a natural tree. The research team is methodically transforming the process into a machine through which carbon dioxide doesn’t get turned into fruit or roots or tree trunks, as it does with a natural tree, but simply gets collected by resin in a reversible chemical process. This article takes a closer look at the process.

## Article

Inside a rectangular Plexiglas tower in a lab on the campus of Arizona State University in Tempe sit the parts for an artificial tree. Not leaves and branches, but a yellowish corrugated filter silently pulling carbon dioxide out of the air. The carbon dioxide doesn’t get turned into fruit or roots or tree trunks, as it does with a natural tree, but it simply gets collected by resin in a reversible chemical process.

Joyce Kilmer might not have approved, but what the artificial tree lacks in poetry, it makes up for in efficiency. According to Klaus Lackner, founder and director of the Center for Negative Carbon Emissions at Arizona State, the artificial tree-right now, a process more than a product-can extract carbon dioxide 1,000 times faster than a natural tree.

Lackner’s team is methodically transforming the process into a machine.

One machine won’t be enough. Considering the 36 gigatons emitted by human activities every year, it will take another and another and maybe 100 million units, each the size of a shipping container, each extracting a ton of carbon dioxide from the air, each day. At that scale, a forest of artificial trees could potentially keep up with the current rate of carbon emission.

One hundred million is a very large number, Lackner admitted, “but we build 80 million cars and trucks a year.” For millions of those machines to be built, getting the cost low and the performance high will be crucial.

When people try to downplay the problem of dumping carbon dioxide pollution into the atmosphere, they often fall back on the idea that the gas is “plant food.” Trees convert carbon dioxide to organic compounds they need to survive, so adding more CO2 should create better conditions for growing trees. It’s a bad argument-like contending that since drinking water is good for us, we don’t need to protect against floods.

According to a study published earlier this year in the Proceedings of the National Academies of Sciences, all the living matter in all the plants on Earth-from the grass in suburban lawns to the rainforests of the Amazon-equals around 150 billion tons. Another 300 billion tons is locked into non-biologically active trunks and stems. Since around 1900, however, the fraction of carbon dioxide in the atmosphere has increased from 280 parts per million to about 412 ppm. The mass of the carbon in that added CO2 is 280 billion tons, far more than plants could lock up over any reasonable time scale.

If humanity is going to draw down the concentration of carbon dioxide in the atmosphere, as scientists recommend to avoid the worst effects of climate change, then it will require something other than plants. An industrial-scale problem calls for industrial-scale solutions.

That recognition has turned carbon capture into a hot topic. In June, a paper in the journal Joule written by researchers at the company Carbon Engineering reported bringing costs of atmospheric carbon dioxide extraction down to between $232 and$94 per ton—a fraction of earlier estimates. Another company, Global Thermostat, claims costs even lower. A cost of \$100 per ton is the equivalent of about 90 cents per gallon of gasoline.

That’s still not cheap, but direct air capture of carbon dioxide may be the best hope for reducing the impact of carbon pollution. “Once the carbon is in the atmosphere,” said Matt Lucas, associate director of the Center for Carbon Removal, a non-profit organization based in Oakland, Calif., “we have fewer solutions for removing it. Direct air capture is one of those solutions.”

## At the Source

Klaus Lackner started his career as a theoretical physicist, eventually landing at Los Alamos National Laboratory in the mid-1980s. In time, however, he began to be drawn to the practical challenges of carbon capture. By the late 1990s, Lackner realized not only that human economic activity was going to rely on fossil fuels for decades to come, but also that “roughly half of the CO2 we put out we can’t capture because nearly half the emissions are coming from non-point sources.”

Much of the focus of carbon capture has been on point sources, such as power plants and cement factories, where industrial processes create a concentrated stream of carbon dioxide. Capturing CO2 from point sources is like pulling water from a hose. “It’s almost certainly cheaper to capture carbon emissions from their source—or never emit them in the first place,” Lucas said.

Unfortunately, efforts to capture and sequester carbon from fossil fuel plants have themselves proven difficult and expensive. Plans to build a carbon-sequestering demonstration power plant called FutureGen were abandoned in 2015, though a coal-fired power plant in Thompsons, Texas, was retrofitted for carbon capture in 2017. Another promising point source is ethanol production, which produces a stream of pure CO2 that is usually vented off; an ethanol plant in Decatur, Ill., is capturing its carbon emissions and storing it underground. Daniel Sanchez, an AAAS Congressional Science and Engineering Fellow, called capture from biorefineries “the best near-term opportunity” for point-source capture.

Even if capture could be deployed at all point sources, it would simply slow the rate of emissions. If we are going to actually reduce the atmospheric CO2 concentration, we can either wait hundreds of years for nature to do it, or we can begin to actively withdraw it from the air.

Unfortunately, unlike point sources, where CO2 concentrations in exhaust streams are relatively high, capturing CO2 from the atmosphere, where it makes up roughly one part per 2,500, is more challenging. This problem is neatly summarized in what’s called Sherwood’s Rule: The cost of extraction scales linearly with the degree of dilution. Combined with the already high cost of flue gas scrubbing, that suggests that the cost of direct air capture of CO2 would be prohibitive. Any attempt to economically capture CO2 directly from the air must find a way around Sherwood’s rule.

To date, the three companies have rolled out air-capture demonstration projects—Carbon Engineering in British Columbia, Climeworks in Switzerland, and New York City-based Global Thermostat—all use blowers or fans to drive air through their capture systems. Though that helps speed the reactions, it’s also energy intensive. And if that energy comes from fossil-fueled power plants, then the whole process threatens to be counterproductive.

Looking for a way to use less energy in the initial collection stage, Lackner came across a Japanese project that collected uranium from sea water using artificial kelp. Uranium concentration in seawater is only 3 parts per billion, more than 100,000 more dilute than atmospheric carbon dioxide. This suggested a way around Sherwood’s rule: relying on passive collection. By avoiding any processing until the carbon dioxide has been substantially concentrated, the amount of energy required to remove the CO2 should be reduced dramatically. Finding a medium that can cycle between capturing and releasing the gas without applying heat or pressure would reduce the energy requirements even more.

## Capture the Gas

Conventional carbon dioxide sorbents absorb the gas when cool and release it when heated, which creates the necessity for a constant energy supply. Lackner’s team found an inexpensive material that operates on a different cycle: it absorbs CO2 when dry and releases it when wet. According to postdoctoral research associate Shahrzad Badvipour, the sorbent consists of “anionic exchange resins, where the resin will bind CO2 as a bicarbonate ion when in a dry state. When wet, a carbonate ion forms, replacing two bicarbonate ions resulting in the release of a single molecule of CO2.”

A prototype carbon capture system running in Lackner’s lab consists of rigid rectangular channels made of sorbent-impregnated polypropylene, roughly an inch wide and a quarter-inch tall. As air passes through, the material picks off the molecules of carbon dioxide. When saturated, the filter is dunked in water and then placed in a terrarium. A gas analyzer tracks the concentration of CO2 in the terrarium, which rises as the wet filter releases the absorbed carbon dioxide, then drops as the gas is taken up by a potted geranium.

Once the filter is dry, it is removed from the terrarium and the cycle starts again.

Though a small demonstration that releases CO2 into an algae pond is in operation near the Arizona State campus, adapting this process into a forest of artificial trees has been challenging. Some of the first renderings depicted enormous tuning fork-shape structures jutting out of the landscape or panels that would fold up accordion-style into a vat of water.

A more recent design involves disks of sorbent loosely stacked into cylinders to allow air to flow over each surface. At regular intervals—roughly every 1,000 seconds—the stack collapses into a canister that then seals itself up and sprays water over the disks, releasing the absorbed gas. Lackner’s team envisions a cluster of 16 of these stacks, with one collapsing and one springing up each minute.

To collect the CO2 from the stacks that are outgassing, a vacuum would be pulled continuously, using a clever ejector system, based on the Bernoulli effect, through a system of valves that switch its input to whatever cylinder is being harvested at the moment. The concept is similar to the system used in steam railroads to draw cold water into pressurized boilers. The flow is directed such that the output of the most depleted cylinder is fed into its neighbor that immediately preceded it in the sequence, which in turn feeds into the one that preceded it. The net effect is that the last cylinder to close is “topped off” at which point its CO2 is harvested.

The lab also has developed a concept that employs a sorbent fabric belt in a continuous loop, with half the loop exposed to the open air collecting CO2 and the other half in an air-tight container where it is wetted to release the gas.

Both of these designs have been sized to produce a CO2 concentration of 5 percent, roughly 125 times higher than what is found in ambient air. That’s rich enough to allow further concentration through conventional means.

## Optimize and Deploy

For decades, Lackner’s vision of artificial trees pulling CO2 from the ambient air seemed closer to a hallucination than a practical engineering project. But that’s changing fast. One sign of the change is the presence of Robert Page, a project manager with a background in product development, working with ASU under a contract funded by the Salt River Project, the Phoenix-area utility.

Page was brought in to help shepherd the artificial tree technology from the laboratory to mass production. One of the first decisions was which concept to build on. In many ways, the closed-loop system seemed to offer some advantages in scalability. In the near term, however, the team decided to go ahead with the stacked-disk system. Though it will require some sophisticated controls to approximate a continuous output from what is essentially a series of batch processes, it was judged to be simpler, less expensive, easier to manufacture, and easier to transport and setup.

The team has now produced a full set of drawings, specifications as well as a bill of materials. The next step is to build prototypes and begin testing. According to Page, the goal is to have a few units ready to deploy by 2020.

Producing a system that operates along the lines of a natural system, while elegant in conception, does involve certain compromises, particularly with regard to maintaining a constant stream of gas, as most production systems would be expected to do. Another challenge is where to site those 100 million shipping container-size devices. Given the sorbent cycle’s dependence on drying, deserts seem ideal, but many windy areas also would be suitable.

Lackner and his team are also looking to further optimize the system, evaluating ways to improve the sorbent or increase the speed at which it pulls in or releases CO2.

The key, however, is to start pulling carbon dioxide out of the air as quickly as possible. Lackner’s artificial trees will never be as elegant as an aspen or as majestic as a sequoia, but if they can help limit the damage from too much CO2 in the atmosphere, they will be beautiful enough.