OUR lungs separate, capture, and transport carbon dioxide (CO2) out of blood and other tissues as part of the normal respiration process. The catalyst that initiates this natural response in the lungs is carbonic anhydrase, the fastest operating natural enzyme known.
Other enzymes play an “energy” role in our bodies as well. For example, ribulose-1,5-bisphosphate carboxylase oxygenase, more commonly known as RuBisCO, catalyzes the first major step of carbon fixation. In that process, molecules of atmospheric CO2 are made available to organisms in the form of energy-rich molecules such as glucose. Methane monooxygenase, or MMO, oxidizes the carbon–hydrogen bond in methane.
Medical researchers have used these enzymes as guides for designing synthetic catalysts that speed up chemical reactions. Now, a collaboration led by Lawrence Livermore is examining carbonic anhydrase as the basis for a new molecule that does for coal-fired power plants what the enzyme does for our bodies: quickly separate CO2. But instead of transporting it out of blood or tissue, the catalyst will remove the greenhouse gas before a power plant emits it to the atmosphere. (See box below.)
“Developing a synthetic molecule to replace CO2 scrubbing processes that use amines could greatly speed up carbon capture,” says geochemist Roger Aines, the principal investigator for the catalyst project. “Current analysis indicates that efficient catalysts might increase the capture rate for CO2 separation by as much as 1,000 times.”
It Takes a Team
Computational biologists and synthetic chemists in Livermore’s Physical and Life Sciences Directorate are working with the Laboratory’s powerful supercomputers to computationally design hundreds of molecular compounds and synthesize the most-promising candidates for laboratory testing. Babcock & Wilcox’s Power Generation Group, a leading supplier of steam-generation and environmental equipment for the electric utility market, will provide benchtop and full-scale testing and process modeling to determine which molecules can be implemented in existing and new processes. The University of Illinois will design and fabricate specific applications for the catalysts selected.
Before the ARPA-E project began, Livermore was already working on methods to design and synthesize such a catalyst through Laboratory Directed Research and Development funding. Results from this early effort showed that key technologies must be improved for global air-capture systems to be feasible. In particular, researchers must speed up the chemistry of the CO2 removal process to keep the capture device to a manageable size. If a new molecule is tough enough, it might also enable the direct capture of CO2 from such emission sources as airplanes and home heating systems.
Supercomputers Aid the First Steps
Felice Lightstone, a computational biologist at Livermore, is leading the initial effort to evaluate potential candidates. “We are developing a computational library of molecules, all of which are designed to protect the zinc ion that activates the catalyst,” says Lightstone. “Using quantum molecular calculations, we can experiment with different configurations, placing nitrogen atoms at various angles and distances from the zinc. The idea is to build scaffold structures that protect the metallic ion in a power plant’s harsh operating environment and optimize the molecule for capturing CO2.” The computational team, which includes Sergio Wong, Lawrence Fellow Yosuke Kanai, and Donghwa Lee, also examined cobalt as a replacement for zinc but found it to be less effective in many configurations.
Quantum molecular calculations can predict activity at the electronic scale—a dimension comparable to about 1/50,000 of a slice of human hair—allowing scientists to examine processes occurring at dimensions that experiments often cannot create. Kanai and Lee are applying density functional theory (DFT) to model the atomic motion and complex dynamics of each candidate’s material interactions. DFT starts from first principles. That is, it uses the laws of quantum mechanics without any ad hoc assumptions to describe the electronic density of a molecular or condensed system.
Kanai notes, however, that at the scale of molecular vibrations, chemical reactions are rare events. A simple application of first-principles molecular-dynamics simulations does not allow researchers to investigate reaction kinetics, which largely determine catalytic reactivity. “To address this challenge, we interface DFT with other computational tools such as the ‘nudged elastic band’ and ‘string’ methods to locate the reaction coordinates and obtain the reaction barrier.”
The ARPA-E team is examining two possible molecular designs. One is a relatively simple dissolved catalyst system that could be applied immediately in industrial practice. This technology, known as regenerable solvent absorption technology, or RAST, is being developed largely by Babcock & Wilcox. The second, a Livermore design, is a “tethered” molecule that holds the catalyst at the air–liquid interface where the CO2 transfer typically takes place. The tethered molecule looks much like mosquito larvae floating just below the surface of water. This approach promises very high efficiency, but using it in power plants may require changes in industrial practices.
Challenges to Overcome
Addressing structural robustness and fast catalytic rates would normally be a slow, expensive process. Because of Livermore’s computational and synthetic chemistry capabilities, the ARPA-E team can quickly evaluate hundreds of candidate compounds computationally, synthesize dozens, and test the most promising ones in the laboratory. Aines estimates that in just two years, the team will be ready to conduct long-term stability experiments on candidate molecules in large-scale testing facilities.
In addition, catalysts for the tethered molecule design must remain within about 100 micrometers of the gas–water interface, where they are most effective. If the catalyst is distributed throughout the solvent, more of it must be produced overall. The team is investigating an approach that adds a hydrophobic molecule to tether the molecule at the gas–water interface. Livermore’s preliminary calculations show that such tethers do not deform the catalyst and should preserve full functionality. Another design possibility uses very small particles containing the catalyst on their surface. These particles move with the solvent and can be easily extracted before thermal desorption.
As candidate molecules move closer to commercialization, team members at Livermore and Babcock & Wilcox will work together to balance the cost of catalyst production with the molecule’s expected lifetime. “For now, we are estimating that a catalyst will live at least a few days, possibly longer,” says Aines. “Surviving the high temperature is the greatest challenge in designing an effective catalyst and will be the limiting factor with this technology.”
Molecules Made to Order
For this project, Valdez is applying what he learned during his postdoctoral research at the Scripps Research Institute under K. Barry Sharpless, who in 2001 won the Nobel Prize in Chemistry. Sharpless is the father of “click chemistry,” a process designed to speed up drug discoveries by building chemical libraries and rapidly screening them for molecules with a desired activity. Click chemistry changed the search for effective drug formulas from a needle-in-a-haystack operation, in which numerous attempts led to only a few successes, into a winnowing process that more quickly identifies a successful candidate.
The ARPA-E researchers want to achieve a similar advance with their combinatorial approach. Their goal is to ensure that each reaction in the multistep synthesis process is fast, efficient, and highly predictable. “The computational aspect of the project involves arbitrarily shuffling and arranging atoms around the metal center,” says Valdez. “For the synthetic chemistry phase, I use an array of scaffolds that have been developed without regard for their linkage stability or synthetic tractability.”
Combinations that are not stable or would require an extensive construction process are eliminated from the list of candidates for synthesis. Only about 2 percent of the computationally derived scaffolds have made it to the synthesis stage, which is currently under way.
Synthesized molecules go through two tests, run by Livermore chemist Sarah Baker, to evaluate how well the candidates actually work. In the first test, a stopped-flow spectrophotometer determines a candidate’s absolute hydration kinetics and stability. The spectrophotometer heats the dissolved catalyst to 80°C and mixes it with CO2 gas in less than 500 microseconds. The color of the resulting solution indicates the rate of CO2 capture. In the second test, a wetted wall apparatus built at Livermore measures the combined rates of all chemical and mass-transfer processes, including the concentration of CO2 emitted by the operation.
Even when combinations of bond length and angles cannot be synthesized, computational methods such as click chemistry allow the team to consider options more quickly. “We originally thought we could synthesize five new molecules a year,” says Aines. “By combining advanced computation and experimental tools, we now expect to test up to 20 specific examples a year. The synergy exhibited in our team’s multidisciplinary approach is critical to our success.”
Aines notes that one day, the technology could be fast enough to remove CO2 directly from the atmosphere. Until then, researchers will focus on capturing this greenhouse gas at the source—removing it from the numerous industrial processes we rely on every day.
Key Words: carbon capture, carbon dioxide (CO2), carbonic anhydrase, density functional theory (DFT), greenhouse gas.
For further information contact Roger Aines (925) 423-7184 (firstname.lastname@example.org).
Lawrence Livermore National Laboratory
Privacy & Legal Notice | UCRL-TR-52000-11-3 | March 15, 2011