Solid Sorbents for Post-Combustion Carbon Capture
There is a lot of interest in post-combustion carbon capture, because it can be (relatively) easily added to existing fossil-powered generating plants. The most mature technology is adsorption of CO 2 by aqueous amine solution. But the energy penalty associated with the aqueous amine cycle — on the order of a 30 percent reduction in electrical output — is considered by most to be too high for wide adoption of the method in power plants. Solid-supported amine sorbents promise to reduce the energy penalty by virtue of the fact that they have a much lower thermal mass per adsorption site, meaning that they don't have to bleed as much steam from the power generating process to desorb CO 2. Additionally, they are less corrosive than the aqueous amines, leading to lower capital costs.
The US-DOE's Carbon Capture Simulation Initiative aims to reduce the development time for the adoption of new carbon capture technologies through multi-scale simulation. The technology of focus for the first phase of the initiative is solid sorbents. The primary role of the ESMS Group within the initiative is to provide physico-chemical models for the sorbents. These models must be physically rigorous -- such that they will be able to benefit from first-principles calculations -- while also being computationally fast enough to operate in the context of coarse-scale CFD and process simulations. (The latter requirement generally calls for the use of reduced-order models.)
The specific sorbent of focus is polyethyleneimine (PEI)-impregnated mesoporous silica. PEI is an amine polymer. These sorbents have a relatively high amine loading (due to the relatively high density of amine sites in PEI) and a potentially high internal surface area due to a strong physical interaction between the polymer and the silica surface, which leads to a coating of the pore walls.
The combination of silica, PEI and pore space within the particle leads to a fairly complex microstructure, and a three-scale, semi-empirical model has been proposed to account for it. The first and largest scale is that of a macroporous agglomerate, composed of many smaller (1-10 microns) mesoporous particles. The assumption is that gas-phase transport on this level is very fast compared with other transport and chemical steps in the process. The intermediate scale is that of the mesoporous particles. Gas-phase transport at this scale is best described by Knudsen diffusion (a type of gas diffusion in which collisions between gas molecules and pore walls are much more frequent than collisions between molecules), through transport pathways coated with PEI. The smallest scale is that of the PEI-silica composite found between the pores within the mesoporous particles. These three scales are illustrated below.
The key mechanistic aspect of the model is the diffusion mechanism. Based on a
thirty-year-old kinetic study of alkanolamines in aqueous solution, many believe
that when carbon dioxide adsorbs to an amine site at the PEI-gas interface, it
forms a stable intermediate species called a zwitterion. The zwitterion can then
donate a proton to another amine site to form a stable ionic pair called carbamate,
transfer its adsorbed CO
2 to a free hydroxide ion (when water is present) to form bicarbonate, or
hop from one free amine site to another, thereby traveling through the PEI bulk.
However, several quantum studies — including one conducted in the context of CCSI
by Joel Kress at Los Alamos National Lab — have shown that the zwitterion is not
stable in a dry polymer environment. Diffusive intermediates involving water have
been found, however, leading to a water-dependent diffusive mechanism in the sorbent.