Center for Nanoscale Control of Geologic CO 2 :

1. Fig. 5 Fig. 2 Figure 6: (a) Arrhenius plot of the temperature dependence of water diffusion coefficients in Vycor-confined H2O (Zanotti et al., 1999) and Vycor-confined H2O-CaCl2 (Mamontov and Cole, 2006). The same data for bulk water and H2O-CaCl2 are shown for comparison. (b) A molecular dynamics simulation snapshot of LiCl confined in a 1.4 nm cylindrical silica pore (section view). Figure 7: Snapshot of MD simulation of supercritical CO2 in contact with 2 M NaCl brine. Periodically-replicated simulation cell contains 572 water molecules, 20 Na+ and Cl- ions, and 288 CO2 molecules (red: oxygen, white: hydrogen, green: carbon, blue: sodium, yellow: chloride). Note that CO2-water interface is not sharp and that few CO2 molecules are dissolved. Center for Nanoscale Control of Geologic CO2: A new U.S. Department of Energy Frontier Research Center Donald J. DePaolo, Earth Sciences Division, E.O. Lawrence Berkeley National Laboratory* ABSTRACT Thrust Area 1: Controlling carbonate nucleation and growth Thrust Area 3: Emergent Processes Thrust Area 2: Structure, Dynamics, and Transport of Fluids in Nanopores and Thin Films The injection of CO2 into the Earth’s subsurface drives the fluid-rock system into “far-from-equilibrium” conditions. The fluxes that return the system to equilibrium are nonlinearly related to the driving forces (e.g., chemical affinities and gradients in the fluid pressures and chemical potentials). The nonlinear response results in emergent structures and self-organization. Flow, solute transport, colloid transport, mineral dissolution and mineral precipitation combine within the mechanical framework of the porous medium to generate precipitate structures in individual pores, precipitate structures correlated over many thousands to millions of pores, and immiscible fluid structures with fractal geometry over scales from millimeters to kilometers (Feder, 1988). The ultimate goal is to develop a new set of phenomenological laws based on insights gathered from the pore scale (and smaller) investigations that will improve our ability to forecast the fate of CO2 in actual reservoirs. The objective of the Nanoscale CO2 Center is to use new investigative tools, combined with experiments and computational methods, to build a next-generation understanding of molecular-to-pore-scale processes in fluid-rock systems, and to develop an ability to control critical aspects of these processes. The stated goal is to use molecular, nanoscale, and pore-network scale approaches to control flow, dissolution, and precipitation in deep subsurface rock formations to achieve the efficient filling of pore space with injected supercritical CO2, with maximum solubility and mineral trapping and near-zero leakage. Advanced knowledge of these small-scale processes is expected to lead to an improved predictive capability for reactive transport of CO2-rich fluid that is applicable for 100–1000 years into the future. The ability to seal deep reservoirs and prevent escape of gases and fluids is one requirement for geological CO2 sequestration. Reservoirs may be sealed, or the seal maintained, by suppressing dissolution of minerals or by causing new minerals to precipitate in existing pore space. To allow the introduced fluid initially to freely enter the formation requires that the pores remain open. But later, to maximize trapping, it is desirable to have the pore space fill in with secondary minerals that incorporate CO2. To manipulate the dissolution, nucleation and growth of minerals in the subsurface requires a detailed understanding of these processes and how they are affected by the special characteristics of porous rock media. The physicochemical properties of fluid CO2-aqueous solution mixtures strongly affect the behavior of fluid CO2 plumes injected into terrestrial geological formations (Fig. 2). The relatively low interfacial tension of fluid CO2-liquid water systems may allow supercritical CO2 to flow through cap rocks or aquitards by buoyancy-driven advection. Fluid CO2-water interfacial properties will determine the importance of “capillary trapping” at the trailing edge of flowing CO2. Fluid CO2-water interfacial properties also may control the uptake of water and ions through the liquid water-fluid CO2 interface, which in turn will strongly affect the reactivity of fluid CO2 with host-rock solid. As the relatively nonwetting CO2 phase is displaced into pores previously filled with subsurface brine, residual saline aqueous films are expected to form (Fig. 5), coating mineral surfaces over time scales that are currently unknown. These thin films of native fluids impose diffusive interfaces between the invading fluid and minerals, thus mediating early stages of fluid CO2-mineral interactions. Over larger than pore-scale distances, mechanical properties of thin films along with capillary forces control the pore-scale advance of fluid fronts, hence the evolution of invading fluid flow paths extending up to reservoir-scale behavior. Exerting molecular level controls. We are exploring novel approaches that may enable control over the spatial and temporal extent of subsurface carbonate mineralization. (1) Using the unique properties of nanoscale materials to modulate fluid composition by introducing trace or minor constituents (Mg, Sr, etc.). (2) Using as inspiration the remarkable level of control that biomolecules can exert over timing, location, and rates of mineralization. The slow adsorption dynamics, strong electrostatic interactions, and tendency towards surface aggregation that are peculiar to highly-charged peptides and proteins lead to controls over nucleation and growth not found in inorganic systems (De Yoreo et al, 2007). (3) Investigating the detailed structure and molecular-scale functioning of individual microorganisms and consortia that thrive in harsh environments, so that it will be possible to manipulate their activities to alter subsurface geochemistry and mineralogy. Understanding the nonlinear dynamics occurring at the pore scale will require a new generation of experimental, imaging and modeling tools. The rapidly developing field of 3D imaging techniques (X-ray, neutron, and magnetic resonance) allows the characterization and observation of dynamic geological and hydrological processes on unprecedented temporal and spatial scales. XCT imaging of reactive flow experiments will be employed, as well as neutron computed tomography (NCT), using the High Flux Isotope Reactor (HFIR) at ORNL, and Small Angle Neutron Scattering (SANS). To complement the imaging, a new generation of pore-scale reactive transport modeling tools, including Lattice-Boltzmann methods, hybrid methods and Direct Numerical Simulation need to be developed and employed. Figure 8: Clipped views of 3D idealized pore geometry represented as a microtube randomly packed with glass beads. Fluid pressure computed by solving the Navier-Stokes equations using Direct Numerical Simulation. Spheres range in diameter from 10 to 30 microns. Figure courtesy of David Trebotich. Thermophysical properties of confined CO2-rich fluids: Since water, carbon dioxide, and hydrocarbons are miscible only to a limited extent, subsurface fluids will be composed of mixed coexisting phases. Equilibria between the components of these phases are critical to system behavior and its evolution in time, since they affect the hydrodynamics of flow through both fractured and porous systems, surface wetting, precipitation and dissolution of solids, and chemical reactivity. To address these issues, a multi-property experimental approach is proposed. Experimental studies of the PVT properties, phase equilibria, fluid conductivities and enthalpies of binary, ternary and more complex CO2-rich fluid mixtures in the 25–100°C, 1–500 bar range will serve as a foundation for exploration of emergent macroscopic behavior originating at the molecular level in close vicinity of solid-fluid interfaces. Figure 9: Tracer diffusion through weathered basalt. A. XRF image of bromide (light) imaged with Beamline 10.3.2 at the ALS, LBNL; B. 3D numerical simulation of front position using a pore network constructed from XCT data using Beamline 8.3.2 at the ALS. From Sitchler et al., 2008. Research Integration Figure 1: Factors controlling CO2 behavior in geologic formations emerge by propagation of molecular and nanoscale phenomena through complex porous networks. Distribution of CO2 at the reservoir scale (a.) results from its propagation through pore networks (b.), which in turn are controlled by pore-scale processes of capillary breakthrough (c.), multiphase viscous flow (d.), diffusive exchanges and reactions across CO2-water interfaces (e.), and in nanopores (f.). Nano-scale processes (g., h., i.) link emergent macroscopic processes to molecular-scale mechanisms. Thrust Areas 1 and 2 of the proposed Center address molecular and nanoscale processes. Area 3 addresses the pore network and larger scales. Virtually every component of the research is dependent on, or must consider results and information from, almost every other component. The dynamics of CO2 solutions in nanopores is linked to both inorganic and organic controls on mineral nucleation, depends on the thermophysical properties of the fluids, and will affect emergent properties and processes. Biomimetic controls on nucleation and growth must consider inorganic controls, the effects of nanophase nutrients, thermophysics of confined fluids, the nature of CO2-H2O interfaces, and thin films. Emergent pore-scale dynamics will depend on virtually everything else. Macroscopic behavior, observed in nature and experiments, or predicted from models, may impact the research agenda and concepts used in all of the other tasks. Other approaches to be investigated include the role of trace elements in controlling mineral growth; the structure and dynamics of the solution-solid interface, and nanoparticle aggregation. The research plan of the Center is predicated on the idea that in order to achieve the full potential for subsurface CO2 storage (10 to 40% of the needed reductions in CO2 emissions over the next century; IPCC, 2005), it will be necessary to use as much of the available subsurface pore space in sedimentary formations as possible. This means that, in addition to the use of depleted oil and gas reservoirs, for which considerable characterization has been done and experience exists, it will also be necessary to use a wide range of other sedimentary rock formations, most of which have not been previously characterized, and most of which have pore space filled with salt-rich aqueous fluid (brine) that would be displaced by injected CO2. Hence the Center’s research is primarily focused on the processes and properties relating to CO2 in saline formations. Figure 3: Schematic of high temperature, pressure vessel to probe microbial growth and geochemical effects under sequestration-like conditions. Molecular Simulations: Molecular-dynamics simulation involving realistic models can be combined with experiment in order to gain atomic-level understanding of the mechanisms underlying the structural and dynamical phenomena of interest, to aid the interpretation of data such as those from neutron scattering, and ultimately, to guide the corresponding macroscopic modeling of thermophysical data. The major technological gaps to controlling and ultimately sequestering subsurface CO2 can be traced to far-from-equilibrum processes that originate at the molecular and nanoscale, but are expressed as complex emergent behavior at larger scales. Essential knowledge gaps involve the effects of nanoscale confinement on material properties, flow and chemical reactions, the effects of nanoparticles, mineral surface dynamics, and microbiota on mineral dissolution/precipitation and fluid flow, and the dynamics of fluid-fluid and fluid-mineral interfaces. The construction of quantitative macroscale process models based on nanoscale process descriptions is a critical additional need. Figure 4: Peptides and proteins as switches, throttles and brakes: In vitro studies of carbonate nucleation and growth with peptides and other organics have established a number of important controls: (1) Aspartic-rich polymers stabilize a dense liquid phase, which transforms to ACC before slowly crystallizing (Gower and Odom, 2000). A–C illustrate representative stages of transformation from ACC nanoparticles to calcite and D shows calcite of various orientations (red, blue, violet) growing out of an ACC film (purple background) formed from a dense liquid phase. (2) Functionalization of surfaces with organic films results in crystallographic control (Aizenberg et al, 1999). In E, a carboxylic SAM directs crystallization on the calcite (012) face. (3) Peptides accelerate growth kinetics at low concentrations and/or high supersaturations by lowering the activation barrier to solute attachment, but inhibit growth at high concentrations (Elhadj et al., 2006a, b). Plot of calcite step speed vs. peptide level (F) shows this effect. The slow adsorption kinetics of macromolecules results in a crossover in the time scales for surface coverage by macromolecules and the exposure time of a terrace before coverage by new crystals (DeYoreo et al., 2007). As supersaturations drops, this competition leads to a fast switch from growth at full speed to complete inhibition (G). Scale bars in mm: A–0.5, B–0.75, C–1.75, D–20, E–30. REFERENCES *Center for Nanoscale Control of Geologic CO2 key personnel: Director - D. DePaolo, Co-Director - J. DeYoreo (LBNL); Research Area Leads – K. Knauss (LBNL), G. Waychunas (LBNL), J. Banfield (UCB/LBNL), A Navrotsky (UC Davis), F.J. Ryerson (LLNL); G. Sposito (UCB/LBNL), T. Tokunaga (LBNL), D. Cole (ORNL), C. Steefel (LBNL), D. Rothman (MIT), S. Pride (LBNL). Department of Energy (DOE) (2007) Basic Research Needs for Geosciences: Facilitating 21st Century Energy Systems, Report from the Workshop Held February 21-23, 2007, Office of Basic Energy Sciences, 186 pp. plus appendixes, available at <http://www.sc.doe.gov/bes/reports/files/GEO_rpt.pdf>. Intergovernmental Panel on Climate Change (IPCC) (2005) Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz, B., O. Davidson, H.C. deConinck, M. Loos, and L.A. Meyer, eds., Cambridge University Press, New York, 442 pp.