Electrical Effects Of Carbon In Rocks
Changes in electrical resistance during rock fracture in the presence of a carbonaceous atmosphere have been investigated using a quartz sandstone with hematite-bearing cement. The experiments were performed in an internally heated gas pressure vessel with a load train that produced strain rates between 10-6 and 10-5/sec. Samples were deformed at temperatures of 350-540 deg.C and pressures of 100-150 MPa in atmospheres of CO2 + 5% CO. These conditions are within the field of graphite stability. An increase in the uniaxial load during the experiment was accompanied by a smooth increase in resistance. This is interpreted as due to displacement of the cement during flattening and packing of quartz grains. As load was increased, small and sudden decreases in load that quickly recovered were recorded, apparently due to formation of microfractures. These events were accompanied by small decreases in resistance. During dilatation preceding and concurrent with catastrophic failure, resistances decreased by up to 40%. Experiments were also conducted on the Westerly granite and on a hematite-free sandstone. Neither exhibited changes in conductance on failure. The samples were examined by x-ray photoelectron spectroscopy, which revealed the presence of thin carbonaceous films on fracture surfaces.
Carbon films, typically tens of nm thick, are common in micro-fractures in rocks and may control electrical conductivity over cm-scale distances. These observations and our experimental results lead to the hypothesis that, as microfractures open in the time leading up to failure along a fault, carbon is deposited as a continuous film on the new mineral surfaces and conductivity increases. Subsequent changes in conductance occur as connectivity of the initial fracture network is altered by continued deformation. Such a process may explain some premonitory electromagnetic effects associated with earthquakes and may significantly affect middle to lower crustal electrical conductivity.
Related work on the effect of pressure on conductivity of metamorphic rocks has identified enhancement by the presence of carbon in rocks from the German KTB.
Collaborators - Catherine McCammon, Falko Langenhorst, Brent T. Poe, David C. Rubie, and Yousheng Xu, Bayerisches Geoinstitut of the Universität Bayreuth; Alfred G. Duba and Jeffrey J. Roberts, Lawrence Livermore National Laboratory; Edward A. Mathez and Rosamond Kinzler, American Museum of Natural History; Gary D. Egbert, Oregon State University
Publications and Presentations, 1998
Electrical Conductivity Of The Crust
The usual geophysical approach for obtaining a quantity in the Earth's interior when it is not directly observable is to obtain related properties by measurements in the field and then use a physical property determined from laboratory experiments (or, when all else fails, from theory) to relate the two properties. An example might be calculating mantle temperatures using electrical conductivity depth profiles and laboratory conductivity temperature relations. However, electrical conductivity of the lower continental crust has presented a persistently irksome problem for interpretation. From geophysical techniques such as magnetotellurics the lower crust appears to be a fairly good electrical conductor having conductivities above 10^-6 S/m, usually in the range 10^-4 to 10^-2 S/m; yet laboratory measurements on typical crustal rocks seem to fall orders of magnitude below these values unless temperatures approach the melting point. We are driven to hypothesize that in situ crustal rocks contain other conducting phases that seem absent in most specimens used for laboratory conductivity measurements. In recent years the principal candidates for the elusive conductive phases are saline fluids or solid carbon (amorphous or graphitic) precipitated from a CO2-rich fluid during cooling. Because seawater and silicate melts have conductivities of the order of 3 S/m and for graphite it is about 104 S/m, only small quantities of either material are required to explain the small conductivities above provided that the conducting phases are interconnected.
Mid-crustal rocks freshly cored from depth in the German continental scientific drilling site (KTB) provide a means to study transport properties in relatively unaltered samples resembling materials in situ Electrical conductivity (s) was measured to 250 MPa pressure on 1 M NaCl-saturated amphibolites from 4 - 5 km depth. The unexpected feature was an increase of s with pressure P that appeared (anisotropically) in most samples. To characterize this behavior we fitted logs vs. P to obtain two parameters: the zero-pressure intercept s0 and the slope dlogs/dP (of order 10-3 MPa-1). Samples of (+) and (-) slopes behave differently. The result for s0 is that samples having negative slopes show strong correlation of s0 with a fluid property (permeability) indicating the usual fluid-dominated behavior correlates with larger s0. In contrast, samples with positive P-shift do not correlate with permeability, indicating that fluids are less important to (+) pressure behavior. As shown in Figure 1 a result for dlogs/dP is that samples of negative P-shift have slopes uncorrelated with initial conductivities. Again in contrast, samples of positive dlogs/dP have greatest P-dependence for lowest initial conductivity s0, i.e., the less fluid the more positive the P-shift. Thus, positive P-shift is consistent with reconnection of solid phases into a conductive texture closer to that of rock at depth. Further, electron microscopy reveals the presence of carbon on internal cleavage surfaces in amphibole, the most abundant mineral phase. Thus, carbon probably dominates the reconnection, but total s still involves fluids as well as ore minerals. For the KTB location it is tempting to say that the reason why mid-to-deep crustal electrical conductivities modeled from geophysical measurements are so much higher than usual conductivities of silicates is the presence of interconnected good conductors in the solid phase.
For some time it has been suspected that transport properties such as electrical conductivity, which are very sensitive to trace phases, may not be properly described by measurements on weathered specimens collected from near-surface outcrops. Thus, the explanation for the lack of positive pressure effects in surface metamorphic rocks is that these rocks have had their aligned, interconnected conductors partially destroyed during denudation. Such rocks have been altered by pressure- and temperature-release cracking and by geological processes, e.g., retrograde metamorphism and oxidation that can easily obliterate coatings of good conductors during their transport to the surface. When saturated with fluid, such a rock then exhibits "normal", negative conductivity decrease with pressure. Despite suffering some alterations, freshly cored rocks better approximate the real materials at depth. Although these cores come from only one site in central Europe, they are of general relevance to the long-standing problem of explaining why crustal electrical conductivities are so high compared to their constituent silicate minerals.
External Reviews - Peer reviews through Office of Basic Energy Sciences, DOE; Office of Program Analysis, DOE; papers peer reviewed.
Collaborators - A. G. Duba, Lawrence Livermore National Laboratory; E. A. Mathez, C. Peach, American Museum of Natural History; G. Nover, S. Heikamp, Mineralogisches Institut, Universitat Bonn.
Shankland, T. J., A. G. Duba, E. A. Mathez, G. Nover, and S. Heikamp, Effects of Solid and Aqueous Phases on Electrical Conductivity of Freshly Cored Metamorphic Rock, Eos, Trans. AGU, 75, Supplement, p. 340 (1994).
Shankland, T. J., A. G. Duba, E. A. Mathez, G. Nover, and S. Heikamp, Evidence for both Fluid and Solid Electrical Conductors in Freshly Cored KTB Rocks, Eos, Trans. AGU, 75, No. 44, Supplement, p. 676 (1994).
Mathez, E. A., A. G. Duba, C. L. Peach, A. Leger, T. J. Shankland, and G. Plafker, Electrical Conductivity and Carbon in Metamorphic Rocks of the Yukon-Tanana Terrane, Alaska, J. Geophys. Res., 100, in press (1995).