Los Alamos National Laboratory
Lab Home  |  Phone
LANL: >> EES Home
EES Group Pages

The Puzzle of Unexpected Disorder: Probing Rocks with Neutrons

James A TenCate, Timothy W. Darling, Thomas Proffen
November, 2005

Neutrons give Earth scientists a unique opportunity to look at the structure of rocks in ways never before imagined. Neutrons can probe the atomic structures and dynamics of ordered and disordered materials such as crystals and glasses (e.g., quartz and obsidian), and combinations of the two. Because they have no charge and scatter only from atomic nuclei, neutrons can penetrate deeply into most materials, including intact rocks. This is important. The grains and bonds at the surface of a rock behave differently from the interior grains because they are exposed to air and humidity and they are not surrounded by other grains. X-rays, a competing way to look at atomic structure, do not scatter effectively from lighter elements like Calcium and Silicon (some of the most abundant elements on Earth) nor can they be readily used on intact samples.

The physics of why rocks behave the way they do is still not well understood. Rocks are peculiar granular materials; they are nonlinear (i.e., not perfect "springs") and hysteretic (changes of length lag changes in force). They also display some unusual memory effects (see "The Memory of Rocks" Physics Today, April 1999). For example, an acoustic wave traveling through a rock can soften the rock. This softness goes away only gradually; the rock can take from several hours to a day to recover (a memory effect). Rocks are some of the easiest and most accessible granular materials to study and their peculiar behavior can easily be measured. Moreover, rocks are of great interest to Los Alamos as a part of a larger class of granular materials, varying from oil- and gas-bearing rocks to concretes to high explosives.

Our previous neutron work [Geophys Res. Lett., 31, L16604, (2004)] shows that it is the "cement" or bond structure that holds a rock together and gives the entire rock its peculiar properties. Contrary to the widely held belief that some rocks are entirely crystalline (e.g., sandstone), we considered the possibility that the contact area responsible for a rock's odd behavior might in fact be non-crystalline/glassy. Why? Glassy materials are known to exhibit peculiar nonlinear behavior. By the way, careful examination of a thin-section photomicrograph like the one shown in the figure suggests that there is no "glass" in the rock.

Thin section (polarized) of a pure Quartz sandstone. Careful examination shows no amorphous/glass present.

If all or some of the atoms are not in an ordered array like found in a crystal, Pair Distribution Function (PDF) analysis of the neutron data can tell us average local distances between atoms – suitable for analyzing glasses or liquids or disordered materials in general. We decided to use the Lujan Center's NPDF beamline at LANSCE to look for disordered/non-crystalline/glassy bits in pure quartz sandstones. The NPDF beamline has been upgraded from a high resolution powder diffraction line to a specialist PDF machine for local structure studies and is arguably the best machine in the world for such a study.

We acquired the best statistics data in existence (24-hour averaging) on an intact rock, a pure quartz (99+%) Fontainebleau sandstone. The techniques for separating crystalline and noncrystalline component fractions from PDF analysis produced a surprising result: in pure silica rocks, there is an unexpectedly large contribution (5-10%) from the "glassy bits" (i.e., silica molecules that are not part of long-range crystal ordering) [Geophys Res. Lett., 31, L24606, (2004)]. Figure 2 shows the data inconsistency that leads to this conclusion. Not surprisingly, several petrologists immediately questioned our measurements and even the validity of our PDF analysis. However, our techniques were validated by (1) measuring the known crystalline and non-crystalline components of a bulk metallic glass (one impregnated with a known quantity of crystalline reinforcements) [Z. Kristallogr. 220 (2005)] and (2) showing that powdered natural quartz crystals are entirely crystalline. Our initial results still stand: natural silica rocks contain a non-crystalline component and this result is a significant finding on the accepted composition of these rocks. No other technique could have revealed this.

PDF plot (atom-pair-separation distance along the bottom axis). The data are red, the model (pure crystal) blue. Note the excess intensity measured in the first two peaks that are NOT part of the crystalline model. These peaks are related to the distance between Si and O atoms and O and O atoms which are not part of the general crystal structure of the grains making up the rock.

Several questions persist and possible applications suggest themselves. How much "disorder" can we see with neutrons and why do other intact natural crystalline rocks show similar results while others don't? If we can use neutrons to see very subtle levels of disorder we might have a new way to date a rock: the transformation of a glassy-like opal to crystalline quartz takes time. On the other hand, we may, as one petrologist still thinks, still find something important has been ignored in the way the neutron analysis is applied to rocks and other similar intact granular materials. Either way, the results are exciting. The mystery persists and more research awaits!

For more information write to tencate@lanl.gov

Funding Acknowledgments: LDRD-ER, NSF (NPDF beamline upgrade), NASA (summer students), OBES (JT)

>> EES Home

Operated by Los Alamos National Security, LLC for the U.S. Department of Energy's NNSA

Inside | © Copyright 2010-11 Los Alamos National Security, LLC All rights reserved | Disclaimer/Privacy | Web Contact