Earth science

Missing link in mantle dynamics

Article metrics

The discovery of crystallographic imperfections known as disclinations in the most profuse mineral in Earth's upper mantle has the potential to solve a problem that has vexed mineral physicists for decades. See Article p.51

The viscosity of Earth's upper mantle controls a wide range of processes, from the attenuation of seismic waves and the rate of surface deformation after earthquakes to the slow, global-scale flow that is associated with mantle convection and the dynamics of tectonic plates. This viscosity is logically interpreted to be dominated by the physical properties of olivine, the most abundant mineral in Earth's upper mantle, as well as in those of the other terrestrial planets (Mars, Venus and Mercury) and the Moon. In this issue, Cordier et al.1 (page 51) report how new techniques to analyse the microstructure of grain boundaries in olivine allowed them to discover crystal defects called disclinations in this mineral. This observation is probably a first for geological materials, and has ramifications for our understanding of the processes that control mantle dynamics.

More than 50 years of effort have gone into the experimental and theoretical characterization of the solid-state flow properties of olivine. Motivated by the principles of materials science, a sizeable fraction of these studies has concentrated on understanding the relationships between the mobility of crystal defects, primarily dislocations, and the creep behaviour (deformation that occurs under continued stress) of both olivine single crystals and olivine aggregates, which are known as peridotites.

Owing to the scales of deformation in the Earth, the results of experiments on olivine must be extrapolated by many orders of magnitude in spatial scale, timescale or stress. Nonetheless, excellent agreement exists between predictions based on such extrapolation and independent geophysical observations. Furthermore, the details of microstructures in mantle samples collected from mountain belts and xenoliths (pieces of rock from great depth that are brought to the surface in volcanic eruptions) show many striking similarities to those observed in experimental samples, providing a first-order validation for the extrapolation. However, the application of these relationships to mantle dynamics remains compromised because a fundamental principle for grain-scale deformation of aggregates, the Von Mises criterion, has been hard to reconcile with these laboratory experiments.

The Von Mises criterion states that five independent slip systems — combinations of crystal planes and directions in which defects move — are required to allow homogeneous grain-scale deformation of polycrystalline aggregates; this requirement can be relaxed to four slip systems for inhomogeneous grain-scale deformation2. However, extensive work on olivine single crystals suggests that only three independent slip systems exist for reasonable stress states in both laboratory and natural conditions. This has led to suggestions that deformation accommodated by the diffusion of mineral components through a process called dislocation climb, or through grain-boundary sliding3,4, might account for the 'missing' grain-scale deformation required by the Von Mises criterion. These hypotheses are challenged by the extremely sluggish diffusion kinetics observed for olivine, and the limited range of grain size where evidence for grain-boundary sliding is observed. Now Cordier et al. suggest that the motion of grain-boundary disclinations could be a solution to the problem.

As noted by Cordier and colleagues, the idea of disclinations was first described5 more than 100 years ago. However, awareness of their significance has been growing in recent years, owing to advances in microscopy6 and continued development of theory for the defect structure of grain boundaries7. Disclinations (and dislocations) can be visualized as linear defects related to the distortion of a cylinder; for example, a twist disclination is defined by a rotation on an axis perpendicular to a cut that runs along the longitudinal axis of the cylinder (see Fig. 1 of the paper1). Stress concentrations around these linear defects allow intracrystalline deformation at stresses much below the theoretical stress required to break a plane of atomic bonds in the crystal. In the present study, the authors took advantage of the high-resolution crystallographic technique of electron backscatter diffraction to identify grain-boundary disclinations in olivine (Fig. 1).

Figure 1: Imaging polycrystalline olivine.

Sylvie Demouchy

Cordier et al.1 have analysed images produced using electron backscatter diffraction to identify grain-boundary disclinations in olivine aggregates. Shown here is a cross-section of a sample deformed at high stress. Different colours denote different crystallographic orientations of the crystals that make up the aggregate. Scale bar, 20 μm.

Many questions remain about the efficacy of the disclination hypothesis for resolving the Von Mises problem in peridotites. For example, how does their density and mobility evolve during deformation and with changes in thermodynamic conditions, and how are the dynamics of disclinations influenced by the anisotropic (direction-dependent) elastic properties of olivine? How are grain-boundary disclinations involved with grain-boundary sliding? Can the inclusion of disclinations in polycrystalline-deformation models help to resolve questions regarding the evolution of lattice-preferred orientations in olivine aggregates deformed to high deformation? The latter problem is crucial for the interpretation of the seismic structure of Earth's mantle, which in turn is currently our best observational technique to explore convective motions inside Earth. Research on these topics promises to provide insights into the viscosity of the terrestrial planets and into the interpretation of a broad range of geophysical observations.


  1. 1

    Cordier, P. et al. Nature 507, 51–56 (2014).

  2. 2

    Hutchinson, J. W. Met. Trans. A 8, 1465–1469 (1977).

  3. 3

    Hirth, G. & Kohlstedt, D. L. J. Geophys. Res. 100, 15441–15449 (1995).

  4. 4

    Hansen, L. N., Zimmerman, M. E. & Kohlstedt, D. L. J. Geophys. Res. 116, B08201 (2011).

  5. 5

    Volterra, V. Ann. Sci. Ecole Norm. Sup. 24, 401–517 (1907).

  6. 6

    Murayama, M., Howe, J. M., Hidaka, H. & Takaki, S. Science 295, 2433–2435 (2002).

  7. 7

    Hirth, J. P., Pond, R. C. & Lothe, J. Acta Mater. 54, 4237–4245 (2006).

Download references

Author information

Correspondence to Greg Hirth.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hirth, G. Missing link in mantle dynamics. Nature 507, 42–43 (2014) doi:10.1038/nature13064

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.