Characterization of porosity in a Callovian-Oxfordian Clay (Tournemire Clay, France)


Supervisors: Dr. Guillaume Desbois and Prof. Janos Urai

Please send your application to: Dr. Guillaume Desbois.


Clay-rich formations are known for their low porosity and low permeability, forming seals for hydrocarbons, confining and separating aquifers, protecting deep groundwater resources from contamination and being a potential host for the long-term storage of radioactive waste. In Europe there are few underground laboratory (Mont Terri – Opalinus Clay, Switzerland; Mol-Dessel – Boom Clay, Belgium; Bure and Tournemire, Callovia-Oxfordian Clay, France), which investigate shale formation properties at different scale for potential application of deep geological storage. The aim is to analyze hydrogeological, geochemical and rock mechanical properties of such claystones and to observe how these properties change during and after gallery excavation, heating and emplacement of buffer material in order to evaluate the long term performance (> 500 000 years) of such a potential host rock. Particularly, the host rock has to delay and attenuate the radionuclides and other contaminants to be released into the biosphere after that the original engineered container starts to deteriorate turning to the decreasing of safety capacity. Nowadays, the host rock is considered as the main barrier against the migration of radionuclides. A major contribution to understanding the sealing capacity, coupled flow, capillary processes and associated deformation in clay is based on detailed investigation of the rock microstructure (Marschall et al., 2005; Esteban et al., 2006; Fanchi, 2010), including pore morphology and pore space.
Picture_1466 - MScPropGuillaumeTournemireClay
Figure 1:
(A) BSE-SEM micrograph showing mineral phase contrast density used to check relative mineral content (B) SE2-SEM micrograph showing surface topography used to detect pores in high quality 2D flat. (C) Segmented pore space imaged in (B). (D) The SUPRA 55 (Zeiss) scanning electron microscope (SEM) used to image pore microstructures.

Pore space is typically divided into size classes (Singh et al., 1985). These are (1) micropores, < 2 nm, close to the unit cell size of minerals, (2) mesopores, 2 to 50 nm, which are formed by the spatial organization of mineral grains, and (3) macropores, >50 nm, between grain clusters (Yven et al., 2007).
Measurements of porosity in claystones depend strongly on the method used, because different methods detect different classes of pores (NAGRA, 2002), and because most methods remove pore water usually by heating to 105°C. Drying of samples at 105°C removes free pore water and partly also the bound water, but it can also change the microstructure of the clay or damage organic particles (Bray et al., 1998). In addition, mercury porosimetry, gas adsorption, and ion exchange capacity all measure the connected porosity only. Porosity of clay are typically measured using densitometry, water loss, mercury injection and ion (Cl-) exchange; details of the different methods and results for the main european clay types are summarized in NAGRA (2002), Sammartino et al. (2003), Boisson et al. (2005) and Yven et al. (2007). If different methods agree that the most abundant pores are between 3 and 60 nm, conventional methods operates as a “black box” and then are mostly “blind”. The majority of pore space in clay has not yet been observed directly, in thin sections and optical microscopy or using a scanning electron microscope on broken or mechanically polished cross-sections (NAGRA, 2002).
Imaging of pore space in fine grained geological materials is a rapidly developing field since ion beam milling was shown to produce smooth and damage free surfaces (Holzer et al., 2007, 2010; Desbois et al., 2008, 2009, 2010b, 2011a; de Winter et al., 2009; Loucks et al., 2009; Holzer and Cantoni, 2011; Heath et al., 2011; Keller et al., 2011). Commercial FIB tools in SEM are based on Gallium ion sources (1pA - >50nA) to produce typical cross-sections of a few μm2; with serial sectioning for the investigation of microstructures in 3D (Holzer et al., 2007, 2010; Desbois et al., 2008; De Winter et al., 2009; Heath et al., 2011). Broad Ion Beam (BIB; up to few mA, Argon source) cross sectioners are commercially available as stand-alone machines, to produce polished cross-sections of much larger area than the FIB. A BIB gun was also recently included in cryo-SEM (Desbois et al., 2011b) but in its actual status, this prototype machine is not yet able to perform productive serial cross sectioning as FIB-SEM tools. However, BIB has two main advantages: (1) it is potentially less damaging since it is based on noble gas source and (2) it is able to produce polished surfaces up to few mm2, about hundred times bigger than those produced by FIB. This last feature is of particular interest for geomaterials since it fits better to the typical length scale range of microstructures and representative elementary area (REA).
In the frame of three PhDs, the characterization of porosity by BIB-SEM combination is actually performed at GED institute of RWTH on two main european clay host rock (Opalinus clay - Houben et al., submit and Boom Clay – Desbois et al., 2009; Hemes et al. 2011), as well as clay-rich gas-shale (Klaver et al., submit). However, no microstructural studies of the French clay have been completed. Thus, this project proposes to study microstructures of one core of Callovian-Oxfordian clay from Tournemire (france) using BIB-SEM approach.


The main objectives of this project are listed below:
  1. Qualitative characterization of the porosity and pore morphologies in dried samples from 2D BIB-polished cross-sections
  2. Classification and mapping of porosity related to mineralogy in 2D BIB-polished cross-sections
  3. Determination of representative elementary area (REA)
  4. Pores segmented in images will provide basis for quantitative characterization of the porosity in REA.
  5. Porosity and pore data obtained from mm2 area will be extrapolated to larger and smaller scales and compared to mercury injection porosimetry performed on similar samples for data upscaling.
  6. Comparison of resulting pore model pore based microstructure with other reference clay (Opalinus and Boom Clay)

Work plan

  1. Review of the relevant literature
  2. Selection of regions of interest and preparation of samples: The raw sample is an air-dried core of 50 cm long and 15 cm in diameter. The first step is to select few samples to be investigated along this core. The samples will be selected in regards to apparent variations in grain size, in mineralogy and layering. About 5 selected samples will be BIB cross-sectioned in order to prepare high quality surface suitable for SEM imaging of pore space down to SEM resolution.
  3. Microstructural study, pore microstructures digitizing and porosimetry: The combination of topography (SE2 images) and chemical analysis (BSE+EDX) information will be combined to describe and classify the porosity from 2D BIB cross-section down to nanometer scale, as well as the determination of REA by point counting method. Porosity will be automatically segmented for further quantifications. 2-3 representative samples will be selected to perform Hg-porosimetry (at Micrometrics company, Aachen)
  4. Microstructures interpretation: Pores segmented in section III will be used to quantify the porosity in term of pore size distribution, pore morphologies and distribution related to mineralogy. Data will be compared with Hg-porosimetry measurement in order to propose a pore model based pore microstructures. Comparison of proposed pore model with previous data structures from other clay-rich host-rocks.


  • Scanning Electron Microscope (SEM)
  • Broad Ion Beam (BIB) sectioning
  • Image processing: Autopano Giga, Adobe Photoshop, Matlab, automatic pore detection with home-developed software, ArcGIS