Engineering Geology 179 (2014) 117–131

Contents lists available at ScienceDirect

Engineering Geology

j ourna l homepage: www.e lsev ie r .com/ locate /enggeo

Nanometer-scale pore fluid distribution and drying damage in preservedclay cores from Belgian clay formations inferred by BIB-cryo-SEM

Guillaume Desbois a,⁎, Janos L. Urai a, Susanne Hemes a, Stephane Brassinnes b, Mieke De Craen c, Xavier Sillen b

a Structural Geology, Tectonics and Geomechanics, EMR, RWTH Aachen University, Lochnerstr. 4-20, D-52056 Aachen, Germanyb ONDRAF-NIRAS, Belgian Agency for Radioactive Waste and Enriched Fissile Materials, Geological Disposal, R&D, Avenue des Arts, 14, 1210 Bruxelles, Belgiumc SCK-CEN, Belgian Nuclear Research Centre, Waste and Disposal Expert Group, Boeretang 200, BE-2400 Mol, Belgium

⁎ Corresponding author. Tel.: +49 241 80 94352.E-mail address: g.desbois@ged.rwth-aachen.de (G. De

http://dx.doi.org/10.1016/j.enggeo.2014.07.0040013-7952/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 19 November 2013Received in revised form 15 May 2014Accepted 6 July 2014Available online 15 July 2014

Keywords:Boom clayImaging of pore waterDryingPore damageBIB–cryo-SEM

This contribution reports on the distribution of in situ pore-water in preserved clay cores from Belgium (BoomClay at HADES URF level and clay from the Pittem Member of the Gent Formation) and identification of typicalpore-scale damage induced by drying.BIB–cryo-SEM (Broad Ion beam–cryogenic-SEM) shows that themajority of pores visible at the resolution of SEM(N50 nm) does not contain water, suggesting that the water content measured by weight loss is located in clayaggregates, in pores b 50 nm.Because water-free pores are not expecting in situ clay formation, we conclude that the procedure used to pre-serve the original clay cores is not fully efficient. Observations of evolution of microstructures before and afterdrying by sublimation of water point to four types of drying-damages driven by the shrinkage of clay particles.On the one hand, quantification of pore size and pore morphologies before and after drying indicates no signifi-cant change in statistical pore characterization, suggesting that quantitative pore morphology studies on a largenumber of pores performed on dried samples are representative of the preserved core samples. On the otherhand, the drying damage observed in this contribution is proposed to be responsible of the changes of fluidflow properties typically measured by other authors on preserved and dried specimens of clays.

sbois).

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Due to their low permeability and possible self-sealing properties innatural conditions, clays are of major importance for society, becausethey are considered to be potential sites for nuclear waste disposal indeep geological formations. A good understanding of their chemical,physical andmechanical properties is also essential for reservoir develop-ment in the unconventional gas industry. Bulkmechanical and fluid flowproperties of clay-rich rocks are widely documented and investigated.However, much less is known about the impact of microstructures onthese properties. Indeed, the coupling between the microstructuralevolution, the poromechanical behavior and the state of hydration ofthe system is still elusive, but required to establish reliable constitutivelaws for the estimation of chemical, physical and mechanical properties(Fanchi, 2010; Savoye et al., 2011) for long time periods (ranging from1 ka to 1 Ma in the case of nuclear-waste geological disposal).

In clays, porosity at grain boundaries is one of the fundamentalparameters, defining the hydro-mechanical properties of the material(Medina et al., 2011). Unfortunately, for clays, most traditional methodsfail to directly image both the porosity in a representative area (orvolume) down to the resolution of pore scale (nm) and the nature ofthe pore filling fluid. Moreover, the microstructures in clays are verysensitive to dehydration, questioning the validity of investigationsdone on dried samples.

Depending on the scale of interest and aim of investigations, thecharacterization of porosity can be performed by different techniques,such as: X-ray computed tomography (micro-CT) scanning (Coleset al., 1998;Wildenschild et al., 2002; Hornapour et al., 2003), magneticresonance imaging (MRI) (Pape et al., 2005), particle size analysis, pointcounting based on petrographic thin sections, environmental scanningmicroscopy (ESEM) (Tiab and Donaldson, 1996; Wang et al.,2013), X-ray diffraction (XRD) (Ward et al., 2005), X-ray fluores-cence (XRF) (McCann, 1998), confocal scanning laser microscopy(CSLM) (Menendez et al., 2001), mercury porosimetry (Favvaset al., 2009) and small angle X-ray (SAXS) and neutron (SANS) scat-tering (Callo and Hall, 2004; Radlinski et al., 2004).

Traditional epoxy-resin impregnation is usually not suitable forpreservation of in situ microstructures in clays. However, some novel

118 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

resin impregnation methods based on 14C-MethylMeth-Acrylate(14C-MMA) resin (Hellmuth et al., 1993; Sammartino et al., 2002;Siitari-Kauppi, 2002; Prêt et al., 2004; Sardini et al., 2009) havebeen developed to preserve in-situ microstructures and imagevery small pore connectivity (down to nm pore diameter; Sardiniet al., 2009). Autoradiography of such impregnated samples allowsquantification of porosity as a function of the mineralogy (Sardini et al.,2009; Prêt et al., 2010), but the quality of images is not sufficient to accessthe morphology of individual pores in 2D cross sections.

Because X-ray μ-tomography does not require dried samples, thistechnique is also good for the 3D investigation of hydrated samples(Onaisi et al., 1993; Otani and Obara, 2004; Van Geet et al., 2005).X-ray tomography can furthermore be applied to deforming samplesto investigate microstructural changes during loading and fluid cir-culation (Renard et al., 2004; Borthwick et al., 2012; Renard, 2012),thus providing 4D microstructural data. Some on-going develop-ments of X-ray-tomography allow tracking the 3D distribution offluids within the porous network (Prodanović et al., 2007; Berget al., 2013) and pave the way for possible studies of the dynamicredistribution of fluids as a function of mechanical loading. However,the resolution provided by X-ray micro-CT imaging remains notgood enough for studying nanoporosity in clays and other tightrock formations and carbonates (resolution down to about amicrometer scale using a laboratory X-ray source [van Geet et al., 2005]and down to about a few hundreds of nanometers using a synchrotronX-ray source [Robinet et al., 2012]).

At a smaller scale, about 15 years ago, a few laboratories started usinga cryogenic SEM on broken samples (cryo-SEM, initially developed for abiological specimen; Echlin, 1978; Sargent, 1988; Erlandsen et al., 1997;Fauchadour et al., 1999; Walther and Muller, 1999; Frederik andSommerdijk, 2005; Marko et al., 2007), to access the fluid distribution(Fassi-Fihri et al., 1992; Mann et al., 1994; Chenu and Tessier, 1995;Robin et al., 1995; Tritlla and Cardellach, 1997; Durand and Rosenberg,1998; Shepherd et al., 1998; Vizika et al., 1998; Barnes et al., 2003;Negre et al., 2004; Kovalevych et al., 2005, 2006; Schenk et al., 2006)related to microstructures, preserved as in natural hydrated conditions.The authors above validated their method, but the observationsmade on broken surfaces were difficult to interpret and therefore ex-periments were stopped. Recently, the development of ion-beamtechniques (Ga+-Focussed Ion Beam [FIB] and Ar-Broad Ion Beam[BIB]), directly implemented within SEM allows investigating repre-sentative, preserved high-quality cross-sections toward microstruc-tural quantification (Holzer et al., 2007, 2010; Desbois et al., 2008,2009, 2011, 2012a,b, 2013; Keller et al., 2011, 2013).

At in situ conditions, Boom Clay (Rupelian, Lower Oligocene) andYpresian clay-rich sands are nearly saturated with water, in therange of 95–100% (SCK-CEN, 1997) and in the range of 94–98%(Van Marcke and Laenen, 2005), respectively. Sample cores storedfor research (e.g. vacuum-packed in Al-coated-poly-ethylene barrierfoil) are assumed to be preserved as in situ, in respect to their hydricaspects. An early contribution by Desbois et al. (2009) showed forthe first time the possibility to image water-filled porosity inpreserved Boom Clay, by using the FIB–cryo-SEM technique. Recentdevelopment of BIB facilities implemented within a cryo-SEM(Desbois et al., 2013) demonstrated that BIB–cryo-SEM is alsosuitable to investigate preserved fine-grained geomaterials, but onmuch larger (representative) high quality cross-sections. However,surprisingly in Desbois et al. (2013), a preliminary result on a differ-ent Boom Clay sample has shown that the majority of the pores werewater-free, contrasting definitely with the early result in Desboiset al. (2009).

This contribution is a follow-up study of Desbois et al. (2013), focus-ing on the study of porosity microstructures in different preserved claycores from Belgium (two cores from Boom Clay at the HADES URF leveland one core from the PittemMember of the Gent Formation), using theBIB–cryo-SEM. The major aim is to describe the distribution of in situ

pore-water in these preserved clay samples. In situ SEM drying ofsamples by sublimation and comparing the same microstructuresbefore and after drying will also contribute to identifying typical poredamage, induced by drying at the scale of the SEM resolution.

2. Samples and methods

2.1. Subsamples from clay drill-core samples

In this contribution, four clay samples fromBelgiumwere investigated(Table 1). Three of these samples are small samples (5× 5×1mm3), sub-sampled from the inner part of three different drill-cores (10 cm in diam-eter and about 30 cm long) using a razor blade in less than 2 min afteropening the drill-core sample bags and before the samples wereplunge-frozen in slushy nitrogen (see Section 2.3). The drill-cores werereceived vacuum-packed in Al-coated-poly-ethylene barrier foil (TypeLOB/39, vapor permeation b0.04 g·m−2·day−1).

Two of the drill cores were collected from two surface drillingcampaigns. The first one was taken at the Kallo-2 borehole (Samplereference: BK/2/1/a1/254.8-255, Table 1). Stratigraphically, this sampleis located in the Pittem Member of the Gent Formation (Ieper Group,Eocene age) and it is known as a sand containing clay with clay-richlenses along the bedding. The subsample studied in this contributionwas collected from one of these clay lenses. The second one comesfrom the Mol-1 borehole and sampled at about the same level as theHADES URF (Underground Research Facility, Mol; sample reference:ON/MOL/1/196, Table 1). Stratigraphically, this sample is located inthe Putte Member of the Boom Formation (Oligocene age) defined asa homogeneous clayey unit. For convenience (Table 1), these samplesare named PC and BC-1, respectively. The depth below the sea level ofthe PC sample is between 254.8 m and 255 m and about 196 m for theBC-1 sample.

The third drill-core is derived from a horizontal borehole direct-ly drilled from the HADES URF in the connecting Gallery Ring 70–71(horizontal borehole toward the west). The sample is taken fromcore 13, at a distance of 12.60–12.65 m from the gallery (intrados),considered to be outside the excavation damage zone (EDZ) of the maingallery. The full sample reference is CG/R70-71W/12.60–12.65, but it isrenamed BC-2 in this contribution. Stratigraphically, the BC-2 sample islocated in the Putte Member of the Boom Formation (Oligocene age) de-fined as a homogeneous clayey unit; and then comparable to the BC-1sample.

Detailed mineralogical composition based on XRD and bulk rockanalysis is only available for the sample BC-1 (Table 1; Zeelmaekers,2011; it is assumed to be also representative for the sample BC-2). Fora comparable sample, Hemes et al. (2013) measured porosity basedon water content (drying at 100 °C) to be in the range of 36–39%, ingood agreement with average water content porosities (37%) knownfor BoomClay (Boisson, 2005). For the PC sample only bulkXRDanalysisis available. As shown in Table 1, the PC sample contains more non-clayminerals than the BC samples. In Boisson (2005), porosity for Clay of theIeper Group is about 40% in average.

The three samples above are considered as preserved in situ samplesand are used to investigate the in situ pore water distribution in drill-core samples.

The time-spans between the date of core drilling and the experimentare between 2 months, 2 years and 15 years for the BC-2, PC and BC-1samples, respectively.

The fourth sample is based on material from the BC-1 drill-core. Apart of the BC-1 samplewas reserved to undergo three successive cyclesof drying in furnace at 100 °C and rewetting in distilledwater within anultrasonic bath. After an additional last drying, this produces a powderof disaggregated clay, which is then rewetted in distilled water toprovide a “remolded” sample (Table 1). Thus, we are sure that theremolded sample is a fully saturated sample to be compared with thenatural “preserved” samples (BC-1, BC-2 and PC).

Table 1Samples description and related water contents.

# Samplenameusedin thisstudy

Samplereference

Formation, memberand geological age

Depthbelow sealevel (m)

State of the analyzedsample

Mineralogical data (%) Watercontent(%)

Waterweightporosity(%)

Non-clay Clay

1 remolded ON/MOL-1/196 – – Water-re-hydratedmud of disaggregatedBoom Clay

– – 39.2 63

2 BC-1 ON/MOL-1/196 Putte member of the Boomformation (HADES level,Oligocene age)

196 Naturally hydratedsample packed inAl-barrier foil

49a

32% quartz, 7% feldspar,2% plagioclase, 4% calcite,2% pyrite, 1.4% organicmatter, rest b 1%

51a

40% illite–smectitegroup, 9% kaolinite,3% chlorite

16.7 34

3 BC-2 CG/R70W-71/13.1-core13.1/12.60–12.65

Putte member of the Boomformation (HADES Level,Oligocene age)

≈196 Naturally hydratedsample packed inAl-barrier foil

– – 13.3 29

4 PC BK/2/1/a1/254.8–255

Pittem member of the Gentformation (Ieper Group,Eocene age)

254.8–255 Naturally hydratedsample packed inAl-barrier foil

76.8b

57.3% quartz, 17.33%feldspar, 1.14% calcite,1.03% pyrite

23.2b

11.84% illite–smectitegroup, 5.49% kaolinite,5.87% chlorite

18.3 38

a XRD and bulk rock analysis from Zeelmaekers (2011).b Bulk XRD analysis (this study).

119G. Desbois et al. / Engineering Geology 179 (2014) 117–131

2.2. Measurement of sample's water content

Subsamples of about 2 cm3 in size, comparable to samples describedin Section 2.1, were dried at room temperature (RH ~ 50–60%) to mea-sure water content and estimate the water weight porosity. Waterweight losses were monitored in time by weighing the samples withan electronic balance (Kern, ABJ-220-4M, resolution of 0.1 mg). Foreach sample, the loss in weight was measured until the water weightloss reached its maximum. The estimation of water weight porosityfrom water content is calculated based on the grain density of thebulk samples (2.65 g/cm3 for BC-1 and BC-2 samples after De Craenet al., 2004) and the density of the pore water (1.019 g/cm3 for BC-1and BC-2 samples after De Craen et al., 2004). For PC samples, robustdata are still scarce in literature. Lima et al. (2012)mention a grain den-sity of 2.76 g/cm3 but water density is not available yet. Thus, density ofpore water for Boom Clay is used as a first order of approximationthough it is expected to be slightly higher because of higher salinity inPC sample (Van Marcke and Laenen, 2005).

2.3. BIB–cryo-SEM

The original concept of the BIB–cryo-SEM allows preparing a repre-sentative high quality cross-section (mm2 in size), inside a scanning elec-tron microscope (SEM), using a noble gas (argon) ion-beam at cryogenictemperature (cryo) to investigate water-saturated microstructures in-situ and at SEM resolution. A detailed description of the BIB–cryo-SEMmachine used in this study (at RWTH Aachen University, Germany) isgiven in Desbois et al. (2013). The following description is a brief summa-ry of the key points of the BIB–cryo-SEM procedure. Fig. 1 gives aschematic drawing of the full BIB–cryo-SEM experiment from the sub-sampling of the drill-core to the imaging of microstructures at cryogenictemperature in BIB cross-sections produced inside the SEM.

2.3.1. Cryogenic quenchingThe PC, BC-2 and BC-1 sub-samples (about 5 × 5 × 1mm3 in size, see

Section 2.1) were: (a) fixed on a copper stub (with a 30° angle) withcryo-gel (O.C.T™ compound, Tissue-Tek®) at room temperature;(b) plunge-frozen and stirred in slushy nitrogen to minimize theLeidenfrost effect (Zhang and Gogos, 1991; Desbois et al., 2008, 2013);(c) transferred from the slushy nitrogen bath to the cryo-preparationchamber, by using the specimen transfer device, operating at cryogenic

temperatures and at vacuumconditions; (d) coatedwith platinum (15A,50 s) at cryogenic temperature; and (e) transferred from the cryo-preparation chamber to the SEM sample cryo-stage, ready to be BIBcut at cryogenic temperature in-situ SEM.

The preparation of the remolded sample for the BIB–cryo-SEM ex-periment follows the same procedure above except that the piece of‘mud’ is self-stabilized without using cryo-gel to fix the sample on thecopper stub.

2.3.2. In situ SEM BIB cross sectioning at cryogenic temperatureIn this contribution, in-situ BIB cross-sectioning is used as de-

scribed in Desbois et al. (2013). After introduction of the Pt-coatedfrozen sample, mounted on the copper stub (see Section 2.3.1)onto the sample stage inside the SEM chamber (−160 °C), the re-gion of interest to be BIB cross-sectioned is placed at the SEMstage rotation axis (Figure 1). Using a three axis Cartesian manipu-lator, the edge of a Ti-mask is first placed at the same working dis-tance as the region of interest (Figure 1) and then approached toless than 40 nm distance from the specimen surface (an ion millingresistant mask, made of titanium, is used to split the beam in two, toproduce a sharp-edged beam for flat cross-sectioning). The SEMstage is tilted to −18° in order to align the BIB with the top surfaceof the Ti-mask. The coincidence point of the electron beam and theBIB at the region of interest is reached for a working distance ofabout 17 mm. To minimize curtaining effects on the cross sectionsurface, the SEM stage oscillates between ±30° around theSEM stage rotation axis, during BIB milling. The Cartesian nano-manipulator is gravity compensated and has a closed loop positionholding system, to avoid drift of the Ti-mask positioning during ro-tation of the SEM stage.

Each BIB cut is performed for 2 h 30min at 10 kV to produce a singlecross-section. During the entire procedure described above, the SEMsample stage is at a temperature of −160 °C.

2.3.3. SEM-imaging of microstructuresAfter BIB milling, microstructures in BIB cross-sections are

SEM-imaged (SE2 detector) at 1.5 kV and a working distance of4 mm, using an aperture of 30 μm. BIB cross-sections are notmetal-coated.

Heating of the cryo-sample stage from −160 °C to +20 °C forabout 2 h at 6 · 10−5 mbar was done after cryo-BIB–SEM in order to

Fig. 1. Schematic procedure of BIB–cryo-SEM investigation of preserved specimenof original BoomClay cores. At room temperature, original BoomClay cores are unpacked fromAl-barrierfoils either to subsample directly to specimens of about 5 × 5 × 1 mm3 in size (i.e. BC-1, PC and BC-2 samples) or to subsample clay material to provide synthetic hydrated mud of disag-gregated Boom Clay (i.e. remolded sample). Both types of specimen are quenched in liquid nitrogen (LN2,−193 K) and then transferred to the BIB–cryo-SEM as detailed in Desbois et al.(2013) to be BIB cross-sectioned at cryogenic temperature (−160 °C) within SEM. After 2.5 h of cross-sectioning, preserved pore microstructures are imaged at cryogenic temperature(−160 °C) within SEM. After sublimation of water within SEM, same microstructures are imaged at 20 °C within SEM. For more details, please check the manuscript.

120 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

sublimatewater (Schenk et al., 2006; Desbois et al., 2008, 2012a,b, 2013).Selected regions of the BIB cross-sections were imaged before and aftersublimation to firstly image pore water and secondly evaluate eventualchanges in pore microstructures.

2.4. Segmentation of porosity in SEM images

Pores present in the regions of the BIB cross-sections (PC, BC-2 andBC-1 samples), imaged before and after sublimation, are segmented in

121G. Desbois et al. / Engineering Geology 179 (2014) 117–131

order to quantify eventual changes in pore morphology with respect todrying.

The pore segmentationwas donewith the help of an image segmen-tation algorithm in MATLAB (Mathworks, 2011) specially designedfor BIB-polished SE2 images (Kempin, 2012). This method allowssegmenting pores in a systematic, fast and reproducible way. Theautomatic segmentation results are afterward checked manually andcorrected, if necessary (up to 15% of the automatically segmentedpores had to be manually corrected in this contribution).

The segmentation of porosity results in binary images with poresshown in black. Each pore is subsequently analyzed in MATLAB(MathWorks, 2011) using PolyLX (Lexa, 2010) to calculate morpholog-ical parameters, like pore perimeter (P), pore area (A), length of thepore's longest axis (L) and length of the shortest axis (W). These param-eters are needed to calculate circularity (4πA/P2) and the inverse of theaxial ratio (1/AR = W/L) (Underwood, 1970; Russ and Dehoff, 2000).Pore area, inverse of axial ratio and circularity are the parameters usedto quantify the possible change in pore morphology reflecting damagefrom drying in each subsample, by comparing the probability of findingsuch a bin-classified parameter in wet vs. dried samples (the bin-sizedoubles for each consequent bin [Adamic and Huberman, 2002] forthe analysis of pore area, whereas a fixed bin-size of 0.1 is used for theanalysis of the inverse of the axial ratio and the circularity), i.e. beforeand after sublimation.

3. Results

3.1. Water contents

All samples investigated in this contribution contain a significantamount of water.

The samples BC-1, BC-2 and PC were measured to contain 16.7%,13.3% and 18.3% of water in weight, corresponding, respectively, tovalues of 34%, 29% and 38% of water-weight porosity (Table 1). Asmentioned in Section 2.2, water weight porosity for the PC sample is arough estimationwith porewater density similar to the porewater den-sity of BC samples.

In comparison, the remolded sample contains about twice theamount of water as the samples BC-1, PC and BC-2. The remolded sam-plewasmeasured to contain 39.2% of water inweight, i.e. 63% of water-weight porosity (Table 1).

Fig. 2. Overview of a typical BIB cross section performed at cryogenic temperature(−160 °C). The polished area results in a 2Dhigh quality flat surfacewith a pseudoGauss-ian shape, which allows investigating microstructures down the resolution of SEM. Highrelief area below the BIB cross section corresponds to sample's region which is only partlyabraded during BIB cross-sectioning.

3.2. Pore-water distribution

Pore water was detected by microstructural comparison betweentwo SEM images of the same region of interest taken before and afterthe sublimation process (Desbois et al., 2009, 2012a, 2013). Each regionof interestwas imaged atmagnifications between 14,000 and 60,000××and located within the polished area of the BIB cross-section (Figure 2).

3.2.1. Remolded sampleBefore the sublimation process, the polished area of the BIB cross-

section shows a relatively flat and homogeneous surface where thewater-free pores, if they exist, are not visible (Figure 3a, c, e). Themost remarkable topography is an oblique curtaining of the surface,which appears in parts of the imaged areas (Figure 3a, c, e). After subli-mation,most areas that exhibited curtaining prior sublimation are pores(Figure 3b, d, f). Because in this experiment only water is sublimated(Desbois et al., 2013), holes appearing in images after sublimationwere filledwith pore-water before sublimation. Sublimation of samplesleft behinddetrital clasts (mainly quartz, calcite andmica in Figure 3b, d,and f, respectively) and clay material, as compacted clay-rich regions orhairy/fibrous clay (Figure 3a–f). We interpret the regions of compactedclay to remnant pieces of the original clay core,whichwere not fully dis-aggregated during the preparation of the remolded sample. In contrastthe hairy/fibrous clay represents the fully disaggregated clay particlesmixed in distilled water (Figure 1).

3.2.2. Preserved core samplesAt first glance (Figure 4), overall microstructures in BIB cross-

sections for BC-1 and BC-2 samples are comparable to typical fabricsof Boom Clay prepared by the BIB method at room temperature(Desbois et al., 2009; Hemes et al., 2011, 2013). In all samples studiedin this contribution, the non-clay minerals (quartz, feldspar, calcite;Figure 4) are embedded in a porous clay matrix. Surprising is that thepores within the clay matrix are clearly visible before the sublimationprocess (Figure 4a, c, e, g, and i). Based on the results obtained for theremolded sample (Section 3.2.1), this indicates that the pores are freeof water before the sublimation. This applies to almost all pores visiblewith SEM in all samples (BC-1, PC, BC-2). Only very few pores in sampleBC-1 (estimated to be less than 1% of the imaged pores) were filledwith a frozen water before sublimation (Figure 4c, d). The largestpores (up to 1 μm wide) were found at interfaces between clasts andthe clay matrix (Figure 4).

3.3. Typical damages induced by drying heating samples within the SEM

A detailed observation of microstructure evolution (Figures 5, 6 and7) indicates microstructural changes at pore scale during drying ofsamples by sublimation up to 20 °C. Figures 5, 6 and 7 show typicalpore-scale damages in sample PC, but they are also representative forsamples BC-1 and BC-2. Four main types of damages were identified.

The three first types involve fracturing of the clay matrix.Damages of Type I are breaks of clay/clay particle interfaces (Figure

5a–f), where the interfaces involve clay particles highly misoriented toeach other, typically a “tip/long axis” contact between two clay particles.

Damages of Type II are detachments (delamination) at clay/clast in-terfaces, where such interfaces involve clay particles orientedwith theirlong axes parallel to the clast surface (Figure 5g–n). We distinguish twokinds of detachment defined by the amplitude of the detachment:a “weak” detachment has maximum amplitudes of a few tens of nano-meters and is localized only at key-points of a clast/clay interface(Figure 5g–j), whereas a “strong” detachment has amplitudes up to1 μm and it affects all the contour of a clast/clay interface (Figure 5k–n).

Damages of Type III are the biggest and most spectacular (Figure 6).Observations show that a damage of Type III results in a crack with anaperture wider than 1 μm (Figure 6a–b), which propagates only withinthe clay matrix but through a large volume of samples (Figure 6b–e).

Fig. 3.Typical BIB–cryo-SEMobservation on remolded sample on same sample's regionbefore (left column) and after (right column)dryingby sublimation.Dashed lines inyellow indicatenon-clay grain boundaries and circles in red point to a region of compacted clay interpreted as a relic of non-disaggregated Boom Clay. Before drying, the space between detrital grains isfilledwith frozenwater (a, c, e). Drying up to 20 °C at a pressure of 6 · 10−5mbar removeswater by sublimation (b, d, f) confirming the presence of early frozenwater by comparisonwithfigures in the left column. Early presence of water is also often suggested by significant curtaining of frozenwater (a, c, e). Hairy illite is typically mixedwith early water. Qtz.: quartz, Cc.:calcite.

122 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

Moreover, most of the crack's length is following clast/clast interfaces(Figure 6e). Damage of Type III involves significant displacement ofsample material. For example, comparison of a few measurementsbefore and after drying (Figure 6f, g) estimates displacements in therange of 20–30%.

The fourth kind of damage (Type IV) does not involve visiblerupture or failure of the clay matrix, but only a local modificationof the pore morphologies, resulting in an increase of pore area,

Fig. 4. Typical BIB–cryo-SEM observation on BC-1, PC and BC-2 samples on same sample's regioparison of images before and after drying shows that themajority of pore visible at resolution ofsquare in red (e, f) indicates to a region imaged at highermagnification in (g, h; respectively) andnon-clay grain. Qtz.: quartz, Cc.: calcite, Felds.: feldspar.

which may also result in a connection of neighboring pores(Figure 7b, c).

3.4. Quantification and comparison of pore morphologies in preserved anddried samples

Quantification of pore morphologies in samples BC-1, BC-2 and PC isbased on more than 2500, 3000 and 1800 singles pores on within the

n before (left column) and after (right column) drying by sublimation. Surprisingly, com-SEMarefluid-free before drying. Dashed lines in yellow indicate non-clay grain boundaries,ellipse in red point to a rarefluid-filled pore (dvs. e). LP indicates a large pore surrounding

123G. Desbois et al. / Engineering Geology 179 (2014) 117–131

Fig. 5. Detailed observations of pores on same sample's region before (left column) and after (right column) drying by sublimation. Images are from sample PC but observations are alsovalid for BC-1 and BC-2 samples. Drying is damaging poremicrostructures; in thisfigure two types of damages are identified (Type I and II; see figure andmanuscript for details); all withinclaymatrix. Damages are interpreted to originate from shrinkage of the claymatrix, which should containwatermeasured bywaterweight before drying because pores at SEM resolutionappear fluid-free. Qtz.: quartz, Cc.: calcite.

124 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

same areas before and after sublimation, respectively. Although sin-gle pores were analyzed within similar areas before and after drying,the total number of pores detected may differ within a range of 2–4%before and after drying, because of noisier images taken at cryogenicconditions (Figures 3, 4, 5, 6 and 7; Desbois et al., 2013) and the

slight modification of pore microstructures due to drying (seeSection 3.3).

Results are compiled in Fig. 8 and show that the probabilities to finda pore within a certain pore-area range (Figure 8a, d, g), a certain in-verse of axial ratio range (Figure 8b, e, h) and a certain circularity

Fig. 6. Detailed observations of a pore damage of Type III before (top row) and after (rows below) drying by sublimation. Images are from sample PC but observations are also valid for BC-1and BC-2 samples. Pore damage of Type IV is a long crackwith aperture above 1 μmpropagating within the claymatrix and interfaces of claymatrix and non-clay grains. Damage of Type III isalso interpreted to originate from shrinkage of the claymatrix. (f andg)depict same regions in (a and b)where reference lines are indicated for evaluation of displacement in (h). Qtz.: quartz.

125G. Desbois et al. / Engineering Geology 179 (2014) 117–131

range (Figure 8c, f, i) are not significantly different in preserved anddried samples, within the standard error.

4. Discussion

4.1. Preservation of wet microstructures in clays and drying damage

The remolded sample, containing 39.2% water in weight, shows thatall porosity is filled with frozen water before the drying-sublimation(Figure 3a, c, e). After sublimation, the porosity appears water-free(Figure 3b, d, f). If the water sublimation process is not activated,

there is no evidence that the sample preparation, BIBmilling or SEM im-aging destabilizes the pore water quenched by cryogeny, in a very goodagreement with Desbois et al. (2013). Therefore, the BIB–cryo-SEMmethod is able to preserve wet microstructures in hydrated samples(typically about 5 × 5 × 1 mm3 in size; Desbois et al., 2013).

The same approach applied to preserved wet natural samples indi-cates that water sublimation induces local damage of the clays' micro-structure, all located within the clay matrix (Figures 5, 6 and 7). This isevidence for the preservation of the wet microstructures by cryogeny,prior to water sublimation. Therefore, drying is interpreted to beresponsible for the observed microstructural changes after water

Fig. 7. Example of segmentation of porosity on the same sample's region before and after drying by sublimation used as basis of pore size and pore morphology quantification in Fig. 8.Damage of Type IV, which corresponds to a pore area increase without bound failure (in contrast to damages of Types I, II and II), is also indicated by a circle in red.

126 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

sublimation. Because elasticity is by far insufficient to result inmajor vol-ume changes, the observed damages are interpreted to be induced by theshrinkage/swelling of the claymatrix during drying (Delage et al., 2007).This is in good agreement with our microstructural observations, whichnever showdamage of non-claymineral grains.With respect to the aper-ture sizes of the damage features: smaller apertures are observed at in-terfaces presenting mechanical incompatibilities due to localmicrostructural anisotropy, involving either highlymisoriented clay par-ticles (Type I, Figure 5a–f) or clay/clast contacts (Type II, Figure 5g–n);larger damages (Type III, Figure 6), running all around clast grains, areinterpreted as the coalescence of damages of Type II. Damages of TypeIV (Figure 7) do not involve failure, but only clay-matrix shrinkage,resulting in an increase of the pore area. Gradients in shrinkage strain be-tween the outside and the inside of a bulk sample and the build-up ofstress, due to grain-scale, non-uniform deformation at interfaces be-tween the clay-matrix and non-clay mineral grains (Klinkenberg et al.,2009; Houben et al., 2013) may localize damage due to drying.

4.2. Hydration state of preserved clay cores

The water contents measured for preserved Boom Clay samples(16.7%, 13.3% and 18.3%, for samples BC-1, BC-2 respectively; Table 1)are within the lower limit of the typical range of water contents givenin reference literature for Boom Clay (19–24%; Boisson, 2005). Thesmall deviations may be explained by the different drying conditionsused; water contents given in Boisson (2005) are from samples driedat 105–110 °C, but samples studied in this contribution were dried atroom temperature. Differences may also be explained by clay-boundwater, partly removed during drying at 105 °C (Pearson et al., 2003;Marschall et al., 2005); although interlayer adsorbed water in clays ismainly removed at a temperature of 250 °C (Plötze et al., 2007).

For the Pittem Clay (PC), themeasured water content is about 18.3%(Table 1) but there is no existing corresponding data in the literature tocompare this value with. However, they are in good agreement withother clay-rich sands of Ypresian Clays,which have a higherweight con-tent than Boom Clay (LTGH, 1996 and INISMA, 1998).

Water contents measured and damages observed after water subli-mation (Figures 5, 6 and 7) both argue for hydrated clay samples. How-ever, the pores imaged at SEM resolution before water sublimation aremainly free of water (Figures 4, 5 and 6). Only rare single pores arewater-filled (Figure 4c, d). At first, this indicates that water measuredin our samples by weighing (Table 1) is residing within the clay aggre-gates, which contain pores not visible at the resolution of the SEM,

possibly trapped by a capillary effect. However, at least for the BoomClay samples (BC-1 and BC-2), water-free pores at the SEM resolutionare in contradiction with a single FIB–cryo-SEM experiment afterDesbois et al. (2009) on a Boom Clay sample from the HADES level,where all observed pores were water-filled.

In the following,we discuss several hypotheses to explainwhyporesimaged in this contribution before drying for Boom Clay samples (BC-1and BC-2) are water-free.

Based on the comparison between this study and Desbois et al.(2009), the first hypothesis is to simply consider that the hydratedstate of an original core varies from core to core, suggesting somewater-unsaturated Boom Clay layers. This hypothesis can be rejected;because pore pressure (Mertens et al., 2004; Bastiaens et al., 2006;Bernier et al., 2007) and measured water contents are consistent withmeasured porosities (De Craen et al., 2004; Boisson, 2005; Delageet al., 2007; Francois et al., 2009). The natural degree of saturation inBoom Clay samples is in the range of 95–100% (SCK-CEN, 1997).

A second hypothesis is the degassing of gas dissolved in pore waterat natural conditions, after sample coring. Depressurization of cores ex-perienced for core drilling may cause vaporization of dissolved gas,resulting in a possible squeezing of water out of the pore space. Boomclay water contains dissolved CH4, CO2, N2, H2O O2; where the CH4

and CO2 concentrations are the highest (De Craen et al., 2004; Keleet al., 2005). Considering Henry's law, the concentration of dissolvedCH4 and CO2 at a temperature T (CCH4

diss:;T and CCO2diss:;T ,, respectively) is a

function of the partial pressure of CH4 or CO2 (pCH4 and pCO2 ,respectively)and the ‘Henry's constant’ for CH4 or CO2 (k

CH4H;T and kCO2

H;T , respectively):

CCH4diss:;T ¼ pCH4

kCH4H;T

CCO2diss:;T ¼ pCO2

kCO2H;T

ð1Þ

where kCH4H;T and kCO2

H;T are given by the Van't Hoff's equation:

kCH4H;T ¼ kCH4

H;298Kexp −CCH4

1T− 1

298

� �� �

kCO2H;T ¼ kCO2

H;298Kexp −CCO2

1T− 1

298

� �� �:

ð2Þ

With log pCH4� � ¼ −8:81 and log pCO2

� � ¼ −2:42 (Dierckx, 1997;De Craen et al., 2004), kCH4

H;298K ¼ 714:29 L � atm �mol−1 and kCO2H;298K ¼

29:41 L � atm �mol−1 (Lide and Frederikse, 1995), at a temperature

Fig. 8.Quantification of pore size (pore area in left column) and poremorphologies (elongation [1/AR] inmiddle columnand circularity on the right column) for samples BC-1, PC andBC-2.All graphs show the compilation of probabilities of the parameter's occurrence after drying (Pdry) in the y-axis as a function of the probabilities of the parameter's occurrence before drying(Pwet) in the x-axis, on same sample regions and binned parameters. Pore areas start from 512 nm2 to 2,097,152 nm2with bin size doubling each bin. Bin size for elongation and circularityis 0.1. Statistics are based onmore than 2500, 3000 and 1800 single pores for samples BC-1, BC-2 and PC, respectively. Dashed black line indicates the 1:1 relationship (i.e. Pdry= Pdry).Weobserved that drying does not modify probabilities of parameter's occurrence suggesting that pore size and poremorphological statistics performed on dried sample are representative ofhydrated Boom Clay cores. Error bars are standard errors.

127G. Desbois et al. / Engineering Geology 179 (2014) 117–131

of T = 289 K (temperature of pore water in in situ Boom ClayBoisson, 2005) and CCH4 ¼ 1600 K and CCO2 ¼ 2400 K (a constant,reflecting the ratio between the enthalpy of solution [for CH4 orCO2] and the gas constant, respectively; Lide and Frederikse,1995), the concentrations of dissolved CH4 and CO2 in in-situBoom Clay pore water are CCH4

diss:;289K ¼ 2:5� 10−12 mol � L−1 andCCO2diss:;289K ¼ 1:7� 10−4 mol � L−1.

The corresponding volume of degassed CH4 and CO2 per unit samplevolume (V

CH4

gaz and VCO2

gaz ) is given by:

VCH4gaz ¼ CCH4

diss:;283K � Vpore

Vsample� Vm ¼ CCH4

diss:;289K � θ� Vm

VCO2gaz ¼ CCO2

diss:;283K � Vpore

Vsample� Vm ¼ CCO2

diss:;289K � θ� Vm

ð3Þ

128 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

where Vm= 24.47 L ⋅mol−1 is themolar volume at standard conditions,Vpore the total pore volume, Vsample the volume of the sample and θ theporosity of the sample (in the range of 35–40%; Boisson, 2005). For aporosity of θ=35 %, we obtain VCH4

gaz ¼ 2:2� 10−11 L �m−3 and VCO2gaz ¼

1:4� 10−3 L �m−3to be compared with the volume of pores per unitsample volume Vpore ¼ θ� 1 ¼ 0:35 L �m−3 . Hence, the volume ofCH4 and CO2 gas potentially available for a full degassing representsabout 0.4% unit of porosity and thus cannot account for a squeezing outof all pore water contained in the visible pore space at SEM-resolution.

In contrast to the second hypothesis, which assumes leakage ofpore-water, a third hypothesis might be to consider poroelastic expan-sion of Boom Clay induced by release of in situ stress after coring. Thisthird hypothesis might involve that some water-free pores visible atSEM resolution result from sample dilation. Giraud and Rousset(1996; after Coussy, 1989) give the variation of the volume as a functionof the variation of the pressure:

ΔVV

¼ Δp� bK0

ð4Þ

where ΔVV is the variation of volume, Δp = peffective,BoomClay − patm is

the variation of pressure between in situ and atmospheric conditionsof Boom Clay (peffective,BoomClay ≅ 2.3 Mpa; Horseman et al., 1987;Delage et al., 2000; Mertens et al., 2004; Bastiaens et al., 2006;Bernier et al., 2007), b the Biot's coefficient (for Boom Clay: b = 1;Giraud and Rousset, 1996) and K0 is the bulk modulus of the porousmedium (for Boom Clay: K0 = 74 MPa; Giraud and Rousset, 1996).Therefore, ΔV

V ≈3%, suggesting that part of the porosity visible atSEM resolution may correspond to volume expansion after coringand therefore be free of water.

A fourth hypothesis is that the 2min time slot between sub-samplingof the core samples and cryogenic quenching of subsamples (seeSection 2.1) is long enough to allow water evaporation from the firstfew micrometer layers of analyzed samples. This hypothesis is unlikely,since our observations never showed a layering of the pore-water contentfrom the top sample surface toward the center of the samples (over thewhole BIB cross sections with a typical depth greater than 200 μm)(Figure 2).

A fifth hypothesis is to consider leakage of porewater during storageof samples in the Al-coated-poly-ethylene barrier foils.

At first, we can simply consider that the sample is packed fullywater-saturated and that only the diffusion of water through the Al-coated-poly-ethylene barrier foil is responsible for sample drying. Com-mercial data available on the use of Al-coated-poly-ethylene barrier foilgives a vapor permeation of b0.04 g/m2/day (at room temperature).Studied cores with a typical length (L) of 30 cm and a radius (R) of5 cm were packed in Al-coated-poly-ethylene barrier foil of about2πR2 + 2πRL = 1.1 × 10−1 m2 area. At the highest magnification (i.e.about 30,000×) used to image the pore space, Hemes et al. (2011,2013) measured porosities at SEM-resolution in representative BoomClay samples of 10–15%. Therefore, the minimum amount of waterthat would have to be removed from a fully water-saturated typical

Table 2Estimation of water weight loss by vacuuming during sample packing in Al-barrier foil.

# Samplename usedin thisstudy

Initial packing condition Waterweightloss (%)

Water weight loss porosity(%; calculation based on DeCraen et al., 2004)

1 BC-1 Repacked one time afterfirst packing done afterinitial core sampling

0.17 0.452 0.64 1.663 0.05 0.124 0.08 0.221 BC-2 Repacked two times after

first packing done afterinitial core sampling.

0.03 0.092 0.03 0.083 0.05 0.124 0.07 0.19

core sample of Boom Clay to observe water-free pores at SEM-resolution is: πR2L = 235.6 cm3 corresponding to 231.2 g of porewater (pore water density = 1.019 g/cm3; De Craen et al., 2004). Thetime needed to remove such an amount of water only by diffusionthrough the Al-coated-poly-ethylene barrier foil is estimated to beabout 146 years. The age of the cores analyzed in the present study isbetween 2 weeks and 15 years, preserved at 7 °C (thus, effectivevapor permeation is expected lower thanmentioned above). Therefore,water diffusion through the barrier foil cannot be the only cause ofwater-free pore space at SEM resolution, though sealing would be theweakest point pertinent to the tightness of the barrier foil but muchmore difficult to quantify.

Secondly, vacuum pumping (P = −750 mbar) of sample bags forstorage after coring may contribute to sample drying. It was unfortu-nately not possible to compare sample weights just after coring andafter vacuum-packing of samples. However, we measured water con-tents in a few samples of the BC-1 (4 samples of about 5 cm3 in volume)and BC-2 (4 samples of about 5 cm3 in volume) cores before and aftervacuum-repacking in Al-barrier foils (Table 2): after the initial packing,right after sample coring, the BC-1 samples were repacked only once,whereas the BC-2 samples were repacked twice. Vacuum (P =−750 mbar) was achieved in fewminutes. Table 2 presents the resultsof water weight loss for the last repacking, indicating a measureablewater weight loss for both experiments, corresponding to a porosity inthe range of 0.12%–1.66% for BC-1 samples and between 0.09% and0.19% for BC-2 samples. Therefore, vacuum-packing, involved in samplestorage, participates in drying samples. However, water-free porositiesreachable by vacuum-packing are clearly below thewater-free porosityobserved in our BIB–cryo-SEM images (Figures 4–7), but this could beexplained by the fact that we did not have access to sample weightsjust after coring and after vacuum-packing of samples, where maximalwater loss is expected.

Therefore, two hypotheses are most likely explaining why pores atthe SEM resolution are water-free before drying sublimation: the vol-ume expansion of samples after coring and the drying by vacuum dur-ing packing of samples in Al-barrier foil.

One paradox remains: Desbois et al. (2009) found water-filled poros-ity at the SEM resolution. So far, we do not have a good explanation forthis. The only difference with this contribution is the size of the originalcore packed in Al-barrier foil. In Desbois et al. (2009), the sample originat-ed from a core of 30 cm diameter, whereas core diameters in this contri-bution are only 10 cm (see Section 2).We can only suggest that in a largercore volume the total volume of pore water remaining inside the coreafter vacuum re-packing is sufficient to buffer the whole core after equil-ibrating with the remaining atmosphere inside the Al-barrier foil bag.

4.3. Impact of drying damage on microstructural studies

Preservation of in situ samples is of special relevance for robust andrepresentative characterization of clay. Unfortunately, this contributionshows that cores considered as “in situ preserved” are not as revealed bywater-free visible porosity at SEM resolution, indicating significantdrying of material occurring before the first opening of the barrier foilcontaining the “in situ preserved” cores. However, up to now, thecores packed in such barrier foils are mainly considered to be “insitu preserved” by the scientific community. For example, “in situ pre-served” cores are widely used as a reference material for porosimetryand permeability analyses. Thus, this study gives a first evaluation ofdrying damages (Figures 5–7) at pore scale in Boom Clay with respectto subsequent drying of “in situ preserved” cores. Drying damages in-volved in the first drying (i.e. after the first opening of the barrier foil)is not accessible yet; it would require the development of new tech-niques to collect genuine in situ preserved core's samples and rapidquenching of microstructures right after sample's collection.

Some recent studies characterized 2D porosity (pore size distribu-tion and pore morphologies) in fine-grained, clay-rich geomaterials

129G. Desbois et al. / Engineering Geology 179 (2014) 117–131

(Desbois et al., 2009; Loucks et al., 2009, 2010;Heath et al., 2011;Hemeset al., 2011, 2013; Curtis et al., 2012; Houben et al., 2013; Klaver et al.,2012) using the BIB–SEM approach at room temperature. In the lightof the present study, it is suggested that such investigations on driedsamples include microstructural changes due to drying (i.e. ‘dryingdamage’; Figures 4–7).

However, a statistical comparison of pore size distributions and poremorphologies, performed on a few thousand single pores, within thesame area before and after drying by sublimation (Figure 8), shows nosignificant difference. Statistical studies of porosity do not account forthe signatures of drying induced damage within the standard errors.Only minor deviations are observed for the circularity (Figure 8c, f, i),but attributed to decreasing image quality at cryogenic conditions(Figures 4–7; the perimeter being more sensitive to image quality).Therefore, statistical characterization of porosity in representative 2DBIB–SEM images, performed at room temperature, can be consideredas representative of pore microstructures in cores considered as “insitu preserved”.

Moreover, detailed observations point to large pores surroundingnon-clay mineral grains (Figure 4) already present prior to drying(Figure 4a, c, e, g, i). Hemes et al. (2011, 2013) also identified suchlarge pores, surrounding non-clay mineral grains in samples compara-ble to Boom Clay sample in this study (BC-1 and BC-2), but their originremained elusive. This contribution suggests that they do not originatefrom drying of cores considered as “in situ preserved”.

In contrast, a number of studies (mainly based onmercury injectionporosimetry) inferred changes of pore microstructures related to thedrying of material with a measurable effect both on total porosity,pore throat distribution and permeability (Galle, 2001; Kang et al.,2003; Cosenza et al., 2007; Gasc-Barbier and Tessier, 2007; Soe et al.,2009; Pineda et al., 2010; Hedan et al., 2012). Only very few of thesestudies checked directly the changes in pore microstructures beforeand after drying (Kang et al., 2003; Renard, 2012); mainly because ofthe challenging task of imaging preserved microstructures below themicrometer scale. Indeed, studies, which observed changes in micro-structure from hydrated to dried clays only refer to large desiccationcracks with apertures above 1 μm (Kang et al., 2003; Hedan et al.,2012; Renard, 2012; Wang et al., 2013), comparable to damage ofType III defined in this contribution. This particular type of damagemay be strongly involved in overestimation of permeability in samples,which have been dried prior to permeability measurement. Our studysuggests that damages of Types I, II and IV may also contribute to over-estimation of total porosity, pore throat distribution and permeabilitybetween hydrated and dried clays but should bemore related to changein capillary porosity (Galle, 2001).

5. Conclusions

This contribution shows how the BIB–cryo-SEM method allowsimaging water-filled pores in cores considered as “in situ preserved”,hydrated clays. Application of BIB–cryo-SEM on samples from originalclay cores shows that porosity is free of water at the SEM resolution,suggesting that water measured by weight loss is only located withinthe clay aggregates.

Water-free pores are unrealistic in natural saturated Boom Clay.Thus, we conclude that the procedure involved to store the originalclay cores in fully preserved natural conditions is not efficient. For afuture study, we plan to sample and stabilize specimens on site directlyafter coring by using cryogenic techniques andwithout intermediate re-packing of samples, to be able to observe fully hydrated porosities.

Microstructural observations on the same areas before and afterdrying by sublimation of water, point to four types of typical dryingdamages induced by the shrinkage of the clay material. However, thequantification of pore size and pore morphologies before and after dry-ing indicates no significant changes in statistical porosity characteriza-tion suggesting, on the one hand, that quantitative studies performed

on a large number of pores even at dry conditions should be representa-tive of core considered as “in situ preserved”. On the other hand, dryingdamages observed in this contribution are proposed to contribute all tothe change of fluid flow properties typically measured by other authorson hydrated and dried specimen of clays.

Acknowledgments

We thank ONDRAF-NIRAS and SCK-CEN for supporting the projectand the Deutsche Forschungsgemeinschaft for funding the BIB–cryo-SEMinstrument (Project UR 64/9-2). We are also grateful to UweWollenberg(GIA at RWTH Aachen University) for his help with maintenance andtechnical aspects of SEM.

References

Adamic, L.A., Huberman, B.A., 2002. Zipf's law and the internet. Glottometrics 3, 143–150.Barnes, P.R.F., Wolff, E.W., Mallard, D.C., Mader, H.M., 2003. SEM studies of the morphol-

ogy and chemistry of polar ice. Microsc. Res. Tech. 62 (1), 62–69.Bastiaens, W., Bernier, F., Li, X.L., 2006. An overview of long-term HM measurements

around HADES URF. EUROCK 2006 Multiphysics Coupling and Long Term Behaviourin Rock Mechanics. 2006, pp. 15–26.

Berg, S., Ott, H., Klapp, S.A., Schwing, A., Neiteler, R., Brussee, N., Makurat, A., Leu, L.,Enzmann, F., Schwarz, J.O., Kersten, M., Irvine, S., Stampanoni, M., 2013. Real-time3D imaging of Haines jumps in porous media flow. Proc. Natl. Acad. Sci. U. S. A. 110(10), 3755–3759.

Bernier, F., Li, X.L., Bastiaens, W., 2007. Twenty-five years' geotechnical observation andtesting in the Tertiary Boom Clay formation. Geotechnique 57 (2), 229–237.

Boisson, J.Y., 2005. Clay Club Catalogue of Characteristics of Argillaceous Rocks, OECD/NEA/RWMC/IGSC (Working Group on Measurement and Physical Understanding ofGroundwater Flow Through Argillaceous Media) August 2005 Report NEA No. 4436(Brochure and CD-Rom Including Data Base). OECD/NEA Paris, France, p. 72.

Borthwick, V.E., Piazolo, S., Schmidt, S., Gundlach, C., Griera, A., Bons, P.D., Jessel, M.W.,2012. The Application of In Situ 3D X-ray Diffraction in Annealing Experiments.First Interpretation of Substructure Development in Deformed NaCl. Material ScienceForum, pp. 715–716 (461–466).

Callo, J.M., Hall, P.J., 2004. The application of small angle scattering techniques to porositycharacterization in carbons. Carbon 42 (7), 1299–1304.

Chenu, C., Tessier, D., 1995. Low temperature scanning electron microscopy of clay andorganic constituents and their relevance to soil microstructures. Scanning Microsc.9 (4), 989–1010.

Coles, M.E., Hazlett, R.D., Muegge, E.L., Jones, K.W., Andrews, B., Dowd, B., Siddons, P.,Peskin, A., Spanne, P., Soll, W.E., 1998. Developments in synchrotron X-raymicrotomography with applications to flow in porous media. SPE Reserv. Eval. Eng.1 (4), 288–296.

Cosenza, P., Ghorbani, A., Florsch, N., Revil, A., 2007. Effects of drying on the low-frequency electrical properties of Tournemire argillites. Pure Appl. Geophys. 164(10), 2043–2066.

Coussy, O., 1989. A general theory of thermoporoelastoplasticity for saturated porous ma-terials. Transp. Porous Media 4 (281), 293.

Curtis, M.E., Cardott, B.J., Sondergeld, C.H., Rai, C.S., 2012. Development of organic porosityin theWoodford Shale with increasing thermal maturity. Int. J. Coal Geol. 103, 26–31.

De Craen, M., Wang, L., Van Geet, M., Moors, H., 2004. Geochemistry of Boom Clay PoreWater at the Mol Site — Status 2004. SCK-CEN Report BLG-990Mol, Belgium.

Delage, P., Sultan, N., Cui, Y.-J., 2000. On the thermal consolidation of Boom clay. Can.Geotech. J. 37, 343–354.

Delage, P., LE, T.-T., Tang, A.-M., Cui, Y.-J., Li, X.-L., 2007. Suction effects in deep Boom clayblock samples. Geotechnique 57 (1), 239–244.

Desbois, G., Urai, J.L., Burkhardt, C., Drury, M.R., Hayles, M., Humbel, B., 2008. Cryogenicvitrification and 3D serial sectioning using high resolution cryo-FIB SEM technologyfor brine-filled grain boundaries in halite: first results. Geofluids 8, 60–72.

Desbois, G., Urai, J.L., Kukla, P.A., 2009. Morphology of the pore space in claystones — ev-idence from BIB/FIB ion beam sectioning and cryo-SEMobservations. eEarth 4, 15–22.

Desbois, G., Urai, J.L., Kukla, P.A., Konstanty, J., Baerle, C., 2011. High-resolution 3D fabricand porosity model in a tight gas sandstone reservoir: a new approach to investigatemicrostructures from mm- to nm-scale combining argon beam cross-sectioning andSEM imaging. J. Struct. Geol. 32 (4), 580–594.

Desbois, G., Urai, J.L., de Bresser, J.H.P., 2012a. Fluid distribution in grain boundaries of nat-ural fine-grained rock salt deformed at low shear stress: implications for rheologyand transport properties. J. Struct. Geol. 43, 128–143.

Desbois, G., Urai, J.L., Kukla, P.A., Wollenberg, U., Pérez-Willard, F., Radí, Z., Sandor, R.,2012b. Distribution of brine in grain boundaries during static recrystallization inwet, synthetic halite: insight from Broad Ion Beam sectioning and SEM observationat cryogenic temperature. Contrib. Miner. Petrol. 163 (1), 19–31.

Desbois, G., Urai, J.L., Pérez-Willard, F., Radí, Z., Offern, S., Bukkart, I., Kukla, P.A.,Wollenberg, U., 2013. Argon broad ion beam tomography in a cryogenic scanningelectron microscope: a novel tool for the investigation of representative microstruc-tures in sedimentary rocks containing pore fluid. J. Microsc. 249 (3), 215–235.

Dierckx, A., 1997. Boom Clay in situ porewater chemistry. SCK•CEN Report BLG-734.Durand, C., Rosenberg, E., 1998. Fluid distribution in kaolinite- or illite-bearing cores:

cryo-SEM observations versus bulk measurements. J. Pet. Sci. Eng. 19, 65–72.

130 G. Desbois et al. / Engineering Geology 179 (2014) 117–131

Echlin, P., 1978. Low temperature scanning electron microscopy: a review. J. Microsc. 112(1), 47–61.

Erlandsen, S.L., Bittermann, A., Krumweide, M., Nelson, R.D., Chen, Y., White, J.G., 1997.High-resolution backscatter electron detection of cell surface molecules on humanplatelets using the double-layer coating method and cryo field emission scanningelectron microscopy. Scanning 19 (5), 356–360.

Fanchi, J.R., 2010. Rock–fluid interactions. Integrated Reservoir Asset Management. GulfProfessional Publishing, Boston, pp. 167–185.

Fassi-Fihri, O., Robin, M., Rosenberg, E., 1992. Wettability of reservoir rock at the porescale: contribution of cryo-scanning electron microscopy. Rev. Inst. Fr. Pétrol. 47(5), 685–701.

Fauchadour, D., Pouget, T., Lechaire, J.-P., Rouleau, L., Normand, L., 1999. Evaluation ofcryotechniques for TEM observation of sols — application to boehmite sols used incatalysts forming. Oils Gas Sci.Technol. Rev. IFP. 54 (No4), 513–524.

Favvas, E.P., Sapalidis, A.A., Stefanopoulos, K.L., Romanos, G.E., Kanellopoulos, N.K.,Kargiotis, E.K., Mitropoulos, A.Ch., 2009. Characterization of carbonate rocks by com-bination of scattering, porosimetry and permeability techniques. Microporous Meso-porous Mater. 120 (1–2), 109–114.

Francois, B., Laloui, L., Clement, L., 2009. Thermo-hydro-mechanical simulation of ATLASin situ large scale test in Boom Clay. Comput. Geotech. 36 (4), 626–640.

Frederik, P.M., Sommerdijk, N.A.J.M., 2005. Spatial and temporal resolution in cryo-electron microscopy — a scope for nano-chemistry. Curr. Opin. Colloid Interface Sci.10, 245–249.

Galle, C., 2001. Effect of drying on cement-based materials pore structure as identified bymercury intrusion porosimetry: a comparative study between oven-, vacuum-, andfreeze-drying. Cem. Concr. Res. 31 (19), 1467–1477.

Gasc-Barbier, M., Tessier, D., 2007. Structural modifications of a hard deep clayey rock dueto hygro-mechanical solicitations. Int. J. Geomech. 7 (3), 227–235.

Giraud, A., Rousset, G., 1996. Time dependent behavior of deep clays. Eng. Geol. 41,181–195.

Heath, J.E., Dewers, T.A., McPherson, B.J.O.L., Petrusak, R., Chidsey, T.C., Rinehart, A.J.,Mozley, P.S., 2011. Pore networks in continental and marine mudstones: characteris-tics and controls on sealing behavior. Geosphere 7, 429–454.

Hedan, S., Cozensa, P., Valle, V., Dudoignon, P., Fauchille, A.-L., Cabrera, J., 2012. Investiga-tion of the damage induced by desiccation and heating of Tournemire argillite usingdigital image correlation. Int. J. Rock Mech. Min. Sci. 51, 64–75.

Hellmuth, K.H., Siitari-Kauppi, M., Lindberg, A., 1993. Study of porosity and migrationpathways in crystalline rock by impregnation with 14C-polymethylmethacrylate. J.Contam. Hydrol. 13, 403–418.

Hemes, S., Desbois, G., Urai, J.L., Craen, M.d, Honty, M., 2011. Comparative study on poros-ity in fine- and coarse-grained Boom Clay samples (Mol-Dessel reference site,Belgium). External Report SCK•CEN, CEN-ER-157, 11/MDC/P-6. . , 46.

Hemes, S., Desbois, G., Urai, J.L., de Craen, M., Honty, M., 2013. Stratigraphic variation inthe morphology of porosity in the Boom Clay formation: insights using 2D high res-olution BIB-SEM imaging and mercury injection porosimetry. Neth. J. Geosci. Geol.Mijnbouw 92 (4), 275–300.

Holzer, L., Gasser, Ph., Kaech, A., Wegmann, M., Zingg, A., Wepf, R., Münch, B., 2007. Cryo-FIB–nanotomography for quantitative analysis of particle structures in cement sus-pensions. J. Microsc. 227, 216–228.

Holzer, L., Münch, B., Rizzi, M., Wepf, R., Marschall, P., 2010. 3D-microstructure analysis ofhydrated bentonite with cryo-stabilized pore water. Appl. Clay Sci. 47, 330–342.

Hornapour, M.M., Djabbarah, N.F., Sampath, K., 2003. Whole core analysis—experienceand challenges. Presented for Presentation at the SPE 13th Middle East Oil Show &Conference, Barhain 5–8 April 2003. , 81575. SPE, pp. 1–16.

Horseman, S.T., Winter, M.G., Entwistle, D.C., 1987. Geotechnical Characterization ofBoom Clay in Relation to the Disposal of Radioactive Waste. EUR 10987 ENCEC Pub-lishers, Luxembourg, p. 86.

Houben, M.E., Desbois, G., Urai, J.L., 2013. Pore morphology and distribution in the shalyfacies of Opalinus Clay (Mont Terri, Switzerland): insights from representative 2DBIB–SEM investigations on mm to nm scale. Appl. Clay Sci. 71, 82–97.

INISMa, 1998. Technical report no. 74.252. Mons, Belgium.Kang, M.-S., Watabe, Y., Tsuchida, T., 2003. Effect of Drying Process on the Evaluation

of Microstructure of Clays using Scanning Electron microscope (SEM) and mer-cury intrusion porosimetry (MIP). Proceedings of the Thirteenth (2003) Inter-national Offshore and Polar Engineering Conference Honolulu, Hawaii, USA,May 25–30, pp. 385–392.

Kele, S., Moors, H., Van Geet, M., De Craen, M., Valkiers, S., Varlam, M., 2005. Geochemicalinvestigations of the dissolved gases in the Boom Clay water. European Society forIsotope Research, VIII Isotope Workshop (June 25 to 30, 2005 in Leipzig), ExtendedAbstract Volume, UFZ-Report 02/2005 (ISSN:0948–9452).

Keller, L.M., Holzer, L., Wepf, R., Gasser, P., 2011. 3D geometry and topology of pore path-ways in Opalinus clay: implications for mass transport. Appl. Clay Sci. 52, 85–95.

Keller, L.M., Holzer, L., Wepf, R., Gasser, P., 2013. Pore space relevant for gas permeabilityin Opalinus clay: statistical analysis of homogeneity, percolation, and representativevolume element. J. Geophys. Res. Solid Earth 118, 2799–2812.

Kempin, D., 2012. A fully automatic microstructure segmentation technique for SE2 de-tector images of stone samples. Bachelor Thesis at RWTH Aache University, Instituteof Imaging & Computer Vision (35 pp.).

Klaver, J., Desbois, G., Urai, J.L., Littke, R., 2012. BIB–SEM study of the pore space morphologyin earlymature Posidonia Shale from theHils area, Germany. Int. J. Coal Geol. 103, 12–25.

Klinkenberg, M., Kaufhold, S., Dohrmann, R., Siegesmund, S., 2009. Influence of carbonatemicrofabrics on the failure strength of claystones. Eng. Geol. 107, 42–54.

Kovalevych, V.M., Carmona, V., Pueyo, J.J., Peryt, T.M., 2005. Ultramicrochemical analyses(UMCA) and cryogenic scanning electron microscopy (cryo-SEM–EDS) of brines inhalite-hosted fluid inclusions: a comparative study of analytical data. Geochem. Int.43 (3), 268–276.

Kovalevych, V.M., Marshall, T., Peryt, T.M., Petrychenko, O.Y., Zhukova, S.A., 2006. Chem-ical composition in Neoproterozoic: results of fluid inclusion study of halite from saltrange (Pakistan) and Amadeus basin (Australia). Precambrian Res. 144, 39–51.

Lexa, O., 2010. PolyLX — the MATLAB™ toolbox for quantitative analysis of microstruc-tures. http://petrol.natur.cuni.cz/~ondro/polylx:home (12-12-2011).

CRC Handbook of Chemistry and Physics, In: Lide, D.R., Frederikse, H.P.R. (Eds.), 76th ed.CRC Press, Inc., Boca Raton, FL (1995).

Lima, A., Romero, E., Pina, Y., Gens, A., Li, X., 2012. Water retention properties of two deepBelgian Clay formations. Unsaturated Soils: Research and Applications. Springer, pp.179–184.

Loucks, R.G., Reed, R.M., Ruppel, S.C., Jarvie, D.M., 2009. Morphology, genesis, and distri-bution of nanometer-scale pores in siliceous mudstones of the Mississippian BarnettShale. J. Sediment. Res. 79, 848–861.

Loucks, R.G., Reed, R.M., Ruppel, S.C., Hammes, U., 2010. Preliminary classification of ma-trix pores in mudrocks gulf coast association of geological societies transactions. , 60,pp. 435–441.

LTGH, 1996. Inventarisatie van de kennis van de leperiaanklei in functie van onderzoeknaar diepe berging van hoogradioactief afval. Laboratorium voor Toegepaste Geologieen Hydrogeologie - LTGH. TGO Report 94/30. Gent Universiteit.

Mann, U., Neisel, J.D., Burchard, W.G., Heinen, V., Welte, D.H., 1994. Fluid-rock interfacesas revealed by cryo-scanning electron microscopy. , 12. First break.

Marko, M., Hsieh, C., Schalek, R., Frank, J., Mannella, C., 2007. Focused-ion-thinning offrozen-hydrated biological specimens for cryo-electron microscopy. Nat. MethodAdv. Online Publ. 4, 215–217 DOI: 10.1038.

Marschall, P., Horsman, S., Gimmi, T., 2005. Characterisation of gas transport properties ofthe Opalinus Clay, a potential host rock formation for radioactive waste disposal. OilGas Sci. Technol. 60 (1), 121–139.

Inc. MathWorks, 2011. MATLAB 7.12.0.635 (R2011a).McCann, T., 1998. The Rotliegend of the NE German Basin: background and prospectivity.

Pet. Geosci. 4, 17–27.Medina, C.R., Rupp, J.A., Barnes, D.A., 2011. Effects of reduction in porosity and per-

meability with depth on storage capacity and injectivity in deep saline aquifers:a case study from the Mount Simon Sandstone aquifer. Int. J. Greenh. Gas Control5 (1), 146–156.

Menendez, B., David, C., Martinez, Nistal A., 2001. Confocal scanning laser microscopy ap-plied to the study of pore and cracks networks in rocks. Comput. Geosci. 27,1101–1109.

Mertens, J., Bastiaens, W., Dehandschutter, B., 2004. Characterisation of induced disconti-nuities in the Boom Clay around the underground excavations (URF, Mol, Belgium).Appl. Clay Sci. 26, 413–428.

Negre, M., Leone, P., Trichet, J., Defarge, C., Boero, V., Gennari, M., 2004. Characterization ofmodel soil colloids by cryo-scanning electron microscopy. Geoderma 121 (1–2),1–16.

Onaisi, A., Audibert, A., Bieber, M.T., Bailey, L., Denis, J., Hammond, P.S., 1993. X-ray to-mography vizualization and mechanical modelling of swelling shale around thewellbore. J. Pet. Sci. Eng. 9 (4), 313–329.

Otani, J.O., Obara, Y., 2004. X-ray CT for Geomaterials Soils, Concrete, Rocks: Proceedingsof the International Workshop on X-ray CT for Geomaterials, GEO2003, 6–7 Novem-ber 2003. CRC Press, Kumamoto, Japan (379p).

Pape, H., Tillich, J.E., Holz, M., 2005. Pore geometry of sandstone derived from pulsed fieldgradient NMR. J. Appl. Geophys. 58, 232–252.

Pearson, F.J., Arcos, D., Bath, A., Boisson, J.-Y., Fernandez, A.M., Gäbler, H.-E., Gaucher, E.,Griffault, L., Hernan, P., Waber, H.N., 2003. Mont Terri Project — geochemistry ofwater in the Opalinus Clay formation at the Mont Terri Rock Laboratory. Reports ofthe FOWG, no. 5, Bern 2003.

Pineda, J., Romero, E., Gómez, S., Alonso, E., 2010. Degradation effects at micro-structural scale and their consequences on macroscopic behaviour of a slightlyweathered siltstone. In: Bolton, Malcolm (Ed.), From micro to macro, two vol-ume set. Geomechanics and Geotechnics. CRC Press, pp. 73–78 (2010).

Plötze, M., Kahr, G., Dohrmann, R., Weber, H., 2007. Hydro-mechanical, geochem-ical and mineralogical characteristics of the bentonite buffer in a heater ex-periment: the HE-B project at the Mont Terri Rock Laboratory. Phys. Chem.Earth 32, 730–740.

Prêt, D., Sardini, P., Beaufort, D., Zellagui, S., Sammartino, S., 2004. Porosity distribution ina clay gouge by image processing of 14C-PolyMethylMethAcrylate (14C-PMMA) au-toradiographs. Case study of the fault of St Julien (Basin of Lodève, France). Appl. ClaySci. 27, 107–118.

Prêt, D., Sammartino, S., Beaufort, D., Meunier, A., Fialin, M., Michot, L., 2010. A newmeth-od for quantitative petrography based on image processing of chemical elementmaps: part I. Mineral mapping applied to compacted bentonites. Am. Mineral. 95,1379–1388.

Prodanović, M., Lindquist, W.B., Seright, R.S., 2007. 3D image-based characterization offluid displacement in a Berea core. Adv. Water Resour. 30 (2), 214–226.

Radlinski, A.P., Mastalerz, M., Hinde, A.L., Hainbuchner, M., Rauch, H., Baron, M., Lin, J.S.,Fan, L., Thiyagarajan, P., 2004. Application of SAXS and SANS in evaluation of porosity,pore size distribution and surface area of coal. Int. J. Coal Geol. 59 (3–4), 245–271.

Renard, F., 2012. Microfracturation in rocks: from microtomography images to processes.Eur. Phys. J. Appl. Phys. 60, 24203.

Renard, F., Bernard, D., Thibault, X., Boller, E., 2004. Synchrotron 3D microtomography ofhalite aggregates during experimental pressure solution creep and evolution ofthe permeability. Geophys. Res. Lett. 31, l07607. http://dx.doi.org/10.1029/2004gl019605.

Robin, M., Rosenberg, E., Fassi-Fihri, O., 1995. Wettability studies at the pore level: a newapproach by use of Cryo-SEM. SPE Form. Eval. 10 (1), 11–19.

Robinet, J.C., Sardini, P., Coehlo, D., Parneix, J.-C., Prêt, D., Sammartino, S., Boller, E.,Altmann, S., 2012. Effects of mineral distribution at mesoscopic scale on solute

131G. Desbois et al. / Engineering Geology 179 (2014) 117–131

diffusion in a clay-rich rock: example of the Callovo-Oxfordian mudstone (Bure,France). Water Resour. Res. 48 (5), W005554.

Russ, J.C., Dehoff, R.T., 2000. Practical Stereology, 2 ed. Kluwer Academic/Plenum, NewYork.

Sammartino, S., Bouchet, A., Prêt, D., Parneix, J.-C., Tevissen, E., 2002. Spatial distribution ofporosity and minerals in clay rocks from the Callovo-Oxfordian formation (Meuse/Haute-Marne, Eastern France)—implications on ionic species diffusion and rocksorption capability. Appl. Clay Sci. 23 (1–4), 157–166.

Sardini, P., El Albani, A., Prêt, D., Gaboreau, S., Siitari-Kaupi, M., Beaufort, D., 2009. Map-ping and quantifying the clay aggregate microporosity in medium- to coarse-grained sandstones using the 14C-PMMA method. J. Sediment. Res. 79, 584–592.

Sargent, J.A., 1988. Low temperature scanning electron microscopy: advantages andapplications. Scanning Microsc. 2 (2), 835–849.

Savoye, S., Goutelard, F., Beaucaire, C., Charles, Y., Fayette, A., Herbette, M., Larabi, Y.,Coelho, D., 2011. Effect of temperature on the containment properties of argillaceousrocks: The case study of Callovo–Oxfordian claystones. J. Contam. Hydrogeol. 125,102–112.

Schenk, O., Urai, J.L., Piazolo, S., 2006. Structure of grain boundaries in wet, synthetic poly-crystalline, statically recrystallizing halite — evidence from cryo-SEM observations.Geofluids 6, 93–104.

Shepherd, T.J., Ayora, C., Cendon, D.I., Chenery, S.R., Moissette, A., 1998. Quantitative sol-ute analysis of single fluid inclusions in halite by LA-ICP-MS and cryo-SEM–EDS.Complementary microbeam techniques. Eur. J. Mineral. 10 (6), 1097–1108.

Siitari-Kauppi, M., 2002. Development of 14C-polymethylmethacrylatemethod for the char-acterisation of low porosity media[PhDthesis] University of Helsinki, Finland (156 p.).

SCK-CEN, 1997. HADES TOUR GUIDE, Notebook, fifth ed.Soe, A.K.K., Osada, M., Takahashi, M., Sasaki, T., 2009. Characterization of drying-induced

deformation behaviour of Opalinus Clay and tuff in no-stress regime. Environ. Geol.58 (6), 1215–1225.

Tiab, D., Donaldson, E.C., 1996. Petrophysics theory and practice of measuring reservoirrock and fluid transport properties. Gulf Publishing Company, Houston, TX (706 pp.).

Tritlla, J., Cardellach, E., 1997. Fluid inclusions in pre-ore minerals from the carbonate-hosted mercury deposits in the Espadan Ranges (eastern Spain). Chem. Geol. 137,91–106.

Underwood, E.E., 1970. Quantitative Stereology. Addison-Wesley, Reading, Massachu-setts, USA (ISBN: 978–0201076509).

Van Geet, M., Volckaert, G., Roels, S., 2005. The use of microfocus X-ray computed tomog-raphy in characterising the hydration of a clay pellet/powder mixture. Appl. Clay Sci.29 (2), 73–87.

Van Marcke, Ph., Laenen, B., 2005. The Ypresian clays as possible host rock for radioactivewaste disposal: an evaluation. Technical Report: NIROND TR-2005/01 (159 pp.).

Vizika, O., Rosenberg, E., Kalaydjian, F., 1998. Study of wettability and spreading impact inthree-phase gas injection by cryo-scanning electron microscopy. J. Pet. Sci. Eng. 20,189–202.

Walther, P., Muller, M., 1999. Biological ultrastructure as revealed by high resolution cryo-SEM of block faces after cryo-sectioning. J. Microsc. 19, 279–287.

Wang, L.L., Bornert, M., Chancole, S., Yang, S., Heripré, E., Tanguy, A., Caldemaison, D.,2013. Micro-scale experimental investigation of the swelling anisotropy of theCallovo-Oxfordian argillaceous rock. Clay Miner. 48, 391–402.

Ward, C.R., Nunt-jarunwong, S., Swanson, J., 2005. Use of mineralogical analysis ingeotechnical assessment of rock strata for coal mining. Int. J. Coal Geol. 64,156–171.

Wildenschild, D., Hopmans, J.W., Vaz, C.M.P., Rivers, M.L., Rikard, D., Christensen, B.S.B.,2002. Using X-ray computed tomography in hydrology: Systems, resolutions, andlimitations. J. Hydrol. 267, 285–297.

Zeelmaekers E. (2011). Computerized qualitative and quantitative clay mineralogy: Intro-duction and application to known geological cases. Unpublished PhD thesis,Katholieke Universiteit (K.U.) Leuven. Groep Wetenschap en Technologie (Heverlee,Leuven), 397 pp.

Zhang, S., Gogos, G., 1991. Film evaporation of a spherical droplet over a hot surface: fluidmechanics and heat/mass transfer analysis. J. Fluid Mech. 222, 543–563.