Contents lists available at ScienceDirect

Journal of Structural Geology

journal homepage: www.elsevier.com/locate/jsg

Microstructures in landslides in northwest China – Implications for creepingdisplacements?

M. Schäbitza, C. Janssena,∗, H.-R. Wenkb, R. Wirtha, B. Schucka, H.-U. Wetzela, X. Mengc,G. Dresena

aGFZ German Research Centre for Geosciences, Telegrafenberg, Potsdam, 14473, GermanybDepartment of Earth and Planetary Science, University of California, Berkeley, CA, 94720, USAc Research School of Arid Environment and Climate Change, Lanzhou University, 222 South Tianshui Road, Lanzhou, 730000, China

A R T I C L E I N F O

Keywords:LandslideMicrostructureTexturePyrophylliteGraphiteGrain size reduction

A B S T R A C T

Microstructures, mineralogical composition and texture of selected landslide samples from three landslides inthe southern part of the Gansu Province (China) were examined with optical microscopy, transmission electronmicroscopy (TEM), x-ray diffraction (XRD) and synchrotron x-ray diffraction measurements. Common sheetsilicates are chlorite, illite, muscovite, kaolinite, pyrophyllite and dickite. Other minerals are quartz, calcite,dolomite and albite. In one sample, graphite and amorphous carbon were detected by TEM-EDX analyses andTEM high-angle annular dark-field images. The occurrence of graphite and pyrophyllite with very low frictioncoefficients in the gouge material of the Suoertou and Xieliupo landslides is particularly significant for reducingthe frictional strength of the landslides. It is proposed that the landslides underwent comparable deformationprocesses as fault zones. The low friction coefficients provide strong evidence that slow-moving landsliding iscontrolled by the presence of weak minerals. In addition, TEM observations document that grain size reductionin clayey slip zone material was produced mainly by mechanical abrasion. For calcite and quartz, grain sizereduction was attributed to both pressure solution and cataclasis. Therefore, besides landslide composition, theoccurrence of ultrafine-grained slip zone material may also contribute to weakening processes of landslides. TEMimages of slip-zone samples show both locally aligned clay particles, as well as kinked and folded sheet silicates,which are widely disseminated in the whole matrix. Small, newly formed clay particles have random orienta-tions. Based on synchrotron x-ray diffraction measurements, the degree of preferred orientation of constituentsheet silicates in local shear zones of the Suoertou and Duang-He-Ba landslide is strong. This work is the firstreported observation of well-oriented clay fabrics in landslides.

1. Introduction

Slip zones of landslides are natural shear zones produced by dif-ferent stress configurations and propagate through different lithologies(Wen and Aydin, 2004). The composition and microstructures oflandslide slip zones are fundamental for understanding the mechanismsof landsliding and shear behavior of soils (Skempton and Petley, 1967;Skempton, 1985; Wen and Aydin, 2004). Slope failure often depends onthe complex interaction of tectonic activity, site morphology, variationsin humidity, and geological factors such as presence of a weak glideplane, as well as texture and mineralogy of both the host materials andthe slip zone. To understand the underlying mechanisms of landslides, alarge number of studies have microstructurally examinated slip zones(e.g. Wen and Aydin, 2003; Bhandary et al., 2005; Chen et al., 2014; Jiaet al., 2014).

Some mechanisms involved in the formation of landslides may besimilar to those that occur in fault zones. Particularly, the abundance ofclay minerals in landslides may affect the mechanical properties of theslip zones similarly to clay-rich fault gouges (e.g. Warr and Cox, 2001;Saffer et al., 2001; Wen and Aydin, 2003; Bhandary et al., 2005; Mooreand Rymer, 2007, 2012; Moore and Lockner, 2008; Collettini et al.,2009; Tembe et al., 2010; Schleicher et al., 2010; Holdsworth et al.,2011; Bradbury et al., 2011; Lockner et al., 2011; Janssen et al., 2014).Also, the critical earthquake magnitude required to trigger landslidesmay be affected by conversion of a flocculated (stable) clay structure toa dispersed (unstable) structure due to changes in pH of the pore fluidor creep-induced textural changes (e.g. Rosenquist, 1966; Loeken,1971; Reves et al., 2006).

Microstructures (e.g. amorphous material/melt, brittle fracturing,dissolution-precipitation processes, intracrystalline plasticity, micro-

https://doi.org/10.1016/j.jsg.2017.11.009Received 10 May 2017; Received in revised form 9 November 2017; Accepted 17 November 2017

∗ Corresponding author.E-mail address: jans@gfz-potsdam.de (C. Janssen).

Journal of Structural Geology 106 (2018) 70–85

Available online 24 November 20170191-8141/ © 2017 Elsevier Ltd. All rights reserved.

T

pores, clay fabric) in the principal slip zone (PSZ) of landslides wereformed by shear deformation due to dynamic slip and/or downslopemovement by creeping. For example, the geometric patterns of themicrostructures in the slip zone of the Shek Kip Mei landslide (HongKong, China) are similar to the S-C fabrics observed in tectonic shearzones (Wen and Aydin, 2003). Particle size reduction by crushing ofgrains and reorientation of grains by shearing generates an extremelyfine-grained gouge in faults as well as landslides. Therefore, some de-formation structures produced during faulting may help identify mi-crostructures resulting from slope failure and landsliding.

Many quantitative studies of clay fabric intensity in natural faultzones, using X-ray texture measurements, have shown that sheet silicatefabrics in fault gouges are predominantly weak (e.g. Warr and Cox,2001; Shimamoto et al., 2001; Solum and van der Pluijm, 2009; Haineset al., 2009; Schleicher et al., 2010; Wenk et al., 2010; Buatier et al.,2012; Janssen et al., 2012). The fabric strengths are generally muchsmaller than in shales, slates, and schists (e.g. Wenk et al., 2010;Haerinck et al., 2015).

So far, preferred orientation, also known as fabric or texture, of clayparticles in landslides has been examined only by microstructural ob-servations (e.g. Kawamura et al., 2007; Chen et al., 2014; Jia et al.,2014) and by a combination of microscopy and image analyses (Wenand Aydin, 2003, 2004, 2005). Chen et al. (2014) suggested thatmacroscopic sliding in the slip zone was most likely dominated bysliding of sheet silicate particles. Wen and Aydin (2003, 2004) re-cognized that porosity, particle-size distribution and particle orienta-tion are key factors controlling the mechanical properties of landslideslip zones. During sliding induced by heavy rainfall, particle movementwithin the slip zone was governed by fluidized particulate flow, re-sulting in weak particle alignment or random particle orientation (Wenand Aydin, 2005). For example, microstructures of the Qinyu landslideslip zone, observed by SEM, mainly show a flocculated structure (Jiaet al., 2014).

In this paper, composition, clay fabrics and microstructures fromsamples of three landslides in the southern part of Gansu Province(China) were examinated. Samples were collected from the main sliplayer and the surrounding damage zone. We analyzed deformationmicrostructures and determined fabric intensities of the clay gouge ofthe slip zone samples. For the first time, lattice-preferred orientation ofconstituent minerals in landslide samples was quantified using syn-chrotron X-ray diffraction measurements. The purpose of this in-vestigation is to determine to what extent microstructures and/orcomposition influence the slip behavior of landslides. Finally, wecompare the results with observations from well-investigated faultzones (San Andreas Fault, Chelungpu Fault) to discuss how the sharedtextural and compositional characteristics are indicative of commondeformation processes. (e.g. Janssen et al., 2011, 2014).

2. Geological setting

The geography of Central China is affected by active mountainbuilding that is associated with frequent occurrences of devastatingearthquakes and mass movements (Bai et al., 2012; Sun et al., 2015).From more than 2000 medium and large landslides reported withinZhouqu and Wudu County, we selected three, Xieliupo, Suoertou andDuang-He-Ba landslide, for our study. Xieliupo and Suoertou landslideshave experienced slow creeping deformation, threatening hundreds ofthousands of people's lives in Zhouqu and Wudu County (Sun et al.,2015). The Duang-He-Ba landslide is a typical loess landslide, with loessmoving on top of Silurian slates and phyllites bedrocks. These facts,together with excellent exposure conditions make them ideal candi-dates to study landsliding.

The three landslides investigated in this study are situated along theBailong River Corridor in the southern part of the Gansu province(Fig. 1a; Jiang and Wen, 2014; Jiang et al., 2014; Yu et al., 2015). TheBailong River Corridor is a tectonically and seismically active region

related to Himalayan orogenic deformation and one of the four mostactive areas of landslides and debris flows in China (Wang et al., 2013).It experienced two giant debris flows on August 8, 2010, as a result of atorrential rain. The debris flows destroyed half of the Zhouqu town andkilled 1756 people (Tang et al., 2011).

The Xieliupo landslide (samples X1-X9) is located on the north bankof the Bailongjinag river (Figs. 1b and 2a). The geology of this area ismainly composed of Silurian slates and phyllites, Permian limestones,Devonian limestones and slates, and Quaternary loess deposits (Tanget al., 2011; Yu et al., 2015). The formation of the landslide is con-trolled by the active Pingding-Huama fault with a mean slip rate of2.6 mm/year (Meng, unpublished data, Fig. 1b). The active Xieliupolandslide is about 2600 m long and 550 m wide with a thickness of 50 m(Jiang et al., 2014). Movement has been observed in the lower part of aslope whose toe was cut off for constructing the S313 provincialhighway (Sun et al., 2015). The landslide moving/creep rate is severalmm per year. In addition, an excavation pit provided a sample of claygouge (slip zone) material from the neighboring Suoertou landslide(sample S1), which is also affected by the same long-term creep de-formation. The lithological structure of both landslides and their linkingto the active Pingding-Huama fault are the same (Figs. 1a–b and 2b), sowe included the Suoertou gouge material in our analysis. The Duang-He-Ba landslide (samples D1-D5) is located in the southeastern tip ofZhouqu County, 20 km north-west of Wudu (Figs. 1a and 2c). Here, ariver eroded a whole profile through a landslide down to its host rockand parallel to its movement direction.

3. Methods

Microstructures of all samples were studied using optical andtransmission electron microscopy (TEM). Optical inspection of thinsections allowed us to identify representative areas characterized byfine-grained, clay-rich material potentially affected by fracturing anddissolution-precipitation processes. These areas were marked on high-resolution optical scans and prepared for TEM using a focused ion beam(FIB) device (FEI FIB200TEM) to avoid preparation-induced damage(Wirth, 2004, 2009). TEM was performed with a FEI Tecnai G2 F20 X-Twin TEM/AEM equipped with a Gatan Tridiem energy filter, a Fi-schione high-angle annular dark field detector (HAADF) and an energydispersive X-ray analyzer (EDX). TEM diffraction patterns also provideda first indication for orientation and distribution of clay particles.

The crystallographic orientation of sheet silicates was determinedfrom synchrotron X-ray diffraction images measured at the high-energybeam line ID-11C of the Advanced Photon Source (APS) at ArgonneNational Laboratory. Because of the high costs and the considerableamount of time necessary for this measurement, we selected only themost promising slip zone samples S1 and D3. The method is describedin detail in a tutorial (Lutterotti et al., 2013; Wenk et al., 2014). Amonochromatic X-ray beam with a wavelength of 0.107863 Å and1 × 1 mm in size was used to penetrate through a 1-mm thick slab ofsample. Diffraction images were recorded with a Perkin Elmer2024 × 2014 image plate detector. Seven images for each sample weremeasured at different sample tilt relative to the incident X-ray, from−45° to 45° in 15° incremental steps. These sample tilts are necessary toprovide adequate pole-figure coverage. The X-ray beam was alsotranslated parallel to the rotation axis over 2 mm to provide sufficientgrain statistics. Images were then deconvoluted for phase fractions,crystallite size, and preferred orientation with the Rietveld method(Rietveld, 1969) implemented in the MAUD (Materials Analysis UsingDiffraction) software (Lutterotti et al., 1997, 2013). Orientation dis-tributions for sheet silicates obtained by the Rietveld method wereexported from MAUD and further processed in BEARTEX (Wenk et al.,1998) to obtain pole figures, and to rotate and plot pole figures.

X-ray diffraction (XRD) was used to analyze fault rock compositionof 11 samples (Table 1a–b). Samples were dried and ground to a finepowder (2g) before analysis. X-ray diffraction analyses were conducted

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

71

Fig. 1. Location of the study area. (a) Satellite image of the Bailong River Corridor in the southern part of the Gansu province of China showing the regional fault system and the positionof landslides. The area of detailed investigation is indicated by rectangular area. (b) Geological map of the Zhouqu County showing the Suoertou and Xieliupo landslides in contact withthe Pingding-Huama Fault (modified after Yu et al., 2015).

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

72

on air-dried sample slides before and after treatment with ethyleneglycol to identify smectite and heating to 550 °C to differentiate be-tween chlorite and kaolinite. X-ray patterns were collected using a

PANalytical Empyrean X-ray diffractometer operating with Bragg-Brentano geometry at 40 mA and 40 kV with CuKα radiation, a step sizeof 0.013 °2Θ from 4.6 to 85 °2Θ and 60 s per step. The Rietveld

Fig. 2. Detailed maps, field photographs and optical images of thin section of investigated (a) Xieliupo, (b) Suertou and (c) Duang-He-Ba landslides. The yellow circles with numbersindicate sample position and sample number. Sample D6 was taken outside of the profile. (For interpretation of the references to colour in this figure legend, the reader is referred to theweb version of this article.)

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

73

algorithm BGMN was used for quantitative analysis (Bergmann et al.,1998; see supplementary material).

4. Results

4.1. Landslides and corresponding XRD analyses

The sliding surface (slip zone) of the Xieliupo landslide is formedbetween grey-brown clayey soil with carbonate rock fragments in theupper part and black clayey soil with carbonaceous slate rock fragmentsof Middle Devonian in the lower part (Fig. 2a). The analyzed sampleswere taken directly from the very fine-grained clayey slip zone mate-rial, the surrounding damage zone and the overlying loess layers. Thefoliated material from the slip zone differs from the adjacent damagezone material due to its grey to brown color and smaller grain sizes. Atlow magnification, slip-zone samples show a weak foliation defined byaligned phyllosilicates that is remarkably similar to clay fault gougestructures.

XRD analyses and corresponding Rietveld fits illustrate a complexcomposition with many contributing phases (Table 1a–b, Fig. 3a–c).The poorly sorted slip-zone samples X4 and X5 are mainly composed ofquartz and calcite/dolomite embedded in a clay-rich matrix of illite,kaolinite, chlorite and a minor amount of pyrophyllite (Table 1a). Da-mage-zone samples (X7, X8, X9) are very heterogeneous. Quartz, cal-cite/dolomite and clay mineral contents vary highly between thesesamples (Table 1a).

The Suoertou landslide represents a deeper section of the same(fault-) slip zone (Fig. 1b). The slip zone (gouge material) of thislandslide is a black seam of very soft and clayey material, containingquartz clasts of different size (Fig. 2b). The black gouge is derived fromneighbouring carbonaceous slates (Jiang et al., 2014). The black gougesample S1 is composed of a very fine-grained phyllosilicate matrix

(mainly pyrophyllite, illite, kaolinite, dickite and chlorite; Table 1a).Some black patches in the gouge matrix are organic material (see sec-tion 4.3). The amount of pyrophyllite in the Suoertou gouge sample(28.9%) is much greater than in the Xieliupo and Duang-He-Ba slip zonesamples (5.4% and 0%; Table 1a–b).

The Duang-He-Ba outcrop provides a complete cross section throughthe whole landslide body. This allowed sampling of the complete profilefrom the surface of the slide down to the host rock. In this outcrop, theslip zone is up to 20 cm thick and located at the boundary betweenSilurian phyllite and the overlaying Quaternary loess deposits with athickness of up to 5 meters (Fig. 2c). The dark brown slip-zone sample(D3) is rich in sheet silicates (chlorite, illite, muscovite) and quartz,additional constituents are plagioclase and calcite/dolomite (Table 1b).Similar to the Xieliupo landslide, the mineralogical composition of thedamage zone (samples D2 and D5) surrounding the slip zone is veryheterogeneous.

4.2. Optical microscopy

Inspection of the samples by optical microscopy revealed a distinctfoliation, grain-scale fracturing and minor faults. The foliation is de-fined by shape-preferred orientation of fragments, oriented fractures,pressure solution seams and aligned phyllosilicates, forming an inter-connected network. However, in contrast to the macroscopic de-formation patterns in the field and in specimens, microscopic ob-servations revealed considerable differences in foliation intensitybetween damage-zone and slip-zone samples and between differentlandslides (Fig. 4a–c). For example, the foliation defined by sub-parallelalignment of sheet silicates in sample S1 (slip zone of Suoertou land-slide) is more intensely developed as compared to the Xieliupo andDuang-He-Ba slip zone samples (samples X4, X5 and D3; Fig. 4a). For asingle landslide, we found that the foliation in samples from the damage

Table 1. Quantitative phase analyses of selected landslide samples based on Rietveld refinements of XRD spectra and high energy x-ray diffraction images. Weight is followed by estimatedstandard deviation. (a) Xieliupo and Suoertou samples. (b) Duang-He-Ba samples.a

(a)

Composition Sample X4 [wt%] Sample X5 [wt%] Sample X7 [wt%] Sample X8 [wt%] Sample X9 [wt%] Sample S1 [wt%] Sample S1 MAUD [wt%]

Quartz 26.7 0.6 27.1 0.6 23.3 0.3 21.1 0.4 34.4 0.7 9.9 0.6 9.5. 0.2Orthoclase 2.3 0.6 2.7 0.5 2.9 0.4 3.7 0.4 2.3 0.6Albite 9.1 0.3 9.8 0.5Calcite 18.4 0.5 18.5 0.4 40.6 0.5 31.4 0.5 4.6 0.4Dolomite 4.7 0.5 8.6 0.7 1.8 0.3 2.0 0.2 6.2 0.6 8.7 0.6 < 1Illite/mica 25.4 0.6 25.3 0.6 15.8 1.1 17.5 0.8 30.6 1.2 20.6 1.4 32.0 0.2Kaolinite 7.7 0.8 6.6 0.7 2.9 0.6 3.9 0.5 1.2 0.5 6.2 1.2 2.1 0.1Chlorite 6.9 0.7 4.2 0.6 2.2 0.5 4.5 0.6 5.1 0.8 1.5 0.6 < 0.1Pyrophyllite 3.7 0.9 5.4 0.9 0.8 0.4 2.5 0.9 4.4 0.8 28.9 1.1 35.8 0.2Hornblende 3.1 0.6 0.7 0.4 0.3 0.2 1.1 0.5Dickite 14.2 1.2 8.4 0.1Anatase 1.1 0.2 0.7 0.2 0.4 0.1 0.2 0.1Siderite 8.5 0.6 11.7 0.1Rutile 1.2 0.2 1.1 0.2

(b)

Composition Sample D1 [wt%] Sample D2 [wt%] Sample D3 [wt%] Sample D3 MAUD [wt%] Sample D5 [wt%] Sample D6 [wt%]

Quartz 42.2 0.8 34.8 0.6 33.0 0.5 8.7 0.4 13.3 0.5 26.5 0.5Orthoclase 3.3 0.5 3.1 0.7 3.0 0.7Albite 16.5 0.7 10.9 0.7 6.4 0.5 4.3 0.1 3.3 0.7 3.5 0.5Microcline 4.9 0.6 4.1 0.7 4.5 0.8Calcite 7.3 0.3 10.6 0.4 4.8 0.3 < 0.1Dolomite 2.3 1.5 1.5 0.3 0.8 0.3Illite/mica 14.2 0.2 28.4 43.5 1.1 68.3 0.4 54.2 0.4 45.5 0.5Kaolinite 5.8 0.1Chlorite 8.7 0.6 9.8 0.6 22.2 0.1 13.0 0.1 13.0 0.1 17.7 0.7Anatase 0.6 0.1

a Samples S1 and D3 were used to prepare two subsamples for XRD and MAUD.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

74

zone is much less developed than in the slip zone (Fig. 4a–b). The hostrock shows a weak foliation that is possibly due to deposition and likelynot a result of older landslide events (Fig. 4c). However, sheet silicatesare difficult to identify accurately in optical thin sections due to theirsmall grain sizes.

Contrary to foliation patterns, the resulting grain sizes displayed inFig. 4a–c indicate only minor differences (with exception of sample S1)between slip-zone material, host rock and damage zone samples. Theslip zone and damaged zone samples of all three landslides consist of anextremely fine-grained, dark reddish-brown to black matrix with em-bedded angular to subrounded quartz grains (Fig. 4a–b). But quartzgrains in slip-zone sample S1 are more rounded to subrounded andmuch smaller than those from the sliding surface samples of the Xieliupoand Duang-He-Ba landslides. Likewise, dark pressure-solution seamsindicating dissolution-precipitation processes are observed in all sam-ples. Other typical pressure-solution features like insoluble material(pressure solution residues) were observed in virtually all thin sections,but more often in the slip-zone samples (Fig. 4a–c). Well-oriented mi-neral fragments suggest local shearing (Fig. 4a).

4.3. TEM observations

The TEM observations are consistent with those from optical mi-croscopy. TEM images display microfabrics that differ between the

selected landslides, but slip zone and reworked material from the da-mage zone show similar microfabrics (Figs. 5a–f and 6a-d). The slip-zone material of the Xieliupo and Duang-He-Ba landslides (samples X4and D3) contains stacks of parallel-oriented detrital clay particles withvarying orientation (Figs. 5a and 6b). Calcite grains and to a lesserextent, quartz grains are partly dissolved (Fig. 5a–b). At other places,TEM observations of the same slip-zone samples reveal intensely com-minuted grains. The extremely fine-grained matrix particles (< 50 nm)are interpreted to be produced by mechanical abrasion (sample D3,Fig. 6c). Quartz grains act as a planer, reducing the grain size of detritalsheet silicate crystals. Textures implying mechanical comminution ofclay particles were also observed in several previous studies (e.g.,Chenu and Guerif, 1991; Baudett et al., 1999; Wen and Aydin, 2003).Wen and Aydin (2003, 2004) concluded that landslide slip zones mayhave experienced grainscale cataclastic deformation. Some clay mi-nerals in slip zone and damage zone samples, protruding into open porespaces, are interpreted as grains, newly formed by dissolution-pre-cipitation processes of grain boundaries (Fig. 5b). They are randomlyoriented. The gouge material of the Souertou landslide (sample S1)contains organic material and is characterized by a strong fabric de-fined by preferred orientation of aligned phyllosilicates (Fig. 6a). Thisstrong fabric covers the entire sample. A significant amount of graphiteand amorphous carbon were only observed in gouge material of sampleS1 by TEM-EDX analyses and high-angle annular dark field images

Fig. 3. XRD spectra for selected samples. (a) Sample X4,Xieliupo; (b) Sample S1, Suertou; and (c) Sample D3, Duang-He-Ba landslide. Some diffraction peaks of correspondingminerals are indexed (compare with Table 1 and diffractionpatterns from synchrotron hard x-ray diffraction in Fig. 10).

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

75

(Fig. 7a–c).The poorly sorted loess and reworked damage zone material of the

Xieliupo and Duang-He-Ba landslides (samples X1, X2, X3, X7 and D2)are characterized by detrital and newly formed clay particles. Similar tothe slip zone samples, randomly oriented clay minerals alternate withaggregates showing preferred orientation (samples X2 and X8;Fig. 5d–e). The latter only occur locally and form face-to-face particlecontacts. These well-oriented stacks form small (< 1 μm) domains(Fig. 5f) and do not produce a macroscopic fabric at the thin-sectionscale. They are widely disseminated in the whole matrix. In some areas,TEM micrographs illustrate clay flakes arranged in a card house fabric(Fig. 5e). Some quartz and/or feldspar grains in the matrix are wrappedby clay minerals (Fig. 6d). Toward the sliding surface, stacked clayparticles are aligned parallel or kinked and folded (sample X3; Fig. 5c).Partly dissolved grains, newly formed clay particles and highly com-minuted grains are observed in damage zone material and in slip zonesamples.

Pores were found in all samples, but unequally distributed. In TEMimages, we mostly observed sub-angular voids located between largergrains and fragments (Fig. 5d). Sub-rounded pores are formed byleaching of carbonates in the sliding surface (Fig. 5a). Inter-clay layerpores occur between illite clay flakes (Fig. 6b). These voids show largeaspect ratios with a long axis of several hundred nm mostly orientedparallel to the clay layers. Illite clay flakes with edge-to-face contactform card house structures with pore spaces between the flakes

(Fig. 5e). The pores were possibly filled with formation water and/orhydrothermal fluids possibly indicating elevated pore-fluid pressurepreventing pore collapse.

4.4. Preferred orientation data

A rough estimate of local fabric strength can be made by inter-preting the TEM diffraction patterns in selected areas. Preferred or-ientation corresponds to rings with distinct diffraction spots whereas aweak fabric is characterized by ring-shaped diffraction patterns. Thepreferred orientation of aligned phyllosilicates in the matrix of sampleX7 indicates a relatively strong fabric (Fig. 8a). In contrast, the randomorientation of clay particles in sample X5 only produces diffractionrings with many spots, suggesting a weak fabric (Fig. 8b). Note that thefault slip sample X5 reveals a weaker fabric compared to the damagezone sample X7.

Quantitative crystallographic preferred orientation (CPO) analyseswere performed on two slip-zone subsamples (from sample S1 withblack clay layer, Suoertou landslide, and sample D3 with brown claylayer, Duang-He-Ba landslide), with synchrotron X-ray diffraction on avolume of about 4 mm3. CPO is immediately evident in 2D synchrotrondiffraction images from the variations of intensity along diffractionrings of sheet silicate minerals (Fig. 9). Greater intensity (darker area)along the rings indicates more lattice plane normals oriented in thatparticular direction.

Fig. 4. Optical photomicrographs of thin sections. The sample number is in the lower left corner. (a) Thin sections from the weakly to strong foliated clayey slip-zone (gouge) matrix withembedded angular fragments and quartz grains. (b) Thin sections from the damage zone material with embedded angular fragments and quartz grains. (c) Thin sections from host rocks.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

76

2D diffraction images were decomposed into 36 diffraction patternsby azimuthal averaging over 10 sectors and these 36 patterns for 7images at different sample tilts for each sample (total of 252 diffractionpatterns) were then used as data for the Rietveld refinement. Fig. 10shows average diffraction patterns for S1 and D3 with measured data(dots) and corresponding Rietveld fit (solid line). The pattern can becompared with XRD results (Fig. 3a–c). Phase fractions from the MAUDRietveld analysis are similar to the XRD analysis of powders(Table 1a–b), with dominating pyrophyllite (35.8 wt% in Sample S1),illite-mica (32 wt% in sample S1 and 68.3 wt% in sample D3), kaolinite(2.1 wt% in sample S1 and 5.8 wtl% in sample D3) and chlorite (13 wt

% in sample D3). The values are not identical because the actualsamples are different and XRD powder analysis averages over muchlarger sample volumes.

The purpose of the synchrotron analysis was not to refine phasefractions, but rather to quantify CPO. Preferred orientation is displayedin diffraction images (Fig. 9) and is even more obvious in stacks ofdiffraction patterns as function of azimuthal angle where sheet silicatesdisplay strong intensity variations in contrast to quartz (Fig. 11). Thegood agreement not only indicates a reliable identification of compo-nent minerals and their volume fractions but also of their orientationpatterns.

Fig. 5. TEM images from microstructures of theXieliupo landslide. The sample number is in thelower left corner. (a) Embayed calcite grain thatis interpreted to be partially dissolved. The grainis surrounded by stacks of parallel-oriented clayparticles. (b) Eroded Quartz grain boundary andopen pore spaces with clay minerals. (c) Kinkedand bent clay particles. (d) Open pore spacesbetween quartz grains and preferred orientationof sheet silicates. (e) Clay flakes in edge-to-facecontact form card house structures with smallopen pore spaces between the clay flakes. (f)Calcite grains embedded in a strong clay fabric.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

77

Fig. 6. TEM images from microstructures of theSuoertou and Duang-He-Ba landslides. The samplenumber is in the lower left corner. (a) Pyropylliteand organic material embedded in aligned phyl-losilicates. (b) Pore spaces between clay flakes.(c) Quartz grain acts as a planer along illitecrystals. Notice that the planed illite particleshave an extremely fine grain size. (d) Quartzgrain is wrapped by clay and poorly sorted matrixparticles.

Fig. 7. TEM microphotographs of sample S1. (a)High-angle annular dark-field (HAADF) imageshowing graphite embedded in sheet silicates. (b)High-resolution transmission electron micro-scopy (HRTM) image of graphite (c) HRTEMimage of carbon.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

78

From orientation distribution patterns of sheet silicate minerals,pole figures were calculated and used to display preferred orientationpatterns (Fig. 12). The coordinate system was rotated to display thecharacteristic (001) maxima of sheet silicates in the center. Pole figuresare plotted in equal-area projection and pole densities are representedin multiples of random distribution (m.r.d.). Note that all samples showconsiderable heterogeneity and the methods used in this study, such asTEM and synchrotron X-ray diffraction, only investigate small volumes.Pole figures are displayed for kaolinite, illite/mica, pyrophyllite (onlysample S1), dickite (only sample S1) and chlorite (only sample D3).

There are strong maxima, up to 4 m.r.d. for pyrophyllite in sample S1and up to 10 mrd for illite in sample D3. Kaolinite has a stronger texturein sample D3 (MaxPf = 4.5 m.r.d.) than in sample S1(MaxPf = 1.6 m.r.d.). In sample S1, pyrophyllite has the strongesttexture (MaxPf = 4.0 m.r.d.), whereas the orientation distribution ofdickite is relatively weak with 1.8 m.r.d.. The texture of chlorite wasmeasured only in sample D3. With 3.3 m.r.d. it is quite strong. Thetexture of quartz in both samples is random.

Fig. 8. Brightfield, and HRTEM TEM images. (a)The spotty rings indicate preferred orientation ofclay crystals (sample X7) and (b) the diffusescattering intensity shows a weak fabric (sampleX5).

Fig. 9. 2D Diffraction images of samples (a) S1and (b) D3 at 0° sample tilt. Some Debye rings areidentified. Azimuthal intensity variations alongDebye rings are indicative of crystallographicpreferred orientation. Note that the intensity ofheterogeneous ring development is more obviousin (b) than in (a).

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

79

Fig. 10. Average diffraction patterns of the diffraction images in Fig. 9 for samples (a) S1 and (b) D3. Dots are experimental data and line is Rietveld fit. Some diffraction peaks areidentified. Below the diffraction pattern are mineral phases and corresponding diffractions.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

80

5. Discussion

5.1. Landslide composition, and the role of fluids in slip-zone mineralogy

Differences in the mineralogical composition of the landslides asrevealed by XRD data (Table 1a–b) may be attributed to different hostrock compositions and provenances of the sample, i.e. if they originatefrom the damage or the slip zone. Quartz and chlorite are the onlyminerals present in all samples. However quartz and chlorite contentsdiffer between landslides due to different host rocks (e.g. Devonianslates vs. Silurian phyllites). The presence of illite/mica in slip zonesamples of the Xieliupo, Suoertou and Duang-He-Ba landslides may bepartly related to the illite content of the respective host rocks. A clearincrease in the abundance of sheet silicates toward the slip zone was notobserved. The absence of smectite in all samples contrasts with thecompositions found in slip zones of other landslides (Shuzui, 2001).This result is consistent with investigations of Mancktelow et al. (2016).The authors suggest that a clay-rich low permeability slip zone mayprevent influx of meteoric water and thus inhibit smectite growth infault gouge.

Fluid-related alteration reactions leading to mineral dissolution andprecipitation are evident by the presence of authigenic phyllosilicates(Fig. 5b) and variations in the mineralogical composition such as theenrichment of kaolinite and pyrophyllite in slip zone samples(Table 1a–b). The lower calcite content of the Xieliupo slip-zone samplescompared to the host rock should also be seen in this context becausefluid-assisted mass transfer by pressure solution potentially led to cal-cite depletion in the slip zone and the relative enrichment of clay mi-nerals (compare Gratier et al., 2011). The interpretation of fluid-as-sisted alteration in slip zone samples is consistent with many fault zonestudies that identified fluid-related alteration assemblages controllingfault mechanics and fluid flow (e.g. Warr and Cox, 2001; Boullier et al.,2009, 2011; Schleicher et al., 2009; Moore and Rymer, 2012).

5.2. Slip zone frictional strength

The presence of graphite and pyrophyllite in gouge material of theSuoertou landslide may indicate that frictional strength of the landslidewas low. Both minerals have extremely low friction coefficients with0.1 for graphite and< 0.2 for pyrophyllite (Al2SiO4). The TEM

Fig. 11. Stack of 36 diffraction spectra averaged over 10°azimuth of diffraction images in Fig. 9 for samples (a) S1and (b) D3 represented as map plots with intensity (grayshade) as function of 2θ. Bottom are measured data, top theRietveld fit. Some diffraction peaks are identified.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

81

evidence of graphite in the slip zone sample of the Suoertou landslide(sample S1, Table 1a) is consistent with fault zone studies. Within afault gouge, graphite possibly serves as a lubricant reducing frictionalstrength (Oohashi et al., 2012; Kuo et al., 2014; Li et al., 2013;Nakamura et al., 2015; Togo et al., 2011). This interpretation is sup-ported by high-velocity friction experiments performed on samplescontaining amorphous carbon. Dramatic weakening was observed whencarbon transformed to graphite (Oohashi et al., 2011, 2013, 2014).Several processes of graphite enrichment have been proposed (Oohashiet al., 2014): (1) graphite may derive from host rock via pressure so-lution or solution transfer processes (Oohashi et al., 2012); (2) hydro-thermal precipitation of graphite from a C-O-H-rich fluid (Zulauf et al.,1999); and (3) graphitization of amorphous carbon due to seismic faultmotion (Oohashi et al., 2011). The latter process was recognized in theprincipal slip zone of several fault zones, which underwent frictionalheating due to rapid sliding (e.g. Kuo et al., 2014; Nakamura et al.,2015). For instance, XRD analyses of coseismic fault gouge samplesfrom the Longmenshan fault zone in Sichuan (China) indicate graphite(Togo et al., 2011). Similar to the outcrop samples, drill core materialfrom the principal slip zone also contains graphite (Kuo et al., 2014).The authors suggested that graphite may have formed due to seismicslip affecting in turn the mechanical properties of the slip layer. Strain-induced amorphization of graphite with increasing frictional slip wasalso reported from fault zones of the Hidaka metamorphic belt(Nakamura et al., 2015). These fault zone studies suggest that the co-existance of graphite and amorphous carbon in gouge material of theSuoertou landslide (Fig. 7) is a strong indicator for graphitization ofamorphous carbon during seismic slip. It is assumed that a formerearthquake (co-seismic event) preceeds the slow-moving (a-seismic)landsliding.

In general, the presence of weak clay minerals and graphite in slip-zone gouge strongly suggests that landsliding may be controlled by thepresence of these weak minerals in agreement with a plethora oflandslide and fault zone studies (e.g. Warr and Cox, 2001; Boullieret al., 2009, 2011; Schleicher et al., 2009; Moore and Rymer, 2012). Forexample, the study of several landslides in Japan indicate that frictionalresistance decreases as clay content (smectite) increases (Shuzui, 2001).Within the creeping portion of the San Andreas Fault (SAF) at shallowdepth, the growth of Mg-rich phyllosilicates may play a key role in themechanical behavior of the SAF and can be directly related to faultweakening (e.g. Moore and Rymer, 2007, 2012; Schleicher et al., 2010;Holdsworth et al., 2011; Bradbury et al., 2011; Lockner et al., 2011).

Unlike Xieliupo and Suoertou landslides, Duang-He-Ba samples do notcontain extremely weak clay minerals. Here, illite (friction coeffi-cient > 0.2) is the most abundant mineral phase in the slip zone(sample D3) and acts as a lubricant reducing frictional strength. Illitemay be formed by fragmentation of muscovite particles (Buatier et al.,2012) that were initially present in the host rock (sample D5). The slipzone sample is depleted in chlorite and muscovite while the quartzcontent increases compared to the host rock. These differences maysuggest that the slip zone material is not only derived from the host rock(phyllite) but also includes material from hanging wall sediments.

5.3. Microstructures

In general, the observed microstructures are very similar to micro-structures in gouge samples from major faults such as the San AndreasFault and the Chelungpu Fault (e.g. Hadizadeh et al., 2012; Holdsworthet al., 2011; Moore and Rymer, 2012; Janssen et al., 2014). For ex-ample, the TEM observations indicate that both fracturing and pressure

Fig. 12. (001) Pole figures of sheet silicates for gouge samples (a) S1 and (b) D3. Equal-area projection, contours in multiples of random distribution. Samples have been rotated from anarbitrary initial orientation to bring the (001) maximum into the center.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

82

solution have led to grain size reduction. Phyllosilicates scraped byrigid clasts such as quartz (Fig. 6b) indicate that the extremely fine-grained clay matrix particles (< 50 nm) in slip zone and damage zonesamples (Fig. 6c–d) were produced by mechanical abrasion rather thanby pressure solution. The opposite is true for calcite and quartz. Thedissolved grain boundaries indicate grain size reduction by pressuresolution (Fig. 5a–b). Microstructural observations of fault gouges haveshown that grain size reduction may occur over a wide range of sliprates and different deformation mechanisms (Stünitz et al., 2010).Strong comminution may be related to dynamic rock fracturing duringpropagation of a single earthquake (Olgaard and Brace, 1983; Wilsonet al., 2005; Brantut et al., 2008) and slow grinding (creeping) pro-cesses (Stünitz et al., 2010), involving pressure solution (Gratier andGamond, 1990; Gratier et al., 2011). However, irrespective of whetherthe ultrafine-grained particles in slip zone samples were produced bycoseismic and/or aseismic slip, the presence of this fine-grained mate-rial implies that frictional strength of the slip zone material is sig-nificantly reduced, which consequently promotes slip (e.g. Marone andScholz, 1989; Chester et al., 2005; Wilson et al., 2005; Ma et al., 2006;Collettini et al., 2009; Han et al., 2010, 2011; Viti, 2011).

5.4. Clay fabric in slip zones

Microstructural analysis, using optical inspection, electron micro-scopy as well as X-ray diffraction techniques, reveal both randomly andpreferentially oriented sheet silicates in the landslide slip-zone samples.TEM images of sample S1 (Suoertou landslide) display well-orientedphyllosilicate stacks for the entire TEM sample (Fig. 6a). These ob-servations are consistent with synchrotron X-ray diffraction measure-ments that reveal relatively strong clay fabrics with very regular or-ientation distributions and pole figures ranging from 1.6 m.r.d. inkaolinite to 4 m.r.d. in pyrophyllite. (Fig. 12). Synchrotron X-ray dif-fraction measurements for sample D3 reveal an even stronger clayfabric than for sample S1. Maximum (001) pole densities range from3.3 m.r.d. in chlorite, 4.5 m.r.d. in kaolinite to 10. m.r.d. in illite. Incontrast, microscopic photographs and TEM images of damaged andslip zone samples X3, X4 and X5 (Xieliupo landslide) and sample D3(Duang-He-Ba landslide) show both locally aligned clay particles, aswell as kinked and folded sheet silicates, which are widely disseminatedin the whole matrix (Figs. 5a, c, 6b). Small, newly formed clay particlesalso have a random orientation (Fig. 5b). These results suggest a cleardistinction between a strong clay fabric in local slip zones (“slip-zonefabric”) and a weak “matrix fabric”. The latter is considered primarily aresult of the growth of new sheet silicates and fault-related deformationas kinking and rotation on the particle scale (Janssen et al., 2012).

Considering microstructures and pole figures, it is probable that thestrong fabrics in samples D3 and S1 are caused by different processes.We suggest two possible explanations for the strong clay fabrics insample S1 (Suoertou landslide). (1) The original fabric of clay in amarine environment will be flocculated (card-house fabric; Bennettet al., 1991). Later, when the saline water is leached out by freshgroundwater, the stable clay fabric is changed to an unstable arrange-ment of a dispersed clay fabric subject to collapse when disturbed orshaken, for example by earthquakes (Black et al., 1999). Second, fo-liated domains with preferred orientation were formed by pressuresolution as described for the damaged zone of the San Andreas Fault(Richard, 2014). Accounting for the microscopic and TEM observations,the microstructural record of the Suoertou sample S1 shows earthquake(co-seismic) related weakening mechanisms, which include amorphouscarbon and graphite. Hence, the first explanation is probably the mostlikely scenario for the Suoertou landslide.

Deformation patterns and the strong fabric in sample D3 (Duang-He-Ba landslide) are difficult to interpret. We found neither indication thatan earthquake acted as an triggering mechanism for ground failure, norstrong evidence for pressure-solution creep. One possible explanation isthat patches of oriented clay particles from the host rock (phyllite) may

be mechanically emplaced into the slip zone. Another reason for thisfinding might be that clay minerals are likely to creep relatively easilybecause of their low friction coefficient (Gratier, 2011).

The detection of a preferred clay fabric orientation in landslidesamples by synchrotron X-ray diffraction measurements is consistentwith microfabrics of the Jin'nosuke-dani landslide (northern centralJapan, Kawamura et al., 2007) but differs markedly from fault gougesamples, which have generally weak fabrics (e.g. Alpine Fault, Warrand Cox, 2001; Moab Fault, Solum et al., 2005; Carbonera Fault, Solumand van der Pluijm, 2009; Bogd fault, Buatier et al., 2012; San AndreasFault, Wenk et al., 2010; Janssen et al., 2014; Chelungpu Fault, Janssenet al., 2014). Solum and van der Pluijm (2009) described clay fabrics infault rocks of the Carboneras Fault (Spain) as isolated and extremelylocalized structures in layers and domains, extending just a few tens ofmicrometers from the slip surfaces into the sample matrix structures.The overall preferred orientation is poor above the scale of a few tens ofmicrometers (Solum et al., 2010). Haines et al. (2009) suggested thatthe very weak fabric in fault gouges could result from growth of claysafter fault slip ceased. In the slip zone samples, we have no clear evi-dence for clay growing after landsliding.

Finally, the question arises about whether or not the composition ofslip zone samples, and/or the orientation of clay particles determine theweakness of landslides. Our study yields no direct information aboutthe shear strength of the analyzed slip-zone samples. Results fromfrictional sliding experiments, however, suggest that thin localizedzones of phyllosilicates along (and below) slip surfaces reduce thefriction coefficient (e.g. Tembe et al., 2006; Collettini et al., 2009;Solum et al., 2010). Studies on clay-bearing soils, on the other hand,document that clay with random orientations tend to be stronger andmore rigid (Mitchell, 1993). Considering these shear experiments, wesuggest that apart from slip-zone composition and extremely fine-grained matrix particles also the identified strong clay fabrics are re-sponsible for a significant reduction of shear strength of the landslides.

6. Conclusions

Microstructures and the presence of strong fabrics in samples de-rived from three landslide slip-zones constrain the interpretation of slipmechanisms that show some similarities with deformation processesdescribed from fault zone gouges. The presence of (weak) clay mineralsand graphite favor weakening in narrow slip zones of landslides andfaults. The process of grain-size reduction depends on the minerals.Fine-grained clay matrix particles were produced by mechanical abra-sion and calcite and quartz grain size reduction occurred by pressuresolution. We suggest that both processes result from sliding events andslow creep episodes. The degree of clay fabric intensity differs betweenthe selected samples. The strong fabric in slip zone samples of theSuoertou and Duang-He-Ba landslide is clearly different from fault gougesamples. To support this statement synchrotron X-ray diffraction mea-surements should be applied to a wider variety of samples from dif-ferent landslides to improve our understanding of landsliding.Speculatively, one of the worst natural disasters in China, the 2010Gansu mudslide, which killed 1756 people, could be caused by groundfailure due to the loss of strength in sensitive clays as described for theSuoertou landslide.

Acknowledgements

Jean-Pierre Gratier and an anonymous reviewer together with theeditorial guidance of William M. Dunne provided very constructivecomments and suggestions that helped to improve this paper. This workis part of the collaborative research program TIPTIMON funded by theFederal Ministry of Education and Research (FKZ 03G0809A). TheLanzhou University provided substantial financial and logistical sup-port. The authors wish to thank to R. Naumann for XRD and XRFanalysis and A. Hendrich and M. Dziggel for help with drafting. HRW

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

83

acknowledges support from NSF (EAR-0836402) and DOE (DE-FG02-05ER15637) and access to beamline 11-ID-C at the Advanced PhotonSource (APS) of Argonne National Laboratory and help from Y. Ren.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jsg.2017.11.009.

References

Bai, S., Wang, J., Zhang, Z., Cheng, C., 2012. Combined landslide susceptibility mappingafter Wenchuan earthquake at the Zhouqu segment in the Bailongjiang Basin, China.Catena 99, 18–25. http://dx.doi.org/10.1016/j.catena.2012.06.12.

Baudett, G., Perrotel, V., Seron, a., Stelletelli, M., 1999. Two dimensional communition ofkaolinite clay particles. Powder Technol. 105, 125–134.

Bennett, R.H., O'Brien, N.R., Hulbert, M.H., 1991. Determination of clay and shale mi-crofabric signatures: processes and mechanisms. In: Bennett, R.H., Bryant, W.R.,Hulbert, M.H. (Eds.), Microstructure of Fine-Grained Sediments. Springer-Verlag,New York, pp. 5–32.

Bergmann, J., Friedel, P., Kleeberg, R., 1998. BGMN- a new fundamental parametersbased Rietveld program for laboratory X-ray sources, it's use in quantitative analysisand structure investigations. ?CPD? Newsletter 20, 5–8.

Bhandary, N., Yatabe, R., Takata, S., 2005. Clay minerals contributing to creeping dis-placement of fracture zone landslides in Japan. In: Sassa, K., Fukuoka, H., Wang,F.W., Wang, G. (Eds.), Landslides – Risk Analysis and Sustainable DisasterManagement. Springer International Publishing, Switzerland, pp. 219–223.

Black, B.D., Solomon, B.J., Harty, K.M., 1999. Geology and Geological Hazards of TooeleValley and the West Desert Hazardous Industry Area, Tooele County, Utah, SpecialStudy 96. Utah Geological Survey.

Boullier, A.-M., Yeh, E.-C., Boutareaud, S., Song, S.-R., Tsai, C.-H., 2009. Microscaleanatomy of the 1999 Chi-Chi earthquake fault zone. Geochem. Geophys. Geosys. 10.http://dx.doi.org/10.1029/2008GC002252.

Boullier, A.M., 2011. Fault-zone Geology: Lessons from Drilling through the Nojima andChelungpu Faults, vol. 359. Geological Society, London, pp. 17–37 SpecialPublication.

Bradbury, K.K., Evans, J.P., Chester, J.S., Chester, F.M., Kirschner, D.L., 2011. Lithologyand internal structure of the San Andreas fault at depth based on characterization ofPhase 3 whole-rock core in the San Andreas Fault Observatory at Depth (SAFOD)borehole. Earth Planet. Sci. Lett. 304. http://dx.doi.org/10.1016/j.epsl.2011.07.020.

Brantut, N., Schubnel, A., Rouzaud, J.-N., Brunet, F., Shimamoto, T., 2008. High-velocityfrictional properties of a clay-bearing fault gouge and implications for earthquakemechanics. J. Geophys. Res. 113. http://dx.doi.org/10.1029/2007JB005551.

Buatier, M.D., Chauvet, A., Kanitpanyacharoen, W., Wenk, R., Ritz, J.F., Jolivet, M., 2012.Origin and behavior of clay minerals in the Bogd fault gouge, Mongolia. J. Struct.Geol. 34, 77–90.

Chen, J., Dai, F., Xu, L., Chen, S., Wang, P., Long, W., Shen, N., 2014. Properties andmicrostructure of a natural slip zone in loose deposits of red beds, southwesternChina. Eng. Geol. 183, 53–64. http://dx.doi.org/10.1016/j.enggeo.2014.10004.

Chenu, C., Guerif, J., 1991. Mechanical strength of clay minerals as influenced by anabsorbed polysaccharide. Soil Sci. Soc. Am. J. 55 (4), 1076–1080.

Chester, J.S., Chester, F.M., Kronenberg, A.K., 2005. Fracture surface energy of thePunchbowl fault, San Andreas system. Nature 437, 133–136.

Collettini, C., Niemeijer, A., Viti, C., Marone, C., 2009. Fault zone fabric and faultweakness. Nature 462. http://dx.doi.org/10.1038/nature08585.

Gratier, J.P., Gamond, J.F., 1990. Transition between seismic and aseismic deformation inthe upper crust. Geol. Soc. Spec. Publ. 54, 461–473.

Gratier, J.P., 2011. Fault permeability and strength evolution related to fracturing andhealing episodic processes (years to millennia): the role of pressure solution. Oil GasSci. Technol. Rev. IFP Energies Nouvelles 66, 491–506.

Gratier, J.-P., Richard, J., Renard, F., Mittempergher, S., Doan, M.-L., Di Toro, G.,Hadizadeh, J., Boullier, A.-M., 2011. Aseismic sliding of active faults by pressuresolution creep: evidence from the San Andreas Fault observatory at depth. Geology12, 1131–1134.

Hadizadeh, J., Mittempergher, S., Gratier, J.P., Renard, F., Di Toro, G., Richard, J., Babie,H.A., 2012. A microstructural study of fault rocks from the SAFOD: implications fordeformation mechanisms and strength of creeping segment of the San Andreas Fault.J. Struct. Geol. 42, 246–260.

Haerinck, T., Wenk, H.-R., Debacker, T.N., Sintubin, M., 2015. Preferred mineral or-ientation of a chloritoid-bearing slate in relation to its magnetic fabric. J. Struct. Geol.71, 125–135. http://dx.doi.org/10.1016/j.jsg.2014.09.013.

Haines, S.H., van der Pluijm, B.A., Ikari, M.J., Saffer, D.M., Marone, C., 2009. Clay fabricintensity in natural and artificial fault gouges: implications for brittle fault zoneprocesses and sedimentary basin clay fabric evolution. J. Geophys. Res. 114. http://dx.doi.org/10.1029/2008JB005866.

Han, R., Hirose, T., Shimamoto, T., 2010. Strong velocity weakening and powder lu-brication of simulated carbonate faults at seismic slip rates. J. Geophys. Res.115http://dx.doi.org/10.1029/2008JB006136. B03412.

Han, R., Hirose, T., Shimamato, T., Lee, Y., Ando, J., 2011. Granular nanoparticles lu-bricate faults during seismic slip. Geology 39, 599–602.

Holdsworth, R.E., van Diggeln, E., Spiers, C., de Bresser, H.J., Walker, R.J., Bowen, L.,2011. Fault rocks from SAFOD core samples. J. Struct. Geol. 33, 132–144.

Janssen, C., Wirth, R., Reinicke, A., Rybacki, E., Naumann, R., Kemmnitz, H., Wenk, H.-R., Dresen, G., 2011. Nanoscale porosity in SAFOD core samples (San Andreas Fault).Earth Planet. Sci. Lett. 301, 179–189.

Janssen, C., Kanitpanyacharoen, W., Wenk, H.-R., Wirth, R., Morales, L., Rybacki, E.,Kienast, M., Dresen, G., 2012. Clay fabrics in SAFOD core samples. J. Struct. Geol. 43,118–127.

Janssen, C., Wirth, R., Wenk, H.-R., Morales, L., Naumann, R., Kienast, M., Song, S.-R.,Dresen, G., 2014. Faulting processes in active faults – evidences from TCDP andSAFOD drill core samples. J. Struct. Geol. 65, 100–116.

Jia, X., Liang, S., Fan, C., 2014. Microstructural Characteristics of Qinyu landslide slipsoil, NW China. In: In: Sassa, K. (Ed.), Landeslide Science for a Safer Geoenvironment,vol. 2 Springer International Publishing, Switzerland. http://dx.doi.org/10.1007/978-3-319-05050-8_29.

Jiang, S., Wen, B.P., Zhao, C., Li, R., 2014. Factors controlling kinematic behaviour of ahuge slow-moving landslide in China. In: In: Sassa, K. (Ed.), Landeslide Science for aSafer Geoenvironment, vol. 2 Springer International Publishing, Switzerland. http://dx.doi.org/10.1007/978-3-319-05050-8_42.

Jiang, X., Wen, B.P., 2014. Creep behavior of slip zone of a giant slow-moving landslide inNorthwest China: the Suoertou landslide as an example. In: In: Sassa, K. (Ed.),Landeslide Science for a Safer Geoenvironment, vol. 2 Springer InternationalPublishing, Switzerland. http://dx.doi.org/10.1007/978-3-319-05050-8_23.

Kawamura, K., Ogawa, Y., Oyagi, N., Kitahara, T., Anma, R., 2007. Structural and fabricanalyses of basal slip zone of the Jin’nosuke-dani landslide, northern central Japan:its application to the slip mechanism of decollement. Landslides 4, 371–380. http://dx.doi.org/10.1007/s10346-007-0094-z.

Kuo, L.W., Li, H., Smith, S.A.F., Di Toro, G., Suppe, J., Song, S.-R., Nielsen, S., Sheu, H.-S.,Si, J., 2014. Gouge graphitization and dynamic fault weakening during the 2008 Mw7.9 Wenchuan earthquake. Geology 42, 47–50. http://dx.doi.org/10.1130/G34862.1.

Li, H., et al., 2013. Characteristic of the fault-related rocks, fault zones, and the principalslip zone in the Wenchuan earthquake fault scientific drilling project Hole-1 (WFSD-1). Tectonophysics 584, 23–42.

Lockner, D.A., Morrow, C., Moore, D., Hickman, S., 2011. Low strength of deep SanAndreas fault gouge from SAFOD core. Nature 472. http://dx.doi.org/10.1038/nature09927.

Loeken, T., 1971. Recent research at the Norwegian geotechnical institute concerning theinfluence of chemical additions on quick clay. Norges Geotek. Inst. Publ. 87,133–147.

Lutterotti, L., Matthies, S., Wenk, H.-R., Shultz, A.J., Richardson, J.W., 1997. Combinedtexture and structure analysis of deformed limestone from time-of-flight neutrondiffraction spectra. J. Appl. Phys. 81, 594–600.

Lutterotti, L., Vasin, R., Wenk, H.-R., 2013. Rietveld texture analysis from synchrotrondiffraction images: I. Basic analysis. Powder Diffr. 29, 76–84.

Ma, K., et al., 2006. Slip zone and energetics of large earthquake from the TaiwanChelungpu-fault Drilling project. Nature 444, 473–476.

Mancktelow, N., Zwingmann, H., Mulch, A., 2016. Timing and conditions of clay faultgouge formation on the Naxos detachment (Cyclades, Greece). Tectonics 35,2334–2344. http://dx.doi.org/10.1002/2016TC004251.

Marone, C., Scholz, C.H., 1989. Particle-size distribution and microstructures within si-mulated fault gouge. J. Struct. Geol. 11, 799–814.

Mitchell, J.K., 1993. Fundamentals of Soil Behavior, 2nd. ed. Wiley, New York 437 pp.Moore, D.E., Lockner, D.A., 2008. The friction in the temperature range 25°-400°C: re-

levance for Fault-one Weakening. Tectonophysics 449, 120–132.Moore, D.E., Rymer, M.J., 2007. Talc-bearing serpentinite and the creeping section of the

San Andreas fault. Nature 448 doi:10.1038.Moore, D.E., Rymer, M.J., 2012. Correlation of clayey gouge in a surface exposure of

serpentinite in the San Andreas Fault with gouge from the San Andreas Fault ob-servatory at depth (SAFOD). J. Struct. Geol. 38, 51–60.

Nakamura, Y., Oohashi, K., Toyoshima, T., Satish-Kumar, M., Akai, J., 2015. Strain-in-duced amorphization of graphite in fault zones of the Hidika metamorphic belt,Hokkaido, Japan. J. Struct. Geol. 72, 142–161. http://dx.doi.org/10.1016/jsg.2014.10012.

Olgaard, D.L., Brace, W.F., 1983. The microstructure of gouge from a mining-inducedseismic shear zone. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 20, 11–19.

Oohashi, K., Hirose, T., Shimamoto, T., 2011. Shear induced graphitization of carbo-naceous material during seismic fault motion: experiments and possible implicationsfor fault mechanics. J. Struct. Geol. 33, 1122–1134. http://dx.doi.org/10.1016/j.jsg.2011.01.007.

Oohashi, K., Hirose, T., Kabayashi, K., Shimamoto, T., 2012. The occurrence of graphite-bearing fault rocks in the Atotsugawa fault system, Japan: origin and implications forfault creep. J. Struct. Geol. 38, 39–50. http://dx.doi.org/10.1016/j.jsg.2011.10.011.

Oohashi, K., Hirose, T., Shimamoto, T., 2013. Graphite as lubricating agent in fault zones:an insight from low- to high-velocity friction experiments on a mixed graphite-quartzgouge. J. Geophys. Res. 118, 2067–2084. http://dx.doi.org/10.1002/jgrb.50175.

Oohashi, K., Han, R., Hirose, T., Shimamoto, T., Omura, K., Matsuda, T., 2014. Carbon-forming reactions under a reducing atmosphere during seismic slip. Geology. http://dx.doi.org/10.1130/G35703.1.

Reves, G.M., Sims, I., Cripps, J.C. (Eds.), 2006. Clay materials Used in Construction, vol.21. Geological Society, London, pp. 1–51 Engineering Geology, Special Publication.

Richard, J., 2014. Large-scale mass transfer related to pressure solution creep-faultinginteractions in mudstones: driving processes and impact of lithification degree.Tectonophysics 612–613, 40–55.

Rietveld, H.M., 1969. A profile refinement method for nuclear and magnetic structures. J.Appl. Crystallogr. 2, 65–71.

Rosenquist, I.T.H., 1966. Norwegian Research into the properties of quick clay – a review.Eng. Geol. 1 (6), 445–450.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

84

Saffer, D.M., Frye, K., Marone, C., Mair, K., 2001. Laboratory results indicating weak andpotentially unstable frictional behavior of smectite clay. Geophys. Res. Lett. 28.http://dx.doi.org/10.1029/2001GL012869.

Schleicher, A.M., Warr, L.N., van der Pluijm, B.A., 2009. On the origin of mixed-layeredclay minerals from the San Andreas Fault at 2.5-3 km vertical depth (SAFOD drillholeat Parkfield, California). Contrib. Mineral. Petrol. 157, 173–187.

Schleicher, A.M., van der Pluijm, B.A., Warr, L.N., 2010. Nanocoatings of clay and creepof the San Andreas fault at Parkfield. Geology 38, 667–670.

Shimamoto, T., Takemira, K., Fujimoto, K., Tanaka, H., Wibberley, C.A.J., 2001. NojimaFault Zone probing by core analyses. Isl. Arc. 357–359.

Shuzui, H., 2001. Process of slip-surface development and formation of slip-surface clay inlandslides in Tertiary volcanic rocks, Japan. Eng. Geol. 61, 199–219.

Skempton, A.W., Petley, D.J., 1967. The strength along structural discontinuities in stiffclays. Proc. Geotec. Conf. Oslo vo 2, 29–46.

Skempton, A.W., 1985. Residual strength of clays in landslides, folded strata and thelaboratory. Geotechnique 35 (1), 3–18.

Solum, J.G., van der Pluijm, B.A., Peacor, D.R., 2005. Neocrystallization, fabrics and ageof clay minerals from the exposure the Moab Fault, Utah. J. Struct. Geol. 27,1563–1576.

Solum, J.G., van der Pluijm, B.A., 2009. Quantification of fabrics in clay gouge from theCarboneras fault, Spain and implications for fault behavior. Tectonophysics 475,554–562.

Solum, J.G., Davatzes, N.C., Lockner, D.A., 2010. Fault-related clay authigenesis along theMoab Fault: implications for calculations of fault rock composition and mechanicaland hydrologic fault zone properties. J. Struct. Geol. 32, 1899–1911.

Stünitz, H., Keulen, N., Hirose, T., Heilbronner, R., 2010. Grain size distribution andmicrostructures of experimentally sheared granitoid gouge at coseismic slip rates -criteria to distinguish seismic and aseismic faults? J. Struct. Geol. 32, 59–69.

Sun, Q., Zhang, L., Ding, X.L., Hu, J., Li, Z.W., Zhu, J.J., 2015. Slope deformation prior toZhouqu, China landslide from INSAR time series analysis. Remote Sens. Environ. 156,45–57.

Tang, C., Rengers, N., van Asch, Th.W.J., Yang, Y.H., Wang, G.F., 2011. Triggeringconditions and depositional characteristics of a disastrous debris flow event inZhouqu city, Gansu Province, northwestern China. Nat. Hazards Earth Syst. Sci. 11,2903–2912.

Tembe, S., Lockner, D.A., Solum, J.G., Morrow, C.A., Wong, T.-F., Moore, D.E., 2006.Frictional strength of cuttings and core from SAFOD drillhole. Geophys. Res. Lett. 33.http://dx.doi.org/10.1029/2006GL02726.

Tembe, S., Lockner, D.A., Wong, T.-F., 2010. Effects of clay content and mineralogy onfrictional sliding behavior of simulated gouges: binary and ternary mixtures ofquartz, illite, and montmorillonite. J. Geophys. Res. 115. http://dx.doi.org/10.29/

2009JB006383.Togo, T., Shimamoto, T., Ma, S., Wen, X., He, H., 2011. Internal structure of

Longmenshan fault zone at Hongkou outcrop, Sichuan, China, that caused the 2008Wenchuan earthquake. Earth Sci. 24, 249–265. http://dx.doi.org/10.1007/s11589-011-0789-z.

Viti, C., 2011. Exploring fault rocks at the nanoscale. J. Struct. Geol. 33, 1715–1727.Wang, L., Wu, Z., Wang, P., Chen, T., 2013. Characteristic causation, and rehabilitation of

Zhouqu extraordinarily serious debris flow in 2010, China. J. Cent. South. Univ. 20,2342–2348. http://dx.doi.org/10.1007/s11771-013-1742-1.

Warr, L.N., Cox, S., 2001. Clay mineral transformation and weakening mechanisms alongthe Alpine Fault, New Zealand. In: In: Holdsworth, R.E., Strachan, R.A., Magloughlin,J.F., Knipe, R.J. (Eds.), The Nature and Tectonic Significance of Fault Weakening, vol.186. Geological Society, London, pp. 85–1001 Special Publication.

Wen, B.P., Aydin, A., 2003. Microstructural study of a natural slip zone: quantificationand deformation history. Eng. Geol. 68, 289–317.

Wen, B.P., Aydin, A., 2004. Deformation history of a landslide slip zone in light of soilmicrostructure. Environ. Eng. Geosci. 2, 123–149.

Wen, B.P., Aydin, A., 2005. Mechanism of a rain-induced slide-debris flow: constraintsfrom microstructure of its slip zone. Eng. Geol. 78, 69–88. http://dx.doi.org/10.1016/j.enggeo.2004.10.007.

Wenk, H.-R., Matthies, S., Donovan, J., Chateigner, D., 1998. BEARTEX: a Windows-basedprogram system for quantitative texture analysis. J. Appl. Crystallogr. 31, 262–269.

Wenk, H.-R., Kanitpanyacharoen, W., Voltolini, M., 2010. Preferred orientation of phyl-losilicates: comparison of fault gouge, shale and shist. J. Struct. Geol. 32, 478–489.

Wenk, H.-R., Lutterotti, L., Kaercher, P., Kanitpanyacharoen, W., Miyagi, L., Vasin, R.,2014. Rietveld texture analysis from synchrotron diffraction images: II. Complexmultiphase materials and diamond anvil cell experiments. Powder Diffr. 29,220–232.

Wilson, B., Dewers, t., Reches, Z., Brune, J., 2005. Particle size and energetics of gougefrom earthquake rupture zones. Nature 434, 749–752.

Wirth, R., 2004. A novel technology for advanced application of micro- and nanoanalysisin geosciences and applied mineralogy. Eur. J. Mineral. 16, 863–876.

Wirth, R., 2009. Focused Ion Beam (FIB) combined with SEM and TEM: advanced ana-lytical tools for studies of chemical composition, microstructure and crystal structurein geomaterials on a nanometre scale. Chem. Geol. 261, 217–229.

Yu, G., Zhang, M., Cong, K., Pei, L., 2015. Critical rainfall thresholds for debris flows inSanyanyu, Zhouqu county, Gansu province, China. Q. J. Eng. Geol. Hydrogeol.http://dx.doi.org/10.1144/qjegh2014-078.

Zulauf, G., Palm, S., Petschick, R., Spies, O., 1999. Element mobility and volumetric strainin brittle and brittle-viscous shear zones of the super deep well KTB. Ger. Chem. Geol.156, 135–149. http://dx.doi.org/10.1016/S0009-2541(98)00189-2.

M. Schäbitz et al. Journal of Structural Geology 106 (2018) 70–85

85