Sedimentary Geology 489 (2025) 106982

Contents lists available at ScienceDirect

Sedimentary Geology
journal homepage: www.elsevier.com/locate/sedgeo

Comparative grain size analysis of fine-grained sediments: A case study of

the Opalinus Clay, northern Switzerland
Géraldine Nicole Zimmerli a,* , Alicia Fantasia a , Stephan Wohlwend b,c , Gaudenz Deplazes d,
Anneleen Foubert a
a
Department of Geosciences, University of Fribourg, Chemin du Musée 6, 1700, Fribourg, Switzerland
b
Institute of Geological Sciences, University of Bern, Baltzerstrasse 1 + 3, 3012, Bern, Switzerland
c
Wohlwend Geologie, Eichrainstrasse 9, 8052, Zürich, Switzerland
d
National Cooperative for the Disposal of Radioactive Waste (Nagra), Hardstrasse 73, 5430, Wettingen, Switzerland

A R T I C L E I N F O A B S T R A C T

Editor: Dr Michal Gradzinski Grain size variability in fine-grained sediments is a critical factor influencing the physical, geomechanical,
geochemical, and hydrological properties of sedimentary rocks. Even small changes in grain size can significantly
Keywords: alter rock behavior. Despite many decades of research on fine-grained sediments in different depositional set­
Grain size analyses tings, studies focusing on the comparative and quantitative calibrated analysis of grain size variations in fine-
Thin sections
grained Mesozoic sediments are scarce. This study provides the first integrative, quantitative, comparative,
Computed tomography
and comprehensive analysis of grain size variability in the detrital fraction of the Opalinus Clay from northern
Laser particle size analysis
Elemental ratios Switzerland (Stadel-2 core), which serves as the designated host rock for radioactive waste disposal. The analyses
Current dynamics include 2D image analysis, X-ray micro-computed tomography, laser particle size analysis, and sieving/decan­
tation analysis, alongside mineralogical (X-ray diffraction, XRD) and geochemical (X-ray fluorescence, XRF)
analyses. Strong correlations between grain size and elemental ratios (Si/Al, Ti/Al, Zr/Al, Zr/Rb) suggest that
these ratios can serve as effective proxies for grain size variations in the Opalinus Clay, primarily controlled by
the presence of quartz and phyllosilicates. Results evidence the dominance of silt-sized particles in the Opalinus
Clay. Distinct trends of grain size coarsening and fining are observed based on XRF-derived grain size proxies,
providing the base for further studies on sediment transport processes, including sediment provenance, flow
strength, and current dynamics associated with the Opalinus Clay. Results highlight the importance of calibrating
grain size variations in fine-grained sediments before their application in the understanding of rock properties
and sedimentary processes.

1. Introduction these applications. Critical rock properties encompass physical (e.g.,

porosity, permeability, density), geomechanical (e.g., cohesion,
Fine-grained sediments play a critical role in numerous environ­ compressibility, shear strength), geochemical (e.g., mineralogical and
mental and industrial applications. They serve as host rocks for radio­ elemental composition, cation exchange capacity, isotopic composi­
active waste disposal (e.g., Bitzer, 1996), reservoirs in hydrocarbon tion), and hydrological parameters (e.g., hydraulic conductivity, capil­
exploration (e.g., Zou et al., 2022), formations for carbon capture and larity, retention, sorption, swelling and shrinkage). Each of these
storage (e.g., Ma et al., 2022), aquitards for groundwater protection and parameters is fundamentally influenced by grain size, which governs
water filtration (e.g., Cai et al., 2023), thermal buffers in geothermal pore structure, surface area, particle interactions, and fluid behavior.
energy systems (e.g., Nunn et al., 1999), and in soil science and agri­ Grain size analyses are therefore vital in characterizing fine-grained
culture (e.g., Braga et al., 2024). Additionally, they are used in con­ sediments. Particle size, shape, and arrangement directly control phys­
struction materials such as bricks, ceramics, and cement (Nweke et al., ical, mechanical, and chemical rock properties. Even within fine-grained
2023). classifications, small variations in grain size can significantly alter rock
A comprehensive understanding of rock properties is essential in all behavior.

* Corresponding author.
E-mail address: geraldine.zimmerli@unifr.ch (G.N. Zimmerli).

https://doi.org/10.1016/j.sedgeo.2025.106982
Received 28 May 2025; Received in revised form 5 September 2025; Accepted 28 September 2025
Available online 2 October 2025
0037-0738/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

For example, porosity and permeability are strongly influenced by sized particles and adsorbed onto fine-grained phyllosilicates (e.g.,
grain size. Clay-rich rocks, with fine particles and small pore throats, Calvert and Pedersen, 2007). Together, these ratios provide a multi-
exhibit low permeability and act as effective barriers, while silt-rich proxy geochemical framework for assessing detrital input and grain
rocks show improved pore connectivity and slightly higher, though size variability within the sedimentary sequence. Often, those proxies
still low, permeability (e.g., Zheng et al., 2015; Verweij et al., 2016; are used without robust calibration against direct granulometric mea­
Carcione et al., 2019; Díaz-Curiel et al., 2024). Geomechanically, fine- surements. Despite the potential of X-ray fluorescence (XRF) data for
grained clay-sized sediments display high cohesion, plasticity, and such applications, proper calibration remains essential.
compressibility but lower shear strength, whereas siltier materials have This work focuses on calibrating the grain size parameter in fine-
greater stiffness and shear resistance (e.g., Burland, 1990; Amann et al., grained sediments, using the Opalinus Clay (OPA) as a case study. The
2017; Crisci et al., 2021). Grain size also affects geochemical and hy­ OPA is a perfect analogue for the study of fine-grained sediments being
drological properties: clay-rich sediments have high cation exchange described mainly as silty claystone formation. The OPA was deposited
capacity (CEC), chemical reactivity, water retention, and sorption ca­ from the latest Toarcian to Early Aalenian and extends from eastern
pacities due to large surface areas and strong capillary forces; silt-rich France to eastern Bavaria (Germany) and from northern Bavaria to the
sediments are less reactive but more resistant to weathering and allow Swiss Plateau (Wetzel and Allia, 2003; Wohlwend et al., 2024; Zimmerli
for increased fluid movement, and provide slightly lower contaminant et al., 2025). Moreover, the OPA is recognized in Switzerland as a
retention (e.g., Delage et al., 2010; Dohrmann et al., 2012; Tournassat suitable host rock for the disposal of radioactive waste, owing to its
et al., 2015; Liu et al., 2018; Kneuker et al., 2023; Fernandes et al., 2024; favorable geotechnical and petrophysical properties (Nagra, 2002,
Ziefle et al., 2024). These variations underscore the importance of ac­ 2014a). While its bulk mineralogical and clay mineralogical composi­
curate grain size characterization in predicting rock behavior across tion is well documented (Mazurek et al., 2023), the grain size distribu­
applications. tion is less understood due to factors such as aggregation, flocculation,
Grain size is also a key parameter for interpreting depositional and the inherent fine-scale heterogeneity of clay mineral-rich sediments.
conditions, even in fine-grained sediments traditionally viewed as ho­ Grain size analyses of the OPA are scarce, and existing studies report
mogeneous. Variations in grain size can signal shifts in energy condi­ diverse, widespread and point-based data without offering a compre­
tions, sediment supply, and transport processes during deposition. hensive dataset (e.g., Nagra, 2014b; Minardi et al., 2016; Seiphoori
Fining-upward or coarsening-upward trends may reflect changes in et al., 2017; Ferrari et al., 2012, 2020; Kneuker et al., 2023).
current dynamics, basin evolution, or sea-level fluctuations. Addition­ This study aims to calibrate grain size analyses in fine-grained sed­
ally, grain size influences the settling behavior of particles and controls iments (OPA) using a comparative and technical approach. The objec­
the stratification and layering of fine-grained successions. As such, tives of this study are threefold: (1) to evaluate grain size variability
detailed grain size analysis contributes to more accurate reconstructions employing four distinct methodologies – 2D Image Analysis based on
of paleoenvironmental conditions and sedimentary processes. thin sections (TS), 3D visualization techniques (computed tomography,
Despite many decades of research on fine-grained sediments across CT), laser particle size analysis (LPSA), and sieving and decantation
various depositional settings, comparative studies on grain size varia­ methods; (2) to calibrate XRF-data as grain size proxy through
tions in fine-grained Mesozoic sediments remain scarce and are often not comparing grain size variability with data obtained from X-ray fluo­
well calibrated (e.g., Nagy et al., 1984; Burkhalter, 1996; Burkhalter rescence (XRF) analyses; and (3) to achieve a comprehensive under­
et al., 1997; Williams et al., 2001; Zhang et al., 2016; Ahmad et al., standing of the overall grain size variability within the OPA (Middle
2017; Fantasia et al., 2019; Ghaznavi et al., 2019; Wengler et al., 2019). Jurassic), emphasizing the role of grain size analysis in elucidating hy­
One of the major challenges lies in the intrinsic difficulty of accurately drodynamic conditions, depositional environments and rock mechanical
measuring grain size in such sediments. This difficulty arises from properties.
several interrelated factors: clay minerals often occur as aggregates or
flocculated particles, making it hard to resolve individual grains; fine- 2. Geological setting
grained consolidated rocks like claystones and siltstones frequently
exhibit cementation in contrast to their unconsolidated counterparts, This study takes the OPA as reference material for studying fine-
further complicating disaggregation and analysis; and grain size pa­ grained sediments. The OPA encompasses the latest Toarcian to Early
rameters often show significant heterogeneity. These issues limit the Aalenian (Opalinum Zone, including the Opalinum and Bifidum Sub­
reliability and comparability of grain size measurements and call for the zones; e.g., Wohlwend et al., 2024; Zimmerli et al., 2024; Fig. 1). The
use of multiple, complementary and calibrated analytical approaches. OPA was deposited in an epicontinental sea being part of the broader
Some studies use geochemical proxies – such as Si/Al, Ti/Al Zr/Al, Laurasian Seaway, which connected the Boreal Arctic Sea in the north to
and Zr/Rb ratios (e.g., Ohta, 2004; Calvert and Pedersen, 2007; the Tethys Ocean in the south (Bjerrum et al., 2001; Korte et al., 2015).
Bloemsma et al., 2012; Liang et al., 2013; Van der Schee et al., 2016; Wu Initially described as tempestite deposits (Wetzel and Allia, 2003),
et al., 2020; Zimmerli et al., 2024) – to infer grain size. Aluminum (Al) is Zimmerli et al. (2024) challenged this interpretation and suggested that
commonly used as a normalizing element for detrital input, given its alternative depositional models cannot be excluded (e.g., contourites,
predominance in fine-grained aluminosilicates and relatively stable tidalites, turbidites, hypopycnites, or hyperpycnites).
geochemical behavior (e.g., Calvert and Pedersen, 2007; Weltje and Today, the OPA is well documented in Switzerland (e.g., Hostettler
Tjallingii, 2008; Löwemark et al., 2011). The Si/Al ratio is frequently et al., 2017; Lauper et al., 2021; Zimmerli et al., 2024) and in southern
applied as a proxy for variations in quartz and aluminum silicate Germany (e.g., Franz and Nitsch, 2009; Dietze et al., 2021). The Early
(detrital) content relative to clay minerals, thereby reflecting shifts in Aalenian, during which much of the OPA was deposited, was generally
grain size (e.g., Calvert and Pedersen, 2007; Craigie and Rees, 2016). characterized by cooler climatic conditions (Nikitenko et al., 2008;
The Zr/Al ratio reflects the relative abundance of zircon, a heavy min­ Gómez et al., 2009; Rogov and Zakharov, 2010; Korte et al., 2015).
eral often enriched in silt-sized sediment fractions, making it a robust The OPA is a silty to fine sandy claystone formation with thin,
indicator of grain size distribution (Craigie and Rees, 2016). Similarly, intercalated carbonate-rich horizons, reaching a total thickness of
the Ti/Al ratio is used as a proxy for the concentration of titanium- approximately 80 to 120 m in northern Switzerland (Wetzel and Allia,
bearing heavy minerals, such as rutile and ilmenite, which are 2003; Zimmerli et al., 2024). Its mineralogical composition is dominated
commonly enriched in the silt to sand-sized fractions (e.g., Calvert and by clay minerals, carbonates, and quartz, with contributions of mica,
Pedersen, 2007). The Zr/Rb ratio could further enhance grain size in­ feldspar, pyrite, siderite and organic matter (Nagra, 2002). Although
terpretations (Wu et al., 2020) by contrasting a coarse-grained indicator fossil-poor, the OPA contains marine micro- and macrofossils including
(Zr) with rubidium (Rb), which is predominantly associated with clay- foraminifers, radiolarians, ostracods, palynomorphs, ammonites, and

2
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 1. A Localization of the drill core Stadel-2 (STA2) in Switzerland. B Core description of STA2 with the corresponding legend (carb. = carbonate; calc. =
calcareous lithologies). The stratigraphic framework encompasses both lithostratigraphy and biostratigraphy, as noted alongside the core (after Wohlwend et al.,
2024; Zimmerli et al., 2024). The sample locations (17 samples) are marked with green circles labeled with corresponding subfacies and sampling depth. C Mi­
crophotographs of the four subfacies (SF1, SF6, SF2, SF3; transmitted light, crossed polars). SF1 is characterized by a homogeneous, clay-rich subfacies. SF6 also
displays a homogeneous fabric but contains dispersed quartz grains. SF2 is rich in lenticular bedding structures within a clay-rich matrix. In contrast, SF3 shows
irregular bedding and a higher abundance of quartz grains within the argillaceous matrix compared to SF2.

3
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

bivalves (e.g., Ohmert et al., 1996; Hostettler et al., 2017; Wohlwend exclude them from the analysis. It should be noted that biogenic silica is
et al., 2024). insignificant in the OPA. Overall, we opted to focus in this study solely
At regional scale, six subfacies (SF1, SF2, SF3, SF4, SF5, and SF6) on the siliciclastic components, which are the dominant components of
characterize the OPA (after Lauper et al., 2018, 2021; Zimmerli et al., the detrital fraction. This allows to compare our grain size results across
2024). While SF4 and SF5 are rather described as calcareous sandy methods and with other studies.
mudstone to muddy sandstone, SF1, SF2, SF3, and SF6 are classical The Udden-Wentworth scale (clay: < 3.9 μm, silt: 3.9–63 μm, sand:
mudstones (Lauper et al., 2018, 2021; Zimmerli et al., 2024), repre­ 63–2000 μm) has been used in this study to allow (1) comparison be­
senting the most fine-grained fractions in the OPA. As such, only those tween the different methods and (2) comparison of grain size variations
fractions have been considered within the framework of this study. SF1 with other studies (Boggs, 2009; after Wentworth, 1922). We are aware
represents a dark gray, homogeneous claystone composed predomi­ that in many technical studies the upper limit of the clay fraction is set at
nantly of clay minerals (64 %) with subordinate quartz (25 %) and 2 μm (e.g., Ferrari et al., 2020; Kneuker et al., 2023). However, to allow
carbonates (11 %). SF2 is a medium-dark gray argillaceous-siliceous (> thorough global comparison in this study the upper limit is set at 3.9 μm.
50 % quartz) claystone, characterized by the presence of quartz- and Due to limitations in spatial resolution, only grains within the size
carbonate-rich lenses, where quartz grains are cemented by carbonate range of 30 to 300 μm were considered for thin section imaging and
(53 % clay minerals, 32 % quartz, 15 % carbonates). SF3 is mineralog­ computed tomography. The 2D images, obtained from thin sections,
ically similar to SF2 but includes a higher proportion of these quartz- have a lateral resolution of 5.17 μm, corresponding to an estimated
and carbonate-rich lenses (34 % clay minerals, 47 % quartz, 19 % car­ spatial resolution of approximately 10–15.5 μm, based on the conven­
bonates). SF6 is interpreted as a homogenized facies resulting from tion that 2–3 pixels are required to resolve a feature. For the CT scans,
mass-wasting processes (Zimmerli et al., 2024), representing a mixture voxel sizes range from 3.5 to 6.5 μm, yielding an estimated spatial res­
of SF1 and SF2 (58 % clay minerals, 30 % quartz, 12 % carbonates). olution of approximately 10.5 to 19.5 μm, following the established
Sedimentary structures such as ripples, discontinuous laminations, and guideline that spatial resolution is roughly three times the voxel
cross-bedding are observed within subfacies SF2 and SF3. Across all dimension. Taking a conservative approach, we have opted to add as
subfacies, quartz grains are consistently subangular in shape (Lauper lower limit for object recognition in the studied images, 30 μm. The
et al., 2018). Microphotograph pictures of the four subfacies are visible maximum grain size detected in the images has been identified at 300
in Fig. 1. Due to its well-characterized lithological variability and fine- μm. This grain size range (30–300 μm) was used for comparison of TS
grained nature, the OPA represents an ideal case for investigating and CT results with the results obtained from laser particle size analysis
grain size variations in fine-grained sediments. In particular, the selected (LPSA). It is acknowledged that this approach excludes the fine silt and
subfacies (SF1, SF2, SF3, SF6) are sufficiently fine-grained to serve as clay fractions. However, LPSA and traditional sieving and decantation
suitable samples for calibrating and testing the grain size parameter. techniques were additionally used to analyze the full grain size spectrum
Additional lithologies observed in the OPA include silty limestone, (0–300 μm) to provide a more comprehensive overview of the fine-
micritic/bioclastic limestone, and iron-rich limestone (after Zimmerli grained nature of the OPA.
et al., 2024; as early diagenetic features). These occur as thin beds,
typically less than 35 cm thick, and are not the focus of this study. 3.2.1. Thin sections
From the 17 representative samples evenly distributed throughout
3. Material and methods the STA2 core (refer to Fig. 1 for location), 17 uncovered 32 μm thin
sections (TS) were prepared at the University of Fribourg (UNIFR),
3.1. Reference core section Department of Geosciences, Switzerland. The TS were examined using
classical optical microscopy (transmitted light) to assess grain distri­
The study focuses on one reference core, the Stadel-2 (STA2) core, bution. Additionally, thin sections were scanned under plane polariza­
localized in northern Switzerland (Fig. 1). The reference core was drilled tion using a Nikon Eclipse Ni petrographic scanning microscope, which
between January and July 2021, achieving a total depth of 1288 m features an automated stage and a DS-Ri2 camera, both housed at
(Hinterholzer-Reisegger, 2022). The OPA is located between 800 and UNIFR. For each thin section, 40 individual images were acquired using
905 m below the surface, with a total thickness of 105 m (Jordan et al., the 2×/0.06 objective with an image resolution of 3x8bit 1636 × 1088
2022; Wohlwend et al., 2022). The lower section of the core is pre­ with 5 % overlap between individual images. Individual images were
dominantly composed of SF1, with minor contributions of SF6. The subsequently stitched together creating one complete mosaic picture of
middle section is mainly dominated by SF2, with minor occurrences of the thin section.
SF1 and SF6. The upper part of the core features a mix of SF1, SF2, minor Grains were segmented from the thin section mosaic pictures (2D –
SF3, and several calcareous beds (Zimmerli et al., 2024; Fig. 1). RGB color images), using the software XamFlow (Lucid; workflow-based
In total, 17 samples were considered for further analysis representing image analysis software; version 1.9.7.0). This software has primarily
the fine-grained fractions in the OPA; SF1 (8 samples), SF2 (5 samples), been utilized in biological and medical fields (e.g., Schmidt et al., 2022;
SF3 (2 samples), and SF6 (2 samples). Samples are evenly distributed Chapman et al., 2025; Lin et al., 2025), with no application in geological
throughout the core section with regular sampling intervals ranging research to date. A key advantage of the software is its capability for
between 5 and 11 m (Fig. 1, Appendix A; online accessible). batch processing, enabling the simultaneous analysis of multiple similar
samples, thereby increasing efficiency and ensuring consistency across
3.2. Grain size analyses datasets. An automatic workflow was developed and applied to all 17
samples using batch processing, consisting of the following steps. First,
Absolute grain size analyses were performed using laser particle size the imported image was cropped to the defined area of interest, in case
analyses, traditional wet sieving and decantation techniques, thin sec­ of the thin sections being 15 mm × 20 mm. The 8-bit RGB images were
tion analyses and CT-scan analyses. The detrital fraction is critical to converted to HSV (Hue, Saturation, Brightness) color space to facilitate
understanding grain size analyses in terms of depositional environments grain segmentation. The brightness component (Value) was utilized for
and rock properties. Previous research has identified clay minerals, threshold-based segmentation. A lower threshold of 70 % and an upper
along with quartz, as the primary detrital constituents of the OPA (see threshold of 100 % brightness were applied to isolate the grains. These
Kneuker and Furche, 2021; Lauper et al., 2021). Calcareous biogenic threshold values were determined based on expert assessment. Label
grains are present within the OPA. However, their nature is ambiguous, analysis (2D labeling) and grain diameter measurements (Equivalent
and it is not sure to which extent the bioclasts are transported and – as Diameter (EqDiameter) were subsequently added to the workflow. The
such – contribute to the detrital fraction. As such, we have decided to labeling was performed using Connected Component Analysis, where all

4
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

segmented voxels connected through at least one of the six cubic sur­ this study but are considered for follow-up studies. Bioclasts and bio­
faces are assigned the same label. The definition used for EqDiameter clastic debris were segmented together but excluded from the quanti­
was squareroot (4 * Area / π). In the TS images, the 2D nature and high tative analysis using the shape factor calculated in XamFlow (same
contrast generally allowed clear identification of grain boundaries approach as for TS). The segmented grains were exported to quantify the
(mainly quartz grains), so aggregation effects were minimal. Bioclasts grain size distribution. The D50 (median grain size) was calculated for
and bioclastic debris, which share similar brightness values with grains, grains within the size range of 30 to 300 μm. Segmented grains were
were segmented together but excluded from the quantitative analysis visualized for each sample using the same software. The representative
using the shape factor calculated in XamFlow. Bioclasts exhibited less volume was identified for each sample with respect to the median grain
rounding compared to the subangular detrital grains in the OPA, with a size (30–300 μm). Analyses show that more than 220 grains are neces­
roundness value of 0 assigned to perfectly spherical grains, allowing for sary to be representative for the respective analytical volume. Between
a comparative analysis of these values (values between 0 and 10 indicate 240 and 20200 grains for CT data were analyzed per sample. This crit­
rounded to slightly irregular grains). This approach ensures consistency ical number has been achieved for all studied volumes of interest (cfr.
and facilitates direct comparison across methods. As a last step, the See Appendix B).
segmented grain images and the resultant data set were exported. The
D50 (median grain size for the detrital fraction between 30 and 300 μm) 3.2.3. Laser particle size analyses
was then calculated for all 17 samples. It should be noted that the All 17 reference samples (see Fig. 1) were prepared for laser particle
Representative Sample (RS) has been determined for each thin section size analyses (LPSA). Small sample cubes (0.5 cm3) were manually
(Appendix B), evidencing that counting >220 grains is necessary to be disaggregated using a pestle and mortar. They were then placed in
representative for the median grain size between 30 and 300 μm. Be­ larger, heat-resistant containers. To remove organic matter, 35 %
tween 630 and 20470 grains for TS were analyzed per sample. This hydrogen peroxide (H2O2) was added. Following this treatment, the
threshold has been reached for each thin section studied (Appendix B). samples were washed and covered with demineralized water. Subse­
quently, 10 % hydrochloric acid (HCl) was added in small increments
3.2.2. X-ray micro-computed tomography until the reaction ceased, indicating the complete dissolution of car­
Quantitative grain size analyses were conducted using 3D X-ray bonates. During the next step, samples were briefly heated to 80–90 ◦ C
micro-Computer Tomography (X-ray CT, non-destructive 3D imaging; e. to facilitate the removal of magnesium and iron-rich carbonate phases.
g., Cnudde et al., 2011; Evans et al., 2015; Safari et al., 2021) on the After cooling, the samples were washed again with demineralized water.
same 17 sample locations referenced above (see Fig. 1). At each sam­ For the final step, all samples were treated with 2 % sodium poly­
pling location, a core segment measuring 0.5 cm × 0.5 cm × 2 cm was metaphosphate and boiled to ensure complete particle separation.
extracted for analysis. Scanning was performed with the Bruker Skyscan The grain size distribution of the prepared bulk samples was
2211 micro-CT scanner, utilizing the SkyScan2211 Control Software measured using the Malvern Mastersizer 3000 laser particle size
version 2.3.0 (CT Image Lab, UNIFR). The instrument employs a tung­ analyzer, connected to a Hydro 3000S module (University of Lausanne,
sten X-ray source. The samples were imaged using a flat panel detector, Institute of Earth Sciences, Switzerland). The analyzer operated within a
with the X-ray source operating at 110 kV and a focus current of 90 μA. particle size range of 0.01 to 2000 μm, with measurements conducted
Scans were performed with a rotation step size of 0.2 degrees, and a 0.5 under standard conditions for particle sizing in aqueous suspensions.
mm copper filter was used to enhance contrast among the main minerals Particle size categories were established according to the Wentworth
present in the OPA, namely quartz, calcite, and clay minerals, each scale (Wentworth, 1922), resp. clay (particles smaller than 3.9 μm), silt
representing similar densities and as such attenuation coefficients. (particles from 3.9 to 63 μm), and sand (particles from 63 to 2000 μm).
Image reconstruction was performed using NRecon software version Grain size data are often questioned as stand-alone technique certainly
1.6.10.5, with post-alignment correction, beam hardening correction, in silt- and clay-sized fraction, especially due to the effect of flocculation
ring artifacts reduction, and smoothing. The final reconstructed CT of clays in the 10–20 μm fractions. The dissolution of organic matter is
datasets are 16-bit images with a spatial resolution of approximately an important step in reducing coagulation and flocculation. It should be
10.5 to 19.5 μm. noted that the very small fractions (0.01 to 2 μm) are often unstable and
The data were further analyzed using the XamFlow software (version less reliable using laser particle size analyses. However, all fractions >2
1.9.7.0). Each sample was cropped to define the volume of interest (4 μm are reliable.
mm × 4 mm × 6 mm) and subsequently filtered using a non-local means To allow comparison with the other methods, both fractions, resp. 30
filter (k-nearest neighbors (knn) mode). Grains were segmented by to 300 μm for comparative analyses and 0 to 300 μm for understanding
employing single thresholding techniques. The 16-bit CT images consist the grain size spectrum including the fine-grained fractions, were used
of 65′535 grayscale values per image, with each value representing the for further analysis.
degree of X-ray attenuation, which is directly related to the material’s
density and composition. The grayscale intensity at each voxel corre­ 3.2.4. Wet sieving and decantation techniques
sponds to the linear attenuation coefficient, providing a basis for ma­ Wet sieving and decantation techniques were applied for the deter­
terial differentiation. For the segmentation of the grains, the global mination of the main fractions (clay, silt, and sand). 17 disaggregated
threshold was defined using the linear attenuation coefficient, where the and pretreated samples (see protocol of sample treatment for LPSA)
lower threshold was set at 0.016 and the upper threshold at 0.030 1/ were used for sieving and decantation. Clay separation was performed
mm. Label analysis (3D labeling) was conducted, including measure­ using the Atterberg extraction method, which relies on sedimentation
ments of the equivalent diameter (EqDiameter) for each segmented and based on Stokes’Law to isolate particles smaller than 3.9 μm. Initially, a
labeled grain. The labeling was performed using Connected Component 63 μm sieve was used to separate the sand fraction (wet sieving). The
Analysis, where all segmented voxels connected through at least one of settling velocity of particles was calculated using Stokes’Law ϑ = (2/9)
the six cubic surfaces are assigned the same label. The definition used for × (ρs -ρf) × g × r2 / η, where ρs is the particle density (~2.65–2.75 g/cm3
EqDiameter was cuberoot (6 * Volume / π). We acknowledge that for quartz, mica), ρf is the fluid density (~1.00 g/cm3 for water), g is
Connected Component Analysis alone cannot always prevent grain ag­ gravitational acceleration (9.81 m/s2), r is the particle radius (1.95 μm
gregation. The authors are aware of this and separate grain algorithms for 3.9 μm diameter), and η is water viscosity (0.001 Pa⋅s at 25 ◦ C). A 10
have been applied but did not result in satisfactory results. More cm water column was used for sedimentation, and the time required for
advanced techniques, including machine learning – based on segmen­ silt particles to settle through this column was calculated using time =
tation, can address this issue but require extensive parameter tuning, distance/velocity. After approximately two hours, the silt had settled
training data preparation, and validation, which are beyond the scope of

5
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

while clay remained in suspension, allowing the supernatant to be measurements were in strong agreement with the data acquired using
carefully removed. This process was repeated four times, with resus­ the handheld XRF core scanner. The XRF core-logging data points closest
pension between each cycle, to ensure complete removal of the clay in depth to the discrete XRF samples were compared, revealing a strong
fraction. The separated sediments (clay, silt, sand) were then dried. The positive correlation between the two datasets. This correspondence
dried main fractions (clay, silt, and sand) were measured and expressed confirms the reliability of the handheld XRF dataset – comprising
as percentages of the corresponding grain size classes. Due to the floc­ approximately 1600 data points over a 105 m interval (depth: 800–905
culation behavior of fine-grained sediments, only the major fractions m) – for capturing the downcore variability of these geochemical proxies
were considered collectively. across the entire sedimentary sequence.

3.3. Mineralogy (X-ray diffraction) 3.5. Statistical analyses

Individual silt and sand fractions from all 17 disaggregated samples The different grain size analyses and the elemental content were
(see LPSA protocol) were subsequently analyzed for their mineralogy compared using the Pearson’s correlation coefficient and corresponding
using X-ray diffraction (XRD), based on the previously dried material. P-values. The XRF-derived elemental ratios of the 17 samples (Si/Al, Ti/
The samples were analyzed using the Rigaku Ultima IV X-ray diffrac­ Al, Zr/Al, and Zr/Rb) were plotted against the calculated D50 values
tometer at the University of Fribourg (Department of Geosciences, obtained from the various grain size analysis methods – TS: 30–300 μm;
Switzerland). This system is equipped with a D/teX detector and a Cu X- CT: 30–300 μm; LPSA: 30–300 μm; and LPSA: 0–300 μm – using x–y
ray tube (Cu-Kα radiation), operated at 40 kV and 40 mA. X-ray scatter plots. The same approach was applied to compare the D50 values
diffraction patterns were recorded from 5◦ to 65◦ 2θ, with a scan rate of across the various grain size measurement techniques. The Pearson’s R-
1◦ /min and a step size of 0.02◦ /step. To minimize spectral interference value ranges from − 1 to 1, with values closer to 1 indicating a strong
and enhance signal quality, a DHL10 mm filter and a BB filter were positive correlation, values closer to − 1 indicating a negative correla­
positioned between the X-ray source and the sample. Peak fitting, phase tion, and values near 0 suggesting no correlation. The P-values provide a
identification, and Rietveld refinement were performed using the measure of statistical significance: P-values <0.05 denote statistically
Rigaku PDXL 2 software and the ICDD database (International Centre for significant correlations, P-values <0.01 indicate very significant corre­
Diffraction Data). lations, and P-values <0.001 denote extremely significant correlations.

3.4. Elemental content 4. Results

3.4.1. Discrete samples 4.1. Grain size distribution

X-ray fluorescence (XRF) data were acquired at UNIFR, using
wavelength dispersive X-ray fluorescence spectrometry (WD-XRF) with The original and segmented TS images for representative subfacies
a Zetium X-ray fluorescence spectrometer (Malvern PANalytical, (SF1, SF2, SF3, SF6), original and segmented CT scan images for
Netherlands). Prior to measurement, 17 samples (sample volume: 0.5 representative subfacies with frequency and cumulative distribution
cm3) from the same reference samples as previous analyses (see Fig. 1) curves, as well as frequency and cumulative distribution curves from
were crushed using a mortar. Elements (major elements: Al, Si, Ca, Fe, LPSA data for all the samples and all studied grain size fractions, are
Mg, Ti, Na, K, P; trace elements: Ba, Cr, Cu, Mn, Nb, Ni, Pb, Rb, Sr, V, Y, presented in Fig. 2 (TS and CT) and Fig. 3 (LPSA). The complete set of
Zn, Zr) were analyzed from produced fused glass beads. Sample prepa­ images from all 17 studied samples is provided in Appendix C for the TS
ration involved 0.7 g of homogeneous whole-rock powder, which has and Appendix D for the CT scans. Additionally, Fig. 2 shows zoomed-in
been calcinated at 1050 ◦ C and grounded using a tungsten carbide TS and CT images visualizing the segmented grains.
oscillatory grinder. This powder was then fused with 0.35 g of lithium For TS, the D50 range from 39 to 63 μm for the considered fractions,
fluoride (LiF) and 6.65 g of lithium tetraborate (Li2B4O7) as flux. The indicating that the samples predominantly represent the coarse silt
uncertainty for major elements is about 1 % and for trace elements 10 % fraction. For the CT scans, the D50 values range from 54 and 69 μm for
determined using the square root method. The detection limit for major the considered fractions, indicating a dominant coarse silt to very fine
elements is from 100 to 1000 ppm and for trace elements from 1 to 10 sand fraction (Figs. 2 and 4).
ppm. In the context of this study, the elemental ratios Si/Al, Ti/Al, Zr/ LPSA allow to consider the full grain size fraction, including the
Al, and Zr/Rb are employed for further analysis, as they are known to be fractions smaller than 30 μm (Fig. 3). The median grain size (D50) was
sensitive to grain size variations (e.g., Ohta, 2004; Calvert and Pedersen, calculated for both the 0–300 μm and 30–300 μm size fractions to enable
2007; Bloemsma et al., 2012; Liang et al., 2013; Van der Schee et al., comparison across methods. The results showed D50 values between 39
2016; Wu et al., 2020). and 56 μm for the 30–300 μm fraction, and between 6 and 15 μm for the
0–300 μm fraction (Fig. 4). These findings suggest that the D50 values for
3.4.2. XRF core logging all samples predominantly represent the silt fraction.
Non-destructive XRF core logging was conducted on the STA2 drill Overall, grain size analysis demonstrates that the facies is predomi­
core at the former core facility in Würenlingen (Switzerland). Scanning nantly characterized by the silt-sized fraction, comprising approxi­
was performed with an Olympus Vanta M handheld XRF analyzer (e.g., mately 19–32 % clay fraction, 62–71 % silt fraction, and 4–16 % sand
Young et al., 2016), mounted on a Multi-Sensor Core Logger (MSCL-S), fraction. Additionally, the results obtained from the sieving and
achieving a logging resolution of 5 cm (Geotek Ltd., UK). The instrument decantation technique exhibit similar trends, with the silt-sized fractions
was calibrated prior to each set of measurements. Analysis utilized both being dominant. The samples comprise approximately 20–30 % clay,
tungsten and rhodium X-ray sources, operating at 40 kV and 10 kV, 64–71 % silt, and 4–15 % sand, as illustrated in Fig. 4 (with clay, silt, and
respectively, with an acquisition count of 20 s. Major elements (Si, Al, sand represented in red, orange and yellow, respectively). These find­
Ca, Ti, Mg, Fe, K, P, S) and trace elements (V, Cr, Mn, Co, Ni, Cu, Zn, As, ings are consistent with the detailed frequency and cumulative distri­
Se, Rb, Sr, Y, Zr, Nb, Mo, Ag, Cd, Sn, Sb, Ba, W, Hg, Pb, Bi, Th, U) were bution curves, which show silt as the dominant fraction in all subfacies
measured with detection limits and uncertainties of 5 to 200 ppm and 2 (Figs. 2 and 3). It should be noted that an offset is apparent between the
% for major elements and 1 to 5 ppm and 4 % for trace elements. TS, CT, and LPSA grain size datasets (frequency and cumulative distri­
In this study, particular emphasis was placed on the elemental ratios bution curves; Figs. 2 and 3). These discrepancies primarily reflect
Si/Al, Ti/Al, Zr/Al, and Zr/Rb, consistent with those analyzed in the XRF methodological constraints and mineralogical effects: (1) phyllosili­
discrete samples. The results obtained from the discrete XRF cates, particularly micas, are not consistently segmented in TS and CT

6
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 2. Grain size representations from TS and CT analyses are illustrated using representative samples from SF1 (sample 16: 895.26 m, red), SF2 (sample 11: 859.51
m, orange), SF3 (sample 2: 806.13 m, green), and SF6 (sample 8: 841.98 m, blue). The TS images are displayed alongside segmented grains highlighted in red, as well
as the final labeled grains color-coded by size: red/orange/yellow for larger grains and green/blue for smaller grains. Bioclasts and bioclastic debris are excluded
from the quantitative analysis. Additionally, the original 2D CT slice, the segmented grains in red, and the 3D visualization of the segmented grains, color-coded by
size, are shown (using the same color code as above). Furthermore, the right side of the TS and CT panels displays the frequency and cumulative distribution curves
for SF1, SF2, SF3, and SF6 on a logarithmic x-axis. At the top of the figure, magnified views from sample 1 (801.03 m; SF3) illustrate the quartz-rich central part of the
section that was segmented.

image analyses, (2) carbonate cement are more readily distinguished in 4.2. Mineralogy
TS than in CT, and (3) the fine fraction (<30 μm) is more comprehen­
sively captured by LPSA. Nevertheless, the overall correlation between The overall mineralogical composition of the 17 analyzed samples,
the methods remains robust (Fig. 4). based on Zwahlen et al. (2022) and Mazurek et al. (2023), shows that

7
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 3. Laser particle size analysis results (all samples, with details provided in the figure legend for individual sample depths and sample numbers) are presented as
frequency distribution and cumulative distribution curves for SF1, SF2, SF3 and SF6 with a logarithmic x-axis (full grain size spectrum). The boundaries of the clay,
silt, and sand fractions are indicated by vertical gray lines. The silt fraction is the dominant fraction in all subfacies.

8
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 4. The calculated D50 values for resp. TS (30–300 μm), CT (30–300 μm) and LPSA (30–300 μm and 0–300 μm) together with the lithology for core STA2. The
legend for the litholog of STA2 is the same as for Fig. 1. Additionally, the determined percentages of clay, silt, and sand fractions for the LPSA and sieving/
decantation are shown. The facies are predominantly characterized by the silt-sized fraction, comprising 19–32 % (LPSA) and 20–30 % (sieving/decantation) clay
fraction, 62–71 % (LPSA) and 64–71 % (sieving/decantation) silt fraction, and 4–16 % (LPSA) and 4–15 % (sieving/decantation) sand fraction.

clay mineral content ranges from 26 % to 69 %, while quartz content increasing quartz and decreasing white mica content is observed along
varies between 12 % and 39 % (Fig. 5). It should be noted that clay- the facies succession from SF1 to SF6, SF2, and SF3 (Fig. 5).
mineral content was not measured directly but was calculated by dif­ In the silt fraction, the primary constituents include quartz, white
ference to 100 %, based on the quantified contents of non-clay minerals mica, feldspar, and clay minerals. The presence of quartz, white mica,
(Zwahlen et al., 2022; see Waber, 2020 for preparation and measure­ feldspar and clay minerals has been also observed by optical micro­
ment protocols). The reported total clay-mineral content in the bulk scopy. Unlike the sand fraction, the silt fraction also contains a notable
fraction thus represents the remainder after subtracting these identified proportion of clay minerals. Separate clay mineralogical analyses have
mineral phases from the total composition. An inverse relationship is been performed by Zwahlen et al. (2022) and Mazurek et al. (2023) on
evident, with higher quartz content – particularly in the uppermost SF3 oriented samples prepared by ultrasonic dispersion, chemical treatment,
samples – corresponding to lower clay mineral content, and vice versa. and conditioning as air-dried, ethylene glycol–saturated, and heated
This reflects a general trend of increasing quartz and decreasing clay specimens (Waber, 2020). The main clay phases quantified include illite,
minerals towards the top of the core. smectite, chlorite, kaolinite, illite–smectite mixed layers, and chlor­
In the sand fraction, quartz is the dominant mineral phase, ite–smectite mixed layers. However, quartz and white mica remain the
comprising between 35 % and 96 % of the assemblage, followed by dominant phases in the silt fraction, with feldspar again representing a
white mica (2–57 %) and feldspar (1–10 %). The presence of quartz, minor component (1–10 %). Coarser subfacies, such as SF2 and SF3, are
mica (muscovite), and feldspar has been also confirmed by optical mi­ marked by higher quartz and lower white mica content, whereas finer
croscopy. The mineralogical characteristics differ between subfacies. subfacies, SF1 and SF6, exhibit elevated white mica and reduced quartz
SF1 samples are characterized by lower quartz content (average ~ 55 %) content. Clay mineral content in the silt fraction ranges from 0.5 % to 39
and higher white mica content (~ 40 %). In contrast, SF2 samples %, with lower values generally associated with SF2 and SF3, and higher
contain higher quartz (average ~ 73 %) and lower white mica (~ 23 %), content present in SF6 and SF1. An exception is sample 7 (classified as
while SF3 samples show the highest average quartz content (~ 87 %) SF2), which exhibits 27 % clay minerals. The three lowermost samples
and the lowest white mica (~ 10 %) (Fig. 5). SF6 displays variable from SF1 are the richest in clay minerals and contain the highest clay
mineralogical compositions depending on the stratigraphic position of content within the silt fraction (Fig. 5). A consistent trend across subf­
the sample within the core. One sample located in the middle of the core acies is observed: SF2 and SF3 exhibit higher average quartz content
and surrounded by SF2 exhibits a higher quartz content, whereas the (75–92 %) and lower content of phyllosilicates (white mica and clay
lower SF6 sample, adjacent to SF1, shows elevated white mica content. minerals, 8–26 %), while SF6 and SF1 are characterized by lower quartz
Overall, the mineralogical composition of SF6 falls between that of SF1 (60–69 %) and higher proportions of white mica and clay minerals
and SF2. Feldspar content remains relatively consistent across all subf­ (30–39 %) (Fig. 5).
acies, ranging between 1 % and 10 % (Fig. 5). A general trend of

9
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 5. Mineralogical data presented in this figure includes the total measured contents of clay minerals and quartz (Zwahlen et al., 2022; Mazurek et al., 2023), as
well as the quartz content in the sand and silt fractions, and white mica concentrations in both fractions. Specifically, the sand fraction includes white mica but lacks
clay minerals, whereas the silt fraction contains both white mica and clay minerals, which are grouped together both being phyllosilicates. All data are displayed
alongside the lithological profile of borehole STA2. In addition, the overall mineral composition of individual samples is shown using pie charts. At the bottom of the
figure, the average mineralogical compositions for each subfacies (SF1, SF2, SF3, and SF6) are illustrated using ternary diagrams: quartz – feldspar – white mica for
the sand fraction, and quartz – feldspar – white mica + clay minerals for the silt fraction. Both fractions exhibit comparable compositional trends across subfacies.
The mineralogical data were first averaged across the selected samples for each of the three main mineralogical groups – quartz, feldspar, and white mica/clay
minerals – and then normalized so that the total sums to 100 %.

10
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

4.3. Elemental content and grain size variability 0.782). P-values for all correlations involving LPSA and TS data were <
0.0001, while those for CT-based comparisons ranged from 0.00016 to
The Si/Al ratios of the 17 XRF-discrete samples range from 1.9 to 4.5 0.00071 (Fig. 7), still within the threshold for statistically significant
(Fig. 6), with the majority (15 samples) clustering between 2.0 and 2.7. results.
Two samples from the uppermost OPA (801.03 m and 806.13 m) exhibit
markedly elevated Si/Al ratios of ~4.5. Comparable patterns are 5. Discussion
observed in the Ti/Al, Zr/Al, and Zr/Rb ratios: Ti/Al values range from
0.050 to 0.080, with the highest values (~0.080) in the uppermost 5.1. Comparison of particle size distribution across different methods
samples and the remaining samples ranging between 0.050 and 0.065;
Zr/Al values span 0.0010 to 0.0055, with only the upper two samples Grain size analyses are consistent across different, independent
reaching 0.0055, while the others remain below 0.0030; Zr/Rb values analytical methods, and align well with subfacies variations (SF1, SF2,
range from 1.0 to 3.7, with the highest values (~3.7) again limited to the SF3, SF6). To quantitatively assess this internal consistency, statistical
uppermost samples, and values between 1.0 and 2.6 elsewhere. comparisons were conducted between the different analytical methods
Comparison of the discrete XRF data with the core-logging XRF applied. Grain size (D50) values derived from different analytical
dataset at the same depth intervals (Fig. 6) reveals a strong positive methods show a high degree of consistency, with the strongest corre­
correlation, confirming the reliability of the downcore XRF logging data lations observed between LPSA and TS data (R > 0.90). CT-derived
for characterizing geochemical variability. Downhole profiles of Si/Al, values exhibit slightly lower, yet still statistically significant, correla­
Ti/Al, Zr/Al, and Zr/Rb all exhibit similar trends (Fig. 6), with elevated tions (R ≈ 0.73–0.78). This reflects the higher complexity of CT-based
ratios in the upper portions of the core (predominantly associated with measurements in fine-grained sediments, where similar mineral den­
SF2 and SF3), and lower ratios in the lower sections (corresponding to sities and challenging segmentation conditions can limit precision,
SF1 and SF6). rather than indicating a fundamental mismatch in grain size trends.
Although the STA2 core spans 105 m, the discrete sampling (17 These results underscore the overall consistency between methods and
samples) may not fully capture the complete grain size heterogeneity. validate the integration of multiple analytical approaches for recon­
Nonetheless, the geochemical proxies (Si/Al, Ti/Al, Zr/Al, Zr/Rb) show structing grain size variability across the STA2 core. The high inter-
strong overall correlation with D50 values from all grain size analysis method consistency supports the robustness of the dataset and pro­
methods (Figs. 6 and 7). Among these, Si/Al displays the highest cor­ vides a strong foundation for future studies utilizing similar multi-
relation with D50 (Fig. 7) and was therefore used to estimate grain size method approaches. These results underline the reliability of LPSA-
variability. This comparison represents the core calibration of XRF- linked measurements as the most robust grain size indicators in this
derived geochemical ratios (Si/Al) against directly measured grain study.
size (D50) values. It demonstrates that grain size trends can be quanti­ While LPSA provides rapid, repeatable, cost-effective data covering
tatively estimated from XRF data. Grain size (D50) values were calcu­ the full grain size spectrum – particularly for fine-grained sediments, it
lated using the regression relationship derived from Si/Al ratios and offers only one single, absolute quantitative measurement without
laser particle size analysis (LPSA) data for the total fraction (0–300 μm), spatial or textural context. The method yields high-resolution grain size
based on 1600 XRF data points (Fig. 6). The resulting D50 values range distributions, but sample preparation is time-consuming, and technical
from 6 to 30 μm, with the majority between 6 and 15 μm. These values, problems such as flocculation or incomplete dispersion can affect ac­
along with coarsening- and fining-upward trends, are illustrated in curacy. Moreover, LPSA does not produce physically separated fractions
Fig. 6. in case further advanced analysis of single fractions are critical. In
contrast, the sieving and decantation method also allows analysis of the
4.4. Statistical analysis complete grain size range, is cost-effective, requires minimal equipment,
and produces physically separated fractions suitable for subsequent
In this study, statistical comparison using Pearson’s correlation co­ mineralogical or geochemical analyses. Nevertheless, it is time-
efficients (R-values) and P-values helped to evaluate the consistency and consuming and labor-intensive, particularly due to the need for
reliability across methods. The strongest correlations in D50 values were repeated resuspension and decantation steps. Besides, flocculation of
observed between TS and LPSA data for the 30–300 μm range (R = fine particles may still occur, complicating the separation steps between
0.926), followed by LPSA total fraction and LPSA 30–300 μm (R = clay and silt. The method is also operator-dependent and may have
0.908), and TS with LPSA total fraction (R = 0.822), with all P-values lower reproducibility compared to automated techniques. Thin section
<0.0001, indicating extremely significant relationships (Fig. 7). analysis does not capture the full grain size spectrum due to spatial
Although slightly lower, CT scan data also showed statistically signifi­ resolution limitations but provides valuable complementary informa­
cant correlations with other methods (R-values between 0.734 and tion on grain texture, including grain shape, fabric, and mineralogical
0.778), with P-values ranging from 0.00023 to 0.00079 (Fig. 7). composition. The ability with TS data to reveal the internal structure,
Additionally, the calculated D50 values from all analytical techniques mineralogical composition, and texture of the grains offers important
(Figs. 4 and 6), including both XRF discrete samples and XRF core insights for understanding the relationship between particle shape and
scanning data, exhibit strong linear correlations with the geochemical depositional environment (Fig. 1 for TS pictures; see also Houghton
ratios Si/Al, Ti/Al, Zr/Al, and Zr/Rb (Fig. 6). Among the methods, the et al., 2024b for further information on TS grain size analysis). On the
highest correlation coefficients (R-values) with Si/Al were consistently other hand, it is important to mention that TS do not capture the com­
observed for the LPSA, with R = 0.955 for the 30–300 μm fraction and R plete morphology of grains and strongly depends on the cutting location
= 0.941 for the 0–300 μm fraction (Fig. 7). Slightly lower, yet statisti­ and orientation randomly slicing grains. The full potential of 3D CT-
cally robust, correlations were obtained for TS (R = 0.883) and CT (R = scanning could expand on this by providing comprehensive insights
0.800). The associated P-values for all methods were < 0.0001, with the into grain shapes, orientations, and potentially grain size variations
exception of the CT data (P = 0.00012) (Fig. 7), still indicating a sta­ within specific sedimentary structures (Houghton et al., 2024b). How­
tistically very significant relationship. Similar trends were observed for ever, it remains limited in its ability to resolve the full grain size dis­
the other geochemical proxies. For Ti/Al, Zr/Al, and Zr/Rb, the stron­ tribution, particularly at the fine-grained end of the spectrum. The cost-
gest correlations were again recorded with LPSA-derived D50 values (Ti/ effectiveness of 3D CT-scanning is limited due to high operational ex­
Al: R = 0.904–0.918; Zr/Al: R = 0.931–0.958; Zr/Rb: R = 0.915–0.947). penses and accessibility constraints, which can hinder its routine
These were followed by TS (Ti/Al: R = 0.812; Zr/Al: R = 0.904; Zr/Rb: application despite its valuable capabilities. Thus, while the superior
R = 0.891) and CT data (Ti/Al: R = 0.739; Zr/Al: R = 0.790; Zr/Rb: R = resolution of LPSA remains the cornerstone of our analysis, integrating

11
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 6. The D50 values from TS, CT and LPSA (30–300 μm and 0–300 μm fraction) are represented in black, gray, light brown and dark brown, respectively. These
data are presented alongside elemental ratios (Si/Al, Ti/Al, Zr/Al, and Zr/Rb) derived from discrete XRF samples and the lithological profile of the core STA2,
following the legend provided in Fig. 1. Elemental ratios are plotted using a 1 m moving average (black line). Scatter plots compare discrete XRF measurements with
the nearest core-logging data points, demonstrating strong linear correlations (high R-values). Additionally, D50 values for the total grain size fraction (0–300 μm),
inferred from a linear regression between Si/Al and D50 (based on LPSA data; see Fig. 7), are included, along with the corresponding 1 m moving average (black line).
The Si/Al ratio was selected due to its highest correlation coefficient with the D50 values. Coarsening- and fining-upward trends are indicated by blue and violet
arrows, respectively.

12
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Fig. 7. x-y plots depicting (1) the D50 values obtained from TS, CT, and LPSA (for both the 30–300 μm and 0–300 μm fractions) presented against Si/Al, Ti/Al, Zr/Al
and Zr/Rb ratios and (2) the comparison between different grain size methodologies. The data for all measurement techniques exhibit similar trends, with strong
correlations to the geochemical ratios, as demonstrated by high Pearson’s correlation coefficient (R-values) and corresponding P-values (first 16 cross-plots and left
table below). For all grain size analytical methods (comparison, 6 cross plots and right table), the observed R-values are indicative of strong positive correlations,
with the associated P-values confirming statistically very to extremely significant relationships.

13
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

TS, CT, and sieving/decantation techniques offers important additional along with clay minerals in the finer size classes, all of which are clas­
insights in grain shape, orientation, texture, and sedimentary structures. sified as phyllosilicates. Feldspar plays a minor role. Geochemical ratios
Last are critical in the understanding of fine-grained rock properties and are primarily governed by the relative abundances of quartz and phyl­
the interpretation of depositional environments. It should be noted that losilicates. In finer-grained subfacies such as SF1, elevated phyllosilicate
TS and CT analyses could be complemented by the acquisition of higher content can result in a decrease in the elemental ratios, as phyllosilicates
resolution datasets (f.e. through the use of FIB-SEM and/or SEM). This are enriched in Al and Rb. Conversely, in coarser-grained subfacies such
would provide a more comprehensive understanding of the full grain as SF2 and SF3, decreased phyllosilicate content leads to higher ratio
size spectrum and allow upscaling from finer-scale textural variations to values. Thus, variations in phyllosilicate and quartz content exert the
larger-scale sedimentary processes. primary control on the observed geochemical ratios, highlighting their
The methodological framework applied in this study aligns closely mineralogical dependence and supporting their utility as grain size
with the comparative approach used by Houghton et al. (2024b), who proxies in this context (e.g., Calvert and Pedersen, 2007; Craigie and
evaluated particle size distributions in artificial samples composed of Rees, 2016).
spherical silica particles using LPSA, optical point counting, 2D image Once calibrated against absolute grain size measurements,
analysis, and CT data. Their findings indicated that CT provided the geochemical proxies are very well suited as grain size proxies for grain
most accurate grain size distribution for coarse, well-defined particles – size estimation in fine-grained sediments. This approach offers several
especially due to its ability to resolve grain morphology and shape in 3D. advantages: it circumvents the challenges associated with physical
They also observed that LPSA tends to overestimate coarse grain abun­ sample preparation and disaggregation, provides continuous high-
dance. However, the authors did not study fine-grained sediments and as resolution data from core-scanning or dense spot analyses, and is espe­
such results cannot be extrapolated to those types of sediments. This cially suited for lithologies where traditional methods like LPSA are
study supports the continued use of LPSA as a reference for grain size time-consuming or sample-destructive. While the reliability of such
calibration in fine-grained sedimentary successions. Findings under­ proxies may vary with mineralogical composition, diagenetic overprint,
score the importance of appropriate method selection which is depen­ or depositional context, these limitations can be addressed through site-
dent upon lithology and grain size range. While for coarser grain sizes specific calibration. Thus, in settings where fine-grained sediment
(medium sand, coarse sand and gravel), CT-scanning techniques are dominates, calibrated XRF-derived proxies offer a practical and scalable
more appropriate, LPSA is the preferred method for fine-grained sedi­ alternative to conventional granulometric techniques, supporting
ments, and especially for the clay, silt and fine sand fractions. However, detailed stratigraphic and paleoenvironmental interpretations.
previous studies have highlighted challenges in accurately resolving
grain size in such fine-grained materials due to e.g. flocculation of clay 5.3. Calibrated grain size parameter in the OPA reveals coarsening- and
minerals, or cementation (e.g., Cui et al., 2023; Houghton et al., 2024a). fining-upward trends
These issues can significantly affect the reliability of results if not
properly addressed. Nonetheless, such limitations may be mitigated Grain size characteristics of the OPA correlate closely with subfacies
through rigorous sample preparation protocols, including effective variations and are relatively homogeneous within individual subfacies.
disaggregation and dispersion procedures tailored to fine-grained sedi­ The grain size distribution indicates that the sediment is predominantly
mentary rocks. This reinforces the need for methodologically robust and silty (see also Zimmerli et al., 2025). In this study, grain size analysis was
lithology-sensitive analytical strategies when characterizing grain size applied to the OPA, a formation in which grain size has long been
distributions in complex sedimentary systems. debated, and where previous results have shown considerable vari­
ability due to the absence of a subfacies-based classification approach (e.
5.2. Grain size data versus elemental content and mineralogy g., Nagra, 2014b; Minardi et al., 2016; Ferrari et al., 2012, 2020;
Kneuker et al., 2023; Appendix E). The methods applied in Nagra
The consistently strong correlations between the D50 values from all (2014b), Minardi et al. (2016), and Ferrari et al. (2012, 2020) involved
analytical techniques with the geochemical ratios (XRF) confirm the wet sieving for particles >0.075 mm and sedimentation techniques for
suitability of Si/Al, Ti/Al, Zr/Al, and Zr/Rb ratios as reliable finer fractions, while Kneuker et al. (2023) utilized a SediGraph particle
geochemical proxies for assessing grain size variability in the OPA. This analyzer. In all cases, carbonate phases were not dissolved prior to
is in line with prior studies that have employed similar proxies for grain analysis. The discrepancies in the results can be attributed to method­
size estimation in a variety of sedimentary contexts (e.g., Ohta, 2004; ological differences, such as incomplete removal of carbonate and/or
Calvert and Pedersen, 2007; Bloemsma et al., 2012; Liang et al., 2013; organic matter. This can lead to flocculation or the formation of clotted
Craigie and Rees, 2016; Van der Schee et al., 2016; Wu et al., 2020). grains, likely due to the presence of organic matter (OM), which can
However, it should be noted that many of these studies, except promote the aggregation of fine particles through electrostatic and
Bloemsma et al. (2012) and Wu et al. (2020), apply these proxies chemical interactions. The internal consistency observed in this study
without site-specific calibration, which may introduce some un­ highlights the importance of systematic sampling, rigorous sample
certainties. Where possible, site-specific calibration enhances the accu­ preparation protocols, and multi-method calibration for accurately
racy and reliability of geochemical grain size proxies. characterizing grain size variability.
It is crucial to critically assess the applicability of Si/Al, Ti/Al, Zr/Al, Distinct vertical trends in grain size were identified within the OPA.
and Zr/Rb ratios as proxies for grain size variations across different Grain size analyses resulted in the identification of different coarsening-
geological settings. While these geochemical ratios have demonstrated and fining-upward trends (Fig. 6, Zimmerli et al., 2025). Eleven coars­
strong correlations with measured grain size distributions in the present ening- and fining-upward trends are observed, with a notable increase in
study, their reliability is context-dependent and may vary significantly the frequency of these trends in the upper section of the STA2 core. This
with changes in mineralogical composition, diagenetic overprint, or can be explained by higher lithological variability in the upper part of
depositional environment. Therefore, the use of these proxies should not the core compared to the lower section, characterized by the coexistence
be generalized without prior validation. It is recommended that their of SF1, SF2, and SF3. However, it should be noted that the precise
application be calibrated and cross-validated against absolute grain size boundaries of these coarsening- and fining-upward trends are interpre­
measurements in each specific paleoenvironmental setting to ensure tative, as smaller-scale variability and the distribution of coarse-grained
interpretive accuracy. facies may influence the apparent trend lines. Zimmerli et al. (2024)
The ratios Si/Al, Ti/Al, Zr/Al, and Zr/Rb are consistent with the previously described coarsening-upward trends within the STA2 core,
measured mineralogical composition of the analyzed samples. The focusing exclusively on coarsening-upward trends and omitting any
dominant components of the silt and sand fractions are quartz and micas discussion on fining-upward trends. They identified two coarsening-

14
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

upward trends in the lower section and five in the upper section, based subfacies (SF1, SF6) with increased phyllosilicate content show
on Si/Al ratios and macroscopic lithological observations. The seven lower ratios.
mapped coarsening-upward trends in the OPA in Zimmerli et al. (2024) • Eleven coarsening- and fining-upward grain size trends are identified
fit well with those presented here. A standard sequence stratigraphic within the Stadel-2 drill core, improving the understanding of hy­
approach was not applicable, and a direct correlation between Si/Al drodynamic conditions during the Opalinus Clay deposition.
ratios and grain size was not established in Zimmerli et al. (2024). This
study provides evidence supporting the use of Si/Al ratios, but addi­ Overall, these findings successfully allow better understanding of the
tionally also Ti/Al, Zr/Al, and Zr/Rb ratios, as a proxies for grain size initial depositional and hydrodynamic conditions during Opalinus Clay
variations in the OPA. formation and support improved predictions of its geotechnical and
The grain size data and the observed grain size distribution in the hydrological behavior relevant to radioactive waste disposal.
STA2 core establish an important foundation for investigating the
physical and dynamic transport processes of the fine-grained succession CRediT authorship contribution statement
of the OPA, including sediment provenance, current sorting, physical
reworking, flow strength, and a detailed characterization of the depo­ Géraldine Nicole Zimmerli: Writing – review & editing, Writing –
sitional environment. Due to the lack of consistent data on grain size, original draft, Visualization, Methodology, Investigation, Formal anal­
those processes have not been adequately studied in the past. The ysis, Data curation. Alicia Fantasia: Writing – review & editing, Vali­
coarsening- and fining-upward trends observed in the STA2 core reflect dation, Methodology, Investigation. Stephan Wohlwend: Writing –
depositional cycles controlled by bottom current dynamics, consistent review & editing, Visualization, Methodology, Investigation, Concep­
with contourite formation. These cycles are characterized by transitions tualization. Gaudenz Deplazes: Writing – review & editing, Validation,
between SF1, SF2, and SF3 facies, highlighting variations in current Supervision, Project administration, Methodology, Investigation,
strength and sediment supply that shape the observed grain size patterns Funding acquisition, Conceptualization, Visualization. Anneleen Fou­
(Zimmerli et al., 2025). bert: Writing – review & editing, Visualization, Validation, Supervision,
The observed coarsening- and fining-upward trends in grain size Project administration, Methodology, Investigation, Funding acquisi­
variability offer critical insights into both the dynamic depositional tion, Formal analysis, Conceptualization.
environment of the OPA and its associated rock properties. Detailed
grain size analysis remains essential for advancing our understanding of Funding sources
these properties, as grain size directly influences the physical, geo­
mechanical, and geochemical behavior of the formation. The predomi­ Financial support for this study is gratefully provided by Nagra and
nance of silt-sized fractions – mainly composed of detrital quartz and the University of Fribourg (Switzerland).
mica – has notable implications: quartz enhances the stiffness and
strength of the rock, while mica, due to its platy morphology and Declaration of competing interest
alignment under compaction, contributes to mechanical anisotropy and
can reduce shear strength (e.g., Crisci et al., 2021). These grain scale The authors declare that they have no known competing financial
mineralogical features, in combination with the grain size distribution, interests or personal relationships that could have appeared to influence
also affect porosity, permeability, and fracture behavior (e.g., Zheng the work reported in this paper.
et al., 2015; Verweij et al., 2016; Amann et al., 2017; Carcione et al.,
2019; Crisci et al., 2021). With a robust and calibrated grain size proxy Acknowledgements
now established, this study provides a solid foundation for improved
prediction and numerical modeling of rock mechanical and transport The authors would like to express their gratitude to Patrick Dietsche
properties within the OPA. (University of Fribourg, Switzerland) for the thin section preparation,
and to Christoph Neururer (University of Fribourg, Switzerland) for his
6. Conclusion support with computed tomography scans. Special thanks are extended
to Brahimsamba Bomou and Thierry Adatte (University of Lausanne,
This study highlights the importance of comparative, systematic, and Switzerland) for their help with the Malvern measurements (LPSA). We
quantitative grain size analyses in the fine-grained detrital fraction of also thank Alexandre Salzmann (University of Fribourg, Switzerland) for
the Opalinus Clay from northern Switzerland – Switzerland’s designated his assistance with X-ray diffraction and X-ray fluorescence measure­
host rock for radioactive waste disposal. The integration of TS, CT, LPSA, ments. Additionally, we are grateful to Peter Suter and Tor Hildebrand
and sieving/decantation analysis provides a robust, comparative for their support with the XamFlow software and the associated work­
framework for assessing grain size in fine-grained sediments. Compar­ flow. Our appreciation goes to Nathan Claret for conducting the initial
ative grain size analyses combined with mineralogical (XRD) and grain size tests of the Opalinus Clay for this Bachelor thesis at the Uni­
geochemical (XRF) analyses capture sediment heterogeneity and vari­ versity of Fribourg (Switzerland). Silvio Giger is thanked for the com­
ability of the Opalinus Clay. ments on an earlier version of the manuscript. Finally, two anonymous
The major outcome of this study can be summarized as follows: reviewers are acknowledged for their very important and helpful com­
ments on the manuscript, as well as the editor Michał Gradziński.
• LPSA provide the most robust grain size distribution, including Financial support for this study is gratefully provided by Nagra and the
particles finer than 30 μm. University of Fribourg (Switzerland).
• TS and CT offer valuable insights into grain morphology and internal
structures but limited spatial resolution exclude the finest fractions. Appendix A. Supplementary data
• Grain size measurements are consistent across methods, with median
grain sizes (D50) closely aligned. Supplementary data to this article can be found online at https://doi.
• Strong correlations between grain size and geochemical proxies (Si/ org/10.1016/j.sedgeo.2025.106982.
Al, Ti/Al, Zr/Al, Zr/Rb) evidence that elemental ratios can serve as
effective proxies for grain size variations and reflect mineralogical Data availability
control by quartz and phyllosilicates.
• Coarser subfacies (SF2, SF3) with higher quartz content correspond The dataset employed in this study is either incorporated within this
to elevated elemental ratios (Si/Al, Ti/Al, Zr/Al, and Zr/Rb); finer manuscript and its supplementary materials or can be requested from

15
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

the authors. Ferrari, A., Favero, V., Manca, D., Laloui, L., 2012. Geotechnical characterization of core
samples from the geothermal well Schlattingen SLA-1 by LMS/EPFL. In: Nagra
Arbeitsbericht NAB 12–50. Nagra, Wettingen, Switzerland (Available at nagra.ch).
References Ferrari, A., Rosone, M., Ziccarelli, M., Giger, S.B., 2020. The shear strength of Opalinus
Clay shale in the remoulded state. Geomechanics for Energy and the Environment
Ahmad, F., Quasim, M.A., Ghaznavi, A.A., Khan, Z., Ahmad, A.H.M., 2017. Depositional 21, 100142. https://doi.org/10.1016/j.gete.2019.100142.
environment of the Fort Member of the Jurassic Jaisalmer Formation (western Franz, M., Nitsch, E., 2009. Zur lithostratigraphischen Gliederung des Aalenium in
Rajasthan, India), as revealed from lithofacies and grain-size analysis. Geologica Baden-Württemberg. Landesamt für Geologie, Rohstoffe und Bergbau -
Acta 15 (3), 153–167. Informationen 22, 123–146 (in German).
Amann, F., Wild, K.M., Loew, S., Yong, S., Thoeny, R., Frank, E., 2017. Geomechanical Ghaznavi, A.A., Quasim, M.A., Ahmad, A.H.M., Ghosh, S.K., 2019. Granulometric and
behaviour of Opalinus Clay at multiple scales: results from Mont Terri rock facies analysis of Middle–Upper Jurassic rocks of Ler Dome, Kachchh, western India:
laboratory (Switzerland). Swiss Journal of Geosciences 110, 151–171. an attempt to reconstruct the depositional environment. Geologos 25 (1), 51–73.
Bitzer, K., 1996. Subseabed disposal of radioactive waste: effects of consolidational fluid Gómez, J.J., Canales, M.L., Ureta, S., Goy, A., 2009. Palaeoclimatic and biotic changes
flow. Environmental Geology 27, 300–308. during the Aalenian (Middle Jurassic) at the southern Laurasian Seaway
Bjerrum, C.J., Surlyk, F., Callomon, J.H., Slingerland, R.L., 2001. Numerical (Basque–Cantabrian Basin, northern Spain). Palaeogeography, Palaeoclimatology,
paleoceanographic study of the early Jurassic transcontinental Laurasian Seaway. Palaeoecology 275, 14–27.
Paleoceanography 16 (4), 390–404. Hinterholzer-Reisegger, P., 2022. TBO Stadel-2-1: Data Report, Dossier I: Drilling. In:
Bloemsma, M.R., Zabel, M., Stuut, J.B.W., Tjallingii, R., Collins, J.A., Weltje, G.J., 2012. Nagra Arbeitsbericht NAB 22–02. Nagra, Wettingen, Switzerland. Available at nagra.
Modelling the joint variability of grain size and chemical composition in sediments. ch.
Sedimentary Geology 280, 135–148. Hostettler, B., Reisdorf, A.G., Jaeggi, D., Deplazes, G., Bläsi, H.R., Morard, A., Feist-
Boggs, S., 2009. Petrology of Sedimentary Rocks, 2nd Edition. Cambridge University Burkhardt, S., Waltschew, A., Dietze, V., Menkveld-Gfeller, U., 2017. Litho- and
Press, Cambridge, p. 21. biostratigraphy of the Opalinus Clay and bounding formations in the Mont Terri rock
Braga, B.B., Costa, C.A.G., Lima, G.D., de Lacerda, C.F., Foerster, S., Brosinsky, A., laboratory (Switzerland). Swiss Journal of Geosciences 110, 23–37.
Medeiros, P.H.A., 2024. Reuse of sediment as a soil conditioner in a semiarid region Houghton, J.E., Nichols, T.E., Worden, R.H., 2024a. Simulating compaction and
dominated by subsistence farming: sediment characterization at the regional scale cementation of clay grain coated sands in a modern marginal marine sedimentary
and effects on maize crop. Journal of Soils and Sediments 24, 1039–1055. system. Geosciences 14, 268. https://doi.org/10.3390/geosciences14100268.
Burkhalter, R.M., 1996. Die Passwang-Alloformation (unteres Aalénien bis unteres Houghton, J.E., Behnsen, J., Duller, R.A., Nichols, T.E., Worden, R.H., 2024b. Particle
Bajocien) im zentralen und nördlichen Schweizer Jura. Eclogae Geologicae size analysis: a comparison of laboratory-based techniques and their application to
Helveticae 89 (3), 875–934 (in German). geoscience. Sedimentary Geology 464, 106607. https://doi.org/10.1016/j.
Burkhalter, R.M., Bläsi, H.R., Feist-Burkhardt, S., 1997. Der «Dogger ß» (oberes Aalénien) sedgeo.2024.106607.
in den Bohrungen Herdern-1, Berlingen-1 und Kreuzlingen-1 (Nordostschweiz) und Jordan, P., Schürch, P., Naef, H., Schwarz, M., Felber, R., Ibele, T., Weber, H.P., 2022.
seine Beziehung zu den gleichaltrigen Schichten im Nordjura. Eclogae Geologicae TBO Stadel-2-1: Data Report, Dossier III: Lithostratigraphy. In: Nagra Arbeitsbericht
Helvetiae 90, 269–291 (in German). NAB 22–02. Nagra, Wettingen, Switzerland (Available at nagra.ch).
Burland, J.B., 1990. On the compressibility and shear strength of natural clays. Kneuker, T., Furche, M., 2021. Capturing the structural and compositional variability of
Géotechnique 40 (3), 329–378. Opalinus Clay: constraints from multidisciplinary investigations of Mont Terri drill
Cai, Y., Yin, Y., Kuang, X., Hao, Y., Liu, J., Zheng, C., 2023. Review: specific storage in cores (Switzerland). Environmental Earth Sciences 80, 421. https://doi.org/
aquitards. Hydrogeology Journal 31, 1999–2019. 10.1007/s12665-021-09708-1.
Calvert, S.E., Pedersen, T.F., 2007. Chapter fourteen: elemental proxies for Kneuker, T., Dohrmann, R., Ufer, K., Jaeggi, D., 2023. Compositional-structural
palaeoclimatic and palaeoceanographic variability in marine sediments: characterization of the Opalinus Clay and Passwang Formation: new insights from
interpretation and application. Developments in Marine Geology 1, 567–644. Rietveld refinement (Mont Terri URL, Switzerland). Applied Clay Science 242,
Carcione, J.M., Gei, D., Yu, T., Ba, J., 2019. Effect of clay and mineralogy on 107017. https://doi.org/10.1016/j.clay.2023.107017.
permeability. Pure and Applied Geophysics 176, 2581–2594. Korte, C., Hesselbo, S.P., Ullmann, C.V., Dietl, G., Ruhl, M., Schweigert, G., Thibault, N.,
Chapman, J., Umebayashi, M., deVet, T., Kulasek, M., Shen, A., Julien, C., Rauch, F., 2015. Jurassic climate mode governed by ocean gateway. Nature Communications 6,
Willie, B.M., 2025. Bone healing response to systemic bisphosphonate-prostaglandin 10015. https://www.nature.com/articles/ncomms10015.
E2 receptor 4 agonist treatment in female rats with a critical-size femoral segmental Lauper, B., Jaeggi, D., Deplazes, G., Foubert, A., 2018. Multi-proxy facies analysis of the
defect. Injury 56, 112269. https://doi.org/10.1016/j.injury.2025.112269. Opalinus Clay and depositional implications (Mont Terri rock laboratory,
Cnudde, V., Boone, M., Dewanckele, J., Dierick, M., Van Hoorebeke, L., Jacobs, P., 2011. Switzerland). Swiss Journal of Geosciences 111, 383–398.
3D characterization of sandstone by means of X-ray computed tomography. Lauper, B., Zimmerli, G.N., Jaeggi, D., Deplazes, G., Wohlwend, S., Rempfer, J.,
Geosphere 7 (1), 54–61. Foubert, A., 2021. Quantification of lithological heterogeneity within Opalinus Clay:
Craigie, N.W., Rees, A.J., 2016. Chemostratigraphy of glaciomarine sediments in the toward a uniform subfacies classification scheme using a novel automated core
Sarah Formation, Northwest Saudi Arabia. Journal of African Earth Sciences 117, image recognition tool. Frontiers in Earth Science 9, 645596. https://doi.org/
263–284. 10.3389/feart.2021.645596.
Crisci, E., Ferrari, A., Giger, S.B., Laloui, L., 2021. Effect of the mineralogical composition Liang, L., Sun, Y., Beets, C.J., Prins, M.A., Wu, F., Vandenberghe, J., 2013. Impacts of
on the elastoplastic hydromechanical response of Opalinus Clay shale. International grain size sorting and chemical weathering on the geochemistry of Jingyuan loess in
Journal of Rock Mechanics and Mining Sciences 143, 104747. https://doi.org/ the northwestern Chinese Loess Plateau. Journal of Asian Earth Sciences 69,
10.1016/j.ijrmms.2021.104747. 177–184.
Cui, Z., Huang, L., Fang, H., Han, D., Bombardelli, F., 2023. Exploring cohesive sediment Lin, M.C., Sihota, P., Kolibová, S.D., Fiedler, I.A.K., Krug, J., Wölfel, E.M., Moritz, M.,
flocculation with surface heterogeneity: a theoretical lagrangian-type flocculation Riedner, M., Ondruschka, B., Citak, M., Klebig, F., von Brackel, F.N.,
model. Water Resources Research 59, e2022WR034374. https://doi.org/10.1029/ Qwamizadeh, M., Jähn-Rickert, K., Busse, B., 2025. Fracture characteristics of
2022WR034374. human cortical bone influenced by the duration of in vitro glycation. JBMR Plus 9,
Delage, P., Cui, Y.J., Tang, A.M., 2010. Clays in radioactive waste disposal. Journal of ziae151. https://doi.org/10.1093/jbmrpl/ziae151.
Rock Mechanics and Geotechnical Engineering 2 (2), 111–123. Liu, J.F., Wu, Y., Cai, C.Z., Ni, H.Y., Cao, X.I., Pu, H., Song, S.B., Pu, S.Y., Skoczylas, F.,
Díaz-Curiel, J., Biosca, B., Arévalo-Lomas, L., Paredes-Palacios, D., Miguel, M.J., 2024. 2018. Investigation into water retention and gas permeability of Opalinus clay.
On the influence of grain size compared with other internal factors affecting the Environmental Earth Sciences 77, 213. https://doi.org/10.1007/s12665-018-7397-
permeability of granular porous media: redefining the permeability units. 3.
Lithosphere 2024 (1), lithosphere_2023_231. https://doi.org/10.2113/2024/ Löwemark, L., Chen, H.-F., Yang, T.-N., Kylander, M., Yu, E.-F., Hsu, Y.-W., Lee, T.-Q.,
lithosphere_2023_231. Song, S.-R., Jarvis, S., 2011. Normalizing XRF-scanner data: a cautionary note on the
Dietze, V., Gräbenstein, S., Franz, M., Schweigert, G., Wetzel, A., 2021. The Middle interpretation of high-resolution records from organic-rich lakes. Journal of Asian
Jurassic Opalinuston Formation (Aalenian, Opalinum Zone) at its type locality near Earth Sciences 40, 1250–1256.
Bad Boll and adjacent outcrops (Swabian Alb, SW Germany). Palaeodiversity 14, Ma, J., Li, L., Wang, H., Du, Y., Ma, J., Zhang, X., Wang, Z., 2022. Carbon capture and
15–113. storage: history and the road ahead. Engineering 14, 33–43.
Dohrmann, R., Genske, D., Karnland, O., Kaufhold, S., Kiviranta, L., Olsson, S., Mazurek, M., Gimmi, T., Zwahlen, C., Aschwanden, L., Gaucher, E.C., Kiczka, M.,
Plötze, M., Sandén, T., Sellin, P., Svensson, D., Valter, M., 2012. Interlaboratory CEC Rufer, D., Wersin, P., Marques Fernandes, M., Glaus, M.A., Van Loon, L.R.,
and exchangeable cation study of bentonite buffer materials: I. Cu (II)- Traber, D., Schnellmann, M., Vietor, T., 2023. Swiss deep drilling campaign
triethylenetetramine method. Clays and Clay Minerals 60 (2), 162–175. 2019–2022: geological overview and rock properties with focus on porosity and
Evans, C.L., Wightman, E.M., Yuan, X., 2015. Quantifying mineral grain size pore-space architecture. Applied Geochemistry 159, 105839. https://doi.org/
distributions for process modelling using X-ray micro-tomography. Minerals 10.1016/j.apgeochem.2023.105839.
Engineering 82, 78–83. Minardi, A., Crisci, E., Ferrari, A., Laloui, L., 2016. Anisotropic volumetric behaviour of
Fantasia, A., Adatte, T., Spangenberg, J.E., Font, E., Duarte, L.V., Föllmi, K.B., 2019. Opalinus clay shale upon suction variation. Géotechnique Letters 6, 144–148.
Global versus local processes during the Pliensbachian–Toarcian transition at the Nagra, 2002. Projekt Opalinuston: Synthese der geowissenschaftlichen
Peniche GSSP, Portugal: a multi-proxy record. Earth-Science Reviews 198, 102932. Untersuchungsergebnisse. Entsorgungsnachweis für abgebrannte Brennelemente,
https://doi.org/10.1016/j.earscirev.2019.102932. verglaste hochaktive sowie langlebige mittelaktive Abfälle. In: Nagra Technischer
Fernandes, M.M., Mazurek, M., Wersin, P., Wüst, R., Baeyens, B., 2024. Cation-exchange Bericht NTB 02–03. Nagra, Wettingen, Switzerland (Available at nagra.ch (in
properties of the Mesozoic sedimentary sequence of Northern Switzerland and German).
modelling of the Opalinus Clay porewater. Applied Geochemistry 162, 105852. Nagra, 2014a. SGT Etappe 2: Vorschlag weiter zu untersuchender geologischer
https://doi.org/10.1016/j.apgeochem.2023.105852. Standortgebiete mit zugehörigen Standortarealen für die Oberflächenanlage:

16
G.N. Zimmerli et al. Sedimentary Geology 489 (2025) 106982

Geologische Grundlagen, Dossier I: Einleitung und Zusammenfassung. In: Nagra Weltje, G.J., Tjallingii, R., 2008. Calibration of XRF core scanners for quantitative
Technischer Bericht NTB 14–02. Nagra, Wettingen, Switzerland (Available at nagra. geochemical logging of sediment cores: theory and application. Earth and Planetary
ch (in German). Science Letters 274, 423–438.
Nagra, 2014b. SGT Etappe 2: Vorschlag weiter zu untersuchender geologischer Wengler, M., Barboza-Gudiño, J.R., Thomsen, T.B., Meinhold, G., 2019. Sediment
Standortgebiete mit zugehörigen Standortarealen für die Oberflächenanlage: provenance of Triassic and Jurassic sandstones in Central Mexico during activity of
Geologische Grundlagen, Dossier IV: Geomechanische Unterlagen. In: Nagra the Nazas volcanic arc. Journal of South American Earth Sciences 92, 329–349.
Technischer Bericht NTB 14–02. Nagra, Wettingen, Switzerland (Available at nagra. Wentworth, C.K., 1922. A scale of grade and class terms for clastic sediments. The
ch (in German). Journal of Geology 30, 377–392.
Nagy, J., Dypvik, H., Bjaerke, T., 1984. Sedimentological and paleontological analysis of Wetzel, A., Allia, V., 2003. Der Opalinuston in der Nordschweiz: Lithologie und
Jurassic North Sea deposits from deltaic environments. Journal of Petroleum Ablagerungsgeschichte. Eclogae Geologicae Helvetiae 96, 451–469 (in German).
Geology 7, 169–187. Williams, C.J., Hesselbo, S.P., Jenkyns, H.C., Morgans-Bell, H.S., 2001. Quartz silt in
Nikitenko, B., Shurygin, B., Mickey, M., 2008. High resolution stratigraphy of the lower mudrocks as a key to sequence stratigraphy (Kimmeridge Clay Formation, late
Jurassic and Aalenian of Arctic regions as the basis for detailed palaeobiogeographic Jurassic, Wessex Basin, UK). Terra Nova 13, 449–455.
reconstructions. Norwegian Journal of Geology 88, 267–277. Wohlwend, S., Bläsi, H.R., Feist-Burkhardt, S., Hostettler, B., Menkveld-Gfeller, U.,
Nunn, J.A., Lin, G., Zhang, L., 1999. Thermal insulation by low thermal conductivity Dietze, V., Deplazes, G., 2022. TBO Stadel-2-1: Data Report, Dossier IV: Microfacies,
shales: Implications for basin-scale fluid flow and heat transport. In: Förster, A., bio- and chemostratigraphic analysis. In: Nagra Arbeitsbericht NAB 22–02. Nagra,
Merriam, D.F. (Eds.), Geothermics in Basin Analysis. Computer Applications in the Wettingen, Switzerland (Available at nagra.ch).
Earth Sciences. Springer, Boston, pp. 117–130. Wohlwend, S., Hostettler, B., Menkveld-Gfeller, U., Feist-Burkhardt, S., Bläsi, H.,
Nweke, O.M., Omeokachie, A.I., Okogbue, C.O., 2023. Characterization, technological Zimmerli, G.N., Dietze, V., Bernasconi, S.M., Deplazes, G., 2024. High-resolution bio-
properties and utilization of clay-rich argillite quarry waste as raw material in and chemostratigraphy of the Toarcian to Aalenian of Northern Switzerland with a
ceramics and other industrial applications. Arabian Journal of Geosciences 16, 506. focus on the Opalinus Clay. Newsletters on Stratigraphy 57 (4), 413–444.
https://doi.org/10.1007/s12517-023-11617-5. Wu, L., Wilson, D.J., Wang, R., Yin, X., Chen, Z., Xiao, W., Huang, M., 2020. Evaluating
Ohmert, W., Allia, V., Arias, C., Baldanza, A., Bergen, J.A., Bucefalo Palliani, R., Zr/Rb ratio from XRF scanning as an indicator of grain-size variations of
Canales, M.L., de Kaenel, E., Garcia Joral, F., Goy, A., Herrero, C., Höhndorf, A., glaciomarine sediments in the Southern Ocean. Geochemistry, Geophysics,
Martinez, G., Mattioli, E., Perilli, N., Riegraf, W., Rolf, C., Ureta, S., Wetzel, A., Geosystems 21, e2020GC009350. https://doi.org/10.1029/2020GC009350.
Wonik, T., 1996. Die Grenzziehung Unter− /Mitteljura (Toarcium/Aalenium) bei Young, K.E., Evans, C.A., Hodges, K.V., Bleacher, J.E., Graff, T.G., 2016. A review of the
Wittnau und Fuentelsaz – Beispiele interdisziplinärer geowissenschaftlicher handheld X-ray fluorescence spectrometer as a tool for field geologic investigations
Zusammenarbeit. LGRB-Informationen 8, 1–52 (in German). on Earth and in planetary surface exploration. Applied Geochemistry 72, 77–87.
Ohta, T., 2004. Geochemistry of Jurassic to earliest cretaceous deposits in the Nagato Zhang, M.M., Liu, Z.J., Fang, S., 2016. Sandstone grain size characteristics of the Upper
Basin, SW Japan: implication of factor analysis to sorting effects and provenance Jurassic Emuerhe Formation in the western region, Mohe Basin (NE China).
signatures. Sedimentary Geology 171, 159–180. Geological Journal 51, 652–669.
Rogov, M.A., Zakharov, V.A., 2010. Jurassic and lower cretaceous glendonite Zheng, J., Zheng, L., Liu, H.H., Ju, Y., 2015. Relationships between permeability,
occurrences and their implication for arctic paleoclimate reconstructions and porosity and effective stress for low-permeability sedimentary rock. International
stratigraphy. Earth Science Frontiers 17 (Special Issue), 345–347. Journal of Rock Mechanics and Mining Sciences 78, 304–318.
Safari, H., Balcom, B.J., Afrough, A., 2021. Characterization of pore and grain size Ziefle, G., Cajuhi, T., Costabel, S., Fourriére, A., Furche, M., Gerowski, J., Kneuker, T.,
distributions in porous geological samples – an image processing workflow. Königer, F., Maßmann, J., 2024. CD-A twin niches in the Mont Terri Rock
Computers & Geosciences 156, 104895. https://doi.org/10.1016/j. Laboratory: characterization and interpretation of hydraulic parameters with regard
cageo.2021.104895. to safety aspects. International Journal of Rock Mechanics and Mining Sciences 174,
Schmidt, F.N., Hahn, M., Stockhausen, K.E., Rolvien, T., Schmidt, C., Knopp, T., 105624. https://doi.org/10.1016/j.ijrmms.2023.105624.
Schulze, C., Püschel, K., Amling, M., Busse, B., 2022. Influence of X-rays and gamma- Zimmerli, G.N., Wohlwend, S., Deplazes, G., Becker, J., Wetzel, A., Francescangeli, F.,
rays on the mechanical performance of human bone factoring out intraindividual Foubert, A., 2024. Facies variability and depositional cyclicity in central Northern
bone structure and composition indices. Materials Today Bio 13, 100169. https:// Switzerland: insights from new Opalinus Clay drill cores. Swiss Journal of
doi.org/10.1016/j.mtbio.2021.100169. Geosciences 117, 18. https://doi.org/10.1186/s00015-024-00463-6.
Seiphoori, A., Whittle, A.J., Krakowiak, K.J., Einstein, H.H., 2017. Insights into Zimmerli, G.N., Wohlwend, S., Deplazes, G., Mann, T., Erbacher, J., Kneuker, T.,
diagenesis and pore structure of Opalinus shale through comparative studies of Eberli, G.P., Foubert, A., 2025. The hidden message of fine-grained sediments:
natural and reconstituted materials. Clays and Clay Minerals 65 (2), 135–153. towards a new depositional model for the Opalinus Clay. Applied Clay Science 276,
Tournassat, C., Bourg, I.C., Steefel, C.I., Bergaya, F., 2015. Chapter 1 - surface properties 107934. https://doi.org/10.1016/j.clay.2025.107934.
of clay minerals. Developments in Clay Science 6, 5–31. Zou, C., Qiu, Z., Zhang, J., Li, Z., Wei, H., Liu, B., Zhao, J., Yang, T., Zhu, S., Tao, H.,
Van der Schee, M., Sierro, F.J., Jiménez-Espejo, F.J., Hernández-Molina, F.J., Flecker, R., Zhang, F., Wang, Y., Zhang, Q., Liu, W., Liu, H., Feng, Z., Liu, D., Gao, J., Liu, R.,
Flores, J.A., Acton, G., Gutjahr, M., Grunert, P., García-Gallardo, Á., Andersen, N., Li, Y., 2022. Unconventional petroleum sedimentology: a key to understanding
2016. Evidence of early bottom water current flow after the Messinian Salinity Crisis unconventional hydrocarbon accumulation. Engineering 18, 62–78.
in the Gulf of Cadiz. Marine Geology 380, 315–329. Zwahlen, C., Aschwanden, L., Gaucher, E., Gimmi, T., Jenni, A., Kiczka, M., Mäder, U.,
Verweij, H., Vis, G.J., Imberechts, E., 2016. Spatial variation in porosity and Mazurek, M., Ross, D., Rufer, D., Waber, H.N., Wersin, P., Traber, D., 2022. TBO
permeability of the Rupel Clay Member in the Netherlands. Netherlands Journal of Stadel-2-1: Data Report, Dossier VIII: Rock properties, porewater characterisation
Geosciences 95 (3), 253–268. and natural tracer profiles. In: Nagra Arbeitsbericht NAB 22–02. Nagra, Wettingen,
Waber, H.N., 2020. SGT-E3 deep drilling campaign (TBO): experiment procedures and Switzerland (Available at nagra.ch).
analytical methods at RWI, University of Bern (Version 1.0, April 2020). In: Nagra
Arbeitsbericht, NAB 20–13, Nagra, Wettingen, Switzerland (Available at nagra.ch).

17