Geochemical fingerprinting of key lithologies and depositional processes across the upper boundary of the Opalinus Clay (Aalenian, Middle Jurassic, northern Switzerland)

The Opalinus Clay is an argillaceous to silty mudstone formation, notable in Switzerland as the selected host rock for deep geological disposal of radioactive waste. Its upper bounding unit (Passwang Formation and eastern equivalents) is composed of successions of mudstone, sandy bioclastic marl and limestone separated by ooidal ironstone beds. The lithostratigraphic transition is diachronous across northern Switzerland and shows high vertical and lateral lithological variability. To constrain this variability into predictive models, and to identify horizons with properties that could potentially influence radionuclide mobility, the sedimentological and diagenetic processes involved in the genesis of this transition have to be investigated. The present study aims at testing the applicability of X‐ray fluorescence chemostratigraphy to characterise the mixed carbonate–siliciclastic units and understand the complex genesis of the lithostratigraphic transition from the Opalinus Clay towards its upper bounding unit. Sediment drill cores from four locations across northern Switzerland (Mont Terri, Riniken, Weiach and Benken) are analysed using high‐resolution X‐ray fluorescence core scanning. Data are compared to petrographic and additional geochemical data sets (inductively coupled plasma mass spectrometry, scanning electron microscopy with energy dispersive X‐ray analysis, micro‐X‐ray fluorescence mapping) obtained from powdered samples, thin section analyses and drill core slabs. The results demonstrate that the combination of these rapid and non‐destructive measurements along with multivariate data analysis allows the fast and objective classification of lithofacies along complex sedimentary successions. Moreover, it provides quantitative means for differentiating between prominent depositional and post‐depositional processes. The lithostratigraphic transition has been traced by the use of specific elemental proxies as a discontinuity, and its genesis linked to sediment bypassing.

To constrain this variability into predictive models, and to identify horizons with properties that could potentially influence radionuclide mobility, the sedimentological and diagenetic processes involved in the genesis of this transition have to be investigated.The present study aims at testing the applicability of X-ray fluorescence chemostratigraphy to characterise the mixed carbonate-siliciclastic units and understand the complex genesis of the lithostratigraphic transition from the Opalinus Clay towards its upper bounding unit.Sediment drill cores from four locations across northern Switzerland (Mont Terri, Riniken, Weiach and Benken) are analysed using high-resolution X-ray fluorescence core scanning.Data are compared to petrographic and additional geochemical data sets (inductively coupled plasma mass spectrometry, scanning electron microscopy with energy dispersive X-ray analysis, micro-X-ray fluorescence mapping) obtained from powdered samples, thin section analyses and drill core slabs.The results demonstrate that the combination of these rapid and non-destructive measurements along with multivariate data analysis allows the fast and objective classification of lithofacies along complex sedimentary successions.Moreover, it provides quantitative means for differentiating between prominent depositional and post-depositional processes.The lithostratigraphic transition has been traced by the use of specific elemental proxies as a discontinuity, and its genesis linked to sediment bypassing.

| INTRODUCTION
The Opalinus Clay is a mudstone succession with very low permeability (Nagra, 2002).This formation has been proposed as the potential host rock for radioactive waste disposal in Switzerland (Nagra, 2014).Since 1996, it has been intensively studied in the Mont Terri rock laboratory, an underground research facility situated in the Jura Mountains, in north-western Switzerland (Bossart et al., 2017;Thury and Bossart, 1999).The Passwang Formation and the 'Murchisonae-Oolith Formation' are lithostratigraphic units overlying the Opalinus Clay and are composed of successions of mudstone, sandy bioclastic marl and limestone, separated by Fe-ooidal intervals (Bläsi et al., 2013;Burkhalter, 1996;Hostettler et al., 2017;Wohlwend et al., 2019).This lithostratigraphic transition is diachronous across northern Switzerland and displays high vertical and lateral facies variability.In order to constrain this lithological variability into predictive models, and to identify horizons with properties that could potentially influence radionuclide mobility, it is essential to understand the different sedimentological and diagenetic processes involved in the formation of this transition.
The aim of the present study is: (a) to explore the applicability of XRF core scanning to differentiate between lithologies; (b) to characterise the depositional and post-depositional processes along a complex, mixed carbonate-siliciclastic lithostratigraphic transition; and (c) to understand better the sedimentary and geochemical processes that shaped the vertical and lateral variability at the lithostratigraphic boundary of the Opalinus Clay with its overlying units in four drill cores across northern Switzerland (Mont Terri, Riniken, Weiach and Benken).

| GEOLOGICAL SETTING
The Opalinus Clay and its upper bounding unit (Passwang Formation and eastern equivalents) were deposited during late Toarcian to early Bajocian times in a shallow-marine epicontinental sea covering central Europe (Ziegler, 1990).At that time, central Europe was situated about 30° north of the equator (Irving, 1977).Considering that no major glaciation is known from the Jurassic, one can infer a subtropical to tropical climate (cf.Etter, 1990) with dense vegetation and lateritic soil formation on the continent (Burkhalter, 1995;Siehl and Thein, 1989;Young, 1989).While sediment provenance and palaeogeography remain unclear, it is suggested that the closest landmasses were situated at least tens to hundreds of kilometres away from the investigated sites (Etter, 1990;Wetzel and Allia, 2003).Situated at the southern end of the South German Basin, the Swiss northern basin was bordered by the Rhenish Massif to the north-west, by the Bohemian Massif to the east, by the Alemannic Islands to the south and by the Burgundy High to the west.
Although the Opalinus Clay has not yet been officially delineated by the Swiss Committee for Stratigraphy (Burkhalter and Heckendorn, 2009;Remane et al., 2005), this informal lithostratigraphic unit has a long history in northern Switzerland (Quenstedt, 1843;1858;Schmidt et al.,1924).In Germany, on the other hand, the term Opalinuston Formation (Franz and Nitsch, 2009) is used and its upper limit is set within the Comptum Subzone (Opalinum Zone), at the base of the Comptumbank.A similar definition was used in northern Switzerland by Burkhalter (1996) to distinguish the Opalinus Clay from the overlying Passwang Formation.However, biostratigraphic studies have demonstrated that this lithological transition is diachronous across northern Switzerland, rendering its precise location and lateral correlation difficult.While the transition is attributed to the upper Opalinum Subzone (lower Opalinum Zone) in the Mont Terri area (north-western Switzerland; see Figure 1; Hostettler et al., 2020), it is sometimes younger towards the east, typically formed during the Comptum Subzone to the Murchisonae Zone (Bläsi et al., 2013;Feist-Burkhardt, 2012;Wohlwend et al., 2019).The lithostratigraphic boundary between the Opalinus Clay and its upper bounding unit (henceforth abbreviated to OPA/UBU) is consequently not always straightforward to delineate.While in most locations the OPA/UBU boundary is marked by a conspicuous calcareous horizon (Bläsi, 1987;Burkhalter, 1996;Burkhalter et al., 1997;Hostettler et al., 2017;Matter et al., 1987;1988;Wohlwend et al., 2019), some successions display gradual lithological changes with successive calcareous horizons (Frickberg; Wohlwend et al., 2019).Hence, within the Benken drill core, the position of the OPA/UBU boundary was reallocated after new biostratigraphic data were acquired (Bläsi et al., 2013;Feist-Burkhardt, 2012).
The 80-130 m thick Opalinus Clay is present in the Swiss Jura and Molasse Basin, as well as in Germany.It is an argillaceous to silty mudstone, which is composed of clay minerals (kaolinite, illite-smectite mixed layers, illite and chlorite), quartz (mainly coarse silt to fine sand) and carbonates (mostly calcite, some siderite, dolomite and ankerite) in varying concentrations.Some minor and accessory components include pyrite, K-feldspar, plagioclase, biotite, muscovite, rutile, zircon, apatite, monazite, glauconite and organic matter (Peters, 1962;Pearson et al., 2003;Lerouge et al., 2014;see Mazurek, 2011 for an overview of the existing mineralogical database).The Opalinus Clay lithology varies regionally.In the Mont Terri area, the formation is divided into five lithostratigraphic sub-units, characterised by three lithofacies: the shaly facies, the carbonate-rich sandy facies and the sandy facies (Bossart and Thury, 2008;Hostettler et al., 2017;Lauper et al., 2018).Further east, the Opalinus Clay is typically divided into five to six subunits, tentatively correlatable across the area (Bläsi, 1987;Matter et al., 1987;1988;Nagra, 2001).A lithostratigraphic overview of the Opalinus Clay across northern Switzerland is provided by Nagra (2002) and Mazurek and Aschwanden (2020).Detailed lithofacies and subfacies descriptions and interpretations are provided by Wetzel and Allia (2003) and Lauper et al. (2018).
In the northern Swiss Jura, the Opalinus Clay is overlain by the Passwang Formation.Originally defined as an alloformation, it is characterised by parasequences of mudstone, sandy bioclastic marl and limestone, separated by ooidal ironstone beds (Burkhalter, 1996).Strong regional variations in lithology and facies distribution occur.East of the Jura (eastwards from the Aare River), the 'Murchisonae-Oolith Formation', the Wedelsandstein Formation and the 'Humphriesioolith Formation' (as defined by Bläsi et al., 2013 and references therein) are time-equivalents of the Passwang Formation.In the Mont Terri area, the basal Passwang Formation, the so-called Sissach Member, shows a thin intraclastic, Fe-ooidal horizon followed by sandy marl and bioclastic limestone (Hostettler et al., 2017).In the eastern Jura, the basal Passwang Formation forms four different lithofacies, varying mainly between spathic, sandy marl and limestone, and Fe-ooidal ironstone (Wohlwend et al., 2019).Further east, the 'Murchisonae-Oolith Formation' shows a Fe-ooidal horizon followed by sandy marl with bioclastic limestone intervals (Bläsi et al., 2013;Burkhalter et al., 1997).
The complex lithostratigraphic and biostratigraphic framework of the OPA/UBU transition is typically attributed to differential subsidence (Bläsi et al., 2013;Wohlwend et al., 2019).During the Lower Jurassic, an extensive stress field caused by the opening of the Tethys and the Atlantic Ocean induced the reactivation of pre-existing basement structures (Permo-Carboniferous troughs; Reisdorf and Wetzel, 2018; Map of Switzerland and location of the studied drill cores.Source: Federal Office of Topography swisstopo.(B) Synthesis of the Aalenian stratigraphy in northern Switzerland (modified after Bläsi et al., 2013;Wohlwend et al., 2019).The boundary delineating the Opalinus Clay from its upper bounding unit fluctuates between the Opalinum and Murchisonae zones (see text for details).The Comptum Subzone constitutes the upper part of the Opalinum Zone, while its lower part is referred to as the Opalinum Subzone.Further subzones are not indicated.The Hauenstein Member (Burkhalter, 1996) is undifferentiated at Mont Terri (Hostettler et al., 2017).'Brauner Dogger' refers to Bläsi et al. (2013).Fm = Formation; Mb = Member Wetzel et al., 2003;Wildi et al., 1989).The rifting was associated with the formation of fault bounded tilted blocks, which led to a series of local basins by differential subsidence.The resulting depositional area was morphologically differentiated into swells and depressions, contributing significantly to the lateral facies variability (Burkhalter, 1996;Lauper et al., 2018;Wetzel and Allia, 2000;2003;Wetzel and Meyer, 2006).

| Sedimentary petrography
The petrography and mineralogy of the Opalinus Clay and its upper bounding unit is documented in several reports and publications (Bläsi, 1987;Burkhalter, 1995;1996;Franz and Nitsch, 2009;Kneuker et al., 2020;Lauper et al., 2018;Lerouge et al., 2014;Matter et al., 1987;1988;Mazurek, 2011;Mazurek and Aschwanden, 2020;Nagra, 2001;2002;Wetzel and Allia, 2003;Wohlwend et al., 2019).Additionally, new petrographic analyses of selected drill core sections were performed at macro-scale and micro-scale.For drill cores acquired more than 30 years ago (Riniken and Weiach), pictures taken shortly after coring were compared to new facies observations to evaluate post-drilling alteration.Existing thin sections from the Riniken, Weiach and Benken drill cores (Nagra internal collection; respectively Matter et al., 1987;1988;Nagra, 2001), as well as thin sections from a drill core crossing the Mont Terri rock laboratory (BDB-1; Reisdorf et al., 2016) were used for microfacies descriptions.Most of the thin sections were stained with potassium ferricyanide and alkaline red S in acid solution (Dickson, 1965).Additional polished thin sections were retrieved from the BPE-2 core (Mont Terri, drill core close to BPE-3; Gygax et al., 2017) and from the Riniken core.Thin sections were studied by classical petrographic microscopy (Leica DM4500 P; Leica).Additional observations, analyses and elemental mapping of selected samples from the Riniken core were performed with a FEI XL30 Sirion FEG (Thermo Fisher Scientific) scanning electron microscope (SEM) coupled with an Energy Dispersive Spectrometer (EDX; X-Max 150 Silicon Drift Detector; Oxford Instruments) at 20 kV (Department of Geosciences, University of Fribourg).Prior to analyses, a thin carbon layer was sputter-coated on the samples.
Siliciclastic-dominated lithologies are described following Lazar et al. (2015), while bedding structures (lenticular and flaser) refer to Reineck and Wunderlich (1968).Carbonate-dominated lithologies are described using the  Dunham (1962) classification.Iron-rich rocks are described as ironstones (abnormally high iron content), following the terms and glossary summarised by Young (1989).

| X-ray fluorescence core scanning
Elemental intensities were measured using an ITRAX XRF core scanner (Cox Analytical Systems) at the University of Bern (see Croudace et al., 2006 for details on the ITRAX core scanner).Different measurement settings (current and voltage) were applied, while exposure time was constantly set at 50 s.
Measurements were performed at 10 mm intervals in relatively homogeneous sediments and at 5 mm in more heterogeneous material (mostly carbonate) for all cores, except the Benken sections, which were measured at 5 mm intervals throughout all lithologies.
The BPE-3 whole-round core sections from Mont Terri were measured with a Mo-tube at 30 kV and 40 mA.The Riniken archive split-core sections were measured with a Mo-tube set to 30 kV and 50 mA, and with a Cr-tube set to 30 kV and 50 mA.The Weiach archive split-core sections were measured with a Mo-tube set to 30 kV and 40 mA.The Benken whole-round core sections (except the OPA/UBU interval that was split; 539.72-539.52 m) were measured with a Cr-tube set to 30 kV and 50 mA.
Raw X-ray fluorescence core scanning (XRF-CS) spectra were processed in the software Q-spec (version 8.6.0).Standard adjustment and refinement of the peak-fitting parameters using representative parts of the core sections was performed.Outliers corresponding to structural defaults were subsequently removed.

| XRF-CS data analysis
Resulting downcore variations of selected elements are presented as ratios of element-intensity counts normalised to Al. Aluminium is assumed to be exclusively of detrital origin, a good representative of the clay fraction, and shows a conservative behaviour throughout changing redox conditions (Calvert and Pedersen, 2007;Löwenmark et al., 2011).
To investigate the relationship between elemental content across the different lithologies, principle component analysis (PCA) was performed on each data set using the freely available PAST software (Hammer et al., 2001).Detailed information on PCA is provided by Wold et al. (1987).Considering the compositional nature of elemental data, a centred log-ratio (clr) transformation was applied to the whole XRF-CS data sets prior to element selection and PCA (Aitchison, 1986;Bloemsma et al., 2012).Biplots were used to determine the type of relationship between selected elements and to visualise their differing trends among distinct lithostratigraphic/lithological units.

| Micro X-ray fluorescence mapping
In order to characterise chemical diversity among the mineralogical, textural and structural phases, three representative areas (different sizes, a few tens of cm 2 ) were mapped along a 1 cm F I G U R E 2 Representative sections of the key lithologies.KL1: argillaceous to argillaceous-siliceous mudstone (Opalinus Clay interval at Riniken).KL2: siliceous-calcareous sandstone to sandy bioclastic packstone/grainstone (basal Passwang Formation at Mont Terri; cross-bedding drilled at 57° angle).KL3: bioclastic sandy wackestone to packstone (lower part of the Weiach OPA/UBU-delineating horizon).KL4 ooidal ironstone (basal Passwang Formation at Riniken).Scale bar is 1 cm thick slab representing the OPA/UBU transition in the Riniken core sections.Micro-XRF mapping was performed at ultra-high resolution (detector spot size = 50 µm) with the Eagle III micro X-ray fluorescence (µ-XRF) spectrometer (Röntgenanalytik Messtechnik GmbH) at the University of Geneva.Current settings were set at 40 kV and 400 mA, and integration time at 10 µs.

| Inductively coupled plasma mass spectrometry
Two vertically adjacent sub-samples of 8 cm 3 of bulk rock were taken from 10 spots along the Riniken core sections (see Figure S1) and ground in an agate mortar.Major and  Matter et al. (1987;1988), Bläsi et al. (2013) and Hostettler et al. (2017).The green brackets indicate the areas illustrated in Figure 4. Fm = Formation; 'Mur.-Oo.Fm' = 'Murchisonae-Oolith Formation'.Ooidal ironstone (basal Passwang Formation at Riniken).Scale bar is 1 cm trace element abundances of the resulting 20 samples were analysed by inductively coupled plasma mass spectrometry (ICP-MS) through a commercial laboratory (Actlabs; package Code Ultratrace 7).Prior to analysis, samples were fused with sodium peroxide in a zirconium crucible and acidified with concentrated nitric and hydrochloric acids.The resulting solutions were diluted and analysed by an Agilent 7900 ICP-MS (Agilent Technologies).Calibration was performed using five synthetic calibration standards.

UBU transition
Four key lithologies (KL) are observed along the studied successions (Figure 2).
Lithofacies KL1 consists of an argillaceous to argillaceous-siliceous mudstone.Occasional to abundant millimetre to centimetre-thick lenses occur within a fine-grained and clay-rich matrix (lenticular bedding).Only a few lenses are continuous over the overall core diameter.They are composed of silt to fine-sand quartz grains and bioclastic debris (mainly mollusc shell and echinoderm fragments).The grains are mostly cemented by Fe-rich calcite.A few pyrite framboids occur throughout the matrix and lenses.This lithology is characteristic of the investigated Opalinus Clay intervals.
Lithofacies KL2 consists of a siliceous-calcareous sandstone to sandy bioclastic packstone/grainstone. Abundant fine-sand quartz grains are mixed with bioclastic fragments (mainly echinoderms and mollusc shells).A few iron-coated grains, dolomite crystals and pyrite framboids occur.The proportion of quartz and bioclasts varies along the core sections.This lithology is typical of the basal Passwang Formation at Mont Terri.
Lithofacies KL3 consists of a bioclastic sandy wackestone to packstone.This lithology is highly heterogeneous and includes calcitic/sideritic nodules and calcareous hardgrounds with frequent intraclasts.Some intraclasts contain ironcoated grains (including Fe-ooids) and ferruginous microbialites.Individual Fe-ooids also occur.Quartz grains are abundant in certain horizons, and heavy minerals such as zircon, monazite and rutile may be present.Pyrite and siderite represent significant components within certain horizons.This lithology typically forms the OPA/UBU-delineating horizons, but is not limited to them.
Lithofacies KL4 consists of an ooidal ironstone.This lithology is characterised by a rust-red colour primarily attributed to the abundant Fe-ooids and limonitic components.The Fe-ooids are mostly spherical or ellipsoidal and less than 2 mm in diameter.Some Fe-pisoids, Fe-oncoids and Fe-spastoliths are also present, although all are named as Fe-ooids henceforth.Their mineralogy was not investigated within this study but is heterogeneous and dominated by goethite and chamosite (Burkhalter, 1995;1996;Hostettler et al., 2017).The limonitised bioclastic fragments and Fe-ooids typically occur within a dominantly sparitic matrix (calcite and dolomite).Some intervals show an important amount of Fe-rich and Al-rich clay minerals.Quartz silt-sized to sand-sized grains are also present in varying quantities.Pyrite is rare to absent.This lithology is typical of the Passwang Formation interval at Riniken and the 'Murchisonae-Oolith Formation' interval at Weiach.

| Sedimentary petrography
Figure 3 summarises the lithology of the four studied core intervals, and Figure 4 shows an expanded view of the lithological variability along the OPA/UBU transitions, such as defined by Matter et al. (1987;1988), Bläsi et al. (2013) and Hostettler et al. (2017).

| Mont Terri
From stratigraphic base to top (11.48-14.32drilling m), the BPE-3 core sections display first the upper sandy facies of the Mont Terri Opalinus Clay (11.48-13.02m; Bossart and Thury, 2008; in this study described as KL1).Thin planar to slightly wavy lenses occur within an argillaceous matrix.They primarily represent starved ripples, but some bioturbation (round burrows) is also visible.A homogeneous micritic bed (mudstone/wackestone; probably a nodule at larger scale; see Hostettler et al., 2020) is observed between 12.21 and 12.26 m.
The OPA/UBU lithostratigraphic boundary is sharp and characterised by an 8-11 cm thick, calcareous hardground (herein referred to as the OPA/UBU-delineating horizon).This horizon is composed of merged and jointly compacted, reworked Fe-ooidal and bioclast-rich intraclasts (wackestone/packstone) and exhumed biomicritic nodules (KL3; Figures 4 and 5A).Stylolitic-like structures are visible and extend over several centimetres.The edges of the F I G U R E 6 Lithological features from the Riniken core.(A) The erosive surface within the Transition Zone (uppermost Opalinus Clay) in the Riniken core is indicated by black arrows.The lower part shows in-situ calcitic/sideritic nodules forming a firm ground at the top of an argillaceous-siliceous mudstone facies.Above the erosive surface, biolithic and lithoclastic intraclasts are present within a quartz-rich, sideritic muddy matrix.(B) Photomicrograph of Fe-ooidal intraclasts (see location in A). (C) Photomicrograph of the ferruginous quartz-rich matrix (see location in A). (D) Calcareous neomorphosed ooids from the uppermost nodular calcareous layer at the top of the sideritic condensed interval (Transition Zone in the Riniken core, below layer 'a'; see Figure 7).(E) Photomicrograph displaying the interface between two calcareous layers at the top of the Transition Zone in the Riniken core (OPA/UBU-delineating horizon).The lower layer ('a' in Figure 7) consists of a sandy wackestone and the upper layer ('b' in Figure 7) consists of a bioclastic packstone.(F) Photomicrograph of the >1 cm pyritised Fe-ooidal layer delineating the OPA/UBU lithostratigraphic boundary at Riniken.(G) Photomicrograph of the typical lithology of the basal Passwang Formation at Riniken.Feooids occur within a calcitic, dolomitic and limonitic matrix.All photomicrographs were acquired in plane-polarised light uppermost lithoclasts are encrusted by ferruginous microbialites with a stromatolitic texture (Figure 5B).Scarce to abundant bioclasts (mainly echinoderms, mollusc shell debris and foraminifers), common Fe-ooids, limonite-coated clasts, quartz silt grains and euhedral pyrite are present.Bioturbation features and boring traces are identified throughout this horizon (Figure 5C,D).

| Riniken
From stratigraphic base to top (335.15-330.06m), the studied core sections display about 4 m of Opalinus Clay (335.15-331.14m).The lower part (335.15-331.90m) is characterised by a lenticular mudstone facies (KL1).Intervals with abundant and thick lenses alternate with intervals with rare and thin lenses.Some of the thicker lenses display current ripples.The level of bioturbation is medium to high, characterised by disrupted intervals and frequent burrows (rounded lenses) throughout the matrix.At a few locations, the lenticular structures are impregnated by rust-red Fe-oxides suggesting post-diagenetic alteration of Fe-minerals.
The latter ends within a <1 cm thick layer composed almost exclusively of Fe-ooids, some of which are fully pyritised (d; Figure 6F).Above this layer follows the Passwang Formation (from 331.14 m onwards; cf.Matter et al., 1987;Burkhalter, 1995).
The basal Passwang Formation (331.14-330-06m) displays a rust-red coloured ooidal ironstone typical of KL4.The lower part is characterised by a 20 cm thick cross-bedded interval followed by a more bioturbated section.Highly dolomitised (possibly ankerite) intervals occur (Figures 6G  and 8).Quartz grains are common to abundant, and bioclasts consist mostly of echinoderm and mollusc shell fragments.Recrystallised ammonite and belemnite remains are observed.

| Weiach
From stratigraphic base to top (555.80-553.16m), the core sections display an argillaceous mudstone facies (KL1; 555.80-554.50m) characteristic of the Opalinus Clay.Only rare quartz/bioclast-rich lenses occur.At the time the core sections were inspected for this study, white post-diagenetic gypsum precipitation was observed around artificial desiccation cracks at the core surface.This alteration feature is associated with the dehydration caused by prolonged core storage.
The OPA/UBU lithostratigraphic boundary is characterised by a 32 cm thick, calcareous unit (OPA/UBUdelineating horizon; 554.82-554.50m; Figures 4 and 9A).It is vertically heterogeneous and limestone beds alternate with muddy sandstone layers.The limestone beds consist of bioclastic, sandy, Fe-ooidal packstone to grainstone (KL3).Bioclasts include mainly mollusc shells, echinoderms and bryozoans.Limonitic components are abundant, but only a few show concentric structures typical of Fe-ooids.Some well-defined, millimetre to over 7 cm-sized intraclasts occur, a few of which are dolomitic.Stylolites are visible within the whole interval.Three 23 cm thick, interbedded muddy sandstone layers are present between the limestone beds.Parts of the overall OPA/UBU-delineating horizon display post-diagenetic precipitation of brownish Ca-sulphate minerals (Figure 9A).Furthermore, a calcite sparitic vein containing disseminated patches of fibrous celestine (identified by comparing with Lerouge et al., 2014 andMazurek andde Haller, 2017; no mineralogical data available; Figure 9A,B) occurs perpendicularly to the bedding within the upper part of the interval.According to Matter et al. (1988), this interval represents the stratigraphical top hardground of the Opalinus Clay.
From 554.50 to 553.16 m, the 'Murchisonae-Oolith Formation' displays a rust-red coloured ooidal ironstone (KL4).It is rich in quartz sand, bioclastic debris (mainly echinoderms and mollusc shell fragments) and Fe-coated (limonitic) grains (including true Fe-ooids).Few intraclasts occur.Some parts show more Fe-rich and Al-rich clay minerals.The matrix is mainly sparitic and consists of calcite and dolomite (possibly ankerite) crystals.This overall facies is highly bioturbated.
According to Bläsi et al. (2013), the OPA/UBU-delineating horizon forms the basal layer of the 'Murchisonae-Oolith Formation'.It is overlain by an argillaceous mudstone resembling the Opalinus Clay facies (KL1), albeit with a possibly higher carbonate content (539.60-539.00m).A sharply delineated sandy bioclastic bed (KL2) occurs between 539.24 and 539.18 m.According to ICP-MS results from Riniken, the averaged major elements constituting the overall lithology are: Si (16.84%),Ca (13.68%), Fe (6.76%), Al (4.54%), S (1.36%), K (1.14%), Mg (1.00%) and Ti (0.30%), with strong differences depending on key lithologies.The remaining elements show values up to a maximum of a few hundred parts per million on average, with Mn as the highest minor element (ca 780 ppm on average).The ICP-MS results for all detectable elements can be found in Table S1.
The sampling methodology for ICP-MS measurements does not allow for direct comparison with XRF-CS data and cannot therefore serve for robust calibration, principally because of the heterogeneity of the sampled bulk volumes.It is particularly true for elements showing high lateral variability.Nevertheless, the most abundant elements show noteworthy and reliable correlations supporting the relative trends in the XRF-CS data (Figure 10).Rigorous quantification of absolute values is not required within the scope of the present study; the relative trends are sufficient for chemostratigraphic purposes, unravelling major lithological boundaries, and to disentangle key depositional and diagenetic processes.
Comparisons between elemental variations measured with the Mo-tube and the Cr-tube in the Riniken core support the reliability of the XRF-CS data sets.Correlation coefficients are higher than 0.75 for all major elements (not provided for Mg; see Table S2).The siliciclastic chemofacies is typically associated with KL1 and partly with KL2.This chemofacies shows a relative enrichment in terrigenous elements, such as Al, Si, K, Ti and Zr.These elements have a negative correlation with the carbonate fraction (Ca; Figure 13A).They are primarily present as building blocks or substituted within the clay minerals, mica, feldspar and quartz lattices.The elements correlate well (Figures 13B,C,D and 14), although small differences exist between them.The occurrence of thin silty/sandy lenses within a predominantly argillaceous matrix, such as within the Opalinus Clay mudstone facies (KL1), is expressed in comparatively higher Si/Al values, which are attributed to the higher quartz over clay minerals content.Hence, within a given lithology, (i.e.KL1 and KL2), the Si/Al ratio along with the Zr/Al ratio may be used to identify subtle grain-size variations (see also Calvert and Pedersen, 2007).Good correlations between Si and Zr are also evident in biplots and  13F) is atypical of the siliciclastic chemofacies and reflects rather the presence of gypsum encrustations on the core surface, disturbing the primary depositional signal of the mudstone facies.
The calcareous chemofacies is identified by the occurrence of dispersed carbonate nodules (KL3; see Mont Terri and Riniken), calcareous hardgrounds (KL3; OPA/UBUdelineating horizons in all studied successions) and spathic limestone beds (KL2; e.g. the Passwang Formation at Mont Terri).This chemofacies is evidence of a significant increase in Ca.While the Ca/Al ratio acts as the main proxy for all types of carbonates (bioclastic and authigenic/diagenetic), the data suggest that strong Mn enrichment supplements the Ca peaks within carbonates of diagenetic origin (nodules and cements).This assumption is supported by µ-XRF mapping displaying high Mn counts within calcite cement and authigenic phases, whereas bioclasts show low Mn intensities (Figure 7).Hence, the two spathic limestone beds within the Passwang Formation at Mont Terri display a significantly lower Mn increase in contrast to the strong Ca increase observed within these two intervals (Figure 15).
Where carbonate intervals are enriched in siderite (e.g.nodules within the Riniken Transition Zone), in Fe-hydroxides and Fe-silicates (typically within Fe-ooids) or in pyrite (e.g.pyritised burrow at Benken), the Fe-related ratios display high values.Pyrite-rich intervals are supplemented by a significant S/Al increase, as pyrite is the main S-bearing phase within the Opalinus Clay (Pearson et al., 2003).Other S-carriers include the late-diagenetic Ba-Sr-sulphates (fibrous celestine; Lerouge et al., 2014;Mazurek and de Haller, 2017), such as identified in the OPA/UBU-delineating horizon at Weiach.The fibrous celestine is nevertheless very limited in extent (restricted to a vein) and displays significant Ba and Sr peaks, although not illustrated within the framework of this study.Interestingly, the OPA/UBU-delineating horizons are associated with higher Zr values compared to other terrigenous elements (Figures 11 and 12).The Mont Terri and Benken successions exhibit Zr/Al peaks at the base of the calcareous hardgrounds, where the OPA/UBU boundary was set by Hostettler et al. (2017) and Bläsi et al. (2013), respectively.In the Weiach succession, the main Zr/Al peak lies at the top of the calcareous unit, where Matter et al. (1987) set the OPA/ UBU lithostratigraphic boundary.At Riniken, the Zr/Al ratio exhibits an enrichment within the overall Transition Zone, before drastically decreasing within the upper bounding unit.In some sections, such as Mont Terri and Benken, the Ti/Al ratio shows similar patterns along the OPA/UBU transition, although with lower magnitude.
The ferruginous chemofacies dominates within the rust-red coloured lithology of the upper bounding unit at Riniken and Weiach (KL4).It is generally rich in Fe and displays a relative increase in Ca compared to the Opalinus Clay mudstone facies.This observation is supported by the mineralogical composition of KL4, which is dominated by Fe-oxyhydroxides, Fe-silicates and carbonate cements (see Figure 8).The Fe/Al and Ca/Al profiles are anti-correlated, suggesting their respective distribution within distinct mineralogical phases (see Figures 7 and 8).At Riniken, the presence of dolomitic cement is revealed by high Mg/Al; Mg was only identified in sections screened with the Crtube.Within this chemofacies, Al displays a trend similar to Fe, and is comparatively enriched with respect to other elemental proxies for clay minerals (see e.g. the Ti/Al record).The relative Al enrichment combined with the Si decrease (fewer quartz grains, but generally larger) within KL4 is the main reason why the Si/Al ratio cannot be used as a proxy for grain size across the OPA/UBU transition at Riniken and Weiach, but only within similar lithologies.Within the ferruginous chemofacies, Mn/Al correlates positively with Ca/Al, suggesting the preferential incorporation of Mn into carbonate cements, whereas Mn/Al correlates predominantly with Fe/Al within the siliciclastic chemofacies, and thus with the fine siliciclastic fraction.The S content is particularly low within the ferruginous chemofacies, which agrees with the very rare pyrite within KL4.
To summarise, the XRF-CS data identify different element associations (chemofacies) depending on key lithologies and on their constituent mineralogical fractions.The (a) siliciclastic, (b) calcareous and (c) ferruginous chemofacies can be identified by comparatively high values of (a) terrigenous elements (Si/Al, Ti/Al, Zr/Al, etc.), (b) Ca/Al (and often Mn/Al) and (c) Fe/Al, respectively.The Opalinus Clay mudstone facies and the 'Murchisonae-Oolith Formation' in Benken (KL1) are characterised by the siliciclastic chemofacies, while some heterogeneities display a characteristic calcareous chemofacies (KL3).The Passwang Formation at Mont Terri (KL2) exhibits the mixed-influence of the siliciclastic and the calcareous chemofacies, while the Passwang The PCA results (Figure 14) illustrate the main correlations among the plotted elements.They show that within siliciclastic-dominated successions, such as those at Mont Terri and Benken, Al, K and Ti correlate strongly together, while Si and Zr differ slightly.This difference is related to their distribution within, respectively, the clay versus the silt/sand fractions, and thus highlights the small grain-size variations (lenticular bedding) exhibited by the downcore elemental profiles.In these successions, Fe is mainly associated with the fine-grained fraction, suggesting a common source.In PCA graphs, Mn plots in-between Ca and Fe (especially at Mont Terri).This feature results from the fact that Mn correlates with Fe within fine-grained, siliciclastic lithologies, while it correlates with Ca within carbonate-rich lithologies.Generally, the Riniken and Weiach successions show similar PCA results.The upper bounding unit (mainly KL4 in these successions) is characterised by high Fe, Mn and Ca, and is negatively correlated with the Opalinus Clay mudstone facies and its associated detrital element (Al, K, Ti, Si and Zr).Sulphur is poorly correlated with other elements and shows no particular relationship with specific lithological fractions.However, at Mont Terri, S correlates well with Fe, suggesting incorporation into dispersed pyrite framboids.In the other successions, the low positive correlation of S with Ca and Zr points to their co-enrichment within specific intervals, such as hardgrounds (i.e.OPA/UBU-delineating horizons).Overall, the PCA results highlight the differences between the distinct lithologies, and underline the influence of grainsize and mineralogical variations on elemental composition.Within this study, the OPA/UBU lithostratigraphic boundary was considered as defined in the current literature (Bläsi et al., 2013;Hostettler et al., 2017;Matter et al., 1987;1988).However, taking into account the present lithological and geochemical data sets, a slightly different definition of the OPA/UBU boundary can be suggested: that is, the OPA/ UBU-delineating horizon could be consistently attributed to either the Opalinus Clay, such as at Riniken and Weiach, or to the upper bounding unit, as at Mont Terri and Benken, allowing for a more homogeneous delineation of these two units.

Mont Terri
From stratigraphic base to top (Figure 11A), the first geochemical heterogeneity (12.21-12.26m) within the Opalinus Clay mudstone facies consists of a calcareous horizon corresponding to a calcitic nodule, inferred from the congruent Ca/Al and Mn/Al peaks (calcareous chemofacies) and confirmed by petrographic analyses.Such calcitic concretions are frequent within the upper part of the Mont Terri Opalinus Clay and are commonly interpreted as products of early diagenesis (Bläsi et al., 1991;Hostettler et al., 2020;Liniger, 2016).Manganese, within the Opalinus Clay mudstone facies, generally correlates with Fe.These two elements are known to exhibit a similar pattern of redox cycling and to be reduced within the suboxic zone (Froelich et al., 1979;Tribovillard et al., 2006).Yet, in the presence of abundant S and reducing conditions, Fe sulphides form, while Mn either migrates back to the water column, or is incorporated into authigenic carbonates (Tribovillard et al., 2006).
The next Ca/Al and Mn/Al peak along the Mont Terri succession (13.00-13.11m) reflects the basal intraclastic hardground of the Passwang Formation (OPA/UBUdelineating horizon).The constituent intraclasts are formed mostly by early diagenetic carbonate phases (i.e.micritic/ sparitic calcite cement), showing a similar Mn/Al enrichment.In contrast to the previous Ca/Al and Mn/Al heterogeneity, this horizon exhibits a congruent Fe/Al increase, which is attributed to the presence of Fe-ooids, oncolitic ferruginous microbialites and limonitised components.Burkhalter (1995) associated the formation of Fe-ooids and ferruginous microbialites within the Passwang Formation with periods of reduced sedimentation, or non-deposition, related to sealevel changes.He suggested that the OPA/UBU transition comprises a sedimentary discontinuity formed in a water depth between storm and fair-weather wave base, characterised by the subtle equilibrium between erosion and sedimentation (sediment bypassing).Turbulent conditions and sediment bypassing may also be inferred from the striking Zr/Al peak at the base of the hardground horizon.Hence, erosive currents might have remobilised the lighter sediment fraction (e.g.quartz grains and clay minerals; see also Si/Al to compare) while heavier minerals such as zircon show a relative increase (Craigie, 2018).Petrographic observations within the hardground show an upwards increase in the size of intraclasts, probably related to the increase in hydrodynamic conditions.Eventually, a period of non-deposition at the sediment-water interface started.This bypassing sedimentary discontinuity induced prolonged stable geochemical conditions within the sub sea floor.Intensive sulphate and iron reduction within the sediment probably explain the higher pyrite content (Kasten and Jørgensen, 2000) revealed by the S/Al peak at the base of the hardground.Enhanced production of alkalinity during sulphate reduction also explains simultaneous carbonate precipitation (Curtis, 1977;Wetzel and Allia, 2000).Subsequent lithification of the reworked carbonate components (intraclasts and biolithoclasts) on the sea floor could then occur.
The Si/Al profile reflects a general increase in average grain size, starting at the onset of the Passwang Formation.Silicon is enriched relative to Al, where the amount of quartz grains increases.A congruent increase in calcitic material is inferred from the Ca/Al ratio.Within this interval, Mn follows the Ca trend, caused by the incorporation of Mn within the authigenic calcitic phases cementing the clastic components throughout this lithology.However, the contrasting Ca and Mn signals displayed by the single element profiles within the sandy spathic limestone beds (Figure 15) are associated with a change in composition, and in particular to the relative amount of authigenic versus bioclastic carbonates.Unlike carbonate cements, bioclasts have low Mn concentrations, while both exhibit high Ca intensities (Figure 7).In turn, these two elements may serve to distinguish between primarily shortterm accumulations (i.e.bioclastic storm-deposit) and comparatively long-term, diagenetic deposits (i.e.hardgrounds and nodules), enabling process-based correlations.

| The OPA/UBU transition at Riniken
The lenticular bedding within the Opalinus Clay mudstone facies is highlighted by small-scale variations in the Ca/Al and Si/Al ratios (Figure 11B).A more important lithological change is identified by a fairly distinct chemofacies within the Transition Zone (331.90-331.14m).The XRF-CS data point to a relative increase in carbonate (Ca/Al), Fe sulphides (S/ Al) and heavy minerals (Zr/Al) within this zone.The Fe/Al and Mn/Al ratios display several peaks that can be linked to F I G U R E 1 5 Calcium and Mn profiles (in total counts and as ratio) in the Mont Terri core.Type of carbonate beds can be distinguished.The higher the Ca/Mn ratio is the more bioclastic content it shows.Within the Passwang Formation, the positive excursions in the Ca record are not mirrored by the Mn record.The OPA/UBU-delineating horizon is highlighted by the blue bracket.OPA = Opalinus Clay; P Fm = Passwang Formation the occurrence of sideritic nodules, and/or horizons of Feoxidation during core storage (mainly from siderite or pyrite).The combination of these proxies and petrographic observations point to a period with slower sedimentation rates, marking perhaps the onset of a regression, as postulated by Burkhalter (1995;1996).An increase in transport energy can be inferred from the relative enrichment in bioclastic debris and heavy minerals, while the presence of several diagenetic concretions and a higher pyrite content are both hints for longer residence time within stable geochemical conditions, typically within the sulphate reduction zone (Wetzel and Allia, 2000).
The erosive discontinuity (331.34 m) within the Transition Zone marks the beginning of a condensed, sideritic interval, which comprises sedimentary signs of intensive erosion and reworking (intraclasts, reworked nodules and coarse bioclasts).Siderite precipitation requires reducing conditions, high concentrations of reactive Fe, high alkalinity and negligible sulphide in pore waters (Berner, 1981;Taylor and Macquaker, 2011).Lerouge et al. (2014) suggested that sideritic concretions within the Opalinus Clay formed close to the sediment/ water interface, at a fluctuating boundary between the suboxic and sulphate reduction zones, and that disseminated siderite precipitated during bacterial methanogenesis and associated production of alkalinity following consumption of HS -by Fe sulphide precipitation.Within this condensed interval pyrite is scarce to absent (excluding intraclasts), an observation supported by the low S/Al.The sideritization event might therefore have occurred before reaching the sulphate reduction zone.Aller et al. (1986), for example, demonstrated that within environments with abundant Fe, siderite might have already precipitated during suboxic diagenesis (also referred to as non-sulphidic-post-oxic sensu Berner, 1981).In fact, the abundance of Fe-oxides might inhibit or postpone sulphate reduction (Chapelle and Lovley, 1992;Lovley and Phillips, 1987), and hence provide prolonged suboxic, non-sulphidic conditions for siderite precipitation.Taylor and Curtis (1995) suggested a similar process for the formation of sideritic ironstone from the Lower Jurassic of England.However, within the scope of this study, it is not possible to exclude the precipitation of siderite during methanogenesis.
The sideritic condensed interval is topped by thin, successive, calcareous, Fe-ooidal beds (high Ca/Al; layers a to d in Figure 7), which are considered by Matter et al. (1987) to be the lithostratigraphic top of the Opalinus Clay.The individual layers, which display small differences in composition, were most probably formed as successive events of synsedimentary lithification of bioclastic debris on the sea floor, associated with sediment starvation (indicated by the presence of drusy, isopachous and microsparitic/micritic cements; Wilson and Palmer, 1992).They mark the top of the Transition Zone and its associated chemofacies, and hence, of the postulated sedimentary discontinuity (Burkhalter, 1995).Layer d shows abundant well-rounded Fe-ooids and coarse bioclastic fragments, which may indicate sediment bypassing under turbulent conditions.Prolonged stable geochemical conditions can be inferred by the high degree of pyritization of the ooids present within the top layer.
The onset of the Passwang Formation is marked by a change in sedimentation regime and sediment type.The Ferich lithology points to a markedly different depositional environment, possibly related to increased input of weathered lateritic soil (see also Burkhalter, 1995).This conclusion is confirmed by the enrichment of low-solubility and mostly incompatible elements Fe and Al, relative to other detrital elements typically present within the Opalinus Clay (Si, Ti, K, etc.).The Si/Al ratio, in particular, shows an unambiguous decrease at the onset of the Passwang Formation that supports a lateritic origin.Deeply weathered lateritic soil profiles are commonly characterised by intense chemical depletion of soluble elements such as Si and K, and residual enrichment of less soluble elements such as Fe and Al often in the form of Fe and Al-oxyhydroxides and clay minerals (typically kaolinite; Tardy and Nahon, 1985).The drastic Mg increase within the Fe-rich lithology is mostly related to the occurrence of dolomite (ankerite) cement.
The basal Passwang Formation is classically interpreted as a depositional environment characterised by predominantly oxic conditions, notably due to the presence of abundant goethite/limonite and to a high level of bioturbation (Bläsi, 1987;Burkhalter, 1995;1996;Matter et al., 1987).Its lowermost cross-bedded unit (uppermost part of the slab on Figure 7) suggests deposition within energetic bottom-water currents.Sediment deposition may occur as sand-wave complexes, such as those interpreted for the Minette ironstone from Luxembourg and Lorraine (Teyssen, 1984).The Minette ironstone shows a similar enrichment in Fe and Al, as well as Cr and V (also observed within this study; Siehl and Thein, 1989).The highly bioturbated lithology of the upper part obscures any sedimentary structures, and might also be responsible for the limited organic matter and the very low pyrite content, such as highlighted by the low S/Al ratio.

| The OPA/UBU transition at Weiach
Core observations and the tight correlation between Ca/ Al and S/Al in Figure 13 suggests that the Opalinus Clay has been strongly affected by alteration during core storage.The precipitation of gypsum, as an alteration product, is common on Opalinus Clay cores acquired a long time ago, tunnel faces and outcrops (Mäder and Mazurek, 1998).It reflects the oxidative breakdown of pyrite (or siderite) and dissolution of calcite (and dolomite) by sulphuric acids (Mäder and Mazurek, 1998).It typically forms in small fractures induced by unloading and partial transient desaturation.
At Weiach (Figure 12A), in contrast to Mont Terri or Benken, the OPA/UBU-delineating horizon (554.82-554.50m) is stratigraphically attributed to the Opalinus Clay (Bläsi, 1987;Bläsi et al., 2013;Matter et al., 1988).However, from a geochemical perspective, it shares similarities with the OPA/UBU-delineating horizons at Mont Terri and at Benken expressed by the significant increase in Ca, Mn, and partly S and Fe.The Weiach interval is nonetheless three times thicker.Two main carbonate beds are separated by a more siliciclastic interval expressing a fining-upwards sequence, as suggested by the Si/Al and Ti/Al ratios.This overall interval can be related to two successive periods of sediment starvation and in-situ lithification of the sea floor (see also two main S/Al peaks indicating periods of stable geochemical conditions within the carbonate beds).The position of the highest Zr/Al peak suggests that the main episode of prolonged sediment bypassing lies at the top of the calcareous interval, where Bläsi (1987) and Matter et al. (1988) set the OPA/UBU lithostratigraphic boundary.These authors postulated that the calcareous unit marks the top horizon of an overall coarsening-upwards sequence, or a so-called 'Dachbank' cycle (see also Wetzel and Allia, 2003), formed by hardening (and reworking) of the top Opalinus Clay sediment.Increasing hydrodynamics and associated reduced sedimentation rates within this interval can be supported by increasing trends of, respectively, Zr/Al (heavy minerals accumulation) and Fe/Al.Iron enrichment, for instance as pyrite, may point to suboxic and/or anoxic conditions within the sub sea floor.Part of the Fe increase may nonetheless be attributed to the higher content of Fe-ooids within this interval.
The lithology and chemofacies of the basal 'Murchisonae-Oolith Formation' exhibit strong similarities with the basal portion of the Passwang Formation at Riniken.A similar mature sediment, composed mainly of the erosive product of continental laterite weathering profiles, can be postulated (Tardy and Nahon, 1985).High-energy bottom-water currents at the base of this unit are, however, not always highlighted (absence of cross-bedding), in contrast to the sequence at Riniken.This sediment type (KL4) differs from its lithostratigraphic equivalents in Mont Terri and Benken, which could be explained by differences in sediment source and/or related to sub-basin topography.The small regional extent of KL4 at the base of the upper bounding unit could point to a single, regional Fe-ooidal sedimentary complex covering the area of Riniken and Weiach, resembling the sand-wave structures interpreted further north, within the South German Basin (Bayer and McGhee, 1985) and in Luxembourg and Lorraine (Teyssen, 1984).

| The OPA/UBU transition at Benken
The succession at Benken (Figure 12B) mostly resembles the Mont Terri transition.Few S/Al, Fe/Al and Mn/Al peaks points to very limited heterogeneities within the Opalinus Clay mudstone facies.They correspond to pyritic and altered sideritic horizons.The uppermost S/Al peak within the Opalinus Clay highlights pyritised burrow traces occurring below the OPA/UBU-delineating horizon (539.70-539.60 m) of the 'Murchisonae-Oolith Formation'.Pyritization might have occurred through early diagenetic replacement of the organic matrix left by the burrower (Schieber, 2002).Abundant infaunal benthos is also consistent with a prolonged sedimentation break at the top of the Opalinus Clay, and the establishment of a benthic hardground community (Wilson and Palmer, 1992).As for Mont Terri, the main heavy minerals enrichment peak (Zr/Al, Ti/Al; Ti being also strongly associated with ilmenite and rutile; Craigie, 2018) occurs at the base of the hardground (OPA/UBU-delineating horizon).The latter shares strong geochemical similarities with the Mont Terri hardground, which all point to similar formation conditions, that is, lithification of reworked intraclastic material during a period of prolonged stable geochemical conditions related to increased hydrodynamics.
Visually, the overlying basal 'Murchisonae-Oolith Formation' differs little from the Opalinus Clay mudstone facies, which was the main reason they were formerly included within same lithostratigraphic unit (Bläsi et al., 2013;Nagra, 2001).The chemofacies, however, reveals slightly coarser material (Si/Al, Zr/Al) within the 'Murchisonae-Oolith Formation' and a higher pyrite content (S/Al).The PCA (Figure 14) provides a further argument for identifying two distinct lithologies, although less differentiated than the equivalent formations within the other core sections.The Ti/ Al and Zr/Al ratios display a strong enrichment a few centimetres above the hardground, which is consistent with higher transport energy and sediment bypassing through the OPA/ UBU transition.

| Genesis of the OPA/UBU transition
While each location has its particularities, a major change in sediment composition and associated chemofacies can be observed at all studied locations throughout the OPA/ UBU transition.The trend towards more carbonate-rich lithologies is also known from the literature (Bläsi, 1987;Burkhalter, 1996;Burkhalter et al., 1997;Hostettler et al., 2017;2020;Wetzel and Allia, 2003).The relative increase in grain-size and bioclastic content points to a depositional environment influenced by shallower water depths and/ or an increase in current strength.Sediment bypassing suggested by petrographic and geochemical proxies indicates a period of enhanced hydrodynamic activity during the OPA/ UBU transition.The sudden lateritic signature of the upper bounding unit at Riniken and Weiach highlights a significant change in sediment source and type, which probably resulted from combined hydrodynamic and palaeoclimate factors (Burkhalter, 1995;Teyssen, 1989).The change in sediment provenance and related hydrodynamic conditions was proposed by Burkhalter (1996), who suggested based on quartz grain-size analysis that terrestrial sediment delivery changed from a (north)west-(south)east direction during the deposition of the Opalinus Clay, to a north(west)-south(east) direction during the deposition of the basal Passwang Formation.While relationships between sediment type, provenance and distribution, and their link with hydrodynamic settings and palaeoclimate are difficult to assess within the scope of this work, it appears that the OPA/UBU boundary falls within a period of major tectonic, hydrodynamic and palaeoclimate perturbations across the epicontinental Central European Basin (Mid-Cimmerian events; de Graciansky and Jacquin, 2003;Underhill and Partington, 1993).The early Aalenian was tectonically active, notably due to the opening of the Atlantic and Tethys oceans (Lemoine et al., 1986;Ziegler, 1988), which had a major effect on the sea floor topography, as well as on sediment deposition (Durlet et al., 1997;Teyssen, 1989;Wetzel et al., 2003).Moreover, a major doming event uplifted the area around the North Sea (Underhill and Partington, 1993), which subsequently modified the overall circulation patterns across the epicontinental Central European Basin.In turn, a generally cooler period was initiated, due in particular to the obstruction of a northward flowing current through the Viking Corridor and consequently reduced heat transfer from the Tethys towards the Boreal Ocean (Korte et al., 2015;Price, 2010).

| XRF-CS for unravelling the formation of sedimentary discontinuities: example of the OPA/UBU transition
The use of XRF-CS as a tool for delineating stratigraphic sequences within Mesozoic rock formations is relatively novel.For instance, Thöle et al. (2020) applied the method to drill cores covering the Berriasian to Aptian interval within the Lower Saxony Basin (northern Germany).They identified several elemental ratio proxies allowing them to correlate proximal and distal basinal environments, and to characterise the studied successions in terms of grain-size and carbonate content variations using Si/Al and Ca/Ti, respectively.The authors used continuous XRF-CS records to identify transgressive and regressive systems tracts, being a suitable tool for establishing sequence stratigraphic frameworks.The present study differs from the work of Thöle et al. (2020), as it focusses on a limited stratigraphic interval, and on the detailed characterisation of one particular sequence boundary (i.e.discontinuity surface).It shows nonetheless that similar proxies can be used to trace variations in grain-size and carbonate content (Si/Al, Zr/Al and Ca/Al).It reveals moreover that element ratios have the potential to detect sedimentary discontinuities and to identify their formation processes, such as illustrated by the combination of Zr/Al, Ti/Al, S/Al, Ca/Al and Mn/Al ratios within the four case studies.
The hypothesis of using geochemical proxies for determining differential accumulation of heavy minerals to trace sedimentary discontinuities (particularly Zr/Al peaks) is supported by petrographic arguments, notably the congruent presence of hardgrounds and typically high-energy deposits (i.e.intraclasts, Fe-ooids) that are classically linked to erosive processes or sediment bypassing within similar successions (Burkhalter, 1995;Teyssen, 1989;Wetzel and Allia, 2000), and the presence of petrographic and geochemical arguments inferring prolonged, stable geochemical conditions within the sub sea floor.The method can hence be used to differentiate among different types of sedimentary discontinuities.Burkhalter (1995) differentiated three types of discontinuity surfaces (comparable to third-order sequence boundaries; Van Wagoner et al., 1988): (a) regressional discontinuities, (b) transgressional discontinuities and (c) omissional discontinuities.Regressional discontinuities separate low-energy highstand deposits from lowstand deposits during regressive phases.They are formed by sediment bypassing or erosion, when the rate of sea-level fall equals or exceeds the rate of sea-floor subsidence.Transgressional discontinuities separate high-energy lowstand deposits from transgressional deposits.They are generated by non-deposition caused by sediment bypassing at the onset of sea-level rise.Omissional discontinuities correspond to maximum flooding surfaces and are formed by sediment starvation.
The geochemical characterisation of the OPA/UBU transition suggests a period of sediment non-deposition due to sediment bypassing during an overall regressive trend, as supported by the lithofacies assemblages and associated chemofacies (i.e. increase in coarser terrigenous and bioclastic content).The results are thus generally in accordance with Burkhalter (1995;1996) and Burkhalter et al. (1997), who interpreted the OPA/UBU boundary as a widespread regressional discontinuity during the early Aalenian.The authors suggested that this boundary marks the top of the low-energy highstand deposit (the Opalinus Clay mudstone facies; KL1), at the onset of a relative sea-level fall, when water energy was at its maximum and prevented sediment deposition (i.e.sediment bypassing), before the overlying lowstand deposit (the upper bounding unit) started to accumulate.The limited vertical extent of the investigated intervals, however, does not allow for confirmation, or rejection, of the interpretation that the OPA/UBU transition reflects a single event throughout northern Switzerland (Burkhalter, 1996) rather than more localised processes related to the palaeogeographic and topographic setting.Similar geochemical signals can be expected within any intra-formational hiatus formed by comparable bypassing processes (but probably of lower intensity).
An overall regressive trend during the early Aalenian is nonetheless suggested from other locations across the epicontinental Central European Basin (Callomon, 2003;de Graciansky and Jacquin, 2003;Sandoval et al., 2002;Teyssen, 1989;Zimmermann et al., 2015;2018), supporting the interpretation of Burkhalter (1996).Zimmermann et al. (2015) proposed a comprehensive framework of second-order, third-order and fourth-order Lower and Middle Jurassic sequences in the North German Basin.They evidenced the occurrence of one third-order sequence (Aal1) and two fourth-order sequences (Aal1a and Aal1b) of predominantly Aalenian age.The third-order sequence started during the Levesquei Zone (beginning of the transgressive phase) and lasted up to the Bradfordensis Zone (end of the regressive phase), with the maximum flooding which lasted about 4.3 Myr taking place within the early Aalenian (Opalinum Zone) Opalinuston shales (i.e.Opalinuston Formation sensu Franz and Nitsch, 2009).It is suggested that the fourth-order sequence (Aal1a) began simultaneously with the onset of the third-order sequence, but terminated within the early Murchisonae Zone, reflecting the maximum flooding within the Opalinuston Formation.Closer to the study area, northwards and westwards from Mont Terri (i.e.Burgundy and Alsace, France), the upper Toarcian and overall Aalenian successions are thin to strongly condensed, or absent in certain locations representing possibly emerged areas (Contini, 1970;Durlet and Thierry, 2000;de Graciansky and Jacquin, 2003;Schirardin, 1960).The scarcity and absence of late Toarcian and Aalenian deposits are interpreted as the result of an overall regressive trend, reflecting the shallow, high-energy water conditions and the associated sediment non-deposition, on an area that used to be a topographic high bordering the Paris Basin, the Burgundy High ( de Graciansky and Jacquin, 2003;Durlet and Thierry, 2000).
Considering the above-mentioned context of a major regressive trend across the epicontinental Central European Basin during the early Aalenian, the OPA/UBU transition can be interpreted as a fourth-order regressive phase, superimposed by an overall third-order regressive half-cycle, where the OPA/UBU boundary reflects the regressional discontinuity of the earliest Aalenian, fourth-order, sea-level fall (Aal1a sensu Zimmermann et al., 2015).The overall deposition of the Opalinus Clay and its upper bounding unit is however more complicated, as it was strongly influenced by differential subsidence (Burkhalter, 1996;Burkhalter et al., 1997;Wetzel et al., 2003;Wetzel and Allia, 2000;2003;Wetzel and Meyer, 2006).Deeper depocenters (topographic lows) formed when the subsidence rate exceeded the eustatic sealevel fall, creating different sub-basins within the depositional area.These depocenters were situated below the storm-wave base, enabling the accumulation of comparatively more parasequences than the neighbouring topographic highs.On the other hand, the highs were situated above the storm-wave base and affected by storm-induced, erosive currents that were responsible for the establishment of sedimentary discontinuities (Burkhalter et al., 1997;Wetzel and Allia, 2000;2003).Depending on location, it resulted in the accumulation of varying numbers of successive high-order parasequences (Bläsi, 1987;Bläsi et al., 2013;Burkhalter, 1996;Burkhalter et al., 1997).It might also explain parts of the diachronous, biostratigraphic framework of the OPA/UBU transition (Bläsi et al., 2013;Hostettler et al., 2020;Wohlwend et al., 2019).Hence, the Mont Terri Opalinus Clay succession exhibits two major coarsening-upwards sequences that can be interpreted as two fifth-order parasequences during the Opalinum Subzone (Lauper et al., 2018).Further to the east, the Opalinus Clay is characterised by one overall shallowing-upwards sequence (third order and fourth order), with higher order, 'Dachbank' cycles occurring within the upper part of the succession, and so, probably younger subsidence pulses provided the new accommodation space (Bläsi, 1987;Burkhalter et al., 1997;Wetzel and Allia, 2003).It has been shown from other locations across the epicontinental Central European Basin that differential subsidence played a major role in the formation of high-order parasequences which imprinted the overall eustatic cyclicity (Durlet et al., 1997;de Graciansky and Jacquin, 2003;Teyssen, 1989).It thus supports the interpretation provided for the Swiss northern basin (Bläsi, 1987;Burkhalter, 1995;1996;Burkhalter et al., 1997;Wetzel and Allia, 2000;2003).
Based on petrographic criteria alone, especially when investigating subsurface successions that are limited to the study of drill cores, it may be difficult to assess the genetic origin of sedimentary discontinuities (given that they can be identified as such), as well as to understand their significance in terms of sequence stratigraphy.It has already been shown that the use of elemental proxies can be convenient for identifying and characterising stratigraphic sequences within mudstone-dominated successions (Thöle et al., 2020;Turner et al., 2016;Ver Straeten et al., 2011).The present study nonetheless highlights the potential of XRF-CS as an additional and complementary tool for unravelling the depositional history of complex and mixed, carbonate-siliciclastic transitions.Discontinuity surfaces and sequence boundaries may be delineated and recognised among different discontinuity types and formation processes, and related to their sequence stratigraphic context.Furthermore, as several geochemical proxies are largely process-based, and in this case can be related to major events (i.e. a major sea-level fall and a longer period of sediment non-deposition), they may enable chemostratigraphic correlations within areas controlled by specific sedimentation regimes, even though they might be overprinted by differential subsidence and associated basin subdivision, and reflect distinct biostratigraphic constraints.
The study of the OPA/UBU lithostratigraphic transition suggests that XRF-CS is successful in identifying, characterising and delineating major lithological units and boundaries in a time-effective, continuous and non-destructive way.It provides additional objective and quantitative support to petrographic and lithostratigraphic studies.Furthermore, XRF-CS elemental proxies allow tracking of subtle grain-size and mineralogical variations that are difficult to identify by visual observations only.They help in deciphering complex depositional environments, allowing a better comprehensive understanding of depositional and diagenetic processes, the last being crucial for correlation studies in complex basins.
The combined use of detailed petrography and XRF chemostratigraphy in this study (a) provides evidence of the change in sediment type and sedimentation regime across the OPA/UBU lithostratigraphic transition, (b) identifies sediment bypassing as the major process responsible for the OPA/UBU sedimentary discontinuity and (c) disentangles the main depositional and diagenetic processes along the overall transition.
The combined use of sedimentary petrography and XRF chemostratigraphy, the latter providing high-resolution elemental variations and element peak-markers, is important for establishing process-based correlations at basin scale, within lithologically heterogeneous and biostratigraphically diachronous sedimentary successions.

F
I G U R E 4 Lithological detail of the OPA/UBU transitional intervals.The studied Mont Terri drill core has a smaller diameter than the Nagra drill cores, and was not drilled perpendicular to bedding.The light blue brackets identify the OPA/UBU-delineating horizons.The dark blue arrow points to the illustrated part of the Transition Zone (T.Z.; Riniken).The green brackets indicate the areas illustrated in Figures 5A, 6A, 7 and 9A,C.Fm = Formation F I G U R E 3 Lithological logs of the studied drill core sections.Dashline indicates the defined OPA/UBU lithostratigraphic boundary based on

F|
Hardground at Mont Terri.(A) Macroscopic view of the hardground delineating the OPA/UBU lithostratigraphic boundary within the BPE-3 drill core.The OPA/UBU-delineating horizon is attributed to the Passwang Formation (Hostettler et al., 2017).Smaller clasts occur at the base, while ferruginous microbialites occur mainly within the upper part.(B) Photomicrograph of ferruginous microbialites with mammilated texture occurring within the upper part of the hardground (from a neighbouring drill core; BPE-2).(C) Typical lithology of the hardground.Fe-ooids, bioclasts and quartz grains occur within a bioturbated, ferruginous, micritic matrix.Right side of the photomicrograph is stained (from BPE-2).(D) Stained photomicrograph showing a boring filled with Fe-spastoliths (plastically deformed Fe-ooids) and fine-grained material (from BPE-2).All photomicrographs were acquired in plane-polarised light Key lithologies along the OPA/

F
I G U R E 7 Photograph of the Riniken slab.The thin calcareous and Fe-ooidal layers topping the Opalinus Clay are referenced by letters (a, b, c and d; see text for descriptions).The numbers display the location of the μ-XRF scans (upper part of the figure).SEM-EDX images, on the right side, display the elemental mapping of neomorphosed ooids at the bottom of the slab (Opalinus Clay), and Fe-rich ooids at the top of the slab (Passwang Formation).Ap. = apatite; Cal.= calcite; Fe-Al-Silic.= ferruginous and aluminous silicates (e.g.chamosite, kaolinite); Feoxyhydr.= iron oxyhydroxides; Py. = pyrite; Qz = quartz; Sid.=siderite Quartz-rich lenses are rare, but a few thin elongated sideritic layers occur.At 539.74 m, near-horizontal burrow traces filled with coarse silty quartz grains and bioclastic debris are visible.
Downcore variations of selected element ratios are displayed in Figures11 and 12as a function of core depth for all studied core intervals.The combination of different element ratios allows the following three chemofacies to be identified: (a) siliciclastic, (b) calcareous and (c) ferruginous.

F
Lithological features from the Weiach and Benken cores.(A) The OPA/UBU transition at Weiach.The lower part consists of the OPA/UBU-delineating horizon (KL3), while the upper part illustrates the beginning of the limonitic lithology of the basal 'Murchisonae-Oolith Formation' (KL4).The calcitic vein with fibrous celestine is indicated by a blue arrow.Yellow arrow shows post-diagenetic alteration, probably Ca-sulphates.(B) Photomicrograph of a calcite/celestine vein from the OPA/UBU-delineating horizon at Weiach.Black arrows point to the fibrous celestine crystals.(C) The OPA/UBU transition at Benken.The lower part displays the OPA/UBU-delineating horizon (KL3), while the upper part illustrates the overlying argillaceous (-calcareous) mudstone of the 'Murchisonae-Oolith Formation' (KL1).(D) Photomicrograph from the OPA/UBU-delineating horizon in the Benken core.Calcitic neomorphosed ooids occur next to an intraclast within a ferruginous matrix.All photomicrographs were acquired in plane-polarised light PCA (Figures 13D and 14) for the Mont Terri and Benken successions, both of which are predominantly composed of siliciclastic sediments.In contrast, the Riniken and Weiach successions are more heterogeneous, and the Al and Zr contents depend mostly on mineralogical composition rather than grain size.Within this chemofacies, Fe/Al, Mn/Al and S/Al are generally low (except at Weiach, where S/Al is high due to alteration products; see below) and correlates positively with the fine siliciclastic fraction.The good correlation between Fe/Al and S/Al within the Opalinus Clay mudstone facies at Mont Terri points to dispersed pyrite framboids (Figure 13E).A few striking peaks may occur, typically when pyritic (Fe/Al and S/Al peaks) or sideritic (Fe/Al and Mn/ Al peaks) concretions are present (see e.g.KL1 in Benken sediments).In the Weiach core, relatively high and well-correlated Ca/Al and S/Al signals are recorded along the overall Opalinus Clay mudstone facies (KL1).Their relative enrichment and tight correlation (see Figure

F
Biplots and correlation coefficients of selected elements measured by ICP-MS (%) versus XRF-CS (in total counts, with the Motube), respectively | 15 LAUPER Et AL.F I G U R E 1 1 Downcore variations of selected element ratios and Al plotted against lithology.Chemofacies are highlighted by different colours.(A) The OPA/UBU transition at Mont Terri.(B) The OPA/UBU transition at Riniken.The OPA/UBU-delineating horizons are highlighted by blue brackets.The beginning of the Transition Zone (T.Z.) is highlighted by the blue dashed line * = Cr-tube; no indication = Mo-tube.OPA = Opalinus Clay; P Fm = Passwang Formation

|
Interpretation of depositional and post-depositional processesBesides showing great potential for geochemically defining key lithologies along the studied OPA/UBU successions, the F I G U R E 1 3 Biplots showing correlation trends among selected element ratios.(A) Calcium is anti-correlated towards lithogenic elements.(B) Good correlation among lithogenic elements is illustrated by K and Ti.(C) Good correlation among lithogenic elements is illustrated by Si and Ti.(D) Good correlation among lithogenic elements is illustrated by Si and Zr.(E) Occurrence of pyrite within the Opalinus Clay is evidenced by the good correlation between S and Fe.(F) Occurrence of gypsum at the surface of the Opalinus Clay is evidenced by the good correlation between Ca and S. Colours of data points correspond to lithology: black = Opalinus Clay mudstone facies; blue = OPA/UBU-delineating horizons (Mont Terri, Weiach and Benken) and the whole Transition Zone in the Riniken core; and red = upper bounding unit, respectively, Passwang Formation at Mont Terri and Riniken, and 'Murchisonae-Oolith Formation' at Weiach and Benkendowncore element ratios provide information on the depositional and post-depositional processes.

F
Principal component analysis (PCA) of selected elements (Si, Ca, Fe, Al, S, K, Mn and Zr) from each data set.The measured elements were treated using a centred log-ratio transformation.The graph axes differ from one PCA to the other.Colours of data points correspond to lithology: black = Opalinus Clay mudstone facies; blue = OPA/UBU-delineating horizons (Mont Terri, Weiach and Benken) and the whole Transition Zone in the Riniken core; and red = upper bounding unit, respectively, Passwang Formation at Mont Terri and Riniken, and 'Murchisonae-Oolith Formation' at Weiach and Benken.See text for interpretations | 19 LAUPER Et AL.