Cerium Anomaly at Microscale in Fossils - Analytical Chemistry (ACS

Page 1 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Cerium anomaly at microscale in fossils Pierre Gueriau,∗,†,‡ Cristian Mocuta,‡ and Loïc Bertrand∗,†,‡ IPANEMA, USR 3461, CNRS, ministère de la Culture et de la Communication, BP48 Saint-Aubin, F-91192 Gif-sur-Yvette, France, and Synchrotron SOLEIL, BP48 Saint-Aubin, F-91192 Gif-sur-Yvette, France E-mail: [email protected]; [email protected]

Abstract

Estimation of rare earth element (REE) concentrations in biogenic apatites (fossil bones, teeth and fish scales), known to be present in significant quantities in fossil bones and teeth (concentration of hundred ppm or greater), lies at the heart of many essential works aiming at inferring paleo-seawater chemistry, redox conditions as well as paleo-environmental and paleo-geographical reconstructions. 1–5 In particular, the so-called cerium anomaly – the assessment of the excursion in relative abundance in cerium compared to the series of REEs 6,7 – is considered as a central proxy to assess paleoenvironmental redox conditions, 8 within the series of redoxsensitive trace elements or ratios (including V/Cr and Cu/Zn) that have been used to decipher paleo-redox conditions. 9,10 By investigating the REE composition of the modern marine environment, German and Elderfield (1990) 8 showed that the measured Ce anomaly is related to the whole redox history of the particular water mass sample. In a review documenting REE distribution patterns in the sediment and sediment porewaters of various modern marine systems, Chen et al. (2015) 11 noticed a pervasive REE fractionation in sediment porewaters largely controlled by adsorption/desorption processes onto various mineral and organic phases, which is commonly influenced by porewater redox conditions. REE patterns measured in vertebrae and scales from deep-sea modern fish displayed an approximate record of the redox condition of contemporaneous seawater. 12 In marine vertebrate fossils REE concentrations have been pointed out to depend on REE uptake during post-depositional recrystallization and diagenesis, 3,5 characterized by a ‘bell-shaped’ REE pattern (a strong middle REE enrichment relative to light and heavy REEs) explained to result from an extensive recrystallization, 5,13 which may no longer preserve the early REE signal of the depositional environment. It led Morad and Felitsyn (2001) 14 to suggest a global oceanic anoxia during the Early Cambrian based on a Ce anomaly in apatite associated to a flat distribution REE pattern. Cerium is the most abundant REE in the Earth crust, and is not a ‘rare’ element stricto sensu as its abundance is comparable to that of transition metals such as copper, nickel or zinc. It displays a particular aqueous chemistry within the REE series as it can be present under the +III and +IV oxidation states, while the other REEs are primarily trivalent.

Patterns in rare earth element (REE) concentrations are essential instruments to assess geochemical processes in Earth and environmental sciences. Excursions in the ‘cerium anomaly’ are widely used to inform on past redox conditions in sediments. This proxy resources to the specificity of cerium to adopt both the +III and +IV oxidation states, while most rare earths are purely trivalent and share very similar reactivity and transport properties. In practical terms, the level of cerium anomaly is established through elemental point quantification and profiling. All these models rely on a supposed homogeneity of the cerium oxidation state within the samples. However, this has never been demonstrated, whereas the cerium concentration can significantly varies within a sample, as shown for fossils, which would vastly complicate interpretation of REE patterns. Here, we report direct micrometric mapping of Ce speciation through synchrotron X-ray absorption spectroscopy and production of local rare earth patterns in paleontological fossil tissues through X-ray fluorescence mapping. The sensitivity of the approach is demonstrated on well-preserved fishes and crustaceans from the Late Cretaceous (ca. 95 Million years (Myr) old). The presence of Ce under the +IV form within the fossil tissues is attributed to slightly oxidative local conditions of burial, and agrees well with the limited negative cerium anomaly observed in REE patterns. The [Ce(IV)]/[Ce(tot)] ratio appears remarkably stable at microscale within each fossil and is similar between fossils from the locality. Speciation maps were obtained from an original combination of synchrotron microbeam X-ray fluorescence, absorption spectroscopy and diffraction, together with light and electron microscopy. This work also highlights the need for more systematic studies of cerium geochemistry at microscale in paleontological contexts, in particular across fossil histologies.

∗ To

whom correspondence should be addressed

† IPANEMA ‡ SOLEIL

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Other notable exceptions are Eu and Yb (+II and +III). 15 This is due to cerium [Xe]4f1 electronic configuration in the trivalent state, Ce4+ being therefore in a stable electronic configuration, with a reasonably moderate standard reduc0 =+1.74 V. In sediments, the Ce anomaly is tion potential Ered related to the insolubility of Ce(IV) that forms under oxidizing conditions – such as CeO2 – at and near the Earth’s surface and thereby induces distinct transport, fractionation and deposition behaviors of cerium compared to purely trivalent REEs. 8,16–18 If the cerium anomaly is used as a standard tool to assess paleo-redox conditions, little is known about the variability of REE signatures at a sample scale. Recent studies have questioned the variability of REE compositions at the (sub) mm scale 19,20 and shown distinct point REE patterns, ratios and Ce anomalies within single bone samples. 20–23 Recently, Herwartz et al. (2013) 24 argued that the fractionation of cerium from the other REE can only be attained under oxidizing conditions; positive and negative Ce anomalies would therefore both point towards oxidized and probably alkaline conditions in proximal surface waters. Together with the observation of variable Ce anomalies over bone profiles, they contested the Ce anomaly in fossil bioapatite as a reliable tool for assessing redox – and especially reducing – conditions. All mentioned studies essentially rely on REEs point quantification and profiling. However, the homogeneity of the distribution of REEs within paleontological samples at micrometric scales has only attracted limited attention up to now. In addition, REE patterns do not directly provide speciation information but allow intercomparison of relative abundances across the series, pointing out to inter-elemental fractionations resulting from complexation with aqueous ions and/or preferential absorption or substitution of specific REE fractions by other minerals and particles. Elemental maps from laser ablation–inductively-coupled plasma–mass spectrometry (LA–ICP–MS) along a few sections of Cretaceous bones showed either distinct gradients in concentrations of rare earth elements and uranium, or intricate patterns related to bone histology. 19 We have recently demonstrated that the distribution of REEs in flat fossils can readily be imaged at microscale with synchrotron X-ray fluorescence (XRF) on entire fossils from the Cretaceous exhibiting both mineralized bioapatite (i.e. bones) and authigenic apatite that have replicated soft-tissues such as muscles. 25 Investigating the behavior of cerium at microscale is likely to provide critical information regarding (a) the length scale at which the cerium signal can be regarded as homogenous, (b) the variation in cerium oxidation state within distinct fossil tissues, and (c) the main physicochemical mechanisms driving the Ce anomaly at local and global level. Here, we report an approach to describe and study the cerium anomaly in fossils at microscale, based on a combination of synchrotron µXRF, µXANES and µXRD. We demonstrate that the oxidation state of Ce can be directly measured at fossil histological scale through point XANES. Speciation mapping through the collection of successive µXRF maps at excitation energies across the Ce L3 -edge allows reconstructing the distribution of the oxidation state of Ce at microscale in

Page 2 of 12

well-preserved fossils. Results highlight variations in cerium concentrations and speciation at fossil histological scales on the basis of speciation maps of the element.

METHODS Analytical developments described herein regarding measurement and imaging of the cerium oxidation state involved two complementary beamlines of the SOLEIL synchrotron. Using the microfocused beam (ca. 10 µm in diameter) of the DiffAbs beamline, which operates in the 3-19 keV energy range, we collected µXRF maps at an excitation energy of 17.2 keV optimized for excitation of fluorescence K-lines from aluminum to yttrium (13 < Z < 39) and L-lines from cadmium to lead (48 < Z < 82) including the whole REE series. The DiffAbs beamline also allowed performing µXRD mapping using this microbeam, and collecting macroscale XANES spectroscopy at the Ce L3 -edge using a degraded beam (ca. 300 µm in diameter). Microscale XANES spectroscopy and speciation mapping at the Ce L3 -edge were collected at the LUCIA beamline, which provide a microfocused beam (ca. 3 µm in diameter) in the 0.8-8 keV energy range. Synchrotron micro-X-ray fluorescence (µXRF) mapping. Synchrotron µXRF raster-scanning was performed at the DiffAbs beamline of the SOLEIL synchrotron source (Saint-Aubin, France). The X-ray beam was monochromatized (∆E/E≈10−4 ) using a Si(111) double-crystal monochromator and focused using bendable mirros in a Kirkpatrick-Baez (KB) geometry down to a size of ca. 11×7 µm2 (H×V, full width at half maximum – FWHM). The photon flux at this energy is estimated at 1.6×109 photons/s in the focused spot from a fit of the fundamental parameters on a fluorescence standard using the PyMca software. 26 The cross-sections were mounted on a xyz scanner stage, allowing ±12 mm movements with µm accuracy. The sample is orientated at 45° to the incident beam and at 45° to the XRF detector, in the horizontal plane (plane of the electron orbit in the synchrotron). Consequently, the isotropically emitted fluorescence is recorded with strongly reduced Thomson scattering (in theory null). The detector used was a photon-counting four elements silicon drift diode (SDD) detector for the shrimp C. berberus (Vortex ME4, Hitachi; total active area: 170 mm2 ), and a mono element SDD detector for all other samples (Vortex EX, Hitachi; 100 mm2 ). A 500 ms counting time was used to attain good statistics on counts from trace elements. µXRF elemental maps were obtained through a full spectral decomposition performed with the PyMca software 26 and corrections for matrix reabsorption using experimental parameters determined from microanalysis of a fluorapatite standard (AS1045-AB, SPI supplies), batch-fitting procedure, Pseudo-Voigt peak shape, and a polynomial baseline approximation. Resulting images are displayed in full range. Synchrotron X-ray absorption near-edge structure (XANES) spectroscopy. Sub-millimetric XANES spectroscopy was perfomed at the DiffAbs beamline at the Ce


2

Page 3 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


L3 -edge (electron binding energy: 5.723 keV) in the 5.700– 5.860 keV range using a Si(111) double crystal monochromator and a beam spot size of about 300×300 µm2 . Energy step sizes were 2 eV between 5.700 and 5.708 keV, 1 eV between 5.710 and 5.727 keV, 0.5 eV between 5.728 and 5.738 keV, 1 eV between 5.739 and 5.760 keV, and 2 eV between 5.762 and 5.860 keV. A 15 s counting time was used per energy step, except for the cerium oxide CeO2 reference sample (2 s). The incident beam intensity signal was measured using a 10 µm thick Si photodiode in transmission, while the fluorescence signal was monitored using a SDD (Vortex EX, total active area: 100 mm2 ). Microscale XANES spectroscopy was performed at the LUCIA beamline of SOLEIL. The X-ray beam was monochromatized using a Si(111) double-crystal monochromator and focused using KB mirrors downstream to a diameter of 3 µm2 (FWHM). Energy calibration was performed by setting the first edge inflection point of an iron foil at 7.112 keV. The photon flux at the energy of 5.900 keV (optimized for the excitation of the Ce L3 -line) was estimated using the procedure mentioned before to 2.8×1010 photons/s in the focused spot. The fossils were mounted on a small copper slab that can hold samples up to ca. 24×18 mm2 large, placed on a micrometric xyz scanner stage under primary vacuum. The sample was orientated at 85° to the incident beam and at 5° to a 4-element SDD XRF detector in the horizontal plane (QUAD SD XFlash Detector 5040, Bruker AXS; total active area: 40 mm2 ), so as to maximise spatial resolution while minimising contribution from scattering. µXANES spectra at the Ce L3 -edge were collected at room temperature in fluorescence mode in the 5.680– 5.860 keV range. Energy step sizes were 2 eV between 5.680 and 5.710 keV, 0.3 eV between 5.7103 and 5.746 keV, 1 eV between 5.747 and 5.790 keV, and 2 eV between 5.792 and 5.860 keV. A 1 s counting time was used per energy step and 2 spectra were cumulated. All XANES spectra were processed through self-written routines using the R statistical environment 27 by removing the pre-edge range background and the absorption background. In order to assess the contributions of Ce(III) and Ce(IV) species, and thus to determine [Ce(IV)]/[Ce(tot)] ratio, the spectra were deconvoluted into a linear combination of spectra from Ce(III)(NO3 )3 ·6H2 O and Ce(IV)(SO4 )2 standards previously measured at the LUCIA beamline by Janots et al. 28 Ce speciation maps were calculated from collection of successive XRF maps in the L3 -edge region at microscale at the LUCIA beamline (beam spot size: 3×3 µm2 FWHM). A 1 s counting time was used to attain good statistics on counts from trace elements. µXRF maps IEi (x, y) were collected at distinct excitation energies Ei corresponding to the edge (5.727 keV), major EXAFS oscillations (5.733, 5.736 and 5.765 keV) and the post-edge regions (5.900 keV). Subtraction at each pixel of an additional spectral map collected before the Ce L3 absorption edge (5.680 keV) clearly reveals the contribution of the characteristic Ce Lα1 line from other fluorescence lines in the same energy domain. Integration of the intensities of the corresponding spectral region of interest (Ce Lα1 ) in each map allowed reconstructing partial

Ce L3 -edge XANES spectrum at each pixel. The contributions of Ce(III) and Ce(IV) species were assessed by deconvoluting the low energy resolution spectra into a linear combination of XANES spectra from Ce(III)(NO3 )3 ·6H2 O and Ce(IV)(SO4 )2 standards. All the procedure was performed through self-written routines using the R statistical environment 27 and the SpectroMicro package. Synchrotron micro-X-ray diffraction (µXRD). µXRD measurements were performed at the DiffAbs beamline at the same energy and beam diameter than in µXRF. The sample was rotated around the vertical axis by 45° with respect to the incident beam, resulting in an effective on-sample beam footprint of ca. 15.5×7 µm2 (H×V, FWHM). XRD images were collected in transmission geometry using a hybrid pixel detector (XPAD3.2 29,30 ) in several images and covering a 3.5◦ –55.8◦ range in 2θ at 17.2 keV incident photon energy for a counting time of 10 s per image (2θ equivalent range at the Cu Kα energy: 7.5◦ –177.7◦ ). The detector is made of several flat tiles (chips) separated by small gaps. Details about detector geometry, calibration, correction of the images and diagrams reconstruction from the corrected images are available in Mocuta et al. 31 and references therein. Small maps were collected in raster-scanning mode (7×10 sample positions, step size: 100 µm). Diffraction diagrams were normalized to the incident beam intensity and the measured background due to sample environment was subtracted. Phase identification was carried out using the Match! software making use of the International Centre for Diffraction Data (ICDD)-PDF 2013 database. Elemental quantifications. ICP-MS and AES quantifications have been carried out at the Institut des Sciences Analytiques (UMR 5280 CNRS, Lyon, France) using ICPAES iCAP 6500 DUO and ICP-MS Q X7 (Thermo Scientific) for major elements and trace elements respectively, and at the Service d’Analyse des Roches et des Minéraux (CRPG, UMR 7358 CNRS-INSU, Vandoeuvre-lès-Nancy) using ICP-AES iCAP 6500 DUO and ICP-MS Thermo X7 for major elements and trace elements respectively.

MATERIAL Fossil specimens. The study material consists of crosssections and isolated body part fragments of fossil specimens that were collected in the Lägerstatte of the Djebel Oum Tkout locality in November 2012 (Muséum d’Histoire naturelle de Marrakesh, Morocco, MHNM). Three series of samples were studied: (i) two cross-sections through the pleon of the shrimp Cretapenaeus berberus Garassino, Pasini & Dutheil, 2006 (MHNM-KK-OT 01c), (ii) three isolated body part fragments and a cross-section from an usual teleost fish which state of preservation (fully articulated specimen with fossilized soft tissues) is representative of that of most actinopterygian fishes from the locality (MHNM-KK-OT 04), and (iii) a cross-section through a frequently observed, but still undescribed, decapod crustacean from the locality (MHNM-KK-OT 05) (Tab. 1). In anticipation of microtaphonomic studies these fossils were carefully


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1:

Page 4 of 12

Designation of the Late Cretaceous (95 Myr) fossil samples studied. Undescribed taxa are provisory in open nomenclature.

Specimen identification

Ref.

Material

Analytical techniques

Decapod crustacean C. berberus MNHN-KK-OT 01c

OT01c-s2f1

Cross-section through the pleon

XRF

OT01c-s1f2

Cross-section through the pleon

macro XANES, µXANES, speciation map

Teleost fish MHNM-KK-OT 03a

OT03a

Entire fossil

XRF

Teleost fish MHNM-KK-OT 04

OT04-is1

Isolated body part

XRF, macro XANES, µXANES, XRD

OT04-is2

Isolated body part

ICP-AES, ICP-MS

OT04-is3

Isolated body part

SEM

OT04-s3f1

Cross-section through the body

XRF, µXANES, speciation map

Decapod crustacean ’crab’ MHNM-KK-OT 05

OT05-s1f1

Cross-section through the carapace

XRF, macro XANES, µXANES, speciation map

Teleost fish MHNM-KK-OT 06

OT06

Cross-section through the caudal fin

SEM

Decapod crustacean C. berberus MNHN-KK-OT 07

OT07

Microsample

ICP-AES, ICP-MS

Sedimentary matrix

OTsm1

Microsample

ICP-AES, ICP-MS

Chondrichthyan Onchopristis MNHN-KK-OT 08 (from a sandstone layer just below the clayey beds)

OT08

Microsample of a rostral tooth

ICP-AES, ICP-MS

sampled as soon as they were found during the paleontological fieldwork. Each was packed in a glass Petri dish, sealed with Parafilm©, and stored in an opaque case to avoid, as much as possible, contamination and light-induced degradation during collection, transport and storage. All necessary permits were obtained for the described study, which complied with all relevant regulations. The Moroccan Ministère de l’Énergie, des Mines, de l’Eau et de l’Environnement delivered permits for the field campaign. This material is currently housed at the Muséum national d’Histoire naturelle (MNHN, Paris, France) for study within an agreement with the MHNM, where it will be returned after study.

hydroxylapatite family (Supplementary Information Fig. S1 & S-2). Replication of decay-prone tissues as authigenic apatite minerals, as observed here, is one of the major processes ensuring exceptional preservation of soft tissues. 37,38 The fidelity of the replication is attributed to the nucleation and growth of apatite nanocrystals of a diameter as small as 30 nm or below. 39–41 Microbial mats containing colonies of very well preserved bacteria and fungi have been identified on the top of each layer. They are assumed to allow the deposition of significant amounts of dissolved phosphorous at the sediment–water interface that are required for an extensive authigenic formation of calcium phosphate. 40,42,43 All fossils are largely covered by a reddish mineral compound, which consists in microspheres (usually 1.0–1.5 µm in diameter) of iron hydroxides that are attributed to past microbial activity.

Geochemistry at the site. The Late Cretaceous (ca. 95 million years old) levels of the Djebel Oum Tkout locality (Morocco) have yielded a well-preserved fauna consisting of mollusks, insects, isopods, malacostracan decapods, elasmobranchs and actinopterygians, 32–36 together with a rich flora represented by gymnosperms and angiosperms, found in 50 cm thick clayey beds at the bottom of the Upper unit of the Kem Kem beds. These organisms were found in 5-cm high successive grey illitic layers. These layers exhibit numerous mudcraks that suggest, together with the fauna as a whole, a peaceful seasonally dried freshwater habitat comparable to a small lake, pool or oxbow lake. Most of the fossils display an exceptional preservation of soft tissues, particularly consisting in finely phosphatized muscles showing the striation of the muscles fibers through SEM imaging. 32 µXRD mapping on the bulk fossil samples shows that both fossilised hard (bones, scales and cuticles) and soft tissues (mineralized muscles and gills) consist in minerals from the

Sample preparation. In the laboratory, all samples were carefully handled with gloves and sterilized tongs. The cross-sections investigated herein were obtained by sawing with a wire saw fossils still embedded in the sedimentary matrix. No organic material, such as resin, adhesive or glue was used during the preparation of samples. This made the preparation significantly more difficult but allowed preventing significant contamination and/or heating of the samples during their preparation.

RESULTS AND DISCUSSION Retrieving local REE patterns from µXRF spectra. XRF spectra collected at each pixel of the maps are processed


4

Page 5 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1:

Partial semi-quantitative REE patterns reconstructed from the processed XRF spectra compared to REE patterns based on ICPMS quantifications. OT01c-s1f2 and OT03a, partial semi-quantitative PAAS-normalized REE patterns reconstructed from the XRF data (16 pixels areas) respectively collected in the peculiar, notched, elongated bone from the teleost fish OT03a, and in the soft tissues of the shrimp C. berberus (cross-section OT01c-s1f2); OT04-is2, partial PAAS-normalized REE pattern based on ICP-MS quantifications from data collected in a sub-millimetric sample from a teleost fish (tissue sample OT04-is2); OT07 and OT08, complete PAAS-normalized REE patterns based on ICP-MS quantifications from data collected in sub-millimetric samples respectively from an another specimen of the shrimp C. berberus (soft tissue sample OT07), and from a sample of an isolated rostral tooth from the chondrichthyan Onchopristis (OT08) that was found in the sandstone layer just below the clayey beds that delivered the other fossils; Elements in grey could not be measured in the partial patterns.

to extract REE contents. Even though full quantification of trace elements using XRF is hampered by the local heterogeneity of the tissues that limits the precision of corrections for matrix X-ray reabsorption, statistics were sufficient to estimate the semi-quantitative contents in REE at the sub-millimetric scale. The estimated abundances of REEs were then normalized to the Post-Archean Australian Shale (PAAS) reference 44 to reconstruct partial local REE patterns at sub-millimetric lateral resolution (Fig. 1). Microscale REE patterns were obtained on two distinct fossils and appear in very good agreement with ICP-MS elemental quantifications carried out in three samples from the site: (i) the sub-millimetric tissue sample OT04-is2 from another teleost fish, (ii) the soft tissue sample OT07 from the shrimp C. berberus, and (iii) a sample from an isolated rostral tooth from the chondrichthyan Onchopristis (OT08) that was found in the sandstone layer just below the clayey beds that delivered our well-preserved fossils. Micrometric to sub-millimetric data from the fossils display no clear middle REE-enriched ‘bell-shaped’ pattern, therefore suggesting a preservation of the early diagenesis REE signal with major REE incorporation mainly via adsorption. 1,5 Furthermore, all REE patterns present a limited negative Ce anomaly (Fig. 1, Tab. 2), which may indicate a slightly oxidative local condition at burial. The homogeneity in REE patterns tends to demonstrate that the overall local redox conditions can be considered as homogeneous at the scale of several successive stratigraphic layers of the site.

lines, only V Kα (4.95 keV) and Ba Lβ (4.83 keV) emissions may significantly overlap, but V and Ba concentrations are respectively 87 and 14 times lower than that of Ce. This is critical for reliable elemental and speciation mapping. At large length scales (step size: 100 µm), the cerium content appears highly heterogeneous. Ce is associated with the calcium distribution of fossil biogenic and authigenic hydroxylapatites (Supplementary Information Fig. S3 B,D,F,H,J, green to orange in the false color overlays). Ce is also found in association with iron-rich layers or spherical structures (Supplementary Information Fig. S-3 D,F,H,J, light blue) from reddish iron hydroxide grains visible in the optical photographs, although this association is significantly weaker (see the weak cerium contribution to the full XRF spectra in these iron-rich layers in Supplementary Information Fig. S-5 D1,G3). One should be careful when interpreting these results as the information depth of the Ce Lα emission in hematite Fe2 O3 is greater (16 µm) than in apatite (9 µm). In our samples, visualisation of Ce and the REE series up to Eu is confirmed by taking advantage of the tunability of the energy of the X-rays of synchrotron: when excitation energy is set below the Fe K-edge, REE L emission lines are straightforwardly identified in XRF spectra (Supplementary Information Fig. S-6). X-ray absorption near-edge structure (XANES) spectroscopy offers a complementary direct way to study the cerium anomaly. Indeed, XANES spectroscopy allows direct probing of the speciation, in particular identification of the oxidation state of the target atom. Spectra collected at the Ce L3 -edge XANES are known to be remarkably different among Ce(III) and Ce(IV) compounds. 18,28,45,46 This approach could therefore represent a powerful tool to directly assess the oxidation state of Ce and [Ce(IV)]/[Ce(tot)] ratios. Only scarce data is available on the XANES study of Ce in comparable samples, such as the work by Takahashi et al. on geological samples including biogenic apatite using millimetric beams (1×0.5 mm2 ). 18,45,46 Recently, Janots et al. used a microbeam (3×3 µm2 ) at the LUCIA beamline to col-

Direct probing of the cerium oxidation state at mesoscale. Ce is present at fractions of weight percent in the fossil tissues (Supplementary Information Tab. S-1), a range straightforwardly accessible using synchrotron µXRF and µXAS. Contrary to other trace elements such as Gd, the Ce Lα emission (4.83 keV) is sufficiently distant from characteristic lines from major and minor elements in fossils, in particular Fe Kα (6.40 keV), to be readily quantified in synchrotron XRF maps. 25 Among neighbouring emission


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 12

Table 2:

Cerium concentration and Ce anomaly obtained from ICP-MS quantification and processed XRF spectra. Data reconstructed from the processed XRF spectra (16 pixels areas) are presented with standard deviation. The cerium anomaly (Ce/Ce*) is calculated as 3CeN /(2LaN +NdN ), where N represents shale-normalized concentration.

Sample

Analytical technique

[Ce] (ppm)

Ce/Ce*

OT04-is2

ICP-MS

2700

0.425

OT07

ICP-MS

2030

0.448

OT08

ICP-MS

1605

0.415

OT01cs1f2

XRF

2067.11 (± 730.22)

0.353 (± 0.107)

OT03a

XRF

1343.14 (± 309.02)

0.259 (± 0.078)

Figure 2: Comparison of typical sub-millimetric XANES spectra collected at the Ce L3 edge (2p3/2 ) in three fossils and two standards. From top to bottom, XANES spectra within the fossilized tissues of the teleost fish OT04-is1, soft tissues of the shrimp Cretapenaeus berberus OT01c-s1f2 and of the new crab OT05-s1f1, and standards (fluorapatite, cerium oxide). Vertical lines A-D denote the main electron transitions.

lect Ce L3 -edge XANES on particular micrometric crystals within a lateritic profile from Madagascar. 28 We collected Ce L3 -edge XANES spectra at both mesoand microscale (300×300 µm2 and 3×3 µm2 H×V FWHM) in our fossils. Typical mesoscale Ce L3 edge XANES spectra collected at a photon flux of 1.6×109 photons.s−1 from the fish, shrimp and crab soft tissues are displayed in Fig. 2, together with spectra from fluorapatite and cerium oxide standards. Three main spectral features are identified: a white line at 5.726 keV resulting from the 2p → (4f1 )5d electron transition and two peaks that partly overlap with the former at the energy resolution of the monochromator attributed to the 2p3/2 → (4f1 L)5d and the 2p3/2 → (4f0 )5d transitions that lead to peaks centered at 5.730 keV and 5.736 keV respectively (see Takahashi et al. 18,45,46 and references therein). Spectra collected at mesoscale indicate that cerium is predominantly under the +III valence state, but low-intensity distinctive features attributed to Ce(IV) are also identified. Linear deconvolution based on Ce(III)(NO3 )3 ·6H2 O and

Ce(IV)(SO4 )2 standard spectra indicates a Ce(IV) content of ca. 23%, 18% and 20% in the fish, shrimp and crab fossil tissues, respectively. The presence of some cerium under the +IV form, e.g. as CeO2 , points to a slightly oxidative local taphonomic context. This is compatible with the limited negative cerium anomaly observed in the REE elemental patterns shown in Fig.1. Mapping the cerium oxidation state at microscale. Although no radiation side-effect is observed with the low flux density synchrotron beam at mesoscale, cerium is known to be sensitive to photo-oxidation, 47 which may represent a possible limitation regarding precise assessment of the [Ce(IV)]/[Ce(tot)] ratio in samples analyzed under greater flux densities. The photon flux density with the microfocused beam exceeds by 5 orders of magnitude that of the unfocused beam (typ. 3×109 vs. 2×104 photons.s−1 .µm-2 at mesoscale). In order to determine whether collection of absorption spectra was reliable at microscale, we collected successively a series of microfocused XANES spectra at the


6

Page 7 of 12

B 215 s Ce(III)(NO3)3·6H2O contribution

(4f1)5d

0.6

0.7

Ce(IV)(SO4)2 contribution 645 s 3010 s

Ce(III), 2p3/2

0.8

A

0.5

Normalized absorption (a. u.)

R2 = 0.997

0

1000

2000

3000

Exposure time (s)

0.5

C (4f0)5d

0.3

0.4

Ce(IV), 2p3/2

0.2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5.700

5.750

5.800

5.850

Energy (keV)

R2 = 0.997

0

1000

2000

3000

Exposure time (s)

Figure 3: Assessment of the photo-oxidation kinetics of Ce in realistic irradiation conditions. A, Ce L3 XANES after 215, 645 and 3010 s of measures in the shrimp cross-section (OT01c-s1f2; photon flux density ca. 3×109 photons.s−1 .µm-2 ). Red dotted and dashed lines respectively show Ce(III) and Ce(IV) contributions to the XANES spectra collected after 215 s. B-C, kinetics behavior of [Ce(III)] (B) and [Ce(IV)] (C) as a function of the accumulation time (s), in real experimental conditions while collecting successive XANES spectra.

LUCIA beamline on a same spot of the shrimp cross-section. A notable increase of the peak corresponding to Ce(IV) is observed that indicates a significant photo-oxidation of cerium (Fig. 3): hν

Ce3+ −→ Ce4+ + e−

These results clearly highlight the presence of Ce under the +IV form within the fossil tissues (Fig. 4E,J,O). The measured Ce(IV) range (20±5%) correlates very well to that assessed at mesoscale from the deconvolution of the sub-millimetric Ce L3 -edge XANES spectra collected at the DiffAbs beamline (Fig. 4K,P). The Ce(IV) distribution appears highly homogeneous between the different fossil tissues (Fig. 4E,J,O), as demonstrated by the Kernel density estimates of the [Ce(IV)]/[Ce(tot)] ratio at the histological level within the tissues of the shrimp and of the new crab (Fig. 4K,P). Such homogeneity in the cerium oxidation state is particularly noticeable because a significant contrast is observed in the total cerium content between distinct fossilized tissues (Fig. 4D,I,N). This behaviour is best exemplified in the cross-section through the carapace of the crab where the [Ce(IV)] distribution is highly homogeneous between the cuticular layer (dorsal green vitreous layer displaying growth lines, "cuticle") and the mineralized soft tissues (ventral brown and white layers, "soft-t1" and "soft-t2" respectively), whereas the cerium concentration significantly varies between them (Fig. 4N, Supplementary Information Fig. S-7). By imaging here for the fist time the microscale speciation of Ce within well-preserved fossils, we demonstrate that the [Ce(IV)] distribution, and thus the [Ce(IV)]/[Ce(tot)] ratio, is remarkably homogeneous at microscale within the different fossils of the locality, without significant tissuedependent speciation behaviour that would have vastly complicated the interpretation of REE patterns routinely used by geochemists.

(1)

The characteristic time associated to photo-oxidation falls within the thousands of seconds time scale, with already measurable changes after a few hundreds of seconds. Except for this series of spectra collected at the same location, all other XANES spectra presented in this paper were therefore collected fast and carefully avoiding previously irradiated sample areas. This ensures minimal artefacts due to photo-oxidation during data collection. However, collection of full XANES spectra at such limited Ce concentrations in a timeframe significantly shorter than 100 s leads to spectra that are too noisy to reliably assess the [Ce(IV)]/[Ce(tot)] ratio at microscale. In order to further limit irradiation while collecting speciation information on a bulk sample in a reflection geometry, a distinct strategy was designed. We constructed Ce L3 -edge speciation maps from successive XRF maps taken at different excitation energies around the Ce L3 absorption edge. Excitation energies were chosen in the pre-edge (5.680 keV), edge (5.727 keV), major EXAFS oscillations (5.733, 5.736 and 5.765 keV) and in the post-edge (5.900 keV) regions. This allowed an accurate inter-comparison of speciationrelated contrasts while irradiating the sample areas under investigation only during 6 seconds. The resulting distributions of Ce(IV) at microscale (scan steps of 1.5×1.5 or 5×5 µm2 ) mapped within transects through the three fossil cross-sections are presented in Fig. 4.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 12

Figure 4:

Speciation maps of cerium in fossil sections. Reconstructions and large views of the fossils, optical photographs of the cross-sections studied, high resolution distributions of total cerium content and [Ce(IV)]/[Ce(tot)] maps from the corresponding box areas from a cross section through a rib of the usual teleost fish OT04-s3f1 (A-E), a cross-section through the fourth somite of the shrimp Cretapenaeus berberus OT01c-s1f2 (F-J), and a cross-section through the carapace of the new crab OT05-s1f1 (L-O). K and P present Kernel density estimates of the [Ce(IV)] distribution (bold lines, N=574 pixels) from areas corresponding to the cuticle and soft tissues of the shrimp (K) and crab (P). Vertical dashed lines identify the [Ce(IV)] contribution as obtained from the deconvolution of the mesoscale XANES spectra. D,E: scan step: 1.5×1.5 µm2 , 3,996 pixels. I,J: scan step: 1.5×1.5 µm2 , 4,387 pixels. N,O: scan step: 5×5 µm2 , 4,141 pixels. Scale bars represent 1 mm in B, 50 µm in C, D, I, and N, 1 cm in G and L, and 500 µm in H and M.

Conclusions and perspectives

entire flat fossils at microscale, and therefore constitutes a promising tool to study REE patterns in fossils. X-ray absorption spectroscopy has been proposed as a way to directly assess the speciation state of cerium in geological applications. 18,45,46 Specific constraints are posed by working at the microscale. In particular, as discussed here, the issue of photoxidation of Ce(III) to Ce(IV) must be considered with great care. Coupling of macro-XANES spectroscopy and derived speciation mapping at characteristic emission energies could be satisfactorily performed at microscale on fossil cross-sections. We have shown here that it provides direct access to the spatial fluctuation of cerium oxidation states across entire fossil sections. Our results demonstrate that the approach is sensitive enough to obtain [Ce(IV)]/[Ce(tot)] maps within fossils at histological scale and correlate well to macroscopic and microscopic REE elemental patterns. In addition to the cerium anomaly observed in REE fractionation patterns, the methodology presented herein provides a new indicator to directly as-

Only a very limited set of approaches is available to study the cerium anomaly (an essential paleo-environmental proxy), namely through measuring REE elemental concentration profiles. The vast majority of existing works were performed using macroscopic measurements through ICP-MS analysis. Analysing the cerium distribution at microscale may provide significant information regarding the biological, physical and chemical mechanisms governing the anomaly, possibly making a clever use of fractionation fluctuations within the REE series. LA-ICP-MS, a micro-destructive approach, allows mapping trace elements (and particularly REEs including cerium) distributions and contents in fossil bones, but remains of limited practical use, particularly when seeking correlation with other information, such as fossil morphology or histology. 19 Synchrotron X-ray fluorescence microscale mapping of major-to-trace elements 25 has shown that REE distributions can straightforwardly be mapped throughout


8

Page 9 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5:

Changes in cerium content and [Ce(IV)]/[Ce(tot)] ratio through time in the acantomorph fish Spinocaudichthys oumtkoutensis (Djebel Oum Tkout Lagerstätte, Upper Cretaceous of Morocco). ↑, ↑↓, and ↔ indicate increasing, fluctuating and stabilized cerium content or [Ce(IV)]/[Ce(tot)] ratio respectively.

sess redox conditions of burial from fossil sites. The observed [Ce(IV)]/[Ce(tot)] ratio is shown to be remarkably constant between distinct fossils from the studied site both at mesoscale and at microscale. No tissue-dependent speciation behaviour was found whatever the original tissue was (bone bioapatite or soft-tissues such as muscles), even though the cerium concentration significantly varies between them. This work therefore highlights the possibility to further clarify the taphonomic pathways of cerium in fossils. The sequence displayed in Fig. 5 can be considered as a starting point: (1) at the death of the animal, the amount of REEs is limited to an in vivo ppt to ppm range; (2) post mortem, bones and newly mineralised tissues concentrate REEs from fluids. Local redox conditions influence the [Ce(IV)]/[Ce(tot)] ratio and the deposition of cerium, and (3) at longer term (thousands to millions of years), both hard tissues and mineralized soft tissues reach a (still debated) closed-pore state, the sediment will keep on being irrigated by Ce-containing fluids. REEs concentrations in fossils are typically is the hundred to thousand ppm range. However, it remains to be shown whether the [Ce(IV)]/[Ce(tot)] ratio within the fossil can be seen as representing an early state, integrated state or final state of the local conditions. This is

also likely to be variable from one site to another, depending on the diagenetic context. Although such methodological developments are described herein against fossils, they deserve to be applied to other paleo- and geo-environmental proxies (e.g. bone crosssections, sediments) in order to better constrain trace elements diffusion, uptake and release. Redox proxies such as the Ce and Eu anomaly could be directly investigated across 2D maps rather than along profiles, therefore bringing more robustness to their analysis. These imaging tools show great promise for studying recrystallization processes that affect early taphonomic signals, and help selecting suitable areas for the quantification and speciation of rare earth in essential probes to our Past. Furthermore, the newly developed methodology is expected to be highly beneficial to the wide community studying rare earth elements in environmental sciences, geochemistry, soil chemistry and cultural heritage. It is also likely to attract attention from the large number of scientists interested on the microscale chemical analysis and speciation of rare earths in fields as diverse as the mining industry, metallurgy or medicine.

Acknowledgement We acknowledge SOLEIL for provision of synchrotron radiation under project no.


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

20130319. We thank D. Thiaudière, S. Réguer and F. Alvès for assistance at the DiffAbs beamline, and D. Vantelon and N. Trcera at the LUCIA beamline (SOLEIL synchrotron). E. Janots (Joseph Fourier University, Grenoble) is warmly acknowledged for providing us Ce(III) and Ce(IV) standard XANES data. P. Cook and S. X. Cohen (IPANEMA) are acknowledged for helpful discussions about XRF and XANES data processing, as well as M. Thoury (IPANEMA) and W. Josse for support in the preparation of samples, M.A. Languille (IPANEMA; CRC) and P. Cook for support in the SEM measurements, at IPANEMA. The SpectroMicro R package is developed by S. X. Cohen. We thank the field workers who collected the fossils (D. B. Dutheil, S. Charbonnier, G. Clément, N.-E. Jalil, F. Khaldoune, H. Bourget and B. Khalloufi (MNHN, Paris), A. Tourani (Cadi Ayyad University, Marrakesh) and P. M. Brito (Rio de Janeiro State University, Rio de Janeiro). Teams at the Institut des Sciences Analytiques (UMR 5280 CNRS, Lyon, France) and Service d’Analyse des Roches et des Minéraux (CRPG, UMR 7358 CNRS-INSU, Vandoeuvre-lès-Nancy) carried out the ICP-MS and AES quantifications. This work is a contribution to the French ANR under the TERRES project (ANR-2010-BLAN-607-03 grant) and has been developed as part of the agreement between IPANEMA and Muséum national d’Histoire naturelle. ICP-MS and AES quantifications were supported by the Muséum national d’Histoire naturelle through the ASTRAF project (ATM Interactions Minéral–Vivant) and by the UMR 7207 CR2P. The IPANEMA platform is a joint unit from CNRS and Ministère de la Culture et de la Communication, and benefits from a CPER grant (MENESR, Région Ile-de-France). 48

Page 10 of 12

(11) Chen, J.; Algeo, T. J.; Zhao, L.; Chen, Z.-Q.; Cao, L.; Zhang, L.; Li, Y. Earth-Science Reviews 2015, (12) Elderfield, H.; Pagett, R. Science of the Total Environment 1986, 49, 175–197. (13) Lécuyer, C.; Grandjean, P.; Barrat, J.-A.; Nolvak, J.; Emig, C.; Paris, F.; Robardet, M. Geochimica et Cosmochimica Acta 1998, 62, 2429–2436. (14) Morad, S.; Felitsyn, S. Sedimentary Geology 2001, 143, 259–264. (15) Earnshaw, A.; Greenwood, N. Chemistry of the Elements; Butterworth-Heinemann Oxford, 1997. (16) Elderfield, H.; Greaves, M. J. Nature 1982, 296, 214–219. (17) de Baar, H. J.; German, C. R.; Elderfield, H.; Van Gaans, P. Geochimica et Cosmochimica Acta 1988, 52, 1203–1219. (18) Takahashi, Y.; Shimizu, H.; Kagi, H.; Yoshida, H.; Usui, A.; Nomura, M. Earth and Planetary Science Letters 2000, 182, 201–207. (19) Koenig, A. E.; Rogers, R. R.; Trueman, C. N. Geology 2009, 37, 511– 514. (20) Hinz, E. A.; Kohn, M. J. Geochimica et Cosmochimica Acta 2010, 74, 3213–3231. (21) Suarez, C. A.; Macpherson, G.; González, L. A.; Grandstaff, D. E. Geochimica et Cosmochimica Acta 2010, 74, 2970–2988. (22) Herwartz, D.; Tütken, T.; Münker, C.; Jochum, K. P.; Stoll, B.; Sander, P. M. Geochimica et Cosmochimica Acta 2011, 75, 82–105. (23) Trueman, C. N.; Kocsis, L.; Palmer, M. R.; Dewdney, C. Palaeogeography, Palaeoclimatology, Palaeoecology 2011, 310, 124–132. (24) Herwartz, D.; Tütken, T.; Jochum, K. P.; Martin Sander, P. Geochimica et Cosmochimica Acta 2013, 103, 161–183.

Supporting Information Available: This material is available free of charge via the Internet at http://pubs.acs.org/.

(25) Gueriau, P.; Mocuta, C.; Dutheil, D. B.; Cohen, S. X.; Thiaudière, D.; consortium, T. O.; Charbonnier, S.; Clément, G.; Bertrand, L. PloS one 2014,

References

(26) Solé, V.; Papillon, E.; Cotte, M.; Walter, P.; Susini, J. Spectrochimica Acta Part B: Atomic Spectroscopy 2007, 62, 63–68.

(1) Trueman, C. N.; Tuross, N. Reviews in mineralogy and geochemistry 2002, 48, 489–521.

(27) R Development Core Team, R Foundation for Statistical Computing, Vienna, Austria 2011,

(2) Wright, J.; Schrader, H.; Holser, W. T. Geochimica et Cosmochimica Acta 1987, 51, 631–644.

(28) Janots, E.; Bernier, F.; Brunet, F.; Muñoz, M.; Trcera, N.; Berger, A.; Lanson, M. Geochimica et Cosmochimica Acta 2015, 153, 134–148.

(3) Reynard, B.; Lécuyer, C.; Grandjean, P. Chemical Geology 1999, 155, 233–241.

(29) Medjoubi, K.; Bucaille, T.; Hustache, S.; Bérar, J.-F.; Boudet, N.; Clemens, J.-C.; Delpierre, P.; Dinkespiler, B. Journal of synchrotron radiation 2010, 17, 486–495.

(4) Kemp, R. A.; Trueman, C. N. Sedimentary Geology 2003, 155, 109– 127.

(30) Le Bourlot, C.; Landois, P.; Djaziri, S.; Renault, P.-O.; Le Bourhis, E.; Goudeau, P.; Pinault, M.; Mayne-L’Hermite, M.; Bacroix, B.; Faurie, D.; Castelnau, O.; Launois, P.; Rouzière, S. Journal of Applied Crystallography 2011, 45, 38–47.

(5) Lécuyer, C.; Reynard, B.; Grandjean, P. Chemical Geology 2004, 204, 63–102. (6) De Baar, H.; Bacon, M.; Brewer, P. Nature 1983, 301, 324–327.

(31) Mocuta, C.; Richard, M.-I.; Fouet, J.; Stanescu, S.; Barbier, A.; Guichet, C.; Thomas, O.; Hustache, S.; Zozulya, A. V.; Thiaudiere, D. Journal of Applied Crystallography 2013, 46, 1842–1853.

(7) Grandjean, P.; Cappetta, H.; Michard, A.; Albarede, F. Earth and Planetary Science Letters 1987, 84, 181–196. (8) German, C. R.; Elderfield, H. Paleoceanography 1990, 5, 823–833.

(32) Dutheil, D. B. Mesozoic fishes 1999, 2, 553–563.

(9) Jones, B.; Manning, D. A. Chemical Geology 1994, 111, 111–129.

(33) Dutheil, D. B. Journal of Vertebrate Paleontology 1999, 19, 243–246.

(10) Pattan, J.; Pearce, N.; Mislankar, P. Chemical Geology 2005, 221, 260–278.

(34) Filleul, A.; Dutheil, D. B. Journal of Vertebrate Paleontology 2001, 21, 774–780.


10

Page 11 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(35) Filleul, A.; Dutheil, D. B. Journal of Vertebrate Paleontology 2004, 24, 290–298. (36) Garassino, A.; Pasini, G.; Dutheil, D. B. Atti della Società italiana di scienze naturali e del museo civico di storia naturale di Milano 2006, 147, 3–17. (37) Allison, P. A.; Briggs, D. E. G. In The processes of fossilization; Donovan, S. K., Ed.; Columbia University Press, 1991; pp 120–140. (38) Briggs, D.; Kear, A.; Martill, D.; Wilby, P. Journal of the Geological Society 1993, 150, 1035–1038. (39) Wilby, P. Kaupia 1993, 2, 99–113. (40) Briggs, D. E. Annual Review of Earth and Planetary Sciences 2003, 31, 275–301. (41) Allison, P. A.; Bottjer, D. J. Taphonomy: process and bias through time; Springer, 2011. (42) Martill, D. M. Palaeontology 1988, 31, 1–18. (43) Wilby, P. R.; Briggs, D. E.; Bernier, P.; Gaillard, C. Geology 1996, 24, 787–790. (44) McLennan, S. Reviews in Mineralogy and Geochemistry 1989, 21, 169–200. (45) Takahashi, Y.; Shimizu, H.; Usui, A.; Kagi, H.; Nomura, M. Geochimica et Cosmochimica Acta 2000, 64, 2929–2935. (46) Takahashi, Y.; Sakami, H.; Nomura, M. Analytica Chimica Acta 2002, 468, 345–354. (47) Shimizugawa, Y.; Handa, K.; Qiu, J. Journal of Materials Science Letters 2003, 22, 15–16. (48) Bertrand, L.; Languille, M.-A.; Cohen, S. X.; Robinet, L.; Gervais, C.; Leroy, S.; Bernard, D.; Le Pennec, E.; Josse, W.; Doucet, J.; Schöder, S. J. Synchrotron Rad. 2011, 18, 765–772.


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

47x26mm (300 x 300 DPI)


Page 12 of 12

Cerium Anomaly at Microscale in Fossils - Analytical Chemistry (ACS

Recommend Documents