-
Chemical Geology 291 (2012) 3854
Contents lists available at SciVerse ScienceDirect
Chemical Geology
j ourna l homepage: www.e lsev ie r .com/ locate /chemgeo
Research paper
A novel extraction chromatography and MC-ICP-MS technique for
rapid analysis ofREE, Sc and Y: Revising CI-chondrite and
Post-Archean Australian Shale(PAAS) abundances
Ali Pourmand a,b,, Nicolas Dauphas a, Thomas J. Ireland a
a Origins Laboratory, Department of the Geophysical Sciences and
Enrico Fermi Institute, The University of Chicago, 5734 South Ellis
Avenue, Chicago IL 60637, USAb Neptune Isotope Laboratory, Division
of Marine Geology and Geophysics, The University of Miami, RSMAS,
4600 Rickenbacker Causeway, Miami, FL 33149, USA
Corresponding author. Tel.: +1 305 421 4384; fax:E-mail address:
[email protected] (A. P
0009-2541/$ see front matter 2011 Elsevier B.V.
Alldoi:10.1016/j.chemgeo.2011.08.011
a b s t r a c t
a r t i c l e i n f o
Article history:Received 2 March 2011Received in revised form 14
August 2011Accepted 24 August 2011Available online 30 August
2011
Edited by: K. Mezger
Keywords:CI-chondriteRare earth elementsLiBO2 flux fusionTODGA
extraction chromatographyMC-ICP-MSPost-Archean Australian Shale
(PAAS)
A new analytical protocol is introduced for rapid measurement of
rare-earth elements (REE), Sc and Y in mete-oritic and geological
materials bymulti-collection inductively coupled plasmamass
spectrometry (MC-ICP-MS).A simple purification step was devised to
reduce REE, Sc and Y abundances in commercial lithium
metaborate(LiBO2) for low-blank flux fusion. Separation of the
analytes from the rockmatrix was achieved by using a singleTODGA
extraction chromatography step. A dynamicmulti-collector cup
configurationwas developed tomeasureREE, Sc and Y using a
desolvating nebulizer and standard-sample bracketing technique. To
test the accuracy ofthis analytical protocol, we analyzed aliquots
of USGS geological reference materials BHVO-1, BIR-1, BCR-2,PCC-1,
W-2, G-2 and G-3, specifically selected to encompass a wide range
of REE, Sc and Y concentrations andmineral compositions. Elemental
abundances in reference materials are indistinguishable within
analytical un-certainties from compilations of literature values
analyzed by various ICP-MS techniques. The average
externalreproducibility on REE, Sc and Y concentrations (reported
as RSD=100standard deviation/average) was~2% based on replicates of
G-3. With the exception of PCC-1, which has low REE concentrations,
adjustmentsfor poly-atomic interferences and procedural blanks in
the reference materials were negligible.In order to re-visit the
terrestrial and cosmic abundances of REE, Sc and Y, aliquots of
nine Post Archean Austra-lian Shales (PAAS), Allende (CV-3), Tagish
Lake (C2-ungrouped), Alais (CI1), Orgueil (CI1) and Ivuna (CI1)
me-teorites were measured using our new analytical procedure. The
REE patterns of PAAS, normalized to the meanof CI-chondrites from
this study, are smoother and show less dispersion compared with
literature measure-ments. Eu/Eu*, LREE/HREE, and La/Sc ratios
remain constant in these samples. The recommended PAAScomposition
based on these new measurements is (in g g1): Sc=15.89, Y=27.31,
La=44.56, Ce=88.25,Pr=10.15, Nd=37.32, Sm=6.884, Eu=1.215,
Gd=6.043, Tb=0.8914, Dy=5.325, Ho=1.052, Er=3.075,Tm=0.4510,
Yb=3.012 and Lu=0.4386. The REE pattern in Allende is similar to
group II-type CaAl-richinclusions (CAIs) that typically show
enrichment in light REE (LREE), depletion in heavy REE (HREE),
andnegative and positive anomalies for Eu and Tm, respectively. The
REE in Tagish Lake and Alais do not showsignificant fractionations
and closely resemble the relatively flat pattern observed in
Orgueil. Based on eighthigh-precision multi-collection ICP-MS
measurements of Orgueil (n=5), Ivuna (n=2) and Alais (n=1),
werecommend a new CI-composition for REE, Sc and Y normalization
and refine the cosmic abundances of theseelements (in g g1):
Sc=5.493, Y=1.395, La=0.2469, Ce=0.6321, Pr=0.0959, Nd=0.4854,
Sm=0.1556,Eu=0.0599, Gd=0.2093, Tb=0.0378, Dy=0.2577, Ho=0.0554,
Er=0.1667, Tm=0.0261, Yb=0.1694 andLu=0.0256.
+1 305 421 4632.ourmand).
rights reserved.
2011 Elsevier B.V. All rights reserved.
1. Introduction
Over the past two decades, inductively coupled plasma mass
spec-trometry (ICP-MS) has increasingly become the preferred method
fordetermination of REE, Sc and Y abundances in meteoritic and
geolog-ical materials (Balaram, 1996; Eggins et al., 1997; Pin and
Joannon,
1997; Potts, 1997; Jarvis, 1988; Robinson et al., 1999; Yu et
al.,2001; Bayon et al., 2009). While most ICP techniques have
utilizedsingle collector, sector-field or quadrupole mass
spectrometers,multi-collector ICP-MS in conjunction with isotope
dilution massspectrometry (IDMS) has only been recently explored
for REE mea-surements (Baker et al., 2002; Kent et al., 2004). The
ability to simul-taneously measure multiple isotopes with MC-ICP-MS
allowstechniques such as isotope dilution and standard-sample
bracketingto achieve more precise and accurate elemental
concentrations. Con-current measurement of multiple isotopes
minimizes the influence of
http://dx.doi.org/10.1016/j.chemgeo.2011.08.011mailto:[email protected]://dx.doi.org/10.1016/j.chemgeo.2011.08.011http://www.sciencedirect.com/science/journal/00092541
-
39A. Pourmand et al. / Chemical Geology 291 (2012) 3854
fluctuations inherent in the plasma source and sample
introductionsystems that otherwise limit precision and accuracy of
the analysiswith single collector ICP-MS.
In order to improve the accuracy and precision of REE, Sc and
Ymeasurements by MC-ICPMS, the following conditions must be
ful-filled: 1) complete digestion of refractory phases and minerals
inthe samples; 2) low procedural blanks; 3) separation of
interferingmatrix, and 4) elimination of molecular (i.e., oxides
and hydrides)and direct isobaric interferences on the mass of the
analytes of inter-est. In the following contribution, we introduce
a novel analytical pro-tocol to reduce REE, Sc and Y concentrations
in commercial LiBO2 flux,which facilitates processing small
aliquots (~1550 mg) of sampleswith concentrations as low as 10 ng
g1. A novel extraction chroma-tography was also developed using a
resin from Eichrom Inc. for REE,Sc and Y purification. The
exceptional affinity of N,N,NN tetraoctyl-1,5-diglycolamide ligand
(hereafter TODGA) for these elements hasmade this resin a preferred
medium for separating the rock matrixthat interferes with REE, Sc
and Y measurements during ICP-MS anal-ysis (Horwitz et al., 2005;
Connelly et al., 2006; Sasaki et al., 2007;Pourmand and Dauphas,
2010). Our proposed procedure takes advan-tage of improved sample
digestion with low-blank, high-temperatureLiBO2 fusion, a single
TODGA extraction chromatography step forquantitative matrix
separation, and a novel, dynamic multi-collectorcup configuration
method that allows measurement of REE, Sc andY byMC-ICP-MS in a
single sample solution. The entire analytical proce-dure, from
sample fusion to MC-ICP-MS analysis, can be completed inless than a
day. In order to examine the accuracy of our protocol, ele-mental
concentrations and CI-normalized REE patterns for seven
USGSreference materials are measured and compared with
high-precisiondatameasured by various ICP-MS techniques from the
literature. Finally,we present REE, Sc and Y abundances and REE
fractionation patterns inaliquots of Post-Archean Australian Shales
(PAAS), Allende (CV3), Tag-ish Lake (CI2-ungrouped), Orgueil (CI1),
Alais (CI1) and Ivuna (CV1)meteorites with the aim to re-evaluate
the terrestrial and cosmic abun-dances of these elements.
2. Method development
2.1. Materials
Concentrated, certified ACS Plus hydrochloric (HCl) and
nitric(HNO3) acids from Fisher Scientific were distilled twice in
sub-boilingquartz and PTFE Teflon stills. The working acid
solutions were titratedwith calibrated sodium hydroxide solutions
before use. High-puritywater from a Millipore Milli-Q system
(resistivityN18 M cm1)was used for rinsing labware, extraction
chromatography and acid di-lutions. Savillex PFA Teflon vials and
beakers used for purification ofalkali flux and chromatography
elutions were first cleaned in 50%HNO3 at 70 C, followed by boiling
in aqua regia (HCl:HNO3 at 3:1).Platinum evaporation dishes were
cleaned in a boiling, 4 mol L1
HCl solution. Pre-packed, 2-mL TODGA cartridges (resin mesh
size:50100 m), connectors and a Plexiglas vacuum chamber are
avail-able from Eichrom Inc. Polypropylene columns from Bio-rad
wereused as loading reservoirs for cartridge chromatography.
High-purity,8-mL graphite crucibles, LiBr non-wetting agent (Pure
grade) and cer-tified multi-element REE, Sc and Y standard
solutions were obtainedfrom SPEX CertiPrep. Commercially
manufactured Puratronic lithiummetaborate powder (99.997% metals
basis, cat# 10739) was pur-chased from Alfa Aesar.
2.2. Sample digestion
2.2.1. Purification of lithium metaborate fluxComplete
dissolution of refractory minerals (such as garnet and
zircon) is a prerequisite for accurate and precise isotopic and
elemen-tal analysis of terrestrial and meteoritic materials with
solution-based
ICP-MS. Conventional methods that use hotplate or
high-pressureParr Bombs and a mixture of HNO3HFHCLO4 (e.g., Eggins
et al.,1997; Baker et al., 2002; Mahlen et al., 2008) are effective
but cantake days to achieve complete dissolution of samples that
containhighly refractory phases. High-temperature alkali flux
fusion offers afaster and more efficient alternative to acid
digestion, and the result-ing solution can be readily used for
matrix separation by extraction orion-exchange chromatography
(Bizzarro et al., 2003; Le Fevre andPin, 2005; Connelly, 2006;
Pourmand and Dauphas, 2010; Dauphasand Pourmand 2011). Flux fusion
also eliminates the need for multi-ple evaporation and acid
conversion steps, which are commonly re-quired in acid dissolution
protocols and can potentially result insample loss, contamination
or elevated blank levels. In practice, how-ever, fewer studies have
chosen flux fusion over acid dissolution be-cause of relatively
higher blank levels associated with commerciallymanufactured flux
material and incomplete recovery of the melt. Ina recent study, for
example, Bayon et al. (2009) reported very highprocedural blanks
(e.g., CeN2.7 ng) when using a Na2O2NaOH fluxmelting technique.
Pourmand and Dauphas (2010) also found highlevels of actinides and
high-field strength elements in three commer-cially available flux
materials. While high blank levels are inconse-quential for the
analysis of most terrestrial materials that have highREE, Sc and Y
abundances, significant blank corrections in bulk mete-orites and
some terrestrial rocks (such as peridotites), with concen-trations
in a few parts per trillion, can lead to inaccurate results.
Ifalkali flux fusion is to be successfully utilized with samples
thathave low levels of REE, Sc and Y, the flux must be further
purified toreduce blank contributions.
In order to address this need, a simple and highly efficient
methodwas developed to remove these elements from commercially
avail-able alkali flux materials using TODGA extraction
chromatography.Approximately 60 g of Puratronic LiBO2 powder from
Alfa Aesar wasdissolved at room temperature in 2 L of 3 mol L1 HNO3
stocked ina PFA Teflon bottle. Adding higher amounts of the flux
will result inincomplete dissolution and precipitation of the flux
on the resin.Three 2-mL Eichrom TODGA cartridges were stacked in
tandem onthe vacuum chamber and 10-mL polypropylene columns were
usedas loading reservoirs. The cartridge assembly was
preconditionedwith 60 mL of 0.05 mol L1 HCl to remove residual REE,
Sc and Y onthe resin, followed by 20 mL of MQ water and 20 mL of 3
mol L1
HNO3 to convert the resin to the loading acid concentration. The
lith-ium metaborate flux solution was subsequently loaded to the
car-tridge array and the eluent was collected in precleaned PFA
Teflonbottles at elution rates of 34 mL min1. A total of five TODGA
car-tridge arrays were used for 400 mL of flux solution per each
array toavoid elution of REE, Sc and Y from the resin due to high
load volume.
The purified LiBO2 was recovered in two steps. The solution
wasfirst evaporated to a thin slab in increments of 200 mL in
precleaned,250 mL PFA Savillex beakers. The beakers were capped
usingthreaded closers with portholes to minimize contamination.
TheLiBO2 slabs (~5 cm in diameter) were subsequently heated in
acapped, 100-mL platinum evaporation dish inside a Thermoline
fur-nace at 720 C for 1 h. This temperature is safely below the
meltingpoint of LiBO2 (845 C), yet it is sufficiently high to allow
complete re-moval of residual HNO3 and water from the flux slabs.
Subsequenttests have shown that high-purity quartz crucibles may
also be usedas a substitute for expensive platinum evaporation
dishes for thisstep. Approximately 57 g of purified flux was
recovered from 60 g oforiginal LiBO2 powder. The flux purification
and recovery proceduresare summarized in Fig. 1. The effectiveness
of this procedure in re-moving REE, Sc and Y from the flux material
is discussed inSection 3.1.
2.2.2. Alkali flux fusionApproximately 1550 mg of USGS reference
materials, PAAS and
homogenized meteorite powders were fused with ~450 mg of
-
1) 60 mL 0.05 M HCl2) 20 mL MQ-Water3) 20 mL 3M HNO3
4) Flux loaded in 3M HNO3
5) Load solution: Purified FluxVacuum range: 0.66-0.79 atm. Flow
rate: 3-4 mL min-1
I. Flux dissolution II. Purification III. Recovery
Furnace dry-down in
capped 100 mL Pt dishes at 720 C for
1 hr
TODGA
TODGA
TODGA
Hot-platedry-down in 250 mL PFA beakers at 220 C for
2-3 hr
~ 60 g LiBO2
dissolvedin 2L
3M HNO3
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
TODGA
Cleaning and conditioning
Fig. 1. Schematic protocol for removal of REE, Sc and Y from
commercial lithium metaborate flux using Eichrom TODGA extraction
chromatography. About 60 g Puratronic LiBO2was dissolved in 2 L of
3 mol L1 HNO3. The flux solution was loaded onto an array of TODGA
cartridges. About 57 g of purified flux powder was recovered by
evaporating theelute in 250 mL PFA beakers at 220 C on a hotplate,
followed by heating the residual flux slabs in Pt evaporation
dishes at 720 C in a furnace.
40 A. Pourmand et al. / Chemical Geology 291 (2012) 3854
purified LiBO2 flux. Previous fusion experiments showed that a
flux:sample ratio of 6 or higher is required for complete digestion
of ter-restrial and meteoritic materials and successful dissolution
of themelt in 3 mol L1 HNO3 (Pourmand and Dauphas, 2010). The
geolog-ical reference materials were specifically selected to cover
a widerange of REE, Sc and Y concentrations, spanning three orders
of mag-nitude. Three chips of the Ivuna meteorite (200800 mg) were
sawnfrom larger pieces and cleaned in an ultrasonic bath for ~10 s
withhigh-purity ethanol. The chips were dried and powdered
separatelyin a sand-cleaned agate mortar under class-100 clean air.
Powderedaliquots of Orgueil, Alais and Tagish Lake were rinsed with
ethanoland digested without further processing. Detailed
information aboutthe terrestrial and meteoritic samples can be
found in Table 1. In ad-dition to the samples, an aliquot of the
multi-element standard solu-tion was also fused with the flux and
processed through the TODGAresin to establish chemical yields for
the REE, Sc and Y on this resin.
Approximately 450 mg of LiBO2 was weighted onto an 8-mL
high-purity graphite crucible. The sample powder was then
transferred to adepression at the center of the flux using clean
weighing paper. Thenext crucial step was the addition of about
60150 mg of a high-puri-ty LiBr non-wetting solution (Spex
Certiprep) to prevent adhesion ofthe fusion melt to the graphite
crucible and facilitate its quantitativetransfer for dissolution.
The crucible containing the flux, sample pow-der and the
non-wetting agent was capped prior to digestion to min-imize
potential contaminations. Fusion was performed at 1070 Cinside a
Thermoline furnace for 12 min and the melt was directlypoured into
a 30-mL Savillex PFA vial containing 15 mL of3 mol L1 HNO3. In
spite of the intense reaction between the meltand the liquid,
sample loss does not occur during this stage as evi-denced by
processing the multi-element standard solution (see Sec-tion 3.1).
The vial containing small fragments of the fusion materialwas then
placed on a Thermolyne Vortex at ~6000 rpm and completedissolution
was usually achieved within a few minutes. The solutionswere
inspected under the microscope and graphite particles
wereidentified as the only remaining solid residues.
2.3. TODGA extraction chromatography
High-precision measurements of REE, Sc and Y with ICP-MS
re-quire that the analytes of interest be free from matrix
elements.
Following dissolution of the fusion melt, REE, Sc and Y were
quantita-tively separated from major and trace elements using a
single-stageTODGA extraction chromatography. The distribution
coefficients(Kd = concentration in TODGA resin/concentration in
equilibratedsolution) for REE on the TODGA resin are extremely high
and exceed10,000 in 3 mol L1 HNO3 for most lanthanides. In
contrast, the Kdsof all REE fall below 1 in 0.05 mol L1 HCl
(Pourmand and Dauphas,2010). Such large differentials in the
affinity of the TODGA resin forREE make it an ideal medium for
separation of these elements fromthe rock matrix. Although Pourmand
and Dauphas (2010) did not re-port distribution coefficients for Pr
and Sc on the TODGA resin, theseelements were found to behave
similarly to other REE, as shown byrecovery of the multi-element
standard solution (see Section 3.1below).
The schematic protocol for high-purity LiBO2 fusion and
TODGAextraction chromatography of REE, Sc and Y is shown in Fig. 2.
Briefly,a 10-mL Bio-Rad polypropylene reservoir, a 2-mL TODGA
cartridgeand two disposable connectors were assembled on the vacuum
cham-ber in tandem. Vacuumwas regulated at 0.790.98 atm with a
single-stage Venturi pump (Mcmaster Carr, Cat# 41605 K13). Flow
rates forload/rinse solutions and elution of REE, Sc and Y were
maintainedbelow 2 and 3 mL min1, respectively, by adjusting the
vacuum.The 2-mL TODGA cartridge was preconditioned with 20 mL
of0.05 mol L1 HCl to remove REE, Sc and Y from the resin,
followedby 10 mL of MQ water and 10 mL of 3 mol L1 HNO3 to convert
tothe loading acid solution. The sample was loaded to the reservoir
in15 mL 3 mol L1 HNO3 and was followed by 12 mL of 3 mol L1
HNO3 to remove matrix elements. Residual major and transition
ele-ments, including Ca, were stripped in 15 mL of 11 mol L1 HNO3.
AllREE, Sc and Y were subsequently eluted in 30 mL of 0.05 mol L1
HClinto 50 mL centrifuge tubes for the reference materials and
PAAS, and30-mL PFA Savillex vials for meteorites. The solutions
containing thereference materials and PAAS were either
gravimetrically diluted to0.45 mol L1 HNO3 or evaporated and
diluted by a factor of 20200for analysis. Solutions containing REE,
Sc and Y for meteorites wereevaporated to 12 L and diluted in 2 mL
of 0.45 mol L1 HNO3prior to MC-ICP-MS analysis. In the case of
direct dilution, care mustbe taken to ensure that the concentration
of nitric acid in the samplesolution remains identical to the
bracketing standard. Otherwise,measured concentrations may be
inaccurate by as much as 10%.
-
Table 1USGS reference materials, Post-Archean Australian Shales
and chondritic meteorites analyzed in this study. Aliquots of Ivuna
were from homogenizing three separate chips thatweighed between 0.2
and 0.8 g. Orgueil AE came from homogenizing a chip of ~0.2 g, and
Orgueil F and G from homogenizing one chip of 219 and two chips of
234, respectively.Digested powders weighed ~0.2 g. Total mass of
Tagish Lake was 0.015 g, which came from homogenizing 34 g of this
meteorite. The dissolved mass for all other samples was
ap-proximately 0.05 g.
Sample type Group Source Collection ID
USGS reference materialBCR-2 (AB, n=2) Basalt, Columbia River,
Oregon, USA BHVO-1 Basalt, Kilauea, Hawaii, USA BIR-1 (AB, n=2)
Basalt, Reykavik Dolerite, Iceland PCC-1 Peridotite, Austin Creek,
California, USA W-2 Diabase, Bull Run Quarry, Virginia, USA G-2
(AC, n=3) Granite. Sullivan Quarry, Rhode Island, USA G-3 (AE, n=5)
Granite. Sullivan Quarry, Rhode Island, USA
Post-Archean Australian Shales (PAAS) AO-6 Amadeus Basin,
Australia AO-7 Amadeus Basin, Australia AO-9 Amadeus Basin,
Australia AO-10 Amadeus Basin, Australia AO-12 Amadeus Basin,
Australia SC-7 Camp Hill, Canberra, Australia SC-8 Camp Hill,
Canberra, Australia PL-1 Canning Basin, Australia PW-5 Perth Basin,
Australia
MeteoritesAlais CI1 Field Museum C3_0067Allende A CV3 USNM,
split 8, position 5 3529Allende B CV3 USNM, split 8, position 5
3529Ivuna A CI1 USNM 6630Ivuna B CI1 USNM 6630Ivuna C CI1 USNM
6630Orgueil A CI1 MNHN 219Orgueil B CI1 MNHN 219Orgueil C CI1 MNHN
219Orgueil D CI1 Universty of Chicago collection C3_1146Orgueil E
CI1 Universty of Chicago collection C3_1146Orgueil F CI1 MNHN
219Orgueil G CI1 MNHN 234Tagish Lake C2-ung. Private collection
41A. Pourmand et al. / Chemical Geology 291 (2012) 3854
2.4. Multi-collection ICP-MS analysis of REE, Sc and Y
Elemental concentration measurements were performed on
Ther-moScientificNeptuneMC-ICP-MS instruments at theOrigins
Laboratoryof the University of Chicago and the Neptune Isotope
Laboratory of TheUniversity of Miami. A comprehensive description
of the Neptune in-strument can be found in Wieser and Schwieters
(2005). The instru-ment's 9 Faraday collectors were utilized for
measuring REE, Sc and Yconcentrations by standard-sample bracketing
technique. The standardsolutions used to determine elemental
concentrations were gravimetri-cally prepared from twobatches of
certifiedmulti-element solution thatcontained all 14 REE, Sc and Y
at ~10 g g1. The concentrations of allREE in the multi-element
solutions were certified by SPEX CertiPrepand the results obtained
using the two standard batches agree. Sampleand standard solutions
were introduced to the plasma in 0.45 mol L1
HNO3 through an Apex-Q+Spiro TMD desolvating nebulizer
(Elemen-tal Scientific Inc.) using a 100 L min1 self-aspirating PFA
nebulizer.Argon and nitrogen were used as carrier and sweep gasses
for the des-olvation inlet system.
One of the main difficulties in measuring elemental
concentra-tions and isotopic ratios of REE by ICP-MS has been the
presenceof polyatomic interferences with similar mass/charge ratios
as REEanalytes, e.g., Ln+, LnO+ and LnOH+ (Evans and Giglio, 1993;
Dulski,1994; Eggins et al., 1997; Baker et al., 2002; Raut et al.,
2003). Wetook specific measures to minimize contributions from
these interfer-ences; the REE isotopes listed in Table 2 were
carefully selected toeliminate the influence of isobars. Following
matrix removal, REE ox-ides remain as the only interfering species
on the selected isotopes ofHREE during MC-ICP-MS analysis. Prior to
measuring the samples,oxide contributions were determined by
analyzing mono-elemental
standard solutions of REE. In order to detect and reliably
measurethe oxides, the standard solutions were analyzed at a high
concentra-tion of ~250 ng g1 with the Apex-Q+Spiro TMD desolvating
sys-tem. As shown in Table 2, intensity ratios of REE oxide/REE
rangedfrom 0.0034% for EuO+/Eu to 0.0976% for NdO+/Nd, with an
averageof 0.03% across all REEs. At these levels, the oxide
corrections on themeasured isotopes in natural samples were
negligible and no correc-tions were required.
In order to measure 14 REE, Sc and Y in a single sample
solution, anovel dynamicmethodwas developed
usingfivemulti-collector config-urations. The arrangement of
representative REE, Sc and Y isotopes andcorresponding collector
configurations are shown in Table 3. A uniqueaspect of this
technique is that 149Sm and 167Er isotopes are measuredin three
collector configurations to allow normalization of LREE andHREE to
middle REE (Sm-Er). This set-up is greatly advantageous
oversingle-collector ICP-MS, as it minimizes the effect of plasma
instabilityand fluctuations in the desolvation inlet system on REE
ratiosmeasuredby standard-sample bracketing. Scandium-45 and 89Y
were also mea-sured in two additional sub-configurations (3 and 4).
Accommodatinga wide mass range that encompass 14 REE in 3 cup
configurations wasmade possible by adjusting the zoom optics (Focus
and Dispersion)and source lens parameters (Focus, Deflection and
Shape) of theNeptune.
A gain calibration was performed for Faraday collectors at the
be-ginning of each session. The analysis sequence began with
measuringion intensities of each analyte in 0.45 mol L1 HNO3
solution. Thesame solution was also used for sample preparation and
standard di-lutions. The data collection method consisted of 1
block of 5 cycleswith 4.2 s integration time and 3 s of idle time
(5 s for Sc) in dynamicmode. A take-up time and signal
stabilization of 90 s was
-
Table 2The REE, Sc and Y isotopes used for measuring
concentrations by MC-ICP-MS. Isotopeswere selected specifically to
avoid isobaric interferences. The contribution of oxide
in-terferences to the intensity signal of REE was assessed by
measuring REEO+/REE ratiosin ~250 ng g1 monoelemental standard
solutions via ESI Apex Q+Spiro TMD deso-lvating nebulizer system.
Average REEO+/REE contribution in standard solutions was0.03% and
demonstrates the effectiveness of desolvation in removing REE
oxides.Oxide interferences were negligible in the samples, which
were analyzed at 10100times lower concentrations compared with the
standards used to determine oxidelevels.
Isotope Isotope abundance (%) Interfering oxide Oxides (%)
45Sc 100.00 89Y 100.00 139La 99.91 140Ce 88.45 141Pr 100.00
146Nd 17.20 149Sm 13.82 151Eu 47.81 157Gd 15.65 141Pr16O
0.0762159Tb 100.00 143Nd16O 0.0976163Dy 24.90 147Sm16O 0.0067165Ho
100.00 149Sm16O 0.0067167Er 22.93 151Eu16O 0.0034169Tm 100.00
153Eu16O 0.0034173Yb 16.13 157Dy16O 0.0127175Lu 97.41 159Tb16O
0.0325
1) 20 mL 0.05 M HCl2) 10 mL MQ water3) 10 mL 3 M HNO3
4) Sample load in 3 M HNO3
5) Load solution: Matrix6) 12 mL 3 M HNO3: Matrix7) 15 mL 11 M
HNO3: Ca8) 30 mL 0.05 M HCl: REEs
TOD
GA
Vacuum range: 0.79-0.98 atm. Flow rates: load and elution at
< 2 mL min-1
Rinse at 3 mL min-1
Dissolve in 15 mL
3 M HNO3
Highpurityflux
fusion
I. Flux Fusion II. Matrix separation
Flux (~ 450 mg) +Sample (~ 50 mg) +60-150 mg LiBr solution in 8
mL graphite crucibles at 1070 C for 12 min
Cleaning and conditioning
Direct transfer of melt
Fig. 2. Schematic protocol for high-temperature LiBO2 flux
fusion and separation of REE, Scand Y from interfering matrix with
TODGA extraction chromatography. A homogenizedaliquot of the sample
powder was fused at 1070 C with purified LiBO2 flux for 12 minin
capped, high-purity graphite crucibles. About 60150 mg of LiBr
non-wetting solutionwas used to facilitate quantitative transfer of
the fusion melt. The melt was directly dis-solved in 15 mL of 3 mol
L1 HNO3 and loaded to a pre-conditioned TODGA cartridge. Fol-lowing
matrix removal, REE, Sc and Y were eluted in 30 mL of 0.05 mol L1
HCl.
42 A. Pourmand et al. / Chemical Geology 291 (2012) 3854
implemented prior to data collection. The inlet system was
rinsedwith 0.45 mol L1 HNO3 for 2 min between each run. This
washtime was sufficient to eliminate memory effects from previous
sam-ple and standard analyses. Signal intensities from the dilution
acidwere subtracted online from procedural blanks, standard and
sampleruns at the beginning of each sequence. Every sample
measurementwas bracketed with two multi-element standard solution
analysesand the REE, Sc and Y concentrations were calculated
according tothe following relationship:
CA CS IS=IA 1
where CA and CS represent the concentrations in the sample and
themulti-element standard, respectively, and IA and IS represent
the in-tensities of the ion beams registered at the faraday
detectors. Concen-trations are reported as the mean of up to 3
measurements in thesame sample solution over the course of the run
session. All uncer-tainties in this study are reported as relative
standard deviation inpercent (100*SD/ x, SD=standard deviation,
x=average of sepa-rate digestions) unless otherwise noted.
3. Results and discussion
3.1. Procedural blanks and elemental recovery
In order to examine the effectiveness of flux purification in
reduc-ing background levels of REE, Sc and Y, concentrations in
purified andcommercial flux materials were determined. About 1 g of
commercialand purified LiBO2 was dissolved in 25 mL of 3 mol L1
HNO3 and theanalytes were separated following the TODGA extraction
chromatog-raphy procedure presented in Fig. 2. The blank results,
shown inTable 4, reveal a dramatic decrease in nearly all REE, Sc
and Y concen-trations from commercial to purified LiBO2. The signal
intensities ofelements such as Gd, Tm and Lu were below the
detection limits of
the Faraday cups in the purified flux solution and improvements
forother elements ranged from a factor of ~6 for Ho to more than
6300for La.
During the course of this study, procedural blanks were
processedand analyzed similarly to the samples using ~0.45 g of
purified flux. Ref-erence materials, PAAS and meteorite samples
were processed in threeseparate batches with blank levels slightly
elevated for the secondand third batches. Total procedural blanks
(Table 4) are based on thegeometric mean for 6 replicate
measurements. The REE and Y abun-dances in the peridotite reference
material PCC-1 are considerablylower compared with most (extra)
terrestrial materials, and requiredrelatively higher blank
adjustments (7.4% on average). Proceduralblank contributions to all
other reference materials were up to twoorders of magnitude smaller
for most elements, with highest adjust-ments made for G-3 at 0.05%
on average. Average blank contributionfor PAAS samples was 0.07%.
Adjustments to Allende (0.14%), TagishLake (0.86%), Ivuna (0.4%),
Orgueil (0.31%) and Alais (1.2%) were alsoimplemented prior to
calculating final concentrations.
In order to test the recovery of REE, Sc and Y, an aliquot of
themulti-element standard solution (~10 g g1, see Table 4 for
concen-tration details) was added to ~0.45 g of purified flux and
processedthrough fusion and extraction chromatography similarly to
the restof the samples. Elemental concentrations, compared against
thesame multi-element standard solution, demonstrate quantitative
re-covery for all elements (Table 4).
3.2. REE in geological reference materials and PAAS
In order to assess the accuracy of our analytical methodology,
REE,Sc and Y concentrations from this study are compared with the
meanof literature values for USGS reference materials BHVO-1
(n=11),BIR-1 (n=10), BCR-2 (n=6), PCC-1 (n=10) W-2 (n=7), G-2(n=10)
and G-3 (n=1) referencematerials in Table 5.
Literaturemea-surements were acquired by quadrupole, sector field
high-resolution(HR-ICP) andmulti-collectionmass spectrometers using
standard addi-tion/bracketing or isotope dilution techniques.With
the exception of G-3, the abundances of elements in geological
reference materials fromthis study are similar to the mean of
literature values within analyticaluncertainties, and demonstrate
the accuracy of our technique (Fig. 3).In the case of G-3, we found
only one study that reports REE and Y
-
Table 3Faraday cup positions and zoom optics of the Neptune
MC-ICP-MS. Measurement of 14 REE, Sc and Y was made possible by
utilizing 9 faraday cups and a dynamic collector con-figuration.
The intensity of REE measured in sub-configurations 1 and 2 were
normalized to 149Sm and 167Er intensities from the Main
configuration. This technique minimizes theeffect of fluctuations
inherent in the plasma and the inlet system and improves
precisions. The use of zoom optics is essential to optimize peak
overlaps for 14 lanthanide isotopes. Atotal of 0.5 mL solution was
consumed during each measurement using a 100 L min1 PFA nebulizer.
Note that Focus and Dispersion voltages can vary depending on cup
config-urations of the Neptune.
Configuration L4 L3 L2 L1 Axial H1 H2 H3 H4 Focus (V) Dispersion
(V)
Main 149Sm 151Eu 157Gd 159Tb 163Dy 165Ho 167Er 2.0 0.0Sub
config. 1 139La 140Ce 141Pr 146Nd 149Sm 15.0 67.5Sub config. 2
167Er 169Tm 173Yb 175Lu 10.0 38.6Sub config. 3 45Sc 0.0 0.0Sub
config. 4 89Y 0.0 0.0
43A. Pourmand et al. / Chemical Geology 291 (2012) 3854
concentrations for this referencematerial; our concentrations
appear tobe higher compared with those reported in Meisel et al.
(2002) for thisreference material (Table 5.g). Considering
quantitative recovery of allelements based on processing a
multi-element standard solutionthrough fusion and TODGA extraction
chromatography (Table 4), accu-rate concentrations for a nearly
identical reference material, G-2, com-pared with values from 10
literature measurements (Table 5.f), andconsistency between five
replicate analyses of G-3 (Table 5.g), we areconfident that the
concentrations of G-3 presented in this study are ac-curate. The
uncertainties on the mean of five replicate analyses of G-3range
between RSD=1.7% for Sm and RSD=3.1% for Sc, with an aver-age of
~2% for all elements, and represent the overall precision of
ouranalytical technique.
Baker et al. (2002) reported high-precision concentrations
ofsome REE in BHVO-1, BCR-2 and BIR-1 by MC-ICP-MS. Those
valueswere later revised by Kent et al. (2004) using a similar
method ofacid digestion, ion exchange chromatography and isotope
dilutionmass spectrometry. In spite of major differences between
the analyt-ical techniques employed by Kent et al. (2004), and the
protocol pre-sented here (flux fusion, extraction chromatography
and standard-sample bracketing), the arithmetic mean of REE
concentrations forBHVO-1 and BIR-1 (Table 5.ab) from this study are
comparablewith those of Kent et al. (2004) within 0.4% and 0.2%,
respectively.In the case of BCR-2 and W-2 (Table 5.c and e), the
average concen-trations of REE, Sc and Y between the two studies
are in agreementwithin 1% and 2%, respectively. Although, in
principle, better preci-sions can be achieved by isotope dilution
compared with standard-
Table 4Blanks and yields after flux fusion and extraction
chormatography separation of REE, Scand Y using the TODGA resin.
Concentrations in a gram of commercial (Alfa Aesar Pur-atronic
LiBO2, 99.997% trace metals) and purified flux materials
demonstrate major im-provements for Y, La, Ce, Gd and Yb after
purification. Procedural blanks are based onthe fusion of
approximately 0.45 g of purified flux in high-purity graphite
crucibles.bDet. = Below the detection limit of the Neptune Faraday
cups. MES: multi-elementstandard solution of REE, Sc and Y.
Element PuratronicLiBO2(pg g1)
PurifedLiBO2(pg g1)
Proceduralblank(n=6)(pg)
MES(g g1)
MES after fluxfusion and TODGAseparation(g g1)
Yield(%)
Sc b Det. b Det. 91 10.00 10.24 102.4Y 3089 13 472 9.95 10.02
100.7La 83,206 13 120 9.88 9.94 100.7Ce 6087 31 247 10.00 10.07
100.7Pr 89 4 53 10.00 10.07 100.7Nd 295 33 225 9.96 10.03 100.7Sm
66 8 75 9.88 9.95 100.7Eu 217 1 28 10.00 10.07 100.7Gd 2499 b Det.
88 9.97 10.04 100.7Tb 97 1 26 10.00 10.07 100.7Dy 98 7 84 10.00
10.06 100.6Ho 7 1 26 9.99 10.05 100.6Er 54 1 57 9.96 10.01 100.5Tm
3 b Det. 16 9.91 9.97 100.6Yb 644 2 58 10.00 10.06 100.6Lu 45 b
Det. 18 9.94 9.99 100.5
sample bracketing technique, preparation and calibration of
REEspikes can be more demanding than the multi-collection
dynamicmethod presented here. Furthermore, mono-isotopic elements
(Pr,Tb, Ho, Tm, Sc and Y) cannot be directly measured by IDMS.
Additionof enriched spikes can also introduce interferences on some
of themeasured isotopes (for example 149Sm and 153Eu spikes can
causeoxide interferences on 165Ho and 169Tm, respectively, see
Table 2).
REE abundances of Post-Archean Australian Shales are often
usedfor normalization of REE patterns (Nance and Taylor, 1976;
Taylorand McLennan, 1985). These abundances were obtained by
spark-source mass spectrometry, which suffers from relatively poor
preci-sion. In addition, some REEs in PAAS (Tm and Yb) were not
actuallymeasured but instead, the abundances of these elements were
de-rived by interpolation between neighbor REEs (Taylor and
McLennan,1985). We have analyzed nine PAAS samples from four basins
acrossAustralia (Table 1) using our new MC-ICP-MS technique and the
re-sults, which include measured abundances of Tm and Yb, are
pre-sented in Table 6. The normalized REE pattern for PAAS from
thisstudy are smoother and overall, our data is of higher quality
com-pared with those reported by Nance and Taylor (1976) (Fig. 4).
Werecommend that these new values of PAAS be used for
normalizationof REE concentration measurements.
Recently, Bendel et al. (2011) studied the REE patterns in a
num-ber of bulk terrestrial rocks and raised the possibility of Tm,
and toa lesser degree, Yb anomalies. On close inspection of the
data, itcould be argued that a small Tm depletion may exist (Fig.
4) butgiven the analytical uncertainty of ~2% for Tm based on
replicates ofG-3 reference material, and the uncertainties in the
mean of CI-chon-drites (see Section 3.4), the potential Tm
depletion cannot be con-vincingly resolved in these PAAS
samples.
3.3. REE in carbonaceous chondrites
With the exception of volatile elements and Li that is burnt in
theSun, the concentrations of most elements in CI carbonaceous
chon-drites represent our best estimate of the solar nebula
composition(Haskin et al., 1966; Evensen et al., 1978; Palme, 1988;
Anders andGrevesse, 1989; Evans and Giglio, 1993; Lodders et al.,
2009). Thesample digestion, extraction chromatography and
multi-collectionICP-MS technique developed during this study were
employed tomeasure REE, Sc and Y abundances in Allende (CV3),
Tagish Lake(CI2-ungrouped), Orgueil (CI1), Alais (CI1) and Ivuna
(CV1). Elemen-tal concentrations determined for these meteorites
are compiled inTables 7 and 8 and their REE fractionations relative
to the mean ofCI-chondrites will be discussed in the following
sections. Resultsfrom this study will be also compared with
previously published,high-precision values where data is
available.
3.3.1. Allende (CV3)Two aliquots of bulk Allende powder from the
Smithsonian Insti-
tute were analyzed and the results are shown in Table 7. A
compari-son between REE abundances (Fig. 5), normalized to the mean
ofCI-chondrites from this study, with bulk analysis of this
meteorite
-
Table 5Rare-earth element, Y and Sc concentrations (g g1) in a)
BHVO-1, b) BIR-1, c) BCR-2, d) PCC-1, e) W-2, f) G-2 and g) G-3
reference material from this study are compared with literature
values measured by Quadrupole (Q), single-collectorhigh-resolution
(SC), and multi-collection (MC) ICPMS techniques. RSD% = Relative
standard deviation (100SD/average) for multiple analyses of the
same sample. The concentrations of REE in PCC-1 were particularly
difficult to measuredue to their exceptionally low abundances and
relatively higher blank corrections. Concentrations of
mono-isotopic elements (Pr, Tb, Ho and Tm) from the literature are
not directly measured by isotope dilution mass spectrometry.
a. BHVO-1
Element [1] RSD% [2] RSD% [3] RSD% [4] RSD% [5] RSD% [6] RSD%
[7] RSD% [8] RSD% [9] RSD% [10] RSD% [11] RSD% Lit. Mean RSD% This
study
Sc 31.00 3.5 31.90 1.8 30.40 0.8 29.70 7.7 32.60 5.2 31.00 4.7
31.10 3.3 31.00Y 24.00 4.2 25.30 1.5 24.90 1.9 26.40 0.4 24.60 4.5
26.30 0.3 22.70 1.7 26.48 1.2 27.40 3.9 25.34 5.8 24.14La 15.50 3.6
15.31 0.5 15.50 1.5 16.00 1.0 15.41 0.3 15.40 0.5 15.60 4.0 15.10
1.1 14.90 0.9 15.13 1.9 16.00 1.4 15.44 2.2 15.08Ce 38.00 3.2 38.40
0.8 38.10 1.5 39.00 3.0 38.38 0.3 37.90 0.4 38.60 4.0 37.70 0.8
37.30 0.5 38.23 1.3 39.40 4.7 38.27 1.5 38.20Pr 5.500 3.7 5.460 0.8
5.590 1.6 5.650 1.7 5.411 0.3 5.300 0.5 5.270 4.0 5.260 0.7 5.200
0.8 5.580 0.7 5.720 1.7 5.449 3.2 5.390Nd 25.00 4.6 24.65 0.7 24.50
2.2 25.10 2.2 24.79 0.2 24.70 0.6 25.10 2.5 24.60 0.9 24.80 0.3
24.76 1.3 26.10 2.8 24.92 1.8 25.07Sm 6.230 4.9 6.070 1.1 6.000 1.9
6.260 3.1 6.097 0.3 6.130 0.8 6.270 3.0 6.130 0.6 6.040 1.3 6.300
1.9 6.360 2.2 6.172 1.9 6.089Eu 2.140 3.4 2.060 0.7 2.100 1.8 2.120
2.0 2.067 0.3 2.100 1.6 2.060 3.0 2.140 0.8 2.060 0.8 2.060 2.4
2.090 4.1 2.091 1.5 2.103Gd 6.350 6.6 6.260 1.8 6.600 2.4 6.260 3.1
6.300 0.3 6.390 2.3 6.300 3.0 6.390 1.3 6.520 0.8 5.710 1.9 6.280
2.2 6.305 3.6 6.317Tb 0.9400 5.2 0.9490 0.9 0.9400 3 0.9700 3.8
0.9590 0.8 0.9500 5.5 0.9260 1.2 0.9450 0.9 0.9300 0.0 1.0000 4.1
0.9509 2.3 0.9360Dy 5.280 5.3 5.260 0.9 5.400 2.3 5.340 2.6 5.363
0.3 5.380 1.8 5.410 3.5 5.340 1.3 5.440 1.2 5.280 1.7 5.510 1.8
5.364 1.4 5.394Ho 1.010 6.4 1.010 0.9 0.980 2.8 1.020 4.0 0.969 1.7
0.970 4.0 0.962 0.7 1.000 1.4 1.040 1.9 1.050 3.2 1.001 3.1 0.984Er
2.570 6.4 2.530 0.5 2.550 1.9 2.590 2.4 2.575 0.3 2.540 0.9 2.560
5.0 2.560 0.4 2.560 1.3 2.510 2.0 2.530 4.4 2.552 0.9 2.562Tm
0.3400 5.4 0.3400 2.9 0.3270 3.2 0.3300 6.5 0.3280 0.9 0.3320 0.9
0.3400 2.9 0.3500 4.8 0.3359 2.4 0.3320Yb 2.000 5.7 1.990 0.7 2.100
4.9 1.990 3.5 1.978 0.3 1.990 1.8 2.040 4.0 1.940 0.7 2.010 1.2
2.060 1.9 2.150 3.5 2.023 3.0 1.998Lu 0.2800 6.8 0.2850 1.2 0.2800
2.4 0.2800 3.7 0.2739 0.3 0.2710 2.1 0.2700 5.0 0.2760 1.4 0.2830
0.9 0.2900 3.4 0.3100 3.4 0.2817 3.9 0.2704[1] Robinson et al.
(1999) SC-ICP-MS, [2] Barrat et al. (2000) ICP-QMS, [3] Dulski
(2001) ICP-QMS, [4] Yu et al. (2001) SC-ICP-MS, [5] Kent et al.
(2004) MC-ICP-MS, [6] Willbold and Jochum (2005) HR-ICP-MS, [7]
Debaille et al. (2006) ICP-QMS, [8] Dia et al. (2006) ICP-QMS, [9]
Makishima and Nakamura (2006) ICP-QMS, [10] Huang et al. (2007)
ICP-QMS, [11] Hu and Gao (2008) ICP-QMS.
b. BIR-1
Element [1] RSD% [2] RSD% [3] RSD% [4] RSD% [5] RSD% [6] RSD%
[7] RSD% [8] RSD% [9] RSD% [10] RSD% Lit.Mean
RSD% BIR-1 A BIR-1 B This study(n=2)
Sc 44.10 1.6 43.80 1.9 42.00 3.1 46.00 5.0 43.98 3.7 42.49 42.98
42.74Y 16.28 0.9 16.20 0.6 14.10 1.8 14.90 3.5 16.11 4.0 15.10 5.3
15.20 2.0 17.00 3.6 15.61 6.0 14.15 14.60 14.38La 0.6030 0.7 0.6040
1.2 0.6240 2.6 0.6300 5.5 0.6200 3.7 0.6330 4.0 0.6050 0.2 0.6100
4.9 0.6040 2.0 0.6000 5.0 0.6133 2.0 0.5968 0.5839 0.5904Ce 1.888
0.2 1.897 0.9 1.890 0.6 1.890 5.3 1.890 2.5 1.892 3.0 1.905 1.920
4.2 1.890 1.9 1.910 3.2 1.897 0.6 1.894 1.926 1.910Pr 0.3750 0.8
0.3780 1.1 0.3680 0.6 0.3800 4.6 0.3900 3.0 0.3704 5.0 0.3680 0.0
0.3800 2.6 0.3740 1.8 0.3720 1.5 0.3755 1.8 0.3685 0.3761 0.3723Nd
2.360 0.9 2.380 1.0 2.390 0.5 2.310 3.8 2.350 3.1 2.371 4.0 2.380
0.1 2.410 2.5 2.370 1.1 2.400 1.2 2.372 1.2 2.380 2.432 2.406Sm
1.115 1.7 1.117 1.5 1.090 0.3 1.070 2.6 1.080 4.9 1.069 1.0 1.092
0.1 1.100 1.8 1.090 1.4 1.102 0.7 1.092 1.5 1.073 1.101 1.087Eu
0.5210 0.9 0.5240 1.0 0.5150 0.3 0.520 3.4 0.5300 2.6 0.5035 3.0
0.5217 0.2 0.5200 5.8 0.5080 1.9 0.5300 1.1 0.5193 1.6 0.5198
0.5341 0.5270Gd 1.864 0.7 1.850 1.0 1.870 0.6 1.770 2.5 1.910 3.0
1.650 5.0 1.887 0.1 1.800 11.1 1.790 0.9 1.810 1.3 1.820 4.1 1.821
1.907 1.864Tb 0.3790 1.4 0.3790 1.3 0.3600 0.4 0.350 3.5 0.3600 3.0
0.3574 4.0 0.3500 5.7 0.3990 0.9 0.3660 1.2 0.3667 4.4 0.3553
0.3661 0.3607Dy 2.520 0.7 2.530 0.9 2.550 0.3 2.430 2.8 2.550 3.5
2.489 0.5 2.592 0.1 2.500 4.0 2.520 0.7 2.590 1.0 2.527 1.9 2.563
2.640 2.602Ho 0.5800 0.9 0.5850 1.1 0.5680 0.3 0.5500 3.5 0.5600
3.9 0.5521 2.0 0.5600 3.6 0.5590 0.6 0.5910 1.3 0.5672 2.6 0.5642
0.5801 0.5721Er 1.731 0.9 1.734 0.8 1.650 0.2 1.640 3.1 1.710 4.8
1.620 5.0 1.738 0.1 1.570 5.1 1.680 0.6 1.740 1.5 1.681 3.5 1.710
1.756 1.733Tm 0.2620 0.2 0.2400 4.7 0.250 4.4 0.2587 11.0 0.250 8.0
0.2400 1.8 0.2501 3.7 0.2486 0.2561 0.2523Yb 1.643 1.2 1.649 1.1
1.670 0.4 1.610 2.7 1.640 3.5 1.670 9.0 1.653 0.2 1.540 3.2 1.620
1.1 1.630 1.8 1.633 2.3 1.642 1.686 1.664Lu 0.2480 1.3 0.2470 1.0
0.2690 0.2 0.2400 3.9 0.2500 3.9 0.2635 3.0 0.2461 0.2 0.220 4.5
0.2410 1.0 0.2430 2.2 0.2468 5.4 0.2421 0.2497 0.2459[1]
Garbe-Schonberg (1993) ICP-QMS, [2] Eggins et al. (1997) ICP-QMS,
[3] Pin and Joannon (1997) ICP-QMS, [4] Robinson et al. (1999)
SC-ICP-MS, [5] Dulski (2001) ICP-QMS, [6] Coogan et al. (2004)
ICP-MS, [7] Kent et al. (2004) MC-ICP-MS, [8] Melluso et al. (2005)
ICP-MS, [9] Willbold and Jochum (2005) SC-ICP-MS, [10] Bayon et al.
(2009) HR-ICP-MS.
c. BCR-2
Element [1] RSD% [2] RSD% [3] RSD% [4] RSD% [5] RSD% [6] RSD%
Lit. Mean RSD% BCR-2 A BCR-2 B This study (n=2)
Sc 32.00 6.3 32.50 4.9 33.41 1.0 36.00 4.9 33.48 5.3 31.81 32.61
32.21Y 35.30 3.8 37.33 3.4 36.15 0.6 37.05 1.4 40.10 2.4 37.19 4.9
33.54 32.66 33.10
44A.Pourm
andet
al./Chem
icalGeology
291(2012)
3854
-
c. BCR-2
Element [1] RSD% [2] RSD% [3] RSD% [4] RSD% [5] RSD% [6] RSD%
Lit. Mean RSD% BCR-2 A BCR-2 B This study (n=2)
La 24.94 0.5 25.60 1.2 25.19 2.4 24.59 1.3 24.36 2.7 26.00 2.7
25.11 2.5 25.14 24.50 24.82Ce 53.52 0.5 55.30 0.5 53.23 4.2 53.30
1.7 52.21 1.3 55.40 2.2 53.83 2.3 54.34 52.61 53.48Pr 6.756 0.7
6.760 0.5 6.900 2.9 7.040 1.0 6.797 1.6 7.070 0.6 6.887 2.0 6.922
6.675 6.798Nd 28.71 0.5 28.60 0.5 29.19 2.0 28.50 1.1 28.25 1.5
29.67 0.4 28.82 1.8 29.43 28.37 28.90Sm 6.534 0.5 6.410 0.2 6.590
1.7 6.680 1.5 6.698 1.3 6.760 2.7 6.612 1.9 6.604 6.441 6.522Eu
1.949 0.6 2.060 2.3 1.920 2.1 1.990 1.5 1.921 0.9 2.010 2.9 1.975
2.8 2.011 1.917 1.964Gd 6.727 0.6 6.680 0.8 6.690 2.2 6.240 1.4
6.838 1.2 7.070 0.5 6.708 4.0 6.863 6.716 6.790Tb 1.060 0.4 1.050
1.0 1.040 1.0 1.059 1.2 1.090 1.0 1.060 1.8 1.062 1.029 1.046Dy
6.441 0.4 6.330 0.7 6.440 2.0 6.200 1.8 6.607 1.2 6.580 0.8 6.433
2.4 6.553 6.335 6.444Ho 1.260 2.4 1.350 1.5 1.330 0.8 1.328 1.8
1.350 1.5 1.324 2.8 1.296 1.272 1.284Er 3.707 0.5 3.620 1.6 3.670
1.1 3.480 0.9 3.650 1.3 3.770 1.0 3.650 2.7 3.642 3.641 3.642Tm
0.5120 1.1 0.5200 1.9 0.516 1.1 0.5149 0.5159 0.5154Yb 3.348 0.5
3.360 2.4 3.350 2.7 3.350 1.5 3.436 0.9 3.400 2.0 3.374 1.1 3.318
3.375 3.347Lu 0.4961 0.4 0.5040 1.0 0.4800 4.2 0.5000 0.5080 0.9
0.5000 3.2 0.498 1.9 0.4817 0.4967 0.4892[1] Kent et al. (2004)
MC-ICP-MS, [2] Willbold and Jochum (2005) SC-ICP-MS, [3] Barrat et
al. (2007) SC-ICP-MS, [4] Huang et al. (2007) ICP-QMS, [5] Mori et
al. (2007) ICP-QMS, [6] Bayon et al. (2009) SC-ICP-MS.
d. PCC-1
Element [1] RSD%
[2] RSD%
[3] RSD%
[4] RSD%
[5] RSD%
[6] RSD%
[7] RSD%
[8] RSD%
[9] RSD%
[10] RSD%
Lit. Mean RSD%
This study
Sc 9.000 7.6 8.0000 4.7 9.650 6.2 8.883 9.4 7.413Y 0.0870 5.3
0.0790 3.4 0.0770 0.7 0.0784 2.4 0.0770 2.3 0.0692 14.0 0.0781 7.3
0.0783La 0.0390 3.8 0.0290 7.4 0.0460 10.7 0.0330 3.0 0.0290 6.9
0.0290 3.4 0.0480 9.0 0.0501 2.2 0.0309 1.5 0.0315 9.0 0.0366 23.3
0.0348Ce 0.0570 1.8 0.0530 9.7 0.0528 1.8 0.0600 1.7 0.0540 3.7
0.0530 4.4 0.0710 6.0 0.0822 4.5 0.0598 0.6 0.0570 9.0 0.0600 15.9
0.0578Pr 0.0085 5.9 0.0068 8.3 0.0076 9.0 0.0076 1.3 0.0070 7.1
0.0076 3.6 0.0098 7.5 0.0088 2.4 0.0081 4.4 0.0065 13.0 0.0078 12.8
0.0075Nd 0.0300 3.3 0.0250 6.4 0.0260 2.5 0.0260 2.7 0.0300 6.7
0.0270 5.1 0.0320 4.0 0.0321 9.0 0.0256 1.5 0.0282 9.0 0.0282 9.5
0.0290Sm 0.0080 18.8 0.0050 50.2 0.0070 19.5 0.0049 8.2 0.0070 21.4
0.0060 13.6 0.0070 8.0 0.0060 4.9 0.0079 8.8 0.0053 19.0 0.0064
17.8 0.0061Eu 0.0018 5.6 0.0011 40.8 0.0009 22.7 0.0009 11.1 0.0017
17.6 0.0010 16.8 0.0020 8.0 0.0010 5.8 0.0006 4.3 0.0011 10.0
0.0012 37.6 0.0012Gd 0.0080 6.3 0.0061 19.7 0.0059 16.9 0.0053 9.4
0.0080 25.0 0.0055 4.1 0.0076 7.7 0.0053 5.9 0.0089 1.1 0.0062 29.0
0.0067 19.8 0.0070Tb 0.0015 6.7 0.0012 25.0 0.0011 16.7 0.0010 10.0
0.0014 14.3 0.0011 8.0 0.0016 6.2 0.0011 8.1 0.0015 5.6 0.0011 21.0
0.0013 17.4 0.0012Dy 0.0130 3.8 0.0087 14.1 0.0110 13.9 0.0085 5.9
0.0110 4.5 0.0095 7.1 0.0120 5.0 0.0096 4.4 0.0109 5.8 0.0099 9.0
0.0104 13.8 0.0100Ho 0.0038 1.3 0.0027 13.8 0.0030 7.6 0.0023 4.3
0.0030 13.3 0.0026 8.1 0.0031 7.6 0.0028 8.0 0.0031 1.3 0.0029 10.0
0.0029 13.5 0.0028Er 0.0123 4.1 0.0113 10.8 0.0117 8.3 0.0100 6.0
0.0130 15.4 0.0115 3.0 0.0120 4.0 0.0111 7.2 0.0110 0.1 0.0115 16.0
0.0115 7.0 0.0115Tm 0.0025 6.0 0.0028 7.4 0.0020 5.0 0.0027 5.6
0.0024 2.4 0.0027 7.7 0.0024 5.3 0.0024 2.2 0.0025 7.0 0.0025 9.5
0.0024Yb 0.0215 3.5 0.0213 6.7 0.0227 4.6 0.0200 3.5 0.0235 3.0
0.0205 2.9 0.0230 4.0 0.0217 5.1 0.0255 0.8 0.0233 10.0 0.0223 7.3
0.0224Lu 0.0049 3.1 0.0046 13.3 0.0047 4.9 0.0045 11.1 0.0054 3.7
0.0045 2.9 0.0050 6.2 0.0044 4.3 0.0049 1.2 0.0049 9.0 0.0048 6.2
0.0045[1] Ionov et al. (1992) ICP-QMS, [2] Eggins et al. (1997)
ICP-QMS, [3] Robinson et al. (1999) SC-ICP-MS, [4] Jain et al.
(2000) ICP-QMS. [5] Takazawa et al. (2000) ICP-QMS, [6] Dulski
(2001) ICP-QMS, [7] Olive et al. (2001) ICP-QMS, [8] Qiet al.
(2005) ICP-QMS, [9] Willbold and Jochum (2005) SC-ICP-MS, [10]
Makishima and Nakamura (2006) ICP-QMS.
e. W-2
Element [1] [2] RSD% [3] [4] [5] RSD% [6] [7] Lit. Mean RSD%
This study
Sc 36.00 36.20 2.6 35.00 36.10 36.10 35.88 1.4 32.94Y 23.00
22.80 0.8 21.90 20.10 20.11 20.19 21.35 6.5 19.58La 10.00 10.59 0.9
10.07 10.52 10.23 0.8 10.52 10.54 10.35 2.4 10.31Ce 23.00 23.08 0.7
22.79 23.20 22.80 0.5 23.22 23.26 23.05 0.8 23.23Pr 3.027 3.027 0.7
3.040 3.030 2.919 2.9 3.030 3.032 3.015 1.4 2.981Nd 13.00 12.95 0.5
12.90 12.91 12.71 0.6 12.91 12.94 12.90 0.7 13.07Sm 3.300 3.310 0.5
3.240 3.270 3.234 0.6 3.270 3.274 3.271 0.9 3.240Eu 1.000 1.0930
0.6 1.100 1.094 1.072 0. 60 1.090 1.096 1.078 3.3 1.109Gd 3.690
3.960 0.7 3.730 3.710 3.692 0.8 3.710 3.714 3.744 2.6 3.741Tb
0.6300 0.6220 0.7 0.6320 0.6500 0.6200 0.6160 0.6283 1.9 0.6077
(continued on next page)
45A.Pourm
andet
al./Chem
icalGeology
291(2012)
3854
-
Table 5 (continued)
e. W-2
Element [1] [2] RSD% [3] [4] [5] RSD% [6] [7] Lit. Mean RSD%
This study
Dy 3.600 3.790 1.1 3.830 3.810 3.825 0.7 3.810 3.822 3.784 2.2
3.887Ho 0.7600 0.7980 0.9 0.8000 0.8030 0.8000 0.8050 0.7943 2.1
0.7825Er 2.500 2.260 0.8 2.170 2.220 2.248 0.5 2.220 2.231 2.264
4.8 2.239Tm 0.3800 0.3270 0.3300 0.3280 0.3413 7.6 0.3170Yb 2.050
2.030 1.4 1.980 2.060 2.020 0.6 2.060 2.058 2.037 1.5 2.041Lu
0.3300 0.2990 1.4 0.3000 0.3010 0.2937 0.5 0.3000 0.3020 0.3037 3.9
0.2951[1] Govindaraju (1994) compilation, [2] Eggins et al. (1997)
ICP-QMS, [3] Kelley et al. (2003) ICP-QMS, [4] Li et al. (2003)
ICP-QMS, [5] Kent et al. (2004) MC-ICP-MS, [6] Bolhar et al. (2005)
ICP-QMS, [7] Jupiter (2008) SC-ICP-MS.
f.G-2
Element [1] RSD% [2] [3] RSD% [4] RSD% [5] [6] RSD% [7] RSD% [8]
[9] [10] Lit. Mean RSD% G-2 A G-2 B G-2 C This study(n=3)
RSD%
Sc 3.900 8.7 3.590 3.640 3.710 4.5 3.448 3.571 3.499 3.506 1.8Y
9.400 3.5 10.00 9.940 3.0 9.400 1.9 10.30 10.00 4.0 9.240 0.8 9.760
9.280 9.980 9.730 3.8 9.268 9.466 9.045 9.260 2.3La 87.00 2.8 89.20
84.36 11.0 88.90 1.0 88.60 89.60 3.5 88.00 89.60 79.70 85.80 87.08
3.6 88.28 88.99 89.31 88.86 0.6Ce 158.0 1.6 160.0 154.0 11.0 164.2
0.8 159.0 164.0 3.8 177.0 161.0 151.0 151.0 159.92 4.8 164.907
166.26 166.41 165.9 0.5Pr 16.30 3.0 17.00 17.35 8.0 16.81 0.7 16.50
16.70 3.8 17.00 2.3 16.90 15.90 17.70 16.82 3.0 16.60 16.73 16.69
16.67 0.4Nd 51.50 2.2 54.70 52.90 7.0 53.24 0.8 54.50 54.80 4.0
54.00 1.7 58.50 49.80 53.70 53.76 4.2 54.04 54.48 54.11 54.21 0.4Sm
6.800 2.1 7.330 7.010 7.0 7.170 1.5 7.260 7.480 4.0 7.130 0.8 7.250
6.950 7.360 7.174 2.9 7.206 7.283 7.096 7.195 1.3Eu 1.410 5.0 1.440
1.400 6.0 1.330 1.8 1.500 1.480 3.8 1.340 2.2 1.300 1.380 1.520
1.410 5.3 1.353 1.366 1.341 1.353 0.9Gd 4.100 4.0 4.490 4.390 6.0
4.030 1.7 5.300 4.080 4.2 4.080 1.1 4.460 4.370 5.360 4.466 10.9
4.514 4.598 4.456 4.523 1.6Tb 0.4900 8.6 0.4800 0.4500 6.0 0.4800
2.4 0.4900 0.4900 4.2 0.5120 0.4 0.520 0.4700 0.5300 0.4912 4.9
0.4754 0.4838 0.4622 0.4738 2.3Dy 2.080 3.3 2.300 2.170 5.0 2.210
1.6 2.410 2.320 4.4 2.240 1.7 2.260 2.040 2.340 2.237 5.2 2.207
2.244 2.131 2.194 2.6Ho 0.3500 3.8 0.3800 0.3500 5.0 0.3800 1.8
0.3900 0.3800 4.5 0.3560 0.9 0.3640 0.3500 0.3700 0.3670 4.1 0.3641
0.3707 0.3530 0.3626 2.5Er 0.8800 5.1 0.9600 0.8400 4.0 0.9100 1.8
1.020 0.9700 4.5 0.9040 0.6 1.020 0.8500 0.9700 0.9324 7.0 0.9220
0.9401 0.8978 0.9200 2.3Tm 0.1170 4.4 0.1200 0.1200 7.0 0.1200 2.4
0.1200 0.1250 5.1 0.1250 2.2 0.1220 0.1200 0.1200 0.1209 2.0 0.1197
0.1253 0.1179 0.1210 3.2Yb 0.7200 5.8 0.7300 0.7200 5.0 0.7400 1.8
0.7800 0.7400 4.4 0.7120 0.3 0.7140 0.6800 0.7300 0.7266 3.5 0.7467
0.7673 0.7335 0.7492 2.3Lu 0.1200 13.0 0.1000 0.1100 6.0 0.1000 1.9
0.1100 0.1070 5.0 0.0990 3.1 0.1100 0.1000 0.1000 0.1056 6.6 0.1021
0.1066 0.1015 0.1034 2.7[1] Dulski (2001) SC-ICP-MS, [2] Jiang et
al. (2007) ICP-MS, [3] Madinabeitia et al. (2008) ICP-QMS, [4]
Bolhar et al. (2005) ICP-QMS, [5] Zhu et al. (2009) ICP-QMS. [6]
Meisel et al. (2002) ICP-QMS, [7] Willbold and Jochum (2005)
SC-ICP-MS, [8] Wang et al. (2007) ICP-QMS, [9] Yu et al. (2001)
SC-ICP-MS, [10] Huang et al. (2007) ICP-QM.
g. G-3
Element [1] RSD% A B C D E This study (n=5) RSD%
Sc 3.253 3.127 3.262 3.178 3.023 3.169 3.1Y 10.30 3.4 8.899
8.719 8.988 8.802 8.562 8.794 1.9La 92.40 1.5 88.74 86.10 89.72
89.36 85.87 87.96 2.1Ce 171.0 2.2 165.2 160.4 167.1 166.0 160.2
163.8 2.0Pr 17.40 2.5 16.55 16.10 16.72 16.70 16.09 16.43 1.9Nd
56.80 2.3 53.65 52.33 54.25 54.05 52.21 53.30 1.8Sm 7.690 2.4 7.049
6.912 7.161 7.029 6.868 7.004 1.7Eu 1.500 3.1 1.316 1.284 1.337
1.311 1.275 1.305 1.9Gd 4.070 3.1 4.403 4.336 4.503 4.403 4.299
4.389 1.8Tb 0.4700 3.1 0.4538 0.447 0.4631 0.4502 0.4393 0.4506
2.0Dy 2.230 2.9 2.072 2.044 2.107 2.052 1.999 2.055 1.9Ho 0.3700
2.8 0.342 0.3354 0.3457 0.3379 0.3286 0.3379 1.9Er 0.9400 3.0 0.871
0.8523 0.8753 0.8607 0.8346 0.8588 1.9Tm 0.1230 3.3 0.116 0.1137
0.1154 0.1137 0.1100 0.1137 2.0Yb 0.7400 3.5 0.729 0.7102 0.7274
0.7206 0.6947 0.7164 2.0Lu 0.1100 3.7 0.103 0.1002 0.1013 0.1013
0.0975 0.1006 1.9
[1] Meisel et al. (2002) ICP-QMS.
Table 5 (continued)
46A.Pourm
andet
al./Chem
icalGeology
291(2012)
3854
-
Mea
n R
EE
co
nce
ntr
atio
ns,
US
GS
ref
eren
ce m
ater
ials
n
orm
aliz
e to
RE
E C
on
cen
trat
ion
s, t
his
stu
dy
0.9
1.0
1.1
PCC-1
0.9
1.0
1.1
1.2BHVO-1
0.8
0.9
1.0
1.1
BCR-2
BIR-1
G-2
La
W-2
0.81.2
0.81.2
LuYbTmErHoDyTbGdEuSmNdPrCeLa LuYbTmErHoDyTbGdEuSmNdPrCe
Fig. 3. Abundances of REE in six reference materials from
literature compilations are indistinguishable from those measured
in this study within analytical uncertainties, and attestto the
accuracy of the measurements. The error bars represent 1 SD on the
mean of literature values.
47A. Pourmand et al. / Chemical Geology 291 (2012) 3854
from literature measurements (Jarosewich et al., 1987;
Shinotsukaand Ebihara, 1997) demonstrates very similar
fractionation patterns,with relative enrichment in LREE and
depletion in HREE. These pat-terns along with a negative anomaly
for Eu and a prominent positiveanomaly for Tm distinguish the
Allende carbonaceous chondrite fromother meteorites. It must be
noted that while the REE pattern inAllende A and B (Table 7)
analyzed in this study are closely replicated,Eu depletion in the
former is not as pronounced. Given that the USNM
Table 6The composition of Post-Archean Australian Shales (PAAS)
based on nine samples. Uncertaisamples, indices reflective of
secular composition changes, such as Eu/Eu* (Eu enrichmentratios,
remain relatively constant for all PAAS samples.
PAAS
Element AO-6 AO-7 AO-9 AO-10 AO-12
Sc 13.55 14.97 14.88 17.15 20.18Y 21.80 23.64 22.01 25.79
27.80La 37.87 40.30 38.65 45.69 44.59Ce 75.23 78.27 74.56 87.48
82.50Pr 8.60 8.96 8.66 10.14 10.02Nd 31.46 32.29 31.50 37.01
37.01Sm 5.728 5.878 5.759 6.935 6.149Eu 0.990 1.031 1.002 1.264
1.115Gd 4.911 5.127 4.944 6.102 5.388Tb 0.720 0.761 0.720 0.861
0.857Dy 4.273 4.557 4.283 5.027 5.291Ho 0.839 0.910 0.844 0.982
1.064Er 2.446 2.687 2.477 2.865 3.121Tm 0.360 0.400 0.365 0.421
0.456Yb 2.419 2.703 2.459 2.819 3.064Lu 0.354 0.399 0.360 0.412
0.451Eu/Eu* 0.56 0.57 0.57 0.59 0.58LREE/ HREE 9.73 9.44 9.67 9.61
9.16La/Sc 2.79 2.69 2.60 2.66 2.21
powder is homogenized from a large piece of Allende, the origin
of thisdiscrepancy remains unknown to us.
Fractionation patterns similar to those presented in Fig. 5
havealso been reported in group II-type CaAl-rich inclusions in
carbona-ceous chondrites (Tanaka and Masuda, 1973; Martin and
Mason,1974; Grossman, 1980; Mason and Taylor, 1982; MacPherson et
al.,1988). Partial removal of refractory condensates and incomplete
con-densation of most volatile elements have been proposed as
possible
nty on the ratios is 2SD. While absolute concentrations are
variable between individualfactor = EuN=
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiSmNGdNp
, N: normalized to CI-chondrites), LREE/ HREE and La/Sc
SC-7 SC-8 PL-1 PW-5 Mean (n=9)
14.18 15.31 14.21 18.61 15.8926.70 27.45 31.61 39.01 27.3143.08
44.52 42.77 63.53 44.5687.61 91.38 90.40 126.8 88.259.90 10.29
10.02 14.72 10.1536.75 38.28 36.82 54.78 37.326.835 7.141 7.240
10.29 6.8841.211 1.192 1.210 1.917 1.2155.983 6.218 6.588 9.127
6.0430.867 0.895 1.013 1.328 0.89145.135 5.299 6.243 7.816
5.3251.009 1.047 1.264 1.515 1.0532.911 3.044 3.814 4.314
3.0750.421 0.442 0.578 0.617 0.45102.774 2.926 3.916 4.030
3.0120.400 0.425 0.569 0.578 0.43860.57 0.54 0.53 0.60 0.570.029.45
9.44 7.81 9.21 9.281.173.04 2.91 3.01 3.41 2.810.67
-
10
100P
AA
S R
EE
co
nce
ntr
atio
ns,
th
is s
tud
y n
orm
aliz
ed t
om
ean
of
CI-
cho
nsd
rite
s, t
his
stu
dy
10
100
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
AO-6 AO-7 AO-10 AO-9 AO-12 SC-7 SC-8 PL-1 PW-5
PA
AS
RE
E c
on
cen
trat
ion
s, N
ance
an
d T
aylo
r (1
976)
no
rmal
ized
to
mea
n o
f C
I-ch
on
sdri
tes,
th
is s
tud
y
? ?
a
b
Fig. 4. Rare earth element patterns in Post-Archean Australian
Shales from this study (a) are compared with results from Nance and
Taylor (1976) (b). All values are normalized tothe mean of
CI-chondrites from this study (Table 9). Normalized REE patterns
from our study are smoother, particularly for HREE. Thulium and Yb,
which are not reported in Nanceand Taylor (1976), are also
presented. Although a small Tm depletion may be present, it cannot
be resolved given the analytical uncertainty for the mean of
CI-chondrites.
Table 7Concentrations of REE, Sc and Y (g g1) in Tagish Lake
(CI2-ung), Allende (CV3), Ivuna (CI1) and Alais (CI1). Allende A
and B are replicate measurements of Allende bulk powderfrom USNM.
Ivuna A, B and C are separate analyses on different chips from this
meteorite.
Element Tagish Lake(C2. ung)
Allende(CV3) A
Allende(CV3) B
Ivuna(CI1) A
Ivuna(CI1) B
Ivuna(CI1) C
Alais(CI1)
Sc 8.39 10.94 11.72 5.56 5.93 5.96 5.93Y 1.768 2.742 2.664 1.434
1.492 1.470 1.560La 0.3182 0.5628 0.5325 0.2720 0.2710 0.2623
0.2670Ce 0.7952 1.407 1.341 0.6998 0.6894 0.6724 0.6858Pr 0.1209
0.2184 0.20603 0.1054 0.1043 0.1017 0.1035Nd 0.6172 1.121 1.052
0.5379 0.5256 0.5164 0.5299Sm 0.1989 0.3562 0.33491 0.1681 0.1682
0.1653 0.1702Eu 0.0781 0.1232 0.11333 0.0638 0.0645 0.0634 0.0656Gd
0.2618 0.4433 0.45021 0.2094 0.2257 0.2225 0.2318Tb 0.0481 0.0801
0.08123 0.0383 0.0407 0.0396 0.0414Dy 0.3300 0.5389 0.54977 0.2480
0.2753 0.2730 0.2858Ho 0.0699 0.1078 0.10975 0.0504 0.0594 0.0581
0.0612Er 0.2113 0.3135 0.31955 0.1453 0.1787 0.1762 0.1847Tm 0.0333
0.0557 0.05457 0.0224 0.0287 0.0272 0.0288Yb 0.2206 0.3258 0.33086
0.1455 0.1833 0.1800 0.1882Lu 0.0319 0.0450 0.04698 0.0205 0.0300
0.0263 0.0292
48 A. Pourmand et al. / Chemical Geology 291 (2012) 3854
-
Table 8Concentrations of REE, Sc and Y (g g1) in seven
replicates of Orgueil (AG) from this study are compared with 6
literature values measured by thermal ionization mass spectrom-etry
(TIMS) and ICP-MS techniques. Orgueil F and G are enriched in LREE
and were not included in calculations of the mean of Orgueil and
CI-chondrites (see Table 9). RSD% =Relative standard deviation
(100SD/average). Isotope dilution analyses of REE from Nakamura
(1974) do not include mono-isotopic elements (Pr, Tb, Ho and
Tm).
Element [1] [2] [3] [4] [5] [6] Lit.Mean
RSD% A B C D E F G This studyMean (n=5)
RSD%
Sc 6.450 5.420 5.200 5.690 11.7 5.852 5.883 5.057 5.097 5.575
6.135 6.516 5.493 7.3Y 1.530 1.780 1.490 1.430 1.558 9.9 1.425
1.344 1.395 1.403 1.413 1.503 1.489 1.396 2.2La 0.2610 0.2440
0.2360 0.2360 0.2450 0.2370 0.2432 4.0 0.2417 0.2454 0.2521 0.2506
0.2454 0.3401 0.3170 0.2471 1.7Ce 0.6680 0.6210 0.6190 0.6230
0.6180 0.6020 0.6252 3.6 0.6240 0.6242 0.6472 0.6371 0.6306 0.8328
0.7797 0.6326 1.5Pr 0.0900 0.0958 0.0980 0.0940 0.0945 3.6 0.0947
0.0957 0.0980 0.0962 0.0964 0.1186 0.1176 0.0962 1.3Nd 0.4900
0.4610 0.4630 0.4490 0.4900 0.4670 0.4700 3.5 0.4864 0.4712 0.4966
0.4843 0.4909 0.5732 0.6004 0.4859 1.9Sm 0.1590 0.1480 0.1440
0.1510 0.1550 0.1510 0.1513 3.5 0.1570 0.1513 0.1589 0.1541 0.1579
0.1726 0.1847 0.1558 2.0Eu 0.0592 0.0581 0.0547 0.0554 0.0590
0.0608 0.0579 4.1 0.0601 0.0590 0.0610 0.0591 0.0613 0.0619 0.0685
0.0601 1.7Gd 0.2100 0.1980 0.1990 0.2010 0.2040 0.2060 0.2030 2.2
0.2062 0.2036 0.2142 0.2078 0.2138 0.2222 0.2359 0.2091 2.2Tb
0.0353 0.0390 0.0395 0.0376 0.0379 5.0 0.0382 0.0375 0.0382 0.0373
0.0387 0.0400 0.0413 0.0380 1.5Dy 0.2590 0.2440 0.2460 0.2100
0.2640 0.2580 0.2468 8.0 0.2636 0.2492 0.2608 0.2545 0.2628 0.2746
0.2782 0.2582 2.4Ho 0.0552 0.0530 0.0574 0.0577 0.0558 3.9 0.0565
0.0543 0.0554 0.0547 0.0568 0.0586 0.0586 0.0555 2.0Er 0.1710
0.1600 0.1620 0.1540 0.1550 0.1600 0.1603 3.8 0.1705 0.1611 0.1662
0.1660 0.1709 0.1787 0.1747 0.1669 2.4Tm 0.0220 0.0270 0.0268
0.0256 0.0254 9.1 0.0263 0.0254 0.0257 0.0258 0.0269 0.0271 0.0261
0.0260 2.2Yb 0.1770 0.1600 0.1660 0.1580 0.1790 0.1710 0.1685 5.2
0.1712 0.1601 0.1681 0.1719 0.1737 0.1838 0.1762 0.1690 3.2Lu
0.0262 0.0244 0.0245 0.0280 0.0262 0.0254 0.0258 5.2 0.0253 0.0241
0.0247 0.0257 0.0260 0.0276 0.0258 0.0252 3.0
[1] Nakamura (1974) TIMS, [2] Nakamura (1974) TIMS, [3]
Shinotsuka and Ebihara (1997) ICP-MS, [4] Friedrich et al. (2002)
ICP-MS, [5] Makishima and Nakamura (2006) ICP-QMS,[6] Makishima and
Nakamura (2006) ICP-QMS.
49A. Pourmand et al. / Chemical Geology 291 (2012) 3854
mechanisms that may be responsible for the REE
fractionationpatterns observed in these objects (Wnke et al., 1974;
Boynton,1975; Davis and Grossman, 1979; Palme and Boynton, 1993;
Palme,2000).
3.3.2. Tagish Lake (C2-ung)The average REE concentrations in
Tagish Lake are about 24% higher
than the mean of CI-chondrites (Tables 7 and 9). Nevertheless,
the REEpattern in this meteorite, shown in Fig. 6, is relatively
flat. Tagish Lakeis a unique meteorite with trace-element patterns
that differ from CMand CI chondrites (Brown et al., 2000; Friedrich
et al., 2002; Mittlefehldt,2002). Rare earth elements in the sample
from this study show slightlyhigher enrichment for all REE but less
deviation from the mean com-pared with previous bulk measurements
of this meteorite (Fig. 6). Al-though the sample quantity analyzed
in this study was relatively small(15 mg), it came from 34 g of
homogenized powder (Jadhav, 2009),which may explain the smoother
REE pattern compared with previousmeasurements of this
meteorite.
1.65
1.70
1.75
1.80
1.85
1.90
1.95
2.00
2.05
2.10
2.15
2.20
2.25
2.30
2.35
Allende A (this study)Shinotsuka and Ebihara
Alle
nd
e R
EE
co
nce
ntr
atio
ns
no
rmal
ized
to
mea
n o
f C
I-C
ho
nd
rite
s, t
his
stu
dy
La Ce Pr Nd Sm Eu
Fig. 5. Rare earth element patterns for two separate analyses of
bulk Allende are compared wto the mean of CI-chondrites (Table 9).
The REE pattern in Allende samples is similar to grouHREE, a
negative anomaly in Eu and positive anomaly in Tm.
3.3.3. Orgueil, Alais and Ivuna (CI1)There have been more
high-precision analyses of REE by ICP-MS and
thermal ionization mass spectrometry in bulk aliquots of Orgueil
thanother primitive meteorites. This is mostly due to sample
availability ofCI chondrites; the recovered mass of Orgueil was 14
kg compared withAlais (6 kg), Ivuna (705 g), Revelstoke (1 g) and
Tonk (7.7 g). As a result,meteoritic estimates of REE, Sc and Y
abundances in the solar system arelargely based on the analyses of
Orgueil (Anders and Grevesse, 1989;Palme and Jones, 2003; Lodders
et al., 2009).
The results for REE, Sc and Y concentrations in seven samples
ofOrgueil homogenized from different chips of MNHN 219, 234 andthe
University of Chicago collection C3_1146, are presented inTable 8.
While five Orgueil measurements (AE) are indistinguishablefrom
literature compilations within analytical uncertainties,
twomeasurements on separate chips of Orgueil (F and G) show clear
en-richment in LREE and Eu relative to the mean of CI-chondrites,
asshown in Fig. 7a. A comparison with CI-normalized REE patternwith
literature compilations (Fig. 7b) shows less scatter in REE
Allende B (this study)(1997) Jarosewich et al. (1987)
Gd Tb Dy Ho Er Tm Yb Lu
ith measurements of this meteorite from previous studies.
Abundances are normalizedp-II CaAl inclusions in carbonaceous
chondrites, with enrichment in LREE, depletion in
-
Table 9Recommended mean of CI-chondrite concentrations (g g1)
for REE, Sc and Y based on MC-ICP-MS measurements of 8
CI-chondrites (five Orgueil (AE), two Ivuna (B and C) andone Alais)
are compared with most commonly cited literature compilations. The
anomalous values for Ivuna A and Orgueil F and G are not included.
The abundances for Alais andIvuna samples are adjusted to the mean
of Orgueil. RSD% = Relative standard deviation (100SD/average)
based on Orgueil AE measurements.
Element [1] [2] [3] [4] Estimatedaccuracy%
[5] Estimatedaccuracy%
This study(n=8)
RSD%
Sc 5.800 5.820 5.920 5.900 3 5.900 5 5.493 5.6Y 1.440 1.560
1.570 1.560 3 1.530 10 1.395 1.9La 0.2340 0.2347 0.2370 0.2450 5
0.2420 5 0.2469 1.5Ce 0.6160 0.6032 0.6130 0.6380 5 0.6220 5 0.6321
1.3Pr 0.0929 0.0891 0.0928 0.0964 10 0.0946 7 0.0959 1.2Nd 0.4570
0.4524 0.4570 0.4740 5 0.4710 5 0.4854 1.5Sm 0.1490 0.1471 0.1480
0.1540 5 0.1520 5 0.1556 1.6Eu 0.0560 0.0560 0.0563 0.0580 5 0.0578
5 0.0599 1.4Gd 0.1970 0.1966 0.1990 0.2040 5 0.2050 5 0.2093 1.8Tb
0.0355 0.0363 0.0361 0.0375 10 0.0384 7 0.0378 1.3Dy 0.2450 0.2427
0.2460 0.2540 5 0.2550 5 0.2577 2.0Ho 0.0547 0.0556 0.0546 0.0567
10 0.0572 7 0.0554 1.6Er 0.1600 0.1589 0.1600 0.1660 5 0.1630 5
0.1667 1.9Tm 0.0247 0.0242 0.0247 0.0256 10 0.0261 7 0.0261 1.8Yb
0.1590 0.1625 0.1610 0.1650 5 0.1690 5 0.1694 2.5Lu 0.0245 0.0243
0.0246 0.0254 10 0.0253 5 0.0256 4.4
[1] Wasson and Kallemeyn (1988), [2] Anders and Grevesse (1989),
[3] McDonough and Sun (1995), [4] Palme and Jones (2003), [5]
Lodders et al. (2009).
50 A. Pourmand et al. / Chemical Geology 291 (2012) 3854
patterns in the data from our study. The uncertainties (RSD%) on
REE,Y and Sc concentration based on five replicate analyses of
Orgueil AEare also generally smaller than those calculated based on
the mean of6 literature measurements.
With the exception of an apparent small enrichment in Lu, the
REEpattern of Alais is quite similar to Orgueil with an overall
enrichmentof about 7% relative to the mean of CI-chondrites (Fig.
7a).
The REE abundances in three chips of Ivuna (A, B and C) are
quite var-iable. As shown in Fig. 8, two of the analyzed chips have
almost flat REEpatterns (B and C) while Ivuna A is characterized by
a significant deple-tion in HREE. Interestingly, the REE abundances
for Ivuna from Lodders(2003) form an entirely different pattern,
with overall depletion inREE, negative anomalies for Tb and Tm and
a positive anomaly for Ho.The deviations from a flat pattern
observed in some Ivuna specimenscould result from REE
redistribution by aqueous fluids on the parent-body of
CI-chondrites. The petrographic texture of CI-chondrites showsgreat
abundance of microbreccias with a wide range of
mineralogicalcompositions. It is indeed well documented that REE
and actinides are
0.95
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
This study Brown
Tag
ish
Lak
e R
EE
Co
nce
ntr
atio
ns
No
rmal
ized
to M
ean
of
CI-
Ch
on
dri
tes,
Th
is S
tud
y
La Ce Pr Nd Sm Eu
Fig. 6. The abundances of REE in Tagish Lake from this study,
normalized to mean of CI-chondLake is geochemically different from
CI-chondrites, it shows a similarly flat pattern. The sawhich may
explain the relatively smooth pattern compared with others.
not distributed homogenously and are concentrated in trace
phasessuch as apatite (phosphates) and merrillite in CI and other
groups ofchondrites (Rocholl and Jochum, 1993; Goreva andBurnett,
2001;Morloket al., 2006; Bouvier et al., 2008).
3.4. Revised REE, Sc and Y in CI-chondrites
AlthoughAlais and two Ivuna chips (B andC) are generally
enriched inall REE relative to the mean of CI chondrites, REE
ratios in these samplesdo not show fractionations that
significantly deviate from the flat patternseen in Orgueil AE and
themean of CI-chondrites from literature compi-lations (Tables 8
and9). The patterns for these CI-chondrites are very sim-ilar to
Tagish Lake (Fig. 6),which came from34 gof homogenizedpowderand is
not affected by inhomogeneity inherent to small sample sizes.
Afteradjusting the abundances of Alais and Ivuna B and C to the
mean ofOrgueil AE, we propose new, fiducial values for the
concentrations ofREE, Sc and Y in CI-chondrites based on the
average of MC-ICP-MS mea-surements in Orgueil (5), Alais (1), and
Ivuna (2), while excluding the
et al. (2000) Friedrich et al. (2002)
Gd Tb Dy Ho Er Tm Yb Lu
rites (Table 9), are compared with previous analyses of this
meteorite. Although Tagishmple analyzed in this study came from
homogenizing 34 g of Tagish Lake meteorite,
-
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
Orgueil A (MNHN 219) Orgueil B (MNHN 219) Orgueil C (MNHN
219)Orgueil D (C3_1146) Orgueil E (C3_1146) Alais FM C3_0067Orgueil
F (MNHN 219,) different chip Orgueil G (MNHN 234)
Org
uei
l RE
E c
on
cen
trat
ion
s n
orm
aliz
ed t
om
ean
of
CI-
Ch
on
dri
tes,
th
is s
tud
y
a
0.80
0.85
0.90
0.95
1.00
1.05
1.10
Nakamura (1974) Shinotsuka et al. (1997) Friedrich et al.
(2002)
Makishima and Nakamura (2006) Makishima and Nakamura (2006)
Nakamura (1974)
Org
uei
l RE
E c
on
cen
trat
ion
s n
orm
aliz
ed t
om
ean
of
CI-
Ch
on
dri
tes,
th
is s
tud
y
b
Fig. 7. a) The REE patterns in Alais and seven samples of
Orgueil are compared. Abundances are normalized to the mean of
CI-chondrites (Table 9). While Orgueil AE show asmooth pattern and
their abundances are closely replicated, Orgueil F and G show
varying degrees of enrichment, particularly in LREE. These samples
were not included in calcu-lation of the mean of CI-chondrites in
this study. Rare earth element abundances in Alais are enriched by
an average of 7% compared with the mean of Orgueil AE with
slightlyhigher enrichment of HREE compared with LREE. b) Six
literature measurements by ICP-MS and TIMS. Isotope dilution
results from Nakamura (1974) do not include mono-isotopicelements
Pr, Tb, Ho and Tm. Literature data for Orgueil shows more scatter
than those measured in this study.
51A. Pourmand et al. / Chemical Geology 291 (2012) 3854
outliers Ivuna A, Orgueil F and G. The values listed in Table 9,
define thecosmic abundances of these elements and are recommended
as referenceconcentrations to normalize REE, Sc and Y abundance
patterns of terres-trial and extraterrestrialmaterials. Our
recommendedREE concentrationsand literature compilations for
CI-chondrites that are commonly used forREE normalization are in
good agreement within estimated accuraciesreported by Palme and
Jones (2003) and Lodders et al. (2009), with im-proved
uncertainties from our study (Table 9). The average ratio
(byweight) for the twin elements Y and Ho for CI-chondrites derived
hereis 25.20.2 (2 of the mean), which is slightly lower but
comparableto previous estimates of Y/Ho for carbonaceous chondrites
(e.g., 25.940.08; Pack et al., 2007).
4. Summary
1. A simple purification procedure is developed to reduce REE,
Sc and Yblank levels in commercial LiBO2 flux by up to 3 orders
ofmagnitude.Homogenized aliquots of terrestrial reference
materials, PAAS and
meteorite samples were digested using low-blank LiBO2 fusion
andREE, Sc and Y were separated from the matrix with a single
TODGAextraction chromatography step. A newmethod was also
developedto take advantage of 9 Faraday collectors, source lens
voltage adjust-ments and zoom optics on the Neptune MC-ICP-MS for
measure-ment of REE, Sc and Y in a single sample solution,
whileminimizing the effect of fluctuations in the Ar plasma and
inlet sys-tems. An Apex-Q+Spiro TMD desolvation system interfaced
withthe Neptune rendered the corrections for polyatomic
interferenceson heavy rare earth elements negligible. The entire
procedure fromsample digestion to data collection can be completed
in a day.
2. The accuracy of the proposed procedurewas tested by comparing
REE,Sc and Y concentrations in USGS reference materials BHVO-1,
BCR-2,BIR-1, PCC-1, W-2 and G-2 with ICP-MS and TIMS data from the
liter-ature. Procedural blank corrections were negligible for
BHVO-1, BCR-2, BIR-1, W-2, G-2 and G-3. Higher blank corrections
for PCC-1 weredue to exceptionally low REE concentrations in this
peridotite refer-ence material. The elemental concentrations and
CI-normalized REE
-
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
Ivuna (CI1) A Ivuna (CI1) B Ivuna (CI1) C Ivuna (CI1) Lodders
(2003)
Org
uei
l RE
E c
on
cen
trat
ion
s n
orm
aliz
ed t
om
ean
of
CI-
Ch
on
dri
tes,
th
is s
tud
y
Ivuna B
Ivuna A
Ivuna C
Ivuna Lodders (2003)
Fig. 8. The CI-chondrite normalized patterns (Table 9) of REE in
three chips of Ivuna are compared. With the exception of an
apparent positive anomaly for Lu in Ivuna B, the REEpatterns in
Ivuna B and C are relatively flat. In contrast, Ivuna A shows
distinct fractionations in HREE. The normalized REE pattern in a
sample of Ivuna from Lodders (2003) showsan entirely different
fractionation pattern, with negative anomalies for Tb and Tm and a
positive Ho anomaly. The observed deviations in some of Ivuna
specimens may be due toheterogeneous distribution of REE in
different mineralogical phases (such as phosphates) in some
CI-chondrites.
52 A. Pourmand et al. / Chemical Geology 291 (2012) 3854
patterns in reference materials from our study were
indistinguishablefrom literature values within uncertainties and
attest to the accuracyof our analytical technique. The REE
concentrations of PAAS were an-alyzed, yielding REE patterns that
are smoother and more completethan previous studies, which did not
report measured values for Tmand Yb. We recommend that these new
values be used to definePAAS for normalization purposes.
3. Rare-earth element concentrations were determined in
primitivechondrites Orgueil (CI1), Alais (CI1) and Ivuna (CI1), as
well asAllende (CV3) powder and Tagish Lake (C2-ung). The
fractionationpattern of REE in Allende was similar to group II-type
CAI, with en-richment in LREE, depletion in HREE, negative Eu and
positive Tmanomalies. Out of the 11 CI-chondrite measurements, 8
show verysimilar REE patterns, while one chip of Ivuna and two
chips ofOrgueil demonstrate significant shifts in LREE and HREE.
Theseanomalous patterns may be the result of REE fractionation
duringparent-body alteration and redistribution of REE in trace
carrierphases such as apatite and merrilite. Excluding these
outliers, werecommend revised values for REE, Sc and Y
concentration of CI-chondrites and refine the solar abundances of
these elements.
Acknowledgements
The authorswould like to thank Junjun Zhang for assistancewith
pu-rification of commercial LiBO2 flux. This work was supported by
a Pack-ard fellowship and NASA through grants NNG06GG75G
andNNX09AG59G to N.D. Discussions with Jean-Alix Barrat, who
sharedwith us a preprint of his work on REE in CI-chondrites, and
AndreasPack regarding Y/Ho ratios, were greatly appreciated. We are
gratefulto Roberta Rudnick and Frdric Moynier for making powder
aliquotsof PAAS and Tagish Lake available for this study,
respectively. Additionalmeteorite samples were generously provided
by the FieldMuseum,Mu-seum National d'Histoire Naturelle and the
Smithsonian Museum.
References
Anders, E., Grevesse, N., 1989. Abundances of the elements:
meteoritic and solar. Geo-chimica et Cosmochimica Acta 53 (1),
197214.
Baker, J., Waight, T., Ulfbeck, D., 2002. Rapid and highly
reproducible analysis of rareearth elements by multiple collector
inductively coupled plasma mass spectrome-try. Geochimica et
Cosmochimica Acta 66 (20), 36353646.
Balaram, V., 1996. Recent trends in the instrumental analysis of
rare earth elements ingeological and industrial materials. TRAC
Trends in Analytical Chemistry 15 (9),475486.
Barrat, J.A., Blichert-Toft, J., Gillet, P., Keller, F., 2000.
The differentiation of eucrites: therole of in situ
crystallization. Meteoritics & Planetary Science 35 (5),
10871100.
Barrat, J.A., Yamaguchi, A., Greenwood, R.C., Bohn, M., Cotten,
J., Benoit, M., Franchi, I.A.,2007. The Stannern trend eucrites:
contamination of main group eucritic magmasby crustal partial
melts. Geochimica et Cosmochimica Acta 71 (16), 41084124.
Bayon, G., Barrat, J.A., Etoubleau, J., Benoit, M., Bollinger,
C., Rvillon, S., 2009. Determi-nation of rare-earth elements, Sc,
Y, Zr, Ba, Hf and Th in geological samples by ICP-MS after Tm
addition and alkaline fusion. Geostandards and Geoanalytical
Re-search 33 (1), 5162.
Bendel, V., Patzer, A., Pack, A., Hezel, D.C., Mnker, C., 2011.
Rare earth elements in bulkchondrites and chondrite components.
42nd Lunar and Planetary Science Confer-ence, 1608, p. 1711.
Bizzarro,M., Baker, J.A., Ulfbeck, D., 2003. A newdigestion and
chemical separation techniquefor rapid and highly reproducible
determination of Lu/Hf andHf isotope ratios in geolog-ical
materials by MC-ICP-MS. Geostandard Newsletter 27 (2), 133145.
Bolhar, R., Kamber, B.S., Moorbath, S., Whitehouse, M.J.,
Collerson, K.D., 2005. Chemical char-acterization of earth'smost
ancient clasticmetasediments from the IsuaGreenstone Belt,southern
West Greenland. Geochimica et Cosmochimica Acta 69 (6),
15551573.
Bouvier, A., Vervoort, J.D., Patchett, P.J., 2008. The Lu-Hf and
Sm-Nd isotopic composi-tion of CHUR: Constraints from
unequilibrated chondrites and implications forthe bulk composition
of terrestrial planets. Earth and Planetary Science Letters273
(12), 4857.
Boynton, W.V., 1975. Fractionation in the solar nebula:
condensation of yttrium andthe rare earth elements. Geochimica et
Cosmochimica Acta 39 (5), 569584.
Brown, P.G., Hildebrand, A.R., Zolensky, M.E., Grady, M.,
Clayton, R.N., Mayeda, T.K.,Tagliaferri, E., Spalding, R., MacRae,
N.D., Hoffman, E.L., Mittlefehldt, D.W., Wacker,J.F., Bird, J.A.,
Campbell, M.D., Carpenter, R., Gingerich, H., Glatiotis, M.,
Greiner, E.,Mazur, M.J., McCausland, P.J.A., Plotkin, H., Mazur,
T.R., 2000. The fall, recovery,orbit, and composition of the Tagish
Lake meteorite: a new type of carbonaceouschondrite. Science 290
(5490), 320325.
Connelly, J.N., 2006. Improved dissolution and chemical
separation methods for Lu-Hfgarnet chronometry. Geochemistry,
Geophysics, Geosystems 7. doi:10.1029/2005GC001082.
Connelly, J.N., Ulfbeck, D.G., Thrane, K., Bizzarro, M., Housh,
T., 2006. A method for pu-rifying Lu and Hf for analyses by
MC-ICP-MS using TODGA resin. Chemical Geology233 (12), 126136.
Coogan, L.A., Thompson, G.M., MacLeod, C.J., Dick, H.J.B.,
Edwards, S.J., Hosford Scheirer,A., Barry, T.L., 2004. A combined
basalt and peridotite perspective on 14 millionyears of melt
generation at the Atlantis Bank segment of the Southwest
IndianRidge: evidence for temporal changes in mantle dynamics?
Chemical Geology207 (12), 1330.
Dauphas, N., Pourmand, A., 2011. Hf-W-Th evidence for rapid
growth of Mars and itsstatus as a planetary embryo. Nature 473
(7348), 489492.
Davis, A.M., Grossman, L., 1979. Condensation and fractionation
of rare earths in thesolar nebula. Geochimica et Cosmochimica Acta
43 (10), 16111632.
Debaille, V., Blichert-Toft, J., Agranier, A., Doucelance, R.,
Schiano, P., Albarede, F., 2006.Geochemical component relationships
in MORB from the Mid-Atlantic Ridge, 2235N. Earth and Planetary
Science Letters 241 (34), 844862.
Dia, A., Chauvel, C., Bulourde,M., Grard,M., 2006. Eolian
contribution to soils onMountCam-eroon: isotopic and trace element
records. Chemical Geology 226 (34), 232252.
http://dx.doi.org/10.1029/2005GC001082http://dx.doi.org/10.1029/2005GC001082
-
53A. Pourmand et al. / Chemical Geology 291 (2012) 3854
Dulski, P., 1994. Interferences of oxide, hydroxide and chloride
analyte species in the de-termination of rare earth elements in
geological samples by inductively coupled plas-ma-mass
spectrometry. Fresenius' Journal of Analytical Chemistry 350 (4),
194203.
Dulski, P., 2001. Reference materials for geochemical studies:
new analytical data byICP-MS and critical discussion of reference
values. Geostandards and GeoanalyticalResearch 25 (1), 87125.
Eggins, S.M., Woodhead, J.D., Kinsley, L.P.J., Mortimer, G.E.,
Sylvester, P., McCulloch,M.T., Hergt, J.M., Handler, M.R., 1997. A
simplemethod for the precise determinationof N= 40 trace elements
in geological samples by ICPMS using enriched isotope in-ternal
standardisation. Chemical Geology 134 (4), 311326.
Evans, E.H., Giglio, J.J., 1993. Interferences in inductively
coupled plasma mass-spec-trometry a review. Journal of Analytical
Atomic Spectrometry 8 (1), 118.
Evensen, N.M., Hamilton, P.J., Onions, R.K., 1978. Rare-earth
abundances in chondriticmeteorites. Geochimica et Cosmochimica Acta
42 (8), 11991212.
Friedrich, J.M., Wang, M.S., Lipschutz, M.E., 2002. Comparison
of the trace elementcomposition of Tagish Lake with other primitive
carbonaceous chondrites. Meteor-itics & Planetary Science 37
(5), 677686.
Garbe-Schonberg, C.-D., 1993. Simultaneous determination of
thirty-seven trace ele-ments in twenty-eight international rock
standards by ICP-MS. Geostandards andGeoanalytical Research 17 (1),
8197.
Goreva, J.S., Burnett, D.S., 2001. Phosphate control on the
thorium/uranium variationsin ordinary chondrites: improving solar
system abundances. Meteoritics & Plane-tary Science 36 (1),
6374.
Govindaraju, K., 1994. Compilation of working values and sample
description for 383geostandards. Geostandard Newsletter 18 (S1),
1158.
Grossman, L., 1980. Refractory inclusions in the Allende
meteorite. Annual ReviewEarth and Planetary Science 8, 559608.
Haskin, L.A., Frey, F.A., Schmitt, R.A., Smith, R.H., 1966.
Meteoritic, solar and terrestrialrare-earth distributions. Physics
and Chemistry of the Earth 7, 167321.
Horwitz, E.P., McAlister, D.R., Bond, A.H., Barrans, R.E., 2005.
Novel extraction of chro-matographic resins based on
tetraalkyldiglycolamides: characterization and po-tential
applications. Solvent Extraction and Ion Exchange 23 (3),
319344.
Hu, Z., Gao, S., 2008. Upper crustal abundances of trace
elements: a revision and up-date. Chemical Geology 253 (34),
205221.
Huang, F., Li, S., Dong, F., Li, Q., Chen, F., Wang, Y., Yang,
W., 2007. Recycling of deeplysubducted continental crust in the
Dabie Mountains, central China. Lithos 96 (12),151169.
Ionov, D.A., Savoyant, L., Dupuy, C., 1992. Application of the
ICP-MS technique to trace-element analysis of peridotites and their
minerals. Geostandard Newsletter 16 (2),311315.
Jadhav, M., 2009. Multi-element isotopic analyses of presolar
graphite grains from theOrgueil meteorite. Washington University in
St. Louis, University Libraries. Pro-quest Dissertations And Theses
2009. Section 0252, Part 0606: 164 pages.
Jain, J.C., Field, M.P., Neal, C.R., Ely, J.C., Sherrell, R.M.,
2000. Determination of the REE ingeological reference materials
DTS-1 (Dunite) and PCC-1 (Peridotite) by ultrasonicand
microconcentric desolvating nebulisation ICP-MS. Geostandards and
Geoana-lytical Research 24 (1), 6572.
Jarosewich, E., CLarke, R.S.J., Barrows, J.N.E., 1987. The
Allende meteorite referencesample. Smithsonian Contributions to the
Earth Sciences 27 (1).
Jarvis, K.E., 1988. Inductively coupled plasma mass-spectrometry
a new techniquefor the rapid or ultra-trace level determination of
the rare-earth elements in geo-logical materials. Chemical Geology
68 (12), 3139.
Jiang, N., Liu, Y., Zhou, W., Yang, J., Zhang, S., 2007.
Derivation of Mesozoic adakiticmagmas from ancient lower crust in
the North China craton. Geochimica et Cosmo-chimica Acta 71 (10),
25912608.
Jupiter, S.D., 2008. Coral rare earth element tracers of
terrestrial exposure in nearshorecorals of the Great Barrier Reef.
Proceedings of the 11th International Coral ReefSymposium, Ft.
Lauderdale, Florida, Session number 4.
Kelley, K.A., Plank, T., Ludden, J., Staudigel, H., 2003.
Composition of altered oceaniccrust at ODP Sites 801 and 1149.
Geochemistry, Geophysics, Geosystems 4 (6).
Kent, A.J.R., Jacobsen, B., Peate, D.W., Waight, T.E., Baker,
J.A., 2004. Isotope dilution MC-ICP-MS rare earth element analysis
of geochemical reference materials NIST SRM610, NIST SRM 612, NIST
SRM 614, BHVO-2G, BHVO-2, BCR-2G, JB-2, WS-E, W-2,AGV-1 and AGV-2.
Geostandards and Geoanalytical Research 28 (3), 417429.
Le Fevre, B., Pin, C., 2005. A straightforward separation scheme
for concomitant Lu-Hfand Sm-Nd isotope ratio and isotope dilution
analysis. Analytica Chimica Acta543 (12), 209221.
Li, B.P., Zhao, J.X., Collerson, K.D., Greig, A., 2003.
Application of ICP-MS trace element analysisin study of ancient
Chinese ceramics. Chinese Science Bulletin 48 (12), 12191224.
Lodders, K., 2003. Solar system abundances and condensation
temperatures of the ele-ments. The Astrophysical Journal 591 (2),
12201247.
Lodders, K., Palme, H., Gail, H.P., 2009. Abundances of the
elements in the solar system. Land-olt-Brnstein, New Series,
Astron. Astrophy. Springer Verlag, BerlinarXiv:0901.1149.
MacPherson, G.J., Wark, D.A., Armstrong, J.T., 1988. Primitive
material surviving inchondrites refractory inclusions. Meteorites
and the Early Solar System. Univer-sity of Arizona Press, Tucson,
AZ, pp. 746807.
Madinabeitia, S.G.d., Lorda, M.E.S., Ibarguchi, J.I.G., 2008.
Simultaneous determinationof major to ultratrace elements in
geological samples by fusiondissolution and in-ductively coupled
plasma mass spectrometry techniques. Analytica Chimica Acta625 (2),
117130.
Mahlen, N.J., Beard, B.L., Johnson, C.M., Lapen, T.J., 2008. An
investigation of dissolutionme