1. Magnesium Isotopic Compositions of International Geological
Reference Materials
Fang-Zhen Teng (1,2)*, Wang-Ye Li (3), Shan Ke (4), Wei Yang (5), Sheng-Ao Liu (4), Fatemeh
Sedaghatpour (2,11), Shui-Jiong Wang (1), Kang-Jun Huang (6), Yan Hu (1,2), Ming-Xing Ling (7),
Yan Xiao (5), Xiao-Ming Liu (8,12), Xiao-Wei Li (4), Hai-Ou Gu (3), Corliss K. Sio (9), Debra A.
Wallace (2), Ben-Xun Su (5), Li Zhao (10), Johnnie Chamberlin (2), Melissa Harrington (1) and
Aaron Brewer (1)
(1) Isotope Laboratory, Department of Earth and Space Sciences, University of Washington, Seattle, WA, 98195, USA
(2) Isotope Laboratory, Department of Geosciences, University of Arkansas, Fayetteville, AR, 72701, USA
(3) CAS Key Laboratory of Crust-Mantle Materials and Environments, School of Earth and Space Sciences, University of Science and Technology
of China, Hefei, Anhui, 230026, China
(4) State Key Laboratory of Geological Processes and Mineral Resources, School of Earth Science and Mineral Resources, China University of
Geosciences, Beijing, 100083, China
(5) State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, P.O. Box 9825, Beijing,
100029, China
(6) School of Earth and Space Sciences, Peking University, Beijing, 100871, China
(7) State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou, 510640,
China
(8) Department of Geology, University of Maryland, College Park, MD, 20742, USA
(9) Origins Laboratory, Department of the Geophysical Sciences, The University of Chicago, 5734 South Ellis Avenue, Chicago, IL , 60637, USA
(10) Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong, China
(11) Present address: Department of Earth and Planetary Sciences, Harvard University, 20 Oxford Street, Cambridge, MA , 02138, USA
(12) Present address: Geophysical Lab, Carnegie Institute of Washington, Washington, DC, 20015, USA
* Corresponding author. e-mail: fteng@u.washington.edu
Magnesium isotopic compositions are reported for
twenty-four international geological reference materials
including igneous, metamorphic and sedimentary rocks,
as well as phlogopite and serpentine minerals. The long-
term reproducibility of Mg isotopic determination, based
on 4-year analyses of olivine and seawater samples, was
≤ 0.07‰ (2s) for d26
Mg and ≤ 0.05‰ (2s) for d25
Mg.
Accuracy was tested by analysis of synthetic reference
materials down to the quoted long-term reproducibility.
This comprehensive dataset, plus seawater data pro-
duced in the same laboratory, serves as a reference for
quality assurance and inter-laboratory comparison of
high-precision Mg isotopic data.
Keywords: magnesium, isotope, MC-ICP-MS, sediments,
silicate reference materials, stable isotopes.
Les compositions isotopiques du magnesium sont fournies
pour vingt-quatre materiaux geologiques de reference
internationaux, comprenant des roches ignees, metam-
orphiques et sedimentaires, ainsi qu’une phlogopite et
des serpentines. La reproductibilite a long terme de la
determination isotopique du Mg, basee des analyses sur
quatre ans d’echantillons d’olivine et d’eau de mer, etait
≤ 0.07% (2s) pour d26
Mg et ≤ 0.05% (2s) pour d25
Mg.
La precision a ete testee par l’analyse de materiaux de
reference synthetiques jusqu’a la reproductibilite a long
terme indiquee. Cette base de donnees complete, ainsi
que des donnees d’eau de mer produites dans le m^eme
laboratoire, servent de reference pour l’assurance qualite
et la comparaison inter-laboratoires de haute precision
des donnees isotopiques du Mg.
Mots-clés : magnesium, isotope, MC-ICP-MS, sediments,
materiaux de reference silicates, isotopes stables.Received 13 Jun 14 – Accepted 09 Oct 14
Significant advances have been made on Mg isotope
geochemistry over the past decade. It was debated whether or
nottheEarthhasachondriticMgisotopiccomposition.Themost
recent studies indicate that the Earth, as well as the Moon, have
Mgisotopiccomposition similartochondriteswithin 0.07‰(2s)
in 26
Mg/24
Mg ratio (Teng et al. 2007, 2010a, Wiechert and
Vol. 39 — N° 3 09
15
P. 329 – 339
3 2 9
doi: 10.1111/j.1751-908X.2014.00326.x
© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

2. Halliday 2007, Handler et al. 2009, Yang et al. 2009,
Bourdon et al. 2010, Chakrabarti and Jacobsen 2010,
Dauphas et al. 2010, Huang et al. 2011, Liu et al. 2011,
Pogge von Strandmann et al. 2011, Sedaghatpour et al.
2013, Xiao et al. 2013). Nonetheless, the causes for the
debate are still not fully understood. One possibility is that the
early studies drew their conclusions from few, unrepresentative
samples. Another explanation is laboratory analytical artefact.
Indeed, studies of some geological reference materials (e.g.,
BCR-1) by different laboratories yielded different Mg isotopic
compositions, ranging over 0.5‰ in 26
Mg/24
Mg ratio (Young
and Galy 2004, Teng et al. 2007, Wiechert and Halliday
2007, Huang et al. 2009, Bourdon et al. 2010, Chakrabarti
and Jacobsen2010).Thisvariationisabouttwicethedifference
(~ 0.3‰) proposed between the Earth and chondrites
(Wiechert and Halliday 2007), indicating the possibility of the
existence of analytical artefacts in some laboratories.
One way to ensure data quality and avoid analytical
artefacts is through analysis of well-characterised geological
reference materials (RMs). However, the database of Mg
isotopic geological RMs is still very limited. Most Mg isotopic
data for geological RMs were reported randomly in literature
and the precision, especially the accuracy, was not assured. In
this study, we report highly precise and accurate Mg isotopic
data for 24 international geological RMs from United States
Geological Survey (USGS), Geological Society of Japan (GSJ),
Centre de Recherches Petrographiques et Geochimiques
(CRPG), France, Canada Center for Mineral and Energy
Technology, Mines and Resources (CCMETMR), and Associa-
tion Nationale de la Recherche Technique (ANRT), France.
These RMs vary greatly in matrices and include igneous rocks
ranging from ultramafic to felsic in composition, metamorphic
and sedimentary rocks, and minerals (phlogopite and serpen-
tine). This dataset, together with Mg isotopic data reported by
the same laboratory for seawater samples, lays the foundation
for using these geological RMs for quality assurance and inter-
laboratory comparison of high-precision Mg isotopic data.
Sample dissolution, column chemistry
and instrumental analysis
All experiments were performed at the Isotope Labora-
tory of the University of Arkansas, Fayetteville. The detailed
procedures for routine sample dissolution, column chemistry
and instrumental analysis of Mg isotopes have been
reported in our previous publications (Teng et al. 2007,
2010a, Yang et al. 2009, Li et al. 2010, Ling et al. 2013b,
Sedaghatpour et al. 2013, Teng and Yang 2014).
Test portions of all reference materials were dissolved by
using concentrated acids in the following sequence: (a) HF-
HNO3, (b) HNO3-HCl, (c) HNO3 or (d) HCl in Teflon vials on a
hot plate. A few drops of concentrated HClO4 were also
added in step (1) for samples containing organic materials.
Separation of Mg was achieved by cation exchange
chromatography, loaded with pre-cleaned Bio-Rad AG
50W-X8 (200–400 mesh) resin. Magnesium was eluted in
1 mol l-1
HNO3 media (Teng et al. 2007, Yang et al. 2009,
Li et al. 2010). The purified sample solution was analysed
using a Nu Plasma MC-ICP-MS with ‘wet’ plasma, consisting
of a Cinnabar spray chamber and a MicroMist micro-uptake
glass concentric nebuliser (Teng and Yang 2014). Magne-
sium isotopic ratios were determined in low-resolution mode,
with 26
Mg, 25
Mg and 24
Mg measured simultaneously in
separate Faraday cups (H5, Ax and L4).
Though our above procedures are the conventional ones
for measuring Mg and other non-traditional stable isotopes by
MC-ICPMS, it is important to point out that the procedure for
cleaning resins can affect the accuracy of Mg isotopic
determination. We typically used 6 mol l-1
HCl, Milli-Q H2O
and 1 mol l-1
HNO3 to clean resins. This procedure is
successful most of the time, except in a few cases. For example,
synthetic RMs and natural samples, which were processed
through the Bio-Rad AG 50W-X8 resin (200–400 mesh) that
had been cleaned using HCl and HNO3, consistently yielded
lower d26
Mg than the expected values (Figure 1). After
treating the resin with 0.5 mol l-1
HF, samples processed
-0.1
0.0
-0.2
-0.4
-1.2
-0.8
δ26
Mg
0.0
0.4
0.1
Δ26
Mgdirty-clean
Dirty resin
Clean resin
Granites Koolau basaltsAllende
Seawater
IL-Mg-1
DTS-2
Figure 1. Effects of resin impurities on Mg isotopic composi-
tions of samples and reference materials. Dirty resin = samples
through columns loaded with resin that was cleaned by using
HCl and HNO3 only. Clean resin = samples through columns
loaded with resin that was cleaned by using HCl, HNO3 and HF.
The top panel shows the difference in Mg isotopic composition
for samples processed through both ‘dirty resin’ and ‘clean resin’
(D26
Mgdirty-clean = d26
Mgdirty – d26
Mgclean). d26
Mg of samples
processed through ‘dirty resin’ were systematically lower than
those through ‘clean resin’ and diverged from the expected true
values. See text for details.
3 3 0 © 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

3. through the HF-cleaned resin yielded the expected values. This
is similar to previous studies that found samples processed
through AG 50W-X12 resin without HF treatment yielded
significant variation in the instrumental fractionation (Chang
et al. 2003). This effect was attributed to the presence of a
significant amount of Si in the resin, causing matrix effects
(Chang et al. 2003). It is worthwhile to note that this effect does
not occur for all batches of resins. Nonetheless, since the
discovery of this effect, each batch of resin was first tested by
processing a pure Mg RM, a synthetic solution and the
Kilbourne Hole olivine. After the expected values of these RMs
were achieved, we started processing samples with unknown
Mg isotopic compositions.
Data reduction and presentation
Magnesium isotopic compositions were determined by
the sample-calibrator bracketing method in a sequence of
calibrator1 – sample1 – calibrator2 – sample2 – calibrator3. . .
The average value of the two bracketing calibrators was
used to correct the sample analysis for instrumental mass
fractionation. The isotopic ratios are then reported in
d-notation:
dX
Mgsamplei
¼ 103
Â
ðX
Mg=24
MgÞsamplei
ðX
Mg=24MgÞcalibratori
þðX Mg=24MgÞcalibratoriþ1
2
À 1
8
:
9
=
;
ð1Þ
where X refers to mass 25 or 26 and calibrator refers to
DSM3 (Galy et al. 2003). Since all samples analysed in this
study follow mass-dependent fractionation, we use d26
Mg in
our discussion exclusively.
The calibrator-sample sequences were repeated n times.
The reported isotopic composition for a given sample is the
average of n repeat analyses. The d26
Mg values for
calibrators were also computed by considering each
calibratori as a sample bracketed by two nearby calibrators
(calibratori-1 and calibratori+1):
dX
Mgcalibratori
¼ 103
Â
ðX
Mg=24
MgÞcalibratori
ðX Mg=24MgÞcalibratoriÀ1
þðX Mg=24MgÞcalibratoriþ1
2
À 1
8
:
9
=
;
ð2Þ
The dispersion of the d26
Mg values for RMs computed by
the calibrator-bracketing method was used to estimate the
uncertainty from instrumental instability by assuming that the
standard deviation (s) of the sample measurements equals
the standard deviation of the calibrator measurements. The
advantage of using calibrator rather than samples to
quantify the instrumental uncertainty is that there are more
calibrator analyses than samples during a batch run. The
standard deviation is thus better known. The standard
deviation derived from calibrators, in most cases, is larger
than that from samples, and hence is more conservative.
Accuracy check and long-term
reproducibility
Many processes can potentially affect the accuracy of
high-precision Mg isotopic determination (Teng and Yang
2014, and references therein). Two approaches are gener-
ally used to evaluate the long-term accuracy of isotopic
determination. One is to process synthetic RMs that mimic the
matrices of natural samples. The other is through analysis of
well-characterised RMs. In this study, we first validate our
instrumental analysis by measurements of a pure Mg RM.
We then evaluated the whole-procedural accuracy by
processing synthetic standard solutions with known isotopic
compositions. Finally, we carried out repeated, full-proce-
dural analyses of two samples with different matrices
(Kilbourne Hole olivine and seawater) over a 4-year period
to examine the long-term measurement reproducibility.
Cambridge-1 has been analysed in different laborato-
ries using different types of instrumental analysis, and has
yielded very consistent values (Galy et al. 2003). Nine
repeated analyses of Cambridge-1 yielded an arithmetic
mean value of -2.623 ± 0.030‰ for d26
Mg and -
1.358 ± 0.030‰ (2s, n = 9) for d25
Mg (Table 1), identical
to values in the literature (Galy et al. 2003). Nonetheless, the
pure Mg reference material Cambridge-1 does not need
sample preparation, hence does not test for uncertainties
associated with sample dissolution and column chemistry as
is needed for natural samples. More importantly, sample
preparation processes are the key steps to monitor since they
have the greatest potential to introduce analytical uncer-
tainties and artefacts. To address this, we doped the pure
Mg RM with different amounts of mono-element RMs to
synthesise various sample solutions that match major
elemental compositions of natural samples. All these
synthetic samples were treated as regular samples and
processed through column chemistry. The purified Mg cuts
were analysed relative to the unprocessed pure Mg
standard solution. Analyses of these various synthetic stan-
dard solutions over 4 years (2009–2012) yielded arithmetic
mean d26
Mg of -0.027 ± 0.065‰ and d25
Mg of -
0.016 ± 0.039‰ (2s, n = 58), falling within the range of
the expected true value of 0 within uncertainties (Figure 2).
3 3 1© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

4. These tests indicate that our measurements are accurate at
our quoted precision.
Our long-term reproducibility was further evaluated by
analysing Kilbourne Hole (KH) olivine and seawater sample
from Hawaii over a period of 4 years. Analyses of KH olivine
yielded an arithmetic mean d26
Mg value of -0.266 ±
0.068‰ and d25
Mg value of -0.137 ± 0.050‰ (2s,
n = 210) (Figure 3). Analyses of Hawaiian seawater gave
an arithmetic mean d26
Mg value of -0.843 ± 0.057‰ and
d25
Mg value of -0.433 ± 0.040‰ (2s, n = 102) (Figure 4).
This value is consistent with Mg isotopic composition of
seawater samples worldwide, measured in both the same
and other laboratories (Foster et al. 2010, Ling et al. 2011,
and references therein).
Table 1.
Magnesium isotopic compositions of Cambridge-1
Cambridge-1 d26
Mg 2s d25
Mg 2s
Replicate -2.612 0.055 -1.337 0.050
Replicate -2.636 0.085 -1.360 0.062
Replicate -2.612 0.078 -1.345 0.061
Replicate -2.621 0.070 -1.350 0.054
Replicate -2.636 0.069 -1.363 0.047
Replicate -2.609 0.093 -1.354 0.056
Replicate -2.648 0.093 -1.382 0.059
Replicate -2.631 0.046 -1.381 0.065
Replicate -2.605 0.071 -1.353 0.050
Recommended -2.623 0.030 -1.358 0.030
2s = 2 times the standard deviation of the population of n (n 20) repeat measurements of the reference materials during an analytical session.
-0.1
-0.2
δ26
Mg
0.0
0.2
0.1
Synthetic solutions
II III IV V VII
Figure 2. Magnesium isotopic compositions of syn-
thetic solutions that were made from the pure Mg
reference material FZT-Mg with various matrices. FZT-
Mg purified by column chromatography from these
mixtures was analysed by bracketing with pure FZT-Mg
not passed through column. I = pure Mg reference
material FZT-Mg. II = IL-granite with concentration
ratios of Mg:Fe:Al:Ca:Na:K:Ti:
Ni = 1:5:30:5:10:20:0.1:0.1; III = IL-Lunar basalt with
concentration ratios of Mg:Fe:Ca:Ti:Ni = 1:4:4:4:0.1;
IV = IL-Chondrite with concentration ratios of Mg:Fe:
Ca:Ni:Al:Na:K:Ti = 1:3:0.2:0.2:0.2:0.1:0.1:0.1; V = IL-
Clinopyroxene with concentration ratios of Mg:Fe:Ca:
Al = 1:0.5:2:0.1; VI = IL-Mg-1 with concentration
ratios of Mg:Fe:Al:Ca:Na:K:Ti = 1:1:1:1:1:1:0.1. The
expected true value of zero (horizontal line) is plotted
for comparison.
-0.1
-0.2
δ26
Mg
0.0
-0.3
-0.4
-0.5
Kilbourne Hole (KH) olivine
δ26
Mg = -0.266 ± 0.068 (2s, n = 210)
Figure 3. Magnesium isotopic compositions of the Kil-
bourne Hole olivine analysed over a period of 4 years.
The horizontal line represents the mean d26
Mg value.
-1.0
-0.9
δ26
Mg
-0.8
-0.7
-0.6
Hawaiian seawater
δ26
Mg = -0.843 ± 0.057 (2s, n = 102)
Figure 4. Magnesium isotopic compositions of a sea-
water sample from south-western Hawaii analysed
over a period of 4 years. The horizontal line represents
the mean d26
Mg value.
3 3 2 © 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

5. Overall, results from the above tests on both pure Mg
RM, synthetic standard solutions and natural samples
demonstrate that Mg isotopes can be measured accurately
with a precision of ≤ ±0.07‰ for d26
Mg and ≤ ±0.05‰ for
d25
Mg (2s).
Magnesium isotopic compositions of
geological reference materials
Magnesium isotopic compositions of twenty-four geolog-
ical RMs analysed in this study are plotted in Figure 5 and
reported in Table 2, together with literature data published by
other groups. Ten out of the twenty-four RMs are measured for
the first time for Mg isotopes. For the other RMs measured
before, our data agree with literature data within uncertainty
exceptthefollowingcases.Thefirstoneisthatd26
Mgvaluesfor
all RMs except seawater reported by Chakrabarti and
Jacobsen (2010) are systematically (~ 0.3‰) lighter than
other groups (Table 2). It should be noted that terrestrial and
extra-terrestrial samples reported in Chakrabarti and Jacob-
sen (2010) are also (~ 0.3‰) lighter than other groups (Teng
et al. 2007, 2010a, Wiechert and Halliday 2007, Handler
et al. 2009, Yang et al. 2009, Bourdon et al. 2010, Dauphas
et al. 2010, Huang et al. 2011, Liu et al. 2011, Pogge von
Strandmann et al. 2011, Sedaghatpour et al. 2013, Xiao
et al. 2013). The second case is that d26
Mg values of the
granite reference material GA and diorite reference material
DR-N analysed by Bolou-Bi et al. (2009) and Brenot et al.
(2008) are 0.5 and 0.3‰ lighter than ours. In fact, their d26
Mg
value for GA is the lightest known value for granites to date (Li
et al. 2010, Liu et al. 2010, Telus et al. 2012, Ling et al.
2013a). The reason for this discrepancy is also unclear
though the matrix effect is the most likely cause since this
granite reference material has a low MgO content (0.95%
m/m) (Govindaraju 1994).
To make an internally consistent database, we report the
recommended values for geological RMs based on data
produced in our laboratory. The only exceptions are the
reference materials BHVO-2 and W-2 that were analysed
only once. In this case, we included literature data to
calculate the recommended values. The recommended Mg
isotopic compositions for all rock, mineral and seawater
reference materials analysed in our study were calculated as
a weighted average of independent replicate analyses
using the following equations:
weighted mean ¼
P x
s2
P 1
s2
; variation on mean ¼
ffiffiffiffiffiffiffiffiffi
1
P 1
s2
s
where x = mean Mg isotopic composition of a RM; s is one
standard deviation.
The calculated weighted 2s in all cases was 0.06‰
for d26
Mg and 0.04‰ for d25
Mg. Since accuracy in our
laboratory has only been tested down to a level of ~ 0.07‰
for d26
Mg and 0.05‰ for d25
Mg, we therefore used the
2s of the population of replicate measurements. In addition
to these rock and mineral reference materials, recom-
mended values are also reported for seawater (Table 2).
As discussed in Ling et al. (2011), seawater may serve as a
good Mg isotopic RM for accuracy checks. Our recom-
mended values for seawater were calculated based on
data measured in our laboratory and results from the
literature (Ling et al. 2011 and references therein).
Overall, the recommended d26
Mg values of geological
RMs vary from -1.371 to -0.107‰. No significant correlation
between d26
Mg and MgO is observed because most of these
igneous rock reference materials fall within the range of
mantle values (Figure 6). The reference materials that are
significantly lighter than mantle values are phlogopite (Mica-
Mg, d26
Mg = -1.371‰), shale SGR-1 (d26
Mg = -0.995‰),
shale SCo-1 (d26
Mg = -0.890‰), syenite SY-2 (d26
Mg = -
0.425‰) and basalt BR (d26
Mg = -0.409‰). Nonetheless,
they fall within the range of crustal and mantle minerals and
rocks (Li et al. 2010, Liu et al. 2010, 2014, Teng et al. 2010a,
2013, Wang et al. 2012, 2014, Huang et al. 2013).
Our study shows that precise and accurate Mg isotopic
data can be routinely achieved in a wide range of natural
rocks, minerals and water samples. Though precision is
usually well established, accuracy may not have been
0.2
0.0
-0.2
-0.4
-0.6
-0.8
0.40.0-1.2 -0.4-1.6 -0.8
δ26
Mg
δ25
Mg
Figure 5. Magnesium three-isotope plot of all reference
samples analysed in this study. The solid line represents
the fractionation line with a slope of 0.515. The larger
grey-filled circles represent recommended values for
reference samples. Data are reported in Table 2.
3 3 3© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

6. Table 2.
Recommended magnesium isotopic compositions of geological reference materials and seawater
RM Name Description d26
Mg 2s d25
Mg 2s
Ultramafic Rocks
DTS-1 Dunite, Twin Sisters
Mountain, Washington,
USA (USGS)
-0.303 0.055 -0.129 0.086
Replicate -0.303 0.055 -0.13 0.086
Replicate -0.308 0.079 -0.138 0.066
Replicate -0.300 0.09 -0.13 0.05
Replicate -0.300 0.05 -0.13 0.09
Recommended -0.302 0.006 -0.132 0.007
Literature-1 -0.25 0.08 -0.12 0.02
Literature-2 -0.30 0.16 -0.13 0.02
Literature-3 -1.03 0.28 -0.52 0.11
DTS-2 Dunite, Twin Sisters
Mountain, Washington,
USA (USGS)
-0.33 0.05 -0.16 0.06
Replicate -0.27 0.10 -0.13 0.07
Replicate -0.30 0.09 -0.15 0.06
Replicate -0.30 0.10 -0.16 0.05
Replicate -0.34 0.06 -0.16 0.06
Replicate -0.34 0.06 -0.18 0.04
Replicate -0.32 0.07 -0.18 0.05
Replicate -0.29 0.06 -0.14 0.05
Replicate -0.38 0.14 -0.19 0.03
Replicate -0.32 0.07 -0.18 0.05
Replicate -0.381 0.141 -0.187 0.028
Replicate -0.267 0.095 -0.129 0.067
Replicate -0.299 0.095 -0.147 0.063
Replicate -0.296 0.095 -0.158 0.052
Replicate -0.327 0.054 -0.161 0.056
Replicate -0.343 0.06 -0.176 0.038
Replicate -0.32 0.066 -0.18 0.054
Replicate -0.292 0.062 -0.141 0.051
Replicate -0.343 0.06 -0.176 0.038
Replicate -0.32 0.066 -0.18 0.054
Recommended -0.32 0.06 -0.17 0.04
Literature-4 -0.23 0.03 -0.12 0.02
PCC-1 Peridotite, Austin Creek,
California, USA (USGS)
-0.197 0.072 -0.096 0.054
Replicate -0.230 0.087 -0.106 0.064
Replicate -0.251 0.06 -0.102 0.038
Recommended -0.229 0.055 -0.101 0.010
Literature-2 -0.22 0.10 -0.15 0.07
Literature-3 -0.51 0.32 -0.26 0.16
Mafic Rocks
BHVO Basalt, Hawaiian
Volcanic Observatory,
Hawaii, USA (USGS)
-0.240 0.076 -0.117 0.047
Replicate -0.216 0.063 -0.142 0.05
Replicate -0.171 0.066 -0.112 0.045
Recommended -0.207 0.070 -0.123 0.033
BHVO-1 Basalt, Hawaiian
Volcanic Observatory,
Hawaii, USA (USGS)
-0.225 0.139 -0.105 0.085
Replicate -0.252 0.064 -0.13 0.041
Replicate -0.185 0.059 -0.123 0.062
Recommended -0.217 0.067 -0.125 0.026
Literature-5 -0.12 0.05 -0.06 0.03
Literature-6 -0.09 0.02 -0.04 0.02
Literature-2 -0.30 0.08 -0.14 0.04
Literature-3 -0.59 0.27 -0.29 0.12
Literature-7 -0.16 0.23 -0.08 0.10
3 3 4 © 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

7. Table 2 (continued).
Recommended magnesium isotopic compositions of geological reference materials and seawater
RM Name Description d26
Mg 2s d25
Mg 2s
BHVO-2 Basalt, Hawaiian
Volcanic Observatory,
Hawaii, USA (USGS)
-0.24 0.08 -0.12 0.05
Literature-4 -0.19 0.02 -0.10 0.01
Literature-8* -0.14 0.21 -0.06 0.11
Literature-9 -0.25 0.11 -0.13 0.08
Literature-10 -0.26 0.06 -0.14 0.04
Literature-11 -0.30 0.62 -0.15 0.32
Literature-12 -0.20 0.08 -0.10 0.03
Recommended -0.20 0.07 -0.10 0.04
BR Basalt, Nancy, France
(CRPG)
-0.430 0.060 -0.200 0.108
Replicate -0.415 0.095 -0.220 0.067
Replicate -0.375 0.078 -0.165 0.056
Replicate -0.372 0.054 -0.200 0.056
Replicate -0.478 0.075 -0.254 0.052
Recommended -0.409 0.087 -0.210 0.065
JB-1 Basalt, Sasebo,
Nagasaki-Ken, Japan
(GSJ)
-0.334 0.047 -0.164 0.068
Replicate -0.257 0.114 -0.144 0.049
Replicate -0.280 0.054 -0.136 0.056
Replicate -0.221 0.050 -0.126 0.049
Replicate -0.201 0.078 -0.132 0.056
Replicate -0.326 0.075 -0.180 0.052
Recommended -0.276 0.098 -0.146 0.041
QMC I3 Dolerite (Queen Mary
College, U.K.)
-0.201 0.093 -0.122 0.059
Replicate -0.197 0.088 -0.081 0.038
Replicate -0.276 0.072 -0.139 0.063
Replicate -0.230 0.084 -0.086 0.048
Recommended -0.232 0.073 -0.098 0.056
W-1 Diabase, Bull Run quarry,
Centreville, Virginia,
USA (USGS)
-0.154 0.063 -0.081 0.050
Replicate -0.139 0.068 -0.065 0.050
Replicate -0.165 0.078 -0.051 0.054
Replicate -0.118 0.078 -0.091 0.056
Replicate -0.209 0.054 -0.113 0.056
Recommended -0.164 0.068 -0.079 0.048
Literature-2 -0.13 0.13 -0.07 0.06
W2 Diabase, Bull Run quarry,
Centreville, Virginia,
USA (USGS)
-0.163 0.082 -0.066 0.102
Literature-13 -0.16 0.10 -0.09 0.06
Recommended -0.16 0.01 -0.08 0.03
Intermediate Rocks
DR-N Diorite, Massif Cham du
Feu, Vosges, France
(ANRT)
-0.193 0.119 -0.09 0.056
Replicate -0.154 0.054 -0.08 0.056
Recommended -0.161 0.056 -0.085 0.014
Literature-14 -0.50 0.04 -0.30 0.04
Literature-15 -0.52 0.04 -0.25 0.04
AGV-1 Andesite, Guano Valley,
Oregon, USA (USGS)
-0.122 0.016 -0.037 0.055
Replicate -0.159 0.095 -0.072 0.052
Replicate -0.103 0.088 -0.076 0.064
Recommended -0.123 0.057 -0.061 0.043
Literature-2 -0.32 0.080 -0.16 0.05
Literature-13 -0.03 0.10 -0.01 0.01
JG-1 Granodiorite, Sori,
Gumma-Ken, Japan
(GSJ)
-0.294 0.054 -0.151 0.056
Replicate -0.340 0.088 -0.173 0.064
Recommended -0.306 0.065 -0.16 0.031
3 3 5© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

8. Table 2 (continued).
Recommended magnesium isotopic compositions of geological reference materials and seawater
RM Name Description d26
Mg 2s d25
Mg 2s
Felsic rocks
G-2 Granite, Bradford, Rhode
Island, USA (USGS)
-0.180 0.042 -0.095 0.036
Replicate -0.198 0.059 -0.076 0.068
Replicate -0.143 0.068 -0.084 0.050
Replicate -0.193 0.065 -0.089 0.053
Replicate -0.105 0.038 -0.062 0.054
Replicate -0.162 0.095 -0.048 0.063
Replicate -0.151 0.081 -0.023 0.055
Replicate -0.079 0.085 -0.034 0.062
Replicate -0.161 0.095 -0.094 0.052
Replicate -0.170 0.088 -0.112 0.064
Replicate -0.128 0.078 -0.054 0.056
Replicate -0.129 0.064 -0.105 0.041
Recommended -0.148 0.071 -0.078 0.057
GA Granite, Andlau, Alsace,
France (CRPG)
-0.240 0.093 -0.129 0.059
Replicate -0.215 0.074 -0.100 0.052
Replicate -0.26 0.07 -0.14 0.05
Replicate -0.29 0.09 -0.12 0.07
Replicate -0.215 0.074 -0.100 0.052
Replicate -0.214 0.095 -0.135 0.074
Replicate -0.19 0.09 -0.08 0.06
Replicate -0.21 0.09 -0.07 0.06
Recommended -0.230 0.07 -0.11 0.05
Literature-15 -0.75 0.14 -0.36 0.08
Literature-2 -0.34 0.15 -0.17 0.11
GS-N Granite, Senones,
Vosges, France (ANRT)
-0.276 0.081 -0.131 0.217
Replicate -0.275 0.085 -0.138 0.075
Replicate -0.258 0.074 -0.137 0.052
Replicate -0.22 0.07 -0.10 0.05
Replicate -0.21 0.09 -0.13 0.07
Replicate -0.258 0.074 -0.137 0.052
Replicate -0.286 0.095 -0.122 0.074
Replicate -0.19 0.04 -0.11 0.07
Recommended -0.23 0.07 -0.12 0.03
Literature-2 -0.24 0.23 -0.12 0.13
SY-2 Syenite, Bancroft, Eastern
Ontario, Canada
(CCMETMR)
-0.407 0.087 -0.213 0.064
Replicate -0.446 0.088 -0.200 0.038
Replicate -0.374 0.072 -0.246 0.063
Replicate -0.493 0.084 -0.220 0.048
Recommended -0.425 0.103 -0.215 0.039
Sedimentary, Metamorphic Rocks and Minerals
MAG-1 Marine mud, Wilkinson
Basin, Gulf of Maine,
USA (USGS)
-0.240 0.058 -0.074 0.063
Replicate -0.251 0.088 -0.116 0.064
Replicate -0.300 0.116 -0.145 0.059
Recommended -0.252 0.070 -0.113 0.041
Literature-1 -0.23 0.09 -0.14 0.04
SCo-1 Shale, Natrona County,
Wyoming, USA (USGS)
-0.845 0.064 -0.444 0.042
Replicate -0.906 0.093 -0.476 0.059
Replicate -0.898 0.088 -0.463 0.038
Replicate -0.937 0.075 -0.499 0.056
Recommended -0.890 0.077 -0.466 0.046
SGR-1 Shale, Green River
Formation, USA (USGS)
-0.933 0.087 -0.508 0.064
Replicate -1.005 0.095 -0.523 0.059
Replicate -0.946 0.088 -0.493 0.064
Replicate -1.015 0.038 -0.513 0.053
Recommended -0.995 0.083 -0.510 0.025
Literature-1 -0.98 0.12 -0.50 0.06
3 3 6 © 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

9. checked at the same level as the precision in other
laboratories. The geological reference materials reported
in this study make it easier for sample calibration and quality
assurance for future Mg isotopic studies.
Acknowledgements
Constructive comments from two anonymous reviewers,
and efficient handling from Michel Gregoire are greatly
appreciated. This work was financially supported by the
National Science Foundation (EAR-0838227, EAR-
1056713 and EAR-1340160) to FZT. MH was partially
supported by the University of Washington Mary Gates
Research Scholarship.
References
Abbey S. (1980)
Studies in “standard samples” for use in the general
analysis of silicate rocks and minerals. Geostandards
Newsletter, 4, 163–190.
Table 2 (continued).
Recommended magnesium isotopic compositions of geological reference materials and seawater
RM Name Description d26
Mg 2s d25
Mg 2s
SDC-1 Mica Schist, Washington
D.C., USA (USGS)
-0.095 0.093 -0.06 0.059
Replicate -0.109 0.084 -0.069 0.048
Replicate -0.099 0.088 -0.045 0.038
Replicate -0.123 0.084 -0.102 0.055
Recommended -0.107 0.025 -0.064 0.048
UB-N Serpentine, Vosges,
France (ANRT)
-0.179 0.05 -0.082 0.037
Replicate -0.181 0.078 -0.08 0.056
Replicate -0.217 0.060 -0.098 0.038
Recommended -0.192 0.043 -0.088 0.019
Literature-1 -0.12 0.08 -0.06 0.06
Literature-13 -0.16 0.09 -0.08 0.07
Phlogopite Phlogopite, Mica-Mg,
Bekily, Southern
Madagascar (CRPG)
-1.374 0.093 -0.750 0.059
Replicate -1.330 0.088 -0.709 0.038
Replicate -1.395 0.072 -0.710 0.063
Recommended -1.371 0.067 -0.719 0.047
Seawater
Recommended Globally distributed -0.83 0.09 -0.43 0.06
2s = 2 times the standard deviation of the population of n (n 20) repeat measurements of the RMs during an analytical session.
The recommended values in bold were calculated as a weighted average of independent replicate analysis. See text for equations. The 2s of recommended
values are calculated based on replicate data.
USGS, United States Geological Survey; GSJ, Geological Society of Japan; CRPG, Centre de Recherches Petrographiques et Geochimiques, France; CCMETMR,
Canada Center for Mineral and Energy Technology, Mines and Resources; ANRT, Association Nationale de la Recherche Technique, France.
Literature: 1 = Wombacher et al. (2009); 2 = Huang et al. (2009); 3 = Chakrabarti and Jacobsen (2010); 4 = Bizzarro et al. (2011); 5 = Bizzarro et al.
(2005); 6 = Baker et al. (2005); 7 = Thrane et al. (2006); 8 = Wiechert and Halliday (2007); 9 = Pogge von Strandmann et al. (2008); 10 = Pogge von
Strandmann et al. (2011); 11 = Schiller et al. (2010); 12 = Bouvier et al. (2013); 13 = Wang et al. (2011); 14 = Brenot et al. (2008); 15 = Bolou-Bi et al.
(2009). Seawater = Ling et al. (2011 and references therein). Data for some geological RMs that were sporadically reported in previous publications from our
group (Yang et al. 2009, Li et al. 2010, Teng et al. 2010a,b, 2011, Huang et al. 2012, Liu et al. 2014) were also included here.
* Data from this literature for BHVO-2 were not used in the calculation.
-0.9
-0.6
δ26
Mg
0.0
-0.3
-1.2
-1.5
050 10 20 4030
MgO (% m/m)
Figure 6. Recommended Mg isotopic compositions of
geological RMs vs. MgO contents. The solid line and
horizontal bar represent the average d26
Mg of -0.25‰
and two standard deviations of ± 0.07‰ for the Earth
(Teng et al. 2010a). d26
Mg data are reported in
Table 2. MgO data are from the literature (Abbey
1980, Govindaraju 1994).
3 3 7© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

10. Baker J.A., Bizzarro M., Wittig N., Connelly J. and Haack
H. (2005)
Early planetesimal melting from an age of 4.5662
Gyr for differentiated meteorites. Nature, 436,
1127–1131.
Bizzarro M., Baker J.A., Haack H. and Lundgaard K.L.
(2005)
Rapid timescales for accretion and melting of differentiated
planetesimals inferred from 26
Al-26
Mg chronometry. The
Astrophysical Journal, 632, L41–L44.
Bizzarro M., Paton C., Larsen K., Schiller M., Trinquier A.
and Ulfbeck D. (2011)
High-precision Mg-isotope measurments of terrestrial and
extraterrestrial material by HR-MC-ICP-MS – Implications
for the relative and absolute Mg-isotope composition of the
bulk silicate Earth. Journal of Analytical Atomic Spectrom-
etry, 26, 565–577.
Bolou-Bi E.B., Vigier N., Brenot A. and Poszwa A. (2009)
Magnesium isotope compositions of natural reference
materials. Geostandards and Geoanalytical Research,
33, 95–109.
Bourdon B., Tipper E.T., Fitoussi C. and Stracke A. (2010)
Chondritic Mg isotope composition of the Earth. Geochi-
mica et Cosmochimica Acta, 74, 5069–5083.
Bouvier A., Wadhwa M., Simon S.B. and Grossman L.
(2013)
Magnesium isotopic fractionation in chondrules from
the Murchison and Murray CM2 carbonaceous chondrites.
Meteoritics and Planetary Science, 48, 339–353.
Brenot A., Cloquet C., Vigier N., Carignan J. and France-
Lanord C. (2008)
Magnesium isotope systematics of the lithologically varied
Moselle river basin, France. Geochimica et Cosmochimica
Acta, 72, 5070–5089.
Chakrabarti R. and Jacobsen S.B. (2010)
The isotopic composition of magnesium in the Solar System.
Earth and Planetary Science Letters, 293, 349–358.
Chang V.T.-C., Makishima A., Belshaw N.S. and O’Nions
R.K. (2003)
Purification of Mg from low-Mg biogenic carbonates for
isotope ratio determination using multiple collector ICP-
MS. Journal of Analytical Atomic Spectrometry, 18,
296–301.
Dauphas N., Teng F.-Z. and Arndt N.T. (2010)
Magnesium and iron isotopes in 2.7 Ga Alexo komati-
ites: Mantle signatures, no evidence for Soret diffusion,
and identification of diffusive transport in zoned oliv-
ine. Geochimica et Cosmochimica Acta, 74, 3274–3291.
Foster G.L., Pogge von Strandmann P.A.E. and Rae
J.W.B. (2010)
Boron and magnesium isotopic composition of seawater.
Geochemistry Geophysics Geosystems, 11, Q08015. doi:
10.1029/2010GC003201.
Galy A., Yoffe O., Janney P.E., Williams R.W., Cloquet C.,
Alard O., Halicz L., Wadhwa M., Hutcheon I.D., Ramon E.
and Carignan J. (2003)
Magnesium isotope heterogeneity of the isotopic
standard SRM980 and new reference materials for
magnesium-isotope-ratio measurements. Journal of Ana-
lytical Atomic Spectrometry, 18, 1352–1356.
Govindaraju K. (1994)
1994 compilation of working values and sample descrip-
tion for 383 geostandards. Geostandards Newsletter, 18
(Special Issue), 158pp.
Handler M.R., Baker J.A., Schiller M., Bennett V.C. and
Yaxley G.M. (2009)
Magnesium stable isotope composition of Earth’s upper
mantle. Earth and Planetary Science Letters, 282, 306–
313.
Huang F., Glessner J.J., Ianno A., Lundstrom C.C. and
Zhang Z. (2009)
Magnesium isotopic composition of igneous rock stan-
dards measured by MC-ICP-MS. Chemical Geology, 268,
15–23.
Huang F., Zhang Z., Lundstrom C.C. and Zhi X. (2011)
Iron and magnesium isotopic compositions of peridotite
xenoliths from Eastern China. Geochimica et Cosmochi-
mica Acta, 75, 3318–3334.
Huang K.-J., Teng F.-Z., Wei G.-J., Ma J.-L. and Bao Z.-Y.
(2012)
Adsorption- and desorption-controlled magnesium isotope
fractionation during extreme weathering of basalt in
Hainan Island, China. Earth and Planetary Science Letters,
359–360, 73–83.
Huang K.-J., Teng F.-Z., Elsenouy A., Li W.-Y. and Bao Z.-
Y. (2013)
Magnesium isotopic variations in loess: Origins and
implications. Earth and Planetary Science Letters, 374,
60–70.
Li W.-Y., Teng F.-Z., Ke S., Rudnick R.L., Gao S., Wu F.-Y.
and Chappell B.W. (2010)
Heterogeneous magnesium isotopic composition of the
upper continental crust. Geochimica et Cosmochimica
Acta, 74, 6867–6884.
Ling M.-X., Sedaghatpour F., Teng F.-Z., Hays P.D.,
Strauss J. and Sun W. (2011)
Homogenous magnesium isotopic composition of seawa-
ter: An excellent geostandard for Mg isotope analysis.
Rapid Communications in Mass Spectrometry, 25, 2828–
2836.
Ling M.-X., Li Y., Ding X., Teng F.-Z., Yang X.-Y., Fan W.-
M., Xu Y.-G. and Sun W. (2013a)
Destruction of the North China craton induced by ridge
subudctions. Journal of Geology, 121, 197–213.
Ling M.-X., Liu Y.-L., Williams I.S., Teng F.-Z., Yang X.-Y.,
Ding X., Wei G.-J., Xie L.-H., Deng W.-F. and Sun W.
(2013b)
Formation of the world’s largest REE deposit by
protracted fluxing of carbonatite by subduction-derived
fluids. Scientific Reports, 3, 1776, doi:10.1038/
srep01776.
Liu S.-A., Teng F.-Z., He Y., Ke S. and Li S. (2010)
Investigation of magnesium isotope fractionation during
granite differentiation: Implication for Mg isotopic compo-
sition of the continental crust. Earth and Planetary Science
Letters, 297, 646–654.
3 3 8 © 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts

11. references
Liu S.-A., Teng F.-Z., Yang W. and Wu F.Y. (2011)
High-temperature inter-mineral magnesium isotope
fractionation in mantle xenoliths from the North China
craton. Earth and Planetary Science Letters, 308,
131–140.
Liu X.-M., Teng F.-Z., Rudnick R.L., McDonough W.F. and
Cummings M. (2014)
Massive magnesium depletion and isotope fractionation in
weathered basalts. Geochimica et Cosmochimica Acta,
135, 336–349.
Pogge von Strandmann P.A.E., Burton K.W., James R.H.,
Van Calstern P., Gislason S.R. and Sigmarsson O.
(2008)
The influence of weathering processes on riverine mag-
nesium isotopes in a basaltic terrain. Earth and Planetary
Science Letters, 276, 187–197.
Pogge von Strandmann P.A.E., Elliott E., Marschall H.R.,
Coath C., Lai Y.-J., Jeffcoate A.B. and Ionov D.A. (2011)
Variations of Li and Mg isotope ratios in bulk chondrites
and mantle xenoliths. Geochimica et Cosmochimica Acta,
75, 5247–5268.
Schiller M., Baker J.A. and Bizzarro M. (2010)
26
Al-26
Mg dating of asteroidal magmatism in the young
Solar System. Geochimica et Cosmochimica Acta, 74,
4844–4864.
Sedaghatpour F., Teng F.-Z., Liu Y., Sears D.W. and
Taylor L.A. (2013)
Magnesium isotopic composition of the Moon. Geochimi-
ca et Cosmochimica Acta, 120, 1–16.
Telus M., Dauphas N., Moynier F., Tissot F.L., Teng F.-Z.,
Nabelek P.I., Craddock P.R. and Groat L.A. (2012)
Iron, zinc and magnesium isotopic fractionation during
continental crust differentiation: The tale from migmatites,
granites and pegmatites. Geochimica et Cosmochimica
Acta, 97, 247–265.
Teng F.-Z. and Yang W. (2014)
Comparison of factors affecting the accuracy of high-
precision magnesium isotope analysis by multi-collector
inductively coupled plasma-mass spectrometry. Rapid
Communications in Mass Spectrometry, 28, 19–24.
Teng F.-Z., Wadhwa M. and Helz R.T. (2007)
Investigation of magnesium isotope fractionation during
basalt differentiation: Implications for a chondritic compo-
sition of the terrestrial mantle. Earth and Planetary Science
Letters, 261, 84–92.
Teng F.-Z., Li W.-Y., Ke S., Marty B., Dauphas N., Huang
S., Wu F.-Y. and Pourmand A. (2010a)
Magnesium isotopic composition of the Earth and chon-
drites. Geochimica et Cosmochimica Acta, 74, 4150–
4166.
Teng F.-Z., Li W.-Y., Rudnick R.L. and Gardner L.R.
(2010b)
Contrasting behavior of lithium and magnesium isotope
fractionation during continental weathering. Earth and
Planetary Science Letters, 300, 63–71.
Teng F.-Z., Dauphas N., Helz R.T., Gao S. and Huang S.
(2011)
Diffusion-driven magnesium and iron isotope fractionation
in Hawaiian olivine. Earth and Planetary Science Letters,
308, 317–324.
Teng F.-Z., Yang W., Rudnick R.L. and Hu Y. (2013)
Heterogeneous magnesium isotopic compostion of the
lower crust: A xenolith perspective. Geochemistry Geo-
physics Geosystems, 14, 3844–3856, doi: 10.1002/
ggge.20238.
Thrane K., Bizzarro M. and Baker J.A. (2006)
Extremely brief formation interval for refractory inclusions
and uniform distribution of 26
Al in the early solar system.
Astrophysical Journal, 646, L159–L162.
Wang G., Lin Y., Liang X., Liu Y., Xie L., Yang Y. and Tu X.
(2011)
Separation of magnesium from meteorites and terrestrial
silicate rocks for high-precision isotopic analysis using
multiple collector-inductively coupled plasma-mass spec-
trometry. Journal of Analytical Atomic Spectrometry, 26,
1878–1886.
Wang S.-J., Teng F.-Z., Williams H.M. and Li S. (2012)
Magnesium isotopic variations in cratonic eclogites: Origins
and implications. Earth and Planetary Science Letters,
359–360, 219–226.
Wang S.-J., Teng F.-Z., Li S. and Hong J. (2014)
Magnesium isotope systematics of mafic rocks during
prograde metamorphism in continental subduction zone.
Geochimica et Cosmochimica Acta. 143, 34–48.
Wiechert U. and Halliday A.N. (2007)
Non-chondritic magnesium and the origins of the inner
terrestrial planets. Earth and Planetary Science Letters, 256,
360–371.
Wombacher F., Eisenhauer A., Heuser A. and Weyer S.
(2009)
Separation of Mg, Ca and Fe from geological reference
materials for stable isotope ratio analyses by MC-ICP-MS
and double-spike TIMS. Journal of Analytical Atomic
Spectrometry, 24, 627–636.
Xiao Y., Teng F.-Z., Zhang H.-F. and Yang W. (2013)
Large magnesium isotope fractionation in peridotite xenolithis
from eastern North China craton: Product of melt-rock interac-
tion. Geochimica et Cosmochimica Acta, 115, 241–261.
Yang W., Teng F.-Z. and Zhang H.-F. (2009)
Chondritic magnesium isotopic composition of the terres-
trial mantle: A case study of peridotite xenoliths from the
North China craton. Earth and Planetary Science Letters,
288, 475–482.
Young E.D. and Galy A. (2004)
The isotope geochemistry and cosmochemistry of magne-
sium. In: Johnson C.M., Beard B.L. and Albarede, F. (eds),
Geochemistry of non-traditional stable isotopes. Reviews
in Mineralogy and Geochemistry. Mineralogical Society
of America (Washington, DC), 197–230.
3 3 9© 2014 The Authors. Geostandards and Geoanalytical Research © 2014 International Association of Geoanalysts