Geochemical Journal, Vol. 34, pp. 11 to 21, 2000
Separation of rare earth elements and strontium from chondritic
meteorites by miniaturized extraction chromatography
for elemental and isotopic analyses
KEIJI MISAWA ,* FUMIE Y AMAZAKI, NAMI IHIRA and N OBORU NAKAMURA
Department of Earth and Planetary Sciences, Faculty of Science, Kobe University, Kobe 657-8501, Japan
(Received April 8, 1999; Accepted July 5, 1999)
A new separation scheme for rare earth elements (REE) and strontium, and its application to REE
determination of chondrites and terrestrial rocks along with isotopic analysis of strontium in chondritic
meteorites are presented. The miniaturized extraction chromatography scheme using RE and Sr Resins
provide simple and effective methods for the isolation of REE and strontium from chondritic meteorites.
Chemical yields of up to 90% of the desired elements and low procedural blanks enable us to use the resins
for small-size chondritic meteorite samples, weighing one to five milligrams. The techniques presented
here are fully applicable to Rb-Sr isotopic systematics and to REE geochemistry and cosmochemistry. A
separation procedure of lead from total rock basalt samples using Sr Resin is also briefly described.
remains in the measured sample. It is well known
that barium poses a potential isobaric interference
on the Eu, LaO, GdO, Yb mass regions (i.e.,
135
Ba 16 O, 137 Ba 16 O, 138 Ba 16 O, 136 Ba 35 Cl,
137
Ba 35 Cl, 138 Ba 35 Cl, 134 Ba 37 Cl, 135 Ba 37 Cl,
136
Ba 37Cl, 137Ba 37Cl) during REE measurements.
Cation exchange chromatography has been commonly used for separation of matrix elements.
Because calcium and strontium possess similar
chemical properties, cation-exchange chromatography is not sufficiently selective in Ca-rich samples for strontium isotope measurement by TIMS.
Birck (1986) proposed a method for the isolation
of strontium from calcium by using DCTA (trans1,2-diaminocyclohexeane-N,N,N′,N′-tetraacetic
acid) together with pyridine. Although the DCTA
method provided high selectivity of strontium as
well as low blanks, the purification of reagents
and separation procedures are complicated.
Horwitz and coworkers developed extraction
chromatographic materials that are highly selec-
INTRODUCTION
Rare earth elements (REE) have been used to
model geochemical processes as well as to constrain high temperature fractionations and redox
conditions in the early solar system (Henderson,
1984). Parent-daughter pairs, 138La-138Ce, 147Sm143
Nd, 146Sm- 142Nd and 176Lu-176 Hf, have been
applied to radiometric age determinations and
petrogenetic tracer studies (e.g., Faure, 1986;
Dickin, 1995). Similarly, the decay of long-lived
radionuclide 87Rb to stable 87Sr has also been applied to radiometric age determinations of rocks
and minerals. In order to obtain stable ion signals
of REE and strontium during thermal ionization
mass spectrometry (TIMS), elements must be
separated from sample matrices and be as clean a
salt as possible. Especially, the isolation of calcium from strontium or REE is essential because
ionization efficiencies of target elements are heavily suppressed if a significant amount of calcium
*Present address: Antarctic Meteorite Research Center, National Institute of Polar Research, Kaga 1-9-10, Tokyo 173-8515,
Japan
11
12
K. Misawa et al.
tive and that sorb strontium or REE (Horwitz et
al., 1991; Dietz and Horwitz, 1993). The aim of
our study is first to scale down procedures previously set up for silicate rock analysis (e.g., Pin et
al., 1994), and then to make the new extraction
chromatography suitable for small-size meteoritic
samples.
EXPERIMENTAL
Reagents and samples
Deionized water was obtained using a Milli-Q
purification system and then by distilling in a
quartz-still. Nitric acid (Wako Pure Chem. Ind.
Ltd., analytical grade) was distilled twice by the
sub-boiling technique in a quartz-still. Hydrochloric acid was obtained by saturation of high-purity
HCl-gas into distilled water. High-purity
perchloric acid and hydrofluoric acid were obtained from National Institute of Standards and
Technology (NIST) and used as received. Chromatographic materials, RE and Sr Resins, were
obtained from Eichrom Industries Inc. Finegrained resins (particle size: 50–100 µ m) were
used for column separation.
A fine-grained sample of the Shaw (L6)
chondritic meteorite, weighing 756.91 mg, was
dissolved using a mixture of concentrated HFHClO4 in a sealed Teflon bomb. The sample solution was evaporated to dryness on a hotplate and
re-dissolved in 2M HCl. This solution was used
as a shelf standard and test sample. After conventional cation exchange chromatography, abundances of potassium, rubidium, strontium, barium,
calcium, and REE of the shelf standard were determined by isotope dilution (ID)-TIMS.
Two chondritic samples, Zaoyang (H5) and
Zhaodong (L4) were analyzed for strontium isotopic compositions. Four ultrabasic metamorphic
rocks, hedenbergite peridotite, orthopyroxenite
and two hedenbergite-eclogites from the LützowHolm Complex, East Antarctica, were analyzed
for REE abundances. Both chondrites and terrestrial rocks were powdered using an agate mortar.
Terrestrial rocks were dissolved using a mixture
of concentrated HF-HClO 4 in sealed Teflon
bombs. Chondrites were dissolved using a mixture of concentrated HF-HNO 3 in PFA Teflon
screw-cap jars.
Rare earth element chemistry
RE Resin is a supported organophosphorous
extractant. The extractant is 1M octyl(phenyl)(N,N-diisobutylcarbamoyl) methylphosphine oxide (CMPO) in tributyl phosphate (TBP), similar
to the extractant used in TRU Resin (Transuranium specific resin: 40% 0.75M CMPO/TBP). The
RE Resin is a 40% loading of 1M CMPO/TBA
onto an inert Amberchrom CG-71ms resin. In the
nitric acid system, the Kd (distribution coefficient)
values for the REE are higher for RE Resin than
for TRU Resin (Huff and Huff, 1993).
Huff and Huff (1993) presented Kd values of
REE and other elements including iron on a RE
Resin column as a function of hydrochloric and
nitric acid concentrations. However, they did not
show Kds for Yb and Lu. Esser et al. (1994) also
reported elution behavior of REE on a RE Resin
column as a function of nitric acid concentration.
These studies show that retention of heavy REE
(HREE) in <4M HNO3 by RE Resin is sensitive
to the nitric acid concentration used to load the
sample solution and wash the column (i.e., loading and washing with <4M HNO3 results in fractional elution). Thus, the calibration was performed in 2M, 2.5M, 3M, and 4M HNO 3 using the
heaviest REE, lutetium.
The RE Resin was washed by 0.1M HCl several times and stored in 0.1M HCl. Standard solution of lutetium (3 µg/ml) was prepared by dissolving high-purity Lu 2O3 in 2.5M HNO3 . The
sample solution containing 1.5 µg of lutetium was
evaporated to dryness and re-dissolved in 2M
HNO3. A 35- µl Teflon column (5 mm × 3 mm ID)
with a polyethylene frit was used in the calibration test. The column was rinsed with distilled
water, and fresh RE Resin was slurried into the
column with 0.1M HCl. One free column volume
(FCV) of RE Resin is 65–75% of the bed volume.
The column was first washed five times with 10
FCV of 0.1M HCl, and then was conditioned with
10 FCV of 2M HNO3. The standard, dissolved with
Separation of REE and Sr from chondrites
200 µl of 2M HNO3, was loaded onto the resin.
Then 50 µl of 2M HNO 3 was added 25 times and
each elution was collected in a Teflon vial. Triethanolamine (TEA) of 1M or 2M solution was
titrated into each fraction to reach a pH of ~7.0
(Shirey et al., 1987). Eriochrom Black T (EBT)
stock solution was prepared by EBT, methanol,
and 1M TEA. One drop of EBT working solution
diluted from the stock solution was added to the
aliquot off the column. The EBT will be blue if
no lutetium is present, and purple or pink depending on the amount of lutetium in the vial. The 2.5M
HNO3, 3M HNO 3, and 4M HNO 3 calibrations proceeded analogously. The results are shown in Fig.
1.
On the basis of the pilot calibration, we prepared an 80-µl column using polyethylene pipette
tip (Quality Science Plastics) with quartz wool at
the bottom (Fig. 2). Resin was first washed with
15 FCV of 6M HCl, and then conditioned with 15
FCV of 4M HNO 3. The sample solution of Shaw
was evaporated to dryness, dissolved in 50 µl of
4M HNO3 and loaded onto the resin. The column
was rinsed with 350 µl of 4M HNO3. The REE
were eluted with 400 µl of 4M HCl. The composite REE spike solution was added, and REE abundances were determined by ID-TIMS. The results
are summarized in Table 1.
The REE recovery rate using this scheme was
very poor, less than 20% for light REE (LREE),
and less than 5% for HREE. As expected from the
acid dependency of its capacity factor, the sorption of Fe(III) rises steadily with increasing nitric
acid concentration (Horwitz et al., 1993; Huff and
Huff, 1993). Thus, it is considered that large
amounts of Fe(III) were present and that this ion
had a significant negative effect on the sorption
of REE as already observed by Pin and Santos
Zalduegui (1997) for TRU Resin. In order to remove iron, pretreatment using a 150- µl anion exchange column (Dowex AG1-X8, 200–400 mesh)
with 6M HCl medium was carried out. When most
of the iron was removed prior to extraction chromatography, chemical yields of REE on RE Resin
chemistry increased to ~60%.
13
Fig. 1. Elution behavior of lutetium through RE Resin
column as a function of nitric acid concentration.
Breakthrough points of lutetium become larger with
increasing concentrations of nitric acid. One free column volume is ~25 µ l.
Finally, we designed a two-stage, single column separation scheme as shown in Fig. 2. The
RE Resin column was first washed with 15 FCV
of 0.1M HCl, and then conditioned with 15 FCV
of 4M HCl. The sample, dissolved in 100 µl of
4M HCl, was loaded onto the resin. The REE were
eluted with 750 µl of 4M HCl, and the effluent
was collected in a Teflon vial. Since the Kd of iron
14
K. Misawa et al.
Table 1. Chemical yields of REE by RE Resin chemistry(a)
1st stage
Pretreatment(b)
2nd stage
Loading ( µ l)
Rinse
( µ l)
Strip
( µ l)
None
4M HNO3 (50)
4M HNO3 (350)
4M HCl (400)
Chemical yields (%)
La
Ce
Nd
Sm
Eu
Gd
Dy
Er
Yb
Lu
n.d.(f)
16
18
23
18
n.d.
n.d.
3
n.d.
1
AG-1 (6M HCl) (c)
7M HNO3 (50)
4M HNO3 (500)
4M HCl (750)
n.d.
60
58
59
60
n.d.
n.d.
n.d.
n.d.
59
RE Resin (6M HCl) (d)
4M HNO3 (50)
4M HNO3 (400)
4M HCl (500)
n.d.
2
1
n.d.
4
30
83
94
86
93
RE Resin (4M HCl) (e)
4M HNO3 (50)
4M HNO3 (400)
4M HCl (500) (g)
89
92
90
92
92
93
92
93
91
90
The 80- µl polyethylene pipette tip column was used (see Fig. 2). Sample solution of the shelf standard (the Shaw chondrite)
was loaded onto the column. REE concentrations (ppm) in the Shaw chondrite: La = 0.425, Ce = 1.05, Nd = 0.799, Sm = 0.251,
Eu = 0.0987, Gd = 0.342, Dy = 0.421, Er = 0.276, Yb = 0.274, Lu = 0.0413.
(b)
Pretreatment chemistry to remove iron.
(c)
The sample, dissolved in 100 µl of 6M HCl, was loaded onto the 150-µl anion exchange column (AG1 X-8, 200–400 mesh) to
retain Fe(III).
(d)
The sample, dissolved in 100 µl of 6M HCl, was loaded onto the RE Resin column. Fe(III) and LREE, but not HREE, were
retained by the RE Resin at this stage. The REE were eluted with 750 µ l of 6M HCl.
(e)
The sample, dissolved in 100 µl of 4M HCl, was loaded onto the RE Resin column. Fe(III) and REE were retained by the RE
Resin at this stage. The REE were eluted with 750 µl of 4M HCl (see Fig. 2).
(f)
“n.d.” denotes not determined.
(g)
Procedural blanks using the RE Resin column, including a pretreatment to remove Fe(III), were determined by ID-TIMS:
La = 2.2, Ce = 4.9, Nd = 2.8, Eu = 0.16 (values are in pg).
(a)
is high (>400) and the Kd values of REE are quite
low, less than 10 (Huff and Huff, 1993), Fe(III),
but not REE, will be retained by the resin at this
stage. The column effluent was evaporated to dryness, and the residue was taken up in 50 µl of 4M
HNO3. The column was first washed out using 30
FCV of 0.1M HCl-0.3M HF, then using 15 FCV
of 0.1M HCl, and finally conditioned with 15 FCV
of 4M HNO 3 . Then sample solution was loaded
onto the column, and the column was rinsed with
400 µl of 4M HNO3 . The REE were finally eluted
with 500 µl of 4M HCl.
Rare earth elements were separated from
ultrabasic metamorphic rocks using the RE column. A powdered sample weighing about 60 mg
was dissolved. After complete dissolution was
achieved, the solution was split into weighed
aliquots. Composite REE tracer solution was
added to an aliquot of the sample solution for IDTIMS (Table 2). In basic samples containing up
to 15% of iron (68032701, 68091201-1, and
68021509), Fe(III) was not completely sorbed in
the resin (80 µl) and was still present in the REE
fraction. When a yellow-colored band, presumably
Fe(III), was not retained by the resin during the
4M HCl elution, a cycle of 4M HCl elution and
0.1M HCl-0.3M HF wash out was repeated.
Strontium chemistry
Sr Resin consists of a 1M solution of bis
4,4′(5′)bis (tert-butylcyclohexano)18-crown-6 in
1-octanol sorbed on an inert Amberchrom
Separation of REE and Sr from chondrites
15
Fig. 2. Two-stage single column method for the isolation of REE from chondrites and terrestrial rocks. At the first
stage iron was adsorbed on the RE Resin, separated from REE using 4M HCl, and washed out by a mixture of
0.1M HCl-0.3M HF. Then the iron-free sample was evaporated to dryness, taken up in 4M HNO 3, and loaded onto
the column after conditioning. The column was rinsed with 4M HNO 3, and the REE were finally eluted with 4M
HCl.
Table 2. Rare earth element data obtained by ID-TIMS and the proposed separation scheme for
metamorphic rocks from the Lützow-Holm Complex, East Antarctica
Sample
84011307J
hb*-peridotite
dissolved (mg)
La (ppm)
Ce
Nd
Sm
Eu
Gd
Dy
Er
Yb
Lu
68.06
12.4
18.3
4.29
1.13
0.456
1.63
1.96
1.18
1.02
0.143
(2.3)**
(2.3)
(1.1)
(1.4)
(2.0)
(1.9)
(2.7)
(2.4)
(2.9)
(2.2)
*“hb” denotes hornblende.
**Values in parenthesis are 2 σ% errors.
68032701
orthopyroxenite
60.74
2.11
3.99
1.61
0.299
0.0356
0.295
0.380
0.300
0.408
0.0686
(0.71)
(0.70)
(0.62)
(1.9)
(2.4)
(0.64)
(2.0)
(2.1)
(2.5)
(2.0)
68091201-1
hb-eclogite
64.30
2.57
3.51
1.44
0.528
0.814
0.881
0.996
0.530
0.468
0.0671
(1.1)
(0.80)
(0.97)
(2.8)
(1.7)
(1.1)
(1.8)
(3.0)
(3.4)
(3.3)
68021509
hb-eclogite
59.11
13.6
27.3
14.4
2.65
1.28
2.26
1.57
0.768
0.672
0.102
(3.1)
(2.2)
(2.0)
(4.9)
(1.9)
(1.2)
(1.2)
(2.0)
(3.6)
(1.7)
16
K. Misawa et al.
Table 3. Chemical yields of strontium and calcium, and strontium
blanks through Sr Resin column chemistry*
Resin volume
( µ l)
Sample weight
(mg)
Sr yield
(%)
Ca yield
(%)
Sr blank
(pg)
150
9.22
98
not determined
80
10
2.19
3.66
4.97
97
54
70
0.26
0.62
0.10
10
30
1.39
87
0.14
33
50
4.33
5.66
98**
93***
0.42
0.038
36
*The Shaw (L6) chondrite was used as a shelf standard. Abundances of alkali and alkaline earth elements in the shelf standard
are: K = 857 (ppm), Rb = 3.56 (ppm), Sr = 12.6 (ppm), Ba = 4.83 (ppm), Ca = 1.48 (%).
**Rinsed with 500 µl of 2M HNO3 , 600 µl of 7M HNO3, and 100 µl of 3M HNO3 .
***Rinsed with 600 µl of 2M HNO 3, 600 µl of 7M HNO3 , and 100 µ l of 3M HNO3 . Chemical yield of Ba (1.5%) was determined
by ID-TIMS with 136Ba-enriched tracer.
CG-71ms resin (Horwitz et al., 1991). The extraction chromatography using this resin provides a
simple and effective method for the separation of
strontium using a nitric acid medium. One FCV
of Sr Resin is ~65% of the bed volume. The Sr
Resin was washed several times with warm
(~50°C) 0.05M HNO 3, and stored in distilled water (column capacity factor for Sr decreases ~1/3
when warm nitric acid are used: M. J. Fern, personal communication, 1994).
In preliminary experiments designed to determine the appropriate conditions for the retention
of strontium from chondritic meteorites using the
Sr Resin with nitric acid, a 150-µl Teflon column
with a polyethylene frit was employed. Using relatively large amounts of sample (~10 mg of the
Shaw chondrite) we performed strontium chromatography. Chemical yield and procedural blank of
strontium chromatography were obtained by IDTIMS (Table 3). In order to apply this method to
additional small chondritic samples containing
less than 50 ng of strontium, the strontium chromatography was miniaturized.
We prepared three types of polyethylene pipette tip columns (10-, 30-, and 50-µl resin) with
quartz wool at the bottom (Fig. 3). The column
was first washed with 25 FCV of warm (~50°C)
0.05M HNO 3, and then conditioned with 10 FCV
of 2M HNO 3. A sample solution of 50 µl of 2M
HNO3 was loaded onto the column. The 10- µl
column was rinsed with 450 µ l of 2M HNO 3 .
Strontium was stripped using 500 µl of warm
0.05M HNO3 . The Kd values of both strontium and
barium increased with increasing nitric acid concentration. When nitric acid concentration exceeds
~3M, the Kd of barium decreases with increasing
acid concentration (Horwitz et al., 1992). The 30µl column was rinsed with 360 µl of 2M HNO3 ,
360 µl of 7M HNO3, and 60 µl of 3M HNO3 . The
strontium fraction was stripped using 300 µl of
warm 0.05M HNO3. The 50-µl column was rinsed
with 600 µl of 2M HNO 3, 600 µl of 7M HNO3 ,
and 150 µ l of 3M HNO3. The strontium fraction
was stripped using 400 µl of warm 0.05M HNO3 ,
and split into two weighed aliquots. Chemical
yields of strontium and calcium were determined
by ID-TIMS with 84Sr- and 44Ca-enriched tracers.
The other was evaporated to dryness and used for
strontium isotope analysis. For spiked runs,
highly-enriched 84Sr spike solution (84Sr/86Sr =
401.4) was added to the Shaw sample before extraction chromatography, and both isotopic compositions and concentrations of strontium were
measured simultaneously. Results are summarized
in Table 4.
Separation of REE and Sr from chondrites
17
Fig. 3. Scheme for the separation of strontium from chondrites. Strontium was effectively separated from barium
by an additional 7M HNO3-3M HNO3 rinse, following a 2M HNO3 rinse.
Table 4. Strontium isotope results for chondritic meteorites
Sample
dissolved (mg)
Separated(a)
(mg)
87
Sr/8 6 Sr (b )
Sr
(ppm) (c)
Shaw (L6) shelf std.
756.91
4.33
5.66
3.98
3.86
0.753679 ± 13(d )
0.753668 ± 13
0.753672 ± 8
0.753667 ±14(e)
12.57
Zaoyang (H5)
88.41
4.62
4.65
4.62
0.743118 ± 25
0.743106 ± 20
0.743111 ± 16
11.48
Zhaodong (L4)
90.46
4.80
4.81
0.776499 ± 21
0.776502 ± 14
NIST SRM 987 (n = 16)
9.987
0.710264 ± 26(f)
Passed through the 50-µ l polyethylene column (see Fig. 3).
Normalized to 86Sr/88Sr = 0.1194.
(c)
About 1 mg aliquots were passed through the 30-µl polyethylene column.
(d)
Uncertainties apply to last digits and are 2σ m.
(e)
Spiked run. Contributions of 84Sr spike were corrected.
(f)
Uncertainty applies to last digits and is 2 σp .
(a)
(b)
18
K. Misawa et al.
Lead chemistry
Because of its high selectivity, the Sr Resin is
useful to separate not only strontium from alkaline earth elements but also lead from iron-rich
matrices in an HCl medium (Gale, 1996; Vajda et
al., 1997). We performed miniaturized extraction
chromatography to separate lead from terrestrial
basalt, BCR-1, using a 35-µl Teflon column. Preliminary results are presented in Appendix.
Instrumentation and measurements
Elemental abundances were determined by IDTIMS using a JEOL 05RB mass spectrometer
equipped with a secondary electron multiplier
(Nakamura, 1974). The REE fraction was dissolved in ~1–2 µl of concentrated HNO 3 and
loaded on a previously outgassed Re-single or ReTa-double filament. Isotopic compositions of
strontium as well as concentrations of strontium
in chondritic meteorites were determined by using a Finnigan MAT 262 mass spectrometer
equipped with five Faraday collectors and a secondary electron multiplier. The strontium fraction
was dissolved in a few µl of concentrated HNO3,
and an amount corresponding to 10–25 ng of strontium was loaded with Ta-activator (Birck, 1986)
onto a previously outgassed W-single filament.
Data acquisition was in the static five collector
mode with 88Sr ion beam intensities between 1 and
1.5 × 10–11 A. Details of strontium isotope measurements on a Finnigan MAT 262 mass
spectrometer will appear elsewhere.
RESULTS AND DISCUSSION
RE Resin chemistry
Pin et al. (1994) and Pin and Santos Zalduegui
(1997) presented chemical separation schemes for
LREE using TRU Resin. Since retention of HREE
in both TRU and RE Resins is very sensitive to
nitric acid concentration, it is hard to separate all
REE from the matrix elements by these methods.
Moreover, chondritic meteorites usually contain
large amounts of iron, more than 15%. Huff and
Huff (1993) proposed another separation scheme
in which the sample solution was loaded onto the
TRU Resin column in 1M HNO3 , and REE were
eluted by 4M HCl. Iron is not strongly sorbed on
the TRU and RE Resins at 1M HNO 3 (Kd = 5).
Passage of 5 FCV of 1M HNO 3 through a TRU
Resin column onto which a sample had been
loaded in 2M HNO 3 -0.5M Al(NO 3 ) 3 removed
~97% of the Fe(III) initially introduced (Horwitz
et al., 1990). The Kd values of LREE on TRU and
RE Resin columns in a nitric acid medium are
more than 200 but those of HREE are less than
20. Thus, the REE separation chemistry proposed
by Huff and Huff (1993) cannot be applied to ironrich samples, such as ultramafic rocks and
chondritic meteorites. Horwitz et al. (1993) concluded that if Fe(III) was reduced to Fe(II) with
0.3M ascorbic acid, its negative effect on sorption of trivalent actinide, americium, becomes
negligible. By analogy ascorbic acid will overcome the problem of loss of trivalent REE. However, Pin and Santos Zalduegui (1997) described
that the amount of ascorbic acid required to reduce iron to its divalent state for mafic rock samples would be excessive.
Our method shown above is simple and does
not use a cation or anion resin column, and effectively removes Fe(III) without any loss of REE.
Because barium is not sorbed on the RE Resin in
4M HNO3, the troublesome element barium is also
effectively removed from the REE fraction. During the measurements of REE-oxide or -metal ions
by ID-TIMS, no detectable interference of barium
or its molecular peaks with LaO, Eu, GdO, or Yb
was observed. The CI-chondrite normalized REE
abundance patterns of ultrabasic metamorphic
rocks from Antarctica are shown in Fig. 4. All
show LREE-enriched patterns (La/Lu ratios of 2–
15) with or without europium anomaly, and HREE
abundances are relatively constant (1–8 ×
chondrites).
Chemical yields are more than 90% for all REE
measured in the chondritic sample (Table 1), and
these values are acceptable for not only meteoritic
samples but also terrestrial samples with low REE
abundances. The blanks using the RE Resin column, including a pretreatment to remove Fe(III),
were obtained for LREE. They are less than 1%
Separation of REE and Sr from chondrites
19
for 1 mg-sized chondritic samples, thus negligible for our present study.
If an additional elution step with 2-methylactic
acid (Eugster et al., 1970) or reverse-phase chromatography using di-(2-ethylhexyl)-phosphoric
acid (HDEHP) coated on an inert resin (Richard
et al., 1976) is employed, the purification of lanthanum, cerium, neodymium, samarium, and lutetium for isotopic analyses on TIMS and ICP-MS
can be achieved.
chondritic materials (1000–1500 ppm). If 20 mgsized chondritic sample is dissolved by 100 µl of
2M HNO3, about 0.01M of potassium is present
in the sample solution. Because the effects of high
concentration of potassium on strontium retention
by the Sr Resin occur when potassium is around
0.01M, it is important to take into account the
potassium concentration of the sample solution.
The 30-µl column is suitable for the determination of strontium concentrations. The effluent
from loading and 2M HNO3 rinse fractions will
contain rubidium, which may be separated from
other elements by conventional cation exchange
chromatography. If the sample size is large enough
and the blank level negligible, a pre-separation of
major elements by conventional cation exchange
chemistry with 2M HCl, and/or additional strontium chromatography will help to obtain highpurity strontium salts without contamination by
other alkaline earth elements.
We confirm that lead is separated effectively
from sample matrix elements including iron and
bismuth when using Sr Resin in an HCl medium
(see Appendix). During isotopic measurements,
ionization efficiency of lead was not suppressed
by iron even if only a single column chemistry
has been carried out. Thus, preliminary results
suggest that Sr Resin is applicable to U-Th-Pb isotopic systematics of silicate rocks and minerals,
and to lead isotope geochemistry.
Sr Resin chemistry
A preliminary experiment using a 150- µl column showed that the chemical yield of strontium
during extraction chromatography was sufficiently
high, but that the procedural blank was not acceptable for isotopic analysis of meteoritic samples. The procedural blank was reduced by decreasing the resin volume. More than 98% of the
barium was removed from the strontium fraction
when successive 2M-7M-3M HMO3 rinses were
employed as described by Pin et al. (1994).
Certain elements (e.g., potassium and lead) do
adversely affect strontium retention onto the Sr
Resin (Dietz and Horwitz, 1993). Potassium is
encountered in appreciable quantities in typical
Memory effects
Both RE and Sr Resins are so selective that
reusing resins has not been recommended (M. J.
Fern, personal communication, 1994; Pin and
Bassin, 1992; Pin et al., 1994; Pin and Santos
Zalduegui, 1997). After passing a chondritic sample containing 110 ng of strontium through the
column of freshly packed Sr Resin, we found an
extremely high “carry-over” of strontium (1.4 ng)
from one sample to the next. The “carry-over” may
affect the measurement of subsequent samples if
the amount of the desired element is quite low.
We confirm that the resin should not be reused in
this type of application and should be discarded
after each single use.
Fig. 4. REE data obtained by ID-TIMS using the
proposed procedure, presented as CI-chondrite
normalized REE patterns for four ultramafic metamorphic rocks from Antarctica. Errors are within symbol
size.
20
K. Misawa et al.
SUMMARY
The results of this study clearly demonstrate
that miniaturized extraction chromatography using RE and Sr Resins provides simple and effective methods for the isolation of the REE and
strontium from chondritic meteorites. Chemical
yields of up to 90% of the desired elements and
low procedural blanks enable us to use the resins
for small-size chondritic meteorite samples,
weighing one to five milligrams. The separation
schemes presented here are fully applicable to RbSr isotopic systems and to determination of REE
abundances in chondrites by the ID-TIMS methods, and are useful for pre-concentrating lanthanum, cerium, neodymium, samarium, and lutetium
for studies of La-Ce, Sm-Nd, Lu-Hf isotopic systematics. Extraction chromatography using Sr
Resin is also applicable to U-Th-Pb isotopic systematics of silicate rocks and minerals, and to lead
isotope geochemistry.
Acknowledgments—A part of this work has been done
at the lead-free laboratory, U.S. Geological Survey in
Denver. We thank the late Dr. Mitsunobu Tatsumoto
for his continuous encouragement and discussions. We
also thank N. Morikawa for his assistance in the laboratory work at Kobe University, S. Hayashi and N.
Shinbo for their help on extraction chromatography at
the first stage. Terrestrial rock samples were provided
by K. Shiraishi. The chondrite samples, Zaoyang and
Zhaodong, were provided by Z. Ouyang. We thank
Wayne Premo, Todd Hinkley, and Mark Fanning for
correcting and refining the English. Constructive reviews of Christian Pin and Hiroshi Hidaka are appreciated.
REFERENCES
Birck, J. L. (1986) Precision K-Rb-Sr isotopic analysis: Application to Rb-Sr chronology. Chem. Geol.
56, 73–83.
Dickin, A. P. (1995) Radiogenic Isotope Geology. Cambridge Univ. Press, Cambridge, 452 pp.
Dietz, M. L. and Horwitz, E. P. (1993) Novel chromatographic materials based on nuclear waste processing chemistry. LC GC 11, 424–436.
Esser, B. K., Volpe, A., Kenneally, J. M. and Smith, D.
K. (1994) Preconcentration and purification of rare
earth elements in natural waters using silica-immo-
bilized 8-hydroxyquinoline and a supported organophosphorus extractant. Anal. Chem. 66, 1736–1742.
Eugster, O., Tera, F., Burnett, D. S. and Wasserburg,
G. J. (1970) Isotopic composition of gadolinium and
neutron-capture effects in some meteorites. J.
Geophys. Res. 75, 2753–2768.
Faure, G. (1986) Principle of Isotope Geology. 2nd
edition, John Wiley & Sons, New York, 589 pp.
Gale, N. H. (1996) A new method for extracting and
purifying lead from difficult matrices for isotopic
analysis. Anal. Chim. Acta 332, 15–22.
Henderson, P. (1984) Rare Earth Element
Geochemistry. Elsevier, New York, 510 pp.
Horwitz, E. P., Dietz, M. L., Nelson, D. M., LaRosa, J.
J. and Fairman, W. D. (1990) Concentration and separation of actinides from urine using a supported
bifunctional organophosphorus extractant. Anal.
Chim. Acta 238, 263–271.
Horwitz, E. P., Dietz, M. L. and Fisher, D. E. (1991)
Separation and preconcentration of strontium from
biological, environmental and nuclear waste samples
by extraction chromatography using a crown ether.
Anal. Chem. 63, 522–525.
Horwitz, E. P., Dietz, M. L. and Chiarizia, R. (1992)
The application of novel extraction chromatographic
materials to the characterization of radioactive solutions. J. Radioanal. Nucl. Chem. 161, 575–583.
Horwitz, E. P., Chiarizia, R., Dietz, M. L., Diamond,
H. and Nelson, D. M. (1993) Separation and
preconcentration of actinides from acidic media by
extraction chromatography. Anal. Chim. Acta 281,
361–372.
Huff, E. A. and Huff, D. R. (1993) TRU. Spec and RE.
Spec chromatography: Basic studies and applications.
34th ORNL/DOE Conf. on Anal. Chem. in Energy
Technology, Gatlinburg, TN.
Nakamura, N. (1974) Determination of REE, Ba, Fe,
Mg, Na and K in carbonaceous and ordinary
chondrites. Geochim. Cosmochim. Acta 38, 757–775.
Pin, C. and Bassin, C. (1992) Evaluation of a strontium-specific extraction chromatographic method for
isotopic analysis in geological materials. Anal. Chim.
Acta 269, 249–255.
Pin, C. and Santos Zalduegui, J. F. (1997) Sequential
separation of light-rare-earth elements, thorium and
uranium by miniaturized extraction chromatography:
Application to isotopic analyses of silicate rocks.
Anal. Chim. Acta 339, 79–89.
Pin, C., Briot, D., Bassin, C. and Poitrasson, F. (1994)
Concomitant separation of strontium and samariumneodymium for isotopic analysis in silicate samples,
based on specific extraction chromatography. Anal.
Chim. Acta 298, 209–217.
Richard, P., Shimizu, N. and Allègre, C. J. (1976) 143Nd/
Separation of REE and Sr from chondrites
146
Nd, a natural tracer: an application to oceanic
basalts. Earth Planet. Sci. Lett. 31, 269–278.
Shirey, S. B., Banner, J. L. and Hanson, G. N. (1987)
Cation-exchange column calibration for Sr and the
REE by EDTA titration. Chem. Geol. 65, 183–187.
Todt, W., Cliff, R. A., Hanser, A. and Hofmann, A. W.
(1996) Evaluation of a 202 Pb-205 Pb double spike for
high-precision lead isotope analysis. Earth Processes: Reading the Isotopic Code (Basu, A. and Hart,
S. R., eds.), AGU Geophysical Monograph 95, 429–
437.
Vajda, N., LaRosa, J., Zeisler, R., Danesi, P. and KisBenedek, Gy. (1997) A novel technique for the simultaneous determination of 210 Pb and 210 Po using a
crown ether. J. Environ. Radioactivity 37, 335–372.
APPENDIX
The 35- µl Sr Resin column (made of heatshrinkable FEP Teflon tube, 5 mm × 3 mm ID)
with a polyethylene frit was first washed with 2
ml of 6M HCl, and then conditioned with 250 µl
of 1M HCl. Powdered samples were dissolved
using a mixture of concentrated HNO 3-HF in PFA
Teflon screw-cap jars. After complete dissolution
was achieved, the samples were dried on a hotplate, re-dissolved in 1 ml of 0.5M HBr, and centrifuged. The supernatant was evaporated to dry-
21
ness, and the residue was taken up in 250 µl of
1M HCl. The sample solution was loaded onto the
column, and the column was rinsed with 900 µl
of 1M HCl. The lead was finally stripped using
450 µl of 6M HCl, because the Kd value of lead
decreases with increasing hydrochloric acid concentration (Vajda et al., 1997). Lead blanks were
47 pg for single column chemistry, and 180 pg for
total procedure including sample loading onto the
filament. Chemical yield of lead was up to 98%.
Isotopic compositions of lead were measured
on Finnigan MAT 262 mass spectrometers
equipped with multi-Faraday collectors and ion
counting systems. The lead fraction was loaded
onto an outgassed Re single-filament with silicagel and phosphoric acid. The raw data were corrected for analytical blank, instrumental mass
fractionation of 1.01 ± 0.29‰ per mass unit (2σ)
based on forty analyses of NIST standard SRM
981 (Todt et al., 1996) during the course of this
study, and spike contribution, if 205Pb spike was
used. Isotopic compositions of lead of BCR-1 are:
206
Pb/ 204 Pb = 18.832 ± 0.004, 207 Pb/ 204 Pb =
15.647 ± 0.004, 208Pb/204Pb = 38.763 ± 0.011 (errors are 95% confidence limits).