FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
Application of K-Ar dating system to be performed by new noble gas mass
spectrometry and its calibration from standard air analysis
Keiko Sato 1, Hajimu Tamura 1, Hidenori Kumagai 2 and Takeshi Hanyu 3
1 Research
Program for Data and Sample Analyses, Institute for Research on Earth Evolution (IFREE)
for Deep Sea Research, Institute for Research on Earth Evolution (IFREE)
3 Research Program for Geochemical Evolution, Institute for Research on Earth Evolution (IFREE)
2 Program
of the principle of this technique is as follows. K-Ar dating is
based on the fact that 40K decays to 40Ar or 40Ca with a half life
of 1.25×109 years [Steiger and Jäger, 1977]. K-Ar age can be calculated by the equation.
1. Introduction
K-Ar dating has been widely used for determination of the
radiometric ages of geological events [e.g., Dalrymple, 1968]. The
reported applicable chronological range of the youngest age
reached to about 10,000 years [e.g., Dalrymple and Lanphere,
1969]. This implies that K-Ar dating technique is very suitable to
make clear the Mesozoic-Quaternary geological events, such as
volcanic and magmatic activities of volcanoes. However, it has
been suggested by several authors that Ar isotope ratios of the
volcanic rocks in recent ages mostly fell on the mass fractionation
line from the atmospheric isotope ratio [Krummenacher, 1970;
Kaneoka, 1980; Matsumoto et al., 1989a]. This means that the
fractionation influences the K-Ar dating of very young samples of
Quaternary ages. Takaoka et al. [1989] suggested that K-Ar dating technique by peak height comparison method (sensitivity
method) is suitable for the very young samples because it is able
to correct the initial Ar isotope ratio assuming mass fractionation
from the atmospheric Ar isotope ratio. Matsumoto et al. [1989b,
1995] performed the method of Takaoka et al. [1989], based on
highly precise 38Ar/36Ar measurements and demonstrated that KAr age of 0.05 Ma of the ideal basaltic rock could be determined
within 15% error (1σ) by K-Ar method for noble gas mass spectrometer. This implies that K-Ar dating method by this “mass
fractionation correction procedure” is very useful to date the very
younger than 1 Ma samples [Takaoka, 1989].
A new Ar analytical system was introduced to JAMASTEC in
2003. The system consists of GV Instruments 5400He mass spectrometer and the sample handling system both for 1st extraction
line with Ta furnace, 2nd purification with correcting Ar gas and
3rd purification of sample noble gases [Tamura, 2005]. In the system, the peak comparison method (sensitivity method) without
38Ar spike was adopted to be dating the Cenozoic geological
events. It has been calibrated by the international biotite mineral
standard of HD-B1, Bern 4B, Bern 4M, Sori 93, which calibrated
by using the radiogenic 40Ar concentration and argon isotope
ratios. The calibration has been examined by analyses of several
of K-Ar age standard minerals and K-Ar age known samples. The
manuscript of Tamura [2005] part is reported that the description
of the noble gas analytical system and we would like to summarize the results of calibration by STD Air (AS) samples in this
manuscript, mainly.
 40 Ar
( rad )
ln 
t=
40

(λ e + λ β )
K

1
 λe + λ β

 λe

 
 + 1
 
 
t: K-Ar age
40
40
-10
λe : Decay constant from g K to Ar [0.581×10 /y]
λβ : Decay constant from
40
40
K to 40Ca [4.962×10-10/y]
Ar [rad]: Concentration of radiogenic 40Ar [mol/g]
It is assumed that the present atomic abundance ratio of 40K to
total potassium in terrestrial rocks and minerals is constant to be
40K/K=1.167×10-4 [Steiger and Jäger, 1977]. Then, the concentration of 40K can be calculated from the total potassium concentration. The radiogenic 40Ar is determined by the sensitivity method.
The concentration of the total 40Ar in a sample is obtained by the
comparison of peak intensity with the known amount of the standard air sample.
The initial 40Ar/36Ar ratio in a sample is assumed to be equal to
that of the atmospheric value in the conventional K-Ar dating.
However, because Ar isotopic ratio of the volcanic rocks in recent
ages mostly fell on the mass fractionation line from the atmospheric ratio, 40Ar/36Ar ratio is estimated by assuming that initial
40Ar/36Ar ratios in all volcanic rocks lie on the fractionation line
from the atmospheric ratio, which follow the procedure of
Takaoka et al. [1989] and Matsumoto et al. [1989b]. The calculation of the uncertainty for age determinations by the sensitivity
method is summarized in Matsumoto et al. [1989b].
2.2 Argon analytical system
The Ar analytical system mainly consists of GV Instruments
5400He mass spectrometer and Ta furnace made by Horiguchi
Iron Works and “the main sample handling system for extraction
and purification of sample gases” of parts made by Horiguchi Iron
Works designed and constructed by us [Tamura, 2005]. Most of
them are made of stainless tubes.
GV Instruments 5400He mass spectrometer has 3 detectors a
high performance and precision magnetic sector mass spectrometer for the analysis of the ultra small amount of noble gases
[Tamura, 2005]. It employs the static vacuum mode of operation
(volume is 1999 cm3 and volume of extension parts is almost
2. Argon analytical system and analytical procedure
2.1 Principle of K-Ar dating
K-Ar dating technique has been summarized or explained by
many authors [e.g., Dalrymple and Lanphere, 1969]. The summary
1
FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
400cm3). The ion source is NIER type GV Instrument Bright
source. The collector assembly is a Faraday collector, and two
high sensitivity collectors, Daly Scintillator and Electron
Multiplier. The 5400He with an axial collector resolution of >600
allows the 3He+ peak to be easily separated from the interference
peaks of HD+ and H3+. It is therefore suitable for the analysis of
He and all the Noble gases. The vacuum inside the mass spectrometer is normally attained to the order of 10-10 torr by pumping
with a Varian 40 l/s ion pump and two NP10 getter pumps at
room temperature. The sample analyses are done by peak jumping
procedure controlled by a personal computer.
The arrangement of the handling sample system and the mass
spectrometer is illustrated in Tamura [2005]. The main sample
handling system is subdivided into four sections: (V1): gas extractions, (V2), (V3): gas purification and (Vstd): the standard air
pipette system. The (V1) extraction of sample gases done with a
tantalum resistance heater by heating samples in a molybdenum
crucible and a tantalum tube in the extraction line (V1). The temperature of the heater is manual controlled by current and voltage.
The inside of the extraction line and the tantalum heater room are
pumped out with Ti-Zr getters, 150 L/s turbo molecular pumps
and rotary pumps, respectively. (V3) The sample gas is purified
by Ti-Zr getter and NP10 getter pumps and pumped out by an ion
pump (40 L/s) or 150 L/s turbo molecular pumps and rotary
pumps, respectively. Getters conditions are that Ti-Zr getters heated almost 800 ºC and a NP10 Zr-Al getter kept at room temperature (activation time is only heated), respectively.
The charcoal trap is also equipped for collecting purified Ar
gas in (V2) and (V3). The inside of the purification line is normally pumped out with a turbo molecular pump or an ion pump. The
vacuum is attained to the order of 10-10 torr in the good conditions. (Vstd) The standard air pipette system consists of the reservoir (1500 and 3000 cm3) of standard air Ar and the pipette (1.4
cm3) with two pipette valves. In this work, the one of the standard
air reservoir (AS) of 1500 cm3 and pipette system was used. The
procedure of making STD air is summarized in Tamura [2005].
tion of mass spectrometer is shown in Table 1.
The relation between the magnetic fields and each Ar isotopic
mass is initially calibrated by a personal computer using mass calibration program of GV Instrument software. The calibration is
sometimes renewed to remove deviations accumulated for a long
term. The analyses and calculations of Ar contents and isotopic
ratios are also done using the program of original isotope calculation software programmed by one of us, Tamura. Three isotopes
of 36Ar (Daly and Faraday), 38Ar (Daly and Faraday) and 40Ar
(Faraday) are analyzed that the each isotope is analysed 21 cyclic
with jumping on the centered flat top peaks. The 40Ar contents are
mostly regressed by the curve fit and are extrapolated to the zero
time when Ar gas is introduced to the mass spectrometer. The
40Ar/36Ar and 38Ar/36Ar ratios during the analysis are calculated
by the equation of Dodson [1978].
3. Calibration of the system
3.1 Analyses of standard air Ar
In the K-Ar dating using sensitivity method the intensity of
radiogenic 40Ar of the unknown sample is compared with the 40Ar
intensity of the standard air Ar sample usually analysed in the
same day, and then the concentration of radiogenic 40Ar is consequently determined. The fundamental assumption is that the sensitivities are almost same between the analyses of an unknown sample and a standard air Ar sample.
Figure 2 indicates the variation of the 40Ar intensity, 40Ar/36Ar
and 38Ar/36Ar of the results of 100 tunes standard air analyses during six months. The fundamental assumption of sensitivity
method is considered to be satisfied. Even if theoretical decrease
of 40Ar is accepted, it seems that the sensitivity is almost stable.
The depletion rate of the standard air Ar will be discussed in the
next section, 3.2. During 6 months (Fig. 2) averaged 40Ar/36Ar
ratio and the standard deviations are 294.7±2.7 [Faraday
/Faraday], 2.663±0.339 (Faraday/Daly) and the averaged
38Ar/ 36Ar ratio and the standard deviations are 0.1887±0.002
(Faraday/Daly), 0.1823±0.002 (Faraday/Daly), respectively.
These 40Ar/36Ar and 38Ar/36Ar ratios are significantly higher than
the atmospheric 40Ar/ 36Ar and 38Ar/ 36Ar ratios of 295.5 and
0.1869 [Nier, 1950; Steiger and Jäger, 1977]. This implies that
mass discrimination corrections are necessary in the sample analyses. Therefore, mass discrimination is corrected in a usual analysis. The equation and the error for mass discrimination correction
are summarized in e.g. Saito [1989], Matsumoto et al. [1989a].
The other important fundamental assumption is the linearity of the
sensitivity between Ar gas volume and the intensity. To examine
the linearity, intensities of 40Ar have been analysed by repeating
the same procedure of dilution of the regular content of standard
air Ar. The good linearity is confirmed between two collectors
that the 40Ar (Faraday) and 36Ar (Daly) intensities are good
regressed at R=0.9977, shown in Table 3.
2-3 Analytical procedure
The extraction, purification and analytical procedure are fundamentally similar to those of Nagao and Itaya [1988], Takaoka et
al. [1989] and Matsumoto et al. [1989b].
The rock, powder or mineral sample of 0.01-1.5g is wrapped by
an Al foil about 0.02-0.03g. The samples are loaded in the sample
holder made of a kind of pyrex glass (Koshitu glass 2nd grade).
Then, the samples are baked out at about 120-130 ºC pumping the
inside of the extraction line with a turbo molecular pump to
remove adsorbed atmospheric Ar on the surface of the samples
during more than a day. After the baking, the extraction line is isolated from the pump. Then, the sample is fallen into the molybdenum crucible using hand magnet and is fused by tantalum heater at
1600 ºC for 15 minutes. The extracted gases are simultaneously
purified by Ti-Zr getter at 800 ºC for 10 minutes to remove active
gases such as O2, N2 and CO2 and decompose H2O, and at room
temperature for 10 minutes. Then, purified gases are collected to
the charcoal trap kept at liquid nitrogen temperature and more purified with a Ti-Zr getter at 800 ºC for 10 minutes and at room temperature for 10 minutes to remove mainly hydrocarbon. Finally,
the gases are purified with Ti-Zr getter at 800 ºC for 10minutes and
at room temperature for 10 minutes to remove mainly hydrogen,
and are introduced to the mass spectrometer. The analytical condi-
3.2 Calibration of standard air Ar
Calibration of the standard air Ar sample has been performed
as follows. The international mineral standards (STD) whose radiogenic 40Ar concentrations are known to be in Table 2, have been
used. The biotite samples and the standard air Ar sample have
been alternatively analyzed 5 times. In each analysis, about
0.0010-0.0003g of STD was used. The 40Ar content of each standard air Ar sample has been calculated by the sensitivity which is
2
FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
the ratio of the radiogenic 40Ar intensity (volt) to the amount
[cm3STP] of that of the STD. The results of determination of 40Ar
content in the standard air Ar are listed in Table 3.
Obtained initial content and depletion rate are listed in Table 2.
The depletion rate of 0.9990 has been adopted because the rate of
0.9995 obtained from the volume (standard air Ar reservoir: 1500
cm3, the pipette: 1.4 cm3) estimated by Horiguchi Iron Works.
The error of determination of 40Ar content is 0.91% (Tab. 3). The
stability of the sensitivity (Volt/cm3STP) has been examined by
the STD analyses data. The twelve sensitivities during three
months are very concordant and the average sensitivity is
5.967×107 Volt/cm3STP (Tab. 3). This confirms the stability of
the system and suggests good reproducibility of sample analyses.
tion of them. The errors of 40Ar intensity, 40Ar/36Ar and 38Ar/36Ar
have been estimated by repeated standard air Ar analyses.
No systematic errors among them. The ages of MesozoicQuaternary STD samples have been obtained within about ±215% errors. This implies that this Ar analytical system may be
sufficiently useful for K-Ar dating of geological samples with
Cenozoic age.
Acknowledgements. We would like to express our grateful thanks
to Drs. Kozo Uto, Akikazu Matsumoto, Tomoaki Sumii and Masafumi
Sudo [now in Potsdam Univ.] of Geological Survey of Japan for offering K-Ar age reference STD samples. Sincere thanks also go to
Professors Kyoichi Ishizaka of Kyoto University, Ichiro Kaneoka of
Tokyo University, Kazuo Saito of Yamagata University and Takahiro
Tagami of Kyoto University for offering K-Ar age STD reference
samples. We also acknowledge to Professors Ichiro Kaneoka of Tokyo
University, Jun-ichi Matsuda of Osaka University and Keisuke Nagao
of Tokyo University for offering noble gas STD reference samples.
We also acknowledge to Professors Nobuo Takaoka of Kyusyu
University, Kazuo Saito of Yamagata University and Dr. Naoyoshi
Iwata of Yamagata University for many technical advices and teaching us important literatures. Thanks also go to IFREE researchers and
IFREE support staffs in JAMSTEC and Supervisor Dr. Yoshiyuki
Tatsumi and for their support and encouragement.
3.3 Blank of the analytical system
A hot blank is analyzed by the same procedure of the unknown
sample analysis without the sample. The 25 times analytical
results of 40Ar during six months distribute from 0.01 to 0.001
volt. A cold blank is analysed by the same procedure of the
unknown sample analysis without operating tantalum resistance
furnace. The three times analytical results of 40Ar during six
months distribute from 0.005 to 0.001V.
In the sample analysis, the sample is wrapped by about 5-20mg
of Al foils. A sheet of Al foil of 5-20 mg has been analyzed. In
usual Ar analyses, only hot blank is corrected because cold blanks
are usually included in hot blanks and the Ar gas derived from Al
foils is negligible. The calculated method of hot blank correction
and the error are summarized in e.g., Matsumoto et al. [1989a]
and Sudo et al. [1996].
Reference
Dalrymple G.B., Potassium-argon ages of Recent rhyolites of the
Mono and Inyo Craters, California, Earth and Planetary Science
Letters, 3, 289-298, 1968.
Dalrymple G.B. and M.A. Lanphere, Potassium-Argon DatingPrincipals, Techniques and Application to Geochronology, In A
series of books in geology, Ed., Jamees Gilluly and A.O.
Woodford, U.S, Geological Survey, 1969.
Dodson M.H., A linear method for second-degree interpolation in
cyclical data collection, Journal of Physics E: Science Instrument,
11, 296, 1978.
Flisch M., Potassium-argon analysis, In Numerical dating in stratigraphy, Ed., S. Odin Gilles, 51-158, John Wiley & Sons, Chichester,
United Kingdom, 1982.
Hurford A.J. and K. Hammerschmidt, (super 40) Ar/ (super 39) Ar
and K/ Ar dating of the Bishop and Fish Canyon tuffs; calibration
ages for fission-track-dating standards, Chemical Geology; Isotope
Geoscience Section, 58, 23-32, 1985.
Itaya T., K. Nagao, K. Inoue, Y. Honjou, T. Okada and A. Ogata,
Argon isotope analysis by a newly developed mass spectrometric
system for K-Ar dating, Mineralogical Journal, 15, 203-221,
1991.
Kaneoka I., Rare gas isotopes and mass fractionation; an indicator of
gas transport into or from a magma, Earth.Planet. Science Lett.,
48, 284-292, 1980.
Krummenacher D., Isotopic composition of argon in modern surface
volcanic rocks, Earth.Planet. Science Lett., 8, 109-117, 1970.
Matsumoto A., Improvement for determination of potassium in K-Ar
dating, Bulletin - Japan, Geological Survey, 40, 65-70, 1989.
Matsumoto A., K. Uto and K. Shibata, Argon isotopic ratios in historic lavas –Importance of correction for the initial argon in K-Ar
dating of young volcanic rocks-, Journal of Mass Spectrometry of
Japan, 40, 565-579, 1989a.
Matsumoto A., K. Uto and K. Shibata, K-Ar dating by peak compari-
4. K-Ar dating results of some age known STD
samples
In order to confirm the calibrated Ar analytical system, several
mineral standards of K-Ar ages and K-Ar age known STD samples have been analyzed in this system. Duplicate Ar analyses of
six samples have been performed (Tab. 2).
The duplicate analyses have been performed to six K-Ar age
mineral standards or K-Ar age known samples of Mesozoic to
Quaternary ages. Obtained ages mostly agree well with those of
literatures. There have been no systematic errors among them.
The ages of Cretaceous sample of Sori93-A and Sori93-B have
been obtained within about ±2-10% errors. The ages of four
Neogene samples (Bern4B, Baba-01, FC-3, HD-B1 biotite) have
been obtained within about ±2-3% errors and that of one
Quaternary sample (YZ-1.4) within about ±5-15% error. This
implies that this Ar analytical system may be sufficiently useful
for K-Ar dating of geological samples with younger ages than
Neogene age.
5. Summary
The new Ar analytical system by sensitivity method of IFREE
has been constructed and calibrated. K-Ar dating by sensitivity
method is suitable for dating of Neogene to Quaternary volcanic
and tectonic events. Calibration has been performed using Sori93
in which the amount of radiogenic 40Ar is known.
The system consists of GV Instrument 5400He mass spectrometer and the main sample handling system for the extraction of
sample gases by tantalum resistance furnace and for the purifica3
FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
son method; new technique applicable to rocks younger than 0.5
Ma, Bulletin - Japan, Geological Survey, 40, 565-579, 1989b.
Matsumoto A. and T. Kobayashi, K-Ar age determination of late
Quaternary volcanic rocks using the "mass fractionation correction
procedure": application to the Younger Ontake Volcano, central
Japan, Chemical Geology, 125, 123-135, 1995.
McDougall I. and T.M. Harrison, Geochronology and thermochronology by the 40Ar/39Ar method, 212pp., Oxford University Press,
Oxford, United Kingdom, 1988.
Nagao K. and T. Itaya, K-Ar age determination, The memoirs of the
Geological Society of Japan, 29, 5-21, 1988.
Nier A.O., A determination of the relative abundance of carbon, nitrogen, oxygen, argon, and potassium, Physics Review, 77, 789-793,
1950.
Saito K., Challenges to Limits - A case for the K-Ar dating -, Journal
of the Mining and Materials Processing Institute of Japan, 105,
1139-1146, 1989.
Steiger R.H. and E. Jäger, Subcommission on geochronology:
Convention on the use of decay constant in Geo-and
Cosmochronology, Earth. Planet. Science Lett., 36, 359-362, 1977.
Sudo M., T. Tagami, K. Sato, N. Hasebe and S. Nishimura,
Calibration of a new analytical system for the K-Ar dating method
and analytical results of K-Ar age known samples, Mem. Fac. Sci.
Kyoto Univ. Geol. Mineral. 58, 21-40, 1996.
Sudo M., K. Uto, K. Anno, O. Ishizuka and S. Uchiumi, Sori 93
biotite: Anew mineral standard for K-Ar dating, Geochemical
Journal, 32, 49-58, 1998.
Takahashi M., K. Saito, H. Umetsu and N. Ichikawa, K-Ar and 40Ar39Ar ages of the Miocene Kitamura and Baba tuffs in the Tomioka
area, Gunma Prefecture, central Japan; with special reference to
the N.13/ N.14 boundary of planktonic foraminiferal zones,
Journal of the Geological Society of Japan, 98, 323-335, 1992.
Takaoka N., K. Konno, Y. Oba and T. Konda, K-Ar dating of lavas
from Zao Volcano, northeastern Japan, Journal of the Geological
Society of Japan, 95, 157-170, 1989.
Takaoka N., Problem in K-Ar dating of Quaternary volcanic rocks,
Journals of the Mass Spectrometry Society of Japan, 37, 343-351,
1989.
Tamura H., K. Sato, T. Hanyu and H. Kumagai, Noble Gas Isotope
Analysis System: Design and Implementation, In Frontier
Research on Earth Evolution, IFREE Report 2003-2004, 2, 2005.
4
FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
Figure 1. The analytical results of variation in 40Ar contents, 40Ar/36Ar and 38Ar/36Ar ratios of standard air AS-tank
argon according to pipetting operations during 6 months. The error bars of all plotting points indicate analytical error
in each analysis. Some error bars are hidden in the plotting points.
Figure 2. Comparing for linearity between intensity of 40Ar (Faraday) vs. 36Ar (Daly). These are good linearity
and good correlation.
5
FRONTIER RESEARCH ON EARTH EVOLUTION, VOL. 2
Table 1. The analytical condition of the GV Instrument 5400He mass spectrometer.
Table 2. Results of calibration standard for AS air Ar using geochemical standard.
Italic names are internal/international geochronological standard, which referenced from, Sudo et al. [1998],
Flish [1982], Hurford and Hammershmidt [1985], Takahashi et al. [1992], McDougaall and Harrison [1988],
Takaoka [1988] and Takaoka et al [1989].
Table 3. Regression results of AS air standard calibration.
SDOM = standard deviation of mean ; volume (cm3STP) ; sensitivity (Volt/cm3STP)
Referrence calculation equation is Sudo et al. [1996].
6