Meteoritics & Planetary Science 36, 1345–1366 (2001)
Available online at http://www.uark.edu/meteor
Argon-40–argon-39 chronology and petrogenesis along the eastern limb
of the Moon from Luna 16, 20 and 24 samples
B. A. COHEN1, G. A. SNYDER1, C. M. HALL2, L. A. TAYLOR1 AND M. A. NAZAROV3
1Planetary
2Department
Geosciences Institute, University of Tennessee, Knoxville, Tennessee 37996, USA
of Geological Sciences, University of Michigan, Ann Arbor, Michigan 48109, USA
3Vernadsky Institute of Geochemistry, Moscow 117975, Russia
*Correspondence author's e-mail address: [email protected]
(Received 2001 February 9; accepted in revised form 2001 June 29)
Abstract–Eighteen new lithic fragments from the Soviet Luna missions have been analyzed with
electron microprobe and 40Ar-39Ar methods. Luna 16 basalt fragments have aluminous compositions
consistent with previous analyses, but have two distinct sets of well-constrained ages (3347 ~ 24 Ma,
3421 ~ 30 Ma). These data, combined with other Luna 16 basalt ages, imply that there were multiple
volcanic events filling Mare Fecunditatis. The returned basalt fragments have relatively old cosmicray exposure (CRE) ages and may have been recovered from the ejecta blanket of a young (1 Ga),
nearby crater. A suite of highlands rocks (troctolites and gabbros) is represented in the new Luna 20
fragments. One fragment is the most compositionally primitive (Mg# = 91–92) spinel troctolite yet
found. Both troctolites have apparent crystallization ages of 4.19 Ga; other rocks in the suite have
progressively younger ages and lower Mg#s. The age and composition progression suggests that
these rocks may have crystallized from a single source magma, or from similar sources mobilized at
the same time. Within the new Luna 24 basalt fragments is a quench-textured olivine vitrophyre
with the most primitive composition yet analyzed for a Luna 24 basalt, and several much more
evolved olivine-bearing basalts. Both new and previously studied Luna 24 very low-Ti (VLT) basalt
fragments have a unimodal age distribution (3273 ~ 83 Ma), indicating that most returned samples
come from a single extrusive episode within Mare Crisium much later than the Apollo 17 VLT
basalts (3.6–3.7 Ga).
INTRODUCTION
The Soviet Luna 16, 20, and 24 missions are so far the only
autonomous missions to have returned samples from another
planetary body. These robotic missions returned drill-core
samples of the lunar surface from northeastern Mare
Fecunditatis, the highlands just south of the Crisium basin, and
southern Mare Crisium, respectively. Our aim in this study is
to relate the chemistry of the rocks from these sites to the source
and evolution of their parent magmas, and to determine the
dates of their emplacement.
The Luna 16 and 24 samples provide information on the
nature and age of the basalt flows that filled Mare Fecunditatis
and Mare Crisium. Although they are geographically close to
each other, the two basins are filled with very different material.
The Luna 24 basalts contain high aluminum but very low titanium
(VLT) (<1.5 wt% TiO2), similar to the Apollo 17 VLT basalts. In
contrast, the Luna 16 basalts are high in aluminum (>11 wt% Al2O3)
and chemically similar to the Apollo 14 aluminous basalts.
Mare Fecunditatis is a relatively old and degraded multiring basin that has been filled by lava flows. Luna 16 sampled
Prelude preprint MS#4508
surficial mare material between the two Copernican-age craters
Langrenus and Taruntius. Basalts comprise most of the coarse
lithic fraction of the core sample, with some anorthositic and
noritic fragments transported to the site by younger impact
events (McCauley and Scott, 1972). As a group, the Luna 16
basalts contain more TiO2 (∼5 wt%) than the low-Ti, aluminous
Apollo 14 basalts (Taylor et al., 1991). The Apollo 14 basalts
may be the oldest basalts known so far (Taylor et al., 1983;
Taylor, 1982), though their origin is disputed (Snyder and
Taylor, 1996; Warren et al., 1997).
Mare Crisium is known to contain a lithologic treasure trove
including some of the most evolved basalts on the Moon. At
least four different flows can be stratigraphically mapped within
this old basin (Head et al., 1978), and these distinctions are
borne out in the Fe and Ti contents as mapped by the Clementine
orbiter (Bussey and Spudis, 2000). The Luna 24 landing site
was in a patch of low-Ti, Fe-rich basalt surrounded by
stratigraphically younger basalts (Head et al., 1978). The
majority of Luna 24 basalt fragments are high-Al, VLT basalts
with similar chemical characteristics to the Apollo 15 green
glass, though the basalts are much less primitive.
1345
© Meteoritical Society, 2001. Printed in USA.
1346
Cohen et al.
Luna 20 is the only highlands site on the eastern limb of the
Moon where samples have been collected; however, the
collection has received proportionately little attention in the
west. Because it is a highlands site, the lithic fragments are
predominantly anorthosites, as well as granulites, Mg-suite
rocks, and impact melt rocks, at least some of which may be
derived from nearby Mare Crisium (Podosek et al., 1973;
Swindle et al., 1991).
We have studied the petrology, mineral-chemistry, and
40Ar-39 Ar ages of 18 "new" fragments from soil returned by
the Luna 16, 20, and 24 missions. Of the six fragments from
Luna 16, ranging in mass from 5.5 to 10.25 mg, four are finegrained basalts and two are regolith agglutinates. Six fragments
from Luna 24, ranging in mass from 2.1 to 9.9 mg, are finegrained VLT basalts. Six fragments from Luna 20, ranging in
mass from 1.25 to 7.55 mg, are Mg-suite plutonic rocks: two
troctolites, two gabbros, and two single anorthite grains.
SAMPLE PREPARATION AND ANALYSIS METHODS
Eighteen fragments were received from the Vernadsky
Institute in Moscow, Russia. Of these, sixteen were basalts or
other igneous rocks, and two (2002 and 2004A) were single
anorthite grains. Chips of allocated fragments that were large
enough to split were thin-sectioned for electron microprobe
(EMP) analysis. The Cameca SX50 electron microprobe at the
University of Tennessee was used for all analyses. Minerals
were used for calibration, and analysis conditions were 15 keV
and 20 nA, with a 20 s collection time for all elements.
Petrographic descriptions and mineral chemistry of the
fragments are included in the next section. Typical textures
are shown in backscattered electron (BSE) images in Fig. 1.
The remaining fragments, ranging in mass from 0.2 to 2.7 mg
(Table 1), were irradiated for 350 h in location L-67 of the
Phoenix-Ford Memorial reactor at the University of Michigan,
producing a J-factor of 5.06 × 10–2. The standard mineral 3gr
hornblende (1071 Ma; Roddick, 1983) was simultaneously
irradiated and analyzed. Stepped-heating experiments were
performed using a continuous-wave laser as described in Snyder
et al. (1996). The laser power varied from 50 to 4000 mW
over each experiment. The temperature at each step was not
measured; however, it scales with power (aP0.25), and the
maximum temperature achieved (at 4000 mW) is ∼1800–
2000 #C. System blanks at masses 36 to 40 were, in order, 1 ×
10–13, 4 × 10–13, 5 × 10–14, 5 × 10–14 and 4 × 10–12 cc STP.
All data was corrected for blanks, mass discrimination, K and
Ca interference, and decay since irradiation. Additionally, data
were corrected for cosmic-ray spallation by deconvolving the
38Ar/36Ar signatures from cosmic ray-spallation (36Ar/38Ar =
0.66; Hohenberg et al., 1978) and terrestrial atmosphere
(36Ar/38Ar = 5.32). An initial 40Ar/36Ar ratio, (40Ar/36Ar)i,
was calculated from the isochron intercept for the high-temperature
FIG. 1. BSE images of representative Luna 16, 20, and 24 samples: (a) 1609A basalt; (b) 1609B basalt; (c) 2003 troctolite; (d) 2004C
troctolite; (e) 2004D gabbro; (f) 24067,3-6A basalt. The scale bar in all images is 100 lm.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1347
TABLE 1. Luna 16, 20, and 24 sample numbers, types, masses, (40Ar/36Ar)i, and whole-rock gas ages.
Sample
Type
1607-1608
1609A
1609B
1609C
1635A
1635B
2002
2003
2004A
2004B
2004C
2004D
24067,3-6A
24067,3-6B
24143,4-6L
24184,4-6A
24184,4-6B
24184,4-6C
Luna 16
Luna 16
Luna 16
Luna 20
Luna 20
Luna 24
basalt
basalt
basalt
glassy fragment
glassy fragment
basalt
anorthite
spinel troctolite
anorthite
gabbro
troctolite
gabbro
basalt
basalt
basalt
coarse basalt
coarse basalt
basalt
bulk soil
bulk soil
bulk soil
bulk soil
bulk soil
bulk soil
Mass (mg)
(40Ar/36Ar)i
1.5
2.3
2.3
0.9
2.4
1.6
1.3
0.9
2.1
0.4
0.3
0.5
2.7
1.6
0.5
1.9
0.5
0.2
–
–
–
–
–
–
1.17
0.67 ~ 0.06
0.97 ~ 0.11
2.22 ~ 0.06 > x > 1.78 ~ 0.04
1.67 ~ 0.05
0.97 ~ 0.19
1.61 ~ 0.03
1.50 ~ 0.02
1.53 ~ 0.08
2.51 ~ 0.02
1.18 ~ 0.05
2.16 ~ 0.28
0.55 ~ 0.01
0.23 ~ 0.01
0.49 ~ 0.06
0.94 ~ 0.13
0.84 ~ 0.09
0.74 ~ 0.09
00.710 ~ 0.003*
00.81 ~ 0.04†
01.96 ~ 0.50‡
2.01‡
01.56 ~ 0.02§
\0.78 ~ 0.14#00
Whole-rock gas age (Ma)
2618
3009
4047
2139
2625
4378
4374
4282
6060
4652
4096
4009
3395
3605
3336
3373
3352
–
~ 20
~ 20
~ 30
~ 20
~ 20
~ 40
~ 40
~ 20
~ 13
~ 13
~ 40
~ 30
~ 40
~ 70
~ 20
~ 50
~ 10
–
–
–
–
–
–
*Kaiser
(1972).
et al. (1973).
‡Agrawal et al. (1974).
§Eugster et al. (1975).
#Average and standard deviation of Bogard and Hirsch (1978).
†Eugster
steps and subtracted from the data. The calculated (40Ar/36Ar)i
values are tabulated in Table 1 and are generally comparable
to (40Ar/36Ar)i values determined for the bulk soils at each
site (Agrawal et al., 1974; Bogard and Hirsch, 1978; Eugster
et al., 1973, 1975; Kaiser, 1972). As shown in Fig. 2, the lowest
temperature steps usually did not fall on the isochron and have
40Ar/36Ar ratios that do not conform to the intercept value.
These steps usually have chronologically meaningless apparent
ages in the argon-release spectra. The whole-gas ages for each
sample are also tabulated in Table 1.
Plateaus were determined as a series of steps where the
apparent age of each step was within 2r of the previous step
and where a general upward or downward trend was not
apparent. The steps used to compute the plateau ages are
indicated on the argon release plots and usually make up more
than 50% of the total 39Ar release; those cases that differ are
discussed individually. The plateau age was computed by
adding the volumes of 40Ar and 39Ar released over the plateau
steps and computing an integrated age. The error quoted in the
plateau age is 2r. The K/Ca ratio was computed from the
39Ar/ 37Ar ratio in each gas-release step and is plotted above the
argon release spectra for comparison. Note that the scale on all
FIG. 2. Isochron correlation diagram for sample 1609B (basalt). The
high-temperature steps are used to fit a line whose intercept defines
the initial 36Ar/40Ar ratio.
1348
Cohen et al.
argon-release plots is the same for apparent age but not for K/Ca,
and that the K/Ca ratio has been stretched by a factor of 100.
MINERAL CHEMISTRY AND PETROLOGY
Luna 16
The six Luna 16 fragments are of two types: fine-grained
basalts (1609A, 1609B, 1635B, 1607-1608) and glassy,
vesicular particles (1609C, 1635A). The most conspicuous
aspect of the basalts is the presence of large (50–300 lm in
length) laths of ilmenite that comprise ∼10% of the samples.
Plagioclases have Ca-rich cores; pyroxenes are also zoned to
more Fe-rich rims. The zoning is typical of rapid crystallization
in an evolving melt. The glassy particles have vesicles, a
"porphyritic" texture, and the appearance of uniformity in the
matrix. Spinels and ilmenites are relatively rare. Pyroxene,
olivine, and spinel compositions are shown in Fig. 3.
FIG. 3. Luna 16 mineral compositions: (a) pyroxene; (b) olivine; (c) spinel.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1609A Basalt–This sample is fine-grained (Fig. 1a). Both
ilmenite and plagioclase are typically acicular and similar in
their longest dimensions (100–250 lm); plagioclase core
compositions vary from An90 to An95 with rims being up to six
An units lower. Pyroxenes and olivines are subhedral to
anhedral and vary in size from 150 to 300 lm. Pyroxenes are
nearly exclusively augites with Fe-rich and Ca-poor rims.
Olivines also vary little. Spinels are minor, fine-grained, and
blocky.
1609B Basalt–This sample is similar to 1609A (Fig. 1b),
but is distinctive in containing a large (350 lm) subhedral
olivine phenocryst (Fo68–69), which is strongly zoned to Fo32
at the rim. Finer grained (50–200 lm) olivines vary in
composition from Fo46 to Fo68. Grain sizes of particular
minerals are identical to 1609A. Pyroxene compositions are,
again, nearly exclusively augite with strong zonation to more
Fe-rich rims. Plagioclase varies in composition from An80 to
An92. Fine-grained chromites generally contain 25–30% Cr2O3
but show a range in compositions. Rims on large grains are
much lower in Cr, Al, and Mg, and higher in Ti.
1635B Basalt–This sample consists of approximately equal
proportions of plagioclase and mafic phases and is slightly
coarser-grained than 1609A or 1609B, indicating slower
cooling. Ilmenite laths vary in composition from Mg# (100×
atomic Mg/Mg + Fe) = 1–9 to Cr# (100× atomic Cr/Cr + Al) =
64–83. Much finer grained and blocky Cr-spinels are minor
and scattered throughout. The plagioclase occurs both as blocky
grains and as more acicular interstitial material and varies in
composition from An78 to An92. The mafic phases include
two pyroxenes, dominantly augite and less prevalent pigeonite.
Larger olivine grains (possibly phenocrysts) typically have
compositions from Fo66 to Fo58. Rims on these olivines are
more Fe-poor, often by up to 5–8 Fo units. Finer-grained
olivine, adjacent to oxides and associated with mesostasis
may have compositions as Fe-rich as Fo35–37. The mineral
chemistry of this sample is consistent with that of a mare
basalt.
1607-1608 Basalt–Plagioclase varies widely in composition
from An 79 to An 92. Uncommon olivine grains range from
Fo49–Fo59. Pyroxenes occur as principally two varieties, augite
and uncommon pigeonite, with occasional ferroaugite grains.
Pyroxenes are often zoned to more Fe-enriched rims. Ilmenite
laths vary in composition from Mg# = 1–9 and Cr# = 59–79.
Finer-grained spinels have a range in compositions.
1609C Glassy Fragment–This sample is a mostly fine
grained, glassy sample with numerous vesicles. In general, the
glass is relatively uniform in composition: SiO2 = 44.1–45.6 wt%;
Al2O3 = 19.3–30.4 wt%; MgO = 4.14–8.91 wt%; CaO = 12.8–
17.1 wt%; FeO = 2.51–10.5 wt%. One area of glass (250 ×
250 lm) appears pseudomorphous (equant and blocky) after
feldspar and is uniform in composition: SiO2 = 45.2 (2)
wt%; Al2O3 = 19.5 (2) wt%; MgO = 8.46 (6) wt%; CaO =
13.2 (1) wt%; FeO = 10.8 (1) wt%; Na2O = 0.286 (16) wt%.
Ilmenites (Mg# = 1–4; Cr# = 44–78) occur as <50 lm angular
1349
grains and spinels were not seen. Plagioclase phenocrysts are
angular (50–150 lm) and vary in composition from An86 to
An 96, although most grains are An 90–94 in composition.
Pyroxenes are typically 50–150 lm in size and vary from
pigeonite to augite in composition. A rare Fe-enriched pigeonite
was also analyzed. Olivines occur as small (generally <50 lm)
grains that vary from Fo56 to Fo65 in composition. However,
several olivine grains are more Fe-rich (Fo41, Fo17, Fo8).
1635A Glassy Fragment–Phenocrysts consist of mostly
pyroxene and olivine and occasional plagioclase. Both
pigeonite and augite occur. Olivine phenocrysts vary widely
in composition from Fo36 to Fo71. Plagioclase phenocrysts
vary in composition from An81 to An95. Finer grained K-feldspar
is minor, associated with the fine-grained, glassy matrix, and
yields a composition of Ab7Or75An18. Ilmenite (Mg# = 5–18;
Cr# = 18–88) is typically anhedral and relatively coarse-grained.
Spinels (Cr# = 56) are rare and fine-grained.
Luna 20
Of the six Luna 20 fragments, two are troctolites, two are
gabbros, and two are single anorthite grains. Due to their small
total-sample size, the anorthite grains were not analyzed by
EMP. The pyroxene compositions within the gabbros and the
spinel and olivine compositions of the gabbros and troctolites
are plotted in Fig. 4.
2004C Troctolite–This sample is a porous troctolite (Fig. 1d)
consisting of unzoned plagioclase (An96–97) and finer grained
globular olivine (Fo76–77).
2003 Spinel Troctolite–This sample (Fig. 1c) consists
mostly of plagioclase and small (<60 lm) grains of olivine and
spinel. This sample is unique in having extremely primitive
olivine (Fo95–96), plagioclase (An95–97), and spinel. A rare
Zn-rich spinel was also analyzed. FeNi metal is also present
(Co = 0.67–0.72 wt%; Ni = 8.81–8.85 wt%).
2004B Anorthositic Gabbro–This sample consists of
mostly anhedral plagioclase (100–300 lm) (An96–97). Olivine
and pyroxene occur as interstitial grains <70 lm. Olivine varies
in composition from Fo83 to Fo73. Pyroxenes occur as pigeonite
and augite. Spinels occur as tiny grains (<30 lm) but also as
larger grains (up to 100 lm) and exhibit two distinct
compositions: Mg-Al chromite and true aluminous spinel. FeNi
metal (Co = 0.37–0.67 wt%; Ni = 1.11–6.55 wt%) is also
present, especially as one large (100 lm) grain.
2004D Gabbro–This sample (Fig. 1e) is conspicuous in
having very coarse-grained (200–400 lm) pyroxene in
combination with acicular plagioclase (<200 lm in longest
dimension). The uniform high-Ca plagioclase composition
(An91–96) is indicative of a highlands affinity. The coarse
pyroxenes vary in composition from pigeonite to augite, both
zoned to more Fe-rich compositions. A traverse across one of
the largest pyroxene grains indicated zonation from a pigeonite
core to a ferroaugite rim. Some finer grained pyroxenes
approach hedenbergite and pyroxferroite in composition.
1350
Cohen et al.
FIG. 4. Luna 20 mineral compositions: (a) pyroxene; (b) olivine; (c) spinel.
Minor, euhedral to subhedral olivine (50–150 lm) is
associated with plagioclase and is exclusively fayalitic in
composition (Fo 1–2 ). These may represent part of the
breakdown assemblage of pyroxferroite. The pyroxenes and
fayalite compositions are more typical of mare basalt than
highlands gabbro.
Luna 24
All six fragments from Luna 24 are basalts. Based on
textures and grain size, these fragments can be subdivided into
fine-grained and coarse-grained (gabbroic) varieties. The
pyroxene, olivine, and spinel compositions are plotted in Fig. 5.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1351
FIG. 5. Luna 24 mineral compositions: (a) pyroxene; (b) olivine; (c) spinel.
24067,3-6A Fine-Grained Quench Basalt–This fragment
is conspicuous in exhibiting a very fine-grained (mostly <50 lm)
texture, including acicular plagioclase, and fine-grained
pyroxenes, rare olivines, and an occasional olivine phenocryst
(Fig. 1f). Plagioclases are exclusively high-Ca (An96 to An91).
A relatively large (150 lm), euhedral olivine phenocryst is
zoned with a core composition of Fo62–64 and a rim composition
of Fo60. Other smaller grains typically fall in the composition
range of Fo63 to Fo50 with occasional grains as low as Fo35.
Pyroxenes are exclusively augites. Glass compositions are
roughly mafic (SiO2 = 49.5–51.6 wt%; TiO2 = 1.28–1.41 wt%;
Al2O3 = 14.7–15.8 wt%; MgO = 2.43–3.35 wt%; CaO = 16.0–
17.2 wt%; FeO = 10.7–12.5 wt%; Na2O = 0.31–0.40 wt%). Minor,
fine-grained (<40 lm) Cr-spinels exhibit uniform compositions.
24067,3-6B Fine-Grained Olivine Basalt–This sample is
very similar to 24143,4-6L. Both samples have a greater
proportion of oxides (ilmenite and Ti-rich spinel) than other Luna
24 samples. Plagioclase (An92–95) is typically acicular to tabular
1352
Cohen et al.
(typically <150 lm in longest dimension) with a much lower aspect
ratio than the plagioclase in quench basalt 24067,3-6A. Olivines
and pyroxenes are similar and uniform in size (100–200 lm).
Olivines exhibit a narrow range in composition. Pyroxenes
occur as both pigeonites and augites. Ilmenites (Mg# = 4; Cr# =
80–85) and spinels (Mg# = 1–2; Cr# = 66–68) are uniform in
composition.
24143,4-6L Olivine Basalt–This sample is a fine-grained
olivine basalt that is very similar in composition to 24067,3-6B.
However, pyroxenes in this sample are exclusively pigeonitic.
Plagioclase compositions vary little from An94 to An91. Olivine
compositions are also uniform. Fe-Ti-rich spinels occur both
as small (<30 lm) grains and large (up to 200 lm) sieve-textured
grains.
24184,4-6B Coarse-Grained Basalt–This sample consists
principally of a few large (200–660 lm), euhedral, tabular,
plagioclase grains which are typically high-Ca in composition
(An91–94). Finer interstitial grains are lower in Ca (An77–80).
Pyroxenes (both pigeonite and augite) are typically euhedral
and coarse-grained (100–400 lm). Rare, fine-grained fayalitic
(Fo7–8) olivines appear associated with ilmenite (Mg# = 1; Cr# =
55–69) as the last crystallizing material.
24184,4-6C Ilmenite-Rich Basalt–This sample is finer
grained than 24184,4-6B, but still coarser grained than the "finegrained" basalts. Plagioclases (An90–94) are typically acicular
(100–200 lm in length) to tabular and often occur in radiating
"clots". Pyroxenes and olivine (100–300 lm) often
poikilitically enclose plagioclase. Olivines vary widely in
composition from Fo3 to Fo54. Pyroxenes occur as pigeonites,
low-Fe augites, and ferroaugites. Rims on pyroxene grains are
more Fe-enriched. Silica occurs in the mesostasis. Ilmenites
(Mg# = 1, Cr# = 42–46) and spinels are finer grained (<50 lm)
and associated with mesostasis and plagioclase. Two different
spinels occur: Ti-rich and chromite.
ARGON-40–ARGON-39 SYSTEMATICS
Luna 16
Argon-release spectra and K/Ca ratios for all six fragments
from Luna 16 are shown in Fig. 6 and ages are reported in
Table 2. Each heating step is shown with a 1r uncertainty,
though the plateau ages are reported with a 2r uncertainty. In
this section, the interpretation of the release patterns is
discussed.
1609A Basalt–This sample is a fine-grained basalt. The
presence of an incompatible-element-enriched, glassy
mesostasis is suggested by the high K/Ca ratios in the lowtemperature steps. A plateau is reached over 38% of the 39Ar
release in the high-temperature steps, yielding an apparent age
of 3155 ~ 4 Ma. The amount of degassing is estimated by
comparing the 40Ar/39Ar ratio of the total sample to that of the
plateau (plus higher temperature) steps. By this method, this
sample appears to have lost ∼60% of its initial 40Ar in an event
at ∼1000 Ma. Such a large amount of degassing has the effect
of decreasing the apparent age of the disturbed sites (lowtemperature release). The effect on the apparent plateau age is
discussed later.
TABLE 2. Cosmic ray exposure ages and depths of Luna rocks.
Sample
Type
1609A
1609B
1635B
1607-1608
1609C
1635A
2004C
2003
2004B
2004D
2002
2004A
24067,3-6A
24067,3-6B
24184,4-6A
24143,4-6L
24184,4-6B
24184,4-6C
basalt
basalt
basalt
basalt
glassy fragment
glassy fragment
troctolite
troctolite
gabbro
gabbro
anorthite
anorthite
basalt
basalt
basalt
basalt
basalt
basalt
38Ar/Ca*
(cc STP/g)
CRE age*†
(Ma)
1.48 × 10–5
3.38 × 10–6
9.09 × 10–6
8.15 × 10–6
1.79 × 10–5
1.27 × 10–5
3.30 × 10–6
5.64 × 10–6
1.59 × 10–5
4.70 × 10–6
3.28 × 10–6
6.43 × 10–6
1.92 × 10–5
8.65 × 10–6
2.22 × 10–6
5.51 × 10–6
2.92 × 10–6
4.11 × 10–6
1060
241
649
582
1281
905
236
403
1133
336
234
459
1370
618
158
393
208
294
PFe/PCa
0.039
0.003
0.039
0.054
0.000
0.000
0.016
–0.016
–0.008
–0.017
–0.004
0.002
–0.007
0.019
–0.001
0.029
0.153
–0.192
Estimated depth
(cm)
<5
>100
<5
<5
–
–
–
–
–
–
–
–
–
∼25
–
<25
–
–
*Excluding Fe-produced 38Ar and high-temperature steps assumed to release 38Ar from pyroxene, as discussed in text.
†Using
nominal production rate of 1.4 × 10–8 cc STP/g Ca/Ma.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1353
FIG. 6. Argon-release spectra and K/Ca ratios for Luna 16 samples. The steps used in the calculation of sample age are indicated by the underbar.
1354
Cohen et al.
1609B Basalt–This sample is similar to 1609A, but is
distinctive in containing a large (350 lm) subhedral olivine
phenocryst. The argon-release pattern of this sample is also
similar to 1609A. The high K/Ca ratios indicate a K-rich glassy
mesostasis that releases argon in the low-temperature steps and
was disturbed at ∼1000 Ma. This disturbance caused ∼60% of
the 40Ar to be lost from the sample. A high-temperature plateau
is reached over 43% of the 39Ar release, yielding an apparent
crystallization age for this sample of 3347 ~ 4 Ma.
1635B Basalt–This sample is slightly coarser grained than
1609A or 1609B, indicating slower cooling. It has a higher
proportion of K-rich mesostasis present than do 1609A or
1609B, as evidenced by its higher K/Ca ratios in the lowtemperature steps. It too was disturbed at ∼1000 Ma, causing
40Ar release, but was more intensely affected, probably because
of the larger amount of glass present. Thus, only a minimum
age is observable at the maximum of the curve in the argonrelease spectrum, at 3421 ~ 7 Ma.
1607-1608 Basalt–This sample is essentially identical to
1635B in its argon release and K/Ca spectra. It experienced an
extreme disturbance at ∼1000 Ma and now only yields a
minimum crystallization age of 3426 ~ 6 Ma, identical within
error to 1635B.
1609C Glassy Fragment–This glassy sample has numerous
vesicles and mineral fragments and the argon spectrum is
problematic. The K/Ca ratio mimics that in the basaltic samples,
but an order of magnitude less K appears to be present. The
anomalously high apparent ages in the low-temperature steps
is probably a result of experimental recoil, where 39Ar is
redistributed within the sample during high-energy conversion
in the reactor, and sometimes completely lost from the lowretentivity sites (McDougall and Harrison, 1999). The downturn
in apparent age in the highest temperature steps also reflects
recoil redistribution. In the low-temperature steps, a negative
apparent age is derived. This is likely the result of an
overcorrection for 40Ar contributed from the lunar environment,
based on the amount of 36Ar in the sample, caused either by
36Ar recoil effects or an inappropriate initial 40Ar/36Ar ratio.
The remainder of the spectrum has a generally flat but ragged
character. The age calculated from this part of the spectrum is
4274 ~ 6 Ma.
1635A Glassy Basalt–This sample is also glassy and vesicular.
The argon release spectrum is as problematic as that in 1609C and
the contribution of excess 36Ar is apparent. However, the calculated
age over the flat part of the spectrum is much different: 2753 ~
4 Ma. Excluding steps with extremely high or low apparent ages
does not significantly change this age or error since only a small
amount of 39Ar is contributed by each step.
Luna 20
Of the six Luna 20 fragments, two are troctolites, two are
gabbros, and two are single anorthite grains. Figure 7 plots the
argon release spectra and K/Ca ratios of all six samples and
ages are reported in Table 3. Each heating step is shown with
a 1r uncertainty, though the plateau ages are reported with a 2r
uncertainty. In this section, the interpretation of the release
patterns is discussed.
2004C Troctolite–This sample is a porous troctolite.
Though the K/Ca ratio in this sample is nearly constant in all
heating steps, high apparent ages are exhibited in the lowtemperature releases. This effect is a recoil artifact, where the
generated 39Ar has enough energy to escape the K-rich phase.
Usually, the 39Ar implants itself into a neighboring Ca-rich
phase and is outgassed at high temperatures, leading to low
apparent ages, but that is not the case in this sample. The 39Ar
may have been lost from the low-retentivity sites altogether,
consistent with the fine-grained, porous nature of the
TABLE 3. 40Ar-39Ar and CRE ages of Luna 16 rocks.
Sample
1635A
1609A
1609B
1635B
1607-1608
1609C
Luna 16 bulk soil
Luna 16 bulk soil
Luna 16 bulk soil
*Minimum
Type
glassy fragment
basalt
basalt
basalt
basalt
basalt
basalt
basalt
basalt
glassy fragment
–
–
–
Age (Ga) ~ 2r CRE age (Ma)
2.753 ~ 0.004
3.155 ~ 0.004
3.347 ~ 0.004
3.35 ~ 0.15
3.39 ~ 0.05
03.421 ~ 0.007*
03.426 ~ 0.006*
03.42 ~ 0.04†
3.50 ~ 0.15
4.274 ~ 0.006
–
–
–
age.
using decay constants of Steiger and Jäger (1977).
n/a = not available.
†Recalculated
905
1060
241
1600
n/a
649
582
455
1000
1281
900
360
890
Reference
This work
This work
This work
Cadogan and Turner (1977)
Fernandes et al. (2000)
This work
This work
Huneke et al. (1972)
Cadogan and Turner (1977)
This work
Kaiser (1972)
Eugster et al. (1973)
Agrawal et al. (1974)
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1355
FIG. 7. Argon-release spectra and K/Ca ratios for Luna 20 samples. The steps used in the calculation of sample age are indicated by the underbar.
1356
Cohen et al.
plagioclase. A well-defined plateau extends over 89% of the
39Ar release, yielding an age of 4191 ~ 22 Ma.
2003 Spinel Troctolite–The argon release spectrum for this
sample is enigmatic. A highly degassed, K-rich phase is
suggested in the low-temperature steps, but no such phase was
found during EMP analysis. Recoil effects or uncorrected 36Ar
may account for the unreasonably high initial apparent ages.
Even if a satisfactory explanation of the low-temperature steps
is lacking, a clear plateau exists over 74% of the 39Ar release,
with an age of 4189 ~ 8 Ma.
2004B Anorthositic Gabbro–This sample consists of
anhedral plagioclase with interstitial olivine and pyroxene
grains. It has a remarkably homogeneous K/Ca ratio, indicating
that degassing is probably all from a single phase, plagioclase.
Therefore, the high initial ages are probably due to recoil loss
of 39Ar. The subsequent step-up in apparent age points to
diffusive 40Ar loss at an extrapolated time of ∼2 Ga. However,
the apparent plateau over 43% of 39Ar loss has an old age of
4114 ~ 25 Ma. If true diffusive loss occurred, this plateau
must be regarded only as a minimum age.
2004D Gabbro–This sample consists of coarse-grained
pyroxene with acicular plagioclase. The argon release profile
shows signs of recoil or uncorrected 36 Ar in the lowest
temperature steps. A K-rich phase is implied by the high K/Ca
ratios in the low-temperature steps as well. In contrast to the
Luna 16 and 24 basalts, this K-rich phase has not been
appreciably disturbed, instead contributing with the more Ca-rich
phases to a plateau over 64% of 39Ar release. The age of
4021 ~ 10 Ma is significantly younger than the 2004B
gabbro.
2002 and 2004A Anorthite–These samples are single
grains of anorthite, possibly coarse grains from other rocks at
the Luna 20 site. Neither was analyzed by EMP. The 2002
anorthite (Fig. 7e) is nearly identical to the 2004B gabbro in
diffusive profile after the first 20% of 39Ar released (Fig. 7c),
with loss at ∼2 Ga, and a plateau over 52% of 39Ar loss yielding
an age of 4114 ~ 10 Ma. The 2004A anorthite has a
significantly higher age of 4265 ~ 5 Ma over 61% of its 39Ar
release, accompanied by recoil and diffusion effects in the lowtemperature steps. Extrapolation of the diffusion profile
between 10% and 40% of 39Ar released points to a disturbance
age of ∼4 Ga.
Luna 24
All six fragments from Luna 24 are basalts. Their argon
release spectra and K/Ca ratios are shown in Fig. 8 and ages
are reported in Table 4. Each heating step is shown with a 1r
uncertainty, though the plateau ages are reported with a 2r
uncertainty. In this section, the interpretation of the release
patterns is discussed.
24067,3-6A Fine-Grained Quench Basalt–This sample
has a very fine-grained (mostly <50 lm), acicular texture. The
argon release spectrum is typical of lunar basalts and
qualitatively similar to the Luna 16 basalts in having a high-K
mesostasis that outgasses at low temperatures. However, this
TABLE 4. 40Ar-39Ar and CRE ages of Luna 20 rocks.
Sample
2004D
2004B
2002
2003
2004C
2004A
Luna 20 bulk soil‡
Luna 20 bulk soil‡
*Recalculated
Type
Age (Ga) ~ 2r
CRE age (Ma)
impact melt
impact melt
impact melt
recrystallized breccia
impact melt
impact melt
metaclastic rock
gabbro
impact melt
highland basalt
gabbro
anorthite grain
spinel troctolite
troctolite
anorthite grain
anorthite grains
anorthite
–
–
0.38 ~ 0.11
0.519 ~ 0.011
3.75 ~ 0.04
03.84 ~ 0.04*
3.879 ~ 0.031
3.895 ~ 0.017
3.90 ~ 0.10
4.021 ~ 0.010
4.087 ~ 0.021
4.1†
4.114 ~ 0.025
4.114 ~ 0.010
4.189 ~ 0.008
4.191 ~ 0.022
4.265 ~ 0.005
4.3 ~ 0.1
4.36†
–
–
180
340
2080
900–1300
1010
1010
600
336
680
n/a
1133
234
403
236
459
400–700
n/a
351
355
using decay constants of Steiger and Jäger (1977).
degassing age.
‡Based on 38Ar.
n/a = not available.
†Mid-temperature
Reference
Swindle et al. (1991)
Swindle et al. (1991)
Swindle et al. (1991)
Podosek et al. (1973)
Swindle et al. (1991)
Swindle et al. (1991)
Cadogan and Turner (1977)
This work
Swindle et al. (1991)
Huneke and Wasserburg (1979)
This work
This work
This work
This work
This work
Cadogan and Turner (1977)
Huneke and Wasserburg (1979)
Agrawal et al. (1974)
Eugster et al. (1975)
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
1357
FIG. 8. Argon-release spectra and K/Ca ratios for Luna 24 samples. The steps used in the calculation of sample age are indicated by the underbar.
1358
Cohen et al.
sample has much less K in the mesostasis than the Luna 16
basalts and the degassing of the mesostasis is masked by high
apparent ages in the lowest temperature steps. This effect is
probably due to uncorrected 36Ar rather than recoil effects,
because a corresponding downturn in apparent age is not seen
in the highest temperature release. A well-defined plateau exists
over 68% of the 39Ar release in the high-temperature steps,
yielding an age of 3224 ~ 7 Ma.
24067,3-6B Fine-Grained Olivine Basalt–This sample is
similar to 24143,4-6L in having a greater proportion of oxides
(ilmenite and Ti-rich spinel) than other samples. The argon-release
spectrum shows evidence for a large amount of uncorrected 36Ar
in the low-temperature steps but a well-defined high-temperature
plateau. The K/Ca ratio shows little evidence of a glassy mesostasis
and correspondingly little degassing. Recoil effects are not evident
in the highest temperature steps. An age of 3279 ~ 8 Ma is derived
from a plateau over 83% of the 39Ar release.
24184,4-6A Gabbro–While EMP analyses were not
performed on this fragment, its argon release spectrum is nearly
identical to that of 24067,3-6B. A plateau over 61% of 39Ar
release yields an apparent age of 3277 ~ 6 Ma.
24143,4-6L Olivine Basalt–This sample is similar to
24067,3-6B in composition and argon-release spectrum. The
K/Ca ratio does not indicate the presence of a K-rich mesostasis,
and recoil effects are not prominent in the highest temperature
steps. The high apparent ages in the low-temperature steps are
probably contributed from 36Ar. A plateau over 65% of the
39Ar release gives an age of 3274 ~ 14 Ma.
24184,4-6B Coarse-Grained Basalt–The predominance
of well-crystallized, large plagioclase grains gives this sample
a uniform K/Ca ratio and flat argon release spectrum, yielding
an age of 3377 ~ 11 Ma from 100% of the 39Ar released.
24184,4-6C Ilmenite-Rich Basalt–This sample is finer
grained than 24184,4-6B, but still coarser grained than the "finegrained" basalts. The Si-rich glassy mesostasis contributes to
argon degassing in the low-temperature steps, but the small size
of this fragment (0.2 mg) limited the number of heating steps,
and therefore, the resolution of the K/Ca spectrum. Though
one heating step degassed >50% of the 39Ar, its apparent age is in
line with the subsequent steps and the plateau is made up of 92%
of the 39Ar release. The apparent plateau age is 3329 ~ 20 Ma.
COSMIC-RAY EXPOSURE AGES
Cosmogenic 38Ar (as well as 36Ar and 37Ar) is produced
by spallation of Ca, K, Ti, and Fe by cosmic rays in the top
∼100 cm of lunar soil in a typical surface production ratio of
1:1:0.1:0.05 (Hohenberg et al., 1978), where 38Ar/36Ar = 1.54
and 37Ar/36Ar >0. Trapped 38Ar is implanted into the soil by
the solar wind, with 38Ar/36Ar = 0.18 and 37Ar/36Ar = 0. Threeisotope correlation plots of argon data, corrected only for blanks,
decay time, and interfering reactions, show that the 38Ar in
these samples is a mixture of cosmogenic and trapped
components (Fig. 9), with no contribution to 38Ar from a Cl-rich
FIG. 9. 38Ar/37Ar/36Ar isochron correlation plots for Luna samples:
(a) Luna 16; (b) Luna 20; and (c) Luna 24. The majority of heating
steps in all samples released 38Ar and 37Ar from a mixture of
cosmogenic (38Ar/36Ar = 1.54) and solar wind (38Ar/36Ar = 0.18)
sources. The slope of the mixing line corresponds to the cosmic-ray
exposure age of each sample.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
source (which would be manifested by 38Ar/36Ar > 1.54). The
low bulk K and Ti contents contribute a negligible amount of
38Ar, but the contribution from Fe may be up to a few percent
in the basalts and gabbros.
The slope of the correlation in the three-isotope plots,
38Ar/ 37Ar, can be directly related to Ca-derived 38Ar (Turner
et al., 1971; Turner, 1972):
38
Ar
= 1.13 × n
Ca
38
37
Ar
× α × J × 7.012 × 10−3
Ar
where 1.13 relates the 38Ar/37Ar ratio to the cosmogenic-only
component, n is the fraction of Ca-only derived 38Ar (taken to
be 97% in the basalts and gabbros; 100% in the troctolites and
anorthites), a is the proportionality factor between 39Ar/37Ar
and K/Ca in the flux monitor (0.545), J is an irradiation
parameter (5.06 × 10 –2), and 7.012 × 10 –3 incorporates
constants of K-decay and unit conversions to make 38Ar/Ca
into units of cc STP/g (Table 2). The 38Ar contributed from Fe
and Ca in pyroxenes is released in the highest temperature steps
(most retentive sites), which generally plot slightly above the
correlation. These points were excluded from the fit to
determine cosmic-ray exposure (CRE) age but were considered
in estimates of the stratigraphic depth, as discussed later.
The CRE age is calculated by multiplying the 38Ar/Ca ratio
by the production function of 38Ar from Ca (P38) on the lunar
surface. A value for P38 of 1.4 × 10–8 cc STP/g Ca/Ma (Turner
et al., 1971) is the most common in determining nominal CRE
ages and is used here in order to compare our calculated values
to those of other samples. In most cases, the calculated CRE
ages of these samples are consistent with previously determined
CRE ages and the CRE age of the bulk soil (Tables 3, 4 and 5).
The assumed value of P38 is a large uncertainty. Measurements
in lunar soils and meteorites (Eugster and Michel, 1995;
Hohenberg et al., 1978) have provided empirical constraints
on P38 vs. depth but the stratigraphic depth of the Luna particles
is not well-known. The production function actually depends
both on depth of irradiation and the target element. Spallationproduced 38Ar from Ca, P38(Ca), initially increases with depth,
as the spallation process produces a cascade of secondary
TABLE 5. 40Ar-39Ar and CRE ages of Luna 24 rocks.
Sample
24067,3-6A
24143,4-6L
24067,3-6B
24184,4-6A
24184,4-6C
24184,4-6B
Luna 24 bulk soil
*Concordant
Type
Age (Ga) ~ 2σ
CRE age (Ma)
dolerite
dolerite
coarse-grained basalt
coarse-grained basalt
fine-grained basalt
coarse-grained basalt
metabasalt
fine-grained basalt
coarse-grained basalt
metabasalt
fine-grained basalt
coarse-grained basalt
fine-grained basalt
metabasalt
coarse-grained basalt
coarse-grained basalt
coarse-grained basalt
coarse-grained basalt
coarse-grained basalt
metabasalt
feldspar from basalt
coarse-grained basalt
fine-grained basalt
coarse-grained basalt
fine-grained basalt
fine-grained basalt
lithic fragment
cataclastic rock
–
2.30 ~ 0.15
2.40 ~ 0.20
2.93 ~ 0.04
3.16 ~ 0.04
3.20 ~ 0.04
3.20 ~ 0.04
3.22 ~ 0.04
3.224 ~ 0.007
3.25 ~ 0.07
3.26 ~ 0.04
3.274 ~ 0.014
3.279 ~ 0.008
3.277 ~ 0.006
3.29 ~ 0.04
03.30 ~ 0.05*
3.329 ~ 0.020
3.33 ~ 0.21
3.34 ~ 0.06
3.36 ~ 0.05
3.36 ~ 0.11
3.37 ~ 0.20
3.377 ~ 0.011
3.39 ~ 0.09
3.43 ~ 0.03
3.49 ~ 0.07
3.52 ~ 0.03
03.60 ~ 0.12†
3.75 ~ 0.07
–
150
400
342
506
249
312
532
1370
308
350
393
618
158
441
215
294
170
196
189
225
217
208
291
426
903
1043
653
15
>300–400
Sm-Nd and 40Ar-39Ar ages on single fragment.
using decay constants of Steiger and Jäger (1977).
†Recalculated
1359
Reference
Shanin et al. (1981)
Shanin et al. (1981)
Burgess and Turner (1998)
Burgess and Turner (1998)
Burgess and Turner (1998)
Burgess and Turner (1998)
Burgess and Turner (1998)
This work
Fugzan et al. (1986)
Schaeffer et al. (1978)
This work
This work
This work
Burgess and Turner (1998)
Wasserburg et al. (1978)
This work
Schaeffer et al. (1978)
Fugzan et al. (1986)
Fugzan et al. (1986)
Hennessy and Turner (1980)
Hennessy and Turner (1980)
This work
Fugzan et al. (1986)
Fugzan et al. (1986)
Fugzan et al. (1986)
Fugzan et al. (1986)
Stettler and Albarede (1978)
Fugzan et al. (1986)
Bogard and Hirsch (1978)
1360
Cohen et al.
particles, but declines with shielding at deeper levels. Because
more energetic cosmic rays are required to produce 38Ar from
Fe, P38(Fe) function declines rapidly with depth. The ratio of
P38 from Fe and Ca has been used (Hennessy and Turner, 1980)
to estimate the depth at which particles received most of their
exposure. A similar calculation was performed for the samples
in this study, where a separate slope was fit to the high-T steps in
the 38Ar/37Ar/36Ar correlation plots, a high-temperature 38Ar/Ca
ratio found and assumed to be from pyroxene, the Fe/Ca ratio for
the sample found from the pyroxene mineralogy (see Mineral
Chemistry and Petrology; 1.52 for Luna 16, 2.2 for Luna 20 and
24), and the P38(Fe)/P38(Ca) calculated. This ratio was compared
with Fig. 5 in (Hennessy and Turner, 1980) to estimate a soil depth
for four Luna 16 basalts and two Luna 24 basalts (Table 2).
The CRE calculations further elucidate some behaviors in
the 40Ar/39Ar spectra. In nearly every sample, the first few
heating steps (up to ∼750 #C) had anomalously high 40Ar/39Ar
ratios. These steps also had anomalously high 38Ar/36Ar ratios
and low 37Ar/36Ar ratios. The excess amount of 38Ar and 40Ar
in the lowest temperature steps is nearly certainly adsorbed
terrestrial atmosphere, contributing argon directly and/or
through addition of adsorbed Cl.
40Ar from all four samples at ∼1 Ga indicates that these samples
DISCUSSION
 −E 
D
D

 2  =  2  exp
 RT 
 a i
 a 0
Volcanism in Mare Fecunditatis
In contrast to the very low-Ti Luna 24 basalts, the four Luna
16 basalts studied contain abundant ilmenite. However, as
pointed out by Semenova et al. (1991), these basalts are not
related to high-Ti basalts from Apollo 11 or 17 by either nearsurface fractionation or melting of similar source regions.
Compared to other mare basalts, the Luna 16 basalts have higher
Al contents. However, the Al contents are not elevated to the
levels suggested by Kurat et al. (1976), as determined by suspect
defocused beam methods. Thus, the need for a source with
affinities to ferroan anorthosites is not warranted. Due to the
high normative olivine content of Luna 16 basalts, it is thought
that the source for the Luna 16 basalts is relatively deep (200–
300 km) and similar in composition to that of the low-Ti basalts
(Steele and Smith, 1972).
The argon release patterns in the basaltic fragments are
typical of lunar basalts (McDougall and Harrison, 1999). The
first degassing is from a very high-K phase, interpreted as a
K-rich, glassy mesostasis. The K/Ca ratio reaches a low,
constant value at about the same heating step as the beginning
of the 39Ar release plateau, indicating that there is an argon
contribution from undisturbed, crystalline plagioclase. Though
the mesostasis was disturbed, the high-temperature plateaus in
samples 1609A and 1609B probably indicate the crystallization
ages for these basalts. In samples 1635B and 1607-1608, more
severe 40Ar loss occurred, and only minimum ages are indicated.
The low-temperature steps indicate a disturbance at ∼1 Ga,
which caused 40Ar loss from each basalt sample. The loss of
were subjected to a thermal event at this time. Comparison of
the extent of calculated 40Ar loss with diffusion models suggests
that the apparent plateau age in these samples is 40–60%
younger than the true age (McDougall and Harrison, 1999),
but this results in a model age older than the age of the solar
system. It is more likely that these samples are not singlesource diffusion domains and that argon loss occurred primarily
from an easily-degassed phase. Following Turner (1971),
Arrhenius plots of the argon data, corrected only for blanks,
decay time, and interfering reactions, were used to further
understand the mechanism of argon loss in these samples. An
effective diffusion parameter, D/a2, was calculated for 39Ar
loss in each step:
D
π 2
Fi ti
=
a 2 36
where Fi is the cumulative fractional loss of 39Ar in step i, and
ti is the cumulative heating time experienced by the sample.
The diffusion parameter D/a2 in each heating step is also related
to the temperature T (estimated from the laser power using a
Stefan–Boltzmann relationship):
where E is the activation energy for the site and R is the gas
constant. A plot of D/a2 vs. 1000/T should show each effective
diffusion domain by a constant slope proportional to E. In the
Luna 16 basalts (Fig. 10), two effective diffusion domains are
evident, one at low temperatures (\700 #C) and one at higher
temperatures. This corroborates the inference that 39Ar loss at
low temperatures is from a high-K glassy mesostasis with
different diffusion parameters than the crystalline plagioclase,
and that the apparently high degree of 40Ar loss is primarily a
result of degassing of this phase and not the entire sample.
The activation energy of the low-temperature domain is
similar in all cases, ∼18 kcal mol–1 K–1. This value is quite
different from the 29.3 kcal mol–1 K–1 suggested by Turner
(1971) for 39Ar loss due to solar heating at the surface of the
Moon. The proximity of two Copernican-age craters, Langrenus
and Taruntius, to the Luna 16 site suggests the basalt fragments
may have been ejected from a young, nearby crater. Simple
shock propagation through a crystalline target fails to cause
complete Ar loss (Deutsch and Schärer, 1994), even at pressures
high enough to convert plagioclase into glass. On the other
hand, shock or burial in an ejecta blanket may be sufficient to
degas a glassy mesostasis.
The four crystalline basalt ages reported here, along with
four additional basalt fragment ages (Cadogan and Turner, 1977;
Fernandes et al., 2000; Huneke et al., 1972), and their CRE
ages, are compiled in Table 3. Each reported age can be
represented as a Gaussian curve, where the width of each
Gaussian is proportional to the uncertainty in the age
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
FIG. 10. Arrhenius plot for Luna 16 basalts. A linear fit was calculated
for the low-temperature (<700 #C) heating steps (solid symbols). The
slope of the line corresponds to the activation energy (E) of the lowtemperature diffusion domain.
determination, and a unit area exists underneath each Gaussian.
Adding the individual Gaussians produces an ideogram, or a
histogram that accounts for measurement uncertainty (Fig. 11).
Two normal distributions were fit to the peaks in Fig. 11 and
are plotted in dashed lines. The two peaks in the ideogram are
made up of (1) two basalts with a combined age of 3347 ~ 24 Ma
and (2) five basalts with a combined age of 3421 ~ 30 Ma.
Both groups are made up of samples dated in multiple
laboratories, strengthening the case that these samples represent
two different flows. The youngest basaltic fragment yields an
age of 3.15 Ga, distinctly younger than the others and possibly
representing a third flow. Because the basalts from this region
have a distinct composition, exotic basalt fragments (emplaced
by ejecta processes from some other region) would be
recognized and do not seem to be represented in this collection
thus far. Therefore, all three basalt age clusters probably
represent volcanism in Mare Fecunditatis from similar sources.
The volcanic period appears to have spanned ∼300 Ma (3.42
to 3.15 Ga), although discontinuously.
The Luna 16 basalt samples have CRE ages
contemporaneous or younger than the 1 Ga 40Ar loss event.
No equivalent stepwise loss of 38Ar is evident, suggesting that
exposure followed the 40Ar loss event. The relatively old CRE
ages and shallow estimated depths of the Luna 16 basalts are
consistent with the suggestion that these samples were affected
by a young (1 Ga) crater, which degassed their easily-disturbed
mesostasis and threw them onto the lunar surface. The Luna
16 basalts with younger CRE ages were probably buried deeper
1361
FIG. 11. Ideogram of crystallization ages in Luna 16 basalts. The
solid curve is a histogram of Gaussian distributions representing each
sample's age and uncertainty. The dashed curves are normal distributions
calculated to fit the ideogram peaks, showing that two basalts have a
combined age of 3347 ~ 24 Ma and five basalts have a combined age of
3421 ~ 30 Ma. Data sources are summarized in Table 2.
in the ejecta blanket and gardened to the near-surface at a later
time. Three of the four Luna 16 basalts were very near the
surface (<5 cm) when they received the bulk of their cosmicray exposure. Coupled with their relatively old CRE ages, this
implies that the Luna 16 regolith is very thin and that not much
unexposed material is available to be gardened.
The two glassy fragments in this study exhibit irregular
release patterns, where recoil effects and the contribution of a
high initial amount of 36Ar are evident in both samples. The
absence of skeletal crystals in the glass and the presence of
many minute, scattered grains of FeNi metal in these samples
makes it probable that these samples are regolith agglutinates
rather than primary basalts. The raggedness of the "plateau"
steps probably reflects K or Ar redistribution in the glassy phase
and uneven release from the embedded grains. In neither sample
does the calculated "plateau" age match the crystallization ages
of the basaltic fragments. In sample 1635A, a young age of
2.8 Ga is recorded, while an interestingly old age of 4.3 Ga
seems to be recorded in sample 1609C. However, the argon
release in an agglutinitic sample reflects contributions from
regolith fragments of all ages, and the combined spectrum is
virtually meaningless in terms of sample age. It is likely that
the difference in ages is due to varying proportions of old
anorthositic material. This is in contrast to impact or volcanic
glass, where clast-free, completely melted samples are known
to yield well-defined Ar plateaus (Culler et al., 2000; Huneke,
1362
Cohen et al.
1978; Spangler et al., 1984). The two glassy Luna 16 samples
have exposure ages >1 Ga, which is to be expected if these
samples are agglutinates which have formed by repeated
micrometeorite bombardment on the lunar surface. No
correlation between 38Ar and a Fe-rich phase was observed,
testifying to the mixed lithologies and grain sizes present.
Magnesium-Suite Plutonism in the Eastern Highlands
The ages derived from the Luna 20 samples are summarized
in Table 4. From the pyroxene and olivine compositions, it
does not appear that any of these samples are fragments of the
same rock. However, the two troctolites have identical 40Ar-39Ar
ages of 4.19 Ga. This appears to be the crystallization age for
both samples. The 2004B gabbro and 2002 anorthite grain
also have identical argon-release profiles (after the first 20%
of 39Ar released) and ages of 4.11 Ga, probably indicating that
this anorthite grain was originally part of this gabbro. The
second anorthite grain is the oldest of this set, at 4.265 Ga. A
similar age was reported by Cadogan and Turner (1977) on a
"group of 5 anorthositic fragments," which they interpreted as
incomplete degassing of primary anorthositic crust in the
Crisium impact. Huneke and Wasserburg (1979) reported
degassing ages of 4.36 and 4.17 Ga on another anorthositic
fragment, while the estimated K-Ar age of the bulk Luna 20
soil is also 4.2 Ga (Eugster et al., 1975). Sample 2004A and
those in the Cadogan and Turner (1977) study were too small
to determine chemistry in order to classify them as ferroan
anorthosites (FANs). The fact that anorthositic grain 2002
appears to be a piece of the same rock as gabbro 2004B means
that one cannot rule out a similar origin for other anorthositic
grains. However, the argon spectrum of 2004A does show
disturbance (Fig. 7f), and an extrapolation of the downturned
trend (from 15–40% of 39Ar released) intersects at ∼3.9 Ga,
the inferred age of the Crisium impact as determined by Luna
20 impact melt rocks (Podosek et al., 1973; Swindle et al.,
1991). Therefore, we favor the interpretation that plagioclase
grain 2004A is an incompletely outgassed anorthosite fragment,
similar to the interpretations of Cadogan and Turner (1977)
and Huneke and Wasserburg (1979). On the other hand, the
troctolite ages of 4.19 Ga indicate the onset of Mg-suite
plutonism on the eastern limb of the Moon.
The Luna 20 soil has been extensively examined for pieces
of Crisium basin impact melt, though they appear to be rare in
the returned sample. For completeness, 40Ar-39Ar ages of Luna
20 impact melt rocks (Cadogan and Turner, 1977; Huneke and
Wasserburg, 1979; Podosek et al., 1973; Swindle et al., 1991)
are included in Table 4. As is the case for other sites on the
lunar nearside, the impact melt ages cluster around 3.9 Ga,
indicating that the mechanism responsible for creating this 3.9 Ga
impact melt, possibly a lunar cataclysmic bombardment, was
operative on the Moon's east limb as well as at the Apollo
landing sites, making the far side the most likely place to find
impact melts with more diverse ages (Cohen et al., 2000).
The ideogram of Luna 20 highlands rocks ages (excluding
impact melts) is shown in Fig. 12. The older samples are
incompletely degassed anorthosites. The range for later
plutonism in this area extends over a wide range, from 4.19 to
4.02 Ga. The depth of the Luna 20 samples could not be inferred
because of their low Fe content. Though some impact melt
rocks can have exposure ages >1 Ga, the CRE ages of Luna 20
plutonic rocks tend to cluster around the bulk soil age of 350 Ma.
In contrast, the Apollo 16 highlands soil CRE age is 125 Ma
(Bogard and Hirsch, 1976), implying a faster gardening rate at
the Apollo 16 site than in the eastern highlands.
There is a trend in age from troctolite to anorthositic gabbro
to more normal gabbro compositions and an accompanying
change to lower Mg#, which may reflect either evolution of a
single source magma for this suite of rocks or melting at
progressively shallower depths, as per cumulate source models
for the lunar upper mantle (Snyder et al., 2000). In particular,
the chemistry of troctolite sample 2003 is extremely primitive,
and a discussion of its petrogenesis is warranted.
Spinel troctolites and cataclastites are thought to represent
samples of the deep crust or mantle. Bence et al. (1974)
discovered an assemblage in sample 73263,1,11 which included
aluminous enstatite, forsteritic olivine (Fo89–91) and pink
pleonaste (Cr# = 0.09 and Mg# = 0.80). Based upon
experimental studies (Kushiro and Yoder, 1966) that showed
that the Al-enstatite-spinel join breaks down to forsteritecordierite at 3 kbar pressure, Bence et al. (1974) estimated that
FIG. 12. Ideogram of crystallization ages in Luna 20 highland rocks
(excluding impact melts). The solid curve is a histogram of Gaussian
distributions representing each sample's age and uncertainty. Data
sources are summarized in Table 3.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
this assemblage would be stable at that pressure in the Moon,
or roughly 60 km deep (near the base of the crust on the
nearside). Dymek et al. (1976) reported two similar spinel
cataclasite clasts in sample 72435, one containing a 30 lm
inclusion of cordierite. The spinels were classified as pleonastes
with compositions ranging from Mg# = 0.44–0.69 and Cr# =
0.03–0.21. Although these samples contained much less
primitive spinel and olivine (Fo71–72) and no aluminous
enstatite, Dymek et al. (1976) postulated that these clasts could
also have been excavated from deep within the lower crust.
Anderson (1973) described a peridotite (15445,10) that
contained the assemblage enstatite-plagioclase-spinel-forsterite
and estimated that this sample came from 40 km deep within
the crust. Snyder et al. (1998) reported a spinel troctolite in a
sample from drive tube 68001/68002. The presence of
pleonaste, with Mg# = 0.65 and Cr# = 0.08, and forsteritic
olivine (Fo75–81) suggested a deep-seated origin. However, neither
cordierite nor Al-enstatite were discovered in this sample.
Roedder and Weiblen (1973) reported a spinel troctolite
fragment from the 250–500 lm fraction of Luna 20 soil.
Although spinel troctolite 2003 has a similar plagioclase
composition (An95–97), it is dissimilar in having higher-Mg
olivine (Fo95–96 vs. Fo91). The spinel compositions in spinel
troctolite 2003 are more magnesian (Mg# = 91–92) than those
analyzed for other Luna 20 samples (Mg# = 82; Roedder and
Weiblen, 1973) and the most Mg-enriched, chromian hercynitic
spinels (Mg# <63) from Apollo 14 (Roedder and Weiblen, 1972;
Steele, 1972). Thus, it appears that spinel troctolite 2003 may
be the most compositionally primitive fragment yet analyzed
from the lower crust of the Moon.
1363
discovery of a quench-textured olivine vitrophyre (24067,3-6A),
with Mg-enriched pyroxene and olivine (up to Fo64), is the
most compositionally primitive Luna 24 basalt analyzed to date.
The ages for the six fragments in this study are within the
range of ages found in other Luna 24 basalt fragments by
40 Ar-39Ar and Sm-Nd techniques (Burgess and Turner, 1998;
Fugzan et al., 1986; Hennessy and Turner, 1980; Schaeffer et
al., 1978; Shanin et al., 1981; Stettler and Albarede, 1978;
Wasserburg et al., 1978). The ages of the 27 basaltic and lithic
fragments dated thus far are summarized in Table 5. When
plotted in an ideogram, which takes into account the uncertainty
in each age, 19 of these fragments fall into a normal distribution
with an age of 3273 ~ 83 Ma (Fig. 13). The number of samples
within this group is significant, but the range within the
distribution spans 160 Ma. This range may indicate either that
this single flow was extruded over a ∼200 Ma timespan (an
unlikely scenario), or that this single group is showing an
artificial range in ages due to differences in individual laboratory
procedures and data reduction. The fragment ages from each
study tend to group together instead of spanning the entire 160 Ma
range. For instance, the shoulder at the older end of the primary
distribution has an age of 3466 ~ 77 Ma derived from four
samples in a single study (Fugzan et al., 1986) and may not
actually be a separate group. The samples at the extreme high
and low ends of the range were poorly described and may not
Volcanism in Mare Crisium
The coarser grained (slower cooled) basalts extend the
pyroxene chemistry of the Luna 24 basalts to much more
Fe-enriched compositions (Fig. 5a), as supported by previous
studies (Nielsen and Drake, 1978; Papike and Vaniman, 1978).
Finer grained basalts show much less variation in pyroxene
chemistry. The first pyroxene to crystallize in the Luna 24
basalts is much more Ca- and Fe-rich than those from Apollo
17 VLTs (which are thought to be from a similar source; see
also Papike and Vaniman, 1978). Coish and Taylor (1978) have
shown that, if the chemistry of monomineralic fragments is
included, the range in pyroxene chemistry is similar to A17
VLTs. However, they also indicate that these more Mg-rich
and Ca-poor pyroxenes could be derived from highland
material.
Nowhere else on the Moon (from lunar meteorites and
returned collections) do we find basalts that are as chemically
evolved as those from Mare Crisium. Whereas olivines
analyzed from Apollo mare basalts rarely fall below Fo30
(Papike and Vaniman, 1978), those from Luna 24 are routinely
this Fe-rich (Fig. 5b). Coish and Taylor (1978) reported olivine
vitrophyres with more Mg-rich compositions. However, our
FIG. 13. Ideogram of crystallization ages in Luna 24 basalts. The
solid curve is a histogram of Gaussian distributions representing each
sample's age and uncertainty. The dashed curves are normal
distributions calculated to fit the ideogram peaks, showing that
nineteen samples have an age of 3273 ~ 83 Ma and four have an age
of 3466 ~ 77 Ma. The older group may not have a real separate age,
as discussed in the text. Data sources are summarized in Table 4.
1364
Cohen et al.
be basalts at all (Shanin et al., 1981; Stettler and Albarede,
1978). All the samples in the 3.27 Ga cluster, for which data
were available, were classified as VLT basalts, though specific
chemistry was usually not reported, making detailed
comparisons impossible. Burgess and Turner (1998) recognized
the single sample with an age of 2.93 Ga as a real outlier (i.e.,
not a problem with monitoring or analysis) and representative
of a younger flow or more extensive thermal resetting.
We favor the interpretation that the Luna 24 core primarily
returned multiple fragments of the same basalt flow at ∼3.27 Ga
within the Crisium basin. This flow was extruded over \200 Ma,
though the exact time and duration is unclear. Grove and
Vaniman (1977) point out that the Apollo 17 and Luna 24 basalts
cannot be related by fractional crystallization. This
interpretation is supported by the difference in age between the
two groups. The tight cluster of Luna 24 basalt ages at 3.27 Ga
contrasts strongly with the much older (3.6–3.7 Ga) ages of the
Apollo 17 basalts (Taylor, 1982). A 400 Ma hiatus in eruption
of a fractionally evolving magma seems quite unlikely. Instead,
the Apollo 17 and Luna 24 VLT basalts may be derived from
similar, but perhaps unrelated, source regions.
The two Luna 24 samples whose depths could be inferred
received their exposure at deeper levels (^25 cm) than the Luna
16 samples. No similar degassing event is indicated in these
samples. The Luna 24 basalt fragments also have overall shorter
CRE ages than both the Luna 16 samples and the Luna 24 bulk
regolith. This suggests that the Luna 24 regolith is deeper than
the Luna 16 regolith, and that vigorous gardening of the basalt
flow itself may be the mechanism for bringing the Luna 24
samples to the surface.
SUMMARY
The Luna 16 basalts have relatively high Al contents but
high normative olivine content, indicating that the source for
the Luna 16 basalts is relatively deep (200–300 km) and similar
in composition to that of the low-Ti basalts. Volcanism within
Mare Fecunditatis appears to have occurred in at least three
different eruptions, though the basalts in each eruption are
chemically similar to each other. The volcanic period appears
to have spanned ∼300 Ma (3.42 to 3.15 Ga). All four basalt
fragments in this study were affected by an intense shock event
∼1 Ga ago. Given the location of the samples between two
Copernican-age craters, it seems probable that these fragments
were emplaced into the Luna 16 landing site area by one of
these young impact events, causing partial 40Ar loss from the
samples. The shallow depth (∼5 cm) and relatively old CRE
ages of the samples and soil imply that the Luna 16 regolith is
thin and turnover is slow, consistent with an interpretation that
the Luna 16 module landed on a young ejecta blanket with
limited regolith development.
The chemistry and age sequence of the Luna 20 samples
2004B, 2004C, and 2004D suggests an evolving magma
source(s) at the Luna 20 site from 4.19 to 4.02 Ga that first
produced troctolites, then anorthositic gabbro, and then gabbro.
The An# of the plagioclase in these samples remains relatively
constant while the olivine becomes fayalitic and the pyroxene
pigeonitic. This sequence may reflect fractional crystallization
in the parent magma or progressively shallower melting of the
source region. The CRE ages of Luna 20 plutonic rocks are
generally the same age or younger than the Luna 20 soil age,
but older than the 125 Ma CRE age of the Apollo 16 highlands
site, implying that the gardening rate is slower in the eastern
highlands than at the Apollo 16 site.
Spinel troctolite 2003 appears to be the most compositionally
primitive fragment from the lower crust in all the Apollo and
Luna collections. Although this sample is one of the oldest
Mg-suite rocks in the Luna 20 collection, it has a much more
magnesian composition than 2004C troctolite, albeit a similar
age. The divergent chemistry of these two fragments seems to
indicate that the Mg-suite rocks in this area are derived from
more than one source, but that these sources were mobilized in
the area at the same time.
Luna 24 basalt 24067,3-6A, with Mg-enriched pyroxene
and olivine, is the most compositionally primitive Luna 24 basalt
analyzed to date. This sample could be an intermediate in the
fractionation of the Apollo 15 green glass to form typically
more evolved Luna 24 basalts. The majority of Luna 24 basalt
fragments dated thus far, including those in this study, have
ages of ∼3.27 Ga, in contrast to the much older (3.6–3.7 Ga)
ages of the Apollo 17 basalts, suggesting that the Apollo 17
and Luna 24 VLT basalts come from similar, but perhaps
unrelated, source regions. The similarity in age of the dated
basalt fragments indicates that the majority of basalts in the
Luna 24 collection are derived from the same flow within Mare
Crisium. The Luna 24 basalt fragments have shorter CRE ages
than both Luna 16 and the Luna 24 bulk regolith, suggesting
that the Luna 24 regolith is deeper and has been overturned
more than at Luna 16. Gardening of the basalt flow may be the
mechanism for bringing the Luna 24 samples to the surface.
Acknowledgements–We thank Eric Galimov and the Vernadsky Institute
for loan of the Luna samples. Allen Patchen was instrumental in obtaining
the EMP analyses. Careful reviews by O. Eugster and G. Turner and
discussions with V. Fernandes greatly improved the manuscript. This
work was supported by NASA grants NAG5-8154 and NAG5-9158 and
made use of NASA's Astrophysics Data System Abstract Service.
Editorial handling: U. Krähenbühl
REFERENCES
AGRAWAL J. K., GOPALAN K. AND RAO M. N. (1974) Solar wind and
cosmogenic rare gases in Luna 16 and 20 soils and their correlation
with cosmic ray produced fossil tracks in lunar samples. Pramana
3, 176–185.
A NDERSON A. T. (1973) The texture and mineralogy of lunar
peridotite, 15445,10. J. Geol. 81, 219–266.
BENCE A. E., DELANO J. W., PAPIKE J. J. AND CAMERON K. L. (1974)
Petrology of the highlands massifs at Taurus–Littrow—An analysis
of the 2–4 mm soil fraction. Proc. Lunar Sci. Conf. 5th, 785–827.
Argon-40–argon-39 chronology and petrogenesis along the eastern limb of the Moon
BOGARD D. D. AND HIRSCH W. C. (1976) Noble gases in 60009–
60010 drive tube samples—Trapped gases and irradiation history.
Proc. Lunar Sci. Conf. 7th, 259–279.
BOGARD D. D. AND HIRSCH W. C. (1978) Noble gases in Luna 24
core soils. In Mare Crisium: The View from Luna 24 (eds. R. B.
Merrill and J. J. Papike), pp. 105–116. Pergamon Press, New York,
New York, USA.
BURGESS R. AND TURNER G. (1998) Laser argon-40–argon-39 age
determinations of Luna 24 mare basalts. Meteorit. Planet. Sci.
33, 921–935.
BUSSEY D. B. J. AND SPUDIS P. D. (2000) Compositional studies of
the Orientale, Humorum, Nectaris, and Crisium lunar basins. J.
Geophys. Res. 105, 4235–4243.
CADOGAN P. H. AND TURNER G. (1977) 40Ar-39Ar dating of Luna 16
and Luna 20 samples. Phil. Trans. Royal Soc. London A284,
167–177.
COHEN B. A., SWINDLE T. D. AND KRING D. A. (2000) Support for
the lunar cataclysm from lunar meteorite impact melt ages.
Science 290, 1754–1756.
COISH R. A. AND TAYLOR L. A. (1978) Mineralogy and petrology of
basaltic fragments from the Luna 24 drill core. In Mare Crisium:
The View from Luna 24 (eds. R. B. Merrill and J. J. Papike), pp.
403–417. Pergamon Press, New York, New York, USA.
CULLER T. S., BECKER T. A., MULLER R. A. AND RENNE P. R. (2000)
Lunar impact history from 40Ar/39Ar dating of glass spherules.
Science 287, 1785–1788.
DEUTSCH A. AND SCHÄRER U. (1994) Dating terrestrial impact events.
Meteoritics 29, 301–322.
DYMEK R. F., ALBEE A. L. AND CHODOS A. A. (1976) Petrology and
origin of Boulders No. 2 and No. 3, Apollo 17 Station 2. Proc.
Lunar Sci. Conf. 7th, 2335–2378.
EUGSTER O. AND MICHEL T. (1995) Common asteroid break-up events
of eucrites, diogenites, and howardites and cosmic-ray production
rates for noble gases. Geochim. Cosmochim. Acta 59, 177–199.
EUGSTER O., GROEGLER N., MENDIA M. D., EBERHARDT P. AND GEISS
J. (1973) Trapped solar wind noble gases and exposure age of
Luna 16 lunar fines. Geochim. Cosmochim. Acta 37, 1991–2003.
EUGSTER O., EBERHARDT P., GEISS J., GROEGLER N., JUNGCK M.
AND MOERGELI M. (1975) Solar-wind-trapped and cosmic-rayproduced noble gases in Luna 20 soil. Proc. Lunar Sci. Conf.
6th, 1989–2007.
FERNANDES V. A., BURGESS R. AND TURNER G. (2000) Laser argon40–argon-39 age studies of Dar al Gani 262 meteorite. Meteorit.
Planet. Sci. 35, 1355–1364.
FUGZAN M. M., GANG V. M., TARASOV L. S., KOLESOV G. M. AND
SHUKOLYUKOV Y. A. (1986) 40Ar-39Ar dating of lunar rocks from
Mare Crisium. Geokhimiia 4, 469–479.
GROVE T. L. AND VANIMAN D. T. (1977) Experimental petrology of very
low Ti basalts and origin of Luna 24 ferrobasalt. In Papers Presented
to the Conference on Luna 24; A Lunar Science Institute Topical
Conference (eds. J. J. Papike, A. L. Albee, T. R. McGetchin, D. S.
McKay, R. B. Merrill, L. E. Nyquist, C. Pieters and P. H. Schultz),
pp. 68–71. Lunar Science Institute, Houston, Texas, USA.
HEAD J. W., ADAMS J. B., MCCORD T. B., PIETERS C. AND ZISK S.
(1978) Regional stratigraphy and geologic history of Mare
Crisium. In Mare Crisium: The View from Luna 24 (eds. R. B.
Merrill and J. J. Papike), pp. 43–74. Pergamon Press, New York,
New York, USA.
HENNESSY J. AND TURNER G. (1980) 40Ar-39Ar ages and irradiation
history of Luna 24 basalts. Phil. Trans. Royal Soc. London A297,
27–39.
HOHENBERG C. M., MARTI K., PODOSEK F. A., REEDY R. C. AND
SHIRCK J. R. (1978) Comparisons between observed and predicted
cosmogenic noble gases in lunar samples. Proc. Lunar Planet.
Sci. Conf. 9th, 2311–2344.
1365
HUNEKE J. C. (1978) 40Ar-39Ar microanalysis of single 74220 glass
balls and 72435 breccia clasts. Proc. Lunar Planet. Sci. Conf.
9th, 2345–2362.
HUNEKE J. C. AND WASSERBURG G. J. (1979) Sliva IZ Piroga: K/Ar
evidence from Luna 20 rocks for lunar differentiation prior to
4.51 AE ago (abstract). Lunar Planet. Sci. 10, 598–600.
HUNEKE J. C., PODOSEK F. A. AND WASSERBURG G. J. (1972) Gas
retention and cosmic-ray exposure ages of a basalt fragment from
Mare Fecunditatis. Earth Planet. Sci. Lett. 13, 375–383.
KAISER W. A. (1972) Rare gas studies in Luna-16-G-7 fines by
stepwise heating technique, a low fission solar wind Xe. Earth
Planet. Sci. Lett. 13, 387–399.
KURAT G., KRACHER A., KEIL K., WARNER R. AND PRINZ M. (1976)
Composition and origin of Luna 16 aluminous mare basalts. Proc.
Lunar Planet. Sci. Conf. 7th, 1301–1321.
K USHIRO I. AND Y ODER H. S. (1966) Anorthite-forsterite and
anorthite-enstatite reactions and their bearing on the basalteclogite transformation. J. Petrology 7, 337–362.
MCCAULEY J. F. AND SCOTT D. H. (1972) The geologic setting of
the Luna 16 landing site. Earth Planet. Sci. Lett. 13, 225–232.
MCDOUGALL I. AND HARRISON T. M. (1999) Geochronology and
Thermochronology by the 40Ar/39Ar Method. Oxford University
Press, New York, New York, USA. 269 pp.
NIELSEN R. L. AND DRAKE M. J. (1978) The case for at least three
mare basalt magmas at the Luna 24 site. In Mare Crisium: The
View from Luna 24 (eds. R. B. Merrill and J. J. Papike), pp.
419–428. Pergamon Press, New York, New York, USA.
PAPIKE J. J. AND VANIMAN D. T. (1978) Luna 24 ferrobasalts and the
mare basalt suite: Comparative chemistry, mineralogy, and
petrology. In Mare Crisium: The View from Luna 24 (eds. R. B.
Merrill and J. J. Papike), pp. 371–401. Pergamon Press, New
York, New York, USA.
PODOSEK F. A., HUNEKE J. C., GANCARZ A. J. AND WASSERBURG G. J.
(1973) The age and petrography of two Luna 20 fragments and
inferences for widespread lunar metamorphism. Geochim.
Cosmochim. Acta 37, 887–904.
RODDICK J. C. (1983) High precision intercalibration of 40Ar-39Ar
standards. Geochim. Cosmochim. Acta 47, 887–898.
ROEDDER E. AND WEIBLEN P. W. (1972) Occurrence of chromian,
hercynitic spinel ("pleonaste") in Apollo-14 samples and its
petrologic implications. Earth Planet. Sci. Lett. 15, 376–402.
ROEDDER E. AND WEIBLEN P. W. (1973) Petrology of some lithic
fragments from Luna 20. Geochim. Cosmochim. Acta 37,
1031–1052.
SCHAEFFER O. A., BENCE A. E., EICHHORN G., PAPIKE J. J. AND
VANIMAN D. T. (1978) 39Ar-40Ar and petrologic study of Luna
24 samples 24077,13 and 24077,63. Proc. Lunar Planet. Sci.
Conf. 9th, 2363–2373.
SEMENOVA A. S., KONONKOVA N. N. AND SOLOV'EVA N. V. (1991)
Mineral and chemical composition of Luna-16 basalts.
Geokhimiia 5, 677–688.
SHANIN L. L., ARAKELIANTS M. M., BOGATIKOV O. A., IVANENKO V. V.,
PUPYREV I. G., TARASOV L. S. AND FRIKH-KHAR F. I. (1981)
Dating samples of lunar soil from the Mare Crisium by the
Ar39-Ar40 technique. Geokhimiia 7, 970–980.
SNYDER G. A. AND TAYLOR L. A. (1996) Apollo 14 high-Al basalts
may not represent earliest volcanism on the Moon: The enigma
of disturbed Sm-Nd and undisturbed Rb-Sr isotopic systems
(abstract). Lunar Planet. Sci. 27th, 1233.
SNYDER G. A., HALL C. M., LEE D-C., TAYLOR L. A. AND HALLIDAY
A. N. (1996) Earliest high-Ti volcanism on the Moon: 40Ar-39Ar,
Sm-Nd, and Rb-Sr isotopic studies of group D basalts from the
Apollo 11 landing site. Meteorit. Planet. Sci. 31, 328–334.
SNYDER G. A., RUZICKA A. M., TAYLOR L. A. AND PATCHEN A. D.
(1998) Journey to the center of the regolith; a spinel troctolite
1366
Cohen et al.
and other clasts from drive tube 68001 (abstract). Lunar Planet.
Sci. 29, #1144, Lunar and Planetary Science Institute, Houston,
Texas, USA (CD-ROM).
SNYDER G. A., BORG L. E., NYQUIST L. E. AND TAYLOR L. A. (2000)
Chronology and isotopic constraints on lunar evolution. In Origin
of the Earth and Moon (eds. R. M. Canup and K. Righter), pp.
361–396. Univ. Arizona Press, Tucson, Arizona, USA.
SPANGLER R. R., WARASILA R. AND DELANO J. W. (1984) Ar-39-Ar40 ages for the Apollo 15 green and yellow volcanic glasses.
Proc. Lunar Planet. Sci. Conf. 14th, B487–B497.
STEELE I. M. (1972) Chromian spinels from Apollo 14 rocks. Earth
Planet. Sci. Lett. 14, 190–194.
STEELE I. M. AND SMITH J. V. (1972) Mineral and bulk composition
of three fragments from Luna 16. Earth Planet. Sci. Lett. 13,
323–327.
S TEIGER R. H. AND J AGER E. (1977) Subcommission on
Geochronology: Convention on the use of decay constants in geoand cosmochronology. Earth Planet. Sci. Lett. 36, 359–362.
STETTLER A. AND ALBAREDE F. (1978) 39Ar-40Ar systematics of two
sub-millimetre rock fragments from Mare Crisium. Earth Planet.
Sci. Lett. 38, 401–406.
SWINDLE T. D., SPUDIS P. D., TAYLOR G. J., KOROTEV R. L. AND
NICHOLS J. (1991) Searching for Crisium Basin ejecta—Chemistry
and ages of Luna 20 impact melts. Proc. Lunar Planet. Sci. Conf.
21st, 167–181.
TAYLOR G. J., WARREN P., RYDER G., DELANO J., PIÉTERS C. AND
LOFGREN G. (1991) Lunar rocks. In Lunar Sourcebook: A User’s
Back
Guide to the Moon (eds. G. H. Heiken, D. T. Vaniman and B. M.
French), pp. 183–284. Cambridge Univ. Press, New York, New
York, USA.
TAYLOR L. A., SHERVAIS J. W., HUNTER R. H., SHIH C-Y., BANSAL
B. M., WOODEN J., NYQUIST L. E. AND LAUL L. C. (1983) Pre4.2 AE mare-basalt volcanism in the lunar highlands. Earth
Planet. Sci. Lett. 66, 33–47.
TAYLOR S. R. (1982) Planetary Science: A Lunar Perspective. Lunar
and Planetary Institute, Houston, Texas, USA. 481 pp.
TURNER G. (1971) 40Ar-39Ar ages from the lunar maria. Earth Planet.
Sci. Lett. 11, 169–191.
TURNER G. (1972) 40Ar-39Ar age and cosmic ray irradiation history
of the Apollo 15 anorthosite, 15415. Earth Planet. Sci. Lett. 14,
169–175.
TURNER G., HUNEKE J. C., PODOSEK F. A. AND WASSERBURG G. J.
(1971) 40Ar-39Ar ages and cosmic-ray exposure ages of Apollo
14 samples. Earth Planet. Sci. Lett. 12, 19–35.
WARREN P. H., KALLEMEYN G. W. AND KYTE F. T. (1997) Siderophile
element evidence indicates that Apollo 14 high-Al mare basalts
are not impact melts (abstract). Lunar Planet. Sci. 28, #1738,
Lunar and Planetary Institute, Houston, Texas, USA (CD-ROM).
WASSERBURG G. J., BROZOLO F. R. D., PAPANASTASSIOU D. A.,
MCCULLOCH M. Y., HUNEKE J. C., DYMEK R. F., DEPAOLO D. J.,
CHODOS A. A. AND ALBEE A. J. (1978) Petrology, chemistry,
age, and irradiation history of Luna 24 samples. In Mare Crisium:
The View from Luna 24 (eds. R. B. Merrill and J. J. Papike), pp.
657–678. Pergamon Press, New York, New York, USA.
Forward
Table of Contents