4.32 Billion Year Old Impact Melts at Apollo 14: Dating the Procellarum Basin?

Introduction: Recent work has improved our understanding
of lunar crustal structure and basin evolution
on the Moon, but its early impact history and implications
for solar system dynamics remain poorly established
[1,2]. In particular, there is a derth of absolute
ages, especially for the period >4.0 Ga, that could be
related to basin-forming events. Here we present U-Pb
isotopic data for Zr-rich minerals found in impact-melt
fragments from Apollo 14 soil sample 14163 that, when
combined with previously published data, dates an impact
event at 4324±15 Ma. The compositions and ages
of these impact-melt fragments provide unique information
about the timing of early impact events and the
composition of the lunar crust.
Results: Fourteen rocklets ranging in size from 1 to
3 mm were extracted from soil 14163. Most are impactmelt
rocks with 10-30% clasts of mostly pyroxene and
plagioclase (100 to 500 μm) in a crystalline matrix. Textures
of the matrix vary from subophitic, formed by intergrowths
of 10-20 μm plagioclase and pyroxene crystals,
to poikilitic, with plagioclase and pyroxene reaching
20-50 μm size. Some fragments that contain no visible
clasts have similar textures are also interpreted as
impact-melt rocks.
All fragments contain notably large proportions of
euhedral to subhedral ilmenite, zircon, apatite/merrillite
and less abundant zirconolite and baddeleyite in the
melt matrices. These grains often form intricate intergrowths
with each other and rock-forming minerals, indicating
their crystallization from the melt (Fig. 1).
However, some slightly larger (~50 μm) zircon and
phosphate grains can be interpreted as relict clasts based
on their granular textures and relationships with the surrounding
phases (Fig. 1). This implies that zircon and
phosphate minerals were present in the target rocks.
Some of these grains were profoundly remelted during
the impact, which resulted in an oversaturation of the
melt in Zr and P and crystallization of new grains of Zrrich
minerals and phosphates during melt solidification.
The presence of zircon and phosphates indicates that the
melts were significantly enriched in KREEP components
and ties their provenance to the Procellarum-
KREEP Terrane (PKT) [3].
Combined U-Pb zircon data obtained for all fragments
form two main clusters on a concordia diagram
(Fig. 2), between about 4.3 and 3.9 Ga. Phosphate data
mostly concentrate on the younger end of this age range
but some analyses are almost as old as the older zircon
grains (Fig. 2).
Figure 1: Zircon (Zr) and phosphate (Apt; Mer) grains in impact
melt fragments from Apollo 14 soil 14161. A-zircon
grains crystallized from impact melt; B-phosphate grain inherited
from the target; C- granular zircon grains
Our interpretation of these data is that all U-rich
minerals experienced variable resetting of the U-Pb system,
with phosphate, where closure temperature is significantly
lower than that in zircon, affected more profoundly
by Pb loss than zircon. Our best estimate of the
time of formation of zircon and phosphate is based on
statistically valid analysis of grains interpreted
texturally as grown from the impact melt (Fig. 2). Ten
of these analyses define an age of 4324±15 Ma
(MSWD=3.0, probability of fit P=0.002).
Our best estimate for the time of resetting is obtained
by combining data from phosphates that are statistically
indistinguishable from 3.9 Ga within the analytical uncertainties.
This group is represented by 18 analyses of
15 phosphate grains from different fragments (Fig. 2)
and defines an age of 3922±6 Ma (MSWD=1.2,
P=0.23). Conversely, a minimum age of the target lithologies,
remelted in the impact that produced the rocklets
studied here (Fig. 2), can be determined from the
five oldest analyses of zircon clasts at 4338±13 Ma
(MSWD=1.5, P=0.2), which is indistinguishable from
the age of the impact melt within the uncertainties. The
obtained ages, combined with textural evidence, imply
that the impact melt was formed at 4324±15 Ma, and
that it occurred in a zircon-rich target with a minimum
age of 4338±13 Ma. Further reworking occurred during
a second impact event at 3922±6 Ma.
Discussion: The younger age of 3922±6 Ma can be
interpreted as the time of the Imbrium impact. The older
age of 4324±15 Ma would then be the time of formation
of the impact melt, which was then caught in the Imbrium
ejecta either at the Apollo 14 landing site or
within the target rocks of the Imbrium impact.
The presence of abundant accessory phases such as
zircons and phosphates is consistent with a substantial
KREEP component in the analyzed particles. The current
distribution of KREEP on the lunar surface appears
to be strongly influenced by Imbrum ejecta [4] so interpretation
of the 4.32 Ga age depends in part on assumed
structure of the pre-impact crust and distribution of
KREEP within the crust.
If KREEP was present only in the deep crust at 4.32
Ga, then a basin-scale impact possibly analogous to Imbrium
or larger may be necessary to excavate a KREEPy
impact melt at this time. However, if KREEPy materials
were present closer to the surface perhaps due to redistribution
related to Mg-suite magmatism, then smaller
impacts might be able to rework KREEPy compositions
at shallower depths. The coherence of the data on the
particles analysed here suggests a large volume of melt
that has been preserved since 4.32 Ga, consistent with a
large impact event. Relict zircons and mineral clasts
suggests that the igneous crust in the vicinity of this impact
was well developed by at least 4.34 Ga, similar to
the model age of KREEP and older than the isochron
ages of many Mg-suite cumulates [5].
A problem that confronts all lunar sample studies using
the current collection is that the pre-Imbrium geology
of the PKT (the source of these 4.32 Ga impact
melts) is not well constrained. The South Pole-Aitken
basin contains regions that are moderately enriched in
Th, but its ejecta is Th-poor [6]; therefore these A14
fragments are probably not SPA ejecta. Alternatively,
these fragments may represent formation of a hypothesized
Procellarum basin [7] although the lack of a
clearly defined basin ejecta signature is a potential problem
with that interpretation. In any case, they provide a
younger limit on the age of lunar differentiation and formation
of KREEP within the lunar crust.
Figure 2: U-Pb data for zircon and phosphate grains from impact
melt fragments. A-all data; B-data used for age calculations.
References: [1] Orgel C., Michael G., Fassett C. I.,
van der Bogert C. H., Riedel C., Kneissl T., and Hiesinger
H. (2018) J. Geophys. Res. Planets 123, 748-
762. [2] Evans A. J., Andrews-Hanna J. C., Head J.
W., Soderblom J. M., Solomon S. C., and Zuber M. T.
(2018) J. Geophys. Planets. 123, 1596-1617. [3] Jolliff,
B.L., Gillis, J.J., Haskin, L.A., Korotev, R.L. and
Wieczorek, M.A. (2000) J. Geophys. Res: Planets 105,
4197-4216. [4] Haskin L. A. (1998) J. Geophys. Res.
Planets 103, 1679-1689. [5] Borg L.E., Gaffney A.M.,
and Shearer C.K. (2015) MAPS 50, 715-732. [6] Moriarty,
D.P., Watkins, R.N., Valencia, S.N., Kendall,
J.D., Evans, A.J., Dygert, N. and Petro, N.E. (2021) J.
Geophys. Res. Planets 126. [7] Zhu, M.H., Wünnemann,
K., Potter, R.W., Kleine, T. and Morbidelli, A.
(2019) J. Geophys. Res: Planets 124, 2117-2140