Petrogenesis of mare basalts: A record of lunar volcanism

Abstract

Returned rock and soil samples from our nearest planetary neighbor have provided the basis for much of our understanding of the origin and evolution of the Moon. Of particular importance are the mare basalts, which have revealed considerable information about lunar volcanism and the nature of the mantle, as well as post-magma-generation processes. This paper is a critical review of the petrogenetic models for the generation of mare basalts formulated over the last twenty years. We have used all available mare basalt analyses to define a six-fold classification scheme using TiO 2 contents as the primary division (i.e., ∗ 1 wt% = very low-Ti or VLT; 1–6 wt% = low-Ti; > 6 wt% = high-Ti). A secondary division is made using Al 2O 3 contents (i.e., ∗ 11 wt% = low-Al; > 11 wt% = high-Al), and a tertiary division is defined using K contents (i.e., ∗ 2000 ppm = low-K; > 2000 ppm = high-K). Such divisions and subdivisions yield a classification containing twelve categories, of which six are accounted for by the existing Apollo and Luna collections. Therefore, we present our discussions in the form of six mare basalt rock types: 1. (1) high-Ti/low-Al/low-K (referred to as “high-Ti/low-K”). 2. (2) high-Ti/low-Al/high-K (referred to as “high-Ti/high-K”). 3. (3) low-Ti/low-Al/low-K (referred to as “low-Ti”). 4. (4) low-Ti/high-Al/low-K (referred to as “high-alumina”). 5. (5) low-Ti /high-Al/ high-K (referred to as “VHK”). 6. (6) VLT/low-Al/low-K basalts (referred to as “VLT”). A variety of post-magma-generation processes have been invoked, such as fractional crystallization, either alone or combined with wallrock assimilation, to explain the compositional ranges of the various mare basalt suites. In order to evaluate these proposed petrogenetic processes, this review is by rock type and is non-site specific, but for each rock type, reference to particular lunar sample return missions is brought forth. This permits a comparison of similarities and differences of broadly similar rock types correlated with geography on the Moon, which, in turn, allows a more thorough petrogenetic evaluation. High-Ti mare basalts (i.e., high-Ti/low-Al/low-K) are found at Apollo 11 and Apollo 17 sites; however, the A-11 basalts contain lower TiO 2 abundances, a considerably larger range in trace-element contents, and the only occurrence of high-Ti/high-K mare basalts. Fractional crystallization and source heterogeneity within each site are the keys to understanding the petrogenesis of the high-Ti basalts. Low-Ti basalts (including both low-Al/K and high-Al/K varieties) are found at Apollo 12, 14, and 15, and Luna 16 sites. The low-Ti basalts exhibit a wide range of major- and trace-element compositions and require source heterogeneity, fractional crystallization, and some assimilation. The high-alumina mare basalts (i.e., low-Ti/high-Al/low-K) are found at Apollo 14 and Luna 16 sites and exhibit a wide range of major-and trace-element compositions. However, in these examples, source heterogeneity is not a major factor. Indeed, fractional crystallization coupled with KREEP assimilation, particularly for the Apollo 14 variants, can explain the compositional ranges of these high-alumina basalts. The VHK mare basalts (i.e., low-Ti/high-Al/high-K) have been sampled only at the Apollo 14 locale and are products of a parental highalumina magma assimilating lunar granite. Very low-Ti (VLT) mare basalts (i.e., VLT/low-Al/low-K) are found at Apollo 17 and Luna 24 sites. Fractional crystallization has had a major influence upon the range in VLT compositions, but Luna 24 VLT basalts have been derived from a source slightly different in composition from that for Apollo 17 VLT varieties. For example, the Luna 24 VLT basalts generally exhibit positive Eu anomalies, a unique property for mare basalts, which almost always have negative Eu anomalies. The concept of a lunar magma ocean (LMO) is generally accepted, and source modelling of all basalts invokes a “mafic LMO cumulate source.” This is the only unifying model for mare basalt petrogenesis, but the semantics and logistics of it are and will be debated for many years. For example, major convective overturn of the LMO appears plausible, but whether this occurred on a local- or planet-wide scale to produce source heterogeneity remains to be determined.