1 Introduction
The Moon has long suffered from the destructive effect of particles from interplanetary space owing to its lack of atmosphere and magnetic field protection. Exploration of the lunar surface magnetic field began in the Apollo era. Using data from lunar samples and orbiting lunar satellites (
Fuller and Cisowski, 1989;
Hood et al., 2001;
Mitchell et al., 2008), the Moon has been shown to possess extensive crustal magnetism. In recent years, a more accurate surface vector mapping of magnetic anomalies over the Moon has been performed using Kaguya and Lunar Prospector observations (
Tsunakawa et al., 2015). This mapping has shown the South Pole–Aitken (SPA) basin on the farside of the Moon to be a typical impact crater with isolated central magnetic anomaly features. However, controversy remains regarding the origin and evolutionary history of lunar magnetic anomalies, as well as the formation and distribution of magnetic minerals on the lunar surface.
Given the paucity of ferromagnetic minerals in lunar igneous rocks, in addition to thermo-remanent magnetization caused by an ancient core dynamo (~3.56 Ga) (
Collinson, 1985;
Weiss and Tikoo, 2014), the magnetism of lunar samples is likely primarily controlled by the quantity of body-centered cubic Fe phases and Fe‒Ni alloys added to the lunar surface by regolith-forming processes and meteorite impacts (
Wasilewski, 1974;
Rochette et al., 2010). In particular, the content of metallic iron is as high as 7 wt.% in the SPA basin on the farside of the Moon. Impact processes involving highly magnetic extra-lunar materials have been shown to be fundamental in generating an external magnetic field on the lunar surface (
Crawford and Schultz, 1988;
Wakita et al., 2021). Distal ejecta from the largest and oldest impact basin (i.e., SPA basin) may even explain the origin of isolated magnetic anomalies far from the basin (
Wieczorek et al., 2012). Many metallic grains on the lunar surface with high Ni and low Co contents originate from chondrites or iron meteorites (
Day, 2020). Common Fe‒Ni alloys in meteorites include kamacite (5‒6 wt.% Ni), taenite (20‒30 wt.% and occasionally 40 wt.% Ni), and awaruite (> 60 wt.% Ni). Tetrataenite is a unique mineral, formed by the ordering of Fe and Ni atoms in taenite during the cooling process (
Pechersky et al., 2012).
Previous studies have revealed substantial natural remanent magnetization in samples returned from the Apollo and Luna missions (
Pearce et al., 1973;
Runcorn et al., 1970;
Rochette et al., 2010). In recent years, evidence of ferromagnetic materials has also been found in Chang’E-5 (CE-5) lunar soils, including native sub-micron magnetite particles, an iron meteorite fragment originating from the outer Solar System, and pentlandite in both basalts and breccias that was exsolved from troilite (
Guo et al., 2022;
Tang et al., 2023;
Liu et al., 2024). A comprehensive study of the magnetic properties of lunar samples will help us to gain a more in-depth understanding of the origin and evolution of the lunar magnetic field. Currently, the Chinese Chang’E-6 (CE-6) mission has successfully returned 1935.3 g of samples from the SPA basin (41.625° S, 153.978° W). This is the first time in human history that lunar samples have been retrieved from the farside of the Moon. The petrological and geochemical characteristics of CE-6 samples indicate that CE-6 soil comprises a mixture of local lunar basalts and non-basaltic ejecta material (
Li et al., 2024b). Combining these data with the geomorphological features of the SPA, the CE-6 returned soils likely contain external materials that were excavated and transported from adjacent impact craters (
Qian et al., 2024). Hence, CE-6 lunar soils provide an excellent opportunity to study the evolutionary history of lunar magnetic anomalies.
Here, we report mineralogical analyses of tetrataenite formed by the phase transition of taenite in CE-6 lunar soil. To the best of our knowledge, this is the first report of tetrataenite formed by impact events in a lunar sample. The presence of tetrataenite, which is a hard magnetic mineral with remanent carrier capability, provides direct evidence of a source of magnetic anomaly carriers on the farside of the Moon; it is also highly likely to increase the local magnetic susceptibility of the lunar surface and may be a critical factor explaining the evolution of magnetic anomalies by space weathering.
2 Materials and methods
Lunar regolith soils, obtained from the farside of the Moon during the CE-6 mission, were allocated by the China National Space Administration (CNSA) for investigation (CE6C0300YJFM001). For comparison, we also selected some CE-5 lunar regolith soils (CE5C1000). We randomly selected 1‒2 mg of lunar soil from each of the CE-5 and CE-6 samples, added a few milliliters of ethanol, and dispersed them via ultrasonic vibration. Each partial suspension containing lunar soil was dispersed on a Si wafer and covered with an Au film for subsequent analysis.
The morphology and composition of the lunar soil samples were observed using scanning electron microscopy (SEM). Foils for subsequent transmission electron microscopy (TEM) analysis were obtained using focused-ion-beam (FIB) technology. This work was performed on a dual-beam system (FEI Scios) equipped with an energy-dispersive spectrometer (EDS). The FIB‒SEM system used a Ga ion source and Pt deposition to extract and fix the foil samples on a movable Cu grid. Observation of nanophases within the FIB foils was performed using a field-emission scanning TEM (200 kV, FEI Talos F200X). High angle annular dark-field images were acquired to identify different mineral phases. Chemical analysis of regions of interest on FIB foils was performed using scanning TEM combined with EDS. The quantification of the TEM- EDS data was performed using an interactive quantitative method in the Bruker Esprit software, which is part of the Talos F200X instrument. High-resolution TEM (HRTEM) images and selected-area electron diffraction (SAED) patterns were acquired to identify nanocrystal structures. Fast Fourier transform (FFT) was applied to HRTEM images of areas of mixed minerals to identify individual mineral structures. Differential phase contrast scanning transmission electron microscopy was used to perform magnetic imaging in an FEI Titan Themis Z TEM operated at 300 kV and t = 293 K. For magnetic imaging, the specimen was placed in magnetic field-free conditions (Lorentz mode) with the conventional objective lens turned off.
3 Results
3.1 Overview of iron-sulfide grains in CE-6 lunar soil
Iron-sulfides are common in lunar soil samples and are susceptible to alteration by space weathering (
Guo et al., 2022;
Cao et al., 2024). The space weathering features of iron-sulfide grains are characterized by a honeycomb-like surface with curved Fe whiskers of varying micro-scale lengths (
Matsumoto et al., 2021). Herein, a number of iron-sulfide grains from CE-5 and CE-6 lunar soils with similar characteristics were selected using scanning electron microscopy observations. Most of these grains had a relatively uniform internal composition without a complex mineral assemblage, suggesting that they were most likely broken off directly from their parent rocks during an impact event and are native lunar iron-sulfides (Supplementary Fig. S1). However, we found a unique iron-sulfide grain in a CE-6 returned sample. Although the main part of the grain had a porous texture, in addition to a curved metal whisker, a large number of nanoscale metal particles were attached to its surface (Fig. 1(a)). A focused-ion-beam foil was extracted from this grain to obtain more detailed information (Fig. 1(b)). Energy-dispersive spectroscopy (EDS) element maps showed that the main hemispherical grain was enriched in Fe and S (Figs. 1(c) and 1(d)) and the curved metal whisker was enriched in Fe (Figs. 1(f) and 1(g)). In addition, an Fe-rich nanoparticle was observed on the surface of the grain (Fig. 1(c), red dotted box). These data indicate that this grain likely experienced a substantial thermal event, a process that typically leads to S escape from iron-sulfides and the formation of nanophase Fe (np-Fe
0) particles and pure Fe whiskers (
Matsumoto et al., 2021;
Guo et al., 2022). Notably, we found a metallic Fe‒Ni particle encased in the middle of the iron-sulfide grain, with several sub-micron Fe-rich particles clustered around it (Figs. 1(c) and 1(e)).
3.2 Identification of tetrataenite in the iron-sulfide grain
The crystal structure and chemical composition of the iron-sulfide grain were analyzed in detail (Fig. 2). The grain contains an irregular spherical metallic particle approximately 500 nm in diameter (Fig. 2(a)). Quantitative EDS element mapping in a transmission electron microscope (TEM) showed that the distribution of Ni across the metallic particle was uneven (Fig. 1(e)). A quantitative TEM‒EDS line profile across the Fe‒Ni metallic particle indicated that the Ni content decreased from left to right, from an average of 50 wt.% to 20 wt.% (Figs. 2(d) and 2(h)). On the basis of high-resolution TEM images, fast Fourier transform (FFT) patterns, and selected-area electron diffraction (SAED) patterns of the metallic particles, the interplanar spacing of the high-Ni phase (~50 wt.%) was determined to be approximately 2.067 Å with (101) lattice fringes. The angle between (101) and (002) was 54.73°, showing a [00] zone axis, consistent with the crystal structure of tetrataenite (Figs. 2(b) and 2(c)). The FFT pattern of the low-Ni phase (20‒40 wt.%), which the angle between (311) and (11) was 58.5°, indicated a face-centered taenite structure (Fig. 2(g)). Therefore, the metallic particle comprised globular tetrataenite with a thick irregular rim of taenite, but the boundary between the two minerals was difficult to distinguish. In addition, the sub-micron Fe-rich particles at the periphery of the Fe‒Ni metallic particle and the Fe-rich whisker were determined to be a pure metallic Fe (α-Fe) phase with (110) lattice fringes (Figs. 2(d)‒2(f)). Quantitative TEM‒EDS analyses of various phases within the iron-sulfide grain are shown in Supplementary Table 1.
3.3 Magnetic properties of different minerals
In Fig. 3, we present a visualization of the magnetic field distribution inside the nanoscale magnetic objects within the iron-sulfide grain obtained from CE-6 lunar soil using differential phase contrast scanning transmission electron microscopy in a magnetic-field-free environment. A bright-field TEM image of the iron-sulfide grain revealed a distinctly mixed zone with different crystal orientations inside the grain (Fig. 3(a)). Although no notable differences were observed in the elemental composition of this zone (Figs. 1(c) and 1(d)), SAED data from the area outlined by the dashed box in Fig. 3(a) revealed a symbiotic relationship between troilite and monoclinic pyrrhotite (Fig. 3(b) and Fig. S2). The troilite showed no evidence of magnetization, but there were some scattered magnetic signals inside the grain, potentially resulting from pyrrhotite in the matrix (Fig. 3(c)). However, because of the extremely small particle size of the pyrrhotite and its incomplete crystalline form, its magnetic characteristics were unclear. The tetrataenite‒taenite (Te + Ta) mixed particle exhibited a continuous in-plane color-wheel relationship, likely due to their intergrowth, and the pure Fe particles and Fe whisker also showed relatively complete color-wheel characteristics (Figs. 3(c)‒3(e)), indicating that they had a typical vortex magnetic structure. Owing to the elongated morphology of Fe whiskers, there may be some limitation to the integrity of the vortex structure. In addition, local magnetic structural illusions caused by irregular pores and uneven boundaries within the sample were observed; these were ignored in our analysis. It is worth noting that although FIB preparation may affect the microscopic details of magnetic domain structures through sample thinning and surface amorphization, this effect is negligible for the submicron mineral particles in this study.
4 Discussion
The droplet-like structure of the iron-sulfide grain, which resembles a hemisphere, indicates that it has experienced thermal events followed by a shedding and sputtering process. Impact-induced thermal events can increase the reaction rate and subsequent diffusion of Fe atoms, promoting the growth of Fe whiskers (
Li et al., 2024a;
Matsumoto et al., 2020). Notably, the farside of the Moon, especially the region in the SPA basin, has a widespread distribution of magnetic anomalies (Supplementary Fig. S3) (
Tsunakawa et al., 2015), likely related to large-scale impacts (
Wieczorek et al., 2012). In addition, previous studies have shown that Ni-rich minerals, especially taenite and tetrataenite, widely occur in iron meteorites, stony-iron meteorites, and ordinary chondrites (
Scott and Clarke 1979;
Albertsen 1981;
Gattacceca et al., 2003;
Clarke and Scott 1980), with Ni contents that far exceed those of substances indigenous to the lunar surface. The high-frequency impact environment on the farside of the Moon may have resulted in particularly complex meteorite fragments being left on the lunar surface. Therefore, we can reasonably infer that the parent body of the analyzed iron-sulfide grain is likely to be a magnetic fragment left after the impact of a meteorite on the lunar surface. By comparing the Ni/Co ratio (approximately 22~25) in the iron-nickel alloy phases, we speculate that its parent material originated from chondrites metal (Supplementary Fig. S4). However, due to the thermal effect during the impact process leading to the recombination of elements, there is a certain degree of uncertainty in this traceability speculation. During the subsequent continuous impact-crushing process, the temperature vastly increases, causing the troilite to wrap around the taenite particle and peel off from the parent body, before finally being sputtered into the CE-6 landing site (Fig. 4). During the cooling process of iron-sulfide droplets, owing to the siderophilic nature of Fe, the internal Fe atoms of the troilite phase are more inclined to enter an Fe‒Ni metallic phase. Given the more evident desulfurization effect experienced by the outer layer of the iron-sulfide grain, a typical honeycomb-like structure is more likely to form in this region, while excess Fe from the surface of the grain forms np-Fe
0 particles and Fe whiskers with specific growth orientations. Owing to the different diffusion rates of Fe and S, a symbiotic relationship between troilite and pyrrhotite is formed in the interior of the grain during the cooling and recrystallization process.
The conventional model posits that tetrataenite typically forms from taenite via a phase transformation under extremely slow cooling conditions, often manifesting as a relatively complete rim structure at the periphery of taenite grains (
Clarke and Scott 1980). However, this study reveals a distinct textural relationship: tetrataenite coexists with taenite in a cloud-like texture and is associated with sub-micron metallic iron. We propose that the tetrataenite did not form during the instantaneous high-temperature and high-pressure stage of the impact. Instead, we argue that it formed within a sustained, regional thermal annealing environment triggered by the impact events. Larger-scale impact events are generally capable of providing a sufficiently prolonged medium- to low-temperature annealing window, which could create more favorable conditions for the sustained growth of tetrataenite. The phase equilibrium diagram for Fe‒Ni alloys indicates that the face-centered cubic taenite phase undergoes a monotectoid reaction (Ta → Te + α–Fe) below approximately 350°C (
Yang et al., 1996). Ferromagnetic tetrataenite and taenite typically exhibit a symbiotic relationship when the Ni content is ~50 wt.% (Fig. 4, upper right illustration) (
Pechersky et al., 2012). Additionally, recent studies have explored the potential of specific elements, such as phosphorus (P), to accelerate Fe-Ni atomic diffusion. Although the specific conclusions of some of these studies remain under debate, we have observed related phenomena in other lunar soil, where regions containing phosphorus appear to be more prone to Ni enrichment (Supplementary Figs. S5 and S6). This observation also provides a clue for the potentially rapid formation of tetrataenite.
Tetrataenite, taenite, pure np-Fe
0 particles, and Fe whiskers are formed through phase transformation and reduction processes. Therefore, the recorded characteristics most likely represent chemical remanence. Although the four aforementioned materials are all ferromagnetic, pure Fe minerals are soft magnets, which are easy to magnetize and demagnetize. Previous studies have shown that the Apollo returned lunar samples exhibited notable superparamagnetic properties (
Fuller, 1998), most likely related to the formation of np-Fe
0 particles via the reduction of impacted glasses and the regolith layer of lunar soils. The particle size of np-Fe
0 particles can be as small as a few nanometers. When the particle size of magnetic nanoparticles is smaller than the critical dimension, their magnetic properties become superparamagnetic. However, tetrataenite is a hard magnet with higher coercivity and remanence stability. Previous studies have shown that mm
3-scale ordinary chondrite samples containing ~40% tetrataenite can increase their coercivity and remanent magnetization at saturation by more than 10 times (
Gattacceca et al., 2003). Although the presence of metallic Fe‒Ni on the Moon is not uncommon, few studies have reported on tetrataenite on the lunar surface.
This study reports the occurrence of strongly magnetic minerals in the southern Apollo basin, which lies in the north-east portion of the SPA. The evolutionary process of tetrataenite reported herein conforms to the evolutionary characteristics of the surrounding terrain and landforms, and also indicates the complexity and multi-source nature of impact events on the farside of the Moon. This study provides strong evidence for the existence of impact debris from strongly magnetic parent bodies on the Moon, as well as potentially providing a new perspective for explaining the process of lunar surface magnetic evolution modified by space weathering. Additionally, this study also offers significant data to bolster future in situ land-surface exploration of magnetic minerals.
PDF
(3334KB)
