2024 Volume 119 Issue 1 Article ID: 231223
We performed a cathodoluminescence (CL) study of Ca-rich plagioclase (An85-86Ab14Or<1) in Stillwater gabbronorite experimentally shocked at 20.1, 29.8, and ∼ 41 GPa, for characterization of the shock effects. Chroma CL image of unshocked plagioclase showed the homogeneous red CL emission. In contrast, experimentally shocked plagioclase showed the heterogeneous CL emission colors in red and blue. The Raman spectra analysis identified that the red and blue portions correspond to plagioclase and maskelynite, respectively. In our observation, plagioclase experimentally shocked at 20 GPa was partially converted into maskelynite. At 30 GPa, most of plagioclase were converted into maskelynite. At 40 GPa, plagioclase was fully converted into maskelynite. Our observations of Ca-rich plagioclase indicated that the maskelynization starts at a slightly lower pressure and completes at a higher pressure than those in the previous studies (∼ 24 and ∼ 28 GPa, respectively). These pressure differences may be due to the high sensitivity of CL, which allows for the detection of small (a few µm in size) and rare phases that may have been overlooked in the traditional methods. The CL spectra of plagioclase showed a continuous change with increasing shock pressure. Hence, the CL imaging method using plagioclase and maskelynite is found to be very effective to estimate precisely shock pressure. In particular, there was a marked decrease in the CL intensity of Mn2+ and Fe3+ centers. Furthermore, the shock-induced center around the UV region was observed in experimentally shocked plagioclase and maskelynite. These CL features reflect the destruction of the framework structure to varying extents depending upon shock pressure. Combined with the Fourier transform infrared spectroscopy (FTIR) analysis in the present study, the transition of plagioclase to maskelynite was clearly illustrated in spectra. The reflectivity decreased continuously with increasing shock pressures during maskelynization. Additionally, the absorption at ∼ 8.6 µm observed in plagioclase was absent in maskelynite. This feature can be used as a diagnostic feature to characterize plagioclase and maskelynite by FTIR. The combination of detailed petrology using CL and FTIR spectra provides valuable insights into the shock scale for achondrites and planetary materials rich in shock-experienced plagioclase.
Plagioclase, ubiquitously found within terrestrial rocks and meteorites, is one of major rock-forming minerals. Many researchers performed shock recovery experiments of plagioclase at various pressures and temperatures to investigate its response to shock metamorphism. The experimentally shocked plagioclase shows structural changes such as undulatory extinction, fracturing, maskelynization, and melting with increasing pressure. These shock effects on plagioclase are often utilized as indicators of shock metamorphism (e.g., Kieffer et al., 1976; Stöffler et al., 1991). Note that maskelynite is considered to be the diaplectic glass of plagioclase formed by a solid-solid phase transition during the passage of a shock wave (e.g., Binns, 1967; Yamaguchi and Sekine, 2000). On the other hand, there are claims that maskelynite can be formed via quenching from a shock melt as a fused plagioclase glass under high pressure (Chen and El Goresy, 2000).
Cathodoluminescence (CL) techniques have been used as pivotal tools, for example, geochemistry, petrology, mineralogy, and crystallography (e.g., Kayama et al., 2009a). CL microscopy enables the observation of various features such as the two-dimensional distribution of silica polymorphs and chemical zoning in plagioclase. Furthermore, CL spectroscopy is utilized to assess the presence of impurities and defects within minerals. Consequently, CL techniques provide valuable information for understanding the formation of rocks and minerals (Götze et al., 2000). Supporting this, Kayama et al. (2009b) used CL techniques to characterize the shock effects in sodic plagioclase experimentally shocked at 20-40 GPa. Their study revealed the presence of shock centers in shocked plagioclase, supporting its usefulness as a new shock barometer for plagioclase analysis (Kayama et al., 2012).
In this study, we have used CL imaging and spectroscopic methods to investigate the shock effects of Ca-rich plagioclase and formation of maskelynite. It is well-established that the Stillwater gabbronorite is a reservoir of Ca-rich plagioclase, characterized by its homogeneous composition (Meurer and Boudreau, 1996). We conducted an evaluation of the shock effects on these Ca-rich plagioclase that was experimentally shocked at room temperature and pressures of 20.1, 29.8, and ∼ 41 GPa. The run-products obtained here were analyzed by various petrographic techniques for comparison with CL results. These results can be applied to various types of rocks containing Ca-rich plagioclase, such as eucrites and lunar meteorites.
Stillwater gabbronorite (Montana, USA) is selected for our measurements. Stillwater gabbronorite is mainly composed of coarse-grained Ca-rich plagioclase and orthopyroxene with thin augite lamellae (Fig. 1; McCallum, 2002). The mineral compositions are shown in Figure 2. The An# (= Ca/Ca + Na + K) of plagioclase are plotted in a limited area (An85-86Ab14Or0-1). Pyroxenes compositions are shown in the pyroxene quadrilaterals, and orthopyroxene (Fs15-16En81-82Wo1-3) and augite (Fs6-8En46-49Wo46-49) also have homogeneous compositions.
The shock recovery experiments were performed by using a single stage 30 mm bore propellent gun at the National Institute for Materials Science. The projectiles are 8 mm thick plate made of stainless-steel SUS 304, which were embedded in the front of a polyethylene sabot. A sample disk was mounted in a stainless container, which is placed in an iron momentum trap. The peak shock pressure produced in the target samples is determined by impedance matching method from the velocity of projectile measured just before impact (Sekine, 1997; Zel’dovich and Raizer, 2002) and the known Hugoniot of stainless (Marsh, 1980). In this study, we assumed the sample pressure reaches the equilibrium with the container by reflections and we successfully recovered run products subjected to peak shock pressures of 20.1 (Run# 603), 29.8 (Ran# 586), and ∼ 41 GPa (Run# 588). For Run# 588, the shock pressure of ∼ 41 GPa was estimated from the amount of smokless powder, since the velocity before impact could not be measured due to device trouble. The set-up of shock recovery experiments of this study is summarized in Table 1. Details of shock experimental procedures are described by Yamaguchi and Sekine (2000) and Sekine et al. (1987). The recovered run products were cut vertically from the impact front, and thin and thick sections were made. We observed the center of these run products because the rim portions of the sample in contact with the stainless container are highly deformed.
| Run# | Sample thickness | Container | Projectile | Flyer | Flyer thickness | Smokless powder | S. Velocity* | Pressure |
| (mm) | (g) | (mm) | (g) | (km/s) | (GPa) | |||
| 603 | 2.50 | SUS304 | 70.382 | SUS304 | 8 | 80 | 0.962 | 20.1 |
| 586 | 3.70 | SUS304 | 91.118 | SUS304 | 12 | 140 | 1.35 | 29.8 |
| 588 | 2.20 | SUS304 | 70.524 | SUS304 | 8 | 200 | - | ∼ 41** |
*S. Velocity is the velocity of projectile measured just before impact.
**Shock pressure estimated by smokeless powder amount.
We studied the recovered samples using an optical microscope with Power Mosaic system (Leika: DM4000 M), a field emission scanning electron microscope (FE-SEM: JEOL JSM-7100) equipped with an energy dispersive spectrometer (EDS) (Oxford Aztec Energy), and a CL imaging system (GATAN: Chroma CL2), an electron probe microanalyzer (EPMA: JEOL JXA-8200), a micro-Raman spectroscope (JASCO: NRS-1000) at National Institute of Polar Research (NIPR). Raman analyses were used to focus the excitation laser beam (beam diameter ∼ 1 µm and wavelength 532 nm). Near- to mid-infrared (2-16 µm) reflectance spectra were measured by FTIR microscope (JASCO: IRT-5000) and Fourier transform infrared spectrophotometer (JASCO: FT/IR-6100) at Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency (ISAS/JAXA). The FTIR analyses were performed with a spot size of 500 µm in diameter. Background spectrum is obtained using the diffuse gold reflectance standard (InfraGold). The sample spectra are normalized to the background spectrum to assume its reflectance value as 100%R. CL spectra were measured by SEM (JEOL: JSM-5410) equipped with a grating monochromator (Oxford: Mono CL2) at Okayama University of Science, ranging from 300 to 800 nm in 1 nm steps. CL spectra measurements were operated at 15 kV accelerating voltage and 2 nA beam current.
Optical microscopic images (Fig. 3), along with BSE and Chroma CL images (Fig. 4), reveals a progressive shock effect in the plagioclase of Stillwater gabbronorite, correlated with the experimental shock pressures.
Starting material (Unshocked). Upon optical microscopic observation (Figs. 3a and 3b), the unshocked plagioclase shows sharp optical extinction under cross-polarized light. The BSE image shows a few fractures on the plagioclase grains (Fig. 4a). The Chroma CL image shows that the plagioclase has a homogeneous red CL emission color across the grains (Fig. 4b). The Raman spectrum of unshocked plagioclase is composed of peaks at ∼ 195, ∼ 288, ∼ 402, ∼ 437, ∼ 488, ∼ 505, ∼ 529, and ∼ 563 cm−1 (Fig. 5a).
Experimentally shocked at 20 GPa. Upon optical microscopic observation (Figs. 3c and 3d), plagioclase experimentally shocked at 20 GPa shows an undulatory extinction under the cross-polarized light. The BSE image reveals an abundance of tiny cracks within the grains (Fig. 4c). The Chroma CL image shows that the plagioclase grain has heterogeneous CL emission color, presenting both red and blue portions on the grains (Fig. 4d). The red portions within plagioclase retain birefringence under the cross-polarized light, whereas the blue portion remain completely extinct. The Raman spectrum from the red portion shows distinct sharp peaks at ∼ 486 and ∼ 504 cm−1, indicating crystalline plagioclase (Fig. 5a). Compared to the unshocked plagioclase, the Raman spectrum shows a heightened background, and some minor peaks are unclear. In contrast, the Raman spectrum from blue portion shows two pronounced broad peaks around ∼ 506 and ∼ 577 cm−1, indicative of maskelynite (Fig. 5b). Intriguingly, we observed a linear structure in crystalline plagioclase embedded within the maskelynite which is characteristic texture for planar deformation fractures (PDFs) (Figs. 6a-6c).
Experimentally shocked at 30 GPa. In the optical microscopic observation (Figs. 3e and 3f), most plagioclase shocked at 30 GPa is completely extinct under the cross-polarized light (i.e., optically isotropic material). Several portions keep the birefringence. The BSE and Chroma CL images also show the heterogeneity of the textures (Figs. 4e and 4f). The Chroma CL image shows red and blue CL emission colors on plagioclase grain. BSE image shows that the red portion has fine fractures, and the blue portion has a smooth surface. The fractured portions are relatively darker than the smooth portions. Raman spectrum of red portion shows remarkable sharp peaks at ∼ 486 and ∼ 504 cm−1 with high background, indicating crystalline plagioclase (Fig. 5). The Raman spectrum of the blue portion shows two broad peaks around ∼ 506 and ∼ 590 cm−1, indicating maskelynite. We also observed PDFs in plagioclase shocked at 30 GPa (Figs. 6d-6f).
Experimentally shocked at 40 GPa. Upon optical microscopic observation (Figs. 3g and 3h), all of the plagioclase experimentally shocked at 40 GPa are completely extinct under the cross-polarized light (i.e., optically isotropic material). The BSE image shows the smooth surface (Fig. 4g). The Chroma CL shows homogeneous blue CL emission color across the grain (Fig. 4h). The Raman spectrum from blue portion shows two broad peaks around ∼ 506 and ∼ 590 cm−1, indicating maskelynite. (Fig. 5b).
CL spectra analysis and emission centerThe CL spectra of plagioclase and maskelynite show a progressive change corresponding to escalating shock pressures (Fig. 7). The unshocked plagioclase has broad CL peaks at ∼ 390, ∼ 520, ∼ 680, and ∼ 720 nm. Experimentally shocked plagioclase also has similar peaks, but the intensities decrease continuously with increased shock pressures, especially in the range from 400 to 800 nm. On the other hand, CL spectra of maskelynite are composed of only a broad peak at ∼ 390 nm that is shouldered in the UV region. We performed a deconvolution analysis of CL spectra based on Kayama et al. (2010). The CL spectra of both plagioclase and maskelynite were converted to energy and subsequently fitted with a Gaussian function (Fig. 8). We identified several emission centers in plagioclase and maskelynite around 3.18 eV (390 nm; Ti4+ impurity center or Al-O−-Al center) (Marfunin, 1979; Lee et al., 2007; Kayama et al., 2012), 2.35 eV (520 nm; Mn2+ center in M site) (Götze et al., 2000), 1.75 eV (680-720 nm; Fe3+ center) (Smith and Strenstrom, 1965), and 3.70 eV (350 nm; shock-induced center) (Kayama et al., 2009b). The intensities of Fe3+ and Mn2+ centers decrease significantly with increasing shock pressures. The shock-induced center becomes evident in both plagioclase and maskelynite in plagioclase and maskelynite above experimentally shocked at 20 GPa.
In-situ µ-FTIR spectra both of plagioclase and maskelynite from near- to mid-infrared range (2-16 µm), are shown in Figure 9. In mid-infrared spectral range, the stretching and bending vibrations of Si-O in the crystal structure responsible for several reflectance features, commonly referred to as the Reststrahlen bands. The unshocked plagioclase shows two pronounced reflectance bands between from 8 to 13 µm, similar to reported plagioclase (e.g., Pernet-Fisher et al., 2017). The most highly reflective band (36%R) shows an asymmetric band with a peak at ∼ 10.8 µm. Conversely, the other reflective band (15%R) shows symmetric band with a peak at ∼ 8.6 µm. Analogous pronounced reflectance bands are observed in the mixture of plagioclase and minor maskelynite experimentally shocked at 20 GPa. However, the reflectance is lower than that of the unshocked plagioclase, and the ∼ 8.6 µm peak is shifted to ∼ 8.9 µm. In contrast, maskelynite from samples experimentally shocked at both 30 and 40 GPa uniformly exhibits a single, more expansive reflectance band (20-22%R) at approximately 10.8 µm. The observed spectral shape and decrease in reflectance of maskelynite are consistent with features reported in previous studies (Pernet-Fisher et al., 2017). No difference in reflectance was observed in maskelynite obtained from samples subjected to shock pressures of 30 and 40 GPa. The Christensen Feature, representing the lowest reflectance portion (denoted as CF), for both plagioclase and maskelynite, is pinpointed at 8.2 µm. In this study, the FTIR spectrum of unshocked plagioclase is obtained from mirror-polished chip, while FTIR spectra of run products are obtained from polished thick sections, with no observed influence from the surrounding epoxy.
CL images provide us a two-dimensional distribution of plagioclase and maskelynite (Fig. 4). Variations in the CL emission of plagioclase and maskelynite can be elucidated by deconvolution analyses of their respective CL spectra (Fig. 8). The CL intensities of plagioclase studied here are decreasing with increasing shock pressures, especially Mn2+ and Fe3+ centers. Additionally, shock-induced center around the UV range increase above 20 GPa. Consequently, it is evident that as shock pressure increases, the CL color undergoes a progressive transition from red to blue until plagioclase converts into maskelynite. A similar phenomenon has been observed in natural samples. Kanemaru et al. (2020) reported that the color CL images and CL spectra of plagioclase in eucrites (basaltic achondrite) are continuously changed with increasing the shock degrees. Moreover, we observed an increase in the luminescence background and a decrease in the intensity of Raman peaks of plagioclase for increase with shock pressures (Fig. 5). Similar results have also been described in previous researches. Kayama et al. (2009b) also report the change of CL property and Raman spectra of experimentally shocked andesine. They report that the experimentally shock pressures are responsible for the decreasing CL emission and Raman intensities, particularly peaks associated with T-O-T atomic vibrations. In addition, Fritz et al. (2005) reports a systematic change in Raman spectra of plagioclase in Martian meteorites with increasing shock pressures. They found that shock-induced deformation in the plagioclase crystal lattice leads to an increase in luminescence background and broadening of Raman bands. Hence, these characteristics of CL and Raman properties are associated with the partial breakdown of the framework structure of plagioclase.
Additionally, in the Raman spectra of maskelynite, we also observed a broadening of peaks associated with the T-O modes at ∼ 1108 and 1190 cm−1 with the increase in shock pressure (Fig. 5b). This implies that in the amorphization process of plagioclase, the T-O bonds tend to be more preserved than the T-O-T bonds, due to their stronger bonding. On the other hand, the analyses of the Raman spectra of maskelynite and shocked plagioclase from previous studies (Fritz et al., 2005; Kanemaru et al., 2020), the peaks at ∼ 1108 and ∼ 1190 cm−1 are not always evident. Therefore, the appearance of these peaks should be strongly dependent on the crystal orientation, and accurate assessment of the amorphization of the T-O bond by Raman spectra will require detailed measurements considering the crystal orientation. The excitation of these peaks may also depend on the excitation wavelength of the laser used. Hence, we only note that the prominent 1108 and 1190 cm−1 peaks in the maskelynite produced by 20 GPa shock pressure indicates that this maskelynite may have formed near the threshold of the maskelynization pressure of Ca-rich plagioclase.
On the other hand, the CL spectra can be representing the amorphization degrees of maskelynite. The CL intensities for plagioclase and maskelynite decrease for peaks other than the shock-induced center under pressures below 20 GPa, whereas for shock pressures 30-40 GPa, there is an increase in CL intensities in the blue to UV region (i.e., Ti4+ impurity center or Al-O−-Al center, and shock-induced center). This feature has been observed in Kanemaru et al. (2020) and Kayama et al. (2009b). This increasing CL intensity may indicate the effect of post shock heating, but the Raman spectra does not display signs of annealing. Hence, likely cause of the increase in CL intensity in the blue to UV region above 20 GPa reflects the amorphization degrees of plagioclase and maskelynite, and these characteristics could be used for more detailed comparisons of shock degrees.
Based on our Chroma CL observations, we concluded that experimentally shocked plagioclase partially converts into maskelynite at 20 GPa, most plagioclase converts into maskelynite at 30 GPa, and plagioclase completely converts into maskelynite at 40 GPa. Previous shock recovery experiments indicate that the Ca-rich plagioclase partially converted into maskelynite at ∼ 24 GPa, and the plagioclase completely converted into maskelynite at ∼ 28 GPa (e.g., Kieffer et al., 1976; Ostertag, 1983; Fritz et al., 2019). We observed maskelynite in samples experimentally shocked at 20 GPa, which is slightly lower than values reported in previous studies. These pressure differences due to the high sensitivity of CL imaging method as well as experimental configurations, which allows for the detection of very small phases (a few µm in size) that have traditionally been overlooked. We observed that in samples (20-30 GPa) where plagioclase and maskelynite coexist, each of them exhibits characteristic occurrences, suggesting the influence of factors other than peak shock pressure during shock recovery experiments. The maskelynite at 20 GPa has thin vein textures (Fig. 4d). This texture may indicate the effects of heating due to friction and/or melting (i.e., shock vein). It is known that the maskelynization pressure decreases with increasing temperature (Yamaguchi et al., 2002; Kubo et al., 2010; Fritz et al., 2019). Hence, the formation of vein shaped maskelynite is likely caused by local heating below the critical maskelynization pressure at room temperature. On the other hand, the presence of plagioclase which shows PDFs in the sample shocked at 30 GPa suggests the presence of heterogeneous shock effects (Pittarello et al., 2020). Even if shock pressure is same in the overall sample, different orientations of grains may deform in different ways. This is more critical to influence to PDFs formation because pressure should get equilibrium much easier than temperature in experiments. On the other hand, relict plagioclase within maskelynite has been reported, and it is considered that local differences in shock impedance are likely the cause (Pittarello et al., 2020). Hence, CL imaging method is effective for more precise determinations of shock pressure estimates using plagioclase and maskelynite.
In the FTIR spectra (Fig. 9), plagioclase displays two broad peaks, whereas maskelynite shows only one. Notably, maskelynite doesn’t have the 9.1 µm absorption seen in plagioclase. Additionally, as the shock pressure increases, the reflectance of each sample decreases. These changes, consistent with previous research (Pernet-Fisher et al., 2017), indicate that they can serve as valuable diagnostic tools for identifying and differentiating these minerals in shocked samples and planetary surfaces. Additionally, we observed that in FTIR spectra of samples shocked at 20 GPa (a mixture of plagioclase and minor maskelynite), the peak initially at 8.6 µm in the unshocked plagioclase seems to have shifted to 8.9 µm. This peak shift reflects the partial distortion of crystal. However, it is known that the peak positions of plagioclase also change with crystal orientation (Jaret et al., 2015; Pernet-Fisher et al., 2017), thus 8.9 µm peak may not be useable as an indicator of shock.
The analysis of shock effects on plagioclase obtained from this study indicates that the results could be applicable to various meteorites containing Ca-rich plagioclase (i.e., eucrites and Lunar samples). For example, the shock pressure of shock degree E (strongly shocked) eucrites has previously been estimated to be ∼ 20-35 GPa (Kanemaru et al., 2020). This is because most of the plagioclase in shock degree E eucrites converted into maskelynite. Here, in comparison with the recent findings, the shock features of shock degree E eucrites are consistent with the sample experimentally shocked at 30 GPa. Hence, our experimental results enable a more detailed estimation of the shock pressure. Moreover, the characteristics of shock effects can be identified by FTIR analysis. This leads to an understanding of the valuable properties of Ca-rich plagioclase in extraterrestrial materials through non-destructive method, and could be applied to screening for shock metamorphic grades.
In this study, we performed CL characterization of shocked Ca-rich plagioclase in Stillwater gabbronorite. Our findings reveal a systematic conversion of plagioclase to maskelynite under varying shock pressures: partial conversion at 20 GPa, nearly complete at 30 GPa, and total transformation by 40 GPa. Planar Deformation Features were evident in plagioclase experimentally shocked 20 and 30 GPa. In a previous study, relatively sodic plagioclase (up to An60) was characterized using CL methods. In this study, we reported the CL feature of Ca-rich plagioclase subjected to shock pressure, and it is possible to apply the CL method to meteorites such as Lunar samples and eucrites that contain Ca-rich plagioclase. Hence, CL microscopy is an effective method for petrographic description because it can easily distinguish between plagioclase and maskelynite. Moreover, CL microscopy can estimate the presence of shock effects in plagioclase from shock-induced center. Complementing these results, FTIR spectra demonstrated a progressive change from plagioclase to maskelynite. The most noticeable change with increasing shock pressure is a decrease in reflectivity. Moreover, the ∼ 8.6 µm peak observed in plagioclase is absence in maskelynite. This feature can be used as a diagnostic feature to characterize plagioclase and maskelynite. This characteristic suggests the potential of FTIR for non-destructive analysis of shocked samples for example eucrites, which can be invaluable for diverse applications.
We are grateful to Astromaterials Science Research Group of ISAS for helping with the FTIR analysis. We are grateful to the Chief Editor, M. Satish-Kumar, Associate Editor, T. Nagai, and to J. Fritz, and an anonymous reviewer for their constructive reviews.
