1:45 PM - 2:00 PM
[SMP25-13] Different entrapment timings of melt inclusions in garnet and quartz in an UHT granulite from Rundvågshetta
Keywords:partial melting, melt inclusion, nanogranitoid, nanogranite inclusion, ultrahigh-temperature metamorphism
The presence of melt during anatexis is considered to have profound effects on the rheology of middle to lower crustal rocks and thus the exhumation mechanisms of them (e.g., Jamieson et al., 2011). Therefore, precisely constraining the timings when melts coexisted in the pressure-temperature-time (P-T-t) evolutions of high-grade metamorphic rocks is important (e.g., Barich et al., 2014).
In this study, we report two types of melt inclusions (MI-g and MI-q) and discuss their entrapment timings in a clock-wise P-T evolution of an ultrahigh-T (UHT) granulite (sample TK2003010309) from Rundvågshetta (Lützow-Holm Complex, East Antarctica). MI-g are enclosed in garnet porphyroblasts, while MI-q are in quartz grains in the matrix.
The occurrences of MI-g have been already reported by previous studies (e.g., Hiroi et al., 2019). They commonly consist of Qtz + Kfs + Pl + Bt. We recently reported that MI-g occur exclusively in the P-poor cores of garnet (Suzuki and Kawakami, 2020). In addition to MI-g, the P-poor cores of garnet commonly include biotite, sillimanite, quartz, rutile, and zircon monomineralic inclusions. Therefore, the P-poor cores of garnet and MI-g were probably formed by the following dehydration melting reaction; Bt + Sil + Qtz = Grt + Kfs + melt (e.g., Spear et al., 1999). By applying the Zr-in-rutile geothermometer (Tomkins et al., 2007) to rutile grains enclosed in the P-poor cores of garnet, the entrapment P-T condition of MI-g was constrained to be 892 ± 11 °C/8.5 kbar to 914 ± 10 °C/12.2 kbar in the sillimanite stability field. This corresponds to the prograde stage in the clock-wise P-T evolution of the UHT granulite. The peak metamorphic condition of 1026 oC/14.6 kbar was obtained by applying the same method to inclusion minerals enclosed in the P-rich mantles of garnet.
We newly found MI enclosed in quartz grains in the intervening matrix between garnet and sillimanite porphyroblasts. These MI-q are not aligned in the host quartz grains. Therefore, we interpret that they are primary ones. They commonly consist of Qtz +Kfs + Pl + Ms, without any mafic minerals. Some internal porosities were observed in MI-q, which were probably formed due to the increase of density during the melt crystallization in a constant cavity volume (e.g., Cesare et al., 2009). Quartz grains in MI-q are less luminescent under cathodoluminescence (CL) images compared to the host quartz grains. This might reflect the differences of Ti content in quartz and thus the crystallization T condition (e.g., Hiroi et al., 2020). However, there still exist some other possible factors that cause CL emissions of quartz (e.g., Okumura, 2005), and thus further analyses are needed to understand the cause.
It should be noted that in order to trap MI, host minerals must grow in the presence of melt. This can occur in two ways; (Ⅰ) growth during melt-producing process and (Ⅱ) melt-consuming process (e.g., Bartoli et al., 2013). As discussed above, MI-g were formed with the peritectic P-poor cores of garnet during the prograde dehydration melting reaction (Ⅰ). On the other hand, MI-q and the host quartz grains in the matrix is not likely to be formed during process (Ⅰ). This is because quartz is always consumed as a reactant mineral and its modal amount decreases during the prograde melt-producing process (e.g., Weinberg and Hasalová, 2015). Therefore, the host quartz grains in the matrix can grow and trap MI-q only during the retrograde melt-consuming process (Ⅱ).
These results may indicate that melts coexisted from prograde into retrograde metamorphic stages and their composition might have changed to less mafic one during fractional crystallization. The melts coexisted in the prograde and retrograde stages of UHT metamorphism probably played an important role in weakening the metamorphic rocks (e.g., Rosenberg and Handy, 2005), and might have triggered and promoted their exhumation, respectively (e.g., Jamieson et al., 2011).
In this study, we report two types of melt inclusions (MI-g and MI-q) and discuss their entrapment timings in a clock-wise P-T evolution of an ultrahigh-T (UHT) granulite (sample TK2003010309) from Rundvågshetta (Lützow-Holm Complex, East Antarctica). MI-g are enclosed in garnet porphyroblasts, while MI-q are in quartz grains in the matrix.
The occurrences of MI-g have been already reported by previous studies (e.g., Hiroi et al., 2019). They commonly consist of Qtz + Kfs + Pl + Bt. We recently reported that MI-g occur exclusively in the P-poor cores of garnet (Suzuki and Kawakami, 2020). In addition to MI-g, the P-poor cores of garnet commonly include biotite, sillimanite, quartz, rutile, and zircon monomineralic inclusions. Therefore, the P-poor cores of garnet and MI-g were probably formed by the following dehydration melting reaction; Bt + Sil + Qtz = Grt + Kfs + melt (e.g., Spear et al., 1999). By applying the Zr-in-rutile geothermometer (Tomkins et al., 2007) to rutile grains enclosed in the P-poor cores of garnet, the entrapment P-T condition of MI-g was constrained to be 892 ± 11 °C/8.5 kbar to 914 ± 10 °C/12.2 kbar in the sillimanite stability field. This corresponds to the prograde stage in the clock-wise P-T evolution of the UHT granulite. The peak metamorphic condition of 1026 oC/14.6 kbar was obtained by applying the same method to inclusion minerals enclosed in the P-rich mantles of garnet.
We newly found MI enclosed in quartz grains in the intervening matrix between garnet and sillimanite porphyroblasts. These MI-q are not aligned in the host quartz grains. Therefore, we interpret that they are primary ones. They commonly consist of Qtz +Kfs + Pl + Ms, without any mafic minerals. Some internal porosities were observed in MI-q, which were probably formed due to the increase of density during the melt crystallization in a constant cavity volume (e.g., Cesare et al., 2009). Quartz grains in MI-q are less luminescent under cathodoluminescence (CL) images compared to the host quartz grains. This might reflect the differences of Ti content in quartz and thus the crystallization T condition (e.g., Hiroi et al., 2020). However, there still exist some other possible factors that cause CL emissions of quartz (e.g., Okumura, 2005), and thus further analyses are needed to understand the cause.
It should be noted that in order to trap MI, host minerals must grow in the presence of melt. This can occur in two ways; (Ⅰ) growth during melt-producing process and (Ⅱ) melt-consuming process (e.g., Bartoli et al., 2013). As discussed above, MI-g were formed with the peritectic P-poor cores of garnet during the prograde dehydration melting reaction (Ⅰ). On the other hand, MI-q and the host quartz grains in the matrix is not likely to be formed during process (Ⅰ). This is because quartz is always consumed as a reactant mineral and its modal amount decreases during the prograde melt-producing process (e.g., Weinberg and Hasalová, 2015). Therefore, the host quartz grains in the matrix can grow and trap MI-q only during the retrograde melt-consuming process (Ⅱ).
These results may indicate that melts coexisted from prograde into retrograde metamorphic stages and their composition might have changed to less mafic one during fractional crystallization. The melts coexisted in the prograde and retrograde stages of UHT metamorphism probably played an important role in weakening the metamorphic rocks (e.g., Rosenberg and Handy, 2005), and might have triggered and promoted their exhumation, respectively (e.g., Jamieson et al., 2011).