Formation processes of fibrolitic sillimanite seams and sillimanite inclusion-rich quartz by fluid-rock interaction (Wazuka-Kasagi area, Ryoke Belt, SW Japan)

Abstract

Fluid-rock interaction between pelitic gneiss and fluids released from a post-tectonic granite in the low pressure/temperature (P/T) type Ryoke metamorphic belt (Wazuka-Kasagi area, SW Japan) has been shown to have formed fibrolitic sillimanite-quartz seams in close association with discordant granitic veins crosscutting the host gneiss. The sillimanite seams are characterized by the presence of cathodoluminescence (CL)-dark quartz grains including sillimanite needles ± muscovite and the absence of feldspars in and around the seams. Thermodynamic calculations showed that the reaction between feldspars and aqueous fluid with low values of log(aK⁺/aH⁺) < ∼ 3.6 and log(aNa⁺/aH⁺) < ∼ 4.1 can produce sillimanite ± muscovite + quartz at 3 kbar, 600 °C. Feldspars close to the fluid path were consumed and CL-dark quartz grains with sillimanite inclusions were formed by the high aH⁺ fluid-rock interaction. The abundance of sillimanite inclusions in quartz decreases with distance from the seam, and quartz grains far from the seam typically have CL-bright and sillimanite-free cores, reflecting the decreasing aH⁺ with distance from the seam as a result of the progress of the fluid-rock interaction. The sillimanite seams associated with CL-dark quartz enclosing sillimanite ± muscovite and the absence of feldspars in the vicinity are microstructural indicators of fluid-rock interaction by high aH⁺ fluids and represent fossilized pathways for such fluids.

INTRODUCTION

The deep crust is composed largely of metamorphic rocks, and thus understanding the fundamental controls exerted by metamorphic fluids on mass and heat transfer, mineral reactions, and rock rheology is critical for determining the geochemical and petrological evolution of the crust (Ague, 2014). The mid- to lower crust of continental arcs undergoes multiple stages of fluid-related events, such as prograde dehydration during regional metamorphism, retrograde fluid release from crystallizing partial melts, and fluid release from crystallizing magmatic bodies during contact metamorphism (e.g., Kawakami et al., 2022; Higashino et al., 2023; Kawakami et al., 2025). In understanding the consequences of such fluid-rock interactions, it is important to recognize the microtextural record of fluid pathways.

Sillimanite and its fibrous variety (fibrolite) are metamorphic indicator minerals commonly found in high-temperature metamorphic rocks at low to medium pressure, usually formed by the breakdown (i.e., dehydration or dehydration melting) of muscovite in the presence of quartz with increasing temperature. However, the origin of some types of sillimanite mats and seams has been ascribed to acidic (or high aH⁺) fluid-related processes (e.g., Vernon, 1979; Kerrick, 1987; Johnson et al., 2003). In such cases, sillimanite seams and their immediate surroundings are interpreted to record former pathways of acidic fluids that were sourced from either internally during partial melting of the host rocks (Johnson et al., 2003) or externally from adjacent magmatic bodies (Kerrick, 1987; McLelland et al., 2002). Despite numerous studies, detailed studies of the formation mechanism of sillimanite seams have not used modern imaging techniques such as cathodoluminescence (CL) imaging and X-ray elemental mapping. Recently, a detailed study of fracture-filling fibrous Al₂SiO₅ minerals using Raman spectroscopy revealed that such Al₂SiO₅ minerals are sometimes mixtures of different polymorphs, such as sillimanite and kyanite, and are useful in constraining the P-T conditions of the final-stage fluid infiltration (Higashino et al., 2025a). Therefore, a detailed study of the formation mechanism of Al₂SiO₅ seams improves our understanding of the microstructural evolution of high-temperature metamorphic rocks in the middle to lower crust.

This study describes and discusses the formation of sillimanite seam in the Ryoke plutono-metamorphic belt (Wazuka-Kasagi area) in SW Japan using CL imaging and X-ray elemental maps with special emphasis on the microstructural evolution of sillimanite-bearing quartz observed in the vicinity of such sillimanite seams. By comparing the samples containing sillimanite seams with unmetamorphosed rocks equivalent to their protoliths and with sillimanite-free pelitic and psammitic schists, the microstructural evolution during fluid-rock interaction between host pelitic gneisses and acidic fluids is discussed.

GEOLOGICAL SETTING

The Ryoke Belt is a well-studied low P/T type plutono-metamorphic belt with an elongated distribution over 800 km in SW Japan (Miyashiro, 1961; Okudaira et al., 1993; Nakajima, 1994; Okudaira, 1996; Brown, 1998; Ikeda, 1998; Suzuki and Adachi, 1998; Kawakami, 2001a, 2001b, 2004; Kawakami et al., 2013; Skrzypek et al., 2016, 2018; Takatsuka et al., 2018a, 2018b; Kawakami et al., 2019, 2022; Okudaira et al., 2024; Fig. 1a). It is mainly composed of abundant syn- to post-tectonic granitoids, pelitic and psammitic metamorphic rocks and metachert. The highest grade metamorphic zone reached peak P-T conditions of ∼ 800 °C and ∼ 0.5 GPa between 96.5 ± 1.9 and 91.5 ± 1.0 Ma, followed by a nearly isothermal decompression P-T path in the Aoyama area, located ∼ 30 km east of the studied area (Fig. 1a; Kawakami, 2001a, 2002; Kawakami and Suzuki, 2011; Kawakami et al., 2019). The metamorphic rocks grade into the unmetamorphosed accretionary complex of the Mino-Tanba Belt to the north.

Figure 1. (a) Simplified geological map of SW Japan showing the distribution of the Ryoke Belt and four well-studied exposure areas of the Ryoke metamorphic rocks. (b) Simplified map showing the sampling locality of Wazuka stop 1-2. The main study area shown in (c) is indicated by a dashed box. (c) Geological map of the Wazuka-Kasagi area (after Ozaki et al., 2000) showing sample localities used in this study. Field distribution of Al₂SiO₅ minerals are also shown. ‘Type 1 And’ represents andalusite with chiastolite texture and ‘Type 2 And’ represents that without chiastolite texture.

In the Wazuka-Kasagi area, the accretionary complex gradually increases in metamorphic grade toward the south. Five metamorphic zones are defined; Chl, Chl-Bt, Bt, Crd, and Sil zones from north to south (Wang et al., 1986; Ozaki et al., 2000; Okudaira et al., 2009; mineral abbreviations after Warr, 2021; Fig. 1c). The Chl, Chl-Bt, Bt, and Sil zones are of regional metamorphic origin, while the Crd zone and part of the Sil zone near the Yagyu Granite are of contact metamorphic origin (Ozaki et al., 2000). The Kizugawa fault defines the boundary between the Crd and Sil zones (Fig. 1c). The main lithologies are pelitic and psammitic schists and metachert in the northern part of the area, while migmatitic gneisses are exposed in the southern part (Fig. 1c). Post-tectonic massive granites such as the Koya Granite and the Yagyu Granite intrude discordantly into the pelitic and psammitic schists, and contact metamorphic aureoles, indicated by the formation of andalusite and cordierite, develop around the granites in the northern part of the area (Wang et al., 1986). On the other hand, gneissose granites such as the Sagawa Granite and the Sugawa Granodiorite intruded concordantly into the migmatitic gneisses in the southern part (Fig. 1c). The intrusion times of the granitoids in this area are not well constrained due to the lack of geochronological studies using high closure temperature geochronometers. The Ao Granite exposed in the Aoyama area is dated at 79.8 ± 3.9 Ma by chemical Th-U-total Pb isochron method (CHIME) monazite dating (Kawakami and Suzuki, 2011) while U-Pb zircon dating by laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) yielded ages of 73-72 ± 1 Ma (Higashino et al., 2025b). The LA-ICPMS U-Pb age of the Joryu Tonalite exposed in the Aoyama area is 88 ± 1 Ma (Higashino et al., 2025b).

ANALYTICAL METHODS

CL imaging and X-ray elemental mapping of thin section samples were performed using a JEOL JXA-8105 superprobe equipped with a panchromatic CL detector XM-Z07030TPCL (Hamamatsu Photonics Co.). The analytical condition for CL observation was acceleration voltage of 15 kV and probe current of 1-3 nA with a focused beam to a defocused beam of 3 µm diameter. Analytical conditions for X-ray elemental mapping were acceleration voltage of 15 kV and probe current of 60 nA with a focused to defocused beam up to 10 µm diameter. All analyses were performed at the Department of Geology and Mineralogy, Kyoto University, Japan.

SAMPLE DESCRIPTIONS AND RESULTS OF CL OBSERVATIONS

More than 80 samples were collected widely in the Wazuka-Kasagi area and three samples from the Crd and Sil zones (Fig. 1) were selected for detailed study using CL images and X-ray elemental maps. These samples are pelitic and psammitic schists and gneisses (metatexite migmatites) with and without sillimanite seams (Figs. 2-4). The microstructures of a sample with sillimanite seams (Figs. 2 and 3) are compared with those without sillimanite seams (Fig. 4) to understand the microstructural characteristics caused by the sillimanite seam formation. In addition, unmetamorphosed Mino-Tanba Belt samples were also collected and examined in detail to identify microtextural features inherited from the protolith, (sample Wazuka stop 1-2; Figs. 1b and 5). These samples are described in detail below.

Figure 2. (a) Slab photo, (b) plane polarized light (PPL) thin section photo, and (c) crossed polarized light thin section photo of sample 1304295. Black boxes in (b) indicate areas observed by CL and X-ray elemental mapping shown in Figure 3. Red dashed lines indicate boundary between a granitic vein and the host gneiss. Blue dashed lines indicate boundary between sillimanite seams and the host gneiss.

Figure 3. (a) Photomicrograph (PPL), (b) backscattered electron (BSE) image, [(c) and (f)] CL images, X-ray elemental maps of (d) Na and (e) Ca, and (g) photomicrograph (PPL) of the boxed area in (f) of sample 1304295. The observed area for (a)-(e) is indicated by a black box in Figure 2b. The false-color CL image of a red box in (a) is shown as (f). Warm color represents higher CL intensity. Note that abundant sillimanite needles are enclosed in quartz grains in (g). F, feldspars.

Figure 4. (a) Photomicrograph (PPL), (b) X-ray elemental map of Na, and (c) false-colored CL image of sample 1210194 (alternating pelitic and psammitic layers in schist without sillimanite), indicating heterogeneous distribution of plagioclase. (c) Warm color represents higher CL intensity. Green- to blue-colored grains are quartz, and white represents plagioclase and sillimanite. (d) Enlargement of red-boxed area d in (b), showing the absence of plagioclase in the center part of the map area. (e) Enlargement of red-boxed area d in (b). White grains are plagioclase and K-feldspar. (f) Enlargement of red-boxed area f in (b), showing abundant plagioclase grains. (g) Enlargement of red-boxed area f in (b). Note that quartz grains in (e) and (g) show similar CL intensity.

Figure 5. (a) Photomicrograph (PPL), (b) X-ray elemental map of Na, and (c) false-colored CL image of sample Wazuka stop1-2 (unmetamorphosed pelitic and psammitic alternation from the Mino-Tanba Belt). (c) Warm color represents higher CL intensity.

Pelitic gneiss with a sillimanite seam

Sample 1304295 was collected about 5 m from the contact between the pelitic gneiss and granite in the Sil zone (Fig. 1c). A granitic vein composed mainly of Qz+Pl+Kfs+Bt+Ms intrudes the gneissose structure discordantly (Fig. 2a). It is coarser-grained than the host pelitic gneiss consisting of Bt+Kfs+Pl+Qz+Ms. A gray ‘sillimanite seam’ composed mainly of Sil+Bt+Ms originates from the granitic vein, which also cuts the gneissose structure discordantly at a low angle. The thickness of the seam varies from a few µm to several mm (Figs. 2 and 3). In thin seams, sillimanite and muscovite envelope matrix minerals, clearly indicating that sillimanite-seam formation postdates gneissose fabric formation (Figs. 3a-3c). Abundant sillimanite needles are included in quartz grains in and around the sillimanite seams, and the amount of sillimanite inclusions in quartz decreases with distance from the sillimanite seams (Fig. 3g). Plagioclase is absent in and around the sillimanite seams, although it is commonly observed in the matrix distant from the sillimanite seams (Fig. 3).

CL images of sillimanite seams and their host gneiss are shown in Figures 3c and 3f. Sillimanite and plagioclase are CL-bright. Quartz away from the sillimanite seam commonly has CL-bright cores, while quartz near the seam is CL-dark. There is a tendency for quartz grains with abundant sillimanite inclusions to be CL-dark (Fig. 3), and for plagioclase grains to be accompanied by CL-bright quartz grains (Fig. 3).

Alternating pelitic and psammitic layers in schist without sillimanite

Both samples show alternation of pelitic and psammitic layers without sillimanite (Fig. 1c). Sample 1210194 is collected from the Crd zone and consists mainly of Bt+Crd+Kfs+Pl+Qz. It characteristically shows a heterogeneous distribution of plagioclase. Patches free of plagioclase are observed in both pelitic and psammitic parts (Figs. 4a-4b). Sample 1210274 was collected from the Sil zone. It is mainly composed of Bt+Kfs+Pl+Qz and shows a relatively homogeneous distribution of plagioclase. In both samples, there is no clear spatial correlation between CL intensity of quartz and the presence or absence of neighboring feldspars (Figs. 4c-4g).

Unmetamorphosed pelitic and psammitic alternation from the Mino-Tanba Belt

A pelitic and psammitic alternation sample Wazuka stop1-2 was collected from the unmetamorphosed Mino-Tanba accretionary complex (stop 1 of Hirajima et al., 1992; Fig. 1b). Both quartz and plagioclase show grain-by-grain variations in CL brightness. CL-dark quartz veins cut the sedimentary layering discordantly and are CL-darker compared to the matrix quartz (Fig. 5). The spatial distribution of plagioclase is not correlated with the location of quartz veins (Fig. 5).

ACTIVITY DIAGRAMS

Activity diagrams were calculated using SUPCRT92 (Johnson et al., 1992) under varying P-T conditions from 0.1 GPa and 500 °C to 0.35 GPa and 700 °C. Quartz saturation and aH₂O = 1 were assumed. Figure 6 is log(aNa⁺/aH⁺)_fluid versus log(aK⁺/aH⁺)_fluid diagrams in the KNASH system showing the stability fields of sillimanite, albite, muscovite, and K-feldspar. The P-T conditions are assumed to be 0.30 GPa and 600 °C for Figure 6a, and 0.30 GPa and 700 °C for Figure 6b. The pressure condition of 0.30 GPa was chosen because both andalusite and sillimanite are found around the sample locality for 1304295 (Fig. 1; Ozaki et al., 2000). No significant difference except for the disappearance of muscovite stability field is observed in Figure 6b by the increase of temperature from Figure 6a. At 600 °C, muscovite stability field disappears below 0.2 GPa and appears above 0.3 GPa.

Figure 6. log(aNa⁺/aH⁺)_fluid versus log(aK⁺/aH⁺)_fluid diagram in KNASH system showing the stability fields of sillimanite, albite, muscovite and K-feldspar for (a) 0.30 GPa and 600 °C, and (b) 0.30 GPa and 700°C.

DISCUSSION

Microstructural constraints on the origin of sillimanite seams originating from the post-tectonic granitic vein

The sillimanite seams in sample 1304295 discordantly cut the gneissose structure of the host pelitic gneiss and originate from the post-tectonic granitic vein (Fig. 2). This suggests that the formation of sillimanite seams postdates the Ryoke regional metamorphism that formed the gneissose structure, and that fluids released from the crystallizing granitic vein are responsible for the formation of the sillimanite seams. CL observation of an unmetamorphosed pelitic and psammitic alternation (sample Wazuka stop1-2; Fig. 5) revealed that quartz grains in the rock equivalent to the inferred protolith of the studied gneisses commonly show CL-bright cores and rarely has CL-dark quartz veins. In spite of the presence of CL-dark quartz, no correlation between CL intensity of quartz and spatial distribution of plagioclase is observed in the unmetamorphosed protolith (Fig. 5). Similarly, a sillimanite-free schist sample with pelitic and psammitic alternation (sample 1210194) has Pl-rich and Pl-poor domains, but quartz in these domains has CL-bright cores (Fig. 4). These observations show that CL-dark quartz layers without feldspars are neither inherited from the protolith nor a product of regional metamorphism.

On the other hand, the CL-dark quartz crystals are commonly distributed in and around the sillimanite seams and contain numerous sillimanite inclusions, especially near the sillimanite seams (Fig. 3). The difference in CL brightness between sillimanite-bearing quartz in and around the sillimanite seams (CL-dark) and sillimanite-free quartz far from the sillimanite seams (generally CL-bright) supports that these quartz grains represent different generations. For quartz to contain sillimanite needles, the quartz would have to be newly formed or recrystallized during the sillimanite seam formation. The decreasing abundance of sillimanite inclusions in quartz and CL-dark rims with increasing distance from the sillimanite seams (Fig. 3) probably reflect the waning effect of fluid-rock interaction further away from the seams.

Thermodynamic calculations show that feldspars are not stable in the presence of acidic fluids (low aNa⁺/aH⁺ or low aK⁺/aH⁺) at the modelled P-T conditions and react to form sillimanite ± muscovite + quartz (Fig. 6). Therefore, sillimanite seams can be formed in originally feldspar-bearing gneisses through infiltration of acidic fluids (e.g., Kerrick, 1987; McLelland et al., 2002). In the case of this study, the post-tectonic granitic vein is a likely source of such fluids as discussed above (Fig. 2). The absence of plagioclase in the vicinity of the sillimanite seams in this study (Fig. 3) suggests a low aNa⁺/aH⁺ nature of the infiltrating fluids. In addition, the coexistence of rare muscovite with sillimanite as inclusions in quartz near the sillimanite seams indicates that some of the muscovite grains in the sillimanite seams were formed simultaneously with sillimanite, and thus the aK⁺/aH⁺ of the infiltrating fluid was also low and the temperature of the fluid infiltration condition did not exceed 700 °C and was possibly around 600 °C under which muscovite can coexist with sillimanite at log (aK⁺/aH⁺)_fluid ∼ 3.2 (Fig. 6). Vernon (1979) proposed a reaction to dissolve albite and form sillimanite in the presence of acidic fluid as follows:

  

\begin{align} &2\text{NaAlSi$_{3}$O$_{8}$} + 2\text{H}^{+} \\&\quad= \text{Al$_{2}$SiO$_{5}$} + 2\text{Na}^{+} + \text{H$_{2}$O} + 5\text{SiO$_{2}$} \end{align} (1).

Similarly, K-feldspar can be a reactant instead of albite as follows:

  

\begin{align*} &2\text{KAlSi$_{3}$O$_{8}$} + 2\text{H}^{+} \\&\quad= \text{Al$_{2}$SiO$_{5}$} + 2\text{K}^{+} + \text{H$_{2}$O} + 5\text{SiO$_{2}$} \end{align*}

These reactions show that acidic fluid infiltration can decompose alkali-feldspars into sillimanite and quartz as a result of fluid-rock interaction, and sillimanite needles can be incorporated into newly grown quartz grains. Thus, CL-dark sillimanite-bearing quartz found together with the sillimanite seams and the absence of feldspars near the sillimanite seams are interpreted as results of fluid-rock interaction consuming feldspars that occurred between high aH⁺ fluid and feldspar-bearing rock. The sillimanite seams probably represent fossil pathways of reactive fluids. The decreasing amount of sillimanite inclusions in quartz with distance from the fluid pathway likely reflects an aH⁺ gradient in the infiltrating fluid produced as a consequence of reaction (1) in which H⁺ is consumed as the reactive fluid moves from the sillimanite seam toward the host gneiss (Fig. 3). The development of sillimanite-free CL-dark rims on the CL-bright quartz cores further from the sillimanite seam, where plagioclase is not consumed (Fig. 3f top and bottom), may represent dissolution-reprecipitation of the quartz rims by reaction with the infiltrating fluid along grain boundaries (e.g., Nakamura and Watson, 2001). A similar case for feldspars is reported by Higashino et al. (2019).

Fluids with low pH are observed in the vicinity of the granite body in the modern geothermal field such as Kakkonda, Japan (Kasai et al., 1998). This supports the interpretation that such fluids were evolved from crystallizing post-tectonic granites in the study area. Thus, it is most likely that the sillimanite and muscovite in the sillimanite seams and the CL-dark quartz containing them were formed simultaneously during the infiltration of the low aNa⁺/aH⁺ and low aK⁺/aH⁺ fluid released from the crystallizing granitic vein. Therefore, the sillimanite seams associated with CL-dark quartz containing sillimanite needles and the absence of adjacent feldspars can be used as microstructural indicators of fluid-rock interactions involving high aH⁺ fluids.

CONCLUSIONS

1. 1.    Detailed microstructural study of sillimanite seams from the low P/T type Ryoke metamorphic belt using CL imaging revealed that these microstructures were formed by fluid-rock interaction involving high aH⁺ fluids derived from crystallizing granites.
2. 2.    The sillimanite seams associated with CL-dark quartz enclosing sillimanite and the absence of neighboring feldspars can be used as microstructural indicators of fluid-rock interaction involving high aH⁺ fluids and represent fossilized pathways of such fluids.
3. 3.    CL imaging is useful in tracing fossilized fluid pathways and fluid-rock interaction zones in metamorphic terrains.

ACKNOWLEDGMENTS

We would like to thank Jonas Kaempf and an anonymous reviewer for constructive comments, Sajeev Krishnan for editorial efforts, and Masaki Takaya for making thin sections. This study was financially supported by the JSPS KAKENHI Grant Numbers JP23740391, JP26400513, JP19H01991, JP23H00145 to T. Kawakami.

REFERENCES

• Ague, J.J. (2014) Fluid flow in the deep crust. Treatise in Geochemistry, 2nd edition, 203-247.
•  Brown,  M. (1998) Unpairing metamorphic belts: P-T paths and a tectonic model for the Ryoke Belt, southwest Japan. Journal of Metamorphic Geology, 16, 3-22.
•  Higashino,  F.,  Kawakami,  T.,  Tsuchiya,  N.,  Satish-Kumar,  M., et al (2019) Brine Infiltration in the middle to lower crust in a collision zone: mass transfer and microtexture development through wet grain-boundary diffusion. Journal of Petrology, 60, 329-358.
•  Higashino,  F.,  Kawakami,  T.,  Adachi,  T. and  Uno,  M. (2023) Multiple post-peak metamorphic fluid infiltrations in southern Perlebandet, Sør Rondane Mountains, East Antarctica. Journal of Mineralogical and Petrological Sciences, 118, S004.
•  Higashino,  F.,  Kawakami,  T.,  Ikeda,  H.,  Adachi,  T. and  Uno,  M. (2025a) Mode of occurrence of fracture-filling microcrystalline quartz and Al₂SiO₅ mineral assemblage in the Sør Rondane Mountains, East Antarctica: Implication for fluid infiltration during retrograde metamorphism. Polar Science, doi:10.1016/j.polar.2025.101209.
•  Higashino,  F.,  Kawakami,  T.,  Kudo,  S.,  Hirata,  T. and  Sakata,  S. (2025b) Zircon crystallization timings of granitoids in the Aoyama area, Ryoke belt, Southwest Japan. Island Arc, 34, e70022.
• Hirajima, T., Wang, G.F., Ikeda, T. and Takeuchi, K. (1992) Low to medium-grade metamorphic rocks in the Wazuka area, southern Kyoto. 29th IGC field guide book vol.5, Metamorphic belts and related plutonism in the Japanese Islands, The Geological Survey of Japan, 255-262.
•  Ikeda,  T. (1998) Progressive sequence of reactions of the Ryoke metamorphism in Yanai district, southwest Japan: the formation of cordierite. Journal of Metamorphic Geology, 16, 39-52.
•  Johnson,  J.W.,  Oelkers,  E.H. and  Helgeson,  H.C. (1992) SUPCRT92, a software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bars and 0° to 1000°C. Computers & Geosciences, 18, 899-947.
•  Johnson,  T.E.,  Gibson,  R.L.,  Brown,  M.,  Buick,  I.S. and  Cartwright,  I. (2003) Partial melting of metapelitic rocks beneath the Bushveld Complex, South Africa. Journal of Petrology, 44, 789-813.
•  Kasai,  K.,  Sakagawa,  Y.,  Komatsu,  R.,  Sasaki,  M., et al (1998) The origin of hypersaline liquid in the Quaternary Kakkonda granite, sampled from well WD-1a, Kakkonda geothermal system, Japan. Geothermics, 27, 631-645.
•  Kawakami,  T. (2001a) Tourmaline breakdown in the migmatite zone of the Ryoke metamorphic belt, SW Japan. Journal of Metamorphic Geology, 19, 61-75.
•  Kawakami,  T. (2001b) Boron depletion controlled by the breakdown of tourmaline in the migmatite zone of the Aoyama area, Ryoke metamorphic belt, southwestern Japan. The Canadian Mineralogist, 39, 1529-1546.
•  Kawakami,  T. (2002) Magmatic andalusite from the migmatite zone of the Aoyama area, Ryoke metamorphic belt, SW Japan, and its importance in constructing the P-T path. Journal of Mineralogical and Petrological Sciences, 97, 241-253.
•  Kawakami,  T. (2004) Tourmaline and boron as indicators of the presence, segregation and extraction of melt in pelitic migmatites: examples from the Ryoke metamorphic belt, SW Japan. Transactions of the Royal Society of Edinburgh: Earth Sciences, 95, 111-124.
•  Kawakami,  T. and  Suzuki,  K. (2011) CHIME monazite dating as a tool to detect polymetamorphism in high-temperature metamorphic terrane: Example from the Aoyama area, Ryoke metamorphic belt, Southwest Japan. Island Arc, 20, 439-453.
•  Kawakami,  T.,  Yamaguchi,  I.,  Miyake,  A.,  Shibata,  T., et al (2013) Behavior of zircon in the upper-amphibolite to granulite facies schist/migmatite transition, Ryoke metamorphic belt, SW Japan: constraints from the melt inclusions in zircon. Contributions to Mineralogy and Petrology, 165, 575-591.
•  Kawakami,  T.,  Horie,  K.,  Hokada,  T.,  Hattori,  K. and  Hirata,  T. (2019) Disequilibrium REE compositions of garnet and zircon in migmatites reflecting different growth timings during single metamorphism (Aoyama area, Ryoke belt, Japan). Lithos, 338-339, 189-203.
•  Kawakami,  T.,  Ichino,  T.,  Kazuratachi,  K.,  Sakata,  S. and  Takatsuka,  K. (2022) Multistage zircon growth recording polyphase metamorphic evolution caused by pulsed granitoid intrusions into a low-P/T type metamorphic belt: P-T-D-t evolution of migmatites in the Ryoke belt, southwest Japan. Island Arc, 31, e12454.
•  Kawakami,  T.,  Satish-Kumar,  M.,  Mitsubori,  T.,  Silpa,  A.S., et al (2025) Origin of retrograde saline fluids during collision zone high-temperature metamorphism constrained from whole-rock sulfur isotope composition, the Sør Rondane Mountains, East Antarctica. Polar Science, doi:10.1016/j.polar.2025.101213.
•  Kerrick,  D.M. (1987) Fibrolite in contact aureoles of Donegal, Ireland. American Mineralogist, 72, 240-254.
•  McLelland,  J.,  Morrison,  J.,  Selleck,  B.,  Cunningham,  B., et al (2002) Hydrothermal alteration of late- to post-tectonic Lyon Mountain Granitic Gneiss, Adirondack Mountains, New York: Origin of quartz-sillimanite segregations, quartz-albite lithologies, and associated Kiruna-type low-Ti Fe-oxide deposits. Journal of Metamorphic Geology, 20, 175-190.
•  Miyashiro,  A. (1961) Evolution of metamorphic belts. Journal of Petrology, 2, 277-311.
•  Nakajima,  T. (1994) The Ryoke plutonometamorphic belt: Crustal section of the Cretaceous Eurasian continental margin. Lithos, 33, 51-66.
•  Nakamura,  M. and  Watson,  E.B. (2001) Experimental study of aqueous fluid infiltration into quartzite: Implications for the kinetics of fluid redistribution and grain growth driven by interfacial energy reduction. Geofluids, 1, 73-89.
•  Okudaira,  T. (1996) Temperature-time path for the low-pressure Ryoke metamorphism, Japan, based on chemical zoning in garnet. Journal of Metamorphic Geology, 14, 427-440.
•  Okudaira,  T.,  Hara,  I.,  Sakurai,  Y. and  Hayasaka,  Y. (1993) Tectono-metamorphic processes of the Ryoke belt in the Iwakuni-Yanai district, southwest Japan. Memoirs of the Geological Society of Japan, 42, 91-120.
•  Okudaira,  T.,  Beppu,  Y.,  Yano,  R.,  Tsuyama,  M. and  Ishii,  K. (2009) Mid-crustal horizontal shear zone in the forearc region of the mid-Cretaceous SW Japan arc, inferred from strain analysis of rocks within the Ryoke metamorphic belt. Journal of Asian Earth Sciences, 35, 34-44.
•  Okudaira,  T.,  Kawakami,  T.,  Ikeda,  T. and  Skrzypek,  E. (2024) Inside the Ryoke Magmatic Arc: Crustal deformation, high-T metamorphism and magmatic pulses. Elements, 20, 96-102.
• Ozaki, M., Sangawa, A., Miyazaki, K., Nishioka, Y., et al. (2000) Geology of the Nara District. With Geological Sheet Map at 1:50,000, Geological Survey of Japan, pp. 162 (in Japanese with English abstract pp. 5).
•  Skrzypek,  E.,  Kawakami,  T.,  Hirajima,  T.,  Sakata,  S., et al (2016) Revisiting the high temperature metamorphic field gradient of the Ryoke Belt (SW Japan): new constraints from the Iwakuni-Yanai area. Lithos, 260, 9-27.
•  Skrzypek,  E.,  Kato,  T.,  Kawakami,  T.,  Sakata,  S., et al (2018) Monazite behaviour and time-scale of metamorphic processes along a low-pressure/high-temperature field gradient (Ryoke belt, SW Japan). Journal of Petrology, 59, 1109-1144.
•  Suzuki,  K. and  Adachi,  M. (1998) Denudation history of the high T/P Ryoke metamorphic belt, southwest Japan: constraints from CHIME monazite ages of gneisses and granitoids. Journal of Metamorphic Geology, 16, 23-37.
•  Takatsuka,  K.,  Kawakami,  T.,  Skrzypek,  E.,  Sakata,  S., et al (2018a) Spatiotemporal evolution of magmatic pulses and regional metamorphism during a Cretaceous flare-up event: constraints from the Ryoke belt (Mikawa area, central Japan). Lithos, 308-309, 428-445.
•  Takatsuka,  K.,  Kawakami,  T.,  Skrzypek,  E.,  Sakata,  S., et al (2018b) Age gap between the intrusion of gneissose granitoids and regional high-temperature metamorphism in the Ryoke belt (Mikawa area, central Japan). Island Arc, 27, e12224.
•  Vernon,  R.H. (1979) Formation of late sillimanite by hydrogen metasomatism (base-leaching) in some high-grade gneisses. Lithos, 12, 143-152.
•  Wang,  G.F.,  Banno,  S. and  Takeuchi,  K. (1986) Reactions to define the biotite isograd in the Ryoke metamorphic belt, Kii Peninsula, Japan. Contributions to Mineralogy and Petrology, 93, 9-17.
•  Warr,  L.N. (2021) IMA - CNMNC approved mineral symbols. Mineralogical Magazine, 85, 291-320.