text
stringlengths 2.01k
330k
| filename
stringlengths 12
156
|
|---|---|
MINERALOGY 2022 Journal of Mineralogical andPetrological Sciences(2022) 117:004 LETTER Geochemical characteristics of an ophiolitic complex from Mt. Tenzan area, Saga Prefecture, northern Kyushu Yusaku TANAKA, Keisuke EsHIMA and Masaaki OwADA Graduate school of Sciences and Technology for Innovation, Yamaguchi University, Yamaguchi 753-8512, Japan The metamorphic complex from the Mt. Tenzan area in northern Kyushu consists mainly of mafic rocks with small amounts of siliceous, calc-silicate, and ultramafic rocks. These lithofacies can be recognized as an ophio- litic complex. Metamorphosed mafic rocks are divided into two types, amphibolites I and II, which are probably derived from supracrustal and intrusive rocks, respectively. The geochemical data of both amphibolites plotted within the field between mid-ocean ridge and island arc basalts; such geochemical features resemble those of back-arc basin basalts. As the metamorphic complex was intruded by Cretaceous granitoids, protoliths of the complex could have been formed prior to the Cretaceous. The protolith lithofacies assemblage and geochemical constraints of the Tenzan metamorphic complex indicate the correlation with the Yakuno ophiolite rather than the Oeyama ophiolite. Keywords: Whole-rock geochemistry, Metamorphosed mafic rocks, Ophiolite complex, Tenzan Area, North- ern Kyushu INTRODUCTION The metamorphic rocks from northern Kyushu occur as large blocks in the Cretaceous granitoids. The protoliths of these metamorphosed rocks are composed of pelitic, calc-silicate, and mafic rocks, with small amounts of lime- stone and ultramafic rocks (Karakida et al., 1969). The Figure 1. (a) Location of northern Kyushu, southwest Japan. (b) zircon uranium-lead (U-Pb) dating of the metamorphosed Regional distribution map of metamorphic rocks from northern clastic rocks from northern Kyushu indicates that the ages Kyushu (modified after Kubo et al., 1993; Hoshizumi et al., of the youngest peaks obtained from detrital zircon cores 2004) with the youngest peak ages obtained from the detrial are approximately 400 and 250 Ma, which correspond to zircon cores from clastic rocks. (c) Simplified geological map the Renge and Suo high-pressure metamorphic belts, re- of the Tenzan area. The zircon U-Pb data are from Tsutsumi et al. (2003, 2011), Adachi et al. (2012), Miyazaki et al. (2017), spectively (Fig. 1; Tsutsumi et al., 2003, 2011; Adachi et and Yuhara et al. (2021). al., 2012; Miyazaki et al., 2017; Yuhara et al., 2021). The Tenzan area in Saga Prefecture is underlain by metamorphosed mafic rocks with small amounts of sili- zan area. Based on the metamorphic pressure-temperature ceous, calc-silicate, and ultramafic rocks (Oshima, 1964) conditions and lithological features, Yamada et al. (2008) (Fig. 1). According to the previous study (Nishimura. correlated the metamorphic rocks from the Tenzan area 1998), the Tenzan area has been thought to be a member with those of the Higo metamorphic rocks without any of the Renge belt because the regional foliations are con- whole-rock analyses. However, the lithological assem- tinued to the high-pressure/low-temperature metamorphic blage of the Tenzan area is equivalent to an ophiolite com- rocks in the Sasaguri area, 40 km north-east from the Ten- plex defined by Ishiwatari (2010), which indicates at least two members of the following lithofacies: 1) mantle peri- doi:10.2465/jmps.210831 dotite, 2) mafic and/or ultramafic cumulate, and 3) mafic M. Owada, owada@ yamaguchi-u.ac.jp Corresponding author Y. Takana, alter8069 @ gmail.com volcanic rocks. This study examined the mineralogical @ 2022 Japan Association of Mineralogical Sciences and whole-rock geochemical characteristics of metamor- 2 Y. Tanaka, K. Eshima and M. Owada phosed mafic and ultramafic rocks from the Tenzan area Amphibolite II is coarse-grained and is composed of and compared our geochemical results with those of other hornblende and plagioclase with small amounts of biotite, Paleozoic ophiolite complexes in southwest Japan. This is potassium feldspar, and quartz, containing ilmenite, titan- the first geochemical analysis of metamorphosed mafic ite, and apatite as accessory minerals (Fig. 2b). Horn- and ultramafic rocks in this area. The data presented here blende rarely developed zonal structure and frequently provide useful information for geotectonic comparisons contains poikilitic plagioclase inclusions. Potassium feld- with Paleozoic ophiolite complexes in southwest Japan. spar appears in the plagioclase grains as an antiperthitic phase. FIELD OCCURRENCE AND PETROLOGICAL Serpentinite is composed of mesh-structured serpen- DESCRIPTIONS tine, talc, tremolite, clinopyroxene, and phlogopite (Fig. 2c). We performed a chemical analysis of the olivine and In terms of field occurrence, the metamorphosed mafic chromian spinel in serpentinite using a JEOL JXA-8230 'I pe I sooque sd on ou ppp are so electron probe micro-analyzer (EPMA) at the Center for Amphibolite I is fine-grained and well-foliated, with thin Instrumental Analyses, Yamaguchi University. Analytical layers and lenses of siliceous and calc-silicate gneisses. procedures of mineral chemistry were performed as de- In contrast, amphibolite II is coarse-grained with weak scribed by Eshima (2021). Olivine (F092-90) appears lo- foliation locally accompanied by serpentinite and tremo- cally as a relict mineral and includes chromian spinel lite rocks as lenticular bodies or lenses (Fig 1c). Based on with a dark brownish color. Its chrome number [Cr# = textural evidence, the mafic and ultramafic rocks in this Cr/(Cr + Al)] ranges from 0.85 to 0.89. The tremolite area underwent a thermal effect at the boundary with the displays columnar shapes with a maximum length of Cretaceous granitoids, but outside of the boundary, the 2.0 cm. The phlogopite exhibits fake crystals surround- rocks were barely affected by contact metamorphism ing the olivine crystals. Mineral data for olivine and chro- (Yamada et al., 2008). mian spinel are available upon request from the corre- Amphibolite I has a fine-grained and nematoblastic sponding authors. texture. It consists mainly of hornblende and plagioclase The tremolite rock primarily consists of tremolite with small amounts of quartz and includes ilmenite, titan- and clinopyroxene with small amounts of serpentine, ite, and apatite as accessory minerals (Fig. 2a). The horn- chlorite, and opaque minerals (Fig. 2d). The tremolite blende has a zonal structure with a light-green core, a is weakly foliated. greenish-brown mantle, and a light-green rim. In addi- tion to these minerals, the calc-silicate thin layers inter- WHOLE-ROCK GEOCHEMISTRY calated with amphibolite I include clinopyroxene, epi- dote, and calcite. The whole-rock chemical compositions of amphibolites I and II, serpentinite, and tremolite rocks were determined by X-ray fuorescence (XRF) analysis. The analyzed samples were free from alteration and veins, and the amounts of 300-600 g were crushed by hand with a W-mortar. Crushed samples were powdered using an au- tomatic W-mortar. Detailed analytical procedures are de- scribed in Eshima and Owada (2018). After ignition in a furnace at 950 °C for 2 h, the samples were measured for loss on ignition; the ignited samples (1.0 g) were mixed with five times the amount of Li2B4O7 as flux. The mixed samples were melted to make glass beads using a bead sampler. The glass beads were analyzed using an XRF analyzer (Rigaku ZSX primus-II) installed at the Center of Instrumental Analysis, Yamaguchi University. The an- 1.0mm alytical conditions included an electric voltage of 50 kV and an electric current of 60 mA, using a Rh anode X-ray Figure 2. Thin section photographs. Plane-polarized light image tube. The analyzed elements were SiO2, TiO2, Al2O3, of (a) amphibolite I, (b) amphibolite I, and (c) serpentinite. Fe2O3, MnO, MgO, CaO, Na2O, K2O, and P2O5 as the Cross-polarized light image of (d) tremolite rock. Hbl, horn- blende; Pl, plagioclase; Ol, olivine; Cpx, clinopyroxene; Tlc, major elements and were Ba, Cr, Nb, Ni, Rb, Sr, V, Y, Zn, talc; Srp, serpentine; Tr, tremolite; Spl, spinel. and Zr as trace elements. To validate the quantitative da- Geochemical characteristics of an ophiolitic complex from Tenzan area 3 Table 1. Chemical compositions of GSJ standard samples, JB-2, Table 2. Representative whole-rock chemical composition of met- JB-3, and JGb-2 amorphosed mafic rocks from the Tenzan area Sample JB-2 JB-3 JGb-2 Type Amp-I Amp-I Amp-I Amp-I Amp-II Amp-II Sample 1903 1903 1903 1903 1905 1905 r.v. m.v. r.v. m.v. r.v. m.v. no. were measured using fused glass beads and a 213 nm 1401 2601 2603 crimination diagram using Nb-Zr-Y (Meschede, 1986) 0302B (wt%) SiO2 52.52 50.77 51.27 47.13 46.73 (wt%) 52.96 SiO2 1.43 0.57 0.55 49.14 50.11 48.09 48.83 50.85 49.28 TiO2 1.18 1.18 1.42 Al2O3 14.56 14.45 17.13 17.09 23.81 23.51 TiO2 1.39 1.58 1.40 1.48 shows a LREE-enriched pattern with an overall 10-350 1.06 Fe2O3 14.17 11.99 Al2O3 15.25 15.06 15.29 15.31 17.33 14.46 11.78 6.78 6.83 18.52 MnO 0.22 0.22 0.18 0.18 0.13 0.13 Fe2O3 9.62 10.46 10.57 10.37 9.70 shows that the IWY-Q2 plot between trachyte to phonolite MnO 4.59 4.57 5.17 5.37 6.27 6.18 0.16 0.17 0.18 0.19 0.16 0.15 MgO 9.77 MgO 7.27 7.44 9.86 9.75 9.76 14.33 8.32 7.29 5.69 5.43 CaO 14.30 Na2O 2.03 1.91 2.72 2.68 0.93 0.92 CaO 13.62 11.82 11.75 12.25 9.39 13.27 K20 Na2O 0.42 0.43 0.78 0.78 0.06 0.06 2.74 Zircon grains were separated from rock powder by 206pb age of 1099.0 ± 0.6 Ma; Paces and Miller, 1993) 2.95 2.97 fields, samples from Akiyoshi-dai plot in basalt field, and P2O5 0.10 0.10 0.29 0.31 0.02 0.03 K2O 0.21 0.20 (with intergrown quartz and Fe-oxide) are anomalous in 0.21 0.80 0.63 100.0099.27 P2O5 0.13 0.17 0.12 0.15 0.10 0.12 Total 100.0099.70 100.00100.85 standard of JEOL was used for the analysis of Hf (HfO2 = 0.44 0.34 0.91 0.75 1.55 (ppm) 1.19 Ba 222 213 245 273 37 60 Total 99.95 100.70 99.62 99.77 99.29 100.62 Cr 28 25 58 63 125 115 (ppm) 2 Ba 45 10 15 59 123 Nb 1 3 3 2 1 131 Ni 17 2 36 31 14 10 Cr 350 254 324 268 100 165 Rb 7 5 15 11 3 3 Ni 77 54 75 83 39 47 IS 180 403 407 438 438 Rb 4 5 4 7 27 178 16 V 575 576 382 V 372 174 174 256 295 271 277 272 212 Y 25 26 27 24 5 Zn 7 67 73 76 73 70 56 ICP-MS Zn 108 110 100 103 49 47 Zr 51 51 98 100 12 13 Sr 255.00 218.00 225.00 274.00 251.00 221.00 Y 26.20 27.60 26.70 25.60 15.10 21.20 r.V., recommended value; m.v., measured value; GSJ, Geological Zr 82.00 94.00 78.00 88.00 41.00 67.00 Survey of Japan. Nb 2.00 15 μm of laser ablation beam diameter. 1.50 3.60 0.40 0.80 La 4.39 4.90 3.53 5.59 5.44 7.90 Ce 11.80 13.80 10.40 14.10 12.40 19.30 ta, we measured the standard rock samples, JB-2, JB-3, Pr 2.01 2.13 1.78 2.28 1.90 2.87 and JGb-2, provided by the Geological Survey of Japan. PN 10.80 11.50 9.74 11.40 8.95 14.30 The results are presented in Table 1. The measured values Sm 3.47 3.75 3.29 3.57 2.47 3.71 of the standard samples were identical to the recommend- Eu 1.23 program Pepi-AGE (Dunkl et al., 2008), and final statis- 1.26 1.27 0.86 1.26 ed values. In addition to XRF analyses, trace elements, Gd 3.89 4.21 3.86 3.80 2.59 3.72 including rare earth elements (REEs), for amphibolites I Tb 0.75 0.77 0.72 0.72 0.45 0.66 and II were determined by inductively coupled plasma Dy 4.91 4.91 4.90 4.74 2.81 3.97 mass spectrometry (ICP-MS) at Activation Laboratory Ho 0.99 1.05 1.03 0.98 0.58 0.80 Ltd., Canada. Er 2.85 2.94 2.99 2.78 1.68 2.36 A total of 34 samples (15 of amphibolite I, 13 of Tm 0.41 0.44 0.43 0.40 0.24 0.35 amphibolite Il, 4 of serpentinites, and 2 of tremolite Yb 2.59 2.83 2.68 2.52 1.48 2.22 rocks) were analyzed by XRF, and 9 samples of amphib- Lu 0.39 0.42 Zircon U-Pb isotope analysis was performed using a 0.38 0.22 0.34 olites I and II were analyzed using ICP-MS. The repre- JH 2.20 2.50 2.10 2.20 1.30 2.00 sentative results are presented in Table 2. All of the Amp-I, amphibolite I; Amp-II, amphibolite II; LOI, loss on igni- whole-rock data used in this study can be requested from tion. the corresponding authors. Figure 3 shows the total alkali versus silica (TAS) and FeO*/MgO versus SiO2 wt% di- agrams (FeO* = 0.9 × Fe2Os*). Amphibolites I and II contents show that the geochemical features of amphib- belong to the subalkaline and tholeite series, respectively olite I are more evolved than those of amphibolite I (Fig. (Figs. 3a and 3b). The FeO*/MgO ratios and Cr (ppm) 3b, Table 2). Y. Tanaka, K. Eshima and M. Owada a Basalt Mg=Mg/ Mg+Fe2)0.2 St Si02 1% Figure 4. Chemical composition of chromian spinel and olivine in SiO2 wt% (b) the serpentinite from the Tenzan area. (a) Compositional rela- TH Cr#. Olivine spinel mantle array (OSMA) is after Arai (1994). CA (b) Compositional relationships between Mg# and Cr# of chro- mian spinel. The data of this study are plotted as the average value from one sample. The compositional ranges of the data are shown as bars accompanied with data plots. The mantle perido- tites from the Oeyama (Arai, 1980; Tsujimori, 1998; Machi and Ishiwatari, 2010) and Yakuno ophiolites (Ishiwatari, 1985a, b) are shown for comparison. Figure 3. (a) SiO2 wt% versus Na2O + K2O wt% (TAS) diagram. (b) FeO*/MgO versus SiO2 wt% diagram (Miyashiro, 1974) those of the Yakuno mantle peridotite (Cr# = 0.6-0.8) showing tholeitic (TH) and calc-alkaline (CA) fields. FeO* = 0.9 × Fe2O3*. than to those of the Oeyama peridotite (Cr# = 0.3-0.5) (Fig. 4). Figure 5a shows the compositional range of primi- GEOTECTONIC SETTING OF TENZAN tive basaltic magmas with their differentiation trends and METAMORPHIC COMPLEX the accumulation directions of specific minerals (Kemp- ton et al., 1997). Kempton et al. (1997) stressed that met- Amphibolite I intercalates with thin layers and lenses of amorphosed mafic rocks up to amphibolite facies grade siliceous and calc-silicate gneisses. In contrast, amphib- should be adopted in this diagram to determine the geo- olite II contacts with lenticular bodies or lenses of ser- chemical compositions of their protoliths. Amphibolites I pentinite and tremolite rocks, and they probably occur as and II studied here underwent amphibolite facies meta- xenoblocks (Fig. 1c). Based on the field occurrence, pro- morphism (Yamada et al., 2008). Therefore, the geo- toliths of amphibolites I and II would, therefore, be su- chemical compositions of the studied samples can be pracrustal and intrusive rocks, respectively. The litholog- adopted to the diagram of Kempton et al. (1997) for es- ical assemblage, including amphibolites I and II and timating the characteristics of the protoliths. The ana- serpentinite in the Tenzan area, is equivalent to an ophio- lyzed samples of amphibolites I and II generally show lite complex as defined by Ishiwatari (2010). Because the evolved compositions, with some samples of amphibolite samples from the Tenzan area were intruded by Creta- I plotted in the primitive basaltic field (Fig. 5a); however, ceous granitoids, they can be recognized as a pre-Creta- the samples do not show accumulation trends. Therefore, ceous ophiolitic complex. Therefore, the metamorphic the geochemical data for amphibolite I and II reflect the complex from the Tenzan area can be compare to geo- liquid composition. The mafic rocks from the Yakuno chemical features of the Oeyama or Yakuno ophiolites. ophiolite also possess the similar compositional ranges, The rock types of the Oeyama ophiolite are dominat- but those from the Oeyama ophiolite are plotted outside ed by mantle peridotite and cumulate rocks (Kurokawa, the liquid compositions. Figure 5b depicts the Mg# ver- 1985). In contrast, the Yakuno ophiolite includes supra- sus TiO2 wt% diagram showing the data from the Tenzan crustal rocks in addition to cumulate rocks and mantle area, the Yakuno and the Oeyama ophiolites (Tsujimori peridotites (Ishiwatari, 1985a). Figure 4 shows the chemi- and Ishiwatari, 2002; Ichiyama and Ishiwatari, 2004; Su- cal composition of chromian spinels and olivines in ser- da and Hayasaka, 2009; Kimura and Hayasaka, 2019). pentinite from the Tenzan area and Oeyama and Yakuno All data were plotted within the field, from mid-ocean ophiolites. Based on the Cr# of chromian spinel and the ridge basalt (MORB) to island arc basalt (IAB). A dia- Fo value of coexisting olivine, the serpentinite in the Ten- gram of the N-MORB-normalized La/Y and Nb/La ratios zan area is a highly depleted mantle peridotite, probably are plotted for amphibolites I and II and the mafic rocks harzburgite or dunite (Fig. 4a). These values are closer to from the Yakuno and Oeyama ophiolites, as well as the Geochemical characteristics of an ophiolitic complex from Tenzan area (b) that the crustal evolution of the Yakuno ophiolite was characterized by intra-oceanic island arc and back-arc basin settings (Figs. 5c and 5d). Amphibolites I and II from the Tenzan area could have experienced the crustal AL evolution in the island arc and back-arc basin settings similar to the Yakuno ophiolite. SiO2 / A1Os (c) ACKNOWLEDGMENTS This work was supported by JSPS KAKENHI Grant Numbers JP15H03748 to M. Owada and JP21J13600 to 0.4 人。 K. Eshima. We wish to thank Dr. K. Itano, two anonymous (La/Y) reviewers, and Dr. Y. Ichiyama, the associate editor. Their comments were useful for improving the manuscript. Rock/C1chondrit (e) REFERENCES Adachi, T., Osanai, Y., Nakano, N. and Owada, M. (2012) LA- ICP-MS U-Pb zircon and FE-EPMA U-Th-Pb monazite dat- <LLD Figure 5. Whole-rock major and trace element compositions of the Mountains, northern Kyushu. Journal of Geological Society <LLD of Japan, 118, 39-52. <LLD Arai, S. (1980) Dunite-Harzburgite-Chromitite complexes as re- Mg# versus TiO2 wt% diagram. (c) N-MORB-normalized La/ fractory residue in the Sangun-Yamaguchi zone, western Ja- Nb versus La/Y diagram (Suda et al., 2014). (d) N-MORB nor- pan. Journal of Petrology, 21, 141-165. malized spider diagram. (e) C1 chondrite-normalized REE pat- Arai, S. (1994) Characterization of spinel peridotites by olivine- 213.4 spinel compositional relationships: Review and interpretation. 2002; Kimura and Hayasaka, 2019) and Yakuno (Ichiyama and Chemical Geology, 113, 191-204. Ishiwatari, 2004; Suda and Hayasaka, 2009) ophiolites are also Eshima, K. (2021) Anatomy of Shaku-dake high-Mg diorite, shown. C1-chondrite and N-MORB of Sun and McDonough southwest Japan: Lithofacies variations and growth process (1989) were used for normalization. OIB, oceanic island basalt; of high-Mg diorite stock. Journal of Mineralogical and Petro- MORB, mid-ocean ridge basalt; IAB, island-arc basalt; BABB, logical Sciences, 116, 83-95. back-arc basin basalt. Compositional felds of BABB in Figure Eshima, K. and Owada, M. (2018) Whole-rock geochemistry of 5c from Lau (Fallen et al., 1992; Pearce et al., 1994), Manus 47.64 216.5 Kyushu. Journal of Geological Society of Japan, 124, 857- 1990) are shown for comparison. Additionally, the compositions 862 (in Japanese with English abstract). of E-MORB, marked “E' (Sun and McDonough, 1989) and IAB Fallen, T.J., Malahoff, A., Zonenshaina, L.P. and Bogdanova, Y. (Tatsumi, 2000) are plotted for comparison. (1992) Petrology and geochemistry of back-arc basin basalts from Lau Basin spreading ridges at 15°, 18°, and 19°S. Min- eralogy and Petrology, 47, 1-35. Hawkins, J.W., Lonsdale, P.F., Macdougall, J.D. and Volpe, A.M. compositions of the IAB, E-MORB with the composi- (1990) Petrology of the axial ridge of the Mariana Trough tional field of back-arc basin basalts (BABB) compiled backarc spreading center. Earth and Planetary Science Letter, 100,226-250 by Suda et al. (2014) (Fig. 5c). Amphibolites I and II Hoshizumi, H., Ozaki, M., Miyazaki, K., Matsuura, H., et al. from the Tenzan area and the mafic rocks from the Yaku- (2004) Geological map of Japan 1:200,000, Kumamoto. Geo- no ophiolite are plotted between N-MORB and IAB logical survey of Japan (in Japanese). compositions, and many samples exhibit BABB affnity Ichiyama, Y. and Ishiwatari, A. (2004) Petrochemical evidence for (Fig. 5c). The spider diagram shows that amphibolites I off-ridge magmatism in a back-arc setting from the Yakuno ophiolite, Japan. Island arc, 13, 157-177. and II have negative Nb anomalies (Fig. 5d). The two Ishiwatari, A. (1985a) Granulite-facies metacumulates of the Yaku- samples of amphibolite I plotted near the composition no ophiolite, Japan: evidence for unusually thick oceanic crust. of IAB in Figure 5c display a significant negative Nb Journal of Petrology, 26, 1-30. anomaly in the spider diagram (Fig. 5d) and rightward Ishiwatari, A. (1985b) Igneous Petrogenesis of the Yakuno Ophio- decrease in the REE patterns similar to those of the met- lite (Japan) in the context of the diversity ophiolites. Contri- butions to Mineralogy and Petrology, 89, 155-167. abasalt from the Yakuno ophiolite (Fig. 5e). This implies Ishiwatari, A. (2010) New developments in ophiolites studies. that amphibolites I and II formed in island arc and/or Journal of Geography, 119, 841-851 (in Japanese with Eng- back-arc basin environments. Suda et al. (2014) inferred lish abstract). 6 Y. Tanaka, K. Eshima and M. Owada Karakida, Y., Yamamoto, H., Miyaji, S., Oshima, T. and Inoue, T. Yakuno Ophiolite, Southwest Japan. Journal of Geological (1969) Characteristics and geological situation of metamor- -qe us m sr ) -9 1 r eo r phic rocks in Kyushu. Memoir of Geological Society of Ja- stract). pan, 4, 3-21 (in Japanese with English abstract). Suda, Y., Hayasaka, Y. and Kimura, K. (2014) Crustal evolution of Kempton, P.D., Downes, H. and Embey-Isztin, A. (1997) Mafic a Paleozoic intra-oceanic island-arc-back-arc basin system granulite xenoliths in Neogene alkali basalts from the western constrained by the geochemistry and geochronology of the Pannonian Basin: insights into the lower crust of a collapsed Yakuno ophiolite, southwest Japan. Journal of Geological Re- orogen. Journal of Petrology, 38, 941-970. search, 2014. doi.org/10.1155/2014/652484. Kimura, K. and Hayasaka, Y. (2019) Zircon U-Pb age and Nd iso- Sun, S.-S. and McDonough, W.F. (1989) Chemical and isotopic tope geochemistry of latest Neoproterozoic to early Paleozoic systematics of oceanic basalts: implications for mantle compo- Oeyama ophiolite: Evidence for oldest MORB-type oceanic sition and processes. In Magmatism in oceanic basins (Saun- ders, A.D. and Norry, M.J., Eds.). Geological Society of Lon- cations. Lithos, 342-343, 345-360. don Special Publication, 42, Geological Society of London, Kubo, K., Matsuura, H., Ozaki, M., Makimoto, H., et al. (1993) In previous studies, Visean fusulinacean, conodont, bra- Geological map of Japan 1:200,000, Fukuoka. Geological sur- Tatsumi, Y. (2000) Continental crust formation by crustal delami- Trace element concentrations of 10 points from 6 zircon nation in subduction zones and complementary accumulation Kurokawa, K. (1985) Petrology of the Oeyama ophiolitic complex of the enriched mantle I component in the mantle. Geochem- in the inner zone of Southwest Japan. Science Report of Nii- istry Geophysics Geosystems, 1, 2000GC000094. https://doi. gata University, Series E (Geology and Mineralogy), 6, 37- org/10.1029/2000GC000094. 113. Tsujimori, T. (1998) Geology of the Osayama serpentinite melange Machi, S. and Ishiwatari, A. (2010) Ultramafic rocks in the Kotaki in the central Chugoku Mountains, southwestern Japan: area, Hida Marginal Belt, central Japan: peridotites of the 320Ma blueschist-bearing serpentinite melange beneath the Oeyama ophiolite and their metamorphism. Journal of Geo- Oeyama ophiolite. Journal of Geological Society of Japan, logical Society of Japan, 116, 293-308 (in Japanes with Eng- 104, 213-231 (in Japanese with English abstract). lish abstract). Tsujimori, T. and Ishiwatari, A. (2002) Granulite facies relics in the Miyashiro, A. (1974) Volcanic rock series in island arc and active Early Paleozoic kyanite-bearing ultrabasic metacumlate in the continental margins. American Journal of Science, 274, 321- Oeyama belt, the Inner Zone of southwestern Japan. Gondwa- 355. oceanic island zircon chemistry in the Nb/Yb versus Sc/ Miyazaki, K., Ikeda, T., Matsuura, H., Danhara, T., et al. (2017) A Tsutsumi, Y., Yokoyama, K., Terada, K. and Sano, Y. (2003) high-T metamorphic complex derived from the high-P Suo SHRIMP U-Pb dating of detrital zircons in metamorphic metamorphic complex in the Omuta district, northern Kyushu, rocks from northern Kyushu, western Japan. Journal of Min- southwest Japan. Island arc, 26. doi 10.1111/iar.12208. eralogical and Petrological Sciences, 98, 181-193. Nishimura, Y. (1998) Geotectonic subdivision and areal extent of Tsutsumi, Y., Yokoyama, K., Terada, K. and Hidaka, H. (2011) the Sangun belt, Inner Zone of Southwest Japan. Journal of SHRIMP dating of detrital zircons from the Sangun-Renge Metamorphic Geology, 16, 129-140. belt of Sangun metamorphic rocks, northern Kyushu, south- Oshima, T. (1964) A preliminary study of the southern metamor- west Japan. Bulletin of National Museum Natural Science, phic rock belt in the Sefuri Mountains, north Kyushu. Re- strata of the limestones that occurred in the Akiyoshi Ter- search paper of Faculty of Education, Saga University, 12, Yamada, S., Owada, M. and Osanai, Y. (2008) Amphibolite-facies Yakawa area and basalt associated with Hina Limestone metamorphic rocks of the Mt. Tenzan area, northern Kyushu, Pearce, J.A., Ernewein, M., Bloomer, S.H., Parson, L.M., et al. Southwest Japan: Correlation with the Higo metamorphic (1994) Geochemistry of Lau Basin volcanic rocks: infuence rocks. Journal of Mineralogical and Petrological Sciences, of ridge segmentation and arc proximity. In Volcanism asso- zawa, 2018). Moreover, Tazaki et al. (1989) reported an ciated with extension at consuming plate margin (Smellie, Yuhara, M., Kiyoura, K., Hidaka, M., Hokada, T. and Hayasaka, Y. J.L., Ed.). Geological Society of London Special Publication, (2021) The metamorphism of the Tagawa metamorphic rocks, 81, Geological Society of London, London, 53-75. eastern North Kyushu. Journal of Geological Society of Ja- Sinton, J.M., Ford, L.L., Chappell, B. and McCulloch, M.T. (2003) pan, 127, 447-459 (in Japanese with English abstract). Magma genesis and mantle heterogeneity in the Manus back- arc basin, Papua New Guinea. Journal of Petrology, 44, 159- Manuscript received August 31, 2021 195. Manuscript accepted April 15, 2022 Suda, Y. and Hayasaka, Y. (2009) Genesis and evolutional process- Released online publication June 4, 2022 es of the Paleozoic oceanic island arc crust, Asago body of the Manuscript handled by Yuji Ichiyama
|
Tanaka (2022) - geochem characteristics of an ophiolite Complex Tenzan area, Kyushu.txt
|
ELSEVIER Tectonophysics 290 (1998) 197±210
Crustal structure and tectonics of the Hidaka CollisionZone, Hokkaido
(Japan), revealed by vibroseis seismicre¯ection and gravity surveys
KazunoriAritaa,,Takashi Ikawab,T anioItob,AkihikoYamamotoc,MatsuhikoSaitoa,
YasunoriNishidaa,HideyukiSatoha,G a k uK imu r ad,1, Teruo Watanabea, TakeshiIkawae,
ToruKurodae
aDepartment of Earth and Planetary Sciences, Hokkaido University, Kita-ku, Sapporo 060, Japan
bDepartment of Earth Sciences, Chiba University, Inage-ku, Chiba 263, Japan
cResearch Center for Earthquake Prediction, Hokkaido University, Kita-ku, Sapporo 060, Japan
dDepartment of Earth Sciences, CIAS, University of Osaka Prefecture, Sakai-shi, Osaka 593, Japan
eJapex Geoscience Institute Inc., Shinagawa-ku, Tokyo 140, Japan
Received 21 November 1996; accepted 16 December 1997
Abstract
This study is the ®rst integrated geological and geophysical investigation of the Hidaka Collision Zone in southern
Central Hokkaido, Japan, which shows complex collision tectonics with a westward vergence. The Hidaka Collision Zone
consists of the Idon'nappu Belt (IB), the Poroshiri Ophiolite Belt (POB) and the Hidaka Metamorphic Belt (HMB) with the
Hidaka Belt from west to east. The POB (metamorphosed ophiolites) is overthrust by the HMB (steeply eastward-dippingpalaeo-arc crust) along the Hidaka Main Thrust (HMT), and in turn, thrusts over the Idon'nappu Belt (melanges) along
the Hidaka Western Thrust (HWT). Seismic re¯ection and gravity surveys along a 20-km-long traverse across the southern
Hidaka Mountains revealed hitherto unknown crustal structures of the collision zone such as listric thrusts, back thrusts,frontal thrust-and-fold structures, and duplex structures. The main ®ndings are as follows. (1) The HMT, which dips steeply
at the surface, is a listric fault dipping gently at a depth of 7 km beneath the eastern end of the HMB, and cutting across
the lithological boundaries and schistosity of the Hidaka metamorphic rocks. (2) A second re¯ector is detected 1 km belowthe HMT re¯ector. The intervening part between these two re¯ectors is inferred to be the POB, which is only little exposed
at the surface. This inference is supported by the high positive Bouguer anomalies along the Hidaka Mountains. (3) The
shallow portion of the IB at the front of the collision zone has a number of NNE-dipping re¯ectors, indicative of imbricatedfold-and-thrust structures. (4) Subhorizontal re¯ectors at a depth of 14 km are recognized intermittently at both sides of the
seismic pro®le. These re¯ectors may correspond to the velocity boundary (5.9±6.6 km =s) previously obtained from seismic
refraction pro®ling in the northern Hidaka Mountains. (5) These crustal structures as well as the back thrust found in theeastern end of the traverse represent characteristics of collisional tectonics resulting from the two collisional events since
the Early Tertiary. 1998 Elsevier Science B.V. All rights reserved.
Keywords: crustal structure; Hidaka Collision Zone; seismic re¯ection; gravity survey; collision tectonics; Hokkaido
Corresponding author. Tel.: C81 11 706-5305; Fax: C81 11 706-5305; E-mail: arita@cosmos.sci.hokudai.ac.jp
1Present address: Geological Institute, University of Tokyo, Bunkyo-ku, Tokyo 113, Japan.
0040-1951/98/$19.00 1998 Elsevier Science B.V. All rights reserved.
PIIS0040-1951(98)00018-3
198 K. Arita et al./Tectonophysics 290 (1998) 197±210
1. Introduction
The northern island of Japan, Hokkaido, is sit-
uated at the conjunction of two active island
arc±trench systems; the Northeast Honshu Arc±
Japan Trench and the Kuril Arc±Trench (Fig. 1a).Hokkaido is divided into three major provinces with
regardtoitspre-Neogenegeology,thetrendofwhich
is oblique to these active arc±trench systems: West-ern,CentralandEasternHokkaido(Fig.1b).Western
Hokkaido, which is a northern extension of North-
east Honshu, consists of Jurassic accretionary com-
plexes intruded by Cretaceous granitic intrusions.
Fig. 1. (a) Sketch map showing plate tectonic setting around Japanese islands. Thick lines and dotted line show present and Tertiary plate
boundaries, respectively. (b) Geologic divisions of Central Hokkaido. Pro®ling lines Fig. 1cFig. 2a,b, and 3 are shown. Modi®ed fromKimura (1994). (c) Schematic geological cross-section of southern Central Hokkaido. POBDPoroshiri Ophiolite Belt, HMBDHidaka
Metamorphic Belt, HWTDHidaka Western Thrust, HMTDHidaka Main Thrust, TFDTokachi Fault System. Cross-section line is
shown in (b). Modi®ed from Kimura (1986).Eastern Hokkaido is composed of Late Cretaceous
to Palaeogene forearc sediments of the palaeo-Kuril
arc±trench system (Kiminami and Kontani, 1983).
Central Hokkaido, which geologically continues to
Sakhalin, has a complicated geological assemblage.
It has been occupied by two subduction±accretionsystems between the palaeo-Eurasian and palaeo-
North American Plates; the one in the west (the
Sorachi-Yezo Belt, the Idon'nappu Belt and the Hi-daka Belt) is a N±S-trending, westward-subduct-
ing system of Late Jurassic to Palaeogene age, and
the other in the east (the Tokoro Belt) is an east-
ward-subducting system of Cretaceous age (Fig. 1b
K. Arita et al./Tectonophysics 290 (1998) 197±210 199
and c; Kiminami and Kontani, 1983; Sakakibara,
1986).
CentralHokkaido,especiallyitssouthernpart,has
undergone two stages of collisions since the Early
Tertiary. The older oblique collision with a right-
lateral sense of motion between the palaeo-Eurasianandpalaeo-NorthAmericanPlatesduringthePalaeo-
gene resulted in the amalgamation of the Hidaka
MetamorphicBeltintheeastandthePoroshiriOphi-olite Belt in the west. The younger collision was
caused by the westward migration of the Kuril fore-
arc plate which started in the Late Miocene due to
the oblique subduction of the Paci®c Plate along the
Kuril Trench (Fig. 1a; Kimura, 1986). The collisionrapidly uplifted the Hidaka Metamorphic Belt as
well as the Poroshiri Ophiolite Belt, producing the
present Hidaka Mountains. The process of collisionbetween the Eurasian Plate and the North American
Plate(Kurilforearcplate)isstillcontinuing.
Although the deformational features due to the
above two collisions are also observed in the So-
rachi-YezoBelt,the HidakaMountains andtheirsur-roundings in southern Central Hokkaido show typi-
callycomplexcollisionalfeatures.Theyaretherefore
collectively termed as the Hidaka Collision Zone,whichprovidesfavourable opportunitiesforstudying
the characteristicsand structure of deep crust similar
to the Ivrea zone of the Alps and the Kohistan±
Ladakh arc of the western Himalaya. It is impor-
tant to elucidate the present deep-crustal structureof the Hidaka Collision Zone in order to understand
the arc±arc collisionalmechanism and the formation
processofcontinentalcrust.Towardthisend,wecar-ried out vibroseis seismic re¯ection pro®ling along
a 20-km-long traverse, and gravity measurements in
the southern Hidaka Mountains.
2. Geophysicalcharacteristicsof theHidaka
CollisionZone
2.1. Seismicrefraction
Seismic refraction experiments and natural earth-
quakedatainandaroundtheHidakaMountainshave
revealed the following complicated tectonic features
(e.g., Den and Hotta, 1973; Okada et al., 1973;Takanami, 1982; Fujii and Moriya, 1983; Moriya,
1983; Miyamachi and Moriya, 1984; Furumura andMoriya, 1990; Miyamachiet al.,1994; Moriya et al.,
1994; Ozelet al.,1996; Iwasaki etal.,1998);
(1) Beneath the Hidaka Collision Zone there are
two seismic zones: a shallower zone (20 to 50 km
deep) likely related to the collision of the Eurasian
and North American Plates, and a deeper zone (over70 km deep) related to the subduction of the Paci®c
Plate.
(2) In the southern Hidaka Mountains a low-
velocity zone (5.5 km =s) dips eastward from the
western coast, and reaches a depth of about 30 km
beneaththe Hidaka Mountains (Fig. 2a).
(3) The seismic velocity structure of both sides
of the Hidaka Mountains is relatively clear, but thatbeneath the mountains is monotonous or not deter-
minable (Fig. 2).
(4) Lateral variation in seismic velocity due to
different geological units is found across the Hidaka
Collision Zone. In particular, seismic waves are at-
tenuated under the Idon'nappu Belt on the western
part of the Hidaka Mountains.
(5) The Moho discontinuity is not clearly visible
as it may be deeper than 50 km (Fig. 2a), although
Miyamachi et al. (1994) reported a crustal thickness
of 32 km beneaththe Hidaka Mountains.
2.2. Magnetotellurics
A recent magnetotelluric survey (Ogawa et al.,
1994) across the Hidaka Collision Zone (about 70km north of the present survey area: Fig. 3) shows
that a 2-km high-resistivity (1000±2000 /DELm) layer
of high-grade Hidakametamorphicrocks is followedby a relatively conductive (500±1000 /DELm) layer of
accretionary prism at depths of 5 to 10 km, which
is underlain again by a high-resistivity (30,000 /DELm)
layer of probably high-grade metamorphic rocks.
This suggests an inter®ngered complex structure orcrustal delamination beneath the Hidaka Collision
Zone probably due to the above-mentioned two col-
lisionevents.
3. Geologicaloutlineof theHidaka collisionzone
with specialreferenceto the surveyedarea
The Hidaka Collision Zone is occupied by the
N±S-trending accretionary and melange complexes
consisting of the Idon'nappu and Hidaka Belts (eg.,
200 K. Arita et al./Tectonophysics 290 (1998) 197±210
Fig. 2. (a) Seismic velocity structures (km =s) across the southern Hidaka Mountains (after Moriya et al., 1994). (b) Seismic velocity
structures (km =s) across Central Hokkaido revealed by seismic refraction pro®ling (after Ozel et al., 1996). HMTDHidaka Main Thrust.
Both pro®ling lines are shown in Fig. 1b.
Kiminami and Kontani, 1983; Kiyokawa, 1992;
Kimura, 1994). The Hidaka Mountains (the Hi-daka Metamorphic Belt and the Poroshiri Ophio-
lite Belt) are situated between the Hidaka Belt and
the Idon'nappu Belt (Fig. 1b). The Hidaka Colli-sion Zone presently shows westward vergence and a
curvature with a westward convex shape due to the
above-mentionedcollisionof the westward-plunging
Kurilforearcplate.
The Hidaka Metamorphic Belt presently displays
steeply eastward-tilted metamorphic and magmatic
sequences with a general NNW trend, but originallyithadawest-side-downgeometrybefore theamalga-
mationof theHidakamagmaticarcandthePoroshiriophiolites. It is considered to have formed in the
western part of the Hidaka Belt during the oblique
collision of the palaeo-Eurasian and palaeo-NorthAmerican Plates during the Palaeogene (Komatsu
et al., 1983); this view, however, has been debated
(e.g., Kimura et al., 1983; Komatsu et al., 1989;
Maeda,1990; Toyoshima,1991; MaedaandKagami,
1996). According to radiometric dating (Arita et al.,1993), thepresentsteeplyeastward-dippingstructure
of the Hidaka Metamorphic Belt had formed before
K. Arita et al./Tectonophysics 290 (1998) 197±210 201
Fig. 3. Interpreted resistivity structure across the Hidaka Collision Zone (after Ogawa et al., 1994). Cross-section line is shown in Fig. 1b.
the Middle Miocene. The Hidaka Metamorphic Belt
consists of a lower metamorphic sequence (gran-
ulite-faciesrocks and orthopyroxene tonalites)in thewest and an upper metamorphic sequence (biotite±
muscovite gneisses and schists) in the east (Fig. 4).
It decreases gradually in metamorphic grade from
the granulite facies in the west to the greenschist fa-
cies eastward, and in general grades into the weaklymetamorphosedturbiditesoftheNakanogawaGroup
(the Hidaka Belt) of Palaeogene age (Fig. 4: Os-
anai et al., 1986; Komatsu et al., 1994). Thesemetamorphic rocks have been intruded by a large
amount of various intrusive rocks, e.g., gabbroic
anddioriticrocksandS-typeorthopyroxene tonalites
in the lower metamorphic sequence and cordierite
tonalite and granite in the upper metamorphic se-quence (Komatsu et al., 1986; Shimura et al., 1992).
The granulite-facies rocks in the western part are
highly mylonitized near the Hidaka Main Thrust(HMT), along which the Hidaka Metamorphic Belt
overthrusts the Poroshiri Ophiolite Belt to the west.
The mylonites suffered dextral ductile shear defor-
mation under the conditions of greenschist facies
(Arita etal.,1986; Toyoshima, 1991).
In the surveyed area the mylonitized granulites
andtonaliteshavestrongmetamorphicandmyloniticfoliations striking N30ë±50ëW and steeply dipping
to the east. The cordierite tonalite, which occupies
the crestline of the Hidaka Mountains, is massive,but has a weak foliation striking N20ë±40ëW and
dipping steeply eastward on the margin. The upper
metamorphic sequence has the same strike and dip
as those of the tonalite. The Nakanogawa Group
shows a monotonous lithology consisting of slateand shale withsome intercalationsof sandstone. The
turbidites dip steeply east, being repeated by folding
and verticalfaulting.
The Poroshiri Ophiolite Belt is composed of
faultedandtightlyfoldedmetamorphosedophiolites,
the original succession of which has been recon-
structed from a basalt to harzburgite tectonite with
a total thickness of at least 5 km (Miyashita, 1983).TherocksofthePoroshiriOphioliteBeltdisplayalot
ofintenseductiledeformationalfeaturesrepresenting
dextral transpression caused by the oblique collision(Jolivet and Miyashita, 1985; Arita et al., 1986; Arai
and Miyashita, 1994). Although the Poroshiri Ophi-
olite Belt is widely distributed in the northern half
of the Hidaka Mountains, it occurs sporadically as a
narrow zone along the HMT (serpentinite of only 80m wide in the surveyed area), and often is missing
inthe southern half(Fig. 4). The Poroshiri ophiolites
202 K. Arita et al./Tectonophysics 290 (1998) 197±210
Fig. 4. Generalized geological map of the central and southern
Hidaka Mountains. HWTDHidaka Western Thrust, HMTD
Hidaka Main Thrust, ROTDRedatoi±Okada Thrust, NOTD
Nitarachi±Oshorobetsu Thrust, HSZDHoroizumi Shear Zone,
TFDTokachi Fault System, PDMt. Poroshiri-dake. Thick line
is a seismic re¯ection line (Fig. 5).
overthrust the Idon'nappu Belt on the west along the
HidakaWesternThrust(HWT)(Fig.1candFig.4).
The Idon'nappu Belt is divided lithotectonically
into two units by the east-dipping Redatoi-Okada
Thrustassociatedwiththinserpentinitebodies(Uedaet al., 1995), namely the Naizawa Complex in the
west and the Horobetsu-gawa Complex in the east
(Fig. 4). Both complexes are composed of melangeand accretionary sediments. Slaty cleavages in these
complexes generally strike NW±SE, and steeply dip
to the east, although their bedding planes are mostly
west-facing (Ueda et al., 1995). The double colli-
sion has made the Idon'nappu Belt a frontal zoneof the Hidaka Collision Zone showing a dextral
strike-slip duplex structure with apparent westwardvergence especially in the Horobetsu-gawa Com-
plex (Kiyokawa, 1992; Ueda et al., 1995). The
Idon'nappu Belt is in tectonic contact with the Cre-
taceous Yezo Supergroup (the Sorachi±Yezo Belt)
and the Miocene formations on the west along the
east-dipping Nitarachi±Oshorobetsu Thrust (Fig. 4).The Miocene sandstones occur along these faults as
well as withinthe Idon'nappu Belt.
4. Vibroseisseismicre¯ectionpro®ling
Vibroseis seismic re¯ection pro®ling was per-
formed along a 20-km-long traverse on Route 236
across the southern Hidaka Mountains. A standardseismic data processing sequence was used, includ-
ing post-stack coherency ®ltering and ®nite-differ-
ence migrations and depth corrections using a 1-Dvelocitymodel (Table1).
Fig. 5 shows an unmigrated depth section in
which the supposed re¯ection phases are indicated
by arrows with numbers. Each phase is interpreted
as follows (numbers correspond to those inFig. 5).
(1)Theintermittentre¯ectorsareclearlytraceable
fromtheHMTonthesurface(aroundRP150)north-
eastward with an angle of 45ë to a depth of 7 kmbelowtheeasternmarginoftheHidakaMetamorphic
Table 1
Field parameters used in the vibroseis seismic re¯ectionexperiment
Source informationSource type 4 vibrators (Y-2400, MK-IV)Interval 50 mSweep frequency 8±45 HzSweep length 16 sNumber of sweeps 10 sweeps =VP
Sweep mode linear up sweepPhase control ground force locking
Receiver information
Natural frequency 8 HzInterval 25 mNumber of geophones 18 geophones =RP (3 series6 parallel)
Layout 1.4 m interval, linear array
Recording information
Number of channels 240Sample interval 4 msRecord length 24 s (after cross-correlation)Low cut frequency 4 Hz, 18 dB =oct
High cut frequency 90 Hz, 72 dB =oct
K. Arita et al./Tectonophysics 290 (1998) 197±210 203
Fig. 5. Unmigrated depth section across the southern Hidaka Mountains. The interpreted re¯ection phases are shown by arrows numbered
1through8which correspond to numbers in the text. Note a sharp eastward bend of the stacking line at RP 660.
204 K. Arita et al./Tectonophysics 290 (1998) 197±210
Fig. 6. An enlarged migrated depth section showing two listric-shaped re¯ectors corresponding to the Hidaka Main Thrust (top) and
Hidaka Western Thrust (bottom) and a duplex structure between them.
Belt.Thesere¯ectionphases areboundaries between
the complex area in the west and the rather trans-
parent area in the east. After migration, the HMT
re¯ectorsshow a listricgeometry (Fig. 6). The HMTis likely to cut across the foliation of metamorphic
rocks and boundaries of lithofaciesat depth.
(2) The HWT re¯ector is not observable at shal-
low levels. At deeper levels, however, a rather clear
re¯ection phase is recognized about 1 km belowthat of the HMT beneath the eastern ¯ank of the
Hidaka Mountains (between RP 550 and 800), and
interpreted to be the HWT. The layer between thesetwo strong events is considered to be the Poroshiri
Ophiolite Belt, and appears to have a duplex struc-
ture in an enlarged migrated depth section (Fig. 6).
In the unmigrated sectionthe west-dipping re¯ectors
(2
0) look to be traced intermittently from a depth of
6.5 km below RP 550 southwestward, although the
tracesbecomevagueaftermigration.
(3) A back thrust dipping west is found around
thenortheasternmostpartof thetraverse.
(4) There is no clear re¯ectivity in the Hidaka
metamorphic rocks. P-wave velocities in the sub-
surface estimated by processing 240-channel data
obtainedfrom the refractionmethodarealmostiden-tical between the different rock units of the Hidaka
Metamorphic Belt. This suggests poor contrast ofimpedance among the rocks of the Hidaka Metamor-
phic Belt.
(5) Beneath the Idon'nappu Belt some complex
re¯ection phases are recognized at depths of severalkm. These re¯ectors may be suggestive of a frontal
thrust-and-fold structure or tectonic stacking of the
Idon'nappu Belt in front of and below the Hidaka
MetamorphicBelt.
(6) Some subhorizontal re¯ection phases can be
observed at a depth of 14 km beneath the eastern
part of the Hidaka Metamorphic Belt and the Hidaka
Belt. A few similar sub-horizontal re¯ection phasesbecome visible beneath the Idon'nappu Belt after
migration.
(7) Possible short re¯ectors at a depth of 20 km
arelikelytobe lowercrustallamination.
(8) Steep re¯ection planes are observed at depths
of over 11 km beneath the western margin of the
seismic line. After migration, these steep planes
are moved outside the traverse. These planes arepresumed tobe of the Nitarachi±Oshorobetsu Thrust
andfaultsintheSorachi±YezoBelt(Fig.4).
5. Gravity survey
Gravity surveys in and around the Hidaka Moun-
tains (Geographical Survey Institute, 1955; Hagi-
K. Arita et al./Tectonophysics 290 (1998) 197±210 205
Fig. 7. Distribution map of the Bouguer anomaly in the southern part of the Hidaka Mountains with a contour interval of 5 mGal after
terrain correction. A±Bis a gravity and seismic traverse. Thick lines are faults. The broken line is the crestline of the Hidaka Mountains.
HWTDHidaka Western Thrust, HMTDHidaka Main Thrust, NOTDNitarachi±Oshorobetsu Thrust, HSZDHoroizumi Shear Zone, TF
DTokachi Fault System, RDMt. Rakko-dake (1472 m).
rawa, 1967; Miyamachietal., 1987) indicatethat the
Bouguer anomalies along the mountains are highly
positive, reaching up to C140 mGal in the north-
ern part (Maruyama et al., 1991). The high positive
anomaly is considered to be due to ma®c and ultra-ma®c rocks in the Poroshiri Ophiolite Belt and the
Hidaka Metamorphic Belt. This is also supported by
ah i g hV
p=Vsratio of over 1.8 in the western ¯ank
of the Hidaka Mountains (Moriya, 1983). The ®rst
precise gravity measurement was performed along a
seismic line (A±B in Fig. 7, more than 200 stations)
and in the surroundings (more than 300 stations) in
ordertoevaluatecrustalstructuremodelsbeneaththeHidaka Mountains.5.1. Bouguer anomaly
The optimum density for gravity reductions in
the study area was estimated to be 2.6615 g =cm
3
using the method proposed by Murata (1993). The
obtained Bouguer anomaly distribution map of the
study area is shown in Fig. 7. A remarkably high
gravity anomaly belt (about 30 km wide) is lo-cated along the crestline of the Hidaka Mountains.
This positiveanomaly reachesits maximum near the
westernperiphery of thecrestline.It isnoted thatthe
correlationof the Bouguer anomalywith topography
is signi®cantly positive in the Hidaka Mountains,which suggests that little crustal `root' exists be-
neath the mountains. In both the easternand western
206 K. Arita et al./Tectonophysics 290 (1998) 197±210
foothills of the mountains, an abrupt decrease in the
Bouguer anomaly is observed, but the patterns of
anomaliesareasymmetric,andtheBougueranomaly
gradients of both sides are different (Fig. 7). The
eastern abrupt gravity decrease corresponds to that
of the Tokachi Fault System which is an active fault(Research Group for Active Faults of Japan, 1991).
Another abrupt decrease is observable on the west-
ern ¯ank of the mountains including the area alongthe seismic line, although no large fault system is
situated there. Along the seismic line, the positive
gravity anomaly increases abruptly around the HMT
on the surface, and reaches its maximum in the
norther half of the seismic line, and then graduallydecreases toward the eastern margin of the Hidaka
Mountains. At the eastern foot of the mountains,
there is a gravity anomaly trough characterized by astrong negative Bouguer anomaly. It corresponds to
the boundary zone between the Hidaka and Tokoro
Fig. 8. A density structural model along the seismic line A±B in Fig. 7 used for the computation (top) and a comparison of observed and
calculated Bouguer anomalies (bottom) computed after Talwani et al. (1959).Belts,whichisburied by athickpileof LateTertiary
to Quaternarysediments.
5.2. Crustal structure
Basedonthepresentseismicre¯ectionresultsand
thegravityandgeologicalconstraints,asimpleblock
model is constructed for crustal structure along the
seismic line (Fig. 8, upper). The eastward-dippingre¯ector, which is interpreted as the HMT, continues
toadepthof 7km.Beneaththeplane,anarrowlayer
of the Poroshiri Ophiolite Belt is probably situated.
A nearly horizontal re¯ection plane is recognized at
a depth of 14 km. In the model construction, thegravitationaleffectsof thedowngoing slabofthe Pa-
ci®c Plate were ignored because they are considered
to be nearlyuniform on the traversewhich is parallelto the Kuril Trench (Fig. 1a). Fig. 8 (lower) shows
a preliminary crustal model along the seismic line
K. Arita et al./Tectonophysics 290 (1998) 197±210 207
based on Talwani's method (Talwani et al., 1959).
The model Bouguer anomaly ®ts well with the ob-
served gravity especially for the western and eastern
parts of the pro®le, whereas in the central part the
computedgravityshowsasystematicincreasewithin
severalmGal.
6. Discussion:crustalstructure of the Hidaka
CollisionZone
On the basis of seismic refraction data, Den and
Hotta (1973) suggested large-scale thrusting of the
crust of Eastern Hokkaido over Western Hokkaido
and vast sedimentation in the foredeep west of thethrust. They attributed the thrust boundary to a plate
boundary between the Okhotsk (the North Ameri-
can) and Eurasian Plates during the Mesozoic. Sucha tectonic scheme around the Hidaka Mountains
(Fig. 2a) has been supported also by recent geophys-
ical studies (e.g., Takanami, 1982; Miyamachi and
Moriya, 1984; Miyamachi etal.,1994; Moriya et al.,
1994).
Fig. 9. An interpreted crustal model of the southern part of the Hidaka Collision Zone. HWTDHidaka Western Thrust, HMTDHidaka
Main Thrust.Although the present integrated study of geologi-
caland geophysicalwork could not detectthe Moho,
it could image the collision tectonics beneath the
Hidaka Mountains such as the listric-shapedHMT, a
back thrust, a fold-and-thrust structure and a subhor-
izontalre¯ectorat a depth of 14 km (Fig. 9).
Asalreadystated,theHidakaMetamorphicBeltis
considered to be an upthrust magmatic arc (palaeo-
Hidaka arc) tilting steeply eastward similar to theHMT on the surface, and therefore the deep rocks of
the palaeo-Hidaka arc occur in its western part. The
granulite-faciesrocksoutcroppinginthewesternpart
of the study area are intensely mylonitized, but, in
general, the thermobarometric analyses of the gran-ulite-facies rocks indicate pressure and temperature
conditions corresponding to those at a depth of 23
km (Osanai et al., 1986). This thickness is almostthe same as the total thickness of the reconstructed
crustal sequences of the Hidaka arc including the
Nakanogawa Group (Komatsu et al., 1983). There-
fore, the depth of the HMT beneath the Hidaka Belt
should be expectedtobe more than23km, assuming
208 K. Arita et al./Tectonophysics 290 (1998) 197±210
that the observed parallel relationship on the surface
between the HMT plane and the lithologic bound-
ary and foliation planes of the Hidaka metamorphic
rocks is maintained at the deeper levels.The seismic
re¯ection pro®ling, however, proves the HMT to be
a listric fault, and by far shallower than that esti-mated before. Hence the HMT most probably cuts
across the lithological boundaries and foliations at
depth (Fig. 9). The Hidaka Main Thrust sheet (Hi-daka Metamorphic Belt) is found to be much thinner
than it has been generally expected. It is noted that
the HMT seems to continue eastward to the velocity
boundary between the 5.9±6.0 km =s layer and the
6.2±6.3 km=s layerdeduced from theseismicrefrac-
tion pro®ling beneath the Tokachi Plain in the north
(Fig. 2b: Ozel et al., 1996; Iwasaki et al., 1998), al-
though both areas are about 60 km from each other.The HMT is the most signi®cant tectonic feature
traceable along the whole Hidaka Mountains, and is
thought to have been a plate boundary until the Ter-
tiary.However,theHMTisjustalistricfaultsituated
at the middle of the upper crust, and consequentlythe true palaeo-plate boundary is expected to exist
beneaththe HMT atdepth.
In the northern half of the Hidaka Mountains, the
Poroshiri Ophiolite Belt reaches up to 5 km in width
(Miyashita, 1983), and large gabbroic bodies occur
in the western part of the Hidaka Metamorphic Belt
(Fig. 4). These ma®c rocks are attributed to the high
positive Bouguer anomaly of up to 140 mGal. Onthe other hand, in the southern half only small bod-
ies of the ophiolitic rocks occur intermittently along
the HMT like in the study area. Nevertheless, thepositive Bouguer anomaly is still high in the present
area (Fig. 7). A 1-km-thick Poroshiri Ophiolite Belt
detected at depth by seismic re¯ection may be re-
sponsible for the positive Bouguer anomaly. This
may be supported by the existence of a conductivelayer between resistive layers beneath the northern
Hidaka Mountains (Fig. 3), although their depth and
thickness are different from each other. The differ-ence in the Bouguer anomaly gradients between the
eastern and western sides of the Hidaka Mountains
and the rapidincreaseof the anomaly from the HMT
eastward (Fig. 7) also suggest that the Poroshiri
ophiolitesgentlydip eastwardand become thickertotheeast.Itisinterestingtonotethatseismicpro®ling
is suggestive of a duplex structure betweenthe HMTand HWTre¯ectors (Fig.6). Theserpentinizedophi-
oliticlayer is considered to play an important role as
a mechanical ¯ow plane for the westward thrusting
of the Hidaka metamorphic rocks along the HMT.
In the unmigrated section the HWT re¯ector is dis-
tinct beneath the northeastern part of the pro®le, andlooks to branch off downward around RP 550, being
traceableintermittentlysouthwestward (2
0in Fig. 5).
If the re¯ection phase is true, this may imply an ex-istence of a tectonic wedge of Indon'nappu melange
which splits the upper crust of the Hidaka arc into
two parts. Further detailed analyses are required to
ascertainthe southwest-dipping re¯ector.
In the shallow part of the Indon'nappu Belt, a
number of NNE-dipping re¯ectors appears, indica-
tive of a fold-and-thrust structure. Ueda et al. (1995)
reportedaduplexstructurewitha right-lateralstrike-slip sense of motion having resulted from dextral
transpression between the Hidaka arc crust and
an oceanic crust (Poroshiri ophiolites) in the Late
Oligocene to Early Miocene (Arita et al., 1986). The
duplex structure probably evolves into the fold-and-thrust structure atdepth.
A west-dipping re¯ector is seen clearly at the
northeastern end of the seismic traverse. On the sur-face some faults are observed, but the sense of their
movement is not clear because of the monotonous
lithology of the Nakanogawa Group. The re¯ector,
however,is presumedtobe abackthrust on thebasis
of a generaltectonic¯ame.
The very weak subhorizontal re¯ectors at 14
km depth at both sides of the seismic pro®le are
signi®cant (Fig. 9). Such a 14-km-deep subhorizon-tal boundary was also detected as a clear velocity
boundary between the 5.9 km =s layer and the 6.6±
6.7 km=s layer by seismic refraction pro®ling in
the northern Hidaka Mountains (Fig. 2b: Ozel et
al., 1996; Iwasaki et al., 1998). These layers aresupposed to be the upper and lower crusts, respec-
tively.In thenorthern HidakaMountains, thedistinct
boundary is traceablewidely from the Sorachi±YezoBelt through the Hidaka Mountains, and dips gen-
tly eastward to the Tokachi Plain (Fig. 2b). Beneath
the Tokachi Plain a middle velocity layer (6.2±6.3
km=s) exists between the layers of 5.9 km and 6.6
km=s in the northern Hidaka Mountains (Fig. 2b).
The boundary between the layers of 5.9 km =sa n d
6.2±6.3 km=s is located at about 8 km depth. It is
K. Arita et al./Tectonophysics 290 (1998) 197±210 209
worth to note that the boundary seems to be the
easterncontinuation of the HMT, although these two
arefarfrom eachother.
Further seismic pro®ling is required eastward for
a detailed imaging of the Hidaka crustal structure
and westward for understanding the tectonics in thecollisionalfront.
Acknowledgements
We would like to thank R. Sorkhabi for criti-
cal reading and improvements of an earlier draft
of the manuscript. This paper has been greatly im-
proved through the efforts of S. Klemperer and twoanomymous reviewers. N. Oshima, H. Yokota and
K. Kameda helped us with the seismicline measure-
ment. Thanks to Y. Murata for providing a computerprogram for calculatingthe Bouguer anomaly by the
ABI method. Financial support for this work was
provided by a grant from the Grant-in-Aidfor Scien-
ti®c Research of the Ministry of Education, Science,
Sports and Culture,Japan (06402018) toK.A.
References
Arai, T., Miyashita, S., 1994. Shear deformation and meta-
morphism of the Poroshiri Ophiolite in the Shunbetsu Riverregion, the Hidaka belt, Hokkaido, Japan (in Japanese, with
English abstr.) J. Geol. Soc. Jpn. 100, 162±176.
Arita, K., Toyoshima, T., Owada, M., Miyashita, S., Jolivet, L.,
1986. Tectonic movements of the Hidaka metamorphic belt,
Hokkaido, Japan (in Japanese, with English abstr.) Monogr.Assoc. Geol. Collab. Jpn. 31, 247±263.
Arita, K., Shingu, H., Itaya, T., 1993. K±Ar geochronological
constraints on tectonics and exhumation of the Hidaka meta-morphic belt, Hokkaido, Japan. J. Min. Pet. Econ. Geol. 88,
101±113.
Den, N., Hotta, H., 1973. Seismic refraction and re¯ection evi-
dence supporting plate tectonics in Hokkaido. Pap. Meteorol.
Geophys. 24 (1), 31±54.
Fujii, S., Moriya, T., 1983. Upper crustal structure in the Hidaka
district by refraction measurements using the quarry blasts.
Geophys. Bull. Hokkaido Univ. 42, 169±190.
Furumura, T., Moriya, T., 1990. Three-dimensional Q structure
in and around the Hidaka Mountains, Hokkaido, Japan. Zisin
(J. Seismol. Soc. Jpn.) 43, 121±132.
Geographical Survey Institute, 1955. Gravity survey in Japan (1),
Gravity survey in Hokkaido district. Bull. Geogr. Surv. Inst. 4,
23±99.
Hagirawa, Y., 1967. Analyses of gravity values in Japan. Bull.
Earthquake Res. Inst. Univ. Tokyo 45, 1091±1228.
Iwasaki, T., Ozel, O., Moriya, T., Sakai, S., Suzuki, S., Aoki, G.,Maeda, T., Iida, T., 1998. Lateral structural variation across
a collision zone in central Hokkaido, Japan, as revealed bywide-angle seismic experiments. Geophys. J. Int. 132, 435±457.
Jolivet, L., Miyashita, S., 1985. The Hidaka shear zone
(Hokkaido, Japan): genesis during a right-lateral strike-slip
movement. Tectonics 4, 289±302.
Kiminami, K., Kontani, Y., 1983. Mesozoic arc±trench systems
in Hokkaido, Japan. In: Hashimoto, M., Uyeda, S. (Eds.),Accretion Tectonics in the Circum-Paci®c Regions. Terra Sci.Publ. Co., Tokyo, pp. 107±122.
Kimura, G., 1986. Oblique subduction and collision: forearc
tectonics of the Kurile arc. Geology 14, 404±407.
Kimura, G., 1994. The latest Cretaceous±early Paleogene rapid
growth of accretionary complex and exhumation of high pres-sure series metamorphic rocks in northwestern Paci®c margin.J. Geophys. Res. 99 (B11), 22147±22164.
Kimura, G., Miyashita, S., Miyasaka, S., 1983. Collision tec-
tonics in Hokkaido and Sakhalin. In: Hashimoto, M., Uyeda,S. (Eds.), Accretion Tectonics in the Circum-Paci®c Regions.
Terra Sci. Publ. Co., Tokyo, pp. 123±134.
Kiyokawa, S., 1992. Geology of the Idon'nappu Belt, central
Hokkaido, Japan: evolution of a Cretaceous accretionary com-plex. Tectonics 11, 1180±1206.
Komatsu, M., Miyashita, S., Maeda, J., Osanai, Y., Toyoshima,
T., 1983. Disclosing of a deepest section of continental-typecrust upthrust as a ®nal event of collision of arcs in Hokkaido,
North Japan. In: Hashimoto, M., Uyeda, S. (Eds.), Accretion
Tectonics in the Circum-Paci®c Regions. Terra Sci. Publ. Co.,Tokyo, pp. 149±165.
Komatsu, M., Miyashita, S., Arita, K., 1986. Composition and
structure of the Hidaka metamorphic belt, Hokkaido Ð his-torical review and present status (in Japanese, with Englishabstr.) Monogr. Assoc. Geol. Collab. Jpn. 31, 189±203.
Komatsu, M., Osanai, Y., Toyoshima, T., Miyashita, S., 1989.
Evolution of the Hidaka metamorphic belt, northern Japan. In:
Daly, J.S., Cliff, R.A., Yardley, B.W.D. (Eds.), Evolution ofMetamorphic Belts. Geol. Soc., Spec. Publ. 43, 487±493.
Komatsu, M., Toyoshima, T., Osanai, Y., Arai, M., 1994. Pro-
grade and anatectic reactions in the deep arc crust exposedin the Hidaka metamorphic belt, Hokkaido, Japan. Lithos 33,31±49.
Maeda, J., 1990. Opening of the Kuril Basin deduced from the
magmatic history of Central Hokkaido, North Japan. Tectono-physics 174, 235±255.
Maeda, J., Kagami, H., 1996. Interaction of a spreading ridgeand
an accretionary prism: implications from MORB magmatismin the Hidaka magmatic zone, Hokkaido, Japan. Geology 24,31±34.
Maruyama, T., Nagasaki, Y., Kitsunezaki, C., 1991. Gravity sur-
vey in and around the Hidaka Mountains, Hokkaido, northernJapan. J. Min. Coll., Akita Univ. 7, 219±281.
Miyamachi, H., Moriya, T., 1984. Velocity structure beneath the
Hidaka Mountains in Hokkaido, Japan. J. Phys. Earth 32, 13±42.
Miyamachi, H., Moriya, T., Maekawa, T., 1987. Gravity survey
210 K. Arita et al./Tectonophysics 290 (1998) 197±210
in the western part of the Hidaka Mountains. Geophys. Bull.
Hokkaido Univ. 48, 45±52.
Miyamachi, H., Kasahara, M., Suzuki, S., Tanaka, K., Hasegawa,
A., 1994. Seismic velocity structure in the crust and uppermantle beneath northern Japan. J. Phys. Earth 42, 269±301.
Miyashita, S., 1983. Reconstruction of the ophiolitesuccession in
the Western Zone of the Hidaka Metamorphic Belt, Hokkaido(in Japanese, with English abstr.) J. Geol. Soc. Jpn. 89, 69±86.
Moriya, T., 1983. Regionality of Vp =Vs in the upper crust of
Hokkaido, Japan. Geophys. Bull. Hokkaido Univ. 42, 145±
154.
Moriya, T., Ozel, O., Nishimiya, T., Miyamachi, H., 1994. Ve-
locity structure model beneath Hidaka mountains, Hokkaido,Japan (in Japanese). Prog. Abstr. Seismol. Soc. Jpn. 2, 373.
Murata, Y., 1993. Estimation of optimum average sur®cial den-
sity from gravity data: an objective Bayesian approach. J.Geophys. Res. 98, 12097±12109.
Ogawa, Y., Nishida, Y., Makino, M., 1994. A collision bound-
ary imaged by magnetotellurics, Hidaka Mountains, centralHokkaido, Japan. J. Geophys. Res. 99, 22373±22388.
Okada, H., Suzuki, S., Moriya, T., Asano, S., 1973. Crustal
structure in the pro®le across southern part of Hokkaido,Japan, as derived fromexplosion seismic observations. J. Phys.Earth 21, 329±354.
Osanai, Y., Arita, K., Bamba, M., 1986. P±T conditions of
granulite- facies rocks from the Hidaka metamorphic belt,Hokkaido, Japan. J. Geol. Soc. Jpn. 92, 793±808.
Ozel, O., Moriya, T., Iwasaki, T., Iidaka, T., Sakai, S., Aoki, G.,
Suzuki, S., 1996. Crustal structure in the central Hokkaido,Japan, from a seismic refraction experiment. J. Fac. Sci.,
Hokkaido Univ., Ser. VII, Geophys. 10, 31±52.
Research Group for Active Faults of Japan, 1991. Active faults
in Japan, sheet maps and inventories, 2nd ed. The Universityof Tokyo Press, Tokyo, 437 pp.
Sakakibara, M., 1986. A newly discovered high-pressure terrane
in eastern Hokkaido, Japan. J. Metamorph. Geol. 4, 401±408.
Shimura, T., Komatsu, M., Iiyama, J.T., 1992. Genesis of the
lower crustal garnet±orthopyroxene tonalites (S-type) of theHidaka Metamorphic Belt, northern Japan. Trans. R. Soc.Edinburgh, Earth Sci. 83, 259±268.
Takanami, T., 1982. Three-dimensional seismic structure of the
crust and upper mantle beneath the orogenic belt in southernHokkaido, Japan. J. Phys. Earth 30, 87±104.
Talwani, M., Worzel, J.L., Landisman, M., 1959. Rapid gravity
computations for two-dimensional bodies with applications tothe Mendocino submarine fracture zone. J. Geophys. Res. 64,49±59.
Toyoshima, T., 1991. Tectonic evolution of the Hidaka meta-
morphic belt and its implication in late Cretaceous±middleTertiary tectonics of Hokkaido, Japan. Sci. Rep. Niigata Univ.,Ser. E (Geol. Miner.) 8, 1±107.
Ueda, H., Kawamura, M., Kato, M., 1995. Structure and meta-
morphism of the Mesozoic accretionary complex in northPaci®c rim Ð a study on the Idon'nappu belt, Hokkaido,northern Japan. In: Hanquan, W., Bai, T., Yiqun, L. (Eds.),IGC Project 294 Int. Symp., Very Low Grade Metamorphism:Mechanism and Geological Application. Beijing, pp. 132±144.
|
Arita (1998) - Crustal structure and tectonics of the Hidaka Collision Zone.txt
|
Geochemical Journal. Vol. 18, pp. 195 to 202,1984
Origin
steamsof gases and chemical equilibrium
from Matsukawa geothermal area,among them in
Northeast Japan
YUTAKA YOSHIDA
Geothermal Development Division, Japan Metals and Chemical Co., Ltd.
24 Ukai, Takizawa-mura, Iwate-gun, Iwate 020-01, Japan
(Received July 19, 1983: Accepted June 5, 1984)
Gas components contained in geothermal steams discharged from wells at the Matsukawa geothermal
areas were examined geochemically. The original deep seated gases of Northeast Japan are suggested to be
uniform with respect to He, Ar and N2 and are emitted through geothermal wells and/or fumaroles after
mixing in various proportions with atmospheric air dissolved in ground water.
Geothermal wells of the Matsukawa area are divided into two groups by the geological structure, of
the area which controls the variation in concentrations of tritium and major gas components occurs. The
influence of the geological barrier can be considered to be limited in a shallow horizon.
The correlation between gas components indicates that the reaction, 2NH3 = N2 + 3H2, is in equi
librium, but the reaction, CH4 + 2H20 = C02 + 4H2, is not in equilibrium under the condition of the
Matsukawa geothermal reservoir.
INTRODUCTION
The Matsukawa geothermal area is a vapor
dominated type geothermal system which is
unique among geothermal systems so far ex
plored in Japan. The first geothermal power
station in Japan was completed in this area by
Japan Metals and Chemicals Co., Ltd. in 1966.
At present, 22MW of electricity is generated
by geothermal steam discharged from eight
production wells. As the geothermal steam is
produced directly from zones deeper than
1,000m through a casing pipe, the contamina
tion of organic material, shallow meteoric water
and atmospheric air does not occur significantly
during the passage of steam through the well.
The components of high-temperature vol
canic gases have been studied from the point of
view of chemical equilibrium (ELLIS, 1957;
MATSUO, 1960; HEALD et al., 1963; STOIBER et
al., 1974; GERLACH, 1979). In recent years,
geothermal development has become active all
over the world, and gochemical studies of
geothermal systems have made a great progress
(D'AMORE et al., 1980; GIGGENBACH, 1980).Moreover, rare gases in volcanic gases are studied
recently in relation to the origin of gases
(MATSUO et al., 1978; NAGAO et al., 1980;
TORGERSEN et al., 1982; Kn'osu, 1983a).
In the present study, an attempt is made to
examine the origin of gases on the basis of the
analytical results of steam discharged from
geothermal wells in the Matsukawa area. In this
paper, steam means the mixture of water vapor
and other gas components such as C02, H2S,
H2, He and so forth.
GEOLOGICAL SETTING
The Matsukawa geothermal power station is
located in the Hachimantai volcanic region
which is one of the most active geothermal areas
in Northeast Japan (Fig. 1). Geological investi
gations in this area have been carried out by
NAKAMURA et al. (1961) and SuMI (1966, 1968),
and abundant geological data have been accumu
lated by Japan Metals and Chemicals Co., Ltd.
The basement of this region consists of the
Miocene Kunimitoge, Takinoue-onsen and Ya
matsuda formations, which are composed of
195
196 Y. YOSHIDA
shale, sandstone, tuff and conglomerate, and is
overlain by the Pliocene andesitic-dacitic Tama
gawa welded tuff. The Tamagawa welded tuff is
covered by the Pleistocene Matsukawa andesite
which is regarded as the' cap rock of the geother
mal system in this area. Several Pleistocene
formations composed of andesitic volcanic rocks
overlie the Matsukawa andesite. Most of the
geothermal steams are derived from the lower
part of the Tamagawa welded tuff formation
and the Yamatsuda formation.
The depth of eight production wells and one
exploration well ranges from 1,000 to 1,600m,
and at present, the steam is perfectly dry and
superheated by 20 to 70°C as compared with
the liquid vapor equilibrium temperature at the
well-head pressure.
The altered rock zone extends along the
Matsukawa river in the direction from ENE to
WSW, comprising an area 7km long and 0.5
1.0km wide. The altered rock zone is calssified
into some subzones as shown in Fig. 1, on the
basis of the mode of occurrence of mineralassemblage. These zones are not only distribut
ed horizontally but also vertically as shown in
Fig. 2, and kaolinite, anhydrite, pyrite and
other alteration minerals are found in boring
core and cutting samples.
SAMPLING AND ANALYSES
Sample collection and chemical analyses of
the geothermal steam from wells of the Matsu
kawa geothermal power station were carried
out in September and December 1982, and
steam condensates were analyzed in April 1982.
Samples of steam condensates for the measure
ment of tritium concentration were collected in
1975 and 1980. Localities of wells are shown in
Fig. 1.
The methods of sampling and analyses
were similar to that of OZAWA (1967), but
partly modified for the sake of convenience.
Samples of steam condensates were collected
through a glass coil condenser. Non-absorbable
gases in alkaline solution were analyzed by the
/
///// / 4atsukawa R. 1_4
/ e~@To Q ® y /'8/
// / /
/ / /
// //AA
/M
/y•// /11\/,M1
16 l/
%4kagaa is R.\
(\ /M9
f//
V/
/ /
7 /
/ / Legendloom
sG~
//Zone of
weak alteration
Zone of
montmorillonite
Zone of
kaolinite
Zone of
alunite
E
SEA OF
JAPAN147E
O~INUAA
KAKKONDA1N
PACIFIC
OCEAN
-4ON
39N
Fig. 1.
zones.Map of the Matsukawa geothermal area showing the localities of wells and distribution of altered rock
•: well-head locality, -*: well-bottom locality .
Origin of gases and chemical equilibrium among them 197
gas chromatographic method (SuGisAKI et al.,
1980; KAWABE et al., 1981). The Hitachi model
164 gas chromatograph combined with a pre
amplifier, Ohkura model AM 1001 B micro-volt
meter was used for He, Ar and N2 measure
ments. Tank oxygen was used as carrier gas at
the flow rate of 5 ml/min. The separation
column consisted of teflon tubing (3 mm inner
diameter) 5 m in length packed by 60/80 mesh
Molecular Sieve 5A. The oven temperature was
set at 40'C. Since this gas chromatograph was
modified to remove hydrogen gas by heated
stainless steel column packed with CuO grains,
the Hitachi model 163 gas chromatograph was
also used for H2, N2 and CH4 measurements.
Tank argon was used as carrier gas at the flow
rate of 30ml/min. The separation column
with 3mm (I.D.) stainless steel tubing 2m in
m 800
400
Sea
bevel
-400LN
NQ
HN
m
()It
cc
2length was packed with 60/80 mesh Molecular
Sieve 5A. The oven temperature was set at
40'C. The analytical error for He was about
10% and for the others less than 5%.
The NH3 concentration in steam condensate
was analyzed colorimetrically. The measure
ment of tritium concentrations in steam con
densates was performed at Gakushuin University
by gas counting method described by YONEDA et
al. (1967), after electrolytic enrichment of
tritium.
®I a3 ®5
®2 ®4
Fig. 2. Schematic cross section of alteration zones in
the Matsukawa geothermal field. 1: Zone of montmoril
lonite and iron-rich saponite, 2: Zone of chlorite, 3:
Zone of kaolinite, 4: Zone of alunite, 5: Zone of
pyrophyrite (after KIMBARA, 1983). RESULTS AND DISCUSSION
He, Ar and N2 concentrations in geothermal
steam Results of chemical analyses of
geothermal steam are listed in Table 1. In this
table, all the gas concentrations are expressed
by volume concentration in the steam. The He/
Ar and N2/Ar ratios are further normalized to
the corresponding atmospheric ratios (Fig. 3).
As seen in Fig. 2, sample points are distributed
mostly along the curve which connects point A
with B. Point A shows the dissolved air in water
which is in equilibrium with the atmospheric
air at 10° C and B indicates the gas (M-3, Decem
ber 13, 1982) with the highest ratios of He/Ar
and N2/Ar. Point B can be considered to re
present the gas derived from a deeper horizon of
this area. The curve is a calculated mixing line
of gases with the composition of points A and
B.
In such a small geothermal area as Matsuka
wa, it can be postulated that the deep seated
gas with a homogeneous composition exists
in a deeper horizon of the area. In other words,
variation of He/Ar and H2/Ar ratios of geother
mal steam is not due to the variation in He/Ar
and N2/Ar ratios of the original deep seated
gases but to the change in mixing ratio of deep
seated gases and dissolved air. The mixing may
occur in the reservoir formation processses.
Corresponding ratios of fumarolic gases from
volcanoes of Northeast Japan (Kiyosu, 1983a)
are also distributed along the extended curve
in Fig. 2. The positive correlation between He/
Ar and N2/Ar ratios common for the original
198 Y. YOSHIDA
Table 1. Composition of geothermal steam from geothermal wells at Matsukawa
Gas concentration in steam (by volume)
Well Depth
mDate Total gas H2 S C02 H2 N2 CH4 Ax He NH3
%ppm ppm ppm ppm ppm ppb ppb ppm
M-1
M-2
M-3
M-5
M-6
M-7
M-8
M-9
T-241006
1080
1170
1190
1203
1280
1406
1599
10509/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
9/28,1982
12/13,1982
12/13,19820.87
0.83
0.30
0.33
0.71
0.74
0.35
0.33
0.32
0.31
0.26
0.24
0.42
0.36
1.14
1.08
0.43487
540
441
492
5 25
555
410
360
368
474
403
384
760
673
616
572
4438010
7550
2480
2730
6350
6660
2960
2810
2630
2510
2110
1910
3360
2830
10600
10100
382022.4
33.0
42.6
44.9
88.0
94.0
32.0
35.0
33.0
55.5
34.6
45.1
30.5
37.1
128
108
12.8115
111
25.9
20.9
88.0
58.1
61.3
65.0
51.2
44.0
36.4
43.2
35.4
38.5
40.2
35.6
19.661.9
64.2
12.2
9.77
50.7
32.5
32.3
34.7
20.7
17.6
14.3
16.7
17.7
17.6
36.7
29.2
6.021220
1120
324
268
930
528
693
726
643
543
523
562
496
479
456
455
28824.0
21.7
3.90
5.02
16.8
14.2
9.21
15.4
8.77
9.05
5.15
7.49
7.48
5.90
6.73
8.42
2.8410.0
15.3
54.4
4.7
5.9
6.2
8.3
51.0
Total gas: Gases other than water vapor.
NH3: Samples were collected on April 20, 1982.
deep seated gas of the Matsukawa area and
fumarolic gases from volcanoes of Northeast
Japan may be related to the fact that the origi
nal deep seated gas of Northeast Japan is uni
form with respect to nitrogen and noble gases
and is emitted through geothermal wells and/
or fumaroles after mixing with dissolved air in
various proportions.
As seen in Fig. 3, He/Ar and N2/Ar ratios of
the original deep seated gas of Northeast Japan
are more than hundred times and three times as
large as those of atmospheric air, respectively.
According to SUGISAKI et al. (1978), four ex
pected sources of nitrogen are as follows: 1.
penetrating atmospheric air, 2. bacterial decom
position of organic matter contained in sedi
ments, 3. pyrolysis of organic matter, 4. release
of inorganic nitrogen from igneous and/or
metamorphic rocks. The most probable origins
of nitrogen which raises the N2/Ar ratio are (3)
and (4) among four possibilities described
above.
On the other hand, MATSUO et al. (1978)
suggested that one of the reasons for high
N2/Ar ratios in volcanic gases from island arcvolcanoes is the contribution of factor (3)
due to sedimentary materials transferred into
the lower crust or upper mantle through subduc
tion.
It can be suggested that N2/Ar ratio of the
original deep seated gas to which there is no
contribution of dissolved air has a fixed value
controlled by sedimentary materials. It can
also be concluded that He, Ar and N2 gases
contained in both of geothermal steams and
fumarolic gases of Northeast Japan are derived
from a common source.
Geothermal reservoir Eight production wells
and one exploration well produce steam at the
Matsukawa geothermal power station. Wells
except M-2 and M-5 were drilled by directional
drilling as indicated in Fig. 1. Table 1 shows
that CO2 concentrations of steam from M-1, 3
and 9 range from 6,350 to 10,600ppm and
those of the others from 1,910 to 3,820ppm.
The H2S concentration of steam has a range
from 360 to 760ppm, and the variation in the
concentration of H2S is smaller than that of
CO2.
Origin of gases and chemical equilibrium among them 199
The relationship among C02, H2S and R
gas (residual gases after the gas is washed with
5 N KOH solution) is shown in Fig. 4. As shown
in Fig. 3, eight production wells are divided
into two groups, i.e., wells M-1, 3 and 9 and
wells M-2, 5, 6, 7, and 8. The exploration well
T-24 does not belong to both of the two groups.
Two production well groups can be distin
guished from each other also by their localities.
Wells M-l, 3 and 9 are located in the zone of
weak alteration, and other wells are located in
the zones with the occurrence of montmoril
lonite, kaolinite and alunite as shown in Fig. 1.
The zonal distribution on the exposed surface
is observed in the vertical section; the alunite,
zone occurs in center, successively surrounded0s
/_Q/_6
00 0°
Ao
00
q)
0°°0
°
>o
1000
100
10
a1.0
0.10
00Q
a
Aeo
0
+B
o ~ 0
00
0 00
0
0.5 1.0 1.5 2.0
( NZ/Ar)sample/(N2/Ar)air
Fig. 3. Relationship between HelAr and N2/Ar rarios of
geothermal gases. •: M-1, 3 and 9, 0: M-2, 5, 6, 7 and
8, o: Fumarolic gas of Northeast Japan (after Kiyosu,
1983a). Point A shows dissolved air in water which is in
equilibrium with atmospheric air at 10°C, and point B
the gas from M-3 with the highest ratios of He/Ar and
N2/Ar. The curve is calculated mixing line of gases of
points A and B. Three fumarolic gases with (N2/
Ar)samplel(N2/Ar)air. > 2 are excluded from this Figure.100 90 80
C02 <
Fig. 4. Gas composition of geothermal steam of Matsu
kawa shown by triangle diagram for C02, H2S and R -gas
(residual gases). •: M-1, 3 and 9, 0: M-2, 5, 6, 7 and 8,
o: T-24 .
by the kaolinite zone and montmorillonite zone
(Fig. 2). Calcite exists in the zones of weak
alteration (SuMI, 1968), and the coincidence of
the localities of wells which discharge CO, rich
steam and the zone of weak alteration suggests
that calcite is one of the sources of CO2 in
steam from wells M-1, 3 and 9.
The tritium concentrations of condensates
of wells M-3 and M-9 are 1.09 1.19 T.U., and
those of wells M-5 and M-8 are 0.28 0.40T.U.
(Table 2).
The variety of the concentrations of major
gas components and tritium as well as He/Ar and
N2/Ar ratios suggests a possibility that the
reservoir or steam channel for wells M-1, 3 and 9
is separated by some barrier (faults, which are
assumed from the upheaval structure of the
Yamatsuda formation) from the other reservoir
Table2. Tritium concentration in condensate of
steam from Matsukawa
Well Date T (T.U.)
M-3
M-5
M-8
M-98/13, 1980
12/ 8.1975
8/13,1980
8/13,19801.09
0.28
0.40
1.19
T. U. = (T/1H) X 1018.
200 Y. YOSHIDA
Table 3. Correlation coefficients(r) between gas components in geothermal steam atMatsukawa
H2S
0.443C02
0.999
0.416 H2
0.682
0.336
0.681N2
0.402
-0.051
0.400
-0.070CH4
0.653
0.107
0.650
0.195
0.944Ar
0.335
-0.083
0.333
-0.147
0.985
0.907He
0.448
-0.009
0.445
-0.005
0.967
0.928
0.935NH3
0.737
0.372
0.735
0.942
0.014
0.275
-0.097
0.094Total gas
H2S
CO2
H2
N2
CH4
Ar
He
or steam channel for wells M-2, 5, 6, 7, 8 and T In this connection, an independent behavior of
24. Since He/Ar and N2/Ar ratios of the original H2S seems to be due to the buffer reaction, e.g.,
deep seated gas of the Matsukawa area are uni sulfide + H20= oxide + H2S . KlYosu (1983b)
form, the influence which causes the variation showed also the possibility that the hydrogen
of concentrations of tritium and chemical com isotopic exchange equilibrium is also established
ponents can be considered to be limited in a between H20 and H2 through the following
shallower horizon than that in which the origi reaction,
nal geothermal gas of the Matsukawa area exists.
2H20= 2H2 +02 (1)
Correlation between individual gas compo
nen is Gas composition of geothermal steam It can be said that weak correlations indicated
gives us two interesting problems. One is the for the pairs of gas components including H2 and
origin of each component and the other is the H2S come from high reactivity of H2 and H2S
chemical reactions taking place in the geother through chemical reactions. Some of possible
mal systems. The origin of He, Ar and N2 has reactions are described above. Beside reaction
been discussed in the previous section. In order (1), following two reactions including H2 can be
to elucidate the origin of gas components such considered among major components to investi
as C02, CH4, H2S, H2 etc., correlation coef gate the reactivity of H2
) ficients(r) between the gas components are cal
culated first, and are given in Table 3. The 2NH3 = N2 + 3H2 (2)
Irl values higher than 0.693 indicate significant
correlation at the confidence level of 99.9%. CH4 + 2H20 = CO2 + 4H2 (3)
Very strong positive correlation (r=0.999)
between C02 and total gas* implies that contri When reaction (2) is in equilibrium, its
bution of CO2 concentration and its variation to equilibrium expression in terms of fugacities fi
those of total gas is great. Strong positive cor becomes
relations are also indicated for the pairs of 3
N2-CH4, N2-Ar, N2-He, CH4-Ar, CH4-He and Ar KN = f (4)
He. N H 3
On the other hand, H2S and H2 have no cor where K
N denotes the equilibrium constant. relation with other gas components, while H2 Since fugacity coefficients can be assumed
shows an appreciabe correlation with NH3. to be close to unity under the condition of KiYosu (1980) showed that the sulfur isotopic common geothermal reservoir, reaction (4) can exchange equilibrium is established at the well be converted using partial pressure, pi, to be bottom temperature between H2S, pyrite and 3
anhydrite in the Matsukawa pN2 PH2 (5) geothermal area . KN = 2
PNH3
* Total gas is defined in this study as all gases except water vapor (refer to Table 1).
Origin of gases and chemical equilibrium among them 201
As the partial pressure can not be measured
directly at present, we have to use the measured
concentrations for discussion. The fact that
neighboring Kakkonda and Ohnuma geother
mal fields (Fig. 1) are water dominated-type
systems suggests that water exists originally in
liquid phase even in the Matsukawa reservoir.
Since the steam is dry, we can assume that a
part of liquid phase in the reservoir evaporates
completely, so that all the chemical components
found at the well-head are originally dissolved in
the liquid. Then, the concentration C; of com
ponent i measured at the well-head can be used
to check the establishment of chemical equi
librium in the liquid reservoir..
Equation (5) can be converted using con
centration of species i in the discharging steam
to equation (6),
CNH3 = A •CN2 • CH2 (6)
where A is a constant at a fixed temperature.
If reaction (2) is in equilibrium, CNH3 should be
proportional to CN2 • CH2. In order to examine
reaction (2), correlation coefficient was cal
culated. A strong positive correlation (r=0.986)
between CNH3 and CN2 • 'CH2 suggests that reaction (2) is in equilibrium.
As the measured well-bottom temperature
at Matsukawa shows a rather narrow range from
223 to 2600C, reservoir temperature is assumed
probably to be constant near 300'C. Rates of
chemical reaction and carbon isotopic exchange
in reaction (3) are considered to be very slow at
temperatures around 300°C (HULSTON, 1977;
SACKETT et al., 1979; NUTI et al., 1980;
GIGGENBACH, 1982). If reaction (3) is in equi
librium at a constant temperature, following
two equations should hold,
2 1 aPCH4 • PH2O = K~ • pco2 ' P H2 (7)
and CCH4 = B • CCO2 • CH2 (8)
where K, and B are an equilibrium constant and
a constant, respectively. In the case of constant
temperature, PH2o (saturated water vapor pressure) can be regarded as constant. Again, the
correlation between CH4 and CC02 • CH2 is examined. The correlation coefficient is ob
tained to be 0.149, which suggests that reaction
(3) is not in equilibrium. It suggests that carbon
isotope exchange reaction between CO2 and
CH4 is also not in equilibrium. This disequi
librium state can be regarded as the result of
contamination of CO2 and/or CH4 from outside
the geothermal system. The contributions of
organic matter to C02 and CH4 production can
not be neglected in some cases (D'AMORE et al.,
1977; WELHAN et al., 1979). It is necessary to
reconsider the applicability of carbon isotopic
geothermometer method.
On the basis of the facts mentioned above,
NH3 is considered to be controlled by chemical
reaction with H2 and N2. It is considered also
that H2 is controlled by chemical reactions
between Fe(II) in minerals, H2O and H2S in the
Matsukawa geothermal reservoir.
Acknowledgement-The author wishes to express his
thanks to Professor N. NAKAI of Nagoya University and
Professor S. MATSUO of Tokyo Institute of Technology
for their critical reading the manuscript. He also ac
nowledges Dr. H. NAKAMURA of Japan Metals and
Chemicals Co., Ltd. for permission and encouragment
which led to the preparation of this paper. Thanks are
also due to the staff member of the geochemical section
of J.M.C. for their help during the sampling and
analyses.
REFERENCES
D'AMORE, F., CELATI, R., FERRARA, G. C. and PANICHI,
C. (1977) Secondary changes in the chemical and
isotopic composition of the. geothermal fluids in
Larderello field. Geochermics 5, 153-163.
D'AMORE, F. and PANICHI, C. (1980) Evaluation of
deep temperatures of hydrothermal systems by a
new gas geothermometer. Geochim. Cosmochim.
Acta 44, 549-556.
ELLIS, A. J. (1957) Chemical equilibrium in mag
matic gases. Am. J. Sci. 255, 416-43 1.
GERLACH, T. M. (1979) Thermodynamic investiga
tion and restoration of the 1970 volcanic gas analyses
from Mt. Etna, Sicily. J. Volcanol. Geother. Res. 6,
165-178.
GIGGENBACH, W. F. (1980) Geothermal gas equi
libria. Geochim. Cosmochim. Acta 44, 2021-2032.
202 Y. YOSHIDA
GIGGENBACH, W. F. (1982) Carbon-13 exchange
between CO2 and CH4 under goethermal conditions.
Geochim. Cosmochim. Acta 46, 159-165.
HEALD, E. F., NAUGHTON, J. J. and BARNES, I. L.,
JR. (1963) The chemistry of volcanic gases: 2.
Use of equilibrium calculations in the interpretation
of volcanic gas samples. J. Geophys. Res. 68, 545
557.
HULSTON, J. R. (1977) Isotope work applied to geo
thermal systems at the institute of nuclear sciences,
New Zealand. Geothermics 5, 89-96.
KAWABE, I., MAKI, I. and SUGISAKI, R. (1981) Geo
chemical study on subsurface gases in the fault zones
of Shikoku Island, Japan-(1): Bubble gas survey
arround the Median Tecotnic Line. Geochem. J.
15, 183-191.
KIMBARA, K. (1983) Hydrothermal rock alteration
and geothermal systems in the eastern Hachimantai
geothermal area, Iwate Prefecture, northern Japan.
Ganseki Kobutsu Kosho Gakkaishi 78, 479-490 (in
Japanese with English abstr.).
KIYOSU, Y. (1980) The abundance and sulfur iso
tope composition of sulfur compounds in the Matsu
kawa geothermal area. Ganseki Kobutsu Kosho
Gakkaishi 75, 353-358 (in Japanese with English
abstr.).
KIYOSU, Y. (1983a) N2 /Ar and He/Ar ratios of
fumarolic gases from volcanoes of Northeast Japan.
1983 Abstract of Annual Meeting of Geochem.
Soc. Japan, 197 (in Japanese).
KIYOSU, Y. (1983b) Hydrogen isotopic composition
of hydrogen and methane from some volcanic areas in
northeastern Japan Earth Planet. Sci. Lett. 62, 41
52.
MATSUO, S. (1960) On the origin of volcanic gases.
J. Earth Sci., Nagoya Univ. 8, 222-245.
MATSUO, S., SUZUKI, M. and MIZUTANI, Y. (1978)
Nitrogen to argon ratio in volcanic gases In Terrestrial
rare gases (ed. E. C. ALEXANDER, JR. and M. OZIMA),
Cent. Acad. Publ. Japan, Tokyo, 17-25.
NAGAO, K., TAKAOKA, N., MATSUO, S., MIZUTANI, Y.
and MATSUBAYASHI, 0. (1980) Change in raregas composition of the fumarolic gases from the
Showa-shinzan volcano. Geochem. J. 14, 139-143.
NAKAMURA, H. and SUMI, K. (1961) Geothermal
investigations of Matsukawa hotspring area, Iwate
Prefecture. Chishitsu Chosasho Geppo 12, 73-84
(in Japanese).
NUTI, S., NOTO, P. and FERRARA, G. C. (1980) The
system H20-CO2-CH4-H2 at Travale, Italy: Tentative
interpretation. Geothermics 9, 287-295.
OZAWA, T. (1967) Chemical analyses of gases at geo
thermal field. Chinetsu 9, 12-16 (in Japanese).
SACKETT, W. M. and CHUNG, H. M. (1979) Experi
mental confirmation of the lack of carbon isotope
exchange between methane and carbon oxides at high
temperatures. Geochim. Cosmochim. Acta 43, 273
276.
STOIBER, R. E. and ROSE, W. I., JR. (1974) Fumarole
incrustations at active Central American volcanoes.
Geochim. Cosmochim. Acta 38,495-516.
SUGISAKI, R. and SHICHI, R. (1978) Precursory
changes in He/Ar and N2/Ar ratios of fault gases
prior to earthquakes. J. Seismol. Soc. Japan 31,
195-206 (in Japanese).
SUGISAKI, R., ANNO, H., ADACHI, M. and UI, H. (1980)
Geochemical features of gases and rocks along active
faults. Geochem. J. 14, 101-112.
SUMI, K. (1966) Hydrothermal rock alteration of
Matsukawa geothermal area, Iwate Prefecture. Kozan
Chishitsu 16, 261-271 (in Japanese).
SUMI, K. (1968) Hydrothermal rock alteration of
the Matsukawa geothermal area, northeast Japan.
Report No. 225, Geol. Survey Japan.
TORGERSEN, T. and JENKINS, W. J. (1982) Helium
isotopes in geothermal systems: Iceland, the Geysers,
Raft River and Steamboat springs. Geochim. Cosmo
chim. Acta 46, 739-748.
WELHAN, J. A., POREDA, R., LUPTON, J. E. and CRAIG,
H. (1979) Gas chemistry and helium isotopes at
Cerro Prieto. Geothermics 8, 241-244.
YONEDA, K. and KIGOSHI, K. (1967) Tritium mea
surement by gas counting. Bunseki Kagaku 16, 561
565 (in Japanese).
|
Yoshida origin of gases and chemical equilibrium from steams of matsukama geothermal area.txt
|
Journal of Mineralogical and Petrological Sciences, Volume 113, page 219-231, 2018 Compositional variation of olivine related to high-temperature serpentinization of peridotites: Evidence from the Oeyama ophiolite Toshio NoZAKA Department of Earth Sciences, Okayama University, Okayama 700-8530, Japan Compositional variation of olivine in serpentinized peridotites provides a significant constraint on modeling the redox conditions of serpentinization and the tectonothermal history of ophiolites. Here I report the variations of Fe, Mg, Mn, and Ni contents of olivine from the Oeyama ophiolite, SW Japan and show textural and chemical evidence for compositional modification of olivine related to high-temperature (T) serpentinization. The Fe- enrichment of olivine adjacent to antigorite without significant magnetite formation indicates a reducing con- dition for high-T serpentinization. Systematic variations of forsterite (Fo) component with distance from anti- gorite suggest Mg-Fe volume diffusion took place in olivine porphyroclasts under the conditions of high-T diagram, which could be a useful indicator of high-T serpentinization. Retrograde antigorite is different from prograde antigorite in having a shape of elongated blade, lacking a significant amount of magnetite inclusion, and being more ferrous than lizardite. The existence of retrograde antigorite provides another piece of evidence for high-T serpentinization even if olivine has been decomposed by intense low-T serpentinization. Approx- imate estimation of time required for the observed Mg-Fe diffusion profiles of olivine porphyroclasts reveals within the serpentinized mantle wedge following rapid exhumation immediately after the amphibolite-facies metasomatism. Volume diffusion INTRODUCTION Magnetite is commonly associated with lizardite and/or brucite in serpentinites. Recent studies have shown Because of the possible relation to the origin of life and the uneven distribution of magnetite in serpentinites, sug- the development of biosphere, molecular hydrogen pro- gesting a local variation in oxygen fugacity, silica activity duced by serpentinization of peridotites have attracted sig- and/or water-rock ratio (e.g., Bach et al., 2006; Beard nificant interests (e.g., Kelley et al., 2001; Fruh-Green et et al., 2009; Katayama et al., 2010; Frost et al., 2013; al., 2003; Sleep et al., 2004; Takai et al., 2004; Kelley et Miyoshi et al., 2014; Schwarzenbach et al., 2016). In con- al., 2005; Schulte et al., 2006; Russell et al., 2010). The trast, Evans (2010) has proposed that magnetite is absent hydrogen production is mainly caused by the oxidation of at relatively high-temperature (T) antigorite-forming ser- Fe in olivine and associated with the formation of magnet- pentinization, because Fe is distributed in olivine with a ite (e.g., Frost, 1985; McCollom and Bach, 2009; Klein considerable Mg-Fe interdiffusion rate, instead of forming et al., 2013, 2015). In this context, it is important to under- magnetite. Such a behavior of Fe during high-T serpentin- stand the formation conditions of magnetite and the effect ization could be a constraint on hydrogen production at of its presence or absence on serpentinization processes. a mantle wedge of supra-subduction zone, at a deep level of oceanic lithosphere, or at the ancient Earth's surface. doi:10.2465/jmps.180420 However, observations of the Fe-enrichment of olivine T. Nozaka, nozaka@ cc.okayama-u.ac.jp Corresponding author 220 T. Nozaka SWJapan RM TM AS OS 100 km Ultramafic complex SM Suo Belt (230-160 Ma) SK Sangun high-P/T OE HP Renge Belt (330-280 Ma) metamorphic belt Figure 1. Distribution of ultramafic complexes and the Sangun high-P/T metamorphic belt (Renge and Suo Belts) in SW Japan (Ishiwatari, 1989, 1990; Nishimura, 1998; Takeuchi, 2002; Isozaki et al., 2010). Abbreviations for ultramafic complexes: AS, Ashidachi; HP, Happo; OE, Oeyama; OS, Ohsayama; RM, Ryumon; SK, Sekinomiya; TM, Tari-Misaka; WS, Wakasa. Color version is available online from https://doi.org/10.2465/jmps.180420. except for some examples (Murata et al., 2009a, 2009b; Nozaka, 2003, 2005; Khedr and Arai, 2012). The tempo- ral sequence of these geologic events was summarized by intense low-T serpentinization or overprinting metamor- Nozaka (2014a). phism. In addition to reporting new occurrence of the retro- ANALYTICAL PROCEDURES grade ferroan olivine in the Oeyama ophiolite, SW Japan, I show here evidence for the volume diffusion of ele- Polished thin sections of serpentinized peridotites from ments in olivine grains under the conditions of high-T the Oeyama ophiolite were prepared for microscopic ob- serpentinization. This compositional modification of oliv- servations and chemical analyses of minerals. Several ine could be a useful indicator of tectonothermal history samples from the Ryumon ultramafic body in the Sanba- of serpentinized forearc peridotites. gawa metamorphic belt exposed in the Kii Peninsula were analyzed for comparison. Quantitative microprobe GEOLOGICAL SETTING analyses were carried out using an electron probe micro- analyzer, JEOL JXA-8230 at Okayama University with The Oeyama ophiolite is a collective name of ultramafic an accelerating voltage of 15 kV and a probe current of complexes exposed in the Renge high-P/T metamorphic s s o s o belt of SW Japan (Fig. 1; Ishiwatari, 1989, 1990; Nishi- Fe, and Mg), 60 s for minor elements (Mn, Ni, and Ca) of mura, 1998; Isozaki et al., 2010). In the present article, olivine, and 10 s for the all element of serpentine; back- the Happo complex is included in the Oeyama ophiolite ground counting time was 5 s for all the cases. Standards despite its long distance from the other complexes, be- used were natural or synthetic oxides and silicates. The cause they are similar in geological and geochemical applied matrix corrections followed the procedures of characteristics (Takeuchi, 2002; Nozaka, 2005; Khedr Bence and Albee (1968), using alpha factors of Naka- and Arai, 2010; Nozaka, 2014a). mura and Kushiro (1970). Identifcation of serpentine The peridotites of the Oeyama ophiolite are consid- minerals of representative samples was confirmed with ered to be residual mantle peridotites and cumulates from Raman shift spectra, using a micro-Raman spectrometer, basaltic melt originated at a Paleozoic supra-subduction JASCO NRS-3100 at Okayama University, with 488 nm zone or forearc mantle wedge (Ishiwatari, 1989; Arai and laser excitation, a 100 × microscope objective lens, and a Yurimoto, 1995; Ishiwatari and Tsujimori, 2003; Khedr diffraction grating with 1800 grooves/mm. and Arai, 2010; Nozaka, 2014a). They were subjected to two stages of mylonitization, which could be associated RESULTS with metasomatic hydration at 700-800 °C and later high-T serpentinization at 400-600 °C (Nozaka, 2005; As I have previously reported the petrographic observa- Nozaka and Ito, 2011; Nozaka, 2014a). Many of the ser- tions of the serpentinized peridotites from the Oeyama pentinized peridotite bodies of the Oeyama ophiolite ophiolite (Nozaka, 2005; Nozaka and Ito, 2011; Nozaka, were intruded by Cretaceous or Paleogene granitic rocks 2014a), I just briefly describe here an outline of the char- and consequently thermally metamorphosed (Arai, 1975; acteristics of olivine and serpentine. Primary olivine of Compositional variation of olivine in high-temperature serpentinites 221 the peridotites shows modes of occurrence with variable porphyroclasts with decreasing distance from antigorite effects of plastic deformation: relatively coarse, equant in serpentinites from the Happo complex. Such a compo- grains with subgrain boundaries; ^cleavable olivine′ with sitional variation is unclear in other ultramafic complexes prominent parting planes; and elongated elliptic porphyr probably because of intense low-T serpentinization and/ oclasts distributed in a matrix of fine-grained recrystal. or contact metamorphism. The plots of olivine composi- lized olivine and lepidoblastic antigorite. The primary ol- tion were obtained avoiding the bright halos and streaks ivine grains are serpentinized more or less from rims and in back-scattered electron images mentioned above but along parting planes or fractures. Antigorite blades pen- could be affected by modification along invisible frac- etrate the porphyroclastic or cleavable olivine and are as- tures or subgrain boundaries to some extent as discussed sociated with chlorite, diopside and/or brucite in the lep- later. Nickel contents show no systematic variation with idoblastic matrix of mylonites. These antigorite blades distance unlike Fo and Mn (Fig. 3). have been interpreted to be a product of high-T serpenti- In the diagrams of NiO versus Fo and MnO versus nization (Nozaka, 2005). Lizardite, which is a product of Fo contents, a cluster of plots from each sample exhibits low-T serpentinization, statically replaces olivine to form a trend of compositional variation from the magnesian mesh texture with tiny grains of opaque minerals (mostly primary olivine to the most ferroan retrograde olivine magnetite) and submicroscopic brucite. The mesh-form- (Fig. 4). The NiO-Fo plots are similar, though more scat- ing lizardite is replaced by metamorphic olivine in contact tered, to the ‘mantle olivine array’ defined by Takahashi aureoles around granitic intrusions. The contact metamor- et al. (1987). The trend is much evident in MnO-Fo dia- phic olivine occurs as discrete small grains or epitaxially gram and designated here as ‘retrograde trend' (Fig. 4b). grown rims over primary olivine grains, coexisting with Prograde olivine occurs in the contact aureoles of antigorite at a low grade zone (Zone I) and with talc, most of the ultramafic complexes of the Oeyama ophio- anthophyllite and orthopyroxene at higher grade zones lite. In a low-grade zone (Zone II), the prograde olivine (Nozaka, 2003, 2005, 2011; Nozaka and Ito, 2011; No- commonly has highly magnesian compositions (Arai, zaka, 2014b). 1975; Nozaka, 2005; Nozaka and Ito, 2011; Khedr and Reflecting the metamorphic history of the Oeyama si s s n n ophiolite, at least two generations of antigorite can be prograde olivine with relatively ferroan compositions observed: i.e., retrograde antigorite formed by high-T ser- (Nozaka, 2003). The ferroan prograde olivine is similar pentinization and prograde antigorite formed by meta- in Fo content to the retrograde olivine, but is different in morphism after low-T serpentinization. Similarly, I will highly variable contents of NiO and MnO (Fig. 4; No- refer to the olivine coexisting with the former antigorite zaka, 2003, 2010). Relatively ferroan olivine formed by or affected by the high-T serpentinization as ‘retrograde regional metamorphism has been reported from the San- olivine’ and that coexisting with the latter antigorite as bagawa metamorphic belt, and it is also different in NiO 'prograde olivine'. and MnO contents from the retrograde olivine (Kunugiza, The retrograde olivine is common in serpentinite 1980, 1982; Nozaka, unpublished data). mylonites from the Happo and Wakasa complexes and Retrograde antigorite coexisting with retrograde ol- locally found outside the contact aureole of the Ohsa- ivine occurs as elongated blades surrounding or penetrat- yama complex. This olivine does not form discrete grains ing primary olivine (Figs. 2e-2i). The antigorite blades but occurs as a part of chemically modifed primary oliv. commonly lack magnetite inclusions unlike lizardite ine in contact with retrograde antigorite blades. It cannot (Figs. 2d, 2g, and 2h) and locally has the intercalation be distinguished from primary olivine under the optical of elongated magnetite, chlorite or brucite. A small microscope but is visible by a high electron back-scatter- amount of awaruite in contact with the antigorite blades ing intensity resulting from an Fe-rich composition (Figs. was observed in one sample (HP21). Retrograde antigor- 2a-2f, 3, and Table 1). Such an enrichment of Fe is lack- ite has aluminous compositions indicative of the incorpo- ing in olivine at any portions in contact with lizardite ration of chlorite components, and lacks evidence for mix- (Figs. 2c and 2f). Back-scattered electron images show ing with brucite unlike the case of lizardite (Figs. 5a and bright halos around some small antigorite blades that are 5c), which shows a brucite-mixing trend as reported from seemingly inclusions in olivine but possibly fracture-fill- many localities (e.g., Beard et al., 2009). However, the ings cut by the thin-section surface (Fig. 2c) and locally Al-enrichment is less significant in antigorite apart from show a small amount of bright streaks ~ 1 μm wide (not matrix chlorite as the case of fracture-filling antigorite in shown) within olivine grains. olivine porphyroclasts (Fig. 2d). Lizardite is commonly Figure 3 shows a systematic decrease of Fo [= 100 x less aluminous than retrograde antigorite, excepting bas- Mg/(Mg + Fe)] and increase of Mn contents in olivine tite-forming Al-rich lizardite (not shown in the figures). 222 T. Nozaka Figure 2. Mode of occurrence of olivine and antigorite. Abbreviatio ite; Ol, olivine; P-Ol, prograde olivine. (a) Photomicrograph of an olivine erpentinite mylonite from Happo (crossed porphy polars, sample# HP40). (b) Back-scattered electron image of the shown in (a). (c) Back scatteredelectron dynamically recrystallized olivine neoblasts and lepidoblastic antigorite of & rpentinite mylonite from Happo (crossed polars, sample# HP11). (f) Back-scattered electron image of a part of (e). Color version is available online from https://doi.org/10.2465/jmps.180420. Prograde antigorite occurs in contact metamorphic ite is characteristically smaller in size and in aspect ratio zones where prograde olivine occurs. This type of antigor- (length/width) than retrograde antigorite, and typically 223 O OI b1t 1.0mm Figure 2 (Continued). Mode of occurrence of olivine and antigorite. Abbreviations for minerals: Atg, antigorite; Chl, chlorite; Liz, lizardite; Mgt, magnetite; Ol, olivine; P-Ol, prograde olivine. (g) Photomicrograph of retrograde antigorite and lizardite replacing olivine in a gently deformed serpentinite from Wakasa (plane-polarized light, sample# WS0301). (h) Back-scatered electron image of a part of (g). (i) Photo- micrograph of écleavable olivine’ surrounded and penetrated by antigorite from a low-grade contact aureole (Zone I) in the Wakasa e d o r ( ( go pd e o s o psa ( (o #s sd pss xs from a low-grade contact aureole (Zone II) in the Oeyama complex (crossed polars, sample# OE0817). (l) Back-scattered electron image of a part of (k). Color version is available online from https://doi.org/10.2465/jmps.180420. 224 T. Nozaka 90.0 91.0- 89.0 I1 (a) HP40 88.0 (b) HP11 87.0 品 1500 I1c dd I1α Figure 3. Variation of forsterite (Fo), M M Mn, and Ni contents of olivine 1000- with the distance from the contact 1000- with (a) fracture-flling antigorite (sample #HP40 shown in Fig. 2b), and (b) antigorite matrix (sam- ple #HP 11 shown in Fig. 2f). 3500 Curves indicate possible diffusion profiles drawn to fit a set of plots 3500 (circle) with high Fo and low Mn values at each distance. Other plots (square), which show a considera- ble deviation from the diffusion profile, probably represent modifi- 口口口 1 cation along invisible fractures or 2500- subgrain boundaries. Color version 100 200 300 400 100 200 300 400 is available online from https:// Distance from antigorite vein (μm) Distance from antigorite matrix (μm) doi.org/10.2465/jmps.180420. forms fine-grained aggregate with subordinate amounts of grade antigorite in the peridotites of the Oeyama ophio- magnetite (Figs. 2k and 2l). Antigorite of regional meta- lite. Further evidence is that plots of Xre of the most morphic origin from the Ryumon ultramafic body in the ferroan olivine-antigorite pair in each sample are consis- Sanbagawa belt is also smaller in grain size than the ret- tent with those of coexisting pairs from other localities rograde antigorite of the Oeyama ophiolite and shows (Fig. 6). The enrichment of Fe in olivine porphyroclasts granoblastic texture with metamorphic olivine (Kunugiza, and the deficiency of magnetite associated with retro- 1980, 1982), suggesting their contemporaneous forma- grade antigorite indicate a reducing condition and support tion. The prograde antigorite, including that of the Ryu- the hypothesis that Mg-Fe interdiffusion in olivine is ef- mon body, typically has more magnesian compositions fective under the conditions of high-T serpentinization (XMg > 0.96) than the retrograde antigorite (XMg = 0.94- proposed by Evans (2010). Although a small amount of 0.96) (Fig. 5d). The antigorite blades that have a typical elongated magnetite is intercalated in some antigorite shape, size, and mode of occurrence of retrograde antigor- blades, they are not necessarily formed contemporane- ite occur in low-grade contact aureoles as well (Figs. 2i ously. Because brucite showing a similar mode of occur- and 2j) and have compositions scattering in both the fields rence to magnetite was observed in some serpentinite of typical retrograde and prograde antigorite (Fig. 5d). mylonites, the possibility that the magnetite replaced brucite, as suggested by Bach et al. (2006) and Beard DISCUSSION et al. (2009), cannot be ruled out. The variations of Fo content with distance from an- Compositional modification of olivine tigorite (Fig. 3) suggest that Mg-Fe interdiffusion took place within the olivine porphyroclasts, i.e., between The textural relationship strongly suggests the contempo- the cores keeping primary compositions and the rims of raneous formation of the retrograde olivine and retro- retrograde olivine coexisting with the retrograde antigor- Compositional variation of olivine in high-temperature serpentinites 225 Table 1. Representative microprobe analyses of olivine and serpentine Sample HP40 HP40 HP40 HP40 HP11 HP11 HP11 WS0615 WS0615 WS0615 WS0615 WS0615 OE0817 Mineral 10 10 Atg Liz 10 10 Atg 10 10 10 Atg Liz Atg Type Pri R R Pri R R-P Pri R P R-P P SiO2 40.51 40.60 42.66 42.36 40.06 39.89 43.02 40.21 40.06 40.82 43.24 39.76 44.94 TiO2 pu nd 0.02 0.03 pu nd 0.01 pu pu nd 0.01 0.00 0.00 nd Al2O3 pu 2.22 0.26 pu pu 2.54 nd nd nd 1.16 0.01 0.08 Cr2O3 nd nd 0.20 0.00 nd nd 0.24 nd pu pu 0.33 0.00 0.00 FeO* 10.22 13.06 3.90 2.10 9.79 11.62 3.59 9.50 10.48 5.81 2.49 1.96 1.53 MnO 0.15 0.25 0.05 0.10 0.14 0.22 0.04 0.13 0.18 0.04 0.05 0.05 0.00 NiO 0.39 0.35 0.22 0.07 0.38 0.37 0.16 0.39 0.40 0.22 0.06 0.25 0.05 MgO 49.16 46.33 36.65 38.80 50.03 47.63 36.76 49.98 48.69 53.12 37.61 39.25 39.57 CaO 0.00 0.01 0.00 0.11 0.00 0.00 0.01 0.00 0.01 0.00 0.01 0.04 0.00 Na2O pu nd 0.00 0.01 nd nd 0.04 pu pu pu 0.00 0.01 0.01 K20 nd nd 0.01 0.02 pu nd 0.00 nd nd nd 0.00 0.01 0.00 Total 100.45 100.59 85.94 83.86 100.40 99.73 86.41 100.21 99.81 100.00 84.96 81.33 86.18 0= 4.000 4.000 7.000 7.000 4.000 4.000 7.000 4.000 4.000 4.000 7.000 7.000 7.000 Si 0.992 1.004 2.023 2.043 0.981 0.991 2.023 0.985 0.989 0.984 2.056 1.988 2.093 Ti pu pu 0.001 0.001 nd nd 0.000 pu pu pu 0.000 0.000 0.000 Al nd nd 0.124 0.015 nd nd 0.141 nd nd nd 0.065 0.000 0.004 Cr nd nd 0.007 0.000 nd pu 0.009 nd nd nd 0.013 0.000 0.000 Fe 0.209 0.270 0.154 0.085 0.201 0.241 0.141 0.195 0.217 0.117 0.099 0.082 0.060 Mn 0.003 0.005 0.002 0.004 0.003 0.005 0.002 0.003 0.004 0.001 0.002 0.002 0.000 Ni 0.008 0.007 0.009 0.003 0.007 0.007 0.006 0.008 0.008 0.004 0.002 0.010 0.002 Mg 1.795 1.709 2.591 2.790 1.827 1.764 2.578 1.825 1.793 1.909 2.666 2.926 2.747 Ca 0.000 0.000 0.000 0.006 0.000 0.000 0.001 0.000 0.000 0.000 0.001 0.002 0.000 Na nd nd 0.000 0.001 nd nd 0.004 nd nd pu 0.000 0.000 0.001 K pu nd 0.001 0.001 pu pu 0.000 pu nd pu 0.000 0.000 0.000 Total 3.008 2.996 4.911 4.949 3.019 3.009 4.904 3.015 3.011 3.016 4.905 5.012 4.906 XMg 0.896 0.863 0.944 0.971 0.901 0.880 0.948 0.904 0.892 0.942 0.964 0.973 0.979 * Total iron as FeO; nd, not determined; XMg = Mg/Mg + Fe) molar ratio. Mineral abbreviations: Atg, antigorite; Liz, lizardite; Ol, olivine. Wakasa (WS) and Oeyama (OE). ite. Relatively low Fo values distributed below the diffu- ism. Since volume diffusion is a process more sluggish sion profile curves (Fig. 3) could be the effects of hidden than grain/subgrain boundary diffusion or pipe diffusion, (not appeared on the thin-section surface) irregular small the serpentinization of the Oeyama ophiolite could pro- fractures filled with antigorite as shown in Figure 2c. ceed at a higher-T or over a longer period than the ser- Stripes of Fe-enriched olivine have been reported and pentinization accompanied by the ferroan olivine stripes. interpreted as a result of Mg-Fe interdiffusion along sub- Because Mg-Fe volume diffusion in olivine looks unre- grain boundaries, or pipe diffusion along dislocations, alistic below 400 °C as discussed later, it probably took within deformed olivine crystals (Kitamura et al., 1986; place under the conditions of high-T serpentinization. Ando et al., 2001; Murata et al., 2009a, 2009b; Plumper Figure 3 indicates that Mn is enriched in retrograde et al., 2012). Although the bright streaks visible in back- olivine as well as Fe. As pointed out by Evans (2010) and scattered electron images suggest that an Fe-enrichment Plimper et al. (2012), this enrichment can be explained along linear defects within olivine crystals took place in by the average distribution coeffcient of Mn established the Oeyama ophiolite as well, their amount and frequen- by Trommsdorff and Evans (1974). For example, the XMn cy are just minor and volume diffusion, which forms a [= Mn/(Mg + Fe + Mn + Ni)] of the most ferroan olivine systematic compositional variation depending on the dis- in two samples (HP11 and HP40) are 0.0019-0.0026, and tance from antigorite, is probably the dominant mechan- applying the Mn distribution coefficient Kp (antigorite/ 226 T. Nozaka 0.6 0.3- (a) b 0.5- 0.4 MOA 0.2 % MOA N 0.3 O!N MnO 壬1o 0.2 PFO 0.1- ●HP40 0.1 HP11 WK0615 壬1α 0.0 0.0- 86 87 88 89 90 91 86 87 88 89 90 91 Fo mol% Fo mol% Figure 4. (a) NiO versus Fo contents and (b) MnO versus Fo contents of olivine in contact with antigorite in serpentinite mylonites (sample# by high-T serpentinization and intracrystalline element diffusion. Compositional fields of olivine from the literature are shown for compar- ison: MOA, mantle olivine array (Takahashi et al., 1987); PFO, prograde metamorphic ferroan olivine from Zone II of contact aureoles, SW Japan (Nozaka, 2003, 2010). Iron-rich prograde metamorphic olivine from the Ryumon peridotites plot within PFO or significantly out of the retrograde trends, or outside the scale ranges (Kunugiza, 1980, 1982; Nozaka, unpublished data). Color version is available online from https:/doi.org/10.2465/jmps.180420. olivine) = 0.18 (Trommsdorff and Evans, 1974) to these ite, as pointed out by Plumper et al. (2012). In fact, the olivine, the calculated Xmn of antigorite are 0.0003- coexistence of awaruite with antigorite, though a small 0.0005. These values are consistent with the XMn of an- amount, was observed in at least one sample from the alyzed antigorite: 0.0002-0.0008 with an average of Oeyama ophiolite. 0.0005 (24 analyses). It is in no doubt that the successive zonal arrange- The Ni contents of olivine porphyroclasts do not ments of olivine compositions in NiO-Fo and MnO- show such a systematic variation as Fo and Mn (Fig. 3). Fo diagrams (Fig. 4) are the results of intracrystalline In the same manner as Mn, using the Kp = 0.65 (Tromms- element diffusion between the most ferroan retrograde dorff and Evans, 1974), XNi was calculated for antigorite olivine and primary olivine. It is the case as well for sam- coexisting ferroan olivine to be 0.0021-0.0027. These val- ples from a low-grade zone (Zone I) of contact aureole ues are not inconsistent with the analyzed values of anti- (e.g., HP11 and WS0615 shown in Fig. 4), in which high- gorite: 0.0015-0.0034, and therefore retrograde olivine ly magnesian prograde olivine also occurs (Fig. 2j), sug- and retrograde antigorite probably tend to keep their equi- gesting the existence of relict retrograde olivine in the librium for Ni as well. However, such a diffusion profile contact aureole. Thermal metamorphism with tempera- as the case of Mn cannot be defined for Ni in the olivine ture conditions below 480 °C, which were estimated for porphyroclasts (Fig. 3). Olivine, which has enrichments of that zone (Nozaka, 2011), looks to have no effect on the Fe and Mn showing deviation from the diffusion profiles compositions of retrograde olivine. at a significant distance from antigorite probably due to We can recognize the effects of high-T serpentiniza- fluid infiltration along fractures or dislocations, shows an tion on the compositional modification of olivine from increase and decrease of Ni from the primary composition textural evidence as shown in some samples (Figs. 2b, (e.g., the three data on the extreme right of Fig. 3a). This 2f, and 3) and as previously reported (Murata et al., opposite behavior of Ni suggests a variation of Ni content 2009a, 2009b; Plimper et al., 2012). However, such a of fluid during high-T serpentinization. The increase of case is uncommon because of the overprinting effects Ni probably results from its distribution between olivine of intense low-T serpentinization and later metamor- and antigorite (Kp = 0.65 according to Trommsdorff phism. The occurrence of ferroan olivine is an indicator and Evans, 1974), whereas the decrease of Ni could be of high-T serpentinization, but we cannot exclude other explained by the formation of Ni-rich phase, e.g., awaru- possibilities for the formation of ferroan olivine in ser- Compositional variation of olivine in high-temperature serpentinites 227 pentinized peridotites of ophiolites: i.e., crystallization (Figs. 2i and 2j). differentiation of basaltic magma, deformation-induced fGakken School,KonoBuilding,27-17Takashima-cho,Numazu410-0056,aan pipe diffusion, metasomatic fuid infiltration, and pro- equilibration compositions are maintained. Retrograde an- grade metamorphism. In these cases, a set of NiO-Fo tigorite is relatively rich in Fe resulting in low Xmg (Figs. and MnO-Fo diagrams (Fig. 4) could be useful to distin- 5b and 5d). The contrastingly high XMg of prograde anti- guish the ferroan olivine of different origin. In a given gorite is likely inherited from lizardite, a reactant from sample from the mantle section of ophiolites, composi- which the prograde antigorite has formed. During low-T tions of olivine produced or affected by high-T serpenti- serpentinization, Fe in olivine is mainly incorporated into nization are expected to be plotted around the retrograde magnetite (e.g., Nozaka, 2003; Evans, 2010), some of trends of Fo-MnO without no significant deviation of Minoo-higashi High School, 5-4-63 Ao-Gein, Minoo 552-0025,Japan NiO from the mantle olivine array (Fig. 4). From this DepartmentfGesciencehmaneiversityatsu90504,an type of olivine, ferroan olivine produced by other pro- nesian compositions of lizardite. Although the XMg of cesses could be different in some points: a progressive lizardite and antigorite varies depending on primary oliv- depletion of NiO and a gentle enrichment of MnO with ine or bulk rock compositions, the retrograde antigorite decreasing Fo contents by magmatic differentiation directly formed after primary olivine should be richer in (Takahashi et al., 1987); no significant change of NiO Fe than prograde antigorite after lizardite in any samples. and MnO contents by high-T pipe diffusion in anhydrous, However, some relict blades of retrograde antigorite in not serpentinized peridotite (Ando et al., 2001); much contact aureoles have a slightly magnesian compositions less extent of Fo, NiO, and MnO variations by amphib- probably resulting from a reequilibration process during olite-facies metasomatism (Nozaka, 2014a); and highly E-mail address: imaoka@yamaguchi-u.ac.jp (T. Imaoka). variable NiO and MnO contents with a large deviation ture of retrograde antigorite of the Oeyama ophiolite is from the retrograde trend and the mantle olivine array an Al-enrichment (Figs. 5c and 5d). This could be an in- by prograde contact/regional metamorphism (Kunugiza, dication of increasing chlorite component in antigorite 1980, 1982; Nozaka, 2003, 2010). under high-T conditions (Padron-Navarta et al., 2013). Before discussion about T conditions, however, a caution Retrograde vs. prograde antigorite is needed for whether or not the antigorite coexists with chlorite; in fact, antigorite in olivine fractures apart from Retrograde antigorite would be easily found if it shows matrix chlorite, tends to be deficient in Al. textural and chemical equilibration with retrograde oliv- * Corresponding author. Tel.: +81 83 933 5765; fax: +81 83 933 5273. ine, but retrograde olivine grains might disappear because ite requires the diagnostic characteristics: a shape of elon- of intense low-T serpentinization. The question is wheth- gated blade, the absence of a significant amount of mag- er it is possible to discriminate between retrograde and netite inclusion, and a composition richer in Fe than that prograde antigorite in the latter case. of lizardite. These characteristics would provide empiri- The two types of antigorite have differences in some cal evidence for high-T serpentinization even in intensely microscopic features. Retrograde antigorite commonly serpentinized peridotites from ophiolites. However, the occurs as relatively large blades with high aspect ratios possibility of chemical disturbance in prograde metamor- and without magnetite inclusions (Figs. 2d, 2g, and 2h). phic rocks should be carefully investigated. - s o gates with minor amounts of magnetite in contact aur- Tectonic implication eoles (Figs. 2k and 2l) or occurs as relatively small blades forming an equilibration texture; e.g., granoblastic tex- The diffusion profiles of olivine porphyroclasts (Fig. 3) ture with prograde metamorphic olivine (e.g., Kunugiza, provide a constraint for tectonic model of the forearc 1980). However, retrograde antigorite also occurs as a mantle wedge (Fig. 7; Schmidt and Poli, 1998; Peacock small blade (Fig. 2d), and the shape or size may not be and Wang, 1999; Stern, 2002; Hacker et al., 2003; Rey- concluding evidence for identification. nard et al., 2011; Nozaka, 2014a). The illustrated diffu- 0024-4937/S - see front matter @ 2013 Elsevier B.V. All rights reserved. sion profiles differ between the two samples (Fig. 3) in a given by metamorphic zonal mapping. In the contact au- reflection of primary olivine composition, and possibly in reoles of SW Japan, prograde antigorite occurs in Zone addition, crystallographic orientation, degree of crystal II, which is defined by the first appearance of prograde deformation and T condition. Nevertheless, similar diffu- metamorphic olivine (e.g., Nozaka, 2003, 2011). How- sion distances are determined: ~ 35 μm for sample HP40 ever, this is not a condition for excluding the occurrence and ~ 45 μm for sample HP11, which allow us to esti- of retrograde antigorite, because it also occurs in Zone II mate the time necessary for the diffusion. 228 T. Nozaka Brc 0.4 been intensely metasomatized (Ishimaru et al., 2009). These 105 Ma (a) Figure 5. Serpentine and chlorite OPAtg compositions calculated on the ba- PAtg (RM) R-P Atg sis of 7 oxygen atoms: (a) Mg + 5 Liz ±Brc Fe* versus Si (Fe*, total iron as Fig. 1. Index map of Southwest Japan (A), geotectonic divisions of post-Triassic accretionary complexes in the western part of SW Japan (B), and distribution of Jurassic and Cretaceous Fe2+); (b) Fe* versus Si; (c) Al ver- II = 7) sus Si; (d) Al versus Mg/(Mg + Fe* O 0.2- Brc trend - Fe*), in which retrograde- and pro- Atg + Cina Tsuchiya et al., 2005). Studies of these hornblende peridotite xenoliths grade-antigorite composition fields rocks can contribute significantly to our understanding of the chemical in serpentinites from SW Japan are 0.1- highlighted. All the plotted serpen- Chl trend tine and chlorite are replacing oliv- ine, coexisting with olivine, or 0.0- forming mylonite matrix, and bas- tite-forming minerals are not plot- Si (0 = 7) Si (0= 7) ted. Arrows indicate mixing trend Atcin with brucite or chlorite. Abbrevia- 1.0- (c) (d) tions: Brc, brucite; Chl, chlorite; 0.2 Cln, clinochlore; Liz, lizardite; and physical processes operating in the subcontinental lithosphere be- 0.8- Retrograde Ol, olivine; P Atg, prograde anti- -antigorite gorite; R Atg, retrograde antigor- ite; R-P Atg, retrograde antigorite = a :() i s o = o e 's = on = m 's sa = ) s () 7 Prograde that has a typical shape, size and = 0) 1995, 2000), as well as the high-Mg, Nb-rich lamprophyres that are antigorite mode of occurrence in prograde 0.4- A contact aureoles. Prograde antigor- ite from the regionally metamor- phosed Ryumon (RM) peridotite 0.2- body (Kunugiza, 1982; Nozaka, unpublished data) plots for com- [Brc Brctrend plement. Color version is avail- 0.0+ 白 2Atg 0.92 0.94 0.96 0.98 1.00 able online from https://doi.org/ Si (O = 7) Mg/(Mg + Fe*) 10.2465/jmps.180420. Arc crust High-T serpentinization 600°℃ Exhumation 800°℃- Mantle wedge characteristics that include Si02 ≥ 56 wt.%, Al203 ≥ 15 wt.%, Mg0 1000℃ interpreted as indicating both the presence of residual garnet and Amphibolite-facies metasomatism. 100°0 0.00 0.10 0.20 Fe*(Mg + Fe*) oline Figure 6. Plots of Fe*/(Mg + Fe*) of the most ferroan olivine- antigorite pair in each sample (stars) from the Oeyama ophiolite Fluidinfiltration (sample# HP11, HP40, HP44, WS0615, OS0502). The field of olivine-antigorite pairs compiled by Evans (2010) is also shown Figure 7. Schematic model of the sequence of metasomatism, hy- for comparison. Color version is available online from https:// dration and exhumation of the forearc mantle peridotites repre- doi.org/10.2465/jmps.180420. sented by the Oeyama ophiolite based on a synthesis of some previous works (Schmidt and Poli, 1998; Peacock and Wang, The diffusion time was calculated on the basis of the 1999; Stern, 2002; Hacker et al., 2003; Reynard et al., 2011; equation: x ~ √Dt, where x, D and t is diffusion distance, Nozaka, 2014a). See Nozaka (2014a) for discussion. Color version is available online from https://doi.org/10.2465/jmps. diffusion coefficient and time, respectively, using a diffu- igneous rocks in and around the Korean Peninsula and the western part of SW Japan (C). Fig. 1c is compiled from Kee et al. (2010), Teraoka and Okumura (2003), and Wu et al. sion distance of 40 μm (Fig. 3) and the diffusion coeff- cient based on the global equation of Dohmen and Chak- raborty (2007a, 2007b). The diffusion coefficient for the equation, and it is necessary to subtract log 6, i.e., direction of [Oo1] of olivine is obtained by the global ~ 0.78 order of magnitude for [100] and [010] according 229 to Dohmen and Chakraborty (2007a, 2007b). However, be useful to distinguish retrograde or retrogressively the difference of D below 700 °C is less than 0.3 order of modified olivine from ferroan olivine of different origin, magnitude (Dohmen et al., 2007). Actually, the difference even if intense low-T serpentinization has obscured the is considered to be much less because the interference early record of the diffusion process. Retrograde antigor- colors under the microscope indicate that the directions and metabasites, and radiometric ages indicate metamorphism in dikes are dominated by hornblende and plagioclase with <2 vol.% of elongated blade, lacking a significant amount of mag- tain the minimum duration of cooling during or after netite inclusion, and having a composition richer in Fe high-T serpentinization, T conditions were supposed to than that of lizardite. These characteristics are expected be close to the maximum estimation for high-T serpenti- to provide empirical evidence for high-T serpentinization nization of the Oeyama ophiolite (Nozaka, 2014a). The even in intensely serpentinized peridotites from ophio- rounding granitoids of the San-yo and Ryoke belts. Hornblende and lites. Approximate estimation of time required for the ob- ~ 4.0 × 107 years at 500 °C and 0.5 GPa. An unreason- served diffusion profiles of olivine porphyroclasts sug- ably long time (> 4.5 × 10? years) was estimated at 400 gests a long residence time of the forearc peridotites °C, suggesting the observed intracrystalline element dif- within the serpentinized mantle wedge following rapid topes of the Early Cretaceous igneous rocks from the Kinki district of exhumation immediately after the amphibolite-facies of high-T serpentinization. metasomatism. In contrast, much shorter diffusion time of tens to hundreds of years has been estimated for the diffusion ACKNOWLEDGMENTS profile caused by the fuid injection that formed oliv- ine-phlogopite veins in the Oeyama ophiolite and inter- I thank Dr. Masaki Mifune for permission to use a Raman preted as a result of rapid exhumation after amphibolite- spectrometer in his laboratory at Okayama University. facies metasomatism of the forearc mantle wedge (No- Comments from reviewers, J. Ando and E. Takazawa zaka, 2014a). That short time was obtained using D of have improved the manuscript. This study was finan- Jaoul et al. (1995), which is one or two orders of magni- cially supported by JSPS KAKENHI Grant Number tude larger than that of Dohmen and Chakraborty (2007a, JP16K05611. 2007b). Using the latter D in this study, the time was recalculated to be ~ 5.0 × 10² to 8.0 × 103 years at 750- SUPPLEMENTARYMATERIALS 800 °C, which is still an extremely short period in the geologic time frame. Although exact time required for Color versions of Figures 1-7 are available online from fusions. Errors and the analytical precision of the ICP-MS analyses are https://doi.0rg/10.2465/jmps.180420. because of the uncertainties of D, T, and pressure condi- tions, the approximate estimation suggests that the cool- REFERENCES ing duration under the conditions of high-T serpentiniza- tion was at least three to four orders of magnitude longer Ando, J., Shibata, Y., Okajima, Y., Kanagawa, K., Furusho, M. and Tomioka, N. (2001) Striped iron zoning of olivine induced by than that of the amphibolite-facies metasomatism in the dislocation creep in deformed peridotites. Nature, 414, 893- Oeyama ophiolite. It is concluded that the peridotites of 895. the Oeyama ophiolite retained for a considerably long Arai, S. (1975) Contact metamorphosed dunite-harzburgite com- period in the region of serpentinized mantle wedge fol- plex in the Chugoku district, western Japan. Contributions to lowing the rapid exhumation immediately after the am- Mineralogy and Petrology, 52, 1-16. Arai, S. and Yurimoto, H. (1995) Possible sub-arc origin of pod- phibolite-facies metasomatism (Fig. 7). iform chromitites. Island Arc, 4, 101-111. Bach, W., Paulick, H., Garrido, C.J., Idefonse, B., Meurer, W.P. CONCLUDING REMARKS and Humphris, S.E. (2006) Unraveling the sequence of ser- pentinization reactions: petrography, mineral chemistry, and petrophysics of serpentinites from MAR 15°N (ODP Leg Iron-enriched retrograde olivine instead of magnetite was 209, Site 1274). Geophysical Research Letters, 33, L13306, formed along with retrograde antigorite by high-T ser- doi:10.1029/2006GL025681. pentinization in the Oeyama ophiolite. Intracrystalline Beard, J.S., Frost, B.R., Fryer, P., McCaig, A., Searle, R., Ilde- volume diffusion of Mg, Fe, and Mn between the most fonse, B., Zinin, P. and Sharma, S.K. (2009) Onset and pro- ferroan retrograde olivine, which coexists with retrograde gression of serpentinization and magnetite formation in oliv- ine-rich troctolite from IODP Hole U1309D. Journal of Pet- antigorite, and primary olivine took place under the con- rology, 50, 387-403. ditions of high-T serpentinization. The diffusion pro- magmatism in SW Japan fits into the regional geotectonic evolution of duced a characteristic zonal arrangement of olivine com- for the electron microanalysis of silicates and oxides. Journal positions in MnO-Fo and NiO-Fo diagrams, which could of Geology, 76, 382-403. Japan. The analytical procedures, accuracies, and reproducibilities T. Nozaka Dohmen, R., Becker, H. and Chakraborty, S. (2007) Fe-Mg diffu- Shank, T.M., Butterfeld, D.A., Hayes, J.M., Schrenk, M.O., sion in olivine I: experimental determination between 700 and Olson, E.J., Proskurowski, G., Jakuba, M., Bradley, A., Lar- 1,200°C as a function of composition, crystal orientation and son, B., Ludwig, K., Glickson, D., Buckman, K., Bradley, oxygen fugacity. Physics and Chemistry of Minerals, 34, 389- A.S., Brazelton, W.J., Roe, K., Elend, M.J., Delacour, A., Ber- PNt+t/PNt1 Pue IS9s/IS8 Pa.1nseaN (9 = u) S000000 F L60ZLS'0 nasconi, S.M., Lilley, M.D., Baross, J.A., Summons, R.E. and Dohmen, R. and Chakraborty, S. (2007a) Fe-Mg diffusion in oliv- Sylva, S.P. (2005) A serpentinite-hosted ecosystem: The Lost ine Il: point defect chemistry, change of diffusion mechanisms City hydrothermal feld. Science, 307, 1428-1434. and a model for calculation of diffusion coeffcients in natural Khedr, M.Z. and Arai, S. (2010) Hydrous peridotites with Ti-rich olivine. Physics and Chemistry of Minerals, 34, 409-430. chromian spinel as a low-temperature forearc mantle facies: Dohmen, R. and Chakraborty, S. (2007b) Erratum: Fe-Mg diffu- evidence from the Happo-O'ne metaperidotites (Japan). Con- sion in olivine II: point defect chemistry, change of diffusion tributions to Mineralogy and Petrology, 159, 137-157. mechanisms and a model for calculation of diffusion coeffi- Khedr, M.Z. and Arai, S. (2012) Petrology and geochemistry of cients in natural olivine. Physics and Chemistry of Minerals, prograde deserpentinized peridotites from Happo-O'ne, Ja- 34,597-598. pan: Evidence of element mobility during deserpentinization. Evans, B.W. (2010) Lizardite versus antigorite serpentinite: Mag- Journal of Asian Earth Sciences, 43, 150-163. netite, hydrogen, and life(?). Geology, 38, 879-882. Kitamura, M., Matsuda, H. and Morimoto, N. (1986) Direct obser- Frost, B.R. (1985) On the stability of sulfides, oxides, and native vation of the Cottrell atmosphere in olivine. Proceedings of metals in serpentinite. Journal of Petrology, 26, 31-63. the Japan Academy, Series B, 62, 149-152. Frost, B.R., Evans, K.A., Swapp, S.M., Beard, J.S. and Mother- Klein, F., Bach, W. and McCollom, T.M. (2013) Compositional sole, F.E. (2013) The process of serpentinization in dunite controls on hydrogen generation during serpentinization of from New Caledonia. Lithos, 178, 24-39. ultramafic rocks. Lithos, 178, 55-69. Frih-Green, G.L., Kelley, D.S., Bernasconi, S.M., Karson, J.A., Klein, F., Grozeva, N.G., Seewald, J.S., McCollom, T.M., Humph- Ludwig, K.A., Butterfield, D.A., Boschi, C. and Proskurow- ris, S.E., Moskowitz, B., Berquo, T.S. and Kahl, W. (2015) ski, G. (2003) 30,000 years of hydrothermal activity at the Experimental constraints on fluid-rock reactions during incip- Lost City vent field. Science, 301, 495-498. ient serpentinization of harzburgite. American Mineralogist, Hacker, B.R., Peacock, S.M., Abers, G.A. and Holloway, S.D. 100,991-1002. (2003) Subduction factory 2. Are intermediate-depth earth- Kunugiza, K. (1980) Dunites and serpentinites in the Sanbagawa quakes in subducting slabs linked to metamorphic dehydra- metamorphic belt, central Shikoku and Ki Peninsula, Japan. tion reactions? Journal of Geophysical Research, 108, doi: The Journal of the Japanese Association of Mineralogists, Pet- 10.1029/2001JB001129. rologists and Economic Geologists, 75, 14-20. Ishiwatari, A. (1989) Ophiolites of Japan. Journal of Geography, Kunugiza, K. (1982) Formation of zoning of olivine with progres- 98, 290-303 (in Japanese with English abstract). - no s Ishiwatari, A. (1990) Yakuno ophiolite and related rocks in Mai- Ryumon peridotite body of the Sanbagawa metamorphic zuru terrane. In Pre-Cretaceous terranes of Japan (Ichikawa, belt, Kii Peninsula. The Journal of the Japanese Association K. et al. Eds.). Osaka City University, Osaka. Publication of of Mineralogists, Petrologists and Economic Geologists, 77, IGCP 224, 109-120. 157-170. Ishiwatari, A. and Tsujimori, T. (2003) Paleozoic ophiolites and McCollom, T.M. and Bach, W. (2009) Thermodynamic constraints blueschists in Japan and Russian Primorye in the tectonic on hydrogen generation during serpentinization of ultramafic framework of East Asia: A synthesis. Island Arc, 12, 190- rocks. Geochimica et Cosmochimica Acta, 73, 856-875. minerals are plagioclase (6-50 vol.%) and amphibole (34%-69%). Miyoshi, A., Kogiso, T., Ishikawa, N. and Mibe, K. (2014) Role of Isozaki, Y., Maruyama, S., Aoki, K., Nakama, T., Miyashita, A. silica for the progress of serpentinization reactions: Con- and Otoh, S. (2010) Geotectonic subdivision of the Japanese straints from successive changes in mineralogical textures of islands revisited: Categorization and definition of elements serpentinites from Iwanaidake ultramafic body, Japan. Amer- and boundaries of Pacific-type (Miyashiro-type) orogen. Jour- ican Mineralogist, 99, 1035-1044. nal of Geography, 119, 999-1053 (in Japanese with English Murata, K., Maekawa, H., Yokose, H., Yamamoto, K., Fujioka, K., with 5 Faraday cups. The analytical methods and the accuracy of Ishi, T., Chiba, H. and Wada, Y. (2009a) Significance of ser- Jaoul, O., Bertran-Alvarez, Y., Liebermann, R.C. and Price, G.D. pentinization of wedge mantle peridotites beneath Mariana (1995) Fe-Mg interdiffusion in olivine up to 9 GPa at T = forearc, western Pacific. Geosphere, 5, 90-104. 600-900°C; experimental data and comparison with defect Murata, K., Maekawa, H., Ishi, K., Mohammad, Y.O. and Yokose, calculations. Physics of the Earth and Planetary Interiors, 89, H. (2009b) Iron-rich stripe patterns in olivines of serpenti- by brown hornblende (Fig. 3B), suggesting a xenocrystic origin for nized peridotites from Mariana forearc seamounts, western Katayama, I., Kurosaki, I. and Hirauchi, K. (2010) Low silica ac- Pacific. Journal of Mineralogical and Petrological Sciences, tivity for hydrogen generation during serpentinization: An ex- 104,199-203. ample of natural serpentinites in the Mineoka ophiolite com- Nakamura, Y. and Kushiro, I. (1970) Compositional relations of plex, central Japan. Earth and Planetary Science Letters, 298, coexisting orthopyroxene, pigeonite and augite in a tholeitic 'L7800 = IS9g/98 'SHOL'0 = ISsg/ISs :saneA (1)PN3 pue (1)1S3 Su! andesite from Hakone volcano. Contributions to Mineralogy Kelley, D.S., Karson, J.A., Blackman, D.K., Fruh-Green, G.L., and Petrology, 26, 265-275. Butterfield, D.A., Lilley, M.D., Olson, E.J., Schrenk, M.O., Nishimura, Y. (1998) Geotectonic subdivision and areal extent of Roe, K.K., Lebon, G.T., Rivizzigno, P. and AT3-60 Shipboard the Sangun belt, Inner Zone of Southwest Japan. Journal of Party (2001) An off-axis hydrothermal vent field near the Metamorphic Geology, 16, 129-140. Mid-Atlantic Ridge at 30°N. Nature, 412, 145-149. Nozaka, T. (2003) Compositional heterogeneity of olivine in ther- Kelley, D.S., Karson, J.A., Frih-Green, G.L., Yoerger, D.R., mally metamorphosed serpentinite from Southwest Japan. 231 American Mineralogist, 88, 1377-1384. ma generation. Earth and Planetary Science Letters, 163, 361- Nozaka, T. (2005) Metamorphic history of serpentinite mylonites 379. from the Happo ultramafic complex, central Japan. Journal of Schulte, M., Blake, D., Hoehler, T. and McCollom, T. (2006) Ser- Metamorphic Geology, 23, 711-723. pentinization and its implications for life on the early Earth Nozaka, T. (2010) A note on compositional variation of olivine and and Mars. Astrobiology, 6, 364-376. pyroxene in thermally metamorphosed ultramafic complexes Schwarzenbach, E.M., Caddick, M.J., Beard, J.S. and Bodnar, R.J. from SW Japan. Okayama University Earth Science Reports, (2016) Serpentinization, element transfer, and the progressive 17, 1-5. development of zoning in veins: evidence from a partially Nozaka, T. (2011) Constraints on anthophyllite formation in ther- serpentinized harzburgite. Contributions to Mineralogy and mally metamorphosed peridotites from southwestern Japan. Petrology, 171, 5, doi:10.1007/s00410-015-1219-3. Journal of Metamorphic Geology, 29, 385-398. Sleep, N.H., Meibom, A., Fridriksson, Th., Coleman, R.G. and Nozaka, T. (2014a) Metasomatic hydration of the Oeyama forearc Bird, D.K. (2004) Hz-rich fluids from serpentinization: Geo- peridotites: Tectonic implications. Lithos, 184-187, 346-360. chemical and biotic implications. Proceedings of the National Nozaka, T. (2014b) Comment on “Dehydration breakdown of anti- Academy of Sciences of the United States of America, 101, gorite and the formation of B-type olivine CPO" by Nagaya et 12818-12823. al. (2014). Earth and Planetary Science Letters, 408, 402-405. Stern, R.J. (2002) Subduction zones. Reviews of Geophysics, 40, Nozaka, T. and Ito, Y. (2011) Cleavable olivine in serpentinite my- 1012,doi:10.1029/2001RG000108. lonites from the Oeyama ophiolite. Journal of Mineralogical Takahashi, E., Uto, K. and Schilling, J.G. (1987) Primary magma and Petrological Sciences, 106, 36-50. compositions and Mg/Fe ratios of their mantle residues along Padron-Navarta, J.A., Sanchez-Vizcaino, V.L., Hermann, J., Con- Mid Atlantic Ridge 29°N to 73°N. Technical Report of Insti- nolly, J.A.D., Garrido, C.J., Gomez-Pugnaire, M.T. and Mar- tute for Study of the Earth's Interior, Okayama University, chesi, C. (2013) Tschermak's substitution in antigorite and Series A, 9, 1-14. consequences for phase relations and water liberation in Takai, K., Gamo, T, Tsunogai, U., Nakayama, N., Hirayama, H., high-grade serpentinites. Lithos, 178, 186-196. Nealson, K.H. and Horikoshi, K. (2004) Geochemical and Peacock, S.M. and Wang, K. (1999) Seismic consequences of warm versus cool subduction metamorphism: Examples from mophilic subsurface lithoautotrophic microbial ecosystem southwest and northeast Japan. Science, 286, 937-939. (HyperSLiME) beneath an active deep-sea hydrothermal Plumper, O., King, H.E., Vollmer, C., Ramasse, Q., Jung, H. and feld. Extremophiles, 8, 269-282. Austrheim, H. (2012) The legacy of crystal-plastic deforma- Takeuchi, M. (2002) The Paleozoic. In Geology of the Shirouma- tion in olivine: high-diffusivity pathways during serpentiniza- dake district: Quadrangle series, 1:50,000 (Nakano, S. et al. tion. Contributions to Mineralogy and Petrology, 163, 701- Eds.). Geological Survey of Japan, Tsukuba, 9-21 (in Japa- 724. nese). Reynard, B., Mibe, K. and Van de Moortele, B. (2011) Electrical Trommsdorff, V. and Evans, B.W. (1974) Alpine metamorphism of conductivity of the serpentinized mantle and fuid flow in sub- peridotitic rocks. Schweizerische Mineralogische und Petrog- duction zones. Earth and Planetary Science Letters, 307, 387- raphische Mitteilungen, 54, 333-352. 394. Russell, M.J., Hall, A.J. and Martin, W. (2010) Serpentinization as a source of energy at the origin of life. Geobiology, 8, 355- Manuscript received April 20, 2018 371. Manuscript accepted July 23, 2018 Schmidt, M.W. and Poli, S. (1998) Experimentally based water Published online October 6, 2018 budgets for dehydrating slabs and consequences for arc mag- Manuscript handled by M. Satish-Kumar
|
Nozaka (2018) - Compositional variation of olivine related to high-temperature serpentinization of peridotites.txt
|
Journal of Mineralogical and Petrological Sciences, Volume 115, page 431-439, 2020 A new occurrence of okhotskite in the Kurosegawa belt, Kyushu, Japan: the okhotskite + Mn-lawsonite assemblage as a potential high-pressure indicator Wataru YABUTA and Takao HIRAJIMA Department of Geology and Mineralogy, Division of Earth and Planetary Sciences, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan We present the first report of okhotskite in a lawsonite-blueschist-subfacies metachert of the Hakoishi subunit, Kurosegawa Belt, Kyushu, Japan, which was metamorphosed at peak temperatures and pressures of 200-300 °C sures than that of previously documented okhotskite with available pressure estimations. Textural relationships indicate that okhotskite formed during peak metamorphism in equilibrium with piemontite, Na pyroxene, mag- nesioriebeckite, braunite, and hematite. Okhotskite shows a significant variation in Fe:Mn ratio (Feto/Mntot = 0.13-0.56) and a following average empirical formula; (Ca7.62Mn2.16)≥7.78(Mn2.71Mg1.29)≥4.00(Mn4.13Fe2.26Al1.36 V323 Tio.02)≥8.0oSi11.86O44.02(OH)16.98. Raman spectra of okhotskite are reported for the first time and show char- acteristic peaks at 362, 480, and 563 cm-1. The stability relationships between okhotskite and other Mn-bearing minerals, such as piemontite, sursassite, spessartine, braunite, and Mn-bearing lawsonite, are examined using a revised Schreinemakers? analysis. The obtained petrogenetic grid provides tight constraints on the P-T relation- ship of natural mineral assemblages observed in Mn-bearing cherts within epidote-blueschist-grade and law- sonite-blueschist-grade. Furthermore, this petrogenetic grid predicts that the assemblage of okhotskite and Mn- bearing lawsonite should be stable at higher pressures. The higher-pressure stability suggests that highly oxi- dized Mn-bearing metacherts can transport water and buffer oxygen in the deeper parts of subduction zones, given that okhotskite and Mn-bearing lawsonite contain high water contents (6.9 and 11.3 wt% H2O, respec- tively) and trivalent manganese. Keywords: Okhotskite, Lawsonite-blueschist subfacies, Metachert, Kurosegawa Belt, Schreinemakers’ analy- sis, Raman spectra INTRODUCTION do, Japan (Togari and Akasaka, 1987). Okhotskite has also been reported from metamor- Pumpellyite group of minerals have a general chemical phosed manganese deposits in the Sanbagawa (Mina- formula of [W&X4YsZ12O56-n(OH)n] (Passaglia and Got- kawa, 1992) and Chichibu (Minakawa et al., 2008) belts tardi, 1973) and are common hydrous minerals in low- in Shikoku, Japan; metamorphosed manganese oxide grade metamorphic rocks. Root-names of the group have ores of the Precambrian Sausar Group, India (Dasgupta been defined based on the dominant cation in the Y-site: et al., 1991); and manganese deposits in the Askiz ore pumpellyite with A13+ (Palache and Vassar, 1925), julgol- district, Khakassia, Russia (Kassandrov and Mazurov, dite with Fe3+ (Moore, 1971), shuiskite with Cr3+ (Ivanov 2009). These reports include the coexistence okhotskite et al., 1981), okhotskite with Mn3+ (Togari and Akasaka with braunite (Mn2+Mn3+SiO12) and/or bixbyite (Mn2O3), 1987), and poppite with V3+ (Brigatti et al., 2006). and the presence of trivalent Mn suggests that the mineral Okhotskite, the Mn3+-dominant member, was first report- records a strongly oxidizing environment. Furthermore, ed from the Kokuriki mine, in the Tokoro Belt, Hokkai- okhotskite is expected to provide a considerable part of the water budget of impure metacherts in the subduction doi:10.2465/jmps.190831 W. Yabuta: yabuta.wataru.25n@kyoto-u.jp Corresponding author zone, because of its high H2O content (6.89 wt% H2O; T. Hirajima: hirajima.takao.4u @ kyoto-u.ac.jp Togari and Akasaka, 1987).
432 W. Yabuta and T. Hirajima In considering phase relations among such highly (a) sw Japan oxidized Mn minerals, oxygen fugacity (fo2) has been considered as a critical factor in addition to pressure (P) and temperature (T) (e.g., Mottana, 1986). Numerous synthetic experiments have been conducted at various fo, values controlled by buffers of certain mineral assem- blages (e.g., Abs-Wurmbach and Peters, 1999). Akasaka et al. (2003) conducted a series of hydrothermal experi- ments at 400-500 °C and 300 MPa to constrain the stabil- ity of Mn-rich pumpellyite. Those authors confirmed Mn- (b) bearing pumpellyite as the precursor to piemontite and that it is stable at lower temperatures and higher oxygen TheYatsushiroArea fugacity, as proposed by Akasaka et al. (1988). Reinecke OD70 (1986) investigated phase relations in Mn-Al-rich quartz- OT19 ites of the upper-blueschist facies from Evvia and Andros Y1811 Serpentinite Tobiishisub-unit islands, Greece, to schematically show the low-T stability Hak koishisub-unit Sedimentarycomplexes of sursassite-bearing assemblages from Evvia when com- Metabasaltic rock Metasiliceous rock OMetachert (this study) pared to spessartine-bearing assemblages from Andros. These previous studies were based on the premise Figure 1. (a) Distribution of the Kurosegawa belt in SW Japan and study area. (b) Geotectonic map of the Kurosegawa Belt in the that oxygen is a perfectly mobile component (Korzhin- Yatsushiro area, Kyushu, Japan (modified after Saito et al., 2010; ski, 1959) and therefore that cations (e.g., Fe and Mn) Sato et al., 2016). The okhotskite-bearing sample was collected can be freely oxidized or reduced in metamorphic rocks. from outcrop OT10 in the western part of the Hakoishi sub-unit, However, some recent studies have found that oxygen which occupies the western half of the Otao unit. Localities of behaves as an inert element in natural systems (Tumiati other Mn-bearing samples, i.e., OT19, KY1811, and OD70, are also shown. et al., 2015; Li et al., 2016). In such cases, the oxidation states of cations are instead fixed, and fo, is lowered as a dependent variable that describes bulk cation oxidation melange that is exposed over an area of 2 × 10 km2, oc- states. Thus the phase relations among Mn-rich minerals cupying the western half of the Otao unit (Saito et al., need to be reconsidered in terms of oxygen mobility. Fur- 2010; Sato et al., 2014; Fig. 1). The Hakoishi subunit con- thermore, during geochemical interaction in subduction sists mainly of serpentinites, lawsonite-bearing blue- zones, oxygen is bound to highly oxidized rocks and is schists (LBSs), metasiliceous rocks, and minor metagran- thus transported deep into the Earth by subduction. Re- ite and is separated from adjacent units by serpentinites fining the relevant phase relations should provide a clear (Saito et al., 2010). Saito et al. (2010) mapped areas of n , ssnd e metabasaltic rocks and metasiliceous rocks, but in the mantle (Thomson et al., 2016). s p o s e In this paper, okhotskite-bearing rocks from a new lated with each other and form a coherent block (Ibuki et locality from the Hakoishi sub-unit of the Kurosegawa al., 2008). Sato et al. (2016) reported that lawsonite + Na belt in Kyushu are described, including the chemical com- -s ru sd o si dnd + auxnd position and Raman spectra of the okhotskite. A provi- semblage in LBS of the eastern part of the study area sional petrogenetic grid is developed for Mn-bearing min- and that lawsonite + Na amphibole + Na pyroxene forms erals under a quartz-excess environment. We also discuss the predominant one in LBS of the western part. Such the behavior of oxygen and the significance of highly oxi- a systematic distribution of low-variance assemblages dized minerals, including okhotskite, in subduction zones. shows that the pumpellyite + Na pyroxene + chlorite + H2O = lawsonite + Na amphibole reaction is recorded in GEOLOGICAL BACKGROUND AND the study area, and an increase in grade to the west is MINERALOGY associated with hydration. The peak metamorphic condi- tions have been estimated to be 200-300 °C and 0.60-0.80 The Usuki-Yatsushiro Tectonic Line (UYTL) trends NE- GPa (Sato et al., 2016). Notably, the Hakoishi LBS (i.e., SW across Kyushu Island, Japan (Fig. 1). In the Yatsu- metabasalt) lacks epidote (Fujimoto et al., 2010; Kami- shiro area, the Kurosegawa Belt lies to the south of the mura et al., 2012; Sato et al., 2014; Sato et al., 2016). UYTL. The studied sample was collected from the west- Metasiliceous rocks in the study area are rich in ern part of the Hakoishi subunit, which is a serpentinite manganese and rare-earth elements (REEs), and are
A new occurrence of okhotskite and its petrological significance 433 a b layer Fe-Mn-rich layer Qtz veins Fe-Mn-rich cm 50um C (d) Okhotskite 30μm 100μm layer Figure 2. Texture of the okhotskite-bearing sample (OT10J). (a) Hand-specimen photograph. (b) Photomicrograph of okhotskite grains within (PPL). found in the abandoned Taneyama mine (Yoshimura, (Figs. 2a and 2d). The Fe-Mn-rich layers are fractured 1969; Yabuta and Hirajima, 2018). The Mn-rich rocks and include quartz-dominated veins that contain okhot- have been interpreted as impure metacherts that formed skite (Figs. 2a-2c). The different parts of the sample con- from siliceous oozes mixed with oceanic sediments, giv- sist of different mineral assemblages maybe reflecting the en that manganese and REEs are common in abyssal local “bulk-rock’ composition (Table 1). The quartz-rich sediments (Yoshimura, 1969; Togari et al., 1988; Kato layers consist of quartz + Na pyroxene + piemontite + et al., 2011). A variety of Mn-bearing minerals have been braunite + albite + apatite. The Fe-Mn-rich layers are reported from metasiliceous rocks related to the aban- composed of hematite + braunite + quartz ± albite ± apa- doned Taneyama mine (e.g., taneyamalite: Aoki et al., tite ± titanite. The quartz veins are composed predomi- 1981; howieite: Ibuki et al., 2008; Mn3+-bearing lawson- nantly of quartz along with Na amphibole + okhotskite ± ite: Ibuki et al., 2010). braunite ± albite ± apatite, and notably developed within The okhotskite-bearing sample (OT101/J) is a band- Fe-Mn-rich layers. ed rock composed of centimeter-thick intercalated quartz-rich and Fe-Mn-rich layers composed of fine- METHODS grained (<30 μm) minerals. Boundary layers (<100 μm thick) composed of Na pyroxene and amphibole (with The chemical composition of okhotskite was determined minor hematite, piemontite and okhotskite) occur at the using a Hitachi S3500H scanning electron microscope contacts between the quartz-rich and Fe-Mn-rich layers equipped with an energy-dispersive X-ray analyzer
434 W. Yabuta and T. Hirajima Table 1. Mineral assemblages observed in sample OT10 Qz Hem Okh Pmt Br Na-Amp Na-Px Ab Ttn Ap Qz-rich layers · · · + 十 Fe/Mn- Matrix · Rich layers Qz-veins · + Boundary layers + O, common; +, rare. (EDAX?) at Kyoto University, Kyoto, Japan. Operating Fe3+, V3+) = 8.00 (after Kato et al., 1981; Togari and Aka- conditions included an accelerating voltage of 20 kV, a saka, 1987). Although H2O content in the Hakoishi okhot- beam current of 500 pA, and a spot size of 5 μm. Fol- skite was not measured in the present study, it is calculated lowing materials were used as standards; albite for Na, as 7.15 wt% H2O by difference of the average chemical synthetic periclase for Mg, synthetic corundum for Al, composition, which is similar to the values (6.89 wt% synthetic quartz for Si, orthoclase for K, synthetic wol- H2O) of the Tokoro okhotskite determined by Togari and lastonite for Ca, synthetic rutile for Ti, vanadium metal Akasaka (1987). From the given formula for the average for V, manganese metal for Mn, and hematite for Fe. A composition of the Hakoishi okhotskite and the nomencla- ZAF correction scheme was used to process the X-ray ture scheme after Passaglia and Gottardi (1973), this min- intensity data. eral should be called okhotskite-(Mn2+). Raman spectra of okhotskite were acquired using Piemontite is present as a major Mn-bearing mineral an NRS-3100 JASCO spectrometer at the Department in the quartz-rich layers and boundary layers, and shows of Geology and Mineralogy, Kyoto University, using characteristic striking pinkish-red to yellow pleochroism a diode-pumped solid-state (DPSS) 532 nm laser. The (Fig. 2d). Some grains show oscillatory zoning, which is wave number was calibrated using the 520.7 cm-1 line attributed to variable REE contents (e.g., La and Nd). The produced by a Si-wafer, and the spectrum of Ne. chemical composition of the REE-poor zones is close to We also conducted thermodynamic calculations to that of ideal piemontite: Ca2Al2Mn3+SisO12(OH) (Table 2). estimate phase relations among Mn-bearing minerals. Na pyroxene occurs as acicular grains with a low Thermodynamic parameters of mineral phases followed molar fraction of diopside (Di < 15 mol%) (Table 2). those in the standard state of Holland and Powell (1998) The proportions of aegirine (Aeg) and jadeite (Jd) are or were calculated by the summation method of Fyfe et variable and reflect the local bulk-rock composition, with al. (1958) based on the database of Robie and Hemi- Aeg45JdsoDis in the quartz-rich layers and Aeg7oJd2oDi10 ngway (1995). Parameters of fuid species were calculat- in the boundary layers (Yabuta and Hirajima, 2018). The ed after Woolley (1953) and Burnham et al. (1969). Na amphibole is Mg-rich, consistent with a high ratio of Schreinemakers’ nets were generated with Calnet ver. ferric to ferrous iron, and is classified as magnesiorie- 2.1 developed by Yoshida and Hirajima (1999). beckite according to the scheme of Leake et al. (1997) Mineral abbreviations follow Whitney and Evans (Table 2). Na amphibole grains in boundary layers and (2010), except for okhotskite (Okh), braunite (Br), and quartz veins show nearly identical compositions. Hema- sursassite (Sus). tite and braunite occur in the Fe-Mn-rich layers, reflect- ing the highly oxidized Si-rich composition (Abs-Wurm- CHEMICAL ANALYSES AND bach, 1980). Braunite also occurs as tiny grains of ~ 20 RAMAN SPECTROSCOPY μm diameter in the quartz-rich layers. Figure 4 compares the Raman spectrum of an okhot- Okhotskite shows characteristic orange to dark-orange skite grain with those of related minerals. This is the first pleochroism, commonly occurs within the quartz veins as report of Raman spectrum of okhotskite, so the character- fine-grained (<30 μm long) crystals, and is uncommon in istic peaks are described. Peaks are observed consistently the boundary layers (Figs. 2b and 2c). The average chemi- close to 362, 480, 537, 563, 593, 684, and 921 cm-1. Even cal formula is (Ca7.62Mn2.16)≥7.78(Mn27Mg1.29)≥4.00(Mn4.13 though the Raman spectra of julgoldite-Fe and Mg-pum- Fe2.26A11.36V323Ti0.02)≥8.00Si11.86O4.02(OH)16.98, -with a Va- pellyite also show a peak near 550 cm-', the spectrum of riety in Fe:Mn ratio (Fetot/Mntot = 0.13-0.56; Table 2 and okhotskite show few similarities to those of other pum- Fig. 3). Total Fe was assumed as trivalent, and the oxida- pellyite group minerals. The peaks close to 362, 480, and tion state of the Mn was calculated based on Z(Mn3+, Al, 563 cm-1 are considered to be diagnostic of okhotskite.
A new occurrence of okhotskite and its petrological significance 435 Table 2. Mineral compositions for sample OT10 Okhotskite Pmt Na-Px Braunite Na-Amp Mineral (∑ cation = 32) (0 = 12.5) (0=6) (0 = 12) (0 = 23) Sample OT10J OT101 Mean OT10J 20 75 24 26 33 N=5 2 94 123 115 SiO2 33.77 33.85 32.69 32.63 33.35 33.26 36.63 55.84 11.44 57.60 TiO2 IPq IPq 0.07 IPq 0.31 0.10 IPq 0.32 0.10 0.48 Al2O3 3.58 2.17 3.93 3.82 2.72 3.24 19.85 11.69 0.33 1.39 V2O3 0.35 0.82 1.10 1.11 0.63 0.80 n.d. n.d. n.d. n.d. Fe2O3* 4.03 10.45 9.84 9.10 8.69 8.42 2.42 14.81 5.23 12.78 Mn2O3** 18.93 14.74 12.36 13.02 15.27 14.98 14.10 0.45 66.88 0.01 MnO** 10.59 8.91 8.70 9.48 11.24 9.68 0.53 IPq 10.14 1.20 MgO 1.31 3.02 3.41 2.90 1.52 2.43 0.20 1.80 0.24 15.52 CaO 19.75 19.59 20.53 20.05 19.79 19.94 22.17 1.51 1.73 1.48 CuO n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.95 n.d. Na2O n.d. n.d. n.d. n.d. n.d. n.d. n.d. 13.66 n.d. 7.29 K2O n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.15 92.31 93.55 92.63 92.11 93.52 92.85 95.90 100.08 97.04 97.90 Total Si 12.308 12.146 11.685 11.783 12.065 12.010 3.019 1.999 1.169 8.012 Ti 0.000 0.019 0.000 0.084 0.026 0.009 0.008 0.050 A1 1.538 0.918 1.656 1.626 1.160 1.381 1.928 0.493 0.040 0.228 V3+ 0.104 0.235 0.316 0.322 0.184 0.191 Fe3+ 1.106 2.821 2.646 2.473 2.366 2.289 0.150 0.399 0.403 1.338 Mn3+ 5.252 4.027 3.363 3.579 4.206 4.120 0.884 0.012 5.204 0.001 Mn2+ 3.268 2.708 2.634 2.899 3.445 2.960 0.037 0.878 0.142 Mg 0.712 1.615 1.817 1.561 0.820 1.309 0.025 0.096 0.037 3.218 Ca 7.712 7.531 7.863 7.757 7.671 7.716 1.957 0.058 0.189 0.221 Cu 0.073 Na 0.948 1.966 K 0.027 Total 32.000 32.000 32.000 32.000 32.000 32.000 8.000 4.014 8.000 15.201 * Total Fe as Fe2O3. ** Recalculated values. For okhotskite, Mn in the X- and Y-sites was assumed to be Mn2+ and Mn3, respectively, and the Y-site was flled with A13+, Mn3t, and Fe3+ in this order, following the method proposed by Akasaka et al. (1997). For piemontite, braunite, and Na-Amp, Mn2+ and Mn3+ were calculated based on the charge balance. Pmt, piemontite; Na-Px, Na pyroxene; Na-Amp, Na amphibole. Bixbyite-bearing Bixbyite-free PROVISIONAL SCHREINEMAKERS' NET (host-rocks) (host-rocks) Tokoro belt (1) Hakoishi this study) Sanbagawabelt(3) Reinecke (1986) performed thermodynamic calculations Okh veins Kamisugai mine Kamogawa mine O Sausar group(2) among Mn-bearing minerals, such as piemontite, Mn- pumpellyite, sursassite, spessartine, and chlorite, in a pseudo-quaternary system of Mg-Mn*(= Mn2+ + Mn3+)- Ca-Al, with excess Qz, H2O, and O2. This is applicable to C the Hakoishi metacherts, as quartz is saturated in the 0& △ whole sample and relevant minerals except in the pres- ence of garnet. 落 However, the Schreinemakers’ net proposed by Re- inecke (1986) has a fundamental shortcoming. The net n n ] =) d l, od ga s ： 紫 senting reactions where the two Ca-bearing phases are ** /Fe3+ Julgoldite Okhotskite Mn3 absent. As Reinecke's analyses do not deal with other Ca-phases, when one occurs as a reactant in a reaction, Figure 3. Y-site occupancy in okhotskite. Sources: (1) Tokoro the other is always present as a product. Consequently, Mn-Fe ore deposits (Togari et al., 1988). (2) Bichhua Forma- tion, Sausar Group, India (Dasgupta et al., 1991). (3) Mn ore either of the two is present on the same side as the invar- deposits, Sanbagawa belt (Minakawa, 1992). iant point [Pm, Pp] across a reaction curve, contravening
436 W. Yabuta and T. Hirajima Table 3. Thermodynamic parameters of Schreinemakers’ analyses Mineral Chemical composition Volume [J/bar·mol] Entropy [J/K·mol] Lawsonite (Lws) Ca(A11.8Mn3+0.2)Si2O7(OH)2:H2O 10.21 (1) 239.0 (2) Sursassite (Sus) (Cao.5Mn2+3.5)(A14Mn3+)Si6O21(OH)7 28.30 (3) 724.7(2) Piemontite (Pmt) Ca2Al2Mn3+Si3O12(OH) 14.02 (4) 307.7(2) Braunite (Br) (Mn2+0.9Ca0.1)Mn3+6SiO12 12.51(4) 415.8 (2) Spessartine (Sps) (Mn2+2.7Ca0.3)Al2Si3O12 11.89 (5) 314.1 (5) Ca4(Mn2+0.5Mg.5)(Mn33.5A1.5)Si6O21(OH)7 Okhotskite (Okh) 30.70 (6) 753.7(2) Chlorite (Chl) MgsAl2Si3O10(OH)8 21.09 (5) 410.5 (5) Quartz (Qz) SiO2 2.269 (5) 41.50 (5) Sources: () Ibuki et al. (2010). (2) calculated in this study (see the text). (3) Nagashima et al. (2009). (4) Anthony et al. (1995) and references therein. (5) Holland and Powell (1998). (6) Togari and Akasaka (1987). fied in the Hakoishi metacherts (Yabuta and Hirajima, （a) Okhotskite 2019), making it suitable for estimating phase relations in the metamorphic rocks of interest. In addition, compo- sitions of garnet, braunite, and sursassite were revised to contain Ca, reflecting natural occurrences in Greece and Kyushu (Table 3). The detail petrography and mineralogy of these Mn-bearing minerals in the Hakoishi sub-unit will be shown elsewhere. The Calnet program yields two possible bundles of reactions, corresponding to the invariant points of [Pmt] and [Okh]. We selected the latter to account for the min- h Julgoldite-(Fe) eral assemblages of Evvia-Andros (Reinecke, 1986). In the provisional petrogenetic grid (Fig. 5), the rep- resentative mineral assemblages of Evvia (Pmt + Sus + Br + Chl) and Andros (Pmt + Sps + Br + Chl) are stable around the invariant point [Lws]. The two assemblages (C) Pumpellyite-(Mg) are separated by the following reaction: (RRUFFDB) Sus + Br = Pmt + Sps + Chl + Qz + H2O + O2 (1) 500 1000 corresponding to the reaction (8) of Reinecke (1986). Raman shift (cm-1) In contrast, mineral assemblages in the Hakoishi metacherts (Sus + Chl + Pmt, Lws + Sus + Chl, and Figure 4. (a) Raman spectra of okhotskite. (b) and (c) Related Lws + Br + Chl; W. Yabuta, unpublished data, 2020) minerals. Julgoldite-Fe (ID R070725) and pumpellyite-Mg (R120172) are from the RRUFF database. are stable around the invariant point [Okh]. This does not contradict the occurrence of okhotskite, which can stably coexist with Br and Pmt on the low-P side of a basic rule of phase diagrams. the reaction curve: In addition, the petrogenetic grid contains viridine, Mn-bearing andalusite. Viridine has been reported from Lws + Okh + O2 = Pmt + Br + Chl + Qz + H2O many Mn-rich metamorphic rocks of relatively lower (2). metamorphic pressure but has not been identified on Ev- via or Andros islands (Greece) or in Hakoishi. Therefore, On the whole, the petrogenetic grid indicates that it is questionable to consider phase relations in blueschist the Evvia-Andros assemblages are stable at higher pres- facies (kyanite stability) with viridine. Therefore, in place sure and temperature compared with the Hakoishi meta- of viridine, we used Mn-bearing lawsonite (Ibuki et al., cherts. This is consistent with estimated P-T conditions 2010). The Mn-bearing lawsonite is ubiquitously identi- of the two areas: 350-450 °C and 1.0-1.2 GPa for south-
A new occurrence of okhotskite and its petrological significance 437 +Qtz Mg ite-(Mn2+) commonly coexist with okhotskite (Togari et +HO al., 1988; Dasgupta et al., 1991). This can be explained in +O2 terms of the higher metamorphic pressure that affected the Hakoishi subunit; i.e., 200-300 °C and 0.60-0.80 GPa as O [Br] AI4 estimated for Zone 2 at Hakoishi (Sato et al., 2016), and 200-230 °C and 0.25-0.35 GPa as estimated for rock units Sps [Chl] Mn* in the Kokuriki mine (Sakakibara, 1991). Instead of pum- pellyite-(Mn?+), sursassite and Mn-lawsonite occur as major minerals in the Hakoishi metacherts. Figure 3 compares the compositions of the okhot- Andros O[Lws] skite described in this study with those of previous stud- ies. A wide range in Mn:Al ratio but a narrow range in Fe:Mn ratio in okhotskite from the Tokoro Belt and Sau- Hakoishi Pmt+Sps+Br+CI sar Group have been documented by Togari et al. (1988) OD70 Evvia and Dasgupta et al. (1991), respectively. In contrast, the okhotskite described in the present study shows a narrow range in Mn:Al ratio and a wide range in Fe:Mn ratio (Fig. 3). Minakawa (1992) and Kassandrov and Mazurov Reinecke(1986) (2009) reported similarly uniform Mn:Al ratios and con- sistently low Fe2O3 contents in okhotskite from the San- P bagawa belt and Askiz ore district. OT19 KY1811 These patterns of Y-site occupancy may be related to the presence or absence of bixbyite in host rocks, whereby A13+-Mn3+ substitution is dominant in the for- T mer case and Fe3+-Mn3+ substitution is dominant in the latter case. As most bixbyite-free rocks contain quartz, Figure 5. A provisional Schreinemakers’ net indicating stability relations among okhotskite and associated Mn-bearing phases assemblages of braunite + quartz and braunite + bixbyite (following a revised methodology after Reinecke, 1986). The may translate into Si-saturation in highly oxidized sys- P-T ratio is arbitrary. tems (Abs-Wurmbach, 1980). However, further investi- gation is required, as some previous studies have reported ern Evvia (Shaked et al., 2000) and 200-300 °C and okhotskite as mono-mineralic veins traversing the Mn- 0.60-0.80 GPa for Hakoishi (Sato et al., 2016). With ref- pumpellyite veinlets with/without piemontite (Togari et erence to the study of Akasaka et al. (2003), who sug- al., 1988); for such cases, equilibria within host rocks gested that Mn-pumpellyite is stable below 500 °C at 300 are not assured. MPa, although schematic, the petrogenetic grid in Figure The provisional Schreinemakers’ net presented here 5 describes the phase relations below that temperature. (Fig. 5) predicts that okhotskite and Mn3+-bearing law- sonite are characteristically stable at higher pressure com- DISCUSSION pared with the P-T conditions determined for Hakoishi and Evvia-Andros. These minerals contain abundant In the studied sample, okhotskite occurrence is limited to Mn3+ and host substantial contents of H2O (Okh: 6.89 an environment closely associated with braunite and he- wt%, Togari and Akasaka, 1987 and 7.15 wt% in this matite, where Mn3+ is abundant. In contrast, quartz-rich study; Lws: 11.3 wt%, calculated from data of Ibuki et layers contain piemontite as the Ca-bearing mineral, and al., 2010). This clearly shows that siliceous sediments Al is relatively concentrated in the presence of albite. can contain high contents of water in the deep Earth These observations suggest that local chemical composi- when impurities such as Mn-rich oceanic sediments are tions strongly control the mineral assemblage, supporting present, although such sediments have generally been the finding of Akasaka et al. (1988) and Togari et al. thought to be water poor (Hacker, 2008). (1988) on the compositional variations in Mn-rich pum- In terms of oxygen transport, the Schreinemakers' pellyite and piemontite. net in Reinecke (1986) assumes perfect mobility of oxy- It is also important to note that no Mn-rich pum- gen (Korzhinski, 1959), whereby oxygen is released as a pellyite (with a high occupancy of Al in the Y-site) is fuid from oxidized solid phases. Under reduced mobility observed in the Hakoishi metacherts, although pumpelly- in a subduction zone, oxygen fuid is transported into the
438 W. Yabuta and T. Hirajima mantle wedge along with highly oxidized minerals (as ite, Mn2+Mn3+Og/SiO4, in the System Mn-Si-O at 1 atm in noted by Tumiati et al., 2015). Indeed, oxygen fuid is Air. Contributions to Mineralogy and Petrology, 71, 393-399. Abs-Wurmbach, I. and Peters, T. (1999) The Mn-Al-Si-O system: highly complicated when considering phase relations, ab experimental study of phase relations applied to paragen- as its high entropy sometimes leads to the prediction of eses in manganese-rich ores and rocks. European Journal of contradictory phase relations. Mineralogy, 11, 45-68. For instance, phase relations between pumpellyite Akasaka, M., Sakakibara, M. and Togari, K. (1988) Piemontite from the manganiferous hematite ore deposits in the Tokoro and epidote groups are discussed as simplified oxidation Belt, Hokkaido, Japan. Mineralogy and Petrology, 38, 105- reactions (e.g., Fe-Pmp + O2 = Ep + H2O; Schiffman and 116. Liou, 1983; Mn-Pmp + O2 = Pmt + H2O; Akasaka et al., Akasaka, M., Kimura, Y., Omori, Y., Sakakibara, M., et al. (1997) 1988, 2003). These reactions, however, would have neg- 57Fe Mossbauer study of pumpellyite-okhotskite-julgoldite ative slope (△S < 0 and △V > 0, owing to high entropy of series minerals. Mineralogy and Petrology, 61, 181-198. Akasaka, M., Suzuki, Y. and Watanabe, H. (2003) Hydrothermal O2-fluid) with pumpellyite group stable at higher P-T Synthesis of Pumpellyite-Okhotskite Series Minerals. Miner- side, even when some other solid phases are present. alogy Petrology, 77, 25-37. Needless to say, such reaction curves on P-T field would Anthony, J.W., Bideaux, R.A., Bladh, K.W. and Nichols, M.C. not explain the high-temperature stability of the epidote (1995) Handbook of Mineralogy volume Il. Tucson, AZ: Mineral Data Publishing, ISBN 9780962209703. group against pumpellyite group in metabasites (e.g., Aoki, Y., Akasako, H. and Ishida, K. (1981) Taneyamalite, a new Nakajima et al., 1977). manganese silicate mineral from the Taneyama mine, Kuma- Such a simple case is sufficient to demonstrate prob- moto Prefecture, Japan. Mineralogical Journal, 10, 385-395. lems in oxygen-mobile models. The effect of oxygen Brigatti, M.F., Caprilli, E. and Marchesini, M. (2006) Poppite, the V3+ end-member of the pumpellyite group: Description and fluid on phase relations will be discussed in detail else- crystal structure. American Mineralogist, 91, 584-588. where. In the present study, it is noted that highly oxi- Burnham, C.W., Holloway, J.R. and Davis, N.F. (1969) Thermo- dized minerals (e.g., Okh and Br) are stable at high P-T dynamic properties of water to 1,000 °℃ and 10,000 bars. even if oxygen is released freely from the system as a Memoirs Geological Society of America, Special Paper 132. perfectly mobile element. This indicates that okhotskite Dasgupta, S., Chakraborti, S., Sengupta, P., Bhattacharya, P.K., serves as a component of the oxygen budget in deep sub- et al. (1991) Manganese-rich minerals of the pumpellyite duction zones, in addition to its role as a water reservoir group from the Precambrian Sausar Group, India. American Mineralogist, 76, 241-245. (Togari and Akasaka, 1987; Togari et al., 1988). Howev- Fujimoto, Y., Kono, Y., Hirajima, T., Ishikawa, M. and Arima, M. er, further investigation is required regarding the fate of (2010) P-wave velocity and anisotropy of lawsonite and epi- highly oxidized Mn-bearing minerals, including okhot- dote blueschists: Constraints on wave transportation along skite, in subduction environments, where their oxidizing subducting oceanic crust. Physics of the Earth and Planetary Interiors, 183, 219-228. capacity may oxidize the mantle hanging wall. Fyfe, W.S., Turner, F.J. and Verhoogen, J. (1958) Metamorphic reactions and metamorphic facies. Memoirs Geological Soci- ACKNOWLEDGMENTS ety of America, 73. 21 51. Hacker, B.R. (2008) H2O subduction beyond arcs. Geochemistry, We sincerely thank Dr. M. Akasaka, Professor Emeritus Geophysics, Geosystems, 9， Q03001, DOI:10.1029/ 2007GC001707. of Shimane University, and Dr. S. Endo for their sincere Holland, T.J.B. and Powell, R. (1998) An internally consistent ther- and constructive comments, which guided the revision of modynamic data set for phases of petrological interest. Jour- an earlier draft of the manuscript. We also thank Profes- nal of Metamorphic Geology, 16, 309-343. sor M. Satish Kumar for his efficient editorial handling of Ibuki, M., Fujimoto, Y., Takaya, M., Miyake, A. and Hirajima, T. (2008) Howieite in meta-manganese siliceous rocks of Kuro- the manuscript. We are grateful for the technical advice segawa belt, western Kyushu, Japan. Journal of Mineralogical offered and the high-quality thin-section samples pre- and Petrological Sciences, 103, 365-370. pared by Mr. M. Takaya and the revision of the Calnet Ibuki, M., Ohi, S., Tsuchiyama, A. and Hirajima, T. (2010) Analy- program by Mr. R. Kato. We acknowledge students of sis of Mn-bearing lawsonite occurring in meta-siliceous rocks Matsubase High School, Kumamoto, for their contribu- in Hakoishi serpentinite melange of Kurosegawa Belt, Central tion to the first report of piemontite in the study area. Kyushu, Japan. Journal of Mineralogical and Petrological Sci- ences, 105, 340-345. This work was supported by JSPS KAKENHI Grant No. Ivanov, O.K., Arkhangel'skaya, V.A., Miroshnikova, L.O. and Shi- JP19H1999 to T.H. lova, T.A. (1981) Shuiskite, the chromium analog of pum- pellyite, from the Bisersk deposit, Urals. Zapiski Vserossiy- skogo Mineralogicheskogo Obshchestva, 110, 508-512 (in REFERENCES Russian). Kamimura, K., Hirajima, T. and Fujimoto, Y. (2012) Finding of Abs-Wurmbach, I. (1980) Miscibility and Compatibility of braun- Prehnite-Pumpellyite facies metabasite from the Kurosegawa
A new occurrence of okhotskite and its petrological significance 439 belt in Yatsushiro area, Kyushu, Japan. Journal of Mineralo- with English abstract). gical and Petrological Sciences, 107, 99-104. Sakakibara, M. (1991) Metamorphic petrology of the Northern Kassandrov, E.G. and Mazurov, M.P. (2009) Magmatogenic Man- Tokoro metabasites, eastern Hokkaido, Japan. Journal of Pet- ganese Ores of the South Minusa Intermontane Trough. Geol- rology, 32, 333-364. ogy of Ore Deposits, 51, No. 5, 356-370. Sato, E., Hirajima, T., Kamimura, K. and Fujimoto, Y. (2014) Kato, A., Matsubara, S. and Yamamoto, R. (1981) Pumpellyite- White mica K-Ar ages from lawsonite-blueschist facies (Mn2t) from the Ochiai Mine, Yamanashi Prefecture, Japan. Hakoishi sub-unit and from prehnite-pumpellyite facies Tobi- Bulletin de Minéralogie, 104, 396-399. ishi sub-unit of the Kurosegawa belt, Kyushu, Japan. Journal Kato, Y., Fujinaga, K., Nakamura, K., Takaya, Y, et al. (2011) of Mineralogical and Petrological Sciences, 109, 258-270. Deep sea mud in the Pacific Ocean as a potential resource Sato, E., Hirajima, T., Yoshida, K., Kamimura, K. and Fujimoto, Y. for rare earth elements. Nature Geoscience, 4, 535-539. (2016) Phase relations of lawsonite-blueschists and their role Korzhinski, D.S. (1959) Physicochemical Basis of the Analysis of as a water-budget monitor: a case study from the Hakoishi the Paragenesis of Minerals. Consultants bureau, New York. sub-unit of the Kurosegawa belt, SW Japan. European Journal Leake, B.E., Woolley, A.R., Arps, C.E.S., Birch, W.D., et al. of Mineralogy, 28, 1029-1046. (1997) Nomenclature of Amphiboles: Report of the Subcom- Schiffman, P. and Liou, J.G. (1983) Synthesis of Fe-pumpellyite mittee on Amphiboles of the International Mineralogical As- and its stability relations with epidote. Journal of Metamor- sociation, Commission on New Minerals and Mineral Names. phic Geology, 1, 91-101. Canadian Mineralogist, 35, 219-246. Shaked, Y., Avigad, D. and Garfunkel, Z. (2000) Alpine high-pres- Li, J., Gao, J., Klemd, R., John, T. and Wang, X. (2016) Redox sure metamorphism at the Almyropotamos window (southern processes in subducting oceanic crust recorded by sulphide- Evvia, Greece). Geological Magazine, 137, 367-380. bearing high-pressure rocks and veins (SW Tianshan, China). Thomson, A.R., Walter, M.J., Kohn, S.C. and Brooker, R.A. Contributions to Mineralogy and Petrology, 171, 72. (2016) Slab melting as a barrier to deep carbon subduction. Minakawa, T. (1992) Study on characteristic mineral assemblages Nature, 529, 76-83. and formation process of metamorphosed manganese ore de- Togari, K. and Akasaka, M. (1987) Okhotskite, a new mineral, an posits in the Sanbagawa belt. Memoirs of the Faculty of Sci- Mn3-dominant member of the pumpellyite group, from the ence, Ehime University, 1, 1-74. (in Japanese with English Kokuriki mine, Hokkaido, Japan. Mineralogy Magazine, 51, abstract) 611-614. Minakawa, T., Fukushima, H., Nishio-Hamane, D. and Miura, H. Togari, K., Akasaka, M., Sakakibara, M. and Watanabe, T. (1988) (2008) Epidote-(Sr), CaSrAlFe3+(SizO7)(SiO4)(OH), a new Mineralogy of manganiferous iron ore deposits and chert from mineral from the Ananai mine, Kochi Prefecture, Japan. Jour- the Tokoro Belt, Hokkaido. Mining Geology Special Issue, nal of Mineralogical and Petrological Sciences, 103, 400-406. 12, 115-126. Moore, P.B. (1971) Julgoldite, the Fe2+-Fe3+ dominant pumpelly- Tumiati, S., Godard, G., Martin, S., Malaspina, N. and Poli, S. ite. Lithos, 4, 93-99. (2015) Ultra-oxidized rocks in subduction melanges? Decou- Mottana, A. (1986) Blueschist-facies metamorphism of manganif- pling between oxygen fugacity and oxygen availability in a erous cherts: A review of the alpine occurrences. Geological Mn-rich metasomatic environment. Lithos, 226, 116-130. Society of America Memoir, 164, 267-299. Whitney, D.L. and Evans, B.W. (2010) Abbreviations for names of Nagashima, M, Akasaka, M., Minakawa, T., Libowitzky, E. and rock-forming minerals. American Mineralogist, 95, 185-187. Armbruster, T. (2009) Sursassite: Hydrogen bonding, cation Woolley, H.W. (1953) Thermodynamic Properties of Molecular order, and pumpellyite intergrowth. American Mineralogist, Oxygen. National Bureau of Standards, Washington, D.C. 94, 1440-1449. Yabuta, W. and Hirajima, T. (2018) Mineralogy of Highly-Oxi- Nakajima, T., Banno, S. and Suzuki, T. (1977) Reactions leading to dized Lawsonite-Blueschist Facies Metachert from the Hako- the disappearance of pumpellyite in low-grade metamorphic ishi sub-unit, the Kurosegawa Belt, Kyushu and Proposal for rocks of the Sanbagawa metamorphic belt in central Shikoku, the Piedmontite Forming Reaction. Abstract of 2018 JpGU Japan. Journal of Petrology, 18, 263-284. Annual meeting, SMP-PO2, Chiba, Japan. Palache, C. and Vassar, H.E. (1925) Some minerals of the Kewee- Yabuta, W. and Hirajima, T. (2019) Blueschist facies impure meta- nawan copper deposits: Pumpellyite a new mineral; sericite; chert as remarkable oxygen- and overlooked water-carriers in saponite. American Mineralogist, 10, 412-418. the subduction zone. Abstract of 2019 Annual Meeting of Ja- Passaglia, E. and Gottardi, G. (1973) Crystal chemistry and no- pan Association of Mineralogical Sciences, R8-08, Fukuoka, menclature of pumpellyites and julgoldites. Canadian Miner- Japan. alogist, 12, 219-223. Yoshida, D. and Hirajima, T. (1999) The algorithm for drawing Reinecke, T. (1986) Phase relationships of sursassite and other petrogenetic grids with Schreinemakers’ method. Journal of Mn-silicates in highly oxidized low-grade, high-pressure Mineralogy, Petrology and Economic Geology, 94, 254-260. metamorphic rocks from Evvia and Andros Islands, Greece. Yoshimura, T. (1969) Supplement to “Manganese Ore Deposits of Contributions to Mineralogy and Petrology, 94, 110-126. Japan", Part I. Science Reports, Department of Geology, Robie, R.A. and Hemingway, B.S. (1995) Thermodynamic Proper- Kyushu University, 9 (Special Issue No. 2), 902 (in Japanese). ties of Minerals and Related Substances at 298.15 K and 1 Bar (10 Pascals) Pressure and at Higher Temparatures. U.S. Geo- logical Survey Bulletin, 2131. Manuscript received August 31, 2019 Saito, M., Takarada, S., Toshimitsu, S., Mizuno, K. et al. (2010) Manuscript accepted August 5, 2020 Geological map of Japan 1:200,000 Yatsushiro and a part of PublishedonlineOctober28,2020 Nomo Zaki. Geological Survey of Japan, AIST (in Japanese Manuscript handled by M. Satish-Kumar
|
Yabuta (2020) high pressure indicator minerals kurosegawa belt.txt
|
"Int J Earth Sci (Geol Rundsch) (2017) 106:1429-1451 CrossMark DOI 10.1007/s00531-016-1437-6 ORIGINA(...TRUNCATED)
| "Tsukada et al. 2017 Early Paleozoic subduction initiation volcanism of the Iwatsubodani Formation, (...TRUNCATED)
|
"OKAYAMA UNIVERSITY Earth Science Report, Vol. 1, No. 1, 1-8, (1994) Petrography of primary peridoti(...TRUNCATED)
| "Nozaka and Shibata (1994) - Petrography of primary peridotites from the Ohsa-yama area, Okayama Pre(...TRUNCATED)
|
"Gravity-Based Fault Mapping:\nThe Ishikari Lowland, Hokkaido, Japan\nA. Yamamoto\nInstitute of Seis(...TRUNCATED)
|
Yamamoto (2005) - Gravity-based fault mapping the Ishikari lowland.txt
|
"Geothermics 39 (2010) 228-241 Contents lists available at ScienceDirect Geothermics ELSEVIER journa(...TRUNCATED)
|
Saibi (2010) Temperature and chemical changes in the fluids of the Obama geothermal field.txt
|
"Geochemical Journal, Vol. 25, pp. 203 to 222, 1991\nGeochemistry of the Nigorikawa geothermal \n (...TRUNCATED)
|
Yoshida (1991) geochemistry of nogorikawa geothermal system.txt
|
End of preview. Expand
in Data Studio
README.md exists but content is empty.
- Downloads last month
- 2