Control factors in Na pyroxene compositions in lawsonite blueschist subfacies rocks: coexisting phases and kinetics

Abstract

Na pyroxene (Napx) of the aegirine-jadeite series occurs as a major metamorphic mineral in metacherts and metabasalts of the lawsonite blueschist (LBS) subfacies in the Kurosegawa belt, Kyushu, Japan. Its jadeite content [X_Jd = ^VIAl/(^VIAl + Fe³⁺ + Ca)] is highly variable at the thin section scale and is crucially influenced by associated minerals, such as albite and sodium amphibole (Namp). Whereas the coexistence with albite and quartz imposes an upper limit on X_Jd in Napx through the equilibrium reaction Ab = Jd + Qz, this paper reveals that uniformly low X_Jd was observed when albite was absent in the metacherts and the coexistence with Namp also effectively reduced X_Jd. In the former case available Na and Al content is limited in metacherts and in the latter case the Al/Fe³⁺ distribution between Napx and Namp controlled the X_Jd in metacherts and metabaslts. Notably, the distribution coefficient [K_D = (Al/Fe³⁺)^Namp/(Al/Fe³⁺)^Napx] systematically varied depending on the mode of growth: K_D ∼ 4 in metabasalts, where Napx developed replacing relict igneous pyroxenes (i.e., topotaxial growth), whereas K_D ∼ 1 in metacherts and metamorphic veins developed in metabasalts, where no relict phase was involved (i.e., nucleation and growth). Previous studies confirmed this trend, suggesting a convergence of K_D to 1 in higher grade rocks, such as epidote-blueschist and eclogite facies. This implies a kinetic control on the distribution process, particularly at lower temperatures (<300 °C). In summary, low-grade Napx associated with igneous pyroxene may represent a metastable state; thus, those in metacherts can serve as a more reliable proxy for the equilibrium conditions of the clinopyroxene system in these grades.

INTRODUCTION

Na pyroxene (Napx) is a major source of pressure-temperature (PT) information in metamorphic rocks. X_Jd has been used as a crucial indicator of metamorphic pressure, assuming the equilibrium of Ab = Jd + Qz, in early works such as Miyashiro and Banno (1958) and Essene and Fyfe (1967). Several authors have proposed thermodynamic models of the aegirine-jadeite-diopside system based on natural occurrences and synthetic experiments to represent equilibrium states over a wide PT range (Carpenter, 1979a, 1979b, 1980; Holland, 1980; Carpenter, 1982; Holland, 1983; Holland and Powell, 1996a, 1996b; Green et al., 2007).

However, the slow reaction rate at lower temperatures prevented such experiments from yielding reliable results. Additionally, natural low-grade metamorphic rocks occasionally contain Na pyroxenes with compositions falling within the miscibility gap expected by these models. Consequently, a close petrological study on natural samples of metamorphic pyroxenes accurately representing the equilibrium of lower grades is desired. In the interpretation, it is crucial to consider additional compositional controls beyond PT that may influence natural Na pyroxene compositions.

One factor is bulk rock composition. Okay (1978) compared the composition of Na pyroxene in metabasite and metachert from a blueschist facies terrain and found relatively higher Ca contents in the former with comparable jadeite contents. Manzotti et al. (2020) confirmed that the bulk Fe³⁺ content determines the composition and stability of Na pyroxene and sodium amphibole by constructing pseudosections on various lithotypes. Another factor, the rock fabric, has been discussed by several authors. Maruyama and Liou (1987) illustrated the role of igneous pyroxene as the nuclei of metamorphic Na pyroxene growth. Furthermore, Morata et al. (1994) associated primary rock composition with textual patterns. In tholeiitic metabasites, Na pyroxene is confined to the contact between igneous augite and veinlets, whereas it overgrows Ti-rich pyroxene (epitaxial growth) in transitional-alkaline metabasites. They argued that the Al contents in the latter cases are the product of Ca-Na replacement; thus, they do not directly represent the metamorphic pressure.

In this study, the Na pyroxene compositional trend from low grade [lawsonite blueschist (LBS) subfacies, <300 °C] metacherts and metabasalts was reported. This demonstrates the crucial effect of the coexisting sodium amphibole (Namp), which effectively reduces X_Jd in association with Na pyroxene. Therefore, Al was concentrated in sodium amphibole through the Al/Fe³⁺ distribution between the two minerals. Furthermore, the extent of the effect systematically differed according to the mode of growth, i.e., topotaxial replacement in metabasalts as K_D = (Al/Fe³⁺)^Namp/(Al/Fe³⁺)^Napx ∼ 4 and nucleation and growth in metacherts as K_D ∼ 1, rather than the bulk rock chemistry or oxidation state. Finally, the documented coexistence of Na pyroxene and sodium amphibole was examined to confirm the kinetic aspect of the effect, which diminished in higher grades, where more ideal behavior of cations is expected.

Abbreviations for mineral names are from the IMA-CNMNC approved symbols (Warr, 2021), except for Napx, Namp for Na pyroxene and sodium amphibole, respectively.

GEOLOGICAL SETTING

On Kyushu Island, Japan, the Kurosegawa Belt lies south of the Usuki-Yatsushiro Tectonic Line and is exposed as a narrow area (<20 km width, Fig. 1a). The Yatsushiro area occupies the western region of the Kurosegawa Belt, which is characterized by a serpentinite mélange, including a Jurassic accretionary complex (Shimotake subunit), Permian-Cretaceous sedimentary basin deposits and pre-Jurassic rocks of various origins and ages (Saito et al., 2010). They include mafic metamorphic rocks of prehnite-pumpellyite facies accompanied by Late Palaeozoic limestone blocks (Tobiishi subunit) and high-P/low-T mafic and siliceous metamorphic rocks of lawsonite blueschist subfacies (Hakoishi subunit) (Fig. 1b). White mica K-Ar ages extracted from metamorphic rocks have been reported as 180 Ma for the Tobiishi subunit and 240-290 Ma for the Hakoishi subunit (Sato et al., 2014), and they can be correlated with the Suo and Renge belts, respectively, in the Inner Zone of SW Japan (Tsujimori and Itaya, 1999; Fig. 1a).

Figure 1. (a) Major high-P/low-T type metamorphic belts in SW Japan and the position of the Yatsushiro area; (b) Geological map of the Yatsushiro area in the Kurosegawa belt (slightly modified from Saito et al., 2010). The Hakoishi subunit is mainly composed of metabasaltic and metasiliceous rocks that suffered lawsonite blueschist subfacies. Westward upgrading (from Zone 1 to Zone 2) is envisaged by Sato et al. (2016) based on the regional variation of low-variant mineral assemblages in metabasaltic rocks.

This study focused on the metamorphic rocks within the Hakoishi subunit. Its peak metamorphic condition is estimated as 200-300 °C and 0.45-0.80 GPa (Sato et al., 2016). The meta-mafic rocks (mentioned as ‘metabasalts’ hereafter) are mainly of basaltic lava or hyaloclastite origin and are characterized as massive or weakly foliated fabric, where relict augite or precursor igneous structures are preserved.

Sato et al. (2016) advocated for westward upgrading based on the spatial distribution of low-variance mineral assemblages in metabasalts. This represents the transition from pumpellyite blueschist (Zone 1) to lawsonite blueschist (Zone 2) subfacies, which is controlled by the reaction (1):

  

\begin{equation} \text{Napx} + \text{Pmp} + \text{Chl} + \text{H$_{2}$O} = \text{Lws} + \text{Namp} \end{equation} (1).

Although the mineral assemblages systematically indicate westward upgrading, the apparent maximum of X_Jd was observed in the middle of the subunit, which will be discussed later.

SAMPLE DESCRIPTION

The siliceous rocks (‘metacherts’) form a series of intercalation with metabasalts and typically display layered or slumping structures with a few centimeters to meter thickness. The matrix is mostly composed of granular quartz (<500 µm in diameter). The major metamorphic minerals in the metacherts [lawsonite, pumpellyite-(Fe³⁺), chlorite, Na pyroxene and sodium amphibole] were also common in the metabasalts. However, the pumpellyites are significantly richer in Fe³⁺ and Sr-bearing epidote (up to 9.44 wt% SrO) occurs in the metacherts (Table 1). As accessory minerals, apatite and titanite occur in most samples. The grains are typically arranged parallel to the layered structures. Certain metacherts are rich in Mn, and various Mn-bearing minerals have been reported (howieite, Ibuki et al., 2008; okhotskite, Yabuta and Hirajima, 2020).

Table 1. Mineral assemblages in the studied metachert samples (●, common; ○, rare)

	Qz	Ab	Napx	Namp	Lws	Pmp	Ep	Phg	Chl	Hem	Ap	Ttn	Others
OD410	●	●	●		●	○	○	●		●	●	●
KY1830a	●		●		●			●	●	●		●	sursassite
OD65	●		●				●	●			●	●
OD70	●		●		●			●	○	●	●	●
KY1815a	●		●	●			○			●	○		howieite, sulphide
KY1822a	●	●	●		●			●		●	●	●
KY1819	●		●		●			●		●	●	○	tourmaline
KY1811	●			●				○	●	●	●	●	piemontite, sursassite, calcite
OT10Q	●	●	●								○		piemontite, braunite
OT10R	●	●	●	●						●			piemontite, okhotskite
OT19	●	●	●		●			●	○	●			tourmaline, braunite
OT19BV	●	●	●							●			braunite
OD311	●		●	●						○	●		piemontite, calcite
OD209	●		●	●	●	●	●		●		●	●	baryte
OT27A-b	●			●		●	●	●			●	●	chalcopyrite, baryte
OD40	●		●	●								●	howieite, sulphide
OD44	●	●	●				○	●	●	●	●	●	calcite

Qz, quartz; Ab, albite; Napx, Na pyroxene; Namp, Na amphibole; Lws, lawsonite; Pmp, pumpellyite; Ep, epidote; Phg, phengite; Chl, chlorite; Hem, hematite; Ap, apatite; Ttn, titanite.

Despite common post-peak veins (quartz, albite, and calcite), both the metabasalts and metacherts were scarcely affected by greenschist facies overprinting during the exhumation stage, represented by the development of actinolite or epidote. Fujimoto et al. (2010) and Sato et al. (2016) provide detailed petrography and mineralogy of the metabasalts.

Na pyroxene occurs commonly both in metabasalts and metacherts but displays contrasting modes of occurrence. In metabasalts, igneous augite is typically replaced along fringes or cracks (Sato et al., 2016, Figs. 2 and 3). It rarely occurs as fillings of albite veins, apart from augite grains, in the metabasalt samples (OD112; Fig. 2a). It formed an interfingered structure with acicular sodium amphibole grains coexisting with lawsonite. In considering elemental partitioning, analyses on the outermost rim of neighboring mineral grains (<100 µm apart) are paired.

Figure 2. The mode of occurrence of Na pyroxenes and associated phases in the Hakoishi subunit: (a) Columnar Napx and Namp grains developed in a Lws/Qz-bearing Ab vein in a metabasalt OD112. A back scattered electron (BSE) image; (b) Folded Napx veins (green color) and later stage Qz veins (black, white, or gray color) in a polished slab surface of a metachert KY1830a; (c) Napx developed in the monomineralic vein and its chemical modification caused by the cross-cut by later stage Qz vein in KY1830a (BSE image); (d) Needle-shaped Napx grains coexisting with Pmt, Br, Ab, and Qz in metachert OT10Q. A photomicrograph under plane polarized light (PPL); (e) Napx aggregate in metachert OD65, in which Namp is not found (BSE image); (f) Napx and fan-shaped Lws developed in a Qz vein of metachert KY1819 (BSE image); (g) Napx and Namp aggregate in metachert OT10R (BSE image); (h) Twin-like grains of Napx and Namp in metachert OD311 (BSE image).

Figure 3. Compositions of Napx on the Jd-Aeg-Di diagram (a) from metachert and (c) metabasalt samples. Longitude (i.e., in the EW trend of the study area) variation of X_Jd in (b) metachert and (d) metabasalt samples. Border colors and fills of symbols and columns represent associated minerals; consult the legends in the figure.

In metacherts, Na pyroxene rarely formed monomineralic folded veins and light green color (Fig. 2b). It commonly occurs as acicular or columnar grains (<100 µm length, Figs. 2c-2f) in the quartz matrix. Textually, it is closely associated with sodium amphibole, forming bi-mineral layers less than 50 µm thick (Fig. 2g); in addition, parallel twin-like composite grains are occasionally observed (Fig. 2h).

While chemical compositions may vary within a thin section, individual grains are largely homogeneous. No textural evidence, such as zoning or core-rim relationships, suggests a multi-stage growth history, except for later-stage growth associated with cross-cutting quartz veins, which are easily distinguishable both chemically and texturally (Figs. 2b and 2c). Consequently, most Na pyroxene grains are interpreted to have formed during the peak metamorphic event. The paragenetic relationships between Na pyroxene and adjacent minerals are thus considered apparent.

ANALYTICAL METHODS

Compositions of minerals were determined by electron microprobe analyses (EPMA), using a Hitachi S3500H scanning electron microscope with an energy-dispersive X-ray analyser (EDAX^®) at the Department of Geology and Mineralogy, Kyoto University. The acceleration voltage was 20 kV, the beam current was 500 pA, and the spot size was 5 µm. Natural and synthetic materials were used as standards, and a ZAF correction scheme was applied to process the X-ray intensity data.

The discussion of cation partitioning may be significantly influenced by the accuracy of Fe³⁺/Fe_total estimation. To evaluate uncertainties arising from different estimation methods, two approaches were applied to Na pyroxene and sodium amphibole. For Na pyroxene, the methods of Droop (1987) and Morimoto et al. (1988) were employed, assuming six oxygen atoms, four total cations, and Na = ^VIAl + Fe³⁺. For sodium amphibole, the methods of Droop (1987) and Matsumoto and Hirajima (2005) were used, assuming 23 oxygen atoms, Σ(Si, Ti, Al, Cr, Fe, Mn, Mg) = 13, or ^BNa (Na in the B site) = ^VIAl + Fe³⁺. For chlorite, total Fe is regarded as divalent.

RESULTS

Compositions of Na pyroxene and sodium amphibole

Many of the analyzed Na pyroxenes were poor in Ca and, thus, belonged to the Aeg-Jd series. The Ca content was lower [<0.25 per formula unit (p.f.u.); Fig. 3a; Table 2] in metacherts than in metabasalts (<0.50; Fig. 3c). X_Jd was variable, particularly in metacherts, ranging from 0.13 to 0.60 within a single hand-specimen (KY1830a). While Ca content varies significantly among samples, it remains relatively constant within individual samples, particularly in metabasalts. Consequently, a negative correlation exists between X_Jd and Ca content across different samples (lower X_Jd corresponds to higher Ca content), as suggested by Endo and Wallis (2017) in the Mikabu-Northern Chichibu accretionary wedge, central Shikoku, whereas the correlation is weak within a single sample. Several Na pyroxenes suffered from later-stage fluid infiltration; however, such alterations could be discerned both textually and chemically. For instance, a mono-mineral vein mainly composed of Na pyroxene with X_Jd ∼ 0.4 was cut by a later-stage quartz vein, and an aegirine-richer part with X_Jd < 0.1 (Fig. 3a) developed at the contact (Fig. 2c).

Table 2. Representative compositions of Na pyroxene

#	KY1819	OT10Q	KY1815a	OD311	KY1830a		KHK31***
Coexistence		Ab	Namp	Namp Ab	Vein	Vein**	Namp Ab
Oxide wt%
SiO2	54.81	55.84	52.73	53.65	54.64	53.37	54.06
TiO2	0.11	0.32	0.25	0.24	n.d.	n.d.	0.29
Al2O3	8.26	11.69	2.71	4.68	7.78	2.25	5.83
Cr2O3	n.d.	n.d.	n.d.	n.d.	n.d.	n.d.	0.03
Fe2O3*	20.78	14.81	27.55	19.93	19.22	24.46	21.27
FeO*	0.00	0.00	1.62	0.94	0.45	0.32	0.00
MnO	0.10	0.40	0.60	1.54	0.49	1.51	0.16
MgO	1.19	1.80	0.37	2.98	1.61	2.62	2.35
CaO	1.43	1.51	1.73	4.89	3.30	4.87	3.96
Na2O	13.15	13.66	12.34	10.58	12.17	10.84	11.96
K2O	0.03	n.d.	n.d.	n.d.	0.03	0.03	0.02
Total	99.86	100.03	99.90	99.43	99.70	100.27	99.93
Cations for O = 6
Si	2.005	1.999	2.003	2.006	2.008	2.007	1.997
Ti	0.003	0.009	0.007	0.007	-	-	0.008
Al	0.356	0.493	0.121	0.206	0.337	0.100	0.254
Cr3+	-	-	-	-	-	-	0.001
Fe3+	0.572	0.399	0.788	0.561	0.532	0.692	0.592
Fe2+	0.000	0.000	0.051	0.029	0.014	0.010	0.000
Mn2+	0.003	0.012	0.019	0.049	0.015	0.048	0.005
Mg	0.065	0.096	0.021	0.166	0.088	0.147	0.129
Ca	0.056	0.058	0.070	0.196	0.130	0.196	0.157
Na	0.933	0.948	0.909	0.767	0.867	0.791	0.857
K	0.001	-	-	-	0.001	0.001	0.001
Total	3.995	4.014	3.990	3.987	3.992	3.993	4.001

^* Recalculated, see details in text.

^** Qz vein contact.

^*** Horokanai Ep blueschist.

Blueish sodium amphibole replaces igneous relicts, i.e., augite and hornblende, or it occurs as needle-shaped grains in metabasalts throughout the study area (Sato et al., 2016, Fig. 3f). The Al/(Al + Fe³⁺) (= Y_Amp) ratio of sodium amphiboles associated with lawsonite and pumpellyite increased westward, from Y_Amp = 0.2-0.3 in Zone 1, through Y_Amp = 0.4-0.7 in the transition zone, to Y_Amp = 0.8-0.9 in Zone 2. This was accompanied with a slight increase in Mg# [= Mg/(Mg + Fe²⁺)] from Zone 1 (Mg# ∼ 0.4) to Zone 2 (Mg# ∼ 0.6) (Sato et al., 2016, Fig. 5). This chemical variation in sodium amphibole from riebeckite (Rbk) or magnesioriebeckite (Mrbk) to glaucophane (Gln) also supported westward upgrading.

In the metacherts, sodium amphibole also occurred as acicular grains similar in size to Na pyroxene (Figs. 2g and 2h), however its distribution was limited to the western half of the study area, i.e., the transition zone and Zone 2 of Sato et al. (2016) (Fig. 1b). Almost all the analyzed sodium amphiboles were classified as Rbk or Mrbk (Ca < 0.2 p.f.u.; Fig. 4a; Table 3), and Gln was not found in the metacherts. The estimated Fe³⁺ content was significantly higher in the metacherts than in the metabasalts. The Mg# ranged from 0.9 to 1.0 in a braunite- and piedmontite-bearing sample (OT10R). In contrast, Ibuki et al. (2010) reported ferro-glaucophane (Mg# = 0.0-0.3) from the Hakoishi metachert. This difference was attributed to the oxidation state (cf. Okay, 1980a; Mottana, 1986; Ubukawa et al., 2007); Mn³⁺-bearing minerals were common in the former, whereas Mn²⁺-bearing howieite occurred in the latter.

Figure 4. Compositions of Namp (square) in (a) metachert and (b) metabasalt samples from the western half of the Transition Zone of Sato et al. (2016), compared with relevant Napx (circle). The composition ranges in representative samples are shown as 3σ ellipses. The Al contents in both Namp and Napx are restricted to lower and narrower ranges when coexisting with each other (shaded ellipse). Al is concentrated in coexisting Namp in metabasalt whereas it is equally distributed in metachert.

Table 3. Representative compositions of sodium amphibole

#	OD209	KY1822	KY1815a	OD311	KHK31**
Coexistence		Ab	Napx	Napx	Napx Ab
Oxide wt%
SiO2	56.32	55.41	53.81	56.69	54.59
TiO2	0.10	n.d.	n.d.	n.d.	0.06
Al2O3	3.31	1.70	2.21	7.47	3.63
Fe2O3*	12.49	15.76	13.52	6.76	13.74
FeO*	7.20	3.83	14.17	12.17	11.47
MnO	1.07	2.92	2.43	n.d.	0.27
MgO	9.52	9.79	4.65	7.61	6.41
CaO	0.23	0.57	0.72	0.73	0.72
Na2O	7.34	7.13	6.57	7.09	7.28
K2O	n.d.	n.d.	0.07	n.d.	0.04
Total	97.58	97.11	98.15	98.52	98.20
Cations for O = 23
Si	8.011	7.995	7.994	7.980	7.926
Ti	0.011	-	-	-	0.006
Al	0.555	0.289	0.387	1.239	0.621
Fe3+	1.470	1.711	1.511	0.716	1.501
Fe2+	0.724	0.462	1.761	1.433	1.392
Mn2+	0.129	0.357	0.306	-	0.033
Mg	2.019	2.106	1.030	1.597	1.388
Ca	0.035	0.088	0.115	0.110	0.112
Na	2.024	1.995	1.892	1.935	2.048
K	-	-	0.013	-	0.008
Total	14.978	15.003	15.010	15.010	15.035

^* Recalculated, see details in text.

^** Horokanai Ep blueschist.

Element partitioning between Na pyroxene and coexisting minerals

Figure 5a illustrates the Al/Fe³⁺ ratios and distribution coefficient [K_D = (Al/Fe³⁺)^Namp/(Al/Fe³⁺)^Napx] of coexisting Na pyroxenes and sodium amphiboles from the Hakoishi metacherts and metabasalts. K_D equals 1.0 on average (0.5-2.0) in the metacherts and 4.0 (1.0-8.0) in the metabasalts. One exceptional metabasalt sample (OD112) contained pairs that displayed an equal distribution (K_D ∼ 1.0). In the samples, Na pyroxene and sodium amphibole occurred in the veinlets, suggesting simultaneous nucleation and growth from fluids. This is unlike other metabasalts (Na pyroxene replaces igneous augite) and similar to metacherts.

Figure 5. Al/Fe³⁺ partition diagram of coexisting Napx and Namp: (a) Lws-BS subfacies of the Hakoishi subunit; (b) Lws/Ep-BS subfacies of Horokanai, the Kamuikotan belt, Hokkaido, Japan; (c) BS and eclogite facies of the Franciscan Complex, California, USA; (d) blueschist and eclogite in other areas (see reference for details).

The Fe²⁺/Mg distribution between Na pyroxene and chlorite (Fig. 6a) and between sodium amphibole and chlorite (Fig. 6b) in metabasalts of the Hakoishi subunit are examined. Despite intra-sample variability in the Fe²⁺/Mg ratio of Na pyroxene, likely due to low Fe²⁺ and Mg contents, Na pyroxene apparently exhibits a slight preference for Mg. In contrast, the nearly equal partitioning is observed in Fe²⁺/Mg ratios of chlorite and sodium amphibole.

Figure 6. Fe²⁺/Mg partition diagram of (a) Napx and Chl and (b) Namp and Chl in the metabasalt samples of the Hakoishi subunit and relevant grade Lws-bearing metabasites.

Figures 5 and 6 present estimation errors (error bars) for microprobe analysis and two estimation methods discussed in the previous chapter. Droop’s method is inappropriate for high-grade sodium amphiboles with significant ^ANa contents, and therefore, was not included in Figures 5c and 5d.

DISCUSSION

General controlling factors

Generally, X_Jd is controlled by metamorphic pressure through the equilibrium reaction Ab = Jd + Qz. In the Hakoishi rocks, the abundance of albite and quartz suggests that the albite association is a primary factor determining the upper limit of X_Jd. Maximum X_Jd values in each sample are comparable to those in relevant grade metabasalts and metacherts, and exhibit a general westward increase (Figs. 3b and 3d). Local variations in X_Jd within individual samples, in which X_Jd should be constant, may be attributed to differences in Al availability or effective variance.

While albite veins are abundant in metabasalts, they are less common in metacherts, and some metacherts even lack matrix albite, resulting in lower X_Jd. Comparing isophysical samples, the range of X_Jd is narrower in more-phased (lower variance) sample (KY1822; Qz + Ab + Lws + Napx + Hem) than in less-phased one (KY1819; Qz + Lws + Napx + Hem; Fig. 3b).

Impact of associated phases and igneous relics on X_Jd

However, a significant decrease in X_Jd in sodium amphibole-bearing rocks compared to that in sodium amphibole-free rocks was observed in both metabasalts and metacherts (Fig. 3). In the equivalent grade, X_Jd was restricted to a lower and narrower range when sodium amphibole coexisted. However, westward upgrading in X_Jd was observed in sodium amphibole-bearing and sodium amphibole-free samples, respectively (Fig. 3). Similarly, the Al content in the sodium amphibole was slightly lower when Na pyroxene coexisted in both lithologies (Fig. 4).

Conversely, the assemblages of other metamorphic minerals, such as pumpellyite, lawsonite and epidote, or their Al/Fe³⁺ ratios, displayed no clear correlation with X_Jd. In addition, the oxidation states did not appear to be a prime factor owing to the Al contents in Na pyroxene and sodium amphibole being similar in the highly oxidized (OT10R, piemontite-bearing) and moderately oxidized samples (KY1815, howieite-bearing). This implies that the Al/Fe³⁺ partitioning between Na pyroxene and sodium amphibole primarily determines X_Jd.

Regarding Na pyroxene, the Hakoishi metacherts were characterized by the absence of igneous augite. This indicated that the Na pyroxene developed through nucleation and growth, in contrast to the replacement of igneous pyroxene in metabasalts. The higher content and westward decrease of Ca of Na pyroxenes in the metabasalts may be regarded as snapshots of the CaMg-NaAl substitution from the igneous precursor to various extents (Figs. 3a and 3c).

Al/Fe³⁺ partitioning between Na pyroxene and sodium amphibole

The effect of the sodium amphibole coexistence on X_Jd is described as the result of competition between the two minerals for a limited amount of Al in the local effective bulk. The loss of one or two may have resulted in a lower Al content under coexistence. This rivalry is characterized by an Al/Fe³⁺ partition.

Onuki and Ernst (1969) introduced it into metasiliceous and metaclastic rocks of the blueschist and eclogite facies to obtain K_D = 1.2. This indicates that the ion-to-ion exchange between the M2 site of sodium amphibole and the M1 site of Na pyroxene has only weak cation-preference. Okay (1980b) reported K_D ∼ 2.1 from low-grade mafic and siliceous rocks (Tavşanlı Region, Northwest Turkey), with Al rather concentrated in sodium amphibole.

By definition (equation 2 in Onuki and Ernst, 1969), the distribution coefficient is determined by the equilibrium temperature. However, two different trends in K_D were recognized within a single metamorphic unit. In the topotaxial growth texture, Al was concentrated in sodium amphibole, whereas nearly equal partitioning were observed in the modes where nucleation and growth were dominant. This demonstrates that the mineral growth process, rather than lithotypes or bulk composition, determines elemental partitioning.

These trends have also been observed in other blueschist regions. For instance, the Horokanai lawsonite/epidote-bearing blueschists (zone II/III of Shibakusa, 1989) in the Kamuikotan belt, Hokkaido, Japan, suffered a metamorphic temperature slightly higher (280-300 °C; Naemura et al., 2022) than that of the Hakoishi lawsonite blueschist. Despite the comparable peak metamorphic temperatures, K_D was mostly equal to 1 (Fig. 5b). This can be attributed to the different rock fabrics. The Horokanai schists display penetrative schistosity and an uncommon occurrence of relict igneous pyroxene.

Furthermore, the documented coexistence of Na pyroxene (aegirine, jadeite, and omphacite of Morimoto et al., 1988) and sodium amphibole (^BNa >1.75, Si >7.5 p.f.u.) was investigated. To exclude errors in the calculation method, Fe³⁺/Fe_tot was recalculated, as mentioned in the text, regardless of the estimates in each document.

Figure 5c compares the K_D in the metamorphic rocks of the blueschist and eclogite facies from the Franciscan Complex. Significantly higher values (K_D ∼ 4) were calculated for rocks in which Na pyroxene replaced relict augite (Brown and Ghent, 1983; Liou and Maruyama, 1987; Maruyama and Liou, 1987; Maruyama and Liou, 1988). Other blueschist facies rocks (Brown and Bradshaw, 1979; Patrick and Day, 1989) and eclogites (other references in Fig. 5c) display K_D values close to 1.

The coexistence with other reports is shown in Figure 5d. Similarly, pairs of blueschist facies overgrown igneous augite or hornblende (Queyras, western Alps, Pognante and Kienast, 1987; Mikabu belt in Kanto mountains, Japan, Hirajima, 1985) show higher K_D (∼ 4), compared to other rocks without such texture (∼ 1) (chlorite zone of the Sanbagawa belt in Shikoku, Sakaguchi and Ishizuka, 2008; Mn nodule from garnet zone in the Sanbagawa belt in Kanto Mountains, Hirajima, 1989; garnet-bearing epidote blueschist in Osayama mélange in Chugoku Mountains, Tsujimori and Liou, 2005). Eclogite facies rocks from Oman (El-Shazly et al., 1990; Searle et al., 1994) and the western Alps (Koons, 1986; Matsumoto and Hirajima, 2005) yielded a K_D of ∼ 1.

Kinetic control on Na pyroxene compositions

In summary, the X_Jd of Na pyroxene in a low-grade environment is restricted by Al availability in the effective bulk compositions. When sodium amphibole is present, competition lowers the Al content of both minerals. However, the outcome of this competition is influenced by growth mechanisms. Topotaxial growth leads to Al enrichment in sodium amphibole, while nucleation and growth processes result in more equitable Al distribution.

Thus, there is a kinetic preference for these elements. Morata et al. (1994) explained the different extents of topotaxial replacement based on the structural affinity between igneous and metamorphic pyroxenes. Ti-rich augite in transitional-alkaline metabasite is structurally closer to Na pyroxene than Ti-poor augite in tholeiitic metabasites and is extensively replaced. Analogously, the replacement of the metamorphic phase of Al-endmembers (Jd, Gln) and Fe³⁺-endmembers (Aeg, Mrbk) with igneous ones (Aug, Mhbl) is considered.

For simplicity, the unit cell parameters for the relevant members of the pyroxene and amphibole groups (data from Anthony et al., 1995) are considered. Comparing the Al- and Fe³⁺- endmembers, Aeg (unit cell volume: 429.1 ų) is closer to Aug (432.5 ų) in relation to Jd (401.2 ų); conversely, Mrbk (895.9 ų) and Gln (880.7 ų) are structurally close to Mhbl (950.1 ų) to a similar extent. A larger departure from the unit cells of the igneous phase translates into a larger distortion produced by topotaxial growth. Therefore, elemental partitioning occurred in the replacing fabric, whereas such constraints were not present during nucleation and growth. This could explain the strong Fe³⁺ preference of Na pyroxene for replacing igneous augite as opposed to the coexisting amphibole.

Fe²⁺-Mg partitioning between Na pyroxene and coexisting minerals in metabasalts of the Hakoishi subunit reveals a slight Mg-enrichment of Na pyroxene relative to coexisting chlorite (Fig. 6a), contrasting with the near-equal partitioning between chlorite and sodium amphibole (Fig. 6b). This trend is consistent with observations from lawsonite-bearing metabasites in Cazadero, California (Maruyama and Liou, 1988) and in the Chugoku Mountains (Tsujimori and Liou, 2007). Given the Mg-rich composition of the relic igneous pyroxenes, this implies a non-equilibrium state in the Fe²⁺-Mg partition in Na pyroxene replacing igneous relic phases.

All these parameters are in the standard state; thus, these kinetic effects are expected to be less pronounced at higher temperatures, where better equilibration is expected. Given the observed equal element partition (K_D ∼ 1) seen in documented higher-grade rocks, the Al preference of replacing phases should have a high temperature limit near 300 °C.

Below this temperature, Na pyroxene developed by topotaxial growth may represent a metastable state by sluggish kinetics. In the quest for equilibrium in a clinopyroxene system, such a texture should be excluded. Consequently, the Na pyroxenes in the metachert samples provide promising examples for constructing a ternary phase diagram of Na pyroxene as they are virtually free of the influence of igneous pyroxenes.

CONCLUSIONS

• •    In the LBS subfacies of the Hakoishi subunit, even under quartz- and albite-saturated environment, X_Jd was crucially affected (lowered) by the coexistence of sodium amphibole rather than by metamorphic pressure.
• •    The Al/Fe³⁺ partitioning between the coexisting Na pyroxene and sodium amphibole varied systematically between the metacherts (K_D ∼ 1) and metabasalts (K_D ∼ 4), reflecting their mode of growth (topotaxial replacement and nucleation and growth, respectively).
• •    Whereas Na pyroxenes in metabasalts may represent a non-equilibrium state arising from kinetics of replacement of igneous phases, ones in metacherts, which are hardly affected by igneous phases, may be a better proxy of the clinopyroxene system equilibrium state at low temperature environments (<300 °C).

ACKNOWLEDGMENTS

We express our sincere gratitude to the reviewers, Dr. S. Endo and Prof. T. Tsujimori, for their constructive comments. We also thank Prof. S. Kumar for his helpful handling of this manuscript. Finally, we gratefully acknowledge the work of past members of our laboratory. This work was partly supported by Grant-in-Aid for Scientific Research (No. JP19H01999) and Bai Xian Asia Institute.

REFERENCES

• Anthony, J.W., Bideaux, R.A., Bladh, K.W. and Nichols, M.C. (1995) Handbook of Mineralogy, Volume II, Silica, Silicates. Mineral Data Publishing, Tucson, Arizona.
•  Brown,  E.H. and  Bradshaw,  J.Y. (1979) Phase relations of pyroxene and amphibole in greenstone, blueschist and eclogite of the Franciscan Complex, California. Contributions to Mineralogy and Petrology, 71, 67-83.
•  Brown,  E.H. and  Ghent,  E.D. (1983) Mineralogy and phase relations in the blueschist facies of the Black Butte and Ball Rock areas, Northern California Coast Ranges. American Mineralogist, 68, 365-372.
•  Carpenter,  M.A. (1979a) Contrasting properties and behaviour of antiphase domains in pyroxenes. Physics and Chemistry of Minerals, 5, 119-131.
•  Carpenter,  M.A. (1979b) Omphacites from Greece, Turkey, and Guatemala; composition limits of cation ordering. American Mineralogist, 64, 102-108.
•  Carpenter,  M.A. (1980) Mechanisms of exsolution in sodic pyroxenes. Contributions to Mineralogy and Petrology, 71, 289-300.
•  Carpenter,  M.A. (1982) Time-temperature-transformation (TTT) analysis of cation disordering in omphacite. Contributions to Mineralogy and Petrology, 78, 433-440.
•  Droop,  G.T.R. (1987) A general equation for estimating Fe³⁺ concentrations in ferromagnesian silicates and oxides from microprobe analyses, using stoichiometric criteria. Mineralogical Magazine, 51, 431-435.
•  El-Shazly,  A.E.,  Coleman,  R.G. and  Liou,  J.G. (1990) Eclogites and Blueschists from Northeastern Oman: Petrology and P-T Evolution. Journal of Petrology, 31, 629-666.
•  Endo,  S. and  Wallis,  S.R. (2017) Structural architecture and low-grade metamorphism of the Mikabu-Northern Chichibu accretionary wedge, SW Japan. Journal of Metamorphic Geology, 35, 695-716.
•  Essene,  E.J. and  Fyfe,  W.S. (1967) Omphacite in Californian metamorphic rocks. Contributions to Mineralogy and Petrology, 15, 1-23.
•  Fujimoto,  Y.,  Kono,  Y.,  Hirajima,  T.,  Kanagawa,  K., et al (2010) P-wave velocity and anisotropy of lawsonite and epidote blueschists: Constraints on water transportation along subducting oceanic crust. Physics of the Earth and Planetary Interiors, 183, 219-228.
•  Green,  E.,  Holland,  T. and  Powell,  R. (2007) An order-disorder model for omphacitic pyroxenes in the system jadeite-diopside-hedenbergite-acmite, with applications to eclogitic rocks. American Mineralogist, 92, 1181-1189.
• Hirajima, T. (1985) Petrological Study of the Sanbagawa Metamorphic Belt in the Kanto Mountains, Central Japan. Ph.D.Thesis, Kyoto University, Japan, 67-109.
•  Hirajima,  T. (1989) Mineralogy of a Fe-Mn rich nodule from the Sanbagawa schist in the Kanto Mountains, Japan. The Journal of the Geological Society of Japan, 95, 817-833.
•  Holland,  T.J.B. (1980) The reaction albite = jadeite + quartz determined experimentally in the range 600-1200 °C. American Mineralogist, 65, 129-134.
•  Holland,  T.J.B. (1983) The experimental determination of activities in disordered and short-range ordered jadeitic pyroxenes. Contributions to Mineralogy and Petrology, 82, 214-220.
•  Holland,  T. and  Powell,  R. (1996a) Thermodynamics of order-disorder in minerals; I, Symmetric formalism applied to minerals of fixed composition. American Mineralogist, 81, 1413-1424.
•  Holland,  T. and  Powell,  R. (1996b) Thermodynamics of order-disorder in minerals: II, Symmetric formalism applied to solid solutions. American Mineralogist, 81, 1425-1437.
•  Ibuki,  M.,  Fujimoto,  Y.,  Takaya,  M.,  Miyake,  A. and  Hirajima,  T. (2008) Howieite in meta-manganese siliceous rocks of Kurosegawa belt, western Kyushu, Japan. Journal of Mineralogical and Petrological Sciences, 103, 365-370.
•  Ibuki,  M.,  Ohi,  S.,  Tsuchiyama,  A. and  Hirajima,  T. (2010) Analysis of Mn-bearing lawsonite occurring in meta-siliceous rocks in Hakoishi serpentinite mélange of Kurosegawa Belt, Central Kyushu, Japan. Journal of Mineralogical and Petrological Sciences, 105, 340-345.
•  Koons,  P.O. (1986) Relative geobarometry from high-pressure rocks of quartzofeldspathic composition from the Sesia Zone, Western Alps, Italy. Contributions to Mineralogy and Petrology, 93, 322-334.
•  Liou,  J.G. and  Maruyama,  S. (1987) Parageneses and compositions of amphiboles from Franciscan jadeite-glaucophane type facies series metabasites at Cazadero, California. Journal of Metamorphic Geology, 5, 371-395.
•  Manzotti,  P.,  Ballèvre,  M.,  Pitra,  P.,  Putlitz,  B., et al (2020) The Growth of Sodic Amphibole at the Greenschist- to Blueschist-facies Transition (Dent Blanche, Western Alps): Bulk-rock Chemical Control and Thermodynamic Modelling. Journal of Petrology, 61, egaa044.
•  Maruyama,  S. and  Liou,  J.G. (1987) Clinopyroxene-a mineral telescoped through the processes of blueschist facies metamorphism. Journal of Metamorphic Geology, 5, 529-552.
•  Maruyama,  S. and  Liou,  J.G. (1988) Petrology of Franciscan Metabasites Along the Jadeite-Glaucophane Type Facies Series, Cazadero, California. Journal of Petrology, 29, 1-37.
•  Matsumoto,  K. and  Hirajima,  T. (2005) The coexistence of jadeite and omphacite in an eclogite-facies metaquartz diorite from the southern Sesia Zone, Western Alps, Italy. Journal of Mineralogical and Petrological Sciences, 100, 70-84.
•  Miyashiro,  A. and  Banno,  S. (1958) Nature of glaucophanitic metamorphism. American Journal of Science, 256, 97-110.
•  Morata,  D.,  Aguirre,  L. and  Puga,  E. (1994) Na-metamorphic pyroxenes in low-grade metabasites from the External Zones of the Betic Cordilleras (southern Spain): influence of rock chemical composition on their formation. Revista Geológica de Chile, 21, 269-283.
•  Morimoto,  N.,  Fabries,  J.,  Ferguson,  A.K.,  Ginzburg,  I.V., et al (1988) Nomenclature of pyroxenes. American Mineralogist, 73, 1123-1133.
• Mottana, A. (1986) Blueschist-facies metamorphism of manganiferous cherts: A review of the alpine occurrences. In Blueschist and Eclogites (Evans, B.W. and Brown, E.H. Eds.). pp. 423, Memoirs, 164, Geological Society of America, Boulder, Colorado, 267-300.
• Naemura, K., Nishida, Y., Horie, Y., Kinoshita, S., et al. (2022) An Application of the Raman carbonaceous material geothermometry to pelitic schists in the Horokanai Area of the Kamuikotan metamorphic belt. Abstract of Annual Meeting of Japan Association of Mineralogical Society, Niigata, Japan, R8-13 (in Japanese).
•  Okay,  A.I. (1978) Sodic pyroxenes from metabasites in the eastern mediterranean. Contributions to Mineralogy and Petrology, 68, 7-11.
•  Okay,  A.I. (1980a) Sodic amphiboles as Oxygen fugacity indicators in metamorphism. The Journal of Geology, 88, 225-232.
•  Okay,  A.I. (1980b) Mineralogy, petrology, and phase relations of glaucophane-lawsonite zone blueschists from the Tavşanli Region, Northwest Turkey. Contributions to Mineralogy and Petrology, 72, 243-255.
•  Onuki,  H. and  Ernst,  W.G. (1969) Coexisting sodic amphiboles and sodic pyroxenes and blueschist facies metamorphic rocks. Special Paper - Mineralogical Society of America, 2, 241-249.
•  Patrick,  B.E. and  Day,  H.W. (1989) Controls on the first appearance of jadeitic pyroxene, northern Diablo Range, California. Journal of Metamorphic Geology, 7, 629-639.
•  Pognante,  U. and  Kienast,  J.R. (1987) Blueschist and Eclogite transformations in Fe-Ti Gabbros: A Case from the Western Alps Ophiolites. Journal of Petrology, 28, 271-292.
• Saito, M., Takarada, S., Toshimitsu, S., Mizuno, K., et al. (2010) Geological map of Japan 1:200,000 Yatsushiro and a part of Nomo Zaki. Geological Survey of Japan, AIST (in Japanese with English abstract).
•  Sakaguchi,  M. and  Ishizuka,  H. (2008) Subdivision of the Sanbagawa pumpellyite-actinolite facies region in central Shikoku, southwest Japan. Island Arc, 17, 305-321.
•  Sato,  E.,  Hirajima,  T.,  Kamimura,  K. and  Fujimoto,  Y. (2014) White mica K-Ar ages from lawsonite-blueschist facies Hakoishi sub-unit and from prehnite-pumpellyite facies Tobiishi sub-unit of the Kurosegawa belt, Kyushu, Japan. Journal of Mineralogical and Petrological Sciences, 109, 258-270.
•  Sato,  E.,  Hirajima,  T.,  Yoshida,  K.,  Kamimura,  K. and  Fujimoto,  Y. (2016) Phase relations of lawsonite blueschists and their role as a water-budget monitor: a case study from the Hakoishi sub-unit of the Kurosegawa belt, SW Japan. European Journal of Mineralogy, 28, 1029-1046.
•  Searle,  M.P.,  Waters,  D.J.,  Martin,  H.N. and  Rex,  D.C. (1994) Structure and metamorphism of blueschist-eclogite facies rocks from the northeastern Oman Mountains. Journal of the Geological Society, 151, 555-576.
•  Shibakusa,  H. (1989) Lawsonite-Pumpellyite-Epidote Stabilities in glaucophane schists in the Horokanai-Kamietanbetsu area of Kamuikotan zone, Hokkaido, Japan. Mineralogy and Petrology, 40, 241-256.
•  Shibakusa,  H. and  Maekawa,  H. (1997) Lawsonite-bearing eclogitic metabasites in the Cazadero area, northern California. Mineralogy and Petrology, 61, 163-180.
•  Tsujimori,  T. and  Itaya,  T. (1999) Blueschist-facies metamorphism during Paleozoic orogeny in southwestern Japan: Phengite K-Ar ages of blueschist-facies tectonic blocks in a serpentinite mélange beneath early Paleozoic Oeyama ophiolite. Island Arc, 8, 190-205.
•  Tsujimori,  T. and  Liou,  J.G. (2005) Eclogite-Facies Mineral Inclusions in Clinozoisite from Paleozoic Blueschist, Central Chugoku Mountains, Southwest Japan: Evidence of Regional Eclogite-Facies Metamorphism. International Geology Review, 47, 215-232.
•  Tsujimori,  T.,  Matsumoto,  K.,  Wakabayashi,  J. and  Liou,  J.G. (2006) Franciscan eclogite revisited: Reevaluation of the P-T evolution of tectonic blocks from Tiburon Peninsula, California, U.S.A. Mineralogy and Petrology, 88, 243-267.
•  Tsujimori,  T. and  Liou,  J.G. (2007) Significance of the Ca-Na Pyroxene-Lawsonite-Chlorite Assemblage in Blueschist-Facies Metabasalts: An Example from the Renge Metamorphic Rocks, Southwest Japan. International Geology Review, 49, 416-430.
•  Ubukawa,  T.,  Hatanaka,  A.,  Matsumoto,  K. and  Hirajima,  T. (2007) Pseudosection analysis for talc-Na pyroxene-bearing piemontite-quartz schist in the Sanbagawa Belt, Japan. Island Arc, 16, 553-574.
•  Warr,  L.N. (2021) IMA-CNMNC approved mineral symbols. Mineralogical Magazine, 85, 291-320.
•  Yabuta,  W. and  Hirajima,  T. (2020) A new occurrence of okhotskite in the Kurosegawa belt, Kyushu, Japan: the okhotskite + Mn-lawsonite assemblage as a potential high-pressure indicator. Journal of Mineralogical and Petrological Sciences, 115, 431-439.