disclaimer - seoul national...
TRANSCRIPT
저 시-비 리- 경 지 2.0 한민
는 아래 조건 르는 경 에 한하여 게
l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.
다 과 같 조건 라야 합니다:
l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.
l 저 터 허가를 면 러한 조건들 적 되지 않습니다.
저 에 른 리는 내 에 하여 향 지 않습니다.
것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.
Disclaimer
저 시. 하는 원저 를 시하여야 합니다.
비 리. 하는 저 물 리 목적 할 수 없습니다.
경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.
이학석사 학위논문
Influence of partial melting and
melt-mantle interaction on
PGE geochemistry of the Mariana
forearc peridotites
부분용융과 멜트-맨틀 상호작용이
마리아나 전호지역 맨틀암석의 백금족 원소
지구화학에 미치는 영향
2019년 2월
서울대학교 대학원 지구환경과학부
박 준 희
Influence of partial melting and
melt-mantle interaction on
PGE geochemistry of the Mariana
forearc peridotites
지도 교수 박 정 우
이 논문을 이학석사 학위논문으로 제출함
2019년 2월
서울대학교 대학원
지구환경과학부
박 준 희
박준희의 이학석사 학위논문을 인준함
2018년 12월
위 원 장 (인)
부위원장 (인)
위 원 (인)
i
Abstract
Mantle compositions can be changed mainly by magmatic processes
such as partial melting and melt-mantle interaction. Direct investigation of
mantle peridotite in the forearc is limited. In this study, forearc mantle
peridotites from the Mariana mud volcanoes (Conical, Asút Tesoru,
Fantangisña, Yinazao, and South Chamorro seamounts) by International
Ocean Discovery Program (IODP) 366, Ocean Drilling Project (ODP) 125
and 195 are examined in the petrographic and geochemical views in order to
investigate the petrogenesis and platinum group element (PGE)
geochemistry of the forearc mantle. The peridotites are divided by
harzburgite, dunite, and serpentinite (completely serpentinized peridotite).
By analyzing the abundance of major elements from whole rocks and
minerals and trace elements from minerals, it is demonstrated that the
peridotites from the Mariana forearc experienced high degrees of melting
with hydrous fluid derived from the slab (more than 20–25%), and some of
them were enriched by interaction with melts. The expected melt agents
reacted with the forearc mantle are suggested as boninitic melt and mid-
ocean ridge basalt (MORB)-like melt, considered as Forearc basalt (FAB)
melt.
It has been accepted that an abundance of PGE is little affected by
serpentinization process. Magmatic processes, however, including partial
melting and melt-mantle interaction can influence on PGE concentration of
ii
the forearc mantle. Peridotites distinguished by the residue of hydrous
melting and interaction with melts has distinct characteristics of PGE
concentrations. The residue of high degrees of hydrous melting shows
variable concentrations of PGE and palladium-like PGE (PPGE)/iridium-
like PGE (IPGE) fractionation compared to that of abyssal peridotites. This
result might be interpreted that PGE changes the host mineral from base-
metal sulfide to PGE alloys with incongruent sulfide melting after/near the
sulfide exhaustion. Whether the melt is saturated with sulfur is the important
factor in determining PGE concentrations in the mantle during melt-mantle
interaction. Whereas boninitic melts can escape the PGE from the mantle
with dissolving sulfide, MORB-like melts precipitate Cu(–Ni)-rich sulfide
which is more concentrated by PPGE than IPGE. Through this study, mantle
heterogeneity beneath the Mariana forearc can occur by hydrous melting
and melt-mantle interaction in terms of PGE.
Keyword : Mariana forearc mantle, mantle heterogeneity, hydrous partial melting,
melt-mantle interaction, platinum group elements
Student Number : 2016-28284
iii
Table of Contents
Abstract ....................................................................................................... ⅰ
Table of Contents ...................................................................................... ⅲ
List of Figures ............................................................................................ ⅴ
List of Tables ............................................................................................. ⅺ
1. Introduction ........................................................................................... 1
2. Geological Background ......................................................................... 4
2.1 Subduction system of the Mariana Arc ............................................. 4
2.2 Mud volcano in the Mariana forearc ................................................. 8
3. Sample selection and Analytical methods .......................................... 11
3.1 Sample selection .............................................................................. 11
3.2 Whole rock major element and sulfur analyses ............................... 12
3.3 Mineral major and trace element analyses by EPMA and LA-ICP-MS
................................................................................................................... 13
3.4 Whole rock PGE analyses ............................................................... 14
4. Results ................................................................................................... 16
4.1 Petrography ...................................................................................... 16
4.1.1 Harzburgite ............................................................................... 16
4.1.2 Dunite ....................................................................................... 18
4.1.3 Petrographic features of serpentinization ................................. 19
4.2 Whole rock major element chemistry .............................................. 26
4.3 Mineral chemisry ............................................................................. 31
4.3.1 Major element chemistry .......................................................... 31
4.3.2 Trace element chemistry ........................................................... 40
iv
4.4 Whole rock PGE and S chemistry ................................................... 47
4.5 Oxygen fugacity .............................................................................. 55
5. Discussion ............................................................................................. 57
5.1 Highly depleted features compared to the abyssal peridotite .......... 57
5.2 Melt-mantle interaction ................................................................... 61
5.2.1 Interaction with boninitic melt ................................................. 62
5.2.2 Interaction with MORB-like melt ............................................ 67
5.3 Effect of serpentinization on mantle peridotite ............................... 68
5.4 PGE geochemistry in the forearc mantle ......................................... 71
5.4.1 Influence of low-T alteration on PGE geochemistry ................ 71
5.4.2 Influence of hydrous melting on PGE geochemistry................ 73
5.4.3 Influence of melt-mantle interaction on PGE geochemistry .... 78
6. Conclusion ............................................................................................ 81
Appendix Figures ...................................................................................... 84
Appendix Tables ........................................................................................ 88
References ................................................................................................. 90
국문 초록 ............................................................................................... 104
감사의 글 ............................................................................................... 106
v
List of Figures
Figure 1. Geological maps of the studied area. (a) The Mariana arc located eastern
boundary of the Philippine oceanic plate. (b) Detailed bathymetric map of sampled
areas in the Mariana forearc. Five mud volcanoes (Conical, Asút Tesoru,
Fantangisña, Yinazao, and South Chamorro seamounts) are located along the
Mariana trench.
Figure 2. Schematic diagram showing the formation of mud volcanoes in the
Mariana forearc region modified from Fryer et al. (1999). ① Dehydration process
of the subducting slab. ② Hydrous fluid drenches mantle wedge, and mantle
wedge is serpentinized. It makes density contrasts between serpentinized peridotite
and surrounding mantle rocks. ③ Serpentinized peridotites and serpentine mud
rise along the normal faults in the overriding plate
Figure 3. Thin section photomicrograph of the peridotites from the Mariana
forearc. (A) Porphyroclastic texture of a harzburgite sample (195-1200A-7R-2/84–
87; XPL). Porphyroclast of Opx is mantled by neoblasts of Ol. (B) Protogranular
texture of a harzburgite sample (366-1498B-7R-4/88–93; XPL). Several mm of Ol
and Opx grain size are characteristics of this texture. (C) A dunite sample with
inclined Ol (366-1493B-9X-1/60–63; XPL). Euhedral spinels are included in
olivine matrix. Olivine grains are stretched as left diagonal patterns. Mineral
Abbreviations from Whitney & Evans (2010): Ol = olivine, Opx = orthopyroxene,
Spl = spinel, Srp = serpentine.
Figure 4. Remarkable petrographic textures of peridotites from the Mariana forearc.
vi
(A) Cleavable olivine grain with kink bands (125-780C-10R-1/18–21; XPL). (B)
Wavy cleavages in Opx (366-1498B-13R-3/21–26; XPL). (C) The texture of
incongruent melting, dissolution of Opx and precipitation of Ol as a replacement
(195-1200A-6R-2/21–25; XPL). Ol embayment is precipitated among Opx grains
with extremely irregular boundaries. (D) Texture of incongruent melting (366-
1498B-13R-3/103–106; XPL). An Opx grain is divided by dissolution, and Ol
replaces opx. The abbreviation is the same as Figure 3.
Figure 5. Plane-polarized photomicrographs of Cr-spinels from the Mariana
forearc mantle peridotites. (A) A symplectic texture of Cr-spinel with Opx and Cpx
in a harzburgite (366-1493B-9X-1/85–87). (B) The holly-leaf texture of Cr-spinel
in a harzburgite (195-1200A-6R-2/21–25). (C) A subhedral spinel in a dunite
sample (366-1493B-9X-1/60–63). (D) Cr-spinel with the highest Cr# in studied
samples in a dunite sample (366-1495B-3G-CC/21-23).
Figure 6. The typical texture in the serpentinization process under cross-polarized
light. (A) Olivine is surrounded by mesh-textured serpentine (125-780C-16R-1/53–
59). (B) Pophryclast Opx is serpentinized into bastite. Relicts of Opx is in the
upper left part of the picture (366-1498B-13R-3/21–26). (C) Antigorite phase of
serpentine with Ol (125-780C-10R-1/18–21). (D) Talc is along the rim of Opx by
steatitization (125-780C-9R-1/42–45). Abbreviations are the same as Figure 3; Tlc
= Talc.
Figure 7. Plots of major element concentrations against wt.% MgO for the Mariana
forearc peridotites. Increasing of MgO content indicates increasing degrees of
partial melting in mantle peridotite. Circles – harzburgite; squares – dunite;
vii
triangles – serpentinites. Colors in harzburgite samples represent groups
distinguished by REE patterns of clinopyroxene (see text for details).
Figure 8. Plots of Al2O3, TiO2, Cr#, and Na2O against Mg# in clinopyroxene from
the Mariana forearc harzburgites. Higher Mg# indicates experience in more degree
of partial melting. Areas of abyssal peridotite (gray) and SSZ peridotite (pale
green) Cpx have originated from Choi et al. (2008). The studied samples which
have refractory nature are included in SSZ peridotite. Only a few samples (red and
yellow; group 2 and 3) have higher Na contents with high Mg# than other samples
(blue; group 1)
Figure 9. Olivine-spinel mantle array from Arai (1991) with the samples from the
Mariana forearc peridotites. The samples experienced higher degrees of partial
melting than 20%. They are gathered in the relatively small area except for a dunite
sample. Gray and pale green fields indicate arrays of abyssal peridotite and SSZ
peridotite, respectively.
Figure 10. A plot of Cr# vs. TiO2 in spinel from the Mariana forearc peridotites.
The partial melting trend was produced by a combination of calculated Cr# from
Hellebrand et al. (2001) and TiO2 from Johnson et al. (1990). Three-fourths of
spinel from abyssal peridotite (Warren et al., 2016) have experienced relatively low
degrees of partial melting around 10%. Data from previous study of forearc
peridotite are depicted as pale green (Parkinson & Pearce, 1998). However, the
studied samples have low Ti contents and high Cr#. Spinel in only group 3 has high
TiO2 (0.18 wt.%). The values of spinel in different basaltic magmas (boninite, IAT
and MORB) are shown with a dark gray area and crosses.
viii
Figure 11. Chondrite-normalized REE patterns of clinopyroxene from the Mariana
forearc harzburgite. Gray field is the area of abyssal peridotites and gray bold line
is the average of the abyssal peridotites (Warren et al., 2016). Data from previous
study of forearc peridotite are depicted as pale green (Parkinson & Pearce, 1998).
According to patterns, harzburgites are divided into three groups. Group 1 shows
similar patterns of abyssal peridotite but has significantly low values of abyssal
peridotite. Group 2 shows quite distinctive flat patterns relative to other groups. A
pattern of group 3 is similar to that of abyssal peridotite such as group 1. However,
the concentrations of REE in group 3 is higher than the contents of group 1.
Figure 12. Primitive upper mantle-normalized trace element patterns of
clinopyroxene from the Mariana forearc harzburgite. Model lines calculated by
anhydrous fractional melting model (Shervais & Jean, 2012) are indicated by gray
dashed line and bold line, 15% and 25%, respectively.
Figure 13. Trace element compositions of Cr-spinels from the Mariana forearc
normalized by EPR MORB Cr-spinel. Gray line indicates compositions of Cr-
spinel from boninite. Group 1 harzburgite is represented in (A), and others are in
(B). Compositions of EPR MORB and boninite is from Pagé and Barnes (2009).
Figure 14. Primitive mantle normalized compositions of PGE from the Mariana
forearc peridotites. (A) group 1, (B) group 2, (C) group 3, and (D) dunite. A gray
area is representative of compositions from abyssal peridotites (Hess Deep from
the Pacific Ocean; Becker & Dale, 2016) and pale blue area is high depleted
abyssal peridotite (Marchesi et al., 2013).
ix
Figure 15. Correlations between Ir and other PGE
Figure 16. Sulfur concentrations from the Mariana forearc peridotites with (A)
LOI, (B) the degree of partial melting, (C) Al2O3 (wt.%) of whole rocks. A dunite
sample with much higher S contents (799 ppm) is indicated by black arrows (366-
1495B-3G-CC/21–23). Symbols are from Figure 15.
Figure 17. Composition variations of PGE from the Mariana forearc peridotites
with S contents. Because 366-1495B-3G-CC/21–23 has much higher S (799 ppm),
it is indicated by black arrows. Symbols are from Figure 15.
Figure 18. Calculated oxygen fugacity of the Mariana forearc peridotites with TiO2
contents of spinel as a proxy of the degree of melting and interaction with melts.
Symbols are the same as Figure 15.
Figure 19. Trace element compositions versus Ti contents in clinopyroxene from
the harzburgites. Dashed lines represent anhydrous melting of the starting point
(PUM; McDonough & Sun, 1995) with points of each 5% as small bars. Gray stars
are residues of 9% partial melting and starting points of hydrous melting. Bold
lines indicate re-fertilized hydrous melting suggested by Bizimis et al. (2000). Dark
gray fields are the areas of clinopyroxene from abyssal peridotite (Warren et al.,
2016).
Figure 21. Calculated melt composition normalized by chondrite (McDonough &
Sun, 1995). Melt compositions are predicted by multiplying concentrations of REE
in clinopyroxene and partition coefficients between basalt and clinopyroxene (see
text). Compositions of group 1 are plotted as blue area. Compositions of boninite
x
and forearc basalt are colored red and gray, respectively.
Figure 22. Concentrations of Pd and Ir against Al2O3 of whole rocks with the
model from Rehkämper et al. (1999a). Simple melting of sulfide (congruent
melting) cannot predict PPGE/IPGE fractionation in the mantle. Partition
coefficients between sulfide melt and silicate melt of PGE are assumed as 10,000
(Ir and Pd are from Rehkämper et al., 1999a). Values of PGE in peridotites from
other studies are plotted together. SSZ – supra-subduction zone; AP – abyssal
peridotite.
Figure 23. Chondrite-normalized concentrations ratios of PGE by those of Ir
against Al2O3 of whole rocks with the model from Rehkämper et al. (1999a).
Simple melting of sulfide (congruent melting) cannot predict PPGE/IPGE
fractionation in the mantle. Partition coefficients between sulfide melt and silicate
melt of PGE are assumed as 10,000 (Ir and Pd are from Rehkämper et al., 1999a).
Values of PGE in peridotites from other studies are plotted together. SSZ – supra-
subduction zone; AP – abyssal peridotite.
xi
List of Tables
Table 1. Mineral modes and degrees of serpentinization
Table 2. Compositions of major elements and trace elements from whole rocks by
XRF (wt.%)
Table 3. Compositions of major elements from olivines by EPMA (wt.%)
Table 4. Compositions of major elements from orthopyroxenes by EPMA (wt.%)
Table 5. Compositions of major elements from clinopyroxenes by EPMA (wt.%)
Table 6. Compositions of major elements from spinels by EPMA (wt.%)
Table 7. Compositions of trace elements in clinopyroxenes from harzburgite
samples analyzed by LA-ICP-Ms
Table 8. Compositions of trace elements from spinels analyzed by LA-ICP-MS
Table 9. Compositions of PGE and S in whole rocks and PM-normalized PGE
ratios
1
1. Introduction
Heterogeneous mantle, the source of different magmas, have been
studied with various temporal and spatial scales (Sobolev et al., 2007; Liang
and Liu, 2018). Although the mantle is commonly homogeneous for major
elements, the mantle has compositional variation in trace elements and
isotopic ratios (Walter, 2014). Regarding the planetary scales of
heterogeneity in the mantle, crustal materials from subduction system play
an important role in causing heterogeneous mantle (Hofmann, 1997). In
the smaller scale, a single tectonic setting, Walter (2014) has argued melt
extraction from the mantle is a hotkey in the discussion of mantle
heterogeneities. In addition, it has been known that melt-rock interaction in
the mantle can generate mantle heterogeneity by recent studies (Kelemen et
al., 1990; Pearce et al., 2000; Dijkstra et al., 2003; Rampone et al., 2008;
Birner et al., 2017).
Platinum group elements (PGE), as highly siderophile and chalcophile
elements, in the mantle have been actively studied because they have
provided new insights of the Earth’s interior, e.g., core-mantle segregation
and late accretion (Lorand et al. 2008; Dale et al. 2012a). Furthermore,
Park et al. (2018) suggested that the Pd/Pt ratio in the magmas plays an
important role in the formation of porphyry Cu±Au deposits. Thus,
understanding the PGE abundance in the mantle is critical to determine how
the primitive magmas are fertile in PGE. For example, Cu, Au, Pt, and Pd
2
from the Giant Ladolam Au deposit on Lihir Island, Papua New Guinea
have originated in the mantle beneath the deposit (McInnes et al., 1999).
Porphyry Cu deposits are particularly distributed above subduction
zones (Park et al. 2018). Supra-subduction zone (SSZ) type peridotites from
ophiolite suites are candidates that can tell how mantle rocks have
experienced essential processes in the subduction zone such as partial
melting, melt-rock interaction, and so on. For example, interaction with arc-
related melt such as boninitic melt would be able to enrich platinum-like
PGE (PPGE), which is more incompatible than iridium-like PGE (IPGE), in
mantle peridotites from ophiolite suites (Büchl et al., 2002; Uysal et al.,
2012). However, Luguet and Reisberg (2016), who reviewed the behavior of
PGE in the mantle, proposed that interaction with any melts do not cause the
enrichment of PPGE in mantle peridotite. They suggested an interaction
between S-saturated melts and mantle peridotites can bring the enrichment
in the mantle.
Abundance or behaviors of PGE in subduction zone mantle are
investigated from mantle xenolith in magmas or ophiolite suite (McInnes et
al., 1999; Kepezhinskas et al., 2002; Lee, 2002; Zhou et al., 2005). Recently,
International Ocean Discovery Program (IDOP) conducted expedition 366
in mud volcano of the Mariana forearc (Yinazao, Asút Tesoru, and
Fantangisña seamounts). Mantle rock fragments from serpentine seamounts
have been newly studied as a new window into the mantle (Parkinson &
3
Pearce, 1998; Savov et al., 2005, 2007; Kodolányi et al., 2012; Debret et al.
2018). Nevertheless, research about HSE chemistry from the forearc mantle
is still barren.
In this study, mantle rock samples from the Expedition 366 and two
more past expeditions of Ocean Drilling Project (ODP), ODP 125 and 195
(Conical and South Chamorro seamounts) are investigated with petrological
and geochemical views. I find out previous magmatic and hydrothermal
processes which are the Mariana forearc peridotite experienced with whole-
rock major elements, PGE, major and trace elements of primary minerals in
mantle rocks. The mantle peridotite from the Mariana forearc experienced
high degrees of partial melting with hydrous fluid and melt-rock interaction,
and serpentinization processes. After the processes are determined, PGE
geochemistry in the subduction zone peridotites from mud volcanoes would
be examined as pioneering work. The behavior of PGE can be affected by
fluxed partial melting and what melts reacted with the peridotite. Saturation
of S is a critical factor of the behavior PGE between subduction zone
peridotite and arc-related magma.
4
2. Geological Background
2.1 Subduction system of the Mariana Arc
The Mariana arc is included the vast tectonic system of the Philippine
oceanic plate. Notably, it is located in the southern part of the Izu-Bonin-
Mariana (IBM) arc and extended from a central part of Honshu, Japan to
Guam, the United States with 2,800 km range of arc volcanoes (Fig. 1). The
oldest basaltic rock sample in the Philippine plate was collected from the
Amami Plateau, northern West Philippine Basin and recorded early
Cretaceous ages (115 Ma from Hickey-Vargas, 2005; 117 and 119 Ma from
Ishizuka et al., 2011b), but this arc basalt was interpreted as a product from
an ancient arc, which had existed between Jurassic– Paleocene (Hickey-
Vargas, 2005). The Izu-Bonin-Mariana subduction system occurred in early
Eocene (52 Ma; Ishizuka et al., 2011a) with subducting of the western part
of Pacific oceanic plate beneath the Philippine plate.
According to Stern (2004), two subduction initiation hypotheses
were suggested: induced and spontaneous initiation. He argued that IBM
subduction is an example of spontaneous initiation which happens due to
different densities of two collided plates. From the IBM subduction system,
5
a number of studies have been conducted for recent years (Stern, 2004;
Ishizuka et al., 2006, 2011a; Reagan et al., 2010, 2017; Shervais &Choi,
2012; Stern et al., 2012; Arculus et al., 2015; Leng &Gurnis, 2015; Stern
&Gerya, 2017). When subduction of the Pacific plate commenced,
decompression melting occurred to fulfill empty room between the
Philippine plate and the Pacific plate (Reagan et al., 2017). Products of the
decompression melting of asthenosphere are similar to MORB (Mid-ocean
ridge basalt), and it is called forearc basalt (FAB; Reagan et al., 2010). After
subduction proceeded more, hydrous fluid caused by dehydration of
subducting plate drenched the mantle wedge above the slab. Hydrous fluid
in the mantle can lower the solidus of the mantle, and this process produces
arc-related magmas (e.g., boninite, andesite, etc.).
The Kyushu-Palau Ridge and the West Mariana Ridge are remnant
arcs from backarc spreading of proto-IBM arc in 30 or 29 Ma and 7 Ma,
respectively (Okino et al., 1998; Stern et al., 2003). First backarc spreading
formed the Shikoku and the Parece Vela Basin. The recent backarc
spreading split the southern part of the previous arc into the West Mariana
Ridge and the Mariana arc leaving the Mariana Trough.
6
In the Mariana forearc, FAB, island arc tholeiite (IAT), and boninite
are yielded by a magmatic process in subduction systems (Ishizuka et al.,
2011a, 2011b; Reagan et al., 2010, 2017). The similarity of lithology and
geochemical characteristics between the Mariana forearc igneous rock and
SSZ type ophiolite suite can consider the Mariana forearc as an optimal
example to study not only subduction initiation but also the formation of
SSZ type ophiolite (Ishizuka et al., 2014).
7
Figure 1. Geological maps of the studied area. (a) The Mariana arc located eastern
boundary of the Philippine oceanic plate. (b) Detailed bathymetric map of sampled
areas in the Mariana forearc. Five mud volcanoes (Conical, Asút Tesoru,
Fantangisña, Yinazao, and South Chamorro seamounts) are located along the
Mariana trench.
8
2.2. Mud volcano in the Mariana forearc
Extension force in the Mariana forearc region has formed numerous
normal faults from the surface of the subducting slab to the sea floor (Fryer
et al., 2000; Stern et al., 2004). The Mariana subduction system is included
in the erosive margin group considering that there is no accretionary wedge
(Clift & Vannucchi, 2004). Both conditions of extension and no accretionary
wedge offer an excellent environment for the formation of serpentine
seamounts which is so-called mud volcanos (Fryer, 1992). Massive mud
volcano chain in the Mariana forearc is solely reported as the only region
where active mud volcanoes are found (Fryer et al., 2000).
Because over half of the dehydration from the subducting slab
occur beneath the forearc region (Jarrard, 2003), mantle wedge peridotite is
highly soaked, leads to serpentinization. Serpentinization lowers the density
of peridotite, and cause density contrasts between serpentinized peridotites
and surrounding mantle rocks. Because of density contrasts, serpentinized
peridotites have buoyancy. Serpentine mud and serpentinized peridotite
fragment rise along normal faults in the overriding plate caused by
extension force and are outpoured to the seafloor (Fig. 2; Lockwood, 1972;
Fryer et al., 1996, 1999). Serpentine seamounts formed by this process
includes not only mud and serpentinized peridotites but also metamorphic
rocks of slab components such as blueschist (Fryer et al., 1999).
Serpentine mud volcanoes are located along the Mariana trench and
9
from 50 to 120 km from the trench (Figure. 1(b); Fryer et al., 1985). The
mud volcanoes have various size; for example, the diameters are 10–50 km
and the heights are 500–2000 m from the sea floor (Ishii et al., 1992). The
boundary between the subducting plate and the overriding plate is 10–25 km
below the seamounts (Fryer et al., 1999; Takahashi et al., 2007). Until now,
reported mantle peridotite fragments are serpentinized spinel harzburgite,
dunite, and pyroxenite (Parkinson & Pearce, 1998: D’antonio & Kristensen,
2004; Savov et al., 2007).
10
Figure 2. Schematic diagram showing the formation of mud volcanoes in the
Mariana forearc region modified from Fryer et al. (1999). ① Dehydration process
of the subducting slab. ② Hydrous fluid drenches mantle wedge, and mantle
wedge is serpentinized. It makes density contrasts between serpentinized peridotite
and surrounding mantle rocks. ③ Serpentinized peridotites and serpentine mud
rise along the normal faults in the overriding plate
11
3. Sample selection and Analytical methods
3.1 Sample selection
Fragments of mantle peridotites from five mud volcanoes on the
Mariana forearc were selected to be analyzed. Samples from Conical and
South Chamorro were acquired in ODP expeditions, Leg 125 and 195
respectively. Based on the core description at that times, we requested the
samples through IODP request system (http://web.iodp.tamu.edu/sdrm)
because these were stored at KCC (Kochi Core Center) in Kochi, Japan.
Yinazao, Asút Tesoru, and Fantangisña seamount were recently investigated
at IODP expedition 366, so samples from the seamounts were directly
attained from a board in the expedition.
Serpentinization is a pervasive process which forearc peridotites
experienced. It can impact on peridotites samples from two aspects. First,
serpentinization has own petrographic characteristics such as mesh-textured
olivine or bastitic texture of pyroxene. However, it can also ruin the pristine
texture of peridotites that can be critical evidence for identifying geological
processes. In addition, given geochemistry, concentrations of fluid-mobile
elements (e.g., B, Sr or Rb, etc.) can be highly influenced by the influx of
fluid into peridotites during serpentinization. On the basis of the above
reasons, we chose relatively less serpentinized samples.
Four samples from each seamount were selected except Fantangisña
seamount, from which seven samples were decided to analyze. All samples
12
from Yinazao seamount were fully serpentinized. In this case, fully
serpentinized peridotites were unavoidably chosen.
3.2 Whole rock major element and sulfur analyses
All major elements of samples were analyzed by XRF from Brigham
Young University (Siemens SRS300 X-ray fluorescence spectrometer) and
NCIRF (National Center for Inter-university Research Facilities at Seoul
National University; Shimadzu XRF-1700). Procedure in BYU was
provided from the website (http://geology.byu.edu/Home/content/Faculty%
20Directory/eric-h-christiansen). SiO2, Al2O3, FeO, MnO, MgO, CaO, K2O,
Na2O were detected by XRF (Table 2). I chose four standard materials,
BHVO-1 and PCC-1 from BYU, and MUH-1 and JP-1 from NCIRF. Data
of references from both institutes showed lower than 10% differences.
Sulfur contents in peridotites were analyzed Leco S-144DR, sulfur
determinator, combusting rock powder and estimating amounts of SO2. JP-1,
BHVO-2 and Coal reference from Leco Corporation were selected as
standard materials for calibration curve because concentrations of S were
variable (26.9 ppm to 0.28%).
13
3.3 Mineral major and trace element analyses by EPMA and
LA-ICP-MS
Major elements analysis of olivine, orthopyroxene, clinopyroxene, and
spinel was determined using SHIMADZU 1600 electron microprobe at
Korea Basic Science Institute (KBSI) and Jeonbuk Technopark (JBTP).
Analytical conditions were 15 kV of accelerating voltage, 20 nA of beam
current and one μm of beam size. Smithsonian microbeam standard minerals
were used as second standards for consistency in each institute. All oxide
contents in this paper were presented as EPMA data, but values of FeO and
Fe2O3 were calculated based on stoichiometry.
Trace element concentrations in clinopyroxene and spinel were
analyzed by laser ablation inductively coupled mass spectroscopy (LA-ICP-
MS) at KBSI and Korea Institute of Ocean Science and Technology
(KIOST). Given that concentrations of trace element in olivine and
orthopyroxene are expected to be near detection limits, the trace element
measurements of olivine and orthopyroxene were not conducted. This
system consists of Thermo scientific Company X2 with New Wave UP213
at KBSI and Agilent 7700X with NWR193 at KIOST. Although the
wavelengths of the laser in the two institutions are different, Gonzalez et al.
(2002) said that results using with 193 nm and 213 nm wavelength are not
different. In-house Excel spreadsheet, used in Park et al. (2015), was also
utilized in this study. Total ablation time was 40–60 sec. Background time
14
was determined for 45–50 sec and analyzing duration is around 35–50 sec.
The following isotopes were selected to estimate each concentration of
elements: 7Li, 11B, 25Mg, 29Si, 43Ca, 45Sc, 47Ti, 51V, 53Cr, 55Mn, 85Rb, 88Sr, 89Y,
90Zr, 93Nb, 133Cs, 137Ba, 139La, 140Ce, 141Pr, 146Nd, 147Sm, 153Eu, 157Gd, 159Tb,
163Dy, 165Ho, 166Er, 169Tm, 172Yb, 175Lu, 178Hf, 181Ta, 208Pb and 232Th in
clinopyroxene and 24Mg, 27Al, 29Si, 44Ca, 45Sc, 49Ti, 51V, 53Cr, 55Mn, 57Fe,
59Co, 61Ni, 65Cu, 66Zn, 71Ga, 101Ru, 103Rh, 105Pd, 185Re, 189Os, 193Ir and 195Pt
in spinel. NIST 612 and 610 glass were used as a primary standard for
spinel and clinopyroxene, respectively. UQAC-FeS-1 were also used as a
primary standard for platinum group elements in spinel grains. For stability
of analysis, BCR-2G was selected as a second standard. In clinopyroxene,
29Si and 43Ca contents of EPMA were decided to correct differences
between the standards and measured samples, which is reduction process. In
spinel, 27Al contents using EPMA were determined for reduction. The
difference between the main standard and BCR-2g is smaller than 10% in
clinopyroxene and 11% in spinel, respectively.
3.4 Whole rock major element and sulfur analyses
Whole-rock PGE contents of the peridotites were analyzed using Ni-
sulfide fire assay-isotope dilution method described in Park et al. (2015).
Contents of core drilling samples were limited, so only 2 g of peridotite
power were used. 2 g of this powder was mixed with 0.2 g of Ni, 0.1 g of S
15
and 2 g of sodium borax. An enriched PGE spike solution was incorporated
(99Ru, 105Pd, 185Re, 190Os, 191Ir, and 195Pt) in the powder. The mixtures were
dried at 100℃ for 30 min and fused in a furnace at 1200℃. The condition
of each crucible was reduced because of ~ 0.5 flour between inner and outer
crucibles. The sulfide beads from the quenched glass were dissolved by 6 M
HCl. The solutions were filtered with filter papers; then the papers were
dissolved by aqua regia. After that, these are dried again to ~ 0.1 ml and
diluted by 2% HNO3.
Elan 6100 ICP-MS from Perkin Elmer was performed to analyze
PGE concentrations in peridotites samples at Seoul National University
(SNU). Potential interference elements such as Ni, Cu, Zn, Co, Hf, Mo, Zr
and Ta with argon and oxygen were also measured as single solutions (e.g.,
Co, Mo and Hf solution) at the last part of each analysis. PGEs
concentrations were corrected by rates of argides or oxides in the solution
analysis. Data reduction is performed with domestic Excel spreadsheet as
used in Park et al. (2015). Procedure blanks were 0.07 ppb for Os, 0.01 ppb
for Ir, 0.01 ppb for Ru, 0.01 ppb for Rh, 0.19 ppb for Pt and 0.12 ppb for Pd.
16
4. Results
4.1 Petrography
Nineteen partly serpentinized peridotites and two fully serpentinites
were microscopically examined for petrography. Peridotites from mud
volcanoes were serpentinized with a varying degree (8% to 100%, Table 1).
The forearc mantle rocks from Yinazao seamount, the closest mud volcano
from the trench (55 km; Fryer, 2012), are completely serpentinized. The rest
four volcanoes are far from the trench (65 to 85 km), and it causes these
seamounts outpoured less serpentinized peridotites than those from Yinazao.
Mineral modes of each sample were determined using primary mantle
minerals and their relicts. Mineral modes were calculated from point-
counting with more than 400 points. Through the calculation, samples were
divided into three groups, harzburgite, dunite, and serpentinite. However,
primary textures in serpentinite samples were not able to be recognized
because spinel was the only relict mineral in serpentinites.
4.1.1 Harzburgite
Harzburgites from the seamounts are moderately to highly
serpentinized (20% to 72%). Harzburgites are composed of 57 – 86% of
olivine, 12 – 38% of orthopyroxene, – 4% of clinopyroxene and – 5% of
spinel. Some are characterized by porphyroclastic texture with 1 – 4 mm
sized orthopyroxene surrounded by olivine neoblasts (Figure 3A). The
17
others have the protogranular texture with similar sizes of olivine and
orthopyroxene (–2 mm; Figure 3B). Some olivine crystals in harzburgites
show kink bands and some orthopyroxene have wavy cleavages, which are
deformation evidence (Figure 4A and B). Along fractures in orthopyroxene,
bastite forms as serpentinization of orthopyroxene. Clinopyroxene is small
(<200 μm) and interstitial among other minerals. Surprisingly, some of the
samples shows the embayment of olivine grains in the orthopyroxene grains,
which means incongruent melting process of mantle (Figure 4C; Dijkstra et
al., 2003; Seyler et al., 2007). If it proceeds severely, an orthopyroxene
seems to be divided into two grains (Figure 4D). However, they were the
same grain because sharing cleavages and the extinction angle. A few
clinopyroxene grains exist in the serpentine matrix composed of complete
serpentinization of olivine and orthopyroxene because Ca-pyroxene cannot
join the serpentinization process. The grain size of Cr-rich spinel from
hundreds of μm to 2 mm. A few of spinel grains has symplectic texture with
ortho- and clinopyroxene (Figure 5A). Spinel grains are subhedral and often
show holly-leaf texture (Figure 5B). Some spinels have olivine inclusions
almost serpentinized.
The representative harzburgite sample is 195-120-7R-2 in South
Chamorro seamount. The result of the modal analysis in this thin section is
that olivine, orthopyroxene, clinopyroxene, and Cr-rich spinel is 82%, 17%,
<1%, and 1%, respectively. Serpentine generally occurs as pale white to
18
brownish veins or veinlets in thin sections, but greenish veins are also found
in orthopyroxene porphyroclasts. Olivine grains are as large as
orthopyroxene (<2 mm), but they are often penetrated by serpentine veinlets
and observed as tiny olivine aggregates. Uniform extinction angles and
same birefringence color variations among the grains support that those
small aggregates were initially a single and larger grain. Orthopyroxene
grains show undulatory extinction.
They are mostly anhedral with corroding boundaries by other minerals
such as olivine, spinel and even another orthopyroxene. Clinopyroxene
occurs only as scarce and small (<200 μm) grains. Spinels are interstitial
and rarely contain partly serpentinized olivine. Also, spinel grains have
ferrit-chromit rims because of low-temperature alteration such as
serpentinization (Ulmer, 1974)
4.1.2 Dunite
Four dunites samples were selected for this study. Only less than 8% of
relict minerals is accounted for clinopyroxene, orthopyroxene, and spinel. A
dunite sample from the Conical seamount (125-780C-10R-1/18–21) is an
exceptional sample because it has cleavable olivine crystals as well as
antigorite by high-temperature serpentinization (~450–550°C; Murata et al.,
2009; Figure 6C). Boundaries of orthopyroxene in 125-780C-8R-1 / 95–98
has talc minerals, which means replacement of serpentine minerals by talc
after serpentinization (steatitization; Bach et al., 2004; Klein and Bach,
19
2009; Figure 6D). 366-1493B-9X-1/60–63 is the least serpentinized
peridotite in our samples. It shows aligned olivine texture and has a large
size of olivine grain (~2 cm; Figure 3C). The dunite samples from Asút
Tesoru has euhedral to subhedral spinels in the serpentinized olivine matrix
(Figure 5C and D).
4.1.3 Petrographic features of serpentinization
Serpentinite classified in this study is completely serpentinized and has
no primary silicate relicts. Only spinel grains survived in severe
serpentinized condition. In this section, typical serpentinization texture of
the Mariana forearc peridotite is dealt with from partially serpentinized
peridotite (Figure 6). The mesh-textured serpentine with relicts olivine
grains is pervasive in the peridotite because mode of olivine is much higher
than other silicate minerals in harzburgite and dunite samples (Figure 6A).
Orthopyroxene grains show quite different products of serpentinization,
named by bastite (Figure 6B). Bastite is pseudomorph after pyroxene
(Dungan, 1979), displaying cleavages and shapes like pyroxene, however,
relief is very low and interference color is dark (Figure 6B). Antigorite,
which has acicular texture, was found only a dunite sample (125-780C-10R-
1/18–21) with cleavable olivine, talc and brucite. This assemblage and
presence of antigorite phase means the high-temperature serpentinization
(~450–550°C; Murata et al., 2009; Figure 6C). Also, talc was detected in
some peridotite from Conical Asút Tesoru, Fantangisña seamounts, which
20
are relatively far from trench (Saboda et al., 1992; Fryer et al. 2018).
21
Figure 3. Thin section photomicrograph of the peridotites from the Mariana
forearc. (A) Porphyroclastic texture of a harzburgite sample (195-1200A-7R-2/84–
87; XPL). Porphyroclast of Opx is mantled by neoblasts of Ol. (B) Protogranular
texture of a harzburgite sample (366-1498B-7R-4/88–93; XPL). Several mm of Ol
and Opx grain size are characteristics of this texture. (C) A dunite sample with
inclined Ol (366-1493B-9X-1/60–63; XPL). Euhedral spinels are included in
olivine matrix. Olivine grains are stretched as left diagonal patterns. Mineral
Abbreviations from Whitney & Evans (2010): Ol = olivine, Opx = orthopyroxene,
Spl = spinel, Srp = serpentine.
22
Figure 4. Remarkable petrographic textures of peridotites from the Mariana
forearc. (A) Cleavable olivine grain with kink bands (125-780C-10R-1/18–21;
XPL). (B) Wavy cleavages in Opx (366-1498B-13R-3/21–26; XPL). (C) The
texture of incongruent melting, dissolution of Opx and precipitation of Ol as a
replacement (195-1200A-6R-2/21–25; XPL). Ol embayment is precipitated among
Opx grains with extremely irregular boundaries. (D) Texture of incongruent
melting (366-1498B-13R-3/103–106; XPL). An Opx grain is divided by
dissolution, and Ol replaces opx. The abbreviation is the same as Figure 3.
23
Figure 5. Plane-polarized photomicrographs of Cr-spinels from the Mariana
forearc mantle peridotites. (A) A symplectic texture of Cr-spinel with Opx and Cpx
in a harzburgite (366-1493B-9X-1/85–87). (B) The holly-leaf texture of Cr-spinel
in a harzburgite (195-1200A-6R-2/21–25). (C) A subhedral spinel in a dunite
sample (366-1493B-9X-1/60–63). (D) Cr-spinel with the highest Cr# in studied
samples in a dunite sample (366-1495B-3G-CC/21-23).
24
Figure 6. The typical texture in the serpentinization process under cross-polarized
light. (A) Olivine is surrounded by mesh-textured serpentine (125-780C-16R-1/53–
59). (B) Pophryclast Opx is serpentinized into bastite. Relicts of Opx is in the
upper left part of the picture (366-1498B-13R-3/21–26). (C) Antigorite phase of
serpentine with Ol (125-780C-10R-1/18–21). (D) Talc is along the rim of Opx by
steatitization (125-780C-9R-1/42–45). Abbreviations are the same as Figure 3; Tlc
= Talc.
25
Table 1. Mineral modes and degrees of serpentinization
*The samples have no relict olivine grains, so the calculation for oxygen fugacity
cannot be applied in the samples. ┼The samples were almost completely serpentinized. Modes of these samples
cannot indicate the primary lithologies. However, these samples were named as
harzburgite because serpentine textures had been distinguished between mesh-
textured serpentine and bastite.
26
4.2 Whole rock major element chemistry
Compositions of major elements in bulk rocks are presented in Table 2.
Major element contents are shown against MgO content. Partial melting of
primitive upper mantle gives rise to increase of olivine portion in the rock
because melting proportion of pyroxene is much higher than that of olivine
during partial melting (20:30:50 = Ol:Opx:Cpx during dry melting; Johnson
et al., 1990). Augmentation of olivine mode entails an increase of MgO
contents in bulk rock. Thus, MgO is used as a depletion indicator in Figure 7
(e.g., Parkinson & Pearce, 1998). The decrease of SiO2 with increasing
MgO results from a lower ratio of Mg:Si in olivine structures (2:1) than that
of in pyroxene structures (1:1 or 1:2). The negative correlation of Al2O3 and
CaO with MgO indicates depletion of pyroxene which is the main host of
the elements. Iron and Mn contents in peridotites have no strong trends, but
FeO in serpentinites from Yinazao seamounts is a robust positive correlation
with MgO. Nickel is partitioned into olivine minerals, so it has a positive
trend during partial melting.
Loss on ignition (LOI) values of the samples have been used as an
index of the degree of serpentinization (e.g., Parkinson & Pearce, 1998;
Agranier et al., 2007; Scott et al., 2017). The degree of serpentinization
from modal analyses has a positive correlation with the value of LOI in
Figure A1 except for a sample (125-780C-10R-1/18–21). Exception sample
may have experienced high-temperature conditions of serpentinization at
27
~450–550 ℃ because of the assemblage of antigorite, cleavable olivine and
talc (Murata et al., 2009). At high-temperature condition of serpentinization,
talc generation (steatitization) can take place in serpentinized peridotite
(Bach et al., 2004). The content of water in talc (~5%) is lower than that of
serpentine minerals (~13%). Thus, LOI means the degree of serpentinization
except for the talc-rich sample.
28
Figure 7. Continued.
29
Figure 7. Plots of major element concentrations against wt.% MgO for the
Mariana forearc peridotites. Increasing of MgO content indicates increasing
degrees of partial melting in mantle peridotite. Circles – harzburgite; squares
– dunite; triangles – serpentinites. Colors in harzburgite samples represent
groups distinguished by REE patterns of clinopyroxene (see text for details).
30
Table 2. Compositions of major elements and trace elements from whole rocks by
XRF (wt.%)
31
4.3 Mineral chemistry
4.3.1 Major element chemistry
Major element composition of olivine, orthopyroxene, clinopyroxene,
and spinel are presented in Table 3, 4, 5, and 6. Each value in a sample is an
average of over three different grains in the sample.
Forsterite (Fo) contents of olivine vary between 89.7 and 92.8.
Contents of other major elements of olivine as well as Fo contents in
samples are little different. The orthopyroxene has Mg# (Mg/(Mg+Fe))
ranging from 91.4–92.3. The Al2O3 (<3 wt.%) and TiO2 (<0.05 wt.%)
contents in both orthopyroxenes are much lower than those from reported
abyssal peridotite but match with those from SSZ peridotite (Choi et al.,
2008; Uysal et al., 2012). Incompatible elements such as Al, Ti, and Na
from clinopyroxene are also depleted like the case of orthopyroxene (Figure
8). However, Na2O contents of some samples are higher than SSZ peridotite
indicating the addition of the elements after the depletion process.
When measuring spinel grains, peripheral rims of the grains are
avoided because Ulmer (1974) argued that altered spinels might have iron-
enriched rims (ferrit-chromit rind). Olivine-spinel mantle array (OSMA,
Figure 9; Arai, 1994) shows the samples experienced a high degree of
partial melting more than 20%. Spinel grains in all samples share similar
compositions, especially regarding Cr# (Cr/(Cr+Al); 0.48–0.59), except a
dunite sample (366-1495-3G-CC, 21–23), has significantly higher Cr# =
32
0.79. Figure 10 also shows refractory characteristics of the samples with low
TiO2 contents as well as high Cr#. Most TiO2 contents in spinel are
significantly low (<0.1 wt.%), whereas TiO2 enriched (~0.2 wt.%) spinels
are found in only one sample (195-1200-6R-2, 21–25). It implies the sample
experienced re-fertilizing process.
According to the nomenclature criteria of spinel group minerals
suggested by Stevens (1944), those spinel compositions cover Al-chromite
and Cr-spinel (Figure A2). However, considering that the high Cr# (> 0.5)
spinels (i.e., Al-chromite) from the Mariana forearc ultramafic xenoliths
have been collectively called "spinel" or "chromian spinel" (e.g. Parkinson
& Pearce, 1998; Savov et al., 2007) and that most samples hosting these
spinels are not distinguishable with regard to Cr#. Thus, in this study, the
term "spinel" will denote both Al-chromite and Cr-spinel hereafter. Start your
thesis.
33
Figure 8. Plots of Al2O3, TiO2, Cr#, and Na2O against Mg# in clinopyroxene from
the Mariana forearc harzburgites. Higher Mg# indicates experience in more degree
of partial melting. Areas of abyssal peridotite (gray) and SSZ peridotite (pale
green) Cpx have originated from Choi et al. (2008). The studied samples which
have refractory nature are included in SSZ peridotite. Only a few samples (red and
yellow; group 2 and 3) have higher Na contents with high Mg# than other samples
(blue; group 1)
34
Figure 9. Olivine-spinel mantle array from Arai (1991) with the samples from
the Mariana forearc peridotites. The samples experienced higher degrees of
partial melting than 20%. They are gathered in the relatively small area except
for a dunite sample. Gray and pale green fields indicate arrays of abyssal
peridotite and SSZ peridotite, respectively.
35
Figure 10. A plot of Cr# vs. TiO2 in spinel from the Mariana forearc peridotites.
The partial melting trend was produced by a combination of calculated Cr#
from Hellebrand et al. (2001) and TiO2 from Johnson et al. (1990). Three-
fourths of spinel from abyssal peridotite (Warren et al., 2016) have experienced
relatively low degrees of partial melting around 10%. Data from previous study
of forearc peridotite are depicted as pale green (Parkinson & Pearce, 1998).
However, the studied samples have low Ti contents and high Cr#. Spinel in only
group 3 has high TiO2 (0.18 wt.%). The values of spinel in different basaltic
magmas (boninite, IAT and MORB) are shown with a dark gray area and
crosses.
36
Table 3. Compositions of major elements from olivines by EPMA (wt.%)
37
Table 4. Compositions of major elements from orthopyroxenes by EPMA (wt.%)
38
Table 5. Compositions of major elements from clinopyroxenes by EPMA (wt.%)
39
Table 6. Compositions of major elements from spinels by EPMA (wt.%)
40
4.3.2 Trace element chemistry
Compositions of trace elements in clinopyroxenes are listed in Table
7. The chondrite-normalized (McDonough & Sun, 1995) clinopyroxene rare
earth element (REE) compositions from harzburgite samples are plotted
with the average value of those from abyssal peridotites (Warren et al.,
2016) in Figure 11. Patterns of REE in clinopyroxenes from each
harzburgite sample are clearly divided three groups; first group (group 1)
shows similar patterns with the average value of abyssal peridotite which
has a feature that more depleted in LREE than HREE because of
incompatibility among the REE (CeN/YbN = 0.01), but group 1 has much
more depleted patterns. Contents of LREE vary as much significance as two
orders of magnitude compared to compositions of HREE. The second group
(group 2) has flat patterns of REE in clinopyroxene which means CeN/YbN =
0.23, and SmN/YbN = 0.49, however, it has comparable contents of HREE
from group 1. Last group (group 3) includes has a characteristic which is a
similar pattern with group 1 (CeN/YbN = 0.01), but less depleted than group
1. Hereafter, data from harzburgite samples are marked with subgroups,
group 1, 2, and 3.
Extended trace elements patterns of harzburgite samples and patterns
of non-modal fractional melting (Johnson et al., 1990) are shown in Figure
12. Through non-modal fractional melting model (Johnson et al., 1990),
samples from the Mariana forearc are expected that they experienced 20% –
41
25% of partial melting. Fluid mobile elements (FME) and LREE in all
groups are enriched compared to the value calculated by the model.
Compositions of high field strength elements (HFSE) in group 1 fall within
the model range. However, HFSE in group 2 and 3 are a little more enriched
than those in group 1, especially Zr.
In Figure 13 and Table 8, patterns and concentrations of trace elements
(Ga, Ti, Ni, V, Zn, Co, Mn, and Sc) in spinel from harzburgite groups and
dunite are represented. Patterns of all samples are similar to that of spinel
from boninite lava rather than MORB. All patterns show similar to each
other, and they have positive Zn anomaly and negative Ti and Sc anomaly
except group 3. Contents of Ti are highly variable in group 1 whereas other
groups and dunites have limited ranges of Ti contents. Spinel grains in
group 3 have a unique characteristic which is positive Ti anomaly. The
concentration of PGE in spinel grains is shown in Table A1. The feature of
PGE concentrations in spinel is that Ru is enriched compared to other PGE.
Spinel from harzburgite has Ru up to 149 ppb (366-1498-7R-4/88–93), and
spinel from a dunite sample has Ru up to 203 ppb (366-1493-9X-1/60–63).
42
Figure 11. Chondrite-normalized REE patterns of clinopyroxene from the
Mariana forearc harzburgite. Gray field is the area of abyssal peridotites and
gray bold line is the average of the abyssal peridotites (Warren et al., 2016). Data
from previous study of forearc peridotite are depicted as pale green (Parkinson &
Pearce, 1998). According to patterns, harzburgites are divided into three groups.
Group 1 shows similar patterns of abyssal peridotite but has significantly low
values of abyssal peridotite. Group 2 shows quite distinctive flat patterns relative
to other groups. A pattern of group 3 is similar to that of abyssal peridotite such
as group 1. However, the concentrations of REE in group 3 is higher than the
contents of group 1.
43
Figure 12. Primitive upper mantle-normalized trace element patterns of clinopyroxene from the Mariana forearc harzburgite. Model lines
calculated by anhydrous fractional melting model (Shervais & Jean, 2012) are indicated by gray dashed line and bold line, 15% and 25%,
respectively.
44
Figure 13. Trace element compositions of Cr-spinels from the Mariana forearc
normalized by EPR MORB Cr-spinel. Gray line indicates compositions of Cr-spinel
from boninite. Group 1 harzburgite is represented in (A), and others are in (B).
Compositions of EPR MORB and boninite is from Pagé and Barnes (2009).
45
Table 7. Compositions of trace elements in clinopyroxenes from harzburgite
samples analyzed by LA-ICP-MS
bdl = below detection limit
46
Table 8. Compositions of trace elements from spinels analyzed by LA-ICP-MS
47
4.4 Whole rock PGE and S chemistry
Harzburgite groups and dunite from the Mariana forearc have PGE
concentrations and patterns from IPGE to PPGE comparable to abyssal
peridotites (Table 9 and Figure 14). Compositions are normalized by
primitive mantle from Becker et al. (2006). Contents of PGE from group 1
not only overlap with abyssal peridotite but also widely vary than abyssal
peridotite (Figure 14A). They are fractionated between IPGE and PPGE
(PdN/IrN = 0.1–0.5 except a sample, 195-1200A-6R-2/21–25 which have
PdN/IrN = 0.9), however, Pt is less fractionated than Pd (PtN/IrN = 0.2–1.0).
195-1200A-7R-2/17–21 have Ir positive anomaly, and 195-1200A-7R-2/84–
87 have extremely depleted in PGE (∑PGE = 2.3 ppb).
Whereas group 2 and dunite have little discrepancy in patterns of group
1, fractionation between PPGE and IPGE is more substantial than group 1
and abyssal peridotite (PdN/IrN = 0.01–0.2). Their patterns are included in a
narrower range than a range of patterns from group 1. Group 3, however,
has an entirely different pattern which is enriched in PPGE (PdN/IrN = 1.2
and PtN/IrN = 1.5).
Figure 15 shows correlations between compositions of PGEs. There are
quite high correlations among Os, Ru, and Rh with Ir (Figure 15(a)–(c)).
However, contents of Pt and Pd have scattered distribution with Ir (Figure
15(d) and (e)). Palladium and Pt do not correlate with their compositions
(Figure 15(f)). The results indicate IPGE and Rh may share their host phases,
48
e.g., alloys, PGM (platinum group mineral), or sulfides.
Sulfur compositions in the Mariana forearc peridotites are represented
with various indicators in Figure 16. Sulfur is known to deplete with
increasing degree of partial melting from 250 ppm (S in PUM = 250 ppm;
McDonough and Sun, 1995). Aluminum contents in peridotites mean how
the sample rocks experience the partial melting. Plots of S against Al2O3 and
degree of melting from HREE show that S contents in the samples are
various (S = 49–276 ppm for harzburgite, S = 181–799 ppm for dunites, and
S = 2979 ppm for serpentinite), but Al2O3 and degree of melting from
HREE are not (Figure 16(a) and (b)). Although serpentinite shows the
highest S content (2979 ppm), sulfur contents are not correlated with the
degree of serpentinization (Figure 16(c)). In Figure 17, concentrations of
PGEs against S show scattered distributions. It might mean that hosts of
PGE are restrictively associated with sulfides.
49
Figure 14. Continued.
50
Figure 14. Primitive mantle normalized compositions of PGE from the Mariana
forearc peridotites. (A) Group 1, (B) Group 2, (C) Group 3, and (D) dunite. A gray
area is representative of compositions from abyssal peridotites (Hess Deep from the
Pacific Ocean; Becker & Dale, 2016) and pale blue area is high depleted abyssal
peridotite (Marchesi et al., 2013).
51
Figure 15. Correlations between Ir and other PGE
52
Figure 16. Sulfur concentrations from the Mariana forearc peridotites with (A)
LOI, (B) the degree of partial melting, (C) Al2O3 (wt.%) of whole rocks. A dunite
sample with much higher S contents (799 ppm) is indicated by black arrows (366-
1495B-3G-CC/21–23). Symbols are from Figure 15.
53
Figure 17. Composition variations of PGE from the Mariana forearc peridotites
with S contents. Because 366-1495B-3G-CC/21–23 has much higher S (799 ppm),
it is indicated by black arrows. Symbols are from Figure 15.
54
Table 9. Compositions of PGE and S in whole rocks and PM-normalized PGE ratios
55
4.5 Oxygen fugacity
Generally, methods from Wood et al. (1990) and Ballhaus et al.
(1991) have been used to determine oxidation states of mantle rocks
(Parkinson & Arculus, 1999; Birner et al., 2017). They are based on the
reaction: 6Fe2SiO4 (olivine) + O2 (fluid) = 3Fe2Si2O6 (orthopyroxene) +
2Fe3O4 (spinel). Wood’s method needs information of olivine,
orthopyroxene, and spinel, however, Ballhaus’ method can estimate
oxygen fugacity from mineral data of olivine and spinel. The studied
harzburgite, as well as dunite samples from the Mariana forearc, are so
highly serpentinized that it is hard to find pairs of three minerals. Thus, I
have employed Ballhaus’ method for calculations of oxygen fugacity.
Harzburgite samples are recorded oxygen fugacities of ΔFMQ =
+0.47 to +2.34 except group 3 with oxygen fugacity of ΔFMQ = -1.42
(Figure 18 and Table 2). Some dunite samples included in the harzburgite
area, and the others have much higher oxygen state (ΔFMQ = +2.99).
Serpentinite samples have no detectable olivine grains, so there is no
oxygen fugacity data from serpentinites.
56
Figure 18. Calculated oxygen fugacity of the Mariana forearc peridotites with
TiO2 contents of spinel as a proxy of the degree of melting and interaction with
melts. Symbols are the same as Figure 15.
57
5. Discussion
5.1 Highly depleted features compared to the abyssal
peridotite
The main processes of mantle melting are decompression melting
primarily occurred in mid-ocean ridges and hydrous partial melting
caused by hydrous fluid from subducting slab. Abyssal peridotite is the
residues of decompression melting in the upper mantle (Kelemen et al.,
1997; Niu, 2004; Warren, 2016). The average degree of melting from
adiabatic decompression melting of abyssal peridotite is 10–15%
estimated by fractional partial melting from the primitive mantle (Niu,
2004).
Hydrous fluid originated from desiccation or dehydration of
subducting slab plays a vital role in the mantle melting in subduction
system. It can make high Mg volcanic rock called boninite because of one
or more melt extraction in the source mantle (Crawford et al., 1998). As a
result, source mantle of boninite become highly refractory. Studied
harzburgites and dunites show large depletion features such as low CaO,
Al2O3 contents of whole rocks, high Mg# of olivine and pyroxenes, high
Cr# and low TiO2 of spinel, and refractory characteristics of
clinopyroxene (Figure 7, 8, 9, and 10). Additionally, the result of non-
modal fractional melting shows that harzburgites have experienced over
58
20% degree of partial melting based on compositions of HREE, which are
not variable relative to that of LREE in metasomatism (Shervais & Jean,
2012)
Hydrous fluid not only lowers solidus of peridotite in mantle
wedge but also reacts with the peridotite. There are two possibilities of
metasomatism generated by hydrous fluid: (1) modal metasomatism
which mean production of new hydrous mineral phases such as amphibole
or phlogopite or (2) cryptic metasomatism which supply elements into the
rock from fluid without any new mineral phases (Bizimis et al., 2000;
McCoy-West et al., 2015; Uenver-Thiele et al., 2017).
In the petrography, any minerals generated by metasomatism did
not be observed in studied thin sections except minerals from
hydrothermal alteration such as serpentinization. Although group 1 is the
most depleted sample in analyzed samples, LREE compositions of
clinopyroxenes from the harzburgites are enriched compared with values
of models calculated from matching with HREE compositions (Figure 12).
Thus, peridotites from the Mariana forearc have experienced cryptic
metasomatism arisen from the reaction of hydrous fluid.
Bizimis et al. (2000) suggested re-fertilized hydrous melting,
which means trace element supplies stemmed from cryptic metasomatism
in clinopyroxene. Generally, anhydrous melting models for spinel
59
lherzolite assumed melting proportion of clinopyroxene is much higher
than other phases (Johnson et al., 1990; Johnson & Dick, 1992), so
exhaustion of clinopyroxene might occur in 20–25% melting of the
primitive mantle, lherzolitic peridotite (McDonough & Sun, 1995).
However, clinopyroxene can survive until 29% melting of primitive
mantle because the proportion of clinopyroxene in melts is reduced
(Gaetani & Grove, 1998; Bizimis et al., 2000). Thus, remained
clinopyroxene from hydrous melting might be more depleted than the
remainder from anhydrous one.
In Figure 19, the re-fertilized hydrous melting model from
Bizimis et al. (2000) are applied to the Mariana forearc peridotite.
Anhydrous and hydrous melting is the incremental melting from primitive
mantle because it is assumed that fluid need to be instilled in each
increment. Incremental melting results in the similar output of non-modal
fractional melting (Johnson et al., 1990). Constants for the model referred
to Shervais and Jean (2012) are shown in Table A2. The starting point of
the hydrous melting is presumed at 9% melting of primitive mantle
because 9% melting point of primitive mantle is the average of abyssal
peridotite (Bizimis et al., 2000). Titanium contents from group 1 are as
low as 50 ppm, cannot be explained through anhydrous melting. Despite
highly depleted in Ti, other trace elements are less depleted. This is
60
because cryptic metasomatism of clinopyroxene caused by the addition of
fluid can account for less depletion of other trace elements. Fluid
composition contented with re-fertilization hydrous melting is shown in
Figure 20 with previous literature data (Bizimis et al., 2000). The trace
element compositions in fluid used hydrous melting from this study are
much lower than previous data. This can be attributed to highly refractory
harzburgite which is easily influenced by low concentrations of trace
element in fluid.
A sample of group 1 is highly enriched in La relative to the low
concentration of Ce (366-1498B-7R-4/88–93). This phenomenon is called
as “La-kick” which sometimes takes place in clinopyroxene analyses, but
whether this effect is an artifact from experiment or real feature of data
has not been articulated yet (Warren et al., 2016). Group 2 shows
depletion of Ti in clinopyroxene. However, they have much higher trace
elements in clinopyroxene, especially LREE (Figure 12 and 19). As it will
be discussed in the next section, this enrichment is interpreted by
interaction with melts after hydrous melting.
61
5.2 Melt-mantle interaction
Composition of mantle peridotite can be influenced by migrating
melt, called as reactive porous flow (RPF), melt-rock interaction or
metasomatism (Kelemen et al. 1990, 1995; Dijkstra et al., 2003; Ishimaru
et al., 2007 Seyler et al., 2007; Paquet et al., 2016; Birner et al., 2017).
Mantle interacted with melts are affected in variable aspects. Depending
on orthopyroxene saturation status, interaction with melt generates
different petrographic features (Dijkstra et al., 2003). If melts are saturated
in orthopyroxene, melts precipitate the cusp of orthopyroxenes between
primary mineral grains. If melts are orthopyroxene-undersaturated,
orthopyroxene from mantle experiences incongruent melting process with
the dissolution of orthopyroxene and precipitation of olivine (Dijkstra et
al., 2003; Seyler et al., 2007). In addition, melt-rock interaction affects the
chemical composition of both melt and rock. Incompatible elements in the
residual mantle are enriched by interaction with melt because melt
extraction makes the elements poor in the mantle (Kelemen et al., 1990).
Because clinopyroxene and spinel are the main hosts of trace elements in
peridotite, they are used to identify melt-mantle interaction (Pearce et al.,
2000; Morishita et al., 2007; Choi et al., 2008; Uysal et al., 2012). In this
study, group 2 and 3 are interpreted as results of interaction with melts
later hydrous melting.
62
5.2.1 Interaction with boninitic melt
Boninite is classified by main features: high SiO2 (>53 wt.%), low
TiO2 (<0.6 wt.%), high MgO contents (>7 wt.%), and phenocrysts or
micro phenocrysts of orthopyroxene (Taylor et al., 1994). It also shows
low concentrations of incompatible elements, which means the source of
boninitic melts is highly refractory (Arndt, 2003). In order to melt highly
refractory mantle, hydrous fluid from subduction is necessary. The genesis
of boninitic magmas is known in forearc regions because hydrous fluid
can lower the solidus of the refractory mantle (Crawford et al., 1989;
Taylor et al., 1994; Arndt, 2003). Also, boninite in Izu-Bonin-Mariana arc
has been reported and progressively studied (Umino et al., 2018; Valetich
et al., 2018). Thus, it is reasonable that peridotite from the Mariana
forearc reacted with boninitic melts.
Group 2 has no difference from other harzburgite groups and
peridotites in the chemical composition of the whole rock, but it shows
significant disparities in trace elements in clinopyroxene. Firstly,
petrographic evidence is included in group 2 harzburgite. Incongruent
melting in the mantle, dissolution of orthopyroxene and precipitation of
olivine can occur in interaction orthopyroxene-undersaturated melt
between mantle rocks (Dijkstra et al., 2003; Seyler et al., 2007) and it
produces olivine embayment in orthopyroxene grain or extreme weird
63
boundaries between orthopyroxene and olivine (Figure 4A and B). The
unique textures of incongruent melting, notably, appear in thin sections
from group 2. Furthermore, Clinopyroxene in group 2 has higher Na2O
contents than that in group 1 despite similar Mg# with group 1 (Figure 8).
Sodium enrichment in the mantle is suggested by interaction with
ascending melts (Kepezhinskas & Defant, 1996; Xiong et al., 2006).
Lastly, group 2 is characterized by flat patterns of REE in clinopyroxene
(Figure 11). Flat patterns are interpreted as an enrichment of LREE
relative to HREE from the composition of group 1. The interacted melt
composition in equilibration with clinopyroxene are predicted by the
simple calculation. The melt compositions are able to be restored by
multiplying well-known partition coefficient between the melt and
clinopyroxene in basalts (DCpx from McKenzie & O’Nions, 1991; Johnson,
1994, 1998) and REE contents of clinopyroxene in mantle samples.
Recovered REE patterns of melts are shown in Figure 21 with estimated
compositions of various melts (Elliott et al., 1997; Reagan et al., 2010,
2013). Hypothetic melt compositions interacted with group 2 are depicted
on the boninite area. It confirms that clinopyroxenes from group 2
harzburgite equilibrated with boninitic melts.
Oxygen fugacity of boninite is recorded higher relative to the quartz-
fayalite-magnetite (QFM) buffer because of contributions from the slab
64
(Brounce et al., 2014). Oxygen fugacity of group 2 is comparable to that
of group 1. However, the average value of the group 2 (+0.9 log units
relative to QFM buffer) is still high fO2. High melt/rock ratio can develop
dunite containing spinel with high Cr# as melt conduits or channels
(Matsumoto & Arai, 2001; Birner et al., 2017), and this dunite with high
Cr# spinel is reported to record high fO2 up to QFM +2 or higher Fe3+ in
spinel than harzburgite samples (Parkinson & Arculus, 1999; Birner et al.,
2017). A dunite sample from this study (366-1495B-3G-CC/21–23) might
originate from interaction with boninitic melt in high melt/rock ratio,
whereas group 2 harzburgite might experience interaction with boninitic
melts under low melt/rock ratio condition.
65
Figure 19. Trace element compositions versus Ti contents in clinopyroxene from
the harzburgites. Dashed lines represent anhydrous melting of the starting point
(PUM; McDonough & Sun, 1995) with points of each 5% as small bars. Gray
stars are residues of 9% partial melting and starting points of hydrous melting.
Bold lines indicate re-fertilized hydrous melting suggested by Bizimis et al.
(2000). Dark gray fields are the areas of clinopyroxene from abyssal peridotite
(Warren et al., 2016).
66
Figure 20. Calculated fluid compositions from group 1 harzburgite normalized
by chondrite (McDonough & Sun, 1995). Gray field is trace element
compositions in fluid calculated by Bizimis et al. (2000).
Figure 21. Calculated melt composition normalized by chondrite (McDonough
& Sun, 1995). Melt compositions are predicted by multiplying concentrations of
REE in clinopyroxene and partition coefficients between basalt and
clinopyroxene (see text). Compositions of group Ⅰ are plotted as blue area.
Compositions of boninite and forearc basalt are colored red and gray,
respectively.
67
5.2.2 Interaction with MORB-like melt
Group 3 seems to be similar with group 1 in patterns of REE in
clinopyroxene except for the degree of depletion (Figure 9). It might be
interpreted that group 3 experienced a low degree of melting. However,
this interpretation is insufficient owing to the following reasons. First of
all, the sample in group 3 shows the greatest depletion in major elements
such as Al2O3 and CaO among the harzburgites. According to Niu (1997),
Al2O3 and CaO contents lower than one wt.% were achieved by over 20%
of fractional melting under 25–28 kbar. In addition, Mg# of clinopyroxene
and Cr# in spinel are high values (94 and 60, respectively) with the
understanding result from a high degree of partial melting.
The restoration of the equilibrated melt with clinopyroxene from
group 3 have a difference of various melts in concentration and pattern.
Titanium, however, as representative of an incompatible element in mantle
melting, in clinopyroxene and spinel is enriched despite higher Mg# and
Cr#, indicating high depletion of the mantle (Figure 6 and 8). The content
of Na2O is the highest value in the sample from group 3, which means
interaction with a melt as illustrated above (Section 5.2.1).
A distinctive feature of group 3 which is the high Ti content in
clinopyroxene and spinel implies that group 3 harzburgite could react with
the high-Ti melt. Reagan et al. (2010) found forearc basalt (FAB) similar
68
with MORB-like tholeiitic basalt in the Mariana forearc. It is interpreted
as products by decompression melting at the initial stage of subduction
initiation (Morishita et al., 2011; Reagan et al., 2017). Forearc basalt has
over an order of magnitude higher TiO2 concentrations compared to
boninite (Talyor et al., 1994; Reagan et al., 2010). Therefore, the
harzburgite in group 3 might interact with FAB, but calculated melt from
the REE pattern of group 3 remains unanswered.
5.3 Effect of serpentinization on mantle peridotite
Serpentinization means hydration of ultramafic minerals such as
olivine and orthopyroxene at relatively low temperature; serpentine group
commonly has three polymorphs: a flat crystal structured lizardite and
cylindrical chrysotile as phases at the lower temperature, and antigorite of
the wavy structure as a high-temperature phase (Evans et al., 2013). If
mantle peridotite is serpentinized, the rocks increase volume and decrease
density. Because the vast mass of water is supplied by the subducting slab,
forearc mantle is serpentinized up to 50% (Hyndman & Peacock, 2003).
Serpentinization also plays a principal role in the formation of mud
volcano (Fryer, 2012). Moreover, understanding the serpentinization
process in the forearc contributes to enlightenment about versatile
research such as elemental recycling in subduction system and even origin
69
of life (Kelley et al., 2005; Evans et al., 2013).
According to the experimental study, olivine is mainly
serpentinized at the lower temperature (<250℃), whereas serpentinization
of orthopyroxene is more stable than that of olivine at the higher
temperature (>250℃; Allen & Seyfried, 2003). Most of the samples from
the Mariana forearc have a remarkable feature of mesh-textured
serpentine, a pseudomorph of olivine, indicating serpentinization at the
lower temperature relative to 250℃. Some samples from Conical
seamount has talc and orthopyroxene rimmed by talc (Figure 6;
steatitization; Klein & Bach, 2009). A dunite sample from Conical
seamount (125-780C-10R-1/18–21) has an extraordinary characteristic
compared to other samples from various seamounts. It presents antigorite,
a high-temperature phase of serpentine (300–600℃ in forearc condition;
Evans, 2004) and cleavable olivine grains (Murata et al., 2009). The
presence of the minerals indicates that peridotite has experienced
serpentinization in deeper part of the mantle wedge due to dragging down
to depth by mantle flow (Debret et al., 2018).
Deschamps et al. (2013) suggested that the enrichment of fluid
mobile elements in serpentinized rocks is attributed to fluid/rock
interaction within the subduction channel after the serpentinization
process. However, the transition from olivine or orthopyroxene to
70
serpentine minerals can effect change in the concentration of major
elements. The typical example of serpentinization effect on major element
chemistry is Mg loss because of differences in Mg:Si ratio of each mineral
(Niu, 2004) although marine weathering might account for Mg loss (Snow
and Dick, 1995). Serpentinization, additionally, is a hydration process of
peridotite, so the loss on ignition (LOI) is major contributor to the mass of
serpentinized peridotite (Table 2 and Figure A1).
Serpentinization can influence on fugacities of sulfur and oxygen,
the main factor of sulfide formation. Because sulfide is considered as the
main host of highly siderophile elements including PGE (Simon & Ripley,
2011), discerning behaviors of sulfide in serpentinization is essential in
following the discussion about PGE geochemistry of the Mariana forearc
mantle. Klein and Bach (2009) proposed two stages of serpentinization.
During the initial stage of serpentinization, desulfurization is dominant in
the peridotite, so it makes the S content decrease and sulfide be removed
from the mantle. However, the late stage of serpentinization with
steatitization brings about high silica activities and high S fugacity. Hence,
sulfide can precipitate and the S content increases in the mantle contrary
to the initial stage of serpentinization.
71
5.4 PGE geochemistry in the forearc mantle
5.4.1 Influence of low-T alteration on PGE geochemistry
Before discussion about effects of the magmatic process on PGE
geochemistry in the forearc mantle peridotite, influences of low-
temperature alteration such as serpentinization or weathering on PGE
abundance in the peridotite need to be deliberated. Although the impacts
of serpentinization and weathering on the behavior of PGE in ultramafic
rocks is still poorly constrained, several researchers have agreed that
platinum group elements are unreactive under the reduced condition
during serpentinization (Snow & Schmidt, 1998; Becker et al., 2006; Liu
et al., 2009; Marchesi et al., 2013). Snow and Schmidt (1998) mentioned
that because the degree of serpentinization can control the density of the
peridotite, the concentration of PGE might be changed, but ratios among
PGE would not be altered. Palladium, however, can have mobility with
combining chlorine as chloride in the fluid under the oxidized condition
(Luguet et al., 2003; Wood & Normand, 2008; Liu et al., 2009). The late
stage of serpentinization with steatitization suggested by Klein and Bach
(2009) increases the fugacities of oxygen and sulfur. Thus, samples
observed steatitization, generation of talc, should be considered when
ratios of PGE are interpreted.
In addition, the low-temperature alteration can disturb assemblages
72
of base-metal sulfide and metal alloys, the main hosts of PGE in
ultramafic rocks (Luguet et al., 2004; Klein & Bach et al., 2009; Marchesi
et al., 2013; Foustoukos et al., 2015). Magmatic sulfide assemblage is
composed of predominant pentlandite, chalcopyrite, and pyrrhotite despite
observing the minor abundance of accessory minerals from the
hydrothermal process such as troilite (FeS), magnetite, and native copper
(Luguet et al., 2004). On the other hand, the assemblage of secondary
sulfides displays Fe-rich assemblage, pyrrhotite being dominant with Co–
Ni-rich pentlandite within the serpentine matrix (Luguet et al., 2004).
Desulfurization in the initial stage of serpentinization generates metal-rich
sulfide phase such as heazlewoodite (Ni3S2) or millerite (NiS) and/or
metal alloy such as awaruite (Ni3Fe–Ni2Fe), replacing the primary sulfide
(Klein & Bach, 2009; Foustoukos et al., 2015; Prichard et al., 2017).
Although the studied samples have scarce abundances of sulfide or
alloy phases, identification of small size sulfide or alloy phases (<100μm)
were perform by using Energy Dispersive Spectrometry (EDS) mounted
on Field Emission Scanning Electron Microscope (FE-SEM) in the six
harzburgite samples. The analyzed sulfide assemblages show effects of
desulfurization such as the presence of awaruite, magnetite, or native
copper. However, there are no Fe-rich assemblages in the samples and no
evidence for precipitation of sulfide during the second stage of
73
serpentinization such as presences of millerite, vaesite (NiS2), polydymite
(Ni3S4) in the inspected samples. Hence, it is thought that their
assemblages and PGE alloys might give evidence for discussing the main
hosts of PGE during magmatic processes in the forearc mantle. I will
discuss it in each next section with influences of magmatic processes on
PGE geochemistry in the forearc mantle.
5.4.2 Influence of hydrous melting on PGE geochemistry
The Mariana forearc mantle is considered to have experienced at
least two events of melting, decompression melting and hydrous melting,
so it has highly refractory features (Ishizuka et al., 2011a, 2011b; Reagan
et al., 2010, 2017). During the partial melting events of the mantle, S,
behaving as incompatible elements, escapes from the mantle. The
concentrations of PGE in the mantle increase in accordance with the
increasing degree of partial melting (Rehkämper et al., 1999b). The
calculated concentrations of PGE in the mantle residue after partial
melting have been predicted by Rehkämper et al. (1999a), and the model
is applied for Ir and Pd as representatives of PGE (Figure 22).
It is assumed that the starting value of S is 250 ppm, which means
714 ppm of sulfide and S is exhausted at 25% of melting (Rekämper et al.,
1999a). The melting model is a near-fractional melting model, combining
the concept of batch melting model and retained melts (Rehkämper et al.,
74
1999a). Partition coefficients of PGE between sulfide and silicate melt are
used from Rehkämper et al. (1999a). Contents of Al2O3 of whole rock is
calculated by the same model of PGE with an equation of partition
coefficient for Al2O3 from Niu (1997). Although calculated Al2O3 content
cannot achieve the real values of the Mariana forearc peridotite, it also
means that all peridotites had experienced sulfide exhaustion or near-
sulfide exhaustion.
Group 1 harzburgite, excluding melt-rock interaction, has patterns
that show the effect of hydrous partial melting. Absolute concentrations of
PGE might interrupt precise meaning because of the nugget effect,
heterogeneous distribution of sulfide, alloy, or PGM (micro) nuggets
(Rehkämper et al., 1999b; Dale et al., 2012b; Park et al., 2013). Thus,
ratios of PGE will be mainly discussed in order to avoid the nugget effect
(Figure 23).
Ratios of PdN/IrN and PtN/IrN from group 1 show decreasing trends
after the sulfide exhaustion. After the complete consumption of base-metal
sulfides, PGM, including Ru-Os ± Ir sulfide and PGE alloys, are main
hosts of PGE (Luguet et al., 2007; Dale et al., 2012b). Prichard et al.
(2017) also suggested that the micron size of PGM is formed during
consumption of sulfide liquid. They mentioned that IPGE-alloy and IPGE-
sulfide are created earlier than Pt-(Pd)-PGM. The correlation among IPGE
75
and Rh from the Mariana forearc peridotite are strongly observed except
for some samples detected with alloys together, so it can mean IPGE and
Rh share the host minerals (Figure 15). However, the fact that correlations
between PGE and S in the studied samples are generally poor is the host
mineral of PGE might be alloy phases (Figure 17). Platinum and Ir
concentrations have two trends; one is Pt:Ir = 2:1 in the group 1, the other
one is Pt:Ir = 1:2 in the others. However, Pd and Ir concentration are
scattered, and it means they are hosted by different phases. Thus, the
retention of IPGE sulfide and alloys in the mantle during hydrous melting
is the essential key of PPGE/IPGE fractionation after/near the sulfide
exhaustion (Figure 23).
Also, incongruent melting of sulfide might be one of the vital factors
for PGE fractionation in the mantle during partial melting (Barnes et al.,
2015; Graham et al., 2017). Platinum and Pd have a preference to the
intermediate solid solution (ISS), Cu-rich residual liquid in sulfide melt,
after crystallization of monosulfide solid solution (MSS) around 1000℃
(Dare et al., 2010; Holwell & McDonald, 2010). Hence, influences of
residual alloys and incongruent sulfide melting should be considered to
predict PGE abundance in the mantle after/near the sulfide exhaustion.
76
Figure 22. Concentrations of Pd and Ir against Al2O3 of whole rocks with the
model from Rehkämper et al. (1999a). Simple melting of sulfide (congruent
melting) cannot predict PPGE/IPGE fractionation in the mantle. Partition
coefficients between sulfide melt and silicate melt of PGE are assumed as 10,000
(Ir and Pd are from Rehkämper et al., 1999a). Values of PGE in peridotites from
other studies are plotted together. SSZ – supra-subduction zone; AP – abyssal
peridotite.
77
Figure 23. Chondrite-normalized concentrations ratios of PGE by those of Ir
against Al2O3 of whole rocks with the model from Rehkämper et al. (1999a).
Simple melting of sulfide (congruent melting) cannot predict PPGE/IPGE
fractionation in the mantle. Partition coefficients between sulfide melt and
silicate melt of PGE are assumed as 10,000 (Ir and Pd are from Rehkämper et al.,
1999a). Values of PGE in peridotites from other studies are plotted together. SSZ
– supra-subduction zone; AP – abyssal peridotite.
78
5.4.3 Influence of melt-mantle interaction on PGE
geochemistry
Group 2 and 3 harzburgite samples are interpreted as products by
interaction with boninitic melt and FAB melt, respectively. Luguet and
Reisberg (2016), who reviewed the petrological history of non-cratonic
peridotite xenoliths and its impacts on the behavior of HSE (PGE and Re),
argued that the condition of percolation melt through the mantle is a
crucial factor in behaviors of HSE during interaction with melts. They
contended that whether the melt is saturated with sulfur can have an
utterly different influence on abundance of PGE and Re in the interacted
mantle.
If S-undersaturated melt interacts with mantle peridotites,
concentrations of PGE in the mantle decrease with an unchanged pattern
of PGE (Luguet and Reisberg, 2016). This is because S-undersaturated
melt dissolves the sulfide, which mainly hosts PGE in the mantle. Group 2
harzburgite shows almost the same patterns and concentrations of PGE
with those of group 1 (Figure 14B and 15). The characteristics of boninitic
melt recorded in group 2 correspond to that of boninite studied before. For
example, primitive boninite is S-undersaturated (Hamlyn & Keays, 1985)
and oxygen fugacity of boninite is higher than QFM +1 (Brounce et al.,
2015). Moreover, IPGE minerals such as Ru-pentlandite, Rh-Fe alloy, and
79
Rh-Ir-Ru alloy are detected with several desulfurized sulfide assemblage
(pentlandite, awaruite, and/or magnetite) in EDS analyses for group 2
(Figure A3).
On the contrary, if S-saturated melt is percolated in and interacted
with the mantle, PPGE might be enriched in the mantle (Luguet and
Reisberg, 2016). Sulfur-saturated melt can precipitate Cu–Ni-rich sulfide
in the mantle because Cu–Ni-rich sulfide (ISS) prefer the liquid condition
with silicate melt rather than mantle rocks (Dare et al., 2010; Holwell et
al., 2010). Thus, the S-saturated silicate melt introduces Cu–Ni-rich
sulfide with PPGE, which have a preference to Cu-rich sulfide, into the
interacted mantle.
The abundance of PGE in group 3 is interpreted by the result from
interaction with MORB-like melt (FAB). Hamlyn & Keays (1985) said
MORB is S-saturated and Brounce et al. (2015) also mentioned oxygen
fugacity of MORB and FAB is lower than that of boninite. Strong
evidence of precipitation of Cu-rich sulfide is the completely different
sulfide assemblage from those of group 2 (Figure A4). There is no
pentlandite grain, but there is Cu-rich FeS, implying crystallization of ISS.
Furthermore, high mass fragment on the iron sulfide has Te, which is
more compatible to ISS than MSS and forms PPGE telluride (Pd-Te or Pt-
Te) at late stage of crystallization of sulfide in the sulfide assemblage
80
(Holwell et al., 2010). Therefore, the mantle interacted with MORB-like
basaltic melt from incipient subduction has enriched PPGE abundance
compared to IPGE.
81
6. Conclusion
Mantle peridotite from the mud volcanos (Conical, Asút Tesoru,
Fantangisña, Yinazao, and South Chamorro seamounts) on the Mariana
forearc has highly refractory features in terms of petrography and
geochemistry. The results of clinopyroxene mode are significantly low
(<4.5 vol.%), high Mg# of silicate minerals and high Cr# of spinel, and
highly depleted patterns of REE in clinopyroxene mean the high degree of
partial melting (>20%). Some of the samples have the striking
characteristics of incongruent melting textures, flat patterns of REE in
clinopyroxene, and enrichment of Na2O in clinopyroxene and TiO2 in
spinel, though other highly refractory features explained above. I divided
the samples into three groups of harzburgite, dunite, and serpentinite.
Group 1 harzburgite is representative of the highly refractory
mantle in the Mariana forearc. Hydrous fluid derived from the slab
enables much higher degrees of partial melting by lowering solidus of
mantle than abyssal peridotite by decompression melting, so products of
arc magmatism such as boninite have more refractory features than
MORB. Trace element contents of clinopyroxene from group 1
harzburgite record evidence of hydrous melting, enrichment of trace
elements from the hydrous fluid with no enrichment of Ti.
Group 2 and 3 harzburgite have different REE patterns of
82
clinopyroxene. Group 2 harzburgite experienced interaction with boninitic
melt after hydrous melting because of following reasons: (1) texture of
incongruent melting, olivine embayment in orthopyroxene, (2) enrichment
of Na2O and LREE contents in clinopyroxene, (3) restoration of
equilibrated melt compositions, and (4) similar oxygen fugacity with
group 1 and boninite. Although group 3 can seem to experience only low
degrees of partial melting in an aspect of the REE pattern from the
clinopyroxene, there is evidence of highly partial melting and enrichment
by the interaction of melt. Low concentration of incompatible major
elements (Al2O3 and CaO), high Mg# in silicate minerals, and high Cr# in
spinels indicates high degree of partial melting, whereas notable
enrichments of Na2O in clinopyroxene and TiO2 in spinel and uniquely
reduced condition display interaction with MORB-like melt such as FAB,
able to be generated in the very beginning stage of subduction initiation.
Influences of highly refractory feature induced by hydrous
melting on PGE geochemistry are a change of the PGE host mineral from
base-metal sulfide to PGM, including PGE sulfide such as laurite and/or
PGE alloys. Fractionation between PPGE and IPGE might occur during
the change of the host mineral by incongruent sulfide melting compared to
the fractionation in abyssal peridotite.
Melt interaction exerts contrary influence on abundance of PGE
83
in mantle peridotites by characteristics of melts. Sulfur-undersaturated
melt such as boninitic melt leads to decrease of PGE contents in the
interacted mantle, while interaction with S-saturated melt such as MORB-
like melt contributes precipitation of Cu–Ni-rich sulfide, which includes
more PPGE than IPGE into the mantle. Thus, interaction with MORB-like
melt yields PPGE enriched features different from highly refractory
forearc mantle.
84
Figure A1. A plot of degrees of serpeninization by point counting method
against LOI (wt.%). Percentage of water in talc mineral (5 wt.%) is lower than
that of serpentine minerals (13 wt.%).
Appendix Figures
85
Figure A2. Ternary diagram of spinel groups suggested by Stevens (1944). Ferric
iron content is significantly low in the Mariana forearc peridotites. They are
classified as Al-chromite or Cr-spinel, but the petrological or petrographic studies
of the mantle have used the term of “spinel”.
86
Figure A3. Back scattered images of sulfide and alloy assemblages from group Ⅱ. (A-C) are from 366-1498B-13R-3/21–26, and (D-E) are
from 366-1498B-13R-3/103–106. Abbreviations are from Whitney & Evans (2010). Pn = pentlandite, Awr = Awaruite, Cct = chalcocite, Hem
= hematite, Spl = spinel
87
Figure A4. A back scattered image of sulfide and alloy assemblage from group Ⅲ
(195-1200A-6R-2/21–25). Intermediate solid solution (ISS) is detected with Te.
Abbreviations are the same as Figure A3, Mag = magnetite, and Po = pyrrhotite.
88
Table A1. Compositions of PGE in spinel grains from the Mariana forearc
peridotite with averages of detection limit in each analysis
Appendix Tables
89
Table A2. Constants used in the partial melting models in the text
The values are from the appendix of Bizimis et al., (2000) and Shervais & Jean
(2012).
90
References
Agranier A., Lee C. T. A., Li Z. X. A. and Leeman W. P. (2007) Fluid-
mobile element budgets in serpentinized oceanic lithospheric mantle:
Insights from B, As, Li, Pb, PGEs and Os isotopes in the Feather River
Ophiolite, California. Chem. Geol. 245, 230–241.
Alexander V. Sobolev, Albrecht W. Hofmann, Dmitry V. Kuzmin, Gregory
M. Yaxley, Nicholas T. Arndt, Sun-Lin Chung, Leonid V.
Danyushevsky, Tim Elliott, Frederick A. Frey, Michael O. Garcia,
Andrey A. Gurenko, Vadim S. Kamenetsky, Andrew C. Kerr,
Nadezhda A M. T., Sobolev A. V, Hofmann A. W., Kuzmin D. V,
Yaxley G. M., Arndt N. T., Chung S.-L. S.-L., Danyushevsky L. V,
Elliott T., Frey F. a, Garcia M. O., Gurenko A. a, Kamenetsky V. S.,
Kerr A. C., Krivolutskaya N. a, Matvienkov V. V, Nikogosian I. K.,
Rocholl A., Sigurdsson I. a, Sushchevskaya N. M. and Teklay M.
(2007) The Amount of Recycled Crust in Sources of Mantle-Derived
Melts. Science. 316, 412–417.
Allen D. E. and Seyfried W. E. (2003) Compositional controls on vent
fluids from ultramafic-hosted hydrothermal systems at mid-ocean
ridges: An experimental study at 4 00°C, 500 bars. Geochim.
Cosmochim. Acta 67, 1531–1542.
Arai S. (1994) Characterization of spinel peridotites by olivine-spinel
compositional relationships: Review and interpretation. Chem. Geol.
113, 191–204.
Arculus R. J., Ishizuka O., Bogus K. A., Gurnis M., Hickey-Vargas R.,
Aljahdali M. H., Bandini-Maeder A. N., Barth A. P., Brandl P. A.,
Drab L., Do Monte Guerra R., Hamada M., Jiang F., Kanayama K.,
Kender S., Kusano Y., Li H., Loudin L. C., Maffione M., Marsaglia K.
M., McCarthy A., Meffre S., Morris A., Neuhaus M., Savov I. P., Sena
C., Tepley F. J., Van Der Land C., Yogodzinski G. M. and Zhang Z.
(2015) A record of spontaneous subduction initiation in the Izu-Bonin-
Mariana arc. Nat. Geosci. 8, 728–733.
Arndt N. (2003) Komatiites, kimberlites, and boninites. J. Geophys. Res.
Solid Earth 108, 1–11.
Bach W., Garrido C. J., Paulick H., Harvey J. and Rosner M. (2004)
Seawater-peridotite interactions: First insights from ODP Leg 209,
MAR 15°N. Geochemistry, Geophys. Geosystems 5.
91
Ballhaus C., Berry R. f. and Green D. H. (1991) High pressure experimental
calibration of the olivine-orthopyroxene-spinel oxygen geobarometer:
implications for the oxidation state of the upper mantle . Contrib. to
Mineral. Petrol. 107, 27–40.
Barnes S. J., Mungall J. E. and Maier W. D. (2015) Platinum group
elements in mantle melts and mantle samples. Lithos 232, 395–417.
Becker H., Horan M. F., Walker R. J., Gao S., Lorand J. P. and Rudnick R.
L. (2006) Highly siderophile element composition of the Earth’s
primitive upper mantle: Constraints from new data on peridotite
massifs and xenoliths. Geochim. Cosmochim. Acta 70, 4528–4550.
Birner S. K., Warren J. M., Cottrell E., Davis F. A., Kelley K. A. and
Falloon T. J. (2017) Forearc peridotites from tonga record
heterogeneous oxidation of the mantle following subduction initiation.
J. Petrol. 58, 1755–1780.
Bizimis M., Salters V. J. M. and Bonatti E. (2000) Trace and REE content
of clinopyroxenes from supra-subduction zone peridotites. Implications
for melting and enrichment processes in island arcs. Chem. Geol. 165,
67–85.
Wood, B. J., Bryndzia, L. T. and Johnson K. E. (1990) Mantle oxidation
state and its relationship to tectonic environment and fluid speciation.
Science. 248, 337–345.
Brounce M. N., Kelley K. A. and Cottrell E. (2014) Variations in Fe3+/∑Fe
of Mariana Arc Basalts and MantleWedge fO2. J. Petrol. 55, 2514–
2536.
Brounce M., Kelley K. A., Cottrell E. and Reagan M. K. (2015) Temporal
evolution of mantle wedge oxygen fugacity during subduction
initiation. Geology 43, 775–778.
Buchl A., Brugmann G., Batanova V. G., Munker C. and Hofmann A. W.
(2002) Melt percolation monitored by Os isotopes and HSE
abundances: a case study from the mantle section of the Troodos
Ophiolite. Earth Planet. Sci. Lett. 204, 385–402.
Choi S. H., Shervais J. W. and Mukasa S. B. (2008) Supra-subduction and
abyssal mantle peridotites of the Coast Range ophiolite, California.
Contrib. to Mineral. Petrol. 156, 551–576.
92
Ciazela J., Koepke J., Dick H. J. B., Botcharnikov R., Muszynski A.,
Lazarov M., Schuth S., Pieterek B. and Kuhn T. (2018) Sulfide
enrichment at an oceanic crust-mantle transition zone: Kane
Megamullion (23°N, MAR). Geochim. Cosmochim. Acta 230, 155–189.
Crawford A. J., Falloon T. J. and Green D. H. (1989) Classification,
petrogenesis and tectonic setting of boninites. Boninites Relat. Rocks,
2–49.
Crossley R. J., Evans K. A., Jeon H. and Kilburn M. R. (2018) Insights into
sulfur cycling at subduction zones from in-situ isotopic analysis of
sulfides in high-pressure serpentinites and ‘hybrid’ samples from
Alpine Corsica. Chem. Geol. 493, 359–378.
Dale C. W., Macpherson C. G., Pearson D. G., Hammond S. J. and Arculus
R. J. (2012) Inter-element fractionation of highly siderophile elements
in the Tonga Arc due to flux melting of a depleted source. Geochim.
Cosmochim. Acta 89, 202–225.
Dale C. W., Burton K. W., Greenwood R. C., Gannoun A., Wade J., Wood
B. J. and Pearson D. G. (2012) Late Accretion on the Earliest
Planetesimals Revealed by the Highly Siderophile Elements. Science.
247, 72–75.
D’Antonio M. and Kristensen M. B. (2004) Serpentine and brucite of
ultramafic clasts from the South Chamorro Seamount (Ocean Drilling
Program Leg 195, Site 1200): inferences for the serpentinization of the
Mariana forearc mantle. Mineral. Mag. 68, 887–904.
Dare S. A. S., Barnes S. J. and Prichard H. M. (2010) The distribution of
platinum group elements (PGE) and other chalcophile elements among
sulfides from the Creighton Ni-Cu-PGE sulfide deposit, Sudbury,
Canada, and the origin of palladium in pentlandite. Miner. Depos. 45,
765–793.
Debret B., Albers E., Walter B., Price R., Barnes J. D., Beunon H., Facq S.,
Gillikin D. P., Mattielli N. and Williams H. (2018) Shallow forearc
mantle dynamics and geochemistry: New insights from the IODP
expedition 366. Lithos, publishing.
Deschamps F., Godard M., Guillot S. and Hattori K. (2013) Geochemistry
of subduction zone serpentinites: A review. Lithos 178, 96–127.
93
Dijkstra A. H., Barth M. G., Drury M. R., Mason P. R. D. and Vissers R. L.
M. (2003) Diffuse porous melt flow and melt-rock reaction in the
mantle lithosphere at a slow-spreading ridge: A structural petrology
and LA-ICP-MS study of the Othris Peridotite Massif (Greece).
Geochemistry, Geophys. Geosystems 4.
Dilek Y. and Furnes H. (2011) Ophiolite genesis and global tectonics:
Geochemical and tectonic fingerprinting of ancient oceanic lithosphere.
Bull. Geol. Soc. Am. 123, 387–411.
Dlingan M. A. (1979) Bastite pseudomorphs after orthopyroxene,
clinopyroxene and Tremolite. Can. Mineral. 17, 729–740.
Elliott T., Plank T., Zindler A., White W. and Bourdon B. (1997) arc
Guguan have and Agrigan have the least negative systematics. J.
geophy 102, 14,991-15,018.
Evans B. W. (2004) The Serpentinite Multisystem Revisited: Chrysotile Is
Metastable. Int. Geol. Rev. 46, 479–506.
Evans B. W., Hattori K. and Baronnet A. (2013) Serpentinite: What, why,
where? Elements 9, 99–106.
Foustoukos D. I., Bizimis M., Frisby C. and Shirey S. B. (2015) Redox
controls on Ni-Fe-PGE mineralization and Re/Os fractionation during
serpentinization of abyssal peridotite. Geochim. Cosmochim. Acta 150,
11–25.
Fryer P., Ambos E. L. and Hussong D. M. (1985) Origin and emplacement
of Mariana forearc seamounts ( Pacific). Geology 13, 774–777.
Fryer P., Wheat C. G. and Mottl M. J. (1999) Mariana blueschist mud
volcanism: Implications for conditions within the subduction zone.
Geology 27, 103–106.
Fryer P., Wheat C. G., Williams T., Albers E., Bekins B., Debret B. P. R.,
Deng J., Dong Y., Eickenbusch P., Frery E. A., Ichiyama Y., Johnson
K., Johnston R. M., Kevorkian R. T., Kurz W., Magalhaes V.,
Mantovanelli S. S., Menapace W., Menzies C. D., Michibayashi K.,
Moyer C. L., Mullane K. K., Park J.-W., Price R. E., Ryan J. G.,
Shervais J. W., Sissmann O. J., Suzuki S., Takai K., Walter B. and
Zhang R. (2018) Expedition 366 summary. Proc. Int. Ocean Discov.
Progr. 366, 1–23.
94
Fryer P. (2012) Serpentinite Mud Volcanism: Observations, Processes, and
Implications. Ann. Rev. Mar. Sci. 4, 345–373.
Fryer P. (1996) Evolution of the Mariana convergent plate margin system.
Rev. Geophys. 34, 89–125.
Fryer P. (1992) Mud Volcanoes of the Marianas. Sci. Am. 266, 26–46.
Fryer P., Lockwood J. P., Becker N., Phipps S. and Todd C. S. (2000)
Significance of serpentine mud volcanism in convergent margins.
Ophiolites Ocean. crust; new insights from F. Stud. Ocean Drill. Progr.
349, 35–51.
Glenn A. G. and Timothy L. G. (1998) The influence of water on melting of
mantle peridotite. Contrib. to Mineral. Petrol. 131, 323–346.
Gonzalez J., Mao X. L., Roy J., Mao S. S. and Russo R. E. (2002)
Comparison of 193, 213 and 266 nm laser ablation ICP-MS. J. Anal. At.
Spectrom. 17, 1108–1113.
Graham S. D., Holwell D. A., McDonald I., Jenkin G. R. T., Hill N. J.,
Boyce A. J., Smith J. and Sangster C. (2017) Magmatic Cu-Ni-PGE-
Au sulfide mineralisation in alkaline igneous systems: An example
from the Sron Garbh intrusion, Tyndrum, Scotland. Ore Geol. Rev. 80,
961–984.
Hamlyn P. R. and Keays R. R. (1986) Sulfur Saturation and Second-Stage
Melts: Application to the Bushveld Platinum Metal Deposits. Econ.
Geol. 81, 1431–1445.
Hickey-Vargas R. (2005) Basalt and tonalite from the Amami Plateau,
northern West Philippine Basin: New early cretaceous ages and
geochemical results, and their petrologic and tectonic implications. Isl.
Arc 14, 653–665.
Hofmann A. W. (1997) Mantle geochemistry: The message from oceanic
volcanism. Nature 385, 218–229.
Holwell D. A. and Mcdonald I. (2010) A review of the behaviour of
platinum group elements within natural magmatic sulfide ore systems.
Platin. Met. Rev. 54, 26–36.
Hyndman R. D. and Peacock S. M. (2003) Serpentinization of the forearc
mantle. Earth Planet. Sci. Lett. 212, 417–432.
95
Ishii T., Robinson P. T., Maekawa H. and Fiske R. (1992) 27. Petrological
studies of peridotites from diapiric serpentinite seamounts in the Izu-
Ogasawara-Mariana forearc, Leg 125. Proceddings Ocean Drill.
Program, Sci. Results 125, 445-.
Ishimaru S., Arai S., Ishida Y., Shirasaka M. and Okrugin V. M. (2007)
Melting and multi-stage metasomatism in the mantle wedge beneath a
frontal arc inferred from highly depleted peridotite xenoliths from the
avacha volcano, Southern Kamchatka. J. Petrol. 48, 395–433.
Ishizuka O., Kimura J. I., Li Y. B., Stern R. J., Reagan M. K., Taylor R. N.,
Ohara Y., Bloomer S. H., Ishii T., Hargrove U. S. and Haraguchi S.
(2006) Early stages in the evolution of Izu-Bonin arc volcanism: New
age, chemical, and isotopic constraints. Earth Planet. Sci. Lett. 250,
385–401.
Ishizuka O., Tani K. and Reagan M. K. (2014) Izu-Bonin-Mariana forearc
crust as a modern ophiolite analogue. Elements 10, 115–120.
Ishizuka O., Tani K., Reagan M. K., Kanayama K., Umino S., Harigane Y.,
Sakamoto I., Miyajima Y., Yuasa M. and Dunkley D. J. (2011a) The
timescales of subduction initiation and subsequent evolution of an
oceanic island arc. Earth Planet. Sci. Lett. 306, 229–240.
Ishizuka O., Taylor R. N., Yuasa M. and Ohara Y. (2011b) Making and
breaking an island arc: A new perspective from the Oligocene Kyushu-
Palau arc, Philippine Sea. Geochemistry, Geophys. Geosystems 12.
Jarrard R. D. (2003) Subduction fluxes of water, carbon dioxide, chlorine,
and potassium. Geochemistry, Geophys. Geosystems 4.
Johnson K. T. M. (1994) Experimental cpx/and garnet/melt partitioning of
REE and other trace elements at high pressures: petrogenetic
implications. Mineralogical Magazine 6, 454–455.
Johnson K. T. M. (1998) Experimental determination of partition
coefficients for rare earth and high-field-strength elements between
clinopyroxene, garnet, and basaltic melt at high pressures. Contrib. to
Mineral. Petrol. 133, 60–68.
Johnson, Kevin T.M., Dick H. J. B. (1992) Open System Melting and
Temporal and Spatial Variation of Peridotite and Basalt at the Atlantis
II Fracture Zone. J. Geophys. Res. 97, 9219–9241.
96
Johnson K., Dick J. B. and Shimizu N. (1990) Melting in the Oceanic Upper
Mantle: An Ion Microprobe Study of Diopsides in Abyssal Peridotites.
J. Geophys. Res. 95, 2661–2678.
Kelemen, P. B., Johnson, K. T. M., Kinzler, R. J., Irving A. J. (1990) High-
field-strength element depletions in arc basalts due to mantle-magma
interaction. Nature 345, 521–524.
Kelemen P. B., Hirth G., Shimizu N., Spiegelman M. and Dick H. J. B.
(1997) A review of melt migration processes in the adiabatically
upwelling mantle. Philos. Trans. R. Soc. London Ser. A 355, 283–318.
Kelley D. S., Karson J. A., Fru G. L., Yoerger D. R., Shank T. M.,
Butterfield D. A., Hayes J. M., Schrenk M. O., Olson E. J.,
Proskurowski G., Jakuba M., Bradley A. S., Larson B., Ludwig K.,
Glickson D., Buckman K., Brazelton W. J., Roe K., Bernasconi S. M.,
Elend M. J., Lilley M. D., Baross J. a, Summons R. E. and Sylva S. P.
(2005) R Esearch a Rticles. Science 307, 1428–1434.
Kepezhinskas P. and Defant M. J. (1996) Contrasting styles of mantle
metasomatism above subduction zones: Constraints from ultramafic
xenoliths in Kamchatka. Geophys. Monogr. Ser. 96, 307–314.
Kepezhinskas P., Defant M. J. and Widom E. (2002) Abundance and
distribution of PGE and Au in the island-arc mantle: Implications for
sub-arc metasomatism. Lithos 60, 113–128.
Klein F. and Bach W. (2009) Fe-Ni-Co-O-S phase relations in peridotite-
seawater interactions. J. Petrol. 50, 37–59.
Kodolányi J., Pettke T., Spandler C., Kamber B. S. and Ling K. G. (2012)
Geochemistry of ocean floor and fore-arc serpentinites: Constraints on
the ultramafic input to subduction zones. J. Petrol. 53, 235–270.
Laboratories B. and Road M. (1991) Partial Melt Distributions from
Inversion of Rare Earth Element Concentrations. J. Petrol. 32, 1021–
1091.
Lee C. T. A. (2002) Platinum-group element geochemistry of peridotite
xenoliths from the Sierra Nevada and the Basin and Range, California.
Geochim. Cosmochim. Acta 66, 3987–4005.
Leng W. and Gurnis M. (2015) Subduction initiation at relic arcs. Geophys.
Res. Lett. 42, 7014–7021.
97
Liang Y. and Liu B. (2018) Stretching chemical heterogeneities by melt
migration in an upwelling mantle: An analysis based on time-
dependent batch and fractional melting models. Earth Planet. Sci. Lett.
498, 275–287.
Liu C. Z., Snow J. E., Brügmann G., Hellebrand E. and Hofmann A. W.
(2009) Non-chondritic HSE budget in Earth’s upper mantle evidenced
by abyssal peridotites from Gakkel ridge (Arctic Ocean). Earth Planet.
Sci. Lett. 283, 122–132.
Lockwood J. P. (1972) Possible Mechanisms for the Emplacement of
Alpine-Type Serpentinite. Stud. Earth Sp. Sci.132, 273–287.
Lorand J. P., Luguet A., Alard O., Bezos A. and Meisel T. (2008)
Abundance and distribution of platinum-group elements in orogenic
lherzolites; a case study in a Fontete Rouge lherzolite (French
Pyrénées). Chem. Geol. 248, 174–194.
Luguet A, Lorand J-P and Seyler M (2003) Sulfide petrology and highly
siderophile element geochemistry of abyssal peridotites: A coupled
study of samples from the Kane Fracture Zone (45°W 23°20N MARK
Area, Atlantic Ocean). Geochim. Cosmochim. Acta 67, 1553–1570.
Luguet A., Lorand J. P., Alard O. and Cottin J. Y. (2004) A multi-technique
study of platinum group element systematic in some Ligurian ophiolitic
peridotites, Italy. Chem. Geol. 208, 175–194.
Luguet A. and Reisberg L. (2016) Highly Siderophile Element and 187Os
Signatures in Non-cratonic Basalt-hosted Peridotite Xenoliths:
Unravelling the Origin and Evolution of the Post-Archean Lithospheric
Mantle. Rev. Mineral. Geochemistry 81, 305–367.
Luguet A., Shirey S. B., Lorand J. P., Horan M. F. and Carlson R. W.
(2007) Residual platinum-group minerals from highly depleted
harzburgites of the Lherz massif (France) and their role in HSE
fractionation of the mantle. Geochim. Cosmochim. Acta 71, 3082–3097.
Marchesi C., Garrido C. J., Harvey J., González-Jiménez J. M., Hidas K.,
Lorand J. P. and Gervilla F. (2013) Platinum-group elements, S, Se and
Cu in highly depleted abyssal peridotites from the Mid-Atlantic Ocean
Ridge (ODP Hole 1274A): Influence of hydrothermal and magmatic
processes. Contrib. to Mineral. Petrol. 166, 1521–1538.
98
Matsumoto I. and Arai S. (2001) Morphological and chemical variations of
chromian spinel in dunite-harzburgite complexes from the Sangun zone
(SW Japan): Implications for mantle/melt reaction and chromitite
formation processes. Mineral. Petrol. 73, 305–323.
McCoy-West A. J., Bennett V. C., O’Neill H. S. C., Hermann J. and Puchtel
I. S. (2014) The interplay between melting, refertilization and
carbonatite metasomatism in off-cratonic lithospheric mantle under
Zealandia: An integrated major, trace and platinum group element
study. J. Petrol. 56, 563–604.
McDonough S. and (1995) The composition of the Earth. Ghemical Geol.
2541.
McInnes B. I. A. (1999) Erratum: Osmium isotope constraints on ore metal
recycling in subduction zones (Science (October 15, 1999) (512)).
Science. 286, 1681.
Mechanics R., Fault N. H., Sci E. and Fault W. G. (2004) Letters To Nature.
Nature 427, 727–730.
Morishita T., Maeda J., Miyashita S., Kumagai H., Matsumoto T. and Dick
H. J. B. (2007) Petrology of local concentration of chromian spinel in
dunite from the slow-spreading Southwest Indian Ridge. Eur. J.
Mineral. 19, 871–882.
Morishita T., Tani K., Shukuno H., Harigane Y., Tamura A., Kumagai H.
and Hellebrand E. (2011) Diversity of melt conduits in the Izu-Bonin-
Mariana forearc mantle: Implications for the earliest stage of arc
magmatism. Geology 39, 411–414.
Murata K., Maekawa H., Yokose H., Yamamoto K., Fujioka K., Ishii T.,
Chiba H. and Wada Y. (2009) Significance of serpentinization of
wedge mantle peridotites beneath Mariana forearc, western Pacific.
Geosphere 5, 90–104.
Nita B., Maurya S. and Montagner J.-P. (2016) Geochemistry, Geophysics,
Geosystems. Geochemistry Geophys. Geosystems 17, 2825–2834.
Niu Y. (1997) Mantle Melting and Melt Extraction Processes beneath Ocean
Ridges: Evidence from Abyssal Peridotites. J. Petrol. 38, 1047–1074.
99
Niu Y. (2004) Bulk-rock major and trace element compositions of abyssal
peridotites: Implications for mantle melting, melt extraction and post-
melting processes beneath Mid-Ocean ridges. J. Petrol. 45, 2423–2458.
Clift, P., & Vannucchi, P. (2004). Controls on tectonic accretion versus
erosion in subduction zones: Implications for the origin and recycling
of the continental crust. Reviews of Geophysics, 42(2).
Okino K., Kasuga S. and Ohara Y. (1998) A new scenario of the Parece
Vela Basin genesis. Mar. Geophys. Res. 20, 21–40.
Park J. W., Campbell I. H. and Arculus R. J. (2013) Platinum-alloy and
sulfur saturation in an arc-related basalt to rhyolite suite: Evidence
from the Pual Ridge lavas, the Eastern Manus Basin. Geochim.
Cosmochim. Acta 101, 76–95.
Park J. W., Campbell I. H., Kim J. and Moon J. W. (2015) The Role of Late
Sulfide Saturation in the Formation of a Cu- and Au-rich Magma:
Insights from the Platinum Group Element Geochemistry of Niuatahi–
Motutahi Lavas, Tonga Rear Arc. J. Petrol. 56, 59–81.
Park J. W., Campbell I. H., Malaviarachchi S. P. K., Cocker H., Hao H. and
Kay S. M. (2018) Chalcophile element fertility and the formation of
porphyry Cu ± Au deposits. Miner. Depos.
Parkinson I. J. and Arculus R. J. (1999) The redox state of subduction
zones: Insights from arc-peridotites. Chem. Geol. 160, 409–423.
Parkinson I. J. and Pearce J. A. (1998) Peridotites from the Izu-Bonin-
Mariana forearc (ODP Leg 125): evidence for mantle melting and melt-
mantle interaction in a supra-subduction zone setting. J. Petrol. 39,
1577–1618.
Pearce J. A., Barker P. F., Edwards S J., Parkinson I J. and Leat P. T. (2000)
Geochemistry and tectonic significance of peridotites from the South
Sandwich arc–basin system, South Atlantic. Contrib. to Mineral.
Petrol. 139, 36–53.
Prichard H. M., Barnes S. J., Dale C. W., Godel B., Fisher P. C. and Nowell
G. M. (2017) Paragenesis of multiple platinum-group mineral
populations in Shetland ophiolite chromitite: 3D X-ray tomography
and in situ Os isotopes. Geochim. Cosmochim. Acta 216, 314–334.
100
Rampone E., Piccardo G. B. and Hofmann A. W. (2008) Multi-stage melt-
rock interaction in the Mt. Maggiore (Corsica, France) ophiolitic
peridotites: Microstructural and geochemical evidence. Contrib. to
Mineral. Petrol. 156, 453–475.
Reagan M. K., Ishizuka O., Stern R. J., Kelley K. A., Ohara Y., Blichert-
Toft J., Bloomer S. H., Cash J., Fryer P., Hanan B. B., Hickey-Vargas
R., Ishii T., Kimura J. I., Peate D. W., Rowe M. C. and Woods M.
(2010) Fore-arc basalts and subduction initiation in the Izu-Bonin-
Mariana system. Geochemistry, Geophys. Geosystems 11, 1–17.
Reagan M. K., McClelland W. C., Girard G., Goff K. R., Peate D. W.,
Ohara Y. and Stern R. J. (2013) The geology of the southern Mariana
fore-arc crust: Implications for the scale of Eocene volcanism in the
western Pacific. Earth Planet. Sci. Lett. 380, 41–51.
Reagan M. K., Pearce J. A., Petronotis K., Almeev R. R., Avery A. J.,
Carvallo C., Chapman T., Christeson G. L., Ferré E. C., Godard M.,
Heaton D. E., Kirchenbaur M., Kurz W., Kutterolf S., Li H., Li Y.,
Michibayashi K., Morgan S., Nelson W. R., Prytulak J., Python M.,
Robertson A. H. F., Ryan J. G., Sager W. W., Sakuyama T., Shervais J.
W., Shimizu K. and Whattam S. A. (2017) Subduction initiation and
ophiolite crust: new insights from IODP drilling. Int. Geol. Rev. 59,
1439–1450.
Rehkämper M., Halliday A. N., Alt J., Fitton J. G., Zipfel J. and Takazawa
E. (1999) Non-chondritic platinum-group element ratios in oceanic
mantle lithosphere: Petrogenetic signature of melt percolation? Earth
Planet. Sci. Lett. 172, 65–81.
Rehkämper M., Halliday A. N., Fitton J. G., Lee D. C., Wieneke M. and
Arndt N. T. (1999) Ir, Ru, Pt, and Pd in basalts and komatiites: New
constraints for the geochemical behavior of the platinum-group
elements in the mantle. Geochim. Cosmochim. Acta 63, 3915–3934.
Saboda K. L., Fryer P. and Maekawa H. (1992) Petrography of Ultramafic
Clasts. Proc. Ocean Drill. Program, Sci. Results 125, 431–443.
Savov I. P., Ryan J. G., D’Antonio M. and Fryer P. (2007) Shallow slab
fluid release across and along the Mariana arc-basin system: Insights
from geochemistry of serpentinized peridotites from the Mariana fore
arc. J. Geophys. Res. Solid Earth 112.
101
Savov I. P., Ryan J. G., D’Antonio M., Kelley K. and Mattie P. (2005)
Geochemistry of serpentinized peridotites from the Mariana Forearc
Conical Seamount, ODP Leg 125: Implications for the elemental
recycling at subduction zones. Geochemistry, Geophys. Geosystems 6,
1–24.
Schwarzenbach E. M., Gazel E. and Caddick M. J. (2014) Hydrothermal
processes in partially serpentinized peridotites from Costa Rica:
evidence from native copper and complex sulfide assemblages. Contrib.
to Mineral. Petrol. 168, 1–21.
Scott S. R., Sims K. W. W., Frost B. R., Kelemen P. B., Evans K. A. and
Swapp S. M. (2017) On the hydration of olivine in ultramafic rocks:
Implications from Fe isotopes in serpentinites. Geochim. Cosmochim.
Acta 215, 105–121.
Seyler M., Lorand J. P., Dick H. J. B. and Drouin M. (2007) Pervasive melt
percolation reactions in ultra-depleted refractory harzburgites at the
Mid-Atlantic Ridge, 15° 20′N: ODP Hole 1274A. Contrib. to Mineral.
Petrol. 153, 303–319.
Shervais J. W. and Choi S. H. (2012) Subduction initiation along transform
faults: The proto-Franciscan subduction zone. Lithosphere 4, 484–496.
Shervais J. W. and Jean M. M. (2012) Inside the subduction factory:
Modeling fluid mobile element enrichment in the mantle wedge above
a subduction zone. Geochim. Cosmochim. Acta 95, 270–285.
Simon A. C. and Ripley E. M. (2011) The Role of Magmatic Sulfur in the
Formation of Ore Deposits. Rev. Mineral. Geochemistry 73, 513–578.
Snow J. E. and Dick H. J. B. (1995) Pervasive magnesium loss by marine
weathering of peridotite. Geochim. Cosmochim. Acta 59, 4219–4235.
Stern R. J., Reagan M., Ishizuka O., Ohara Y. and Whattam S. (2012) To
understand subduction initiation, study forearc crust: To understand
forearc crust, study ophiolites. Lithosphere 4, 469–483.
Stern, R. J., Fouch, M. J., & Klemperer, S. L. (2004). An overview of the
Izu‐Bonin‐Mariana subduction factory. Inside the subduction factory,
138, 175-222.
Stern R. J. and Gerya T. (2017) Subduction initiation in nature and models:
A review. Tectonophysics.
102
Stern R. J., Klemperer S. L. and Fouch M. J. (2004) An overview of the izu-
bonin-mariana subduction factory. Geophys. Monogr. Ser. 138, 175–
222.
Stevens R. E. (1944) Composition of some chromites of the western
hemisphere. J. Mineral. Soc. Am. 29, 1–34.
Takahashi N., Kodaira S., Klemperer S. L., Tatsumi Y., Kaneda Y. and
Suyehiro K. (2007) Crustal structure and evolution of the Mariana
intra-oceanic island arc. Geology 35, 203–206.
Taylor R. N., Nesbitt R. W., Vidal P., Harmon R. S., Auvray B. and
Croudace I. W. (1994) Mineralogy, chemistry, and genesis of the
boninite series volcanics, Chichijima, Bonin Islands, Japan. J. Petrol.
35, 577–617.
Uenver-Thiele L., Woodland A. B., Seitz H. M., Downes H. and Altherr R.
(2017) Metasomatic processes revealed by trace element and redox
signatures of the lithospheric mantle beneath the Massif Central,
France. J. Petrol. 58, 395–422.
Umino S., Kanayama K., Kitamura K., Tamura A., Ishizuka O., Senda R.
and Arai S. (2018) Did boninite originate from the heterogeneous
mantle with recycled ancient slab? Isl. Arc 27, 1–16.
Ulmer G. C. (1974) Alteration of chromite during Serpentinization in the
Pennsylvania-Maryland District. Am. Mineral. 59, 1236–1241.
Uysal I., Ersoy E. Y., Karslĩ O., Dilek Y., Sadĩklar M. B., Ottley C. J.,
Tiepolo M. and Meisel T. (2012) Coexistence of abyssal and ultra-
depleted SSZ type mantle peridotites in a Neo-Tethyan Ophiolite in
SW Turkey: Constraints from mineral composition, whole-rock
geochemistry (major-trace-REE-PGE), and Re-Os isotope systematics.
Lithos 132–133, 50–69.
Valetich M. J., Mavrogenes J., Arculus R. and Umino S. (2018) Evolution
of chalcophile elements in the magmas of the Bonin Islands. Chem.
Geol., 1–16.
Walter M. J. (2013) Melt Extraction and Compositional Variability in
Mantle Lithosphere. Treatise on Geochemistry 2nd Edition. 393–419.
Warren J. M. (2016) Global variations in abyssal peridotite compositions.
Lithos 248–251, 193–219.
103
Wood S. A. and Normand C. (2008) Mobility of palladium chloride
complexes in mafic rocks: Insights from a flow-through experiment at
25℃ using air-saturated, acidic, and Cl-rich solutions. Mineral. Petrol.
92, 81–97.
Xiong X. L., Xia B., Xu J. F., Niu H. C. and Xiao W. S. (2006) Na depletion
in modern adakites via melt / rock reaction within the sub-arc mantle.
Chem. Geol. 229, 273–292.
Zhou M. F., Robinson P. T., Malpas J., Edwards S. J. and Qi L. (2005) REE
and PGE geochemical constraints on the formation of dunites in the
Luobusa ophiolite, Southern Tibet. J. Petrol. 46, 615–639.
104
국문 초록
맨틀의 화학조성은 부분용융, 멜트-맨틀 상호작용과 같은
화성활동에 의하여 주로 변화한다. 특히 전호지역 맨틀암석에 대
한 연구는 오피올라이트의 형성, 섭입대 발생 모델에 대한 연구
등과 같이 지질학적 중요 논의들에 대한 연구에 필수적이지만, 접
근이 용이하지 못하다는 단점을 가지고 있다. 본 연구에서는 국제
해양연구 프로그램(IODP와 ODP)를 통하여 마리아나 전호지역 이
화산에서 채취된 전호지역 맨틀암석의 성인과 전호 맨틀암에서의
백금족 원소 거동에 대하여 연구하기 위하여 암석 기재학적, 지구
화학적 분석을 진행하였다. 맨틀암석은 하즈버가이트, 더나이트,
사문암으로 분류할 수 있었다. 전암분석과 광물분석을 통하여 마
리아나 전호지역 맨틀암은 함수유체로 인해 20–25% 이상의 부분
용융을 거쳤으며 몇몇 시료는 멜트와의 상호작용을 겪었음을 확인
하였다. 맨틀암과 상호작용한 멜트로는 보니나이트질 멜트와 중앙
해령 현무암과 유사한 특징을 가진 전호 현무암질 멜트로 생각된
다.
이화산에서 채취한 맨틀암에 만연한 사문화 과정은 맨틀암
내의 백금족 원소 함량에 큰 영향을 미치지 못하는 것으로 알려져
있다. 그러나 부분용융과 멜트와의 상호작용은 맨틀암 내 백금족
원소 함량에 영향을 줄 수 있다. 함수부분용융으로 인해 높은 정
도의 용융을 겪은 잔여맨틀암은 해양저 페리도타이트와 비교하였
을 때에 백금족 원소의 함량이 다양하게 나타났으며 팔라듐 그룹
백금족 원소(PPGE)와 이리듐 그룹 백금족 원소(IPGE) 사이에 분별
의 정도가 더 크게 나타났다. 이는 부분용융에 따라 황화물이 부
조화 용융을 거치고 황화물 소진(sulfide exhaustion)에 가까워지면서
105
백금족 원소의 주 포용광물이 황화물에서 백금족 원소 합금으로
바뀌었기 때문으로 해석된다. 멜트와의 반응은 멜트 내 황포화도
에 따라 상호작용한 맨틀의 백금족 원소에 영향을 다르게 미친다.
만약 전호 현무암과 같은 황이 포화된 멜트와 맨틀이 상호작용을
하게 된다면 부분용융의 잔여 맨틀과는 달리 팔라듐 그룹 백금족
원소가 이리듐 그룹 백금족 원소보다 더 부화되는 것을 알 수 있
었다. 이를 통해 마리아나 전호지역 맨틀 내 백금족 원소의 불균
질성은 섭입대 발생 시 생성되는 전호 현무암과의 상호작용으로
나타날 수 있음을 밝혔다.
Keywords : 마리아나 전호지역 맨틀암석, 맨틀의 불균질성, 함수부분용
융, 멜트-맨틀 상호작용, 백금족 원소
Student Number : 2016-28284
106
감사의 글
길지도 않고 그렇다고 짧은 기간도 아닌 2년 반의 석사과
정 기간동안 주변분들의 정말 많은 사랑과 도움을 받으며 학위과
정을 진행할 수 있었습니다. 박정우 교수님, 다양한 대학원 수업과
실험, 디스커션 등을 통해 연구의 “연”자도 모르던 저에게 연구
의 즐거움을 알게 하셨습니다. 학업에만 집중할 수 있도록 연구실
환경과 분위기에 물심양면으로 항상 신경써주신 것 감사합니다.
무엇보다도 학업적 측면뿐만 아니라 삶에 대하여 어떤 관점을 가
져야하고 실천할 수 있을지, 삶에 멘토로 삼고 항상 존경하는 마
음을 가지게 되었습니다. 사랑합니다. 이인성 교수님, 인자한 미소
와 따뜻한 말씀으로 건설적인 조언을 해주시고 미래에 대해 함께
고민하고 조언해주신 것 감사히 기억하겠습니다. 이상묵 교수님,
교수님과의 디스커션을 통해 항상 제가 생각하던 것을 넘어선 통
찰력을 배울 수 있었습니다. 감사합니다.
해양암석지구화학 연구실 식구들! 2년 넘게 동고동락을 함
께한 우리 친구들 정말 감사하고 사랑합니다. 항상 반가운 얼굴로
볼때마다 커피를 사주셨던 선기형, 매력적인 이야기꾼 지혁이형,
사랑으로 가득한 사랑이, 새로운 클린랩 마스터 지원이, 301호 지
킴이 규승이, 해암지의 쉐프 성준이, 한국어실력을 숨기고 있는
Austin. 우리 연구실의 귀염둥이들, 대학원생보다 더 열심히인 우리
인턴 성환이, 규찬이, 다영이, 그리고 2년간 실험실의 엄마 노릇을
했던 졸업생 유진이, 모두가 주었던 물질적, 정신적 도움이 없었다
면 석사과정을 이렇게 잘 마무리하지 못했을 것 같습니다. 여러분
의 사랑 평생 기억하겠습니다.
2015년부터 함께하며 저의 삶을 바꾸어 놓은 서울대학교회
107
와 지체들 감사합니다. 홍종인, 박동열, 김종서 교수님과 김난주,
이화연, 김수민 사모님, 신상주 목사님, 제 어리광을 들어주시고
제가 성장할 수 있게 격려와 기도로 항상 함께 해주시고 연약한
저의 모습들을 감싸주고 사랑해주신 것 감사합니다. 나이를 막론
하고 우리 교회 형제, 자매들로부터 항상 많이 배워왔습니다. 석사
과정을 마치며 잠깐 떨어지겠지만, 여러분과 다시 함께할 날들을
소망합니다. 힘든 일이나 기쁜 일이나 아무 일이 없더라도 언제나
만나면 즐거웠던 10학번 동기들 고맙습니다. 함께 대학원 생활을
했던 훈, 성중, 준하, 지훈이와 학부 졸업 후에도 잊지않고 만났던
지상, 호석, 민구, 규식. 만날 때 마다 에너지를 주고 즐거운 추억
을 만들어 주었던 친구들 항상 고맙습니다. 특히 가까이에서 대학
원 생활의 선배로서 조언해주고 격려해주었던 진우, 고맙습니다.
항상 저를 막내로서 큰 사랑을 부어주시고 칭찬과 격려를
아끼지 않으셨던, 감사하다는 말로는 부족한 아버지, 어머니, 진심
으로 감사합니다. 저의 선택을 항상 믿고 존중해주시며 넘치는 지
원을 해주신 것을 알고 있습니다. 부모님이 알려준 큰 사랑 베풀
며 살아가겠습니다. 타지에 나와 힘들지만 서로에게 힘이 되어주
고 응원해주었던 누나, 형과 형수님, 감사합니다. 2010년에 시작했
던 사랑을 변치 않고 지키며 키워간 채현이, 사랑합니다. 함께할
앞날을 기대하며 기도합니다.
끝으로 지금의 저를 있게 하시고 이 결실을 맺게 하신 전
능한 주님께 전심을 다해 감사와 찬양을 올려드립니다.