Quick Search:    advanced search

GSW Home   GeoRef Home   My GSW Alerts   Contact GSW   About GSW   Journals List   Help

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Hwang, S.-L. Articles by Xu, Z. GeoRef GeoRef Citation

Hematite and magnetite precipitates in olivine from the Sulu peridotite: A result of dehydrogenation-oxidation reaction of mantle olivine?

¹ Department of Materials Science and Engineering, National Dong Hwa University, Hualien, Taiwan, ROC
² Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan, ROC
³ Central Geological Survey, P.O. Box 968, Taipei, Taiwan, ROC
⁴ Institute of Materials Science and Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan, ROC
⁵ Department of Earth Sciences, National Cheng-Kung University, Tainan, Taiwan, ROC
⁶ Key Laboratory of Continental Dynamics, Ministry of Land and Resources. Institute of Geology, Chinese Academy of Geological Sciences, Beijing 100037, China

Correspondence: ^* E-mail: tfyui{at}earth.sinica.edu.tw

	ABSTRACT

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
Analytical electron microscopic observations have been carried out on a garnet peridotite from the Maobei area, Sulu ultrahigh-pressure terrane. The results showed that olivine in this garnet peridotite (5.3–6.6 GPa; 853–957 °C), contains precipitates of chromian magnetite and chromian-titanian hematite at dislocations and (001) faults. Specific crystallographic relationships were determined between these precipitates and the olivine host, viz. [101]_Mt//[001]_Ol, [110]_Mt//[01]_Ol, and [01]_Mt//[011]_Ol; and [0001]_Hm//[100]_Ol and [100]_Hm//[001]_Ol. These oriented oxides are not associated with silicate/silica phases and therefore cannot be accounted for by the mechanism of olivine oxidation. It is postulated that these magnetite and hematite precipitates most likely have resulted from dehydrogenation-oxidation of nominally anhydrous mantle olivine during rock exhumation. In view of the contrasting diffusion rates of H and Fe in the olivine lattice, it is suggested that the formation process might actually take place in steps. Hydrogen diffusion with concomitant quantitative oxidation of Fe²⁺ to Fe³⁺ in olivine occurred early during initial rock exhumation and was followed by slow Fe diffusion forming magnetite/hematite at stacking faults and dislocations within the olivine lattice. Two requirements are essential under such a scenario: an ample amount of H content of the olivine, and an appropriate exhumation rate, probably in the range of 6–11 mm/year, of the host rock. It is also noted that such dehydrogenation-oxidation processes may hamper a correct estimate of the actual P-T conditions and mantle oxidation state based on mineral chemistries present in mantle eclogite/peridotite. The present study demonstrates that oriented mineral inclusions may not necessarily form through exsolution processes sensu stricto, but may form through a series of more complicated reaction mechanisms.

Key Words: Magnetite • hematite • olivine • dehydrogenation-oxidation • UHP peridotite

	INTRODUCTION

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
Lamellae or oriented needles/rods of a single oxide/silicate phase evenly distributed within host silicates have been generally regarded as precipitates formed upon unmixing of parental minerals. Albite lamellae in K-feldspar are a well-known classical example (e.g., Deer et al. 1966). Recently, numerous oriented needles/rods of oxides/silicates have been found within silicate minerals from ultrahigh-pressure (UHP) rocks and have been suggested to be of exsolution origin (Zhang and Liou 1999). Since exsolution, by definition, takes place in the solid state within a closed system (in terms of material transfer) as a result of changing physical conditions, the phenomena or the recalculated compositions of the presumed parental phase can theoretically be employed to yield quantitative estimates of the physical conditions at which the exsolution process occurred. Some examples of such practices include pressure estimates based on the recalculated Ti content in olivine with oriented ilmenite rods from the Alpe Arami peridotite (Dobrzhinetskaya et al. 1996), and the recalculated Si content in titanite with oriented coesite needles from the Kokchetav UHP marble (Ogasawara et al. 2002). On the other hand, recent studies have also shown that oriented mica + quartz + talc in phengite from the Dora-Maira whiteschist (Ferraris et al. 2000) or oriented rutile needles in garnet from the Sulu UHP eclogite (Hwang et al. 2007) may not have necessarily formed in a closed system (in terms of material transfer). Such open-system reactions would not be ascribed to exsolution sensu stricto and would, in some cases, cast doubts on the quantitative estimation of the physical conditions mentioned above. Detailed studies of the formation mechanisms for such lamellae or oriented needles/rods are therefore required before any geological implications can be confidently proposed.

In the present study, two kinds of oriented Fe(-Cr-Ti) oxide precipitates have been identified within olivine crystals from a garnet peridotite within the Sulu UHP terrane. With the help of analytical electron microscopic (AEM) studies, possible formation mechanisms for these oxides are then discussed.

	BACKGROUND GEOLOGY AND SAMPLE DESCRIPTION

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
The Dabie and the Sulu UHP metamorphic terranes in east-central China are a result of a subduction-collision process whereby the Yangtze block was subducted beneath the North China block during 240–220 Ma (e.g., Ames et al. 1993; Liou et al. 1996; Liu et al. 2004a, 2004b, 2006). These two terranes were then separated by the left-lateral Tanlu fault for a distance of at least 500 km (Fig. 1), probably during Cretaceous and/or Cenozoic time. The UHP metamorphic rocks mainly include eclogite, marble, gneisses, amphibolite, and ultramafic rocks. Inclusions of coesite and quartz pseudomorphs in zircon, garnet, omphacite, and kyanite, etc. are common in all rock types (e.g., Enami and Zang 1989; Wang et al. 1989; Ye et al. 2000). Microdiamond inclusions, however, have only been reported from a few eclogites (Xu et al. 2005).

View larger version (30K): [in this window] [in a new window]   FIGURE 1. Simplified geologic map of the Dabie and the Sulu UHP terranes, showing the sample drilling site of Maobei. Ep = epidote; Ec = eclogite; GS = greenschist; UHP Metamorphic belt = ultrahigh-pressure metamorphic belt.

The studied garnet peridotite is a core sample from Maobei drilled by the local geological teams in the Sulu UHP terrane, China (Fig. 1). The sample corresponds to a type A peridotite of Zhang et al. (2000). The peridotite consists mainly of olivine, orthopyroxene, clinopyroxene, and garnet, with small amounts of chromite. Phlogopite inclusions are present within some clinopyroxene grains. In addition, amphibole reaction rims are observed around some garnet and olivine grains, whereas serpentine and chlorite are also common along grain boundaries or along fractures in olivine and pyroxene grains.

	ANALYTICAL TECHNIQUES

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
Electron probe microanalysis (EPMA)
Quantitative chemical analyses were made by a JEOL electron probe microanalyzer (EPMA) JXA-8900R, equipped with four wave-length dispersive spectrometers (WDS), located at the Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan.

Back-scattered electron images were used to guide the analysis to target positions in the minerals. A 2 µm defocused beam was used for analysis at an accelerating voltage of 15 kV with a beam current of 10 nA. The measured X-ray intensities were corrected by the ZAF method and calibrated using the following synthetic and natural mineral standards: wollastonite for Si (TAP) and Ca (PET), rutile for Ti (PET), corundum for Al (TAP), Cr₂O₃ for Cr (PET), fayalite for Fe (LiF), tephroite for Mn (PET), nickel olivine for Ni (LiF), albite for Na (TAP), and adularia for K (PET). Peak intensity for each element and both the upper and lower background positions were counted for 20 and 10 s, respectively. The analytical uncertainties for elements are estimated to be ~1% at the >10 wt% level, ~5% at the 1–10 wt% level, and ~10% at the <1 wt% level.

Analytical electron microscopy (AEM)
Olivine crystals with precipitates were selected from petrographic thin sections and argon-ion milled to electron transparency for AEM observations. We used a JEOL 3010 instrument, located at the National Sun Yat-sen University, Kaohsiung, Taiwan, at 300 kV for imaging, selected area electron diffraction (SAED), and point-count energy dispersive X-ray (EDX) analysis of the milled thin foils. The foils were clamped between two copper rings to ensure sample integrity. The beam size was 10 nm and the acquisition time was 120 s. Semi-quantitative chemical analysis was based on the Cliff-Lorimer thin-film approximation using k values determined by mineral standards. Matrix X-ray subtraction was performed before composition calculation. The measurement errors were ±0.5% for Fe and ±1.5% for Ti and Cr.

	MINERALOGY AND MINERAL CHEMISTRY

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
Optical microscopy under plane-polarized light indicates that the olivine grains range from 0.5 to 5 mm in size. Plate-like (reddish) and rod-like (opaque) precipitates are always found in the interior of the olivine grains in contrast to a precipitate-free zone ranging in width from 100 to 200 µm along the grain rims (Fig. 2a). In a few cases, fractures filled with retrograde minerals, such as serpentine or chlorite, crosscut these precipitates.

View larger version (39K): [in this window] [in a new window]   FIGURE 2. Photomicrographs in plane-polarized light showing precipitates in an olivine crystal from the Maobei peridotite: (a) precipitate-free zone near the olivine rim with the (010) plane edge on, (b) plate-like chromian-magnetite (Mt) and rod-like chromian titanian hematite (Hm) precipitates, (c) discrete bands, i.e., palisades or rod-like precipitates parallel to (001)Ol. Scale bars are 100 µm. The insert in b is a back-scattered electron image of one hematite rod in olivine. Here the scale bar is equal to 10 µm.

View larger version (43K): [in this window] [in a new window]   FIGURE 3. (a, b) Transmission electron micrograph and corresponding SAED pattern inset of plate-like Cr-magnetite precipitates in olivine along the [310]Ol/[01]Mt zone axis. The precipitate edges intersect at ~126°. An interfacial dislocation is also arrowed in b. (c) Magnified SAED pattern. (d) Schematic indexing of the olivine/magnetite crystallographic relationship as compiled in the stereographic projection in e. Scale bars: a = 1 µm, b = 500 nm.

Hematite is associated either with a (001) fault in the olivine, in particular at its superpartial dislocation (Fig. 4a) and ledges (Fig. 4b), or with free dislocation (Fig. 4c). Using g·R = integer criteria in diffraction-contrast imaging, the displacement vector of the (001) defects was identified to be consistent with either R₁ = <0 3/11 1/4> or R₂ = <0 0 1/2>. As noted by Drury (1991), these are the only two <0 v w> translations in olivine that produce a coincident site lattice with half of the divalent cation sites exactly coincident. Based on the SAED pattern and schematic indexing (Figs. 4d–4e), the typical precipitate rod was found to extend along the [110]_Hm//[010]_Ol direction following a specific crystallographic relationship [0001]_Hm//[100]_Ol and [100]_Hm// [001]_Ol. These precipitates are typically bounded by hematite faces such as (0001), (100), and (104) and form palisades along the (001) fault of olivine (Figs. 2c and 4a–4b).

View larger version (47K): [in this window] [in a new window]   FIGURE 4. Transmission electron micrograph of rod-like chromian-titanian hematite precipitates in an olivine crystal: (a) at the edge or superpartials of a stacking fault, (b) at the ledge of a stacking fault, and (c) at isolated dislocations (arrowed). The corresponding SAED pattern and the schematic indexing of the olivine/hematite crystallographic orientation are shown, respectively, in d and e. Scale bars: a = 1 µm, b = 500 nm, c = 200 nm. Note that, because a was not taken along any zone-axis, only the trace of the (001)Ol plane was included.

	DISCUSSION

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

In the present study, two types of Fe oxides with different shapes, i.e., plate-like chromian magnetite and rod-like chromian titanian hematite, were identified having specific crystallographic relationships with the olivine host in association with stacking faults or dislocations. There is no oriented silicate/silica phase associated with these Fe oxides within olivine. Although both magnetite and hematite are present only in the interior part of the olivine grains, the olivine interior and inclusion-free rim are chemically indistinguishable, despite the fact that Ti and Cr are both too low to show meaningful variations. This similarity may indicate that the inclusion-free olivine rim is most probably a result of nucleation site saturation at the grain boundary. It is noted that both magnetite and hematite are only present as inclusions in olivine, but not in other constituent minerals nor in the rock matrix. Consequently, this negates the possibility that these two oxides are primary inclusions in olivine. The different mineral associations between these oxides and those secondary minerals after olivine, such as serpentine and chlorite, also indicate that they are genetically unrelated.

The different shapes of these two oxides could be rationalized by minimization strain energy under a specific crystallographic relationship, as has been demonstrated by Risold et al. (2003) for rod-like ilmenite precipitates in olivine. The ambient lattice parameters for chromian-titanian hematite have been determined by electron diffraction and lattice parameter refinement (based on the lattice parameters of Fo₉₂Fa₈ olivine being a = 4.760 Å, b = 10.221 Å, c = 5.991 Å) to be a = 5.031 Å and b = 13.680 Å. This would lead to a misfit (d_ppt-d_Ol)/d_Ol level of –1.5, –4.2, and –3.5% for (120)_Hm//(010)_Ol, (0001)_Hm//(100)_Ol, and (100)_Hm// (001)_Ol, respectively. In contrast, using a lattice parameter of 8.40 Å for chromian magnetite, the calculated misfits on the (11)_Mt// (100)_Ol habit plane are very small: –0.9%, +0.2%, and +0.2%, respectively, for [101]_Mt//[001]_Ol, [110]_Mt//[01]_Ol, and [01]_Mt// [011]_Ol. In fact, the compressibility or thermal expansion difference between the precipitates and the olivine is not large enough to affect the shape in terms of the anisotropic lattice mismatch, which is in good agreement with the reported data from the equation of state (Smyth et al. 2000a, 2000b) (see also the Appendix). The precipitate morphology, in this case, is therefore not a good indicator of the formation temperature and pressure.

It is well known that the oxidation state of the (metaso-matized) mantle is close to the quartz-fayalite-magnetite buffer (Wood et al. 1990) and that mantle olivine has a very low Fe³⁺ content (Nakamura and Schmalzried 1983; McCammon et al. 2004). In the present study, the presence of both magnetite and hematite precipitates in olivine clearly indicates that oxidation must have been an integral part of the formation mechanism. Magnetite/hematite lamellae/needles have been reported in oxidized olivines from laboratory experiments (Champness 1970), as well as in volcanic rocks or in mantle xenoliths enclosed within volcanic rocks (e.g., Haggerty and Baker 1967; Kohlstedt and Vander Sande 1975). Based on these experimental studies and natural occurrences, it has been demonstrated that olivine could be oxidized at high temperatures. Magnetite/hematite as oxidation products in such olivines, however, is invariably associated with pyroxene and/or SiO₂ phases. Such pyroxene and/or SiO₂ phases are required according to stoichiometric reactions such as:

(1)

(2)

in which SiO₂ would further react with olivine to form pyroxene. It has also been shown that such oxidation processes generally take place inward from olivine grain boundaries. In addition, Khisina et al. (2000) recently conducted similar high-temperature oxidation experiments on olivine from a peridotite nodule sampled from a kimberlite and observed nanometer-scale inclusions as oxidation products, which include mixtures of feroxyhite (FeOOH), bernalite [Fe(OH)₃], and β-cristobalite. These oxidation products were suggested to result from the decomposition of a hydrous magnesian silicate, probably hydrous olivine (Khisina and Wirth 2002). They were interpreted to be metastable phases that would eventually transform to hematite and quartz. The lack of silicate/silica phases in association with magnetite/hematite precipitates in olivine from the Maobei peridotite, the presence of Fe oxides in the interior of the olivine grains (Fig. 1a), and the occurrence of the Maobei peridotite within UHP metamorphic rocks do not support this oxidation mechanism. It should be noted that the crystallographic relationships between oriented hematite and host olivine in the present study are the same as those reported for experimentally oxidized olivine (Champness 1970) as well as in natural occurrences of oxidized olivine (Kohlstedt and Vander Sande 1975). On the other hand, although the crystallographic relations between oriented magnetite and olivine in the present study are the same as those reported from experimental results (Champness 1970), they are different from those of natural occurrence given by Kohlstedt and Vander Sande (1975) (Table 3). The reason for the latter difference is not clear.

(3)

Although magnetite and hematite precipitates in olivine from the present study are closely associated with stacking faults and dislocations, laihunite layers were not observed within olivine. If the absence of such laihunite layers is due to later decomposition in which magnetite/hematite formed during subsequent cooling, silicate/silica phases should be present as additional reaction products (Dyar et al. 1998). Such silicate/silica phases, as mentioned above, were not observed in the present study. In this regard, the stacking faults and dislocations in the studied olivines may only be rapid diffusion pathways and sites for heterogeneous nucleation and not be an inherited structure from the decomposition of laihunite layers.

Recognizing the difficulties in applying various oxidation scenarios for olivine (due to the lack of associated silicate/silica phases as described above), Zhang et al. (1999) hypothesized a phase change scenario to explain the observed oriented magnetite lamellae in olivines from the Mauwu harzburgite, Dabie UHP terrane. They suggested that the harzburgite may have been sub-ducted into the deep mantle with the olivine being converted to a high-pressure phase. The magnetite lamellae could have been exsolved from the olivine during subsequent exhumation after back-transformation from wadsleyite (or other high-pressure phases such as phase A) that contains a significant Fe³⁺ content (Richmond and Brodholt 2000). Although such a phase change scenario could also be employed to explain the magnetite/hematite precipitates in the olivine grains from the present study, other possibilities should still be investigated.

Since the pioneering work of Martin and Donnay (1972), it has now been well established that H can be regularly contained in nominally anhydrous minerals, such as olivine, pyroxene and garnet, under mantle conditions (e.g., Rossman 1996; Ingrin and Skogby 2000). Taking olivine as an example, H is normally incorporated within the crystal lattice as a hydroxyl ion in association with point defects, such as Si and Mg cation vacancies (e.g., Kohlstedt et al. 1996; Kohn 1996). In the presence of Ti, H may associate with planar defects forming clinohumite-like layers within the olivine (Kitamura et al. 1987). Decomposition of such clinohumite-like layers was also proposed to be responsible for the formation of oriented ilmenite rods within olivine from orogenic peridotites by Risold et al. (2001).

The reported H content in mantle olivines may, expressed in ppm H₂O, reach up to 400 ppm, but typically is <100 ppm (Bell and Rossman, 1992; Beran and Libowitzky 2006). Based on experimental data, the H content in olivine may be as high as ~5000 ppm H₂O under mantle conditions (Hirschmann et al. 2005; Mosenfelder et al. 2006). If the release of H from such mantle olivines during exhumation takes place through dehydrogenation-oxidation process (Ingrin and Skogby 2000), Fe³⁺ may form in the following manner:

(4)

Since very little Fe³⁺ can be accommodated within olivine under mantle conditions (Nakamura and Schmalzried 1983; McCammon et al. 2004), and Fe³⁺ may be released from the olivine lattice to form hematite and magnetite. The whole reaction can be described as:

(5)

where represents Fe or Si cation vacancies. These allow for charge compensation due to the presence of H ions in the olivine lattice (Ingrin and Skogby 2000). It should be noted that since the diffusion rate of H is much faster than that of Fe in olivine (Kohlstedt and Mackwell 1998; Dohmen and Chakraborty 2007), reaction 5 should be an integrated expression. The real processes would have taken place in steps, with dehydrogenation-oxidation occurring first, followed by magnetite-hematite formation. Unfortunately, the AEM observations were not capable of distinguishing between the formation sequences for magnetite and hematite. Under the estimated metamorphic pressure and temperature conditions (i.e., 53–66 kbar and 853–957 °C) for the Maobei peridotite, calculations indicate that it would take less than one day for H to diffuse out of olivine crystals with a size of 5 mm, but it would take about 10 Ma for Fe to diffuse out of olivine with the same grain size. Although the above estimations can only be semi-quantitative, due to uncertainties such as the effect of site vacancies or oxygen fugacity on the diffusion rate, the results clearly demonstrate the contrasting time frames for the two diffusion processes. It is therefore not surprising to see that, in the absence of silicate/silica phases, oriented magnetite and/or hematite has only been reported from orogenic peridotites (Zhang et al. 1999; this study). For mantle xenoliths within volcanic rocks, although the eruption rate (or the cooling rate) is slow enough for H to diffuse out of the olivine within the xenoliths, it is too fast for Fe diffusion to form magnetite/hematite. It is also worth mentioning that Khisina and Wirth (2002) reported exsolved nanometer-size hydrous olivine within the olivine from a peridotite nodule from the Udachnaya kimberlite pipe, Siberia. This observation would indicate an extremely fast extrusion rate for the Udachnaya kimberlite such that even H was trapped within the olivine crystal and could only exsolve as hydrous olivine. The exhumation rate for the host rock is therefore one of the critical conditions for whether such precipitates will form within the olivine. The rock exhumation rate should not be too fast such that Fe³⁺ in the olivine has enough time to diffuse within the lattice and to nucleate, thus forming magnetite/hematite at favorable lattice sites under high temperatures. The exhumation rate should also not be too slow, such that the Fe³⁺ diffuses out of the olivine and enters into adjacent minerals allowing for re-equilibration.

Similar to other UHP terranes, the exhumation history of the Dabie-Sulu terrane can be divided into an early near isothermal decompression and a later decompression cooling stage (Liou et al. 1996; Hacker et al. 2000; Ratschbacher et al. 2000; Lin et al. 2005). The first-stage exhumation was mainly driven by buoyancy bringing the rock massif from mantle depths >100 km (750–970 °C) to the base of the crust ~35 km (550–650 °C). The second-stage exhumation has been suggested to be related to extension, which resulted in the massif being brought to the surface. The magnetite and hematite precipitates in olivine of the Maobei peridotite most probably were formed during the first stage of rock exhumation, because temperature conditions for the second stage of exhumation may be too low for Fe diffusion. The exhumation rate for the first stage is suggested to be in the range of 6–11 mm/year (Liu et al. 2004a, 2004b, 2006). This rate may be taken as a reference appropriate for the formation of magnetite and hematite precipitates in olivine from mantle peridotite through a dehydrogenation-oxidation process in the temperature range 750–970 °C. It is difficult, however, to further explore the limits of the formation conditions for these oxide precipitates because the available information is rather limited.

The distribution density of rod-like chromian-titanian hematite inclusions varies both within and among the olivine grains. By counting the number of rods within the olivine crystals, specifically rods end-on in the olivine [010] section (Fig. 2c), the volume percentage of rods is estimated to be 0.025–0.050%. Given a composition of Fo_91–93 for the olivine host, the dopant levels in the primary olivine were estimated to be ~700 ppm for Fe³⁺. If the plate-like chromian magnetite precipitates are also taken into consideration, the Fe³⁺ content could reach ~1000 ppm (which roughly corresponds to a Fe³⁺/Fe^total ratio of 0.0125). If all the Fe³⁺ resulted from a dehydrogenation-oxidation process, that would correspond to a loss of a 160 ppm H₂O from a typical olivine grain in the peridotite. Recent studies of orogenic peridotites from the UHP terrane indicate that olivine still preserves 30–150 ppm H₂O (Xu et al. 2006). Theoretically, there should be ample amounts of H₂O in the North China-Yangtze UHP subduction-collision zone for substantial hydrogenation of olivine when the Maobei peridotite was incorporated into the subduction-collision system (e.g., Peacock 1990). Mantle metasomatism could be another alternative (Banfield et al. 1992) as was indicated by the reported chemical nature of some metasomatized Sulu garnet peridotites (Zheng et al. 2005).

It should be pointed out that the above dehydrogenation-oxidation process is also a convincing alternative with regard to the formation mechanism for the oriented magnetite lamellae in olivines from the Mauwu harzburgite, Dabie UHP terrane, as reported by Zhang et al. (1999). The difference between the presence of only magnetite vs. the presence of both magnetite and hematite precipitates in olivine would be due to the different amounts of Fe³⁺ present, which in turn would result from different H contents in the olivine before exhumation of the peridotite. In fact, the postulated dehydrogenation-oxidation process may not necessarily be incompatible with the phase transformation scenario proposed by Zhang et al. (1999) since wadsleyite is known to be capable of containing up to 3 wt% H₂O (Inoue et al. 1995), in addition to a probable high content of Fe³⁺ (Richmond and Brodholt 2000).

	IMPLICATIONS

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
If the magnetite and hematite precipitates in olivine from the Maobei peridotite were indeed formed through a dehydro-genation-oxidation process, it may lead to some interesting and important implications. For example, in the case of mantle xenoliths brought up by volcanic eruptions, although H could totally diffuse out of the olivine crystal, the slower diffusion rate for Fe³⁺ would not allow it to do so such that it would be trapped within the lattice. Subsequently no oxide inclusions would form. Detecting such small amounts of Fe³⁺ in mantle olivine might be difficult (e.g., Dyar et al. 1989). However, the inverse relation between H⁺ and Fe³⁺ contents in kaersutite megacrysts from basaltic flows is a well-known example of the dehydrogenation-oxidation process (Dyar et al. 1992). In this regard, the absence of magnetite/hematite precipitates in pyroxenes and garnet from UHP peridotites as well as mantle xenoliths, may also be accounted for by the fact that Fe³⁺ is more easily accommodated within the lattice structure of either mineral (Ingrin and Skogby 2000). Consequently, the effect of the dehydrogenation-oxidation process must be taken into consideration during estimation of the mantle oxidation state as based on the chemical composition of typical mantle minerals (Wood et al. 1990). Such a process also makes the correction for Fe³⁺ in P-T estimations for eclogite and mantle peridotite more complicated, if not impossible. This latter issue has been thoroughly reviewed by Krogh Ravna and Paquin (2003) for UHP rocks. It has been shown that employing different Fe³⁺ correction schemes (including no Fe³⁺ correction, Fe³⁺ correction by different theoretical calculations, or Fe³⁺ correction by Mössbauer analysis or micro-XANES work), the temperature estimates may differ by 100–200 °C, but the estimated error in pressure, caused by ignoring Fe³⁺ in the garnet-orthopyroxene geobarometers, is completely unknown (Krogh Ravna and Paquin 2003; Schmidt et al. 2003). The temperature difference between estimations with and without the Fe³⁺ correction for clinopyroxene in this study is about 135 °C, in agreement with previous studies. This would be the maximum error in temperature estimation that could be induced by the dehydrogenation-oxidation process.

The present study provides an additional example that topotactically oriented mineral inclusions in UHP minerals may not necessarily be the result of simple exsolution (sensu stricto) process. Ferraris et al. (2000) noticed oriented talc, quartz, and celadonitic mica inclusions occurring within phengite in the pyrope white schist from Dora Maira. They ascribed the observation by the decompression reaction:

(6)

and commented that it is not an exsolution mechanism sensu stricto because K was removed from the system. Both reactions 5 and 6 allow for mass to be removed from the system, but differ in that the latter takes place through mineral decomposition whereas the former takes place through ion diffusion. In contrast, Hwang et al. (2007) demonstrated that oriented rutile needles in garnet from UHP eclogite most probably form through interaction between infiltrating fluids and the garnet host, indicating Ti mobility under UHP metamorphic conditions despite the very low solubility of rutile in H₂O at 1000–1100 °C and 1–2 GPa (Tropper and Manning 2005). This suggestion is further supported by the presence of rutile-bearing veins in eclogites probably resulting from an increase in Ti solubility due to the presence of F- and Na-aluminosilicate polymers in the fluid (e.g., Gao et al. 2007).

It should be noted that such complicated formation mechanisms need not be restricted to oriented mineral inclusions in UHP rocks. They may also operate in rocks formed under lower P-T conditions. For example, oriented monazite inclusions within fluorapatite have been reported in a metasomatized mica schist (Pan et al. 1993) as well as magnetite-apatite ores of magmatic origin (Harlov et al. 2002a). In both cases, formation of the monazite inclusions was interpreted to be the result of fluid-fluor-apatite interaction. Based on a series of systematic experimental studies, the formation conditions (P-T and fluid chemistry) have been constrained for monazite inclusion formation in fluorapatite (Harlov et al. 2002b, 2005; Harlov and Förster 2003). All these examples clearly demonstrate that oriented mineral inclusions may form through a variety of reaction mechanisms other than the process described in this study.

	APPENDIX

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
Equation of state of Fe-Ti-Cr oxides and silicate olivine
According to the compilation of Smyth et al. (2000a, 2000b), the ilmenite (FeTiO₃) structure is relatively more compressible than hematite (-Fe₂O₃), eskolaite (Cr₂O₃), and Ti₂O₃ with bulk moduli of 170, 225, 238, and 186 GPa, respectively. In contrast to thermal expansion, compression of ilmenite is anisotropic with the a- and c-axes having linear compressibilities of 1.34 and 2.63 x 10^–3/GPa, respectively. The bulk modulus of the FeO₆ octahedron is about 140 GPa, whereas that of the TiO₆ octahedron is 290 ± 50 GPa.

As for silicate olivine, the compression of the unit cell is strongly anisotropic, with the b-axis being by far the most compressible in all natural compositions (Smyth et al. 2000a, 2000b). This compression behavior is consistent with ultrasonic measurements of ferromagnesian olivines, which indicate that b is the slowest direction, whereas a and c are nearly equal and fast. This anisotropy is due to the fact that the M2 polyhedra, the most compressible of the structural units, form continuous layers in the a-c plane so that compression parallel to b depends only on compression of M2, whereas compression in other directions requires compression of both M1 and M2. Furthermore, the M2 polyhedron shares an edge with the silicate tetrahedron, but this shared edge is parallel to c so it does not affect compression in the b-direction. The axial compressions were reported to be 1.6, 4.3, and 0.8 x 10^–3/GPa along the a-, b-, and c-axes of synthetic forsterite (Mg₂SiO₄), respectively. The general effect of increasing pressure on the silicate olivine structure is observed to be similar to that of decreasing the temperature, so that mantle olivine at 100 km depth is predicted to have a crystal structure similar to that of forsterite at 1 atmosphere and 600 °C.

	ACKNOWLEDGMENTS

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 
We thank L.C. Wang for technical assistance and J. Chu for reading the manuscript. Helpful suggestions/comments by D.E. Harlov, K.N. Bozhilov, and an anonymous reviewer are greatly appreciated. This research was supported by National Science Council, Republic of China.

	Footnotes

 
MANUSCRIPT HANDLED BY DANIELH ARLOV

MANUSCRIPT RECEIVED September 10, 2007; MANUSCRIPT ACCEPTED February 5, 2008

	REFERENCES CITED

Top Abstract Introduction Background geology and sample... Analytical techniques Mineralogy and mineral chemistry Discussion Implications Appendix Acknowledgments References cited

 

Ames, L., Tilton, G.R., and Zhou, G. (1993) Timing of collision of the Sino-Korean and Yangtze cratons: U-Pb zircon dating of coesite-bearing eclogites. Geology, 21, 339–342.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Ashworth, J.R. (1979) Two kinds of exsolution in chondritic olivine. Mineralogical Magazine, 43, 535–538.[CrossRef][ISI][GeoRef]

Banfield, J.F., Dyar, M.D., and McGuire, A.V. (1992) The defect microstructure of oxidized mantle olivine from Dish Hill, California. American Mineralogist, 77, 977–986.[Abstract][ISI][GeoRef]

Bell, D.R. and Rossman, G.R. (1992) Water in Earth’s mantle: the role of nominally anhydrous minerals. Science, 255, 1391–1397.[Abstract/Free Full Text][CrossRef][ISI][Medline][GeoRef]

Beran, A. and Libowitzky, E. (2006) Water in natural mantle minerals II: olivine, garnet and accessory minerals. In H. Keppler and J.R. Smyth, Eds., Water in Nominally Anhydrous Minerals, 62, p. 169–191. Reviews in Mineralogy and Geochemistry, Mineralogical Society of America, Chantilly, Virginia.

Berman, R.G., Aranovich, L.Ya., and Pattison, D.R.M. (1995) Reassessment of the garnet-clinopyroxene Fe-Mg exchange thermometer: II. Thermodynamic analysis. Contributions to Mineralogy and Petrology, 119, 30–42.[ISI][GeoRef]

Bozhilov, K., Green, H.W., and Dobrzhinetskaya, L.F. (2003) Quantitative 3D measurement of ilmenite abundance in Alpe Arami olivine by confocal microscopy: Confirmation of high-pressure origin. American Mineralogist, 88, 596–603.[Abstract/Free Full Text][ISI][GeoRef]

Brey, G.P. and Köhler, T. (1990) Geothermobarometry in four-phase lherzolite II. New thermobarometers, and practical assessment of existing thermobarometers. Journal of Petrology, 31, 1353–1378.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Champness, P.E. (1970) Nucleation and growth of iron oxides in olivines, (Mg,Fe)₂SiO₄. Mineralogical Magazine, 37, 790–800.[ISI][GeoRef]

Deer, W.A., Howie, R.A., and Zussman, J. (1966) An introduction to the rock-forming minerals, 528 p. John Wiley and Sons, Inc., New York.

Dobrzhinetskaya, L.F., Green, H.W., and Wang, S. (1996) Alpe Arami: A peridotite massif from depths of >300 km. Science, 271, 1841–1845.[Abstract][CrossRef][ISI][GeoRef]

Dohmen, R. and Chakraborty, S. (2007) Fe-Mg diffusion in olivine II: Point defect chemistry, change of diffusion mechanisms and a model for calculation of diffusion coefficients in natural olivine. Physics and Chemistry of Minerals, 34, 409–430.[CrossRef][ISI]

Drury, M.R. (1991) Hydration-induced climb dissociation of dislocations in naturally deformed mantle olivine. Physics and Chemistry of Minerals, 18, 106–116.[CrossRef][ISI][GeoRef]

Dyar, M.D., McGuire, A.V., and Ziegler, R.D. (1989) Redox equilibria and crystal chemistry of coexisting minerals from spinel lherzolite mantle xenoliths. American Mineralogist, 74, 969–980.[Abstract][ISI][GeoRef]

Dyar, M.D., McGuire, A.V., and Mackwell, S.J. (1992) Fe³⁺/H⁺ and D/H in kaersutite; misleading indicators of mantle source fugacities. Geology, 20, 565–568.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Dyar, M.D., Delaney, J.S., Sutton, S.R., and Schaefer, M.W. (1998) Fe³⁺ distribution in oxidized olivine: A synchrotron micro-XANES study, Part I. American Mineralogist, 83, 1361–1365.[Abstract][ISI][GeoRef]

Enami, M. and Zang, Q.J. (1989) Quartz pseudomorph after coesite in eclogites from Shandong province, east China. American Mineralogist, 75, 381–386.[ISI]

Ferraris, C., Chopin, C., and Wessicken, R. (2000) Nano- to micro-scale decompression products in ultrahigh-pressure phengite: HRTEM and AEM study, and some petrological implications. American Mineralogist, 85, 1195–1201.[Abstract/Free Full Text][ISI][GeoRef]

Gao, J., John, T., Klemd, R., and Xiong, X. (2007) Mobilization of Ti-Nb-Ta during subduction: Evidence from rutile-bearing dehydration segregations and veins hosted in eclogite, Tianshan, NW China. Geochimica et Cosmochimica Acta, 71, 4974–4996.[CrossRef][ISI][GeoRef]

Hacker, B.R., Ratschbacher, L., Webb, L., and McWilliams, M.O. (2000) Exhumation of ultrahigh-pressure continental crust in east central China: Late Triassic-Early Jurassic tectonic unroofing. Journal of Geophysical Research, 105, 13339–13364.[CrossRef]

Haggerty, S.E. and Baker, I. (1967) The alteration of olivine in basaltic and associated lavas. Contributions to Mineralogy and Petrology, 16, 233–257.[CrossRef][GeoRef]

Harlov, D.E. and Förster, H.-J. (2003) Fluid-induced nucleation of REE-phosphate minerals in apatite: Nature and experiment. Part II. Fluorapatite. American Mineralogist, 88, 1209–1229.[Abstract/Free Full Text][ISI][GeoRef]

Harlov, D.E., Andersson, U.B., Förster, H.-J., Nyström, J.O., Dulski, P., and Broman, C. (2002a) Apatite-monazite relations in the Kiirunavaara magnetite-apatite ore, northern Sweden. Chemical Geology, 191, 47–72.[CrossRef][ISI][GeoRef]

Harlov, D.E., Förster, H.-J., and Nijland, T.G. (2002b) Fluid-induced nucleation of REE-phosphate minerals in apatite: Nature and experiment. Part I. Chlorapatite. American Mineralogist, 87, 245–261.[Abstract/Free Full Text][ISI][GeoRef]

Harlov, D.E., Wirth, R., and Förster, H.-J. (2005) An experimental study of dissolution-reprecipitation in fluorapatite: Fluid infiltration and the formation of monazite. Contributions to Mineralogy and Petrology, 150, 268–286.[CrossRef][ISI][GeoRef]

Hermann, J., Gerald, J.D.F., Malaspina, N., Berry, A.J., and Scambelluri, M. (2007) OH-bearing planar defects in olivine produced by the breakdown of Ti-rich humite minerals from Dabie Shan (China). Contributions to Mineralogy and Petrology, 153, 417–428.[CrossRef][ISI][GeoRef]

Hirschmann, M.M., Aubaud, C., and Withers, A.C. (2005) Storage capacity of H₂O in nominally anhydrous minerals in the upper mantle. Earth and Planetary Science Letters, 236, 167–181.[CrossRef][ISI][GeoRef]

Hwang, S.L., Yui, T.F., Chu, H.T., Shen, P., Scheretl, H.P., Zhang, R.Y., and Liou, J.G. (2007) On the origin of oriented rutile needles in garnet from UHP eclogites. Journal of Metamorphic Geology, 25, 349–362.[CrossRef][ISI][GeoRef]

Iishi, K., Torigoe, K., and Han, X.J. (1997) Oriented precipitate complexes in iron-rich olivines produced experimentally in aqueous oxidizing environment. Physics and Chemistry of Minerals, 25, 8–14.[CrossRef][ISI][GeoRef]

Ingrin, J. and Skogby, H. (2000) Hydrogen in nominally anhydrous upper-mantle minerals: Concentration levels and implications. European Journal of Mineralogy, 12, 543–570.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Inoue, T., Yurimoto, H., and Kudoh, Y. (1995) Hydrous modified spinel, Mg_1.75SiH_0.5O₄: A new water reservoir in the mantle transition region. Geo-physical Research Letters, 22, 117–120.[CrossRef][ISI][GeoRef]

Khisina, N.R. and Wirth, R. (2002) Hydrous olivine (Mg_1–yFe²⁺ _y )_2–xV_xSiO₄H_2x—a new DHMS phase of variable composition observed as nanometer-sized precipitations in mantle olivine. Physics and Chemistry of Minerals, 29, 98–111.[CrossRef][ISI][GeoRef]

Khisina, N.R., Khramov, D.A., Kleschev, A.A., and Langer, K. (1998) Laihunitization as a mechanism of olivine oxidation. European Journal of Mineralogy, 10, 229–238.[Abstract/Free Full Text][ISI][GeoRef]

Khisina, N.R., Langer, K., Andrut, M., Ukhanov, A.V., and Wirth, R. (2000) Nano-scale microstructure of Fe³⁺-, OH^–-bearing crystalline inclusions in experimentally oxidized olivine from a mantle nodule. Mineralogical Magazine, 64, 319–335.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Kitamura, M., Kondoh, S., Morimoto, N., Miller, G., Rossman, G.R., and Putnis, A. (1987) Planar OH-bearing defects in mantle olivine. Nature, 328, 143–145.[CrossRef][GeoRef]

Kohlstedt, D.L. and Mackwell, S.J. (1998) Diffusion of hydrogen and intrinsic point defects in olivine. Zeitschrift für Physikalische Chemie, 207, 147–162.[ISI]

Kohlstedt, D.L. and Vander Sande, J.B. (1975) An electron microscopy study of naturally occurring oxidation produced precipitates in iron-bearing olivines. Contributions to Mineralogy and Petrology, 53, 13–24.[CrossRef][ISI][GeoRef]

Kohlstedt, D.L., Keppler, H., and Ruby, D.C. (1996) Solubility of water in the , β, phases of (Mg,Fe)₂SiO₄. Contributions to Mineralogy and Petrology, 123, 345–357.[CrossRef][ISI][GeoRef]

Kohn, S.C. (1996) Solubility on H₂O in nominally anhydrous minerals using 1H MAS NMR. American Mineralogist, 81, 1523–1526.[Abstract][ISI][GeoRef]

Krogh Ravna, E.J. and Paquin, J. (2003) Thermobarometric methodologies applicable to eclogites and garnet ultrabasites. In D.A. Carswell and R. Compagnoni, Eds., Ultrahigh pressure metamorphism, p. 229–259. Eötvös University Press, Budapest.

Lee, H.Y. and Ganguly, J. (1988) Equilibrium compositions of coexisting garnet and orthopyroxene: Experimental determinations in the system FeO-MgO-Al₂O₃-SiO₂, and applications. Journal of Petrology, 29, 93–113.[Abstract/Free Full Text][CrossRef][ISI][GeoRef]

Lin, L.H., Wang, P.L., Lo, C.H., Tsai, C.H., and Jahn, B.M. (2005) ⁴⁰Ar-³⁹Ar Thermochronological constraints on the exhumation of ultrahigh-pressure metamorphic rocks in the Sulu terrain of Eastern China. International Geology Review, 47, 872–886.[CrossRef][ISI][GeoRef]

Liou, J.G., Zhang, R.Y., Wang, X., Eide, E.A., Ernst, W.G., and Maruyama, S. (1996) Metamorphism and tectonics of high-pressure and ultra-high-pressure belts in the Dabie-Sulu region, China. In A. Yin and T.M. Harrison, Eds., Tectonic Evolution of Asia, p. 300–344. Cambridge University Press, U.K.

Liu, F.L., Xu, Z.Q., Liou, J.G., and Song, B. (2004a) SHRIMP U-Pb ages of ultra-high-pressure and retrograde metamorphism of gneisses, south-western Sulu terrain, eastern China. Journal of Metamorphic Geology, 22, 315–326.[CrossRef][ISI][GeoRef]

Liu, F.L., Xu, Z.Q., and Xue, H.M. (2004b) Tracing the protolith, UHP metamorphism, and exhumation ages of orthogneiss from the SW Sulu terrain (eastern China): SHRIMP U-Pb dating of mineral inclusion bearing zircons. Lithos, 78, 411–429.[CrossRef][ISI][GeoRef]

Liu, F.L., Gerdes, A., Liou, J.G., Xue, H.M., and Liang, F.H. (2006) SHRIMP U-Pb zircon dating from Sulu-Dabie dolomitic marble, eastern China: Constraints on prograde, ultrahigh-pressure and retrograde metamorphic ages. Journal of Metamorphic Geology, 24, 569–589.[ISI][GeoRef]

Martin, R.F. and Donnay, G. (1972) Hydroxyl in the mantle. American Mineralogist, 57, 554–570.[ISI][GeoRef]

McCammon, C.A., Frost, D.J., Smyth, J.R., Laustsen, H.M.S., Kawamoto, T., Ross, N.L., and van Aken, P.A. (2004) Oxidation state of iron in hydrous mantle phases: Implications for subduction and mantle oxygen fugacity. Physics of the Earth and Planetary Interiors, 143-144, 157–169.

Moseley, D. (1981) Ilmenite exsolution in oli