Research Article

Rocks from the Mantle Transition Zone: Majorite-Bearing Xenoliths from Malaita, Southwest Pacific

See allHide authors and affiliations

Science  19 May 2000:
Vol. 288, Issue 5469, pp. 1215-1223
DOI: 10.1126/science.288.5469.1215


Rocks containing high-pressure mineral assemblages derived from the mantle transition zone between depths of about 400 and 670 kilometers occur as xenoliths and megacrysts on the island of Malaita in the southwest Pacific on the Ontong Java Plateau. Observed ultrahigh pressure mineral chemistries include majorite, calcium- and magnesium-perovskite, aluminous silicate phases, and microdiamond. Based on an empirical barometer, majoritic garnets in these xenoliths record pressures of up to 22 gigapascal. The occurrence of material with perovskite chemistry and several enigmatic aluminous phases indicates pressures of up to 27 gigapascal. Samples were brought to the surface at about 34 million years ago by potassic ultramafic magmas, which evidently originated in the lower mantle.

Xenoliths, the main direct source of information about mantle mineralogy, are predominantly derived from the top of the upper mantle, from depths of ∼150 to 200 km (1,2), although some have been reported from depths of ∼400 km (3, 4). Mineral inclusions in diamond are generally interpreted to yield a record of phase relations at depths of 150 to 200 km, equivalent to pressures of ∼5.5 to 7 GPa (5,6). Extremely rare single-mineral inclusions of ferropericlase, Mg-Si perovskite, Ca-Si perovskite, and magnesiowustite have recently been discovered in diamond derived from the lower mantle (7). Most deep xenoliths and diamonds are transported in kimberlites and lamproites that intrude continental lithosphere (1, 8). Here, we describe an extensive suite of garnetite and rare garnet peridotite xenoliths that contain majoritic and other ultrahigh pressure mantle assemblages, indicating that these rocks sample the mantle transition zone in an oceanic environment.

Geology and Locality

The xenoliths and macrocrysts occur in ∼34 Ma (million years ago) alnöite pipes and sills (9), which cut Cretaceous Ontong Java Plateau basalts, Miocene limestones, and mudstones on the island of Malaita, southwest Pacific (Fig. 1). The oceanic Ontong Java Plateau covers an area of 1.28 × 106 km2 to the northeast of the Solomon Island archipelago and Bougainville Island (Fig. 1) and is one of the largest plume-generated intraplate igneous provinces preserved on the planet. Volcanic activity on the Ontong Java Plateau commenced at ∼122 Ma, and a second major pulse of magmatism occurred at 90 Ma (10). At this time, the Ontong Java Plateau was located several thousand km farther east in the central Pacific Ocean (11). Igneous rocks on Malaita have a plume origin, distinct from the arc-generated islands in the Solomon Archipelago (10).

Figure 1

Map showing the location of Malaita on the Ontong Java Plateau. Igneous rocks on Malaita have a mantle plume origin, distinct from the arc generated islands in the Solomon Archipelago (10). Collision of Malaita with the Indo-Australian plate occurred ∼10 Ma (13), before initiation of Solomon Islands arc magmatism (12,14).

Seismic reflection profiles show that Ontong Java Plateau basalts are cut by numerous distinctive plug-like bodies that vary in diameter up to ∼2.5 km (11) and may be kimberlite pipes (12). Alnöitic magmatism at ∼34 Ma occurred before collision of Malaita with the Indo-Australian plate ∼10 Ma (13), and before initiation of geologically unrelated Solomon Islands arc magmatism (12, 14).

We obtained mantle xenoliths and garnet macrocrysts from stream gravels in several drainage systems from the north central and eastern parts of the island (15), where a number of pipes have been mapped. The studied ultramafic xenoliths are ovoid in shape and range in diameter from <1 to 30 cm. Compositions include spinel lherzolite, pyroxenite intergrowths, spinel bearing eclogite, and rare garnet lherzolite. The macrocrysts suite includes pyrope, subcalcic diopside, augite, orthopyroxene (bronzite), ilmenite, clinopyoxene-ilmenite intergrowths, phlogopite, olivine, and zircon (14, 16, 17). Many of the macrocrysts exhibit a distinctive surficial polish that is interpreted to have been caused by abrasion during turbulent magmatic emplacement. Garnet macrocrysts are typically elongate and range in size from ∼1 cm by 0.5 cm by 0.5 cm to larger than 20 cm by 10 cm by 10 cm. Except for olivine, all of the other macrocrystal phases are extremely fresh.

Garnet and Majorite

We obtained more than 600 analyses (18) of garnets in 19 samples. Results are given in Tables 1 through 3 and in Web table 1 (18). Most of the garnets (∼99%) are derived from eclogite protoliths (E-type). They are pyrope-rich with Mg/(Mg+Fe) = 0.69 to 0.82 and Cr2O3 < 0.5 weight percent (wt %). Rare (∼1%) peridotite derived garnets (P-type) have Mg/(Mg+Fe) = 0.82 to 0.92 and Cr2O3 = 1.0 to 2.8 wt % (Table 1). The TiO2 and Na2O contents in P- and E-type pyropes are <0.2 and 0.1 wt % and 0.4 to 0.7 wt %, respectively, and some majoritic garnets contain 3.2 wt % TiO2 and 4.3 wt % Na2O (Tables 1 and2).

Table 1

Analyses of garnet standard and representative compositions of Malaita garnets and majorites.

View this table:
Table 2

Compositions of prograde reactants in sample KC-99-M3.

View this table:
Table 3

Compositions of ultrahigh pressure minerals from Malaita compared to literature data.

View this table:

Both E- and P-type garnets show an extreme range of covariation between Si and Al+Cr (Fig. 2A). They show excellent stoichiometry and define a well-correlated trend between pyrope, with Si4+ = 3 per formula unit (pfu; based on 12 oxygens) and Al+Cr = 2 pfu, and ideal clinopyroxene (19), with Si4+ = 4 pfu and Al+Cr = 0 pfu (Fig. 2A). The slope of the combined vector defines continuous solid solution in majoritic compositions from 3.2 Si4+ pfu to 3.8 Si4+ pfu, which we interpret to reflect the incorporation of Si4+ in octahedral coordination in garnet.

Figure 2

(A) Chemical structural data based on 12 oxygens pfu for garnets from Malaita. Garnets exhibit an extreme range of pressure induced covariation between Si and Al+Cr and define a well correlated trend between pyrope (open squares) with Si = 3 pfu and Al+Cr = 2 pfu and ideal clinopyroxene (19) with Si4+ = 4 pfu and Al+Cr = 0 pfu. Slope of the combined vector defines continuous solid solution in majoritic compositions (solid circles) from 3.2 Si pfu to 3.8 Si pfu, reflecting the presence of Si4+ in octahedral coordination. A small population of ∼3 Si pfu “garnets” (crosses) are clearly deficient in Al+Cr (1.85 to 1.55). These may be the ultradeep “majorite” composition reported in experiments at 27 GPa (23, 24), which forms at depths of ∼670 km. (B) Data relative to 12 oxygen atoms for aluminous pyroxene diamond inclusions (open squares) from kimberlites (5, 6, 28, 53, 59) and representative majorite compositions (solid circles) from Tables 1 and 2. The clinopyroxene data show the expected jadeite substitution vector yielding a slope similar to unity. By contrast, Al and Cr in the majoritic garnets are not balanced by Na and K substitutions. Majorite substitution into garnet mainly involves the M2+ cations (Fe, Ca, and Mg). (C) Si versus Al+Cr variation in diamond inclusion clinopyroxenes (open squares) and orthopyroxenes (solid circles) (5, 6, 28, 53, 59). The clinopyroxenes have an average Si4+ pfu of 3.94 ± 0.08 and do not plot on the majorite vector. Rather they define a much steeper vector because Al incorporation has evidently no effect on the Si content of pyroxene. Orthopyroxenes from diamond inclusions or kimberlite-hosted xenoliths are invariably much more siliceous than Malaita majoritic garnets. They have low contents of Al and Cr, and there is no evidence for solid solution toward a more aluminous orthopyroxene.

An important consideration is whether compositional data can be used to discriminate between majoritic garnet-derived material and simple aluminous pyroxene (such as omphacite). Hypothetical end-member majorite and Al-free clinopyroxene have identical compositions (4 Si4+ pfu relative to 12 oxygen atoms), and it would be impossible to distinguish between them based solely on mineral chemistry. This argument is not relevant to the Malaita majorites, as the most silicic compositions have <3.82 Si4+ pfu. Data for aluminous pyroxene diamond inclusions compiled from worldwide kimberlite occurrences and representative majorite compositions from Tables 1 and 2 are compared in Fig. 2B. The clinopyroxene data show the expected jadeite substitution vector yielding a slope similar to unity. By contrast, Al and Cr in the majoritic garnets are not balanced by Na and K substitutions. Majorite substitution into garnet mainly involves the M2+ cations (Fe, Ca, and Mg) which explains the horizontal vector in Fig. 2B. A further consideration is that the clinopyroxene diamond inclusions, when recalculated to 12 oxygens, have an average Si4+ pfu of 3.94 ± 0.08 (Fig. 2C), which is significantly greater than the most silicic majoritic garnet. These clinopyroxenes do not plot on the majorite vector, instead they define a very different, much steeper vector, because Al incorporation has evidently no effect on the Si content of pyroxene.

The only pyroxene compositions that could be confused with majoritic garnet are those that experienced tschermak substitution (Al + Al = Si + M2+). Such tschermakitic pyroxenes are either orthorhombic or monoclinic and are thus birefringent. Except for the most siliceous majorite analyses, the Malaita majorite compositions are invariably isotropic. It is significant that the Malaita majorite has the same low Ca content as coexisting metastable pyrope (much lower than the elevated Ca values that are typical of tschermak-rich clinopyroxene). Only Mg-tschermakitic orthopyroxene could be compositionally similar to majoritic garnet. However, the maximum solubility of Al in orthopyroxene is insufficient to explain the full range in composition observed in the Malaita majoritic material (Fig. 2A). Experiments in the pure MgO-Al2O3-SiO2 system have shown that the maximum Al-solubility in orthopyroxene is ∼18 mole percent (20). This only translates to 3.55 to 3.60 Al ions pfu (per 12 oxygens). Therefore, more than half of the analyses of Malaita majoritic material cannot be explained as aluminous orthopyroxenes. Importantly, in a similar study using natural starting compositions (21), Al and Cr contents in orthopyroxenes were lower by ∼50% than in corresponding material produced from end-member reagents.

Orthopyroxene and spinel can form as a breakdown product of garnet and olivine in the shallow (<80 km depth) mantle. Such high Al-orthopyroxene is only found in very high-temperature reaction products [i.e., at >1500°C (20)] far above the diamond stability field. Orthopyroxenes from within the diamond stability field are more siliceous and have even lower Al contents (Fig. 2C), in agreement with experimental constraints (20). The complete absence of olivine in all Malaita garnetites, the presence of phases of exclusively high-pressure origin (i.e., diamond and aluminous silicate phases), and Al content exceeding maximum solubility in orthopyroxene together argue strongly that the phase with majoritic chemistry is indeed majorite.

Majorite is preserved in several habits. E-type majorite occurs in single veins or sets of anastomosing veins up to 300 μm wide in garnet macrocrysts and megacrysts, and also in composite garnetite xenoliths. In some xenoliths, it also occurs randomly distributed as amoeboid areas up to 100 μm in diameter (Fig. 3A). Garnetites consist of garnet (both majoritic and nonmajoritic) and minor spinel. Neither orthopyroxene nor clinopyroxene are present in these rocks, possibly reflecting bulk composition of the protolith. P-type (Cr-bearing) majorite occurs in a rare garnet peridotite xenolith as crystals up to 30 μm in diameter that form part of a symplectite reaction microstructure with grains of Al-Mg spinel (Fig. 3, B and C). The spinel is similar in composition to high-pressure spinel reported by Liu (22). This microstructure is interpreted to have developed from breakdown of an even higher pressure prograde assemblage of Ca-perovskite, Mg-perovskite, and a phase low in Si and high in Ti and Na that is similar to the ultrahigh pressure “majorite” composition reported in experiments at 27 GPa (Fig. 3B) (23, 24). Development of the majorite-spinel symplectite microstructure may reflect slow decompression during long time-scale convection in the mantle transition zone. This process and evidence for metastable survival may explain the range of majorite compositions within single xenoliths (18). Relationships shown in Fig. 3, B and C demonstrate that formation of the symplectite predated development of glassy kelyphite veins which formed by decompression melting during rapid ascent of the host alnöite. At high scanning electron microscope (SEM) magnification, all high-pressure phases, including those with majoritic chemistry, appear as a very fine grained (submicrometer) polycrystalline aggregate (Fig. 3, D and E). This groundmass also appears to contain relics of resorbed reactant phases. Formation of the fine-grained polycrystalline groundmass is interpreted to be caused by rapid decompression during eruption that induced partial reversion and subsequent reordering.

Figure 3

(A) BSE image of garnetite xenolith KC-99-M5 (60). Scale bar, 100 μm. Picture shows randomly distributed amoeboid areas of prograde E-type majorite (Mj) and Al-Mg spinel (Spl) up to 100 μm in diameter in metastable pyrope (Prp). (B) BSE image of garnet peridotite xenolith KC-98-16 showing relatively large discrete grains of high-pressure majorite? (HP-Mj) and Ca-perovskite? (Ca-pv) as well as relict grains of Cr-pyrope (Prp) and olivine (Ol). Majorite occurs as crystals up to 30 μm in diameter that form part of a symplectite reaction microstructure with grains of Al-Mg spinel. Formation of the symplectite predated development of 100-μm-wide glassy kelyphite veins, which formed by decompression melting during rapid ascent of the host alnöite. Scale bar, 100 μm. (C) BSE image of garnet peridotite xenolith KC-98-16 at higher magnification showing details of the Mj-Spl symplectite microstructure and discordant relationships with the decompression induced kelyphytic glass that contains remnants of reactant phases. Scale bar, 100 μm. (D) SE image of phases with Mg-perovskite chemistry in garnet peridotite xenolith KC-98-16. Scale bar, 10 μm. (E) SE image of phases with Mg-perovskite chemistry at high magnification showing relics of resorbed reactant phases in a submicrometer polycrystalline groundmass that is interpreted to be caused by rapid decompression during eruption that induced partial reversion and reordering. Scale bar, 1 μm.

Majorite is a complex garnet solid solution between pyrope-almandine and orthopyroxene or clinopyroxene that forms at pressures between ∼12 and 26 GPa in both basaltic and peridotitic protoliths. With increasing pressure, garnet with ∼3 Si4+atoms pfu becomes progressively depleted in Al and enriched in Si (up to 3.83 Si4+ pfu), M2+ (Ca2+ + Mg2+ + Fe2+) cations, and Na+(25–27). Majorite has occasionally been observed as rare inclusions in diamond (28). Intergrowths of garnet and clinopyroxene or orthopyroxene in eclogite and peridotite xenoliths have been reported from a number of cratonic kimberlite pipes (3,4) and from units of garnet-pyroxenite and garnet peridotite in obducted ultramafic units in orogenic belts. These intergrowths have been interpreted to be pyroxene exsolved from majoritic garnet (29). The occurrence of majorite is significant, because this phase is considered to be a dominant constituent of the transition zone between depths of 400 and 650 km (25, 26, 30–35).

Compared to all other reported natural majorites, the Malaita majorites include significantly more silicic (Si4+ up to 3.83 pfu) compositions and are thus interpreted to have been derived from much greater depths. These majorites appear to have low first-order (gray) birefringence, possibly indicating that their structure deviates from cubic symmetry (36). Malaita xenoliths provide the first opportunity to directly study the composition and mineralogy of the mantle transition zone and its lower mantle boundary in the pressure range of ∼12 to 27 GPa.

Other High-Pressure Phases

Experiments imply that in the pressure range between 15 and 22 GPa, E-type majorite is accompanied by metastable pyrope and stishovite (25). At even greater depths between 22 and 27 GPa, both E- and P-type majorite react to form a less silicic phase with 3.18 to 3.32 Si atoms for 12 oxygens that was still termed majorite by (23) [but see (24)]. A small population of “garnets” (Fig. 2A) with ∼3 Si pfu, but clearly deficient in Al+Cr (1.85 to 1.55), may be this ultradeep aluminous phase that forms at depths of ∼670 km. In experiments at pressures >22 GPa, these “ultrahigh pressure majorites” coexist with Ca silicate perovskite. At pressures >24 GPa, several enigmatic aluminous phases appear, and at pressures >25.5 GPa, Mg-Al silicate perovskite becomes a stable phase (22, 23, 26, 31, 32, 34, 37,38). All of these phases occur in the Malaita samples (Table 3) and are comparable in composition to those synthesized by (23) between 24 and 27 GPa.

Other high-pressure phases in Malaita samples include pyroxene-ilmenite-rutile intergrowths (39), jadeite-rich clinopyroxene, picroilmenite (MgO, 3.55% to 8.31%), large (∼2 cm by 1 cm) “kimberlitic” zircons, Fe-Ni sulfides, a SiO2phase, and microdiamond surrounded by narrow rims of graphite. A SEM image of one of the microdiamonds that occurs as an inclusion in E-type pyrope and compositional profiles for Si, Mg, Ca, and Fe are shown in Fig. 4. The presence of Fe and Ca in the analytical profile of the diamond is interpreted to arise from the transmission of x-rays through the diamond into the garnet substrate. The preference for Fe and Ca rather than Si and Mg is related to the x-ray transmission characteristics of diamond, which transmits shorter wavelength x-rays more efficiently than those with longer wavelengths (40).

Figure 4

(A) An SE image of a microdiamond from KC-98-M35 that occurs as an inclusion in E-type pyrope. Scale bar, 10 μm. (B) Electron microprobe compositional profiles for Si, Mg, Ca, and Fe across microdiamond. The Fe and Ca counts are likely produced in the garnet substrate and are observed relative to the lighter elements, because of preferential transmission through diamond of shorter wavelength x-rays (40). In addition to the probe results, Raman spectroscopy has also confirmed that this phase is carbon.

Infrared Spectroscopy of Ultrahigh Pressure Materials

The infrared reflectance spectrum of the phase with Mg perovskite-chemistry in the garnet peridotite xenolith (Table 3) is shown in Fig. 5A, where it is compared with spectra of diopside and enstatite (41). The perovskite-chemistry phase has broad reflectance peaks near 960 and 880 cm–1, which are associated with stretching vibrations of silica tetrahedra, and weaker bands between 600 and 700 cm–1, which are probably associated with Si-O-Si bending vibrations (42, 43). The peaks in the spectrum of the phase with Mg-perovskite chemistry are substantially broadened relative to the peaks that lie at grossly similar frequencies in the pyroxene samples. Moreover, there are mutually exclusive features between the spectra, such as the absence of a discrete peak at 990 to 1010 cm–1 in the analyzed sample and the presence of the feature near 740 cm−1. Indeed, the reflectance spectrum of the perovskite chemistry phase is substantially different from those of the pyroxenes. This difference cannot be attributed to particle-size effects, because the change in the reflectance spectra of pyroxenes due to particle size reduction is different from those documented in Fig. 5A (44). At higher frequencies (>∼1100 cm–1), there is clear evidence for a scattering background in the spectrum, implying that local heterogeneities are present at scale lengths of several micrometers and smaller (Fig. 3E). Given the breadth of the features in the spectrum, the most plausible interpretation is that the phase with Mg-perovskite chemistry represents a disordered material with pyroxene stoichiometry. This is consistent with the likely presence of silicate chains within the material, based on the gross similarity of the positions of the reflectance peaks above 800 cm−1 with those of pyroxenes. The presence of such a disordered, silicate chain-containing material in these samples is not surprising: the single other observation of naturally occurring reverted silicate perovskite shows the clear presence of a disordered phase of pyroxene stoichiometry (45). We believe that the present observations are analogous to the meteoritic results (albeit with reversion having likely occurred on a substantially longer time scale in the xenolith samples).

Figure 5

(A) Unpolarized infrared reflectance spectra of thin sections of natural samples of diopside (from Bird Creek, Ontario), enstatite (Bamble, Norway) and of a 50-μm-diameter spot in the vein of perovskite chemistry within sample KC-98-16. Although the relative amplitudes of peaks may vary depending on crystallographic orientation of the sample material, both the magnitude of broadening and the comparatively small number of reflectance peaks in the spectrum of the vein of perovskite chemistry indicate that substantial disorder is present within this sample. (B) Unpolarized infrared reflectance spectrum of a 50-μm-diameter spot of pyropic garnet from sample KC-98-M35 compared with a comparable spot size in the majorite-chemistry region of sample KC-98-16. The spectrum of pyropic garnet is in accord with previous spectra of garnets near the pyrope-almandine join (61). In contrast, the spectrum of the majorite-chemistry region shows some similarity to those of chain silicates (43), indicating that some high temperature reversion of this material has taken place.

It is well known that at ambient pressures and laboratory time scales, magnesium silicate perovskite reverts to an amorphous phase or enstatite at temperatures of 750 to 850 K (46). Similarly, small particles of glasses of pyroxene composition have been observed to form from perovskite in ∼300 hours at 1073 K, and in 7 min at 1273 K (47). Over shorter time scales, broadened pyroxene-like peaks are observed in disordered samples quenched from high-temperature conditions (47). These results are in general accord with those shown in Fig. 5A and provide a likely genetic history for the formation of high-pressure phases in the xenoliths. We interpret this to involve initial reversion (during ascent) of perovskite to an amorphous phase, followed by subsequent (and partial) temperature-induced ordering of the amorphous material. The degree of ordering/microstructural development within these samples may ultimately provide a constraint on their time-temperature-transformation history during ascent from depth.

The spectra of the zones of majorite chemistry have some similarities with that of MgSiO3-clinoenstatite (43), although as with the zones of perovskite-chemistry, the peaks are significantly broadened and moderately shifted relative to the end-member. Specifically, MgSiO3-clinoenstatite has peaks at 1074, 1017, 951, 900, 860, and 690 cm−1. In the spectrum shown in Fig. 5B, the comparatively sharp features at 1017 and 900 cm–1 are absent (or broadened into weak shoulders), and the peaks at 1055, 960, and 858 cm–1 are significantly broadened relative to the end-member spectrum (this broadening is not a consequence of comparing the reflectance spectra of Fig. 5B with the absorbance spectra of (43), as Kramers-Kronig transforms of the spectra of Fig. 5B retain substantially broadened peaks). In this instance, the broadening could be produced by a combination of Al substitution in the clinopyroxene and significant structural disorder within the material induced by reversion from a high-pressure garnet phase. Unfortunately, few data exist on the effect of Al on the infrared spectra of clinopyroxenes, and the resultant degree of “transformational” disorder is difficult to assess.

Majorite Geobarometer

Substitution of Si for Al+Cr with increasing pressure can be used as a geobarometer of equilibration in the mantle (Fig. 6A) (4, 30). Exchange between Si and Al+Cr for the experimental data is well correlated with the range of substitutions shown by the Malaita majorites (Fig. 6B). The highest pressure experimental compositions show the highest Si pfu and lowest Al+Cr pfu. We have developed empirical geobarometers for both substitutions from published data (48). Very good agreement between Si and Al+Cr substitution in the natural and synthetic datasets demonstrates that these barometers are applicable to both P- and E-type majorites. The barometers are P (GPa) = −50.7 + 18.97(Si) [r 2 = 0.886] (Fig. 6C) and P (GPa) = 23.7 − 9.06(Al+Cr) [r 2 = 0.890] (Fig. 6D).

Figure 6

(A) Si for Al+Cr variation in garnet and majorite. Data from relevant high-pressure experimental data for basaltic and ultramafic systems in the pressure interval 5 to 18 GPa: open triangles, Herzberg and Zhang (57); open circles, Akaogi and Akimoto (32); solid circles, Kato et al. (37); solid squares, Canil (32); open squares, Walter (57). (B) Si and Al+Cr exchange exhibited by the experimental data (open triangles) is well correlated and overlaps the range of substitutions shown by the Malaita majorites, with the highest pressure experimental compositions showing highest Si pfu and lowest Al+Cr pfu. (C) Correlation between experimentally determined Si4+ pfu and quoted experimental pressure estimates. Linear regression of the data yields the following empirical barometer:P (GPa) = −50.7 + 18.97(Si). (D) Correlation between experimentally determined Al+Cr pfu and quoted experimental pressure estimates. Linear regression of the data yields the following empirical barometer: P (GPa) = 23.7 – 9.06(Al+Cr). (E) Comparison of pressures calculated using the empirical Si and (Al+Cr) barometers. Methods agree to better than 1 GPa [r 2 = 0.987]. (F) Comparison of experimental pressure estimates, and pressures calculated using the Si barometer (solid circle) and (Al+Cr) barometer (open triangle). (G) Pressure estimates for Malaita majorite-bearing xenoliths and macrocrysts. Shown for comparison are pressures calculated for majorite diamond inclusions reported from Monastery Mine (28) (open triangles) and Jagersfontein (53) (open circles) and majorite reconstituted from garnet-clinopyroxene (orthopyroxene) intergrowths in xenoliths from Jagersfontein (3, 4) (open squares). Pressures calculated for majorite diamond inclusions lie along the same vector as the Malaita majorites and yield a maximum of 15 GPa. Note that the majorite compositions from reconstituted xenoliths yield systematically higher pressures by the Si barometer than the (Al+Cr) barometer. This suggests that SiO2 or (Al2O3+Cr2O3) in the garnet and pyroxene mixing proportions for the xenoliths (3,4) were overestimated.

Pressures calculated from the experimental data using each barometer agree to better than 1 GPa [r 2 = 0.987] (Fig. 6E). The barometers are apparently independent of temperature, and calculated pressures are in good agreement with those determined in experiments (Fig. 6F). For the total Malaita population (Fig. 6G), average agreement between P(Si) andP(Al+Cr) is ±0.5 GPa. The uncertainty is within analytical error and reflects largely the quality of mineral chemical data. The barometer implies that the majorite-bearing Malaita xenoliths and macrocrysts are derived from the deep upper mantle in the pressure-depth interval 9.5 GPa (260 km) to 22 GPa (570 km).


A paleogeotherm at ∼34 Ma for the Ontong Java Plateau has previously been determined from pyroxene geothermobarometry for samples from Malaita (14, 17). The paleogeotherm is considerably hotter than typical cratonic geotherms and indicates that diamond could not have been stable in the Ontong Java Plateau lithospheric mantle, or oceanic lithosphere in general, during either the eruption of the Ontong Java Plateau basalts or that of the ∼34 Ma alnöites. Diamond could, however, have been stable at depths >150 km. In fact, microdiamonds do occur in pyrope and majorite-bearing Malaita garnet macrocrysts and majorite-bearing E-type garnetite xenoliths. This observation confirms that diamonds are present at least locally in the transition zone (49). E-type diamond in this tectonic setting could only have formed during subduction of the eclogitic precursor or during (transient) storage of slabs in the transition zone (33, 50).

Previous studies have interpreted Malaita alnöites to have formed by fractional crystallization of a proto-alnöitic magma in the asthenosphere (16, 17, 51). Malaita peridotite xenoliths, rare eclogite xenoliths, and a diverse suite of ilmenite, garnet, pyroxene, and zircon macrocrysts were interpreted to indicate maximum source depths of 140 km (10, 12, 14, 16,17, 51, 52). However, the occurrence of majoritic and microdiamond-bearing xenoliths and macrocrysts forming from Mg-Si perovskite-bearing precursors, clearly requires the source of the host alnöite to be deeper than 670 km in the lower mantle (50) or some major upward transport of material before eruption. Thus, macrocrysts do not form by an unconstrained fractional crystallization process of the alnöitic magma at the interface between the lithospheric and asthenospheric mantle (16).

Malaita's Si-rich majorite-bearing xenoliths and macrocrysts are derived from depths of up to 570 km (22 GPa), which is significantly deeper than majorite diamond inclusions reported from Monastery Mine (28) and Jagersfontein (53), namely, 9 GPa (262 km) to 15 GPa (413 km) and 7 GPa (204 km) to 11 GPa (303 km) determined using the barometer we present here (Fig. 6G). Additional high-pressure phases from Malaita such as Ca and Mg-Si perovskite indicate substantially greater depths (770 km).

The presence of TiO2- and Na2O-rich eclogitic pyropes in the Malaita garnetite xenoliths, with compositions identical to E-type garnets in diamond inclusions, indicates significant metastable preservation of pyrope garnet in the transition zone. Such preservation could occur because temperatures in subduction zones are relatively low. In addition, the ultradeep slabs may be anhydrous, and the resultant amount of chemical interaction between pyrope and pyroxene could be limited by lithology and/or kinetics. The existence of such metastability in the transition zone suggests that inferences based on silicate minerals that occur as diamond inclusions may underestimate the depth of formation of diamond. In other words, the apparent absence of majorite does not preclude formation of diamond in the mantle transition zone.

Metastability may also be significant for the interpretation of the seismic structure of the mantle. The region above the base of the upper mantle is defined seismically by extremely rapid seismic velocity increases (54) and by compressional and shear-wave velocities that progressively increase from ∼9.5 to 10 km/s for P waves and 5.0 to 5.5 km/s and S waves, respectively (55). The progressive change in seismic velocities could reflect the fact that kinetically inhibited phase transformations are unlikely to occur at specific depths. Also, the preponderance of E-type ultradeep xenoliths indicates that the mantle transition zone which was first cut by the Ontong Java Plateau plume and subsequently sampled by lower mantle-derived alnöites is likely to be dominated by subducted oceanic crust. This appears to be confirmed by seismic tomographic images for the southwestern Pacific region that clearly show a zone of high seismic velocities lying east of New Guinea (56). This may represent the site of accumulation of subducted Pacific oceanic crust. The other significant interpretation from the southwestern Pacific/south Asian mantle tomography (56) is that at 400 km depth, the transition zone is not uniformly cool. Instead, it exhibits very pronounced lateral gradients in temperature and possibly also in mineralogy. The transition zone may therefore represent a thermally and compositionally variable, yet volumetrically extremely significant geochemical mantle reservoir.

  • * To whom correspondence should be addressed. E-mail: k.collerson{at}


View Abstract

Navigate This Article