REVIEW

Perovskite in Earth’s deep interiorKei Hirose,1,2* Ryosuke Sinmyo,1,2 John Hernlund1

Silicate perovskite-type phases are the most abundant constituent inside our planet andare the predominant minerals in Earth’s lower mantle more than 660 kilometers belowthe surface. Magnesium-rich perovskite is a major lower mantle phase and undergoes aphase transition to post-perovskite near the bottom of the mantle. Calcium-rich perovskiteis proportionally minor but may host numerous trace elements that record chemicaldifferentiation events. The properties of mantle perovskites are the key to understandingthe dynamic evolution of Earth, as they strongly influence the transport properties of lowermantle rocks. Perovskites are expected to be an important constituent of rocky planetslarger than Mars and thus play a major role in modulating the evolution of terrestrialplanets throughout the universe.

Perovskite-type crystals, or perovskites, arefamous for their chemical flexibility andwide range of industrial applications asferroelectrics, piezoelectrics, superconduc-tors, and photovoltaic materials. The min-

eral perovskite, named for Russian mineralogistLev Perovski in 1839, has a chemical formulaCaTiO3 and was discovered in the southern UralMountains. Perovskite, for research-ers studying Earth, refers to the iso-structural MgSiO3 or CaSiO3 minerals(called Mg-perovskite or Ca-perovskitehereafter) believed to make up thebulk of the lower mantle. Intact Mg-perovskite crystals have also beendiscovered in shocked veins in theTenham meteorite (1, 2) and named“bridgmanite” after pioneering high-pressure physicist Percy Bridgman.In ABX3-type materials with ideal

perovskite structure, ions A and X com-prise close-packed face-centered cu-bic structures, with A ions sited in thelarger spaces between corner-linkedBX6 octahedra (Fig. 1, left). These well-packed structures are favored at highpressure, and major elements in ter-restrial planets (Mg, Si, and O) formMg-perovskite under the conditionsprevailing in their deep interiors. Mg-perovskite is stable above ~23 GPaand was first synthesized in the 1970sin the laboratory (3, 4). Mg-perovskiteis a predominant phase in Earth’slower mantle (depth of 660 to 2600to 2890 km) and is also expected tobe an abundant mineral in other terrestrialplanets larger than Mars. The formation of Mg-perovskite develops an important dynamicalboundary inside planets because it greatly in-creases density and viscosity (5).

Mineralogy of Earth’s lower mantle andpost-perovskite phase transition

Seismology shows a layered structure in Earth’smantle, with marked velocity differences be-tween the upper mantle (to 410-km depth), tran-sition zone (410- to 660-km depth), lower mantle

(660- to 2600-km depth), and lowermost mantle(2600- to 2890-km depth). Pyrolite is a theoret-ical chemical composition and similar to those ofleast-differentiated upper mantle rocks. Pyroliteis often assumed to represent a typical compo-sition for the entire mantle because phase transi-tions in primary minerals that occur in a pyroliticcomposition align with the layered structure of

the mantle (Fig. 2A). However, seismic-wavespeeds are not well explained by a pyrolitic man-tle, possibly suggesting that Earth’s mantle is aninhomogeneous mixture of basaltic and harzbur-gitic rocks (6).The occurrence of Mg-perovskite at about

660-km depth is readily imaged by seismic wavesas a global discontinuity. In a pyrolitic mantle,Mg-perovskite is formed from both the primaryand secondary minerals in the transition zone,(Mg,Fe)2SiO4 ringwoodite (a spinel-type mineral)and (Fe,Al)-bearing MgSiO3 majorite (from the

garnet family), upon increasing pressure (Fig.2B). As a consequence, Mg-perovskite constitutesa large proportion of the lower mantle, co-existing with (Mg,Fe)O ferropericlase and Ca-perovskite (~80, ~15, and ~5 volume % in pyrolite,respectively). The formation of Mg-perovskite isassociated with an increase in the coordinationnumber of Si from four to six. This leads to anegative Clapeyron (dP/dT) slope, where P ispressure and T is temperature, because the in-crease in the coordination number increasesfirst-neighbor Si–O distance and thus increasesentropy (7). The Clapeyron slope of the post-spinelphase boundary (ringwoodite to Mg-perovskitewith ferropericlase) was first determined to be–3 MPa K–1 with a conventional “quench”methodin a multianvil press (Fig. 2B), consistent withthe seismologically determined dP/dT slope ofthe 660-km boundary (8). This value has beencontroversial (–0.4 to –3 MPa K–1), because pre-cise determination of pressure at high temper-ature is usually dependent on the P-V-T equationof state of an internal standard, but the accuracyand consistency of the equation of state for var-ious kinds of standards are matters of debate (9).The incorporation of a relatively small Mg2+

ion between the SiO6 octahedra in Mg-perovskitecauses an orthorhombic distortion ofthe typical cubic perovskite structure(Fig. 1, left). This distortion is enhancedwith increasing pressure (fig. S1) andeventually leads to a phase transitioninto “post-perovskite” (Fig. 1, right) above120 GPa (Fig. 2C) (10–12). The lower-most mantle (D″ layer) is marked bya shear velocity jump intermittentlyaround 2600-km depth that matches120 GPa, mainly at relatively coldregions. The calculated sound velocityof post-perovskite is consistent withthe seismological observations of an~3% increase in shear velocity andalmost no change in compressionalvelocity for discontinuities imaged inthe D″ layer (13). Nevertheless, theeffects of minor components such asFeO, Fe2O3, and Al2O3 on the pres-sure and the width of the transitionare controversial (Fig. 2C). Ohta et al.(14) and Fiquet et al. (15) reported asharp (≤5-GPa width) boundary inpyrolite at 120 to 138 GPa and 2500to 3500 K, consistent with the seis-mic velocity jump observed around2600-km depth. Conversely, the phase

transition was found to occur over a very wide(>20 GPa) pressure range with the incorporationof Fe and Al (16). Grocholski et al. (17) foundthat the post-perovskite phase transition doesnot occur in pyrolite at mantle pressures, butmay occur in basalt and harzburgitic mantle bulkcompositions.

PEROVSKITES

Hirose et al., Science 358, 734–738 (2017) 10 November 2017 1 of 5

1Earth-Life Science Institute, Tokyo Institute of Technology,Meguro, Tokyo 152-8550, Japan. 2Department of Earth andPlanetary Science, Graduate School of Science, TheUniversity of Tokyo, Bunkyo, Tokyo 113-0033, Japan.*Corresponding author. Email: kei@elsi.jp

Fig. 1. Crystal structures of MgSiO3 perovskite and post-perovskite.SiO6 coordination polyhedra of oxygen atoms (white spheres) aroundsilicon atoms are shown as blue octahedra, and the Mg ions are shownas yellow spheres. Both are orthorhombic crystal lattices with spacegroup Pbnm for Mg-perovskite and Cmcm for post-perovskite.

on Novem

ber 9, 2017

http://science.sciencemag.org/

Dow

nloaded from

Dynamics in the lower mantleThe negative Clapeyron slope of the post-spineltransition resists penetration of mantle convec-tion between the transition zone and the lowermantle. This, combined with a viscosity increase,causes the stagnation of subducting pieces offormer oceanic lithosphere called “slabs” around660 km depth (8). Slab stagnation occurs also at~1000 km depth, corresponding to a viscosity“hill” (18) possibly related to iron spin crossover(19) or iron depletion in Mg-perovskite (20). Slabsmay also stagnate on the top of bridgmanite (Mg-perovskite)–enriched ancient mantle structures(BEAMS) (21) (Fig. 3). By contrast, upwelling hotplumes do not seem to stagnate at the 660-kmboundary, possibly because Mg-perovskite trans-forms into majorite at high temperatures (>2100 K)with a positive Clapeyron slope (22) (Fig. 2B).Conversely, the post-perovskite phase transfor-mation in the lowermost mantle exhibits a largepositive dP/dT slope and thus assists mantleconvection (11, 12, 23). If post-perovskite is pres-ent in plume-upwelling regions, the phase tran-sition to Mg-perovskite plays an important rolebecause it occurs near the lowermost mantlethermal boundary layer. The phase transitiontherefore destabilizes the lowermost mantle andenhances the number of plumes, possibly alsoincreasing the temperature of the whole mantleby a few hundred degrees (24).Laboratory experiments have demonstrated

that Mg-perovskite is much more rigid thanupper mantle and transition-zone minerals andis three orders of magnitude more viscous thancoexisting MgO periclase (5). Indeed, the viscos-ity of the lower mantle based on geodetic mea-surements has been estimated to be at least oneorder of magnitude higher than shallower levelsand is often assumed to be more viscous by afactor of ~30 (25). If periclase can be intercon-nected between stronger Mg-perovskite grainsto weaken the whole rock, this would lead tocomplex and interesting behaviors. In a well-connected state, changes in the iron spin statein (Mg,Fe)O ferropericlase could influence theviscosity of the weak phase and manifest them-selves as a higher-viscosity layer in the lower man-tle (19). The transformation into a post-spinelassemblage (Mg-perovskite with periclase) re-veals little interconnectivity between periclasegrains (26). However, recent deformation thepost-spinel assemblage suggest that periclasegrains might become interconnected and weakenthe assemblage at larger strains, and could lead todramatic weakening. This kind of strain-weakeningbehavior is important because it may give riseto shear localization in the lower mantle, in whichnarrow zones of intense deformation separatelarge quiescent regions that undergo relativelylittle deformation (see below). Strain causes weak-ening that in turn begets more strain, leadingto a runaway effect similar to what is thoughtto accommodate plate tectonics. In the deepestportions of Earth’s mantle, post-perovskite isexpected to be much weaker than perovskiteowing to its layered crystal structure (28). Be-cause it is stabilized at lower temperatures, such

weak post-perovskite may weaken rocks thatwould otherwise be highly viscous owing tothe usual influence of cooler temperatures onviscosity.

Chemical and physical properties ofmantle perovskite

Impurities, such as Al and Fe, can strongly in-fluence the behavior of Mg-perovskite in Earth’slower mantle. Although iron is predominantlyFe2+ in the upper mantle and the transition zone,it is able to coexist as Fe0, Fe2+, and Fe3+ in thelower mantle. This is in part because Al3+ ispreferentially accommodated into the small Si4+

site (B site) in Mg-perovskite, which induces thedisproportionation of Fe2+ (→Fe3+ + Fe0) as Fe3+

enters the Mg2+ site (A site) in order to maintaina charge balance. Indeed, relics of natural Mg-perovskite found in diamonds contain a largeamount of Fe3+ (29), although the diamonds areformed under environments that may not rep-resent the typical redox conditions of the man-tle (30). Recent experiments found that ironin Mg-perovskite is dominantly Fe2+ at 40 to70 GPa, whereas Fe3+/(Fe2+ + Fe3+) = ~0.6 inother pressure ranges (20). Because ferrous ironis preferentially accommodated into coexisting

ferropericlase, it may cause iron depletion in Mg-perovskite at 1100- to 1700-km depth. Shim andothers speculate that the observed mid–lowermantle viscosity anomaly (20) might be a resultof iron depletion in Mg-perovskite.The spin state of iron in Mg-perovskite changes

from high to low spin with increasing pressure.Such an iron spin crossover was first discoveredin (Mg,Fe)O ferropericlase at the pressure rangeof the lower mantle (31) and has important geo-physical implications (32). The iron spin state inMg-perovskite is more complicated owing to mul-tiple valence states and crystallographic sites. Fe2+

is accommodated only in the A site and undergoesthe spin crossover above ~30 GPa from high spinto what was identified as intermediate spin (33, 34).However, a component that was assigned tointermediate-spin Fe2+ in Mössbauer spectra maybe explained by high-spin Fe2+ in the distorted Asite (35). Conversely, Fe3+ that occupies the B siteundergoes high-spin to low-spin crossover above30 GPa (36). The valence and spin states controlthe partitioning of iron between Mg-perovskiteand ferropericlase (31). Recent work suggeststhat it changes around 60 GPa because of thespin crossover (37), but the effect is much smallerthan previously thought. Recent approaches, such

Hirose et al., Science 358, 734–738 (2017) 10 November 2017 2 of 5

Fig. 2. Phase proportions and phase transition boundaries in Earth’s lower mantle. (A) Mineralproportions in pyrolytic mantle rocks. Ol, olivine; Wd, wadsleyite; Rw, ringwoodite; Cpx, clinopyroxene;Opx, orthopyroxene; Mj, majorite (garnet). (B) Phase transition boundaries near 660-km depth inpyrolitic mantle and basalt compositions (22, 88). Mg-perovskite is formed in pyrolite by both post-spinel(ringwoodite to Mg-perovskite and ferropericlase) (green) and majorite-perovskite transition (blue).Majorite transforms into Mg-perovskite in basalt (orange). Two-phase coexisting regions are shown asbands. Pv, perovskite. (C) Phase boundary between Mg-perovskite and post-perovskite is illustratedfor MgSiO3 (yellow line) (23), pyrolytic mantle (blue), and subducted basalt (orange) (14, 17). (D) Phasetransition boundary in Ca-perovskite. Both theory (72) and experiment (71) demonstrated tetragonal tocubic phase transition at high temperatures.

on Novem

ber 9, 2017

http://science.sciencemag.org/

Dow

nloaded from

as theoretical predictions (38) and the observationof natural samples (39), provide further insightabout changes in iron partitioning.Mg-perovskite is much less compressible (bulk

modulus at ambient condition, K0 = 258 GPa)(40) than coexisting ferropericlase (K0 = 158 GPa)(41). The measured Vickers hardness of Mg-perovskite is 18 GPa, several factors higher thanthat of olivine, enstatite, and periclase (42). Theeffects of impurities (Fe2+, Fe3+, and Al3+) on Kare not yet clear (43, 44), likely because of themultiple site distributions and electronic statesof iron. Unlike the remarkable volumetric andbulk modulus decrease of ferropericlase from thespin crossover, even at high temperatures, theFe3+ spin transition in Mg-perovskite has littleeffect upon its compression behavior (43, 44).The combination may explain mid–lower mantleseismic anomalies without introducing compo-sitional heterogeneity (32).The thermal properties of Mg-perovskite are

key to understanding the thermal structure ofthe lower mantle and the heat transported awayfrom the core by mantle convection. The thermalexpansion coefficient and heat capacity of Mg-perovskite largely determine the adiabatic tem-perature profile in the lower mantle (45). Althoughthe heat capacity is difficult to measure at highP-T, theory predicts an almost constant valueat >1000 K regardless of pressure (46). Recentprogress also includes the measurement of ther-mal diffusivity of Mg-perovskite (47, 48), from

which the thermal conductivity of the lower-most mantle was determined to be about 10 Wm−1 K−1. Lattice thermal conductivity is dimin-ished at higher temperature but enhanced athigher pressure (47, 48), and thus, the thermalconductivity may remain similar throughoutthe lower mantle. The contribution of radiativethermal transport through Mg-perovskite still re-mains controversial (49, 50). Combining suchthermal conductivity with temperature gradientsat the base of the mantle, which may be steeperthan the post-perovskite phase transition bound-ary (51), a global heat flow across the core–mantleboundary (CMB) is estimated to be more than7 TW (23), suggesting rapid cooling of Earth’score. Depending on the pressure scale (52), theClapeyron slope of the boundary can be smaller,leading to the larger CMB heat flow.The sound velocity of Mg-perovskite has been

determined at lower mantle pressures by Brillouinscattering measurements (53, 54) and theoreticalcalculations (55, 56). A comparison of shear andcompressional velocities of Mg-perovskite andferropericlase with seismological observations cantest the pyrolitic mantle compositional model.Murakami et al. (53) argued that Mg-perovskiteconstitutes more than 90% in lower mantle rocks,although this may be an overestimate (57). Itindicates a chondritic-like Mg/Si ratio close to1.0, which is different from the observed valueof 1.3 in the pyrolitic upper mantle. Consideringthe large volume of the lower mantle, a smaller

Mg/Si ratio would help reconcile the bulk sili-cate mantle with the chondritic compositionalmodels for Earth’s formation (Box 1).Electrical conductivity is an observable prop-

erty of Earth’s interior obtained from temporalfluctuations in the magnetic field (58). Conductivityis sensitive to water and iron content, so it pro-vides specific information on the chemistry ofthe mantle. The conductivity of (Fe,Al)-bearingnatural Mg-perovskite measured in a multianvilpress is higher than those of transition-zoneminerals because of a greater sensitivity to tem-perature (59). Although the influence of the spincrossover in Fe3+ is controversial (60, 61), it shoulddiminish electrical conductivity as is found inferropericlase, owing to the reduction in thenumber of unpaired electrons. The discrepancysurrounding the impact of the spin crossovereither allows for the simple pyrolytic composi-tion (61) or requires the inclusion of subductedbasaltic crust in the lower mantle (60). Post-perovskite with a layered crystal structure (Fig.1B) exhibits more than two orders of magnitudehigher conductivity than Mg-perovskite (62).This should enhance electromagnetic couplingbetween the lowermost mantle and the liquidouter core, which may help to explain variationsin day length on decadal time scales (63).The presence of H2O in crystal lattice affects

melting temperature, viscosity, and electricalconductivity to larger extents. Multianvil experi-ments (64) reported 8000 parts per million (ppm)

Hirose et al., Science 358, 734–738 (2017) 10 November 2017 3 of 5

Fig. 3. Mg-perovskite–enriched ancient mantle structures in thelower mantle. Numerical simulations (21) demonstrated that Mg-perovskite–enriched (low Mg/Si) domains are strong and organize mantleconvection patterns. They are encapsulated by weaker pyrolitic (high Mg/Si) material, which circulates between the shallow and deepest mantle

through channels between the strong domains. The Mg-perovskite–richdomains are unmixed throughout the history of Earth, which may reconcilethe fixed positioning of plume centers and the slab stagnation at~1000-km depth and provide a reservoir to host primordial geochemicalsignatures.

PEROVSKITES on N

ovember 9, 2017

http://science.sciencem

ag.org/D

ownloaded from

H2O in Al-bearing Mg-perovskite, suggesting thatthe lower mantle is potentially a large water res-ervoir (65). These experiments are consistentwith geochemical estimates of high volatile con-centrations with up to ~1000 ppm H2O in thesource mantle for oceanic island basalts, possi-bly originating from the lower mantle (66). Con-versely, earlier experiments (67) and calculationsfound a much lower solubility of water in Mg-perovskite (a concentration of 100 ppm or less)and suggested that the high concentrations ofH2Omeasured in other experiments were pres-ent in fluid inclusions instead of dissolvedinto the mineral itself. Al-bearing post-perovskite was predicted to accommodate5 to 10 timesmoreH2O thanMg-perovskite(68). Townsend and others (68) arguedthat the dehydration of post-perovskiteupon transformation to Mg-perovskite athigh-temperature regions induces partialmelting, which is observed as an ultralow-velocity zone (69). Experiments also sug-gested high solubilities of noble gases inMg-perovskite: up to ~1 weight % (wt %)He and Ar and some Kr, but not Xe (70).

CaSiO3 perovskite: A host oftrace elements

Because the ionic radius of Ca2+ is muchlarger than that of Mg2+, Mg-perovskiteforms a very limited solid solution withCaSiO3 component, and therefore, Ca-perovskite is present as a separate phase.The ratio between the ionic radii of A andB in ABX3-type perovskites, called “toler-ance factor,” determines the type of pe-rovskite structures proposed by VictorGoldschmidt. The structure of Ca-perovskiteexhibits cubic symmetry at the high tem-peratures of the lower mantle (>1900 K)(45) but is possibly distorted in subductedcold materials (71, 72) (Fig. 2D). Theorypredicted that Ca-perovskite transformsinto post-perovskite–type phase or disso-ciates into CaO and SiO2, depending ontemperature, above 500 GPa (73).Ca-perovskite is proportionally minor

but is likely the predominant host for large ionlithophile (silicate-loving) elements (LILE) in thelower mantle (74, 75). Because of the high Ca-perovskite/melt partition coefficients (>10 formany of the LILE), partial melting in the lowermantle causes strong enrichment of LILE in Ca-perovskite–bearing solid residues. Although thestability of Ca-perovskite above the solidus tem-perature is very small in a pyrolite composition,it is one of themajor constituents (~20 volume%)and the last melting phase in basaltic crusts sub-ducted into the lower mantle. Indeed, the melt-ing temperature of basaltic materials is lower thanthat of pyrolite (76), and if they undergo partialmelting at the CMB, Ca-perovskite–rich residuesare formed with LILE-enriched characters. Oncethey are upwardly entrained in plumes, meltingof such blocks in the shallow mantle could pro-duce characteristic geochemical signatures inmagmas; strong enrichment in Th andU relative

to Pb in Ca-perovskite will result in radiogenicPb isotopes. Incorporation of U and Th into Ca-perovskite has also been examined by using the-oretical calculations (77).

Perspective: Challenges and new views

The CMB temperature is about 4000 K, butmany chemical and physical properties of Mg-perovskite are known only for moderate tem-peratures. For example, thermal conductivity wasdetermined to 1073 K (47), and sound velocity(53), electrical conductivity (60), and iron parti-

tioning (37) were obtained only below ~2700 K.In addition, most measurements at lower mantleP-T conditions were performed in a laser-heateddiamond-anvil cell (DAC), with an intrinsicallylarge temperature gradient and different sam-ple geometries, which might be major sourcesof inconsistency for results reported by differentgroups. Advanced experiments with different heat-ing techniques—such as multianvil apparatus withsintered-diamond cubes (26), resistance-heatedDAC, or shockwaves—will help improve under-standing of the nature of the bottom of the mantlethat controls the dynamics and evolution of bothmantle and core. Rheological properties of Mg-perovskite will also be examined at deep lowermantle P-T conditions in the near future, withadvanced experimental techniques using, for ex-ample, the rotational Drickamer apparatus (27),the Kawai-type deformation-DIA apparatus (78),and rotation DAC (79).

Geodynamical models suggest that the wholemantle becomeswell-mixed bymantle convectionover ~100 million–year time scales with the as-sumption of a simple viscous rheology (21). Themantle is in a fluid dynamical regime characterizedby relatively high convective vigor and a close bal-ance between viscous and buoyancy forces, andforwhich viscous entrainment andmixing is highlyeffective. However, geochemical evidence directlycontradicts this, as it points to the existence ofisotopically distinct domains in Earth’s interiorthat should have been isolated since Earth’s for-

mation (80–82). Many clues suggest thatthese isolated reservoirs reside in Earth’slower mantle, and therefore the dynam-ics of the lower mantle, and by extensionthe transport properties of Mg-perovskite,must be reexamined in greater detail.A simple model to delay mixing of geo-

chemically distinct domains is through var-iations in theMg/Si ratio of rocks circulatingthrough the deep mantle, which give riseto viscosity variations that organize thelarge-scale pattern of convection. The vis-cosity variations relate to the relative pro-portions of Mg-perovskite to ferropericlaseas the Mg/Si ratio changes. The resultingdynamical regime results in isolated strongBEAMS separated by channels of higher-Mg/Si (relatively weak) rocks that readilydeform and circulate between the shallowand deepmantle (Fig. 3) (21). Such BEAMSmayhelp to solve the “missing siliconparadox”(Box 1), in addition to hosting other isotopi-cally distinctmaterial since the time of earlyEarth. The potential for strain weakeningthrough the interconnection of the weakerpericlase (27) may amplify this picture byshear-localizing rheology that would helpisolate compositionally distinct domains.Perovskite should also be an important

constituent in other terrestrial planets iftheir rocky mantles contain substantial vol-umes above ~23 GPa. Venus, being similarin size and mass to Earth, should also hostMg-perovskite as its most abundant min-eral phase. The base ofMars’mantle, on the

other hand, is estimated to be 19 to 27GPa (83), andit is unknown whether any Mg-perovskite is sta-ble inside the red planet (84). Larger terrestrial-like planets that exist around other stars may findlarge portions of their mantles above ~100 GPa,well beyond the stability field of Mg-perovskite.Nevertheless, theMg-perovskite–bearing portionof every terrestrial planet could play a limiting rolein the dynamical evolution of the planet, partic-ularly if it is highly enriched in Mg-perovskite, be-cause the high viscosity of this phase moderatesthemantle convection conveyor belt that circulatesmaterial between the surface and deep interior.If a planet hosts too much Mg-perovskite, and

its rheology is thereby too strong to permit ef-ficient cooling by mantle convection, heat willbuild up in the interior, owing to ambient radio-activity, and raise the temperature, thus soften-ing the rock until it can permit convective flowsand release the excess heat (85). Thus, terrestrial

Hirose et al., Science 358, 734–738 (2017) 10 November 2017 4 of 5

Characteristics of MgSiO3 perovskite.

PHOTO

CREDIT:T

AKAYUKIISHII

on Novem

ber 9, 2017

http://science.sciencemag.org/

Dow

nloaded from

mantles containing a higher proportion of Mg-perovskite are expected to be hotter than thosebearing enough periclase to allow for substantialweakening of the rock. If the elevated temper-atures are high enough to cause partial meltingof an assemblage of Mg-perovskite with a minoramount of periclase, the eutectic melt will havea relatively higher Mg/Si ratio that is close topyrolite (86). Segregation of this pyrolitic meltfrom the residual solid, followed by freezing ofthe melt at shallower depths, could lead to large-scale segregation of a mantle into a low-Mg/Siregion above 23 GPa and a high-Mg/Si regionat shallower depths. The latter rock will bear ahigher abundance of periclase if it is returnedto depth (>23 GPa) and would be substantiallyweaker than the rock that produced it, possiblypenetrating to form weak channels or conduitsto accommodate flow through the surroundingMg-perovskite–rich rocks. This leads to a sce-nario very much like BEAMS, described earlier.However, such an escape clause via partial melt-ing and segregation is only available if Mg/Si >1 initially, because a planetary mantle with Mg/Si < 1 is expected to host Mg-perovskite alongwith a SiO2 phase that is also expected to be high-ly viscous (87). In the latter case, eutectic meltingand segregation do not lead to the productionof any composition rock that is remarkably softer;thus, elevated temperatures would need to bemaintained indefinitely to soften the rock suf-ficiently to release the internal heat of the planet.Such internal dynamicsmay determinewhether aplanet can sustain something like plate tectonics,which accommodates the interaction of internalmatter with the surface environment, and sustaindisequilibrium that is thought to be an importantsource of energy for sustaining a biosphere.

REFERENCES AND NOTES

1. N. Tomioka, K. Fujino, Science 277, 1084–1086 (1997).2. O. Tschauner et al., Science 346, 1100–1102 (2014).3. L. Liu, Geophys. Res. Lett. 1, 277–280 (1974).4. E. Ito, Geophys. Res. Lett. 4, 72–74 (1977).5. D. Yamazaki, S. Karato, Am. Mineral. 86, 385–391 (2001).6. W. Xu, C. Lithgow-Bertelloni, L. Stixrude, J. Ritsema,

Earth Planet. Sci. Lett. 275, 70–79 (2008).7. A. Navrotsky, Geophys. Res. Lett. 7, 709–711 (1980).

8. Y. Fukao, M. Obayashi, T. Nakakuki, Annu. Rev. Earth Planet.Sci. 37, 19–46 (2009).

9. D. M. Trots et al., J. Geophys. Res. 118, 5805–5813 (2013).10. M. Murakami, K. Hirose, K. Kawamura, N. Sata, Y. Ohishi,

Science 304, 855–858 (2004).11. A. R. Oganov, S. Ono, Nature 430, 445–448 (2004).12. T. Tsuchiya, J. Tsuchiya, K. Umemoto, R. A. Wentzcovitch,

Earth Planet. Sci. Lett. 224, 241–248 (2004).13. R. M. Wentzcovitch, T. Tsuchiya, J. Tsuchiya, Proc. Natl. Acad.

Sci. U.S.A. 103, 543–546 (2006).14. K. Ohta, K. Hirose, T. Lay, N. Sata, Y. Ohishi, Earth Planet. Sci.

Lett. 267, 107–117 (2008).15. G. Fiquet et al., Science 329, 1516–1518 (2010).16. D. Andrault et al., Earth Planet. Sci. Lett. 293, 90–96 (2010).17. B. Grocholski, K. Catalli, S. H. Shim, V. Prakapenka,

Proc. Natl. Acad. Sci. U.S.A. 109, 2275–2279 (2012).18. M. L. Rudolph, V. Lekić, C. Lithgow-Bertelloni, Science 350,

1349–1352 (2015).19. H. Marquardt, L. Miyagi, Nat. Geosci. 8, 311–314 (2015).20. S.-H. Shim et al., Proc. Natl. Acad. Sci. U.S.A. 114, 6468–6473 (2017).21. M. D. Ballmer, C. Houser, J. W. Hernlund, R. M. Wentzcovitch,

K. Hirose, Nat. Geosci. 10, 236–240 (2017).22. K. Hirose, J. Geophys. Res. 107, ECV 3-1–ECV 3-13 (2002).23. S. Tateno, K. Hirose, N. Sata, Y. Ohishi, Earth Planet. Sci. Lett.

277, 130–136 (2009).24. T. Nakagawa, P. J. Tackley, Geophys. Res. Lett. 31, L16611

(2004).25. B. H. Hager, J. Geophys. Res. 89, 6003–6015 (1984).26. D. Yamazaki et al., Phys. Earth Planet. Inter. 228, 262–267 (2014).27. J. Girard, G. Amulele, R. Farla, A. Mohiuddin, S. Karato,

Science 351, 144–147 (2016).28. D. P. Dobson et al., Nat. Geosci. 6, 575–578 (2013).29. C. A. McCammon, T. Stachel, J. W. Harris, Earth Planet. Sci. Lett.

222, 423–434 (2004).30. E. M. Smith et al., Science 354, 1403–1405 (2016).31. J. Badro et al., Science 300, 789–791 (2003).32. Z. Wu, R. M. Wentzcovitch, Proc. Natl. Acad. Sci. U.S.A. 111,

10468–10472 (2014).33. C. McCammon et al., Nat. Geosci. 1, 684–687 (2008).34. J. F. Lin et al., Nat. Geosci. 1, 688–691 (2008).35. A. Bengtson, J. Li, D. Morgan, Geophys. Res. Lett. 36, L15301

(2009).36. R. Sinmyo, C. McCammon, L. Dubrovinsky, Am. Mineral. 102,

1263–1269 (2017).37. H. Piet et al., Proc. Natl. Acad. Sci. U.S.A. 113, 11127–11130 (2016).38. J. M. R. Muir, J. P. Brodholt, Phys. Earth Planet. Inter. 257, 12–17 (2016).39. F. V. Kaminsky, J. F. Lin, Am. Mineral. 102, 824–832 (2017).40. Y. Tange, Y. Kuwayama, T. Irifune, K. Funakoshi, Y. Ohishi,

J. Geophys. Res. 117, B06201 (2012).41. Y. Fei, L. Zhang, A. Corgne, H. Watson, A. Ricolleau, Y. Meng,

V. Prakapenka, Geophys. Res. Lett. 34, L17307 (2007).42. S. Karato, K. Fujino, E. Ito, Geophys. Res. Lett. 17, 13–16 (1990).43. T. B. Ballaran et al., Earth Planet. Sci. Lett. 333-334, 181–190 (2012).44. Z. Mao et al., Am. Mineral. 102, 357–368 (2017).45. T. Katsura, A. Yoneda, D. Yamazaki, T. Yoshino, E. Ito,

Phys. Earth Planet. Inter. 183, 212–218 (2010).46. J. Tsuchiya, T. Tsuchiya, R. M. Wentzcovitch, J. Geophys. Res.

110, B02204 (2005).47. G. M. Manthilake, N. de Koker, D. J. Frost, C. A. McCammon,

Proc. Natl. Acad. Sci. U.S.A. 108, 17901–17904 (2011).

48. K. Ohta et al., Earth Planet. Sci. Lett. 349-350, 109–115(2012).

49. H. Keppler, L. S. Dubrovinsky, O. Narygina, I. Kantor, Science322, 1529–1532 (2008).

50. A. F. Goncharov et al., Phys. Earth Planet. Inter. 247, 11–16(2015).

51. J. W. Hernlund, C. Thomas, P. J. Tackley, Nature 434, 882–886(2005).

52. Y. Ye, V. Prakapenka, Y. Meng, S. H. Shim, J. Geophys. Res.122, 3450–3464 (2017).

53. M. Murakami, Y. Ohishi, N. Hirao, K. Hirose, Nature 485, 90–94 (2012).54. A. Kurnosov, H. Marquardt, D. J. Frost, T. B. Ballaran, L. Ziberna,

Nature 543, 543–546 (2017).55. R. M. Wentzcovitch, B. B. Karki, S. Karato, C. R. S. Da Silva,

Earth Planet. Sci. Lett. 164, 371–378 (1998).56. X. L. Wang, T. Tsuchiya, A. Hase, Nat. Geosci. 8, 556–559 (2015).57. S. Cottaar, T. Heister, I. Rose, C. Unterborn, Geochem. Geophys.

Geosyst. 15, 1164–1179 (2014).58. A. Kuvshinov, Surv. Geophys. 33, 169–209 (2012).59. T. Yoshino, S. Kamada, C. C. Zhao, E. Ohtani, N. Hirao,

Earth Planet. Sci. Lett. 434, 208–219 (2016).60. K. Ohta et al., Earth Planet. Sci. Lett. 289, 497–502 (2010).61. R. Sinmyo, G. Pesce, E. Greenberg, C. McCammon,

L. Dubrovinsky, Earth Planet. Sci. Lett. 393, 165–172 (2014).62. K. Ohta et al., Science 320, 89–91 (2008).63. R. Holme, Geophys. J. Int. 132, 167–180 (1998).64. S. Kakizawa, T. Inoue, H. Yurimoto, Goldschmidt 2015

Conference Abstracts, no. 1494 (Geochemical Society and theEuropean Association of Geochemistry, 2015).

65. E. R. Hernández, D. Alfe, J. Brodholt, Earth Planet. Sci. Lett.364, 37–43 (2013).

66. J. E. Dixon, D. A. Clague, J. Petrol. 42, 627–654 (2001).67. N. Bolfan-Casanova, H. Keppler, D. C. Rubie, Earth Planet. Sci.

Lett. 182, 209–221 (2000).68. J. P. Townsend, J. Tsuchiya, C. R. Bina, S. D. Jacobsen,

Earth Planet. Sci. Lett. 454, 20–27 (2016).69. K. Yuan, B. Romanowicz, Science 357, 393–397 (2017).70. S. S. Shcheka, H. Keppler, Nature 490, 531–534 (2012).71. T. Komabayashi, K. Hirose, N. Sata, Y. Ohishi, L. S. Dubrovinsky,

Earth Planet. Sci. Lett. 260, 564–569 (2007).72. L. Stixrude, C. Lithgow-Bertelloni, B. Kiefer, P. Fumagalli,

Phys. Rev. B 75, 024108 (2007).73. T. Tsuchiya, J. Tsuchiya, Proc. Natl. Acad. Sci. U.S.A. 108,

1252–1255 (2011).74. K. Hirose, N. Shimizu, W. van Westrenen, Y. Fei, Phys. Earth

Planet. Inter. 146, 249–260 (2004).75. A. Corgne, C. Liebske, B. J. Wood, D. C. Rubie, D. J. Frost,

Geochim. Cosmochim. Acta 69, 485–496 (2005).76. D. Andrault et al., Science 344, 892–895 (2014).77. S. N. Perry, J. S. Pigott, W. R. Panero, Am. Mineral. 102, 321–326

(2017).78. N. Tsujino et al., Nature 539, 81–84 (2016).79. R. Nomura et al., Rev. Sci. Instrum. 88, 044501 (2017).80. S. Mukhopadhyay, Nature 486, 101–104 (2012).81. L. J. Hallis et al., Science 350, 795–797 (2015).82. A. Mundl et al., Science 356, 66–69 (2017).83. C. M. Bertka, Y. Fei, Science 281, 1838–1840 (1998).84. F. Sohl, G. Schubert, in Treatise on Geophysics, G. Schubert,

Ed. (Elsevier, ed. 2, 2015), pp. 23–64.85. D. C. Tozer, in The Earth’s Mantle, T. F. Gaskell, Ed. (Academic

Press, 1967), pp. 325–353.86. C. Liebske, D. J. Frost, Earth Planet. Sci. Lett. 345-348, 159–170 (2012).87. F. Xu et al., Earth Planet. Sci. Lett. 459, 332–339 (2017).88. T. Ishii, H. Kojitani,M. Akaogi,EarthPlanet. Sci. Lett.309, 185–197 (2011).89. C. J. Allègre, J. P. Poirier, E. Humler, A. W. Hofmann,

Earth Planet. Sci. Lett. 134, 515–526 (1995).90. A. Shahar et al., Geochim. Cosmochim. Acta 75, 7688–7697 (2011).91. N. Dauphas, F. Poitrasson, C. Burkhardt, H. Kobayashi,

K. Kurosawa, Earth Planet. Sci. Lett. 427, 236–248 (2015).92. S. Tateno, K. Hirose, Y. Ohishi, J. Geophys. Res. 119, 4684–4694

(2014).

ACKNOWLEDGMENTS

We thank the reviewers for their comments, which helpedimprove the manuscript. This research was supported by JapanSociety for the Promotion of Science research grants (16H06285to K.H. and 16H06023 and JP16H01115 to R.S.).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/358/6364/734/suppl/DC1Fig. S1References (93–97)

10.1126/science.aam8561

Hirose et al., Science 358, 734–738 (2017) 10 November 2017 5 of 5

Box 1. The chemical composition of Earth.

ThebulkEarth is considered tobe similar in composition to chondrites (primitivemeteorites),with theexception of volatile elements (89), because the chondrites coming from the asteroid belt havecompositions that are almost the same as the solar composition. However, the typical composition ofEarth’s uppermost mantle (pyrolite) is different from that of the silicate part of the chondrites; Mg/Simolar ratio is ~1.3 for the formerand 1.0 for the latter. Indeed, aprimarymineral isMg2SiO4-richolivine inthe upper mantle, whereas a primary mineral is MgSiO3-rich enstatite in chondrites.This is the missingsiliconparadox. In order to reconcile thebulkEarthMg/Si ratiowith the chondritic and solar values, it hasbeen argued that 6 to 12 wt% Si is present in the core (89, 90), as supported by the 30Si/28Si isotopiccomposition (d30Si) of Earth’smantle,which is higher than those of chondrites (90).The recent work byDauphas et al. (91) estimated a more modest amount of Si (3.6 wt %) in Earth’s core, considering thebulk planetary-scale variations in Mg/Si and d30Si.They argued that such variations were caused by thefractionation of forsterite (Mg/Si = 2) in the solar nebula,which can vary theMg/Si ratio, and thusmantleviscosity, in planetary bodies. Alternatively, it is also possible that Earth’s lower mantle has achondritic Mg/Si ratio different from that observed in the upper mantle.This means that the lowermantle is composed almost entirely ofMg-perovskite (21, 53),which could happen if amagmaoceanextended to the deep mantle and crystallized it as a liquidous phase (92).

PEROVSKITES on N

ovember 9, 2017

http://science.sciencem

ag.org/D

ownloaded from

0 20 40 60 80 100 120 14016

18

20

22

24

26

Tilti

ng a

ngle

(deg

ree)

Pressure (GPa)

Fig. S1. Tilting of SiO6 octahedra in Mg-perovskite under pressure. Black, pure

MgSiO3 perovskite (40, 43, 93-96); blue, Fe2+-bearing Mg-perovskite (43); red, (Al, Fe2+,

Fe3+)-bearing Mg-perovskite (96, 97). Solid line and open symbols are based on unit-cell

parameters assuming rigid octahedron. Broken line and closed symbols are obtained from

the atomic positions of oxygen. Post-perovskite phase transformation may occur at tilting

angle larger than 24 degrees.

Supplementary materials

2

References and Notes

1. N. Tomioka, K. Fujino, Natural (Mg,Fe)SiO3-ilmenite and -perovskite in the Tenham

meteorite. Science 277, 1084–1086 (1997). doi:10.1126/science.277.5329.1084

Medline

2. O. Tschauner, C. Ma, J. R. Beckett, C. Prescher, V. B. Prakapenka, G. R. Rossman,

Discovery of bridgmanite, the most abundant mineral in Earth, in a shocked

meteorite. Science 346, 1100–1102 (2014). doi:10.1126/science.1259369 Medline

3. L. Liu, Silicate perovskite from phase transformations of pyrope-garnet at high pressure

and temperature. Geophys. Res. Lett. 1, 277–280 (1974).

doi:10.1029/GL001i006p00277

4. E. Ito, The absence of oxide mixture in high-pressure phases of Mg-silicates. Geophys.

Res. Lett. 4, 72–74 (1977). doi:10.1029/GL004i002p00072

5. D. Yamazaki, S. Karato, Some mineral physics constraints on the rheology and geothermal

structure of Earth’s lower mantle. Am. Mineral. 86, 385–391 (2001). doi:10.2138/am-

2001-0401

6. W. Xu, C. Lithgow-Bertelloni, L. Stixrude, J. Ritsema, The effect of bulk composition and

temperature on mantle seismic structure. Earth Planet. Sci. Lett. 275, 70–79 (2008).

doi:10.1016/j.epsl.2008.08.012

7. A. Navrotsky, Lower mantle phase transitions may generally have negative pressure-

temperature slopes. Geophys. Res. Lett. 7, 709–711 (1980).

doi:10.1029/GL007i009p00709

8. Y. Fukao, M. Obayashi, T. Nakakuki, Stagnant slab: A review. Annu. Rev. Earth Planet.

Sci. 37, 19–46 (2009). doi:10.1146/annurev.earth.36.031207.124224

9. D. M. Trots, A. Kurnosov, T. B. Ballaran, S. Tkachev, K. Zhuravlev, V. Prakapenka, M.

Berkowski, D. J. Frost, The Sm:YAG primary fluorescence pressure scale. J.

Geophys. Res. 118, 5805–5813 (2013). doi:10.1002/2013JB010519

10. M. Murakami, K. Hirose, K. Kawamura, N. Sata, Y. Ohishi, Post-perovskite phase

transition in MgSiO3. Science 304, 855–858 (2004). doi:10.1126/science.1095932

Medline

11. A. R. Oganov, S. Ono, Theoretical and experimental evidence for a post-perovskite phase

of MgSiO3 in Earth’s D″ layer. Nature 430, 445–448 (2004).

doi:10.1038/nature02701 Medline

12. T. Tsuchiya, J. Tsuchiya, K. Umemoto, R. A. Wentzcovitch, Phase transition in MgSiO3

perovskite in the earth’s lower mantle. Earth Planet. Sci. Lett. 224, 241–248 (2004).

doi:10.1016/j.epsl.2004.05.017

13. R. M. Wentzcovitch, T. Tsuchiya, J. Tsuchiya, MgSiO3 postperovskite at D″ conditions.

Proc. Natl. Acad. Sci. U.S.A. 103, 543–546 (2006). doi:10.1073/pnas.0506879103

Medline

14. K. Ohta, K. Hirose, T. Lay, N. Sata, Y. Ohishi, Phase transitions in pyrolite and MORB at

lowermost mantle conditions: Implications for a MORB-rich pile above the core–

mantle boundary. Earth Planet. Sci. Lett. 267, 107–117 (2008).

doi:10.1016/j.epsl.2007.11.037

3

15. G. Fiquet, A. L. Auzende, J. Siebert, A. Corgne, H. Bureau, H. Ozawa, G. Garbarino,

Melting of peridotite to 140 gigapascals. Science 329, 1516–1518 (2010).

doi:10.1126/science.1192448 Medline

16. D. Andrault, M. Muñoz, N. Bolfan-Casanova, N. Guignot, J.-P. Perrillat, G. Aquilanti, S.

Pascarelli, Experimental evidence for perovskite and post-perovskite coexistence

throughout the whole D″ region. Earth Planet. Sci. Lett. 293, 90–96 (2010).

doi:10.1016/j.epsl.2010.02.026

17. B. Grocholski, K. Catalli, S. H. Shim, V. Prakapenka, Mineralogical effects on the

detectability of the postperovskite boundary. Proc. Natl. Acad. Sci. U.S.A. 109, 2275–

2279 (2012). doi:10.1073/pnas.1109204109 Medline

18. M. L. Rudolph, V. Lekić, C. Lithgow-Bertelloni, Viscosity jump in Earth’s mid-mantle.

Science 350, 1349–1352 (2015). doi:10.1126/science.aad1929 Medline

19. H. Marquardt, L. Miyagi, Slab stagnation in the shallow lower mantle linked to an

increase in mantle viscosity. Nat. Geosci. 8, 311–314 (2015). doi:10.1038/ngeo2393

20. S.-H. Shim, B. Grocholski, Y. Ye, E. E. Alp, S. Xu, D. Morgan, Y. Meng, V. B.

Prakapenka, Stability of ferrous-iron-rich bridgmanite under reducing midmantle

conditions. Proc. Natl. Acad. Sci. U.S.A. 114, 6468–6473 (2017).

doi:10.1073/pnas.1614036114 Medline

21. M. D. Ballmer, C. Houser, J. W. Hernlund, R. M. Wentzcovitch, K. Hirose, Persistence of

strong silica-enriched domains in the Earth’s lower mantle. Nat. Geosci. 10, 236–240

(2017). doi:10.1038/ngeo2898

22. K. Hirose, Phase transitions in pyrolitic mantle around 670-km depth: Implications for

upwelling of plumes from the lower mantle. J. Geophys. Res. 107, ECV 3-1–ECV 3-

13 (2002). doi:10.1029/2001JB000597

23. S. Tateno, K. Hirose, N. Sata, Y. Ohishi, Determination of post-perovskite phase

transition boundary up to 4400 K and implications for thermal structure in D″ layer.

Earth Planet. Sci. Lett. 277, 130–136 (2009). doi:10.1016/j.epsl.2008.10.004

24. T. Nakagawa, P. J. Tackley, Effects of a perovskite-post perovskite phase change near

core-mantle boundary in compressible mantle convection. Geophys. Res. Lett. 31,

L16611 (2004). doi:10.1029/2004GL020648

25. B. H. Hager, Subducted slabs and the geoid: Constraints on mantle rheology and flow. J.

Geophys. Res. 89, 6003–6015 (1984). doi:10.1029/JB089iB07p06003

26. D. Yamazaki, E. Ito, T. Yoshino, N. Tsujino, A. Yoneda, X. Guo, F. Xu, Y. Higo, K.

Funakoshi, Over 1 Mbar generation in the Kawai-type multianvil apparatus and its

application to compression of (Mg0.92Fe0.08)SiO3 perovskite and stishovite. Phys.

Earth Planet. Inter. 228, 262–267 (2014). doi:10.1016/j.pepi.2014.01.013

27. J. Girard, G. Amulele, R. Farla, A. Mohiuddin, S. Karato, Shear deformation of

bridgmanite and magnesiowüstite aggregates at lower mantle conditions. Science 351,

144–147 (2016). doi:10.1126/science.aad3113 Medline

28. D. P. Dobson, N. Miyajima, F. Nestola, M. Alvaro, N. Casati, C. Liebske, I. G. Wood, A.

M. Walker, Strong inheritance of texture between perovskite and post-perovskite in

the D′′ layer. Nat. Geosci. 6, 575–578 (2013). doi:10.1038/ngeo1844

4

29. C. A. McCammon, T. Stachel, J. W. Harris, Iron oxidation state in lower mantle mineral

assemblages. Earth Planet. Sci. Lett. 222, 423–434 (2004).

doi:10.1016/j.epsl.2004.03.019

30. E. M. Smith, S. B. Shirey, F. Nestola, E. S. Bullock, J. Wang, S. H. Richardson, W.

Wang, Large gem diamonds from metallic liquid in Earth’s deep mantle. Science 354,

1403–1405 (2016). doi:10.1126/science.aal1303 Medline

31. J. Badro, G. Fiquet, F. Guyot, J. P. Rueff, V. V. Struzhkin, G. Vankó, G. Monaco, Iron

partitioning in Earth’s mantle: Toward a deep lower mantle discontinuity. Science

300, 789–791 (2003). doi:10.1126/science.1081311 Medline

32. Z. Wu, R. M. Wentzcovitch, Spin crossover in ferropericlase and velocity heterogeneities

in the lower mantle. Proc. Natl. Acad. Sci. U.S.A. 111, 10468–10472 (2014).

doi:10.1073/pnas.1322427111 Medline

33. C. McCammon, I. Kantor, O. Narygina, J. Rouquette, U. Ponkratz, I. Sergueev, M.

Mezouar, V. Prakapenka, L. Dubrovinsky, Stable intermediate-spin ferrous iron in

lower-mantle perovskite. Nat. Geosci. 1, 684–687 (2008). doi:10.1038/ngeo309

34. J. F. Lin, H. Watson, G. Vankó, E. E. Alp, V. B. Prakapenka, P. Dera, V. V. Struzhkin, A.

Kubo, J. Zhao, C. McCammon, W. J. Evans, Intermediate-spin ferrous iron in

lowermost mantle post-perovskite and perovskite. Nat. Geosci. 1, 688–691 (2008).

doi:10.1038/ngeo310

35. A. Bengtson, J. Li, D. Morgan, Mössbauer modeling to interpret the spin state of iron in

(Mg,Fe)SiO3 perovskite. Geophys. Res. Lett. 36, L15301 (2009).

doi:10.1029/2009GL038340

36. R. Sinmyo, C. McCammon, L. Dubrovinsky, The spin state of Fe3+ in lower mantle

bridgmanite. Am. Mineral. 102, 1263–1269 (2017). doi:10.2138/am-2017-5917

37. H. Piet, J. Badro, F. Nabiei, T. Dennenwaldt, S.-H. Shim, M. Cantoni, C. Hébert, P.

Gillet, Spin and valence dependence of iron partitioning in Earth’s deep mantle. Proc.

Natl. Acad. Sci. U.S.A. 113, 11127–11130 (2016). doi:10.1073/pnas.1605290113

Medline

38. J. M. R. Muir, J. P. Brodholt, Ferrous iron partitioning in the lower mantle. Phys. Earth

Planet. Inter. 257, 12–17 (2016). doi:10.1016/j.pepi.2016.05.008

39. F. V. Kaminsky, J. F. Lin, Iron partitioning in natural lower-mantle minerals: Toward a

chemically heterogeneous lower mantle. Am. Mineral. 102, 824–832 (2017).

doi:10.2138/am-2017-5949

40. Y. Tange, Y. Kuwayama, T. Irifune, K. Funakoshi, Y. Ohishi, P-V-T equation of state of

MgSiO3 perovskite based on the MgO pressure scale: A comprehensive reference for

mineralogy of the lower mantle. J. Geophys. Res. 117, B06201 (2012).

doi:10.1029/2011JB008988

41. Y. Fei, L. Zhang, A. Corgne, H. Watson, A. Ricolleau, Y. Meng, V. Prakapenka, Spin

transition and equations of state of (Mg, Fe)O solid solutions. Geophys. Res. Lett. 34,

L17307 (2007). doi:10.1029/2007GL030712

42. S. Karato, K. Fujino, E. Ito, Plasticity of MGSIO3 perovskite: The results of

microhardness tests on single crystals. Geophys. Res. Lett. 17, 13–16 (1990).

doi:10.1029/GL017i001p00013

5

43. T. B. Ballaran, A. Kurnosov, K. Glazyrin, D. J. Frost, M. Merlini, M. Hanfland, R.

Caracas, Effect of chemistry on the compressibility of silicate perovskite in the lower

mantle. Earth Planet. Sci. Lett. 333-334, 181–190 (2012).

doi:10.1016/j.epsl.2012.03.029

44. Z. Mao, F. Wang, J.-F. Lin, S. Fu, J. Yang, X. Wu, T. Okuchi, N. Tomioka, V. B.

Prakapenka, Y. Xiao, P. Chow, Equation of state and hyperfine parameters of high-

spin bridgmanite in the Earth’s lower mantle by synchrotron X-ray diffraction and

Mössbauer spectroscopy. Am. Mineral. 102, 357–368 (2017). doi:10.2138/am-2017-

5770

45. T. Katsura, A. Yoneda, D. Yamazaki, T. Yoshino, E. Ito, Adiabatic temperature profile in

the mantle. Phys. Earth Planet. Inter. 183, 212–218 (2010).

doi:10.1016/j.pepi.2010.07.001

46. J. Tsuchiya, T. Tsuchiya, R. M. Wentzcovitch, Vibrational and thermodynamic properties

of MgSiO3 postperovskite. J. Geophys. Res. 110, B02204 (2005).

doi:10.1029/2004JB003409

47. G. M. Manthilake, N. de Koker, D. J. Frost, C. A. McCammon, Lattice thermal

conductivity of lower mantle minerals and heat flux from Earth’s core. Proc. Natl.

Acad. Sci. U.S.A. 108, 17901–17904 (2011). doi:10.1073/pnas.1110594108 Medline

48. K. Ohta, T. Yagi, N. Taketoshi, K. Hirose, T. Komabayashi, T. Baba, Y. Ohishi, J.

Hernlund, Lattice thermal conductivity of MgSiO3 perovskite and post-perovskite at

the core–mantle boundary. Earth Planet. Sci. Lett. 349-350, 109–115 (2012).

doi:10.1016/j.epsl.2012.06.043

49. H. Keppler, L. S. Dubrovinsky, O. Narygina, I. Kantor, Optical absorption and radiative

thermal conductivity of silicate perovskite to 125 gigapascals. Science 322, 1529–

1532 (2008). doi:10.1126/science.1164609 Medline

50. A. F. Goncharov, S. S. Lobanov, X. Tan, G. T. Hohensee, D. G. Cahill, J.-F. Lin, S.-M.

Thomas, T. Okuchi, N. Tomioka, Experimental study of thermal conductivity at high

pressures: Implications for the deep Earth’s interior. Phys. Earth Planet. Inter. 247,

11–16 (2015). doi:10.1016/j.pepi.2015.02.004

51. J. W. Hernlund, C. Thomas, P. J. Tackley, A doubling of the post-perovskite phase

boundary and structure of the Earth’s lowermost mantle. Nature 434, 882–886 (2005).

doi:10.1038/nature03472 Medline

52. Y. Ye, V. Prakapenka, Y. Meng, S. H. Shim, Intercomparison of the gold, platinum, and

MgO pressure scales up to 140 GPa and 2500 K. J. Geophys. Res. 122, 3450–3464

(2017). doi:10.1002/2016JB013811

53. M. Murakami, Y. Ohishi, N. Hirao, K. Hirose, A perovskitic lower mantle inferred from

high-pressure, high-temperature sound velocity data. Nature 485, 90–94 (2012).

doi:10.1038/nature11004 Medline

54. A. Kurnosov, H. Marquardt, D. J. Frost, T. B. Ballaran, L. Ziberna, Evidence for a Fe3+-

rich pyrolitic lower mantle from (Al,Fe)-bearing bridgmanite elasticity data. Nature

543, 543–546 (2017). doi:10.1038/nature21390 Medline

55. R. M. Wentzcovitch, B. B. Karki, S. Karato, C. R. S. Da Silva, High pressure elastic

anisotropy of MgSiO3 perovskite and geophysical implications. Earth Planet. Sci.

Lett. 164, 371–378 (1998). doi:10.1016/S0012-821X(98)00230-1

6

56. X. L. Wang, T. Tsuchiya, A. Hase, Computational support for a pyrolitic lower mantle

containing ferric iron. Nat. Geosci. 8, 556–559 (2015). doi:10.1038/ngeo2458

57. S. Cottaar, T. Heister, I. Rose, C. Unterborn, BurnMan: A lower mantle mineral physics

toolkit. Geochem. Geophys. Geosyst. 15, 1164–1179 (2014).

doi:10.1002/2013GC005122

58. A. Kuvshinov, Deep electromagnetic studies from land, sea, and space: Progress status in

the past 10 years. Surv. Geophys. 33, 169–209 (2012). doi:10.1007/s10712-011-9118-

2

59. T. Yoshino, S. Kamada, C. C. Zhao, E. Ohtani, N. Hirao, Electrical conductivity model of

Al-bearing bridgmanite with implications for the electrical structure of the Earth’s

lower mantle. Earth Planet. Sci. Lett. 434, 208–219 (2016).

doi:10.1016/j.epsl.2015.11.032

60. K. Ohta, K. Hirose, M. Ichiki, K. Shimizu, N. Sata, Y. Ohishi, Electrical conductivities of

pyrolitic mantle and MORB materials up to the lowermost mantle conditions. Earth

Planet. Sci. Lett. 289, 497–502 (2010). doi:10.1016/j.epsl.2009.11.042

61. R. Sinmyo, G. Pesce, E. Greenberg, C. McCammon, L. Dubrovinsky, Lower mantle

electrical conductivity based on measurements of Al, Fe-bearing perovskite under

lower mantle conditions. Earth Planet. Sci. Lett. 393, 165–172 (2014).

doi:10.1016/j.epsl.2014.02.049

62. K. Ohta, S. Onoda, K. Hirose, R. Sinmyo, K. Shimizu, N. Sata, Y. Ohishi, A. Yasuhara,

The electrical conductivity of post-perovskite in Earth’s D″ layer. Science 320, 89–91

(2008). doi:10.1126/science.1155148 Medline

63. R. Holme, Electromagnetic core–mantle coupling–I. Explaining decadal changes in the

length of day. Geophys. J. Int. 132, 167–180 (1998). doi:10.1046/j.1365-

246x.1998.00424.x

64. S. Kakizawa, T. Inoue, H. Yurimoto, “Water solubility of Al-bearing bridgmanite at the

lower mantle condition,” Goldschmidt2015 Conference Abstracts, no. 1494

(Geochemical Society and the European Association of Geochemistry, 2015).

65. E. R. Hernández, D. Alfe, J. Brodholt, The incorporation of water into lower-mantle

perovskites: A first-principles study. Earth Planet. Sci. Lett. 364, 37–43 (2013).

doi:10.1016/j.epsl.2013.01.005

66. J. E. Dixon, D. A. Clague, Volatiles in basaltic glasses from Loihi seamount, Hawaii:

Evidence for a relatively dry plume component. J. Petrol. 42, 627–654 (2001).

doi:10.1093/petrology/42.3.627

67. N. Bolfan-Casanova, H. Keppler, D. C. Rubie, Water partitioning between nominally

anhydrous minerals in the MgO–SiO2–H2O system up to 24 GPa: Implications for the

distribution of water in the Earth’s mantle. Earth Planet. Sci. Lett. 182, 209–221

(2000). doi:10.1016/S0012-821X(00)00244-2

68. J. P. Townsend, J. Tsuchiya, C. R. Bina, S. D. Jacobsen, Water partitioning between

bridgmanite and postperovskite in the lowermost mantle. Earth Planet. Sci. Lett. 454,

20–27 (2016). doi:10.1016/j.espl.2016.08.009

69. K. Yuan, B. Romanowicz, Seismic evidence for partial melting at the root of major hot

spot plumes. Science 357, 393–397 (2017). doi:10.1126/science.aan0760 Medline

7

70. S. S. Shcheka, H. Keppler, The origin of the terrestrial noble-gas signature. Nature 490,

531–534 (2012). doi:10.1038/nature11506 Medline

71. T. Komabayashi, K. Hirose, N. Sata, Y. Ohishi, L. S. Dubrovinsky, Phase transition in

CaSiO3 perovskite. Earth Planet. Sci. Lett. 260, 564–569 (2007).

doi:10.1016/j.epsl.2007.06.015

72. L. Stixrude, C. Lithgow-Bertelloni, B. Kiefer, P. Fumagalli, Phase stability and shear

softening in CaSiO3 perovskite at high pressure. Phys. Rev. B 75, 024108 (2007).

doi:10.1103/PhysRevB.75.024108

73. T. Tsuchiya, J. Tsuchiya, Prediction of a hexagonal SiO2 phase affecting stabilities of

MgSiO3 and CaSiO3 at multimegabar pressures. Proc. Natl. Acad. Sci. U.S.A. 108,

1252–1255 (2011). doi:10.1073/pnas.1013594108 Medline

74. K. Hirose, N. Shimizu, W. van Westrenen, Y. Fei, Trace element partitioning in Earth’s

lower mantle and implications for geochemical consequences of partial melting at the

core–mantle boundary. Phys. Earth Planet. Inter. 146, 249–260 (2004).

doi:10.1016/j.pepi.2002.11.001

75. A. Corgne, C. Liebske, B. J. Wood, D. C. Rubie, D. J. Frost, Silicate perovskite-melt

partitioning of trace elements and geochemical signature of a deep perovskitic

reservoir. Geochim. Cosmochim. Acta 69, 485–496 (2005).

doi:10.1016/j.gca.2004.06.041

76. D. Andrault, G. Pesce, M. A. Bouhifd, N. Bolfan-Casanova, J.-M. Hénot, M. Mezouar,

Melting of subducted basalt at the core-mantle boundary. Science 344, 892–895

(2014). doi:10.1126/science.1250466 Medline

77. S. N. Perry, J. S. Pigott, W. R. Panero, Ab initio calculations of uranium and thorium

storage in CaSiO3-perovskite in the Earth’s lower mantle. Am. Mineral. 102, 321–326

(2017). doi:10.2138/am-2017-5816

78. N. Tsujino, Y. Nishihara, D. Yamazaki, Y. Seto, Y. Higo, E. Takahashi, Mantle dynamics

inferred from the crystallographic preferred orientation of bridgmanite. Nature 539,

81–84 (2016). doi:10.1038/nature19777 Medline

79. R. Nomura, S. Azuma, K. Uesugi, Y. Nakashima, T. Irifune, T. Shinmei, S. Kakizawa, Y.

Kojima, H. Kadobayashi, High-pressure rotational deformation apparatus to 135 GPa.

Rev. Sci. Instrum. 88, 044501 (2017). doi:10.1063/1.4979562 Medline

80. S. Mukhopadhyay, Early differentiation and volatile accretion recorded in deep-mantle

neon and xenon. Nature 486, 101–104 (2012). doi:10.1038/nature11141 Medline

81. L. J. Hallis, G. R. Huss, K. Nagashima, G. J. Taylor, S. A. Halldórsson, D. R. Hilton, M.

J. Mottl, K. J. Meech, Evidence for primordial water in Earth’s deep mantle. Science

350, 795–797 (2015). doi:10.1126/science.aac4834 Medline

82. A. Mundl, M. Touboul, M. G. Jackson, J. M. D. Day, M. D. Kurz, V. Lekic, R. T. Helz,

R. J. Walker, Tungsten-182 heterogeneity in modern ocean island basalts. Science

356, 66–69 (2017). doi:10.1126/science.aal4179 Medline

83. C. M. Bertka, Y. Fei, Implications of Mars Pathfinder data for the accretion history of the

terrestrial planets. Science 281, 1838–1840 (1998).

doi:10.1126/science.281.5384.1838 Medline

84. F. Sohl, G. Schubert, in Treatise on Geophysics, G. Schubert, Ed. (Elsevier, ed. 2, 2015),

pp. 23–64.

8

85. D. C. Tozer, in The Earth’s Mantle, T. F. Gaskell, Ed. (Academic Press, 1967), pp. 325–

353.

86. C. Liebske, D. J. Frost, Melting phase relations in the MgO–MgSiO3 system between 16

and 26GPa: Implications for melting in Earth’s deep interior. Earth Planet. Sci. Lett.

345-348, 159–170 (2012). doi:10.1016/j.epsl.2012.06.038

87. F. Xu, D. Yamazaki, N. Sakamoto, W. Sun, H. Fei, H. Yurimoto, Silicon and oxygen

self-diffusion in stishovite: Implications for stability of SiO2-rich seismic reflectors in

the mid-mantle. Earth Planet. Sci. Lett. 459, 332–339 (2017).

doi:10.1016/j.epsl.2016.11.044

88. T. Ishii, H. Kojitani, M. Akaogi, Post-spinel transitions in pyrolite and Mg2SiO4 and

akimotoite–perovskite transition in MgSiO3: Precise comparison by high-pressure

high-temperature experiments with multi-sample cell technique. Earth Planet. Sci.

Lett. 309, 185–197 (2011). doi:10.1016/j.epsl.2011.06.023

89. C. J. Allègre, J. P. Poirier, E. Humler, A. W. Hofmann, The chemical composition of the

Earth. Earth Planet. Sci. Lett. 134, 515–526 (1995). doi:10.1016/0012-

821X(95)00123-T

90. A. Shahar, V. J. Hillgren, E. D. Young, Y. Fei, C. A. Macris, L. Deng, High-temperature

Si isotope fractionation between iron metal and silicate. Geochim. Cosmochim. Acta

75, 7688–7697 (2011). doi:10.1016/j.gca.2011.09.038

91. N. Dauphas, F. Poitrasson, C. Burkhardt, H. Kobayashi, K. Kurosawa, Planetary and

meteoritic Mg/Si and δ30 Si variations inherited from solar nebula chemistry. Earth

Planet. Sci. Lett. 427, 236–248 (2015). doi:10.1016/j.epsl.2015.07.008

92. S. Tateno, K. Hirose, Y. Ohishi, Melting experiments on peridotite to lowermost mantle

conditions. J. Geophys. Res. 119, 4684–4694 (2014). doi:10.1002/2013JB010616

93. T. Komabayashi, K. Hirose, E. Sugimura, N. Sata, Y. Ohishi, L. S. Dubrovinsky,

Simultaneous volume measurements of post-perovskite and perovskite in MgSiO3 and

their thermal equations of state. Earth Planet. Sci. Lett. 265, 515–524 (2008).

doi:10.1016/j.epsl.2007.10.036

94. D. Nishio-Hamane, Y. Seto, K. Fujino, T. Nagai, Effect of FeAlO3 incorporation into

MgSiO3 on the bulk modulus of perovskite. Phys. Earth Planet. Inter. 166, 219–225

(2008). doi:10.1016/j.pepi.2008.01.002

95. S. Tateno, K. Hirose, N. Sata, Y. Ohishi, Structural distortion of CaSnO3 perovskite under

pressure and the quenchable post-perovskite phase as a low-pressure analogue to

MgSiO3. Phys. Earth Planet. Inter. 181, 54–59 (2010).

doi:10.1016/j.pepi.2010.03.003

96. K. Glazyrin, T. Boffa Ballaran, D. J. Frost, C. McCammon, A. Kantor, M. Merlini, M.

Hanfland, L. Dubrovinsky, Magnesium silicate perovskite and effect of iron oxidation

state on its bulk sound velocity at the conditions of the lower mantle. Earth Planet.

Sci. Lett. 393, 182–186 (2014). doi:10.1016/j.epsl.2014.01.056

97. L. Ismailova, E. Bykova, M. Bykov, V. Cerantola, C. McCammon, T. Boffa Ballaran, A.

Bobrov, R. Sinmyo, N. Dubrovinskaia, K. Glazyrin, H.-P. Liermann, I. Kupenko, M.

Hanfland, C. Prescher, V. Prakapenka, V. Svitlyk, L. Dubrovinsky, Stability of Fe,Al-

bearing bridgmanite in the lower mantle and synthesis of pure Fe-bridgmanite. Sci.

Adv. 2, e1600427 (2016). doi:10.1126/sciadv.1600427 Medline

9