HAL Id: insu-01124396
https://hal-insu.archives-ouvertes.fr/insu-01124396
Submitted on 6 Mar 2015
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Megacrystals track magma convection between reservoir and surface
Yves Moussallam, Clive Oppenheimer, Bruno Scaillet, Iris Buisman, Christine Kimball, Nelia Dunbar, Alain Burgisser, C. Ian Schipper, Joan Andújar,
Philip Kyle
To cite this version:
Yves Moussallam, Clive Oppenheimer, Bruno Scaillet, Iris Buisman, Christine Kimball, et al..
Megacrystals track magma convection between reservoir and surface. Earth and Planetary Science
Letters, Elsevier, 2015, 413, pp.1-12. �10.1016/j.epsl.2014.12.022�. �insu-01124396�
Contents lists available atScienceDirect
Earth and Planetary Science Letters
www.elsevier.com/locate/epsl
Megacrystals track magma convection between reservoir and surface
Yves Moussallama,b,∗,Clive Oppenheimera,Bruno Scailletb,Iris Buismanc,
Christine Kimballd, Nelia Dunbare, Alain Burgisserf,g, C. Ian Schipperb,h, Joan Andújarb, Philip Kyled
aDepartmentofGeography,UniversityofCambridge,DowningPlace,Cambridge,CB23EN,UK bISTO,7327Universitéd’Orléans-CNRS-BRGM,1AruedelaFérollerie,45071Orléanscedex2,France cDepartmentofEarthSciences,UniversityofCambridge,DowningStreet,Cambridge,CB23EQ,UK
dDepartmentofEarthandEnvironmentalScience,NewMexicoInstituteofMiningandTechnology,801LeroyPlace,Socorro,NM87801,USA eNewMexicoBureauofGeologyandMineralResources,NewMexicoInstituteofMiningandTechnology,801LeroyPlace,Socorro,NM87801,USA fCNRS,ISTerre,F-73376LeBourgetduLac,France
gUniversitédeSavoie,ISTerre,F-73376LeBourgetduLac,France
hSchoolofGeography,EnvironmentandEarthSciences,VictoriaUniversity,POBox600,Wellington,NewZealand
a r t i c l e i n f o a b s t ra c t
Articlehistory:
Received1April2014
Receivedinrevisedform9December2014 Accepted11December2014
Availableonline13January2015 Editor:T.Elliott
Keywords:
convection bi-directionalflow megacrystal anorthoclase crystalzoning meltinclusion
Active volcanoes are typically fed by magmatic reservoirs situated within the upper crust. The developmentofthermaland/orcompositionalgradientsinsuchmagmachambersmayleadtovigorous convection as inferredfrom theoretical models and evidencefor magma mixingrecorded involcanic rocks. Bi-directional flow is also inferred to prevail in the conduits of numerous persistently-active volcanoes basedonobserved gas and thermal emissions atthe surface,as well as experiments with analogue models. However, moredirect evidence for suchexchange flows hashitherto been lacking.
Here,weanalysetheremarkableoscillatoryzoningofanorthoclasefeldsparmegacrystalseruptedfrom the lava lake of Erebus volcano, Antarctica. A comprehensive approach, combining phase equilibria, solubilityexperimentsandmeltinclusionandtexturalanalysesshowsthatthechemicalprofilesarebest explainedasaresultofmultipleepisodesofmagmatransportbetweenadeeperreservoirandthelava lakeatthesurface.Individualcrystalshaverepeatedlytravelledup-and-downtheplumbingsystem,over distancesofuptoseveralkilometers,presumablyasaconsequenceofentrainmentinthebulkmagma flow. Ourfindingsthus corroboratethe modelof bi-directionalflow inmagmatic conduits. Theyalso implycontrastingflowregimesinreservoirandconduit,withvigorous convectionintheformer(regular convectivecyclesof∼150daysataspeedof∼0.5 mm s−1)andmorecomplexcyclesofexchangeflow andre-entrainmentinthelatter.Weestimatethattypical,1-cm-widecrystalsshouldbeatleast14years old,andcanrecordseveral(from1 to3)completecyclesbetweenthereservoirandthelavalakeviathe conduit.ThispersistentrecyclingofphonoliticmagmaislikelysustainedbyCO2fluxing,suggestingthat accumulationofmaficmagmainthelowercrustisvolumetricallymoresignificantthanthatofevolved magmawithintheedifice.
©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Many volcanoes, such as Etna and Stromboli in Italy, persis- tently emit prodigious quantities of gas and heat at the surface withoutsignificantaccompanying lavaflowsortephraproduction (Francis etal., 1993). The decoupling of gas and thermalenergy fromthemagmaefflux hasbeen investigatedinseveraltheoreti- cal andexperimental treatments that considerexchange flow be-
* Correspondingauthorat:ISTO,7327Universitéd’Orléans-CNRS-BRGM,1Arue delaFérollerie,45071Orléanscedex2,France.
E-mailaddress:yves.moussallam@cnrs-orleans.fr(Y. Moussallam).
tween two fluids of contrasting density and viscosity (Kazahaya et al., 1994; Stevenson and Blake,1998; Huppert andHallworth, 2007; Beckett et al., 2011). These studies suggest that a sta- ble bi-directional flow can develop in volcanic conduits, and re- inforce interpretations of the magma dynamics of a number of persistently degassing volcanoes (e.g., Oppenheimer et al., 2009;
Shinohara and Tanaka, 2012). Conduit convection suggests sub- stantial endogenous growth of volcanoes (Francis et al., 1993;
Allard, 1997) and may explain melt inclusion trends associated with degassing volcanoes (Witham, 2011). While the conclusion that magma convectsinconduits feeding manyopen vent volca- noesseemsinescapable,directevidencehasbeenlacking.
http://dx.doi.org/10.1016/j.epsl.2014.12.022
0012-821X/©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).
2 Y. Moussallam et al. / Earth and Planetary Science Letters 413 (2015) 1–12
Thepotentialofzonedcrystalstorecordchemicalandphysical changes experienced by their host magma has long been recog- nised (see review by Ginibre et al., 2007). Oscillatory zoning in feldspar is a commonphenomenon withseveral inferredorigins.
Twomainschoolsofthoughthaveprevailed,originatingfromstud- iesinthe1920s,andbroadlyevokeeitherextrinsic(Bowen,1928) or intrinsic (Harloff, 1927) mechanisms. The extrinsic school ar- gues that zonation reflects changing pressure, temperature, com- positionand volatilecontent of themelt surrounding the crystal via processes such as convection (e.g., Singer et al., 1995) and episodicfluctuationsinmagmasupply.Theintrinsicschoolpoints tothewidely-observedlackofcorrelationofoscillatorylayersbe- tween crystals (e.g., Wiebe, 1968; Shore and Fowler, 1996) and argues insteadthat kinetic processesatthe crystal-meltinterface are responsible for zonation (e.g., Allegre etal., 1981; L’Heureux andFowler, 1996). The kinetic modelsimply highdegreesofun- dercooling atthe boundary layer andreproduce thetypical saw- toothpatternsrevealedinelectronmicroprobe(EMP)traversesof oscillatory-zonedfeldsparcrystals.However,theseintrinsicmodels cannot reproduce low amplitude variations. Nor do they explain multiple resorption episodes, evident in many oscillatory-zoned crystals.Recentstudieshavehighlightedhowbothmechanismsfor generatingoscillatory zoning canbe manifested insingle crystals (e.g. Viccaro et al., 2010), and have used mineral compositional zoningtoidentifyavarietyofmagmaticprocessessuchasmagma recharge, mixing and degassing (e.g., Humphreys et al., 2006;
Charlieretal.,2008; Kahletal.,2013).
Erebus volcano (Antarctica) is renowned for its long-lived phonoliticlavalake,whosebehaviourhasbeenexplainedinterms of the conduit exchange-flow model (Oppenheimer et al., 2009).
It is also remarkable for the impressive size to which its most abundantmineralphase,anorthoclasefeldspar,grows(Kyle,1977;
Dunbar et al., 1994). These anorthoclase megacrystals (up to 10-cm-long)displayexceptionaloscillatoryzoningandlarge(upto 600-μm-diameter)meltinclusions(Fig. 1). Inthisstudy,weanal- ysethemajorelementcompositionofnaturalanorthoclasecrystals andmajorelementandvolatilecompositionsofenclosedmeltin- clusions,andcomparethemtoanorthoclasecrystalsandassociated meltsderivedbyphaseequilibriumandsolubilityexperiments.The experimental data provide a tightly-constrained framework with which to interpret the naturalzoning andto retrace the growth historyofindividualanorthoclasemegacrystals.
We startwith abrief overviewofthe Erebus phonolitephase assemblage, including a description of the natural anorthoclase megacrystals, and then describe the different experimental and analytical methods used in thisstudy. We present the resultsof phaseequilibriumexperiments,andcomparethechemistryofsyn- theticandnaturalanorthoclase.Wethendescribesolubilityexper- imentsandtheir relationtoanalyses ofmeltinclusionsandtheir hostzonewithinnaturalcrystals.Lastly,welinkthechemicalzon- inginnaturalanorthoclasetothetimescalesofmagmaascentand descent.
2. Backgroundinformation 2.1. Mineralassemblage
Erebus volcano(3794 m,77.58◦S, 161.17◦E), hoststhe world’s onlyphonoliticlavalake.Thislavalakeappearstohavebeenper- sistently degassingsincethe volcanowas first observedby James Ross in 1841. This current passive activity is sporadically inter- ruptedbyStrombolianexplosionsthatejectfreshbombsontothe crater rim. All bombs analysed since 1972 are virtually identical in terms ofmineral assemblage (with one exception) andwhole rockandmatrixglassmajor,minorandtraceelements(Kellyetal., 2008).Thischemicalstabilityextendstoolderlavaflows,suchthat
alllavaseruptedfromErebusinthelast20 kahavethesamecom- position(Kellyetal.,2008) makingthevolcanoanidealsystemto investigateusingexperimentalpetrologytoolsatequilibriumcon- ditions.
Phonolite bombs are composed chiefly of vesicular matrix glass (∼67 vol%; fragile, easily disintegrated and microlite-free) and anorthoclase feldspar (∼30 vol%) with minor amounts of titanomagnetite (∼1.1 vol%), olivine (∼0.8 vol%), clinopyroxene (∼0.6 vol%) andfluorapatite (∼0.5 vol%),andlesserquantities of pyrrhotiteblebs(Kyleetal.,1992; Kellyetal.,2008).Ofallmineral phases(describedindetailedbyKellyetal.,2008)anorthoclaseis theonlymineralwithcompositionalzoning.
2.2. Anorthoclasemegacrystals
Euhedral anorthoclasefeldspar in phonolite bombs are zoned withrespecttomajor(Fig. 1)andtrace(Sumner, 2007) elements.
Thecompositionalzoning occursatavarietyofscales:thelowest frequency variations shown by some crystals (Fig. 2) have wave- lengthsof∼5mmandamplitude of∼7 mol%(Or) whilehigher- frequency variations havea typical wavelengthof 800 μmand amplitude of 4 mol%Or. Inter-zone variations are in the range of Ab61–66, An21–10, and Or14–28 (with Ab, An and Or referring to thealbite,anorthiteandorthoclase endmembersrespectively;
Fig. 2). Small-scale elemental maps (Fig. 1) reveal embayment at some zone boundaries, indicating resorption episodes. Hosted in the anorthoclase crystals are large (usually <600 μm across) phonolitic meltinclusions, typically trappedinsingle growthlay- ersoftheanorthoclase(Fig. 1). Naturalanorthoclasemegacrystals presented inthisstudywere manually separatedfrom phonolitic bombseruptedin1984and2005(Fig.S1;TableS1).
3. Methods
3.1. Phaseequilibriaandsolubilityexperiments
The startingmaterial(ERE 97018)usedinall experimentsisa phonolitic bomberupted in 1997andcollected at thecrater rim (describedinMoussallam etal., 2013). Allphaseequilibriaexper- imentsare describedinMoussallam etal. (2013) andthepresent studyusesasubsetofthoseexperimentswhoseconditionsclosely reproducedthenaturalmineralassemblage.
Water- andCO2-solubilityexperimentswereconductedacrossa rangeofpressures(50,100,200,and300 MPa),temperatures(950 and1000◦C), andXH2O(molefractionofwaterinthefluidphase, from near0 to 1) under reduced conditions(fO2≈QFM or be- low; withQFMbeingthequartz–fayalite–magnetiteredoxbuffer).
We usedinternally-heatedpressurevessels attheISTOlaboratory in Orléans, which can reach pressures of up to 400 MPa under controlled temperatures(upto1200◦C)andoxygenfugacitycon- ditions. The vessel was pressurised using an argon–hydrogen gas mixtureasthepressuremediumtocontrolredoxstate(Scailletet al., 1992). Heating was applied by a double-wound molybdenum furnace creatingastable“hot-spot”zone.TwoS-typethermocou- ples locatedoneithersideofthis5-cm-long“hot-spot”permitted precise control oftheheating resistances thus preventingthees- tablishmentofthermalgradients.
Experimental charges consisted of natural anhydrous sample powder (30 mg) withXH2O,loaded [= mole fractionof H2Oadded to the capsule, H2O/(H2O+CO2)] varying from 0 to 1 (i.e. pure CO2 to pure H2O) and loaded in gold capsules (2 cm in length, with 2.5 mm inner diameter and 2.9 mm outer diameter). The capsules were wrapped in liquid nitrogen-soaked tissue to pre- ventwaterlossandweldedshut.Foreachexperiment,sixcapsules were placedinasampleholderhungbya thinPtwire.Thetem- perature gradient along the “hot-spot” zone where the capsules were located wasalways <2◦C.Rapidquenchingwas assured by
Fig. 1.Upperpanel:PotassiumX-raymapsofasectionedanorthoclasecrystal(sampleeb05027a2)showingremarkableoscillatoryzoning.Irregularly-shapedmeltinclusions areclearlyvisibleasbrighter(K-rich)areas.Centralpanel:EMPtransectreportedintermsoforthoclaseendmembercontent(in%).Thelocationofthetransectinshownon theX-raymap;thewhitecirclesshowcorrespondingpointsbetweentheanalysisandthecrystal.Thestartingpointofthetransectistheupperrightportionofthecrystal.
Gapsinthetransectindicatethepresenceofmeltinclusions.Lowerpanel:lefthandside:Close-uppotassiumX-raymapsofarea(a),thelocationofwhichisshownon upperpanelmap.Noticethepatchinessofthecentralzones(labelledptch)andtheembayedzoneboundaries(labelledemb),bothtexturesindicativeofoutofequilibrium, resorptionevents.Righthandside:Close-uppotassiumX-raymapsofarea(b),thelocationofwhichisshownonupperpanelmap.Thismapshowsthestrongcontrast betweenthelasttwozonesclosesttotherimofthecrystal,therelativelystraightboundaryandgradualchangeofcomposition.
passinganelectricalcurrenttotheholdingPtwire(DiCarloetal., 2006),suchthatthesampledroppedintothecoldpartoftheves- sel,providingacoolingrateof>100◦C s−1.Aftereachexperiment, capsules were weighed to verify that no leakage had occurred.
Theywere then openedandpart ofthecharge (asa single frag- ment) was embedded in an epoxy resin and polished for SEM, secondaryionmassspectrometry(SIMS) andelectronmicroprobe analyses(seeSection3.3below).
3.2.Micro-tomography
X-raymicro-tomography(μ-CT)analysesofindividualanortho- clasecrystalswereperformedwithaPhoenixNanotom180atISTO
inOrléansinordertoconfirmwhetherornotthemeltinclusions they hosted were isolated fromeach other.We used a molybde- num target, tungsten filament, operating voltage of 100 kV, fil- ament currentof 80 μA, and2 s exposure time. Whole crystals (from 1 to 5 cm in length) were mounted on to carbon fibre rodsusingAralditeepoxyresin.Thesampleswererotatedthrough 360◦ during exposure. About 2300 imageswere collected during each analysis, each ofthembeingan average offourraw images takenat2 sintervals. Reconstructionoftheseimagesintoa stack of greyscaleimages representing differentphases was performed with a PC micro-cluster running commercial software.The voxel edge length for most scanswas 6–11 μm. This was sufficient to
4 Y. Moussallam et al. / Earth and Planetary Science Letters 413 (2015) 1–12
Fig. 2.EMPtransectreportedintermsoforthoclasecontent(in%)forcrystals(fromtoptobottom):Ere_1984_03_S2,Ere_1984_XC_06_S2,eb05027a2andeb05042b2.Gaps areduetomeltinclusions.
resolvethegeometryofmostmeltinclusions;however,inclusions smallerthan11 μmcouldnotbeimaged.
Image processing was handled using the ImageJ freeware (http://rsbweb.nih.gov/ij/).Originalimageswereconvertedto8 bits and a median filter was applied to reduce noise. Image con- trastwas enhancedusingafuzzycontrastenhancementalgorithm tuned tohighlight theinclusions. Imageswere then convertedto binary; outlier pixels were removed; anda second median filter was applied. Thisthresholded image set was combinedwith an- other set of images thresholded so as to characterise the melt outsidethe crystal. This enabled a particleanalyser algorithm to identify any melt inclusions that were connected with the glass surrounding the crystal. The algorithm was alsoused to identify anyinterconnectedinclusions.Finalimagesetswerehandledusing theVGStudioMAXsoftwaretoobtain3Dimagerepresentations.
3.3. Analyticaltechniques(SIMSandEMPA)
Electronmicroprobeanalysesofsyntheticanorthoclasecrystals, andnaturalmeltinclusionsandtheirhostcrystalzoneswereper- formedonaCamecaSX-100attheDepartmentofEarthSciences, University of Cambridge, and on a Cameca SX-100 at New Mex-
icoInstituteofMiningandTechnology.Rim-to-rimandcore-to-rim profileswereacquiredat10 μmstepspacingalong7000–8000 μm distances. For glass analyses, we used an accelerating voltage of 15 kV, a beamcurrentof 2 nAanda defocused beamof10 μm tomeasureSi,Ti,Al,Fe,Mn,Mg,Ca,NaandK.Forsyntheticmin- eralphases,thesameelementswereanalysedwithanaccelerating voltageof10 kV,a beamcurrentof6 nAanda focusedbeamat 1 μm.X-raymapswereacquiredusinga100 nAbeamcurrentand 15 kVaccelerating voltagewitha10–13 μmstepsize and12 ms dwelltime.NaandKwereanalysedfirstinordertominimiseal- kalilossduringanalysis.
SIMS analyses of volatiles (CO2 and H2O) in natural melt in- clusions and synthetic melts produced by solubility experiments were performedon a Cameca ims-4f anda Cameca ims-1280at the NERC IonMicroprobe Facility atthe University of Edinburgh.
OntheCamecaims-4f,measurementsweremadeusingaprimary O− ion beamwithan accelerating voltageof 15 kV anda beam currentof5 nA,asecondaryacceleratingvoltageof4500 Vminus a50 Voffset,anda25 μmimagefield.OntheCamecaims-1280, measurements were madeusing aprimary O− ionbeamwithan accelerating voltage of 10 kV and a beam current of 5 nA. On both instruments,theionbeamwas rasteredover an areaof ap-
