Internet & Web

Evolution of porosity and geochemistry in Marcellus Formation black shale during weathering

Description
Evolution of porosity and geochemistry in Marcellus Formation black shale during weathering
Categories
Published
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institutionand sharing with colleagues.Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third partywebsites are prohibited.In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further informationregarding Elsevier’s archiving and manuscript policies areencouraged to visit:http://www.elsevier.com/authorsrights  Author's personal copy Evolution of porosity and geochemistry in Marcellus Formation blackshale during weathering Lixin Jin a,b,c, ⁎ , Ryan Mathur d , Gernot Rother e , David Cole f  , Ekaterina Bazilevskaya b , Jennifer Williams b ,Alex Carone c , Susan Brantley b,c a Department of Geological Sciences, University of Texas at El Paso, El Paso, TX 79968, United States b Earth and Environmental Systems Institute, Pennsylvania State University, University Park, PA 16803, United States c Department of Geosciences, Pennsylvania State University, University Park, PA 16803, United States d Department of Geology, Juniata College, Huntingdon, PA 16652, United States e Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, United States f  Department of Geological Sciences, Ohio State University, Columbus, OH 43210, United States a b s t r a c ta r t i c l e i n f o  Article history: Received 9 November 2012Received in revised form 15 July 2013Accepted 16 July 2013Available online 27 July 2013Editor: J. Fein Keywords: Pyrite dissolutionOrganic matterShale gasTrace metalsNeutron scatteringComputed tomography Soils developed on the Oatka Creek member of the Marcellus Formation in Huntingdon, Pennsylvania were an-alyzed to understand the evolution of black shale matrix porosity and the associated changes in elemental andmineralogicalcompositionduringin fi ltrationofwaterintoorganic-richshale.Makingthereasonableassumptionthatsoilerosionratesarethesameasthosemeasuredinanearbylocationonalessorganic-richshale,wesuggestthat soil production rates have on average been faster for this black shale compared to the gray shale in similarclimate settings. This difference is attributed to differences in composition: both shales are dominantly quartz,illite, andchlorite, but the Oatka Creek memberatthislocation has moreorganic matter(1.25 wt.% organic car-bon inrockfragments recovered from thebottom of theauger cores and nearby outcrops) and accessory pyrite.Duringweathering,theextremelylow-porositybedrockslowlydisaggregatesintoshalechipswithintergranularpores and fractures. Some of these pores are either  fi lled with organic matter or air- fi lled but remainunconnected,andthusinaccessibletowater.Basedonweatheringbedrock/soilpro fi les,disintegrationisinitiatedwith oxidation of pyrite and organic matter, which increases the overall porosity and most importantly allowswater penetration. Water in fi ltration exposes fresh surface area and thus promotes dissolution of plagioclaseand clays. As these dissolution reactions proceed, the porosity in the deepest shale chips recovered from thesoil decrease from 9 to 7% while kaolinite and Fe oxyhydroxides precipitate. Eventually, near the land surface,mineral precipitation is outcompeted by dissolution or particle loss of illite and chlorite and porosity in shalechips increases to 20%. As imaged by computed tomographic analysis, weathering causes i) greater porosity,ii) greater average length of connected pores, and iii) a more branched pore network compared to theunweathered sample.This work highlights the impact of shale – water – O 2  interactions in near-surface environments: (1) black shaleweathering is important for global carbon cycles as previously buried organic matter is quickly oxidized; and(2) black shales weather more quickly than less organic- and sul fi de-rich shales, leading to high porosity andmineral surface areas exposed for clay weathering. The fast rates of shale gas exploitation that are ongoing inPennsylvania, Texas and other regions in the United States may furthermore lead to release of metals to theenvironment if reactions between water and black shale are accelerated by gas development activities in thesubsurface just as they are by low-temperature processes in our  fi eld study.© 2013 Elsevier B.V. All rights reserved. 1. Introduction Black shales cover only a small percentage of continental land areabut are economically and environmentally important (e.g., Tourtelot,1979; Falk et al., 2006; Shpirt et al., 2007; Piper and Calvert, 2009;Pollack et al., 2009). Speci fi cally, these shales often host signi fi cantamounts of methane that can be exploited pro fi tably. For example, theDevonian black shales of the northeastern U.S.A. are being developedfor natural gas by increasing shale porosity using hydraulic fracturing(Engelder et al., 2009). Here, we explore the major mineral – water – gasreactionsatEarthsurfacethatleadtoblackshalealterationandporositychanges.Analysis of shale weathering is also of importance for geologicallylong-term C and O balances. Speci fi cally, mineral weathering reactions, Chemical Geology 356 (2013) 50 – 63 ⁎  Correspondingauthorat:DepartmentofGeologicalSciences,UniversityofTexasatElPaso, El Paso, TX 79968, United States. E-mail address:  ljin2@utep.edu (L. Jin).0009-2541/$  –  see front matter © 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.chemgeo.2013.07.012 Contents lists available at ScienceDirect Chemical Geology  journal homepage: www.elsevier.com/locate/chemgeo  Author's personal copy release of volcanic gases, burial of organic matter, and weathering-drivenre-oxidationoforganicmatterovergeologictimeframesallcon-tribute to the global balance of atmospheric CO 2  and O 2  (Petsch et al.,2005). Furthermore, black shales are not only rich in organic material,but also in metals (Vine and Tourtelot, 1970; Shpirt et al., 2007).Whentheyareexposedatearth'ssurfaceorbydrilling,theycanreleasesigni fi cant amounts of metals to solution ( Jaffe et al., 2002; Tuttle andBreit, 2009; Tuttle et al., 2009). Indeed, black shale weathering world-wide is a major contributor to the global cycles of Mo, Os, carbon, andotherelementsandshalesareasourceofmetalcontaminantsinsurfacewaters ( Jaffe et al., 2002; Huh et al., 2004; Wilde et al., 2004; Petschet al., 2005; Pollack et al., 2009; Tuttle and Breit, 2009; Tuttle et al.,2009; Miller et al., 2011). Finally, the opening of fractures in shales byhydraulic fracturing stimulates the return to the surface of a variety of organic and inorganic constituents in  fl owback water that are also of signi fi cant environmental concern (Kargbo et al., 2010; Gregory et al.,2011).The target of this study is the black shale of the Middle DevonianMarcellusFormationwithintheHamiltongroup.Thisformationunder-liesmuchofPennsylvania,extendingintoOhio,WestVirginia,andNewYork(Obermajeret al.,1997; Faill,1998;Harper,2008).Depositeddur-ing a period of rapid transgression in anoxic waters less than 100 mdeep,theMarcellusShaleischaracterizedbyitsblackcolor,highpyriticcontent, organic-rich nature, enrichment of trace metals, and lack of fossils (Potter et. al., 1980; Roen, 1983; Obermajer et al., 1997; Shultz,1999; Sageman et al., 2003). In Pennsylvania, natural gas is currentlyexploited in the northeastern, north central, and southwestern regionsusing horizontal drilling followed by hydraulic fracturing to stimulateproduction (Harper, 2008; Soeder and Kappel, 2009). The opening of fractures stimulates the methane sorbed to organic or mineral matteror contained in pores to be released into the fracture network. The un-derlyinghypothesisofourworkisthatmineral – waterreactionsstudiedin the soil zone can contribute to our understanding of deep shale – water interactions and the potential for contamination during shalegas exploration and recovery. Fractures formed during hydraulicfracturing and the porosity network created during weathering arevery different but both expose fresh mineral surface area to reactive fl uids and a better understanding of the low-temperature processescan help provide fundamental understanding for the prediction of water – rock reactions during shale – gas development.We investigate weathering of the Marcellus shale where it is ex-posedinHuntingdon,Pennsylvania,usingcombinedneutronscattering,computed tomography, elemental analysis, and X-ray diffraction. Thislocation is one of satellite sites for the Susquehanna Shale Hills criticalzone observatory, SSHO. The goals of this work are to: (1) identify theimportant mineralogical reactions during weathering; (2) determinethe primary porosity of the Marcellus shale and how the pores openor close in the rock as weathering proceeds; and (3) compare andcontrast reaction rates with a similar study on organic-poor gray shale(Rose Hill shale in SSHO) located within 20 miles ( Jin et al., 2010).This study complements our previous study on Cu mobility andtransport during shale weathering at the same location presented byMathur et al. (2012). 2. Methods  2.1. Rock and soil sampling  Weinvestigatedmineralweatheringandelementalmobilityonmid-Devonian Marcellus black shale on a forested northwest-facing planarhillslope located in Jackson Corner, Huntingdon County, Pennsylvania(Fig. 1). Our focus is an exposed section of the Oakta Creek unit thatforms the basal portion of the Hamilton Group (Obermajer et al., 1997;Faill, 1998; Soeder, 2010). This unit is currently being explored inten-sively for shale gas in other parts of Pennsylvania where the unit re-mains buried and still contains gas. The Huntingdon area was logged2 – 3 times in the past 200 years like most of central Pennsylvania, andiscurrentlyunderprivateownership.Thecurrentvegetationisamaturemaple – pine forest.The  fi eld sites, previously described in Mathur et al. (2012), arebrie fl y summarized here. The location is a convex-upward hillslopealong a zero th -order catchment (i.e. water  fl ows only ephemerally inthe low area). A total of three soil pro fi les were sampled: two alongthe ridgetop within 10 m of one another (RT 1 , RT 2 ), and one at thebaseofthehillslope(VFS).TheelevationdropfromRTtoVFSisapprox-imately 30 m. At these sites, soils were hand-augered until it wasimpossible to auger further (refusal). The three cores varied in totaldepth, described here as soil thickness, ranging from 84 cm at thebase of the hill to 134 cm at the ridgetop. Soil samples were collectedevery 5 to 15 cm, with depth 0 de fi ned as the organic soil mineral soilinterface (O – A interface).A total of 8  “ parent rock ”  samples from the Oakta Creek memberwerealsocollectedandanalyzed.Parentrockhypotheticallyrepresentstheinitialcompositionofthebedrockbutasdiscussedlater,thesesam-ples have been slightly altered. Four of these putative parent sampleswere rock fragments collected from the bottom of the three augeredcores.Inaddition,fourhandsampleswerecollectedfromaroadoutcropat the base of the zero th -order watershed within 5 m of the augeredcore VFS (Fig. 1). Visual inspection of the outcrop reveals that beddingin the unit is subhorizontal. Sub-perpendicular jointing was also ob-served that is typical of Alleghanian fracture patterns throughout theregion.Porosity and permeability are known to vary within the MarcellusFormation due to variations in texture, chemistry, mineralogy, organiccontent, thermal maturity, fracture spacing, veining, and stratigraphicrelationships (Soeder, 1988; Arthur et al., 2008). The most organic-rich member of the Marcellus Formation that is currently targeted forshale gas is the Union Springs member. To place our study of mineral –   1  0  0  0  f  t 9 0 0 f  t  RT 1 RT 2 VFS N 100 m   O  u  t c  r o  p   f o  r   s  a  m  p   l  i  n g    p  a  r e  n  t       E    a    s     t       B    r    a    n    c      h      R     d Philadelphia Pennsylvania, USA MRTMMSMVF Bulk density coresSoil chemistry cores Fig.1. StudysitesonMarcellusshaleincentralPennsylvania(modi fi edfromMathuretal.,2012). Soils werecollectedfrom theridge top(RT 1 , RT 2 ) and valley fl oor(VFS) sites;bulkdensity cores were collected at the ridge top (MRT), mid-slope (MMS) and valley  fl oor(MVF). All sites are situated along a hillslope which is roughly planar. Rock fragmentswere sampled at an outcrop at the base of the planar hillslope and also from the bottomof soil pro fi les and were used to approximate parent Marcellus shale.51 L. Jin et al. / Chemical Geology 356 (2013) 50 – 63  Author's personal copy fl uid reactions in the Oatka Creek member in context of the moreorganic-rich Union Springs, we also report the chemistry of 10 coresamples from the Purcell, the Union Springs, and the Onondoga mem-bers of the Marcellus Shale drilled at Howard, PA (Bald Eagle core).Three more cores were also collected along the planar hillslope(de fi ned by soil sites RT 1 , RT 2  and VFS) at the ridgetop (MRT), middleslope (MMS), and valley  fl oor (MVF) on November 22, 2011 for mea-surement of bulk density following the method of  Blake and Hartge(1986). To measure bulk density, several samples were collected fromaugeredholesatvariousdepthsusingasamplerofknownvolume.Sam-ples were then transferred to polyethylene bags, weighed, dried at105 °C in the laboratory, and weighed again after drying. Bulk densitywascomputedusingthemassafterdrying.Thevolumetricsoilmoisturecontent was calculated from the difference in the before- and after-drying mass, using the volume ratio of soil moisture (converted frommass of soil moisture) and bulk soil, expressed as %.  2.2. Solid sample preparation and analysis After collection,thesoil androcksamples were dried and ground topass through a 100-mesh sieve ( b 150  μ  m). Total carbon, nitrogen andhydrogen contents were measured on a CHNS-O Elemental Analyzer(modelEA1110,Leco, St.Joseph, MI)atPennState AgriculturalAnalyt-ical Services Laboratory using ground samples. Total sulfur concentra-tions were measured by a LECO Sulfur coulometer. For this analysis,about 0.4 g of each soil and rock sample were weighed into ceramiccruciblesandcombustedinafurnaceusingtinandironbeadsasacom-bustionadditive.ThereleasedSO 2 gaswastitratedwithaKIO 3 solution.One standard was used for calibration.To analyze for elemental chemistry of soils and parent rock, groundsamples were fused directly with lithium metaborate at 1000 °C andre-dissolved in 5% nitric acid following standard procedures (Suhr andIngamells,1966).Solutionswerethendilutedandanalyzedforelementalconcentrations using inductively coupled plasma atomic emission spec-trometry (ICP-AES) on a Perkin-Elmer Optima 5300 at the PennsylvaniaState University. USGS and NIST reference rocks of known compositionswererunusingthesameprotocolandusedascalibrationstandards.Thedetection limit is about 0.05 wt.% and precision is ±5% for concentra-tions of major elements. For all samples, loss on ignition (LOI) wasnot measured by combustion but is reported here simply as the differ-ence between the summation of major element oxides and 100% andgenerally includes volatiles from combustion of carbonates, sul fi des,organic matter, and clays.Elemental concentrations were used to determine enrichment ordepletion throughout the soil pro fi le relative to elemental concentra-tionsintheparentrockmaterialbycalculatingthemasstransfercoef  fi -cient  τ  (Brimhall and Dietrich, 1987; Anderson et al., 2002): τ  i ;  j  ¼ C   j ; w C  i ;  p C   j ;  p C  i ; w − 1  ð 1 Þ Here  C   refers toconcentrationsof immobile ormobile elements( i ,  j ,respectively)inparent(  p )orweatheredmaterial( w ).WeselectedTiasthe immobile element ( i ), as Ti was mainly present as rutile andappeared inert in the soil pro fi le according to concentration data(Mathur et al., 2012). Parent values were calculated as the average of the 8 putative parent samples, and uncertainties were presented asonestandarddeviationforeachelement.Theerrorsin τ wereevaluatedby propagating uncertainties from elemental concentration analyses of the soils (the analytical uncertainty) and the parent rock (the standarddeviation around the mean of the 8 parent samples) as described in anearlier publication ( Jin et al., 2010).Soil mineralogy has been previously reported in Mathur et al.(2012). For this study, samples were further characterized to revealthe detailed clay mineralogy using standard heat and chemical treat-ments for clays as described in Jin et al. (2010).Because of the higher S, organic matter, and carbonate contents inthe Bald Eagle core, the analytical technique was modi fi ed. Speci fi cally,thesamples were ground and then ashed at900 °Cto measure theLOI.However,Limetaboratedigestionoftheseashedsamplesyieldedchem-icalanalyses whosetotals did notadd up to 100 wt.%. Suchanobserva-tioniscommonwhenallvolatileshavenotbeenreleasedduringashing.WeinferredthatShadcombinedwithCainthesamplesduringheatingand that the ashed samples therefore still contained CaSO 4 . To correctfor this, the S content of the ash was analyzed and added to the totals(as SO 3 ). After that correction the totals equaled 100 ± 2.3 wt.%. Allthe oxide concentrations were corrected for this S and reported on asreceived basis.To measure organic carbon content of Bald Eagle core samples,carbonate minerals (inorganic carbon) were  fi rst removed by 2Nhydrochloric acid and the residue was dried and measured on aCHNS-O Elemental Analyzer. To measure inorganic carbon content of Bald Eagle core, hydrochloric acid was injected into a sealed serumvial containing a pre-weighed powder sample. The amount of CO 2  re-leased was calculated by multiplying the headspace volume by theCO 2  concentration (analyzed by a CO 2 /H 2 O Analyzer (LiCor-7000) attheBiogeochemistryLaboratoryintheDepartmentofCropandSoilSci-ences,PennState).AcalibrationcurvewascreatedusingCO 2 standardsfrom 970 to 10,010 ppm.  2.3.Porosityandinterfacialcharacteristicsassessedwithneutronscattering  To investigate the development of porosity during weathering,we carried out neutron scattering experiments on nine shale rockfragments collected from different depths of the ridgetop soil pro- fi le RT 1  (Fig. 1): RT 1 -10, RT 1 -26, RT 1 -34, RT 1 -52, RT 1 -82, RT 1 -98,RT 1 -115, RT 1 -119A, and RT 1 -119B. The numbers refer to the top of the sampling depth below land surface in cm. The last two chips(RT 1 -119A, B) were recovered from the bottom of the augered core(thesesamples werealsoanalyzed asputativeparentrockasdescribedpreviously).Rock fragments were generally angular with one dimension signi fi -cantly smaller than the other two dimensions: in other words, thesefragmentsaretypical fl atshalechipswherethebeddingplaneisparallelto the plane de fi ned by the two larger dimensions of the chip, and per-pendiculartothesmallestdimension.Topreparesamplesforscattering,thechipsweredriedatroomtemperatureandwerecutparalleltobed-ding. The samples were uniformly cut to be 150  μ  m thick to preventmultiple scattering (Anovitz et al., 2009). Thin sections were commer-ciallypreparedonquartzslides(standard boron-containingglassslidesabsorb neutrons).Theprinciplesofsmall-angleneutronscatteringanditsapplicationtorock nanostructure studies are discussed in the literature (e.g., GuinierandFournet,1955;Radlinski,2006),andwehaverecentlyusedthistech-nique to characterize weathering pro fi les of a gray, organic-poor shale( Jinetal.,2011b).Insmall-angleneutronscattering(SANS),acollimatedneutron beam is elastically scattered by the sample. Position-sensitivedetectors measure the scattering intensity  I  ( Q  ) as a function of the scat-tering angle,whichisde fi ned asthe angulardeviationfrom theincidentbeam. The scattering vector  Q   in units of Å − 1 , is related to scatteringangle  θ  by  Q   = (4 π / λ ) sin( θ /2), where  λ  is the wavelength of the neu-trons. Thus, the size range of features accessible with neutron scatteringdepends on the neutron wavelength  λ  and the range in scattering angle θ . SANS data over a given  Q  -range (Q  min , Q  max ) contain informationabout particles with approximate dimensions of 1/ Q  max  b  r  b  1/ Q  min .Thescatteringintensityofstructuralfeaturesisdirectlyproportionalto the scatteringcontrast:  I  ( Q  ) =  ϕ V   p (  ρ 1 ⁎ −  ρ 2 ⁎ ) 2 ×  P  ( Q  ) ×  S  ( Q  ). Here,  ρ 1 ⁎  and  ρ 2 ⁎  are the coherent scattering length densities (SLDs) for neu-trons for phases 1 and 2, respectively. The terms  P  ( Q  ) and  S  ( Q  ) denotethe form and structure factors, for which analytical expressions existfor different geometries of scatterers, including mass and surface frac-tals (Radlinski, 2006). The terms  ϕ  and  V   p  are, respectively, the volume 52  L. Jin et al. / Chemical Geology 356 (2013) 50 – 63  Author's personal copy fractionofthedispersedphaseandthevolumeperscatterer.TheSLDof phase  j isgivenby:  ρ   j  ¼ ∑ N i ¼ 1 b i  ρ  j N   A M   j ,with b i equaltotheboundcoherentscatteringlengthofatom i , N  thetotalnumberofatomsinthecompound,  ρ  j  the mass density,  M   j  the molar mass, and  N   A  the Avogadro constant.The SLD is weakly negative for hydrocarbons, but the contrast betweenmineralsandhydrocarbonisapproximatelythesameasthatofmineralsand air- fi lled pores (Hall et al., 1983). Thus air- fi lled and organic- fi lledpores both have SLD values of about 0. Most minerals and rocks haveSLDs on the order of (3.5 – 4.5) ∗ 10 − 6 Å − 2 , creating strong scatteringfrom mineral – pore interfaces, regardless of air- or hydrocarbon in fi lling.In contrast, mineral – mineral interfaces usually scatter at larger lengthscales and possesssigni fi cantlylessscatteringpower.Therefore,allscat-tering in rocks can be attributed to mineral – pore features, regardless of whether the pores are  fi lled with hydrocarbons or air. This so-calledtwo-phase approximation has been commonly used in the literature tointerpretneutronscatteringforrocks(Porod,1952).TheSLDofthemin-eral matrix was calculated from the elemental composition followingprocedures described in our previous study on shale ( Jin et al., 2011b),using the NIST SLD calculator (http://www.ncnr.nist.gov/resources/sldcalc.html). Here, we assumed the grain density of the Marcellusshale to be 2.6 g/cm 3 . The formula of an  ‘ arti fi cial ’  compound was thencalculatedwiththestoichiometrydeterminedfromtheelementalchem-istry of the parent (SupplementTable 1), including the total carbon andhydrogen contents. Errors in SLDfrom the assumption that all the chipshavethesamechemicalcompositionlikelyarenegligiblewithin5%.Thisestimatewasbasedonourpreviousstudy:SLDvaluescalculatedforasetofgrayshalechipsfromasimilarsoilpro fi leintheShaleHillscatchmentwithin30 milesofthesitestudiedheredifferedfromonetotheotherbyonly 0.03 × 10 − 6 Å − 2 (average =3.71 × 10 − 6 Å − 2 ) despite a relative-ly large variation in weathering extent ( Jin et al., 2011b).SANS and USANS experiments were carried out at the NIST Centerfor Neutron Research (NCNR). A Cd mask was af  fi xed to the front of the sample to ensure that identical regions were illuminated in theSANS and USANS measurements. Measurements were carried out atthe NG7 30-m SANS (Glinka et al., 1998) and BT5 USANS (Barker et al., 2005) beamlines. Details about these instruments are given onthe NCNR website (http://www.ncnr.nist.gov/instruments). Scatteringdata were corrected for emptycellscattering, background, andneutrontransmission. SANS and USANS raw data were normalized to absoluteintensities (Kline, 2006).  2.4. Computed tomography Neutron scattering (NS) yields statistical information about the po-rosity, but the information is obtained in inverse space and thus the in-formation about pore shapes is indirect. In addition, SANS and USANSonly can detect pore features in the range of approximately 0.5 nm to3  μ  m. In order to determine pore shapes and probe larger pores, wealso used computed tomography to investigate one representative fria-ble and weathered shale chip (~3 × 3 × 5 mm) recovered from 40 cmsoil depth from the MMS site. The images for the chip were collectedby synchrotron X-ray-based 3-D  μ  -computed tomography ( μ  -CT) atBeamline 8.3.2 at the Advanced Light Source, Lawrence BerkeleyNational Laboratory. The air-dried sample was mounted on aspeci fi cally designed sample holder and was illuminated with25 keV X-rays. These X-rays penetrate the sample but are attenuatedas a function of density. Lower density regions or phases (pores withair, water, or organic matter) attenuate X-rays to a lesser degree.Phase-contrast imaging was used to visualize small density variationsand to differentiate organic matter and low-density minerals from theair- fi lled pores (Groso et al., 2006; Langer et al., 2008). We also usedthe X-ray detector with 0.9  μ  m pixel size resolution to make non-destructive measurements of porosity. We were able to detect poresof diameter  N ~3  μ  m (3 pixel resolution).Filteredback-projectiontomographicreconstruction(KakandSlaney,1988) was accomplished using Octopus software (Masschaele et al.,2005). As a result, 2044 image slices, each 0.9  μ  m thick, were acquiredas output. Further image analysis included sub-sampling, anisotropicdiffusion  fi ltering, image smoothing, threshold-based segmentationand skeleton analysis. Skeleton analysis, also called medial axis, allowsvisualization of the network of interconnected one-dimensional paths(branches) for transport. We completed all image analysis with ImageJsoftware (Abramoff et al., 2004). 3. Results  3.1. Rock and soil chemistry 3.1.1. Parent rock The 8 rock fragments of putative parent Marcellus shale (OaktaCreek member) showed relatively small variations in concentrationsof K, Mg, Al and Si oxides. However, concentrations of the low-abundance elements Ca, Fe and Na vary signi fi cantly (SupplementTable 1). Among all the major oxides, Fe 2 O 3  contents vary the most,ranging from 2.0 to 7.0 wt.%. Of the 4 samples of putative parent rockfragments analyzed, we observed 1.16 – 1.39 wt.% C, 0.19 – 0.20 wt.% N,and 0.68 – 0.70 wt.% H. Duplicate analyses of the rock fragment recov-ered from the bottom of RT 1  showed 0.025 ± 0.005 wt.% S (Table 1).  3.1.2. Deep core samples The 10 samples from the Bald Eagle core from the Purcell, UnionSprings, and Onondoga members of the Marcellus formation (from767 to 923 ft depth) differed in chemistry from the 8 rock fragmentsfrom the Oakta Creek member exposed at the land surface near ourweathering study (Supplement Table 1). Speci fi cally, CaO contents inmost of the samples of the Union Springs range from 1.5 to 5.6%, withone sample containing a calcite vein with CaO at 36.4%. Values of LOIvaried between 8.6 and 15.5% (the veined sample had 30.0%). Asdiscussedbelow,thesehigherCaOandLOIconcentrationsareattributedtothepresenceofhighcarbonatecontentsincertainsamples.Quantita-tive XRD of the core samples reveals the presence of quartz, illite, andchlorite ± calcite. Minor minerals include K-feldspar, plagioclase, py-rite and rutile. These 10 deep core samples have about 0.26 – 1.53 wt.%S (average: 0.93 ± 0.42%), much higher than the rock fragments fromthe Oakta Creek member near Huntingdon.  3.1.3. Soils All soils except for the two shallowest samples from RT 1  and RT 2 have lower total C and N concentrations than the local parent rock(Table 1; Fig. 2A and B). Likewise, the S concentrations in the RT 1  soilsare ~0.01 wt.%, lower than the putative parent and the samples fromthedeepcoreofMarcellusFormation(Fig.2C).Incomparisontoparentrock,thesoilisalsodepletedinK,Al,andMg(Table1).Furthermore,theshallower soils closer to land surface become generally even moredepleted in K, Al, and Mg compared to deeper samples. In contrast,the Ca and P concentrations and LOI are relatively higher in shallowsoils than those in the parent. Concentrations of Fe oxides, however,show no obvious variations with depth.The τ pro fi lesforAl,Mg,NaandKcanbeclassi fi edasdepletionpro- fi les (Brantley and White, 2009) for all three sites, i.e., these elementshave beendepleted throughout theregolith compared tothe immobileelementTiintheparent(Fig.3).Thefractionaldepletionwithrespecttothe average assumed parent composition (= − τ ) extrapolated to theland surface is approximately 70% for Al, Mg and K. Interestingly, ridgetopsoilsRT 1 andRT 2 showsharpdepletionfrom0to30 cm,butremainrelatively constant at ~10% depletion below that. All soils, however,have  τ Si  and  τ Fe  values around 0, documenting little to no loss or gainofSiandFe.Incontrasttotherelativelysimpleandconsistentdepthbe-haviorofAl,Mg,Na,Fe,andSi,thelowconcentrationelementsCaandPexhibit variable behavior (Table 1; Fig. 3). 53 L. Jin et al. / Chemical Geology 356 (2013) 50 – 63
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x