lable at ScienceDirect

Marine and Petroleum Geology 27 (2010) 116ndash125

Contents lists avai

Marine and Petroleum Geology

journal homepage wwwelsevier comlocate marpetgeo

Microstructures of structure I and II gas hydrates from the Gulf of Mexico

Stephan A Klapp a Gerhard Bohrmann a Werner F Kuhs b M Mangir Murshed b Thomas Pape aHelmut Klein b Kirsten S Techmer b Katja U Heeschen a Friedrich Abegg a1

a MARUM Center for Marine Environmental Sciences and Department of Geosciences University of Bremen Klagenfurter Strasse GEO Building PO 330440 D-28334 Bremen Germanyb GZG Abt Kristallographie Universitat Gottingen Goldschmidtstr 1 D-37077 Gottingen Germany

a r t i c l e i n f o

Article historyReceived 20 November 2008Received in revised form25 March 2009Accepted 27 March 2009Available online 5 April 2009

KeywordsGas hydrateGulf of MexicoMicrostructureStructure IIStabilityElectron microscopyX-ray diffractionPorosity

Corresponding author Tel thorn49 421 218 65055 fE-mail address sklappmarumde (SA Klapp)

1 Present address Leibniz-Institut fur MeereswiWischhofstrasse 1-3 D-24148 Kiel Germany

0264-8172$ ndash see front matter 2009 Elsevier Ltddoi101016jmarpetgeo200903004

a b s t r a c t

Gas hydrate samples from various locations in the Gulf of Mexico (GOM) differ considerably in theirmicrostructure Distinct microstructure characteristics coincide with discrete crystallographic structuresgas compositions and calculated thermodynamic stabilitiesThe crystallographic structures were established by X-ray diffraction using both conventional X-raysources and high-energy synchrotron radiation The microstructures were examined by cryo-stage Field-Emission Scanning Electron Microscopy (FE-SEM) Good sample preservation was warranted by the lowice fractions shown from quantitative phase analysesGas hydrate structure II samples from the Green Canyon in the northern GOM had methane concen-trations of 70ndash80 and up to 30 of C2ndashC5 of measured hydrocarbons Hydrocarbons in the crystallo-graphic structure I hydrate from the Chapopote asphalt volcano in the southern GOM was comprised ofmore than 98 methane Fairly different microstructures were identified for those different hydratesPores measuring 200ndash400 nm in diameter were present in structure I gas hydrate samples no such poresbut dense crystal surfaces instead were discovered in structure II gas hydrate The stability of the hydratesamples is discussed regarding gas composition crystallographic structure and microstructureElectron microscopic observations showed evidence of gas hydrate and liquid oil co-occurrence ona micrometer scale That demonstrates that oil has direct contact to gas hydrates when it diffusesthrough a hydrate matrix

2009 Elsevier Ltd All rights reserved

1 Introduction

Gas hydrates are crystalline compounds of light hydrocarbons(methane ethane etc) or non-hydrocarbon compounds (carbondioxide hydrogen sulfide nitrogen) and water molecules whichcombine to build various hydrate structures (Sloan and Koh 2007)The formation of gas hydrates is controlled by an array of physicaland chemical parameters Availability of hydrate-forming gases isthe prerequisite hydrates only crystallize when the gas concen-trations at least locally and temporarily exceed solubility (Kven-volden 1993) Hydrostatic pressure as given by the water depthsediment temperature and pore water salinity are the limitingformation factors for marine gas hydrates which also govern thethickness of the gas hydrate stability zone

ax thorn49 421 218 65099

ssenschaften IFM-GEOMAR

All rights reserved

Gas hydrates are considerable methane reservoirs (Collett2002 Max et al 2006) and may turn into future sources of fossilfuels The global methane hydrate inventory can be considered asa chemostat which may partly control climate cycles (Kennettet al 2002) The total amount of carbon stored in gas hydrates isstill uncertain Buffett and Archer (2004) figured 3000 Gt carbonbeing stored in gas hydrates instead Klauda and Sandler (2005)came up with 74400 Gt of methane being stored as hydrates

Hydrates form different crystallographic structures Structure I(sI) gas hydrate is often addressed as methane hydrate becausemethane is the main guest molecule sometimes amended by slightfractions of ethane and hydrogen sulfide In many natural depositslike at several sites at the Hydrate Ridge offshore Oregon (Suesset al 2001 Torres et al 2004) or at the Blake Ridge (Dickens et al1997 Matsumoto et al 2000) methane originates predominantlybut not exclusively from microbial methane production Becausemethane from biogenic origin is far more profuse than higherhomologues predominantly derived from thermogenetic catalysisgas hydrate sI is considered as the most abundant hydrate structureon earth (Kvenvolden 1995) despite its greater limitationsin stability conditions compared to the other two structures

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 117

(Sloan and Koh 2007 Lu et al 2007) Structure II and structure Hhydrates (sII and sH) derive hydrocarbons primarily from thermo-genic origin such compounds are methane through butanes for sIIand methane through pentanes for sH Thermogenic origin ofhydrocarbons in gas hydrates has been reported for instance in thenorthern Gulf of Mexico (GOM) (Sassen et al 2001a) and in theBarkley CanyonCascadia margin (Lu et al 2007)

In general hydrate structures are related to the molecule diam-eters of present hydrate-forming gases however laboratory exper-iments have shown that before a long-term stable structure isequilibrated transient existences of metastable structures are found(eg Klapproth et al 2003) Gas hydrate structures I and II are quitedistinguishable by their X-ray diffraction (XRD) pattern Rietveldanalysis (Toby 2001) of the XRD data allow for quantitative analysisof different phases occurring within a sample ie to assess theamounts of ice different gas hydrate structures and other phases

Microstructural observations of gas hydrates can be bestobtained by electron microscopy For convenience and in contrastto the usual nomenclature in material science lsquolsquomesoporesrsquorsquo areunderstood here as pores in the range of hundreds of nanometers indiameter A mesoporous microstructure is known for laboratoryproduced synthetic hydrates made of argon (Ar) nitrogen (N2)carbon dioxide (CO2) and methane (Kuhs et al 2000 Klapprothet al 2003 Staykova et al 2003) as well as natural sI hydrates(Kuhs et al 2004 Techmer et al 2005) Mesoporous and nonpo-rous hydrate microstructures are shown by Stern et al (20042005) although there the microstructure was not directly relatedto the crystallographic structure or the gas chemistry of thehydrates Interestingly for all investigated mesoporous hydratesthe pore diameter has been observed in the same range of 200ndash400 nm (with the notable exception of CO2 hydrates with poresbelow 100 nm as discussed in Kuhs et al 2000) It did not matterwhether synthetic or natural hydrates were observed nor whethera hydrate sample stems from deep within the gas hydrate stabilitynor whether it is close to the phase boundary Macropores in themicrometer range were identified as originating from gas hydratedecomposition (Kuhs et al 2004 Stern and Kirby 2007) Pores oftens of micrometers in diameter were found in methane-domi-nated sI gas hydrates from the Hydrate Ridge and were alsoattributed to gas hydrate dissociation upon recovery (Bohrmannet al 2007) Each such decomposition pore has a different diameterin the few micrometer range moreover the size-normalized poredensity is much smaller for decomposition pores than for meso-pores Some pores or rather bubbles could be inclusions of free gasinto the gas hydrate matrix Such free gas bubbles can be consid-ered as the original part of the in situ structure or microstructure(Bohrmann et al 1998) whereas decomposition pores are

Table 1Gas hydrate samples presented in this paper TVG TV-guided grab GC gravity corer In

Latitude [N] Longitude [E] Water depth [m] Cruise sampli

Northern Gulf of MexicoGreen Canyon RegionGC185 (Bush Hill)274695 913044 547 SO 174 station274697 913047 547 SO 174 station

GC415273261 905955 1049 SO 174 station273347 905885 951 SO 174 station

Southern Gulf of MexicoChapopote Region215400 932624 2902 SO 174 station215395 932621 2904 M672 GeoB 1215395 932621 2904 M672 GeoB 1215395 932621 2904 M672 GeoB 1

secondary alterations These gas bubbles would add to the porosityestimation (eg Suess et al 2002)

This paper assesses the relation between gas hydrate crystallo-graphic structure and microstructure We present new insights onthe microstructure of natural sI and sII hydrates formed frommethane through butane gases in the GOM We combine resultsfrom XRD gas chromatography and Field-Emission ScanningElectron Microscopy (FE-SEM) for individual samples fromdifferent locations in the GOM and attribute diverse microstructureobservations to the corresponding gas chemical and crystallo-graphic findings

2 Geological setting and sampling in the Gulf of Mexico

Fluid migration in the GOM transports hydrocarbons fromsubsurface reservoirs to the seafloor at numerous sites Thicksediment layers on the continental margin cover Early Jurassic saltsheets (Humphris 1979) which create salt ridges and basinsthrough halokinetic mobility (Rowan et al 1999) Deeply buried oiland gas reservoirs are tapped by salt-tectonic-induced faults whichserve as pathways for hydrocarbons to shallower reservoirs andeventually to the seafloor The depletion of heavy hydrocarbons orthe stripping of light hydrocarbons (Chen et al 2004 Whelan et al2005) results in gases ascending to near-surface sediments wheregas hydrates can form as part of seep related processes (Sassenet al 2001a MacDonald et al 1994 2003 Chen and Cathles2003) Gas hydrate presence in the northern GOM was firstreported in 1984 (Brooks et al 1984) and has been extensivelystudied henceforward Collett and Kuuskraa (1998) calculated thetotal amount of hydrate-captured gas in the GOM area to be 701 GtEstimations should be treated carefully though since at individualsites elevated salinities and temperatures prevent hydrate forma-tion despite high gas fluxes (Ruppel et al 2005)

Samples investigated in this study were retrieved from thenorthern and southern GOM during the RV Sonne cruise SO 174 infall 2003 (Bohrmann and Schenck 2004) and RV Meteor cruiseM672b in spring 2006 (Bohrmann and Spiess 2008 Table 1)During the Sonne cruise gas hydrates were recovered from threeareas (Fig 1) From two different sites in the Green Canyon offshoreLouisiana in the northern GOM (block CG415 as used by the UnitedStates Mineral Management Service) in about 1050 m water depth(named TVG-3 and TVG-12 in this paper see Table 1) from twosites in the Bush Hill area (block GC185) which is part of the GreenCanyon (these locations are called TVG-10 and GC-2) and from onesite at the Chapopote asphalt volcano (sample TVG-6) in theCampeche Knolls area in the Southern GOM On the M672b cruiseadditional hydrates were sampled at one gravity coring station

the text the short name is used

ng station tool Short name (in text) Total core recovery [m]

47 GC-2 GC-2 28157 TVG-10 TVG-10

96 TVG-3 TVG-3169 TVG-12 TVG-12

140 TVG-6 TVG-60618 GC-6 1 GC-6 1 10618 GC-6 2 GC-6 2 (sub-sample 1) 10618 GC-6 2 GC-6 2 (sub-sample 2) 1

Fig 1 Map of sampling locations in the Gulf of Mexico For station numbers see Table 1Lines indicate the cruise track

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125118

from the Chapopote asphalt volcano (Fig 1) Samples from thiscruise are named GC-6 with sub-samples

The Bush Hill area is located in relatively shallow water depths(500ndash600 m below sea level mbsl) and is fed by gas from the Jollietgas field (Chen and Cathles 2003) from which free gas and oildischarges into the water (Sassen et al 2001b MacDonald et al2003) The oil-stained gas-laden sediments host chemosyntheticcommunities (MacDonald et al 2003) Compositions of lighthydrocarbons (C1ndashC4) should result in the formation of sII gashydrates as suggested by many authors especially for the Bush Hillregion (eg Sassen et al 1994 2001b Chen and Cathles 2003MacDonald et al 2005) Structure II gas hydrate was recentlyidentified by XRD in Green Canyon block GC234 (Rawn et al 2008)Gas hydrates occur both as white precipitates as well as brownish-yellowish colored from oil perfusion through the hydrates (Fig 2see MacDonald et al 1994 2003 Bohrmann and Schenck 2004)

Fig 2 Gas hydrate specimen associated to oil-s

The Chapopote asphalt volcano (in the following only Chapo-pote) is located in deep water (3000 mbsl) northwest of theYucatan peninsula (Fig 1) The Chapopote is part of the CampecheKnolls salt diapirs which are special due to the outflow of naturalasphalt at the seafloor on top of the salt domes (MacDonald et al2004) The salt-tectonic-induced faults allow for fluid migrationfrom subsurface hydrocarbon reservoirs (Ding et al 2008) andseafloor seepage Sufficient gas supply at the Chapopote seep sitesevokes the formation of shallow gas hydrates

3 Methods

Gas hydrate samples were recovered either by TV-guided grabor gravity corer (see Table 1) The Bush Hill samples were coveredonly by a very thin sediment layer the other hydrate samples wereretrieved from shallow sediment depths such that the hydrateswere not directly exposed to seawater Immediately after recoverythe samples were quenched in liquid nitrogen (LN2) to avoid furtherdecomposition of the hydrates Henceforward the hydrates werealways kept in LN2 in a continuous cooling chain when transportedor prepared for measurements in the home lab

Bulk hydrate samples from each location were split into severalsub-samples They were prepared in LN2 in a cold room (10 C to20 C) with air conditioning providing dry air to avoid conden-sation of atmospheric water as ice crystals on the surface of thespecimen The sub-samples were then used for electron micros-copy and XRD For the latter sub-samples from each location weremeasured both with a laboratory diffractometer at the University ofGottingen Germany as well as at high-energy synchrotron beamlines in Hamburg Germany

31 X-ray diffraction

XRD analysis was used for quantitative assessment of the majorphases in the samples For the in-house measurements in Gottin-gen a Bragg-Brentano focusing Philips MRD (Materials ResearchDiffractometer) cryo-goniometer was used which employsa wavelength of 07093 Aring from a molybdenum anode Thesynchrotron measurements were conducted with a wavelength ofw012 Aring at the Beam Line BW 5 as well as at the HARWI II beam lineat the Hamburg Synchrotron Laboratory (HASYLAB) of the DeutscheElektronen Synchrotron (DESY) in Hamburg Germany The pene-tration power was sufficient to study larger samples in aluminumcans of 7 or 12 mm inner diameter which were located in a cryostatwith aluminum vacuumheat shields The short wavelength of thehigh-energy synchrotron radiation allowed for bulk sedimentsample investigations All synchrotron scans were performedexclusively at distinct locations of uncrushed gas hydrate pieces

tained sediments (TVG-10 from Bush Hill)

Fig 3 Rietveld plots of the synchrotron X-ray diffraction data of the hydrate samples(wavelength 01197314 Aring) The solid black lines open circles and solid grey linesrepresent the calculated pattern the observed intensity data and the differencebetween observation and model calculation respectively The vertical solid and dashedlines show possible positions of Bragg reflections of hexagonal ice Ih and hydraterespectively Note different scaling for the x-axis (a) The sample contains mainly sIhydrate (w88 the remainder being ice Ih) collected from the Chapopote (SO 174station 140 TVG-6) (b) Sample contains about 95 sII hydrate collected from the BushHill (SO 174 station 47 GC-2)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 119

comprising crystals of ca 200ndash300 mm (Klapp et al 2007) and noton powdered samples Accordingly structural information wasobtained for the scanned volumes containing coarse crystals andnot for the whole powder sample In contrast the cryo-MRDmachine gave averages on the complete powder sample Details ofthe XRD measurements are given in Bohrmann et al (2007) andKlapp et al (2007) Phase fractions of the samples are then calcu-lated from Rietveld refinement using the GSAS software (Larsonand van Dreele 1994) and the EXPGUI user interface (Toby 2001)

32 FE-SEM investigations

Gas hydrate pieces which were previously characterized byXRD were analyzed by SEM For electron microscopic investiga-tions two electron microscopes were applied one is a Cryo-Field-Emission Scanning Electron Microscope (FE-SEM) Zeiss Leo 1530Gemini the other one is a Cryo-Field-Emission-Environmental-Scanning Electron Microscope (FEI Quanta 200 FEG) The FE-SEMsare designed for work at low acceleration voltages of less than 2 kVwhich is well suited for ice or gas hydrates in order to minimizesample alteration due to beam damage (Kuhs et al 2004 Sternet al 2004) The samples were kept in a LN2 cooled sample stageinside a vacuum chamber at temperatures of about 188 C anda pressure of about 1 106 bar during measurements

The uncoated specimens were pieces of gas hydrate withdimensions of about 05 05 05 cm The SEMs are equippedwith thin window energy dispersive X-ray analysis (EDX) detectorsallowing for small-scale analysis of low-Z elements Both SEMsfacilitate EDX analyses within distinct areas of interest as drawninto the SEM screen image In standard practice images are taken ata working distance ranging from 6 to 10 mm by secondary electronswith voltages between 17 and 24 kV this is because gas hydratesare easily sublimation-etched even by slightly higher voltages Toobtain good results with the EDX unit 50 kV voltages were appliedthe working distance was raised to about 10ndash12 mm due to theconstruction of the EDX detector in the sample chamber The spotsize for imaging normally is set to 30 for EDX investigations thespot size was raised to 40 or 50

33 Gas chemical analyses

For all hydrate pieces recovered during cruise SO 174 gaschemical analyses were performed onboard by gas chromatog-raphy using a Thermo-Electron gas chromatograph (GC) equippedwith a capillary column (Alltech AT-Q 30 m length 032 mm ID)and a flame ionization detector (FID) as described in Heeschen et al(2007) Hydrate pieces from the gravity corer station GC-6 (sub-samples 1 and 2) from cruise M672 were dissociated onshore andanalyzed with a two-channel GC (Agilent Technologies 6890N)C1ndashC5 hydrocarbons were separated detected and quantified witha capillary column (OPTIMA-5 032 mm ID 50 m length) connectedto a FID In addition permanent gas constituents (O2 N2 CO2) aswell as methane and ethane were determined using a packed(molecular sieve) stainless steel column coupled to a ThermalConductivity Detector While gas analyses of specimen GC-6 1 andsub-sample 1 of specimen 2 were done after X-ray inspectionsspecimen GC-6 2 (sub-sample 2) was not investigated by XRD Inthe following we understand C1ndashC5 hydrocarbons which might behydrate forming as lsquolightrsquo hydrocarbons Abundances of lighthydrocarbons are given as relative portion of the sum of C1ndashC5

hydrocarbons in molThe results of the gas chemical analyses together with salinity

data and water temperature were used to calculate the stabilityboundaries of gas hydrates at the Bush Hill area (block GC185) andat the Chapopote The salinity and temperature data were recorded

on a CTD device by the Remotely Operated Vehicle Quest 4000 m ofthe MARUM Bremen during dive no 81 on the Chapopote onMeteor cruise M672 (Bohrmann and Spiess 2008) The phaseboundaries were calculated using the software HWHYD of theHeriot-Watt-University Edinburgh UK (HWHYD 2005)

4 Results

41 Phase analyses

XRD measurements revealed that all samples were comprised ofcubic gas hydrates ie either crystallographic structure type I or IIAdditionally all samples had fractions of ice as evidenced by theirdistinctive Bragg peaks Structure I gas hydrate is identified by themajor peaks (321) and (320) (Fig 3a) and type II by the major peaks(511) (531) and (440) (Fig 3b) Hexagonal ice (Ih) is present in allsub-samples as confirmed by the three major ice Ih peaks (100)(002) and (101) (Fig 3)

Bush Hill gas hydrates were entirely composed of sII gas hydrate(Table 2) The samples GC-2 as well as TVG-10 were particularlywell preserved with average ice fractions of lt15 and w13respectively (Table 2) Because of the small ice fractions thesesamples allowed for observations of sII gas hydrate without muchinterfering ice Also samples from Green Canyon (TVG-3 and TVG-12) gave no indication on sI gas hydrate in this area instead all gashydrates in those samples is sII

Table 2Gas hydrate structures found at individual sites in the Gulf of Mexico Structures andphase fractions were determined by quantitative phase analyses of X-ray diffractionexperiments (sI sII structure I structure II respectively) Numbers are in weightpercent

Region Sample station sI Ice sII

Bush Hill GC-2 00 77 92300 108 89200 345 65500 46 954

average 00 144 856st-dev 118 118

TVG-10 00 246 75400 120 88000 86 91400 74 926

average 00 132 868st-dev 68 68

GC415 TVG-3 00 620 38000 772 22800 747 253

average 00 713 287st-dev 67 67

TVG-12 00 417 583

Chapopote TVG-6 531 469 00875 125 00

average 703 297 00st-dev 172 172

GC-6 1 669 331 00GC-6 2 (sub-sample 1) 729 271 00average 699 297 00st-dev 30 30

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125120

In turn the samples from the Chapopote contained sI gashydrate (Table 2) The gas hydrate sI fractions range in thesesamples between 73 and 67wt with the balance being ice (Ih)

42 Gas chemical composition of the gas hydrates

The gas chemical compositions allowed for clear distinctionbetween samples from different sites Gas hydrates from the Bush

Table 3Mol percent of C1ndashC5 hydrocarbons measured from purposefully decomposed gas hymeasurements C5-Isomers Sum of unidentified C5-isomers

Region Sample station C1 C2

Bush Hill (GC185) GC-2 7218 663024 064

TVG-10 7169 1309108 043

GC415 TVG-3 7461 1110545 213

TVG-12 8224 887005 023

Chapopote TVG-6 9871 112007 013

GC-6 1 9808 165013 006

GC-6 2 (sub-sample 1) 9762 098005 065

GC-6 2 (sub-sample 2) 9410 261033 011

Hill area yielded a methane portion of about 720 mol of themeasured light hydrocarbons ethane portions of 7ndash130 anda propane fraction of 120 to almost 160 (Table 3 Heeschen et al2007) In addition these gas hydrates had a significant fraction ofbutane isomers with 20ndash45 of iso-butane and 09ndash14 of n-butane

Light hydrocarbons from the Green Canyon hydrates werecomposed of 750ndash820 methane 90ndash110 ethane and 60ndash100propane The samples contained about 12ndash19 of iso-butane and14ndash24 of n-butane

Gas hydrate samples from the Chapopote differed significantlyin the composition of enclathrated light hydrocarbons they con-tained about 987 of methane and 11 of ethane and 01 ofpropane while butane isomers were found in much smalleramounts in the hydrates recovered in sample TVG-6 A similar gascomposition was obtained for specimen 1 from GC-6 Specimen ofGC-6 2 contained two hydrate sub-samples One sub-sample wascomposed similarly to both specimen 1 and the hydrates fromsample TVG-6 (see Table 3) The second sub-sample of hydratespecimen 2 comprised ca 941 methane ca 26 ethane 23propane and ca 10 butane isomers (Table 3) XRD analysis was notdone for this sub-sample of specimen 2

43 Microstructure inspections by Scanning Electron Microscopy

Electron microscopic observations showed clear differences inthe microstructure between hydrate samples comprising sII andmore than 200 of C2ndashC5 compared to methane-dominated sI gashydrates (Fig 4 see Tables 2 and 3)

The surface of sII hydrate from Bush Hill (Fig 4A) displays a solidand dense tier-like array of rectangular units which becomesdistinct at about 500magnification (Fig 4B) Here dense is used todescribe the nonporous microstructure of the hydrate No meso-pores are discernable in samples which contain larger fractions ofnon-methane light hydrocarbon gases (C2ndashC5) and form sII gashydrate At somewhat greater magnification (ca 1000) sI hydratefrom Chapopote (Fig 4C) shows a porous microstructure which canbe clearly discerned as 200ndash500 nm pores at extreme magnification(Fig 4D) Evidence of porosity is not visible in images of 5000magnification of sII hydrate (Fig 4E) At very high magnificationssII hydrate displays a tier-like array of layers (Fig 4F) The tier-like

drate samples (in bold below standard-deviation) n gives back the number of

C3 iC4 nC4 C5-Isomers

1573 450 092 005 n frac14 2069 017 003 002

1185 201 135 002 n frac14 7126 025 007 000

995 188 243 003 n frac14 2393 012 070 003

626 118 144 001 n frac14 2029 006 007 001

012 002 002 000 n frac14 3005 001 001 000

016 003 006 003 n frac14 5006 001 001 001

093 015 026 005 n frac14 2042 007 010 001

225 032 064 007 n frac14 3022 003 004 001

Fig 4 Electron microscopic pictures of gas hydrate samples (A B E F TVG-10 Bush Hill sII hydrate C D GC-6 Chapopote asphalt volcano sI hydrate) A Overview image of a BushHill sII gas hydrate The surface appears dense and solid B Tier-like microstructure is clearly identifiable on a 20 mm scale C At about the same scale mesopores are identifiable inChapopote sI gas hydrates D Magnification of mesopores from Chapopote sI hydrates E Dense surface and tier-like shell are the typical and general microstructure of Bush Hill andGreen Canyon sII gas hydrates F On very small scales tiers in geometrical patterns become evident

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 121

structure mimics a section across individual crystal planes andthereby forms tiers at the surface (Fig 4E F) Geometrical patternsare visible at large magnifications (Fig 4F) ie parallel planes anddistinct angles between them In the porous hydrate samples incontrast geometrical patterns are not identifiable (Fig 4D)

EDX measurements reveal carbon signals in all gas hydratesamples It turns out that the semi-quantitative EDX investigationsshow a constantly higher carbon content in the dense and tier-likestructured material of the Bush Hill and Green Canyon sII hydratesthan in the mesoporous sI hydrates from the Chapopote or in anyother mesoporous hydrates which were investigated in our earlierworks (Kuhs et al 2004 Techmer et al 2005 Bohrmann et al2007) These results were expected as on average the sII gashydrates presented in this study have more carbon atoms perhydrate cage than the sI hydrates However a fully quantitative

evaluation is impossible due to the specimenrsquos uneven surfacewhich deflects the electron beam

SEM investigations also revealed that gas hydrates occur indirect contact with heavy hydrocarbons ie they are soaked withpetroleum Fig 5A shows a SEM image of a sample with more than85 wt gas hydrate content Dense gas hydrate is visible in theright part of the image in the left a shimmering dark-lookingmaterial is discernable The former liquid state of the material issuggested by its fabric as the flowing was stopped by quenching ofthe sample in liquid nitrogen The dark-looking material is char-acterized by a clear carbon peak as evidenced by EDX measure-ment Although semi-quantitative the EDX carbon signals of theoily dark patches in the images were much higher than those of thedense and tier-like material Fig 5B shows a high magnification ofthe contact between a patch of higher hydrocarbon homologues

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

Fig 1 Map of sampling locations in the Gulf of Mexico For station numbers see Table 1Lines indicate the cruise track

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125118

from the Chapopote asphalt volcano (Fig 1) Samples from thiscruise are named GC-6 with sub-samples

The Bush Hill area is located in relatively shallow water depths(500ndash600 m below sea level mbsl) and is fed by gas from the Jollietgas field (Chen and Cathles 2003) from which free gas and oildischarges into the water (Sassen et al 2001b MacDonald et al2003) The oil-stained gas-laden sediments host chemosyntheticcommunities (MacDonald et al 2003) Compositions of lighthydrocarbons (C1ndashC4) should result in the formation of sII gashydrates as suggested by many authors especially for the Bush Hillregion (eg Sassen et al 1994 2001b Chen and Cathles 2003MacDonald et al 2005) Structure II gas hydrate was recentlyidentified by XRD in Green Canyon block GC234 (Rawn et al 2008)Gas hydrates occur both as white precipitates as well as brownish-yellowish colored from oil perfusion through the hydrates (Fig 2see MacDonald et al 1994 2003 Bohrmann and Schenck 2004)

Fig 2 Gas hydrate specimen associated to oil-s

The Chapopote asphalt volcano (in the following only Chapo-pote) is located in deep water (3000 mbsl) northwest of theYucatan peninsula (Fig 1) The Chapopote is part of the CampecheKnolls salt diapirs which are special due to the outflow of naturalasphalt at the seafloor on top of the salt domes (MacDonald et al2004) The salt-tectonic-induced faults allow for fluid migrationfrom subsurface hydrocarbon reservoirs (Ding et al 2008) andseafloor seepage Sufficient gas supply at the Chapopote seep sitesevokes the formation of shallow gas hydrates

3 Methods

Gas hydrate samples were recovered either by TV-guided grabor gravity corer (see Table 1) The Bush Hill samples were coveredonly by a very thin sediment layer the other hydrate samples wereretrieved from shallow sediment depths such that the hydrateswere not directly exposed to seawater Immediately after recoverythe samples were quenched in liquid nitrogen (LN2) to avoid furtherdecomposition of the hydrates Henceforward the hydrates werealways kept in LN2 in a continuous cooling chain when transportedor prepared for measurements in the home lab

Bulk hydrate samples from each location were split into severalsub-samples They were prepared in LN2 in a cold room (10 C to20 C) with air conditioning providing dry air to avoid conden-sation of atmospheric water as ice crystals on the surface of thespecimen The sub-samples were then used for electron micros-copy and XRD For the latter sub-samples from each location weremeasured both with a laboratory diffractometer at the University ofGottingen Germany as well as at high-energy synchrotron beamlines in Hamburg Germany

31 X-ray diffraction

XRD analysis was used for quantitative assessment of the majorphases in the samples For the in-house measurements in Gottin-gen a Bragg-Brentano focusing Philips MRD (Materials ResearchDiffractometer) cryo-goniometer was used which employsa wavelength of 07093 Aring from a molybdenum anode Thesynchrotron measurements were conducted with a wavelength ofw012 Aring at the Beam Line BW 5 as well as at the HARWI II beam lineat the Hamburg Synchrotron Laboratory (HASYLAB) of the DeutscheElektronen Synchrotron (DESY) in Hamburg Germany The pene-tration power was sufficient to study larger samples in aluminumcans of 7 or 12 mm inner diameter which were located in a cryostatwith aluminum vacuumheat shields The short wavelength of thehigh-energy synchrotron radiation allowed for bulk sedimentsample investigations All synchrotron scans were performedexclusively at distinct locations of uncrushed gas hydrate pieces

tained sediments (TVG-10 from Bush Hill)

Fig 3 Rietveld plots of the synchrotron X-ray diffraction data of the hydrate samples(wavelength 01197314 Aring) The solid black lines open circles and solid grey linesrepresent the calculated pattern the observed intensity data and the differencebetween observation and model calculation respectively The vertical solid and dashedlines show possible positions of Bragg reflections of hexagonal ice Ih and hydraterespectively Note different scaling for the x-axis (a) The sample contains mainly sIhydrate (w88 the remainder being ice Ih) collected from the Chapopote (SO 174station 140 TVG-6) (b) Sample contains about 95 sII hydrate collected from the BushHill (SO 174 station 47 GC-2)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 119

comprising crystals of ca 200ndash300 mm (Klapp et al 2007) and noton powdered samples Accordingly structural information wasobtained for the scanned volumes containing coarse crystals andnot for the whole powder sample In contrast the cryo-MRDmachine gave averages on the complete powder sample Details ofthe XRD measurements are given in Bohrmann et al (2007) andKlapp et al (2007) Phase fractions of the samples are then calcu-lated from Rietveld refinement using the GSAS software (Larsonand van Dreele 1994) and the EXPGUI user interface (Toby 2001)

32 FE-SEM investigations

Gas hydrate pieces which were previously characterized byXRD were analyzed by SEM For electron microscopic investiga-tions two electron microscopes were applied one is a Cryo-Field-Emission Scanning Electron Microscope (FE-SEM) Zeiss Leo 1530Gemini the other one is a Cryo-Field-Emission-Environmental-Scanning Electron Microscope (FEI Quanta 200 FEG) The FE-SEMsare designed for work at low acceleration voltages of less than 2 kVwhich is well suited for ice or gas hydrates in order to minimizesample alteration due to beam damage (Kuhs et al 2004 Sternet al 2004) The samples were kept in a LN2 cooled sample stageinside a vacuum chamber at temperatures of about 188 C anda pressure of about 1 106 bar during measurements

The uncoated specimens were pieces of gas hydrate withdimensions of about 05 05 05 cm The SEMs are equippedwith thin window energy dispersive X-ray analysis (EDX) detectorsallowing for small-scale analysis of low-Z elements Both SEMsfacilitate EDX analyses within distinct areas of interest as drawninto the SEM screen image In standard practice images are taken ata working distance ranging from 6 to 10 mm by secondary electronswith voltages between 17 and 24 kV this is because gas hydratesare easily sublimation-etched even by slightly higher voltages Toobtain good results with the EDX unit 50 kV voltages were appliedthe working distance was raised to about 10ndash12 mm due to theconstruction of the EDX detector in the sample chamber The spotsize for imaging normally is set to 30 for EDX investigations thespot size was raised to 40 or 50

33 Gas chemical analyses

For all hydrate pieces recovered during cruise SO 174 gaschemical analyses were performed onboard by gas chromatog-raphy using a Thermo-Electron gas chromatograph (GC) equippedwith a capillary column (Alltech AT-Q 30 m length 032 mm ID)and a flame ionization detector (FID) as described in Heeschen et al(2007) Hydrate pieces from the gravity corer station GC-6 (sub-samples 1 and 2) from cruise M672 were dissociated onshore andanalyzed with a two-channel GC (Agilent Technologies 6890N)C1ndashC5 hydrocarbons were separated detected and quantified witha capillary column (OPTIMA-5 032 mm ID 50 m length) connectedto a FID In addition permanent gas constituents (O2 N2 CO2) aswell as methane and ethane were determined using a packed(molecular sieve) stainless steel column coupled to a ThermalConductivity Detector While gas analyses of specimen GC-6 1 andsub-sample 1 of specimen 2 were done after X-ray inspectionsspecimen GC-6 2 (sub-sample 2) was not investigated by XRD Inthe following we understand C1ndashC5 hydrocarbons which might behydrate forming as lsquolightrsquo hydrocarbons Abundances of lighthydrocarbons are given as relative portion of the sum of C1ndashC5

hydrocarbons in molThe results of the gas chemical analyses together with salinity

data and water temperature were used to calculate the stabilityboundaries of gas hydrates at the Bush Hill area (block GC185) andat the Chapopote The salinity and temperature data were recorded

on a CTD device by the Remotely Operated Vehicle Quest 4000 m ofthe MARUM Bremen during dive no 81 on the Chapopote onMeteor cruise M672 (Bohrmann and Spiess 2008) The phaseboundaries were calculated using the software HWHYD of theHeriot-Watt-University Edinburgh UK (HWHYD 2005)

4 Results

41 Phase analyses

XRD measurements revealed that all samples were comprised ofcubic gas hydrates ie either crystallographic structure type I or IIAdditionally all samples had fractions of ice as evidenced by theirdistinctive Bragg peaks Structure I gas hydrate is identified by themajor peaks (321) and (320) (Fig 3a) and type II by the major peaks(511) (531) and (440) (Fig 3b) Hexagonal ice (Ih) is present in allsub-samples as confirmed by the three major ice Ih peaks (100)(002) and (101) (Fig 3)

Bush Hill gas hydrates were entirely composed of sII gas hydrate(Table 2) The samples GC-2 as well as TVG-10 were particularlywell preserved with average ice fractions of lt15 and w13respectively (Table 2) Because of the small ice fractions thesesamples allowed for observations of sII gas hydrate without muchinterfering ice Also samples from Green Canyon (TVG-3 and TVG-12) gave no indication on sI gas hydrate in this area instead all gashydrates in those samples is sII

Table 2Gas hydrate structures found at individual sites in the Gulf of Mexico Structures andphase fractions were determined by quantitative phase analyses of X-ray diffractionexperiments (sI sII structure I structure II respectively) Numbers are in weightpercent

Region Sample station sI Ice sII

Bush Hill GC-2 00 77 92300 108 89200 345 65500 46 954

average 00 144 856st-dev 118 118

TVG-10 00 246 75400 120 88000 86 91400 74 926

average 00 132 868st-dev 68 68

GC415 TVG-3 00 620 38000 772 22800 747 253

average 00 713 287st-dev 67 67

TVG-12 00 417 583

Chapopote TVG-6 531 469 00875 125 00

average 703 297 00st-dev 172 172

GC-6 1 669 331 00GC-6 2 (sub-sample 1) 729 271 00average 699 297 00st-dev 30 30

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125120

In turn the samples from the Chapopote contained sI gashydrate (Table 2) The gas hydrate sI fractions range in thesesamples between 73 and 67wt with the balance being ice (Ih)

42 Gas chemical composition of the gas hydrates

The gas chemical compositions allowed for clear distinctionbetween samples from different sites Gas hydrates from the Bush

Table 3Mol percent of C1ndashC5 hydrocarbons measured from purposefully decomposed gas hymeasurements C5-Isomers Sum of unidentified C5-isomers

Region Sample station C1 C2

Bush Hill (GC185) GC-2 7218 663024 064

TVG-10 7169 1309108 043

GC415 TVG-3 7461 1110545 213

TVG-12 8224 887005 023

Chapopote TVG-6 9871 112007 013

GC-6 1 9808 165013 006

GC-6 2 (sub-sample 1) 9762 098005 065

GC-6 2 (sub-sample 2) 9410 261033 011

Hill area yielded a methane portion of about 720 mol of themeasured light hydrocarbons ethane portions of 7ndash130 anda propane fraction of 120 to almost 160 (Table 3 Heeschen et al2007) In addition these gas hydrates had a significant fraction ofbutane isomers with 20ndash45 of iso-butane and 09ndash14 of n-butane

Light hydrocarbons from the Green Canyon hydrates werecomposed of 750ndash820 methane 90ndash110 ethane and 60ndash100propane The samples contained about 12ndash19 of iso-butane and14ndash24 of n-butane

Gas hydrate samples from the Chapopote differed significantlyin the composition of enclathrated light hydrocarbons they con-tained about 987 of methane and 11 of ethane and 01 ofpropane while butane isomers were found in much smalleramounts in the hydrates recovered in sample TVG-6 A similar gascomposition was obtained for specimen 1 from GC-6 Specimen ofGC-6 2 contained two hydrate sub-samples One sub-sample wascomposed similarly to both specimen 1 and the hydrates fromsample TVG-6 (see Table 3) The second sub-sample of hydratespecimen 2 comprised ca 941 methane ca 26 ethane 23propane and ca 10 butane isomers (Table 3) XRD analysis was notdone for this sub-sample of specimen 2

43 Microstructure inspections by Scanning Electron Microscopy

Electron microscopic observations showed clear differences inthe microstructure between hydrate samples comprising sII andmore than 200 of C2ndashC5 compared to methane-dominated sI gashydrates (Fig 4 see Tables 2 and 3)

The surface of sII hydrate from Bush Hill (Fig 4A) displays a solidand dense tier-like array of rectangular units which becomesdistinct at about 500magnification (Fig 4B) Here dense is used todescribe the nonporous microstructure of the hydrate No meso-pores are discernable in samples which contain larger fractions ofnon-methane light hydrocarbon gases (C2ndashC5) and form sII gashydrate At somewhat greater magnification (ca 1000) sI hydratefrom Chapopote (Fig 4C) shows a porous microstructure which canbe clearly discerned as 200ndash500 nm pores at extreme magnification(Fig 4D) Evidence of porosity is not visible in images of 5000magnification of sII hydrate (Fig 4E) At very high magnificationssII hydrate displays a tier-like array of layers (Fig 4F) The tier-like

drate samples (in bold below standard-deviation) n gives back the number of

C3 iC4 nC4 C5-Isomers

1573 450 092 005 n frac14 2069 017 003 002

1185 201 135 002 n frac14 7126 025 007 000

995 188 243 003 n frac14 2393 012 070 003

626 118 144 001 n frac14 2029 006 007 001

012 002 002 000 n frac14 3005 001 001 000

016 003 006 003 n frac14 5006 001 001 001

093 015 026 005 n frac14 2042 007 010 001

225 032 064 007 n frac14 3022 003 004 001

Fig 4 Electron microscopic pictures of gas hydrate samples (A B E F TVG-10 Bush Hill sII hydrate C D GC-6 Chapopote asphalt volcano sI hydrate) A Overview image of a BushHill sII gas hydrate The surface appears dense and solid B Tier-like microstructure is clearly identifiable on a 20 mm scale C At about the same scale mesopores are identifiable inChapopote sI gas hydrates D Magnification of mesopores from Chapopote sI hydrates E Dense surface and tier-like shell are the typical and general microstructure of Bush Hill andGreen Canyon sII gas hydrates F On very small scales tiers in geometrical patterns become evident

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 121

structure mimics a section across individual crystal planes andthereby forms tiers at the surface (Fig 4E F) Geometrical patternsare visible at large magnifications (Fig 4F) ie parallel planes anddistinct angles between them In the porous hydrate samples incontrast geometrical patterns are not identifiable (Fig 4D)

EDX measurements reveal carbon signals in all gas hydratesamples It turns out that the semi-quantitative EDX investigationsshow a constantly higher carbon content in the dense and tier-likestructured material of the Bush Hill and Green Canyon sII hydratesthan in the mesoporous sI hydrates from the Chapopote or in anyother mesoporous hydrates which were investigated in our earlierworks (Kuhs et al 2004 Techmer et al 2005 Bohrmann et al2007) These results were expected as on average the sII gashydrates presented in this study have more carbon atoms perhydrate cage than the sI hydrates However a fully quantitative

evaluation is impossible due to the specimenrsquos uneven surfacewhich deflects the electron beam

SEM investigations also revealed that gas hydrates occur indirect contact with heavy hydrocarbons ie they are soaked withpetroleum Fig 5A shows a SEM image of a sample with more than85 wt gas hydrate content Dense gas hydrate is visible in theright part of the image in the left a shimmering dark-lookingmaterial is discernable The former liquid state of the material issuggested by its fabric as the flowing was stopped by quenching ofthe sample in liquid nitrogen The dark-looking material is char-acterized by a clear carbon peak as evidenced by EDX measure-ment Although semi-quantitative the EDX carbon signals of theoily dark patches in the images were much higher than those of thedense and tier-like material Fig 5B shows a high magnification ofthe contact between a patch of higher hydrocarbon homologues

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

Fig 3 Rietveld plots of the synchrotron X-ray diffraction data of the hydrate samples(wavelength 01197314 Aring) The solid black lines open circles and solid grey linesrepresent the calculated pattern the observed intensity data and the differencebetween observation and model calculation respectively The vertical solid and dashedlines show possible positions of Bragg reflections of hexagonal ice Ih and hydraterespectively Note different scaling for the x-axis (a) The sample contains mainly sIhydrate (w88 the remainder being ice Ih) collected from the Chapopote (SO 174station 140 TVG-6) (b) Sample contains about 95 sII hydrate collected from the BushHill (SO 174 station 47 GC-2)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 119

comprising crystals of ca 200ndash300 mm (Klapp et al 2007) and noton powdered samples Accordingly structural information wasobtained for the scanned volumes containing coarse crystals andnot for the whole powder sample In contrast the cryo-MRDmachine gave averages on the complete powder sample Details ofthe XRD measurements are given in Bohrmann et al (2007) andKlapp et al (2007) Phase fractions of the samples are then calcu-lated from Rietveld refinement using the GSAS software (Larsonand van Dreele 1994) and the EXPGUI user interface (Toby 2001)

32 FE-SEM investigations

Gas hydrate pieces which were previously characterized byXRD were analyzed by SEM For electron microscopic investiga-tions two electron microscopes were applied one is a Cryo-Field-Emission Scanning Electron Microscope (FE-SEM) Zeiss Leo 1530Gemini the other one is a Cryo-Field-Emission-Environmental-Scanning Electron Microscope (FEI Quanta 200 FEG) The FE-SEMsare designed for work at low acceleration voltages of less than 2 kVwhich is well suited for ice or gas hydrates in order to minimizesample alteration due to beam damage (Kuhs et al 2004 Sternet al 2004) The samples were kept in a LN2 cooled sample stageinside a vacuum chamber at temperatures of about 188 C anda pressure of about 1 106 bar during measurements

The uncoated specimens were pieces of gas hydrate withdimensions of about 05 05 05 cm The SEMs are equippedwith thin window energy dispersive X-ray analysis (EDX) detectorsallowing for small-scale analysis of low-Z elements Both SEMsfacilitate EDX analyses within distinct areas of interest as drawninto the SEM screen image In standard practice images are taken ata working distance ranging from 6 to 10 mm by secondary electronswith voltages between 17 and 24 kV this is because gas hydratesare easily sublimation-etched even by slightly higher voltages Toobtain good results with the EDX unit 50 kV voltages were appliedthe working distance was raised to about 10ndash12 mm due to theconstruction of the EDX detector in the sample chamber The spotsize for imaging normally is set to 30 for EDX investigations thespot size was raised to 40 or 50

33 Gas chemical analyses

For all hydrate pieces recovered during cruise SO 174 gaschemical analyses were performed onboard by gas chromatog-raphy using a Thermo-Electron gas chromatograph (GC) equippedwith a capillary column (Alltech AT-Q 30 m length 032 mm ID)and a flame ionization detector (FID) as described in Heeschen et al(2007) Hydrate pieces from the gravity corer station GC-6 (sub-samples 1 and 2) from cruise M672 were dissociated onshore andanalyzed with a two-channel GC (Agilent Technologies 6890N)C1ndashC5 hydrocarbons were separated detected and quantified witha capillary column (OPTIMA-5 032 mm ID 50 m length) connectedto a FID In addition permanent gas constituents (O2 N2 CO2) aswell as methane and ethane were determined using a packed(molecular sieve) stainless steel column coupled to a ThermalConductivity Detector While gas analyses of specimen GC-6 1 andsub-sample 1 of specimen 2 were done after X-ray inspectionsspecimen GC-6 2 (sub-sample 2) was not investigated by XRD Inthe following we understand C1ndashC5 hydrocarbons which might behydrate forming as lsquolightrsquo hydrocarbons Abundances of lighthydrocarbons are given as relative portion of the sum of C1ndashC5

hydrocarbons in molThe results of the gas chemical analyses together with salinity

data and water temperature were used to calculate the stabilityboundaries of gas hydrates at the Bush Hill area (block GC185) andat the Chapopote The salinity and temperature data were recorded

on a CTD device by the Remotely Operated Vehicle Quest 4000 m ofthe MARUM Bremen during dive no 81 on the Chapopote onMeteor cruise M672 (Bohrmann and Spiess 2008) The phaseboundaries were calculated using the software HWHYD of theHeriot-Watt-University Edinburgh UK (HWHYD 2005)

4 Results

41 Phase analyses

XRD measurements revealed that all samples were comprised ofcubic gas hydrates ie either crystallographic structure type I or IIAdditionally all samples had fractions of ice as evidenced by theirdistinctive Bragg peaks Structure I gas hydrate is identified by themajor peaks (321) and (320) (Fig 3a) and type II by the major peaks(511) (531) and (440) (Fig 3b) Hexagonal ice (Ih) is present in allsub-samples as confirmed by the three major ice Ih peaks (100)(002) and (101) (Fig 3)

Bush Hill gas hydrates were entirely composed of sII gas hydrate(Table 2) The samples GC-2 as well as TVG-10 were particularlywell preserved with average ice fractions of lt15 and w13respectively (Table 2) Because of the small ice fractions thesesamples allowed for observations of sII gas hydrate without muchinterfering ice Also samples from Green Canyon (TVG-3 and TVG-12) gave no indication on sI gas hydrate in this area instead all gashydrates in those samples is sII

Table 2Gas hydrate structures found at individual sites in the Gulf of Mexico Structures andphase fractions were determined by quantitative phase analyses of X-ray diffractionexperiments (sI sII structure I structure II respectively) Numbers are in weightpercent

Region Sample station sI Ice sII

Bush Hill GC-2 00 77 92300 108 89200 345 65500 46 954

average 00 144 856st-dev 118 118

TVG-10 00 246 75400 120 88000 86 91400 74 926

average 00 132 868st-dev 68 68

GC415 TVG-3 00 620 38000 772 22800 747 253

average 00 713 287st-dev 67 67

TVG-12 00 417 583

Chapopote TVG-6 531 469 00875 125 00

average 703 297 00st-dev 172 172

GC-6 1 669 331 00GC-6 2 (sub-sample 1) 729 271 00average 699 297 00st-dev 30 30

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125120

In turn the samples from the Chapopote contained sI gashydrate (Table 2) The gas hydrate sI fractions range in thesesamples between 73 and 67wt with the balance being ice (Ih)

42 Gas chemical composition of the gas hydrates

The gas chemical compositions allowed for clear distinctionbetween samples from different sites Gas hydrates from the Bush

Table 3Mol percent of C1ndashC5 hydrocarbons measured from purposefully decomposed gas hymeasurements C5-Isomers Sum of unidentified C5-isomers

Region Sample station C1 C2

Bush Hill (GC185) GC-2 7218 663024 064

TVG-10 7169 1309108 043

GC415 TVG-3 7461 1110545 213

TVG-12 8224 887005 023

Chapopote TVG-6 9871 112007 013

GC-6 1 9808 165013 006

GC-6 2 (sub-sample 1) 9762 098005 065

GC-6 2 (sub-sample 2) 9410 261033 011

Hill area yielded a methane portion of about 720 mol of themeasured light hydrocarbons ethane portions of 7ndash130 anda propane fraction of 120 to almost 160 (Table 3 Heeschen et al2007) In addition these gas hydrates had a significant fraction ofbutane isomers with 20ndash45 of iso-butane and 09ndash14 of n-butane

Light hydrocarbons from the Green Canyon hydrates werecomposed of 750ndash820 methane 90ndash110 ethane and 60ndash100propane The samples contained about 12ndash19 of iso-butane and14ndash24 of n-butane

Gas hydrate samples from the Chapopote differed significantlyin the composition of enclathrated light hydrocarbons they con-tained about 987 of methane and 11 of ethane and 01 ofpropane while butane isomers were found in much smalleramounts in the hydrates recovered in sample TVG-6 A similar gascomposition was obtained for specimen 1 from GC-6 Specimen ofGC-6 2 contained two hydrate sub-samples One sub-sample wascomposed similarly to both specimen 1 and the hydrates fromsample TVG-6 (see Table 3) The second sub-sample of hydratespecimen 2 comprised ca 941 methane ca 26 ethane 23propane and ca 10 butane isomers (Table 3) XRD analysis was notdone for this sub-sample of specimen 2

43 Microstructure inspections by Scanning Electron Microscopy

Electron microscopic observations showed clear differences inthe microstructure between hydrate samples comprising sII andmore than 200 of C2ndashC5 compared to methane-dominated sI gashydrates (Fig 4 see Tables 2 and 3)

The surface of sII hydrate from Bush Hill (Fig 4A) displays a solidand dense tier-like array of rectangular units which becomesdistinct at about 500magnification (Fig 4B) Here dense is used todescribe the nonporous microstructure of the hydrate No meso-pores are discernable in samples which contain larger fractions ofnon-methane light hydrocarbon gases (C2ndashC5) and form sII gashydrate At somewhat greater magnification (ca 1000) sI hydratefrom Chapopote (Fig 4C) shows a porous microstructure which canbe clearly discerned as 200ndash500 nm pores at extreme magnification(Fig 4D) Evidence of porosity is not visible in images of 5000magnification of sII hydrate (Fig 4E) At very high magnificationssII hydrate displays a tier-like array of layers (Fig 4F) The tier-like

drate samples (in bold below standard-deviation) n gives back the number of

C3 iC4 nC4 C5-Isomers

1573 450 092 005 n frac14 2069 017 003 002

1185 201 135 002 n frac14 7126 025 007 000

995 188 243 003 n frac14 2393 012 070 003

626 118 144 001 n frac14 2029 006 007 001

012 002 002 000 n frac14 3005 001 001 000

016 003 006 003 n frac14 5006 001 001 001

093 015 026 005 n frac14 2042 007 010 001

225 032 064 007 n frac14 3022 003 004 001

Fig 4 Electron microscopic pictures of gas hydrate samples (A B E F TVG-10 Bush Hill sII hydrate C D GC-6 Chapopote asphalt volcano sI hydrate) A Overview image of a BushHill sII gas hydrate The surface appears dense and solid B Tier-like microstructure is clearly identifiable on a 20 mm scale C At about the same scale mesopores are identifiable inChapopote sI gas hydrates D Magnification of mesopores from Chapopote sI hydrates E Dense surface and tier-like shell are the typical and general microstructure of Bush Hill andGreen Canyon sII gas hydrates F On very small scales tiers in geometrical patterns become evident

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 121

structure mimics a section across individual crystal planes andthereby forms tiers at the surface (Fig 4E F) Geometrical patternsare visible at large magnifications (Fig 4F) ie parallel planes anddistinct angles between them In the porous hydrate samples incontrast geometrical patterns are not identifiable (Fig 4D)

EDX measurements reveal carbon signals in all gas hydratesamples It turns out that the semi-quantitative EDX investigationsshow a constantly higher carbon content in the dense and tier-likestructured material of the Bush Hill and Green Canyon sII hydratesthan in the mesoporous sI hydrates from the Chapopote or in anyother mesoporous hydrates which were investigated in our earlierworks (Kuhs et al 2004 Techmer et al 2005 Bohrmann et al2007) These results were expected as on average the sII gashydrates presented in this study have more carbon atoms perhydrate cage than the sI hydrates However a fully quantitative

evaluation is impossible due to the specimenrsquos uneven surfacewhich deflects the electron beam

SEM investigations also revealed that gas hydrates occur indirect contact with heavy hydrocarbons ie they are soaked withpetroleum Fig 5A shows a SEM image of a sample with more than85 wt gas hydrate content Dense gas hydrate is visible in theright part of the image in the left a shimmering dark-lookingmaterial is discernable The former liquid state of the material issuggested by its fabric as the flowing was stopped by quenching ofthe sample in liquid nitrogen The dark-looking material is char-acterized by a clear carbon peak as evidenced by EDX measure-ment Although semi-quantitative the EDX carbon signals of theoily dark patches in the images were much higher than those of thedense and tier-like material Fig 5B shows a high magnification ofthe contact between a patch of higher hydrocarbon homologues

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

Table 2Gas hydrate structures found at individual sites in the Gulf of Mexico Structures andphase fractions were determined by quantitative phase analyses of X-ray diffractionexperiments (sI sII structure I structure II respectively) Numbers are in weightpercent

Region Sample station sI Ice sII

Bush Hill GC-2 00 77 92300 108 89200 345 65500 46 954

average 00 144 856st-dev 118 118

TVG-10 00 246 75400 120 88000 86 91400 74 926

average 00 132 868st-dev 68 68

GC415 TVG-3 00 620 38000 772 22800 747 253

average 00 713 287st-dev 67 67

TVG-12 00 417 583

Chapopote TVG-6 531 469 00875 125 00

average 703 297 00st-dev 172 172

GC-6 1 669 331 00GC-6 2 (sub-sample 1) 729 271 00average 699 297 00st-dev 30 30

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125120

In turn the samples from the Chapopote contained sI gashydrate (Table 2) The gas hydrate sI fractions range in thesesamples between 73 and 67wt with the balance being ice (Ih)

42 Gas chemical composition of the gas hydrates

The gas chemical compositions allowed for clear distinctionbetween samples from different sites Gas hydrates from the Bush

Table 3Mol percent of C1ndashC5 hydrocarbons measured from purposefully decomposed gas hymeasurements C5-Isomers Sum of unidentified C5-isomers

Region Sample station C1 C2

Bush Hill (GC185) GC-2 7218 663024 064

TVG-10 7169 1309108 043

GC415 TVG-3 7461 1110545 213

TVG-12 8224 887005 023

Chapopote TVG-6 9871 112007 013

GC-6 1 9808 165013 006

GC-6 2 (sub-sample 1) 9762 098005 065

GC-6 2 (sub-sample 2) 9410 261033 011

Hill area yielded a methane portion of about 720 mol of themeasured light hydrocarbons ethane portions of 7ndash130 anda propane fraction of 120 to almost 160 (Table 3 Heeschen et al2007) In addition these gas hydrates had a significant fraction ofbutane isomers with 20ndash45 of iso-butane and 09ndash14 of n-butane

Light hydrocarbons from the Green Canyon hydrates werecomposed of 750ndash820 methane 90ndash110 ethane and 60ndash100propane The samples contained about 12ndash19 of iso-butane and14ndash24 of n-butane

Gas hydrate samples from the Chapopote differed significantlyin the composition of enclathrated light hydrocarbons they con-tained about 987 of methane and 11 of ethane and 01 ofpropane while butane isomers were found in much smalleramounts in the hydrates recovered in sample TVG-6 A similar gascomposition was obtained for specimen 1 from GC-6 Specimen ofGC-6 2 contained two hydrate sub-samples One sub-sample wascomposed similarly to both specimen 1 and the hydrates fromsample TVG-6 (see Table 3) The second sub-sample of hydratespecimen 2 comprised ca 941 methane ca 26 ethane 23propane and ca 10 butane isomers (Table 3) XRD analysis was notdone for this sub-sample of specimen 2

43 Microstructure inspections by Scanning Electron Microscopy

Electron microscopic observations showed clear differences inthe microstructure between hydrate samples comprising sII andmore than 200 of C2ndashC5 compared to methane-dominated sI gashydrates (Fig 4 see Tables 2 and 3)

The surface of sII hydrate from Bush Hill (Fig 4A) displays a solidand dense tier-like array of rectangular units which becomesdistinct at about 500magnification (Fig 4B) Here dense is used todescribe the nonporous microstructure of the hydrate No meso-pores are discernable in samples which contain larger fractions ofnon-methane light hydrocarbon gases (C2ndashC5) and form sII gashydrate At somewhat greater magnification (ca 1000) sI hydratefrom Chapopote (Fig 4C) shows a porous microstructure which canbe clearly discerned as 200ndash500 nm pores at extreme magnification(Fig 4D) Evidence of porosity is not visible in images of 5000magnification of sII hydrate (Fig 4E) At very high magnificationssII hydrate displays a tier-like array of layers (Fig 4F) The tier-like

drate samples (in bold below standard-deviation) n gives back the number of

C3 iC4 nC4 C5-Isomers

1573 450 092 005 n frac14 2069 017 003 002

1185 201 135 002 n frac14 7126 025 007 000

995 188 243 003 n frac14 2393 012 070 003

626 118 144 001 n frac14 2029 006 007 001

012 002 002 000 n frac14 3005 001 001 000

016 003 006 003 n frac14 5006 001 001 001

093 015 026 005 n frac14 2042 007 010 001

225 032 064 007 n frac14 3022 003 004 001

Fig 4 Electron microscopic pictures of gas hydrate samples (A B E F TVG-10 Bush Hill sII hydrate C D GC-6 Chapopote asphalt volcano sI hydrate) A Overview image of a BushHill sII gas hydrate The surface appears dense and solid B Tier-like microstructure is clearly identifiable on a 20 mm scale C At about the same scale mesopores are identifiable inChapopote sI gas hydrates D Magnification of mesopores from Chapopote sI hydrates E Dense surface and tier-like shell are the typical and general microstructure of Bush Hill andGreen Canyon sII gas hydrates F On very small scales tiers in geometrical patterns become evident

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 121

structure mimics a section across individual crystal planes andthereby forms tiers at the surface (Fig 4E F) Geometrical patternsare visible at large magnifications (Fig 4F) ie parallel planes anddistinct angles between them In the porous hydrate samples incontrast geometrical patterns are not identifiable (Fig 4D)

EDX measurements reveal carbon signals in all gas hydratesamples It turns out that the semi-quantitative EDX investigationsshow a constantly higher carbon content in the dense and tier-likestructured material of the Bush Hill and Green Canyon sII hydratesthan in the mesoporous sI hydrates from the Chapopote or in anyother mesoporous hydrates which were investigated in our earlierworks (Kuhs et al 2004 Techmer et al 2005 Bohrmann et al2007) These results were expected as on average the sII gashydrates presented in this study have more carbon atoms perhydrate cage than the sI hydrates However a fully quantitative

evaluation is impossible due to the specimenrsquos uneven surfacewhich deflects the electron beam

SEM investigations also revealed that gas hydrates occur indirect contact with heavy hydrocarbons ie they are soaked withpetroleum Fig 5A shows a SEM image of a sample with more than85 wt gas hydrate content Dense gas hydrate is visible in theright part of the image in the left a shimmering dark-lookingmaterial is discernable The former liquid state of the material issuggested by its fabric as the flowing was stopped by quenching ofthe sample in liquid nitrogen The dark-looking material is char-acterized by a clear carbon peak as evidenced by EDX measure-ment Although semi-quantitative the EDX carbon signals of theoily dark patches in the images were much higher than those of thedense and tier-like material Fig 5B shows a high magnification ofthe contact between a patch of higher hydrocarbon homologues

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

Fig 4 Electron microscopic pictures of gas hydrate samples (A B E F TVG-10 Bush Hill sII hydrate C D GC-6 Chapopote asphalt volcano sI hydrate) A Overview image of a BushHill sII gas hydrate The surface appears dense and solid B Tier-like microstructure is clearly identifiable on a 20 mm scale C At about the same scale mesopores are identifiable inChapopote sI gas hydrates D Magnification of mesopores from Chapopote sI hydrates E Dense surface and tier-like shell are the typical and general microstructure of Bush Hill andGreen Canyon sII gas hydrates F On very small scales tiers in geometrical patterns become evident

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 121

structure mimics a section across individual crystal planes andthereby forms tiers at the surface (Fig 4E F) Geometrical patternsare visible at large magnifications (Fig 4F) ie parallel planes anddistinct angles between them In the porous hydrate samples incontrast geometrical patterns are not identifiable (Fig 4D)

EDX measurements reveal carbon signals in all gas hydratesamples It turns out that the semi-quantitative EDX investigationsshow a constantly higher carbon content in the dense and tier-likestructured material of the Bush Hill and Green Canyon sII hydratesthan in the mesoporous sI hydrates from the Chapopote or in anyother mesoporous hydrates which were investigated in our earlierworks (Kuhs et al 2004 Techmer et al 2005 Bohrmann et al2007) These results were expected as on average the sII gashydrates presented in this study have more carbon atoms perhydrate cage than the sI hydrates However a fully quantitative

evaluation is impossible due to the specimenrsquos uneven surfacewhich deflects the electron beam

SEM investigations also revealed that gas hydrates occur indirect contact with heavy hydrocarbons ie they are soaked withpetroleum Fig 5A shows a SEM image of a sample with more than85 wt gas hydrate content Dense gas hydrate is visible in theright part of the image in the left a shimmering dark-lookingmaterial is discernable The former liquid state of the material issuggested by its fabric as the flowing was stopped by quenching ofthe sample in liquid nitrogen The dark-looking material is char-acterized by a clear carbon peak as evidenced by EDX measure-ment Although semi-quantitative the EDX carbon signals of theoily dark patches in the images were much higher than those of thedense and tier-like material Fig 5B shows a high magnification ofthe contact between a patch of higher hydrocarbon homologues

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

Fig 5 Electron microscopic pictures of gas hydrate samples A Oil patches are adjacent to dense gas hydrate (DGH) The flowing of the oil prior to deep freezing is discernable Inthe image the oil is flowing downwards see the highly viscous flow in the central lower part of the image (sample TVG-10) B Detail of the border between oil (top left) and gashydrate (below) (sample TVG-3)

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125122

and dense material Fig 2 shows photographs of sediment sampleson a centimeter scale illustrating that hydrate-bearing sedimentsare soaked with petroleum and discolor the hydrate in parts

Fig 6 Stability models for gas hydrate occurrences in the Gulf of Mexico Bush Hillstructure II hydrate (TVG-10) Chapopote structure I hydrate (TVG-6) The phaseboundaries were calculated using the Heriot-Watt-Hydrate software (HWHYD 2005)The temperature was measured during ROV-dive 81 on M672

5 Discussion

51 Hydrocarbon composition and crystal structure

Gas chemical analyses revealed high abundances of lighthydrocarbons and in particular of methane in all samples (Table 3Heeschen et al 2007) The fractions of methane though differsignificantly between the samples from the Bush Hill (GC185) andGreen Canyon (GC415) area where roughly a fourth to a fifth of themeasured hydrocarbons consist of C2 to C5 compounds and Cha-popote asphalt volcano hydrates which host mostly more than 98methane in hydrocarbons The hydrocarbon compositions suggestgas hydrate sII formation in the Bush Hill and Green Canyon areasbecause the larger molecules would not fit into the smaller cages ofthe structure I hydrate (Sloan and Koh 2007) This agrees well withour results of XRD measurements and other published gas chemicaldata for the Bush Hill and Green Canyon Sassen et al (1994 2001b)and Chen and Cathles (2003) reported on selective stripping ofhydrocarbons from the thermogenic vent gas which crystallize tosII gas hydrate They found hydrate-bound light hydrocarbons inthe same range of composition as reported here

For Chapopote both XRD investigations and compositions oflight hydrocarbons reveal sI hydrate formation with almost exclu-sively methane as the hydrate-forming gas The exception is onehydrate piece which contained 941 methane and 59 C2ndashC5 ofhydrate-forming hydrocarbons This suggests that sII hydrate maysometimes occur in this area

The dominance of strongly methane-rich sI forming hydrateswas surprising because at Chapopote seeping fluids compriseheavy oil asphalt and light hydrocarbons (Bohrmann and Spi-ess 2008 MacDonald et al 2004) and in the sediments theC1C2ndashC5-ratio varied considerably with depth (Bohrmann andSchenck 2004) This melange of hydrocarbons would suggest thatalso light hydrocarbons other than methane were incorporated intothe hydrates Yet we found only one hydrate piece in which thisseems to be the case On the other hand this piece was next toother hydrate pieces which contained more methane and crystal-lized to sI hydrate This observation points to an inhomogeneousdistribution of gases in the gas hydrates at Chapopote To furtherexplore the origin and distribution of Chapopote hydrate gasesmore information on the hydrocarbon isotopic compositions and

extend of biodegradation is needed For this work it is important tobe aware of the hydrocarbon composition in the actual sampleswhich were investigated also by XRD and SEM

XRD analyses confirm the interpretation of the gas analyses ofthe individual hydrate pieces To the best of our knowledge theresults presented here are the first direct crystallographic structureidentifications of gas hydrates from these locations in the GOM byXRD together with a preliminary report by Rawn et al (2008) whofound sII hydrates by XRD in block GC234 In the many years of gashydrate research in the GOM often the hydrate structures wereinferred from gas compositions By interpreting compositions ofhydrate-bound hydrocarbons the most probable crystallographichydrate structure could be predicted on thermodynamic groundsbut an experimental confirmation is still indicated due to remain-ing uncertainties in particular close to stability limits and possiblemetastable formations

52 Hydrate stability

The onset of gas hydrate stability differs significantly at the BushHill area and Chapopote (Fig 6 calculation see Section 33) At the

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 123

sampling site in the Bush Hill area in about 550 mbsl sII hydratesare stable until about 310 mbsl where the water is about 121 C AtChapopote which is in about 2900 mbsl hydrates would be stableup to a water depth of 620 m where the water has a temperature of74 C (Fig 6) so Bush Hill sII hydrate would require an increase ofabout 10ndash11 C above ambient and the Chapopote sI hydrates anincrease of 14ndash15 C above ambient to destabilize The watertemperature at Bush Hill is barely within the sI stability limit whichmay explain why no sI hydrate is observed there Moreover Mac-Donald et al (2005) show that temperature variations at Bush Hillwould regularly exceed the stability of sI hydrate Little is knownabout temperature variations at Chapopote but they are unlikely tochallenge the stability of sI hydrate and even less likely to desta-bilize sII hydrate Structure II gas hydrate with the given gascomposition is clearly the more stable hydrate structure the lack ofsII hydrate at Chapopote is most probably a result of gas compo-sition not formation dynamics

53 Relation between crystal structure microstructurehydrocarbon composition and gas hydrate stability

It is noteworthy that the thermodynamically more stablehydrates lack mesoporosity but instead feature a dense and tier-likemicrostructure as evidenced by SEM They form from a hydro-carbon mixture of about 70ndash80 of C1 plus 20ndash30 C2ndashC5 andcrystallize to structure II hydrate Mesopores (200ndash500 nm diam-eter) have been considered as a typical characteristic for gashydrate (Kuhs et al 2000 2004 Techmer et al 2005 Bohrmannet al 2007) and were identified in all Chapopote hydrates Thereonly sI gas hydrates were found by XRD analyses and of thosesamples the gas composition turns out to be 98 methane

These observations point to a tight connection between signif-icantly large fractions of light hydrocarbons (C2ndashC5) sII hydrateformation dense microstructure and greater gas hydrate stabilityHence it follows that the relation of microstructure and gas hydratestability needs to be discussed

It is likely that hydrate dissociation starts on the hydrate surfaceTherefore a larger surface area would lead to an accelerateddecomposition Given that the mesopores (200ndash500 nm) were notjust a surface characteristic but an element of the internal structureof the mesoporous hydrates it would be reasonable to assumea larger surface area for the mesoporous hydrates because then thewalls of the pores would add to the surface area Hydrauliccompression and density experiments suggest that almost 20 volof bulk methane hydrate samples exist as mesopores (200ndash500 nm)(Suess et al 2002) However area estimations from SEM inspec-tions indicate that also larger pores of about 1 mm size or bubblesmight have added to the calculated mesoporous volume (unpub-lished data) such that the true volume of mesopores might besignificantly lower Kuhs et al (2004) argue from specific surfacearea experiments that the mesopores of both laboratory-grown andsubmarine gas hydrates are predominantly closed Thus they maybe a feature related to the grain boundary network of gas hydratesas was suggested by recent synchrotron microtomographic work(Murshed et al 2008) Hence it seems unlikely that the mesoporescould contribute significantly to an accelerated bulk decompositionof the hydrates

Most importantly the stability of gas hydrates is primarilygoverned by the cage filling of adequate gases (Sloan and Koh2007) and consequently by the crystallographic structure typedepending on the available gases at a geological site The datapresented here show that hydrocarbon mixtures comprisingcomparatively less volatile hydrocarbons like propane and butaneprefer a structure II gas hydrate and form dense microstructuresLower pressure or lower subcooling is needed to shift a gas

assortment consisting of a larger fraction of less volatile hydro-carbons into the hydrate stability This suggests that the volatility ofthe hydrate formers could be an important factor accounting for theobserved hydrate microstructure This suggestion is underlined bythe fact that volatile non-hydrocarbon gases like N2 O2 or Aralso form sII hydrate but feature a mesoporous microstructure(Kuhs et al 2000)

54 Mesoporous microstructure

The mesoporosity seems to be a function of the gas chemicalcomposition in that only highly volatile gases feature that micro-structure But also dense surfaces have been produced formethane-dominated sI gas hydrate Stern et al (2003 2004 2005)and Klapproth et al (2007) undertook gas hydrate formationexperiments above the ice point and produced dense surfaces Kuhset al (2000) showed both dense and mesoporous hydrate surfacesfor samples reacted below the ice point The samples often showa uniform mesoporous microstructure which sometimes isamended by dense parts The mesoporous microstructure can infact be used as a rapid mean of identifying gas hydrates in thescanning electron microscope as it uniquely is formed when gashydrates are present in the sample Indeed estimates of the relativeproportion of icewater and gas hydrates (as identified by themesoporosity) in the SEM compares very favorably with resultsfrom quantitative phase analysis by means of X-ray diffraction(Bohrmann et al 2007)

55 Intrinsic dense microstructure

The dense microstructure was clearly observed in samples thatwere well preserved with little ice fractions Accordingly the SEMimages should show the intrinsic and unaltered microstructure Forvarious reasons it can be ruled out that the dense surface derivesfrom ice of decomposed gas hydrate Firstly detailed microstruc-ture studies were performed and the ice fractions of the samplesTVG-10 as well as GC-2 established by diffraction are too low toaccount for the many SEM observations of dense materialSecondly carbon signals as measured by SEM-mounted EDX unitsof dense samples cannot derive from ice as water molecules do notcontain carbon Thirdly decomposing gas hydrate does not formsmooth and dense surfaces but macropores of several mm indiameter as shown in Bohrmann et al (2007) from Hydrate Ridgesamples Such macropores are phenomena due to partial decom-position Larger than those dissociation-derived macropores areintrinsic characteristics like gas-filled voids (Bohrmann et al 1998)which are shown by aragonite precipitates that image the bubblesby the surface morphology (Bohrmann and Torres 2006)

An alternative explanation for dense surface structure imagedby electron microscopy would be that they originate from meso-pores filled with oil Indeed the co-occurrence of the dense hydratemicrostructure with oil patches could support such a suppositionHowever dense surfaces of sII gas hydrate were found also inspecimen without any visible oil patches indicating that the densesurface is truly intrinsic In addition the mesoporous microstruc-ture is discernable in oil-stained sI gas hydrates retrieved from oilseeps in the Eastern Black Sea (unpublished data)

56 Gas hydrate and liquid hydrocarbons

SEM images show shimmering oil coatings on gas hydratesurfaces (Fig 5) Although the coexistence of oil and gas hydrate inthe GOM is known since long (Brooks et al 1984) the microscopicscale of the vicinity has not been explored so far MacDonald et al(2003) showed massive deposits of Bush Hill yellowish gas

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125124

hydrates on a meter scale which are colored by oil Sassen et al(2001c) report on oil-stained sediments with gas hydrates as vein-fillings from Atwater Valley (Block GC425) in the GOM Liquid crudeoil is described to occur as fluid inclusions and along fractureswithin the crystalline gas hydrate (Sassen et al 2001c) Our workshows that oil diffuses through gas hydrate occurrences as can beinferred macroscopically from Fig 2 and massive decolorization ofgas hydrate deposits (MacDonald et al 1994 2003) as well asmicroscopically from Fig 5

The presence of hydrocarbons outside the gas hydrate crystalstructure is relevant for hydrate characterization In interpretingchemical data from hydrate-associated light hydrocarbons it nor-mally is assumed that those gases are entrapped in the hydratecages ie as crystal-bound guest molecules (Sassen et al 19942001b Chen and Cathles 2003 MacDonald et al 2004 2005)These data are commonly used to deduce the crystal structure andthe thermodynamic stability and on those grounds to estimate theabundances of hydrates in a region (Sassen et al 2001d) Our SEMimages show that oil sticks to the hydrate surface As it turns outthat heavy hydrocarbons occur along with the hydrates the ques-tion arises whether light hydrocarbons also accompany petroleumand hydrates for instance by being sequestered in long-chainedhydrocarbons without being included in the structure One possi-bility would be that volatile hydrocarbon molecules stick to thehydrate surface by adsorption (eg Larsen et al 1998) Liquidhydrocarbons are visible by FE-SEM but gaseous hydrocarbonscannot be detected The cage filling of the hydrate structures bydistinct gas species can be best obtained from Raman spectroscopy(eg Hester et al 2007 Chazallon et al 2007 Bourry et alsubmitted for publication) Also gas exchange reactions in the cagescan be traced by Raman measurements (Murshed and Kuhs 2007)

6 Conclusions

The electron microscopically depicted microstructure of gashydrates from the Green Canyon (GC415) and Bush Hill (GC185)features smooth and dense surfaces Mesopores are not discernablein these samples These hydrates form sII gas hydrate proved by XRDand comprise 70ndash80 methane and 20ndash30 C2ndashC4 hydrocarbons

There is clear evidence that hydrocarbons with relatively lowvolatility tend to form dense gas hydrate microstructures whilemethane and light non-hydrocarbons gases often form mesoporoushydrates This suggests that the volatility of the gases playsa significant role for the microstructural appearance and supportsalso the earlier finding that a mesoporous microstructure isa characteristic signature of more volatile gas hydrates in particularof methane hydrate (Kuhs et al 2004) Dense microstructures areless readily assigned to a certain structure type or composition asmethane hydrate sometimes also can appear in a dense micro-structural form (Stern et al 2004 2005 Klapproth et al 2007)

Oil patches together with dense gas hydrate occur in directcontact with each other Insight into the formation process and thedetailed chemistry at the micro-scale cannot be deduced from thepresent data however

It will not be possible to relate the porous nature of hydrates toany specific geological phenomenon until a full understanding of itsformation is achieved However in microstructural SEM work ongas hydrates the appearance of mesopores can still be used as anindicator for gas hydrates and ndash as we have shown here ndash to someextent also for their chemical nature

Acknowledgements

We thank the shipboard scientific parties and crews on RVSonne SO 174 and RV Meteor M672 for their help Likewise we

thank the HASYLABHamburg for beam time and support con-cerning the synchrotron XRD work We thank two anonymousreviewers and Keith Hester for helpful comments and suggestionsSA Klapp thanks the Vollkammer Scholarship for funding as wellas GLOMAR Graduate School for Marine Sciences This is publica-tion GEOTECH - 920 of the RampD-Programme GEOTECHNOLOGIENfunded by the German Ministry of Education and Research (BMBF)and German Research Foundation (DFG) Funding for this researchwas provided by the METRO grant 03G0604A (BMBF)

References

Bohrmann G Greinert J Suess E Torres M 1998 Authigenic carbonates fromthe Cascadia subduction zone and their relation to gas hydrate stabilityGeology 26 647ndash650

Bohrmann G Torres M 2006 Gas hydrates in marine sediments In Schulz HDZabel M (Eds) Marine Geochemistry Springer pp 481ndash512

Bohrmann G Schenck S 2004 RV SONNE Cruise Report SO174 OTEGA II Balboa ndashCorpus Christi ndash Miami (1 Octoberndash12 November 2003) GEOMAR Report No117 116 pp

Bohrmann G Kuhs WF Klapp SA Staykova DK Techmer KS Klein HMurshed M Abegg F 2007 Appearance and preservation of natural gashydrate from Hydrate Ridge sampled during ODP Leg 204 drilling MarineGeology 244 (1ndash4) 1ndash14

Bohrmann G Spiess V Cruise Participants 2008 Report and Preliminary Resultsof RV Meteor Cruise M672a and 2b Balboa ndash Tampico ndash Bridgetown 15Marchndash24 April 2006 Fluid Seepage in the Gulf of Mexico Berichte Fachber-eich Geowissenschaften Universitat Bremen Bremen No 263 119 pp

Bourry C Chazallon B Charlou JL Donval JP Ruffine L Henry P Geli LCagatay MN Moreau M the Marnaut Scientific Party Free gas and gashydrates from the Sea of Marmara Chemical and structural characterizationChemical Geology in press doi101016jchemgeo200903007

Brooks JM Kennicutt II MC Fay RR McDonald TJ Sassen R 1984 Thermo-genic gas hydrates in the Gulf of Mexico Science 225 (4660) 409ndash411

Buffett B Archer D 2004 Global inventory of methane clathrate sensitivity tochanges in the deep ocean Earth and Planetary Science Letters 227 185ndash199

Chazallon B Focsa C Charlou JL Bourry C Donval JP 2007 A comparativeRaman spectroscopic study of natural gas hydrates collected at differentgeological sites Chemical Geology 244 (1ndash2) 175ndash185

Chen DF Cathles III LM 2003 A kinetic model for the pattern and amounts ofhydrate precipitated from a gas steam application to the Bush Hill vent siteGreen Canyon Block 185 Gulf of Mexico Journal of Geophysical Research 108(B1) 2058 doi1010292001JB001597

Chen DF Cathles III LM Roberts HH 2004 The geochemical signatures ofvariable gas venting at gas hydrate sites Marine and Petroleum Geology 21 (3)317ndash326

Collett TS Kuuskraa VA 1998 Hydrates contain vast store of world gas resourcesOil and Gas Journal 96 (19) 90ndash95

Collett TS 2002 Energy resource potential of natural gas hydrates AAPG Bulletin86 (11) 1971ndash1992

Dickens GR Castillo MM Walker JCG 1997 A blast of gas in the latest Paleo-cene simulating first-order effects of massive dissociation of oceanic methanehydrate Geology 25 (3) 259ndash263

Ding F Spiess V Bruning M Fekete N Keil H Bohrmann G 2008 A conceptualmodel for hydrocarbon accumulation and seepage processes around Chapopoteasphalt site southern Gulf of Mexico from high resolution seismic point ofview Journal of Geophysical Research-Solid Earth 113 (B8) B08404doi1010292007JB005484

Heeschen KU Hohnberg HJ Haeckel M Abegg F Drews M Bohrmann G2007 In situ hydrocarbon concentrations from pressurized cores in surfacesediments Northern Gulf of Mexico Marine Chemistry 107 498ndash515

Hester KC Dunk RM White SN Brewer PG Peltzer ET Sloan ED 2007 Gashydrate measurements at Hydrate Ridge using Raman spectroscopy Geo-chimica et Cosmochimica Acta 71 2947ndash2959

Humphris CC 1979 Salt movement in continental slope northern Gulf of MexicoAAPG Bulletin 63 782ndash798

HWHYD 2005 HWHYD Heriot-Watt-Hydrate Version 11 Multiphase Hydrate andPVT Prediction Software for Microsoft Windows Operating Systems Centrefor Gas Hydrate Research Institute of Petroleum Engineering Heriot-Watt-University Edinburgh UK

Kennett J Cannariato K Hendy I Behl R 2002 Methane Hydrates in QuaternaryClimate Change the Clathrate Gun Hypothesis AGU Books Board 216 pp

Klapp SA Klein H Kuhs WF 2007 First determination of gas hydrate crystallitesize distributions using high-energy synchrotron radiation GeophysicalResearch Letters 34 (13) L13608 doi1010292006GL029134

Klapproth A Goreshnik E Staykova DK Klein H Kuhs WF 2003 Structuralstudies of gas hydrates Canadian Journal of Physics 81 503ndash518

Klapproth A Techmer KS Klapp SA Murshed MM Kuhs WF 2007 Micro-structure of gas hydrates in porous media In Kuhs WF (Ed) Physics andChemistry of Ice Royal Society of Chemistry Publishing Cambridge UK pp321ndash328

SA Klapp et al Marine and Petroleum Geology 27 (2010) 116ndash125 125

Klauda JB Sandler SI 2005 Global distribution of methane hydrate in oceansediment Energy amp Fuels 19 459ndash470

Kuhs WF Klapproth A Gotthardt F Techmer KS Heinrichs T 2000 Theformation of meso- and macroporous gas hydrates Geophysical ResearchLetters 27 (18) 2929ndash2932

Kuhs WF Genov G Goreshnik E Zeller A Techmer KS Bohrmann G 2004The impact of porous microstructures of gas hydrates on their macroscopicproperties International Journal of Offshore and Polar Engineering 14 (4)305ndash309

Kvenvolden KA 1993 Gas hydrates ndash geological perspective and global changeReviews of Geophysics 31 (2) 173ndash187

Kvenvolden KA 1995 A review of the geochemistry of methane in natural gashydrate Organic Geochemistry 23 997ndash1008

Larsen R Knight CA Sloan ED 1998 Clathrate hydrate growth and inhibitionFluid Phase Equilibria 150ndash151 353ndash360

Larson AC van Dreele RB 1994 General Structure Analysis System (GSAS) LosAlamos National Laboratory

Lu H Seo Y Lee J Moudrakovski I Ripmeester JA Chapman NR Coffin RBGardner G Pohlman J 2007 Complex gas hydrate from the Cascadia marginNature 445 (7125) 303ndash306

MacDonald IR Guinasso NL Sassen R Brooks JM Lee L Scott KT 1994 Gashydrate that breaches the sea floor on the continental slope of the Gulf ofMexico Geology 22 699ndash702

MacDonald IR Sager WW Peccini MB 2003 Gas hydrate and chemosyntheticbiota in mounded bathymetry at mid-slope hydrocarbon seeps northern Gulfof Mexico Marine Geology 198 133ndash158

MacDonald IR Bohrmann G Escobar E Abegg F Blanchon P Blinova VBruckmann W Drews M Eisenhauer A Han X Heeschen K Meier FMortera C Naehr T Orcutt B Bernard B Brooks J de Farago M 2004Asphalt volcanism and chemosynthetic life in the Campeche Knolls Gulf ofMexico Science 304 999ndash1002

MacDonald IR Bender LC Vardaro M Bernard B Brooks JM 2005 Thermaland visual time-series at a seafloor gas hydrate deposit on the Gulf of Mexicoslope Earth and Planetary Science Letters 233 45ndash59

Matsumoto R Uchida T Waseda A Takeya S Hirano T Yamada K Maeda YOkui T 2000 Occurrence structure and composition of natural gas hydratesrecovered from the Blake Ridge ODP Leg 164 Northwest Atlantic In Proceed-ings of the Ocean Drilling Program Scientific Results vol 164 pp 13ndash28

Max MD Johnsen HK Dillon WP 2006 Economic Geology of Natural GasHydrate Coastal Systems and Continental Margins Springer Dordrecht 341 pp

Murshed MM Kuhs WF 2007 Hydrate phase transformations imposed by gasexchange In Kuhs WF (Ed) Physics and Chemistry of Ice Royal Society ofChemistry Publishing Cambridge UK pp 427ndash433

Murshed MM Klapp SA Enzmann F Szeder T Huthwelker T Stampanoni MMarone F Hintermuller C Bohrmann G Kuhs WF Kersten M 2008Natural gas hydrate investigations by synchrotron radiation X-ray cryo-tomo-graphic microscopy Geophysical Research Letters 35 L23612 doi1010292008GL035460

Rawn CJ Sassen R Ulrich SM Phelps TJ Chakoumakos BC Payzant EA2008 Low temperature X-ray diffraction studies of natural gas hydrate samplesfrom the Gulf of Mexico In Englezos P Ripmeester J (Eds) Proceedings of the6th International Conference on Gas Hydrates (ICGH 2008) Vancouver BritishColumbia Canada

Rowan MG Jackson MPA Trudgill BD 1999 Salt-related fault families and faultwelds in the northern Gulf of Mexico AAPG Bulletin 83 1454ndash1484

Ruppel C Dickens G Castellini DG Gilhooly W Lazarralde D 2005 Heat andsalt inhibition of gas hydrate formation in the northern Gulf of MexicoGeophysical Research Letters 32 L04605 doi1010292004GL021909

Sassen R Mac Donald IR Requejo AG Guinasso Jr NL Kennicutt II MCSweet ST Brooks JM 1994 Organic geochemistry of sediments fromchemosynthetic communities Gulf of Mexico slope Geo-Marine Letters 14110ndash119

Sassen R Sweet ST Milkov AV DeFreitas DA Kennicutt II MC 2001a Ther-mogenic vent gas and gas hydrate in the Gulf of Mexico slope is gas hydratedecomposition significant Geology 29 (2) 107ndash110

Sassen R Losh SL Cathles III L Roberts HH Whelan JK Milkov AVSweet ST DeFreitas DA 2001b Massive vein-filling gas hydrate relation toongoing gas migration from the deep subsurface in the Gulf of Mexico Marineand Petroleum Geology 18 551ndash560

Sassen R Sweet ST DeFreitas DA Morelos JA Milkov AV 2001c Gas hydrateand crude oil from the Mississippi Fan Foldbelt downdip Gulf of Mexico SaltBasin significance to petroleum system Organic Geochemistry 32 999ndash1008

Sassen R et al 2001d Stability of thermogenic gas hydrates in the Gulf of Mexicoconstraints on models of climate change In Paull CK Dillon WP (Eds)Natural Gas Hydrates Occurrence Distribution and Detection GeophysicalMonograph American Geophysical Union Washington DC pp 131ndash143

Sloan ED Koh CA 2007 Clathrate Hydrates of Natural Gases CRC Press BocaRaton 752 pp

Staykova DK Kuhs WF Salamatin AN Hansen T 2003 Formation of porousgas hydrates from ice powders diffraction experiments and multistage modelJournal of Physical Chemistry B 107 10299ndash10311

Stern LA Circone S Kirby SH Durham WB 2003 Temperature pressure andcompositional effects on anomalous or lsquolsquoselfrsquorsquo preservation of gas hydratesCanadian Journal of Physics 81 271ndash283

Stern LA Kirby SH Circone S Durham WB 2004 Scanning Electron Micros-copy investigations of laboratory-grown gas clathrate hydrates formed frommelting ice and comparison to natural hydrates American Mineralogist 891162ndash1174

Stern LA Kirby SH Durham WB 2005 Scanning electron microscope imagingof grain structure and phase distribution within gas-hydrate-bearing intervalsfrom JAPEXJNOCGSC et al Mallik 5L-38 what can we learn from comparisonswith laboratory-synthesized samples In Dallimore SR Collett TS (Eds)Scientific Results from the Mallik 2002 Gas Hydrate Production Research WellProgram Mackenzie Delta Northwest Territories Canada Geological Survey ofCanada Bulletin vol 585 p 19

Stern LA Kirby SH 2007 Grain and Pore Structure Imaging of Gas Hydratefrom Core MD02-2569 (Mississippi Canyon Gulf of Mexico) a First Look byScanning Electron Microscopy (SEM) US Geological Survey Open-File Report2004ndash1358

Suess E Torres ME Bohrmann G Collier RW Rickert D Goldfinger CLinke P Heuser A Sahling H Heeschen K Jung C Nakamura K Greinert JPfannkuche O Trehu A Klinkhammer G Whiticar MC Eisenhauer ATeichert B Elvert E 2001 Sea floor methane hydrates at Hydrate RidgeCascadia Margin In Paull C Dillon WP (Eds) Natural Gas Hydrates Occur-rence Distribution and Detection Geophysical Monograph vol 124 AmericanGeophysical Union pp 87ndash98

Suess E Bohrmann G Rickert D Kuhs WF Torres ME Trehu A Linke P 2002Physical properties and fabric of near-surface methane hydrates at hydrateridge Cascadia margin In Proceedings of the Fourth International Conferenceon Gas Hydrates Yokohama pp 740ndash744

Techmer K Heinrichs T Kuhs WF 2005 Cryo-electron microscopic studies onthe structures and composition of Mallik gas-hydrate-bearing samples InDallimore SR Collett TS (Eds) Scientific Results from the Mallik 2002 GasHydrate Production Research Well Program Mackenzie Delta NorthwestTerritories Canada Geological Survey of Canada Bulletin vol 585 p 12

Toby BH 2001 EXPGUI a graphical user interface for GSAS Journal of AppliedCrystallography 34 210ndash213

Torres ME Wallmann K Trehu AM Bohrmann G Borowski WS Tomaru H2004 Gas hydrate growth methane transport and chloride enrichment at thesouthern summit of Hydrate Ridge Cascadia margin off Oregon Earth andPlanetary Science Letters 226 225ndash241

Whelan J Eglinton L Cathles III L Losh S Roberts H 2005 Surface andsubsurface manifestations of gas movement through a NndashS transect of the Gulfof Mexico Marine and Petroleum Geology 22 479ndash497