35. PETROLOGY AND GEOCHEMISTRY OF SANDSTONES IN THE …

35. PETROLOGY A N D GEOCHEMISTRY OF SANDSTONES IN THE LOWER CRETACEOUSSUBMARINE FAN COMPLEX, DEEP SEA DRILLING PROJECT HOLE 603B 1

D. R. Prezbindowski and Edward D. Pittman, Amoco Production Company2

ABSTRACT

Diagenesis of the fine-grained, feldspathic sandstones in the Lower Cretaceous submarine fan complex cored inDSDP Hole 603B can be considered to have occurred in three stages: (1) replacement of matrix and framework grainsby pyrite, siderite, phillipsite (?), and particularly by ferroan calcite; (2) dissolution of ferroan calcite and feldspars toproduce secondary macroporosity; and (3) development of sparse feldspar and quartz overgrowths, and authigenicmodification of remnant matrix. Only ferroan calcite is a volumetrically important diagenetic mineral phase (up to50 vol.%).

Matrix in thin sandstone turbidite deposits has been extensively replaced by ferroan calcite. Carbon stable isotopedata suggest that organic diagenesis had only a minor influence on calcite precipitation. Oxygen stable isotope data in-dicate that the minimum average calcite precipitation temperature was 40° C. Preliminary calculations show that steady-state diffusion of C a + + from the dissolution of nannoplankton skeletal material in the interbedded pelagic marls to theassociated sandstones is a feasible transport mechanism.

A thick sandstone unit from 1234-1263 m sub-bottom is extensively replaced by calcite near the upper and lowercontacts. Farther into the sand body away from the contacts, the sandstone has good secondary porosity resulting fromthe dissolution of ferroan calcite that partially replaced matrix and framework grains. The central portion of the thicksand appears to be a channel with high-energy clean sand. We believe that the channel provided a conduit for focusedflow of diagenetic compactional fluids responsible for dissolution. Focused flow may be the result of the earlier lithifi-cation of the pelagic limestones and thin-bedded sandstones which, then formed vertical permeability barriers. Calcitedissolution has occurred and may still be occurring at temperatures less than 65°C.

INTRODUCTION

Hole 603B penetrated and continuously cored a pre-viously unknown Lower Cretaceous submarine fan com-plex (lithologic Subunit VA) on the lower continental rise270 mi. (435 km) off Cape Hatteras (Fig. 1). This com-plex consists of terrigenous sandstone and claystone tur-bidite deposits interbedded with pelagic marls (Site 603chapter, this volume). Selected sandstones are porous andin places friable (Fig. 2). The geologic setting and his-tory preclude the possibility of meteoric water, soil-de-rived organic acids, or surface weathering affecting thesesediments after deposition. Therefore, diagenetic eventsare believed to be related to in situ controls. Our pur-pose is to provide a survey study of the general diagenet-ic history of the sandstones and limestones using opticaland electron microscopy, X-ray diffractometry, micro-probe analytical techniques, stable isotopes, and organ-ic geochemistry. The database (30 samples) is inadequateto address every diagenetic problem. We hope that thisstudy will provide a basis for future, more definitive work.

METHODS

Oxygen and carbon stable isotope compositions of carbonate min-erals were determined using the phosphoric acid technique first docu-mented by McCrea (1950). All analyses were made on a Finnigan MAT251, constant-ratio mass spectrometer. The 20-100 mg whole-rock sam-

van Hinte, J. E., Wise, S. W., Jr., et al., Init. Repts. DSDP, 93: Washington (US.Govt. Printing Office).

2 Address (Prezbindowski, present address): International Petrology Research, 3228 East15th St., TUlsa, OK 74104; (Pittman) Amoco Production Co., Research Center, P.O. Box3385, Tlilsa, OK 74102.

pies were first vacuum roasted at 360° C for 1 hr. to remove volatilecomponents. Results are referenced to the Chicago PDB standard un-less otherwise indicated. Analytical error was found to be no greaterthan ±0.2‰

Analyses of feldspars were made using a Kevex 8000-III energy-dis-persive spectrometer attached to an Applied Research LaboratoriesSEMQ microprobe. This permitted use of a low-beam current to pre-vent loss of alkalis from the analysis area. Beam operating parameterswere: excitation voltage, 20 kV; beam current, 1 nA (measured in aFaraday cage); beam diameter 10 µm; and emission current, 125 µA.Counts were accumulated to a preset precision of 0.5% in a windowset for SiKα. A minimum of two points per grain were analyzed andaveraged. Analyses whose totals were less than 95% or more than 105%were rejected from the data set.

Calcite cement was also analyzed on an Applied Research Labora-tories SEMQ microprobe with the following operating conditions: ex-citation voltage, 20 kV; emission current, 125 µA; beam current, 1 nA(measured in a Faraday cage); spectrum acquisition time, 100 s. Theanalytical totals suggest, and cross analyses of standards confirm, thatanalytical errors under these conditions are within 3-5% of the amountpresent.

All rock samples were pressure impregnated with a blue-dyed epoxyresin before thin sections were prepared. Thin sections were stained tofacilitate recognition of carbonates and K-feldspar using the techniquesof Dickson (1966) and Bailey and Stevens (1960), respectively.

Vitrinite reflectance was determined on four samples using stan-dard reflective light techniques discussed by Stach (1975). Elementalanalytical techniques used for kerogen are comparable to those of Rob-inson (1969).

PETROLOGY

Sandstones

Texture and Composition

Sandstones of Subunit VA are predominantly fine-grained, but range from very fine to medium-grained.

961

D. R. PREZBINDOWSKI, E. D. PITTMAN

45°N

40'

35 '

1 T I 1 I I 1 I I [

DSDP SitesLegs 2, 1 1 , 4 3 , 4 4 , 7 6 , 9 3

Location of DSDP Site 603 about 435 km (270 n. mi.) off the east coast of North America.

Typically, the subangular sand-sized components are em-bedded in an argillaceous matrix, which constitutes 0-35% of the total rock and varies inversely with ferroancalcite that replaces the matrix. We interpret the matrixto be detrital. The clay particles are arranged in a face-to-face packing and appear to "flow" around frame-work grains because of compaction (Fig. 3). There is noevidence of delicate fibers or euhedral crystals that char-acterize authigenic clays (Wilson and Pittman, 1977).We also find no evidence of the type suggestcu by Dick-inson (1970) to support phyllosilicate cement or pseudo-matrix origins. If the matrix is diagenetic, it would haveto be called epimatrix (Dickinson, 1970), because it lacksthe clear evidence needed to classify the interstitial clayas phyllosilicate cement.

No attempt was made to separate the matrix for X-raydiffractometry because we felt the disaggregation pro-cess would also affect the sand-size lithic fragments. Wedid obtain X-ray diffraction data for the claystone "tails"of the turbidites, which should be representative of thedetrital matrix in the sandstone (see section on fine-grained rocks). Mixed-layer, randomly interstratified il-lite/smectite (80% expandable) is the dominant clay, al-

though minor kaolinite, illite, and chlorite are also pres-ent.

Point-counts for 19 samples indicate the sandstonesare arkoses and subarkoses in a modified Dott (1964)classification (Fig. 4). Framework grains and their meancomposition as a percentage of the total rock includemonocrystalline quartz (29.6 vol.%), polycrystallinequartz (4.7 vol.%), lithic fragments (5.0 vol.%), K-feld-spar (8.7 vol.%), plagioclase (4.8 vol.%), muscovite (1.9vol.%), biotite (1.5 vol.%), and other (1.1 vol.%). Lith-ic fragments include plutonic igneous, schist, phyllite,shale, and altered volcanic. Zircon, tourmaline, pyrox-ene, glauconite, and collophane are the most commonaccessory minerals. Skeletal grains and woody debris arefragmented and sparse.

Composition of detrital plagioclase in Subunit VAbased on microprobe analyses in three samples is shownin Figure 5. No plagioclase is more calcic than An26,and 45% of the grains are in the Ano_s modal class. De-trital plagioclase in modern deep-sea sands from passiveand active tectonic settings has a wide range in composi-tion: about one-third of the grains are more calcic thanAn40; 19% are essentially pure albite (Maynard, 1984).

962

PETROLOGY AND GEOCHEMISTRY OF SANDSTONES, HOLE 603B

Sandstone"channel"29 m thick(8.7 m lostcore)

1235-

1240-

1245-

ε 1 250 - $£•:

1255-

1260-

1265-

Thin interbedssandstone andpelagic marl

Extensive calcite replacement;no macroporosity

Friable(no matrix or cement?)

Secondary porosity —20%5% Matrix; 1% Calcite

Secondary porosity — 10%21.5% Matrix; < 1 % Calcite

Extensive calcite replacement;no macroporosity

Thin sandstones haveextensive replacementcalcite; no macroporosity

Figure 2. Diagrammatic representation of thick sandstone near the topof the Lower Cretaceous turbidite section in Hole 603B. Arrowsindicate locations of samples studied.

Selective removal of the more calcic plagioclase duringdiagenesis, without modifying the composition of theremnant plagioclase, would increase the percentages ofthe remaining modal classes. A proportional recalcula-tion of these classes produces the lower histogram ofFigure 5. This histogram also has Ano_5 as the primaryclass (37%), which compares favorably with the upperhistogram (Fig. 5) for the Lower Cretaceous sandstones.This suggests that provenance and selective removal ofcalcic plagioclase is a possible explanation for the abun-dance of albitic plagioclase in the Lower Cretaceous sand-stones. Albitization may play a role, but more micro-probe data over a considerable depth range would beneeded to substantiate the process.

Diagenesis

Most of the sandstone samples are argillaceous, withessentially no macroporosity. Authigenic components inthese samples are restricted to replacements because ofthe lack of open pores needed for authigenic precipita-tion. Pyrite occurs ubiquitously as a replacement of ar-gillaceous material and within chambers of skeletal grains.Trace amounts of titanium minerals (ilmenite-leucoxene)are associated with volcanic lithic fragments. The mostsignificant authigenic event is the replacement of argil-laceous matrix and framework grains by medium crys-talline ( — 75 µm) calcite (Fig. 6). Based on 25 micro-probe spot analyses, the calcite cement in Sample 603B-46-1, 55 cm is ferroan, having an average composition(mol%) of Ca94.o6 Mßo.75 Fe2.34 Mn2.85. Zonal variationof iron and/or manganese was not detected in the cal-

B

Figure 3. A. Scanning electron micrograph showing detrital matrix(m). B. Energy-dispersive X-ray spectrum for a spot analysis at"m" suggests a mixed-layer illite/smectite clay. There is a sugges-tion of a small magnesium peak at the left margin, which may in-dicate the presence of chlorite. Unlabeled peaks correspond to ele-ments in the conductive coating. Sample 603B-62-2, 47-50 cm.

Quartz + chert+ quartzite

Quartz arenite

Subarkose Sublitharenite

> 10% Porosity

o < 10% Porosity

Feldspar Lithic

Dotfs (1964)classification

(modified)

Figure 4. Classification of sandstones in the Lower Cretaceous subma-rine fan complex by thin-section modal analysis. Sandstones repre-sent arkoses and subarkoses.

963


50 η

4 0 -

~. 3 0 -

20-

10-

Lower Cretaceous sandstoneDSDP Hole 603Bn= 78

5 10 15 20 25

Plagioclase compostion (wt. % An)

30

40 -i

30-

E 20-

O

1 0 -

Reconstructed from modern(Maynard, 1984), assumingloss of plagioclase > Aπ3 Q

10 15 20 25 30

Plagioclase composition (wt. % An)

Figure 5. The upper histogram depicts plagioclase composition in LowerCretaceous sandstones, based on microprobe analyses in three sam-ples. The lower histogram is based on a study of plagioclase com-position in modern sediments (Maynard, 1984). For this histogram,plagioclase more calcic than An 3 0 was assumed to be lost by dia-genesis and the remaining plagioclase was redistributed proportion-ately among the remaining classes. In both histograms, albitic ma-terial constitutes the primary mode, which suggests that albitiza-tion need not play a major role in explaining the compositionaldistribution of the Lower Cretaceous sandstones.

cite. The replacement origin for the ferroan calcite is sup-ported by its anhedral crystal morphology (loaf-shaped)and the lack of enfacial junctions at triple contact points(Bathurst, 1971). In some samples, the replacement isessentially complete, but in other samples the replace-ment carbonate is less than 1 vol.%. Within a thin sec-tion, a transition from unreplaced to advanced replace-ment of the matrix by calcite sometimes occurs (Sample603B-68-2, 92-95 cm). The scanning election microscopereveals remnant clay within and among the replacementferroan calcite crystals (Fig. 7A). Energy-dispersive X-ray(EDX) substantiates the ferroan calcite identification(Fig. 7B) and indicates that the entrapped mixed-layerillite/smectite is also iron-bearing (Fig. 7C). The iron inthe smectite may be the source of the iron in the replace-ment calcite, but the current database is inadequate to

0.125 mm

Figure 6. Transmitted-light photomicrograph showing ferroan calcite(c) replacing darker matrix (m) in which light-colored sand grainsare embedded. Sample 603B-62-3, 47-50 cm.

substantiate this hypothesis. Ferroan calcite also replac-es lithic fragments, feldspars, and quartz (Fig. 8).

Excluding the friable sand, macroporosity is predom-inantly secondary and has formed from the dissolutionof ferroan calcite replacements (Fig. 9). This has pro-duced oversized, moldic, and intragranular pores, andcockscomb textures in feldspars, all of which have beencited as evidence for the presence of secondary porosity(Schmidt and McDonald, 1979). Patches of unreplacedargillaceous matrix are common in porous sandstoneswhen viewed with a scanning electron microscope (Fig.10A) and appear as brownish material in thin section.Silicate framework grains can dissolve directly without acarbonate replacement stage (Surdam et al., 1984), butthe textural evidence suggests that replacement of sili-cate grains is so common that intragranular-moldic po-rosity probably has formed from dissolution of ferroancalcite (Fig. 10B). Ferromagnesian heavy minerals showevidence of dissolution. Lithic fragments are often mi-croporous, but rarely contain secondary macropores.

The core descriptions (see Site 603 chapter, this vol-ume) reported that core from the interval 1234.4-1263.4m sub-bottom was soft, soupy, or disturbed and in plac-es unconsolidated. Approximately 22.5 m of this coredinterval was not recovered. Our evaluation of the porousinterval is based on seven samples, including one sampleof friable sand (Fig. 2). Judging by the samples studied,this is the only interval with significant macroporosity(> 10%). The other sandstone samples have, at most, afew percent porosity, but typically have < 1 vol.% mac-roporosity.

Secondary porosity development is partly dependenton the interplay of two processes: (1) the amount of fer-roan calcite that replaced the matrix and frameworkgrains; and (2) the amount of ferroan calcite dissolvedto create secondary porosity. The following discussionof selected samples will illustrate the interrelationship.

964

PETROLOGY AND GEOCHEMISTRY OF SANDSTONES, HOLE 6O3B

! I

10µm

Ca

Figure 7. A. Scanning electron micrograph showing sand grain in lowerright bound by ferroan calcite (c) containing wisps of unreplacedclay matrix (m). B. EDX spectrum for calcite substantiates the pres-ence of iron. Unlabeled peaks are Au and Pd contained in theevaporative coating. C. EDX pattern for matrix shown in A sug-gests an iron-bearing mixed-layer illite/smectite; Sample 603B-66-4,45-49 cm.

Sample 603B 48-1, 47-51 cm (-1253.3 m sub-bot-tom), has 21.5 vol.% argillaceous matrix, < 1 vol.% fer-roan calcite and 10 vol.% macroporosity. Calcite replace-ment of matrix was not extensive, but essentially all ofthe replacement calcite was dissolved to create second-ary porosity. In contrast, Sample 603B-46-1, 60 cm(1234.2 m sub-bottom) has essentially no matrix, 43.5vol.% ferroan calcite, and no macroporosity. Althoughcalcite replacement is extensive, the calcite was not dis-solved to create secondary porosity. If the ferroan cal-cite was removed from this sample, it would yield a sam-ple analogous to the friable material from 603B-46-2,

50µm

Figure 8. Transmitted-light photomicrograph of ferroan calcite (c) re-placing feldspar grains (f). Ferroan calcite has also replaced the ar-gillaceous matrix. Sample 603B-68-2, 92-95 cm.

0.125 mm

Figure 9. Transmitted-light photomicrograph of secondary porosity(p) resulting from partial dissolution of ferroan calcite (c), whichhas extensively replaced matrix. Remnant matrix appears as darkpatches. A plagioclase feldspar (0 grain is partially replaced bycalcite. Sample 603B-62-2, 47-50 cm.

30-32 cm (-1235.41 m sub-bottom), which has essen-tially no matrix, a trace of ferroan calcite, and a poros-ity of perhaps 35-40%. This presents a dilemma. Wehave interpreted the friable sand in Sample 603B-46-2,30-32 cm as having only patchy calcite cement that waslater dissolved. It would however, be possible to producethe same result through total replacement of matrix bycalcite, followed by complete dissolution of this calcite.Other porous samples typically have about 5 vol.% ma-trix, 1 vol.% ferroan calcite and about 20 vol.% macro-porosity. Calcite replacement of the matrix was not com-plete in these intervals, but essentially all of the calcitewas dissolved. The exact relationship of noncementation

965


40 µm

0.125 mm

Figure 10. A. Scanning electron micrograph of patches of remnantmatrix following replacement by and dissolution of ferroan calcite;Sample 603B-47-1, 69-73 cm. B. Dissolution of feldspar or fer-roan calcite that replaced feldspar (f) has produced a cockscombstructure and secondary porosity (sp). Intergranular porosity (p) isalso secondary and formed from dissolution of ferroan calcite thatreplaced matrix. Sample 603B-47-1, 35-39 cm.

to complete dissolution of calcite cement in the friablesand remains problematical. However, there is no doubtabout the importance of calcite dissolution in creatingmacroporosity in other associated samples.

Siderite occurs as sparse 25-35 µm "wheat"-shapedrhombohedra replacing ferroan calcite in a few samples.Siderite also occurs in the form of silt- and sand-sizedframework grains, probably as a complete replacementof an unknown precursor. These grains are composed ofa mosaic of crystals approximately 15 µm in diameter.The siderite grains occur in fine-grained laminae alongwith concentrations of opaques and micas. It is possiblethat these replacement siderite grains were derived fromthe shelf and transported to the continental rise. We be-lieve that these siderite grains could not have survived asubaerial weathering cycle so must have originated in

the marine environment. Siderite also occurs interspersedamong the cleavage flakes of some muscovite grains. Al-though this is evidence of an authigenic origin, it doesnot necessarily mean the siderite formed in situ.

Rare ankerite rhombohedra occur as va replacementof ferroan calcite. These crystals are too uncommon tobe paragenetically related to siderite.

Porous samples have authigenic minerals that are notpresent in matrix-rich or calcite-rich samples. The mostcommon is secondary feldspar, which occurs as over-growths on the surface of detrital grains (Fig. 11 A) andless commonly, overgrowths on the remnants of partlydissolved feldspar (Fig. 1 IB). The authigenic feldsparhas essentially pure sodium or potassium end-membercompositions, based on a microprobe survey. This is ex-pected from the work of Kastner and Siever (1979). In

20 µm

• • ; <

20µmFigure 11. Scanning electron micrographs. A. K-feldspar overgrowth;

Sample 603B-47-1, 69-73 cm. B. Tabular authigenic K-feldspar crys-tals associated with a partly dissolved feldspar. These minute crys-tals appear to have a common orientation, suggesting growth onremnant feldspar; Sample 603B-47-1, 35-39 cm. Energy-dispersivespectra show the authigenic feldspar to be potassic.

966


thin section, and when viewed with the SEM, small, eu-hedral K-feldspar crystals are seen associated with part-ly dissolved feldspar. These crystals are not always alignedwith the remnant feldspar and may have nucleated with-out a template.

Other authigenic growths include sparse quartz over-growths on framework grains and modification of clayminerals in the remnant matrix. The remnant matrix insandstones of Subunit VA often takes on an authigeniccharacter, with a more open packing and delicate wispsof clay (Fig. 12) forming as regenerated or neoformedclay minerals. Nannofossils occur in this material. Anenergy-dispersive X-ray spot analysis for an authigenicfiber of clay indicates the presence of calcium and po-tassium, suggesting a mixed-layer illite/smectite (Fig. 12).

A minor amount of zeolite was noted in several thinsections, associated with shaly laminae or present as ce-ment in very fine grained sandstone. Based on the detec-tion of potassium by EDX, we believe this is phillips-ite, although we were unsuccessful in confirming this byX-ray diffraction (XRD). This zeolite occurs as radiat-ing crystals with individual crystals up to 120 µm inlength.

Tabular crystals 10-15 µm long commonly are adheredto sand grains in the friable sample from 603B-46-2, 30-32 cm (~ 1235.4 m sub-bottom). EDX indicates the pres-ence of calcium and sulfur in these crystals, suggesting a

Figure 12. Scanning electron micrograph showing authigenic characterimposed on matrix remaining after replacement and dissolution offerroan calcite. Note nannofossils associated with the remnant ma-trix. Energy-dispersive X-ray spectrum shows presence of calciumand potassium suggestive of a mixed-layer illite/smectite. Sample603B 57-4, 50-56 cm.

sulfate such as anhydrite or gypsum. These crystals havebeen tentatively identified as gypsum by XRD, althoughthe crystal morphology appears to be more orthorhom-bic than monoclinic. Formation of the gypsum(?) is be-lieved to be due to desiccation during core storage, aprocess discussed by Briskin and Schreiber (1978).

Fine Grained Rocks

Claystones

The claystones probably represent the very fine com-ponents ("tails") of turbidity currents. X-ray diffractom-etry of the clay minerals (< 2 µm fraction) was made byJ. C. Hoffman for one sample (603B-56-1, 123-126 cm)of claystone. The dominant clay mineral is a mixed-lay-er, randomly interstratified illite/smectite with approxi-mately 80% smectite layers (Fig. 13). The compositionof this mixed-layer clay was determined following thetechnique of Reynolds and Hower (1970). Illite, kaolin-ite, and chlorite also occur in the claystone. Very fineflakes of muscovite are visible in thin section. X-ray dif-fractometry of the whole rock indicates the presence ofquartz, feldspar, and pyrite.

Pelagic Limestones

XRD indicates that calcite constitutes a significantproportion of the marls and marly limestones that com-prise the pelagic facies. The clay mineral assemblage isthe same as in the claystones. Other minerals includequartz and feldspar. In thin section, trace amounts ofphosphatic material of organic origin and muscovite arevisible.

Only the more robust nannofossils are present in thepelagic limestones from Subunit VA and some of theseare partly dissolved (Fig. 14). Delicate forms, such asholococcoliths, are largely absent (Covington and Wise,this volume).

In some thin sections, ferroan calcite in the form ofovoid to elliptical bodies up to 35 µm long occurs. Indi-vidual ferroan calcite crystals that compose the bodiesrange from 6-30 µm. These bodies probably representcalcite-replaced radiolaria (R. W. Scott, personal com-munication, 1985).

STABLE ISOTOPE GEOCHEMISTRY

The small size of the skeletal and authigenic carbon-ate cement components (<200 µm) made it impossibleto physically separate and analyze them individually. Thusthe stable isotopic composition results reported are whole-rock carbonate values (Fig. 15 and Table 1). Calcite skel-etal material and iron-rich calcite cement are the majorcarbonate components in the samples analyzed. Anker-ite and siderite are volumetrically insignificant. Theseminor carbonate components are believed not to con-tribute a distinct stable isotopic signal to the whole-rockanalyses.

The stable isotopic analyses were undertaken in orderto address the following questions. Is there a significantδ 1 8θ shift away from normal marine biogenic values forthe authigenic calcite? Is this shift a result of precipita-

967


<2µ,m air-dried

I = Illite and micaI/S - Mixed-layer

illite/smectiteC = ChloriteK = KaoliniteQ = Quartz

14 20

Degrees

26 32 38 44

< 2 µ m Ethylene-glycol-solvated

14I

20

Degrees

26

20 CuKα

32T38

144

Figure 13. X-ray diffractograms for air-dried and glycol-treated <2 µm fraction in claystone. Mixed-layer in-terstratified illite/smectite with 80% expandable layers is the dominant clay mineral. Sample 603B-56-1,123-126 cm.

tion during burial? If so, can the approximate tempera-ture of precipitation be determined? Do the δ13C com-positions indicate that organic diagenetic processes sig-nificantly influence carbonate replacement?

The oxygen and carbon stable isotope compositionof the carbonate components contained in the sandstone

and pelagic limestone samples (Fig. 15) shows wide range.This variability is directly related to the diagenetic modi-fication of the primary stable isotopic signature of thesediment. This primary signal is largely controlled bythe biogenic carbonate component in the sediment, most-ly nannoplankton skeletal debris. A compositional range

968


Table 1. Stable isotopic data of carbonate fractionfrom Hole 603B.

2 µ m

Figure 14. Scanning electron micrograph of partially dissolved Cocco-lith in marl; Sample 6O3B-5O-1, 96-104 cm.

+ Limestone• SandstoneO Sandstone, >3% porosity "PDB v '

x + 6

+3

- 1 0 - 6

3

6

9

J--12

Figure 15. Bivariate plot of δ 1 8 θ and δ 1 3C compositions of the calcitecontained in sandstones and pelagic limestones. The calcite con-tained in the sandstones apparently occurs in two trends. Samplesin the trend containing porous sandstones are from the same strati-graphic interval. The area enclosed by the dashed line representsthe expected stable isotopic compositional range of nannoplank-ton material at the time of deposition (Stout, 1985).

of - 1 to +1 for the oxygen stable isotope compositionand of + 0.5 to +3.0 for the carbon stable isotope com-position can be considered the primary signal range (Stout,1985). Diagenetic alteration of the carbonate componentswith increasing temperature will tend to shift the δ 1 8 θsignature toward more negative values. The magnitudeof this depletion trend will be dependent on the changesin the oxygen stable isotope composition of the intersti-tial water. This trend has been commonly documentedin other DSDP studies (Garrison, 1981). The δ 1 8 θ com-positional range of -2.5 to -4.5 for the limestones an-alyzed indicates that a 2-5%o negative diagenetic shifthas occurred during burial diagenesis. We discount the

Core-Section(interval in cm)

46-1, 5546-1, 6046-2, 30-3247-1, 35-3947-1, 69-7347-1, 54-6648-1, 47-5150-1, 96-10451-1, 16-2351-3, 88-9253-1, 41-4353-1, 47-5053-1, 51-5555-2, 15-2055-4, 93-9557-4, 55-5657-4, 61-6657-4, 69-7558-1, 61-6561-1, 86-9162-2, 50-5662-2, 56-5766-1, 7-1066-1, 7-1066-2, 110-11366-3, 109-11466-4, 45-4966-5, 28-3366-5, 49-5466-5, 49-5468-2, 92-9571-4, 68-71

Importance ofcarbonate

components (°/o)

Calcite

2030

< 5< 5<5<5< 525

5201515851525

<55

3020808085

<5

45304035

<5

255

Siderite

tr

tr

<5

< 5

tr

Isotopic values(‰ PDB)

δ 1 3 C

-6.9-10.1-0.8-3.0-0.1-0.4+ 0.7+ 0.6+ 1.2+ 1.0-0.6-0.8-0.3+ 1.4+ 1.7+ 1.2-0.6-0.4-0.9-0.1+'0.2+ 0.2+ 0.7-1.7+ 1.4+ 1.1+ 0.2+ 0.6+ 1.4-1.1+ 1.0+ 2.0

δ 1 8 θ

-5.6-5.9-3.0-3.1-1.3-3.0-0.8-3.0-2.6-3.3-8.4-6.3-3.7-2.4-2.6-2.3-8.3-8.5-7.9-4.4-4.5-4.2-2.6-6.7-7.4-7.5-8.3-8.4-3.1-10.3-8.0-2.8

Note: the approximate volumetric importance of the car-bonate components as determined by X-ray diffrac-tion is given. All analyses are referenced to the PDBstandard. Reproducibility was better than ±0.2‰ forcarbon and oxygen, tr = trace.

possibility of meteoric water diagenesis affecting the δ 1 8θcomposition of these sediments, strictly because of theirdepositional setting.

Using a - l.O‰ (SMOW) estimate for the oxygen sta-ble isotope composition of the Cretaceous ocean water,the approximate calcite cementation temperatures for thesands can be determined. Using the oxygen isotope cal-cite-water-temperature relationship established by O'Neilet al. (1969), a maximum carbonate precipitation tem-perature of 62°C was determined for a δ 1 8 θ composi-tion of -9.5%o. This temperature is comparable to thepresent-day temperature range of the sampled interval,55-60°C (based on a geothermal gradient of 4.2°C/100m, determined from temperatures measured in the bore-hole). A δ 1 8 θ composition of -6.0‰ approximates themost enriched value for the extensively calcite replacedsandstones (Fig. 15) and yields a temperature of 40°C.

A calcite replacement temperature range of 40-62° isvalid only if our assumption of a constant δ 1 8 θ com-position for the diagenetic water responsible for dissolu-tion-precipitation is correct. Interstitial water will inter-act (solution-precipitation) with the carbonate, feldspar,and clay components in the sediment at increasing tem-peratures. This interaction will shift the δ 1 8 θ composi-tion of the water toward a more enriched or positive val-ue. The magnitude of this positive δ 1 8θ shift is controlled

969


by the initial δ 1 8 θ composition of the minerals, effectivemolar ratio of the oxygen contained in the water relativeto the reactive minerals, rate of carbonate recrystalliza-tion, and fluid flow. A maximum positive δ 1 8 θ shift ofO%o to +1.3%o was documented for interstitial waterscollected at depths of 500 m sub-bottom from DSDPSites 572-574, Leg 85 (Stout, 1985). These interstitialwaters were associated with biogenically dominated sed-iments similar to the marls and limestones present inHole 603B. If a similar δ 1 8 θ enrichment trend is devel-oped in the interstitial water in Hole 603 B, the calcu-lated calcite precipitation temperatures would be too low.The alteration of basement basalts to form authigenicclays and/or zeolites can produce pore waters depletedin 18O (Brenneke, 1977). Basement lies at least 250 mdeeper in the section than the base of Hole 603B, sothat although the influence of 18O-depleted water fromthe alteration of basement basalts cannot be totally ruledout, it is considered an unlikely possibility. The vast nor-mal marine δ 1 8 θ signal contained in the sediments over-lying the basement would rapidly enrich any depleteddiagenetic water moving upward. Determination of theδ 1 8θ composition of the interstitial water at maximumdepth from this site would resolve this question. It is im-portant to note that an enrichment trend in the intersti-tial water will elevate the calculated average calcite re-placement temperatures. The oxygen stable isotope com-position of the calcite suggests that the bulk of thecalcite replacement of the matrix in these sandstones hasoccurred near present-day depths. This tentative conclu-sion assumes that past geothermal gradients were notsignificantly different from the present gradient.

The δ1 8θ composition of the total carbonate containedin the sandstones with greater than 3% visible porosityare significantly more enriched than that contained inthe nonporous sandstones (Fig. 15). This δ 1 8θ differencecan be directly related to the variation in the amounts ofauthigenic and detrital carbonate components. Calcitereplacements dominate the total carbonate in the non-porous sands, whereas detrital carbonate is the most im-portant carbonate component in the porous sands. As aresult, the whole-rock δ 1 8 θ composition of the poroussandstones is strongly biased toward a more enrichedcomposition. The amount of this enrichment is deter-mined by the amount and type of remnant carbonateskeletal material.

The majority of the carbonate samples analyzed havea marinelike δ 1 8θ composition, that is, 0 to +3.1%o(Fig. 15). However, carbon from the alteration of organ-ic material was incorporated into the diagenetically al-tered calcite component of a small number of sandstonesamples. The importance of organically derived carbonin the calcite can be related to the degree of 13C deple-tion. The more negative the δ13C composition, the larg-er the contribution of organically derived carbon to thecalcite. Irwin et al. (1977), and Pisciotto and Mahoney(1981) have discussed the importance of organic oxida-tion, sulfate reduction, fermentation, and thermocata-lytic decarboxylation on carbonate cementation process-es. In their studies, the carbon stable isotope composi-tion of the authigenic carbonate was used as a measure

of importance of these processes. Oxidation, sulfate re-duction, and thermocatalytic reactions of organic mate-rial produce 13C-depleted carbon in the form of CO2 andHCO3-. Fermentation processes produce 13C-enriched car-bon in the form of CO2 (Pisciotto and Mahoney, 1981).The negative δ13C shift observed in the samples analyzedmust be attributed to a minor contribution of carbonfrom either oxidation, sulfate reduction, or thermocata-lytic reactions involving organic components in the sedi-ment. In these sediments, sulfate reduction is consideredto be the most important of these processes.

ORGANIC GEOCHEMISTRY

Total organic carbon content averages 1.55 wt.% forfour dark gray pelagic marls ranging in depth from1423.94-1442.0 m sub-bottom. The gas chromatogra-phy/mass spectrography (GC/MS) and elemental analy-ses indicate that the type III kerogen associated with thesesediments is immature (Fig. 16). Interpretation of thevitrinite reflectance data is less clear. Based on four sam-ples, the vitrinite reflectance (Ro) is approximately 0.6%,although there is also a substantial reworked population(unshaded) shown on the histograms (Fig. 17). Workersdo not agree on the Ro level that coincides with the startof hydrocarbon generation, although generally it is inthe 0.5 to 0.7% range. This is partly a semantic prob-lem. Some workers use the term "generation" but reallymean "expulsion," and as Lewan (1985) has shown, themaximum bitumen generation does not coincide withoil expulsion. Vitrinite reflectance measurements appearto be a good index of the magnitude of thermal stressexperienced by kerogen. Vitrinite reflectance, however,may not always be a good index of the stage of petrole-

Stage of hydrocarbon generation

I Prehydrocarbon generation EHI3 Oil and gas generation

Oil-generatingtype I kerogen

I Gas generation

1.50-

1.00-

0.75-

0.50-

0.25-

0.00

Oil-generatingtype II kerogen

Gas-generatingtype III kerogen

0.00 0.06 0.12 0.18 0.24 0.30 0.36 0.42

Atomic O/C ratio of kerogen

Figure 16. Van Krevelen plot showing results of elemental analyses forfour samples of marls and marly limestones from Hole 603B. (Seealso Fig. 17).

970


0.0

1 1 -

10-

9-

8-

7 -

6-

5-

4 i

3 -

2 -

1 -i

0- P

j

603B-66-1, 77-80

1 —

i i i0.5 1.0 1.5 2.0 2.5 3.0

1 6 -

1 5 -

1 4 -

1 3 -

1 2 -

1 1 -

1 0 -

9 -

8 -

7 -

6 -

5 -A _£\ —

3 -

2 -

1 -

0-

603B-66-5, 147-150

-

r "

-

r

πη , ,

11-

10-

9 -

7 "

6 -5 -

4 -

C0

C

M

1 -

o - I

If

10.0

603B-66-3, 131-134

D0.5 1.0 1.5

i

2.0I

2.5 3.0

1 2 -

1 1 -

00

CD

O

7"

6 "

5 "4 -

3 "2 -

1 -

o -

π 603B-68-2, 57-60

r

In

π

ΛΠ,0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0

Figure 17. Vitrinite reflectance analysis data for four samples of marls and marly limestones (see also Fig. 16) indicatean average Ro of approximately 0.6%. The nonpatterned part of the histograms is interpreted as the reworked por-tion of the sample population.

um generation because the kinetics for oil generation arenot the same for all kerogens (Lewan, 1985). We believethe kerogen in these sediments has not generated expelledhydrocarbons.

Type III kerogen is more oxygen-rich than type II ke-rogen (Tissot and Welte, 1978) and therefore capable ofgenerating more organic acids and carbon dioxide to cre-ate acidic conditions. Carboxylic acids generated by thethermal maturation of kerogen are preserved as short-chain aliphatic-acid anions in the formation waters be-tween temperatures of about 80-100°C (Carothers andKharaka, 1978). This does not mean that the organicacids formed at these temperatures, but rather that theyare preserved today in the 80-100°C temperature range.At temperatures below approximately 80°C, bacteria de-stroy the aliphatic-acid anions, and at temperatures aboveapproximately 140°C, these anions are destroyed by ther-mal decarboxylation (Carothers and Kharaka, 1978). Rob-in and Rouxhet (1978) show that most of the carboxyloxygen is lost prior to the generation of hydrocarbons.This suggests that these aliphatic acids may be availableto create secondary porosity quite early. On the otherhand, carbon dioxide is generated between 50 and 150°C,

with a peak at 100°C (Hunt, 1979). If the borehole tem-perature is correct, then carbon dioxide is not near peakgeneration.

DISCUSSION

Calcite Replacement

Carbonate replacement is volumetrically the most im-portant diagenetic event affecting the sandstone units stud-ied. Ferroan calcite comprises up to 50.4% of the wholerock. It occurs as a replacement of silicate frameworkgrains and matrix material. Because of its volumetricimportance, the origin and timing of this replacementevent is of major interest. The petrographic observationsand stable oxygen isotope data suggest that calcite re-placement is an intermediate diagenetic process, followedby a dissolution event and the precipitation of minoramounts of silicates.

Source of Calcite

Dissolution of nannoplankton skeletal material is be-lieved to be the primary source for the calcite that re-places the matrix in the sandstones studied. The pre-

971


dominant carbonate component in the interbedded marlsis nannoplankton skeletal material. Dissolution of eventhe robust Coccolith skeletal forms is evident (Fig. 14).The absence of delicate nannoplankton forms suggeststhat they may have been removed by dissolution, whichfor these forms, is initiated in the water column duringsedimentation and continues during burial (Wise, 1977).The carbonate skeletal material dissolved in the watercolumn is lost as a potential source for authigenic cal-cite during burial diagenesis. The amount of calcite lostto the water column versus that dissolved during the buri-al of these Cretaceous sediments cannot be determined.Previous studies (Morse, 1977, and Wise, 1977) have sug-gested that loss of carbonate from the sediment into thewater column by diffusion or advection after burial isminimal. The abundance of partially dissolved nanno-plankton skeletal material in the marl samples analyzedsuggests that an adequate source of carbonate for lithi-fication of the marls and replacement of the clay matrixin the sands is present.

Mass Transport

The precipitation and/or dissolution of mineral com-ponents in a sediment during burial diagenesis is depen-dent not only on chemical equilibrium but also on chem-ical transport. Ions formed at sites of dissolution mustbe transported to sites of precipitation and out of thediagenetic system in order to maintain cementation anddissolution processes. Five general chemical transportmechanisms have been used to explain dissolution-pre-cipitation trends in the subsurface. Meteoric water in-flux and mineralogical dewatering arising from clay trans-formations are two of these five mechanisms that can bediscounted because of the depositional setting and lowburial temperatures. The remaining mechanisms, com-pactional dewatering (unchanneled and channeled), dif-fusion, and convective fluid flow are evaluated individu-ally.

Because this is a survey study using a limited data-base, we intend only to discuss the general transport ofcalcium as a means of delineating the most plausiblemass transport mechanism. In subsequent paragraphs,we develop the following scenario regarding the originof Ca+ + for replacement calcite. Briefly, calcium de-rived from pelagic limestones by the dissolution of nan-noplankton skeletal material was transported by diffu-sion into the associated sandstones, where it formed re-placement calcite. This process would be most effectivefor thin sandstone units adjacent to marls. Based on oursamples, the thick sandstone unit (-1234-1263 m sub-bottom) is calcitic and tight at the top and base, butfriable in the center. This friable sand may represent aturbidite-channel sandstone without matrix that was atone time lightly cemented by calcite. Some of the pri-mary porosity was retained because the transport mech-anism was not adequate for moving significant amountsof Ca+ + through a thick sandstone unit.

We believe that the bulk of the replacement calcite inthe sandstone beds was externally derived. This is sug-gested by the general lack of detrital carbonate compo-nents in the sandstones to serve as an internal carbonate

source. The high sedimentation rate for these turbiditedeposits generally precludes a significant nannoplank-ton contribution from the water column. A close associ-ation between pelagic limestones and the diagenetic cal-cite in the sandstones suggests a possible source-sink re-lationship.

The burial diagenesis of nannoplankton-rich sedimentsto form chalks has been extensively studied (see Schlang-er and Douglas, 1974; Wise 1977, for a review). Thesestudies have documented the importance of dissolution-diffusion-reprecipitation phenomena in chalks. Variabil-ity in the stability of the skeletal components containedin the carbonate oozes is the driving mechanism pro-posed for lithification (Wise, 1977). These differencesare thought to be related to variation in mineralogy, mi-nor and trace element content, organic content, ultra-structure, and morphology of tests (Wise, 1977). Obser-vations of selective removal and/or the development ofcalcite overgrowths on different species and componentsof nannoplankton skeletal material are common (Adel-seck et al., 1973; Roth et al., 1975).

Regardless of the transport mechanism, calcite solu-bility must be different between the calcareous sediments(source) and the sandstone unit (sink) in order to drivecementation and/or replacement. The calcareous sedi-ments may be more highly soluble, not only because ofthe variable skeletal stability documented in studies ofchalk diagenesis, but also because of the much highersurface-area/volume ratios of this material (Wise, 1977).

Adamson (1976) indicates that in a system of a mixedmass of crystals, the finest-particle-size group determinesthe solubility. Generally, particle solubility does not in-crease until particles of less than 0.1 µm are encountered(Adamson, 1976). This particle size is well within therange expected from the disaggregation of some cocco-liths (Wise, 1977). Between the finer and coarser crystal-line Coccolith forms (accounting for calcite overgrowthsand cementation within the limestone unit) and the ad-jacent coarser-grained sandstone units, a concentrationgradient can be established. Larger crystalline carbonatenucleation grains are likely to concentrate in the sand-stone. The replacement of these sandstone units may beself-limiting. The progressive calcite replacement of thematrix in the sandstone units from the exterior to the in-terior of the bed could significantly affect the rate oflater carbonate transport to the interior. As permeabilitydecreases in the sandstones adjacent to the source lime-stones, interior replacement would be restricted. This re-striction is related to the increase in tortuosity and de-crease in permeability and the effective cross-sectionalarea for diffusion and fluid flow in the sand undergoingreplacement. Diffusional rates will decrease with increas-ing tortuosity and lower permeabilities (Berner, 1971 and1980; Pandey et al., 1974). This self-limiting feedbackcould help to account for the distribution of replace-ment carbonate in the Hole 603B sandstone units; thinsand units are extensively replaced with no macroporosi-ty, whereas the thick sand unit is extensively replaced bycalcite only on the exterior margins.

Vertical compactional fluid flow (unchanneled) is be-lieved not to be an important mass transport mechanism

972


for carbonate replacement in these sandstones. A simplemass transport calculation helps support this conclusion.The following assumptions were made in these calcula-tions:

1. Porosity reductions of 60% occurred during bur-ial of the carbonate sediments; that is, an initial porosi-ty of 70% was reduced to 10% during burial.

2. All of the interstitial fluid was expelled verticallyupward.

3. The Ca+ + concentration in the water was 0.4 g/Kg, similar to the concentration reported by Manheimand Sayles (1974) for interstitial waters in Pacific bio-genic deposits.

4. All of the Ca+ + contained in the interstitial wateris used to form calcite in the sandstone during its verti-cal migration.

5. Replacement calcite makes up 10% by volume ofthe sandstone.

Approximately 600,000 cm3 of interstitial water con-taining 6 moles of Ca+ + would be released from 1 m3 ofcarbonate sediment during porosity reduction, and 2720moles of Ca+ + would be required to form calcite thatmakes up 10 vol.% of a 1 m3 sandstone body. Assumingperfect efficiency (i.e., all of the Ca+ + in solution goesto form calcite), it would require 450 m3 of carbonateooze to provide the Ca+ + for 1 m3 of sandstone, or a450:1 carbonate ooze/sandstone ratio. If we decreaseCa+ + transport efficiency (i.e., only a small percentageof the Ca+ + in solution is utilized for calcite precipita-tion) to more realistic values of a few percent or less, therequired carbonate ooze/sand ratio increases by almosttwo orders of magnitude, to a value of 45,000/1. Theapproximate carbonate ooze/sand ratio for the Creta-ceous sediments from Hole 603B is 9:1, at least threeorders of magnitude too small to account for the re-placement calcite. Even if one adds the upward move-ment of interstitial fluids from below the deepest corepoint (1576 m sub-bottom) to the underlying basement(1826 m sub-bottom), a total increase of only 250 m ofsection, compaction fluid transport of the carbonate isthree orders of magnitude too small to explain the dia-genetic carbonate.

Can sufficient material be transported by diffusionover short distances of several meters to account for thecarbonate replacement in the thin sandstone units? Pre-vious work (Wood and Hewett, 1982) has suggested thatdiffusional mass transport is effective only at a grain-to-grain scale. They suggest that convectional fluid flowdominates mass transport on a larger scale (Wood andHewett, 1982, 1984). Recent work by Hanor (1984) sug-gests that diffusional transport of ions over distances ofkilometers can help control the vertical chemical varia-tion of formation waters. Leythaeuser et al. (1982) haveshown that diffusion can be an effective process for pri-mary migration of gas to form commercial-sized gasfields. Diffusion of CO2 between hydrocarbons and theunderlying formation waters in oil reservoirs is thoughtto be responsible for extensive carbonate cementationin the water-saturated interval of hydrocarbon reservoirs(Rácz, 1971). At this time, the effectiveness of diffu-sional mass transport for cementation in the subsurface

is still uncertain. In order to help answer this question,particularly in reference to the interbedded Cretaceoussandstones and marls of Hole 603B, a simple diffusion-al model was constructed. This model was designed totest the possible effectiveness of diffusional transport ina marl-sandstone cycle on a meter scale. The marl wasviewed as the source of calcium, the sand as the sink forthe calcium in the form of calcite replacement or ce-ment. A constant diffusion gradient or steady-state sys-tem was assumed to be maintained between the calciumsource and sink; that is, dissolution of nannoplanktonskeletal material maintained the calcium concentrationin the marl and calcite precipitation maintained a lowerconcentration in the sandstones. The relative concentra-tion difference was established as 20 mg Ca + + /1 . Thisuses a concentration of 400 mg/1 of Ca+ + in the marls,a reasonable assumption based on Manheim and Sayles(1974), and a concentration of 380 mg/1 of Ca+ + in thesandstones. The projected Ca+ + concentration in thesandstones was chosen to reflect a reasonable differencein solubility arising from particle size differences in the0.1 µm and larger size range. Note that the work ofAdamson (1976) on particle size controls on solubilitydid not include carbonate minerals. Our value of a 5%difference in solubility should be viewed as an estimate.

The application of Fick's diffusion equations for thedetermination of diffusional fluxes is dependent on theestablished boundary conditions for the specific case be-ing studied. Variation of these boundary conditions willalways change the solution to the diffusion equation(Crank, 1956).

The boundary conditions assumed for the sandstone-marl model are:

1. The contact between the marl and sandstone unitscan be approximated as a membrane of thickness (£).

2. The faces of the membrane are maintained at aconstant but different Ca+ + concentration after steady-state conditions are established; that is, there is a con-stant concentration gradient.

3. At t = 0 the concentration in the membrane is equalto the concentration in the sink (sandstone unit).

The determination of the total mass of diffusing sub-stance (Qt) which has passed through the 1-m-thick marl-sandstone membrane in time (t) can be made by usingthe following equation derived by Crank (1956).

Qt =D(C2 - C

I H-B— 1 when / oo

where:

D = diffusion coefficient (remains constant),C2 = relative C a + + concentration in marl,C\ = relative Ca+ + concentration in sandstone,

£ = thickness of membrane,/ = time in seconds.

We will test the effectiveness of diffusional transportto provide 10,800 mg of Ca+ + per cm2 through a 1-m-thick membrane. This is enough calcium to provide 10%by volume calcite cement in a 1-m-thick sandstone unit.

973


Given that

D = l x 10-6 cm2/s (Berner, 1980),Q = O.38mgCa++/cm3,C2 = 0.40mgCa+ + /cm3,

i = 1 meter,Qt = 10,800 mgCa++/cm2 ,

it would take approximately 1.7 m.y. for enough Ca+ +

to pass through a 1-m-thick marl-sandstone contact(membrane) under steady-state conditions to replace (orcement) a 1-m-thick sandstone unit with 10 vol.% cal-cite. This is a reasonable amount of time in light of a70 Ma minimum age for these sediments. Even if our as-sumption regarding the Ca+ + concentration gradient isoveroptimistic by an order of magnitude, diffusionaltransport can account for the cementation pattern. Dif-fusion in this closely coupled source-sink system (1 m)with a constantly maintained concentration gradient po-tentially can be an effective mass transport mechanismover geologic time. Other geological studies have sug-gested that diffusion can be an important interforma-tional transport mechanism (Füchtbauer, 1972).

Fluid convection can be an important mass transferprocess in the subsurface (Peck, 1967). It can theoreti-cally increase mass transfer rates by at least several or-ders of magnitude over diffusion. This mechanism hasrecently gained attention as a process that could theoret-ically account for major mass transfer events (cementa-tion-dissolution) in the subsurface over major distances(Wood and Hewett, 1982, 1984). This process circum-vents the traditional constraint on the availability of dia-genetic fluids in unidirectional flow models—e.g., com-paction fluid drive systems. Land (1984) discusses theadvantages of convective fluid flow in reference to thediagenesis of the Frio sandstones of the Texas Gulf Coast.He, however, suggests caution in directly transferring atheoretical model to different subsurface diagenetic sys-tems. The convective fluid flow model proposed by Woodand Hewett (1982 and 1984) has not been documentedin subsurface sedimentary systems. This lack of "proofin the rocks" may only reflect inadequacies in the cur-rent data base.

The high permeabilities (1-10 darcies) and homoge-neity assumed for the theoretical sand unit used in Woodand Hewetfs model do not approximate the low perme-abilities suggested by the clay matrix and general inho-mogeneity of the Cretaceous sediments found in Hole603B. These major differences suggest that a thermallydriven convectional mass transfer system may be unim-portant in these sediments. The model, as put forth byWood and Hewett (1982), cannot in our opinion be ap-plied to these sediments to demonstrate the importanceof convectional mass transfer.

Origin of Secondary Porosity

Petrographic evidence indicates that iron-rich, replace-ment calcite in selected samples in the 1234-1263-m sand-stone interval has dissolved to form secondary macro-porosity. Interpretation of the origin of the porosity is

limited by poor core recovery and the fact that too fewsamples from this interval were examined. Although anexact estimate of the amount of calcite removed fromthe major sandstone unit cannot be made, petrographicobservations suggest that in some samples it was a sig-nificant amount (up to 22 vol.%).

Why are the thinner sandstone units apparently unaf-fected by this dissolution event? We believe that part ofthe thick sandstone unit was not effectively cemented bycalcite at the time of dissolution. This may be the resultof time and/or mass transport constraints. The cleanhigh-energy channel portion of the sandstone unit at1234-1263 m sub-bottom would have been more porousand permeable, which would have allowed for the pref-erential movement of the diagenetic fluids responsiblefor dissolution. How did these diagenetic fluids enterthe porous and permeable sand? Waters from the shalesand marls may have moved vertically along microfrac-tures and/or through the development of high pressure-gradients across the zones of low permeability.

We can only speculate that compactional waters havebeen channelized into this sandstone unit, perhaps asthe result of early lithification of pelagic limestone andthin sandstone. These units with low permeabilities wouldtend to deter vertical fluid movement. As a result, theturbidite channel sandstone may serve as the major path-way for updip compactional fluid movement.

The causative agent (or agents) for dissolution can-not be established with the current data. We are, how-ever, confident: that meteoric water, because of deposi-tional setting and history, was not responsible for thissecondary porosity. The interplay of pelagic carbonatesand the sandstone units during lithification suggest thata time-temperature-dependent interaction among theshales, limestones, and sandstones has occurred. Such arelationship has been suggested by Loucks et al. (1984)to account for the distribution of secondary porosity inthe Tertiary elastics of the Gulf Coast. The organic mat-ter associated with DSDP Site 603 Lower Cretaceous sed-iments contains immature type III structured kerogenwith a vitrinite reflectance of Ro — 0.6%. The kerogenhas not reached the hydrocarbon generation-expulsionstage, but has started to generate prehydrocarbon altera-tion products such as organic acids (Robin and Roux-het, 1978).

The key aspect of the development of this secondaryporosity is the low temperature at which it has been ini-tiated, <60°C. Support for the relatively low tempera-ture comes from the correspondence between present bore-hole temperatures and estimated temperatures of precip-itation of calcite from oxygen stable isotopes. The claymineralogy suggests that the sedimentary section has notbeen substantially hotter in the past. Clay in the clay-stones is predominantly mixed-layer, randomly interstrati-fied illite/smectite with 80% smectite layers. Based onHower et al. (1976) this corresponds to a temperature ofapproximately 57°C. If the temperature had been high-er, the smectite layers would have decreased at the ex-pense of illitic layers until they reached the limit of 20%smectite, which corresponds to 100°C according to astudy in the Gulf Coast by Hower et al. (1976). The tem-

974


perature suggested is much lower than that proposed byFranks and Forester (1984) for CO2-generated porosity,

100°C, and for organic-acid-generated porosity, >80°C(Surdam et al., 1984). We cannot bring these agents upfrom significantly deeper in the section unless we sug-gest a basement origin.

Are we seeing the initial phase of secondary porosityformation caused by the introduction of CO2 (Franksand Forester, 1984) and/or organic acids (Surdam et al.,1984) generated from the alteration of associated organ-ic material? Do heat flow or alteration products fromthe basement trigger the formation of secondary porosi-ty? Or can formation water in the less permeable marlsand sandstones be mixing with that in the major sand-stone unit, causing dissolution as generally outlined byRunnels (1969)? These questions cannot be answered atthis time.

CONCLUSIONS

A dynamic diagenetic history is recorded in the LowerCretaceous turbidite sandstones recovered from Hole603B. Pyrite, siderite, ankerite, feldspar, quartz, zeolite,and calcite have been precipitated. Iron-rich calcite re-placement of the detrital clay matrix in the sandstones isvolumetrically the most important constructive diagenet-ic event. The average calcite precipitation temperaturesrange from 40-62°C. The source of calcium carbonateis believed to be the dissolution of nannoplankton skel-etal material contained in the interbedded deep-watermarls and limestones. Transport of the carbonate fromthe source (marls and limestones) to the site of precipi-tation in the sandstone units can be accomplished bysteady-state diffusion. Unchanneled, compactional flu-id flow and thermal convection transport mechanismsare constrained by restrictions on fluid volume, perme-abilities, and vertical sediment heterogeneity and are be-lieved to be insignificant.

Secondary macroporosity is well developed in portionsof the thick sandstone unit (-1234-1263.4 m sub-bot-tom). This macroporosity was formed by the dissolutionof the iron-rich replacement calcite. The specific agentresponsible for dissolution cannot be identified, but me-teoric water, organic acids associated with soil develop-ment, or surface weathering can be excluded. Channeled,compactional fluid flow through the incompletely ce-mented turbidite channel axis is a possible answer.

The early diagenetic processes are dominated by me-chanical compaction and diffusional transport in a re-stricted system. The evolution of a restricted diageneticsystem from a cementation to a dissolution phase (sec-ondary porosity formation) suggests that the agent ofdissolution is internally evolved. If the products of or-ganic diagenesis, such as CO2 and organic acids, are re-sponsible for this dissolution, we must revise early esti-mates of the effective thermal windows for porosity gen-eration to lower temperatures. These previous estimatessuggest that temperatures of 80 °C or greater are requiredto generate and initiate the formation of significant sec-ondary porosity.

This study represents a survey of the petrographic andgeochemical analyses of a limited number of core sam-

ples from the Lower Cretaceous section of Hole 603B.We emphasize that it should be viewed as a basis for fu-ture, more detailed studies and as a forum for present-ing speculative ideas. We made no attempt to cover allaspects of the diagenesis of these sediments and hopethat this survey will stimulate further research.

ACKNOWLEDGMENTS

We would like to acknowledge the invaluable assistance of MaryAnn Holmes of the University of Nebraska, Jay Muza and SherwoodWise of Florida State University, and Janet Haggerty of the Universityof Tulsa in obtaining material and information for this study. The datafrom microprobe studies and X-ray diffractometry of clay mineralswere provided by Tom Fernalld and J. C. Hoffman, respectively. Weare grateful to R. E. Larese and M. A. Miller for critiquing this paper,to Amoco Production Company for their support in this study, and toR. C. Surdam and S. Fisher for their helpful and thought-provokingreviews of the manuscript.

REFERENCESAdamson, A. W., 1976. Physical Chemistry of Surfaces: New York

(John Wiley and Sons).Adelseck, C. G., Jr., Geehan, G. W., and Roth, P. H., 1973. Experi-

mental evidence for the selective dissolution and overgrowth of cal-careous nannofossils during diagenesis. Geol. Soc. Am. Bull., 84:2755-2762.

Bailey, E. H., and Stevens, R. E., 1960. Selective staining of K-feld-spar and plagioclase on rock slabs and thin sections. Am. Min.,45:1020-1025.

Bathurst, R. G. C , 1971. Carbonate Sediments and Their Diagenesis:Amsterdam (Elsevier).

Berner, R. A., 1971. Principles of Chemical Sedimentology: New York(McGraw-Hill).

,1980. Early Diagenesis—A Theoretical Approach: Prince-ton, NJ (Princeton University Press).

Brenneke, J. C , 1977. A comparison of the stable oxygen and carbonisotope composition of Early Cretaceous and Late Jurassic car-bonates from DSDP Sites 105 and 367. In Lancelot, Y., Siebold,E., et al., Init. Repts. DSDP, 41: Washington (U.S. Govt. PrintingOffice), 937-955.

Briskin, M., and Schreiber, C , 1978. Authigenic gypsum in marinesediments. Mar. Geol., 28:37-49.

Carothers, W. W., and Kharaka, Y. K., 1978. Aliphatic acid anions inoil field waters—implications for origin of natural gas. Am. As-soc. Pet. Geol. Bull, 62:2441-2453.

Crank, J., 1956. The Mathematics of Diffusion: Oxford (Oxford Univ.Press).

Dickinson, W. R., 1970. Interpreting detrital modes of graywacke andarkose. J. Sed. Petrol., 40:695-707.

Dickson, J. A. D., 1966. Carbonate identification and genesis as re-vealed by staining. J. Sed. Petrol., 36:491-505.

Dott, R. H., Jr., 1964. Wacke, graywacke and matrix—what approachto immature sandstone classification? J. Sed. Petrol., 34:625-632.

Franks, S. G., and Forester, R. W., 1984. Relationships among sec-ondary porosity, pore-fluid chemistry and carbon dioxide, TexasGulf Coast. In McDonald, D. A., and Surdam, R. C. (Eds.), Clas-tic Diagenesis. Am. Assoc. Pet. Geol. Mem., 37:63-80.

Füchtbauer, H., 1972. Diagenesis of arenaceous deposits. In Craw-ford, F. D. (Ed.), Arenaceous Deposits: Sedimentation and Dia-genesis: Edmonton (Univ. Alberta, Dept. of Extension), pp. 205-286.

Garrison, R. E., 1981. Diagenesis of oceanic carbonate sediments: areview of the DSDP perspective. In Warme, J. E., Douglas, R. C ,and Winterer, E. L. (Eds.), The Deep Sea Drilling Project: A Dec-ade of Progress. Soc. Econ. Paleontol. Mineral., 32:181-207.

Hanor, J. S., 1984. variation in the chemical composition of oil-fieldbrines with depth in northern Louisiana and southern Arkansas:implications for mechanisms and rates of mass transport and dia-genetic reaction. Trans. Gulf Coast Assoc. Geol. Soc, 34:55-61.

Hower, J., Eslinger, E. V., Hower, M. E., and Perry, E. A., 1976.Mechanism of burial metamorphism of argillaceous sediment: 1.Mineralogical and chemical evidence. Geol. Soc. Am. Bull., 87:725-737.

975


Hunt, J. M., 1979. Petroleum Geochemistry and Geology: San Fran-cisco (W. H. Freeman and Co.).

Irwin, H., Curtis, C , and Coleman, M., 1977. Isotopic evidence ofsource of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269:209-213.

Kastner, M., and Siever, R., 1979. Low temperature feldspars in sedi-mentary rocks. Am. J. Sci., 279:435-479.

Land, L. S., 1984. Frio sandstone diagenesis, Texas Gulf Coast: a re-gional isotopic study. In McDonald, D. A., and Surdam, R. C.(Eds.), Clastic Diagenesis. Am. Assoc. Pet. Geol. Mem., 37:47-62.

Lewan, M. D., 1985. Evaluation of petroleum generation by hydrouspyrolysis experimentation. Phil. Trans. R. Soc. London, 315:123-134.

Leythaeuser, D., Schaefer, R. G., and Yukler, A., 1982. Role of diffu-sion in primary migration of hydrocarbons. Am. Assoc. Pet. Geol.Bull., 66:408-429.

Loucks, R. G., Dodge, M. M., and Galloway, W. E., 1984. Regionalcontrols on diagenesis and reservoir quality in Lower Tertiary sand-stones along the Texas Gulf Coast. In McDonald, D. A., and Sur-dam, R. C. (Eds.), Clastic Diagenesis. Am. Assoc. Pet. Geol.Mem., 37:15-46.

McCrea, J. M., 1950. On the isotopic chemistry of carbonates and apaleotemperature scale. J. Chem. Phys., 18:849-857.

Manheim, F. T , and Sayles, F. L., 1974. Composition and origin ofinterstitial waters of marine sediments, based on deep-sea drill cores.In Goldberg, E. D. (Ed.), The Sea (Vol. 5). Marine Chemistry:New York (Wiley-Interscience), 527-568.

Maynard, J. B., 1984. Composition of plagioclase feldspar in moderndeep-sea sands: relationship to tectonic setting. Sedimentology, 31:493-501.

Morse, J. W., 1977. The carbonate chemistry of North Atlantic oceandeep-sea sediment pore water. In Anderson, N. R., and Malahoff,A. (Eds.), The Fate of Fossil Fuel CO2 in the Oceans: New York,(Plenum), pp. 323-343.

O'Neil, J. R., Clayton, R. N., and Mayeda, T. K., 1969. Oxygen iso-tope fractionation in divalent metal carbonates. J. Chem. Phys.,51:5547-5558.

Pandey, G. N., Tek, M. R., and Katz, D. L., 1974. Diffusion of fluidsthrough porous media with implications in petroleum geology. Am.Assoc. Pet. Geol. Bull., 58:291-303.

Peck, A. J., 1967. Mass transport in porous rocks. Mineralium Depos-ita, 2:243-247.

Pisciotto, K. A,, and Mahoney, J. J., 1981. Isotopic survey of diage-netic carbonates. In Yeats, R. S., Haq, B. U., et al., Init. Repts.DSDP, 63: Washington (U.S. Govt. Printing Office), 595-609.

Rácz, D., 1971. The role of diffusion transfer in the evolution of het-erogeneous reservoir systems. Chem. Geol., 8:171-196.

Reynolds, R. C , and Hower, J., 1970. The nature of interlayering inmixed-layer illite/montmorillonites. Clays Clay Min., 18:25-36.

Robin, P. L., and Rouxhet, P. G., 1978. Characterization of kerogensand study of their evolution by infrared spectroscopy: carbonyland carboxyl groups. Geochim. Cosmochim. Ada, 42:1341-1349.

Robinson, W. E., 1969. Kerogen of the Green River Formation. In Eg-linton, G., and Murphy, M. T. J. (Eds.), Organic Geochemistry:Berlin-Heidelberg-New York (Springer), pp. 619-637.

Roth, P. H., Wise, S. W , and Thierstein, H., 1975. Early chalk dia-genesis and lithification: sedimentological applications of paleon-tological approaches. IXme Congr. Int. Sedimentol., 7:187-192.

Runnels, D. D., 1969. Diagenesis, chemical sediments and the mixingof natural waters. J. Sed. Petrol., 39:1188-1201.

Schlanger, S. O., and Douglas, R. G., 1974. The pelagic-ooze-chalk-limestone transition and its implication for marine stratigraphy. InHsü, K. J., and Jenkyns, H. (Eds.), Pelagic Sediments on Landand Under the Sea: Oxford (Blackwell and Int. Assoc. Sedimen-tol.), pp. 117-148.

Schmidt, V., and McDonald, D. A., 1979. Texture and recognition ofsecondary porosity. In Scholle, P. A., and Schluger, P. R. (Eds.),Aspects of Diagenesis: Tulsa (Soc. Econ. Paleontol. Mineral.), pp.209-225.

Stach, E., 1975. Textbook of Coal Petrology: Berlin-Stuttgart (Ge-brüder Borntraeger).

Stout, P. M., 1985. Chemical diagenesis of pelagic biogenic sedimentsfrom the equatorial Pacific [Ph.D. dissert]. Univ. California, SanDiego.

Surdam, R. C , Boese, S. W , and Crossey, L. J., 1984. The chemistryof secondary porosity. In McDonald, D. A., and Surdam, R. C.(Eds.), Clastic Diagenesis. Am. Assoc. Pet. Geol. Mem., 37:127-149.

Tissot, B. P., and Welte, D. H., 1978. Petroleum Formation and Oc-currence: New York (Springer-Verlag).

Wilson, M. D., and Pittman, E. D., 1977. Authigenic clays in sand-stones: recognition and influence on reservoir properties and pa-leoenvironmental analysis. J. Sed. Petrol., 47:3-31.

Wise, S. W., Jr., 1977. Chalk formation: early diagenesis. In Ander-son, N. R., and Malahoff, A. (Eds.), The Fate of Fossil CO2 in theOceans: New York (Plenum), pp. 717-739.

Wood, J. R., and Hewett, T. A., 1982. Fluid convection and masstransfer in porous sandstones—a theoretical model. Geochem. Cos-mochim. Acta, 46:1707-1713.

, 1984. Reservoir diagenesis and convective fluid flow. InMcDonald, D. A., and Surdam, R. C. (Eds.), Clastic Diagenesis.Am. Assoc. Pet. Geol. Mem., 37:99-110.

Date of Initial Receipt: 16 April 1985Date of Acceptance: 16 December 1985

976

Documents

35. PETROLOGY AND GEOCHEMISTRY OF SANDSTONES IN THE …