46. ORGANIC MATTER IN LEG 96 SEDIMENTS: CHARACTERIZATION BY PYROLYSIS1

Jean K. Whelan and Martha Tarafa, Woods Hole Oceanographic Institution2

ABSTRACT

Sediments from Deep Sea Drilling Project Sites 615, 617, 618, 619, and 620-623 were subjected to pyrolysis. Thesediments are immature with respect to petroleum generation as determined by production index values of less than 0.1and rm a x values of 460-480°C. The amount of pyrolyzable organic matter was moderately low as compared to typicalpetroleum source rocks. The immature organic matter present does not appear to contain a significant proportion ofwoody material as shown by the low gas-generating potential. Typical overbank sediments from Sites 617 and 620 gener-ally show higher P2 values (500-800 µg hydrocarbon per g dry weight sediment) than typical channel-fill sediments fromSites 621 and 622 (P2 = 450-560 µg/g). Tmax for both types of sediment remained very constant (462-468 °C) with aslight elevation (+ 15°C) occurring in samples containing lignite. The highest P2 values occurred in sections described asturbidites. Very low P2 values (about 50 µg/g) occurred in sands. P2 values for shallower sections of basin Sites 618 and619 tended to be higher (900-1000 µg/g) and decreased in deeper, more terrigenous sections of Site 619. Preliminary ex-periments indicate that microbiological degradation of sediment organic matter causes a decrease in P2. Pyrolyzable or-ganic matter from lower fan Site 623 appears to increase with depth in two different sediment sequences (40-85 and 95-125 m sub-bottom). Organic matter type, as shown by pyrolysis capillary gas chromatography (GC) patterns, was gener-ally the same throughout the well, with much more scatter occurring in the deepest sections (130-155 m sub-bottom).One major and two minor organic matter types could be recognized in both fan and basin sites drilled on Leg 96.

INTRODUCTION

Thermal techniques are now used routinely for deter-mining the amount of petroleum produced and the pe-troleum-generation potential of source rocks (Espitaliéet al., 1977; Claypool and Reed, 1976; Hue and Hunt,1980; Barker, 1974; Dembicki et al., 1983). These tech-niques were applied in this study to immature marinesediments to examine high molecular weight organic ma-terial and to correlate pyrolysis properties with other ge-ological characteristics. The rationale for using these mea-surements on immature and surface sediments was thatpyrolysis gives a measure of hydrogen-rich (reduced) or-ganic matter in a sample (Espitalié et al., 1977). Analy-sis of this material could provide information on thesource and early diagenetic history of the sediment whichmight be a useful paleontological tool.

Hydrogen-rich organic material can be lost from sed-iments in either of two general ways: (1) thermally (inthe petroleum-generating process) or (2) by biologicaldegradation which normally accompanies sediment re-working—such as that which occurs under oxic condi-tions in the water column or at the sediment/water in-terface. Rapidly buried sediments would be removed morequickly from this oxic zone where rapid biodegradationcould occur and, therefore, should contain a higher con-tent of pyrolyzable organic matter than more slowly de-posited sediments (assuming bottom-water oxygen con-ditions remain constant). In addition, higher tempera-tures might be required to break down more reworked(hydrogen-poor) organic matter. These ideas were tested

Bouma, A. H., Coleman, J. M., Meyer, A. W., et al., Init. Repts. DSDP, 96: Washing-ton (U.S. Govt. Printing Office).

2 Address: (Whelan, Tarafa) Chemistry Department, Woods Hole Oceanographic Insti-tution, Woods Hole, MA 02543.

on Deep Sea Drilling Project (DSDP) Leg 95 sediments(Tarafa, Whelan, and Mountain, in press) and by sub-jecting sediments from DSDP Leg 96 (Gulf of Mexicoincluding the lower Mississippi Fan) to pyrolysis analy-ses as described below. In addition, preliminary labora-tory experiments were carried out to determine whethermicrobiological degradation can cause a decrease in pyro-lyzable organic matter.

METHODS

Wet sediment was ground with a mortar and pestle and 50-100 mgsamples were placed in a quartz tube, held in place with quartz wool,and pyrolyzed by procedures previously described in detail (Whelan etal., 1983). Briefly, the samples were placed in a desorption probe inthe cooled interface of a Chemical Data Systems (CDS) Model 820Geological Reaction System and heated at 40°C per minute from 200to 500° C. Organic compounds desorbed during the heating processwere swept out of the system in a helium stream. The stream was splitwith part going directly to a flame ionization detector. During the heat-ing process, the compounds evolved by heating elute as two peaks—thelower temperature peak (P{) evolving at 150-250°C represents sorbedorganic compounds naturally present in the sediment; the higher tem-perature peak (P2) evolving at 400-500° C represents compounds crackedfrom (thermally broken off) the sediment kerogen or protokerogen(high molecular weight organic) matrix. The temperature maximum ofthis peak, called Tmax, is measured via a thermocouple placed next tothe sediment sample. It should be noted that this procedure givessomewhat different absolute Tmax values than (but the same profile as)the more commonly used Rock Eval apparatus (J. K. Whelan, K. Pe-ters, and M. E. Tarafa, unpublished results). Profiles of Pu P2, andTmax as a function of depth are shown in Figure 3 for the extensivesuite of Site 623 samples analyzed in this work. Error bars show lσstandard deviations for replicate grinds. Error bars for duplicate sam-ples taken from the same grind (not shown) were generally smallerthan the width of the data points. The production index (PI), P\/(P\+ P2), is also shown. This is a parameter commonly used in petrole-um source rock analysis which gives the ratio of generated to total pe-troleum-generating potential of the sediment (Espitalié et al., 1977).

The pyrolyzed material evolved in P2 was also trapped and ana-lyzed by capillary GC in some cases, as shown in Figures 4 and 5.

Carbon content was determined by combustion, using a Leco Ap-paratus. Organic carbon was determined on ground, dried, and decar-bonated (by HC1 vapor) samples (Whelan et al., 1984).

757

J. K. WHELAN, M. TARAFA

RESULTS AND DISCUSSION

General

The production index values (PI) of sediments exam-ined in this work are all less than 0.1, which shows thatthey are immature with respect to petroleum generationand cannot contain more than traces of migrated petro-leum. The latter conclusion differs from that of Kenni-cutt et al. (this volume), who conclude that migratedhydrocarbons are generally present in fan and basin sed-iments. These authors base their conclusions on quali-tative (rather than quantitative) data consisting of sol-vent extraction hydrocarbon compositions. These solventextractable hydrocarbons correspond to the Pj compo-nents in the current work (representing levels of hydro-carbons naturally present in the sediments) which are lowand fairly constant throughout all samples ranging fromabout 10 to 20 ppm. These low levels are more consis-tent with a source from recycled terrigenous sedimentsthan from migrated petroleum. Kennicutt et al. (this vol-ume) point out that these two alternatives cannot be dis-tinguished from their data.

In contrast, P2 values, reflecting the potential of or-ganic matter in the sediment for generating hydrocar-bons if subjected to further burial and higher tempera-tures, is more variable. The P2 values are generally onthe low side for samples from all sites as compared to atypical good petroleum source rock (Table 1) rangingfrom about 50 to 800 for fan sites and from 300 to 1000for basin sites. For comparison, values from a numberof Alaskan North Slope wells, classified as having lowto moderate generating potential, fall in the 200-2000 ppmrange (Claypool and Magoon, 1985).

Fan Sites 615, 617, 620, 621, and 622

A few samples from characteristic fan sediments wereanalyzed in order to determine if pyrolyzable organicmatter (P2) and Tmax differed between channel-fill andoverbank sediments. (P\ values were usually very lowand are, therefore, generally not reported.) P2 and Tmax

values were obtained from midfan sites for representa-tive overbank sequences from Sites 617 and 620, andfrom representative channel-fill sequences from Sites 621and 622 (Table 1). In addition, a number of sampleswere also examined from lower fan Sites 615 and 623where overbank and channel sequences are not as welldefined and tend to interbed (Coleman, Bouma, et al.,this volume). Bottom sections of Site 615 represented theoldest sediments recovered on this leg. A very extensivesuite of Site 623 samples were analyzed, as discussed ina separate section.

For samples containing only silt and mud (and nosand), the midfan overbank sediments (indicated as OBfrom Sites 617 and 620) generally show significantly high-er P2 values (540 to 800 µg/g or ppm) than the midfanchannel-fill sequences (P2 = 450-560 µg/g from Sites621 and 622 indicated by CF in Table 1). Tmax values forboth types of samples are very constant (462-468 °C).Exceptions appear to occur in samples containing smallamounts of lignite, such as Sample 622-6-1, 50-51 cmwhich shows Tmax to be higher by about 15°C with more

scatter in replicate values. This sample also shows a slight-ly lower P2 (possibly resulting from greater reworking ofthe organic matter) than found for the other two sam-ples from this site.

In addition, P2 also appears to vary with grain sizewith sections containing sands showing much lower val-ues than those with only mud or silty mud (Table 1). Forexample, Samples 615-3-3, 8-9 cm, and 615-5-3, 99-100 cm, which are sands or silty sands, both show verylow P2 values (57 and 63 µg/g, respectively), as com-pared to Samples 615-3-2, 30-31 cm and 615-3-3, 63-64 cm, which are muds or silty muds and have P2 valuesalmost an order of magnitude higher (637 and 624 µg/g,respectively). This difference is much larger than thatbetween channel fill and overbank sites which was de-scribed above. No consistent difference is observable be-tween muds and silts. For example, P2 values for over-bank samples from Site 620, which are generally homo-geneous clays, show almost the same P2 values asoverbank sediments from Site 617, which generally con-tain a small amount of silt as well. Tmax values for all ofthese sections remain very constant (except for those con-taining lignite as discussed above).

Several slightly elevated P2 values appear to correlatewith sections described as turbidites. Examples includeSamples 615-33-1, 46-47 cm (P2 = 773 µg/g as com-pared to values for silty muds in the rest of the hole ofabout 600) and Sample 617-5-4, 49-50 cm (867 µg/g ascompared to other Site 617 values generally around700 µg/g). These examples are similar to results fromLeg 95, where it was found that high P2 values could beused to identify turbidites or slumps in some cases (Tara-fa et al., in press).

It should also be noted that P2 values tend to de-crease somewhat (compared to the two samples just dis-cussed) at the bottom of Site 615 (below 485 m) in amore marine section containing redeposited carbonates(Kohl et al., this volume).

Sites 619 and 618

Sediments from the two basin sites which contain morehemipelagic sediments than the fan sites were also inves-tigated (Table 1 and Figures 1 and 2). P2 values for shal-lower sections of both these holes are generally higherthan for any of the Mississippi Fan sections. For exam-ple, values range from about 900 to 1000 µg/g from 28to 73 m sub-bottom at Site 619 with the surface (4 m)and deeper sections (below 78 m) having values morecomparable to those of the fan sediments. The deeperlower values (generally less than 500 µg/g below 110 ft.sub-bottom) correlate with sediments in which methanegas (and possibly microbiological methanogenic activi-ty) was found (Whelan et al., this volume). However,some of the sections containing particularly low P2 val-ues as compared to other values for this site (such asSamples 619-13-4, 130-150 cm and 619-17-3, 130 cm withP2 values of 414 and 343 µg/g, respectively) were alsoburrowed or bioturbated (Table 1, Bouma, Stelting et al.,this volume).

The deeper sections of Site 619 below 100 m sub-bot-tom, where P2 values decrease, may also contain a larger

758

ORGANIC MATTER IN LEG 96 SEDIMENTS

Table 1. Summary of pyrolysis data and lithology for Leg 96 samples.

Core-Section(level or interval in cm)

Hole 615

3-2, 30-313-2, 70-713-2, 723-2, 933-3, 8-93-3, 63-645-3, 99-1005-4, 40-416-46-5, 120-1216-6, 10-119-1, 779-3, 0-210-1, 10-1110-1, 58-5933-1, 46-4747-2, 42-4349-1, 2-350-1, 3-451-1, 5-6

Hole 617

1-3, 23-243-1, 28-295-4, 49-508-1, 49-5011-4, 41-42Average

Sub-bottomdepth (m)

13.9014.3014.3214.5315.1815.7333.5934.5043.6046.3046.7068.0070.2076.8077.28

305.16458.62485.22494.73504.25

3.2317.8841.7965.7999.01

Type3

OBOBOBOBOB

P2

h'c

(fg/g)

637 ± 91571 ± 331

10024557 ± 2

624 ± 363 ± 1.7

241 ± 110783668 ± 471139 ± 15474 ± 11276701 ± 89177 ± 13773 ± 83600 ± 42437 ± 52575 ± 7.6599 ± 40

701 ± 56678 ± 59867 ± 37695 ± 31540 ± 92696 ± 116

^max

463470475455454463464470467463464471478461470

468.5471.5

467467

464.5

465462

462.5463.5

462

b

±

±

±±±±±±±±±±±±±±+

±

±±±±±

(°O

2.8

1.4

122.89.26.412.84.2751.43.50.712.12.82.86.4

1.41.43.52.85.2

Description of sediment

Mud with silt; laminatedDark grey sandy silt; laminatedDark grey sandy silt; laminatedDark grey sandy silt; laminatedSilty sand — fining upward sequence?Silty mudSilty sandMud with deformed silty layersDark grey silt-sand (drilling disturbance)Muddy silt — fine-grained turbiditeSorted silty sandSilt laminated mudSilty sandSilty mudSilty sand (with chert and lignite)Silty turbidite mudSilty mud in layers with ligniteNannofossil ooze (homogeneous)Nannofossil oozeForaminifer nannofossil ooze

Rapid nephloid sedimentationSlump on back side of leveeSilty mud turbidite on leveeMud with silt laminaeMud with silt laminae; fine-grained turbidite

Hole 618

10-2, 40

Hole 618A

1, top1-2, 131-4, 130-1501-5, 130-1502-5, 0-203-4, 2-80

Hole 619

1-2, 131-2, 130-1504-5, 130-1505-4, 130-1506-5, 130-1507-5, 1298-4, 130-1509-3, 9510-4, 12411-3, 130-15012-3, 10513-4, 130-15014-2, 4416-4, 8317-3, 130-15019-2, 120-150

Hole 620

2-1,41-427-1, 50-51Average

81.00 591 ± 48 ndα

9.2010.8215.0016.5034.0040.00

2.634.00

28.0036.0047.0057.0065.0072.6578.0092.00

102.00113.00119.00141.00150.00167.00

3.4151.50

OBOB

1061500930625462702

401523929

16721027872

1033939744731540414688444343419

689587638

±

±±±±

±±±±±±±±±±±±±±±

±±±

23

1398251273

2061292195927311057457376025553028

1032772

nd480 ± 2.8

ndndndnd

ndndndndndndndndndndndndndndndnd

464 ± 6.4468 ± 4.2

Dark grey homogeneous (?) mud

Dark green-grey homogeneous mudDark green-grey mudVery dark mud — gassyVery dark mud — gassyDark grey mud with black blebsDark grey gassy mud — grey blebs

Dark grey mud — bioturbated — thin silt laminaeDark grey mud — bioturbated — thin silt laminaeLaminated dark mud with streaksDark grey laminated mudDark grey homogeneous mudDark grey mud with laminae?Dark grey laminated mudDark grey laminated mudDark grey laminated gassy mudDark grey laminated mudDark grey homogeneous gassy mudDark grey burrowed mudDark grey homogeneous mudDark green-grey mud — silt laminated?Bioturbated mud — reworked clastsMud with silt and sand layers

Clay (homogeneous)Clay (drilling disturbance)

Hole 621

2-1, 50-515-1, 50-517-1, 50-51Average

Hole 622

2-1, 50-516-1, 50-518-1, 50-51Average

4.0032.8046.70

4.0042.2061.30

CFCFCF

CFCFCF

457497541498

543477545522

±±±±

±±±±

42323542

16233539

466468468

464482466

±±±

±±±

1.41.49.2

2.88.42.8

Clay with dark organic-rich zonesDark grey mud (homogeneous)Dark grey mud (gas pockets)

Dark grey mud — laminae?Dark grey — lignite-gasDark grey — homogeneous

[ O B = overbank; CF = channel fill as identified by Stelting et al. (this volume).lσ standard deviations are shown.

c µg hydrocarbon per g dry wt. sediment.nd = not determined.

759

J. K. WHELAN, M. TARAFA

40 60 80 100 120 140

Sub-bottom depth (m)160 180 200

Figure 1. A. Pyrolysis P2 yields for Hole 619A (µg hydrocarbon/g drywt. sediment). B. Pyrolysis P2 yields for Hole 619 (mg hydrocar-bon/g organic carbon).

1200

1000-

^ 800

600

ü 400

200-

I -

I I

10 20 30 40 50 60 70

Sub-bottom depth (m)

80 90 100

Figure 2. Pyrolysis P2 yields for Site 618 (µg hydrocarbon/g dry wt.sediment).

proportion of terrigenous organic matter as suggestedby increases in terrigenous over marine lipid biomarkers(such as lower n-Cl7/n-C29 ratios and higher odd-evenpreference index [OEPI] values, Requejo et al., this vol-ume). This decrease in P2 is not accompanied by a corre-sponding decrease in organic carbon which remains rel-atively constant at about 0.7-0.8% throughout the hole(Whelan et al., this volume). To test this point further, aplot of P2 normalized to organic carbon is shown in Fig-

ure IB and, except for one elevated point at 119 m, alsoexhibits a sharp decrease below 100 m sub-bottom.

Site 618 sediments produce more scatter in P2 values(Table 1 and Fig. 2). For example, an exponential^) de-crease in P2 values occurs between 15 and 16.5 m, eventhough these two samples are from the same core, whichvisually consisted of the same homogeneous dark graygassey mud.

Two of the shallowest Site 618 sections with P2 valuesof 930 and 1060 µg/g are comparable to those of thericher sections from Site 619. The rapid P2 decrease be-low 15 m may be real and related to the sharp increase incore gas microbiological methane which occurs in thesame interval (Whelan et al., this volume). Alternative-ly, this scatter could be a result of sediment heterogene-ity caused by surface slumping (see Site 619 chapter).

Effect of Microbiological Degradation on P2

Some very preliminary experiments were carried outto test effects of microbiological degradation on P2. Analiquot of sediment from the top of Site 618 was allowedto stand in a covered beaker at room temperature on thebench top for several months. A portion was analyzedperiodically for P2 as shown in Table 2. It can be seenthat a significant decrease in pyrolyzable organic matteroccurred over the course of the experiment with the larg-est decrease occurring during the first 7 days of the ex-periment. Because of the preliminary nature of the ex-periment, no attempt was made either to exclude air fromthe sediment or to mix it well with air to avoid develop-ment of anaerobic zones. Thus, the organisms causingthe P2 decrease could have been either aerobes or anaer-obes or a combination of both. Experiments are now inprogress to see if similar decrease can be observed understrict anaerobic conditions. For purposes of the currentwork, the experiment does show that microbiological deg-radation either at or shortly after sediment depositioncould have caused the observed changes in P2 in bothLeg 96 sediments, as discussed above, and in Leg 95sediments (Tarafa et al., in press).

Site 623Results from pyrolysis of Site 623 samples are shown

in Table 3 and Figure 3. There appears to be a small de-crease in P2 with depth from 10 to 25 m. A gradual in-crease then occurs from about 40 to 85 m sub-bottomwhere error bars on replicate samples (separate grinds,see Methods section) are approximately the width of the

Table 2. Changes in pyrolyzable carbon (P2) and Tmax in sedi-ment (618A, top) left standing in covered container at roomtemperature for time indicated.

Time (days)

07

3876

137

P2 ± \σ standard deviation(µg hydrocarbon/g dry wt. sediment)

607 ± 10 (duplicate samples)503482 ± 18 (triplicate)396 ± 13 (duplicate)410 ± 5 (duplicate)

1 max <• M

456-461n.d. a

466-470n.d.

470-473

a n.d. = not determined.

760

ORGANIC MATTER IN LEG 96 SEDIMENTS

Table 3. Summary of pyrolysis Pi capillary GC patterns and percentage organic carbon for Leg 96 sediments.

Core-Section(level or interval in cm)

Sub-bottomdepth (m) Organic C (%)

P2 capillaryGC pattern8 PI

Hole 623

2-2, 1152-3, 453-1, 45-473-1, 1153-5, 455-2, 1156-3, 1157-2, 428-2, 699-2, 4310-3, 3-611-3, 1-312-1, 1-3712-3, 91-9312-4, 1712-4, 11013-1,714-2, 9814-3, 5015-1, 17-1915-1, 9816-1, 4516-1, 11516-2, 4516-3, 4517-1, 4517-1, 11517-2, 115

Hole 615

3-2, 723-2, 936-49-1, 779-3, 0-2

Hole 618A

1-2, 13

Hole 619

1-2, 13

Hole 621

21-1

9.210.0n.d.16.222.63849.2

66.475.786.497.0

103105.6107.3108.2112.2123.9124.9131131.8140.6141.2142143.6150150.6151.4

14.3214.5343.668.070.2

10.83

2.63

131

n.d.c

n.d.1.311.31n.d.1.06n.d.1.12n.d.1.03n.d.0.97n.d.0.86n.d.n.d.1.161.27n.d.n.d.n.d.n.d.0.75n.d.n.d.n.d.n.d.0.67

n.d.n.d.n.d.n.d.n.d.

n.d.

n.d.

0.86

AAn.d.AAAAn.d.ABAAA

n.d.AAAABCDAEDAAAA

En.d.EAA

B

A

0.0340.034n.d.0.0450.0620.0440.039n.d.0.0320.0550.0290.0400.058n.d.0.0260.0400.0290.0230.0490.1030.0580.0250.0130.0770.0210.0420.0560.030

0.0120.120.019

0.025 ± 0.0060.053

0.028

0.048

479556n.d.485351425440n.d.545662481350471n.d.4734236146648055116145767993673443306516

100245783

474 ± 11276

500

401

470469n.d.462470465466n.d.464464472472464n.d.468468468462468475479462478471469464470468

475 ± 1.4455 ± 0467 ± 1471 ± 7478 ± 5

480 ± 2.8

468 ± 3.5

n.d. n.d. n.d.

j* See Figure 2.ng hydrocarbon per mg ash wt.

c n.d. = not determined.

data points. At 85 m, the large error bar for P2 suggestsan increase in organic matter heterogeneity. From 95 to125 m sub-bottom, the P2 values undergo a second in-crease, very similar in magnitude and slope to that foundin the 40-85 m interval. Below 125 m, a second drop inP2 occurs and much more scatter is found in the P2 val-ues from the deeper samples.

The hydrocarbons evolved during pyrolysis were alsotrapped and the concentrations of individual pyrolysisproducts measured. This analysis has been carried outfor Site 623 sediments as shown by the different capil-lary GC P2 patterns (A through E) of pyrolyzed organicmatter shown in Figure 4; these patterns correspond tothe different symbols shown in Figure 3. The total C7 toC25 compound P2 depth profile (shown as "tot. cap." inFig. 3) is almost identical in shape and the concentra-

n.d.

tions are almost equal to the total pyrolysis P2 values(which represent Q to about C32) described above. There-fore, most of the hydrocarbons evolved in the crackingprocess are in the C7-C25 range rather than in the gas(Q-Q) range, because these lighter compounds are nottrapped and analyzed by the capillary GC. The lack ofthe light hydrocarbon generation potential implies thatthe organic matter in these sediments, while being fairlylean in terms of hydrocarbon-generating potential, is moreprone to oil than to gas generation (Dembicki et al.,1983). This indicates that these sediments do not con-tain a large proportion of immature terrigenous woodyorganic matter, which tends to be more gas (methane)prone (Tissot and Welte, 1978; Hunt, 1979).

Two intervals show approximately linear increases inP2 with depth (from 40 to 85 m and 95 to 125 m sub-

761

J. K. WHELAN, M. TARAFA

(µg HC/g (µg HC/gdry wt. sediment) dry wt. sediment)

0 50 0 190 570 0 .06 .12 462 476

Tot. cap.(µg HC/g

dry wt. sediment)

800

5 0 -

100H

150H

•H

M

®B D C

i r

•I T T

OE

Figure 3. Pyrolysis data for DSDP Leg 96, Site 623. Letters (A-E) and corresponding symbols refer to dif-ferent organic carbon types identified by typical P2 capillary GC patterns as shown in Figure 4. Tot.cap. is total capillary GC P2 values.

bottom). Preliminary shipboard data (lithology and gam-ma-ray logs) suggested that these intervals might be two"fining upward" sequences of channel-overbank depos-its—from 65 to 83 m and from 92 to 120 m sub-bottom.However, sedimentological and well-logging data suggestthat complex interbedding of channel-fill and overbanksequences make it difficult to distinguish these sequenc-es on the lower fan sites such as 623. A trend of increas-ing P2 with depth within a fining-upward channel-over-bank sequence would not be expected. If P2 values arecorrelatable with organic carbon content, as has beenobserved for some immature marine sediments (Whelanand Hunt, 1982; Jasper et al., 1984), then P2 values wouldbe expected to be higher for the fine-grained sedimentsat the top of the sequence than for the coarser grainedsediments at the bottom, as is generally observed for re-deposited organic carbon.

In deeper sections of Site 623 from 126 to 160 m sub-bottom, P2 values show more scatter and no apparenttrend with depth (Fig. 3). The capillary GC patterns ofthese pyrolysis products also show more scatter in or-ganic carbon types in this part of the hole as comparedto shallower sections (Fig. 4). For example, types A andB in Figure 4 appear to have very similar compositions,with B being richer. Note that the richer type B organicmatter appears at the bottom of both the postulated chan-nel-overbank sequences (Samples 623-9-2, 43 cm and623-14-3, 50 cm at 75.7 and 124.9 m, respectively). How-ever, the deeper 126-156 m section contains leaner P2

types C and D as well as a richer type E. Note that thericher type, E, is very similar to that found in Sections615-3-2 and 615-6-4 (F and G in Fig. 4). The predomi-nant organic matter type at Site 623, type A in Figure 4,

is also seen in two samples from Site 615 (Samples 615-9-1, 77 cm, Fig. 4H, and 615-9-3, 0-2 cm at 68 and70.2 m, respectively) as well as in the surface section ofone of the basin sites (Sample 619-1-2, 13 cm). The cap-illary P2 GC pattern of pyrolyzed organic matter in asurface section of a more anoxic basin site, Sample 618A-1-4, 130 cm (Fig. 51), is richer than A and closely re-sembles type B (the type detected at the bottom of thetwo postulated channel-overbank sequences, as discussedabove).

The scatter in organic matter types in the deepest sec-tions of this hole also correlates with generally higher andbroader Tmax values (Fig. 3). In petroleum source rocks,the more hydrogen-rich sediment samples just above orentering the petroleum-generation zone give a sharp Tmax

(generally in the range of 480-520°C on the CDS instru-ment used in this work) with the width of the peak, asshown by the error bars, being in the range of 2-5°Cmaximum. The rmax values then generally increase withdepth and show a broader temperature range as petrole-um is generated and the remaining pyrolyzable groupsbecome more difficult to break off. A broad Tmax canalso indicate a mixture of organic carbon types, result-ing in a mixture of reworked and unreworked material.

The TmaK values for the Site 623 samples are low andshow a broad peak range (Fig. 3) as might be expectedfor somewhat organic-lean immature sediments contain-ing organic matter from more than one source. The broad-ness of the Tmax peaks and the scatter in the values ismuch more pronounced in the deepest section (126-156 msub-bottom) suggesting an increased complexity in or-ganic sources—consistent with increased complexity ofP2 capillary GC patterns (as discussed above).

762

ORGANIC MATTER IN LEG 96 SEDIMENTS

Paleontological data can be used together with thepyrolysis data to constrain the possible organic sources.Below 40 m sub-bottom, planktonic fauna are rare. Mostof those present are reworked Cretaceous benthic fora-minifers (see Site 623 chapter, this volume), consistentwith our relatively low P2 values which indicate the pres-ence of reworked organic matter. Radiolarians are foundin the 0-40 m section, suggesting deposition of marinematerial and generally lower sedimentation rates. Thepyrolysis data do not show a clear difference betweenthe 0-40 m as compared to the 40-160 m interval sedi-ment, although there is possibly a trend to slightly high-er P2 and Tmax values in the surface sections. The sedi-ments from 126 to 156 m sub-bottom were not analyzedin detail by the paleontologists, so there is no paleonto-logical data to test our hypothesis of more mixed organ-ic sources for this deeper section.

Interpretation of Lower Fan Sites 623 and 615Pyrolysis Results

Sedimentological data suggest that complex interbed-ding of channel-fill and overbank sequences may be typ-ical of lower fan sites and that interpretation of the twoP2 increases at Site 623 as channel-fill sequences is anoversimplification. Reexamination of the pyrolysis datataking this complexity into account provides the follow-ing information:

1. Some process(es) of unknown nature are causingan overall increase (preservation?) of pyrolyzable organ-ic matter within the depth intervals from 40 to 85 m andfrom 95 to 125 m sub-bottom.

2. The nature (but not the amount) of high molecu-lar weight organic matter (as determined by pyrolysiscapillary GC—Figs. 3, 4) remains fairly constant through-out the hole, except in sections below 125 m. This resultimplies either that the source of high molecular weightorganic matter remains the same but is delivered in vary-ing amounts throughout the hole (which seems unlikely),or that processes (biological?) reworking the sedimentorganic matter during deposition have remained fairlyconstant (which seems more likely based on lipid bio-marker data—from many geographic areas—Ourisson etal., 1984).

Based on the very limited set of midfan overbank andchannel-fill test samples described previously (Table 1),ranges (to be tested in future work) can be proposed forP2 as related to depositional setting in fan sequences.For immature sediments containing only mud and siltand rmax values in the range of 462-468 °C (indicatingabsence of lignite and/or more mature types of carbon),P2 values of less than 500 µg hydrocarbon per g dryweight sediment are suggested to be typical of channel-fill sequences, while P2 values greater than 600 representoverbank deposits, slumps, or turbidites. Furthermore,because very low P2 values occur for sands (< 300 µg/gand often < 100), it is suggested that sections containingsand which also show very high P2 values (>500) havebeen mixed with finer-grained unreworked organic-richsediments. Such mixing might be expected from debrisflows and might also result from drilling disturbance.Slumps or turbidites were postulated to be the cause of

abnormally high P2 values in some Leg 95 sediments(Tarafa et al., in press). Using these criteria, the sedi-ments from lower fan Sites 615 and 623 can be classifiedas channel-fill (CF), overbank (OB), or slumps (labeledS, includes turbidites) on the basis of pyrolysis as shownin Table 4. It will be interesting to test in future workwhether these classifications are realistic in terms of othergeological parameters.

CONCLUSIONS

The conclusions reached from pyrolysis analyses ofLeg 96 sediments can be summarized as follows:

1. For fan depositional environments, limited and pre-liminary data suggest that P2 values of less than 500 µg/gare typical of channel fills; values of greater than 600 µg/gare typical of overbank, slump, or turbidite deposits withthe highest values occurring in the latter two types.

2. Microbiological degradation of sedimentary organ-ic matter caused a decrease in P2.

3. The quantity of pyrolyzable organic matter at fansites is generally low with values in some basin sectionsbeing somewhat higher. The production index and Pvalues are low with respect to petroleum generation andare not typical of sediments containing migrated hydro-carbons.

4. Capillary GC analyses of the pyrolyzable organicmatter shows the type (but not the amounts) to be verysimilar at lower fan Site 623 from 0 to 126 m sub-bot-tom. Higher yields of pyrolyzable organic matter are pos-tulated to correlate with depositional history. Specifically,more rapid deposition of organic matter would provideless opportunity for oxic biodegradation of sediment or-ganic matter in the water column assuming bottom-wa-ter oxygen conditions remained constant.

5. The composition of pyrolysis products suggests anorganic matter type that is more prone to oil than to gasgeneration. This implies that the immature organic mat-ter in these sediments does not contain a significant quan-tity of woody terrigenous material.

6. rmax values are very constant for most sediments.The generally low, broad Tmax peaks suggest the pres-ence of immature organic matter, possibly reworked bybiogenic processes, before deposition. Elevated Tmax val-ues appear to occur in samples containing lignite.

7. P2 capillary GC patterns from samples taken fromthe bottom sediments of Site 623 (126-156 m sub-bot-tom) show a mixture of organic matter types.

8. Most of the organic matter types found at Site 623,as identified by pyrolysis capillary-GC of P2 could alsobe detected at other Leg 96 sites from which data wereavailable at time of this writing (i.e., from Site 615 onthe lower fan, and Sites 618 and 619 located in the in-traslope basins).

9. A decrease in P2, which occurs in deeper sectionsof Basin Site 619, may be related to an increase in terrig-enous material and/or to increased bioturbation.

ACKNOWLEDGMENTS

We thank the crew and staff of DSDP for samples, Christine Bur-ton and Joan Brazier for organic carbon analyses, Richard Sawdo forkeeping our instruments running, and John Farrington for his helpfulcomments about the manuscript. K. Peters and an anonymous review-

763

J. K. WHELAN, M. TARAFA

623-17-1, 115 cm

623-15-1, 17 cm

co(J

z

D

jhil l ' R

ON

t= Φ

m+

p

Xyl

to-

Xyl

er

ILc Φ σ>α> c o

9

o

NC

1JC

10

]

CM

δ

L

JC1

2!

δz

JC1

4i

^.

δz

δz

623-15-1,

JC15

j t.

slC

17

CO

δz

O

ü

z

98 cm

_*->'—

JC18

slC

20

623-16-1, 115 cm

Figure 4. (A-E) Typical P2 capillary GC patterns for sediment types A through E recognized at Site 623. (F-I) P2 capillary GC patterns for Leg 96sediments other than those from Site 623.

er reviewed an earlier version of this manuscript. This work was sup-ported by National Science Foundation grant OCE82-00485. WoodsHole Oceanographic Institution Contribution No. 6028.

REFERENCES

Barker, C , 1974. Pyrolysis techniques for source-rock evaluation. Am.Assoc. Pet. Geol. Bull., 58:2349-2361.

Claypool, G. E., and Magoon, L. B., 1985. Comparison of oil-sourcerock correlation data for the Alaskan North Slope: techniques, re-sults, and conclusions. In Magoon, L. B., and Claypool, G. E.(Eds.), Alaska North Slope Oil/Source Rock Correlation Study.Am. Assoc. Pet. Geol., Spec. Stud. Geol., 20:49-81.

Claypool G. E., and Reed, P. R., 1976. Thermal-analysis techniquefor source-rock evaluation: quantitative estimate of organic rich-ness and effects of lithologic variation. Am. Assoc. Pet. Geol. Bull.,60:608-626.

Dembicki, H., Jr., Horsfield, B., and Ho, T. T. Y., 1983. Source rockevaluation by pyrolysis-gas chromatography. Am. Assoc. Pet. Geol.Bull., 67:1094-1103.

Espitalié, J., Madec, M., Tissot, B., Mennig, J. J., and Leplat, P.,1977. Source rock characterization method for petroleum explora-tion. Proc. Annu. Offshore Technol. Conf., 9(3):439-444.

Hue, A. Y., and Hunt, J. M., 1980. Generation and migration of hy-drocarbons in offshore South Texas Gulf Coast sediments. Geo-chim. Cosmochim. Acta, 44:1081-1089.

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

Jasper, J. P., Whelan, J. K., and Hunt, J. M., 1984. Migration of Cto Cg volatile organic compounds in sediments from the Deep SeaDrilling Project, Leg 75, Hole 53OA. In Hay, W. W., Sibuet, J . -C,et al., Init. Repts. DSDP, 75: Washington (U.S. Govt. Printing Of-fice), 1001-1008.

Ourisson, G., Albrecht, P. A., and Rohmer, M., 1984. The microbialorigin of fossil fuels. Scient. Amer., 251(8):44-51.

Tarafa, M. E., Whelan, J. K., and Mountain, G. S., in press. Sedi-ment slumps in the middle and lower Eocene of Deep Sea DrillingProject Holes 605 and 613: Chemical detection by pyrolysis tech-niques. In Poag, C. W., Watts, A. B., et al., Init. Repts. DSDP,95: Washington (U.S. Govt. Printing Office).

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

Whelan, J. K., Fitzgerald, M. G., and Tarafa, M., 1983. Analysis oforganic particulates from Boston Harbor by thermal distillation-pyrolysis. Environ. Sci. Tech., 17:292-298.

764

ORGANIC MATTER IN LEG 96 SEDIMENTS

615-9-1, 77 cm

Figure 4. (Continued).

Whelan, J. K., and Hunt, J. M., 1982. Organic matter in Deep SeaDrilling Project Site 504 and 505 sediments studied by a thermalanalysis-gas chromatography technique. In Cann, J. R., Langseth,M. G., Honnorez, J., Von Herzen, R. P., White, S. M., et al., Init.Repts. DSDP, 69: Washington (U.S. Govt. Printing Office), 443-450.

Whelan, J. K., Hunt, J. M., Jasper, J., and Hue, A., 1984. Migrationof Cj-Cg hydrocarbons in marine sediments. Org. Geochem., 6:683-694.

Date of Initial Receipt: 4 March 1985Date of Acceptance: 6 September 1985

765

J. K. WHELAN, M. TARAFA

Table 4. Prediction of channel-fill (CF) versus overbank (OB) versus turbidite or slump (S) deposits based on pyrolysis data.

Core-Section(level or interval in cm)

Sub-bottomdepth (m) Description of sediment

Predicted1-type

Hole 615

3-2, 30-33

3-2, 70-77

3-2, 72

3-2, 93

3-3, 8-93-3, 63

5-3, 99-100

5-4,406-46-5, 120

6-6, 109-1, 77

9-3, 0-2

10-1, 10

10-1, 5833-1, 46

47-2, 42

49-1, 2-3

50-1, 3-4

51-1, 5-6

Hole OZJ

2-2, 115

2-3, 45

3-1, 115

3-5, 45

5-2, 115

6-3, 115

8-2, 69

9-2, 43

10-3, 3-6

11-3, 1-312-1, 1-3

12-4, 17

12-4, 110

13-1

14-2, 98

14-3, 50

15-1, 17

15-1,9816-1, 45

16-1, 11516-2, 45

16-3, 45

17-1,4517-1, 115

17-2, 115

13.90

14.30

14.32

14.53

15.1815.73

33.59

34.50

43.60

46.30

46.70

68.00

70.20

76.80

77.28

305.16

458.62

485.22

494.73

504.25

9.210.0

16.2

22.6

38.0

49.2

66.4

75.786.4

97.0

103.0

107.3

108.2

112.2

123.9

124.92

131.0

131.8

140.6

141.2

142.0

143.6

150.0

150.6151.4

6406001002

455762463241783700139474276700177770600440575600

4795564853514254405456624813504714734236146648055116145767993673443306516

463470475455454463464470467463464471478461470468471467467465

470469462470465466464464472472464468468468462468475479462478471469464470468

Mud with silt laminaeDark grey sandy silt — laminatedDark grey sandy silt — laminatedDark grey sandy silt — laminatedSilty sand — fining upward sequence?Silty mudSilty sandMud with deformed silty layersDark grey silt — sand (drilling disturbance)Muddy silt — fine-grained turbidites?Sorted silty sandSilty laminated mudSilty sandSilty mudSilty sand (with chert; lignite)Silty turbidite mudSilty mud in layers with ligniteNannofossil ooze (homogeneous)Nannofossil oozeForaminifer nannofossil ooze

Mottled mud with minor silt laminaeMottled mud with less color bandingMottled, no bands — silt scour basesMottled, no bands — scour basesMottled — faint color bandingDark grey mud, silt laminaeMud with silt laminae, color bandsInterbedded mud and siltContorted mud and dark grey siltHomogeneous mud with silt blebsDark grey mud with silt laminaeDark grey mud with silt laminaeDark grey mud with silt laminaeDeformed mixture silt, sand and mudColor-banded dark grey mud with silt laminaeColor-banded/dark grey mud/between two sand layersDark grey silt — drilling disturbanceDark grey silt — drilling disturbanceDark grey mud with silt-sand laminaeMassive dark grey siltMassive dark grey siltDark grey mud with silty-sand laminaeDark grey mud — drilling disturbanceDark grey mud — drilling disturbanceDark (no sand?) — drilling disturbance

OB or SSS

CFCF

OB or SCFCF

OB or SOB or S

CFCFCF

OB or SndS

OB or SMarineMarineMarine

CF?

CFCFCFCF?

OB or SCFCFCFCFCFS

OB or SOB or SCFCFCF

OB or SCFSCFCF9

µg hydrocarbon per g dry wt. sediment.1 OB = overbank; S = slump or turbidite; CF = channel-fill. "Marine" means sediment probably deposited under more marine rather than fan depositional condi-tions. ? means not classifiable on basis of 7*2

766