25. CHEMICAL AND MINERALOGICAL STUDIES, SITES 322 AND 325

Edward A. Perry, Jr., Edward C. Beckles, and Robert M. Newton,Department of Geology, University of Massachusetts, Amherst, Massachusetts

INTRODUCTION

The results in this report represent part of a combinedgeochemical research program into the diagenesis andgeochemical evolution of deep-sea sediments. Otherworkers have determined pore water chemistry(Gieskes, this volume) and isotopic compositions(Anderson and Lawrence and Lawrence et al., both thisvolume), and our complementary studies detail the sedi-ment mineralogy and chemistry at Site 322, the sedimentmineralogy at Site 325, and the alteration products ofthe basal basalts encountered at Site 322.

EXPERIMENTAL METHODSThe sediment samples were repeatedly washed in dis-

tilled water until flocculation no longer occurred andthen separated into the < 1 µm and > 1 µm fraction bycentrifugal sedimentation. The >l µm fraction wasdried, gently ground in acetone, and analyzed on a glassslide by X-ray diffraction (XRD). Oriented XRDspecimens of the <l µm fractions were prepared byallowing a distilled water suspension to dry on a glassslide at room temperature.

The determination of clay mineral abundances byXRD is notoriously unreliable because of variations inchemical composition, particle size, "crystallinity," andsample preparation effects such as orientation (see alsoTowe, 1974). We have, therefore, presented our claymineral data only in terms of the relative XRD peak in-tensity. The strongest XRD peaks were chosen in orderto minimize the effects of differing degrees of preferredorientation which most strongly modify the intensity ofthe weaker, higher order XRD peaks. The 10Å peak ismeasured for illite (a discrete mica phase), and the 7Åpeak is used for chlorite, as kaolinite was found to beabsent or present only in trace amounts. The 17Å peakwas chosen for mixed-layer=illite/smectite (hereafterreferred to simply as illite/smectite), but its peak intensi-ty varies as a function of the proportion of smectitelayers relative to illite layers (Reynolds and Hower,1970). We, therefore, also report the proportion ofsmectite layers in the illite/smectite as determined by themethods of Reynolds and Hower (1970).

Chemical analyses of the washed < 1 µm and > 1 µmfractions were performed by automated electronmicroprobe analysis. Loss on ignition was determined at950°C and the ignited sample was fused with a 1:1 mix-ture of Li2B4θ7. Six analyses (with matrix corrections)were performed on each of the resulting fused glassesand the mean of the six analyses for each sample isreported in Tables 1 and 2.

In Core 11 of Site 322, several closely spaced sampleswere analyzed. For the purposes of plotting ourmineralogic and chemical data in the figures of this text,we have averaged three sets of data and plotted each as a

single point. The three sets of data averaged are in-dicated by brackets in Tables 1 and 2, and it can be seenthat no major chemical differences exist within each setof averaged data.

SITE 322 RESULTS

Mineralogy

The upper 295 meters of the sediments sampledappear to be dominantly terrigenous in origin. Thecoarse fraction has abundant quartz, feldspar, andhornblende. The clay mineralogy consists of illite,chlorite, and illite/smectite (Figure 1). Kaolinite is ab-sent or present only in trace amounts. The illite/smectiteis somewhat heterogeneous in its proportion of smectitelayers, but consistently averages about 60% smectitelayers in the upper 295 meters. The major proportion ofthese clays appear to be detrital in origin. This sedimentsection corresponds to lithologic Unit 1 and has beendescribed as a mature, unconsolidated mixture of sand,silt, and clay of late Miocene to early Pliocene age (Site322 report, this volume).

Two samples (4-2 and 5-1) were examined from theunderlying Unit 2 which has been described as an im-mature claystone of probable late Miocene age. Thesamples were taken at 355 and 392 meters and have thesame detrital illite, chlorite, and illite/smectites (60% ex-pandable) as Unit 1. However, clinoptilolite is now pres-ent in both the coarse and fine fractions even though theclay mineralogy does not seem to indicate any abundantalteration of volcanics. The clinoptilolite may haveprecipitated from the pore waters, perhaps incor-porating ions which have diffused upwards from themore volcanic-rich sediments below.

Section 6-1 (438 m) is the first sample encounteredthat has a drastically different clay mineralogy from theoverlying sediments. Clinoptilolite is still present, but il-lite and chlorite are now present in minor amounts(Figure 1). Also, the illite/smectite is now essentially apure smectite with nearly 100% smectite layers. As canbe deduced from arguments detailed elsewhere (Perry etal., in press), this mineralogy represents the submarinealteration of volcanics. This first evidence of volcanicalteration coincides with the highest Siθ2 pore watervalue of Gieskes and Lawrence (this volume). The claymineralogy is dominated by nearly pure smectites to adepth slightly greater than 512 meters, although clinop-tilolite only persists in any abundance to a depth of 509meters. This section of sediments from 438 to 512 metersmust be very enriched in altered volcanics, althoughdetrital minerals are still found in the coarse fraction ofthe sediments.

The basal 2 meters of sediment are not as enriched involcanics. The clays have more abundant detrital illiteand chlorite, and the illite/smectite now has only 60%

465

E. A, PERRY, JR., E. C. BECKLES, R. M. NEWTON

TABLE 1

Sample(Interval

in cm)

1-1,0-151-2, 135-1402-2,103-1073-1,0-124-2,135-1505-1,72-766-1,18-2210-0,0-1511-1,148-15011-2,125-12711-3,125-128H-4,40-4311-4,62-6411-4,122-12411-5,18-2011-5, 110-11211-6, 30-3211-6,103-10511, CC11, CC11, CC

Approx.Depth

(m)

76.579.5

193295355391.5438.5486505.5506.5508509509.5509.8510510.5512512.5513.8513.9514

K2O

2.071.761.912.672.322.723.092.780.991.101.182.624.072.632.954.214.254.214.414.453.96

CaO

0.450.320.320.540.700.390.390.610.240.150.050.57n.d.b

n.d.0.15n.d.n.d.n.d.n.d.n.d.n.d.

TiO2

0.860.820.820.931.101.170.841.261.391.821.831.990.800.870.980.900.770.780.710.750.78

Na2O

1.261.081.131.271.381.101.321.320.830.740.901.270.860.891.091.160.980.950.810.890.77

MgO

5.043.643.854.864.594.583.455.53

10.4510.4810.164.984.314.104.233.373.583.353.713.273.72

A12O3

14.2712.1312.6115.5516.0917.1917.1817.1614.0314.0014.1517.8116.8316.3616.6017.9219.3718.6019.3119.1719.10

SiO2

66.2372.0771.4063.0263.4663.0162.3959.9862.2160.6860.2463.1662.6063.0762.1159.8659.2058.5358.2857.6258.63

MnO2

0.380.170.090.290.280.151.380.370.350.350.370.550.390.640.710.851.751.681.711.651.79

Fe2O3a

9.448.037.88

10.8510.089.699.96

11.029.52

10.2311.127.06

10.1311.4411.1711.7410.1411.8811.0912.1911.25.

Note: Brackets indicate closely spaced samples which have been averaged together for graphical purposes inFigures 3 and 4.

aTotal iron as Fe2O3 ."n.d. = not detected.

TABLE 2

Sample(Intervalin cm)

1-1,0-151-2, 135-1402-2, 103-1073-1,0-124-2, 135-1505-1,72-766-1, 18-2210-0,0-1511-1, 148-15011-2,125-12711-3,125-128H-4,40-4311-4,62-6411-4,122-12411-5,18-2011-5,110-11211-6,30-3211-6,103-10511, CC11, CC11, CC

Approx.Depth

(m)

76.579.5

193295355391.5438.5486505.5506.5508509509.5509.8510510.5512512.5513.8513.9514

K2O

1.711.571.611.871.662.022.652.011.531.631.632.033.633.133.383.584.724.774.774.474.49

CaO

2.683.163.373.634.653.912.893.003.733.403.153.240.591.062.121.010.620.530.520.701.06

TiO2

0.760.750.710.810.710.840.800.730.670.670.740.920.880.981.080.880.870.790.750.740.72

Na2O

2.972.813.343.013.693.753.203.273.793.983.643.961.931.872.852.261.922.001.842.002.00

MgO

2.282.281.912.762.011.741.421.622.202.301.861.493.122.903.102.172.201.752.142.521.89

A12O3

14.6213.4114.5815.6816.4116.9516.1416.5415.5415.1014.8617.3415.7416.3917.3616.0516.3315.3215.8616.3915.42

SiO2

70.0872.2170.6066.7366.0266.2869.5067.5368.7168.9670.0666.3267.5265.8064.0266.9367.1469.0967.6164.7067.89

MnO2

0.090.070.080.130.130.240.260.390.260.050.420.430.440.190.391.170.870.461.231.371.16

Fe2O3a

4.793.523.815.404.724.273.144.933.573.903.644.266.177.675.705.945.32 '5.295.297.105.34

Note: Brackets indicate closely spaced samples which have been averaged together for graphical purposesin Figures 3 and 4.

aTotal iron as Fe2O3 .

smectite layers as opposed to being a nearly pure smec-tite. These lowermost sediments have much more of adetrital component than the altered volcanic claystonesabove them.

In addition to the sediment samples, four volcanicdrilling breccias were also examined by XRD to deter-

mine the mineralogy of the alteration products. Sample12-1, 90-100 cm has clay pebbles mixed with a basaltbreccia. The clay is mainly authigenic smectite plusclinoptilolite, but does have minor amounts of detritalillite and chlorite. Sample 12-1, 130-140 cm is a volcanicbreccia which is altering to smectite and phillipsite; the

466

CHEMICAL AND MINERALOGICAL STUDIES

50

100 -

150 -

200 -

RELATIVE0.2

XRD PEAK INTENSITIES0.4 0.6 0.8 1.0

250 -

α.LüQ

; \ \

\ \ j

I

/ Hole 322 <fMm~

Illite/smectite

v Chlorite

\ 11 lite

300 -

o00 350GO J J U

Z>C/>

400 -

450 -

500 -

550

Figure 1. Relative XRD peak intensity of the clay mineralsin the <l µm fractions of sediments from Site 322plotted against sample subbottom depth. The intensity ofa given XRD peak is normalized to the total intensity ofthe 7, 10, and 17k XRD peaks.

same is true of Sample 13-1, 25-26 cm. Sample 13-1, 130cm is composed of a mixture of hard volcanic pebblesand soft pebbles. The clays found are authigenic smec-tite, clinoptilolite, and detrital illite plus chlorite. Thesoft pebbles are apparently a mixture of volcanic altera-tion products and detrital sediments. With the exceptionof phillipsite, all the basalt fragments appear to be alter-ing to the same smectite plus minor clinoptilolitemineralogy as found in the overlying sediments.

Sediment ChemistryTables 1 and 2 give the major element chemistry of the

washed, ignited < 1 µm and > 1 µm sediment fractions.Ca, Na, and Si are much more abundant in the coarsefraction as its mineralogy is dominated by quartz plusplagioclase. Mg is much more abundant in the fine frac-tion due to its presence in both chlorite and illite/smec-

tite. Illite (K-mica) is present in both size fractions andthere is K-feldspar in the coarse fraction so that K doesnot exhibit any consistent differences in abundancebetween the coarse and fine fractions.

SITE 325 RESULTSOnly mineralogy has been determined for the

sediments of Site 325. Two sedimentary units arerecognized with the boundary between the two oc-curring at about 570 meters (Site 325 report). The upper570 meters of sediments are a mixture of silty clay, siltyclaystone, and claystone ranging in age from Miocene toQuaternary. The coarse fraction is dominated by quartzand plagioclase, and the fine fraction has a clay mineralsuite of illite, chlorite, and illite/smectite (Figure 2). Theillite/smectite is similar to that found in the upper partof Site 322 and also has an average of about 60% smec-tite layers.

The lower sediment unit (early Miocene) is richer insandstones and siltstones than the upper unit, but theclay mineralogy of two clay stones (642 and 710 m) fromthis unit indicates altered volcanics. Some clinoptiloliteis found from 615 to 622 meters and again could haveprecipitated from pore waters and incorporated ions

RELATIVE ×RD PEAK INTENSITIES0.2 0.4 0.6 0.8 1.0

Q_UJO

100 -

200 -

300 -

400 -

i

t

- 4

ii

i i

99 A

I '

I1

i Hole 325 <:fum

\ Illite/smectite

\ o- k Chlorite

-

Illite

<> :I 500DCO

600 -

700 -

800

Figure 2. Relative XRD peak intensity of the clay mineralsin the <l µm fraction of sediments from Site 325 plottedagainst sample subbottom depth. The intensity of a givenXRD peak is normalized to the total intensity of the 7,10, and 17k XRD peaks.

E. A. PERRY, JR., E. C. BECKLES, R. M. NEWTON

which may have diffused upward from these volcanics.Detrital illite and chlorite are less abundant (Figure 2),and the clays are dominated by illite/smectite with near-ly 100% smectite layers. Clinoptilolite is also present.

DISCUSSION

The above mineralogic and chemical data can at leastpartially explain some of the pore water chemicalgradients found by Gieskes and Lawrence (this volume).At Site 322, the following pore water trends were ob-served with increasing core depth: (1) a depletion in dis-solved Mg, (2) an enrichment in Ca which is larger thanthe Mg depletion, and (3) a depletion in K. The porewaters also exhibit a depletion in O18 (Lawrence et al.,this volume).

Our chemical data on the fine and coarse fractions(Figures 3 and 4) do not seem to indicate any majorchemical or mineralogical changes in the upper 400meters of detrital sediments. Illite and chlorite covaryclosely in abundance which would imply a commonprovenance, while the illite/smectite exhibits a largervariation in its abundance (Figure 1). The illite/smectiteof upper sediment section is somewhat heterogeneous inits composition so that it is possible that minor amountsof nondetrital smectite may be present and could haveformed from the submarine alteration of small amountsof volcanic material. As we discuss below, this alterationwould remove Mg and add Ca to the sediment porewaters.

The strongest Ca addition and Mg removal occur inthe pore waters of the sediments from below 400 meters

50

1OO

150

200

250

300

350

400

450

500

0.1 0.2

MeO/ΔI2θ3 ATOMIC

0.3 0.4 0.5 0.6 0.7 0.8

550

Hole 322

Fraction (washed)

Figure 3. Atomic cation/Al ratios versus sample subbottomdepth for the washed < i µm fraction of sediments fromSite 322.

Q.LüQ

50

100

150

200

250

300

MeO/AI2O3 ATOMICO.I 0.2 0.3 0.4 0.5

I 350

400

450

500

550

Hole 322>1µ.m Fraction

Cα/AI

K/AI

Mg/AI

Figure 4. Atomic cation /Al ratios versus sample subbottomdepth for the washed >l µm fraction of sediments fromSite 322.

(Gieskes and Lawrence, this volume). The sedimentsbelow this depth change in their clay mineralogy from adetrital suite to one dominated by authigenic smectite(Figure 1) which has formed as an alteration or sub-marine "weathering" product of volcanics. The silicatefractions of these sediments (which are enriched in smec-tites) have higher O18 contents than the above detritalsediments (Lawrence et al., this volume). These isotopicdata are best explained by the submarine weathering ofvolcanics to a smectite phase (Savin and Epstein, 1970;Lawrence et al., in press).

The smectite formation can also contribute to thepore water gradients through the following reactionscheme. Volcanics are weathered to release both Ca andMg to the pore water. The formation of smectite as aweathering product then removes not only the Mgreleased by the weathered volcanics but also the Mg in-itially present in the pore water itself. The smectite for-mation would also leave the pore waters depleted in O18,and the pore water gradient exhibits O18 depletion which

468

CHEMICAL AND MINERALOGICAL STUDIES

is strongest in these deeper sediments. The alteration ofthe basal basalt underlying the sediments appears toaffect the pore water gradients in a similar fashion and isalso producing smectite as an alteration product.

The chemistry of the sediments does not strongly sup-port the above hypothesis as it has done in earlierstudies (Perry et al., in press). Both the coarse and thefine fractions of the deeper sediments exhibit a depletionin Ca (Figures 3 and 4), as do the altered basalts at thebottom of Hole 322 (Donnelly and Wallace, thisvolume; Drever, this volume). The Mg shows a generalincrease in the clay fraction (Figure 3) which correlatesroughly with the smectite abundance, but the correla-tion is not as striking as it was for Site 149 sediments. AtSite 149 the Mg uptake by authigenic smectite was veryclearly demonstrated, but these were tropical sedimentsfrom the Venezuela Basin with no chlorite present andsmectite was the only clay phase having an appreciableMg content (Perry et al., in press). At Site 322 we havetwo Mg clay phases, smectite and detrital chlorite, sothat the Mg content of the sediments can be controlledby detrital and/or authigenic phases. The detritalsediments at Site 322 have a much higher Mg/Al ratiothan the detrital sediments at Site 149. The pattern ofMg enrichment in the volcanic sediments is thereforenot as clearly delineated from the detrital sediments.The volcanic sediments still exhibit some higher Mgcontents than the terrigenous sediments, including thetwo data points in the basal 2 meters (Figure 3) whichare also enriched in detrital components.

The pore waters also exhibit a slight depletion in K,which may be reflected in the chemistry of the coarsefraction of the sediments (Figure 4). Clinoptilolite andphillipsite were observed in the deeper sediments andbasal basalts. These zeolites appear to be the most likelysink for the pore water potassium, and they are moreabundant in the coarse fraction of the sediments.

Hole 325 pore waters exhibit chemical and isotopictrends which are similar to those of Hole 322, but whichare of an even larger magnitude (Gieskes and Lawrence,this volume; Lawrence et al., this volume). Basalt base-ment was not reached in this hole, but sediments en-riched in authigenic smectite were again encountered atthe bottom of the hole (Figure 2). Apparently the samereaction mechanisms discussed above for Hole 322 areoperative at Site 325. The deeper sediments, which pre-sumably formed much closer to a ridge crest spreadingcenter, are enriched in a volcanic component which isaltering to smectite plus zeolite phases. Again, thealteration of the basement basalts may be proceeding ina similar fashion and also contributing to the chemicalgradients observed in the pore waters.

SUMMARYSite 322 has an upper 400-meter section of sediments

which is dominantly of a terrigenous origin. Except forthe basal 2 meters of sediments which are also enrichedin detrital components, the deeper sediments have a claymineralogy which is dominated by a smectite and lesseramounts of clinoptilolite. These minerals have formedfrom the submarine weathering of volcanics whichcauses a net Mg removal and a Ca addition to the porewaters. This weathering also causes the pore waters tobecome depleted in O18 and the smectites to becomeenriched in O18 relative to terrigenous illite/smectites.The pore water depletion in K is most probably causedby the formation of zeolites. The basal basalts weatherby a mechanism similar to that of the volcanics in theoverlying sediments.

Although the basalt basement was not reached at Site325, a similar, but more pronounced alteration ofvolcanics is indicated by the pore water chemical andisotopic data. The mineralogic data support thishypothesis in that the upper 630 meters of sediments aredetrital in origin, but smectites are abundant in thedeeper sediments. The sediment section is thereforesimilar to Site 322.

The sediment chemical data generally indicate a Mgenrichment in the altered volcanics. The Mg enrichmentis not as clearly delineated as has been found elsewherebecause the upper sediment sections are already relative-ly enriched in Mg due to abundant chlorite derived froman Antarctic provenance.

ACKNOWLEDGMENTSThis research was supported by NSF Grant 40051 and ACS-

PRF Grant 7237-AC2.

REFERENCESLawrence, J.R., Gieskes, J.M., and Broecker, W.S., in press.

Oxygen isotope and cation composition of DSDP porewaters and the alteration of layer II basalts: Earth Planet.Sci. Lett.

Perry, E.A., Gieskes, J.M. and Lawrence, J.R., in press. Mg,Ca and O18/O16 exchange in the sediment—pore water sys-tem, Hole 149, DSDP: Geochim. Cosmochim. Acta.

Reynolds, R.C. and Hower, J., 1970. The nature of interlayer-ing in mixed-layer illite/montmorillonites: Clays ClayMineral., v. 18, p. 25-36.

Savin, S.M. and Epstein, S., 1970. The oxygen and hydrogenisotope geochemistry of ocean sediments and shales:Geochim. Cosmochim. Acta, v. 34, p. 43-63.

Towe, K.M., 1974. Quantitative clay petrology: The trees butnot the forest: Clays Clay Mineral., v. 22, p. 375-378.

469