20. AGE DETERMINATIONS ON IGNEOUS ROCKS IN HOLE 373A

20.1. POTASSIUM-ARGON AGE DETERMINATION OF BASALT SAMPLES FROM LEG 42A,HOLE 373A, CORE 7

H. Kreuzer, M. Mohr, and I. Wendt, Bundesanstalt für Geowissenschaften und Rohstoffe, Hannover, WestGermany

ABSTRACT

Eight basalt samples from Sections 2, 4, and 5 of Core 7, DSDPHole 373A, were investigated petrographically and chemically inorder to determine their suitability for potassium-argon dating. Allof the samples were at least partially altered, and thus partial lossof argon may have occurred. Close to the lower boundary of thesampled interval the basalt shows petrographic and chemicalindications of a higher degree of alteration. Here the potassium-argon dates are around 3.5 million years while in central parts ofthe interval the dates are around 5 million years. Consideration ofthe petrographic and chemical analyses together with K/Ar datingresults in a minimum age determination for this interval of 4million years.

INTRODUCTION

During Leg 42A of the Deep Sea Drilling Project, asite was drilled on the flank of a seamount on theTyrrhenian Abyssal Plain (latitude 39° 43.68'N, longi-tude 12°59.56'E) in a water depth of 3517 meters. A270-meter-thick Pliocene to Quaternary sequence ofmarls, tuffs, and zeolitic sediments was penetratedbefore drilling reached the acoustic basement. Penetra-tion into this basement reached a total of 187.5 meters.The sequence comprised: an upper 30 meters of basal-tic breccia with a limestone matrix, approximately 105meters of porphyritic and amygdular basalts interlay-ered with basaltic breccia, and these were followeddownwards by aphyric basalts flows to a terminaldepth of 457.5 meters subbottom.

Eight samples were received at this laboratory forpotassium-argon dating. These were taken from Core 7within the interlayered sequence of basalt and basaltbreccia. The core itself contains only an upper 50 cm ofbasaltic breccia while the remainder (approximately 5m) is entirely made up of basalt with inclusions ofpalagonite and tuffaceous sediment or veins of sparrycalcite.

METHODSEach sample was initially subjected to detailed

petrographic and chemical analysis in order to assessalteration and thus the degree of accuracy to beexpected from the dating technique. In particular, wewished to determine the likelihood of argon lossthrough alteration of the basalt.

The K/Ar dating was carried out on whole-rocksamples. These were ground to less than 800 µm,treated with 10% formic acid for 4.5 hours to removecalcite, soaked overnight in 0.4 N sodiumpyrophos-phate solution, and intensively treated ultrasonically(after dilution to 0.017V) to remove weakly attachedparticles. Afterwards the samples were washed withdistilled water and menthanol and dried at 50°C.

Argon was determined by isotope dilution and potas-sium by flame photometry. The relative analytical errorof the potassium determination (approximately +2%) isnegligible compared with the large error of the radio-genic argon concentration. This is due to the small ab-solute amount and the correspondingly low fraction ofradiogenic argon in the total argon.

PETROGRAPHY

The upper five samples of the sequence show nomarked petrographic differences. Table 1 lists ourgeneral petrographic observations derived from themicroscopic analyses and serves to distinguish an uppergroup of five porphyritic, amygdaloidal basalt samplesfrom three altered types of varying texture which makeup the bottom of the sequence.

In the upper five samples plagioclase phenocrystsare generally free of alteration, but the laths of thegroundmass may show strong alteration to carbonate.Augite phenocrysts have suffered alteration of varyingintensity to smectite, with the smectite growth begin-ning at the grain boundaries. Olivine phenocrysts aregenerally completely altered to a mixture of limonite,smectite, and carbonate. However, in Samples 7-2, 17-25 cm, and 7-4, 0-9 cm, isolated crystal shapes arepresent containing relic unaltered olivine. Olivine isbelieved also to have been part of the groundmass inthese rocks, but is not identifiable because of the levelof alteration. Smectite limonite aggregates occur in anyinterstices which are not filled with pyroxenes. Sample7-2, 34-41 cm, is especially rich in limonite. It isgenerally impossible to determine whether this mineralhas formed during late-stage development of a smec-tite-limonite mixture as an interstitial filling or whetherit is simply an alteration product of mafic minerals.Amygdules are usually coated with smectite and filledwith carbonate.

Alteration extends even to the larger phenocrysts inSample 7-4, 74-83 cm. Plagioclase phenocrysts up to 1mm long, with the exception of their rims, are entirely

531

TABLE 1General Petrographic Observations on Basalt Samples From Hole 373A, Core 7

Sample(Interval in cm)

7-2,17-25

7-2, 34-41

7-2,131-136

7-2, 142-150

7-4, 0-9

7-4, 74-83

7-4,136-141

7-5,0-7

Texture

Porphyritic, amyg-daloidal; groundmassintergranular tointersertal; amyg-dules generally < 2mm length

Porphyritic, amyg-daloidal; groundmassintersertal to sub-op hitic

Aphyric, subparallel;Amygdules 0.5 mmlength; 2 zones, 1mm wide, show ofcoarsening matrix

Fine to extremelyfine grained, partlymicrolitic

Notes on Minerals Present

Plagioclase

As phenocrysts:single oragglomerated,

idiom orphicmorphic, tabular;An ~80%, zoned ortwinned, up to 5 mm

lengthIn groundmass: laths,twinned or zoned

Few stout, tabular, al-tered, idiomorphicphenocrysts, up to 1mm length; An -80%;laths in groundmass0.1 to 0.4 mm lengthLaths in groundmass,0.05-0.2 mm length;one altered micro-phenocryst (0.3 mmlong); some orienta-tionAcicular, partly skele-tal, 0.05 to 0.1 mmlength

Augite

Hypidiomorphic, ingroundmass betweenplagioclase laths, lightbrown, zoned, ormosaic structure,0.05-0.3 mm length

Hypidiomorphic toxenomorphic; lightishbrown, zoned ormosaic structure, 0.1-0.4 mm length

Hypidiomorphic toxenomorphic; lightbrown (Ti rich?),zoned or mosaicstructure

Hypidiomorphic (?),highly altered;brownish, up to 1mm length

Olivine

Stout phenocrysts;idiomorphic, up to 2mm length; probablyin groundmass also

Idiomorphic pheno-crysts, reduced size;probably in ground-mass alsoPhenocrysts idiomor-phic, up to 2 mmlength; probably ingroundmass alsoFew altered pheno-crysts, 1-3 mm

Idiomorphic, 0.6mm length; highlyaltered

Opaques

Skeletal grainsas inclusionsin augite andsmectite

Idiomorphicgrains, evenlydistributed,mainly as in-clusions insmectiteAs inclusionsin augite

Limonitepresent

Carbonate

Generallypresent

Present in in-terstices be-tween plagio-clase laths ofgroundmass

Plagioclaselaths alteredto carbonate

As veins 0.5mm broad

Smectite

Aggregates ininterstices

Very rich insmectite/limoniteaggregatesInterstitial fill-ing, very finegrained aggre-gates or subpar-allel fibers, 0.1mm long.

Fine-grainedsmectite aggre-gates, fibers raregreen color

Smectite as in-clusions inplagioclaselaths

Smectite asinclusions inplagioclaselaths

RockName

Porphyriticamygdaloidbasalt

Porphyriticamygdaloidalbasalt(altered)

Fine-grainedamygdaloidalbasalt(altered)

Fine-grainedamygdaloidalbasalt(stronglyaltered)

POTASSIUM-ARGON AGE DETERMINATION OF BASALT SAMPLES

replaced by carbonate. Plagioclase laths in the matrixof this sample are generally penetrated by smectite andcarbonate which also occur as interstitial filling. Therims of augite phenocrysts are frequently altered tosmectite and the few larger (1-3 mm) olivine pheno-crysts are completely altered to limonite, smectite, andcarbonate. Smectite also occurs as cross-laminatedaggregates and sometimes as parallel fibers (0.1 mmlong) which may have formerly been pyroxenes. Thematrix smectite in this sample frequently contains fine-grained leucoxene which may suggest a former glassyground mass.

Plagioclase laths in the fine-grained amygdaloidalbasalt sample, 7-4, 136-141 cm, contain inclusions ofsmectite which again might have formed by alterationor devitrification of glassy matrix inclusions. Augitecrystals in the groundmass are partially altered tosmectite, and the latter also fills most of the intersticesbetween these and the plagioclase laths. The amyg-dules present are coated with smectite and filled withcarbonate.

In the most altered and lowermost sample, 7-5, 0-7cm, plagioclase laths usually contain smectite, and bothaugite and olivine are almost entirely replaced by amixture of smectite and limonite. The groundmassconsists of smectite with limonite and small idio-morphic ore grains. Pseudomorphs after mafic mineralsare difficult to recognize. Amygdules in this sample, upto 0.5 mm long, are again filled with carbonate.

CHEMICAL ANALYSIS

The results of the chemical analyses (X-ray fluores-cence, H2O, FeO, CO2 = wet chemical analyses) areshown in Table 2, with those for trace elements inTable 3. These chemical analyses agree rather wellwith the results obtained by Dietrich et al. (this vol-ume), especially if one compares the average valuesthey report for samples from Core 7 with the averagesof our results.

As can be seen in Table 2, the five samples from theupper part of Core 7 have significantly higher SiO2 and

A12O3, and slightly higher K2O and Na2O concentra-tions than the three samples closest to the base.However, MnO, MgO, and FeO(?), as well as H2O andCO2, are all significantly lower in the samples of theupper group of five samples.

The lower SiO2, A12O3, and alkali concentrations(including the trace element Sr) in the lower parts ofthe sampled sequence correspond with an absence ofplagioclase phenocrysts as reported in the petrographysection. The increase in MgO (and also Mn and V,Table 3) in these samples suggests a relative enrich-ment in mafic components (see Figure 1). However,because of the strong alteration detectable in thesesamples, it is not possible to determine whether olivineor enstatite (resp. pigeonite) could be responsible forthis increase.

Trace elements which are resistant to alteration,such as Zr, Nb, Y, and Sc, appear in both groups. Inthe Ti-Y-Zr triangle-plot from Table 3 all of thesamples are located in the "Ocean-Floor-Basalt"(OFB) area (Figure 2).

Table 4 shows the CIPW-normative composition ofthe samples as calculated from the chemical analyses inTable 2. The resulting plagioclase plot (Figure 3)yields an An content of about 55% to 65% (except forSample 7-5, 0-7 cm, which is close to the base of thecore). Again this is in agreement with the results ofDietrich et al. (this volume). However, in contrast, ourcalculated hypersthene contents are generally higher,and the olivine contents (Figure 4) are lower thanthose given by these workers. All of the samples plot inthe olivine tholeiite field.

SUITABILITY OF THE CORE 7 SAMPLES FORK/AR DATING

Petrographically we recognize the eight samples ashaving come from the central part (the upper group offive) and lower margin (the lower three) of a basalticsill. The rock is classified as an amygdaloidal basalt, itsmain constituents being plagioclase (phenocrysts andmatrix), augite and smectite, with minor olivine (phe-

TABLE2Chemical Analysis of Samples From Hole 373 A, Core 7 (in %)

SiO2

A12O3

TiO2Fe2O3

FeOMgOCaOMnONa2OK2OP2O5SO3

co2H2OLOITotal

7-2,17-25 cm

46.8718.740.8744.972.558.02

10.430.1082.950.300.120.02

<0.033.240.33

99.55

7-2,34-41 cm

47.8118.560.9155.502.557.63

10.090.1032.990.370.130.010.142.800.53

100.13

7-2,131-136 cm

48.3618.930.9304.772.417.49

10.250.0913.050.330.120.020.061.631.60

100.04

7-2,142450 cm

47.0718.710.8445.092.487.95

10.520.1082.810.270.100.030.213.600.24

100.03

7-4,0-9 cm

46.2918.830.8274.752.718.39

10.700.1142.680.230.100.110.423.97

-0.04100.08

7-4,74-83 cm

45.1017.010.8824.903.31

10.159.990.1332.580.260.100.030.884.620.76

100.71

7-4,136-141 cm

44.4216.680.9575.463.09

10.0710.660.1762.570.220.110.050.893.991.07

100.41

7-5,0-7 cm

41.4116.240.8934.873.43

11.1410.590.1782.370.250.110.821.534.862.57

101.26

533

H. KREUZER, M. MOHR, I. WENDT

TABLE 3Trace Elements of Samples From Hole 37 3A, Core 7 (in ppm)

SrRbNbZrYPbNiCoMnVZnCuCrCeLaBaSc

7-2,17-25 cm

212125722114843686214847812743057032

7-2,34-41 cm

2011047835271123083115061852593019933

7-2,131-136 cm

210347827101003573215151702731628033

7-2,142-150 cm

214141682118883187914949782792648237

7-4,0-9 cm

21140613412973690015247593112707438

7-4,74-83 cm

1908266251114042

106016160732702309335

7-4,136-141 cm

190756921239241

1440177731292363508638

7-5,0-7 cm

17512265235

10240

1474163811962402527537

A = Na2O + K2OF = Total Fe as FeOM MgO

Figure 1. A-F-M plot of Core 7 basalt samples. Key tosample numbers: 1 = 7-4, 0-9 cm; 2 = 7-4, 74-83 cm;3 = 7-4,136-141 cm; 4 = 7-5, 0-7 cm;5 = 7-2,17-25 cm;6 = 7-2, 34-41 cm; 7 = 7-2, 131-136 cm; 8 = 7-2, 141-150 cm.

nocrysts and ground mass), opaques (including limo-nite), and carbonate. Sporadically, pigeonite andquartz are identifiable. The phenocryst content de-creases to zero towards the lower selvage while smec-tite and carbonate increase due to the alteration ofplagioclase and mafic minerals. The mafites may alsobe partially altered in the central parts of the sill.

Samples 7-4, 74-83 cm; 7-4, 136-141 cm; and 7-5,0-7 cm; are consequently unsuitable for dating becauseof intense alteration whereas the other five samplesappear moderately suitable for radiometric dating onthe basis of their petrography.

Chemically, few of the samples are sufficiently freshfor confident dating. Generally water contents are less

Figure 2. Ti-Zr-Y triangle plot of Core 7 samples. LKT =low potassium tholeiite; OFB = ocean floor basalt; CAB =calc alkali basalt; WPB = within plate basalt. (Samplenumbers same as in Figure 1.)

than 2%. The three samples from the lower part of thebasalt layer (7-4, 74-83 cm; 7-4, 136-141 cm; 7-5, 0-7cm), which on petrographic grounds are already un-suitable for K/Ar dating, have the highest H2O con-tents. SiO2 and H2O (Figure 5) as well as TiO2 (Figure6) and H2O, are negatively correlated for the upperfive samples, while the lower three samples deviateconsiderably from the regression lines defined by theupper group. On the other hand, the degree of oxida-tion (Fe3+/Fetotai) is negatively correlated with H2O(Figure 7).

POTASSIUM-ARGON DATING RESULTSThe radiometric dating results are listed in Table 5.

The ages of the eight samples range from 3 to 5 m.y.

534

POTASSIUM-ARGON AGE DETERMINATION OF BASALT SAMPLES

TABLE 4CIPW - Normative Composition of Hole 373A Core 7 Samples (degree of oxidation as determined)

OrAbAn

Ne(Wo

DNEnIFs

Mt11ApThHmCcH 2 OTotal

7-2,17-25

1.7925.0337.37

-(5.76

10.74J4.98

7.4 9 |7.49

5.36J5;3 6

6.101.660.290.040.800.073.24

100.01

7-2,3441

2.1925.3436.31

-(5.10

9.50 J4.40

12.75 { 1 2

:

7 5

1.34 | ^ 3 4

5.901.750.310.021.450.322.80

99.98

7-2,131-136

1.9826.0837.64

-(5.36

9.99 \ 4.63

1 2 . 7 9 1 1 2 : 7 9

1.07 | ^ ° 7

5.451.790.290.041.090.141.63

99.98

7-2,142-150

1.6023.6337.82

HM(l.Ol)(5.21

9.72 j 4.51

11.92 f1;92

2.39 | 2 ; 3 9

5.931.600.240.051.010.483.60

99.99

74,0-9

1.3621.9339.01

_(4.472

8.336 | 3.862

12.42 {12;_42

3.21 | 3 2 1

6.681.570.2370.1950.1380.9543.97

100.00

74,74-83

1.5421.6534.19

-( 3.83

7.17 { 3.23I 0.10

7 n I 6.881 0.25

7.1091.670.240.05—2.004.62

100.01

74,136-141

1.3121.5633.72

-(5.48

10.21 \ 4.74

8.93-j8-93

S.llf:11

7.811.830.260.09—

2.043.99

99.98

7-5,0-7

1.5014.8836.26

-( 2.69

5.04 \ 2.25I 0.09

14-64 {14;06

8.64 | 8 2 7

7.15 "1.710.261.47-

3.534.86

99.94

Figure 3. Normative plagioclase (Or, Ab, An) plot of Core 7samples. (Sample numbers same as in Figure 1.)

with individual errors of 20% to 50%. The lowest agescorrespond to the most altered samples in the lowerpart of the sampled sequence (i.e., the lower margin ofthe sill). They yield an average age of 3.4+0.5 m.y.while the less altered upper group of five samples yieldan average age of 4.2+0.5 m.y., and a weighted meanage of 4.0+0.3 m.y.

DISCUSSION

The average deviation from the weighted mean agequoted above exceeds the mean analytical error (1.6σ),so there is evidence here for disturbance of the K/Arsystem. This is indeed likely in view of the alterationdetected in all of the samples. Thus we regard theK/Ar dates of Table 5 as minimum ages. The basalt ofCore 7 consequently should be older than 4.0 +0.3 m.y.

Savelli and Lippanni (this volume) analyzed twosamples from Core 7. Their samples from Section 2(5-12 cm and 67-76 cm) are from the vicinity of our sam-ples in that section (17-25 cm, 34-41 cm, and 131-136cm). However, their K/Ar ages are significantly higherthan ours (mean age 7.2 ±0.9 [2σ] m.y.). It is possiblethat the formic acid, sodium phosphate, and ultrasonictreatments applied in our laboratory are capable them-selves of reducing K/Ar ages and this is to be checked.

ACKNOWLEDGMENTS

We wish to thank Prof. Harre and Dr. Raschka and theirlaboratory staff for the chemical analyses, and Dr. P. Müllerfor critical and stimulating discussion of our results.

REFERENCE

Harland, W. B. and Francis, E. H. (Eds.), 1971. The Phan-erozoic time scale: Geol. Soc. Lond. Spec. Publ., v. 5.

535

H. KREUZER, M. MOHR, I. WENDT

Figure 4. Ne-Di-Ol-Hy-Q-plot of Core 7 samples.Figure 1.)

(Sample numbers same as in

48 r

46 -

CM 44OCO

42 -

40

-

-

_

-

-

1

\

1 1

3

\\• 2

• 4

I

H 2 0

Figure 5. Si02 - H2O correlation of Core 7 basalt samples.(Sample numbers same as in Figure 1.)

TABLE 5K/Ar Data of Samples From Hole 373A, Core 7

Samplesa

(Interval in cm)

7-2, 17-25

7-2, 34-41

7-2,131-1367-2,142-1507-4, 0-97-4, 74-837-4,136-1417-5,0-7

Kb

0.1620.1670.2050.2120.1610.1430.1420.1560.1450.247

Radiogenic Argon

(ccSTP/g) 107

0.239 ± 0.0470.315 ±0.0620.282 ± 0.0400.279 ± 0.0950.279 ± 0.0500.287 ± 0.0560.275 ± 0.0540.224 ± 0.0460.176 ± 0.0390.35 ±0.17

(%)

4.95.53.72.76.25.35.02.7

10.63.0

K/Ar-Age(m.y.)

3.7 ± 0.74.7 ± 0.9c

3.4 ± 0.53.3 ± l . l c

4.3 ± 0.85.0 ± 1.04.9 ± 1.03.6 ± 0.83.0 ± 0.73.6 ± 1.7

Note: All errors given are intervals of 95% analytical confi-dence. For age calculations the constants of the "PhanerozoicTime Scale" (Harland and Francis, 1971), were used, i.e.,λß = A.I2 10"10 year"1; λe = 0.584 lO~1 0 year"1; 40K/K

=

1.19 10-2 atom %.aGenerally, whole rock ground to 400-250 µm, preheated over-night to < 150°C except where noted.

^Relative analytical error of K analysis = ± 2%.cWhole-rock fraction 800-400 µm, preheated overnight to<80°C

536

K/AR AGE DETERMINATIONS ON BASALT ROCKS

1.5 r-

1.0

OP

0.9

0.8

0.70r

0.65

• 3

• 6

• 40.60

0.55

% H2O HoO

Figure 6. T1O2 - H2O plot of Core 7 samples. (Samplenumbers same as in Figure 1.)

Figure 7. Fe3+/Festal vs H2θ. (Sample numbers same asin Figure 1.)

20.2. K/Ar AGE DETERMINATIONS ON BASALT ROCKS FROM HOLE 373A

C. Savelli and E. Lipparini, Laboratorio di Geologia Marina, CNR, Bologna, Italy

Three basalt rock specimens from Hole 3 73A wereanalyzed. Two are from lithologic Unit IV (Core 7,Section 2). Sample 1 (Table 1) was from the interval67-76 cm and Sample 2 from 5-12 cm. The sampleswere chosen as being the freshest of a total of fivesamples from Core 7 at our disposal. Two othersamples (from cores 12 and 4, respectively) are notanalyzed because of obvious alteration. The basalt inthe analyzed samples is porphyritic and has pheno-crysts of plagioclase, "iddingsitized" olivine, and rarebrown spinel in a groundmass consisting of plagioclase,pyroxene, olivine, ore, and alteration products (mainlythose of glass, smectites, and limonite). Further, asecondary mineral is calcite, which fills some of the

void spaces of the rock. Plagioclase and pyroxenecrystals are fresh, as is shown in Figure 1.

The third specimen was from lithologic Unit V(Core 12, section 1, 120 cm; Sample 3 of Table 1). Itspetrographic composition corresponds fairly well withthe average composition of the whole unit as givenabove: plagioclase 45% to 60%, pyroxene 20% to 25%,and alteration minerals 20% to 30%. Plagioclase andpyroxene crystals are significantly altered as is shownin Figure 2.

For argon isotopic determinations high-frequencyinduction heating and an all-metal mass spectrometerwere used. Samples were preheated in vacuum atabout 160°C for'20 hr. Potassium was analyzed by

537