Neodymium and strontium isotopic dating of diagenesis and low-grade metamorphism of argillaceous sediments

Neodymium and strontium isotopic dating of diagenesis and low-grade metamorphism of argillaceous sediments

- Geochimica Pergamon et Cosmochimica Acta. Vol. 58. No. 5. DD. 1471-1481. 1994 dopyright 0 1994 ~lsevier Science Ltd Printed in the USA.All right...

1MB Sizes 0 Downloads 21 Views

-

Geochimica

Pergamon

et Cosmochimica

Acta. Vol. 58. No. 5. DD. 1471-1481. 1994 dopyright 0 1994 ~lsevier Science Ltd Printed in the USA.All rights reserved

OOl6-7037/94 $6.00 + .OO

Neodymium

and strontium isotopic dating of diagenesis and low-grade metamorphism of argillaceous sediments

URS SCHALTEGGER, ’ PETER STILLE, ’ NAOUAL RAIS, 2 ALAIN PIQUE, 2 and NORBERT CLAUER ’ ‘Centre de Gtochimie de la Surface (CNRS), I, rue Blessig, 67084 Strasbourg, France *Dkpartment des Sciences de la Terre, UniversitC de Bretagne Occidentale, 6, avenue Le Gorgeu, 29287 Brest, France (Received December 9, 1992; accepted in revised form October 4, 1993)

Abstract-The behaviour of the Rb-Sr and Sm-Nd isotopic systems with increasing degree of Hercynian metamorphic overprint was studied along a transect in Cambrian shales of northwestern Morocco. Clay fractions of ~0.2 to 2-6 pm size from five samples were investigated, representing a range from nonmetamorphic to epizonal metamorphic conditions. The samples were washed in cold 1 N HCl prior to digestion to separate soluble/exchangeable Rb, Sr, Sm, and Nd from amounts of these elements fixed in the crystallographic sites of the minerals and to analyze both components separately. The results reveal that the Rb-Sr isotopic system is dominated by Sr hosted by clay mineral phases (both detrital and authigenic illite and chlorite) and carbonate-hosted soluble Sr. Isotopic homogenization of Sr occurred during Hercynian metamorphism, yielding ages between 309 and 349 Ma. The Sm-Nd isotopic system, on the other hand, is dominated by cogenetic apatite and Fe oxide/ hydroxide, both having high contents of leachable REEs. The leachates yield a Sm-Nd isochron age of 523 + 72 Ma, indicating diagenetic equilibrium between apatite and Fe-oxide/hydroxide. Fine-grained clay fractions of <0.2 pm size plot onto this reference line, suggesting isotopic equilibrium with the leachates. Size fractions >0.2 Frn show inheritance of a detrital Nd component. The study demonstrates that the diagenesis of the investigated argillaceous sediments can be dated by the Sm-Nd chronometer in authigenic cement phases. The isotopic system of these minerals (apatite, Fe hydroxide/oxide) was homogenized during authigenic mineral growth in a sediment that was flushed by diagenetic fluids and had abundant primary or secondary interconnected pore space. The Hercynian metamorphic overprint caused partial isotopic rehomogenization of the adsorbed and clay-hosted portion of the Sr as well as of the carbonate-hosted Sr. The Sm-Nd system in the cement phases survived this metamorphism. This results in decoupling of the two isotopic systems and allows the dating of diagenesis on the one hand ( Sm-Nd) and metamorphism on the other hand (Rb-Sr). INTRODLXXION VARIOUSEFFORTSHAVEBEEN madeinthelasttwo decades to determine the age of deposition, diagenesis, and low-grade metamorphism of sediments by various isotopic methods. The use of the K-Ar and Rb-Sr methods to date sediments has been widely applied and accepted (ARONSON and HOWER, 1976; CLAUER, 1979, 1982; HUNZIKER, 1979; CLAUER et al., 1993). There are many cases, however, where neither method yields unequivocal results. The discrepancy in K-Ar and Rb-Sr age data of sediments has many possible causes, among which are ( 1) inheritance of detrital components; (2) mineralogical and isotopic heterogeneity of the analyzed material; (3) and mixing of different components with different isotopic characteristics within the same clay fraction, which are not isotopically equilibrated during lowtemperature overprinting. Different approaches were developed by different investigators to cope with these difficulties: the use of sized fractions from the same sample ( ARONSON and HOWER, 1976; REUTER, 1987), careful disaggregation ofthe rocks to avoid fragmentation of the mineral grains ( LIEWIG et al., 1987), and the combination of mineralogical, morphological, and geochemical investigations ( CLAUER, 1979, 1982 ). These procedures allow the determination of which grain size in a given elastic sediment would be free of detrital components or

mineral phases ( CLAUER et al., 1993). A series of grain-size fractions from the same sample usually displays decreasing K-Ar and Rb-Sr ages for decreasing grain size. Very small grain sizes ( ~0.2 pm or ~0.1 pm) often represent the maximum amount of authigenic clay and minimal detrital contribution. The use of the Sm-Nd method to date sedimentary systems is a recent development. The two REEs Sm and Nd are known to convey information on the age and nature of a detrital component of sedimentary rocks and even on the source region of sediments, because they are relatively immobile during weathering, transportation, deposition, diagenesis, and low-grade metamorphism ( MCCULLOCH and WASSERBURG, 1978, CHAUDHURI et al., 1992). The preservation of the isotopic signature of the source allows the calculation of depleted mantle model ages (T,,; DEPAOLO, 1988) or crustal residence ages (O'NIONS et al., 1983). tNd values and mantle model ages reflect the integrated crustal history of the sedimentary material and allow recognition of important crustforming events ( MICHARD et al., 1985 ). It has been shown only recently that the Sm-Nd system might be used to date the diagenesis of fine-grained elastic sediments ( STILLEand CLAUER, 1986; OHR et al., 199 1; BROS et al., 1992; STILLEet al., 1993). New evidence has recently been presented that shows considerable fractionation of the Sm-Nd system during early diagenesis of a elastic sedimentary 1471

1412

U. Schaltegger

system (AWWILLER and MACK, 1991; BROS et al., 1992; CHAUDHURI et al., 1992; ZHAO et al., 1992). Results of isotopic dating in sediments generally have to be treated with caution, because they may reflect mixtures of two or more components with different origins, which thus yield mixing lines with apparent ages usually in excess of the depositional age of the sediment ( NAGLER et al., 1992). The Rb-Sr and Sm-Nd isotopic systems are expected to behave differently during metamorphic overprinting, due to either their different geochemical behaviour or to differences in the chemical resistance of their principal host mineral phases. The present study was undertaken to determine the mineral components in argillaceous sediments, whose isotopic system has been disturbed during a low-grade metamorphic overprint. Knowledge of the mineralogical distribution of the Rb, Sr, Sm, and Nd inventories in the different size-fractions of a sediment is crucial to the understanding of the mechanisms of isotopic homogenization during diagenesis and lowgrade metamorphism. The target of such an investigation is to assess the geological and mineralogical limitations for the successful dating of sediments. STUDY

AREA

AND SAMPLE

et al.

(b

SELECTION

The study was carried out on greywackes from the Coastal Block of the Western Meseta, northwestern Morocco (Fig. la). The greywackes comprise alternating psammitic and pelitic layers, with a high amount of volcanogenic material ( PIQLJk, 1979). The shales contain the fossil Pnrudoxides ( LECOINTRE, 1926), which implies a Middle Cambrian sedimentation age (536-5 17 Ma, according to HARLAND et al., 1990). The sedimentary sequence of the Moroccan Meseta was overprinted by Hercynian metamorphism. The northern part of the investigated area experienced diagenetic conditions only, while the southern part reached anchizonal to epizonal metamorphic conditions (PIQUE, 1979; WYBRECHT, 1985). The age of metamorphism was estimated at 370 Ma in the eastern part of the Meseta ( CLAUER et al.. 1980) and at 330-290 Ma in the western part ( HUON et al., 1987) It was associated with a penetrative foliation, which was cut in the study area by the ca. 265 Ma old Rehamna granite ( MRINI et al., 1992). thus defining a minimum age for the deformation. A profile southwest of Casablanca (Fig. I b) was chosen for our study. It covers the whole range of thermal overprinting including the nonmetamorphic sample 2988, the anchizonal samples 2990 and 299 1, and the epizonal samples 2995 and 3014. A short description of the analyzed material is given in Table I. A metamorphic assemblage of biotite, garnet, staurolite, and kyanite was recognized in psammitic rocks at the southern end of the profile, indicating upper greenschist facies conditions (PIQUE, 1979). The metamorphic zonation used here was based on illite crystallinity values of the <2 Frn fractions ( RAIS, 1992), which were determined following the method of KOBLER ( 1966). Potassium-Argon investigations revealed the existence of a post-Hercynian hydrothermal overprint that caused the formation of smectitic clays and some secondary chlorite and the lowering of the K-Ar ages to around 200 Ma (RAIS, 1992). Triassic ages of 184-220 Ma are scattered throughout the Meseta and the adjacent High Atlas and Anti Atlas regions in the South (HUON et al., 1993). The ages are thought to be related to crustal extension during the opening of the Northern Atlantic, which is also reflected by coeval basin development and crust-contaminated MORB-type basaltic dike magmatism (e.g., FIECHTNER et al., 1992). METHODS The samples were gently crushed after removing all signs of meteoric weathering. A size fraction of 163 pm was obtained by dry sieving, followed by washing in distilled water. Grain size fractions ~2 pm and 2-6 pm were separated using differential settling in distilled water according to Stokes’ law, and smaller fractions <0.2,0.2-

25 kilometers

FIG. 1. (a) Major tectonic units of northwestern Morocco. Ruled: Paleozoic rocks of the Coastal Block, study area. (b) Geological sketch map of the investigated area in the Coastal Block of the Western Meseta, Morocco (after PIQU& 1979). Degree of Hercynian metamorphism is estimated from illite cristallinity ( WYBRECHT, 1985). Stippled: Outcrops of the investigated Paradoxides schists (Middle Cambrian). Crosses: Rehamna granite. Sample localities are indicated by stars and sample numbers.

0.4 (or <0.4), 0.4-1, and l-2 pm were further separated by ultracentrifugation. The final grain-size distribution was examined using a scanning electron microscope. Grain-size separation procedures are described by RAIS ( 1992). Most of the analyzed samples have been leached in 1 N HCI at room temperature for I5 minutes. The leachate and the corresponding residues were separated by centrifugation and the residues rinsed in distilled water to remove the salts. The leachates, residues, and untreated aliquots were analyzed independently for Rb-Sr and Sm-Nd isotopic compositions. Residues and untreated samples were ignited at 8OO“C for 24 hours to oxidize the organic matter, because the presence of organic matter during the acid digestion seems to prevent complete spike equilibration and leads to erroneous isotopic results. The ignited powders were spiked with “Rb, ‘%r, and ‘50Nd-‘49Sm tracers and subsequently digested with HF and HNOJ for 5-7 days in closed Savillex vials, dried, and treated with 6 N HCI for several hours. The chemical separation of Rb and Sr was carried out on AGSOW-Xl2 (200-400 mesh) cation-exchange resin using I.5 N and 4 N HCI as eluents, respectively. The REE were eluted with 6 N HCI, and subsequently separated on the same type of resin using alpha-hydrobutyric acid as eluent. Total procedural blanks are < 1.5

Dating argillaceous

sediment

TABLE 1. Short description

Sample

III [%I 2988

2990

2991

2995

3014

Sm [%I

80

20

0

unmetamorphlc.

80

20

0

green

0.4.lprn

75

25

0

sedimentary not disturbed

1-2pm

70

30

0

2.6pm

70

30

0

20

0

anchimetamorphlc,

15

0

(diagenesis);

0.4.lpm

80

20

0

develops

l-2pm

75

25

0

2.6pm

60

20

0

70

20

10

low aplzonal

70

25

5

extenwe

1-2pm

65

30

5

penetratwe

2.6pm

55

40

5

grains and oriented

sample;

pressure

<0.4pm

50

35

15

epizonal

55

35

10

(epizone);

1.2pm

55

40

5

Hydrothermal

2.6vm

55

40

5

smectlte.

sample,

15

10

75

epvonal

sample

30

30

40

thermal

alteration;

1.2pm

45

45

10

(anchzone).

2.6pm

45

40

15

secondary

I.C. (
of a of detrltal

= 2.5

metamorphic

wlh

forms

follatlon.

secondary

strong hydro-

I.C. (
Hydrothermal smectite.

= 3 (eplzone);

formation

new growth.

alteration

<0.4vm

layertng

due to rotation

penetrative

0.4.lprn

lnterstrallf!ed

= 6.5

I.C. (c2Fm) solution;

cleavage

0.4.lpm

(and

(c2p)

solution cleavage

and forms a secondary

<0.4pm

= 5

aIteratIon

which lowers

clorite-vermiculite).

forms

the I.C. value.

Sm=smectlte,

(1992)

ng for Sr and ~0.4 ng for Nd. Strontium, neodymium, and samarium isotopic ratios and concentrations were analyzed on a VG Sector multicollector mass spectrometer. Strontium was deposited on W single filaments with TazOS as an activator, whereas Nd and Sm were loaded on a Ta-Re-Ta triple filament assembly with hydrochloric and phosphoric acids. The NBS 987 standard yielded an 87Sr/s6Sr value of 0.7 10266 + 3 (2 sigma mean, n = l8), and the ‘43Nd/ ‘44Nd ratio of the La Jolla standard solution was measured at 0.5 11852 t 3 (2 sigma mean, n = 10). The BCR standard yielded a 14jNd/ ‘44Nd ratio of 0.5 I2638 & 9 (2 sigma mean) and a Sm/Nd ratio of 0.2296. The 85Rb/s7Rb ratio was determined on a TSN 206 single collector mass spectrometer using Ta single filaments. The accuracy of the 87Rb/86Sr ratios is better than +l%, whilst that of the ‘?jrn/ ‘“Nd ratios is better than +0.5%. Reference lines were calculated using model 3 of ISOPLOT (LUDWIG, 1988). After the separation of leachates and residues by centrifugation, some of the leachates were diluted to approximately 30 ml with I N HCI and Ca, Mg, Fe, and P determined by spark emission spectroscopy, K by atomic absorption spectroscopy, Ti and Ba by ICP, and REEs Pb, U, and Y by ICP-MS. The precision for trace and REE concentrations is better than + 10%. The mineralogy and geochemistry of the studied material are described in detail in RAIS ( 1992). Accessory mineral phases in a few clay fractions were investigated using a Philips CM I2 scanning transmission electron microscope (STEM) equipped with an EDAX-EDS system. MINERALOGICAL

I.C

pressure

0.4.lpm

RAIS

strattflcatlon

by metamorphism

65

Chl=chlorite

bloturbation

psammtticlpelltlc

60

after

= 9 (dlagenesls);

rock; extenwe

cO.Zpm

Ill=lllite,

descrlptlon

I.C. (
compact

0.2.0.4pm

cnstallintty;

material

Short

Chl [%I

cO.Zpm 0.2.0.4vm

Abbreviations: I.C.=lll!te

of the investigated

Clay mineralogy

size fractions

1473

by Rb-Sr and Sm-Nd

CONSTRAINTS

The mineralogical composition of the studied material was investigated by standard X-ray diffraction (XRD) methods ( RAIS, 1992). The analyzed samples consist mostly of illite, chlorite (and possibly interlayered chlorite-vermiculite), and smectite. The shape of the grains indicates that newly formed illite is concentrated in grain-size fractions smaller than about 1 pm. Coarser illites show diffuse boundaries reflecting a probable detrital origin. Trace amounts of quartz and K-feldspar grains were found in the coarser fractions. Because the XRD method has detection limits around a few volume percent, some of the clay separates were examined by a scanning

transmission electron microscope (STEM). This revealed the presence of minor amounts of K-feldspar, albite, a Fe-Mg carbonate, Fe-oxide/ hydroxide, T&oxide, apatite, and monazite. Two semi-quantitative chemical EDS analyses of apatite and monazite are given in Table 3. The Triassic hydrothermal alteration at around 200 Ma produced smectite and secondary chlorite, which are enriched in the finest grain-size fractions ( ~0.4 pm). The two northernmost samples 2988 and 2990 are free of secondary smectite. The smectite content in fraction x0.4 pm increases toward the south-east from 5% (sample 299 I ) to 75% (30 14). The addition of hydrothermal smectite caused a shift in illite cristallinity from a value around 2.5 (epizone) to 5 (anchizone) in the <2 wrn fraction of the highest grade sample 30 I4 (Table I ). The K-Ar system of the fine-grained fractions of sample 3014 was severely affected by this process ( RAIS, 1992). CHEMICAL

COMPOSITION

OF THE

LEACHATES

The chemical compositions of the leachates are of interest, because they may be indicative of mineralogical assemblages, which were dissolved or leached during the acid treatment. This treatment is known to remove Rb, Sr, Sm, Nd, and other elements from sites accessible to the reagent, i.e., soluble accessory minerals, adsorbed complexes or ions in exchangeable interlayer positions, but it does not preferentially remove radiogenic isotopes (CLAUER, 1979; CLAUER et al., 1993). The removal of the leachable Sr and Nd results in an increase in s7Sr/86Sr and s7Rb/s6Sr ratios and a decrease in ‘43Nd/ ‘44Nd and ‘47Sm/‘44Nd ratios of the residues. Mass balance requires that the leachates, residues, and untreated samples must plot on straight lines in Rb-Sr and Sm-Nd isochron diagrams, respectively. Poor correlations were obtained for most of the l-2 pm and 2-6 wrn fractions, which indicates isotopic inhomogeneity. The leachates usually have low and uniform 87Sr/BhSr ratios, which may be interpreted to be close to the value of the last interstitial fluid phase that was in equilibrium with the newly forming authigenic clay minerals ( CLAUER, 1979). Leachate, residue. and untreated sample define a line whose slope corresponds to the age of isotopic equilibration in such a case, if no inherited minerals nor inherited radiogenic Sr are present. For the Sm-Nd system, an inverse relationship for leachate

1474

U. Schaltegger et al. TABLE 2. Chemical analyses of major, minor and rare earth ele-

ments in the leachate solutions elements

2988 2968 2988 ~0.2pn 0.2.0.4pm0.4.lpm leachate

leachate

leachate

2988

2990

2.fJym

2.6pm

leachate

leachate

2991 c0.4pl leachats

2991 0.4-l pin leachate

2995 2+m leachate

2995

3014

6.20pm

sO.4pm

leachate

leachate

3014 2.6pm leachate

ca I%1

8.58

9.72

14.79

9.52

14.92

3.98

3.52

6.99

8.56

10.46

11.04

K

3.55

3.36

5.09

1.90

2.42

2.12

1.31

1.35

1.16

1.16

Al

lO.fi

9.89

14.79

n.d.

12.28

8.98

5.01

n.d.

3.72

1.20 6.11

Fe

6.71

6.56

10.10

4.76

5.41

8.47

3.60

3.26

1.96

3.59

WI Mfl

2.11

2.51

3.73

2.13

2.25

2.31

1.20

1.14

0.92

3.21

4.00 2.76

0.17

0.23

0.46

0.24

0.07

0.11

0.06

0.70

0.50

0.44

2.91

0.26

0.54 1.19

0.14

0.53

n.d.

0.42

0.79

0.96

1.93

1.25

0.61

2.05 n.d.

1.10

1.35

n.d.

436

8a

2.34

2.94

0.99 3.74

TI ItwJ F?J

175

233 2739

167

180

218

n.d.

639

139 1142

369

6646

237 1785

95

1288

n.d.

211

602

n.d.

U Th

33

47

98

n.d.

34

79

73

n.d.

39

102

n.d.

10

12

20

o.d.

146

68

72

n.d.

60

28

Y

163

277

365

kd.

375

294

296

n.d.

983

118

n.d. n.d.

222

359

708

rid.

102

n.d.

4049

1498

n.d.

891

n.d.

n.d. n-d.

n.d.

n.d.

7650 to37

1683

130

375 f3

6342

79

1833 271

4321 7812

4324

554

1025

248

n.d.

342

565

1240

n.d.

429

3956

n.d.

4091

947

n.d.

83

136

323

n.d.

195

3852 633

652

n.d.

782

153

n.d.

19

30

71

n.d.

96

101

n.d.

169

25

n.d.

113 11

n.d.

P

% leachate

n.d, =

0.55

n.d.

63

114

302

n.d.

54 229

7.9

15

33

n.d.

30

38

69 12

177

n.d.

6.7

29

n.d.

17

29

69

2.5

3.3

1017

72

n.d.

407

435

n.d.

698

37

n.d.

78

127

36 119

112

n.d.

307

37

n.d.

19

17

17

n.d.

45

5.3

n.d.

n.d.

40

49

43

n.d

91

12

n.d.

7.9

n.d.

4.1

10

7.5

n.d.

5.4

1.1

o.d. n.d.

15

25

48

n.d.

23

35

24

n.d.

34

6.7

2.8

4

7.6

n.d.

3.1

4.4

2.8

n.d.

3.3

1.3

n.d.

1450

2362

5126

n.d.

1703

18408

18398

17720

4741

n.d.

2.9

1.9

0.97

1.68

1.4

1.5

0.97

0.70

7.7

r1.d. 1.32

1.4

not delermined

and residue has been obtained by OHR et al. ( 199 I ) and BROSet al. (1992) with the leachates having higher Sm/Nd and ‘43Nd/‘44Nd ratios than the residues. The chemical analyses of the leachates in this study suggest that most of the leachable Sr and Nd is derived from soluble minerals rather than from exchangeable sites of the clay minerals. The Ieachates are characterized by high concentrations of Ca, Al, K, Mg, and Fe (Table 2), with each sample having its own, distinct leachate chemistry. The high Mg, Fe, and Ca contents may originate from dissolved or leached ankeritic carbonate; Fe is also probably derived from Feoxides/hydroxid~. The origin of the relatively high K contents remains unclear, because K in interlayer sites of clay minerals is inaccessible to weak acid solutions. Leaching of altered accessory grains of K-feldspar and plagioclase, as revealed by the STEM investigations, would account for the high K and Pb and extremely high Ba concentrations (Table 2). Barium seems to be a constituent of a detrital and/or diagenetic mineral phase in sample 2988 and decreases with increasing metamo~llic and/or hydrothermal overprint. The Ieachates contain high levels of At(up to 14.7%), which suggests dissolution of an Al-bearing silicate phase. Phosphorous concentrations of up to 3% are probably caused by the dissolution of apatite and possibly leaching of monazite. Positive correlations between PZOs and Nd concentrations for untreated fractions of samples 2988/2990 and 2995 (Fig. 2a) point to a phosphate phase as host for the REEs. During the examination of the clay fractions using the STEM, one monazite grain was found in fraction 0.2-0.4 pm of sample 2988 and semi quantitatively analyzed (Table 3). Monazite is usually resistant to the weak acids as used for the acid leaching and may, therefore, be of little importance for the soluble/exchangeable portion of the REEs. It can be assumed, however, that at least some of the apatite would be dissolved during our leaching procedure in cold I N Hfl. A compa~son of phosphorous concentrations in untreated fractions and leachates suggests that about lo-40% of the apatite was

dissolved. We suggest, therefore, that the phosphate host of the REEs in samples 2988, 2990, and part of the fractions of 2995 is apatite. Some size fractions from samples 30 14 and 2995 show no correIations in Fig. 2a. These samples and sample 299 I show a positive correlation for Fe and Nd concentrations in their leachates (Fig. 2b), which suggests that large amounts of REEs are absorbed or adsorbed by a Fe-oxide/hydroxide in the southern higher grade samples (see TUREKIAN et al.,1973; SHAW and WASSERBURG, 1985; PALMER and ELDERFIELD, 1986). REEs are extremely concentrated in the Ieachates (Tables 2 and 4): up to 70% of the total Nd inventory is contained in 1-2 ~01% leached material. The REE distribution patterns normalized to NASC

TABLE3. Semi-quantitative STEM-EDS analyses of monazite and apatite Element [%I

PZOS CaO La203 Ce203 Nd& ThO2

W t

O.Z-0.4um 2988 monazite

0.4-lurn 2995 apatite

35.2 <1.0 15.6 29.0 13.0

45.8 54.2
6.5

0.7 700.0

Dating argillaceous sediment by Rb-Sr and Sm-Nd

(4

0.24 -

0.20 016-

0" a"

0.12 0.08 -

+,*

pd,6**e"

,,&

s

,'

0 ,*'

,,,,,/'

0

,,*"' ,Q*

0

P' ,' ,' ,' ,' ,' ,' ,'

/

untreated n 2388 0 2990 + 2995 0 3014

0.04 -,**'

100

0

\ W 2388 + 2990 A 2331 + 2335 0 3014 L

’ /

I?!

0.2 t'

.+ 0

cl

Leachates

0.6 -

0.4 -

3 OC

[ppml

(b) 0.8 -

J

200 Nd

s

,/$ ,'

+,_ A,YA, 20

40

60

f

Nd [ppml FIG. 2. (a) PZOs vs. Nd diagram of untreated samples showing positive correlations for samples 2988/2990 and 2995, which suggests a phosphate phase as host of the rare earth elements (data from RAIS, 1992). (b) Fe vs. Nd diagram of leachates showing a positive correlation between the two elements in samples 299 1, 2995, and 30 I4 (data from Table 2). This indicates that the major part of (leachable) Nd is adsorbed onto or absorbed by Fe-oxide/hydroxide in these samples.

(North American Shale Composite, GROMET et al., 1984) are humpshaped (Fig. 3). This, together with high REE concentrations exceeding 1000 times chondritic values, may be a typical feature of diagenetic apatite known to scavenge large amounts of REEs ( SHAW and WASSERBURG, 1985). Samples of higher metamorphic grade and with hydrothermal alteration show a slight enrichment of the light REEs, which also means lower Sm/Nd ratios, and a slight negative Ce-anomaly (Fig. 3b). Cerium is known to be fractionated in a strongly oxidizing environment due to the oxidation to Ce4’, causing the negative anomaly ( MARIANO, 1989). ISOTOPIC

RESULTS

Rb-Sr The Rb-Sr results are compiled in Table 4 together with the Sm-Nd isotopic data. Grain-size fractions from <0.2 pm (co.4 pm) to 2-6 pm were analyzed independently for their leachates, residues, and the untreated aliquots. The leachates are rather uniform in composition: the “Rb/% ratios are always below unity and the 87Sr/86Sr ratios range from 0.722 to 0.7 I7 for the nonmetamorphic or the anchimetamorphic samples and from 0.717 to 0.713 for the epimetamorphic and hydrothermally altered samples. The leachate-residue- (untreated) tie-lines yield decreasing apparent Rb-Sr ages for decreasing grain size within each

1475

sample and also for the same grain size of different samples of increasing metamorphic grade (Table 5 ). The resulting apparent ages range from 440 Ma (coarse fractions of sample 2988) to 227 Ma ( ~0.4 pm of sample 30 14), which is similar to the range of the K-Ar ages (RAIS, 1992; CLAUER et al., 1994). Sample 30 14 is severely affected by the post-Hercynian hydrothermal alteration; the leachate-residue tie-line of the ~0.4 pm size fraction has a slope that yields an apparent age of 227 Ma (Table 4). This age is in the same range as the KAr age for the same sample and lower than that of Hercynian metamorphism. The untreated fractions of the same sample yield a cluster of points in a Rb-Sr evolution diagram and indicate postmetamorphic disturbance of the isotopic system, too. The least metamorphic samples 2988 and 2990 are not hydrothermally altered and do not contain young smectite, and thus may give precise information on the age of Hercynian metamorphism. Untreated clay fractions of both samples define a reference line with an apparent age of 309 f 43 Ma and an initial “Sr Ja6Sr ratio of 0.7 18 + 0.0 12. An apparent age of 348 + 23 Ma and an initial Sr isotopic ratio of 0.720 * 0.009 (Fig. 4a) is yielded by their residues only. ratio and 1 /Sr (Fig. Linear relationships between *‘Sr/%r 4b) suggest that the reference lines are the result of two-component mixing and consequently that the apparent ages defined by either reference lines may not be geologically meaningful. Very fine fractions <0.2 pm might yield the best estimate for the age of the Hercynian metamorphic overprint, because they usually contain the maximum amount of newly formed illite and are strongly depleted in inherited detrital components. Leachates, residues, and untreated aliquots of samples 2988 and 2990 define a reference age of 342 f 21 Ma, with an initial “Sr/*%r ratio of 0.7 15 -t 0.0 IO (Fig. 4~). Sm-Nd

Most of the samples were analyzed for samarium and neodymium isotopic compositions of leachates, residues, and untreated clay fractions (Table 4). Untreated samples show variable concentrations of Sm and Nd, the latter reaching values of up to 267 ppm ( <0.4 pm, 2995 ), which is unusually high for clay fractions. The REE are enriched in the leachates during the acid treatment, where Nd reaches concentrations of several thousands of ppm in the dissolved material (see Tables 2 and 4). The leachates have variable ‘43Nd/‘44Nd and ““Sm/ ‘44Nd ratios, the latter ranging from 0.09-0.2 1, that are in most cases higher than their residues and the untreated sample. Leachate-residue(untreated) tie-lines of the two low-grade samples, 2988 and 2990, in the northern part of the profile, yield apparent ages between 5 13 and 868 Ma, up to 250 million years in excess of the depositional age of the greywacke. The initial +d values of the tie-lines are around -8. Higher grade samples show disturbed Sm-Nd systems: the leachate-residue tie-lines are tilted to a zero or negative slope in some fractions of samples 299 1, 2995, and 30 14 (Table 5), coupled with slightly lower Sm/Nd and similar Nd isotopic ratios for the leachates compared to the residues. The three components are positively correlated in the case of the 0.4-l pm fraction of 2991, but show the inverse relationship with the leachate having the lowest 143Nd/‘44Nd

U. Schaltegger ct al.

I476

TABLE 4.

Sample

Size fraction

2990

<0,2

u R L 0.2-0.4 U R L 0.4-I u R L l-2 u R L 2-6 U R L u R L 0.2-0.4 U R L

l-2

2-6

2991

2.87

1.97

0.97

2.11

1.24

<0.2

0.4-I

<0.4

0.4-l

u R L u R L u R L u R L u R

3.04

0.36

0.40

0.45

1.61

1.86

0.58 1-2

2-6

2995

<0.4

0.4-l

l-2

2-6

3014

CO.4

0.4-l

l-2

2-6

=Rb/86Sr

37sr/mr

i2s

k25

eNd(0)

ENd(Si0)

0.511877 0.51 1886 OS12006 0.511838 0.511856 0.512005 0.511866 0.511878 o.51zoss n.d. 0.511882 0.512000 0.511861 0.511834 0.512070

8 4 6 I5 5 6 3 16 6

-14.9 -14.7 -12.4 -15.7 -15.3 -12.4 -l!il -14.9 -11.3

-8.8 -8.3 .7.4 -3.4 -8.6 -8.1 -8.7 -8.1 -8.0

6 3 3 4 5

-16.0 -12.5 -15.2 -75.7 -11 1

-8.3 -8.1 -2.0 -2.1 -7.9

0.0979 0.0862 0.2067 0.1043 n.d n.d. 0.0972 n.d. md. 0.1004 n.d. n.d. 0.1039 n.d. n.d

0.511902 0.511905 0.512232 0.511910 n d. n.d. 0.511893 n.d. n.d. 0.512050 n.d. n.d 0.511907 n.d. ".a.

7 8 7 2

-14.4 -14.3 -6.8 -14.2

-8.0 -7.1 -7.5 -8.2

5

-14.5

-30

5

-11.5

-5.2

3

-14.3

-8.3

88.99 n.d. 2352.33 n.d. 58.57 4572.37 99.53 72.36 3956.77 30.47 n.d. 3188.52

0.0324 n.d. 0.0938 n.d. 0.1051 0.0264 0.1035 0.1074 0.0387 0.1021 n.d. 0.1026

0.511410 n.d. 0.511888 n.d. 0.511771 0.511928 0.511852 0.511900 0.511876 0.511340 n.d. 0.511827

20

-24.0

-17.7

49.88 9.04 108S.37 27.52 5.26 489.27 41.86 10.71 323.69 8.73

267.29 47.65 5936.85 148.47 32.55 2700.04 221.73 53.70 5010.63 46.08

0.1129 0.1147 0.1100 n. d. 0.1178 0.1094 0.1135 0.1206 0.1103 0.1146

PPmSm

PPmNd

5 5 5 7 5 5 7 6 5 6 4 4 5 5 4

3.22 2.53 43.96 4.83 3.94 76.80 6.23 5.36 194.18 n.d. 9.17 179.0s 6.55 5.60 115.39

18.84 15.62 219.74 29.33 25.29 356.56 38.02 34.83 806.63 n.d. 62.82 842.02 3366 35.64 474.43

0.1035 0.0279 0.1211 0.1008 0.0942 0.1303 0.0930 0.0931 0.1457 n.d. 0.0883 0.1238 0.1024 0.0949 0.1471

0.780288 0.795843 0.722539 0.764153 0.764149 0.721550 0.741566 0.743751 0.721534 0.728613 0.731497 0.720474 0.738293 0.747545 0.721309

7 6 4 5 5 4 6 s 4 4 4 5 6 5 4

3.52 2.72 23.75 7.73 n.d n.d. 11.74 ".d. n.d. 21.34 n.d. n.d. 1013 n.d. n.d.

21.73 19.07 6949 44.70 n.d. n.d. 72.58 n.d. n.d. 127.82 n.d. n.d 58.67 n.d. n.d.

0.746588 n.d. 0.713590 0.748350 0.750344 0.712754 0.744938 0.748543 0.711844 0.746208 0.747388 0.714535

5

14.75 n.d. 364.87 n.d. 10.20 732.67 17.16 12.94 649.94 5.51 n.d. 541.18

0.730450 0.734665 0.714657 0.735371 0.738117 0.714270 0.734713 0.736677 0.714218 0.738645

6 4 5 7 4 5 4 5 5 5

wm Rb

PPrn Sr

263.80 300.09 60.19 267.20 278.67 83.41 247.10 240.55 155.30 227.30 223.80 130.10 194.60 2os.50 71.63

17.92 13.57 417.00 24.20 19.46 495.00 53.85 37.29 728.80 53.85 48.21 495.20 19.73 55.91 600.60

43.430 66.080 0.418 32.360 43.040 0.488 12.610 18.880 0.617 12.290 13.540 0.761 2.894 10.710 0.34s

0.221131 1.042283 0.720331 0.840315 0.937477 0.717550 0.803434 0.828381 0.717428 0.772734 0.794525 0.714376 0.772932 0.780427 0.716532

242.82 235.44 58.53 265.44 261.63 133.08 227.20 223.23 232.78 227.76 217.56 239.19 220.18 197.23 107.85

48.72 39.24 367.49 83.83 75.34 490.26 123.22 117.34 775.23 199.93 189.42 742.42 95.76 97.62 573.31

14.521 17.505 0.465 8.595 10.100 0.722 5.352 5.426 0.877 3.302 3.330 0.940 6.671 5.867 0.543 6.710 n.d. 0.237 6.719 10.339 0.221 8.326 8.060 0.158 15.107 10.109 0.253

1‘%m/l44Nd

143Nd/144Nd

Leachate

WI 2988

96

Rb-Sr and Sm-Nd isotopic data

: R L u R L u R L " R L u R L U U R L u R L u R L u R L u IJ R L

0.34

0.32

4.27

4.55

3.52

1.32

7.73

5.38

5.50

5.57

297.49 128.74 n.d. n.d. 70.18 855.51 447.77 133.54 511.47 135.82 117.46 1534.34 272.47 20.89 326.31 117.51 71.13 1305.55 284.58 54.70 258.86 74.36 1256.18 1410.68

6 5 5 4 5 4 5 4 5 4

173.20 195.50 74.1, 188.00 212.50 47.50 201.10 174.20 53.49 182.60

103.00 103.00

92.03 103.80 185.40 117.54 107.20 268.10 66.70

4.876 5.507 0.421 5.326 5.944 0.742 5.273 4.715 0.578 8.249

139.63 65.30

87.82 380.50

6.528 0.427

0.740904 0.715925

5 4

3.26 595.48

16.36 3198.47

n.d. 0.1122

178.20 188.32 47.98 497.28 218.47 133.10 212.30 203.47 30.27 201.30

58.29 22.97 436.80 63.33 SO.75 409.67 66.33 67.05 153.30 57.84

8.313 23.320 0.318 22.605 12,523 0.341 9.291 8.821 0.572 10.110

0.746377 0.732175 0.715652 0.753144 0.764102 0.715837 0.750788 0.755648 0.716811 0.747833

6 5 5 8 S 4 6 6 5 6

11.93 7.10 70.27 18.75 9.40 o.d 23.01 14.33 17.75 13.32

66.75 36.50 439.08 106.30 50.50 n.d 124.39 72.42 107.33 70.80

0.108s 0.1176 0.0362 0.1067 0.1167 n-d. 0.1116 0.1189 0.0933 0.1137

n.d. 0.715228

4

9.46 119.01

47.41 687.83

0.1206 0.1038

n.d. 7.25

SOS.60

md. 154.46

n.d. 0.136

4

-14.7

-8.0

30 3 7 5 8 6

-16.3 -13.2 -15.4 -14.4 -14.9 -13.7

-11.0 -7.4 -9.3 -8.6 -8.5 -8.0

8

-14.5

-8.4

0.511942 0.511944 O.Sl1973 n. d. 0.511985 0.511966 0.511967 0.511367 0.511990 0.511266 0.511356 n.d 0.511963

6 12 3

-13.6 -13.6 -12.3

-8.1 -8.2 -7.3

11 5 4 4 4 4

-12.8 -13.1 -13.1 -13.1 -12.7 -13.1

-7.7 -7.5 -7.7 -8.2 -7.0 -7.8

6

-13.1

-7.6

0.511890 0.511884 0.511903 0.511917 0.512002 n.d. 0.511343 0.511887 0.511891 0.511854 0.511871 O.Sl18S3 0.511862

5 5 3 7 5

-14.6 -14.7 -14.3 -14.1 -12,s

-8.2 -9.6 -7.7 -8.3 -7.2

6 5 5 3

-15.5 -14.7 -14.6 -15.4

-10.0 -3.6 -8.3 -2.9

4 3

-15.2 -15.2

-10.3 -2.1

Abbrenatiofis:U=untreated, R-residue, L-leachate; n.d.=notdetenmned

and 14’Sm/ ‘44Nd ratios (indicated verse co~eIation).

in Table 5 as positive

in-

The Sm-Nd isotopic analyses of the leachates define a linear array in a Sm-Nd isochron diagram (Fig. 5a). A best-fit line was calculated through the leachates of samples 2988, 2990, 2991, and 2995, and yields an age of 523 + 72 Ma, with an

initial tNd of -7.6. Sample 3014 was excluded chron calculation because of its high degree

morphic hydrothermal

from the isoof postmeta-

alteration. The residues of <0.4 Frn

size fraction plot onto the feachate reference

line within their

analytical error, whereas the coarser-grained residues and three of four residues of sample 30 14 plot below the line (Fig.

1477

Dating argillaceous sediment by Rb-Sr and Sm-Nd

o cO.2pm

2988

l 0.2.0.4pm

0 0.4.lpm .

2.6pm

2988 2986

1 .o

2990 & : \ ;;

0.9

2

ntreated 0.8

r,

Samples:

: 0.718

309

? 0.12;

i: 43

Ma

MSWD=lOZO

0.7 0

20

40

60

80

8’Rb/86Sr

ii-La

.Ce

Nd

iI?

Sm

Eu

Tb

/

/

Yb

Lu

-

FIG. 3. Rare earth element (RI%) distribution patterns of leachates (a) from lowest grade metamorphosed samples 2988 and 2990, (b) from higher grade samples 299 1,2995and 3014, no~alized to North American Shale Composite (NASH, Gromet et al.. 1984).

0.71

0.02

0.04

0.06

1 /St

5b). The untreated samples scatter in a very limited range of Sm/Nd and Nd isotopic ratios (Fig. 5c) and plot mainly below the leachate reference line. The finest fractions ~0.2 Grn from samples 2988 and 2990 were plotted separately in a Sm-Nd isochron diagram to test the degree of isotopic equiIib~um (Fig. 6). These fractions

I;; : \ Ik

0.9

-

2

age: 342

? 2

Sr,: 0.715

5. Summary of leachate-residue tie-line relationships for RkSr and Sm-Nd. depleted mantle Nd model ages and K-Ar ages TABLE

1Ma

+ 0.010

MSWD=3220

0

20

.

’ 40



.

’ 60

1

a0

87Rb/*6Sr 2988

Z990

2991

299s

3014

0 2-0.4 0.4-l l-2 2-6

343 352 404 437 426

0.77159 0.7121 07137 0.7102 0.7143

701 566 445 698

10.2 0.2-O 4 04-I i-2 2-6

298 342 327 324 346

0.7203 0.7183 0.7174 Cl.7161 0.7186

n. n. n n


<0.4

0.4-I i-2 2-6

dO.4

0 4-l J-2 2-6

CO.4 0.4-l 1-2 2-6

358 no alignment 326 234 266 304 361 387 227 no alignment 331 no alignment

07124

868

513

d. d. d. d.

05113 0.5114 05115 05116 0.5114

n. d

2.09

07129 0.71 0.7111 07140

negatwe slope porltw, l”“er$e correlation negatwe ape n.d. “.d.

07158

negawe slope n. d. no aflgnment honzontal

07141

1.33 1.33 1.12 1.35

n. d. negatwe slope no alignment n. d.

” d.

n d.

387 110 fl I t12 313 fl2

i A4 408 441 141 440 1.42 453 1.39

0.5116 n. d. n. d. n. d n. d.

07111 0.7137

II

2.91

362 i t 415*9 416 f. 428 *

1.35 386

I 34

10 10

303 f6 299 + 6 t 6 317t7

1.31 32

266 286 295 298

1.49

206 t 7 208 * 5 246 * 6 236t6

1 140 135 149

t,Of

I

8 8

1.44 296

1.33

(d)

clay

A

~0.2

m

m 0 A + Cl

0.9 -

n

0.8

f 9 f 8 * 9 zt 6

-

.A &:+A+

0.71

I

0.2

fractions to Z-6pm

CE

i

0.4

+

CJ

8

0.6

2988 2990 2991 2995 3014

+

I3

8

.

0.8

ch~orite/(chiorite+i~lite) FIG. 4. (a) Rb-Sr isochron diagram and (b) 87Sr/86Sr vs. I /Sr diagram of untreated samples and residues of the least metamorphic samples 2988 and 2990; (c) Rb-Sr evolution diagram for fractions ~0.2 pm of same samples; (d) *‘Sr/*% vs. chlorite content of clay fractions, indicating mixing of chlorite and illite.

147x

U. Schaltegger et al.

The source ofthe neodymium

(a)

'18

.

0.5116'

I

0.10



0.1 2

.





0.14

.

0.16



0.18

n

2988

0

2990

A

2991

+

2995

0

3014



0.20

1

0.22

147~m/144Nd (b)

Two-stage model ages were calculated to elucidate the possible sources of Nd in the investigated Cambrian sediment, using today’s 14’Sm/ ‘44Nd ratio until 520 Ma and an average crustal ‘47Sm/‘44Nd ratio of 0.115 (TAYLOR et al., 1983) from 520 Ma backwards to a depleted mantle evolution curve. The resulting crustal provenance ages are in the range of 1.31.5 Ga for most of the untreated samples (Table 5). These model ages are intermediate to those of the two major crustforming events in Northern Africa, the Ebumean (2.1-1.9 Ga), and Pan-African (730-600 Ma) orogenies, respectively (see KROGH and KEPPIE, 1990, for references). This suggests that the detrital Nd was derived from magmatic and volcanic rocks formed during these orogeneses.

0.5124

INTERPRETATION 0.5122 2 :

O.SlZC

s z $ _

0 0.51lf

0.511f

0.10

0.12

0.14

0.5124

sample 3014

0.18

0.20

0.22

44Nd

-

. x

fractions>0.4pm

0

0.16

t47Sm/’ (c)

fractions<0.4vm

0

[Untreated)

“““i

o

___j

d

0.5116'





0.10



0.12





0.14

1 47Sm/1





0.16





0.18



0.20



0.22

OF THE RB-SR REFERENCE

LINES

It has been demonstrated that the correlation in Fig. 4a is the result of two-component mixing on the basis of the s7Sr/ *%r vs. I /Sr relationship (Fig. 4b). The two endmembers can be denoted by plotting the strontium isotopic composition vs. the mineralogical composition of the clay fraction (e.g., the chlorite content, Fig. 4d). These two components are (A) an illite with high Rb/Sr, which is partly detrital, partly newly formed, and (B) a chlorite-type mineral with a low Rb/Sr ratio. This mineralogical mixture yields best-fit-lines with consistent ages of 309 +- 43 Ma (for untreated fractions) and 348 ? 23 Ma (for residues), in agreement with the age of 342 +- 2 1 Ma for leachates, residues, and untreated aliquots of ~0.2 pm fractions of samples 2988 and 2990. The addition of similar amounts of detrital illite to newly formed illite would result principally in a steepening of mixing lines. The scatter of the points for untreated samples and residues around the reference line in Fig. 4a, however, rather reflects variable admixture of inherited detrital illite, leading to large errors in the calculated 8’Sr/86Sr intercepts. The Rb-Sr ages of 309 +- 43 Ma to 342 + 21 Ma are in agreement with a KAr age of 300 t- 10 Ma, given by RAIS (1992), CLAUER et al. ( 1994), and HUON et al. ( 1987) for the age of Hercynian metamorphism. We suggest, therefore. that illite and chlorite formed contemporaneously during Hercynian metamorphism and are in isotopic equilibrium.

44Nd

FIG. 5. (a) Sm-Nd isochron diagram of leachates. The leachates define a reference line of 523 k 72 Ma, dating the diagenesis reflecting isotopic equilibrium among phosphate and Fe-oxides/hydroxides; (b) Same diagram with residues. Residues of CO.2 pm size plot onto the leachate reference line within analytical errors, suggesting isotopic near-equilibrium, whereas residues >0.2 wrn plot below the line; (c) Same diagram with untreated fractions, plotting mostly below the leachate reference line.

~

0.5122

z :

2 = T

indicate equilibrium for the Sr isotopes during Hercynian metamorphism (Fig. 4~). Leachate, residue, and the untreated of the ~0.2 pm fraction from sample 2990 plot onto the 523 Ma reference line for all leachates (Fig. 5a), reflecting an isotopic equilibrium 523 Ma ago. Residue and untreated aliquot of sample 2988 plot slightly below the leachate reference line.

0.5120 t

0.5118""



0.10





I”““’

0.15

1 47Sm/1

0.20

* ’

1

0.25

44Nd

FIG. 6. Sm-Nd isochron diagram for leachates, residues and untreated

~0.2 Frn fractions

of samples 2988 and 2990.

Dating argillaceous sediment by Rb-Sr and Sm-Nd The leachates have uniform and low Rb/Sr and 87Sr/86Sr ratios from 0.722 to 0.7 13, decreasing with increasing degree of metamorphism and/or hydrothermal alteration. The geochemical composition of the leachates suggests that most of the leachable Sr is hosted in carbonate, some possibly in phosphate. The considerably lower 87Sr/86Sr ratios of the leachates of the samples 299 1, 2995, and 30 14 compared to those of samples 2988 and 2990, suggests a change in the isotopic composition from a diagenetic leachate Sr reservoir to a Sr reservoir of metamorphic or hydrothermal origin with a lower strontium isotopic composition. The reference line of the fractions ~0.2 pm of the two least metamorphosed samples (2988, 2990; Fig. 4c) demonstrates that in the finest size fractions newly formed illite, chlorite, and carbonate/ phosphate are in isotopic equilibrium, produced during Hercynian metamorphism over a distance of 30 kilometers. The widespread, uniform geochemical and isotope geochemical character of the leachates suggests that the metamorphic carbonate was precipitated from a fluid, which was buffered by a pre-existing soluble phase with regionally homogeneous strontium isotopic composition, such as diagenetic carbonate cement. DIAGENETIC EQUILIBRATION SM-ND SYSTEM

OF THE

The Sm-Nd data of the leachates define a best-fit line of 523 Ma in a Sm-Nd isochron diagram (Fig. 5a), which most likely dates the diagenetic homogenization of the leachable portion of Sm and Nd. The ‘43Nd/ ‘44Nd ratios of the leachates do not show a simple linear relationship with 1 /Nd values; thus, no simple two-component mixture can be invoked. The chemical composition of the two least metamorphic samples 2988 and 2990 indicates that the REE are hosted by a phosphate phase, probably apatite, which yields a hump-shaped REE-pattern (Fig. 3a) and high Sm/Nd ratios. The leachates of the higher grade samples 299 1, 2995, and 3014 are characterized by high concentrations of Fe and REE patterns that are enriched in LREEs, implying low Sm/Nd ratios. The leachate reference line is, therefore, the result of mixing apatite and Fe-hydroxide-hosted Sm and Nd, yielding the range in Sm/Nd from 0.09-0.21, but having the same initial neodymium isotopic composition. Samarium and neodymium in the residues are considered to reside in crystallographic positions of the insoluble silicate phases. A considerable portion is, however, included in phosphate, since only lo-40% of the apatite is dissolved during the acid treatment. The initial ‘43Nd/‘44Nd ratios of residues from size fractions ~0.4 pm are very similar to those of the leachates (Fig. 5b). We suggest that fine-grained illite may represent the isotopic composition of the diagenetic fluid, from which it precipitated, but we are aware that the residues may still contain considerable amounts of apatite. We conclude, therefore, that the diagenesis of argillaceous sediments can be dated by the Sm-Nd chronometer in authigenic accessory mineral phases that probably grew during diagenetic compaction in the pore space of the sediment (hereafter termed cement phases). The isotopic system of these minerals (apatite, Fe-hydroxide/oxide) was homogenized during authigenic mineral growth in a sediment that was flushed by diagenetic fluids and had abundant primary

1479

or secondary interconnected pore space. Early diagenetic dewatering seems to be an adequate process to explain largescale isotopic homogeneity as well as late diagenetic episodic dewatering of overpressured basin compartments (HUNT, 1990). The Hercynian metamorphic overprint caused the rehomogenization of the Rb-Sr system, indicated by ca. 300-340 Ma reference lines for residues, leachates, and untreated aliquots. The Sm-Nd isotopic system of apatite and Fe-hydroxide/oxide remained undisturbed during the metamorphic overprint, because ( 1) the fluid/rock ratios during metamorphism was small compared to that during the diagenetic dewatering; (2) the metamorphic fluid did not introduce large amounts of Nd with different isotopic composition; and (3) the cement minerals are highly enriched in REE and are, therefore, insensitive for contamination by Nd with different isotopic composition. Any (probable) dissolution /reprecipitation or recrystallization during Hercynian metamorphism did, therefore, not change the isotopic Nd composition of these minerals. The decoupling of the two isotopic systems is, therefore, mainly a mineralogical effect (carbonate vs. apatite and Fehydroxide). Fine-grained clay minerals ~0.2 pm, however, reflect isotopic homogenization for Nd 523 Ma ago and for Sr 300 Ma ago. This implies different crystallographic retentivity for Nd and Sr in sheet silicates such as illite and chlorite. CONCLUSIONS The Sm-Nd isotopic system of clay material from Cambrian shales collected in Northwestern Morocco is dominated by the leachable portion of Sm and Nd, which is mainly hosted in authigenic, cogenetic apatite, and Fe-oxide/ hydroxide. The leachates representing these minerals yield a 523 f 72 Ma reference isochron, which agrees with a depositional age of 536-5 17 Ma. A variation in Sm/Nd ratios is crucial for the calculation of a best-fit line and results from mixing cogenetic diagenetic minerals with different Sm/Nd ratios. Clay size fractions ~0.2 ym (and in some cases 0.20.4 rm) are in isotopic equilibrium with the leachates, whereas coarser fractions contain inherited components. The Rb-Sr isotopic system achieved near-equilibrium conditions during Hercynian metamorphism, yielding reference isochron ages between 309 and 348 Ma. Fine-grained clay fractions ~0.2 pm of the two least metamorphic samples are in equilibrium with their leachates and reflect strontium isotopic homogeneity among the clay minerals (illite, chlorite) and soluble/exchangeable Sr, mainly in carbonate. The metamorphic fluid was probably buffered by the pre-existing diagenetic carbonate and exhibits apparently constant isotopic composition over a large area. The present study clearly demonstrates that the Sm-Nd system of authigenic accessory minerals such as phosphates and oxides/ hydroxides in clay fractions can be used as a tool to date diagenesis in argillaceous sediments. Isotopic equilibrium conditions over large areas imply the presence of an isotopically homogeneous diagenetic fluid over a large area, high water/rock ratios and the existence of abundant interconnected pore space. The Hercynian metamorphic overprint caused isotopic reequilibration of the clay-hosted Sr with the metamorphic

1480

U. Schaltegger et al.

fluid, whereas the neodymium isotopic composition in apatite, Fe-oxide/hydroxide, and clay minerals remained unchanged. The Sm-Nd isotopic system was not disturbed during the metamorphic overprint due to small fluid/rock ratios and low REE contents of the metamorphic fluid compared to the authigenic minerals. Low-grade metamorphic overprinting, therefore, results in the decoupling of the two isotopic systems: Sm-Nd was not affected by the metamorphic fluids and still reflects diagenetic equilibrium, while Rb-Sr was re-equilibrated during metamorphism 200 Ma after diagenesis. Acknowledgmenfs-We would like to thank J. L. CCzard, B. Kiefel, G. Krempp, Ph. Larqut, G. Morvan, R. Rouault, J. Samuel, D. Tisserant, and Robert Wendling for help and technical assistance provided throughout this study. M. Steinmann and H. Zwingmann helped with data evaluation and computer programming. S. Fortier is thanked for correcting the English. The paper benefits from comments of F. Gauthier-Lafaye and S. Chauduri and careful reviews of S. Hemming, J. X. Zhao, and an anonymous referee. The very helpful and positive comments of J. X. Zhao and the editorial handling of S. R. Taylor are particularly acknowledged. The research of US at the Centre de Gkochimie de la Surface in Strasbourg is supported by a fellowship of the Schweizerischer Nationalfonds zur FGrderung der wissenschaftlichen Forschung, which is kindly acknowledged. Editorial handling: S. R. Taylor REFERENCES ARONSONJ. L. and HOWERL. ( 1976) Mechanisms of burial metamorphism ofargillaceous sediment. 2. Radiogenic argon evidence. Grol. Sac. Amer. Bull. 87, 738-744. AWWILLERD. N. and MACK L. E. ( 1991) Diagenetic modification ofSm-Nd model ages in Teriary sandstones and shales, Texas Gulf Coast. Geology 19, 3 I l-3 14. BROSR., STILLEP., GAUTHIER-LAFAYE F., WEBERF., and CLAUER N. ( 1992) Sm-Nd dating of Proterozoic clay materials: An example from the Francevillian sedimentary series (Gabon). Earth Planet. &i. Lell. 113, 207-2 18. CHAUDHURIS., ST~LLEP.. and CLAUERN. ( 1992) Sm-Nd isotopes in fine-grained elastic sedimentary materials: Clues to sedimentary processes and recycling growth ofthe continental crust. In Isolopic Signatures and Sedimentar!: Records (ed. N. CLAUER and S. CHAUDURI)pp. 287-319, Springer-Verlag. CLAUERN. ( 1979) A new approach to Rb-Sr dating of sedimentary rocks. In Lectures in Isotope Geology (ed. E. J.&GERand J. C. HUNZIKER).pp. 30-5 I. Springer-Verlag. CLAUERN. ( 1982) The rubidium-strontium method applied to sediments: certitudes and uncertainties. In Numerical Dating in Sfra/igraphy (ed. G. ODIN), pp. 245-276. Wiley. CLAUERN., JEANNETTE D., and TISSERANT D., et al. ( 1980) Datation isotopique des cristallisations successives d’un socle cristallin et cristallophyllien ( Haut-Moulouya, Moyen-Maroc). Geol. Rundschau 169, 63-83. CLAUERN., CHAUDURIS., KRAL~KM., and BONNOT-COURTOIS C. ( 1993) Effects of experimental leaching on Rb-Sr and K-Ar isotopic systems and REE contents of diagenetic illite. Chem. Geol. 103, I-16. CLAUERN., RAE N., SCHALTEGGERU., and PIQUE A. (1994) KAr isotope systematics of clay fractions in greywacke-type sediments during a multi-stage low-grade metamorphic evolution. Chem. Geol. (submitted). DEPAOLOD. J. ( 1988) Neodymium Isolope Geochemistry. Miner& und Rocks, Vol. 20. Springer-Verlag. FIECHTNERL., FRIEDRICHSEN H., and HAMMERSCHMIDT K. ( 1992) Geochemistry and geochronology of Early Mesozoic tholeiites from Central Morocco. Geol. Rundschau 81, 45-62.

GOLDSTEINS. L., O’NIONS R. K.. and HAMILTONP. J. (1984) A Sm-Nd isotopic study of atmospheric dust and particulates from major river systems. Earth Planet. Sci. Lett. 70, 22 l-236. GROMETL. P., DYMEKR. F., HASKINL. A., and KOROTEVR. L. ( 1984) The “North American shale composite”: its compilation. major and trace element characteristics. Geochim. Cosmochim. Acta 48,2469-2482. HARLANDW. B., ARMSTRONG R. L., Cox A. V., CRAIGL. E., SMITH A. G., and SMITH D. G. (1990) A geologic,al time scale 1989. Cambridge Univ. Press. HUNZIKERJ. C. ( 1979) Potassium argon dating. In Lectures in Isotope Geo/ogJJ(ed. E. J;~GERand J. C. HUNZIKER),pp. 53-76. SpringerVerlag. HUON S., PIQUE A., and CLAUERN. (1987) Etude de l’orogen&se hercynienne au Maroc par la datation K-Ar de I%volution mttamorphique de schistes ardoisiers. SC?. G&l. BuII. 40, 273-284, Strasbourg. HUON S., CORN~E J. J.. PIQUEA., RAIS N., CLAUERN., LIEWIGN., and ZAYANER. ( 1993) Mise en evidence au Maroc d’kvenements d’age Triasico-Liasique lib g I’ouverture de 1’Antlantique. Bull. G&01.Sot. France 164, 165-176. HUNT J. M. (1990) Generation and migration of petroleum from abnormally pressured fluid compartments. Amer. Assoc. Petrol. Geol. Bull. 74, I - 12. KROGH T. E. and KEPPIE J. D. (1990) Age of detrital zircon and titanite in the Meguma Group, southern Nova Scotia, Canada: Clues to the origin of the Meguma Terrane. Tectonophysics 177, 307-323. KOBLERB. ( 1966) La cristallinitt de I’illite et les zones tout g fait supkrieures du mCtamorphisme. In Colloque sur les Etages Tecfonique.p,pp. 105-122. A la BaconniZre, NeuchPtel. LECOINTREG. ( 1926) Recherches g&ologiques dans la Meseta maroCaine. Mem. Sue. SC. Nat. Maroc 14. LIEWIGN., CLAUERN., and SOMMERF. (1987) Rb-Sr and K-Ar dating of clay diagenesis in Jurassic sandstone reservoirs, North Sea. Amer. Assoc. Perrol. Geol. Bull. 11, 1467-1474. LUDWIGK. ( 1988) Isoplot, Version 2.03. b’SGS Open-File Rep. 51% 55 7. MARIANOA. N. ( 1989) Economic geology of rare earth minerals. In Geochemisrrv and Mineralogy ofRare Earth Elements (ed. B. R. LIPINand G. A. MCKAY); Rev. Mineral. 21, 309-337. MCCULLOCHM. T. and WASSERBURG G. J. ( 1978) Sm-Nd and RbSr chronology of continental crustal formation. Science 200, 1003101 1. MICHARDA., GURRIE~P., SOUDANTM., and ALBAR~DEF. ( 1985) Nd isotopes in French Phanerozoic shales: external vs. internal aspects of crustal evolution. Geochim. Cosmochim. Acta 49,60 I 610. MRINIZ., RAFIA., DUTHOUJ. L., and VIDALP. ( 1992) Chronologie Rb-Sr des granitoides hercyniens du Maroc: constquences. Bull. SOL..G&01.France 163, 28 l-29 I N.&CLERTH., SCHAFERH. J., and GEBAUERD. ( 1992) A Sm-Nd isochron on pelites 1 Ga in excess of their depositional age and its possible significance. Geochim. Cosmochim. Acta 56, 789-795. OHR M., HALLIDAYA. N., and PEACORD. R. ( 199 I ) Sr and Nd isotopic evidence for punctuated clay diagenesis, texas GulfCoast. Earth Planet. Sci. Lett. 105, I 10-126. O’NIONS R. K., HAMILTONP. J., and HOOKERP. J. (1983) A Nd investigation of sediments related to crustal development in the British Isles. Earth Planet. Sci. Left. 63, 229-240. PALMERM. R. and ELDERFIELDH. ( 1986) Rare earth elements and neodymium isotopes in ferromanganese oxide coatings of Cenozoic foraminifera from the Atlantic Ocean. Geochim. Cusmochim. Acla 50,409-4 17. PIQUE:A, ( 1979) Evolution structurale d’un segment de la chaine hercynienne: La Meseta Marocaine Nord-Occidentale. Sci. Ggol. M&m. 56. RAIS N. ( 1992) CaractCrisation min&ralogique, cristallochimique et isotopique ( K-Ar) d’un mCtamorphisme polyphas& de faible intensitt. Le cas des grauwackes Cambriennes du Maroc Occidental. Unpubl. PhD thesis, Univ. of Brest. REUTER A. ( 1987) Implications of K-Ar ages of whole-rocks and

Dating argillaceous sediment by Rb-Sr and Sm-Nd grain-size fractions of metapelites and intercalated metatuffs within an anchizonal terrane. Contrib. Mineral. Petrol. 97, 105-l 15. SHAW H. F. and WASSERBURG G. J. ( 1985) Sm-Nd in marine carbonates and phosphates: Implications for Nd isotopes in seawater and crustal ages. Geochim. Cusmochim. Acta 49, 503-5 18. STILLEP. and CLAUERN. ( 1986) Sm-Nd isochron-age and provenance of the argillites of the Gunflint Iron Formation in Ontario, Canada. Geochim. Cosmochim. Acta SO, 1141-I 146. STILLEP., GAUTHIER-LAFAYE F.. and BROSR. ( 1993) The neodymium isotope system as a tool for petroleum exploration. Geochim. Cosmochitn.

Acta 57, 4521-4525.

TAYLORS. R., M~LENNAN S. M., and MCCULLOCHM. T. (1983)

1481

Geochemistry of loess, continental crustal composition and crustal model ages. Geochim. Cosmochim. Acta 47, 1897- 1905, TUREKIANK. K., KATZ A., and CHAN L. (1973) Trace element trapping in pteropod tests. Limnol. Oceanogr. 18, 240-249. WYBRECHTE. J. ( 1985) Front supCrieur de la schistositt et cristallisations associCes. Exemple des Grauwackes Cambriennes du Maroc Occidental. PhD thesis, Univ. Strasbourg. ZHAOJ. X., MCCULLOCHM. T., and BENNETV. C. ( 1992) Sm-Nd and U-Pb zircon isotopic constraints on the provenance of sediments from the Amadeus Basin. central Australia: Evidence for REE fractionation. Geochim. Cosmochim. Acta 56, 921-940.