Karson, J.A., Cannat, M, Miller, D.J., and Elthon, D. (Eds.), 1997Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 153

26. DATA REPORT: GEOCHEMISTRY AND MINERAL CHEMISTRY OF ULTRAMAFIC ROCKS

FROM THE KANE AREA (MARK)1

Carl-Dietrich Werner2 and Joachim Pilot2

ABSTRACT

A suite of serpentinized peridotites from Leg 153 was analyzed for major and trace elements. All the rocks are less-refrac-tory peridotites. The mineral chemistry of olivines (Mg# = 90.3), orthopyroxenes (Mg# = 90.4), clinopyroxenes (Mg# = 91.6),and some Cr spinels (Mg# = 68.4; Cr# = 29.7) shows the characteristic features of minerals from less-depleted mantle perido-tites. Some geothermometric calculations gave temperatures for the pyroxene system of about 1300°C, reflecting the prekine-matic to synkinematic crystallization of the pyroxenes, and gave a range from 1180° to 1075°C for the mineral pairorthopyroxene and spinel. Oxygen-isotope determinations show δ 1 8 θ values of about 6%c for olivine, orthopyroxene, and cli-nopyroxene; δ 1 8 θ values from chromian spinel/magnetite vary between 2.9%e and 3.8%o, values from serpentine vary between2.8‰ and 5.3%e, and whole-rock values vary between 2.9%e and 4.6‰. Two temperatures could be calculated from the oxygen-isotope exchange for the mineral pair clinopyroxene and spinel: about 1100° and about 900°C. The 1100°C value nearly corre-sponds to the data obtained using the orthopyroxene-spinel mineral thermometer, and may indicate subsolidus deformation andrecrystallization. The temperature range of the serpentinization could only be estimated. It depends upon the rock/water ratio,and may not have exceeded ~300°C. Finally, a calculation of possible rock/water ratios is given.

INTRODUCTION SAMPLE MATERIAL

Serpentinized peridotites are exposed in a belt about 2 km wideand more than 20 km long on the western median valley wall of theMid-Atlantic Ridge south of the Kane Fracture Zone (MARK) at23°20' N. This outcrop is interpreted as a part of the oceanic uppermantle uplifted during tectonic extension of the slowly spreadingMid-Atlantic Ridge (Cannat et al, 1995; Karson and Winters, 1992;Purdy and Detrick, 1986).

Shipboard documentation has shown that the ultramafic complexfrom Site 920 (Leg 153) can be divided into several "units," whichdiffer especially in their fabrics and slightly in their degrees of ser-pentinization. All samples are moderately to strongly serpentinizedharzburgites with 15%—30% relict minerals (Cannat, Karson, Miller,et al., 1995).

Polished thin sections were used for microscope investigationsand also for microprobe work, which was conducted on the "CAME-BAX S 50" at the Ruhr Universitàt, Bochum.

Major-element oxides and trace elements (Ga, Pb, Cu, Zn, Y, Nb,Zr, Ti, V, Ni, Co, Cr, Se, S, and Cl) were measured by X-ray fluores-cence spectroscopy (XRF) using a Philips PW 1480 sequential spec-trometer. The XRF was used to analyze both pressed pellets and glassdiscs prepared from powdered, but not ignited, material with lithiumtetraborate. Na, K, Li, Rb, and Zn were analyzed by flame photome-try, Sr and Ba by inductively coupled plasma-atomic emission spec-troscopy (ICP-AAS) using a Perkin-Elmer PLASMA 1000 spectrom-eter, and As and Sb by inductively coupled plasma-mass spectrome-try (ICP-MS) using a Perkin-Elmer ELAN 5000 spectrometer. Boronwas determined by optical spectroscopy, with boron-free spectralcarbons, using a Carl Zeiss PGS 2 spectrograph.

δ 1 8 θ was determined from whole-rock samples and from mineralseparates of olivine, orthopyroxene, clinopyroxene, serpentine, andchromian spinel/magnetite.

Two samples from Hole 920B and 8 samples from Hole 920D, to-gether with an Ocean Drilling Program (ODP) internal laboratorystandard ("Ultra"), were chosen for geochemical investigation. Twofurther samples from Hole 920D were used only for microprobe in-vestigations. The special sample designations correspond to the ship-board documentation (Cannat, Karson, Miller, et al., 1995), but sim-pler numbers, defined in Table 1, are used in most of the tables andparts of the text.

WHOLE-ROCK GEOCHEMISTRY

The results of the chemical investigations are compiled in Table2. The CIPW (after the authors Cross, Iddings, Pearson, Washington)norm calculations (converted to 100% and with total Fe as Fe2+) arelisted in Table 3. Generally, the chemical and mineralogical compo-sitions of the investigated ultramafic rocks are within the ranges pre-viously reported for the Kane region by Aumento and Loubat (1971),Cannat et al. (1992), Dick et al. (1984), Michael and Bonatti (1985),and Miyashiro et al. (1969). The chemical and mineralogical datashow the characteristic features of serpentinized harzburgitic ultra-

Table 1. Identification of sample numbers.

'Karson, J.A., Cannat, M, Miller, D.J., and Elthon, D. (Eds.), 1997. Proc. ODP, Sci.Results, 153: College Station, TX (Ocean Drilling Program).

2TU Bergakademie Freiberg, Institut für Mineralogie, D-09596 Freiberg, FederalRepublic of Germany.

Samplenumber

1

68

10111213141516

Hole, core, section, piece

920B-8R-4, Piece 11902B-13R-1, Piece 2C920D-3R-l,Piece6920D-5R-l,Piece4920D-14R-4, Piece 6920D-16R-1, Piece 5D920D-16R-6, Piece 10920D-19R-l,Piece5920D-20R-5, Piece 4920D-21R-2, Piece 5B920D-22R-2, Piece ID920D-22R-4, Piece IB

Depth(mbsf)

75.5117.618.237.1

119.4134.4140.3162.7177.6183.9192.1193.7

457

DATA REPORT

Table 2. Chemical composition of ultramafic rocks from Leg 153.

Sample no.:

Hole:

Core:

Section, piece:

Depth (mbsf):

1

920B

8R

4, 11

75.5

Major elements (wt%)SiO2TiO2A12O3

Cr2O3Fe2O3

MnOMgONiOCaONa2OK2OP2O5LOITotal

39.50.021.130.347.50.119

38.40.270.240.090.0270.011

12.399.95

Trace elements (ppm)

BLiRbSrBaGaPbCuZnYNbZrTiVNiCoCrSeSAsSbCl

255

<l2.850.31_

~31636~l_4

12046

2150102

235014

8650.220.24

1095

3

920B

13R

1,2C

117.6

38.150.0111.070.357.920.112

38.80.270.690.0560.0090.010

12.8100.25

101_2.891.62

<l4

1738<l

_3

6544

2135101

241010

5250.710.25

3795

5

920D

3R

1.6

18.2

38.70.021.20.348.150.087

38.50.270.140.060.0320.010

12.5100.01

24__2.470.70

<24

2436<l~ 1~3

12047

210598

233511

8900.420.32

855

6

920D

5R

1.4

37.1

37.90.0181.210.397.920.12

38.90.270.060.060.0180.011

13.099.88

21__0.930.53

<2_

4354_

<l3

11048

2130102

269011

8700.400.07

830

8

920D

14R

4.6

119.4

39.30.0081.30.388.310.12

38.650.280.940.050.0200.011

10.599.87

91_1.451.25_6

1947~ 1<l

45048

2180110

262014

4251.380.34

2210

10

920D

16R

1,5D

134.4

38.30.0131.120.368.090.11

38.70.290.830.090.0090.011

12.2100.12

5__1.210.43_

~227

~l_

~38043

2250108

249011

5400.21_

2310

12

920D

19R

1.5

162.7

39.00.0141.260.398.260.124

37.70.290.300.070.0040.007

12.599.92

211_4.650.74

<l5

1540<l<l

48548

2250112

269012

10502.00.04

1000

13

920D

20R

5.4

177.6

39.20.0241.310.377.930.113

37.80.270.650.080.0120.012

12.6100.37

81

<l2.281.01-

<21737-

<l4

14549

2145101

250012

8852.30.40

2455

14

920D

21R

2,5B

183.9

38.10.021.150.377.950.12

38.80.280.350.070.0120.010

12.799.93

242_2.590.98

<27

2144<l<l

412051

220599

25508

7700.420.52

1480

16

920D

22R

4, IB

193.7

39.40.0081.20.418.040.122

38.80.271.180.080.0090.010

10.7100.23

9

~l2.470.95_

~2

46-_5

5048

215099

27809

3801.180.22

2610

Ultra

38.80.021.180.3658.340.103

37.60.2860.130.0450.0040.005

13.099.82

18

_2.241.9-

<31240

-<l

312050

2250109

250011

9900.71-

735

Notes: - = not detected.

Table 3. CIPW norm of the ultramafic rocks from Leg 153.

Sample no.:

COrAbAnOpxCpxOlCm11ApTTIAn*Mg#Ni/CoCr/NiTi/ZrTi/CrV/CrCu/Zn

1

0.630.180.871.28

27.31-

69.080.58(1.040.031.05

6090.921.1

1.09300.0510.0200.44

3

_0.060.553.05

17.710.66

77.330.590.020.030.61

8590.521.1

1.13220.0270.0180.45

5

0.960.220.590.74

24.48-

72.360.580.040.030.81

5690.221.5

1.1140

0.0510.0200.67

6

1.160.120.590.29

21.37-

75.740.660.040.030.71

3390.520.9

1.26360.0410.0180.80

8

_0.130.483.68

19.821.22

73.990.630.020.030.61

8890.119.81.20

270.0190.0180.40

10

_0.060.873.00

16.521.31

77.270.910.030.030.93

7890.320.8

1.11260.0320.0170.64

12

0.690.030.691.70

26.59-

69.590.660.020.030.72

7189.920.1

1.20210.0320.0180.38

13

_0.080.773.61

21.33-

73.500.630.050.030.85

8290.521.2

1.16360.0580.0200.46

14

0.480.080.691.93

19.75-

76.370.630.040.030.77

7490.522.3

1.16300.0470.0200.48

16

_0.060.763.23

18.312.61

74.300.680.020.030.82

8190.421.7

1.29100.0180.0170.48

Ultra

1.060.030.240.71

28.73-

68.540.630.050.010.27

7589.820.6

1.1140

0.0480.0200.30

Notes: C = corundum, Or = orthoclase, Ab = albite, An = anorthite, Opx = orthopyroxene, Cpx = clinopyroxene, Ol = olivine, II = ilmenite, Ap = apatite, An* = normative anorthitecontent in Plagioclase, - = not determined, other abbreviations are defined in the text.

mafic rocks, with Mg# = (Mg/[Mg + Fe]) between 89.8 and 90.9. TheThornton-Tuttle Index (TTI) is generally below 1 (Table 3), and thedominant norm mineral of the H2O-free calculated rocks is olivine(ca. 70-80 wt%), followed by orthopyroxene (18-29 wt%). Com-pared to the modal data (between 2 and 5 vol%), normative clinopy-roxene contents are generally too low. Plagioclase was not observedin any of the investigated thin sections. Actually, the relatively smallamounts of Na, Ca, and Al in normative albite (Ab) and anorthite

(An) are incorporated within the pyroxenes, including parts of nor-mative corundum. Another part of Al is a constituent of the spinels.

Geochemical data from our samples suggest a moderate degree ofdepletion by partial melting, clearly lower than in more northern partsof the Mid-Atlantic Ridge, for instance, the Atlantic or Oceanogra-pher Fracture Zones (Michael and Bonatti, 1985; Dick et al., 1984).The water-free, calculated average of all our analyses (n = 11) is near-ly identical to the mean value of analyses from the less-refractory

458

DATA REPORT

peridotite of Michael and Bonatti (1985) for the major-element ox-ides SiO2, Fe2O3, and MgO. This indicates such an insufficient mo-bility of Si, Fe, and Mg that the bulk chemistry of the protolith wasessentially not altered by partial melting and serpentinization. Someminor oxides, however, show lower concentrations in comparisonwith the Michael and Bonatti (1985), less-refractory peridotite: Cr2O3

is 86%, A12O3 is 62%, TiO2 is 60%, CaO is only 31%, and Na2O isclearly higher, at 155%. The striking differences among A12O3, CaO,and Na2O can be explained mainly by serpentinization; this fact is notsurprising, because Michael and Bonatti (1985) considered the lossof some components through hydration in their model calculations.

As expected, the data from the three essential oxides (MgO, SiO2,and Fe2O3) show a very low relative variance (V): MgO (V = 1.25%),SiO2 (V = 1.5%), and Fe2O3 (V = 2.95%). The subordinate compo-nent A12O3 shows a V = 6.3%. The most variable oxide is CaO (V =75%), followed by K2O (V = 64%), TiO2 (V = 34%), Na2O (V =23%), and P2O5 (V = 21%). Of the trace elements, Ni, Co, V, and Crshow low variability (V = 2.5%-5.7%). The highest variance valuesare from Cl (V = 56%), B (48%), S (32%), As and Sb (V ~ 80%), andto Ba, Cu, and Sr (V = 40%-65%). The latter elements were especial-ly influenced by subseafloor metamorphism, which leads to theirsupply by serpentinization, as for Ca, K, and Na (which are suppliedor lost, depending upon the governing temperature conditions, andthe varying rock/water relations). Conversely, the concentrations ofthe compatible elements are very uniform in all of the samplesthroughout the profile, and even ratios such as Ni/Co, Cr/Ni or V/Crare very similar (Table 2). However, there is no good correlation be-tween loss on ignition (LOI) and Cl (r = -0.333, where r = correlationcoefficient). Cl is most probably incorporated in serpentine minerals,whereas Ca was irregularly lost during alteration of clinopyroxeneand orthopyroxene. The result of this Ca loss is reflected in the valuesof the CIPW norm, with only four clinopyroxene-bearing samples(Table 2).

In general, the incompatible-element contents are very low, andare strongly depleted relative to primitive mantle composition (Hof-mann, 1988). The following percentages were calculated from aver-ages of 11 ultramafic rocks (Hofmann, 1988): Na = 21%, K = 46%,Sr = 13%, Ba = 16%, Ti = 9%, and Zr = 38%. For instance, an averageof 2.4 ppm Sr is much lower than the Sr content of seawater. The de-pletion of the incompatible elements may result not only from hydra-tion, but also, to a certain extent, from partial melting processes. Ob-viously, primary deviations in the MARK mantle peridotite compo-sition relative to the worldwide mean composition cannot be ruledout.

MINERAL CHEMISTRY

Olivine

This study dealt only with the relict minerals of samples fromHole 920D. All thin sections investigated contained a comparablyhigh percentage (10-20 vol%) of relatively fresh olivine grains thatcan be divided into two grain-size groups: (1) elongated, "augen"-like grains, 3-5 mm long, and (2) more or less rounded grains, witha diameter of 1-2 mm.

Both olivine groups are prekinematic or synkinematic. Accordingto Cannat et al. (1992), the elongation is the result of a strong plasticstrain with translation, kink banding, and granulation under astheno-spheric conditions. Recrystallization of the small grains and their par-tial kink-band formation, however, occurred in a lower temperatureenvironment.

Microprobe analyses of 83 olivines from both groups (Table 4)show no significant differences in their chemical composition. For-sterite (Fo) varied only between 89.4% and 90.2% or, if Ni2(Si04) isincluded, between 89.8% and 90.6% (mean values are 89.8% and90.2%, respectively). Mg# varied between 89.9 and 90.7 (average90.26). NiO varied between 0.29 wt% and 0.49 wt% (average 0.39wt%). This variation is considerably higher than that of the MgO con-tent, and was probably caused by internal displacement during tec-

tonic deformation and/or a hydrothermal influence. Some olivinegrain profiles show unsystematic patterns: the core can be Ni rich orNi poor, the rim may show the opposite behavior, but irregular or di-rected distributions across the grain also occur. FeO, MgO, and MnO,as well as Mg#, show equivalent features. Contents of TiO2, A12O3,and Cr2O3 are very low, and the oxides are sometimes absent alto-gether. The CaO concentration varies between 0.01 wt% and 0.09wt%. All of these features are characteristic of olivines from oceanicultramafic rocks, according to Hamlyn and Bonatti (1980).

The mean values of our olivines (including Mg#) are identical tothose reported by Michael and Bonatti (1985) for olivines from less-refractory peridotites from the North Atlantic part of the Mid-Atlan-tic Ridge, and, therefore, correspond well with the geochemical fea-tures of the whole rocks.

The more-refractory olivines from some more northern parts ofthe Mid-Atlantic Ridge contain higher Fo contents and, therefore, ex-hibit higher Mg# values. The same is true for olivines from a moresouthern region (Cannat et al., 1992). The host rocks of these olivinesare more-refractory peridotites than in the MARK region.

Orthopyroxene

Three main groups of orthopyroxene can be distinguished micro-scopically:

1. More or less rounded porphyroclasts (up to 10 mm long), withkink bands and clinopyroxene exsolution lamellae. Generally,the lamellae are partly, sometimes totally, transformed to ser-pentine. The outer rims are often replaced by talc and/orchrysotile.

2. Single grains (up to 2 mm diameter) within the matrix. Thegrains show undulatory extinction, kink banding, and containvery small clinopyroxene exsolutions.

3. Polygonal grain aggregates without deformation.

Orthopyroxene exsolution lamellae also occur in clinopyroxene.A total of 74 prekinematic and postkinematic orthopyroxenes in

thin sections from Hole 920 D was investigated using the electron mi-croprobe (Table 5). The data show only small differences in Mg#(90.0-90.8), but they show a considerable variability of A12O3 (3.5-4.9 wt%), Cr2O3 (0.55-1.15 wt%), and especially CaO (0.8-4.5 wt%)contents. TiO2, MnO, and NiO contents also vary considerably.

No striking differences exist among the samples investigated,aside from the orthopyroxene in thin section 11 (Sample 920D-16R-6, Piece 10), in which the contents of A12O3 and Cr2O3 are clearlyhigher.

Calculation of the quadrilateral molecular pyroxene proportionsaccording to the International Mineralogical Association proposal(Morimoto et al., 1989) shows in the more or less "normal" orthopy-roxene (n = 40; Mg# = 90.2, Ca# = 2.8, where Ca# = Ca/[Ca + Fe +Mg]):

Wollastonite (Wo) = 1.42-4.74 wt%, mean (x) = 2.83, standarddeviation (s) = ±0.95, V = ±33.6%

Enstatite (En) = 86.0-88.9 wt%, x = 87.60, s = ±0.91, V =±1.04%

Ferrosilite (Fs) = 8.9-11.3 wt%, x = 9.57, s = ±0.37, V = ±3.87%.

A few orthopyroxenes with higher Wo proportions (n = 8; Mg# =90.4, Ca# = 6.8) show the following:

Wo = 5.1-8.8 wt%, x = 6.81, s = ±1.44, V = ±21.1%En = 82.2-85.7 wt%, x = 84.26, s = ±1.38, V = ±1.64%Fs = 8.7-9.2 wt%, x = 8.93, s = ±0.166, V = ±1.86%

In both cases, standard deviation and variance are very low for Enand Fs, but Wo varies over wide ranges. The orthopyroxenes richerin Wo show an increase in their A12O3 content.

459

DATA REPORT

Table 4. Chemical composition of olivines from Hole 920D.

Sample no.:

No.:

n:

SiO2

TiO2A12O3Cr2O3FeOMnOMgONiOCaOTotalMg#FaFoNis

Sample no.:

No.:

n:

SiO2TiO2A12O3Cr2O3

FeOMnOMgONiOCaOTotalMg#FuFoNis

Sample no.:

No.:

n:

SiO2TiO2A12O3Cr2O3FeOMnOMgONiOCaOTotalMg#FaFoNis

Sample no.:

No.:

n:

SiO2TiO2A12O3Cr2O3FeOMnOMgONiOCaOTotalMg#FaFoNis

1

1

41.820.010.020.019.390.17

49.710.410.08

101.6290.4

9.7089.90

0.40

11

1

40.64-__9.910.18

49.470.410.02

100.6389.910.2389.37

0.40

21

1

41.98_0.010.019.110.06

49.620.430.02

101.2490.7

9.3690.22

0.42

31

1

40.83_0.020.029.250.12

49.040.390.02

99.6990.4

9.6589.97

0.38

2

Profile 1

2

41.59_

0.02_

9.660.11

49.260.310.06

101.0190.1

9.9889.710.31

1

12

3

1

41.46___9.330.19

49.030.360.07

100.4490.4

9.7989.86

0.35

0

13

Profile 4

1

41.46_

0.04-

9.250.17

48.700.300.05

99.9790.4

9.7789.93

0.30

22

Grains

4

41.200.010.020.019.680.14

49.170.400.04

100.6790.010.0389.57

0.40

13

32

Profile

6

40.690.010.020.019.500.13

49.210.410.03

100.0190.2

9.8689.750.39

2

41.34_0.030.029.110.13

49.350.350.03

100.3690.6

9.4390.25

0.32

12

23

1

41.29_0.04-9.700.14

49.450.350.05

101.0290.110.0189.65

0.34

33

1

40.800.020.020.039.270.15

49.500.320.02

100.1390.5

9.6290.06

0.32

4

2

40.92_0.030.029.040.13

49.460.410.06

100.0790.7

9.3890.22

0.40

14

1

41.660.020.010.029.470.15

49.630.320.02

101.3090.3

9.6490.01

0.35

24

Profile

2

41.510.010.01_9.500.13

49.590.380.04

101.1790.3

9.7689.86

0.38

14

34

1

40.570.020.030.059.260.13

49.130.340.03

99.5690.4

9.6690.00

0.34

5

10

6

Profile 2

1

41.41_0.02_9.580.19

49.460.490.08

101.2390.2

9.9489.58

0.48

15

4

41.360.010.030.029.420.12

49.470.410.06

100.9090.3

9.7289.87

0.41

25

1

41.300.020.03_9.730.11

49.530.390.05

101.1690.1

9.9989.63

0.38

35

Grains

1

40.850.020.02_9.070.14

48.850.460.03

99.4490.6

9.5290.02

0.46

2

41.37_0.020.039.660.13

49.360.440.07

101.0890.1

9.9789.59

0.44

1

16

7

3

41.23_0.030.069.440.16

49.320.390.07

100.7090.3

9.8289.80

0.38

1

17

Grain Profile

2

40.970.010.020.019.580.10

49.050.370.06

100.1790.1

9.9389.70

0.37

26

1

40.69_0.010.039.380.17

49.220.430.03

99.9690.3

9.7889.80

0.42

36

1

41.200.010.020.039.210.15

49.880.410.01

100.9290.6

9.5090.10

0.40

1

41.15___9.280.09

49.350.390.06

100.3290.4

9.5990.03

0.38

27

1

40.24_0.03-9.600.10

49.680.430.04

100.1290.2

9.8389.75

0.42

37

4

40.98_0.020.019.550.14

49.350.420.01

100.4890.2

9.8889.71

0.41

8

1

41.25_0.010.039.410.17

49.560.380.05

100.8690.4

9.7589.880.37

18

2

41.14_0.02_9.6!0.17

49.730.420.04

101.1390.2

9.8989.70

0.41

13

28

Grains

1

40.220.010.02-9.540.11

49.770.380.03

100.0890.3

9.7889.85

0.37

15

38

10

40.830.010.010.019.540.12

49.160.400.03

100.1190.2

9.9089.71

0.39

9

Profile 3

1

41.10_0.020.049.320.10

49.680.350.05

100.6690.5

9.5890.07

0.35

12

19

10

1

41.18_0.010.029.470.10

49.270.320.06

100.4390.3

9.8089.88

0.32

20

Grains

1

41.480.030.01_9.350.15

49.460.380.02

100.8890.4

9.7089.93

0.37

29

1

40.81__-9.410.11

49.090.390.03

99.8490.3

9.7789.84

0.39

39

Profile

3

40.920.010.01-9.440.15

49.690.410.02

100.6590.4

9.7389.87

0.40

1

41.35_0.010.029.560.18

48.920.290.04

100.3790.110.0389.68

0.29

30

1

40.93__0.029.460.12

49.170.370.02

100.0990.3

9.8289.81

0.37

40

7

40.700.010.02-9.640.14

49.180.410.03

100.1390.110.0089.60

0.40

41

3

40.820.020.01-9.420.15

49.250.420.05

100.1490.3

9.7989.790.42

Notes: No. = run number, n = number of analyses performed, Fo = forsterite, Fa = fayalite, Nis = Ni silicate (Ni2[SiO4]), - = not detected.

460

Table 5. Chemical composition of orthopyroxenes, Hole 920D (prekinematic and Porphyroclastic grains).

DATA REPORT

Sample no.:

No.:

IV.

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaOTotalMg#WoEnFs

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaOTotalMg#WoEnFs

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaOTotalMg#WoEnFs

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaOTotalMg#WoEnFs

10

1

5

55.970.023.680.766.090.16

32.680.111.63

101.1090.5

3.1487.48

9.38

11

2 3

Grains

2

55.850.034.050.926.050.17

32.460.111.66

101.3090.5

3.2187.41

9.38

12

Grains

2

54.890.053.550.695.760.15

31.760.082.43

99.3690.8

4.7486.26

9.00

21

1

55.150.073.830.756.150.16

32.900.051.10

100.1690.5

2.1288.37

9.51

32

1

55.000.083.870.725.890.05

32.030.132.11

99.8890.6

4.1086.87

9.03

1

55.06-3.600.556.350.16

32.810.181.40

100.1190.2

2.6887.58

9.74

22

Profile 1

1

55.110.033.840.735.760.13

30.970.174.16

100.9090.6

8.0183.14

8.85

14

33

Profile 3

2

55.250.043.830.766.200.18

32.730.081.23

100.3090.4

2.3788.02

9.61

2

55.440.044.241.026.020.16

32.160.111.37

100.5690.5

2.6887.85

9.47

13

2

55.160.033.760.746.140.13

32.630.091.80

100.4890.4

3.4587.16

9.39

23

1

55.620.023.830.705.960.13

32.860.091.07

100.2890.8

2.0888.71

9.21

34

2

54.890.053.880.845.730.11

31.920.092.42

99.9390.9

4.7186.42

8.87

4

2

55.250.034.461.006.340.16

32.670.120.73

100.7690.2

1.4288.68

9.90

14

1

55.690.023.600.736.250.14

33.430.050.84

100.7590.5

1.6188.86

9.53

24

1

55.450.053.820.726.120.18

32.870.131.06

100.4090.5

2.0588.46

9.49

35

1

55.540.073.770.656.060.13

32.940.091.14

100.3990.6

2.2088.48

9.32

11

5

2

55.450.034.470.996.410.14

32.580.140.86

101.0790.0

1.6788.38

9.95

13

15

2

55.210.043.770.786.160.11

32.980.161.06

100.2790.5

2.0488.53

9.43

25

1

55.300.064.070.886.370.12

32.420.121.33

100.6790.1

2.5787.60

9.83

36

2

55.320.043.650.806.240.15

32.430.101.45

100.1890.2

2.8187.52

9.67

6

1

55.420.084.601.066.240.15

31.750.152.12

101.5790.1

4.1386.16

9.71

16

Profile

2

54.720.013.820.866.370.13

32.980.080.81

99.7890.2

1.5688.65

9.79

14

26

1

55.010.084.190.906.120.15

32.710.060.98

100.2090.5

1.9188.57

9.52

37

7 8

Profile

1

55.21_4.621.155.830.13

31.120.102.57

100.7390.5

5.0885.72

9.20

17

2

54.170.013.620.686.280.12

32.780.111.24

99.0190.3

2.3987.99

9.62

27

Profile 2

1

54.640.044.000.875.950.17

31.730.122.32

99.8490.5

4.5286.17

9.31

38

Profile

1

55.020.053.670.775.900.15

31.600.132.94

100.2390.5

5.6885.19

9.13

3

55.410.033.640.736.260.14

32.750.111.36

100.4390.3

2.6287.77

9.61

1

54.550.084.931.035.730.09

30.010.124.48

101.0290.3

8.8182.25

8.94

18

2

54.230.053.770.786.090.12

31.760.082.32

99.2090.3

4.5286.05

9.43

28

1

54.210.074.390.945.720.14

30.770.123.66

100.0290.5

7.1683.89

8.95

15

39

5

55.530.043.560.726.390.15

32.700.091.34

100.5290.1

2.5887.60

9.82

9

3

55.450.044.490.986.330.14

32.650.101.18

101.3690.2

2.2887.95

9.77

19

1

54.360.063.910.845.680.11

30.600.094.26

99.9190.6

8.2982.91

8.80

29

1

54.560.064.160.965.920.17

32.230.192.14

100.3990.7

4.1386.70

9.17

40

2

55.790.073.610.766.260.12

32.480.141.71

100.9490.2

3.2987.13

9.58

13

10

Grain

3

54.800.033.710.736.230.16

32.900.081.18

99.8290.4

2.2688.14

9.60

14

20

Profile 1

1

55.100.033.850.766.260.17

32.620.091.60

100.4890.3

3.0787.29

9.64

30 31

Profile 3

2

54.900.073.760.825.800.07

31.740.102.80

100.0690.7

5.4285.70

8.88

41

Grains

2

55.760.033.720.756.140.13

32.580.091.68

100.8890.4

3.2487.34

9.42

2

55.130.063.830.776.210.18

32.850.111.07

100.2190.4

2.0788.33

9.60

42

1

55.130.033.860.866.050.19

31.830.172.16

100.2890.4

4.2086.33

9.47

Note: Abbreviations are defined in the text and in Table 4.

461

DATA REPORT

For both the orthopyroxene groups, no significant correlation be-tween CaO/Cr2O3 and Al2O3/CaO could be calculated. On the otherhand, A12O3 and Cr2O3 are very strongly correlated:

orthopyroxene (n = 40): r = +0.875; (3.5% Al2O3)/(0.69% Cr2O3)-(5.0% A12O3)/(1.14% Cr2O3),

orthopyroxene (n = 8): r = +0.883; (3.5% Al2O3)/(0.72% Cr2O3)-(5.0% A12O3)/(1.12% Cr2O3).

High CaO and Al2O3/Cr2O3 contents in orthopyroxene can best beexplained as the effects of high temperature in a dry system, onlyslightly influenced by pressure (Anastasiou and Seifert, 1972; Lind-sley and Dixon, 1975; Obata, 1976; Sachtleben and Seck, 1981). Us-ing the "Cr-Al-orthopyroxene" thermometer of Witt-Eickschen andSeck (1991), temperatures between 1245° and 1340°C, with an aver-age of 1275°C, were calculated for 8 selected orthopyroxenes. TheLindsley (1983) two-pyroxene thermometer, applied to 8 orthopy-roxene/clinopyroxene pairs, resulted in a range from 1250° to1300°C, and even the "classic" Kretz (1963; 1982) two-pyroxenethermometer showed values between 1170° and 1380°C, with an av-erage of ~1300°C. Only with the Wood and Banno (1973) methodwere higher temperatures (1450° ± 25°C) calculated.

Five grain profiles clearly show a tendency in the Fe and Mg dis-tribution toward higher contents in the marginal zone and lower con-tents in the core. The inverse is true for Ni, to some degree. The be-havior of Ca, Al, and Cr is irregular.

The microprobe results for a few polygonal grains and orthopy-roxene exsolutions in clinopyroxene are listed in Table 6. The ortho-pyroxene exsolutions have a relatively wide range in Mg#s, as wellas in the contents of A12O3, FeO, NiO, and CaO. In contrast, however,the polygonal grains more or less correspond to the Porphyroclasticorthopyroxenes, aside from their comparatively low CaO content.

A comparison with the data of Michael and Bonatti (1985) for or-thopyroxenes from less-refractory and more-refractory peridotitesfrom the northern section of the Mid-Atlantic Ridge shows a goodcorrespondence between the mean value of all our orthopyroxenesand that from less-depleted peridotite; this is also consistent with thegeochemical characteristics of MARK ultramafic rocks.

The orthopyroxene compositions of rocks from the 15°37'N re-gion in the Mid-Atlantic Ridge axial valley (Cannat et al., 1992),which include strongly depleted peridotites, for instance, are clearlymore refractory, with Mg# = 91.8 compared to a mean of Mg# = 90.2in orthopyroxene from the Kane area.

Table 6. Chemical composition of orthopyroxenes, Hole 920D (postkine-matic grains and exsolutions).

Sample no:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaOTotalMg#WoEnFs

43

12

44

Polygonal grains

3

56.060.033.530.726.370.13

33.100.040.99

100.9790.2

1.9088.36

9.74

3

56.340.043.630.736.260.14

32.970.171.13

101.4190.4

2.1788.24

9.59

45 46

13

47 48

Exsolutions in clinopyroxene

1

53.530.025.201.077.210.16

31.120.101.90

100.3188.5

3.7384.9811.29

1

54.690.033.740.786.380.18

32.740.131.13

99.8090.1

2.1887.95

9.87

1

54.620.073.860.816.050.11

32.340.071.55

99.4890.5

3.0187.63

9.36

1

54.990.073.790.835.580.16

31.600.053.12

100.1991.0

6.0485.26

8.70

Note: Abbreviations are defined in text and Table 4.

Table 7. Chemical composition of clinopyroxenes, Hole 920D (prekinematic and Porphyroclastic grains).

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaONa2OTotalMg#Ca*#WoEnFs

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaONa2OTotalMg#Ca*#WoEnFs

1

4

52.040.074.671.152.930.07

17.300.03

22.470.08

100.8191.348.345.9549.26

4.79

11

2

Grains

2

52.420.134.451.112.600.07

16.540.06

23.660.06

101.1091.950.748.4947.24

4.27

11

12

Profile

6

51.31-0.155.511.492.420.10

15.730.07

24.480.10

101.3692.052.850.6045.33

4.07

2

50.940.165.731.692.670.08

16.300.04

23.890.09

101.5991.651.3

49,0146.58

4.41

3

1

51.820.114.591.162.600.07

16.480.11

23.910.06

100.9191.951.048.8046.94

4.26

13

3

51.330.105.031.332.610.13

15.800.11

24.540.06

101.0491.852.750.3545.27

4.38

4

3

52.150.084.641.163.300.11

18.460.06

20.680.07

100.7190.944.642.1552.42

5.43

12

14

Grains

2

51.820.105.021.292.610.08

15.910.05

24.260.08

101.2291.652.350.0045.684.32

10

5

2

52.360.074.611.063.290.10

18.240.08

20.670.05

100.5390.844.942.4052.17

5.43

15

1

52.270.074.681.232.890.12

16.850.06

23.350.06

101.5891.249.947.4847.754.77

6

Profile

2

52.020.094.751.222.710.10

16.650.06

23.430.06

101.0991.650.347.9847.534.49

16

1

52.210.085.011.143.510.10

19.130.02

19.620.07

100.8990.742.439.9954.265.75

7

3

51.830.104.911.262.610.12

16.250.06

24.070.08

101.2991.751.649.2846.364.36

17

2

50.650.094.981.262.460.10

16.290.06

24.020.08

99.9992.251.449.3046.604.10

8

2

51.930.124.741.183.040.10

17.280.07

22.110.07

100.6491.047.945.4549.51

5.04

1

18

9

1

52.450.114.851.173.530.13

18.630.07

20.880.04

101.8690.444.642.0152.25

5.74

3

19

Grains

2

50.330.085.111.362.430.10

15.800.12

24.460.09

99.8892.152.750.4345.504.07

3

50.610.104.461.172.490.09

16.600.07

23.940.10

99.6392.250.948.7647.134.11

11

10

Profile

2

51.470.195.751.572.200.10

15.800.04

24.310.11

101.5492.852.550.5345.74

3.73

20

3

51.340.134.571.132.440.08

16.180.05

24.380.08

100.3892.252.049.8646.114.03

DATA REPORT

Clinopyroxene

Clinopyroxene is mainly developed as laminated porphyroclasts(up to 3-4 mm in size) with orthopyroxene exsolutions. Postkinemat-ic recrystallization formed a smaller proportion of fresh and colorlessgrains, which are sometimes twinned, and occur subordinately in anintragranular position. They might represent products of a subsoliduscrystallization process, but Nicolas et al. (1980) have discussed forthat a formation by a limited local melting event. Exsolution lamellaein orthopyroxene constitute a third group of clinopyroxenes.

Microprobe data from 82, mainly prekinematic or synkinematic,clinopyroxenes are listed in Table 7, and show a range in Mg# from90.0 to 92.8. These values are clearly greater than those from ortho-pyroxene. The high A12O3 contents vary from 3.75 to 5.75 wt%, andare reflected by normative Ts (= tschermakite, Mg or Ca tschermak-ite) from 7.9% to 16.9%. The main components, CaO (19.1-24.7wt%) and MgO (15.7-19.6 wt%), show a considerable variability, asdo FeO (2.25-3.85 wt%) and Cr2O3 (0.85-1.7 wt%). Contents of theminor-element oxides TiO2, MnO, NiO, and Na2O vary quite sub-stantially. The clinopyroxenes of thin section 11 (Sample 920D-16R-6, Piece 10) are distinguished by above average A12O3 and Cr2O3 con-tents, as in the corresponding orthopyroxene. All clinopyroxenes inthe samples investigated can be described as tschermakitic (chromi-an) diopsides.

The quadrilateral molecular pyroxene proportions, according tothe IMA proposal (Morimoto et al., 1989), show a main group of cli-nopyroxenes with higher, and a subgroup with clearly lower Wo con-tents:

Main group (n = 33; Mg# = 91.6):

Wo = 46-50.6 wt%, x = 49.03, s = ±1.35, V = ±2.75%En = 45.3-49.4 wt%, x = 46.71, s = ±1.22, V = ±2.61%Fs = 3.7-4.8 wt%, x = 4.26, s = ±0.27, V = ±6.39%.

Subgroup (n = 12; Mg# = 90.6):Wo = 38.6-45.5 wt%, x = 42.62, s = ±2.45, V = ±5.76%En = 49.5-55.1 wt%, x = 52.01, s = ±2.07, V = ±3.98%Fs = 4.8-6.3 wt%, x = 5.37, s = ±0.43, V = ±7.95%.

Standard deviations and variance for Wo and En are somewhatlower than for Fs, and the subgroup generally shows a greater com-positional range. The clinopyroxene grains of this group show a dis-memberment along cleavage planes and some dissolution features atthese interruption zones, with the formation of antigorite, tremolite/actinolite, and, in some places, brown amphibole. This has led to acertain loss of Ca from the central parts of the grains, but the meanA12O3 contents are nearly equal in both groups.

Peters (1968) has described a good correlation between Ca/(Ca +Mg) (defined as Ca*#, in Tables 7, 8) and the A12O3 contents in cli-nopyroxene from Alpine serpentinites. The following correlationswere calculated for the two clinopyroxene groups from Hole 920D:

clinopyroxene (n = 33): r = +0.481; Ca*# ranges from 46 at 3.94wt% A12O3 to 52 at 4.89 wt% A12O3,

clinopyroxene (n = 12): r = -0.204; Ca*# ranges from 40 at 4.83wt% A12O3 to 48 at 4.63 wt% A12O3.

This opposite behavior can be explained by the partial alteration ofthe clinopyroxene subgroup.

A12O3 and Cr2O3 exhibit a positive correlation, but not as strong asthat in the orthopyroxene system, and both groups differ noticeably:

Table 7 (continued).

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaONa2OTotalMg#Ca*#WoEnFs

Sample no.:

No.:

n:

SiO2

TiO2

A12O3

Cr2O3

FeOMnOMgONiOCaONa 2OTotalMg#Ca*#WoEnFs

21

2

51.000.124.921.372.420.07

16.020.04

24.120.08

100.1692.252.049.8646.12

4.02

31

2

52.190.125.181.333.850.11

19.530.07

19.080.06

101.5290.041.238.6355.10

6.27

14

22

Grains

2

51.510.103.770.872.740.13

16.920.06

23.530.08

99.6991.750.047.6947.78

4.53

32

2

51.300.105.201.392.990.09

17.340.09

22.170.13

100.8091.247.945.4649.60

4.94

23

2

51.430.144.751.132.540.10

17.270.08

22.570.09

100.1092.448.446.3349.43

4.24

33

2

51.410.125.101.352.480.09

15.820.06

24.350.11

100.8991.952.550.3045.55

4.15

24

2

51.630.104.851.242.610.13

15.930.06

24.500.11

101.1691.652.550.1745.46

4.37

15

34

Grains

1

50.990.115.161.442.790.13

15.920.02

24.380.12

101.0691.152.449.9445.40

4.66

25

Profile

2

50.980.124.481.212.390.08

15.920.07

24.720.12

100.0992.252.750.6145.44

3.95

35

3

51.810.094.441.213.010.04

17.770.03

22.480.11

100.9991.347.645.3349.874.80

26

1

50.290.054.271.132.710.07

16.780.08

23.980.13

99.4991.750.748.3947.234.38

36

3

51.390.105.021.392.700.10

16.260.09

24.100.10

101.2591.551.649.2146.324.47

15

27

4

51.420.114.671.282.460.10

15.940.07

24.690.11

100.8592.052.750.4945.434.08

37

51.670.114.651.242.600.08

16.280.04

24.190.10

100.9691.851.749.4246.314.27

28 29

Grains

3

51.890.114.711.182.390.07

16.050.04

24.580.13

101.1592.352.450.3145.75

3.94

1

52.270.094.821.262.460.05

16.680.11

23.500.13

101.3792.450.348.2147.774.02

30

1

52.450.094.121.192.670.10

17.420.08

22.800.11

101.0392.148.546.2949.314.40

Note: Abbreviations are defined in Table 4 and in the text.

463

DATA REPORT

Table 8. Chemical composition of clinopyroxenes, Hole 920D (postkinematic grains and exsolutions).

Sample no.:

No.:

n:

SiO 2

TiO 2

A12O3

C r 2 O 3

FeOMnOMgONiOCaONa 2OTotalMg#C a * #WoEnFs

14

38 39 40

Exsolutions in orthopyroxene

1

51.950.133.910.992.250.08

16.690.04

24.290.08

100.4193.051.149.2247.10

3.68

1

51.680.124.49

1.152.480.10

16.390.03

23.970.05

100.4692.251.349.1246.76

4.12

1

50.970.135.111.342.700.08

16.130.11

23.830.14

100.5491.451.549.1146.42

4.47

41

1

51.770.094.411.133.330.11

18.900.09

20.560.08

100.4791.043.941.4453.13

5.43

15

42 43 44

Postkinematic, intergranular

1

52.170.084.781.203.250.09

18.130.15

21.780.08

101.7190.946.343.8150.95

5.24

3

51.730.124.431.162.880.10

17.310.08

23.020.10

100.9391.448.846.5248.78

4.70

2

51.210.074.231.073.000.07

17.580.10

22.250.09

99.6791.347.645.2549.87

4.88

13

45

1

51.900.014.391.103.450.08

19.65_

19.650.08

100.3191.041.839.5154.95

5.54

Note: Abbreviations are defined in the text and in previous tables.

clinopyroxene (n = 33): r = +0.333; (3.5% Al2O3)/(0.83% Cr2O3)-(5.5% Al2O3)/(1.50% Cr2O3),

clinopyroxene (n = 12): r = +0.768; (3.5% Al2O3)/(0.89% Cr2O3)-(5.5% A12O3)/(1.37% Cr2O3).

This mineral-chemical similarity could be the expression of com-patible conditions during crystallization.

Three grain profiles demonstrate the opposite distribution of MgOand CaO: the rims are MgO rich and CaO poor, whereas the cores areMgO poor and CaO rich. FeO seems to follow the trend of MgO, butA12O3, Cr2O3, and NiO show an indistinct behavior. The marginal de-crease of CaO may be caused by aqueous fluids at higher tempera-tures, and the subsequent loss of a Ca-rich solution.

The correspondence between the chemical composition of theKane clinopyroxene and the data of Michael and Bonatti (1985) forless-depleted peridotites is not very satisfactory, but the Mg# valueslie close together (Kane: Mg# = 91.3; Michael and Bonatti [1985]:Mg# = 90.8). Clinopyroxene from more-refractory ultramafic rocksshow Mg# values between 92.5 and 94 (Michael and Bonatti, 1985;Cannat et al., 1992). This is not surprising, when one considers thesensitivity of the relatively complex clinopyroxene system duringpartial melting, subsolidus, and hydration processes.

Some of the postkinematic intragranular clinopyroxenes (Table 8)exhibit relatively high A12O3, Cr2O3, and FeO contents, as well as alower TiO2 content and nearly equal Mg#s, but low Ca*# values.These features are consistent with the formation of the postkinematicclinopyroxenes in a lower temperature regime than that of the pre-kinematic clinopyroxenes.

Only three exsolution lamellae were investigated. They all havelow FeO but high CaO contents (high Ca*# values), and are relativelyrich in TiO2. The clinopyroxene exsolutions are enriched in TiO2,A12O3, and Cr2O3, in contrast with their host mineral, orthopyroxene.The lamellae are also somewhat enriched in MgO relative to FeO.Consequently, Mg# is slightly higher than it is in orthopyroxene. Wecan therefore state that clinopyroxene exsolution leads to a partial im-poverishment of Ti, Mg, Cr, and Al in orthopyroxene. The composi-tion of clinopyroxene exsolutions does not differ essentially from thecomposition of the prekinematic to synkinematic clinopyroxene,aside from some smaller Cr proportions.

Chromian Spinel

Cr spinel is an integral, but irregularly distributed, accessory min-eral in ultramafic rocks. It occurs mostly as larger, lobate, but moreor less fractured, grains in more or less altered olivines and orthopy-roxenes. The grains always show some dissolution and commonly

show a thin overgrowth of secondary magnetite, resulting from ser-pentinization. Results of a limited number of analyses from onlythree samples are compiled in Table 9, and the hypothetical mineralend members are calculated.

All grains investigated are picotites, with only small varianceswithin individual thin sections. Spinels from sample 11 (Sample920D-16R-6, Piece 10) are lower in Cr2O3 and somewhat higher inA12O3 than those from sample 15 (Sample 920D-22R-2, Piece ID). Inthin section 11, the spinels show normative hercynite. Calculationsfor sample 15, however, result in magnesiochromite (MCm) and fer-rochromite (Cm), with few exceptions. These differences do not sup-port the opinions of Christiansen (1985) and Nicolas et al. (1980)about a metamorphic origin of both types of Cr spinel. Generally, theCr spinels investigated are primary peridotite minerals, but in thinsection 15 (Cr# >30, where Cr# = Cr/[Cr + Al]), a certain recrystalli-zation at lower pressure should have modified the original composi-tion, because the negative correlation between Cr# in spinel andA12O3 in orthopyroxene is disturbed here. However, the two samplesare not sufficient for a substantial statement to be made.

The mean element ratios for the Cr spinel groups are as follows(Fe3+# = Fe3+/[Fe3+ + Cr + Al]):

sample 11 (n = 11): Cr# = 28.3, Mg# = 67.6, Fe3+# = 1.96,sample 15 (n = 18): Cr# = 30.6, Mg# = 68.2, Fe3+# = 1.37.

A geothermometric estimation of orthopyroxene-spinel pairs afterequation 6 of Sachtleben and Seck (1981) gave ~l 180°C for sample11, and ~1075°C for sample 15. Both temperatures are lower thanthose calculated using pyroxene thermometry, and may reflect a cer-tain element exchange between the spinels and the surrounding sili-cate minerals during subsolidus deformation and recrystallization.

OXYGEN-ISOTOPE INVESTIGATIONS FROMAN ULTRAMAFIC PROFILE, HOLE 920D:

ALTERATION BY SEAWATER

Oxygen-isotope investigations from oceanic environments werefirst performed on samples dredged from the ocean floor (for exam-ple, Muehlenbachs and Clayton, 1971, 1972a, 1972b; Wenner andTaylor, 1973), and later on samples from borehole material (Mueh-lenbachs, 1976, 1977, 1978, 1979). The main results are the follow-ing: (1) Unaltered mid-ocean ridge basalt (MORB) samples give δ 1 8 θ= 5.7%O ± 0.3%o (i.e., δ 1 8 θ = 5.7%O ± 0.2%0 from Muehlenbachs andClayton, 1972a, 1972b; Kyser et al. 1982; Ito et al., 1987; for the bulkmantle δ 1 8 θ = 5.5%o, from Mattey et al., 1994). (2) Alteration pro-

DATA REPORT

Table 9. Chemical composition of Cr spinels, Hole 920D.

Sample no.:

No.:

SiO2

TiO2

A12O3

Cr 2 O 3

Fe 2 O 3

FeOMnOMgONiOCaOTotalCr#Mg#Fe ? + #CmMCmMtHeSp

Sample no.:

No.:

SiO2

TiO2

A12O3

Cr 2 O 3

Fe 2 O 3

FeOMnOMgONiOCaOTotalCr#Mg#Fe 5 +#CmMCmMtHeSp

Sample no.:

No.:

SiO,TiO2

A12O3

Cr 2 O 3

Fe 2 O 3

FeOMnOMgONiOCaOTotalCr#Mg#Fe ? +#CmMCmMtHeSp

1

0.04-

41.3724.87

2.7413.90

0.2015.530.27_

98.9128.766.6

2.9227.95

_2.932.68

66.44

11

11

0.040.05

42.7224.78

1.2613.260.15

16.040.190.02

98.5028.068.3

1.3427.63

_1.333.20

67.84

21

0.070.04

40.7427.13

1.6713.550.17

15.780.220.01

99.3630.967.5

1.7830.50

-1.780.34

67.38

2

0.020.06

41.6324.48

3.4113.640.14

15.840.290.01

99.4928.367.4

3.6127.36

_3.621.61

67.41

12

12

0.040.01

41.3825.93

2.2612.230.15

16.670.23_

98.9029.670.8

2.4026.86

2.112.40_

68.63

22

0.070.09

41.2126.87

2.0213.220.17

16.100.250.01

100.0030.468.4

2.1329.49

0.532.15_

67.83

3

0.040.03

41.1425.10

2.2713.450.20

15.650.21_

98.0729.067.5

2.4428.42

_2.441.85

67.29

13

0.010.07

41.4826.90

1.5813.570.17

16.030.18_

100.0130.367.8

1.6629.92

_1.670.76

67.65

23

0.020.07

41.2227.52

1.2413.490.16

15.990.270.03

100.0130.967.9

1.3230.67

_1.320.21

67.80

4

0.020.07

42.0125.08

1.7013.710.21

15.640.250.02

98.6928.667.0

1.8228.20

_1.823.11

66.87

14

0.070.05

41.2527.23

1.6213.320.13

16.150.18_

100.0030.768.4

1.7030.95

0.341.77_

66.94

24

0.040.08

41.0427.19

1.0613.050.20

16.010.230.01

98.9130.868.6

1.1330.35

0.191.13_

68.33

11

5

0.050.04

41.5925.36

1.9613.720.14

15.700.130.01

98.7029.067.1

2.0928.57

_2.102.36

66.97

15

0.030.08

41.2826.85

1.7513.010.16

16.190.33_

99.6830.468.9

1.8529.24

0.731.86_

68.17

15

25

0.080.09

41.3726.79

1.7012.920.12

16.270.190.01

99.5430.369.2

1.8029.06

0.941.82_

68.18

6

0.060.05

42.7024.12

1.7313.180.18

15.980.20_

98.2027.568.4

1.8427.13

_1.842.83

68.20

15

16

0.070.09

40.9627.48

1.3713.130.19

16.070.20_

99.5531.068.6

1.4530.17

0.641.46_

67.73

26

0.020.06

41.0326.89

1.6912.840.14

16.250.31_

99.2130.569.3

1.7928.93

1.161.80_

68.11

7

_

0.0243.1424.83

1.4113.180.24

16.330.22_

99.3727.868.8

1.4827.45

_1.482.47

68.60

17

0.080.09

41.1726.93

1.6512.910.15

16.140.320.02

99.4430.569.0

1.7429.24

0.981.76_

68.02

27

0.070.09

41.2026.91

1.6313.220.11

16.080.20_

99.5130.568.4

1.7229.86

0.301.74_

68.10

8

0.030.08

42.4024.15

2.3712.740.16

16.240.260.04

98.4727.669.4

2.5227.12

_2.530.98

69.37

18

0.060.05

41.1327.13

1.7012.720.20

16.280.310.07

99.6630.769.5

1.8028.79

1.581.81_

67.82

28

0.060.06

41.0927.03

1.5112.960.16

16.160.24_

99.2630.669.0

1.5929.52

0.791.60_

68.09

9

0.050.03

42.7324.96

1.5513.230.14

16.180.24_

99.1028.268.6

1.6427.80

_1.642.06

68.50

19

0.050.05

41.1126.95

1.5213.150.14

16.150.13_

99.2530.568.6

1.6229.89

0.301.62_

68.19

29

0.080.07

41.6326.81

1.3713.230.14

16.090.210.04

99.6730.268.4

1.4430.00_1.450.19

68.36

10

0.050.06

42.9725.44

0.8513.560.17

15.970.240.02

99.3228.467.7

0.8928.35

_0.903.13

67.62

20

0.030.11

41.5426.75

1.5912.840.12

16.340.30_

99.6130.269.4

1.6828.92

0.921.68_

68.48

30

0.010.08

40.9927.29

1.1113.360.16

15.930.22-

99.1430.968.0

1.1830.64

-1.180.27

67.91

Note: Abbreviations are defined in the text and in previous tables.

cesses by seawater at low temperatures (lower than 100°C) producesδ 1 8 θ > 5.7%o, partly a strong O18 enrichment that depends on temper-ature and degree of alteration (whole-rock δ 1 8 θ up to more than 1C‰,smectite up to more than 20%o, Muehlenbachs and Clayton, 1972a).Basalts show a correlation between δ 1 8 θ and water content. (3) Alter-ation processes by seawater at higher temperatures decrease the 18Ocontent. The latter is the case for seawater-formed serpentine miner-als (Wenner and Taylor, 1971, 1973; Margaritz and Taylor, 1974;Ikin and Harmon, 1983; Tsen-Fu Yuei et al., 1990).

In this study, δ 1 8 θ was determined from samples of a harzburgiticserpentinite profile from Hole 920D. To separate relict minerals fromeach profile segment, samples with the lowest degree of alterationwere selected. Insofar as possible, olivine, orthopyroxene, and cli-

nopyroxene were separated, in addition to serpentine and spinel/mag-netite.

All analyses were made in a nickel-glass line using C1F3 to liber-ate oxygen, which was converted to CO2, and measured in a delta Emass spectrometer. The δ 1 8 θ data are related to SMOW (standardmean ocean water), as National Bureau of Standards quartz no. 28was used. All analyses were made in duplicate. The reproducibilitywas better than 0.2%o.

The δ 1 8 θ values are shown in Table 10. We note the followingfeatures:

1. The δ18θ-values of the whole-rock samples decrease from thebottom of the profile (4.6%o) to the top (3.0%o).

465

DATA REPORT

Table 10. Oxygen-isotope data (δ 1 8θ) from harzburgites, Hole 920D

Core, section,piece

3R-1.65R-1.414R-4, 616R-1,5D19R-1.520R-5, 421R-2, 5B22R-4, IB

Depth(mbsf)

18.237.1

119.4134.4162.7177.6183.9192.1

Wholerock

3.02.94.33.83.34.03.84.6

Olivine

5.96.1

6.15.65.5

Orthopyroxene

5.2

6.16.16.06.15.85.9

Clinopyroxene

_

-5.96.1_5.95.95.8

Serpentine

_

2.84.74.23.23.73.85.3

Magnetite

2.93.13.83.0-3.13.13.6

Note: - = mineral fraction is not present.

2. The deepest sample 16 (Sample 920D-22R-4, Piece IB) has aδ 1 8 θ distinctly smaller than the mean δ 1 8 θ value of freshMORB (5.7%c). We conclude that all these whole-rock sam-ples were altered by circulating ocean waters, and the lowered618O values indicate that these ocean waters were heated.

3. This tendency toward lower 18O contents is not uniform: thethird sample from the top (sample 8: Sample 920D014R-4,Piece 6), has the second highest δ 1 8 θ (4.3‰), much higherthan values from samples taken nearby. The third sample fromthe bottom (sample 13: Sample 920D-20R-5, Piece 4) has thenext highest value, 4.C‰ (Fig. 1). Low 18O contents (relativeto those of their neighbors) are obtained from samples 14(Sample 920D-21R-2, Piece 5B), 12 (Sample 920D-19R-1,Piece 5), and the top samples 6 (Sample 920D-5R-1, Piece 4),and 5 (Sample 920D-3R-1, Piece 6), which must indicate lo-calities of rechanneled flow of ocean water (i.e., in faultzones). The mineral separates can be divided into a group ofrelict minerals including olivine, orthopyroxene, and clinopy-roxene, and most of the spinel/magnetite, and another group ofconsisting of alteration products, mainly serpentine and tracesof extremely fine-grained magnetite formed during the serpen-tinization.

4. The primary formed, relict minerals (olivine, orthopyroxene,clinopyroxene) almost all have similar δ 1 8 θ, which is similarto MORB (5.7%o); they seem to be slightly enriched in 18O inthe middle part of the profile. At the top of the profile, the or-thopyroxene (sample 5: Sample 920D-3R-1, Piece 6; δ 1 8 θ =5.2%o) is distinctly lowered.

5. Of the secondary minerals, serpentine has a δ 1 8 θ range from5.3%o to 2.8%c, which is in the lower part of the range for oce-anic serpentines reported by Tsen-Fu Yui et al. (1990; 3.0%c to12.4%o), and in the middle of the range reported by Wennerand Taylor (1973; +0.8%o to 6.7%O).

Of all minerals investigated, serpentine δ 1 8 θ is most similar towhole-rock δ 1 8 θ. The whole-rock δ 1 8 θ and the serpentine δ 1 8 θ arecorrelated (Fig. 1). The whole-rock peak of sample 8 (Table 1) is alsoconspicuous in the serpentine profile.

The spinel/magnetite δ 1 8 θ changes by only 0.9%e, and also exhib-its a striking peak at the level of sample 8. Furthermore, two groupsof samples have a constant δ 1 8 θ (within the measurement error of0.2%o), one with 3.6%o and 3.8%o, and the second from 2.9%o to3.1%0.

The difference (serpentine - spinel/magnetite) is very small anddecreases from +1.7%o to -0.3 %o (within the measurement error).Using the temperature calibration of Wenner and Taylor (1973) for18O in coexisting serpentine and magnetite, we get (also with the larg-est difference, \.l%ò) unreasonably high temperatures, effectively <×>.This means that our spinel/magnetite samples contain nearly no sec-ondary magnetite, formed during serpentinization (in isotope-exchange equilibrium with serpentine).

The isotope pattern in Figure 2 shows an obvious trend betweenwhole rock and serpentine, but a great difference between olivine, or-

7 -

6 -

α

(fi 4 -

3 -

2 ~

1 -

0

<tf

—O— wh-r sample—D— ol sample— -- opx sample- v - - cpx sample•••O••• s p t s a m p l e

—O- • mgt sample

3 4 5

δ 1 8 θ

Figure 1. δ 1 8 θ vs. harzburgite profile. Sample 1 = Sample 153-920D-22R-4(Piece IB), sample 2 = Sample 153-920D-21R-2 (Piece 5B), sample 3 =Sample 153-920D-20R-5 (Piece 4), sample 4 = Sample 153-920D-19R-1(Piece 5), sample 5 = Sample 153-920D-16R-1 (Piece 5D), sample 6 = Sam-ple 153-920D-14R-4 (Piece 6), sample 7 = Sample 153-920D-5R-1 (Piece4), and sample 8 = Sample 153-920D-3R-1 (Piece 6). Abbreviations aredefined as follows: wh-r = whole-rock, ol = olivine, opx = orthopyroxene,cpx = clinopyroxene, spt = serpentinite, mgt = magnetite.

thopyroxene, clinopyroxene, and spinel/magnetite, and the twogroups of magnetite: (1) samples 8 and 16, mean δ 1 8 θ = 3.7%e, whichare the least altered whole-rock samples; (2) samples 14, 10, 6, and5, mean δ 1 8 θ = 3.0%e. Data from each group exhibit a surprisinglysmall degree of scatter.

If these alterations are due to the influence of circulating, hot, andpossibly acidic ocean waters, then a relation between δ 1 8 θ and sometrace-element concentrations (especially of those trace elements thatare more concentrated in the original ocean water) is possible. Figure3 and Table 11 show a clearly negative correlation between whole-rock δ 1 8 θ and sulfur. In this regard, the results of sulfur-isotopeanalyses, which are in progress, may prove very exciting, because theocean sulfate δ34S of 20.3%e is very different from the mantle value(near (Mo).

The relations with other elements are not as convincing (Table11): boron seems to be enriched with increasing alteration, similar tosulfur, and is characterized by greater scatter. Chlorine seems to beimpoverished with increasing alteration, and strontium exhibits nosignificant correlation with alteration.

The water content (expressed as LOI) shows no correlation withthe whole-rock δ 1 8 θ. Instead, we have two groups, containing exactly

466

DATA REPORT

6 -

5 -

P S

3 —

oDΔ

V

O

wh-rwh-rwh-rwh-rwh-r

olopxcpxsptmgt

Δ Δπo

O

ov

v

v OV

>O

o

.18.δ lö0(wh-r)

Figure 2. Whole-rock δ 1 8 θ values plotted vs. δ 1 8 θ values from different min-erals. Alteration processes decreased whole-rock δ 1 8 θ values.The relict min-erals olivine, clinopyroxene, and orthopyroxene (in contrast to serpentiniteand magnetite) underwent almost no oxygen alteration in the different levelsof the profile. Abbreviations are defined in Figure 1.

wh-r 0-18 sample nrB ppm sample nrS ppm/10 sample nr(LOI-10)10 sample nrCl ppm/100 sample nr

100

Figure 3. Whole-rock δ 1 8 θ , B, S/10, LOI - 10, and CU10 plotted vs. the sam-ple profile of harzburgites (see Fig. 1). The alteration seems to increase (in adifferent manner) with the contents of B, S, and Cl, and with LOI. Abbrevia-tions are defined in Figure 1.

Table 11. Whole-rock δ1 8θ and the content of B, S/10, Cl/100, Sr, and (LOI - 10) 10 from harzburgites, Hole 920D.

Core

3R5R

14R16R19R20R21R22R

Whole-rockδ 1 8 o

3.02.94.33.83.34.03.84.6

B (ppm)

2421

95

218

249

S/10(ppm)

89.087.042.554.0

105.088.577.038.0

(LOI-10) 10

25305

222526277

Cl/100(ppm)

8.558.30

22.1023.1010.0024.5514.8026.10

Sr

2.470.931.451.214.652.282.592.47

the same samples as the two groups of magnetite δ 1 8 θ: high magne-tite δ 1 8 θ (3.7‰) is associated with low LOI, and vice versa. The dif-ference in the LOI values is small.

Isotope Temperatures

Assuming isotope-exchange equilibrium, we have calculated iso-tope temperatures for the mineral pairs clinopyroxene (diopside) andmagnetite (spinel). We used the relation of Bottinga and Javoy (1975)for anhydrous minerals for temperatures (T) higher than 500°C andthe coefficient for these minerals given by Matthews et al. (1983):

Δ18O (clinopyroxene - magnetite) = 4.03 • 106/T2

We get reasonable temperatures (Table 12), which cluster around1100° and 900°C. However, one must consider that we do not have

"pure" magnetite, but Cr spinel with a different composition, but thesame crystal structure. Therefore, the calculation is affected by a par-tial uncertainty.

It is also difficult to obtain temperatures from the clinopyroxene-olivine pair. The olivine δ 1 8 θ values (Table 10) form two groups, andthe group around 6.0%o does not differ from the δ 1 8 θ of orthopyrox-ene and clinopyroxene. It is known that the δ 1 8 θ values of clinopy-roxene and orthopyroxene at such high temperatures do not differ(Bottinga and Javoy, 1975). The lack of a difference between pyrox-ene and olivine raises the question whether, at this level, these min-erals were formed in isotope-exchange equilibrium. Examples of co-existing minerals of oceanic samples with clear disequilibrium havebeen reported (Muehlenbachs and Clayton, 1971).

The two deepest samples, samples 16 (Sample 920D-22R-4,Piece IB), and 14 (Sample 920D-21R-2, Piece 5B), have lower δ 1 8 θin olivine, with a mean of 5.55%o; using the mean of orthopyroxeneand clinopyroxene of 5.85%c, we get Δ18O = 0.30%e. (If we tentatively

467

DATA REPORT

Table 12. Calculated temperatures (T) for clinopyroxene-magnetite, forthe formation of the serpentines using δ 1 8 θ = 0 and final δ 1 8 θ of waterequilibrated by formation of antigorite, Hole 920D.

Core,section, piece

3R-1,65R-1,414R-4, 616R-1,5D19R-1.520R-5, 421R-2, 5B22R-4, IB0 group 1

0 group 2

Depth(mbsf)

18.237.1

119.4134.4162.7177.6183.9192.1

T(°C)(cpx-mgt)

1112867

927927

10801096

906

T(°C)(serpentine),

δ 1 8 O w = 0

>183>134>146>171>158>155>121

δ 1 8 θ w

T = 235°C

_

1.453.352.851.852.352.453.95

δ 1 8 o wT = 250°C

_

1.803.703.202.202.702.804.30

Note: W = water, cpx = clinopyroxene, mgt = magnetite. 0 group 1 = mean of Sample14R-4, (Piece 6) and 22R-4 (Piece IB); 0 group 2 = mean of Samples 16R-1 (Piece5D), 20R-5 (Piece 4), and 21R-2 (Piece 5B).

calculate a temperature with the relation and coefficient B = 1.240 ofBottinga and Javoy [1975], assuming an error of 0.2%c, we get a tem-perature of 1860°, with a variation between -450° and +1200°).

To calculate the isotope temperature of serpentinization, we needa second mineral, with a δ 1 8 θ value lower than that of serpentine, thatformed in isotopic equilibrium with serpentine. Unfortunately, it wasnot possible to separate the very fine grained magnetite, which occursonly as traces and is intensively intergrown with serpentine. We ten-tatively calculated temperature, using the relation for serpentine-wa-ter from Wenner and Taylor (1971, 1973), and assuming δ 1 8 θ =0.0%o for water as an extreme case (Table 12). We obtained temper-atures between 121° and 183°C, increasing from bottom to top! Wen-ner and Taylor (1971; 1973) estimated 125°C for lizardite and 180°Cfor chrysotile in Mid-Atlantic Ridge environments. Our samples,however, also contain antigorite (and some additional minerals,formed up to the transitional zone between greenschist and amphib-olite facies). Wenner and Taylor (1971; 1973) estimated 225° and235°C for antigorite. Pressure-temperature diagrams for theharzburgite system in equilibrium give a range from 250° to 550°Cfor antigorite at a pressure of 3 kbar (Bucher and Frei, 1994, p. 153).We calculated the resulting δ 1 8 θ for water that has attained isotopicequilibrium with the forming antigorite, assuming temperatures of235° and 250°C (Table 12), and obtained ranges from 1.45%e to3.95%O, and 1.80%o to 4.30 ‰. The seawater with δ 1 8 θ = 0%c had tak-en up I8O by the formation of serpentine from olivine and other min-erals.

Water/Rock Ratios

For the calculation of water/rock or better water/mineral (serpen-tine) ratios, we used the expressions of Taylor (1978). We assumedtemperatures of 235°, 250°, and 300°C, and calculated ratios forclosed and open systems in the sense of Taylor (Table 13). The finalΔ18O (serpentine - water) is calculated for these temperatures usingthe equation of Wenner and Taylor (1973). For the deepest sample,only a small amount of water was available and active. The higher thesample position within the profile, the more water could penetrate tothat point and cause isotope exchange (and, of course, mineral reor-ganization).

CONCLUSIONS AND SUMMARY

The ultramafic rock samples investigated are serpentinizedharzburgites that contain 15%—30% relict minerals. Whole-rockgeochemistry shows the normal pattern of less-depleted, refractorymantle peridotites. Si, Mg, and Fe contents are very uniform, as are

the compatible trace elements, but contents of K, Na, Ti, and espe-cially Ca, vary considerably. Partial anatexis of the primary mantlematerial and hydrothermal metamorphism of the depleted rocks leadto these irregularities, especially for Ca.

The relict minerals olivine, orthopyroxene, and clinopyroxeneshow different morphological and genetic varieties, but their chemi-cal compositions are relatively uniform within and between the indi-vidual samples, especially with respect to their Mg/Fe ratios (Mg#)(Table 14).

Synkinematic or prekinematic olivine (with Fo = 89.8%-90.6%)and orthopyroxene show nearly identical Mg#, but different NiOcontents: in olivine, NiO = 0.39 wt% (0.29-0.49 wt%), and in ortho-pyroxene, NiO = 0.11 wt% (0.05-0.19 wt%). All these features arecharacteristic of less-refractory mantle minerals.

The contents of A12O3 and Cr2O3 in orthopyroxene are stronglycorrelated (r = +0.88). CaO locally shows strikingly high concentra-tions (up to 4.5 wt%), leading to a subgroup with Wo contents be-tween 5 and 8.8 wt%. Exsolution of clinopyroxene leads to a selec-tive impoverishment of Ti, Mg, Cr, and Al in orthopyroxene.

The Porphyroclastic clinopyroxene can be classified as tscherma-kitic diopside because of its high A12O3 content (3.75-5.75 wt%).A12O3 is well correlated with Cr2O3 in two groups (r = +0.33; r2 =+0.77), but not as strongly as in orthopyroxene. The general coinci-dence of these features with the equivalent data for orthopyroxeneseems to reflect comparable conditions during the crystallization ofboth the pyroxenes. The behavior of Mg and Ca is clearly the oppo-site, and the scatter of Ca*#s is remarkable. Microprobe profilesthrough clinopyroxene grains show MgO-rich and CaO-poor rims, aswell as MgO-poor and CaO-rich cores. The marginal decrease inCaO may be caused by reaction with fluids at higher temperatures,and probably also by partial melting.

In addition to the main group of clinopyroxenes (Wo = 49, En =46.7, Fs = 4.3), a subgroup exists that is clearly poorer in Ca and rich-er in Mg (Wo = 2.6, En = 52, Fs = 5.4). Both groups differ stronglyin the correlation between Ca*# and A12O3: the Wo-rich group has r= +0.48, and the Wo-poor group has r = -0.20. This opposite behav-ior should be the result of partial alteration with a loss of Ca.

The chromian-bearing spinels are typical picotites, but with cer-tain differences. Hercynite-bearing samples are slightly poorer inCr2O3, grains with somewhat higher Cr2O3 contents show some Mgchromite by calculation. All Cr spinels investigated must be consid-ered as primary constituents of the peridotite, but they are altered tosome degree.

Different temperatures were calculated using several geothermo-metric methods:

1245°-1340°C, mean value = 1275°C ("Cr-Al-orthopyroxene"geothermometer; Witt-Eickschen and Seck, 1991),

1250°-1300°C (two-pyroxene geothermometer; Lindsley, 1983),1170°-1380°C, mean value 1300°C (two-pyroxene geother-

mometer; Kretz, 1963, 1982),1425° ± 25°C (Wood and Banno, 1973).

Aside from the Wood and Banno (1973) data, the temperature~1300°C may reflect the primary formation of prekinematic to synk-inematic orthopyroxene and clinopyroxene in the protolith at pres-sures not greater than 10 kbar (Anastasiou and Seifert, 1972; Wyllie,1981).

In the orthopyroxene-chromian spinel system, applying the meth-od of Sachtleben and Seck (1981) on the basis of XMg, XA1, and XCr,the two Cr-spinel-bearing samples yield temperatures of 1180° and1075°C, respectively. With a certain restriction, the temperature esti-mate for the clinopyroxene-spinel/magnetite pair, using I8O ex-change, gave 1100° and 900°C. All these values may reflect subsoli-dus deformation and recrystallization processes, possibly in morethan a single step.

DATA REPORT

Table 13. Calculated water/rock ratios for serpentine for closed and open systems for different temperatures, Hole 920D.

Core,section, piece

3R-1,65R-1.414R-4, 616R-1,5D19R-1.520R-5, 421R-2, 5B22R-4, IB

Depth(mbsf)

18.237.1

119.4134.4162.7177.6183.9192.1

(w/r)cs

T = 235°C

_

(3.67)0.611.14

(2.69)1.751.260.09

(w/r)os

T = 235°C

_

(1.96)0.530.88

(1.62)1.210.940.08

(w/r)cs

T = 250°C

_

(2.95)0.561.02

(2.26)1.521.100.08

(w/r)os

T = 250°C

_

(1.72)0.480.80

(1.44)1.090.850.08

(w/r)cs

T = 300°C

_

(1.93)0.440.79

(1.58)1.130.820.07

(w/r)os

T = 300°C

_

(1.29)0.390.651.120.870.670.06

Notes: Numbers in parentheses represent values from serpentine with low δ 1 8 θ , - = not calculable, w/r = water/rock ratio, cs = closed, os = open.

Table 14. Variation of mineral chemistry in the profile of Hole 920D.

Sample no.

101112131415

Core,section, piece

16R-1,5D16R-6, 619R-1.520R-5, 421R-2, 5B22R-2, ID

Depth(mbsf)

134.4140.3162.7177.6183.9192.1

n

209

11126

28

Olivine

Mg#

90.3390.2790.1990.2990.3590.21

Orthopyroxene

n

514_182016

Mg#

90.5390.28

_90.4290.5390.26

n

20107

106

29

Clinopyroxene

Mg#

91.2992.0891.3492.1892.0691.60

Ca#

48.2652.4750.7551.8750.1350.82

Note: Symbols and abbreviations are defined in previous tables.

The δ 1 8 θ values are very uniform in the relict minerals olivine, or-thopyroxene, and clinopyroxene, and values are similar to those ofMORB, although slightly higher than for the bulk mantle. Whole-rock samples, serpentine, and magnetite show clearly lower δ 1 8 θ val-ues, resulting from hydrothermal alteration at relatively high temper-atures. The temperature determination of the serpentinization is re-stricted because only serpentine is available. Using some calculationsand estimations about the water/rock ratio, a temperature range of±250°C could be probable.

ACKNOWLEDGMENTS

We are grateful to the Deutsche Forschungsgemeinschaft, Bonn,for considerable financial support of this work. H.-J. Bernhardt (Bo-chum) has greatly supported the microprobe investigations, B. Knip-ping (Clausthal), and G. Bombach, E. Rüdiger, and K. Volkmann(Freiberg) were responsible for the analytical procedure. C.-D.W. isgrateful to all these colleagues and their institutions. The nickel-glassline was constructed, and the samples were separated and prepared byW. Tikhomirov; the mass spectrometric analyses were performed byR. Liebscher. For critical reviews and constructive comments, wethank E. Bonatti, J. Karson, and especially P. Kelemen and B. Wa-then.

REFERENCES

Anastasiou, R, and Seifert, R, 1972. Solid solubility of A12O3 in enstatite athigh temperatures and 1-5 kb pressure. Contrib. Mineral. Petrol,34:272-287.

Aumento, E, and Loubat, H., 1971. The Mid-Atlantic Ridge near 45°N. XVI.Serpentinized ultramafic intrusions. Can. J. Earth Sci., 8:631-663.

Bottinga, Y, and Javoy, M, 1975. Oxygen isotope partitioning among theminerals in igneous and metamorphic rocks. Rev. Geophys. Space Phys.,20:250-265.

Bucher, K., and Frei, M., 1994. Petrogenesis of Metamorphic Rocks: Berlin(Springer-Verlag).

Cannat, M, Bideau, D., and Bougault, H., 1992. Serpentinized peridotitesand gabbros in the Mid-Atlantic Ridge axial valley at 15°37'N and16°52'N. Earth Planet. Sci. Lett, 109:87-106.

Cannat, M., Karson, J.A., Miller, D.J., et al., 1995. Proc. ODP, Init. Repts.,153: College Station, TX (Ocean Drilling Program).

Cannat, M., and ODP Shipboard Scientific Party, 1995. Probing the founda-tion of the Mid-Atlantic Ridge. Eos, 76:129-133.

Christiansen, F.G., 1985. Deformation fabric and microstructures in ophi-olitic chromitites and host ultramafics, Sultanate of Oman. Geol. Rund-sck, 74:61-76.

Dick, H.J.B., Fisher, R.L., and Bryan, W.B., 1984. Mineralogic variability ofthe uppermost mantle along mid-ocean ridges. Earth Planet. Sci. Lett.,69:88-106.

Hamlyn, P.R., and Bonatti, E., 1980. Petrology of mantle-derived ultramaficsfrom the Owen fracture zone, Northwest Indian Ocean: implications forthe nature of the oceanic upper mantle. Earth Planet. Sci. Lett., 48:65-79.

Hofmann, A.W., 1988. Chemical differentiation of the Earth: the relationshipbetween mantle, continental crust, and oceanic crust. Earth Planet. Sci.Lett., 90:297-314.

Ikin, N.P., and Harmon, R.S., 1983. A stable isotope study of serpentiniza-tion and metamorphism in the Highland Border suite, Scotland, U.K.Geochim. Cosmochim. Acta, 47:153—167.

Ito, E., White, W.M., and Göpel, C, 1987. The O, Sr, Nd and Pb isotopegeochemistry of MORB. Chern. Geol., 62:157-176.

Karson, J.A., and Winters, A.T., 1992. Along-axis variations in tectonicextension and accommodation zones in the MARK area, Mid-AtlanticRidge 23°N latitude: ophiolites and their modern oceanic analogues. InParsons, L.M., Murton, B.J., and Browning, P. (Eds.), Ophiolites andTheir Modern Oceanic Analogues. Geol. Soc. Spec. Publ. London,60:107-116.

Kretz, R., 1963. Distribution of magnesium and iron between orthopyroxeneand calcic pyroxene in natural mineral assemblages. J. Geol., 71:773-785.

, 1982. Transfer and exchange equilibria in a portion of the pyrox-ene quadrilateral as deduced from natural and experimental data.Geochim. Cosmochim. Acta, 46:411-421.

Kyser, T.K., O'Neil, J.R., and Carmichael, I.S.E., 1982. Genetic relationsamong basic lavas and ultramafic nodules: evidence from osygen isotopecompositions. Contrib. Mineral. Petrol., 81:88-102.

Lindsley, D.H., 1983. Pyroxene thermometry. Am. Mineral., 68:477-493.Lindsley, D.H., and Dixon, S.A., 1975. Diopside-enstatite equilibria at 850°

to l,400°C, 5 to 35 kbar. Am. J. Sci., 276:1282-1301.Margaritz, M., and Taylor, H.P., Jr., 1974. Oxygen and hydrogen isotope

studies of serpentinization in the Troodos ophiolite complex, Cyprus.Earth Planet. Sci. Lett, 23:8-14.

469

DATA REPORT

Mattey, D., Lowry, D., and Macpherson, C , 1994. Oxygen isotope composi-tion of mantle peridotite. Earth Planet. Sci. Lett, 128:231-241.

Matthews, A., Goldsmith, J.R., and Clayton, R.N., 1983. Oxygen isotopefractionations involving pyroxenes: the calibration of mineral-pair geo-thermometers. Geochim. Cosmochim. Acta, 47:631-644.

Michael, PJ., and Bonatti, E., 1985. Peridotite composition from the NorthAtlantic: regional and tectonic variations and implications for partialmelting. Earth Planet. Sci. Lett, 73:91-104.

Miyashiro, A., Shido, F., and Ewing, M., 1969. Composition and origin ofserpentinites from the Mid-Atlantic Ridge, 24° and 30°N latitude. Con-trib. Mineral. Petrol, 23:117-127.

Morimoto, N., et al., 1989. Nomenclature of pyroxenes. Can. Mineral,27:143-156.

Muehlenbachs, K., 1976. Oxygen isotope geochemistry of DSDP Leg 34Basalts. In Yeats, R.S., Hart, S.R., et al., Init. Repts. DSDP, 34: Washing-ton (U.S. Govt. Printing Office), 337-339.

, 1977. Oxygen isotope geochemistry of rocks from DSDP Leg 37.Can. J. Earth Sci., 14:771-776.

1978. Oxygen isotope geochemistry of rocks from Leg 46. InDmitriev, L., Heirtzler, J., et al., Init. Repts. DSDP, 46: Washington (U.S.Govt. Printing Office), 257-258.

1979. The alteration and aging of the basaltic layer of the seafloor: oxygen isotope evidence from DSDP/IPOD Leg 51, 52, and 53. InDonnelly, T, Francheteau, J, Bryan, W., Robinson, P., Flower, M., Salis-bury, M., et al., Init. Repts. DSDP, 51, 52, 53: Washington (U.S. Govt.Printing Office), 1159-1167.

Muehlenbachs, K., and Clayton, R.N., 1971. Oxygen isotope ratios of sub-marine diorites and their constitutent minerals. Can. J. Earth Sci.,8:1591-1595.

, 1972a. Oxygen isotope studies of fresh and weathered submarinebasalts. Can. J. Earth Sci., 9:172-184.

-, 1972b. Oxygen isotope geochemistry of submarine greenstones.Can. J. Earth Sci., 9:471-478.

1976. Oxygen isotope composition of the oceanic crust and itsbearing on seawater. J. Geophys. Res., 81:4365-4369.

Nicolas, A., Boudier, F, and Bouchez, J.L., 1980. Interpretation of peridotitestructures from ophiolitic and oceanic environments. Am. J. Sci.,280:192-210.

Obata, M., 1976. The solubility of A12O3 in orthopyroxenes in spinel and pla-gioclase peridotites and spinel pyroxenite. Am. Mineral, 61:804-816.

Peters, T, 1968. Distribution of Mg, Fe, Ca and Al in coexisting olivine,orthopyroxene and clinopyroxene in the total serpentinite (Davos, Swit-zerland) and in the Alpine metamorphosed Malenco serpentinite (N.Italy). Contrib. Mineral Petrol, 18:65-75.

Purdy, G.M., and Detrick, R.S., 1986. Crustal structure of the Mid-AtlanticRidge at 23CN from seismic refraction studies. J. Geophys. Res.,91:3739-3762.

Sachtleben, T., and Seck, H.A., 1981. Chemical control of Al-solubility inorthopyroxene and its implication on pyroxene geothermometry. Contrib.Mineral. Petrol, 78:157-165.

Taylor, H.P., Jr., 1978. Oxygen and hydrogen isotope studies of plutonic gra-nitic rocks. Earth Planet. Sci. Lett., 38:177-210.

Wenner, D.B., and Taylor, H.P., 1973. Oxygen and hydrogen isotope studiesof the serpentinization of ultramafic rocks in oceanic environments andcontinental ophiolitic complexes. Am. J. Sci., 273:202-239.

Wenner, D.B., and Taylor, H.P., Jr., 1971. Temperatures of serpentinization ofultramafic rocks based on 16O/18O fractionation between coexisting ser-pentine and magnetite. Contrib. Mineral. Petrol, 32:165-185.

Witt-Eickschen, G., and Seck, H.A., 1991. Solubility of Ca and Al in ortho-pyroxene from spinel peridotite: an improved version of an empiricalthermometer. Contrib. Mineral Petrol, 106:431-439.

Wood, B.J., and Banno, S., 1973. Garnet-orthopyroxene and orthopyroxene-clinopyroxene relationships in simple and complex systems. Contrib.Mineral Petrol, 42:109-124.

Wyllie, PJ., 1981. Plate tectonics and magma genesis. Geol. Rundsch.,70:128-153.

Yui, T.-F, Yeh, W.-H., and Lee, C.W., 1990. A stable isotope study of serpen-tinization in the Fengtien ophiolite, Taiwan. Geochim. Cosmochim. Acta,54:1417-1426.

Date of initial receipt: 24 August 1995Date of acceptance: 27 February 1996Ms 153SR-018

470