A normative calculation technique for determining opal in deep-sea sediments

Geochlmica

et Cosmochunica

Acta. 1977.

Vol. 41. pp 671 10 676 Prrgamon

Prcs

Pnnted

$11Great Bntaln

NOTES

A normative calculation technique for determining opal in deep-sea sediments MARGARET LEINEN* Oregon (Recked

State University.

29 April

Corvallis.

1976; uccepted

Oregon

in reoisrd

97330. U.S.A.

form 22 No~~mher

1976)

Abstract-The opal content of deep-sea sediment can be estimated by subtracting non-biogenic silica, calculated from the aluminum and magnesium concentrations in the sediment, from the total silica content of the samples. Unlike most previously described methods. this calculation is capable of yielding reliable estimates of opal in pre-Pleistocene sediments because it is unaffected by structural changes that take place in opal as it ages

INTRODUCHON THE OPALINE skeletal remains of marine plankton such as diatoms, radiolarians and silicoflagellates are important constituents of marine sediments. A quantitative estimate of the amount of this biogenic silica is an important parameter for geologic, geochemical and paleoceanographic studies. The determination of opal in deep-sea sediment. however, has been a frustrating problem to geologists. The methods commonly used are X-ray diffraction (CALV~RT. 1966; ELLIS, 1972), chemical dissolution (BEZRUKOV, 1955; ELLIS, 1972; HURD, 1972) and i.r. spectroscopy (CHESTER and ELDERFIELD. 1968). These techniques have generally been developed for use in the study of a particular deep-sea region; within such individual regions the methods give consistent results. When used to determine opal in sediments from other areas, however. the methods do not give predictable results. For example, ELLIS (1972) was able to obtain internally consistent opal contents for South Atlantic sediments using the X-ray technique, but using the same technique in the equatorial Pacific HEATH et al. (1977) found that opal determinations had to be corrected for overestimation of up to SOTi. The reason for this inconsistency is probably that the opal in deep-sea sediments is very poorly ordered silica with a variable degree of ordering and of hydration. This has made the application of existing analytical techniques for opal to all sediments unreliable because they depend on physical and structural characteristics of the opal. Experiments with the dissolution techniques illustrate the problems of relying on measurement of physical properties of opal (in this case, solubility). In a test of the sodium hydroxide leach technique ELLIS (1972) found that an appreciable quantity of opal was undissolved after four leaches and that up to seven dissolutions were required to remove all the opal. HURD (1972) had similar problems obtaining consistent results with the sodium carbonate dissolution and recommended that a complete standard curve be run with each set of samples to monitor variability in dissolution. There are indications that this would be insufficient: ELLIS (1972) found that even paired leachings of duplicate samples removed markedly different proportions of opal in sediment samples. Not unexpectedly, the opal content of sediments determined by two techniques in the same area is often different.

LISITZIF; (1971) was able to map patterns of surface productivity reflected in the opal content of South Atlantic sediments determined by sodium carbonate leaching, but the opal contents were different from those measured by the X-ray technique (ELLIS and MOORE, 1973). This severely restricts the value of these measurements of biogenic silica for geochemical balances or sediment budgets. CHESTER and ELD~RFIELD (1968) used i.r. spectroscopy to determine biogenic opal. This technique measures Si stretching and the author’s results agreed well with parallel determinations made by the X-ray technique. The presence of more than 5’?” quartz in sediment interferes with the absorption band, however, and requires a time-consuming modification to the technique. Further, if the opal:quartz ratio is less than 3: 1, a common situation in pelagic sediments. it is not possible to determine the opal content by this method. A much more serious problem for opal determination is caused by the behavior of opal as it ages. Over periods of tens of millions of years. opal in sediments can invert to opal-CT (JONES and S~GNI~, 1971), it can dissolve and be redeposited on other sediment grains, it can diffuse out of the locus of deposition or it can persist relatively unchanged. This highly variable aging behavior makes any opal determination method based on physical or structural characteristics suspect for sediments older than about one million years. A more promising approach to the problem of determining opal in deep-sea sediments. the one used in this study. is based on the chemistry of the biogenic opal. The silica in biogenic opal is very pure, with an average of less than 0.1 wt “,” of other elements (MARTIS and KNAIIER, 1973). If this silica can be distinguished from silica in the non-biogenie fraction of the sediment. the opal content can be determined. The technique developed here is to assume that the amount of opal in sediments of any age can be estimated by a normative calculation in which some of the analytical silica concentration is subtracted as non-biogenie in proportion to the amount of alumina or other chemical constituents in the sample. The first use of such a calculation for deep-sea sediments was made by B~s~R~M et al. (1972) who argued that since the SiO,:Al,O, ratio of the continental crust is about 3: 1. opaline silica in sediments could be estimated from the relation: SiOZ ‘opal’ = SiOz measured

*Present Universtty

address: Graduate School of Rhode Island, Kmgston,

of Oceanography. RI 028X1. U.S.A.

where 671

(3 Al>O,)

is the

estimate P +

- 3 AlzOj measured. of non-biogenic

silica.

612

Notes

[BOSTROMand FISHER(1971) chose a ratio of 4: 1 to represent the oceanic and continental crust.] However, because the non-biogenic fraction of marine sediments is made up of clay mineral assemblages with various admixtures of volcanic ash, authigenic minerals, manganese nodules and hydrated ferromanganese oxides, it cannot be assumed to have constant elemental ratios equal to those of the average crust. In clays, which are generally the major non-biogenic sediment component, the Si02:A1,0, ratio ranges from 2: 1 to 7: 1 (WEAVERand POLLARD,1973) and varies with structure and cation substitution. The amount of non-biogenic silica could, therefore, be a function of both the dominant clay mineral group in the sediment and the various cations which substitute in the clay structure. In order to determine whether the method of subtracting non-biogenic silica is feasible and capable of giving reliable opal concentrations, three techniques based on subtracting non-biogenic silica have been developed for sediment from Deep Sea Drilling Project (DSDP) cores from the central equatorial Pacific and tested on data from other areas of the World Ocean.

(1969) indicate that the precision of the quartz determination is high, with a coefficient of variation of 1.9%. Opal determinations were made as a check on the presence of opal not detected in smear slides. Although the X-ray technique is probably not quantitative for these samples, most of which are pre-Pleistocene. it does indicate whether or not opal is present. Finally, a portion of the total sample was dissolved with aqua regia and hydrofluoric acid in Teflon-lined bombs, neutralized with boric acid after dissolution (BERNAS,1968) and analyzed for Si, Al, Mg, Mn and Fe by atomic ahsorption spectrometry. The standard error of the atomic absorption analyses was &3% of measured value for Si and Al (samples with very low concentrations had standard error of up to *5?;) and +2:/i of measured value for Mg, Mn and Fe. Comparison with USGS rock standards indicated no systematic error.

RESULTS Quartz and opal

METHODS To assess the variability of the SiO,:Al,O, ratio and chemical composition of non-biogenic sediment, 42 samples which had no recognizable skeletal opal in smear slides and which spanned the entire Cenozoic time period were chosen from DSDP cores from the central equatorial Pacific. The clay minerals present in these samples were determined by X-ray diffraction to test the validity of a normative calculation based purely on the clay mineralogy. After calcium carbonate was removed with acetic acid buffered to pH = 5, ferromanganese hydroxyoxides were removed by the dithionateecitrate-bicarbonate technique of MEHRAand JACKSON{1960). The samples were Mg2+ saturated and the <2pm fraction was mounted as oriented aggregates and X-rayed in duplicate. In addition, quartz and opal were determined by X-ray diffraction (TILL and SPEARS,1969; ELLIS, 1972) in carbonate-free aliquots of the same samples. TILL and SPEARS

Quartz accounted for l-3 wt ‘l/i of all samples except those from the last one million years. which had quartz contents of up to 7 wt %. In a study of equatorial Pacific sediment by HEATH(1969) and in a study of North Pacific sediment by HEATH et al. (1976) similar quartz contents were measured. X-ray diffractograms of the < 2 pm size fraction indicated that 21 of the samples chosen for study contained appreciable amounts of opal in that size range even though none was apparent in smear slides of the samples. These samples were not used for further work. Of the remainder. none showed opal in diffractograms of the i 2 pm size fraction and none contained opal as determined by the X-ray diffraction technique of CALVERT (1966) and ELLIS(1972). In addition, the paleolatitudes of all SaEIpkS [taken from VAN ANDEL er af. (197s)j were well outside the zone of high biological productivity associated with the equator. There is, therefore, no reason to suspect that there is undetected opal in these samples.

Table 1. Semi-quantitative clay mineral abundance, quartz and non-biogenic silica, and SiO,:AI,O, in opal-free samples Sample** 40(0-l) 71(x3-29) 72(27-B) 73(20-U) 74(33-34) 74(36-37) 74(&i-45) 75(1-Z) 75(16-18) 75(21-22) 75(31-321 77/37-385 80(21-22) 81(16-17) 159(13-l‘?) 159(19-20) 159(22-23) 159123-241 16Oil-2) 162149-50)

smectite*

69 83 45 44 76 88 :?I 90 53 15 85 1:: :: 91 82

11iite*

::

40 39 19 0 :: 2: 63 11

;

16

:

1: 5

Kaolinite'

Chlorite*

IQuartz

SiOp,:Al203

:

3.8 3.6

6.6 2.8

3 5

3.2

5.7

4 2

3.2 9.5 3.7 5.4 3.9 4.3

z 3

3.5 3.8 3.9

1 5 :

%Carbonate-free Non-biogenic Silica

1

12

2.9 3.5 3.0

3.3 3.1 3.9 3.8

2.5 2.3 10.2 8.2

: 1

4.0 3.6 3.9

7.6 2.1 3.0

:

3.8 3.9

5.7 5.8

:

3.9 3.7

7.3 2.8

ratios

65 6 8 11 10 26 ::

2

:: 32 33

27 6

z: 22 33 21 5

li

4$

1:

::

i

67 66 14 8 66 10

i

59 :i, 25 50 23

* Per cent of total clay minerals present (BEAYE, 1964). ** Sample number indicates DSDP site; age interval of sample (m.y.) in parentheses. SiO,:Al,O,: (1) ratio calculated from clay mineralogy: SiO, :Al,O, = C4.0 (% smectite) + 2.5 (% ill&e) + 2.0 (% kaolinite) + 4.0 (% chlorite)] (l/IOO). (2) ratio determined by atomic absorption spectrometry: SiOz:AlzOs = (total Si02 measured - quartz/total A1203 measured). Carbonate-free non-biogenic silica: (1) estimated from alumina content and Si02:A120, (column 1): non-biogenic silica = (SiO,:Al,O,) (total Al,O, measured). (2) actual content measured by atomic absorption: non-biogenic silica = total SiO, measured - quartz.

673

Notes

The relative proportions of clay mineral groups present in the samples based on peak areas and using the weighting factors of BISCAYE (1965) are given in Table 1. The BI~CAYE (1965) method assumes that these clays account for all the sediment finer than 2 pm. Although the assumption is undoubtedly incorrect (HEATH et al.. 1976) the method does give a reasonable distribution of the clays present among the four major clay groups. Except in rare instances the amounts of feldspars and quartz were insignificant in the samples.

The analytically determined SiOZ:AI,O, ratios of the 21 opal-free samples ranged from 2.1 to 10.2 (Table 1). These have been corrected for quartz content.

DISCUSSION The wide range of Si02:AI,0, ratios in opal-free sediments implies that A120s alone is not a reliable predictor of non-biogenic silica. Using Bostrom’s calculation for non-biogenic silica (assumes SiO,: Al,Os = 3: 1 in nonbiogenic sediment) a value in excess of the total measured silica is predicted for more than half of his southeast Pacific samples for which analyses have been published (BosTROM and PETERSON. 1969). Opal estimates for all of these samples are, therefore. negative. A plot of non-biogenic silica measured in the opal-free samples in this study against non-biogenic silica estimated by a SiOZ :Al,O, ratio of 3:l is shown in Fig. 1 (A). For these samples the ratio underestimates structural silica by up to 30 wt “;. A second approach to the discrimination of non-biogenie silica is to establish an SiO,:AI,O, ratio on the basis of the clay mineralogy of the sediment. An SiO,:AI,O, ratio was estimated for the clay phase of each of the opalfree samples used in this study. This was done by assigning each clay group an average ratio based on compiled analyses of clay mineral compositions (WEAVER and POLLARD, 1973). The ratios used were: smectite4: 1. illite-2.5: 1. kaolinite-2: 1 and chlorite4: 1. Using these values, theoretical Si0,:A1203 ratios were calculated for the total clay fraction in each sample. These ratios ranged from about 4: 1 for samples deposited near the East Pacific Rise to about 3: 1 for samples deposited furthest from the rise crest (Table 1). These estimates are probably conservative since oceanic smectites which have been analyzed have SiO,:AI,O, ratios of 6 to 19 (EKLUND, 1974; AOKI ct al., 1974). The calculated ratios vary significantly from those actually measured (Table 1). Some of the high values for chemically determined ratios may be due to the presence of more ‘oceanic’ smectites, but the range of ratios is large and many of the analytically determined ratios are much lower than those estimated from the clay mineralogy. Thus, this approach is inadequate to characterize the non-biogenie silica content of the sediments. The measured nonbiogenic silica contents and those estimated from a ratio based on clay mineralogy for the opal-free sediments from this study are compared in Fig. I (B). The values differ by up to 30 wt “‘, silica. Stepwise regression analysis was used to determine whether the inclusion of elements other than Al could generate an improved equation to estimate the non-biogenic silica content of deep-sea sediment. Because variations in the Si0,:A1,03 ratio of clay minerals result from substitution of other cations (especially Mg and Fe) for octahedral Al, as well as from Si:Al variations in tetrahedral sites, inclusion of other octahedral cations should lead to better predictions of non-biogenic silica. Using this empirical approach several models were tested by considering different combinations of elements and by changing the amount of variance which an element had to explain to be included

NON-SIOGENIC

SILICA

ESTIMATED

Fig. 1. Comparison of non-biogenic silica measured in opal-free sediments with that predicted by (A) a SiO,:Al,Os ratio of 3:1, (B) a SiO,:Al,OJ ratio based on the clay mineralogy of the sample, and (C) regression equation (1) described in text. Straight line indicates perfect correspondence between non-biogenic silica predicted and measured. in the equation. The results follow what would be intuitively expected from the mineralogy of marine clay minerals. In all cases, Al was the most important element and accounted for 6@75% of the variance in non-biogenic silica content. Mg or Mg’ was usually the second variable, accounting for up to 10% of the variance. This relation is expected since magnesium substitution for trivalent cations in the octahedral sheets of dioctohedral clay Table 2. Regression equations for predicting non-biogenic silica content of deep-sea clays. Units are in wt “;,

1)

4.33 Al + 1.36 Mg2 = non-biogenic

2)

4.04 Al = 2.27 Mg2

3)

1.1 MgMn

3.57 Al - 0.3 Fe2 - non-biogenic

silica

non-biogenic silica

silica

5.8%Si02

?a

5.7% Si02

2.257

7.5% SiO2

1.780

lb

IRI = mean absolute value of regression residuals. t = t statistic for significance of final variable, II = 21.

Notes

614

Table 3. Deep-sea sediments. lithologies, measured or described opal content, and opal content predicted by normative calculation and by ratio techniques m

Location

Sediment Type

Qescribed Opal Content

Predicted O&l

1

2

Content 3

Carnegie Cruise 7 Samples 10

5%9’N

3Z056'W

green hemipelagic mud

rad. in 125Um fraction

9

9

2

13

2o3o's

9SD43'W

siliceous globigerina ooze

abund. rad. in
18

55

52

3015'S

sso48'u

siliceous globigerina ooze

abund. rad. in <125um

51

13

55

1G

14007'5 1lloSo'u

ferruginous calcrreous ooze

scarce

1

7

6

17

23O16'S 114°45'W

ferruginous calcareous ooze

"one

0

0

-1

18

29oO6'S 114o48'U

calcareous ooze

none

1

8

7

19

31028'S llZ"Sl'W

ferruginous calcareous ooze

none

0

D

-.?

none

0

0

3

14

21

23059'5 10@43'W

ferruginous calcareous ooze

22

39051'S loloo4'w

ferruginous calcareous ooze

some rad. in <125um sire

3

5

23

40024'S

calcareous ooze

sponge spit. in <125um

1

s

-.l 0

97033'11

1

25

34"3S'S 91o52'x

ferruginous calcareous ooze

l-Z% siliceous remains

1

5

26

32olo's

S9%'U

calcareous ooze

some radiolaria

1

13

9

27

31°54'U

88ol7'U

calcareous ooze

some radiolaria

1

I

-4

29

24o57'W 82'15'W

calcareous ooze

0

3

2

31

16"49'$ 78o39'W

siliceous ooze, red clay

many i-ad.in <125um size

16

20

12

none

34

9'58's

8Z010'W

green mud

many rad., diatoms in 125pm fraction

20

20

9

35

ll"oo's

87o24'w

siliceous red clay

abund. rad. in <125un,

22

22

19

9FJO5'U

36

14015'5

calcareous ooie

a few rad. in <125iin

1

11

6

41

12O39'S 117o22'W

ferruginous calcareous ooze

a few rad. in <125um

1

13

12

44

17ooo's 129o451w

ca1careous ooze

a few rad. in <125w1

0

9

-2

45

17olZ'S 136o37'W

ferruginous ca1careous ooze

a few rad. in <125um

0

0

-2

56

34O44'N 141*04'E

gray siliceous volcanic mud

rad. = 60% of sand sizes

26

27

20

57

37O40'N 145'26'E

volcanic radiolarian ooze

rad. i 60% of sand sizes

27

27

20

58

3E041'N 147O41'W

volcanic diatom or rad. ooze

rad. abund. in sand sizes

23

23

15

59

40'20'N 150'58'E

didtinnooze

abund. diat. in sand sizes

28

28

21

60

45'24'N 169'36'E

diatom ooze

abund. diat. in sand sires

29

29

23

61

44O16'N 137o37'W

gray clayey mud

rad. in sand sizes

10

10

62

40037'M 132o23'w

red clay

rad. in sand sizes

10

64

33'49'N 126??O'W

red clay

rad. in sand sizes

65

31°38'N 128oZE'W

red clay

rad. insandsizes

66

29'21'N 132oZO'W

red clay

i-ad.= 70% of sand sizes

69

26'13'N 142*02'W

red clay

rad. * 70% sand sizes

0

10

2

8

8

3

7

8

2

7

7

2

10

10

-1

70

24OOZ'N 145'33'15

red clay

rad.. diat. in sand sizes

6

8

-2

72

32O42'N 160'44'W

red clay

rad. in sand sizes

10

10

-2

73

33'27'N 145o3O'W

red clay

rad. in sand sizes

8

8

-2

77

21°18'N 13E036'W

red clay

rad. in sand sizes

10

10

1

79

12O40'N 137'32'W

radiolarian ooze

abund. rad. in said sires

21

21

17

80.

7O45'N 141'24'W

radialarian ooze

abund. rad. in sand sizes

22

22

19

81

3'0l'N 149o46'W

siliceous calcareoos oo*e

rad. in sand sizes

22

36

32

82

1048'5 152Q?'W

siliceous calcareous ooze

i-ad.

3

18

13

85

10'54'5 16lo53'W

ca1careous ooze

a few radioiaria

0

0

-1

88

12ooo's

gray mud

a few radiolaria

12

13

5

a9

27ooo's 109ooo'w

volcanic calcareous sand

none

0

0

-1

77QOO'W

in sand

sizes

Predicted oual content: (1) Opal ‘content of bulk sediment calculated from equation (1). (2) Opal content of carbonate-free sediment calculated from equation (1). (3) Opal content of carbonate-free sediment calculated from a non-biogenic SiO,:AiZO,

of 3~1.

Notes

minerals generally correlates with the amount of aluminum substitution for silicon in the tetrahedral layers. Magnesium is also a trace constituent in calcium carbonate supplied to the sediment in the tests of calcareous plankton; however, the concentrations of magnesium and carbonate do not vary systematically with calcium carbonate in this or other similar areas (SAYLES and BISCHOFF, 1972). When Fe. Fe’ or any of its cross products was chosen by the regression program as a second or third variable, the residuals increased indicating poorer fit to the data. Since iron is an important structural constituent of marine clays, especially authigenic clays, and can even substitute for silicon in tetrahedral sheets (BISCHOFF, 1972; EKLUND, 1974), it should account for some of the variance in the silica content of the clay fraction. However, the large fraction of the total iron present as amorphous hydrated oxides and as oxide coatings on sediment grains apparently obscures the relation between silicon and structural iron in the clay minerals. The three best equations are given in Table 2. Tests on additional data described below indicate that the first equation predicts the most reasonable non-biogenic silica contents for most sediments. Figure 1 (C) shows the non-biogenic silica predicted by this equation plotted against that measured in the opal-free sediments. Since many opal determination techniques are unpredictable when used outside the area for which they were developed. the regression equations were tested on additional sediment data for which chemical analyses and estimates of opal content were available. These data represent a wide range of sediment types including siliceous and calcareous oozes. terrigenous sediment, red clay and metalliferous sediment. They also cover an extremely wide geographic area. A summary of the samples, their lithologies, estimates of their opal contents, and the opal contents predicted by equation (1) (above) and a simple Si02:A1,0, of 3: 1 are given in Table 3. The data are from surface samples taken by the Carnryie (REVELLE. 1944). Qualitative estimates of the opal content of these sediments are taken from notes on the bulk sediment and the coarse fraction (> 125 mm), The carbonate-free opal contents predicted by the normative calculation are always consistent with the qualitative description of the sediment. The opal content of the calcareous oozes, in which siliceous remains are described as scarce or absent, is always less than 2 wt ;; (less than 10 wt “; of carbonate-free sediment). Most of the red clay samples (from the North Pacific) had some radiolarians in the coarse fraction. The normative calculation consistently indicated about 7-10 wt % opal in the carbonate-free fraction. This value is in good agreement with other estimates of biogenic opal in Holocene North Pacific red clays based on microscopic examination (HORN rt al., 1970). Estimates of the carbonate-free opal content of siliceous oozes sampled by the Carnegie ranged from about 30 to 50 wt ‘I,, opal. In contrast, the opal content estimated by a Si02:A1203 ratio of 3: 1 for siliceous oozes averaged Z&30 wt “,,. The negative opal contents predicted for many calcareous oozes and red clays by the 3: I ratio confirms that it tends to overestimate structural silica. Two possible types of error inay exist in the opal concentration predicted by the normative calculation. The first is systematic error due to inaccuracies in the regression equation itself. The second is poor reproducibility due to limitations in the precision of the chemical analyses. The mean absolute value of residuals for the regression equation used to estimate non-biogenic silica in this study is 7 wt “/, non-biogenic silica (see Table 2). The accuracy of equation (1) for determining opal is, therefore, about +7 wt ‘>A,The error in precision of the chemical analyses can be determined from the standard error of multiple determinations of the same element. The largest standard error was +5”<, of the measured value for Al and Si in low concentrations, but in practice duplicate determinations of opal content on the same sample never differed

675

by more than 2 wt 7” opal. The k7 wt “i, estimate of error based on samples with no opal is probably a more realistic estimate of error than one based on a standard curve of known amounts of opal in an opal-free matrix. Such a curve would not reflect the errors caused by using the equation on a wide range of sediment compositions. The accuracy of the technique is comparable to that of the most widely used opal determination method, the X-ray technique. ELLIS (1972) indicates that the accuracy of the X-ray technique, based on a standard curve of opal mixed with an opal-free matrix, is about +5 wt “A opal.

CONCLUSIONS The normative calculation method developed for the central Pacific gives reasonable estimates of opal content for a wide variety of deep-sea sediments. Most importantly, because it is unaffected by diagenetic changes which take place in opal as it ages, the method may be used to determine the opal contents of pre-Pleistocene deep-sea sediments. No technique will adequately estimate the original opal content of sediments from which most of the opal has escaped by diffusion. but such losses may be restricted to sediments near extensive hiatuses (DONNELLY and MERRILL, 1977). The equation used in this study is clearly an improvement over the simple Si0,:A120, ratio used by BOSTROM et al. (1972) which can yield errors of as much as 30wt 7” opal. The inclusion of iron in the equation used to predict non-biogenic silica might improve the opal estimates, but would necessitate the separation of amorphous iron from that incorporated into silicates, representing a substantial increase in the amount of laboratory preparation and analysis. Although the equation used here to determine opal is applicable to a wide range of sediments from the deep ocean, it has not been validated for nearshore sediments which are rich in terrigenous debris. Such sediments contain greater quantities of illitic and kaolinitic clays than the samples discussed here. Such clays have significantly more aluminum than most pelagic deposits. Given a set of opal-free samples from nearshore sediments. however, it should be possible to develop a similar relationship by which to determine their opal contents. Acknowledgemmts~Samples for the project were provided by the Deep Sea Drilling Project. Carbonate determinations were made by C. RATHBUN. P. PRICE and M. CLAUSON gave advice in making X-ray determinations of opal. In addition, the manuscript has benefited from the suggestions and review of TJ. H. VAN ANDEL and G. Ross HEATH to whom I would like to express sincere thanks, Financial support was provided by National Science Foundation Grant No. Fa-31468 and a fellowship from the Amoco Production Company. REFERENCES AOKI S., OINUMA K. and SUDO T. (1974) The distribution of clay minerals in the Recent sediments of the Japan Sea. Deep-Sea Res. 21, 299-310. BERNAS B. (1968) A new method for decomposition and comprehensive analysis of silicates by atomic absorption spectrometry. Anal. Chem. 40, 1682-1686. BISCHOFF J. L. (1972) A ferroan nontronite from the Red Sea geothermal system. Claqls Claq’ Minerals 20, 217-223. BEZRUKOV P. L. (1955) Distribution and rate of deposition of silicate sediments in the Sea of Okhotsk. Dokl. Akad. Nauk SSSR 103,473476. BI~CAYE P. E. (1965) Mineralogy and sedimentation of Recent deep-sea clay in the Atlantic Ocean and adjacent seas and oceans. Bull. Geol. Sot. Amer. 76, X03-832.

NOkS

676

BOSTROMK. and PETERSONM. N. A. (1969) The origin of a aluminum-poor ferromanganoan sediments in areas of high heat flow on the East Pacific Rise. Mar. Geol. I, 421447. B~STROMK. and FISHER D. E. (1971) Volcanogenic uranium, vanadium and iron in Indian Ocean sediments. Earrh Plunet. Sci. Lett. 11, 95-98. BOSTROM K., JOEWJU O., VALDESS. and RIERAM. (1972) Geochemical history of South Atlantic Ocean sediments since late Cretaceous. Mar. Geol. 12. 85-122. CALV~RTS. E. (1966) Accumulation of diatomaceous silica sediments in the Gulf of California. Bull. Geol. Sot. Amer. 77. 569-596. CHESTERR. and ELDERFIELD H. (1968) The infrared determination of opal in siliceous deep-sea sediments. Gro. chim. C~s~~~c~~rn~Acta 32. 1128-I 140. DONNELLYT. W. and MERRILLL. (f977) The scavenging of magnesium and other chemical species by biogenic opal in deep-sea sediments. C/rem. Geol. (in press). EKLUNI~W. A. (1974) A microprobe study of metaliiferous sediment components. MS. thesis, Corvallis, Oregon State University, 77 pp. EI.LISD. B. (1972) Holocene sediment of the South Atlantic ocean: the calcite ~mpensation depth. and concentration of calcite, opal and quartz. MS. thesis, Corvallis, Oregon State University, 77 pp. ELLISD. B. and MOORF.T. C. JR. (1973) Calcium carbonate, opal and quartz in Holocene pelagic sediments and the calcite compensation level in the South Atlantic Ocean. J. Mar. Res. 31, 210-221. HL~THG. R. (1969) Mineralogy of Cenozoic deep-sea sediments from the equatorial Pacific Ocean. Buil. Geoi. Sot. Amrr. 80. 1997-2018. HCA~ G. R.. PIZZASN. G.. CORLISS J. B. and DYMOND J. D. (1976) Pelagic sedimentation in the North Pacific. 1. Quantitative estimates of the clay mineralogy of less than 2 micron fractions. In Program and Abstracts 13th Annual Meeting, The Clay Minerals Society. 25th Clay Mineral Conference, Corvallis, Oregon, 1976. HEATHG. R.. MOORET. C. JR., DAUPHINJ. P. and OPDYKE N. D. (in preparation) Quartz, organic carbon, opal and

calcium carbonate in Holocene, 6~,~yr, and Brunhes-Matayama age pelagic sediments of the North Pacific. HORND. R., HORN B. M. and DELACHM. N. (1970) Sedimentary provinces of the North Pacific. In Geological Investigations of the North Pacific. (editor J. D. Hays). Geol. Sot. Amer. Men?. 126, I-22. HOARD D. C. (1972) Interactions of biogenic opal, sediment and seawater in the central equatorial Pacific. Ph.D. dissertation, Honolulu, University of Hawaii, 81 pp. JONESJ. B. and SEGNITE. R. (1971) The nature of opal 1. Nomenclature and constituent phases. d. G&. Ser. Azutrizliu 18. 57-68.

I_~~~TzIN A. P. (1971) Geochemical, mineralogical, paleontologic studies, Leg 6. In Initial Reports of the Deep Sea ~rj~~j~g Projcef, (editors A. G. Fisher et al.), Vol. 6, pt. I. pp. 829-961. U.S. Government Printing Office. MARTIE J. H. and KNAUERG. A. (1973) The elemental composition of plankton. Geochim. Cosmochim. Actu 37, 1639-1653. MEHRA0. P. and JACKSONM. L. (1960) Iron oxide removal from soils and clays by a dithioniteecitratc system buffered with sodium bicarbonate. In Proc. 7th National Cotlf: Clays and Clay Mi~~ru/s~ (editor A. Swineford), pp. 31l--327. REVELLER. (1944) Marine bottom samples collected in the Pacific Ocean by the Carnegie on its seventh cruise. Carnegie fnst. Wusii. Puhi. 556, 182 pp, SAYLESF. and BISCHOFF J. L. (1973) Ferromaneanoan sediments in the equatorial East Pacific. Earfh Planet. Sci. Lrtt. 19, 33s-336. TILL R. and SPEARSD. A. (1969) The determination of quartz in sedimentary rocks using an X-ray difiactometer. Clays CIay Minerals 17. 323.-327. V,lv A~nr-I_ TJ. H.. HEATHG. R. and MOORE T. C. JR. (I 975)Cenozoic history and Paleoceanography of the central equatorial Pacific. Geol. Sot. Am. Mrm. 143, 134 p, WEAVERC. E. and POLLARDL. D. (1973) The Chemistry qfCIa>> ~j~er~z~~, Vol. 15.Dez,~~~pn~e~~.~ itz S~~im~nt~~og~. Elsevier, New York.

Goochimica etCovnochlmtca Acta. 1977. Vol.41.pp.676to6x1.PergamonPress. PrmtedinGreatBritain

Mass spe~trometrie isotope di~utiou analyses of tellurium in meteorites and standard rocks C.L. SE/~(?H,*J. R. DE LAETEKand K. J. R. ROSMAN Department of Physics, Western Australian Institute of Technology, South Bentley, Western Australia (Rrcriard

9 August

1976; uccepted in rezised form 26 ~I~~~~nher 1976)

Abstract-A stable isotope dilution technique using solid source mass spectrometry has been used to determine the elemental abundance of Te in 25 chondrites. 3 achondrites, 1 tektite, and I:! standard rocks. Mean values for the Cl, C2, and CV3 meteorites are 2.34. 1.48, and 1.03 ppm, respectively; or atomic abundances for Te (normalized to Si = 10h atoms) of 4.84, 2.49, and 1.46. The atomic abundance obtained for the Cl chondrite Orgueil is significantly lower than the accepted value of 6.42. As a consequence we recommend that the ‘cosmic’ abundance of Te and Xe should be re-examined. The depletion ratio for Te in ordinary chondrites of 0.10, is about the same as that for Zn. Elemental abundances of Te in 12 standard rocks are in the ppb range.

1. lNTRODUCTION

‘COSMIC’ abundance tables (e.g. CAMERON, 1973) provide the basic data for theories of nucleosynthesis and cosmo*Present address: School of Education, Murdoch University, Murdoch, Western Australia.

chemistry. These data have been largely based on the Type I carbonaceous chondrites (Cl), because it is believed that they closely approximate the condensable fraction of primordial solar system material (ANDERS.1971). KRKHENBDHL et al. (1973) have determined the abundance of 17 trace elements in three Cl and three C2 chondrites. As a result of these data, CAMERON(1973) has estimated the