Chlorine and fluorine abundances in ultramafic rocks

Be0~himi~atC~~ochimicaAccta,lQ6gVol.S2,pp.853

to 968. Pergamon Press. PrintedinNorthernIreland

GEOCHEMXAL NOTES

Chlorine and fluorine abtzndancesin ulWc

rocks

A. M. i%UEBlER* W. H. HWANO and W. D. JOEXS Department of Earth S&noes, Washington University, St. Lou&, Missouri (&ze&ed 26 dray 1967; -p&d

& reu&wdform23 iVovtrnber 1967)

Abstract-A

total of 66 ultrama& rocks of various modes of occurrence haa been analysed for chlorine and fluorine by a rapid speotrophotometrio method. The different modea of ocourrence seem to have different abundanoe levels of Cl and F, but this may be due to secondary effeotts.

of the concentrations of the halogens chlorine and fluorine in uhramefic rocks and &N-DELL (1963) reported an have yielded variable and inconsistent results. KURODA average of 0.06 per sent Cl for 13 ultramafio ro&a (not including serpent&&es), but stated that bemuse of the small number of ~amplea and the variability in the chlorine contents the exact figure was of little significance. E-Y (1968) determined the chlorine contents of samples of aerpentinised dunites, peridot&es and pyroxenit- from eight diamond drill holes from Ontario, Canada, and found an average of 0.23 per cent Cl for the serpentiniaed dunites but an average of only O-06 per cent Cl for the serpentinised peridotitea. EARLIX made fluorine determinations on three of the eamples, and found that none aontained more than O-01 per oent F. HESS and OT.&OI~~ (1964) found that Cl varied from 0604 to O-031 per cent in 13 serpentinitea from the AMSOC oore hole near Mayaguez, Puerto Riao, and that F was leee than 0.006 per cent for 12 of the 13 samples. HOERING and P-R (1961) determined the CI aontents of two dunites to (1963) reported analyses of 8 peridotite inolusiona be O-014 and O*Ol? per oent Cl. GREENUND from South African kimberlite pipes which averaged 650 ppm Cl and 320 ppm F; Cl ranged from 140 to 1000 ppm, F ranged from 220 to 460 ppm. Five eoiogites from the same pipes averaged 160 ppm Cl and 330 ppm F; Cl ranged from 100 to 270 ppm, F ranged from 140 to 690 ppm. Due018 (1963) reported fluorine analyses for 41 serpentinites from a single serpentine mass near Mariposa, California, which averaged O-063 per oent F. These studies of chlorine and fiuorine in ultrama& rooks have been either confined to a few isolated UltramaSo bodies or limited to a small number of sampies. In nearly every ease the analyses have been made only on alpine-type ultramafic intrusions or their serpentinised equivalents. We have analyzed a total of 66 ultramafio rocks for chlorine and fluorine by a epectrophotometric method desoribed by HUANO and JOHNS (1967). The 55 samples, which are of world-wide geographic distribution, include alpine-type intrusions, both fresh and serpentinised, ultramafic inclusions from basalt8 and kimberlite pipes, and ultramafic zonea from Rtratiform sheets. We have attempted to define the levels of chlorine and fluorine in these various types of ultramafics rooks. Other geoehemical data are available for most of thttsesamples (S~~JEBER and MURTBX, 1966; STTJESERand GOLES, 1967). &~EEWINATXOXS

* Present address: Department of Geology;y/Miami University, Oxford. Ohio 46066. 6

363

354

Geochemical

notes

EXPERIMENTALMETHOD AND RESULTS The experimental method employed was that described by HUAN~ and JOHNS (1967). This is a rapid spectrophotometrio method employing separate aliquots of a single sample dissolution for simultaneous determination of Cl and F. The method involves fusing the powdered rock material in Na&Os-ZnO, subsequent digestion in water, and spectrophotometric determination of Cl (using the stable oolored ferris thiocyanate complex) and F (using the ziroonium Eriochrome cyanine R complex). Table 1. Chlorine and fluorine analyses of ultramaflc rocks (ppm) Sample Ultralnafic in&&one

Ant-1 EPI-1 PI-I Mex-1 sws-1 sws-2 PCNA-6 PCNA-14 Eur-1

Eur-2 At-11 Af-19

Af-7 Af-8 B-9 Af-20 Af-21 WPI-6-1

F

in ba.aalta

Ross Island, Antarctica dun&e Guadalupe Island peridotite Hawaiian 1801 Flow pyroxenite Chihuahua, Mexico peridotite Peridot Cove, Arizona peridotite San Carlos, New Mexioo peridotite Ludlow, California peridotite Alaska-Canada Border peridotite Kapfenstein, Austria peridotite Eifel Dietriot, Germany peridot& Monduli, Tanzania peridotite Rift Zone, Uganda peridotite Mean

Ultrama$c inc.&&s

CI

28 87 18 34 9 45 16 24 31 7 63 23 32

-* 19 62 14 24 8 22 86 21 f

in kimberl~te pipes

Weseelten Pipe, South Afriaa garnet peridotite Wesselten Pipe, South Africa garnet peridotite Roberts Vi&or Pipe, South Afrioa garnet peridot&e Bultfontein Pipe, South Afrioa garnet peridotite Visser Pipe, Tanzania eologite Kakanui, New Zealand peridotite Mean

299 282 142 257 52 109 200

190 236 390 348 96 254 252

123 184 33 53 65 94 68 63 72 103 106 88

16 12 -

Alpine ultramccfio intmaiona Fresh (O-5 % serpentine) Mt. Albert, Quebec peridotite AppM-1 Webster-Addie, N. Carolina dunite AppM-2- 1 Twin Sisters, Washington duuite PCNA-1 Tulameen, British Columbia dunite PCNA-5-l Sonoma Co., Calif. peridotite PCC-1 Preaoher Creek, Wyoming peridotite PCW-l-l Tinaquillo, Venezuella peridotite Car-l Dun Mtn., New Zealand dunite WPI-3 Shikoku, Japan dunite WPI-5 Ahnklovadalen, Norway dunite Scan-4-2 Standard dunite NBS 4999 Mean

20 40 58 48 19*

356

Geochemiaal notes Table 1 (continued) Cl

I?

Partially Serpentinised (10-75 % serpentine) Tulameen, British Columbia serpentinite PCNA-5-3 Canyon Mtn., Oregon dunite PCNA-7-l Cantwell, Alaska dunite PCNA-9 Med-2 Livorno, Italy serpentinite Konya, Turkey dunite Med-6-2 Yate, New Caledonia peridotite WPI-2 Ghost Mtn., Ontario serpentinite PCC-12-S Car-2-3 Mayagues, Puerto Rico serpentinite Scan-4-4 Ahnklovdalen, Norway garnet peridotite Mean

232 139 305 335 199 187 388 673 214 304

-

Serpentinised (95-100 % serpentine) Aus- 1 Queensland, Australia serpentinite AusNew South Wales, Australia serpentinite Kalgoorlie, W. Australia serpentiuite AusMed-4 Bou Aszer Mine, Morocco serpentinite Car-2-4 Mayaguea, Puerto Rico serpentinite Guatemala serpentinite PCNA-3-2 Glenarm, Maryland serpentinite AppM-4 Orford Lake, Quebec serpentinite AppM-8-3 Thompson River, Manitoba serpentinite PCC-6-l Bird River, Manitoba serpentinite PCC-IO-2 PCC-14-3 Crow Lake, Ontario serpentinite Mean (excluding PCC-6-1)

88’7 81 179 780 444 93 146 133 2870 35 173 296

Sample

8 <8

106 320 40 20 20 1660 48f

UltramaJic zontwfrom stratiform ah&a Man-l-l BI-1 A-l PCC-1-l

Stillwater, Montana peridotite Isle of Rhum peridotite Kola Peninsula, Russia pyroxenite Muskox, N.W.T., Canada serpentmite

185 111 71 2080

138 88

1890 1790

204 66

Other oaaurrenatw I&W-l Mar-2

Mid-Atlantic Ridge serpentinite dredge St. Paul’s Rocks peridotite

* Below detection limit of 8 ppm F. In order to determine the precision of the analyses at the levels commonly found in ultramafia rocks, the U.S.C.S. standard peridotite, PCC-1, was analyzed a total of six times during the course of the work. The chlorine content was found to be 85 f 5 ppm Cl. The fluorine content was below the lower limit of detection, 8 ppm F. Thus the precision of the chlorine analyses is better than 10 per cent. Several other samples were analyzed in duplicate with a comparable precision. JOHANSEN and STEINNEB (1967) reported 66 ppm Cl for PCC-1, as determined by neutron activation analysis. The results of the determinations of chlorine and fluorine for 66 ultrama rocks m presented in Table 1.

356

Geoohemicalnotes

A total of twelve &ram&o inclusionsfrom basalt host rooks from continental and oceanio environments was 8naly5ed for chlorine and fluorine. The ohlorine oonaentrationswere found to be oonsistently low, even though the samples were very wide-spread on 8 geographio basis (Table 1). The average (arithmetricmean) chlorinecontent of these twelve inclusionsis 32 ppm Cl. EARLBY (1958) reported 0.01 per cent chlorine in a single okvine bomb in Keoent basalt from H8waii. Of the various modes of occurrenceof uItramaflorocks examined in the present study, the basalt inclusionsshowed the lowest abundance level of ohlorine. This may be related to the fact that the basalt in&sions, 8s a group, show the Ieast amount of alteration of any type of ultmmafio rock studied here. Most of the fluorinecontents of the basalt inclusionsare either below or very near the lower limit of deteotion (8 ppm F) of the experimental method employed. An everage of approximately 20 ppm F is indicated. The two samples which showed relatively high fluorine conoentrations (SWS-I and Af-19) were found through X-ray diffraction studies to contain a mica and an amphibole, respectively, and it is reasonable to suppose that the additional fluorine is located in these minerals, eubstituting for hydroxyl ions. Kim&w&

pipe i&W

Five ultramafic inclusions and one eclogite inch&on from kimberlite pipes were analyzed in the present study (Table I). The average chlorine content of the kimberlite inclusions (290 ppm Cl) is six times higher than that of the basalt inolusions. The kimberlite inclusions consistently contain more fluorine than 8ny other &ram&o rocks, averaging 262 ppm F, and are the only group of uhramafio rocks studied in which fluorine exceeds chlorine. The high fluorine concentrations probably are dabed to the abundance of pblogopite and/or amphibole which were found by X-ray diffraotion. The relatively high concentrationsof both chlorine and ftuorinein the kimberlite ultramaflo inclusionswere probably to be expeated bemuse of their ooonrreneein the kimberlite matrix, which is a brecciated, serpentinised, peridotitic rook, showing the effeats of volatiles. The extensive hydration and carbonation of the kimberlite suggests the former presence of large amounts of gases. Alpine intrwiosa A total of 31 alpine-type ~tr~~~ rooks, varying widely in degree of ~~enti~~tion, was analyzed (Table 1). On the basis of thin section examination the samples were conveniently divided into three groups according to the amount of serpentinepresent. The “‘fresh” samples, which contain at the most only minute traces of serpentine, have an average chlorine oontent of 88 ppm Cl. The “partially serpentinised” samples whioh contain from about 10 to 76 per cent serpentine, average 304 ppm Cl, indioating a oonsiderableintroduction of chlorine during serpent&&&ion. The “serpentmites”, which are in most oases totally altered to serpentine, have an average chlorine content of 296 ppm Cl. Most of the fluorine abundances of the alpine intrusions are either below or very near the lower limit of detection (8 ppm F) of the experimentalmethod employed. There was apparently no introduction of fluorine upon serpentinisation. An overah average of about 28 ppm F for 811the alpine ultramafio intrusionsis indicated, whioh is comparable to the fluorine abundance level found in ultramafio inclusionsin basrtlt. Only two prior studies have reported at any length on chlorinein uhramaf%~rooks. &JRODA and SANDELL(1953) reported a pronounced v&ability in chlorine content of ~t~rnafi~ rocks (Tom O-006 to 0.24 per cent in 13 mmpkza). There was some indicationthat the chlorinecontent tended to rise with the water content, but the relationship ~8s not dose. From the H,Of contents tabulated by KWRO~Aend SANDNLL,it would appear that most of their ultram&o samples were largely serpentinised.

Geochemical notes

367

EABLEY (1968) reported chlorine analyses for a total of 66 ultramafio rocks, all highly serpentinised, from eight diamond drill holes in Ontario, Canada. The amount of ohlorine in serpentiniseddunite ranged from about 0.1 to 0.7 per sent and averaged O-23 per cent for all of the samples, whereasthe serpentinisedperidotitesaveraged only 0.049 per cent ohlorine.From determinationsof the water oontent of oomposite samples, EABLEYsuggestedthat a high chlorine content was associated with a high water aontent, but a high water oontent did not necessarily indioate a high chlorine oontent. The chlorine analyses of the present study indicate a substantial introduction of chlorine on serpentinisation. There is, however, no close relationshipbetween dwee of serpentinisation and oblorinecontent; the totally serpentinisedsamples contain no more chlorine,on the everage, than the partially eerpentinisedsamples. A small amount of ohlorinemay have been introduoed even into the “fresh” alpine samples, as most of them inevitably show inoipientserpentinisation. Thus the chlorine content of the serpentine-freealpine ultrama6o rook may be just about the same as that of the ultramafioinclusionsin basalts. Other occwrencea Four samplesfrom ultramaficzones in stratiformsheetswereanalyzed for chlorineand fluorine (Table 1). The results show no signi&& differencesfrom the ranges of values found in the alpine ultramafie occurrences. Here again, the Muskox serpentinite has a markedly higher chlorine content than the other relatively unaltered samples within the group. Two samples from the Mid-Atlantio Ridge, one a dredge haul and the other from St. Paul’s Rooks, were also analyxed (Table 1). They both have a relatively high content of chlorine, which might be expected in view of their contaot with sea water. CONCZUSIONS The elements ohlorine and fluorine are rather easily introduced, probably during secondary alteration of ultrama& rocks. This may aooount for the wide variability, especially of chlorine, in previously reported analyses. The different modes of occurrenceof ultrama& rocks seem to have different abundance levels of chlorine and fluorine, but this may be related to secondary effects. We suggest that the amounts of primary chlorine and fluorine in ultrama6o rooks of all kinds are most nearly represented by the abundance levels found in ultrama6o inalusions in basalts (32 ppm Cl, 20 ppm F). These values may be compared with those listed for ultrama& rocks in recently compiled tables of the crustal abundances of the elements. TUREIIIANand WEDEPOHL(1961) suggested85 ppm Cl and 100 ppmF for ultrama& rocks, whereasVINOQRAIIOV (1962) suggested 50 ppm Cl and 100 ppm F. The present study thus indicates that a downward revision of previously compiled chlorine and fluorine abundanaesin ultramafic rocks would be neoessary if serpentinisationis not a magmatic process. AotiZed~This work was partially supported by a grant from the Petroleum Research Fund of the American Chemical Sooiety.

DUBOISR. S. (1963) Remnent, induced, and total magnetism of a suite of serpentinitespecimens from the Sierra Nevada, California. J. Ceophya. Res. 68, 267-2’78. E-Y J. W. (1968) On chlorine in serpentiniseddunite. Amer. Minem;Z. 43, 149-155. GREENLANII L. P. (1963) Unpublished Ph.D. Dissertation, Australian National University. HESS H. H. and OT~LORAG. (1964) Mineralogicaland chemioal composition of the Mayaguez serpentinitecores. In A Study of Serpentinite, pp. 152-168. NAS-NRC Pub. 1188. HOERIN~T. C. and PARIZERP. L. (1961) The geochemistry of the stable isotopes of chlorine. Ueochim. Cormaochim.Acta 28, 186199.

Geoohemioal notee

368

HUANO W. H. and JOHNS W. D. (1967) Simultaneous determinations of fiuoriue and chlorine in &&ate rooks by a rapid ~~~ophoto~~e method. Ad. C&m. Acta 87,606-616. JOHANSEN 0. and STBIINNIBS E. (1967) Determination of &kwin5 in U.S.G.S. standard rooke by neutron aotivation analysis. aeochipn. Oo~o&&n. Acta 81,1107-f 109. KIINODAP. K. and SANDEI,LE. B. (1953) Chlorine in igneous rocks. BulE. &oZ. XOG.Amw. @, 879-896. STUEBER A. M. and MURTHY V. R. (1966) Strontium isotope and alkali element abundanoes in ultramaflct rocks. @eu&nt. coenwrchim. A~da 80,1243-1269. Sm A. M. and GOLES G. G. (1967) Abundanozw of Ka, l’&n, Cr, Se and Co in ultrama& m&s. G%octim. ~o~&h~rn. A&a 81,75-93. ‘PUREKIANK. K. and WEDEPOHL K. H. (1901) Distribution of the elements in some major units of the earth’s crust. &ol. SOG.Amw. 232621. 7R, 176-192. VINOUR~DOV A. I?. (1962) Average contents of ohemical elements in the prinoipal types of igneous rocks of the earth’s crust. geochemistry, 7, 641-664.

CIeochlmica et floemoohtmios Acta,1968,Vol. 32, pr. 3% to 361. PergamonPress. Printedin NorthernIreland

PAT HAWU, H. K. SCIINOES* and A. L. BURLINUAME Department of Chemistry and Space S&noes Laboratory, University of California, Berkeley, California 94720

Ab&t&---Several eerier of aromatic carboxylic acids (in&ul.ing mono-, di-, and trimethyl benzoic; mono-, di-, and trimethyl propanoio; mono- and dimethyl butanoic; indanoio and tetrahytinaphthoio; and naphthoio aoids) have been isolated from the extra& of Colorado Green River Shale. Gas liquid chromatographio and both low and high resolution maas apeotrometric techniques were utilized for separation and oharaoterization reapeotively.

We wish to report on a maas speotrometric study of the carboxylio acids in the Green River Shale (Eocene), specifically on the occurre~me and type of aromatio acids. Although normal &o, ant&o and isoprenoid acids have been reported in this sediment [ABELSON and PARKER (1962), LAWLOR and ROBINSON (1965), LEO and PARKER (1966), E~LINTON et al. (1966), RAXSAY (1966), DOUGLASet al., (1968) HAUL et d. (1967)], aromatic aoids have received little attention thus far. In faot, from geological so~weesno individual aromatio acids appear to have been isolated [Loc~~~andLrrr~~~N (19561, WKITE~IAD~~~BREUER (1963)lwith theexGeption of benzoio acid, which has been isolated from soil [KONONOVA( IBSl)]. The presence of aromatic hydrocarbons hae been established and, of course, aromatic hydrocarbons have been studied extensively using mass spectrometry (RIJIXIXNoAXEand SCNNOES,1968; SCIINOESand BUIUNOAME, 1966). We find that the Colorado sediment contains several eeries of aromatie aoids whose methyl eaters range in molecular weight from 150 (methy substituted benzoate) to 242 (trimethyl naphthoate). For each mol. wt. a series of ieomers is present. Although the maas s~t~rnet~~ data do not oompbtely detie positions of aroma& substitution for individual a&da they + Present Wisoonsin.

address:

Department

of Bioohemistry,

University

of Wisconsin,

Madison,