Variations in the chemical and mineralogical composition of rim and plains specimens of the Cañon Diablo meteorite

Variations in the chemical and mineralogical composition of rim and plains specimens of the Cañon Diablo meteorite

Geochhics et CosmochimicsActa,1967,Vol.51.pp.1885to 1892.Pergamon Press Ltd. Printed in Northern Ireland Variations in the chemical and mineralogical...

587KB Sizes 1 Downloads 49 Views

Geochhics et CosmochimicsActa,1967,Vol.51.pp.1885to 1892.Pergamon Press Ltd. Printed in Northern Ireland

Variations in the chemical and mineralogical composition of rim and plains specimens of the Caiion Diablo meteorite* CARLETON B. MOORE, PAMELA JOST BIXRELL and CHARLES F. LEWIS Center for Meteorite Studies, Arizona State University, Tempe, Arizona (Received2 May 1967; accepted in revbedform

12 June 1967)

Abstract-Quentitative chemicaland modal mineralogicalanelyaeswere made on samplesrepresenting “rim” and “plains” populations of the Ca on Diablo meteorite. These investigations indicated that the volume per cent of the taenite plus plessitephase, the volume per cent of oxide phaee, the total weight per cent nickel and the density are different to a level of signif%canoe of O-05 for each of the populations. Other phases and chemical constituents determined showed no sign&ant differences. The values obtained for the components showing diffemnoes were: taenite plus pleseite, ‘%im” 3.98 and “plGns” l-53 vol.%; oxide (lswrencite), “rim”
Diablo ootahedrite has a.lways been of great interest because of its association with the Meteor Crater, Arizona, its relatively great abundance and wide distribution in collections, and the discovery of diamonds in some fragments. A recent burst of aotivity in the investigation of the origin of the meteoritic diamonds has supported NININQER’B (1960, 1956) earlier observation that the diamonds are found in shocked specimens from the rim of the crater, with a single exception (HEYMANN et al., 1966), and not from specimens found on the plains surrounding the crater. Typical speoimens from these two populations have also been shown (NININQER, 1956 ; LIPSCWTZ and ANDERS, 1961~1,b) to have other noticeable differences. The “rim” specimens quite commonly show evidence of reheating while the “plains” specimens were not reheated. These observations have led several investigators, notably NININOER (1956) and LIPSCHUTZand ANDERS (1961 a, b) to theorize that the diamonds were formed by shock waves produced in the parent body upon collision with the Earth rather than by primary hydrostatic pressure in a large celestial body. The following investigation was undertaken to examine the chemical and mineralogiocll makeup of the Canon Diablo meteorite, particularly the “rim” and “plains” populations. A number of investigators have reported chemical analyses of Caiion Diablo. The moat notable studies have been by GOLDBERG et al. (1951) and WASSON (1967). Both of these investigators reported determinations of selected major chemical constituents and the trace elements germanium and gallium. THE CA%ON

* Presented at the XXth Internationel Congressof Pure and Applied Chemistry, Division C, Cosmio Chemistry, Moscow (1965). Contribution No. 25, Center for Meteorite Studies, Arizona State University. 1885 23

1886

C. B. MOORE,P. J. BIRRELLand C. I’. LEWIS

(~KDBERO'S average value from eleven diverse specimens of 7.11 wt.;,; nickel id commonly taken as a typical value. He obtained a nickel value of 8.22 for Canyon Diablo (1936) while WASSONfound 7.93 for Cafion Diablo (1936) and 8.20 for Canyon Diablo (1949). Valid statistical data of this type are important in considerations of chemical and mineralogical variations in large iron meteorite bodies, of whether conditions suitable for the formation of diamonds are localized, and whether other iron meteorites such as Odessa, Camp Verde, hshfork and Bloody Basin are related to the Canon Diablo fall. It may also enable the specimens known as Canyon Diablo (1936) and Canyon Diablo (1949) to be properly evaluated with respect to the majority of “common” Canon Diablo specimens.

EXPERIMENTAL PROCEDURE Thirty-twoCtion Diablospecimensfrom the rim of the Meteor Crater, Arizona and fourteen specimens from the plains surrounding the crater were selected as a first suite to be investigated. These specimens from the Nininger Meteorite Collection at Arizona State University were identified by HI. H. NININUERas “rim” and “plains” specimens but the exact location of find was not recorded. It is probable that the rim specimens came from the north-east corner of the crater where the large majority of rim specimens seem to be fouud. In order to assure statistical sampling each specimen selected was a whole uncut individual. Two parallel cuts approximately 1 cm apart were made through the center of each specimen. The average size of the rim specimens available is smaller than the average of the plains specimens and, therefore, a larger number are required to provide an equal area of cut surface for study. The total area studied for the rim and plains specimens was about 310 cm2 each. One side of the center slice was ground and polished for metallographic analysis. The polished face was etched with a 5 % nital solution for 10 sec. A modal (point count) analysis was made with a reflected light microscope equipped with a mechanical stage to move the specimen. The phase present under the crosshairs was reoorded for each movement of the stage. Approximately 2500 points were counted for each specimen. The mineral phases recorded in the investigation were kamacite, taenite plus plessite, cohenite, schreibersite, troilite, graphite and oxides. A semple from each specimen was prepared for chemica1 analysis by driIhng two grams of metal from the face of one of the side pieces. Care was taken to avoid large inclusions in the samples. For each specimen iron was determined volumetrically by titration with potassium dichromate, nickel gravimetrically by precipitation with dimethylglyoxime, cobalt colorimetrically with nitroso-R salt and carbon by a combustion technique. For selected samples the other end piece was prepared for density determination by grinding all the oxide coating from the surface of the meteorite fragment. The density of these specimens was then determined by an immersion technique using n-propyl alcohol. RESULTS

AND DISCUSSION

The results of the modal analysis are given in Table 1. For each of the phases counted two methods of summarizing the data are given. In the &St, a mean of the points counted (equivalent to volume per cent) is computed ; in the second, a sum of the points oounted for each individual specimen multiplied by the area of each specimen and divided by the total area is made. Both of the values oomputed for each phase counted appear to be fairly close. In order to compare the values obtained from the rim and plains populations, standard deviations from the mean were calculated and using Student’s t distribution, tests of significance were run for each of the phases. Only the taenite plus plessite and oxide were found to be significantly different at a level of signi&ance of O-05. The taenite plus pIessite was higher in the rim specimens and the oxide in the plains specimens. It should be pointed out that

Variationsin the chemicalandmineralogicalcompositionof the C&ionDiablometeorite 1887 the presence of oxide in the interior of the slices probably is an indication of the presence of the mineral lawrencite. There appeared to be no structural control of the looation of the oxide along oracks or phase boundaries. The reaction of some of the plains specimens with the atmosphere to form oxide was so rapid that noticeable amounts formed on some freshly polished surfaces during the approximately 4%min period required for an analysis. MICHAEL LIPSCWTZ (personal communication) has recently oonflrmed this observation that the rim specimens are oxidized much less also pointed out at the time of presentation rapidly than plains samples. LIPSCHUTZ of this paper in Moscow that it is often very difficult to distinguish between schreibersite and cohenite in polished section. In the preparation of our specimens schreibersite had a greater tendency to produce small fractures thus giving one distinguishing characteristic. Acknowledging the possibility that there might be some confusion of these two phases, tests of significance were run on the difference between the totals of schreibersite plus cohenite in both populations. No significant difference was found. An interesting feature noted in the rim specimens, but not in the plains specimens, was the iron phosphide eutectio structure in the schreibersite as illustrated in PERRY (1944). Also, around the edges of larger specimens and throughout the smaller specimens rhabdites were changed into shapeless particles. Diamond was found in two of the rim specimens but not in any of the plains specimens. Each specimen was investigated for evidence of shock following the methods of HEYBXANN, LTPSCHUTZ, NIELSENand ANDERS (1966). The criteria used were as follows : (1) Heavily shocked: Completely recrystallized kamacite throughout the specimen, Carbon diffusion zone in at least part of the specimen. (2) Moderately shocked : &recrystallized kamacite in a part of the specimen. (3) Lightly shocked: Only Neumann lines. The results for individual meteorites are indicated in Table 1. The rim specimens were 74% heavily shocked, 20% moderately shocked and 0% lightly shocked. The plains specimens were 0% heavily shocked, 45% moderately shocked and 66% lightly shocked. Of the two diamond-bearing rim speoimens one was heavily shocked, the other moderately shocked. It is interesting to note that the shock effects found for the rim speoimens in this study are essentially identical with those found by HEYMA- et al. (1966). The plains specimens show some difference in degree of shocking between the present investigation and that of HEM~~NNet al. but in neither did any of the plains specimens exhibit evidence of heavy shocking. The results of the chemical analyses are given in Table 2. Statistical studies using Student’s t distribution indioate that the difference between both the niokel and carbon contents are signi&antly higher in the rim specimens to a level of signiflaance of 0.05. Note that the carbon values are for metallio samples in whioh an effort was made to avoid large cohenite areas. This nickel difference coincides with the significant difference observed between the two populations for the high niukel tsenite plus plessite. Also given in Table 1 are the values for the density determinations. They too show a significant difference at a level of significance of O+Ob,with the rim specimens having a higher density.

344001 344002 344004 34.4005 344006 34.4007 344008 34.4009 34.4010 34.4019

“Rim” specimens:

SUm Mean a

34.4011 34.4012 34.4013 34.4114 34.40 17 34.4018 344024 34.4025 34.4026 34.4046 344047 344048 344049 34.4050

“Plaine” specimens

Number

:

98.4 94.7 89.4 97.9 86.4 85.9 85.0 96.8 73.6 87.6

303.7 -

1907 2014 2084 1825 2288 2253 2271 2521 1978 2364

88.0 87.4 11.8

-

15.1 17.5 17.5 13.9 10.1 IO.6 12.5 12.7 12.6 9.9

96.6 58.4 93.4 83.4 90.5 91.9 94.6 97.7 85.0 97.1 63.7 93.1 87.3 90.7

1681 3828 1999 1680 2410 1679 2273 2060 2480 2106 2607 2276 3005 2677

% Kamacite

28.8 25.4 19-o 20.1 18.9 24.7 18.8 24.7 17.4 20.4 14.5 30.8 17.6 22.0

counted

No. pts.

11.2 9.8 8.8 0.4 16.1 9.0

0.4 7.6 -

-

4.8 5.2 6.0

10.7 1.1 8.7 4.3

11.1 --

16.2 1.1 9.5 4.3 3.8 2.0 -

-

% Cohenite _-

0.8 3.8 1.9 0.8 2.4 3.7 5.2 2.4 2.8 2.8

I.4 1.5 0.6

1-O 1-G 0.7 2.0 1.7 1.5 1.0 I.0 3.2 0.8 1.5 0.9 2.3 I*1

% Plessite

0.8 1.2 1.1 1.2 0.9 0.6 1.1 1.0 7.5 0.6

0.8 0.8 0.5

1.0 0.1 1.9 1.3 0.6 0.4 0.7 1.2 0.1 1.4 0.3 1.1 0.7 0.1

0.2 -_

0.1 -

1.2 1.6 4.4

0.5

0.3 17.1 1.3 _..-

2.8 -

-

-

2.6 2.6 2.0

1.5 7.5 2.9 3.9 2.9 2.5 I.7 0.1 0.4 0.4 6.3 2.6 1.1 3.1

% Troilite

_________-_.-.

% Oxide

specimens of the C&on

% Schreibersite

Table 1. Modal analysis of fourteen plains and thirty-two

-

1.1 0.9 3.5

0.X

13.4 -_ -_--

-

..-__

% Graphite

7.87 7.80 7.85

7.87 7.72 7.84 7.89 7.86 7.88 -

7.81 0.05

-

7.76 7.79 7.79 7.87 7.85 7.80 7.75 -

7.79 7.85 7.76 -

7.86 -

Density

Diablo meteorite

HeWJ Heaty Heavy He&q Heavy Heavy Heavy Moderate Moderate HeFWv

-~-

Light Light, Light Moderato Moderate Moderate Light Moderate Moderltt c Light Light Light Light’ 1,ight

Shock effect

5

313.5 -

9.2 13.2 14.2 4.1 9.7 7-O 6.8 6.2 6.8 5.7 4.8 8.06 7.2 6.1 5*6 Il.6 8.6 5‘9 9.7 8-B 14.1

7.6

81.5 81.0 SO.9 86.6 87.6 80.2 83.1 96.8 80.6 87.2 97-4 86.3 81.4 89.9 84.6 96.2 92.2 66.4 92-7 30.6 86.8 83.4

87.0 86.7 7-o

2961 2668 3012 2644 2289 2062 2276 1909 1967 2969 1906 1330 1849 2023 2768 2025 2577 2026 1522 2026 2941 2073

-

* Contained diamonds.

&lm Me8n

344020 34.4021 34.4022 34.4023 34-4027 34.4028 34.4029 34.4030 344031 34.4032 34.4033 34.4034 34.4036 344036 34.4037 34.4038 34.4040 34.4041 344042 34.4043 34*4044* 34*404ti* 8.1 8.3 5.5

12.3 13.9 14.1 9.7 6.3 14.9 12.5 14.7 8.0 10.2 14.3 3.0 11‘7 0.9 4.7 10.2 I.2 16.0 10.7 11.9 3.3 2.9 1.3

5.8 4.6 4.0 3.2 1.1 2.8 2.0 2.5 4.7 1.7 0.8 2-l 4.2 5.9 3.1 0.5 2.4 3.8 1.4 2.8 2.4 3.2 I.1 1.0 1.3

0.03 0.6 0.9 0*5 o-4 2.0 2.4 o-7 0.1 0.3 2.6 o-2 0.1 0.6 2.8 0.7 1-I 0.04 0.1 1.3 0.3 0.4 0.9

2; 3.6 0‘04 -

0.4 0.2 2.8 2.1 1.2 0.5

0.6 O"7 3.3

4.7 lEb3 0.5 0.2 -

7.85 0.05

7.91 7.82 7.79 7.87 -

7.88 7.89 7.83 -

-

H-W Modexate Moderate HtWy

H-V He=Y HeWY Heavy Moderate Moderate

He=Y HMV H-7 H-V H-V Moderhe

[email protected]=Y Modfx&te Heavy

HfJaV$l [email protected]=Y Moderate

1890

4’ _/. B. NOORE, I’. J. BIRRELL and C. k‘. LEWIS Table Sprc. “Pleins”

I?. Chemical No.

Diablo

c

Fe

0.008 0.148 o-074 0.166 0.124 0.017 0.013
90.6 91.0 91.5 90.1 91.4 91.3 90.0 90.1 85.2 89.8 90.4 89.8 87.4 89.2

7-O 7.20 6.37 7.16 7.03 6.82 7.08 7.56 7.23 7.14 5.19 7.09 7.41 7.16

0.38 Cl.49 u.44 o-42 0.47 0.33 0.35 0.33 0.41 0.43 0.39 0.38 0.37

0.109 0.16

89.7 2.0

7.10 o-33

0.40 0.0.5

0.017 0.013 0.16 0.012 0.474 0.078 0.368 0.008 0.185 0.134 0.119 0.345 0.657 0.203 0.075 0.900 0.918 co.01 0.573 0.703 0.041 0.285 0.194 0.039 1.48 0.021 0.395 1.33 0.039 0.239 0.406 0.264

92.5 91.4 91.4 91.4 87.3 92-6 91.0 91.3 91.3 90.8 90.2 88.2 88.9 86.4 88.2 88.2 87.5 89.3 87.8 86.0 88.6 91.3 90.0 88.9 87.1 90-l 88.9 89-4 90.8 90.5 87.9 88-2

7.19 8.0 7.21 7.13 6.83 7.88 6.97 7.76 7.51 7.15 7.21 7.85 7.57 7.00 8.23 7.25 9.13 7.70 7.45 7.08 7.07 7.23 7.23 7.14 7.00 7.03 7.49

7.33 7.42 7.37

o-37 O-50 O-26 0.44 0.41 O-52 0.41 0.45 0.29 0.35 0.34 0.38 0.38 o-39 o-30 0.35 0.38 0.36 0.42 0.41 0.39 0.41 0.33 0.33 0.44 0.37 0.37 0.34 1.35 o-41 0.48 0.48

0.334 0.37

89.5 1.7

7.40 0.45

0.39 0.06

ivi -

---

C’O

specimens:

34.4011.1 34.4012.1 34.4013.3 34.4014.3 34.4017.1 34.4018.1 34.4024.1 34.4025.1 34.4026.1 34.4046.1 34.4047.3 34.4048.3 34.4049.3 34.4050.1 Mean fJ “Rim”

rtnalysis of met,al phase of C&on met,eorit#es

0.37

specimens:

34.4001.1 34.4002.1 34.4004.1 34.4005.1 34.4006.1 34.4007.1 34.4008.1 34.4009.1 34.4010.1 34.4019.1 34.4020.3 34.4021.2 34.4022.1 34.4023.1 34.4027.2 34.40282 34.4029.2 34.4030.1 34.4031.1 34.4032.2 34.4033.1 34.4034.3 34.4035.1 34.4036.1 34.4037.3 34.4038.2 34.4040.3 34.4041.3 34.4042.2 34.4043.1 34.4044.1 34.4045.1 Mean a

i.27 6.98

Variations in the chemical and minerelogical composition of the C&ion Diablo meteorite

1891

CONCLUSIONS

The Cafion Diablo iron meteorite appears to be significantly inhomogeneous with respect to its mineralogical and ohemical composition between the rim and plains speoimens seleoted for this investigation. The rim specimens are significantly higher in the high nickel phases taenite plus plessite. This is paralleled by a significantly higher nickel content in the metal phase of the rim specimens. The carbon content of the metal phase and the density are also significantly high for the rim specimens. In their papers, HEY~NN (1964), HEYMANN et aZ. (1966), LIPSCWTZ (1965) and WASSON (1967) have all noted that there are variations in the Cafion Diablo meteorite specimens but have presented strong evidence based upon rare gas, shock, and trace element studies to show that the meteorites found in the vicinity of Meteor Crater are the products of a single fall. Whether the inhomogeneity is of a transitional nature or possesses sharp boundaries is difficult to assess. Even though the rim specimens are quite numerous and common they come from a rather limited area primarily on the north-east crater rim and may represent fragments from a small area within the parent body. The Canyon Diablo (1936) and (1949) specimens which WASSON has shown to have large nickel contents of 7.93 and 8.20% have also been found on the crater rim. Specimens with nickel contents transitional to these values have been observed in the rim specimens analyzed in this study. The rim specimens are different from the plains specimens in their chemical and mineralogical composition as well as in their degree of shocking and diamond-bearing a5ities. Whether the compositional variations have any control on the formation of diamonds is not apparent at this time but should be taken into consideration when strong geographical (i.e. rim location) arguments are used in discussing the origin of the diamonds. The derivation of a weighted average for the bulk composition as WASSON (1967) has done is rather difficult based upon the observed variations, the small sample that survived the impact, and the poor overall sampling of the survivors. The fact that. the majority of collected specimens are from the plains may give them a weak bias as representative of the original mass. On this basis the most suitable analysis for the main mass of the Cafion Diablo meteorite would then be: Fe-89.7 wt.%, Ni7*10%, Co-0.40% and a density of 7.81. Ackrcowledgnx&+-This study was partially supported by NASA Grant 399. Pm JOST acknowledges the support of the NSF-UBP grants GE-120 and GE-4067. The advice of H. H. NININOERin selecting specimens, and H. B. Wrnr is gratefully acknowledged. The kindness and support of the Barringer Crater Company in pertially supporting one of the author’s (C. B. M) trip to the IUPAC in Moscow and H. B. WIIK’S visit to Arizona State University is sincerely appreciated. REFERENCES GoL~B~~R~E. A., UCEIY~~~LLA. and BROWN H. (1951) The distribution of Ni, Co, Ge, Pd and Au in iron meteorites. Geochim. Comnochim. Acta 2, l-25. HED. (1984) Origin of the Canyon Diablo No. 2 and No. 3 meteorites. Nature 294, 819-820. HED., Lmscnu~z M. E., NIELSEN B. and ANDERS E. (1966) Canyon Diablo meteorite: metallographic and mass spectrometric study of 50 fragments. J. Qeophy8. Ree. 71,61Q-640.

1892

C. B. MOORE, P. J. BIRRELL and C. F. LEWIS

L~PSOHTJTZ M. E. (1965) Origin of atypical meteorites from the arizona Meteorite Crater. Katuro Z&$636-638. LIPSCHUTZM. E. and ANDERS E. (1961a) The record in the meteorites--IV. Origin of diamonds in iron meteorites. Geochim. Comochim. Acta 24, 83-105. LIPSOHUTZM. E. and ANDERS E. (196lb) On the mechanism of diamond formation. Science 134, 2095-2099. NININQER H. H. (1950) Structure and composition of Canyon Diablo meteorites as related to zonal distribution of fragments. Popular Astron. 58, 169-173. NININQER H. H. (1956) Arizona’s Meteorite Crater. Sedona, Arizona. PERRY H. H. (1944) The metallography of meteorite iron. Bull. U.S. Nat. Mu.Y. No. 184. WASSON J. T. (1967) Concentrations of Ni, Ga and Ge in a series of Canyon Diablo and Odessa meteorite specimens. J. Geophya. Res. 72, 721-730.