Authors Copy - Semantic Scholar
Authors Copy
Proceedings of the Apollo 11 Lunar Science Conference. Vol. 1. pp. 499 to 512.
Investigation of lunar metal particles J. I. GOLDSTEIN Department of Metallurgy and Materials Science, Lehigh University, Bethlehem, Pennsylvania 18015 E. P. HENDERsoN U.S. National Museum, Washington, D.C. 20560 and H. YAK0wITz National Bureau of Standards, Washington, D.C. 20234 (Received 31 January 1970; accepted in revisedfarm 23 february 1970)
. .
Abstract—Several metallic particles from lunar fine samples 10084 and 10085-17M and from breccia sample 10046-18A were investigated by means of optical microscopy, scanning etectron microscopy and electron probe microanalysis. These particles consisted of two large gtobules, metal spread on glassy spheres, fragments and metal in the foamy vesicular-like fragments. The largest globule had a structure consisting of 40 tm Fe—Ni dendrites in a matrix of troilite. The dendrites showed typical Ni segregation, the composition at the outside being about 16 wt.% Ni with 13 wt.% Ni at the center. The troilite contained 01—15 wt.% Ni and appears to be in dis-equitibrium. A high Ni rim region was found at the troilite—dendrite interface. This region is taenite containing 29—46 wt.% Ni and about 03 wt.% S and was created by the rejection of Ni from the troilite. We were able to synthesize this globule; the same structure was obtained. From this simulation, the cooling rate was determined as 2.5CC/sec. This globule probably was created by the impact of a chondrite on the moon. A globule separated from the breccia rock apparently solidified quickly and then cooled slowly, the rock acting as a kind of crucible. This globule’s metal regions contain 2 wt% Ni, 1 wt.%P, 03 wt.% Co, balance Fe. There are also eutectic regions of phosphide with fine intergrowths of kamacite, troilite and carbides. This particle shows a striking similarity to spheroids from the Canyon Diablo iron. The evidence indicates that this globule was probably created by the impact of an iron meteorite on the moon. The analysis of the metal particles makes it clear that both meteoritic and lunar Fe—feS intergrowths are present in the lunar fines. The lunar metal is similar to the meteoritic material in that it is surrounded by sulfide; both metal constituents are of comparable structure and Co content. The two types of material can be differentiated by the significant Ni and P content of the transformed original meteoritic material.
INTRODUCTION
THERE are two types of metallic particles found in the lunar material. One type is found in the crystalline rocks and consists of a fine intergrowth of Fe, containing about 07 wt.% Co, and troilite. This intergrowth was probably derived from a homogeneous sulfide liquid which was immiscible in the magma that produced the igneous rocks (SKINNER, 1970). Other types of metallic particles often contain Fe, Ni and Co with varying amounts of P, S and C. These were probably derived from meteoritic materials (irons, chondrites, etc.) of different sizes which impacted with varying degrees of intensity on the lunar surface. The purpose of this paper is to describe several types of metallic particles which were separated from the lunar fines and breccia, to develop criteria for differentiating between metal particles of lunar 499
Authors Copy
1. I.
500
GOLDSTEIN,
I. P.
HENDERSON
and H. YAKowrrz
origin and of meteoritic origin, and to discuss some of the cratering and remetting processes occurring on the surface of the moon. DESCRIPTION AND ORIGIN OF METAL PARTICLES According to RAMDOHR and EL GoREs’s’ (1970) the lunar breccia rocks are made up of 5 main components: (1) globules of melted metal, melted troilite, or combina tions of both, (2) glassy gtobules, (3) fragments of meteorites, (4) foamy vesicular-like fragments with internal gas bubbles and (5) fragments of igneous rocks. In the lunar fines, the relative abundances of components 1—5 varies from sample to sample. We have examined metal particles in these 5 components of our samples; breccia 10046-l$A and soil samples 10084 and 10085-17M. Optical metatlography, scanning electron microscopy and electron probe microanalysis were used. Two globular particles were synthesized in the laboratory in an attempt to determine their cooling histories.
A. Metal globules Two relatively large metal globules were studied. The largest metallic particle studied was dubbed the “Mini-Moon” because its external features mimic the lunar image familiar to man. This particle was separated from lunar fines in sample 1008517M. The mini-moon resembles an asymmetric discus which cooled rapidly; it has dimensions 35 x 32 x 23 mm, weighs $8 mg and has a specific gravity of 736 as recovered. The mini-moon contains Ni—Fe, troilite, chromite and some siliceous material. The general aspect of the mini-moon is shown in Fig. 1. Siliceous material has adhered to the surface. Some of this is probably splattered lunar material. One side is smoother and appears similar to troilite. The opposite side had a matte finish. Cratering occurred on both surfaces probably as the sample cooled in the meteoritic atmosphere created by the impact of the parent meteorite. The craters can be roughly classified into three groups: (1) Saucer-like craters having shallow wide depressions. (2) Cup or bowl craters which are deeper with steeply dipping inside walls. (3) Flask-like craters with a bowl-like opening leading to a neck which connects with a spherical cavity within the mini-moon. Some of these features are shown in Fig. 2. In some craters, loose lunar fragments partly filled the opening. We do not know when these fragments entered the craters. In some saucer-type craters, lunar fragments the are firmly attached to the inside wall indicating that they were splashed into shattered. are underside the on depression. Often the sidewalls and bottoms of craters A closer examination of these fragments suggests some were preferentially shocked— even shock melted. Figure 3 shows a typical underside crater. Dendrites are standing free since the interdendritic material has flowed out. In the majority of the craters, calcium, aluminum, silicon and titanium as well as Fe and Ni have been identified. No silicate mineral was identified but several individual ilmenite pieces were found.
Authors Copy
Investigation of lunar metal particles
501
The lunar material on the surface of the mini-moon was probably obtained by impact processes and adhesion to the molten surface by lunar dust. The mini-moon was polished so as to reveal six consecutive sections. This distance between sections was unequal; the first section was made starting at the wrinkles at
Fig. 1. a (top). Overall view of mini-moon. Lunar material is adhering to surface especially the bottom. Note semi-polished top and craters. SEM Field is 43 x 34 mm. b (lower left). Optical view of top surface. Note wrinkles, craters and lunar material. Diameter of mini-moon is 35 mm. c (lower right). Optical view of bottom. Note matte finish and craters. Diameter of mini-moon is 35 mm.
the upper right of Fig. lb. Cross-sections were examined in the as-polished and in the etched condition (etchant-freshly prepared 1% nital). A cross-section near the center is shown in Fig. 4. The structure shows dendrites of Fe—Ni with an average spacing of 40 jum in a matrix of troilite. Note the presence of a thin light colored rim region
Authors Copy
J. I.
502
Gousmn’,
F. P.
HENDERsoN
and H.
.
YAK0wITz
—
_w_.
>/‘ --..
Fig. 2 (left). Two types of craters in the top surface. The paired craters are saucer-like. They contain fragments of lunar material. The other craters are bowl-like. Field is 600 m x 500 tim.
Fig. 3 (right). Underside crater. Note frag mented dendritic material left after the impact which created the crater. Field is 830 tim x 665 tim.
is typical of alt (5 4um) at the Fe—Ni and troilite interface (Fig. 4b). This structure (Fig. 4b). A phase troilite the in six cross-sections. Chromite inclusions were found troilite in the in found also were few localized inclusions which may be schreibersite one cross-section.
Approximate relative amounts of troilite as a function of depth are indicated in Table 1. This trend shows that the outside is richer in Fe—Ni than is the center. Since Fe—Ni solidifies first, the particle solidified from the outside in. Etched sections
)
I
•i ,
t”
I s •
r
---S-----
Fig. 4. a (left). Cross-sectional view of mini-moon. (Etched in 1% nital.) Troilite (gray), Fe—Ni dendrites and voids (black) are visible. Note cross-sections of craters. view. Top of micrograph is top of mini-moon. Width of field, 1 8 mm. b (right) Enlarged Note rim region at Fe—Ni and troilite interface. Chromite inclusions (trigonal gray is areas) are in the troilite. The Fe—Ni dendrites show deformation bands. Field 460 urn x 300 urn.
T
.0
.
Authors Copy
Investigation of lunar metal particles
•
503
show evidence of shock melting coupled with deformation due to shock as a result of impact (Fig. 4b). The electron probe microanalyzer was used to obtain quantitative compo sitional information. Data for Fe, Ni and S or for Fe, Ni and P were taken simultaneously. ,
,
----.‘.
•1
—
-:c: --
•
-
-
‘-c,
.
-
4. -‘
—
.-
-k
•)
-‘“
/41
V
-
‘
.‘
:..
4
* -
Fig. 5. a (left) Simulated mini-moon crosssection. Note similarity to Fig. 4a. Dendrite spacings are comparable. Width of field I 8 mm.
‘
\..
••
J
Fig. 5. b (right). Simulated mini-moon cross section. Note rim phase at Fe—Ni and troitite interface. Field is 460 m x 300 ttm.
Some 52 individual points were probed in order to establish compo sitions in the Fe—Ni, troilite and light colored rim region. Analytical runs for 50 of the 52 points totaled between 99 and 101 wt.°/. Raw X-ray data was reduced to concen tration by means of a computer program developed by HEN0c and HE1NRICH (1970) . The resuLts Table 1. Relative abundances of Ni—Fe and FeS in the mini-moon Order of section 1 2 3 4 5 6
Abundance of phases FeS Ni—Fe 25 23 32 32 39 38
75
77 68 68 61 60*
* On this section the relative abun dance of the Ni-rich rim region was counted and found to be between I and 2 per cent. Abundances optically deter mined with a grid ($00 counts per section).
may be summarized as follows: (I) troilite, 353 wt.% S, contains variabl a e amount of Ni (0l—15 wt.%, average 09 wt.%), 005 wt.% P, balance Fe. (2) Fe—Ni dendrites. The dendrites have an average composition of 15 wt.% Ni. The Ni varies from the center of the dendrite increasing towards the edge; a typical range is 13— 16 wt.% Ni. Phosphorus follows the Ni content being about 015 wt.% at the center,
Authors Copy
504
J. I. Gousri, E. P. HENDERSON
and
H. YAK0wITz
the dendrites 03O wt.% at the edge averaging 025 wt.%. Sulfur is constant within Ni is variable averaging about OO6 wt.%. (3) In the light colored rim region, the s the Ni follow ranging from 29 to 46 wt.% with an average of 35 wt.%. Phosphorus 01 to from g e content and averages 07 wt.%. The sulfur content is variabl rangin low toward biased 0.8 wt.% with an average of 03 wt.°/0. Perhaps the Ni content is the of ility possib values and the S content is biased toward high values due to the electron probe being partially in the troilite phase. solidification The rim region probably formed in the following manner. During cooling pro As . of the mini-moon, dendrites of taenite containing Fe—Ni formed g dendrite the growin t ceeded, the dendrites increased in size and the nickel conten of surround melt i—S Fe—N increased. At the same time, the Ni content of the remaining for the sible respon was nism ing the dendrite also increased. The foregoing mecha ses proces ion (Diffus tes. center-to-edge segregation of Ni in the mini-moon dendri ation segreg Ni e to-edg reduce but do not eliminate this segregation.) The center(1970) for iron—nickel observed was consistent with that predicted by FLEMINGS et a!. ec. 10°C/s 1 and n alloys cooled to room temperature at rates betwee of taenite. The tes dendri the After solidification troilite is found surrounding to CRAIG et a!. ing Accord Ni. troilite in the mini-moon contained up to 15 wt.% in equilibrium troilite t of conten (1968) and DOAN and GoLDsTEIN (1970b), the nickel is indicated fact this ce of Eviden with taenite decreases with decreasing temperature. ore after Theref . troilite itic meteor in by the low nickel concentrations (
Proceedings of the Apollo 11 Lunar Science Conference. Vol. 1. pp. 499 to 512.
Investigation of lunar metal particles J. I. GOLDSTEIN Department of Metallurgy and Materials Science, Lehigh University, Bethlehem, Pennsylvania 18015 E. P. HENDERsoN U.S. National Museum, Washington, D.C. 20560 and H. YAK0wITz National Bureau of Standards, Washington, D.C. 20234 (Received 31 January 1970; accepted in revisedfarm 23 february 1970)
. .
Abstract—Several metallic particles from lunar fine samples 10084 and 10085-17M and from breccia sample 10046-18A were investigated by means of optical microscopy, scanning etectron microscopy and electron probe microanalysis. These particles consisted of two large gtobules, metal spread on glassy spheres, fragments and metal in the foamy vesicular-like fragments. The largest globule had a structure consisting of 40 tm Fe—Ni dendrites in a matrix of troilite. The dendrites showed typical Ni segregation, the composition at the outside being about 16 wt.% Ni with 13 wt.% Ni at the center. The troilite contained 01—15 wt.% Ni and appears to be in dis-equitibrium. A high Ni rim region was found at the troilite—dendrite interface. This region is taenite containing 29—46 wt.% Ni and about 03 wt.% S and was created by the rejection of Ni from the troilite. We were able to synthesize this globule; the same structure was obtained. From this simulation, the cooling rate was determined as 2.5CC/sec. This globule probably was created by the impact of a chondrite on the moon. A globule separated from the breccia rock apparently solidified quickly and then cooled slowly, the rock acting as a kind of crucible. This globule’s metal regions contain 2 wt% Ni, 1 wt.%P, 03 wt.% Co, balance Fe. There are also eutectic regions of phosphide with fine intergrowths of kamacite, troilite and carbides. This particle shows a striking similarity to spheroids from the Canyon Diablo iron. The evidence indicates that this globule was probably created by the impact of an iron meteorite on the moon. The analysis of the metal particles makes it clear that both meteoritic and lunar Fe—feS intergrowths are present in the lunar fines. The lunar metal is similar to the meteoritic material in that it is surrounded by sulfide; both metal constituents are of comparable structure and Co content. The two types of material can be differentiated by the significant Ni and P content of the transformed original meteoritic material.
INTRODUCTION
THERE are two types of metallic particles found in the lunar material. One type is found in the crystalline rocks and consists of a fine intergrowth of Fe, containing about 07 wt.% Co, and troilite. This intergrowth was probably derived from a homogeneous sulfide liquid which was immiscible in the magma that produced the igneous rocks (SKINNER, 1970). Other types of metallic particles often contain Fe, Ni and Co with varying amounts of P, S and C. These were probably derived from meteoritic materials (irons, chondrites, etc.) of different sizes which impacted with varying degrees of intensity on the lunar surface. The purpose of this paper is to describe several types of metallic particles which were separated from the lunar fines and breccia, to develop criteria for differentiating between metal particles of lunar 499
Authors Copy
1. I.
500
GOLDSTEIN,
I. P.
HENDERSON
and H. YAKowrrz
origin and of meteoritic origin, and to discuss some of the cratering and remetting processes occurring on the surface of the moon. DESCRIPTION AND ORIGIN OF METAL PARTICLES According to RAMDOHR and EL GoREs’s’ (1970) the lunar breccia rocks are made up of 5 main components: (1) globules of melted metal, melted troilite, or combina tions of both, (2) glassy gtobules, (3) fragments of meteorites, (4) foamy vesicular-like fragments with internal gas bubbles and (5) fragments of igneous rocks. In the lunar fines, the relative abundances of components 1—5 varies from sample to sample. We have examined metal particles in these 5 components of our samples; breccia 10046-l$A and soil samples 10084 and 10085-17M. Optical metatlography, scanning electron microscopy and electron probe microanalysis were used. Two globular particles were synthesized in the laboratory in an attempt to determine their cooling histories.
A. Metal globules Two relatively large metal globules were studied. The largest metallic particle studied was dubbed the “Mini-Moon” because its external features mimic the lunar image familiar to man. This particle was separated from lunar fines in sample 1008517M. The mini-moon resembles an asymmetric discus which cooled rapidly; it has dimensions 35 x 32 x 23 mm, weighs $8 mg and has a specific gravity of 736 as recovered. The mini-moon contains Ni—Fe, troilite, chromite and some siliceous material. The general aspect of the mini-moon is shown in Fig. 1. Siliceous material has adhered to the surface. Some of this is probably splattered lunar material. One side is smoother and appears similar to troilite. The opposite side had a matte finish. Cratering occurred on both surfaces probably as the sample cooled in the meteoritic atmosphere created by the impact of the parent meteorite. The craters can be roughly classified into three groups: (1) Saucer-like craters having shallow wide depressions. (2) Cup or bowl craters which are deeper with steeply dipping inside walls. (3) Flask-like craters with a bowl-like opening leading to a neck which connects with a spherical cavity within the mini-moon. Some of these features are shown in Fig. 2. In some craters, loose lunar fragments partly filled the opening. We do not know when these fragments entered the craters. In some saucer-type craters, lunar fragments the are firmly attached to the inside wall indicating that they were splashed into shattered. are underside the on depression. Often the sidewalls and bottoms of craters A closer examination of these fragments suggests some were preferentially shocked— even shock melted. Figure 3 shows a typical underside crater. Dendrites are standing free since the interdendritic material has flowed out. In the majority of the craters, calcium, aluminum, silicon and titanium as well as Fe and Ni have been identified. No silicate mineral was identified but several individual ilmenite pieces were found.
Authors Copy
Investigation of lunar metal particles
501
The lunar material on the surface of the mini-moon was probably obtained by impact processes and adhesion to the molten surface by lunar dust. The mini-moon was polished so as to reveal six consecutive sections. This distance between sections was unequal; the first section was made starting at the wrinkles at
Fig. 1. a (top). Overall view of mini-moon. Lunar material is adhering to surface especially the bottom. Note semi-polished top and craters. SEM Field is 43 x 34 mm. b (lower left). Optical view of top surface. Note wrinkles, craters and lunar material. Diameter of mini-moon is 35 mm. c (lower right). Optical view of bottom. Note matte finish and craters. Diameter of mini-moon is 35 mm.
the upper right of Fig. lb. Cross-sections were examined in the as-polished and in the etched condition (etchant-freshly prepared 1% nital). A cross-section near the center is shown in Fig. 4. The structure shows dendrites of Fe—Ni with an average spacing of 40 jum in a matrix of troilite. Note the presence of a thin light colored rim region
Authors Copy
J. I.
502
Gousmn’,
F. P.
HENDERsoN
and H.
.
YAK0wITz
—
_w_.
>/‘ --..
Fig. 2 (left). Two types of craters in the top surface. The paired craters are saucer-like. They contain fragments of lunar material. The other craters are bowl-like. Field is 600 m x 500 tim.
Fig. 3 (right). Underside crater. Note frag mented dendritic material left after the impact which created the crater. Field is 830 tim x 665 tim.
is typical of alt (5 4um) at the Fe—Ni and troilite interface (Fig. 4b). This structure (Fig. 4b). A phase troilite the in six cross-sections. Chromite inclusions were found troilite in the in found also were few localized inclusions which may be schreibersite one cross-section.
Approximate relative amounts of troilite as a function of depth are indicated in Table 1. This trend shows that the outside is richer in Fe—Ni than is the center. Since Fe—Ni solidifies first, the particle solidified from the outside in. Etched sections
)
I
•i ,
t”
I s •
r
---S-----
Fig. 4. a (left). Cross-sectional view of mini-moon. (Etched in 1% nital.) Troilite (gray), Fe—Ni dendrites and voids (black) are visible. Note cross-sections of craters. view. Top of micrograph is top of mini-moon. Width of field, 1 8 mm. b (right) Enlarged Note rim region at Fe—Ni and troilite interface. Chromite inclusions (trigonal gray is areas) are in the troilite. The Fe—Ni dendrites show deformation bands. Field 460 urn x 300 urn.
T
.0
.
Authors Copy
Investigation of lunar metal particles
•
503
show evidence of shock melting coupled with deformation due to shock as a result of impact (Fig. 4b). The electron probe microanalyzer was used to obtain quantitative compo sitional information. Data for Fe, Ni and S or for Fe, Ni and P were taken simultaneously. ,
,
----.‘.
•1
—
-:c: --
•
-
-
‘-c,
.
-
4. -‘
—
.-
-k
•)
-‘“
/41
V
-
‘
.‘
:..
4
* -
Fig. 5. a (left) Simulated mini-moon crosssection. Note similarity to Fig. 4a. Dendrite spacings are comparable. Width of field I 8 mm.
‘
\..
••
J
Fig. 5. b (right). Simulated mini-moon cross section. Note rim phase at Fe—Ni and troitite interface. Field is 460 m x 300 ttm.
Some 52 individual points were probed in order to establish compo sitions in the Fe—Ni, troilite and light colored rim region. Analytical runs for 50 of the 52 points totaled between 99 and 101 wt.°/. Raw X-ray data was reduced to concen tration by means of a computer program developed by HEN0c and HE1NRICH (1970) . The resuLts Table 1. Relative abundances of Ni—Fe and FeS in the mini-moon Order of section 1 2 3 4 5 6
Abundance of phases FeS Ni—Fe 25 23 32 32 39 38
75
77 68 68 61 60*
* On this section the relative abun dance of the Ni-rich rim region was counted and found to be between I and 2 per cent. Abundances optically deter mined with a grid ($00 counts per section).
may be summarized as follows: (I) troilite, 353 wt.% S, contains variabl a e amount of Ni (0l—15 wt.%, average 09 wt.%), 005 wt.% P, balance Fe. (2) Fe—Ni dendrites. The dendrites have an average composition of 15 wt.% Ni. The Ni varies from the center of the dendrite increasing towards the edge; a typical range is 13— 16 wt.% Ni. Phosphorus follows the Ni content being about 015 wt.% at the center,
Authors Copy
504
J. I. Gousri, E. P. HENDERSON
and
H. YAK0wITz
the dendrites 03O wt.% at the edge averaging 025 wt.%. Sulfur is constant within Ni is variable averaging about OO6 wt.%. (3) In the light colored rim region, the s the Ni follow ranging from 29 to 46 wt.% with an average of 35 wt.%. Phosphorus 01 to from g e content and averages 07 wt.%. The sulfur content is variabl rangin low toward biased 0.8 wt.% with an average of 03 wt.°/0. Perhaps the Ni content is the of ility possib values and the S content is biased toward high values due to the electron probe being partially in the troilite phase. solidification The rim region probably formed in the following manner. During cooling pro As . of the mini-moon, dendrites of taenite containing Fe—Ni formed g dendrite the growin t ceeded, the dendrites increased in size and the nickel conten of surround melt i—S Fe—N increased. At the same time, the Ni content of the remaining for the sible respon was nism ing the dendrite also increased. The foregoing mecha ses proces ion (Diffus tes. center-to-edge segregation of Ni in the mini-moon dendri ation segreg Ni e to-edg reduce but do not eliminate this segregation.) The center(1970) for iron—nickel observed was consistent with that predicted by FLEMINGS et a!. ec. 10°C/s 1 and n alloys cooled to room temperature at rates betwee of taenite. The tes dendri the After solidification troilite is found surrounding to CRAIG et a!. ing Accord Ni. troilite in the mini-moon contained up to 15 wt.% in equilibrium troilite t of conten (1968) and DOAN and GoLDsTEIN (1970b), the nickel is indicated fact this ce of Eviden with taenite decreases with decreasing temperature. ore after Theref . troilite itic meteor in by the low nickel concentrations (
