Dorian G. W. Smith

Reflectance Spectra of Mafic Silicate-Opaque Assemblages with Applications to Meteorite Spectra

Abstract The reflectance spectra of wustite and mixtures of mafic silicates plus carbon or magnetite can be used to interpret meteorite and asteroid spectra. Mafic silicate + magnetite spectra show many features characteristics of magnetite-bearing meteorites— an overall decline or constant reflectance and lown overall reflectance. Mafic silicate + amorphous carbon spectra show low overall reflectance and a red slope unlike that seen in CV and CO carbonaceous chondrite spectra, probably because the meteoritic carbon is in a more ordered form. The reflectance spectra of ureilites are largely consistent with an assemblage of mafic silicates and abundant carbon. Ordinary chondrite reflectance spectra cannot be reproduced by any of the laboratory mixture spectra. The reflectance spectrum of wustite is a reasonable match to the spectrum of ordinary chondrite metal, suggesting that most ordinary chondrite metal grains are probably coated with an optically thick layer of an oxide. Ordinary chondrite and mafic silicate reflectance spectra are consistently less red-sloped than S-class asteroid spectra. The various spectral criteria use to deco0nvolve mafic silicate spectra are also applicable to CV and CO carbonaceous chondrites, ureilites, and ordinary chondrites, because the opaque phases present in these meteorites are spectrally neutral.

Spectroscopy of the cation distribution in the schorlomite species of garnet

Abstract A homogeneous megacryst of schorlomite was investigated to determine the valence states of Fe and Ti and the crystallographic sites occupied by these elements. The chemical composition of the specimen was analyzed by electron microprobe, wet-chemical analysis, FTlR, and INAA. The results from X-ray absorption near-edge structure spectroscopy (XANES) are consistent with exclusively Ti4+ occupying the octahedral site only. The tetrahedral site is deficient in Si and the results of low-temperature 57Fe Mossbauer spectroscopy indicate that the remainder of the site is occupied by Fe3+ and substantial Fe2+. A spin-allowed intensified crystal-field transition of [4]Fe2+ is present in the near-infrared spectrum. The optical absorption spectrum is dominated by an intense band centered at 500 nm with a full width of 8000 cm-1 at half maximum peak height; this band is assigned to an Fe2+-Ti4+ intervalence charge transfer transition between ,[4]Fe2+ and [6]Ti. The cation site occupancies in this specimen of schorlomite can be expressed by the following formula: {Ca2.866Mg0.800Na0.038Mn0.019}Σ3.003 [Ti4+1.058Fe3+0.631 Al0.137Fe2+0.057 Mg0.055Zr0.039V3+0.014Mn0.013] Σ2.004 - (Si2.348Fe3+0.339Fe2+0.311 [4H]0.005)Σ3.003O12.

Reflectance spectra of glass-bearing mafic silicate mixtures and spectral deconvolution procedures

Abstract The reflectance spectra of lunar, and perhaps asteroidal and Mercurian, soil analogs can be used to ascertain the effectiveness of continuum-removal procedures designed to isolate specific mineral absorption features. Spectra differences between natural and laboratory-produced glasses preclude the direct application of the latter to the analysis of remote sensing data. Nevertheless, it appears that currently used straight line continuum-removal procedures yield generally reasonable reconstructions of mafic silicate absorption bands in terms of band center wavelength positions. Better results would be achieved by removing a more accurately derived glass continuum from observational spectra.

An electron microprobe study of some Alberta placer gold

The detailed characteristics of 435 placer gold grains from the North Saskatchewan and Athabasca rivers have been investigated using a combination of optical microscopy, energy dispersive electron microprobe analysis and SEM techniques. Most grains show high-fineness rims surrounding lower-fineness cores (giving an extreme range of 4.3%–46.1% Ag within an individual grain). Characteristics of the rims lend support to the hypothesis that they are the products ofin situ leaching of silver followed by transport which largely destroys a spongy texture produced by the leaching. There is no change in the average Ag content of grains of any given size range with distance downstream, which also indicates that all significant leaching takes place before transportation. However, differences are apparent in the average Ag content with grain size.

MinIdent: An application to the identification and classification of amphiboles

SummaryAmphibole data in the MinIdent database (Smith andLeibovitz, 1986) were initially entered using species names quoted in the original source. The database has been updated by reclassifying these early data using the program AMPHTAB supplied by N. M. S. Rock and by adding supplemental data from the more recent literature, with the species names again checked using AMPHTAB. Associated MinIdent mineral identification software was utilized to determine which minerals in the database most closely resemble a series of unknown specimens chemically, as expressed in the Chemical Matching Index, CM, a relative figure-of-merit. Chemical data fromMogessie and Tessadri (1982) and Hawthorne (1983) were used to check the agreement between MinIdent and AMPHTAB for the classification of 221 unknown amphiboles.With 450 amphibole analyses entered and compiled in MinIdent, the name assigned by AMPHTAB showed the highest value of CM in MinIdent for 127 of the 221 unknown amphiboles (57.5°/x) and the second highest value for another 32 (14.5%). A chemically adjacent amphibole field had the highest value of CM for 59 of the 221 unknowns (26.7%), where “chemically adjacent” refers to a change in one chemical parameter. The greatest discrepancy between the two programs occurred in the hornblendes, with an agreement of just 20%, although for 58% of the unknowns the species with the highest CM in MinIdent was in a chemical field adjacent to the species name assigned by AMPHTAB. In many cases the disagreement between MinIdent and AMPHTAB could be ascribed to a lack of data in MinIdent.A comparison of the two programs suggests that the assignment of a single name to an unknown amphibole by AMPHTAB with no direct indication of its reliability may be; misleading. Standard analytical errors are frequently sufficient to overlap the arbitrary boundaries between amphibole species fields. In such cases it may be preferable to use a program such as MinIdent which, rather than assigning an arbitrary amphibole name, presents a list of 20 amphiboles with the degree of similarity between them and the unknown amphibole indicated. MinIdent offers the additional benefit of allowing input of other than chemical data and bases the match between unknown and standard data upon all input data. This will become more of an advantage as instruments such as automated refractometers become available for routine use.ZusammenfassungAusgangspunkt war das Amphibol-Datenmaterial (Smith und Leibovitz, 1986) mit den dort verwendeten Artnamen. Diese Basisdaten wurden vervollständigt und erneuert durch Reklassifizierung mittels des AMPHTAB Programms, ergänzt durch N. M. S. Rock, und durch Hinzufügung weiterer Daten aus der neuesten Literatur, deren Speciesnamen wiederum mit AMPHTAB überprüft wurden. Außerdem wurde eine MinIdent Mineralidentifizierungs-Software verwendet, um die Minerale zu bestimmen, die in ihrem Chemismus am ehesten einer Serie von unbekannten Amphibol-Species entsprechen, wie sie im Chemical Matching Index (CM) aufscheinen. Zur Klassifikation von 221 unbekannten Amphibolen wurden chemische Daten von Mogessie und Tessadri (1982) verwendet um die Übereinstimmung zwischen MinIdent und AMPHTAB zu überprüfen.Unter den 450 in MinIdent zusammengestellten und eingegebenen Amphibolanalysen zeigen die bei AMPHTAB angegebenen die höchsten CM Werte, nämlich 127 von 221 unbekannten Amphibolen (57,5%) und weitere 32 (14,5%) die zweithöchsten Werte. Innerhalb eines chemisch benachbarten Amphibolfeldes hatten 59 der 221 unbekannten Amphibole (26,7%) die höchsten CM Werte, wobei unter „achemisch benachbart” die Änderung eines chemischen Parameters zu verstehen ist. Die größten Unterschiede zwischen den beiden Programmen traten bei den Hornblenden auf. Die Übereinstimmung lag bei nur 20%, obwohl bei 58% der unbekannten Amphibole die Species mit dem höchsten CM Wert in MinIdent in ein chemisches Feld zu liegen kamen, welches zu den bei AMPHTAB angegebenen Speciesnamen eine benachbarte Position einnimmt. Die Unterschiede zwischen MinIdent und AMPHTAB könnten in vielen Fällen auf ein Fehlen von Daten in MinIdent zurükzuführen sein.Ein Vergleich beider Programme deutet an, daß die Angabe eines Einzelnamens für ein unbekanntes Amphibol im AMPHTAB Programm ohne Angaben über die Zuverlässigkeit zu Mißverständnissen führen kann. Normale analytische Fehler können bereits dazu führen, daß die Grenzen zweier willkürlicher Amphibolfelder überlappen. In derartigen Fällen emphiehlt sich die Anwendung des MinIdent Programmes, welches eben nicht einen willkürlichen Amphibolnamen angibt, sondern eine Liste von 20 Amphibolen mit dem Grad ihrer Ähnlichkeit, und einem Hinweis auf den unbekannten Amphibol. MinIdent bietet den zusätzlichen Vorteil, daß man außer chemischen auch andere Daten eingeben kann, und stellt dann sämtliche Daten des unbekannten Amphibols den Standard Daten gegenüber. Dieser Klassifizierungsvorgang wird mit der zunehmenden Routineanwendung von automatischen Refraktometern verstärkte Anwendung finden.

Minident — Some Recent Developments

Minldent is a command-driven program developed originally on a mainframe and now ported to a PC. Since its initial description, a substantial amount of new data and several new features have been added. In addition to information published on new mineral species, data have been included for about 700 presently unnamed minerals. A subset facility also has been added, as well as an extensive synonymy which provides abbreviated information and source references for some 1500 synonyms, varieties, discredited minerals and species of dubious authenticity. A full mineral classification scheme also has been included. Minldent-PC requires at least 560 kbytes of usable RAM, one 32 Mbyte HD drive and a math coprocessor.