PACROFI VI - Electronic Program

Possibilities and limits of infrared microscopy applied to studies of fluid inclusions in sulfides and other opaque minerals

Volker Lüders* & Christian Reutel**

*GeoForschungsZentrum Potsdam, PB 4.3 Lagerstättenbildung, Telegrafenberg A 50, D - 14473 Potsdam (Germany)

**Geolink Inc., Düstere-Eichen-Weg 1, D - 37073 Göttingen (Germany)

IR equipment
In 1984 Campbell et al. introduced infrared microscopy to the study of fluid inclusions (FIs). Previous studies of FIs in wolframite were performed by using a Research Devices Model F infrared microscope (e.g. Campbell et al., 1988). It allows observations in the near infrared light in the wavelength range from 0.8 to 1.2 microns ( = 800 - 1200 nm) with a maximum magnification of 400x. A detailed description of this IR microscope is given by Campbell (1991).

Other IR microscopes such as the Olympus BHSM-IR or Leica IR use an infrared sensitive TV camera to display the infrared image on a monitor. For FI studies long working distance IR objectives with magnifications of 5x, 10x, 20x, 50x and 80x are available. Hence, preference is given to the Olympus BHSM-IR, because the Leica IR microscope only allows the use of a 20x long working distance IR objective.

The transformation of the IR image to a monitor can either be done by an usual videocamera with removed IR-filter (e.g., Hitachi) or by special high-resolution IR cameras with extreme sensitive IR tubes (Electrophysics, Hamamatsu) which are about 10 times more expensive than usual video observation cameras. The latter have a resolution capacity for the near infrared in the wavelength range from 0.8 to 1.3 microns whereas high-resolution IR cameras are sensitive to longer wavelengths (lambda ~= 2.2 microns/Electrophysics, (lambda ~= 2.3 microns/Hamamatsu). The resolution of an usual high-resolution IR camera can be enhanced up to nearly 2.6 microns by a selected tube which is available on request (Hamamatsu).

The Olympus BHSM-IR microscope can be run with any commercial heating/freezing system. To be able to perform quick cyclic heating and/or freezing runs which is often necessary for watching phase transitions on a monitor screen, the U.S.G.S. heating/freezing system seems to be the most useful. For FI studies in the near infrared it is necessary to substitute all glass/silica windows of the metal chamber of the heating/freezing system by fluorite windows, since glass or silica windows have an absorbtion for near infrared light of about 50 %. For fluorite windows an IR transmittance of about 95% in the near infrared wavelength range 0.8 to 6.0 mm is given.

Transparency of opaque minerals
The quality of IR transmittance of several opaque minerals (oxides, sulfides, sulfosalts) was tested by FTIR spectroscopy (Ziemann & Lüders, this issue). It turned out that some of the investigated minerals such as arsenopyrite and some pyrites are opaque to near infrared light. For arsenopyrite this phenomenon can be explained by the very low band gap energy of the mineral (Campbell et al., 1984).

The IR transparency of sulfides and sulfosalts strongly depends on the chemical composition and the amounts of trace elements. Pyrites which contain elevated amounts of As (~= 0.5 wt.% As) are opaque in the near infrared up to l = 1.9 microns (Richards & Kerrich, 1993). On the other hand, pyrites with low As and/or trace element content can exhibit up to 40 % IR transmittance at 1.8 to 2.5 microns and are suitable for fluid inclusion studies.

A correlation between As content and IR transparency is obviously also given for fahlores. Whereas tetrahedrites show an excellent IR transmittance even at low IR wavelengths ( l ~= 0.9 microns), As-rich fahlores and real tennantites are opaque for the IR equipment.

Fig. 1 (left) Two-phase FI in pyrite: Quartz cupola, 330m level, Pechtelsgrun pegmatite, Saxony, Germany
Fig. 2 (right) Two-phase FI in tetrahedrite, Silverton, Colorado, USA

IR microthermometric FI studies require an extremely good polish of both surfaces of thick sections. The optimal thickness of the sections was found to lie between 90 and 130 microns. Kryometric investigations can be performed on any FI in an opaque mineral without any dramatically loss in contrast. By the use of a monitor with a big screen phase transitions in the low-temperature range can be observed more or less unproblematically.

Difficulties may appear when heating the mineral. Campbell et al. (1988) were not able to measure homogenization temperatures of FIs in wolframite from the Panasqueira deposit (Portugal) with the Research Devices Model F IR microscope. During heating the samples became opaque prior to homogenization of the FIs. FTIR spectra of wolframites indicate two maxima of transmittance of different degree, a lower one at about 1.0 mm and a higher one well above 1.8 microns (Ziemann & Lüders, this issue). During heating, the band gap energy decreases and, consequently, the IR adsorption increases. For wolframite it can be assumed that the lower transmittance maximum at 1.0 mm will disappear at lower temperatures than the higher maximum above 1.8 mm. To prove this assumption FIs in wolframite samples from the oxide-silicate stage of the Panasqueira deposit have been studied. It turned out that the observed phenomenon of decreasing IR-transparency with progressive heating can be compensated by boosting the sensivity of the IR camera, and measurements of the homogenization temperatures are possible. The data obtained indicate salinities of primary and pseudosecondary FIs between 7-19 wt.% NaCl equiv. and mean homogenization temperatures between 300 to 330oC. The data obtained clearly differ from those of inclusions in associated quartz ( mean salinity 8.5 wt.% NaCl equiv., mean Th 280oC) which is assumed to have deposited cogenetically (Kelly & Rye, 1979, Campbell et al., 1988).

The problem of decreasing IR-transparency during heating particulary appears in sulfides. Due to the varying quality of IR-transmittance, many pyrites become opaque upon heating mostly prior to homogenization of contained FIs. Until now, combined measurements of Tm ice (-3.3 to -2.2oC) and Th (211 - 256.5oC) were only obtained from FIs in pyrite octahedrons from the Murgul Cu deposit (NE Turkey). Pentagondodecahedrons from the same deposit became opaque at temperatures between 250 and 270oC before the homogenization temperature of two phase inclusions is reached. Homogenization temperatures of these inclusions were obtained by a cyclic technique similar to that described by Goldstein & Reynolds (1994). The precision of the obtained homogenization temperatures (298 to 324oC) lies in a range of ± 2.0oC.

The infrared technique is a powerful tool to obtain information on internal features and fluid inclusions in some opaque minerals. Especially in the field of the evolution of pegmatitic/ hydrothermal fluids and the formation of tin-tungsten deposits this method has been shown to be applicable. For some sulfides such as dark sphalerite, stibnite and As-poor sulfosalts FIs, if contained can be measured routinely with a high-resolution IR-equipment. The applicability of IR microthermometry to fluid inclusion studies in pyrite (chalcopyrite) is controlled by the quality of the specific IR-transmittance of the samples.