The analysis of fibrous amphiboles by FTIR spectroscopy in the OH-stretching region

The mechanism through which fibrous amphiboles may give rise to mesothelioma is still not completely understood. Several factors have been taken into account, and these include the morphological aspect (length:width ratio), the chemical composition, and a variety of surface properties which are ultimately responsible for the mineral-cellule interactions (e.g. van Oss et al., 1999). Studies in vitro demonstrate that the morphology of the fiber has a strong role in determining its biological danger, because very thin and long crystallites are hardly eliminated by the alveolar macrophages. However, a second factor which seems to influence the biological attack of a fiber is its Fe content; iron can in fact be involved in complex biochemical reactions with oxygen and cause DNA damages. It follows that the rapid determination of the chemical composition of a fiber, and of its Mg/Fe ratio in particular, is extremely important in environmental studies. The direct analysis can be achieved only by micro-chemical tools such as by EDS- or WDS-equipped electron microscopes. However, these techniques are extremely expensive and often unsuitable when dealing with extremely fibrous (diameter < 3 µm) materials. Therefore a rapid and easily accessible method is highly desirable in large scale environmental monitoring. The best alternative to EMPA is provided by FTIR spectroscopy, a technique which can be used on both single crystals (down to few µm in dimension) and powders. Here we present the results of a new calibration based on the analysis of a large set of well-characterized fibrous and prismatic natural amphiboles spanning a very large variety of chemical compositions and geological occurrences. All samples were previously studied using X-ray diffraction and EMPA. FTIR spectra in the principal OH-stretching region were collected on KBr disks prepared with a mineral:matrix = 5:150 mg ratio. Most spectra show four prominent bands which can be assigned to the combination of Mg and Fe²⁺ at the OH-coordinated M(1,3) sites (Della Ventura et al. 1996, 1997, Iezzi et al. 2005, 2007). The digitized spectra were fitted by interactive optimization followed by least-squares refinement; all bands were modelled as symmetric Gaussians. Della Ventura et al. (1996, 1997) showed that the binary site-occupancies at M(1) and M(3) in the amphibole structure are related to the observed intensities of the four (A to D) components in the principal IR OH-stretching spectrum. Using the original equations of Burns and Strens (1966): ^M(1,3)Mg = 3I_A + 2I_B + I_C and ^M(1,3)M²⁺ = I_B + 2I_C + 3I_D (with M²⁺ = Fe²⁺) where I_A-I_D are the intensities measured for the corresponding A to D bands, one can derive the (Mg, M²⁺) site populations at M(1,3) with a high degree of confidence. This method is particularly suitable for asbestiform materials which cannot be properly characterized by EMP. The present work shows that the above spectroscopic tool can be applied to a large variety of amphibole types. For species were significant (Mg, Fe) are present at M(4) (i.e. anthophyllite –cummingtonite - grunerite) an additional information (e.g. Mössbauer) is however required for a complete characterization of the sample. References Burns, R.G. and Strens, R.G.J. (1966) Science, 153, 890-892. Della Ventura, G., Robert, J.-L., Hawthorne, F.C. (1996) Geochimica and Cosmochimica Acta, vol. spec. 5, 55-63. Della Ventura, G., Robert, J.-L., Raudsepp, M., Hawthorne, F.C., Welch, M. (1997) American Mineralogist, 82, 291-301. Iezzi, G., Della Ventura, G., Hawthorne, F.C., Pedrazzi, G., Robert, J.-L., Novembre, D. (2005) European Journal of Mineralogy, 17, 733-740. Iezzi, G., Della Ventura, G., Bellatreccia, F., Lo Mastro, S., Gunther, M., Bandly, (2007) Mineralogical Magazine, in press. van Oss, C.J., Naim, J.O., Costanza, P.M., Giese, R.F. Jr., Wu, W., Sorling, A.F. (1999) Clays and Clay Minerals, 47, 697-707.