Characterization of Goethite and Hematite in a Tunisian Soil Profile by Mossbauer Spectroscopy

-As part of the characterization of a Tunisian red soil profile, six samples, taken at different depths, were investigated by MSssbauer spectroscopy at room temperature and at 80 K to obtain information about the various types of Fe oxides present. By considering magnetic hypertine field distributions, the spectra of goethite and hematite were well resolved. Chemical analyses of the samples revealed a partial substitution of Fe by A1 and Mn. The spectral behavior of the goethite was predominantly influenced by crystaUinity and amount of A1 substitution which resulted in a reduction of the magnetic hyperfme field. The effect of Mn substitution was much more pronounced in the hematite spectrum as a consequence of a stronger suppression of the Morin transition by Mn than by AI. Key Words--Aluminum, Goethite, Hematite, Iron, Manganese, M6ssbauer spectroscopy, Soil. I N T R O D U C T I O N MSssbauer spectroscopy is a useful tool for the characterization of iron minerals in soils and clays, and, when used in conjunction with chemical and X-ray powder diffraction analyses, can provide a quanti tat ive estimate of the iron content of various species present in an unknown sample (see, e.g., Bancroft, 1973). One source of error in such analyses is that the MSssbauer recoilless fraction of goethite (a-FeOOH) may differ from specimen to specimen and seems to be indirectly related to the water content of the mineral (Chambaere et al., 1984). An even larger error may result from an incorrect interpretation of the spectrum. For example, many iron species in soils, goethite in particular, are poorly crystalline, and effects due to small crystallite size lead to an asymmetric distr ibution of the magnetic hyperfine field and reduce its average value (Morup et al., 1983). Al-substi tution also introduces line broadening and a reduction of the average magnetic hyperfine field (Golden et al., 1979; Goodman and Lewis, 1981; Fysh and Clark, 1982; Murad and Schwertmann, 1983). Although goethite and hemati te have different average hyperfine fields, the area under the sextets of both phases cannot be accurately determined from an analysis of only a few sextets having Lorentzian shape. The outer steep flanks of the asymmetric goethite peaks are usually not satisfactorily described, resulting in an underest imation of the hemati te content of the sample. Only for samples containing poorly crystalline goethite or goethite with substantial A1 substitution, resulting in a goethite doublet at room temperature, is the hemati te sextet well resolved. For all other samples, meaCopyright 9 1986, The Clay Minerals Society surements must be made of low-temperature spectra (4 K) in which asymmetric line broadening due to smallcrystallite effects are largely reduced. Unfortunately, most applied MSssbauer spectroscopy is performed at room temperature or at about 80 K using liquid nitrogen. At 80 K the asymmetry of the goethite peaks is perceptible and increases with increasing temperature. At room temperature the broadening may be so pronounced that a conventional fit with a l imited number of Lorentzians is not practicable. To obtain valid resuits from such spectra the distr ibutions of the hyperfine parameters must be considered. Murad (1982) showed that the spectra ofgoethi te can be satisfactorily described by a distr ibution of magnetic hyperfine fields, assuming that the isomer shift and quadrupole splitting are not affected by the crystalline imperfections. In the present paper, a Mrssbauer study of Tunisian soil samples is presented wherein it is shown that even from room-temperature spectra containing broad and asymmetric peaks, reasonable qualitative and quantitative results can be obtained i f the spectra are analyzed in terms of hyperfine field distributions. The meri t of such an approach has also been demonstra ted for synthetic Al-substi tuted goethite-hematite mixtures by Amarasir iwardena et al. (1986). MATERIALS A N D METHODS The samples were collected at different depths in a red soil in northwestern Tunisia (near Tamera, about 100 km west of Tunis). The area is presently covered with a degraded cork oak forest. The parent material (sample 198) is a red hemati t ic iron core, covered with a yellow crust that consists mainly ofgoethi te and clay. The macroscopic characteristics of the samples are: