Page 272 - Microtectonics
P. 272

10.2  ·  Techniques to Study Deformation Fabrics  265
                 point defects. These may include ions of 3d elements (tran-  both types of analysis is monochromatic CL imaging. Here,
                                         2+
                                             3+
                                                    3+
                 sition group metals, such as Mn , Cr  and Fe ), ions of  distribution patterns are obtained separately for certain
                                                  3+
                 4f elements (rare earth elements, such as Dy ), other ions  wavelengths. Most studies apply panchromatic (or integral)
                        –
                 [e.g., (S )  in sodalite and haüyne] and vacancies and re-  imaging, which uses the entire detected CL signal independ-
                      2
                 lated defect centres in the crystal lattice (Richter and  ent of its wavelength composition. One advantage of OM-
                 Zinkernagel 1981; ten Have and Heynen 1985; Solomon  CL imaging is that colour variations within minerals are
                 1989; Barbin and Schvoerer 1997; Götze 2000; Pagel et al.  easily observed, which provides a rough indication of the
                 2000; Nasdala et al. 2004a).                  wavelength. A combination of trace element analysis and
                   A large number of mineral species has been studied us-  CL spectroscopy promises to become a powerful tool for
                 ing CL, including carbonate minerals (Spötl 1991; El Ali et al.  mapping of trace element distribution in microstructures
                 1993; Yang et al. 1995; Habermann et al. 1996, 1998; Barbin  (Yang et al. 1995; Barbin and Schvoerer 1997).
                 and Schvoerer 1997; Gillhaus et al. 2000), diamond
                 (Smirnova et al. 1999; Mineeva et al. 2000; Davies et al. 2003),  10.2.2
                 zircon (Ohnenstetter et al. 1991; Cesbron et al. 1993; Hanchar  Raman Spectroscopy
                 and Miller 1993; Hanchar and Rudnick 1995; Nasdala et al.
                 2003), monazite (Seydoux-Guillaume et al. 2002), apatite  Raman spectroscopy is based on the interaction of the (vi-
                 (Barbarand and Pagel 2001; Kempe and Götze 2002), quartz  brating) electric field vector of light with vibrations in the
                 and silica (Fig. 10.1; Matter and Ramseyer 1985; Shimamoto  sample (molecule, crystal lattice). Light can be absorbed by a
                 et al. 1991; Gorton et al. 1996; Götze et al. 2001), feldspar  sample upon the excitation of a vibration, provided that its
                 group minerals (Mora and Ramseyer 1992; Mariano and  photon energy corresponds to the phonon quantum energy
                 Roeder 1989; Finch and Klein 1999), and clay minerals  of an allowed vibration in the sample (vibrational quanta
                 (Götze et al. 2002). Cathodoluminescence is a well-estab-  are called phonons). This is the case for light in the infrared
                 lished technique in sedimentology to study mineral growth  range, and such absorption is used in infrared absorption
                 processes and provenance of sediments (Marshall 1988;  analysis. In contrast, the phonon energy of visible light is
                 Augustson and Bahlburg 2003) but is also useful in micro-  too high to be transferred to the sample through the excita-
                 tectonic studies. The intensity distribution pattern and col-  tion of a vibration. However, it is possible that a small frac-
                 our of CL may be independent of optically visible micro-  tion of the photon energy of visible light is taken to excite
                 structures and can reveal mineral growth from solutions,  the sample to vibrate. After such interaction, a scattered light
                 e.g. in microcracks (Narahara and Wiltschko 1986; Laubach  photon will be lowered in energy and shifted in wavelength
                 et al. 2004), in veins and along grain boundaries and details  towards the red end of the electromagnetic spectrum, a
                 of the microstructure of cataclasites (Stel 1981; Blenkinsop  process known as Stokes-type Raman scattering. Alterna-
                 and Rutter 1986). In veins, CL can reveal growth surfaces  tively, it is possible that the vibrational state of the sample
                 that are invisible in ordinary light (Dietrich and Grant 1985;  be lowered in which case the scattered light has gained en-
                 Urai et al. 1991). Fracture patterns can be easily distin-  ergy and is blue-shifted. This is called anti-Stokes Raman
                 guished (Kanaori 1986). Overgrowth structures along grain  scattering. In the case of Rayleigh scattering no energy shift
                 boundaries, which are invisible in normal transmitted light,  of the light results. Raman analyses are mostly done in the
                 can be used for strain analysis (Onasch and Davis 1988).  Stokes region because of the higher intensity of Stokes-type
                 Dislocation networks can be made visible in some cases  Raman scattering compared with anti-Stokes scattering.
                 (Grant and White 1978).                         Raman scattering is a weak effect. Only a small frac-
                   Cathodoluminescence analysis can be carried out using  tion of the incident light interacts with vibrations in the
                 either “hot cathode” or “cold cathode” systems attached to  sample as described above. The Raman shift is measured
                 an optical microscope (then called OM-CL) and in a SEM  as the wave number difference between incident and scat-
                 or electron microprobe (called SEM-CL; (Fig. 10.9a,  tered light. The wave number is defined as the reciprocal
                                                                                        –1
                 ×Photo 10.9; Yacobi and Holt 1990; Shimamoto et al. 1991;  wavelength and is given in cm  (McMillan and Hof-
                 Lloyd 1994; Götze 2000; Nasdala et al. 2003). In both cases  meister 1988; Marfunin 1995; Smith and Carabatos-
                 ordinary polished sections covered with a thin conductive  Nédelec 2001; Loudon 2001; Nasdala et al. 2004b).
                 layer (carbon or gold) are used. The CL technique involves  Raman spectroscopy is nowadays generally done with
                 the analysis of spectra or the generation of images. Catho-  a laser beam, which is directed at a sample using high-
                 doluminescence spectroscopy provides information about  resolution objective lenses. Powerful Raman microprobe
                 the spectral composition of the emitted light, which is  systems have a depth resolution as good as 2 µm when
                 needed for the sound interpretation of causal CL-active  the beam is focused at the sample surface and a real lat-
                 defects etc. Cathodoluminescence imaging, in contrast, is  eral resolution of 1–1.5 µm, which may result in a real
                                                                                            3
                 predominantly used to reveal internal structures of miner-  volume resolution better than 5 µm  (Markwort et al.
                 als, for which it may not be absolutely necessary to know  1995). The scattered light is analysed in a spectrometer.
                 the causes for local variations of the CL. A combination of  Apart from the analysis of single spectra, Raman is also
   267   268   269   270   271   272   273   274   275   276   277