**4. Gallery**

processing include making several scans of the same area using different environmental

In all the above maps, the increase in information due to the statistics of a great number of spectra is implemented visually by plotting the appropriate classic spectral parameters. When the components of the sample are unknown, the reference spectra are not available, and the spectral sources are not well identified from spectrum to spectrum, or to extract very fine spectral modification of a single compound, mathematical treatment based on the statistical structure of the hyperspectral data can be implemented. Nowadays, numerous mathematical treatments exist and are tested by spectroscopists in various kinds of applications [9, 16, 17]. As an illustration, **Figure 10** shows the principal component analysis extraction of spectral sources of interest in the Raman mapping of a uranium dioxide ceramic and the corresponding

**Figure 10.** Principal component analysis of a Raman mapping of a uranium dioxide ceramic. At right, the spectral

sources with their percentages and at left, the corresponding maps [9].

parameters (different laser wavelengths, external stress, temperature, etc.).

**3.5. Statistical treatment**

174 Raman Spectroscopy and Applications

**Figure 11** shows various examples of Raman maps made on polished thin sections of various rocks. As described below, the mineralogical determination and the interpretation of minerals in their context are greatly facilitated by Raman mapping.

#### **4.1. Sedimentary samples**

The sample in **Figure 11a** is a sandstone with an oxide matrix coming from the Hettangien formation (203.6–199.6 My), Chéniers France (see Ref. [18] for more information). This formation was occasionally exploited for iron and manganese since the Iron Age and also contains barite mineralisation. The grains consist of quartz and plagioclase (**Figure 11a1**). The different phases of iron oxide crystallisation in the holes of the rock, where hematite is altered to goethite, are highlighted by Raman mapping (**Figure 11a2**). These minerals are difficult to distinguish in transmitted light by optical microscopy. A small grain of barite is also visible in the Raman map.

#### **4.2. Magmatic sample**

Magmatic rocks are in most cases relatively easy to study using optical microscopy (the crystals are large and euhedral). However, Raman mapping can reveal small variations in mineralogical phases or crystal orientation. **Figure 11b1** shows an optical view of a granite from Autun, Saône-et-Loire, France (see Ref. [19] for more information). This rock is composed of large grains of quartz and labradorite associated with biotite, phlogopite, plagioclase and accessory minerals (titanite, apatite, rutile and brookite; **Figure 11b2**).

#### **4.3. Volcanic samples**

The sample shown in **Figure 11c** is a basalt from El Teide volcano, Tenerife, Spain. Basaltic rocks containing glass and microlitic crystals tend to be rather opaque in thin section, and identification of the individual minerals using optical microscopy is difficult, as shown in **Figure 11c1**. On the other hand, Raman mapping reveals easily all the microlitic paragenesis (labradorite, augite, analcime, apatite, anatase and hematite; **Figure 11c2**). Analcime occurs in vesicles and is a secondary mineral in this basalt. **Figure 11c3** shows the crystalline orientation of the plagioclases phases with polysynthetic twinning obtained from the intensity ratio of the spectral peaks at 195 and 514 cm−1. With this type of sample, a statistic treatment of the image could provide information on the orientation of the minerals and, consequently, lava flow direction.

**Figure 11.** Image gallery made on polished thin sections of different samples. (a) Sandstone from the Chéniers mine, Sacierges-Saint-Martin, Indre, France, (a1) optical view in reflected light, (a2) Raman map. (b) Granite from Autun, Saône-et-Loire, France (b1) optical view of the rock, (b2) Raman map. (c) Basalt from El Teide volcano, Tenerife, Spain (c1) optical view in transmitted light, (c2) Raman map and (c3) detail of the orientation in a plagioclase crystal obtained from the intensity ratio of the spectral peaks at 195 and 514 cm−1. (d) Garnet-kyanite granulite xenoliths in gneiss from La ferme des Saugères, Allier, France (d1) optical view in transmitted light, (d2) Raman map. (e) Highly deformed micaschist from the Sierra Alhamilla, Alméria, Spain (e1) optical view of a pressure shadow around a garnet in transmitted light, (e2) Raman map. (f) Microfossil from the Draken formation, Svalbard (f1) optical view in transmitted light, (f2) Raman map. (g) "Calamine" (Zn ore rocks) sample from BeniTajite mine, Haut Atlas, Morocco (g1) optical view, (g2) Raman map and (g3) detail of two generations of cerussite.

#### **4.4. Metamorphic samples**

Metamorphic rocks are characterised by very different mineral paragenesis, which can be difficult to identify in optical microscopy but easily resolvable using Raman mapping.

The sample in **Figure 11d** is a garnet-kyanite granulite from the gneissic upper unit (USG), La Ferme des Saugères in the Sioule metamorphic series, Allier, France (see Refs. [20] and [21] for more information). This rock is composed of quartz and granoblastic orthoclase that are readily visible in the Raman map (**Figure 11d2** A particularity of this sample is that it presents a retromorphosed facies associated with the destabilisation of the garnet and kyanite to phlogopite. Identification of the Al2SiO5 polymorphs (andalousite, sillimanite and kyanite) by optical microscopy is difficult but not with Raman mapping. The information thus obtained could be helpful for establishing the conditions of formation of the rock (e.g. in a P-T-t diagram). **Figure 11d2** is a Raman map based on peak parameters that have been modified by changes in the crystal orientation of kyanite. Note that carbon is present because the sample was coated for analysis by scanning electron microscopy and electron microprobe.

The second metamorphic sample is a highly deformed micaschist from the Sierra Alhamilla, Alméria, Spain (see Ref. [22] for more information). **Figure 11e1** is an optical micrograph demonstrating the high degree of deformation of the rock, making identification of the individual phases difficult. Raman mapping, on the other hand, reveals that the rock contains chloritoïde relics as well as various accessory minerals such as rutile, anatase, apatite, graphite and hematite in a small alteration vein. A pressure shadow filled by clinochlore and a new generation of muscovite can be observed. With its ability to distinguish structural changes in individual minerals, Raman mapping also highlights the direction of deformation, as indicated by the quartz and muscovite crystals (**Figure 11e2**).

#### **4.5. Application for micropaleontology**

**Figure 11f** is a dolomitised conglomerate from the Draken Formation (−800 to −700 My), Svarlbard. This conglomerate includes cherty lenses rich in microbial mat/planktonic microfossil assemblages (**Figure 11f1**) [23]. Two colourless phases that are not distinguishable in optical microscopy are revealed by Raman mapping to be opal and hydroxyapatite (**Figure 11f2**) (see Ref. [10] for more information). Some small spots of pyrite are also observed.

#### **4.6. Application for metallogeny**

**Figure 11.** Image gallery made on polished thin sections of different samples. (a) Sandstone from the Chéniers mine, Sacierges-Saint-Martin, Indre, France, (a1) optical view in reflected light, (a2) Raman map. (b) Granite from Autun, Saône-et-Loire, France (b1) optical view of the rock, (b2) Raman map. (c) Basalt from El Teide volcano, Tenerife, Spain (c1) optical view in transmitted light, (c2) Raman map and (c3) detail of the orientation in a plagioclase crystal obtained from the intensity ratio of the spectral peaks at 195 and 514 cm−1. (d) Garnet-kyanite granulite xenoliths in gneiss from La ferme des Saugères, Allier, France (d1) optical view in transmitted light, (d2) Raman map. (e) Highly deformed micaschist from the Sierra Alhamilla, Alméria, Spain (e1) optical view of a pressure shadow around a garnet in transmitted light, (e2) Raman map. (f) Microfossil from the Draken formation, Svalbard (f1) optical view in transmitted light, (f2) Raman map. (g) "Calamine" (Zn ore rocks) sample from BeniTajite mine, Haut Atlas, Morocco (g1) optical view,

(g2) Raman map and (g3) detail of two generations of cerussite.

176 Raman Spectroscopy and Applications

Mineral identification in metallogeny is very important for understanding the mineralisation process. **Figure 11g1** shows calamine deposit from the BeniTajite mine, Haut Atlas, Morocco, developed in a karstic area (see Ref. [24] for more information). The Raman map in **Figure 11g2, 3** shows the presence of cerussite formed by the alteration of galena. The cerussite has a botryoidal banded texture and forms a crust shielding the galena from oxidation. Some cerussite recrystallised during evolution of the karst, resulting in two generations of the mineral, each characterised by very specific textures that are invisible to optical or electronic microscopy but are readily visible in Raman mapping. Other minerals, such as calcite, cosalite (a Pb-Bi-sulphur) and phosophohedyphane (a Pb-phosphate), can also be identified by Raman mapping.
