**2.3 Collection of fluorescence spectra**

Taking into account that the emission spectra of organic compounds can provide information about them and about their concentration in mixed liquids, this spectroscopic technique appears as a complementary tool allowing the evaluation of the authenticity of matured alcoholic beverages (Gaigalas et al., 2001; Martínez et al, 2007; Navas & Jimenez, 1999; Walker, 1987). Fluorescence spectra were collected in the 540-800 nm spectral range, using a spectrofluorometer model USB4000-FL from the Ocean Optics company, coupled to a laser of 514 nm wavelength and cuvette holder by optical fibers. The spectral resolution was ~10 nm. The mezcal samples were put into 3.0 mL quartz cuvettes, in a 90 degrees configuration between the excitation source and the detector. This orientation remained constant during the collection of all the spectra. The laser power on the samples was 45 mW.

#### **2.3.1 PCA-scores**

Fig. 3 (a) and (b) depict the scores plots obtained from the mean centred spectra. According to Fig. 3 (a), PC1 explains 90 % of the variance. Two groups can be observed along PC1-axis, one of them including white mezcals and ethanol, and the other one, including rested, abocado and distilled mezcals. This indicates that data structure is mainly influenced by the presence or absence of certain organic molecules (not necessarily extracted from wood), all of them having similar emission features.

Fig. 3. PCA-scores plots obtained for the mean centred fluorescence spectra. Samples correspond to different stages of maturation. (a) Scores plot before the removal of outliers, (b) scores plot after the removal of outliers. (**□**) White or young, (△) abocado, (**○**) rested, (▽) aged, (◁) distilled and (◇) ethanol.

Three isolated objects, corresponding to rested, abocado and aged, can also be observed along PC1-axis. Among them, the first two can be considered as outliers. On the contrary, in 28 Principal Component Analysis

the spectral features belonging to authentic matured mezcals. On the other hand, onedimensional loadings plot obtained from standardized spectra [Fig. 2 (c)], lacks in the spectral information provided, thus limiting its uses for interpretation purposes. In spite of that, standardization may be useful for certain applications (*i.e.* calibration of

Taking into account that the emission spectra of organic compounds can provide information about them and about their concentration in mixed liquids, this spectroscopic technique appears as a complementary tool allowing the evaluation of the authenticity of matured alcoholic beverages (Gaigalas et al., 2001; Martínez et al, 2007; Navas & Jimenez, 1999; Walker, 1987). Fluorescence spectra were collected in the 540-800 nm spectral range, using a spectrofluorometer model USB4000-FL from the Ocean Optics company, coupled to a laser of 514 nm wavelength and cuvette holder by optical fibers. The spectral resolution was ~10 nm. The mezcal samples were put into 3.0 mL quartz cuvettes, in a 90 degrees configuration between the excitation source and the detector. This orientation remained constant during the collection of all the spectra. The laser power on the samples was 45 mW.

Fig. 3 (a) and (b) depict the scores plots obtained from the mean centred spectra. According to Fig. 3 (a), PC1 explains 90 % of the variance. Two groups can be observed along PC1-axis, one of them including white mezcals and ethanol, and the other one, including rested, abocado and distilled mezcals. This indicates that data structure is mainly influenced by the presence or absence of certain organic molecules (not necessarily extracted from wood), all

Fig. 3. PCA-scores plots obtained for the mean centred fluorescence spectra. Samples correspond to different stages of maturation. (a) Scores plot before the removal of outliers, (b) scores plot after the removal of outliers. (**□**) White or young, (△) abocado, (**○**) rested, (▽)

Three isolated objects, corresponding to rested, abocado and aged, can also be observed along PC1-axis. Among them, the first two can be considered as outliers. On the contrary, in

**-600 -400 -200 0 200 400 <sup>600</sup> b**

**PC2 (3%)**

**-4500 -3000 -1500 0 1500 3000 4500**

**PC1 (97%)**

prediction/classification by PLS or PLS-DA) (Esbensen, 2005).

**2.3 Collection of fluorescence spectra** 

of them having similar emission features.

aged, (◁) distilled and (◇) ethanol.

**-4500 -3000 -1500 0 1500 3000 4500**

**PC1 (90%)**

**3 2 1 outliers??**

**Outliers**

**2.3.1 PCA-scores** 

**-2100 -1400 -700 0 700 1400 2100 a**

**PC2 (7%)**

the case of the aged sample, the higher concentration of wood extracts in comparison to the rested samples originates a noticeably different spectrum, thus explaining the observation of this aged sample as an isolated object. In order to improve the distribution of objects, samples detected as outliers were removed. Fig. 3 (b) shows the scores plot after outliers removal. Similarly to Fig. 3(a), two main groups can be observed. However, the percentage of explained variance for PC1 increases to 97%. This fact indicates that, even when the object distribution does not depict significant variations, the removal of these outliers, allows the model to describe the data structure in a more efficient way.

There are other samples that can be considered as outliers [indicated with numbers 1-3 in Fig. 3(a)]. Among them, number 1 corresponds to white mezcal with worm (some mezcal producers add a worm to their product as a distinctive), the worm probably providing certain organic molecules. These organic molecules would have similar emission properties than those of the rested ones. Hence, number 1 can be taken as outlier or not. We decide to take it, as correctly classified. On the contrary, objects 2 and 3, were considered as outliers, because they do not have any particular characteristics like object 1. Furthermore, when they were removed from the PCA (data not shown), distribution of objects and explained variance percentages remained similar. For this reason, we decided not to remove them.

In conclusion, only the joint analysis of the scores plot, the raw spectra, the loading plots and all other available information of the samples can give a correct interpretation of our results.

Figure 4 (a) and (b) shows the PCA-scores plots obtained from standardized spectra before and after the removal of outliers. As it was described before, the similar object distribution observed in Fig. 3 and 4 indicates that standardization does not provide any additional benefit for the spectral data analysis.

Fig. 4. PCA-scores plots obtained for the mean centred and standardized fluorescence spectra. Samples correspond to different stages of maturation. (a) Scores plot before the removal of outliers, (b) scores plot after the removal of outliers. (**□**) White or young, (△) abocado, (**○**) rested, (▽) aged, (◁) distilled and (◇) ethanol.

Application of Principal Component Analysis

artificially matured mezcals.

al., 2009; Zolotov et al., 1987).

Jimenez Sandoval, 2000).

reduce the predictive ability of the model.

**3.1 Collection of spectral data set** 

**bacterial surfaces** 

Holan, 1995).

to Elucidate Experimental and Theoretical Information 31

necessarily correlated with maturation in the wood casks. Thus, fluorescence spectroscopy did not demonstrate to have enough sensibility to discriminate between authentic and

**3. PCA of Raman spectra for the determination of heavy metals attached to** 

Toxic metals are not degradable and tend to accumulate in the exposed organisms causing serious health effects. The use of biological agents to remove or neutralize contaminants (biorremediation) is a very important tool for the removal of such toxics. In particular, the use of inactivated microorganisms as adsorbents (biosorption) has been suggested as an effective and economical way to remove heavy metals from water and food intended to human or animal consumption (Davis et al., 2003; Haltunen et al., 2003, 2007, 2008; Ibrahim et al. 2006; Mehta & Gaur, 2005; Mrvčić et al., 2009; Shut et al., 2011; Volesky &

Metal biosorption is usually evaluated by means of analytical methods or atomic absorption spectrometry. These methods allow the quantification of free metal ions in the supernatants of bacterial/metal samples (Ernst et al., 2000; Haltunen et al., 2003, 2007, 2008; Velazquez et

In this sense, a method involving vibrational spectroscopic techniques (*i.e.:* Raman spectroscopy) and multivariate methods (both unsupervised and/or supervised), would represent an advantage over the standard procedures, due to the possibility of quantifying the metal ions directly from the bacterial sample, and at the same time, obtaining structural information (Araujo-Andrade et al., 2004, 2005, 2009; Ferraro et al., 2003; Gerbino et al. 2011;

PCA carried out on the Raman spectra represents the first step in the construction of a calibration model allowing the quantification of metal ions attached to bacterial surfaces. This analysis allows obtaining a correlation between the spectral variations and the property of interest (*i.e.* the metal ion concentration), identifying the optimal spectral region/s for the calibration of quantification models, and also detecting erroneous measurements leading to

In this section, we present the PCA results obtained from the Raman spectra corresponding to bacterial samples (*Lactobacillus kefir*) before and after the interaction with four heavy

Even when the main objective of this study was to calibrate models for the quantification of metal ions attached to bacterial surfaces using supervised methods (*i.e*.: PLS), the calibration of prediction models goes beyond of the intention of this chapter. For this reason, we

The Raman spectra of the bacterial samples before and after the interaction with metal ions were measured by placing them onto an aluminum substrate and then under a Leica microscope (DMLM) integrated to a Renishaw micro-Raman system model 1000B. In order

metals (Cd2+, Pb2+, Zn2+, Ni2+) in three different concentrations each.

focused this section just on the discussion of the PCA results.
