**2.1 Fluorescence characteristics of olive oil and its components**

Conventionally, two basic types of spectra characterize the fluorescent properties: excitation and emission spectra. For a system containing a single fluorophore, the shape and location of the excitation and emission spectra are independent of respective chosen emission and excitation wavelengths. However, for a system containing several fluorescent components, the excitation and emission spectra depend on particular emission and excitation wavelength used for measurements. Therefore, in systems containing several fluorophores, single-wavelength spectra are insufficient for a comprehensive description of fluorescent properties, thus multidimensional measurement methods should be used.

The most comprehensive characterization of a multicomponent fluorescent system is obtained by measurement of an excitation-emission matrix, known also as a total luminescence spectrum or fluorescence landscape. This technique was first introduced by Weber (1961). After the first application to edible oils by Wolfbeis & Leiner (1984), it has been intensively used for exploring oil fluorescence. Total luminescence spectra are usually obtained by measurement of emission spectra at several excitation wavelengths. They may be presented as a three dimensional plot, with the fluorescence intensity plotted in function of the excitation and the emission wavelengths (Ndou and Warner, 1991; Guilbault, 1999). Another representation of the total luminescence is obtained using two-dimensional contour maps, in which one axis represents the emission and another – the excitation wavelength, and the contours are plotted by linking points of equal fluorescence intensity, Fig.1. The total luminescence spectrum gives a comprehensive description of the fluorescent components of the mixture and may serve as a unique fingerprint for identification and characterization of the sample studied. The acquisition of contour maps at sufficient resolution (determined by the number of individual emission spectra recorded) on conventional spectrofluorometers is time-consuming, requiring a large number of scans for each sample (Guilbault, 1999 ).

Alternatively, multicomponent fluorescent systems may be investigated by the synchronous fluorescence techniques, proposed by Lloyd, (1971). This technique involves simultaneous

techniques, relying on recording of single emission or excitation spectra, are often insufficient if directly applied. In such cases, total luminescence or synchronous scanning fluorescence techniques are used, improving the analytic potential of the fluorescence measurements. With contributions from numerous analytes, the autofluorescence of olive oil exhibits numerous overlapping bands. Such complex spectra should be analyzed using

Analytical applications of fluorescence to olive oils include discrimination between the different quality grades, adulteration detection, authentication of virgin oils, quantification of fluorescent components, monitoring thermal and photo-oxidation and quality changes

In this chapter the application of fluorescence spectroscopy to qualitative and quantitative analysis of olive oils is reviewed. Methodological aspects of fluorescence measurements and

Conventionally, two basic types of spectra characterize the fluorescent properties: excitation and emission spectra. For a system containing a single fluorophore, the shape and location of the excitation and emission spectra are independent of respective chosen emission and excitation wavelengths. However, for a system containing several fluorescent components, the excitation and emission spectra depend on particular emission and excitation wavelength used for measurements. Therefore, in systems containing several fluorophores, single-wavelength spectra are insufficient for a comprehensive description of fluorescent

The most comprehensive characterization of a multicomponent fluorescent system is obtained by measurement of an excitation-emission matrix, known also as a total luminescence spectrum or fluorescence landscape. This technique was first introduced by Weber (1961). After the first application to edible oils by Wolfbeis & Leiner (1984), it has been intensively used for exploring oil fluorescence. Total luminescence spectra are usually obtained by measurement of emission spectra at several excitation wavelengths. They may be presented as a three dimensional plot, with the fluorescence intensity plotted in function of the excitation and the emission wavelengths (Ndou and Warner, 1991; Guilbault, 1999). Another representation of the total luminescence is obtained using two-dimensional contour maps, in which one axis represents the emission and another – the excitation wavelength, and the contours are plotted by linking points of equal fluorescence intensity, Fig.1. The total luminescence spectrum gives a comprehensive description of the fluorescent components of the mixture and may serve as a unique fingerprint for identification and characterization of the sample studied. The acquisition of contour maps at sufficient resolution (determined by the number of individual emission spectra recorded) on conventional spectrofluorometers is time-consuming, requiring a large number of scans for

Alternatively, multicomponent fluorescent systems may be investigated by the synchronous fluorescence techniques, proposed by Lloyd, (1971). This technique involves simultaneous

multivariate and multiway methods.

**2. Fluorescence of olive oils** 

each sample (Guilbault, 1999 ).

analysis of fluorescence spectra are also discussed.

**2.1 Fluorescence characteristics of olive oil and its components** 

properties, thus multidimensional measurement methods should be used.

during storage.

scanning of both excitation and emission wavelengths, keeping a constant difference between them. Synchronous scanning fluorescence spectroscopy is very useful for the analysis of mixtures of fluorescent compounds, because both excitation and emission characteristics are included into a single spectrum. Although it provides less information than the excitation-emission matrix, it may still present a viable alternative to the total luminescence measurements due to its inherent simplicity and rapidity. A set of synchronous spectra recorded at different wavelength intervals may be concatenated into a total synchronous fluorescence spectrum. In such spectra fluorescence intensity is plotted as a function of the excitation wavelength and the wavelength interval. Both single wavelength interval and total synchronous fluorescence spectra were used for studies of olive oils (Sikorska et al. 2005a; Poulli et al. 2005). The relation between various kinds of fluorescence spectra of a virgin olive oil is presented in Fig. 1.

Fig. 1. Different types of fluorescence spectra; fluorescence spectra of a diluted sample of extra virgin olive oil (1%, v/v, in n-hexane) are shown as an example.

Numerous factors affect measured fluorescence intensity and spectral distribution. These factors are related to the nature and the concentration of fluorophores, their molecular environment, and scattering and absorption effects. They may be immeasurably important in complex natural systems, such as oils, and have to be taken into account when measuring and interpreting the fluorescence spectra. Fluorescence intensities are proportional to the concentration over only a limited range of optical densities (Lakowicz, 2006). To obtain proportionality between the fluorescence intensity and the fluorophore concentration, the absorbance at the excitation wavelength should be below 0.05 and close to zero in the emission spectral region. At higher concentrations, the inner filter effects have to be taken into account. These effects may decrease the observed fluorescence intensity by either

Analysis of Olive Oils by Fluorescence Spectroscopy: Methods and Applications 67

Typical fluorescence spectra of extra virgin and refined olive oils are shown in Fig. 3, (Sikorska et al., 2011). The fluorescence depends on sample concentration; therefore spectra for diluted and intact samples are shown. Both total fluorescence and total synchronous

Based on the published data, one may conclude that the fluorescence properties depend on the quality grade of olive oils (Kyriakidis&Skarkalis, 2000; Poulli et al., 2006; Guimet et al., 2004a). For a selected quality category, the spectra may show minor differences between samples, however, the general features remain similar, permitting identification and

The total fluorescence spectrum of diluted extra virgin olive oils, measured with the use of right angle geometry, exhibits two intense bands, one with excitation at about 270–330 nm and emission at about 295–360 nm and the second with excitation at about 330–440 nm and emission at about 660–700 nm, Fig. 3. An additional band appears in spectra of refined olive oil, located in the intermediate range, with excitation at 280-330 nm and emission at 372-480 nm. The long-wavelength band has a lower intensity in refined as compared to virgin olive

The spectra of the same undiluted oils measured with the front face geometry method show a clearly different fluorescence pattern. The spectra are not affected by the inner filter effect, because front face geometry was used for measurement of undiluted samples. Additional bands are observed in the spectra of extra virgin olive oil at about 310–390 nm in excitation and 440–580 nm in emission. The ratio of fluorescence intensity of short- and longwavelength bands is lower as compared to the spectrum of the diluted sample. The spectrum of undiluted refined olive oil exhibits a broad band with emission at 350-600 nm, two maxima at 320/420 nm and 365/450 nm in excitation/emission, and a long wavelength emission at 650-700 nm. Only a trace of the short-wavelength emission is observed with the

The differences in the spectra between diluted and undiluted samples may result from the high fluorophore concentrations in the intact oil samples and a variety of molecular interactions, such as quenching and energy transfer, which alter fluorescence characteristics. The effect of concentration on the total and synchronous fluorescence spectra of vegetable

The total synchronous fluorescence spectra of undiluted oils showed dependence of spectral shape and intensity on the wavelength interval () used in the measurements, with the presence of particular bands dependent on . At lower values of the bandwidths are reduced and the spectrum is simplified as compared to the total fluorescence spectra. Appearance of new bands or splitting of existing bands is typically observed with increasing . Emission bands are present in the excitation region below 310 nm, 310-350 nm, 350- 380 nm, and above 550 nm in spectra of virgin olive oils (Sikorska et al., 2011). Similar spectral characteristics for virgin olive oil were reported by Poulli et al., (2006). The bands in total synchronous fluorescence spectra were observed in the 270–325, 347–365 and 602–685 nm excitation wavelength ranges with the respective wavelength intervals of 20–120, 30–50

Refined oils are characterized by a relatively weak band between 290-320 nm, a very broad band spreading to about 500 nm, and a band above 550 nm. All of these bands equally

spectra are presented for the same oils, to enable comparison.

authentication of oil samples.

oil (Sikorska et al., 2011).

and 20–76 nm.

maximum at 300/331 nm (Sikorska et al., 2011).

oils including olive oils was reported by Sikorska et al. (2004; 2005b).

appear in the total fluorescence spectra (Sikorska et al., 2011).

reducing the intensity of the excitation or by absorbing the emitted radiation. To avoid the inner filter effects due to the high optical densities or sample turbidity, appropriate geometry of sample illumination should be used. The most common geometry using rightangle observation of the center of a centrally illuminated sample is only appropriate for diluted solutions with low optical densities. For opaque samples, front-face illumination is achieved using either triangular or square cuvettes oriented at 30 or 60 relative to the incident beam.

The effects of concentration and sample geometry on oil spectra were addressed by several authors. Zandomeneghi et al. (2005) in the very detailed studies compared the emission fluorescence spectra of undiluted extra virgin olive oil obtained with the traditional setup (right-angle fluorescence) and using front face fluorescence. The absorption of undiluted olive oil samples was from 2 up to 12 absorbance units, on passing from 325 to 260 nm, and therefore the inner filter phenomena affected the right angle spectra considerably. Excluding the long-wavelength chlorophyll region, significant differences in the number, shape, intensity, and position of the bands in spectra of the same undiluted oil obtained with right angle and front face geometry were observed, Fig. 2. The right angle fluorescence spectra showed considerable distortions, even after the mathematical corrections for the inner filter effects due to the absorption of both the excitation and emission radiation were applied. The front-face fluorescence spectra were less affected by self-absorption and thus provided reliable information about type of fluoropores and their concentration. It was also demonstrated that analysis of spectra affected by inner filter effects may lead to spectral misinterpretation and invalid assignments of origin of some fluorescent bands (M. Zandomeneghi & G. Zandomeneghi, 2005; Zandomeneghi et al., 2006).

Fig. 2. A. Emission spectra of olive oil with *λex* = 320 nm. (a) front face spectrum; (b) right angle spectrum with the absorption correction; (c) absorption spectrum of the oil multiplied by 106; (d) right angle spectrum without the absorption correction. B. Emission spectra of olive oil with *λex* = 280 nm. (a) front face spectrum; (b) right angle spectrum with the absorption correction and multiplied by 3.7; (c) front face spectrum, second-order corrected; (d) right angle spectrum. Reprinted with permission from Zandomeneghi et al. 2005. Copyright 2005 American Chemical Society.

reducing the intensity of the excitation or by absorbing the emitted radiation. To avoid the inner filter effects due to the high optical densities or sample turbidity, appropriate geometry of sample illumination should be used. The most common geometry using rightangle observation of the center of a centrally illuminated sample is only appropriate for diluted solutions with low optical densities. For opaque samples, front-face illumination is achieved using either triangular or square cuvettes oriented at 30 or 60 relative to the

The effects of concentration and sample geometry on oil spectra were addressed by several authors. Zandomeneghi et al. (2005) in the very detailed studies compared the emission fluorescence spectra of undiluted extra virgin olive oil obtained with the traditional setup (right-angle fluorescence) and using front face fluorescence. The absorption of undiluted olive oil samples was from 2 up to 12 absorbance units, on passing from 325 to 260 nm, and therefore the inner filter phenomena affected the right angle spectra considerably. Excluding the long-wavelength chlorophyll region, significant differences in the number, shape, intensity, and position of the bands in spectra of the same undiluted oil obtained with right angle and front face geometry were observed, Fig. 2. The right angle fluorescence spectra showed considerable distortions, even after the mathematical corrections for the inner filter effects due to the absorption of both the excitation and emission radiation were applied. The front-face fluorescence spectra were less affected by self-absorption and thus provided reliable information about type of fluoropores and their concentration. It was also demonstrated that analysis of spectra affected by inner filter effects may lead to spectral misinterpretation and invalid assignments of origin of some fluorescent bands (M.

A B

Fig. 2. A. Emission spectra of olive oil with *λex* = 320 nm. (a) front face spectrum; (b) right angle spectrum with the absorption correction; (c) absorption spectrum of the oil multiplied by 106; (d) right angle spectrum without the absorption correction. B. Emission spectra of olive oil with *λex* = 280 nm. (a) front face spectrum; (b) right angle spectrum with the

absorption correction and multiplied by 3.7; (c) front face spectrum, second-order corrected; (d) right angle spectrum. Reprinted with permission from Zandomeneghi et al. 2005.

Copyright 2005 American Chemical Society.

Zandomeneghi & G. Zandomeneghi, 2005; Zandomeneghi et al., 2006).

incident beam.

Typical fluorescence spectra of extra virgin and refined olive oils are shown in Fig. 3, (Sikorska et al., 2011). The fluorescence depends on sample concentration; therefore spectra for diluted and intact samples are shown. Both total fluorescence and total synchronous spectra are presented for the same oils, to enable comparison.

Based on the published data, one may conclude that the fluorescence properties depend on the quality grade of olive oils (Kyriakidis&Skarkalis, 2000; Poulli et al., 2006; Guimet et al., 2004a). For a selected quality category, the spectra may show minor differences between samples, however, the general features remain similar, permitting identification and authentication of oil samples.

The total fluorescence spectrum of diluted extra virgin olive oils, measured with the use of right angle geometry, exhibits two intense bands, one with excitation at about 270–330 nm and emission at about 295–360 nm and the second with excitation at about 330–440 nm and emission at about 660–700 nm, Fig. 3. An additional band appears in spectra of refined olive oil, located in the intermediate range, with excitation at 280-330 nm and emission at 372-480 nm. The long-wavelength band has a lower intensity in refined as compared to virgin olive oil (Sikorska et al., 2011).

The spectra of the same undiluted oils measured with the front face geometry method show a clearly different fluorescence pattern. The spectra are not affected by the inner filter effect, because front face geometry was used for measurement of undiluted samples. Additional bands are observed in the spectra of extra virgin olive oil at about 310–390 nm in excitation and 440–580 nm in emission. The ratio of fluorescence intensity of short- and longwavelength bands is lower as compared to the spectrum of the diluted sample. The spectrum of undiluted refined olive oil exhibits a broad band with emission at 350-600 nm, two maxima at 320/420 nm and 365/450 nm in excitation/emission, and a long wavelength emission at 650-700 nm. Only a trace of the short-wavelength emission is observed with the maximum at 300/331 nm (Sikorska et al., 2011).

The differences in the spectra between diluted and undiluted samples may result from the high fluorophore concentrations in the intact oil samples and a variety of molecular interactions, such as quenching and energy transfer, which alter fluorescence characteristics. The effect of concentration on the total and synchronous fluorescence spectra of vegetable oils including olive oils was reported by Sikorska et al. (2004; 2005b).

The total synchronous fluorescence spectra of undiluted oils showed dependence of spectral shape and intensity on the wavelength interval () used in the measurements, with the presence of particular bands dependent on . At lower values of the bandwidths are reduced and the spectrum is simplified as compared to the total fluorescence spectra. Appearance of new bands or splitting of existing bands is typically observed with increasing . Emission bands are present in the excitation region below 310 nm, 310-350 nm, 350- 380 nm, and above 550 nm in spectra of virgin olive oils (Sikorska et al., 2011). Similar spectral characteristics for virgin olive oil were reported by Poulli et al., (2006). The bands in total synchronous fluorescence spectra were observed in the 270–325, 347–365 and 602–685 nm excitation wavelength ranges with the respective wavelength intervals of 20–120, 30–50 and 20–76 nm.

Refined oils are characterized by a relatively weak band between 290-320 nm, a very broad band spreading to about 500 nm, and a band above 550 nm. All of these bands equally appear in the total fluorescence spectra (Sikorska et al., 2011).

Analysis of Olive Oils by Fluorescence Spectroscopy: Methods and Applications 69

250

250

100

F

20

40

 [nm]

Fig. 3. Fluorescence spectra of extra virgin and refined olive oil: A. Total fluorescence spectrum of diluted olive oil, 1% in n-hexane, B. Total fluorescence spectrum of diluted refined olive oil, 1% in n-hexane, C. Total fluorescence spectrum of undiluted extra virgin olive oil, front face geometry, D. Total fluorescence spectrum of undiluted refined olive oil, front face geometry, E. Total synchronous fluorescence spectrum of undiluted extra virgin olive oil, front face geometry, F. Total synchronous fluorescence spectrum of undiluted

60

80

300

350

ex [nm]

400

450

500

D

300

350

ex [nm]

400

450

500

B

300 400 500 600 700

300 400 500 600 700

em [nm]

300 400 500 600 700

ex [nm]

em [nm]

300 400 500 600 700

300 400 500 600 700

em [nm]

300 400 500 600 700

refined olive oil, front face geometry (Sikorska et al., 2011).

ex [nm]

em [nm]

250

250

100

E

20

40

60

[nm]

80

300

350

ex [nm]

400

450

500

C

300

350

ex [nm]

400

450

500

A

The identification of origin of the particular emission bands relies mainly on comparison to the spectra of chemically pure fluorescent components. The fluorescence properties of compounds occurring in oils or suggested to contribute to their emission are listed in Table 1. The short wavelength band in total fluorescence spectra, which covers the region of 270– 330 nm in excitation and 295–360 nm in emission, corresponds to the band at 280–310 nm in the total synchronous fluorescence spectra and is assigned to tocopherols and phenols. This assignment has been confirmed by several observations. Firstly, it was shown that a similar band appears in various vegetable oils, either cold-pressed or refined, and not only in olive oils (Sikorska et al., 2004). Olive oils contain considerable amounts of phenolic compounds, with their concentrations significantly reduced in refined oils. This observation seems to confirm that tocopherols also contribute to the emission observed in this wavelength range. In fact, tocopherols are present in most vegetable oils in widely variable amounts, from 70 to 1900 mg/kg (Cert et al., 2000). The vitamin E group includes four natural tocopherols (-, - , -, -) and four tocotrienols (T3, T3, T3, T3), all — in the R-configuration at the three double bonds in the side-chain of tocotrienols. Due to their structural similarity, all of these compounds exhibit very similar UV-absorption spectra and have similar fluorescence properties, Table 1. Of all tocopherols, -tocopherol is predominant in olive oils. Indeed, the band in olive oil spectra being discussed is similar to the one in the total luminescence spectrum of -tocopherol dissolved in *n*-hexane. Moreover, conventional excitation and emission spectra of the olive oils in the wavelength range mentioned are also similar to those of -tocopherol, and the excitation spectra are in good agreement with the absorption spectrum of -tocopherol in *n*-hexane (Sikorska et al. 2004). Still, the detailed analysis of excitation and emission spectra suggests contributions from several other fluorophores.

There still remain some inconsistencies concerning the assignment of vitamin E (tocopherol) bands in olive oil spectra. In one of the pioneering papers, where the emission spectra of various oils were reported, it has been suggested that the bands in the emission spectrum (ex=365 nm) with the maximum at 525 nm may partly originate from compounds of the vitamin E group, or their derivatives formed upon oxidation (Kyriakidis & Skarkalis, 2000). However, this interpretation is based on spectra of undiluted olive oils measured using right angle geometry, and therefore strongly affected by inner filter effects, and in some cases referring to the spectral region where no emission of tocopherols is present (Zandomeneghi et al. 2005). It should be underlined that the emission of vitamin E in *n*-hexane has its maximum at about 320 nm, with a similar maximum appearing in the spectra of oils. Moreover, it has been stated (Zandomeneghi et al. 2006) that the known products of oxidation of R--, -, -, -tocopherols, the R--, -, -, -tocopherolquinones, are all nonfluorescent substances (Pollok & Melchert, 2004).

Note that a considerable number of minor components belonging to different classes of phenolic compounds such as phenolic acids, phenolic alcohols, hydroxyisochromans, secoiridoids, lignans, and flavonoids are present in virgin olive oils (Servili et al., 2004). Most of polyphenols are fluorescent substances, absorbing in the 260-310 nm range and emitting in the near-UV range, with their bands centered at 310-370 nm (M. Zandomeneghi & G. Zandomeneghi, 2005). These phenolic compounds can be detected by fluorescence after separation by HPLC, using excitation/emission wavelengths of 264/354, 310/430 or 280/320 nm (Dupuy et al., 2005). Fluorescence typical for phenolic components of olive oils was reported recently by Tena et al., (2009), using excitation at 270 nm with the fluorescence maxima appearing in the 362-420 nm range, Table 1.

The identification of origin of the particular emission bands relies mainly on comparison to the spectra of chemically pure fluorescent components. The fluorescence properties of compounds occurring in oils or suggested to contribute to their emission are listed in Table 1. The short wavelength band in total fluorescence spectra, which covers the region of 270– 330 nm in excitation and 295–360 nm in emission, corresponds to the band at 280–310 nm in the total synchronous fluorescence spectra and is assigned to tocopherols and phenols. This assignment has been confirmed by several observations. Firstly, it was shown that a similar band appears in various vegetable oils, either cold-pressed or refined, and not only in olive oils (Sikorska et al., 2004). Olive oils contain considerable amounts of phenolic compounds, with their concentrations significantly reduced in refined oils. This observation seems to confirm that tocopherols also contribute to the emission observed in this wavelength range. In fact, tocopherols are present in most vegetable oils in widely variable amounts, from 70 to 1900 mg/kg (Cert et al., 2000). The vitamin E group includes four natural tocopherols (-, - , -, -) and four tocotrienols (T3, T3, T3, T3), all — in the R-configuration at the three double bonds in the side-chain of tocotrienols. Due to their structural similarity, all of these compounds exhibit very similar UV-absorption spectra and have similar fluorescence properties, Table 1. Of all tocopherols, -tocopherol is predominant in olive oils. Indeed, the band in olive oil spectra being discussed is similar to the one in the total luminescence spectrum of -tocopherol dissolved in *n*-hexane. Moreover, conventional excitation and emission spectra of the olive oils in the wavelength range mentioned are also similar to those of -tocopherol, and the excitation spectra are in good agreement with the absorption spectrum of -tocopherol in *n*-hexane (Sikorska et al. 2004). Still, the detailed analysis of excitation and emission spectra suggests contributions from several other fluorophores.

There still remain some inconsistencies concerning the assignment of vitamin E (tocopherol) bands in olive oil spectra. In one of the pioneering papers, where the emission spectra of various oils were reported, it has been suggested that the bands in the emission spectrum (ex=365 nm) with the maximum at 525 nm may partly originate from compounds of the vitamin E group, or their derivatives formed upon oxidation (Kyriakidis & Skarkalis, 2000). However, this interpretation is based on spectra of undiluted olive oils measured using right angle geometry, and therefore strongly affected by inner filter effects, and in some cases referring to the spectral region where no emission of tocopherols is present (Zandomeneghi et al. 2005). It should be underlined that the emission of vitamin E in *n*-hexane has its maximum at about 320 nm, with a similar maximum appearing in the spectra of oils. Moreover, it has been stated (Zandomeneghi et al. 2006) that the known products of oxidation of R--, -, -, -tocopherols, the R--, -, -, -tocopherolquinones, are all

Note that a considerable number of minor components belonging to different classes of phenolic compounds such as phenolic acids, phenolic alcohols, hydroxyisochromans, secoiridoids, lignans, and flavonoids are present in virgin olive oils (Servili et al., 2004). Most of polyphenols are fluorescent substances, absorbing in the 260-310 nm range and emitting in the near-UV range, with their bands centered at 310-370 nm (M. Zandomeneghi & G. Zandomeneghi, 2005). These phenolic compounds can be detected by fluorescence after separation by HPLC, using excitation/emission wavelengths of 264/354, 310/430 or 280/320 nm (Dupuy et al., 2005). Fluorescence typical for phenolic components of olive oils was reported recently by Tena et al., (2009), using excitation at 270 nm with the fluorescence

nonfluorescent substances (Pollok & Melchert, 2004).

maxima appearing in the 362-420 nm range, Table 1.

Fig. 3. Fluorescence spectra of extra virgin and refined olive oil: A. Total fluorescence spectrum of diluted olive oil, 1% in n-hexane, B. Total fluorescence spectrum of diluted refined olive oil, 1% in n-hexane, C. Total fluorescence spectrum of undiluted extra virgin olive oil, front face geometry, D. Total fluorescence spectrum of undiluted refined olive oil, front face geometry, E. Total synchronous fluorescence spectrum of undiluted extra virgin olive oil, front face geometry, F. Total synchronous fluorescence spectrum of undiluted refined olive oil, front face geometry (Sikorska et al., 2011).

Analysis of Olive Oils by Fluorescence Spectroscopy: Methods and Applications 71

The fluorescence spectra of a vitamin E standard in hexane and an oil polyphenol extract in methanol–water at the same concentration as found in the extra virgin olive oil were investigated separately (Cheikhousman et al., 2005). The maximums in the fluorescence excitation spectrum measured at the 330 nm emission wavelength of oil polyphenol extract and vitamin E were observed at 284 and 290 nm, respectively. The spectral contribution of both the tocopherols and phenolic compounds to the fluorescence of extra virgin olive oil was confirmed by the similarity between the reconstructed spectrum of the mixture and the spectrum of extra virgin oil (Cheikhousman et al., 2005). Recently the fluorescence intensity at 280/320 nm in excitation/emission was successfully used to determine phenol contents in

Thus, both tocopherols and phenolic compounds contribute to the short-wavelength emission of the olive oils, with the tocopherol contribution dominating in refined oils. The exact positions of the maxima of the short-wavelength emission vary slightly between various oils, which may result from differences in the respective tocopherol composition. Note that in oils obtained by physical methods (cold pressing), including olive, linseed and rapeseed oils, this fluorescence maximum was blue-shifted as compared to refined oils, pointing out the difference between fluorescence of refined and cold-pressed oils in this

Due to the similar fluorescence properties of tocopherols and some phenolic compounds (Table 1), their emission appears as a single broad band, therefore a detailed study of excitation and emission spectra in this region is required to reveal presence of various species (Sikorska et al., 2008a). The excitation and emission spectra of virgin olive oil measured respectively at em=330 nm and ex=295 nm agree very well with the respective spectra of -tocopherol, Fig. 4. On the other hand, the excitation and emission spectra measured respectively at em=300 nm and ex=280 nm are blue shifted as compared to tocopherol, being attributed to the phenolic compounds, according to Cheikhousman et al.

0

284 301

1x107

2x107

I

Fig. 4. (A) Excitation spectra (em=300 and 330 nm) and emission spectra (ex=280 and 295 nm) of extra virgin olive oil (green and blue) and tocopherol (red); the respective

spectrum ( = 10 nm) of extra virgin olive oil, (Sikorska et al., 2011).

excitation and emission wavelengths are shown in the picture; (B) synchronous fluorescence

F [a.u.]

3x107

4x107

B

300 400 500 600 700

[nm]

665

methanol/water extracts of olive oils (Papoti & Tsimidou, 2009).

spectral region (Sikorska et. al., 2004).

2.5x107 ex

 em 300 nm em 330 nm

278 295 310 318

250 275 300 325 350 375

[nm]

295 nm

ex 280 nm

(2005).

0.0

5.0x106

1.0x107

I

F [a.u.]

1.5x107

2.0x107

A


abs, exc, em – absorption, excitation and emission maxima, - molar absorption coefficient, a (Eitenmiller et al., 2008), b (Ward et al., 1994), c (Undenfriend, 1962), d(Diaz et al., 2003), e(Tena et al., 2009).

Table 1. Fluorescence properties of olive oil components.

 [dm3 mol-<sup>1</sup> cm-1]

> 94700 75000

131000 47100

101800 44500

145000 27800

119200 55200

Solvent λex

ether acetone 9:1 acetone/water

ether acetone 9:1 acetone/water

ether 9:1 acetone/water

ether 9:1 acetone/water -

> - -

282 - ethanol/n-hexane 270 310

[nm]

436 405 430e

436 406 e

436 435e -

> - -

668 669 669

648 652 653

673 671

661 658

> - -

λem [nm]

[nm]

**Vitamin E**

**Chlorophylls**

663

645

666

654

667

Vanillic acid e methanol 270 349 Syringic acid e 270 361 Gallic acid e 270 382 p-Coumaric acid e 416 o-Coumaric e 270 426 Cinnamic acid e 270 420 Tyrosol e 270 420 Caffeic acid e 270 457 abs, exc, em – absorption, excitation and emission maxima, - molar absorption coefficient, a (Eitenmiller et

**Phenolic compounds**

al., 2008), b (Ward et al., 1994), c (Undenfriend, 1962), d(Diaz et al., 2003), e(Tena et al., 2009).


Solvent λabs

Chlorophyll *a* b,c,d acetone 430

Chlorophyll *b* b,c,d acetone 455

Pheophytin *a* b,c,d acetone 409

Pheophytin *b* b,c,d acetone 434

Pheophorbide *a* b,c,d acetone 409

hexane

Table 1. Fluorescence properties of olive oil components.

Oleuropein ethanol/n-

The fluorescence spectra of a vitamin E standard in hexane and an oil polyphenol extract in methanol–water at the same concentration as found in the extra virgin olive oil were investigated separately (Cheikhousman et al., 2005). The maximums in the fluorescence excitation spectrum measured at the 330 nm emission wavelength of oil polyphenol extract and vitamin E were observed at 284 and 290 nm, respectively. The spectral contribution of both the tocopherols and phenolic compounds to the fluorescence of extra virgin olive oil was confirmed by the similarity between the reconstructed spectrum of the mixture and the spectrum of extra virgin oil (Cheikhousman et al., 2005). Recently the fluorescence intensity at 280/320 nm in excitation/emission was successfully used to determine phenol contents in methanol/water extracts of olive oils (Papoti & Tsimidou, 2009).

Thus, both tocopherols and phenolic compounds contribute to the short-wavelength emission of the olive oils, with the tocopherol contribution dominating in refined oils. The exact positions of the maxima of the short-wavelength emission vary slightly between various oils, which may result from differences in the respective tocopherol composition. Note that in oils obtained by physical methods (cold pressing), including olive, linseed and rapeseed oils, this fluorescence maximum was blue-shifted as compared to refined oils, pointing out the difference between fluorescence of refined and cold-pressed oils in this spectral region (Sikorska et. al., 2004).

Due to the similar fluorescence properties of tocopherols and some phenolic compounds (Table 1), their emission appears as a single broad band, therefore a detailed study of excitation and emission spectra in this region is required to reveal presence of various species (Sikorska et al., 2008a). The excitation and emission spectra of virgin olive oil measured respectively at em=330 nm and ex=295 nm agree very well with the respective spectra of -tocopherol, Fig. 4. On the other hand, the excitation and emission spectra measured respectively at em=300 nm and ex=280 nm are blue shifted as compared to tocopherol, being attributed to the phenolic compounds, according to Cheikhousman et al. (2005).

Fig. 4. (A) Excitation spectra (em=300 and 330 nm) and emission spectra (ex=280 and 295 nm) of extra virgin olive oil (green and blue) and tocopherol (red); the respective excitation and emission wavelengths are shown in the picture; (B) synchronous fluorescence spectrum ( = 10 nm) of extra virgin olive oil, (Sikorska et al., 2011).

Analysis of Olive Oils by Fluorescence Spectroscopy: Methods and Applications 73

Recently, fluorescence from conjugated fatty acids including -eleostearic acid, *cis*-parinaric acid and 8,10,12,14,16-octadecapentaenoic acid, was reported for Borage oil (Smyk et al., 2009). The fluorescence spectrum of methyl ester of the *cis-*parinaric acid has the maximum at 416 nm. It was also suggested that the low intensity emission with the maximum at 524 nm may originate from riboflavin (vitamin B2) in virgin olive oils. Riboflavin is a polar, water soluble compound and no quantitative data regarding the presence of riboflavin in olive oils are available. This vitamin was found in olive pulp (Zandomeneghi et al., 2005). Additional emission in the intermediate region has been detected as a result of oxidation (Cheikhousman et al., 2005; Poulli et al., 2009a, 2009b; Tena et al., 2009; Sikorska et al., 2008).

In past decades improvements in both spectroscopic instruments and computers contributed to the extensive application of fluorescence spectroscopy in food analysis, including olive oils. Although even the simple conventional analysis of fluorescence spectra may produce valuable data, most of the successful applications rely on multivariate methods for extracting useful

According to Christensen et al. (2006), several conditions should be met in an ideal system for fluorescence measurements: 1) the concentration of the fluorophores must be sufficiently low, to be approximately linearly related to the fluorescence intensity, 2) signals from each of the fluorescent components must beindependent of each other, 3) the signal contribution from interfering species must be insignificant compared to the target fluorophore signal. Most of these conditions, however, are routinely violated in intact food samples. Therefore, due to the complex character of the spectra, they are rarely used for direct analysis, being rather used as spectral patterns or fingerprints of particular samples. The vast amount of spectral information

Multivariate and multiway methods are specifically suited for treatment of such complex spectral data. The multivariate analysis has several advantages over the univariate approach. It enables analysis of nonselective signals in the presence of spectral interferences, providing diagnostic tools for detection of the outliers. Their application to spectral data has provided important tools for food analysis, where they can be used for exploration,

Traditional multivariate analysis of fluorescence data is performed on a series of emission, excitation or synchronous spectra arranged into a matrix. It starts usually with data exploration that is aimed at discovering structures in the data set, clustering of objects and outlier detection. This analysis does not require any prior knowledge of the explored data, employing unsupervised pattern recognition methods, including principal component analysis (PCA). Other methods used to explore the food fluorescence data include hierarchical cluster analysis, non-negative matrix factorization, common components and specific weights analysis, and canonical correlation analysis (Sadecka & Tothova, 2007).

Three-way models are used for analysis of sets of fluorescence excitation-emission matrices, including parallel factor analysis (PARAFAC) and the Tucker model. The PARAFAC model decomposes the fluorescence data into a number of components. These components correspond to the distinct fluorophores present in the samples. The analysis provides relative concentrations of each of the fluorophores in the mixture, accompanied by the

contained in such fingerprints could be used in qualitative and quantitative analysis.

**2.2 Methods of analysis of fluorescence data** 

analytical information from the measured fluorescence signals.

classification and calibration purposes (Christensen et al., 2006).

Synchronous fluorescence spectroscopy offers a superior solution to the problem of the overlapping spectral bands, by reducing their spectral widths. The synchronous fluorescence spectrum of virgin olive oil measured at =10 nm shows two separate bands, with the band at 301 nm attributed to tocopherol. Linear correlation was found between the fluorescence intensity at 301 nm and the total tocopherol concentration as determined chromatographically, for a set of different diluted vegetable oils (1% v/v, in n-hexane) (Sikorska et. al., 2005b).

The band at 284 nm could originate from phenolic compounds such as phenolic aglycons, based on the molecules of tyrosol and hydroxytyrosol, derived from phenolic glycosides present in the olives. Oleuropein aglycons are present in high amounts in virgin olive oil. The synchronous spectrum (=10 nm) of pure oleuropein exhibits a maximum at about 289 nm and is shifted by about 5 nm as compared to the fluorescence band observed in the virgin olive oil. This shift may result from the solvent, as the phenolic compounds are poorly soluble in nonpolar solvents, the spectrum was recorded in the n-hexane – ethanol mixture. On the other hand, the emission observed in the oil may originate from oleuropein derivatives, with slightly different emission properties (Sikorska et al., 2008).

The band splitting, observed in synchronous spectra only at small values, seems to be characteristic for virgin olive oils and has been not observed for refined oils (Sikorska et. al., 2005a). Synchronous fluorescence spectra acquired for virgin olive and sunflower oils at =20 nm were reported to have one double band at around 275 and 297 nm, and one more band at 660 nm, in contrast to sunflower oil that shows an intense band at around 300 nm and a weak one at 325 nm (Poulli et al., 2006).

The synchronous fluorescence spectra of phenolic compounds (tyrosol, p-coumaric and caffeic acids) present in virgin olive oils were measured at =30 nm (Dupuy et al., 2005). The spectra present a fluorescence band in the 275-350 nm spectral range, with the maxima for tyrosol and tocopherol observed respectively at 276 and 295 nm. Interestingly, it was shown that the mathematically calculated emission spectrum from a mixture of tyrosol, pcoumaric and caffeic acids, and -tocopherol was very similar to the experimentally obtained spectrum of olive oil. Therefore, we once more conclude that the fluorescence emission between 275 and 400 nm has contributions from both tocopherols and phenolic compounds (Dupuy et al., 2005).

A long-wavelength band is observed in the olive oil spectra, with excitation at about 350– 420 nm and emission at about 660–700 nm, corresponding to the band above 550 nm in total synchronous fluorescence spectra. This band was attributed to pigments of chlorophyll group, based on its excitation and emission characteristics (Zandomeneghi et al., 2005; Diaz et al., 2003). This group includes chlorophylls *a* and *b*, and pheophytins *a* and *b*, derived from chlorophylls by loss of magnesium. The emission spectra of these four chlorophyll derivatives are very similar, with their maxima in the range of 653 to 671 nm in 9:1 acetone/water (Diaz et al., 2003). Their presence is characteristic for virgin olive oils, being reduced to trace amounts in refined oils.

The origin of the emission in the intermediate region for both virgin and refined olive oils is unclear. It seems that various components may be responsible for bands appearing in this region. Wolfbeis & Leiner (1984), suggested that in addition to tocopherols and chlorophylls, parinaric acid, a conjugated 18:4 (n-3) acid, may also contribute to vegetable oil emission. Recently, fluorescence from conjugated fatty acids including -eleostearic acid, *cis*-parinaric acid and 8,10,12,14,16-octadecapentaenoic acid, was reported for Borage oil (Smyk et al., 2009). The fluorescence spectrum of methyl ester of the *cis-*parinaric acid has the maximum at 416 nm. It was also suggested that the low intensity emission with the maximum at 524 nm may originate from riboflavin (vitamin B2) in virgin olive oils. Riboflavin is a polar, water soluble compound and no quantitative data regarding the presence of riboflavin in olive oils are available. This vitamin was found in olive pulp (Zandomeneghi et al., 2005). Additional emission in the intermediate region has been detected as a result of oxidation (Cheikhousman et al., 2005; Poulli et al., 2009a, 2009b; Tena et al., 2009; Sikorska et al., 2008).

#### **2.2 Methods of analysis of fluorescence data**

72 Olive Oil – Constituents, Quality, Health Properties and Bioconversions

Synchronous fluorescence spectroscopy offers a superior solution to the problem of the overlapping spectral bands, by reducing their spectral widths. The synchronous fluorescence spectrum of virgin olive oil measured at =10 nm shows two separate bands, with the band at 301 nm attributed to tocopherol. Linear correlation was found between the fluorescence intensity at 301 nm and the total tocopherol concentration as determined chromatographically, for a set of different diluted vegetable oils (1% v/v, in n-hexane)

The band at 284 nm could originate from phenolic compounds such as phenolic aglycons, based on the molecules of tyrosol and hydroxytyrosol, derived from phenolic glycosides present in the olives. Oleuropein aglycons are present in high amounts in virgin olive oil. The synchronous spectrum (=10 nm) of pure oleuropein exhibits a maximum at about 289 nm and is shifted by about 5 nm as compared to the fluorescence band observed in the virgin olive oil. This shift may result from the solvent, as the phenolic compounds are poorly soluble in nonpolar solvents, the spectrum was recorded in the n-hexane – ethanol mixture. On the other hand, the emission observed in the oil may originate from oleuropein

The band splitting, observed in synchronous spectra only at small values, seems to be characteristic for virgin olive oils and has been not observed for refined oils (Sikorska et. al., 2005a). Synchronous fluorescence spectra acquired for virgin olive and sunflower oils at =20 nm were reported to have one double band at around 275 and 297 nm, and one more band at 660 nm, in contrast to sunflower oil that shows an intense band at around 300 nm

The synchronous fluorescence spectra of phenolic compounds (tyrosol, p-coumaric and caffeic acids) present in virgin olive oils were measured at =30 nm (Dupuy et al., 2005). The spectra present a fluorescence band in the 275-350 nm spectral range, with the maxima for tyrosol and tocopherol observed respectively at 276 and 295 nm. Interestingly, it was shown that the mathematically calculated emission spectrum from a mixture of tyrosol, pcoumaric and caffeic acids, and -tocopherol was very similar to the experimentally obtained spectrum of olive oil. Therefore, we once more conclude that the fluorescence emission between 275 and 400 nm has contributions from both tocopherols and phenolic

A long-wavelength band is observed in the olive oil spectra, with excitation at about 350– 420 nm and emission at about 660–700 nm, corresponding to the band above 550 nm in total synchronous fluorescence spectra. This band was attributed to pigments of chlorophyll group, based on its excitation and emission characteristics (Zandomeneghi et al., 2005; Diaz et al., 2003). This group includes chlorophylls *a* and *b*, and pheophytins *a* and *b*, derived from chlorophylls by loss of magnesium. The emission spectra of these four chlorophyll derivatives are very similar, with their maxima in the range of 653 to 671 nm in 9:1 acetone/water (Diaz et al., 2003). Their presence is characteristic for virgin olive oils, being

The origin of the emission in the intermediate region for both virgin and refined olive oils is unclear. It seems that various components may be responsible for bands appearing in this region. Wolfbeis & Leiner (1984), suggested that in addition to tocopherols and chlorophylls, parinaric acid, a conjugated 18:4 (n-3) acid, may also contribute to vegetable oil emission.

derivatives, with slightly different emission properties (Sikorska et al., 2008).

and a weak one at 325 nm (Poulli et al., 2006).

compounds (Dupuy et al., 2005).

reduced to trace amounts in refined oils.

(Sikorska et. al., 2005b).

In past decades improvements in both spectroscopic instruments and computers contributed to the extensive application of fluorescence spectroscopy in food analysis, including olive oils. Although even the simple conventional analysis of fluorescence spectra may produce valuable data, most of the successful applications rely on multivariate methods for extracting useful analytical information from the measured fluorescence signals.

According to Christensen et al. (2006), several conditions should be met in an ideal system for fluorescence measurements: 1) the concentration of the fluorophores must be sufficiently low, to be approximately linearly related to the fluorescence intensity, 2) signals from each of the fluorescent components must beindependent of each other, 3) the signal contribution from interfering species must be insignificant compared to the target fluorophore signal. Most of these conditions, however, are routinely violated in intact food samples. Therefore, due to the complex character of the spectra, they are rarely used for direct analysis, being rather used as spectral patterns or fingerprints of particular samples. The vast amount of spectral information contained in such fingerprints could be used in qualitative and quantitative analysis.

Multivariate and multiway methods are specifically suited for treatment of such complex spectral data. The multivariate analysis has several advantages over the univariate approach. It enables analysis of nonselective signals in the presence of spectral interferences, providing diagnostic tools for detection of the outliers. Their application to spectral data has provided important tools for food analysis, where they can be used for exploration, classification and calibration purposes (Christensen et al., 2006).

Traditional multivariate analysis of fluorescence data is performed on a series of emission, excitation or synchronous spectra arranged into a matrix. It starts usually with data exploration that is aimed at discovering structures in the data set, clustering of objects and outlier detection. This analysis does not require any prior knowledge of the explored data, employing unsupervised pattern recognition methods, including principal component analysis (PCA). Other methods used to explore the food fluorescence data include hierarchical cluster analysis, non-negative matrix factorization, common components and specific weights analysis, and canonical correlation analysis (Sadecka & Tothova, 2007).

Three-way models are used for analysis of sets of fluorescence excitation-emission matrices, including parallel factor analysis (PARAFAC) and the Tucker model. The PARAFAC model decomposes the fluorescence data into a number of components. These components correspond to the distinct fluorophores present in the samples. The analysis provides relative concentrations of each of the fluorophores in the mixture, accompanied by the

Analysis of Olive Oils by Fluorescence Spectroscopy: Methods and Applications 75

refining methods that do not alter the initial glyceridic structure; pure olive oil (or simply olive oil) consists of a blend of virgin and refined olive oil. The potential of fluorescence to discriminate olive oils of different quality was the subject of several studies. Both total luminescence and synchronous fluorescence spectra combined with various chemometric

Scott et al. (2003) used total luminescence spectra of four different types of edible oils: extra virgin olive, non-virgin olive, sunflower and rapeseed oils. The spectra of undiluted oil samples were measured in the excitation range from 350 to 450 nm with 10 nm intervals and in the emission range from 400 to 720 nm with 5 nm interval. Three supervised neural network algorithms were used for sample classification: simplified fuzzy adaptive resonance theory mapping, traditional back propagation and radial basis function. The 100%

Guimet et al. (2004a) in a series of studies investigated possibility of application of total fluorescence spectra for discrimination between various quality grades of olive oils. The excitation-emission matrices of undiluted oils were measured using right angle geometry. The hierarchical agglomerative clustering method with the Euclidean distance as a similarity measure and the average linkage method were applied to discriminate between three classes of commercial Spanish olive oils (virgin olive oils, pure olive oils, and olivepomace oils). To optimize the sample grouping into clusters, different preprocessing methods and two spectral ranges were tested, which either included or not the fluorescence peak of chlorophylls. The oils were distinguished using the unfolded excitation–emission fluorescence matrices in the 300-400 nm excitation range and 400-600 nm emission range, thus excluding the chlorophyll band, Fig. 5. The large variations in the chlorophyll band intensity, even between samples of the same type, tend to deteriorate oil discrimination. The optimal preprocessing included normalization of the unfolded spectral excitation–emission fluorescence matrices, followed by column autoscaling. The comparison of the results obtained from the excitation–emission fluorescence matrices to those from a single emission (ex=345, 360, 390 nm) and excitation (em=410 nm) fluorescence spectrum analysis showed the advantage of the total fluorescence data, which result in a significantly better

Other studies used unfold PCA and PARAFAC to explore the excitation–emission fluorescence matrices of virgin and pure olive oils (Guimet, 2004b). The spectral ranges studied were ex=300–400 nm, em=400–695 nm and ex=300–400 nm, em=400–600 nm. The first range included chlorophylls, whose peak was much more intense than those of the other components. The second range did not include the chlorophyll peak, being limited to the fluorescence spectra of the oxidation products and vitamin E. The three-component PARAFAC model on the second range (chlorophylls excluded) was found to produce the most useful results. With this model, it was possible to distinguish well between the two groups of oils and to calculate the underlying fluorescent spectra of the three families of compounds. Both unfold PCA and PARAFAC applied to the excitation–emission matrices showed clear differences between fluorescence of the two main groups of olive oils (virgin and pure). Chlorophylls had a strong influence on the models because of their high fluorescence intensity and high variability. Differentiation between the two types of oils was better when the chlorophyll fluorescence region was excluded from the models. The oxidation products are the species that most contribute to the separation between the two

approaches were successfully used for this purpose.

correct classification was obtained using all methods.

discrimination.

respective excitation and emission loadings, which correspond to the respective excitation and emission spectra, facilitating identification of the fluorescent constituents. This approach is called mathematical chromatography, enabling qualitative and quantitative analysis of the individual mixture components (Bro, 2003; Christensen et al., 2006).

The analytical problem of the food quality assessment often involves assignment of a particular product to a specific category. To perform such classification, supervised pattern recognition methods are used. In these methods the information about the class membership of the samples in a certain category is used to derive classification rules, which are next applied to classify new samples into correct categories on the basis of patterns present in their measurements (Berrueta et al., 2007). A number of classification techniques were used for analysis of food fluorescence data in the supervised mode: linear discriminate analysis, factorial discriminate analysis, k - nearest neighbors, discriminate partial least squares regression (DPLS), soft independent modeling of class analogy, and artificial neural networks.

Multivariate calibration is the most important and successful combination of chemometrics with spectral data used in analytical chemistry. The calibration consists of building a relationship between a desired chemical, biological or physical property of a sample, and its spectrum. The advantage of such approach is the replacement of the wet chemical measurements of a concentration, which are usually slow and expensive, by the spectral measurements, which are nondestructive and fast, requiring little or no sample preparation and producing no waste chemicals. The multivariate regression methods most frequently used in fluorescence analysis are partial least-squares regression (PLS) and principal component regression (PCR). N-way partial least-squares regression (N-PLS) is used for calibration analysis of fluorescence excitation-emission matrices (Geladi, 2003).
