Analytica Chimica Acta 598 (2007) 128–134

Preliminary study on application of mid infrared spectroscopy for the

evaluation of the virgin olive oil “freshness”
Nicoletta Sinelli a , Maria Stella Cosio a , Carmen Gigliotti b , Ernestina Casiraghi a,∗
aDepartment of Food Science and Technologies, University of Milan, Via Celoria 2, 20133 Milan, Italy
b Department of Biomedical Sciences and Biotechnologies, University of Brescia, Viale Europa 11, 25123 Brescia, Italy
Received 23 March 2007; received in revised form 9 July 2007; accepted 11 July 2007
Available online 14 July 2007

Abstract
The freshness of virgin olive oils (VOO) from typical cultivars of Garda regions was evaluated by attenuated total reflectance (ATR) and
Fourier transform infrared (FTIR) spectroscopy, in combination with multivariate analysis. The olive oil freshness decreased during storage mainly
because of oxidation processes. In this research, 91 virgin olive oils were packaged in glass bottles and stored either in the light or in the dark
at room temperature for different periods. The oils were analysed, before and after storage, using both chemical methods and spectroscopic
technique.
Classification strategies investigated were partial least square discriminant analysis (PLS-DA), linear discriminant analysis (LDA), and soft
independent modelling of class analogy (SIMCA).
The results show that ATR-MIR spectroscopy is an interesting technique compared with traditional chemical index in classifying olive oil
samples stored in different conditions. In fact, the FTIR PCA results allowed a better discrimination among fresh and oxidized oils, than sam-
ples separation obtained by PCA applied to chemical data. Moreover, the results obtained by the different classification techniques (PLS-DA,
LDA, SIMCA) evidenced the ability of FTIR spectra to evaluate the olive oil freshness. FTIR spectroscopy results are in agreement with
classical methods. The spectroscopic technique could be applied for the prediction of VOOs freshness giving information related to chemical
modifications. The great advantages of this technique, compared to chemical analysis, are related to rapidity, non-destructive characteristics
and low cost per sample. In conclusion, ATR-MIR represents a reliable, cheap and fast classification tool able to assess the freshness of virgin
olive oils.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Virgin olive oil; Freshness; Oxidation; Storage; Fourier transform infrared (FTIR) spectroscopy; Attenuated total reflection mid infrared (ATR-MIR)
spectroscopy

1. Introduction
Extra virgin olive oil shows excellent sensorial and nutritional
properties, due to antioxidant compounds and to an interest-
ing unsaturated to saturated fatty acids ratio. These properties
decrease with time because of oxidation processes affected by
air, heat, light and metals [1]. Oil oxidation, is a very com-
plex phenomena that regards mainly unsaturated fatty acids
and modifies oil sensorial characteristics through off-flavours
development and secondary reaction products formation. Con-
sequently, it is matter of great concern for the olive oil industry
∗ Corresponding author. Tel.: +39 02 5031 9184; fax: +39 02 5031 9191.
E-mail address: ernestina.casiraghi@unimi.it (E. Casiraghi).
to preserve the positive attributes of extra virgin oil over time
from production to bottling, up to purchasing.
Nowadays, there are different methods used and/or pro-
posed for evaluating the oxidative deterioration of olive oil.
Among the routine methods, there are the peroxide value (PV),
which determines the amount of primary oxidation products,
UV absorbance at 232 and 270 nm, that measures the forma-
tion of conjugated dienes and trienes due to the formation of
secondary oxidation products and the Rancimat Oxidative Sta-
bility Instrument, determining the oxygen uptake [2]. According
to European Commission (EC) [3] and International Olive Oil
Council (IOOC) regulations [4], the extra virgin olive oil oxi-
dation level is assessed by the PV and spectrophotometric
absorbance, defining the following limits: PV ≤ 20 meq kg−1
and K270 ≤ 0.22. The EC legislation also considers the value

0003-2670/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2007.07.024
 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134
of K232 that must be ≤2.4. Nevertheless, these methods sup-
ply only limited information about the level of olive oil
oxidation.
In recent years, some other techniques have been applied
to study fatty acids oxidative processes. In particular, elec-
tron spin resonance spectroscopy (ESR) was applied to detect
free radicals produced during the oxidation process [5], high-
performance liquid chromatography (HPLC) with different
detection systems was used for hydroperoxide analysis [6] and
high-performance liquid chromatography/gas chromatography
with mass (HPLC/GC–MS) was applied for identifying triglyc-
eride oxidation products [7,8]. Moreover, in a recent study, the
triglyceride oligopolimers have been considered good indices of
olive oil oxidation level [9].
However, these last techniques are complex to perform or to
interpret, expensive and time consuming and generally highlight
only one or few aspects of the oxidation process, giving only
partial information about the extent of the process. On the other
hand, the olive oil industry needs to be able to know quickly
the level of oil oxidation in order to predict its remaining shelf
life. Moreover, consumers expect manufactures and retailers to
provide products of high quality and look for quality seals and
brands.
Therefore, the development of innovative analytical tools able
to execute fast and reliable quality checks on extra virgin olive
oil is required.
IR spectroscopy can be performed quickly with advantages
in terms of cost per sample.
Spectroscopy coupled with chemometric methods is one of
the most convenient and straightforward analytical tools to study
food products. The IR spectrum is obtained by a change in
the molecular dipole moment during vibration. Mid-IR (MIR)
spectroscopy (4000–400 cm−1 ) applied to a food sample gives
information on the molecular bonds and therefore on the func-
tional groups of the molecule. Spectrum bands involve many
of the general stretching, bending and wagging motions of the
molecule, and the overall spectrum may act as a fingerprint for
a given compound or compounds. Thus, the olive oil spectrum
may be considered to represent its overall chemical composition
and therefore have the potential to detect the olive oil quality
[10,11].
Fourier transform infrared spectroscopy with attenuated total
reflectance (ATR) or transmission cell accessories has been used
to quantify chromanols in vegetable oils [12], to authenticate,
identify or classify fats and vegetable oils [13–16], but little
has been reported about its utility for the evaluation of olive oil
oxidation.
This paper reports the potential of ATR-MIR spectroscopy,
in combination with multivariate statistical analysis for evaluat-
ing the freshness, i.e. oil quality at bottling time, stored in real
life conditions and not applying an accelerated thermo-oxidation
process. For this study, virgin olive oils were stored in the light
for 1 year and in the dark for 1 or 2 years; 1 year being gen-
erally considered the maximum storage period from bottling to
consumption. This approach could represent a faster recognition
tool for monitoring olive oil freshness since it is characterized
by simplicity of sample preparation.
129
2. Experimental
2.1. Oil sample
Fresh and mature olive fruits were harvested from typical
cultivars of Garda region and within 24 h were processed. From
these olives, samples of monovarietal VOOs were obtained using
a micro-oil press equipped with a hammer crusher, a vertical
mixer and a two-phase decanter (Alfa Laval, Firenze, Italy).
The data set was composed of 91 samples of VOOs treated as
follows:
- 35 oil samples were analysed within a month from the pro-
duction (fresh);
- 32 oil samples were divided in two aliquots and stored for 1
year in dark conditions (dark 1 year) and light (light 1 year),
respectively;
- 24 oil samples were stored for 2 years in dark conditions (dark
2 years).
All the samples were put into 100 mL amber (dark) and
transparent (light) closed glass bottles and stored at room tem-
perature. The average room temperature was between 17 and
25 ◦ C considering both the variation of day and night and sea-
son. A headspace equivalent to 10% of the bottle volume was
maintained to avoid oxidation during storage.
2.2. Chemical analyses
The chemical analysis included the measurement of several
parameters: the free acidity (acidity), which is indicative of the
free fatty acid content of the oil, expressed as oleic acid (%);
the peroxide value, which is a measure of the amount of the
hydroperoxides (meqO2 kg−1 ) formed through oxidation during
storage, and finally, the UV absorbance at 232 and 270 nm (K232 ,
K270 ). The chemical analyses were performed in triplicate on
each oil sample, before and after storage, according to Official
Methods of the European Regulation [3]. All the used chemicals
and solvents were of analytical grade.
2.3. FTIR spectroscopy
ATR FTIR experiments were performed on a spectrometer
(TENSOR 27, Bruker Optics S.r.l., Milano, Italy) equipped with
a deuterated triglycine sulfate (DTGS) detector. For both back-
ground and sample, 16 cm−1 asymmetric interferograms with a
spatial resolution of 4 cm−1 were co-added and Fourier trans-
formed. The oil samples (fresh, dark and light 1 year, dark 2
years) were positioned on a horizontal germanium crystals ATR
plate with 11 internal reflections. Single beam spectra of the
samples were collected and rationed against a background of
air. For each sample, spectral data were collected in the range
4000–550 cm−1 at room temperature (20 + 0.5 ◦ C) and stored as
log (1/R). Opus software (v. 5.5, Bruker Optics S.r.l., Milano,
Italy) was used for spectral acquisition, instrument control and
preliminary file manipulation. In fact to eliminate disturbing
H2 O and/or CO2 bands in the ratio spectra, the spectra were
130 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134

compensated with the Atmospheric Compensation command,
using Opus software.
2.4. Data processing
Statistical analysis of the chemical parameters was performed
by SPSS statistical software (version 13.0 Inc., Chicago, IL,
USA) with General Linear Model (GLM) procedure, using
Student–Newman–Keuls test.
To reduce the baseline shift, the spectral data were trans-
formed into second derivative (Savitzky–Golay method, number
of smoothing points = 10).
Principal component analysis (PCA) was performed as an
explorative analysis to each data set (chemical data and FTIR
spectra) in order to display the data structure and to reduce the
dimensionality of the spectral data to a small number of com-
ponents. PCA is a mathematical procedure for resolving sets
of data into orthogonal components whose linear combinations
approximate the original data to any desired degree of accuracy
[17,18]. PCA was performed using The Unscrambler software
package (version 9.6, CAMO, ASA, Norway).
To apply classification methods VOO samples were divided
into four classes: fresh (class 1), dark 1 year (class 2), dark 2
years (class 3) and light 1 year (class 4). Classification tech-
niques investigated were Partial Least Square Discriminant
Analysis (PLS-DA), linear discriminant analysis (LDA), and
soft independent modelling of class analogy (SIMCA).
PLS-DA is a variant of partial least square (PLS) regression,
in which each sample, in the calibration set, is assigned to a
dummy variable (0,1) as a reference value. The classification of
the olive oils according to VOOs freshness was on the basis of
the 0.5 cut-off value. For each class, olive oils samples were split
into calibration and prediction sample sets, assigning samples to
each set on the basis of their position of the class file, i.e. 50% of
sample was used for calibration and the remainder for prediction
[18]. The PLS-DA regression used in this study was performed
by using the PLS II model regression of statistical package The
Unscrambler (version 9.6, CAMO, ASA, Norway).
Linear discriminant analysis is a supervised classification
technique where the number of categories and the samples
belonging to each category are previously defined. The criterion
of LDA for selection of latent variables maximizes the differ-
ences between categories and minimizes the variances within
categories. The method produces a number of orthogonal linear
discriminant functions, equal to the number of the categories
minus one, that allow the samples to be classified in one or
another category [19,20].
LDA was carried out using SPSS (version 13.0 Inc., Chicago,
IL, USA) on the PCA samples scores of the significant compo-
nents (PCs), using full cross-validation (leave one out method),
both for chemical and spectral data set.
SIMCA method considers different classes which are mod-
elled individually by a separate principal component model. The
number of significant PCs is determinate for each class and as
many models as the number of classes are obtained by training.
The test set is used to test all class models [18,21]. SIMCA model
was developed using first four classes and then three classes
(fresh, dark and light) and was performed using the software
The Unscrambler (version 9.6, CAMO, ASA, Norway). Class
cut-off limits in SIMCA were set at the 5% level.
3. Results and discussion
3.1. Chemical analyses
The quality of virgin olive oils was determined by analytical
parameters: acidity, PV, K232 and K270 . Table 1 shows the mean,
the standard deviation, the minimum and maximum value of
these analytical parameters. Statistically significant differences
among VOO samples were observed for acidity, PV and UV

Table 1
Range of chemical parameters of virgin olive oil samples
Fresh (n = 35) Dark 1 year (n = 16) Dark 2 years (n = 24) Light 1 year (n = 16)

Acidity (%) F = 24.79*** Mean 0.12a 0.22b 0.22b 0.23b

S.D. 0.025 0.090 0.064 0.067
Min 0.09 0.11 0.11 0.11
Max 0.17 0.37 0.39 0.34

PV (meqO2 kg−1 ) F = 191.92*** Mean 4.93a 23.63b 32.02c 55.56d

S.D. 1.18 3.36 11.21 10.47
Min 3.00 17.50 13.00 34.50
Max 7.80 29.00 53.00 66.50

K232 F = 258.91*** Mean 1.30a 2.34b 2.82c 3.30d

S.D. 0.26 0.51 0.11 0.17
Min 0.82 1.06 2.60 3.00
Max 2.02 2.98 3.00 3.57

K270 F = 145.49*** Mean 0.11a 0.32b 0.32b 0.81c

S.D. 0.034 0.12 0.12 0.18
Min 0.05 0.10 0.16 0.56
Max 0.24 0.54 0.60 1.16

PV, peroxide value; K232 and K270 UV absorbance at 232 and 270 nm; S.D., standard deviation; min, the minimum value, max, the maximum value.
*** Significant difference (p < 0.001).
 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134
values. In particular, fresh oils result always different from stored
samples.
All fresh samples, immediately analysed after production
(class 1), show the analytical values within the law limits (acid-
ity ≤ 0.8%, PV ≤ 20 meqO2 kg−1 , K232 ≤ 2.4, K270 ≤ 0.22);
consequently, they can be considered as extra virgin olive oils
according to European legislation [22].
The samples analysed after 1 and 2 years of storage in dark
conditions (class 2 and class 3, respectively), and after 1 year
in light conditions (class 4), presented an acidity value lower
than 0.4. The peroxide values were in the range from 17.5 to
29 meqO2 kg−1 for class 2, from 13 to 53 meqO2 kg−1 for class
3, and from 34.5 to 66.5 meqO2 kg−1 for class 4. All stored
samples showed UV values higher than those established by
legislation.
Results obtained by PCA applied to correlation matrix of the
entire set of chemical data, are reported in Fig. 1, which shows
the bi-plot of the two first principal components, explaining 89%
of the total variance. A sample distribution indicative of olive
oil freshness was found, showing freshness decay associated
with storage conditions. In fact, a separation between fresh and
oxidized olive oil samples was found along PC1; the fresh olive
oils, belonging to class 1, were located mainly on the left zone of
the plane where PC1 scores are negative, while the oxidized olive
oil samples were located on the right zone of the PC1. On the
PC plane, samples stored in dark conditions are located between
fresh and light samples. Acidity, PV, and UV values were all
responsible for a difference between fresh and oxidized olive
oils; stored samples showing higher values of these parameters
than fresh olive oil samples. In particular PV and UV had the
higher eigenvectors on PC1.
LDA classification model was developed using the PCA
scores (first five PCs). An overall rate of 87% of correct clas-
sification was obtained by the model; in particular the samples
of class 1 were 100% correctly classified, while the samples of
class 2, 3 and 4 were 50%, 88% and 94%, correctly classified.
3.2. FTIR spectroscopy
Fourier Transform infrared (FTIR) spectra of olive oils stored
in different conditions are shown in Fig. 2. The spectra are
Fig. 1. Biplot of the first two PCs of olive oil samples using chemical data.
131
Fig. 2. FTIR spectra of olive oil samples, fresh and stored in different conditions
and periods.
dominated by some peaks at 2924, 2852, 1743, 1463, 1377,
1238, 1163, 1114, 1099 and 721 cm−1 . Absorbance at 2924 and
2852 cm−1 are due to bands arising from CH2 stretching vibra-
tions, asymmetric and symmetric, respectively. The major peak
at 1743 cm−1 arises from C O stretching vibrations; the bands
at 1463 and 1377 cm−1 arise from CH2 and CH3 scissoring
vibration, while those at 1238, 1163, 1114, 1099 are associated
with the C O stretching vibration. A small peak at 721 cm−1
corresponds to CH2 rocking mode [11–13,23–25].
A preliminary examination of the spectra was performed by
PCA applied to the second derivative of the spectra in the range
4000–550 cm−1 . A scatter plot of sample scores on the first two
components is shown in Fig. 3. The first component (89% of the
total variance) was able to separate the fresh oils, characterized
by negative PC1 values, from all other samples stored in different
conditions. Furthermore, the samples stored in light conditions
for 1 year were discriminated from those stored in the dark for
1 and 2 years, these latter partially separated along PC2 (5% of
the total variance). Thus, a fairly good sample separation was
possible on the basis of different storage conditions, even if, oils
stored in darkness for different periods (1 or 2 years) were not
well divided.
The loading plot was analysed in order to show which vari-
ables influenced the separation among oils. As can be seen in
Fig. 3. Score plot of the first two PCs obtained by PCA applied to the second
derivative of virgin olive oils spectra.
132 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134

Classification of the four oil classes was first attempted using

Fig. 4. Loading plot of PC1 and PC2 obtained by PCA applied to the second
derivative of virgin olive oils spectra: (a) first principal component, (b) second
principal component.
PC1 and PC2 loading plots (Fig. 4), the major peak arising from
C O stretching vibrations (1743 cm−1 ), CH3 scissoring vibra-
tion (1363 cm−1 ) and C O stretching vibration (1218 cm−1 )
were identified as principal wavenumbers explaining variation
during the storage of olive oils [11–13,23–25]. In particular on
PC1:
- 1743 cm−1 , corresponds to C O stretching vibration of alde-
hydes and ketons;
- 1363 cm−1 corresponds to CH3 scissoring vibration of ethers;
- 1218 cm−1 corresponds to C O stretching vibration of epox-
ides.
The variation in the above absorption bands could be related
to a higher concentration of aldehydes, ketons, epoxides in the
stored oil samples than in the fresh oils.
PLS-DA: the results are summarized in Table 2. The best overall
classification was obtained by using second derivative of spectral
data in the wavenumber range 4000–550 cm−1 . The score and
the loading plots of the first two PCs of the PLS II model gave
the same information as those obtained by PCA analysis and
described above.
The r and RMSEP for the PLS II model (three PLS latent
variables) were 0.98 and 0.08 for class 1 (fresh samples), 0.88
and 0.18 for class 2 (dark 1 year), 0.92 and 0.17 for class 3 (dark
2 years), and 0.93 and 0.15 for class 4 (light 1 year). The PLSII
model was able to correctly classify 100% of oil samples on the
basis of storage conditions.
Table 3 shows the results obtained by LDA classification
based on the scores of the first and second PCs, which accounted
for more than 90% of the total variance. In cross validation,
fresh olive oils and oxidized olive oils in the light were 100%
correctly classified, while the samples belonging to class 2 and
class 3 were 75% and 92% correctly classified, respectively. In
fact, only 12 out of 16 samples stored in the dark for 1 year
and 22 out of 24 stored in the dark for 2 years were correctly
identified. In any case, an overall rate of 93% of cross-validated
grouped cases was achieved using LDA. The misclassified sam-
ples belonging to class 2 (1 year in the dark) are classified as
belonging to class 3 (2 years in darkness) and vice versa. LDA
results also evidenced the ability of MIR spectra to discriminate
among VOOs stored in different conditions.
Finally, the spectral data was subjected to analysis by
SIMCA, developing models for fresh and oxidized VOOs (dark
1 year, dark 2 years and light 1 year). The obtained results are
shown in Table 4.
For each of the four classes, PCA models were built using half
of a sample belonging to each category and were tested with the
remanding samples. The best models were derived from second
derivative spectral data between 4000 and 550 cm−1 and were
able to identify correctly 18 fresh olive oils out of 18 samples and
8 oils oxidized in the light out of 8 samples, using four and six
principal components, respectively. On the contrary, the models
were incapable to classify oils oxidized in dark conditions (1
and 2 years). In fact, only 5 out of 8 oxidized samples (1 year
in the dark) and 5 out of 12 samples stored for 2 years in the
dark were correctly classified. However, an overall rate of 78%
correct classification was obtained by SIMCA model. Classes
overlapping regard only stored samples in darkness.
Due to the poor ability of these models to classify olive oils
stored in the dark, the SIMCA analysis was repeated consider-

Table 2
PLS-DA classification results of virgin olive oil samples using FTIR spectral data (class 1 = fresh; class 2 = dark 1 year; class 3 = dark 2 years; class 4 = light 1 year)
Class Calibration Prediction

No. samples Total % samples False negative False positive No. samples correctly % samples correctly False negative False positive
correctly classified correctly classified classified classified

1 17/17 (100%) 0 0 18/18 (100%) 0 0

2 8/8 (100%) 0 0 8/8 (100%) 0 0
100 100
3 12/12 (100%) 0 0 12/12 (100%) 0 0
4 8/8 (100%) 0 0 8/8 (100%) 0 0
 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134 133

Table 3
LDA classification results of virgin olive oil samples using FTIR spectral data (class 1 = fresh; class 2 = dark 1 year; class 3 = dark 2 years; class 4 = light 1 year)
Classes 1 2 3 4 Total
Fitting 1 35 (100%) 0 0 0 35
2 0 14 (87%) 2 6 16
3 0 2 22 (92%) 0 24
4 0 0 0 16 (100%) 16
Cross-validation 1 35 (100%) 0 0 9 35
2 0 12 (75%) 4 6 16
3 0 2 22 (92%) 0 24
4 0 0 0 16 (100%) 16

Table 4
SIMCA classification results using four classes of virgin olive oil samples (class 1 = fresh; class 2 = dark 1 year; class 3 = dark 2 years; class 4 = light 1 year)
Classes PCs No. of samples correctly classified % Samples correctly classified False negative False positive

1 4 18/18 (100%) 0 0
2 6 5/8 (62%) 3 7
78%
3 6 5/12 (42%) 7 3
4 6 8/8 (100%) 0 0

Table 5
SIMCA classification results using three classes of virgin olive oil samples (class 1 = fresh; class 2 = dark 1 and 2 years; class 3 = light 1 year)
Classes PCs No. of samples correctly classified % Samples correctly classified False negative False positive

1 4 18/18 (100%) 0 0
2 6 20/20 (100%) 100% 0 0
3 6 8/8 (100%) 0 0

ing only three classes, one of them including all olive oils stored
in darkness (both for 1 and 2 years). The improved results are
shown in Table 5. The new total rate of samples correctly classi-
fied was 100%. In fact, the fresh olive oils, the olive oils oxidized
in the dark for 1 and 2 years, and in the light were correctly
identified.
4. Conclusions
These results demonstrated that ATR-MIR spectroscopy is
able to classify virgin olive oil samples stored in different con-
ditions and to evaluate the freshness of olive oils (i.e. their
oxidation). In fact, the FTIR PCA results allowed a better
discrimination among fresh and oxidized oils, than samples sep-
aration obtained by PCA applied to chemical data. Moreover,
FTIR spectroscopy gave sufficient information to describe the
different storage conditions and storage periods. The ability of
FTIR spectra to discriminate among VOOs stored in the light
and in the dark was also demonstrated by the different classifi-
cation technique. It was also interesting to note the absence of
any false positive in PLS-DA and SIMCA models, using three
classes.
Therefore, the spectroscopic technique could be applied for
the prediction of olive oil freshness providing information com-
parable to chemical modifications, as FTIR spectroscopy results
are in agreement with those obtained by classical methods
results. The success of this technique depends on the concen-
tration and ratio between the major chemical components of the
olive oil samples. In addition, the same spectra can be used to
obtain quantitative information on several parameters used for
routine quality control of oils.
A higher number of samples might be necessary to develop
a more robust method to be used commercially as an indus-
trial application. More samples are also needed to improve the
precision of the classification models developed.
Anyway, the work points out the great advantages of using
ATR-MIR spectroscopy instead of chemical analysis. Advan-
tages are related to rapidity, non-destructive characteristics and
low cost per sample of MIR spectroscopy as well as need-
ing neither particular sample preparation nor specially qualified
laboratory personnel.
In conclusion, ATR-MIR could represent a reliable, cheap
and fast classification tool, able to verify the freshness of virgin
olive oils, not requiring chemical analyses for discrimination
among freshness classes.
References
[1] A. Bendini, L. Cerretani, S. Vecchi, A. Carrasco-Pancorbo, G. Lercker, J.
Agric. Food Chem. 54 (2002) 4880.
[2] F. Shahidi, U.N. Wanasundara, In: Fereidoon Shahidi (Ed.) Natural Antiox-
idant, Chemistry, Health Effect and Applications, 1996.
[3] E.C. Off. J. Eur. Communities, Regulation 2568/1991, 11 July 1991.
[4] COI/T.15/NC no. 2/REV. 10, 2001.
[5] V. Papadimitriou, T.G. Sotiroudis, A. Xenakis, N. Sofikiti, V. Stavyian-
noudaki, N.A. Chaniotakis, Anal. Chim. Acta 573 (2006) 453.
[6] T. Oshima, A. Hopia, B. German, E.N. Frankel, Lipids 31 (1996) 1091.
[7] P. Rovellini, N. Cortesi, E. Fedeli, Riv. Ital. Sostanze Grasse 75 (1998) 57.
134 N. Sinelli et al. / Analytica Chimica Acta 598 (2007) 128–134
[8] M.T. Morales, R. Aparicio, J.J. Rios, J. Chromatogr. A 668 (1994) 455.
[9] F. Caponio, M.T. Bilancia, A. Pasqualone, E. Sikorska, T. Gomes, Eur.
Food Res. Technol. (2005) 92.
[10] D.A. Skoog, F.J. Holler, T.A. Nieman, Principles of Instrumental Analysis,
fifth ed., Saunders College Pub., Philadelphia, 1998.
[11] N.B. Colthup, L.H. Daly, S.E. Wiberly, Introduction to Infrared and Raman
Spectroscopy, third ed., Accademic Press, Inc., New York, NY, 1990.
[12] M. Khalique Ahmed, J.K. Daun, R. Przybylski, J. Food Compos., Anal. 18
(2005) 359.
[13] H. Yang, J. Irudayaraj, M.M. Paradkar, Food Chem. 93 (2005) 25.
[14] M. Safar, D. Bertrand, P. Robert, M.F. Devaux, C. Genot, J. Am. Oil Chem.
Soc. 71 (1994) 371.
[15] D.B. Dahlberg, S.M. Lee, S.J. Wenger, J.A. Vargo, Appl. Spectrosc. 51
(1997) 1118.
[16] Y.W. Lai, E.K. Kemsley, R.H. Wilson, J. Agric. Food Chem. 42 (1994)
1154.
[17] K.R. Beebe, R.J. Pell, M.B. Seasholtz, Chemometrics, A Practical Guide,
John Wiley and Sons, New York, NY, 1998.
[18] T. Naes, T. Isaksson, T. Fearn, A.M.C. Davies, A User-friendly Guide to
Multivariate Calibration and Classification, NIR Publication, Chichester,
UK, 2000.
[19] M. Meloun, J. Militky, M. Forina, Chemometrics for Analytical Chemistry,
Ellis Horwood, New York, NY, 1992.
[20] M. Blanco, J. Pages, Anal. Chem. 463 (2002) 295.
[21] Q. Chen, J. Zhao, H. Zhang, X. Wang, Anal. Chem. 572 (2006) 77.
[22] E.C. Off. J. Eur. Communities, Regulation 1513/2001, 23 July 2001.
[23] M.D. Guillen, N. Cabo, J. Agric. Food Chem. 46 (1998) 1788.
[24] J. Coates, in: R.A. Meyers (Ed.), Encyclopaedia of Analytical Chemistry,
John Wiley & Sons Ltd., Chichester, 2000, p. 10815.
[25] J. Workman, Handbook of Organic Compounds, Academic Press, London,
2001, p. 209.