Full Text

REVIEW PAPER

Fluorescence Spectroscopy Measurement for QualityAssessment of Food Systems—a Review

Romdhane Karoui & Christophe Blecker

Received: 13 March 2010 /Accepted: 26 April 2010# Springer Science+Business Media, LLC 2010

Abstract The present review gives an overview of the useof fluorescence spectroscopy (i.e., conventional, excitation–emission matrix, and synchronous fluorescence) for deter-mining changes in food products and their quality duringtechnological process and storage. From the present review,it was shown that fluorescence spectroscopy is able todetermine several properties (functional, composition, nutri-tional) without the use of chemical reagents. This is due tothe use of chemometric tools (descriptive and predictivemethods). The review focuses on the use of fluorescencespectroscopy for the determination of the quality of animal(i.e., dairy, meat, fish, and egg) and vegetable (oils, cereal,sugar, fruit, and vegetable) products as well as the identifi-cation of bacteria of agro-alimentary interest.

Keywords Fluorescence spectroscopy . Food systems .

Quality . Chemometrics

Introduction

Public interest in food quality and production has increasedin recent decades, probably related to changes in eatinghabits, consumer behavior, and the development andincreased industrialization of the food supplying chains(Christensen et al. 2006). The demand for high quality andsafety in food production obviously calls for high standardsfor quality and process control, which in turn requires

appropriate analytical tools to investigate food. Fluores-cence spectroscopy is an analytical technique whose theoryand methodology have been extensively exploited forstudies of molecular structure and function in the disciplineof chemistry and biochemistry (Strasburg and Ludescher1995). Even though fluorescence is one of the oldestanalytical methods used (Valeur and Bochon 2001), it hasjust, recently, become quite popular as a tool in biologicalscience related to food technology. An indication for thispopularity is the increasing number of research publicationsabout fluorescence as well as the introduction of newcommercially available instruments for fluorescence analy-sis, particularly, front-face fluorescence spectroscopy(FFFS). Traditional right angle fluorescence spectroscopictechnique cannot be applied to thick substances due to largeabsorbance and scattering of light. Indeed, when theabsorbance of the sample exceeds 0.1, emission andexcitation spectra are both decreased and excitation spectraare distorted (Karoui et al. 2003). To avoid these problems,a dilution of samples (when it is possible) was performed sothat their total absorbance would be less than 0.1. However,the results obtained on diluted solution of food samplescannot be extrapolated to native concentrated samples sincethe organization of the food matrix is lost. To reply withthis request, FFFS could be utilized. The use of onlyexcitation and emission wavelengths could limit the abilityof fluorescence spectroscopy to determine the quality offood systems. To comply with this requirement, thevariation in the excitation and emission wavelengths allowssimultaneous determination of compounds in several food-stuffs. Therefore, it would be interesting to use for eachfood product different excitation wavelengths simulta-neously. This could be realized by using synchronousfluorescence spectroscopy (SFS) which presents two inter-esting advantages from our point of view: it (1) allows the

R. Karoui (*) :C. BleckerDepartment of Food Technology, Gembloux Agro-Bio Tech,University of Liège,Passage des Déportés, 2,5030 Gembloux, Belgiume-mail: [email protected]

Food Bioprocess TechnolDOI 10.1007/s11947-010-0370-0

consideration of the whole fluorescence landscape, i.e.,spectra recorded at different offsets and (2) retainsinformation related to several fluorophores compared to aclassical emission spectrum, which is mainly specific to asole fluorophore. Synchronous fluorescence spectra areobtained by simultaneously scanning both the excitationand emission monochromators keeping a fixed wavelengthinterval, named offset between them. It gives a narrowerand simpler spectrum. For the SFS technique, the selectionof a wavelength interval is one of the most importantexperimental parameter and the parameter should beoptimized, which is carried out by measuring the spectraat various offsets.

This review will provide the reader with the basicprinciples of fluorescence including the use of thistechnique, especially the use of the most common FFFSand synchronous fluorescence for the assessment of thequality of several food systems that will be discussed indetail (Table 1).

Fluorescence Spectroscopy

Definition

Fluorescence is the emission of light subsequent toabsorption of ultraviolet or visible light of a fluorescentmolecule or substructure, called a fluorophore. Thus, thefluorophore absorbs energy in the form of light at a specificwavelength and liberate energy in the form of emission oflight at a higher wavelength. The general principles can beillustrated by a Jablonski diagram (Papageorgiou andGovindjee 2004; Rost 1995; Zude 2008), as shown inFig. 1.

The first step (1) is the excitation, where light isabsorbed by the molecule, which is transferred to anelectronically excited state, meaning that an electron goesfrom the ground singlet states, S0, to an excited singletstate, S′1. This is followed by a vibrational relaxation orinternal conversion (2), where the molecule undergoes atransition from an upper electronically excited state to alower one, S1, without any radiation. Finally, the emissionoccurs (3), typically 10−8 s after the excitation, when theelectron returns to its more stable ground state, S0,emitting light at a wavelength according to the differencein energy between the two electronic states. This expla-nation is somewhat simplified. In molecules, each elec-tronical state has several associated vibrational states. Inthe ground state, almost all molecules occupy the lowestvibrational level. By excitation with ultraviolet or visiblelight, it is possible to promote the molecule of interest toone of several vibrational levels for the given electroni-cally excited level. This implies that absorption and

fluorescence emission does not only occur at one singlewavelength, but rather over a distribution of wavelengthscorresponding to several vibrational transitions as compo-nents of a single electronic transition. This is whyexcitation and emission spectra are obtained to describethe detailed fluorescence characteristics of molecules. Infact, fluorescence is characterized by two wavelengthparameters that significantly improve the specificity of themethod, compared to spectroscopic techniques based onlyon absorption.

Quantum Yield (Efficiency)

Each molecule presents a specific property, which isdescribed by number, named quantum yield or quantumefficiency (�).

f ¼ number of quanta emitted

number of quanta absorbed¼ quantum yield ð1Þ

As illustrated in Eq. 1, the higher the value of �, thegreater the fluorescence of a compound (e.g., chlorophyll).A practically non-fluorescent molecule (e.g., carotenoids) isone whose quantum efficiency is zero or so close to zerothat the fluorescence is not measurable. All energyabsorbed by such a molecule is rapidly lost by collisiondeactivation.

Excitation and Emission Spectra

Excitation Spectrum

The excitation spectrum is defined as the relative efficiencyof different wavelengths of exciting radiation in causingfluorescence. The shape of the excitation spectrum shouldbe identical to that of the absorption spectrum of themolecule and independent of the wavelengths at whichfluorescence is measured. However, this is seldom the casebecause the sensitivity and the bandwidth of the spectro-photometer (absorbance spectrum) and the spectrofluorim-eter (excitation spectrum) are different. In addition, formany food samples, scattering properties and energytransfer between neighboring molecules could contributeto this difference. A general rule of thumb is that thestrongest (generally the longest) wavelength peak in theexcitation spectrum is chosen for excitation of the sample.This minimizes possible decomposition caused by theshorter wavelength, higher energy radiation.

Emission Spectrum

The emission spectrum of a compound results from theradiation absorbed by the molecule. The emission spectrum

Food Bioprocess Technol

is the relative intensity of radiation emitted at variouswavelengths. In theory, the quantum efficiency and theshape of the emission spectrum are independent of thewavelength of the excitation radiation. In practice, this isnot the case. Indeed, it has been shown that fluorescenceof chlorophyll from a green leaf has a lower shortwavelength emission maximum when excited with greenlight than when excited with blue light (Buschmann2007). Green light penetrates more deeply into the leafsince it is less absorbed than blue light and the green light-excited fluorescence from more inside the leaf is morereadily re-absorbed by the chlorophylls on its way to thesample surface. The re-absorption of fluorescence isparticularly high in the short wavelength fluorescencewhere it overlaps with the absorption spectrum ofchlorophyll. If the exciting radiation is at wavelength thatdiffers from the wavelength of the absorption peak, lessradiant energy will be absorbed and hence less will beemitted.

Stokes Shift

According to the Jablonski diagram (Fig. 1), the energy ofemission is lower than that of excitation. This implies thatthe fluorescence emission occurs at higher wavelengthsthan the absorption (excitation). The difference betweenthe excitation and emission wavelengths is known asStokes shift (Valeur and Bochon 2001), as indicated withthe arrow in Fig. 2, marking the difference between theexcitation and emission spectrum of tryptophan fluores-cence spectra scanned on milk submitted to ultra-hightemperature.

Stokes shift cm�1� � ¼ 107

1

lex� 1

lem

� �; ð2Þ

where λex and λem are the maximum wavelengths(nanometer) for excitation and emission, respectively.

Normally, the emission spectrum for a given fluoro-phore is a mirror image of the excitation spectrum, asseen to some extent in Fig. 2 for tryptophan. The generalsymmetric nature is a result of the same transitions beinginvolved in both absorption and emission and thesimilarities of the vibrational levels of S0 and S1.However, there are several exceptions, since severalabsorption bands can be observed in the excitationspectrum but only the last peak is observed in the emissionspectrum, representing the transition from S1 to S0. Thefluorescence of vitamin A, as seen in Fig. 3, is an exampleof this, with three absorption peaks and only oneemission peak. Normally, only emission or excitationspectra (i.e., one excitation or emission wavelength) arerecorded when investigating the fluorescence of a

sample. However, it can be beneficial and informativeto obtain the entire fluorescence landscape (also knownas two-dimensional fluorescence spectroscopy) in orderto find the exact excitation and emission maxima, as wellas the correct structure of the peaks. Furthermore, itfacilitates more appropriate analysis of fluorescence datafrom complex samples with more fluorophores present(Christensen et al. 2006).

Factors Affecting Fluorescence Intensity

Several factors related to the nature and the concentrationof fluorophores of food samples influence the fluorescenceintensity.

Quenching

Fluorescence quenching represents any process leading to adecrease in fluorescence intensity of the sample (Lakowicz1983). It is related to the deactivation of the excitedmolecule by either intra- or intermolecular interactions.There are two types of quenching: statistic and dynamic.The first occurs when the formation of the excited state isinhibited due to a ground state complex formation inwhich the fluorophore forms non-fluorescent complexeswith a quencher molecule. Dynamic or collisional quench-ing refers to the process when quenchers deactivate thebehavior of the excited state after its formation. Theexcited molecule will be deactivated by either intramolec-ular interaction (collision) or intermolecular activity(interaction with other molecules). One of the best-known quenchers is oxygen. A higher temperature alsoresults in larger amount of collisional quenching due to theincreased velocities of molecules. Resonance energytransfer can also be considered as dynamic quenching,since an interaction between the donor and acceptormolecules could take place inducing a full or partialdeactivation of the excited fluorophore (donor). Theenergy transfer does not involve emission of light, but adipole–dipole interaction between the donor and acceptormolecule.

Concentration and Inner Filter Effect

The equation defining the relationship of fluorescenceintensity to concentration is:

If ¼ fI0 1� 10�"lc� �

;

where If is the fluorescence intensity, f the quantum yield,I0 the intensity of the incident light, ε the molarabsorptivity, l the optical depth of the sample, and c themolar concentration of the fluorophore.


Tab

le1

Overview

oftheliteraturesurvey

onfluo

rescence

ofdifferentfood

system

s

Fluorop

hore

Dairy

prod

ucts

Meat

Fish

Egg

Edibleoils

Cereal

Sug

arFruitandvegetable

Bacteria

Aminoacids

andnu

cleic

acids

Bou

bello

utaandDufou

r(200

8),Dufou

rand

Riaub

lanc

(199

7),

Dufou

ret

al.(200

0),

Ham

mam

iet

al.(201

0),

Herbertet

al.(199

9,20

00),Karou

iand

Dufou

r(200

3,20

06),

Karou

iet

al.20

04a,

b,20

05a,

b,c,

2006

b,e,

2007

c,Kulmyrzaev

etal.(200

5),Liu

and

Metzger

(200

7),

Mazerolleset

al.

(200

1),Rou

issiet

al.

(200

8),Scham

berger

andLabuza(200

6),

Zaïdi

etal.(200

8)

Allaiset

al.(200

4),

Dufou

randFrencia

(200

1),Frencia

etal.(200

3),Lebecqu

eet

al.(200

3),Møller

etal.(200

3),

Sahar

etal.(200

9a)

Dufou

ret

al.

(200

3),Karou

iet

al.(200

6f)

Karou

iet

al.

(200

6c,

d200

7d,e)

Karou

ietal.(200

6a),

Zando

meneghi

(199

9)

Baunsgaardet

al.

(200

0a,b),Bro

(199

9),Karou

iet

al.(200

7a),

Mun

cket

al.

(199

8),Nørgaard

(199

5),Ruo

ffet

al.(200

6)

Noh

andLu(200

7),

Seidenet

al.(199

6)Ammor

etal.

(200

4),Leblanc

andDufou

r(200

2,20

04),

Leriche

etal.

(200

4),

Tou

rkya

etal.(200

9)

Collagen

Egeland

sdal

etal.

(199

6,20

02,20

05),

Skjervo

ldet

al.

(200

3),Swatland

(198

7),S

watland

and

Findlay

(199

7),

Swatland

etal.

(199

5a,b)

Chlorop

hyll

Eng

elsen(199

7),

Guimet

etal.(200

4,20

05),Kyriakidis

andSkarkalis

(200

0),Pou

lliet

al.

(200

5,20

07),

Zando

meneghi

etal.

(200

5)

DeEllet

al.(199

6),

Hagen

etal.(200

6),

Lötze

etal.(200

6),

Mosho

uetal.(20

05)

Ferulic

acid

Sym

onsandDexter

(199

1,19

92,19

93,

1994)

Maillard

prod

ucts

Liu

andMetzger

(200

7),

Scham

berger

and

Labuza(200

6)

Karou

iet

al.

(200

6c,d)

Baunsgaardet

al.

(200

0a,b)

NADH

Kulmyrzaev

etal.

(200

5)Dufou

ret

al.

(200

3),Karou

iet

al.(200

6f)

Gho

shet

al.(200

5)Ammor

etal.

(200

4),Leblanc

andDufou

r(200

2,20

04),

Tou

rkya

etal.(200

9)


FADH

Kulmyrzaev

etal.(200

5)

Oxidatio

nprod

ucts

Karou

iet

al.(200

7b),

Woldet

al.(200

2,20

05)

Gatellieret

al.(200

7),

Mølleret

al.(200

3),

Olsen

etal.(200

5),

Sahar

etal.(200

9b),

Veberg(200

6),

Veberget

al.(200

6)

Hasegaw

aet

al.

(199

2),O

lsen

etal.(200

6)

Eng

elsen(199

7),

Guimet

etal.(200

4,20

05),Sikorskaet

al.(200

4,20

05)

Polyp

heno

lsSikorskaet

al.

(200

4,20

05),

Zando

meneghi

etal.(200

5)

Baunsgaardet

al.

(200

0a,b),Karou

ietal.(200

7a),Ruo

ffet

al.(200

6)Retinol

(vitamin

A)

Bou

bello

utaandDufou

r(200

8),Dufou

rand

Riaub

lanc

(199

7),Dufou

ret

al.(200

0),Ham

mam

iet

al.(201

0),Herbertet

al.(199

9,20

00),Karou

iandDufou

r(200

3,20

06),

Karou

iet

al.(200

4a,b,

2005

a,b,

c,20

06b,

e,20

07c),Kulmyrzaev

etal.

(200

5),Liu

andMetzger

(200

7),Mazerolleset

al.

(200

1),Rou

issiet

al.

(200

8),Zaïdi

etal.(200

8)

Karou

iet

al.

(200

6c,d,

2007

d,e)

Riboflavin

(vitamin

B2)

Bou

bello

utaandDufou

r(200

8),Ham

mam

iet

al.

(201

0),K

arou

iandDufou

r(200

3,20

06),Karou

iet

al.

(200

5c,20

06b,

e,20

07c),

Liu

andMetzger

(200

7),

Rou

issiet

al.(200

8),

Zaïdi

etal.(200

8)

Zando

meneghi

etal.(200

3)

Tocoph

erols

(vitamin

E)

Guimet

etal.(200

4,20

05),Kyriakidis

andSkarkalis

(200

0),Pou

lliet

al.

(200

5,20

07),

Sikorskaet

al.

(200

4,20

05),

Zando

meneghi

etal.

(200

5)


According to Lakowicz (1983), for low absorbance(<0.05), the equation can be written as:

If ¼ 2:3f I0 " l c;

where If is the fluorescence intensity, f the quantum yield,I0 the intensity of the incident light, ε the molarabsorptivity, l the optical depth of the sample, and c themolar concentration of the fluorophore. The decrease is in apart caused by an attenuation of the excitation beam in theareas of the solution in front of the detection system and bythe absorption of the emitted fluorescence within thesolution. This is defined as the inner cell or inner filtereffect. The equation expressing the fluorescence intensityindicates that there are three major factors other thanconcentration that could affect the fluorescence intensity:

1. The quantum efficiency �; the greater the �, thegreater the fluorescence intensity.

2. The intensity of incident light I0; a more intense sourcewill yield greater fluorescence. In actual practice, a veryintense source can cause photodecomposition of thesample. Hence, one compromise is a source of moderateintensity (such as a mercury or xenon lamp).

3. The molar absorptivity of the compound, ε. In order toemit radiation, a molecule must first absorb radiation.Hence, the higher the ε, the better the fluorescenceintensity of the compound will be.

It should be remembered that the overall fluorescenceintensity of a given sample is expressed as the sum of thefluorescence contribution from each of the inherent fluoro-phores present in the sample. However, due to the complexsystems of food products, the fluorescence intensity may notbe additive because the quenching phenomenon and inter-actions with the molecular environment of the fluorophoresmay take place.

Molecular Environment

The local environment of a fluorophore has an importanteffect on the shape of the fluorescence spectra. In more

polar environments, the fluorophore in excited state willrelax to a lower energy state of S1. This means that theemission of polar fluorophores will be shifted towardslonger wavelengths (lower energy) in more polar solvents.

The structure of macromolecules and the location inmacromolecules can also have a large effect on thefluorescence emission and quantum yield of a fluorophore.Temperature, pH, and color strongly affect the fluorescencesignal. Increased temperature leads to increased movementof the molecules, and thereby more collisions, thusinducing a reduced fluorescence signal. It is thereforeimportant that all samples in an experiment present thesame temperature. The pH value affects the fluorescence,and most hydroxyl aromatic compounds fluoresce better athigh pH (Guilbaut 1989). The color of the sample can affectboth the shape and the intensity of the spectra. Darksamples will reabsorb more of the fluorescence than brightsamples.

Scatter

Scattering of the incident light affects the fluorescencesignal. As mentioned in the previous section, the absor-bance of the sample measured plays an important role influorescence measurements. Especially in turbid solutionsand solid opaque samples (like most foods), the amount ofscattered and reflected light affects the measurementsconsiderably, with respect to both the sampling (i.e., theoptical depth of the sampling) and the obtained (fluores-cence) signal. Scattered light can be divided into Rayleighand Raman scatter.

Rayleigh scatter refers to the scattering of light byparticles and molecules smaller than the wavelength of thelight. Rayleigh is so-called elastic scatter, meaning that noenergy loss is involved, so the wavelength of the scatteredlight is the same as that of the incident light. Rayleighscatter can be observed as a diagonal line in fluorescencelandscapes for excitation wavelengths equaling the emis-sion wavelength. The signal from fluorophores with littleStokes shift will be situated close to the scattering line, andtherefore be most affected by Rayleigh scatter. Due to theconstruction of grating monochromators used for excitationin most spectrofluorometers, some light at the doublewavelength of the chosen excitation will also pass throughto the sample. For this reason, an extra band of Rayleighscatter, so-called second-order Rayleigh, will typicallyappear in fluorescence measurement for emission wave-lengths at twice the given excitation wavelength. Rayleighscatter can be disregarded by measuring and consideringthe fluorescence signal only between the first- and second-order Rayleigh scatter.

Raman scatter is inelastic scatter due to absorption andre-emission of light coupled with vibrational states. A

1 hνex

S0

S’1

Energy

S

3 hνem

2

Fig. 1 Jablonski diagram showing the basic principle in fluorescencespectroscopy


constant energy loss will appear for Raman scatter, meaningthat the scattered light will have a higher wavelength thanthat of the excitation light, with a constant difference inwavelength. In liquid samples, the solvent is decisive in theamount and nature of Raman scatter, while for solidsamples it will typically be an expression of the bulksubstances. Raman scatter can in most cases be neglectedbecause of its weak contribution to the fluorescence signal.

Instrumentation

The basic setup for an instrument for measuring steady-state fluorescence is shown in Fig. 4. The spectrofluorim-eter consists of a light source (generally xenon or mercurylamp); a monochromator and/or filter(s) for selecting theexcitation wavelengths; a sample compartment; a mono-chromator and/or filter(s) for selecting the emission wave-

lengths; a detector, which converts the emitted light to anelectric signal; and a unit for data acquisition and analysis.The sampling geometry can have a substantial effect on theobtained fluorescence signal. If absorbance is less than 0.1,the intensity of the emitted light is proportional to thefluorophore concentration and excitation and emissionspectra are accurately recorded by a classical right-anglefluorescence device. In this case, the excitation light travelsinto the sample from one side, and the detector ispositioned at right angles to the center of the sample.When the absorbance of the sample exceeds 0.1, theintensity of emission and excitation spectra decreases andexcitation spectra are distorted. To avoid these problems,dilution of samples (when it is possible, i.e., liquid samples)is currently performed so that their total absorbance will beless than 0.1. However, the results obtained on dilutedsolutions of food samples cannot be extrapolated to nativeconcentrated samples since the organization of the food

0

0.04

0.08

0.12

0.16

245 265 285 305 325 345 365 385

Wavelength (nm)

Flu

ore

scen

ce in

ten

sity

(a.

u.)

Fig. 2 Excitation (full line) andemission (dotted line) trypto-phan fluorescence spectrarecorded on UHT milk

0

0.04

0.08

0.12

0.16

250 300 350 400 450 500

Wavelength (nm)

Flu

ore

scen

ce in

ten

sity

(a.

u.)

Fig. 3 Excitation (full line) andemission (dotted line) vitamin Afluorescence spectra recorded onUHT milk


matrix is lost. In addition, the dilution may change theconcentration of other relevant fluorescent species below orclose to the detection limit of fluorescence. Moreover, forsolid samples, the dilution cannot be realized (e.g., meat,cheese). To avoid these problems, FFFS can be used(Fig. 4). In this manner, it is possible to measure more turbidor opaque samples, since the signal becomes more indepen-dent of the penetration of the light through the sample.However, when front-face sampling is used, the amount ofscattered light detected will increase due to the higher level ofreflection from the surface topology of the sample and sampleholder. To minimize these effects, it is recommended that thesample is not placedwith its surface oriented at an angle of 45°to the incident beam, but rather at 30°/60° to the light sourceand the detector (Lakowicz 1983).

Data Analysis

Data analysis of fluorescence spectra has been wellestablished by Smilde et al. (2004). Fluorescence isinherently multidimensional. Indeed, multidimensionalfluorescence signals recorded from a sample can conve-niently be presented as a matrix of fluorescence intensitiesas a function of excitation and/or emission wavelengths.Due to the neighboring wavelengths, highly correlated datapresent in emission and excitation spectra have beenpointed out (Smilde et al. 2004). In this case, principalcomponent analysis (PCA), common component andspecific weights analysis (CCSWA), partial least squares

regression (PLS), factorial discriminant analysis (FDA),parallel factor analysis (PARAFAC), etc. have proven to bepowerful methods for the extraction of valuable information(Boubellouta and Dufour 2010; Christensen et al. 2006;Hammami et al. 2010; Kulmyrzaev and Dufour 2010).

Applications of Fluorescence in Foods and Drinks

Recently, the application of fluorescence spectroscopy incombination with multidimensional statistical techniquesfor the evaluation of food quality has increased. In most ofthe research papers, the obtained fluorescence signal wasassigned to specific fluorophores after fixing the excitationor the emission wavelength.

Dairy Products

Dairy products contain several intrinsic fluorophores, whichrepresent the most important area of fluorescence spectrosco-py. They include the aromatic amino acids and nucleic acids(AAA+NA) tryptophan, tyrosine, and phenylalanine inproteins; vitamins A and B2; nicotinamide adenine dinucle-otide (NADH) and chlorophyll; and numerous other com-pounds that can be found at a low or very low concentrationin food products.

Dufour and Riaublanc (1997) investigated the potentialof FFFS to discriminate between raw, heated (70 °C for20 min), homogenized, and homogenized and heated milks.The authors applied PCA to the tryptophan and vitamin A

Excitation

Monochromator Interference filter

Monochromator Interference filter

Sample holder

Emission

Right-angle Front-face

Photodetector

Data acquisition and analysis

Light Fig. 4 Basic setup of aspectrofluorimeter


fluorescence spectra, and good differentiation between milksamples as a function of homogenization and heat treatmentapplied to milk samples was observed. They concluded thatthe treatments applied to milk induced specific modifica-tions in the shape of the fluorescence spectra. Recently,Kulmyrzaev et al. (2005) confirmed these earlier findings.In their research, the emission and excitation spectra ofdifferent intrinsic probes (i.e., AAA+NA, NADH, andFADH) were used to evaluate changes in milk followingthermal treatments in the range of 57–72 °C for 0.5–30 min. The PCA applied to the normalized spectra allowedgood discrimination of milk samples subjected to differenttemperatures and times; the obtained results were con-firmed recently by Boubellouta and Dufour (2008) whodiscriminated milk samples according to heating in 4–50 °Cand acidification (pH ranging from 6.8 to 5.1). Boubelloutaet al. (2009) determined the effect of the adding calcium,phosphate, and citrate at different levels (i.e., 3, 6, and9 mM) on the molecular structure of skimmed milks; theaddition of phosphate induced molecular change (observedby mid-infrared and synchronous fluorescence) that wasdifferent than that of calcium and citrate; the origin of thisdifference was not depicted by the authors. In these threelatter research studies, milk samples were heated at onlyrelatively low temperatures allowing the non-monitoring ofthe development of Maillard browning reaction; this wasrealized after by Schamberger and Labuza (2006); thefluorescence spectra of milks were processed for 5, 15, 20,25, and 30 s in 5 °C increments from 110 to 140 °C andfound to be well correlated with hydroxylmethylfurfural.Indeed, the R2 values of 0.95 were found continuouslythroughout the emission wavelength range of 394 to447 nm. In addition, the fluorescence levels increased withhigher time–temperature combinations. One of the mainconclusions of this study was that FFFS could be considered apromising method for measuring Maillard browning in milk;the authors encouraged the use of this technique for on-linemonitoring of thermal processing of milk. Later, Liu andMetzger (2007) confirmed the aforementioned results follow-ing the use of FFFS for monitoring changes in non-fat drymilk (n=9) collected from three different manufacturers andstored at four different temperatures (4, 22, 35, and 50 °C)for 8 weeks. Different intrinsic probes (fluorescent Maillardreaction products (FMRP), riboflavin, tryptophan, andvitamin A) were used, and each of the considered spectraldata sets allowed good discrimination of milk samples keptat 50 °C from the others. In addition, good discrimination ofmilk samples as a function of the storage time was observed.In a similar approach, Feinberg et al. (2006) used fluores-cence spectroscopy to identify five types of heat treatments(pasteurization, high pasteurization, direct ultra-high temper-ature (UHT), indirect UHT, and sterilization) of 200commercial milk samples stored at 25 and 35 °C for 90 days.

By applying FDA, Feinberg et al. (2006) found thattryptophan fluorescence spectra could be considered well-adapted to discriminate sterilized milks and probablypasteurized milks from the other milk samples. However,this intrinsic probe failed to discriminate the other types ofmilk. An explanation could be that fluorescence spectra wererecorded in the pH 4.6 soluble fraction of the milk sample,inducing a loss of information contained in milk samples.

Regarding milk coagulation, Herbert (1999) and Herbertet al. (1999) used FFFS to monitor milk coagulation at themolecular level. Three different coagulation processes havebeen studied: the glucono-δ-lactone (GDL), the rennet-induced coagulation system, and a mixed GDL and rennet-induced coagulation system. Emission fluorescence spectraof the tryptophan were recorded for each system during themilk coagulation kinetics. By applying the PCA to normal-ized fluorescence spectra data sets of the three systems,detection of structural changes in casein micelles duringcoagulation and discrimination of different dynamics of thethree coagulation systems was achieved. Herbert et al. (1999)concluded that FFFS allows the investigation of networkstructure and molecular interactions during milk coagulation.

Most of the aforementioned studies regarding discrimi-nation of milk were performed at a laboratory scale onextreme and controlled samples. Milk products frommountain areas are reputed to have specific organolepticand nutritional qualities (Bosset et al. 1999; Coulon andPriolo 2002; Renou et al. 2004), and the tracing of milkproduction sites is therefore important in order to avoidfraud. In this context, the potential of FFFS to discriminatebetween milks according to their geographical origin wasexplored. Forty milk samples—8 produced in lowland areas(430–480 m), 16 produced in mid-level areas (720–860 m),and 16 produced in mountain (1,070–1,150 m) areas—fromthe Haute-Loire department in France at key periods ofanimals feeding were analyzed (Karoui et al. 2005c).Tryptophan fluorescence spectra, AAA+NA spectra, andriboflavin spectra were recorded directly on milks, withexcitation wavelengths set at 290, 250, and 380 nm,respectively. The excitation spectra of vitamin A were alsorecorded, with the emission wavelength set at 410 nm. Byapplying FDA to the spectral collection, a trend to a goodseparation between milks as a function of their origins wasobserved. The best results were obtained with AAA+NAfluorescence spectra, since 81.5% and 76.9% of thecalibration and validation spectra, respectively, were cor-rectly classified. However, some misclassification occurredbetween milks produced in mid-level areas and the othermilk samples. In the same context, FFFS has demonstratedits potential for monitoring ewe’s milk samples accordingto the feeding system, i.e., scotch bean versus soybeanmeals (Hammami et al. 2010; Rouissi et al. 2008) andgenotype (Zaïdi et al. 2008).


Regarding the use of FFFS for monitoring the quality ofcheeses during ripening, Dufour et al. (2000) and Mazerolleset al. (2001) used FFFS to monitor 16 semi-hard cheesesproduced and ripened under a controlled scale. By applyingPCA to the normalized tryptophan fluorescence spectra,good discrimination of cheeses presenting a ripening time of21, 51, and 81 days was observed, while an overlap wasobserved between cheeses aged 1 day and those aged21 days. The spectral pattern of tryptophan indicated a redshift of aged cheeses suggesting that the environment ofripened cheeses was more hydrophilic than that of young (1-day-old) cheeses. This phenomenon was assigned to partialproteolysis of casein as well as to the salting phenomenon,which may induce some changes in the tertiary andquaternary structures of casein micelles. Regarding thefluorescence spectra of vitamin A, two shoulders located at295 and 305 nm and a maximum located at 322 nm wereobserved (Dufour et al. 2000). In addition, the shape of thespectra changed with ripening time. By applying PCA to thenormalized vitamin A spectra, better discrimination ofcheeses aged 21, 51, and 81 days from those aged 1 daywas achieved. The authors determined the link between thedata recorded by mid-infrared (MIR) and fluorescence byusing canonical correlation analysis (CCA). Correlationcoefficients of 0.58 or more were obtained suggesting thatfluorescence and MIR spectra might provide a commondescription of the investigated cheese samples duringripening.

Karoui et al. (2006b) continued this work by recordingtryptophan, vitamin A, and riboflavin spectra of 12 semi-hard cheeses (Raclette) of 4 different brands, which wereproduced during summer period at the industrial level. Byapplying CCSWA to the spectral data sets and physico-chemical data, good discrimination of the four brands wasobserved. The same research group (Karoui and Dufour2006) evaluated the potential of FFFS to predict therheological parameters of 20 semi-hard cheeses at the endof their ripening stage (60 days) from fluorescence spectrarecorded at a young stage (2 days old). By using tryptophanfluorescence spectra scanned on cheeses aged 2 days and at20 °C, the storage modulus (G′), loss modulus (G″), strain,tan (δ), and complex viscosity (η*) were predicted by usingPLS regression with leverage correction with R higher than0.97. The obtained results were confirmed recently byBoubellouta and Dufour (2010), reporting that synchronousfluorescence spectroscopy could be used for the determi-nation of fat melting and cheese melting of two cheesevarieties (i.e., Comté and Raclette).

In another context, FFFS has been used for theauthentication of different varieties of soft, semi-hard, andhard cheeses during ripening at the retail stage (Herbert1999; Herbert et al. 2000; Karoui and Dufour 2003; Karouiet al. 2003, 2004a, b, 2005a, b). Herbert et al. (2000)

explored the potential of FFFS to discriminate betweendifferent soft cheese varieties. Tryptophan and vitamin Aspectra were acquired on the cheese samples. The environ-ment of the tryptophan residues was found to be relativelymore hydrophilic for the ripened cheeses than for those atthe young stage. This phenomenon was attributed to partialproteolysis of caseins during ripening, resulting in anincrease of tryptophan exposure to the solvent. To test theaccuracy of FFFS in differentiating between the eight softcheeses, the authors applied FDA to the most relevant PCs,and good discrimination of cheeses was observed, withbetter results obtained with vitamin A spectra (96% and93% for the calibration and validation samples, respective-ly) than with tryptophan spectra (95% and 92% for thecalibration and validation samples, respectively). However,in their investigations, samples were studied at the center ofthe cheese, which could induce some misclassificationbetween the investigated cheeses. In the case of softcheeses, protein breakdown, lipolysis, pH, etc. differsignificantly between the surface and the center. To replywith this request, the matrix structure of three retailed softcheeses, produced each with different manufacturingprocess, was studied from the surface to the center of thecheese using FFFS, among other techniques (Karoui andDufour 2003). Cheese slices, 5-mm thick, were cut from thesurface to the center of the samples. PCA applied to thetryptophan fluorescence spectra recorded for each cheesevariety showed good discrimination of cheese samples as afunction of their location. The environment of tryptophanresidues was found to be more heterogeneous in the surfacesamples than in the center samples; this was attributed tothe changes in the extent and type of protein–proteininteractions in the protein network depending on thesampling zone. One of the limiting points of this studywas the low number of cheeses. In addition, onlytryptophan and vitamin A fluorescence spectra were studiedin this research. To reply with this request, later, Karoui etal. (2006e, 2007c) used FFFS to investigate changes at themolecular level of both the external (E) and central (C)zones of 15 ripened soft cheeses produced according totraditional and stabilized cheese-making procedures. TheCCSWA was applied to the tryptophan, vitamin A, andriboflavin spectral data sets, and the plane defined by thecommon components 1 (q1) and 3 (q3) showed cleardiscrimination between the cheese varieties and samplingzones. From the obtained results, it was concluded thatCCSWA allowed very efficient management of all thespectroscopic information collected on the investigated softcheeses. Each of the fluorophore provides information,which can be used for recognizing the cheese variety andsampling zone. The CCSWA method sums up thisinformation using two common components (q1 and q3),taking into account the relation between the different


fluorescence data sets. The result obtained from CCSWAdid not observed with the PCA performed separately oneach of fluorescence spectra, illustrating that CCSWAmethodology allowed efficiently the use of all the spectro-scopic information provided by the three intrinsic probes. Inanother study, Karoui et al. (2005a) attempted to discrim-inate 25 Gruyère and L’Etivaz Protected Designation ofOrigin cheeses. Emission spectra were scanned followingexcitation at 250 and 290 nm, and excitation spectra followingemission at 410 nm. By applying FDA, 100% correctclassification was obtained from the emission and excitationspectra, suggesting the use of FFFS as an accurate techniquefor the determination of the geographic origin of cheeses.These findings were fully supported on Emmental cheesesoriginating from different European countries and manufac-tured during both winter and summer seasons (Karoui et al.2004a, b, 2005b). One of the strong conclusions of thesestudies was that FFFS allows a good discrimination betweenEmmental cheeses produced from raw milk and those madewith thermized milk. This was supported by the accuracy ofFFFS to predict chemical parameters collected from the Eand C zones of 15 soft cheeses being assessed (Karoui et al.2006e). The PLS regression applied to the normalizedvitamin A fluorescence spectra provided the highest valuesof R2, since values of 0.88, 0.86, 0.86, and 0.84 were foundfor fat, dry matter, fat in dry matter, and water-solublenitrogen, respectively.

Finally, FFFS was used for monitoring the oxidation ofdairy products. Good estimation of the sensory analysesfrom FFF spectra of both sour cream (Wold et al. 2002) andcheese (Wold et al. 2005) has been achieved. Wold et al.(2005) showed that naturally occurring porphyrin andchlorophylls play an important role as photosensitizers indairy products. The degradation of these componentsshowed higher correlation with sensory measured lipidoxidation. Recently, the feasibility of the use of FFFS as anon-destructive technique for monitoring oxidation at themolecular level of semi-hard cheeses, made with cow’smilk and collected during both the grazing period (summer)and the stabling period (autumn), was examined at thesurface (20 mm from the rind) and the inner (40 mm fromthe rind) layers throughout the ripening stage—i.e., 2, 30,and 60 days old by Karoui et al. (2007b). By applying FDAto the 400–640-nm emission fluorescence spectra recordedat the surface layer, correct classification was observed for100% and 91.7% for the calibration and the validationspectra, respectively. With regard to the samples cut fromthe inner layers, the authors stated that the 400–640-nmemission fluorescence spectra failed to discriminate cheesesthat were either 2 or 30 days old. The main conclusion ofthis study was that throughout ripening the riboflavincomponent was affected primarily by oxygen and light(Marsh et al. 1994), while the physicochemical modifica-

tion that takes place during ripening seemed to presentlesser effect than did light and oxygen.

Meat and Meat Products

Research regarding the application of FFFS for the evalua-tion of meat products has focused on the measurements offluorescence from collagen, adipose tissues, and protein(Newman 1984; Jensen et al. 1989; Frencia et al. 2003).

The collagen in connective tissue is known to be animportant parameter of meat quality, as it is related to thetenderness and texture of the meat. Collagen exists in severaldifferent genetic forms, four of which have been found to bepresent in muscle: types I, III, IV, and V. Types I, III, and IVpresented similar fluorescent characteristics when they wereexcited in the 330–380-nm spectral region (Hildrum et al.2006). Elastin, another important fluorophore in meat,presents quite similar fluorescence properties to those ofcollagen types I, III, and IV (Egelandsdal et al. 2005).

Adipose tissue contains fluorescent molecules that arespecific for fat. Indeed, it has been shown by severalauthors (see, for example, Ramanujam 2000; Skjervold etal. 2003) that the fat-soluble vitamins A, D, and K exhibitfluorescence in the 387–480-nm spectral region afterexcitation at 308–340 nm. Swatland (Swatland 1987;Swatland et al. 1995a, b; Swatland and Findlay 1997)could be considered the pioneer in the field of meat, with aseries of papers on different aspects of the use of FFFS,starting in 1987. His work focused on measuring collagenand elastin fluorescence from the connective tissues inmeat. The author reported that after excitation at 365 nm,the fluorescence emission spectra of adipose tissueexhibited a maximum at 510 nm with a secondary plateauvarying from 430 to 450 nm (Swatland 1987). The obtainedfluorescence spectra of various meats were found to becorrelated with biochemical and sensory analyses such aschewiness (Swatland et al. 1995a), palatability (Swatland etal. 1995b), and toughness (Swatland and Findlay 1997).Most of the studies realized by Swatland and co-workerswere analyzed using univariate data analytical approach, bythe comparison of single wavelength or extracted fluores-cence peak features. Although interesting results wereachieved, the use of multivariate statistical analyses isneeded for curve resolution and useful quantitative meas-urements due to the complexity of fluorescence spectra.Egelandsdal et al. (1996) applied PCA and PLS regressionto the fluorescence spectra scanned on meat products afterexcitation at between 300 and 400 nm for studying isolatedperimysial sheets from a type I muscle and found a highcorrelation between perimysial breaking strength andfluorescence emission spectra recorded after excitation at335 nm. Wold et al. (1999a, b) later confirmed theseprevious investigations and suggested that it would be


possible to measure the amount of connective tissue inground meat by using an excitation wavelength of 380 nm.They have also showed that both connective tissue andintramuscular fat content could be measured using anexcitation wavelength of 332 nm. Egelandsdal et al.(2002) studied six different batches of beef longissimusdorsi samples originating from 151 animals by using FFFSand Warner–Bratzler (WB) peak values. By applying PLSregression, poor to good (R=0.45–0.84) correlations be-tween WB peak values and the emission spectra wereobtained. In addition, minor difference in predictability wasobserved using excitation wavelengths at 332 or 380 nm.The emission wavelengths containing the most relevantinformation about WB peak values were found in the 360–500-nm spectral range. Emission wavelengths around375 nm, following excitation at 332 nm, were found to berelated to a component in the perimysial tissue, most likelypresent in collagen I or III. In another study, FFFS wasexplored to predict the age of Parma hams produced fromtwo muscles (semi-membranosus and biceps femoris) andaged 3 months (young), 11 and 12 months (matured), and15 and 18 months (aged) by Møller et al. (2003). UsingPLS regression, prediction of the age was considered asgood, with a relative error of prediction of approximately1 month; also, a good correlation between fluorescence andchemical, sensory, and physical parameters was found.

Egelandsdal et al. (2005) tried to throw more light on thephenomena affecting the fluorescence signal and the abilityof the FFFS technique to quantify collagen contents. Beefmasseter and latissimus dorsi and pork glutens mediusmuscles, among others, were chosen for their wide differ-ences in color and connective tissue quality and content.PLS regression was applied to their fluorescence spectra inorder to predict collagen (measured as hydroxyproline), andgood results were obtained (root mean square error ofprediction (RMSEP) of 0.55%). Similar prediction resultswere obtained with complex sausage batters consisting ofdifferent kinds of muscles and presenting a large span inmyoglobin and realistic ranges in collagen and fat. One ofthe interesting results obtained in this study was that FFFSgave lower prediction errors for collagen content than didnear-infrared (NIR) reflectance when applied to the samebatters. Recently, Sahar et al. (2009a) tried to predict someparameters in meat with excitation set at 290 nm (emission305–400 nm) and the obtained results were not successfulsince only 53% and 55% of correct classification for proteinand cooking loss, respectively, in the validation data setswere obtained.

Differences in the level of collagen within a muscle orbetween different muscles led to a huge difference intenderness (Light et al. 1985). Using excitation wave-lengths set between 332 and 380 nm, FFFS found apromising technique for estimating tenderness in such

muscles (Hildrum et al. 2006). Recently, tryptophan fluo-rescence spectra scanned on two beef muscles (longissimusthoracis and infraspinatus) 2 and 14 days post mortemshowed a maximum located around 336 nm (Dufour andFrencia 2001). In addition, the maximum emission of agedmuscles showed a shift to higher wavelengths (red shift).The PCA performed on spectra showed good discriminationof samples according to the muscle type and aging.However, with only a limited number of meat samples, themodels suffered from over-fitting and consequently were notvery robust against the inclusion or exclusion of samples.Further analyses with more samples are necessary tosubstantiate these models. This would allow more variabilityof the chemical properties and thus development of generalmathematical models for better accuracy of the FFFStechnique. Frencia et al. (2003) therefore assessed thepotential of FFFS to discriminate between five muscle typespresenting different levels of collagen contents at two pointsin time (2 and 14 days post mortem). Applying FDA to thetryptophan fluorescence spectra, correct classification rate of82% was obtained. The authors concluded that FFFS is apowerful technique that allows a relatively good identifica-tion of muscle types according to maturation. Preliminaryresults have shown that results obtained at the molecularlevel by FFFS are related to the macroscopic levels (sensoryand rheology data sets). Indeed, by applying CCA to thefluorescence spectra (spectrofluorimeter with a front-facedevice or coupled to a fiber-optic) and mechanical properties(texturometer) or sensory attributes, a strong correlationbetween the different methods was found. The authorsconcluded that a common description of the samples waspossible from both the fluorescence and the rheology orsensory data, which was later confirmed by the investigationof Lebecque et al. (2003) on 25 longissimus thoracis samplesfrom animals presenting different ages (between 2.5 and8 years) and sexes. The authors depicted that the phenomenaobserved at the molecular and macroscopic levels wererelated to the changes in the texture of meat during aging.The obtained results were confirmed by Allais et al. (2004)on meat emulsion and frankfurter reporting that highcorrelation was obtained from fluorescence spectra andrheology methods.

Lipid oxidation is one of the factors limiting the qualityand acceptability of meat and meat products (Veberg 2006).As explained above, lipid oxidation can be determined byseveral methods, such as 2-thiobarbituric acid, sensoryanalysis, and dynamic headspace gas chromatographycombined with mass spectrometry. Veberg et al. (2006)used FFFS and other destructive methods to determine thelevel of lipid oxidation and explored the usefulness of thistechnique for detection of low levels of lipid oxidation inturkey meat stored during 9 months at −10 and −20 °C. Byapplying PLS regression between fluorescence spectra and the


values obtained by the other traditional techniques, goodcorrelations between fluorescence spectra and thiobarbituricacid reactive substances (TBARS), hexanal, 1-penten-3-ol,total components and sensory measured rancidity andintensity were found. One of the interesting conclusionsof this study was that FFFS could be considered as asensitive and non-destructive method for early lipidoxidation determination in turkey meat. Sahar et al.(2009b) confirmed the obtained results recently onchicken meat since FFFS was found to be able todetermine heterocyclic aromatic amines in grilled meat.The obtained results confirmed the previous findings ofOlsen et al. (2005) who monitored poultry meat freeze-stored in air at −20 °C for 26 weeks and stated that FFFSand gas chromatography coupled with mass spectrometrycould detect oxidative changes in pork back fat earlierthan the sensory panel and the electronic nose could.However, the correlation of fluorescence spectra withsensory analyses was found to be poorer and less than thatobserved between sensory analysis and gas chromatogra-phy with mass spectrometry. In a similar approach,Gatellier et al. (2007) monitored lipid oxidation of chickenmeat by using FFFS and TBARS over 9 days. In theirresearch study, three chicken genotypes representative ofFrench production were compared (i.e., Standard, Certi-fied, and Label). The samples were stored in darkness at4 °C, and a good correlation between emission spectrarecorded after excitation at 380 nm and TBARS wasfound. In addition, the genotype meat was found topresent an effect on the shape of fluorescence emissionspectra, since the fluorescence intensity of the emissionspectra of Certified and Label animals after 7 days ofrefrigerated storage was significantly higher than that ofStandard chicken meat samples. The obtained resultsconfirmed previous findings of Munck (2001) reportingthat fluorescence could be used in slaughtering and cuttingby robots.

Fish

The lipid fraction of marine fish has been shown to containa high level of polyunsaturated fatty acids. During storageand/or processing, the degradation of polyunsaturated fattyacids can lead to the development of primary and secondaryproducts, resulting in the formation of fluorescent com-pounds and the loss of essential nutriments. Indeed,Gardner (1979) has shown that relatively higher fluores-cence intensity was observed at longer wavelength maximaas the quality of fish decreased.

The 493/463- and 327/415-nm ratios have proved to be amore effective index of fish quality than other commonassessment methods. Hasegawa et al. (1992) used FFFS forthe quantitative assessment of oxidative deterioration in

freeze-dried fish. Two excitation wavelengths, i.e., 370 and450 nm exhibiting maximum emissions at 460 and 500 nm,respectively, were used. For fish samples stored at 25 °C inthe dark, an increase in the fluorescence intensity at 500 nmwas noted, while that at 460 nm remained unchanged. Thefluorophores observed at 460 nm have been attributed tothe reaction between reducing sugars (e.g., glucose andribose) and amino acid compounds inducing activation ofthe Maillard reaction, while that observed at 500 nm hasbeen attributed to the lipid oxidation products. In a similarapproach, Olsen et al. (2006) used an excitation wavelengthof 382 nm to record spectra in the 450–750-nm region onfour different batches of salmon pâté stored at 4 °C for 4, 8,and 13 weeks. Citric acid or calcium disodium ethylene-diamine tetraacetate was added as metal chelators to twobatches, whereas no chelator was added to the third batch.The three investigated batches contained oil, while a fourthone was made with the same amounts of ingredients butwithout any oil. The obtained results showed an increase inthe fluorescence intensity with increased storage time for allthe batches. In addition, the shape of the spectra changedlargely between samples containing oil and those withoutoil. Indeed, samples containing oil exhibited the highestintensity in the range of 470–475 nm, while those with noadded oil presented maxima between 440 and 450 nm. Byapplying PCA to the collection of spectral data, a cleardifference between samples according to the storage timewas observed; the largest variation in the data sets wasattributed to whether the sample contained oil or not; thestorage time was the second most important factor that ledto this discrimination. The authors applied PLS regressionto estimate the age of salmon pâté. The correlationcoefficients for the sensory attributes, dynamic head-spacedata, fluorescence spectra, and electronic nose sensorresponses were 0.64, 0.94, 0.93, and 0.70, respectively.The corresponding RMSEP were 3.8, 1.7, 1.8, and 3.5,respectively, illustrating that FFFS could be a suitabletechnique to measure lipid oxidation. This observation isconfirmed by the highest level of correlation found betweenfluorescence spectra and sensory attributes, among theother analytical techniques. Recently, FFFS has been usedto monitor fish freshness for four different species—cod,mackerel, salmon, and whiting fillets—at 1, 5, 8, and13 days of storage (Dufour et al. 2003). Emission spectra ofAAA+NA, tryptophan, and NADH were scanned afterexcitation at 250, 290, and 336 nm, respectively. The spectraof the first two excitationwavelengths showedmaxima locatedat 338 and 336 nm, respectively, while the last excitationwavelength showed two maxima located at 414 and 438 nm.For all three excitation wavelengths, the shape of the spectraillustrated some differences according to storage time, suggest-ing that a fluorescence spectrum may be considered as afingerprint. Applying FDA to the tryptophan fluorescence


spectra allowed 56% of samples to be correctly classified.Better classification was obtained fromAAA+NA andNADH,since 92% and 74% correct classification was observed,respectively. The authors concluded that AAA+NA fluores-cence spectra could be considered as fingerprints that mayallow discrimination between fresh and aged fish fillets.

Aubourg et al. (1998) used fluorescence spectroscopy tomonitor changes that occurs in sardines stored at −18 and−10 °C. Sardines stored at −18 °C were sampled after 0.5,2, 4, 8, 12, and 24 months, and those stored at −10 °C weresampled at 3, 10, 25, 60, and 120 days. Fluorescence wasmeasured at two excitation–emission maxima, 327/463 and393/463 nm. The authors used the fluorescence ratio,defined as the fluorescence intensity at 327/463 nm overthe fluorescence intensity at 393/463 nm, determined inaqueous and organic solutions (phases resulted from lipidextraction). This ratio was found to increase throughout thewhole storage time at the two temperatures when it wasdetermined in the aqueous solution. One of the mostlimiting points of this study was that the use of onlymaxima of emission and excitation wavelengths couldinduce some loss of information contained in the fluores-cence spectra. Recently, Karoui et al. (2006f) explored thepotentiality of FFFS to differentiate between fresh and frozen-thawed fish. A total of 24 fish (12 fresh and 12 frozen-thawed)were analyzed by using excitation wavelengths set at 290(tryptophan) and 340 nm (NADH). The emission spectra ofNADH of fresh fish showed a maximum at 455 nm and ashoulder at 403 nm, while frozen-thawed fish was character-ized by a maximum located at 379 nm and a shoulder at455 nm. By applying PCA to NADH spectra, gooddiscrimination between fresh and frozen fish samples wasobserved, which was confirmed by applying FDA to the firstfive PCs of the PCA performed on the NADH spectra. Indeed,100% correct classification was obtained for the calibrationand validation data sets, respectively. One of the interestingconclusions of this research was that NADH fluorescencespectra may be considered a promising tool for differentiatingbetween fresh and frozen-thawed fish samples.

Eggs and Egg Products

In order to ensure high and consistent egg quality, anattractive and alternative strategy for determining the state ofegg freshness can be achieved by sensor technologies. Thesetechniques (such as NIR, MIR, fluorescence spectroscopies,etc.) appear to be promising tools for non-destructivelydetermining egg freshness. Such methods cannot eliminatethe need for more detailed physicochemical analyses, butthey may help to screen for samples that require furtherexamination. Freshness makes a major contribution to thequality of eggs and egg products. One of the main concernsof the egg industry is the systematic determination of egg

freshness because consumers may perceive variability infreshness as lack of quality.

The changes that occur in eggs during storage are manyand complex and affect the functional properties of eggyolk and egg albumen. These changes include thinning ofalbumen, increase of pH, weakening and stretching of thevitelline membrane, and increase in the water content of theyolk. Freshness can be explained to some extent byobjective sensory, (bio)chemical, microbial, and physicalparameters and can therefore be defined as an objectiveattribute. Knowledge of the various descriptors of proper-ties that are encountered in eggs immediately after layingmust be known, as well as the changes in properties thattake place over time. This information can be gained byperforming controlled storage experiments that extend fromthe time after laying; loss in freshness and spoilage can thusbe monitored. Posudin (1998) assessed the potential ofFFFS to determine egg freshness by using ultravioletradiation for the quality evaluation of eggs with differinglevels of pigmentation. The emission spectra of differenteggs showed two maxima located at 635 and 672 nm(ascribed to the pigments of porphyrin and porphyrinderivatives of florin and oxoflorin) after excitation at 405,510, 540, and 557 nm. The intensity at 672 nm depends onthe egg freshness. The autofluorescence of fresh egg isstronger than that of old one, since the intensity ofautofluorescence depends on the amount of porphyrin onthe shell surface. From these preliminary results, the authorconcluded that fluorescence spectroscopy could be apromising approach for quantitative estimation of porphyrinin eggs and thus determine egg freshness. Recently, FFFSwas used to monitor egg freshness during storage (Karouiet al. 2006c, d). The authors found that FMRP (excitation360 nm; emission 380–580 nm) recorded on thick and thinalbumens and vitamin A scanned on egg yolk (emission410 nm; excitation 270–350 nm) could be considered aspowerful tools for the evaluation of egg freshness stored atroom temperature, while tryptophan fluorescence spectrarecorded on thick and thin albumens and egg yolk failed todiscriminate between fresh and aged eggs. Using excitationat 360 nm, the emission spectra recorded on fresh thick eggalbumen exhibited two maxima located at 410 and 440 nm,respectively. Similar results were obtained on thin albumenof fresh eggs. The very characteristic fluorescence spectra ofthick and thin albumen of eggs stored for a long time (i.e.,18 days or more) at room temperature showed a shoulderlocated at 414 nm and a maximum at approximately 438 nm.In addition, as the spectra showed large differences betweenfresh thin/thick egg albumens and those stored for a longtime (29 days), the authors considered the spectra asfingerprints for freshness identification. Indeed, thick andthin albumens of fresh eggs within 2–3 days of laying had thehighest intensity at 410 nm, while aged eggs had the lowest


one. The authors concluded that the shape of FMRP iscorrelated with storage time: thick albumen of fresh eggshad the lowest ratio of fluorescence intensity FI440 nm/FI410 nm (i.e., 1.0), while that of eggs stored for 29 dayshad the highest (i.e., 1.30). The changes in the FI440 nm/FI410 nm ratio has been ascribed to the change in theviscosity of both thick and thin egg albumens and theformation of furosine during storage (Birlouez-Aragon etal. 1998; Kulmyrzaev and Dufour 2002). Indeed, it hasbeen reported that during egg storage, a decrease in theviscosity of thick albumen was observed (Lucisano et al.1996). This phenomenon has been attributed to theseparation of the β-fraction of ovomucin, rich in carbo-hydrate, and from the ovomucin–lysozyme complex.However, in the study by Karoui et al. (2006c, d), eggswere stored at room temperature, and little attention hasbeen given to the influence of temperature and relativehumidity variations on fluorescence measurements, al-though Stadelman et al. (1954) observed a linear decreaseof −1.15 Haugh Units per 10 °C increase in testingtemperature. Therefore, Karoui and co-worker havecontinued this work by investigating changes at themolecular level of 126 eggs stored at 12.2 °C and 87%RH for 1, 6, 8, 12, 15, 20, 22, 26, 29, 33, 40, 47, and55 days (Karoui et al. 2007d). Of the intrinsic fluoro-phores tested, only PCA applied to the vitamin Afluorescence spectra allowed good identification of eggsas a function of their storage time. By applying FDA tothe AAA+NA spectra, correct classification rates of 69.4%and 63.9% were observed for the calibration and valida-tion sets, respectively. Quite similar results were obtainedwith AAA+NA scanned on egg yolks. The best resultswere obtained with vitamin A fluorescence spectra, wherecorrect classification rates of 97.7% and 85.7% in thecalibration and validation sets were obtained, respectively.The authors concluded that vitamin A fluorescence spectraprovide useful fingerprints allowing the identification of eggsduring storage at low temperature and could be considered asa powerful intrinsic probe for the evaluation of eggfreshness. Karoui and co-workers have continued this workby testing the ability of vitamin A fluorescence spectra tomonitor changes at the molecular level of 225 eggs stored at12.2 °C and 87% RH in an atmosphere containing 2% (n=108) and 4.6% (n=99) of CO2 for 55 days (Karoui et al.2007e). Again, vitamin A fluorescence spectra allowed gooddiscrimination of eggs according to both storage time andconditions, while more overlapping between egg sampleswas observed when the other intrinsic probes were investi-gated: eggs aged 22 days or less were separated from thoseaged 26 days or more. In addition, eggs were found to bewell separated for each storage time, except for thosesamples aged 20 and 22 days and those aged 26 and 29 days,where some overlapping was observed.

Edible Oils

Olive oil is an economically important product of Mediter-ranean countries. It has a fine aroma and pleasant taste, withexcellent health benefits. The quality of olive oil rangesfrom the high-quality extra-virgin olive oil (EVOO) to thelow-quality olive-pomace oil. EVOO is obtained from thefruit of the olive tree by mechanical pressing and withoutrefining processes. Owing to its high quality, it is the mostexpensive type of olive oil. For this reason, it may bemislabeled or adulterated for economic reasons. Mislabel-ing often involves false information regarding the geo-graphic origin or oil variety (Aparicio et al. 1997).Adulteration involves the addition of cheaper oils; the mostcommon adulterants found in virgin olive oil are refinedolive oil, residue oil, synthetic olive oil–glycerol products,seed oils, and nut oils (Baeten et al. 1996; Downey et al.2002; Sayago et al. 2004). Owing to the low price of olive-pomace oil, it is sometimes used to adulterate EVOO. Forthis reason, a rapid method to detect such a practice isimportant for quality control and labeling purposes. In thiscontext, Zandomeneghi et al. (2005) recorded fluorescencespectra of EVOO using right angle and FFFS. The formermethod showed considerable artifacts and deformation,while the latter provided spectra that are much less affectedby self-absorption. The authors attributed this to the self-absorption phenomena when using right-angle fluores-cence, even when the spectra are corrected for inner filtereffects. In another study, Sayago et al. (2004) appliedfluorescence spectroscopy for detecting hazelnut oil adul-teration in virgin olive oils. Virgin olive, virgin hazelnut,and refined hazelnut oil samples and a mixture of them at5%, 10%, 15%, 20%, 25%, and 30% adulteration wereanalyzed after excitation at 350 nm. By performing lineardiscriminant analysis (LDA), 100% correct classificationwas achieved. In a similar approach, Kyriakidis andSkarkalis (2000) used excitation wavelength of 360 nm todifferentiate between common vegetable oils, including oliveoil, olive residual oil, refined olive oil, corn oil, soybean oil,sunflower oil, and cotton oil. All the oils studied showed astrong fluorescence band at 430–450 nm, except for virginolive oil, which exhibited a low intensity at both 440 and455 nm, a medium band around 681 nm and a strong one at525 nm. The latter two bands have been ascribed tochlorophyll and vitamin E compounds, respectively. Thevery low intensity of the peaks at 445 and 475 nm is due tothe high content of phenolic antioxidants, which providemore stability against oxidation. All refined oils showed onlyone intense peak at 445 nm, which is due to fatty acidoxidation products formed as a result of the large percentageof polyunsaturated fatty acids present in these oils.

Oil (and food systems in general) could be considered asa complex system, which presents a set of different


properties and contains many fluorescent molecules. Theuse of only excitation or emission wavelengths could limitthe ability of this technique to determine the quality of oilsamples. To comply with this requirement, the variation inthe excitation and emission wavelengths allows simulta-neous determination of compounds present in oils. Thiscould be realized by using synchronous fluorescencespectroscopy. In this context, Sikorska et al. (2004, 2005)used synchronous fluorescence spectroscopy with excita-tion wavelengths from 250 to 450 nm and emission spectrain the range 290 to 700 nm. The peak located at 320 nmafter excitation at 290 nm has been attributed to tocopher-ols, while the band located at 670 nm in emission and405 nm in excitation belongs to pigments of the chlorophyllgroup. In order to compare the set of synchronousfluorescence spectra of different oils, Sikorska et al.(2005) applied the k-nearest neighbors method, and gooddiscrimination between oil samples with a very lowclassification error ranging between 1% and 2% and a lowstandard deviation value was obtained. Guimet et al. (2004)explored the use of FFFS to discriminate between virginand pure olive oils; the ranges studied were at excitationwavelength λex=300–400 nm and emission wavelength(λem) 400–695 nm and λex=300–400 nm and λem 400–600 nm. The first range was found to contain chlorophylls,whereas the second range contained only the fluorescencespectra of the remaining compounds (oxidation productsand vitamin E). Later, Guimet et al. (2005) appliedPARAFAC to detect the adulteration of EVOO with olive-pomace oil at low levels (5%). Discrimination between thetwo types of oils was achieved by applying both LDA anddiscriminant multi-way PLS regression; the latter methodgave 100% correct classification. The same technique(synchronous fluorescence) was used to analyze 73 sam-ples, including 41 edible and 32 lampante virgin olive oilscollected in October and November 2002 (Poulli et al.2005). PCA and hierarchical cluster analysis applied to theemission spectra in the range 350–720 nm (at excitationwavelengths varying from 320 to 535 nm) showed goodseparation between the two types of oils. Recently, the sameresearch group assessed the potential of synchronousfluorescence spectra to detect adulteration of virgin oliveoil (VOO) with other oils (Poulli et al. 2007). By applyingPLS regression to the excitation spectra recorded 250–720 nm with a wavelength interval of 20 nm, the authorsstated that FFFS could be useful for the detection of olive-pomace, corn, sunflower, soybean, rapeseed, and walnutoils in VOO at levels of 2.6%, 3.8%, 4.3%, 4.2%, 3.6%,and 13.8%, respectively. Frying oil deterioration has alsobeen measured by using five selected excitation wave-lengths varying from 395 to 530 nm (Engelsen 1997). Byapplying PLS regression, the author showed good correla-tion between fluorescence spectra and quality parameters

describing the deterioration (e.g., anisidine value, iodinevalue, oligomers, and vitamin E).

Cereals and Cereal Products

The potential of fluorescence spectroscopy for monitoringcereals has increased over the past few years with thepropagated application of chemometric tools and withtechnical and optical developments of the spectrofluorim-eter. Zandomeneghi (1999) used FFFS (excitation 275 nm;emission 280–575 nm) to differentiate between differentcereal flours (i.e., rice, creso, maize, pandas). The sameresearch group also utilized visible excitation set at 445 nm(emission 460–600 nm) to differentiate between flours offive different wheat varieties and a good discrimination wasobserved. In another study, excitation wavelengths set at275, 350, and 450 nm presenting fluorescence emissionmaxima at 335, 420, and 520 nm, respectively, wereutilized to classify botanical tissue components of complexwheat flour and rye flour; the bands were attributed toAAA, ferulic acid, and riboflavin components, respective-ly. The last fluorescent component was confirmed byZandomeneghi et al. (2003), who attributed the bandobserved at 520 nm to riboflavin, while the band between430 and 530 nm was found to be proportional to the luteincontent of the flour (Zandomeneghi et al. 2000).

Ferulic acid and riboflavin spectra have been reported tohave good accuracy for monitoring wheat flour refinementand milling efficiency following the use of fluorescenceimaging. Successful classification was obtained, suggestingthat FFFS may be used to classify wheat cultivars (Symonsand Dexter 1991, 1992, 1993, 1994). These results haverecently been confirmed by Karoui et al. (2006a) wheretryptophan fluorescence spectra of 59 samples (20 completeKamut®, semi-complete Kamut®, and soft wheat flours, 28pasta and 11 semolinas manufactured from complete Kamut®,semi-complete Kamut®, and hard wheat flours) were scannedafter excitation at 290 nm. PCA performed on the flours’spectra clearly differentiated complete Kamut® and semi-complete Kamut® samples from those produced fromcomplete and semi-complete soft wheat flours, while gooddiscrimination of pasta samples manufactured from completeKamut® and complete hard wheat flours from those madewith semi-complete Kamut® and semi-complete hard wheatflours was achieved. The best discrimination was obtainedwith tryptophan spectra recorded on semolinas, since the fourgroups were well discriminated. Indeed, by applying FDA tothe spectral collection, 86.7% and 87.9% correct classificationrates were obtained for the calibration and validation samples,respectively. In a similar approach, emission spectra (370–570 nm) were recorded after excitation at 350 nm on red andwhite wheat kernels and a clear difference was observedbetween the two group samples (Ram et al. 2004); this


difference has been attributed to the morphologicalvariation in the pericarp and nuclear organization of thetwo varieties of wheat.

Sugar

In combination with multivariate statistical analyses, fluo-rescence spectroscopy has proved to be a promisingscreening method for predicting quality parameters in beetsugar samples (Munck et al. 1998). Indeed, it has beenshown that commercial sugars exhibit characteristic fluo-rescence, which can be used to obtain informationregarding minor constituents in the sugar. Fluorescencehas successfully been applied to the beet sugar manufac-turing process with the use of multivariate data analysis(Munck et al. 1998). The same approach with multipleexcitation and emission wavelengths used by Carpenter andWall (1972) has also been employed, and interesting resultswere obtained. In a study of beet sugar samples, it waspossible to classify white sugar samples according tofactory and to predict quality parameters such as aminonitrogen, color, and ash from the fluorescence data of thesesamples (Nørgaard 1995). The fluorescence data of thickjuice samples showed more ambiguous results owing to themore complex sample composition. Another study of beetsugar samples utilized the three-dimensional structure ofthe fluorescence excitation–emission landscapes to resolvespectral excitation and emission profiles of fluorophorespresent in sugar with a multi-way chemometric model,PARAFAC (Bro 1999). Four fluorescent components werefound to capture the variation in the fluorescence data of268 sugar samples collected from a beet sugar factory in asingle campaign, and two of them showed spectra with aclose similarity to the pure fluorescence spectra of theamino acids tyrosine and tryptophan. The concentrations ofthe four components estimated from the sugar samplescould be correlated with several quality and processparameters, and they were characterized as potentialindicator substances of the chemistry in the sugar process,which has been confirmed by the use of HPLC analysiscombined with fluorescence detection on thick juicesamples and evaluation by PARAFAC (Baunsgaard et al.2000a). Seven fluorophores were resolved from thick juice.Apart from tyrosine and tryptophan, four of the fluoro-phores were identified as high molecular weight com-pounds, which were related to colorants absorbing at420 nm. Three of the high molecular weight compoundswere found to be possible Maillard reaction polymers. Thelast of the seven fluorophores indicated a compound withpolyphenolic characteristics. In a fluorescence study of 47raw cane sugars collected from many different locationsand campaign years, three individual fluorophores werefound; one of them, representing maximum excitation and

emission at 275 and 350 nm, respectively, was character-ized as an ultraviolet color precursor that participates incolor development during storage. The other two (340,420 nm and 390, 460 nm excitation/emission in the visiblewavelength area) are considered to be potential colorants,which shows a link with their fluorescence behavior(Baunsgaard et al. 2000b). Recently, FFFS has been usedto monitor adulteration of honey with cane sugar syrup(Ghosh et al. 2005). Using an excitation wavelength of340 nm, pure honey samples were characterized by twoprominent features—a shoulder located at around 440 nmand a maximum located at 510 nm, which has beenascribed to flavins—while cane sugar syrup samplesexhibited a maximum located around 430 nm. The peakslocated at 440 and 430 nm in pure honey and sugar syrupsamples have been attributed to NADH. Synchronousfluorescence was then applied to differentiate between purehoney and sugar syrup samples, and good discriminationbetween the two groups was observed. The spectra for canesugar syrup were characterized by a shoulder around305 nm and a prominent band around 365 nm, while honeysamples had a strong peak around 460 nm and a muchweaker peak around 365 nm. The authors observed anincrease in the intensity at 365 and 425 nm, as well as theratio of FI365/FI425, with the increase of cane sugar syrupconcentration; thus the ratio of FI365/FI425 has beensuggested as a potential method to monitor adulteration ofhoney with cane sugar syrup. In another study, fluorescencespectra were scanned on 62 honey samples belonging toseven floral origins after excitation at 250 nm (emission280–480 nm), 290 nm (emission 305–500 nm), and 373 nm(emission 380–600 nm) and emission set at 450 nm(excitation 290–440 nm) by Ruoff et al. (2005) and Karouiet al. (2007a). By applying FDA to the four data sets(concatenation), correct classification rates of 100% and90% were observed for the calibration and validationsamples, respectively. In addition, the seven honey typeswere well discriminated, indicating that the molecularenvironments, and thus the physicochemical properties, ofthe investigated honeys were different. One of the mainfindings of these studies is that FFFS might be a suitableand alternative technique to classify honey samples accord-ing to their botanical origins; this was confirmed recentlyby Ruoff et al. (2006), who studied 371 honey samplesoriginating from Switzerland, Germany, Italy, Spain,France, Slovenia, and Denmark. By using chemometrictools, the error rates of the discriminant models ranged from0.1% to 7.5%.

Fruit and Vegetables

Chlorophyll fluorescence has been used as an intrinsicprobe to determine the physiological status of whole plants


and plant organs (Song et al. 1997). This component hasbeen considered an efficient probe for monitoring applesduring maturation, ripening, and senescence (Song et al.1997). Recently, fluorescence spectroscopy has been consid-ered to have the potential for assessing the mealiness ofapples (Moshou et al. 2005), since relatively good correlationwas obtained between mealiness and fluorescence spectra.Other authors have used chlorophyll fluorescence of applesas a potential predictor of superficial scald developmentduring storage (De Ell et al. 1996) and for the estimation ofanthocyanins and total flavonoids in apples (Hagen et al.2006). In another study, Lötze et al. (2006) used fluorescenceimaging as a non-destructive method for the pre-harvestdetection of bitter pit in apples; the same technique hadalready been utilized to determine apple juice quality (Seidenet al. 1996; Noh and Lu 2007). Two excitation wavelengths,set at 265 and 315 nm, were chosen as they yielded therichest spectra of two juice-apple varieties (Jonagold andElstar). The spectra showed two excitation–emission maxima(315/440 and 265/350 nm) that have been not attributed toany component in apple juice (Seiden et al. 1996). ApplyingPCA to the two juice-apple varieties, good discriminationwas observed—which was not achieved with titratableacidity or soluble solids data. The authors pointed out thatan increase in the ripening process of apples involves anincrease in the soluble fluorescent compounds. Goodcorrelation between soluble solids and fluorescence spectrawas observed independent of the apple varieties, indicatingthe possibility of modeling the progression in maturity withinformation obtained from spectra, while fluorescence wasfound to correlate poorly to the amount of titratable acids injuice.

Identification of Bacteria of Agro-alimentary Interest

The identification of microorganisms of agro-alimentaryinterest in food and food products by conventionalphenotypic procedures based on morphology and biochem-ical tests involves a large quantity of reagents and, in somecases, is unable to discriminate microorganisms at the strainlevel. In this context, Leblanc and Dufour (2002) assessedthe potential of different intrinsic probes (i.e., tryptophan,AAA+NA, and NADH) to discriminate between 25 strainsof bacteria in dilute suspensions. The best results wereobtained by using AAA+NA spectra, where correctclassification rates of 100% and 81% were observed forthe calibration and validation samples, respectively. Theauthors noted that fluorescence spectroscopy is able todiscriminate and identify bacteria at genus, species, andstrain levels. This assumption was later demonstrated by thesame research group (Leblanc and Dufour 2004). In theirstudies, the spectra of three bacteria strains (Lactococcuslactis, Staphylococcus carnosus, and Escherichia coli) were

recorded at different growth phases. By applying PCA tothe spectra scanned on each bacteria, three groupscorresponding to three main phases of growth wereidentified (lag phase, exponential phase, and stationaryphase). The authors then gathered the spectra recorded onthe three bacteria into one matrix, and this new matrix wasanalyzed by PCA. The obtained results showed gooddiscrimination of spectra according to bacteria and meta-bolic profile. Recently, Ammor et al. (2004) utilized thesame technique for the identification of lactic acid bacteriaisolated from a small-scale facility producing traditional drysausages. Again, fluorescence spectroscopy demonstratedits ability to discriminate between Lactobacillus sakeisubsp. carnosus and Lactobacillus sakei subsp. sakei. Inanother approach, Leriche et al. (2004) isolated 30Pseudomonas spp. strains from milk, water, cheese center,and cheese surface belonging to two traditional workshopsmanufacturing raw milk Saint Nectaire cheese. By applyingFDA to the data sets, clear linkages between groups ofisolates were noted. In the first workshop, the milk wasimplicated being as the sole source of cheese contamina-tion, whereas in the second workshop the milk and cheesecenter isolates were found to be similar, but different fromsurface cheese isolates. The authors attributed this contam-ination at the cheese surface to the water used during theripening process (washing of the cheese surface). From theresults obtained, it was stated that it is possible tocharacterize, differentiate, and trace Pseudomonas spp.strains using the fluorescence technique. These findingswere reinforced by the high correlation (using CCA)observed between the data sets obtained from the metabolicprofiling and fluorescence spectroscopy. The obtainedresults were fully supported by the same research group(Tourkya et al. 2009) since good discrimination, even forstrains for which ambiguity still remained after PCR andAnalytical Profile Index 20 NE identification.

Saito (2009) depicted that Napa cabbage (Brassica rapaL.) laser-induced fluorescence (LIF) spectra of a normalcore and a rotten core show greater intensities in the greenwavelength range of the LIF spectrum of the rotten core incomparison to that of the normal core. The integrated peakarea between 450 and 600 nm for the rotten core was morethan twice that for the normal core. Regarding pear (Pyruscommunis L.), differences were found in the LIF spectrafrom the surface of a young pear and a ripe pear. Theintensity of the blue region (400–500 nm) of the LIFspectrum in the ripe pear surface was increased incomparison to the pear picked at harvest time in a stillunripe stage. In another approach, (Kim et al. 2003)developed a LIF imaging food system (i.e., meat pork andapple). Multispectral fluorescence emission images wererecorded after excitation set at 355 nm and the systemfluorescence emission images were captured in the blue,


green, red, and far-red regions of the spectrum centered at450, 550, 678, and 730 nm, respectively, from a 30-cmdiameter target area in ambient light. Images of apples andof pork meat artificially contaminated with diluted animalfeces with excitation at 355 nm and emission at 450, 550,678, and 730 nm have also demonstrated the versatility offluorescence imaging techniques for potential applicationsin food safety inspection. The most interesting conclusionwas that regions of contamination that were not readilyvisible to the human eye were found to be easily identifiedfrom the fluorescence images.

Conclusions

As illustrated in the present review, the environment ofintrinsic fluorophores recorded on intact food systemscontains valuable information regarding the compositionand nutritional values of food products. The huge potentialfor the application of fluorescence spectroscopy combinedwith multivariate statistical analyses for the evaluation offood quality has also been demonstrated; the great differencebetween food systems has been related to the differences inthe molecular structure of the samples resulting in a variationof the optical pathway of excitation light and fluorescenceinside the optically complex natural food systems. Themethod is suitable as an effective research tool and can be apart of evaluation procedure for food quality.

Calibration stage and development of the calibrationequations are the limiting steps for adopting the fluores-cence as a technique for the determination of the qualityand authenticate food products, as it is time-consuming andcostly procedure. However, when the calibration stage isaccomplished successfully, the determination of a chemicalproperty or the geographical origin can be carried out veryrapidly with a single analysis for minimal cost.

This development could make the FFFS a powerful toolfor its use for on-line process control. In this context, FFFSsensors would give valuable information in comparison tothe NIR since the information given by the latter are basedon molecular overtone vibrations, which are less sensitiveand specific. More recent improvement in the fluorescenceinstrument passes through the development of spectrome-ters that allows on-line measurements. The main innovationin this area is the use of fiber optics for the connectionbetween the spectrometer and the sensing device. Eventhough the review focuses on examples from the foodindustry, the principles are broader and fluorescence couldbe applied to other fields (pharmaceutical, biotechnology,etc.). It is therefore expected that in the coming years, FFFScombined with chemometric tools would be a reliable toolfor understanding the bases of food molecular structure and,as a consequence, for their qualities.

References

Allais, I., Viaud, C., Pierre, A., & Dufour, E. (2004). A rapid methodbased on front-face fluorescence spectroscopy for the monitoringof the texture of meat emulsions and frankfurters. Meat Science,67, 219–229.

Ammor, S., Yaakoubi, K., Chevallier, I., & Dufour, E. (2004).Identification by fluorescence spectroscopy of lactic acid bacteriaisolated from a small-scale facility producing traditional drysausages. Journal of Microbiological Methods, 59, 271–281.

Aparicio, R., Morales, M. T., & Alonso, V. (1997). Authentication ofEuropean virgin olive oils by their chemical compounds, sensoryattributes, and consumers’ attitudes. Journal of Agricultural andFood Chemistry, 111, 3–10.

Aubourg, S. P., Sotelo, C. G., & Pérez-Martín, R. (1998). Assessmentof quality changes in frozen sardine (sardine pilchardus) byfluorescence detection. Journal of the American Oil ChemistrySociety, 75, 575–580.

Baeten, V., Meurens, M. T., Morales, R., & Aparicio, R. (1996).Detection of virgin olive oil adulteration by Fourier transformRaman spectroscopy. Journal of Agricultural and Food Chem-istry, 44, 2225–2230.

Baunsgaard, D., Andersson, C. A., Arndal, A., & Munck, L. (2000).Multiway chemometrics for mathematical separation of fluores-cent colorants and colour precursors from spectrofluorometry ofbeet sugar and beet sugar thick juice as validated by HPLCanalysis. Food Chemistry, 70, 113–121.

Baunsgaard, D., Nørgaard, L., & Godshall, M. (2000). Fluorescenceof raw cane sugars evaluated by chemometrics. Journal ofAgricultural and Food Chemistry, 48, 4955–4962.

Birlouez-Aragon, I., Nicolas, M., Metais, A., Marchond, N., Grenier,G., & Calvo, D. A. (1998). A rapid fluorimetric method toestimate the heat treatment of liquid milk. International DairyJournal, 8, 771–777.

Bosset, J. O., Jeangros, B., Berger, Th, Bütikofer, U., Collomb, M.,Gauch, R., et al. (1999). Comparison de fromages à pâte dure detype Gruyère produits en région de montagne et de plaine. RevueSuisse d’Agriculture, 31, 17–22.

Boubellouta, T., & Dufour, E. (2008). Effects of mild heating andacidification on the molecular structure of milk components asinvestigated by synchronous front-face fluorescence spectroscopycoupled with parallel factor analysis. Applied Spectroscopy, 62,490–496.

Boubellouta, T., & Dufour, E. (2010). Cheese-matrix characteristicsduring heating and cheese melting temperature prediction bysynchronous fluorescence and mid-infrared spectroscopies. Foodand Bioprocess Technology. doi:10.1007/s11947-010-0337-1.

Boubellouta, T., Galtier, V., & Dufour, E. (2009). Effects of addedminerals (calcium, phosphate, and citrate) on the molecularstructure of skim milk as investigated by mid-infrared andsynchronous fluorescence spectroscopies coupled with chemo-metrics. Applied Spectroscopy, 63, 1134–1141.

Bro, R. (1999). Exploratory study of sugar production usingfluorescence spectroscopy and multi-way analysis. Chemometricsand Intelligent Laboratory Systems, 46, 133–147.

Buschmann, C. (2007). Variability and application of the chlorophyllfluorescence emission ratio red/far-red of leaves. PhotosynthesisResearch, 92, 261–271.

Carpenter, F. G., & Wall, J. H. (1972). Fluorescence in commercialsugar. In F. G. Carpenter & J. H. Wall (Eds.), Proceedings of the1972 technical session on cane sugar refining research (pp. 47–61). New Orleans: USDA-ARS.

Christensen, J., Nøgaard, L., Bro, R., & Engelsen, S. B. (2006).Multivariate autofluorescence of intact food systems. ChemicalReviews, 106, 1979–1994.


http://dx.doi.org/10.1007/s11947-010-0337-1

Coulon, J. B., & Priolo, A. (2002). La qualité sensorielle des produitslaitiers et de la viande dépend des fourrages consommés par lesanimaux. INRA Production Animales, 15, 333–342.

De Ell, J. R., Prange, R. K., & Murr, D. P. (1996). Chlorophyllfluorescence of delicious apples at harvest as a potential predictorof superficial scald development during storage. PostharvestBiology and Technology, 9, 1–6.

Downey, G., McIntyre, P., & Davies, A. N. (2002). Detecting andquantifying sunflower oil adulteration in extra virgin olive oilsfrom the eastern Mediterranean by visible and near-infraredspectroscopy. Journal of Agricultural and Food chemistry, 50,5520–5525.

Dufour, E., & Riaublanc, A. (1997). Potentiality of spectroscopicmethods for the characterisation of dairy products, I—Front-facefluorescence study of raw, heated and homogenised milks. LeLait, 77, 657–670.

Dufour, E., & Frencia, J. P. (2001). Les spectres de fluorescencefrontale: une empreinte digitale de la viande. Viandes et ProduitsCarnés, 22, 9–14.

Dufour, E., Mazerolles, G., Devaux, M. F., Duboz, G., Duployer, M.H., & Mouhous Riou, N. (2000). Phase transition of triglyceridesduring semi-hard cheese ripening. International Dairy Journal,10, 87–99.

Dufour, E., Frencia, J. P., & Kane, E. (2003). Development of a rapidmethod based on front-face fluorescence spectroscopy for themonitoring of fish freshness. Food Research International, 36,415–423.

Egelandsdal, B., Kvaal, K., & Isaksson, T. (1996). Autofluorescencespectra as related to tensile properties for perimysium frombovine masseter. Journal of Food Science, 61, 342–347.

Egelandsdal, B., Wold, J. P., Sponnich, A., Neegaård, S., & Hildrum,K. I. (2002). On attempts to measure the tenderness of Long-issimus Dorsi muscles using fluorescence emission spectra. MeatScience, 60, 187–202.

Egelandsdal, B., Dingstad, G., Tøgersen, G., Lundby, F., & Langsrud,Ø. (2005). Autofluorescence quantifies collagen in suasagebatters with a large variation in myoglobin content. Meat Science,69, 35–46.

Engelsen, S. A. (1997). Explorative spectrometric evaluations offrying oil deterioration. Journal of the American Oil ChemistrySociety, 74, 1495–1508.

Feinberg, M., Dupont, D., Efstathiou, T., Louâore, V., & Guyonnet, J.P. (2006). Evaluation of tracers for the authentification of thermaltreatments of milks. Food Chemistry, 98, 188–194.

Frencia, J. P., Thomas, E., & Dufour, E. (2003). Measure of meattenderness using front face fluorescence spectroscopy. Sciencedes Aliments, 23, 142–145.

Gardner, H. W. (1979). Lipid hydroperoxide reactivity with proteinsand amino acids: A review. Journal of Agricultural and FoodChemistry, 27, 220–229.

Gatellier, P., Gomez, S., Gigaud, V., Berri, C., Le Bihan-Duval, E., &Santé-Lhoutellier, V. (2007). Use of fluorescence front facetechnique for measurement of lipid oxidation during refrigeratedstorage of chicken meat. Meat Science, 76, 543–547.

Ghosh, N., Verma, Y., Majudmer, S. K., & Gupta, P. K. (2005). Afluorescence spectroscopic study of honey and cane sugar syrup.Food Science and Technology Research, 11, 59–62.

Guilbaut, G. G. (1989). Principles of fluorescence spectroscopy in theassay of food products. In L. Munck & A. Francisco (Eds.),Fluorescence analysis in foods (pp. 33–58). Copenhagen: Long-man Group.

Guimet, F., Ferré, J., Boqué, R., & Rius, F. X. (2004). Application ofunfold principal component analysis and parallel factor analysisto the exploratory analysis of olive oils by means of excitation–emission matrix fluorescence spectroscopy. Analytica ChimicaActa, 515, 75–85.

Guimet, F., Ferré, J., & Boqué, R. (2005). Rapid detection of olive-pomace oil adulteration in extra virgin olive oils from theprotected denomination of origin « Siurana » using excitation–emission fluorescence spectroscopy and three-way methods foranalysis. Analytica Chimica Acta, 544, 143–152.

Hagen, S. F., Solhaug, K. A., Bengtsson, G. B., Borge, G. I. A., & Bilger,W. (2006). Chlorophyll fluorescence as a tool for non-destructiveestimation of anthocyanins and total flavonoids in apples. Posthar-vest Biology and Technology, 41, 156–163.

Hammami, M., Rouissi, H., Salah, N., Selmi, H., Al-Otaibi, M.,Blecker, C., & Karoui, R. (2010). Fluorescence spectroscopycoupled with factorial discriminant analysis technique to identifysheep milk from different feeding systems. Food Chemistry.doi:10.1016/j.foodchem.2010.03.107.

Hasegawa, K., Endo, Y., & Fujimoto, K. (1992). Oxidative deterio-ration in dried fish model systems assessed by solid samplefluorescence spectrophotometry. Journal of Food Science, 57,1123–1126.

Herbert, S. (1999). Caractérisation de la structure moléculaire etmicroscopique de fromages à pâte molle. Analyse multivariée desdonnées structurales en relation avec la texture. PhD Thesis,Ecole Doctorale Chimie Biologie de l’Université de Nantes,Nantes, France.

Herbert, S., Riaublanc, A., Bouchet, B., Gallant, D. J., & Dufour, E.(1999). Fluorescence spectroscopy investigations of acid- andrennet-induced milk coagulation of milk. Journal of DairyScience, 82, 2056–2062.

Herbert, S., Mouhous, R. N., Devaux, M. F., Riaublanc, A., Bouchet,B., Gallant, J. D., et al. (2000). Monitoring the identity and thestructure of soft cheeses by fluorescence spectroscopy. Le Lait,80, 621–634.

Hildrum, K. I., Wold, J. P., Segtnan, V. H., Renou, J. P., & Dufour, E.(2006). New spectroscopic techniques for online monitoring ofmeat quality. In L. M. L. Nollet & F. Toldra (Eds.), Advancedtechnologies for meat processing (pp. 88–129). London: Taylor& Francis.

Jensen, S. A., Reenberg, S., & Munck, L. (1989). Fluorescenceanalysis in fish and meat technology. In A. D. Francisco (Ed.),Fluorescence analysis in foods (pp. 181–192). London: LongmanGroup.

Karoui, R., & Dufour, E. (2003). Dynamic testing rheology andfluorescence spectroscopy investigations of surface to centredifferences in ripened soft cheeses. International Dairy Journal,13, 973–985.

Karoui, R., & Dufour, E. (2006). Prediction of the rheologyparameters of ripened semi-hard cheeses using fluorescencespectra in the UV and visible ranges recorded at a young stage.International Dairy Journal, 16, 1490–1497.

Karoui, R., Mazerolles, G., & Dufour, E. (2003). Spectroscopictechniques coupled with chemometric tools for structure andtexture determinations in dairy products: A review. InternationalDairy Journal, 13, 607–620.

Karoui, R., Dufour, E., Pillonel, L., Picque, D., Cattenoz, T., & Bosset,J. O. (2004a). Determining the geographic origin of Emmentalcheeses produced during winter and summer using a techniquebased on the concatenation of MIR and fluorescence spectroscopicdata. European Food Research and Technology, 219, 184–189.

Karoui, R., Dufour, E., Pillonel, L., Picque, D., Cattenoz, T., & Bosset,J. O. (2004b). Fluorescence and infrared spectroscopies: A tool forthe determination of the geographic origin of Emmental cheesesmanufactured during summer. Le Lait, 84, 359–374.

Karoui, R., Bosset, J. O., Mazerolles, G., Kulmyrzaev, A., & Dufour,E. (2005a). Monitoring the geographic origin of both experimen-tal French Jura hard cheeses and Swiss Gruyère and l'Etivaz PDOcheeses using mid-infrared and fluorescence spectroscopies.International Dairy Journal, 15, 275–286.


http://dx.doi.org/10.1016/j.foodchem.2010.03.107

Karoui, R., Dufour, E., Pillonel, L., Schaller, E., Picque, D., Cattenoz,T., et al. (2005b). Determination of the geographic origin ofEmmental cheeses by combining infrared and fluorescencespectroscopies. International Dairy Journal, 15, 287–298.

Karoui, R., Martin, B., & Dufour, E. (2005c). Potentiality of front-facefluorescence spectroscopy to determine the geographic origin ofmilks from Haute-Loire department (France). Le Lait, 85, 223–236.

Karoui, R., Cartaud, G., & Dufour, E. (2006a). Front-face fluorescencespectroscopy as a rapid and non-destructive tool for differentiatingvarious cereal products: A preliminary investigation. Journal ofAgricultural and Food Chemistry, 54, 2027–2034.

Karoui, R., Dufour, E., & De Baerdemaeker, J. (2006b). Commoncomponents and specific weights analysis: A tool for monitoringthe molecular structure of semi-hard cheese throughout ripening.Analytica Chimica Acta, 572, 125–133.

Karoui, R., Kemps, B., Bamelis, F., De Ketelaere, B., Mertens, K.,Schoonheydt, R., et al. (2006c). Development of a rapid methodbased on front face fluorescence spectroscopy for the monitoringof egg freshness: 1—Evolution of thick and thin albumens.European Food Research and Technology, 223, 303–312.

Karoui, R., Kemps, B., Bamelis, F., De Ketelaere, B., Mertens, K.,Schoonheydt, R., et al. (2006d). Development of a rapid methodbased on front face fluorescence spectroscopy for the monitoringof egg freshness: 2—Evolution of yolk. European FoodResearch and Technology, 223, 180–188.

Karoui, R., Mouazen, A. M., Dufour, E., & De Baerdemaeker, J.(2006e). Utilisation of front face fluorescence spectroscopy forthe determination of some chemical parameters in soft cheeses atexternal and central zones. Le Lait, 86, 155–169.

Karoui, R., Thomas, E., & Dufour, E. (2006f). Utilisation of rapidtechnique based on front-face fluorescence spectroscopy fordifferentiating between fresh and frozen-thawed fish fillets. FoodResearch International, 39, 349–355.

Karoui, R., Dufour, E., Bosset, J. O., & De Baerdemaeker, J. (2007a).The use of front face fluorescence spectroscopy to classify thebotanical origin of honey samples produced in Switzerland. FoodChemistry, 101, 314–323.

Karoui, R., Dufour, E., & De Baerdemaeker, J. (2007b). Front facefluorescence spectroscopy coupled with chemometric tools formonitoring the oxidation of semi-hard cheeses throughoutripening. Food Chemistry, 101, 1305–1314.

Karoui, R., Dufour, E., Schoonheydt, R., &De Baerdemaeker, J. (2007c).Characterisation of soft cheese by front face fluorescence spectros-copy coupled with chemometric tools: Effect of the manufacturingprocess and sampling zone. Food Chemistry, 100, 632–642.

Karoui, R., Schoonheydt, R., Decuypere, E., Nicolaï, B., & DeBaerdemaeker, J. (2007d). Front face fluorescence spectroscopyas a tool for the assessment of egg freshness during storage at atemperature of 12.2°C and relative humidity of 87%. AnalyticalChimica Acta, 582, 83–91.

Karoui, R., Schoonheydt, R., Decuypere, E., Nicolaï, B., & DeBaerdemaeker, J. (2007e). Monitoring the egg freshness duringstorage under modified atmosphere by fluorescence spectroscopy.Food and Bioprocess Technology, 1, 346–356.

Kim, M. S., Lefcourt, A. M., & Chen, Y. N. (2003). Multispectrallaser-induced fluorescence imaging system for large biologicalsamples. Applied Optics, 42, 3927–3934.

Kulmyrzaev, A., & Dufour, E. (2002). Determination of lactulose andfurosine in milk using front-face fluorescence spectroscopy. LeLait, 82, 725–735.

Kulmyrzaev, A., & Dufour, E. (2010). Relations between spectral andphysicochemical properties of cheese, milk, and whey examinedusing multidimensional analysis. Food and Bioprocess Technol-ogy, 3, 247–256.

Kulmyrzaev, A., Levieux, D., & Dufour, E. (2005). Front-facefluorescence spectroscopy allows the characterization of mild

heat treatment applied to milk. Relations with the denaturation ofmilk proteins. Journal of Agricultural and Food Chemistry, 53,502–507.

Kyriakidis, N. B., & Skarkalis, P. (2000). Fluorescence spectrameasurement of olive oil and other vegetable oils. Journal ofthe American Oil Chemistry Society, 83, 1435–1439.

Lakowicz, J. R. (1983). Fluorophores. In J. R. Lakowicz (Ed.),Principles of fluorescence spectroscopy (pp. 63–93). New York:Plenum.

Lebecque, A., Laguet, A., Chanonat, M., Lardon, S., & Dufour, E.(2003). Joint analysis of sensory and instrumental data applied tothe investigation of the texture of Charolais meat. Science desAliments, 23, 172–175.

Leblanc, L., & Dufour, E. (2002). Monitoring the identity of bacteriausing their intrinsic fluorescence. FEMS Microbiology Letters,211, 147–151.

Leblanc, L., & Dufour, E. (2004). Monitoring bacteria growth usingtheir intrinsic fluorescence. Sciences des Aliments, 24, 207–220.

Leriche, F., Bordessoules, A., Fayolle, K., Karoui, R., Laval, K.,Leblanc, L., et al. (2004). Alteration of raw-milk cheese byPseudomonas spp.: Monitoring the sources of contaminationusing fluorescence spectroscopy and metabolic profiling. Journalof Microbiological Methods, 59, 33–41.

Light, N., Champion, A. E., Voyle, C., & Baiely, A. J. (1985). The roleof epimysial, perimysial and endomysial collagen in determiningtexture in 6 bovine muscles. Meat Science, 13, 137–149.

Liu, X., & Metzger, L. E. (2007). Application of fluorescencespectroscopy for monitoring changes in nonfat dry milk duringstorage. Journal of Dairy Science, 90, 24–37.

Lötze, E., Huybrechts, C., Sadie, A., Theron, K. I., & Valcke, R. M.(2006). Fluorescence imaging as a non-destructive method forpre-harvest detection of bitter pit in apple fruit (Malus domesticaBorkh). Postharvest Biology and Technology, 40, 287–294.

Lucisano, M., Hidalgo, A., Comelli, E. M., & Rossi, M. (1996).Evolution of chemical and physical albumen characteristicsduring the storage of shell eggs. Journal of Agricultural andFood Chemistry, 44, 1235–1240.

Marsh, R., Kajda, P., & Ryley, J. (1994). The effect of light on thevitamin B2 and the vitamin A content of cheese. Nahrung, 38,527–32.

Mazerolles, G., Devaux, M. F., Duboz, G., Duployer, M. H., MouhousRiou, N., & Dufour, E. (2001). Infrared and fluorescencespectroscopy for monitoring protein structure and interactionchanges during cheese ripening. Le Lait, 81, 509–527.

Møller, J. K. S., Parolari, G., Gabba, L., Christensen, J., & Skibsted,L. H. (2003). Monitoring chemical changes of dry-cured Parmaham during processing by surface autofluorescence spectroscopy.Journal of Agricultural and Food Chemistry, 51, 1224–1230.

Moshou, D., Wahlen, S., Strasser, R., Schenk, A., De Baerdemaeker, J.,& Ramon, H. (2005). Chlorophyll fluorescence as a tool for onlinequality sorting of apples. Biosystems Engineering, 91, 163–172.

Munck, L. (2001). Spectroscopy: Fluorescence (review In: Encyclo-pedia of Food Science, Food Technology and Nutrition. 2nd Ed.revised from 1st Ed.). London: Academic Press, 5431–5441.

Munck, L., Nørgaard, L., Engelsen, S. B., Bro, R., & Andersson, C.A. (1998). Chemometrics in food science—A demonstration ofthe feasibility of a highly exploratory, inductive evaluationstrategy of fundamental scientific significance. Chemometricand Intelligent Laboratory Systems, 44, 31–60.

Newman, P. B. (1984). The use of video image-analysis forquantitative measurements of fatness in meat: 2. Comparison ofvia, visual assessment and total chemical fat estimation in acommercial environment. Meat Science, 10, 161–166.

Nørgaard, L. (1995). Classification and prediction of quality andprocess parameters of thick juice and beet sugar by fluorescencespectroscopy and chemometrics. Zuckerindustrie, 120, 970–981.


Noh, H. K., & Lu, R. (2007). Hyperspectral laser-induced fluores-cence imaging for assessing apple fruit quality. PostharvestBiology and Technology, 43, 193–201.

Olsen, E., Vgot, G., Ekeberg, D., Sandbakk, M., Pettersen, J., &Nilsson, A. (2005). Analysis of the early stages of lipid oxidationin freeze-stored pork back fat and mechanicallly recovered poultrymeat. Journal of Agricultural and Food Chemistry, 53, 338–348.

Olsen, E., Veberg, A., Vogt, G., Tomic, O., Kirkhus, B., Ekeberg, D.,et al. (2006). Analysis of early lipid oxidation in salmon pâtéwith cod liver oil and antioxidants. Journal of Food Science, 71,S284–S292.

Papageorgiou, G. C., Govindjee (Ed.) (2004). Chlorophyll a fluores-cence: a signature of photosynthesis. Dordrecht: Springer

Posudin, Y. I. (1998). Lasers in agriculture. New York: Science.Poulli, K. I., Mousdis, G. A., & Georgiou, C. A. (2005). Classification

of edible and lampante virgin olive oil based on synchronousfluorescence and total luminescence spectroscopy. AnalyticaChimica Acta, 542, 151–156.

Poulli, K. I., Mousdis, G. A., & Georgiou, C. A. (2007). Rapidsynchronous fluorescence method for virgin olive oil adulterationassessment. Analytica Chimica Acta, 105, 369–375.

Ram, M. S., Seitz, L. M., & Dowell, F. E. (2004). Naturalfluorescence of red and white wheat kernels. Cereal Chemistry,81, 244–248.

Ramanujam, N. (2000). Fluorescence spectroscopy in vivo. In R. A.Meyers (Ed.), Encyclopedia of analytical chemistry (pp. 20–56).London: Wiley.

Renou, J. P., Deponge, C., Gachon, P., Bonnefoy, J. C., Coulon, J. B.,Garel, J. P., et al. (2004). Characterization of animal productsaccording to geographic origin and feeding diet using nuclearmagnetic resonance and isotope ratio mass spectrometry: Cowmilk. Food Chemistry, 85, 63–66.

Rost, F. W. D. (1995). Fluorescence microscopy (Vol 1 und 2).Cambridge: Cambridge University Press.

Rouissi, H., Dridi, S., Kammoun, M., De Baerdemaeker, J., & Karoui,R. (2008). Front face fluorescence spectroscopy: A rapid tool fordetermining the effect of replacing soybean meal with scotchbean in the ration on the quality of Sicilo-Sarde ewe’s milkduring lactation period. European Food Research and Technol-ogy, 226, 1021–1030.

Ruoff, K., Karoui, R., Dufour, E., Luginbühl, W., Bosset, J. O.,Bogdanov, S., et al. (2005). Authentication of the botanical originof honey by front-face fluorescence spectroscopy, a preliminarystudy. Journal of Agricultural and Food Chemistry, 53, 1343–1347.

Ruoff, R., Luginbühl, W., Künzli, R., Bogdanov, S., von der Ohe, K.,von der Ohe, W., et al. (2006). Authentication of the botanicaland geographical origin of honey by front face fluorescencespectroscopy. Journal of Agricultural and Food Chemistry, 54,6858–6866.

Sahar, A., Boubellouta, T., Lepetit, J., & Dufour, E. (2009). Front-facefluorescence spectroscopy as a tool to classify seven bovinemuscles according to their chemical and rheological character-istics. Meat Science, 83, 672–677.

Sahar, A., Boubellouta, T., Portanguen, S., Kondjoyan, A., & Dufour,E. (2009). Synchronous front-face fluorescence spectroscopycoupled with Parallel Factors (PARAFAC) analysis to study theeffects of cooking time on meat. Journal of Food Science, 74,E534–E539.

Saito, Y. (2009). Monitoring raw material by laser-induced fluores-cence spectroscopy in the production. In M. Zude (Ed.), Opticalmonitoring of fresh and processed agricultural crops (pp. 319–336). New York: CRC.

Sayago, A., Morales, M. T., & Aparicio, R. (2004). Detection ofhazelnut oil in virgin olive oil by a spectrofluorimetric method.European Food Research and Technology, 218, 480–483.

Schamberger, G. P., & Labuza, T. P. (2006). Evaluation of front-facefluorescence for assessing thermal processing of milk. Journal ofFood Science, 71, 69–74.

Seiden, P., Bro, R., Poll, L., & Munck, L. (1996). Exploringfluorescence spectra of apple juice and their connection toquality parameters by chemometrics. Journal of Agriculturaland Food Chemistry, 44, 3202–3205.

Sikorska, E., Górecki, T., Khmelinskii, I. V., Sikorski, M., & DeKeukeleire, D. (2004). Fluorescence spectroscopy for character-ization and differentiation of beers. Journal of the Institute of theBrewing, 110, 267–275.

Sikorska, E., Górecki, T., Khmelinskii, I. V., Sikorski, M., & Kozioł, J.(2005). Classification of edible oils using synchronous scanningfluorescence spectroscopy. Food Chemistry, 89, 217–225.

Skjervold, P. O., Taylor, R. G., Wold, J. P., Berge, P., Abouelkaram,S., Culioli, J., et al. (2003). Development of intrinsic fluorescentmultispectral imagery specific for fat, connective tissue, andmyofibers in meat. Journal of Food Science, 68, 1161–1168.

Smilde, A., Bro, R., & Geladi, P. (2004). Multi-way analysis withapplication in the chemical sciences. England: Wiley.

Song, J., Deng, W., & Beaudy, R. M. (1997). Changes in chlorophyllfluorescence of apple fruit during maturation, ripening, andsenescence. Hort Science, 32, 891–896.

Stadelman, W. J., Ziegler, F., & Darroch, J. G. (1954). The effect ofegg temperature on its broken-out albumen quality evaluation.Poultry Science, 33, 1082–1083.

Strasburg, G. M., & Ludescher, R. D. (1995). Theory and applicationsof fluorescence spectroscopy in food research. Trends in FoodScience & Technology, 6, 69–75.

Swatland, H. J. (1987). Autofluorescence of adipose tissue measuredwith fibre optique. Meat Science, 19, 277–284.

Swatland, H. J., & Findlay, C. J. (1997). On-line probe prediction ofbeef toughness, correlating sensory evaluation with fluorescencedetection of connective tissue and dynamic analysis of overalltoughness. Food Quality and Preference, 8, 233–239.

Swatland, H. J., Gullet, E., Hore, T., & Buttenham, S. (1995). UVfiber-optic probe measurements of connective tissue in beefcorrelated with taste panel scores for chewiness. Food ResearchInternational, 28, 23–30.

Swatland, H. J., Nielsen, T., & Andersen, J. R. (1995). Correlations ofmature beef palatability with optical probing of raw meat. FoodResearch International, 28, 403–416.

Symons, S. J., & Dexter, J. E. (1991). Computer analysis offluorescence for the measurement of flour refinement asdetermined by flour ash content, flour grade color, andtristimulus color measurements. Cereal Chemistry, 68, 454–460.

Symons, S. J., & Dexter, J. E. (1992). Estimation of millingefficiency: Prediction of flour refinement by the measurementof pericarp fluorescence. Cereal Chemistry, 69, 137–141.

Symons, S. J., & Dexter, J. E. (1993). Relationship of flour aleuronefluorescence to flour refinement for some Canadian hardcommon wheat classes. Cereal Chemistry, 70, 90–95.

Symons, S. J., & Dexter, J. E. (1994). Aleurone and pericarpfluorescence as estimators of mill stream refinement forvarious Canadian wheat classes. Journal of Cereal Science,23, 73–83.

Tourkya, B., Boubellouta, T., Dufour, E., & Leriche, F. (2009).Fluorescence spectroscopy as a promising tool for a polyphasicapproach to Pseudomonad taxonomy. Current Microbiology, 58,39–46.

Valeur, B., & Bochon, J. C. (Eds.). (2001). New trends in fluorescencespectroscopy—Applications to chemical and life sciences. Berlin:Springer.

Veberg, A. (2006). Fluorescence spectroscopy of food lipid oxidation.PhD Thesis, Department of Chemistry, Biotechnology and FoodScience, Norwegian University of Life Science, Norway, Norwege.


Veberg, A., Olsen, E., Vogt, G., Mielnik, M., Nilsen, A. N., & Wold, J.P. (2006). Front face fluorescence spectroscopy—A rapid methodto detect early lipid oxidation in freeze stored minced Turkeymeat. Food Science, 71, 364–370.

Wold, J. P., Kvaal, K., & Egelandsdal, B. (1999). Quantification ofintramuscular fat content in beef by combining autofluorescencespectra and autofluorescence images. Applied Spectroscopy, 53,448–456.

Wold, J. P., Lundby, F., & Egelandsdal, B. (1999). Quantification ofconnective tissue (hydroxyproline) in round beef by autofluor-escence spectroscopy. Journal of Food Science, 64, 377–383.

Wold, J. P., Jørgensen, K., & Lundby, F. (2002). Nondestructivemeasurement of light-induced oxidation in dairy products byfluorescence spectroscopy and imaging. Journal of DairyScience, 85, 1693–1704.

Wold, J. P., Veberg, A., Nilsen, A., Iani, V., Juzenas, P., & Moan, J.(2005). The role of naturally occurring chlorophyll and porphyr-ins in light-induced oxidation of dairy products. A study basedon fluorescence spectroscopy and sensory analysis. InternationalDairy Journal, 15, 343–353.

Zaïdi, F., Rouissi, H., Dridi, S., Kammoun, M., De Baerdemaeker,J., & Karoui, R. (2008). Front face fluorescence spectroscopy

as a rapid and non destructive tool for differentiating betweenSicilo-Sarde and Comisana ewe’s milk during lactation period:A preliminary study. Food Bioprocess and Technology, 1, 143–151.

Zandomeneghi, M. (1999). Fluorescence of cereal flours. Journal ofAgricultural and Food Chemistry, 47, 878–882.

Zandomeneghi, M., Festa, C., & Carbonaro, L. (2000). Front-surface absorbance spectra of wheat flour: Determination ofCarotenoids. Journal of Agricultural and Food Chemistry, 48,2216–2221.

Zandomeneghi, M., Carbonaro, L., Calucci, L., Pinzino, C., Galleschi,L., & Ghiringhelli, S. (2003). Direct fluorometric determinationof fluorescent substances in powders: The case of riboflavin incereal flours. Journal of Agricultural and Food Chemistry, 51,2888–2895.

Zandomeneghi, M., Carbonaro, L., & Caffarata, C. (2005). Fluores-cence of vegetable oils: Olive oils. Journal of Agricultural andFood Chemistry, 53, 759–766.

Zude, M. (Ed.). (2008). Optical methods for monitoring fresh andprocessed food—Basics and applications for a better under-standing of non-destructive sensing. Boca Raton: Taylor &Francis.


Full Text

Documents

food quality

food production

food products

food matrix

synchronous fluorescence

fluorescence analysis

high quality

quality of animal