XXI Meeting of the Portuguese Electrochemistry Society &

XVIII Iberian Electrochemistry MeetingAbstract Book

XXI Encontro da Sociedade Portuguesa de Eletroquímica&

XVIII Encontro Ibérico de EletroquímicaLivro de Resumos

Title XXI Meeting of the Portuguese Electrochemistry Society & XVIII Iberian Electrochemistry Meeting

Título XXI Encontro da Sociedade Portuguesa de Eletroquímica & XVIII Encontro Ibérico de Eletroquímica

Event Abbreviation / Abreviatura do Evento SPE2016

Coordination / Coordenação António M. Peres (Instituto Politécnico de Bragança, Portugal) Conceição Angélico (Instituto Politécnico de Bragança, Portugal) Luís G. Dias (Instituto Politécnico de Bragança, Portugal) Maria José Arabolaza (Instituto Politécnico de Bragança, Portugal) Miguel Vilas-Boas (Instituto Politécnico de Bragança, Portugal)

Edition / Edição Instituto Politécnico de Bragança · 2016 5300-253 Bragança · Portugal Tel. (+351) 273 303 200 · Fax (+351) 273 325 405 http://www.ipb.pt

Imaging services / Serviços de imagem Atilano Suarez (Instituto Politécnico de Bragança, Portugal)

URI http://hdl.handle.net/10198/12931

ISBN 978-972-745-213-2

�iii

��

PC13Qualita:veevalua:onofTunisianoliveoilsusinganelectronictongueandchemometrictools:aprospec:vestudy.

Souihli Slima,b, Nuno Rodriguesc,d, Luís G. Diasa,e, Ana C.A. Velosof,h, José A. Pereirai, Souheib Ouesla5b, António M. Peresj

aEscolaSuperiorAgrária,Ins)tutoPolitécnicodeBragança,CampusSantaApolónia,5300-253Bragança,Portugal.bUniversitéLibredeTunis,32BisavenueKheireddinePacha,1002Tunis,Tunisia.cREQUIMTE-LAQV/CIMO,EscolaSuperiorAgrária, Ins)tutoPolitécnicodeBragança,CampusSantaApolónia,5300-253Bragança,Portugal.dUniversidaddeLéon,DepartamentodeIngenieríaAgrária,Av.Portugal,nº41,24071Léon,España.eCQ-VR,CentrodeQuímica–VilaReal,UniversityofTrás-os-MonteseAltoDouro,Apartado1013,5001-801VilaReal,PortugalfIns)tutoPolitécnicodeCoimbra,ISEC,DEQB,RuaPedroNunes,QuintadaNora,3030-199Coimbra,PortugalhCEB-CentreofBiologicalEngineering,UniversityofMinho,CampusdeGualtar,4710-057Braga,PortugaliREQUIMTE-LAQV,EscolaSuperiorAgrária,Ins)tutoPolitécnicodeBragança,CampusSantaApolónia,5300-253Bragança,Portugal.j LaboratoryofSepara)onandReac)onEngineering -LaboratoryofCatalysisandMaterials (LSRE-LCM),EscolaSuperiorAgrária,Ins)tutoPolitécnicodeBragança,CampusSantaApolónia,5300-253Bragança,Portugal

* peres@ipb.pt

Olive oil commercializa:on has a great impact in the regional economy of several countries includingTunisia. It is a high-value food product, quite prone to frauds. So, it is important to establish analy:caltechniques that can ensure labeling correctness regarding olive oil quality as well as its origin, namelyconcerning theolive(s) cul:var(s)used in theproduc:on,which isofmajor importance formonovarietaloliveoils.Tradi:onalanaly:caltechniqueslikethosebasedonchromatographyarequiteexpensive,:me-consuming,notportableanddifficulttoimplementin-situ,consideringtheusualharshenvironmentsoftheolive industry. In this work, the feasibility of using an electronic tongue as a classifica:on tool fordiscrimina:ngTunisianoliveoilsaccordingtotheirqualitylevel(i.e.,extravirginoliveoil,virginoliveoilorlampanteoliveoil)andolivecul:var(i.e.,Chetoui,Shaliandothers,accordingtothelabelinforma:on)wasevaluatedforthefirst:me.Oliveoilqualitywasassessedqualityparameters(freeacidity,peroxidevalues,K232 and K270 ex:nc:on coefficients) and on the organolep:c evalua:on carried out by a sensory panelparameters (according to the Interna:onal Olive Council direc:ves). The poten:ometric signal profilesrecorded with the electrochemical mul:-sensor device during the electrochemical analysis of olive oils’hydroethanoliocextracts(≈5min),coupledwithlineardiscriminantmodels,establishedbasedonthemostinforma:ve sub-sets of sensors, selected by a simulated annealing algorithm,were able to sa:sfactorilyperformoliveoilsdiscrimina:onaccordingto:(i) Olivecul:var:sensi:vi:esof88%for leave-one-outcross-valida:onandmeansensi:vi:esof79%fortherepeatedK-foldscross-valida:on(4foldswith10repeats),achievedwithamul:variatemodelbasedontheinforma:ongatheredby20sensorsoftheelectronictongue;and,(ii) Quality level:sensi:vi:esof91%for leave-one-outcross-valida:onandmeansensi:vi:esof84%fortherepeatedK-foldscross-valida:on(4foldswith10repeats),achievedwithamul:variatemodelbasedontheinforma:ongatheredby26sensorsoftheelectronictongue.

Overall, the results show the sa:sfactory performance of a poten:ometric electronic tongue containingcross-sensi:vity lipidmembranes as sensors, whichmay be tenta:vely a\ributed to the capacity of theelectrochemical device in discrimina:ng olive oils with different polar compounds contents, which arerelatedtospecificsensorya\ributesofoliveoilssuchasbi\ernessandpungency.Furthermore,thepresentstudy, concerning Tunisian olive oils analysis using an electronic tongue, confirms the results previouslyreportedintheliteratureforoliveoilsfromothergeographicalorigins.

�44

Souihli Slim1,2, Nuno Rodrigues3,4, Luís G. Dias1,5, Ana C.A. Veloso6,7, José A. Pereira8, Souheib Oueslati2, António M. Peres9

1: ESA, Instituto Politécnico de Bragança, Bragança, Portugal; 2: Université Libre de Tunis, Tunis, Tunisia; 3: Universidad de Léon, Departamento de IngenieríaAgrária, Léon, España; 4: REQUIM,TE-LAQV/CIMO, ESA, Instituto Politécnico de Bragança, Bragança, Portugal; 5: CQ-VR, Centro de Química – Vila Real, Universidade de Trás-os-Montes e Alto Douro, Vila Real, Portugal; 6: Instituto Politécnico de Coimbra, ISEC, DEQB, Coimbra, Portugal; 7: CEB, University of Minho, Braga, Portugal; 8: REQUIMTE-LAQV, ESA, Instituto Politécnico de Bragança, Bragança, Portugal; 9: LSRE-LCM, ESA, Instituto Politécnico de Bragança, Bragança, Portugal (*) peres@ipb.pt

CONCLUSIONS

• The potentiometric E-tongue coupled with a LDA-SA procedure demonstrated to be a fast and cost-effective tool for :

à Tunisian olive oils’ predictive classification according to olive cultivar;

à Tunisian olive oils’ predictive classification according to quality grade.

• The overall results achieved confirmed the E-tongue potential for olive oil analysis, previously reported by our research group

[1-6] as well as other reearch groups [7-12].

1

2

3

4

Figure 1 – Multi-sensor analytical system:1 - PC for data acquisition;2 - DataLogger Agilent;3 – Electronic tongue;4 – Magnetic stirrer.

3

Double multi-sensor system:

40 sensors

INTRODUCTION

üOlive oil commercialization has a great impact in the

regional economy of several countries including Tunisia.

üIt is a high-value food product, quite prone to frauds.

üTraditional analytical techniques are quite expensive, time-

consuming, not portable and difficult to implement in-situ,

due to the harsh environments of the olive industry.

ü In this work, the feasibility of using an electronic tongue as

a classification tool for discriminating Tunisian olive oils

ELECTRONIC TONGUEê

Potenciometric system(all-solid-state electrodes)

ê20 lipidic polymeric membranes (�2)

Ag/AgCl reference electrodeData acquisition with DataLogger Agilent

Each lipidic polymeric membrane contains:

32% of PVC;65% of plasticizer;3% of additive compound.

___________________________________________________________________________________

Additive compound___________________________________________________________________________________

[1] Octadecylamine[2] Oleyl alcohol[3] Methyltrioctylammonium chloride[4] Oleic acid

___________________________________________________________________________________

Plasticizer___________________________________________________________________________________

[A] Bis(1-butylpentyl) adipate[B] Dibutyl sebacate[C] 2-Nitrophenyl-octylether[D] (2-ethylhexyl)phosphate[E] Dioctyl phenylphosphonate

___________________________________________________________________________________

Electronic tongue analysis

ê

TUNISIAN OLIVE OILS

Quality: EVOO (3); VOO (4) & LOO (36)(according to physico-chemical & sensory analysis)

Cultivar: Chetoui cv. (11); Sahli cv. (26) & others cvs. (4)(according to the label information)

ê

Extraction with H2O:EtOH (80:20 v/v), to obtain a polar compounds rich-solution

AcknowledgementsThis work was co-financed by Project POCI-01-0145-FEDER-006984 – Associated Laboratory LSRE-LCM funded by FEDER funds through COMPETE2020 - Programa Operacional Competitividade e Internacionalização (POCI), by national funds through FCT - Fundação para a Ciência e a Tecnologia; and by the strategic funding of UID/BIO/04469/2013 unit. Nuno Rodrigues thanks FCT, POPH-QREN and FSE for the Ph.D. Grant (SFRH/BD/104038/2014).

OBJECTIVES

QUALITATIVE EVALUATION OF TUNISIAN OLIVE OILS USING AN ELECTRONICTONGUE AND CHEMOMETRIC TOOLS: A PROSPECTIVE STUDY

RESULTS

Establishment of the best E-tongue-LDA-SA models:à variable selection with simulated annealing (SA) algorithmà sub-set with minimum number of sensors => maximum correct classification,

- LOO-CV

- repeated K-folds-CV: K=4 folds × 10 repeats

i) Olive oils’ quality discrimination: EVOO, VOO & LOOà model with 26 potentiometric sensor signals à 91% of predictive correct classifications for LOO-CVà 84% of predictive correct classifications for repeated

K-folds-CV

ii) Olive oils’ discrimination according toTunisian cultivars: Chetoui, Sahli & othersà model with 20 potentiometric sensor signals à 88% of predictive correct classifications for LOO-CVà 79% of predictive correct classifications for repeated

K-folds-CV

ELECTRONIC TONGUE

(multi-sensor system)ê

Chemical sensors with high stability and cross-sensitivity to different substances in solution

ê

TO OBTAINOlive oils’ signals profiles corresponding to a

fingerprint of sample matrix

TO APPLY

Chemometric methods:- Linear discriminant analysis (LDA)

- Simulated annealing (SA) variable selection algorithm- Leave-one-out cross-validation (LOO-CV)

- Repeated K-folds-CV (repeated K-folds-CV)ê

TO ALLOWOlive oils’s identification/classification according to:

- Olive oil quality (EVOO, VOO & LOO)-Olive cultivar (Chetoui, Sahli & others)

75% data for training

25% data for internal validation

REFERENCES[1]Pereset al.(2014)ProcediaEngineering,87:192-195;[2]Diaset al.(2014)Food Chemistry,160:321–329;[3]Velosoet al.(2016)Talanta,146:585–593;[4]Diaset al.(2016)European Food ResearchandTechnology,242:259-270;[5]Rodrigueset al.(2016)LWT- Food Science and Technology,73:683-692;[6]Rodrigueset al.(2016)European Food Researchand Technology (DOI:10.1007/s00217-016-2773-2);[7]Apetrei et al.(2010)Analytica Chimica Acta,663,91-97;[8]Apetrei &Apetrei (2013)Food ResearchInternational,54,2075-2082;[9]Cosio etal.(2007)FoodChemistry,101,485-491;[10]Haddi etal.(2013)FoodResearchInternational,54,1488-1498;[11]Oliveri etal.(2009)AnalyticalandBioanalytical Chemistry,395,1135–1143;[12]Rodríguez-Méndez etal.(2008)Electrochimica Acta,53,5867-5872

-5.0

-2.5

0.0

2.5

-15 -10 -5 0 5First discriminant function (97.2%)

Seco

nd d

iscr

imin

ant f

unct

ion

(2.8

%)

EVOO

LOO

VOO

-2.5

0.0

2.5

5.0

-10 -5 0 5First discriminant function (94.2%)

Seco

nd d

iscr

imin

ant f

unct

ion

(5.8

%)

Chetoui

Other

Sahli