Chemical analysis of food techniques and applications edited by yolanda picó

The progress of food science and its concepts have driven change of classic analytical methods titrimetric or gravimetric analysis to instrumental and biochemical ones chromatog-raphy, b

CHEMICAL ANALYSIS OF FOOD: TECHNIQUES AND

APPLICATIONS

YOLANDA PICO´

Department of Medicine Preventive, Faculty of Pharmacy, University of Valencia, Spain

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an imprint of Elsevier

525 B Street, Suite 1900, San Diego, CA 92101-4495, USA

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands

First edition 2012

Copyright Ó 2012 Elsevier Inc All rights reserved

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or

by any means electronic, mechanical, photocopying, recording or otherwise without the prior writtenpermission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology RightsDepartment in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email:permissions@elsevier.com Alternatively you can submit your request online by visiting the Elsevier web site athttp://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Notice

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter

of product liability, negligence or otherwise, or from any use or operation of any methods, products,

instructions or ideas contained in the material herein Because of rapid advances in the medical sciences,

in particular, independent verification of diagnoses and drug dosages should be made

Library of Congress Cataloging-in-Publication Data

Chemical analysis of food: techniques and applications/edited by Yolanda Pico´

A catalogue record for this book is available from the British Library

For information on all Academic Press publications

visit our web site atelsevierdirect.com

Printed and bound in USA

12 13 14 15 16 10 9 8 7 6 5 4 3 2 1

ISBN: 978-0-12-384862-8

Ouissam Abbas Walloon Agricultural Research

Centre (CRA-W), “Henseval” building, Chausse´e

de Namur, 24, 5030 Gembloux, Belgium

Eugenio Aprea IASMA Research and Innovation

Centre, Food Quality and Nutrition, Area, Via E

Mach, S Michele all’Adige (TN), Italy

Kavita Arora Advanced Instrumentation Research

Facility, Jawaharlal Nehru University, New Delhi

110067, India

Vincent Baeten Walloon Agricultural Research

Centre (CRA-W), ’Henseval’ building, Chausse´e

de Namur, 24, 5030 Gembloux, Belgium

Damia` Barcelo´ Departmento of Environmental

Chemistry, IDAEA-CSIC, Barcelona, Spain

Catalan Institute for Water Research (ICRA),

Girona, Spain

Simona Benedetti Department of Food Technology,

University of Milan, Via Celoria, Milan, Italy

Carlo Bicchi Dipartimento di Scienza e Tecnologia

del Farmaco, Universita` degli Studi di Torino,

Via Pietro Giuria n9, Turino, Italy

Pierre-Antoine Bonnet Laboratories and Control

Department, Agence Franc¸aise de Se´curite´

Sanitaire des Produits de Sante´ (AFSSAPS), 635

rue de la Garenne, 34740 Vendargues, France

Monique Bremer RIKILT e Institute of Food Safety,

Wageningen University and Research Centre,

Wageningen, The Netherlands

Franca Carini Institute of Agricultural and

Environmental Chemistry, Universita` Cattolica

del Sacro Cuore, Piacenza, Italy

Alejandro Cifuentes Laboratory of Foodomics,

Institute of Food Science Research CIAL (CSIC),

Madrid, Spain

Chiara Cordero Dipartimento di Scienza e

Tecnologia del Farmaco, Universita` degli Studi di

Torino, Via Pietro Giuria n9, Turino, Italy

M.S Cosio Department of Food Technology,University of Milan, Via Celoria, Milan, ItalyBarbara d’Acampora Zellner Dipartimento Farmaco-chimico, Facolta` di Farmacia, Universita` di Messina,Viale Annunziata, Messina, Italy

Photis Dais NMR Laboratory, Department ofChemistry, University of Crete, Voutes campus,Heraklion, Crete, Greece

Pierre Dardenne Walloon Agricultural ResearchCentre (CRA-W), ’Henseval’ building, Chausse´e

de Namur, 24, 5030 Gembloux, BelgiumPaola Dugo Dipartimento Farmaco-chimico,Facolta` di Farmacia, Universita` di Messina, VialeAnnunziata, Messina, Italy Universita` Campus-Biomedico, Via Alvaro del Portillo, Roma, ItalyGiovanni Dugo Dipartimento Farmaco-chimico,Facolta` di Farmacia, Universita` di Messina, VialeAnnunziata, Messina, Italy

Lisa Elviri Dipartimento di Chimica Generale eInorganica, Chimica Analitica, Chimica Fisica,Universita` degli studi di Parma, Parco Area delleScienze 17/a, Parma, Italy

Marinella Farre´ Departmento of EnvironmentalChemistry, IDAEA-CSIC, Barcelona, SpainMichele Forina Department of Drug and FoodChemistry and Technology, University of Genova,Via Brigata Salerno, 13, Genova, Italy

Virginia Garcı´a-Can˜as Laboratory of Foodomics,Institute of Food Science Research CIAL (CSIC),Madrid, Spain

Maria Groot RIKILT e Institute of Food Safety,Wageningen University and Research Centre,Wageningen, The Netherlands

George Kaklamanos Veterinary Laboratory ofSerres, Terma Omonoias, Serres, Greece

Lina Kantiani Departmento of EnvironmentalChemistry, IDAEA-CSIC, Barcelona, Spain

vii

James M Karlinsey Department of Chemistry,

Penn State Berks, Reading, Pennsylvania, 19610

USA

Romdhane Karoui Universite´ d’Artois, Faculte´ des

Sciences Jean Perrin, Rue Jean Souvraz, Lens

Cedex, France

Esther Kok RIKILT Institute of Food Safety,

Wageningen University and Research Centre,

Wageningen, The Netherlands

Jozef L Kokini University of Illinois at Urbana,

Champaign College of Agriculture and

Consumer Sciences, Food Science and Human

Nutrition Department

Varinder Kaur Department of Chemistry, Punjabi

University, Patiala, Punjab, India Department of

Chemistry, Panjab University, Chandigarh, India

Sumati Kumar Department of Chemistry, Ch Devi

Lal University, Sirsa Haryana, India

Erica Liberto Dipartimento di Scienza e Tecnologia

del Farmaco, Universita` degli Studi di Torino, Via

Pietro Giuria n9, Turino, Italy

Myriam Malet-Martino Biomedical NMR Group,

SPCMIB Laboratory (UMR CNRS 5068),

Universite´ Paul Sabatier, 118 route de Narbonne,

31062 Toulouse cedex, France

Ashok Kumar Malik Department of Chemistry,

Punjabi University, Patiala, Punjab, India

Department of Chemistry, Panjab University,

Chandigarh, India

Vicky Manti RIKILT e Institute of Food Safety,

Wageningen University and Research Centre,

Wageningen, The Netherlands

Robert Martino Biomedical NMR Group, SPCMIB

Laboratory (UMR CNRS 5068), Universite´ Paul

Sabatier, 118 route de Narbonne, 31062 Toulouse

cedex, France

Monica Mattarozzi Dipartimento di Chimica

Generale e Inorganica, Chimica Analitica,

Chimica Fisica, Universita` digital studi di

Parma, Parco Area delle Scienze 17/a, Parma,

Italy

Linda Monaci Institute of Sciences of Food

Production (ISPA), National Research Council of

Italy (CNR), Bari, Italy

Luigi Mondello Dipartimento Farmaco-chimico,Facolta` di Farmacia, Universita` di Messina, VialeAnnunziata, Messina, Italy Universita` Campus-Biomedico, Via Alvaro del Portillo, Roma, ItalyPaolo Oliveri Department of Drug and FoodChemistry and Technology, University ofGenova, Via Brigata Salerno, 13, Genova, ItalyYolanda Pico´ Food and Environmental SafetyResearch Group, Faculty of Phamacy, University

of Valencia,Theo Prins RIKILT e Institute of Food SafetyWageningen University and Research Centre,Wageningen, The Netherlands

Lourdes Ramos Department of InstrumentalAnalysis and Environmental Chemistry, IQOG-CSIC, Juan de la Cierva 3, Madrid, Spain

Patrizia Rubiolo Dipartimento di Scienza eTecnologia del Farmaco, Universita` degliStudi di Torino, Via Pietro Giuria n9, Turino,Italy

Mattheo Scampicchio Faculty of Science andTechnology, Free University of Bolzano, PiazzaUniversita`, Bolzano, Italy

Barbara Sgorbini Dipartimento di Scienza eTecnologia del Farmaco, Universita` degliStudi di Torino, Via Pietro Giuria n9, Turino,Italy

Varsha Sharma School of Life Sciences, JawaharlalNehru University, New Delhi 110067, IndiaAnu Singh Advanced Instrumentation ResearchFacility, Jawaharlal Nehru University, New Delhi

110067, India Department of Biotechnology,School of Life Sciences, Jaipur NationalUniversity, Jaipur, Rajasthan 302025, IndiaManoj Pratap Singh Advanced InstrumentationResearch Facility, Jawaharlal Nehru University,New Delhi 110067, India

Nesli Sozer University of Illinois at Urbana,Champaign College of Agriculture andConsumer Sciences, Food Science and HumanNutrition Department

Apostolos Spyros NMR Laboratory, Department ofChemistry, University of Crete, Voutes campus,Heraklion, Crete, Greece

Georgios Theodoridis IASMA Research and

Innovation Centre, Food Quality and Nutrition

Area, Via E Mach, S Michele all’Adige (TN),

Italy Department of Chemistry, Aristotle

University, Thessaloniki, Greece

Ine van der Fels RIKILT e Institute of Food Safety,

Wageningen University and Research Centre,

Wageningen, The Netherlands

Marjolein van der Spiegel RIKILT e Institute of

Food Safety, Wageningen University and

Research Centre, Wageningen, The

Netherlands

Leo van Raamsdonk RIKILT e Institute of Food

Safety, Wageningen University and Research

Centre, Wageningen, The Netherlands

Saskia van Ruth RIKILT e Institute of Food Safety,Wageningen University and Research Centre,Wageningen, The Netherlands

Hridya Narayan Verma Department ofBiotechnology, School of Life Sciences, JaipurNational University, Jaipur, Rajasthan 302025, IndiaAngelo Visconti Institute of Sciences of FoodProduction (ISPA), National Research Council ofItaly (CNR), Bari, Italy

Ya.I Yashin Scientific Development & ProductionCenter “Khimavtomatika,” SelskohozyaistvennayaMoscow, Russia

A.Ya Yashin Scientific Development & ProductionCenter “Khimavtomatika,” SelskohozyaistvennayaMoscow, Russia

It is a great pleasure for me to introduce a new

book from an old friend and colleage, Yolanda

Pico´, full professor at the University of Valencia

I have known Yolanda since her PhD thesis and

postdoctoral stay at the Free University of

Amsterdam Her research interests have always

been devoted to develop advanced analytical

chemistry methods for determining trace

organic contaminants in food and

environ-mental samples A few years ago I was able to

convince her to edit one of her first books on

Food Contaminants and Residue Analysis that

was published in 2008 as volume 51 of the

Comprehensive Analytical Chemistry series I

now reaffirm what I wrote in 2008 about

Yolanda’s book: that its content is again

extremely comprehensive and therefore will

solve most of the problems encountered in

food residue analysis In addition, it will be

a useful guide for either newcomers and/or

expert food laboratories seeking to solve the

traceability of a broad range of contaminants

and residues in food using the most advanced

analytical instruments

In this respect this new book describes the

incredibly large amount of the latest analytical

instruments and applications in food analysis

It is certainly a good exercise for the reader tocompare both books to better appreciate theprogress that has taken place in this field inthe past 4 years This book contains 22 chaptersdevoted to more general aspects such as qualityassurance issues and analytical techniquesinvolving state-of-the art sample preparation,chromatographic-mass spectrometric combina-tions, biosensors, nanotechnology, electropho-resis, molecular techniques, and other newtools The last part of the book reports a broadspectrum of applications including, amongothers, fraud, food proteomics, nutritionalsupplements, GMO, allergens, and emergingcontaminants

Overall this book covers most of the aspects

on the recent analysis of food contaminantsand residues, and I expect it will be a key refer-ence in the community of food residue special-ists on global scale Finally, I would like tothank Yolanda for the incredible amount ofwork, time, and expertise devoted as editor ofthe book My gratitude goes also to the variouswell-known authors for their contributions incompiling such a world-class and timely book

Food products are analyzed for a variety of

reasonsde.g., compliance with legal and

labeling requirements, assessment of product

quality, determination of nutritive value,

detec-tion of adulteradetec-tions, research and

develop-ment, etc Food analysis is an area in

continuous evolution, which is especially

impelled by the increasing demand of the

consumers for food safety and quality, the

concern of food authorities to ensure safe food

of the highest nutritional quality, and the effort

of producers and industry to meet these

demands It is also particularly complex because

it integrates and applies principles of biology,

chemistry, microbiology, biochemistry,

nutri-tion, and engineering to characterize new

ingre-dients and food products, detect the food

processing techniques used, and ensure the

safety and nutritional value of the food supply

The progress of food science and its concepts

have driven change of classic analytical

methods (titrimetric or gravimetric analysis) to

instrumental and biochemical ones

(chromatog-raphy, biosensors, spectroscopy) because of the

new quantitative and qualitative information

provided In this context, in addition to the

many excellent comprehensive descriptions of

historical and already well-established classical

methods, this book addresses the most recent

advances in analytical and bioanalytical

tech-niques and their application in innovative and

emerging areas within food science

Chemical analysis of foods presents what is

new or challenging within this subject through

multiple topics: reviewing novel technologies

increasingly applied to food analysis;

describing and analyzing in depth several

specific approaches, and providing a picture

of the most pioneering applications with aninsight into future trends The purpose ofthis book is to offer an updated and high-quality original contribution on new develop-ments in food analysis and its emergingapplications

The book contains twenty-three chapterswritten by experts on the subject and is struc-tured in two parts: the first one describes therole of the latest developments in analyticaland bioanalytical techniques, and the secondone deals with the most innovative applica-tions and issues in food analysis The two firstintroductory chapters about sampling andsample preparion and data analysis and che-mometrics are followed by a review of themost recently applied techniques in process(on-line) control and in laboratories for theanalysis of major or minor compounds offood These techniques ranged from the non-invasive and non-destructive ones, such asinfrared spectroscopy, magnetic resonance,and ultrasounds, to emerging areas as nano-technology, biosensors, and electronic nosesand tongues, including those already wellestablished in food analysis, such as chromato-graphic and electrophoretic techniques Thesechapters also include two important tools forsolving problems in chemical and biologicalanalysis: mass spectrometry and molecular-based techniques

The second part of the book looks at the areas

of food authenticity, safety, and traceability.Important and innovative issues, such as fraud-ulent practices, biological active components,flavors and odors, novel foods including those

xiii

modified genetically, dietary supplements, food

proteomics, metal speciation and radionuclides,

are covered

This book attempts to fill a void in

informa-tion on recently developed analytical techniques

for professionals, students, and academics in

food analysis by offering information on

modern instrumentation, techniques, and

appli-cations It is hoped that it will be helpful to learn

more on chemical analysis of food and of

partic-ular interest to those involved in food research

and development, as well as food product

char-acterization and analysis It is also intended to

serve as general reference for post-graduate

students, which are not exposed to many of

the emerging technologies and applications in

food analysis, as well as a practical reference

guide for a wide range of experts: biologists,

biochemists, microbiologists, food chemists,

toxicologists, chemists, agronomists, hygienists,

and everybody who needs to use analytical

techniques for evaluating food quality andsafety The techniques and applications dis-cussed in this book are not only emerging nowbut they also will be in the future critical forcontinued assurance of an affordable, safe, andavailable food supply

I would like to thank the authors that haveagreed to participate in this initiative for theirinsight and stimulating chapters and for thetime and effort devoted to them They providethe perfect blend of knowledge and skills thatwent into authoring this book I would alsoreally like to thank Prof Damia` Barcelo´ forproviding me with the opportunity to becomethe editor of this book as well as to the projectmanagers and all the staff from Elsevier foroffering excellent support and advice Finallyand foremost, I hope that the book lives up tothe expectations of the readers You are theones who will make the book an integral part

of food analysis

C H A P T E R1

Basics and Advances in Sampling

and Sample Preparation

L Ramos

Department of Instrumental Analysis and Environmental Chemistry, IQOG-CSIC,

Juan de la Cierva 3, Madrid, Spain

The first problem faced when dealing with

food science is probably the statement of the

concept of food A number of possible

defini-tions for this concept can be found in the

specialized literature Some of them focus on

its composition (typically, carbohydrates, fats,

protein and water), others in the way used by

humans to seek food items (which, in most

cultures, has nowadays changed from hunting

and gathering to farming, ranching, and

fishing) In other cases, definitions focus on the

nature of the matter itself and/or the expected

benefices associated to its consumption Finally,

one should recognize that, above definitions, theconcept food is also highly cultural dependent.Items considered food may be sourced fromwater, minerals, plants, animals, or other cate-gories such as fungus, fermented, elaborated,and processed products Taking into consider-ation some of these viewpoints, food could bedefined as any substance or product, liquid orsolid, natural, elaborated, or processed that,because of their characteristics, applications,components, preparation, and conservationstate, is eaten or drunk by humans as nourish-ment and enjoyment

Whatever the definition adopted, it is

a general consensus that, almost without

Chemical Analysis of Food: Techniques and Applications

DOI: 10.1016/B978-0-12-384862-8.00001-7 3 Copyright Ó 2012 Elsevier Inc All rights reserved.

exception, food is a complex heterogeneous

mixture of a relatively wide range of chemical

substances Also, it is agreed that the two key

aspects regarding food are its chemical

compo-sition and its physical properties The reason is

that these feature categories determine the

nutritional value of the considered food item

and its sanitary state, as well as its acceptation

by consumers and functional activity This

explains why both food analysis and legislation

focus on these two aspects

Foodstuffs are analyzed for a number of

diver-gent reasons Governmental and official agencies

watch over the accomplishment of legal, labeling,

and authenticity requirements This includes

early detection of possible adulterations and

fraudulent practices that could result in economic

losses or consumers damage Food analysis is

also of primary importance for the food industry,

which assesses the quality of the original raw

materials and its maintenance through the

complete processing, transportation, and

conser-vation process Scientific researchers are involved

in the constant update of the methodologies used

to control all the above-mentioned aspects as well

as in the development of new analytical

proce-dures that allow the lowering of the allowed

maximum residue levels (MRLs) of toxic

compo-nents and the inclusion of new ones in current

legislation, the detailed characterization of fooditems, and the development of new foodstuffswith added value Finally, in recent years, therehas been an increasing concern by consumersregarding the quality of food This has partiallybeen motivated by the different scandals origi-nated by food contamination with toxicantsand/or forbidden products but, also and moreimportant, by the nowadays accepted relation-ship between diet and health and the increasingdemand of foodstuffs with added nutritionalproperties The latter frequently results in thedevelopment and addition of new ingredients,whose effect on the original food item at shortand long time should also be tested

It is evident from previous considerations thatfood analysis is an extremely wide field inconstant evolution involving analysis and chem-ical determinations of very different nature andwith widely divergent goals These differencestranslate also to the methods in use for food anal-ysis As shown in Fig 1.1, these methods rangefrom subjective (e.g., organoleptic determina-tions) to objective procedures based on physical,chemical, microscopic, and microbiologic deter-minations Other approaches based on, forexample, biological determinations and personalquestionnaires are also used This volumereviews the current state-of-the-art and last

• OTHER methods:

– Biological methods – Nutritional questionnaire

– Nitrogen content – Carbohydrates – pH, acidity, alcohol, redox…

• Instrumental methods

FIGURE 1.1 Different types of

methods applied for food analysis.

developments regarding chemical methods and

will pay special attention to those based on the

use of modern instrumental analytical

tech-niques that, in many instances, have only

recently started to be applied in this dynamic

research field

1.2 TYPES OF SAMPLES AND THE

ANALYTICAL PROCEDURE

Food analysis demands chemical

determina-tions at very different levels and for different

purposes As previously indicated, for

conven-tional foods, chemical analysis and controls are

applied from independent ingredients and raw

materials to the processed products and

end-products and, when required, to all

interme-diate items to guarantee food quality These

types of determinations become especially

rele-vant during the development and

implementa-tion of new processing and conservaimplementa-tion

procedures, or when developing new formula

and products

As in any other analytical process, the

chem-ical analysis of foodstuffs involves a number of

equally relevant steps with a profound effect

on the validity of the data generated (Fig 1.2)

Although in some cases on-site

determina-tion is possible, most samples have to be

transported to the laboratory for chemical ysis Thereby, in many instances, the first issue

anal-to consider is how many samples (or ples) should be taken, of which size and fromwhere to guarantee the representativeness ofthe subsamples Whether random or purpose-ful, significant consideration needs to be given

subsam-to the sampling prosubsam-tocol in order subsam-to obtain atthe end of the analytical process data meaning-ful and interpretable Sampling is a complexprocess that firstly depends on the nature ofthe matrix to be sampled (solid or liquid), itssize (as a whole or as subsamples), and thegoal of the analysis (e.g., determination ofmain components or trace analysis), just tomention a few parameters In some cases, theprocedure and minimum amount of samplenecessary to develop a particular analysis isclearly stated in current legislations [see, e.g.,(90/642/EEC, 1993) and (2002/63/EC, 2002)for the determination of pesticides residues inproducts of plant and animal origin] In othercases, protocols similar to those set in legal textscan be followed or alternative procedures can beadopted as far as they guarantee the representa-tiveness of the sampling process In-depthdiscussion on this complex matter is out of thescope of this chapter Therefore, the reader isreferred to texts of a more specialized naturefor a detailed discussion on this topic [see, e.g.,

Separation

Chemical reaction

Qualitative analysis Quantitative analysis

Data acquisition

Data reduction

Data interpretation

process.

1.2 TYPES OF SAMPLES AND THE ANALYTICAL PROCEDURE 5

I ANALYTICAL TECHNIQUES

Curren et al., (2002); Woodget and Cooper,

1987]

Samples should remain unaltered during

transportation and storage until the moment of

the analysis As a rule of thumb, samples must

be stored for the shortest possible time When

applicable, stabilization procedures that, for

example, retard biological action, hydrolysis of

chemical compounds, and complexes, and

reduce the volatilization of components and

adsorption effects, should be adopted

Once in the laboratory, samples are typically

subjected to a number of operations and

manip-ulations before instrumental analysis of the

target compounds These several treatments

are grouped under the generic name of sample

preparation The number and nature of these

operations and treatments typically depend on

the nature and anticipated concentration level

of the target compounds, and also on those of

the potential matrix interfering components

and on the selectivity and sensitivity of the

analytical technique selected for final separation

and/or detection Sample preparation would

include from the labeling and mechanical

pro-cessing and homogenization of the received

samples, to any type of gravimetric or

volu-metric measurement carried out Sample

prepa-ration also includes all treatments conducted to

decompose the matrix structure in order to

perform the fractionation, isolation, and

enrich-ment of the target analytes Treatenrich-ments

devel-oped to make the tested analyte(s) compatible

with the detector (e.g., change of phase and

derivatization reactions) and to enhance the

sensitivity of the detector are also considered

part of the sample preparation protocol

Table 1.1presents a simplified overview on

food components and food contaminants

typi-cally considered for chemical analysis In most

instances, these analytes are also the subject of

routine controls Target compounds include

from metals and organometallic species to

vola-tile components The latter include flavor

and fragrances, but also off-flavors that can

create problems with unacceptable food ucts Many main and minor components withnutritional or added functional value, such aslipids, proteins, carbohydrates, vitamins, andantioxidants, are also analyzed for legal,quality, or research reasons In addition, foodadditives, residues, and a large variety ofcontaminants of different origin and natureare nowadays matter of continuous monitoringand control to ensure the accomplishment ofcurrent legislations The increasing social pres-sure for safe foods contributes to support theconstant research efforts carried out to improvethe accuracy and sensitivity of the analyticalmethodologies used to determine these partic-ular compounds

prod-Except for the few cases in which direct mination is feasible (e.g., spectroscopy determi-nation of main food components in combinationwith chemometrics, see Chapter 2;control process by low intensity ultrasounds, seeChapter 5; use of sensors, see Chapter 7), the

deter-TABLE 1.1 Overview of the Typical Food Components

• Food additives and contaminants:

• Pesticides and veterinarian drugs

• Contaminants PCBs, PCDD/Fs, PAHs, PBDEs, phthalates, mineral oils.

• Mico- and phyto-toxins

• Migrants from packaging materials

• Process and/or storage residues

• Metallic and organometallic species

determination of the analytes mentioned inTable

1.1 requires some type of sample preparation

before instrumental analysis, almost irrespective

of the technique selected for final

separation-plus-detection In the simplest case, this consists

of the usually quantitative (i.e., exhaustive and

nonselective) extraction of the compound(s) of

interest from the matrix in which they are

entrap-ped, a fractionation or clean-up step to isolate

them from other coextracted materials, and

a final concentration of the purified extracts to

ensure analyte(s) accurate detection As in other

application areas, in food analysis, the several

analytical steps involved in such procedures

are most frequently carried out off-line, which

make them tedious and time consuming In

general, the complexity of the procedures

increases as the concentration of the target

compound decreases and so the possibility of

loss and contamination of the analyte due to

the continual manual manipulation of the

extracts In recent years, much effort has been

devoted to eliminating these drawbacks This

has led to the development of faster and more

powerful and/or versatile extraction techniques,

often incorporated from other research areas,

such as environmental and molecular analysis

(see e.g., Chapters 6, 7 and 13) These include,

for example, automated purge-and-trap (P&T),

solid-phase microextraction (SPME), and

stir-bar-sorptive extraction (SBSE) for the analysis

of volatile components (Table 1.2); a number of

solvent-based microextraction techniques

espe-cially adapted for the determination of

semi-and nonvolatile analytes in liquid sample; other

techniques suitable for the treatment of

viscous and (semi-) solid samples, such as matrix

solid-phase dispersion (MSPD), widely used

enhanced fluid/solvent extraction techniques,

such as supercritical fluid extraction (SFE),

pressurized liquid extraction (PLE), subcritical

water extraction (SWE), and

microwave-assisted extraction (MAE) and

ultrasound-as-sisted extraction (USE); and also microfluidic

devices, DNA arrays, real-time PCR, and other

molecular techniques The latter approacheswill be the matter of subsequent chapters withinthis volume Thereby, in this chapter, the lasttrends in the use of some of the modern analyt-ical techniques previously mentioned for foodanalysis will be revised through selected repre-sentative application examples

1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD

ANALYSIS

Every single physico-chemical treatmentcarried out to isolate the analytes from othermatrix components that could interfere duringtheir instrumental determination and/or toincrease their concentration in the extract sub-jected to analysis is considered a step of thesample preparation protocol According to thisconsideration, one can conclude that most ofconventional and official sample preparationmethods (AOAC, 1990; Nollet, 1996) in use forfood analysis are long, laborious, and highlymanipulative multistep procedures prone toloss, degradation, and/or contamination of thetarget analytes Therefore, in this field, samplepreparation is a key part of the analyticalprocess with a profound effect on (i) the timerequired to complete the analytical process, (ii)the cost of the determination in terms ofsolvents and sorbents consumption, and (iii)the validity of the final result

Again as in other application areas, sampletreatment is considered the bottleneck of theanalytical methodologies in use for food anal-ysis It is estimated that 60e80% of the workactivity and operating costs in the analyticallaboratories is spent in preparing samples forintroduction into the analytical system selectedfor instrument determination It is also esti-mated that this part of the analytical process isresponsible for more than 50% of the error asso-ciated to the final reported data These figuresexplain the efforts carried out during the last1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD ANALYSIS 7

I ANALYTICAL TECHNIQUES

TABLE 1.2 Overview of Selected Analytical Techniques in Use for Food Analysis

Base of the technique Name of the technique (acronym)

Purge of volatile compounds Static and dynamic headspace (S/D HS)

Purge-and-trap (P&T) Programmed thermal vaporization (PTV) Direct thermal desorption (DTD) Simultaneous distillationeextraction (SDE) Solvent extraction Liquideliquid extraction (LLE)

In-vial liquideliquid extraction (in-vial LLE) Single-drop microextraction (SDME) Liquid-phase microextraction (LPME) Dispersive liquideliquid microextraction (DLLME) Extracting syringe (ESy)

Sorption extraction

Liquid desorption Solid-phase extraction (SPE)

Open-tubular-coated capillaries Solid-phase dynamic extraction In-tube solid-phase microextraction (in-tube SPME) Fiber-in-tube solid-phase extraction ( fiber-in-tube SPE) Single short column (SSC)

Solid-phase microextraction (SPME) Dispersive solid-phase extraction (dSPE) Molecular imprinted solid-phase extraction (MISPE) Restricted access medium (RAM)

Thermal desorption Solid-phase microextraction (SPME)

Stir-bar-sorptive extraction (SBSE) Matrix solid-phase dispersion Matrix solid-phase dispersion (MSPD)

Enhanced fluid/solvent extraction Supercritical fluid extraction (SFE)

Pressurized liquid extraction (PLE) Subcritical water extraction (SWE) Microwave-assisted extraction (MAE) Ultrasound-assisted extraction (USE)

decades to develop analytical approaches that

represent a faster, more automated,

cost-effec-tive, and greener alternative to the previously

mentioned traditional protocols

Solid-phase microextraction (SPME) is

a miniaturized technique that fulfills most of

these requirements In SPME, the analyte(s)

is(are) adsorbed onto a fused-silica fiber coated

with an appropriate sorbent layer by simple

exposure of the fiber for a preselected time to

the headspace (HS) of the sample or by direct

immersion in a liquid sample Since its

intro-duction in 1990 by Pawliszyn’s group (Arthur

and Pawliszyn, 1990) as a (virtually)

solvent-free preconcentration technique, SPME has

profusely been used in many application fields

including food analysis Here, its primary use

has been the preconcentration of volatile

analy-tes from liquid, semi-solid, and solid samples,

for which it has been demonstrated to be

a simple, rather selective, and relatively fast

(under nonequilibrium conditions) technique

SPME has been used for different application

studies such as lipids oxidation and protein

degradation during storage of soup powder(Raitioa et al., 2011), and the evaluation of thetraceability of grapes origin (Rocha et al.,

2007) In this latter work, a fused SPME silicafiber coated with Carbowax-divinylbenzenewas used in the HS-mode to establish the mono-terpenoid profile of Vitis vinifera L cv ‘Fernao-Pires’ white grape The use of HS-SPMEcoupled with comprehensive two-dimensionalgas chromatography with time-of-flight massspectrometry (GC
GC-ToF MS) alloweddetermining 56 monoterpenoids in grapes.Among them, 20 were reported for the firsttime in this fruit A typical example of theresults obtained is shown in Fig 1.3 Theauthors concluded that, as monoterpenoidsare secondary metabolites whose synthesis isencoded by variety-related genes, the terpe-noid profile may be used as a way to tracegrape varietal origin

Recently, stir-bar-sorptive extraction (SBSE)has been found to be advantageous as comparedwith conventional extraction techniques likesimultaneous distillationeextraction (SDE) or

18

19

36

10 3

20

22 25

27 34

37 42 45

44 43

49

47 48

29 30

2

22 25

2

3

3 4

20

22 26

40 46 49

48

53

51 52 32

30

29

27 24

17 16

15 14 13

12

11 10 8 6 3

1

Esters

Aldehydes Terpendiols

Oxides Tertiary Monoterpenols

25

37 43

45

54

Primary Monoterpenols

20

22 25

27 34

37 42 45

44 43

49

47 48

29 30

20

22 26

40 46 49

48

53

51 52 32

30

29

27 24

17 16

15 14 13

12

11 10 8 6 3

1

Esters

Aldehydes Terpendiols

Oxides Tertiary Monoterpenols

25

37 43

45

54

Primary Monoterpenols

FIGURE 1.3 GC
GC contour plot corresponding to ions m/z 93, 121 and 136 Bands or clusters formed by structurally related compounds are highlighted.

1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD ANALYSIS 9

I ANALYTICAL TECHNIQUES

direct HS, and more modern sample

prepara-tion techniques, such as SPME, for the

determi-nation of unknown taints in food (Ridgway

et al., 2010) SDE uses larger volumes of solvent

than SBSE, which provides improved

detect-ability as compared with HS and SPME and

also minimizes the potential for contamination

from external laboratory sources In general,

SBSE provided better results than these

estab-lished techniques, although the optimized

method was not feasible for the determination

of methyl methacrylate and hexanal Other

examples of the use of SBSE and a discussion

of the advantages and limitations of this

tech-nique as compared with SPME, SPE, and other

conventional sample preparation techniques

can be found inOlariu et al (2010)

Liquideliquid extraction (LLE) is the

tech-nique of choice in most official methods

However, some of these procedures are

frequently revisited in an attempt to expand their

application field by incorporating new target

compounds into the analysis (Mol et al., 2007)

The straightforward nature of most LLE methods

would suggest that their adaptation for

imple-menting some of the newly developed solvent

microextraction techniques is a relatively easy

goal, attainable by simple scaling down of the

original procedures Depending on the

applica-tion, practice can be slightly more complicated

However, the high sensitivity provided by

many modern instrumental techniques and the

increased use of these miniaturized techniques

in food analysis demonstrate the feasibility of

the approach [see, e.g., Asensio-Ramos et al.,

(2011)]

Single-drop microextraction (SDME) was the

first solvent-based microextraction technique

introduced and has up to now been one of the

most profusely used for food analysis Typical

applications involving SDME are presented in

Table 1.3

SDME can be used as a two-phase system, as

in the direct-immersion (DI) and drop-to-drop

microextraction (DDME) approaches, or as a

three-phase system, as in the HS mode or inthe more recently introduced liquideliquideliquid microextraction (LLLME) In its simplestconfiguration, a single microdrop of a water-insoluble solvent suspended at the tip of a GCsyringe is either immersed in an aqueoussample (DI mode) or exposed to the HS of

a sample contained in a vial Strategies such asstirring, heating, and/or salting out the solu-tion, and derivatization of the target compoundsare frequently used to speed up the extractionprocess Once the extraction time is completed,the drop is withdrawn into the syringe and theenriched solvent is transferred to the systemselected for instrumental analysis without anyadditional treatment HS has been used for pre-concentration of volatile analytes or derivatives.Meanwhile, the two-phase approaches areparticularly suitable for the analysis of less vola-tile and relatively polar compounds in pristinesamples

DDME is a modification of the DI-SDMEprocedure that has been used for the fast, inex-pensive clean-up and quantitative preconcen-tration of different analytes from aqueoussolutions with minimum sample consumption

In a representative application,Shrivas and Wu(2007)used DDME with chloroform (0.5 mL) forthe rapid determination of caffeine in one drop

of beverages and foods, i.e., 7 mL The tion took only 5 min and was carried at roomtemperature and without salt addition The

gas chromatographyemass spectrometry(GCeMS) method exhibited good linearitybetween 0.05 and 5.0 mg/mL with correlationcoefficient of 0.980, recoveries above 97%,

a relative standard deviation (RSD) of 4.4%,and a limit of detection (LOD) of 4.0 ng/mL.DDME avoided the main shortcomings ofconventional methods of caffeine extraction,like large amount of organic solvent andsample consumption and long sample pretreat-ment process The authors proposed the opti-mized DDME-based procedure as a simple,

TABLE 1.3 Selected Applications of Solvent Microextraction Techniques

Sample type Analytes Extraction solvent (mL) Extractionmode Extractiontime (min) LOD

a (mg/L,ng/g) Reference SDME

Two-phase system

(2008)

Mousavinia (2006) Mineralized rice flour Cadmium Dithizone (0.01 M) in

chloroform (3)

Digested defatted milk

powder

(2009) Degassed and filtrated

beverages, chocolate

(2007) Three-phase system

L extract

26 Essential oil compounds

10 6 Wang et al (2009) Mussel extract 3 Butyltin

TABLE 1.3 Selected Applications of Solvent Microextraction Techniques (Cont’d)

Sample type Analytes Extraction solvent (mL) Extractionmode Extractiontime (min) LOD

a (mg/L,ng/g) Reference HF(2/3)LPME i

Alcoholic beverages 51 Multiclass

pesticides

(2008) Filtrated orange juice 2 Fungicides 2-Octanone þ HCl,

10 mM (20)

(2010) Aqueous green tea

and tea leave extracts

6 Organosulfur pesticides

(2008) Filtrated fruit juices 7 Phenolic acids Hexyl acetate þ NaOH,

0.02 M (8)

(2010) Diluted milk, beer, juice Volatile organic

selenium species

10 3 Ghasemi et al (2011) Mineralized oyster

(reference material)

(2010) Buffered bovine milk 3 Tetracycline

antibiotics

Aliquat 336 in 1-octanol þ H 3 PO 4 , 0.1 M (pH ¼ 1.6); NaCl, 1 M (24)

Banana extract 8 Multiclass

Extracted and purified

milk extract

Extracted and purified

food extracts (milk, egg

yolk, olive oil)

(2009)

Extracted and

purified porcine tissue

Mineralized rice, tea,

defatted milk powder

i Hollow-fiber liquid-phase microextraction.

j For this technique, the extraction solvent column corresponds to the acceptor phase (followed, when applicable, by the back-extraction phase) Otherwise specified, the donor phase is the corresponding buffered sample or sample extract.

fast, and feasible diagnosis tool for caffeine in

food and beverages

Application of SDME to the analysis of polar

compounds required a modification that

resulted in a three-phase SDME system named

LLLME In this approach, the deionized polar

analytes were preconcentrated from the

aqueous sample in a few microliters of organic

phase and subsequently back-extracted in an

aqueous microdrop that acted as receiving

phase Up to now, the technique has mainly

been used for the analysis of aqueous samples

and biological fluids To the best of our

knowl-edge, only one study has reported on its

applica-tion to food analysis The study (Zhu et al., 2010)

proposed the combined use of LLLME with

capillary electrophoresis (CE) for the on-line

purification and preconcentration of adenine

from green tea extracts

Hollow fiber-protected two-phase liquid

mi-croextraction (HF(2)LPME) was introduced by

He and Lee (1997)with the name of liquid-phase

microextraction In its simplest version, the

tech-nique involves a small-diameter microporous

polypropylene tube (the hollow fiber), typically

sealed at one end, to contain the organic

extract-ing solvent The open end of the hollow fiber is

attached to a syringe needle used to fill the fiber

with the organic solvent Once filled, the fiber is

immersed in the vial containing the investigated

aqueous sample to allow analyte migration

through its walls After a preselected extraction

time, the solvent is withdrawn with the syringe

and transferred to the instrument selected for

analyte determination, typically gas

chromatog-raphy (GC) The hollow fiber can be considered

to act as a membrane Consequently, this

tech-nique is more appropriate for the analysis of dirty

aqueous samples than SDME Due to the higher

stability of the solvent, contained in the hollow,

it also allows higher stirring rates than SDME

On the contrary, HF-LPME typically used to

involve larger extractant volumes (Table 1.3)

and longer extraction times than SDME

(20e60 min vs 5e15 min with SDME) In its

three-phase format (HF(3)LPME), the analytespreconcentrated in the water-immiscible organicsolvent used to fill the pores of the hollow fiberpolymer are subsequently extracted to anaqueous acceptor phase that is placed in thelumen of the fiber The HF(3)LPME technique istypically used to extract water-soluble analytesfrom aqueous matrices and, because the finalacceptor solution is aqueous, liquid chromatog-raphy (LC) and CE are usually preferred for finalinstrumental determination of the tested analy-tes During the last few years, a number of studieshave demonstrated the feasibility of the tech-niques for the determination of analytes of verydifferent nature in food and beverages Applica-tions include the analysis of micropollutants inalcoholic drinks (Plaza Bolan˜os et al., 2008),orange juice (Barahona et al., 2010), and otherbeverages (Xiong and Hu, 2008); phenoliccompounds in fruit juices (Saraji and Mousavi,

2010), antibiotics in bovine milk (Shariati et al.,

2009), and metallic (Abulhassani et al., 2010)and organometallic (Ghasemi et al., 2011) species

in complex foodstuffs

In dispersive liquideliquid microextraction(DLLME), the investigated aqueous sample(up to 10 mL) is extracted with a small amount

of a water-immiscible extraction solvent cally 10e50 mL) dissolved in 0.5e2 mL of

(typi-a w(typi-ater-soluble solvent The technique c(typi-an beconsidered a modification of a miniaturizedLLE in which extraction is favored by the forma-tion of small microdrops of the water-immis-cible solvent by fast injection of the mixture oforganic solvents into the water with a syringe.The enriched organic phase is then separatedfrom the aqueous sample by centrifugation orfreezing (depending on its density) and directlysubjected to instrumental analysis, typically by

GC Application to polar analytes requiresprevious pH adjustment and/or in situ derivati-zation, which can be accomplished by directaddition of the derivatization agent to thesample or by dispersion together with theextraction solvent Since its introduction in

2006 (Rezaee et al., 2006), this miniaturized and

green, but highly manipulative technique, has

profusely been used in different application

areas In food analysis, the DLLME has been

demonstrated to be a valuable alternative to

large-scale conventional procedures for the

determination of relatively abundant food

components, such a cholesterol (Daneshfar

et al., 2009), and also for the analysis of trace

organic (Cunha et al., 2009; Liu et al., 2011b)

and inorganic (Wen et al., 2011) contaminants

and other illegal substances (Liu et al., 2011a)

Several recent studies have reported on the

use of ionic liquid as extractant in DLLME,

a trend also observed on SDME and HF(2/3)

LPME (Table 1.3) These examples demonstrate

that room-temperature ionic liquids are a

valu-able alternative to classical organic volatile

solvents for the extraction of both organic and

inorganic compounds that, apart from greening

the analytical process, efficiently contribute to

reduce the exposure of the analyst to toxic

solvents Ionic liquids can directly be applied

to aqueous samples The analysis of solid

matrices is only possible after extraction of the

target analytes from the matrix and dilution of

the extract in water Ravelo-Pe´rez et al (2009)

used this approach for the determination of

eight pesticides belonging to classes different

from bananas In this method, the homogenized

fruit sample (1 g) was extracted with acetonitrile

and, after evaporation and reconstitution of the

extract in 10 mL of water, the target compounds

were preconcentrated by DLLME using

[HMIM][PF6] (88 mg) as extractant and

meth-anol (714 mL) as disperser solvent The ionic

liquid was recovered after centrifugation at

4000 rpm (20 min), diluted in acetonitrile, and

analyzed without any further treatment by

LC-DAD Figure 1.4 shows the typical

chro-matograms obtained for (A) a spiked and (B)

a nonspiked banana Acceptable mean

recov-eries in the 53e97% range, with RSD values

lower than 9%, and LODs (0.32e4.7 mg/kg)

below the MRLs set in current legislations

were obtained in all instances These analyticalfigures of merit would prove the validity ofthe optimized method for the intended determi-nation, although the observed severe matrixeffect made the use of matrix-matched calibra-tion mandatory

Solid-phase extraction (SPE) is the mostwidely used technique for the treatment ofaqueous samples and extracts in laboratories

A large variety of sorbents, ranging from sical sorbents, such as silica, florisil, and C8 orC18, to modern cross-linked polymers arenowadays commercially available in differentformats, including conventional SPE cartridgesand disks for off-line and on-line analysis aswell as 96-well plates As illustrated in severalreviews (Beyer and Biziuk, 2008; Buldinia

clas-et al., 2002; Ihnat, 2003; Kinsellaa clas-et al., 2009;Ridgway et al., 2007; Rostagno et al., 2010), all

of them have been used for food analysis.Current trends in the use of SPE for foodanalysis agree with those observed in closelyrelated research areas, such as environmentalanalysis These include the preference for theso-called universal sorbents, i.e., those able tosimultaneously retain polar and nonpolaranalytes, in an attempt to increase the number

of analytes monitored in a single analysis;the use of highly cross-linked polymers toimprove the retention of very polar analytes;and the use of very selective sorbents based

on restricted access media (RAM) or molecularimprinted polymers (MIPs) (Turiel andMartı´n-Esteban, 2010) Food analysis is atpresent benefited by the development experi-enced in the last decade in the field ofnanotechnologies In a representative study,Lo´pez-Feria et al proposed the use of carbonnanotube-based solid-phase extraction for thecontrol of multiclass pesticides in virgin oliveoils (Lo´pez-Feria et al., 2009) Carboxylatedsingle-walled carbon nanotubes (SWCNs)were preferred to multiwalled carbon nano-tubes for the application Once optimized,the method consisted of the direct elution of1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD ANALYSIS 15

I ANALYTICAL TECHNIQUES

the investigated olive oil diluted with n-C6

(1:5,v/v) through an SPE column containing

30 mg of the selected nanotubes After

washing the column with 3 mL of n-C6, the

analytes were eluted with 0.5 mL of ethyl

acetate The extract was finally concentrated,

reconstituted on methanol, and analyzed byGCeMS Complete sample preparation wascarried out in less than 8 min and the SPEcolumn could be reutilized more than 100times The low LODs achieved (in the 1.5and 3.0 mg/L) allowed the application of the

8(a)

method to control the target pesticides in very

restrictive samples, such as the ecological

virgin olive oil

Probably the most successful development

introduced in the last few years in the field of

SPE has been the method known as QuEChERS

The acronym applies to quick, easy, cheap,

effec-tive, rugged, and safe, which is supposed to

describe the main merits of the analytical

proce-dure introduced by Anastassiades et al (2003)

for the determination of pesticides in fruits

and vegetables The method is a multistep

procedure based on dispersive solid-phase

extraction (dSPE) In its basic scheme for

pesti-cide analysis in fruits and vegetables (Fig 1.5)

(Wilkowska and Biziuk, 2011), the method

involves the initial sample treatment with

magnesium sulfate to promote water separationfrom the organic solvent, followed by treatmentwith primary secondary amine (PSA) to removepolar components, such as organic acids, somesugars, and polar pigments Other protocolsinclude sample shaking with graphitizedcarbon black (GCB) to eliminate sterols andpigments such as chlorophyll

The rapid acceptation of this fast and efficientsample preparation protocol promoted its quickadaptation for other types of analysis, includingdifferent application such as the analysis ofnonpolar microcontaminants (Ramalhosa et al.,

2009) and acrylamide in different food items(Mastovska and Lehotay, 2006), drugs in animaltissues (Stubbings and Bigwood, 2009), andblood (Plossl et al., 2006)

vortexing immediately for 1 min

shaking by hand or with the vortex mixer for 30 s and centrifugation of extract (or a batch of extracts) for about 1 min

vortexing for 30 s and centrifugation of extract (or a batch

of extracts) for about 1 min

addition 10 mL of acetonitrile and shaking the sample vigorously for 1 min using the vortex mixer at maximum speed

Weighing 10 g of the well-chopped sample into a 40 mL Teflon centrifuge tube

Addition of 4 g anhydrous MgSO4 1 g and NaCl

Addition of ISTD solution

Transfering a 1 mL aliquot of the upper acetonitrile layer into

a microcentrifuge vial containing 25 mg PSA sorbent and 150

mg anhydrous MgSO4 and capping tightly

Addition of 5% aq formic acid (if necessary)

Final determination (usually GC–MS) FIGURE 1.5 Main steps in QuEChERS procedure for determining pesticides in fruits and vegetables.

1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD ANALYSIS 17

I ANALYTICAL TECHNIQUES

dSPE has also benefited for the development

of new materials Chen et al (2009) prepared

a magnetic molecularly imprinted polymer for

the separation of tetracycline antibiotics from

egg and tissue samples by dSPE The

satisfac-tory results obtained with this method as

compared with more conventional

configura-tions such as MIP-SPE and MIP-SPME (Table

1.4), together with the simplicity of the

opera-tion methodology and the possibility of

recov-ering the magnetic particles with a simple

magnet, make this novel approach an

inter-esting alternative for sample preparation

Matrix solid-phase dispersion (MSPD) is

a widely accepted technique for the treatment

of liquid, viscous, and (semi-) solid samples In

MSPD, the extraction and (preliminary)

clean-up of the target analytes is carried out in a single

step and in a column format The column

config-uration simultaneously contributes to simplify

the analytical process and to avoid the emulsion

problems associated to most of the conventional

LLE-based procedures When the sorbent

disper-sant and the extraction solvent protocol are

properly selected, MSPD can yield analyze extracts that, in the case of foodstuffs,are usually processed by GC or LC

ready-to-In food analysis, MSPD has mainly been usedfor the determination of trace organic micropol-lutants and, in particular, of pesticides (Barker,2007; Bogialli and Di Corcia, 2007; Gilbert-Lopez

et al., 2009; Kristenson et al., 2006) For this type

of application, sorbents used for sample sion range from classical ones (e.g., alumina, flo-risil, carbon, or C8) to new materials likemultiwalled carbon nanotubes (Guan et al.,

disper-2011) or highly selective dedicated sorbents(Yan et al., 2011)

Most recent trends focus on the tion of the MSPD process (Kristenson et al.,2001; Ramos et al., 2009) and/or the combineduse of MSPD with one or several of the previ-ously described novel sample preparation tech-niques in order to improve the efficiency and/

miniaturiza-or selectivity of the MSPD process In an tive example,Yan et al (2011)proposed the use

illustra-of a new synthesized kind illustra-of aniline-naphtholmolecularly imprinted microspheres (0.2 g)

TABLE 1.4 Comparison of QuEChERS Method with Magnetic MIP with the Results Obtained by Using MIP-SPE

and MIP-SPME for the Determination of Tetracycline Antibiotics

MIP-SPME 5 or 10 min for

LC-72e94 3e6 1.5e3.5 100 Hu et al.

for SPE clean-up

LC-UV 66e69 <8 Not

mentioned

Not mentioned

Caro et al (2005)

Adapted from Chen et al (2009)

selective for Sudans as dispersant for

miniaturi-zed MSPD of 0.1 g of egg yolk After washing

the MSPD column with 4 mL of methanol:water

(1:1, v/v), analytes were quantitatively extracted

with 3 mL of acetone:acetic acid (95:5, v/v) The

concentrated eluent (1 mL) was used as

disper-sive solvent for DLLME The mixture was

shaken and ultrasonicated to form a

homoge-neous cloudy solution Phase separation was

subsequently performed by centrifugation at

4000 rpm for 10 min The four studied Sudan

dyes were simultaneously determined by

LC-UV after concentration of the corresponding

enriched phase Figure 1.6 shows a schematic

diagram of the complete sample preparation

procedure (Yan et al., 2011) The method showed

a good linearity for all target analytes in theinvestigated 0.02e2.0 mg/g range (r2 0.9990),with recoveries better than 87% and RSDsbelow 6%

The main application fields of the techniquesbased on the use of compressed fluids, namelysupercritical fluid extraction (SFE) and pressur-ized liquid extraction (PLE), so-called subcrit-ical water extraction (SWE) when water isused as extractant, in food analysis are the isola-tion of relevant natural compounds and of func-tional products (Mendiola et al., 2007)

Probably some of the most widely knownindustrial applications of SFE in food analysis

Frit

Conic tube Vacuum

Withdraw Syringe

Eluent

Deionized water

(k) (j)

(i) (h)

(g) (f)

(b) (a)

Ultrasonic cleaner

Water

Centrifugation

FIGURE 1.6 Schematic of the MIP/MSPD combined with DLLME proposed for the simultaneous determination of four Sudan dyes in egg yolk (a) Blending of the sample with the selective MIP (MIM); (b) transfer of the blended sample to the column; (c) completed MSPD column; (d) washing of the MSPD column and elution of the test analytes; (e) eluent to be evaporated, (f) injection of the extractant into the eluent for DLLME; (g) addition of deionized water into the DLLME extractantedispersant mixture; (h) formation of the emulsion assisted by ultrasounds; (i) emulsion of the ternary mixture; (j) phase separation by centrifugation; and (k) collection of the high-density extractant.

1.3 TRENDS IN SAMPLE PREPARATION FOR FOOD ANALYSIS 19

I ANALYTICAL TECHNIQUES

are the extraction of caffeine from coffee and tea

and of cholesterol from, e.g., egg However, the

particular features of this technique make it

suit-able for many other applications Thereby, SFE

with carbon dioxide modified with 35% of

methanol and combined off-line with GCeMS

has been used for obtaining the amino acid

profiles of genetically modified maize and

soybean (Bernal et al., 2008) Comparison of

these profiles with those obtained for their

cor-responding isogenic nontransgenic varieties

proved that the latter seemed to have higher

content of several amino acids

The distinguished advantages of SFE for

automatic sample treatment and its relatively

simple at-line or on-line coupling with different

separation-plus-detection instruments have

made SFE the technique of choice for a variety

of application studies

SFE coupled at-line with CE equipped with

fluorimetric detection (CE-FD) has been used

for the determination of flavin vitamins in food

samples (Zougagh and Rı´os, 2008) The nonpolar

nature of supercritical carbon dioxide was used

for the initial elution of the nonpolar interference

compounds existing in the matrix; then, the

extraction of the studied water-soluble vitamins

was achieved by modification of the polarity of

the extracting agent with 5% methanol Extracts

were clean enough to allow direct CE-FD

anal-ysis In another interesting study, SFE-LC was

used for the determination of air- and

light-sensitive food components, such as lycopene

(Po´l et al., 2004) Here, a single monolithic

column was used for trapping and subsequent

chromatographic separation of target analytes

The method showed a linear response over the

studied range of 0.1e2.5 mg, a good repeatability

(RSD, 3.9%), and sensitivity (LOD, 0.5 ng)

Complete analysis was done in only 35 min

A typical chromatogram demonstrating the

performance of the SFE-LC method proposed

for real-life applications is shown inFig 1.7

Despite the increasing acceptance of PLE

(and SWE) as fast and relatively green

techniques for food extraction (nez et al., 2005; Mendiola et al., 2007), the devel-opment of equivalent hyphenated systems withcommercial PLE system can still be considered

Carabias-Martı´-an unachieved goal The main reason is the tive large volume of extractant used by thesePLE devices (typically more than 35 mL), whichmakes difficult the coupling with both thesubsequent clean-up step (if required) and theselected instrumental chromatographic or sepa-ration technique Probably, the most plausiblestrategy to circumvent this limitation could bethe use of a miniaturized PLE setup [see, e.g.,Ramos et al (2007)] Although these types ofdevices represent also the best alternative forthe PLE treatment of size-limited samples, tothe best of our knowledge, no commercially

rela-FIGURE 1.7 Typical chromatogram obtained for

a tomato extract Peak identification: (1) b-carotene; (2) lycopene; (3) trans-lycopene; and (4) cholesterol (internal standard) Residual carbon dioxide inside the trapping- separation monolithic column showed as a peak eluting at 1.2 min.

cis-available system of these characteristics is

avail-able yet

Apart from miniaturization, the main recent

highlights concerning PLE are the application of

sequential elution protocols during PLE and the

preference for the so-called selective PLE The

former approach is usually selected when

the study aimed to obtain information as

complete as possible regarding sample

composi-tion Sequential PLE has been used in

combina-tion with either Fourier transformeion cyclotron

resonanceemass spectrometry (FTeICReMS)

or capillary electrophoresisetime-of-flightemass

spectrometry (CEeTOFeMS) to study the

metab-olomics of genetically modified organisms (Leon

et al., 2009) Using this sophisticated strategy, the

authors found differences in the metabolite levels

of three transgenic maize varieties compared with

their wild isogenic lines suggesting specific

metabolic pathways

In selective PLE, sorbent(s) used for

purifi-cation of the food extracts in conventional

sample preparation are packed at the bottom

of the extraction cell to perform in-cell

purifica-tion In most instances, ready-to-analyze

extracts are obtained with concentration as

the only required treatment; before final

instru-mental determination of the target compounds,

at least a miniaturized PLE system (Ramos

et al., 2007) or a very sensitive detector was

used

1.4 CONCLUSIONS

Foodstuffs are complex mixtures of volatile,

inorganic, and organic components at very

different concentration levels Food chemical

characterization requires the analysis of all

a variety of macromolecules such as proteins,

other macronutrients like carbohydrates and

lipids, natural bioactive components such as

polyphenols, aroma and flavor components,

inorganic micronutrients and organometallic

compounds, as well as undesirable residues of

other small molecules introduced duringproduction, processing, storage and/or trans-port of food, including contaminants, i.e.,plasticizers, pesticides, persistent organicpollutants, veterinary drugs, and toxins.Sample preparation is, in one way or another,required almost in all these types of analyses.Because of the still rather traditional protocolsused for many of these determinations, thereare many opportunities for improvement andanalytical development The new analyticaldemands derived from current legislations con-cerning (and constantly affecting) food routinecontrol and monitoring programs to ensurehuman health protection also contribute topromote new improvements and developments

in the field, with increasing automation ably being one of the main requirements.Finally, as in many other application fields,the large amount of (frequently toxic) wastesgenerated during sample preparation of food-stuffs demands the development of alternativegreener analytical procedure also in thisresearch area

prob-Acknowledgments

Author thanks MICINN for project AGL2009-09733 and CM for program S-2009/AGR-1464 (ANALISYC-II).

References

90/642/EEC, D., 1993 Annex I to Directive 90/642/EEC [Part

of products to which maximum limits apply], as tioned by Directive 93/58/EEC (OJ L 211, 23.8.1993, p.6) 2002/63/EC, C d, 2002 Commission directive 2002/63/EC

men-of 11 July 2002 establishing community methods men-of sampling for the official control of pesticides in and on products of plants and animal origin and repealing Directive 70/700/EEC Official Journal of European Communities L 187, 30e43 16/07/2002.

Abulhassani, J., Manzoori, J.L., Amjadi, M., 2010 Hollow fiber based-liquid phase microextraction using ionic liquid solvent for preconcentration of lead and nickel from environmental and biological samples prior to determi- nation by electrothermal atomic absorption spectrometry Journal of Hazardous Materials 176, 481e486.

I ANALYTICAL TECHNIQUES

Anastassiades, M., Lehotay, S., Stajnbaher, D., Schenk, F.,

2003 Fast and easy multiresidue method employing

acetonitrile extraction/partitioning and “dispersive

solid-phase extraction” for the determination of

pesti-cide residues in produce Journal of AOAC International

86, 412e431.

AOAC, 1990 In: Helrich, K (Ed.), Official Methods of

Analysis of the AOAC AOAC, USA.

Arthur, C.L., Pawliszyn, J., 1990 Solid phase

micro-extraction with thermal desorption using fused silica

optical fibers Analytical Chemistry 62, 2145e2148.

Asensio-Ramos, M., Ravelo-Pe´rez, L.M.,

Gonza´lez-Curbelo, M.A., Herna´ndez-Borges, J., 2011 Liquid

phase microextraction applications in food analysis.

Journal of Chromatography A doi:10.1016/j.chroma.

2011.05.096.

Barahona, F., Gjelstad, A., Pedersen-Bjergaard, S.,

Rasmussen, K.E., 2010 Hollow fiber-liquid-phase

microextraction of fungicides from orange juices Journal

of Chromatography A 1217, 1989e1994.

Barker, S.A., 2007 Matrix solid phase dispersion (MSPD).

Journal of Biochemical and Biophysical Methods 70,

151e162.

Bernal, J.L., Nozal, M.J., Toribio, L., Diego, C., Mayo, R.,

Maestre, R., 2008 Use of supercritical fluid extraction

and gas chromatographyemass spectrometry to obtain

amino acid profiles from several genetically modified

varieties of maize and soybean Journal of

Chromatog-raphy A 1192, 266e272.

Beyer, A., Biziuk, M., 2008 Applications of sample

prepa-ration techniques in the analysis of pesticides and PCBs

in food Food Chemistry 108, 669e680.

Bogialli, S., Di Corcia, A., 2007 Matrix solid-phase

disper-sion as a valuable tool for extracting contaminants from

foodstuffs Journal of Biochemical and Biophysical

Methods 70, 163e179.

Buldinia, P.L., Riccib, L., Sharma, J.L., 2002 Recent

appli-cations of sample preparation techniques in food

anal-ysis Journal of Chromatography A 975, 47e70.

Carabias-Martı´nez, R., Rodrı´guez-Gonzalo, E.,

Revilla-Ruiz, P., Herna´ndez-Me´ndez, J., 2005 Pressurized liquid

extraction in the analysis of food and biological samples.

Journal of Chromatography A 1089, 1e17.

Caro, E., Marce´, R.M., Cormack, P.A.G., Sherrington, D.C.,

Borrull, F., 2005 Synthesis and application of an

oxytetracycline imprinted polymer for the solid-phase

extraction of tetracycline antibiotics Analytica Chimica

Acta 552, 81e86.

Chen, L., Liu, J., Zeng, Q., Wang, H., Yu, A., Zhang, H.,

Ding, L., 2009 Preparation of magnetic molecularly

imprinted polymer for the separation of tetracycline

antibiotics from egg and tissue samples Journal of

Chromatography A 1216, 3710e3719.

Cunha, S.C., Fernandes, J.O., Oliveira, M.B.P.P., 2009 Fast analysis of multiple pesticide residues in apple juice using dispersive liquideliquid microextraction and multidimensional gas chromatographyemass spec- trometry Journal of Chromatography A 1216, 8835e8844 Curren, M.S.S., King, J.W., 2002 Sampling and sample preparation for food analysis In: Pawliszyn, J (Ed.), Sampling and Sample Preparation for Field and Labo- ratory Fundamentals and New Directions in Sample Preparation Comprehensive Analytical Chemistry, vol XXXVII Elsevier, Amsterdam.

Daneshfar, A., Khezeli, T., Lotfi, H.J., 2009 Determination

of cholesterol in food samples using dispersive liquideliquid microextraction followed by HPLCeUV Journal of Chromatography B 877, 456e460.

Fan, Z., Zhou, W., 2006 Dithizoneechloroform single drop microextraction system combined with electrothermal atomic absorption spectrometry using Ir as permanent modifier for the determination of Cd in water and bio- logical samples Spectrochimica Acta Part B: Atomic Spectroscopy 61, 870e874.

Gao, S., You, J., Zheng, X., Wang, Y., Ren, R., Zhang, R., Bai, Y., Zhang, H., 2010 Determination of phenylurea and triazine herbicides in milk by microwave assisted ionic liquid microextraction high-performance liquid chromatography Talanta 82, 1371e1377.

Ghasemi, E., Sillanpa¨a¨, M., Najafi, N.M., 2011 Headspace hollow fiber protected liquid-phase microextraction combined with gas chromatographyemass spectros- copy for speciation and determination of volatile organic compounds of selenium in environmental and biological samples Journal of Chromatography A 1218, 380e386.

Gilbert-Lopez, B., Garcia-Reyes, J.F., Molina-Diaz, A., 2009 Sample treatment and determination of pesticide resi- dues in fatty vegetable matrices: a review Talanta 79, 109e128.

Guan, S., Yu, Z., Yu, H., Song, C., Song, Z., Qin, Z., 2011 Multi-walled carbon nanotubes as matrix solid-phase dispersion extraction adsorbent for simultaneous anal- ysis of residues of nine organophosphorus pesticides in fruit and vegetables by rapid resolution LC-MS-MS Chromatographia 73, 33e41.

He, Y., Lee, H.K., 1997 Liquid-phase microextraction in

a single drop of organic solvent by using a conventional microsyringe Analytical Chemistry 69, 4634e4640.

Hu, X., Pan, J., Hu, Y., Huo, Y., Li, G., 2008 Preparation and evaluation of solid-phase microextraction fiber based on molecularly imprinted polymers for trace analysis of tetracyclines in complicated samples Journal of Chro- matography A 1188, 97e107.

Ihnat, M., 2003 Sample preparation for food analysis Comprehensive Analytical Chemistry 41, 765e856.

Kinsellaa, B., O’Mahonya, J., Malonec, E., Moloneya, M.,

Cantwella, H., Fureyb, A., Danaher, M., 2009 Current

trends in sample preparation for growth promoter and

veterinary drug residue analysis Journal of

Chroma-tography A 1216, 7977e8015.

Kristenson, E.M., Haverkate, E.G.J., Slooten, C.J., Ramos, L.,

Vreuls, R.J.J., Brinkman, U.A.T., 2001 Miniaturized

automated matrix solid-phase dispersion extraction of

pesticides in fruit followed by gas

chromatographic-mass spectrometric analysis Journal of Chromatography

A 917, 277e286.

Kristenson, E.M., Ramos, L., Brinkman, U.A.T., 2006 Recent

advances in matrix solid-phase dispersion Trends in

Analytical Chemistry 25, 96e111.

Leon, C., Rodriguez-Meizoso, I., Lucio, M.,

Garcia-Can˜as, V., Iban˜ez, E., Schmitt-Kopplinb, P., Cifuentes, A.,

2009 Metabolomics of transgenic maize combining

Fourier transform-ion cyclotron resonance-mass

spec-trometry, capillary electrophoresis-mass spectrometry

and pressurized liquid extraction Journal of

Chroma-tography A 1216, 7314e7323.

Liu, B., Yan, H., Qiao, F., Geng, Y., 2011a Determination

of clenbuterol in porcine tissues using solid-phase

extraction combined with ultrasound-assisted

disper-sive liquideliquid microextraction and HPLCeUV

detection Journal of Chromatography B 879, 90e94.

Liu, X., Zhao, A., Zhang, A., Liu, H., Xiao, W., Wang, C.,

Wang, X., 2011b Dispersive liquideliquid

micro-extraction and gas chromatographyemass spectrometry

determination of polychlorinated biphenyls and

poly-brominated diphenyl ethers in milk Journal of

Separa-tion Science 34, 1084e1090.

Lo´pez-Feria, S., Ca´rdenas, S., Valca´rcel, M., 2009 One step

carbon nanotubes-based solid-phase extraction for the

gas chromatographicemass spectrometric multiclass

pesticide control in virgin olive oils Journal of

Chro-matography A 1216, 7346e7350.

Manzoori, J.L., Amjadi, M., Abulhassani, J., 2009 Ultra-trace

determination of lead in water and food samples by

using ionic liquid-based single drop

microextraction-electrothermal atomic absorption spectrometry

Analyt-ical Chimica Acta 644, 48e52.

Martendal, E., Budziak, D., Carasek, E., 2007 Application of

fractional factorial experimental and Box-Behnken designs

for optimization of single-drop microextraction of

2,4,6-trichloroanisole and 2,4,6-tribromoanisole from wine

samples Journal of Chromatography A 1148, 131e136.

Mastovska, K., Lehotay, S.J., 2006 Rapid sample preparation

method for LCeMS/MS or GCeMS analysis of

acryl-amide in various food matrices Journal of Agricultural

and Food Chemistry 54, 7001e7008.

Mendiola, J.A., Herrero, M., Cifuentes, A., Iban˜ez, E., 2007.

Use of compressed fluids for sample preparation: food

applications Journal of Chromatography A 1152, 234e246.

Mol, H.G.J., Rooseboom, A., van Dam, R., Roding, M., Arondeus, K., Sunarto, S., 2007 Modification and re-validation of the ethyl acetate-based multi-residue method for pesticides in produce Analytica and Bio- analytical Chemistry 389, 1715e1754.

Nollet, L.M.L., 1996 In: Nollet, L.M.L (Ed.), Handbook of Food Analysis Marcel Dekker, USA.

Olariu, R.I., Vione, D., Grinberg, N., Arsene, C., 2010 Sample preparation for trace analysis by chromato- graphic methods Journal of Liquid Chromatography & Related Technologies 33, 1174e1207.

Plaza Bolan˜os, P., Romero-Gonza´lez, R., Garrido Frenich, A., Martı´nez Vidal, J.L., 2008 Application of hollow fibre liquid phase microextraction for the multiresidue determination of pesticides in alcoholic beverages by ultra-high pressure liquid chromatography coupled to tandem mass spectrometry Journal of Chromatography

A 1208, 16e24.

Plossl, F., Giera, M., Bracher, F., 2006 Multiresidue ical method using dispersive solid-phase extraction and gas chromatography/ion trap mass spectrometry to determine pharmaceuticals in whole blood Journal of Chromatography A 1135, 19e26.

analyt-Po´l, J., Hyo¨tyla¨inen, T., Ranta-Aho, O., Riekkola, M.L., 2004 Determination of lycopene in food by on-line SFE coupled to HPLC using a single monolithic column for trapping and separation Journal of Chromatography A

1052, 25e31.

Raitioa, R., Orliena, V., Skibsted, L.H., 2011 Storage stability

of cauliflower soup powder: the effect of lipid oxidation and protein degradation reactions Food Chemistry 128, 371e379.

Ramalhosa, M.J., Paı´ga, P., Morais, S., Delerue-Matos, C., Oliveira, M.B., 2009 Analysis of polycyclic aromatic hydrocarbons in fish: evaluation of a quick, easy, cheap, effective, rugged, and safe extraction method Journal of Separation Science 32, 3529e3538.

Ramos, J.J., Dietz, C., Gonza´lez, M.J., Ramos, L., 2007 Miniaturised selective pressurized liquid extraction of polychlorinated biphenyls Journal of Chromatography

A 1152, 254e261.

Ramos, J.J., Gonza´lez, M.J., Ramos, L., 2009 Comparison of gas-chromatography-based approaches after fast mini- aturised sample preparation for the monitoring of selected pesticide classes Journal of Chromatography A

1216, 7307e7313.

Ravelo-Pe´rez, L.M., Herna´ndez-Borges, J., Ramos, M., Rodrı´guez-Delgado, M.A., 2009 Ionic liquid based dispersive liquideliquid microextraction for the extraction of pesticides from bananas Journal of Chro- matography A 1216, 7336e7345.

I ANALYTICAL TECHNIQUES

Rezaee, M., Assadi, Y., Milani-Hosseini, M.R., Aghaee, E.,

Ahmadi, F., Berijani, S., 2006 Determination of organic

compounds in water using dispersive liquideliquid

microextraction Journal of Chromatography A 1116, 1e9.

Ridgway, K., Lalljie, S.P.D., Smith, R.M., 2007 Sample

preparation techniques for the determination of trace

residues and contaminants in foods Journal of

Chro-matography A 1153, 36e53.

Ridgway, K., Lalljie, S.P.D., Smith, R., 2010 An alternative

method for analysis of food taints using stir bar sorptive

extraction Analytica Chimica Acta 677, 29e36.

Rocha, S.M., Coelho, E., Zrostlı´kova´, J., Delgadillo, I.,

Coimbra, M.A., 2007 Comprehensive two-dimensional

gas chromatography with time-of-flight mass

spec-trometry of monoterpenoids as a powerful tool for grape

origin traceability Journal of Chromatography A 1161,

292e299.

Rostagno, M.A., D’Arrigo, M., Martı´nez, J.A., 2010.

Combinatory and hyphenated sample preparation for

the determination of bioactive compounds in foods.

TrAC Trends in Analytical Chemistry 29, 553e561.

Saraji, M., Mousavi, F., 2010 Use of hollow fibre-based

liquideliquideliquid microextraction and

high-performance liquid chromatographyediode array

detection for the determination of phenolic acids in

fruit juices Food Chemistry 123, 1310e1317.

Saraji, M., Mousavinia, F., 2006 Single-drop microextraction

followed by in-syringe derivatization and gas

chroma-tographyemass spectrometric detection for

determina-tion of organic acids in fruits and fruit juices Journal of

Separation Science 29, 1223e1229.

Shariati, S., Yamini, Y., Esrafili, A., 2009 Carrier mediated

hollow fiber liquid phase microextraction combined

with HPLCeUV for preconcentration and determination

of some tetracycline antibiotics Journal of

Chromatog-raphy B 877, 393e400.

Shrivas, K., Wu, H.F., 2007 Rapid determination of

caffei-nexe caffeine in one drop of beverages and foods using

drop-to-drop solvent microextraction with gas

chroma-tography/mass spectrometry Journal of

Chromatog-raphy A 1170, 9e14.

Shrivas, K., Wu, H.F., 2008 Ultrasonication followed by

single-drop microextraction combined with GC/MS for

rapid determination of organochlorine pesticides from

fish Journal of Separation Science 31, 380e386.

Stubbings, G., Bigwood, T., 2009 The development and

validation of a multiclass liquid chromatography

tandem mass spectrometry (LCeMS/MS) procedure for

the determination of veterinary drug residues in animal

tissue using a QuEChERS (QUick, Easy, CHeap,

Effec-tive, Rugged and Safe) approach Analytica Chimica

Acta 637, 68e78.

Tankeviciute, A., Kazlauskas, R., Vickackaite, V., 2001 Headspace extraction of alcohols into a single drop Analyst 126, 1674e1677.

Turiel, E., Martı´n-Esteban, A., 2010 Molecularly imprinted polymers for sample preparation: a review Analytica Chimica Acta 668, 87e99.

Wang, L., Wang, Z., Zhang, H., Li, X., Zhang, H., 2009 Ultrasonic nebulization extraction coupled with head- space single drop microextraction and gas chromatog- raphy-mass spectrometry for analysis of the essential oil

in Cuminum cyminum L Analytica Chimica Acta 647, 72 Wen, X., Yang, W., Yan, Z., Deng, Q., 2011 Determination of cadmium and copper in water and food samples by dispersive liquideliquid microextraction combined with UVevis spectrophotometry Microchemical Journal 97, 249e254.

Wilkowska, A., Biziuk, M., 2011 Determination of pesticide residues in food matrices using the QuEChERS meth- odology Food Chemistry 125, 803e812.

Woodget, B.W., Cooper, D., 1987 Samples and Standards John Wiley and Sons, Chischester.

Xiao, Q., Hu, B., He, M., 2008 Speciation of butyltin compounds in environmental and biological samples using headspace single drop microextraction coupled with gas chromatography-inductively coupled plasma mass spec- trometry Journal of Chromatography A 1211, 135e141 Xiong, J., Hu, B., 2008 Comparison of hollow fiber liquid phase microextraction and dispersive liquideliquid microextraction for the determination of organosulfur pesticides in environmental and beverage samples by gas chromatography with flame photometric detection Journal of Chromatography A 1193, 7e18.

Yan, H., Wang, H., Qiao, J., Yang, G., 2011 Molecularly imprinted matrix solid-phase dispersion combined with dispersive liquid-liquid microextraction for the deter- mination of four Sudan dyes in egg yolk Journal of Chromatography A 1218, 2182e2188.

Zhao, E., Han, L., Jiang, S., Wang, Q., Zhou, Z., 2006 Application of a single-drop microextraction for the analysis of organophosphorus pesticides in juice Jour- nal of Chromatography A 1114, 269e273.

Zhu, Z., Zhou, X., Yan, N., Zhou, L., Chen, X., 2010 On-line combination of single-drop liquideliquideliquid micro- extraction with capillary electrophoresis for sample cleanup and preconcentration: a simple and efficient approach to determining trace analyte in real matrices Journal of Chromatography A 1217, 1856e1861 Zougagh, M., Rı´os, A., 2008 Supercritical fluid extraction as

an on-line clean-up technique for determination of riboflavin vitamins in food samples by capillary elec- trophoresis with fluorimetric detection Electrophoresis

29, 3213e3219.

C H A P T E R2

Data Analysis and Chemometrics

Paolo Oliveri, Michele Forina

Department of Drug and Food Chemistry and Technology, University of Genoa,

Via Brigata Salerno, 13, Genoa, Italy

O U T L I N E

2.1.1 From Data to Information 25

2.2 From Univariate to Multivariate 27

2.3.4 Supervised Qualitative Modeling 392.3.5 Supervised Quantitative Modeling 452.3.6 Artificial Neural Networks 48

2.1 INTRODUCTION

2.1.1 From Data to Information

Advances in technology and the increasing

availability of powerful instrumentation now

offer analytical food chemists the possibility for

obtaining high amounts of data on each sample

analyzed, in a reasonable e often negligible e

time frame (Valca´rcel and Ca´rdenas, 2005)

Often, in fact, a single analysis may provide

a considerable number of measured quantities,

generally of the same nature For instance, gaschromatographic (GC) analysis of fatty acidmethyl esters allows us to quantify, with a singlechromatogram, the fatty acid composition of

a vegetable oil sample (American Oil Chemist’sSociety, 1998) Spectroscopic techniques as wellmay supply, with a single and rapid analysis

on a sample, multiple data of homogeneousnature: in fact, a spectrum can be considered as

a data vector, in which the order of the variables(e.g., absorbances at consecutive wavelengths)has a physical meaning (Oliveri et al., 2011)

Chemical Analysis of Food: Techniques and Applications

DOI: 10.1016/B978-0-12-384862-8.00002-9 25 Copyright Ó 2012 Elsevier Inc All rights reserved.

In other cases, a set of samples can be

described by a number of heterogeneous

chem-ical and physchem-ical parameters at the same time

For example, a global analytical

characteriza-tion of a tomato sauce may involve the

quanti-fication of color and rheological parameters as

well as pH and chemical composition and e

possibly e a number of sensorial responses

(Sharoba et al., 2005) Also in such cases, each

sample may be described by a data vector,

but without any implication with respect to

the order of the variables Instead, differences

in variable magnitude and scale between

different variables may affect data analysis

if a proper pre-processing approach is not

followed

The availability of large sets of data does not

mean at the immediate time availability of

infor-mation promptly accessible to the sample

analyzed: usually, in fact, a number of steps

are required to extract and properly interpret

the potential information embodied within the

data (Martens and Kohler, 2008)

A deep understanding of the nature of

analytical data is the first basic step for any

proper data treatment, because different data

types usually require different processing

strat-egies, which closely depend on their nature and

origin For this reason, the data analyst should

always have a complete awareness of the

problem under study and about the whole

analytical process from which data derive e

from the sampling to the instrumental analysis

Such knowledge is fundamental: it makes the

difference between a chemometrician and

a mathematician A chemometrician is, first of

all, a chemist, who is acquainted with his

data, and utilizes mathematical methods for

the conversion of numerical records into

rele-vant chemical information

The analytical food chemist William Sealy

Gosset (1876e1937), who worked at the

Arthur Guinness & Son brewery of Dublin,

can be considered as one of the fathers of

chemometrics In fact, he studied a number

of statistical tools and adapted them to bettersolve actual chemical problems He had topresent his studies using a pseudonym, sincehis company did not permit him to publishany data Considering himself as a modestcontributor in the field, rather than a statisti-cian, he adopted the pen name Student Hismost famous work was on the definition ofthe probability distribution that is commonlyreferred to as the Student’s t distribution(Student, 1908)

The term chemometrics was used for the firsttime by Svante Wold, in 1972, for identifying thediscipline that performs the extraction of usefulchemical information from complex experi-mental systems (Wold, 1972)

Statistics offers a number of helpful tools thatcan be used for converting data into informa-tion Univariate methods, which consider onevariable at a time, independently of the others,have been and are still extensively used forsuch purposes Nonetheless, they usuallysupply just partial answers to the problemsunder study, since they underutilize thepotential for discovering global informationembodied in the data For instance, they arenot able to take into account inter-correlationbetween variables e a feature that can be veryinformative, if recognized and properlyinterpreted

Multivariate strategies are able to take intoaccount such an aspect, allowing a morecomplete interpretation of data structures.However, in spite of their big potential, multi-variate methods are generally less used thanunivariate tools

On the other hand, a number of people trymultivariate analysis as the last-ditch resort,when nothing seems to provide the desiredresults, pretending that chemometrics providevaluable information from data that do notcontain any informative feature at all

Such demeanor is very hazardous especiallywhen complex methods are being used, becausethere may be the risk of employing chance

correlations to develop models with good

performances only on appearance e namely,

on the same samples used for model building e

but with very poor prediction ability on new

samples: this is the so-called overfitting To

over-come such a possibility, a proper validation of

models is always required In particular, the

more complex the technique applied, the deeper

the validation recommended

For these reasons, a good understanding of

the characteristics of the methods employed

for data processing is always advantageous

as well

In this chapter, an overview of the

chemomet-ric techniques most commonly used for data

analysis in analytical food chemistry will be

pre-sented, highlighting potentials and limits of

each one

2.2 FROM UNIVARIATE

TO MULTIVARIATE

A bidimensional table is probably the most

typical way to arrange, present, and store

analytical data: conventionally, in

chemomet-rics, each row usually represents one of the

samples analyzed, while each column

corre-sponds to one of the variables measured

As an example, Table 2A.1 reports the

red-wine data set, which consists of 27 chemical

and physical parameters measured on 90 wine

samples, belonging to three Italian

denomina-tions of origin from the same region (Piedmont):

Barolo, Grignolino, and Barbera The original

data set was composed of 178 samples (Forina

et al., 1986)

Table 2A.1contains also additional

informa-tion, which is usually not processed but which

may be extremely helpful in the final

under-standing and interpretation of the results In

particular, the two heading lines contain the

numbers and the names of the variables, which

are additional information for the columns,

while the two heading columns include the

names identifying the samples and their class,which represent additional information for therows

It is easy to guess that such data enclose

a great deal of potential information Anyway,the simple visual inspection of the table, whichcontains a considerable number of records,does not provide directly any valuable informa-tion about the samples analyzed A conversionfrom data into information is necessary

Univariate methods are still the most used

in many cases, although they generally offeronly a very limitative vision of the globalsituation

2.2.1 Histograms

A good way to extract information from data

is to use graphical tools Among them, grams are probably the most widely employed(Chambers et al., 1983)

histo-To build a histogram, the range of interest

of the variable under study is divided into

a number of regular adjacent intervals Foreach interval, the contribution of the measuredsamples is graphically displayed by a verticalrectangle, whose area is proportional to thefrequency (i.e., the number of observations)within that interval Consequently, the height

of each rectangle is equal to the frequencydivided by the interval width, so that it hasthe dimension of a frequency density

Frequently, such frequency values arenormalized, dividing each of them by the totalnumber of observations, thus obtaining relativevalues It follows that, in such cases, the sum ofthe areas of all the rectangles e i.e., the sum ofall the relative frequencies e is equal to 1.The frequency distribution visualized by

a histogram can be used to estimate the bility distribution of the variable under studyand to make deductions about the samples.Figure 2.1shows examples of histograms for

proba-a portion of the dproba-atproba-a given inTable 2A.1, namelyfor variables number 13, 21, and 26

I ANALYTICAL TECHNIQUES

Three typical patterns are noticeable In

particular, variable 13 (phosphate) shows

a unimodal and almost symmetric shape, which

may suggest that such variable follows a normal

probability distribution (Fig 2.1a)

Conversely, variable 21 (OD280/OD315 of

diluted wines) presents a bimodal distribution,

which may suggest that this variable to be

charac-terized by different average values for diverse

sample classes (Fig 2.1b) In such cases,

histograms could be drawn for each class rately, to verify the trend of the within-classdistributions

sepa-Instead, the histogram shape for variable

26 (proline) reveals an underlying asymmetricdistribution (Fig 2.1c) It is possible toconvert such behavior into an almost normalone, simply by applying a logarithmic trans-formation to the variable, as it is shown inFig 2.2

OD280/OD315 of diluted wines

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

2x 10-3

2.2.2 Normality Tests

Assessing for compatibility with a normal

distribution is a basic issue in data analysis,

because many methods require variables to be

normally distributed As observed, frequency

distributions may be employed for this purpose

Visual examination of histogram shapes may

supply a preliminary evaluation Besides, the

cumulative empirical frequency distributions

(EFDs) constitute the basis for a family of

statis-tical normality tests, which are usually referred

to as KolmogoroveSmirnov tests (Kolmogorov,

1933; Smirnov, 1939)

One of the most effective and employed

among them is the Lilliefors test, which may

be used for generally assessing how well an

empirical distribution fits with a theoretical

one (Lilliefors, 1970) In the case for normality

verification, the null hypothesis (H0) is that the

observed empirical frequency distribution for

a given variable is not significantly different

from the theoretical normal probability

distribu-tion, at a given significance level The alternative

hypothesis (H1) is that the observed EFD is not

compatible with the theoretical normal

distribu-tion, at that significance level

The test procedure consists in ordering thevalues of the variable to be tested and normal-izing them by means of a Student’s transforma-tion (or autoscaling):

x
i;v ¼ xi;vs xv

The variable is corrected by subtracting itsmean (xv) from each of its values and thendividing by its standard deviation (sv) Theautoscaled variable is dimensionless andpresent mean equal to 0 and standard deviationequal to 1

Then, the corresponding cumulative ical probability distribution is estimated fromthe statistical parameters computed, and themaximum distance from such hypothesizeddistribution and the empirical one is calcu-lated This value is compared with a criticaldistance value, at a predetermined significancelevel, and such comparison determines theacceptance/rejection of the null hypothesis.The critical values, which depend on thesample size, were obtained by Monte Carlosimulations and are available on tables orstatistical software

theoret-The Lilliefors test can be performed also in

a graphical way (Iman, 1982), as it is illustrated

inFigs 2.3 and 2.4for the same cases ofFigs 2.1and 2.2

Charts for the Lilliefors test report the tive empirical frequency distributions (EFDs) forvariables number 13, 21, and 26 of Table 2A.1,after column autoscaling (polygonal curves inFigs 2.3 and 2.4), together with the cumulativetheoretical probability distribution (sigmoidsolid curves), and the distance limits according

cumula-to the Lilliefors test, at a 5% significance (sigmoiddot curves) When the EFD curve intersects atleast one of the limits individuated bythe critical distance, the null hypothesis isrejected As for the examples reported inFig 2.3, the null hypothesis is accepted only forthe variable phosphate, while for both the othervariables examined, it is rejected at the same

FIGURE 2.2 Histogram for the log-transformed variable

proline of Table 2A.1

I ANALYTICAL TECHNIQUES

significance level In fact, only in the first case

(Fig 2.3a), the polygonal EFD curve does not

intersect the critical distance lines in any point

In addition, it can be easily verified e

con-firming the deductions made by looking at the

histogram of Fig 2.2 e that the logarithmic

transformation applied to the variable proline

makes it compatible with the normal

distribu-tion (seeFig 2.4)

2.2.3 ANOVA

Analysis of variance (ANOVA) is the name of

a group of statistical methods based on Fisher’s

F tests, generally aimed at verifying theexistence/absence of significant differencesbetween groups of data The null hypothesisH0 is that all the data derive from the samestochastic population, i.e., there is no significantdifference between the groups considered In

OD280/OD315 of diluted wines (autoscaled)

order to verify this hypothesis, the final F test

compares the variability between groups with

the variability within groups (Box et al., 1978)

The simplest case is the one-way ANOVA,

whose procedure is described with a real

numerical example The two columns ofTable

2.1report the values of the alcoholic degree for

Barolo and Barbera wine samples of the

red-wine data set (Table 2A.1), respectively, and

some basic descriptive parameters A summary

of all the parameters computed for the ANOVA

test is given in Table 2.2 The aim is to assess

whether there is a significant difference between

the alcohol content of the two wines or not In

fact, although the mean Barbera alcoholic

percentage (13.07% abv) is noticeably less than

the corresponding Barolo value (13.83% abv),

the two respective ranges overlap, so that it

might be suspected the observed difference to

be due to chance variations

The within-columns variance can be computed

as a pool variance, under the hypothesis that the

variances of the different groups are

homogeneous When only two groups e and,consequently, two variances e are beingcompared, a Fisher’s F test is suitable to verifythis preliminary hypothesis In the given numer-ical example, the test value is computed as

Ft ¼ s2Barolo

s2 Barbera

¼ 0:2740:254 ¼ 1:08 (2.2)

The F critical value, at a 5% right significance leveland for 29 degrees of freedom (d.o.f.) both at thenumerator and at the denominator, is 1.86 So it

is possible to conclude that the variances of thetwo groups considered are not significantlydifferent, at a 5% right significance level

In problems involving more than twogroups, the comparison among variances can

be performed with multiple F tests on all thepossible pairs, or by means of the Cochran’stest or the Bartlett’s tests (Snedecor andCochran, 1989) The former is valid when there

is an equal numbers of data in each group,while the latter has a wider applicability

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Log proline (autoscaled)

I ANALYTICAL TECHNIQUES

TABLE 2.1 Alcohol Content (% abv) for Barolo and

Barbera Samples of Red-Wine Data Set,

and Basic Statistical Parameters

Barolo Barbera 14.23 12.86 13.20 12.88 13.16 12.81 14.37 12.70 13.24 12.51 14.20 12.60 14.39 12.25 14.06 12.53 14.83 13.49 13.86 12.84 14.10 12.93 14.12 13.36 13.75 13.52 14.75 13.62 14.38 12.25 13.63 13.16 14.30 13.88 13.83 12.87 14.19 13.32 13.64 13.08 14.06 13.50 12.93 12.79 13.71 13.11 12.85 13.23 13.50 12.58 13.05 13.17 13.39 13.84 13.30 12.45 13.87 14.34 14.02 13.48

TABLE 2.2 Full ANOVA Parameters for the Data

given inTable 2.1 Computed F ratio(from variances of columns Barolo andBarbera) ¼ 1.08 Critical F value (at 5%significance) ¼ 1.86 F test on variances

of columns Barolo and Barbera:

significance ¼ 41.8%

Source of variation d.o.f Sum of squares Variance Total 60 10874.485

Mean 1 10850.384 Between columns 1 8.786 8.786 Within columns 58 15.314 0.264 Computed F ratio ¼ 33.28; Critical F value (at 5% significance) ¼ 4.01; ANOVA F test: significance ¼ 0.0%.

In the numerical example discussed, thewithin-columns variance e computed as pooledvariance e corresponds to

s2 within ¼

of the ANOVA null hypothesis:

FANOVA ¼ s2between

s2 within

¼ 8:7860:264 ¼ 32:28 (2.5)The F critical value, at a 5% right signifi-cance level, for 1 degree of freedom at thenumerator and 58 degrees of freedom at thedenominator, is 4.01 From the comparison

with the computed test value, it follows that

the null hypothesis is rejected at a 5%

signif-icance level The conclusion is that the

differ-ence between the alcoholic content of Barolo

and Barbera samples is significantly larger

than the variability within each of the two

groups

ANOVA tests can be applied also when the

effect of two variability sources (e.g., type of

wine and vintage year) is to be verified: such

a scheme is usually called a two-way ANOVA

When a number of replicate measurements are

available for each level combination of the two

factors (nested two-way ANOVA), the model

obtained also allows an estimation of the

inter-action between the factors, together with its

significance

2.2.4 Radar Charts

Radar charts e also known as web charts,

spider charts, star charts, cobweb charts, polar

charts, star plots, or Kiviat diagrams e are

a data display tool that can be considered as

a sort of link between univariate and

multivar-iate graphical representations (Chambers et al.,

1983)

They consist of circular graphs divided into

a number of equiangular spokes, called radii

Each radium represents one of the variables

A point is individuated on it, whose distance

from the center is proportional to the magnitude

of the related variable for that datum Finally, all

the data points e corresponding to all the

vari-ables measured on a sample e are connected

with a line, which represent a sort of sample

profile

Usually, each plot represents a single

sample, and multiple observations are

compared by examining different plots It is

also possible to overdraw several lines on the

same chart, although the outcome will be

legible only for small data sets As a matter of

fact, when the number of samples is large,

such graphical representation is generally not

very functional

Within radar charts, variables can be sented without any previous scaling, revealingwhat variables are dominant for a given dataset Nonetheless, when variables are character-ized by considerably different scales (as in thecase for red-wine data ofTable 2A.1), a prelimi-nary transformation may be helpful in order tomake visible within the graph the contribution

repre-of all repre-of them, by assuring the same a prioriimportance

For instance, by looking atFig 2.5, it clearlyappears that, without any scaling, fourfeatures are dominating, corresponding tothe variables number 10, 13, 24, and 16, whichare characterized by the highest mean values(seeTable 2.1) The contribution of the remain-ing 23 variables is not recognizable withinthese graphs Furthermore, it is not possible

to draw many valuable considerations aboutthe sample profiles In particular, it can benoticed that Grignolino wines (Fig 2.5a) arecharacterized, on average, by smaller valuesfor the four observable variables It can also

be deduced that Barolo has a higher tion from variable 26, while Barbera (Fig 2.5c)has higher contributions from variables 24and 10

contribu-On the other hand, Fig 2.6 illustrates that,after application of column autoscaling (seeEqn (2.1)), the a priori differences in locationand dispersion among the original variablesare eliminated, thus showing the contribution

of all of them and highlighting the differencesamong the observations In fact, in this secondgraph, the profiles of the three wines appearmuch more dissimilar than in the previousone By a joint examination of the three radarcharts ofFig 2.6, it can be deduced that Baroloand Barbera samples present two rathercomplementary profiles, while the Grignolinoprofile is somewhat intermediate In particular,Barolo (Fig 2.6a) is characterized by higheraverage values of variables 1, 2, 13, 15, 16, 18,

21, 22, 23, and 26 Instead, Grignolino(Fig 2.6b) presents average lower values of allthe variables, except for the number 20 Finally,

I ANALYTICAL TECHNIQUES