Evaluating the flavor authenticity of thermal and high pressure apple juice by headspace gc ms fingerprinting

Faculty of Bioscience Engineering Evaluating the flavor authenticity of thermal and high pressure processed apple juice by headspace GC-MS fingerprinting Promoter : Prof.. i KU Leuven

Faculty of Bioscience Engineering

Evaluating the flavor authenticity of thermal and high pressure processed apple juice by headspace

GC-MS fingerprinting

Promoter : Prof Dr Ir Ann Van Loey Erasmus Dissertation Co-promoter: Prof Dr Ir.Tara Grauwet

Supervisor : Dr Biniam Kebede

Department of Microbial and Molecular Systems (M2S)

Center for Food and Microbial Technology Doan Ngoc Hai Dang

2014 - 2015

This dissertation represents the document presented to the evaluation commission but has not been corrected after the defense Reference can be made to this dissertation provided that written permission is obtained from the promoter(s) of this dissertation

i

KU Leuven

Faculty of Bioscience Engineering

Evaluating the flavor authenticity of thermal and high pressure processed apple juice by headspace

GC-MS fingerprinting

Promoter : Prof Dr Ir Ann Van Loey Erasmus Dissertation Co-promoter: Prof Dr Ir.Tara Grauwet

Supervisor : Dr Biniam Kebede

Department of Microbial and Molecular Systems (M2S)

Center for Food and Microbial Technology Doan Ngoc Hai Dang

2014 - 2015

ii

ACKNOWLEDGEMENT

Firstly, I would like to express my deepest gratefulness to my supervisor, Dr Biniam Kebede, for his patient guidance, encouragement and advice Without his help and support, this dissertation would not have been completed I would also like to thank my Co-promoter, Prof Tara Grauwet, for her useful comments, remarks for making the dissertation more focused

I am extremely thankful to my promoter, Prof Ann Van Loey, and the head of the Laboratory of Food Technology at KU Leuven, Prof Marc Hendrickx, for kindly giving the opportunity to carry out my work in this Laboratory

I would like to express deepest sense of honor to Erasmus Mundus Scholarship committee, for giving me an opportunity to study in Belgium I will never forget valuable time when I have been here

My sincerest gratitude to Hochiminh City University of Technology for allowing me to study in Belgium by Erasmus Mundus exchange program I would like to especially thank to Assoc Prof Le Van Viet Man, Department of Food Technology, for being truly supportive and encouraging

My deep appreciation gives to Ms Junjie Yi who helped me perform several laboratory tasks and gave helpful comments on my writing with her enthusiasm and generous knowledge I am also thankful to Ms Carolien Buvé, Ms Heidi Roba and Ms Margot De Haes for the technical assistances

Last but not least, I would like to thank Ms Katrien Verbist, Ms.Lut Vancuyck and all the laboratory staff and my fellow students in the Laboratory of Food Technology for their continuous support and encouragement

Doan Ngoc Hai Dang

June 2015

iii

ABSTRACT

Although thermal processing has been commonly used to preserve food, the intensive thermal load damages the quality of food products This has led to the investigation of innovative, non-thermal processing techniques, among which high pressure (HP) processing has received an attention Currently, there is a wide application of HP processing, for instance for food juicing In the juice industry, one of the main challenges

is to produce juices with a flavor close to the original freshly-squeezed ones Hence, there is a need to evaluate the potential of HP pasteurization to maintain fresh juice qualities The aim of this study is to evaluate the flavor authenticity of HP processed juices in comparison to thermal processed juices As a case study, three commercially

relevant apple varieties were selected: Pink Lady, Granny Smith and Jonagold The apple

juices were pasteurized with HP and thermal processing on an equivalent microbial

basis (5-log reduction of Escherichia coli O157:H7) Next, the volatile fraction of fresh

and pasteurized apple juices was analyzed with a microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS) method The effect of HP processing on key-aroma compounds seems to clearly differ among the

headspace-solidphase-apple varieties In Pink Lady headspace-solidphase-apple juice, many key aroma ester compounds are detected

in low amount after processing, and even in much lower amount in HP pasteurized

juices In Granny Smith, few compounds were detected by the VID procedure and thus

only one key aroma ester compound is detected in higher amount in HP processed

juices In Jonagold, key aroma ester compounds are detected in a comparable amount in

HP processed and fresh samples However, these compounds are detected in higher amount in thermal treated juices Furthermore, in all apple varieties, high amount of aldehydes and alkenes are observed in thermally treated juices compared to other juices However, in literature these compounds are not reported as key aroma compounds In general, based on the applied HS-SPME-GC-MS procedure, it is less straightforward to discuss about the effect of HP processing on apple flavor in comparison to thermal processing The contribution of a volatile compound to the juice odor is not only dependent on its concentration but also the odor threshold In the future, there is a need to investigate the aroma threshold value of these key aroma compounds and to combine it with sensory analysis before making any conclusion about the impact HP processing on apple aroma

Figure 3: Characteric pressure–temperature history of a food product during HP processing 20

Figure 4: Juicer (a) and its gears (b) (Angel Juicer 8500S, the Netherlands) 28

Figure 5: High pressure equipment (Resato, the Netherlands) showing the main components: (a) housing enclosing the six high pressure vessels, (b); (c) temperature control unit; (d) housing enclosing the intensifier (e); and (f) data logging and process monitoring device 30 Figure 6: GC-MS (7890N GC system, Agilent technologies, Diegem, Belgium) indicating the following components: (a) cooled tray of the auto sampler; (b) sample incubator; (c) fiber conditioning and regeneration oven; (d) GC-injection port; (e) GC oven housing the column (f) and (g) MS unit 32

Figure 7: Total peak area detected of (a) Pink Lady and (b) Granny Smith using different

incubation time at 40oC Following incubation, the extraction was fixed at 40oC during 10 min 36

Figure 8: Number of peaks detected of (a) Pink Lady and (b) Granny Smith using

different incubation time at 40oC Following incubation, the extraction was fixed at 40oC during 10 min 36

Figure 9: The peak areas of some Pink Lady’s key aroma compounds using different

incubation time: (a): 2-methyl-1-butanol, (b): 1-propanol and (c): 1-butanol 37

v

Figure 10: Total peak area detected of (a) Pink Lady and (b) Granny Smith using different

extraction time at 40oC 38

Figure 11: Number of peaks detected of (a) Pink Lady and (b) Granny Smith using

different extraction time at 40oC 38

Figure 12: The peak areas of some Pink Lady’s key aroma compounds using different

incubation time at 40oC: (a): ethyl acetate, (b): 1-butanol 1-propanol and (c): 1-butanol 39

Figure 13: The effect of different types of fiber coatings on the total peak area of: (a)

Pink Lady and (b) Granny Smith 40

Figure 14: The effect of different types of fiber coatings on the number of peaks of: (a)

Pink Lady and (b) Granny Smith 40

Figure 15: The effect of different types of fiber coatings on some key aroma compounds

of Pink Lady 41

Figure 16: The effect of different types of fiber coatings on some key aroma compounds

of Granny Smith 42

Figure 17: A bi-plot based on a PLS-DA model visualizing the differences among the

volatile fraction of apple varieties: Pink Lady (), Granny Smith (), Jonagold (), Red

Chief () and Golden Delicious () 43

Figure 18: Total ion chromatograms of volatile fraction of reference (fresh) Pink Lady,

Granny Smith and Jonagold apple juices obtained by HS-SPME-GC-MS procedure Per

chromatogram, the key aroma compounds are identified with number as indicated in table 4 47

Figure 19: PLS-DA bi-plots visualizing impact differences between of reference/ fresh (), thermal () and HP () treatments for three apple varieties namely: Pink Lady,

Granny Smith, Jonagold The small empty circles represent the headspace components,

where only the selected headspace components (bold open circles) are named To clearly show the most important ones, the key aroma compounds are put in italic The

vi

correlation loadings for the categorical Y-variables are represented as vectors The X- and Y- variances explained by LV1 and LV2 are indicated in the respective axes 49

Figure 20: The relative peak area of key aroma compounds detected in fresh and

thermal and HP processed Pink Lady juices 54

Figure 21: The relative peak area of key aroma compounds detected in fresh and

thermal and HP processed Granny Smith juices 54

Figure 22: The relative peak area of key aroma compounds detected in fresh and

thermal and HP processed Jonagold juices 56

Figure 23: PLS-DA bi-plots visualizing the volatile fraction of the three apple varieties

Pink Lady (), Granny Smith () Jonagold () per condition: fresh, thermal, HP processing The small empty circles represent the headspace components, where only the selected headspace components (bold open circles) are named To clearly show the most important selected headspace components, the key aroma compounds are put in

italic The correlation loadings for the categorical Y-variables are represented as vectors The X- and Y- variances explained by LV1 and LV2 are indicated in the respective axes.59

Figure 24: The relative peak area of key aroma compounds in each variety: Pink Lady (PL), Granny Smith (GS) and Jonagold (JG) 64

vii

LIST OF TABLES

Table 1: Volatiles identified to be produced by apple fruit and apple juice to date 4 Table 2: Volatile 'character-impact' compounds in apple fruit and its products 7

Table 3: Commercially Available SPME fiber (Manini and Andreoli, 2001) 14

Table 4: Key aroma compounds that were detected in the volatile fraction of Pink Lady,

Granny Smith and Jonagold apple varieties by HS-SPME-GC-MS procedure (“X” indicates

the compounds were detected in the apple variety and “-“ means the compounds were not detected) 45

Table 5: Selected headspace components in Pink Lady, Granny Smith and Jonagold for

each class (fresh, thermal and HP) The headspace components are listed in a decreasing order of VID coefficient, where a positive VID coefficient illustrates a higher concentration of a component for that class and vice versa The retention index (RI) of components is listed To clearly show the most important ones, the key aroma components are put in italic 50

Table 6: The headspace components selected in fresh and thermal and HP processed

samples for each class (Pink Lady, Granny Smith and Jonagold) The headspace

components are listed in a decreasing order of VID coefficient, where a positive VID coefficient, illustrates a higher concentration of a component for that class and vice versa The retention index (RI) of components is listed The key aroma components were put in italic 59

viii

TABLE OF CONTENTS

ABSTRACT iii

LIST OF FIGURES iv

LIST OF TABLES vii

TABLE OF CONTENTS viii

GENERAL INTRODUCTION xi

PART I: LITERATURE REVIEW

1 Apple flavor 1

1.1 Apple juice production 1

1.2 Apple flavor 2

1.2.1 Apple taste 2

1.2.2 Apple aroma 3

1.2.1.1 Volatile compounds identified in apple aroma 4

1.2.2.2 Volatile compounds important for apple aroma 6

1.3 Alternative apple juice processing technology 9

1.4 Advanced analytical and data analysis tool in analyzing apple flavor 12

1.4.1 Advanced analytical techniques 12

1.4.1.1 Sample preparation methods 12

1.4.1.2 Gas chromatography- mass spectrometry 15

1.4.2 Data analysis techniques 17

2 High pressure processing 18

ix

2.1 Basic principles of HP processing 18

2.1.1 Le Chatelier principle 18

2.1.2 Pascal principle 19

2.1.3 Adiabatic heat of compression 19

2.2 Intrinsic and extrinsic processing parameters 20

2.2.1 Intrinsic processing parameters 20

2.2.2 Extrinsic processing parameters 22

2.3 Effect of HP on color, flavor, texture of apple juice 23

2.4 Commercial application of HP proceesing on apple juice 24

PART II: EXPERIMENTAL WORK 26

3 Objective 27

4 Materials and methods 28

4.1 Sample preparation 28

4.2 Treatments 29

4.2.1 Thermal treatment 29

4.2.2 High pressure treatment 29

4.2.3 Post treatment sample handling 30

4.3 HS-SPME-GC-MS analysis 31

4.4 Data pre-processing and multivariate statistical data analysis 32

5 Results and discussion 34

5.1 Optimizing HS-SPME-GC-MS method and screening of apple varieties 35

x

5.1.1 Optimization of incubation time 35

5.1.2 Optimization of extraction time 37

5.1.3 Selection of fiber coating 39

5.1.4 Screening of apple varieties 42

5.1.5 Conclusion 43

5.2 Flavor authenticity evaluation by HS-SPME-GC-MS fingerprinting 44

5.2.1 Process impact comparison per apple variety 44

5.2.1.1 From chromatogram to data table (data pre-processing) 44

5.2.1.2 Qualitative classification of impact differences 47

5.2.1.3 Quantitative of selected headspace components 49

5.2.1.4 Interpretation of selected volatile compounds 52

5.2.1.5 Conclusion 56

5.2.2 Variety impact comparison per processing condition 57

5.2.2.1 Qualitative classification of the apple varieties per processing 57

5.2.2.2 Quantitative selected headspace components 59

5.2.2.3 Interpretation of selected volatile compounds 62

5.2.2.4 Conclusion 65

GENERAL CONCLUSION 66

REFERENCES 69

xi

GENERAL INTRODUCTION

Apples are amongst the most widely grown and widely consumed of temperate fruit crops, taking second place only to grape The common domesticated apple is putatively

an interspecific hydrid complex, usually designated Malus x domestica Borkh or M

domestica Borkh It is of high economic importance in the international due to its

pleasant sensory qualities such as color, taste and odor The fruits are eaten fresh, dried

or tinned or processed into juice, preserves or alcoholic beverage Fresh apple juice is a most unstable material from both a chemical and microbiological point of view For a long time, thermal processing has been and still is the preferred technology to achieve microbial inactivation and extend the shelf-life of these products However, thermal processing, particularly under severe conditions, may damage the nutritional and sensory properties of products It has been reported that a heat treatment leads to a cooked off-flavor as observed in different apple juices or to increased nonenzymatic browning

Today’s consumers desire high quality foods that are safe, convenient and nutritious, with freshly prepared flavor, texture and color, with minimal or no chemical preservatives High pressure (HP) is reported as a new food processing technology developed to achieve consumer demands for fresher products with reduced microbiological levels and retain flavor The reason for being called non-thermal technology is because HP treatment does not use heat to preserve the food In several countries, HP processing is being used on an increasing commercial basic, including in the juice industry The HP treated apple juices are labeled as “fresh”, “natural”, “highest quality premium” and guaranteed to provide real flavor of freshly squeezed juice In that context, it is important to evaluate the capacity of HP processing to maintain fresh juice qualities in comparison to thermal processing

Considering the importance of aroma components for apple juice quality, it is crucial to investigate volatile fraction on HP and thermal treated apple juice Previously, some studies used targeted analytical approach, which focuses on particular compounds of interest However, it is known that perceived odor/ flavor is not due to a single (group of) volatile compound(s) but rather as a result of mixture of odorants Therefore, an untargeted multi-response approach (through exploiting the potential of integrated

xii

analytical and data analysis tools) can be more suited to explore the relationship between volatile compounds and (bio-)chemical changes in complex systems like apple juice In addition, since apple juice aroma components differ from variety to variety, to paint a complete picture, different apple varieties should also be taken into consideration

The general objective of this dissertation is to compare the effect of thermal and HP processing on the volatile fraction of different apple varieties An HS-SPME-GC-MS fingerprinting is used as an untargeted multi-response approach to investigate the volatile fraction of pasteurized apple juice

This dissertation paper is subdivided into two parts: A literature review (Part I) and a summary of the experimental work performed (Part II) In Part I, there are two chapters

In Chapter 1, starting with the apple production followed by apple flavor, alternative apple juice processing technology and finally analysis tools for apple flavor will be described In Chapter 2, more focus will be given to HP processing including basic principle, processing parameters, effect of HP on apple juice and commercial status of

HP processing

In Part II, the performed experimental work is described starting with the objective of this dissertation (Chapter 3) followed by the description of the materials and methods (Chapter 4) used to achieve the objective In Chapter 5, firstly the results of a headspace GC-MS fingerprinting method optimization will be provided and secondly the results of the performed two types of comparative data analyses will be respectively discussed: (i) process impact comparison per apple variety; (ii) variety impact comparison per processing condition Finally, general conclusions and future perspective of this work will be presented

xiii

PART I:

LITERATURE REVIEW

1

1 Apple flavor

Apple juice is one of the popular fruit juices widely marketed in the European countries

A great number of apples are harvested for processing to juice every year The high commercial value of apple juice is ascribed to its pleasant sensory qualities such as

color, taste and odor (Komthong et al., 2007) Under this chapter, a general overview of

apple flavor is discussed in four sections Section 1.1 deals with apple juice production,

section 1.2 describes apple flavor including apple taste, apple aroma In the next

sections (1.3 and 1.4), the alternative apple juice processing and analysis tool in

analyzing apple flavor are summarized

1.1 Apple juice production

In apple juice processing, fruits are milled, producing pomace, which is pressed or extracted, and a cloudy juice is obtained After adding ascorbic acid, this juice is screened, filtered, and pasteurized prior to distribution In some processes, filtered juice can be further clarified by enzyme or gel treatment to obtain clear apple juice (Rivas, 1995; Lea, 1994) Ascorbic acid is added in apple juice to control enzymatic browning The action of ascorbic acid in the prevention of enzymatic browning is to reduce the intermediate o-quinones to the original phenolic compounds before they can undergo further reaction to form the pigments However, the effectiveness of ascorbic acid in antibrowning is temporary The enzymatic browning can re-generate after the ascorbic acid has been completely reduced to dehydroascorbic acid, therefore, this effect is more

pronounced during the storage (Komthong et al., 2007)

Fresh apple juice is a most unstable material from both a chemical and microbiological point of view For a long time, thermal processing has been and still is the preferred technology to achieve microbial inactivation and extend the shelf-life of these products Juice pasteurization is based on a 5-log reduction of the most resistant microorganisms

of public health significance (USFDA, 2004) Different time-temperature combinations could accomplish the targeted process Initially, pasteurization of fruit juices was done

at low temperatures 63-65°C for relatively long time (D'Amico et al., 2006) However

extended heat treatment cause undesirable quality changes during this process

2

(Moshonas and Shaw, 2000) Current pasteurization method in practice is high temperatures-short time (HTST) HTST, or flash pasteurization, is the most commonly

used method for heat treatment of fruit juice (David et al., 1996) The time/temperature

requirement for pasteurization of juice depends on the initial microbial load, pH of the product and if enzyme inactivation is required Apple juice is treated by HTST at 77 to 88°C for 25 to 30 s (Moyer and Aitken, 1980) However, thermal processing, particularly under severe conditions, may damage the nutritional and sensory properties of products It has been reports that a heat treatment leads to a cooked off-flavour as observed in different apple juices (Su and Wiley, 1998) or to increased nonenzymatic browning (Markowski, 1998) Therefore, alternatives to conventional thermal processing, which do not involve direct heat, have been investigated in order to obtain safe products, but with fresh-like quality attributes (eg high pressure, pulsed electric fields, ultrasound and membrane filtration) Among these technologies, due to the objective of this work, more focus will be given to high pressure processing This

technology will be discussed in detail in section 1.3

1.2 Apple flavor

Flavor is one of the key drivers of consumers' appreciation of apple (Daillant-Spinnler et

al., 1996) Despite its common usage, there is no universally accepted definition of the

term “flavor” Perhaps, one of the best definitions of flavor is proposed by Heymann and others, 1993: “Flavor is the biological response to chemical compounds (the physical stimuli) by the senses, interpreted by the brain in the context of human experience” Relative to the physical stimuli, Kader, 2008 stated apple flavor depends upon taste, which is mainly related to sugar and organic acid content and other compounds such as phenolic acids; and aroma, related to the odor – active volatile compounds (esters, alcohols, aldehydes, and ketones) In that context, each section is discussed step by step:

apple taste (section 1.2.1), apple aroma (section 1.2.2)

1.2.1 Apple taste

Taste implies the detection of nonvolatile compounds using several types of receptors that are mainly on the tongue There are five basic sensations determine taste: sweet, sour, salty, bitter, and umami; of which sweetness and sourness are major contributors

3

to the taste of fresh apple and apple juice (Poll, 1981; Watada et al., 1980) Sweetness in

apples is caused by three sugars: fructose, glucose and sucrose (Yahia, 1994) Fructose constitutes 50% of the total sugars, which account for 10 to 15% of apple fruit fresh

weight (Rouchaud et al., 1985) while glucose and sucrose vary between 2 to 4% (Fourie

et al., 1991) Apple taste is believed to mainly associate with balance between sugar and

acid content (Poll, 1993) Malic acid is the dominant acid in apples and citric is present

in lower amounts (10% of malic acid) Titratable acid is an indicator of the acid content and this parameter is used as an important quality factor of apple juice Poll, 1993 demonstrated that a juice with a given sugar concentration and a low acid concentration was flat and tasteless while a very high acid juice was sour and unpleasant to taste Polyphenols are primary constituents to astringency and bitterness of apples and apple juice (Poll, 1981; Williams, 1979) On the one hand, the amount of phenolic compounds present in ripe fruit is normally thought to be too low to contribute significantly to such astringency On the other hand, Poll, 1981 reported that polyphenol concentration did influence taste of apple juice and found a relatively good relationship between bitter-astringent taste and polyphenol concentration

1.2.2 Apple aroma

Although taste is crucially important to its perception, it is the presence of trace amounts of volatile compounds, which is responsible for odor that gives much of the character to fruits and their processed products (Williams, 1979) Aroma volatile compounds can be detected in parts per trillion (ppt) concentrations, and aroma stimuli usually reach the olfactory epithelium via two separate pathways: the nose, during

sniffing, and the mouth, during eating (Negoias et al., 2008) Three major chemosensory

systems are involved in the perception of these stimuli: smell, taste, and trigeminally mediated sensations Gas chromatography (GC) and mass spectrometry (MS) have made

it possible to identify more than 300 volatile compounds present in different apple cultivars (Paillard, 1990; Maarse, 1991), but only a few of these, about 20–40

compounds, have been shown to be responsible for fruit aroma (Cunningham et al.,

1986)

4

1.2.1.1 Volatile compounds identified in apple aroma

Maarse, 1991 has compiled extensive lists of volatiles extracted from apples and apple juice Apple volatiles are characterized as being of relatively low molecular weight, belonging to many different chemical classes and present in trace amount (Maarse, 1991) In details, these compounds include alcohols, aldehydes, ketones, acid and esters Although there is a great range of compounds in the volatile profile of apples, the majority are esters (78-92%) and alcohols (6-16%) (Paillard, 1990) Ester volatiles in apples are classified as ethyl esters, butyric esters, propanoic esters and hexanoic esters The alcohols mostly are ethyl, butyl and hexyl alcohols (Paillard, 1990) Higher molecular weight volatiles, often with one or two hydrophobic aliphatic chains, are likely to be trapped by skin waxes and are generally not found in the headspace (Paillard, 1990)

Table 1: Volatiles identified to be produced by apple fruit and apple juice to date

ALCOHOLS

methanol, ethanol, n-propanol, i-propanol, isopropanol, n-butanol, 2-butanol, pentanol, 2-methyl butanol, 2-methylbutan-1-ol, 2-methy1-2-butanoI, 3-methyl butanol,2-pentanol, 3-pentanol, 4-pentanol, 2-methyl-2-pentanol, 3-methyl pentanol, 4-methyl pentanol, hexanol, (E)-2-hexanol, (E)-3-hexanol, (E)-2-hexanol, (E)-3-hexanol, 1-hexen-3-ol, 5-hexanol, (Z)-3- hexen-1-ol, (E)-2-hexen-1-ol, (E)-hex-2- en-1-ol, heptanal, 2-heptanol, 6-methyl-5-heptanol, octanol, 2-octanol, 3-octanol, (Z)-3-octen-1-ol, (Z)-5-octenol, nonanol, 2-nonanol, 6-methy1-5-heptanol, decanol, 3-octenol, a-terpinol, terpen-4- ol,benzyl alcohol, 2-phenethanol, terpinert-4-ol, isobareol, citranellol, geraniol, eugenol

ALDEHYDES

formaldehyde, acetaldehyde, propanal, 2-oxopropanal, 2-propenal, butanal, isobutanal, 2-methyl butanal, (E)-2-butanal, pentanal, isopentanal, hexanal, (E)-2-hexenal, (Z)-3-hexenal, (E)-3-hexenal, octanal, nonanal, decanal, undecanal, dodecanal, benzaldehyde, phenyacetaldehyde

KETONES

2-propanone, 2-butanone, 3-hydroxybutan-2-one, 2,3-butanedione, 2-pentanone, 3- pentanone, 4-methylpentane-2-one, 2-hexanone, 2-heptanone, 3-heptannne, 2-octanone, 7-methyloctan-4- one, acetophenon, gamma-undecalactone

ACIDS

formic acid, acetic acid, propanoic acid, butanoic acid, isobutanoic acid, 2-methyl butanoic acid, methyl butanoic acid, pentanoic acid, pentanoic acid, 4 -methyl pentanoic, hexanoic acid, (E)-2- hexanoic acid, heptanoic acid, (E)-3-heptanoic acid, octanoic acid, (Z)- octenoic acid, nonanoic acid,

3-5

(Z)-3-nonenoic acid, decanoic acid, decenoic acid, undecanoic acid, undecenoic acid, dodecanoic acid, dodecenoic acid, tridecanoic acid, tridecenoic acid, tetradecanoic acid, tetradecenoic acid, pentadecanoic acid, pentadecenoic acid, hexadecanoic acid, hexadecenoic acid, heptadecanoic acid, heptadecenoic acid, octadecanoic acid, 9-octadecenoic acid, 9,12-octadecadienoic acid, 9,12 ,15-octadecatriertoic acid, nonenoic acid, decanoic acid, decenoic acid, nonadecanoic acid, nonadecenoic acid, eicosanoic acid, benzoic acid

ESTERS

methyl formate, ethyl formate, propyl formate, butyl formate, 2-methyl butyl formate, 3-methyl butyl formate, pentyl formate, i-pentyl formate, hexyl formate, methyl acetate, ethyl acetate, propyl acetate, butyl acetate, isobutyl acetate, t-butyl acetate, pentyl acetate, 2-methyl butyl acetate, 3- methyl butyl acetate, hexyl acetate, heptyl acetate, octyl acetate, benzyl acetate, (Z) -3-hexenyln acetate, (E)-hex-3-enyl acetate, (E)-2-hexenyl acetate, 2-phenyl ethyl acetate, n-octyl acetate, nonyl acetate, decyl acetate, methyl n-propionate, ethyl propionate, ethyl 2-methyl propionate, ethyl hydroxy propionate, propyl propionate, butyl propionate, isobutyl propionate, 2-methyl butyl propionate, 3-methyl butyl propionate, hexyl propionate, methyl butyrate, ethyl butyrate, propyl butyrate, isopropyl butyrate, butyl butyrate, isobutyl butyrate, pentyl butyrate, isopentyl butyrate, hexyl butyrate, cinnamyl butyrate, ethyl crotonate, methyl isobutyrate, ethyl isobutyrate, butyl isobutyrate, pentyl isobutyrate, hexyl isobutyrate, methyl 2-methylbutyrate, ethyl 2- methylbutyrate, propyl 2-methylbutyrate, butyl 2-methylbutyrate, isobutyl 2- methylbutyrate, penty12-methylbutyrate, hexyl 2-methylbutyrate, 2-methylbutyl butyrate, 3- methylbutyl butyrate, 2-methylbutyl oxtanoate, methyl pentanoate, ethyl pentanoate, propyl pentanoate, butyl pentanoate, amyl pentanoate, isoamyl pentanoate, hexyl pentanoate, methyl isopentanoate, ethyl isopentanoate, isopentenyl isopentanoate, methyl hexanoate, ethyl hexanoate, propylhexanoate, butyl hexanoate, isobutyl hexanoate, butyl-(E)-hex-2-enoate, pentyl hexanoate, 2- methyl butyl hexanoate, 3-methyl butyl hexanoate, butyl trans-2-hexanoate, ethyl heptanoate, hexyl hexanoate, propyl heptanoate, butyl heptanoate, ethyl octanoate, propyl octanoate, butyl octanoate, isobutyl octanoate, pentyl octanoate, isopentyl octanoate, hexyl octanoate, ethyl nonoate, ethyl decanoate, butyl decanoate, isobutyl decanoate, pentyl decanoate, isopentyl decanoate, hexyl decanoate, ethyl dodecanoate, butyl dodecanoate, hexyl dodecanoate, diethyl succinate, ethyl 2- pbenylacetate, dimethylphthalate, diethylphthalate, dipropylphthalate

MISCELLANEOUS

diethyl ether, methyl propyl ether, dibutyl ether, 2-methyl butylether, methyl propyl ether, methyl butyl ether, dihexyl ether, methyl phenyl ether, 4-methoxyallyl benzene, (Z)- linalool oxide, (E)-lonalool oxideethylamine, butylamine, isoamylamine, hexylamine, diethoxymethane, dibutoxymethane, dihexoxymethane, dihexoxyethane, 1-ethoxy-1- propoxyethane, 1-butoxy-1- ethoxyethane, 1-etoxy-1-hexoxyethane, 1-ethoxy-1-octoxye thane, 1,1-diethoxyethane, 1,1- dibutoxyethane, 1-butoxy-1-2-methylbutoxy ethane, 1,1-diisobutoxyethane, 1-butoxy-1- hexoxyethane, 1,1-di-2-methyl butoxyethane, 1,2-me-thyl butoxy-1-hexoxy ethane, 1,1-di- hexoxyethane, 1,1-diethoxypropane, 1,1- diptb.oxyp entane, 4-methoxyallyl benzene, furan, furfural, 5-hydroxymethylfurfural, 2,4,5-trimethyl-1,3-dioxolane, ethane, ethylene, a-farnescene, l3-farnescene, benzene, ethyl ben-zene, 1-methylnaphthalene, 2-methylnaphthalene, 13- damascenone, a-pinene

3-Source: Nursten (1970); Dimick and Hoskin (1983); Acree and McLellan (1989); Maarse (1991)

6

1.2.2.2 Volatile compounds important for apple aroma

Volatiles do not contribute equally to odor, and in fact, the odor of apple seems to

depend on the presence of only few volatiles (Cunningham et al., 1986) Studies of

aroma have been directed towards assessing the sensory impact of individual volatile

compounds found in fruit (Maga, 1990; Teranishi et al., 1987) Additionally, a few groups

of researchers have assessed the flavor significance of chemicals analyzed by sniffing

effluents at the outlet of the column (Guadagni et al., 1966; Acree et al., 1984; Cunningham et al., 1986; Grosch, 1993; McDaniel et al., 1990) Acree and co-workers,

1984 and Grosch, 1993 inject the aroma extract into the GC after successive dilutions: compounds that are perceived by the human subject at the highest dilution level are believed to be the character impact volatiles of the sample Those compounds present in the food at concentrations above their odor threshold contribute to the food aroma Odour threshold has been defined as the minimum physical intensity detection where the subject is not required to identify the stimulus but just to detect the existence of the

stimulus (Teranishi et al., 1987) About 20 individual volatile compounds have been

reported to be 'major contributors' or 'character-impact' compounds to the aroma of

apples and apple products (Table 2)

The large number of these character-impact compounds reported is possibly due to several causes The existence of a large number of apple cultivars makes it difficult to define only one or a few character-impact compounds to apples Study results also differ greatly in the way of determining the sensory significance of different volatile compounds (Durr, 1994; Yahia, 1994) Furthermore, there is a frequent lack of separation between 'primary' and 'secondary' volatiles (Berger, 1991) 'Primary' volatile compounds are those produced by controlled enzymatic reactions in intact tissue, while 'secondary' volatiles are formed by various uncontrolled enzymatic reactions when plant tissues are disrupted, such as by homogenization, cutting, chewing, or heating

(Roberts and Acree, 1995) Several volatile compounds listed in Table 2, such as

(E)-2-hexenal, (E)-2-hexenol, (Z)-3-(E)-2-hexenal, and (Z)-3-hexenol, are secondary volatiles (Hatanaka, 1993) that contribute significantly to aroma of apple juice and essence (Flath

et al , 1967)

7

Table 2: Volatile 'character-impact' compounds in apple fruit and its products

sweet, perfumy and fruity odour

burning taste sweet

highly diluted-pleasant overall flavour, aroma, sweet aroma

earthy, unpleasant harmonious, fruity

red apple aroma Cox-like aroma harmonious nail polish

Golden Delicious Golden Delicious Mclntosh

Delicious

many

Golden Delicious Mclntosh

Delicious

many

many

Pink Lady Pink Lady Royal Gala, Golden Delicious

Golden Delicious

many

Royal Gala Cox's Orange Pippin

many

Gala

Apple fruit Apple juice Apple juice Apple juice Apple juice Apple juice Apple juice Apple juice Apple juice Apple juice

Apple fruit Apple fruit Apple fruit Apple fruit Apple fruit Apple fruit Apple juice

Apple fruit Apple juice Apple juice Apple fruit

Rizzolo et al (1989) Rizzolo et al (1989) Panasiuk et al (1980) Flath et al (1969)

Duerr ( 1979)

Rizzolo et al (1989) Panasiuk et al (1980) Flath et al (1969)

Duerr (1979)

Cunningham et al (1986)

Karlsen et al., 1999 Lopez et al., 2007 Lopez et al., 2007 Young et al (1996) Rizzolo et al (1989) Rizzolo et al (1989)

8

Table 2: Volatile 'character-impact' compounds in apple fruit and its products

banana like fruity, estery harmonious, fruity fruity

apple like sweet strawberry spicy, aniseed green apple apple ethereal-fruity fruity

- pungent and penetrating

many

Golden Delicious Delicious

Gala

many

Gala Gala Pink Lady Pink Lady Golden Delicious

many

Apple fruit

Apple juice Apple fruit

Apple fruit

Apple fruit Apple juice Apple fruit Apple juice Apple fruit Apple essence Apple fruit Apple juice Apple fruit Apple fruit Apple fruit Apple fruit Apple fruit Apple fruit

Rizzolo et al., 1989

Deurr, 1979

Rizzolo et al., 1989 Flath et al., 1967

Plotto, 1998

Williams et al., 1977

Plotto, 1998 Plotto, 1998

Lopez et al., 2007 Lopez et al., 2007 Lopez et al., 2000 Karlsen et al., 1999

9

Esters have been identified to be primarily responsible for apple aroma Acetates, propanoates, butanoates and hexanoates are the most important esters contributing to

apple odor in such cultivars as “Golden Delicious”, “Royal Gala”, “Cox’s Orange Pippin” and

“Gala” Hexyl acetate was considered to characterize Cox's Orange Pippin and Golden

Delicious apples The sensory descriptor given to hexyl acetate was " sweet fruity, apple”,

“Gala, ripe, pear" Rizzolo et al (1989) and Flath et al (1967) found that ethyl 2-methyl

butanoate, which had a very low odour threshold, had an intense apple odour that has

been characterized as “apple like”, “ fruity ” in Delicious and Gala apples The compound

that had the most effect on the sensory attributes, which are particularly important for Royal Gala apple, is 2-methylbutyl acetate

Alcohols are the second most important group of organic compounds in terms of contribution to apple flavor 1-butanol which posses a sweet aroma was suggested to be

one of four important contributors to the flavour of Royal Gala apples besides

methylbutyl acetate, butyl acetate, hexyl acetae E-2-hexenol is considered as an important constitute of apple juice, provides “harmonious, fruity “ aroma

Hexanal and (E)-2-hexenal were the most abundant aldehydes identified in apple juice where they contribute to the green, fresh apple aroma note (Durr and Schobinger, 1981 ;

Flath et al , 1967) These two volatiles were formed quickly after apple tissues were

crushed and homogenized They might be also important during cutting and chewing of fresh apples Hexanal and (E)-2-hexenal were produced by crushed apples, with a maximum after 5 minutes (Paillard and Rouri, 1984) Ketones are thought to have low aroma value but there is an exception for β-Damascenon β-Damascenon which has a fruity odour is reported to be an important aroma impact compound in most apple juice and essence This compound is present in fresh apple and freshly extracted juice in very low concentration l µl/l but present in very high concentration in heated apple juice, hence is probably a product resulting from thermal processing (Roberts and Acree, 1995)

1.3 Alternative apple juice processing technology

As mentioned in section 1.1 and 1.2, although conventional thermal processing of apple

juices remains the most widely adopted technology for shelf life extension and

10

preservation, thermal pasteurization tends to reduce the apple juice’s quality and freshness Therefore, some non-thermal pasteurization methods have been proposed during the last couple of decades, including membrane filtration, ultrasound (US), pulsed electric field (PEF) and high pressure (HP) processing (Rupasinghe and Yu, 2012) These emerging techniques do not involve direct heat and are claimed to produce safe foods with improved quality attributes In this section, these alternative technologies will be introduced shortly

Membrane filtration is a non-thermal technique that uses a physical barrier, a porous membrane or filter, to separate particles in a fluid Particles are separated on the basis of their size and shape with the use of pressure and specially designed membranes with different pore sizes In apple juice processing, microfiltration (MF) and ultrafiltration (UF) membranes retain microorganisms while the smaller molecules like vitamins, sugar and antioxidants pass through the membranes with the water The effectiveness depends on many factors including processing factors such as types of membrane, pore size, transmembrane pressure and medium factors such as type of juice and microorganism For example, an ultrafiltration (UF) unit, with polysulphone membranes

of 10 kDa and 50 kDa pore sizes and trans-membrane pressures of up to 155 kPa, were used to treat apple juices (Rupasinghe and Yu, 2012) However, MF and UF is generally more applicable for clear apple juices that are filtered anyway Since these methods cause color changes and some flavor changes, they are not widely used for cloudy apple juices

Power ultrasound (US) is also a potential non-thermal technique for preservation of apple juice The principle of ultrasonic processing could be explained as follows: Firstly, the ultrasonic transducers convert electrical energy to sound energy Secondly, when the ultrasonic waves propagate in liquid, small bubbles will be formed and collapsed thousands of times per second This rapid collapse of the bubbles (cavitation) results in high-localized temperatures and pressure, causing breakdown of cell walls, disruption of

cell membranes and damage of DNA (Manvell, 1997; Knorr et al.,2004; O’Donnell et al

2010) A few studies have been conducted to examine the effect of ultrasound on treated apple juices For example, Abid, 2003 found that sonication treatment significantly improved the phenolic compounds, ascorbic acid, and total antioxidant capacity without

11

any significant effect on the physicochemical parameters (pH, titratable acidity, °Brix) of apple juice Furthermore, sonication induced significant reduction in microbial population, but unfortunately it could not achieve the 5-log reduction of microorganisms The results suggested that sonication technology might successfully be employed for the processing of apple juice with improved safety and quality from consumer’s health point of view

Pulsed electric field processing (PEF) is an innovative technology that applies short bursts of high voltage electricity for microbial inactivation and causes no or minimum effect on food quality attributes Briefly, the foods being treated by PEF are placed between two electrodes, usually at room temperature The applied high voltage is usually in the order of 20-80 kV for microseconds It is generally believed that electric fields induce structural changes in the membranes of microbial cells based on generation of pores of the cell membrane, leading consequently to microbial destruction

and inactivation (Tsong, 1991; Barbosa-Cánovas et al., 1999) Because the mechanism of

electroporation is based mainly on a mechanical electrocompressive force affecting the cell membrane, the PEF technology is considered a nonthermal preservation process There are some studies have proven that compared with thermal processing, PEF processing can extend the shelf-life and preserve the original sensory and nutritional characteristics of apple juice due to the very short processing time and low processing

temperatures Endilek et al., 1999 revealed that PEF treatment was effective in

activating E coli O157:H7 and E.coli 8739 in apple juice samples A 5 log reduction was observed with electric field strengths 30 kV/cm and total treatment time of 172 Ms

Aguilar et al., 2007 reported that PEF-treated juice retained better most of the volatile

compounds responsible for colour and flavour of the apple juice while conventional HTST pasteurization produced significant losses in phenolic compounds and in volatiles responsible for flavour

High pressure (HP) processing, also described as high hydrostatic pressure (HHP), or ultra high pressure (UHP) processing, subjects liquid and solid foods, with or without packaging, to pressures between 100 and 800 MPa The application of HP processing in food area started from 1900s when Hite and other researchers applied HP processing on the preservation of milk, fruits and vegetables However, it takes a long time for the

12

commercial products to emerge in the market In 1990, the first HP processed fruit jams were sold in the Japanese market Subsequently, HP processed commercial products including fruit juices and beverages, vegetable products have been produced in North

America, Europe, Australia, and Asia (Balasubramaniam et al., 2008) HP processing is

considered a technology with the most promising perspective of industrial utilization (Farkas and Hoover, 2000) One of the main advantages of this process is the almost instantaneous and isostatic pressure transmission to the product, independent of size, shape, and food composition yielding highly homogenous products HP processing is said to extend shelf-life, guarantee safety and maintain fresh quality (Cheftel, 1995) A

detailed discussion of HP processing can be found on Chapter 2

1.4 Advanced analytical and data analysis tool in

analyzing apple flavor

1.4.1 Advanced analytical techniques

In general, major analytical steps include sample preparation, separation, quantification

and data analysis The sample preparation techniques will be described in section

1.4.1.1 For separation and quantification, gas chromatography-mass spectrometry

(GC-MS) will be introduced in section 1.4.1.2

Typical methods for sample preparation use liquid/liquid or liquid/solid extraction with organic solvents Most of these solvents are toxic, difficult for disposal and potential to

cause environmental and occupational damage (Stewart et al 2009) As a result, many

solvent free extraction methods have been developed These can be classified according

to the nature of extraction phase, ie., gas-phase extractions, membrane extraction and sorbent extractions Static headspace and dynamic headspace techniques are categorized under gas-phase extraction In membrane extraction methods, analytes are extracted from the sample with the use of a polymer membrane Solid phase microextraction (SPME) is a sorbent type extraction method (Prosen and Zupancic-Kralj, 1999)

13

Static headspace is a technique, which is used to sample and concentrate volatile

compounds from the headspace (gaseous phase above the liquid matrix) (Figure 1)

Once at equilibrium, the analytes are transferred to the chromatograph for analyses (Prosen and Zupancic-Kralj, 1999) The advantages of static headspace are: minimal sample preparation, rapidity, use of little or no solvent and inexpensive However, static headspace may lack the sensitivity for the determination and may not detect poorly volatile compounds (Wampler, 2001)

Figure 1: Schematic representation of static headspace sampling using gas-tight syringe (www.restekcorp.com)

In dynamic headspace, the analytes are exhaustively extracted from the sample, equilibrium is never reached since the gases in the headspace are continuously removed from the vial The volatile analytes are swept to a trap where they are held until analyses (Grob, 2014) Compared to static headspace, dynamic headspace is more sensitive, but dynamic headspace still has some drawbacks such as the limitation to compounds that

do not co-elute with solvent and the presence of solvent impurity peaks

SPME is a relatively new method that offer several advantages for sample preparation including reduced time per sample, less sample manipulation resulting in an increased sample throughput and, in addition, the elimination of organic solvents and reduced analyte loss

In SPME analysis, a fused-silica fiber, which is coated with a polymer (stationary phase)

is used to isolate and concentrate analytes from the liquid phase or from the headspace

(Miller et al., 1999) The selectivity of SPME depends on the phase coating of the silica

fiber The volume of the coating determines the capacity and the sensitivity of the method The choice of the appropriate stationary phase should take into account both

14

volatility and polarity of analytes Using thick coatings, both extraction and desorption become slower and the analysis time longer Several kinds of stationary phases with different polarity and thickness are now commercially available

Table 3: Commercially Available SPME fiber (Manini and Andreoli, 2001)

SPME analysis can be applied in two ways, direct and headspace extraction The direct extraction mode is based on direct immersion of the coated fiber into the sample and direct mass transfer from the matrix to the polymeric phase In headspace mode, analytes should pass into the vapor phase before the extraction The choice between direct and headspace extraction mainly depends on the volatility of the analytes and on

the kind of matrix to be analyzed In the case of the headspace sampling (Figure 3b),

first, the sample in a vial is heated and agitated to facilitate the release of volatile

15

compounds to the headspace After equilibrium state is achieved (between the liquid matrix and the headspace), the fiber is only inserted in the gaseous phase directly above

the liquid matrix in the vial (Wardencki et al., 2004)

Headspace SPME sampling mode has a higher selectivity than direct sampling for extracting volatile and semi-volatile compounds, since only these compounds are

released to the headspace (Wardencki et al., 2004) Following an extraction process

(Figure 2c), the fiber is transferred, with the help of the syringe-like handling device, to

analytical instruments where it will be desorbed at high temperature (i.e in the hot

GC-injection port) (Figure 2d) (Muller et al., 1999)

Figure 2: Schematic representation of headspace sampling with SPME: (a) direct extraction; (b) headspace extraction; (c) extraction; (d) desorption http://www.labsphere.biz/pdf/CTC/SPME- option-brochure.pdf

The efficiency of SPME analysis is determined by factors, such as sample volume, salt type and amount, type of fiber, extraction mode (direct or headspace), extraction time

and temperature, and desorption condition (Wardencki et al., 2004; Muller et al., 1999)

16

Gas chromatography-mass spectrometry (GC-MS) is the synergistic combination of two powerful analytic techniques The GC part of this integration enables high-resolution separation of volatile compounds found in a mixture of a gas phase (Dickes and Nicholas, 1976; Grob, 2004) Each compound exiting the GC system is transferred into a MS, in which identification of the chemical nature of the compound and quantification will take place (Downard, 2004; Masucci and Caldwell, 2004) Therefore, with the GC-MS technique both qualitative and quantitative information about the target compounds is obtained (McNair and Miller, 1998)

In gas chromatography, there is a physical separation in which the components in a mixture are selectively distributed between the mobile phase, which is an inert carrier gas, and a stationary phase, which is present as a coating of either column packing particles or the inner column wall The chromatographic process occurs as a result of repeated sorption/desorption steps during the movement of the analytes along the stationary phase by the carrier gas The separation is due to the differences in distribution coefficients of the individual components in the mixture Compounds with a weak interaction with the stationary phase will be eluted quickly, while compounds with

a strong interaction are eluted slowly The time elapsed between sample injection to the GC-injection port and elution is called the retention time (Dickes and Nicholas, 1976; Grob, 2004)

Individual molecules, which are eluted at different retention times, will pass through a heated transfer line to a detector; in this case the mass spectrometer (Downard, 2004) The vaporized analytes are then ionized, producing molecular and/or fragment ions, which are then mass resolved utilizing a mass filter and detected The resulting mass spectrum is displayed as a plot of the relative intensity of these ions versus their mass-to-charge ratio (m/z) Since most ions produced are singly charged, their m/z values are indicative of their masses Atomic mass units are defined as daltons (μ) As the gas chromatographic separation proceeds, the mass analyzer is repeatedly scanned The ion intensities for all m/z values for each scan can then be summed to generate a chromatographic trace commonly called a total-ion current chromatogram Since the

17

mass spectrum generated by a given pure chemical compound is the same every time, the mass spectrum is actually a fingerprint for the molecule (Downard, 2004) A library, which is found in the data system, can be useful to identify an unknown compound by comparing the mass spectrum of this unknown compound to mass spectra in the library (McNair and Miller, 1998)

1.4.2 Data analysis techniques

Since a huge amount of data can be generated using GC-MS, data analysis techniques to convert the complex raw data obtained into useful information have been developed A brief description of two important data analysis steps for GC-MS data will be provided (i) data pre-processing and (ii) multivariate data analysis (Katajamaa and Oresic, 2007) Data pre-processing involves transforming the raw data into a format that allows easy access to characteristics of each observed ion and ensures subsequent data analysis steps to be simpler Data pre-processing includes deconvolution, filtering, alignment and normalization Through these processes, raw measurement signal is transformed into a format that does not contain any measurement noise (or baseline) or systematic variations between samples with measurements that have been clustered across

different samples (Cevallos-Cevallos et al., 2009; Katajamaa and Oresic, 2007)

Currently, advanced softwares, such as Automated Mass Deconvolution and Identification System (AMDIS) and Mass Profiler Professional (MPP), have been developed to convert the three-dimensional raw data, which consists of m/z, retention time, ion current, into a two-dimensional data expressing time-aligned and mass-aligned

abundances of chromatographic peaks (Kebede et al, 2013; Antignac et al., 2011;

Katajamaa and Oresic, 2007), which will be further analyzed with multivariate data analysis

The multivariate techniques transform the large number of original variables into new few manageable variables that can maximally explain the variation in the data so that analytical information of importance is emphasized By definition, multivariate data analysis (MVDA) aims to study samples across multiple dimensions taking into account the information of all variables detected (X) (e.g., headspace components if the gas food fraction is investigated) to describe the response variable of interest (Y) (e.g., process

18

impact) Two techniques are commonly used: principal component analysis (PCA) and partial least squares (PLS) regression PCA is an unsupervised statistical tool that is often applied as an exploratory technique while PLS and PLS-DA are supervised techniques that are applied to describe response variable and the separation between

the sample classes in the multivariate space (Grauwet et al., 2014) In PLS-DA, to

visualize the multidimensional data structure, projections of the samples/objects and variables can be graphically represented in score and loading plots, respectively By combining scores and loadings, bi-plots graphically illustrate how samples relate to each other (scores) and what was the importance of each variable to the separation

(difference) between classes (loadings) (Herrero et al., 2012; Grauwet et al., 2014)

2 High pressure processing

As mentioned in section 1.4, HP processing has received the largest attention among

alternative technology In the juice industry, it is important to produce juice with a flavor close to that of freshly squeezed one For that reason, HP technology is considered for apple juice processing In this Chapter, high processing is discussed in more details

starting from basic principles of HP processing (section 2.1), intrinsic and extrinsic parameter (section 2.2) to effect of HP on apple juice (section 2.3) and finally commercial application of HP processing on apple juice (section 2.4)

2.1 Basic principles of HP processing

The effect of HP processing is governed by three general principles: the Le Chatelier

principle (section 2.1.1), the Pascal principle (section 2.1.2) and the principle of

adiabatic heat of compression (section 2.1.3) (Balasubramaniam et al 2008)

2.1.1 Le Chatelier principle

Le Chatelier’s Principle states that under equilibrium conditions, processes associated with volume decrease are encouraged by pressure, whereas processes involving volume increase are inhibited by pressure (Balasubramaniam et al., 2008) Due to this fact, at a relatively low temperature (0 - 40°C) covalent bonds are almost unaffected by HP where

19

the tertiary and quaternary structures of molecules which are maintained chiefly by

hydrophobic and ionic interactions are altered by high pressure >200 MPa (Hendrickx et

al., 1998) Covalent bonds are resistant to pressure, which means that small

macromolecules responsible for flavor and odor characteristics in a food are not changed by pressure and foods that are subjected to HP processing at ambient or chill temperatures do not undergo substantial changes to flavor or color Likewise, the molecular structure of vitamin and availability of minerals is largely unaffected and HP causes minimal changes to the nutritional value of foods (Fellows, 2009) This is the reason why there is a hypothesis HP processing has a better flavor retention However, flavor authenticity of HP processed products need to be investigated

2.1.2 Pascal principle

The Pascal principle states that during HP processing, pressure is applied uniformly around the food product and is instantaneously and uniformly distributed throughout

the system (Balasubramaniam et al., 2008) This application of high pressure to all parts

of a food is a significant advantage compared to thermal processing because the food is treated evenly throughout and package size, shape and compositions are not factors in determining the process conditions (Fellows, 2009)

2.1.3 Adiabatic heat of compression

To investigate the impact of HP processing and to evaluate process uniformity, it is

essential to consider the combined effect of pressure and temperature Figure 3 shows a

typical pressure–temperature history of a food product during a HP treatment During the compression phase (t1~t2), foods increases (T1~T2) as a result of physical compression (P1~P2) The magnitude of this increase is dependent on factors such as final pressure, product composition, and initial temperature The temperature of water increases about 3°C for every 100 MPa pressure increase at room temperature (25°C)

On the other hand, fats and oils have a heat of compression value of 8–9°C/100 MPa, and proteins and carbohydrates have intermediate heat of compression values

(Rasanayagam et al., 2003, Patazca et al., 2007) Before decompression, the product is

held under pressure (P2~P3) for a certain time (t2~t3) Upon decompression (P3~P4),

20

in a perfectly insulated (adiabatic) system, the product will return to its initial temperature In practice, however, the product will return to a temperature (T4) slightly lower than its initial temperature (T1) as a result of heat losses during the holding time The rapid heating and cooling resulting from HPP treatment offer a unique way to increase the temperature of the product only during the treatment, and to cool it rapidly

thereafter (Balasubramaniam et al., 2008) In practical applications, the combination of

adiabatic heating and HP could potentially be used for the rapid, high-temperature

sterilization of low-acid foods (Knoerzer et al., 2010) However, during HP

pasteurization (which is the case in this study), the aim is to maintain the processing temperature as low as possible (below ambient temperature), this principle is less important

Figure 3: Characteric pressure–temperature history of a food product during HP processing

2.2 Intrinsic and extrinsic processing parameters

The overall impact of HP processing is determined by the integrated effect of intrinsic and extrinsic parameters On the one hand, the major intrinsic parameters are: the target microorganism and its physiological state and food properties, such as pH, water activity, and composition On the other hand, the major extrinsic parameters are: pressure, process time and temperature, vessel material and parts and packaging material A good understanding of the influences of these factors is essential for optimizing the HP treatment and ensuring microbiological safety (Hogan, 2005)

21

2.2.1 Intrinsic processing parameters

The resistance of microorganism is highly variable, depending mainly on the type of organism and food matrix, as discussed below

Generally, HP can produce greater destructive effect in organisms with a greater degree

of structural complexity As a result, yeast and molds are more susceptible to pressure than bacteria, and can be inactivated using relatively low pressure (Hogan, 2005) In most case, treatment at pressure from 300 to 400 MPa for a few minutes is sufficient to

inactivate most yeast and molds (Daryaei et al 2008) While for most forms of

vegetative bacteria, significant reduction (usually higher than 4 units) in the population

are achieved when 400 – 600 MPa at room temperature are used (Rendueles et al

2011) In addition, gram positive organisms are more pressure resistant than gram negatives while spores are most resistant and can survive pressures greater than 1000

Mpa (Ledward et al., 1995) The physiological status of microbiological populations

subjected to HP processing can also influence pressure resistance Bacteria are more susceptible during the logarithmic phase as compared to stationary phase Also, differences between pathogenic strains belonging to the same genus or species have

been described (Rendueles et al., 2011; Ledward et al., 1995)

The microbial susceptibility to HP inactivation is clearly influenced by the condition of the environment where microorganisms are present such as pH, water activity, and composition (Hogan, 2005) The pH of food is one of the main factor affecting the growth and survival of microorganisms Yeast and molds are relatively resistant to low pH, and

a pH of 4.0 has little effect on these microorganisms when they are subjected to heat or pressure Vegetative cells of bacteria become more sensitive to pressure and heat in low

pH conditions (Ritz et al., 1998) On the other hand, compression of foods during HP

treatment may shift the pH of food as a function of imposed pressure Taking organic acids as an example, they are reliable to be presented in the dissociated form during pressurization and the pH can increase temporarily The direction of pH shift and its magnitude must be determined for each high pressure treatment (Hogan, 2005)

Lowering the water activity of a food can significantly influence the microbial growth

(Black et al 2007) Oxen and Knorr, 1993 showed that a reduction of water activity

22

(measured at one atmosphere) from 0.98-1.0 to 0.94-0.96 resulted in a marked reduction in inactivation rates for microbes suspended in a food Reducing the water activity appears to protect microbes against inactivation by HP processing; however, it is

to be expected that microbes may be sublethally injured by pressure, and recovery of sublethally injured cells can be inhibited by low water activity Consequently, the net effect of water activity may be difficult to predict (Hogan, 2005)

Food composition plays an important role in determining the susceptibility of microorganisms to HP The presence of fats, proteins, minerals and sugars serves as a

protector and increases microbial resistance to pressure (Rendueles et al 2011)

Nutrient-rich foods, such as milk or poultry meat can protect microorganisms (Ledward

et al 1995)

2.2.2 Extrinsic processing parameters

The critical extrinsic parameters include pressure, process time and temperature and packaging material The extent and duration of HP treatment influence the microbial inactivation An increase in pressure increases microbial inactivation However, the duration of treatment is increased that does not necessary leads to an increase in the

lethal effect (Ledward et al., 1995) As mentioned above, the microbial response to

high-pressure treatments depends on the type of microorganism For each of them, there is a pressure-level threshold beyond which no effects are detected by increasing the exposure time There also exists a pressure level at which increasing the treatment time causes significant reductions in the initially inoculated microbial counts (Yordanov and Angelova, 2014) The 600 MPa pressure is considered by many authors as threshold value and also is considered to be economical and microbiologically safe for achieving the pasteurization level if it is combined with temperatures in the range 35-55 oC (Stoica

et al., 2013) An increase in process temperature above ambient temperature, and to a

lesser extent, a decrease below ambient temperature, increases the inactivation rates of microorganisms during high pressure processing Temperatures in the range 45 to 50◦C appear to increase the rate of inactivation of pathogens and spoilage microorganisms

(Rastogi et al., 2007)

Literature review: high pressure processing

23

For in-pack products, during the pressurization stage, packaging material such as polypropylene has been shown to act as an insulating barrier It prevents heat loss from the product and slows down the heating rate of the product (Juliano, 2005) Therefore, careful selection of a packaging material with good heat transfer properties can improve

heat transfer and thus reduce process non-uniformity (Knoerzer et al., 2010)

2.3 Effect of HP on color, flavor, texture of apple juice

HP processing is an attractive non-thermal process because the pressure treatments require to inactivate bacterial cells, yeast and moulds have a minimal effect on the sensory qualities associated with “fresh like” attribute such as texture, color and flavor

HP processing has been widely reported to keep color properties closer to the made juice than thermal processes, as well as other organoleptic and nutritional

fresh-features (Cheftel, 1995; Oey et al., 2008) However, most of studies were constructed on

citrus juices, studies on apple juice were limited

Donsi et al., 2010 evaluated the effectiveness of a pulsed high HP treatment on Annurca

apple juice at pressure levels of 150 to 300 MPa, temperature levels of 25 to 50oC They reported there were no difference with the brightness and a* values of the juice immediately after the treatment, however a significant increase of the brightness and a* values was detected during storage under refrigerated conditions (4oC) for 21 days

It is generally assume that the flavor of fruits is not altered by HP processing, since the

structure of small molecular flavor is not directly affected by HP Kim et al (2012)

demonstrated that treatment of apple juice at 500 MPa, 25°C, 3 min did not cause

significant changes in vitamin C content Novotna et al (1999) compared the aroma of

apple juice treated by high pressure or pasteurized at 80°C for 20 min and reported that high pressure treated samples were better than pasteurized samples

On the other hand, the effects of pressure treatment on enzyme need to be considered as enzymatic reactions can alter the content of some flavor compounds change the texture

in fruits Weemaes et al., 1998 reported a pressure inactivation of polyphenoloxidase

(PPO), enzyme responsible for fruit browning and flavor loss, in apples treated at 600 MPa and at room temperature (25°C) Riahi and Ramaswamy, 2003 evaluated the HP