View
217
Download
1
Category
Preview:
Citation preview
INFORMATION TO USERS
This manudpt has been reproduced from the miuofilrn master. UMI films the
text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be fmm any type of
computer printer.
The quality of this reproduction is dependent upon the quality d the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleedthrough, subsiandard margins, and improper alignment
can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manusuipt and
there are missing pages, these will be noted. Also, if unauthorized copyright
material had to be removed, a note will indicate the deletkm.
Oversize materials (a-g., maps, drawings, charts) are reproduced by sectioning
the original, beginning at the upper lefthand comer and continuing from left to
right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 8' x 9" Mack and white photographic
prints are available for any photographs or illustrations appearing in this copy for
an additional charge. Contact UMI directly to order.
8 8 H & Howell Information and Learning 300 North teeb Road, Ann Arbor. MI 481061346 USA
THE EFFECT OF DIFFERENT STORAGE CONDITIONS ON THE
QUALITY OF ORANGE JUICE
Mary lene Lagace
Department of Food Science and Agricultural Chemistry
McGill University, Montreal
October, 1 998
A thesis submitted to the Faculty of Graduate Studies and Research in partial Fulfilment
of the requirements of the degree of Master of Science
National Library ($1 of Canada Bibliotheque nationale du Canada
Acquisitions and Acquisitions et Bibliographic Services services bibliographiques
395 Wellington Street 395. rue Wellington OttawaON K1AON4 Ottawa ON KIA ON4 Canada Canada
The author has granted a non- L'auteur a accorde une licence non exclusive licence allowing the exclusive permettant a la National Library of Canada to Bibliotheque nationale du Canada de reproduce, loan, distribute or sell reproduire, priiter, distribuer ou copies of this thesis in microform, vendre des copies de cette these sous paper or electronic formats. la forme de microfiche/film, de
reproduction sur papier ou sur format electronique.
The author retains ownership of the L'auteur conserve la propriete du copyright in this thesis. Neither the droit d'auteur qui protege cette these. thesis nor substantial extracts &om it Ni la these ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent &re imprimes reproduced without the author's ou autrement reproduits sans son permission. autorisation.
ABSTRACT
Unpasteurized (condition A) and pasteurized (condition 8) orange juice samples
were stored frozen for eight months. In addition, pasteurized samples were also
aseptically packaged and stored at +1 OC in polyethylene bags (condition C). Nine quality
parameters were monitored during the eight months of storage: sedimentation of the pulp.
cloud measurement. aroma volatiles, ascorbic acid concentration. viscosity, density,
colour. sugar content (sucrose. glucose and fructose). organic acids (malic and citric), in
addition to sensory analysis. The optimum storage condition for freshly processed orange
juice was the unpasteurized frozen storage method (condition A). The juice retained most
of its chemical and physical properties and was rated by a sensory panel to have the
highest sensory score.
Des echantillons de jus d'orange non-pasteurises (condition A) et pasteurises
(condition B) sont congeles pour huit mois. De plus, des echantillons pasteurises sont
aussi emballCs aseptiquement et entreposes a +I OC dam des sacs de polyethyhe
(condition C). Neuf paramktres de qualite sont ttudies durant les huit rnois
d'entreposage: la sedimentation de la pulpe, la mesure de l'opacite, Ies arhes volatiles.
I'acide ascorbique. la viscosite, la densite. la couleur. le contenu en sucre (sucrose,
glucose and fructose). les acides organiques (malique et citrique), de mOme que
l'evaluation sensorielle. La condition d'entreposage optimale pour un jus d'orange frais
presse est la methode de congelation sans pasteurization (condition A). Ce jus retient la
plupart de ces proprietes. chimiques et physiques. et obtient la note la plus haute lon de
l'evaluation sensorielle.
ACKNOWLEDGEMENTS
I would like to thank my supervisor Dr. V.A. Yaylayan for his advice and his supervision,
and my co-supervisor Dr. E. Farnworth, researcher at the Food Research and
Development Center at Saint-Hyacinthe, for his availability and his patience.
Dr. R. Couture and A. Lassonde Inc. company who permitted me to do a master's degree
while working hl l time.
Jean Ledoux and Diep Vu froin A. Lassonde company for their help with HPLC analyses
and viscosity measurements. Guylaine Dery and Carole Marchand in charge of sensory
evaluation.
Brian Stewart for his availability and his expertise in gas chromatography analysis.
And finally thank you to my family for their suppon, my boyfriend Francois Brion (for
his patience), and all my colleagues at FRDC, especially Pascal Daigle. Julie Barrette.
Denise Chabot. Franqois St-Germain. Christine Gendron. Claude Gagnon. etc.. who
helped me in many ways during my master's degree.
TABLE OF CONTENTS
ABSTRACT ii
iii
ACmO WLEDGEMENTS iv
TABLE OF CONTENTS v
LIST OF TABLES ilr
LIST OF FIGURES xiii
GENERAL INTRODUCTION I
LITERA TURE REVIEW 3
2.1 Principal parts o f an orange 3
2.2 Pasteurization I
2.2.1 Purpose of pasteurization 4
2.2.2 Pasteurization Time-Temperature 5
2.2.3 ChemicaI changes during pasteurization 5
2.3 Storage 6
2.3.1 Frozen juice 6
2.3.2 Aseptically packaged juice 7
2.3.3 Storage studies of orange juice 8
2.3.4 Indicators of chemical degradation 9
2.4 Principal compositional analyses performed during storage studies 9
2.42 Ascorbic acid IS
2.4.3 Sugars and organic acids 18
2-43. I Sugars I 8
2-4-32 Organic acids 19
2.4.4 Colour 20
2.4.5 Sedimentation of pulp, cloud. viscosity and density -- 7 7
2.1.5.1 Sedimentation of pulp 12
2.4.5.2 Cloud 23
2.4.5.3 Viscosity and density 23
2.4.6 Sensory analysis 24
CHAPTER 3
M4 T E W S AND METHODS 26
3.1 Experimental design 26
3 5 Sampling of otange juice at Mexico 28
3.3 Sample preparation for analytical analyses 29
3.5 Analytical methods 29
3 . 4 Sedimentation of the pulp 29
3.4.2 Cloud measurement 3 0
3.4.3 Volatiles 5 0
5-43.! Gas chromatography coupled with a mass spectrometer (GC-MS) 30
A. Internal standard mixture 3 0
B. Volatile standards 3 1
Primary stock solution 3 I
Secondary standard solutions 51
C. Sample preparation 34
D. GC-MS conditions 34
3.4.3.2 Gas chromatography with a headspace injector (Headspace-GC-FID) 35
A. Internal standard mixture 35
B. Primary stock solution 35
C. Sample preparation 36
D. Headspace-GC conditions 36
3-44 Viscosity and density measurements 36
3.4.5 Ascorbic acid 37
3.4.6 Colour measurement 38
3.47 Sugars and organic acids 38
3 N . 1 High performance liquid chromatography (HPLC) 38
A. Standard sugar soiutions 38
vii
B- Standard organic acid solutions 38
C. Standard calibration cume 39
D. Sample preparation 39
E. HPLC conditions 39
3.4.7.2 "Brix 40
3.4.7.3. Titratable acidity 40
3.4.8 Sensory evaluation 41
3.4.9 Statistical treatment 42
RESUL TS AND DISCUSSION 43
4.1 Effect of storage condition on individual parameters 43
4.1 . I Density 43
4.1.2 Cloud and sedimentation of the pulp 45
4.1.3 Sugars 47
4.1.3.1 Brix measurement 47
4. 1.3.2 HPLC measurements 48
4. I .-I Organic acids 52
4. I A. 1 Titratable acidity measurements 5 2
4.1 A.2 HPLC measurements 54
4.1.5 Ascorbic acid 57
4.1.6 Viscosity 60
4.1.7 Colour 63
4.1.8 Volatiles 67 - - - --
4. 1.9 Sensory evaluation 83
4.2 Summary of all the parameters 90
4 3 Correlation between chemical and physical changes and sensory evaluation 92
CHAPTER 5
CONCL LlSIOlV 94
REFERENCES 96
APPErnLXA
PROFIL DESCRIPTIF QUANTITA TIF* I05
viii
APPENDIX B
EXAlMPLES OFSTATISTlCAL CALCULATIONS 108
APPErnIX C
CONCENTRATION OF VOLA TILES IN ORANGE JUICE FOR
THE MONTHS 2-6 112
LIST OF TABLES
Table number Page
Volatile components which contribute to fresh orange flavour 10
Thermally degraded compounds derived from essential oil constituents 12
Degradation products in canned single strength orange juice after
12 weeks storage at 35 O C 15
Parameters measured 27
Volatile standards for the primary stock solution ( 100 mL) 32
Density value of orange juice over an eight month storage period 44
Equation of the regression curves and coefficients of regression (R')
of the density of orange juice over an eight month storage period 44
Sedimentation of the pulp (measured at 3 time intervals) of orange
juice over an eight month storage period 45
Cloud stability (measured at 3 time intervals) of orange juice over
an eight month storage period 16
%rix value of orange juice over an eight month storage period 47
Equation of the regression curves and coeficients of regression (R')
of the "Brix of orange juice over a eight month storage period 38
Sucrose concentration of orange juice over an eight month
storage period 39
Glucose concentration of orange juice over an eight month
storage period 49
Fructose concentration of orange juice over an eight month
storage period 50
Results obtained from the t-test for each sugar contained in orange juice-5 1
Table number
X
Page
Equation of the regression curves and coefficients of regression
(R') of the three sugars of orange juice over an eight month
storage period 52
Titratable acidity measurements of orange juice over an eight month
storage period 53
Equation of the regression curves and coefficients of regression (R') of the
titratable acidity measurement of orange juice over an eight month
storage period. 53
Citric acid concentration of orange juice over an eight month
storage period 54
Malic acid concentration of orange juice over an eight month
storage period 55
Results obtained from the t-test for each organic acid contained in orange
juice 56
Equation of the regression curves and coefficients of regression (R')
of the two organic acids of orange juice over an eight month
storage period 56
Ascorbic acid concentration of orange juice over an eight month
storage period 57
Equation of the regression curves and coefficients of regression (R')
of the ascorbic acid concentration of orange juice over an eight month
storage period 57
Viscosity measurement of orange juice over an eight month
storage period 61
Equation of the regression curves and coefficients of regression (R')
of the viscosity measurement of orange juice over an eight month
storage period 61
Table number
xi
Page
Colour measurement (L value) of orange juice over an eight month
storage period 64
Equation of the regression curves and coefficients of regression (R')
of the colour measurement (L value) of orange juice over an eight
month storage period 65
Volatile concentration of orange juice for the first month of storage 68
Methanol concentration of orange juice over an eight month
storage period 70
Equation of the regression curves and coefficients of regression (R')
of the methanol concentration of orange juice over an eight month
storage period 70
I -Hexan01 concentration of orange juice over a six month
storage period 72
Equation of the regression curves and coefficients of regression (R')
of the 1-hexanol concentration of orange juice over an eight month
storage period 73
P-Myrcene concentration of orange juice over a six month
storage period 75
Equation of the regression curves and coefficients of regression (R')
of the P-myrcene concentration of orange juice over an eight month
storage period 75
Limonene concentration of orange juice over a six month
storage period 78
Equation of the regression c w e s and coefficients of regression (R')
of the limonene concentration of orange juice over an eight month
storage period 78
a-Terpineol concentration of orange juice over a six month
storage period 8 1
xii
Table number Page
39. Equation of the regression curves and coefficients of regression (R')
of the a-terpineol concentration of orange juice over an eight month
storage period 8 1
40. Mean scores of sensory evaluation of orange juice over eight months
of storage 85
4 1. Summary of the changes during the period of storage 90
LJST OF FIGURES
Figure number Page
Cross section of a citrus fruit 3
Volatile quality indicators 13
Principal formation pathway of alpha-terpineol 14
Anaerobic and aerobic degradation of vitamin C (AA) in orange
juice: DKA, diketogulonic acid: HF. hydroxyfurfural 16
Experimental design of the three storage conditions 27
GC-MS chromatogram of the 36 volatile standards used in the analysis
of the orange juice samples 33
HPLC Chromatogram of the three sugars in orange juice processed by
condition A at the last month (8) of the storage period 50
HPLC chromatogram of the two organic acids in orange juice
processed by condition A at the last month ( 8 ) of the storage period 55
Ascorbic acid concentration of orange juice over an eight month
storage period 58
Viscosity of orange juice over an eight month storage period 61
Colour measurement (L value) of orange juice over an eight month
storage period 66
Methanol concentration of orange juice over an eight month
storage period 7 1
1-Hexanol concentration of orange juice over a six month
storage period 74
Beta-myrcene concenvation of orange juice over a six month
storage period 76
Limonene concentration of orange juice over a six month
storage period 79
Figure number
xiv
Page
Alpha-terpineol concentration of orange juice over a six month
storage period 82
Evolution of the attribute "orange taste" in orange juice over an eight
month storage period 86
Evolution of the attribute "fresh pressed" in orange juice over an eight
month storage period 87
Evolution of the attribute "homogeneity" in orange juice over an eight
month storage period 88
Evolution of the attribute "presence of pulp" in orange juice over an
eight month storage period 89
CHAPTER I
GENERAL INTRODUCTION
Orange juice is the most popular fruit juice but it is also the least stable (Shaw.
1996a) due to its delicate and complex flavor which is easily altered during heat treatment
and storage (Shaw er al., 1993; Shaw. 1996a). In Canada, the refrigerated juice market is
worth 190 millions dollars and more than 56 millions dollars only in Quebec (Auger.
1997). Since World War 11. the emergence of new technologies has permitted processing
industries to make orange based products market favorites. Canned, dry crystals, frozen
concentrate, chilled ready-to-serve, "not from concentrate" and fresh squeezed are
possible products for orange juice developed through the years (Brown er al., 1993). The
consumers are now considering orange juice "not from concentrate" as a premium quality
product. The composition of orange juice can depend on many factors such as growing
condition. various treatments and practices. maturity. rootstock. variety and climate.
These factors are associated with the orange fruit whereas each of them have intrinsic
properties (Velduis. 1971). Different cultivars of sweet oranges such as Harnlin. Mam,
Parson. Brown. Pera. Pineapple. Sharnouti and Valencia are available. The juice made
from freshly squeezed Valencia oranges has an excellent quality and is considered by
many as a standard by which all other juices must be compared with (Nagy, 1 996).
The shelf life of citrus juices and related beverages are primarily determined by
microbial growth and by chemical changes. Many factors have an impact on shelf life
such as processing conditions (pasteurization time and temperature), oxygen content,
properties of and type of container, storage temperature and type of product (concentrate.
single-strength, chilled. unpasteurized) (Shaw el al., 1993).
The most important factor in determining the shelf life of aseptically packaged
orange juice, concentrate and fresh squeezed juice is the storage temperature (Graumlich
et a/., 1986; Fellen, 1988). The juice can be stored aseptically (stainless steel tank or
polyethylene bag) or in frozen blocks. The pasteurized product stored aseptically above
freezing temperatures is more economicd than storage of frozen product. However
aseptically stored product can lose some of its desirable flavour top-notes and develop
non-desirable flavours (off-flavour) if it is stored for a long period (Shaw et al., 1993).
Many studies have been done on the storage of concentrated juice since the 1940's
(Curl, 1947) by measuring parameters that have been shown to be related to orange juice
quality, whereas the "not from concentrate" juice has not been studied much. This fact
will probably change in the next few years considering the popularity of this type of
product. since its quality depends in part or. the mode of storage used after pressing. This
study is the most complete research done on the applicable storage condition of the juice
"not from concentrate". A variety of instrumental and sensory evaluation techniques were
used to measure changes in the juice composition and properties. The micriobiological
aspect was not a part of this study because with low temperature and pasteurization, no
change was expected.
The objective of this project was to identify the optimum storage
conditions (time, temperature and type of container) for freshly
squeezed orange juice samples (pasteurized or not pasteurized).
CHAPTER 2
LITERATURE RE MEW
2.1 Principal parts of an orange
SEGMENT MEMBRANE
Figure 1 . Cross section of a citrus fruit (Ting and Rouseff. 1986).
The flavedo (or epicarp) is the colored portion of the peel. This part contains the
carotenoids which give the colour of the citrus fruits. Under the flavedo, is the albedo or
mesocarp. This fraction is a thick. white and spongy layer. A large part of the pectic
substance is found in the albedo. The juice vesicules are found in 9 to 13 segments
separated by segment membranes (Ting and Rouseff, 1986). During the pressing, the
fruit is cut in two parts and squeezed to obtain a mix of serum juice and pulp. Thus
components arc found from all the parts of the fruit.
2.2 Pasteurization
The juice is subject to quality degradation especially at ambient temperature due
to microbiological, enzymatic activities and chemical reactions (Chen er of., 1993). All
these changes reduce the shelf life of the orange juice because of the effects on the
nutritional quality, colour and flavour.
2.2.1 Purpose of pasteurization
Pasteurization of orange juice is required for two reasons: inactivation of pectic
enzymes and destruction of microbial population capable of causing spoilage (Carter
198 1 : Graumlich er a/., 1986). The pectic enryrne, pecdnmethylesterase (PME), naturally
present in orange juice. catalyses chemical reactions with the pectin molecule (Carter
198 1). This reaction produces a loss of cloud (clarification) in juices or gelation of
concentrates (Graumlich et (11.. 1986). Maintaining cloud is important to eye appeal and
retention of certain flavour compounds associated with the cloud matrix (Sadler et a/.,
1992). Once inactived. the PME cannot regenerate or increase in concentration (Carter.
198 1). Pasteurization is used to achieve commercial stability of cloud of various citrus
products not just orange juice (Graumlich el al.. 1986).
Shorter time-temperature pasteurization is required for microbial than for PME
inactivation (Carter, 1981). Heat pasteurization procedures established for enzyme
inactivation are therefore sufficient for microbial inactivation (Chen et al., 1993), but due
to non-aseptic transfer of the juice a second pasteurization is required for complete
microbial destruction.
2.2.2 Pasteurization Time-Tern perature
The time-temperature relation is the most important factor when pasteurization of
orange juice is carried out. At lower temperatures the time required is longer. and shorter
for higher temperatures. Most of the time two pasteurization steps are done. The first is
performed when the juice is extracted. The pectin enzymes are inactived by immediate
heat treatment at 90 C for 7 seconds. This treatment is often used in the production of
Florida orange concentrate according to Carter ( 198 1). The second pasteurization is done
prior to packaging for commercial purposes on reconstitued juices. which don't have
residual enzyme activity. A temperature ot about 74 O C for 16 seconds is usually
recommended (Carter 198 1 ).
Many time-temperature pasteurization procedures exist: the combination chosen
depends on the product.
2.23 Chemical changes during pasteurization
Even when the heat treatment has been used to increase the shelf life of orange
juice. certain chemical processes can continue that will adversely affect the flavor of the
product due to the formation of off-flavour compounds during storage. The most
significant changes due to pasteurization are increases in the following known oxidation
products of the major oil component d-Iirnonene: alpha-terpineol, 4-terpineol. carve01
and pans-mentha-2.8-dien- 1-01 (Schreier er a!.. 1977) The most effective way to
minimize the degradation of flavour has been to store the juice at refrigerated temperanue
(Shaw el ai.. 1993).
2.3 Storage
The major factor affecting the quality of orange juice concentrate is the
temperature of storage and the composition of the storage container (Bokhari er al., 1995).
The permeability of the container to oxygen is an important factor for the shelf life of
stored pasteurized orange juice (Marshall et al., 1986), because oxidization is the primary
reason for flavour and vitamin C degradation during storage. Glass bottles are the
containers that provide the longest shelf life (Shaw. 1996b) compared to other types of
packaging. like plastic and polyethylene. If they are sealed correctly. no oxygen can pass
through the glass and no absorptioil of aroma by the packaging occurs.
The juice "not from concentrate" can be stored frozen in blocks. and thawed when
the final product is required, or stored aseptically for the needed period of time above
freezing temperature (Shaw et al.. 1993; Shaw. 1996b).
2.3.1 Frozen juice
Freezing large blocks of single-strength juice requires more energy than
aseptically stored products and therefore the fieezing method involves more cost (Shaw er
al., 1993). Many reactions are stopped at fieezing temperatures and others are slowed
down. Freeze preservation is known to keep degradative processes to a minimum during
long periods of storage (Olson, 1968). The quality of natural juice preserved by fieezing
without heat treatment and consumed immediately after thawing, is the closest to that of
freshly squeezed orange juice (Merin and Shorner, 1984). Certain studies on storage of
juice concentrate have used as a reference control juice stored at -18 OC (Tatum er al.,
1975; Marcy e l ul., 1984; Marcy et al., 1989). However, even at this temperature,
reactions between the components are not prevented. For example, Kefford et al. (1959)
and Kanner et al. (1982) reported small losses of ascorbic acid during storage of
concentrate at -18 OC. Bokhari ef a!. (1995) found little change for vitamin C and acidity
for a concentrate stored at -15 OC. Therefore, changes can still occur in frozen orange
juice. However. the effect of freezing on juice "not fiom concentrate" has not been
studied thoroughly and the application of the results obtained for juice fiom concentrate,
probably are not applicable.
According to Merin and Shomer (1984) the rate of separation of insoluble
particles from the serum provides an indication of physical changes which occur in juice
during frozen storage. The primary freezing rate is an important factor. If the juice is
quickly frozen. the separation of thawed juice after a relatively long storage is delayed.
The tendency for separation is greater after a slow freezing rate because large ice crystals
formed cause compression of insoluble particules causing aggregation. Aggregation
appears in thawed juices for all type of treatments after a long period of storage in the
frozen state. Particular attention must be taken when the juice is frozen to minimize the
effect of crystallization of water especially for not concentrated juice.
2.3.2 Aseptically packaged juice
Low density polyethylene (LDPE) is the most commonly used material for the
inner layer of a multilayer package. Aseptic packaging produces a sterile product. On the
other hand. certain factors affect the shelf life of product packaged aseptically such as
oxygen which can permeate along the seams of the package and react with orange juice
components (Varsel, 1980). Another problem is the absorption by LDPE of nonpolar
volatiles particularily terpene hydrocarbons (such as d-lirnonene) and aldehydes. This
phenomenon is called "scalping" and it decreases the level of flavour components (Shaw
et a!., 1993). Another major factor affecting the shelf life of the product is storage and
distribution temperature. It is essential to maintain the juice at refrigerated temperatures
during storage, transportation and retail marketing in order to keep the quality high (Shaw
et a/., 1993).
2.3.3 Storage studies of orange juice
Many studies have been carried out on the effects of temperature of storage on
orange juice concentrate. The studied parameters are always related to these three
important criteria of quality: the flavour, the colour and the texture.
Kanner ef a!., (1982) studied the stability of orange juice concentrate. Their
samples were stored in metal cans at -18, 5 , 12, 17. 25 and 36 "c. Nonenzymatic
browning, ascorbic acid degradation. furfural accumulation and sensory quality were
measured over a period of 18 months. In 1984. Marcy er ai. stored in metal cans orange
juice concentrate at -12.2, -6.6, -1.1 and 4.4 OC and analyzed O~rix . % acidity, ascorbic
acid, furfural. serum viscosity. apparent viscosity. browning, Hunter colour values and
taste panel scores at monthly intervals for one year. Later, a study on processed
pasteurized orange juice was done by Kaanane el al. (1988). Total solid. pH and acidity.
form01 index. total sugar. ascorbic acid content and furfural production were investigated
on samples stored at 4. 22.5, 35 and 45 OC respectively in glass bottles for 14 weeks.
Marcy el ai. (1989) undertook another study on aseptically packaged orange juice
concentrate and orange drink concentrate. The juice was stored in laminated bags at 4.
15, 22 and 30 O C for 6 months. Ascorbic acid, nonenzymatic browning and sensory
quality were measured monthly. More recently, Bokhari er al. (1995) examined the
quality of packaged and stored Kinnow (mandarin or tangerine oranges) juice concentrate.
Samples were placed at three temperatures: -15,4 and 32 "C in high density polyethylene
bags and low density polyethylene bags. The target parameters were @Brix, acidity,
ascorbic acid. pectinesterase activity, organoleptic evaluation, total bacterial plate count
and yeast.
2.3.4 Indicators of chemical degradation
Some chemical compounds are useful tools to evaluate the quality of a product.
Lee and Nagy (1996) have drawn a list of seven chemical markers for the citrus industry:
ascorbic acid. dehydroascorbic acid. hydroxymethylfurfura1, furfural, 2.5-dimethyl-4-
hydroxy-3(2H)-hone, 2,3-dihydro-3.5-dihydroxy-6-methyl-4H-py-4-one. 4-vinyl
guaiacol and alpha-terpineol. These compounds can give some indications of temperature
and storage abuse.
2.4 Principal compusilionai analyses performed during storage studies
Freshly pressed orange juice has a full. fruity flavour quality that has not been
completely duplicated in any orange juice product. The overall total quality of a juice
depends on many criteria. Various analyses are done in order to have as much
infomation as possible on the changes in composition of the juice as a hc t ion of the
storage temperature. Flavour degradation is considered as the most important factor in
quality loss of citrus juice products.
2.4.1 Volatiles
A volatile fluid is a liquid with the tendency to become vapour at low temperature
and pressure values. Volatiles contribute to the flavour, and the flavour degradation is
considered as the most important factor in quality loss of citrus products. The volatiles
have been studied to understand the degradation process due to storage temperature. The
scarce information available on the contribution of volatile compounds to the sensory
impression of a complex aroma is vague and partially contradictory (Ziegler, 1970).
Orange flavour is the result of a combination of volatile components in specific
proportions. The relative contribution of individual components to the overall juice
flavour is influenced by the threshold values and by possible synergy between
components. Also, volatile and non-volatile components present in orange juice may
interact with each other and influence its flavour (Amed er al., 1 978). There are several
classes of compounds that contribute to the distinct flavor of orange: terpenes, aldehydes.
esters and alcohols (Shaw. 199 1). Shaw (1 W6a) summarized volatile components known
to contribute to fresh orange flavour (Table 1).
Table 1.
1 996a).
Volatile components which contribute to Fresh orange flavour (Shaw.
Hydrocarbons d-limonene
I citronella1 I methyl butanoate I
myrcene a-pinene valencene
I sinensal I I
Aldehydes acetaldehyde
The hydrocarbons limonene, myrcene and pinene are the three major constituents
of peel oil (Shaw er al.. 1977). These components are easily absorbed and/or adsorbed by
the low density polyethylene (LDPE) layer used in packaging due to the affinity of the
LDPE for the non-polar components. Halek and Meyer (1989) found approximately 30%
total loss of d-limonene and alpha-pinene from the studied solution. Concerning the
myrcene. it seems to be the most affected with 40% loss due to sorption. The absorption
of d-limonene increases the permeation of oxygen through the packaging material. Sadler
and Braddock (1990) have shown that the permeability of the oxygen by the LDPE was
proportional to the absorbed mass of limonene. In addition, the temperature also has an
effect on the absorption. The absorption of volatiles and the permeability to oxygen were
substantially lower at 4 O C than at 25 O C (Pieper et al., 1992).
oc tanal nonanal decanal
Esters ethyl acetate
Alcohols ethanoI
ethyl butanoate ethyl 2-methylbutanoate
ethyl 2-hydroxyhexanoate
(E)-2-hexenol ' (2)-3 -hexen01
linalool
Two categories of volatile flavour compounds can be identified. Depending on
the polarity of the components, a volatile can be found in the oil-soluble constituent
present in peel oil and in juice oil, or in the water-soluble constituents present in the juice
(Nagy. 1996). The peel oil of citrus fruit is located in small, ductless glands present in the
flavedo, or the outer portion of the peel. During the extraction, a small amount of peel oil
ends up in the juice and provides the characteristic. pleasant citrus aroma of intact h i t
(Nagy, 1996). However, some of the flavour components. including the "top-notes", are
provided by volatiles present in the juice sac. including oil soluble components in juice oil
located within the juice sac (Shaw. 199 1). The water-soluble volatile components present
in the juice sac are considered to contribute most to the characteristic flavour of fresh
citrus fruit (Huet, 1969). However. the flavour of orange juice is a complex combination
of aqueous essence, essence oil and peel oil. The major component of orange and peel
oil. d-limonene, is generally considered as non-essential to orange flavour but it acts as a
precunor of alpha-terpineoi. which is know as an o ff-flavour compound (Dm, 1 980).
The changes occuring in orange juice after processing and storage can be divided
in two classes: ( I ) loss of original flavour and (2) development of off-flavours and off-
colours (Shaw er al.. 1993). Both of these events involve volatiles.
The storage is an important factor to consider. especially at ambient temperature.
The formation of compounds which impart off-flavour is more pronounced under these
conditions. A summary of off-flavour compounds derived from the thermal
decomposition of citrus oil constituents is presented in Table 2. The formation of off-
flavours during storage is often due to chemical changes o c c u ~ g during the initial heat
treatment of the juice (Shaw er al.. 1993).
Some volatile components are recognized as degradative indicators:
hydrox ymethy 1 fiufural commonly cal ied HMF, furfural, 4-viny lguaiacol. 2,5-dimethyl-4-
hydroxy-3(2H)-bone ( h e o l ) and alpha-terpineol (Lee and Nagy , 1996). 4-
Vinylguaiacol, h e 0 1 and alpha-terpineol were judged to be most responsible for
malodorous properties of time-temperature abused juice (Shaw et a[., 1993). Figwe 2
shows the structure of different chemical indicators (Lee and Nagy, 1996).
Table 2. Thermally degraded compounds derived from essential oil constituents
(Shaw et al., 1993).
Precursor I
d-limonene ' linaloo 1 '
I cis- 1.8-p-rnenthanediol* 1 I 1 .
1.8-cineole I Pungent. cam~horaceous a-terpineol
p-mentha- 1 (7),2-dien-8-01
Derived compounds a-terpineol a-terpineo 1
nerol
'Present naturally in fresh j~
Flavour response Stale, musty, piney Stale, musty, piney Sweet. rose fruity
geraniol cis4.8-p-rnenthanediol
*Also know as cis-terpineol.
Sweet, floral rose Sweet. carn~horaceous
p-cymene I Terpiney off-flavour
1.4-cineole p-mentha- l.5-dien-8-01
p-mentha- 1 (7).2-dien-8-01 cis-p-mentha-2.8-dien- 1-01 trans-p-mentha-2.8-die* 1 -
0 1 p-cymen-8-01
p-c ymene a,p-dimethy lstyrene
ice.
-- ------ Not characterized Not characterized Not characterized Not characterized
Nonspeci fied off- flavour Terpiney off-flavour Terpiney off-flavour
**Isomeric mixture of nerd and geranid.
Figure 2. Volatile quality indicators.
HMF Furfural
4-vinyl guaiacol
Furaneol
HMF is the primary breakdown product during the dehydratation of glucose or
fructose in an acid medium. In citrus products, the presence of HMF is suggested as the
first indication of change due to storage even though it is not responsible for the off-flavor
characteristic of stored citrus juice (Berry and Taturn. 1965). High concentration of HMF
in juice is considered to be due to excessive heat treatment during concentration or
pasteurization.
The second component. the furfurall, is produced by the aerobic and anaerobic
degradation of ascorbic acid (Bauemfeind and Pinken, 1970; Kaanane et al., 1988).
FurfUral and HMF may be related to the darkening of juice and are also useful indicators
of temperature abuse or storage time in diverse foods (Pompei er a[., 1986; Beeman.
1987). The furfiual and HMF content of tieshly processed citrus juice is essentially zero
but increases to significant amounts, during high-temperature storage (Nagy and Randall,
1973; Askar, 1984). F u r W does not contribute directly to the flavour changes but its
accumulation parallels to that of other compounds that alter flavour (Nagy and Smoot,
1977).
4-Vinyl guaiacol has been identified as the most detrimental off-flavour
compound in aged canned orange juice ( T a m et a/., 1975). It is formed by the
carboxylation of fermlic acid (Fiddler et al., 1967). The Cvinyl guaiacol imparts an old
h i t or rotten flavour to orange juice at a level of 0.075 ppm (Tatum et al.. 1975).
Like HMF, fbmneol is also a degradative product of sugar (Shaw et a!.. 1968).
Furaneol content gradually increased as a function of storage time and temperature (Lee
and Nagy, 1987). Futaneol gives a pineapple-like aroma to orange juice (Tatum ef al..
1975).
Concerning the alpha-terpineol, it is derived from the degradation of some
essential oil components of orange juice (Figure 3). I t s precursors are mainly linalool and
d-limonene, which are the major volatile organic constituents of orange juice.
Approximatively 2-3 ppm of alpha-terpineol is sensorially detectable as deterioration
(Diirr. 1980). Tatum et al. (1975) were more precise in indicating as little as 2.5 ppm of
alpha-terpineol added to freshly expressed juice caused a stale. musty or piney aroma to
orange juice. This compound is very important because of the large presence of limonene
in juice.
Figure 3. Principal formation pathway of alpha-terpineol.
Limonene
Hydration
Alp ha-terpineol
The volatiles are generally extracted fiom orange juice by distillation or by liquid-
liquid extraction with different solvents such as methyl chloride or diethyl ether. Gas-
chromatography (GC) is the method most often used for separation and quantification.
Mass spectroscopy (MS) and nuclear magnetic resonance (NMR) permit identification of
organic compounds. The most common GC system used for volatiles determination is
flame ionization detection. To measure very volatile components. a headspace injector
can be coupled with the GC-FID (flame ionization detector). Depending on the
composition of volatiles. either a polar or non-polar column can be used. Tatum el ul.
(1975) analyzed a canned single-strength orange juice and extracted the volatiles with
methylene chloride and used a gas chromatographic separation with a Carbowav column.
IR and MS spectroscopy was used for identification of the different peaks. Ten
degradation compounds were isolated from the canned juice stored at 35 O C for 12 weeks
(see Table 3).
Table 3. Degradation products in canned single strength orange juice after 12 weeks
storage at 35 "C (Tatum et al.. 1975).
FurfUral I
a-terpineo 1
2.4.2 Ascorbic acid
Degradation of ascorbic acid (vitamin C) in orange jl
cis- 1 .&p-menthanediol trans- 1 .8-D-menthanediol
I I
3-hydroxy-2-pyrone 1 4-vinylguaiacol
lice is well do'
2-hydroxyacetyl furan 2.5-dimethyl-4-hydroxy -3 (2H)-hanone
I
cumented,
probably because of its nutritional value. In 1980, Nagy studied the variability in the
vitamin C content of citrus fruits and their products as influenced by variety, cultural
practice, maturity, climate. fresh h i t handling, processing factors, packaging and storage
conditions. Storage time and temperature were found to be important factors in the loss
of vitamin C. Low temperature is imperative for the retention of vitamin C during storage
I
benzoic acid '1
5-hydroxymethy 1 Mimi b
(Nagy, 1980). The degradation of ascorbic acid can be effected by enzymatic or
nonenzymatic processes. Enzymatic degradation is controlled by pasteurization at high
temperature which disables the oxidative enzymes. The greatest loss of vitamin C in
processed product is due to aerobic and anaerobic reactions of a nonenzymatic nature
(Nagy, 1980). Both aerobic and anaerobic degradations occur in the same juice system
(Kennedy et d., 1992). Aerobic and anaerobic pathways for degradation of vitamin C in
aqueous medium (see Figure 4) have been proposed by Bauernfeind and Pinkert (1970).
Figure 1. Anaerobic and aerobic degradation of vitamin C (AA) in orange juice: DKA,
diketogulonic acid; HF, hydroxyfurfUml (Bauemfeind and Pinkert, 1970).
A A DHA OKA
HG-OH GHO
After oxygen is consumed (aerobic), vitamin C is degraded anaerobically but at
rates lower than the aerobic process (Nagy, 1980). However, according to Kennedy et ul.
(1992). the aerobic process generally predominates and the anaerobic process takes place
when the level of dissolved oxygen has reached equilibrium.
The processing of fruit to make juice products results in minimal loss of vitamin C
potency, but subsequent storage of the finished product at high temperatures results in
considerable losses. Shaw et al. (1993) summarized the factors correlating with
degradation rates of vitamin C reported by various studies. Solid content and storage
temperature, pH, presence of metal ions and fiee oxygen (headspace and dissolved in
juice) were considered important. Kennedy et al. (1992), studied the degradation of
ascorbic acid as a function of time and temperature ( -20, 4. 20. 37. 76 and 105 OC).
Samples stored at 20, 37, 76 and 105 OC showed dramatic decreases in vitamin C levels
after the tint few days of storage. This fact coincided with an initial drop of the dissolved
oxygen level. This correlation between dissolved oxygen and ascorbic acid stability was
not demonstrated in frozen samples (stored at -20 OC), but small loses of vitamin C were
observed after 23 days in the samples with the lowest losses of dissolved oxygen. In
1982, Kanner et al., demonstrated that the loss in ascorbic acid was limited at - 1 8, 5 and
12 'C but important at 25 and 36 OC. Generally all the authors (Kanner er 01.. 1982;
Marcy er a/., 1984; Kaanane et 01.. 1988; Marcy er a!., 1989. Bokhari et al., 1995) have
found a decrease in ascorbic acid at all temperatures (even at frozen temperatures); losses
increase with the temperature and storage time. Except Bokhari et af. (1995) who used a
titration with 0.1 N iodine. all the studies were perfomed with the AOAC method of
titration with 2.6-dichlorophenol indophenol to measure the ascorbic acid in solution
(Carter. 1981). Titration has been the most commonly used method. but new
developments in instrumental methodology provide the possibility of alternative methods
of analysis. Vitamin C can be quantified for example by fluorimetric procedures, the
dinitrophenylhydrazine method, or HPLC reverse-phase ion pair. These analyses have
been reviewed by Ting and Rouseff (1 986).
2.4.3 Sugars and organic acids
The popular flavour of orange juice is the result of a natural combination of
volatile compounds in a well-balanced system of sugars, acids and insoluble solids
(Shaw, 1986).
Sugars
The economically most important juice parameters. for the citrus grower or the
processor. are the total soluble solids ('Brix) and Brixacid ratio. Besides sugars, the
remainder of total soluble solids in the juice are organic acids. nitrogenous C O ~ Q O U ~ ~ S .
soluble pectic substances and other minor constituants (Ting and Rouseff. 1 986).
The main sugars found in orange juice are D-glucose. D-fructose and D-sucrose. D-
glucose and D-fructose are reducing sugars, while D-sucrose is non-reducing. The sugars
in orange juice are subject to change because of juice acidity. The ratio of sugar content
in processed juice can change during storage due to acid-catalysed hydrolysis of D-
sucrose to D-glucose and D-fructose (Chen et al.. 1993). When the orange is mature. total
reducing sugars ( D - ~ C ~ Q S ~ and D-glucose) are approximatively equal to D-sucrose. and
they are in almost equal proportion.
Sugar content of juice can be measured by several methods. Total sugar is often
measured as 'Brix. Sugars comprise 70-85 % of the total soluble solid material in orange
juice (Chen el al., 1993; Miller and Hendrix, 1996). O ~ r i x is a unit used to designate per
cent dissolved sugar (Carter. 1981). The official test for O ~ r i x is performed by a Brix
hydrometer. This instrument measures the specific gravity and it is calibrated to read
directly in degrees Brix, or percent of sucrose, at a standard temperature of 20 O C (Millers
and Hendrix, 1986). Unofficial method for reading of Brix content can be performed with
a rehctometer. Marcy et al. (1984) and Bokhari et al. (1995) used rehctometers in their
methods for determining of "Brix. Both methods are rapid and the results are ohen
expressed with acidity in the form of a ratio ('Bridacid). Brix and its acid ratio are
criteria used in determination of h i t maturity and juice quality (Carter. 198 1).
2.4.3.2 Organic acids
Organic acids in orange juice consist primarily of citric acid (85-95 Oh) and malic
acid. Smaller amounts of tartaric and succinic acids may be present (Chen et al.. 1993:
Miller and Hendrix. 1996). The two major organic acids in citrus juice (citric and malic)
can be separated and quantified using a reverse phase HPLC system (T ing and Rouse&
1986). However. the measure used to express acidity is often titratable acidity. A known
volume of juice is titrated to a phenolphthalin end point or pH of 8.1 using sodium
hydroxide. Results are expressed as g of citric acid / I OOmL (Carter. 198 1). This method
seems to be the most widely used for acidity (Marcy et al.. 1984: Kaanane er ol.. 1988:
Bokhari a ui.. 1995). All citric and malic acids in orange juice are not in their fiee form:
a part is in the salt form. The combination of acid and salt imparts great buffering
capacity to citrus juice (Chen et a!.. 1993). The brixacid ratio is first an index of legal
h i t maturity because acidity decreases when the soluble solids content increases (Chen
er a!.. 1993). This ratio plays an important part in acceptance (both by processors and
consumers) of the orange juice (Miller and Hendrix. 1996).
According to studies performed on stored orange juice concentrates. %rix and
acidity showed no significant change during the test period (Marcy et ai.. 1984; Kaanane
er al., 1988). Bokhari et al. (1995) found no significant change in OBrix but there was a
gradual decrease in the acidity throughout the entire storage period in all samples stored at
different temperatures. This fact caused a gradual increase in the "Brkacid ratio.
2.4.4 Colour
The two important factors affecting the appearance of the orange juice are opacity
and colour (Stewart, 1980). The orange colour of orange juice is due to carotenoids.
These C-40 compounds are primarily responsible for the red, orange and yellow pigments
found in citrus juices (Shaw et af.. 1993). The typical colour of orange juice is an
important factor for acceptance by the consumer. During storage, darkening of citrus
juices occurs. This phenomenon is called "browning". Browning reactions may be either
enzymatic or nonenzyrnatic (Paul. 1972). Browning increases with increasing storage
time and storage temperature. Enzymatic browning is usually not considered important
because of high temperatures used during pasteurization.
During storage studies of citrus juices, the nonenzymatic browning is often
measured. The reactions are very complex and many compounds found in orange juice
can undergo browning reactions. Nonenzymatic browning reactions frequently involve
reducing sugars or sugar-related compounds with amino acids in Maillard-type reactions
(Paul. 1972). Maillard was the first to describe the development of a brown color in
mixtures containing amino acids and reducing sugars. Varsel (1980) considered the
oxidation of ascorbic acid as the major factor in browning of citrus products. Besides
ascorbic acid. large amounts of other organic acids and their salts create favourable
conditions for degradation of sugars. amino acids, and phenolic compounds during
processing and upon storage via browning reactions (Lee and Nagy, 1988). The
formation of melanoidar pigments result fiom reaction of furfUral with amino acids or
fiom furfural polymerization. The interaction of nonemymatic browning and oxidation
reaction products with juice constituents is very complex and catalytic behavior of one
type of reaction on the other may reduce the predictability of the quality degradation that
can occur (Adams, 1989). Robertson and Sarnaniego (1986) found a highly significant
correlation between browning index, HMF and fumval formation. They suggested that
all three serve as chemical indicators of storage temperature abuse in lemon juices.
Handwerkand and Coleman (1988) studied the role of amino acids. sugars, ascorbic acid.
buffer and catalysts, sulfur containing amino acids and thiols in the browning reaction.
They concluded that reducing sugars, ascorbic acid and probably other carbonyl
compounds are reactants in browning reactions and are the precursors to compounds of
taste significance. Amines and amino acids present in juice catalyse the initial reaction
and also take part in later sequences of the reaction that occur.
The measurement of colour was initially performed visually by difference. Today.
instruments are used for colour determination. Meydav et a/. (1977) proposed a
procedure to measure browning in citrus products. Centrihgation. dilution and filtration
are followed by a transminance spectra at 420 nrn. This method gives a "browning
index". A simpler method which measures the colour rehctance is done with an
instrument such as the Hunter Color Difference. This intrument is based on the
tristimulus principle of X, Y and Z (Nickenon, 1946) related with LAB values where L
represents the lightness. A the redness and B the yellowness. For browning, L is the most
important value. This method is faster than that of Meydav et at. (1977).
The tristirnulis calorimeter was used by Kanner et al. (1982) and Marcy et a!.
(1989). The latter used absorbance at 420 nm to confirm the colour change. They found
changes in the tristimulus attribute L occured more rapidely for juice concentrate stored at
30 O C more than 2 months. The colour of juice stored at 4 O C was not significantly
different after 6 months of storage. Kanner et UL (1 982) found an increase in the attribute
L for juice concentrate stored at 25 OC for at least 200 days of storage. But the
concentrates stored at 5 and 12 O C for 18 and 12 months were stable. Marcy et ol. (1984)
used absobance at 420 run and observed no change in colour of orange concentrate kept at
-12 O C for 1 year. However, slight changes in samples stored at -6.6. -1.1 and 4 O C were
found. The rate of browning depends in large part on the temperature of storage. Low
temperature is recommended to delay the development of brown colour.
2.4.5 Sedimentation of pulp, cloud, viscosity and density
Sedimentation of pulp, cloud, viscosity and density S e c t the appearance and the
mouth feel of orange juice and are therefore important for acceptance by consumen.
2.4.5. I Sedimentation of pulp
In addition to contributing to apparence and mouth feel. pulp also contributes to
the perceived flavour because it preferentially absorbs nonpolar vola~iles. The presence
of pulp modifies not only the intensity but also the balance of the overall aroma (Itadfort
er al.. 1971). Most of the aroma volatiles are anached to the pulp panicles: therefore an
orange juice should contain a sufkient amount of pulp material (Diim. 1980). The
separated pulp may coalesce and float to the top or settle at the bottom of the bottle
leaving a clear or slightly hazy serum resulting in an unattractive non-homogeneous
appearance of the product (Rangarma and Raghuramaiah. 1970). The sedimentation of
pulp becomes parricularily imponant when the frozen juice is thawed. The separation of
insoluble particles (pulp) From the juice serum increases with increasing fiozen storage
time and is especially apparent after a slower primary fieezing rate.
Men a juice is frozen, depending on the primary freezing rate. the insoluble
panicles are compacted between ice crystals. This factor causes flocculation. the pulp
settles more rapidly and the separation is more rapid. Merin and Shomer (1984)
measured the height of separated serum Eom a suspension of thawed nanual Valencia
orange juice after 0, 30, 60 and 90 days. Measuremenu were made at 0.5. 1, 2 and 3
houn after thawing. Three different methods of freezing were used: liquid nitrogen
fieezing (10 rnin), blast fieezing (3 hr) and chamber fieezing (24 hr). The results showed
that the primary rate of fieedng is very important. Thc senling rate was advanced with
the frozen storage time and in accordance with the sequence of the primary fmdng rate.
The samples frozen by the liquid nitrogen had the lowest rate of sedimentation among the
three methods. Quick primary freezing delayed the separation of thawed juice over a
relatively long storage period.
2.4.5.2 Cloud
In o m g e juice. an opaque or cloudy appearance is considered a desirable
characteristic. The cloud is the opaque appearence of citrus juice due to colloidal
suspension of particles (Ting and Rouseff, 1986), and the pectin is the protective colloid
(Carter. 198 1). The partideb can be ce!lulose. protein or lipids ( h e and Olssofi, 1994).
The pectin substances found in juice are complex carbohydrates and consist of
galacturonic acid and its methyl ester in chains of undetermined length (Kertesz, 195 1).
Clarification results from the natural enzymatic effect of pectinmethylesterase (PME).
Cloud is primarily stabilized by pasteurization which causes the deactivation of PME.
Without pasteurization, a part of methyl ester linkages are split, forming pectinic acids or
possibly pectic acid. which. with the calcium naturally present in the juice. forms a
precipitate (insoluble calcium pectate) and causes clarification or gelation of concentrated
juice (Rangma and Raghurarnaiah, 1970). In additions, other cloud forming
components are entrapped in this precipitation process, resulting in total loss in single-
strengh juices (Irwe and Olsson. 1994). No studies of the effect of frozen storage seem to
have been reported on the stability of juice cloud. According to Carter ( 198 1 ), measuring
cloud in citms juice provides an indication of abuse and stability. This can be performed
with a calorimeter (at 650 run).
2.4.5.3 Viscosity and density
Viscosity is associated with pectin - the most important contributor to juice
viscosity. Serum viscosity affects the mouth feel and the body of the juice. It is usually
measured with a Brookfield viscosimeter (Ting and Rouseff. 1986). Marcy er ul. (1984)
found no signifiant change in serum viscosity during one year of storage.
The density measurements have not been reponed in literature.
2.4.6 Sensory analysis
Sensory analysis is a very complex domain. Many studies on organoleptic
evaluation have been performed on various materials using a wide variety of methods and
producing a variety of results. Sensory analysis is an important method for evaluating
consumer acceptance of a product. The panel can consist of untrained. semi-trained or
trained (expert) members. The perception of flavour depends on two classes of
compounds: (1) the non-volatile compounds that possess taste attributes and are sensed
by taste buds in the mouth. and (2) the volatile substances that are odorous and sensed by
olfactory receptors in the nasal passage. Flavour is the combination of these two classes
(Hendrix and Hendrix. 1996). Sensory evaluation is a way of making a connection
between chemical changes and acceptability of the juice.
Kaanane et al. (1982) used at least 25 untrained panalists from among university
personnel. Samples were compared by a triangle test. Results showed no statistically
significant difference between concentrate stored at -1 8 O C and those stored at 5, 12 and
1 7 O C for 1 7, 10, and 8 months, respectively. After these periods, off-flavours developed
which were associated with a caramel-like taste. Marcy ei al. (1984) asked six trained
panelists to score samples in comparison with a reference juice (store at -1 7 OC). Panelists
were unable to detect a significant difference in orange juice concentrate stored at -12.2
and -6.6 "C during the 12 month of storage. Concerning the juice concentrate stored at - 1.1 and 4.4 O C , the scores of the panel were significantly different after 9 and 5 months
respectively. In 1989, Marcy et ui. determined sensory quality with an experienced 12
member panel by a difference test where ( 0 ) indicates no significant difference and (+)
indicates a significant difference from the control (stored at -1 8 OC). Analysis of the juice
concentrate showed that there was no significant difference from the control when stored
up to 6 months at 4 "C. On the other hand. the samples stored at 15 and 22 OC were
significantly different after 2 months of storage. Bokhari et a/. (1995), used a panel of
eight judges for evaluating the juice for colour and flavour. They found that more loss in
colour and flavour was observed at high temperatures as compared to low temperatures in
all the packaging materials.
CHAPTER 3
MA TERIALS AND METHODS
3.1 Experimental design
The research project started in May 1997 using Valencia orange juice from
Mexico. The samples were divided into three categories: (A) unpasteurized orange juice
stored at -1 8 O C in glass bonks, (£3) pasteurized orange juice stored at -18 O C in glass
bottles and. (C) pasteurized orange juice aseptically packaged in polyethylene bags
containing 1 1 85 Kg of juice stored at +lUC.
Pasteurized and unpasteurized samples were stored frozen (-18 O C ) for eight
months (May 1997 - December 1997). In addition. pasteurized samples were also
aseptically packaged and stored at + I OC in polyethylene bags (see Figure 5). At specific
time intervals, samples were analyzed for quality parameters.
To have a complete study of the effect of the three storage conditions; nine quality
parameten were analyzed as showed in Table 4. Analyses were carried out once a month
for a period of eight months. Certain of the nine parameters changed rapidly. especially
with the thawed unpasteurized juice. An order was established for analyses according to
relative importance.
Figure 5. Experimental design of the three storage conditions.
FRESH ORANGE JUICE
I Unpasteurized I
Frozen Glass bottle
-1 8OC
I Aseptically packagedl
Table 3. Parameters measured.
I PARAMETERS MEASURED I ORDER OF ANALYSIS 1 1 Sedimentation of the pulp
I
I 1 I . . 1 Cloud
I
I - 7 I
I Ascorbic acid I 5 I
b 1
1 Colour I 6 I
Volati les Viscosity and density
1 I
Sugar I 7 Organic acids 8
I
Sensory evaluation* 9 J
*The sensory analysis were perfirmed in parallel at ALessonde Inc. with on expert panel.
3 4
Each month. 4200 rnL (12 bottles of 350 mL) of frozen pasteurized, frozen
unpasteurized and aseptically packaged (sampled from a bin) orange juice were necessary
for the analysis. Sensory evaluation required 2 100 mL (6 bottles of 350 mL), and 2 100
mL (6 other bottles of 350 rnL) were required for the physico-chemical analysis to
perform one series of analysis on one juice. The 2100 mL (6 bottles) of each storage
condition for physico-chemical analyses were mixed by two to produce a total of 3 bonles
of 700 mL. On each bottle, the analysis were performed two times for a total of six
measurements per parameter, per storage condition, per month.
3.2 Sampling of orange juice at Mexico
The fruits were harvested in Nuevo Leon (Mexico), hand picked fruits were
brought by trucks. fiom the groves to the processing plant and extracted by Brown Citrus
Machinery. These extractors use a reaming action to extract juice from citrus h i t . except
for the model 1 100. The Brown model 1100 has a feeder that places the fruit into three
single lanes as it enters the extractor. For the production 8 emacton; model 400 and one
extractor model 1 100 were used. The capacity of the plant was 15 tons of fruits per hour.
After extraction. the juice was finished, to separate cloudy but othenvise clean
juice. fiom pulp. rag, seed and pips. After finishing, the juice was pumped into a holding
tank. The process was well defined by Rebeck (1995). Part of the juice was placed in a
250 liters tank and the sample of unpasteurized juice was placed in bottles. Three
hundred (300) bottles of 350 mL were prepared for the experiment. The balance of juice
was pasteurized (95 OC, 15 seconds) with a TETRA ASEPT system and transferred into
three stainless steel tanks of 5500 liters capacity connected together. The juice was mixed
and the bins were filled. This system was used to fill aseptically (4 O C ) the 1 1 85 Kg bins.
The bins were kept at 1 OC . In total, 12 bins were prepared for the experiment. One of
the bins was opened in a stainless steel tank and then the pasteurized juice was transferred
into glass bottles. The bottles were then frozen at -18 O C . This juice was kept at this
temperature before and during shipping by truck. All the tanks used in this process are
washed, sterilized and aseptized. The juice remained more than one month in Mexico
before its arrival at Rougemont. The juice was processed on the 13" of March and
arrived on the 24" of April at A. Lassonde Inc. (Rougemont. Qc.). The storage
experiment began on the 14" of May 1997 (month 1).
3.3 Sample preparatim for analytical analyses
Frozen orange juice samples were immersed in a warm water bath at 45 OC
(Shaker Bath. Orbit LAB-LINE. model no 3540) with continuous shaking in order to
achieve 25 O C (Merin and Shomer. 1984). About 30-35 minutes was necessary with a 100
RPM shaking to reach the desired temperature. All the nine analyses were carried out at
ambient temperature. The aseptically packaged juice in bins was homogenized for 5-7
minutes and was equilibrated at ambient temperature prior to analysis.
3.4 Artal)ticai methods
Ail the analyses were performed with water obtained from a Milli-Q Water
System (Millipore Corporation. Beddford. MA.).
3.4.1 Sedimentation of the pulp
A well mixed orange juice (100 mL) was placed in a 8" graduated centrifbge tube
(100 rnL, Kimble Glass Inc., US). After 0, 1,2 and 3 hours, a visual measurement of the
height of sedimentation of the pulp in mL was performed (Merin and Shomer, 1984).
3.4.2 Cloud measurement
About 2 mL of the supernatant above the pulp sedimentation was taken with a
pasteur pipette and placed in a colorimeter cell. The % transmittance at 650 nm on a
colorirneter (Spectrophotorneter DU 640, Beckman, US) was measured. Water was w d
for the blank (100 % T). Six readings were collected for each sample. to compensate for
the variation of the response lamp. The measure was taken at 0. 1.2 and 3 hours with the
same sample to follow the evolution of the cloud. After the measurement, the liquid was
carehlly transferred back in the graduated sedimentation tube to avoid changes in the
measurement of sedimentation of the pulp. This method was adapted from Carter. 198 1.
3.4.3 Volatiles
The volatile aroma were measured by GC-MS. and the very volatile components
were quantified by headspace coupled with a GC-FID.
3.4.3.1 Gas chromatography coupled with a mass spectrometer (GC-MS)
A. Jnternal ~~d mixture
Phenol-dS (9.2 mg) was accurately weighed into a 1.5 rnL vial with a screw cap
and 1.0 mL of diethyl ether (Nanograde, Mallinckrodt, Kentucky, US) was added and the
vial was capped immediately. The vial was agitated until all the solids were dissolved.
With a clean 10 pL syringe, 10 pL of cyclohexanone-d4 and 10 pL o f butanoLd9 were
added through the septum. The syringe was washed before and after the addition to avoid
cross contamination. The cap was changed (new septum) and the vial was lightly mixed.
The internal standard was stored at 4 "C.
The three internal standards were purchased from CIL, Cambridge Isotope
Laboratory (Andover, MA, US). This internal standard mixture was used to correct for
the extraction efficiency of the ether and to check the variation of the retention times in
the GC-MS (Finnigan MAT software).
Primary srock solution
Diethyl ether (10 rnL) was added to a 100 mL volumetric flask. According to
Table 5. the appropriate amount of high purity standards were added to the volumetric
flask with a clean 10 pL syringe. Diethyl ether (1 mL) was used to wash the standards
into the flask. The 100 rnL flask was capped after each addition. When all the volatile
standards were added. the volume was diluted with diethyl ether to 100 mL and labeled as
standard #4.
Secondary standard sohions
The primary stock solution (standard #4) was diluted as foliows using a 10 mL
volumetric flask:
standard # 1 1.0 mL of the primary stock solution diluted to 10 mL with diethyl ether:
standard #Z 2.0 mL of the primary stock solution diluted to 10 mL with diethyl ether;
standard #3 5.0 mL of the primary stock solution diluted to 10 mL with diethyl ether.
For the analyses on GC-MS, 1.0 rnL of each standard (standard 1 to 4) was
pipetted into an 1.5 mL auto-sampler vial and 10 pL of the intemal standard were added.
The vials were immediately capped and mixed thoroughly on a Vortex for 10 seconds.
The external standards were used to establish a calibration curve (GC responses vs
concentration) to quantify the unknown concentration in the samples. A Finnigan MAT
software quantified automatically each volatile component with associated calibration
curve. As shown in Table 5, the standards were composed of 36 compounds. An
example of the GC-MS chromatogram showing the 36 volatile standards is illustrated in
Figure 6 .
Table 5. Volatile standards for the primary stock solution (100 mL).
Compounds I Density I pL I Std #J I Std #3 I Std#2
Methyl butyrate 99%*
tdmL) 0.898
( I R)(+)Alpha- pinene 98%* Ethyl butyrate 99%* Hexanal 98 %* (+)Sabinene 99%' 3-Carene 90%* Alpha- phellandrene* Beta-myrcene* Heptanal95 %* d-Limonene (G-R Inc. Om.. CA)
?-methy I-butan01 99+9/0* 3-methyl-butanol 99+%* Trans-2-hexenal 95+%* (S)(7)2-Hexanol 99%' Gamma-terpinene 9 5-e0h Octanal99%* LHexanol98+%* 3-Hexen-l -ol 98%. Nonanal95%* 2-4- dimethylstyrene 97 %* Furfiml* Decanal95 %@ (f )LinalooI 97 Oh* 1 -0ctanoP Terpinen-4-01 96?/0* Ethyl-3-hydroxy- hexanoate 98+%*
0.858
0.878
0.834 0.844 0.860 0.852
0.80 1 0.820 0.844
0.815
0.809
0.846
0.8 14
0.849
0.82 1 0.8 14 0.846
.-
4
17.16 6.864
17.56 7.024
16.68 6.672 16.88 6.752 2 1.50 8.600 I 7 .04 6.8 16
17.96 7.184 16.36 6.544 63 3 253.2
16.30 6.520
16.18 6,472
16.92 6.768
16.28 6.5 12
16.98 6.792
1 6.42 6.568 16.28 6.5 12 16.92 6.768
L
1
0.827 0.910
1.160 0.830 0,870
4
4
4
(IJdmL) 35.92
34.32
35.12
33.36
I
4 4
4 4 4
( W m L) ( ~ g J m L) 1 7.96 7.184
4 I 33.76
1
33.08 3624
46.40 33 2 0 34.80
0.827 0.933
0-974
5 4
4 4
150
4
4
4
4
4
4 4 4
I
16.54 18.12
23 20 16.60 17-40
33.08 37.32
38.96
4 4
4
43 .OO 34.08
35.92 32.72 1266
32.60
32.36
33.84
32.56
3 3.96
32.84 32.56 33 -84
16.54 18.66
19.48
Figure 6. GC-MS chromatogam of the 36 volatile standards used in the analysis of
orange juice samples.
*From Aldrich Chemical Company Inc, Milwaukee, WI, US; unless otherwise specified. **Solid or viscous sample were weighed. The order in the Table 5 is the order of the retention time of the elution of the compoundr fiont the G{ column (DB W.L't;3.
18.66
19.18
7.464
7.672
7.1 12
10.00
10.00
14.72
1.98
2.00
2.00
37.32
38.36
3.732
3.836
3.556
5 .OO
5 .OO
7.360
0.99
1-00
I .OO
Alpha-terpineoi 98%* (RK-)Cawone 98%. Ciaal (cis and trans) 95 %' (S)(-)perilladchyde 92a/08 24decadicnal 85%. Valcnccne ( A ~ S Organics. N-J.US) 4-Viny 1-guaiacol 97'10 ( h c a s t t r Sy nthcsis fnc.N-HUS) HMF 99% (Sigma Chcrnicds Co. MO. US) Vanillin (RP chirnic .France1
0.930
0.959
4
4
3 5.56
50.00
50.00
0-889
1 -00
1 .OO
0.920
liq*
solida*
solid* *
17.78
25.00
25.00
4
5
5
8
-
-
73.60
9.90
10.0
10.0
36.80
4.95
5.00
5.00
C. a m ~ l e preparat~on
NaCl (5.0 g) (U.S.P., BDH Inc., Toronto, Canada) was added to a screw cap tube
(1 5 mL), followed by well mixed orange juice (10.0 mL). A 10 pL of the internal
standard solution was added into the tube before capping. The tube was shaken until the
salt is totally suspended. The juice was extracted by shaking vigorously for 60 seconds in
diethyl ether (1.0 mL) and centrifuged at 3410 x g for 2 minutes (Fisher Centrific
Centrifuge. Model 225, rotor 04-978-54). The tube was shaken another 60 seconds and
centrifuged again for 5 minutes. The supernatant (200 pL) was transferred to a 1.5 rnL
auto-sampler vial with a reduced volume insert. The vial was capped immediately.
Depending on the compound. retention time (GC) and specific ions (MS) were used ro
identify the peaks.
0. * .
c o n d ~ t ~ o n ~
A Varian 3400 GC coupled with a Finnigan Mat Incos 50 MS (scan range: 35-
250 m u . scan rate: 0.43 scans/sec) was used for the GC-MS analysis. The GC column
used was a 0.25 mm X 30 m with 0.25 pm DBWAX film. The flow rate was 1.0 mL/min
for a split ratio of 50: 1. The injection volume was 1 .O pL and the carrier gas was helium.
The temperature of the injector was 250 O C and 225 "C for the transfer line. The column
initial temperature was 40 O C and was increased to 90 O C at a rate of 3 '&in. After 1
rnin the temperature was M e r increased to 120 O C at a rate of 10 "%in and kept at 120
OC for 2 min. Then. the temperature was increased to 200 OC at a rate of 20 O h i n and
kept at 200 OC for 5 min. The total run time was 30 minutes.
3.4.3.2 Gas chromatography with a headspace injector (Headspace-GC-FID)
A* lnternal standard mixture
A 100 pL of 1-propanol(99.5+%, ACS reagent Aldrich Chemical Company Inc,
Milwaukee, WI, US) was added to a 1.5 screw cap vial with 900 pL of water. The final
concentration of 1 -propano1 was 160 ppm when 10 pL were added to 5 mL of juice.
8.
Water (5.0 mL) was added to a volumetric flask (25 mL) and appropriate amounts
of the following compounds were added and diluted to volume with water:
a. 8000 pL of ethanol (anhydrous Accusolv. Anachemia. N-Y. US)
b. 400 pL of methanol (optima, Fisher Scientific. Nepean, Ont.. CA)
c. 80 pL of acetaldehyde ( 99%. Aldrich Chemical Company Inc. Milwaukee. US)
d. 3 pL of ethyl acetate (HPLC. Aldrich Chemical Company Inc. Milwaukee. US)
When the juice samples ( 5 mL) spiked with 25 pL (#I) and 50 pL (#2) of the
primary stock solution, the final concentration obtained (pL/mL) were as follows:
Primary stock solution
Acetaldehyde
Methanol
Ethanol
Ethyl acetate
C. Sample pre~antion
A well mixed juice (5.0 mL) was added to each of three vials (8 mL) followed by
10 pL of the internal standard solution (I-propanol). The first vial was capped. A 25 pL
and 50 FL of the primary stock standard solution were added respectively to the second
and the third vials and then they were capped. A calibration curve was drawn (area-
response factor vs concentration of the standard) with the three values and the unknown
concentrations (no spike) is the point which crosses the x-axis.
D. + *
eads~ace-GC condttms
A Hewlett Packard 5890 GC with a HP 7694 headspace injector (oven at 65 OC,
loop at 200 "C and 15 min for vial equilibration) was used for the Headspace-GC analysis.
The GC column used was a 0.25 mm X 30 m with 0.25 pm DBWAX film. The GC
column flow rate was 1.0 mL/min for a split ratio of 30:l. The injection volume
(headspace gas) was 1.0 rnL and the carrier gas was helium (nitrogen for the make-up
gas). The temperature of the injector was 250 'C and 150 OC for the m s f e r line. The
initial column temperature was 20 O C for 5 min and was increased to 50 O C at a rate of 5
'Ginin. Immediately, the temperature was further increased to 220 "C at a rate of 20
'Omin and kept at 220 'C for 5 rnin. The total run time was 35 minutes.
3.4.4 Viscosity and density measurements
In a 50 mL centrifuge tube, 40 mL of well mixed juice was centrifuged at 1925 x g
for 10 min (GS-6 Centrifuge, Beckman Instruments Inc., Palo Mto, CA). For the
viscosity measurement, 10.0 mL of the supernatant was transferred to a viscosirneter
(100, 484-485-4861, Cannon-Fenske routine viscosimeter, Cannon Instrument Co., PA.
US). The viscosimeter had been conditionned in a thennostated bath at 40 OC. The time
that it took the juice to pass between two ftved points was noted.
Viscosity = constant of the viscosimeter at 40 OC (cSt/s) X Time (s).
For the density, a pycnometer (25.0 mL) was weighed accurately (1) and 25.0 rnL
of the centrifuged juice was placed in it. The flask and the juice were re-weighed
accurately (2). The weight difference divided by the volume gives the density of the
centrifuged juice, according the following equation.
Density = Weight (2)(g) - Weight (l)(g) / Volume of pycnometer (25.0 mL)
3.4.5 Ascorbic acid
A 1% ( W m starch (2.5 mL) (Starch. soluble. BDH. Analar) solution was added
to a well mixed juice (20.0 rnL) in a 100 mL flask. From a burette, Iodine solution
(0.02N) (Iodine. N/50 solution certified 0,0205-0.0195 N. Fisher Scientific) was added in
a slow stream while gently swirling the flask. At the point where the solution showed a
strong biue-purple coloration that fades. the iodine was added one drop at a time. The
endpoint was indicated when one drop produced a blue-purple colour that persisted for at
least 30 seconds. In orange juice, the endpoint produced a light green coloration.
The following equation was used to determine the mg of ascorbic acid per 100 mL of
juice:
ascorbic acid concentration (mgA00rnL) = rnL of 0.02 N iodine X 8.8 1
To confirm the validity of the method, a standard solution of 30 mg of ascorbic
acid/100 mL of water was prepared (L-ascorbic acid, Sigma Chemical Co, MO, US) and
titrated with the solution of iodine.
Note: O.OZN Iodine should nor be storedfor more than 2 months.
3.4.6 Colour measurement
A juice sample (50.0 mL) was transferred to a specimen cell (1.5" Round?
LYSACK SALES, Associated LTD., Etobicoke, Ont., CA). A reading on a Hunter LAB
instrument (Hunter Associated Laboratory, Inc.. Virginia, US) was taken after calibration
of the instrument with black and white plates (HunterLAB color standard). The
HunterLAB was selected for a ON5 mode, 0.5 area view and 1.75 port size. The L. A and
B values were noted. Three readings were carried out per sample.
3.4.7 Sugars and organic acids
3.4.7.1 High performance liquid chromatography (HPLC)
A. tandard suwr solmng
D-Sucrose (500.0 mg), D-glucose (250 mg) and of D-fructose (250 mg) (AnaIar.
BDH Inc.. Toronto) were added to a 100 mL volumetric flask. The sugars were dissolved
and the flask was filled to the mark with water.
B.
Citric acid (150.0 mg) (Analar, BDH Inc., Toronto) and L(-)malic acid (50 mg)
(Sigma Chemical Co. MO. US) were added to 100 mL volumetric flask. The organic
acids were dissolved and the flask were filled to the mark with water.
C.
Standard calibration curves were drawn for each sugar and organic acid. A Waten
s o b a r e calculated calibration c w e s with accuracy limit for each quantified
components.
D.
Well mixed juice (10.0 mL) was centrihged for 5 minutes at 3410 x g (Fisher
Centrific Centrifuge, Model 225. rotor 04-978-54). The supernatant (5.0 mL) was
pipetted into a 100 mL volumetric flask and diluted with water. A SepPak C 18 (Sep-Pak
Plus cartridges. Waters Corporation, Milford. MA.) was conditioned with 5 mL of
methanol followed by 10 mL of water. After the conditioning. 5 mL of the diluted
supernatant juice was passed through the SepPak Cl8. The first 2 mL of the juice was
rejected to avoid dilution of sugars and organic acids by water. The recovered juice was
filtered through a 0.45 pm filter (Millex-HVl3. Millipore. Millipore Corporation.
Bedford. MA ) and placed in a sample vial for HPLC analysis. The standards of sugars
and organic acids were processed in the same manner as the juice samples to measure the
recovery. The percent of recovery for the sucrose, glucose and fructose were >99%.
>99% and 98% respectively. And for the citric and malic acids. the percent of recovery
were 99% and >97% respectively.
E. 1 C cond~t~on~ .
A Waten HPLC (Milford. MA, US) equipped with a Model 510 pump and an
autosampler (717 plus) was used for the KPLC analysis. The mobile phase was a
degassed solution of sulfuric acid 0.01 N (290 pL of H2SOdliter of water) (Analar, BDH
Inc.. Toronto) filtered through a 0.45 pm filter. The HPLC column used was an
Interaction polymetric Ion 300 (Inter Action Chromatography) with a BioRAD cation H+
as a precolumn. The HPLC column flow rate was 0.4 W m i n and the injection volume
was 1 5.0 pL. Solute elution was monitored using a fixed wavelength at 2 14 nm (Waters
484 tunable Absorbance) for the organic acids, and a differential rehctometer (Water
410) for the sugars. The total run time was 20 minutes. This method was adapted from
Doyon er a/.. 199 1.
Well mixed juice (1.0 mL) was placed in the cell of a digital rehctometer
(RFMM 80.0-95 % sugar, Bellingham & Stanley Limited. England). The reading of the
per cent of sugPr and the temperature was noted. The instrument was adjusted to zero with
water before each measurement.
3.4.7.3. Titratable acidity
Well mixed juice (35.0 mL) was added to a 250 mL beaker. The juice was diluted
with about 50 rnL of water. The juice was then titrated with NaOH (1.0 N) (Sodium
hydroxide. 1.000 * 0.002 N. VWR Scientific) using a pH meter (Coming pHmeter 140).
The titration with 1 .O N sodium hydroxide was followed until the equivalence point or pH
of 8.1 was reached. The rnL of NaOH used was noted. Because of the difficulty in
obtaining a precise endpoint (pH 8.2) with the pH meter method. a solution of
phenolphthalein (ACS, Fisher) I % H in ethanol as an indicator was used to verify the
endpoint. With the small variability in the density measurements, a standard value of
1.04 was used in all calculations of the titratable acidity to avoid a source of variation.
This method was adapted from Carter (198 1). Percent of acid was calculated
using the following equation:
% acid = {(N NaOH X volume (mL))/1000) X {100g/(mL juice X density)) X G.E.W.
(g/ 1 OOmL)
N = Normaliiy of NaOHsolution
G. E. W. = 6404for orange juice (molecular weight oj'the citric acid}
Measured density or a standard density of I . 04 was used for the calculation.
3.4.8 Sensory evaluation
The sensory evaluation was performed in parallel with the chemicd analysis by an
expert panel (12 persons) under the supervision of a sensory evaluation expert. A
descriptive quantitative profile with continuous scale limited to 0 to 130 was used. A
sample questionnaire is presented in the Appendix A.
The tests were performed on Tuesdays and Thursdays of the same week that the
samples were analyzed for the other nine parameters. Each panel member received a 50
rnL of juice in a glass (1 bottle for 4 persons). The samples were kept at 1 O C and at -18 0 C depending on the initial condition of processing. Twenty four hours (24 hrs) before
analysis, the samples were stocked in a refrigerator at 8.5 O C to equilibrate the
temperature to 10 O C for the sensory evaluation. The bin samples were frozen, if
necessary, until Monday and Wednesday to be thawed at the same time of the froten
samples. The frozen samples were immersed in a thennostated water bath at 40 'C for 30
minutes (Waterbath 183, Precision Scientitic Inc.) before the test.
3.4.9 Statistical treatment
Statistical treatments were performed on the data obtained under each condition
for months 1 to 8. A One-way analysis of variance (ANOVA) was not possible because
this method assumes populations with equal standard deviations and equal number of
samples. In most cases. in this study, a difference was detected by the Bartlen's test for
homogeneity of variances and some data were rejected according the Chauvenet's
criterion for rejecting a reading (Holman and Gajda. 1984). The Kruskal-Wallis non-
parametric ANOVA test was performed with the Instat software to establish significantly
different measurements. Statistical treatment (ANOVA) were performed on all data.
However. only relevant data are presented. Because of the variations in data among the
months of analysis, a t-test was performed on each parameter to compare the means of the
first and the last month of analysis. The critical value of the 1 t I for P value of 0.0 1 (99%
confidence interval) for a 5 degrees of freedom was 4.03 (Miller and Miller. 1988).
Appendix B shows different examples of the statistical calculations.
CHAPTER 4
RESULTS AND DISCUSSION
4.1 Effect of storage condition on individual parameters
Parameters monitored during storage including density, cloud, sedimentation of
the pulp, sugars, organic acids. ascorbic acid, viscosity, colour. volatiles and sensory
analysis. are presented and discussed separately. Parameters which did not show
significant changes during the storage period are discussed first.
4.1.1 Density
Measurements of the den sity of the orange juice were performed monthly as
described in the Materials and Methods section. The results of the density measurements
are presented in Table 6.
A t-test analysis of the density data showed significant differences for juices
stored under conditions A and C ( I t I equaled 5.1 9 (A), 2.63 (B) and 6.75 (C)). However.
when the data were analyzed for a linear trend. no significant constant increase or
decrease were found. The data points were varied within one standard deviation of the
mean. Close to zero slopes (Y z b) were calculated and poor regression coeffcients
indicated no trend in the data among the eight months of analysis (see Table 7).
Table 6. Density value' of orange juice over an eight month storage period.
Condition A Condition B Condition C
Mean (g/mL) Standard Mean (g/mL) Standard Mean (g/mL) Standard Month deviation deviation deviation
I 1 .090* 0.002 1.105 0.005 1 SO4 0.002 2 1.1 14 0.0 14 1.120 0.0 13 1.107** 0.002 3 1.1 12 0.016 1.129** 0.0 19 1.121 0.02 1 4 I. 105 0.003 1.107 0.002 1.104** 0.00 1 5 1.105 0.005 1.107'" 0.003 1.104 0.005 6 1 SO0 0.007 1.097 0.006 1.103** 0.002 7 1.1 14 0.022 1.1 13 0.034 1.122 0.024 8 1.106 0.005 1.1 lo** 0.002 1.1 10 0.00 1
*Calculated with only three values. **Calculated with only five values. Chauvenet's criterion. I Average of six measurements. unless otherwise specified. Condition A: Not pasteurized orange juice stored at -1 8°C in glass bottles. Condition B: Pasteurized orange juice stored at 48*C in glass bottles. Condition C: Pasteurized orange juice stored at +l°C in polyethylene bag.
Table 7. Equation of the regression curves and coefficients of regression (R') of the
density of orange juice over an eight month storage period.
Condition Equation R' A Y = 0.0009 X + 1.1013 0.08320 B Y =-0.001 X + 1.1161 0.07663 C Y = 0.0008 X + 1,1058 0,06 195
The data tabulated in Table 6 did not present changes over time and showed no
difference at the beginning of the study between the juice stored under the three
conditions of storage (A, B and C). No significant increase or decrease in juice density
was found among the dam Changes of density due to different processes or storage times
have not been reported also in the literature.
4.1.2 Cloud and sedimentation of the pulp
Sedimentation of the pulp and cloud were measured by visual observation and
spectrophotometric readings. These two parameters are important for the appearance and
mouth feel of the orange juice. The results are presented in Tables 8 and 9. Orange juice
stored under conditions B and C did not show any sedimentation or loss of cloud.
Table 8. Sedimentation of the (measured at 3 time intervals) of orange juice
over an right month storage period.
Condition A I hour 2 hours 3 hours
Mean (mL) Standard Mean (mL) Standard Mean (mL) Standard Month deviation deviation deviation
1 8 3 9 0.7 62.01 0.0 55' I .4 2 66.7 9 -3 58.3 6.1 55.2 3 -8 3 58.6** 0.5 55.8 I .3 54.0 1.1 4 62.0 4.7 56.2 2.8 54.2 1.6 5 69.0 4.7 60.8 2.8 57.2 1.9 6 58.8 2.5 54.8 1.7 54.0 1.4 7 56.3 1.6 55.2 1 .Z 54.4* * 1.1 8 S 8 . P 0.5 55 .O 1.4 54.7 1 .4
*Calculated with only three values. **Calculated with only five values. Chauvenet's criterion. L Average of six measurements, unless otherwise specified.
Table 9. Cloud stability1 (measured at 3 time intervals) of orange juice over an
eight month storage period.
Condition A 1 hour 2 hours 3 hours
Mean (%T) Standard Mean (%T) Standard Mean (%T) Standard Month deviation deviation deviation
1 44.32* 1.84 55.76' 3.63 6 1.67* 2 -94 2 54.89 3 .OO 54.93 1.08 59.39 2.6 1 3 44.53 4.32 50.19 4. 14 52.84 4.25 4 48 2 2 0.70 55.12 2.70 58.29 2.93 5 56.78 7.95 61.01 3.85 63.24 1.54 6 56.2 1 5.57 55.39 3 .03 58.86 3.8 1 7 53 -63 1.51 58.78 1.22 60.49* * 0.75 8 57.07 - 3 r 61 57-89" 0.3 1 60-64 1 .58
*Calculated with only three values. **Calculated with only five values. Chauvenet's criterion. ' Average of six measurements. unless otherwise specified.
The data shown in Tables 8 and 9 are for orange juice samples stored under
condition A. Despite the variations in the data, a decreasing trend during the three hours
of measurements was observed. Moreover. after 3 hours. the monthly measurements o f
the unpasteurized orange juice showed a sedimentation of almost 50 %. A similar
phenomenon was observed for the measurement of the cloud. After three h o r n of
observation. the percent of transmittance increased by about 10 %. These observations
may be important to predict the stability of orange juice after thawing and processing.
Sedimentation of the pulp and cloudiness are important factors for mouth feel and
appearance. The presence of pulp modifies not only the intensity but also the balance of
the overall aroma.
4.1.3 Sugars
Brix measurement
Sugars comprise 70-85 % of the total soluble solid material in orange juice. Total
sugar was measured as 'Brix. The OBrix measurement was performed with a digital
refiactometer. The results of the O~rix measurements are presented in the Table 10.
Table 10. 0 Brix value' of orange juice over an eight month storage period.
Condition A Condition B Condition C
Mean (%) Standard Mean ( O h ) Standard Mean (%) Standard Month deviation deviation deviation
1 11.5 0.3 12.0 0.1 11.4 0.1 2 11.1 0.3 11.5 0.2 11.8 0.3 3 11.1 0.4 1 1.2** 0.3 10.6 0.5 4 1 1.5** 0.2 11.7 0.2 1 1.9** 0.1 5 11.1 0.5 11.6** 0.1 10.8 0.5 6 10.5 0.5 1 f .O** 0.4 11.4 0.3 7 11.2 0.3 1 I.8** 0.3 11.8 0.2 8 1 1.9** 0.1 12.2 0.1 12.1S* 0.0
**Calculated with only five values. Chauvenet's criterion. I Average of six measurements. unless otherwise specified.
Table 10 shows the results obtained over an eight month period. For the OBrix
value, no change in concentration with time of storage or method of processing was
observed. Linear equations of the regression curve and coefficients of regression were
calculated from the data (see Table 1 1).
Table 1 1. Equation of the regression curves and coefficients of regression (R') of the
'Brix of orange juice over an eight month storage period.
Condition Equation RL A Y = 0.0131 X + 11.2 0.006 1 1 B Y = 0.2620 X + 11.5 0.0263 1 C Y = 0.0738 X + 11.1 0.1 1355
A t-test analysis of the "Brix data showed significant difference only for the juice
stored under condition C ( 1 t 1 equaled 2.1 1 (A). 2.08 (B) and 13.58 (C)). A standard
deviation of zero for the last month can explain this result. However. when the data were
analyzed for a linear trend. no significant constant increase or decrease was found for any
of the stonge conditions. There was no linear correlation between the time of stonge and
the " ~ r i x value. The variability in the data may be due to random errors associated with
the refkctometer used to measure the 'Brix.
The literature also indicates no change in "Brix values of orange juice during
storage.
HPLC measurements
The main sugars found in orange juice (D-sucrose. D - ~ ~ U C O S ~ and D-fructose) were
measured by HPLC according to procedure described in the Materials and Methods
section. The results of the HPLC measurements for each sugar (D-sucrose, D - ~ ~ U C O S ~ and
D-fructose) are presented in Tables 12, 13 and 14. An example of an HPLC
chromatogram showing the three sugars is illustrated in Figure 7.
Table 12. Sucrose concentration' of orange juice over an eight month storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (g/lOOrnL) deviation (g/100mL) deviation (g/100mL) deviation
1 4.89 0.09 5.15 0.07 4.90 0.07 7 4.92 0.04 5.17 0.07 5.15 0.08 3 5.17 0.14 5.37 0.12 5.09 0.14 4 5.2 1 0.08 5.32 0.08 5.36 0.06 5 4.71 ** 0.07 4.85 0.12 4-50 0.12 6 4.96*+ 0.02 5.06 0.14 5.08 0.1 1 7 4.87 0.08 5.02 0.06 4.9 1 ** 0.03 8 4.90 0.02 5.10 0 .04 4.9 1 0.05
**Calculated with only five values, Chauvenet's criterion. I Average of six measurements. unless otherwise specified.
Table 13. Glucose concentration' of orange juice over an eight month storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Mon tb (g/100mL) deviation (gf 100mL) deviation (g/100rnL) deviation
1 1.97 0.0 1 1.99 0.05 1.90 0.02 3 - 1.85 0.04 1.89 . 0.02 1 -90 0.04 3 1.98 0.04 2.00 0.05 1.91 0.06 4 1.97 0.03 2.00 0.03 2.0SS* 0.0 1 5 1.82 0.06 1.84 0.03 1.77 0.04 6 1.91 0.02 1.93 0.07 1.99 0.07 7 1.95 0.04 1 -97 0.02 2.0 1 ** 0.00 8 1.88** 0.0 1 1.93 0.02 1.97 0.03
**Calculated with only five values, Chauvenet's criterion. I Average of six measurements, unless otherwise specified.
Table 14. Fructose concentration1 of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (g/lOOmL) deviation (g1100mL) deviation (gI100mL) deviation
1 2.34 0.04 2.40 0.33 2.3 1 0.02 2 2.25 0.04 2.29** 0.0 1 2.34 0.07 3 2.36 0.03 2.40 0.04 2.30 0.04 4 2.33** 0.03 2.3 4 0.03 -. 7 40 0.05 5 2.27 0.02 2.29 0.04 - 7 .-- 77** 0.04 6 2.3 8 0.03 2.40 0.09 2.52** 0.04 7 2.42 0.03 2.47* * 0.03 2.50 0.04 8 2.3 1 0.02 2.34** 0.0 1 2.3 8 0.02
"Calculated with only five values, Chauvenet's criterion. I Average of six measurements, unless othenvise specified.
Figure 7. HPLC chromatogram of the three sugars in orange juice processed by condition
A at the last month (8) of storage (D-sucrose tR = 1 2.8 76 min. o-glucose rR = 1 5.550 min
and D-fructose tR = 17.1 1 7 min).
Data fiom the HPLC analysis of sugars (Tables 12- 14) showed some variability
over the eight months of analysis. The data appeared to increase or decrease as a block
regardless of the storage condition. A t-test was performed (see Table 15) with the data
tabulated in Tables 19 and 20.
Table IS. Results obtained from the t-test for each sugar contained in orange juice.
Condition I t 1 value
SUCROSE A 1.68 B 4.65 C 6.02
GLUCOSE A 14.4 1 B 2.8 1 C 4.93
FRUCTOSE
Significant differences between the means of D-sucrose concentration was found
for the juices stored under conditions B and C and similarly for D-glucose concentration
in juices stored under conditions A and C. However. when the data were analyzed for a
linear trend. no significant constant increase or decrease was found in all cases. Close to
zero slopes (Y z b) and poor regression coefficients were calculated for the data obtained
for the three storage conditions (see Table 16).
Table 16. Equation of the regression curves and coefficients
three sugars of orange juice over an eight month storage period.
of regression (R') of the
Condition Equation R SUCROSE
A Y = -0.0156 X + 5.02 0.05463 B Y = -0.0298 X + 5.26 0.19376 C Y = -0.0240 X + 5.10 0.0550 1
GLUCOSE A Y = -0.006 X + 1.94 0.05499 B Y = -0.005 X + 1.96 0.03903 C Y = 0.012 X + 1.88 0.1 1157
FRUCTOSE A Y = 0.008 X + 2.30 0.1 1107 B Y = 0.005 X + 2.34 0.04082 C Y = 0.021 X + 2.28 0.25797
The conclusion was the same for the three sugars. No linear trend could be
observed (almost zero slope). and the variation in the data could not be correlated linearly
with the time (R' very low). Therefore. there was no linear association between time of
storage and sugar concentrations.
4.1.4 Organic acids
Titratable acidity measurements
The measurement of the acidity of the orange juice is ofien expressed as titratable
acidity. This method is rapid and gives a general percentage of organic acids in juice.
Titration with NaOH was used to perform this measurement. The results of titratable
acidity measurements are presented the Table 17.
A t-test analysis of the titratable acidity data showed significant difference only
for juice stored under condition A ( I t 1 equaled 7.35 (A), 3.34 (B) and 2.75 (C)).
However, when the data were analyzed for a linear trend, no significant constant increase
or decrease was found in each condition.
Table 17. Titratable acidity measurement1 of orange juice over an eight month
storage period.
Condition A Condition B Condition C
Mean (%) Standard Mean (%) Standard Mean (%) Standard Month deviation deviation deviation
1 0.78 0.0 1 0.78 0.0 1 0.74 0.0 1 2 0.76 0.02 0.78 0.02 0.77** 0.0 1 3 0.77** 0.00 0.77" * 0.00 0.74" * 0.00 4 0.75 0.0 1 0.76 0.0 1 0.77 0.0 1 5 0.76 0.0 1 0.77 0.0 1 0.73 0.0 1 6 0.74** 0.0 1 0.76 0.0 1 0.77" * 0.00 7 0.75 0.0 t 0.75** 0.00 0.77 0.0 1 8 0.75 0.0 1 0.76 0.0 1 0.76 0.0 t
**Calculated with only five values. Chauvenet's criterion. 1 Average of six measurements. unless othenvise specified.
Table 18. Equation of the regression curves and coefficients of regression (R') of the
titratable acidity measurement of orange juice over an eight month storage period.
Condition Equation R A A Y = -0.004 X + 0.78 0,59834 B Y = -0.004 X + 0.78 0.72637 C Y = 0.002 X + 0.75 0.10812
Linear equations of the regression curve and coefficients of regression were
calculated from the data (see Table I I). The slopes of the regression curves were close to
zero and no real trend was observed. Some variability appeared in the three conditions
over the eight months. Regression coefficients were low and indicated that no linear
trend can be observed.
Table 17 is the data obtained fiom the colorimetric titration. The end point was
more stable with the method using phenolphthalein than the method with a fixed pH end
point (pH = 8.2).
The literature has also indicated that there is no change in titratable acidity of juice
during storage period.
4.1.4.2 HPLC measurements
Organic acids in orange juice consist primarily of citric and malic acids. These
two acids were quantified by HPLC measurement. The results of the HPLC analysis are
presented in Tables 19 and 20. An example of a HPLC chromatogram including the two
organic acids is illustrated in Figure 8.
Table 19. Citric acid concentration' of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (mg/lOOrnL) deviation (mg/lOOmL) deviation (mg/lOOrnL) deviation
1 958.70 27.83 958.85 97.17 965.20 1 3 -69 3 - 894.48 19.76 922.63" 5 .07 932.16 25.06 3 989.25 * * 3.88 I 005.52** 6.12 957.36 10.85 4 1055.33 12-86 1074.60 12-79 1089.7 1 1 1.29 5 947.78 33.05 971.57 20.54 9 18.63 18.95 6 998.33 9.3 3 1012.80 33.69 1033.74** 15.62 7 962.42 18.13 980.89 1 3.23 980.77 7.66 8 979.57 8.70 1006.95 7.6 1 99 1.58 4.93
"Calculated with only five values, Chauvenet's criterion. Average of six measurements. unless otherwise specified.
Table 20. Malic acid concentration' of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (mg/lOOmL) deviation (mg/lOOmL) deviation (mgI100mL) deviation
1 77.96* 2.27 162.23 4.85 158.55 2.95 2 157.40 4.4 1 160.80** 1.39 163.88 6.36 3 160.85 3 -03 164.3 7 2.09 159.42 2.07 4 165.67 1.89 170.16 5.37 1 72.72* * 0.95 5 1 65.69 6.99 170.55 3.32 158.18 7.05 6 167.78 3.66 171 -66 7.40 177.35** 3.30 7 140.10 4.55 164.6 1 2.83 164:94 2.65 8 161.78 1.85 165.84 1.33 156.02 11.14
*Calculated with only two values. **Calculated with only five values. Chauvenet's criterion. I Average of six measurements. unless otherwise specified.
Figure 8. HPLC chromatogram of the two organic acids in orange juice processed by
condition A at the last month (8) of the norage period (citric acid tR = 14.167 min and
malic acid tR = 17.267 min).
Data from the HPLC analysis of organic acids (Tables 19 and 20) showed some
variability over the eight months of analysis. The data appeared to increase or decrease as
a block regardless of the storage condition. A t-test was performed (see Table 21) with
the data tabulated in Tables 19 and 20. Significant differences between the means of
citric acid concentration was only found for the juice stored under condition C. However,
when the data were analyzed for a linear trend, no significant constant increase or
decrease was found (see Table 22).
Table 2 1.
juice.
Results obtained from the t-test for each organic acid contained in orange
Condition I t 1 value
CITRIC ACID
MALIC ACID
Table 22. Equation of the regression c w e s and coefficients of regression (R') of the
two organic acids of orange juice over an eight month storage period.
Condition Equation R A CITRIC ACID
A Y = 4.8271 X + 951.5 1 0.06593 B Y = 6.5096 X + 962.43 0.12610 C Y = 5.7830 X + 957.62 0.0646 I
MALIC ACID
The conclusion was the same for the two organic acids. No linear trend could be
observed (almost zero slope), and the variability in the data could not be correlated
linearly with the storage time (R' very low). Therefore, there was no linear association
between time and organic acids concentrations.
4.1.5 Ascorbic acid
The ascorbic acid was measured by titration of juice samples with a solution of
iodine (0.02 N) according to the experimental section. The results of the measurements of
the concentration of ascorbic acid are presented in Table 23 and illustrated in Figure 9.
Table 23. Ascorbic acid concentration of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (mg/lOOmL) deviation (mgA00mL) deviation (mg/lOOmL) deviation
1 54.77 0.23 54.9 1 0.23 50.95 0 -23 7 - 52.94* * 0.20 52.05 1.35 50.44 0.72 3 52.42 0.56 51.91 0.43 49.04 0.45 4 53.08 0.46 52.42 0.28 48.82 0.94 5 52.79 0.30 52.41 ** 0.00 49.0 1 0 .27 6 53.37 0.5 1 5 1.83 0.87 48.71 ** 0.50 7 52.49 0.33 52.20 0.82 50.00 1.03 8 53.59 0.53 52.49 0.33 48.16 0.66
**Calculated with only five values. Chauvenet's criterion. I Average of six measurements, unless otherwise specified.
Table 24. Equation of the regression curves and coefficients of regression (R') of the
ascorbic acid concentration of orange juice over an eight month storage period.
Condition Equation R; A Y = -0,095X + 53-6 1 0.0939 1 B Y =-0.19SX + 53.41 023295 C* Y = -0,373X + 50-85 0.80 167
excluding data for the seventh month
Figure 9. Ascorbic acid concentration of orange juice over an eight month storage period
56
0 1 2 3 4 5 6 7 8 9
Month
Linear equations for the regression curves and coefficients of regression were
calculated from the data (see Table 24). A plot of the concentration of ascorbic acid as a
hction of time (months) is shown in Figure 9. Inspection of Table 24 and Figure 9
indicates that if the data for the first month is ignored, no trend can be observed in
ascorbic acid concentration for juices stored under conditions A and B. Although the
efficiency of the measurement of ascorbic acid concentration by titration is about 99 %
(measured with a standard of ascorbic acid). However, the titration method is associated
with difficulty of observing the end point in the orange juice. In fact, the juice becomes
lightly greenish. The higher values of the first month were probably due to experimental
enon of determining correctly the end point.
A t-test was performed with the data tabulated in Table 23. For juices stored
under conditions A and B, the first month was neglected. The comparison of the two
means showed a significant difference only for juice stored under condition C ( / t 1 equaled 2.78 (A), 0.78 (B) and 9.76 (C)). The data From juice stored under the third
condition (C) was also analyzed by a non-parametric ANOVA. Significant changes
(P<0.0001) were also found between different months of storage under condition C. The
Durn's test indicated that the first month data was significantly different from the fourth.
sixth and eighth months. Similarly, the second and seventh months data were also
significantly different. Regression equations were calculated using the data from all the
months (Y = -0.1796X + 50.07 and R squared equaled 0.24431) and also when the
seventh month was excluded (Y = -0.373 1X + 50.8499 and the coefficient of regression
was 0.801 67: Table 24). The Figure 9 shows the data without the seventh month for juice
stored under condition C. The slope presented a decreasing trend and the coefficient of
regression showed a negative relation between the time and the concentration of ascorbic
acid. indicating that the level of ascorbic acid present in the juice changes (decreases)
during the storage period. These changes may be due to the degradation of ascorbic acid.
primarily by the oxygen contained in the juice and by anaerobic non-enymatic reactions.
in the polyethylene bag stored at +1 OC. The ascorbic acid concentration dropped by 5 %
during the eight months of storage. The low temperature of storage (+l°C) minimized the
loss of ascorbic acid. A retention of 95 % of the original ascorbic acid concentration
during the storage period is sufficient to keep the nutritive quality of the juice.
The second month of each storage condition was examined to determine the
presence of any initial differences between the three conditions before the study period.
The D m ' s test indicated significant differences only between conditions A and C (PC
0.0 1). The initial significant difference between juice stored under condition A versus the
condition C is probably due to the degradation of ascorbic acid by oxygen contained in
the juice in the polyethylene bag at the beginning of the experiment. The decrease of
ascorbic acid probably occurred in the first week after the packaging. In the frozen juices
(conditions A and B), the reaction with the oxygen and the ascorbic acid molecule was
very slow. Pasteurization can also have a negative initial impact on the concentration of
ascorbic acid.
4.1.6 Viscosity
The viscosity measurements were performed on the orange juice on a monthly
basis during the eight months of storage. A 10.0 mL of centrifuged juice was placed in a
viscosimeter according to Chapter 4. The results were expressed in centi-Stoke (cSt =
mrn2/s). The results of the viscosity measurements are presented in Table 25. and
illustrated in Figure 10.
Unfortunately, the juice was not handled properly during the analysis of the first.
sixth and the seventh month samples. As a result the samples were refrozen and analyzed
a few months later. Consequently, the data obtained for these months were affected by re-
freezing and the results were not reliable (see Table 25). Data from these three
measurements (shown between brackets) were removed From the analysis and not
included on Figure 10 in the subsequent equations or calculations.
Table 25. Viscosity measurement' of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean (eSt) Standard Mean (cSt) Standard Mean (cSt) Standard Month deviation deviation deviation
1 (1.1777*)~ 0.0047 (1.6407)' 0.1 101 (1.5 1 57)L 0.0575 2 0.93 19 0.0077 1.3 138 0.0 106 1 -234 1 0.0053 3 0.950 1 0.0 1 52 1.3336 0.0095 1.1876 0.0 157 4 0.9449 0.01 10 1.3 180 0.0082 1.2 157** 0.009 1 5 0.9367 0.0050 1.3141 0.0 100 1.1700 0.0048 6 (0.9094**)' 0.0004 (1.1763)' 0.0 158 (1.0882)' 0.0 192 7 (0.8969**)' 0.0068 (1.2096**)' 0.01 33 (1.1 9 8 ) 0.0193 8 0.9458 0.0092 1.3 109 0.0060 1 .I09 1 * * 0.0200
*Calculated with only two values. **Calculated with only five values, Chauvenet's criterion. I Average of six measurements, unless othenvise specified. ' Data unreliable due to improper manipulation.
A t-test analysis of the viscosity measurement data showed no significant
differences for all three conditions of storage ( 1 t I equaled 2.85 (A). 0.58 (B) and 2.97
(C)).
Table 26. Equation of the regression curves and coefficients of regression (R~) of the
viscosity measurement of orange juice over an eight month norage period.
Condition Equation R' A Y = 0.00 1 X -t 0.9373 0.10678 B Y = -0.002 X + 1.3263 0.227 18 C Y = -0.003X + 1.2150 0.05937
Figure 10. Viscosity of orange juice over an eight month storage period
1 2 3 4 5 6 7 8 9
Months
Linear equations of the regression curve and coefficients of regression were
calculated from the data (see Table 25). The graph of the viscosity (cSt) as a function of
time (months) shows a close to zero slope for the regression curve for dl the three
conditions. With the low significant differences in time and the slopes of the regression
curve of the three conditions of storage (A, B and C), no linear trend could be detected
over time. The t-test has confirmed this with t values less than 4.03 for all three
conditions. No increase or decrease in juice viscosity was found in this study. This
results are consistent with the literature data.
The Kruskal-Wallis ANOVA indicated a significant difference (P < 0.0001) in the
initial viscosities of the three conditions of storage (see Figure 10). The D u d s test
indicated that the difference was between the juices stored under conditions A and B
(considering the second months). The viscosity of juice stored under condition A was 30
% lower than the condition B. Lower viscosity for the unpasteurized orange juice is
expected because of the enzyme pectin methyl esterase was not deactivated and continued
to be effective before the complete freezing of the juice. A faster freezing could minimize
this effect. The two pasteurized orange juices differed only by 10 % in viscosity.
Probably. freezing increased the viscosity of orange juice however, no reference to this
phenomenon was found in the literature. The freezing storage could cause. on a long time
basis. problems in texture and mouth feel.
4.1.7 Colour
A phenomenon named browning can arise when orange juice is stored under
inappropriate conditions. The colour of the orange juice was measured with a Hunter
LAB instrument. The L value measures the lightness of the juice. This value is the most
important reading to measure a possible darkening of the juice. The results of the colour
measurements are presented in Table 27 and illustrated by Figure 1 1.
Table 27. Colour measurement1 (L value) of orange juice over an eight month
storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month deviation deviation deviation
1 63-16 0.28 67.22 0.07 66.25 0.1 1 2 62.74 0.39 67.44 0.1 1 67.3 1 0.26 3 61.94 1.19 66.98 0.53 65.27 0.66 4 62.56 0.33 67.28 0 2 0 66.60 0.14 5 62.62 0.4 1 67.23 0.10 65.54** 0.19 6 61.70 0.87 65.96 0.6 1 64.4 1 0.57 7 61.98 0.59 67.19 0.3 1 63.74 0.30 8 6 1.98* 0.43 66.08 0.76 64.10** 0.36
*Calculated with only two values. **Calculated with only five values, Chauvenet's criterion. 1 Average of' six measurements. unless otherwise specified.
a A 1-test analysis of the L value data showed significant differences for juices
stored under conditions A and C ( ( t 1 equaled 5.33 (A). 3.65 (B) and 14.06 (C)).
However, when the data was analyzed for a linear trend. only juice stored under condition
C showed a decreasing trend.
Table 27 and Figure 1 I showed little variation in the L value for juices stored
under conditions A and B, but no major trends could be observed. On the other hand.
juice stored under condition C showed a decrease in the L value with time. The plot of L
value as a function of time showed a slope close to zero for the regression c w e for
condition A and B (Table 28).
Table 28. Equation of the regression curves and coefficients of regression (R') of the
colour measurement (L value) of orange juice over an eight month storage period.
Condition Equation RL A Y =-0.151X + 63.02 0.53870 B Y = -0.147 X + 67.58 0.39585 C Y = -0.435 X + 67.36 0.7050 1
Linear equation of the regression curves and coefficients of regression were
calculated from the data in Table 28. The juices stored under conditions A and B had
lower slopes and lower values of coefficients of regression than the juice stored under
condition C. The storage conditions A and B did not show any significant changes during
the storage. However, the juice stored under condition C revealed a decreasing trend over
time in the value of L. The coeficient of regression indicated a variability in the
measurement of this parameter.
The value of L of the juice stored under condition C showed a slight trend to
decrease as seen with the t-test. An analysis by Kruskall- Wallis non-parametric ANOVA
showed a significant difference in the data (P < 0.000 1). The D m ' s test indicated larger
differences between months. Comparison of the first month's value with the last month's
indicates that the juice had retained its colour almost at 97 % value. This change was not
considered important for the overall colour of orange juice.
Figure 11. Colour measurement (L value) of orange juice over an eight month storage period
0 1 2 3 4 5 6 7 8 9
Months
The first month's data was analyzed to find initial differences between the three
conditions of storage. The Kruskd-Wallis ANOVA indicated significant differences (P <
0.0001) between the three conditions of storage at the first month of analysis. The
Dunn's test indicated a significant difference between juices stored under conditions A
and B. This difference is obvious in Figure 1 I. The unpasteurized juice (condition A)
was darker than juices stored under conditions B and C. This fact was contrary to the
expected results because the heat treatment is more favorable to darkening of the orange
juice. However, this may be due to the interference of the pulp during the measurement.
4.1.8 Vola tiles
Volatiles were the most important parameters measured in this study. Some of the
volatile compounds were degraded during heat treatment and produced undesirable
components during the storage period. Changes in flavour compounds were monitored by
gas chromatography (GC) analysis. One GC system was coupled with a mass
spectrometer (MS) and the other with headspace injector. These two systems combined
permitted the quantification of 40 volatile compounds. The GC-MS analysis of volatiles
were performed during the first six months (problems with power failures during analyses
in the last two months caused loss in sensitivity in the MS multiplier and data were
therefore discarded). However. data from the six months were sufficient to observe trends
in the volatiles during storage. Only the volatiles that are most relevant are presented in
this chapter. Some volatiles did not change such as Iinalool, some others were on the
limit of detection such as vinyl guaiacol, and others were not detected like furfural. HMF
and h e o l . An example of the data obtained for the first month is presented in Table 29
(see Appendix C for huther examples).
Table 29. Volatile concentration of orange juice for the first month of storage.
CONDITION A B C
methyl-butymte a- pinene ethy 1-butyrate hexanal sabinene 3 -carene a-phellandrene P-myrcene heptad limonene 2&3 - methy 1-butano 1 2-hexenal 2-hexanol y-terpinene octanal 1 -hexand 3-hexen-1 -01 nonanal dimethyl-styrene fUrfirra1 decanal linalool octanol terpinene-4-0 1 hydroxy -ethy 1-hexanoate a-terpineo 1 valencene geranial & carvone perillaldehyde 2-2-decadienal vinyl guiaicol hydroxymethyl furfUral vanillin
CONDITION A B C
Headspace-GC @%mL) (ug/mL) (ug/mL) acetaldehyde 9.67 8.45 8.56 ethyl acetate 0.33 0.23 0.20 methanol 50.55 35.37 29.00 ethano 1 607.35 636.69 592.52
ND, not detected. *limit of detection.
The following section discusses the five volatiles which showed the most
interesting differences. Volatile that showed significant differences under different
storage conditions were methanol, 1-hexanol. P-myrcene. limonene and a-terpineol.
Methanol was the second most abundant alcohol in orange juice after ethanol. Its
contribution to the overall taste of juice has not been determined according to the
literature. The results of the methanol analyses are presented in Table 30 and illustrated
in Figure 12. Data obtained for the methanol concentration of the juice samples showed
some variability, especially for juice stored under condition A. the standard deviation of
the concentrations were large. This fact may be explained by the interference of the pulp
present in some samples. The pulp appears to have complicated the measurement of
volatile compounds. The dispersion of the volatile components between the pulp and the
serum can be affected by the aggregation of the pulp found particularly in samples stored
under condition A.
Table 26. Methanol concentration' of orange juice over an eight month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (ug/mL) deviation (ug/mL) deviation (uglmL) deviation
1 50.55** 7.29 35.37** 0.98 29.00* 7.19 2 49.86** 3.64 36-21 1.80 35.62** 5. I8 3 53.01S* 2.30 33.91 2.08 32.19** 0.29 4 52.54** 2.42 3 5 -54 3.93 36.65 1.49 5 50.99 12.83 37.09* * 0.44 3 3.40 1 .58 6 55.12 6.82 37.05 0-94 38.03"" 0.52 7 59.46 12.29 35.88 1. I5 3 5.3 3 2.04 8 47.05** 3.79 35.80** 1.99 3 5 .27+ * 1.14
"Calculated with only five values. Chauvenet's criterion. I Average of six measurements. unless otherwise specified.
The t-test analysis of the methanol data showed no significant differences for all
the three conditions of storage ( I t I equaled 0.95 (A), 0.43 (B) and 1.93 (C)).
Table 3 1. Equation of the regression curves and coefficients of regression (R') of the
methanol concentration of orange juice over an eight month storage period.
Condition Equation R; A Y = -0.004 X + 0.78 0.59834
The graph of the concentration of methanol as function of time showed close to
zero slopes and poor coefficients of regression for all the three conditions of storage (see
Figure 12) indicating no linear trends. The t-ten has confirmed this with values less than
4.03 for all the three conditions of storage. No increase or decrease in methanol
concentration in orange juice was found in this study. Despite the addition of an internal
standard. the headspace data were found to be very variable.
A difference was observed between the unpasteurized (A) and the pasteurized (B
and C) orange juices. A non-parametric ANOVA (Kruskal-Wdlis) was performed on the
initial data. Significant changes were found at P < 0.0001. The Dunn's test showed a
significant difference between juices stored under conditions A and C. Methanol
concentrations in the samples processed by condition C were approximately 45 % that of
those processed by the condition A. Microbial activity before the complete freezing of
the orange juice may be responsible for these differences.
Figure 12. Methanol concentration of orange juice over an eight month storage period
I 1 I 1 1 I I I
0 1 2 3 4 5 6 7 8 9
Months
Although the role of methanol in the overall flavour of orange juice is not
described in the literature, it could be speculated that it may enhance some of the
desirable flavour notes of the orange juice.
1-Hexanol is another volatile whose contribution to the orange flavour has not
been clearly determined. Diin et al. (198 1) have classified it as a detrimental alcohol. No
other references were available in the literature. The results of the I-hexanol analyses are
presented in Table 32 and illustrated at Figure 13.
Table 32. I-Hexanol concentration' of orange juice over a six month storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (ugImL) deviation (ug/mL) deviation (ug/mL) deviation
1 0.88 0.06 0.34 0.0 1 0.08 0.06 2 0.94 0.04 0.28** 0.03 0.12 0.04 3 0.85 0.03 0.41** 0.0 1 0. 19** 0.0 1 4 0.88 0.10 0.35** 0.0 1 0.28 0.02 5 1 .OO 0.05 0.42 0.0 I 0.19 0.0 Z 6 0.72 0.07 0.35 0.03 0.35 0.02
**Calculated with only five values. Chauvenet's criterion. I Average of six measurements, unless otherwise specified.
A t-test analysis of the I-hexanol data showed significant differences for juices
stored under conditions A and C ( 1 t 1 equaled 4.18 (A), 1.29 (B) and 9.98 (C)). No linear
trend was observed in juices stored under conditions A and B even though the t-test
showed a significant difference in the two means for juice stored under condition A.
Table 33. Equation of the regression curves and coefficients of regressior. (R') of the
1-hexanol concentration of orange juice over an eight month storage period.
Condition Equation R; A Y = -0.0 17X + 0.93 0.1 1179 B Y =O.OllX + 0.31 0.18355 C Y = 0.047X + 0.04 0.77968
The graph of the concentration of 1-hexanol as hc t ion of time (months) showed
close to zero slopes and poor coefficients of regression for the three conditions. No linear
correlation or trend between the concentration of 1-hexanol and time was found (Figure
15). The juice stored under condition C seemed to increase slightly with time, but the
concentrations were at the detection limit, and the little increase was not significant.
Differences were observed between the initial concentrations of 1 -hexan01 in
juices processed under the three conditions of storage. A non-parametric ANOVA
(Kruskai-Wallis) performed on the first month data indicated the presence of significant
changes (P < 0.0001). The D m ' s test also showed significant differences between
samples of juice stored under conditions A and C. Pasteurization may be responsible for
the initial difference of 90 % between the conditions C and A (see Figure 13). Samples
processed by condition B had approximately 60 % less I -hexan01 than samples processed
by condition A. The pasteurization and/or microbial activity may interfere with the levels
of I-hexanol. Although the literature indicates that 1-hexanol is undesirable in orange
juice. However. sensory analysis (see section 4.1.9) favored storage condition A.
Figure 13. 1-Hexanol concentration of orange juice over a six month storage period
0 1 2 3 4 5 6 7
Months
The P-myrceae was the second most abundant terpene found in the orange juice
samples. It is recognized to be important in orange juice flavour with its fruity aroma and
sweet balsamic-herbaceous taste at below 10 ppm levels (Arctander, 1969). The results of
the P-myrcene analyses are presented in Table 34 and illustrated in Figure 14.
Table 34. P-Myrcene concentration of orange juice over a six month storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (ug/mL) deviation (ug/mL) deviation (ug/mL) deviation
I 2,17** 0.10 2.02 0.06 1.89 0.02 2 2.02 0.22 1 -79'' 0.05 1.64* * 0 .03 3 2.37* * 0.25 2.04* * 0.02 1.65 0.09 4 2.58 0.50 1.94 0.08 I .58* * 0.0 1 3 2.3 5 0.26 1.97** 0.05 1.39 0.26 6 1.90 0.10 1.60** 0.02 1.22 0.03
*Calculated with only five values. Chauvenet's criterion. 1 Average of six measurements, unless otherwise specified.
The P-myrcene is a non-polar component associated with the pulp rather than with
the serum of juice. As mentioned above, the aggregation of the pulp seems to interfere
with the quantification of pulp-associated components such as P-myrcene. As a result the
data obtained from the analysis were variable. especially for juices stored under condition
A. A t-test analysis showed significant differences for all the three conditions of storage
(A. B and C) ( 1 t 1 equaled 4.39 (A), 15.40 (B) and 45.06 (C)). However. these
differences between the two means were affected by the variability in the data. No linear
trend was observed in juices stored under conditions A and B even if the data for the t-test
presented a significant difference in the two means.
Table 35. Equation of the regression curves and coefficients of regression (R') of the
P-myrcene concentration of orange juice over an eight month storage period.
Condition Equation R' A Y = 0.004 X + 2.25 0.00 103 B Y = 0.047X + 2.06 0.27657 C Y =-0.1 19X + 1.98 0.92593
The graph of the concentration of P-myrcene as function of time (months) showed
close to zero slopes and poor coefficients of regression indicating the absence of a linear
correlation between the concentration of P-myrcene and the storage time for conditions A
and B. The concentration of P-myrcene in samples processed by the condition C
decreased slightly with time. A non-parametric ANOVA (Kurskal-Wallis) was performed
on the data processed by juice stored under condition C. Significant changes were found
between the different months (P < 0.0001). The Dunn's Multiple Comparison test
indicated differences between the data of the first month and the fifth and similarly
between the first and the sixth.
Figure 14. Beta-myrcene concentration of orange juice over a six month storage period
0 1 2 3 4 5 6 7
Months
A non-parametric ANOVA (Kruskal-Wallis) was performed on the initial data
from the first month of analysis to see if any initial difference can be observed due to the
method of processing. Significant differences were found (P < 0.0001). The D m ' s test
showed significant differences between juices stored under conditions A and C. Sorption
of the myrcene by the LDPE (low density polyethylene) was probably responsible for this
difference. The sorption of the P-rnyrcene continued with time of storage until the sixth
month. No degradation mechanism of P-myrcene has been reported in the literature, and
for this reason ail the losses are attributed to the sorption by the LDPE. A decrease of
about 35 % was noted between concentrations of P-myrcene at the beginning and at the
end of the analysis for condition C. The loss of 35 % by sorption due to LDPE was in
agreement with the literature (up to 40 % sorption by the LDPE). The temperature of
storage may have influenced the time taken for this reaction.
The limonene was the most abundant volatile component in orange juice. This
terpene accounted for almost 95 % of all the volatiles measured. The limonene has a low
aroma of citrus. Its contribution to the total flavour of orange juice is low and is
recognized to be an important precursor of ot'flavours through degradation to a-
terpineol. The results of the limonene analyses are presented in Table 36 and illustrated
in Figure 15.
The analysis of limonene was complicated due to relatively high levels found in
the orange juice compared to the rest of volatiles. The simultaneous extraction of all the
volatiles presented problems during quantification. In order to have sufficient
concentrations of the less abundant volatiles, the extracts were concentrated ten times.
The analysis by GC-MS of the extract showed saturation of the signal for the limonene.
As a result a new calibration curve was developed for limonene with lower
concentrations. A quadratic equation was fitted to calculate the concentration of limonene
in the juice samples.
Table 36. Limonene concentration1 of orange juice over a six month storage period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (ug/mL) deviation (ughL) deviation (ug/mL) deviation
1 193.73** 11.41 1 76.03 9.92 159.12** 4.88 2 181.1 1 17.2 1 173.29 5.82 158.1 1 4.22 3 1 92.00* * 27.22 169.34** 4.32 1 3 7.22 13.65 4 1 83 .OO 25.23 153 -67 8.69 128.69 4.78 5 (234.20)? 26.62 (202.34)' 1 1.26 (146.16)' 5.3 1 6 189.62 9-77 172.30 6.62 129.98 5.10
**Calculated with only five values, Chauvenet's criterion. I Average of six measurements, unless otherwise specified.
Data unreliable due to instrumental malfunction.
A t-test analysis of the limonene data showed significant differences for juice
stored under condition C ( I t I equaled 0.93 (A). 1.25 (B) and 11.00 (C)).
Table 37. Equation of the regression curves and coefficients of regression (R') of the
lirnonene concentration of orange juice over an eight month storage period.
Condition Equation R A Y = -0.3 11X + 188.89 0.0 1 155 B Y = -1.60X + 174.05 0.12010 C Y = -6.78X + 164.32 0.7600 1
The graph (Figure 15) of the concentration of limonene as bct ion of time
(months) shows no linear wnd of limonene concentration with time for juices stored
under conditions A and B. Regression analysis also indicated close to zero slopes and
poor coefficients of regression for juices stored under conditions A and B (Table 37).
(The limonene
calculations - spectrometer).
concentration obtained for the fifth
Figure 15 and Table 37 - due to
month was excluded in the above
technical problems with the mass
Figure 15. Limonene concentration of orange juice over a six month storage period
0 I 2 3 4 5 6 7
Months
Like P-rnyrcene, the Iimonene is a non-polar component associated with the pulp
rather than with the serum of the juice. Differences found in the pulp consistency may
explain the difficulty to obtain reproducible measurements. sIts appears to be easier to
extract the volatile component associated with the serum than with the pulp. The
concentration of limonene in juice stored under condition C decreased with time
according the Figure 15, and confirmed by the t-test. A non-parametric ANOVA
(Kurskal-Wallis) was also performed on this data. Significant differences were found
between the different months (P < 0.0001). The Dunn's Multiple Comparison test
indicated differences between the first and the fourth and also between the first and the
sixth months among others.
The first month data were analyzed to detect initial differences among the three
sample types due to sorption or pasteurization. A non-parametric ANOVA (Kruskal-
Wallis) was performed on the fiat month data From each group. Significant differences
were found (P c 0.0001). The D m ' s test showed significant differences between juices
stored under conditions A and C. Sorption of the limonene by the LDPE and
pasteurization were probably responsible for these differences. The differences in
limonene concentration between juices stored under conditions A and B can be explained
by the pasteurization. A difference of about 8 % was found between the two means for
the two processes (A and B). The initial sorption of the LDPE can be estimated by
comparing data from juices stored under conditions B and C. A difference of about 8 %
was found between the means of the first months of juices stored under conditions B and
C. The sorption of the limonene probably continued with time until the sixth month.
Another phenomenon was probably occurring during the six months of storage: the
degradation of the limonene. especially to a-terpineol. Adsorption of limonene may lead
to a loss of about 25 % according to the literature but the temperature of storage may
affect the rate of this reaction. The difference in concentration of limonene between the
first month and the sixth month was about 18 Oio. About a total of 30 ppm of limonene
was adsorbed or degraded at a rate of about 5 ppm per month..
The a-terpiaeol is one of the most detrimental degradation products in orange
juice. Its gives a stale, musty or piney aroma at concentrations around 2-3 ppm. In
general, the a-terpineol concentration increases with heat treatment and storage. The
results of the a-terpineol analyses are presented in Table 38 and illustrated in Figure 16.
Table 38. a-Terpineol concentration' of orange juice over a six month storage
period.
Condition A Condition B Condition C
Mean Standard Mean Standard Mean Standard Month (ug/mL) deviation (ug/mL) deviation (ughnL) deviation
1 4.09 0.33 1.43 0.09 1.67** 0.05 2 3.56* * 0.05 1.80 0.06 2.50 0.08 3 4.34* * 0.13 t -55 0.05 2.33** 0.08 4 3.95 0.26 1.40 0.04 3. I5 0.14 5 4.10** 0.08 1.37 0.06 3.13 0.10 6 4.49 0.13 1.15** 0.03 3.46 0.08
*+Calculated with only five values. Chauvenet's criterion. I Average of six measurements, unless otherwise specified.
A t-test analysis showed significant differences in the a-terpineol data for juices
stored under conditions B and C ( I t 1 equaled 2.79 (A), 6.40 (B) and 43.50 (C)).
However, no linear trends were observed in juices stored under conditions A and B
although the t-test showed a significant difference in the two means for condition B.
Table 39. Equation of the regression curves and coefficients of regression (R') of the
a-terpineol concentration of orange juice over an eight month storage period.
Condition Equation R' A Y = 0.093 X + 3.77 0.28493 B Y =-O=OSI X + 1.73 0.49708 C Y = O X 3 X + 1.54 0.88 18 1
The graph (see Figure 16) of the concentration of a-terpineol as h c t i o n of time (months)
showed close to zero slopes and poor coefficients of regression for juices stored under
conditions A and B (Table 39). No decreasing or increasing trends were observed for the
concentration of a-terpineol with time for juices stored under conditions A and B.
However, the concentration of a-terpineol in juice samples stored under condition C
increased with time according the Figure 16 and confirmed by the t-test.
Figure 16. Alpha-terpineol concentration of orange juice over a six month storage period
Months
An non-parametric ANOVA (Kurskal-Wallis) was performed on the data of juice
stored under condition C. Significant differences were found between the different
months (P < 0.000 1). The Dunn's Multiple Comparison test indicated differences
between most of the months. An increasing mnd could be observed for the concentration
of a-terpineol in orange juice after six months of storage. An increase of about 50 % in
a-terpineol concentration was observed (about 1.80 ppm) at the end of the sixth month.
This is equal to the formation of about 0.30 ppm per month and this value represents only
the 6 % of the drop in limonene concentration.
A non-parametric ANOVA (Kruskal-Wallis) was performed on the first month
data to detect any initial differences. Significant differences were found (P c 0.0001).
The D u d s test showed significant differences between juices stored under conditions A
and B. The percent difference between the concentrations of a-terpineol in juices stored
under conditions A and B at the first month was 65 % higher in A. This difference in a-
terpineol concentration between unpasteurized orange juice (A) and pasteurized orange
juice (B) was unexpected because a-terpineol concentration is supposed to increase with
increasing heat treatment, such as during pasteurization. However, the sensory evaluation
data indicated (see the following section) that the juice stored under condition A had the
highest scores for the "fresh pressed" note. This might indicate that the off-flavour of a-
terpineol was masked due to the presence of other unknown components.
4.1.9 Sensory evaluation
Sensory evaluation is essential to correlate chemical and physical changes to the
overall organo leptic properties of the juice. Without sensory evaluation. it is difficult to
assen that one condition was better than the others. The sensory evaluation was
performed by a panel of twelve orange juice experts. The panel judged twelve orange
juice attributes (acidity, sweetness, orange perfume. orange taste. homogeneity, presence
of pulp. fresh pressed, oily. maturity, plastic, oxidation and peely) over the eight month
period. The scores for the first month were not included in the analysis due to
adjustments introduced after the first session such as some attributes were added and the
scale was changed from 100 to 130. The mean scores of each attribute of the sensory
evaluation are presented in Table 40.
Table 40 shows the monthly means (20 to 24 evaluations) of the scores by the
expert judges. The standard deviations are not included in Table 40. Deviations between
the scores of the expert panel were sometimes very large but the data seemed to be
consistent during the period of analysis.
Analysis of the sensory scores indicated that bbmaturity" and "peely" attributes did
not show any significant changes during the storage period for all the conditions. The
results were initially the same and no trend in time was observed. Concerning the
attributes that showed variations. the changes can be divided into two classes: initial
changes and changes due to storage (8 months). The "acidity " attribute was constant in
time but juice stored under condition A seems to have slightly less acidic taste thm the
pasteurized juices (conditions B and C). Similar pattern was observed for the "oily"
attribute. The opposite phenomenon was noted for the "sweetness" attribute. The
unpasteurized juice (condition A) had slightly higher scores for the sweet taste. The
condition A was also noted to have a better "perfine of orange" compared to the
conditions B and C. Figure 17 showed the results obtained for the anribute "orange
taste".
Table 40. Mean scores of sensory evalunliun' of orange j juice over eight months o f storage.
Month Acidity S d n e s s Onnw Orange Homogeneity Presence of Fresh oily Maturity Plastlc Oxidation Pwly pedume taste PU~P P - U ~
Condition A: Not pasteurized orange juice stored at - 18°C' in glass bottles. a
Condition 13: Pasteurized orange juice stored ill - 1 8 " ~ in ~ I ~ I S S bottles. Condition C: Pasteurized orange juice stored at + I"C in polyethylene bag. 'scores are on a scale of*O- 130.
Figure 17. Evolution of the attribute "orange taste" in orange juice over an eight month storage period
1 2 3 4 5 6 7 8 9
Months
The "orange taste" was more pronounced in orange juice stored under condition A
which showed about 30 % higher score compared to condition C. A more appreciable
difference was found for the "fresh pressed" attribute (Figure 18). Orange juice stored
under condition A showed 40 % higher score than condition B and a 50 % higher score
than condition C. Therefore, the juice stored under condition A had significantly better
fresh pressed orange taste. On the other hand, a decreasing freshness was observed in
juice stored under condition C to the extent of 25 %. Consistent with this observation. the
juice stored under condition C also appeared to have higher scores in two attributes:
"oxidation" and "plastic". The former showed a tendency to increase during the storage
period ( 6 5 %). On the other hand, the "plastic" attribute was five times higher in juice
stored under condition C compared to A.
Figure 18. Evolution of the attribute "fresh pressed" in orange juice over an eight month storage period
0 , I i 1 I I 1 I I
1 2 3 4 5 6 7 8 9
Months
The last two attributes, "homogeneity" and "presence of pulp" were significantly
different in the pasteurized and unpasteurized juices. Figures 19 and 20 show
respectively the variation in the "homogeneity" and the "presence of pulp" over a period
of eight months.
Figure 19. Evolution of the attribute "homogeneity" in orange juice over an eight month storage period
1 2 3 4 5 6 7 8 9
Months
The juice stored under condition A was about 50 % less homogenous and had 50
% more pulp than the two other pasteurized orange juices (B and C). The pasteurization
seems to have a stabilizing effect on the pulp matrix. According to Figure 20, a
decreasing trend for the presence of pulp in juice stored under condition C was observed
(decrease of 25 %). This fact could be caused by a variation in texture due to pulp
aggregation.
Figure 20. Evolution of the attribute "presence of pulp" in orange juice over an eight month storage period
1 2 3 4 5 6 7 8 9
Months
4.2 Summary of all the parameters
All changes found in the ten parameters are summarized in Table 4 1.
Table 41. Summary of the changes during the period of storage.
PARAMETER COMMENTS
Density Cloud a
Sedimentation a
Sugars Organic acids Ascorbic acid
Viscosity a
8 Colour
Volatile Methanol a
P-myrcene
Limonene
No change An increase of 10 % in the transmittance of the juice stored under condition A, after 3 hours of monitoring. Sedimentation of 50 % of the pulp particles after 3 hours for juice stored under condition A. No change No change A decrease of 5 % in concentration of ascorbic acid for juice stored under condition C over the storage period. Initial differences: (1) viscosity of juice stored under condition A was 30 % lower than in juice stored under condition B. (2) A difference of 10 % between the two pasteurized orange juices (conditions B and C). 97 % of the colour was retained under condition C (loss of 3 %). Juice stored under condition A was the darkest (screening effect of the pulp particles).
Initial difference in concentration of methanol: orange juice stored under condition A showed 45 % higher concentration than juices stored under conditions B and C. Initial difference in concentration of I -hexanol: orange juice stored under condition A showed 60 and 90 % higher concentrations than juices stored under conditions B and C respectively. Decrease of about 35 % in concentration of p-myrcene for juice stored under condition C. Decrease of 18 % in concentration of limonene for juice stored under condition C (5 ~ ~ d r n o n t h ) .
a-terpineol
10 Sensory analysis Maturity Peely Acidity
Oily
Sweetness
Perfume of orange
Orange taste
Fresh pressed
a
a
0
0
Oxidation
Plastic
Homogeneity
Presence of pulp
a
Increase of about 50 % in concentration of a-terpineoi for juice stored under condition C (0.30 ppdmonth). Initial difference in concentration of a-terpineol: orange juice stored under condition A showed 65 % (4 ppm) higher concentrations than juices stored under conditions B and C*
No change No change Initial difference: slightly less acid taste in juice stored under condition A than in juice stored under conditions B and C. Initial difference: slightly less oily taste in juice stored under condition A than in juice stored under conditions B and C. Initial difference: slightly higher sweet taste in juice stored under condition A than in juice stored under conditions B and C. Juice stored under condition A had a better p e h e of orange than juices stored under conditions B and C. Juice stored under condition A showed a higher initial difference of about 30 % compared to condition C. A 50 % initial difference was observed between the conditions A and C. Loss of freshness of 25 % in juice stored under condition C during storage. Tendency to increase (35 %) in juice stored under condition C over the storage period. Five times higher in juice stored under condition C than in juice stored under condition A. Juice stored under condition A was about 50 % less homogenous than juices stored under both conditions B and C. Unpasteurized orange juice (condition A) contained 50 % more pulp. A wnd to decrease (25 %) in the presence of the pulp in iuice stored under condition A over the storage wriod.
4.3 Correlation between chemical and physical changes and sensory
evduation
Correlation between chemical, physical and sensory analysis was complicated due
to the number of variables that could affect the taste of the orange juice. However some
suggestions can be put forward.
First. lower viscosity and higher rate of sedimentation of orange juice stored under
condition A (unpasteurized orange juice) can be associated with the attributes
"homogeneity" and "presence of pulp". The activity of the PME enzyme could affect the
cloud which suppons the pulp particles. Most of the particles contributing to the
viscosity of the orange juice are pectin molecules. The lack of homogeneity could be due
to a low freezing rate which is associated with large ice crystal formation and
agglomeration of the pulp into larger particles. This phenomena was evident for the
expert panel. The presence of pulp could be advantageous or disadvantageous depending
on the consumer's taste. To minimize the global effect of the enzymes (PME and others).
a faster freezing rate of orange juice could be suggested to overcome this problem in
unpasteurized orange juice.
The attributes "acidity" and "sweetness" were perceived slightly differently in the
unpasteurized orange juice (condition A). It could not be correlated to the chemical
analysis since no significant changes were observed in the sugar and organic acid
contents. The attributes "plastic" and "oxidation" also could not be correlated to chemical
analysis. These off flavours probably leaked, in large part, fiom the packaging material
(LDPE), which could impart the undesirable attributes noted..
The most interesting data were fiom the "fiesh pressed" and "orange taste"
attributes. The high scores of the expert panel given for juice stored under condition A
versus the pasteurized juices (conditions B and C) were important for the evaluation of
this study and for fisher research. These attributes were probably related, in large part,
to the volatile components identified in the orange juice. According to section 4.1.8
(volatiles), some important differences were found among the juices stored under the
three conditions. For orange juice stored under condition A, the concentrations of
methanol and I-hexanol were 45 and 90 % higher respectively than in C. In addition
condition A showed no sorption of volatiles by the packaging material. On the other
hand, the juice stored under condition C which scored low in these attributes, had 18 %
sorption or degradation of limonene and an increase of 0.30 ppm per month in alpha-
terpineol concentration. This obsentation could explain the low scores obatined in the
"fresh pressed" attribute. However. the finding of a high content of an off-flavour
compound alpha-terpineol in orange juice (condition A) which as also rated high in
sensory evaluation was unexpected.
CHAPTER 5
CONCLUSION
Orange juice is prone to rapid chemical deterioration which is highly dependent on
storage temperature. The condition A (unpasteurized frozen orange juice) seems to be the
optimum storage condition with the retention of some important volatile components and
the highest sensory evaluation scores. The condition B (pasteurized frozen orange juice)
is completely discarded as a method of storage because of its high cost and the low
sensory evaluation scores associated with the process. Condition C (the aseptically
packaged orange juice stored at +1 "c) is economically favorable but losses of important
volatile components by sorption on LDPE and degradation during long term storage led to
its lower rating in sensory evaluation. It could be proposed, as a practical alternative that
mixing aseptically packaged pasteurized orange juice with unpasteurized frozen juice. can
generate an orange juice with good sensory qualities and at a lower cost to the consumers.
However, a more rapid freezing can avoid the effect of enzymes and agglomeration of the
pulp particles. Moreover. a method to stabilize the pulp in the thawed unpasteurized
orange juice should also be developed before commercial application of the freezing
method of storage.
Although, in this study, some differences were found in the concentrations of the
volatile components in the three conditions of storage (A, B and C). However, undetected
and very low concentrations of low threshold compounds could influence the aroma
profile. In addition, synergetic effect of different volatiles are still unknown. The effect,
on the flavour profile of orange juice, of the individual volatile components such as
methanol, 1-hexanol, a-terpineol, limonene and myrcene, should be known in order to
have a better understanding of the complex aroma of the orange juice.
Future studies on the possible effects of the microorganisms and enzymes on the
quality of fresh unpasteurized orange juice. could lead to a better understanding of the
reactions that could occur before the complete Freezing of the orange juice. Higher
concentrations of certain volatiles, such as methanol and 1-hexanol. found in the
unpasteurized orange juice could be due to the activity of enzymes and microorganisms.
Finally, research could also be undertaken, to verify the possibility of masking of off-
flavours such as a-terpineoi.
REFERENCES
Adams, J.P. 1989. Proceedings of the 22nd Annual IFT Short Come on the Food
Industry, University of Florida, Gainsville, FL. pp. 245-257.
-4hmed. E.M., Dennison, R..4., Dougherty, R.H. and Shaw P.E. 1978. Flavor and odor
thresholds in water of selected orange juice components. J. Agric. Food
Chem. %(I): 187-191.
Arcander, S. 1969. P e h e and Flavor, Vol. I and 11, (Arctander, S.. ed), Montclair. N.-
Askar. A. 1984. Flavor changes during production and storage of h i t juice. Fluss.
Obst. 5 1(11): 564-569.6 10-6 12.
Auger. M. 1997. La guene des jus d'orange refkigeres: Tropicana en tEte au Quebec.
Les Affaires (sarnedi 1 '' fevrier), pp. 34-3 5.
Bauernfeind, LC. and Pinkert, D.M. 1970. Advances in Food Research. vol.18.
(Chichester, C.O., Mark E.M., Stewart G.F., eds), Academic Press, New-
York, p. 219.
Beeman, C.P. 1987. Improved efficiency for colorimetric determination of fumual in
citrus juices. L Assoc. Off. Anal. Chem. 70(13): 601603.
Berry. R.E. and Tatum, J.H. 1965. 5-Hydroxyrnethylfurfural in stored foam-mat orange
powders. J. Agric. Food Chem. 13: 588-590.
Bokhari. A.A. Shah, T.H. and Ahmad. I. 1995. Effect of packaging and storage on the
quality of Kinnow (Citrus reticdata) juice concentrate. Fruit Proc. 1 1 :
368-3 7 1.
Brown, M., Kilmer, R.L. and Bedigian, K. 1993. Chapter 1. Overview and trends in the
h i t juice processing industry. In: Fruit Juice Processing Technology, (S.
Nagy, C.S. Chen, P.E. Shaw, eds), Agscience, Inc. FL, pp. 1-22.
Carter. P.D. 198 1. Reconstituted Florida Orange Juice. ProductionPackaging
Distribution. Technical Manual. Florida Department of Citrus, Lakeland
FL.
Chen. C.S.. Shaw. P.E. and Parish. M.E. 1993. Chapter 5. Orange and tangerine juices.
In: Fruit Juice Processing Technology. (S. Nagy. C.S. Chen. P.E. Shaw.
eds). Agscience. Inc., FL.
Curl. A.L. 1947. Concentrate orange juice storage studies. The effect of degree of
concentration and temperature of storage. Canner 105: 14.
Doyon. G.. Gaudreau. G.. St-GeIais. D.. Beaulieu. Y. and Randall. C.J. 1991.
Simultanecus HPLC determination of organic acids. sugars and alcohols.
Can. Inst. Sci. Technol. 24(12): 87-94.
D . P 1980. Aroma quality of orange juice, a brief review. Alimenta 19: 35-36.
Dlin: P.. Schobinger. U. and Waldvogel, B. 1981. Aroma quality of orange juice after
filling and storage in soft packages and glass bottles. Alimenta 20(4): 91 - 94.
Fellen, P.J. 1988. Shelf life and quality of freshly squeezed. unpasteurized,
polyethylene-bottled citrus juice. J. Food Sci. 53(6): 1699- 1702.
Fiddler, W., Parker, W.E.9 Wassennan, A.E. and Doerr, R.C. 1967. Thermal
decomposition of ferulic acid. J. Agric. Food Chem. 15: 757-76 1.
Graumlich. T.R., Marcy, J.E. and Adams, J.P. 1986. Aseptically packaged orange juice
and concentrate: A review of the influence of processing and packaging
conditions on quality. J. Agric. Food Chem. 34: 402-405.
Halek. G. W. and Meyers, M.A. 1 989. Comparative sorption of citrus flavor compounds
by low density polyethylene. Pack. Tech. Sci. 2: 14 1 - 146.
Handwerk. R.L. and Coleman. R.L. 1988. Approaches to the citrus browning problem.
A review. J. Agric. Food Chem. 36: 23 1-236.
Hendrix. C.M.. Jr. and Hendrix. D.L. 1996. Flavor-general concepts. In: Quality
Control Manual for Citrus Processing Plants. Vol. 111. (S. Nagy. C.S.
Chen. P.E. Shaw. eds), Agscience. Inc., FL. pp. 3-61.
Holman. J.P. and Gajda. W.I. 1984. Chapter 3. Analysis of experimental data. In:
Experimental Methods for Engineers, McGraw-Hill Book Company. New-
York, pp. 46-99.
Huet, R. 1969. The aroma of citrus juice. Fruits 23(9): 453-47 1.
Irwe, S. and Olsson, I. 1994. Chapter 3. Reduction of pectinesterase activity in orange
juice by high pressure treatment. In: Minimal processing of Foods and
Process Optimization. An Interface, (RP. Singh and F.A.R. Oliveira, eds.),
CRC Press. New-York, pp. 35-42.
Kaanane, A., Kane, D. and Labuza T.P. 1988. Time and temperature effect on stability
of Moroccan processed orange juice during storage. I. Food Sci. 53 (5 ) :
1470-1473, 1489.
Kanner, J., Fishbein, I., Shalom, P., Harel. S. and Ben-Gem I. 1982. Storage stability of
orange juice concentrate packaged aseptically. J. Food Sci. 47: 42943 1,
436.
Kefford. J.F., McKenzie. H.A. and Thompson. P.C.O. 1959. Effects of oxygen on
quality and ascorbic acid retention in canned and frozen oranges juices. J.
Sci. Food Agric. 10: 5 1-63.
Kennedy. J.F.. Rivera. Z.S., Lloyd L.L.. Warner, F.P. and Jumel. K. 1992. L-ascorbic
acid stability in aseptically processed orange juice in TetraBrik cartons and
the effect of oxygen. Food Chem. 45: 327-33 1.
Kertest 2.1. 1 95 1 . The pectic substances. Interscience Publishers. Inc.. New-York and
London.
Lee. H.S. and Nagy, S. 1988. Relationship o f sugar degradation to detrimental changes
in citrus juice quality. Food Tech. Q( 1 1 ): 9 1-94.97.
Lee, H.S. and Nagy. S. 1996. Chapter 9. Chemical degradative indicators to monitor the
quality of processed and stored citrus products. In: Chemical Markers for
Processed and Stored Foods, (Lee, LC. and Kim, HA., eds.), American
Chemical Society, Washington, DC, pp. 86-1 06.
Marry. J.E., Graumlich, T.R, Crandall. P.G. and Marshall, M.R. 1984. Factors affecting
storage of orange concentrate. L Food Sci. 49: 1628- 162%
Marcy, J.E., Hansen. A.P. and Graurnlich T.R. 1989. Effect of storage temperature on
the stability of aseptically packaged concentrated orange juice and
concentrated orange drink. J. Food Sci. 54(1): 227-230.
Marshall, M., Nagy, S. and Rouseff R.L. 1986. Factors impacting on the quality of
stored citrus fruit beverages. In: The Shelf Life of Food and Beverage.
(G. Charalambous, ed.), Elsevier, Amsterdam, pp. 23 7-255.
Merin. U. and Shomer I. 1984. Structural stability of fresh and frozen-thawed "Valencia"
(C. sinensis) orange juice. J. Food Sci. 49: 148991493. 15 13.
Meydav. S.. Saguy. I and Kopeiman. I.J. 1977. Browning determination in citrus
products. J. Agric. Food Chem. 25(3): 602-604.
Miller. J.C. and Miller. J.N. 1988. Statistics for Analytical Chemistry. second edition.
Ellis Horwood Limited. New-York. pp. 55.58.
Miller. W.M. and Hendrix. C.M.. Jr. 1996. Fruit quality, inspection. handling. sampling
and evaluation. In: Quality Control Manual for Citrus Processing Plants.
Vol. 111. (S. Nagy. C.S. Chen. P.E. Shaw. eds). Agscience. Inc.. FL. pp.
233-25 1.
Nagy, S. 1980. Vitamin C contents of citrus h i t and their products: A review. J. Agric.
Food Chem. 28: 8-18.
Nagy, S. 1996. Factors affecting the flavor of citrus fruits and their juice products. In:
Quality Control Manual for Citrus Processing Plants. Vol. 111, (S. Nagy.
C .S. Chen, P.E. Shaw, eds), Agscience, Inc., FL. pp. 102-1 33.
Nagy, S, and Randall, V. 1973. Use of furfUral content as an index of storage
temperature abuse in commercially processed orange juice. J. Agric. Food
Chem. 32: 979-98 1.
Nagy, S. and Smoot, J.M. 1977. Temperature and storage effects on percent retention
and percent U.S. recommended dietary allowance of vitamin C in canned
single-strength orange juice. J. Agric. Food Chem 25(1): 13 5- 1 3 8.
Nickerson. D. 1946. Color measurement and its application to the grading of agriculttlral
products. U.S. Dept. Agric. Misc. Publ. 580, Washington. D.C.
Olson. R.L. 1968. Objective tests for frozen food quality. In: Low Temperature
Biology of Foodstuffs. (J. Hawthorn. €..I Rolfe. eds.). Pergarnon Press,
London. p. 138.
Paul. P.C. 1972. Chapter 1. Basic scientific principles. sugars. and browning reactions.
In: Food Theory and Application, (P.C. Paul. H.H. Palmer. eds.). John
Wiley and Sons. Inc.. New-York. pp. 1-75.
Pieper. G.. Borgd. L.. Ackermann. P. and Fellers. P.J. 1992. Absorption of aroma
voiatiles of orange into laminated carton package did not affect sensory
qualtiy. J. Food Sci. 57(6): 1408-1 41 1.
Pompei, C.. Rossi. M. and Baroui, E. 1986. Rectified concentrated grape must: HPLC
analysis of phenolic substances and hydroxyrnethyfiufural. J. Food Sci.
5 l(6): 1498-1 500.
Radford,. T.. Kawoshirna, K., Friedel. P.K., Pope, L.E. and Gianturco M.A. 1974.
Distribution of volatile compounds between the pulp and serum of some
fruit juices. J. Agric. Food Chem. 22: 1066-1070.
Rangarma, S. and Raghuramaiah, B. 1970. Stabilization of cloud in orange squash.
Indian Food Packer 24(2): 14-2 1.
Roberstson, G.L. and Samaniego C.M.L. 1986. Effect of initial dissolved oxygen levels
on the degradation of ascorbic acid and the browning of lemon juice during
storage. J. Food Sc. 51(1): 184-1 87, 192.
Sadlrr. G.D. and Braddock, R.J. 1990. A research note. Oxygen permeability of low
density polyethylene as a function of limonene absorption: An approach to
modeling flavor "scalping". J. Food Sci. 55(2): 587-588.
Schreier. P.. Drawert, F.. Junker, A. and Mick. W. 1977. The quantitative composition
of natural and technologically changed aromas of plants. I1 Aroma
compounds in oranges and their changes during juice processing. 2.
Lebens. Unters. Forsh. 164: 188-193.
Shaw. P.E. 1986. The flavour of non-alcoholic h i t beverages. In: Developments in
Food Science. vo1.3. I.D. Morton. A.J. Macleod. eds). Elsevier Science
Publishers. Amsterdam, The Netherlands.
Shaw. P.E. 199 1. Chapter 9. Fruits 11. In: Volatile compounds in Foods and Beverages.
(Henk Maarse.ed), New-York. pp. 305-327.
Shaw. P.E. 1996a. Volatile components important to citrus flavors. In: Quality Control
Manual for Citrus Processing Plants, Vol. 111, (S. Nagy, C.S. Chen, P.E.
Shaw, eds). Agscience, Inc., FL. pp. 134-1 72
Shaw, P.E. 1996b. Shelf life and ageing of citrus juices, juice drinks and related soft
drinks. In: Quality Control Manual for Citrus Processing Plants, Vol. 111,
(S. Nagy, C.S. Chen. PI. Shaw, eds), Agscience, Inc., FL. pp. 173499.
Shaw. P.E., Ahmed, E.M. and Dennison, R.A. 1977. Orange juice flavor: contribution of
certain volatile components as evaluated by sensory panels. Proc. Int. Soc.
Citriculture 3 : 804-807.
Shaw, P.E., Nagy, S. and Rouseff R.L. 1993. The shelf life of citrus products. In: Shelf
Life Studies of Foods and Beverages, Chemical, Biological, Physical and
Nutritional Aspects, Developments in Food Science, No 33. (G.
Charlambous, ed.), Eisevier, Amsterdam, pp. 755-778.
Shaw, P.E.. Taturn, J.H. and Berry, R.E. 1968. Base-catalysed fructose degradation and
its relation to non-enzymic browring. J. Agric. Food Chem. 16(6): 978-
982.
Stewart. I. 1980. Color as related to quality in citrus. In: Citrus Nutrition and Quality.
Nagy. S. and Attaway. J.A. (Eds.), ACS Symposium Series 143, American
Chemical Society. Washington. DC, p. 129.
Tatum. J.H.. Nagy. S. and Berry R.E. 1975. Degradation products formed in canned
single-strength orange juice during storage. J. Food Sci. 40: 707-709.
Ting. S.V.and Rouseff, R.L. 1986. Citrus Fruits and Their Products. Analysis and
Technology, (Marcel Dekker, Inc.. ed), New-York.
Vanel. C. 1980. Citrus juice processing as related to quality and nutrition In: Citrus
Nutrition and Quality, (S. Nagy, J.A. Attaway, eds.). ACS Symposium
Series 143, American Chemical Society: Washington, D.C. pp. 225-27 1.
Veldhuis, M.K. 1971. Orange and tangerine juices. In: Fruit and Vegetable Juice
Processing Technology. 2nd ed., (Tressler D.K, kslyn M.A., eds.), AVI
Publishing Co.. Westport, CT. pp. 3 1-9 1.
Ziegler, E. 1970. Assessment of citrus oils. Zur Beurteilung von Zitrusolen Deutshe
lebensm. Rundschau 66(9): 290-296.
APPENDIX A
PROFIL DESCRlPTIF Q W T I T A TIF*
NOM:
Prenom: Date:
Nous dons evaluer un ichantillon de jus d'orange. en fonction de l'intensite de ses
caracteristiques do srveurs ET d'arhes. Pour chaque descripteur. on notera l'intensite
perGue sur l'echelle bomee correspondante. Une echelle d'ajout de termes se trouve en
fin de questionnaire. en cas de besoin.
MDIQUEZ LE CODE DE L'ECHANTILLON D E G U S ~ DANS LA CASE
SUIVANTE:
Descri pteur: AC IDE
MTENSI'T$ nulle tres forte
Descripteur: SUC&
MTENSIT~. nulle tres forte
Descripteur: PARFUM D'ORANGE
INTENS 1s nulle tres forte
* The test was conducted in French.
Descripteur: G O ~ T (AROME) D'ORANGE
M T E N S I ~ nulle tres forte
Descripteur: HOMOGENEITE DU N S (TEXTURE)
M T E N S I ~ nulle tres forte
Descripteur: P R ~ E N C E DE PULPE (TEXTURE)
INTENS ITE faible forte
Descripteur: FRAIS PRESSE
MTENS ITE pas du tout tres
Descripteur: HUILEUX
[NTENSITE pas ciu tout tres
Descripteur: MATUN* DE L'ORANGE
MTMSITE pasdu tout
Descripteur: LAMINEPLASTIQUE
INTENSITE pas du tout tres
Descripteur: OXYDATION
IN TENS IT^. pas du tout tres
Descripteur: ZESTE
IN TENS IT^. pas du tout tres
E W P L E S OF STATISTICAL CALCULATIONS
on (Holman and Gajda, 1984)
-
In applying Chauvenet's criterion to eliminate dubious data points. one first calculates the
mean value and standard deviation using all data points. The deviations of the individual
points are then compared with the standard deviation in accordance with the following
information, and the dubious points are eliminated.
Number of readings
n
6
Ratio of maximum acceptable deviation
to standard deviation. dm,&
1.73
The mean (x,,,) and the o value is given by:
Example: Methanol for the processing condition A
Data 40.67.45.35,54.13,59.01,46.35,58.73
Mean 50.7 1
a 7.66
di = 1 40.67 - 50.7 1 1 = 10.04
dmJm = 10.04/7.66 = 1.3 1
The data was not rejected.
Kruskal- Wallis non-~ametric ANOVA (Ins tat software)
Example: Methanol for the processing condition A
Kruskal-Wallis Nonparamedc .4,VOV-4 Ten
Number Sum Mean of of of
Group Points Ranks Ranks
Kmkal-Wallis Statistic KW = 8.279 The exact P value calculation would have taken too Long. - so the ch-square approximate P value is s h o w innesd. The P value is 0.3086. considered not sigzificulr. Variation among column medians is nor sigiircmtiy grever rhan cspcc: by chance.
D m ' s Multiple Comparisons Test
Mean Comparison DiEerence P value
-III-LI---- - ----.-.. 1 vs. 2 3.100 ns P X M i L VS. j -5.000 P>O.OI L vs* 4 -2.000 ns P>0.05 I vs. 5 2.100 ns P>O.Oj L vs. 6 -6.900 ns P>O.O5 I vs. 7 -9.133 n~ P>0.03
I vs. 8 8.600 ns PO.05 2 vs. 3 -8200 n~ PW.05 2 VS. 4 -5.200 ns PO.05 2 vs. 5 -1.100 ns PW.05 2 vs. 6 -10.100 n~ P>O,OS 2 VS. 7 -12.433 ns P>O.OS 2 VS. 8 5.400 ns EW.05 3 vs. 4 3.000 ns bO.05 3 vs. 5 7.100 ns PW.05 3 VS. 6 -1.900 ns PW.05 3 vs- 7 4233 us P~0.05 3 vs. 8 13.600 ns p~I.05 4 1-s. 5 4.100 ns D0.05 4 vs. 6 4900 ns P>O.OS 4 vs. 7 -7.233 ns P>O.OZ 4 vs. 8 10.600 ns P>O.OI' 5 VS. 6 -9.000 n~ P>O,OZ 5 vs. 7 -I 1.323 n~ P>O.O5 5 vs. Q 6.500 ns P>O.OT 6 vs. ' -1.333 ns P>O.OT 6 vs. 8 15.500 ns P>O.OZ 7 vs. 8 17.8; S ns P>O.OZ
These trsis are based on a Gaussian approximation. n e y are only accur for large sarnpie sizes.
Sunmap- of Dan Surnber of
Group Poinrs Median Minimum 4Ia~rnum
t-test (Miller and Miller, 1988)
Taking the null hypothesis that the two methods give the same results. The difference
was measured from the two individual standard deviation sl and s2 by using the equation:
It can be show that "t" is given by:
t = (u,, - I,) / s(l/n, - 1/nz)'"- where 'T' has n, + n2 - 2 degrees of freedom.
Value of "t" for a confidence interval of 99 %, for 5 degree of freedom = 4.03.
Example: Methanol for the processing condition A. B and C between months 1 and 8.
METHANOL First month Last month n Average std n Average std
s Value
t value Value
All the results were below 4.03, therefore no difference were found between the fist and
the last month.
APPENDIX C
CONCENTRATlON OF VOLATILES IN ORANGE JUICE FOR THE
MONTHS 2-6
Month 2 CONDITION A B C
GC-MS (uWmL) ( u g h L) (ughL) methyl-butyrate 0.01 0.0 I 0.0 1 a-pinene ethy Lbutyrate hexanal sabinene 3-carene a-phellandrene Q-myrcene heptanal limonene 2&3-methy I-butanol 2-hexenal 2- hexano 1 y-terpinene octanaI I -hexan01 3-hexen- 1-01 nonanal dimeth y I-styrene firfkral decanai Iinatoo t octanol terpinene-4-ol hydroxy-ethyl-hexanoate a-terpineol valencene geranial& carvone perillaldehyde 2-2-decad ienal viny t guiaicol hydroxymethyi hfiral
acetaldehyde ethyl acetate methanol ethanol 543 -02 580.07 599.23 ND, not detected, *limit of detection.
Month 3
CONDITION A B C
a-pinenc ethy 1-buryrate hexanal sabinene 3-carene a-phellandrene P-rn yrcene heptanal limonene 2&3-methyl-butanol 2-hexenal 2-hexanol y-terpinene octanal 1 -hexan01 3 -hexen- 1-01 nonanal dimethyl-styrene brfural decanal linalool octanol terpinene-4-ol hydroxy-ethyl-hexanoate a-terpineol valencene geranial & carvone perillaldehyde 2-2-decadienal vinyl guiaicol hydroxyrnethy I f irfiral vanillin
acetaldehyde ethyl acetate methanol ethanol ND, not detected.
Iimit of detection.
Month 4
CONDITION A B C
GC-MS (u%m L) ( W m L) ( W m L ) methy 1-butyrare 0.09 0.07 0.07 a-pinene ethy I-butyrate hexanal sa b inene 3-carene a-phellandrene P-rn yrcene heptanaI limonene 2&3-methy I-butanol 2-hexenal 2-hexanol y-terpinene octanal I -hexan0 1 3-hexen- 1-01
nonanal dimethy [-styrene furfural decanal linaloo t octanol terpinene-4-01 hydroxy-eth y I-hexanoate a-terpineol valencene geranial & carvone perillaldehyde 2-2-decadienal vinyl guiaicol h ydroxyrnethy l hrfura1
0.67 0.34 0.34 0.04* N D ND 1 -94 ND
153.37 ND ND ND 0. IO* 0.08* 0.35 0.28 ND ND ND 0.30 1.09 0.23 0.37 2.44 1 .JO 7.1 1 ND
0. Id* 0.02* 0.18 0.02.
acetaldehyde ethyl acetate methanol ethanol 624. 16 627.13 632.23 ND, not detected. +Iimit of detection.
Month 5
CONDlTION A B C
a-pinene ethyl-butyrate hexanal sabinene 3-carene a-phellandrene ernyrcene heptanal limonene 2&3-rnethy i-butanol 2-hexend ?-hexan01 y-terpinene 0. IS* 0. IS* 0.12 octanal I - hexanol 3-hexen- 1-01 nonanal dimethy f-styrene hrfiJmI decanal linalool octanol terpinene-l-01 hydroxy-ethy I-hexanoate a-terpineol valencene geranial dk carvone
vinyl guiaicol hydroxyrnethyl furfirral vanillin Headspace-GC W m L ) (uwm L) (ug/mL) acetaldehyde 11.10 9.62 9.66 ethyl acetate methanol ethanol 648.96 63025 6 19.02 ND, not detected. *limit of detection.
Month 6
GC-MS (uglm L) (u%mL) ( W m L ) methy I-butyrate ND ND ND a-pinene ethy I-butyrate hexanal sabinene 3-carene a-phellandrene P-myrcene heptanal t imonene 2&3-methy l-butanol 2-hexenal 2-hexano l y-terpinene octanal I-hexanol 3-hexen-l-01 nonanal dimethy I-styrene furfirral decanal IinaIool octanol terpinene-4-o l hydroxy-ethyl-hexanoate a-terpineol valencene geranial& carvone perillaldehyde 2-2-decadienal vinyl guiaicol hydroxyrnethyl firfirat vanillin Heads paceGC (~g/mL) ( ~ % m L) (WvmL) acetaldehyde 10.77 10.28 10.33 ethyl acetate methanol ethanol 656.60 628.47 639.06 ND, not detected. *limit of detection.
Recommended