You are on page 1of 12

ARTICLE IN PRESS

International Dairy Journal 16 (2006) 750–761


www.elsevier.com/locate/idairyj

Changes in flavour volatiles of whole pasteurized milk as affected by


packaging material and storage time
Andreas E. Karatapanis, Anastasia V. Badeka, Kyriakos A. Riganakos,
Ioannis N. Savvaidis, Michael G. Kontominas
Laboratory of Food Chemistry and Technology, Department of Chemistry, University of Ioannina, Ioannina 45110, Greece
Received 10 November 2004; accepted 23 June 2005

Abstract

Changes in volatile profiles of whole pasteurized milk stored under fluorescent light at 4 1C, packaged in different containers were
monitored for a period of 7 days in a study designed to differentiate between light-induced oxidative and purely autoxidative effects
related to packaging material. Changes in volatiles were measured using solid-phase microextraction–gas chromatography/mass
spectroscopy, while microbiological and sensory analyses were used to assess milk quality. Two distinct patterns of milk flavour
deterioration were observed. In light-exposed samples, a light-induced oxidation mechanism prevailed while in light-protected
samples, an autoxidation mechanism was apparent. Under both conditions, the concentration of selected odorous compounds
increased with storage time. Microbiological data correlated poorly with both sensory and GC/MS data. Sensory data correlated
well with selected volatile compounds pointing to dimethyl disulphide, pentanal, hexanal and heptanal as potential markers of fresh
milk quality. Based on sensory analysis, the optimal shelf life of the whole pasteurized milk used in this study was approximately 5
days.
r 2005 Elsevier Ltd. All rights reserved.

Keywords: Milk flavour; Packaging material; Storage time

1. Introduction described as rancid, butyric, bitter flavour, (d) ‘‘micro-


bial flavour’’, described as acidic, bitter, fruity, malty,
Determination of milk off-flavours has been the putrid unclean flavour, (e) ‘‘oxidized flavour’’, described
subject of numerous investigations (Badings & Neeter, as papery, cardboardy, metallic, oily, fish flavour, (f)
1980; Contarini & Povolo, 2002; Kim & Morr, 1996; ‘‘transmitted flavour’’, described as weedy, cowy, barny
Marsili, 2003; Toso, Procida, & Stefanon, 2002; Urbach, flavour and (g) ‘‘miscellaneous flavour’’, described as
1990). According to the committee on off-flavour flat, chemical, foreign, lack of freshness, salty flavour.
nomenclature and reference standards of the American Modern techniques employed to isolate and concen-
Dairy Science Association (Shipe et al., 1978) off- trate odour active compounds from the volatile fraction
flavours in milk may be categorized as follows: (a) of milk products prior to gas chromatography/mass
‘‘heated flavour’’, described as cooked, caramelized and spectroscopy (GC/MS) analysis include: vacuum dis-
scorched flavour, (b) ‘‘light-induced flavour’’, described tillation, simultaneous steam distillation and extraction,
as burnt sunlight flavour, (c) ‘‘lipolyzed flavour’’, static headspace (HS), dynamic HS/purge and trap,
direct thermal desorption (Contarini & Povolo, 2002)
Corresponding author. Tel.: +302651098342; for solid and semi-solid samples and solid-phase
fax: +302651098795. microextraction (Jung, Yoon, Lee, & Min, 1998;
E-mail address: mkontomi@cc.uoi.gr (M.G. Kontominas). Marsili, 1999a; Toso et al., 2002). In addition to the

0958-6946/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.idairyj.2005.06.008
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 751

above techniques for the analysis of semi- and non- refrigerated storage. The six packaging materials in-
volatile flavour compounds, considerable work has been cluded: (1) a 0.5 L clear glass bottle, 800 mm in thickness,
done with HPLC, supercritical fluid extraction and (2) a 0.5 L clear PET bottle, 300–350 mm in thickness; (3)
supercritical fluid chromatography (Marsili, 2003). a 0.5 L pigmented [PET+2% TiO2] bottle, 300–350 mm
Packaging can directly prevent the development of in thickness; (4) a 0.5 L pigmented [HDPE+2% TiO2/
light-induced off-flavour by protecting the product from HDPE+4% carbon black pigment/HDPE+2% TiO2]
both light and oxygen (Schröder, 1982; Skibsted, 2000; 3-layer coextruded bottle, 550–600 mm in thickness; (5) a
Vassila, Badeka, Kondyli, Savvaidis, & Kontominas, 0.5 L pigmented [HDPE +2% TiO2] monolayer
2002). Apart from traditional glass bottles and coated bottle and (6) a 0.5 L coated paperboard carton
paperboard cartons, all-plastic containers have been consisting of polyethylene/board/polyethylene (20/395/
used for pasteurized milk packaging (Defosse, 2000; 35 mm). Freshly pasteurized milk stored in clear glass
Schröder, Scott, Bland, & Bishop, 1985). Problems with bottles (day 0 of storage) served as the control sample.
all-plastic containers used in the studies cited above The white (TiO2) pigmented bottles were chosen on the
included light transmission and oxygen permeability. It basis of consumer preference (White, 1985) with regard
should be noted, however, that oxidative reactions to colour and their potential protective effect against
(auto-oxidation) have been reported to take place in harmful incident light. The carbon black pigment has
milk packaged even in coated paperboard cartons which been suggested (Schröder, 1982) for maximum protec-
were found to be more or less impermeable to oxygen tion against UV–VIS radiation, with both pigments
(Rysstad, Ebbesey, & Eggestad, 1998; Schröder et al., being approved by the Food and Drug Administration,
1985). More recently, polyethylene perephthalate (PET) USA. Bottle (4) containing inner and outer white
and coextruded high-density polyethylene (HDPE) coloured layers was white in colour. Monolayer and 3-
bottles have been used for fresh milk packaging (Bakish layer HDPE and PET bottles were produced by Argo
& Hatfield, 1997; Cladman, Scheffer, Goodrich, & Ltd. (Athens, Greece) on an extrusion molding line
Griffiths, 1998; van Aardt, Duncan, Marcy, Long, & (Soplar SA, Switzerland) and an injection stretch blow
Hackney, 2001). PET has excellent mechanical proper- molding line (Nissei ASB, USA) respectively. Paper-
ties, a good barrier to oxygen and reduces the adverse board carton blanks were supplied by Dodoni Ltd.
effects of light on milk quality in the form of pigmented (Ioannina, Greece) and were produced by Variopack
bottles. On the other hand, pigmented HDPE bottles, (Homburg/Saar, Denmark).
both monolayer and multilayer, at a higher thickness
than current PET, are finding their way into the fresh
milk packaging market. Both provide excellent conve- 2.2. Sample preparation and handling
nience through easy opening and reclosing, thus
minimizing recontamination (Cladman et al., 1998). Homogenized whole fat milk (3.5% fat) was obtained
Many studies have reported off-flavour development from the dairy plant Dodoni Ltd. (Ioannina, Greece)
in milk at a given storage time, usually at the end of shelf immediately after pasteurization (75 1C, 15 s). The milk
life of the product (Contarini & Povolo, 2002; Cormier, was aseptically dispensed into sterilized bottles and
Raymond, Champagne, & Morin, 1991; Jenq, Bassette, coated paperboard cartons in the plant laboratory.
& Crang, 1988; Leong, Harte, Partridge, Ott, & Bottles were sealed using polypropylene screw caps
Downes, 1992; Marsili, 1999b; Urbach & Milne, 1987; while coated paperboard cartons were sealed on the
Vallejo–Cordoba & Nakai, 1994). Limited information production line sealer. HS for both cartons and bottles
exists in the literature on changes in flavour compounds was between 50 and 60 mL. The filled containers were
of milk as affected by storage time and packaging transported to the laboratory in polystyrene foam ice
conditions. Thus the objectives of the present study were boxes packed in ice within 30 min of the filling process.
(1) to monitor HS volatiles of whole pasteurized milk Bottles were stored at 470.5 1C for a period up to 7
stored under refrigeration and fluorescent light as a days under perpendicular fluorescent light provided by
function of storage time and type of packaging material one Philips 55 W ‘‘cool white’’ fluorescent tube in a
and (2) to possibly relate changes in volatile compounds commercial display floor type cabinet 70  200  70 cm3
to sensory and microbiological attributes of milk. (w  h  d) in dimensions bearing a glass door. The
fluorescent lamp produced 825750 lux on the side
surface of the containers, measured with a General
2. Materials and methods Electric light meter (Fairfield, CT, USA). Milk samples
(100 mL) were sampled from sealed containers at 1, 3, 5
2.1. Packaging materials and 7 days after initial packaging for physicochemical,
microbiological and sensory testing. Testing at day 0
Six different packaging materials were evaluated with was carried out on milk samples immediately after
respect to their potential effect on milk quality during packaging. On each sampling day, samples of the six
ARTICLE IN PRESS
752 A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761

different treatments were assayed with 3 replicates for were measured with a Shimadzu model 210 UV–VIS
each treatment (n ¼ 3). recording spectrophotometer (Rysstad et al., 1998).

2.3. SPME– GC/MS analysis 2.5.2. Determination of oxygen transmission rates


Oxygen transmission rates for all packages (bottles
Milk samples (8 mL), along with 10 mL of an internal and cartons) were measured using the Oxtran 2/20
standard solution (20 mg mL1 4-methyl-2-pentanone) oxygen permeability tester (Mocon Controls, USA)
and a micro-stirring bar were placed in a 20 mL glass at a relative humidity (RH) of 60%, temperature of
serum vial (20  72 mm) which was sealed with an 22 1C (ASTM D-3985) and were expressed as
aluminum crimp cap provided with a needle-pierceable mL m2 day1 atm1.
polytetrafluroethylene/silicone septum. Solid-phase mi-
croextraction (SPME) was performed with a 75-mm
Carboxen/Polydimethylsiloxane (PDMS) fibre mounted 2.5.3. pO2 measurements
to a SPME manual holder assembly (Supelco, Belle- Dissolved O2 in the milk sample (pO2) was carried out
fonte, USA). The sample vial was placed in a 50 1C using a microprocessor oximeter (WTW, OXI 96,
water bath and stirred at high speed. After allowing Weilheim, Germany) via direct emersion of the oximeter
5 min for the sample to equilibrate at 50 1C, the needle of probe into the milk sample.
the SPME device was inserted into the vial through the
vial septum, and the plunger of the SPME apparatus 2.6. Microbiological tests
was pushed down to expose the Carboxen/PDMS fibre
to the vial HS. After 15 min of exposure time with To monitor possible microbial post-pasteurization
constant stirring, the fibre was retracted into the needle contamination of milk in the various containers,
assembly and removed from the vial. standard plate counts were determined according to
the International Dairy Federation (IDF) standard No.
2.4. GC/MS analysis conditions 100 B (IDF, 1991a). Psychrotrophic counts were
determined according to the IDF standard No. 101 A
GC analysis of volatile compounds adsorbed onto the (IDF, 1991b).
SPME fibre was carried out on a 60-m  0.25-mm-
internal diameter DB-5 (5% phenyl–94% dimethyl
arylene siloxane) fused-silica capillary column with a 2.7. Sensory evaluation
film thickness of 1 mm (J&W Scientific, Folsom, USA).
Flow rate of the helium carrier gas was 0.8 mL min1. A A panel of 17 individuals consisting of faculty and
Hewlett-Packard 6890 series gas chromatograph graduate students were trained, in two sessions, to
equipped with a Hewlett-Packard 5973 mass selective differentiate between various off-flavours in milk
detector was used. During the injection phase, the samples by preparing and evaluating common off-
injector was operated in the split mode (2:1 split ratio) at flavour reference standards (light oxidized, plastic-
a temperature of 250 1C. For thermal desorption, the acetaldehyde and lack of freshness) (van Aardt et al.,
SMPE fibre remained in the injector for 5 min. The 2001). Sensory data were collected between days 0 and 7
initial column temperature was 40 1C. The column was of storage. Milk samples (30 mL, 1474 1C) were
maintained at 40 1C for 10 min, heated to 140 1C at a presented to panelists in individual sensory booths as
rate of 5 1C min1, heated to 220 1C at a rate of sets of three or four with a resting interval of 1 or 2 min
8 1C min1 and held at 220 1C for 5 min. MS conditions between sample sets. Panelists rated each sample for
were as follows: source 230 1C; acquisition performed in intensity and type of off-flavour on a scale between 0
electron impact (EI) mode (70 eV) by 2.92 scans s1 and and 5, where a score of 5 corresponded to very good
the mass range used was mass/charge (m/z): 33–300. The flavour milk and a score of 0 corresponded to milk unfit
temperature of the transfer line was held constant at for consumption due to very strong off-flavours (IDF,
220 1C. Peak identification was performed by compar- 1987).
ison of retention times and mass spectra of eluting
compounds to those of the Wiley library (Wiley275, J. 2.8. Statistical analysis
Wiley & Sons Ltd., West Sussex, England).
Data were subjected to analysis of variance using the
2.5. Physical tests Excel 97 software program (Microsoft, CA, USA).
Where statistical differences were noted, differences
2.5.1. Light transmission measurements among packages were determined, using the least
Spectral transmission characteristics over the wave- significant difference (LSD) test. Significance was
length range of 350–780 nm for all packaging materials defined at Po0.05.
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 753

3. Results and discussion compounds, namely dimethyl disulphide and heptanal


were produced from day 1 of storage, while 2,3-
3.1. Volatile compound determination octanedione, 2-heptanone, octanal and nonanal were
formed toward the end of the product shelf life.
Volatiles produced in milk as a function of storage Carbonyl compounds are the secondary products of
time and packaging material are shown in Tables 1–6. lipid photoxidation while dimethyl disulphide is both a
Given that the detector response factor for each product of light exposure and a microbial metabolite
compound was not determined, the amounts of volatiles (Jung et al., 1998). Of the above compounds, 2,2,4-
reported in Tables 1–6 are only semi-quantitative and trimethyl pentane, dimethyl disulphide, pentanal, hex-
will be therefore used to show only increasing/decreas- anal and heptanal increased proportionally with time of
ing trends for each compound. storage.
Freshly pasteurized milk (Table 1, day 0) in contact A similar volatile profile was observed for milk stored
with glass resulted in the measurement of 19 com- in clear and TiO2 pigmented PET bottles (Tables 2 and
pounds, including hydrocarbons (7), ketones (4), esters 3). This was expected as the mechanism of milk
(3) alcohols (2), aldehydes (2) and sulphur compounds deterioration in the above packaging materials is
(1). A similar profile of volatiles has been reported for photo-oxidation. Photo-oxidation occurs in milk in the
low fat cows’ milk at day 0 of storage by Marsili and presence of a photosensitizer such as riboflavin which
Miller (2001). excites oxygen to its singlet state, a requirement for
As shown in Table 1, in addition to the compounds oxygen to react with the unsaturated lipid moiety to
identified in freshly pasteurized milk, a number of produce a hydroperoxide (Skibsted, 2000). The TiO2

Table 1
Volatile compound profile (mg mL1) of whole pasteurized milk packaged in clear glass as a function of storage time at 4 1C

Volatile compound tra (min) Quality factorb Days of storage Trendc

0 1 3 5 7
d e
Ethanol 6.53 72 0.69 0.72 0.84 0.63 1.00
Acetone 7.52 86 117.27 130.44 129.64 112.32 136.01
Dimethyl sulphide 8.72 97 1.52 1.57 1.53 1.13 1.79
1-Propanol 10.43 80 0.61 0.73 0.68 0.40 0.76
3-Methyl pentane 11.91 91 0.44 0.44 0.56 0.61 0.93
2-Butanone 12.63 86 17.85 21.32 20.26 18.05 20.97
Ethyl acetate 13.80 86 0.73 0.73 0.64 0.60 0.82
Chloroform 14.13 96 1.09 1.06 1.36 1.63 0.88
Propanoic acid ethyl ester 14.87 86 7.76 4.46 5.01 4.73 3.94
3-Methyl-2-butanone 16.59 72 0.46 0.55 0.64 0.48 0.53
Benzene 16.97 91 2.16 2.89 3.87 3.73 3.94
2-Pentanone 18.19 70 1.78 2.38 2.00 1.81 2.15
2,2,4-Trimethyl pentane 18.41 83 4.31 6.49 15.36 36.03 41.37 mm
Pentanal 18.78 80 2.10 2.13 2.63 4.31 7.02 m
Butyric acid methyl ester 20.27 95 2.70 3.94 3.10 2.73 2.90
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 UD 0.30 0.43 0.58 0.89 m
Toluene 23.01 94 0.74 0.81 2.59 2.61 2.91
Hexanal 24.35 90 0.28 0.73 1.67 4.27 5.55 mm
1-Hexanol 27.46 72 UD UD UD UD 0.25
Ethyl benzene 27.67 94 0.26 0.54 0.47 0.41 0.86
1,4-Dimethyl benzene 28.00 97 0.51 0.92 0.72 0.68 1.00
2-Heptanone 28.52 87 UD UD UD UD 0.18
Heptanal 29.03 88 UD 0.14 0.21 0.44 0.80 m
2-Heptenal 31.31 75 UD UD UD UD 0.17
2,3-Octanedione 32.10 71 UD UD UD 0.18 0.49
Octanal 32.92 73 UD UD UD UD 0.16
Nonanal 36.01 72 UD UD UD 0.10 0.18

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard, UD ¼ under detection limit.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).
ARTICLE IN PRESS
754 A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761

Table 2
Volatile compound (mg mL1) profile of whole pasteurized milk packaged in clear polyethylene terephthalate (PET) as a function of storage time at
4 1C

Volatile compound tra (min) Quality factorb Days of storage Trendc

1 3 5 7
d e
Ethanol 6.53 72 0.77 0.82 0.76 0.51
Acetone 7.52 86 114.16 133.66 123.71 121.98
Dimethyl sulphide 8.72 97 1.49 1.61 1.34 1.40
1-Propanol 10.43 80 0.44 0.72 0.35 0.46
3-Methyl pentane 11.91 91 0.56 0.53 0.64 0.84
2-Butanone 12.63 86 18.75 19.89 19.72 20.31
Ethyl acetate 13.80 86 0.74 0.67 0.65 0.93
Chloroform 14.13 96 1.04 1.46 3.31 1.42
Propanoic acid ethyl ester 14.87 86 4.14 4.81 3.59 4.29
3-Methyl-2-butanone 16.59 72 0.59 0.50 0.60 0.51
Benzene 16.97 91 3.79 3.82 3.85 3.98
2-Pentanone 18.19 70 2.02 1.99 1.77 2.07
2,2,4-Trimethyl pentane 18.41 83 7.77 17.88 31.41 40.16 m
Pentanal 18.78 80 2.17 2.74 6.43 8.41 m
Butyric acid methyl ester 20.27 95 3.53 2.25 2.44 2.38
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 0.33 0.54 0.77 1.25 m
Toluene 23.01 94 1.17 1.65 2.36 2.41
Hexanal 24.35 90 0.79 1.84 4.91 10.67 mm
1-Hexanol 27.46 72 UD UD UD 0.24
Ethyl benzene 27.67 94 0.40 0.54 0.31 0.35
1,4-Dimethyl benzene 28.00 97 0.73 1.06 0.22 0.33
2-Heptanone 28.52 87 UD UD 0.15 0.23
Heptanal 29.03 80 0.13 0.23 0.58 1.79 mm
2-Heptenal 31.31 75 UD UD UD 0.21
2,3-Octanedione 32.10 71 UD UD 0.39 1.01
Octanal 32.92 73 UD UD UD 0.19
Nonanal 36.01 72 UD UD 0.09 0.23

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard, UD ¼ under detection limit.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).

pigmentation in the white coloured bottle only partly not detected in the present study. Furthermore, neither
protects the product from photo-oxidation resulting acetaldehyde, a degradation product of PET, nor its
in lower amounts of pentanal, dimethyl disulphide, oxidation product, acetic acid, was identified in milk
hexanal and heptanal, as well as 2-heptanone, octanal, packaged in the PET bottles. This is in contrast to the
and nonanal after day 7 of storage. This is in agree- findings of van Aardt et al. (2001), Nijssen, Kamper-
ment with the findings of de Man (1978) who reported man, and Jetten (1996) and Bosset, Gallmann, and
that inorganic fillers (TiO2) used to block harmful Sieber (1994). The reason for this is unclear. Cadwalla-
light wavelengths were not effective in preventing der and Howard (1998) attributed off-flavours at the
light-induced deterioration of milk. This may be end of milk shelf life to compounds such as 2,3-
due to the fact that filler particles may act as butanedione. In the present study, a similar compound
strong scattering centers causing a large proportion of in structure (2,3-octanedione) was identified on day 5 of
incident light to find its way into the milk at scattered milk storage (Tables 1–3). Marsili (1999a) used SPME
angles. for isolation of milk volatiles and reported a linear
The present results are in agreement with earlier relationship between hexanal formation in 2% cows’
reports in the literature (Cadwallader & Howard, 1998; milk packaged in half gallon HDPE jugs exposed to
Cladman et al., 1998) that showed the most common fluorescent light for 2 days. A plot of increasing
volatile compounds in light-oxidized milk to be dimethyl pentanal concentration with increasing time of exposure
disulphide, pentanal, hexanal, acetaldehyde, 1-hexen-3- more accurately fitted a second-order polynomial
one and 1-octen-3-one. The latter two compounds were equation.
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 755

Table 3
Volatile compound profile (mg mL1) of whole pasteurized milk packaged in monolayer pigmented polyethylene terephthalate (PET) as a function of
storage time at 4 1C

Volatile compound tra (min) Quality factorb PET days of storage Trendc

1 3 5 7
d e
Ethanol 6.53 72 0.91 0.76 0.88 0.85
Acetone 7.52 86 131.65 123.16 134.32 123.42
Dimethyl sulphide 8.72 97 1.66 1.40 1.49 1.44
1-Propanol 10.43 80 0.77 0.70 0.62 0.73
3-Methyl-pentane 11.91 91 0.38 0.60 0.51 0.33
2-Butanone 12.63 86 21.26 19.66 19.48 19.50
Ethyl acetate 13.80 86 0.82 0.72 0.65 0.69
Chloroform 14.13 96 1.10 1.48 1.87 1.61
Propanoic acid ethyl ester 14.87 86 4.25 4.33 3.88 3.34
3-Methyl-2-butanone 16.59 72 0.69 0.58 0.53 0.57
Benzene 16.97 91 3.49 3.88 3.28 2.81
2-Pentanone 18.19 72 2.49 1.92 1.85 1.86
2,2,4-Trimethyl-pentane 18.41 83 11.79 16.61 35.41 41.86 m
Pentanal 18.78 80 2.05 2.61 4.08 6.43 m
Butyric acid methyl ester 20.27 95 3.65 2.36 2.53 2.05
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 0.30 0.36 0.45 0.78 m
Toluene 23.01 94 1.02 1.84 2.06 1.89
Hexanal 24.35 90 0.71 1.54 3.96 5.14 mm
1-Hexanol 27.46 72 UD UD UD 0.26
Ethyl benzene 27.67 94 0.30 0.42 0.29 0.29
1,4-Dimethyl benzene 28.00 97 0.37 0.52 0.40 0.40
2-Heptanone 28.52 87 UD UD UD 0.17
Heptanal 29.03 80 UD 0.17 0.36 0.74 m
2-Heptenal 31.31 75 UD UD UD 0.15
2,3-Octanedione 32.10 71 UD UD 0.25 0.62
Octanal 32.92 73 UD UD UD 0.14
Nonanal 36.01 72 UD UD UD 0.15

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard, UD ¼ under detection limit.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).

The present results differ significantly from those of milk samples packaged in carton packs were tentatively
Urbach and Milne (1987) who found major differences identified (Tables 4–6). Milk samples (4–6) in light-
in the concentration of ethanol, acetaldehyde, propan-2- protected containers had significantly lower levels of
ol and acetone between freshly pasteurized milk and oxidation than those packaged in the light-exposed
milk held for 7 days at 7 1C. In the present study, containers (1–3).
acetone concentrations remained quite constant An important observation was that dimethyl disul-
throughout the 7-day storage period while the ethanol phide, a product mainly of milk light exposure was not
concentration increased on day 7 of storage and in O2 detected (Table 4), indicative of the protective effect of
permeable packages (Tables 4–6). Acetaldehyde and the multilayered HDPE bottle provided by the carbon
propan-2-ol were not detected. Finally, Jenq et al. (1988) black pigment. The same compound was formed in
reported that both n-hexanal and n-pentanal were the small amounts in pigmented HDPE bottles (Table 5)
compounds in milk most affected by fluorescent light and even lower amounts in carton packs (Table 6). The
(51 h exposure) and added copper. carton pack provided a somewhat better protection of
A different profile of volatiles produced during milk milk than the pigmented HDPE bottle from incident
storage was observed (formation of non-oxidation light. This is supported by light transmittance data of
products) in light protected containers; 33 compounds packaging materials (Fig. 1). Another important ob-
in milk samples packaged in 3-layer pigmented HDPE servation (Table 4) was that carbonyl compounds,
bottles, 32 compounds in milk samples packaged in formed as secondary autoxidation products such
TiO2 pigmented HDPE bottles and 36 compounds in as pentanal, hexanal, 2-heptanone, 2-pentanone and
ARTICLE IN PRESS
756 A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761

Table 4
Volatile compound profile (mg mL1) of whole pasteurized milk packaged in 3-layer pigmented high-density polyethylene (HDPE) as a function of
storage time at 4 1C

Volatile compound tra (min) Quality factorb Days of storage Trendc

1 3 5 7

Ethanold
6.53 72 0.83 e
0.87 1.04 13.20 mm
Acetone 7.52 86 126.74 133.87 132.76 146.91
Dimethyl sulphide 8.72 97 1.66 1.60 2.19 19.33 mm
Methyl acetate 9.04 70 0.46 0.19 0.53 0.83 m
Dichloro-methane 9.25 97 UD UD 7.52 18.67
Carbon disulfide 9.66 83 UD 1.29 1.16 1.62
2-Methyl-propanol 10.48 70 0.75 1.03 1.75 1.81
3-Methyl-pentane 11.91 91 0.94 0.80 0.61 1.29
2-Butanone 12.63 87 20.44 26.63 21.39 21.31
Ethyl acetate 13.80 86 0.89 0.91 0.73 3.61 mm
Chloroform 14.13 96 0.89 1.35 3.70 4.42 m
Propanoic acid ethyl ester 14.87 86 4.64 9.05 15.21 12.68
3-Methyl-butanal 16.30 97 UD UD UD 1.68
3-Methyl-2-butanone 16.59 72 0.55 0.67 0.66 0.76
Benzene 16.97 91 4.54 4.71 4.03 4.88
2-Pentanone 18.19 72 2.16 2.55 1.93 2.49
2,2,4-Trimethyl-pentane 18.41 83 9.61 39.53 148.28 23.20 mmkk
Pentanal 18.78 80 1.85 2.28 2.33 2.91
Butyric acid methyl ester 20.27 97 2.96 3.01 2.47 6.17 m
3-Methyl-1-butanol 20.86 78 UD UD UD 1.24
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 UD UD UD UD
Toluene 23.01 94 1.25 1.92 2.47 4.24 m
Hexanal 24.32 90 0.48 0.54 0.80 UD
Butyric acid ethyl ester 24.35 97 UD UD UD 6.57
Ethyl benzene 27.67 94 0.53 0.45 0.38 0.72
1,4-Dimethyl benzene 28.00 97 1.20 0.86 0.66 1.31
2-Heptanone 28.52 87 — 0.12 0.10 0.24
1,3-Dimethyl benzene 29.12 97 0.63 0.31 0.39 0.43
Hexanoic acid methyl ester 29.87 90 0.22 UD UD 1.52
Hexanoic acid ethyl ester 32.61 97 UD UD UD 2.83
Limonene 34.19 98 UD 0.10 UD 0.20
Octanoic acid methyl ester 36.40 75 UD UD UD 0.19
Nonanal 36.01 72 0.09 0.11 0.11 0.19

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard, UD ¼ under detection limit.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).

3-methyl-2-butanone, remained constant in concentra- increasing storage time observed for milk packaged in
tion equivalent to concentrations in freshly pasteurized the TiO2 pigmented HDPE bottle. This effect could be
milk (day 0 of storage). This was expected since photo- explained in terms of the partial protection of milk from
oxidation did not progress to any significant extent light provided by this packaging material (Fig. 1).
owing to the excellent protection provided by the carbon Excess production of ethanol in milk packaged in PE
black pigment. Christy, Amantea, and Irwin (1981) bottles and carton packs (Tables 4–6) was responsible
demonstrated that co-extruded black/white polyethylene for the formation of large amounts of esters such as
totally blocked diffuse transmittance of 400–500 nm to butyric acid methyl ester, butyric acid ethyl ester,
less than 5% thus protecting milk from off-flavour hexanoic acid methyl ester, hexanoic acid ethyl ester
development. The same observations hold more or less and octanoic acid methyl ester. Such esters were not
for the other two samples packaged in TiO2 pigmented found in milk packaged in glass and PET bottles. This
HDPE and paperboard carton containers (Tables 5 route of ester formation has been suggested by Hosono,
and 6). A partial exception to this trend was the slight Elliot, and McGugan (1974) and is related to growth of
increase in pentanal and hexanal concentrations with Pseudomonas spp., P. fragi in particular. Toluene may
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 757

Table 5
Volatile compound profile (mg mL1) of whole pasteurized milk packaged in monolayer pigmented high-density polyethylene (HDPE) as a function
of storage time at 4 1C

Volatile compound tra (min) Quality factorb Days of storage Trendc

1 3 5 7

Ethanold
6.53 72 0.75 e
0.69 0.46 10.62 mm
Acetone 7.52 86 126.12 127.75 114.58 154.55
Dimethyl sulphide 8.72 97 1.61 1.60 1.26 37.17 mm
Methyl acetate 9.04 70 0.45 0.58 0.37 1.53 m
Dichloro-methane 9.25 97 UD UD 2.08 1.85
Carbon disulfide 9.66 83 UD 0.53 0.27 0.62
2-Methyl-propanol 10.48 70 0.74 0.67 0.52 0.51
3-Methyl-pentane 11.91 91 0.62 0.60 0.39 0.49
2-Butanone 12.63 87 20.96 20.58 18.23 16.08
Ethyl acetate 13.80 86 0.89 0.67 0.58 5.61 mm
Chloroform 14.13 96 1.04 0.98 2.89 1.26
Propanoic acid ethyl ester 14.87 86 4.36 4.81 3.81 4.75
3-Methyl-butanal 16.30 97 UD UD UD 0.57
3-Methyl-2-butanone 16.59 72 0.63 0.63 0.57 0.63
Benzene 16.97 91 4.13 4.15 3.08 4.20
2-Pentanone 18.19 72 2.26 2.05 1.54 2.74
2,2,4-Trimethyl-pentane 18.41 83 11.17 19.73 146.92 2.36 mmkk
Pentanal 18.78 80 2.06 2.93 3.42 4.06 m
Butyric acid methyl ester 20.27 97 2.53 2.17 2.40 9.19 m
3-Methyl-1-butanol 20.86 78 UD UD UD 3.06
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 UD 0.34 0.42 0.60
Toluene 23.01 94 1.40 2.53 1.74 2.15
Hexanal 24.32 90 0.58 0.87 1.45 —
Butyric acid ethyl ester 24.35 97 UD UD UD 2.33
Ethyl benzene 27.67 94 0.46 0.34 0.56 0.84
1,4-Dimethyl benzene 28.00 97 0.43 0.60 1.59 1.86 m
2-Heptanone 28.52 87 UD UD UD 0.38
1,3-Dimethyl benzene 29.12 97 0.43 0.23 0.49 0.76
Hexanoic acid methyl ester 29.87 90 0.32 UD UD 3.21
Hexanoic acid ethyl ester 32.61 97 UD UD UD 1.26
Octanoic acid methyl ester 36.40 75 UD UD UD 0.43
Nonanal 36.01 72 0.09 0.12 0.13 0.28

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard, UD ¼ under detection limit.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).

originate from the breakdown of b-carotene as well as statistically significant differences (P40:05) in mesophi-
other sources (Marsili & Miller, 1998) while chloroform lic counts were recorded for milk samples in all
may be from chlorinated sanitizers used to wash teats packaging materials for a given sampling day during
prior to milking. The formation of 3-methyl butanal is the entire 7-day storage period (data not shown). After 7
believed to be through the action of bacterial enzymes days for storage, mesophilic counts were between 6.00
on leucine (Marsili & Miller, 1998). and 6.41 log cfu mL1, values close to the upper limit
(106–107 cfu mL1) for keeping quality of milk (IDF,
3.2. Microbiological analysis 1986).

3.2.1. Mesophilic counts 3.2.2. Psychrotrophic counts


Initial (day 0) counts were high, 4.36 log cfu mL1, Initial (day 0) counts were 3.43 log cfu mL1. No
indicative of the current poor milk collection practices statistically significant (P40:05) differences in psychro-
throughout Greece. Such high microbial loads are trophic counts were recorded for milk samples in all
responsible for the short shelf life of domestic refriger- packaging materials up to day 3 of storage
ated milk, which does not normally exceed 5 days. No (3.67 log cfu mL1). Statistically significant (Po0:05)
ARTICLE IN PRESS
758 A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761

Table 6
Volatile compound profile (mg mL1) of whole pasteurized milk packaged in paperboard carton as a function of storage time at 4 1C

Volatile compound tra (min) Qulaity factorb Days of storage Trendc

1 3 5 7

Ethanold 6.53 72 0.59e 0.79 0.45 3.98 mm


Acetone 7.52 86 128.61 132.97 134.51 121.34
Dimethyl sulphide 8.72 97 1.60 1.42 1.16 13.57 mm
Methyl acetate 9.04 70 0.53 0.49 0.42 1.69 m
Dichloro-methane 9.25 97 — — 18.27 21.14
Carbon disulfide 9.66 83 0.61 1.81 2.20 2.81
2-Methyl-propanol 10.48 70 0.73 0.81 0.65 1.16
3-Methyl-pentane 11.91 91 0.94 0.51 0.67 0.71
2-Butanone 12.63 87 20.90 20.27 20.00 15.68
Ethyl acetate 13.80 86 0.46 0.55 0.74 44.06 mm
Chloroform 14.13 96 2.50 3.23 7.56 6.38
Propanoic acid ethyl ester 14.87 86 6.46 10.51 11.24 14.81 m
3-Methyl-butanal 16.30 97 — — — 2.08
3-Methyl-2-butanone 16.59 72 0.57 0.56 0.46 0.84
Benzene 16.97 91 4.31 4.11 3.46 5.03
2-Pentanone 18.19 72 2.53 2.08 1.69 2.64
2,2,4-Trimethyl-pentane 18.41 83 5.55 17.14 37.75 17.57 mmkk
Pentanal 18.78 80 1.80 2.31 3.12 2.61
Propanoic acid methyl ester 19.76 72 — — 0.13 1.84
Butyric acid methyl ester 20.27 97 4.24 3.09 3.06 6.25
3-Methyl-1-butanol 20.86 78 — — — 2.96
4-Methyl-2-pentanone (i.s.) 21.25 91 25 25 25 25 i.s.
Dimethyl disulphide 22.59 90 — — 0.14 0.23
Toluene 23.01 94 1.66 2.59 5.34 4.57
Hexanal 24.32 90 0.60 0.61 1.12 —
Butyric acid ethyl ester 24.35 97 — — — 10.74
Ethyl benzene 27.67 94 2.08 1.08 1.21 1.87
1,4-Dimethyl benzene 28.00 97 7.00 4.15 4.01 6.04
2-Heptanone 28.52 87 — — — 0.54
1,3-Dimethyl benzene 29.12 97 3.28 3.38 1.97 2.55
Hexanoic acid methyl ester 29.87 90 0.41 0.19 0.14 1.44
a-Pinene 30.95 97 0.70 0.60 0.67 1.38
Hexanoic acid ethyl ester 32.61 97 — — 0.12 3.80
1,3,5-Trimethyl benzene 33.04 93 — — — 0.32
Delta.3carene 33.65 97 0.38 0.32 0.41 0.76
Limonene 34.19 98 — — 0.09 0.52
Octanoic acid methyl ester 36.40 75 — — — 0.22

m ¼ slight increase, k ¼ slight decrease, mm ¼ large increase, kk ¼ large decrease, i.s. ¼ internal standard.
a
Retention time.
b
Quality factor ¼ % matching of the experimental mass spectra against those found in the Wiley database.
c
Trends are designated as follows:
d
Tentatively identified on the basis of the Wiley275 library (J. Wiley & Sons Ltd., West Sussex, England).
e
Values are the mean of three determinations (n ¼ 3).

differences among packages were recorded during days 5 milk sample in the multilayer (white/black/white)
and 7 of storage (data not shown). After 7 days of HDPE bottle received a score of 4.1 (best protection)
storage, relatively high psychrotrophic counts between followed by the sample in the paperboard carton (score
5.31 and 6.10 log cfu mL1 were recorded for all packa- 3.6), the sample in the monolayer pigmented HDPE
ging materials. bottle (score 3.4) and the sample in the clear glass bottle
(score 3.0), with the samples in clear and pigmented PET
3.3. Sensory evaluation scoring 2.5 and 2.7, respectively. Flavour defects were
described as ‘‘slightly stale’’ to ‘‘stale’’ and ‘‘fruity’’ for
Flavour evaluation results are shown in Table 7. In samples packaged in paperboard carton and HDPE and
general, off-flavours increased more rapidly in milk ‘‘plastic’’, ‘‘oxidized’’ to ‘‘burnt taste’’ for samples
packaged in light-exposed containers compared with packaged in PET and glass bottles. After 7 days of
light-protected containers. After 5 days of storage, the storage, the milk samples in all containers were
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 759

unacceptable. The present sensory results are in general 3.4. Relationship between sensory, microbiological and
agreement with those of van Aardt et al. (2001) who volatile compound data
reported better milk keeping quality in amber PET and
UV compounded PET compared with clear PET bottles There was a rather poor relationship between sensory
as well as the results of Cladman et al. (1998) who and microbiological data with respect to both total
reported superior milk protection against light oxidation mesophile and psychrotroph counts. As previously
when milk was packaged in green PET in comparison documented (Bishop & White, 1986; Marsili, 2000;
with clear PET bottles. Even though the total microbial Urbach & Milne, 1987), spoilage of a specific food
count (106–107 cfu mL1) was in all cases below or equal substrate is the result of growth of specific spoilage
to the approximate upper limit for acceptability after 7 organisms and not of total microflora. Furthermore,
days of storage, the sensory characteristics of the there was a poor relationship between microbiological
product had clearly deteriorated resulting in an unac- and volatile compound data since for a large number of
ceptable product. This has previously been noted by volatiles there was no change in concentration with
Labuza (1982) and Maxcy and Wallen (1983) who increasing microbial populations (i.e. between days 5
reported that the most reliable method to judge the and 7 of storage, Tables 1–6).
quality of fresh milk is sensory evaluation. With regard to the relationship between sensory and
volatile compound data several observations can be
made. In containers where light-induced changes pre-
dominate (containers 1, 2 and 3) there was a gradual
100 increase in the concentration of pentanal, hexanal,
90 1 heptanal, dimethyl disulphide and 2,2,4-trimethyl
80 2
70 Table 8
Oxygen transmission rate of packaging materials
60
Packaging materiala O2 transmission rateb
%T

50
(mL m2 d1 atm1)
40
Clear glass bottle 0
30 Clear PET bottle 15.471.1a
3
20 Pigmented PET bottle 13.271.2a
3-Layer pigmented, coextruded 42.272.4b
10 5 HDPE bottle
6 Monolayer pigmented HDPE 44.472.5b
0 4
300 350 400 450 500 550 600 650 700 750 800 bottle
λ (nm) Coated paperboard carton 824742c

Fig. 1. Spectral transmission curves of the milk packaging materials. a


HDPE ¼ high-density polyethylene, PET ¼ polyethylene ter-
(1) Clear glass, (2) clear polyethylene terephthalate (PET), (3) ephthalate, temperature ¼ 22 1C, relative humidity ¼ 60%.
pigmented PET, (4) 3-layer pigmented high-density polyethylene b
Values within a column followed by different superscript letters are
(HDPE), (5) monolayer pigmented HDPE and (6) coated paperboard significantly different (Po0:05). Values reported are the mean of six
carton (T ¼ transmittance, l ¼ wavelength). replicates 7SD.

Table 7
Flavour evaluation of whole pasteurized milk packaged in various containers during storage at 4 1C

Packaging materiala Scoreb/comments


Days of storage at 4 1C

0 1 3 5 7

Clear glass bottle 5a,* 5a 3.5c/Slightly stale 3.0c/Slight oxidized flavour 1.8b/Oxidized flavour
Clear PET bottle 5a 4.4b 3.3b/Plastic taste 2.5c/Plastic taste, oxidized flavour 1.1b/Plastic taste, oxidized flavour
Pigmented PET bottle 5a 4.5b 3.5c 2.7c/Plastic taste, oxidized flavour 1.2b/Plastic taste, oxidized flavour
3-Layer pigmented, coextruded HDPE bottle 5a 5a 5a 4.1a 2.9a/Stale
Monolayer pigmented HDPE bottle 5a 5a 4.3b 3.4b/Slightly stale 2.9a/Stale
Coated paperboard carton 5a 5a 4.6b 3.6c/Slightly stale 3.1a/Stale, fruity
a
HDPE ¼ high-density polyethylene, PET ¼ polyethylene terephthalate. Values within a column followed by different superscript letters are
significantly different (Po0:05), asterix denotes mean of three determinations (n ¼ 3).
b
Numerical scale of scoring: very good ¼ 5, good ¼ 4, fair ¼ 3, poor ¼ 2, very poor ¼ 1, unfit for consumption ¼ 0.
ARTICLE IN PRESS
760 A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761

Table 9
Concentration of dissolved oxygen (mg mL1) in whole pasteurized milk as a function of packaging material and storage time

Packaging materiala Days of storageb

1 3 5 7

Clear glass 9.170.5a 8.370.5a 5.870.3a 4.370.3a


Clear PET 7.470.4b 6.370.4b 5.570.5a 4.570.2a
Pigmented PET 7.370.3b 6.970.6b 5.470.4a 3.770.2b
3-Layer pigmented HDPE 8.370.4c 8.070.4a 7.570.3b 6.670.4c
Pigmented HDPE 7.570.3b 7.470.5b 6.170.4a 5.670.4d
Coated paperboard carton 9.470.6a 8.770.6a 8.070.5b 6.670.3c
a
HDPE ¼ high-density polyethylene, PET ¼ polyethylene terephthalate.
b
Values within a column followed by different superscript letters are significantly different (Po0:05). Values reported are the mean of six replicates
7SD.

pentane with storage time. The accumulation of these (50–60 mL). Schröder (1982) reported that even a small
compounds seems to correlate rather well with decreas- HS in a container provides a considerable O2 reservoir
ing flavour scores of milk (Table 7). Beginning with day relative to the quantity dissolved in milk. In a container
3 of storage, off-flavour notes such as ‘‘slightly stale’’ with a 15% HS (a common proportion for a commercial
and ‘‘slightly plastic’’ were recorded in milk packaged in milk carton), the ratio between dissolved and gaseous
containers 1, 2 and 3. O2 in milk at 10 1C is approximately 1:5. Thus the
In containers that provided good protection against contribution of dissolved oxygen in milk (Table 9) to
light (containers 4, 5 and 6) there was a significant oxidation phenomena is rather insignificant.
increase in the concentration of ethanol, dimethyl No single compound is a dependable marker of the
sulphide, ethyl acetate and a smaller increase in keeping quality of milk given the relatively poor
concentration of esters such as methyl acetate with relationship between the amount of compounds formed
increasing storage time between days 5 and 7. An (most aldehydes, sulphides, esters, hydrocarbons) and
interesting observation was the formation of a series of storage time under fluorescent light. This is in agreement
esters such as butyric acid methyl and ethyl ester, with the work of Marsili (1999a) who showed a poor
hexanoic acid methyl and ethyl ester and octanoic acid correlation between the extent of light exposure,
methyl ester towards the end of the milk’s shelf life. aldehyde levels produced and perception of off-flavours
These esters were formed in the 2 PE bottles and in partly defatted and skim milk.
particularly in the carton pack. Finally, there was an
initial increase in concentration of 2,2,4-trimethyl
pentane up to day 5 of storage which was followed by 4. Conclusion
a large decrease in this compound’s concentration on
day 7 of storage. Dimethyl disulphide, a typical product Based on sensory evaluation, the shelf life of the Greek
of milk light exposure was not detected in the white/ whole pasteurized milk in this study was approximately 5
black/white HDPE container while its concentration in days. The best overall protection in terms of flavour
the pigmented HDPE bottle (0.6 mg L1) was 3-fold that retention was provided by the multilayer pigmented
in the paperboard carton (0.23 mg L1). This trend in HDPE bottles followed by the paperboard carton and the
dimethyl disulphide production is supported by spectral pigmented HDPE bottle. The onset of milk flavour
transmission curves of the various packaging materials deterioration, as determined by sensory assessment,
tested (Fig. 1). Carbonyl compounds, though present in correlated relatively well with the formation of carbonyl
light-protected containers, (containers 4, 5 and 6) were compounds (hexanal, pentanal, heptanal) and dimethyl
found in substantially lower amounts than in light- disulphide, products of light-induced oxidation in light-
exposed containers (containers 1, 2 and 3). Such exposed containers. Correlation of sensory attributes to
compounds were formed in the light-protected contain- levels of individual volatiles produced in the light-
ers through auto-oxidation, rather than photo-oxida- protected containers was relatively poor.
tion, from oxygen present in the HS of the containers.
Within the storage period tested, and for the given
packaging materials, oxygen permeability (Table 8) did Acknowledgements
not seem to significantly influence the degree of lipid
oxidation in the milk (Tables 4–6). Part of the The authors wish to thank the Food Certification
explanation for this may be the relatively large HS Unit of the University of Ioannina for providing assess
(containing air) which was similar in all packages to the GC/MS instrumentation.
ARTICLE IN PRESS
A.E. Karatapanis et al. / International Dairy Journal 16 (2006) 750–761 761

References Marsili, R. J. (1999b). SPME–MS–MVA as an electronic nose for the


study of off-flavours in milk. Journal of Agricultural and Food
Badings, H. T., & Neeter, R. (1980). Recent advances in the study of Chemistry, 47, 648–654.
aroma compounds of milk and dairy products. Netherlands Milk Marsili, R. T. (2000). Shelf life prediction of processed milk by SPME,
and Dairy Journal, 34, 9–30. MS and multivariable analysis. Journal of Agricultural and Food
Bakish, R., & Hatfield, E. (1997). Coextrusions for flexible packaging. In Chemistry, 48, 3470–3475.
A. L. Brody, & K. S. Marsch (Eds.), Wiley encyclopedia of packaging Marsili, R. (2003). Flavours and off-flavours in dairy foods. In H.
technology (2nd ed., pp. 237–240, 629–638), New York, USA. Roginski, J. Fuquay, & P. Fox (Eds.), Encyclopedia of dairy
Bishop, J. R., & White, C. H. (1986). Assessment of dairy product science, Vol. 2. Amsterdam, The Netherlands: Elsevier.
quality and potential shelf-life—a review. Journal of Food Protec- Marsili, R. T., & Miller, N. (1998). Determination of the cause of off-
tion, 49, 739–753. flavours in milk by dynamic headspace GC/MS and multivariate
Bosset, J. O., Gallmann, P. U., & Sieber, R. (1994). Influence of light data analysis. In C. Mussinan, E. Contis, C. T. Ho, T. Parliament,
transmittance of packaging materials on the shelf life of milk and A. Spanier, & F. Shaidi (Eds.), Food flavour formation, analysis, and
dairy products—a review. In M. Mathlouthi (Ed.), Food packaging packaging influences (pp. 159–171). Amsterdam, The Netherlands:
and preservation (pp. 222–268). London, UK: Blakie Academic & Elsevier Science Publishers.
Prof. Marsili, R. T., & Miller, N. (2001). SPME–GC/MS testing for
Cadwallader, K. R., & Howard, C. L. (1998). Analysis of aroma active monitoring off-flavours in processed milk: Application of odour
components of light-activated milk. In M. Morrello, & C. whey plots. In R. T. Marsili (Ed.), Flavour, fragnance and odour
Mussinan (Eds.), Flavour analysis: Development in isolation and analysis. New York, USA: Marcel Dekker.
characterization. ACS Symposium Series No. 705 (pp. 343–358). Maxcy, J. R., & Wallen, S. E. (1983). Heterogeneity of samples as a
Washington, DC, USA: American Chemical Society. problem in shelf life prediction. Journal of Food Protection, 46,
Christy, G. E., Amantea, G. F., & Irwin, R. E. T. (1981). Evaluation of 543–544.
effectiveness of polyethylene overwraps in preventing light-induced Nijssen, B., Kamperman, T., & Jetten, J. (1996). Acetaldehyde in
oxidation of milk in pouches. Canadian Institute of Food and mineral water stored in polyethylene terephthalate (PETE) bottles:
Science Technology Journal, 14(2), 135–138. Odour threshold and quantification. Packaging Technology and
Cladman, W., Scheffer, S., Goodrich, N., & Griffiths, M. W. (1998). Shelf- Science, 9, 175–185.
life of milk packaged in plastic containers with and without treatment Rysstad, G., Ebbesey, A., & Eggestad, J. (1998). Sensory and chemical
to reduce light transmission. International Dairy Journal, 8, 629–636. quality of UHT milk stored in paperboard cartons with different
Contarini, G., & Povolo, M. (2002). Volatile fraction of milk: oxygen and light barriers. Food Additives and Contaminants, 15,
Comparison between purge and trap and solid phase microextrac- 112–129.
tion techniques. Journal of Agricultural and Food Chemistry, 50, Schröder, M. J. A. (1982). Effect of oxygen on the keeping quality of milk.
7350–7355. 1. Oxidized flavour development and oxygen uptake in milk in relation
Cormier, F., Raymond, H., Champagne, C. P., & Morin, A. (1991). to oxygen availability. Journal of Dairy Research, 49, 407–424.
Analysis of odour-active volatiles from Pseudomonas fragi grown in Schröder, M. J. A., Scott, K. J., Bland, K. J., & Bishop, D. R. (1985).
milk. Journal of Agricultural and Food Chemistry, 39, 159–161. Flavour and vitamin stability in pasteurized milk in polyethylene-
Defosse, M. (2000). Novel milk bottle designs keep blow molders on coated cartons and in polyethylene bottles. Journal of the Society of
top. Modern Plastics, 77, 114–116. Dairy Technology, 38, 48–52.
de Man, J. M. (1978). Possibilities of prevention of light-induced Shipe, W. F., Bassette, R., Deare, D. D., Dunkley, W. L., Hammoud,
quality loss of milk. Canadian Institute of Food Science and E. G., Harper, W. J., et al. (1978). Off flavours of milk:
Technology Journal, 11, 152–154. Nomenclature, standards and bibliography. Journal of Dairy
Hosono, A., Elliot, J. A., & McGugan, W. A. (1974). Production of Science, 61, 855.
ethylesters by some lactic acid and psychrotrophic bacteria. Journal Skibsted, L. H. (2000). Light-induced changes in dairy products.
of Dairy Science, 57, 535–539. International Dairy Federation, Doc. No. 346 (pp. 4–9).
Jenq, W., Bassette, R., & Crang, R. (1988). Effects of light and copper Toso, B., Procida, G., & Stefanon, B. (2002). Determination of volatile
ions on volatile aldehydes of milk and milk fractions. Journal of compounds in cow’s milk using headspace GC/MS. Journal of
Dairy Science, 71, 2366–2372. Dairy Research, 69, 569–577.
Jung, M. Y., Yoon, S. H., Lee, H. O., & Min, D. B. (1998). Singlet Urbach, G. (1990). Headspace volatiles from cold-stored raw milk.
oxygen and ascorbic acid effects on dimethyl disulfide and off-flavour Australian Journal of Dairy Technology, 45, 80–85.
in skim milk exposed to light. Journal of Food Science, 63, 408–412. Urbach, G., & Milne, T. (1987). The concentration of volatiles in
IDF. (1986). Monograph on pasteurized milk. Bulletin No. 200. pasteurized milk as a function of storage time and storage
IDF. (1987). Sensory evaluation of dairy products. Doc. 99A. temperature. A possible indicator of keeping quality. Australian
IDF. (1991a). Enumeration of microorganisms: Colony count Journal of Dairy Technology, 42, 53–58.
technique at 30 1C. Standard 100B. Vallejo-Cordoba, B., & Nakai, S. (1994). Keeping quality assessment
IDF. (1991b). Estimation of numbers of psychrotrophic microorgan- of pasteurized milk by multivaride analysis of dynamic headspace
isms: Colony count technique, 6.5 1C. Standard 101A. GC data. 1. Shelf life prediction by principal component regression.
Kim, Y. D., & Morr, I. V. (1996). Dynamic headspace analysis of light Journal of Agricultural and Food Chemistry, 42, 989–993.
activated flavour in milk. International Dairy Journal, 6, 185–193. van Aardt, M., Duncan, S. E., Marcy, J. E., Long, T. E., & Hackney,
Labuza, T. P. (1982). Open shelf-life dating of foods. Westport, CT, C. R. (2001). Effectiveness of poly(ethylene terephthalate) and high
USA: Food and Nutrition Press. density polyethylene in protection of milk flavour. Journal of Dairy
Leong, C. M. O., Harte, B. R., Partridge, J. A., Ott, D. B., & Downes, Science, 84, 1341–1347.
T. W. (1992). Off-flavour development in milk packaged in Vassila, E., Badeka, A., Kondyli, E., Savvaidis, I., & Kontominas, M.
polyethylene-coated paperboard cartons. Journal of Dairy Science, G. (2002). Chemical and microbiological changes in fluid milk as
75, 2105–2111. affected by packaging conditions. International Dairy Journal, 12,
Marsili, R. T. (1999a). Comparison SPME and dynamic headspace 715–722.
methods for the GC/MS analysis of light-induced lipid oxidation White, C. H. (1985). Consumer reaction to colored plastic milk jugs.
products in milk. Journal of Chromatography Science, 37, 17–23. Journal of Dairy Science, 68, 261–264.

You might also like