Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Quality control of raw cows’ milk by headspace analysis

Quality control of raw cows’ milk by headspace analysis

Ratings: (0)|Views: 11 |Likes:
Published by yuki_akitsu
quality control in cows milk,
quality control in cows milk,

More info:

Categories:Types, Research
Published by: yuki_akitsu on Jun 23, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

05/14/2014

pdf

text

original

 
International Dairy Journal 18 (2008) 506–513
Quality control of raw cows’ milk by headspace analysis
K.A. Hettinga
Ã
, H.J.F. van Valenberg, A.C.M. van Hooijdonk
Dairy Science and Technology Group, Wageningen University and Research Centre, P.O. Box 8129, 6700 EV Wageningen, The Netherlands
Received 30 June 2007; accepted 24 October 2007
Abstract
This study investigated whether headspace analysis of volatile components can be used for monitoring the quality of raw cows’ milk.The detection of different quality defects caused by cows’ feed, microbiological and chemical contamination, as well as enzymaticdeterioration was studied. Fresh raw milk without quality defects was shown to always contain the same seven volatile components.It was also shown that treatments like heating and homogenization of raw milk may drastically change this basic pattern resulting in asmuch as a 10-fold increase in the number of volatile compounds. The growth of 
Pseudomonas
could not be detected in an early stageusing headspace analysis. Feed was shown to have an effect on the volatile composition if specific vegetable byproducts were fed to thecow. Chloroform contamination was quantified using the method. Also, the extent of lipolysis could be quantified by measuring the freefatty acids. For quantification of both chloroform and lipolysis, the sensitivity and reproducibility of the method were sufficient forquality control purposes. The method was thus able to detect several quality defects with a single analysis and may therefore be a usefulsupplementary method for raw milk quality control.
r
2007 Elsevier Ltd. All rights reserved.
Keywords:
Milk; Headspace; Solid-phase microextraction; Gas chromatography-mass spectrometry; Quality monitoring
1. Introduction
Chemical analyses are an important tool to monitor thequality of food products. One of the fastest and easiestchemical analyses performed with this goal is based onheadspace analysis of volatile components. Analysingvolatile components has been proven to be useful forquality monitoring of a wide range of food products, suchas oils (Aprea et al., 2006), vegetables (Barra et al., 2007), fruits (Beaulieu&Lea, 2006), ready to eat products (Limbo, Sardi, Farina,&Rubin, 2005), wine (Carrillo& Tena, 2006), juices (Ros Chumillas, Belissario, Iguaz,& Lopez, 2007), cheese (Andersen, Wold,&Mortensen, 2006), and milk (Contarini&Povolo, 2002;Marsili, 2000; Valero, Sanz,&Martinez Castro, 1999). Headspace analysis of milk and milk products has been performedusing static headspace, solid-phase microextraction(SPME), and purge and trap (P&T). To identify theindividual volatile components, the headspace samplingtechniques were usually coupled to gas chromatography/mass spectrometry (GC/MS) (Contarini, Povolo, Leardi,&Toppino, 1997;Fabre, Aubry,&Guichard, 2002;Vallejo- Cordoba&Nakai, 1994a). Although quality control of milk based on headspaceanalysis has been focused on (heated) milk products(Marsili, 2002;McSweeney, Nursten,&Urbach, 1997; Vallejo-Cordoba&Nakai, 1994a), also changes in raw milk quality may be detectable by analysing the volatilecomposition. Cows’ diet, microbiological and chemicalcontamination, as well as enzymatic deterioration maychange the volatile composition of raw milk (Azzara&Campbell, 1992;Shipe et al., 1978). As headspace analysis can detect a whole range of volatile components at thesame time, it may be able to detect a wide range of qualitydefects with a single analysis. Headspace analysis may thusbe a useful supplementary method for raw milk qualitycontrol.Differences within the normal cows’ diet as well as feed-ing of specific vegetable byproducts may alter the volatile
ARTICLE IN PRESS
www.elsevier.com/locate/idairyj0958-6946/$-see front matter
r
2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.idairyj.2007.10.005
Ã
Corresponding author. Tel.: +31317482286; fax: +31317483669.
E-mail address:
 
composition of raw milk. The differences in volatilecomposition between cows that received either fresh grassor silage have been compared before (Bendall, 2001).This showed almost no effect of the ration on the volatilecomponents detected in milk. However, the difference involatile composition of milk from cows receiving either astarch-rich (e.g., maize) or crude fiber-rich (e.g., grass)diet has not been studied before. These differences in diethave an influence on the overall composition of the milk,by influencing the fermentation in the rumen (Ekern et al.,2003). Also, the transfer of volatile components fromspecific vegetable byproducts to the milk may be studiedusing a headspace analysis. The transfer of terpenes fromcows’ feed to the milk has been studied before (Fernandez,Astier, Rock, Coulon,&Berdague, 2003). The transfer of a wider variety of flavor components to the milk,causing specific off-flavors, has also been describedbefore (Shipe et al., 1978). Whether this transfer of volatile components from the feed to the milk can bedetected using a headspace analysis has, however, not beenstudied.Microbiological contamination has a major influence onraw milk quality. The major group of bacteria causingproblems in cold stored raw milk are psychrotrophic bac-teria, the most important group of which are
Pseudomonas
species. In the literature, the production of volatilecomponents by
Pseudomonas
has only been described forpasteurized milk (Cormier, Raymond, Champagne,&Morin, 1991;Reddy et al., 1968). Headspace analysis may thus also be a useful method to detect
Pseudomonas
inraw milk as a marker for microbiological contaminationand growth.Chemical contamination of raw milk is mainly due to thecleaning and disinfection of the milking equipment. Themain volatile component associated with this problem ischloroform (Resch&Guthy, 1999). Static headspace combined with GC/ECD has been used before to quantifychloroform (Miller&Uhler, 1988;Resch&Guthy, 1999). It is, however, not known whether our headspace analysisis sensitive enough to quantify chloroform in raw milkcorrectly.One of the main enzymatic reactions causing qualitydeterioration in raw milk is lipolysis. Lipolysis is thebreakdown of triglycerides to free fatty acids (FFAs). Thiscan cause a rancid off-flavor in milk (Santos, Ma, Caplan,&Barbano, 2003). A method by the Bureau of DairyIndustries (BDI) is the current method to determine theextent of lipolysis (Deeth, 2006). This BDI method islaborious, expensive, and does not always correspond verywell with the off-flavor in milk. A headspace analysis may,however, correspond better with the off-flavor and is easierand cheaper (Evers, 2003;Gonzalez-Cordova&Vallejo- Cordoba, 2003). A headspace method may thus be a usefulalternative to the BDI method.The aim of this study was to show which quality defectsof raw milk can be studied simultaneously with a simple,fast, and robust headspace analysis.
2. Materials and methods
 2.1. Milk samples 2.1.1. Fresh raw milk 
Fresh raw cowsmilk samples from individual milktrucks were frozen at
À
20
1
C (
n
¼
10; 250mL sample pertruck). Loss of volatile components is not specificallyprevented in this experiment, as sampling was performed asis currently usual for raw cows’ milk.Also, 10mL fresh raw cows’ milk samples were analyzedfrom 10 different individual Dutch farms (bulk tank milksample) as well as from 46 individual cows of two experi-mental farms of Wageningen University, ‘‘De Ossekam-pen’’ and ‘‘Zegveld’’.
 2.1.2. Feeding experiment: grass/maize
Ten cows were selected at the experimental farm of Wageningen University, ‘‘De Ossekampen’’. The cowswere randomly distributed over five groups of two cowsin a latin square design. Every group of cows received adlibitum a diet containing 55% silage and 45% concentrate,based on dry matter (DM). Grass silage (GS), maize silage(MS), starch-rich concentrate (SC), and crude fiber-richconcentrate (CC) were fed in different ratios. Cows werefed one of five diets:1. 55% GS, 22.5% SC, and 22.5% CC2. 27.5% GS, 27.5% MS, 45% SC3. 27.5% GS, 27.5% MS, 22.5% SC, 22.5% CC4. 27.5% GS, 27.5% MS, 45% CC5. 55% MS, 22.5% SC, and 22.5% CCEach diet was given for 3 weeks. After 3 weeks, a 10mLmilk sample was taken.
 2.1.3. Feeding experiment: specific vegetable byproducts
The effect of feeding onions, green cabbage, orange peel,and spent beer brewers barley to cows was investigated.These products were chosen based on their usage as cowfeed in the Netherlands, as well as their possible off-flavoreffect. Eight cows were selected at the experimental farm of Wageningen University, ‘‘De Ossekampen’’. The cowswere randomly distributed over four groups. Next to anormal diet (mixture of maize and grass silage), every cowreceived a total of 0.5kgDMday
À
1
of one vegetablebyproduct for 5 days. The vegetable byproducts weremixed with the normal diet. Every day, for 8 days, a 50mLmilk samples was taken twice a day.
 2.1.4. Microbiological contamination: pseudomonas
One liter fresh raw cows’ milk was obtained from theexperimental farm of Wageningen University, ‘‘De Osse-kampen’’ and divided in aliquots of 250mL in sterile250mL bottles. The samples were spiked with
Pseudomo-nas fragi 
(supplied by the Laboratory of Food Microbio-logy of Wageningen University) at approximately 10
5
ARTICLE IN PRESS
K.A. Hettinga et al. / International Dairy Journal 18 (2008) 506–513
507
 
colony forming units (cfu)mL
À
1
in triplicate. The bottleswere stored at 7
1
C. Samples were taken every day forcounting of the number of 
Pseudomonas
according toHayes, White, and Drake (2002). Also, 10mL samples weretaken daily for headspace analysis.
 2.1.5. Chemical contamination: chloroform
Fifty milliliters of fresh raw cows’ milk was obtainedfrom the experimental farm of Wageningen University,‘‘De Ossekampen’’. The sample was divided over 15 vialscontaining 2mL of milk. For determination of linearity,two samples were analyzed without spiking and eightsamples were spiked in duplicate with chloroform at a levelof 2.5, 5, 12.5, or 25
m
gL
À
1
. For determination of reproducibility, one sample was spiked in five-fold at alevel of 12.5
m
gL
À
1
.
 2.1.6. Enzymatic deterioration: lipolysis
Fresh raw cowsmilk (2L) was obtained from theexperimental farm of Wageningen University, ‘‘De Osse-kampen’’. Within 1h after obtaining the milk, 250mL milkwas heated to 40
1
C and subsequently mixed with a blenderfor 1min. The milk sample was cooled with cold runningwater. Various amounts (0–5mL) of the milk mixed with ablender were added to 200mL fresh raw milk. Afterkeeping the samples for 72h at 4
1
C, 0.2mL hydrogenperoxide was added to stop lipase activity. Samples wereanalyzed for extend of lipolysis with the BDI methodaccording toDriessen, Jellema, Van Luin, Stadhoudersn,&Wolbers (1977).All samples were kept frozen at
À
20
1
C for a maximum of 1month before analysis. A vial with 10mL demineralizedwater was used as blank for SPME analysis and a glassflask with 25mL demineralized water was used as blank forP&T analysis.
 2.2. SPME analysis
Milk samples of 10mL were heated in 20mL vials sealedwith silicone/Teflon septa and magnetic caps for 1min at60
1
C. Volatiles were extracted from the headspace for5min with a 75
m
m PDMS-carboxen SPME fiber (Supelco,Zwijndrecht, The Netherlands) using the combiPALautosampler (CTC Analytics AG, Zwingen, Switzerland).The volatiles were thermally desorbed from the fiber byheating for 3min at 250
1
C. The fiber was subsequentlycleaned for 10min at 275
1
C. GC/MS analysis wasperformed on a Finnigan Trace GC gas chromatographcoupled to a Finnigan DSQ mass spectrometer (Thermo-Finnigan, San Jose, CA, USA). Volatiles were separated ona BPX-5 column of 30m length, 0.15mm internal diameter(i.d.), and 0.25
m
m film thickness (SGE, Bester, Amstelv-een, The Netherlands). Oven temperature was held at
À
30
1
C for 3min, raised to 190
1
C at 20
1
Cmin
À
1
, followedby 1min holding. Helium was used as the carrier gas at aflow rate of 0.6mLmin
À
1
. The MS interface and the ionsource were kept at 250
1
C. Acquisition was performed inelectron impact mode (70eV) with 2scanss
À
1
; the massrange used was
m
/
z
33–250. For the determination of theextent of lipolysis, the acquisition was performed in SIMmode, quantifying the ions with
m
/
z
60 and 73.
 2.3. Purge and trap analysis
Milk samples of 25mL were heated at 40
1
C in a closedglass flask (70mL). The flask was placed in a water bath of 37
1
C and a flow of purified nitrogen (50mLmin
À
1
) waspassed through the sample for 30min. The sample wasconstantly mixed by a magnetic stirrer. Volatiles wereadsorbed on a glass tube (length 100mm, 3.0mm i.d.) filledwith 90mg of Tenax TA (20/30 mesh, Alltech NederlandBV, Zwijndrecht, The Netherlands) using P&T. A cold trapwith ethanol of 
À
10
1
C was used to prevent water vaporfrom entering the Tenax tube. Volatiles were desorbed ontothe column using a thermal desorption unit for 10min at250
1
C (Chrompack TCT injector 16200, Chrompack,Middelburg, The Netherlands). GC/MS analysis wasperformed on a Varian 3400 gas chromatograph (Varian,Bergen op Zoom, The Netherlands) coupled with aFinnigan MAT95 mass spectrometer (Thermo Electron,Bremen, Germany). Volatiles were separated on a AgilentJ&W DB-5 column of 60m length, 0.25mm i.d., and0.25
m
m film thickness (VWR International B.V., Amster-dam, The Netherlands). The oven temperature was held at40
1
C for 4min, raised to 200
1
C at 6
1
Cmin
À
1
, followed by4min holding. Helium was used as carrier gas at a constantpressure of 125kPa. The mass spectrometer was operatedin electron impact mode (70eV) with 1scans
À
1
; the massrange used was
m
/
z
24–300.
 2.4. Data analysis
The results of the measurements were analyzed using theAutomated Mass Spectral Deconvolution and Identifica-tion System (AMDIS) software (NIST, Gaithersburg, MD,USA). Identification of milk volatiles was based onmatching deconvoluted mass spectra from AMDIS withspectra from the NIST/EPA/NIH mass spectral database.Whenever possible, the retention times were matched tostandards, spiked to milk. Analytical standards wereobtained from Sigma-Aldrich (Zwijndrecht, The Nether-lands), with the exception of FFA which were obtainedfrom Supelco (Zwijndrecht, The Netherlands) and ethanol,chloroform, and acetone which were obtained from Merck(Darmstadt, Germany). Subsequent peak integration wasperformed using the XCalibur software package.SPSS for Windows version 12.0 (SPSS Inc., Chicago,IL, USA) was used for comparisons between groups(ANOVA). SPSS was also used to determine the mean,correlation coefficient, and relative standard deviation(RSD) of the results.
o
0.05 was considered statisticallysignificant.
ARTICLE IN PRESS
K.A. Hettinga et al. / International Dairy Journal 18 (2008) 506–513
508

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->