You are on page 1of 5

DIETRICH G.

QUAST and MARCUS KAREL


Dept. of Nutrition & Food Science, Massachusetts institute of Technology, Cambridge, MA 02139

EFFECTS O F ENVIRONMENTAL FACTORS O N THE


OXIDATION O F POTATO CHIPS

INTRODUCTION months and then decreased to a level Table l-Effect of equilibrium relative
lower than the starting value. Samples humidity (RH) on the rate of oxygen uptake of
POTATO CHIPS are a deep-fat-fried packed under vacuum showed signifi- ootato chios (from run lla
snack food containing a high percentage cantly smaller peroxide values than those
of oil; in common with other fat-contain- packaged at atmospheric oxygen concen-
ing dry foods, they frequently deteriorate tration. Unfortunately, the water activity RH
due to oxidative rancidity. Loss of ac- of the samples also increased during the (%I
ceptability often occurs because of pro- experiments. The authors also concluded 0.1 0.15
duction of objectionable off-flavor that high moisture content (slightly above 11.0 0.13
compounds resulting from reactions in- monolayer value) resulted in high levels 32.0 0.07
volving oxygen absorption. Knowledge of of peroxides. 40.0 0.06
the factors which affect the rate of Cavaletto et al. (1966) Dela Cruz et 62.0 0.19
oxygen uptake of food products is im- al. (1966), and Cavaletto and Yamamoto 75.0 1.60
portant for process and product develop- (1968, 197 1) studied the factors affecting alnitialheadspace concentration, 12.2% 0, ;
ment, packaging and storage. macadamia nut stability. They found that concentration at which rate was determined
The effect of oxygen and moisture the stability of the product decreased 10% 0, ; temperature, 37’C; extent of oxida-
tion, 57 ~1 0, STP/g
content on the quality of dry foods has with increasing moisture content. The
been investigated by many authors, in- lowest moisture content (1.4%) resulted
cluding Tamsma et al. (1961, 1967) and in the most stable product. Light had no
Tamsma and Pallansch (1964) who made effect on the stability of the product.
storage stability studies of whole milk Similar results were observed for raw and oil did not increase the storage stability
powder under different oxygen concen- roasted kernels. The authors found that of the chips.
trations. Their results are particularly the stability of the nuts could be in- Berger (1971) found that accelerated
significant since they show that for cer- creased by adding antioxidants to the tests were useful for predicting storage
tain products even oxygen concentrations frying oil and by vacuum packaging. life of several products during the initial
lower than 1% can have a marked effect Another factor which can affect signi- slow period of oxidation.
on the quality of the product. ficantly the rate of oxygen uptake of We have undertaken a study with the
Tuomy et al. (1969) and Tuomy and food products is light. Smits et al. (1970) aim of predicting the storage behavior of
Hinnergardt (1968a, b) made an extensive and Radtke et al. (1970) made a com- oxidation-susceptible dry foods on the
study of the effect of headspace oxygen prehensive theoretical and experimental basis of laboratory evaluation of kinetics
concentration on the oxygen uptake and study of this factor for edible oils. of deteriorative reactions and of data for
organoleptic properties of freeze-dried Fuller et al. (197 1) studied the effect oxygen and water transfer into con-
combination foods. They obtained good of frying oils and light on the storage tainers. An initial phase of the investiga-
correlations between the amount of oxy- stability of potato chips. They found that tion dealt with a survey of various types
gen absorbed and the panel rating for fluorescent light accelerated the quality of foods with respect to their oxidation
flavor. Approximate analysis of their data deterioration significantly. Addition of behavior (Quast and Karel, 197 1). Potato
indicates that the rate of oxygen uptake antioxidant (propyl gallate) to the frying chips were then selected for further study
is a strong function of oxygen concentra-
tion in the range of O-5% for these
combination foods.
More recently, the effect of water
activity or equilibrium relative humidity
on the rate of oxygen uptake of dry
foods and model systems was studied by
Labuza et al. (1969, 1971) and Heidel-
baugh et al. (197 1). The results indicate
that usually the rate of oxidation is high
at very low water activity and decreases
with water activity until it reaches a
minimum. At higher water activities
(usually larger than 0.5, corresponding to
the so-called intermediate moisture
foods) the rate of oxidation increased
with water activity.
McWatters et al. (197 1) determined
the peroxide value of roasted peanut
kernels as a function of time at different
temperatures. In most casesthe peroxide
value increased during the initial six Fig. l-Oxidation ofpotato chips at 37OC up to a very high extent.

584JOURNAL OF FOOD SCIENCE-Volume 37 (1972)


OXIDATION OF POTATO CHIPS-585

as representative of a moisture- and space oxygen concentration could be adjusted this manner the gas previously present in the
oxygen-sensitive group of products. to a desired level. The change of headspaceoxy- system was flushed out and after some time the
This paper presents the first phase of gen concentration was determined with the aid oxygen concentration inside the flasks was
of an oxygen probe with a tip 10 m m long equal to that of the standard gas mixture. To
our investigation and is concerned with which could be inserted through the 6 m m establish when the system had been sufficiently
the influence of environmental factors on thick rubber stopper (Quast and Karel, 1972). flushed, the oxygen concentration of the gas
rates of oxidation of the products. Sub- The net headspace volume was calculated by leaving the Warburg manometer was monitored
sequent papers will deal with the develop- subtracting the volume occupied by the sample. continuously with an oxygen probe. Then the
ment of mathematical models for storage From the headspace oxygen concentration, manometer valve and the sidearm valve were
behavior and with evaluation of these measured at regular intervals, the extent of closed and, after thermal and diffusional equi-
models. oxidation as well as the rate of oxidation at any librium were established, the rate of oxygen
time could be calculated. uptake was determined at the new oxygen con-
centration over a short period.
EXPERIMENTAL Oxygen uptake by Warburg technique After this the oxygen concentration was
Samples This method has been fully described by changed to some other value in a similar man-
Potato chips were obtained from the John Umbreit et al. (1964). The flasks used had ner. After determining the rates at all the
E. Cain Co., Cambridge, Mass., on the same day volumes from 15-60 ml, and samples of 2-5g desired oxygen concentrations, the procedure
they were produced. Two experimental runs were used. Two types of fluids were employed was repeated through new cycles of oxygen
were made in which the sampleshad the follow- for the manometers. For long-term experiments concentrations until the critical extent of oxi-
ing characteristics: at constant oxygen concentration and water dation was reached. In this manner the water
activity, mercury was used. The use of mercury activity of a given sample stayed constant dur-
Run 1: results in low sensitivity, and readings were ing the whole process while the oxygen concen-
Oil used for frying: Sunflower oil usually taken once every 5-10 days. The ad- trations were changed frequently and the ex-
Peroxide value: 0.55 meq/kg chips vantage of mercury is that the levels never tent of oxidation increased.
Fat content: 34% exceed the physical limits of the manometer Since the amount of calculations is formi-
Density: 1.21g/cm3 even during very large changes in atmospheric dable, only the manometric readings were
Moisture content: 0.9Og/lOOg solids pressure. recorded, and the rates together with their
For the accurate determination of rates ovex standard deviations were calculated by a com-
Run 2: small time intervals, Apiezon oil (Associated puter program using least-squaresdata-fitting.
Oil used for frying: Cottonseed oil Electrical Industries, Ltd., distributed by Shell
Peroxide value: 0.40 meq/kg chips Oil Corp.) was used in the manometers.
Fat content: 39% A special technique was developed to allow RESULTS & DISCUSSION
Density: 1.165g/cm3 changing the oxygen concentration (PO2) in- Preliminary experiments
Moisture content: 1.3Og/lOOgsolids side the flasks: Triplicate manometer flasks
containing samples were transferred to desicca- In order to establish the significant
The peroxide values were determined ac- range of oxygen uptake in potato chips, a
tors containing saturated solutions of given
cording to the “Official Methods” of the preliminary experiment was made over a
equilibrium RH. Adequate equilibration was
American Oil Chemists’Society after fat extrac- long time and with extensive oxidation.
tion (Karel and Labuza, 1969). The results were obtained after 2 days at 37°C under high
vacuum. Then the flasks were connected to After approximately 1200 ~1 OaSTP/g
always expressed as meq/kg of dry product.
Each equivalent of peroxide, as calculated from their manometers and the rate of oxygen up- were absorbed, the rate increased very
the titration, corresponds to half a mole of take at atmospheric oxygen concentration and rapidly to a level roughly 100 times the
at certain RH was determined over a period of rate observed in the previous period (see
oxygen taken up by the product if all of the l-2 days.
oxygen is incorporated into hydroperoxides. Fig. 1). The rapid increase of the rate
Consequently, a peroxide value of one cor- Next the oxygen concentration inside the
flasks was adjusted to some other value by pass- coincides with a very marked develop-
responds stoichiometrically to an oxygen up- ment of rancid odor.
ing a stream (20-50 cm3/min) of an oxygen-
take of 11.2 ~10, STP/g (or 0.5 pmole 0, /g). nitrogen mixture of known composition
Oxygen uptake by measuring headspace through the system. First, this gas was bubbled
concentration through a wash bottle which contained a satu-
In these experiments, 20.0-g samples of rated solution of the same RH as a given sample
crushed potato chips were transferred to wide- and which was at the same temperature. This
mouth glass bottles of a total volume of 72 prevented addition or removal of moisture from
cm’. These flasks were provided with rubber the samples. Next, the gas stream was con-
stoppers such as are used in packaging of nected to the opened sidearm of the manom-
pharmaceuticals. The equilibrium relative eter flask. The valve on top of the manometer
humidity (RH) of the samples and the head- was opened to allow the gas to flow through. In

W
0 60.

HOURS Fig. J-Effect of light and RH on the oxidation


Fig. 2-Change of peroxide value of potato chips at 37’C of potato chips at 37OC.
566-JOURNAL OF FOOD SCIENCE- Volume 37 (1972)

Even though rancid odor appears be- at an extent of 1200 to 1500 ~1 Oz STP/g, considerably, possibly due to increase in
fore the critical extent of 1200 ~1 independent of water activity, light, tem- mobility of the reactants. The product
OaSTP/g is reached, it was assumed that perature and oil in which the chips were kept at 75% RH also showed strong
this is the maximum allowable extent. At fried. nonenzymatic browning.
this point the product is of very low The effect of equilibrium RH on the Furthermore, the product with a water
quality from the point of view of flavor. rate of oxidation is shown in Table 1. It activity of 0.40 (RH=40%) was found to
In later investigations it was found that can be seen that the rate is lowest around be unacceptable because of loss of crisp-
the period of rapid oxidation was reached 40% RH. At high RH the rate increases ness. Kaghan (1968) indicated that pota-
to chips are considered commercially un-
acceptable at a moisture content of 3%
(RH=32%). Consequently subsequent
Table 2-Effect of light on oxygen uptake of potato chips at 23°C. work concentrated on RH below 40%.
0.1% RH (from run 2) The pattern of oxidation of the potato
Oxygen cone chips over time followed the usual autoxi-
Time itio,, EXT dation kinetics (Fig. 2 and 3), in which
(days) insideflasks Pva EXTb .PV most of the oxygen absorbed in early
Dark: flaskswrappedin aluminumfoil stages of oxidation is present in the
0 21 0.4 ‘LO - hydroperoxides (Labuza et al., 1969). If
15 19.0 - - - all of the absorbed oxygen is present in
30 17.4 - - - hydroperoxides, the EXT/PV ratio is
45 17.0 - - equal to 11.2. Larger values indicate
60 - 20.6 - significant oxygen content in other oxida-
75 12.6 - - - tion products. In experiments in which
92 10.5 23.1 210 11.7 peroxide value as well as oxygen absorp-
116 7.5 30.8 350 11.4 tion were studied under identical condi-
Room light: no sunlight,strongartificial light tions, the ratios of oxygen absorbed to
0 21 0.4 ‘LO - peroxide value were close to the theo-
15 18.5 - - retical value (Table 2). The subsequent
30 15.4 - - discussion of individual environmental
45 12.0 - - factors is based primarily on oxygen up-
60 - 25.1 - - take, which will hereafter be referred
75 1.5 - 417 25.4
to as “extent” 011 OaSTP/g). The rate of
oxygen uptake will be given in units of 1.11
92 4.8 16.4 497 13.2
O2 STP/g. hr.
116 1.7 31.5
Sunlight: flasksnearwindow, receivingsomesunlightduring the day Effect of light
0 21 0.4 %O - The effect of light on the rate of
15 16 - - oxygen uptake is shown in Figure 3 and
30 8 - - - Table 2. From the figure one can see that
45 0.7 - - even diffuse artificial room light can have
60 0.0 38.9 540 13.9 a significant effect on the rate of oxygen
15 - - - - uptake. This effect becomes more impor-
tant at higher water activity.
92 0.0 38.8 540 13.9 Table 2 shows results obtained by
116 - 38.6 540 14.0
flask experiments at 23°C where the
aPV = peroxidevalue peroxide values were also determined.
bEXT = extent Similar results were also obtained at
37°C. In addition, the very strong acceler-
ating effect of occasional sunlight is
shown. At higher water activities an even
more significant effect is to be expected,
according to Figure 3. In practice, it is
difficult to assessaccurately the effect of
light on packaged foods since it depends
on the area exposed to the light.
Effect of temperature
z The rate of oxygen uptake was deter-
,b
P
1 _ RH=II.
mined at three temperatures at an RH of
I- 0.1%. From an Arrhenius plot of the
; o++----
rates, we found the activation energy for
oxygen uptake of potato chips to be 10
3
kcal/g-mole at low RH. This is a relatively
low value as compared with 19 kcal/g-
mole calculated from the results of Berger
(197 1). Thus, the effect of temperature
----a on the rate of oxidation of potato chips is
0&d-
500 1000 1500 2oocJ not very strong. Consequently, relatively
EXTENT (pl 02STP/g)
high temperatures have to be used in
Fig. 4-Rate of oxidation of potato chips at 37OC as a function of accelerated stability tests if significant
extent at PO2=0.0055 atm. acceleration is desired. It is possible,
O X IDATION OF POTATO CHIPS-587

P function of the extent but the slope of


the straight line portion of plots is a
function of the other variables.A t P O 2 =
0.21 the straight line approximation is
g o o d only u p to a n extent of
1 2 0 0 - 1 5 0 0 , while at low o x y g e n concen-
tration this approximation is valid for
slightly h i g h e r extents. T h e initial d e -
crease of the rate at low water activity
could b e d u e to a very small increaseof
the water activity d u r i n g the experiment.
E ffect of equilibrium relative humidity
W-0
D u e to the experimental p r o c e d u r ein
which the rates w e r e obtained, it w a s not
possibleto determinet h e m at regulara n d
E X T E N T tcl O,STP/g) pre-establishedintervals. Therefore, the
Fig. & R a t e of oxidation of potato chips at 3 7 O C as a function of
rates at five extents (0, 300, 600, 9 0 0 a n d
extentat P O 2 = 0 . 2 1 arm.
1 2 0 0 ) at e a c h of the values of the other
i n d e p e n d e n tvariables( R H a n d P 0 2 ) w e r e
taken from the plots, such as s h o w n in
Figures 4 a n d 5. This technique allows
however, that at very h i g h temperatures until a m a x i m u m extent of a b o u t 2 0 0 0 ~ 1 crossplotting the experimental results.
the m e c h a n i s m of the reaction changes, O z S T P /g w a s reached. In this m a n n e r , Figure 6 s h o w s h o w the rate of o x y g e n
thus limiting the application of acceler- rate determinations w e r e m a d e in tripli- uptake d e p e n d s o n the R H at selected
ated tests. cate at five different values of RH, six valuesof the other i n d e p e n d e n tvariables
values of P O 2 (oxygen partial pressure, over the w h o l e r a n g eof interest.
E ffect of extent of oxidation including P O 2 = 0 .0),a n d b e t w e e nfive a n d
T h e rates of o x y g e n uptake w e r e 1 5 different values of extent, d e p e n d i n g It c a n b e s e e n that in early stagesof
d e t e r m i n e d in triplicate over small tim e o n the relative humidity. A b o u t 1 0 0 0 oxidation a n d at low humidities the rate
intervals at the following equilibrium experimental rates at the combinations of is a very strong function of RH. A s the
relative humidities: 0.1, 11, 20,. 3 2 a n d the three i n d e p e n d e n tvariablesw e r e o b - reaction progressesto the critical extent,
4 0 % . T h e o x y g e n concentrations w e r e tained. however, the rate b e c o m e s almost inde-
cycled as described previously over the A limited n u m b e r of the experimental p e n d e n t of the equilibrium RH. Also, at
following concentrations (PO2): 0.0055, results at s o m e of the valuesof R H a n d h i g h o x y g e n concentration the rate is a
0.035, 0.072, 0 . 1 2 2 a n d 0.21 atm of P O 2 is s h o w n in Figures4 a n d 5. It c a n strongerfunction of RH.
partial pressure.T h e rate determinations b e s e e n in all cases that the rate of E ffect of o x y g e n concentration
w e r e continued with increasing extent o x y g e n uptake is approximately a linear T h e experimental results c a n also b e
crossplotted as a function of P O 2 at
selectedvalues of the other two i n d e p e n -
dent variables, as s h o w n in Figure 7. It
c a n b e s e e n that in all casesthe rate is a
very strong function of o x y g e n concen-
tration w h e n the concentration is low.
T h e s h a p e of the curves is similar for all
valuesof the other i n d e p e n d e n tvariables
except for R H = O .l, w h e r e the curve is
slightly different. This could indicate a
different m e c h a n i s m of reaction at this
low water activity. It must also b e m e n -
tioned, however, that the error in the
determinations at this water activity is
likely to b e larger d u e to the h i g h
sensitivity of the system to very small
c h a n g e sin moisture content.
All the environmental factors studied
h a d a strong effect o n the rate of o x y g e n
absorption of potato chips. Similar ef-
fects a r e to b e expected for a large
n u m b e r of snack foods. U n d e r typical
commercial storage conditions the light
intensity a n d temperature c a n b e quite
variable. Completely o p a q u e p a c k a g i n g
m a terials should b e u s e d to eliminate the
effect of light o n oxidation rates.
In a typical flexible p a c k a g e contain-
Fig. 6-Rate of oxidation of potato chips at ing a food product, the equilibrium R H
3 7 ° C as a function of equilibrium relative h u - Fig. I-Rate of oxidation of potato chips at a n d the h e a d s p a c eo x y g e n concentration
midity. 3 7 ’C as a function of o x y g e n partial pressure. c h a n g e d u r i n g storage. This c h a n g e is a
!i%%-JOURNAL OF FOOD SCIENCE- Volume 37 (1972)

function of package parameters and of stages of oxidative deterioration of the gen diffusion on oxidation of some dry
foods J. Food Tech. (British) 6: 95.
the rate at which oxygen is absorbed by potato chips used in this study. Quast, D.G. and Karel, M. 19721 Technique for
the product. In future work on this determining oxygen concentrations inside
subject we will explore the possibilities of packages. J. Food Sci. 37: 490.
REFERENCES
Radtke. R.. Smits. P. and Heiss. R. 1970. Uber
using the experimental results for the den.Einfluss vbn Licht verschiedener Inten-
Berger, K.G. 1971. Practical application of an
development of mathematical models for accelerated stability test to rancidity prob- sit& und WeIIenIiinaenbereiche auf den
storage life prediction. lems in food processing. J. Food Tech. oxydativen Verderb v&i Speiseolen II: Ver-
(British) 6: 253. suchsergebnisse und Diskussion Fette,
Cavaletto: C.G. and Yamamoto, H.Y. 1968. Seifen. Anstrichmittel 72(6I: 497.
Criteria for selection of packaging materials Smits, P.-Becker, K. and He&, R. 1970. Uber
CONCLUSIONS for roasted macadamia kernels. J. Food Sci. den EinfIuss van Licht verschiedener Inten-
22: 97. sit& und WeIIenIiingenbereiche auf den
CONSIDERING the spoilage of potato Cavaletto. C.G. and Yamamoto, H.Y. 1971. oxvdativen Verderb van SoeiseGlen 1. Fette.
chips due to oxidative rancidity, there are Factors affecting macadamia nut stability. Seifen. Anstrichmittel 72t63: 490. .
3. Effect of roasting oil quality and antioxi- Tamsma, A. and PaIIansch. M.J. 1964. Factors
several approaches for improving the stor- dants. J. Food Sci. 36: 81. related to the storage stability of foam spray
age stability of this product. Since the Cavaletto, C.. Dela Cruz, A., Ross, E. and dried whole miIk. 4. Effect of oowder mois-
Yamamoto, H.Y. 1966. Factors affecting ture content and in-pack oxygen at different
bulk density of this product is typically macadamia nut stability. 1. Raw kernels. sSt;;ge temperatures. J. Dairy Sci. 47(S):
0.056g/cm3, it has a very large headspace Food Technol. 20: 1084.
volume per unit weight of product. If the Dela Cruz, A., Cavaletto. C., Yamamoto, H.Y. Tamsma, A.. Pallansch, M.J.. Mucha, T.P. and
and Ross, E. 1966. Factors affecting maca- Patterson, W.I. 1961. Factors related to the
product is packaged at atmospheric oxy- damia nut stabihtv. 2. Roasted kernels. fhwor stability of foam-dried whole milk. 1.
gen concentration, then the headspace J. Food Sci 20: 12i7. fiH$ of oxygen level. J. Dairy Sci. 44(S):
Fuller, G., Guadagni, D.G., Weaver, M.L., Not-
oxygen is enough to cause oxygen uptake ter G. and Horvat, R.J. 1971. Evaluation of Tamsma: A., Kurtz, F.E. and PaIIansch. M.J.
in excess of 3000 ~1 OsSTP/g. Conse- ole:c safflower oil in frying of potato chips. 1967. Effect of oxygen removal technique
on flavor stability of low-heat foam spray-
quently, inert gas packaging would result Hei~e~a~~~+r3~: ,
43.
. ., Yeh, C.P. and Karel, M. id; whole miIk. J. Dairy Sci. 50(10):
in a very significant increase of the 1971. Effects of model system composition
storage life of potato chips, provided the on autoxidation of methyl Iinoleate. Agric. Tuomy,‘J.M. and Hinnergardt, L.C. 1968a.
Food Chem. 19(l): 140. Effect of headspace oxygen on the quality
headspace oxygen concentrations at- Kaghan, W.S. 1968. Olin’s new PCG, a unique of freeze-dried beef and chicken stew.
tained were below 1% and the package two-side coated ghrssine. 30th Annual Technical Report 6865FL, U.S. Army
Packaging Forum. The Packaging Institute, Natick Labs.
permeability to oxygen was chosen ac- Ny Tuomy, J.M. and Hinnergardt, L.C. 1968b.
cordingly. The package should also be Karei M. and Labuza. T.P. 1969. Optimization Effect of headspace oxygen on the quality
of protective packaging of space foods. of freeze-dried raw beef patties. Technical
designed to avoid light penetration. Final report. USAF School of Aerospace Report 6%54-FL, U.S. Army Natick Labs.
From the results shown, it can also be Medicine Contract No. F 41-609-68-C-0015. Tuomy, J., Hinnergardt, L.C. and Hehner. R.L.
concluded that an excessively dry prod- Labuza. T.P., Tsuyuki, .H. and Karel, M. 1969. 1969. Effect of oxygen uptake on quality of
Kinetics of Iinoleate oxidation in model cooked, freeze-dried combination foods.
uct is not desirable for good storage systems. J. Am. Oil Chem. Sot. 46(8): 409. Agric. Food Chem. 17(6): 1360.
stability. The activation energy for oxy- Labuza. T.P., Heidelbaugh, N.D.. Silver, M. and Umbreit. W.W., Burris. R.H. and Stauffer, J.F.
Karel, M. 1971. Oxidation at intermediate 1964. “Manometric Techniques.” Burgess
gen uptake of potato chips is not very moisture contents. J. Am. Oil Chem. Sot. Publishing Co. Minneapolis. Mimi.
hiah comnared with that of other mecha- 4812): 86. M S received 2113172; revised 412172; accepted
McWatters, K., Heaton, E.K. and Cecil. S.R. 416172.
nisms of food quality deterioration. 1971. Storage stability of peanut pie mixes. This investigation was supported by Public
The peroxide value appears to be a Food Prod. Dev. (Feb-March): 69. Health Service Grant No. FD00140 from the
fairly reliable index of the extent in early Quast, D. and Karel, M. 1971. Effects of OXY- Food and Drug Administration.

You might also like