Food Chemistry 119 (2010) 1143–1149

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Vacuum frying reduces oil uptake and improves the quality parameters
of carrot crisps
V. Dueik a, P. Robert b, P. Bouchon a,*
a
Department of Chemical and Bioprocess Engineering, Pontificia Universidad Católica de Chile, P.O. Box 306, Santiago 6904411, Chile
b
Department of Food Science and Chemical Technology, Universidad de Chile, P.O. Box 223, Chile

a r t i c l e i n f o a b s t r a c t

Article history: Recent consumer trends towards healthier and low fat products have had a significant impact on the
Received 28 April 2009 snack industry. The objective of this study was to examine the most important quality parameters of vac-
Received in revised form 30 June 2009 uum and atmospheric fried carrot slices in order to identify the specific advantages of vacuum technol-
Accepted 20 August 2009
ogy. Said parameters include oil uptake, colour changes, and trans a and b-carotene degradation.
Equivalent thermal driving forces were used (DT = 60 °C and 80 °C) to compare the processes, maintain-
ing a constant difference in temperature between the oil and the boiling point of water at the working
Keywords:
pressure. The results showed that vacuum frying can reduce oil content by nearly 50% (d.b.) and preserve
Vacuum frying
Carrot
approximately 90% of trans a-carotene and 86% of trans b-carotene. This process also allowed for the raw
Oil uptake carrot colour to be preserved, which was reflected by good correlations between a* and trans b-carotene
Carotenoids content, b* and trans a-carotene content, and hue and total carotenoid content.
Colour Ó 2009 Elsevier Ltd. All rights reserved.
Crisp
Frying

1. Introduction Several processes have been developed in order to allow com-

panies to manufacture reduced-fat products that possess the de-
Frying is a complex unit operation that is widely used in the sired quality attributes of deep fat fried food whilst preserving
food industry. During the process, food is immersed in an oil bath their nutritional properties. These include alternative technologies
at a temperature above the boiling point of water. This results in such as extrusion, drying, and baking, which may be applied to raw
counterflow of water vapour (bubbles) and oil at the surface of food or formulated products. Unfortunately, none of them has been
the product (Bouchon, Aguilera, & Pyle, 2003). Oil uptake is one as successful as expected because they are still unable to impart
of the most important quality parameters of fried food, but this the desired quality attributes of deep fat fried food, such as flavour,
is incompatible with recent consumer trends towards healthier texture, appearance, and mouthfeel.
food and low fat products (Bouchon & Pyle, 2004). The consump- Vacuum frying is a promising technology that may be an option
tion of oil and saturated fat in particular is related to significant for the production of novel snacks such as fruit and vegetable
health problems, including coronary heart disease, cancer, diabe- crisps that present the desired quality attributes and respond to
tes, and hypertension (Saguy & Dana, 2003). Other undesirable new health trends. This deep frying process is carried out in a
effects derived from the high temperatures involved in the frying closed system under pressures well below atmospheric levels,
process and exposure to oxygen are the degradation of important which makes it possible to substantially reduce the boiling point
nutritional compounds and the generation of toxic molecules in of water and thus the frying temperature (Garayo & Moreira,
the foodstuff or the frying oil itself (Fillion & Henry, 1998). This 2002). In fact, most of the benefits of this technology are the result
information has raised a red flag regarding human consumption of the low temperatures employed and minimal exposure to oxy-
of fried food. As a result, healthy, low fat snack products have ac- gen. Said benefits include: (i) the reduction of adverse effects on
quired a new level of importance in the snack industry (Moreira, oil quality (Shyu, Hau, & Hwang, 1998), (ii) preservation of natural
Castell-Perez, & Barrufet, 1999). However, even health-conscious colour and flavours (Shyu & Hwang, 2001), (iii) decreased acrylam-
consumers are unwilling to sacrifice organoleptic properties, and ide content (Granda, Moreira, & Tichy, 2004), and (iv) preservation
intense full-flavour snacks continue to play an important role in of nutritional compounds (Da Silva & Moreira, 2008).
the salty snack market (Mariscal & Bouchon, 2008). The oil uptake mechanism of vacuum frying is still not fully
understood. During a normal operation, the product is placed in-
* Corresponding author. Tel.: +56 2 3547962; fax: +56 2 3545803. side the frying basket once the oil reaches the target temperature.
E-mail address: pbouchon@ing.puc.cl (P. Bouchon). The lid is then closed and the chamber is depressurised.

0308-8146/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2009.08.027
1144 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149

Subsequently, the basket is immersed in the oil bath, where it re- stainless steel (316L) lid, which was thermostatically controlled to
mains for the required amount of time. It is then lifted out and maintain the set frying temperature (±2 °C) using a temperature
the vessel is pressurised using a pressure release valve. This results control system (Micro-controller X, model PXR4, Fuji Electric
in a sudden increase in the surrounding pressure, which may force Instruments, Japan). The fryer basket rod was connected to a rotary
the vapour inside of the pores to condense, which means that oil system, which was used to stir the oil (40 rpm) before frying in or-
absorption may precede cooling (Garayo & Moreira, 2002). How- der to minimise temperature gradients.
ever, as these authors explain, the low pressure may allow air to In vacuum frying experiments, the frying vessel was connected
diffuse faster into the porous structure, obstructing oil passage to a two-stage high vacuum pump (model DVR-140, Dosivac,
and leading to lower oil absorption than is observed in atmo- Argentina) with the capacity to generate a vacuum up to 1.92 in.
spheric frying. This hypothesis has been supported by the results Hg (which corresponds to a water boiling point of approximately
of experiments that suggest that vacuum fried potato crisps absorb 38 °C). In order to prevent water vapour from the product from
less than half the oil of crisps fried under atmospheric conditions. mixing with the oil of the vacuum pump and damaging it, a con-
In order to compare vacuum and atmospheric frying, Mariscal and denser was installed between the devices. (It consisted of a 4-m
Bouchon (2008) defined the term ‘equivalent thermal driving stainless steel pipe that was immersed in an ice–water mixture
force’, which is the difference between the oil temperature and at 0 °C.) In atmospheric frying experiments, the vessel lid was
the boiling point of water at the working pressure. The authors vented by opening a 2 in. metal piece. That piece was tightly af-
used several equivalent thermal driving forces (40 °C, 50 °C and fixed to the lid during the vacuum frying experiments.
60 °C) when frying apple slices under atmospheric or vacuum con-
ditions (4.4 in. Hg). Their experiments showed that vacuum fried 2.3. Frying conditions
apple slices absorbed slightly less oil, and presented better results
for colour preservation than atmospheric fried samples. In both sets of experiments, the fryer vessel was filled with 3 l of
Scant information is available on this topic because scientists are high-oleic acid sunflower oil (Camilo Ferrón Chile S.A, Chile), which
only just beginning to perform research in this field. However, it is was preheated to 160 °C for 1 h prior to frying and discarded after
clear that vacuum frying of non-traditional fruits and vegetables 3 h of frying.
has great potential. In this respect, carrots are an interesting raw Equivalent thermal driving forces were used in both processes.
material to be studied because they are the most important source The thermal driving force was defined by Mariscal and Bouchon
of dietary carotenoids. In addition to acting as natural pigments, (2008) as the difference between the oil temperature and the boil-
these compounds have been linked to the prevention of certain ing point of water at the working pressure (that is, 100 °C under
types of cancer and degenerative and chronic diseases as well as atmospheric conditions and 38 °C for vacuum frying). We used
pro-vitamin A activity (Rodríguez-Amaya, 2001). The processing two driving forces (60 °C and 80 °C), which resulted in frying tem-
of foods may cause major carotenoid degradation due to the isom- peratures of 160 °C and 180 °C for atmospheric frying and 98 °C
erisation of trans-carotenoids, which represent almost 100% of total and 118 °C for vacuum frying.
carotenoid content in raw carrots (Kopas-Lane & Warthesen, 1995),
and oxidation, with a subsequent loss of biological activity and
2.3.1. Vacuum frying experiments
colour deterioration (Rodríguez-Amaya, 2001). Trans–cis isomerisa-
Once the oil reached frying temperature, ten carrot slices
tion affects pro-vitamin A activity, bioavailability, and the antioxi-
(20 g) were placed in the frying basket in order to minimise the
dant capacity of carotenoids. Also, there is some evidence that
drop in temperature. The slices were then covered with a grid to
all-trans-isomers are absorbed preferentially by humans as com-
prevent them from floating. The vessel lid was fastened and the
pared to cis isomers in the case of b-carotene. However, it should
vessel depressurised. When the pressure inside the vessel reached
be noted that isomerisation is only one of the factors that may deter-
1.92 in. Hg, the basket was immersed in the frying oil for increasing
mine their bioavailability; carotenoid bioavailability may be en-
frying times until it reached the bubble-end point. The time re-
hanced by the food matrix, the interaction between carotenoids,
quired to reach said point varied depending on the frying temper-
the presence of dietary fat, and processing conditions, particularly
ature. After each frying time, the basket was lifted out and left to
mechanical disruption and heat treatment (Schieber & Carle, 2005).
stand for 3 min and the vessel was pressurised. The samples were
The objective of this study is to compare oil content, moisture
then removed from the fryer and allowed to cool to room
loss, trans a-and b-carotene retention, and colour development in
temperature.
atmospheric and vacuum deep fat fried carrot crisps made using
equivalent thermal driving forces in order to identify the technol-
ogy’s potential for producing novel snacks that present the desired 2.3.2. Atmospheric frying experiments
quality attributes whilst reflecting new health trends. Once the oil reached the frying temperature, ten carrot slices
were placed in the frying basket and covered with a grid. The bas-
ket was immersed in the frying oil for increasing frying times until
2. Materials and methods it reached the bubble-end point. The time required to reach said
point varied depending on the frying temperature. The basket
2.1. Sample preparation was lifted out and left to stand for 3 min. The samples were then
removed from the fryer and allowed to cool to room temperature.
Carrots (Daucus carota cv. Abaco) were purchased from a local
supermarket and stored at 7 °C and 85–95% relative humidity.
2.4. Experimental considerations
The carrots were washed and cut into 2 mm-thick slices using a
Mandolin Slicer (Danesco International Inc., USA) from which
In order to determine the total amount of oil absorbed by the
3.8 cm discs were extracted.
samples, no de-oiling system was used in any experiment. Re-
ported results correspond to the arithmetic mean of three
2.2. Frying equipment batches ± standard deviation. All batches were handled indepen-
dently and fresh oil was used for each batch. All determinations
Both atmospheric and vacuum frying were carried out using an and measurements were carried out in triplicate on each batch un-
electrically heated, 10-l stainless steel (316L) vessel covered with a less otherwise specified.
 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149 1145

3. Analytical methods at room temperature. The mixture was poured through a separato-
ry funnel and the carotenoid solution was washed five times with
3.1. Oil content water to remove the alkali. It then was filtered through anhydrous
sodium sulphate. The same process was followed for the fresh
Total oil content of ground carrot crisps was determined gravi- sample extracts for HPLC analysis.
metrically by Soxhlet extraction with petroleum ether (AOAC,
1995). 3.5.2. Preparation of standard solutions
Both all-trans-a- and all-trans-b-carotene standards were ob-
3.2. Solids content tained from carrots (Daucus carota cv. Abaco). Concentrations of
standards in petroleum ether were determined by spectrophotom-
Each extracted oil-free sample was placed in a Petri dish, dried etry using their respective A1%
1cm (Rodríguez-Amaya, 2001). Calibra-

in a forced air oven at 105 °C to constant weight, and cooled in a tion curves were obtained for each carotenoid, diluting standard
desiccator (AOAC, 1995). Solid content of raw carrot slices was solutions and measuring associated peaks with HPLC.
determined using the same procedure.
3.5.3. Chromatographic procedure
3.3. Moisture loss Carotenoid analysis was carried out by HPLC using a Waters
symmetry column (C18, 5 lm particle size, 4.6 mm i.d  25 cm;
Moisture loss was expressed on a dry basis and calculated using Waters, USA). An isocratic mobile phase of methanol:acetoni-
the difference between the original moisture content and the mois- trile:ethyl acetate (20:65:15 by vol) was used at a rate of 1 ml/
ture content at time t. min. The HPLC system consisted of a Merck-Hitachi L-6200 pump
and a Waters 996 Photodiode Array detector, which were hooked
up to a computer with Millennium 32 software. Carotenoids were
3.4. Colour analysis using a computer vision system
detected at 450 nm and identified by comparing peak retention
times with standards. Cis-b-carotene appeared as a peak following
A colour digital camera model PowerShot A70 (Canon, USA)
its trans-isomer.
connected to a computer via USB interface IFC-300PCU (Canon,
USA) was mounted on a stand inside a large opaque box with black
3.5.4. a and b carotene retention
interior surfaces. The iris was operated using the manual mode,
The absolute retention factor of each carotenoid was expressed
with lens aperture at f = 8 and speed 1/3(1/6) (no flash, no zoom)
according to Fillion and Henry (1998) on a dry-weight fat free basis
in order to achieve high uniformity and repeatability. The samples
as follows:
were illuminated using four CIE source D65 lamps (60 cm length
and 18 W; Model TLD/965, Phillips, Singapore) at 45° angle in or- nutrient content per g cooked food
Absolute retention ¼
der to maximise diffuse reflection, which is responsible for colour. nutrient content per g raw food
The angle between the camera lens axis and the sample was set at  ðdry weight; fat freeÞ  100 ð1Þ
around 90° in an effort to reduce gloss. The camera was grey bal-
anced before each imaging session with a Kodak grey card with
18% reflectance (L* = 50). Afterwards, calibration samples were 3.6. Water activity (aw) determination
placed in the camera’s view using a white card. A 1600  1200 pix-
el image was acquired and stored in high resolution and superfine The aw of vacuum and atmospheric fried carrot crisps was mea-
quality JEPG (Joint Photographic Experts Group) format in RGB col- sured using a Lufft aw-wert-Messer (Durotherm, Germany) at
our coordinates. The RGB colour images were converted to CIELAB 20 °C.
or Lab values using Adobe Photoshop 6.0 software (Adobe Systems
Inc., USA), which were normalised to L*, a*, b* as explained in 3.7. Texture
Mariscal and Bouchon (2008).
The texture of the fried carrot crisps was analysed using the
3.5. a and b carotene determination procedure described by Da Silva and Moreira (2008), which con-
sists of a three-point bending test. The sample was supported using
3.5.1. Carotenoid extraction from raw carrot slices and carrot crisps two parallel edges in order to apply the load centrally. The system
Carotenoid content was determined using the methodology was mounted in a TA.XT2 Texture Analyser (Texture Technologies
described by Robert, Carlsson, Romero, and Masson (2003). Carote- Corp., USA) using a support span of 16 mm. A 2.5 mm-thick steel
noids were extracted from raw carrot slices by homogenising blade with flat edge was used to fracture the sample at a constant
about 1 g of sample with Celite and cold acetone with a mortar rate of speed of 10 mm s1. The force (N) at the fracture point
and pestle. The slurry was vacuum filtered and the solid residue (highest value in the plot) was used as the resistance to breakage.
was re-extracted with cold acetone. Four extractions were made
until no orange colour remained in the solid residue. The combined 3.8. Statistical analysis
acetone extracts were transferred into petroleum ether. The organ-
ic phase was washed with distilled water to remove residual ace- Statistical analysis was executed using Statgraphics 5.0 soft-
tone and filtered through anhydrous sodium sulphate. Then, the ware (Manugistic Inc., USA). One-way variance analysis was car-
solvent of an aliquot of 200 ll was evaporated to dryness under ried out in order to confirm that there were no significant
a nitrogen stream. Extracts were dissolved in 1000 ll of acetone differences amongst fried sample measurements under specific
prior to analysis by HPLC. conditions. Disparities between samples fried under different con-
Carotenoids from fried carrot crisps were extracted from about ditions were determined through confidence interval analysis
1 g of ground sample with petroleum ether for 4 h and then vac- using the Bonferroni test. All significant differences were deter-
uum filtered. The extract was made up to 100 ml with petroleum mined with a confidence level of 95%. Linear correlation of colour
ether and an aliquot of 5 ml was saponified overnight with an parameters and carotenoid content was also examined by carrying
equal volume of 10% potassium hydroxide in methanol solution out a linear regression between a specific colour parameter and
1146 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149

carotenoid content (total carotenoids, trans a-carotene, or trans b-

carotene content). Then, the coefficient of determination (r2) for
the linear regression model was calculated.

4. Results and discussion

4.1. Moisture loss

Fig. 1 shows moisture loss for each frying time during vacuum
and atmospheric frying of carrot slices using thermal driving forces
of 60 °C and 80 °C. The loss of moisture during vacuum and atmo-
spheric frying presented a classical drying profile. There was an ini-
tial rapid decrease in water content, which was mainly due to the
loss of surface and unbound inner water, followed by a gradually
decreasing gradient due to crust formation.
Fig. 2. Oil uptake during atmospheric (AF) and vacuum frying (VF) of carrot crisps
All samples were dried to the same final moisture content (bub- when using thermal driving forces of 60 °C (Toil (VF) = 98 °C and Toil (AF) = 160 °C) and
ble-end point, 2% w.b.) The frying temperature and technology 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180 °C). Points are means ± standard deviations
(pressure) significantly affected the rate of moisture loss and the (n = 3).
time required to achieve the desired level of dehydration (see
Fig. 1). Samples fried under atmospheric conditions using a ther-
mal driving force of 80 °C achieved bubble-end point after 2 min,
whilst samples fried under vacuum conditions needed 3 min.
Something similar was observed when the samples were fried
using a thermal driving force of 60 °C. Samples fried under atmo-
spheric and vacuum conditions needed 3 and 5 min, respectively,
to achieve bubble-end point. The differences are thought to be
mainly associated with micro-structural changes. Though the same
thermal driving force is used in both processes, samples fried
under vacuum conditions are exposed to lower temperatures. As
a result, micro-structural changes/damage are impaired. (This is
one of the main advantages of vacuum technology.) Also, during
the initial depressurisation step of vacuum frying, micro-structural
surface changes may occur, which may prevent water from escap-
ing. Furthermore, even though dehydration is mainly limited by
heat transfer, diffusion may play a role. Diffusion slows down at Fig. 3. Oil content versus moisture loss during atmospheric (AF) and vacuum frying
lower temperatures, a factor that may preclude moisture loss. (VF) of carrot crisps when using thermal driving forces of 60 °C (Toil (VF) = 98 °C and
Toil (AF) = 160 °C) and 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180 °C). Points are
Similar results were found by Mariscal and Bouchon (2008) when
means ± standard deviations (n = 3).
frying apple slices.

4.2. Oil absorption gradient, with a considerable amount absorbed during the initial
period. There were great differences between samples fried under
Fig. 2 shows the evolution of the oil uptake of carrot crisps for atmospheric and vacuum conditions and between those fried using
increasing atmospheric and vacuum frying times using thermal different thermal driving forces.
driving forces of 60 °C and 80 °C. The general pattern of oil uptake Oil absorption has long been claimed to be a surface-related
shows an initial rapid increase followed by a gradually decreasing phenomenon resulting from the competition between drainage
and suction into the porous crust once the food is removed from
the oil bath and begins to cool (Bouchon et al., 2003; Gamble, Rice,
& Selman, 1987; Moreira, Sun, & Chen, 1997). Consequently, oil
absorption is heavily linked to moisture loss since it determines
the extent of crust formation and therefore the volume that is
available for oil infiltration. Fig. 3 shows oil uptake versus moisture
loss during vacuum and atmospheric frying of carrot slices. The
strong relationship between them suggests that the water vapour
replacement mechanism, which is said to occur during atmo-
spheric frying, would also take place during vacuum frying, as
determined by Mariscal and Bouchon (2008). Statistical analysis
showed that both the frying method and the thermal driving force
had a significant effect (p < 0.05) on oil uptake (g oil/g dry solids)
for the same dehydration level. In fact, vacuum fried carrot slices
absorbed 47% and 50.5% (d.b.) less oil than atmospheric fried ones
at bubbled-end point when using thermal driving forces of 60 °C
and 80 °C, respectively. When comparing vacuum fried carrot
Fig. 1. Moisture loss during atmospheric (AF) and vacuum frying (VF) of carrot
crisps when using thermal driving forces of 60 °C (Toil (VF) = 98 °C and Toi-
crisps, those fried using a thermal driving force of 60 °C absorbed
l (AF) = 160 °C) and 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180 °C). Points are a significantly lower amount of oil compared to those fried using
means ± standard deviations. a thermal driving force of 80 °C. These results concur with those
 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149 1147

obtained by Garayo and Moreira (2002), who determined that the In addition, b-carotene all-trans-isomer is absorbed preferentially
oil content of vacuum fried potato crisps (37% d.b.) was signifi- by humans compared to cis isomer (Schieber & Carle, 2005).
cantly lower than that of potato crisps fried under atmospheric Our results suggest that trans b-carotene degradation is faster
conditions (66% d.b.) The authors explained that during a normal than trans a-carotene degradation in all frying conditions. The deg-
vacuum frying operation, the product is removed from the oil bath radation rate of carotenoids depends on their structure. Key as-
and the vessel is vented before cooling takes place. This results in a pects are the number of double bonds and the maximum overlap
sudden increase in the surrounding pressure at a constant temper- of carbon–carbon double bond molecular orbitals, which give the
ature, which may force the vapour inside of the pore to condense, highest susceptibility (Miller, Sampson, Candeias, Bramley, &
decreasing Ppore and therefore initiating oil absorption before cool- Rice-Evans, 1996). Both trans a- and trans b-carotene have 11 dou-
ing begins (that is, Psurroundings  Ppore > 0). They further explain that, ble bonds. Their structural difference lies in the cyclisation of their
because of the low pressure, air may diffuse faster into the porous ends: trans b-carotene has one b-ring on each side of the molecule
space, obstructing oil passage and therefore leading to a reduction whereas trans a-carotene has one b-ring and a e-ring. Steric hin-
in oil uptake in vacuum fried snacks as compared to atmospheric drance reduces the orbital overlap between the chain double bonds
fried ones. Mariscal and Bouchon (2008) also observed lower oil and the b-ring double bonds in the molecules of trans a- and trans
uptake in vacuum fried apple crisps than in atmospheric fried ones. b-carotene. This leaves trans b-carotene with 9 fully overlapping
They attributed this to the lower vapour-pressure of water during double bonds plus 2 b-ring double bonds with reduced overlap,
vacuum frying and to the higher temperatures reached during and trans a-carotene with 9 fully overlapping double bonds plus
atmospheric frying, which induce added structural changes like only one b-ring conjugated double bond, as the double bond of
tissue/constituents degradation. However, the decrease was less the e-ring is not part of the conjugated double bond system (Ang-
pronounced in apples than in carrot slices. This is probably due uelova & Warthesen, 2000). Thus, trans b-carotene has a higher
to the high porosity of apples. susceptibility to oxidation and isomerisation because it has 11 con-
jugated double bounds, whereas trans a-carotene only has 10.

4.3. Carotenoid content

4.4. Colour development
The most abundant carotenoid found in fresh and fried carrot
samples was trans b-carotene, which represents about 60% of total Carrots’ colour is largely related to the presence of carotenes. As
carotenoid content. The second-most abundant was trans a-caro- reported by Sulaeman et al. (2001), the orange colour of carrots
tene, which accounts for nearly 40%. These results are in agreement and carrot crisps can be described by lightness (L*), redness (a*),
with the literature that has reported that trans b-carotene content yellowness (b*), and Hue parameters. Table 2 shows changes in
ranges from 60% to 80%, whereas trans a-carotene content ranges L*, a* and b* values of carrot crisps for increasing frying times dur-
from 10% to 40% (Chen, Peng, & Chen, 1995). ing atmospheric and vacuum frying. Hue values are not included
Table 1 shows absolute retention of trans a- and trans b-caro- since we did not find a significant relationship between this
tene after vacuum and atmospheric frying of carrot slices using parameter and processing conditions.
thermal driving forces of 60 °C and 80 °C. A thermal driving force L* is a critical parameter in the frying industry as it is usually
of 60 °C during vacuum frying results in minor degradation of both the first quality attribute evaluated by consumers when determin-
trans a- and trans b-carotene (only 10.5% and 13.6% degradation, ing product acceptance. Low L* values indicate a dark colour and
respectively). When the thermal driving force was increased are mainly associated with non-enzymatic browning reactions.
(DT = 80 °C), absolute retention was about 60%, which is still much As shown in Table 2, L* value decreased significantly during atmo-
higher than the absolute retention found in atmospheric fried spheric frying for both thermal driving forces. For instance, L* value
crisps. diminished from L0* = 65.2 to L* = 34.2 at bubble-end point (much
Thermal processing may result in losses of all-trans-carotenes darker) when frying at 180 °C. On the other hand L* remained
and the formation of cis isomers. In fact, a reduction in the percent- nearly constant during vacuum frying, only decreasing to
age of all-trans-b-carotene with the concomitant increase of 13-cis L* = 58.4 and L* = 55.9 when using thermal driving forces of 60 °C
and 9-cis isomers has been observed during processing (Chandler & and 80 °C, respectively.
Schwartz, 1988). Losses of all-trans-a- and all-trans-b-carotene The degradation of trans a- and trans b-carotene had a big im-
were greater during atmospheric frying, probably as a result of pact on a* and b* values, as shown in Table 2. Crisps fried under
their isomerisation at elevated temperatures. In fact, formation of vacuum conditions using a thermal driving force of 60 °C main-
13-cis-b-carotene was observed in carrot crisps fried at 160 °C tained their orange-yellow chromaticity, showing a slight decrease
and 180 °C, representing 12.5 ± 1.1% and 28.6 ± 1.4% of total caro- in their colour coordinates, which varied from a0* = 49.9 to
tenes, respectively. Chandler and Schwartz (1988) noted that cis a* = 45.4 and from b0* = 65.4 to b* = 65.1. Carrot slices fried under
isomers could account for as much as 29% of total b-carotene con- atmospheric conditions showed a significant decrease in their col-
tent in processed sweet potatoes. Trans–cis isomerisation affects our coordinates, reaching final values of a* = 15.3 and b* = 25.3,
pro-vitamin A activity and the antioxidant capacity of carotenoids. when frying at 180 °C. Our results correspond to those of Shyu,

Table 1
Trans a and trans b-carotene absolute retention, water activity and kinetic constants of textural changes during vacuum and atmospheric frying of carrot crisps when using
thermal driving forces of 60 °C (Toil (VF) = 98 °C and Toil (AF) = 160 °C) and 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180 °C).

Frying pressure Thermal driving Trans a-carotene Trans b-carotene absolute Water activity at Texture kinetic constants
(inch Hg) force (°C) absolute retention at retention at bubble-end point
Softening kinetic Hardening kinetic
bubble-end point (%) bubble-end point (%)
constant ks (s1) constant kh (s2)
1.92 60 89.5 ± 2.5 86.4 ± 4.3 0.442 ± 0.015 1.8582 0.0441
80 60.4 ± 1.7 58.6 ± 3.0 0.452 ± 0.01 1.9655 0.1209
29.92 60 35.7 ± 1.4 31.6 ± 1.8 0.439 ± 0.015 3.5218 0.1322
80 33.9 ± 1.3 27.9 ± 1.5 0.436 ± 0.012 4.1194 0.2775
1148 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149

Table 2
Evolution of lightness (L*), green–red chromaticity (a*) and blue–yellow chromaticity (b*) during atmospheric (AF) and vacuum frying (VF) of carrot crisps when using thermal
driving forces of 60 °C (Toil (VF) = 98 °C and Toil (AF) = 160 °C) and 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180 °C).

Colour coordinate Frying pressure (in. g) Thermal driving force (°C) Frying time (min)
0 1 2 3 5
L* 1.92 60 62.6 ± 0.7 60.7 ± 1.1 59.1 ± 1.7 58.4 ± 1.8
80 65.2 ± 0.47 61.1 ± 0.7 59.3 ± 1.7 55.9 ± 0.7 –
29.92 60 55.5 ± 1.3 47.9 ± 0.8 43.1 ± 0.8 –
80 52.4 ± 1.6 34.2 ± 1.0 – –
a* 1.92 60 47.8 ± 1.5 46.7 ± 2.5 45.9 ± 2.5 45.4 ± 2.2
80 49.9 ± 1.0 48.1 ± 1.9 45.9 ± 2.4 37.1 ± 3.2 –
29.92 60 45.2 ± 1.9 38.1 ± 2.8 24.9 ± 2.0 –
80 34.4 ± 3.3 15.3 ± 3.9 – –
b* 1.92 60 64.0 ± 3.5 63.1 ± 3.2 64.1 ± 2.5 65.1 ± 2.5
80 65.4 ± 0.64 62.3 ± 2.6 61.5 ± 2.7 60.1 ± 3.5 –
60 62.8 ± 1.3 53.6 ± 0.9 41.8 ± 3.1 –
80 61.0 ± 1.4 25.3 ± 2.0 – –

Hau, and Hwang (2005), who reported that the values of a* and b*
decrease significantly as the frying temperature increases above
100 °C because of carotenoids’ instability. In fact, we found good
linear correlations between a* and trans b-carotene content
(r2 = 0.95), b* and trans a-carotene content (r2 = 0.78) and hue
and total carotenoids content (r2 = 0.91) when comparing values
of fried crisps at bubble-end point.

4.5. Water activity (aw)

As shown in Table 1, samples reached a water activity (aw) aver-

age value of approximately 0.44. This value matches the results ob-
tained by Katz and Labuza (1981), who found that potato crisps’
sensory acceptability decreases markedly above a critical value of
0.47. Furthermore, no significant differences were found amongst Fig. 4. Textural changes during atmospheric (AF) and vacuum frying (VF) of carrot
the samples. crisps when using thermal driving forces of 60 °C (Toil (VF) = 98 °C and Toi-
l (AF) = 160 °C) and 80 °C (Toil (VF) = 118 °C and Toil (AF) = 180°C). Points are
Water activity affects the stability of dehydrated foods because
means ± standard deviations (n = 10). Continuous lines represent the model from
it determines both chemical reaction rates and microbial activity. Eq. (2).
The limiting value of aw for the growth of any microorganism is
around 0.6. Below this value, food spoilage is mainly due to enzy-
where MF* = normalised maximum force; ks = kinetic constant for
matic and chemical reactions, such as oxidation (Adams & Moss,
softening of carrot tissue during frying (s1); kh = kinetic constant
1995). The rate of oxidation may rise as water content increases
for the crust hardening process during frying (s2) and t = frying
since water can enhance reactant mobility and bring catalysts into
time (min).
solution. In relation to carotenoid stability. Lavelli, Zanoni, and
The final maximum force (MF) of fried carrot crisps was not af-
Zaniboni (2007) showed that both trans a- and b-carotene from
fected by the frying technology and the thermal driving force used.
dehydrated carrots present their maximum stability at aw values
All samples reached final values of between 5.01 and 5.68 N. Nev-
between 0.34 and 0.54. These results confirm previous findings
ertheless, the frying technology significantly affected the rate of
by other authors (Baloch, Buckle, & Edwards, 1977).
softening and hardening. For instance, when using a thermal driv-
ing force of 60 °C, ks values were 1.8582 and 3.5218 s1 and kh val-
4.6. Texture
ues were 0.0441 and 0.1322 s2 when frying under vacuum and
atmospheric conditions, respectively. Also, higher frying tempera-
Frying of raw vegetables induces major changes in their micro-
tures resulted in faster softening of the tissue and subsequent
structure, which in turn determine their final physical and sensory
hardening (see Table 1 and Fig. 4).
properties. The most important textural attribute of crisps and
chips is crispness, which denotes freshness and high quality. A
crisp should be firm and snap easily when bent, emitting a crunchy 5. Conclusions
sound (Krokida, Oreopoulou, Maroulis, & Marinos-Kouris, 2001).
During frying, vegetable tissues show an initial softening that is Vacuum-fried carrots may be a promising snack category due to
followed by hardening due to the progressive development of a the fact that carrots are the most important source of dietary
dehydrated crust. In order to incorporate both phenomena, Pedre- carotenoids, and this technology makes it possible to overcome
schi and Moyano (2005) modelled textural changes (normalised major carotenoid degradation pathways due to isomerisation and
maximum force) during frying by means of two terms (see Eq. oxidation and thus preserve biological activity. The results showed
(2)). The first describes the softening of the tissue for short frying that vacuum fried crisps (driving force of 60 °C) may reduce the oil
times (fast phase) and the second describes a hardening dependent content of carrot crisps by nearly 50% (d.b.) compared to atmo-
component (slow phase), which is a function of the square of the spheric fried crisps produced using the same driving force. Further-
frying time and gains relevance at longer frying times. more, they preserve around 90% of trans a-carotene and 86% trans
MF  ¼ eks t þ kh  t2 ð2Þ b-carotene, which leads to the preservation of the colour of raw
carrots. This is reflected by L*, a*, b* colour coordinate analyses,
 V. Dueik et al. / Food Chemistry 119 (2010) 1143–1149 1149

excellent linear correlations between a* and trans b-carotene con- Granda, C., Moreira, R., & Tichy, S. (2004). Reduction of acrylamide formation in
potato chips by low-temperature vacuum frying. Journal of Food Science, 69,
tent (r2 = 0.95), b* and trans a-carotene content (r2 = 0.78), and hue
405–411.
and total carotenoids content (r2 = 0.91) when comparing values of Katz, E. E., & Labuza, T. P. (1981). Effect of water activity on the sensory crispness
fried crisps at bubble-end point. As a result, vacuum frying may be and mechanical deformation of snack food products. Journal of Food Science, 46,
a useful process in the production of novel snacks that present de- 403–409.
Kopas-Lane & Warthesen (1995). Carotenoids photostability in raw spinach and
sired quality attributes and respond to new health trends. carrots during cold storage. Journal of Food Science, 60, 773–776.
Krokida, M. K., Oreopoulou, V., Maroulis, Z. B., & Marinos-Kouris, D. (2001). Effect of
Acknowledgements pre-treatment on viscoelastic behaviour of potato strips. Journal of Food
Engineering, 50, 11–17.
Lavelli, V., Zanoni, B., & Zaniboni, A. (2007). Effect of water activity on carotenoids
The authors appreciate the financial support received through degradation in dehydrated carrots. Food Chemistry, 104, 1705–1711.
FONDECYT Project No. 1070764 and Grants from Conicyt and the Mariscal, M., & Bouchon, P. (2008). Comparison between atmospheric and vacuum
frying of apple slices. Food Chemistry, 107(4), 1561–1569.
PUC School of Engineering. Miller, N. J., Sampson, J., Candeias, L. P., Bramley, P. M., & Rice-Evans, C. A. (1996).
Antioxidant activities of carotenes and xanthophylls. FEBS Letters, 384, 240–242.
References Moreira, R., Castell-Perez, M. E., & Barrufet, M. A. (1999). Deep-fat frying:
Fundamentals and applications. Gaithersburg, MD: Aspen Publishers Inc..
Moreira, R., Sun, X., & Chen, Y. (1997). Factors affecting oil uptake in tortilla chips in
Adams, M. R., & Moss, M. O. (1995). Food microbiology. Cambridge: The Royal Society
deep-fat frying. Journal of Food Engineering, 31, 485–498.
of Chemistry.
Pedreschi, F., & Moyano, P. (2005). Oil uptake and texture development in fried
Anguelova, T., & Warthesen, J. (2000). Degradation of lycopene, a-carotene, and b-
potato slices. Journal of Food Engineering, 70, 557–563.
carotene during lipid peroxidation. Journal of Food Science, 65(1), 71–75.
Robert, P., Carlsson, R. M., Romero, N., & Masson, L. (2003). Stability of spray-dried
AOAC (1995). Official methods of analysis (16th ed.). Washington: Association of
encapsulated carotenoid pigments from Rosa Mosqueta (Rosa rubiginosa)
Official Analytical Chemists.
Oleoresin. Journal of the American Oils Chemical Society, 80(11), 1–6.
Baloch, A. K., Buckle, K. A., & Edwards, R. A. (1977). Stability of b-carotene in model
Rodríguez-Amaya, D. (2001). A guide to carotenoid analysis in foods. <http://
systems containing sulphite. Journal of Food Technology, 12, 309–316.
www.ilsi.org/file/carotenoid.pdf>. [ISBN 1-57881-072-8].
Bouchon, P., Aguilera, J. M., & Pyle, D. L. (2003). Structure oil-absorption
Saguy, S., & Dana, D. (2003). Integrated approach to deep fat frying: Engineering,
relationships during deep-fat frying. Journal of Food Science, 68, 2711–2716.
nutrition, health and consumer aspects. Journal of Food Engineering, 56,
Bouchon, P., & Pyle, D. L. (2004). Studying oil absorption in restructured potato
143–152.
chips. Journal of Food Science, 69, E115–E122.
Schieber, A., & Carle, R. (2005). Occurrence of cis-isomers in food: Technological,
Chandler, L. A., & Schwartz, S. J. (1988). Isomerization and losses of trans-beta-
analytical, and nutritional implications. Trends in Food Science and Technology,
carotene in sweet potatoes as affected by processing treatments. Journal of
16, 416–422.
Agricultural and Food Chemistry, 36, 129–133.
Shyu, S., Hau, L., & Hwang, L. (1998). Effect of vacuum frying on the oxidative
Chen, B. H., Peng, H. Y., & Chen, H. E. (1995). Changes of carotenoids, color, and
stability of oils. Journal of the American Oils Chemical Society, 75(10), 1393–1398.
vitamin A content during processing of carrot juice. Journal of Agricultural and
Shyu, S., Hau, L., & Hwang, L. (2005). Effects of processing conditions on the quality
Food Chemistry, 43, 1912–1918.
of vacuum-fried carrot chips. Journal of the Science of Food and Agriculture, 85,
Da Silva, P., & Moreira, R. (2008). Vacuum frying of high-quality fruit and vegetable-
1903–1908.
based snacks. Lebensmittel – Wissenschaft und Technologie, 41(10), 1758–1767.
Shyu, S., & Hwang, L. (2001). Effect of processing conditions on the quality of
Fillion, L., & Henry, C. J. K. (1998). Nutrient losses and gains during frying: A review.
vacuum fried apple chips. Food Research International, 34(2–3), 133–142.
International Journal of Food Sciences and Nutrition, 49(2), 157–168.
Sulaeman, A., Keeler, L., Taylor, S. A., Giraud, D. W., Taylor, S. L., Wehling, R. L., et al.
Gamble, M. H., Rice, P., & Selman, J. D. (1987). Distribution and morphology of oil
(2001). Carotenoid content and physicochemical and sensory characteristics of
deposits in some deep fried products. Journal of Food Science, 52, 1742–1743.
carrot chips deep-fried in different oils at several temperatures. Journal of Food
Garayo, J., & Moreira, R. (2002). Vacuum frying of potato chips. Journal of Food
Science, 66(9), 1257–1264.
Engineering, 55, 181–191.