Journal of the Science of Food and Agriculture J Sci Food Agric 80:777±782 (2000)

Development and application of soy-proteinfilms to reduce fat intake in deep-fried foodsM Rayner, V Ciolfi, B Maves, P Stedman and GS Mittal*School of Engineering, University of Guelph, Guelph, Ontario, Canada, N1G 2W1

(Rec

* CoE-ma

# 2

Abstract: A soy protein ®lm coating was developed and evaluated to reduce fat transfer in deep-fried

foods during frying. Soy protein isolate solutions (10% SPI) with 0.05% gellan gum as plasticizer cooled

after being held at 80°C for 20min provided suitable ®lms. There was a signi®cant fat reduction

(55.12(�6.03)%db) between fried uncoated and coated discs of doughnut mix. The same ®lms were

used on potato fries. Some panellists observed a slight difference between the coated and uncoated fries

but many preferred the coated fries over the uncoated ones. Penetration test on potato fries showed no

signi®cant difference between the texture of coated (SPI with gellan gum) and the uncoated fried

samples. A solution of 10% SPI with 0.05% gellan gum is recommended for coating foods to reduce fat

intake during deep-fat frying.

# 2000 Society of Chemical Industry

Keywords: soy-protein ®lm; edible ®lm; frying; fat reduction; low-fat food

INTRODUCTIONIn the deep-fat frying process, the oil serves as the heat

transfer medium. However, the oil also migrates into

the food thereby causing health concerns. Despite the

health risks associated with high fat diets, fried foods

are still being consumed because of their great taste. In

the United Kingdom, potato chips represent approxi-

mately one-third of the total processed potato market.

Potato products vary from 30% to in excess of 40% by

weight in oil content.1 A reduction in fat content is a

goal many producers are striving for. In many cases

reducing the fat content also affects organoleptic

properties such as taste and mouthfeel. Recent

research by Williams and Mittal2 has successfully

utilized edible ®lm coatings to substantially reduce fat

content in fried foods while maintaining food quality.

Extensive research work has been done in recent

years on the physical and mechanical properties of

edible ®lms but not much has been reported on their

applications. Williams and Mittal2 tested gellan gum

and cellulose ®lms for fat and moisture barrier

properties. Results showed that the ®lm coating

reduced fat absorption by as much as 50±90% in

foods that were deep-fat-fried. The ®lm or coating can

be applied to food products by spraying, immersion, or

direct application.3 Williams and Mittal2 experimen-

ted with both the spraying and immersion methods of

application. The immersion method was preferred

because coatings were much more evenly spread and it

was easier to implement.

Other researchers4,5 also demonstrated the effec-

tiveness of edible coatings for low-fat starchy and

eived 8 June 1999; revised version received 1 December 1999; acce

rrespondence to: GS Mittal, School of Engineering, University of Gueil: gmittal@uoguelph.ca

000 Society of Chemical Industry. J Sci Food Agric 0022±5142/2

cowpea products. Coated samples were soggy and less

brown compared to uncoated samples possibly due to

the non-optimization of the coatings applied. Use of

coatings on marinated chicken strips has decreased

free fatty acid in frying oil by 25%6 Edible ®lms on

chicken strips were not objectionable to the con-

sumers.7

Soy protein (SP) ®lms can be produced by heating

aqueous soy protein isolate (SPI) dispersions to form

surface ®lms or by depositing and drying SP-based

®lm-forming solutions in molds. Bates and Wu8

formed ®lms from SPI solutions which were heated

in stainless steel pans at 85°C. Films formed on the

surface and were removed using a glass rod. They

considered the concentration of aqueous SPI disper-

sions on ®lm characteristics. Some of their optimal

combinations included: 4.3% SP at pH 8.5 and 5.3%

SP at pH 9.5. At higher protein concentrations (above

7%), ®lm formation was hindered as gelation of the

mixture occurred. Gennadios et al9 made ®lm-forming

solvents using 5% SPI and 3% glycerin to act as

plasticizer to overcome ®lm brittleness. Sodium

hydroxide was used to adjust the pH to 11. The

solution was heated to 70°C and held for 30min.

Excessive temperatures are undesirable when form-

ing a ®lm because they result in an excessive rate of

evaporation during ®lm drying. This may result in ®lm

defects such as pinholes or non-uniformity in the ®lm.3

The temperature of heating is also important to the

®lm colour. Gennadios et al9 concluded that the heat

treatment of a ®lm induced an increase in ®lm

yellowness. Heating the SPI ®lm increased the tensile

pted 20 January 2000)

lph, Guelph, Ontario, Canada, N1G 2W1

000/$17.50 777

M Rayner et al

strength with time. This also caused a decrease in ®lm

elongation and thus elasticity. Heat-induced cross-

linking in the ®lm structure contribute to the increase

in toughness and the decrease in ®lm ¯exibility. Water

solubility decreased with a temperature increase from

80 to 95°C. The increase in heat treatment of SPI

®lms signi®cantly reduced their water vapour perme-

ability. This decrease was due to the formation of

covalent links within the ®lms during heating and the

lower water solubility of heated ®lms indicating a

decrease in protein hydrophobicity.

The formation of a homogenous free standing SPI

®lm could be achieved in the pH ranges of 1±3 and

6±12.10 The ®lm did not form between pH 4 and 5 but

rather coagulated around its isoelectric point (pH 4.5).

When moving away from the isoelectric point, the SPI

proteins denature, unfold and solubilize exposing

sulfhydryl and hydrophobic groups. These groups

associate during drying creating hydrophobic and

disul®de bonding forces which form a ®lm structure.

SPI ®lms in pH ranges of 6±11 had signi®cantly higher

tensile strengths, higher percentage elongation at

break and lower water vapour permeability than ®lms

formed at pH ranges of 1±3. Gennadios et al10

concluded that SPI ®lm-forming solutions did not

require ethanol. Less heat energy was required to dry

the ethanol solvent ®lm than a 100% aqueous ®lm.

Glycerin is commonly used as plasticizing agent in

3±5% w/v concentrations. It acts to increase the intra-

molecular spacing of the protein matrix allowing a

greater degree of ¯exibility. However, it reduces

overall tensile strength. Gennadios et al10 reported

that glycerin also affects the water vapour permeability

as glycerin is hydrophilic and thus favours water

adsorption through ®lms. It also increases the ®lm

boiling point. The boiling point of glycerin is 290°Cbut drops substantially in presence of water. This

improves ®lm quality as glycerin decreases the rate of

®lm formation and thus avoids the formation of

pinholes.3

This paper investigates the development and appli-

cation of soy protein ®lm coatings on fried foods (discs

and balls of doughnut mix, fries and discs of potatoes)

to reduce the fat absorption during deep-fat frying.

MATERIALS AND METHODSFilm formulation and formationThe formation of soy protein ®lms revolve around the

Table 1. Soy protein ingredient sources for makingedible films

Name

Low fat soy ¯our

Supro M9

Arcon S

Supro M57

Supro 710

Ardex F

Pro Fam 981

778

selection of formulation ingredients for the soy protein

dispersion. The alternative formulations considered

different sources of soy protein, application method,

and the use of plasticizing agents. Three types of soy-

protein were tried: low fat soy ¯our (SPF), soy protein

concentrates (SPC), and soy protein isolates (SPI)

(Table 1). Once the feasible alternative(s) were

determined the best application method was chosen.

Plasticizing agents can impart bene®cial physical

properties to the ®lm. Glycerin and gellan gum were

tried as plasticizing agents.

Soy-protein was weighed out to represent 5% SP

(w/w) in a 200ml solution, mixed for 20min using a

mixer at medium speed avoiding ®lm formation. The

soy protein solution was titrated with 0.1N NaOH

until pH read 8.0. The solution was heated to 75±

85°C by a hot plate and the temperature was

monitored with a thermocouple probe and a data

logger. After waiting 4-min the ®lm formed on the

surface was removed using a glass slide and inspected.

This was repeated ®ve times at 4-min intervals using

the same solution. The ®lms were dried in the con-

vection oven on `dehydration' heat level (<200°C),

visually inspected, and observations were made.

Thermal effects on film formationThe alternatives (SPI materials Profam 981 and Ardex

F) that passed the preliminary ®lm formation tests

were used in further testing. In this procedure, the

solution was continually mixed and the ®lm formation

occurred after dipping, as the slide was dried in the low

heat convection oven. The ®lm formed as the water

evaporated. This method was more desirable than

others as it has a greater ease of scale-up and allows for

a more consistent product. Further, the purpose was

to determine the effect of heat treatment of the

solution on the resulting ®lm formation. SPI was

weighed out to represent 10% solute (w/w) in a 200ml

solution, and mixed using a mixer at medium speed

avoiding ®lm formation for 20min. The soy protein

solution was titrated with 0.1N NaOH until pH read

8.0. The sample was placed on a heating/mixing plate

and mixing continued at low speed throughout the

experiment. The ®rst slide dip was made when the

solution was at room temperature (21°C). The solu-

tion was then heated to 80±85°C by a hot plate and the

temperature was monitored. The second slide dip

sample was made after 20min at the heated tempera-

ture. The solution was allowed to cool to room

Type Source company Soy protein (%)

SPF Bulk Barn 46%

SPC Protein Technologies International 57%

SPC Protein Technologies International 70%

SPC Protein Technologies International 57%

SPI Protein Technologies International 90%

SPI UFL Foods 90%

SPI UFL Foods 90%

J Sci Food Agric 80:777±782 (2000)

Table 2. Triangle test set-up for sensory evaluation on deep-fatfried potato fries

Test Sample 1 Sample 2 Sample 3

A Uncoated SPI Uncoated

B SPI�gellan gum SPI SPI

C SPI�gellan gum SPI�gellan gum Uncoated

Soy-protein ®lms to reduce fat in fried foods

temperature (21°C) and then the third dip slide

sample was made. The ®lms were dried in the con-

vection oven on `dehydration' heat level (<200°C),

visually inspected, and observations were made.

With the dispersion preparation method developed

above, the refrigeration stability of the SPI (Profam

981 and Ardex F) solution was also determined. Film

dispersions of both types of SPI were made and were

stored at 6°C for 48h. Refrigeration stability of the

solution is important for commercial applications as

prepared or left-over solution has to be stored under

refrigerated conditions to avoid spoilage.

Foods usedDiscs (6cm diameter and 0.5cm thick) and balls (3cm

diameter) from doughnut mix (Bisquick make), and

potato discs (3.5cm diameter and 2.5cm thick) and

fries (5cm long and 0.5cm diameter) were used in the

following studies.

Plasticizer effect on fat uptakeThe fat permeability of different ®lm formulations was

tested by coating doughnut mix disc-shaped samples

(6cm in diameter and 5mm thick). Three formula-

tions were prepared using plain SPI (Ardex F), SPI

with 3% glycerin, and SPI with 0.05% gellan gum. SPI

was weighed out to represent 10% solute (w/w) in a

500ml solution, mixed at medium speed avoiding ®lm

formation for 20min using both the magnetic stir bar

and a propellor type mixer. The solution was titrated

with 0.1N NaOH until the pH read 8.0. The

dispersion was placed on heat/mix plates and mixing

continued at low speed. The solution was heated to

80°C by a hot plate and the temperature was moni-

tored. It was held at this temperature for 20min. The

solution was divided into three portions. To one

portion 3% glycerin was added, and to the other

portion 0.05% gellan gum was added. All were mixed

throughly. The ®lm was applied to the food samples

using the dipping method and the ®lms were dried in

the convection oven on `dehydration' heat level

(<200°C). Seven replicates were made. Chemical

composition was measured before and after frying for

coated and uncoated samples. Moisture content was

determined using ASAE11 method, and fat content by

Soxhlet fat extraction method.12 The frying time was

15min at 160°C in a laboratory fryer.

Sudan red dye testsThe purpose was to make visible the amount of fat that

entered into the food product during frying. Sudan red

dye was placed in the oil prior to the start of frying.

Samples were prepared and fried in the fryer. Fried

samples were put into a cooler at 6°C for 30min and

cut in half axially so that the cross-section could be

observed. Visual observations were made on the depth

of oil penetration. This was performed on potato fries

(5cm long and 0.5cm diameter), doughnut mix discs

and balls (3cm diameter). These samples were also

coated with each of the ®lm formulations to see if there

J Sci Food Agric 80:777±782 (2000)

were signi®cant differences between coated and un-

coated foods. Experiments were replicated three times.

Penetration or puncture testMechanical properties of the ®lms are important to

consider as they affect the ability of the ®lm to remain

on the product during handling and the texture or

mouth feel of the product. Ateba and Mittal13 used

penetration test to measure the texture of fried

meatballs. Puncture tests with different sizes of probes

were also performed on the potato fries by Choi et al14

for objective measurement of texture. Pedreschi and

Aguilera15 used puncture test on potato strips at

different frying times to follow the kinetics of texture

development and structural changes. Similar tests

were used in this study to evaluate the effects of edible

®lm coatings on the fried product texture. A universal

testing machine (model 4204, Instron Corp., Burling-

ton, Ontario, Canada) was used to perform this test.

Samples were prepared using potato discs 2.5cm thick

and at least 3.5cm diameter. Three formulations of

®lm were tested: plain 10% SPI (Ardex F), 10% SPI

with 3% glycerin, and 10% SPI with 0.05% gellan

gum, as well as the non-dipped potato discs. All were

fried for 15min at 160°C and 10 replicates were made

of each. A plunger 3.1mm in diameter penetrated the

sample placed on the platform, at a rate of 1cmsÿ1

using a 50N load cell set at 1% scale (ie 0.5 N). The

plunger was pressed through the sample to a depth of

1.2cm, and repeated on 10 replicates for each

treatment.

Sensory evaluationsAn untrained sensory panel consisting of under-

graduate students evaluated the performance of the

various coated potato fries. Three triangle tests were

performed to determine if consumers could distin-

guish between coated and uncoated potato fries (Table

2). Each sample was randomly numbered and

presented to the panel member with the instructions

to pick the two different potato fries out of three.

Uniform product size was presented. preference and

any comments were invited.

RESULTS AND DISCUSSIONFilm formationBoth SPI materials (Ardex F and Profam 981) were

successful ®lm-forming solutions. Both contained

90% protein. No ®lm was formed using soy ¯our

since the protein concentration was only 46%. Soy

779

M Rayner et al

protein concentrate (Supro M57, Supro M9 and

Arcon S) initially showed promising results, but the

®lms were very weak and did not hold together when

immersed. Although the Supro 710 SPI was high in

protein (90%), ®lm formation could not be achieved as

it was intended for nondairy coffee whiteners, desserts,

dips and yogurts. The results con®rm the previous

®ndings of Bates and Wu8 and Gennadios et al.9,10

Heat treatment of SPIThe cold-dipped slide had a rough texture and many

visible particles that were granular in nature. The hot-

dip slide had almost no visible particles and had a

smoother texture. The re-cooled slide which had been

exposed to the longest heat treatment, had very few

particles and was also almost completely transparent.

There was no appreciable difference between the hot

and the re-cooled slides with respect to texture or

transparency. In terms of colour, the cold-dip slide

®lm had an amber tint, whereas the heat-treated slide

was relatively colourless. The Ardex F samples were

lighter in colour than the Profam 981 samples. The

best method of achieving an almost transparent ®lm

coating was to hold the dispersion at 80°C for 20min

and then cool the solution to room temperature.

Previously, Bates and Wu8 had used 85°C for 4.3%

and 5.3% SPI; and for 5% SPI solution, Gennadios etal9 used 70°C for 30min to get satisfactory ®lms.

However, they did not use the ®lms for coating of fried

foods to reduce fat content.

Refrigeration effectResults revealed that the Profam 981 dispersion did

not hold up well to the lower temperatures. After

refrigerating both dispersions for 48h, gelation oc-

curred in the Profam 981 while the Ardex F remained

in solution. Due to the instability of the Profam 981

dispersion, it was eliminated and the ingredient used

for further soy protein ®lm testing was Ardex F.

Fat uptake in doughnut mix discsResults (Table 3) show that the coated samples with

plasticized ®lm had a higher mean percentage fat

reduction than the samples coated with non-plasti-

cized ®lm. Duncan ranking and t-test (not shown) data

Table 3. Fat absorption and moisture change in fried doughnut mix discs

10% SPI 1

Uncoated (control)

Moisture content, db 0.3865�0.0174

Fat content, db 0.4262�0.0585

Coated

Moisture content, db 0.4009�0.0095

Fat content, db 0.2527�0.0567

Moisture gain (�) or loss (ÿ), %db �3.19a

Fat reduction, % db 40.71�17.82b

The means and standard deviations of seven replications are listed. Mea

780

show that there was a signi®cant difference between

the mass fraction of fat in the coated versus the

uncoated samples, and thus the application of the soy

protein ®lm has a signi®cant effect in reducing the food

fat uptake. The use of plasticizer (glycerin or gellan

gum) signi®cantly reduced the food fat uptake. Hence

3% glycerin or 0.05% gellan gum should be used when

preparing SPI ®lm solution for food coatings. The

change in moisture content was insigni®cant between

coated and uncoated fried foods. Earlier, Williams and

Mittal2 evaluated gellan gum and cellulose ®lms to

reduce fat absorption in fried doughnut mix discs. Fat

absorption was reduced by 50% to 90% in coated

products. Similarly, Chinnan et al6 reported that the

coated chicken strips with hydroxypropyl methylcellu-

lose (HPMC) had the lowest fat content (23%) and

the highest moisture (36%) in the surface layers.

According to Huse et al,5 the methylcellulose coating

of akara balls reduced the total fat content by 49%

during frying in peanut oil for 100s at 193°C.

Sudan red dye testsThe Sudan red test proved to be an effective method

for a quick comparison between coated and uncoated

samples with the exception of the french fries. The

french fries show no signi®cant difference between

coated and uncoated samples due to the small crust

thickness of the fried potato. The surfaces of the french

fries were red but the cross-sections of the samples

were too small to see any signi®cant differences. In the

discs of doughnut mix with no ®lm, the dye penetrated

through to the core (a layer 7±9mm thick), while the

coated discs showed the red dye being restricted to a

thin crust around the surface (3±5mm thick). There

were no signi®cant differences among three types of

coated samples (SPI ®lm with and without plastici-

zers). This test was also repeated using doughnut mix

balls. The uncoated doughnuts produced a red layer

approximately 3±5mm thick, while the doughnut

coated with 3% glycerin and 10% SPI or 0.05% gellan

gum and 10% SPI, showed almost no penetration of

the dye into the crust (<1mm thick). In fact, the dye

did not coat more than half surface of the coated

doughnuts.

with and without coatings

Coatings

0% SPI�3% glycerin 10% SPI�0.05% gellan gum

0.3944�0.0103 0.3900�0.0067

0.3930�0.0924 0.3939�0.0234

0.3936�0.0113 0.3935�0.0059

0.1792�0.026 0.1768�0.0321

ÿ0.20a �0.90a

54.40�7.51a 55.12�6.03a

ns with the same letters are not signi®cantly different at 95% level.

J Sci Food Agric 80:777±782 (2000)

Table 4. Results of penetration test on deep-fat fried potato discs

Peak force during penetration (N)

Uncoated, control 10% SPI with 3% glycerin 10% SPI 10% SPI with 0.05% gellan gum

Avg. (�sd) 0.043 (�0.0018)a 0.029 (�0.0013)b 0.027 (�0.0018)b 0.039 (�0.0016)a

Columns with same letter (a,b) are not signi®cantly different at 95% level. Means of 10 samples.

Soy-protein ®lms to reduce fat in fried foods

Penetration test on potato discsResults (Table 4) show that the fried samples coated

with plain 10% SPI and 10% SPI plasticized with 3%

glycerin were softer compared to uncoated fried

samples (control). It may be due to the formation of

softer crust when coating the samples with SPI

dispersions with or without glycerin plasticizer. The

food coated with 10% SPI formulation plasticized with

0.05% gellan gum did not have a mean force of

penetration signi®cantly different from that of the

control. This is favourable since it was desired that the

®lm coating would not have a signi®cant impact on the

textural aspects of the fried product. Earlier, whey

protein concentrate-based ®lm of 0.07mm thickness

did not impose any observable texture or appearance

change on the frankfurters.16

Table 6. Sensory panellists’ preference results on deep-fat fried potato fries

Test Preference No of responses

Sensory evaluationsTable 5 shows the results which were analysed using

the cumulative binomial probabilities.17 Each volun-

teer had a 33% chance of selecting a correct sample

based only on chance. For this the number of

minimum correct responses (X) required at a

con®dence level (z) is given by: X =0.4717

zp

n�[(2n�3)/6], where z =1.64 for 5% level and

n =number of panel members. Since the correct

responses of the sensory panel are less than X, the

panel could not tell the difference between the samples

at 5% level. The difference between actual correct

responses and X was more for A and B tests. Hence it

was dif®cult to separate coated and uncoated fried

samples. Results from the panellists' comments on

preferences are given in Table 6. In test A, consumers

found it more dif®cult to differentiate the taste

between the coated and uncoated fries, than in testC,

though some people did prefer the uncoated samples.

Tests A and C showed that there was a large group that

preferred the coated fries, or liked both the coated and

uncoated fries equally. Earlier, Mallikarjunan et al7

reported the effects of various coating methods on the

sensory attributes of fried chicken nuggets and

Table 5. Triangle sensory evaluation results on deep-fat fried potato fries

Test

No of

volunteers

No of correct

responses

Minimum no of correct

responses at 5% level a

A 10 5 7

B 9 3 6

C 12 7 8

From Roessler et al, Ref 17.

J Sci Food Agric 80:777±782 (2000)

marinated chicken strips. HPMC coating was applied

using solution or by incorporating into the breading

mix. Consumers preferred the products breaded with

the mix.

Overall, the results showed that the best alternative

is the ®lm dispersion plasticized with gellan gum. This

was because the dispersion excelled over the non-

plasticized and glycerin-plasticized dispersions parti-

cularly in terms of textural properties.

CONCLUSIONSResults showed that the higher the protein content of

the material, the better the ®lm formation capabilities

of the solution. Although protein content can be

increased by increasing the amount of material

dispersed, the increased non-protein content such as

fat, dietary ®bre and minerals appears to hinder ®lm

formation. Materials with high protein contents, and

thus low levels of other ingredients, had the optimum

®lm formation characteristics. The supplier's proces-

sing of the soy protein material also affected the ®lm

formation. Not all of the soy protein isolates formed

stable ®lms.

Heating the solution to 80°C, holding for 20min

and then allowing the solution to cool under well-

mixed conditions, was the optimum heat treatment.

Compared to other heat treatments (none and a hot

dip) this had the lowest amount of particulates visible

in the ®lm. A well-dispersed solution allowed for more

protein unfolding, and thus a stronger ®lm matrix.

This allowed a thicker coating to be applied to the

product as the solution did not readily run off. SPI

(Ardex F) provided a strong, stable ®lm and solution.

The 10% SPI ®lm without a plasticizer had a fat

reduction of 40.7 (�17.8)%, the ®lm with 3% glycerin

had a fat reduction of 54.4 (�7.5)% and the ®lm with

A Liked uncoated 4

Liked SPI coated 1

Liked both equally 5

B Liked SPI�gellan gum coated 2

Liked SPI coated 5

Liked both equally 1

C Liked SPI�gellan gum coated 8

Liked uncoated 3

Liked both equally 1

781

M Rayner et al

0.05% gellan gum had a fat reduction of 55.1

(�6.0)%. The Sudan red demonstrated signi®cant

fat content differences between the coated and un-

coated food products.

The penetration test showed that the uncoated

samples were not signi®cantly different from the

samples coated with SPI plasticized with gellan gum

but the samples coated with SPI plasticized with

glycerin and non-plasticized SPI were both signi®-

cantly different from the uncoated samples. This

signi®es that the samples coated with SPI plasticized

with gellan gum had the same hardness as the

uncoated samples.

The untrained sensory panel could not distinguish

between the samples with coating with or without

plasticizer. However, seven members in the panel of 12

successfully picked out the difference between the

coated samples with SPI�plasticizer and uncoated

samples. Most of the panel members preferred the

samples coated with SPI plasticized with gellan gum.

Thus the recommended alternative is the 10% SPI

with 0.05% gellan gum as plasticizer. The material

costs for the addition of the ®lm to french fries was

calculated to be $0.01kgÿ1.

REFERENCES1 Gamble MH, Rice P and Selman JD, Relationship between oil

uptake and moisture loss during frying of potato slices from c v

Record UK. tubers. Int J Food Sci Technol 22:233±241 (1987).

2 Williams R and Mittal GS, Water and fat transfer properties of

polysacchride ®lms on fried pastry mix. Food Sci Technol

(LWT) 32:440±445 (1999).

3 Kester JJ and Fennema OR, Edible ®lms and coatings: a review.

Food Technol 40(12):47±59 (1986).

4 Mallikarjunan P, Chinnan MS, Balasubramaniam VM and

Phillips RD, Edible coatings for deep-fat frying of starchy

products. Food Sci Technol (LWT) 30:709±714 (1997).

782

5 Huse HL, Mallikarjunan P, Chinnan MS, Hung YC and Phillips

RD, Edible coatings for reducing oil uptake in production of

akara (deep-fat frying of cowpea paste). J Food Proc Preserv

22:155±165 (1998).

6 Chinnan MS, Holownia K, Mallikarjunan P and Erickson MC,

Frying oil and product quality in response to application of

edible ®lms to marinated foods, IFT Annual Meeting Technical

Program Abstracts, p 35 (1999).

7 Mallikarjunan P, Chinnan MS, Bheem Reddy R, Holownia K

and McWatters KP, Consumer acceptance of edible ®lm

coated fried breaded poultry products. IFT Annual Meeting

Technical Program Abstracts, pp 131±132 (1999).

8 Bates RP and Wu LC, Protein quality of soy protein-lipid (yuba)

and derived fractions. J Food Sci 40:425±426 (1975).

9 Gennadios A, Ghorpade VM, Weller CL and Hanna MA, Heat

curing of soy protein ®lms. Trans ASAE 39:575±579 (1996).

10 Gennadios A, Brandenburg AH, Weller CL and Testin RF,

Effect of pH on properties of wheat gluten and soy protein

isolate ®lms. J Agric Food Chem 41:1435±1439 (1993).

11 ASAE, ASAE Standards ±Standards Engineering Practices Data.

American Society of Agricultural Engineers, St. Joseph, MI,

p 556 (1997).

12 AOAC, Of®cial Methods of Analysis, Assoc. Of®cial Analytical

Chemists, Washington, D C (1990).

13 Ateba P and Mittal GS, Dynamics of crust formation and kinetics

of quality changes during frying of meatballs. J Food Sci

59:1275±1278 (1994).

14 Choi H, Blumenthal MM and Schaich KM, Texture of french

fry crust±relationship to frying oil chemistry, IFT Annual

Meeting Technical Program Abstracts, p 88 (1999).

15 Pedreschi F and Aguilera JM, In situ textural characterization of

potato strips during frying. IFT Annual Meeting Technical

Program Abstracts, p 50 (1999).

16 Chen H, Zutara MS, Li J and Donnelly CW, Quality of beef

frankfurters affected by whey protein-based edible coatings

during refrigerated storage, IFT Annual Meeting Technical

Program, Abstracts, p 59 (1999).

17 Roessler EB, Pangborn RM, Sidel RM and Stone H, Expanded

statistical tables for estimating signi®cance in paired-prefer-

ence, paired-difference, duo-trio and triangle tests, J Food Sci

43:940±943 (1978).

J Sci Food Agric 80:777±782 (2000)