Utilization of Commercial Pea Protein Isolate and Commercial

Hydrogenated Canola 0ils in the Development of a Non-Dairy Frozen

Dessert

by

Àlbert Shiu Man Chan

À thesis
presented to the University of Manitoba
in partial fulfillment of the
requirements for the degree of
Master of Science
1n
Food Science

Winnipeg, Manitoba
(c) ¿lbert Shiu Man Chan, 1990
 National Library Bibliothèque nationaf e
ffi@ffi of Canada du Canada
Canadian Theses Service Service des thèses c¿nadiennes
Ottawa. Canada
Kl A ON4

The author has granted an ¡ffevocable non- Lauteur a accordé une licence irrévocable et
exclusive,llcence allowing the National Ubrary non exclusive p€rmettant à la Bibliothèque na-
of Canada to reproduce. t<¡an. distribute or sell tionale du Canada de reproduire. prêter. dis-
copies of his/her thesis by any means and in any tribuer ou vendre des copies de sa thèse de
form or format. making this thesis available to in- quelque rnanière et sous quelque forme que ce
terested persons. soit pour mettre des exemplaires de cette thèse
à la disposition des personnes intéressées.

The author retains ownership of the copyright L auteur conserve la propriété du droit d'auteur
in his/her thesis. Neither the thesis nor substan- qui protège sa thèse. Nila thèse ni des extraits
tial extracts from it may be printed or otherwise substant¡els de celleci ne doívent être im¡imés
reproduced without his/her permissircn. ou autrement reproduits sans son autorisation.

rsBN ø-315-633ø9-3

f-\ Eã'(
LA NACÅA
UTILIZATION OF COM}IERCIAL PEA PROTEIN ISOLATE

AND CO}ûIERCIAL HYDROGENATED CANOLA OILS IN THE

DEVELOPMENT OF A NON-DAIRY FROZEN DESSERT

BY

ALBERT SHIU MAN CHAN

A thesis subnrined to thc Faculty of Craduate Studies of

tl¡e University of Marritoba in part¡al fulfìllment of the requirentents
of the degree of

MASTER OF SCIENCE

o L990

Permission has been granted to the LIBRARY OF THE UNIVER-

S¡TY OF MANITOBA to lend or sell copies of this thesis. to
the NATIONAL LIBRARY OF CANADA to rnicrofilm this
thesis and to lend or sell copies oí the film, and UNIVERS¡TY
MICROFILMS to publish an abstract of this thesis.

The author res€ryes other publication rights, and neither tltc

thesis nor extensive extracts frorn it may be printeC or other'
wise reproduced without the author's written permission.
I hereby declare that I am the sole author of this thesis.
I authorize the University of Manitoba to lend this thesis to other
institutions or individuals for the purpose of scholarly research"

Albert Shiu Man Chan

I further authorize the University of Manitoba to reproduce this thesis

by photocopying or by other means, in total or in part, at-!he request
-other - institutions
oI or individuals for the purpose of scholarly
r esea rch.

Albert Shiu Man Chan

- 11 -
The University of Manitoba requires the signatures of aIl persons using
or photocopying this thesis. P1ease sign below, and give address and
date.

- 111 -
 ÀBSTRÀCT

Commercial pea protein isolate (ppl) prepared from yeIlow field peas

(pisum sativum t.) and blends of commercial hard (H) and soft (S) marga-
rine oils (Canola) in various weight ratios were used as the sources of

protein and fat, respectively, in the development of a non-dairy frozen

dessert. Four PPI levels ß"5%, 5.0%, 6.5% and 8.0%) and five oil
blends &HI%S: 30170, 40l60, 50/50, 60/40 and 70130) were selected for
developing the product formulation. All the frozen desserts were evalu-

ated by a sensory panel using a nine-point hedonic scale for their

overall acceptability and to identify the optimum oil blend/ppt leveI
conbination. It was found that a formulation containing by weight
'1
0.50% fat, , 11 .00% sugar, 1 .00% glucose solids , 0.35%
3.50% PPI 1

emulsifier/stabilizer and 63,65% water yielded a frozen dessert which

r,las more preferred to the others as judged by the sensory panel. The

optimum composition of the oil blend conlained 60% hard and 40% soft
'100C, 21.10C
margarine oi1. The solid fat index of this oil blend at
and 33.30C were 24,29, '13.20 and 2.97, respectively. The f rozen
desserts made with oil blends containing less than 50% hard margarine
oil scored significantly lower than the most acceptable one. The use of
ppI leveIs exceeding 5.0% resulted in frozen desserts that were signifi-
cantly less acceptable than the most preferred one. This indicated that
b.0% was the maxinum leve1 of PPI which could be used in the manufacture

of a satisfactory Product.

- IV -
 ACKNOWLEDGEMENTS

I Dr. Ronald Roy Pereira for his guidance and patience

wish to thank
throughout this study, and his understanding and encouragement. in my
times of trouble. I r+ould also Like to express my gratitude to the
following individuals who have made valuable contributions to this
study: Dr. Sue Àrntfield for performing the DSC analysis, George
Davidson of Canada Packers, Inc. for carrying out the test for Solid Fat
Index, Dr. Costas Biliaderis for the interpretation of the viscometry
data, Gregg Morden and Val Huzel of the Soil Science Department for
their assistance in the Kje1dah] analysis, Blaine Snell for photography
in the melting quality study, Linda Neden of the Statistical Advisory
Service for her advice on the analysis of the sensory evaluation data,
Paul Stephen for the abundant supply of computer units, and the staff
and students of the Departments of Food Science and Foods and Nutrition
for their participation in the sensory panels. For their encouragement,
I am grateful to Dr. Don Murray, Ðr. Greg Blank, Ðr. Costas Biliaderis
and Arnie Hydamaka.

On a personal noter my Sincere appreciation goes to my friends, both

within and outside the department, who helped me through some very
painful and trying times. To Francis, Philip, Hansen, Eva, Sue-Ellen,
tucia, Lucinda, Rick, Janine, l"laureen, Pamela, Bill r Vien, Georgina,

Mickey, Norman, Diana, Chris and Joe, i thank them all for makíng my

burden rnore endurable. Finally, my deepest gratitude resides with my

parents, to whom I everything. I thank them for their unfailing
owe

faith and support through the years, and above aII, for setting to me an
example of commitment and perseverance in the face of adversity.

- vI -
 DEDi CATI ON

To My Parents

and

To World Peace

- vl] -
 CONTENTS

ABSTRÀCT 1V

ACKNOWLEDGEMENTS

DEDI CÀTI ON vii

Chapter paqe

I. INTRODUCTION

II. LITERATURE REViEW

Classification of frozen de sse r ts

Background of the use of pI ant proteins and/or oils in
the manufacturing of dairy-like products
Use of Plant Proteins and F ats in the Development of lce
Cream-like Products I
Useof P1ant Proteins in Ice Cream-Like Products I
Useof Vegetable Fats/Oi1s in Ice Cream-tike Products 0
Mellorine and Parevine 1

Ingredients in Ice Cream and Non-Dairy Frozen Desserts 3

Protein. 4
Milk Proteins 4
Functions and Sources of Milk Proteins in Ice
Cream 17
'18
Pea Proteins.
Manufacture of Pea Protein Preparations 19
Differential Scanning Calorimetry (nSC) 21
Fat 22
Milkfâr 22
Functions and Sources of Mi lkfat inIce Cream 24
Sources and Composition of Cano1a 0i1 25
So1id Fat Index (SrI ) 28
Sources and Functions of Sweetening Àgents 29
Sources of Sucrose and Corn Syrup So1ids " 29
Functions of Sweeteners . 30
Functions and Sources of Stabilizers 32
Functions of Stabilizers 32
Sources of Stabilizers JJ
Rheology 34
Functions and So urces of Emulsifiers 36
Functions of Emulsifiers 36
Sources of Em ulsifiers 40
Processing Steps of Ice Cream Manufaclure 40
Blending of Ingr edients 40
Pasteurization 41

- vlrl -
 Homogenization 41
Aging 43
F reez i ng 44
Hardening 48
Melting Quality De fects of Ice Cream 48
Does-Not-Melt 49
Foaminess 49
Watery 49
Whey Leakag e. 50
Curdiness 50

TII METHODS AND MÀTERIALS 5l

Introduction 51
Testing of the Pea Protein Isolate . 55
Flavour 56
Mic robi olog i ca I Tests 57
Differential S canning Calorimetry (osc) 58
SoIid Fat Index 59
Sample Preparation 59
Melting of Margarine 60
Moisture Removal Þl
Preparation of 0il Blend Samples 61
Processing of non-dairy frozen dessert nixes 62
Pasteurization 64
Homogenization 65
CooJ.ing and Àging 6s
Freez i ng 65
Testing of the Mixes 66
Titratable Acidity and pH 66
Total Solids 67
Microbíological h--l^
LC5L5 67
Stability Test 68
Fat Content 68
Protein Content 69
Viscometry Studies 69
Tests Performed on the Frozen Desserts 70
Sensory Ànalysis of the Frozen Desserts 70
Melting Quality of the Frozen Desserts 71
Microstructure of the Frozen Desserts 72

iv. RESULTS ÀND DISCUSSION 73

Results of Tests Performed on the Pea Protein isolates. 73

Flavour Evaluation of PPI 73
Microbiological Tests of PPI '14
DSC Analysis of PPI 75
Solid Fat Index of the Margarine 0i1 Blends 78
Results of Tests Performed on the Frozen Dessert Mixes 80
Microbiological Tests 80
pH and Titratable Acidity 81
Protein ConLent of the Mi.x es 84
Fat Content of the Mixes 86
Total Solids Content of the Mixes
Viscosity of the Frozen Dessert Mi ;";....:::::.33
- IX -
 Results of Tests Performed on the Frozen Desserts . 94
Overrun of the Frozen Desserts . 94
Sensory Ànalysis of the Non-Dairy Frozen Desserts . "102
97
Melting Quality of the Frozen Desserts .
t'licrosÈructure of the Frozen Desserts 113

V. STJMMÀRY ÀND CONCLUSIONS 123

REFERENCES 126

Appendi x paqe

À. FLAVOUR EVALUÀTION DATA OF THE PPT SLURRIES. 13'7

B. SENSORY EVALUÀTION DÀTÀ OF THE NON-DAIRY FROZEN DESSERTS

MÀDE WITH FIVE MARGÀRINE OIL BLENDS. 138

c. RÀNKED SENSORY EVALUATTON DATA OF NON_DAIRY FROZEN DESSERTS

MADE WTTH FIVE MÀRGARTNE OIL BLENDS AND THE
CATCULATION OF THE COEFFICIENT OF CONCORDANCE. 140

Ð. SENSORY EVÀLUÀT]ON DÀTA OF NON_DÀIRY FROZEN DESSERTS

PREPARED AT FOUR LEVELS OF PPT. 142

E
RÀNKED SENSORY EVALUATION DÀTA OF NON-DAIRY FROZEN DESSERTS
PREPARED A1 FOUR PPI TEVELS ÀND THE CÀLCUIÀTION OF
THE COEFF]CiENT OF CONCORDÀNCE. 144

-x-
 LIST OF TABLES

Table pêgq

'1.
Typical Formulation of imitation Ice Cream (t¿ellorine). 11

a
Typical Formulations of Parevine. 12

'13
')
Formulations of Àll-vegetable Frozen Desser t
^ Àpproximate Composition of Dif ferent Grades of Ice Cream. . 14

Approximate Composition of Bovine Mi Ik.

.,t
IJ
tr

'19
6. General Composition of Peas.

7. Concentration and Composition of Milkfat in Each Melting

Range Region. 24

8. Solid Fat Index of Hydrogenated Canola 0i1, Hydrogenated

Soybean Oil and Palm OiI Used as Ingredients for
Formulating Hard and Soft Margarine Base 0ils. 26

g, Formulae for Hard (print) Margarine (p¡,l) Oits. 27

10. Formulae for Soft Margarine (Su) Oits. 27

11. Major Fatty Acids of Canola oil and High Erucic Àcid
Rapeseed (HEAR) 0i1. 28

12. Typical Chemical Composition and Microbiological Quality of

Four Grades of Pea Protein Isolate. 56

13. Basic Formulation of the Non-Dairy Frozen Dessert Mixes. . 63

14, Formulation of an lce Cream Mix, 10% B.F. 64

15. Results of Microbiological Tests of PPI. 75

tb" Results of DSC Analysis of PPi. 77

17, Alcohol Coagulation Test Results on the Non-Dairy Frozen

Dessert Mixes from the Two Slages of the Research. . 77

18. SoIid Fat index of Soft and Hard Margarine 0iIs, Margarine
Oil Blends and Butterfat. . , 78

- xl -
19. Results of Standard PIate Count (SPC) and Presumptive
Coliforms Test of the Non-Da iry Frozen Dessert Mixes made
r,¡ith Five Different Margarin e Oil Blends. 81

20 Results of Standard Plate Count (SPc) and Presumptive

Coliforms Test of the Non-Da iry Frozen Dessert Mixes
Prepared at Four Different P PI. LEVEIS. 81

21 pH and Titratable Acidity of th e Non-Dairy Frozen Dessert

Mixes made with Five Differe nt Margarine 0il Blends. . 83

22, pH and Titratable Acidity of the Non-dairy Frozen Dessert

at Four Levels of PPI.
Mixes Prepared 83

23. Protein Content of the Non-Dairy Frozen Dessert Mixes made

with rive Margarine 0i1 Blends. 85

24. Protein Content of the Non-Dairy Frozen Dessert Mixes

Prepared at Four Levels of PPI. 86

25. Fat Content of the Non-Dairy Frozen Dessert Mixes nade with
Five Margarine 0il Blends. 8'7

26, Fat Content of the Non-Dairy Frozen Dessert Mixes Prepared

at Four PPI Leve1s. 87

27. Total Solids Content of the Non-Dairy Frozen Dessert Mixes

made with Five Different Margarine 0i1 slends. 8B

28. Total solids content of the Non-Dairy Frozen Dessert Mixes

Prepared at Four PPI tevels. 89

31. Overrun of Non-Dairy Frozen Dessert Mixes made with Five

Different Margarine 0i1 Blends. 97

32. Overrun of Non-Dairy Dessert Mixes Prepared at Four

Different PPI tevels. 97

33. Sums Ranks and Tukey lest Results on the Mean Scores of
of
the Non-Dairy Frozen DesserLs made wiLh Five Different
Margarine 0i1 BIends. 99

34. Sums of Ranks and Tukey Test Results on the Mean Scores of
the Non-Da i ry Fr ozen Desserts Prepared at Four Di f ferent
PPI Levels. 101

- xII -
 LIST OF FiGURES

Fiqure page

1. Milkfat Globule Membrane in Homogenized lce Cream Mix. 38

2. À Schematic representation of the Structure of Ice Cream. . 45

3. Schematic Presentation of the Research Plan. 52

4" Melting Behaviour of the Frozen Desserts Prepared with nive

Margarine Oil Blends - Trial . '1
105

5. Melting Behaviour of the Frozen Desserts Prepared with Five

Margarine 0i1 Blends - Trial 2. 107

6. Melting Behaviour of lce Cream and the Frozen Desserts Made

wi tñ r'our Levels of PPI - Tr ial 1 . 1 09

7. Melting Behaviour of the Frozen Desserts Made with Four

Levels of PPi - Tria1 2. 111

8. The internal Structure of the Frozen Dessert Made with the

HS 64 0i1 B1end and of Ice Cream. 115

9. The Internal Structure of the Frozen Dessert Made with the

HS 46 Oit B1end and of Ice Cream. . 117

'10. The Internal Structure of the Frozen Dessert Containing 3.5%

PPi (P ¡S) and of Ice Cream. 119

11. The Internal Structure of the Frozen Dessert Containing 8.0%

PPI (P 80) and of lce Cream. 121

- x111 -
 ChaPter I
INTRODUCTION

Many dairy-1ike food products such as margarine and non-dairy coffee

creamers have been developed over the years by the food industry
utilizing materials of plant origin. The development of such products
is in response to several factors. First of which is the rising costs
and limitations on supply of dairy ingredients which prompted the food
industry to turn to other sources of supply (Simmons et â1., 1980).
Second, the economics, availability, functional and nutritional proper-
ties offered by plant materials made potential product developnent
possible (Morr, 19'19; Rajor et a1., 1983). Lastly, consumer demands for
lower-priced and convenient food products, advances in food science

technology and the increasing popularity of non-dairy products also

promoted this trend (National Dairy Counci1,1983; Lawhon et aI., 1980).

In addition to these reasons, various segments of the population

constitute a market for such products. One such segment is made up of
individuals who are intoleranl to lactose and/or milk proteins. Lactose
intolerance is a common adverse reaction to cows' milk (taylor and
Cumming, 1985). Malabsorption of lactose due to a deficiency in the
enzyme lactase in the human host may frequently lead to gastrointestinal
discomfort and other symptoms (taylor, 1986; Sandine and Daly, 1979).
Nearly 7O% of the world's adult population is lactase-deficient,
however, such individuals may or may not develop the symptoms of lactose

I
-t
 2

intolerance (Savaiano and Levitt, 1 987 ) . Ànother commonly reported

adverse reaction is cows' milk allergy (Emerson and Johnson, 1985).

Bovine milk contains 18 lo 25 proteins that are antigenic and can

provoke allergic reactions in humans (Taylor,'1986). The prevalence of

cor,ls' milk allergy has been estimated as 0.3-7.5% of the population
(Emerson and Johnson, 1985; Taylor, 1986).

Ànother group is comprised of individuals who are concerned with

saturated fat intake in their diets. Wor1d-wide interest in physical
fitness of the possible Iink
(Lusas and Rhee, 1986) and the recognition

between coronary heart disease and saturated fat intake in the diet
(Porter, 1980) have raised consumer awareness in dietary planning in
North Àmerica. Although dairy products are not the only foodstuffs
which contain saturated fat and cholesterol, some have been targeted as
undesirable. Such public attitude is reflected by the Àgriculture
Canada ( 986) food consumption statistics between 978 and 985. À
'1
1 1

steady decline in consumption is shown in some dairy products such as

homogenized whole milk (down tron 42.70 to 31.87 L/person/year) and

butter (dropped from 4.47 Lo 3.39 kg/person/year). An opposite trend,

however, is observed in low-fat products such as partly skimmed 2% mil.k
(up from 51.29 to 60.96 L/person/year) and skim milk (rose from 3.75 to
4"31 Llperson/year ).

Individuals who are unable to consume dairy products due to religious

or ethical reasons make up the third group. laws, for
Jewish dietary
instance, prohibit the consumption of dairy food and meat food at the
same meal (Fram, 1974; Sacharow, 1974). Vegetarianism, whether it is

based onreligion, ethics, health concern or perSonal reasons, may also

occasionally exclude the use of food of animal origin (Simoons , 1982).

 3

The rising of these individuals represent a steadily growing

numbers

market sector which the food industry has come to recognize. In order
to meet the needs and demands of these consumers, increasing effort has
been channelled into the development of new food products by the food
industry. The dairy sector, for instance, has put forth many new prod-
ucts or substitutes for traditional products. Some examples are

Iactose-reduced mi1k, calorie-reduced ice cream, Nutri-whip and non-

dairy frozen desserts. Of particular interest is the non-dairy frozen
desserts because plant proteins and oils are usually utilized in these
products to partially or completely replace milk proteins and butterfat.
One such products which gained wide commercial success in the U.S. in

the mid-1980's was Tofutti, an ice cream-like frozen dessert which

contained soy protein and soy oil (Shurtleff and Àoyagi, '1985).

Due to the interest shown towards these non-dairy frozen desserts, an

investigation was conducted to determine the feasibility of utilizing
other sources of plant protein and oil for the development of such a
product. Commercial pea protein isotate and commercial margarine oils
(hydrogenated Cano1a oils) were used in this study for several reasons.

These ingredients were selected for several reasons. Firstr green peas
1.las the most widely cultivated horticultural crop in Manitoba
(Àgriculture Canada, 986 ) .1 If this trend continues, and since the
manufacturer of the pea protein isolate is a Manitoba-based company, a
steady suppJ.y of this relatively new plant protein isolate should be
available if demands increase. Thus farr pêâ protein has been used as a
meat extender in sausage (Delaquis, 1983) and as a protein fortifier for
cereal products such as bread (Grant, 1983). Second, Cano1a is the most
 4

important oilseed crop in Canada since it accounted for approximately

62% of the total vegetable oil production in JuIy of 1989 (Statistics

Canada, 1989). At present Canola oil is widely used as shorLening oi1,

salad oil and in margarine production" Any attempts to develop more

means of utilizing these domestic products are apparently important and
logical in econonic terms. FinallY, at the academic 1eve1, this project
is undertaken since no research has thus far been conducted to investi-
gate the feasibility of developing a non-dairy frozen dessert based on
hydrogenated Canola oils and pea protein isolate.
 Chapter II
LITERATURE REVIEW

2.1 CLASSIFICÀTION OF FROZEN DESSERTS

the many types of frozen dessert available in North America,

Àmong

some can be grouped under the broad category of ice cream and related

products. These include ice cream, ttozen custard, ice milk, sherbet,
water ice, frozen confections, and mellorine- and parevine-type products
(erbuckle,'1986). The latter two resemble ice cream and may be regarded
as its imitations.

In the U.S., dairy-1ike foods which resemble or substitute for tradi-

tional dairy foods are referred to as imitation or substitute dairy
products by the FDA definition (National Dairy Council, 1983). The
imitation labet generally implies nutritional inferiority. Dairy-like
products are commonly manufactured by partially or completely replacing
the milk proteins and/or faÈ with their vegetative counterparts. They
are not considered as dairy products and can be divided into two catego-
ries: non-dairy and fi1led products. Filled products such as mellorine

and coffee creamers are basically manufactured by combining non-fat milk

solids with vegetable fat (Rahman, 1974; Weiss, 1983), whereas non-dairy

products such as parevine and non-dairy whip toppings do not conlain any
dairy ingredients.

-5-
)) BACKGROUND OF THE USE OF PTANT PROTEINS ÀND/OR OitS iN THE
MÀNUFACTURI NG OF DAIRY-LIKE PRODUCTS

The history of the use of vegetable proteins and oils in food produc-
tion has been a long one. Tofu, for instance, a soybean product devel-
oped by the Chinese nearly two millenia ago, has remained in their diet

until today. A margarine which contained predominantly coconut .and palm

kernel oils had appeared on Lhe market by the turn of the century
(eritchett, 1974). In nodern times, the interest in plant materials
originated from the effort in the 1960's to combat global hunger prob-
lems (Lusas and Rhee, 1986). Some dairy-like products were proposed or
developed as part of an effort to alleviate the extent of hunger and
n'ralnutrition in third world countries. Swaminathan and Papria (1967)
gave details of the preparation of oilseed- and nut-based milk substi-

tutes in the hope of overcoming milk shortage and improving the nutri-
tion of the very young in developing countries. Chandrasekhara et al.
(1971) developed Miltone, a toned milk product prepared by mixing peanut
protein isolate with buffalo or corl's milk in an attempt to extend the

milk supply for children in India. I n the early 1 970' s, a whey-soy

drink mix was also developed for preschool feeding programs in devel-
oping countries sponsored by U.S. aid (liildingr 1979). The mix was made
up of sweet cheese whey solids (41 .7%), f uII f at soy f lour ß6.9%) ,
soybean oil (12.3%), corn syrup solids (9.1%), and was supplemented lvith
vitamins and minerals. AIso in the 70's, Sosulski et al. (1978) experi-

mented in formulating imitation and blended milk products based on

legume protein isolates. Their study evaluated the protein isolates

prepared by alkaline extraction and acid precipitation from ten legume

species, including field pea, for their performance as the protein
 7

source in imitation milks. However, all the imitation milks were

significantly inferior to cor.v's milk in flavour and odour.

Currently, there are many dairy-ì.ike products on the market. Some

widely known examples are coffee creamers and non-dairy whip toppings.
These fabricated products contain vegetable fat and the protein usuaIJ-y

used is sodium caseinate or isolated soy protein (Kolar et'al. , 1979).

Soy-based infant milk replacements for

formulas have been used as
infants allergic to milk proteins (Thomson, 1979). Commercial produc-
tion of sterilized and packaged soya milk in Hong Kong has been a
successful venture since 945 (Jonas , 19751 Steinkraus, 978 ) .
1 1

lmitation and substitute cheeses containing vegetable fat have gained

substantial popularity in the early 1980's in the U.S. (l¡ational Dairy
CounciI, 1983). Commercial butter containing vegetable oil is currently
produced in Sweden, Finland, Àustralia, the U.K. , Switzerland, Japan,

Canada and the U.S. (Uadsen, 1985). Margarine, a butter-like product

made from vegetable oil, has also gained a strong foothold in the
market. ÀIso available are ice cream-like products such as parevine and

mellorine. is a non-dairy frozen dessert which is devoid of

Parevine
dairy and meat ingredients (Àrbuck1e, 1969). Mellorine is a filled
product which resembles ice cream or ice milk (weiss, 1983). Shurtleff
and Àoyagi (1984) published the production methods for many soymilk
products, some of which are manufactured commercially. Some examples
are soymilk ice cream, soymilk mayonnaise, soy dressings, soy chip dipt
soy popsicles, soy shakes, soymilk yoghur! and soymilk cheeses.
2,3 USE OF PLÀNT PROTEINS AND FATS iN THE DEVELOPMENT OF TCE CREÀM-
LI KE PRODUCTS

This section reviev¡s some of the recent research conducted on the use

of plant protein and/or fat in the manufacture of imitation ice cream.

¿"3. t Use of Plant Proteins in Ice Cream-Like Products

Soy protein has been utilized extensiveLy in the food industry. Some

atlempts have been made to explore the possibility of utilizing other

plant proteins. Lawhon et al. (1980) attempted to partially substitute
membrane-produced oilseed isolates for milk-solids-not-fat at leveIs

f.ron 20% to 80% in soft-serve frozen desserts. The study found that soy
, peanut and cottonseed storage protein isolates at substitution levels
of 60%, 60% and 20%, respectivel-y, did not cause any loss in the overall
acceptability of the frozen desserts.

protein ingredients were tried by Simmons

Cottonseed protein and soy
et a1. (1980) to partially replace the milk-solids-not-fat at levels
from 5% to 80% in soft-serve frozen desserts. CotLonseed and soy
flours, and soy protein concentrate were found to affect the overall
acceptability of the frozen desserts at 40% substitution level and
higher. protein isolate at 80% substitution affected the overall
Soy

acceptability of the final product. G1andless cottonseed storage

protein isolate showed no significant effects on the acceptability of
the final product and these researchers concluded that it might be used
to substitute up to 80% of. the milk-solids-not-fat in soft-serve frozen
desserts.
 9

El-Deeb and Sa1am (1984) used defatted glandless cottonseed flour,

defatted soybean flour and faba bean flour to partially substitute milk-
solids-not-fat in ice mixes. Their results showed that faba bean
cream

flour could be used at the 10% level to replace milk-solids-not-fat

without adversely affecting the quali.ty of the ice cream. Vanilla and
cholocate ice cream with acceptable qualities could be made with
defatted glandless cottonseed flour at 10% and '15%, respectively. The
chocolate flavour vras more effective than the vanilla in masking the
defects in flavour and therefore permitted a higher substitution 1evel.
However, soybean flour incorporation was only successful at 10% substi-
tution level. Higher levels resulted in flavour defects in the ice
c ream.

In India, a soft-serve ice cream vras developed using soybean and

buttermilk as part of an effort to provide lower-priced nutritious foods
to the populace (najor and Gupta,1982; Rajor et al., 1983). The soft
ice cream mix rvas prepared by incorporating hydrogenated vegetable oil
and sugar into a soybean-buttermilk slurry. The slurry was manufactured
by grinding previously soaked, bJ.anched and dehulled soybean in fresh
sweet-cream buttermilk. The optimal mix fornulation contained 9% fat'
'15%
sugar and a slurry with a soy solids to buttermilk solids ratio (on
a fat-free basis) of 1.3:1.

Ànother indian research group attempted to manufacture ice cream by

incorporating groundnut protein isolate (Cabriel et al., 1986). Their

study revealed that replacement of milk-solids-not-fat up Lo 40% did not

affect the sensory qualities of the final product which rvas reported to
contain more protein and cost less lo produce.
 10

2.3.2 Use of Veqetable Fats/0ils in ice Cream-Like Produc t,s

Partial substitution of milk fat by vegetable fat in ice cream manu-

facturing has been tried in several studies. youssef et a1. ( 1 981 )

substituted the fat in fresh cream with anhydrous butteroil, coltonseed

and corn oils in ice cream production. Anhydrous butteroil at a substi-
tution level of 50% did not affect the flavour, body and texture of the
ice cream. Substitution level of. 25% for cottonseed and corn oils was
recommended as the upper limit at which no adverse effect was noticeable

in the final product.

Partial replacement of milk fat with hydrogenated vegetable oil in

ice cream making was attempted by El-Deeb et aI. (1983). Their results
showedthat substitution of milk fat by hydrogenated oils decreased the
overrun, the score on body and texture, and the rate of melting of the
ice cream. Substitution levels higher than 10% in vanilla mix imparted
an off-flavour to the ice cream. However, chocolate was found to be

more effective than vanilla in improving the flavour of mixes containing

hydrogenated oils. The researchers concluded that up to 35% of milk fat

could be replaced by hydrogenated oils in chocolate ice cream with no

adverse effects on the product quality.

Laustsen (1985) and Lautsen (1986) reported that two specialty vege-

table fats, E 31 and Confao 5, were developed in Denmark.

Polawar
Polawar E 3'1 $¡as a fat based on fractíonated palm kernel oiI. The

nature of Confao was not disclosed. Laustsen (1985) claimed that these
specialty fats offered lower prices, uniform quality, longer shelf lifet
improved consistency aL room temperature, and neutral colour and taste.
 11

Suggested appl ications of these fats were in filled milk, f illed cream,

filled coffee cream, yoghurt and ice c ream.

2.3.3 Mellorine and Parevine

Mellorine and parevine are commercial ice cream-like products which

are legal-Iy availabLe only in of the states in the U.S. Mellorine
some

was first granted a lega1 status in Texas in 195'1 and its sales are only

permitted in several states (Borgstrom, 1976). The sales of mellorine

are prohibited in Canada. Mellorine is a fil-1ed product in which
milkfat has been replaced, in whole or in part, with vegetable fat
(Shurtleff and Àoyagi, 1985). Regular mellorine contains a minimum of
6-10% fat while the low-fat version contains a minimum ot 3-4%, Refined

and hydrogenated vegetable fats are used in imitation ice cream' a

mellorine-type product. The fats may be that of coconut, soybean' corn'
cottonseed, or their combinations (nrbuckle, 1986). À typical formula-
tion of imitation ice cream (mellorine) is shown in Tabl-e . '1

Parevine is a non-dairy frozen dessert which does not contain milk

and meat products, or any of their derivatives (¡rbuckle , 1959).
However: eggs and egg products are permitted. Parevine is not federally
regulated in the U.S., however, state standards do exist (Shurtleff and
Aoyagi, 1985). The non-dairy frozen desserts permitted to be sold in
Canada are in essence parevine-type products.

A variation of parevine is the all-vegetable frozen dessert. À11 the

ingredients in this type of products are of plant origin. Typical

formulations of parevine and all-vegetable frozen dessert are shot+n in
Table 2 and Tab1e 3, respectively.
 12

TABTE 1

Typical Formulation of Imitation Ice Cream (t'tellorine).

============================= ===========================================
Const i tuen t s Perc en ta ge

Water 60 .60
lk-sol ids-not-fat '12.00
Mi
Suc r ose 12.00
Vegetable fat 10.00
Corn syrup solids 5.00
Stabilizer 0.30
Emulsifier 0.10
Vitamins À and D opt i ona 1

Total solids 39.40

Source: Àrbuckle (1986)

TABLE 2

Typical Formulations of Parevine.

==== ========== === == = = = = = = = === === === = === = = = = === === ======== = == == == ==== === =
Economy Med i um DeIuxe
Mix #'1 Mix #2 Mix #3
Const i tuents (%\ (%) (%,

Water 68. 00 67.00 64.50

Vegetable fat (pure) 10.00 10.00 10.00
Suga r 15.00 16.00 16.00
Who1e egg solids or
dry whole eggs 4.501 9.00 2

Lor+ D.E. corn syrup solids or

hydrolized cereal solids 3.00 2.00
Vegetable protein 2.00
Microcrystalline ce1lulose 1.50
Stabilizer 0.35 0.35 0.3s
Salt 0.15 0.15 0.15

Total solids 32.00 33.00 35. 50

Source: Àrbuckle ( 1986).

1 Can be replaced by using 18 fresh whole eggs per 100 Ib of mlx.
2 Can be replaced by the use of 26 fresh whole eggs per 100 lb of mix.
 '13

TÀBLE 3

Formulations of All-vegetable Frozen DesserL.

2% f.aL nix 4% fat mix 10% fat mix

i ngredi ents (%) (%) (%\

Water 67.50 67.25 66.70

Corn sweetener (36 D"E. ) 14.00 13.00 10.00
'10.00
Fruc tose (,5/S ) 14.00 12.00
Vegetable fat 2.00 4.00 10"00
Soy protein 2.00 4.00 3.00
Stabi 1 izer/emulsi f i er 0. 50 0.45 0.30

Total solids 32. 50 32,75 33.30

Source: Arbuckle (1985).

2.4 INGREDIENTS TN iCE CREAM ÀND NON-DÀIRY FROZEN DESSERTS

Ice cream is a frozen foam prepared by simultaneous aeration and

freezing of a pasteurized mix. The basic components of the unflavoured

mix are milk-solids-not-fat (proteins), milkfat, svleeteners, emulsifiers
and stabilizers (fabte ¿). Canadian federal standards (r'ood and Drugs
Act, 1982) require that an ice cream mix must have a minimun total
solids of 36% and a fat content of 10%, or 8% if chocolate syrup is
added. This section will discuss the functions and sources of these
ingredients.
 14

TÀBLE 4

Approximate Composition of Different Grades of Ice Cream.

========================== ==============================================
Constituents &)
I ce cream Stabi lizersl Àpprox imate
Grade MiLkfat MSNF Suga r Emulsifiers total solids
10.0 10.0-11.0 '15.0 U.J
Econonly
12.0 9.0-10.0 13.0-16.0 0 "2-0 "4 35.0-37.0

Trade Brand 12.0 1'1 .0 15.0 0.3

14. 0 8. 0-9. 0 13.0-16.0 0.2-0.4 37.5-39.0

Del uxe 16.0 7.0-8.0 1 3.0-1 6.0 0.2-0.4 40. 0-4 1 .0

Pr em i um- 1 8.0-20.0 6.0-7.5 16.0-17.0 0.0-0.2 42 "0-45.0

Super premium 20.0 5.0-6.0 14.0-17.0 0,25 46. 0

Source: Àrbuckle ( 1 986 )

2.4.1 Protein
2.4.1 ,1 Milk Proteins

Bovine of water, fat, protein, lactose and minerals

milk iS composed
(fabte S). The constituents other than water and fat are collectively
referred to as milk-solids-not-fat (¡¿S¡tr). These materials are also
termed skim milk solids or serum solids. Milk-solids-not-fat in the
form of nonfat dry milk contains approximately 3.0% moisture, 0.9% f.aL,
37.0%protein and 51% lactose (¡rbuck1e, '1986). The most important
components of MSNF are the milk proteins, namely the casein and whey
proteins.

Of atl the proteins in nilk, casein accounts for approximately 80%

, 1978). This protein contains many components, but û5r-r ûs2-r

(Brunner

ß- and k-caseins have been identified as the major ones (nalg1eish,

1982).
 15

TABTE 5

Approximate Composition of Bovine Milk.

Componen t s Perc en ta ge

Water .43
B'1
Fat 3.70
Prote i n 3.50
Lac tose 4.90
Minerals 0.70

Sol ids-not-fat 9.10

Source: Henderson (1971).

in milk as colloidal particles termed micelles (rox and

Casein exists
Mulvihill, 19821. The micelles vary from 20 to 300 nm in diarneter
(Schmidt, 1980) and assume a highly porous and hydrated spherical struc-
ture (Schmidt and Morris, 1984). This structure is an aggregate of
numerous submicelles which range f rom '1 0-20 nm in diameter (t¡orr , 1 985 ) .

The submicelles are in turn nade up of molecules of the different

caseins ie. ds1-r &s2-t ß- and k-caseins. Both the micelles and submi-
celles are not static structures. There are continuous exchange reac-
tions between micelles and serum submicelles (Schmidt and Morris, 1984);

and between submicelles ahd serum casein molecules (Mcl¿ahon and Brown,
1984).

All casein monomers are low in nolecular weight in comparison to

other food proteins. In general the casein polypeptide chain possesses
a fair number of uniformLy distributed proline residues which limits the
a-helix or ß-sheet formation (l¡odler,1985) and this results in an open
random coil structure (l,torr, 1982), The polypeptide chains contain very
 16

fer+ sulphydryl groups and do not rely on disulphide linkages to provide

structural stability (nalgIeish, 1982). This also renders casein a high

degree of heat stability. Clusters of acidic (carboxyl and ester phos-
phate) and hydrophobic residues are distributed unevenly along the poly-
peptide chains, leading to the formation of distinctive hydrophilic and

hydrophobic regions , 1982; Kinsella, 984 ) . This molecular

(Morr 1

arrangement gives the molecule its anrphiphilic properties.

Whey or serum proteins ot the total proteins in milk

make up 20-25%
(Borst , 1978). The two principal fractions of whey proteins are
ß-lac toglobul in and a-lactalbumi n which const i tute 70-80% of thi s
protein group.

in milk, ß-lactoglobulin exists as a dimer with a nrolecular r¡¡eight of

approximately 35,000 (lim, 1980). Each monomer contains two disulphide
bonds and one free sulphydryl group (nilara and Sharkasi, 1986), and the
tr+o are joined by disulphide linkages (Schmidt and Morris,
rnonomers

1984). It is because of the presence of the sulphydryl groups that

ß-lactoglobulin is sensitive to heat denaturation and intermolecular
interaction (t¡orr, 1985). Beta-lactoglobulin has a compact and rather
spherical structure which owes it to a relatively low proline content
that permits a substantial helical content in the conformation (lim,
1980 ; Morr , 982 ) .
1 Ac idic and basic residues r ôs well as hydrophobic

and hydrophilic ones, are uniformly distributed along its polypeptide

chain (Evans, '1986). The result is the lack of amphiphilic properties
in the molecule.
 17

Àlpha-Iactalbumin is a very compact, nearly spherical single-chain

globulin with a molecular weight of approximately 16,000 (Schmidt and

Morris, 1 984 ) . AIpha-lactalbumin contains four intramolecular disul-

phide bonds and no sulphydrylgroups, it is therefore less susceptible
than ß-Iactoglobulin to heat denaturation (Schmidt and Morris, '1984;
Kinsella, 1984).

2.4.1.2 Functions and Sources of MiIk Proteins in ice Cream

The proteins in the MSNF serve several functions in ice cream. The

most obvious one is their high nutritional qualities (Swartz and Wong,
1985) which increase the nutritional value of ice cream. Casein, due to
its amphiphilic properties, acts as an emulsifying agent and helps to
stabilize the oil in water emulsion of ice cream (Julien, 1985). The
mi.1k proteins also act as stabilizers as they are able to hold water

through hydration (Hamilton, 1983). The results are an increase in the

mix viscosity and the retardation of the forming of large ice and
lactose crystals during the freezing process and during storage. The
increased viscosity improves the whipping quality of the mix, and once
foaming begins the entrapped air bubbles are stabilized by a layer of
fat which may have complexed with denatured milk proteins (Berger et
a1., 1972b), The presence of large ice and lactose crystals in ice
cream gives the product an undesirable texture which is manifested by an

icy, gritty and sandy mouthfeel. On the whole, the milk proteins in lhe
mix give body, structure and consistency to the finished ice cream.

The lactose and minerals in the MSNF also serve some functions in ice
cream. Lactose, though not a potent sweetening agent, nevertheless
contributes a small degree of sweetness to ice cream. But crysLalliza-
 18

tion of lactose during the freezing process due to excessive amounts of

lactose; and during storage because of fluctuating temperatures gives
the finished ice cream a sandy texture. The mineralsr oD the other
hand, help to the flavour of ice cream. Both the lactose and
enhance

the minerals also depress the freezing point of the mix (Iversen, 1983),
al-though not to the extend as shown by the sugars in the mix.

in the production of ice cream, milk is commonly used as a dispersion

medium for other ingredients. Milk also provides some of the MSNF

required in the ice cream mix" About 10-11% ot MSNF is needed in a

typical ice cream mix in order to produce a good ice cream (Rothwell,
1984). Since milk contains only 9% of. MSNF, the use of a concentrated
source(s) of MSNF is necessary in order to reach the desired levels.
These sources include buttermilk, nonfat dry milk/skim milk powder,

whole milk powder, condensed/evaporated mi1k, whey protein powder/

concentrale and caseinates.

2.4.1.3 Pea Proteins

L.) belong to the family Leguminosae

Peas (pisum sativum (Pomeranz,

1985). The general composition of peas is shown in Table 6.

Pea seed protein is composed of mainly globulins and a smaller frac-

tion of albumins. The albumins account for of the total protein
13-14%

(Grant et aI. , 1976) and are largely of cytoplasmic origin (pate, 1977).
Pea albumins consist of many subunits which range from 18,000 to 90,000

daltons in molecular weight (Mosse and PernoJ.l-et, 1983). The albumins

contain more sulphur amino acid residues than the globulins.
 19

TABLE 6

General Composition of Peas.

Const i tuents Perc en ta ge

Water 78.2
Protein 5.8
Fat 0.4
Ca rbohydrates 8.1
Fiber 5.6
0rganic ac ids 0.19
Àsh 0.7

Source: WiLIs et al. (1984)

The globulins are storage proteins predominantly located in the

cotyledons (pate, 1977). They account for nearly 80% of the total
protein in mature pea seeds (Boulter, 1983). Three types of these glob-
ulins have been recognized thus far: legumin, vicilin and convicilin
(Boulter, 1983).

2.4"1.4 Manufacture of Pea Protein Preparations

Pea protein preparations can be conventionally manufactured by either

dry or wet process. In the dry process as described by Sosulski and

Sosulski (1986), dehulled pea seeds are reduced by pin milling to a

flour of desired particle size which is then separated into a protein-
rich and a starch-rich fraction by air-classification. The starch frac-
tion is reclassified to separale the residual protein which is combined
vrith the first protein fraction to yield pea protein concentrate. The
wet processing method is a slightly more complicated procedure (Sumner
et al., 1981 Sosulski and SosuLski, 1986). In general, the proLeins are
 20

extracted from pea flour under alkaline conditions. Non-protein

materials are removed and extracted for residual proteins. The protein
extractions are combined and centrifuged to separate the starch. By
adjusting the purified protein extraction to pH 4.5 r+ith acid, the
proteins are precipitated and then separated from the whey. To prepare
a salt of the protein (proteinate) the curd is neutralized with a.n
alkali (usually NaOH) and then dried. if an isoelectric protein isolate
is desired, the curd is redissolved in aIkali, reprecipitated with acid,
washed and then dried.

The commercial PPIs'used in this study were prepared from the seeds
of yellow fiel-d pea (pisum sativum L. var. Century or var" Trapper) by
Woodstone Foods Ltd., Portage Ia Prairie, Manitoba, using an acid

extraction method patented by Nickel ( 98 1 ) . Due to the propr ietary

1

nature of the process, detailed information on the manufacturing process

is not avail-ab1e. BriefIy, dehulled peas are r+et-miIIed in water to
give a slurry from vrhich the proteins are extracted by lowering the pH
Lo 2.5-3.0 with acid. The solubilized proteins are then separated from

the non-protein mater.ials, precipitated at pH 4.5, neutralized and

spray-dr i ed .

Due to the severe extraction conditions, the structure of the

proteins which influences the functionality of the protein isolate can

be drastically altered. analytical method which evaluates the

One

extent of denaturation as a reflection of the functional integrity of a

protein extraction is Differential Scanning Calorimetry (OSC). This
method was also used in this research to determine the severity of dena-

turation of the pea protein isolate. It is therefore relevant tc

briefly review the principles of this method in the following section.
 21

2.4.1.5 Differential Scanning Calorimetry (DSC)

Differential (¡SC) is a thermoanalytical tech-

scanning calorimetry
nique which measures the thermal energy absorbed or released when a
substance/system undergoes a change in state, be it as the result of a
physical or chemical process (t¡right,1982). In the operation of DSC, a
samp).e and a thermally inert reference material are maintained in the

same thermal environment rvhile heat is applied to both at a programmed

rate (Sleeter, 1985). The thermal energy absorbed or evolved i.e. an

increase or decrease in the enthalpy by the sample causes its tempera-
ture to exceed (exotherm) or lag (endotherm) that of the reference.
Such a differential heat flow is recorded as AT (temperature difference)

or millicalories per second as a function of the temperature program

range or time and is normally shown as a peak on a thermogram.

The DSC thermogram directly yields some important information

concerning a given sample. Readily obtainable from the thermogram is

the enthalpy change (AH) associated with the transition (wright , 1982).

The area under the peak is directly proportional to the enthalpic change
and the direction of the peak indicates whether the transition is endot-
hermic or exothermic (Biliaderis, 1983). A downward peak indicates an
endothermic process while an upward one depicts an exothermic process
(Àrntfield and Murray, 1981).

The temperature at a transition takes place is also shown

which
directly on the thermogram. This temperature is commonly indicated by
the peak naximum temperature which has been denoted by the symbols Td,
T'm, Tm, Td and Tmax (wrigirt, 1982). in the case of proteins, this is
referred to as the denaturation temperature.
 22

Processing conditions as temperature, pH and ionic strength

such
influence the molecular conformation of protein molecules and thus
determine the degree of denaturat ion of the protein preparat ion.
Partial denaturation of a protein resulted from processing will be
reflected by the lowering of the AH while complete denaturation will not
yield an endothermic transition (niÌiaderis, 1983).

2.4.2 Fat
2.4"2.1 Milkfat

Cow's milk contains an average of 3.70% f.al (Henderson, 1971). The

fat exists in milk in the form of globules held in a state of emulsion

(oil in water). The size of the fat globules ranges from 0.1 to 10 um
in diameter and varies with the breed of cow, stage of lactation and
individual cow (¡therton and Newlander, 977) . The globules are
1

enclosed by a thin membrane known as the milk fat globule membrane

(unCu). This membrane is approximately 10 nm thick and is composed of
proteins, phospholipids, glycoproteins, triglycerides, cholesterol,
enzymes and other components (McPherson and Kitchen, 1983). The MFGM

functions as a barrier which protects the milkfat from the action of

milk lipase and as a natural emulsifier that maintains the fat globules
in the aqueous phase. Nevertheless, fat gLobules tend to cluster
because of the protein agglutinins on the MFGM surface and separate from
the milk serum due to gravity and difference in density i.e. creaming.

The creaming phenomenon can be retarded by means of homogenization.

The process reduces the fat globules to less than 1 um in diameter vrhich

largely eliminates the tendency of creaming. However, lhe reduction of

fat globule size results in a lremendous expansion in surface area and
 23

the original MFGM materials are no longer adequate to surround all the

newly formed fat particles. As a result, MFGM fragments and the milk
proteins adsorbed onto the fat particle surface and form the new (arti-
ficial) membrane which is less prone to clustering. In ice cream mixes,
the added emulsifiers are aÌso a component of this new membrane" Its
formation and composition are results of random interactions among the
materials surrounding each fat globule as it is formed (Berger and

white, 1976a), Upon cooling, the membrane moves into a more stable
state through structural rearrangements.

Milkfat is a diverse mixture of glycerides of fatty acids (Atherton

and Newlander, 1977). The physical properties of milkfat therefore
the proportions and properties of the component fatty acids.
depend on

The proportions of the fatty acids and triglycerides are in turn

strongly influenced by factors such as feeding conditions and lactation
stage (Frede et al., 1985). Several hundred of fatty acids have been
identified in milkfat but only ten make up the majority (90%) of the
fatty acid composition (Munro and IlIingworth, 1986) and have signifi-
cant influence on the behaviour of the fat (najair,.1986). These include
five short chain saturated fatty acids (4:0, 6:0, 8:0, 10:0 and 12:0),
three long chain saturated fatty acids (14:0, 16:0 and 18:0) and two
tong chain monoenoic fatty acids ('18:1, cis- and trans-) (uunro and
i 11 i ngworth, 1 986 ) .

Milkfat has a wide melting range instead of a sharp melting point due
to its heterogeneous composition (Mortensen, 1983). Àt or above 400C
milkfat exists as a fiquid, and as a solid at or below -400C. within
this temperature range, milkfat contains both solid and Liquid fats.
 24

The melting range of mitkfat can be divided into three regions: <00C'

00-200C, and >200C (Munro and illingworth, 1986). Table 7 shows the
concentration and composition of the milkfat in each region.

TABLE 7

Concentration and Composition of Milkfat in Each Melting Range Region.

=========================================================== =============
Melting Range (0C) Concentration (%') Composition
<0 35 - 40 low molecular weight unsaturat-
ed triglycerides
0-20 45 - 50 low molecular weight saturated
triglycerides or high molecular
weight monoene triglycerides
>20 10 - 15 high molecular weight saturated
triglycerides
Source: Munro and I llingworth ( 1 986 )

2,4"2.2 Functions and Sources of Milkfat in Ice Cream

Milkfat plays several roles in ice cream. The most important one is
to impart the unique rich, creamy flavour to the product (potLer, 1980).
The characteristic milkfat flavour is such a complex phenomenon that
thus far it has been impossible to duplicate. Milkfat also improves the
body and texture of ice cream as the fat crystals play an integral part
in the structure of ice cream. The lubricating effect of fat in the
mouth also contributes a smooth texture to ice cream (lerger and l^thite,

1979). À high-fat ice cream is remarkably smoother than one with less
fat (routs and Freeman, 1948). But fat is a foam depressant, high
levels of fat and excessive churning of fat during the freezing process
impair the whipping quality of ice cream mixes (Berger et a1., 1972a),
 25

In nutritional terms, milkfat is a source of calories, fatty acids and

four fat-soluble vitamins.

As indicated earlier, milk serves as a source of Ì'ISNF. However,

liquid whole milk also contains an average of 3.'10% nilkfat. This fat
must be taken into consideration when formulating an ice cream mix. In
Canada, the J.ega1 minimum fat content of plain ice cream mixes is 10%.
Other sources of milkfat must therefore be furnished in order to meet
this requirement. These sources include fresh cream, frozen cream,
plastic cream, butter and anhydrous milkfat.

2.4.2.3 Sources and Composition of Canola 0i1

The name "Canola" is a registered trademark of the Canola Council of

Canada (Ackman, 1983). The name is to designate newly developed

used

rapeseed (Brassica napus and Brassica campestris) cultivars that are low
in both erucic acid (<5% of total fatty acids) and glucosinolates (<3 mg
equivalents of 3-butenyl isothiocyanate per gram of oil-free dried meal)
(Boulter, 1983). These cuLtivars are also termed as "double-low"
because they contain low quantities of the two aforementioned factors.

Canola oil is the oil extracted from the whole seeds of such rapeseed
cultivars. Between 1971 and 1981, six rapeseed varieties of the Canola
type had been licensed. These included four varieties of Brassica napus
(Tower, Regent, Altex and Andor) and two of Brassica campestris (Cand1e

and Tobin), all of r+hich is characterized by a high oi1 content (>40%,

dry basis) (paun, 1983). Àlmost all of the rapeseed grown in Canada is
of the Canola type (Boulter, .1983). In JuIy of 1989, Cano1a accounted
for about 62% of the total vegetable oil production in Canada
(Statistics Canada, 1989) "
 26

The Canola oil used in margarine production requires it to be

specifically formulated in order to render the end product the desired

characteristics. Margarines made with pure CanoIa oil have been
reported to develop a grainy or chalky texture after a fev¡ months of

storage because of the fine crystal structure of oil (Anon., Cano1a

1981). Such defects can be rectified by blending Canola oil with other
oils of varying hardness to improve lhe smoolhness of margarine. The
base oils for hard (print or stick) and soft (tuU) margarines are there-
fore made by blending base stocks whose number and sources vary among

margarine manufacturers (Teasdale and Mag, '1983). Tables 8, 9 and 10

show the solid fat index of sone of these base stocks and the formulae
for hard and soft margarine oils.

TABTE 8

Sotid Fat Index of Hydrogenated Canola 0i1, Hydrogenated Soybean 0il and
Palm Oil Used as Ingredients for Formulating Hard and Soft Margarine
Base 0i1s.

Hydrogena Led Hydrogenated

Cano1a 0i1 Soybean 0i1 Palnr 0il
0i1 Stock c1 c2 c3 c4 SB4 SB5

Temperature ( oc ) Solid Fat Index (ml/kg)

10 4 12 38 50 50 60 22-28
¿t.5 2 52040 40
'ls 45 15-20
33.3 0 0 2 15 30 7-10

Source: Teasdale and Mag (1 983 ) .

The composition of high erucic acid rapeseed (Hn¡n) oils

Canola and
IS shown in Table 11. Canola oi1 is characteristic among vegetable oils

in two respects. First, the leve1s of the major saturated fatty acids,
 27

TÀBIE 9

Formulae for Hard (print) Margarine (pt',t) oits"

Types of Print Margarine (Pu)

I ngredien L (%) PM1 PM2 PM3 PM4 PMs PM6 PM7 PM8

Liquid Canola oil 20 50

c'1 60 60 51 51 55 45
C2 65
c3 20
c4 34 35
sB4 19_ 40 34 25
sB5 JJ 50
Palm 0i l- l5 15

Source: Teasdale and Mag (1983)

'10
TABLE

Formulae for Soft Margarine (Su) Oits.

Types of Soft Margarine (SM)

Ingredient (%) sM1 SM2 SM3 SM4

tiquid Canola oil

c1 T- 85 75 ::_
c4
sB5 20 1: 17
Palm oil ?2_ 15

Source: Teasdale and Mag ( 1 983 ) .

palmitic and stearic, which amount to 5-6% (Àckman, 1983), are the
lowest among the rnajor edibleoils (Downey, 1983). Second, the mono-
unsaturated fatty acid content is quite high (ca. 60%) with oleic (18:1)
being the predominant species (Ackman, '1983). This phenonmenon is the
result of the inabiliLy of low erucic acid rapeseed (fpen) plants to
 28

synthesize erucic acid (22:1 ) from oleic acid which is consequently

accumulated (Downey, 1983). However, hydrogenated Canola oils are used
in the formulation of margarine oils. Às hydrogenation lowers the unsa-
turated fatty acid content of an oiI, the margarine oils will as a
result contain different proportions of these fatty acids.

TÀBtE 1'1

Major Fatty Àcids of Canola oil and High Erucic Àcid Rapeseed (gn¡n)
0i1.

Canola oil HEÀR oil

Fatty acids (%) (%)

Palmitic '15
0) 2-5 3- 5
Stearic 18 0) 1- 3 1- 3
0leic 18 1) 53- 58 18-27
Linoleíc 18 2) 19-23 14-18
Linolenic 18 3) 8-12 8-9
Gondo i c 20 1) 1-2 12-1 4
Eruc ic 22 1) <1- 4 25-45

Source: Teasdale and Mag ( 1 983 ) .

2.4.2,4 Solid Fat Index (srl )

The SoIid rat Index (l¡ethod cd '10-57) of the American 0i1 Chemists'
Society (AOCS) has been widely accepted as an empirical tool to charac-
terize fats according to their solid-Iiquid contents (walker and Bosin,
1971). This method is a dilatometric technique which measures the
change in volumes as sLabilized fats expand or contract when being
melted or solidified at a given temperature (Rossel1, 1986). In the
A0CS method, solid fat indices are determined at 10.00Cr 21.10Cr 26.70C,

33.30C and 37.80C to characterize shortenings and margarine oils. SFI

 29

is expressed in units of milliliters solid fat per kilogram of fat

(mf/kg). SFi values range from 0 for fuIly liquid oils to 80-100 for
nearly solid fats.

SFi is indicator of the amount of solid fat present in a fat

an
sample (waddington,1986) or is used to indicate the extent of crystal-
lization of fats and oils (¡irker and Pad1ey, 1987). It does not
measure the true solids content of a fat which can be determined by

Differential Scanning Calorimetry or Nuclear Magnetic Resonance

Spectroscopy (walker and Bosin , 19'11). SFI is also not an indicator of

the melting point of a fat as the relationship between SFi and melting
point is dependent on the nature of the fat in question (weiss,'1983).

2.4 .3 Sources and Functions of Sweeteninq Aqents

The Canadian Food and Drug Regulations (1982) dictate that ice cream

shall be sweetened with sugar, liquid sugar, invert sugar, honey, dext-
rose, glucose, corn syrup solids, or any combination of these sweet-
eners. in practice, the most widely used ice cream sweetening agent is
the combination of sugar(sucrose) and corn syrup solids (;ulien, 1985).
The following discussion will therefore focus on these two sweeteners.

2.4.3.1 Sources of Sucrose and Corn Syrup Solids

Sucroseis a dissacharide composed of a molecule of glucose and fruc-

tose. The sources of sucrose are cane and beet. Some advantages of
using sucrose in ice cream are its availability, simple production
methods, relative low cost and long history of use (Inglett, 1981 ).
Sucrose can be used in either a syrup or granular form. Granulated
 30

suqar must contain a minimum of 99.8% sucrose as required by the

Canadian Food and Drug Regulations (1967Ì'.

Corn sweeteners are manufactured by converting corn starch molecules

Lo fragments of different chain J.engths through hydrolysis by acid,
acid-enzyme and enzyme processes (gobbs, 1986; Newsome, 1986). Corn

syrup (gl-ucose syrup) and dried corn syrup (dried glucose syrup) must

have a minimum of 20 dextrose equivalent (D.8.) as required by the

Canadian Food and Drug Regulations (1984). Dextrose equivalent is an

indicator of the extent to which starch has been converted to glucose

(Hobbs , 1 986 ) . I t measures the percentage of reduc ing sugars, calcu-

lated as dextrose, total dry substance (Hoynak, 19'74)" Corn
based on

syrups in liquid or dry form are commonLy classified according to their

D.E. val-ues (;ulien, 1985): low conversion (30-35 D.E. ), medium conver-
sion (-sO ¡.E.), and high conversion (¿gO o.E.). The properties of corn
syrups are determined by the degree of hydrolysis of corn starch i.e.
D.E. (cotf et at., 1983). As the D.E. value increases, the glucose
content increases while the content of water-binding dextrins decreases.
In other word, with increasing D.E., corn syrups become sweeter and more
effective in lowering the freezing point of the mix but less effective
in imparting viscosiLy to it. Low and medium D.E. corn syrups are

usually used in ice cream manufacture.

2,4.3 "2 Functions of Sweeteners

The obvious role of sweeteners in ice cream is to give sweetness and

to flavour. Sweeteners are also a major and economic source of

enhance

solids which, if adequate in quantity, give body and a fine texture to

 31

ice cream (Fouts and Freeman,'1948). A good average ice crean conlains
14-16% of sweeteners. Further, sweeteners lower the treezing point of

ice to prevent them from becoming a rigid frozen mass inside

cream mixes

the freezer (potter, 1980). Sweeteners also improve the texture of the
finished ice creams by controlling the amount of ice crystals formed

during f.reezing and storage (Harper and Shoemaker, 1983). But excessive
amounts of sugar in the mix will impair its ability to absorb air during

the freezing process.

Às mentioned earlier, a combination of sucrose and corn syrup solids

is widely accepted aS a Sweetening agent in ice cream production. The
purpose of this practice is to partially replace sucrose with corn
sweeteners. Between one-guarter to one-third of the sweetening agents
in ice cream can be corn s$reeteners (Àrbuckle, 1986). There are two
reasons for such practice. First of which is to increase the total
solids of the mix without increasing its sweetness and at the same time
lower the cost of the mix (tqanai and Bradley, 1969). Corn syrups are
ideal for this purpose because they are less sweet and are compatible
with sucrose and food flavours used in ice cream manufacture (Pomeranz,
1985). Second, the use of lower D.E. corn syrups helps to control
sucrose crystallization (Hoynak , 1974), increases the mix viscosity
which in turn gives a firmer and smoother ice cream (Coft et aI., 1983).
Ice cream containing corn syrups also show increased resistance to heat
shock during storage and more uniform melting (Pomeranz, 1985).
 32

2.4.4 Functions and Sources of Stabilizers

2.4"4.1 Functions of Stabilizers

Stabilizers used in ice cream production are generally hydrophilic

colloids or hydrocolloids, commonly known as gums (Sharma, 1981). Gums

are usually high molecular weight polysaccharides that dissolve or

disperse in v¡ater to give a thickening and/or gelling effect (Igoe,
1982). This phenomenon is due to the ability of gums to bind and immo-
bilize large amounts of water (Andreasen, 1985). Stabilizer gums are
used in very low quantities but are effective in performing several
important functions in ice cream. First of which is to stabilize the
oil/water emulsion in ice cream mix (GIicksman, 1984). This is made
possibJ.e by: (i) the ability of stabilizers to increase the viscosity of

the aqueous phase, thereby lowering the frequency of collision of the

emulsion droplets that results in agglomeration and ( ii ) the film-
forming properties of some gums which enable their molecules to surround
the fat gJ-obules to form a barrier against coalescence.

The second function of stabilizers in ice cream is to retard the

growth of lactose and ice crystals during storage. Tipson (1956) theor-
ized that a stabilizing agent can inhibit ice crystal growth by: (i)
adsorbing on ice crystal surfaces to prevent further expansion, (ii)
limiting the mobility of free water by its water-binding properties,
(iíi) complexing with crystallizing materials, and (iv) enhancing the
solubility of crystalline materials. The size and number of lactose and
ice crystals can increase during prolonged storage. Fluctuations in
storage temperatures cause ice cream to thaw and re-freeze, resulting in
the formation of large ice crystaLs. Substantial amounts of large ice
crystals in ice cream give the product an undesirable coarse texture.
 33

Furthermore, the presence of excessive lactose crystals in ice crean is

responsible for a sandy and gritty texture. This is best controlled by
limiting the amount of MSNF used in the ice cream mix as it is the major
source of lactose in ice cream, as well as, preventing t.emperature fluc-
tuat ion dur ing storage.

Lastly, stabiLizers to the mix. Àn increase in

impart viscosity
viscosity facilitates the incorporation and retention of air during the
freezing process and also improves the body and the melting qualities of
the f inished ice cream (Cottrell et a1. , '1980
).

Blends ofare widely used in ice cream mixes as they are found
gums

to be more effective due to synergism than individual gums to provide

stabitization (clicksman, 1 985) .

Combined and integrated emul-sifiers and stabilizers are also popular

in the ice cream industry for two reasons (Hielsen, '1978). First, a
combined emulsifier and stabilizer system is like1y to eliminate the

defects of individual components or improve the overall performance of

the system through synergism" Second, the use of integrated emulsifiers
and stabilizers simplifies the production process and improves its effi-
c iency.

2.4"4.2 Sources of Stabilizers

Stabilizer gums can be classified by chemical nature or by source

(Graham, 1978). On the basis of chemical nature, gums can be categor-

ized as anionic, cationic, or neutral. when classified by sourcer gums

can be grouped as plant exudates, seed gums, seaweed extract, pJ.ant
 34

extract, cel1ulose derivatives, microbial gums, proteins and synthetic

gums.

2.4 .4.3 Rheology

Since hydrocolloids alter the rheological properties of the system in

which they are dispersed or dissolved, it is appropriate to briefly
review the concepts of rheology. By definition, rheology is the study
of deformation and flow (¡icfinson and Stainsby, 1982\. Àlthough the
science of rheology encompasses both solids and fluids, the rheology of
the latter will be discussed because only liquid samples were tested for
their rheological properties in this research. Viscometry and oscilla-
tory testing are two methods of investigation which can be carried out
to study the rheological behaviour of fluids. On1y visconetry will be
discussed briefly because the frozen dessert mixes in this research were
not subjected to oscillatory testing.

Viscometry measures the viscosity, which is the internal friction or

the resistance to flow, of a fluid (Glicksman, 1982). Some concepts are
fundamental to the understanding of viscometry: shear stress, shear
strain, shear rate and shear viscosity (Morris, 1984). The force per
unit area applied laterally to a sample is known as shear stress r+hich
has units of Pasca1 (pa). The amount of deformation of the sample
induced by the applied force ís called shear strain which is dimension-
less i.e. without units. Shear rate is the rate of strain (strain per
unit time) and is expressed in reciprocal time (s-1) . The shear
viscosity of the sample is the ratio of shear stress to shear rate and
has units of Pascal seconds (pas). Plotting shear stress against shear
 35

rate yields a flow line on the plot (a rheogram). This line can be
linear or curvilinear. On this basis, fluids can be classif ied as
Newtonian or non-Newtonian (Mohsenin, 19i8).

In Newtonian (or ideal) ftuids the shear stress is directly propor-

tional to the shear rate. This relationship is demonstrated by a Iinear
flow line which passes through the origin. The viscosity of a Newtonian
fluid is constant and is sometimes referred to as absolute viscosity.
The shear stress of fluids, oD the other hand, is not
non-Newtonian

directly proportional to lhe shear rate (Glicksman, 982 ) . The

1

viscosity varies with the shear stress or shear rale and may be tinre-
dependent. The flow line is curvilinear and the viscosity at any point
on lhe curve (called the apparent viscosity) is the slope of a line
connecting that point and the origin (Mohsenin, 978 ) . Non-Newtonian
1

fluids fa11 into five categories: pseudoplastic, dilatant, Bingham

plastic, thixotropic and rheopectic materials. The apparent viscosity
of the latter two is time-dependent, but such a relationship does not
exist for the others.

Pseudoplastic and dilatant flor+s are two opposite phenomena

(oickinson and Stainsby, 1982). A pseudoplastic (shear-thinning) fluid
decreases in apparent viscosity with increasing shear rate. This

constrasts with a dilatant (shear-thickening) material whose apparent

viscosity increases with increasing shear rate.

A Bingham plastic is a fluid which rvill not flow unless a minimum

force (the yield stress) is applied to overcome the initial resistance
(Glicksman, 1982). The flow follovls a Newtonian pattern i.e. a linear
fLow line once it has commenced.
 36

Thixotropic and rheopectic flows are pseudoplastic and dilatant in

nature, respectively, except that their rheological behaviour is influ-
enced by time (Mohsenin, 1978). Thixotropic fluids thin out as the

shear rate increases, however, the original viscosity iS restored if a

period of rest is permitted. In a rheopectic flow, the fluid thickens
with increasing shear rate, but the original viscosity is also restored
after a period of rest.

2.4.5 Functions and Sources of Emulsifiers

2.4.5.1 Functions of Emulsifiers

An emulsion is a dispersion of at least two immiscible liquids, one

of which being dispersed in another in the form of droplets or globules
(Ctrristiansen, 1985). The dispersed liquid is usually referred to as
the dispersed or internal phase, while the other liquid is the closed,
conlinuous or external phase (Schneider, 1986). The emulsion system is
verl' unstable due to the surface tension between the two phases. If
altowed to stand the dispersed phase will rapidly coalesce and separate
from the continuous phase i.e. the breakdown of the emulsion.

In order to prevent phase separation, emulsifiers are introduced into

the system. Emulsifiers are interfacially active substances capable of
reducing the at the interfacial region and thus allow
surface tension
the droplets to remain dispersed in the continuous phase. This capa-
bility is due to the hydrophilic-lipophilic nature of an emulsifier
nolecule which permits it to position itseLf at the interface. Two
of action of emulsifiers have been suggested (risher and Parker,
nodes
1985): (i) once settled in lhe interfacial layer, the emulsifier
 37

molecules provide electrostatic or steric forces which repel emulsion

droplets from one another and (ii) some emulsifiers form liquid crystal-
line multilayers at the interface to prevent the coalescence of emulsion

droplets.

Emulsifiers in ice cream, ironically, are not used primarily to

provide emulsion stability. The oi1 in water emulsion in the ice cream

mix prior to freezing is largely stabilized by the milk proteins and

stabilizers, although the emulsifiers do provide additional stabilizing
effect and improve the extent of fat dispersion in the mix (F1ack,
1983b). The major function of emulsifiers is to promote and control the
destabilization or de-emulsification of the milkfat globule membrane
formed during homogenization when the mix is being frozen (Gregory,
1982). Figure 1 shows the orientation of emulsifier molecules in homog-
enized ice cream mix. Other important functions of emulsifiers in ice
cream are to promote fat-protein interactions, to facilitate air incor-
poration, to impart dryness at extrusion, to impart smoothness and
consistency, to increase resistance to shrinkage and to improve melting
properties (Hielsen, 1978).
Figure 1: MiLkfat GIobuIe Membrane in Homogenized Ice Cream Mix.

Source: Flack (1983a).

 39

Cosein l"licelle s

\.uo -unrIs
/\
_/\
,/\
o bo
a)
-o (J
@
\o
,':>-\
/J St
rt7)-(t ^
\1-ì1r/ (./
\o
.()
n Crystottized
O/ %@
o
o
L iqu id aì
/( Éol ()
Whey l6
protei NS

\
*Ð

o
oo \------.
@" -oo
 40

2"4"5"2 Sources of Emulsifiers

Emulsifiers are classified according to their electrical charge or

solubility properties (Schneider, 1986). Emulsifiers can be anionic,
cationic, amphoteric or non-ionic if categorized by electrical charge.
When solubility is used as the classification criteria the hydrophilic-

lipophilic balance (Hrg) index is often used.

Commercial emulsifiers for food uSe can be either natural or

synthetic, but the synthetic types are used predominantly. Emulsifiers
fa11 into several categories: lecithin, monoglycerides and derivatives,
propylene glycol esters, polyglyerol esters, sorbitan and polysorbate
'1983).
esters, and surcose esters (weiss,

2"5 PROCESSING STEPS OF ICE CREAM MÀNUFACTURE

Ice cream is produced by the simultaneous freezing and aeration of a

liquid mix base. The preparation of Lhis mix begins with blending of
ingredients, follows by pasteurization, homogenization, cooling and
aging. The mix is then flavoured, frozen, packaged and hardened. The

functions of each processing step are discussed briefly as follows.

2.5.1 Blendinq of Inqredients

The first step of ice cream manufacture is to prepare a mix base by

blending the ingredients in proportions as determined by the desired
formula. Blending is carried out in a tank or vat in which the liquid
ingredients are hetd. The dry ingredients are incorporated r+hiIe the
liquids are agitated and heated. To aid dissolution and dispersion, the
 41

dry materials can be either premixed or sifted to avoid cJ-umping prior

to blending (ÀrbuckIe, 1986). The temperature at which the solid ingre-
dients are added depends l-argely on their properties.

2.5.2 Pasteurization

Pasteurization of the mix is required by law to render the product

free from pathogenic organisms. Ice cream mixes are usually pasteurized
by the continuous high-tenperature-short-time (HrSt) or the holding
method (¡uIien, 1985). In the continuous HTST pasteurization, the mix
is heated to at least 800C for 25 sec in Canada. The holding or batch
method requires the mix to be heated to a minimum of 690C and held at

that temperature for not less than 30 min. Pasteurization offers

benefits other than rendering the mix safe for consumption. The keeping
qual.ity of the mix is improved due to a reduced microbial load. The
heat treatment also furthers the dissolution and hydration of the ingre-
dients, this in turn promotes their unifornr distribution and interac-
tions.

2,5.3 Homosenization

The principal objective of homogenizing the mix is to produce a

uniform and stabLe emulsion. Homogenization breaks up the fat globules

into much snaller ones (<2 um in diameter), thereby increasing the total
surface area onto which the stabilizing materials (mostly caseins) can
adsorb. This results in a fine dispersion of emulsified fat globules
which greatly enhances the stability and thus the life span of the emul-
sion.
 42

Homogenization is accomplished by forcing the mix under pressure

through a small adjustable valve called the homogenizing head (Gravlund,
1984). Fat globules are disintegrated by shear force, turbulence, cavi-
tation and ullrasonic vibration as they pass through the valve system of

the homogenizer (nees, 1974; Reuter, 1978). The temperature and pres-

sure at which homogenization is carried out largely determine the effec-

tiveness of the operation. Best results of homogenization are obtained
at temperatures which milkfat is in the liquid state (nees and Pando1fe,
1986). Homogenization pressures vary with the type of homogenizing
valve and mix composition. In general, the required homogenizing pres-
sure decreases with increasing fat content in the mix (l,fitten and

Neirinckx, .1986). Ice cream mixes should be homogenized in a two-stage

operation in order to give proper viscosity control. Àfter passing the

homogenizing valve under high pressures the disintegrated fat globules

tend to will impart excessive viscosity to the mix.
form clusters which
The clusters are dispersed if the mix is passed through a second valve

in series with the first one at lower pressures (Rees, 1974; Brennan,
1974). ice cream mixes are usually homogenized at pasteurization temp-
eratures and at pressures of 25C0-3000 psi on the first stage and 500
psi on the second (Arbuckle, 1986).

Homogenization also serves some other functions in addition to stabi-

lizing the emulsion. This process improves the distribution of mix
ingredients and promotes the interactions among emulsion components.
Ice cream made from a homogenized mix is in texture because
smoother

smaIl fat globules allow the incoproration of fine air cells and also
obstruct the formation of large ice crystals (White,1981). The fat in
a homogenized mix is less prone to churning during the freezing process,
 43

this in turn improves lhe whipping quality of the mix and results in a

smoother ice cream (Thomas, 1 98'1 ) .

2.5 "4 Àq i nq

After pasteurization and homogenization, mixes are cooled to s40C to

control microbial growth and to induce proper crystallization of the

liquid fat. It rlas mentioned earlier that milk fat is a diverse mixture
of glycerides of fatty acids (¡therton and Newlander, 1977). As temper-
ature is lowered, the high melting glycerides solidify at the fat
globule surface to form a shell surrounding a still-liquid core of low-
melting glycerides (Evans, '1986). The crystallization of fat is impor-
tant because the of fat churning in the freezer increases when
degree
the liquid fraction in the fat is high (Berger and White, 1971) "
Excessive churning of fat impairs the whipping quality of the mix and
causes defects such as fat and ice separation, buttery texture and
resistant to melt with leakage of serum.

The cooled mix is held at the cooling temperatures for 4 to 24 h

before the process. This holding period is known as aging
freezing
during which some changes take place in the mix. The most important of
which is the continuation of fat crystallization and the adsorption of
milk proteins onto the fat globules (Mitten and Neirinckx, '1986). The
milk proteins and the stabilizers also become fully hydrated, and the
emulsifiers orientated at the fat-serum interface during this period.
At this stage, the artificial fat globule membrane assumes its final
form (rigure 1 ). The overall result in the mix is a stable emulsion

which has increased viscosity and improved whipping properties.

 44

2 "5.5 Freezinq

of freezing is to produce a stable foam through partial

The purpose

destabilization of the ice cream mix (oickinson and Stainsby, 1982).

Freezing is accomplished with ice cream freezers which can be divided
into two basic types, the batch and the continuous freezer (Hatt et al.,
1986). Both types of freezer are usually consisted of a horizontal
tubular freezing chamber fitted vrith a motor-driven scraper/dasher to
aid freezing and aeration. The batch freezer can only process a defi-
nite amount of mix at a time and aeration is done at atmospheric pres-
sures. In the continuous freezing, ice cream is fed to the freezer at a
steady rate and is frozen and aerated under pressure. The major advan-
tages of the continuous freezer over the batch freezer are: (i) a lower
freezing temperature and a shorter freezing time, (ii) more accurate
control of the freezing process and thus a more uniform finished
product, and (iii) improved efficiency and economics of pJ.ant opera-

tions.

The freezing process may be regarded as a structurization process

(Monzini and Maltini, 1983) through which the liquid mix is transformed

into a complex system composed ofliquid, solid and gaseous phases

(Keeney, 1982). Structurally, ice cream is a partly frozen foam in
which air is dispersed in the form of tiny bubbles (air cells) in a
concentrated and unfrozen solution (continuous phase). In this solution
are dissolved sugars and salts; colloidaI milk proteins and stabilizers;
solidified fat, lactose and ice crystals (erbuckle, 1986). Figure 2
shows a schematic representation of ice cream structure.
Figure 2: À Schematic representation of the Structure of Ice Cream.

Source: FIack (1983a).

 .{o

lce crystols Stobitizers etc

20 - 60¡r

Wote( ----ts-

Air
10 - l00fr

Açgtomero led f ot
íorming
o skin oround crriroPPed
o ir cells
Fot gtobutes 0'2 -2'0¡r
¡\t -20oC (-4't=)
B5 -90"/. of the t'¿oter
r5 f-rozen
 4'l

Several important changes which take place in the mix during lhe

treezing process to give ice cream its unique structure are the freezing
of water, aeration and partial destabilization or de-emulsification of
fat. Às liquid water in the mix is frozen into ice crystals, the
dissolved materials become increasingly concentrated in the unfrozen
portion of the mix, lowering the freezing point continuously. The

liquid mix is therefore gradualJ.y turned into a viscous plastic mass

instead of a frozen solid throughout the f.reezing process (Thomas,
1981). This is important in the aeration of the mix because Hall et al.
(1986) indicate that the mix onJ.y incorporates air readily at a tempera-

ture of about -50C. This suggests that a partially frozen mix is more
effective in trapping air than the unfrozen mix. The incorporated air
causes an increase in the voÌume of the frozen mix. This phenomenon is
known as the overrun which is defined as the volume of ice cream
obtained in excess of the volume of the mix (Àrbuckle, 1986). Overrun
is expressed in percentage.

partial destabilization of the fat is essential in the retention of

À
air in the mix. During the freezíng process, some faL globules are
ruptured by the combined effect of low temperature and agitation (Berger
and WhiLe, 1979). The liquid fat in these disrupted globules leaks out

and acts like a cement which binds other fat globules together to form a
skin that encloses and stabilizes the entrapped air cells (Berger et
aÌ., 1972b). This effect also strengthens the structure between the air
cells (the lamellae).
 48

2.5.6 Hardeninq

Àfter ice cream has been extruded from the freezer and packaged, it
is at temperatures between -200 and -300C for hard-
immediately stored
ening (Julien, 1985). This practice is extremely important since it is
at this stage that ice cream aquires it definite structure and becomes
sufficiently strong to be handled. During hardening, more water will be
frozen into ice crystals until the concentration of the solutes in the
continuous phase elevates to such an extent that freezing can no longer
occur. Up Lo 95% of the water in ice cream is frozen at the hardening
temperatures (Monzini and Mattini, 1983). Àt this point the structure
of ice cream beocmes rigid. In order to avoid coarseness and iciness in
the ice cream, hardening should be carried out rapidly so as to prevent
the formation of large ice crystals.

2.5.'l Meltinq Oualitv Defects of Ice Cream

The melting behaviour of ice cream can be used as an objective

criteria in of formulation and processing condi-
determining the effects
tions on the quality of ice cream. Since the melting behaviour of the
frozen desserts was also studied in this research, the meLting quality
defects will be briefly reviewed. These defects have been covered
extensiveJ.y by Nelson and Trout (1965) and Àrbuckle (1986).

À high quality ice cream should melt readily at room temperature.

The melted ice cream should be a smooth and homogeneous liquid which

resembles the original mix. The most common defects in the melting
quality of ice cream are: does-not-me1t, foamy, watery, whey leakage and
c urdy .
 49

2"5"7 "1 Does-Not-Melt

The does not melt or slow melting defect is indicated by an ice cream

which retains or tends to retain its shape or is slow to melt when

allowed to stand at room temperature. FacLors which are responsible for
this defect are excessive viscosity in the mix resulting from high fat
content, improper homogenization, overstabilization and slow cooling;
lovr extrusion temperature from continuous freezers, and high freezing
point. This defect is often associated with a soggy, gummy, doughy or
sticky body.

2.5.7 "2 Foami ness

À foamy or frothy meltdown is shown by the presence of numerous fine

air in the completely melted ice cream. This defect is caused
bubbles
by excessive overrun, large air cells, large amount of egg solids,
excess emulsifiers or gelatin in low-solids mixes and high mix viscosi-
ties.

2.5.7.3 Watery

A watery meltdown is characterized rapid

by melt ing that yields a
thin, watery liquid. The common cause of this defect is a low solids
content in the mix. A coarse, weak body may accompany a watery melt-
down.
 50

2.5"7 "4 Whey Leakage

T.ihey leakage or wheying-off is shown by a watery liquid seeping out

of the ice cream during melting. Major causes of this defect are poor
quality ingredients, improperly balanced mixes and ineffective stabili-
zation. Whey leakage is closely associated with curdiness.

2.5.7 .5 Curd i ness

The curdy meltdown is characterized by a feathery scum on the surface

of the watery melted ice cream and the presence of fine particles in
this liquid. This defect is a result of protein destabil-ization in the
ice cream caused by excess acidity, imbalanced salt content, excessive
homogenization pressures, melting and refreezing in the freezer,
prolonged low temperature storage and shrinkage of ice cream.
 Chapter II i
METHODS ÀND MÀTERIALS

3.1 INTRODUCTION

The objective of this research was to develop a non-dairy frozen

dessert with acceptable sensory qualities using commercial pea proLein
isolate and margarine oil blend as the of protein and fat,
sources
respectively. Other ingredients used include water' sugar, glucose

solids, flavouring materials and stabilizer/emulsifier.

The research was divided into two phases in order to arrive at an oil
blend/protein l-eveL combination which would result in an acceptable
non-dairy frozen dessert, ôS determined by Sensory evaluation. The

content of sugar I Corrì Syrup solids , flavouring materials and

stabilizer/emulsifier remained constant among batches of non-dairy

frozen dessert throughout the entire research. Two variables, the fat
composition which rlas manipuLated by blending hard and soft margarine
oils in various weighl proportions, and the protein level were studied
for their effects on the quality and acceptability of the final prod-
ucts. A schematic layout of the research is shown in Figure 3.

- f,t
Figure 3: Schematic Presentation of the Research p1an.
 [-)
.JJ

?e¿ píoie n isoiares lJarC ¿nc' soii mareàrrn3s

and 5 oiL bIenCs
FLavour, nicrociolcAical :esrs ¿ni
Dif fereniial scannino calorime.'iv So.Lid fel index
I

WaIer, suqar, corn syrup scì.ids, slabilizer/enuisifier

Fi rst Siage

5 oil blends
- Mix pÍocessinq --- Tesling of nixes anci
1 oroie in level frozen ciesserLs

Sensory evaLuation: consuner paneI, 9-point hedonic scaLe

I
I

idenl i fy the oi L bLend that

produces ihe most acceptabì.e proouct

Seconci Siage

1 oi L blend
-- Mix Drocessing --- Testing of ¡¡ixes and
4 protein leveLs lrozen ciesseris

Sensory eval-ualion: consuner pane)., 9-poini hedonic scaLe

identi fy the protein Ievel ihat

produces the ncst accepiab)-e proCuct
 54

The first of the research rvas aimed at identifying a margarine

stage
oil blend which was able to create an acceptable frozen dessert. Five
oil blends were used to process the frozen desserts. The oil blend
which resulted in the most acceptable final product was identified by
subjecting the frozen desserts to a sensory panel. This oil blend alone
was to be used in the processing of frozen desserts in the next phase of

the research.

The goal of the of the investigation was to determine a

second stage

protein level which would result in an acceptable frozen dessert.

Batches of frozen dessert were processed at four protein leve1s. The
most acceptable protein level was identified by subjecting the frozen

desserts to a sensory panel.

Prior to stage one of the research, the pea protein isolates were

tested for their flavour and microbiological quality. The extent of

denaturation of the pea protein isolates was measured by Differential
Scanning Calorimetry (OSC). Both the hard and soft margarine oils and
alt the oil blends made up of the two were measured for their solid fat
indices. The non-dairy frozen dessert mixes were tested for their
microbiological quality, pH, tiLratable acidity, fat conteht, protein
content, total solids, stability and rheological properties. The frozen
products were evaluated for their overrun, Sensory quality, melting
quality and microstructure.
 LL

3"2 TESTTNG OF THE PEA PROTEIN ISOLÀTE

The pea protein isolates (ppl) used in this research were obtained
from Woodstone Foods, Portage la Prairie, Manitoba. Four grades of PPI
are generally nanufactured by the company: Iltoodstone GoId, Propulse
9858, Propulse 985R and Propulse 980. The four grades of PPI were manu-

factured for various applications. Primary use of each grade of PPI, as

recommended by the manufacturer, r¡ras in diet and health

Woodstone GoId

foods; Propulse 985g in baking; Propulse 985R in replacing other

proteins; and Propulse 980 in animal feed.

Samples taken from three grades of PPI (Cota, Propulse 985s and
Propulse 985R) were tested for their flavour, nicrobiological quality
and the extent of denaturation. Chemical analyses vrere not performed on

the PPI samples. However, information regarding the chemical composi-

tion of the PPI was obtained from the company (raUte lZ).
 56

'12
TABLE

Typical Chemical Composition and Microbiological Quality of Four Grades

of Pea Protein Isolate.

-;;;-il;;;-;;;i;;;-------
Gold Propulse 9858 Propulse 985R Propulse 980

Moisture, % 5.0 5.0 6.0 6.0

pH (10% solution) 6.5 6.5 6.5 6.5
Protein, %1 85.0 83.0 81.0 80.0
îaL, %2 3.0 2.0 3.5 N/A
Ash, %2 4.0 N/A 4.5 N/À
Crude f.ibre, %2 0. 5 0.4 0.5 N/A

Standard Plate Count

( )
organ i sms/g <50 , 000 <50 ,000 <50 ,000 <50 ,000
forms
CoI i -ve -ve _ve -ve
Salmonella -ve -ve _ve -ve

Source: Woodstone Foods Ltd., Portage la Prairie, Manitoba.

1 %ProLei.n = % Kjeldahl nitrogen x 6.25, dry weight basis.
2 dry weight basis.
N/A measurements not available.

3 .2.1 Fla vour

À flavour evaluation was carried out to determine the acceptability

and the difference in the flavour of the three grades of PPI (GoId,
Propulse 9858 and 985R). For this purpose a sensory panel consisted of
'16 judges using a 9-point hedonic scale was employed. A 10% slurry of

each grade of PPI was prepared by heating the appropriate quantity of

PPI in deionized water to a temperature of 79.40C (1750r) and maintained

at such for with constant stirring. The pasteurized slurries

30 min
were cooled and slored at 40C in sealed flasks for nol more than 24 h

before the taste panel began.

 57

The panel r,ras carried out in a panel room under red lighting to elim-
inate the effect of colour on the panelists'performance. The slurry of
each grade of PPI ltas assigned a 3-digit random number. Àpproximately

20 m1 of the slurry vlas poured into a paper cup J.abelled with the corre-
sponding 3-digit number. A set of samples in random order was presented
to each panelist. The panelists were instructed to chew a piece of
unsalted cracker and rinse their mouths with water between samples. The
analysis of variance was used to determined the presence of difference
among samples.

3.2.2 Microbiolooical Tests

The PPi samples were tested for Standard P1ate Count (SpC) and the
presence of Coliforms and Salmonella. Standard plate count and the
Presumptive test for coliforms were carried out as outlined in the
Standard Methods for the Examination of Dairy Products (¿pH¡, 1978).
Difco SPC agar and Violet Red Bile (Vnn) agar rlere prepared according to
manufacturer's instructions. E1even grams of PPI sanple were l+eighed
aspectically into a dilution bottle containing 99 nl sterile buffer
solution and mixed thoroughly. Dilution factors of 122 and 1:10 were
used for both tests. P1ates were poured in duplicates and incubated at
320C for 48 h for SPC and at 320C for 24 h f.or the presumptive coliform

test. Typical coliforms colonies on the VRB plates were transferred to

2% Brilliant Green Lactose Bile Broth and incubated at 320C for 48 h.

Lack of gas production indicated a negative test"

The presence of Salmonella in the PPI samples was detected by the

ÀOAC method (40ÀC, 1984) with modifications. Twenty-five grams of PPI
 58

samples were aseptically weighed into a blender jar containing 225 nI oÍ.

sterilized Nutrient Broth. of the jar were blended thor-

The contents

oughly using an Osterizer blender. The jar was then incubated at 350C
f.or 24 h. One ml of the jar contents was transferred to a tube of
Selenite Cystine Broth and a tube of Tetrathionate Broth which r+ere

incubated at 430C f.or 24 h. A loopful of each of the two broths was

streaked onto the surface of one plate of Brilliant Green Sulfadiazine
agar and one plate of Salmonella Shigella agar. The plates were incu-
bated at 350C f.or 24 h and examined for suspicious colonies. Suspected
colonies were inoculated onto Triple Sugar Iron agar slant and Lysine
Iron agar slant and incubated at 350C for 24 h. Colonies from positive
slants were tested with the ÀPI 208 system (¡pt Produits de Laboratoire
Ltee, St. Laurent, Quebec) to determine their identity.

3,2,3 Differential Scanninq Calorimetrv (DSC)

The degreeof denaturation of the PPI samples was determined by DSC.

Each PPI sample was first freeze-dried and then used to prepare a 20%
(wiw) slurry in water. À 10 to 15 mg sample of the slurry vlas weighed
into a tared DSC pan which was then hermetically sealed. À Dupont Mode1

9900 thermal analyzer with a Model 910 DSC ceII base and pressure DSC

celI in the test. The sealed DSC pan r¡as heated from 250 to
was used

1200C in the celI at a rate of 100C/min. The measurements were

processed by the DSC standard analysis software. Downward peaks indi-

cate an endothermic process. The temperature of denaturation (ra)
equals the peak temperature. The enthalpy of the reaction is calculated
from the peak area $¡ith the equation:
 59

AH = (¡/ucP) (60BEAqs)

where! AH enthalpy of the reaction (mcal'mg-

=
1
)

fi= area ( in 2 )
!t= sample mass (mg)
U- sampJe concentration (% wlw)
p= protein concentration of sample (%)
l= time base (min'in-1)
[= cell calibration coef f icient
aqs = Y axis range (mcaI's-f in-1)

3.3 SOLiD FAT INDEX

The physical characteristics of the fat samples involved in this

research were studied by measuring their solid fat index (srl). The
ÀoCS Official Method Cd 10-57 (¡OCS, 1974) was used for this purpose.

The work was performed by the chemical laboratory of Canada Packers

I nc. , Winnipeg. Instead of measuring the solid fat indices at five

different temperatures (100, 21,10 26.70 , 33.30 and 37.80C), âs
'
outlined in the nrethod, only three ( 00 , 21 .10 and 33.30C ) were used
1

routinely by the laboratory. This practice is common in the oils and

fats industry as Àllen (1982) pointed out that most margarine and short-
ening manufacturers have made modifications of the Iengthy official
method to suit quality controL purposes. The use of these three temper-

atures has been generally accepted (weiss, 1983).

3.3.1 Sanple Preparation

Hard (print or stick) margarine (20 kg blocks) and soft (tuU) marga-
rine ( 1 3 kg plastic tubs) were purchased from Canada Packers Inc. ,
Winnipeg for this study. The formulalion of the hard margarine oil- was
pus (rabte g) and that of the soft margarine oil r+as SM2 (tab1e 10).

The margarines were first nelted lo recover the oiI fractions. 0il
 60

samples were collected and treated for moisture removal before being

tested or to prepare various oil blends for testing. A milkfat

used
sample was also prepared from a commercial buLter in an identical manner

and subjected to the test.

3.3.1.1 Melting of Margarine

The soft margarine was melted by placing the tubs in a hot water tank

thermostatically maintained at 600C (1400r). The tubs remained inside

the tank for 5 h with periodic checking to ensure complete melting and

separation of the aqueous phase. The aqueous phase had to be removed

since it usually contains milk or whey solids, salt, and flavour
(Teasdale and Mag, 1983) which may provide unwanted interference in the

assessment of the quality of the finished frozen dessert. The oil was

then collected in sanitized plastic containers, sealed and refrigerated.

Àpproximately 700 mI of the well-mixed oil was collected separately for
further study.

UnIike the soft margarine, each block of hard margarine was packaged
in a paper carton, therefore its melting to be done in a different
had

manner. À block of hard margarine was heated in a stainless steel steam

vat which had been sanitized with a 200 ppm chlorine solution and thor-
oughly drained. The margarine was heated slowly with agitation to 600C

to ensure complete melting. oil was allowed to cool and separate

The

from the non-fat materials. The clear oi1 was then decanted carefully
into a sanitized pl-astic tub, sealed and refrigerated. Approximately
700 ml of the well-mixed oil was collected separately in a 1000-mI
Erlenmyer flask for testing and the preparation of oil blend samples.
 61

3"3.1.2 Moisture Removal

Each margarine oil sample collected after the melting step was trans-
ferred into a 1000-m1 glass beaker on a hot plate with stirrer capa-

bility. The oil was heated to a temperature of 600C and two teaspoon-

fuls of diatomaceous earth the Chemical Lab of Canada

(obtained from
Packers, Winnipeg) were blended in. The oil and diatomaceous earth
mixture was agitated for 15 min before the latter vlas removed by filtra-
tion using Whatman No. 42 tilter paper. The filtrate was then used to
prepare various oil blends.

3.3.1.3 Preparation of 0i1 Blend Samples

The two types of margarine oil and five oiI blends prepared from the

two and a butterfat sample were tested for SFI. The oil blends of 100 g
each were madeup of the hard (H) and soft (S) margarine oils in the
following proportions by weight i ,
30%-7 0% (HS 37) 40%-60% (HS 46)
'
50%-50% (HS 55), 60%-40% (Hs g+) and 70%-30% (HS 73). These oil blends

were chosen because preliminary studies showed that non-dairy f.rozen

dessert made from PPI and pure hard or soft margarine oil did not

possess acceptable quality. À11 oi1 samples were kept in 250-ml- glass

bottles sealed with plastic caps and stored under refrigeration until
testing.
 62

3"4 PROCESSING OF NON-DÀIRY FROZEN DESSERT MiXES

PIain (unflavoured) non-dairy frozen dessert mixes were prepared by

blending and pasteurizing the ingredients, followed by homogenization,
cooling, aging. The aged mixes were flavoured immediately before
and
the freezing process. All cleaned equipment and utensils involved in
the processing of the mixes were sanitized with a 200 ppm chlorine solu-
tion and drained thoroughly prior to use.

kg duplicate batches !¡ere processed using each oil blend in

Twenty
the first stage of the research. The mixes and the resultant frozen
desserts are designated by the code of the oiI blend they contained.
The basic formulation of the mixes is shown in Table 13.

In the stage, the mixes were prepared in 20 kg dupJ.icate

second

batches at four protein levels of 3.5% (P 35), 5.0% (P 50), 6"5% (P 65)
and 8.0% (P 80). These batches of mix and the resultant frozen desserts
will be hereafter referred to by the respective codes designated in the
parentheses. 0n1y the oil blend which re5ulted in the most acceptable
frozen dessert in the first stager âs judged by a sensory panel, rlas
used in this stage. The formulation used was modified from the one
shown in Table 13. with the levels of sugar, glucose solids, fat and

stabilizer/emulsifier maintained as in Table 13, the water content in

each batch was adjusted to correspond to the PPI level so that the total
percentage of ingredíents remained at '100.

Since the processing steps, conditions and equipment for the non-
dairy frozen dessert were identical to those for ice cream, one 20 kg
batch of ice cream mix was processed as to familiarize with the produc-
 63

TÀBLE 13

Basic Formulation of the Non-Dairy Frozen Dessert Mixes"

I ngredi ent Percentage Source

Water 6¿. t5

Sugar, fine granulated 11.00 Manitoba Sugar Co., I'iinnipeg,

Man i toba .

Glucose solidsr 1'1.00 Casco Inc., Toronto, Ontario"

(39-43 D.E. )

Fat 10.50 Canada Packers Inc., Winnipeg.

Pea protein isolate (powder) S.OO woodstone Foods, Portage Ia

(woodstone Gold) prairie, Manitoba.
Stabilizer/emulsifier
2 0.35 Germantown Manufacturing Ltd.,
(Beatamix 1 143) Scarborough, Ontario.
Tota l 37.85
'1 Contains 18.4% monosaccharides, 14"2% disaccharides, 12.5% trí-
saccharides and 54.9% higher saccharides.
2 Contains mono- and di-glycerides, cellulose gum, guar gumr poly-
sorbate 80 and carrageenan.

tion procedures and equipment involved. The formulation of the ice

cream mix used was obtained from the University of Manitoba Dairy Pilot

Plant (raUte l¿).

 64

TÀBtE 1 4

Formulation of an lce Cream Mix, 10% B.F.

Ingredient Percentage Source

Milk, 3.3%B.F. 55.93 University of Manitoba

Cream , 42% B.F. 21,80 Dairy Pilot Plant

Sugar, fine granulated 12.75 Manitoba Sugar Co., winnipeg,
Man i toba

Skim milk powder, spray-dried 4.68 Modern Dairies, St. Claude,

Man i toba .

Glucose solidsl 4.50 Casco Inc., Toronto, Ontario.

(39-43 D.E. )

Stabilizer/emulsifier 0.30 GermanLownManufacturing Ltd.,

(Beatamix 143) 2
1 Scarborough, Ontario.
Salt 0.04

'L'ot â 1 1 00.00

1 Contains 18.4% monosaccharides, 14.2% disaccharides, 12.5% tri-

saccharides and 54.9% higher saccharides.
2 Contains mono- and di-glycerides, cellu).ose gumr guar gum, poly-
sorbate 80 and carrageenan.

3,4.1 Pasteurization

AI1 dry ingredients were blended into the water in a sanitized stain-
less steel steam Mfg. Co., rtlinois) with a capacity of 40
vat (Groen

kg. Heat and agitation were slowIy applied until all materials were
dissolved. The oil was added when the vat contents reached a tempera-
ture of 48.90c ( 200r) . The mix was pasteurized at 79.40C ( 1 750F) for
1

30 min with constant gentle agitation.

 65

3.4 "2 Homoqenization

Homogenization of the pasteurized mix v¡as performed with a Cherry

Burrell Stellar Series 200 Superhomo homogenizer. The mix was homogen-

ized at of 2000 psi

pressures on the first stage and 500 psi on the
second. The homogenized mix was collected in a sanitized stainless

steel milk can

3.4.3 Coolinq and Aqinq

Cooling of the mix was done by immersing the milk can into a water

tank with circulating cold water. The mix was gently agitated to facil-
itate cooling to a temperature of (800f) or below. The can was
26.70C

then sealed and kept in a cold room at 4.40C (¿00n) for further cooling.
Its contents were stirred every 20 min to aid the cooling process. The
mix, when reached a temperature of 4.40C (400F), was allowed to age
overnight at the same temperature.

3,4.4 Freezinq

Prior to freezing, 15 kg of the aged mix was coloured and flavoured

with'1.4 kg of butterscotch sundae topping (Bowes Co. Ltd., Toronto) and

40 ml of butterscotch flavour (Givaudan Ltd., Toronto). Preliminary

work showed that butterscotch was an acceptable flavour for the type of
non-dairy frozen dessert being studied. Freezing rvas carried out in a
Cherry Burrell Fr 40-B Duo-Dash batch freezer. The frozen mix was pack-
aged into six 2 L paper cartons which were then weighed and stored in an

ice cream hardening room at -340C. One dozen of 150 nrl styrofoam cups
 66

were âIso filled with the frozen mix and kept frozen at the hardening
temperature for the study of the microstructure of the frozen desserts.
Overrun of the frozen mix was calculated using the formula of Arbuckle
(1e86):

weight of. 2 L of mix - weight of 2 L of. frozen mix

% Overrun = ------- X 100
weight ot 2 L of. frozen mix

3.5 TESTING OF THE MIXES

Mixes were tested for their pH, titratable acidity, total solids,
microbiology, stability, faL content, protein contenL and viscosity.
Samples were collected after the cooling stage and kept refrigerated for
not more than 24 h before being tested.

3.5.1 Titratable Aciditv and pFI

The titratable acidity and pH in the mixes were determined by the

modified procedures of the Standard Methods for the Examination of Dairy
Products (epH¡, 1978). Thirty six grams of mix were weighed directly
into a tared 100-mI glass beaker. The pH of the mix was measured by a
Corning pH meter.

Titratable acidity of the rnix was determined after the pH measure-

ments were taken. Thirty-six ml of boiled and cooled deionized water

was added and then blended in with a magnetic sti rrer bar. Twelve drops

of 1% alcoholic phenolphthalein solution were dispensed by an eye

dropper and mixed into the diluted mix. The beaker contents were
titrated to the phenoLphthalein end point of pH 8.3, as monitored by the
 67

pH meter, using certified 0.'1 N sodium hydroxide solution (goli)

di spensed by a Kimax Nafis acidometer. The titratable acidity r+as

obtained by dividing the acidometer reading bv 4.

3 "5 "2 Total Sol ids

Total solids of the mixes were determined by the À0ÀC method (À04C,

1984). Àpproximately 1-2 g of mix rvas rveighed directly into an aluminum

drying dish (5 cm diameter) which had been heated in an oven at 1000C
for '15 min and subsequently cooled in a dessicator. The sample inside
the dish r¡as spread over the entire bottom of the dish and then dried on
a steam bath for 30 min before being transferred into a forced air oven
for drying at 1000C for 3.5 h. Percentage total solids was calculated
by the following formula:

residue weight
% Total Solids = ------- X '100
sample weight

3"5.3 Microbioloqical Tests

The mixes were tested for standard plate count and the presence of
coliforms. Standard Plate Count method and the Presumptive test for
Coliforms using Violet Red Bile agar as described in the Standard
Methods for the Examination of Dairy Products (¡pH¡, 1978) were used.

Eleven grams of mj.x were transferred asceptically into a dilution blank

bottle containing of sterilized buffer solution. Difco Standard

99 mI

Plate Count agar and Violet Red Bile agar were prepared according to
manufacturer's instructions. Dilution factors of 1:10 and '1:100 r+ere
 68

used for the standard plate count. Duplicate plates were incubated at
320C for 48 h before reading. Dilution factors of 1:2 and 1:10 were

used in the presumptive test for coliforms. The plates vrere incubated
aL 320C for 24 h. Suspicious colonies were inoculateð, in 2% BrilIiant
Green Lactose Bile Broth and incubated at 320C for 48 h and examined for
gas production. A1l media, glassware and dilution blanks vtere routinely
sterilized in an autoclave al 1210C for 15 min at 15 psi pressure and
cool-ed bef ore use.

3.5.4 Stabilitv Test

The procedure used was described in the Methods of Analysis of MiIk

and Its (ulfl Industry Foundation, 1959). Five mI of mix was
Products
placed in a test tube, followed by the addition of 5 ml of alcohol solu-
tion. The alcohol solution was nade up by diluting 72 nI of. 95% ethanol
to a volume of 100 ml with deionized water. The tube contents were
mixed by gently inverting the tube several times. Instability of the
mix was indicated if flakes of curd appear.

3
"
5.5 Fat Content

The Pennsylvania test outlined in the Standard Methods for the

Examination of Dairy Products (¡pH¡, 1978) was used to determined the

fat content of the mixes. Nine-gram, 50% cream test bottles were used
instead of the ice cream test botlle, as outlined in the method.
 69

3"5"6 Protein Content

of the mixes r,läs measured by the Kjeldahl method

The nitrogen content
(¡O¡C, 1984) wittr modifications. The digestion catalysts listed in the
method vlere repì-aced by 2 Kjeltab tablets (risher Scientific ) .
Digestion was proceeded as in the method. The digested
described
materials were cooled, diluted r,lith 300 ml of distilled water and made
alka1i by the addition of 50 m1 of certified'10 N sodium hydroxide solu-
tion (fisher Scientific ) prior to distillation. Approximately 200 ml of
distillate was collected in a 500-m1 Erlenmyer flask containing 50 ml of
.1.0%
boric acid with methyl red and bromocresol green added as pH indi-
cators. The distillate was titrated $rith 0.1 N sulphuric acid. The

nitrogen content was calculated as follows:

(ml acid added - ml acid blank) x normality of acid x

o/rr
/'N -- ---- -----;;;;;-;;-;;;J;-i;;t-
14.007
____ X. r00

3.5.7 Viscometrv Studies

of viscosity of the mixes vras carried out by a Bohlin VOR

The study
rheometer system (sohtin Reologi, Lund, Sweden). The system vras
connected to a Bohlin temperature control unit (type 200123) , a
compressed air supply (2 bar), a Data Train DC-200S monitor screen, âD
IBM personal computer (uodel 5150), and an Epson FX-85 printer" Test

parameters vlere selected from the rheometer software. Viscosity meas-

urements of the mixes were taken using the C25 concentric cylinder
system at 20oC and the torque element setting at 95.8 g'cm.
Approximately 1 1 ml of sample was introduced into the concentric
 70

cylinder unit. The sample lvas subjected to every second one of the 48

shear rates within a range of 0.03682 to 1465 s-1, both inclusive. The

sensitivity was set at 1x. Initial delay time, constant delay time and
integration time were set at 30 sr 5 s and 5 s, respectively. One meas-
urement was taken at each shear rate.

3.5 TESTS PERFORMED ON THE FROZEN DESSERTS

3.6. 1 Sensory Analvsis of the Frozen Desserts

Frozen desserts from of the research were evaluated by a

each stage
sensory panel using a 9-point hedonic scale (Larmond, 1982), Panelists
were staff members and graduate students from the departments of Foods
and Nutrition and Food Sc ience of the University of Manitoba.
Fifty-three panelists participated in the first panel and forty eight in
the second.

Each panel carried out in a panel room under red lighting to

was

eliminate the effect of colour on the panelists' performance. Each

sample of the frozen desserts r+as assigned a 3-digit random number.

Approximately 50 g of frozen dessert was placed in a styrofoam cup

labelIed with the corresponding 3-digit number. À set of samples in
random order was presented to each panelist. Five non-dairy frozen

dessert samples were tested in the first panel and four in the second.
Panelists were instructed to chew a piece of unsalted cracker and rinse
their mouths with water belr¡een samples. The samples were allowed to be
swallowed.
 71

Several statistical methods were used to analyze the data from each
pane1. The presence of differences among samples and among judges was
detected by the analysis of variance. The significance of the varia-
lions among the samples was determined by the Tukey test (0'Mahony,
1986). In order to confirm whether there vlas a general agreement among
panelists upon the ranking of the samples in each panel and to establish
a reliable order of preference, the Kendall's coefficient of concordance
was used (cibbons, 1971).

3,6.2 Meltinq Oualitv of the Frozen Desserts

The arrangement and procedures for the recording of the melting

behaviour of the frozen desserts are described as foIlows.
Approximately 50 g of forzen dessert was placed in a plastic petri dish.
Frozen desserts fromthe same trial of each stage of the research were
lined up against a black background and were allowed to melt at room
temperature. Photographs of the dish contents were taken at twenty-
minute intervals in an one-hour period, beginning at zero minute. A
Nikon camera with a 24 mm wide-angl-e lens and Kodak Ektachrome ED 135-20

slide film for the photography. Lighting was provided by four

were used

Sylvania 82 Superflood EB W bulbs, two on each side of the background.

The lamps which generated heat were activated only when photographs were

being taken in order to avoid possible accelerated melting of the dish

contents.
 72

3.6.3 Microstructure of the Frozen Desserts

Frozen dessert samples were sl-icedinto sections of 8-10 um thick at

-300C using an Àmerican Optic Model 851c Cryo-Cut II Cryostat Microtome.

Tissue-Tek embedding medium (t"liles Scientific ) was used for mounting.

Each sample section r+as transferred onto a glass slide chilled inside

the microtome and a chilled glass slide cover slip was placed on top of
the specimen. The slide was then placed on a cooling stage located on

top of the viev¡ing stage of a Zeiss Universal transmitted-light reserach

microscope. The cooling stage r,Ias a sealed hollow brass square (approx-
imately 6.5 cm by 5.5 cm and 1 cm thick) with an opening in the center
which al-lows the passaqe of liqht through the specimen. Two nozzles

located on one edge of the square were used to connect Lhe stage to a

Haake F3-C digital refrigerated bath and circulator via two rubber

tubings. The coolant, a 50/50 mixture of automotive engine antifreeze

and distilled water, was cooled to -'15 to -200C and circulated inside
the cooling stage to prevent rapid melting of the specimen. The struc-
ture of the speciment was photographed by a 35 mm camera attached to the

microscope using Kodak Ektachrome ED'135-20 slide film.

 Chapter IV
RESULTS AND DISCUSSTON

4, 1 RESULTS OF TESTS PERFORMED ON THE PEA PROTEIN ISOLÀTES.

4. t.t Flavour Evaluation of PPI

On a scale of one to nine (dislike extremely to like extrenely), the

average scores for the three grades of PPI tested: Gold, 9858 and 985R

were 3.69, 3.69 and 3.38, respectively (nppendix a). These scores indi-
cated the general acceptability of the PPI slurries by the panelists
ranged from dislike slight1y to dislike moderately. However, the
flavour evafuation did not point out which aspect(s) of the PPI slurries
was objectionable to the judges. The for the PPI slur-
average scores
ries were 1ow perhaps because rather concentrated (10%) PPI slurries
were used for flavour evaluation. This concentration was used because
the manufacturer's guidelines suggested that a 10% PPI solution rlas
bland in flavour. The undesirable qualities of the PPI, on the
contrary, became acuLe at this concentration and were readily detected
and penalized by the panelists.

However, the PPI levels used in preparing the non-dairy frozen

desserts were lower (3.5%,5.0%, 6.5% and 8,0%). Therefore if the 10%
PPi slurry had been found acceptable qualities, the use
in its sensory

of these concentrations would not then be expected to pose any problent

j.n the flavour of the frozen desserts. In addition, the frozen desserts

were completely different food systems in which the presence of other

- 73 -
 74

ingredients i.e. sweeteners, fat,emulsifiers, stabilizers and

flavouring materials may conceal or modify the shortcomings of the PPI.
Consequently, these average scores may not be regarded as a definite
guide to the performance of the PPi with respect to flavour in food
applications"

Ànalysis of variance did not reveal any significant difference in the

general acceptability between the three grades of PPI at the 5% signifi-
cant leveI (Àppendix A). Às a result, Woodstone GoId PPI was selected

for this research because it was indicated by the manufacturer as the

premium grade.

4.1 .2 Uic¡pþjplggicai Tests of PPI

tests vrere carried out to assess the general microbiological

These
quality of the PPI and to ascertain that these products were free of
pathogenic organisms. The presence of Salmonella was not detected in
any sample (table 15). A negative presumptive test for coliforms was
also noted for all the protein isolates. There were, however, varia-
tions in the Standard Plate Count results. of the GoId grade,
One batch

the 9858 grade and the 985R grade PPI showed counts in conformance to
the <50,000 organisms/g value as suggested by the manufacturer's typical
analysis (Table '12). Two batches of the Gold grade PPIr oD the other
hand, gave counts exceeding 50,000 organisms/g. These counts are not

significant since they are in the same log cycle as the manufacturer's
standard (fable 12). It is like1y that the high counts were contributed
by the manufacturing process during the drying and packaging operations.
 75

TABLE 'I5

Results of Microbiological Tests of PPI

SampIe

GoId

c20 c38 c 87 985 B 985 R

sPc (cnu/s) 7,3001 53,000 60,000 34 ,000 7 ,200

Coliforms Count
( organ i sms/g ) <1 <1 <1 <1 <.1

Salmonella 2
-ve -ve -ve _ve -ve

1 Àverage of results from duplicate plates.

2 Performed in duplicates.

4.1.3 DSC Ànalysis of PPI

The denaturation temperatures (fd) of the PPI are

and enthalpies (Att)
shown in Table 16. Bhatty (1982) cited that DSC analysis of pea
proteins isolated by the salt solubilization method of Murray et al.
(1978) shor+ed a Td of approximately 900C and AH of 4.0 - 4.5 cal/g. All

the PPI samples tested had slightly lower Td but much lower AH readings,
implying that they were severely denatured. It may be speculated that
the cause of denaturation is the harsh treatment that the proteins
received during the isolation process (Section 2.4.1.4).

significantly affect their functional prop-

Denaturation of proteins
erties. 0f particular interest to this research are the solubility and
emulsifying properties of the PPi. Voutsinas et aI. (1983) found that
the solubiJ.ity of PPI decreased with increasing heating time at 800C as
the proteins becanie progressively denatured. The emulsifying activity
of the protein isolates also showed a similar trend. This is in agree-
 76

ment vrith the suggestion of Kinselta (1982) that the solubility of

proteins was crucial to their emulsifying ability. it is because a
protein which stabilizes an emulsion does so by forming a film at the
interfacial region. High solubilities therefore permit rapid migration
of large amounts of protein molecules to the interface to effect stabi-
lization.

Despite the extent of denaturation shownin the PPI tested, the

frozen dessert mixes containing these proteins showed satisfactory
stability. The alcohol coagulation test failed to induce destabiliza-
tion of any of the mixes (tab1e 17). This is because the pea proteins
were not the only source of stabilization in the mixes. Other ingredi-
ents such as the stabilizers, emulsifiers and glucose solids also
contributed to emulsion stability which was further enhanced by the
homogenization process. This may also be illustrated by the disappear-
ance after homogenization of a very thin film of oil which tended to
collect on the surface of the mixes during pasteurization as agitation
was slowed down or ceased. 0i1 separation in the homogenized mixes was

not observed even after the aging process"

 77

TABLE 1 6

Results of DSC Ànalysis of PPI.

====================== =================== ===============================

SampIe Td1 ( oc) AH2(ca1/g protein)

Gold: c 204 83.753 0. 32 3

G 38s 81.73 0.51
G 876 63.51 0.39
985 B 86.03 0. 59
985 R 85 .47 0 "32

1 Denaturation temperature.
2 Denaturation enthalpy.
3 Average of duplicate measurements.
4 Used in Trial- 1 of stage one.
5 Used in Trial 2 of stage one.
6 Used in both trials of stage two.

TÀBLE 17

Alcohol Coagulation Test Results on the Non-Dairy Frozen Dessert Mixes

from the Two Stages of the Research.

Results

Sample Trial 1 Trial 2

HS 37 -ve 1
-ve
HS 46 -ve -ve
HS 55 -ve -ve
HS 64 -ve -ve
HS 73 -ve -ve
P35 -ve -ve
P50 -ve -ve
P65 -ve -ve
P80 -ve -ve
1 À negative was indicated by the absence of flakes/curds
result in the
mix/alcohol mixture; result of duplicate observations.
 78

4,2 SOTID FAT INDEX OF THE MARGÀR]NE OIL BLENDS

The So1id Fat Index (Sfl) values at 100C and 21.'10C were seen Lo
increase with oil in the oil
increasing proportions of hard margarine
blends (ta¡te lg). However, at 33.30C the SFI values fluctuated. They
should have followed a similar trend as observed at the other two temp-
eratures. Further, the SFI values of the soft margarine oil at 10aC and
210C, and ihat of the hard margarine oi1 at'100C exceeded the usual SFI

range aimed for by the production line. Such anomalies may be contrib-
uted by: (i) errors during testing or (ii) the altered
made
crystallization/melting behaviour caused by variations in the oil manu-
facturing process.

TABLE 1 8

Solid Fat Index of Soft and Hard Margarine 0ils, Margarine 0il Slends
and Butterfat.

SoIid fat index

Sample 1 00c 21 ,10c 33.30C

S (normal range) 1 0-'1 4 6-9 2-4

s 14.212 9. 28 3. 56
HS 37 20,16 1 0.89 3.26
HS 46 20.93 11 .28 3.32
'1
HS 55 22.07 1 .9'1 3.39
HS 64 24.29 13.20 2
"97
HS 73 25.05 '13.33 2,91
H 29.55 14,96 2
"99
H (normal range)t 26-28 13-15 2-3 "5

Butterfat 36. 60 21 .10 4,25

1 Source: G. Davidson (PersonaI communication ).

2 Result of a single measurement.
 79

In order to explain the relationship between the oil blend composi-

tion and the SFI values, the crystallization behaviour of the margarine
oils must be understood. soft margarine oils were formulated to
The
give a low degree of crystallization i.e. low SFI values at refrigera-
t.ion temperatures so that a spreadable margarine can be obtained,
whereas the hard margarine oils were formulated so as to yield a firm
product. A progressive increase of hard margarine oil content in the
oil blends is therefore matched by an increase in SFI values - a rise in
the degree of crystallization at a given temperature.

Berger and White (1971 ) pointed out that excessive churning of the

fat causedby over destabilization of the oi1/water emulsion would

impair the whipping quality of ice cream mixes and adversely affect the
quality of the finished products. One of the factors which promotes
churning is the presence of soft fat (Iow in crystallinity) in the emul-
sion. Such fat contains a greater liquid oi1 content which is conducive
to excessive churning. This phenomenon is also seen in the non-dairy
frozen dessert mixes. It is probable that a similar influence of fat
crystallinity on fat destabilization in ice cream mixes may have also
existed in the frozen dessert mixes during the freezing process.

It is logical to suggest that the excessive fat destabilization in

the frozen desserts can perhaps be managed by the use of vegetable fats
which show crystallization behaviour similar to that of milkfat. But
this may not be feasible. it is because even though vegetable oil can
be hydrogenated to such an extent that it becomes similar to milkfat in
the SFI profiJ.e (oil stock C3, Table 8), the oil is processed as an
ingredient in formulating margarine oils (fables 9 and 10) and is not
available in the market place. This research, on the other hand, r,las
 80

aimed at.utilizing only commercially available materials for the devel-

opment of a non-dairy frozen dessert. Further, it was found in the
preliminary studies that frozen desserts prepared with pure hard marga-

rj.neoil were not judged as acceptable by a taste panel. The use of

oils with higher SFI values is not likel-y to improve the quality of the
frozen desserts.

4.3 RESUTTS OF TESTS PERFORMED ON THE FROZEN DESSERT MIXES

4.3.'1 Microbioloqical Tests

These tests were carried out to ensure that the mixes were produced
under sanitary conditions. All the Standard Plate Counts were low
(tables 19 and 20). The coliforms test also showed negative results for
all samples. results conformed to the Canadian Food and Drug
These
regulations (1982) of a standard plate count of less than 100,000
organsims/g and a coliforms count of less than 10 organisms/g. The test
results also indicated that the pasteurization process was effective in
controlling the bacterial content of the mixes and that the subsequent
processing steps had been carried out with proper sanitation measures.
 81

TABLE 1 9

Results of Standard Ptate Count (SpC) and Presumptive Colifornrs Test of

the Non-Dairy Frozen Dessert Mixes made with Five Different Margarine
0il Blends.

sPc (cru/g) Coliforms Count (organisms/g)

Sample Trial'1 Trial 2 mean Trial 1 Trial 2 mean

HS 37 1
,600
1
1 ,400 ,500 <'l <1 <'l
HS 46 1 ,400 1 ,300 ,400 <1 <1 <l
HS 55 1 ,400 2,100 ,800 <1 <1 <t
HS 64 1 ,400 1 ,400 ,400 <'1 <1 <1
HS t5 I
,500 1 ,700 ,600 <1 <1 <1

1 Average of duplicate plates.

TÀBtE 20

Results of Standard Plate Count (SpC) and Presumptive Coliforms Test of

the Non-Dairy Frozen Dessert Mixes Prepared at Four Different PPI
LeveIs.

sPc (cru/g) Coliforms Count (organisms/g)

Sample Trial 1 lria1 2 mean Trial 1 TríaI 2 mean

P35 1 ,600
1
2,000 1,800 <1 1
<1 <1
P50 2,600 3,000 2 ,800 <1 <1 <1
P65 1,700 2,100 1 ,900 <1 <,1 <1
P 80 2,400 1 ,700 2,100 <1 <1 <1

1 Average of duplicate plates.

4.3 .2 Ë and Titratable Aciditv

No references t+ere found in the literature regarding the pH and

titratable acidity of non-dairy frozen dessert mixes contai ning PPI and

canola oiI. However, the ingredients of these mixes are similar to

those used in ice cream mixes except that the protein and fat are from
 82

plant sources. Therefore, factors which influence the pH and titratable

acidity of ice cream mix may also be applicable to the mixes in this
research. Àrbuckle (1986) pointed out that an increase in MSNF content
raises the titratable acidity and lowers the pH of ice cream mix. Milk
components r+hich contribute to the acidity are the proteins, mineral

salts and dissolved gases. But the bulk of the acidity in milk is
attributed to the phosphates and caseins (niel, 1985). An ice cream mix
containing 10% f.at and '10 Lo 11% MSNF (rabte ¿) will have an acidity
between 0.180 and 0.1g8%, and a pH of 6.32 to 6.31 (Arbuckle, 1986).

The pH of the frozen dessert mixes was likeIy to be influenced by the

PPI and, to a lesser extent, the water which may have contained minerals

and dissolved gases used in preparing the mixes (Tables 21 and 22).
Further, residual acid or alkali in the PPI froni the manufacturing
process (Section 2.4.1.4) may have also affected the pH of the mixes.
This factor may help to explain the consistency of the pH readings
within each trial and the discrepancies between the two trials in stage
one of the research (rabte Zl ) because the PPI used in each trial were
not from a common lot. It may also explain the difference in the pH
readings between the mixes of stage one and two. Pea protein isolate
from a different batch was used during stage two and hence there !¡as no
difference in the pH readings between trials (taUte ZZ). Such confusion
stems from the fact that a large quantity of PPi from a common lot could
not be secured to meet the needs of the entire research.

In milk, the majority of its titratable acidity is attributed to the

phosphates and caseins (niet, 1985). The principal source of acidity of

the frozen dessert mixes would therefore be Èhe pea proteins since they
 83

TABLE 21

pH and Titratable,{cidity of the Non-Dairy Frozen Dessert Mixes made

with Five Different Margarine 0il Blends.

pH Titratable acidity (%)

Sample Trial'1 Trial 2 mean Trial 1 Trial 2 mean

HS 37 6.351 6.10 6.23 0.1 0. 3 0.12

HS 46 6.35 6.10 6.23 0"1 0. 3 0.12
HS 55 6"35 6.10 6.23 0.1 0. 3 0.12
HS 64 6.40 6.10 6.25 0.1 0. 3 0.12
HS 73 6.3s 6.10 6.23 0.'1 0. 3 0.12

1 Àverage of duplicate measurements.

TABLE 22

pH and Titratable Acidity of the Non-dairy Frozen Dessert Mixes Prepared

at Four Levels of PPI.

pH Titratable acidity (%)

Sample Trial 1 Trial 2 mean TriaI 1 Trial 2 mean

P 35 6.70 1
6.80 6.75 0.061 0.06 0.06
P50 6.80 6.80 6.80 0 .07 0,08 0.08
P65 6"70 6.80 6.7s 0.10 0.10 0.10
P80 6.7 0 6.70 6.70 0.'13 0.13 0.13

1 Average of duplicate measurements.

were the only mix component capable of reacting with alkali to a great
exlent. This is because proteins are amphoteric in nature. If any
residual acid from the manufacturing process existed in the PPI it
'
would have also increased the titratable acidity of the mixes.

in stage one of the research the titratable acidities within each

trial are fairly constant because all the mixes contained one level of
 84

PPI (raUte Zl). The difference between trials was likely caused by the
use of PPI of separate lots in t.rial. In the second stage, ro
each
significant difference tras noted in the titratable acidity between and
within trials because all the mixes were made from a common batch of PPI
(tabte ZZ). However, the titratable acidity of the mixes in this stage
increased with increasing PPi 1evel. This observation is similar to
that seen in ice cream mixes where the acidity is raised by increasing
MSNF content. As the PPI level in the frozen dessert mixes increased,

so did the protein content and hence an increase in the mix acidity.
The pH and titratable acidity of the frozen dessert mixes were not
adjusted to approximate those of ice cream mix because this study was
aimed at utilizing the PPI on an as-is basis so that its performance in

this type food system could be studied.

4.3.3 Protein Content of the Mixes

In the first stage of the research, only one level (5%) of PPI was
used in the preparation of the mixes because it yielded approximately 4%
of protein on a dry weight basis. This protein content was required in

the frozen dessert its formulation was modified from the general
because

composition of an economy grade ice cream (table ¿). This type of ice
cream contains 9.0 to 11.0% of. MSNF which translates into a protein

level of 3.3 to 4.1% since nearly 37.0% ot MSNF is made up of milk

proteins (Arbuckle, 1986).

The protein content of all the mixes was higher than 4% (Tab1e 23).
This is because the PPi powder was used on an as-is basis, no weight
adjustment was made in order to give all the mixes exactly 4% protein.
 85

Hov¡ever, there are some fluctuations in the protein content of the

mixes. Normal variations in the precision and repeatability of the
measurement of the PPi powder and the Kjeldahl t.est may have also
contributed to these observations.

In addition the 5% PPI level used in the first stage, three different
levels r¡ere i.ntroduced into stage two. The purpose of this design was

to study the feasibility and the effects of incorporating higher or

lower levels of PPI powder into the frozen desserts, A higher PPI level
would increase the food value of the frozen desserts because of an

increase in the protein content, while a lower PPI leve1 may help to
improve the sensory quality of the products. There r+as a progressive
increase in the protein content of the mixes with increasing PPI levels
(taUte Z¿). Factors which may have contributed to the variations in the
protein content between the two trials were pointed out in the previous
paragraph.

IÀBIE 23

Protein Content of the Non-Dairy Frozen Dessert ì4ixes made with Five
Margarine 0il Blends.

Protein Contenl &)1

Sample Trial 1 Trial 2 mean

HS 37 4.27 1
4. 18 ¿. )?
HS 46 4,22 L 1L 4.13
HS 55 4.16 4.24 4.20
HS 64 4.29 4.27 4.28
HS 73 4. 18 4.12 4.15

1 % Protein = % N x 6.25 (wet weight basis); average of duplicate

measurements.
 86

TABLE 24

Protein Content of the Non-Dairy Frozen Dessert Mixes Prepared at Four

Levels of PPI.

Protein Content (%)1

Sample Trial 1 Trial 2 mean

P35 2.991 2.78 2 "83

P50 4.08 4.15 ô. 1)
P 65 5. 38 5.45 5.42
P80 6. 55 6.82 6 .69

1 % ProLein = % N x 6.25 (wet weight basis); average of duplicate

measurements.

4.3"4 Fat Content of the Mixes

Since only one level of fat (10.50%) in the preparation of

was used

all the mixes, the fat test results of the mixes should be fairly
consistent, but the fat content vras seen to vary between 10.25% anð
10.75% in stage one (tabte ZS) and between 10.25% and 10.50% in stage

two (rabte Z6). This could be the result of variaLions in the measure-
ment of ingredients during the processing of the mixes, although it
shoul-d be noted that the fat content of the mixes also included the fat

contributed by the PPI (ta¡te lZ). Nevertheless, these fluctuations are

usually observed in food production lines due to the method and type of
equipment used to measure the quantity of ingredients and; due to losses
through handling. Such variations are also generally tol-erated by the
formulation of the product. An example is that an economy grade ice
cream may have a fat content of '1010.5%. Set against this 0.5% toler-

ance range, the variations seen among the mixes of both stages are
acceptable. As a result, it is believed that these fluctuations were
not sufficiently large to affect the quality of the frozen desserts.
 87

TÀBLE 25

Fat Content of the Non-Dairy Frozen DesserL Mixes made with nive
Margarine 0i1 Blends"

Fat Content (%)

SampJ.e Tr iaI 1 Trial 2 mean

HS 37 10.501 10.75 10.63

HS 46 10.25 10.50 10.38
HS 55 10.25 10.50 10.38
HS 64 10.50 10.25 10.38
HS 73 10. 50 10.75 1 0.63

1 Average of duplicate measurements.

TABLE 26

Fat Content of the Non-Dairy Frozen Dessert Mixes Prepared at Four PPI
LeveIs.

Fat Content &)

SampIe Trial 1 '1'rra-L Z mean

P 35 0.50
'1 1
0.2s 10.38
P50 10.50 0.25 10.38
P65 10.50 0.50 10.50
P80 10.25 0.25 10.25
'l Average of duplicate measurements.

4.3.5 Total Solids Content of the Mixes

total solids conLent of the mixes in stage one is quite consis-

The
tent because these mixes were prepared according to a common formulation
(tabte Zz). À11 the nixes at this stage should contain approximately
37,60% total solids according to the formulation. However, Table 27

sho¡vs that the total solids content of the mixes was higher. Two
factors may have contributed !o these observations. First of which is
 88

the loss of moisture during the pasteurization of the mixes by the batch
method in which they were heated for 30 min. The second factor
at 1750F

may have originated from the variations in the measurement of ingredi-

ents. Àn excess of dry ingredients or insufficient water used during

processing could have altered the total solids content of the mixes.

In stage two, the mixes v¡ere prepared at four levels of PPI and hence
an increase in PPi levels was matched by an increase in the total solids
content (table 28)" The PPI level at this stage increased in 1"5%
increments. Since the PPI contained 5% moisture, 1"5% of. PPI would
yield 1.43% dry matter. The total solids content of a mix should there-
fore differ from the one containing a higher or lower level of PPI by
1.43%. However, this r+as not observed in Table 28. Factors which could
have contributed to these variations were discussed in the previous
paragraph.

TÀBLE 27

rotal solids Mixes made çith

'"';ii: ;T,!::"i;';3?åll,l:"ãiT iiffi::

Total Sotids Content (%)

SampIe Trial 1 Trial 2 mean

HS 37 38 "221 38.27 38.25

HS 46 38. 13 37"85 37.99
HS 55 38. 57 38.65 38.61
HS 64 38.32 38.27 38.30
HS 73 38. 50 37.88 38.19
'1 Àverage of duplicate measurements.
 89

TABLE 28

Total solids content of the Non-Dairy Frozen Dessert Mixes Prepared at

Four PPI Levels.

Total SoIids Content (%)

Sample Trial 1 Trial 2 mean

P 35 36.25 1
36.26 36.26
P50 37.56 37.97 37.77
P65 39.74 40.00 39.87
P 80 40.78 41.12 40.95

1 Average of duplicate measurements

4.3"6 ViscositL of the Frozen Dessert Mixes

The apparent viscosities of the mixes of stage one and two are shown
in 30, respectively. Both tables show that these mixes
Tables 29 and
rvere pseudoplastic in nature, âs indicated by a decline in viscosi.ty

with increasing shear rate (Mohsenin, 1978). The flow of homogenized

ice cream mix r,ras also pseudoplastic as observed by Dickinson and
Stainsby (1982). it can be speculated that the internal structure of
the frozen dessert mixes might have been very similar to that of ice
cream mix in order for them to exhibit such a close resemblance in rheo-

logical behaviour" However, all the frozen dessert mixes tlere more
viscous than the ice cream mix prepared in this research (tables 29 and
30). This could be the result of the preparation of the ice cream mix
according to a different formulation (faUte ZS). Às seen from Tables 29
and 30, the viscosity readings of the two trials of a sample did not
always agree cJ.osely at the same shear rate. This is very likely caused
by the variations in the fat, protein and total solids content of the
mixes from the two trials. It was also shown in the two tables that in
 90

some instances such as HS 64, the apparent viscosity increased shor t 1y

after the measurement had begun. This is a usual phenomenon as the mix

was undergoing a stabi lization process in reaction to the appl i ed

stress.
 TARLN 29
ApÞarent viscosities of the ¡lon-tlairy Fro¿en Dessert Mixes l4¿de with five Dittcrent Mðrçðrine Oil. B!ends.

Apparent viscosity (pas) at 20oC

Sarrrple
HS 37 ils 46 l.{s 55 r(s 64 IIS 73

Shear Rate (s'r) Trial. 't'r ia I Trial '1'riaI 2 'l'r i a I Tr i a L 'I'r i a I lrial. Trial Tri¿i
0.03682 59 .52 |
r4 . 58 29.56 7.334 r5.82 8.405 14.61 10.19 l5.92 r0.58
0.05833 5l .0s I3.44 26.00 8.206 I2. BB 8.335 rB.9l 1? ,19 16. B0 r5.lB
0.09246 30.94 8.609 14.46 6.38 r 7.721 6.469 t5.04 10.3r r0. 89 r3.94
0. r4 65 16. 58 5.139 7 .892 4.485 4.655 4.660 1.037 7.3?2 6.332 r0.50
0.232t 9.821 3.869 5 ,342 3.73 r 3.59 r 3.683 7.559 s.289 4.404 7.117
0.3686 5.050 2.449 3.255 2 .492 2.404 2.sgg 9. y) 3.387 2,7 S4 3.699
0. s8 3 3 3 . 164 r.6tB 2.079 1.778 r .652 t.82t 3.3r5 2. t33 t.761 2. 155
0 .9232 2.041 1 .
102 1 ,467 1.255 I . 1u7 1.29?_ 2 .237 I . 4 01 1 . 165 r.373
1 .467 r .382 0 . 7 68 l .020 0.9083 0. B5l0 .599
3 O.9BBB 1 I .007 0.Bl3s 0.9310
2 .322 r.018 0.5591 0.7085 0.6699 0.6sBB 0.7850 L 204 0.7788 0.6140 0.6716
3.682 0.7590 0.4298 0.5 r96 0.4790 0.4966 0.6083 0.9249 0.5908 0.4316 0.4954
s.833 0.5900 0.3403 0.39s7 0.4 I 58 0.3996 0.5r43 0,7425 0.4BtB 0.3389 0,3B9s
9.24 6 0.4795 0.2875 0.3r86 0.3473 0.32s8 0.4446 0.62s4 0.4113 0.2663 0.3150
t4 .65 0.39 r5 0.2s21 0.2633 0.2850 0.2684 0.3775 0. s l BB 0.3467 0 .2226 0 ,27 6t)
23 .26 0.3251 0.208 0.2191 0.23s8 0.2230 0.3218 0.4 366 0.2926 0. r861 0.2349
36.86 0 .27 37 0. rB09 0. tB66 0. r 955 IJ.IB9O 0 .27 49 0.3686 0,2495 0. r59l 0 .2û22
58.33 0.2344 0. l585 0.r619 0. 163 l 0.1622 0.2339 0,3071 0.2r33 0. t391 0. r764
oa îa 0.2029 0. 1393 0.1417 0. 1369 0. 1396 0. r 990 0.2s2?- 0.1820 0.1233 0 . r 54 4
146.7 0.1719 0 .1226 0. r235 0. I 156 0. l I 84 0.1705 0.2048 0. rs3B 0. 1086 0. rl50
232.2 0. 1381 0. 1084 0. 1077 0.0985 0.09279 0 . I 4 5B 0. 1662 0,1296 0.09638 0. r t30
3 68 . 2 0. r 168 0.09557 0.09394 0.08487 0.07849 0.1247 0. r352 0. r092 0.08634 0. r031
583.3 0.09947 0.08409 0.08 r 98 0.07423 0.07003 0. r 068 0. r 107 0.09229 0.07756 0.09028
924.6 0.08459 0.07392 0.07238 0.06578 0.06234 0.09 r 63 0.09 r 50 0.07668 0.06992 0.079fi3
r465 0.06859 0.06539 0.064 rs 0.0s866 0.0538 r 0.078 r9 0.07639 0.0683ì 0.06272 0.0/u28
1 Àver,lge oI dupl icate me.rsuremenLs.

Lo
 TADLE 3O

Àpp,lrent viscosiIies of llon-Dairy Prozen Dessert Mixes Prcparec] at Four ppl LevcIs aDcì of Ice crca¡ìr I.iix.

Apparent viscosiry (eas) a¡ 20oC

Sarnpl e

35 P 50 P65 P BO

Shear Rate (s-r) Trial Trial Trial Trial 2 Tr ial Trial Trial 'lr ia 1 fce Crea¡lr Mir.
0.03682 4.8621 3. r83 t1 .ql 15.72 r 6.84 23 .28 s9. 6 r 59.34 r5.40
0.05833 5.191 3.070 t4.Bl t3.4t 28.91 IB.B9 56.32 56.04 8.6?5
0.09246 4.0r3 2.593 l0.lB 9.3ss 2t.36 13.66 39. B4 4 0.65 1 . 165
0. ¡4 65 2.730 1.9.1 9 6.787 6.63 l 13.70 9.78 r 2s .97 21 .99 2. ì49
0.2326 2 .328 t .922 4.969 5.073 9.534 7.428 r8.57 20.08 l . 54 9
0.3686 t.616 r .323 3.369 3.566 6.214 5.136 12.75 r3.65 0.9630
0.s833 I . ls2 r .062 2.305 2.541 4.283 3.680 IJ . BB 6 9. 4 64 0.6228
0.9232 0. B I 87 0.7693 I .640 1 .857 2.965 2.673 6.209 6.664 0.4040
t .467 0. 6199 0.5697 I . 199 I .32t1 2 .132 l .967 4.386 ,1 .7 96 0.25i2
2 .322 O.4BBB 0 .4 4 54 0.8952 0.9954 r.589 1.462 3.209 3 . 4 5B 0. rBr7
3.682 0 .3622 0.3495 0.6401 0 .7 626 | . 140 r .087 2.345 1 C11
0. r10l
s.833 0.29 r6 0.2900 0. s00 r 0.5909 0.8758 0.8034 1.764 r.895 0. r062
9.246 0 .237 9 0.2396 0.39 |6 0 .4 697 0 .67'¿2 0.6533 l .339 1.447 0.0852
r4.65
r

0.2008 0.204 l 0. 3 r 59 0.3776 0.5346 0.5193 I . 04 2 r.12t 0.0689?

23 .26 0. r 700 t 0. t770 tJ.2577 0.3088 0.4266 0,4 r 97 0.824s 0.BrJ60 0.0s929
36.86 0.1443 0. r 55s 0. 2 t 53 0.2594 0.3478 0.3458 0.6683 0.7 l59 0.05201
58.33 0. 129 r 0. l37B 0. | 832 0.22t6 0.2899 0.2906 0.55 r7 0.5909 0.0q 617
92 .32 0.1140 0.1227 0. r 59 r 0.1916 0 .247 t 0.2493 0.4649 0 .497 7 0.u42J9
r46.7 0. ¡ 007 0. r09 | 0. r3B2 0. r669 0.2142 0. 2 I 69 tì ?ot1 0.4234 0. u390s
232.2 0.08927 0.09706 0 ,1232 0. r463 0.lB7B 0.1913 0.3399 0.363 r 0.03641
368.2 0. 07915 0.08640 0. r0BB 0.1279 0. 1654 0. r69l 0.2908 0.3107 0.03416
sB 3 . 3 0.07014 0.07667 0.09579 0.1il2 0. r452 0. r48? 0.2485 0.2647 0.0328rJ
924.6 0.06204 0.06779 0.08398 0.09609 0. 'l 267 0. 1300 0.2r33 0 .2255 0.03155
t465 0.05343 0.05881 0.07292 0.08l2l 0. l09l 0. r r02 0.0l0l l
Average of duplicale nreasuretirenLs.

\o
t-.)
 93

Although the fat content of the mixes fluctuated between 10.25% and
10.75% (rable 25), it was not observed to have a direct relationship

with the apparent viscosity of the mixes (rabte Z9). The SFI values of
the fat did not have any noticeable effect on the mix viscosity either.
But it appears that the type of fat used in frozen dessert mixes may
have some influence on the mix viscosity. Youssef et al. (1981) found
that the mix viscosity decreased with increasing level of butterfat
substitution by cottonseed and corn oils in their ice cream mixes. In a
similar study, E1-Deeb et al. (1983) observed that the mix viscosity
increased with increasing substitution leveI of butterfat by hydrogen-
ated oils in their mixes. The resulLs of these studies suggested that
whether a vegetable oil would differ from butterfat in the ability to
affect mix viscosity may be dependent on its source and type.

The protein level of t.he in this study, however, showed a

mixes
direct relationship with the apparent viscosity. TabIe 30 indicates
that as the PPI level increasedr so did the apparent viscosity of the
mixes. Such a trend is expected because proteins dispersed in water
exhibit a gelling and/or thickening effect (tgoe, 1982) due to their
ability to bind and immobilize large amounts of water (Àndreasen, 1985).
Ðuring the aging period, the milk proteins in ice cream mix become fulIy
hydrated (¡titten and Neirinckx, '1986). The result is an increase in the
mix viscosity after aging. It is reasonable to suggest that a similar
phenomenon had also occurred in the frozen dessert mixes. The combina-

tion of Lhese effects were intensified by the increasing pea protein

content of the mixes. The increase in the protein content also corre-
sponded to a steady decline in the overrun of the frozen desserts (table
 94

31). In other word, the overrun decreased with rising viscosity of the
mixes. This agrees with Arbuckle (1986) who stated that the whipping
rate of ice cream mix was generally retarded as the mix viscosity
increased.

Plant protein preaprations in general seem to be more effective than

milk proteins in imparting viscosity to the mix. Lawhon et aI. (1980),
El-Deeb and Salam (1984) and Gabriel et al. (1986) found that as the
level of MSNF replacement with plant protein preparations increased in
their frozen dessert mixes, so did the mix viscosity. The viscosities
of these mixes were always greater than that of the control. Bul
Simmons et al. (1980) observed that the increase in mix viscosity was

not always proportional to increasing replacement level and that the

mixes containing plant protein preparations were not, as a rule, more
viscous than the control. It appeared that whether plant proteins were
more effective than milk proteins in imparting viscosity to the mix was

dependenL on the replacement leveI, and the source and type of the

protein preparations involved.

4"4 RESULTS OF TESTS PERFORMED ON THE FROZEN DESSERTS

4 "4.1 Overrun of the Frozen Ðesserts

The overrun of the non-dairy frozen desserts from the first stage and
the second stage of the research is shown in Tables 31 and 32, respec-
tively. In stage one, all the mixes were prepared with identical formu-
lations. The only variable at this stage was the composition of the fat
used in the mixes, although some fluctuations in the fat content were
observed (table 25) " These variations may have affected the whippa-
 95

blility of the mixes because foam formation during freezing is depressed

as the level of fat rises (Berger et al. 1972a), but they were perhaps
too small to influence the overrun of the frozen desserts as the overrun
was not seen to have fluctuated with the fat content of the mixes.

However, the overrun of the frozen desserts was affected by the

composition of the fat the mixes contained. It to increase

was observed

with the rising SFI values of the fat in the mixes. This is because
excessive fat destabilization which retards the whipping rate of the mix
is promoted by fats with large liquid oil fractions (Berger and White,
1971). Àn increase in SFI values implies a lowering of the liquid oil
content of a fat (waddington, l986). As the fat in the frozen desserts
contained a greater fraction of hard margarine oil i.e. a higher SFI and
a lower liquid oil content, the detrimental effect of fat destabiliza-
tion on mix whippabiLity was lessened. This resulted in an increase in
the overrun. Nevertheless, the overrun of the frozen desserts was not
as high as that of the ice crearn 191.7%) prepared in this research.
This is because the frozen dessert mixes and the ice cream mix were not
similar in composition (tables 13 and 14). Àlso, hydrogenated vegetable
oils have been shown to depress the overrun of ice cream. El-Deeb et
al. (1983) observed that the overrun of ice cream declined steadily as
the leveL of milkfat substitution by hydrogenated vegetable oils
increased. Further, the overrun of these ice cream was lower than that
of the control. But the overrun-depressing effect appeared to be
related to the type of oils involved. In a similar study using cotton-
seed and corn oils, Youssef et aI. (1981) did not observed any detri-

mental effect of milkfat substitulion by cottonseed and corn oils on

overrun.
 96

In the secondstage, the overrun of the frozen desserts decreased

with rising PPI levels in the mixes (fa¡te ¡2)" A study by Simmons et
aI. (1980) using plant protein isolates to partiall-y replace MSNF in
frozen desserts showed that the overrun declined with increasing level
of replacement. Ànother study by El-Deeb and Salam (1984) using seed
flours also resulted in similar findings. Both investigations found
that the overrun of the controls was higher. it is possible that the
ability to impair overrun v¡as an inherent property of plant protein
preparations. But this property may be dependent on the type and source
of the protein involved because Gabriel et aI. (1986) found that
increasing level of MSNF substitution with groundnut protein isolate did
not result in a corresponding decrease in overrun.

Às was mentioned earlier that the apparent viscosity of the nixes

increased with increasing level of PPI (rabte ¡O). Àrbuckle (1985)
pointed out that as the viscosity of ice cream mix increases, the whip-
ping rate becomes retarded. But the studies by Simmons et al. (1980)
and Gabriel et al. (1986) showed that an increase in the viscosity of
their frozen dessert mixes containing plant protein preparations did not
necessarily result in a decline in the overrun. Perhaps whether the
overrun rqould drop with rising mix viscosity was again a function of the
type and source of protein preparations used in the mixes.
 97

TÀBLE 31

Overrun of Non-Dairy Frozen Dessert Mixes made with Five Different

Margarine 0il Blends.

Overrun (%)

Sample Trial 1 Trial 2 mean

HS 37 50.0 1
51 .7 50 .9
HS 46 52.4 54 "2 53 .3
HS 55 54. 9 59 "7 57 .3
HS 64 60. 1 62.0 61 .1
HS 73 63.2 65.4 64 .3

1 Average of six measurements.

TÀBLE 32

Overrun of Non-Dairy Dessert Mixes Prepared at Four Different PPI

Levels.

Overrun (%)

Sample Trial 1 Trial 2 mean

P 35 63.4 1
57 .2 65.3
P50 57.1 55.6 56.4
P65 45.8 44 "2 45.0
P80 35"4 39.4 37 ,4

1 Àverage of six measurements.

4"4,2 Sensory Ànalvsis of the Non-Dairy Frozen Desserts

Regular ice cream was not used as a control in the sensory evaluation
of the frozen desserts because Lhe objective of the sensory testing in
this research was to judge the frozen desserts based on their orvn quali-
ties per se, rather than in comparison to those of ice cream" Simp1e
preference test alone was used because of the exploratory nature of this
 98

investigation due to the lack of information on the properties of this

PPi.
type of products which contain On1y the frozen desserts from the
second trial of each stage of the research were subjected to sensory

evaluation and each panel was not repeated. This is because replication
in preference testing is not a typical practice since the objective of
the test is to provide generalized results on sample differences (Stone

and Side1, 1985). Furthermore, replication in hedonic testing is not

effective as a consistency check on the panel results (O'Mahony, 1986)"

The analysis of variance of the sensory evaluation data of the frozen

desserts of the first stage is shown in Appendix B. The analysis showed
that there were significant variations among the samples and among pane-
lists at the 5% level of significance. Despite the significant varia-
tions among the panelists, analysis of the sensory evaluation data by
the coefficient of concordance (a = 0.05) showed that the ranking of the
samples was not a random event (Appendix C). This implied that there

vras a general consensus among the panelists regarding the order of pref-
erence for the samples.

The order of preference for the frozen desserts was oblained by

ranking the mean scores (taUte ¡¡). On hand, the order of

the other
preference established by the sums of ranks is slightly different, but
it is considered to be an accurate estimate of the preferential order of
the samples (Kenda11, 1970). The degree of acceptabiLity or preference
increases with the magnitude of the mean score or the sum of ranks.
Despite the slight discrepancy in the ranking of the samples by the two
orders of preference, they both ranked HS 64 as the most acceptable
product" The Tukey test results (fable 33) revealed that HS 64 was
 99

significantly more acceptable lhan HS 3746, but did not differ

and HS

from HS 55 and HS 73 in acceptability at the 5% level of significance.

HS 55 and HS 73 did not in turn differ significantly from HS 37 and HS

46, The differences among the samples may therefore be marginal and
were not very distinctive to the panelists. This may help to explain
the slight variation. seen in the ranking of the frozen desserts based on
the mean scores and on the sums of ranks.

TÀBLE 33

Sums of Ranks and Tukey Test Results on the Mean Scores of the Non-Dairy
Frozen Desserts made with Five Different Margarine 0i1 Blends.

Sample HS 64 HS 55 HS 73 HS 37 HS 46

Sum of Ranks 1 19'l .5 1 58.5 163.5 1s1.0 130"5

Mean Score2 6.5.1 a 5"94 ab 5" 87 ab 5.77 b 5"40 b

Honeslly significant difference for the Tukey test = 0.69 (s = 0.05).

Mean scores with the same letter(s) are not significantly different at
the 5% level of significance.
'1from Àppendix C.
2 f.ron Àppendix B.

Although the fat content of the frozen dessert mixes varied (lable
25) , there vras no direct relationship between these variations and the
sensory scores of the end products. 0n the contrary, the composition of
the fat may be the dominant factor which had influenced the quality of
the frozen desserts at this stage. This can be seen in Table 25 that
although HS 73 contained more fat than HS 55, the mean scores of the two
samples did not differ significantly (table 33). However, both HS 73

and HS 37 had identical leveIs of fat (fa¡te 2S) and yet the latter r+as
 100

significantly less preferred by the panelists to HS 64 based on the mean

scores (fable 33). These observations suggest that the composition

rather than the content of fat which may have affected the quality of
the frozen desserts. This point is reflected in two studies. Youssef
et al. (1981) aia not find any direct relationship between the overall
sensory score and the level of substitution of butterfat with corn and
cottonseed oils in Íce cream. The controls scored higher. On the

contrary, E1-Ðeeb et al. (1983) found that as the level of milkfat

substitution with hydrogenated oils increased in their mixes, the
overall score of the resultant ice cream declined accordingly. The
findings of these studies suggested that vegetable oils may affect the
sensory qualities of ice cream to which they were added, but the effect
appeared to be influenced by the type as well as the amount of oil used.
Since the frozen desserts in this research contained only vegetable
oils, it is safe to infer that they are inferior to ice cream with
respect to their sensory quality.

As for the first stage of this study, the oil blend HS 64 was found
to be the optimum fat composition which could be used to rnanufacture an
acceptable frozen dessert. The mean .scores of the frozen desserts
decreased as the oil blend composition move toward HS 55 or HS 73 (table
33). This in essence provides a fat composition/Srl range for formu-
lating a fat, in conjunction with the PPI, which is able to produce a
satisfactory frozen dessert. Àccording to the findings of this stage of
the study, the optimum SFI for a fat to be use in the frozen dessert mix
would be 24.29 at 100c (HS 64), within a range ot 22,07 (gs SS) and
25.05 (ss Zg) (ra¡re 1g).
 101

The analysis of variance of the sensory evaluation results of stage

two is shown in Àppendix D. Significant variations (a = 0.05) among the

panelists and among the samples tested were revealed. Despite the vari-
ations among the panelists, the analysis by the coefficient of concor-
dance (q = 0.05) indicated that the paneJ-ists generally rank the samples
in the same order according to some characteristics of the samples
(Appendix E). The mean scores and the sums of ranks of the frozen
desserts are shown in Table 34. Both the mean scores and the sums of
ranks dropped steadily as the PPI levels increased. The Tukey test
results (rabte g¿) showed that P 35 was significantly higher than P 65
and P 80, but not P 50 in the mean score. Both P 50 and P 65 did not
differ from each other but scored significantly higher than P 80. The
order of preference for the samples according to both the mean scores
and the sums of ranks was P 35 first, P 50 second, P 65 third, and P 80
last. The panelists'order of preference for the samples at this stage
clearly indicated that high levels of PPI in the mixes negatively affect
the sensory qualities of the frozen desserts.

TÀBtE 34

Sums of Ranks and Tukey Test Results on the Mean Scores of the Non-Dairy
Frozen Desserts Prepared at Four Different PPI Levels.

SampIe P 35 P50 P65 P80

Sum of Ranks 1 1 52.0 131.5 115.5 81.0

Mean Score 2
6.65 a 6.19 ab 5.73 b 4.60 c

Honestly significant difference for the Tukey test = 0.69 (a = 0.05).

Mean scores with the same letter are not significantly different at the
5% level of significance.
1 from Appendix E.
2 from Appendix D.
 102

that the incorporation of plant protein preparations

The suggestion
into ice cream would affect the finished product quality have been
reflected in two studies. E1*Deeb and Salam (1984) and Gabriel et al.
( 986) found that increasing leveIs of MSNF substitution with plant
1

protein preparations in ice cream mix was matched by a steady decline in

the sensory/overall score of the resultant ice cream. However, two
earlier studies produced slightly different results. Simmons et a1.
(1980) found that increasing 1eve1s of MSNF substitution with plant
proteins did decrease the overall acceptability of ice cream. Lawhon et
al. (1980) also obtained similar results with some exceptions that the
substitution level did not show any direct relationship with overall
acceptability. But in both studies it appeared that whether the lower
scores of the experimental products were significantly different from
that of the control was influenced by the substitution level and the
type and source of protein preparations used. Às the present investiga-
tion is concerned, the PPI level of 3.5% yielded the most acceptable
frozen dessert as compared to those containing higher levels of PPI.
Significantly less acceptable products resulted from the use of PPI
leve1s higher than 5%. This indicated that the 5% level was the upper
limit to which PPI could be used in the mix r+ithout negatively affecting
the quality of the frozen dessert.

4.4.3 Meltinq Oualitv of the Frozen Desserts

The melting quality of the frozen desserts was studied because it is

an objective method of assessing the influence of the formulation and
processing conditions on their quality. Figures 4 and 5 show the
 103

melting behavíour of the frozen desserts from trial one and two of the

first stage of the research, respectively. it was shown Lhat the frozen
desserts in the two trials at room tempera-
melted in a uniform manner
ture r,¡ithin a period of one hour. No melting quality def ects were
noted. This suggested that the frozen desserts had been properly formu-
lated and processed.

The meì-ting quality of the frozen desserts of the first and second
trial of stage two of the study is depicted in Figures 6 and 7, respec-
tively. While P 35 and P 50 in both trials melted completely at the end
of the one-hour period, P 65 and P 80 behaved differently as both showed
a slow rate of melting. It can be seen lhat P 65 still contained some
materials which had failed to melt after being exposed to room tempera-
ture for an hour. 0n the other hand, P 80 obviously suffered from the
does-not-melt defect as frozen materials were cLearl.y visible at the end
of the observation period. One of the causes of this melting defect is
overstabilization of the rnix. The mixes Í¡ere overstabilized as a resulL
of the high leve1s of PPI used in P 65 and P 80. Pea protein isolates
contain fiber and oligosaccharides (table lZ). These materials, as well
as the pea proteins, are able to bind water and therefore act like
stabilizers. Consequently, high levels of PPI, combined with the amount
of stabilizer in the formulation, resulted in an excessive stabilization
effect in the mixes.

The melting behaviour of the frozen desserts at this stage was also
seen to be related to the sensory scores. Às the PPi level in the
frozen desserts increased, so did their resistance to melting which was
in turn matched by a decline in the mean scores and sums of ranks (tabte
 104

34). The does not melt deiect is usually accompanied by a soggy, gummy

or sticky body. Figures 6 and 7 show that both P 65 and P 80 appeared

to be soggy. Àlthough the panelists were asked to judge the frozen
desserts on the basisof general preference, the negative qualities of
the samples must have been sufficiently noticeable to be reflected by

the low scores.

Figure 4: Melting Behaviour of the Frozen Desserts Prepared with Five
Margarine 0i1 BIends - Trial 1.

From top to bottom:

0 minute
20 minutes
40 minutes
50 minutes
106
Figure 5: Meì-ting Behaviour of the Frozen Desserts Prepared with five
Margarine 0iL BIends - Trial 2.
From top to bottom:
0 minute
20 minutes
40 minutes
60 minutes
10¡J
Figure 6: Melting Behaviour of Ice Cream and the Frozen Desserts Made
with Four Leve]s of PPI - Tria1 1.
From top to bottom:
0 minute
20 minutes
40 minutes
60 minutes
110
Figure 7: Melting Behaviour of the Frozen Desserts Made with Four
Levels of PPI - Tria1 2.
From top to bottom:
0 minute
20 minutes
40 minutes
50 minutes
IL2
 113

Lû.4. Microstructure of the Frozen Desserts

This part of the research was primarily carried out. to provide a

general comparison of the internal structure of the frozen desserts. It
was not intended to be a detail and in-depth study due to the many tech-

nical difficulties involved. 0nIy representative illustrations are

presented in this section. Photographs of the most and the least
preferred samples from each stage of the study were selected. Figures
8,9, 10 and'11 show the internal structure of HS 64' HS 46' P 35 and P
80, respectively. Each figure also includes a photograph of the struc-
ture of ice cream for comparison purposes. The photographs did not have
high resolution since the image of the frozen desserts lvere magnified 63

times. Higher magnification factor could not be achieved because the

cooling stage employed to prevent rapid melting of specimen made the use

of the more powerful objectives of the microscope impossible.

Figures 8, 9, and 11 shows that the ice cream specimen had a

'10

comparatively more defined structure than the frozen desserts. There

r.las no attempt made to identify the various features of the microstruc-

ture of the frozen desserts because such an undertaking is beyond the

objective of this study. Further, due to the low nìagnification and the
rudimentary technique used to prepare these photographs, it was not

possible to obtain accurate identification of the various structural

components. The specimens were noL dyed as the colours of the photo-
graphs may have suggested. The various colours in the photographs are

the combined result of the age of the slides from which the photographs
were prinled and the printing process. Nevertheless, the structure of
the frozen desserts Ì{as very obscured and ill-defined with the exception
 114

of P 35 which appeared similar to ice cream but its structural integrity

may be inferior to ice cream as indicated by the fractures observed.

These photographs suggest that both the vegetable fat and pea proteins
were notable to give a structure that is similar to ice cream to the
frozen desserts. This may in turn negativeLy affect the product
qual i ty/acceptabi 1 i ty.
Figure B: The Internal Structure of the Frozen Dessert Made r+ith the HS
64 0i1 Blend and of lce Cream.
Magnification: 63X

Top: HS 64
BotLom: Ice Cream
Figure 11: The Internal Structure of lhe Frozen Dessert Containing 8.0%
PPI (P gO) and of Ice Cream.
Magnification: 63X

Top: P 80

Botton: Ice Cream

 ChaPter V

SUMMÀRY ÀND CONCLUSIONS

The objective of this study was to combine commercial pea protein

isolate (ppI) and oil blends prepared from hard (H) and soft (S) marga-

rine oils to produce a non-dairy frozen dessert of acceptable quality.

To achieve this goal, the study was divided into two stages. In the

f irst, oil of f ive weight proportions 1nlfs: 30170, 40160,

blends
150150, 60140 and 70/30) and a common leveL ß%) of PPI were used in
order to identify the oil blend which would produce the most acceptable
frozen dessert. This oil blend was subsequently used in the second
stage of the study in conjunction with four PPI levels ß.5%,5"0%r 6.5%
and 8.0%) in order to determine the optimum oil blend/PPI level combina-
tion that would result in a frozen desserL with satisfactory accept-
ability. The major findings of this study are summarized in the

following paragraphs.

flavour of the 10% PPI slurries was found to be unacceptable,

The
contrary to the manufacturer's claim. DSC analysis indicated that the
PPI was severely denatured, most likely the result of harsh processing
condiLions. ÀIthough the functionality of the PPi may have diminished
due to the denaturation, the results of the alcohol coagulation test
showed that the pea proteins were able to remain stable in the frozen

mixes. Increases in PPi levels in the mixes were matched by rising

total solids content and mix viscosity, and by increasing resistance to

-123-
 124

melting in the frozen desserts. The frozen desserts containing 6.5% and

8,0% Pil showed a melting defect (does-not-melt ).

The margarine oil blends all showed lower SFI values than butterfat.
Mix viscosity was not affected by the composition of the oil blends.

Overrun, on the contraryr. increased with increasing proportions of hard

margarine oit in the oil blends. No difference in the melting behaviour

among the frozen desserts containing various oil blends rvas observed.

In the first stage of this study, it was found that at a common leve1
of PPi (5.0%), the oil blend HS 64 was able to produce a frozen dessert
with the highest sensory score in comparison to the others. The oil
blends HS 55 and HS 73 did not differ significantly from HS 64, HS 73
and HS 46 in infJ-uencing the acceptability of the frozen desserts.
However, f rozen desserts containing the oi1 blends HS 37 and HS 46 r,¡ere
significantly less preferred to the one prepared ¡"ith HS 64. The SFI
values of the oil blends which were able to produce a satisfactory
frozen dessert ranged f.ron 22.07 to 25.50 at 100C.

Results fron the second that the PPi level of 3.5%

stage showed

produced a frozen dessert with the highest score as compared to the

others. The use of PPI levles of 6.5% and 8.0% resulted in signifi-
cantly less acceptable products. The maximum level of PPi which could
be used to manufacture a satisfactory frozen dessert was found to be
5.0%.

In conclusionr âD oil blend composed of 60% (wlw) of hard margarine

oil and a0% (.wl.w) of soft margarine oiI, used at a level of 10.50% in

conjunction with 3.5% PPI, was found to be the optimum combination in

producing a non-dairy frozen dessert wilh an acceptable quality. It was

 125

not feasible to incorporate high levels of PPI into the frozen dessert
mix in order to raise the nutritional value of the finished product

rvhiLe maintaining its acceptability.

Due to the exploratory nature of this research, iLs scope has been
restricted to identifying the oil blend/ppl level combination at which a
generally satisfactory non-dairy frozen'dessert could be produced.
Recommendations for future research include:

f. investigation of the possibility of improving the acceptability

and functionality of PPI.

2. detail study of the characteristics of the frozen desserts.

3. feasibility studies of the development of other dairy-like prod-
ucts utilizing PPI and margarine oil blends"
4. feasibility studies of utilizing PPi and margarine oil blends in
other types of food system.
 REFERENCES

ÀCKMÀN, R. G. 1983. Chemical composition of rapeseed oil. in: High and

Low Erucic Àcid Rapeseed 0i1s. Production, Usage, Chemistry, and
Toxiological Evaluation. J"K.G. Kramer, F.Ð. Sauer and l^1.J. Pigden
(eds. ). p.85-129. Academic Press, Toronto.

AGRICULTURE CÀNÀDA. 1986. Handbook of Selected Agricultural statistics.

Policy Branch, Àgriculture Canada, 0ttawa.
ALLEN, R. R. 1982. Hydrogenation. in: Bailey's lndustrial 0il and Fat
Products. Vol. 2.4th ed. D. Swern (ed.). p. 1-95. John Wiley and
Sons, New York, N. Y.
ANDREÀSEN, T. G. 1985. Sorbet and sherbet with Fructodan. Nordeuropaeisk
Mejeri-tidsskr. 51 ( 5) :125-129.
ANONYMOUS. 1981. Canola oil - a nutritional and commercial profile. J.
Am. Oil Chem. Soc. 58 (9):'127A-728A,
AOÀC. 1984. 0fficial Methods of Ànalysis. 14th ed. Association of
0fficial Ànalytical Chemists, Washington, D.C.
AOCS. 1974.0fficiaI
Method Cd 10-57. Solid Fat Index. Àmerican 0i1
ChemisLs' Society, Champaign, I11.
APHA. 1978. Standard Methodsfor the Examination of Dairy Products. 14th
ed. E.H. Marth (ed.). Àmerican Public Health Association, Washington,
D. C.

ARBUCKLE, W. S. 1969. How to make parevine. Am. Dairy Rev.

31(8) :19,22,74.

ARBUCKLE, I,T. S. 1986. Ice Cream. 4th ed. Àvi Publishing Co., I,lestport,
Conn.

ARNTIFIELD, S. D. and E. D. 1981. The influence of processing

MURRÀY,
parameters on food protein functionality. I. Differential scanning
calorimetry as an indicator of protein denaturation. Can. Inst. Food
Sci. Technol. J . 14(a) :289-29a.
ATHERTON, H. V. and NEWLANDER, J. A. 1977. Chemistry and Testing of
Dairy Products. 4th ed. AVI Publ. Co. Inc., Westport, Conn.
BERGER, K. G., BULLIMORE, B. K., I,IRIGHT, Ì,1. B. 1972a. The structure of
ice cream - Part 1. Dairy Inds. 37(8) i419-425"
BERGER, K. G., BULLIMORE, B. K., WHITE, G. W. and WRIGHT, W. B. 1972b.
The structure of ice cream - Part 2. Dairy Inds. 37(9) i493-497.
-126-
 127

BERGER,K. G. and I,iHITE, G. W. 1971. An electron microscopical

investigation of fat destabilization in ice cream, J. Food TechnoI.
6(3):285-294.
BERGER,K. G. and WHiTE, G. W. 1976. The fat globule membrane in ice
cream - Part I. Dairy Ind. InL. 41 (6) :199-202,212.

BERGER, K. G. and l^lHITE, G. W. 1979. Ice cream. in: Food Microscopy. J.

G. Vaughan (ed.). p. 499-530. Academic Press, London.

BHÀTTY,R. S. 1982. In vitro hydrolysis of skim milk and pea proteins by

pepsin and renin. Can. Inst. Food Sci. Technol. J. 15(2):101-'108.
BILiÀDERIS, C. G. 1983. Differential scanning calorimetry in food
research - a review. Food Chem. 10(4):239-265.
BIRKER, P. J. M. W. L. and PÀDLEY, F. B. 1987. Physical properties of
fats and oils. in: Recent Àdvances in Chemistry and Technology of
Fats and Oils. R. J. Hamilton and A. Bhati (eds.). p. 1-11. Elsevier
Applied Science Publ. Ltd., New York, N. Y.

BORGSTROM, G. 1976. Principles of Food Science. Vol.1. Food Technology.

Food and Nutrition Press, Inc., Westport, Conn.
BORST, J. R. 1978. Whey proteins. in: Encyclopedia of Food Science. M.
S. Peterson and À. H. Johnson (eds.). p. 875-877. AVi Publ. Co. Inc.,
Westport, Conn.

BOULTER, D. 1983. Regulation of storage protein synthesis and deposition

in developing pea seeds. in: Perspectives for Peas and Lupins as
Protein Crops. R. Thompson and R. Casey (eds.). p. 221-224. Martinus
Nijhoff Publishers, The Hague.
BOULTER, G. S. 1983. The history and marketing of rapeseed oil in
Canada. in: High and tow Erucic Acid rapeseed 0ils. Production,
Usage, Chemistry, and Toxicological Evaluation. J. K. G. Kramer, F
D. Sauer and l,t. J. Pigden (eds. ). p. 61-83. Àcademic Press, Toronto.
BRENNÀN, J. G. 1974. Mechanical emulsification procedures. in:
Encyclopedia of Food Technology. À. H. Johnson and M. S. Peterson
(eds.). p.351-365. AvI Publ. Co. Inc., I.Iestport, Conn.
BRUNNER, J. R. 1978. Casein. in: Encyclopedia of Food Science. M. S.
Peterson and A. H. Johnson (eds.). p. 126-136. AVI Publ. Co. Inc.,
Westport, Conn.

CHANDRASEKHARA, M. R., RÀMÀNNÀ, B. R., JÀGÀNNATH, K. S. and RAMÀNATHAN,

P. K. 1971. Miltone vegetable toned milk. T. Food Technol. 25232-34.
COTTRELL, J. I. L., PASS, G. and PHILLIPS, G" 0. 1980. The effect of
stabilisers on the viscosity of an ice cream mix. J. Sci. Food Agric.
31:1066-1070.

CHRISTIANSEN, A. J. 1985. Emulsifiers in food applications. Food

Technot. N. z. 20 (1 ) :13-.15,17 ,
 128

DÀLGLEISH, D, G. 1982. Milk proteins - chemistry and physics. in: Food

Proteins. P. F. Fox and J. J. Condon (eds.). p" 155-178. Applied
Science Publishers, London.
DAUN, J. K. 1983. Introduction of LEAR varieties in Canada. in: High
and Low Erucic Àcid Rapeseed 0i1s. Production, Usage, Chemistry, and
Toxicological Evaluation. J. K. G. Kramer, F. D. Sauer and W. J.
Pigden (eds. ). p. 161-180. Academic Press, Toronto.
DELÀQUIS, P. J. 1983" Physical, Chemical, Sensory and Microbiological
Properties of Pork Sausage extended with Pea Protein Isolates. M. Sc.
Thesis. University of Manitoba, winnipeg, MB.
DICKINSON, E. and STÀINSBY, G.1982. Colloids in Foods. Àpplied Science
Publ. Ltd., London.
R. K. 1983. The origin and description of the Brassica oilseed
DOI,¡NEY,
crops. in: High and tow Erucic Acid Rapeseed Oils. Production,
Usage, Chemistry, and Toxicological Evaluation. J. K. G. Kramer, F.
D. Sauer and W. J. Pigden (eds.). p. 1-20. Àcademic Press, Toronto.
EI-DEEB, S.4., SALÀM, A. E. and MASHÀLY, R. i. 1983. Use of
hydrogenated oils in ice cream making. Alex. Sci. Exch. 4(1 ):115-123.
EL-DEEB, S. À. and SÀLÀM, A. E. 1984. The use of flour vegetable
proteins in ice cream manufacturing. Àlex. Sci. Exch. 5(1):87-'100.
EMERSON, and JOHNSON, J. L. 1985. Àdverse reactions
J. L. to foods - an
overvi ew. Food Technol . Àust . 37 ( '1 1 ) :496-502 ,505.
EVANS, E. W. 1986. Interactions of milk components in food systems. in:
Interactions of Food Components. G. G. Birch and M. G. tindley
(eds.). p. 241-275. Elsevier Applied Science Publishers Ltd.,
Barking, Essex, U. K.
FISHER, L. R. and PÀRKER, N. S. 1985. How do food emulsion stabilizers
work ? cSI Ro Food Res. Q. 45Q) :33-39.
FLÀCK, E. A. 1983a. Stabilizing and emulsifying agents. Papers given at
a Two-day Symposium held at the Scientific Societies Lecture Theatre,
Savile Row, London W1 on'19 and 20 October, 1981. p. 18-29. Society
of Dairy Technology, Huntingdon, Canbs.
FLÀCK, E. A. 1983b. The use of emusifiers to modify the texture of
processed foods. Food Flavour. Ingredients Process. Packag.
5(8) : 32-33,35,37,38-39.
FOUTS, E. L. and FREEMÀN, T. R.'1948. Dairy Manufacturing Processes.
John Wiley and Sons, Inc., New York.

FOX, P. F. and MULVIHILL, D. M. 1982. Milk proteins: molecular,

colloidal and functional properties. J. Dairy Res. 49( a):679-693,
 129

FRAM, H. 1974. Kosher food. in: Encyclopedia of Food Technology. A. H.

Johnson and M. S. Peterson (eds.)" p. 535-537. AVI Publ. Co. Inc.,
I,lestport, Conn.

FREDE, E., D. and TIMMEN, H. 1985. Melting and solidification

PRECHT,
behaviour and varying compositions. in: Milkfat and it
of milk fat
Modifications. Contributions at a Lipidforum Symposium, Goteborg,
January 26-27, '1984. R. Marcuse (ed.). p. 32-42. SIK, Gothenburg,
Sweden.

GABRIEL, K., RAMA RAO, M. and RANGÀNADHAM, M. 1986. Ice cream made by
incorporation of different Ievels of groundnut protein isolate.
Indian J. dairy Sci. 39(4) :492-495.
GIBBONS, J. D. 197 '1. Nonparametric Statistical Inference. McGraw-Hi11
Book Company, New York.

GTICKSMÀN, M. 1982. Functional properties of hydrocolloids. in: Food

hydrocolloids. Vol. 1. M. Glicksman (ed.). p. 47-99. CRC Press, Boca
Baton, Florida.
GLICKSMÀN,M. 1984. Role of hydrocolloids in food Processing. in: Gums
and Stabilisers for the Food Industry. 2. Applicationsof
Hydrocolloids. G.0. Phillips, D. J. Wedlock and P. À. WiIliams
(eds.). p. 297-320. Pergamon Press, Oxford, U. K.

GLICKSMÀN, M. 1985. Hydrocolloids in fabricated foods. Food Technol. N.

z. 20(1 0) :75,79-82,85.

GCFF,D. and PEÀRSON, A. M. 1983. Aspartame and corn syrup solids as

sweeteners for ice cream - Part 1. Modern Dairy 62(3):'1'1-'12,14-15r24"
GRAHAM,H. D. 1978. ColloidaI dispersions: polysaccharide gums. in:
Encyclopedia of Food Science. M. S. Peterson and À. H. Johnson
(eds.). p. 161-174. ÀVI Publ. Co. Inc., Westport, Conn.
GRANT, D. R. 1983. Pea protein used as a protein fortifier for cereal
products: chemistry and nutrition. in: Developments in Food Science.
Vol. 58. Progress in Cereal Chemistry and Technology. Proceedings of
the 7th World cereal and Bread Congress, Praguer 1982. J. Holas and
J. Kratochvie (eds. ). p. '1065-'1068. Elsevier, Amsterdam.
GRANT, D. R. SUMNER, A. K. and JOHNSON, J. 1976. An investigation of pea
seed albumins. Can. I nst. Food Sc i . Technol . J. 9Q) :84-91 .

GRAVLUND, K.
1984. Homogenization. Nordeuropaeisk Mejeri-Tidsskr.
50(9):212-215.
GREGORY,D. J. H. 1982. Emulsifiers - their influence on the production,
microstructure and quality of toffees, chocolate, coatings and ice
cream. Part'1. Confectionery Production 48(9) :366,368-369.
 130

HALL, C. W., F'ARRALL, A. W. and RIPPEN, À. L. 1986. Ice cream f reezers.

in: Encyclopedia of Food Engineering.2nd ed. C. I^l. Ha11, A. W.,
FarraIl and À. L. Rippen (eds.). p. 404-408. AVI Publ. Co. Inc.,
Westport, Conn.

HAMILTON, M. P. 1983. Use of milk and miJ-k products in ice cream. in:
Ice Cream. Papers given at a Two-day Symposium held at the Scientific
Societies Lecture Theatre, Savile Row, London w1 on 19 and 20
October, '1981. p. 9-1'7. Society of Dairy Technology, Huntingdon,
Cambs.

HARPER, E. K. and SHOEMAKER, C. F. 1983. Effect of locust bean gum and

selected sweetening agents on ice recrystallization rates. J. Food
Sci. 48(6) :1801-1803, 1806.

HENDERSON, J. L. 197 1. Composition, physical and chemical properties of

milk. in: The FIuid-Milk Industry. 3rd ed. p. 18-48. AvI Publ. Co.
I nc . , Westport , Conn .

HOBBS, L. 1986. Corn syrups. Cereal Foods WorId 31(12):852,854,856,858.

HOYNÀK, P. X. 1974. Sugars and sweeteners. in: Encyclopedia of Food

Technology. À. H. Johnoson adn M. S. Peterson (eds. ). p. 867-873. ÀVI
Publ. Co. Inc., Westport, Conn.

iGOE, R. S. 1982. Hydrocolloid interactions useful in food systems. Food

Technol. 36( 4):72-74.
iNGIETT, G. E. '1981. Sweeteners - a review. Food Technol.
35(3):37-38,40-41.
IVERSEN, E. K. 1983. Scoopable ice cream. Nordeuropaeisk Mejeri-Tidsskr"
49(5):116-122.
JONÀS, J. J. 1975. impact of vegetable proteins on dairy products. J.
Milk Food Technol. 38 ( 1 ) :39-43.
JULIEN, J. P. 1985. Ice Cream. in: Dairy science and Technology. La
Fondation de Technologie Laitiere du Quebec' Inc. p. 315-335. Les
Presses de I'Universite Laval.
KEENEY, P. G, 1982. Development of frozen emulsions. Food Technol.
36(11):65-70.
KENDAIL, M. G. 1970. Rank Correlation Methods. 4th ed. Charles Griffin &

Co., London.
KItÀRÀ, A. and SHARKASI, T. Y. 1986. Effects of temperature on food
proteins and its implications on functional properties. CRC Crit.
Rev. Food Sci. Nutr. 23( ):323-395.
KINSELLA,J. E. '1984. Milk proteins: physicochemical and functional
properties. CRC Crit. Rev. Food Sci. Hutr. 21(3):197-262.
 131

KOLAR,C. W., CHO, I. C. and I,¡ATROUS, W. L. 1979. Vegetable protein

application in yogurt, coffee creamers and whip topping. J. Àm.0i1
Chemists' Soc. 56(3) :389-391.
LARMOND, E. 1982. Laboratory Methods for Sensory Eval-uation of Food.
Publication 1637. Communication Branch, Àgriculture Canada, 0ttawa.
LAUSTEN,K. 1985. Use of vegetable fat based on palm-kernel oil in dairy
products. in: Mitkfat and its Modifications. Contributions at a
Lipidforum Symposium, Goteborg, January 26-27, 1984. R. Marcuse
(ed. ). p. 1 1 1-1 19. SIK Golhenburg, Sweden.

LAUTSEN, K. 1985. Vegetable fats in the dairy industry. Ðairy Ind. Int.
51 (2) :12-13.

LAWHON,J. T., GOLIGHTLY, N. H. and IUSÀS, E. W. 1980. Utilization of

membrane-produced oilseed isolates in soft-serve frozen desserts. J.
Am. Oil Chemists' Soc. 57(9):302-306.

LIM, E. M. 1980. Functional Properties of Milk Proteins with Particular

Reference to Confectionery Products. Scientific and Technical Surveys
No. 120. The British Food Manufacturing Industries Research
Àssociation. Leatherhead, Surrey, England.
LUSAS, E. W. and RHEE, K. C. 1986. Applications of vegetable food
proteins in traditional foods. in: ÀCS Symposium Series 312. PIant
Proteins: Applications, Biological Effects, and Chemistry. R. L.0ry
(ed.). p. 32-44. Àmerican Chemical Society, Washington, D.C.
MADSEN, 1985. The use of vegetable fats in dairy products. in:
J.
Milkfat and its Modifications. Contributions at a tipidforum
Symposium, Goteborg, January 26-27, 1984. R. Marcuse (ed.). p.
120-127. SiK, Gothenburg, Sweden.
MAHDI, S. R. and BRADLEY, Jr., R. L. 1969. Fat destabilization in frozen
desserts containing Iow dextrose equivalent corn sweeteners. J. Dairy
Sci. 52(11) :1738-1741.

McMAHON, D. J. and BROWN, R. J. 1984. Composition, structure, and

integrity of casein micelles: a review. J. Ðairy Sci. 67(3) i499-512.
McPHERSON, A. V. and KiTCHEN, B. J. 1983. Reviews of the progress of
dairy science: the bovine milk fat globule membrane - its formation,
composition, structure and behaviour in milk and dairy products. J"
Dairy Res. 50(1 ):107-133.
MILK INDUSTRY FOUNDATION. 1959. Laboratory Manual. Methods of Ànalysis
of Milk and its Products. p. 449-450. Milk Industry Foundation,
Washington, D. C.
MITTEN,H. L. and NEIRINCKX, J. M. 1986. Developments in frozen
products. in: Modern Dairy Technology. VoI. 2. R. K. Robinson (ed.).
p. 215-259. Elsevier Applied Science Publishers, London.
 132

MODtER, H. I.t. 1985. Functional properties of nonfat dairy ingredients -

a review. Modification of products containing casein. J. Dairy Sci.
68 ( 9 ) :2195-2205 .

MOHSENIN, N. N.1978. Rheology. in: Encyclopedia of food Science. M. S.

Peterson and A. H. Johnson (eds.). p. 663-671. ÀvI Publ. Co. Inc.,
l.iestport, Conn.
MONZINI, À. and MALTINI, E. 1983. Structural- modification by freezing in
Iiquid foods. in: Progress in Food Engineering. C. Cantarelli and C.
Peri (eds.). p. 435-469. Forster Publ. Ltd., Kusnacht, Switzerland.
MORR, C. V. 1979. Technical problems and opportunities in using
vegetable proteins in dairy products. J. Am.0il Chemists'Soc.
56(3):383-385.
MORR, C. V. 1982. FunctionaL properties of milk proteins and their use
as food ingredients. in: Developments in Dairy Chemistry. VoI" 1. P.
F. Fox (ed.). p. 375-399. Àpplied Science Publishers, London.
MORR, C. V. 1985. Functionality of heated milk proteins in dairy and
related foods. J. Dairy. Sci. 68(10):2773-2781.
MORRIS, E. R. 1984. Rheology of hydrocolloids. in: Gums and Stabilisers
for the Food Industry.2. Àpplications of Hydrocolloids. G.0.
Phillips, D. J. Wedlock and P. A. Williams (eds.). p. 57-78. Pergamon
Press,0xford, U. K.

MORTENSEN, B. K. 1983. Physical properties and modification of milkfat.

in: Developments in Dairy Science. Vol.2. tipids. P. F. Fox (ed.).
p. 159-194. Applied Science Publishers Ltd., London.
MOSSE, J. and PERNOLLET, J. C. 1983. Storage proteins. in: Chemistry
and Biochemistry of Legumes. S. K. Àrora (ed. ). p. '111-'193. 0xf ord &
IBH Publ. Co., New Delhi.
MUNRO, D. S. and ILLINGWORTH, D. 1986. Milkfat-based food ingredienLs:
present and poLentiat products. Food Technol. Aust. 38(8):335-337.
MURRAY, E.0., MYERS, C. D. and BARKER, L. D,1978. Protein product and
process for preparing same. Canadian Patent 1
'028 '552.
NÀTIONÀL DÀIRY COUNCIt. 1983. imitation and substitute dairy products.
Dairy Council Digest 54(1).
NEWSOME,R. L. 1986. Sweeteners: nutritive and non-nutritive. Food
Technol. 40(8) : 1 95-206.
NICKEL, G. B. 1981. Process for preparing products from legumes.
Canadian Patent 1 ,1 04 ,871 .

NIELSEN, B. J. 1978. Combined emulsifer/stabilizers for ice cream.

Gordian 78(6):176.
 133

0'MAHONY, M. 1986. Sensory Evaluation of Food. Marcel Dekker, Inc., New

York, N" Y.
PÀTE, J. S" 1971, The pea as a crop p1ant. in: The Physiology of the
Garden Pea. J. F. Sutcliffe and J. S. Pate (eds.). p. 469-484.
Àcademic Press, London.

POMERANZ, Y. 1985. Functional Properties of Food Components. Academic

Press, London.
PORTER, J. I,t. G. 1980. The role of milk and milk constituents in the
human diet. Int. Dai.ry Fed. Annu. 8u11. 125:14-21.
POTTER, N. N. 1980. Milk and milk products. in: Food Science. 3rd ed.
p. 374-421. ÀVI Publ. Co., Westport, Conn.
PRiTCHETT, W. 1974. Margarine. in: Encyclopedia of Food Technology. A.
H. Johnson and M. S. Peterson (eds.). p. 574-580. ÀVi Publ. Co.,
Westport, Conn.

RAHMAN, A. R. 1974. Fabricated foods. in: Encyclopedia of Food

Technology. À. H. Johnson and M. S. Peterson (eds.). p. 388-389. AVI
Publ. Co. Inc., I.iestport, Conn.

RAJAH, K. K. 1986. Milkfat products. J. Soc. Dairy Technol.

39(4):104-109.
RÀJOR, R. B. and GUPTA, S. K. 1982. Soft-serve ice cream from soybean
and buttermilk. i. Method of manufacture. Indian J. Dairy Sci.
35(4):454-459.
RAJOR, R.8., GUPTÀ, S. K", PÀTEL, À. À. and PATIL, G. R. 1983. À new
'softy' from buttermilk and soybean. Indian Dairyman 35( 1 ) :29-32.
REES, L. H. 19'14. Homogenizer and colloid milIs. in: Encyclopedia of
Food Technology. A. H. Johnson and M. S. Peterson (eds.). p. 499-501.
ÀVI Publ . Co. I nc . , I,lestport , Conn .
REES, L. H. and PÀNDOLFE, W. D. 1986. Homogenizers. in: Encyclopedia of
Food Engineering. 2nd ed. C. W. Hall, À. W., Farrall and À. t.
Rippens (eds.). p.467-472. AVI Publ. Co. Inc., Westport, Conn.
REUTER, H. 1978. Homogenization. in: Encyclopedia of Food Science. M.
S. Peterson and À. H. Johnson (eds.). p. 374-376. ÀVi Publ. Co. Inc.,
Westport, Conn.

RIEL, R. '1985. Composition and physicochemical structure of mi1k" in:

Dairy Science and Technology. La Fondation de Technologie Laitiere du
Quebec, Inc. p. 1-51. Les Presse de I'Universite taval.
ROSSELL, J. B. 1986. CIassicaI analysis of oils and fat. in: Ànalysis
of Oils and Fats. R. J. Hamilton and J. B. Rossell (eds.). p. 1-90.
Elsevier Applied Science Publishers, London.
 134

ROTHI^IELL, J. 1984. Uses for dairy ingredients in ice cream and other
frozen desserts. J. Soc. Dairy Technol" 37(4\:119-121"
SÀCHAROW, 1974. Taboos and food prejudices.
S. in: Encyclopedia of food
TechnoJ-ogy. À. H. Johnson and M. S. Peterson (eds.). p. 884-887. AVI
Publ. Co. Inc., Westport, Conn.

SÀNDINE, W. E. and DALY, M. 1979, Milk intolerance. J. F'ood Prot.

42$):435-437 .
SAVÀIÀNO, D. A. and LEVITT, M.D. 1987. Milk intolerance and microbe-
containing dairy foods. J. Dairy Sci. 70Q):397-406.
SCHMIDT, D. G.'1980. Colloidal aspects of casein. Neth. Milk Dairy J.
34:42-64.
SCHMIDT, R. H. and MORRIS, H. À. '1984. Gelation properties of milk
proteins, soy proteins, and blended protein systems. Food Technol.
38 ( 5 ) :85-88 ,90 ,92-94 ,96.

SCHNEIDER, M. 1986. Emulsifying agents. Confectionery Manufacture and

Marketing 23( ) : 1 1,13-1 4,16.
SHARMA,S. C. 1981. Gums and hydrocolloids in oil-water emulsions. Food
Technol. 35(1):59-67.
SHRUTLEFF, W. and AOYAGI, À. 1984. Dairylike products made from soymilk.
in: Tofu and Soymilk Production: The Book of Tofu. voI. II. 2nd ed.
p. 235-255. The Soyfoods Center, Lafayette, California.
SHURTLEFF, W. and ÀOYAGI, A. 1985. Toffutti and other Soy Ice Cream.
Vol. I and II. The Soyfoods Center, Lafayette, California.
SIMMONS, R. G., GREEN, J. R., PAYNE, C.4", WAN, P. J. and LUCÀS, E. W.
1980. Cottonseed and soy protein ingredients in soft-serve frozen
desserts. J. Food Sc i . 45 (6 ) : 'l 505-1 508.
SIMOONS, F. J.1982. Problems in the use of animal products as human
food: some ethnographical and historical problems. in: ÀnimaI
Products in Human Nutrition. D. C. Beitz and R. G. Hansen (eds.)" p.
19-34. Àcademic Press, New York.
SLEETER, R. T. 1985. InsLrumental analysis for quality control and
quality assurance. in: Bailey's Industrial 0i1 and Fa! Products.
Vol.3. T. H. Àpplewhite (ed.). p. 167-242. John Wiley and Sons, New
York.

SOSUISKI, F. W., CHÀKRABORTY, P. and HUMBERT, E. S. 1978. Legume-based

imitation and blended milk products. Can. Inst. Food Sci. Technol. J.
11(3):117-123.
 '135

SOSUtSKI, F. W. and SOSULSKI, K. 1986. Composition and functionality of

protein, starch, and fiber from wet and dry processing of grain
legumes. in: ACS Symposium Series 312. Plant Proteins: Applications,
niõlogical nffects, and Chemistry. R. L. Ory (ed.). p. 176-190.
Àmerican Chemical Society, Washington, D.C.

Statistics Canada. 1989. Oils and fats. 40(7).

STEINKRAUS, K. H. 1978. imitation milks. in: Encyclopedia of Food
Science. M. S. Peterson and n. H. Johnson (eds.). p.405-408. AVI
Publ" Co. 1nc., Westporl, Conn.
STONE, H. and SIDEL., J. L. 1985. Sensory Evaluation Practices. Academic
Press, New York, N. Y.

SUMNER, A. K., NEILSEN, M. A. and YOUNGS' C. G. 198'1 . Production and

evaluation of pea protein isolate. J. Food Sci. 46Q):364-366,372"
SWAMINÀTHÀN, M. and PÀRPIA, H. A. B. 1967. Milk substitutes based on
oilseeds and nuts. World Rev. Nutr. Diet.8:'184-206.
M. and WONG, C. 1985. Milk proteins:
SI,IARTZ, nutritional and functional
uses. Cereal Foods World 30(2) :173-175.
TÀYLOR, S. L. and CUMMING, D. B. 1985. Food allergies and sensitivities.
Food Technol. 39(9):65-7'1 .

TAYIOR, S. L.'1986. Immunologic and allergic properties of cow's milk

proteins in humans. J. Food Prot. 49(3):239-25t,
TEÀSDAIE, B. F. and MAG, T. K. 1983. The commercial processing of low
and high erucic acid rapeseed oils. in: High and tow Erucic Àcid
Rapeseed Oils. Production,
Usage, Chemistry, and Toxicological
Eváluation. J. K. G. Kramer, F. D. Sauer and W. J. Pigden (eds.)' p.
197-229. Academic Press, Toronto.

THOMÀS, E. L. 198'1. Structure and properties of ice cream emulsions.

Food Technol. 35 ( 1 ) i441 r48.
'44,46
THOMSON, W. A. B. 1979. Infant formulas and the use of vegetable
protein. J. Am. Oil Chemists' Soc. 55(3):386-388.
TIPSON, R. S. 1956. Technique of Organic Chemistry. Part I. Vol.3.2nd
ed., p.395. Wiley, N. Y. Quoted in Harper, E. K. and Shoemaker, C. F.
1983. Effect of locust bean gum and selected sweetening agents on ice
recrystallization rates. J. Food Sci. 48(6) :1801-1803,1806.
VOUTSINAS, L. P., CHEUNG, E. and NÀKAI, S. 1983. Relationships of
hydrophobicity to emulsifying properties of heat denatured proteins.
J. Food Sci. 48(1):26-32"
I^¡ADDINGTON, D. '1986.
Application of wide-line NMR in the oils and fats
industry. in: Analysis of 0iIs and Fats. R.J. Hamilton and J. B.
Rossetl (eds. ). p. 341-399. Elsevier Applied Science Publishers,
London.
 136

WAIKER,R. C. and BOSIN, W" 4.1971. Comparison of SFI, DSC and NMR
methodsfor determining solid-liquid ratios in fats. J. Àm.0i1
Chemists' Soc. 48/2):50-53.
}lEISS, T. J. 1983. Food 0i1s and their Uses. 2nd ed. AVI Publ. Co. Inc.,
Westport, Conn.

wHITE, G. W. 1981. Homogenisation of ice cream mixes. Dairy Ind. Int.

46 Q) :29,3'1 -33, 35-35.

D. 1979. Vegetable protein application in whey soy drink

I^¡iLDiNG, M. mix
and ice cream. J. Am. Oil Chemists' Soc. 56(3):392-396.
WILLS, R. B. H., EVANS, T. J., LIM, J. S. K., SCRIVEN, F. M. and
GREENFIELD, H. 1984. Composition of Australian foods. 25. Peas and
beans. Food Technol. Aust. 36(1 1 ):512:514"
ttRIGHT,D. J. 1982. Àpplication of scanning calorimetry to the study of
protein behaviour in foods. in: Developments in Food Proteins. Vol.
1. B. J. F. Hudson (ed.). p.61-89. Àpplied Science Publishers,
London.

YOUSSEF, À. M., SÀLAM, A. E. and SALÀMÀ, F. A. 1981. Manufacture of a

high quality ice cream. ii. Substitution of fresh creamfat by butLer,
cottonseed and corn oil-s. Egyptian J. Dairy Sci. 9(1):51-58.
 Appendix A

FLAVOUR EVALUATION DATA OF THE PPi SLURRIES.

=== === === == == = = == = = = === = = == = == = = = = == == = == = = = = = == == = = = = = === = ==== === ===== =

Sc ore

Samples (Grade)

Judges Baking Gold RepIac ement Total

1 5 6 3 14
2 3 2 4 9
3 6 6 4 16
4 4 5 2 11
tr
2 3 3 I
'10
6 5 2 3
7 2 2 3 7
I 2 2 3 7
9 5 6 4 15
10 1 3 4 I
1'1 4 3 3 10
12 6 T 17
12 6 4 5 15
tr
14 3 4 12
15 2 4 3 9
,1

16 2 1 4

Tota I 59 59 54 172

Mean Score 3.69 3.59 3.38

Table of Analysis of Variance

== == == ===== === == == === == = = ==== = == = == = = = = == == == == = = = = ======= = = = = == = === === =
Source of Variation df SS MS

SampLes 2 1 .05 0.53 0. 14 ns1

Judges 15 70.34 4.69 1 .28 ns
Er ror 30 1 09.67 3.66

Tota 1 47 181.09

F(0.05, 2, 30) = 5.3¿

F(0.05, 15, 30) = 2.01
1 non-significant at o = 0.05

- t3t -
 Àppendix B

SENSORY EVALUATION DATA OF THE NON_DAIRY FROZEN DESSERTS

MADE WITH F]VE MARGARINE OIL BLENDS.

== = ==== = === === == = = === ====== == === == ==== == = == = ====== === =i = = == = = == = = === = = ==
Score

Samples

Judges HS 37 HS 46 HS 55 HS 64 HS 73 Tota l
1 JJ
2 40
J 20
+ 28
5 27
6 30
,]
28
I 31
9 32
0 32
1 40
2 40
35
4 42
q 28
6 32
7 24
I 38
9 21
20 24
21 24
22 26
23 35
24 40
25 28
26 28
27 26
28 19
29 27
30 27
31 21
32 27
33 19
34 32
35 38
36 25
37 25

- 138 -
 139

38 7 T 6 I
,7
I 33
1 30
39 6 6 4
40 ð 7 I I I 39
41 * 6 6 6 6 28
42 5 7 7 7 I 34
43 4 4 c.
7 7 28
44 5 7 I 6 4 30
45 6 4 7 3 27
46 7 2 6 6 3 24
47 5 7 6 4 3 25
48 6 5 7 6 29
49 T 4 4 7 6 25
50 7 7 7 6 6 33
51 6 5 5 7 6 29
52 ö 6 9 7 T 34
53 4 4 7 ¿ 6 23

Tota I 306 286 315 345 311 1 553

Mean Score 5.77 5.40 5.94 6"51 5.87

Table of Ànalysis of Variance.

==================================================== ====================
Source of Variation df SSMSF
Samples 4 .14
34 8.s3 4. 99't 1

Judges 52 353.85 6.80 3.98*

Error 208 356.26 1
"71

Tota I 264 7 44.25

F(0.05' 4, 208) = 2,41

F(0"05 , 52, 208) = 1,42
'1 significant at s = 0.05
't
 Àppendix C

RANKED SENSORY EVÀLUATION DATA OF NON-DÀiRY FROZEN

DESSERTS MADE WITH FIVE MÀRGARINE OIL BLENDS AND THE
CALCULATION OF THE COEFFICIENT OF CONCORDANCE.

=================== =========================================== ==== ======

Rank

Samples

Judges HS 37 HS 46 HS 55 HS 64 HS 73 NTl ¿z

1 3.5 1.0 3.5 5.0 2.0 '1

2
2 3.0 3.0 3.0 3.0 3.0 1 5
4.0
,1
? 2"0 2.0 2.0 5.0 J
4 2"0 1.0 3.5 5.0 3.5 I 2
5
'1.5 3.5 '1
.5 5.0 3.s 2 212
6 2.5 1.0 ÀÈ 2.5 4.5 2 212
7 4.0 t.þ '1
.5 4.0 4.0 2 213
I 2"5 2.5 1.0 5.0 ¿n 1 2
9 2"5 2.5 1.0 4"0 5.0 1 2
0 1.0 3.5 3.5 5.0 2.0 1 2
,1
3.0 3.0 3.0 3.0 3.0 1 5
2 3.0 3.0 3.0 3.0 3.0 1 5
3 3.0 1.0 3.0 5.0 3.0 1 J
14 i.q 4.5 2,0 2.0 2.0 2 213
15 3.s 2"0 3.5 1.0 5.0 I 2
16 1.0 3.5 2.0 3.5 5.0 1 2
1'l 4.0 '1
.5 4.0 4.0 1.5 2 213
'18
1.5 4.0 ¿n 4.0 '1.5 2 213
19 3"0 1.0 3.0 3.0 5.0 1 3
20 4"5 2"0 2.0 2.0 4.5 2 213
21 4.5 2.0 4.5 1.0 3.0 2
22 2.0 1.0 5.0 3.5 3.5 2
23 5.0 3.0 1.0 3.0 3.0 J
,)tr 4.5
24 2.5 '1
.0 4.5 2 212
25 1.0 2.0 4.0 5.0 3.0 0 U

26 3.0 1.5 ^tr 4.5 t.5 2 212

27 4.5 1.5 ¿q 3.0 t.5 2 212
28 2,0 4.0 2.0 5.0 2.0 1 3
4.0 3.0 5.0 '1.5
29 1.5 1 2
30 4.5 1.0 4.5 3.0 2.0 1 2
31 4.0 5"0 1.5 3.0 1.5 1 2
32 1.5 3.5 5.0 3.5 1.5 2 212
33
'1
.0 5.0 2.0 3.5 3.5 1 2
34 4.5 2.0 2.0 2.0 2.J 2 213
35 1.5 4.0 4.0 4.0 2 213
36 3.0 2.0 1.0 ^É. ¿q 1 2

- 140 -
 141

3.0 1.5 i.q t.þ 2 212

37 ^ç.
38 3.0 1.0 2.0 4.5 ',ltr 1 2
atr 1.0 4.5 4.5 aa
39 2.5 2 LsL
40 3.5 1.0 3.5 3.5 2
41 1.0 3.5 3.5 3.5 3"5 4
42 1.0 3"0 3.0 3.0 5.0 3
43 1tr 1C 3.0 4.5 ¿E 2 212
44 2.0 4.0 5.0 3.0 1.0 U 0
45 3.0 4.5 2.0 4.5 1.0 1 2
46 5.0 1.0 J.J 3.5 2.0 1 2
47 3.0 5.0 4"0 2"0 1.0 0 U
Itr 1.5 5.0 3.5 212
48 3.5 toJ 2
49 2.0 2.0 2"0 5.0 4.0 1 3
50 4.0 4.0 4.0 1.5 1"5 2 213
51 3.5 t"5 1.s 5"0 3.5 2 212
5¿ 4.0 2.0 5.0 3.0 1.0 0 0
53 2.5 2"5 5.0 1n 4.0 1 2

Sum of
130.s 158.5 .5 '163"5
Ranks 3
151.0 191

1 number of tied observations

2 number of ranks in each tie
3 the degree of preference increases rllith increasing Sum of ranks.

Formula for calculating the coefficient of concordance:

n r k ( n + 1 ) r2
r lRi-----
j=1L 2
|
r

kn(n2-1)-EEt(12-1)
12

where I^l' corrected coefficient of concordance

k number of observers
n number of treatments
Ri sum of ranks of each treatment
t number of ranks in each tie

I nterpretat ion :
than the chi-square value at (n -
If I{''k(n - 1) is greater , û), 1

reject the null hypothesis that there tvas no consensus among the judges
regarding the order of preference for the samples.
In stage one, W' 'k(n - 1 ) = 17.91 and the chi-square value at (4,
0.05) = 9.49, the null hypothesis is therefore rejected.
 Appendix D

SENSORY EVÀLUATION DATA OF NON_DAIRY FROZEN DESSERTS

PREPARED ÀT FOUR LEVELS OF PPi.

==================================== ========= ===========================

Score

Samples

Judges P 35 P50 P55 P80 Total

1 16
2 19
28
4 26
21
6 25
7 24
I 30
9 17
0 22
'1
24
2 26
3 28
4 31
5 20
16 21
17 24
18 22
19 28
20 30
'19
21
22 33
¿s 25
24 25
¿5 20
¿6 23
27 18
28 28
29 19
30 21
31 27
32 26
JJ 25
34 18
35 23
36 7
37 13

- 142 -
 143

38 7 7 4 3 21
39 7 4 3 6 20
40 7 6 4 3 20
41 I 7 6 5 ¿b
42 6 6 I I 28
43 7 I 7 9 31
44 7 7 I I 30
45 7 4 4 4 19
46 6 7 7 3 23
47 6 7 5 6 24
48 7 5 4 2 18

Tota 1 319 297 2't5 221 1112

Mean Score 6.65 6.19 5.73 4.60

Table of Ànalysis of Variance.

Source of Variation df SS MS

Samples 3 110.42 36.81 )î ))*1

Judges 47 301.2'7 6 .41 ? q?*
Error 141 2s6. 08 1 "82

'l 667 .77

Total 91

F(0.05, 3, 141) = 2.67

F(0.05,47r 141) = 1.44
1 * significant at a = 0.05
 Appendix E

RANKED SENSORY EVÀLUÀTION DÀTA OF NON-DAIRY TROZEN

DESSERTS PREPARED ÀT FOUR PPI LEVELS AND THE CÀICULATION
OF THE COEFFICIENT OF CONCORDANCE.

- === === == == == == = == = = == == === == ==== ======== == = = == ==== === ==== === ==

Rank

Samples

Judges P 35 P50 P65 P80 NTl t'

1 2.5 2.5 4.0 .0 1 2
2 4.0 3.0 1.5 .5 1 2
? Lf\ t.þ 3.0 .5 1 2
4.0 3.0 2.0 .0 0 n
4
5 4.0 3.0 2.0 .0 0 0
6 2.0 ') Ê.
3.5 .0
,1
2
7 2.0 4.0 3.0 .0 0 0
I 3.5 1.5 '1
.s 3.5 2 2,
9 3.0 4.0 '1
.5 t.þ 1 2
0 2.0 3.5 ?tr 1.0 1 2
1 3.5 ¿.v 3.5 1.0 1 2
2 4.0 2.0 2.0 2"0 '1
3
3 3.5 1.5 1.5 3.5 2 2,
14 4.0 J.U t"þ 1.5 1 2
15 3.0 4.0 2.0 1.0 0 0
16 4.0 1.5 3.0 1.5 1 2
'1 ,1

17 4.0 2.5 2.5 .0 2

18 ¿n 3.0 2.0 '1
.0 0 0
'19 4.0 1.0 2.5 2.5 1 2
20 1.5 3.5 3.5 1.5 2 2,
21 4.0 2.0 3.0 1.0 0 0
22 2.0 2,0 2.0 4.0 1 3
23 4.0 2.5 2.5 .0 1 2
24 3.0 4.0 2.0 .0 0 0
25 4.0 3.0 2.0 .0 0 0
26 1.5 4.0 3.0 .5 1 2
27 2.0 1.0 4.0 3.0 0 0
1.0
,1

28 2.5 4.0 2.5 2

29 1.0 4.0 3.0 2.0 U 0
30 3.0 2.0 4.0 1.0 n 0
3'1 3.0 3.0 3.0 1.0 '1
3

32 3.5 3.5 l¡J 1.5 2 2,

33 4.0 )(\ )n )(\ 1 3
AF
34 4.0 2,5 L.J 1.0 1 2
35 4.0 J.U 2.0 1.0 0 0
36 4.0 2.0 2.0 2,0 1 J

-144-
 145

,1

37 3.5 3.5 1.0 2"0 2

'1
38 3.5 3.5 2.0 .0 1 2
39 4.0 2.0 1.0 3.0 0 0
40 4.0 3.U 2.0 1.0 0 0
41 4.0 3.0 2.0 1.0 0 0
.1
42 1.5 tr 3.5 2 212
43
'1
.5 3.0 t"þ 4.0 1 2
44 1.5 1.5 3.5 3"5 2 212
45 4.0 2.0 ¿.u 2.0 3

46 2.0 3.5 3.s 1.0

,)tr
2
47 2.5 4"0 1.0 2
48 4.0 3.0 2.0 1.0 0

Sum of
Ranks 3
152.0 131 .5 1'15.5 81.0

1 number of tied observations

2 number of ranks in each tie
3 the degree of preference increases with increaSing sum of ranks.

Formula for calculating Lhe coefficient of concordance:

n k(n+1)
1 l=
T
I
[*' l'
vlt = ---,
k kn ( nz - 1) - EE t ( t2 - 1 )
12

where Wt= corrected coefficient of concordance

l--
i(- number of observers
n= number of treatments
Ri= sum of ranks of each treatment
!-
L- number of ranks in each tie

I nterpretat i on :
chi-square value at (n - 1 , 0),
If I^t''k(n - 1) is greater lhan the
reject the nulI hypothesis that there $¡as no consensus among the judges
regarding the order of preference for the samples.
In stage two, W''k(n - 1) = 38.17 and the chi-square value at (3,
0.05) = 7.81 ¡ the null hypothesis is therefore rejected.