Although great care has been taken to provide accurate and current information, neither the
author(s) nor the publisher, nor anyone else associated with this publication, shall be liable for any
loss, damage, or liability directly or indirectly caused or alleged to be caused by this book. The
material contained herein is not intended to provide specific advice or recommendations for any
specific situation.
Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress.
ISBN: 0-8247-4712-7
This book is printed on acid-free paper.
Headquarters
Marcel Dekker, Inc.,
270 Madison Avenue, New York, NY 10016, U.S.A.
tel: 212-696-9000; fax: 212-685-4540
Distribution and Customer Service
Marcel Dekker, Inc.,
Cimarron Road, Monticello, New York 12701, U.S.A.
tel: 800-228-1160; fax: 845-796-1772
Eastern Hemisphere Distribution
Marcel Dekker AG,
Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
tel: 41-61-260-6300; fax: 41-61-260-6333
World Wide Web

The publisher offers discounts on this book when ordered in bulk quantities. For more information,
write to Special Sales/Professional Marketing at the headquarters address above.
Copyright # 2004 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, microfilming, and recording, or by any
information storage and retrieval system, without permission in writing from the publisher.

Current printing (last digit):
10987654321
PRINTED IN THE UNITED STATES OF AMERICA
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
Preface
For centuries, frozen foods have been available to consumers in countries that experience
cold winters. In some areas with severe winters such as Alaska, Russia, and others, foods
are routinely frozen by leaving them outside. Since 1875, with the development of
mechanical ammonia freezing systems, the frozen food industry has grown steadily,
especially in the past two decades.
Frozen foods have the advantages of being very close in taste and quality to fresh
foods as compared with other preserved or processed foods. Frozen foods are popular and
accessible in most developed countries, where refrigerators and freezers are standard home
appliances. Nowadays, frozen foods have become essential items in the retail food
industries, grocery stores, convenience food stores, fast food chains, food services, and
vending machines. This growth is accompanied by the frequent release of new reference
books for the frozen food industry.
Several updated books on freezing preservation of foods or frozen foods have been
available in the past decade, and most of them are excellent books. The science and
technology of food freezing can be viewed from several perspectives:
Food engineering principles. These principles explain such phenomena as heat and
mass transfer, freezing time, convective and conductive processes, and other
processes and principles relevant to understanding the dynamics of freezing.
Food science and technology principles. These principles explain the chemistry and
biology of food components, their interactions during processing, and other
principles relevant to understanding how foods behave before, during, and
after the frozen stage.
Food manufacturing principles. These principles explain how we can start with a raw
ingredient and end with a finished frozen product.
Food commodities, properties and applications. This approach takes an individual

commodity of food (e.g., fruits, vegetables, dairy, muscle foods) and explains
the whole spectrum of factors that involve cooling, refrigeration, freezing, and
thawing unique to that category of food and its properties. Although the
underlying principles are the same, freezing carrots is definitely different from
freezing salmon. These data are a combination of the three principles above
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
and are the basis of our ability to enjoy winter vegetables during summer and
100 flavors of ice cream all year round.
Over the past two decades, books have been published that cover some or all of the
topics above. When it comes to books on frozen foods, it is an endless venture. The reason is
simple: Every month and every year, food scientists, food technologists, and food engineers
witness rapid development in the science and technology of frozen foods. We continually see
new knowledge, new equipment, and new commercial applications emerging.
Based on the above premises of principles and applications, the Handbook of Frozen
Foods uses the following approaches to covering the data:
Principles. Chapters 1 through 8 cover principles applicable to the processing of
frozen foods, such as science, technology, and engineering. Topics include the
physical processes of freezing and frozen storage, texture, color, sensory
attributes, and packaging.
Meat and poultry. Seven chapters (Chapters 9–15) discuss freezing beef and poultry
meat, covering operations, processing, equipment, packaging, and safety.
Seafoods’ Chapters 16 through 21 discuss frozen seafoods, covering principles,
finfish, shellfish, secondary products, HACCP (Hazards Analysis and Critical
Control Points), and product descriptions.
Vegetables. Five chapters (Chapters 22–26) discuss frozen vegetables, covering product
descriptions, quality, tomatoes, French fries, and U.S. grades and standards.
Fruits. Chapters 27 through 29 discuss frozen fruits and fruit products, covering
product descriptions, tropical fruits, and citrus fruits.
Special product categories. Chapters 30, 31, and 32 provide details on some popular
products: frozen desserts, frozen dough, and microwavable frozen foods.

Safety. Chapters 33 through 36 discuss the safety of processing frozen foods covering
basic considerations, sanitation of a frozen food plant, risk analysis in
processing frozen desserts, and U.S. enforcement tools for frozen foods.
This volume is the result of the combined effort of more than 50 contributors from
10 countries with expertise in various aspects of frozen foods, led by an international
editorial team. The book contains eight parts and 36 chapters organized into eight parts.
In sum, the approach for this book is unique and makes it an essential reference on frozen
food for professionals in government, industry, and academia.
We thank all the contributors for sharing their experience in their fields of expertise.
They are the people who made this book possible. We hope you enjoy and benefit from the
fruits of their labor.
We know how hard it is to develop the content of a book. However, we believe that
the production of a professional book of this nature is even more difficult. We thank the
production team at Marcel Dekker, Inc., and express our appreciation to Ms. Theresa
Stockton, coordinator of the entire project.
You are the best judge of the quality of this book.
Y. H. Hui
Paul Cornillon
Isabel Guerrero Legarreta
Miang H. Lim
K. D. Murrell
Wai-Kit Nip
iv Preface
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
Contents
Preface
Contributors
PART I. FREEZING PRINCIPLES
1.FreezingProcesses:PhysicalAspects
Alain Le Bail

2.PrinciplesofFreeze-ConcentrationandFreeze-Drying
J. Welti-Chanes, D. Bermu
´
dez, A. Valdez-Fragoso, H. Mu
´
jica-Paz,
and S. M. Alzamora
3.PrinciplesofFrozenStorage
Genevie
`
ve Blond and Martine Le Meste
4.FrozenFoodPackaging
Kit L. Yam, Hua Zhao, and Christopher C. Lai
PART II. FROZEN FOOD CHARACTERISTICS
5.FrozenFoodComponentsandChemicalReactions
Miang H. Lim, Janet E. McFetridge, and Jens Liesebach
6.FlavorofFrozenFoods
Edith Ponce-Alquicira
7.FoodSensoryAttributes
Patti C. Coggins and Roberto S. Chamul
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
8.TextureinFrozenFoods
William L. Kerr
PART III. FROZEN MEAT AND POULTRY
9.FrozenMuscleFoods:Principles,Quality,andShelfLife
Natalia F. Gonza
´
lez-Me
´
ndez, Jose

´
Felipe Alema
´
n-Escobedo,
Libertad Zamorano-Garcı
´
a, and Juan Pedro Camou-Arriola
10.OperationalProcessesforFrozenRedMeat
M. R. Rosmini, J. A. Pe
´
rez-Alvarez, and J. Ferna
´
ndez-Lo
´
pez
11.FrozenMeat:ProcessingEquipment
Juan Pedro Camou-Arriola, Libertad Zamorano-Garcı
´
a,
Ana Guadalupe Luque-Alcara
´
z, and Natalia F. Gonza
´
lez-Me
´
ndez
12.FrozenMeat:QualityandShelfLife
M. L. Pe
´
rez-Chabela and J. Mateo-Oyagu

¨
e
13.ChemicalandPhysicalAspectsofColorinFrozenMuscle-BasedFoods
J. A. Pe
´
rez-Alvarez, J. Ferna
´
ndez-Lo
´
pez, and M. R. Rosmini
14.FrozenMeat:PackagingandQualityControl
Alfonso Totosaus
15.FrozenPoultry:ProcessFlow,Equipment,Quality,andPackaging
Alma D. Alarcon-Rojo
PART IV. FROZEN SEAFOODS
16.FreezingSeafoodandSeafoodProductsPrinciplesandApplications
Shann-Tzong Jiang and Tung-Ching Lee
17.FreezingFinfish
B. Jamilah
18.FreezingShellfish
Athapol Noomhorm and Punchira Vongsawasdi
19.FreezingSecondarySeafoodProducts
Bonnie Sun Pan and Chau Jen Chow
20.FrozenSeafoodSafetyandHACCP
Hsing-Chen Chen and Philip Cheng-Ming Chang
21.FrozenSeafood:ProductDescriptions
Peggy Stanfield
vi Contents
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
PART V. FROZEN VEGETABLES

22.FrozenVegetables:ProductDescriptions
Peggy Stanfield
23.QualityControlinFrozenVegetables
Domingo Martı
´
nez-Romero, Salvador Castillo, and Daniel Valero
24.Production,Freezing,andStorageofTomatoSaucesandSlices
Sheryl A. Barringer
25.FrozenFrenchFriedPotatoesandQualityAssurance
Y. H. Hui
26.FrozenPeas:StandardandGrade
Peggy Stanfield
PART VI. FROZEN FRUITS AND FRUIT PRODUCTS
27.FrozenFruitsandFruitJuices:ProductDescription
Peggy Stanfield
28.FrozenGuavaandPapayaProducts
Harvey T. Chan, Jr.
29.FrozenCitrusJuices
Louise Wicker
PART VII. FROZEN DESSERTS, FROZEN DOUGH, AND
MICROWAVABLE FROZEN FOODS
30.IceCreamandFrozenDesserts
H. Douglas Goff and Richard W. Hartel
31.EffectofFreezingonDoughIngredients
Marı
´
a Cristina An
˜
o
´

n, Alain Le Bail, and Alberto Edel Leon
32.MicrowavableFrozenFoodorMeals
Kit L. Yam and Christopher C. Lai
PART VIII. FROZEN FOODS SAFETY CONSIDERATIONS
33.SafetyofFrozenFoods
Phil J. Bremer and Stephen C. Ridley
34.FrozenFoodPlants:SafetyandInspection
Y. H. Hui
Contents vii
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
35.FrozenDessertProcessing:Quality,Safety,andRiskAnalysis
Y. H. Hui
36.FrozenFoodsandEnforcementActivities
Peggy Stanfield
Appendix A: FDA Standard for Frozen Vegetables: 21 CFR 158. Definitions:
21 CFR 158.3; FDA Standard for Frozen Vegetables: 21 CFR 158. Frozen Peas:
21CFR158.170
Appendix B: Frozen Dessert Processing: Quality, Safety, and Risk Analysis.
SpecialOperations
viii Contents
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
Contributors
Alma D. Alarcon-Rojo Universidad Auto
´
noma de Chihuahua, Chihuahua, Mexico
Jose
´
Felipe Alema
´
n-Escobedo Centro de Investigacio

´
n en Alimentacio
´
n y Desarrollo,
A. C., Hermosillo, Sonora, Mexico
S. M. Alzamora Universidad de Buenos Aires, Buenos Aires, Argentina
Marı
´
a Cristina An
˜
o
´
n Universidad Nacional de La Plata, La Plata, Argentina
Sheryl A. Barringer Department of Food Science and Technology, The Ohio State
University, Columbus, Ohio, U.S.A.
D. Bermu
´
dez Universidad de las Ame
´
ricas—Puebla, Puebla, Mexico
Genevie
`
ve Blond ENSBANA–Universite
´
de Bourgogne, Dijon, France
Phil J. Bremer Department of Food Science, University of Otago, Dunedin, New
Zealand
Juan Pedro Camou-Arriola Centro de Investigacio
´
n en Alimentacio

´
n y Desarrollo, A.C.,
Hermosillo, Sonora, Mexico
Salvador Castillo Miguel Hernandez University, Orihuela, Spain
Roberto S. Chamul California State University, Los Angeles, Los Angeles, California,
U.S.A.
Harvey T. Chan, Jr. HI Food Technology, Hilo, Hawaii, U.S.A.
Philip Cheng-Ming Chang National Taiwan Ocean University, Keelung, Taiwan
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
Hsing-Chen Chen National Taiwan Ocean University, Keelung, Taiwan
Chau Jen Chow National Kaohsiung Institute of Marine Technology, Kaohsiung,
Taiwan
Patti C. Coggins Department of Food Science and Technology, Mississippi State
University, Mississippi State, Mississippi, U.S.A.
J. Ferna
´
ndez-Lo
´
pez Miguel Hernandez University, Orihuela, Spain
H. Douglas Goff Department of Food Science, University of Guelph, Guelph, Ontario,
Canada
Natalia F. Gonza
´
lez-Me
´
ndez Centro de Investigacio
´
n en Alimentacio
´
n y Desarrollo,

A.C., Hermosillo, Sonora, Mexico
Richard W. Hartel Department of Food Science, University of Wisconsin–Madison,
Madison, Wisconsin, U.S.A.
Y. H. Hui Science Technology System, West Sacramento, California, U.S.A.
B. Jamilah University Putra Malaysia, Selangor, Malaysia
Shann-Tzong Jiang National Taiwan Ocean University, Keelung, Taiwan
William L. Kerr Department of Food Science and Technology, University of Georgia,
Athens, Georgia, U.S.A.
Christopher C. Lai Pacteco Inc., Kalamazoo, Michigan, U.S.A.
Alain Le Bail ENITIAA–UMR GEPEA, Nantes, France
Tung-Ching Lee Department of Food Science, Rutgers University, New Brunswick, New
Jersey, U.S.A.
Martine Le Meste ENSBANA–Universite
´
de Bourgogne, Dijon, France
Alberto Edel Leon Universidad Nacional de Co
´
rdoba, Co
´
rdoba, Argentina
Jens Liesebach Department of Food Science, University of Otago, Dunedin, New
Zealand
Miang H. Lim Department of Food Science, University of Otago, Dunedin, New
Zealand
Ana Guadalupe Luque-Alcara
´
z Centro de Investigacio
´
n en Alimentacio
´

n y Desarrollo,
A.C., Hermosillo, Sonora, Mexico
Domingo Martı
´
nez-Romero Miguel Hernandez University, Orihuela, Spain
x Contributors
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
J. Mateo-Oyagu
¨
e Universidad de Leo
´
n, Leo
´
n, Spain
Janet E. McFetridge Department of Food Science, University of Otago, Dunedin, New
Zealand
H.Mu
´
jica-Paz Universidad Auto
´
noma de Chihuahua, Chihuahua, Mexico
Athapol Noomhorm Asian Institute of Technology, Pathumthani, Thailand
Bonnie Sun Pan National Taiwan Ocean University, Keelung, Taiwan
J. A. Pe
´
rez-Alvarez Miguel Hernandez University, Orihuela, Spain
M. L. Pe
´
rez-Chabela Universidad Auto
´

noma Metropolitana, Mexico City, Mexico
Edith Ponce-Alquicira Universidad Auto
´
noma Metropolitana, Mexico City, Mexico
Stephen C. Ridley College of Agriculture, Food, and Environmental Science, University
of Wisconsin–River Falls, River Falls, Wisconsin, U.S.A.
M. R. Rosmini Universidad Nacional del Litoral, Santa Fe, Argentina
Peggy Stanfield Dietetic Resources, Twin Falls, Idaho, U.S.A.
Alfonso Totosaus Universidad Auto
´
noma del Estado de Hidalgo, Hidalgo, Mexico
A. Valdez-Fragoso Universidad Auto
´
noma de Chihuahua, Chihuahua, Mexico
Daniel Valero Miguel Hernandez University, Orihuela, Spain
Punchira Vongsawasdi King Mongkut’s University of Technology Thonburi, Bangkok,
Thailand
J. Welti-Chanes Universidad de las Ame
´
ricas—Puebla, Puebla, Mexico
Louise Wicker Department of Food Science and Technology, University of Georgia,
Athens, Georgia, U.S.A.
Kit L. Yam Rutgers University, New Brunswick, New Jersey, U.S.A.
Libertad Zamorano-Garcı
´
a Centro de Investigacio
´
n en Alimentacio
´
n y Desarrollo,

A. C., Hermosillo, Sonora, Mexico
Hua Zhao Rutgers University, New Brunswick, New Jersey, U.S.A.
Contributors xi
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
1
Freezing Processes: Physical Aspects
Alain Le Bail
ENITIAA–UMR GEPEA, Nantes, France
I. INTRODUCTION
This chapter presents the freezing process. Selected models permitting estimates of the
freezing time are proposed and discussed. These models are based on the classical model
established by Plank; improvements of this model are presented and discussed. The
different types of freezing processes used in the industry are then presented. Blast freezing
is probably the most popular freezing process, but other concepts such as contact freezing
are also used in a wide range of applications. The thermal contact resistance existing
between the refrigerated surface and the product is often neglected; a focus is proposed on
this aspect. The discussion ends with an evaluation of the freezing rate.
II. FREEZING PROCESS
A. Heat Transfer During Freezing
The heat transfer phenomenon involved in freezing of biological material is basically
nonlinear heat transfer. The latent heat of water represents a large amount of heat that has
to be removed from the foodstuff. Generally, a high freezing rate is desired in order to
obtain numerous small ice crystals. Nevertheless, this is not always the case. For example,
consider frozen dough, for which a slower freezing gives a better preservation of yeast
activity. Freezers can be classified in two families; batch freezers, for which a given amount
of product will be frozen in the same batch, and continuous freezers, which can be
operated in a production line. The refrigeration system used allows classifying freezers in
two other subfamilies: freezers using cryogenic fluids such as carbon dioxide or liquid
nitrogen, and freezers using a mechanical refrigeration unit and a secondary refrigeration
fluid (air, brine, etc.). Mechanical refrigeration units are used for a large majority of

industrial freezers. Cryogenic fluid will be used for special applications requiring (a)
minimal investment, (b) specific use (i.e., meat grinding), or (c) high freezing rate. Heat
transfer conditions and thus the freezing rate are closely related to the type of freezer.
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
B. Freezing Time
A basic analytical model has been proposed by Plank (1941) assuming that (a) the initial
temperature of the product is equal to the phase change temperature, (b) phase change
occurs at constant temperature, and (c) all thermophysical properties and heat transfer
coefficients are constants. Consequently, the initial cooling and final cooling after freezing
are not taken into account. The freezing time given by the Plank formula is proposed in
Eq. (1).
t ¼
DH ? r ? X
ðT
a
À T
f
Þ? N
?
1
h
þ
X
4l
F
!
ðsÞð1Þ
where t is freezing (thawing) time (s); DH is enthalpy difference over the freezing plateau
(J ? kg
À1

); r is density of the frozen food (kg ? m
À3
); T
a
is medium temperature (K or 8C);
T
f
is initial freezing temperature (K or 8C); l
F
is thermal conductivity of the food in the
frozen state (W ? m
À1
? K
À1
); H is heat transfer coefficient (W ? m
À2
? K
À1
); X is
characteristic dimension (m); and N is coefficient (see Table 1).
Based on this first approach, several authors attempted to improve the accuracy of
the freezing (or thawing) time calculation. Ramaswamy et al. (1984) proposed a review of
these equations. Nagoaka et al. (1955) proposed Eq. (2), which takes into account the
amount of heat to be removed during the pre- and postfreezing periods. In this equation,
DH represents the enthalpy difference between the initial temperature (T
i
) of the product
and the final temperature at the end of freezing (J ? kg
À1
).

t ¼½1 þ 0:008
*
T
i

DH ? r
T
f
À T
a
ðÞ
?
PX
h
þ
RX
2
l
F
!
ðsÞð2Þ
with P and R geometric factors as defined in the table below.
Geometry XPR
Slab Thickness 0.5 0.125
Cylinder Diameter 0.25 0.0625
Sphere Diameter 0.167 0.0416
Levy (1958) proposed an expression extrapolated from the model of Nagoaka that
differs mainly in that the temperature difference between initial and final conditions is
explicitly taken into account, the enthalpy difference being considered between the initial
temperature (T

i
) of the product and the final temperature at the end of freezing (J ? kg
À1
)
Table 1 Coefficient of the Plank Formula
Geometry LN
Slab Thickness 2
Cylinder Diameter 4
Sphere Diameter 6
2 LeBail
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
(Eq. 3). The International Institute of Refrigeration (1972) proposed Eq. (3), which is once
again very similar to the model of Nagoaka. This time it is the enthalpy difference that is
taken into account between the initial freezing temperature and the final temperature T
c
(DH ¼enthalpy between T
f
and final temperature of the product T
c
in J ? kg
À1
) (Eq. 4).
t ¼½1 þ 0:008ðT
i
À T
f
Þ
DH ? r
F
T

f
À T
a
ðÞ
?
PX
h
þ
RX
2
l
F
!
ðsÞð3Þ
t ¼
DH ? r
F
T
ifp
À T
a
ÀÁ
?
PX
h
þ
RX
2
l
F

!
ðsÞ t ¼
DH
0
? r
0
ðT
f
À T
a
Þ
?
Pd
h
þ
Rd
2
l
F
!
ðsÞð4Þ
Cleland and Earle (Cleland et al., 1979) used an approach based on numerical models and
experimental results. The Plank equation is proposed in nondimensional form using the
Fourier number Fo based on the P and R factors and on the Biot number (Bi) and Stefan
number (Ste).
F
0
¼
P
BiSte

þ
R
Ste
ð5Þ
Ste ¼
DH between T
f
and T
a
DH between T
f
and T
final
ð6Þ
Bi ¼
hX
2l
F
ð7Þ
These authors introduced a dimensionless Plank number [Pk, Eq. (8)] and defined new
coefficients P
*
and R
*
of the new Plank equation for a slab.
Pk ¼
DH between T
i
and T
f

DH between T
f
and T
final
ð8Þ
P
*
¼ 0:5072 þ 0:2018Pk þSteð0:3224Pk þ
0:0105
Bi
þ 0:0681Þð9Þ
R
*
¼ 0:1684 þ Steð0:2740Pk þ 0:0135Þð10Þ
t ¼
DH ðbetween T
f
and T
final
Þ? r
ðT
f
À T
a
Þ
?
R
*
X
h

þ
R
*
X
2
l
F
!
ðsÞð11Þ
This equation was acceptable for initial temperature T
i
< 40

C and medium temperature
between À15

C < Ta < 40

C, heat transfer coefficients between 10 < h < 500 Wm
À2
? k
À1
and maximum slab thickness of 12 cm. For an infinite cylinder, they found Eqs. (12) and
(13).
P
*
¼ 0:3751 þ 0:0999Pk þSteð0:4008Pk þ
0:0710
Bi
À 0:5865Þð12Þ

R
*
¼ 0:0133 þ Steð0:0415Pk þ 0:3957Þð13Þ
Freezing Processes: Physical Aspects 3
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
And for a sphere,
P
*
¼ 0:1084 þ 0:0924Pk þSteð0:2310Pk À
0:3114
Bi
þ 0:6739Þð14Þ
R
*
¼ 0:0784 þ Steð0:0386Pk À 0:1694Þð15Þ
These latter were applicable for 0:155 < Ste < 0:345; 0:5 < Bi < 4:5, and 0 < Pk < 0:55.
The numerical parameters were obtained from experiments realized with thylose gels (77%
w.c.). The accuracy of these relationships for freezing time prediction were within +3%
for slabs, +5.2% for infinite cylinders and +3.8% for spheres Ramaswamy and Tung
(1984) established a new model extrapolated from the previous one. They use a regression
approach:
t ¼½0:3022CðT
i
À T
f
ÞþL þ2:428C
0
ðT
f
À T

c
Þ
r
0
ðT
f
À T
a
Þ
?
Pd
h
þ
Rd
2
l
F
!
ðsÞð16Þ
With C and C
0
¼ specific heat of the foodstuff respectively before and after freezing
(J ? kg
À1
? K
À1
) and L ¼enthalpy over the freezing plateau (J ? kg
À1
). This formula was
established in the following conditions:

1

C < T
i
< 25

C; À18

C < T
c
< À10

C; À178

C < T
a
< À18

C;
13:9 < h < 68:4W? m
2
? k
À1
:
Other expressions are available in the literature but the one proposed above can be
considered as a good basis. The accuracy can be greatly improved by using numerical
models for which extensive studies have been done (Cleland, 1990). These models allow us
to take into account the time-dependent heat transfer coefficient and the temperature.
Modern software is now available to realize this type of modeling without major
difficulties.

III. CONVECTIVE PROCESSES: AIR FREEZING, BRINE FREEZING,
CRYOGENIC FREEZING
In the case of convective freezing, air, a cryogenic fluid (mainly liquid nitrogen), or a brine
can be used as refrigerant. In the case of air, it can be admitted that the air velocity is in the
range of 1 to 5 m ? s
À1
for most industrial application, leading to the effective heat transfer
coefficient in the range of 10 to 50 W ? m
À2
k
À1
between the medium and the product.
Large-scale spiral freezers have been developed by equipment companies and are widely
used in the industry. Individual quick freezing (IQF) consists of freezing small products
individually with a high air speed (i.e., 1–5 m ? s
À1
). Freezing of larger products can be
realized with blast air but will yield a low freezing rate and thus a low quality in terms of
ice crystal size; plate freezers are preferred. Some specificity in terms of air flow pattern
have been developed in order to reduce water loss by dehydration (i.e., counter flow or
partial counter flow with air inlet a mid position between entrance and exit). Higher air
velocity can also be imposed in a local section such as the entrance of the freezer. This has
been developed for small products (few centimeters thick). It improves heat transfer and
permits a superficial freezing which will minimize mass losses by reducing partial vapor
pressure at the surface of the product. This partial superficial freezing is also called crust
freezing or cryomechanical freezing (Macchi, 1995; Agnelli et al., 2001); it can be realized
4 LeBail
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
by using a liquid nitrogen bath in which the products are floating for a short period before
traveling either into a conventional belt freezer at follow or in the vicinity of the cryogenic

bath to gain the benefit of the vaporized gas for the final freezing (see Mermelstein, 1997).
Freezing in a brine will yield a much higher heat transfer coefficient than in blast air. The
product can be wrapped; in this case, a brine made of CaCl
2
, NaCl, propylene glycol,
ethanol or mixtures of them can be used (Venger et al., 1990). In the case of unwrapped
products (Lucas et al., 1999) the freezing process can be combined with a soaking effect.
Soaking resulted in a salt concentration at the surface of the product and prevented
freezing in an external layer (a nonfrozen layer of ca. 1 mm has been observed [Lucas et al.,
1999b]). Brine freezing is also used for small products such as shrimp to prevent excessive
water loss by dehydration and eventually to enhance solute intake; in the case of shrimp,
for example, a brine solute is generally a mixture of salt and sugar. One major drawback of
the immersion technique is that the brine concentration is changing during the process,
requiring a specific adjustment of it during the treatment. Undesirable side effects may
occur such as spoilage of the product by the brine (requiring filtering and cleaning of the
brine) and cross-contamination of pathogenic microorganisms, as shown by Berry et al.
(1998).
The use of solid carbon dioyde as a refrigerant is an intermediate solution between
contact freezing (contact of the solid flakes of CO
2
) and convective freezing (convective
heat transfer between the sublimated CO
2
and the product). An optimal design will result
in a temperature of the gaseous CO
2
as high as possible.
IV. CONDUCTIVE PROCESSES: CONTACT FREEZERS
Contact freezing can be considered as a mass production process that has a relatively high
freezing rate. IQF and cryogenic refrigeration will yield yet higher freezing rates. It is

widely used in the industry to produce slabs of frozen foods such as fish filets and mashed
vegetables. As for any freezing process, the geometry of the product will rule the freezing
rate as described by the Plank equation (Plank, 1941). Two classes of contact freezers can
be defined, continuous and batch systems. In batch systems, the product is usually frozen
from both sides, by plane heat exchangers applying a certain pressure against the product
(Fig. 1). Continuous systems are usually operated by applying a thin product on a
refrigerated surface and by scraping it off after freezing. Two concepts have been
developed, namely rotating drum freezers and linear belt freezers (Marizy et al., 1998).
In batch plate freezers, the product is usually installed in a cardboard box containing
a plastic film or pouch. It can eventually be installed directly against the refrigerated
surface, but this will create a problem in removing the frozen product at the end of the
process. Rotating drum freezers (Fig. 2) have been developed in the industry to freeze
Figure 1 Contact freezer: batch system.
Freezing Processes: Physical Aspects 5
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
liquid or viscous products or even solid products such as fish filets. In this case, the
product is directly applied against a rotating metallic drum refrigerated from the inside by
a brine, for example. Such a process has been modeled by Marizy et al. (1998). The
product is applied on one side (thickness between 1 mm and a few cm) and is scraped after
a rotation. Madsen (1983) studied drum freezing of codfish and showed that this process
improved the storage stability of cod relative to those frozen in a plate freezer. Recent
literature on drum freezer technology mainly concerns patents. Some of these patents are
related to heat transfer improvement between the refrigerant and the drum, such as
Reynolds (1993) in the case of boiling refrigerant (R22 type). Specific patents are related to
food preparation and processing such as a patent (Hoogstad, 1988) for preparing tea or
coffee extracts destined to freeze drying, a patent (Dalmau, 1987) for citrus fruit freezing, a
patent (Roth, 1982) for freezing and forming meat patties, and a patent (W.a.A., 1969) for
shrimp processing. Several refrigeration techniques are used: boiling refrigerant (i.e.,
Reynolds (1993)), cryogenic refrigerant (Anonymous, 1980) or brine. Cryogenic
refrigerant such as liquid nitrogen is an expansive solution, as gas is emitted to the

ambience. Nevertheless, its very low phase change temperature permits it to achieve high
heat flux and thus fast freezing. Wentworth et al. (1968) presented a development for
increasing the efficiency of the cryogenic fluid distribution thanks to a jacket. Anonymous
(1980) describes a process in which the disadvantage of the stagnant cryogenic fluid at the
lower part of the drum is used as an advantage to remove the product from the drum
(owing to the thermal shock caused by the sudden cooling).
Machinery using the linear design (Fig. 3) has been recently developed and proposed
on the market, while the rotating design has been used several years to freeze liquid or
semiliquids foods. The linear continuous contact freezer consists of a refrigerated surface
on which a plastic film is sliding, or on which food is frozen on a mobile refrigerated
surface. In the former, the product is applied onto the film and is frozen during its
translation on the refrigerated surface. After freezing, the product is removed from the
film, which is discarded. Additional refrigerating effect is usually added by allowing
Figure 2 Contact freezer: rotating drum freezer for liquid and viscous food.
Figure 3 Contact freezer: linear system with a sliding plastic film.
6 LeBail
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
refrigerated air on top of the product. This kind of process is not extremely efficient owing
to thermal contact resistance between the product and the refrigerated surface. The plastic
film represents a first thermal resistance. Moreover, the uncontrollable deformation of the
product will result in the apparition of an air film between the cold surface and the plastic
film. Nevertheless, this process remains very handy in achieving a superficial freezing of
the product preventing dehydration during conventional freezing.
Single-side contact freezing is of course less efficient than double-side contact
freezing. Nevertheless, a continuous process is highly desirable in the industry in order to
minimize contamination by handling and also for productivity. Maltini (1984) studied the
application of solid foods to contact freezing processes and did a comparison with air
freezing. He suggested that the food should be regular in shape to ensure contact of greater
than 20 mm
2

? g
À
1
. Donati (1983) did a similar study and compared the drum freezing
technique with several other freezing processes.
V. SCRAPED FREEZERS
The scraped heat exchanger has been adapted to the case of freezing ice cream. It typically
consists of a rotating drum equipped with one or two blades. The rotating drum is
installed in a refrigerated vessel. The blades allow the scraping of the ice crystals formed
onto the inner surface of the refrigerated drum. They also permits whipping of the air
inside the mix. Indeed, a certain overrun (amount of air entrapped in the final ice cream)
must be obtained to ensure an acceptable texture of the ice cream. For this purpose, a
certain back pressure must be applied at the exit of the system (between ca. 100 and
500 kPa). Fat and air structures in ice creams have been investigated by several authors
(Bolliger et al., 2000). The temperature of the refrigerated surface, the formulation, the
back pressure, and the rotating speed will interact with the degree of fat destabilization
and foam structure.
Straight blades are usually used. Helical blades aiming to propel the ice cream mix
toward the exit of the system has been evaluated by Myerly (1998). More recently, a new
design of continuous freezer has been developed by Windhab et al. (1998). This system has
been developed from a twin screw extruder that has been adapted for the freezing of ice
creams. The enhanced local shear stresses acting in the extrusion channel resulted in
improved microstructure in comparison with conventional scraped heat exchanger. This
process, known as cold extrusion, can yield an ice cream at a much lower temperature than
a conventional scraped heat exchanger. Thus the ice cream obtained with such a freezer
cannot be used to fill forms or mold but does not need any further conventional hardening.
VI. THERMAL PERFORMANCE OF FREEZING PROCESSES
The thermal performance of a given freezing process is related to the overall energy
consumption required to cool down a given product from an initial temperature down to a
final one. An accurate evaluation of this parameter is difficult because it has to take into

account the type of refrigerating system (mechanical compression, cryogenic) being
considered, the geometry of the product, the freezing rate, the final temperature, and the
balance that will be considered between the refrigeration in the freezing process per se and
the refrigeration load that will be held by the storage system. A first approach can be
realized by comparing the heat transfer coefficient between the refrigerated medium
Freezing Processes: Physical Aspects 7
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
(convective freezing) and the surface (contact freezing). The order of magnitude of the
effective heat transfer coefficient as defined by Eq. (17) for convective process or by Eq.
(18) for contact freezing are summarized in Table 2.
U ¼ h ¼
F
ðT
F
À T
S
Þ
with T
S
andT
F
¼ T @ surface and fluid ð17Þ
U ¼
1
TCR
¼
F
ðT
RS
À T

FS
Þ
with TCR ¼ thermal contact resistance ð18Þ
T
RS
and T
FS
¼ T @ refrigerated and food surface.
One can see from Table 2 that assuming a ‘‘perfect’’ thermal contact in the case of
contact freezing is not acceptable. If the product is applied directly onto the refrigerated
surface, heat transfer coefficients as high as 500 to 1000 can be expected, but are a function
of process parameters as detailed by LeBail et al. (1998). A drum freezer used to freeze
mashed vegetables (broccoli in the present case) was studied. A parameter study showed
that the thermal contact resistance was higher for lower surface temperature. This
supposes that the mechanical stress in the product during freezing (stretching owing to ice
formation) interacts with the quality of the mechanical and therefore the thermal contact
between the surface and the frozen product. The roughness of the metallic surface is also
an important parameter. Specific study of LeBail (unpublished data) showed that a factor
of 10 can be observed between a smooth and a rough surface of stainless steel (ca.
70 W ? m
À2
? K
À1
and 700 W ? m
À2
? K
À1
for the rough and the smooth surface,
respectively). The presence of packaging drastically deteriorates the heat transfer. A
plastic film seems to reduce slightly the heat transfer coefficient, whereas the presence of

cardboard results in a heat transfer coefficient that can be as low as 20 W ? m
À2
? K
À1
(Creed et al., 1985), which is comparable to blast air freezing.
The thermal efficiency of the freezing process is thus highly related to the geometry
of the product, its physical state (solid, liquid), and the process that is considered. A high
Table 2 Effective Heat Transfer Coefficients U for Convective and Conductive Freezing Processes
as Defined by Eqs. (17) and (18)
Process Conditions U (W ? m
À2
? K
À1
) Ref.
Convective Blast air 10–50
Convective Brine 50–500
Convective Liquid nitrogen, smooth
cylinders, warming regime
U ¼
1860
ðT
F
ÀT
S
Þ
þ 125 W ? m
À2
? K
À1
ÂÃ

120–200 Macchi, 1995
Convective Liquid nitrogen, strawberry, meat
balls (experimental)
170–230 Agnelli and
Mascheroni, 2001
Conductive Drum freezer with direct contact,
sample ¼mashed broccoli mean
value ranging between (vs. process
parameters)
214
1000–166
Marizy et al.,
1998; LeBail et
al., 1998
Conductive Plate freezer: sample: copper block;
310 kPa pressure
Creed et al., 1985
No wrapping 481
1 layer polyethylene film 278
Corrugated fiberboard þpolyethylene 20
8 LeBail
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
freezing rate is usually desired except for some specific products (i.e., bread dough). The
packaging plays a major role. It permits a reduction of water loss but has a negative effect
on the heat transfer rate. At present, individual quick freezing of meat products, for
example, chicken breast, is much used in the industry. Recent studies have pointed to
immersion freezing in brine, even though some problems will necessarily occur for the
treatment of the brine. Contact freezing is also widely used and is well adapted for mass
production. The presence of packaging is required for obvious handling reasons (removal
of the frozen product from the refrigerated surface) but has a very negative impact on the

efficiency of the process. Direct freezing in cardboard should be avoided.
VII. FREEZING RATE
The freezing rate is central to the final quality of frozen foods. A slow rate results in cell
dehydration and large ice crystals that might damage the texture of a food. A fast freezing
rate prevents the migration of water into the extracellular spacing and yields fine and
numerous ice crystals. Side effects such as the increasing of the concentration of the
remaining aqueous solution might affect the integrity of cell membranes or of proteins.
The freezing rate is a very general statement used most of the time to compare freezing
conditions. The freezing rate is numerically presented in two ways in the literature: Plank
(1941) proposed an expression of the freezing rate evaluated as the velocity of the phase
change front (dimension/time). The International Institute of Refrigeration (IIR, 1972)
defined the nominal freezing time as the duration between 0

C and 10

C above the initial
freezing temperature. Based on this definition, several researchers calculated the freezing
rate by a ratio of temperature difference and the respective duration (Eq. (19) in K/time).
This approach, which can be called temperature formulation, yields a freezing rate unit in
K ? s
À1
or in K ? min
À1
(practical unit). The approach proposed by Plank (1941) will be
called the Plank formulation and yields freezing rate in m ? s
À1
or cm ? h
À1
(practical unit).
Plank calculated the velocity of the phase change front by deriving the expression of the

freezing time. This yielded Eq. (19a–c), respectively, for slab, cylinder, and sphere with
x ¼distance from center (slab), r ¼radius, and r
o
¼outer radius of the geometry.
wðxÞ¼
T
a
À T
f
ðÞ
r ? DH ?
1
h
þ
x
l
F
hi
m ? s
À1
ÀÁ
ð19aÞ
wðrÞ¼
T
a
À T
f
ðÞ
r ? DH ?
r

r
o
? h
þ
r
l
F
Ln
r
o
r
ÀÁ
hi
ðm ? s
À1
Þð19bÞ
wðrÞ¼
T
a
À T
f
ðÞ
r ? DH ?
r
2
r
2
o
? h
þ

rðr
2
=r
o
ÀrÞ
l
F
hi
ðm ? s
À1
Þð19cÞ
F
T
ðrÞ¼
T
1
À T
2
t
1
À t
2
ðK ? min
À1
Þð19dÞ
In the case of the temperature formulation, a beginning criterion and an ending criterion
for freezing must be defined [subscripts 1 and 2 in Eq. (19)]. LeBail et al. (1996, 1998b)
showed that the freezing rate value is dependent on these criteria. Thus the use of the
freezing rate from Plank expression or the evaluation of the freezing rate from the ratio of
the lower thickness by the corresponding freezing time [i.e., determined by the nominal

freezing time (IIR, 1972)]. In this latter case, a mean freezing rate will be obtained.
Freezing Processes: Physical Aspects 9
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
A high freezing rate might result in cracks in the product. Shi et al. (1999) reported
that stress as high as 2 MPa (20 atm) can be reached during the freezing of biological
tissue. This result was obtained from a mathematical model (viscoelastic model coupled to
a thermal model) developed to study the freezing of a sample of potato (17.8 mm
diameter). A frozen mantle first appears at the surface of the product. Meanwhile, the
formation of ice at the core will yield an increase of the pressure. Radial and
circumferential stresses develop during freezing. Rapid temperature drop (i.e., cryogenic
freezing with surface temperature down to À196

C for liquid nitrogen) will induce higher
stress, which can be as high as 1.5 MPa, whereas freezing in a medium at À40

C yields
stress in the range of 1 to 0.5 MPa (Shi et al., 1999). Even though the temperature of the
frozen external mantle is far from the glass transition, the tensile failure strength, which
was around 0.5 MPa for potato, might be passed, leading to cracks. On the other hand, a
depression of the initial freezing point due to a pressure increase will result in a partial
thawing (Otero et al., 2000) leading to a release of the stress.
VIII. CONCLUSION
This chapter offers a general presentation of the freezing processes including evaluation of
the freezing time. A focus proposed on the surface heat transfer coefficient includes
contact freezing. It shows that an infinite heat transfer coefficient can’t be assumed in this
process, which is widely used in the industry.
REFERENCES
IIR (1972). Recommendations for the processing and handling of frozen foods. International
Institute of Refrigeration, 2nd Ed. Paris, 1972.
M, Agnelli, et al. (2001). Cryomechanical freezing. A model for the heat transfer process. Journal of

Food Engineering 47:263–270, 2002.
Anonymous (1980). Method and Apparatus for Cooling and Freezing. Patent UK-2023.789A.
ED Berry, et al. (1998). Bacterial cross-contamination of meat during liquid-nitrogen immersion
freezing. Journal of Food Protection 61(9):1103–1108.
S Bolliger, et al. (2000). Correlation between colloidal properties of ice cream mix and ice cream.
International Dairy Journal 10(4):303–309.
A Cleland, et al. (1979b). A comparison of methods for predicting the freezing times of cylindrical
and spherical foodstuffs. Journal of Food Science 44:958.
AC Cleland, (1990). Food refrigeration processes. Analysis, design and simulation. E. Sciences,
p. 284.
D Cleland, et al. (1986). Prediction of thawing times for food of simple shape. International Journal
of Refrigeration 10:32–39.
PG Creed, et al. (1985). Heat transfer during the freezing of liver in a plate freezer. Journal of Food
Science 50:285–294.
G Dalmau, (1987). Method for freezing citrus fruit portions. Patent EP.0248.753.A2.
L Donati, (1983). Freezing of foods. Effects of freezing on thermophysical properties of foods.
Technologie-Alimentari 6(6):21–31.
B Hoogstad, (1988). Method of preparing a freeze-dried food product. Unilever. Patent EP-
0256.567.A2.
A LeBail, et al. (1996). Application of freezing rate expressions and gassing power to frozen bread
dough. Proceedings of the International ASME Congress, Atlanta, GA, USA.
10 LeBail
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
A LeBail, et al. (1998a). Continuous Contact Freezers for Freezing of Liquid or Semi-Liquid Foods.
Influence of the Thermal Contact Resistance Between Food and Refrigerated Surface.
Symposium of the International Institute of Refrigeration, Nantes, France.
A LeBail, et al. (1998b). Influence of the freezing rate and of storage duration on the gassing power
of frozen bread dough. Symposium of the International Institute of Refrigeration, Nantes,
France.
F Levy, (1958). Calculating freezing time of fish in airblast freezers. Journal of Refrigeration 1(55).

T Lucas, et al. (1999a). Mass and thermal behaviour of the food surface during immersion freezing.
Journal of Food Engineering 41(1):23–32.
T Lucas, et al. (1999b). Factors influencing mass transfer during immersion cold storage of apples in
NaCl/sucrose solutions. Lebensmittel Wissenschaft und Technologie 32(6):327–332.
H Macchi, (1995). Conge
´
lation alimentaire par froid mixte. Proce
´
de
´
avec pre
´
traitement par
immersion dans l’azote liquide. ENGREF, Paris.
A Madsen, (1983). Drum-freezing and extrusion of fish. Boletim de Pesquisa, EMBRAPA Centro de
Technologia Agricola e Alimentar (Brazil) 1:204–205.
E Maltini, (1984). Contact freezing, Industrie-Alimentari. 23 218:573–580.
C Marizy, et al. (1998). Modelling of a drum freezer. Application to the freezing of mashed broccoli.
Journal of Food Engineering 37(3):305–322.
NH Mermelstein, (1997). Triple-pass immersion freezer eliminates need for separate mechanical
freezer. Food Technology 51(7):133.
RMS Myerly, (1998). Stepped helical scraper blade for ice cream maker. United States Patent US-
845349.
J Nagaoka, et al. (1955). Experiments on the freezing of fish by the air-blast freezer. Journal of
Tokyo University of Fischery 42(1):65.
L Otero, et al. (2000). High pressure shift freezing. Part 1. Amount of ice instantaneously formed in
the process. Biotechnol. Prog. 16:1030–1036.
R Plank, (1941). Beitrage zur berechnung und bewertung der gefriergeschwindigkeit von
lebensmittel. Beiheft zur Zeitschrift fu
¨

r die gesamte Ka
¨
lte-industrie 3(10):1–16.
HS Ramaswamy, et al. (1984). A review on predicting freezing times of foods. Journal of Food
Process Engineering 7:169–203.
M Reynolds, (1993). Drum Contact Freezer System and Method. US Patent US-5199.279.
E Roth, (1982). Method of Freezing and Forming Meat Patties. US Patents US-4849.575.
X Shi, et al. (1999). Thermal fracture in a biomaterial during rapid freezing. Journal of Thermal
Stresses 22:275–292.
KP Venger, et al. (1990). Freezing of fish by immersion in non-boiling liquid (in Russian).
Kholodil’naya Tekhnika 5:30–32.
Wentworth and Associates Inc. (1969). Shrimp Processing. Patent UK-1.173.348.
A Wentworth, et al. (1968). Quick Freezing Apparatus. US Patent US-3410.108.
EJ Windhab, et al. (1998). Low temperature ice-cream extrusion technology and related ice-cream
properties. European Dairy Magazine 1:24–29.
Freezing Processes: Physical Aspects 11
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
2
Principles of Freeze-Concentration and
Freeze-Drying
J. Welti-Chanes and D. Bermu
´
dez
Universidad de las Ame
´
ricas–Puebla, Puebla, Mexico
A. Valdez-Fragoso and H. Mu
´
jica-Paz
Universidad Auto

´
noma de Chihuahua, Chihuahua, Mexico
S. M. Alzamora
Universidad de Buenos Aires, Buenos Aires, Argentina
In freeze-concentration and freeze-drying processes, water is first frozen in the material.
Ice is removed by mechanical means during freeze-concentration, leaving a concentrated
liquid, while ice is removed by sublimation in freeze-drying, yielding a dried material. The
removal of water by these methods yields high-quality products, but in both processes it is
a very expensive operation owing to the high consumption of energy. Knowledge of the
theoretical principles behind these processes is necessary for minimization of detrimental
changes, operating strategies, and optimization purposes. Thus the fundamental aspects of
freeze-concentration and freeze-drying are presented in this chapter.
I. FREEZE-CONCENTRATION
A. Introduction
Freeze-concentration is the term used to describe the solute redistribution in an aqueous
solution with an initial relatively low concentration by the partial freezing of water and
subsequent separation of the resulting ice [1,2]. Freeze-concentration is based on the
freezing temperature-concentration diagram (Fig. 1) [3].
It is necessary briefly to review the physicochemical changes that occur during a
freezing process before relating them to the freezing of foods. The phase diagram (Fig. 1)
allows identifying different phase boundaries in a mixture. It consists of the freezing curve
(AB), solubility curve (CE), eutectic point (E), glass transition curve (DFG), and
conditions of maximal freeze-concentration. The freezing curve corresponds to solution–
ice crystals equilibrium. Along this curve, as water is removed as ice, the concentration of
solute increases during the freeze-concentration process. The solubility curve represents
equilibrium between the solution and supersaturated solution in a rubbery state. The
freezing and solubility curves intersect at the eutectic point E (C
e
,T
e

), which is defined as
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
the lowest temperature at which a saturated solution (liquid phase) can exist in equilibrium
with ice crystals (solid phase). The water content at point E is the unfreezeable water.
Below T
e
only ice crystals embedded in a solute–water glass exist. The point F (C
0
g
; T
0
g
)
lower than point B (C
0
g
; T
0
m
) represents a characteristic transition in the state diagram. The
glass transition curve (DFG) represents the glass–rubber transition of the solute–water
mixture, and the type and concentration of the solute and the temperature define it. Above
the DFG curve, solutions are in an unstable rubbery or liquid state; below the DFG curve,
solutions transform into the glassy state (amorphous solid). The maximum freeze-
concentration (maximum ice formation) only occurs in the region above T
0
g
, but below the
equilibrium ice melting temperature of ice (T
0

m
) [4,5]. The liquid solute–water mixture is
the maximum freeze-concentrated and has become glassy. The glass transition
temperature of this unfrozen glassy mixture is designated T
0
g
, and C
0
g
is the solid content
of this glass [3–7]. Figure 1 also shows the aqueous solution with initial concentration and
temperature C
i
and T
i
undergoing freeze-concentration.
B. Freeze-Concentration System
A typical freeze-concentration system (Fig. 2) consists of three fundamental components:
(a) a crystallizer or freezer, (b) an ice–liquid separator, a melter–condenser, and (c) a
refrigeration unit. In the freeze-concentration system, the solution is usually first chilled to
a prefreezing temperature in a cooler (Fig. 2), and then the solution enters the crystallizer
where part of the water crystallizes. Cooling causes ice crystal growth and an increase in
solute concentration. The resulting mixture of ice crystals and concentrated solution is
pumped through a separator where crystals are separated and the concentrated solution is
drained off. Ice crystals are removed and melted by hot refrigerant gas. The final products
are cold water and concentrated solution, which flow separately [1,8].
Figure 1 Typical solid–liquid state diagram for a food system.
14 Welti-Chanes et al.
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
1. Crystallizers

The heat of crystallization can be taken out directly or indirectly. In direct-contact
crystallizers, the original solution is allowed to get in contact with the refrigerant, and heat
is withdrawn by vacuum evaporation of part of the water, usually at pressures below 3 mm
Hg, and by evaporation of the refrigerant. The refrigerants (CO
2
,C
1
–C
3
hydrocarbons)
form icelike gas hydrates, which sequester water at temperatures above 0

C. A
disadvantage of this method is that part of the aromas will be lost during the evaporation.
Direct heat removal is applied in seawater desalinization but is not suitable for liquid
foods, owing to the aroma losses and deterioration of the product by the refrigerant. In
crystallizers with indirect heat removal, the refrigerant (R22 or ammonia) is separated
from diluted solution by a metal wall. So crystallization takes place on chilled surfaces,
from which ice crystals are removed by a scraper. This kind of process has been used
commercially for orange juice and coffee concentration [1,9].
2. Separators of Ice-Concentrated Solution
The separation of ice crystals from concentrated solutions can be performed by the use of
presses, centrifuges, and wash columns, operating in either batch or continuous mode.
Hydraulic and screw presses are used for pressing ice-concentrated slurries to form
an ice cake. Pressures around 100 kg/cm
2
are needed to avoid occlusion of solids in the
cake, which is the limiting factor of this method. Since the presses are completely closed,
aroma loss is negligible [1,9].
Figure 2 Schematic diagram for the freeze-concentration process of foods.

Principles of Freeze-Concentration and Freeze-Drying 15
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.
Ice and concentrated solutions may be separated by centrifugation at about 1000 G.
Centrifugation must be conducted under inert atmospheres to reduce oxidation and aroma
loss. Solute losses may occur if concentrated solution remains adhered to the crystal
surface, but washing of the cake with water will minimize such losses. This washing stage
renders the centrifugation operation more efficient than pressing [1,2,9].
In washing columns, the ice–solution mixture is introduced at the bottom of the
tower, and the solution is drained off. The crystals move toward the top of the column in
countercurrent to the wash liquid, which is obtained by melting part (5–3%) of the washed
crystals leaving the column. In this process the loss of dissolved solids with the ice is less
than 0.01%, and aroma losses are negligible. Wash columns are preferred in freeze-
concentration of low-viscosity liquids such as beer and wine [1,8,9].
C. Influence of Process Parameters
Crystallization is the main step in freeze-concentration, so it is very important to obtain
large and symmetrical crystals. Large crystals can be more easily separated from the
concentrated solution. Large crystals also reduce the loss of solutes due to occlusion and
adherence to the small crystals [1,8]. During crystallization, two kinetic processes take
place: the formation of nuclei and the growth of crystals. Nucleation is the association of
molecules (at some degree of subcooling) into a small particle that serves as a site for
crystal growth. Once a nucleus is formed, crystal growth is simply the enlargement of that
nucleus. Nucleation and growth of crystals are dependent on solute concentration, bulk
supercooling, residence time of the crystals in the crystallizer, freezing rate, molecular
diffusion coefficient of water, and heat transfer conditions. These factors should be
carefully controlled to regulate crystal formation [2,10].
1. Solute Concentration
In general, an increase in solute concentration produces an increase in nucleation and a
decrease in the growth velocity of the ice crystals and in the mean diameter of the crystal.
At critical concentration, solutes may solidify along with ice and are difficult to separate.
Practical maximum concentrations for freeze-concentration are between 45–55% range

[1,9,10].
2. Bulk Supercooling
Supercooling is the driving force responsible for the creation of crystal nuclei and their
growth. The nucleation rate is proportional to the square of the bulk supercooling. At high
bulk supercooling values, the nucleation rate decreases, owing to the inhibition of
molecular mobility. Crystal growth exhibits a first-order dependence of the bulk
supercooling [1,9,10].
3. Residence Time of the Crystals in the Crystallizer
At constant bulk supercooling and solute concentration, the crystal size is proportional to
the crystal residence time. At short residence times the crystals produced are very small
[1,10].
4. Freezing Rate
16 Welti-Chanes et al.
Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.