(In . 1‘ ,
, fififlfiflanfiMM-ufiv. .

rife:
z!
.loC-wI-‘O'

1"; ,
‘1‘: .
9115
§np§l

:2...- EP. 4?... 3.
v :3 . v) a . [P.JI 9 .5.

. ‘ V ‘ .. . ..

:IL w‘1—C‘EI In. '. I.
.. a} :Q 4...: 3-1.”. . .

. . I up; .

l

0 V

i
a

35$?
"i2; ,

)1;1¢.{1.3!v o‘ttfiulu!
5-: k351i 2?

its. 3.:
II t {#0949}.
_) x
9.

In.
Loy . Ly .1 2:97:01 ‘1:
139.3,»... . . prisii. 11.1:
.5‘.‘ . QKII;B2.....!. {:5
:5. V2.32: f. . :zolnz $3
nVelnh I. 377.231}: .1?
. 53:7!-

‘r'.;t“{50"1 Ovlt
wk}! r. {Sr
99".

.DII. It‘ll-D .
fera‘ . ;fa.l!"l

:53. 4 ‘ V
r. .&.t‘..¥l5nv?.t.rnz$
f!.»¥(£ :- ‘s; 7].}. fir

317':
I ,uitut ‘!!77.X|3I.§
i.«.li..~’)t§"f I. f- ,V.

. Orv 11.111 .AIE I'LL '5’3r’1’.
~\:.v~ ‘0: .9. n}? Var).

. .XI-rv- u! L Vll.vv
.ivtfl btr .1???

 

. ..E.
;. ..u|v\’>§¢i Z...
[in
u}

 

TATE UNIVERSITY LIBRARIES

Iiii‘iili‘“ W \lli““|il|HH

 

 

 

 

 

 

MW“ ‘ in

1
Hill
3 1293 00910 3601

 

I"?

This is to certify that the

dissertation entitled

Volatile Retention Kinetics and
Microstructure of Microwave Freeze-Dried
Model Foods

presented by

Su-Der Chen

has been accepted towards fulfillment
of the requirements for

 

 

Ph.D. degree in Food Science and Human

Nutrition/Agricultural
Engineering

M‘i'W‘

Major pffessor Z) V V

 

Date (flit/1.7.” /?q~2

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY
MIchIgan State
UnIverslty

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

'7 " “- :- t- TI
‘Ifbh ":E C" I

all; ’3 S 199E

 

 

 

 

1

 

 

 

L_______
————T

 

fl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Inahution
cums-annea-

 

VOLATILE RETENTION KINETICS AND MICROSTRUCTURE
OF MICROWAVE FREEZE-DRIED MODEL FOODS

By

Su-Der Chen

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition
Department of Agricultural Engineering

1993

ABSTRACT

VOLATILE RETENTION KINETICS AND WCROSTRUCIURE OF MICROWAVE
FREEZE-DRIED MODEL FOODS

By

Su-Der Chen

Volatile compounds are an important component of foods. Even though they are
usually present only in low concentration, they contribute strongly to the overall acceptance
of a food product. As a result, their retention during processing is a key objective. However,
the mechanisms and factors which influence volatile retention in microwave freeze-dried
products have received little attention. Since product temperature and moisture content
change continuously during dehydration, modeling quality degradation kinetics is very
complicated. In general, the rate constants are functions of both temperature and moisture
content, and require several parameters to fit. This study presents a simplified model for
the kinetics of volatile retention during microwave freeze-drying, and should provide

important information for industrial processing and optimization.

The objective of this project were to (1) study the effect of sample composition, freezing
temperature, and chamber pressure on volatile retention and rnicrostructure of both con-
ventionally and microwave freeze-dried model foods, (2) analyze the temperature, moisture
and volatile retention profiles of samples during microwave freeze—drying, and (3) develop

a kinetic model for volatile retention.

A pregelatinized corn starch solution of 8% solids was mixed with 50 or 500 ppm each
of limonene, l-hexanol and l-decanol, and subjected to several sample treatments: control

or null treatment, addition of 2% or 4% B~cyclodextrin (CD), and addition of 2% or 4%

sucrose. The samples were frozen at -15°C, -40°, -60°C, or -l98°C (in liquid nitrogen), and
freeze-dried in a conventional freeze-drier, or in a TM012 mode cavity at 10 W microwave
power input and 1 or 2 torr chamber pressure. The microstructure and volatile retention
were analyzed by scanning electron microscopy (SEM) and gas chromatography (GC),

respectively.

Microwave freeze-drying led to significantly higher volatile retention and faster drying
rates than conventional freeze-drying. Sample treatment was an important factor in the
microstructure of the freeze-dried samples. Samples frozen in liquid nitrogen had small
pores and narrow gaps, which generally led to slower drying rates and lower levels of volatile
retention. l-Decanol retention was higher than l-hexanol retention, probably due to selective
diffusion. Limonene retention in samples containing B-cyclodextrin was very high, probably

due to inclusion complex formation.

Samples that differ by only 4% in formulation had very different temperature, moisture,
and volatile retention profiles during microwave freeze-drying. The sample containing 4%
B-cyclodextrin had the highest volatile retention and fastest drying rate. The rate of volatile
retention depended strongly on moisture content, and each volatile compound reached a
different equilibrium retention level. Using the differential method, volatile. retention was

confirmed to be a first order reaction, with rate constants ranging from 0.01 to 0.06 min".

ACKNOWLEDGMENTS

The author would like to express her sincere gratitude to her advisor, Dr. Robert Y.

Ofoli, for his encouragement, guidance and financial support during this graduate program.

Sincere appreciation is extended to my guidance committee, Dr. Jes Asmussen,
Department of Electrical Engineering, Dr. Jerry N. Cash, Department of Food Science and
Human Nutrition, Dr. James F. Steffe, Department of Food Science and Human Nutrition
and Department of Agricultural Engineering, and Dr. Elaine P. Scott, Department of
Mechanical Engineering, Virginia Polytechnic Institute and State University, for their
assistance and advice of this research. Further acknowledgments are also due to Mr. Ronald

E. Fritz, and Dr. Hoojjat Harold for their technical assistance.

Finally, the author wishes to acknowledge her family and friends for their encour-
agement and assistance. The author especially thanks her husband, Ching-Tung Kuo for

his love, support and understanding throughout this graduate study.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ x
LIST OF FIGURES ............................................................................................... xii
NOMENCLATURE ............................................................................................. xvi
CHAPTER 1 INTRODUCTION AND OBJECTIVES ......................................... 1
CHAPTER 2 LITERATURE REVIEW ............................................................... 3
2.1 Application of microwave energy in freeze-drying ................................... 3
2.2 Factors affecting microwave freeze-drying ............................................... 8
2.2.1 Electric field strength ........................................................................ 8
2.2.2 Chamber pressure ............................................................................. 10
2.2.3 Other factors ..................................................................................... 12
2.3 Problems associated with microwave freeze-drying ................................. 14
2.3.1 Gas ionization and breakdown ......................................................... 14
2.3.2 Corona discharge .............................................................................. 15
2.3.3 Non-uniform heating ......................................................................... 15
2.3.4 Entrainment ...................................................................................... 16
2.4 Modeling of transport phenomena during microwave freeze-drying.’.‘....... 16
2.5 Mechanism of volatile retention during freeze—drying .............................. 18
2.5.1 Adsorption ........................................................................................ 18
2.5.2 Encapsulation ................................................................................... 19
2.5.3 Selective diffusion ............................................................................ 19
2.5.4 Microregion formation ..................................................................... 20
2.5.5 DeItaTtheory ................................................................................... 22
2.6 Factors influencing volatile retention of freeze-dried foods ..................... 22

2.6.1 Solid composition and concentration ............................................... 24

2.6.2 Cyclodextrin inclusion ..................................................................... 24
2.6.3 Freezing rate ..................................................................................... 25
2.6.4 Chamber pressure ............................................................................ 28
2.7 Kinetics of volatile retention ..................................................................... 29
2.8 Microstructure of freeze-dried products .................................................... 29

CHAPTER 3 MICROSTRUCTURE AND VOLATILE RETENTION OF

CONVENTIONALLY FREEZE-DRIED MODEL FOODS ............................... 31
3.1 Introduction ............................................................................................... 31
3.2 Materials and Methods ............................................................................... 34

3.2.1 Preparation of starch gel model foods ............................................. 34
3.2.2 Volatile retention analysis ................................................................ 35
3.2.3 Microstructure analysis .................................................................... 36
3.2.4 Statistical analysis ............................................................................ 36
3.3 Results and Discussion .............................................................................. 36
3.3.1 Volatile retention ............................................................................. 36
3.3.2 Microstructure .................................................................................. 42
3.4 Conclusron ....... 52

CHAPTER 4 VOLATILE RETENTION OF MICROWAVE FREEZE-DRIED

MODEL FOODS .................................................................................................. 53
4.1 Introduction ............................................................................................... 53
4.2 Materials and Methods ............................................................................... 56

4.2.1 Preparation of starch gel model foods ............................................ 56
4.2.2 The microwave freezedrying system .............................................. 57

4.2.2.1 Single mode internally tunable resonance cavity applicator.... 57

vi

4.4.2.2 The microwave generation system ............................................ 60

4.4.2.3 Temperature measurement system ............................................ 62
4.4.2.4 Vacuum chamber ...................................................................... 62
4.4.2.5 Data acquisition system ............................................................. 63

4.2.3 Volatile retention analysis ................................................................ 63
4.2.4 Moisture content measurement ........................................................ 64
4.2.5 Statistical analysis ............................................................................ 64
4.3 Results and Discussion .............................................................................. 65

4.3.1 Volatile retention and final moisture content of microwave

freeze-dried samples ................................................................................. 65

4.3.2 Effect of sample composition on volatile retention ............................. 67

4.3.3 Effect of freezing temperature on volatile retention ........................... 72

4.4 Conclusions ............................................................................................... 76
CHAPTER 5 MICROSTRUCTURE OF MICROWAVE FREEZE-DRIED

MODEL FOODS .................................................................................................. 77

5.1 Introduction ............................................................................................... 77

5.2 Materials and Methods ............................................................................... 78

5.2.1 Preparation of starch gel model foods ..... 78

5.2.2 Microstructure analysis .................................................................... 79

5.3 Results and Discussion .............................................................................. 79

5.3.1 Effect of food composition on microstructure ................................. 79

5.3.2 Effect of freezing temperature on rnicrostructure ............................ 82

5.3.3 Effect of chamber pressure on rnicrostructure ................................. 87

5.4 Conclusions ............................................................................................... 93

vii

CHAPTER 6 KINETICS OF VOLATILE RETENTION DURING MICRO-

WAVE FREEZE-DRYIN G .................................................................................. 94
6.1 Introduction. .............................................................................................. 94
6.2 Theory ....................................................................................................... 95
6.3 Materials and Methods .............................................................................. 100

6.3.1 Preparation of starch gel model foods. ............................................. 100
6.3.2 Experimental set-up .......................................................................... 100
6.3.3 Moisture content measurement. ....................................................... 101
6.3.4 Volatile retention analysis ................................................................ 101
6.3.5 Statistical analysis ............................................................................ 102
6.4 Results and Discussion. ............................................................................. 102
6.4.1 Temperature profiles ........................................................................ 102
6.4.2 Moisture profiles ............................................................................... 103
6.4.3 Volatile retention profiles ................................................................ 109
6.4.4 Kinetics of volatile retentin. ........................................................... 113

6.4.4.1 Determination of kinetic parameters by the integral
method. .................................................................................................. 1 13
6.4.4.2 Determination of the kinetic parameters by the differential

method. .................................................................................................. 116

6.5 Conclusions ............................................................................................... 129
CHAPTER 7 OVERALL CONCLUSIONS ........................................................ 130
CHAPTER 8 SUGGESTIONS FOR FUTURE RESEARCH ............................ 132
CHAPTER 9 BIBLIOGRAPHY .......................................................................... 133
APPENDICES ...................................................................................................... 140

Appendix A. Molecular formulae and chromatography of volatile

compounds ...................................................................................................... 141
Appendix B. A modified Likens-Nickerson apparatus for the simultaneous
distillation/solvent extraction ......................................................................... 142
Appendix C. Basic program to record voltage signals through the data
acquisition system during microwave freeze>drying ...................................... 143
Appendix D. Basic program to compute temperature, pressure, nitrogen flow

rate, and microwave power during microwave freeze-drying ......................... 144
Appendix E. Typical data during microwave freeze-drying ............................ 145
Appendix F. Tukey’s test for determining the significance of factors

influencing volatile retention and final moisture content ................................ 146

Table 2.1

Table 2.2

Table 3.1

Table 3.2

Table 4.1

Table 4.2

Table 4.3

Table 6.1

Table 6.2
Table 6.3

Table 6.4a

Table 6.4b

LIST OF TABLES

AT values, specific heat and dielectric constants of selected

volatiles ....................................................................................... 23
The properties of cyclodextrins (CD) ............................................ 27
Volatile retention of conventionally freeze-dried model foods... 38
AN OVA table showing the effect of type of volatile, sample
composition and freezing temperature on volatile retention ...... 39
Volatile retention and moisture content of microwave
freeze-dried model foods ........................................................... 66
ANOVA table showing the effect of sample composition,
freezing temperature and chamber pressure on the final moisture
content ......................................................................................... 68
AN OVA table showing the effect of type of volatile compound,
sample composition, freezing temperature and chamber pressure
on volatile retention ..................................................................... 69
Quality degradation kinetic models for dynamic processing or
storage ................................................................................ l I": ..... 98
Empirical models for moisture content profiles .......................... 108
Parameters in the kinetic models of Eqs. (6.12) and (6.13) for
limonene retention in the control (integral method) ................... 117
A correlation between moisture content and the rate of volatile
retention in microwave freeze-dried samples ............................. 118
A correlation between average temperature and the rate of

volatile retention in microwave freeze-dried samples ................ 118

Table 6.5 Empirical models for volatile concentration and rate of retention
as a function of drying time ........................................................ 120

Table 6.6 Volatile retention kinetics ........................................................... 125

Figure 2.1

Figure 2.2

Figure 2.3
Figure 2.4

Figure 2.5
Figure 2.6

Figure 2.7

Figure 2.8

Figure 3.1

Figure 3.2
Figure 3.3

Figure 3.4

Figure 3.5

LIST OF FIGURES

Temperature dependence of t»? and e"for various foods at 2.8 GHz. 6

Dielectric constant 8’ and dielectric loss factor 8" of several

aqueous mixtures at 3.0 GHz and 25°C ....................................... 7
Effect of field strength on drying rate ......................................... 9
Effect of pressure on drying rate during microwave freeze-

drying .......................................................................................... 11
Effect of chamber pressure on interface temperature .................. 13
Effect of water concentration on the diffusion coefficient of
alcohols relative to water in maltodextrin at 25°C ....................... 21
Temperature—water concentration dependence of the diffusion

coefficient of propanol relative to water in maltodextrin ............ 21

Chemical formula and molecular shape of B-cyclodextrin .......... 26

Volatile retention in freeze-dried model foods frozen at three
different temperatures .................................................................. 40
Effect of sample treatment on volatile retention ................. - .. ......... 41
Scanning electron micrographs of freeze-dried control samples
frozen at different temperatures .................................................. 43
Scanning electron micrographs of samples containing 2% and
4% B—cyclodextrin and frozen at -15°C ......................................... 46
Scanning electron micrographs of samples containing 2% and

4% sucrose and frozen at-15°C ..................................................... 48

Figure 3.6

Figure 4.1
Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5

Figure 4.6

Figure 4.7

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Scanning electron micrographs of three model foods frozen at
-l98°C .........................................................................................
Experimental setup ......................................................................
Cross-sectional view of the cylindrical tunable microwave cavity
Microwave system with an external circuit .................................
Effect of sample treatment on volatile retention ..........................
Volatile retention of the control frozen at three different tem-

peratures ......................................................................................
Volatile retention of samples containing 4% B-cyclodextrin
frozen at three different temperatures .........................................

Volatile retention of samples containing 4% sucrose frozen at

three different temperatures .........................................................

Microstructure of three samples frozen at -60°C, then

freezeadriedat lOWand 1 torr ......................................................
The surface micrographs of three samples frozen at ~15°C and
-198°C, then freeze-dried at 10 W and 1 torr ..............................

The cross-section micrographs of three model foods systems
frozen at -198°C, then freeze-dried at 10 W and 1 torr .......... L .......
Microstructure of three model foods frozen at -60°C, then
freeze-dried at 10 W and 2 torr ....................................................
The higher magnification cross-section micrographs of the
control and samples containing 4% B—cyclodextrin frozen at

-l98°C, then freeze-dried at 10 W and l torr or 2 torr .................

50
58
59
61
70

73

74

75

80

83

85

88

90

Figure 6.1a

Figure 6.1b

Figure 6.1c

Figure 6.2
Figure 6.3

Figure 6.4

Figure 6.5

Figure 6.6

Figure 6.7

Figure 6.8

Figure 6.9

Figure 6.10

Surface and center temperature profiles of control samples
freeze-dried at 10 W and 1 torr .....................................................
Surface and center temperature profiles of samples containing
4% B—cyclodextrin freeze-dried at 10 W and 1 torr ......................
Surface and center temperature profiles of samples containing
4% sucrose freeze-dried at 10 W and 1 torr ..................................
Moisture profiles of three microwave freeze-dried samples .......
Moisture and volatile retention profiles of the control
freeze-dried at 10 W and l torr ....................................................
Moisture and volatile retention profiles of samples containing
4% B—cyclodextrin freeze-dried at 10 W and 1 torr .....................
Moisture and volatile retention profiles of samples containing
4% sucrose freeze-dried at 10 W and 1 torr ................................
First order limonene retention kinetics in the control using the
integral method ............................................................................
Rate of volatile retention versus average volatile concentration
during microwave freeze—drying of the control ...........................
Rate of volatile retention versus average volatile concentration
during microwave freeze-drying of samples containing 4% B-

cyclodextrin .................................................................................
Rate of volatile retention versus average volatile concentration
during microwave freeze~drying of samples containing 4%
sucrose .........................................................................................
First order volatile retention and rate constant of the control

(integral method) ..........................................................................

xiv

104

105

106
107

110

111

112

115

121

122

123

126

Figure 6.11 First order volatile retention and rate constant of samples
containing 4% B‘cyclodextrin (integral method) .......................... 127
Figure 6.12 First order volatile retention and rate constant of samples

containing 4% sucrose (integral method) ..................................... 128

XV

NOMENCLATURE

[A] concentration of volatile or quality factor (normalized with respect to initial con-
centration)

[A]0 initial concentration of volatile or quality factor (normalized with respect to initial
concentration)

[ALEl equilibrium concentration of volatile (normalized with respect to initial concentra-
tion)

a0 amplitude of sine wave

°Bx solid content

C water vapor concentration, g m’3

C, specific heat, J g" °C’l

D mass diffusivity, m2 s'1

E electric field strength, v m"

E, activation energy, cal/g-mol

f microwave frequency, Hz

H, molar heat of sublimation of ice, cal mole'l

ID index of deterioration, dimensionless

IMC initial methionine content, %

K microwave energy dissipation coefficient, W m‘1 V'2
thermal conductivity, kW rn'l °C"

k reaction rate constant, min'1 for a first order reaction, min“ mol m'3 for a zero order
reaction

k0 pre-exponential factor, min‘1 for a first order reaction, min'l mol m'3 for a zero order
reaction

M or MC moisture content, dimensionless

N, water vapor flux, 3 rn‘2 s"

P microwave power absorption, W in3

P pressure, torr

Pl-P7 parametric constants

PC protein content, %

R ideal gas constant, 1.987 cal mole" K"

t time, s

T temperature, K

T“, mean temperature, K

W weight of sample at any time, g

W, initial weight of sample, g

W, weight of dried sample, g

Greek letters

8* complex dielectric constant, dimensionless
8’ relative dielectric constant, dimensionless
e" dielectric loss factor, dimensionless

e, dielectric constant of free space (8.854 X 10‘"), F rn'l
Q density. 8 m'3

a product porosity, dimensionless
Subscripts

d dried region

frozen region
volatile compound

water

CHAPTER 1
INTRODUCTION AND OBJECTIVES

Freeze-drying is a sublimation process in which ice is transformed to water vapor
without passing through the liquid phase. Both heat and mass transfer occur during
freeze-drying, because heat must be transferred through the dried region to the frozen region,
and the sublimated water vapor removed from the product to a condenser. Since the thermal
conductivity of the dried region of most foods is low, conventional freeze-drying encounters
a large thermal resistance, leading to longer drying times. Microwave energy is a volumetric
heat source which penetrates the dried layer to reach the frozen region of the product. In
addition, the dielectric loss factor of the frozen region is much larger than that of the dried
region; therefore, most of the microwave energy is absorbed by the frozen region to induce
ice sublimation. Thus, microwave freeze-dryin g helps overcome the heat transfer limitation

of conventional freeze-drying, accelerates the drying rate, and could reduce cost.

Since freeze-drying is a low temperature process, it can minimize thermal degradation,
reduce shrinkage of foods, and provide better flavor retention and product quality. The
factors which affect volatile retention in conventionally freeze-dried products have been
studied extensively. They include solid composition, freezing rate, type of volatile, ice core
temperature, structural collapse, and drying rate. However, no comparable studies have

been conducted on volatile retention during microwave freeze-drying.

The microstructure of freeze-dried food products can be analyzed by scanning electron
microscopy (SEM). The resulting information can be helpful in interpreting the effect of

sample composition, freezing temperature, and chamber pressure on volatile retention,

1

drying rate, and pore size distribution. In general, a slower freezing rate leads to larger ice
crystals and results in larger pores, which may improve water vapor transfer during

freeze-drying.

There is no information on volatile retention profiles during a dynamic dehydration
process such as freeze-drying. However, the retention kinetics of other quality factors and
nutrients have been studied and characterized as a first order reaction, with the kinetic
parameters determined by the integral method. Since the rate constants are functions of
temperature and moisture content during drying, many researchers have modeled the pre—
exponential factor and activation energy in the Arrhenius relationship as functions of
moisture content and temperature, by statistical regression analysis. The result is that the
kinetic models are very complicated. Since kinetic models provide important information
on the mechanism of volatile retention, they need to be simplified for industrial utilization

and to enable optimization studies.
The objectives of this project were to:

(1) Study the volatile retention and microstructure of conventionally freeze-dried

model foods.

( 2) Study the effect of sample composition, freezing temperature, and chamber pressure

on both the drying rate and volatile retention.

(3) Analyze the effect of sample composition, freezing temperature, and chamber

pressure on the surface and cross-section rnicrostructure of the products.

(4) Analyze the temperature, moisture, and volatile retention profiles of samples during

microwave freeze-drying, and develop a kinetic model for volatile retention.

CHAPTER 2
LITERATURE REVIEW

2.1 Application of microwave energy in freeze-drying

Freeze-drying is a sublimation process which involves the transformation of ice to
water vapor without passing through the liquid phase. For ice to sublimate, the temperature
and pressure must be held below the triple point of water (0.0098°C and 4.567 torr). The
sublimation latent heat is 2920 I per gram of ice (Decareau, 1985). Since freeze-drying is
a low temperature process, heat sensitive foods, pharmaceuticals and other biological
materials can be freeze-dried to obtain better product quality. Freeze-drying can minimize
thermal degradation, Maillard reactions, enzymatic reactions and protein denaturation, and
lead to better flavor retention. Moreover, it can reduce product shrinkage and produce a

porous structure which enables rapid and easy rehydration.

Both heat and mass transfer occur during freeze-drying, because heat must be trans-
ported through the dried region to the frozen core, and the sublimated water'vapor must be
removed from the material to a condenser. However, thermal conductivity in the dried
region of most foods and biological materials is low; therefore, a large thermal resistance
is encountered in convectional freeze-drying, and requires much time to accomplish. This

fact provides the basic attraction for microwave freeze-drying.

Several investigators have studied ways to improve heat transfer during freeze-drying.
Kan and de Winter ( 1968) introduced inert and high conductivity gases into the freeze-dryin g
chamber to increase heat transfer to the frozen region. However, Sunderland (1980) thought

the high velocity of sublimated water vapor leaving the product during freeze-drying could
3

purge these inert gases and weaken heat conduction. Sunderland (1980) reported that
increasing the surface temperature of materials can increase the drying rate and reduce drying
time in conventional freeze-drying, but it may also cause thermal degradation of the outer
surface and melt the frozen region. Copson (1958) reported that microwave energy can
penetrate the dry region to reach the frozen region of biological products, thereby overcoming

the heat transfer limitations of conventional freeze-drying operations.

Microwave frequencies lie between 300 MHz and 300 GHz. Only two frequencies
(915 MHz and 2450 MHz) are presently used in industrial food processing in the United
States. Microwave devices require a magnetron, a cathode located in the center and an
anode around the circumference. When power is supplied, the cathode emits electrons into
the vacuum space, and the anode has resonant cavities that act as oscillators to generate
electric fields (Copson, 1962). Microwave energy is generated by molecular vibration, and
internal friction between molecules and absorbed as heat. Due to rapid and volumetric
heating, microwave energy is used in several food processing operations such as cooking,

dehydration, blanching, sterilization, pasteurization and tempering (Mudgett, 1986).

The behavior of a material in a microwave field is characterized by the complex

dielectric constant 8':

I

e*=e —ie" (2'1)

which consists of a real part (8’) and an imaginary part (8"): 8’ is the relative dielectric

constant, and represents the capacity of the material to store electrical energy; a" is the

dielectric loss factor, and represents the loss of electrical energy in the material.

The ratio of the loss factor to the dielectric constant is the loss tangent:

tan 8 = e"/e’ (2-2)

which is related to the penetration of products by an electric field and the dissipation of
electrical energy as heat (Harper et al., 1962). Dielectric pr0perties are functions of tem-

perature, frequency and moisture content (Roebuck et al., 1972).

Risman and Bengtsson (1971) determined the dielectric properties of foods by the
cavity perturbation technique. They observed that both the dielectric constant and the loss
factor of frozen foods are very small, but that dielectric properties show a large step increase
around the melting point of water (0°C) (Figure 2.1). Mudgett et al. (1979) also obtained
small values for the dielectric properties of frozen meat, and found that dielectric properties
at -20°C are larger than those at -40°C, because the free water content of frozen meat decreases

with freezing temperature.

Roebuck et a1. (1972) reported the dielectric properties of several carbohydrate-water
mixtures at 3.0 GHz, and showed that the dielectric properties of glucose and sucrose
solutions are higher than those of starch solutions (Figure 2.2). In addition, gelatinized

potato starch-water mixtures have higher 8" than granular potato starch-water mixtures.

The power generated by an electromagnetic field is (Ma and Peltre, 1975a):

P = nfeo my (2.3)

 

    

 

 

I
'04 “33$" 11 cooxeo seer-
\ "
/
5. I / DISTILLED
‘/ warren
in)?” - - . . , - L
-20 -to o 20 (.0 so

TEMPERATURE C°
Figure 2.1 Temperature dependence of e’ and e"for various foods at 2.8 GHz.

(Source: Risman and Bengtsson, 1971)

 

   
 

— amen route era-cu
”" “Llflllllb acme 81““
oo- stucou
-a- seem:
-o- IYNMOOL
-o- stretnot

      
   

 

 

 

 

 

‘0 I00
armrest
ioo‘éo‘éo‘q‘o‘z‘o‘é
xsouos
as:
24.
m?
I6

m.

 

 

  

 

 

 

'zb
3L a —cuuuun route suacu
-- --- «urinate route snacu
w -o-stucos£
—a- sucnost
.F} “O'EVNAIOL
-o- stream
0 A L a a A A l
O 20 4O 60 80 IOO
% WATER
L A A a + 4* J J n 4 _l
IOO IO 60 40 0
% SOLIDS

Flamzzmebeuiccmmme'anddidecuiclossfme'ofsevaflaquemsmixnnes
at 3.0 GHzand 25°C. (Source: Roebucket al., 1972)

where P is the microwave power absorbed, f is the microwave frequency, go is the dielectric
constant of free space (8.854 x 10 "2 Fm") and E is the electric field strength. Microwave

power absorption was approximated by Harper (1962) as:
P = 0.556.1510"2f8"E2 (2.4)

The dielectric loss factor of the frozen region is much larger than that of the dried
region; therefore, most of the microwave energy is absorbed by the frozen region for ice
sublimation. Thus, microwave freeze-drying can accelerate the drying rate by 3 to 13 times

the convectional freeze-drying rate (Ma and Peltre, 1975b) .

Sunderland (1982) investigated the costs of freeze-drying food by conventional and
microwave methods. The total estimated costs per kilogram of frozen food product for
conventional and microwave freeze-drying are $0.27 and $0.18, respectively. In addition
to drastically reducing drying time, microwave freeze-drying also reduces requirements for

equipment, floor space, labor and maintenance (Peltre et al., 1977).

2.2 Factors affecting microwave freeze-drying

2.2.1 Electric field strength

The typical drying curve is a function of electric field strength (Figure 2.3). Increasing
the electric field strength will accelerate the drying rate; however, a microwave power input

that is too high will cause melting of the frozen core, overheating of the material to develop

FRACTION INITIAL MOISTURE

 

Press. : 250 microns
Sample .- I? in. cube

 

 

 

 

ammo rm: . to

Figure 2.3 Effect of field strength on drying rate.
(Source: Ang et al., 1977b)

10

a dark spot, and puffing of the sample. Ang et al. (1977b) reported that corona discharge
will occur when the electric field strength exceeds 170 V/cm. They suggested an optimal
electric field strength of between 100 and 135 V/cm, at a chamber pressure of between 0.2
and 0.4 mmHg, to ensure a faster drying rate without overheating and melting. During
freeze-drying, the moisture content of the sample decreases with drying time; therefore, the
energy requirement for ice sublimation will decrease with time, especially in the falling
drying rate period. Therefore, Ang et al. (1978) adjusted the electric field strength from

170 V/cm to 110 V/cm during microwave freeze-drying to help maintain product quality.

2.2.2 Chamber pressure

The vacuum chamber pressure is another important process variable in microwave

freeze-drying. As shown by the Clausius-Clapeyron equation:

£15_AH,
P ’er

 

dT (2-5)

a low operating chamber pressure corresponds to a lower temperature in the frozen core
(Dyer et al., 1966). Diffusivity of water vapor from the transition region also depends on
pressure: a lower chamber pressure leads to lower mass transfer resistance. In addition, at
a low chamber pressure, microwave power input can be increased to shorten the drying time
without overheating and/or melting of samples (Ang et al., 1977a; Ang et al., 1978). They
studied the effect of chamber pressure on drying rate (Figure 2.4), and reported that melting
and vaporization occurred at a chamber pressure of 3.5 torr. Ma and Peltre (1975b) suggested
that a low chamber pressure can also reduce the probability of corona discharge, ionization

or gas plasma formation.

11

8

 

Mama

In

Corrected mm cm. .I
In

 

 

 

17,-.2 unite
.4 .
firm uni-lg
W ) III "0.
I) I 21 3 4‘ 5 6
Tune, the

Figure 2.4 Effect of pressure on drying rate during microwave freeze-drying.
(Source: Ma and Peltre, 1975b)

12

Introducing inert gases into the chamber may assist in transporting water vapor out of
the vacuum chamber, and reduce the chance of corona discharge during microwave
freeze-drying. The water vapor flow regimes in the capillary channels can be classified as
continuum, transition or free-molecule flow, depending on the ratio of the mean-free path
of the vapor molecules to the capillary diameter, and the Knudsen number (Hill and Sun-
derland, 1971). Hill and Sunderland (1971) analyzed the relationship between the interface
temperature, position and chamber pressure using numerical methods (Figure 2.5) and
reported that slower drying rates for beef slab occur at a chamber pressure of 0.01 and 0.1
torr, because the maximum mass flux is limited by Knudsen diffusion. At operating chamber
pressures between 0.5 and 4 torr, the transport of water vapor is considered to fall in the

transition region. The fastest freeze-drying rate occurs at a chamber pressure of 1 torr.

2.2.3 Other factors

A thick or large sample will limit microwave penetration and cause most of the
microwave power to be absorbed in the surface layer. This leads to non-uniform heat
distribution. Decreasing the dimension of the sample reduces mass transfer resistance,
allowing operation at a higher microwave power without overheating or melting (Ma and

Peltre 1975b; Ang et al., 1977b).

Besides processing conditions, the chemical composition, size, and geometry, as well
as the physical, electrical, and thermal properties of food products will influence their
microwave heating characteristics (Mudgett, 1985). Ma and Peltre (1975b) and Sunderland
(1980) suggested that a slow freezing rate causes larger ice crystal formation. This may

result in wider channel structures which facilitate water vapor removal during freeze-drying.

 

13

 

 

«4

400

 

Internet I'M” ,

i

 

 

 

gaseo-a

   

 

464
7; NIOO'R
W8 '05
4w 1 r m r
0 02 0-4 0" 00 IO

 

Musicales: Matinee position. XII.

Figure 2.5 Effect of chamber pressure on interface temperahue.
(Source: Hill and Sunderland, 1971) l

14

2.3 Problems associated with microwave freeze-drying

2.3.1 Gas ionization and breakdown

The range of microwave frequencies (0.3 - 30 GHz) is also the range for gas collisional
frequencies. This results in a resonance effect. At a low pressure of 0.1 - 5 torr, electrons
can be stripped from the molecules due to the high velocity of collision and can be accelerated
to cause ionization. Gas ionization disrupts the electric field configuration, consumes a
large fraction of microwave energy, limits the magnitude of microwave power during
freeze-drying, and may cause discoloration and flavor loss to foods (Arsem and Ma, 1985;
Sunderland, 1980; Gould and Kenyon, 1971). The microwave gas breakdown curve is a
function of frequency, pressure, temperature, gas composition, shape and geometry of the
cavity and the electric field strength. Gould and Kenyon (1971) reported that a frequency
of 2450 MHz is more suitable than 915 MHz, due to a higher allowable breakdown field

and greater dielectric properties.

Sunderland (1980) suggested that gas ionization can be reduced by' lowering the
chamber pressure to about 0.05 torr. On the other hand, Arsem and Ma (1985) noted that
it is hard to keep the chamber pressure below the breakdown level, because water vapor
generated during freeze-drying may also increase the probability of gas breakdown in the
microwave cavity. However, several investigators suggested that the ionization of gases
can be avoided by variable microwave power control, introducing a fresh load to absorb the
microwave power, or reducing vacuum pressure and electric field strength (Arsem and Ma,

1985; Jackson et al., 1957; Slater, 1957; Sunderland, 1980).

 

15

2.3.2 Corona discharge

Once the gas becomes ionized and develops the ability for electrical conduction, the
microwave power will be reflected from the cavity, and plasma or corona discharge will
occur. This shows up as a visible colored glow, and consumes microwave power within
the cavity (Tetenbaum and Weiss, 1979; Arsem and Ma, 1985; Gould and Kenyon, 1971).
Corona discharges occur at pressure levels of 0.2 - 0.4 torr, and are discharges of water
vapor occur at pressure levels of 0.025 - 0.1 torr, with a high electric field intensity of 400

V/cm (Rosenberg and Bogl, 1987; Peltre et al., 1977).

2.3.3 Non-uniform heating

Non-uniform absorption is caused by variations in the composition or thickness of the
material, or by uneven electric fields. It results in unwanted early corona discharge, local

melting, or overheating in the high energy zone (Harper et al., 1962).

Furthermore, if the microwave power supply is too high, and the water vapor transfer
from the food is slow, the water vapor pressure at the ice front will rise above the triple
point of water to cause a meltback problem. Since water has a higher loss factor than ice,
the liquid water will absorb most of the microwave energy and cause intense localized
heating (Gould and Kenyon, 1977; Sunderland, 1980). To avoid this, Sunderland (1980)
suggested that the heating rate must be matched to the maximum allowable water vapor
flow rate and load changes. The microwave input should be adjusted on the basis of chamber

pressure, temperature, humidity and sample weight to avoid non-uniform heating.

 

 

16

2.3.4 Entrainment

When the chamber pressure is reduced to 0.1 torr in order to avoid corona discharge,
entrainment will occur during microwave freeze-drying of bulky foods (Arsem and Ma,
1985). This can account for about 20% of the total moisture removal for the entire drying
process. It occurs because the expansion ratio of solid to gas is great enough to generate a
high velocity, expanding vapor that is capable of entraining a significant quantity of solids
under low temperature and pressure conditions (Arsem and Ma, 1985; 1990). Entrainment
of hydrated salts and water soluble proteins will affect microwave energy consumption,
drying rate and food quality. Arsem and Ma ( 1985) compared the microstructure and sensory
evaluation of both conventional and microwave freeze-dried steaks, and reported that there
was a noticeable difference in the texture and microstructure of muscle fibers, but there was

no significant difference in sensory characteristics.

2.4 Modeling of transport phenomena during microwave freeze-drying

The earliest mathematical analysis of microwave freeze-drying was done by Copson
(1958), who assumed quasi-steady state, with only heat conduction and volumetric heat
generation occurring in both the frozen and dried regions. A general transient analysis, with
a moving sublimation front of heat and mass transfer in an infinite slab was investigated by
Ma and Peltre (19753). They assumed that heat transfer in the dried region is by internal
heat generation, conduction and convection, but that only volumetric heat generation and
conduction occur in the frozen region. Water vapor transport in the dried region was con-

sidered to fall in the transition region, with the mass diffusivity only dependent on pressure.

 

 

17

Ma and Peltre (1975a) also simulated a drying curve for beef from a numerical model
and compared it with experimental results. They concluded that the physical, thermal and
electromagnetic properties of beef influenced microwave volumetric heat generation, heat
and mass transfer. Therefore, more specific and comprehensive data on the properties of
various biological materials must be obtained before simulation models will accurately fit

real situations.

Ang et al. (1977a) incorporated the anisotropic character of beef muscle fiber and
developed two-dimensional heat and mass transfer models for microwave freeze-drying in
rectangular coordinates. Ma and Arsem (1982) observed entrainment of ice crystals and
frozen particles during rapid water vapor flow, and developed the concept of an effective
heat of sublimation for modeling radiant and microwave freeze-drying of heterogeneous
materials. Chang and Ma ( 1985) developed three-dimensional heat and mass transfer models
for analysis of microwave and radiant heat freeze-drying processes. Arsem and Ma ( 1990)
presented a more detailed mathematical model to describe combined microwave and radiant

freeze-drying.

The general models of heat and mass transfer in microwave freeze-drying are based

on energy and mass balances (Chang and Ma, 1985).

Energy balance in the dried layer:

3T
pded[a—td] = V - (deTd)+KdE2- CPWV(NwTd) (2.6)

Mass balance-in the dried region:

 

18

ac
5(5) = v . (DVC) (2.7)

Energy balance in the frozen region:

pfcp (337:1 ] = V - (kfvrf) + [<sz (2.8)
Because of the complex nature of the above equations, analytical solutions are not
easily obtained, requiring the use of dimensional analysis and numerical techniques (Ofoli
and Komolprasert, 1988). For example, Komolprasert and Ofoli (1987) used dimensional
analysis to obtain a predictive mathematical model for the microwave heating of water and
starch solutions. Dimensional analysis was also used to calculate the transient temperature
distribution and the position of the sublimation front for freeze-drying slabs, cylinders and

spheres, with a radiation boundary condition (Dyer and Sunderland, 1966; 1968; 1971).

2.5 Mechanism of volatile retention during freeze-drying

Several theories have been proposed to explain the mechanism of volatile retention in
freeze-dried products: adsorption, encapsulation, selective diffusion, microregion formation

and Delta T theory.

2.5.1 Adsorption

Rey and Bastien (1962) found that acetone retention in freeze-dried glucose solutions
was strongly influenced by the initial glucose content. They explained that volatile retention
depends on volatile adsorption on the dried layer of freeze-dried products, with glucose

playing the role of an adsorbent.

 

19

2.5.2 Encapsulation

Some volatiles can be trapped in miniature, sealed capsules to prevent volatile loss
during food processing (Dziezak, 1988). Materials such as protein or starch gels, malto-
dextrin, and cyclodextrin, can encapsulate volatile compounds by molecular interactions
such as hydrogen bonding, polar-polar interaction, and nonpolar-nonpolar interaction
between volatile molecules and solid components to provide higher volatile retention

(O’Keefe et al., 1991; Mills and Solms, 1984; Wyler and Solms, 1981).

Chung (1984) reported that higher dextrin equivalent (D.E.) maltodextrins led to higher
volatile retention, because they provided more binding sites. He explained that the inter-
actions may occur between hydroxyl groups of maltodextrin molecules and functional
groups of volatile compounds. The addition of gum can produce higher volatile retention,

because crosslinked networks may trap volatile compounds.

The interactions of flavor compounds with soy protein or whey protein were studied
by O’Keefe et al. (1991), and Mills and Solms (1984). Crowther et al. (1980) also reported
that denatured protein exposes its nonpolar regions, which may increase binding with
nonpolar flavor compounds. Wyler and Solms (1981) studied the inclusion complex for-
mation of pregelatinized potato starch with decanal, methone, and limonene volatile com-

pounds under different temperature conditions.

2.5.3 Selective Diffusion

The selective diffusion theory for volatile retention during drying was proposed by
Thijseen (1975). He defined the relative volatility of a volatile compound as the ratio of

the volatility of the compound to the volatility of water at the same temperature. Generally,

 

 

 

20

volatile compounds have high relative volatilities; once they diffuse to the evaporation or
sublimation surface, they quickly evaporate. Therefore, the selective diffusion of volatiles
and water to the sublimation surface during freeze-drying becomes the controlling and
limiting factor in volatile loss. The ratio of the diffusion coefficient of the volatile compound
to that of water (Dj/Dw) is not constant, and decreases with moisture content (Figure 2.6)
and temperature (Figure 2.7). It is also related to the molecular size of the volatile compound

and the pore size of the dried region (Thijseen, 1979).

Thijseen (1975) reported that volatile compounds can only escape during a critical
time period, t,, during freeze-drying. The critical time t, is defined as the time required for
drying a specimen from the initial moisture content at the sublimation front to the critical
moisture content. When the water vapor pressure in the pores of the dried region is low, or
the resistance to heat and water vapor diffusion is less, I, is short, leading to higher aroma
retention. Below the critical moisture content, only water can diffuse to the surface of the
product because Dj/Dw is less than 1, and residual volatile compounds remain in the spec-

imen.

2.5.4 Microregion formation

The microregion theory for volatile retention was proposed by Flink and Karel (1970a,
1970b). During freezing, the unfrozen solution between the growing ice crystals will be
concentrated by crystallization of the water, and the volatile compounds will be pushed by
the growing ice crystals and trapped by molecular interaction in the concentrated carbo-
hydrate to form microregions (Flink, 1975b). Therefore, freezing conditions are very

important in determining the properties of the microregions: slow freezing results in fewer,

 

21

 

 

 

   
  
  
 

100 ,
a :
W .
‘ AI hl ' It -d t‘
'10”? coosmrnao exnn
e — -—methanol
'2 I—
10 5 —-propanol
: —pentanol
10’3 :-
)-
10_4 r r r I 4 1 L J J

 

 

 

010 20 30 4O 50 60 70 80 90100
——-’ wt‘tcwater

Figure 2.6 Effect of water concentration on the diffusion coefficient of alcohols
relative to water in maltodextrin at 25°C. (Source: Thijseen. 1979)

 

10°

1 V'IY

V V V I'V'I'

10'? .-

propenol in malto-dextrin

   

10'3 r

 

 

. .1 J 1 L J J l l
o 0'"

0102030405060708090100
___.pwt%water

 

Figure 2.7 Temperature-water concentration dependence of the diffusion
coefficient of propanol relative to water in maltodextrin. (Source: Thijseen, 1979)

 

22

larger, and more concentrated microregions, with a lower permeability that improves volatile
retention. Moreover, there is no loss of volatile compounds from the microregion as long

as it is in the frozen region (Flink, 1975b; Flink and Karel, 1970b).

During freeze—drying, volatiles can diffuse from a microregion just within the frozen
zone, through the ice front, and up to the dried region. When volatiles pass the ice interface
through the microregion, volatile loss occurs continuously until the critical moisture content
is reached; at this time the microregion is sealed, and volatile loss ceases. However, water
loss may still occur after reaching the critical moisture content, because of the plasticizing

action of water and the small size of the water molecules (Flink and Karel, 1970b).

2.5.5 Delta T theory

Shaath and A220 (1989) developed the Delta T theory for microwave heating of volatile
compounds. This theory may represent another volatile retention mechanism during
microwave freeze-drying. They defined AT as the ratio of the temperature increase in the
volatile compound to the temperature increase in water at the same microwave power level.
They found that the more polar alcohols generally have a higher AT than do the correspondin g
aldehydes. The AT values of some volatiles, along with specific heat and dielectric constants,

are given in Table 2.1.

2.6 Factors influencing volatile retention of freeze-dried foods

Volatile retention during freeze-drying is influenced by solid composition, concen-
tration, initial volatile concentration, freezing rate, drying rate, sample dimensions, frozen
layer temperature, structural collapse, and heat input (Flink, 1975a; Gero and Smyrl, 1982;
Gerschenson et al., 1979).

 

23

Table 2.1 AT values, specific heat and dielectric constants of selected volatiles.

(Source: Shaath and A220, 1989)

 

 

 

 

 

 

 

 

Chemical AT Value Cp 8’
C-2 alcohol 1.73 0.59 24
C-6 alcohol 1.13 0.59 13
C-8 alcohol 0.97 0.59 10
C-10 alcohol 0.75 0.54 8
Benzyl alcohol 1 .81 0.44 13
Carvone 2.38 0.40 15
Water 1.00 1.00 78

 

 

 

 

 

 

24

2.6.1 Solid composition and concentration

An increase in carbohydrate concentration results in increased volatile retention
(Ofcarcik and Burks, 1974; Gero and Smyrl, 1982 ). Food composition influences the
molecular interaction with volatiles. Low molecular weight carbohydrates, such as sucrose,
glucose, and lactose, are more effective for volatile retention than polymeric substances like
polyvinylpyrrolidone (PVP), starch and cellulose, which have less available binding sites

(Bartholomai et al., 1975; Chirife et al., 1973; Chung, 1984).

The emulsified oil phase in a food system can extract some nonpolar volatiles or
increase volatile retention by hindering the diffusion of volatile molecules to the evaporating
surface (Etzel and King, 1980; Chung, 1984). Chung (1984) also indicated that gum can
be added to increase the viscosity of the solution and to form a network which can trap
volatile molecules to improve retention. In addition, different food compositions have
different dielectric loss factors and will influence microwave power absorption and drying

time during freeze-drying.

Sample dimensions also affect drying time. Thick specimens have a longer drying
time, resulting in high aroma loss (Thijseen, 1975; Flink, 1975a). Thin specimens usually

get uniform heating, with reductions in mass transfer resistance and drying time.

2.6.2 Cyclodextrin inclusion complexes

Cyclodextrin can be used as an encapsulating medium for volatile compounds. a-, B-

and y- cyclodextrins are cyclic oligosaccharides consisting of six, seven and eight glucose

molecules, respectively, linked by a (1-4)-g1ycosidic bonds. Cyclodextrin has a relatively

 

25

rigid, doughnut-shaped ring structure, and a hollow cavity of a specific volume. The
chemical structure of B-cyclodextrin is shown in Figure 2.8. The properties of a-, 13-, and

'y- cyclodextrins are given in Table 2.2.

The specific structure of cyclodextrin includes a nonpolar cavity and a polar group on
the outer surface of the cavity. Therefore, nonpolar guest molecules of suitable size and
shape can be encapsulated into the cavity, where a favorable nonpolar-nonpolar interaction
is established. This forms inclusion complexes which promote flavor retention during food
processing. For example, B-cyclodextrin is able to form inclusion complexes with flavor

compounds of molecular weights between 80 and 250 (Pszczoal, 1988).

Cyclodextrin inclusion complexes can protect guest compounds from oxidation, light
induced reaction, thermal decomposition, and loss due to evaporation or sublimation. They
can be used to control volatility and retain the potency of spices and flavors, and eliminate
undesired tastes and odors. They can also be used to modify the taste, aroma, texture, and
color of guest compounds, and therefore offer a variety of potential uses for the food and

non-food industries (Pszczoal, 1988).

2.6.3 Freezing rate

A low freezing rate increases the size of ice crystals and pore diameter, and promotes
the formation of fewer, but larger microregions. Flink and Gejl-Habsen (1972) predicted
that larger microregions lead to a lower permeability of volatile compounds in comparison

with ice sublimation, and improves volatile retention.

26

 

 

Figure 2.8 Chemical formula and molecular shape of B-cyclodexhin.
(Source: Pszczoal, 1988)

27

Table 2.2 The properties of cyclodextrins (CD).
(Source: Pszczoal, 1988)

 

 

 

 

 

 

 

 

 

 

 

 

Molecular Weight 973 1136 1297
Inside Diameter (Ii) ~ 5.7 ~ 7.3 ~ 9.5
Diameter (.51) ~ 13.7 ~ 15.3 ~ 16.9
Height (Ii) ~ 7.0 ~ 7.0 ~ 7.0
Water Solubility (gllOOrnl) 25°C 14.50 1.85 23.20
[ot]3,o (H20, 1%) 150.5° 162.5° 117.4°
__,_._-‘,,,,z »~ — ,,,, -__, ~— —-—- -— bra—anal:

   

28

Slow freezing of pectin, dextrin and corn syrup solutions before freeze-drying shows
significantly greater amounts of volatiles than fast freezing (Flink and Gejl-Hansen, 1972;
McPeak et al., 1987; Gero and Smyrl, 1982; Thijseen, 1975). However, McPeak et al. (1987)
reported that there was no significant difference in volatile retention between gelatin gel
solutions that were frozen rapidly and slowly, because the formation of a gel can entrap
volatiles within the solids matrix before microregion formation during freezing. Chirife
and Karel (1973) also reported that alcohol retention in rapidly frozen, freeze-dried PVP

solutions is due to both entrapment in microregions and adsorption.

2.6.4 Chamber pressure

A decrease in chamber pressure results in improved moisture transfer, decreasing ice
front temperature and leading to better volatile retention (Thijseen 1975). Petersen et al.
(1973) showed that volatile retention is hardly influenced by chamber pressures between
0.3 and 0.7 torr, but a large decrease in volatile retention occurs at a chamber pressure of
0.8 torr. They explained that the ice front temperature of coffee extracts, at a chamber
pressure of 0.8 torr, is higher than the temperature which causes the freeze-dried matrix to

collapse, and results in substantial loss of aroma (Flink, 1975b).

Chung (1984) also explained that a high vapor pressure at the center of the specimen
may cause cracking of the dried region and the release of trapped volatiles in microwave-
aided systems. During microwave freeze-drying, water vapor is quickly sublimated by
absorption of microwave energy, immediately increasing chamber pressure and ice front

temperature. Therefore, the control of chamber pressure is a very important factor.

29

2.7 Kinetics of volatile retention

Flavor compounds are generally found in low concentrations within a given food
product; however, flavor characteristics contribute greatly to the overall acceptance of the
food product. Flavor retention is often reported as a percentage of the original content after
food processing. There is little information on the changes that take place during the unsteady
thermal process. The kinetic parameters of flavor stability have received little attention;
actual kinetic data for flavor retention and development during thermal processing of real

food systems are virtually non-existent (Villota and Hawkes, 1986).

Chung (1984) studied the kinetics of ethanol, acetone and ethyl acetate retention in
model food systems at constant temperatures of 35, 60, 80 and 95°C during microwave and
conventional heating. He reported that volatile retention is a first order reaction, and that
temperature dependent rate constants follow the Arrhenius relationship. He also compared
volatile retention of model food systems during both microwave heating and conventional
heating, and found that volatile retention in microwave heating was higher than that in
conventional heating. He explained that this is because microwaves generate heat within
the sample itself, and the surrounding atmosphere can be kept cool without a heat transfer

medium in order to obtain better volatile retention.

2.8 Microstructure of freeze-dried products

Scanning electron microscopy (SEM) can be used to observe the microstructure of
food products as it is changed by food processing. The microstructure may assist in inter-
preting volatile retention in freeze-dried products; it also helps to determine the bulk density

and porosity of the product (Rosenberg, et al. 1988). Freezing rates can control ice crystal

30

size, and affect the microstructure of freeze-dried specimens. Generally, fast freezing causes
minimum cell damage and minimum isolation of tissue components (deMan, 1986 and Bello
et al, 1982). Slower freezing leads to more damage in microstructure (Li-Shing-Tat and

Jelen, 1987; Woodward and Cotterill, 1985; Bello et al., 1982).

Flink and Gejl-Hansen (1972) investigated the effects of freezing rate on the micro-
structure of a freeze-dried maltodextrin-hexanal system. They observed that there were
smaller droplets of volatiles in the more quickly frozen samples, corresponding to the
formation of smaller microregions during quick freezing. Arsem and Ma (1985) studied
the microstructures of both microwave and conventionally freeze-dried steaks, and found
that there is a noticeable difference in the texture of muscle fibers. The microwave
freeze-dried steak had a broken and open structure, due to the rapid expansion of moisture
vapor which also sweeps the channels clear of all materials. The conventional freeze-dried

steak had a sealed surface and hardened fibers due to the high temperature treatment.

CHAPTER 3
MICROSTRUCTURE AND VOLATILE RETENTION OF
CONVENTIONALLY FREEZE-DRIED MODEL FOODS

3.1 Introduction

Freeze-drying is a sublimation process that transforms ice to water vapor without
passing through the liquid phase. Since it is a low temperature process, it can minimize
thermal degradation, reduce shrinkage of foods, and provide for better flavor retention and
product quality. Several theories have been advanced to explain the mechanisms of volatile
retention in freeze-dried products: adsorption, encapsulation, selective diffusion, and
microregion formation. Rey and Bastien (1962) found that acetone retention in freeze-dried
glucose solutions was strongly influenced by the initial glucose content. They explained
that volatile retention depends on volatile adsorption on the dried foods, with glucose playing

the role of an adsorbent.

Some volatile compounds can be encapsulated in miniature, sealed capsules by
molecular interaction to prevent loss (Dziezak, 1988). For example, cyclodextrins can be
used as encapsulating media for volatile compounds, because nonpolar guest molecules of
suitable size and shape can be incorporated into the cavity to establish a favorable
nonpolar-nonpolar interaction to promote flavor retention during food processing (Pszczoal,
1988). Materials such as protein or starch gel can also promote volatile retention by
molecular interactions such as hydrogen bonding, polar-polar interaction and nonpolar-
nonpolar interaction between volatile molecules and solid components. The interaction of

flavor compounds with soy protein or whey protein has been studied by O’ Keefe et al. ( 1991)
3 l

32

and Mills and Solms (1984). Wyler and Solrns (1981) have also studied the inclusion
complex formation of pregelatinized potato starch with decanal, methone and limonene

under different temperahrre conditions.

The selective diffusion theory for volatile retention during drying was proposed by
Thij seen (1 975). Once volatile compounds diffuse to the evaporation or sublimation surface,
they quickly evaporate due to their high volatility. Therefore, selective diffusion of volatiles
and water to the sublimation surface becomes the conh'olling and limiting factor in volatile
loss during freeze-drying. However, the ratio of the diffusion coefficient of the volatile
compound to that of water is not constant, and decreases with decreasing moisture content
and temperature (Thijseen, 1979). The ratio is also related to the molecular size of the
volatile compound as well as the pore size dishibution in the dried region.

Another theory proposed for volatile retention is associated with microregion for-
mation, and was investigated by Flink and Karel (1970a, 1970b), and Flink (1975a). During
freezing, the volatile compounds are pushed by the growing ice crystals and tapped by
molecular interaction in the concentrated carbohydrate to form microregions. There is no
loss of volatiles from the microregion as long as it is in the frozen core. During freeze-drying,
volatile compounds pass the ice front and continuously evaporate until the critical moisture
content is reached. At this time, the microregion is sealed and volatile loss ceases. According
to this theory, slow freezing improves volatile retention, because it results in fewer, larger,

and more concentrated microregions with lower permeability.

The factors that influence volatile retention of freeze-dried foods include composition,
concenh'ation of solids, initial volatile concentration, freezing rate, drying rate, sample

dimensions, frozencore temperature, struchrral collapse, and rate of heat input (Flink, 1975b;

 

from

33

Gero and Smyrl, 1982; Gerschenson et al., 1979). Slow freezing of pectin, dextrin and corn
syrup solutions before freeze-drying showed significantly higher volatile retention than fast
freezing (Flink and Gejl-Hansen, 1972; McPeak et al., 1987; Gero and Smyrl, 1982; Thij seen,
1975). However, McPeak et al. (1987) showed that there was no significant difference in

volatile retention between gelatin solutions frozen rapidly and slowly.

An increase in carbohydrate concentration results in increasing volatile retention
(Ofcarcik and Burks, 1974; Gero and Smyrl, 1982). Low molecular weight carbohydrates
such as sucrose, glucose, and lactose are more effective for volatile retention than polymeric
substances like polyvinylpyrrolidone (PVP), starch and cellulose, which provide fewer
binding sites (Bartholomai et al., 1975; Chirife et al., 1973). Also, a thick specimen causes
higher volatile loss because of the larger heat transfer resistance in the dried region, which
delays heat transport to the frozen core and increases drying time (Thijseen, 1975; Flink,

1975b).

Chamber pressure also affects volatile retention during freeze-drying. As is evident

from the Clausius-Clapeyron equation:

15-“):
P 'er

 

dT (3.1)

a lower chamber pressure corresponds to a lower temperature in the frozen core (Dyer et
al., 1966). Thijseen (1975) also suggested that decreasing chamber pressure results in

decreasing moisture vapor transfer resistance, and leads to better aroma retention.

Scanning electron microscopy (SEM) can be used to provide information on the

microstructure of freeze-dried food products. The microstructure can assist in interpreting

34

volatile retention, and may be used to determine product bulk density and porosity
(Rosenberg et al., 1988). Freezing temperature affects the freezing rate and the size of ice
crystals; thus, it also influences the pore size distribution and microstructure of freeze-dried

products. Generally, slow freezing rates lead to larger ice crystals, resulting in bigger pores.

The objectives of this project were to (1) study the retention of selected volatile
compounds in freeze-dried model foods, and (2) assess the effect of freezing temperature

and sample composition on the microstructure of freeze-dried specimens.

3.2 Materials and Methods

3.2.1 Preparation of starch gel model foods

The model foods were prepared by mixing pregelatinized corn starch solutions (8%
solids) with 50 ppm each of l-hexanol, l-decanol and limonene. The resulting gels were
subjected to five treatments: control or null treatment, addition of 2% or 4% B-cyclodextrin
(CD), and addition of 2% or 4% sucrose. Three volatile compounds were dissolved in
2-methylbutane to prepare a 500 ppm volatile stock solution. The samplesuvmre homoge-
nized and 10 g of each sample was poured into an aluminum pan. Samples were then frozen

at -15°C, -40°C or -198°C (in liquid nitrogen) before freeze-drying.

Samples were freeze-dried for about 24 hours in a Virtis freeze-drier with a conducting
shelf (Virtis Company, Gardiner, NY). The shelf temperature and condenser temperature

were maintained at 10°C and -60°C, respectively.

Pregelatinized corn starch and B—cyclodextrin were obtained from American Maize

Products Co. (Hammomd, Indiana). Limonene, l-hexanol, l-octanol and l-decanol were

35

obtained from the Sigma Chemical Co. (St. Louis, MO). 2-Methylbutane was obtained from
the Aldrich Chemical Co. (Milwaukee, WI). Sucrose and anhydrous sodium sulfate were
obtained from the J .T. Baker Chemical Co. (Philipsburg, NJ).

3.2.2 Volatile retention analysis

Two replications of five sample treatments frozen at three temperatures were used to
study effect of sample composition, freezing temperature on retention of each volatile in

conventionally freeze-dried model foods.

Before volatile distillation and extraction, an internal standard of 1 ml 400 ppm
l-octanol solution and 150 ml distilled water were added to each freeze-dried sample. The
volatile compounds were extracted by 10 ml of 2-methylbutane for 1 hour in a modified
Likens-Nickerson apparatus. Trace water in the extracted volatile solution was removed
by anhydrous sodium sulfate, and the extracted volatile solution was concentrated to 3 ml

by flashing nitrogen gas through it before gas chromatography analysis.

The 0.5 ul of extracted volatile solution was injected into a Hewlett-Packard 5890 gas

chromatograph (GC) equipped with a flame ionization detector. The separation of volatile
compounds was achieved in a fused silica capillary column (30m x 0.32 mm ID.) with
SupelcowaxTM 10 stationary phase. The oven temperature was held at 40°C for 5 min and
increased at a rate of 5°C/min to 165°C. The injector and detector temperatures were
maintained at 220°C and 230°C, respectively. The carrier gas was helium at a flow rate of
30 mllmin. The peak area of each volatile compound was recorded on a Hewlett-Packard
3390A integrator. The peak area response of each volatile compound relative to the internal

standard was obtained by injection of standard solutions. The retention level of each volatile

36

was calculated as a percent of the volatile concentration of the frozen specimen.

3.2.3 Microstructure analysis

Each freeze-dried specimen was cut to a 5 mm x 5 mm piece, mounted on an aluminum
stub (Electron Microscopy Sciences, Ft. Washington, PA) using epoxy glue, and coated
with a thin layer of gold in a sputter coater (Emscope model SC 500, Kent, England).
Scanning electron micrographs were obtained by a model JSM-35C scanning electron
microscope (JEOL, Osaka, Japan) at an accelerating voltage of 15 kV and condenser lens

setting of 400.

3.2.4 Statistical analysis

Volatile retention data were analyzed with a MSTATC statistics software package
(Version 4.0, Michigan State University, E. Lansing, MI). Two replications were carried
out for all volatile retention analyses. Statistical significance of the degree of variation was
determined by a three-factor (sample treatment x freezing temperature x volatile compound)
completely randomized design experiment of analysis of variance (AN OVA); Tukey’s test
and standard error of the means were used to evaluate the significance of differences between

the means of treatments at P S 0.05.

3.3 Results and Discussion

3.3.1 Volatile retention

The retention of volatile compounds in samples frozen at three temperatures is shown

in Table 3.1. The analysis of variance is presented in Table 3.2. The most significant factors

37

in volatile retention were composition of the sample and the type of volatile compound.
The retention of individual volatiles in this study was not significantly influenced by the
freezing temperature (Figure 3.1). McPeak et al. (1987) also reported that freezing rate did
not significantly influence volatile retention in gelatin solutions. They explained that the
volatile compounds were probably trapped within a gel matrix before the formation of
microregions during the freezing process, so retention was not significantly influenced by

the different freezing temperatures.

The retention of volatile compounds was significantly influenced by sample treatment
(Figure 3.2). In general, the retention of the two alcohols (l-hexanol and l-decanol) sig-
nificantly increased with molecular weight. This may be the result of selective diffusion,
whereby diffusion of volatiles from the concentrated amorphous region to the sublimation
surface becomes a controlling factor in volatile loss. Since the diffusion of each volatile
compound is governed by Fick’s law, the smaller molecular weight or less soluble volatile
compounds have larger diffusivities and therefore diffuse faster to the evaporation front

(Gero and Smyrl, 1982; Thijseen, 1979).

Bartholomai et al. (1975) and Chirife et al. (197 3) reported that low molecular weight
carbohydrates have more available binding sites and were more effective for volatile
retention than polymeric compounds. However, in this study, the addition of low molecular
weight sucrose did not significantly influence volatile retention (Figure 3.2). This was
probably due to the fact that 8% pregelatinized corn starch already provides adequate binding

sites for the 50 ppm volatile compounds, making additional sites unnecessary.

38

Table 3.1 Volatile retention of conventionally freeze-dried model foods.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

Sample m 1~Decanol

Treatments Temp.(°C) % % %

(1) 8% starch ~15 2.2 63.6 85.6

~40 0 52.5 76.8

~198 0.9 48.2 87.6

(2) 8% starch with ~15 87.3 85.7 89.1

2% cyclodextrin ~40 76.7 72.3 93.0

~198 80.4 79.3 88.3

(3) 8% starch with ~15 90.7 76.8 97.6

4% cyclodextrin ~40 99.9 91.6 92.5

~198 95.1 89.1 86.2

(4) 8% starch with ~15 2.4 57.4 77.1

2% sucrose ~40 0.3 65.8 74.1

~198 0.8 57.0 81.1

(5) 8% starch with ~15 4.4 57.5 68.8

4% sucrose ~40 0.8 63.2 85.1

0 57.7 75.0

  

 

 

   

39

Table 3.2 ANOVA table showing the effect of type of volatile, sample composition

and freezing temperature on volatile retention.

 

!

 

 

 

 

 

 

 

 

 

 

 

 

 

Source Degrees of Sum of Mean F Probability I
» Freedom N _ Squares 7 Square Value ___ V _ J
Factor A 4 37134. 3 9283.6 21.8 68 0.000 I
FactorB 2 31.8 15.9 0.37 l
FactorC 2 35389.3 17694.6 416.81 0.000
AB 8 542.3 67.8 1.60 0.153 1
AC 8 23754.5 2969.3 69.94 0.000
BC 4 38.6 9.7 0.23 1
ABC 16 1046.3 65.4 1.54 0.127
Error 1910. 4 42. 5 l

-“--—

Key: Factor A = sample composition; Factor B = freezing temperature; Factor C = type
of volatile compound

 

 

 

.
.
.
.
.

 

g Secozom

 

~198 C

C
Temperature

~15C

Ing

Freez

1 -Decenol

Hexenol

m Limonene 8&3 1

Figure 3.1 Volatile retention in freeze—dried model foods frozen at three different temper-

atures.

41

 

 

 

 

 

at eoeeaom

Sample Treatment

I Limonene

1-Hexanol 1-Deeenol

Figure 3.2 Effect of sample treatment on volatile retention.

42

The addition of B-cyclodextrin significantly increased volatile retention for all compounds

(Figure 3.2). This is due to the fact that B-cyclodextrin can encapsulate suitably sized volatile
compounds, enhancing retention. Except for 1-decanol, volatile retention in samples con~
taining 4% B-cyclodextrin is significantly higher than that of samples containing 2%
B-cyclodextrin. Limonene retention in samples containing B-cyclodextrin was significantly
higher than that in all other treatments, each of which had less than 5% limonene retention
(Table 3.1, Figure 3.2). This is probably because the nonpolar ring structure of the limonene

molecule fits within the cavity of B-cyclodextrin to form an inclusion complex (Chang and

Reineccius, 1990), leading to high retention.

The retention of l-hexanol and 1~decanol in samples containing B-cyclodextrin was

also significantly higher than that in other treatments. The diameter of the B-cyclodextrin
cavity is 7.8 A (Pszczoal, 1988). Using the averagelength of a carbon-carbon chain, the
length of 1~hexanol and l-decanol can be estimated at about 9 A and 15 A , respectively.
Since these are linear molecules, some part of the nonpolar structure can be included in the

B—cyclodextrin cavity to improve retention.

3.3.2 Microstructure

As expected, freezing temperature was a significant factor in the pore size of all
freeze-dried samples. For the control (8% pregelatinized corn starch), the pore shape at a
freezing temperature of ~15°C was elliptical with an average size of 600 um x 300 um
(Figures 3.3A and 3.3B). Pores of samples frozen in liquid nitrogen were nearly spherical

(Figures 3.3C and 3.3D) with an average diameter of 2.5 um. The starch gel structure

43

Figure 3.3 Scanning electron micrographs of the freeze-dried control samples frozen at
different temperatures. (A), (B): ~15°C, and (C), (D): ~198°C.

 

44

)l samplt‘ lg

 

45

appeared to be that of a three dimensional network, with a smooth surface.

The rnicrostructure of samples containing B-cyclodextrin frozen at ~15°C is shown in

Figure 3.4. The starch gel structure was covered by aggregated B-cyclodextrin which
appeared to block the pores. Also, some small encapsulated volatile droplets appeared to
be embedded on the aggregated B-cyclodextrin structure. The aggregated B-cyclodextrin
not only blocked the original hole in the starch gel structure, but also appears to have
encapsulated the volatile dr0plets to prevent loss. The microstructure of samples containing
sucrose and frozen at ~15°C is shown in Figure 3.5. The filamentous network structures
observable in Figures 3.5A and 3.5C were thinner than those in Figure 3.3A, and flakes

were visible above the web-like fibrous gel structure, which also contained many large pores.

The microstructure of three different freeze-dried samples frozen at ~198°C (in liquid

nitrogen) is shown in Figure 3.6. As already discussed, quick freezing causes the formation
of very small ice crystals. The freeze-dried control had many small pores within a three
dimensional gel network structure with a smooth surface (Figure 3.6A). This gel matrix
also contained numerous tiny globular lumps, which may incorporate volatiles. The sample
containing 4% B-cyclodextrin frozen at ~198°C is shown in Figure 3.68. The volatile
compounds were encapsulated in B-cyclodextrin spherical droplets with crude surfaces.
This structure prevented volatile loss during freeze-drying to obtain significantly higher
volatile retention than other sample treatments (Table 3.1). A flaky structure and large
number of small pores are evident in samples containing 4% sucrose (Figure 3.6C). From
these micrographs, it is clear that the type of additive significantly affects freeze-dried starch
gel structure. This fact helps explain the different levels of volatile retention observed in

this study.

46

Figure 3.4 Scanning electron micrographs of samples containing 2% and 4% B-cyclodextrin
and frozen at ~15°C. (A), (B): 2% B-cyclodextrin, and (C), (D): 4% B~cyclodextrin.

 

ll ’1

 

48

Figure 3.5 Scanning electron micrographs of samples containing 2% and 4% sucrose and
frozen at ~15°C. (A), (B): 2% sucrose, and (C), (D): 4% sucrose.

 

40

 

50

Figure 3.6 Scanning electron micrographs of three model foods frozen at ~198°C. (A): 8%
pregelatinized corn starch (control), (B) 8% pregelatinized corn starch with 4% B—cyclo—
dextrin, (C) 8% pregelatinized corn starch with 4% sucrose.

 

 

 

 

 

 

 

52

3.4 Conclusions
Volatile retention significantly increased in samples containing B—cyclodextrin,

apparently due to the formation of inclusion complexes. Encapsulation in cyclodextrin
inclusion complexes appears to provide an important volatile retention mechanism for
freeze-dried foods, and should be studied further. The presence of sucrose did not signif-
icantly influence volatile retention. Retention of the two alcohols increased with molecular
weight, and appears to be based on selective diffusion. Freezing temperature had no effect

on volatile retention, although it significantly influenced the pore size distribution of the

freeze-dried samples.

 

CHAPTER 4
VOLATILE RETENTION OF MICROWAVE FREEZE-DRIED
MODEL FOODS

4.1 Introduction

Since the thermal conductivity of the dried region of most foods is low, a large thermal
resistance is encountered in conventional freeze-drying processes, leading to longer drying
times. Copson (1958) suggested that microwave freeze-drying can be used to overcome
the heat transfer limitations of conventional freeze-drying. Since microwave energy is a
volumetric heat source, it penetrates the dried layer to reach the frozen region of the sample.
In addition, the dielectric loss factor of the frozen region is much larger than that of the dried
region; therefore, most of the microwave energy is absorbed by the frozen region to induce
sublimation of ice. As a result, microwave freeze-drying can increase the drying rate by a
factor of 3 to 13 times the conventional freeze-drying rate (Ma and Peltre, 1975b). Sun-

derland (1982) investigated the cost of freeze-drying by microwave energy, and reported

that this method is also cheaper than the conventional method.

The factors which influence microwave freeze-drying, such as electric field strength,
Chamber pressure, and sample dimension, have been studied by Ma and Peltre (1975b) and
Ang et al. (1977b). Although increasing the electric field strength can increase the drying
rate, a field strength that is too high can cause overheating or melting of the sample (Ang

et al., 1977b). Ang et al. (1978) adjusted the electric field strength from 170 V/cm to 110
V/cm to avoid overheating and to maintain product quality during the freeze-drying process.

53

”infirlfll‘f w

54

Ma and Peltre (1977b) studied the effect of chamber pressures between 0.02 and 3.5
torr on the drying rate during microwave freeze-drying. They reported that melting and
vaporization occurred at 3.5 torr, and suggested that low pressure operation can help avoid
overheating or melting of the sample, and corona discharge. In general, a lower chamber
pressure corresponds to a lower temperature in the frozen core (Dyer et al., 1966). Thij seen
(1975) suggested that decreasing chamber pressure results in decreasing moisture vapor
transfer resistance and leads to better volatile retention. Petersen et al. (1973) showed that
coffee extracts had less volatile retention at a chamber pressure of 0.8 torr than 0.2 torr,

because the higher ice front temperature caused the freeze-dried matrix structure to collapse.

However, there is a limit to how low the chamber pressure can be. Entrainment of
hydrated salts and water soluble proteins has been observed at 0.03 torr during microwave
freeze-drying of beef (Arsem and Ma, 1985). Under 0.1 torr and low temperature conditions,
the expansion of ice to water vapor upon sublimation is large enough to generate a high
Velocity. The resulting vapor expansion is capable of entraining a significant amount of

solids which are then lost to the product being processed.

Hill and Sunderland (1971) analyzed the influence of chamber pressure between 0.01
and 4 torr on the drying rate of beef during conventional freeze-drying. The slower drying
rates occurred at a chamber pressure of 0.01 and 0.1 torr, because the mean free path of gas
molecules is larger than the pore diameter, and the gas molecules collide with the wall and

decrease the mass flux. They also reported that water vapor transport falls in the transition
I‘egion at chamber pressures between 0.5 and 4 torr, and that the fastest freeze-drying rate

Occurs at a chamber pressure of 1 torr.

55

Most of the microwave freeze-drying research has focused on the development of heat

and mass transfer models (Ma and Peltre, 1975a; Ang et al., 1977a, Chang and Ma, 1985;

Arsem and Ma, 1990). The quality of microwave freeze-dried products has received rela-

tively little attention. Arsem and Ma (1985) evaluated the microstructure and flavor of both

conventional and microwave freeze-dried steaks, and reported that there was a noticeable

difference in the texture and microstructure of muscle fibers, but there was no significant

difference in sensory characteristics. Cohen and Yang (1992) studied the rehydration of

microwave and conventional freeze-dried green peas, and concluded that microwave
freeze-dried green peas had a larger rehydration ratio than conventional freeze-dried green
peas. However, there was no significant difference in sensory panel scores between these

two methods of freeze-drying.

The factors which influence volatile retention in conventional freeze-dried samples
have been extensively studied. They include solid composition and concentration, freezing
rate, drying rate, sample dimensions, frozen core temperature, structural collapse, and rate
0f heat input (Flink, 1975a; Gero and Smyrl, 1982; Gerschenson et al., 1979). An increase
in carbohydrate concentration results in increasing volatile retention (Ofcarcik and Burks,
1 974; Gero and Smyrl, 1982). Also low molecular weight carbohydrates are more effective
for volatile retention than polymeric substances like starch and cellulose, which provide
fewer available binding sites (Bartholomai et al., 1975; Chirife et al., 1973). Materials such
as protein or starch gel, and cyclodextrin can encapsulate volatile compounds by molecular

interactions to provide high retention (O’Keefe et al., 1991; Mills and Solms, 1984; Wyler
and Solms, 1981).

Slow freezing of pectin, dextrin and corn syrup solutions before freeze-drying shows

Significantly higher volatile retention levels than rapid freezing (Flink and Gejl-Hansen,

56

1972; McPeak et al., 1987; Gero and Smyrl, 1982; Thijseen, 1975). However, there was
no significant difference in volatile retention between gelatin and starch gel solutions frozen
rapidly and slowly (McPeak et al., 1987; Chapter 3). Sunderland (1980) suggested that
freezing temperature affects the pore size of freeze-dried samples, and may influence water
vapor transfer during microwave freeze-drying. However, no comparable studies have been

conducted on volatile retention of microwave freeze-dried products.

The objectives of this work were to:
(1) study the effect of sample composition on the drying rates and volatile retention during
microwave freeze-drying,

and

(2) investigate the effect of freezing temperature and chamber pressure on volatile retention

of microwave freeze-dried samples.

4.2 Materials and Methods

4.2.1 Preparation of the model foods

Pregelatinized corn starch solutions of 8% solids were mixed with 500 ppm each of
1~hexanol, l-decanol and limonene to form a gel. Samples were then subjected to one of
three treatments: (1) control or null treatment, (2) addition of 4% B—cyclodextrin (CD), and
(3) addition of 4% sucrose. Each sample was homogenized and poured into a cylindrical

mold 2.5 cm high and 1.5 cm in diameter. Before microwave freeze-drying, samples were

frozen at ~15°C, ~60°C or ~198°C (in liquid nitrogen).

57

Two replications of three sample treatments frozen at three temperatures and freeze-
dried at two chamber pressure were used to study effect of sample composition, freezing
temperature and chamber pressure on final moisture content and retention of each volatile

of microwave freeze-dried samples.

Pregelatinized corn starch and B-cyclodextrin were obtained from American Maize

Products Co. (Hammomd, Indiana). Limonene, 1-hexanol, 1~octanol and l~decanol were
obtained from the Sigma Chemical Co. (St. Louis, MO). 2-Methylbutane was obtained from
the Aldrich Chemical Co. (Milwaukee, WI). Sucrose and anhydrous sodium sulfate were
obtained from the J .T. Baker Chemical Co. (Philipsburg, NJ).

4.2.2 The microwave freeze-drying system

The microwave freeze-drying system consists of eight basic components (Figure 4.1):
a tunable microwave cavity, a microwave generator with an external circuit, a temperature
measurement system, a vacuum system with a pressure gauge, nitrogen gas with a needle

Valve and flow meter, a condenser, a platform for the frozen sample, and a data acquisition

System.

4.2.2.1 Single mode internally tunable resonance cavity applicator

A cylindrical tunable brass cavity (17.8 cm o.d., 23.6 cm height, and 0.3175 cm thick)

With an internal transverse brass circular shorting plate (Figure 4.2) was used for all
experiments. The shorting plate inside the cavity is adjustable to provide a variable cavity
length from 6.0 cm to 19.0 cm, corresponding to different field modes. Two cylindrical

brass tubes (1.372 cm id and 7.62 cm long) are located 3.81 cm above the surface of the

bottom plate and are perpendicular to each other. A semirigid 50 ohm impedance copper

58

 

microwave power input

Figure 4.1 Experimental setup.

Components: 1. microwave cavity, 2. sample, 3. pirani gauge, 4. controller, 5. condenser,
6. desiccant trap, 7. mist filter, 8. vacuum pump, 9. nitrogen tank, 10. flow meter, 11. data
acquisition system, 12. Luxtron fluoroptic thermometer. '

59

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1
f 1211
1 .1/
.. 1o _
LT
2
l J F 7 l
4 I 1
8
L” Le
___a 141 -3
J.
16 7

 

 

 

—52‘1_‘ 15 3 if 311*

to vacuum system é—I:

%

9

 

 

 

 

 

Figure 4.2 Close-up view of the cylindrical tunable nricrowave cavity.

Components: 1. conducting cylindrical shell, 2. sliding short, 3. end plate, 4. silver finger
stock, 5. coaitial input port, 6. adjustable coupling probe, 7. screen viewing port, 8. quartz
tube, 9. rubber stopper, 10. aluminum cover with O—ring, 11. compression gland feed-
through, 12. Luxtron MIW temperahue probe, 13. nitrogen gas inlet port, 14. sample, 15.
Teflon sample holder.

coaxial probe serves as a field excitation probe (a coupling probe), and has an outer conductor
of 1.27 cm diameter and an inner conductor of 0.44 cm diameter. The coupling probe
provides a variable inner-conductor depth from 0.0 to 3.4 cm to couple microwave power
into the cavity. Adjustments of the cavity length (LC) and coupling probe depth (LP) are

made manually to within 0.1 m, using micrometer indicators.

4.2.2.2 The microwave generation system

The microwave system utilizes an external circuit and is shown in Figure 4.3. The
microwave generator (Model MPG-4M, Opthos Instruments, lNC., Rockville, Maryland)
can provide variable power from 1 to 100 W at a single frequency of 2.45 GHz. In this
study, the microwave power input level was kept constant at 10 W. The cylindrical frozen
samples were excited in a TM.)12 mode cavity (Asmussen et al., 1987). To accommodate
the reduction in sample moisture content with drying time, the cavity length (L) and coupling
probe depth (LP) were adjusted to keep the lowest reflected power. The circulator (Model
MAHC 7238, Microwave Associates) has three ports which are connected to the microwave
generator and two coaxial directional couplers each with a 30 dB attenuator port (Model
3000-30, Narda Microwave Corp., Plainview, New York). Two directional couplers are
connected to the microwave cavity and a dummy load (Model 369 NF, Narda Microwave
Corp., Plainview, New York) and the attenuator ports were connected to power sensor. The
input and reflected power were measured by power sensors (Model HP 84881A, Hewlett
Packard) and power meters (Model HP 435B, Hewlett Packard).

61

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

power
meter temperature
(30 dB probe
attenuator
circulator port)
Hi» 4 «H
variable ower directional microwave
- p l P coupler cavrty
microwave generator
sample
directional
power J
VVV cou ler
mete (30 dB P
attenuator
port)
dummy load

Figure 4.3 Microwave generation system with an external circuit.

62

4.2.2.3 Temperature measurement system

A four-channel fluoroptic thermometer system (Model 755, Luxtron Corp., Mountain
View, California) was used to measure the sample surface temperahrre in situ . The fluoroptic
probe, protected by a 3 mm o.d. Pyrex capillary tube, was placed in the center of the Teflon
holder. The MIW probes used have a temperahrre range of ~190 to 300 °C. They are
electrically nonconductive and do not interfere with the electromagnetic fields. The probe
consists of a sensing tip (1.0 mm diameter), a silica fiber with a jacket of Teflon, and a
connector attached to the optical head in the main system unit. Measurements were made

at the rate of one every 10 seconds.

4.2.2.4 Vacuum chamber

The diagram of the microwave freeze-drying system (Figure 4.1) shows the quartz
tube (4.8 cm o.d., .14 cm wall thickness and 52.4 cm height) inserted through the microwave
cavity to form the vacuum chamber. The bottom of the quartz tube is blocked with a rubber
cork on which is placed a Teflon sample support. A small tube on the side connects the
chamber to the condenser, a vacuum pump (Model E2M2, Edwards High. Vacuum Inc.,
Grand Island, New York) with an inlet desiccant trap (Model ITD20K, Edwards High
Vacuum Inc., Grand Island, NY), and an outlet mist filter (Model EMFlO, Edwards High
Vacuum Inc., Grand Island, NY). The chamber pressure is measured by a model PRMlOK
Pirani Gauge Head (Edwards High Vacuum Inc., Grand Island, NY) which is connecmd to
a model 1101 conh'oller (Edwards High Vacuum Inc., Grand Island, NY).

A continuous flow of nitrogen is circulated in the vacuum chamber by adjusting the
integral bonnet needle valve (Whitey Co., Highland Heights, OH) to help maintain a constant

63

vacuum pressure. The nitrogen gas flow rate was measured by a flow meter (Model
FMA-300, Omega, Stamford, Connecticut) with a power supply. The sublimated water
vapor was transported to a liquid nitrogen condenser where it was condensed to keep the

water vapor partial pressure inside the chamber as low as possible.

4.2.2.5 Data acquisition system

Temperature, vacuum pressure, input and reflected microwave power, and nitrogen
flow rate were recorded by the data acquisition system every ten seconds. The system
consists of a 20-channel relay multiplexer (Model HP 44708A, Hewlett Packard), voltmeter
(Model HP 3852A, Hewlett Packard), and computer. The HP 44708A consists of a com-
ponent module and a 20~channel terminal module. The wires from each voltage source are
connected to the terminal module, and the recording frequency of each voltage source is

controlled by a computer.

4.2.3 Volatile retention analysis

Before volatile distillation and extraction, an internal standard of lur'nl 2000 ppm
l~octanol solution and 150 ml distillate water were added to each freeze-dried sample. The
volatile compounds were extracted by 15 ml of 2~methylbutane for 1 hour in a modified
Likens-Nickerson apparatus. Trace water in the extracted volatile solution was removed
by anhydrous sodium sulfate, and the extracted volatile solution was concentrated to 6 ml

by flashing nitrogen gas through it before gas chromatography analysis.

The 0.5 pl of extracted volatile solution was injected into a Hewlett-Packard 5890 gas

chromatograph (GC) equipped with a flame ionization detector. The separation of volatile

compounds was achieved in a fused silica capillary column (30m x 0.32 mm ID.) with

64

SupelcowaxTM 10 stationary phase. The oven temperature was held at 40°C for 1 min, then
increased at a rate of 7.5°C/min to 165°C. The injector and detector temperatures were
maintained at 220°C and 230°C, respectively. The carrier gas was helium at a flow rate of
30 mllmin. The peak area of each aroma compound was recorded on a Hewlett-Packard
3390A integrator. The peak area response of each volatile compound relative to the internal
standard was obtained by injection of standard solutions. The retention of volatiles in each
freeze—dried model food was calculated as a percent of the initial concentration of volatile
compounds in the frozen specimen.

4.2.4 Moisture content measurement

The initial weight (W,) of the frozen sample was measured by a Mettler balance. After

pro-selected periods of microwave freeze-drying, the weight (W) of the sample was mea-

sured. The dimensionless moisture content was calculated by

 

MC = _ (4.1)

The weights of dried samples (W4) were 0.08W, for 8% starch gel, and 0.12W, for 8% starch

gel mixed with 4% B-cyclodextrin or 4% sucrose.

4.2.5 Statistical analysis

Volatile retention and final moisture content data were analyzed using a MSTATC
statistics software package (Version 4.0, Michigan State University, E. Lansing, MI). Two
replications were carried out for all volatile retention and moishrre content analyses. The

statistical significance of the degree of variation in volatile retention and final moishrre

65

content was determined by a four-factor (sample treatment x freezing temperature x chamber
pressure x volatile compound) completely randomized design experiment and a three-factor
(sample treatment x freezing temperature x chamber pressure) completely randomized
design experiment of analysis of variance (AN OVA), respectively. Tukey’ s test and standard
error of the means were used to evaluate the significant difference between the means of

treatments at P S 0.05.

4.3 Results and Discussion

4.3.1 Volatile retention and final moisture content of microwave freeze-

dried samples

The volatile retention and final moisture content for the microwave freeze-dried
samples are shown in Table 4.1. The samples containing 4% B-cyclodextrin and 4% sucrose
required 50 and 60 minutes, respectively, to obtain a final dry sample; the control on the
other hand, required 100 minutes. Sample composition affects dielectric properties and
microwave power absorption (Roebuck et al., 1972), resulting in different processing time

requirements to achieve a given dried sample, even at the same microwave power input.

As expected, the time required for microwave freeze-drying was significantly shorter
than that for conventional freeze-drying, which normally requires overnight operation to
obtain a dry sample. As a result, the retention of each volatile compound during microwave
freeze-drying was significantly higher than that obtained during conventional freeze-drying.
For example, limonene retention during conventional freeze-drying of the control was less

than 5% (Chapter 3); however, microwave freeze-drying of the same material resulted in a

Table 4.1 Volatile retention and moisture content of microwave freeze-dried model foods.

66

(A) 8% Corn starch gel freeze-dried at 10 W for 100 min.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WWEmonene l-Herxanol 1~Decanol
Temp.(°C) (torr) % % % %
~15 l 6.8 69.3 89.4 90.0
~15 2 6.3 62.0 88.9 88.0
~60 l 5.8 58.8 79.6 88.4
~60 2 3.3 47.8 74.5 86.5
~198 l 8.2 39.4 78.7 85.0
~198 _2 7.2 40.0 72.6 79.5

 

 

 

(B) 8% Corn starch gel with 4% B-cyclodextrin freeze-dried at 10W for 50 min.

 

 

 

 

 

 

 

 

 

M.C.

 

Limonene

 

 

 

1~Hexanol

 

  

Freezing Pressure M.C. Limonene l-Hexanol 1-Decanol
Temp.(°C) (torr) % % % %
-15 1 8.8 95.4 93.6 95.4 ||
~15 2 4.2 98.6 95.3 96.0
~60 1 3.5 90.7 84.1 95.8
~60 2 3.0 95.5 75.7 93.7
~198 1 10.1 99.6 84.3 94.5
~198 2 5.9 94.6 77.2 85:1.

  

     

l ~Decanol

  

 

 

 

 

 

Temp.(°C) (torr) % % % %
-15 1 6.7 71.2 77.2 94.1
-15 2 2.2 74.2 69.5 90.0 H
~60 1 7.0 59.5 67.7 94.2 J
~60 2 4.4 58.2 59.7 94.1
1

 

 

 

N

 

 

 

 

~198 15.8 38.9 79.2 94.0
~198 9.3 38.9 68.9 99.0

67

retention of over 40%.

It was detem1ined, through analysis of variance, that freezing temperature and chamber
pressure were significant factors in the final moisture content of the freeze-dried products
(Table 4.2). The samples frozen in liquid nitrogen had significantly higher final moisture
contents than those frozen at higher temperatures. The rapid freezing induced by liquid
nitrogen causes smaller ice crystal formation in the samples, leading to smaller pores and
void spaces in the dried region (Chapter 3). This increases the mass transfer resistance, and
results in higher moisture contents. The samples freeze-dried at 1 ton chamber pressure
had higher final moisture contents than those at 2 torr (Tables '4. l and 4.2). A higher chamber
pressure corresponds to a higher ice front temperature (Dyer et al., 1966), and since dielectric
properties increase with temperature, the higher chamber pressure induces a faster drying

rate (Risman and Bengtsson, 1971).

The analysis of variance for volatile retention by a four-factor completely randomized
design is presented in Table 4.3. Sample composition, freezing temperature, chamber
pressure and the type of volatile compound were all significant factors in volatile retention.
In general, the samples freeze-dried at 1 ton had higher volatile retentions than those at 2
ton, although the difference was not significant. Only the retention of 1-hexanol in the
samples containing 4% sucrose freeze-dried at 1 ton chamber pressure is significantly higher

than at 2 torr (Table 4.1).

4.3.2 Effect of sample composition on volatile retention

The retention of individual volatile compounds was a strong function of sample
treatment (Figure 4.4). The retention of 1~decanol was higher than that of 1~hexanol.

According to the selective diffusion mechanism proposed by Thijseen (1979), the diffusion

68

Table 4.2 ANOVA table showing the effect of sample composition, freezing temperature
and chamber pressure on the final moisture content.

 

 

 

 

 

 

 

 

 

 

 

'-- “f “a“ *1
Freedom Squares Square Value

Factor A 18.3 9.1 M 6. 0010 1
Factor B 2 154.6 77.3 50.7 0.000 1
Factor C 1 81.0 81.0 53.2 0.000
AB 4 63.9 16.0 10.5 0.000 1

AC 2 15.4 7.7 5.1 0.018

BC 2 6.5 3.2 2.1 0.149 1
ABC 4 13.3 3.3 2.2 0.112 1
Error 18 27.4 1.5 1

1 ma! --“i

 

 

Key: Factor A = sample composition; Factor B = freezing temperature; Factor C = chamber
pressure.

Table 4.3 ANOVA table showing the effect of type of volatile compound, sample

composition, freezing temperature and chamber pressure on volatile retention.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source Degrees of Mean F Probability 1
Freedom Squares Square Value 1

FactorA i 2 1 17672.8 3836.4 309.00 W 0.000 *1 1
FactorB 2 2092.9 1046.5 84.29 0.000
Factor c 1 260.7 260.7 21.00 0.000 1
Factor D 2 9387.9 4694.0 378.08 0.000 1
AD 4 7606.2 1901.5 153.16 0.000

BD 4 1277.0 319.3 25.71 0.000

CD 2 94.6 47.3 3.81 0.028

1 AB 4 265.6 66.4 5.35 0.001 '
AC 2 19.4 9.7 0.78

BC 2 38.1 19.0 1.53 0.225 .
ABC 4 100.4 25.1 2.02 0.104 1
ABD 8 1522.3 190.3 15.33 0.000 f

I ACD 4 135.7 33.9 2.02 0.038 1
BCD 4 31.9 8.0 0.64
ABCD 8 149.2 18.6 1.50 0.178 1
Error 670. 4 12.4 1
-_-_:

Key. Factor A: sample composition; Factor B: freezing temperature; Factor C= chamber
pressure; Factor D= type of volatile compound

 

70

 

 

 

 

g cozceom

+ 4% sucrose

+ 4% cyclodextrin
Sample Treatment

@ Limonene 9$

control

Hexanol 7‘ 1-Decanol

1-

atile retention.

Figure 4.4 Effect of sample treatment on vol

71

of volatiles from the concentrated amorphous region to the sublimation surface is a con-
trolling factor in volatile loss. The smaller molecular weight or less soluble volatile com-
pounds have larger diffusivities and diffuse faster to the sublimation front, leading to lower

retention (Gero and Smyrl, 1982; Thijseen, 1979).

Each volatile compound has different specific heat and dielectric properties which
influence the internal temperatures for individual volatiles during microwave heating. Based
on this observation, Shaath and Azzo (1989) developed the so called Delta T theory. This
theory provides a possible mechanism that may be used to explain volatile retention in
microwave freeze-dried products. When the theory is applied to this study, the order of
temperature increases would be limonene > l-hexanol > l-decanol. With the exception of
samples containing 4% B-cyclodextrin, the order of volatile retention was l-decanol >

l-hexanol > limonene, as would be predicted by this theory.

The retention of l-decanol and limonene was significantly higher in samples containing
4% sucrose than in the control. Bartholomai et a]. (1975) and Chirife et al. (1973) reported
that low molecular weight carbohydrates have a larger number of binding-sites and were
more effective for aroma retention than polymeric compounds. However, the retention of

l-hexanol in samples containing 4% sucrose was not higher than in the control.

The addition of B-cyclodextrin led to significantly higher levels of volatile retention

(Figure 4.4). In particular, limonene retention in samples containing 4% B-cyclodextrin
was about 95%, which was significantly higher than in other treatments. This is due to the
fact that B—cyclodextrin can encapsulate suitably sized volatile compounds and promote
retention (Chang and Reineccius, 1990). The retention of l-hexanol in samples containing

4% B—cyclodextrin was also significantly higher than that in other sample treatments, and

72

l-decanol retention in samples containing 4% B—cyclodextrin was significantly higher than
that in the control. Encapsulation in B~cyclodextrin inclusion complexes appears to provide

an important mechanism for volatile retention.

4.3.3 Effect of freezing temperature on volatile retention

The effect of freezing temperature on volatile retention in three microwave freeze-dried
samples are shown in Figures 4.5, 4.6 and 4.7. The retention of l-decanol was generally
independent of freezing temperature. The retention of l-hexanol in the control and in samples
containing 4% B-cyclodextrin frozen at -15°C was significantly higher than those frozen in
liquid nitrogen (Figures 4.5 and 4.6). Limonene retention in the control and samples con-
taining 4% sucrose significantly increased with higher freezing temperature (Figures 4.5
and 4.7). The retention of limonene in samples containing 4% B-cyclodextrin was very

high; however, the effect of freezing temperature was not significant (Figure 4.6).

In general, samples frozen at -15°C had higher volatile retention than those frozen at

lower temperatures. In contrast, freezing temperature did not influence volatile retention
during conventional freeze-drying (Chapter 3). During microwave freeze-drying, ice is
more quickly sublimated to water vapor than in conventional freeze-drying; therefore, mass
transfer resistance becomes a limiting factor (Sunderland, 1980). Slower freezing rates
cause larger ice crystal formation, leading to larger void spaces in the freeze-dried samples.
The larger void spaces reduce mass transfer resistance, thus avoiding high water vapor
pressure build-up inside the samples. If this pressure build-up occurs, puffing or melting

of samples may take place, leading to the release of encapsulated volatiles.

73

 

 

 

 

at cozceom

-198 C

-GOC

Freezing Temperature

Hexanol % 1-Decanol

1-

 

e
n
e
n
o
m

@L

 

e retention of the control frozen at three temperatures

atil

Figure 4.5 Vol

74

 

 

 

 

é conceom

C -198 C

Freezing Temperature

-15C

Hexanol 7‘ 1-Decanol

1

   

Limonene '.

trin frozen at three

4% B—cyclodex

ning

e retention of samples con

Figure 4.6 Volatil

temperatures.

75

 

0’
O

Retention (%)
8

20

 

 

-15 C -60 C -198 C
Freezing Temperature

@ Limonene $9 1-Hexanol 1-Decanol

Figure 4.7 Volatile retention of samples containing 4% sucrose frozen at three tempera—

tures.

 

76

4.4 Conclusions

The drying time and volatile retention of microwave freeze-dried samples are shorter
and higher, respectively, than those of conventional freeze-dried samples. The samples
containing 4% B-cyclodextrin had significantly higher volatile retention, apparently due to
the formation of inclusion complexes. They also required relatively shorter drying times.
The retention of 1—decanol was higher than l-hexanol, and appeared to be based on selective
diffusion. Samples frozen in liquid nitrogen had significantly higher final moisture contents
than those frozen at higher temperatures, because the smaller pores in the dried region
increased water vapor transfer resistance. In general, samples frozen at -15°C had signif-
icantly higher volatile retentions. Operating the camber at 2 torr pressure resulted in lower
final moisture contents than at l torr; however chamber pressure was not a significant factor

in volatile retention.

CHAPTER 5
MICROSTRUCTURE OF MICROWAVE FREEZE-DRIED MODEL
FOODS

5.1 Introduction

The microstructure of microwave freeze-dried food products may be helpful in
interpreting volatile retention and drying data, and enable the determination of product bulk
density and pore size distribution (Margaritis and King, 1971, Rosenberg et al., 1985).
Several researchers have studied the effect of freezing temperature on the microstructure of
freeze-dried products (Bello et al., 1987; Consolacion and Jelen, 1986; Koonz and Rams-
bottom, 1939; Chapter 3). In general, a slower freezing rate leads to larger ice crystals and
results in larger pores and higher void volumes. This can improve water vapor transfer
during freeze-drying (Margaritis and King, 1971; Sunderland, 1980; Chapter 4). Conso-
lacion and Jelen (1986) found that during the slow freezing step, the growing ice crystals
mechanically compress protein molecules to promote fiberization. and extended

cross-linking protein structure.

Arsem and Ma (1985) observed the microstructure of microwave and conventionally
freeze-dried beef muscle fibers and reported a noticeable difference between the two. The
microstructure of microwave freeze-dried samples had broken pores, and the channels were
cleanly swept out by the rapid expansion of vapor. On the other hand, the conventionally
freeze-dried muscle fiber suffered case hardening and sealed pores.

78

The objectives of this research were:
(1) to study the effect of composition on the surface and cross-section microstructure of
microwave freeze-dried model foods,
and
(2) to investigate the effects of chamber pressure and freezing temperature on the micro-

structure of the freeze-dried samples.

5.2 Materials and Methods
5.2.1 Preparation of starch gel model foods

Three sample treatments were used in this study. The control was 8% pregelatinized
corn starch mixed with 500 ppm each of l-hexanol, l-decanol, and limonene. The other
two treatments involved addition of 4% B-cyclodextrin (CD) and 4% sucrose, respectively,
to the control. Each sample was homogenized and poured into a cylindrical mode 2.5 cm
high and 1.5 cm in diameter. Before microwave freeze—drying, samples were frozen at
- 15°C, -60°C or - 198°C (in liquid nitrogen). Samples were then freeze-dried at a microwave
power input of 10 w and a chamber pressure of 1 or 2 torr. A description of the microwave
freeze-drying systems used for this work is given in Chapter 4.

Pregelatinized corn starch and B-cyclodextrin were obtained from the American Maize

Products Co. (Hammomd, Indiana). Limonene, 1-hexanol, 1-octanol and l-decanol were
obtained from the Sigma Chemical Co. (St. Louis, MO). 2-Methylbutane was obtained from
the Aldrich Chemical Co. (Milwaukee, WI). Sucrose and anhydrous sodium sulfate were
obtained from the J .T. Baker Chemical Co. (Philipsburg, NJ).

79

5.2.2 Microstructure analysis

Each microwave freeze-dried specimen was cut into two small pieces (5 x 5 mm), one
from the surface of the sample and the other from the cross-section of the sample. Each
piece was mounted on an aluminum stub (Electron Microscopy Sciences, Ft. Washington,
PA) using epoxy glue and coated with a thin layer of gold in a sputter coater (Emscope
model SC 500, Kent, England). Micrographs were obtained by a scanning electron
microscope (IEOL model ISM-35C, Osaka, Japan) at an accelerating voltage of 15kV and

condenser lens setting of 400.

5.3 Results and Discussion

5.3.1 Effect of sample composition on microstructure

Surface and cross-section micrographs of three sample treatments that were frozen at
—60°C and freeze-dried at 10 W microwave power and 1 torr chamber pressure are shown
in Figure 5.1. The dark regions are empty pores and void spaces left after the sublimation
of ice crystals. The control (8% starch) shows a uniform three-dimensional. gel network
structure (Figures 5.1A and 5.1B). The surface rnicrostructure is noticeably different from
the cross-section rnicrostructure. A larger gel structure and larger void spaces are evident

in the cross-section micrograph (Figure 5.1B).

The surface microstructure of 8% starch gel with 4% B—cyclodextrin featured parallel

fibers and channels, with broken thin flakes. Cross—linkages due to molecular interaction
are evident; they connect the parallel fibers and support the fiber structure (Figure 5.1C).

The channels were formed during the freezing step when the growing ice crystal

 

551.31: L
F1. 1181;:
011:: 111;:
1111:; it:

it 01 1:13:

.
”A ‘9.”

it?

11: 2:

80

Figure 5.1 Microstructure of three samples frozen at -60°C, then freeze-dried at 10 W and
l torr. (A) surface of 8% starch gel, (B) cross-section of 8% starch gel, (C) surface of 8%
starch with 4% B-cyclodextn'n, (D) cross-section of 8% starch with 4% B—cyclodextrin,
(B) surface of 8% starch with 4% sucrose, (F) cross-section of 8% starch with 4%
sucrose.

On

410011111:

. (C1 strict-'3
1161:1011;

rh 111111111

 

~t.w
.C

.3

82

mechanically compressed the gel structure and pushed the solid components closer together
to concentrate the fiber structure between the parallel-oriented ice crystals (Consolacion
and Jelen, 1986). There were deposits of aggregated B—cyclodextrin on the cross-section
channel structure, and they appear to seal or cover up some of the pores and void spaces
(Figure 5.1D). This is quite different from what was observed in the surface structure. The
aggregated deposits blocking the pore structure may be responsible for the higher volatile

retention in this sample (Chapter 4).

The samples containing 4% sucrose featured broken flakes deposited on the fibrous
structure (Figures 5113 and 5.1F). The sample containing 4% sucrose had significantly
larger pores than both the control and the sample containing 4% B—cyclodextn’n (Figure 5.1).
These micrographs clearly show that food composition significantly influences the micro-

structure of freeze-dried products.

5.3.2 Effect of freezing temperature on microstructure

The surface micrographs of three samples frozen at -15°C and -198°C and freeze-dried

at l torr are shown in Figure 5.2. As expected, the samples frozen in liquid nitrogen contained
smaller pores (Figure 5.2B) and narrower gaps between the fibrous structure (Figures 5.2D
and 5.2F) than the samples frozen at - 15°C (Figures 5.2A, 5.2C and 5.213) and -60°C (Figures
5.1A, 5.1C and 5.1E). Void spaces generally increased with freezing temperature.

The cross-section micrographs of samples frozen in liquid nitrogen are shown in Figure
5.3. These micrographs are clearly different from the samples frozen at -60°C (Figures
5.1B, 5.1D and 5.1F). The cross-section of the 8% starch gel frozen in liquid nitrogen shows

a three dimensional gel structure with many small pores (Figure 5.3A). In

83

Figure 5.2 Surface micrographs of three samples frozen at -15°C and -l98°C, then freeze-
dried at 10 W and 1 torr. (A) 8% starch gel at -15°C, (B) 8% starch gel at -198°C, (C)
8% starch gel with 4% B-cyclodextrin at -15°C, (D) 8% starch gel with 4%

B-cyclodextrin at -198°C, (E) 8% starch gel with 4% sucrose at -15°C, (F) 8% starch gel
with 4% sucrose at -l98°C.

 
   

' ~. is 55,1. "¥~~ig=-~ -
~ ., .“ o‘l1'_.‘~. "
. 1' d‘ ; ‘fifi;!;vv.,i; I!" N)%

‘ 1 ' e 'eriic é;

‘ 1" 9.7., '-' 1 é;

 

   
 

  

 

C, [11:13":
1981.11

11151.10;

agar?! . ..-I...-._......E

85

Figure 5.3 Cross-section micrographs of three model foods frozen at -l98°C, then freeze-
dried at 10 W and 1 torr. (A) 8% starch gel, (B) 8% starch gel with 4% B«cyclodextrin,
(C) 8% starch gel with 4% sucrose.

1"11‘11‘.v.‘.~%_fl f‘fi‘rfl’fi
.‘ I“. 1 !

198°C. 1E 15
1 1141ch

 

87

contrast, the cross-section of the samples containing 4% B—cyclodextrin and 4% sucrose

frozen in liquid nitrogen featured small lumps of aggregated material or pieces of flakes

deposited on narrow gaps or channels (Figures 5.3B and 5.3C).

Ice crystals occupy larger volumes than the liquid water from which they originate.
Therefore, mechanical injury of cells and tissue occurs during the freezing process. In
general, rapid freezing favors the formation of small intracellular ice crystals, which causes
a minimum dislocation of tissue and less structural damage (Li-Shing-Tat and Jelen, 1987;
Bello et al., 1982). However, rapid freezing leads to smaller pores in the frozen products,
which causes larger mass and heat transfer resistance, and increases drying time. For
example, over the same drying time, the final moisture content of samples frozen in liquid
nitrogen was higherthan that of samples frozen at -15°C and -60°C (Chapter 4). In addition,
volatile retention in freeze-dried samples frozen in liquid nitrogen was lower than that of
samples frozen at higher temperatures (Flink, 1975; Gero and Smyrl, 1982; McPeak et al.,
1987; Chapter 4). Therefore, smaller ice crystals are not desirable for microwave freeze-

drying.
5.3.3 Effect of chamber pressure on rnicrostructure

The surface and cross-section micrographs of the control frozen at -60°C and

freeze—dried at 2 torr are shown in Figures 5.4A and 5.4B, respectively. The surface
microstructure of 8% starch gel freeze-dried at 2 torr had larger pores than that at 1 torr
(Figures 5.1A and 5.4A) . The average pore diameters of the samples frozen in liquid nitrogen
and freeze-dried at 1 torr and 2 torr chamber pressure were 4.81.1m and 6.01.1m, respectively
(Figures 5.5A and 5.58). These pores are larger than the average diameter of 2.5 pm observed

88

Figure 5 .4 Microstructure of three model foods frozen at -60°C, then freeze-dried at 10 W
and 2 torr. (A) surface of 8% starch gel, (B) cross-section of 8% starch gel, (C) surface
of 8% starch with 4% B-cyclodextrin, (D) cross-section of 8% starch with 4% B-cyclo-
dextrin, (E) surface of 8% starch with 4% sucrose, (F) cross-section of 8% starch with
4% sucrose.

88

F‘fii‘ffiflfi '2 r'
. i {’1‘ at"?!

“43%?“

(

   

81841119513"
gel. 10 tit
14111-1
‘7: 513111113

Figure 5.5 Higher magnification cross-section micrographs of the control and samples
containing 4% B-cyclodextrin frozen at -l98°C, then freeze-dried at 10 W and l torr or 2
torr. (A) control at 1 torr, (B) control at 2 torr, (C) 4% B«cyclodextrin at l torr, (D) 4%
B—cyclodextrin at 2 torr.

91

 

92

when the same sample was conventionally freeze-dried (Chapter 3). It is apparent that
different freeze-drying processes influence the pore size distribution.

The surface microstructure of samples containing 4% [3»cyclodextrin frozen at -60°

and freeze-dried at 2 torrhad a more broken and open structure than those at 1 torr (Figures
5.1C and 5.40. The average pore diameters of the samples frozen in liquid nitrogen and
freeze-dried at 1 torr and 2 torr were 5.61m and 8.0um, respectively (Figures 5.5C and
5.5D), and are slightly larger than those of the control processed under the same conditions
(Figures 5.5A and 5.5B). The channels in samples freeze-dried at 2 torr (Figures 5.53 and
5.5D) are clearly swept, in contrast to those at 1 torr (Figure 5.5A and 5.5C). A smaller
number of larger pores with thicker gel matrices are also evident in the samples freeze-dried
at 2 torr (Figure 5.5D).

Chamber pressure had a noticeable influence on the microstructure of samples con-
taining 4% sucrose. The surface and cross-section microstructure of the samples freeze-dried
at 2 ton had more void spaces and less flakes covering the gel structure than those at 1 torr
(Figures 5.1B, 5.1F and 5.4153, 5.4F, respectively). Samples freeze-dried at a chamber
pressure of 2 torr have higher ice front temperatures (Dyer et al., 1966) and larger ice
sublimation rates than those at 1 torr. However, the water vapor cannot immediately transport
to the outside of the sample due to mass transfer resistance. As a result, there is a build-up
of vapor pressure inside the sample. Once the water vapor pressure inside the sample is
high enough, the vapor expands and sweeps out a channel or pore and produces a more
broken surface structure. If it also causes puffing, then volatile retention in the puffed sample
may decrease due to the release of entrained volatiles (Chapter 4).

93

5.4 Conclusions

Sample treatment was a significant factor in final product rnicrostructure. The control
(8% starch) had a uniform three-dimensional network gel structure. The surface micro-
structure of samples containing 4% B-cyclodextrin featured parallel fiber and channels, with
broke thin flakes. There were deposits of aggregated materials which sealed the pores and
void spaces on the cross-section channel structure of samples containing 4% B-cyclodextiin.
Samples containing 4% sucrose featured broken flakes deposited on the fiber structure.
Samples frozen in liquid nitrogen contained smaller pores and narrow gaps. Generally,
samples freeze-dried at 2 torr had larger pore size and more broken structures than those at

1 torr.

CHAPTER 6
KINETICS OF VOLATILE RETENTION DURING MICROWAVE
FREEZE-DRYING

6.1 Introduction

Even though they are usually in low concentration, volatile compounds contribute
very strongly to the overall acceptance of a given food product. As a result, their retention
during food processing is an important quality factor. However, there is little information
on volatile retention kinetics during dynamic thermal processes (Villota and Hawkes, 1986).
Because the temperature and moisture profiles of products continuously change during
dehydration, the kinetics of volatile retention during drying of products are complicated.
However, modeling the kinetics of volatile retention is necessary to understand the retention

mechanism; it also provides important information for the optimization of thermal processes.

Chung (1984) studied the kinetics of ethanol, acetone and ethyl acetate retention in
model food systems at constant temperatures of 35°C, 60°C, 80°C and' 95°C during
microwave and conventional heating. He reported that volatile retention is a first order
reaction, and that the temperature dependent rate constant follows the Arrhenius relationship.
He also found that volatile retention in products heated by microwave energy was higher

than those processed by conventional heating.

The objectives of this project were to analyze the temperature, moisture and volatile
retention profiles of model food gels during microwave freeze-drying, and to develop a

model for the kinetics of volatile retention.

94

6.2 Theory

95

Nutrient and quality degradation reactions are generally assumed to be irreversible,

noncyclic, first or zero order reactions. The kinetic parameters are usually determined by

the integral (rather than the differential) method.

A zero order quality degradation kinetics is governed by the equation:

——=—k

dt

which may be integrated to obtain:

[A] = [Ala-kt

A first order quality degradation kinetics is:

m1

dt =44“

which, upon integration gives:

ln[A] = ln[A 1,, - kt

(6.1)

(6.2)

(6.3)

(6.4)

Under isothermal conditions, the zero order rate constant can be obtained from the slope of

a plot of [A] versus time. Similarly, the first order rate constant can be obtained from the

slope of a plot of In [A] versus time. The effect of temperature on the reaction rate constant

generally follows an Arrhenius relationship:

1 I] qui I 6 1‘, I.
ll .1' “1"“

96

—Ea
k =koexp[ RT ) (6.5)

Of

 

lnk-lnk E“ 1
__ ,R T (6.6)

A plot of In k versus the inverse temperature (1“) gives a straight line with a slope of -E,/R

and an intercept of In k0. Several researchers have applied the Arrhenius relationship to
derive kinetic models for dynamic thermal processes (Wanninger et al., 1972; Saguy et al.,
1978a,b; Laing et al., 1978; Villota and Karel, 1980; Mishkin et al., 1984; Choi and Villota,
1989a,b; Franzen et al., 1990) and for non-isothermal storage (Mizrahi and Karel, 1977;
Labuza, 1979). For instance, to accommodate non-isothermal conditions, Eq. (6.5) can be

substituted into a first order quality degradation expression (Eq. 6.3) to obtain:

w th]_ . 4;,
Jul, - [A] -10 k" exleTQJdt (6'7)

which, upon integration, gives:

lntAl=1nlA],-1o'k, exp1 $121144 (6.8)

In addition to variations in temperatrue, most quality degradation kinetic parameters

 

 

are also influenced by the water activity of the product (Saguy and Karel, 1980). For instance,
the rate of ascorbic acid degradation decreased with drying time, and ascorbic acid retention
seemed to approach some equilibrium level during the final drying step (Mishkin et al.,
1984; Villota and Karel, 1980). In order to obtain rate constants for quality degradation

 

97

during dynamic processing, many researchers have modified the pre-exponential factor and
the activation energy in the Arrhenius equation to obtain functions of moisture and/or
temperature (Mizrahi and Karel, 1977; Saguy et al., 1978a,b; Villota and Karel, 1980;
Mishkin et al., 1984; Choi and Villota, 1989b; Franzen et al., 1990). This has usually been
accomplished through statistical regression analysis or numerical algorithms (Table 6.1).
Most of these kinetic models work very well but are quite complicated, and require con-
siderable simplification to be of practical use for industrial application or process optimi-

zation.

Thompson (1982) reported that isothermal kinetics for the growth and inactivation of
some microorganisms cannot be directly applied to non-isothermal conditions, and that
using isothermal kinetics may over-predict quality degradation during dynamic dehydration.
In addition, the quality degradation may appear to be a first order reaction in the excess
nutrients, but the reaction rate and order may change as the reaction proceeds (Thompson,
1982). If the reaction shifts from first to zero order during food processing, then the rate
constants are very difficult to determine by the integral method. However, the differential
method can be used to easily analyze this reaction mechanism and obtain the two rate
constants (Levenspiel, 1972). Therefore, the differential method of analyzing quality
degradation retention kinetic data may be more suitable than the integral method for dynamic
thermal or dehydration processes. Using the differential method, the reaction rates can be
obtained from the slope of concentration versus time curve at some selected concentrations.
The reaction rate data must then be plotted against an evaluated concentration function. If
the resulting plot is linear and goes through the origin, then the rate equation is consistent
with the experimental data, and the reaction order, rate constant and reaction mechanism

can be determined (Levenspiel, 1972).

98

Table 6.1 Quality degradation kinetic models for dynamic processing or storage.

 

   

   
 

     

 

    

  
        
 

    

 

 

 

 

 

 

 

 

 

 

 

Quality Reaction Model Process Reference
Order Condition
ascorbic first —Ea blending Wanninger
acid (integral) Ink = In k. +-§T-+ Pr 1111" JR. (1972)
ascorbic first In k0 = 111(1)] M .1. PZMZ) permeable Mizrahi
acid (integral) package and Karel
Ea 2 storage (1977)
F=P3+P4M+P5M (dynamicM
& const. T)
ascorbic . first [10011)]=-1"k,(M)exp1'E'(M)1(-4—'- air-drying Saguy et al.
acrd (integral) "- RT ‘1” (dynamic M (1978a)
k.(M)=exp<P.+P.M) 84'”
E.(M) = exp<Pa + P4M)
ascorbic first in k, = P, + P,(Br/10)+ P3(Bx/10)2 + mix/10)3 heating- Saguy et al.
acid (integral) 5, = P,+P,(Bx/10)+P,(Bx/10)’+P,(Bx/10)’ concentration (1978b)
(dynamic M
& T)
ascorbic zero _ i=n high temp. Laing et al.
acid (integral) [A 1.. - [A l. 1230”?” extrusion 1 (1978)
(dynamic T & .
const. M)
reduced first In k = pl M + P2/T3 + [231143 .1. P4M2/T air-drying Villota and
ascorbic (integral) (dynamic M Karel
acid +195 M/T2 .1. P6M3/T3 + p7 & const. T) (1980)
total first m k = pl M + p 21142/]:2 + P3/T3 air-drying Villota and
ascorbic (integral) (dynarrric M Karel
acid +P4M3/T2 .1. 195 M /T3 + 136 & const. T) (1980)

 

 

 

 

Table 6.1 (cont’d)

99

 

 

 

 

 

 

 

 

 

 

ascorbic first In k0 = p1+ P2M .1, P3M2 dehydration Mishkin et
acid (integral & (dynamic M al. (1984)
differential) Ea = 194 .1, 135 M + P 6M2 .1. P7M3 & T)
methionine first ln k = P1 + P2T + P3/T + [M *IMC/ extrusion Choi and
(integral) ([24 + pvac) + pGM/TZ] [197 + p8(pc)°-5] (const. M, Villota
IMC, PC & (l989a,b)
dynamic T)
Browning pseudo-zero p2 rehydration Franzen et
reaction (integral) 1n k0 = P1 +fi & heating al. (1990)
(dynamic M
Ea P4 & const. T)
'R— = P3 + 3;
quality loss zero & _Ea storage Labuza
first k = k0 exp1 R(T _1_ a sin 21m] (sine wave (1979)
(integral) m 0 temperature)

 

 

meme-rm“;

100

6.3 Materials and Methods
6.3.1 Preparation of starch gel model foods

Pregelatinized corn starch solutions of 8% solids were mixed with 500 ppm each of
1-hexanol, 1-decanol and limonene, and subjected to three treatments: (1) null treatment or
control, (2) addition of 4% [ii-cyclodextrin (CD), and (3) addition of 4% sucrose. Each
sample was homogenized and poured into a cylindrical mode 2.5 cm high and 1.5 cm in
diameter. Samples were frozen at -60°C in a freezer before microwave freeze-drying.

Pregelatinized corn starch and [Ev-cyclodextrin were obtained from the American Maize

Products Co. (Hammomd, Indiana). Limonene, l-hexanol, 1-octanol and l-decanol were
obtained from the Sigma Chemical Co. (St. Louis, MO). 2-Methylbutane was obtained from
the Aldrich Chemical Co. (Milwaukee, WI). Sucrose and anhydrous sodium sulfate were
obtained from the J .T. Baker Chemical Co. (Philipsburg, NJ).

6.3.2 Experimental set-up

The microwave freeze-drying system consisted of several components: a single mode
tunable resonance microwave cavity applicator, a microwave generator with an external
circuit, a temperature measurement system, a vacuum system with a pressure gauge, a system
for flushing nitrogen gas through the vacuum chamber, and a condenser. The surface and
center temperatures of each sample, vacuum pressure, input and reflected microwave power,
and nitrogen flow rate were recorded by a data acquisition system every ten seconds. A
detailed description of the microwave freezedrying system used for this work is given in
Chapter 4.

101

6.3.3 Moisture content measurement

The initial weight (W) of the frozen sample was measured by a Mettler balance. After
preselected periods of microwave freeze-drying, the weight (W) of the sample was mea-
sured. The dimensionless moisture content was calculated by

_ (iv—w.)

MC — W (6.9)

The weights of the dried samples (W4) were 0.08W1 for 8% starch gel, and 0.12W,1 for 8%

starch gel mixed with 4% B—cyclodextrin or 4% sucrose.

6.3.4 Volatile retention analysis

Before volatile distillation and extraction, an internal standard of 1 ml 2000 ppm
1-octanol solution and 150 ml distilled water were added to each freeze-dried sample. The
volatile compounds were extracted by 15 ml of 2-methylbutane for 1 hour in a modified
Likens-Nickerson apparatus. Trace water in the extracted volatile solution was removed
by anhydrous sodium sulfate, and the extracted volatile solution was concentrated to 6 ml
by flashing nitrogen gas through it before gas chromatography analysis.

The 0.5 ul of extracted volatile solution was injected into a Hewlett-Packard 5890 gas

chromatograph (GC) equipped with a flame ionization detector. The separation of volatile
compounds was achieved in a fused silica capillary column (30 m x 0.32 mm ID.) with
SupelcowaxTM 10 stationary phase. The oven temperature was held at 40°C for 1 min, then
increased at a rate of 7.5 °C/min to 165°C. The injector and detector temperatures were

maintained at 220°C and 230°C, respectively. The carrier gas was helium at a flow rate of

 

1’. 24.1..
.11

1“ "I.“

102

30 mllmin. The peak area of each volatile compound was recorded on a Hewlett-Packard
3390A integrator. The peak area response of each volatile compound relative to the internal
standard was obtained by injection of standard solutions. The retention of volatiles in each
freeze-dried model food system was calculated as a percent of the initial concentration of
volatile compounds in the frozen specimen.

6.3.5 Statistical analysis

PLOTIT (Version 1.1, Interactive Graphics and Statistics, Scientific Programming
Enterprises, Haslett, MI) and MSTATC statistics software package (Version 4.0, Michigan
State University, E. Lansing, MI) were used for regression analysis. After preselected
periods of microwave freeze-drying, two replications were carried out for volatile and

moisture content analyses.
6.4 Results and Discussion
6.4.1 Temperature profiles

The center and surface temperature profiles of the three treatments recorded during
freeze-drying at a microwave power of 10W and a chamber pressure of 1 torr are shown in
Figure 6.1. The surface temperature profiles of samples gradually increased with drying
time. The center temperature profiles of samples containing 4% B—cyclodextrin and 4%
sucrose showed some fluctuation. The center temperatures of samples containing 4%
B—cyclodextrin were generally higher than that of the control. Risman and Bengtsson (1971)
reported that both the dielectric constants and the dielectric loss factors of frozen foods are

small, but show a great increase around the melting point of water. Therefore, as the

Nil-“E1. 1’" my

103

temperature of the frozen samples approaches the melting point of water, the samples attain
higher dielectric loss factors and absorb larger quantities of microwave power to cause

further increases in temperature.

Most heat and mass transfer analysis for microwave freeze-drying have been done for
a slab geometry. The slab is often divided into an exterior dried region and a frozen center
region (Ma and Peltre, 1975a,b; Ang et al., 1977a,b; Arsem and Ma, 1990). These studies
have generally produced temperature profiles similar to the ones seen in the control sample
(Figure 6.1a), and have reported that the center temperature in the frozen region remained
below the melting point until the final drying step. However, in this study, the center
temperature of the samples containing 4% [S—cyclodextrin increased rapidly, and exceeded
the melting point within about 30 minutes (Figure 6.1b).

6.4.2 Moisture profiles

The moisture profiles of samples freeze-dried at 10 W microwave power input and 1
torr chamber pressure are shown in Figure 6.2. The data can be modeled by a second-order
polynomial equation (Table 6.2) of the form: ' 1

MC =10l +P2t +sz (6.10)

Dielectric properties are a function of sample composition (Roebuck et al., 1972); the
resulting influence on microwave power absorption within diverse samples leads to different
drying rates. For example, the control required a significantly longer time to dry than the
samples containing 4% B-cyclodextrin and 4% sucrose. It is clear that even a small difference
in formulation of samples significantly influences the drying rate and the temperature

profiles.

104

 

50 j I I I 1 I l I I r I I l I I I T I T
o—o surface
40 H center ,,

 

 

 

 

 

30

3 20

2 10 .

3 ~ ' __

o 0 ~ __

3 " __

E 40 1' 1 ..

o -20 f . I ' * ‘ ' " ‘ ‘

F‘ " -1
-30 -- __
-40
“-501 I l 1 l 1 r

l ' l ' l l j l ' l '. l ' I
0 10 20 30 40 50 6O 70 80 " - 90 100
Time (min)

Figure 6.1a Surface and center temperature profiles of control samples freeze-dried at a
microwave power of 10 W and 1 torr.

105

50 l l l I l l l l I
o—o surface

40 A—A center

 

30
20
10

 

0

I
g
C

r
,1 d
I, C
q
" d
I
'I -
_' d
, d
.t
.1

Temperature (°C)

1 LL11 llll

 

“501 l I I I I I I I_. T .
O 5 10 15 20 25 30 35 40.45 50

 

Time (min)

Figure 6. lb Surface and center temperature profiles of samples containing 4% B-cyclo-
dextrin freeze-dried at a microwave power of 10 W and 1 torr.

m. . f1?fi1fl3"

106

 

50. I I I l I l I I I I I
:e—osurlace

40f A—d center

 

 

 

 

 

 

 

 

 

 

 

30%
E: zo-f
2 10-j
3 = --
g 0'. __ ..
a) _ _ _‘
E1 10: I,"
o ’20; j
P u
—30 ,_
-4o _ '—
l
-50

F l l l l l I F l l, l l
l 5 10 15 20 25 30 35 40 45 50. 55 60
Time (min)

Figure 6.1c Surface and center temperature profiles of samples containing 4% sucrose
freeze-dried at a microwave power of 10 W and 1 torr.

 

107

 

 

 

 

a . i
0 '3 E“
+9 1 E
t: ; ,
0 1 f.--
U 1
0’ '3
In .
:1 .:
+’ :
ta .
"III .-
O 7.
2 : I
02‘. — regression lines ° '.
f o 87astarch a ' o
0-1'3 6 8%starch with 47:01) ° 11
2 A 87astarch with 4% sucrfie I
0.0 I I T T I I 1 I I I I I I. I I

 

0 10 20 30 40 5'0 60 70 801-90 100
Time (min)

Figure 6.2 Moisture profiles of three microwave freeze-dried samples.

108

Table 6.2 Empirical models for moisture content profiles.

 

      

 

Regression of moisture profile data

 

 

4% sucrose

 

 

 

 

3% starch M.C. = 1.015 — 0.019: + 0.94x10'4t2 998 0-20

8% starch with M.C. = 1.023 — 0.026t + 0.1 1x10‘3r2 994 0-12
4% cyclodextrin

8% starch with M.C. = 1.037 - 0.033: + 0.29r10‘3t2 ~996 0- 14

 

 

11m:
11

109

Often, two different periods of drying rate occur: a constant rate period and a falling
rate period. The critical moisture content is defined as the moisture content at which the
drying rate changes from a constant rate to a falling rate. The critical moisture contents of
each sample can be estimated from the moisture profile, and are given in Table 6.2. The
control required a longer time (60 minutes) to reach the critical moisture content than the

other two sample treatments.
6.4.3 Volatile retention profiles

The volatile retention and moisture profiles of samples are shown in Figures 6.3, 6.4
and 6.5. As moisture content decreased, the rate of volatile loss also decreased, with each
volatile in the three samples approaching a different final retention level. The retention
profiles are similar to those obtained by King et a1. (1984) during a spray drying process.
Due to their high volatilities, once volatiles diffuse to the sublimation front, they quickly
evaporate. Therefore, the diffusion of a volatile to the sublimation front becomes the
controlling step for volatile loss (Thij seen, 1979). Thijseen (1 979) showed that the diffusion
coefficients of water and volatiles are strongly dependent on moisture content, and the ratio
of volatile diffusivity to water diffusivity decreases quickly with decreasing moisture
content. Thijseen (1975) also reported that volatile compounds can escape only until the
critical moisture content is reached; that is, volatile loss only occurs during the constant
drying rate period. Below the critical moisture content, only water can diffuse to the surface
of the product. Since microwave freeze-drying occurs through volumetric heating, only a
short time is required to complete the constant drying rate period, thus leading to higher

levels of volatile retention.

110

 

 

 

 

 

 

100 ‘ ‘ I I I l I I I l l
\1 1 l I I I I I I I j
09-: ' ° ‘
0.8-j . ,
07-:
S 0.6-: 1
:3 a
i: 0.5:
Q) .
I» 04:
DC: ' E
0.3-:
— regression lines
0.21 it moisture content '
E A decanol ,
0-1‘: c1 hexanol
0.0 . o limonene

I I . I ' I ' I I I ' I . I ' I ', I ' I
0 10 20 3O 40 50 60 70 80 " .90 100
Time (min)

Figure 6.3 Moisture and volatile retention profiles of the control freeze-dried at 10 W and
l torr.

 

111

 

 

 

1.0 C - . i 1
3 1‘ j l l l
09-: ‘3“
0.8-3 1 - g .
- E

llllLLLLllll I ll

Retention
0
U"
I
‘- .

: — regression lines

0.2-1 a! moisture content

decanol .
hexanol

limonene
l

I I I I I I I . I
0 5 10 15 20 25 30 35 40".45 50

llLl+llllll

 

O
g—a
llllll
O D D

 

Time (min)

Figure 6.4 Moisture and volatile retention profiles of samples containing 4% Becyclodex-
trin freeze-dried at 10 W and 1 torr.

112

 

10 1 l I I I l l l l l

 

 

 

+ H— a a
0.9 \°
0.8 1
0.7
5 0.6
:3 E
13 0.5 1
a) u
‘6 0 4 1
0: ' E
0.8 -I
— regression lines
0.2 a! moisture content '-
A decanol r .
0-1 El hexanol ' '-
00 I 0 Ilimonene

l l l l l l l l l, r l
O 5 10 15 20 25 30 35 40 45 50. 55 60
Time (min)

Figure 6.5 Moisture and volatile retention profiles of samples containing 4% sucrose
freeze-dried at 10 W and 1 torr.

. - f". '.

113

6.4.4 Kinetics of volatile retention

6.4.4.1 Determination of kinetic parameters by the integral method

There is no information in the technical literature on volatile retention kinetics during
freeze-drying. There are, however, several studies related to the retention kinetics of ascorbic
acid and other nutrients. The kinetic parameters are generally determined by the integral
method, with the rate constant modeled as a function of temperature and moisture (Table
6.1). Though the models provide a good set forthe quality degradation data, many parameters

are required to determine the rate constant.

For instance, in modeling the kinetics of ascorbic acid retention during dehydration,

Mishkin et al. (1984) used the following expressions for the pre-exponential factor and the

activation energy:
lnko =P1+P2M+P3M2 (6.11)
and
E, =P4+P5M+P6M2+P7M3 (6.12)

Eqs. (6.1 1) and (6.12) can be substituted into the Arrhenius relationship Eq. (6.6) to obtain:

_ 2 P4 1 P5 M P6 £2 P7 L4:
lnk-P,+P2M+P,M {EL-(EL {'15) T {3] T (6.13)

Equation (6.13) is rather complicated, and would clearly require numerical integration of

Eq. (6.8) to determine the concentration profiles.

IT”

114

In an atternative approach, Villota and Karel (1980), and Choi and Villota (1989a)
developed empirical nutrient degradation rate constants which did not follow the Arrhenius
relationship (Table 6.1). Villota and Karel (1980) estimated the rate constant of ascorbic
acid retention during air-drying by:

1 M2 M M 3
1nk=PlM+Pz(;5)+-P3M3+P 7J+P5(-T—Z]+P ?]+P7 (6.14)
Before using the differential method to evaluate the rate constant for volatile retention

during freeze-drying, an attempt was made to determine the kinetic parameters for limonene
retention during microwave freeze-drying of the control (8% starch gel) by the integral
method. Assuming that volatile retention is a first order reaction (Chung, 1984), and that

the rate constant is a function of both temperature and moisture profiles during microwave
fieewdrying, Eq. (6.2) becomes:

1111-11:- f 'kdt (6.15)
0

[AL

The plot of ln([A ]/[A ]0) versus time is shown in Figure 6.6. The left hand side of Eq. (6.15)

can be determined from the data of [A] as a function of drying time by a nonlinear regression
The resulting equation (coefficient of determination of 0.996) is:

[A]

lnl-Af]: = 0.541 exp(—0.0337t) "' 0.543 (6.16)

Equation (6.16) can be differentiated with respect to time to obtain the rate constant for

limonene degradation as a function of time:

115

 

0~00:'I*I‘I'I'I'I'I'I'I‘
-o.05-:
-o.10§
-o.15-:
-o.2o-§
-o.25-j
-o.3o-j
4.351:
-o.4o-:
-o.45-j
-o.5o-j .

-O.55-: — regression line
_ 0.60 - o limonene

'l'l'l'l'l'l'l'T'l'l
0 10 20 30 40 50 60 70 80790100

ln([A]/[A]0)

 

 

Time (min)

Figure 6.6 First order limonene retention kinetics in the control using the integral method.

116

k = 0.0182 exp(—0.0337t) (6.17)

The rate constant of limonene degradation in Eq. (6.17) decreases with drying time.
If the rate constant is modeled as a function of the average temperature and moisture content
of 8% starch gel during microwave freeze-drying, and it is assumed to follow Eq. (6.13) or
(6.14), then the constants can be obtained by multiple regression analysis (Table 6.3). The
parametric constants of Eq. (6.13) tabulated in the first column of Table 6.3 can be used to
compute the pre-exponential factor and activation energy. Unfortunately, the resulting
activation energy (based on Eq. 6.13) is negative over all values of time, except at the
initiation of the freeze-drying process. Therefore, although the correlation between the
kinetic model and the empirical data is nearly perfect, this approach seems to be entirely
one of curve fitting with no physical basis. In addition, a kinetic model that requires the
determination of seven parameters is entirely too complicated for routine optimization of
food processes and for industrial applications. Therefore, in this study, the differential
method was used in an attempt to develop a simple model for volatile retention kinetics for

microwave freeze-drying.
6.4.4.2 Determination of the kinetic parameters by the differential method

Both moisture content and the rate of volatile retention decreased with drying time,
with nearly perfect correlation between the two variables in all cases (Table 6.4a). This is
due to the fact that the ratio of volatile diffusivity to water diffusivity is strongly dependent
on moisture content, and decreases sharply with decreasing moisture content (Thij seen,
1979). On the other hand, the average temperature increases gradually with drying time,

while the rate of volatile retention decreases with the average temperature. The correlation

 

117

Table 6.3 Parameters required to fit the kinetic models of Eqs. (6.13) and (6. 14) for limo-
nene retention in the control (integral method).

Eq. (6.13)
P1

 

      

 

 

 

 

 

 

 

 

42.24

P2 5.25 2.1x108

P3 34.67 16.09 ll

P4 - 19720 -2928

P5 1915 34895

P6 21378 -1.56x108

P7 -2756 -16.83
Coeff. of determination (R2) 0.999 0.999

 

 

118

Table 6.4a Correlation between moisture content and the rate of volatile retention in
microwave freeze-dried samples.

 

 

 

 

 

 

 

Sample \ Volatile m l-hexanol l-decanol _
8% starch 0.97 1.00 0.99
8% starch with 0.97 1.00 0.99
4% cyclodextrin
8% starch with 0.99 0.99 0.98
4% sucrose

 

Table 6.4b Correlation between average temperature and the rate of volatile retention in
microwave freeze-dried samples.

SampleWolatile ---}
8% starch -O.96 -0.98 -0.98

8% starch with
4% cyclodextrin

8% starch with -0.88 -O.88 -0.89
4% sucrose

   
 

 

 

 

 

 

 

119

of the rate of volatile retention to the average temperature (Table 6.4b) is not as high as that
with moisture content (Table 6.4a). Therefore, the rate of volatile retention depends more
strongly on moisture content than on the average temperature of the sample. In addition,
the activation energy for volatile retention which involves physical processes such as
evaporation and diffusion is very small (Thijseen, 1979; Chung, 1984); therefore, the rate
constant is less dependent on temperature. The center temperature profiles of the samples
were always lower than room temperature during microwave freeze-drying. Due to this
low temperature gain, the effect of temperature on volatile retention is small and can be

neglected in comparison to the effect of moisture content during low temperature freeze-

drying.

A relationship for volatile retention as a function of drying time was developed by a

nonlinear regression analysis (Table 6.5) using the equation:
[A] = P1 exp(Pzt) + P3 (6.18)
The rate of volatile retention can be determined by differentiating Eq. (6.18) to obtain:

-d[A]
dt

 

= —Pllr’2 exp(Pzt) (6- 19)

Plots of the rate of volatile retention versus volatile concentration yielded linear profiles
for all samples (Figures 6.7, 6.8 and 6.9), confirming that volatile retention is a first order
reaction. As the equilibrium volatile retention is approached, the rate of volatile retention
approaches zero. Each volatile retention profile approached an equilibrium retention level

that depended on sample make-up and the type of volatile compound. The rate constants

120

Table 6.5 Empirical models for volatile concentration and rate of retention as a function
of drying time.

limonene [A] = 0.407 6104-0942!) + 0590 '995 #:‘l = 0.0171 exp(-0.042t)

     

 

8% starch hexanol [A] = 0.216CXP(-0-023t) + 0775 '975 11%.! = 0.005 CXp(-0.023t)

decanol [A]=O.186exp(-0.0llt)+0.818 .992 -d[A]
d:

 

= 0.0021 exp(—0.01 1t)

 

limonene [A]=0.106exp(—0.05t)+0.897 .994 —d[A] _
8% starch T: —0.0053exp(—0.05t)

 

 

with hexanol [A] =0.408exp(—0.0lt)+0.594 .998 —d[A] __
4% CD —dt —0.0041exp(—0.01t) ‘

decanol [A] = 0.073 exp(-0.023t) + 0.931 .974 _-% = 0.0017exp(-0.023t)

 

 

limonene [A] = 0.473 cxp(—0.045:) + 0.551 .986 —d[A]
8% starch d:
with hexanol [A] = 0.35 exp(-0.05t) + 0.66 -993 -d [A]

—= . .05: 5
4% sucrose d: 00175 CXPH) )

= 0.0213 exp(-0.045t)

 

 

decanol [A]=0.056exp(—0.062t)+0.944 .994 -d[A]
dt

 

 

 

 

= 0.0035 exp(—0.062t) .

 

   

—d[A]/dt.

121

 

0.010
0010-j
0.014;
0012-:
0.010:

0.000{

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIWIIIIIII

o limonene (k= 0. 042min-1, 12240 995)
o hexanol (k: 0.023min 1 .122: -0. 975)
A decanol (k=0.011min‘1,R2=0.992)
regression lines

   
     

 

 

Figure 6.7 Rate of volatile retention versus average volatile concentration during micro-

wave freeze-drying of the control.

122

 

 

 

 

 

0.006 ...........................
j o limonene (k: -0.050min'1, 122:0. 994)
1 n hexanol (k: 0.010min 1.R2=0.99a)
0,005- A decanol (k=0.023min-1, R2=0.974)
I — regression lines
0004
g .
c
\ .
'2 0.003:
T" .
0002-:
1
0.001-
0.000‘...fi....,.. ...... W
0.7 0.8 0.9 1.0

Figure 6.8 Rate of volatile retention versus average volatile concentration during micro-

wave freezedrying of samples containing 4% B-cyclodextrin.

123

 

   
  

 

 

 

0.022. ......... . ......... WWW, ......... , .........
3 o limonene (k=0.044min'1, R2=0.906)
0020-; n hexanol (k=0.050min'1. R2=0.993)
0 018; A decanol (k=0.061min"1, R2=0.994)
' I — regression lines
0016-:
.5 0.014{
'9 :
> 0.012:
<1: :
,5 0.010?
' 0.000-j
0006-:
0004-;
0002-: .
0.000d Ilrlr1rlrlrrrrrrrr
0.5 0.6

[A]

Figure 6.9 Rate of volatile retention versus average volatile concentration during micro-
wave freeze-drying of samples containing 4% sucrose.

 

124

and equilibrium volatile retention values, [AL], are shown in Table 6.6. The first order

volatile retention rate constants ranged from 0.01 to 0.06 min".

To enable the plot of the rate of volatile retention versus volatile concentration to be

extended through the origin, Eq. (6.3) was modified to obtain the volatile retention kinetics

 

by the integral method:
113—Lamina...) (6.21)
or
_d[A] - kdt (6.22)

([A1-[A1..,) ' .
Eq. (6.22) can be integrated to obtain:

([A] —[A1..,) _ kt

-1“ ([A ]o -' [A leq) '—

 

(6.23)

If the left hand side of Eq. (6.23) is plotted versus time, the reaction rate constant can be
obtained from the slope of the line, with the lines extending through the origin (Figures
6.10, 6.11 and 6.12).

125

Table 6.6 Volatile retention kinetics.

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

“I_Volaule —Reaction Rate Equation _k(min ) R2 ‘
limonene fi:fl_ _ 0 042([A]— —0 587) .042 .587
8% starch hexanol —(:i[tA] = 0023““ _ 0.769) .023 .769
decanol .5254] = 0.012(114] _ 0.825) .01 l .825
3% Stamh limonene .1284] = 0.050([A] _ 0.895) .050 .895
42:31) hexanol —(:1[;4] = 0.010([A] _ 0.586) .010 .586 .998
decanol .1254] = 0.023([A] _ 0.930) .023 .930 .974
8% Stan!) limonene .1284] = 0045“!“ -0.547) .044 .547 .986
with hexanol _d [A] .050 .657 .993
4% sucrose dt = 0.050([A] - 0.657)
decanol 1%]; = 0.061([14] _ 0.942) .061 9423* _ .994

 

 

 

 

 

 

 

126

 

 

 

3'5 . ' T ' I ' I ' I ' a ' l ' l u l . l r
j: 0 limonene (k=0.039min’ . R2=0.997)
0' a hexanol (k=0.023min"1, R2=0.992) .
.2. 3'0". A decanol (k=0.011min'1. R2=0.997) ° j
.3 I — regression lines 1
| L _.
O 2.5 . o
I-‘l - -
.3 3 j.
v 2.0d -
\ I n
A
O‘ ‘ °
.2. 1.51 .. _
< .
H ..
,i. 1.0-; . _
<12 . a
H . .
:1 - a .
5 0.5: ' j
| 2 . ‘

0.0 4 ‘

 

r 1 l ' I ' 1 ' I 1 fi' 1 ' I ' l '
0 10 20 30 40 50 60 70 80 90 100
Time (min)

Figure 6.10 First order volatile retention and rate constant of the control (integral
method).

127

2.5

 

o llimonrlane (kl=0.040min‘a, ill-30.996)l
. n hexanol (k=0.010min-1, R2=0.999)
. A decanol (k=0.021min-1. R2=0.984)
2'0: — regression lines

1.53 . _
1’0: o a in

0.5‘

—ln(([A]-[A]eq)/([A]o-[A]eq))

0-0 4' I I I I I I T. I
0 5 10 15 20 25 30 35 40", 45 50

 

 

 

Time (min)

Figure 6.11 First order volatile retention and rate constant of samples containing 4%
B—cyclodextrin (integral method).

128

 

 

 

 

3'0 I I I I I a I I I I I
2 j o limonene (k=0.041min' .R2=0.990)
c‘ 3 o hexanol (k=0.040min-1. R2=0.996) 3
,2 2,5- A decanol (k=0.054min-1. R2=0.994) . .J
:1, : — regreSsion lines 1)
I . 1
r-ofi 2.0“ A "
5.. 3 ° :
‘\’ - . :
’0“ 1.5‘ o 1
'2. A
< . .
H 1.0- -
| . I
.3 i - 3
t: 0.5“ ' —I
E. I ‘
I :
0'0 I 7 I I I l I I I I I
0 5 10 15 20 25 30 35 40 45 50, 55 50

Figure 6.12 First order volatile retention and rate constant of samples containing 4%

Time (min)

sucrose (integral method).

129

6.5 Conclusions

Three model foods which differed by only 4% in formulation had significantly different
temperature, moisture and volatile retention profiles during microwave freeze-drying. The
rate of volatile retention was strongly dependent on moisture content. If the effect of
temperature is assumed to be negligible for the case of low temperature microwave
freeze-drying, then the volatile retention can be modeled simply as a first order reaction by
the differential method. The first order volatile retention constants ranged from 0.01 to 0.06
min".

While there is excellent agreement between existing kinetic models and empirical data,
the approach seems to be purely one of curve fitting, and the parameters convey no physical
meaning. In addition, the number of parameters that must be determined to obtain the rate
constant is large, and further complicates analysis as well as industrial utilization. The
approach adopted here provides a simple yet accurate method for modeling the kinetics of
volatile retention during freeze-drying, and should be useful in process design and process

optimization.

CHAPTER 7
OVERALL CONCLUSIONS

1. Volatile Retention:

(1)

(2)

(3)

(4)

(5)

Sample composition significantly influenced volatile retention. For example, the
samples containing B-cyclodextrin had significantly higher volatile retention,

probably due to inclusion complex formation.

Freezing temperature did not influence volatile retention in conventionally freeze-
dried starch gel systems. However, the highest freezing temperature (-15°C) gen-

erally led to higher volatile retention during microwave freeze-drying.

Microwave freeze-drying resulted in significantly higher volatile retention than

conventional freeze-drying.

Operating the chamber at 1 and 2 torr did not cause any significant differences in

volatile retention.

The retention of l-decano] was higher than that of l-hexanol, and'appeared to be

based on selective diffusion.

2. Microstructure:

(1)

Sample composition was a significant factor in the rnicrostructure of freeze-dried
products. The control (8% starch) had a uniform three-dimensional network gel
structure. There were deposits of aggregated materials which sealed the pores and
void spaces on the cross-section channel structure of samples containing 4%
B—cyclodextrin. The samples containing 4% sucrose featured broken flakes

deposited on the fiber structure.
130

(2)

(3)

131

The samples frozen in liquid nitrogen had smaller pore distribution than those frozen
at -15°C and -60°C. The average pore diameters of samples freeze-dried by

microwave were larger than those freeze-dried conventionally.

Generally, samples freeze-dried at 2 torr had larger pore sizes and more broken

structures than those at 1 torr.

3. Kinetics of Volatile Retention:

(1)

(2)

(3)

(4)

Three starch gel systems which differed by 4% in formulation had significantly
different temperature, moisture and volatile retention profiles during microwave

freeze-drying.

The individual volatile compounds in each sample reached an equilibrium final
retention level, and the rate of volatile retention strongly depended on moisture

content.

The volatile retention was modeled simply as a first order reaction, using the dif-

ferential method. The rate constants ranged from 0.01 to 0.06 min".

The differential method is a simpler method than the integral method in determining

volatile retention kinetics during freeze-drying.

CHAPTER 8
SUGGESTIONS FOR FUTURE RESEARCH

( 1) Evaluate the kinetic models developed here for volatile or other quality retention

in actual food products during microwave freeze-drying.

(2) Develop a method to measure dielectric loss factors and dielectric constants of

samples during microwave freeze-drying.

(3) Evaluate the effect of sample composition on dielectric properties and drying rate

of microwave freeze-dried products.

(4) Evaluate the effect of sample dimension on drying rate and volatile retention of

microwave freeze-dried samples.

132

CHAPTER 9
BIBLIOGRAPHY

Ang, T.K., Ford, J .D. and Pei, D.C.T. 1977a. Microwave freeze-drying of foods: a theoretical
investigation. Int. J. Heat Mass Transfer, 20: 517-526.

Ang, T.K., Pei, D.C.T. and Ford, J .D. 1977b. Microwave freeze drying: an experimental
investigation. Chem. Eng. Sci., 32: 1477-1489.

Ang, T.K., Ford, J .D. and Pei, D.C.T. 1978. Optimal modes of operation for microwave
freeze drying of food. J. Food Sci., 43: 648-649.

Arsem, H.B. and Ma, Y.H. 1985. Aerosol formation during the microwave freeze dehy-
dration of beef. Biotechnol. Progress, 1: 104-110.

Arsem, H.B. and Ma, Y.H. 1990. Simulation of a combined microwave and radiant freeze
dryer. Drying Technol., 8(5): 993-1016.

Asmussen, J ., Lin, H.H., Manring, B. and R. Fritz. 1987. Single-mode or controlled mul-
timode microwave cavity applicators for precision materials processing. Rev. Sci.
Instrum., 58(8): 1477-1486.

Bartholomai, G.B., Brennan, 1.6. and Jowitt, R. 1975. Influence of cellulose and starch on
the retention of volatiles during freeze-drying of a liquid extract of mushroom.
Lebensm. - Wiss. u. - Technol., 8: 174-176.

Bello, R.A., Luft, J .H. and Pigott, GM. 1982. Ultrastructural study of skeletal fish muscle
after freezing at different rates. J. Food Sci., 47: 1389-1394.

Chang, TN. and Ma, Y.H. 1985. Application of optimal control strategy to hybrid microwave
and radiant heat freeze drying system. In: Drying’85. Hemisphere Pub. Corp.,
Washington, 249-253.

Chang, Y.-I. and Reineccius, GA. 1990. Interaction of B—cyclodextrin with enantiomers of
limonene and carvone. J. Food Sci., 55: 1686-1688, 1695.

Chirife, J., Karel, M. and Flink, J. 1973. Studies on mechanisms of retention of volatile in
freeze-dried food models: the system PVP-n-propanol. J. Food Sci., 38: 671-674.

Chirife, J. and Karel, M. 1973. Contribution of adsorption to volatile retention in a
freeze-dried food model containing PVP. J. Food Sci., 38: 768-771.

Choi, Y. and Villota, R. 1989a. Development of kinetic models for methionine degradation
in fortified soybean model system. J. Food Process. Preserv., 13: 355-384.

133

Theater-37:37“ “I
I 1.:

 

134

Choi, Y. and Villota, R. 1989b. Prediction of methionine retention during extrusion pro-
cessing of fortified soybean model systems. J. Food Process. Preserv., 13: 385-406.

Cohen, J. and Yang, T.C.S. 1992. Microwave-assisted freeze-drying of green peas. Paper
no. 174, presented at 53th Annual Meeting of Institute of Food Technologists, New
Orleans, Louisiana, June 20-24.

Consolacion, RI. and Jelen, P. 1986. Freeze texturization of proteins: effect of the alkali,
acid and freezing treatments on texture formation. Food Microstructure, 5: 33-39.

Copson, D.A. 1958. Microwave sublimation of foods. Food Technol., 12: 270-272.

Copson, D.A. 1962. Microwave Heating in Freeze-Drying, Electronic Ovens, and Other
Applications. AVI Publishing Company, INC. Westport, Connecticut, USA.

Crowther, A., Wilson, LA. and Glatz, CE. 1980. Effects of processing on adsorption of
off-flavors onto soy protein. J. Food Processing Eng., 4: 99-115.

Chung, S. 1984. Mechanisms of Volatile Retention During Conventional and Microwave
Heating. MS Thesis, University of Illinois Urbana-Champaign, USA.

Decareau, R.V. 1985. Freeze-drying. In: Microwaves in the Food Processing Industry.
Decareau, R.V. (Ed.) Academic Press, Inc., Orlando Florida, USA.

deMan, J .M., deMan, L. and Gupta, S. 1986. Texture and rnicrostructure of soybean curd
(tofu) as affected by different coagulants. Food Microstructure, 5: 83-89.

Dyer, DP. and Sunderland, J .E. 1966. Bulk and diffusional transport in the region between
molecular and viscous flow. Int. J. Heat Mass Transfer, 9: 519-526.

Dyer, D.F., Carpenter, D.K. and Sunderland, J .E. 1966. Equilibrium vapor pressure of frozen
bovine muscle. J. Food Sci., 32: 196-201.

Dyer, DP. and Sunderland, J .E. 1968. Heat and mass transfer mechanisms in sublimation
dehydration. J. Heat Transfer, 90C: 279-384.

Dyer, DP. and Sunderland, J .E. 1971. Freeze-drying of bodies subject to radiation boundary
conditions. J. Heat Transfer, 93C: 427-431.

Dziezak, J .D. 1988. Microencapsulation and encapsulated ingredients. Food Technol.,
42(4): 136-153.

Etzel, MR. and King, C.J. 1980. Retention of volatile components during freeze drying of
substances containing emulsified oils. J. Food Technol., 15: 577-588.

Flink, J. and Gejl-Hansen, F. 1972. Retention of organic volatiles in freeze-dried carbo-
hydrate solutions: microscopic observations. J. Agric. Food Chem., 20(3): 691-694.

135

Flink, J. and Karel M. 1970a. Retention of organic volatiles in freeze-dried solutions of
carbohydrates. J. Agric. Food Chem, 18(2): 295-297.

Flink, J. and Karel M. 1970b. Effects of process variables on retention of volatiles in
freeze-drying. J. Food Sci., 35: 444-447.

Flink, J .M. 1975a. Process conditions for improved flavor quality of freeze dried foods. J.
Agric. Food Chem., 23(6): 1019-1026.

Flink, J .M. 1975b. The retention of volatile components during freeze drying: a structurally
based mechanism. In: Freeze Drying and Advanced Food Technology. Goldbith, S.A.,
Rey, L. and Rothmayr, W.W. (Ed.) Academic Press, San Francisco, USA.

Franzen, K., Singh, R.K. and Okos, MR. 1990. Kinetics of nonenzymatic browning in dried
skim milk. J. Food Eng, 1 1: 225-239.

Gero, L. and Smyrl, T.G. 1982. Behavior of low molecular weight organic acids during
freeze drying. J. Food Sci., 47: 954-957.

Gerschenson, L.N., Bartholomai, GB. and Chirife, J. 1979. Retention of volatiles during
freeze drying of tomato juice. J. Food Technol., 14: 351-360.

Gould, J .W. and Kenyon, EM. 1971. Gas discharge and electric field strength in microwave
freeze-drying. J. Microwave Power, 6(2): 151-167.

Harper, J .C., Chichester, CO. and Roberts, TE. 1962. Freeze-drying of foods. Agricultural
Eng., 43: 78-81, 90.

Hill, J .E. and Sunderland, J .E. 1971. Sublimation-dehydration in the continuum, transition
and free molecule flow regimes. Int. J. Heat Transfer, 14: 625-638. ’

Jackson, S., Rickter, S.L. and Chichester, CO. 1957. Freeze-drying of fruit. Food Technol.,
1 1: 468-470.

Kan, B. and de Winter, F. 1968. Accelerating freeze-drying through improved heat transfer.
Food Technol., 22: 1269-1278.

King, C.J., Kieckbusch, T.G. and Greenwald, CG. 1984. Food quality factors in spray
drying. In: Advances in Drying. Vol. 3. Mujumdar, A.S. (Ed.) Hemisphere Co., New
York, USA.

Komolprasert, V. and Ofoli, R.Y. 1987. Mathematical modeling of microwave heating by
the method of dimensional analysis. J. Food Process. Preserv., 13: 87-106.

Koonz, CH. and Ramsbottom, J .M. 1939. A method for studying the histological structure
of frozen products. 1. Poultry. Food Research, 4: 117-128.

136

Labuza, T.P. 1979. A theoretical comparison of losses in foods under fluctuating temperature
sequences. J. Food Sci., 44: 1162-1168.

Laing, B.M., Schlueter, D.L. and Labuza, T.P. 1978. Degradation kinetics of ascorbic acid
at high temperature and water activity. J. Food Sci., 43: 1440-14-43.

Levenspiel, O. 1972. Chemical Reaction Engineering. 2nd Ed. John Wiley & Sons, INC.
N .Y., USA.

Li-Shing-Ta, B. and Jelen, P. 1987. The rnicrostructure and rehydration properties of phoenix
oyster mushroom (Pleurotus Sajor-Caju) dried by three alternative process. Food
Microstructure, 6: 135-142.

Ma, Y.H. and Arsem, H. 1982. Low pressure sublimation in combined radiant and micro-
wave freeze drying. In: Drying’82. Hemisphere Pub. Corp., Washington, 196-200.

Ma, Y.H. and Peltre, P.R. 1975a. Freeze dehydration by microwave energy: Part 1. Theo-
retical investigation. AIChE J ., 21(2): 335-344.

Ma, Y.H. and Peltre, P.R. 1975b. Freeze dehydration by microwave energy: Part 2.
Experimental study. AIChE J ., 21(2): 344-350.

Margaritis, A. and King, C.J. 1971. Factors governing terminal rates of freeze-drying of
poultry meat. Chem. Eng. Progr. Symp. Ser., 67(108): 112-121.

McPeak, D.W., Idziak, E. and Smyrl, T.G. 1987. The effect of freezing rate on the retention
of volatile fatty acids during freeze drying of solutions containing high molecular
weight solutes. Can. Inst. Food Sci. Technol. J ., 20(3): 176-178.

Miller, H.L. and Sunderland, J .E. 1963. Thermal conductivity of beef. Food Technol., 17(4):
124-126.

Mills, OE. and Solms, J. 1984. Interaction of selected flavor compounds with whey proteins.
Lebensm. -Wiss. -u. -Technol., 17: 331-335.

Mishkin, M., Saguy, I. and Karel, M. 1984. A dynamic test for kinetic models of chemical
changes during processing: ascorbic acid degradation in dehydration of potato. J. Food
Sci., 49: 1267-1270, 1274.

Mizrahi, S. and Karel, M. 1977. Accelerated stability test of moisture-sensitive products in

permeable packages at high rates of moisture gain and elevated temperatures. J. Food
Sci., 42: 1575-1579, 1589.

Mudgett, R.E., Mudgett, D.R., Goldblith, S.A., Wang, D.I.C. and Westphal, W.B. 1979.
Dielectric properties of frozen meats. J. Microwave Power, 14(3): 209-216.

am." I . in-p,

 

137

Mudgett, RE. 1985. Dielectric properties of food. In: Microwaves in the Food Processing
Industry. Decareau, R.V. (Ed.) Academic Press, Inc., Orlando Florida, USA.

Mudgett, RE. 1986. Microwave properties and heating characteristics of foods. Food
Technol., 40(6): 84-93.

Ofcarcik, RP. and Burks, EB. 1974. Carbonyl retention in model systems and Bermuda
onion juice during lyophilization: effect of simple carbohydrates, binary carbohydrate
mixtures and sucrose inversion. J. Food Sci., 39: 350-353.

Ofoli, R.Y. and Komolprasert, V. 1988. On the thermal modeling of foods in electromagnetic
fields. J. Food Process. Preserv., 12: 219-241.

O’Keefe, S.E., Resurreccin, A.P., Wilson, LA. and Murphy, RA. 1991. Temperature effect
on binding of volatile flavor compounds to soy protein in aqueous model system. J.
Food Sci., 56: 802-806.

Peltre, R.P., Arsem, H.B. and Ma, Y.H. 1977. Applications of microwave heating to freeze
drying: perspective. AIChE Sym. Ser., 73(163): 131-133.

Petersen, E.E., Lorentzen, J. and Flink, J. 1973. Influence of freeze-drying parameters on
the retention of flavor compounds of coffee. J. Food Sci., 38: 119-122.

Pszczoal, DE. 1988. Production and potential food applications of cyclodextrins. Food
Technol., 42(1): 96-100.

Rey, LR. and Bastien, M.C. 1962. Biophysical aspects of freeze-drying. In: Freeze-Drying
of Foods. Fisher, F.R. (Ed.) National Academy of Science - National Research Council,
Washington, DC. USA. " ,

Risman, PO. and Bengtsson, NE. 1971 . Dielectric properties of food at 3 GHz as determined
by a cavity perturbation technique. 1. Measuring technique. J. Microwave Power, 6(2):

101-123.

Roebuck, B.D., Goldblith, SA. and Westphal, W.B. 1972. Dielectric properties of
carbohydrate-water mixtures at microwave frequencies. J. Food Sci., 37: 199-204.

Rosenberg, M., Kopelrnan, I.J., Talmon, Y. 1985. A scanning electron microscopy study of
microencapsulation. J. Food Sci., 50: 139-144.

Rosenberg, U. and Bogl, W. 1987. Microwave thawing, drying and baking in the food
industry. Food Technol., 41(6): 85-91.

Rosenberg, M. Talmon, Y. and Kopelrnan, 1.]. 1988. The microstructure of spray-dried
microcapsules. Food Microstructure, 7: 15-23.

138

Saguy, 1., Mizrahi, R., Villota, R. and Karel, M. 1978a. Accelerated method for determining
the kinetic model of ascorbic acid loss during dehydration. J. Food Sci., 43: 1861-1864.

Saguy, 1., Kopelman, U. and Mizrahi, S. 1978b. Simulation of ascorbic acid stability during
heat processing and concentration of grapefruit juice. J. Food Process Eng, 2: 213-225.

Saguy, I. and Karel, M. 1980. Modeling of quality deterioration during food processing and
storage. Food Technol., 34(2): 78-85.

Shaath, NA. and Azzo, NR. 1989. Design of flavors for the microwave oven. The Delta T
theory. In: Thermal Generation of Aromas. Parliment, T.H., McGorrin R]. and Ho,
C-T. (Ed.) ACS Symp. Series.

Slater, LE. 1975. Advanced technology: innovating with microwaves. Food Eng, June:
51-53.

Sunderland, J .E. 1980. Microwave freeze drying. J. Food Process Eng, 4: 195-212.

Sunderland, J .E. 1982. An economic study of microwave freeze-drying Food Technol.,
36(2): 50-56.

Tetembaum, SJ. and Weiss, J .A. 1979. Microwave breakdown of water vapor. IEEE
Transactions of Plasma Science VOL PS-7, No.2, June: 109-111.

Thij seen, H.A.C. 1975. Effect of process conditions in freeze drying on retention of volatile
components. In: Freeze Drying and Advanced Food Technology. Goldbith, S.A., Rey,
L. and Rothmayr, W.W. (Ed.) Academic Press, San Francisco, USA.

Thijseen, H.A.C. 1979. Optimization of process conditions during dryingwith regard to
quality factors. Lebensm. -Wiss. u. -Technol., 12: 308-317. '

Thompson, DR. 1982. The challenge in predicting nutrient changes during food processing.
Food Technol., 36(2): 97-108,1 15.

Villota, R. and Karel, M. 1980. Prediction of ascorbic acid retention during drying. II.
Simulation of retention in a model system. J. Food Process. Preserv., 4: 141-159.

Villota, R. and Hawkes, J .G. 1986. Kinetics of nutrients and organoleptic changes in foods
during processing. In: Physical and Chemical Properties of Food. Okos, M.R. (Ed.)
American Society of Agricultural Engineers, St. Joseph, Michigan, USA.

Wanninger JR., LA. 1972. Mathematical model predicts stability of ascorbic acid in food
products. Food Technol., 25(6): 42-45.

Woodward, SA. and Cotterill, O.J. 1985. Preparation of cooked egg white, egg yolk, and
whole egg gels for scanning electron microscopy. J. Food Sci., 50: 1624-1628.

 

139

Wyler, R. and Solms, J. 1981. Inclusion complexes of potato starch with flavor compounds.
In: Flavor’81 3rd Weurman Symposium Proceedings of the International Conference.
Schreier, P. (Ed.) Walter de Gruyter, Berlin, N .Y., USA.

140

APPENDICES

141

Appendix A. Molecular formulae and chromatography of volatile compounds.

 

 

 

 

 

 

 

 

 

 

 

 

 

Volatile Formula M.W. Density Retention Time (min) F
limonene CmI-I“, Q 136 0.844 5.94 1.098
l-hexanol CH3(CH2)4CHZOH 102 0.822 8.85 0.880
l-decanol CH3(CH,_)8CH20H 158 0.829 15.83 0.996
l-octanol CH3(CH2)6CHZOH 130 0.830 12.49 1.000

Response Factor (F) = (peak area of volatile)(density of internal standard)

 

(peak area of internal standard)(density of volatile)

(weight of internal standard)(peak area of volatile)
F(peak area of internal standard)

 

Volatile Content =

volatile content of freeze-dried sample
original volatile content of frozen sample

 

Volatile Retention = x 100%

Example: Gas chromatography of 0.511.] of 400 ppm volatile solution. (l-octanol as an
internal standard)

E

 

 

 

 

 

 

__ “I 133;
QRERZ

RI QRER TYPE AREAZ
1.38 2.1981E+87 388 57.614
584 1.35 8533288 TBP 6.664
1.49 13586 TVB 8.836
1.76 775348 T89 3.841
_ 3.33 1.21566+87 TVP 31.978
335 5.94 1??698 P8 8.467
g 8.85 138768 PB 8.365
4,53 12.49 159910 PB 0.418
15.83 158568 PB 8.417

134°

TOTAL RREQ= 3.8813E+87
HUL FRCTOR= 1.8888988

IIIIII

.1 ”out
w)

 

STOP

142

Appendix B. A modified Likens-Nickerson apparatus for simultaneous distillation/solvent

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

extraction.
I"?
I—H
— carbon dioxide - acetone condenser
water out
logging
4’
{\‘1 r: , { 7 ’—-—.
. I
u >1 m
( ‘ )
II . i
I I
water in
/
2-methy1butane

Sample

143

Appendix C. Basic program to record voltage signals through the data acquisition system

during microwave freeze-drying.

’FILE NAME: SUDER.5

’This program configures the Data Acquisition/Control unit HP3852A

’to scan a list of channels, wait for a period of time and scan again.

’This procedure could be repeated as many times as the user wants.

’The number of scans could be specified in line 100 (Example: 1 to 60
’indicates 60 scans). The time delay between two scans could be set

’in line 120. Also, the channel list could be:

’RD#(1)= surface temperature, RD#(2) = center temperature, RD#(3) = pressure,
’RD#(4) = nitrogen flow rate, RD#(S) = input power, RD#(6) = reflected power.
’The measured DC voltages are saved in a file called OUTPUTS.

10 CLS

20 PRINT "EXPERIMENT IN PROGRESS: DO NOT TOUCH"

30 PRINT "STOP PROCESS BY PRESSING CONTROL BREAK"

4O OPEN "OUTPUTS" FOR OUTPUT AS #5

50 OPEN "GPIBO" FOR OUTPUT AS #1

60 OPEN "GPIBO" FOR INPUT AS #2

70 PRINT #1, "ABORT"

80 PRINT #1, "STATUS"

90 PRINT #1, "REMOTE 9"

100 FOR J% = 1 TO 961

110 TIME#= 1% * 10

120 PRINT #1, "OUTPUT 9#9; WAIT 9. 745" ‘ _
130 PRINT #1, "OUTPUT 9#31,CONFMEAS DCV, 100-105, USE 600’
140 PRINT #1, "ENTER 9"

150 FORI%=1TO6

160 INPUT #2, RD#(I%)

170 NEXT 1%

180 PRINT 5#, TIME#, RD#(l), RD#(2), RD#(3), RD#(4), RD#(S), RD#(6)
190 NEXT J %

200 CLOSE #5

210 END

144

Appendix D. Basic program to compute temperature, pressure, nitrogen flow rate, and
microwave power during microwave freeze-drying.

’FILE NAME: NCHEN.5

’TO CALCULATE AVERAGE OUTPUT SIGNALS PER MINUTE
’TIME! = Time(min), NT!(l) = Surface temperature (C), NT!(2) = Pressure (torr), ’NT!(3)
= N2 Flow Rate (cm3/min), NT!(4) = Power (W)

10 OPEN "B:SEM2CY.2" FOR INPUT AS #5

20 OPEN "B:SEM2CY2.PRN" FOR OUTPUT AS #6

30 FOR 1% = 1 TO 200

40 NT!(1)= 0

41 NT!(2) = 0

42 NT!(3) = 0

43 NT!(4) = 0

50 FOR J% = 1 TO 6

60 INPUT #5, TIME, RD!(1), RD!(2), RD!(3), RD!(4), RD!(S), RD!(6)
70 RT!(1)=10*RD!(1)

80 RT!(2) = 4.138661 + 2.108456 * RD!(3) + .31180716 * RD!(3)"2
90 RT!(3) = .0863502 * RD!(4)"7 - 1.694572 * RD!(4)"6+ 13.2883 * RD!(4)"5 - 53.28604
* RD!(4)"4 + 116.55 * RD!(4)"3 - 133.3751 * RD!(4)"2+95.01064 * RD!(4) +0.03045233#
100 RT!(4) = (RD!(S) - RD!(6)) * 30

110 NT!(1)=NT!(1)+RT!(1)

120 NT!(2) = NT!(2) + RT!(2)

130 NT!(3) = NT!(3) + RT!(3)

140 NT!(4) = NT!(4) + RT!(4)

150 NEXT J%

160 TIME! = 1% - 1

170 NT!(1)=NT!(1)/6

180 NT!(2) = NT!(2) / 6

190 NT!(3) = NT!(3) / 6

200 NT!(4) = NT!(4) / 6

210 PRINT #6, TIME!, NT!(l), NT!(2), NT!(3), NT!(4)

220 NEXT 1%

230 CLOSE #6

240 CLOSE #5

250 END

145

Appendix E. Typical data obtained during microwave freeze-drying.

SAMPIEE: 8% STARCH & 4% B—CYCLODEXTRIN W/ 500 PPM VOLATILES FROZEN
AT -15 c

CONDITION: 10w, 2 TORR, 50 MIN

WEIGHT = 4.458g ---> 0.603g

TIME SURFACE PRESSURE N2 FLOW Power CAVITY PROBE
(min) TEMP. (torr) RATE (W) LENGTH LENGTH
(°C)

 

 

 

(cm3/min) Lc (cm) Lp (m)
-24.30 .54 28.27 1.71 13.892 24.32
-17.37 .72 28.03 10.11 13.892 24.32
- 10.58 .93 28.04 9.87 13.892 24.32
-9. 18 .95 27.97 9.75 13.892 24.32
-8.42 .95 27.93 10.06 13.894 24.32
-7.54 .96 27.88 10.22 13.894 24.32
~6.74 .96 27.84 10.03 13.894 24.32
-5.84 .96 27.82 1.012 13.894 24.32
-4.56 .97 27.82 10.26 13.894 24.32
-3.87 .97 27.82 10.26 13.885 24.32
-3. 15 .97 27.79 9.85 13.865 24.32
-2.38 . 27.80 9.29 13.854 24.32
-1.57 27.75 9.16 13.836 24.32
-.60 27.73 9.57 13.824 24.32
.66 27.71 9.46 13.815 24.32
1.66 27.70 9.46 13.812 24.32
2.65 27.70 9.56 13.808 24.32
3.43 27.68 9.57 13.805 24.32
4.45 27.68 9.73 13.803 24.32
5.36 27.65 9.75 13.803 24.32 '
6.27 27.64 9.79 13.832 24.32
6.63 38.60 9.45 13.843 24.32
7.61 27.63 9.57 13.856 24.32
8.18 27.64 9.68 13.856 24.32
8.70 27.64 9.30 13.872 24.32
8.77 27.63 9.71 13.882 24.32
10.18 27.64 9.56 13.892 24.32
11.00 27.64 9.64 13.912 24.32
11.96 27.63 9.45 13.925 24.32
12.93 27.63 10.06 13.927 24.32
14.52 27.63 9.81 13.928 24.32
15.27 27.63 9.73 13.930 24.32
16.20 27.71 9.80 13.932 24.32
17.47 27.65 9.57 13.933 24.32
18.59 28.05 9.88 13.934 24.32
19.32 27.75 10.01 13.935 24.32
19.84 27.72 9.90 13.936 24.32
20.02 27.68 9.88 13.935 24.32

Appendix E. (cont’d)

21.27
21.20
22.22
22.62
23.69
24.37
24.87
24.14
25.85
27.30
28.03
28.93
31.35

1.82
1.92
2.28
2.32
2.07
2.05
2.03
2.02

1.99
1.97
1.96
1.95

27.68
29.47
31.00
31.25
29.78
29.73
29.72
29.71
29.68
29.69
29.66
29.71
29.67

146

10.03
9.75
9.73
10.00
9.97
29.73
29.72
29.71
29.68
29.69
29.66
27.71
29.67

13.933
13.934
13.936
13.936
13.936
13.936
13.938
13.938
13.937
13.937
13.938
13.937
13.936

24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32
24.32

147

Appendix F. Tukey’s test for determining the significance of factors influencing volatile
retention and final moisture content.

Table F. 1 The effect of sample treatment and type of volatile compound on volatile retention

148

of conventionally freeze-dried samples.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Treatment Volatile Compound Retention (%)
control limonene 1.0J
control l-hexanol 54.8l
control l-decanol 83.3DE

+ 2% B-cyclodextrin limonene 81.5”:
+ 2% B-cyclodextrin l-hexanol 79.1EFG
+ 2% [El-cyclodextrin l-decanol 90.1BC
+ 4% B-cyclodextrin limonene 95.3A
+ 4% B-cyclodextrin l-hexanol 85.8”
+ 4% B-cyclodextrin l-decanol 92.1AB
+ 2% sucrose limonene 1.2J
+ 2% sucrose 1-hexanol 60.1H
+ 2% sucrose l-decanol 77.4FG
+ 4% sucrose limonene 1.7J
+ 4% sucrose l-hexanol 59.5H
+ 4% sucrose l-decanol 76.3G

 

where, control = 8% pregelatinized corn starch with 50 ppm volatile solution.
A" Different superscripts within a column indicate significant difference (P<0.05).

 

149

Appendix F. (cont’d)
Table F2 The effect of freezing temperature and type of volatile compound on volatile
retention of conventionally freeze-dried samples.

 

 

 

 

 

 

 

 

 

 

Temperature (°C) Volatile Compound Retention (%)
-15 limonene 37.4C
-15 l-hexanol 68.2B
-15 l-decanol 83.6A
-40 limonene 35.5C
-40 l-hexanol 69.1B
~40 l-decanol 84.3A
-198 limonene 35.4C
-198 1-hexanol 66.2B
-198 l-decanol 83.6A

 

 

 

 

 

“'5 Different superscripts within a column indicate significant difference (P<0.05).

150

Appendix F. (cont’d)
Table F3 The effect of sample treatment and freezing temperature on final moisture content
of microwave freeze-dried samples.

 

 

 

 

 

 

 

 

 

 

Sample Treatment Temperature (°C) Moisture Content (%)
control -15 6.50BC
control -60 4.55CD
control -l98 7.68B

+ 4% B—cyclodextrin -15 6.48BC

+ 4% B-cyclodextrin -60 3.23D

+ 4% B-cyclodextrin -l98 7.95B
+ 4% sucrose -15 4.450)
+ 4% sucrose -60 5.6580)
+ 4% sucrose -198 12.52A

 

 

 

 

 

where, control: 8% pregelatinized corn starch with 500 ppm volatile solution.
DDifferent superscripts within a column indicate significant difference (P<0. 05).

Table F. 4 The effect of sample treatment and chamber pressure on final moisture content
of microwave freeze-dried samples.

 

 

 

 

 

 

 

Sample Treatment Pressure (torr) Moisture Content (%)
control 1 6.92AB
control 2 5.57B
+ 4% B—cyclodextrin 1 7.43AB
+ 4% B-cyclodextrin 2 4.33B
+ 4% sucrose 1 9.82"
+ 4% sucrose 2 5.27B

 

 

 

 

 

where, control: 8% pregelatinized corn starch with 500 ppm volatile solution.
'BDifferent superscripts within a column indicate significant difference (P<0. 05).

151

Appendix F. (cont’d)
Table F5 The effect of sample treatment and type of volatile on volatile retention of
microwave freeze-dried samples.

 

 

 

 

 

 

 

 

 

 

Sample Treatment Volatile Compound Retention (%)
control limonene 52.9G
control l-hexanol 80.6D
control l-decanol 86.2C

+ 4% B—cyclodextrin limonene 95.7A
+ 4% B-cyclodextrin l-hexanol 85.0C
+ 4% B—cyclodextrin 1-decanol 93.4B
+ 4% sucrose limonene 568':
+ 4% sucrose l-hexanol 70.3E
+ 4% sucrose l-decanol 94.2AB

 

 

 

 

 

where, control = 8% pregelatinized corn starch with 500 ppm volatile solution.
“3 Different superscripts within a column indicate significant difference (P<0.05).

152

Appendix F. (cont’d)
Table R6 The effect of freezing temperature and type of volatile on volatile retention of
microwave freeze-dried 8% starch gel.

 

 

 

 

 

 

 

 

 

 

Temperature (°C) Volatile Compound Retention (%)
-15 limonene 65.7D
-15 l-hexanol 89.1“
-15 l-decanol 89.0“
-60 limonene 53.3E
-60 l-hexanol 77.0C
-60 l-decanol 87.4“
-198 limonene 39.7F
-198 l-hexanol 75.6C
-l98 l-decanol 82.2B

 

 

 

 

 

“'13 Different superscripts within a column indicate significant difference (P<0.05).

Table F7 The effect of chamber pressure and type of volatile on volatile retention of
microwave freeze-dried 8% starch gel.

 

 

 

 

 

 

 

 

 

 

Pressure (torr) Volatile Compound Retention (%)
1 limonene 55.8C
1 l-hexanol 82.5“B
1 l-decanol 87.8“
2 limonene 49.9C
2 1-hexanol 78.6B
2 l-decanol 84.7AB

 

“'C Different superscripts within a column indicate significant difference (P<0.05).

 

153

Appendix F. (cont’d)
Table F8 The effect of freezing temperature and type of volatile on volatile retention of
microwave freeze-dried 8% starch gel with 4% B-cyclodextrin.

 

 

 

 

 

 

 

 

 

 

 

Temperature (°C) Volatile Compound Retention (%)
-15 limonene 97.0“
-15 l-hexanol 94.4“B
-15 l-decanol 95.7“
-60 limonene 93. 1“B
-60 l-hexanol 79.9C
-60 l-decanol 94.7“
- 198 limonene 97. 1“
-198 l-hexanol 80.8C
-198 l-decanol 89.8B

 

 

 

 

“'C Different superscripts within a column indicate significant difference (P<0.05).

Table F9 The effect of chamber pressure and type of volatile on volatile retention of
microwave freeze-dried 8% starch gel with 4% B-cyclodextrin.

 

 

 

 

 

 

 

Pressure (torr) Volatile Compound Retention (%)
1 limonene 95.2“
1 l-hexanol 87.3"
1 1-decanol 95.2“
2 limonene 96.2“
2 l-hexanol 82.7C
2 l-decanol 91 .6“B

 

 

 

 

 

“'C Different superscripts within a column indicate significant difference (P<0.05).

Appendix F. (cont’d)

Table F. 10 The effect of freezing temperature and type of volatile on volatile retention of

154

microwave freeze-dried 8% starch gel with 4% sucrose.

 

 

 

 

 

 

 

 

 

 

 

 

 

Temperature (°C) Volatile Compound Retention (%)
-15 limonene 72.7B
-15 l-hexanol 73.3B
-15 l-decanol 92.0“
-60 limonene 58.9C
-60 l-hexanol 53.7C
-60 1-decanol 94.1“
-198 limonene 38.9D
-198 l-hexanol 74.0B
-l98 l-decanol 96.5“

 

“‘0 Different superscripts within a column indicate significant difference (P<0.05).

Table FM The effect of chamber pressure and type of volatile on volatile retention of

microwave freeze-dried 8% starch gel with 4% sucrose.

 

 

 

 

 

 

 

 

 

 

Pressure (torr) Volatile Compound Retention (%)
1 limonene 56.5D
1 1-hexanol 74.7B
1 1-decanol 94. 1“
2 limonene 57. 1°
2 l-hexanol 66.0C
2 l-decanol 94.3“

 

“'0 Different superscripts within a column indicate significant difference (P<0.05).

 

 

                                                 

NSTRTEU

III!IIIIINIIIIIIIIIIIIIIll