EXPLORATION ON THE STABILITY OF
INTERMEDIATE MOISTURE FOODS
by
J. Antonio Torres
B.S. Mathematics, Cath. U. of Chile, 1970
B.S. Industrial and Chemical Engineering, Cath. U. of Chile, 1973
Sc.M. Food Science, Mass. Institute of Technology, 1978
Submitted in Partial Fullfillment
of the Requirements for the Degree of
Doctor of Philosophy
at the
Massachusetts Institute of Technology
May,
1984
Z~I3ACHSETS iNSTij TE
OF TECHNOLOGY
Q
J. Antonio Torres
JUN 261984
LIBRARI E3
The author hereby grants to
to distribute copies of thi
Signature
of
T. permission to reproduce and
e is document in whole or in part.
Autho
r
Certified
np of Nutrition & Food Science
84
8a
by
M rcus Karel, Thesis Supervisor
Accepted by
Stevel Tannenbaum, Chairman of the
Department Committee on Graduate Students
This doctoral thesis has been examined by a Committee
of the Department of Nutrition and Food Science as follows:
Professor Robert S. Langer
Chairtw
Professor Marcus Karel
Thesis Supervisor
7Th*O
Professor ChoKyun Rha
Professor Anthony J. Sinskey
w~g'
-
-~
EXPLORATION ON THE STABILITY OF
INTERMEDIATE MOISTURE FOODS
by
J. Antonio Torres
Submitted to the Department of Nutrition and Food Science on
May 18, 1984 in partial fullfillment of
the requirements for the Degree of
Doctor of Philosophy in
Food Science and Technology
ABSTRACT
Microbial
(IMF's)
has
environmental
been
stability
found
conditions.
localized surface
to
of
be
intermediate
affected
Temperature
condensations resulting
by
moisture
foods
unsteady
state
fluctuations
induce
in potential
surface
outgrowth before moisture diffuses into the food product.
Due to
is not
always
organoleptic, cost and/or
safety restrictions it
possible to include safety margins by lowering water activity, or
increasing preservative content in IMF formulations.
The solution explored in our current research has
to consider surface a
separate region of
the food and,
been
develop
processes that enhance its growth inhibiting properties providing
a much needed safety margin on the surface.
The first approach
microbial growth was
to enhanced
to maintain,
for as long
uneven preservative concentration:
in food bulk.
surface resistance
high
as possible,
on the surface and
to
an
low
We found that this was possible by the use of zein
reduced diffusion into
based coatings that
food bulk of
sorbic
acid applied on the surface.
surface treatment was
shown
in Staphytococcus auteu6 S-6 surface challenge experiments
under
The effectiveness of this
water
activity
(aw) of 0.88, stored at 30C under constant 88% relative
humidity
extreme testing
conditions.
for 16
(RH), remained stable
were stable for only 2
days.
Samples
or more
with bulk
days.
Uncoated
controls
Samples with bulk aw=0.85
exposed
at 30C to cycles of 12 hours at 85% RH and 12 at 88% RH, remained
stable for 28 or
more days.
Uncoated
controls were stable
for
only 3 days.
Sorbic
mechanism for the
acid
distribution
improvement was
studies
showed
diffusion control.
that
the
Apparent
diffusion coefficients for sorbic acid were 3-7 x 10~9 cm 2 /sec in
zein coatings and 1 x 10-6 cm 2 /sec
in the IMF model used in
this
study.
The second
approach
to microbial
stability
was
establishment of a pH difference between coated surface and
bulk.
the
food
with
Preliminary experimental work showed that samples
pH
surface
reduced
to
resistance
increased
associated with
the
lipophilic acids
growth
microbial
as
sorbic
reduction increases availability
the preservative, it results
soon
occurred.
The
acid.
pH
is
effectiveness
of
at
increased bacteriostatic
such
However, as
rapid growth
disappeared
pH difference
growth
microbial
controls.
significantly longer than untreated
as the
of
free
remained
reduced
surface
pH
effective form
of
Since
of the most
a
in significant microbial
stability
to
establish
improvements.
It has
been shown
that it
is possible
permanent and controlled pH differences between surface and
A negatively charged macromolecule can be immobilized in
values.
the
bulk
of
form
molecules,
a
food
surface
can
electrolytes,
particularly
while
component
coating
move
other
The
freely.
immobilization of the macromolecule can be described best using a
This model
Donnan equilibrium model.
differences between
that the key
surface and
allowed us to estimate
It also
food bulk.
parameters were electrolyte
pH
predicted
concentration and
the
number of charged groups of the macromolecule.
To prove the validity of
an IMF
model with
low
total electrolyte
0.005M) and coated it with
and agarose.
units.
the ApH concept we
concentration
a deionized mixture of
Measured pH differential was
Such a reduction resulted in
formulated
(about
X-carrageenan
in the 0.3 to 0.5
a 2.5 fold increase in
pH
the
surface availability
of
the
active
form
of
sorbic
acid
as
compared to food bulk.
The effectiveness of this
in S.auteus S-6 challenge
surface treatment was
tests at aw= 0.88,
We found a stability period of about 20 days.
to obtain
even
longer
periods.
In
our
shown
RH = 88% and
It seems
tests
the
30C.
possible
loss
of
stability was most probably due to growth occurring in the sample
bulk and not on the surface.
ACKNOWLEDGEMENTS
appreciation to
to express my
I would like
Professor
Marcus Karel, for his constant advice, criticism and support;
Dr.
questions;
and patience
support
generous
Wetherilt, for
important
Donald
Mr.
particularly,
Inc.,
Westreco,
to
research on the
to best focus
recommendations on how
invaluable
their
Sinskey, for
and Dr.
Rha
Dr.
Langer,
reading
in
to
and
analyzing preliminary reports.
Demain, Drs.
Holick,
Langer and Marletta for
honoring
I would also like to thank
Krieger and Rosenberg and Drs.
me with Teaching
Physiology
Human
respectively.
My deepest
students:
for
appreciation should also
enjoying
industrial microbiology
understanding that it is
Biochemistry,
Analytical
and
working
Industrial
courses in
Assistantships in their
Microbiology,
"my"
Dr.
harder
experiments earlier
be given
finishing
and
than expected;
difficult to organize
to
a new course
for
in
Human Physiology, with a new, more up-to-date orientation without
organizational difficulties;
analytical
biochemistry
carefully planned
class
for
understanding
equipment
may
experiments.
break
I
hope
that
even
and
down
that
they
new
ruin
all
learned from these experiences as much as I did.
My ever increasing gratitude to Professor Lechtman, for
her calm
but intense,
confident but
humble, hard
working
always available attitude whenever I needed her advice.
but
My grateful appreciation to Dr.
friendship but
also for
the many
Bouzas, mostly for his
hours working
together
late
nights and weekends.
I would
like to
express my
gratitude to
Sinskey and Holick and
their many students,
share so much of their
Drs.
Rha,
for allowing me
to
resources.
My gratitude belongs also
to
Barbara Sbuttoni, Peggy Foster and
Cheryl O'Brien for "tons"
of
contaminated glassware
microbial media and
disposal; to
Norman
Soule and Paul Steinberg for their help with Nomarsky microscopy;
to Ann Armitage for help and advice beyond duties; to Judy Quimby
for general administrative help and for letting me use the master
key, whenever my door decided to lock me out.
Many years have
passed since
I first
came to
M.I.T.
I was
lucky,
Like everything in life, there were ups and downs.
by my side I had not
only my advisor and an outstanding
but also a large number of
be always in my
of sadness and
that only my Brazilian friends have a word for -to think that I
will
Their names and faces
friends.
memory with a mixture
painful as it is
family,
happiness
"saudade".
might forget someone, I
As
would
like to thank:
Spiro Agathos, Jose
Cal-Vidal, Eduardo and Roxana
Jose and
Tina
Diskin, Dr.
Antoniano, Bobby Burke,
Jose
Cartaya, Osvaldo and Denisse
Chu,
Luis and Gail
Converti, John
and
Karen Damtoft,
Dr.
Martin
Jorge Funes, Jorge and Ivonne Godoy, Jorge and Mirta
Gurlekian, Arturo
and Carmen
Inda, Dr.
Edward S.
Josephson,
Margarita Jimenez, Louis and Susan Kacyn, Reza and Zari
Kamarei,
Carlos and Dolores Kienzle, Jorge and Gladys Lopez, Carmen
Mora,
Nicolas and Lichi
Masao
Majluf, Nana Mensa,
Yuichiro Miyasaka,
and Michiko Motoki, Chris Paola, Jorge and Cristina Retamal, Aida
Rios, Rene Rios, Amin and Alba Salim, Dr.
and
Marta
Savio,
Guillermo
Dr.
Aurora
Rodrigo and
Silverman, Toshiaki Tazawa,
Tolentino, Mariana Torres,
and
Julio Salinas, Ricardo
Schaeffeld,
Lucia Teixeira,
Michael Tracy, Uri
and Dafna
Judith
Nicole
Tsach,
Felipe Vera-Solis, John and Angelica Volker, Ujual and
Bal,
Enzo and Monica, Jose Luis and Leonor and Anita and Olga Yanez
Last but not at all the least I would like to thank my family and
my wife's
family
for
their
appreciation is beyond words.
confidence,
help
and
love.
My
This thesis is dedicated to my children,
Marcia and Toni
for love, so intense and so immense, I thought it could only be
found in impossible dreams; to my wife,
Anita
without her love, patience and hard work to support our little
family, I could have never realized my dream; to my parents
Amario and Aida
from whom I learned how strong and beautiful family love can be.
TABLE OF CONTENTS
Page
Title
Page
Abstract
........................................
..........................
Acknowledgements
Table of Contents
..................................
.................................
. .
List of Figures ...................................
. 15
List of Tables .................................... ............
18
1.
20
Introduction .................................
1.1
1.2
2.
..
. ..
. . 20
. .
. .23
..
. ..
.
..
. ..
. . 24
..
. ..
. . 24
..
. ..
. . 26
..
. ..
. . 26
Aims of the project .......................... ..
. ..
. . 28
2.1
2.2
3.
Approach to IMF product development .....
Effect of unsteady-state environmental
conditions on microbial stability .......
1.2.1 Examples of surface condensation
problems .... ......................
1.2.1.1 Temperature fluctuations during
production and distribution .....
1.2.1.2 Temperature fluctuations inside
a food warehouse ................
1.2.1.3 Lack of temperature uniformity
within a product ................
1.2.2 Research problem and solution
proposed ..........................
General aims ............................
Specific aims ...........................
.
. 23
. .
.29
Literature survey and theoretical background . ..
.
..
. . 31
..
.
..
.
..
..
---------- 31
. .. . . 32
. .. . . 34
..
..
.
.
..
..
. . 34
. . 38
..
..
..
..
.
.
.
.
..
..
..
..
.
.
.
.
3.1
3.2
3.3
Water activity ..........................
3.1.1 Thermodynamic basis for the
concept of aw .....................
3.1.2 Equations for the prediction of aw
Microbiological aspects of IMF's ........
3.2.1 Microbiology of IMF's and the
'hurdle' concept .. .................
3.2.2 Surface microbial load............
Preservative applications on the surface
of foods.................................
3.3.1 Application examples..............
3.3.1.1 Dairy products..................
3.3.1.2 Bakery products ................ .
. 31
.
.
.
.
41
41
41
43
3.3.1.3 Dried fruits ..................
3.3.1.4 Meat products ....................
3.3.2 Discussion
3.4
.........................
45
.....------
0
The reduced preservative diffusion
approach
47
.............................----
3.4.1.1 Mass transfer estimations ........
3.4.1.2 Experimental device for
potential coating tests
3.5
...............
3.4.2 Formulation of effective coatings
The reduced surface microenvironment pH
approach
*
..................
.----------
47
...........
47
..........
..........
0
3.5.2.1 IMF model requirements ...........
3.5.2.2 Selection of a polyelectrolyte ...
Experimental Procedures ....................... .....-----4.1
Preliminary experiments .....................
4.1.1 IMF model: composition,
preparation and analysis ..............
4.1.2 Preliminary microbial studies .........
4.1.2.1 Determination of inoculation
conditions .......................... .....--4.1.2.2 Growth response in the presence
of high K-sorbate concentrations at
various pH levels ...................
4.2
4.1.3 Stability of the uncoated IMF
model .................................
The reduced preservative diffusion
approach ....................................
4.2.1 Permeability experiments ..............
4.2.1.1 Preliminary tests
4.3
..............
..........
70
70
75
75
77
..------ 82
4.2.2 Determination of sorbic acid
stability .............................
4.2.3 Sorbic acid distribution studies ......
4.2.3.1 Microscopy studies ..................
4.2.3.2 Sorbic acid determinations ..........
4.2.4 Microbial challenge tests .............
The reduced surface microenvironment pH
approach ....................................
4.3.1 IMF model reformulations ..............
4.3.2 Coating procedures ....................
Experimental results ......................... ----...--5.1
70
77
4.3.3 Microbial stability tests .............
5.
.49
.50
...................
4.2.1.2 Permeability tests on selected
ingredients
0
.52
.53
.55
.66
.66
.67
.--
......
3.5.1 Theoretical considerations .........
3.5.1.1 The Donnan equilibrium model .....
3.5.2 Applications to an IMF model .......
4.
46
.............................
3.4.1 Theoretical considerations on mass
transfer
~
Preliminary microbial challenge studies
5.1.1 Bacterial challenge studies .......
.85
.85
.87
.87
.87
.97
.97
097
100
101
5.1.2 Mold challenge studies ........
5.2
...............
101
5.1.3 Conclusions ...................
The reduced preservative diffusion
107
approach ............................
.107
.107
.108
.115
.115
.115
.119
.119
.120
5.2.1 Permeability experiments ......
5.2.2 Sorbic acid stability studies .
5.2.3 Sorbic acid distribution studie
5.2.3.1 Microscopy studies
..........
5.2.3.2 Sorbic acid determinations ..
5.2.4 Microbial challenge tests .....
5.2.4.1 Background microbial load ...
5.2.4.2 Location of microbial growth
5.2.4.3 Determination of coating
effectiveness
5.3
...............
5.2.5 Conclusions ...................
The reduced surface microenvironment pH
. ..
..
... 122
.........
125
.130
.130
.130
.136
.136
approach ............................
5.3.1 Dialysis experiments ..........
5.3.2 Microbial challenge tests
5.3.3 Sorbic acid stability .........
5.3.4 Conclusions
6.
...................
Discussion .............................................. 138
6.1
6.2
6.3
6.4
138
139
General ................................
Sorbic acid stability ..................
The reduced preservative diffusion
approach ...............................
The reduced pH surface microenvironment
approach ...............................
.
7.
Summary and conclusions ..................... ............
8.
Suggestions for future work ................. ..
8.1
143
. ..
. . 147
. ..
. . 147
. ..
. . 148
The reduced preservative diffusion
..
microenvironment
The reduced pH surface
..
approach ...............................
approach
8.2
142
...............................
References .. ................................................. 150
Appendix A.
An example of surface condensation problems ......
............
158
Appendix B.
Measured and predicted equilibrium moisture
content comparison ............................... ............
163
Appendix C.
Mass transfer estimations ........................ ............
166
Appendix D.
1. Use of permeability values to estimate coating
effectiveness ...............................
2. Estimation of zein coating effectiveness ....
............
Appendix E.
Preparation and composition of coatings tested
.168
0 0 .170
...........
172
Appendix F.
Sorbic acid determinations as percentage of
initial concentrations .............
............
Appendix G.
Sorbic acid surface application
.176
..........
0
.178
...........
Appendix H.
Determination of apparent sorbic acid diffusion
coefficients using experimentally measured sorbic
acid distributions ..............................
0
.............
180
LIST OF FIGURES
Figure No
1.
Development of the intermediate moisture
cheese analog used as a model system in
this
2.
Page
Title
thesis
.........................................
22
Temperature fluctuations in a food
warehouse
...........................................
25
3.
A measure of the significance of good
sanitation .......................................... 37
4.
Schematic representation of microbial
stability enhancement by high
concentration of K-sorbate deposited on
a highly impermeable edible coating
.................
48
5.
Experimental approach to the selection
of coating compositions .............................. 51
6.
Preliminary microbial tests and pH
determination. IMF model No.1 coated
with pH 4 zein solution .............................. 54
7.
Schematic representation of a Donnan
theory model for the establishment of a
pH difference between surface and food
bulk ................................................
56
8.
Schematic representation of a
semipermeable membrane. Analysis for an
impermeable charged polyelectrolyte and
univalent permeable ions ............................. 57
9.
Schematic representation of a
semipermeable membrane, effect on pH ................ 62
10.
Equilibrium deviations for the
equilibrium analysis of semipermeable
membrane ............................................ 63
11.
Molecular structure of agarose,
furcellaran and carrageenans ........................ 69
12.
Preparation of IMF model system ..................... 72
13.
Sorbic acid analysis by HPLC ............ ............
74
14.
Surface microbial challenge with
S.auteus at high aw and neutral pH .........
15.
Effect on S.auteus viability of high
K-sorbate concentrations
16.
...................
......
78
Inoculation procedure for uncoated
..............
.............................
samples
.81
17.
Permeability cell used in this thesis
18.
Experimental procedures to determine
K-values for selected ingredients ...
.........
84
19.
Experimental procedures for the
determination of sorbic acid stability .....
.........
86
20.
Experimental procedures for the
determination of sorbic acid
distribution
88
..................
.....
21.
Sample holder for zein coating, sorbic
acid spraying, solvent removal and
inoculation
22.
Sample preparation: cutting, zein
coating and sorbic acid surface
application
23.
surface
pH
............
10 2
.............
Effect of X-carrageenan on S.auteu
growth
26.
95
..............................................
Experimental procedures for microbial
stability studies on the effect of
reduced
25.
inoculation
Sample preparation:
procedure
24.
92
................................
.....................................
Effect of deionized X-carrageenan on
S.auAeu s
growth
. . . . . . . . . . . . . .
0
.
0
0 . *
.............
....
105
27.
Microbial stability of the uncoated IMF
(model No.1) ............................... -------- 106
28.
Variation of K-values as a function of
initial K-sorbate concentration
difference
.................................
..------ 110
29.
Sorbic acid determinations as percentage
of initial concentration ................... ........
112
30.
Microbial growth response at high
K-sorbate concentrations ..................
113
.......-
31.
Nomarsky microscopy studies ........................ 116
32.
Sorbic acid distribution studies ................... 118
33.
zein dipped coated
Stability test:
samples, bulk aw = 0.88, challenged with
S.aueu&s
34.
37.
(constant RH)
.*.
*..
(RH cycles)
.......
..
Stability test: zein sprayed coated
samples, bulk aw = 0.85, challenged with
S.auteus (RH cycles) ...................
...........
* 00 0.127
Measurement of the pH differential
2 and
established on IMF model No.
estimation of its effect on the surface
availability of undissociated sorbic
o...........132
.....................
Microbial challenge studies:
effect of
Measurement of the pH differential
established on IMF model No. 2 and
estimation of its effect on the surface
availability of undissociated sorbic
acid, extended test
41.
0.126
0.0.0.0.0....128
reduced surface pH ..................... ............
40.
.
Schematic representation of the highly
dynamic and heterogeneous conditions on
acid
39.
.
....
the surface of a coated food ...........
38.
124
.....
*.**...
.
zein dipped coated
Stability test:
= 0.85, challenged with
aw
bulk
samples,
S.auieuz
36.
123
......................
zein sprayed coated
Stability test:
samples, bulk aw = 0.88, challenged with
S.au4eus
35.
(constant RH)
....................
...........
134
0 0 .135
Sorbic acid stabilitv in the presence of
X carrageenan ...................................... 137
LIST OF TABLES
Table No.
1.
Effect on vapor pressure of a within
product temperature differential caused
by the heat generated by an illumination
source.
2.
3.
Page
Title
..............
.
Effectiveness ratio between the
dissociated (Ki) and undissociated (K2 )
form of sorbic acid ..................... ..
.
Effect of pH on the % undissociated form
of lipophilic acids used as
preservatives ...........................
..
..
..
.
. 35
...
.36
4.
Hurdles used in IMF .................................. 39
5.
Recommended methods for surface
treatment of cheese and cheese products ............. 42
6.
English muffins, days of mold free
shelf-life of coated and uncoated
samples ............................................. 44
7.
Apparent diffusion coefficients required
to achieve significant microbial
stability improvements .............................. 49
8.
D values for diffusion thru a solid
matrix
9.
....................
.
.
.......................
Values of mM+/1 and mX-/1 for two values
of mM+/2 and a range of values for
--.--.-.-.-..
zmp-z ------.-.-.--- --- ---.-.-.-.--
50
. .60
10.
pH values as affected by the
concentration of charged groups and the
concentration of electrolytes present in
a food product ...................................... 67
11.
Composition of the IMF model system ................. 71
12.
Comparison of K-values ...
13.
Composition of IMF model, coating
solutions and sorbic acid solutions ................ 90
14.
IMF surface treatments ....................
.......................... 83
......... 96
electrolyte
IMF
97
15.
Low
16.
Protein fraction dialysis requirements .............. 98
17.
X-carrageenan coating composition....-................99
18.
Media for preliminary tests on the
possible effect of X-carrageenan on
microbial growth
19.
20.
21.
22.
****e....103
..........................
Number of days for visual detection of
mold growth .. . . . . . . . . . . . . . .
Summary of K-values for various coating
materials as determined by a diffusion
.........
cell ........................
.
. . .1
...-------
7
109
Sorbic acid stability study, initial
concentrations ..................................... 111
Effect of pasteurization on sorbic acid
-storage stability
23.
...............................
..............
....................
Sorbic acid deposited on each individual
sample piece
........................................
117
24.
Microbial background load tests .................... 120
25.
Location of growth tests ........................... 121
26.
Electrolyte elimination by
27.
Microbial stability improvements as a
function of various surface treatments
dialysis
................ 131
...........
141
-
1.
20
-
Introduction.
Foods
with
improved
microbial
stability
controlled
presence of
by
solutes.
adjusting moisture
These products,
so
be
This parameter
fabricated by reducing their water activity (aw).
can be
can
and
content
by
the
intermediate
called
moisture foods (IMF's), are microbiologically stable as long as:
(1)
aw < aw*
where:
aw = aw(t,
x,
y. z)
= aw at any time (t) and any location (x,y,z)
within
the IMF
aw* = critical aw value
Above aw*, microbial growth is possible.
depend upon
parameters such
Its value will
concentration and
as pH,
type
of
preservative(s), expected microbial contamination, etc.
This stability condition may be temporarily lost due to
changing environmental conditions.
The purpose of this thesis is
to analyze the resulting microbial stability problems.
1.1
Approach to IMF product development
The process of developing an
IMF can be simplified
if
shelf life and organoleptic quality are represented as a function
-
21
-
of the various fabrication parameters, such as water content, use
of solutes, pH, concentration
conditions, and/or many of the
processing
of preservatives, heat
other hurdles used to obtain
the
desired microbial stability.
Figure 1 shows a
was used (Motoki,
specific example where this
It corresponds
1979).
approach
to the development
of
the intermediate moisture cheese analog used as a model system in
this thesis.
product formulations
of acceptable
A region
was
defi.ned by the identification of the following key factors:
a. an acceptable texture, a region identified through
mechanical
measurements using an Instron;
b. an acceptable
taste, a
region identified
thru
organoleptic
tests; and
c. a minimum aw,
a region identified
thru measurements with
an
electric hygrometer.
Figure
difficulties
1
illustrates
surrounding
IMF
the
product
technology.
The
development
region
of
acceptable combinations, represented by the shaded area, is quite
limited.
Therefore, it is not possible to include considerations
on possible and/or
out-of-control situations
their microbial stability.
that could
impair
-
22
-
Figure 1
DEVELOPMENT OF THE INTERMEDIATE MOISTURE CHEESE ANALOG
USED AS A MODEL SYSTEM IN THIS THESIS
PROTEIN-OIL
WA TER
HUMECTA NTS
-
1.2
-
23
Effect of unsteady-state environmental conditions on
microbial stability
Microbial
unsteady-state environmental
been mostly ignored
have tested
by
possibility
has
working in
this field,
who
constant temperature
and
1982;
1982; Erickson,
and Mudahar,
Hanseman et
Prior, 1980;
affected
also
This
conditions.
conditions of
them under
is
IMF's
by researchers
humidity conditions (Bhatia
Theron and
of
stability
al., 1980;
Flora et
al.,
1979; Pavey, 1972; Anonymous, 1972).
Most
environmental relative humidity
packaged
generally
temperature
conditions.
moisture
in
changing
in
external
These products
are
However,
materials.
proof
result
can
fluctuations
by
affected
not
IMF's- are
surface
extensive
condensation problems.
1.2.1
Examples of surface condensation problems
Examples
abound in
of
the production
following ones are given
surface
unsteady
condensations
stability balance,
state
temperature
and commercialization
only to emphasize
will
disrupt
expressed as
aw <
the
aw*.
situations
of IMF's.
the fact that
delicate
As
surface microbial growth is then often observed.
a
The
these
microbial
consequence,
-
24
-
Temperature fluctuations during production and
1.2.1.1
distribution
Let us consider an IMF (aw
and stored
in a
warehouse
10OF
0.80), packed at T(1) =
=
at T(2)
= 70F.
this
Furthermore,
product could be loaded into a
truck at T(3) = 10OF (hot
day) or a railroad car at T(3')
=
summer
35F (winter day) to be unloaded
into another warehouse at T(4) = 70F.
This type of situation was analyzed in Appendix A.
only
showed that while headspace humidity condensation will play
as 10 ml
result in as much
drops
Water
package.
condensing surfaces
spots
localized
with
condensation for a
package
on
the product
high
aw
provide
that
estimated
We
levels.
cold
as
will
itself,
can
weight
1 lb net
inside, acting
on the
and
pieces
warm food
evaporation from
role, moisture
a minor
We
condensations on food surfaces would allow microbial growth, i.e.
aw > 0.85, for several hours.
1.2.1.2
Temperature fluctuations inside a food warehouse
Kuklov
Grundke and
inside
a
commercial
(1980)
Temperature
2).
(Figure
warehouse
temperature
the
measured
fluctuations were especially severe during the winter season --to
save energy a
heating system is
situation could be expected in
conditioning or
heating
constant repetition
of
system
turned on and
off.
similar
A
a consumer kitchen, where an
is
turned
these temperature
off
cycles
and
on.
could
air
The
create
-
25
-
Figure 2
TEMPERATURE FLUCTUATIONS IN A FOOD WAREHOUSE
tj0 C
t,
25
20
p
t,
a.
Week in summer
b.
Week in winter
C
-
severe
-
problems
stability
microbial
26
to
due
repetitive
the
creation of surface conditions with high aw-
1.2.1.3
Lack of temperature uniformity within a product
A typical
is
a
The
shelf (Reid, 1976).
on a display
refrigerated IMF sitting
situation
of
this kind
of
example
product is heated up by a light source while it is cooled down by
a cold source.
it is under
product with aw = 0.90 and
Let us assume a
the influence
of a 2C
that
As
temperature difference.
shown in Table 1 the vapor pressure on the warm side will be even
The
higher than the saturation vapor pressure on the cool side.
vapor
pressure
transferred from
differential
the warm
will
result
cool side
to the
in
being
water
equilibrium
until
conditions are established.
1.2.2
Research problem and solution proposed
The above examples have shown that localized surface aw
conditions play
stability of IMF products.
treatments
that
role
important
an
can
in
the
overall
microbial
Therefore, there is a need to develop
IMF
protect
surfaces
against
unstable
temperature environments.
The solution
consider surface a
explored
in
this
separate food region
that enhance specifically its
thesis
has
and, develop
been
to
processes
growth inhibiting capacity.
This
-
27
-
Table 1
EFFECT ON VAPOR PRESSURE OF A WITHIN PRODUCT TEMPERATURE
DIFFERENTIAL CAUSED BY THE HEAT ORIGINATED FROM AN ILLUMINATION
SOURCE
/
Light source
Moisture transfer
Cold
source
Cold side (T)
mm Hg
T, C
Pv0
Warm side (T + 2C)
mm Hg
Pv
Pv0
Pv
0
4.6
4.1
5.3
4.8
5
6.5
5.9
7.5
6.8
margin on
the surface,
would provide
a much
needed safety
option not available when considering the product as a bulk.
an
-
2.
28
-
Aims of the project
The
solution
problem resulting
surface layer
growth.
from
with
Thus,
an
if a
proposed
to
the
unstable
temperature
microbial
improved capacity
growth allowing
stability
conditions
to
resist
is
a
microbial
condition appears
on
the
surface microenvironment, the surface properties will inhibit, or
at
least
reduce,
growth
equilibration will
occur
rate
to
such
before high
an
cell
extent,
that
aw
concentrations
are
reached.
2.1
General aims
Increase surface microbial growth resistance by the use
coatings
of edible
with
specific
this
Maintain an unequal distribution of preservative,
i.e.
approaches
been
followed
to
different
satisfy
experimental
have
Two
properties.
enhanced stability goal:
a.
Approach 1
start with a higher (initial) concentration of preservative(s) on
the surface
and
difference for
use a
as
long
"reduced preservative
coating
as
the
to maintain
possible.
This
diffusion", required
concentration
approach,
the selection
called
of
a
coating capable of reducing preservative diffusion rate from food
surface into food bulk.
-
b.
-
29
Approach 2
than that of
food bulk.
sorbic acid,
are
called
required the
the
"reduced
identification
pH
itself
is
This
Leistner and Rodel, 1976).
strongly affected by pH (e.g.
approach
Moreover, growth
1973).
et al.,
is
effectiveness
acids whose
lipophilic
dependent (Freese
particularly
food preservatives,
Most
lower
is
where pH
surface microenvironment
Create a
surface
of
microenvironment",
pH
the conditions
which
under
a
reduced
charged macromolecule could be used to maintain a stable
surface pH.
2.2
1.
Specific aims
Analysis
unsteady
of
surface
where
situations
state
condensations are significant and cause unexpected
microbial
stability problems.
2.
where
the
measured.
Particularly,
microbial outgrowth
effectiveness
modification
surface
as
collect background
affected
model
testing of an IMF
Development and microbial stability
by pH,
be
will
information
aw
and
on
K-sorbate
concentration.
3.
On the reduced preservative diffusion approach:
experimental approaches for
appropriate
coating
mathematical expressions
the rapid
composition;
and
(1)
develop
identification of
and,
analytical
develop
(2)
methods
an
to
study
-
K-sorbate distribution
-
30
surface
between
and
food
bulk
of
the
pH
treated and untreated samples.
4.
On the reduced surface microenvironment pH approach:
(1) develop
a model
and
expression for
derive an
food bulk as a
difference between food surface and
of experimental
guideline
for
conditions;
the
(2)
selection
of
use
this expression
coating
material(s)
application conditions; and, (3) execute microbial
tests and determine
function
stability improvements
as
a
and
stability
to evaluate
effectiveness of the surface modification procedures.
the
-
3.
31
-
Literature survey and theoretical background
Modern IMF's are an
products.
Their
success
example of highly engineered
or
failure
depends
on
food
a
clear
understanding on their limitations.
3.1
Water activity
3.1.1 The concept of water activity
The concept of
the formulation of
IMF.
water activity occupies
Scott
a key role
(1957) showed that
in
there was
a
value aw
is
thermodynamically defined as the ratio at a given temperature
of
correlation between aw
and microbial
growth.
The
the fugacity, f, of water in some given state, and its fugacity f
in a state which
for convenience, has
state (Reid, 1976).
In
the case of
been chosen as
reference
aqueous solutions the
convenient reference state is pure water.
Fortunately,
most
fugacity
can be substituted by vapor pressure.
Since aw is defined at a fixed temperature it is a function
of temperature.
Fortunately, in
function of it (van
some cases this
most cases it
den Berg and Bruin,
may not
be true as
authors (Mazza and LeMaguer,
is not a
1978; Reid, 1976).
has been
many
1978;
et al., 1976; Rasekh
al., 1971; Saravacos and Stinchfield, 1966; Heldman et al.,
and Fenton, 1941).
In
reported by
1978; Troller and Christian,
Iglesias and Chirife, 1976a,b; Iglesias
strong
et
1965
However, for practical purposes we can assume
-
32
-
assumption
that aw is independent of temperature and will be the
used in this project.
3.1.2
Equations for the prediction of aw
number of equations
A very large
predict the aw of foods (Chirife,
1980; van den Berg and
able to describe the
None has been
1978).
has been derived
accommodate hysteresis phenomena nor
Bruin,
whole range, nor
the effect of
to
temperature.
ways in which water interacts with
This is due to the many
to
food
components.
The theoretical estimation
of Gibbs; theories
free energy model
theory based on the
physical sorption if
a surface can
structure formed
the
by
solutions
have included
of aw
about
the
be distinguished (i.e.
and
proteins
polymers:
common food
carbohydrates); or a combination of both if the particular system
requires so.
In applying these theories, it has been found
physical sorption
B.E.T
classical
theories work
better at
and
solution
equation)
low aw,
that
the
(e.g.
more
are
theories
succesful at higher aw (Chirife and Ferro Fontan, 1980; Benmergui
1979; Ferro
et al.,
1979).
Fontan
et al.,
An excellent review on
1979a,b; Chirife
al.,
et
the theoretical prediction of
aw
was presented by van den Berg and Bruin (1978).
The subject of empirical and semi-emipirical
has
been
Argentina.
extensively
covered
His excellent
by
the
group
review on fitting
of
equations
Chirife
equations to
in
water
-
-
33
be considered a
sorption isotherms should
primary reference
on
this field and was used extensively in this project (Iglesias and
Chirife, 1982, Chirife and Iglesias, 1978; Boquet et al., 1978).
Other published work of this
group include the determination
the Halsey
equation (Iglesias
al., 1976a);
comparison of
and
(Iglesias and Chirife, 1976c)
the parameters for the B.E.T.
and Chirife,
the Halsey
1976b; Iglesias
and Henderson
and
(Iglesias
isotherms
Chirife,
et
equations
(Iglesias and Chirife, 1976c); the application of the concept
local
of
and
1976d);
of
the
application of the Smith equation for the high aw region (Chirife
They also
et al., 1979).
mixtures of dehydrated food
equation (Iglesias et
the B.E.T.
precision of Ross equation
range
1978).
(Chirife,
calculate the
aw
of a
low aw range
products in the
given
equation
has
the
mixtures in the
IM
derived
to
been
with known
system
using
1979); and evaluated
al.,
to predict aw of
This
for
of aw
studied the determination
composition.
Therefore the formulation of an IMF is a trial-and-error process.
Recently an equation has been proposed to estimate
final moisture content of complex
moisture content for
and also the
aw value
the desired
(Lang and
trial-and-error
the
systems, given the desired
aw
the individual components
at
Steinberg, 1980).
process
associated
This
would
with
Ross'
eliminate
the
equation.
Therefore, it was decided to test the accuracy claimed
by the authors, using
model system
calculations.
used
We
ten products from
in this
study.
were unable
the literature and
the
summarizes
our
Appendix B
to predict
equilibrium
moisture
-
34 -
the
when
was found
(5% error)
good agreement
recognized that
be
It should
content with the accuracy reported by the authors.
equation was applied to our model system (product 11).
3.2
Microbiological aspects of IMF's
3.2.1 Microbiology of IMF's and the 'hurdle' concept
IMF's range in aw from 0.7 to 0.9 and in water content from
appropriate preventive
the scope
fall under
enzymatic
of this
growth,
browning
unless
Since
taken.
measures are
mold
to
be susceptible
non
or
degradation,
enzymatic
may
These foods
50%.
20 to
project, changes
not
they do
to
not related
microbial growth will be ignored.
with
The value aw is only one of the environmental factors
which the IMF technologists can control growth of microorganisms.
Their growth
factors
and
such
as
pH,
temperature,
environment, and the presence
preservatives.
Moreover,
it
Food
gas
many
by
composition
other
the
of
of inhibitory substances, such
has
been shown
when
that
as
other
the inhibitory effects of aw
factors are less than optimal,
enhanced (e.g.
influenced
will be
survival
are
Troller and Christian, 1978).
preservatives
are
used
mainly
fungistatic
as
bacteriostatic
activity.
Those most commonly in use today are lipophilic acids.
Generally
agents, although
they
they inhibit growth of
shown in Table
do
have
some
microorganisms without killing them.
2 their activity
results from the
As
undissociated
-
form.
Therefore, they are not
Freese et
al.,
1973).
Based
percentage has been calculated
Table 3.
However,
undesirable.
This
35
-
active at high pH (Eklund,
on
their pKa
the
as a function of
acidification
incompatibility
stability and organoleptic quality
is
usually
between
undissociated
pH as shown
is typical of IMF
microbial
technology
limitations.
Table 2
EFFECTIVENESS RATIO BETWEEN THE DISSOCIATED (K2 ) AND
UNDISSOCIATED (K1 ) FORM OF SORBIC ACID [1]
Organism
70 (0.96
Bacillus cereus
15 (0.44)
Escherichia coli
100 (0.96)
Pseudomonas aeruginosa
15 (0.97)
Staphylococcus aureus
600 (0.89)
Candida albicans
in
organoleptically
enhanced
Bacillus subtilis
1983;
[2)
10 (0.84)
[1]
Eklund, 1983.
[2]
Fraction of total variance explained by analyzing data
model.
36
-
-
Table 3
EFFECT OF pH ON THE % UNDISSOCIATED FORM OF
LIPOPHILIC ACIDS USED AS PRESERVATIVES
Compound
pKa [1]
3
4
5
6
7
benzoate
4.2
94
61
14
2
.2
sorbate
4.8
98
86
39
6
.6
propionate
4.9
99
88
43
7
.8
parabens
8.5
99
99
99
99
[1]
Freese et al., 1973; Sauer, 1977.
A measure
of the
illustrated by Figure 3 which
ce'evitaie,
30C.
97.0
significance of
good sanitation
shows the growth of
E chenichia coZi and BacilZla
is
Sacchatomyce4
coagatans at pH 5.5 and
The lower the initial contamination, the longer the sorbate
protection (Anonymous 1978a).
In the
case of IMF's it would
preferable to decrease microbial load before aw is reduced.
resistance of microorganisms increases
as aw is lowered
be
Heat
(Corry,
1974, 1975, 1976).
Table 4
technologist.
We
summarizes the
will
add
hurdles available
another factor
to
this
to the
list:
IMF
a
-
37
Figure 3
A MEASURE OF THE SIGNIFICANCE OF GOOD SANITATION
1 09.
O.8.
0.70.60.5-
01% POTASSIUM
SORBATE
0.40
C 0.3-
0
0.2.
10'
102
0.1
Time (Hours)
Time (Hours)
a.
1.
Sacchanomyces cekeviiae, inoculum levels as indicated
SORBATE
0 1% POTASSIUM
1 .
0.90.8
0.7
0.6
0.5
09.
0 8.
0.7.
0 6.
05,
0.4
4Ci00.3-
0 0.3
0.2
0.1
Time (Hours)
b.
Time (Hours)
Eschekichia coLi, inoculum levels as indicated
100 -
-
38
-
SORBATE
0.1%POTASSIUM
0.3
0.3
0.2
0.2-
10'
10
102
Time (Hours)
Time (Hours)
c.
-
Bacitlts coagutans,
inoculum levels as indicated
coating with enhanced resistance
to microbial growth.
the
From
above presented facts it can be concluded that the application of
a coating with a higher concentration of preservatives, and/or
pH lower
than the
optimal for
closer to the range where
bacterial growth
and
the preservatives are most
a
therefore
effective,
would constitute a major contribution to the stability of IMF's.
Evidently, such a coating would only enhance the stability of the
surface, but we have already shown that there is a need to
overprotect the surface.
3.2.2
Surface microbial load
In a previous section the idea of IMF ingredients
pretreatment was suggested
to reduce
microbial load.
subsequent handling can cause recontamination.
products are cooked or
heat
However,
Even though
pasteurized, slicing and packaging
some
gives
-
39 -
Table 4
HURDLES USED IN IMF [1,2]
Hurdle
Bacteria
Yeasts
Molds
+
-
+
-
+
-
+
-
aw
+
pH
+
redox potential (Eh)
+ -
+ -
+ -
storage temp. [3]
+ -
-
-
thermal process
+
+
+
preservatives [4]
+ -
+ -
+ -
competing microflora
+ -
+ -
+ -
-
[1]
Adapted from Leistner and Rodel
[2]
+
= Inhibition
+ -
= Some inhibition
-=
[3]
[4)
(1976)
No inhibition
See also Farrel and Upton (1978), Theron and Prior (1980)
(1979)
See also Raccach et al.
ample opportunity for recontamination and shelf life usually ends
as a result of microbial growth (Stiles and Ng., 1979).
Another problem of surface contamination is that viable
counts can
1979).
be
highly variable
(Anderson
et al.,
This is particularly important in IMF's.
1980;
Gill,
We have already
-
shown that bacteriostatic
-
40
barriers can
be overcome
by a
large
number of cells.
We should also consider
auteuA,
whose
minimal
concentration.
while
under
aw
the problem of
requirement
Staphytococcu4
depends
on
oxygen
Under anaerobic conditions its minimal aw is 0.91
aerobic
Therefore, the
conditions
surface
it
(Scott,
readily available is the region we should be more concerned
with
this
ubiquitous
oxygen could
1953).
more
of
where
0.86
be
potential outgrowth
of foods,
is
organism
(Pawsey
and
Davies, 1976; Lubieniecki-von Schelhorn, 1975).
Therefore, we
see
in
surface
contamination
another
reason to overprotect IMF surfaces.
Microbial challenge
studies
on
the
surface
require
In the case
of
meats it has been found that about 106 cells/cm 2 is the level
at
knowledge of cell concentrations at the surface.
odors can
which spoilage
Newton and Rigg, 1979).
on adipose fresh
be detected
(Gill and
Newton,
1980;
Gill and Newton (1980) found that growth
meat ceased
at 108
cells/cm 2 .
Other
authors
reported that no definite changes occur in the meat until surface
bacterial count exceeded 108 cells/cm 2 (Ingram and Dainty, 1971).
Results by
Sauter et
carcasses depended
the
supports
growth
on
strongly on
generally
food
(1979) showed
al.
the fat
accepted
surfaces
fermentable substrates from
is
that growth
layer thickness,
that
observation,
limited
on
by
within the food.
the
lamb
which
bacterial
diffusion
Therefore, it
of
is
-
41
-
potential
advisable to specifically determine the surface growth
of our particular model system.
3.3
Preservative applications on the surface of foods
basis of
the
selected on
or
dipping, spraying
or
dusting,
will
Examples
surface
the
on
into
incorporation
by
be
only
given
food
on
of
direct
in relatively close contact
wrapping material when this is
the product.
type
by
evenly
distributed
concentrated
mixture,
the
into
addition
methods
of
variety
convenience and
processing
either
be
can
They
product.
a
by
applied
Preservatives are
by
the
with
surface
applications.
3.3.1
Application examples
3.3.1.1 Dairy products
Solid cheeses are sprayed or dipped.
they are
to
allowed
sealing when wrapped.
solutions
ranging
(Anonymous,
1978a).
recommended
by
As
from
Pfizer
Standards of Identity
Anonymous, 1976b).
avoid
to
dry thoroughly
shown in Table
20
More
to
40%
is
specific
Chemicals
in
regulations can be
treatment,
After
imperfect
heat
5, K-sorbate in
water
commonly
used
most
details
procedures
on
accordance
to
found elsewhere
Federal
(e.g.
-
42
-
Table 5
RECOMMENDED METHODS FOR SURFACE TREATMENT
OF CHEESE AND CHEESE PRODUCTS
Product
Cheddar Colby
Monterrey Jack
Blue, etc.
Muenster
Edam, Gouda
Solution
concentration
%w/v
20-40
20
Solids
deposited [2]
Directions
%w/w
.10-.30
Dry sprayed or
dipped before
wrapping to
avoid poor seals.
.05-.10
Because of bland
flavor, .1%
deposition should
not be exceeded.
Provolone
Pasta filata
Caciocavallo
Siciliano
20-40
.10-.30
Preservative
treatment should
follow 24-hour
hold period after
brine dip. 4-7 days
before wax dipping
or film packaging.
Mozzarella
20-40
.10-.20
Processing varies.
No specific treatment recommended.
Swiss, Gruyere,
Emmentaler
20-40
.10-.30
Not dipped because
"eye-flooding".
Important to spray
eyes.
[1]
[2]
Anonymous, 1978a.
Concentration of dip, exposure time and cheese surface
volume ratios pieces will determine amount picked up.
Solution should be changed regularly to prevent
contamination by resistant microorganisms.
-
43
-
Propionic acid and sodium
and calcium propionate
are also effective in preventing mold growth on cheese
but high
concentrations lead
to bitterness
too,
surfaces,
(Kosikowski,
1977,
p.552) which is not a problem with sorbates (Anonymous, 1978b).
3.3.1.2
Bakery products
Bakery products are
oven,
except
temperatures.
for
bacteria
some
they come
out of
survive
which
present in the
Mold spores
and
air then contact
the
sometimes
storage (Anonymous,
consumer
the
high
the
during cooling, packaging, and
surface of the product
during distribution
sterile as
1977a).
Therefore addition of preservatives is desirable.
Calcium propionate and
even less active
preservatives
have been used in an attempt to minimize the inhibiting effect of
on
preservatives
leavening action.
yeast
and,
fermentation
therefore
on
Sorbic acid and potassium sorbate, used widely
in chemically leavened snack cakes, are considered too active
be added to yeast raised doughs.
by Monsanto
been proposed
baked or griddled
in the
usual way, but
depanning, the
surfaces with
a
sorbate.
the hot
Tests
shown in Table
sorbate
A solution to this problem
(Anonymous, 1977b).
oven or after
seconds from
the
The
surface, leaving
on English
muffins
6 (Anonymous, 1978b).
has
is
immediately after
the
evenly on
all
product is covered
solution.
The product
to
water
evaporates
in
shield
of
encouraging
as
a protective
were quite
Similar tests on
variety
-
44
-
Table 6
ENGLISH MUFFINS, DAYS OF MOLD FREE SHELF-LIFE OF
COATED AND UNCOATED SAMPLES [1]
days in commercial bakery
Muffins with:
no preservative
5
0.5% Ca propionate in dough
7
1.0% Ca propionate in dough
9
0.1% Sorbate in dough
and 0.1% as a surface spray
12
and 0.2% as a surface spray
26
[1]
Ref.: Anonymous, 1978b
bread, hamburger
buns, rolls
and tortillas
positive
were also
(Anonymous, 1977a).
3.3.1.3
Dried fruits
Dried fruits are often moisturized to attain
texture for consumer appeal.
desirable
Potassium sorbate has been used
in
these high moisture dried fruits to protect them against mold and
yeast spoilage (Nury et al.,
1960).
In another study the
sorbic
-
acid gradient was
45
also measured
-
Thirty
al., 1980).
(Bolin et
four percent moisture prunes (skin:pulp = 37:73) dipped in a 1.7%
potassium sorbate solution showed that after one day of
the skin contained 708
the end of 30
initial
final equilibration but rapid
the
204.
ppm and pulp only
weeks, skin contained 312
explained by
At
fruit pulp 207.
ppm sorbic acid and
This discrepancy, lack of
diffusion was
dipping,
authors as
due to
a
probable
addition compound with either the waxy coating materials or other
components.
3.3.1.4
Meat products
has
The relative short life of fresh, unfrozen poultry
Fresh broilers in retail
been an industry problem for some time.
outlets normally have
level of 104
an initial contamination
10 5 cells/cm 2 . Normally they can only be stored for 1 or 2
at 3 - 5C
and still maintain their
Robach, 1980).
1978; To and
shelf life
of
fresh, whole
freshly chilled carcasses
sorbate for 30
seconds.
In a
broilers
into a
The
the
Robach (1979)
was extended
5% w/v
days
Ivey,
freshness (Robach and
study by
control birds were
dipping
by
solution of
to
potassium
stored for
10
were
days when spoilage was evident, while sorbate-treated birds
stored for 19 days, at which time spoilage was evident.
The sensory properties
dipping in 5 or
of cooked
10% sorbate solutions
sensory panels (Cunningham, 1979).
chicken parts
after
were found acceptable
by
-
In a study
46
-
of turkey products,
Robach et al.
(1980)
found that dipping and spraying were more effective than
pumping
it into the product, even when the total amount sprayed or dipped
was lower than
increasing storage time microbial
probably
lost, most
It was also
the total pumped.
due
to
with
noted that
protection at the surface
was
into
the
migration
preservative
product.
3.3.2
Discussion
Uneven distribution of preservatives has been proposed,
proved to be effective, shown
and in
some cases,
to pose no organoleptic
has been
its use
agencies.
Therefore FDA acceptance
diffusion
proposed
to
preservative
can
stability
microbial
enhance
regulatory
authorized by
of the reduced
problems,
be
predicted.
Uneven distribution
very succesful for
poultry) or
where
production process
products with
very short
(e.g.
proved
life (e.g.
fresh
interfere
preservative could
the
has been
of preservatives
and
cheese ripening
yeast
with
the
leavened
bakery products).
also that the
The examples described showed
stability improvement is
the preservative
into
generally limited by
the product.
technique to other products it
In
microbial
the diffusion
order to
expand
would be necessary to reduce
rate of diffusion as proposed in this study.
of
this
the
-
3.4
47
-
The reduced preservative diffusion approach
We have proposed to enhance the microbial stability
surface.
The
overall concentration
can still
be within
limits by lowering the concentration in the bulk, e.g.
K-sorbate to 0.1%.
its
high concentration of preservative on
IMF's by the use of a
The
amount corresponding to this
of
legal
from 0.3%
difference
will be the amount used on the surface as described in Figure 4.
3.4.1
Theoretical considerations on mass transfer
To facilitate the selection of coating procedure(s) and
ingredient(s) we need:
the mass
(1) a mathematical expression to estimate
transfer properties
stability; and, (2)
required
to achieve
an experimental device
the
where the
desired
potential
coatings could be easily tested.
3.4.1.1 Mass transfer estimations
Before measuring
materials, it
was
properties of
mass transfer
necessary to
determine
what rate
coating
would
be
required to achieve significant microbial stability improvements.
required
apparent
diffusion coefficients have been summarized in Table 7.
In order
Based on
expressions derived
in
Appendix C
to evaluate the magnitude of these values, D values for different
systems have been summarized in Table 8.
seems that an
appropriate coating will
From this comparison it
have to allow
diffusion 1,000 times slower than food-like matrices.
K-sorbate
-
48
-
Figure 4
SCHEMATIC REPRESENTATION OF MICROBIAL STABILITY ENHANCEMENT BY
HIGH CONCENTRATION OF K-SORBATE DEPOSITED ON A HIGHLY
IMPERMEABLE EDIBLE FOOD COATING
aw of a surface region may
*_
.
over aw* as a result of
_,_increase
unsteady state environments
.
bulk food stays at aw*
bulk and surface contains 0.3%
K-sorbate
a.
Uncoated food
edible diffusion barrier coating
K-sorbate deposited on the surface,
concentration >> 0.1%
surface is protected against
increases over aw*
bulk food contains 0.1% K-sorbate
homogeneously distributed
b.
Coated food
-
49
Experimental device for potential coating tests
3.4.1.2
The search
appropriate coating
for an
As
shown in
Appendix D
the use
of a
regenerated cellulose (dialysis membrane
30
microns,
Union
Carbide,
Chicago,
A
effectiveness.
diffusion
reference
value,
=
type 30F0, thickness
Illinois)
opportunity to obtain rapid, even though approximate
of a coating
was
procedure
film permeability in a
facilitated by measurements of
cell.
-
direct measurement
provided
an
measurement
on the
food
model itself would have been much more laborious.
Table 7
APPARENT DIFFUSION COEFFICIENTS REQUIRED TO ACHIEVE SIGNIFICANT
MICROBIAL STABILITY IMPROVEMENTS
D,
Number of days
h = 0.01 mm
x10~ 1 0
1
5
10
30
14.7
2.9
1.5
0.5
cm 2 /sec
h = 0.03 mm
h = 0.05 mm
x10~ 9
x10~ 9
13.3
2.7
1.3
0.4
36.8
7.4
3.7
1.2
-
50
-
Table 8
D VALUES FOR DIFFUSION THRU A SOLID MATRIX
Diffusing specy
D, cm 2 /sec
Solid
Reference
Water
fish muscle
3-30x10- 7
Jason (1965)
Water
alfalfa wafer
3- 8x10- 6
Bakker-Arkema
et al. (1964)
Water
tree periderm
tissues
3-45x10- 8
White (1978)
Water
sugar beet
3.4.2
Formulation of effective coatings
3x10- 7
Vaccarezza
et al. (1974)
The approach used in this study for the formulation
effective coatings
have been
summarized
could be
specific information
found in
in Figure
5.
Little
our literature
survey.
Particularly scarce was the information on transport
Formulation was facilitated by
simple tests
using
These guidelines
the
concerned
the
cell previously
food
diffusing molecule and the intended
summarized as follows:
properties.
the following guidelines and
diffusion
of
surface
described.
conditions,
effect sought.
the
They can
the
be
-
51
-
Figure
EXPERIMENTAL APPROACH TO THE SELECTION OF COATING COMPOSITIONS
Literature survey
Materials of interest
Selection of alternatives
Composition range and procedures
Preliminary experiments
(solution stability, microscope
slide coating, water swelling,
mechanical properties, etc.)
Specific compositions and
procedures
Diffusion cell studies
Permeability values
I
Comparison with previously
obtained values
-
a.
All
had
coatings
be
to
52
-
grade
food
organoleptically
and
acceptable.
b. The surface where the coating would be applied is at an aw
in
were interested in properties
in
Therefore we
the IM range.
this
aw
eliminated
This
range.
would
that
materials
experience severe water swelling.
c. Mass transfer would be slower in a coating with a very 'tight'
microstructure, i.e.
leaving little free
one
ones or
with bulky
over
would
that
side groups
the desired
hinderances to
present steric
the
This favored linear polymers
diffusion of sorbic acid.
highly branched
space for
structure.
tight
Numerous charged groups would have the same effect.
Once a
coating
composition is
selected,
preliminary
trial and error experiments can be used to determine formulations
films
that produce
adhesivity.
with reasonable
strength,
flexibility
The final and critical step were the measurement
and
of
K values.
3.5
The reduced surface microenvironment pH approach
to IMF
A second approach
surface microbial
stability
enhancement proposed in this study was the establishment of a
difference between
resistance
at
food surface
reduced
pH
is
and food
bulk.
associated
bacteriostatic effectiveness of lipophilic
The
with
pH
increased
increased
acids such as
sorbic
-
acid at low
pH.
53
-
predict
that
it
should
surface
form of the preservative,
availability of the most effective
can
reduction increases
surface pH
Since
result
in
we
surface
significant
stability improvements.
Preliminary
samples with
surface
a
growth
experimental
reduced
of
surface pH
S.auxeus
untreated controls, e.g.
work
are
S-6
that
confirmed
capable
significantly
of
food
resisting
longer
than
These tests showed also that
Figure 6.
as soon as the pH
difference disappeared rapid growth
Experimentally, a
non permanent
pH difference
occurred.
was achieved
inclusion of low molecular weight acids in a zein based
by
coating.
Diffusion was assumed to be the mechanism for pH equilibration.
3.5.1
Theoretical considerations
The Donnan model
that it
is possible
differences.
for semipermeable membranes
permanent
pH
the following discussion will
be
conditions for
to establish
The objective of
suggests
to describe:
a. how
this
theoretical
model
can
be
interpreted
for
our
particular application, and
b. how can it be used to identify the parameters which will allow
us to
establish
difference.
the
necessary
conditions
for
maximum
pH
-
54
-
Figure 6
PRELIMINARY MICROBIAL TESTS AND pH DETERMINATIONS
IMF MODEL No.1 COATED WITH A pH 4 ZEIN SOLUTION
a.
b.
pH
log N,
cells/cm
8
6
7
*UNCOATED
6
*SURFACE
V CLOSE TC
ACENTER
5
A ZEIN, pH 4
5
2
4
a. cell counts on coated
and uncoated samples
4 1-
-
2
6 days
37* C, 87%RH
4 days
b. pH values as function of
location and time for
coated samples
-
-
55
3.5.1.1 The Donnan equilibrium model
The
equilibrium
Donnan
model
ionic
the
describes
concentration differentials created by the presence of a membrane
could
move freely from the
it
macromolecule
charged
other
coating while
electrolytes,
other solutes, particularly
components, water and
would be able to
a
a food surface
form of
immobilized in the
of
situation
the
represent
charged
is that
behind this modelling
The rationale
macromolecule.
to a
electrolytes but impermeable
low molecular weight
to
membrane is assumed permeable
The
separating two solutions.
food bulk to the
surface
and viceversa as schematized in Figure 7.
The first
system to
consists of water as the
be
dealt with
(Hiemenz,
1977a)
(P-z)
solvent, a charged macromolecule
and a low molecular weight uni-uni-valent electrolyte (M+X-).
The
specific
described in
analysis is that
designate
the
situation
macroion
be
Figure 8.
p-z,
macromolecule,
negatively charged
to
with
We shall
i.e.
a valence
this
in
considered
arbitrarily
of
consisting
number
-z.
a
This
macroion will be placed on side 1 of the chamber separated by the
semipermeable membrane M-M',
but permeable to ions M+ and
be found
on
both
sides
The
X-.
of the
concentrations because of the
membrane.
This membrane is impermeable to
condition for
At equilibrium M+ and Xmembrane,
but
presence of P-z on
equilibrium is
potentials should be equal, therefore:
not
in
side 1 of
that the
p-z
will
equal
the
chemical
-
56
-
Figure
SCHEMATIC REPRESENTATION OF A DONNAN THEORY MODEL FOR THE
ANALYSIS OF A pH DIFFERENCE BETWEEN SURFACE AND FOOD BULK
ELECTROLYTES IN
COATING
OHNa+
,-CHARGED
MACROMOLECU
Cl~ ( ','
7
HYDRATED
OH~ Na*
is
Nac
-4'
C
f
ELECTROLYTES
IN FOOD
H*
-
57
-
Figure 8
SCHEMATIC REPRESENTATION OF A SEMIPERMEABLE MEMBRANE
ANALYSIS FOR AN IMPERMEABLE CHARGED POLYELECTROLYTE
AND UNIVALENT PERMEABLE IONS
SIDE 2
SIDE 1
M
p-z
--- >
M+
__.--->
M+
X-
----..--- >
X~
-
MX/l
58 -
(2)
MX/2
=
But:
MX/1
MX/2
+
RT
MX/2
+
RT ln
in a
(3)
MX/1
0
MX/2
a
MX/
(4)
2
Therefore:
=
a
(5)
a
MX/2
MX/l
Equation (5) can be turtner expanded using:
=
a
(6)
+ a
a
MX
M
X
Therefore:
a
+
=M ++/2 aM +
a _
M /1 X /l
(7)
/2
Expression (7) states that the ion activity product
is
constant on both sides of the membrane, although the activity
of
ions M+ and X- is not.
a
+
M
+ m
-
M
a
Remembering that:
+
(9)
y
X
X
Y
=
+ /MX
(10)
+Y
Y
M
X
where:
Y.
(8)
M
= activity coefficient, component i
-
59
-
m= molal concentration, component i
Assuming:
Y
~
Y
+/MX/1
(11)
+/MX/2
we obtain:
m +
m _
SM/
X /
=
mn +
M
i(2
(12)
X
/2
Another restriction on the system is that both sides of
the chamber have to be electrically neutral.
+ m
z m
P-Z
X~/l
mn
m
=
(13)
M /1
=
(14)
inm
M /2
X-/2
be
the
for
obtained
unknowns
ions
weight
macromolecule, i.e.
in
compartment
the
side 1,
mM+/1
and
These
mX-/'.
the concentrations of the
expressions were used to evaluate
molecular
expressions
14 two quadratic
From expressions 12 thru
can
Therefore:
in
terms of
with
the
the valence
low
charged
z,
the
concentration of the macromolecule and the ionic concentration in
side 2.
1
The dependence of the ionic concentrations in chamber
is easier to visualize when numerical values are used as shown in
Table 9.
These
charge z = -10
calculations
assumed a
and a concentration
macromolecule
ranging from 10-4
to 10-3.
This is equivalent to a molecule with a m.w.= 10,000 used in
1-10% concentration range.
a
with
the
-
60
-
Table 9
VALUES OF mM+/1 AND mx-/1 FOR TWO VALUES OF mM+/2 AND A RANGE OF
VALUES FOR z mp-z.
ALSO INDICATED ARE THE LOW MOLECULAR
WEIGHT ELECTROLYTE RATIOS BETWEEN SIDE 1 AND 2.
a)
mM+/2 = 10-3
z mp-z
x10
3
1
2
4
6
8
10
a)
mM+/1
x10
mM+/1/mM+/2
3
1.62
2.41
4.24
6.16
8.12
10.10
mx-fi
04
m-/2
m1-
x104
1.6
2.4
4.2
6.2
8.1
10.1
6.18
4.14
2.36
1.62
1.23
0.99
0.62
0.41
0.24
0.16
0.12
0.10
mM+/1/mM+/2
mx-/1
mX-/1/mX-/2
mM+/2 = 10-2
z mp-z
x10
1
2
4
6
8
10
3
x10 2
1.05
1.11
1.22
1.34
1.48
1.62
x10
1.05
1.11
1.22
1.34
1.48
1.62
3
9.51
9.05
8.20
7.44
6.77
6.18
0.95
0.91
0.82
0.74
0.68
0.62
-
61
-
concentration of the cation
Table 9 shows that the
M+
is higher in the phase containing the charged macromolecule while
becomes
The difference
for the anion X-.
the reverse is true
larger when the concentration of the charged macromolecule and/or
is lower.
concentration of the permeable electrolyte
groups present,
charged
more immobilized
words, the
the
higher and when
groups it carries is
the number of charged
In
other
the
more
asymmetrically the simple electrolyte will be distributed.
not
have
calculations
previous
The
considered
the
presence of water and the proton and hydroxyl ions that it forms,
should include
interested in pH values
Since we are
9.
as shown in Figure
expressions (Grignon
our
them in
and
we
Scallan,
1980; Scallan and Grignon, 1979; Donnan, 1934, 1924, 1911; Donnan
and Guggenheim, 1932).
considering the six
the
determine
now
Let us
condition
deviations
All mirror image equilibrium
conditions
were not considered since they
would yield identical equations.
We should also remember that:
a
_
OH /l H /1
+
a _a
OH /2 H /2
(15)
KW
W
=
(16)
=
Applying the equilibrium condition to Case 1:
o
= dG
=
P
)
(dn
x~/1
by
of equilibrium
particular cases
depicted in Figure 10.
equilibrium
X~/1
+ y
_
OH /1
(dn
_
OH /1
) +
-
-
62
Figure 9
SCHEMATIC REPRESENTATION OF A SEMIPERMEABLE MEMBRANE
EFFECT ON pH
COMPONENT
MOLAL CONCENTRATION
MOLAL CONCENTRATION
SIDE 1
SIDE 2
Kw/Y
OHp-Z
+ n
+
zmp-z
-
y
Kw/x + m -
n
m
y
x
Kw/y
Kw/X
np- Z
x
-
63
Figure 10
EQUILIBRIUM DEVIATIONS USED FOR THE EQUILIBRIUM ANALYSIS
OF A SEMIPERMEABLE MEMBRANE
i >------>
<-C---
Case 1
OH
<-----
H+
Case 2
>
----M
Case 3
Case 4
OH~------>
OH~------ >
Case 5
Case 6
-
64
-
(17)
(dn
X~/2
X /2
OH /2
OH /2
From mass balance and electroneutrality considerations:
dn _
x /l1
= dn
_ = dn
OH /1
= dn
X /2
= dn
(18)
OH /2
Therefore:
OH /l
= P
+ 11
+ y
X /2
OH /2
X /l
(19)
+ RT
(20)
But:
0
1
Ii.
=
in a.
Substituting in (19):
=
X~/2
a
OH /i
(21)
a
X /l
OH /2
Similarly for cases 2 thru 6 described in Figure 10:
(22)
=
aH /2
aM /i
=
aX-/l
aM+/l
(23)
=
aX-/l
aH/l
(24)
=
aM
aM /2
aH /i
aX-/2
aM /2
aX-/2
aH+/2
a M+/2
aOH /2
aH /2
aOH /2 =
aH /
l
aOH/l
aOH-/i
(25)
(26)
Not all of the above equations are independent of
other.
26.
It is possible to eliminate four of them:
each
21, 22, 25 and
Only for algebraic simplicity we shall substitute activities
-
From
by molal concentrations.
-
65
the
23 and 24
equations 15, 16,
following series of equalities was obtained:
m +
M /
m _
X /2
_
M+
H /
m
_
refer
now
considerations
electroneutrality
-(27)
OH /2
_
back
to
mH/2
m
us
Let
m +
_
expressions for mM+/1 and mM+/2-
where
determine
to
used
were
9
Figure
expressions
Substituting these
in (27):
K
+ n + zm
/y
- y
=
-
K W/y + M -
n
x
KWX
y
m
x
(28)
-=
K/y
This expression can be rearranged to obtain:
Kw/y
+ n + zm
(29)
KW/X + m
Substituting in (29) the following expressions derived from (28):
x
(30)
m =X n
(31)
y =
we obtain:
zm
1 ++
(32)
KW/y + n
This expression gives the distribution constant for all
permeable solutes
in terms
of
the number
of charges
and
the
-
concentration of the
66
-
charged macromolecule, the
anion concentration, all in side 1.
represents the pH
that log X
i.e.
between coated
proton and
the
It is important to emphasize
1 and
difference between side
surface and food
bulk in the
2,
our
case of
From equation (30) we obtain:
intended application.
log X = log y - log x = pH(food bulk) - pH(coating) (33)
3.5.2
Applications to an IMF model
3.5.2.1 IMF model requirements
As
shown
in
Figure
macromolecule can be immobilized
other molecules,
coating, while
move freely.
The
key parameters
7,
negatively
a
in the form
of a food
groups on the surface.
surface
particularly electrolytes,
in the expression
difference between coated surface and food bulk are,
concentration in the food and
charged
for the
charged
emphasized in Table
which reports the calculated pH differences as affected by
two parameters.
pH
electrolyte
the number of immobilized
This is further
can
10
these
It is quite apparent that significant ApH values
achieved in
an
IMF system
with
low
electrolyte
can only
be
content.
To prove the validity and potential application of
the
pH difference concept, an IMF model had to satisfy this condition
(IMF model No.2).
-
67
-
Table 10
pH VALUES AS AFFECTED BY THE CONCENTRATION OF CHARGED GROUPS AND
THE CONCENTRATION OF ELECTROLYTES PRESENT IN A FOOD PRODUCT
Charged groups, [M]
Electrolytes, [M]
log X = ApH
0.001
0.0100
0.0010
0.0001
0.02
0.21
1.00
0.010
0.0100
0.0010
0.0001
0.21
1.00
2.00
3.5.2.2 Selection of polyelectrolyte
Several food grade polyelectrolytes were considered for
Among them
we
had pectic
acid,
formulation.
a coating
carriers into
incorporation as negative
xanthan gum,
and
furcellaran
carrageenans.
A linear
chain
connected by a -(1-4)
substances.
Some of
linkages
is the basic structure of
the carboxylic acid
groups are
and others are in the form of free acids or salts.
charged groups
methylation.
can
be
acid
of anhydro-D-galacturonic
increased
by
altering
units
pectic
methylated
The number of
the
degree
of
-
68
-
The molecular structure
main chain built up
and 4 positions.
of xanthan gum
of 8 -D glucose
The side
consists of
a
units linked through the
1
chains consists of two mannose
units
and a glucuronic acid unit, shielding the backbone of xanthan gum
and could be the
major reason for
the enzymatic resistance
and
the uniformity of chemical and physical properties.
Figure 11
shows the
furcellaran and carrageenans.
these three
seaweed galactans
residues linked alternatively
The presence
of
molecular structure
them with
show
sulfate
electronegative
groups
their
p-(1-4)
by B-(1-3) and
hydrophobicity making
whereas sulfate
agarose,
The idealized representations
anhydro ring give them different properties.
ring confers
of
groups
and
of
galactose
linkages.
the
3,6
The presence of the
less
the galactan
confer hydrophilicity,
making
soluble,
it
more
soluble.
The ideal polyelectrolyte should have a large number of
strongly
dissociated
groups.
Solubility
application should also be considered.
and
easiness
of
-
69
-
Figure 11
MOLECULAR STRUCTURE OF AGAROSE, FURCELLARAN AND CARRAGEENANS
Sulfate content
ye
agarose
H
H
OH
H
HM
0-3
H
C O
furcellaran
K -carrageenan
H
A
/H
CP
.H
L
C
12-16
25
H
SO1 H
AH
-
I*
OH
30
carrag eenan
0isC H OH0
0
CHOS
OH0
H'
X- carrageenan
35
Sulfate content % is given as an indicator of substitution, e.g.
X-carrageenan has on the average about one sulfate group per
repeating unit.
-
70
-
4. Experimental Procedures
4.1
Preliminary experiments
4.1.1 The IMF model: composition, preparation and analysis
An IMF model
was required to
test the feasibility
the surface modification procedures proposed in this study.
product
chosen
was
an
IM
cheese
laboratory (Motoki et al., 1982).
11.
A
brief preparation
Glycerol (Certified
NJ), sorbitol
model
A.C.S.; Fisher
(Pfizer Co.,
this
Components are shown in
Table
is shown
(about
mycostatic agent.
50C)
The
New York,
with K-sorbate
proteins used
protein (Ajinomoto USA.,
New York,
in Figure
Scientific Co.,
NY) and
(Baker Analytical Reagents, Phillipsburg,
warm water
The
in
description
developed
of
12.
Fair
sodium
Lawn,
chloride
NJ) were dissolved
(Pfizer
were:
NY) and
Co.)
as
in
the
isolated
soybean
sodium and
calcium
caseinates (Erie Casein Co, Erie, IL).
An emulsion paste
was obtained by
preblending 30%
the proteins in a Waring blender with the aqueous solution.
of
The
remainder of the protein was mixed with an oil phase composed
of
hydrogenated vegetable oil (Durkex 500, SCM Corp., New York,
NY)
and decaglycerol decaoleate (Glyco Chemical Inc., Greenwich,
CT)
and then
emulsified using
Kitazawa Sangyo,
a food
Co., Tokyo,
cutter (2
Japan).
blades, 1450
This emulsion
rpm,
was
then
blended with the emulsion paste using the same food cutter.
After emulsification, the mixture
was adjusted to
desired pH with glucono-delta-lactone (GDL) (FMC Co.)
or
the
sodium
-
71
-
Table 11
COMPOSITION OF THE IMF MODEL SYSTEM
Ingredient
isolated soy protein
%w/w
26.1
Na-caseinate
5.9
Ca-caseinate
2.0
hydrogenated vegetable oil
decaglycerol monooleate
34.0
.4
salt
4.8
glycerol
5.9
sorbitol
19.3
.4
K-sorbate
cheese flavor
monosodium
glutamate
1.0
.2
GDL/SAL
[1]
water
[2)
[1]
[2]
Glucono-delta-lactone (GDL) or Sodium Aluminum phosphate
(SAL) is used to adjust the mixture to the desired pH.
The initial water content is 100ml/100 g solids; this
amount is reduced by drying the mixture w/ hot air so
as to achieve the desired final aw.
-
-
72
Figure 12
PREPARATION OF IMF MODEL SYSTEM
Oil, Emulsifier
Proteins
Water & other
ingredients
I
I
I
Mix by hand
Magnetic stirrer
\1
113
Heat to 50C
3
Mix w/food cutter
Mix w/blender
Emulsify w/
food cutter
pH measurement
Dry w/ hot air while in
food cutter
Casing
I
Pasteurize, 45 minutes in
80C water bath
I
I
final
Cooling
Check
pH and aw
-
-
73
Co., Westport, CT).
aluminum phosphate (Stauffer Chemical
of the water
food cutter
bowl.
When
emulsion was filled
the desired
into a
aw level
the
was reached
the
(Type 30F0,
cellulose casing
Union
were
Thereafter they
Carbide, Tarrytown, NY) with a hand press.
placed. into seamless vinylidene
air into
by blowing warm
was then evaporated
Some
(diameter
chloride casing tubes
40 mm, Kreha Chemical Co., Tokyo, Japan).
The emulsion was then pasteurized at about 85C for
minutes in a water bath.
After cooling, pH was determined with a
Beckman
39507,
(combination electrode
probe
surface electrode
45
Instruments, Inc., Cedar Grove, NJ).
parameters
Other important
electric hygrometer
moisture
relatively
low
drying
by
the
temperature
oven
vacuum
was
selected
method.
to
an
Zurich,
Sina,
Beckman Instruments);
the USA by
measured
content
using
measured
Nova
Equihygroscope,
(SINA
Switzerland; marketed in
aw,
were
and,
A
minimize
humectant losses.
Finally, concentration of the preservative used in this
model, K-sorbate
was determined
by
HPLC.
A
procedure
simple
based on the paper by Park and Nelson (1981) was found to satisfy
our needs
(Figure 13).
prepared with 0.2%
Analysis
K-sorbate showed
(0.199 + 0.01)%, n = 8.
of an
IMF model
No.1
excellent recovery
batch
values,
-
74
Figure
SORBIC ACID ANALYSIS BY HPLC
Food sample, x g (<25 g)
5 g Celite
100 ml MeOH
Blender, 4 minutes
Filtration (Whatman No.2)
Wash solids w/ 50 ml MeOH twice
-
discard solids
Dilute filtrate to 250 ml w/ MeOH
Transfer to separatory funnel
Add 100 ml 0.5 N NaOH
50 ml 1:1 Petroleum/ethyl ether mixture
Organic phase:
Wash gently w/ 25 ml 0.5 N NaOH
Aqueous phase:
Transfer to 500 ml Erlenmeyer
Organic:
Discard
Titrate to pH 2 w/ HCl (1:1)
Aqueous
Add to aqueous phase
I
Pass thru Millipore filter
HPLC, 30% acetonitrile,
C-18 column
-
4.1.2
75
-
Preliminary microbial studies
The microorganisms
based on
resistance to
their
microbiology.
used in
These
are
low
this study
aw and
were
in
food
niget
and
importance
Aspe gillus
S.auteuA S-6,
selected
A.tonophitus.
4.1.2.1
Determination of inoculation conditions
Preliminary tests were
surface
Food
outgrowth.
required to determine
S.auteus
with
optimal
fabricated
samples
microbial growth conditions (i.e.
neutral
pH and high aw)
were
inoculated at several levels by dipping them in glycerol solution
(Difco,
Growth was followed by plating on BHI
cell suspensions.
Detroit, MI) agar plates and expressed as viable counts/cm 2 .
cells/cm 2 .
future
This value
experiments.
determination of
cell
was used
A
cell
guide for
as a
complement
to
concentrations in
solutions used to inoculate
desired initial
observed was about
the maximum growth
shown in Figure 14
tests
the
aqueous
food samples, so
concentration.
Similar
3x10 8
the design
these
was
As
of
the
glycerol
as to achieve
the
preliminary
test
were also required for the mold challenge studies.
It is important
to note that
only in the surface modification
inoculated.
the surface.
since we are
interested
of IMF's, only the surface
Later on, we will show that outgrowth affected
Therefore results
or spores/cm 2 .
could be expressed as
was
only
cells/cm 2
-
76
-
Figure 14
SURFACE MICROBIAL CHALLENGE WITH S.aaueua
AND NEUTRAL pH
S-6 AT HIGH aw
-
-
77
4.1.2.2 Growth response in the presence of high K-sorbate
concentrations at various pH levels
This was
preliminary experiment
another
the ability
found on
of K-sorbate
1982; Lahellec et al.,
Robach,
1978).
at
Potter,
1981; Reinhard and Radler, 1981; Ivey
However,
effectiveness
growth
and
1982; Lynch
et al.,
1982; Blocher
(Gordon Greer,
microbial
to retard
no
information
above
levels
the
was
found
to
was
Much information
information.
provide necessary background
required
and
on
its
authorized
legally
concentrations.
Various amounts
(Difco, Detroit,
MI)
of
and
pH
was
Results have been represented in
results are well
pH =
that above
and high
incapable of growth inhibition.
HCl
with
adjusted
Figure 15.
At 0.3%
in
aw
level, the
TSB
(0.1N).
K-sorbate,
states
published work, which
in accordance to
5.5
dissolved
K-sorbate were
preservative
From this experiment, it
is
could
be inferred, that an effective K-sorbate surface concentration at
pH values closer to neutral should be approximately 0.5 to 1.0%.
4.1.3
Stability of the uncoated IMF model
been
Procedures for the microbial challenge tests have
summarized in
Initial
Figure 16.
inoculation and
growth
death) of S.auteus was determined by plating on BHI agar
Initial inoculation
obtained by
plating
levels
on
of
A.niget
and
Saboureau dextrose
plates.
A.tonophitz
agar
(SDA,
(or
were
Difco)
-
78
-
Figure
HIGH K-SORBATE CONCENTRATION EFFECT ON S.autea
VIABILITY
Cell
w/0.1N HCl.
pH adjusted
was TSB,
Media utilized
concentrations were determined at 19, 39, 49, 97 and 216 hours.
Values reported correspond to increases over inoculation level
(+), decreases (-) and no viable cells found (k).
pH
6.5
-
* KK
6.0 1O-KKKK *-[(IK
I&.KK
5.5
5.0
.0---KK
00.3
KKIKK
1.0
%
3.0
5.0
K-sorbate
10.0
-
79
-
Figure 16
INOCULATION PROCEDURES FOR UNCOATED SAMPLES
S.auteaA
A. nigeA, A. tonophitus
BHI slant
24 h, 37C
SDA slants, 3-5 days,
room temperature
BHI broth
4-5 h, 37C
SDA Petri dishes
1 week, room temperature
I
I
I
suspend spores in sterile water
centrifugation
resuspend pellet in
glycerol/water solns.
at desired aw
i
Dip samples in cell or
spore suspensions
To determine inoculation
level plate on BHI or SDA
If challenging w/ cells store
at 37C and sample periodically.
If challenging with spores store
at room temperature and examine
periodically visually for growth.
-
plates.
Mold
was
outgrowth
-
80
by
determined
direct
visual
examination of inoculated samples.
Inoculated samples were placed in sterile Petri
and stored
in
a
dessicator
at
a
aw
adjusted
dishes
with
glycerol-water mixtures to that of the specific sample.
sterile
Whenever
possible, saturated salt solutions were used instead.
of
The objective
tests was
these
to
the
determine
microbial growth potential of our particular IMF model surface as
a function
the
of
fabrication parameters:
pH
and
The
aw.
results served both as a guide for the design of experiments
coated samples and as the associated uncoated controls for
for
these
samples.
4.2
The reduced preservative diffusion approach
4.2.1 Permeability experiments
4.2.1.1 Preliminary tests
Figure 17 shows the diffusion cell.
magnetic stirrer
provided
to
high concentration
the
to load the cell with
the
solution, eliminate
eliminate the
permeability.
at
reduce
The side tubes were used
interfaces.
levels to
were
influence
A mechanical and a
resistance
air bubbles
of hydrostatic
and
adjust
pressure
Samples were taken from the upper chamber and
preservative concentration was determined
on
the
spectrophotometrically
(at 268 nm, absorption maximum determined experimentally) with no
need for an extraction procedure.
-
81
-
Figure 17
PERMEABILITY CELL USED IN THIS THESIS
CHAMBER 2
CELLULOSE
CHAMBER I
-
The solvent used
82
-
was 50% glycerol
(w/w) which
should
give:
a. a aw
level
in
important since
the intermediate
we want
moisture
a sorption
range.
status for
This
the
is
coating
similar to actual use conditions.
b. a solvent viscosity in the
in IMF's (Perry, 1963, p.
range of solutions commonly
3-199).
The
advantages
of
availability,
inertness,
mechanical
Dialysis tubing
made
homogeneous, thus
out
found
of
strength
facilitating
the
and
low
cellulose
regenerated
obtention
of
its
were
cellulose
regenerated
cost.
are
very
reproducible
results.
The diffusion cell
determine K values.
agreed
very
well
was tested for
its suitability
As shown in Table 12, diffusion cell
with
those
obtained
with
a
to
values
dialysis
sac
arrangement.
4.2.1.2 Permeability tests on selected ingredients
Permeability tests
on
selected
coating
compositions
were determined according to procedures schematized in Figure 18.
Measurements on plain cellulose were often repeated to check
experimental artifacts (e.g.
agitation speeds).
for
-
83
-
Table 12
COMPARISON OF K-VALUES
Experimental device
K-values x10 2
(mg/hr cm 2 )/(mg/ml)
dialysis sac
5.7
4.1
2.6
4.2
4.2 + 1.3
dialysis cell
4.8
4.8
4.6
4.7
4.6
4.7 + 0.1
-
84
-
Figure 18
EXPERIMENTAL PROCEDURES TO DETERMINE K-VALUES FOR SELECTED
COATINGS
Prepare test coating solution
Cast on cellulose support
Mount on dialysis cell
I
Fill chamber 1 with K-sorbate
in 50% glycerol solution
Fill chamber 2 with pure solution
Start mechanical and magnetic stirrers
Take samples from chamber 2
Measure UV absorbance
(268 nm)
Examine coating under magnifying
magnifying glass. If deffective,
discard experiment
-
4.2.2
-
85
Determination of sorbic acid stability
tests was
these
of
The objective
to
intended
the high concentrations of
stability of sorbic acid at
the
determine
use and to check for interactions with the IMF model.
model
Sorbic acid stability was assessed using the IMF
prepared as outlined
were selected to
values
expressed
0.3 to
5% (w/w,
potassium
salt
cover the
as
the
conditions existing in
Initial
4.1.1.
in section
concentrations
range,
wet basis)
form.
simulate
To
were
tests, samples
microbial stability
also inoculated with S.auteu. S-6.
Experimental procedures outlined
sorbic acid determinations by
HPLC, pH by
in Figure 19
include
a Beckman surface
pH
electrode, and S.auteus growth response as described previously.
It should be
noted that these
procedures included
uninoculated
pasteurized controls to determine if preservative losses could be
related
to
microbial
growth,
as
well
as
uninoculated
not
affect
pasteurized controls to ascertain if heat treatment could
sorbic acid stability.
4.2.3
Sorbic acid distribution studies
The objective
of these
tests was
coatings reduce preservative diffusion
into food bulk.
to show
that
from coated food
Therefore we had to determine how zein
zein
surface
coatings
affected the distribution of sorbic acid sprayed on the IMF model
-
86
-
Figure 19
DETERMINATION OF SORBIC ACID STABILITY PROCEDURE
IMF model fabrication
Non pasteurized -
pH determination
Pasteurization
Uninoculated
w- pH determination
Moisture equilibration
Inoculation stock
I
1:100 dilution w /
aqueous glycerol
Inoculation
I
I
Storage
Determination of:
Sorbic acid by HPLC
Microbial counts
-
surface.
87
-
Moreover, we wanted to show that the distribution was a
function of a coating
mass transfer controlling parameter,
e.g.
thickness.
4.2.3.1 Microscopy studies
Nomarsky microscopy (Peil,
the general appearance of
their thickness.
sections and
1982) was
zein coatings as
used to
well as to
observe
estimate
CO 2 frozen samples were sliced into 10
microns
To
prevent
placed immediately
glass slides.
on
artifacts caused by dehydration, samples were kept in a
constant
and
examined
relative humidity
chamber (over
saturated BaCl 2 )
within 24 hours.
4.2.3.2 Sorbic acid determination
The effectiveness
of zein
diffusion controlling barrier was
into a
core
and
a
surface
coatings as
a sorbic
samples
studied by separating
piece,
Figure
acid
acid
Sorbic
20.
concentration in each fraction was then determined by HPLC.
4.2.4
Microbial challenge tests
Microbial challenge
situation condition.
studies
were
worse
pH was close to neutral, 6.4, and food bulk
contained only a minimum amount of preservative, 0.1%
Pasteurization, 2 hours
insure low microbial
a
on
based
at BOC in
a water bath,
background counts.
K-sorbate.
was needed
Two different
to
storage
-
88
Figure
EXPERIMENTAL PROCEDURES FOR THE DETERMINATION OF
SORBIC ACID DISTRIBUTION
IMF sample preparation
Freezing
Storage
Microtome slicing
Slice into a surface
and a core fraction:
I
Microscopy
Continuity,
thickness
1 cm
E.8cm
SURFACE
=
REST
CORE
(
I
Extraction
HPLC analysis
89 -
-
conditions were tested.
In the first one, samples with a bulk aw
= 0.88 were stored at 30C and 88% RH.
with bulk aw = 0.85, were stored
In the second one, samples
hours
at 30C with cycles of 12
at 85% RH and 12 hours at 88% RH.
Table 13
shows
the
IMF
model
used for the
composition of the solutions
ingredients
and
zein and sorbic
the
acid
application on the model surface.
Preparation of surface treated samples required extreme
care to:
microbial
(i) maintain constant sample aw; (ii) reduce
contamination; and, (iii) eliminate all the residual alcohol used
as a solvent to apply zein and sorbic acid.
To maintain constant aw, samples were kept in a constant RH
chamber more
that
chamber was kept
controlled by
95% of
in circulation
the use
Sr(N0 3 )2 for 88%
by the use
of saturated
and 85% RH,
Air
in
the
fan.
RH
was
BaC1 2
and
preparation time.
the
of a
salt solutions,
respectively.
The chamber
itself
was placed in a constant temperature room.
To facilitate sample
placed on a
sterile inoculation
handling, individual pieces
needles kept
on needle
were
holder
(Figure 21).
To eliminate residual ethanol, samples were kept for at
least 12 hours in the constant RH chamber after samples had
coated and after application of the preservative:
been
sorbic acid.
-
90
-
Table 13
COMPOSITION OF IMF MODEL, COATING SOLUTIONS AND SORBIC ACID
SOLUTIONS
Component
1.
IMF model (1,000 g dry basis)
A. water
salt
glycerol
sorbitol
K-sorbate
Amount
1050
48
59
193
1.5
B. isolated soybean protein
Na-caseinate
Ca-caseinate
260
60
20
C. hydrogenated vegetable oil
emulsifier
340
2. Coating solution (100 g)
zein
glycerol
Myvacet 7-00
ethanol 190 proof
3. Sorbic acid solutions (100 g)
sorbic acid
ethanol 190 proof
x
0.25x
1
(balance)
10
90
-
91
-
Figure 21
FOOD SAMPLE HOLDER FOR ZEIN COATING, SORBIC ACID SPRAYING,
SOLVENT REMOVAL AND INOCULATION
-
92
-
Figure 22
SAMPLE PREPARATION:
CUTTING, ZEIN COATING AND SORBIC ACID
SPRAYING
food sample
cut into disks
each piece placed on
a dissecting needle
samples stored in constant
RH chamber on needle holder
zein solution (50C)-.--
samples dipped or sprayed
3 times with fresh solution
uncoated
control
solvent removal in constant
RH chamber on needle holder
(about 12 h)
sorbic acid
spraying
snr vin
no surface
sorbic acid
control
solvent removal in constant
RH chamber on needle holder
(about 12 h)
I
determination of aw
-
93
-
based
procedures
Experimental
the
on
above
considerations have been schematized in Figure 22.
Casings
were
thus obtained
was cut
into
carefully removed and
disks.
the cylinder
on a sterile dissecting needle
Each disk was placed
and
stored in the constant RH chamber for at least 12 hours.
Zein solutions
at
and constant
(190 proof)
in
various concentrations
were
ratio (4:1)
zein:glycerol
ethanol
then
prepared to coat samples by dipping or spraying them three times.
A fresh
Thereafter, coated and
separated for control purposes.
were
samples
Uncoated
time.
each
was used
solution
uncoated
samples received a surface application of sorbic acid by spraying
10% preservative
them with a
ethanol was
of residual
Again removal
constant RH chamber
solution in
for at
least 12
control purposes.
were kept for
ethanol (190
done by
hours.
proof).
storage in
the
samples
Untreated
were used
Various samples
to
determine whether aw changes had occurred.
Figure 23 shows the inoculation procedure used for
the
microbial challenge of samples with various surface treatments.
S.auteus S-6 stocks were kept in BHI tube slants at refrigeration
Fresh
temperature.
two weeks.
S.auteta
cultures were
An inoculation solution was prepared by
from the slants to
(about 120
to mid-log phase
diluted 100
prepared approximately
times
concentration.
with
40%
BHI broth.
- 160
transferring
After incubation at
30C
the broth
was
Klett units)
glycerol to
Samples were sprayed
every
reduce
with this cell
aw
and
cell
suspension,
-
94
-
then placed in individual Petri dishes and stored in
dessicators
at 85 or 88% RH and 30C.
Table 14.
summarized in
have been
Sample treatments
These samples were used in the following studies:
used to
core of
maintain the
have no
our
Otherwise,
sterility.
validity, i.e.
close as
outgrowth would
possible
not necessary
enhance
to
to
would
studies
surface challenge
failed to
had
coating
that the
samples as
our
the
that it was possible
we had to prove
nature of our tests
to
Due
microbial load.
background
determine
were
various surface treatments
a. Not inoculated samples after
mean
microbial
surface
stability.
b. Samples were also included to determine if it was possible
bulk of inoculated
assume that the
numbers
and therefore
cell
growth (i.e.
cells/cm 2 ).
to
samples remained
sterile
as
surface
with
various
could be
expressed
Inoculated samples
surface treatments and bulk aw 0.88 were stored at 30C and 88%
RH.
occurred the top layer
After significant growth had
removed.
cell
Separate
surface fraction
used to
were
the
obtained for
counts
determine
core
core/surface
was
and
cell
numbers ratio.
c. Inoculated, uncoated
used
surface
were
samples
without
samples with
as
added
positive
surface
available from section 4.1.3.
added to
sorbic acid
controls.
Information
preservative
is
the
on
already
-
95
-
Figure
SAMPLE PREPARATION:
Stock culture
-
INOCULATION PROCEDURE
BHI broth inoculation
t
incubate at 30C to
mid log phase
dilute w/ 40% glycerol
(1:100 dilution)
sample on needle holder -
I
inoculate by spraying
w/ glycerol cell
suspension
i
I
place food pieces in
individual Petri dishes
store at 30C in dessicators
at constant 88% RH or use 12 h
cycles of 85 and 88% RH
I
determine survival curve
-
96
-
Table 14
IMF SURFACE TREATMENTS
1. Zein coating:
none
coating by dipping
coating by spraying
2. Surface sorbic acid (1]:
none
surface sprayed with 10% solution
3.
Inoculation:
uninoculated
inoculated
4. Bulk water activity:
0.88
0.85
5. Storage conditions:
constant RH
RH cycles
[1] All samples contained 0.1% K-sorbate in the bulk
e. Inoculated, coated
surface were
sets of
samples
used to
experiments
with
determine coating
will be
reported.
samples with a bulk aw = 0.88
The second
one was
a test
added
to
effectiveness.
In the
first
were stored at 30C and 88%
on samples
stored at 30C with cycles of 12
at 88% RH.
sorbic acid
with bulk
aw =
hours at 85% RH and 12
the
Two
one,
RH.
0.85
hours
-
97
-
4.3 The controlled surface microenvironment pH approach
4.3.1
IMF model reformulations
After a theoretical analysis
we concluded that it
possible to establish permanent reduced surface pH conditions
the IMF model had low total electrolyte concentration.
was
if
Therefore
Table 15
LOW ELECTROLYTE IMF (IMF No.
2)
Amount, g
Component
26.1
7.9
34.0
1.6
5.9
47.0
0.4
100 to 56.2
ISP
Caseinates
Hydrogenated vegetable oil
Emulsifier
Glycerol
Sorbitol
Sorbic acid
Water [1]
[1] Moisture content reduced by air drying
4.3.2
Coating procedures
we had to reformulate our model as shown in Table 15 (IMF No.2).
Sodium chloride
was
substituted
by
sorbitol
with
equivalent
osmolality, while electrolytes present in the soybean isolate and
caseinates were eliminated by dialysis.
We had shown in Table 10
-
98
-
Table 16
PROTEIN FRACTION DIALYSIS REQUIREMENTS [1]
[M] electrolytes
Permeate
IMF No.2
[3]
[2]
protein
concentration
(g/g water)*100
IMF No.2
Dialysis
46.4
7.0
equivalent permeate
conductivity [4]
micromho
<1.5
<0.00015
<0.001
[1] The objective of these calculations is to estimate permeate
conductivity values based on protein concentration in the
IMF model and the solution to be dialyzed (46.4 and 7.0%).
(2] Assuming that we are using a coating with 0.01 M charged
groups and that the goal is a pH differential = 1
(3] Calculated using the ratio 46.4:7.0 to account for the
concentration difference between IMF No.2 and dialysis
solution (46.4 and 7.0%, respectively).
[4] Estimated using a NaCl standard curve.
Electrolyte
removal
conductivity measurements.
was
A NaCl
followed
electrical
by
standard curve
was used
to
estimate electrolyte concentrations from electrolyte conductivity
measurements.
were not
Measurements on
possible
the protein solutions
because proteins
measurements were used instead.
are
conductive.
themselves
Permeate
-
99
-
Table 17
X-CARRAGEENAN COATING COMPOSITION
Component
Amount, g
Deionized agarose
Deionized carrageenan
Sorbic acid
Propylene glycol, 40% solution
1
1
1.2
to 100
pH differences
that significant
surface
were
concentration
possible
only
kept
below
was
food and
between bulk
when
the
total
approximately
0.001
treated
electrolyte
This
M.
restriction was used to estimate protein dialysis requirements as
shown in Table 16.
The polyelectrolyte chosen
for experimental tests
was
X-carrageenan incorporated in an agarose gel as shown (Table 17).
This polyelectrolyte
percentage of
was
for its
chosen
sulfate groups
--
a
solubility
and
strongly dissociated
high
group.
Agarose was chosen as the coating forming matrix for its lack
which explains
charged groups
independent
of
pH
and
considerations in our study.
why
salt
its gelling
properties
concentration,
of
are
important
-
Therefore, X-carrageenan
the formulation until
detected
Although we will show that
not affect sorbic
heating.
As
electrolytes
in
groups can
experiments.
present
in
add it
is safer to
dialysis
was
to
used
commercial
to
cause
pasteurization (2 hours at 80C)
proteins,
normally
could not be added
preliminary
acid stability, it
with
steamed
agarose was
Sulfate
after heating.
as
hydrolysis
agarose
-
obtained when
Best coatings were
for 30 minutes.
100
does
after
eliminate
reagent
grade
X-carrageenan and agarose.
4.3.3
Microbial stability tests
Microbial challenge studies were again based on a worse
situation condition.
bulk pH was 6.1.
A
The
aw of
IMF model No.2
was 0.88
standard homogeneous amount of
was incorporated (0.22%
sorbic acid).
while
preservative
Storage conditions
were
30C and 88% RH (over saturated BaCl 2 )The effectiveness of a reduced surface pH was tested by
inoculating the surface of coated samples with S.au Aeu
S-6.
As
outlined in Figure 24, complementary tests included determination
of pH
difference between
coated top
and uncoated
bottom,
sorbic acid stability studies using HPLC determinations.
and
-
5.
-
101
Experimental results
5.1 Preliminary microbial challenge studies
5.1.1 Bacterial challenge studies
The possibility that X-carrageenan
effect on microbial
shown in Table 18.
growth was
could have an
tested using
media prepared
The variables of interest were
concentration and initial pH.
inherent
as
X-carrageenan
As indicated in Figures 25 and
26
neither one had a deleterious effect on S.auteas.
Challenge tests were conducted at various aw and pH levels.
Results shown in Figure
tested increase
in cell
27 indicate that even
numbers
increase may be due in part
at the lowest
were detected.
However
by decreasing pH
this
to recovery rather than growth.
is also interesting to note that stability was greatly
showing the increased
aw
It
increased
effectiveness of
sorbic
acid.
5.1.2 Mold challenge studies
Results of the challenge study are summarized in
19.
Table
Nine Petri dishes, each containing three sample pieces, were
A tenth dish was used to determine
observed for visible molding.
initial inoculation
20-30 for
levels:
A.tonophita.
30-70
Table
19 shows
significantly to the stability of the
the samples was
decreased, the
molding increased.
spores/cm 2 for
pH
that
and
contributes
As the pH
of
for detection
of
IMF tested.
time required
A.nigex
-
102
-
Figure 2
EXPERIMENTAL PROCEDURES FOR MICROBIAL STABILITY STUDIES ON
THE EFFECT OF REDUCED SURFACE pH
Petri dish with IMF No.2
Pour agarose-carrageenan solution
Store over saturated BaCl2 , 30C
Inoculate after pH equilibration
Store under previous conditions
Measure pH of coated top and uncoated bottom
Homogenize
HPLC
microbial counts
-
103
-
Table 18
MEDIA FOR PRELIMINARY TESTS ON THE POSSIBLE EFFECT OF
X-CARRAGEENAN ON MICROBIAL GROWTH
Media (Results in Figure)
BHI (26)
BHI + carrageenan (25)
BHI + deionized carrageenan (26)
% carrageenan
0
0
initial pH
5.5
7.3
0.1 - 2.0
5.5 -
5.9
0.1 -
7.2 -
7.3
1
1
2.0
5.5
7.2
-
104
-
Figure 25
EFFECT OF X-CARRAGEENAN ON S.auxeu4 GROWTH
Experiments were ran with media prepared according to Table 18.
At both pH, carrageenan concentration had no effect on S.auetuA
growth. Lines have been drawn through maximum and minimum
values.
CARRAGEENAN: 0.1-2.0 %
9
E
U0
pH:- 7.2-7.3
CARRAGEENAN:0.1-2.0 %
8/
z
0
7
2
3
4
DAYS
-
105
-
Figure 26
EFFECT OF DEIONIZED X-CARRAGEENAN ON S.autea
GROWTH
Experiments were ran with media prepared according to Table 18.
(initial pH) = 5.5 showed a minor growth
1% carrageenan at pHi
delay, which was not further investigated.
DEIONIZED SYSTEM
10
Es
8
0
6
BHI
4
+1%
CARRAGEENAN
pHI
pHf
5.5
8.6
7.3
8.7
5.5
8.6
D2
DAYS
3
-
106
Figure 27
MICROBIAL STABILITY OF THE UNCOATED IMF (MODEL No.1)
pH- 55
pH- 6.4
log N, cells/cmt
8
8
7
log N,
cells/cmt
.84-85 *
U
.86
7
6
6
5
e 24-25
5
4
2
4
6
8
days
10
24
2
4
6
24
8
days
-
107
-
Table 19
NUMBER OF DAYS FOR MOLD GROWTH VISUAL DETECTION
aw
A.tonophitua
pH
A.nigex
0.85
6.5
6.0
5.5
2
40
>220
5
9
40
0.83
6.5
6.0
5.5
12
>220
>220
12
15
>220
5.1.3
Conclusions
From both
effect of pH
challenges it
on stability is
could be
concluded that
In
highly significant.
practice,
however, substantial pH decrease is not possible, due to loss
acceptable taste and texture (Motoki et al.,
5.2
5.2.1
the
of
1982).
The reduced preservative diffusion approach
Permeability experiments
It was very
difficult to
obtain good
reproducibility
within a particular treatment with coatings, even though
with uncoated cellulose gave good reproducibility.
results
This did
not
-
-
108
invalidate conclusions based on "order of magnitude" differences.
All results
obtained
were
summarized in
Table
20.
Specific
sample preparation conditions have been outlined in Appendix E.
Best
determinations
values
were
were
obtained
repeated
experimental modifications.
for
several
zein
films.
These
with
slight
times,
Particularly,
repeated using different sorbic acid concentrations.
As shown in
Based
Figure 28, a slight concentration dependence was observed.
on these results and
zein films had
as indicated in Appendix
preservative diffusion 1,000
was
experiment
the
D we showed
that
times smaller
than
the uncoated cellulose film.
5.2.2
Sorbic acid stability test
Table 21 shows that initial preservative affected pH of
the model
values in
system giving
the 6.3
to 6.8
range.
No
attempt was made to correct for this pH variation.
as percentage
of
initial concentrations are shown in Appendix F and summarized
in
Sorbic acid
Figure 29.
determination expressed
Surface growth at 0.57 and 1.06% K-sorbate have
plotted in Figure
30.
4.24%, total counts per
At
the higher
been
concentrations, 2.10
sample remained below detectable
during the length of the test (about 40 days).
and
levels
-
109 -
Table 20
SUMMARY OF K-VALUES FOR VARIOUS COATING MATERIALS AS DETERMINED
BY A DIFFUSION CELL
Coating material
Initial concentration
differential, (mg/ml)
uncoated cellulose
K-values x10 4
(mg/hr cm 2 )/(mg/ml)
470
10
+ 10
cellulose coated with [1]:
zein (2]
7.9
-
110.2
3.4 +
2.2
gelatin hot melts
100
170
210
210
caustic amylose
coagulated with ammonium
sulfate
100
80 (HVII-20-40)
90 (HVII-20-SA)
120 (HVII-20G-40)
caustic amylose
coagulated by acid
salt mixture
100
120 (HVII-20-S)
110 (HVII-20G-PS)
140 (HVII-20G-S)
low DS amylose ester
100
260 (dipped)
320 (sprayed)
[1] specific coating compositions are given in Appendix E.
[2] individual values have been plotted in Figure 28
-
110
-
Figure 28
VARIATION OF K-VALUES AS A FUNCTION OF INITIAL K-SORBATE
CONCENTRATION DIFFERENCE
mg/hr cm 2
K' mg/mi
10F
AA
60
100
Initial concentration difference , mg/ mi
-
111
-
Table 21
SORBIC ACID STABILITY STUDY, INITIAL CONCENTRATIONS
pH [1,4]
% K-sorbate
at time 00 [2]
%K-sorbate
at time 0 [3,4]
6.3 + 0.06 (n=4)
0.55
0.57
6.5 + 0.03 (n=4)
0.95
1.06 + 0.11 (n=2)
6.6 + 0.02 (n=4)
1.98
2.10 + 0.07 (n=2)
0.04 (n=4)
4.02
4.24
6.8
+
+
0.03 (n=2)
[1] corresponds to pasteurized and not pasteurized samples which
showed no significant differences
[2] time 00 denotes time 0 before pasteurization
[3] time 0 denotes time after pasteurization and moisture
equilibration. During sample preparation, moisture removal
was stopped before reaching final desired value. Final
moisture conditions were then achieved by storage over
saturated BaCl 2 solution. This explains the slightly lower
values for unpasteurized samples.
[4] n = number of replicates
Results shown in Figure 29 indicate that sorbic acid is
stable under
testing conditions.
storage at 37C and 88% RH, in
After approximately
40
the dark, losses of less than
were detected at all four concentrations tested.
days
25%
Moreover, there
is no initial concentration effect on preservative stability.
-
112
Figure 29
SORBIC ACID DETERMINATION, PERCENTAGE OF INITIAL CONCENTRATION
Except for controls, all samples were inoculated with 106 to 107
cells/cm 2 . Controls were either pasteurized [o] or not [A].
Initial concentrations, 0.57 to 4.24%, were determined after
pasteurization and moisture equilibration. This is a summary of
the values reported in Appendix F, by drawing lines thru average
points.
100
NOT INOCULATED
Co: 0.57%
Co: 4.24%
Co: 1.06 %
Co: 2.1%
CONTROLS
A
NOT
0
PASTEURIZED
PASTEURIZED
la
90
0
*0
10
20
DAYS
30
-
113
-
Figure 30
MICROBIAL GROWTH RESPONSE AT HIGH K-SORBATE CONCENTRATIONS
S
aM
E
2F
7
6
4
S
12
4
S
12
16
20
24
28
32
24
28
32
36
doYs
S
E
o
7
6
16
20
36 dys
Experiment was ran in duplicates. Lines have been drawn through
Samples with higher sorbic acid
maximum and minimum values.
concentrations, inoculated at the same level were bactericidal.
No viable counts were detected after inoculation.
-
114
-
Table 22
EFFECT OF PASTEURIZATION ON SORBIC ACID STORAGE STABILITY
REMAINING %SORBIC ACID AFTER 38 DAYS STORAGE AT 37C, RH = 87.5%
%K-sorbate
Pasteurized [1)
Non-pasteurized
0.57
1.06
2.10
4.24
84.7
83.4
91.8
84.7
84.7
84.7
86.0
76.8
86.1 + 3.8
83.1 + 4.2
[1] 2 hours in a 80C water bath
As shown in Table 21,
heat treatment (2 hours at
resulted in no detectable sorbic
80C)
acid losses.
Moreover, it
had
uninoculated
with
inoculated
S.auteun
plays
a role
in
response was
a
function
of
no effect upon storage (Table 22).
Finally,
samples to
sorbic acid
we
compared
determine if
degradation.
presence of
Growth
outgrowth
initial sorbate concentration.
At 0.57%, significant
was detected, at 1.06%, initial
decrease with later growth
observed (Figure 30), while 2.10 and 4.24% were lethal.
None
were
of
-
115
an effect on
these behaviors had
-
(Figure
sorbic acid stability
29).
Sorbic acid distribution studies
5.2.3
5.2.3.1
Microscopy studies
Photomicrographs in Figure 31 show a continuous coating
with minimal
show
Thickness measurements
thickness variation.
that each application resulted in similar thickness increments.
5.2.3.2
Sorbic acid determinations
determinations
Sorbic acid
sorbic acid
amount of
deposited on
showed
each individual
Therefore data were normalized for
shown in Table 23.
in the
variation
sample
analysis.
amount
Sorbic acid core concentrations were divided by the total
deposited on each individual piece.
then
plotted
as
a
function
of
as
These normalized values were
time
sampling
and
surface
treatment.
As shown in Figure 32 zein coatings reduced sorbic acid
core concentrations
significantly.
This
represents
the
rate
reduction in the process of diffusion of sorbic acid deposited on
the surface
of these
samples into
barrier properties of zein films.
the number of zein spray
core, due
to
Moreover, the thickness,
applications had also a strong
A more quantitative analysis will
paragraph.
the food
be presented in the
the
i.e.
effect.
following
-
116
Figure 31
NOMARSKY MICROSCOPY STUDIES
12.0 t 2.3 microns
a. Uncoated control (OX)
b. One zein spray application (X)
"We4
I,
26.5 1 0.9 microns
c. Two zein spray applications (2X)
38.0 1 1.8 microns
d. Three zein spray applications (3X)
-
117
-
Table 23
SORBIC ACID DEPOSITED ON EACH INDIVIDUAL SAMPLE PIECE
Time, h
2
18
44
68
118
168
228
Pooled samples [1,3]
mg
Recovery controls [2]
mg
18.9 + 6.8 (n=5)
22.6
17.7
15.4
20.0
21.5
14.8
+ 5.3
+ 2.1
+ 6.1
+ 3.9
+10.2
; 4.0
13.7
14.9; 16.4
12.9; 15.0
13.6
12.3; 9.1
(n=6)
(n=6)
(n=6)
(n=5)
(n=5)
(n=6)
18.6 + 6.1 (n=39)
[1] Corresponds to uncoated and coated samples
(2] These controls were included to determine if losses
occurred when separating samples into a core and a surface
fraction.
[3] n = number of samples
Apparent diffusion coefficients for sorbic acid in
the
food model and zein coating were evaluated using data from Figure
32.
10-6,
=
As shown in Appendix H, the average values were (1.0 + 0.1 x
n = 6),
(3.3 + 0.7 x 10- 9 , n = 5) and (6.8 + 0.9 x 10~9,
n
6) cm 2 /sec for samples OX, 1X and 3X.
The value obtained for the uncoated IMF model, 1 x 10-6
cm 2 /sec, agrees
well with
the one
obtained by
Guilbert et
al
-
118
-
Figure 32
SORBIC ACID DISTRIBUTION STUDIES
Effect of coatings on normalized core concentrations defined as
core concentrations divided by total amount deposited on each
individual piece. Experiment was ran in duplicates, lines have
been drawn through average values.
25
20
15
10
0
20
40
60
80
100
120
140
160
HOURS
-
(1983).
119
-
Their value, determined in an IM agar model at the
same
aw, 0.88, was 2.0 x 10-6 cm 2 /sec.
The calculated Df/Dc
values of
compared with
ratios were
respect to
1,000 with
The surface
and 150
(3X),
cellulose
film
The difference may be
measured with the permeability cell.
to several factors.
300 (1X)
of the
IMF model
topographical features such as pores, valleys, etc.
it
more
difficult
to
cover
assumptions made to analyze
than
cellulose.
due
is rich
which
in
makes
the
Second,
diffusion
Figure 32, unidimensional
and core concentration measured representing the situation in the
piece center, introduce
error leading
to lower
The
D values.
values are however quite comparable given the orders of magnitude
of diffusion reduction with respect to cellulose.
5.2.4 Microbial challenge tests
5.2.4.1
Background microbial load
Background microbial
treatments have been
load
cores were free
did
after approximately a
week storage
for
(2 to 3 weeks or more)
for
The difference
seems to
of contaminants and
indicate that
that surface
of coated
samples
had enhanced
sample
contamination
occurred during sample preparation and/or sample removal.
the surface
surface
Cell counts
uncoated samples and only much later
coated samples.
various
Table 24.
summarized in
reach detectable counts
tests for
growth
Since
inhibiting
-
120
-
Table 24
MICROBIAL BACKGROUND LOAD TESTS
NUMBER OF DAYS BEFORE VARIOUS NOT-INOCULATED
SAMPLES REACHED DETECTABLE COUNTS [1]
days
Treatment [2]
Uncoated
6.5 (or less?)
Uncoated
w/.2% sorbic acid added as dip
6.5 (or more)
Coated w/ 3 20% zein solution dips
w/.2% sorbic acid added as spray
16
coated w/ 3 12% zein solution dips
w/.2% sorbic acid added as spray
16
(1] About 1,000 cells/cm 2
[2] All samples were stabilized with 0.1% K-sorbate. Some
samples received a surface application equivalent to an
additional 0.2%. Bulk aw was 0.88 and storage conditions
were 88% RH and 30C.
properties, it is to be expected that they will have lower counts
for longer periods of time.
5.2.4.2
Location of microbial growth
Coated and uncoated inoculated samples that had shown
3 to 4 log cycles increases above inoculation level were used
to
-
-
121
Table 25
LOCATION OF GROWTH TESTS
COMPARISON OF SURFACE AND CORE FRACTION COUNTS [1]
Treatment
Surface
counts
x 10 7
uncoated
6.7
2.6
.40
uncoated
w/ .2% sorbic acid as dip
5.6
0.14
.03
Core
counts
x 10 5
% total
counts
coated w/8-12-20% zein as dips
w/ .2% sorbic acid as dip
24.0
140
5.80
coated w/3 12% zein as dips
w/ .2% sorbic acid as dip
16.0
25
1.60
(1] samples taken after 11 days storage at 30C and 88% RH.
Samples were peeled to separate a surface and a core
fraction, a slightly more difficult procedure for coated
samples. All samples contained 0.1% K-sorbate. Some
samples received a surface application equivalent to an
additional 0.2%.
determine if growth
samples.
limited
detected
artifact.
was occurring
only on the
surface of
food
Table 25 shows that it is safe to assume that growth is
to
sample
in
the
surfaces.
core,
but
Significant
they
represent
cell
an
numbers
were
experimental
It is very difficult to separate a sterile core from a
-
heavily contaminated surface
122
-
without cross-contamination.
This
operation was slightly more difficult for coated samples.
5.2.4.3
a.
Determination of coating effectiveness
Tests at high relative humidity
with bulk aw = 0.88
These tests correspond to samples
stored at
between different
follows:
RH.
constant 88%
30C and
To
allow for
stability limit
treatments a
and
comparisons
was defined
samples are no longer considered stable if cell
as
counts
are one cycle above inoculation level.
At this
numbers, reported
point it
is important
as time
0 inoculation
obtained
levels, were
treatments, those that
bactericidal effect, are lower
cell
That is why, inoculation levels
about 4 hours after inoculation.
for effective surface
that
to mention
initial
even show
than those obtained for
positive
That is also why we have chosen as initial inoculation
controls.
values those measured for positive controls.
Figure 33 shows the effectiveness of samples coated
dipping in 20% zein solutions.
as being
between 5
and 8
by
Stability period can be estimated
days.
There
is also
a
significant
bactericidal effect during the initial storage period.
Figure 34 shows the
by spraying
them 3
times with
stability tests of samples
12% zein
solutions.
period has increased to about 10 to 16 days.
coated
Stability
-
123
-
Figure 33
STABILITY TEST:
ZEIN DIPPED COATED SAMPLES, BULK aw = .88,
CHALLENGED WITH S.auteuA (CONSTANT RH)
log N, cells/cm
8
7
6
.IMIT
5
4
W/.2%SORBIC ACID
ADDED AS SPRAY
* UNCOATED
3
U COATED W/3 20% ZEIN DIPS
2 4 6 8 10 12 14 16
days
30*C, 88%RH
Samples were taken in duplicates or triplicates. Empty symbols
are used to indicate whenever no counts were detected, <1,000
counts/cm 2 . Lines have been drawn through maximum and minimum
values. The two parallel lines represent the stability limit
defined in the text.
-
124
-
Figure 34
STABILITY TEST:
ZEIN SPRAYED COATED SAMPLES, BULK aw = 0.88,
CHALLENGED WITH S.auteus (CONSTANT RH)
W/.2% SORBIC ACID
ADDED AS SPRAY
0 UNCOATED
E COATED W/ 312%
tog
N'ZEIN SPR AY S
8cellsktm2
7
6
5
4
g
2 4 6 8 10 12 14 16 days
30*C, 88%RH
Samples were taken in duplicates or triplicates. Empty symbols
are used go indicate whenever no counts were detected, <1,000
counts/cm . Lines have been drawn through maximum and minimum
values. As indicated in Figure 33, the two parallel lines
represent the stability limit defined in the text.
-
b.
125
-
Tests with cycles of low and high relative humidity
bulk aw = 0.85 stored
at
hours at 85% RH and 12 hours
at
These tests correspond to samples with
30C and exposed to cycles of 12
Figure 35 shows the effectiveness of samples coated with
88% RH.
20% zein dips, a repetition of preparation conditions reported in
The
Figure 33.
only
difference was
that
conditions
storage
The change in bulk aw and storage RH
tested were less demanding.
stability period:
conditions resulted in an increased
10 to
15
days.
Figure 36 shows the
with 12% zein
stored under
sprayed
stability test of samples
the less
In
demanding conditions.
this case the stability period seems to be longer than, or
about
the length of the testing period, 28 days.
5.2.5
Conclusions
stability
Improved microbial
coatings
were
applied
to
the
IMF
consistent with predictions based
was achieved
model.
This
on diffusion cell
when
zein
result
is
experiments
and sorbic acid distribution studies.
At this point
microbial
response
predicted with
to
accuracy.
important to
it is
the
An
surface
attempt
emphasize that
the
not
be
environment
to describe
can
the
surface
environment has been schematized in Figure 37, where we show
highly dynamic
conditions faced
by microorganisms.
We have
the
a
-
126
-
Figure 35
STABILITY TEST:
ZEIN DIPPED COATED SAMPLES, BULK aw = 0.88,
CHALLENGED WITH S.auteus (RH CYCLES)
Samples were taken in duplicates or triplicates. Empty symbols
are used to indicate whenever no counts were detected, <1,000
counts/cm 2 . Lines have been drawn through maximum and minimum
values. As indicated in Figure 33 the two parallel lines
represent the stability limit defined in the text.
-
127
Figure 36
STABILITY TEST:
ZEIN SPRAYED COATED SAMPLES, BULK aw = 0.88,
CHALLENGED WITH S.a teu6
(RH CYCLES)
W/.2% SORBIC ACID
ADDED AS SPRAY
* UNCOATED
I COAT ED W/ 3 12%.
ZEIN SPRAYS
j og N'
30*C, 12h at 85 1 12 h at 88%RH
Samples were taken in duplicates or triplicates. Empty symbols
are used go indicate whenever no counts were detected, <1,000
Lines have been drawn through maximum and minimum
counts/cm .
vplues. *As indicated in Figure 33 the two parallel lines
represent the stability limit defined in the text.
-
128
-
Figure 3
SCHEMATIC REPRESENTATION OF THE HIGHLY DYNAMIC AND
HETEROGENEOUS CONDITIONS ON THE SURFACE OF A COATED FOOD
various cell types
how injured is this one?
preservative concentration
gradient
localized water condensation
moisture diffusing into food
cell growth due to favorable
transient aw condition
NUTRIENT DIFFUSION
PRESERVATIVE DIFFUSION
MOISTURE DIFFUSION
-
condensation.
Therefore, a
Moreover,
the
stability.
effective coating
had
to
The
be the
final test
of a
not
potentially
with
challenge
microbial
their
model could
precise mass transfer
predict microbial
physiological
depend on the history of
status of the microorganism will
locations.
an
diffusing into the food after
food bulk to surface and water
localized
from
diffusing
nutrients
gradient,
preservative concentration
event of
-
129
a
microorganism of interest.
Further analysis of these
microbial tests allowed
the
following statements:
a. Spraying of zein solutions on food samples is a more effective
treatment than dipping.
The
exact nature of this
difference
between
individual
was not determined.
b. Microbial test
values.
showed
The cause
large differences
behind this variation
variability in the amount of
has been traced
sorbic acid deposited on
to
coated
surfaces as shown in Appendix G.
c. Samples with bulk aw = 0.88
88% RH
and 30C
and pH = 6.4, stored at
represent extreme
constant
testing conditions.
Such
conditions were chosen to accelerate the obtention of results.
Samples with lower
low and
high RH,
bulk aw, 0.85,
85 and
88% RH,
challenged with cycles
cycles of
significantly longer stability periods.
closer to
situations
found
in
12 hours,
of
gave
This cycling test
is
distribution
of
commercial
-
IMF's.
from
Unfortunately, no
these
conditions.
results,
130
-
model is
the
available to
consequences
of
extrapolate
other
abuse
Only qualitative comments are possible.
d. The severeness of the above test should not be underestimated.
challenge
Surface
hurdle was
aw,
of
and
pH
support outgrowth of S.auteuA.
temperature were capable to
The only
terms
in
conditions,
the high
concentration of
sorbic
acid
existing on the surface of treated samples.
5.3
The reduced surface microenvironment pH approach
5.3.1
Dialysis experiments
Dialysis experiments yielded protein fractions with the
equation (32)
samples
with
26.
in Table
characteristics summarized
we
predicted that
a
food
we
surface/food
data
and
coated
IMF
difference
of
Using this
could prepare
pH
bulk
approximately 0.5 pH units.
Dialyzed protein solutions were steamed for 15
and then freeze dried.
Three
IMF sample runs were prepared
labelled according to the ISP batch source, i.e.
5.3.2
minutes
and
A, B and C.
Microbial challenge tests
Using ISP from
subjected to
batch A,
microbial challenge.
IMF No.
2
As shown
equilibration was not achieve instantaneously.
was prepared
in Figure
38,
and
pH
About 3 to 4 days
-
131
-
Table 26
ELECTROLYTE ELIMINATION BY DIALYSIS
FINAL PERMEATE CONDUCTIVITY MEASUREMENTS
Protein
Conductivity, micromho
Equivalent NaCl
x 10-3
30 - 40
20
18
ISP, batch A [1,2]
batch B [3]
batch C [4]
Na-caseinate [5]
Ca-caseinate [5]
3 - 4
2.0
1.8
2
2
0.2
0.2
[1] Each batch consisted of a 7 g/1 (100 g total protein)
solution dialyzed against 20 1 deionized distilled water,
which was changed as frequently as required.
[2] Protein solution as is dialyzed in a cold room (about
7 days).
[3] Protein solution adjusted to pH 8.0, dialysis in a cold room,
pH neutralization by dialysis against 0.0001 M HCl (about
4 days).
[4] Dialysis at 60C (about 12 hours).
(5] Each batch consisted of a 7 g/1 (50 g total protein) solution
dialyzed against 6 1 deionized distilled water, which was
changed as frequently as required.
were
needed
to
reach
a
stable
difference.
difference was 0.3 to 0.5 pH units, i.e.
value:
0.5.
The
final
pH
close to the calculated
-
132
-
Figure 38
MEASUREMENT OF THE pH DIFFERENTIAL ESTABLISHED ON IMF MODEL
No.2 AND ESTIMATION OF ITS EFFECT ON THE SURFACE AVAILABILITY
OF UNDISSOCIATED SORBIC ACID
[ S , A ] are ApH values measured as described in the text.
[ 0 , A ] are calculated % undissociated sorbic acid at
surface pH conditions. This value should be compared to the
pH conditions [ ---
calculated for the bulk
4
].
As
the
one
indicated,
samples were inoculated four days after coating application.
pH
Experiment was ran twice in duplicates or triplicates.
gave
which
electrode,
pH
surface
a
with
done
measurements were
ApH with a maximum error ± 0.1 pH units.
28
pH food: 6.1
BATCH A
0.6
0
200
a
0.0
-12
L
SURFACe'i
0
0
A0.2
Z
----------------------------------INoC
2
4
FOOD
BULK*
4UL
LATION
6
a
10
12
14
16
DAYS
4
-
In
38
Figure
133
have
we
The difference between these
reduced pH surface conditions.
reducing surface
surface
undissociated, i.e.
S.auteus
with
the
the
doubled
than
by
on
established
conditions
pH
two
achieved
effect
stabilizing
the
the
of
concentration
the active form of the preservative.
in
shown
results
with
inoculated
were
samples
equilibrium,
ApH
After
The
pH.
more
have
strong
the
of
localized
and
to bulk
%
the
represented
also
acid corresponding
undissociated sorbic
curves visualizes
-
Inoculated
39.
Figure
uncoated samples were used as positive controls.
ISP from batches B
ApH values shown
microbial tests with the
extended
used for more
and C were
Again
in Figure 40.
samples were inoculated after an equilibration period of about
days.
Based
on
conductivity
the
measurements
similarity
Figure 39 was expected, and confirms the prediction power of
4
to
the
Donnan equilibrium model.
21 days
Growth occurred
days after surface
and fungi
Multiple
The
food bulk.
could
not
i.e.
at day
be
an "equilibration"
the pH
25
bacteria
of unidentified
same time,
samples taken
observation
disappearance caused by
Growth
At the
was observed.
disappeared.
finding.
treatment.
after inoculation,
difference
this
27 confirmed
a
ApH
between surface
and
explained
as
As shown in Figures 38 and 40 it is clear that stable
ApH conditions had already been established.
in section 5.3.3
sorbic acid was
Moreover, as
stable at testing
shown
conditions.
-
134
-
Figure 39
EFFECT OF REDUCED SURFACE pH
MICROBIAL CHALLENGE STUDIES:
S.auteus counts on uncoated controls and
samples coated with X-carrageenan
7
5
0e
4
2
4
S
5
10
12
DAYS
Lines have been drawn
Experiment was ran twice in duplicates.
values
and
should
be compared with
through maximum and minimum
two uncoated controls.
-
135
-
Figure 40
MEASUREMENT OF THE pH DIFFERENCE ESTABLISHED ON IMF MODEL
No.
2 AND ESTIMATION OF ITS EFFECT ON THE SURFACE AVAILABILITY
OF UNDISSOCIATED SORBIC ACID, EXTENDED TEST
Experiment was ran in duplicates or triplicates. pH measurements
were done with a surface pH electrode, which gave ApH with a
maximum error £ 0.1 pH units.
0.6
pH food:6.1
0
0 0.
0
BATC
27
21
4
s
12
16
DAYS
-
Most
likely
process.
explanation
the
collected
contamination
136
-
during
unavoidable
in
lies
the
sample
long
microbial
preparation
It should be noted that at testing conditions food bulk
is not stable.
5.3.3
Sorbic acid stability
The possibility
affected by the
that sorbic
acid stability
presence of X-carrageenan
was also
As shown in Figure 41 no major sorbic acid losses were
could
be
considered.
detected.
We should note that the outgrowth of a mixed microbial population
interfered with sorbic acid recovery determinations, as shown
by
samples taken at days 21 and 33.
5.3.4
Conclusions
in these
We have shown
experiments that:
(1) it
possible to establish a pH differential between surface and
bulk.
in the range predicted
Moreover, measured values were
the Donnan theory model;
in improved microbial
treated food
food
by
(2) surface microenvironment pH resulted
stability as
surfaces to
is
shown by
outgrowth of
the resistance
S.auxeus S-6;
and,
sorbic acid was stable under our experimental conditions.
of
(3)
-
137
-
Figure 41
SORBIC ACID STABILITY IN THE PRESENCE OF X-CARRAGEENAN
5
2
Initial
cells/cm .
All samples were inoculated with 10
to
concentration was 0.22% and values reported correspond
individual values obtained in two experimental runs.
100A
aA
4A
SoS
0
0
A
z
A
*
lu
C
*BATCH
B
A BATCH
C
80
8
16
24
32
D AYS
-
6.
-
138
Discussion
6.1 General
The general aim of this thesis was to find solutions to
microbial stability problems associated with unstable temperature
environments normally found in commercialization and consumer use
of IMF products.
affect localized aw
showed that they
We
food
This would explain why shelf life would
surface conditions.
end
unexpectedly due to surface microbial growth.
The stability
studied as a function
untreated
of an
(No.1)
IMF model
(section 5.1) and
of aw, pH
was
preservative
concentration (section 5.2.2).
expected
As
instability.
of product
aw
pH, showing
the
low
than
higher
The specific aw value
Unfortunately
acid.
levels
stability.
and
1%
K-sorbate
resulted
increased effectiveness
pH
products
levels
also
are
in
function
limit was a strong
organoleptically acceptable (Motoki et al.,
Higher
0.85
of
not
sorbic
always
1982).
microbial
increased
Tests at aw = 0.88, showed that somewhere between 0.5
K-sorbate
provided
stability.
reasons, levels above 0.3% are not allowed.
However,
for
safety
-
Therefore we
-
surface conditions
that
approaches
to investigate
decided
to create
would allow us
139
of high
K-sorbate
concentration or reduced pH.
6.2
Increased stability due to increased undissociated sorbic
acid concentration in the surface
Improved stability was achieved by adding more K-sorbate
on
the surface and reducing preservative diffusion into the food, or
by making more of the
pH.
surface
active form available by reducing
Therefore sorbic acid stability studies under our conditions
of use were required.
We showed that, in samples with 0.57 to 4.24% K-sorbate
25% were detected
losses less than about
showed no
detectable
5.2.3.2).
Finally
losses
after
acid
sorbic
about 30
10
days
stability in
X-carrageenan was also investigated.
detected after
mg/cm 2
(section
storage
the
presence
Less than 10%% losses
storage (section
days
with published findings
results were consistent
storage
surface applications of 1
Samples with
(section 5.2.2).
after 40 days
5.3.3).
(Saxby et
of
were
These
al.,
1982; Bolin et al., 1980; Arya, 1980; Heintze, 1974, 1971).
6.3
The reduced preservative diffusion approach
The
required the
use
of
high
development
surface
of a
K-sorbate
coating
acting as
concentrations
a
diffusion
-
barrier.
cell
140
A coating composition was identified thru
experiments.
Zein
films
diffusion constants 1,000 times
film used
as a
reference value
found
to
permeability
show
apparent
smaller than cellulose
dialysis
were
(section 5.2.1).
confirmed the
distribution studies
films.
-
Sorbic
barrier properties
acid
zein
of
We found that sorbic acid apparent diffusion coefficients
150 and 300 times
in zein were between
smaller than in the
x 10~ 9 cm 2 /sec and 1 x
IMF
10-6
itself.
The values were 3.3 to 6.8
cm 2 /sec,
in zein and IMF model respectively (section 5.2.3).
films acting as barrier
was
Zein coated
and
uncoated IMF model samples were challenged with S.auteas S-6
and
The effectiveness of zein
confirmed by extensive
stored
under
conditions.
constant
microbiological tests.
high
and
under
low-high
RH
cycles
As shown in Table 27 zein coated samples were stable
for 28 or more days, while
uncoated models were stable only
for
2-3 days.
As shown in
from our
Figure 37 it
testi'ng conditions
found in commercial
practice.
is difficult to
to other
product abuse
There are
extrapolate
situations
no models, nor
information, to account for the highly heterogeneous and
conditions existing on the surface.
enough
dynamic
-
141
-
Table 27
MICROBIAL STABILITY IMPROVEMENTS AS A FUNCTION
OF VARIOUS SURFACE TREATMENTS
Number of days to reach stability limit defined as the number of
days for a tenfold increase over
initial S.auteus
inoculation
level.
Days
Treatment
a.
Tests w/samples w/bulk aw = 0.88
stored at 30C and 88% RH
uncoated
w.2% sorbic acid added as spray
coated w/3 20% zein as dips
w/.2% sorbic acid as spray
coated w/3 12% zein sprays
w/.2% sorbic acid as spray
b.
2
5 - 8
10 - over 16
Tests with samples w/bulk aw = 0.85
exposed to cycles of 12 hours at
85% RH and next 12 hours at 88% RH
Uncoated
w/.2% sorbic acid as spray
3
coated w/3 20% zein as dips
w/.2% sorbic acid as spray
10 -
15
coated w/3 12% zein sprays
w/.2% sorbic acid as spray
20 -
over 28
-
6.4
142
-
The reduced pH surface microenvironment approach
The use of reduced
development of
feasibility
objective.
surface pH conditions required
the
of predicting
the
model capable
a mathematical
a
for
and
experimental
conditions
This
information
was provided
the use
by
the
of
shown
As
semipermeable membranes.
Donnan equilibrium model for
ApH
desired
in sections 4.3.1 and 4.3.2 the requirements were an IMF with low
total
electrolyte
polyelectrolyte.
a
and
To satisfy these
a
immobilizing
coating
large
about
conditions an IMF with
0.005 M total electrolyte concentration and a coating composed of
deionized
and
X-carrageenan
deionized
agarose
to
used
was
establish a pH differential between surface and bulk in the order
Such a product was found to
of 0.5 pH units (section 5.3.1).
to microbial contamination of the food bulk which was not
pH = 6.1,
under testing conditions:
0.22%
(section 5.3.2).
stability
aw = 0.88,
sorbic acid
by
the
increased
availability
undissociated sorbic acid due to reduced surface pH.
only 4.8% corresponded to
the
value
availability
was
around
increased
stable
=
As indicated in Figures 38 and 40 surface
achieved
was
due
when stability was lost apparently
stable for up to 20 days
be
the active form
12%.
2.5
Therefore
times
without
preservative use in the IMF formulation.
In the bulk
while on the
surface
of
surface
preservative
increasing
total
-
7.
1.
143
-
Summary and conclusions
Changing environment conditions
exposed during
to which
production, storage,
are important
microbial
stability
IMF products
are
distribution and
factors.
An
use,
analysis
We
showed that they result in local surface condensations.
we
concluded that
could
solve
by
problem
this
surface
treatments leading to improved surface stability.
2.
An
IMF
was
model
tested
for
Information on microbial stability as
and K-sorbate
with S.aukeua
3.
Sorbic acid
concentration was
S-6, A.
was
used
as the
conditions
experimental
a function of pH,
collected when
major
zein and
treatments
had
X-carrageenan
an
effect
surface
our
on
growth,
microbial
including
conditions and IMF formulation
use of
of
the acid is stable under
homogeneous
versus
challenged
improver
mode of
application
incorporation),
processing
preservative concentration variations,
(surface
aw
nigeL and A.tonophituz.
It was shown that
stability.
stability.
microbial
variations (IMF No.1 and
coatings).
its
of
None
stability,
2,
these
confirming
findings reported in the literature.
4.
An approach
to
improved
preservative surface
surface stability
concentration maintained
impermeable food coating was developed.
using
by a
a
high
highly
General guidelines,
-
that allowed us to
144
-
select materials to
be tested for
mass
transfer properties, were established.
5.
Permeability tests showed that zein based coatings were
the
selected
for
most promising
and were
coating alternative
coated IMF surface microbial challenges with S.ateu.
6.
Sorbic
acid
confirmed
studies
distribution
films.
Estimated
coefficients ranged
between
3-7
values were 150-300
times smaller than
zein
properties of
for IMF model, 1 x 10-6 cm 2 /sec.
diffusion
apparent
10-9
x
barrier
the
cm 2 /sec.
the value
These
measured
The latter one agreed well
with published work.
7.
Several combinations
of
coating and
preservative
surface
application were challenged with 105 cells/cm 2 under various
conditions.
These tests showed that:
a. Periods of
conditions.
stability
were strong
function
of
testing
No model is available to extrapolate results
from one condition to another.
b. When the IMF model was prepared
with aw = 0.88 and pH
-
6.4, challenged with S.awteus and then stored at 30C, 88%
RH, it showed 5 - 8 days stability for samples coated
by
dipping in 20% zein.
c. When the same test was
repeated on samples sprayed
12% zein we doubled stability to 10 - 16+ days.
with
-
145
-
with aw = 0.85 and pH
d. When the IMF model was prepared
=
cycles
6.4, challenged with S.auteus and then exposed to
of 12 hours at 85% RH and 12 hours at 88% RH, all at 30C,
stability periods were 10
days
20 to 28
to 15 days and
for zein dipped and sprayed samples, respectively.
8.
In these experiments pH reduction
microbial stability.
not permanent
enhanced
surface pH
that reduced
Preliminary tests showed
well with
and correlated
was
microbial
loss of
stability.
9.
permanent pH
conditions
to
create
(>0.5
concentration was
were coated with
An alternative to
pH
equilibrium
Donnan
differences:
semipermeable membranes.
possible
the establishment
was analyzed for
A theoretical model
reduced
significantly
units)
if
reduced below
1% of
the
total
0.005 [M]
a polyelectrolyte:
electrolyte reduction, use
model
for
it
was
surface
pH
that
model showed
This
of
electrolyte
and if
samples
X-carrageenan.
of a
coating
highly impermeable to electrolytes, was not explored.
10.
Experimental tests
confirmed the
Donnan equilibrium model.
prediction power
of
the
Surface pH reduction in the order
of 0.3 to 0.5 pH units were measured experimentally.
Such a
pH reduction increased surface availability of undissociated
sorbic acid 2.5 times.
-
11.
The effectiveness of
thru microbial
146
reduction was
surface pH
with S.aukteuA
challenge tests
extreme testing conditions, aw
%sorbic acid = 0.22,
days.
-
=
0.88, RH =
stability was greater
established
S-6.
Under
88%, T =
or equal to
30C,
20
-
8.
-
147
Suggestions for future work
what work
(a)
exploration of:
relate
important suggestions
that the
We believe
is needed
to
solutions
to improve
found; and, (b) what work is needed to utilize our findings.
8.1
The reduced preservatives diffusion approach
Food
coatings
as
acting
we
Therefore,
applications.
provide interesting
barriers
diffusion
could
find
should
answers to the following questions:
1.
What are the
mechanical properties of
under
zein coatings
use conditions?
should
Although some literature information is available we
find out whether zein coatings can withstand the
mechanical
abuse normal to food distribution operations.
2.
What
is
the
coating
maximum
thickness
is
that
organoleptically acceptable?
Succesful microbial challenge studies here reported utilized
zein coatings only
0.03 to
coatings would extend
should
find
out
what
0.04 mm
thick.
Since
thicker
microbial stability significantly
is
the
maximum
acceptable
to
we
a
consumer.
3.
What would be consumers attitude towards coated product?
If we tell consumers that, product x has been "corn
protein
-
for
coated
better
148
-
would
coatings
thicker
stability",
probably be accepted.
4.
Would it be possible to make zein coatings more acceptable?
Preliminary microbial tests
using sorbic acid
incorporated
Although we did
in zein films were found to be ineffective.
not investigate the reason for this behavior, we can
assume
that a major sorbic acid fraction remained encapsulated, and
microbial growth.
was unavailable to affect
This
suggests
that we could be able to encapsulate flavoring agents making
the coating not only acceptable, but even desirable, from an
organoleptic point of view.
5.
Are there other products that could benefit from our surface
treatment?
The possibility of applying zein coatings to other
be ignored.
should not
fresh poultry
stability of
improving
For example,
be of
would
great
products
microbial
commercial
interest.
8.2
The reduced pH surface microenvironment approach
The
technology
limitation.
difficulty
main
lies
The
in
the
total
low
in
the
electrolyte
total
potential
application
of
this
of
this
concentration
approach
will
realized only if we find answers to the following questions:
be
-
1.
be
products can
What food
-
149
electrolyte
low
produced with
concentrations?
Most IMF products
salts
contain large amounts
that can be
pH
used to
create an IMF
can
reductions
with total
be
not
components
find food
be interesting to
It will
achieved.
surface
therefore
and
other
of NaCl and
electrolyte
concentrations below 0.005 M.
2.
What coating could be used to isolate eletrolytes present in
IMF's?
If it is not possible
to eliminate electrolytes present
foods, an alternative
would be
in
samples with
to precoat
a
barrier impermeable to electrolyte transport.
3.
What is
the
effect
organoleptic
of
surfaces
food
with
reduced pH?
The organoleptic
of
response
differences between surface
with
pH
will have to
be
food
consumers to
and food bulk
determined before continuing further studies.
4.
Are there other products that could benefit from surface
pH
reductions?
As
stated
previously,
applications to IMF's.
be explored.
we
see
no
need
to
restrict
Other food applications should
also
-
150
-
REFERENCES
and Stringer,
Sebaugh, J.L., Marshall, R.T.
Anderson, M.E.,
A method for decreasing sampling variance in
W.C., 1980.
bacteriological analysis of meat surfaces. J.Food Protection
21.
43:
sorbate.
potassium
and
Sorbic
acid
1978a.
Anonymous,
St.
Publication IC/FI-13, Monsanto Industrial Chemicals Co.,
Louis, Mo.
Anonymous, 1978b. Here is the easy economical dependable way...
to double, perhaps triple, the shelf life of your bakery
Industrial
Monsanto
IC/FI-18,
Publication
products.
Chemicals Co., St. Louis, Mo.
Anonymous, 1977a. Potassium sorbate surface treatment for yeast
Monsanto
Publication IC/FI-21,
raised bakery products.
Chemicals Co., St. Louis, Mo.
Anonymous, 1977b. Outlook expands for use of sorbate to preserve
"natural freshness" of foods. Food Processing 46 (11): 78
Technical
Food Preservatives.
Anonymous, 1976b.
Pfizer, Chemicals Division, New York, N.Y.
Information,
Technical
Kelacid, a Kelco algin product.
Anonymous, 1972.
Inc.,
Kelco, Division of Merck & Co.,
Bulletin DB#8.
Chicago, Ill.
Arya, S.S., 1980. Stability of sorbic acid in aqueous solutions.
28 :1246
J.Agric. Food Chem.
Bakker-Arkema, F.W. and Hall, C.W., 1964. Diffusion in alfalfa
wafer treated as semi-inifinte solids. Michigan Quarterly
Bulletin 47 (2):231.
The
Benmergui, E.A., Ferro Fontan, C. and Chirife, J., 1979.
in
of aqueous solutions
prediction of water activity
connection with intermediate moisture foods. I-aw prediction
in single aqueous electrolyte solutions. J.Food Technology
14 :625.
Private
milk proteins.
Hydration of
1979.
Berlin, E.,
communication. Lipid Nutrition Laboratory, USDA, Beltsville,
Md 20705.
Bhatia, B.S. and Mudahar, G.S., 1982. Preparation and
studies on some Intermediate Moisture Vegetables.
Science and Technology (India) 19 (1):40.
storage
J.Food
-
151
-
Transport
Bird, R.B., Stewart, W.E. and Lightfoot, E.N., 1960.
N.Y.
York,
New
Inc.
Sons
and
Wiley
John
Phenomena.
Blocher, J.C., Busta, F.F. and Sofos, J.N., 1982. Influence of
potassium sorbate and pH on ten strains of Type A and B
J.Food Science 47 (6):2028.
CtoAt..idium botatinum.
Bolin, H.R., King, A.D. and Stafford, A.E., 1980. Sorbic acid
loss from high moisture prunes. J.Food Science 45 :1434.
Boquet R., Chirife, J., and Iglesias, H.A., 1978. Equations for
fitting water sorption
isotherms of
foods.
Part
2:
models.
J.Food
of
various
two-parameter
Evaluation
Technology, 13 :319.
Bowes, A.de P.,
Church, C.F.C., 1975.
Food Values of
Commonly Used. J.B. Lipincott, Philadelphia, Pa.
Chirife, J., 1978. Prediction of water activity in
moisture foods. J.Food Technology 13 :417.
Portions
intermediate
Chirife, J., Boquet, R., Iglesias, H.A., 1979. The mathematical
description of water sorption isotherms of foods in the high
range of water activity. Lebensm. -Wiss. und -Technologie
12 :150.
Prediction of water
Chirife, J. and Ferro Fontan, C., 1980.
activity of aqueous solutions in connection with intermediate
aw
investigation of the
experimental
moisture foods:
related
lactate and some
lowering behavior of sodium
compounds. J.Food Science 45 :802
The
Chirife, J., Ferro Fontan, C. and Benmergui, E.A., 1980.
in
of aqueous solutions
prediction of water activity
IV-aw
foods.
moisture
with
intermediate
connection
prediction in aqueous non electrolyte solutions.
J.Food
Technology 15 :59.
and Iglesias, H.A., 1978.
Chirife, J.,
water sorption isotherms of foods.
J.Food Technology 13 :159.
Equations for fitting
Part I - A 'review.
Cooper, R.M., Knight, R.A., Robb, J. and Seiler, D.A., 1968.
Equilibrium relative humidity of cakes. Food Trade Review.
38 :40
The stability of intermediate moisture
Corry, J.E.L., 1976.
foods with respect to Salmonella. In "Intermediate Moisture
Foods," p. 215, Davies, R., Birch, G.G. and Parker, K.J.
(eds.), Applied Science Publishers Ltd., London.
-
152
-
Corry, J.E.L., 1975. The effect of aw on the heat resistance of
bacteria. In "Water Relations of Foods," p.235, Duckworth,
R.B. (ed.), Academic Press, London.
The effect of sugars and polyols on the
Corry, J.E.L., 1974.
J.Applied Bacteriology 37
heat resistance of salmonellae.
:31.
diffusion
1975.
Methods of solution when the
Crank, J.,
In "The Mathematics of
Ch.2.
coefficient is constant.
Oxford University Press,
Diffusion". 2nd Edition. p.11.
Ely House, London.
Shelf-life and quality characteristics
Cunningham, F.E., 1979.
of poultry parts dipped in potassium sorbate. J.Food Science
44 :863
Donnan, F.G., 1934. Die genaue Thermodynamik der Membrangleichgewichte. II. Z.Physikal Chem. Abt.A 168 (516): 24.
und
1911.
Theorie der Membrangleichgewichte
Donnan, F.G.,
Membranopotentiale bei Vorhandesein von nicht Dialysiereden
Physikalisch-Chemisch
zur
Beitrage
ein
Elektrolyten,
Physiologie. Zeitschrift fur Electrochemie 17 :572.
1932.
Die
genaue
Donnan, F.G.
and
Guggenheim, E.A.,
Physical
I.Z.
Thermodynamik der Membrangleichgewichte.
Chem. Abt.A.
162 :346.
The antimicrobial effect of dissociated and
Eklund, T., 1983.
J.Appl.
undissociated sorbic acid at different pH levels.
Bacteriology, 54 :383.
Recent developments in
1982.
Erikson, L.E.,
Moisture Foods. J.Food Protection 45 (5):484.
Intermediate
low
The effect of
1978.
and Upton, M.E.,
Farrel, G.M.
temperature on the growth and survival of Staph4,ococcuz
aujLea4 and SaZmonetla typhimulum when inoculated on to
bacon. J.Food Technology 13 :15
1941. Storage
Fenton, F.C.,
Engineering 22 :185.
of grain
sorghums.
Agricultural
The
and Chirife, J., 1979a.
Ferro Fontan, C., Benmergui, E.A.
in
of aqueous solutions
prediction of water activity
connection with intermediate moisture foods. II - On the
choice of the best aw lowering single strong electrolyte.
J.Food Technology 14 :639
The
and Chirife, J., 1979b.
Ferro Fontan, C., Benmergui, E.A.
in
of aqueous solutions
prediction of water activity
III - aw
connection with intermediate moisture foods.
-
prediction
solutions.
153
-
in multicomponent strong
J.Food Technology 14 :47.
electrolyte
aqueous
Preparation
and Rao, V.N.M., 1979.
Flora, L.F., Beuchat, L.R.
of a shelf stable, intermediate moisture food product from
muscadine grape skins. J.Food Science 44 (3):854.
Freese, E.,
Sheu, C.W.
and Galliers, E., 1973.
Function of
lipohilic acids as antimicrobial food additives. Nature 241
:321
Gill, C.O., 1979.
A review.
J.Appl. Bacteriology 47 :367
Intrinsic
bacteria
in
meat.
Gill, C.O., and Newton, K.G., 1980.
Development of bacterial
spoilage at adipose tissue surfaces of fresh meat.
Appl.
and Env. Microbiology 39 :1076.
Gordon Greer, G., 1982. Mechanism of beef shelf life
by sorbate. J.Food Protection 45 (1):82
extension
Grignon, J. and Scallan, A.M.,
1980. Effect of pH and neutral
salts upon the swelling of cellulose gels. J.Applied Polymer
Science 25 :2829.
Grundke, G.
and
Kuklov, K.,
1980.
Das Kryptoklima
in
Lebensmittelverpackungen. Lebensmittelindustrie 27 (1):13.
1983.
and Cheftel, J.C.
Guilbert S., Giannakopoulos, A.
Diffusivity of sorbic acid in food gels at high
and
intermediate water activities.
Paper presented at
the
International Symposium on the Properties of Water, Beaume,
France, September 11-16.
Hanseman, J.Y.,
Guilbert, S.,
Richard, N. and Cheftel, J.C.
1980. Influence conjointe de l'activite de l'eau et du pH
sur la croissance de S.auxeus dans un aliment canne a
-Wiss. und -Technologie
humidite intermediaire. Lebensm.
13 :269.
Hydration of
Hansen, J.R., 1976.
24 (6):1136.
Food Chem.
soybean
protein.
J.Agric.
1976. Uber die gegenseitige Beeinflussung von
Heintze, V.K.,
Die industrielle ObstSorbinsaure und schwefliger Saure.
und Gemuseverwertung 61 :555
and Hendrick,
Heldman, D.R., Hall, C.W.
equilibrium relationships of dry milk.
:845.
Vapor
1965.
T.I.,
J.Dairy Science 48
-
154
-
4.
Osmotic and Donnan equilibrium. Ch.
Hiemenz, P.C., 1977.
In "Principles of Colloid and Surface Chemistry".
1st.
Edition. p. 125. Marcel Dekker, Inc., New York, N.Y.
Handbook
and Chirife, J.,
1982.
Iglesias, H.A.
Isotherms: Water Sorption Parameters for Food
Components. Academic Press, New York, N.Y.
of
and
Food
Food
Iglesias, H.A. and Chirife, J., 1976a. Prediction of the effect
of temperature on water sorption isotherms of food material.
J.Food Technology 11 :91.
Equilibrium moisture
and Chirife, J., 1976b.
Iglesias, H.A.
content of air dried beef. Dependence on drying temperature.
J.Food Technology 11 :565.
1976c. Monolayer values in
and Chirife, J.,
Iglesias, H.A.
und
dehydrated foods and food components. Lebensm. -Wiss.
-Technologie 9 :107.
On the local isotherm
Iglesias, H.A. and Chirife, J., 1976d.
concept and modes of moisture binding in food products.
24 (1):77
J.Agric. Food Chem.
1976e. A model for describing
Iglesias, H.A. and Chirife, J.,
the water sorption behavior of foods. J.Food Science 41 :984
P.,
Viollaz,
and
J.
Chirife,
H.A.,
Iglesias,
Thermodynamics of water vapor sorption by sugar beet
J.Food Techonology 11 :91.
Iglesias, H.A., Viollaz, P. and Chirife, J. 1979.
for predicting moisture transfer in mixtures
dehydrated foods. J.Food Technology 14 :89
1976.
root.
A technique
of packaged
1971. Changes caused by Microbes
Ingram, M. and Dainty, R.H.,
34 :21.
in spoilage of heat. J.Appl. Bacteriology.
Ivey, F.J., Robach, M.C., 1978. Effect of sorbic acid and sodium
toxin
outgrowth and
Costidium botuLinum
nitrite on
J.Food Science 43
production in canned comminuted pork.
(6):1782.
Jason, A.C., 1965. Effects of fat content on diffusion of
16 :281.
in fish muscle. J.Sc.Fd.Agric.
water
Kosikowski, F.V., 1977. Cheese and fermented milk foods.
Kosikowski and Associates, Brooktondale, N.Y.
F.V.
Growth
1981.
Lahellec, C., Fung, D.Y.C., Cunningham, F.E.
effect of sorbate and selected antioxidants on toxigenic
J.Food Protection 44 (7):531
strains of S.auteus.
-
155
-
Lang, K.W. and Steinberg, M.P. 1980. Calculation of moisture
content of a formulate food system to any given water
activity. J.Food Technology 45 :1228.
Leistner, L. and Rodel, W., 1976. The stability of intermediate
moisture foods
with
respect
to
microorganisms.
In
Birch,
20, Davies, R.,
"Intermediate Moisture Foods," p.
(eds.), Applied Science Publishers
G.G. and Parker, K.J.
Ltd., London.
SauerstoffM.
,
1975.
Die
Lubieniecki-von Schelhorn,
fur
mikrobielle
konzentration als
bestimmender Faktor
Vorgange in Lebensmittel unter besonderer Berucksichtigung
einer sauerstofffreien Verpackung. Verpackungs Rundschau 26
(1):1.
Lynch, D.J., Potter, N.N., 1982. Effects of potassium sorbate on
J.Food
normal flora and on S.auteub added to minced cod.
Protection 45 (9): 824.
Zur Unbestandigkeit
and Sabalitschka T., 1965.
Marx, H.
Sorbinsaure. Seifen, Ole, Fette, Wachse 91 (21):720
der
Mazza, G. and LeMaguer, M., 1978. Water sorption properties of
yellow globe
onion
(Allium cepa
L.).
Can.Inst.Food
11 (4):189
Sci.Technol.
McAdams, W., 1954. Heat Transmission, 3rd Edition,
Book Company, Inc., New York.
McGraw-Hill
Personal communication. Visiting Scientist.
Motoki, M., 1979.
Massachusetts Institute of Technology, Cambridge, MA.
1982.
Development and
Motoki, M.,
Torres, J.A., Karel, M.,
from
stability of intermediate moisture cheese analogs
and
Processing
J.Food
proteins.
soybean
isolated
Preservation 6 (1):41.
film
The effect of
and Rigg, W.I., 1979.
Newton, K.G.
of
life and microbiology
permeability on the storage
vacuum-packed meat. J.Appl. Bacteriology 47 :433.
Preservative
and Brekke, J.E., 1960.
Nury, F.S., Miller, M.W.
effect of some antimicrobial agents on high-moisture dried
fruits. J.Food Technology 14 (2):113.
Park, G.L. and Nelson, D.B., 1981. HPLC analysis of sorbic acid
in citrus fruit. J.Food Science 46 (5):1629.
Controlling the amount of internal aqueous
Pavey, R.L., 1972.
Swift and Co.,
solution in Intermediate moisture Foods.
Contract
Research and Development Center, Oak Brook, Ill.
-
156
-
No. DAAG 17-70-C-0077, Food Laboratory U.
Laboratories, Natick, MA 01760.
S.
Army
Natick
1976. The safety of intermediate
Pawsey, R. and Davies, R.,
In
moisture foods with respect to Staphytococcus anueua.
"Intermediate Moisture Foods," p.182, Davies, R., Birch, G.G.
(eds.), Applied Science Publishers Ltd.
and Parker, K.J.
London.
Development and characterization of fortified
Peil, A., 1982.
polymer coatings for rice. Ph.D.
Thesis.
Massachusetts
Institute of Technology, Cambridge, MA.
Perry, J.H. (ed.), 1963. Chemical Engineer's Handbook.
Edition, McGraw Hill Book Co., Tokyo.
Fourth
Raccach, M., Baker, R.
C., Regenstein, J.
M. and Mulnix, E.
J., 1979. Potential application of microbial antagonism to
extended storage stability of a flesh type food.
J.Food
Science 44 :43.
Rasekh, J.G., Stillings, B.R. and Dubrow, D.L., 1971.
Moisture
adsorption of fish protein concentrate at various relative
humidities and temperatures. J.Food Science 36 :705
Reid, D.S.,
1976.
Water activity concepts in intermediate
p.54,
In "Intermediate Moisture Foods,"
moisture foods.
Davies, R., Birch, G.G. and Parker, K.J. (eds.), Applied
Science Publishers Ltd., London.
Reinhard, L., Radler, F., 1981. Die Wirkung von Sorbinsaure auf
I. Beeinflussung des Wachstums
Sacchauomyces ceAevL&Zae.
Glucoseverwertung.
und
anaeroben
aeroben
sowie
der
172 :278.
Z.Lebensm.Unters.Forsch.
Robach, M.C.,
1979. Extension
broilers, using a potassium
42 :855.
of shelf-life of fresh, whole
sorbate dip. J.Food Protection
Robach, M.C. and Ivey, F.J., 1978. Antimicrobial efficacy of a
J.Food
potassium sorbate dip on freshly processed poultry.
Protection 41 :284.
Robach, M.C., To, E.C., Maydav, S. and Cook, C.F., 1980.
of sorbates on microbiological growth in cooked
products. J.Food Science 45 :638.
Effect
turkey
1966.
Effect
R.M.,
and Stinchfield,
Saravacos, G.D.
temperature and pressure on the sorption of water vapor
freeze-dried food materials. J.Food Science 31 :779.
of
by
Sauer, F., 1977. Control of yeasts and molds with preservatives.
J. Food Technology 31 (2):66.
-
157
-
Sauter, E.A., Jacobs, J.A.,
Parkinson, J.F.
and Ercanbranch,
S.K., 1979. Effect of carcass weight and fat thickness of
lamb carcasses on surface bacteria carcasses on surface
bacteria counts. J.Food Science 44 :1430
Degradation
and Reid, R.G., 1982.
Saxby, M.J., Stephens, M.A.
of sorbic acid in model food systems. Food Chemistry 9 :283
Scallan, A.M. and Grignon, J.,
pulp and paper properties.
1979. The effect of cations on
2 :40.
Svensk papperstidning.
of
food
relations
1957.
Water
Scott, W.J.,
microorganisms. Advances in Food Research 7 :83.
spoilage
Scott, W.J., 1953. Water relations of StaphyZococcus atueus
6 :549.
30C. Austr.J.Biol.Sci.
at
Stiles, M.E. and Ng., L.K., 1979. Fate of pathogens inoculated
onto vacuum-packaged sliced hams to simulate contamination
during packaging. J.Food Protection 42 :464.
Theron, D.P. and Prior, B.A., 1980.
Effect of water activity
growth
and
on
Staphytococcus auteuz
and temperature
thermonuclease production in smoked snoek. J.Food Protection
43 :370.
Potassium sorbate dip as a
To, E.C. and Robach, M.C. 1980.
method of extending shelf life and inhibiting the growth of
whole
in fresh,
StaphyZococcuA auAeus
Salmonella and
broilers. Poultry Science 59 :726.
Water activity and Christian, J.H.B., 1978.
Troller, J.A.
basic concepts. In "Water Activity and Food," p.1, Academic
Press, Inc., London.
Tsai, C.Y., 1980. Note on the effect of reducing agent on
preparation. Cereal Chemistry 57 :288.
zein
1974.
and Chirife, J.,
Vaccarezza, L.M., Lombardi, J.L.
beet
of
sugar
air
drying
Kinetics of moisture movement during
root. J.Food Technology 9 :317.
Water activity and its
Van den Berg, C. and Bruin, S., 1978.
Paper
Theoretical aspects.
estimation in food systems:
presented at the Second International Symposium on Properties
of Water in Relation to Food Quality and Stability, Osaka,
Japan, September 10-16.
White, M.S., 1978. Water vapor diffusion through eastern hemlock
periderm. Wood and Fiber 11 :171
-
-
158
APPENDIX A
An example of surface condensation problems
As an example
condensation problems
of surface
let us
analyze
is loaded from
a
warehouse, T(1) = 65F, into a railroad car in the winter, T(2)
=
individual package when it
what happens to an
35F.
The objective of these calculation is to show that cooling of the
food pieces is
to evaporation.
mostly due
walls will accumulate
The package
moisture condensations that
drops on a few food surface
spots.
inside
will lead
This will lead to
to
localized
conditions of high aw and therefore to microbial outgrowth.
Calculations will be
separately, i.e.
paragraphs.
simplified by treating
steps
a thru
e
all rate
processes
the
following
discussed in
This will allow us to identify which one is the rate
limiting process and
how long growth
promoting conditions
last.
Let us assume the following values:
package dimensions:
5"x3"x2"
net weight/package:
1 lb
food pieces, spheres, r :
0.5"
food pieces/package:
66
product density
:
50 lb/ft 3
will
-
a.
159
-
product conductivity:
0.4 Btu/hr ft F
specific heat:
0.7 Btu/lb F
package heat conductivity:
0.05 Btu/hr ft F
package thickness:
0.08"
air, natural convection, h:
0.8 Btu/hr ft2 F
air, specific heat:
0.25 Btu/lb F
air, conductivity:
0.016 Btu/hr ft F
air, density:
0.081 lb/ft
3
Assume that we have a package filled only with air
air inside = 65F
air outside = 35F = constant
Further, assume that
which would
be the
there is no
worst
convection inside the
case situation.
graphs (McAdams, 1954) it can
package
Gurney-Lurie
Using
be estimated that air
temperature
inside the package will reach 36F in less than a minute.
b. We can now
assume that air
food pieces, is at 35F.
inside the package,
surrounding
Let us now calculate how long it takes a
food particle to cool under down these conditions if there is no
moisture evaporation
at all.
In this
case the
heat
transfer
resistance ratio (surface/bulk) is 12 which allows us to use
the
-
simplified equation
160
described in
-
McAdams (p.
41, 1954).
The
following values were thus obtained:
product - air temperature
time
F
h
25.5
.1
c.
1.0
5.7
2.0
1.2
Using a heat balance we calculated the amount of water needed
individual food piece =
to cool down an
9.24 g/package).
This
amount
(i.e
0.14 g/food piece
is large
to
cause
significant
localized aw increases but small enough to be provided by only
thin surface layer.
It should be
noted that while
a
evaporation
will come from the surface of all food pieces, condensation
will
occur only on a few localized spots.
d.
Following
al., 1960)
calculations similar
the instantaneous
to example
initial
estimated as 0.72
mg water/sec.
0.36 mg water/sec
we estimated
rate of
Assuming an
that it would
21.2-1 (Bird
evaporation
average equal
take the
et
was
to
surface
only about 6 minutes to evaporate all the water needed to provide
this amount
and therefore
the necessary
cooling effect.
This
value should be compared to the one obtained in part b, 2 hours.
-
e.
the results
Based on
161
-
c and d
simplify the
we can
cooling
analysis by assuming that due to evaporation the surface will
at 35F.
food
We
piece
to
cool
again
Using
down.
whole
will take the
how long it
can now calculate
graphs
Gurney-Lurie
(McAdams, 1954) we estimated that under this assumption the
stage
an underestimation caused by assuming that every preceding
however, that
It shows
food
This is obviously
piece core will reach 36F in about 10 minutes.
occurred instantaneously.
be
cooling
and
therefore surface evaporation will be relatively fast.
f.
Finally we
a
food
surface at
aw
leading
to
drop could
keep
outgrowth.
Disappearance of the water
processes:
diffusion into the food, spreading over the
evaporation to the headspace, etc.
water
that a
the time length
tried to estimate
microbial
drop will be due to
many
surface,
The simplest one to calculate
the
only
mechanism and that we were dealing with 0.2 ml drop, an IMF
with
was the
aw = 0.80
Assuming
process.
diffusion
and microorganisms capable
that
this was
to grow down
to 0.85,
could
estimated that disappearance of growth allowing conditions
take up
to
50
Inclusion of the
other rate
order of magnitude.
is
This
hours.
obviously
an
we
overestimation.
processes should reduce
it by
However, even a few hours will have
one
serious
microbial stability consequences.
Summarizing
we
can
conclude
that
heat
processes involved in food cooling are much faster than
reequilibration.
transfer
moisture
Therefore every time that the product goes thru
-
temperature cycles
released in an
162
large amounts
-
of
analogous process to
surface moisture
will
evaporative cooling.
will lead to significant moisture condensations and therefore
microbial stability difficulties.
be
This
to
-
163
-
APPENDIX B
Measured and predicted equilibrium moisture content comparison
[1]
g
Mi
Mc
error
[2]
[3]
(4]
22.1
26%
52.1
56%
93.1
40%
79.0
19%
69.2
50%
47.5
55%
Product
[1]
Composition
1
IM meat
aw = 0.83
meat solids
meat fat
Frodex, 42DE
water
59.9
10.6
14.5
15.0
23.6
meat solids
meat fat
glycerol
NaC1
water
51.3
9.1
25.0
2.6
25.0
23.6
meat solids
meat fat
glycerol
NaCl
26.6
6.7
25.0
3.7
fish solids
fish fat
glycerol
NaCl
fat
water
22.0
.7
20.1
4.1
13.2
39.9
5
IM deep
fried
aw = 0.83
fish solids
fish fat
glycerol
NaCl
water
20.9
.1
21.5
3.6
31.5
6
IM irish
stew gravy
aw = 0.85
glycerol
corn syrup
sucrose
NaCl
egg yolk
seasoning
fat
water
12.5
5.5
5.0
3.0
10.0
2.1
38.4
23.5
2
IM meat
aw = 0.83
3
IM meat
aw = 0.83
4
IM deep
fried
cat fish
aw =0.83
32.0
142.0
381.0
23.6
142.0
381.0
15.0
142.0
381.0
15.0
142.0
381.0
133.0
36.0
48.0
424.0
16.0
50.0
-
7
IM carrots
aw = 0.77
8
IM carrots
aw = 0.77
9
IM sauce
(ham)
aw = 0.86
10
IM sauce
(chicken)
aw = 0.86
11
IM cheese
analog
aw = 0.86
164
-
carrot solids
glycerol
NaCl
prop. glycol
water
6.3
51.1
2.1
.9
39.3
29.0
78.0
295.0
96.0
carrot solids
glycerol
NaC1
prop. glycol
water
34.2
51.1
2.1
.7
27.5
28.0
78.0
295.0
92.0
n/fat dry milk
glycerol
egg yolk
corn syrup
prop. glycol
NaCl
others
fat
water
12.8
9.8
10.0
2.1
1.0
1.0
1.3
47.7
14.2
30.0
116.0
16.0
37.0
163.0
449.0
50.0
n/fat dry milk
glycerol
egg yolk
corn syrup
NaCl
others
fat
water
12.8
11.8
10.0
2.0
1.0
1.2
47.2
14.1
30.0
116.0
16.0
37.0
449.0
50.0
isol.soy prot. 26.1
5.8
Na caseinate
Ca caseinate
2.0
oil phase
34.4
4.8
NaCl
glycerol
5.9
sorbitol
19.3
others
1.7
water
34.0
29.0
29.0
80.3
24%
56.7
59%
28.5
12%
29.6
80%
449.0
116.0
58.0
50.0
52.4
[1] Product and compositions reported by Chirife (1978) with the
exception of No. 11, the model system used in this study.
Fat was estimated using food composition tables (Bowes and
Church, 1975).
[2) Equilibrium content for each ingredient was obtained from
published isotherms (Iglesias and Chirife, 1976b; Chirife et
al., 1980; Troller and Christian, 1978; Benmergui et al.,
1979; Rasekh et al., 1971; Copper et al.,1968; IglesTis ~nd
Chirife, 1976~3; Heldman et al., 1965; Hansen, 1976; Berlin,
-
165
-
1979).
[3] Calculated moisture content (Lang and Steinberg, 1980):
(b-1)
Mc = 1wi Mi/I Mi
[4] % error was defined as:
(Mc - Mm )/Mm
where
MrP
=
100 = %error
measured
composition data.
moisture
(b-2)
content
calculated
from
-
166
-
APPENDIX C
Mass transfer estimations
Approximations for required D-values
can be based
con
the following equation (Crank, 1975, equation 2-7):
-- exp(-x 2 / 4Dt)
C(x) =
(c-l)
(rrDt)2
Since we are interested in surface concentrations, we can
assume
small values for x, therefore:
exp(-x 2 /4Dt)
x 2 /4Dt
(c-2)
x 2 /4Dt)
(c-3)
l 1
which gives us:
C(x)
An average
(TDt)12
=
(1
'surface'
-
concentration
can
be
calculated
follows:
C(O,Ax) = average between x=O and x=Ax
C(x=O)
= 1/2 ( C(x=O) + C(x=Ax)
(c-4)
= M/(TDt)1/2
(c-5)
as
as
-
167
-
Therefore:
1
C(0, Ax)
=
M
~
(7TDt)
2
(c-6)
1/2
The average value at time 0 is given by the following expression:
Co(0,Ax) = M/Ax
(c-7)
Let us define reduction in average surface concentration as:
f = C(0,Ax)/C(O,Ax)
(c-8)
AX
=~~ t 1 / 2
AX
(2--
(c-9)
)
4Dt
(tDt)
This equation can be
2
simplified to yield:
Ax 2
D=
(c-10)
nt
Assuming that
Ax is
the
coating thickness
required D-values for a given
h we
can
estimate
desired stability period with
results shown in Table 7 (f = 0.05).
the
-
168
-
APPENDIX D
1.
Use of permeability values to estimate coating effectiveness
Under well stirred conditions, the resistance to mass transfer in
a permeability
constant
cell will
concentration
be
due only
difference
reservoirs facilitates the analysis
to film
between
A
properties.
diffusion
cell
of experimental data.
This
was done by using
cl(t=O) = concentration in reservoir 1 = 0
c2(t=0) = concentration in reservoir 2 = large
The use of
large reservoirs
compared to
of
the area
mass transfer allowed us to measure a constant permeability
rate
equal to:
N = DA(c' 2 - c'i)/x
(d-1)
where:
N
= flux of K-sorbate
D
= apparent diffusion constant
A
= area of mass transfer
c'i= concentration in the film which is in
equilibrium with bulk solution concentration ci,
equilibrium expressed as
x
=
a diffusion distance
c'i= kci
-
169 -
Assuming constant k we obtain:
N = K A (c2 - ci)
(d-2)
K = kD/x
(d-3)
where:
If we
evaluate this
coated reference material, i.e.
the uncoated
expression for
r = cellulose
and rc =
and
coated
cellulose, we obtain:
Kr = krDr/xr
(d-4)
Krc = krc Drc/xrc
(d-5)
Assuming no interface
resistance between
coating and
and k values independent of testing material, i.e.
krc = kr = kc
(c = coating alone)
(d-6)
We know that:
xrc
xr
Xc
Drc
Dr
Dr
xrc
= Xr + Xc
(d-7)
where:
(d-8)
We can express x as:
xc = f xr
(d-9)
cellulose
-
170
-
Thus:
1 + f
fDr + Dc
Drc
Dc Dr
(d-10)
We are interested in Dc << fDr, thus:
(d-11)
1 + f = fDrc/Dc
From (d-4), (d-5) and (d-9) we obtain:
Krc
_
Kr
Drc
xr
Dr
xrc
Drc
_
(1
(d-12)
+ f)Dr
From (d-11) and (d-12) we obtain:
Dc
f Krc
_
Dr
2.
(d-13)
Kr
Estimation of zein coating effectiveness
The
effectiveness
of
zein
comparing experimental K-values obtained
conditions (10 mg
Krc =
xr =
at the same initial
Ac
was
for zein coated
The values were:
4.7x1g- 2 (mg/hr cm 2 )/(mg/ml)
1.5x10- 4 (mg/hr cm 2 )/(mg/ml)
0.003 cm
by
sorbic acid/ml solution)
uncoated cellulose films.
Kr =
evaluated
films
(Table 12)
(Figure 28)
(product specification)
and
-
xc
171
-
= 0.00088 + 0.00012 cm (SEM determination)
Substituting in equation (d-13)
was 1070 smaller than D (cellulose).
we found that
D(zein)
-
172
-
APPENDIX E
Preparation and composition of coatings tested
a.
Zein films with the following compositions were tested:
zein, x = 8 to 20
glycerol
Myvacet 7-00
ethanol 190 proof
x g
0.25x g
1 g
balance to 100 g
Ethanol, glycerol and Myvacet were heated up to 50-60C.
Zein was then slowly
added under constant agitation.
Cellulose
films were then sprayed or dipped with this solution while
hot
(50C).
b.
Gelatin
hot
melts
with
the
g
Gelatin
following
compositions
tested:
Polyol
g
propylene
glycol
80
Type B, 175 Bloom
20
ethylene
glycol
80
Type B, 175 Bloom
20
propylene
glycol +
Myvacet 7-00
80
Type B, 175 Bloom
20
1
still
were
-
173
-
Polyol and plasticizer were heated to 90C under medium agitation,
and then gelatin was
slowly added to
was mixed for 30 minutes at 90C.
the vortex.
The
solution
Cellulose films wer then coated
and let dry at room temperature for approximately 12 hours.
c.
Caustic amylose coagulated
with ammonium sulfate films
with
the following compositions were also tested:
Code (2]
Starch soln
Amylose water
ml
g
106
106
96[2]
30
HVII-20-40
HVII-20-Sat 30
HVII-20G-Sat 30
Coagulation bath
(NH4 )2 SO4
NaOH soln
NaOH water
ml
g
40% soln
saturated soln
saturated soln
14.4
14.4
14.4
4.8
4.8
4.8
[1]
Amylose types used in preliminary trials: Amylose AVII
(American Maize products, Co.); Hylon V and Hylon VII
(National Starch and Chemicals Corp.). Hylon VII (HVII)
showed the best film forming properties.
[2]
contained 5 g glycerol were added as a plasticizer.
Amylose,
temperature for 1
added.
water
hour.
and
glycerol
While stirring,
Stirring was continued for
were
stirred
at
the NaOH solution
about 3 hours.
The
room
was
solution
was then cloth filtered and let rest for 12 hours for elimination
of air bubbles.
Cellulose was coated with this solution and then
dipped in the coagulation bath.
The
final step was a wash
with
-
50% glycerol.
174
-
They were then let dry for about 4 to 12 hours
at
room temperature.
d.
to
Caustic amylose coagulation by acid salt mixture was used
prepare the films shown in the following table:
Amylose
g
Code
Water
ml
106
Coagulation
bath
NaOH
g
Water
ml
4.8
14.4
S=H 3 PO 4 (15 g)+
Na 2 HPO 4 (30 g)+
H2 0 (55 ml)
HVII-20-S
30
HVII-20G-PS
30
96[1] 4.8
14.4
P=H 2 SO 4 (12 g)+
Na 2 SO4 (30 g) +
H 2 0 (64 ml)
and then S too
HVII-20G-S
30
96[1] 4.8
14.4
S
(1]
5 g glycerol were added as a plasticizer
The preparation
of amylose
coated cellulose
supports
was identical to the one described in part c.
e.
Water soluble amylose esters with low degree of
were also tested.
High amylose
starch may be
substitution
made more
water
soluble by esterification of some of the free hydroxyl groups.
A
limited amount of branching disrupts the linearity of the amylose
molecule producing a derivative which can be easier dispersed
water.
The
conditions
used
in
this
study
to
prepare
in
this
-
as follows.
material were
sulfoxide were
mixed
175 -
g amylose
10
2
for about
g
and 56.7
hours at
room
temperature.
Thereafter 0.6 g of triethyl amine and 0.6 g of acetic
were added.
temperature.
Mixing
The
times with ethanol.
was continued
mixture was
then
for another 2
filtered and
dimethyl
anhydride
hours at
room
washed
three
Supernatant was discarded while solids
vacuum dried (50C, 18 hours).
were
A 10% solution in water was heated
to boiling point and used to form films on the cellulose support.
-
176
-
APPENDIX F
Sorbic acid as percentage of initial concentration.
100
n:2
:2
90
Coo: 0.55% Co: 0.57%
pH: 6.3
s0
to
10
100
:2
n:2
90
Coo: 0.95%
Co:
30
DAYS
-
177
-
100
90
80
10
20
30
DAYS
z
00
80
100
10
20
30
DAYS
-
178
-
APPENDIX G
Sorbic acid surface application
Sorbic
acid
spraying
for
a
surface
application
equivalent to 0.2% sample weight was determined experimentally as
follows.
Several samples were coated with zein and then
sprayed
with 10% sorbic acid in ethanol 190 proof.
Sorbic acid determinations have been represented in Figure G-1 as
a function of spraying time.
The following additional information was used:
sample weight average = 6.2 g
desired % sorbic acid on surface = 0.2
calculated average sorbic acid/sample piece = 12.4 mg
calculated sorbic acid /cm 2 = 0.65
When this
last
value
was used
in
conjunction
with
Figure G-1 an average spraying time of 17 sec was obtained.
Figure
samples:
large
G-1
shows
standard
also
a
deviations
correlation coefficient, r = 0.81.
large
and
variation
a
between
relatively
low
-
179
-
Figure G-1
DETERMINATION OF SPRAYING CONDITIONS
mg sorbic acid/cm2
1.5
1.0
.6
.5
5
10
15
20 25 30
SPRAY TIME, SEC
180
-
-
APPENDIX H
Determination of apparent sorbic acid diffusion coefficients
from experimentally measured sorbic acid distributions
Sorbic acid apparent diffusion coefficients in our
model and
in zein
Figure 32.
1.
coatings
can be
evaluated using
data
from
concentrations
were
in the center of
the
The following assumptions were necessary:
Experimentally determined
average core
assumed to represent the concentration
This is
food piece.
small.
IMF
In our case
when the
valid
it was 1/5 of
core sample
is
the food piece, or
very
about
1g.
2.
Graphs and
equations obtained
for unidimensional
diffusion
for the case of an infinite
slab were assumed valid for
the
analysis of our disk shaped
samples (r = 1.3
cm)
with
diffusion
simplification
occurring
was
from
possible
every
because
cm, h = 1
surface.
edge
effects
This
were
eliminated by cutting food pieces as shown in Figure 20, i.e.
obtaining a core piece shaped as a smaller disk (r = 0.8
cm,
h = 0.5 cm).
Center conditions can
graphs.
a.
(Adams, 1954, p.36).
be evaluated using
Therefore we defined:
Y = an unaccomplished core concentration change.
Gurney-Lurie
-
Y =
181 -
(C - C)/C
with C = core concentration
C = average concentration for total food piece
X = a relative time
b.
X = Dft/rf 2 with
rf = half thickness = 0.5 cm.
t = time, seconds
Df = food apparent sorbic acid diffusion constant, cm2 /sec
m = resistance ratio
c.
M = (Df rf)/(Dc rc)
with
rc
=
coating thickness = 0.0012 cm (1x); 0.0038 cm (3x)
Dc
=
coating apparent sorbic acid diffusion constant, cm2 /sec.
and
The following values were obtained for uncoated food, m=0:
time
h
10
20
40
60
Y
X
0.84
0.61
0.37
0.12
0.15
0.26
0.50
0.95
Df x 106
cm 2 /sec
0.9
0.9
1.1
1.1
1.0
+
0.1
-
182
-
The following values were obtained for 1x coated
food.
The value obtained for Df was used to evaluate X which allowed us
to determine m.
time
h
Y
X
m
Df x 106
cm 2 /sec
60
80
100
120
140
0.46
0.39
0.35
0.31
0.18
0.85
1.14
1.42
1.71
1.99
0.55
0.70
0.75
1.00
0.75
4.4
3.4
3.2
2.4
3.2
3.3 + 0.7
Similarly for samples coated 3 times:
time
h
Y
x
m
Df x 106
cm 2 /sec
60
80
100
120
140
160
0.60
0.47
0.43
0.40
0.32
0.23
0.85
1.14
1.42
1.71
1.99
2.28
1.00
0.95
1.10
1.35
1.30
1.10
7.6
8.0
6.9
5.6
5.8
6.9
6.8 + 0.9