J . FJ Techno/.

(1984) 19,45-63

Osmotic concentration of potato. I. Criteria for the

end-point of the osmosis process
A . LENART* AND J . M . FLINK?

Summary
A study was conducted to determine what conditions define the equilibrium state
between potato and osmosis solution for an osmosis concentration process. It was
shown that at equilibrium, there is an equality of water activity and soluble solids
concentration in the potato and in the osmosis solution. Rinsing the surface of the
potato after osmotic concentration was shown to significantly reduce solids gain and
soluble solids concentration in the potato, thus resulting in a sizeable increase in the
potato water activity.
When water loss, solids gain, change of water activity and economics are con-
sidered, the optimal conditions for an equilibrium osmosis with sucrose would use a
50% solution at a solution/solids ratio of 4. Uptake of solids during sucrose-based
osmosis results in 75% of the soluble solids in the equilibrated potato coming from the
osmosis solution. A comparison of various osmosis solutions at a 60% total solids level
shows that mixed sucrose-salt solutions give a greater decrease of water activity than
the pure sucrose solution, even though the mass transport data are similar, this un-
doubtedly being due to the uptake of salt,
A model has been developed for calculation of osmosis mass transport data and
water activity for osmotic concentration to equilibrium in sucrose solutions for the
concentration range 10-70% and solutionlsolids range of 1-10. The mass transport data
can be calculated with an average error < 4%. Water activity can also be predicted with
good accuracy for the range of parameters normal for osmosis concentration processes.
The proposed model was also able to predict osmosis mass transport data and water
activity data for short, non-equilibrium osmosis times.

Introduction
Osmotic concentration is a water removal process in which cellular materials (such as
fruits or vegetables) are placed in a concentrated solution of soluble solutes. Concen-
tration results primarily from an osmotic water flow, caused by the water and solute
activity gradients across the cell’s semi-permeable membrane. In natural food systems,
there is also some leakage of solute across the semi-permeable membrane, so osmotic
concentration is, in actuality, simultaneous water and solute diffusion processes (Lee &
Salunke, 1968; Dixon, Jen & Paynter, 1976; Hawkes & Flink, 1978; Islam & Flink,
1982). Studies on osmotic concentration of potato have been reported by Garcia &
McFeeley (1978) and Islam & Flink (1982).
Authors’ address: Department for the Technology of Plant Food Products. Royal Veterinary and
Agricultural University, Copenhagen, Denmark.
*Permanent address: Department of Food Technology, Warsaw Agricultural University, Warsaw,
Poland.
?To whom correspondence should he addressed.
46 A. Lenart and J . M. Flink

The majority of work on osmotic concentration has concerned fruits (primarily

apples) treated in concentrated sucrose solutions. In such a system, Dixon et uf. (1976)
and Dixon & Jen (1977) showed that in addition to the increase in total solids concen-
tration, the sucrose content in the apple solids also increased substantially after
osmosis. Hawkes & Flink (1978) showed similar large uptake of solids following
osmosis in sucrose or sucrose-maltodextrin mixtures. Contreras & Smyrl (1981)
studied osmosis concentration of apples using corn syrup solids. They demonstrated
that effective osmosis could be achieved with little uptake of the corn syrup solids
molecules, most likely due to their larger molecular size. Flink (1980) and Islam & Flink
(1982) showed that osmosis with sucrose-salt binary mixtures resulted in significant
solids uptake, the level being higher than that found for sucrose solutions at the same
concentration. Further, Islam & Flink (1982) showed sizeable uptake of solids for
osmosis in either NaCl or KC1 solutions. It seems evident that, in most cases, osmotic
concentration gives an uptake of osmosis solutes into the plant tissue. It has also been
shown that loss of natural solutes during osmosis also occurs. Ponting (1973) notes that
fruit acids will be removed along with the water, even under osmosis conditions, giving
a product of higher sugar content and decreased acid content. Dixon et af. (1976) and
Dixon & Jen (1977) demonstrated that the organic acid content of apple slices fell by
29% during osmosis in 70% sucrose syrup. Lerici, Pepe & Pinnavaia (1977) showed the
time dependence of concentration for various solutes in apple slices during osmosis.
They note that while the total sugar content rises (due to uptake of sucrose), the content
of reducing sugars present naturally in apples falls, as does the malic acid content.
It can be seen that osmotic concentration results in a complex simultaneous dif-
fusion process in which sample and solution change continuously in both composition
and concentration. It has been proposed (Hughes, Chichester & Sterling, 1958; Lee &
Salunke, 1968) that osmosis results from osmotic pressure gradients and that water flow
will continue until water activity is equal in the sample and osmosis solution. Hughes et
al. (1958) studied transfer of solutes between canning syrups (lower solute concen-
trations than normal in osmosis) and peaches. They noted that even though fluid
transfer was essentially completed after 36 hr, continued uptake of solids indicated that
true equilibrium had not been achieved even after a 60 day storage period. Lee &
Salunke (1968) reported that after 20 hr osmosis, their apple slices had not attained the
equilibrium state. While Farkas & Lazar (1969) did not investigate the equilibrium
state, they did develop an empirical relationship giving simple calculations for osmosis
time to achieve a given weight reduction.
The current investigation was undertaken to determine if the equilibrium state for
osmotic concentration was defined in terms of equality of water activity or of equality of
some concentration factors. Besides giving information regarding the mechanism of
osmosis, knowledge of the end-point criteria should allow development of theoretical
models allowing calculation of the sample and solution compositions and water
activities at various stages of the osmosis process.

Materials and methods

Materiuls
Reagent grade sucrose and sodium chloride were used to prepare osmosis solution.
Hansa potatoes were purchased at a local supermarket.
 Osmosis of potato. I . 47

Preparation methods
Osmosis solutions were prepared by blending the desired solute(s) on a w/w basis
with deionized water. For each experiment, an amount of potato cubes adequate for all
test conditions was prepared. Potatoes were peeled and cut into 10mm cubes. As cubes
were cut they were held in water until the entire batch was prepared. Approximately 20
min was required to prepare the cubes. Just prior to starting the osmotic concentration
step, the cubes were removed from the water and their surfaces gently blotted with
absorbent paper. Samples of the blotted cubes were removed for initial solids content
determination. In each osmotic concentration, 100 g of potato cubes were held sub-
merged in the osmosis solution by a plastic screen. A magnetic stirrer provided gentle
circulation of the osmotic solution. The glass vessel was sealed with aluminium foil to
prevent evaporation from the free surface. For each experiment, the variables included
osmosis time, osmosis solute type and concentration, and solution to potato (w/w)
ratio. Osmosis was conducted at room temperature (23°C). Following the osmotic
concentration, the potato was removed from the solution and its surface gently blotted
with paper to remove adhering solution. (In a few experiments, tests were conducted to
determine the influence of a quick rinse of the surface before blotting on calculated
osmosis behaviour.) The osmosed sample was weighed and its total solids content
determined by air drying at 90°C for 24 hr. As the air oven was initially at room
temperature, the initial phase of drying (about 0.5-1 hr) was conducted at a temper-
ature lower than 90°C. The total solids content of the used osmosis solution was also
determined by air drying. In addition, refractometer measurements were conducted as
a quick method for determining the concentration change of the osmosis solution
following osmosis.
The water activity of potato samples and osmotic solution, both before and after
osmosis was determined using a modified version of the McCune, Lang & Steinberg
(1981) Proximity Equilibration Cell (Lenart & Flink, 1983). In the PEC method, filter
paper is equilibrated with the sample. The equilibrium moisture content of the filter
paper is related by a calibration curve to the water activity of the filter paper, which is
equal to the water activity of the sample.
All determinations were conducted in duplicate.

Calculution methods
In accordance with the terminology of Hawkes & Flink (1978), the term ‘Mass
Transport Data’ is used in this study to indicate the overall exchange of solutes and
water between the potato and the osmosis solution. The individual components of mass
transport data are defined below. Mass transport data for the osmotic concentration of
potato were determined from gravimetric measurements by assuming that under the
conditions used, insoluble solids present initially in the potato cubes remain in the
potato cubes throughout the osmotic concentration. It can be noted here that gravi-
metric measurements of solids content of the osmosis solutions following osmosis
agreed very well with refractometer measurements. As the refractometer measures
soluble solids, this agreement of values indicates that the total solids obtained by the
gravimetric measurement are all soluble solids, thus supporting the above hypothesis
that insoluble solids are not transported to the osmosis solution during osmosis.
Solids gain (SG) can be defined as the net uptake of solids by potato on an initial
potato weight basis:
48 A . Lenart and J . M . Flink

and the water loss ( W L ) as the net loss of water from potato on an initial potato weight
basis:

WL= [ ( M p i ) (1 - T S p i )- ( M p f ) (1- TSpf)

Mpi
Normalized solids content ( N S C ) is defined as the ratio of the total solids at the end
of osmosis to the initial total solids:

NSC -
TSPf
1

TSpi ' (3)

Mass transport data of the type given in equations 1-3 can be used for following
the course of osmosis. However, in order to use a diffusional model to calculate
'end-point criteria' data, it is necessary to first determine the soluble solids concen-
tration in the liquid phase of the potato, This soluble solids concentration (SSC) can be
defined as:

In this study, the amount of insoluble solids present in the initial potato was assumed to
be a fixed fraction (76%, Schwimmer & Burr, 1967) of the total initial solids ( h S p =
0.76), and the amount of insoluble solids in the potato sample is assumed to remain
constant throughout osmosis.
Normalized soluble solids content ( N S S C ) is defined as the ratio of the soluble solids
concentration in potato after osmosis to the soluble solids concentration in potato
before osmosis:
NSSC =
SSCf
SSCi '

Solid gain content (SGC) can be defined as the soluble solids uptake as a percentage
of the soluble solids content of potato after osmosis:

A Hewlett-Packard 982SA calculator was used to evaluate regression line slopes and
intercepts and correlation coefficients ( r ) . In addition, the average percent error
(labelled Ave. 96 error) for each regression line was determined using the method
described by Lang & Steinberg (1981).

Results and discussion

Analysis o j t h e approach to equilibrium
In an analysis aimed at determining the end-point criteria of osmotic concentration,
it was necessary to first determine the osmosis time required for ensuring that
equilibrium between potato and osmosis solution had been attained. The first test was
conducted with a sucrose/salt blend (45% sucrose/lS% salt) which in other studies
(Islam & Flink, 1982) had demonstrated good osmosis characteristics. The overall mass
transport data and water activity (a,) values for osmosis at 1,4 and 20 hr with this mixed
system are given in Table 1.
 Osmosis of potato. I. 49
Table 1. Influence of osmosis time on mass transport data and
water activity (a,) for osmotic concentration of potato cubes
(23°C: R = 4: solution = 45/15% suaoselsalt)

Time (hr)

O* 1 4 20 201

Potato
SG 0 10.4 12.4 17.3 13.7
WL 0 45.4 61.2 59.3 57.6
TS 19.9 46.6 63.0 64.2 59.8
NSC 1.0 2.3 3.1 3.2 3.n
SSC 5.6 30.4 47.5 51.6 45.0
SGC 0 68.5 72.1 78.4 74.1
a, 0.991 n.91 0.810 0.800 0.849
Osmosis solution
SSC 62.5 55.0 51.8 52.0 52.0
n, 0.m 0.774 0.780 0.795 0.795

*Initial condition before osmosis.

+Rinsed and blotted.

After a short time of 1 hr, a significant increase of solids gain and water loss can be
noted. Soluble solids concentration in the potato is about 30%, which is much lower
than the soluble solids concentration in osmosis solution (about 55%). The effect of the
above mentioned changes in these parameters is the reduction of a, of potato (from
0.991 to 0.921), but the final a , is still much higher than the a , of the solution (0.774). It
is obvious that the system has not reached equilibrium after I hr osmosis. After 4 hr, the
soluble solids concentration in potato has almost reached the value found for the
solution, and the a , values of potato and solution differ by only 0.03. After 20 hr
osmosis, water loss is the same as after 4 hr, but solids gain has significantly increased. It
can be seen that soluble solids concentrations and water activities in the potato and in
solution are now equal.
While after 4 hr osmosis with this system, equilibrium was apparently not achieved,
it is noted that mass transport data (except for solids gain) are not significantly changed
by increasing the osmosis time from 4 to 20 hr. On this basis, it was decided to
investigate if osmosis solutions of other solutes or concentrations would attain
equilibrium after 4 hr. Results of these experiments are given in Table 2. In deionized
water, equilibration of a , is attained at a level 0.002 higher than in the fresh potato. This
presumably is related to the sizeable water uptake by the potato. After 4 hr osmosis in
40 and 60% sucrose, there are significant increases in solids gain and water loss, but the
soluble solids concentration in potato is still much lower than in the solution and a ,
values are still quite different. In 10% salt solution, 4 hr osmosis results in the equili-
bration of a , for potato and solution. It is noted that the water loss and solids gain levels
are much lower than with the 60% sucrose, this difference between sugar and salt
solutions resulting from differences in the a, lowering behaviour of the solutes. Uptake
of salt and sucrose in equal amounts will result in a stronger a, lowering effect for the
salt. Thus, when the 5% uptake of salt is added to the 7% initial soluble solids
concentration, one obtains a a, value of 0.955. As this a, corresponds to the a, in a
8.8% salt solution, it would appear that the 7% initial soluble solids concentration in
potato is equivalent in a, effectiveness to 3.8% salt (8.8-5%). These results are similar
to the effect noted by Islam & Flink (1982).
4
50 A , Lenart and J . M . Flink

Of the osmosis solutes tested, levels of water loss and solids gain after 4 hr osmosis
are highest for the mixed sucrose/salt systems. While the soluble solids concentrations
and a, values are only slightly different in potato than in the respective solutions, it
must be noted that equilibrium has not been achieved in 4 hr osmosis.
The results shown in Tables 1 and 2 would indicate that equilibrium is not attained
after 4 hr in most systems. While the degree of approach to equilibrium would indicate
that equilibration will not require the full 20 hr period tested, 20 hr (overnight equili-
bration) was chosen as a matter of convenience, as it assured attainment of equilibrium
and was convenient with respect to conduct of the experimental measurements.
Table 2. Influence of osmosis solution composition on mass transport data and water activity (a,) for osmotic
concentration of potato cubes (4 hr; 24°C; R = 4)

Osmosis solution

Initial 40 % 60% 10% 50/10% 45/15%

potato Water sucrose sucrose salt sucr./salt sucr./salt

Potato
SG 0 -1.5 6.7 10.4 3.2 13.2 14. I
WL 0 13.4 40.1 54.6 24.6 61.2 61.0
TS 19.5 16.1 39.3 53.5 28.8 62.9 63.3
NSC 1 .o 0.8 2.0 2.8 1.5 3.2 3.3
SSC 5.5 3.3 22.0 36.7 12.3 48.1 49.1
SGC 0 -32.5 58.9 68.9 40.3 73.9 75.1
*W 0,992 0.994 0.974 0.966 0.955 0.852 0.802
Initial solution
ssc 0 40.1 60.6 10.1 62.4 62.3
flW 0.994 0.960 0.909 0.924 0.766 0.662
Final solution
ssc 0 35.5 52.2 8.8 sr.9 51.4
flW 0.994 0.968 0.950 0.955 0.83Y 0.783

Additional test series were included to evaluate the effect of surface rinsing on mass
transport data and a,. One set of typical results are given in Table 1. It can be seen that
while a final rinse prior to surface blotting does not give a significant decrease in water
loss, rinsing does give a significant decrease of solids gain and soluble solids concen-
tration, which results in a sizeable rise of the a,. This appears to result in a large change
of the potato’s status from that which had been attained at equilibrium. It was observed
that the effect of rinsing on a, was greater for osmosis in mixed sucrose/salt systems
than in pure sucrose solutions. Awareness of this sizeable change in sample
concentrations and a, is especially important, as in many reports on osmotic
concentration, a rinsing step prior to further handling is recommended. While this will
have a positive effect with regard to surface film removal, it must be recognized that it
will also greatly effect sample concentrations and a,. As it is important in this study to
achieve equilibrium conditions, surface rinsing prior to blotting was not used.

Studies at equilibrium conditions

In the following subsections, the effect of a number of parameters on the osmosis
behaviour of potato was investigated for systems which have attained equilibrium
between the potato and the osmosis solution. These samples were osmosed for 20 hr at
room temperature.
 Osmosis of potato. I . 51

Effect of solutionlsample ratio. Mass transport data and water activity data for
varying ratios ( R )of 60% sucrose solution and potato are given in Table 3. The soluble
solids concentration (SSC) for solution and potato and solids gain content (SGC) for
potato are shown in Fig. 1 as a function of solution-potato ratio ( R ) .It is noted that for
R = 1-6, there is a significant rise in SSC. After 20 hr osmosis, the soluble solids
concentration (SSC) is essentially the same for potato and solution for each ratio, which
means that osmotic concentration has gone to equilibrium. Figure 1 shows that solids
gain content (SGC) is approximately 75% at a solution-potato ratio of 2, and that over
the range of ratios 2-10, SGC shows only a small linear increase (about 3%).

Table 3. Influence of solution to potato ratio on mass transport data and water activity (a,) for osmotic
concentration potato cubes (20 hr; 21°C; solution = 60.6% sucrose)

Initial Ratio (solution :potato)

potato
conditions 1 2 4 6 10

Potato
SG 0 13.0 14.2 14.7 15.5 16.9
WL 0 50.4 54.0 58.7 60.4 61.0
TS 21.4 55.0 59 2 64.5 66.9 68.4
NSC 1 .0 2.6 2.8 3.0 3.1 3.2
ssc 6.1 39.2 44.1 49.9 53.1 55.5
SGC 0 71.7 73.5 74.1 75.1 76.7
aw 0.991 f0,001 0.964k 0.001 0.956f 0.001 0,93050.004 0.907 t0.003 0.876 f0.009
Solution
ssc 60.5 34.5 44.9 51.4 54.0 56.8
0w 0.905 f0.002 0.9702 0.001 0.955 f0.000 0.929 +0.W1 0.908 +O.O3 0.882 f0.QO8

While it can be seen in Fig. 2 that solution-potato ratio has little effect on solids
gain, it is noted that water loss ( W L )and total solids show significant increases over the
range 1-4. In Table 3 it can be noted that after 20 hr osmosis, water activity values are
the same for the potato and the final osmosis solution. Furthermore, in Fig. 3 it is noted
that the relationship between water activity and soluble solids concentration are the
same for both potato and the final osmosis solution. The water activity is about 0.96
when the solution-potato ratio is 2. Near this point, it can be seen that the slope of
water activity versus soluble solids concentration makes a large change. At a, values
below 0.96, it seems that a, is roughly linearly related to soluble solids concentration,
with a large decrease in a, resulting from small increases in soluble solids concen-
tration. On the basis of the dependence of water activity with ratio (Table 3), the best
ratio for osmosis would seem to be 10. However, when results of a, and solids gain are
considered together, the best osmotic effect is achieved at ratios 4-6. When all things
are considered, including economics, it would seem likely that a ratio of 4 should be
optimum.
It can be mentioned at this point that the relationship, water activity versus total
solids, can be made equivalent to the traditional water sorption isotherm by converting
the total solids values to moisture content values. When the data in Fig. 3 was
converted, a sorption isotherm was obtained which fit well with the high end of the
water sorption data of Islam & Flink (1982) obtained by humidification of dried Hansa
potato slices.
52 A. Lenart and J . M . Flink

%
0
- "
,-. - SGC
60
50,

10,

I I 1 I I I I I I

1 2 5 10
R
Figure 1. Effect of solution-potato ratio on osmosis 'end-point criteria' data for equilihrium
osmosis in 60% sucrose (00 .potato; x. solution).

10

I I

1 2 5 10
R
Figure 2. Effect of solution-potato ratio on osmosis mass transport data for equilibrium
osmosis in 60% sucrose.
 Osmosis of potato. I . 53

I 1
1 I 1 r 1
0.87 0.90 0.95 0.99
ow

Figure 3. Relationship of water activity on equilibrium solids concentrations for a 60% sucrose
osmosis conducted at varying solution-potato ratios (00 .potato; x, solution).

Effect of solution concentration. The mass transport data and water activity data for
osmosis at various sucrose concentrations (0-70% w/w; ZSs = 0.00-0.70) are given in
Table 4.The soluble solids concentration (SSC) of potato and of osmosis solution are
directly proportional to initial sucrose concentration (ZSs)(Fig. 4). Over the range of
concentration of 40-70%, SSC of potato and osmotic solution are essentially the same
after 20 hr osmosis. For experiments with initial sucrose concentrations of 10 or 20%,
potato after osmosis has less SSC than the osmosis solution, which means that at low

Table 4. Influence of sucrose concentration o n mass transport data and water activity (u,) for osmotic
concentration of potato cubes (20 hr; 23°C; R = 4)

Sucrose solution ('36)

Osmosis Initial
data potato Water 10 20 40 50 60 70

Potato
SG 0 -2.7 -0.2 4.6 13.7 16.3 18.6 15.8
WL 0 17.7 3.6 25.8 40.7 49.6 59.9 67.9
TS 20.3 20.6 21.1 32 .0 47.0 55.3 64.6 73.4
NSC 1.0 1.o 1.0 1.6 2.3 2.7 3.3 3.8
ssc 5.8 3.1 6.2 15.5 32.8 41.9 52.8 61.5
SGC 0 -56.1 -3.8 47.0 72.4 75.7 80.0 77.2
(I ~ 0.991 0.992 0.990 0.984 0.970 0.959 0.928 0.890
1_0.000 +o.ooo ?0.000 tO.OOO +.o.oOo -C0.0o1 io.000 tO.OO1
Initial solution
ssC 0 10.3 20.4 40.2 50.4 60.9 71.1
(1% 0.993 0 ,987 0.981 0.964 0.946 0.899 0.823
Final solution
SSC 0.7 9.7 17 3 34.5 42.8 51.8 59 9
a, 0.993 0.986 0.983 0.971 0.960 0.933 0.895
54 A . Lenart and J . M . Flink

I
1
d
10 20 50 70
4,%
Figure 4. Effect of sucrose concentration on osmosis ‘end-point criteria’ data for equilibrium
osmosis at R = 4 ( 0 ,potato: x. solution).

sucrose concentration, potato and solution have not attained equilibrium. The solids
gain content ( S G C )curve for osmosis in sucrose shows that SGCrapidly increases when
sucrose concentration increases up to the 50% level. Above this concentration, SGC
remains at the same level as that found in the ratio experiments (about 75%). From
results shown in Fig. 5 , it is seen that the water loss ( W L ) and total solids are directly
prcportional to the initial concentration of the sucrose in solution over the range
20-70%. A sucrose solution of 10% does not cause any significant increase in solids
gain, and has little effect on W L from potato. This would indicate that solutions of
about 10% sucrose are isotonic to the potato tissue, a fact noted by Garcia & McFeeley

Figure 5. Effect of sucrose cnncentratinn on osmosis mass transport data tor equilibrium
osrnosisat R = 4.
 Osmosis ofpotato. I . 55

(1978). Solids gain increases proportionally with increasing sucrose concentration up to

a solids gain of 16%. After 20 hr in water (leaching), potato loses about 18% water and
more than 55% of the soluble solids.
Table 4 shows that after 20 hr osmosis in varying concentrations of sucrose, water
activity is the same for potato and solution, while in Fig. 6 it is noted that the
relationship of water activity with soluble solids concentration (SSC) is the same for
potato and the osmosis solution. A water activity of about 0.96 corresponds to an initial
sucrose concentration of 50%. Near this a, level, there is a sizeable change in the slope
of the water activity versus soluble solids concentration curve, the same behaviour as
that which was noted for the influence of ratio. Again, at a, levels below 0.96, a small
increase in soluble solids concentration results in a sizeable decrease in a,. When
osmosis effectiveness is analysed with respect to response of water activity to sucrose
concentration and on water loss and solids gain, the best results are achieved by using
the maximum possible concentration of sugar in solution.

O
O
’
80
i

0.87
I 0.90 0.95
0,

0.99
X
ssc

OW

Figure 6 . Relationship of water activity on equilibrium solids concentrations for osmosis

conducted at varying sucrose concentrations ( 0 ,potato: x, solution).

Conversion of total solids values to moisture contents again allowed preparation of

water sorption isotherms which fit well with the data of Islam & Flink (1982).

Effect of solution composition. Mass transport data and water activity data for
osmosis with solutions of varying composition are given in Table 5. Results show that
sucrose alone and in combination with salt at high concentration (about 60% total
solids) have very similar values for the mass transport data. These solutions all give high
amounts of water loss together with high solids gain, resulting in high normalized solids
content. The sucrose containing solutions all showed about the same values for soluble
solids concentration (SSC) or solids gain content (SGC), with about 79% of the soluble
56 A . Lenart and J . M . Fiink

Table 5. Influence of osmosis solution composition on mass

transport data and water activity (a,) for osmotic concentration
of potato cubes (20 hr: 23°C; R = 4)

Solution (%)

OsmosisInitial 60% 10% .50/10% 45/1.5%

data potato sucrose salt sucr./salt sucr./salt

Potato
SG 0 18.6 5.2 17.1 18.2
WL 0 59.9 20.0 62.3 60.0
TS 19.4 64.6 28.8 66.5 64.5
NSC 1.0 3.3 1.5 3.4 3.3
SSC 5.5 52.8 13.9 54.2 52.5
SGC 0 80.0 52.7 78.6 79.6
a, 0.991 0.928 0.939 0.820 0.789
~0.001 ?O.OOo ?O.oOo -c0.001 ~0.001
Initial solution
ssc 60.9 10.2 61.9 60.7
M 2.2 1.8 4.2 5.2
a, 0.899 0.919 0.770 0.691
Final solution
ssc 51.8 9.2 53.1 5o.x
a, 0.933 0.939 0.827 0.790

solids in potato after osmosis being taken up from the solution. It appears that both
sucrose and sucrose/salt mixtures can be used for osmotic concentration of potato.
Among other things, the choice of solution will depend upon the extent to which the
solute infusion and water loss helps in preservation of the product. At the 60% total
solids level, sucrose/salt mixtures are the most effective osmosis solutions, giving
comparatively low levels of water activity for the potato cubes. Sucrose alone at 60%
solids results in a much higher water activity for the potato cubes. The large decrease in
water activity obtained with the sucrose/salt solutions compared to the same concen-
tration of sucrose alone may be attributed to the higher molar concentration of the
sucrose/salt solution. While the level of SGC in potato is the same for all three
sucrose-containing osmosis solutions, the composition of the solids gained must be
different (this reflected in the fact that salt is present in only two of the three solutions).
It is noteworthy that it is the two sucroselsalt solutions that give the marked lowering of
water activity for the potato. In the mixed systems, sucrose is the higher molecular
weight material which acts to give the high concentration gradients which result in high
water loss, while salt can much more easily penetrate tissue and give a larger decrease of
water activity at the same level of total solids content.
NaCl has been tested alone at 10% solids concentration. The mass transport data
show that this salt solution is not effective for concentration of potato cubes (Table 5).
The large decrease of water activity obtained with salt solution compared to the same
concentration (w/w) of sucrose alone (Table 4) may be attributed to the higher molar
concentration of the salt solution. A 10% salt solution has about the same molar
concentration as a 50% sucrose solution, but gives a larger decrease in water activity
than 50% sucrose. This effect appears to result from the vast difference in molecular
weight, and the resultant difference in the penetration of the potato tissue.
 Osmosis ofpotato. I . 57

During the course of osmosis, potato cubes pick up solute. It has been shown that
after an equilibrium time osmosis, soluble solids concentration (SSC) and water activity
(a,) on the surface of potato cubes are similar to these values for the osmosis solution.
In an experiment at equilibrium conditions (20 hr) in which osmosed cubes were
sectioned into an inner half and an outer half, results with three osmosis solutions
(sucrose, salt and a sucrose/salt mixture) (Table 6) showed that water activity is
essentially constant for the cube surface and at a layer halfway to the centre of the
cubes. The equality of a , values for the whole cubes and the various sections when
osmosed to equilibrium in various osmosis solutions would indicate that there was an
even distribution of solids gain and water loss throughout the cubes.

Table 6. Water activity distribution in potato cubes after osmotic

concentration (20 hr: 23°C; R = 4: raw potato: ISp = 19.37%;
a, = 0.991)

Water activity of potato after osmosis

60% 10% so/1 0 %

sucrose salt sucr./salt

Whole cubes 0.933 0.941 0.827

= 10 rnm on edge
Surface Layer of cubes 0.931 0.939 0.828
= 2 mm thick
Centre of cubes 0.931 0.940 0.825
= 6 mrn on edge

Prediction of the osmosis end-point equilibrium condition

A prediction of the osmosis end-point equilibrium condition has been developed for
the following range of initial process parameters (potato and solution conditions):
R = 1-10;
ZSS = 0.00-0.70;
TSpi = 0.18-0.22;
InSp = 0.76;
Time = 20 hr (time assuring equilibration for 10 mm cubes).
This prediction was developed on the basis of a sucrose osmosis solution.
Beyond the experimentally determinable ‘end-point criteria’ data given earlier in
equations 1-6, it is also possible o n the basis of the initial concentrations in potato and
osmosis solution alone to develop equations for the equilibrium soluble solids concen-
tration, mass of potato after osmosis and potato total solids. If it is assumed that the
equilibrium condition can be defined as the soluble solids concentration of potato
( S S C p ) being equal to the soluble solids concentration of the used osmosis solution
(SSCs),then this equilibrium value, the calculated soluble solids concentration (SSCc)
can be defined

sscc = [ (TSpi) (1-InSp)+(R) (I&)

1+ R (TSpi) (InSp )
-
(7)

When the calculated equilibrium SSCc value (determined exclusively on the basis of
initial experimental conditions) is correlated with the measured equilibrium SSC value
58 A . Lenart and J . M . Flink

(determined from the various experimental conditions investigated), the relationship

shown in Fig. 7 is obtained for the potato cubes. When correlations are tested for
experiments which investigated the influence of ratio alone or initial sucrose concen-
tration alone, the correlation coefficient obtained between SSCc and SSC is greater
than 0.995. The best Ave. % error for these individual correlations was about 1.7%.
When the correlations were conducted with results from both ratio and sucrose concen-
tration experiments as parameters, the correlation coefficients remained high (Table
7). but on the basis of the Ave. % error analysis, the multiple correlation for R = 1-10
and 1% = 0.40-0.70 was chosen as giving the best overall relationship for potato cubes:
SSC = -3.6248+1.0701 (SSCC). (8)
A similar correlation analysis for calculated SSCc and measured SSC for the sucrose
osmosis solution gives a very high correlation coefficient ( I = 0.999) for the complete
range of parameters. and Ave. % error is only about 1.3%.

ssc

50

20

1C

Figure 7. Relationship of measured experimental SSC values and calculated SSCc values based
on initial solution and potato properties (@. influence of ratio; 0. influence of concentration;
A.influence of mixed solute).
 Osmosis ofpotato. I . 59

Table 7. Results of correlation analysis between experimental and calculated potato osmosis data

Solution Potato

Range of sucrose SSC ssc TS MI, “w

solution parametcrs
used in correlation r* Et r E r E r E r E

R = 1-10
I S Y = 0.10-0.70 0.9990 1.27 0.9957 7.37 0.9943 2.81 0.9407 3.73 0.8038 1.1I
R = 1-10
ISs = 0.40-0.70 0.9967 1.00 0.9985 2.49 0.9812 1.59 0.8911 3.12 0.9730: 0.353
~~ ~

* r = correlation coefficient.
t E = average o/c error.
4R = 1-4 and ISs = 0.10-0.70.

If one wants t o calculate the remaining final equilibrium parameters for the osmosis
of potato, it is necessary to know the relationship between total solids of potato and the
soluble solids concentration of potato. Since the initial solids varies between potato
samples, it is necessary to determine the following correlations on the basis of the
normalized solids content ( N S C ) and normalized soluble solids content (NSSC). For
determining the correlation of NSCand NSSC, results for all R and ISs parameters were
used. A very high correlation coefficient ( r = 0.9982) was found for the relationship
NSC = 0.771+0.270 (NSSC), (9)
where NSSCcould vary from 1 to 12. The Ave. % error was less than 1.7%. (It can be
noted as an aside that when NSSCis less than 1 (i.e. leaching in water) the Ave. % error
rise to lo%, and thus this equation is not useful for calculation of leaching behaviour.)
It is now possible to use equation 9 as an intermediary step in calculating final total
solids from the final soluble solids concentration. As the final soluble solids concen-
tration can be calculated in two ways (either equation 7 alone, or equations 7 and 8
together), it is possible to determine two sets of values for the final mass transport data.
Specifically, in the first case, knowing the initial SSC value (fresh potato) allows the
equilibrium SSCc value (equation 7) to be converted to the equilibrium NSSC value
(equation 5 ) . This NSSC value can then be converted via equation 9 to the cor-
responding equilibrium NSC value. Knowing the initial total solids value, this NSC
value can then be converted back to the final total solids at equilibrium (equation 3 ) . In
the second case, equation 8 is used to convert the SSCc value from equation 7 to a
‘corrected’ SSC value which can then be used to calculate the NSSC value. and
thereafter, as above, the final total solids (with ‘correction’).
In calculating the next step-the weight of potato after osmosis-it is necessary to
assume that the amount of the insoluble solids in the potato does not change during the
osmosis. Assuming this, it is then possible to determine the weight of potato after
mmosis ( M p f c ) from:
(TSpi) ( M p i ) (1nSp) (1 -SSCc)
Mpfc = (10)
TSpfc-SSCC
The results of all these mathematical manipulations are shown in Table 7, where the
experimental equilibrium mass transport data are correlated with the calculated values
obtained using equations 7 and 9. It can be seen that for total solids. while there is a high
60 A. Lenart and J . M . Flink

correlation for both ranges of initial sucrose concentration, the Ave. % error is much
lower when the initial sucrose concentration (ZSs) is limited to 0.40-0.70. With respect
to the final weight of potato ( M p f ) , the correlation is less good as the value of Mpfis
very sensitive to total solids value. In both cases tested. however. the Ave. % error is
less than 4%. In Table 8. the influence of calculating SSCc from equation 7. or from
equations 7 and 8 on mass transport data is shown. It is seen that the correlation
coefficients are not different and that the Ave. % errors values are also not very
different,

Table 8. Results of correlation analysis between experimental and calculated potato

osmosis data ( R = 1-10 and I.% = 0.10-0.70)

Correlation coefficient r Ave. o/c error

Potato
osmosis Calculated Calculated Calculated Calculated
data with Eqn* (7) with Eqn (7.8) with Eqn (7) with Eqn (7.8)

TS 0.9943 0.9946 2.81k4.26 I .81?1.75

Mp 0.9407 0.9420 3.73 22.70 3.9723.20
U W 0.8038 0.8053 1 . 1 1 2 1.84 I . 13tl.73

*Eqn: equation

In summarizing the above results, it can be noted that the above model is useful for
calculating mass transport data over the range R = 1-10 and ISs = 0.10-0.70 with an
Ave. % error of less than 4%. When the ISs range was limited to 0.40-0.70, and when
equation 8 was used as a ‘correction’ to SSCc values, the Ave. % error is reduced by
only about 1%. It is felt that satisfactory results are obtained for the mass transport data
when using an ISs range = 0.10-0.70 and by calculating SSCc from equation 7 (e.g.
without ‘correction’). The fact that SSCc can be calculated from equation 7 indicates
again that the equilibrium osmosis state can be defined in terms of equality of soluble
solute concentration in the solution and potato. Through correlation of the final
osmosis parameters and the measured water activity values, it was found that a linear
relationship (r = 0.9935) existed between NSC and log (1-a,):
NSC = 6.1056+2.499Olog ( l - ~ ~ ) , (11)
with an average error of 2.4%. This will give a difference between calculated and
experimental water activity values of less than t0.003.
When evaluating the correlation between measured final potato a , and calculated
final potato a, (from equation ll), it was noted that inclusion of the full R range gave a
poor correlation coefficient (see Table 7, upper line). When the R values were limited to
1-4 (with ZSs still ranging from 0.10 to 0.70), a good correlation was found with a low
Ave. % error (see Table 7, lower line). In Table 8, it is seen again that use of the
equation 8 ‘correction’ for SSCc when calculating a, values does not result in a better
correlation coefficient or a lower Ave. % error.
It has been shown above that by having knowledge of the initial osmosis parameters,
the proposed model can be used to calculate osmosis mass transport data and water
activity for the equilibrium condition with an average error of under 4%. If one limits
the range of initial sucrose concentrations to thoge reasonable for osmotic
concentration (I,% of 0.4-0.7). the average error will fall to about 3% or less.
 Osmosis of potato. I. 61

As in practice it is seldom that osmosis is conducted to equilibrium, an evaluation

was conducted to determine if the above model could also be used to calculate the
osmosis mass transport data and water activity for a time period which is shorter than
the equilibration time. In this case, 4 hr was chosen as it is a time that is often found in
the literature (Islam & Flink, 1982; Hawkes & Flink, 1978).
Table 9. A comparison of experimental and calculated mass transport data and water activitv for osmotic
concentration of potato cubes in sucrose solution (4 hr: 24°C; R = 4: TSpi = 19.52: SSC = 0.7xSSCequil.)

Sucrose concentration in osmosis solution (%)

40 60

Potato Calculated Calculated Calculated Calculated

osmosis with Eqn (9) with Eqn (12) with Eqn (9) with Eqn (12)
data Measured (error %) (error %) Measured (error %) (error ".)

MP 67.0 80.6 (+20.0) 65.6 (-2.0) 56.0 70.6 (+26.0) 56.8 (+1.4)
TS 39.3 37.8 ( -3.8) 41.1 (+4.6) s3.s 49.1 ( -8.2) 52.4 (-2.0)
NSC 1.02 1.94 ( -4.0) 2.11 (+4.5) 2.7.5 2.52 ( -8.4) 2.69 (-2.2)
0 \v 0.974 0.978 ( +0.4) 0.975 (+o. I ) 0.9% 0.963 ( -0.3) 0.957 (-0 9)

Mass transport data for a 4 hr osmosis was calculated from the above model by
noting that after 4 hr, the soluble solids concentration in the potato was 70% of the
soluble solids concentration in potato at equilibrium (Tables 1, 2 and 4). Very high
correlation was found between calculated and experimental water activity values, but
mass transport data showed a sizeable error for the weight of final potato (Table 9).
When it was noted that the correlation for NSC with NSSC differed for 4 hr (non-
equilibrium) and 20 hr (equilibrium) osmosis, a new correlation for 4 hr was developed:
NSC = 0.943+0.270 ( N S S C ) , (12)
to replace equation 9 when calculating non-equilibrium mass transport data. Compar-
ing equations 9 and 12 shows that the slopes are the same, while the intercept is higher
for the 4 hr equation. Mass transport data calculated with equation 12 showed a much
lower error than that calculated with equation 9 (Table 9). The basic for the higher
intercept value for the relationship NSC = f ( N S S C ) is that the soluble solids concen-
tration in the potato is increasing with time, while the insoluble solids content was taken
t o be constant. As NSC includes the effect of both soluble and insoluble solids. while
NSSC includes only the soluble solids, the relative influence of the insoluble solids on
NSC will fall with increasing soluble solids uptake. Thus after a short osmosis time (4
hr), where the solids gain is only about 70% of the equilibrium level (Tables 2 and 4),
the influence of insoluble solids on NSC is greater than at the equilibrium time, and thus
the intercept is higher.
It appears that correcting the NSC versus NSSC equation for short time will give
improved results. It should also be possible to develop an equation to express the
change in this relationship with time for short osmosis times, but further experimental
data are required to investigate this point.

Summary
It has been shown that the equilibrium condition between potato and sucrose osmosis
solution is an equality of water activity and soluble solids concentration in the potato
62 A . Lenart and J . M . Flink

and in the osmosis solution. It was shown that to achieve equilibrium with 10 mm potato
cubes requires between 4 and 20 hr of osmosis. For sucrose and sucrose/salt mixtures,
the water loss after 4 hr is essentially at the equilibrium level, while solids gain is still
increasing at this time. Rinsing the surface of the potato after osmotic concentration
was shown to significantly reduce solids gain and soluble solids concentration in the
potato, this resulting in a sizeable increase in the potato water activity.
When water loss, solids gain, change of water activity and economics are
considered, osmosis to equilibrium in a 50% sucrose solution at a solution/solids ratio of
4 is judged to be optimal. Increasing the solution/\olids ratio above 4 gives a greater
decrease of water activity than increasing the sucrose concentration above 50%.
Uptake of solids during sucrose-based osmosis results in 75% of the soluble solids in the
equilibrated potato coming from the osmosis solution. A comparison of various osmosis
solutions at the 60% total solids level shows that mixed sucrose-salt solutions give a
greater decrease of water activity than pure sucrose solution. even though the mass
transport data are similar. This is undoubtedly due to the uptake of salt from the mixed
solutions.
A model has been developed for calculation of osmosis mass transport data and
water activity for osmotic concentration to equilibrium in sucrose solutions in the range
10-70% and solution/solids range of 1-10. The soluble solids concentration at equilib-
rium can be calculated with an average error of 2.5% in potato and 1% in solution. The
corresponding mass transport data can be calculated with an average error < 4%. The
range of osmosis parameters over which water activity can be calculated with good
accuracy is more limited than those given above, though those which are omitted fall
outside the range for normal mmosis processes. The proposed model was able to
predict osmosis mass transport data and water activity for short, non-equilibrium
osmosis times for sucrose concentrations of 40-7095 and solution/solid ratios of 4-6.

List of symbols
a, -water activity:
ZnSp -insoluble solids (fraction of initial solids weight);
ZSs -initial solids of solution (fraction);
Mpi -initial weight of potato (g);
Mpf -final weight of potato (g);
Mpfc -calculated final weight of potato (8);
NSC -normalized solids content (fraction of initial total solids);
NSSC-normalized soluble solids content (fraction of initial soluble solids);
R -solution to sample ratio (wlw);
SG -solids gain (g solids/100 g initial potato);
SGC -solids gain content (% of final soluble solids);
SSC -soluble solids concentration (g soluble solids/100 g solution in potato);
SSCc -calculated soluble solids concentration (g soluble solids/100 g solution in
pot at o );
T S -total solids of potato (%);
T S p i -initial total solids of potato (fraction);
TSpf -final total solids of potato (fraction);
TSfc -calculated final total solids of potato (fraction):
W L -water loss (g water/100 g initial potato).
 Osmosis of potato. 1. 63

Acknowledgments
T h e authors would like to thank the Governments of Poland and Denmark for their
support of the study visit by AL under the provisions of the Polish-Danish Cultural
Exchange Programme. We would also like to thank the Danish Veterinary and Agri-
cultural Research Council for their support of this project.

References
Contreras, J.E. & Smyrl, T.G. (1981). An evalu- Islam, M.N.& Flink, J.M. (1982). Dehydration of
ation of osmotic concentration of apple rings using potato. 11. Osmotic concentration and its effect on
corn syrup solids solutions. Canadian Institute of air drying behaviour. Journalof Food Technalogy,
Food Science Technology Journal, 14,301-314. 17,387-403.
Dixon, G.M. & Jen, J.J. (1977). Changes of sugars Lang, K.W. & Steinberg, M.P. (1981). Linearization
and acids of osmovac-dried apple slices. Journal of of the water sorption isotherm for homogeneous
Food Science, 42,1126-1127. ingredients over A , 0.30-0.95. Journal of Food
Dixon, G.M., Jen, J.J. & Paynter, V.A. (1976). Science, 46,1450-1452.
Tarty apple slices result from combined osmotic- Lee, C.Y. & Salunke, D.K. (1968). Sucrose pen-
dehydration and vacuum-drying process. Food etration in osmo-freeze dehydrated apple slices.
Product Development, 10,634-637. Current Science, 37,297.
Farkas, D.F. & Lazar, M.E. (1969). Osmotic de- Lenart, A. & Flink, J.M. (1983). An improved
hydration of apple pieces. Food Technology, 23, proximity equilibration cell method for measuring
688-690. water activity of foods. Lebensmitrel-Wissenschafr
Flink, J.M. (1980). Dehydrated carrot slices: In- und Technologie, 16,84-89.
fluence of osmotic concentration on drying Lerici, C.R., Pepe, M. & Pinnavaia, G. (1977). La
behaviour and product quality. In: Food Process disidratazione della drutta mediate osmosi diretta.
Engineering, Vol. I (Ed. by P. Linko et al.). Pp. Industria Conserve, 52, 125-129.
412-418. Academic Press, London. McCune, T.D., Lang, K.W. & Steinberg, M.P.
Garcia, R. & McFeeley, J.C. (1978). Determining (1981). Determination of water activity with the
the cellular concentration and osmotic potential of proximity equilibration cell. Journal of Food
plant tissues. American Biology Teacher, 40,119- Science, 46, 1978-1979.
120. Ponting, J.D. (1973). Osmotic dehydration of fruits.
Hawkes, J. & Flink, J.M. (1978). Osmotic concen- Process Biochemistry, 8, 18-20.
tration of fruit slices prior to freeze dehydration. Schwimmer, S. & Burr, H.K. (1967). Structure and
Journal of Food Processing and Preservation, 2, chemical composition of the potato tuber. In:
265-284. Potato Processing (Ed. by W. F. Talburt & 0.
Hughes, R.E., Chichester, C.O. & Sterling, C. Smith), Chap. 2. Pp. 12-43. Avi Publishing,
(1958). Penetration of maltosaccharides in pro- Westport, CT, U.S.A.
cessed Clingstone peaches. Food Technology, 12,
111-1 15.

(Received 24 January 1983)