Biotechnol. Prog. 1992, 8, 424-428
424
Mechanical Properties of Hydrocolloid Gels Filled with Internally Produced C02 Gas Bubbles A. Nussinovitch, R. Velez-Silvestre, and M. Peleg' Department of Food Science, University of Massachusetts, Amherst, Massachusetts 01003
Agar (2 % ), alginate (1% algin), and K-carrageenan (1.5% ) gel specimens were prepared from mother solutions that contained 0-2.5 % sodium bicarbonate (agar and carrageenan) or calcium carbonate (alginate). Upon immersion in a citric acid bath (0-2%), the carbonate reacted with the diffusing acid to produce numerous carbon dioxide bubbles. The compressive strength and deformability of the gas-filled gels so produced were determined using a Universal testing machine and compared with those of pure gels and gels containing the carbonate but not subjected to the process after various immersion times. While the agar and alginate gels retained considerable mechanical integrity even after several hours, the carrageenan gels disintegrated after about 2-5 h. Under similar conditions, the number of bubbles produced in the agar gels was about twice that in the alginate gels, an observation that cannot be explained solely by stoichiometric considerations.
Introduction Gas-filled gels have primarily been studied as a medium of bubble formation in attempts to simulate the effects of decompression on divers (1-3). Consequently, the gels were prepared under pressure to increase the gas solubility. The gases so studied were nitrogen, carbon dioxide, or helium, whose bubbles were formed upon decompression of the gel. Although less studied, gas-filled gels, or spongy gels, can also have other potential technological uses. For example, the mere presence of gas bubbles affects the gel density and can be used for its flotation in a liquid medium. It also creates internal space that can be infused with an active ingredient, or can simply increase the internal surface area of the gel. It can also result in unique textural characteristics that can be exploited in gel-based food products (3) or in a dehydrated form, to produce biodegradable sponges with potential medical applications. Introduction of gas bubbles into a gel matrix can be done by foaming the hydrocolloid solution prior to the setting of the gel (4). An alternative way is to form the bubbles in situ. This can be done by dissolving a bicarbonate or carbonate salt in the hydrocolloid solution and then dipping the set gel in acid to produce the carbon dioxide gas bubbles. Calculation shows, however, that even a 0.5% solution of NaHC03 or CaC03, for example, can release about 1.3 or 1.1 mL of COdmL of gel at ambient temperature, potentially enough to disrupt the gel to such an extent that it will have no appreciable strength. If, however, part of the formed gas can escape by diffusion or through cracks in the gels, then the gel can retain some of its mechanical integrity and its original shape. Another potential problem is that the mere exposure of the gel to an alkaline and/or acid environment can have an additional disruptive effect on its mechanical properties. Information on the conditions needed to produce this kind of gas-filled gels and how they affect the gels mechanical properties was unavailable from the literature.
* Author to whom correspondence should be addressed. 8756-7938/92/3008-0424$03.00/0
The objectives of this work were to find such conditions for agar, alginate, and K-carrageenan gels and to establish the effects of the bubble formation and presence on the physical and mechanical properties of these gels.
Materials and Methods Food-grade commercial agar (U.S. Biochemical Corp.), algin (Kelgin LV of Kelco), and K-carrageenan (Gelcarin HWG of Marine Colloids) were used for gel preparation. The agar powder, 2 % w/w, was dispersed in distilled water and heated to boiling. After the solution was cooled to 45-50 "C, NaHC03 powder was added in a concentration of 0.5-2.5% in 0.5% increments. %-Carrageenanpowder, 1.5% w/w, was dispersed in a 1%KC1 solution heated previously to about 65-70 "C. The dispersion was heated to boiling, and after it was cooled to about 75 "C, NaHC03 in concentrations of 0.5-2.5 (in 0.5% incrementa) was added. Alginate powder, 1% w/w, calcium hydrogen orthophosphate (CaHP04), 1%,and CaC03,0.5-2.5% in 0.5% incrementa, were added, slowly, to stirred cold distilled water (10"C),until complete dissolution of the ingredients. A freshly prepared solution of 1% glucono-&lactonewas then admixed with this solution using vigorous stirring. The hot solutions of the agar and carrageenan were poured into a special split metal mold shown schematically in Figure 1 (5). The mold consisted of metal rings held with an adhesive tape while mounted on a special rod with the rings' internal diameter. After being cooled for 1h at 20 "C, the tape and the gel were cut with a sharp blade and the cylindrical specimens (1.5 cm X 1.5 cm) were removed from the rings as shown also in the figure. The method to produce the cold set alginate was different. Since alginate gels undergo considerable shrinkage and syneresis, the algin solution was poured into a plastic container (10 X 10 X 8 cm) and allowed to set there. After 48 h (61,specimens were taken from the slab using a cork borer. The exact dimensions of each specimen were determined with a caliper. The agar and carrageenangels were aged for 24 h (6),before immersion in 0.5-2% citric acid solutions. The volume of citric solutions was about 100times the volume of a single gel specimen to guarantee
0 1992 American Chemical Society and American Institute of Chemical Engineers
425
Blotechnol. Rug., 1992, Vol. 8, No. 5
r
m
HOT GUM SOLUTION
1
FORCE APPLIED
TAPED' JOINTS
W R I N G
(0) (b) (C) Figure 1. Schematic view of the split mold used for making cylindrical gel specimen (from ref 5).
2
0
6
4
8
10
t 112 (minll2) Figure 4. Acid diffusion in a carrageenangel. ( X is the distance penetrated after time t; different symbols represent different replicates.)
-
sgar+ alginate -&canageenan
l-loh 1.06
1
E
5>. 1.02 k
0.98 W
0
Figure 2. Appearance of a C02-filled gel. 8-
0.94
r2 = .993
m
J
0
2
3
4
5
6
Figure 5. Density changesin carbonatecontaining agar, alginate, and carrageenangels during immersion in a 2% citric acid bath.
n
E ' E .4 -
x . 22% Citric acid 1% NaHC03 .
0
1
TIME (hrs)
6-
2
-
I
4
.
'
r
6
.
-
r
-
8
.
10
Figure 3. Acid diffusion in an agar gel. ( X is the distance penetrated after time t; different symbols represent different replicates.)
excessacid. The acid diffusion into the gels was monitored for about 2.5 h using phenolphthalein as an indicator. Specimens of pure gels as well as gels with added carbonate, and before the acid diffusion,were mechanically tested in 30-min intervals. They were compressed to failure, between parallel lubricated plates, at a constant deformation (displacement) rate of 10 mm/min, corresponding to an initial 0.011 s-l strain rate using an Instron Universal Testing Machine Model 1OOO. The Instron was connected to a Macintosh I1 computer by an analog to digital conversion interface card from Strawberry Tree Computers. A specially developed program written in Quick Basic by M. D. Normand, Food Science Department, University of Massachusetts, performed the data acquisition and conversion of the Instron's continuous voltage vs time output into digitized force-deformation, force-time, stress-strain, or
stress-time relationships with any desired definition of the stress and strain. All the mechanical tests were performed in triplicate. Counting of bubble number per unit volume was done using a Bausch and Lomb light microscope. Pictures were taken using a Pentax K 1000 camera and a Ashai Pentax microscope adapter. A cork borer with a 4-mm internal diameter was used to bore circular specimens from sliced gels 1mm thick. Six specimens were taken from each gel, three cut longitudinally (axially) and three transversally (parallel to the diameter). The reported results are the mean values of these six counts.
Results and Discussion The appearance of a COz-filled gel is shown in Figure 2. The picture was taken after the process has been completed, that is, after the acid had reached the gel center through diffusion. That the acid motion in the gel was diffusion controlled was evident from the linearity of the penetrated distance (X) vs t1/2plots. Examples of such plots are shown in Figures 3 and 4. (As previously mentioned, the gels used in these experiments had phenolphthalein in them so that the distance could be measured directly with a caliper after the specimen was dissected.) The slope of the X vs plots of all the gels was about 0.7 m m ~ m i n ~This . ~ . suggeststhat the hydrocolloid species and concentration, in the range tested, had little effect on the acid diffusion rate.
428
Blotechnol. prog., 1992, VOI.
a, NO. 5
AGAR 50
40
2% cmc rM rol”
30
Gel acgr immersion in 1 % CBlc wid soludon NaHC03+2% C I I K Lcld
20
2
h
10
Y Y
Figure 6. Typical stressatrain relationships of
agar gels before and after internal COz bubble formation.
The number of bubbles formed depended on the immersion time and the carbonate concentration, but the pattern was not the same in the three types of gels. In agar, after about 2.5 h, there were already about 4500 bubbles/cm3. Afterward, the number slowly increased with continued exposure and reached about 5500 bubbles/cm3 after 54 h of immersion. Compared to agar, the alginate gels had a considerably smaller number of bubbles, on the order of 900/cm3 after 2.5 h. It increased to about 20002700 after 24 or 36 h depending on the carbonate concentration. Since the theoretical COZyield of sodium bicarbonate and calcium carbonate is about 51% and 44 % respectively, the difference in the number of bubbles cannot be solely attributed to stoichiometric considerations. It was more probably the result of differences in the nucleation mechanism, and possibly in the diffusion or escape rate of the formed C02 gas. Confirmation of this conjecture, however, requires a different kind of experiment not performed in the reported study. The carrageenan gels had about 300 bubbles/cm3 after being exposed to the acid for about 2 h. But since they disintegrated soon after, this information serves no useful purpose (see below). As could be expected the bubble formation lowered the density of the gels (Figure 5) and caused their flotation. In the agar gels, however, the bubbles were gradually filled withliquid and after about 2.5 h they started to sink again. Mechanical Propertiesof the Gas-FilledGels. Agar Gels. Typical compressive stress-strain relationships of regular and gas-filled agar gels are shown in Figure 6.They demonstrate that the incorporation of the carbonate alone, or dipping the original gel in acid by itself, does not have a significant effect on the compressive behavior of the gel.
Once bubbles were formed, however, there was a drastic decrease in the compressive strength (stress at failure) and in the strain the specimen could sustainbefore it failed. A significant reduction in stiffness expressed by the slope of the initial and practically linear part of the stress-strain relationship was only observed in gels with a high carbonate content (2.5 5% ). The weakening of the gel was probably due to local rupture produced during bubble formation and growth, which acted as sites for failure propagation. But although the gas-filled agar gels lost strength and deformability, they still maintained considerable mechanical integrity and could be handled as regular solid gels. The extent to which the mechanical properties were altered depended on the carbonatecontents, the immersion time in the acid bath, and the acid concentration. In the range tested, the combined effect could result in a drop of strength from about 50 to 30 kPa after 30 min of immersion and to about 15 kPa after 2 h. Alginate Gels. Typical compressive stress-strain relationships of ordinary and gas-filled alginate gels are shown in Figure 7. In contrast with the agar gels, immersion in acid actually increased the pure gel strength and deformability. This was due to acid-induced crosslinking, which helped the gel retain ita mechanical strength even in the face of structural disruption caused by the bubbles formation. The carbonate presence, however, had a disruptive effect, primarily manifested in lower stiffness. This was mainly due to the pH increase to a level beyond that required for optimal cross-linking. As could be expected, the gel strength depended on both the acid and calcium carbonate concentrations. At the
Biotechnol. h g . , 1992, Vol. 8, No. 5
427
ALGINATE
6C 50
Gel s f b i~ m m m In 1% cimc Kia ~ o l u w n
IO 5 8 NsHC03*2% Cnnc acid1
40
(0 5% NaHC03+1% Clvlc
ncdl
30 IC
20
acid solullon
n
2
% NeHC03 Illled gel
10
Y
Y
cnO v) 6 1 0 1
1
,“i
(2 5% NeHC03+2% Clmc acid1
G85n1h3w1
ca\ flIW *I
(2.5% NsHCO3+1% CtWc acid
20 1 5% NaHC03 fillsd pel
10
2 5% NaHC03 filled gel
n 1
0.0
0.2
0.4
0.6
0.0
1.0 .0
0.2
0.4
0.6
0.8
.o
HENCKY’S STRAIN Figure 7. Typical stress-strain relationships of alginate gels before and after internal COz bubble formation.
CARRAGEENAN I
50 ‘T 40
-
30 20
sner immnlon in scla solullon
-
1 % CIVIC
v
pura gel
40
v,
f
2 5 % NnHC03 lilled gel
1 5 8 NaH-03 1, led 901
0.0
0.1
0.2
0.3
0.4
0.5
HENCKY’S STRAIN Figure 8. Typical stress-strain relationships of K-carrageenan gels before and after internal COz bubble formation.
Biotechnol. Rog., 1992, Vol. 8, No. 5
428
higher calcium carbonate concentration (2.5%), it also increased with the immersion time by up to about 10%. K-Carrageenan Gels. The mere immersion of the K-carrageenan gel in an acid bath weakened it considerably (Figure 8). The added sodium bicarbonate also weakened the gel and reduced its deformability but not to the same extent as the acid. With the added structural disruption caused by the bubble formation, the strength and deformability were further reduced, leading to total disintegration upon long immersion in the acid bath.
Conclusions Carbon dioxide-filled gels with a considerable mechanical integrity can be produced by the simple process of incorporating a bicarbonate or carbonate salt in the hydrocolloid mother solution and immersion of the set gel in an acid bath. The structure and mechanical properties of the resulting gels can, to a great extent, be manipulated and controlled through adjustment of the bicarbonate, or carbonate, the acid concentration,and the immersion time. Among the three types of gas-filled gels tested, the agar and alginate by far outperformanced the K-carrageenan which appears to be unsuitable for the purpose. All the described gels were prepared and tested in a pure form; that is, they did not contain any ingredient not absolutely necessary for setting and gas production. But, even under these circumstances, different bubble formation and growth patterns were clearly observed. The objectives of the work were to demonstrate that it is possible to produce these kinds of gels and to find conditions under which they are mechanically stable. The interplays between the nucleation pattern, the bubble internal pressure, and the gel matrix flexibility and strength were not studied. The same also applies to impurities that can, at least theoretically, influence the nucleation pattern and the origin and quality of the hydrocolloid used. But if and when gas-filled gels find a commercial use in the future, especially in the biotechnology or foods, it is unlikely that they will be prepared in a pure form. How other ingredients such as sugars, fruit pulp particles, enzymes, or nutrients affect the mechanical and physical properties of the resulting gels
will have to be learned independently, and in a given specific system. The described methodology will most probably be a convenient means to study such effects. Similarly, other methods of internal bubble formation can also be tried, by replacing the sodium bicarbonate by another soluble carbonate, for example. At least in principle, the process can also be reversed by incorporating the acid in the gel’s mother solution and immersing the set gel in a bicarbonate solution, for example. But again, specific effects of the process on the resulting properties of the gels will have to be studied independently by methods that are the same as or similar to those described in this work.
Acknowledgment This paper is a contribution of the Massachusetts Agricultural Experiment Station at Amherst. The project was supported in part by BARD, the United States-Israel Binational Agricultural Research and Development Fund (Grant 1-1166-86). Literature Cited (1) Yount, D. E.; Straw, R. H. Bubble formation in gelatin: A model for decompression sickness. J. Appl. Phys. 1976’47, 5081-5089. (2) D’Arrigo, J. S.The surface chemistry of bubble formation. Aviation, Space Env. Med. 1978, (2),358-361. (3) DArrigo, J. S. Stable gas-in-liquid emulsions; Elsevier: Amsterdam, 1986. (4) Morris, V. J.Multicomponent gels. In Gums and stabilizers for the food industry; Phillips, G. O., Wedlock, D. J., Williams, P. A., Eds.; Elsevier: London, 1986;pp 81-99.
(5) Nussinovitch, A.; Peleg, M.; Normand, M. D. A modified Maxwell and non-exponential model for characterization of the stress relaxation of agar and alginate gels. J. Food Sci. 1989,54, 1013-1016. (6) Nussinovitch, A.; Peleg, M. Strength-time relationships of agar and alginate gels. J. Texture Stud. 1990,21, 51-60.
Accepted April 1, 1992. Registry No. COZ,124-38-9; agar, 9002-18-0; K-carrageenan, 11114-20-8; Kelgin LV, 9005-38-3.
424
Mechanical Properties of Hydrocolloid Gels Filled with Internally Produced C02 Gas Bubbles A. Nussinovitch, R. Velez-Silvestre, and M. Peleg' Department of Food Science, University of Massachusetts, Amherst, Massachusetts 01003
Agar (2 % ), alginate (1% algin), and K-carrageenan (1.5% ) gel specimens were prepared from mother solutions that contained 0-2.5 % sodium bicarbonate (agar and carrageenan) or calcium carbonate (alginate). Upon immersion in a citric acid bath (0-2%), the carbonate reacted with the diffusing acid to produce numerous carbon dioxide bubbles. The compressive strength and deformability of the gas-filled gels so produced were determined using a Universal testing machine and compared with those of pure gels and gels containing the carbonate but not subjected to the process after various immersion times. While the agar and alginate gels retained considerable mechanical integrity even after several hours, the carrageenan gels disintegrated after about 2-5 h. Under similar conditions, the number of bubbles produced in the agar gels was about twice that in the alginate gels, an observation that cannot be explained solely by stoichiometric considerations.
Introduction Gas-filled gels have primarily been studied as a medium of bubble formation in attempts to simulate the effects of decompression on divers (1-3). Consequently, the gels were prepared under pressure to increase the gas solubility. The gases so studied were nitrogen, carbon dioxide, or helium, whose bubbles were formed upon decompression of the gel. Although less studied, gas-filled gels, or spongy gels, can also have other potential technological uses. For example, the mere presence of gas bubbles affects the gel density and can be used for its flotation in a liquid medium. It also creates internal space that can be infused with an active ingredient, or can simply increase the internal surface area of the gel. It can also result in unique textural characteristics that can be exploited in gel-based food products (3) or in a dehydrated form, to produce biodegradable sponges with potential medical applications. Introduction of gas bubbles into a gel matrix can be done by foaming the hydrocolloid solution prior to the setting of the gel (4). An alternative way is to form the bubbles in situ. This can be done by dissolving a bicarbonate or carbonate salt in the hydrocolloid solution and then dipping the set gel in acid to produce the carbon dioxide gas bubbles. Calculation shows, however, that even a 0.5% solution of NaHC03 or CaC03, for example, can release about 1.3 or 1.1 mL of COdmL of gel at ambient temperature, potentially enough to disrupt the gel to such an extent that it will have no appreciable strength. If, however, part of the formed gas can escape by diffusion or through cracks in the gels, then the gel can retain some of its mechanical integrity and its original shape. Another potential problem is that the mere exposure of the gel to an alkaline and/or acid environment can have an additional disruptive effect on its mechanical properties. Information on the conditions needed to produce this kind of gas-filled gels and how they affect the gels mechanical properties was unavailable from the literature.
* Author to whom correspondence should be addressed. 8756-7938/92/3008-0424$03.00/0
The objectives of this work were to find such conditions for agar, alginate, and K-carrageenan gels and to establish the effects of the bubble formation and presence on the physical and mechanical properties of these gels.
Materials and Methods Food-grade commercial agar (U.S. Biochemical Corp.), algin (Kelgin LV of Kelco), and K-carrageenan (Gelcarin HWG of Marine Colloids) were used for gel preparation. The agar powder, 2 % w/w, was dispersed in distilled water and heated to boiling. After the solution was cooled to 45-50 "C, NaHC03 powder was added in a concentration of 0.5-2.5% in 0.5% increments. %-Carrageenanpowder, 1.5% w/w, was dispersed in a 1%KC1 solution heated previously to about 65-70 "C. The dispersion was heated to boiling, and after it was cooled to about 75 "C, NaHC03 in concentrations of 0.5-2.5 (in 0.5% incrementa) was added. Alginate powder, 1% w/w, calcium hydrogen orthophosphate (CaHP04), 1%,and CaC03,0.5-2.5% in 0.5% incrementa, were added, slowly, to stirred cold distilled water (10"C),until complete dissolution of the ingredients. A freshly prepared solution of 1% glucono-&lactonewas then admixed with this solution using vigorous stirring. The hot solutions of the agar and carrageenan were poured into a special split metal mold shown schematically in Figure 1 (5). The mold consisted of metal rings held with an adhesive tape while mounted on a special rod with the rings' internal diameter. After being cooled for 1h at 20 "C, the tape and the gel were cut with a sharp blade and the cylindrical specimens (1.5 cm X 1.5 cm) were removed from the rings as shown also in the figure. The method to produce the cold set alginate was different. Since alginate gels undergo considerable shrinkage and syneresis, the algin solution was poured into a plastic container (10 X 10 X 8 cm) and allowed to set there. After 48 h (61,specimens were taken from the slab using a cork borer. The exact dimensions of each specimen were determined with a caliper. The agar and carrageenangels were aged for 24 h (6),before immersion in 0.5-2% citric acid solutions. The volume of citric solutions was about 100times the volume of a single gel specimen to guarantee
0 1992 American Chemical Society and American Institute of Chemical Engineers
425
Blotechnol. Rug., 1992, Vol. 8, No. 5
r
m
HOT GUM SOLUTION
1
FORCE APPLIED
TAPED' JOINTS
W R I N G
(0) (b) (C) Figure 1. Schematic view of the split mold used for making cylindrical gel specimen (from ref 5).
2
0
6
4
8
10
t 112 (minll2) Figure 4. Acid diffusion in a carrageenangel. ( X is the distance penetrated after time t; different symbols represent different replicates.)
-
sgar+ alginate -&canageenan
l-loh 1.06
1
E
5>. 1.02 k
0.98 W
0
Figure 2. Appearance of a C02-filled gel. 8-
0.94
r2 = .993
m
J
0
2
3
4
5
6
Figure 5. Density changesin carbonatecontaining agar, alginate, and carrageenangels during immersion in a 2% citric acid bath.
n
E ' E .4 -
x . 22% Citric acid 1% NaHC03 .
0
1
TIME (hrs)
6-
2
-
I
4
.
'
r
6
.
-
r
-
8
.
10
Figure 3. Acid diffusion in an agar gel. ( X is the distance penetrated after time t; different symbols represent different replicates.)
excessacid. The acid diffusion into the gels was monitored for about 2.5 h using phenolphthalein as an indicator. Specimens of pure gels as well as gels with added carbonate, and before the acid diffusion,were mechanically tested in 30-min intervals. They were compressed to failure, between parallel lubricated plates, at a constant deformation (displacement) rate of 10 mm/min, corresponding to an initial 0.011 s-l strain rate using an Instron Universal Testing Machine Model 1OOO. The Instron was connected to a Macintosh I1 computer by an analog to digital conversion interface card from Strawberry Tree Computers. A specially developed program written in Quick Basic by M. D. Normand, Food Science Department, University of Massachusetts, performed the data acquisition and conversion of the Instron's continuous voltage vs time output into digitized force-deformation, force-time, stress-strain, or
stress-time relationships with any desired definition of the stress and strain. All the mechanical tests were performed in triplicate. Counting of bubble number per unit volume was done using a Bausch and Lomb light microscope. Pictures were taken using a Pentax K 1000 camera and a Ashai Pentax microscope adapter. A cork borer with a 4-mm internal diameter was used to bore circular specimens from sliced gels 1mm thick. Six specimens were taken from each gel, three cut longitudinally (axially) and three transversally (parallel to the diameter). The reported results are the mean values of these six counts.
Results and Discussion The appearance of a COz-filled gel is shown in Figure 2. The picture was taken after the process has been completed, that is, after the acid had reached the gel center through diffusion. That the acid motion in the gel was diffusion controlled was evident from the linearity of the penetrated distance (X) vs t1/2plots. Examples of such plots are shown in Figures 3 and 4. (As previously mentioned, the gels used in these experiments had phenolphthalein in them so that the distance could be measured directly with a caliper after the specimen was dissected.) The slope of the X vs plots of all the gels was about 0.7 m m ~ m i n ~This . ~ . suggeststhat the hydrocolloid species and concentration, in the range tested, had little effect on the acid diffusion rate.
428
Blotechnol. prog., 1992, VOI.
a, NO. 5
AGAR 50
40
2% cmc rM rol”
30
Gel acgr immersion in 1 % CBlc wid soludon NaHC03+2% C I I K Lcld
20
2
h
10
Y Y
Figure 6. Typical stressatrain relationships of
agar gels before and after internal COz bubble formation.
The number of bubbles formed depended on the immersion time and the carbonate concentration, but the pattern was not the same in the three types of gels. In agar, after about 2.5 h, there were already about 4500 bubbles/cm3. Afterward, the number slowly increased with continued exposure and reached about 5500 bubbles/cm3 after 54 h of immersion. Compared to agar, the alginate gels had a considerably smaller number of bubbles, on the order of 900/cm3 after 2.5 h. It increased to about 20002700 after 24 or 36 h depending on the carbonate concentration. Since the theoretical COZyield of sodium bicarbonate and calcium carbonate is about 51% and 44 % respectively, the difference in the number of bubbles cannot be solely attributed to stoichiometric considerations. It was more probably the result of differences in the nucleation mechanism, and possibly in the diffusion or escape rate of the formed C02 gas. Confirmation of this conjecture, however, requires a different kind of experiment not performed in the reported study. The carrageenan gels had about 300 bubbles/cm3 after being exposed to the acid for about 2 h. But since they disintegrated soon after, this information serves no useful purpose (see below). As could be expected the bubble formation lowered the density of the gels (Figure 5) and caused their flotation. In the agar gels, however, the bubbles were gradually filled withliquid and after about 2.5 h they started to sink again. Mechanical Propertiesof the Gas-FilledGels. Agar Gels. Typical compressive stress-strain relationships of regular and gas-filled agar gels are shown in Figure 6.They demonstrate that the incorporation of the carbonate alone, or dipping the original gel in acid by itself, does not have a significant effect on the compressive behavior of the gel.
Once bubbles were formed, however, there was a drastic decrease in the compressive strength (stress at failure) and in the strain the specimen could sustainbefore it failed. A significant reduction in stiffness expressed by the slope of the initial and practically linear part of the stress-strain relationship was only observed in gels with a high carbonate content (2.5 5% ). The weakening of the gel was probably due to local rupture produced during bubble formation and growth, which acted as sites for failure propagation. But although the gas-filled agar gels lost strength and deformability, they still maintained considerable mechanical integrity and could be handled as regular solid gels. The extent to which the mechanical properties were altered depended on the carbonatecontents, the immersion time in the acid bath, and the acid concentration. In the range tested, the combined effect could result in a drop of strength from about 50 to 30 kPa after 30 min of immersion and to about 15 kPa after 2 h. Alginate Gels. Typical compressive stress-strain relationships of ordinary and gas-filled alginate gels are shown in Figure 7. In contrast with the agar gels, immersion in acid actually increased the pure gel strength and deformability. This was due to acid-induced crosslinking, which helped the gel retain ita mechanical strength even in the face of structural disruption caused by the bubbles formation. The carbonate presence, however, had a disruptive effect, primarily manifested in lower stiffness. This was mainly due to the pH increase to a level beyond that required for optimal cross-linking. As could be expected, the gel strength depended on both the acid and calcium carbonate concentrations. At the
Biotechnol. h g . , 1992, Vol. 8, No. 5
427
ALGINATE
6C 50
Gel s f b i~ m m m In 1% cimc Kia ~ o l u w n
IO 5 8 NsHC03*2% Cnnc acid1
40
(0 5% NaHC03+1% Clvlc
ncdl
30 IC
20
acid solullon
n
2
% NeHC03 Illled gel
10
Y
Y
cnO v) 6 1 0 1
1
,“i
(2 5% NeHC03+2% Clmc acid1
G85n1h3w1
ca\ flIW *I
(2.5% NsHCO3+1% CtWc acid
20 1 5% NaHC03 fillsd pel
10
2 5% NaHC03 filled gel
n 1
0.0
0.2
0.4
0.6
0.0
1.0 .0
0.2
0.4
0.6
0.8
.o
HENCKY’S STRAIN Figure 7. Typical stress-strain relationships of alginate gels before and after internal COz bubble formation.
CARRAGEENAN I
50 ‘T 40
-
30 20
sner immnlon in scla solullon
-
1 % CIVIC
v
pura gel
40
v,
f
2 5 % NnHC03 lilled gel
1 5 8 NaH-03 1, led 901
0.0
0.1
0.2
0.3
0.4
0.5
HENCKY’S STRAIN Figure 8. Typical stress-strain relationships of K-carrageenan gels before and after internal COz bubble formation.
Biotechnol. Rog., 1992, Vol. 8, No. 5
428
higher calcium carbonate concentration (2.5%), it also increased with the immersion time by up to about 10%. K-Carrageenan Gels. The mere immersion of the K-carrageenan gel in an acid bath weakened it considerably (Figure 8). The added sodium bicarbonate also weakened the gel and reduced its deformability but not to the same extent as the acid. With the added structural disruption caused by the bubble formation, the strength and deformability were further reduced, leading to total disintegration upon long immersion in the acid bath.
Conclusions Carbon dioxide-filled gels with a considerable mechanical integrity can be produced by the simple process of incorporating a bicarbonate or carbonate salt in the hydrocolloid mother solution and immersion of the set gel in an acid bath. The structure and mechanical properties of the resulting gels can, to a great extent, be manipulated and controlled through adjustment of the bicarbonate, or carbonate, the acid concentration,and the immersion time. Among the three types of gas-filled gels tested, the agar and alginate by far outperformanced the K-carrageenan which appears to be unsuitable for the purpose. All the described gels were prepared and tested in a pure form; that is, they did not contain any ingredient not absolutely necessary for setting and gas production. But, even under these circumstances, different bubble formation and growth patterns were clearly observed. The objectives of the work were to demonstrate that it is possible to produce these kinds of gels and to find conditions under which they are mechanically stable. The interplays between the nucleation pattern, the bubble internal pressure, and the gel matrix flexibility and strength were not studied. The same also applies to impurities that can, at least theoretically, influence the nucleation pattern and the origin and quality of the hydrocolloid used. But if and when gas-filled gels find a commercial use in the future, especially in the biotechnology or foods, it is unlikely that they will be prepared in a pure form. How other ingredients such as sugars, fruit pulp particles, enzymes, or nutrients affect the mechanical and physical properties of the resulting gels
will have to be learned independently, and in a given specific system. The described methodology will most probably be a convenient means to study such effects. Similarly, other methods of internal bubble formation can also be tried, by replacing the sodium bicarbonate by another soluble carbonate, for example. At least in principle, the process can also be reversed by incorporating the acid in the gel’s mother solution and immersing the set gel in a bicarbonate solution, for example. But again, specific effects of the process on the resulting properties of the gels will have to be studied independently by methods that are the same as or similar to those described in this work.
Acknowledgment This paper is a contribution of the Massachusetts Agricultural Experiment Station at Amherst. The project was supported in part by BARD, the United States-Israel Binational Agricultural Research and Development Fund (Grant 1-1166-86). Literature Cited (1) Yount, D. E.; Straw, R. H. Bubble formation in gelatin: A model for decompression sickness. J. Appl. Phys. 1976’47, 5081-5089. (2) D’Arrigo, J. S.The surface chemistry of bubble formation. Aviation, Space Env. Med. 1978, (2),358-361. (3) DArrigo, J. S. Stable gas-in-liquid emulsions; Elsevier: Amsterdam, 1986. (4) Morris, V. J.Multicomponent gels. In Gums and stabilizers for the food industry; Phillips, G. O., Wedlock, D. J., Williams, P. A., Eds.; Elsevier: London, 1986;pp 81-99.
(5) Nussinovitch, A.; Peleg, M.; Normand, M. D. A modified Maxwell and non-exponential model for characterization of the stress relaxation of agar and alginate gels. J. Food Sci. 1989,54, 1013-1016. (6) Nussinovitch, A.; Peleg, M. Strength-time relationships of agar and alginate gels. J. Texture Stud. 1990,21, 51-60.
Accepted April 1, 1992. Registry No. COZ,124-38-9; agar, 9002-18-0; K-carrageenan, 11114-20-8; Kelgin LV, 9005-38-3.
