ARTICLE IN PRESS

International Dairy Journal 15 (2005) 495–500

www.elsevier.com/locate/idairyj

On fat destabilization and composition of the air interface in ice

cream containing saturated and unsaturated monoglyceride
Z. Zhang, H.D. Goff
Department of Food Science, University of Guelph, Gordon Street, Guelph, Ont., Canada N1G 2W1
Received 23 October 2003; accepted 25 August 2004

Abstract

The influence of glycerol monostearate (GMS, saturated), glycerol monooleate (GMO, unsaturated) and type of milk proteins on
fat destabilization and the composition of the air interface in ice cream was investigated by particle size distribution, de-emulsified
fat content and immuno-gold labelled transmission electron microscopy . When ice cream was made from skim milk powder and
GMS, casein micelles, b-casein and b-lactoglobulin were all detected at the air interface. Fat was also found adsorbed or residing at
the air interface. GMO strongly displaced caseins from the fat interface, introduced more partially coalesced fat adsorbed to the air
interface and seemed to displace more proteins from the air interface. However, GMO was slightly less effective at displacing b-
lactoglobulin from the fat interface than GMS. Fat globules were seldom found adsorbed to the air interface when whey protein was
used as the only source of protein because ice cream emulsions were too stable to undergo desirable fat destabilization. In this case,
the air interface predominantly comprised b-lactoglobulin, regardless of emulsifier type.
r 2004 Elsevier Ltd. All rights reserved.

Keywords: Ice cream; Glycerol monooleate; Glycerol monostearate; Milk proteins: Fat destabilization; Air interface; Immunolabelling

1. Introduction Leeder, 1971). During freezing/aeration, emulsifiers

such as mono- and diacylglycerols (mono- and diglycer-
Ice cream is a frozen aerated emulsion in which ides) or polyoxyethylene sorbitan esters are added to
surfactants are required to stabilize both the oil–water promote partial coalescence of fat globules, i.e.,
and the air–water interfaces. Surfactants in ice cream are destabilization of the protein stabilized-emulsion (Goff,
categorized as either large molecule surfactants, i.e., Liboff, Jordan, & Kinsella, 1987). Partial coalescence of
milk proteins, or small molecule surfactants, i.e., fat contributes to the formation of a fat network, which
emulsifiers. The competitive adsorption of proteins is essential for dryness, creaminess and melting-resis-
and emulsifier to the two interfaces is of great tance properties of ice cream, and is also important for
importance with respect to the texture and quality of the stability of air bubbles.
ice cream. A large amount of proteins is involved in the Fat globules become more susceptible to destabiliza-
initial production of the oil-in-water emulsion during tion after the low-temperature (
4 1C) ageing step,
mixing and homogenization. Emulsifiers are also present during which a considerable amount of protein is
in the protein-stabilized emulsion, although they are not displaced from the fat interface by emulsifiers, and the
required for the purpose of emulsification, as sufficient emulsifiers themselves undergo crystallization (Gelin,
protein is available (Goff & Jordan, 1989; Govin & Poyen, Courthaudon, Le Meste, & Lorient, 1994). The
magnitude of destabilization is related to the oil/serum
Corresponding author. Tel.: +1 519 824 4120 ext. 53878; fax: interfacial tension in the presence of emulsifier and the
+1 519 824 6631. extent of protein displacement (Goff & Jordan, 1989;
E-mail address: dgoff@uoguelph.ca (H.D. Goff). Bolliger, Goff, & Tharp, 2000). The shearing stability of

0958-6946/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.idairyj.2004.08.014
 ARTICLE IN PRESS
496 Z. Zhang, H.D. Goff / International Dairy Journal 15 (2005) 495–500

protein-stabilized emulsions varies. Sodium caseinate- lactoglobulin was from Bethyl Laboratories (Montgom-
stabilized emulsions exhibit greater stability than those ery, TX, USA). Primary antibody polyclonal mouse
stabilized by whey proteins or skim milk. Skim milk- anti-bovine b-casein was produced at the Animal Care
stabilized emulsions are the least stable (Van Camp, Van Facility, University of Guelph. Secondary antibodies
Calenberg, Van Oostveldt, & Huyghebaert, 1996; Segall were conjugated to 10 nm gold particles and obtained
& Goff, 1999). Higher levels of destabilization were from Sigma-Aldrich Canada (Oakville, Ont), i.e. goat
obtained with more hydrophilic emulsifiers (Govin & anti-rabbit immunoglobulin antibody (G7402) and goat
Leeder, 1971; Lin & Leeder, 1974; Euston, Singh, anti-mouse immunglobulin antibody (G7652). Purified
Munro, & Dalgleish, 1995). The destabilization power water, free of inorganic and organic impurities, was
of emulsifiers can be further classified, with unsaturated produced using a Milli-Q Ultrapure water purification
monoacylglycerols being better destabilizing agents than system (Millipore Corp., Bedford, MA, USA). All
saturated monoacylglycerols (Barfod, Krog, Larsen, & experiments were done in triplicate.
Buchheim, 1991; Goff & Jordan, 1989; Pelan, Watts,
Campbell, & Lips, 1997). However, emulsifiers with 2.2. Ice cream preparation
saturated hydrocarbon chains are better initiators of fat
crystallization than those with unsaturated chains, Ice cream mixes with the following compositions were
especially for the use of non-dairy fats (Berger, 1997). prepared: (1) 10% milk fat, 11% SMP, 15% sucrose,
Although the role of emulsifier in shear-induced emul- 0.12% guar gum, 0.015% carrageenan and 0.3% GMS
sion destabilization is widely discussed, it is still not well (GMS–SMP ice cream); (2) 10% milk fat, 11% SMP,
understood how different emulsifiers influence the 15% sucrose, 0.12% guar gum, 0.015% carrageenan and
composition of the air interface and the resting 0.3% GMO (GMO–SMP ice cream); (3) 10% milk fat,
microstructure of ice cream. 4.4% WPI, 6.4% whey permeate, 15% sucrose, 0.12%
In earlier papers, we showed using transmission guar gum, 0.015% carrageenan and 0.3% GMS
electron microscopy (TEM) that the presence of (GMS–WPI ice cream); and (4) 10% milk fat, 4.4%
emulsifier in ice cream led to direct fat adsorption to WPI, 6.4% whey permeate, 15% sucrose, 0.12% guar
air interfaces (Goff, Verespej, & Smith, 1999; Zhang & gum, 0.015% carrageenan and 0.3% GMO (GMO-WPI
Goff, 2004). In this work saturated and unsaturated ice cream). All the ingredients were put into water and
monoacylglycerols as well as skim milk powder (SMP) brought to a temperature of 65 1C, held at that
and whey protein isolate (WPI) were evaluated in terms temperature for 30 min, homogenized with a Gaulin
of their influence on fat destabilization and the V15-8 T (APV Gaulin, Everett, MA, USA) two-stage
composition of air interface in ice cream, with an single-piston homogenizer using 172 MPa pressure on
emphasis on the latter. Results were obtained using the first stage and 34 MPa on the second, and cooled to
particle size distribution, de-emulsified fat content and 4 1C overnight. Batch freezing of 2 L lots of ice cream
immuno-gold labelled TEM to determine fat structure mix was carried out in a Taylor Freezer (B733-32,
and distribution of b-casein and b-lactoglobulin. Teknicraft, Rockton, IL, USA). Ice cream was whipped
and frozen to a draw temperature of 5 1C and the total
whipping time was approximately 15 min. Ice cream
2. Materials and methods samples were then hardened at 35 1C for at least 24 h
before further testing was performed.
2.1. Materials
2.3. Measuring destabilization of fat
Butteroil (99.9% anhydrous) and SMP, (with total
protein 36 g 100 g1, calcium 12 mg g1, phosphorous Ice cream was melted in the refrigerator (4 1C) under
17 mg g1 and sodium 4.5 mg g1) were obtained locally static conditions. Two phases were evident in the ice
(Gay-Lea Foods, Guelph, Ont., Canada). Whey protein cream container after melting. On the top was a fluffy,
isolate (WPI, with total protein 92 g 100 g1, calcium relatively rigid air-rich phase that could be removed
1 mg g1, phosphorous 0.5 mg g1 and sodium from the container with a spoon. The lower layer was a
6.4 mg g1) was industrial grade and was produced by liquid which had the appearance of regular mix,
ion exchange processing (Protose Separations, Tees- hereafter referred to as the drainage. The fluffy air-rich
water, Ont., Canada). Whey permeate was obtained layer was moved into a centrifuge tube and centrifuged
from Parmalat Canada (London, Ont., Canada). Guar at 100 g for 4 min at 4 1C. The drainage was removed
gum, carrageenan, glycerol monostearate (GMS) and by inserting a siphon tube into the bottom of the
glycerol monooleate (GMO) were obtained from Da- centrifuge tube.
nisco (Scarborough, Ont., Canada). LR-gold resin was Destabilization of fat in the air-rich phase was
obtained from Polysciences (London, Ont., Canada). determined by a combination of particle size analysis
Primary antibody polyclonal rabbit anti-bovine b- and free oil methods. The volume-surface weighted
 ARTICLE IN PRESS
Z. Zhang, H.D. Goff / International Dairy Journal 15 (2005) 495–500 497

diameter distribution of fat globules (d3,2) was determined degree of fat coalescence in the ice cream made from
by integrated light scattering using a Mastersizer X SMP. In addition, particle size in ice cream made from
(Malvern Instruments Ltd., Malvern, UK). Measure- whey protein was not dependent on the saturation state
ments were performed at room temperature and dilution of the monoacylglycerol. However, GMO promoted
in the sample chamber was approximately 1:1000. A more coalescence of fat than was the case for GMS
45 mm lens was used for the measurements. This lens can when ice creams were made from SMP.
detect particles with size between 0.1 and 80 mm. There- The percentages (w/w) of the drained liquid in the
fore highly coalesced fat globules in the air-rich phase melted ice cream accounted for 47%, 90%, 88% and
may not be included in the measured data. Fat droplets in 90% for ice cream made from GMO–SMP, GMO–WPI,
the air-rich phase in melted ice cream were separated by GMS–SMP and GMS–WPI, respectively. This suggests
centrifugation at 15,000 g for 60 min at 40 1C. After that GMO–SMP ice cream gave the best melting
centrifugation the fat layer demonstrated two physical resistance. Barfod (2001) also reported that for ice
states at room temperature, a gel-like state and liquid oil. cream made from the mixture of skim and whey powder,
The liquid oil contained no protein and comprised of fat unsaturated emulsifiers gave better melting resistance
globules with their adsorbed proteins displaced by than their saturated counterparts, but did not mention
emulsifiers. The gel-like fat comprised of fat globules the ratio of skim milk to whey powder in the ice cream
that retained a certain amount of protein. A higher recipe. The data obtained in the present study did not
degree of fat destabilization led to a higher proportion of indicate a pronounced influence of the saturation state
protein-free fat in the total fat of the air-rich phase. of emulsifiers on the melting resistance of ice cream
made with whey proteins alone.
2.4. Immuno-gold labelling TEM Remaining air bubbles, surrounded by fat, were
visible in melted ice cream containing GMO–WPI,
Sample preparation for TEM and immuno-gold GMS–SMP and GMS–WPI. For ice cream made from
labelling procedures were the same as those described GMO–SMP, the whole top layer of the melted ice cream
by Zhang and Goff (2004). did not look fluffy, but was rather creamy. This was
probably because many air bubbles collapsed while the
fat network surrounding air bubbles maintained struc-
3. Results and discussion tural integrity during melting and draining of the ice
cream. Air bubbles in melted ice cream made from whey
3.1. Fat destabilization in air-rich phases protein were more stable, as demonstrated by higher
volume of air trapped by the intact air bubbles. The
All ice creams were melted after holding at 4 1C for density of air-rich phase in the melted ice cream made
24 h. The d3,2 of fat globules in the air-rich phase and from whey protein was 0.19 g mL1, and that for SMP
drained liquid in melted ice cream are shown in Fig. 1. was approximately 0.85 g mL1.
Particle size of the air-rich phase was larger than that of The protein-free fat in the air-rich phase was the
the drained liquid. The larger particles were believed to coalesced fat spreading at or close to the air interface.
be partially coalesced fat globules that were either Coalesced fat was melted down to liquid oil during
adsorbed to the air interface during shearing aeration centrifugation at 40 1C. Again, the saturation state of
or, due to their lower density, rose up to the air-rich the mono- or diacylglycerols did not obviously affect fat
phase during melting of the ice cream. The average destabilization in the ice cream made from whey
particle size in SMP ice cream was larger than that in ice proteins; the percentages of protein-free fat in the total
cream made from whey protein. This indicated a higher fat fraction of the air-rich phase were similar (
11%) in
GMO–WPI and GMS–WPI ice creams. However,
1.2 compared with GMS, GMO caused a much higher
percentage of protein-free fat in the air-rich phase in ice
1
cream made from SMP. Protein-free fat accounted for
0.8 46% and 22% in the total fat fraction of the air-rich
d3,2 (µm)

0.6 phase in GMO–SMP and GMS–SMP ice creams,

respectively. This indicated that the ability of unsatu-
0.4
rated monoacylglycerols to displace caseins from the fat
0.2 interface and to promote an extensive destabilization of
0 fat was much stronger than that of saturated mono-
GMO-SMP GMO-WPI GMS-SMP GMS-WPI acylglycerols. However, too much partially coalesced fat
Fig. 1. Average surface-weighted droplet size (d3,2) of the air-rich spreading on the air interface may have caused the
phase (white bar) and drained liquid (black bar) in melted ice creams. collapse of air bubbles in the GMO–SMP ice cream
See text for formulation variables. during melting.
 ARTICLE IN PRESS
498 Z. Zhang, H.D. Goff / International Dairy Journal 15 (2005) 495–500

3.2. Microstructure and interfacial composition substituted by resin but can de discriminated based on
shape and interface (Goff et al., 1999). In this study, we
The microstructure of ice cream as viewed by freeze- ignored ice crystals and focussed on fat, air and the
substitution TEM shows the cross-section of ice crystals, unfrozen phase. Fat globules were more discrete and
air bubbles, fat globules and clusters and the unfrozen intact and more casein micelles could be seen at the fat
phase containing casein micelles. Air, ice and fat are all interface in GMS–SMP ice cream (Figs. 2a and 3a) than
in GMO–SMP ice cream (Figs. 2b and 3b). This
microstructure suggests that GMO also displaced
adsorbed protein at the air interface, which is consistent

Fig. 2. TEM of ice cream immunolabelled for b-casein. Ice cream was Fig. 3. TEM of ice cream immunolabelled for b-lactoglobulin. Ice
made from 11% SMP, 15% sugar, 10% fat, 0.12% guar gum, 0.015% cream was made from 11% SMP, 15% sugar, 10% fat, 0.12% guar
carrageenan and (a) 0.3% GMS or (b) 0.3% GMO. The small black gum, 0.015% carrageenan and (a) 0.3% GMS or (b) 0.3% GMO. The
dots indicate the location of b-casein. A: air bubble, F: fat globules, small black dots indicate the location of b-lactoglobulin. A: air bubble,
CM: casein micelles. Bar ¼ 500 nm: F: fat globules, CM: casein micelles. Bar ¼ 500 nm:
 ARTICLE IN PRESS
Z. Zhang, H.D. Goff / International Dairy Journal 15 (2005) 495–500 499

with the previous emulsifier effect studies (Goff & Immuno-gold labelling suggested that b-casein was
Jordan, 1989; Barfod et al., 1991; Barfod, 2001). more preferentially associated with interfaces than b-
Partially coalesced fat was seen to adsorb to the air lactoglobulin, however the extent of labelling was lower
interface, moreso in GMO–SMP ice cream compared to than anticipated, making quantification impossible. This
GMS–SMP ice cream. was perhaps due to a low level of reaction of the
Casein micelles, b-casein and b-lactoglobulin were all antibody with chemically fixed proteins (Zhang & Goff,
more evident at the air interface of GMS-SMP ice cream 2004). It is thus possible to suggest that GMO also acted
(Figs. 2a and 3a) compared to GMO–SMP ice cream into displacing protein adsorption at the air interface, in
(Figs. 2b and 3b). Protein adsorption at the air interface addition to protein displacement from the fat interface,
was more disrupted in GMO–SMP ice cream by the either through direct adsorption or via enhanced
presence of partially-coalesced fat (Figs. 2b and 3b). adsorption of partially coalesced fat at the air interface.

Fig. 4. TEM of ice cream immunolabelled for b-lactoglobulin. Ice cream was made from 15% sugar, 10% fat, 4.4% WPI, 6.6% whey permeate,
0.12% guar gum, 0.015% carrageenan and (a,b) 0.3% GMS or (c,d) 0.3% GMO. The small black dots indicate the location of a-lactoglobulin. A: air
bubble, F: fat globules. Bar ¼ 500 nm:
 ARTICLE IN PRESS
500 Z. Zhang, H.D. Goff / International Dairy Journal 15 (2005) 495–500

A considerable number of intact fat globules were monoacylglycerol at promoting a limited degree of
found in the TEM pictures of WPI ice cream regardless partial coalescence of fat, and, as a result, slightly more
of the emulsifier used. Most of the fat interface was fat was adsorbed to the air interface. This demonstrated
covered with whey protein, which included labelled b- the interrelationship between the composition of the fat
lactoglobulin (Figs. 4a and c). Discrete fat globules interface and the composition of the air interface.
suggested that neither emulsifier promoted protein
desorption nor fat destabilization in the WPI ice creams
as much as they did in SMP ice creams. Some labelled b- Acknowledgements
lactoglobulin was also found in the serum phase.
The air interface of GMO–WPI ice cream comprised We thank Unilever Research Colworth, UK for
mainly whey protein, and b-lactoglobulin was detected. financial support and Dr. Sandy Smith for assistance
Fewer fat globules were adsorbed to the air interface with the microscopy studies.
directly with GMO compared to GMS (Fig. 4d
compared to Fig. 4b). This suggested that GMO was
less powerful at displacing whey protein from the fat References
interface than GMS, country to what was found with
SMP. In GMS–WPI ice cream a few fat globules close to Barfod, N. M. (2001). The emulsifier effect. Dairy Industries
International, 66, 32–33.
the air interface were occasionally seen to have part of Barfod, N. M., Krog, N., Larsen, G., & Buchheim, W. (1991). Effects
the adsorbed protein displaced (Fig. 4b). This was not of emulsifiers on protein-fat interaction in ice cream mix during
found in GMO–WPI ice cream (Fig. 4d). The mechan- aging 1: Quantitative analyses. Fat Science Technology, 93, 24–29.
ism of the different effect of saturated and unsaturated Berger, K. G. (1997). Ice Cream. In S. E. Friberg, & K. Larsson (Eds.),
monoacylglycerol on the de-emulsification of whey Food Emulsions, (3rd ed.) (pp. 413–490). New York: Marcel
Dekker.
protein-stabilized emulsions is not well understood. Bolliger, S., Goff, H. D., & Tharp, B. W. (2000). Correlation between
The work of Clark, Wilde, Wilson, and Wustneck colloidal properties of ice cream mix and ice cream. International
(1992) demonstrated a higher affinity of b-lactoglobulin Dairy Journal, 10, 303–309.
to saturated emulsifiers, e.g., lauryl and stearoyl sucrose Clark, D. C., Wilde, P. J., Wilson, D. R., & Wustneck, R. (1992). The
interaction of sucrose esters with b-lactoglobulin and b-casein from
esters, than to an unsaturated emulsifier, i.e., mono-
bovine milk. Food Hydrocolloids, 6, 173–186.
unsaturated oleic sucrose ester. The stronger interaction Euston, S. E., Singh, H., Munro, P. A., & Dalgleish, D. G. (1995).
of the saturated emulsifiers with b-lactoglobulin may Competitive adsorption between sodium caseinate and oil-soluble
have caused the b-lactoglobulin–emulsifier complex to and water-soluble surfactants in oil-in-water emulsions. Journal of
enter the serum phase. Generally, the air interface of Food Science, 60, 1124–1131.
GMS–WPI ice cream still mainly comprised of whey Gelin, J. L., Poyen, L., Courthaudon, J. L., Le Meste, M., & Lorient,
D. (1994). Structural changes in oil-in-water emulsions during the
proteins, but a few partially coalesced fat globules could manufacture of ice cream. Food Hydrocolloids, 8, 299–308.
be found directly attached to the air interfaces (Fig. 4b). Goff, H. D., & Jordan, W. K. (1989). Action of emulsifiers in
promoting fat destabilization during the manufacture of ice cream.
Journal of Dairy Science, 72, 18–29.
4. Conclusions Goff, H. D., Liboff, M., Jordan, W. K., & Kinsella, J. E. (1987). The
effects of polysorbate 80 on the fat emulsion in ice cream mix:
evidence from transmission electron microscopy studies. Food
Fat destabilization and the composition of air inter- Microstructure, 6, 193–198.
face in ice cream depended on the sources of emulsifiers Goff, H. D., Verespej, E., & Smith, A. K. (1999). A study of fat and air
and proteins. Unsaturated monoacylglycerol was more structures in ice cream. International Dairy Journal, 9, 817–829.
Govin, R., & Leeder, J. G. (1971). Action of emulsifiers in ice cream
powerful than saturated monoacylglycerol at displacing
utilizing the HLB concept. Journal of Food Science, 36, 718–722.
caseins from the fat interface and brought more partially Lin, P. M., & Leeder, J. G. (1974). Mechanism of emulsifier action in
coalesced fat to air interface, giving the final product an ice cream system. Journal of Food Science, 39, 108–111.
better melting resistance. Emulsifiers increased air Pelan, B. M. C., Watts, K. M., Campbell, I. J., & Lips, A. (1997). The
bubble stability through promoting partial coalescence stability of aerated milk protein emulsions in the presence of small
of fat, but too much partially coalesced fat led to molecule surfactants. Journal of Dairy Science, 80, 2631–2638.
Segall, K. I., & Goff, H. D. (1999). Influence of adsorbed milk protein
unstable air bubbles, as was seen during the melting of type and surface concentration on the quiescent and shear stability
ice cream. of butteroil emulsions. International Dairy Journal, 9, 683–691.
When whey protein was used as the only source of Van Camp, J., Van Calenberg, S., Van Oostveldt, P., & Huyghebaert,
protein, ice cream emulsions were too stable to undergo A. (1996). Aerating properties of emulsions stabilized by sodium
caseinate and whey protein concentrate. Milchwissenschaft, 51,
desirable fat destabilization. Fat globules did not
310–315.
adsorb to the air interface to the extent that they did Zhang, Z., & Goff, H. D. (2004). Protein distribution at air interfaces
when SMP was the protein source. Saturated mono- in dairy foams and ice cream as affected by casein dissociation and
acylglycerol was slightly more effective than unsaturated emulsifiers. International Dairy Journal, 14, 647–657.