My+Articles+LSF


 * Effects of overrun on structural and physical characteristics of ice cream**

Abstract
Air is an important component in ice cream, affecting both physical properties and storage stability. The objective of this study was to measure the effects of air incorporation in ice cream, at overrun levels of 80%, 100% and 120%, on the growth of air cells and ice crystals, as well as on the hardness and melt-down rate of the product. Ice creams with different overruns were stored either in bulk containers (at −10°C with normal refrigeration cycling) or on microscope slides (at −6°C, −10°C or −20°C) for analysis. In bulk storage, mean air cell size initially increased during hardening, decreased during the early stages of storage and ultimately increased to larger sizes at longer (up to 3 months) storage times. Initial air cell size was smaller in ice creams with higher overrun, potentially due to the higher shear stresses during manufacture. Ice creams with lower overruns (80%) were harder than those made with 120% overrun but melted more rapidly. For samples stored on the microscope slide, lower storage temperature (−20°C) limited the mobility and solubility of air cells within the serum phase so that disproportionation was inhibited and primarily coalescence occurred for air cells in close proximity. At high storage temperature (−6°C or −10°C), disproportionation and coalescence were enhanced due to the higher mobility of the serum phase. Higher overrun led to slightly more stable air cells during storage.

Air in ice cream provides a light texture and influences the physical properties of melt down and hardness. However, it is not just the amount of air incorporated, or overrun, but also the distribution of sizes of the air cells that influences these parameters. The manufacture of high quality ice cream requires careful control of both overrun and air cell size distribution. However, very little attention has been paid to the air cell size distribution during manufacture and storage of ice cream. There are numerous factors that influence development of air cells in ice cream (Marshall & Arbuckle, 1996). Upon freezing, shearing forces during mixing break larger bubbles into smaller ones. To break air cells down into small sizes, a high local shear stress is required. This shear stress is governed by the mixing impeller and the viscosity of the ice cream slurry as it is forming and is related to the amount of ice formed and the viscosity of the continuous (freeze concentrated) phase.
 * Introduction**

As new air cell surface is formed during freezing of ice cream, it must be stabilized in some way to prevent coalescence. In electron microscope images, the air–serum interface contains numerous fat globules (both singly and in clusters) with a smooth surface between the fat globules. The partially coalesced fat is thought to form a network within the serum that separates air cells and keeps them from recombining (Walstra, 1989; [Flores and Goff (1999a) suggested that the amount of air cells at (2002a)] and [Chang & Hartel (2002b)] ). The presence of ice crystals and a viscous serum phase also help to stabilize the air cells. According to Wilson (1989), coalescence is also enhanced when air cells move close together by Brownian motion. In principle, coalescence involves two or more bubbles and results in coarsening of the foam structure. Thus, when coalescence occurs, the air cell size distribution is shifted to larger sizes (Ronteltap & Prins, 1989). Moreover, it is usually air cells larger than about 20 μm that are involved in coalescence since larger air cells have greater interfacial area. The probability of coalescence is reduced as temperature decreases due to the increased viscosity of the serum phase. Chang and Hartel (2002b) showed that partial coalescence was observed at low temperatures and during the latter stages of storage, leading to formation of irregularly shaped air cells that led to a network of interconnected air pockets. Generally, drainage involves the rise of air cells and subsequent downward flow of serum phase due to gravity. The larger the air cell, the faster it rises. Drainage by itself does not change the air cell distribution (Lees, 1991), rather it changes the film thickness between the air cells (Ronteltap & Prins, 1989) and promotes coalescence. Increasing the viscosity of the serum phase, which may be achieved by addition of stabilizer (Aguilera & Stanley, 1999) or by decreasing storage temperature (Chang & Hartel, 2002b), are ways to retard drainage (Walstra, 1989).

Disproportionation, which occurs due to differences in Laplace pressure between air cells, may also be controlled by increasing viscosity of the serum phase and forming a thick film on the surface of the air cells. However, Chang and Hartel (2002b) observed that an increase in viscosity of the serum phase and an increase in fat destabilization were not sufficient to prevent disproportionation from occurring. Rather, the rate of disproportionation was decreased as temperature decreased from −6°C to −28°C. In general, the existence of a difference in pressure between air cells promotes the diffusion of air from the smaller to the larger cells. Thus, disproportionation results in a bimodal distribution with both smaller and larger air cells being present in the ice cream (Ronteltap & Prins, 1989). In the early stages of disproportionation, a net decrease in mean size may be observed. As more air cells disappear, however, a gradual increase in the mean size may be observed with time (Lees, 1991). The amount of air incorporated during freezing affects the size of the ice crystals, with larger ice crystals observed at lower overrun (Arbuckle, 1977). Flores and Goff (1999a) suggested that the amount of air cells at 70% overrun is just enough to prevent collisions among ice crystals and to disperse the serum phase around each crystal. Lower overruns (and subsequently larger ice crystals) in ice cream lead to decreased hardness, (Prindiville, Marshall, & Heymann, 1999; Abd El-Rahman, Madkor, Ibrahim, & Kilara, 1997). In contrast, Wilbey, Cooke, and Dimos (1997) found that the presence of air (high % overrun) decreased the hardness of ice cream. Thus, contradictory results relating air content and hardness in ice cream have been observed, most likely due to differences in secondary effects (ice crystals, etc.). According to Pelan, Watts, Campbell, and Lips (1997), it is the stability of air cells that slows down the meltdown rate of ice cream, as determined by adding saturated monoglycerides to the mix. Campbell and Pelan (1998) found that melt-down resistance increased as draw temperature from the freezer decreased due to increased overrun and fat destabilization, although ice crystals may also have influenced the melt-down rate. Although much is known about the effects of air on the properties of ice cream, contradictory results may be due to secondary effects (e.g., ice crystals). Furthermore, the effects of overrun on the air cell size distribution and the subsequent effects on the physical properties of the ice cream have not been studied. Here, the effects of overrun and the air cell size distribution on the structural and physical properties of ice cream are studied in more detail.

**Production and storage of ice cream**
All ice creams were manufactured in the Babcock Hall Dairy Plant at the University of Wisconsin-Madison. The dry ingredients were mixed with liquid ingredients in the mix vat, heated to 38°C, pasteurized (APV Crepaco, Tonawanda, NY) for 19 s at 82°C, homogenized in a 2-stage homogenizer (6.89×103 and 2.43×104 kPa, respectively), and cooled and aged at 1–2°C for 24 h. After aging, vanilla flavor was added and the mix was frozen to −5.5°C to −6°C in a continuous, scraped-surface freezer (Waukesha Cherry Burrell, Model WS106GS, Louisville, KY). During freezing, air was injected under pressure to the desired overrun, and the ice creams were drawn from the freezer into 0.0019 m3 (half-gallon) cylindrical containers. Ice creams with 80% overrun were first produced, followed by those with 100% and 120% overruns, respectively. Ice cream samples were hardened in an air-blast freezer at −27°C for 24 h. After hardening, ice creams were transferred to a storage freezer at −30°C for 8 days before two containers of each ice cream were placed in a storage freezer held at −10°C with normal temperature variations due to cycling of the mechanical refrigeration system. The ice cream samples were stacked tightly in two rows in the storage freezer at −10°C. Ice creams were held for up to 3 months for analyses. In addition to bulk storage, ice cream samples immediately from the draw of the freezer also were stored on microscope slides at −6°C, −10°C and −20°C to follow changes to individual air cells. A small amount of ice cream was placed on a depression in a microscope slide, carefully spread into a thin layer, covered with a glass cover slip and sealed to prevent moisture loss. Samples on the microscope slide were made in duplicate for each overrun level and storage condition.

2.3. Analyses
Measurements of ice crystal and air cell size distributions were performed in a custom-built refrigerated glove box (Donhowe, Hartel, & Bradley, 1991) with a light microscope (Model FS-35DX, Nikon, Inc., Garden City, NY) located within. All utensils, including microscope slides and cover slips, needed to prepare samples were held in the glove box prior to use to prevent melting. For storage of bulk containers, core samples of ice cream (in duplicate) were taken 6 cm from the surface and 4 cm from the sides of the cups after drawing, after hardening, once every 5 days during the 1st month, once a week during the 2nd month and once every 2 weeks during the 3rd month of storage. Ice cream samples upon drawing from the continuous freezer were taken 10 min after the 1st cup was drawn from the freezer. According to Chang (2000), at least 250 or more air cells must be observed to ensure statistical analysis. Thus, 2–4 microscope slides of each sample were prepared to get at least 300 individual air cells and ice crystals for analysis. For calibration of size determination at a later time, a micrometer ruler (Nikon) was also imaged at the same time. For air cell analysis, a small amount of ice cream was placed in a depression between 2 cover slips arranged on a slide. A cover slip was placed over the top of the sample and pressed gently with tweezers. Duplicate samples were prepared for air cell analysis. The temperature of the glove box was set to −6°C to slightly melt the ice cream and allow observation of the air bubbles arising from the sample (Chang, 2000). The magnification used for observations of air cells was 40×. Chang and Hartel (2002c) have shown that this method gives air cell sizes consistent with those measured using the more laborious cryo-stage electron microscopy method. For ice crystal analysis, 3–5 drops of butanol were added to a small amount of ice cream placed on a similar microscope slide in order to disperse the crystals without melting or dissolving (Donhowe, 1993). A cover slip was pressed on top of the ice cream and gently pressed back and forth with tweezers to disperse the sample into a thin layer. By doing this, the air cells were destroyed, but the ice crystals were spread without overlapping. The temperature of the glove box was set to −10°C for observation of ice crystals. The magnification used for observation of ice crystals in fresh-drawn and hardened and cycled ice creams were 80× and 40×, respectively. In order to observe the change in size of air cells and ice crystals, the diameter of each was determined from the projected area in the micrographs. The perimeter of each air cell and ice crystal was traced manually on a digitizing board (Summasketch Professional, Summagraphics Corp., Fairfield, CT) connected to image analysis software (Sigma-Scan, Jandel Scientific, Sausalito, CA) on a personal computer (IBM TX, Los Angeles, CA). Each projected area was converted to an equivalent circular diameter (the diameter of a circle with the same area). Mean size and distribution statistics for both air cells and ice crystals were calculated with a custom-written macro in Excel (Microsoft Corp., Redmond, WA). Melt-down rates of duplicate ice cream samples were measured in a controlled temperature (30°C) chamber. Ice cream samples were tempered in a freezer at −14°C for 24 h prior to analysis. For melt-down rate, 160 g of ice cream were placed on a 2-mm stainless-steel screen with a funnel and graduated cylinder beneath to collect the melt. Timing of the melt-down rate began when the first drop of melt (after 30–60 s) touched the bottom of the cylinder. Volumes were recorded once every 10 min for 60 min. Hardness of ice creams after 8 days at −30°C (before cycling) was measured by using a Precision Penetrometer (Precision Scientific Co., Chicago, IL) with a 60° Cone probe. Before the measurements were taken, the ice creams were held at −14°C for 24 h. Ice cream containers (in duplicate) were cut 6 cm from the bottom to allow the bottom sections to be used for hardness measurement. Penetration of the probe was conducted 4 cm from the side of each cup and was repeated six times (three times for each of two containers). Hardness was measured as the depth (in mm) of penetration of the cone into the ice cream after the cone was released for 2 s at room temperature. Three measurements were made on each of the two duplicates and the results averaged. Statistical analyses (standard //t//-tests) were conducted for the results (mean values) by using the SAS Program (SA Institute Incorporation, Cary, NC). Statistical significance is given in terms of //p//-values, with differences at the 95% confidence interval (//p//<0.05) being considered statistically significant.

3.1. Storage in bulk containers
During manufacture and through the early stages of storage in bulk containers, the mean air cell size, in general, decreased (//p//<0.05) with increased overrun, as shown in Fig. 1. An increase in overrun promoted break-up of larger air cells into smaller ones during freezing so that a higher percentage of smaller air cells was observed in ice cream made with 120% overrun than in ice cream made with 80% overrun. This effect may have been related to an increase in apparent viscosity of the ice cream slurry during freezing as more air was incorporated. Chang and Hartel (2002a) showed that higher apparent viscosity led to more efficient break down of air cells and to smaller air cell sizes.

[|Full-size image] (30K) Fig. 1. Comparison of mean air cell sizes in ice creams with 80%, 100%, and 120% overrun upon drawing (draw), after storage at −30°C for 8 days (hardened), after hardening and storage at −10°C for 3 days (3 days cycled), 9 days (9 days cycled), 2 weeks (2 weeks cycled), 4 weeks (4 weeks cycled), 8 weeks (8 weeks cycled), and 9 weeks (9 weeks cycled).

View Within Article Upon hardening, the mean air cell size in each ice cream had increased although the increase was dependent on the overrun. Ice cream made with 80% overrun had the largest increase in mean size, with the ice cream made with 120% overrun exhibiting the smallest increase. Upon storage at −10°C, mean air cell size decreased in all ice creams and continued to decrease for up to 4 weeks in the ice cream made with 80% overrun. An increase in temperature at constant pressure should result in an increase in volume (size) based on the ideal gas law. Thus, this decrease during warm temperature storage must be due to changes in the overall distribution of air cell sizes (see below). After this initial period of reduction in size, the air cells began to grow once again. Over longer terms of storage at −10°C, the air cells increased in size until, after nine weeks of storage, mean air cell sizes were about the same regardless of the overrun content. Overruns of 100% and 120% generally showed the same trends, although not as distinctly. The overrun level is thought to influence ice crystal formation through secondary effects (Arbuckle, 1977). In this study, the mean ice crystal size generally decreased with increasing overrun (Fig. 2), although the effect was relatively small. Upon drawing from the freezer, ice creams made with 80% and 100% overrun had about the same mean ice crystal size (//p//=0.869), with the mean size being significantly smaller in the ice cream made with 120% overrun (//p//=0.033 for comparison with 80% and //p//=0.0423 for comparison with 100% overrun). However, after storage at −30°C for 8 days (labeled “hardened” in Fig. 2) and storage (at −10°C), the ice cream made with 80% overrun generally had the highest mean ice crystal size. That is, the ice cream with lowest overrun showed the highest recrystallization, although the effects were relatively small. These results corroborate those reported in Arbuckle (1977) and found by Flores and Goff (1999b) who showed that increased overrun caused a slight decrease in mean ice crystal size. These effects may be related to the change in heat transfer rates from the ice cream upon increased aeration.

[|Full-size image] (19K) Fig. 2. Comparison of mean ice crystal sizes in ice creams with 80%, 100%, 120% overrun. See Fig. 1 for description of storage conditions.

View Within Article Hardness of ice creams was measured after 8 days at −30°C. Ice creams made with 80% and 100% overrun had the same (//p//=0.9735) hardness (7.5±0.7 mm) and these were significantly harder than the ice cream made with 120% overrun (9.8±0.9 mm) (//p//=0.0003). These results confirmed those of Wilbey et al. (1997), where increased overrun caused a decrease in hardness of ice cream. However, Prindiville et al. (1999) and Abd El-Rahman et al. (1997) found the opposite effect. Perhaps other factors (air cell and ice crystal size distributions) had more effect on hardness than total overrun in these studies. Melt down rates of ice creams were also measured after 8 days of storage at −30°C. Ice cream made with 80% overrun melted more rapidly than those made with 100% and 120% overrun (Fig. 3). No significant differences (//p//>0.05) were observed between the volume of melted ice cream with 100% and 120% overrun. These results may be explained by the differences in fat destabilization or air cell and ice crystal sizes. Although the extent of fat destabilization was not measured in this study, it is possible that higher shear stress occurred with higher overrun (higher air content leading to higher apparent viscosity), leading to greater fat destabilization and thus slower melt down. Furthermore, the ice cream made with 80% overrun had larger air cells and ice crystals after hardening (Fig. 1 and Fig. 2) than the ice creams made with 100% and 120% overrun, and perhaps these factors influenced melt down rates. Another potential cause of the slower melt down with higher overrun may be the difference in heat transfer rate due to the greater presence of air. Air is a good insulator and undoubtedly slowed the rate of heat transfer into the ice cream with higher overrun.

[|Full-size image] (5K) Fig. 3. Meltdown of ice creams with 80%, 100%, and 120% overrun after storage at −30°C for 8 days and 1 day at −14°C.

View Within Article

3.2. Storage on microscope slides
To study the effects of storage temperature and overrun on the changes in air cell size distribution, ice cream directly from the exit of the scraped-surface freezer at −6°C was held on a microscope slide for up to 3 days at temperatures of either −6°C, −10°C or −20°C. Air cell size distributions after 3 h are compared in Fig. 4 for ice cream with an overrun of 80%. At all temperatures, there was a significant increase in mean size after 3 h (Fig. 5). At draw, the mean air cell size for the ice with 80% overrun was 22.9±0.6 μm. Mean air cell size had increased to 28.7±2.0 (//p//=0.0203), 27.6±1.8 (//p//=0.0039), and 34.4±2.5 (//p//<0.001) μm after 3 h at −6°C, −10°C and −20°C, respectively. No difference (//p//=0.4556) between mean air cell size for 3 h at −6°C or −10°C was observed. However, the nature of the changes in the air cell size distribution was dependent on the storage temperature. At −6°C and −10°C, there was a significant shift of air cells to smaller sizes for the smaller air cells (less than about 30 μm) in addition to the shift to larger sizes for larger air cells (larger than about 30 μm), indicative of more than one mechanism for change in air cells. At –20°C, only an increase in air cell size, over the entire size range, was observed. The shift to smaller sizes indicates that disproportionation occurred when temperatures were below or at approximately −10°C and the shift to larger sizes indicates that coalescence occurred at the same time. However, only coalescence occurred for storage at −20°C. The net result was an increase in mean size for all conditions.

[|Full-size image] (5K) Fig. 4. Comparison of the air cell size distributions of ice cream with 80% overrun stored on thin slides at different temperatures (−6°C, −10°C, and −20°C) for 3 h.

View Within Article

[|Full-size image] (5K) Fig. 5. Comparison of the mean air cell sizes of ice cream with 80% overrun upon drawing from the continuous freezer at −6°C (fresh drawn), after storage on thin slides at different temperatures (−6°C, −10°C, and −20°C) for 3 h.

View Within Article The effects of overrun on the change in mean size after 3 h at −6°C on a microscope slide is shown in Fig. 6. Significant (//p//<0.05) increases in mean size were seen for ice creams with 80% and 120% overrun, but not for the ice cream made with 100% overrun. The changes in air cell size distribution for these conditions are shown in Fig. 7. At 80% overrun, both disproportionation and coalescence occurred since the number of small air cells decreased at the same time the number of large air cells increased. At 100% overrun, no significant changes in air cell size distribution were observed and at 120% overrun, air cell size increased for the entire size range (no disproprortionation).

[|Full-size image] (7K) Fig. 6. Comparison of the mean air cell sizes of ice creams with 80%, 100%, and 120% overrun upon drawing from the continuous freezer at −6°C (draw) and after storage on thin slides at −6°C for 3 h.

View Within Article

[|Full-size image] (4K) Fig. 7. Comparison of air cell size distributions of ice creams with 80%, 100%, and 120% overrun upon drawing from the continuous freezer at −6°C (draw) and after storage on thin slides at −6°C for 3 h (3 h).

View Within Article Extended storage time on the microscope slide led to further changes in air cell size distribution. As noted previously for storage of ice cream with 80% overrun at −10°C for 3 h, some disproportionation occurred as there was a shift in the air cell size distribution to smaller (less than 30 μm) sizes. However, the main mechanism of change for longer times (up to 3 days) was coalescence, with an increase in numbers of air cells at all sizes observed. Interestingly, at 3 days of storage at −10°C, the air cell size began to decrease, in a manner similar to that observed in the bulk storage samples described in the previous section. The slight decrease in mean size from 24 h to 3 days of storage was observed as an increase in air cell numbers for small and intermediate sizes (up to abut 50 μm) along with a slight decrease in number of air cells greater than about 50 μm. At −20°C, changes in mean air cell size were observed for all levels of overrun (Fig. 8). At 80% overrun, mean air cell size was initially the largest and increased the most dramatically during storage. For the ice cream with 100% overrun, the mean air cell size increased slowly over time. At 120%, the ice cream had the smallest mean air cell size at the draw and remained small throughout the storage period. Air cell size distributions (not shown) showed that coalescence was the main mechanism of change in air cells with very little evidence of disproportionation at any level of overrun. This is not surprising since the mobility of gas molecules in the ice cream matrix at −20°C is greatly inhibited.

[|Full-size image] (12K) Fig. 8. Comparison of the mean air cell sizes of ice creams with 80%, 100%, and 120% overruns upon drawing from the continuous freezer at −6°C (draw), after storage on the thin slides at −20°C for 3 h, 24 h, and 3 days.

View Within Article

4. Conclusions
Increasing the overrun in ice cream (from 80% to 100% or 120%) led to formation of slightly smaller air cells and ice crystals, probably due to the higher shear stresses exerted in the freezer barrel due to the higher air content. Two different mechanisms led to changes in air cells, disproportionation and coalescence. Both mechanisms were observed during storage in both bulk containers and in sample sections on microscope slides. These structural changes as overrun increased resulted in ice creams that were slightly softer (higher penetration depth) and slightly more resistant to melt down (slower melting rate).