SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

Butter and other milk fat-based products are valuable products for the dairy industry due to their unique taste, their textural characteristics, and nutritional value. However, an increased consumer demand for low-fat-based products increases the need for an increased essential understanding of the effective factors governing the structure of milk fat-based products. Today, 2 manufacturing techniques are available: the churning method and the emulsification method. The first is typically used for production of butter with a globular structure, which has become increasingly popular to obtain low-fat-based products, typically without presence of milk fat globules. The microstructure of milk fat-based products is strongly related to their structural rheology, hence applications. Structural behavior is not determined by one single parameter, but by the interactions between many. This complexity is reviewed here. Parameters such as thermal treatment of cream prior to butter making, water content, and chemical composition influence not only crystal polymorphism, but also the number and sizes of fat crystals. The number of crystal–crystal interactions formed within the products is related to product hardness. During storage, however, postcrystallization increases the solid fat content and strengthens the fat crystal network. The fat crystal network is strengthened by the formation of more and stronger crystal–crystal interactions due to mechanically interlinking of fat crystals, which occurs during crystal growth. Postcrystallization is directly linked to chemical composition. The initially observed microstructural difference causing different rheological behavior will disappear during storage due to postcrystallization and formation of more crystal–crystal interactions.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

Butter and other milk fat-based products are economically important to the Danish dairy industry with an export value of 1.73 billion DKK or 231.90 million EUR in 2011 (LF 2011). This corresponds to a market share of 4.9% of the global export of butter and butteroil (2005 values; AGR 2006). During the last decades, the variety of milk fat-based products has increased in response to the increased demand for low-fat-based products fulfilling the consumers’ need for functionality and nutritional value. A recent field experiment has shown that consumers are willing to pay more for butter with good nutritional content (Saulais and Ruffieux 2012). However, the 1st step to develop such products is to gain an essential understanding of the physical properties of the fat, as they affect not only flavor perception and health characteristics, but also the underlying material functionality and rheological characteristics of the products, such as spreadability, hardness, and mouthfeel (Narine and Marangoni 1999a; Marangoni and others 2012).

Spreadability is a rheological phenomenon determined by the chemical composition, solid fat content (SFC), and functionality of the fat crystal network (deMan and Beers 1987; Campos and others 2002; Tang and Marangoni 2007). The properties of the fat crystal network depend on the interactions between the crystals. Crystal size, shape, and polymorphic structure, as determined by processing conditions and chemical composition, affect the crystal–crystal interactions. This review primarily focuses on published information addressing how chemical composition, manufacturing, water content, and storage conditions affect the structure of butter and other milk fat-based products, published within the last 10 y. The most recent review dealing with this topic dates back to 2001 (Wright and others 2001). Other recent reviews within related areas focus more specifically on the structural behavior of lipid shortenings (Ghotra and others 2002), the relationship between morphology and the physical properties of fat crystal network (Tang and Marangoni 2006), and the effect of a reduced amount of saturated fat on the rheological properties of fat-based products (Berger and Idris 2005; Pernetti and others 2007; Rogers 2009; Wassel and others 2010). Also, the effect of minor components and other additives on fat crystallization in milk fat-based products have been reviewed (Smith and others 2011). Moreover, the fundamental aspects of triglyceride crystallization have recently been reviewed (Acevedo and Marangoni 2010; Sato and Ueno 2011; Marangoni and others 2012).

Milk fat

Milk is secreted from the mammary glands of female mammals, containing all the essential nutrients (Riccio 2004) except for iron and vitamin C, and is a widely used food ingredient due to its unique flavor, high nutritional value, and rheological properties. Milk fat occurs naturally in milk and cream, forming an oil-in-water emulsion. In the emulsion, milk fat is contained within milk fat globules, surrounded by the milk fat globule membrane (Figure 1). The milk fat globule membrane, acting as an emulsifier, stabilizes the fat globules in the emulsion by lowering the surface tension. In addition, it protects the fat from chemical reactions such as hydrolysis and oxidation. The globule membrane consists primarily of phosphorous and glycolipids (30%) together with proteins (41%). In addition, the membrane contains cholesterol (2%), natural glycerides (14%), and water (13%) (Belitz and others 2009). Milk fat, on the other hand, contains a wide range of triglycerides, with a broad thermal range of melting points ranging from –50 to 80 °C (Table 1). The triglycerides consist of a glycerol backbone to which 3 fatty acids moieties are attached (Figure 2). The fatty acid composition for milk, as reported by Couvreur and others (2006), is listed in Table 1 and the corresponding melting points have been added (Knothe and Dunn 2009). For a detailed description of the triglyceride composition in milk fat, the reader is referred to a study by Gresti and others (1993), identifying 223 individual molecular species of even-numbered triglycerides. In addition to the triglycerides forming 97.5 weight% of the milk fat, diglycerides (0.4%), monoglycerides (0.03%), phospholipids (0.6%), cholesterol (0.3%), and traces of fat-soluble vitamins are also present (Walstra 1999; Fox and McSweeney 2006).

Table 1. Fatty acid (FA) composition (g/100g, grams of total fatty acid) of milk from cows fed with either corn silage or fresh grass, as further described by Couvreur and others (2006), from where data are adapted. The right column lists the literature data for melting points of the fatty acids (Knothe and Dunn 2009)
Fatty acidFatty acid structureCorn silage (%) Couvreur and others (2006)Fresh grass (%) Couvreur and others (2006)Melting point (°C)
  1. a

    Vaclavik and Christian 2008;

  2. b

    Sigma 2013;

  3. c

    Weast 1985;

  4. d

    Lide 1999;

  5. e

    Gunstone and others 2007;

  6. f

    Knothe and Dunn 2009.

ButyricC4:04.33.9−7.9a
HexanoicC6:02.52.2−4b
CaprylicC8:01.51.316.5c
CapricC10:03.33.131.5c
CaproleicC10:10.270.24
UndecanoicC11:00.030.0528.6c
LauricC12:03.73.444c
TridecanoicC13:00.090.1441.5d
MyristicC14:011.810.958c
MyristoleicC14:10.870.83−4d
PentadecanoicC15:01.482.3153–54c
PalmiticC16:031.324.363c
PalmitoleicC16:11.921.6832–33e
HeptadecanoicC17:01.251.6762–63
StearicC18:010.311.271.2c
OleicC18:122.5928.1512.82–52.38f
LinoleicC18:22.283.52−7.15f
LinolenicC18:30.220.70−11.48f
ArachidicC20:00.140.0277c
EicosatrienoicC20:30.100.12
ArachidonicC20:40.100.10−4949b
BehenicC22:00.0260.05879.54f
DocosapentaenoicC22:50.0790.0538−20b
image

Figure 1. Schematic illustration of the microstructure of cream and milk. To the left is cream, an oil-in-water emulsion, with the continuous water phase being blue and the milk fat globules dispersed within the water phase. Butter is obtained by churning of the cream, facilitating a phase inversion to a water-in-oil emulsion (middle image). The microstructure of the butter is illustrated by a white continuous fat phase with water droplets (blue) and milk fat globules dispersed within together with fractions of ruptured milk fat globule membranes. The fat crystals are present both in the continuous fat phase as well as in the milk fat globules. Upon storage, the solid fat content increases and a more dense fat crystal network is formed (right image).

Download figure to PowerPoint

image

Figure 2. Triglyceride structure with the glycerol backbone marked in red. The 3 fatty acid moieties are denoted R1, R2, and R3.

Download figure to PowerPoint

The chemistry of milk fat is thereby very complex, and is affected by the feed and breed of the cow, the state of lactation, the production process, and also the interactions within the milk itself (Couvreur and others 2006; Buldo and others 2012). The proportion of fresh grass in the feed appears to be linearly correlated with a decreasing amount of milk fat: the more fresh grass, the less fat and the smaller the milk fat globules. When the proportion of grass contributes 30% or more of the total feed, the average size of the milk fat globules slightly decreases to 3.28 μm (3.50 μm below 30% fresh grass) and the amount of unsaturated fatty acids increases at the expense of saturated fatty acids (Couvreur and others 2006). The unsaturated fatty acids play an important role in the physical as well as nutritional functionality of the milk fat (Marangoni and others 2012). An increased degree of unsaturation of the fatty acid moieties and/or a decreasing chain length results in a lower melting point compared to saturated fat with a long chain. Such changes in the chemical composition of milk fat are likely to alter the rheological behavior of the butter products made from it.

Anhydrous milk fat (AMF) is mostly used for recombination of various dairy products, but has applications in chocolate and ice cream production. AMF is produced from fresh cream or butter by removal of water, proteins, and other minor components. It contains, by definition, a minimum of 99.8% milk fat, of which at least 98.8% is triglycerides and not more than 0.1% water (World Health Organization 2010). Two other qualities of milk fat exist: anhydrous butter oil that also contains a minimum of 99.8% milk fat, but can be made from cream and butter of different ages, and butteroil that must contain 99.3% milk fat. Butteroil can also be obtained from cream and butter of various ages (FAO 1983). AMF is often fractionated into 3 fractions: the high-melting fraction, the middle-melting fraction, and a low-melting fraction, as seen in Figure 3.

image

Figure 3. Thermogram obtained upon melting of anhydrous milk fat (AMF) using Differential Scanning Calorimetry. AMF was cooled from 70 to –15 °C at 7.5 °C/min and heated at a scan rate of 4 °C/min. From the thermogram, the characteristic high-, low-, and middle-melting fraction of the AMF can be identified.

Download figure to PowerPoint

Milk fat-based products

Milk fat-based products, such as butter, are water-in-oil emulsions consisting of crystallized fat in a continuous phase, in which water droplets (2.3–10.6 μm) (van Dalen 2002; van Lent and others 2008), milk fat globules (2.5 μm) (Rønholt and others 2012a), and partially damaged fat globules are dispersed (Figure 1 and 4). The size distribution of the water droplets is important in respect to microbiological growth, and chemical and physical stability together with organoleptic properties (Frede and Buchheim 1994; Delmarre and Batt 1999; van Lent and others 2008). Large water droplets favor aroma and flavor effect due to a delayed oxidation of the fat phase as a consequence of the reduced fat/water interface (Frede and Buchheim 1994). However, as large water droplets can serve as growth zones for bacterial growth, their size must be kept as small as possible (Frede and Buchheim 1994; Delmarre and Batt 1999; van Lent and others 2008).

image

Figure 4. Images of butter structure obtained using confocal laser scanning microscopy. The continuous fat phase is the red/purple color, while water droplets are green (marked with green arrow). As fat crystals are color-negative they appear as dark black/gray shadows (shown in the right image with a white arrow). The phospholipids surrounding the milk fat globules are blue (shown in the left image with a red arrow). The image to the left is focused on the milk fat globule membrane, while the right image is optimized to visualize the fat crystal network. The liquid fat within the milk fat globules can be observed in the right image, as they appear as bright red spherulite-shaped objects (shown with a red arrow in the right image).

Download figure to PowerPoint

The fat crystal network is held together by crystal–crystal interactions, characterized by strong irreversible (primary) bonds and weak reversible van der Waals (secondary) bonds (Haighton 1965). The irreversible are formed upon crystal growth, as the crystals get mechanically interlinked (van den Tempel 1958). Most often, the rheological properties of the fat crystal network are characterized using rheological measurements (Wright and others 2001).

Structural characteristics of milk fat-based products

The amount of the different phases is dependent on production parameters such as heat treatment of the cream, degree of processing, and the microstructure (Juriaanse and Heertje 1988). The ratio between liquid and solid fat is of utmost importance for the rheological properties of butter and spreads. Without solid fat, a milk fat-based product is fully liquid. Without liquid fat, it would appear hard and brittle (Narine and Marangoni 1999a). Marangoni and others have studied the relationship between SFC, hardness, and textural behavior in milk fat-based products (Narine and Marangoni 1999a,b, 2001; Awad and others 2004; Batte and Marangoni 2005; Marangoni and Ollivon 2007) and cocoa butter (Marangoni and McGauley 2003). The properties of a fat crystal network can be quantified by the elastic modulus, G’ (Narine and Marangoni 1999a, 2001). Moreover, the relationship between SFC and G’ can describe the microstructure of a fat crystal network using the fractal dimension (Narine and Marangoni 1999b; Awad and others 2004; Tang and Marangoni 2006), which describes the microstructures within the fat crystal network (Meakin 1988). However, other studies have shown that even though the SFC is the same, fat can have very different rheological properties (Haighton 1965; Shama and Sherman 1970; Narine and Humphrey 2004; Rønholt and others 2012a,b).

Fat crystallization

Chemical composition

The rheological properties of fat-based products are influenced by the fat crystal structure, solidification, and transformation behavior (Sato 1999; Mazzanti and others 2004). The fat crystals consist of triglyceride molecules (Figure 2), forming nanoscale structured elements (150- to 350-nm long and 10- to 60-nm thick) assembled into network of fat crystals (20 to 100 μm) (Acevedo and Marangoni 2010). In naturally occurring triglycerides, such as those present in milk fat, the 3 fatty acid moieties are most often of different chain length, resulting in a potentially asymmetrical structure generating stereoisomers of the triglyceride derivatives (Sato and Ueno 2001). The degree of symmetry within the triglycerides is related to the free edge energy, hence the crystallization mechanism. A triglyceride with a high structural symmetry results in spherulitically grown long needle-like crystals, while a lower degree of symmetry is related to lozenge-shaped crystals (Hollander and others 2003). A microstructure dominated by numerous small crystals is shown to increase the hardness by up to 20% compared to a microstructure with few, but large crystals, a difference that from consumer perspective, would be noticed by different mouthfeel, brittleness, and spreadability (Fedotova and Lencki 2010). For milk fat, the broad range of triglycerides results in different polymorphic forms due to varying chain length and degree of saturation. Consequently, the presence of mixed fatty acids might result in segregation during either solidification or aging (Sato 1999). The chemical composition of the fat is therefore likely to affect the rheological behavior of the fat crystal network. However, even though the chemical composition is identical, 2 very similar milk fat-based products can have different crystal structure (polymorphism), depending on thermal and mechanical treatment. The term polymorphism describes the type of lateral packing of the aliphatic triglyceride chains (Sato 1999; Mazzanti and others 2004). The lateral packing is influenced by factors such as cooling rate, agitation temperature, and the mechanical treatment (Heertje and others 1988; Heertje 1993; ten Grotenhuis and others 1999; Sato 1999; Herrera and Hartel 2000a,b,c; Mazzanti and others 2004; Marangoni and Ollivon 2007).

Crystallization mechanism

For milk fat-based products, the crystallization mechanism differs between bulk milk fat, such as AMF, and emulsions, such as natural cream (Skoda and van den Tempel 1963; Walstra and van Beresteyn 1975; Fredrick and others 2011). The crystallization mechanism relevant for crystallization of milk fat is the nucleation and growth mechanism. Here, the 1st step in the crystallization process is the formation of nuclei. Depending on the surroundings, primary nucleation occurs either homogeneously or heterogeneously. If no foreign solid surface is present, homogeneous nucleation will occur, as is the case within intact milk fat globules or AMF. Heterogeneous nucleation, on the other hand, is favored by the presence of a foreign surface, such as the inner surface of the processing equipment, dust particles, or simply nontriglyceride components in the milk product. Upon phase inversion of cream or emulsification of spreads, foreign surfaces are introduced to milk fat previously isolated within the milk fat globules (for cream) or in the bulk state (for spreads). The newly formed interfaces will act as catalytic impurities, causing heterogeneous nucleation. In terms of the physical properties of the final product, the different crystallization mechanism does not seem to influence the final rheological behavior (Lopez and others 2001a,b; Fedotova and Lencki 2010).

Polymorphism

In milk fat, the 3 main crystal polymorphs are denoted α, β’, and β according to Larsson (1966) (Figure 5), in increasing order of stability (Lopez and others 2002; Mazzanti and others 2004). While the triglyceride chains pack hexagonally in the α-crystals, they are orthorhombic in the β’-crystals and triclinic in the β-form (Chapman 1962). Each polymorph is identified by its characteristic set of short and long d-spacings (interplanar spacing) of the crystal lattice (Vaeck 1960; Larsson 1966; ten Grotenhuis and others 1999), using X-ray diffraction (XRD). The short spacings range from 3 to 6 Å, while the long spacings range from 38 to 72 Å (ten Grotenhuis and others 1999). The long spacings are either triple-chain-length (3L), formed when the chemical properties of 1 or 2 of the 3 chain moieties are markedly different from the others, or double-chain-length (2L) when the chemical properties are identical or very similar (Sato and Ueno 2001) (Figure 5 bottom).

image

Figure 5. Illustration of fat crystallization in milk fat. While α-crystals (hexagonal subcell structure) form directly from the melt, β’-crystals (orthorhombic subcell) form either via recrystallization of α- to β’-crystals or directly from the melt. β-Crystals (triclinic subcell) are primarily formed via recrystallization from β’-crystals. The chain length structure of the fat crystals is in double or triple arrangement, depending on the chemical properties of the fat (modified figure from Rønholt and others 2012b).

Download figure to PowerPoint

Being the least stable polymorph with the lowest melting point, the α-crystals easily transform to either β’ or β, depending on the thermal treatment (Rousseau and others 2005). The transformation pathway is dependent on the thermodynamic stability, as expressed in Gibbs free energy. High activation energy is required to obtain the most stable crystal polymorph, the β-crystal. The driving force needed to overcome the activation energy is temperature-dependent. For crystallization to occur, the temperature of the liquid phase must be well below the melting point of the triglycerides, it must be subjected to super-cooling (Sato and Ueno 2001).

The polymorphic form, however, is important for milk fat-based products, as it is related to rheology and the textural behavior of the products. β-Crystals tend to form large, platelet-like crystals resulting in a grainy macroscopic structure (Madsen 1971; Sato 1999; Saadi and others 2012) with a sandiness-like mouthfeel (Madsen 1971; Sato 1999). The formation of β-crystals is preferred to be eliminated in milk fat-based products by using well-controlled thermal treatment during manufacturing and by manipulating chemical composition (Sato 1999; Rousseau and others 2005; Saadi and others 2012). A prerequisite to understand the rheological behavior of milk fat-based products is, therefore, a detailed knowledge of the crystal polymorphism combined with a macroscopic understanding of the interactions within the fat crystal network.

Chemical Composition

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

The chemical composition impacts the rheological behavior of milk fat in several ways. One factor is the chemistry of milk fat as defined by nature. But methods to modify the rheological behavior of milk fat-based products by either chemical interesterification, blending with vegetable oils, emulsifiers, or specific milk fat fractions have also been an area of interest for researchers during the last 50 y (Wright and others 2001; Pernetti and others 2007). Today, seasonal variation in milk quality can be avoided by using a standardized feeding regime year-round (Shi and others 2001; Couvreur and others 2006; Kamleh and others 2010; Buldo and others 2012). However, field studies have shown that seasonal variations in the feeding of cows influence the fat–water ratio in butter, ranging from 86:14 (fat:water) in winter time to 76:24 during summer (Kamleh and others 2010), and also crystallization behavior due to different triglyceride compositions (Shi and others 2001). Moreover, it is possible to relate different feeding regimes to the fatty acid composition of the milk fat (Buldo and others 2012). However, due to individual cow variation, it is complicated to obtain a clear correlation between feeding regime and melting behavior of cream (Buldo and others 2012). Nevertheless, the breed of the cow can be linked to differences in chemical composition of the milk fat (Shi and others 2001; Kamleh and others 2010).

Minor components

In addition to the triglyceride composition, milk contains other minor components such as phospholipids originating from the milk fat globule membrane. Those minor components are shown to have a significant influence on crystallization in milk fat-based products (Wright and others 2000a; Vanhoutte and others 2002; Mazzanti and others 2004; Fedotova and Lencki 2010). Mazzanti and others (2004) used synchrotron radiation to study the effect of polar lipids (2.9%) on polymorphism in AMF. The study shows that the presence of polar lipids affects both the crystallization process and the transformation of β-crystals. A similar conclusion was drawn by both Vanhouette and others (2002) and Wright and others (2000a), studying the effect of phospholipids on isothermal crystallization of milk fat. Using differential scanning calorimetry (DSC) and low-resolution nuclear magnetic resonance (LR-NMR), Vanhoutte and others (2002) found that addition of up to 0.07% phospholipids delays the crystallization onset. Moreover, phospholipids are shown to influence the microstructure of milk fat-based products (Fedotova and Lencki 2008) and the sensory properties of such products (Fedotova and Lencki 2010). At concentrations lower than 2%, phospholipids act as nucleation sites, hence increase the spherulite size of the crystals. However, for concentrations above 2%, the microstructure consisted of small, uniformly distributed fat crystals. Such concentrations of phospholipids increased the hardness as perceived orally and also measured using cone penetrometry (Fedotova and Lencki 2010). For more details on how the presence of minor components and additives affect the crystallization of fats, the reader is referred to a recent review by Smith and others (2011).

Chemical modification

Using enzymatic or chemical interesterification allows alteration of SFC, polymorphism, and elastic behavior, depending on the degree of interesterification (Zhang and others 2004; Zárubová and others 2010). The main drawback associated with chemical modification is, however, decreased flavor perception of the products (Rousseau and Marangoni 1999; Kontkanen and others 2011). An alternative way to obtain milk fat-based products with a “healthier” profile is the addition of vegetable oils and/or fats (Martini and Herrera 2008), as that decreases the overall melting point and increases the proportion of unsaturated fatty acids (Wright and others 2001). The addition of vegetable oils also lowers the cost compared to pure butter, and the butter flavor is to a large extend kept intact (Danthine 2012). A number of recent studies focus on the effect on crystallization and rheological behavior upon addition to AMF of vegetable oils such as rapeseed oil (Buldo and Wiking 2012; Danthine 2012; Kaufmann and others 2012a,b), palm oil (Danthine 2012), and canola oil (Wright and others 2005). Addition of such oils to a milk fat matrix decreases the hardness of the blends more than an order of magnitude (Figure 6) (Kaufmann and others 2012a; Buldo and Wiking 2012). The rheological effect is due to solubilization of fractionated milk fat in the oil phase (Wright and others 2005), causing a decrease in both SFC and melting point (Wright and others 2005; Buldo and Wiking 2012; Kaufmann and others 2012a). Yet, deviations from this behavior have been reported (Rousseau and others 1996; Danthine 2012). Upon blending of rapeseed oil and AMF at very low temperatures (5 °C), up to 20% of rapeseed oil can be accommodated in the fat matrix without affecting the rheological properties of the blend (Rousseau and others 1996). This behavior is likely ascribed to a strong network formation upon blending at such a low temperature (Buldo and Wiking 2012). The effect of adding a solvent such as canola oil, low-melting fraction of AMF, hexane, or ethyl acetate to fractionated milk fat has been studied in relation to crystallization behavior and rheological properties (Wright and others 2000b; Wright and others 2005). The studies show that the choice of solvent affects both SFC and the induction time for crystallization to occur. The effect of solvent on crystal polymorphism is not unambiguously determined. While Wright and others (2000b) did not observe any polymorphic transition due to solvent addition, Danthine (2012) observed polymorphic changes in fractionated milk fat upon RO and palm oil addition. It appears that mixtures of palm oil (40% to 80%) and the middle-melting fraction of AMF or palm oil (40% to 90%) and AMF are more β-tending, compared to blends containing either the very low-melting AMF fraction, the low-melting fraction or the high-melting fraction that are all β’-tending (Danthine 2012). Similarly, addition of hydrogenated cottonseed oil to AMF or addition of coconut oil to zero-erucic rapeseed oil or palmstearin has been shown to enhance the β’-tending behavior of a fat-based crystal network (Rousseau and others 2005; Piska and others 2006). Care must be taken to compare the obtained results directly, as chemical composition, induction time, thermal treatment, crystallization time and temperature, shear during cooling, and crystallization vary from study to study. Moreover, a number of studies focus on the addition of emulsifiers to alter the structure of fat-based products (Litwinenko and others 2002; Kalnin and others 2004; Martini and Herrera 2008; Shiota and others 2011; Saadi and others 2012), as both crystallization and polymorphism is influenced (Kalnin and others 2004). Furthermore, the rheological properties can also be manipulated upon emulsifier addition (Saadi and others 2012).

image

Figure 6. The elastic modulus (G’) for blends of anhydrous milk fat (AMF) and rapeseed oil (RO) in different blending ratios. The blends were measured at 5 °C using corrugated parallel plates.

Download figure to PowerPoint

Manufacturing

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

The manufacturing process strongly affects the rheological behavior of the final products, as cooling rate, shear, and temperature during processing all affect fat crystallization, hence, network formation and microstructure. Today, 2 commercial methods are used for manufacturing milk fat-based products: churning technology and emulsification technology. The churning technology, often referred to as the Fritz method named after the inventor (Frede and Buchheim 1994), has been developed from the traditional batch churning of precrystallized cream. The churning can be done either in continues process or with the simpler batch churning process (Frede and Buchheim 1994). Initially, the cream is separated from the milk by centrifugation to reach a fat content of 23% to 35% for the batch churning method and 36% to 46% for the continues method. After pasteurization of the cream it undergoes cooling and ripening.

Thermal treatment

Polymorphism

A number of studies have focused on the effect of cooling rate on the physical properties of milk fat-based products (Vasic and deMan 1965; Heertje and others 1988; Herrera and Hartel 2000b,c; Campos and others 2002; Martini and others 2002; Wiking and others 2009; Kaufmann and others 2012a; Rønholt and others 2012a,b). Upon fast cooling (strong super-cooling), fat crystallization equilibrium will not be reached, and nucleation predominates over crystal growth. Consequently, many small homogeneous crystals will be formed, primarily of the α-form that eventually transform to β’ (Heertje and others 1988; Campos and others 2002). Contrary to slow cooling, this system will be closer to crystallization equilibrium. Thus, crystals of the β’-polymorph will form in a more stable lamellar structure. The nuclei formed upon slow cooling are of the high-melting triglyceride fraction. During continuous slow cooling, the middle- and low-melting fractions of the triglycerides will crystallize, forming layers surrounding the nuclei (Bunjes and others 2007). Consequently, mixed heterogeneous crystals are formed (Mulder 1953; Cantabrana and deMan 1964; Breitschuh and Windhab 1998). In a recent study by Bouzidi and Narine (2009), it is further stated that for mixtures of cocoa butter and 1,3-dilauroyl-2-stearoyl-sn-glycerol, the cooling rate also affects not only the crystallization mechanism but also the crystal growth rate. Moreover, they showed that for triglycerides the main barrier of nuclei formation is the chemical structure hence symmetry of the triglyceride molecule. Additionally, chemical changes in the composition of the milk fat are shown to cause deviations from the previous observed effect of cooling rate on crystal polymorphism (Mazzanti and others 2004).

Minor components

Upon addition of polar lipids to a model system of AMF, a fast cooling rate (5 °C/min) combined with low-crystallization temperature (17.5 °C) results in increased formation of β-crystals, a behavior that might be a consequence of the polar lipids acting as nucleation sites, hence inducing crystallization of the high-melting fraction of the triglycerides. Such crystallization results in the formation of α-polymorphs that subsequently transform to first β’-crystals and then to β-crystals (Mazzanti and others 2004). The formation of β-polymorphs is thereby highly dependent on the initial fraction of α-crystals (Figure 5) (Mazzanti and others 2004). For slow cooling (0.3 °C/min), however, the more stable β’-phase is formed directly from the melt. Due to the slow cooling, the polar lipids are segregated within the milk fat and hence do not affect the crystallization process (Mazzanti and others 2004). Rønholt and others 2012a studied the effect of fast (7.5 °C/min) and slow cooling (0.4 °C/min) on the crystal polymorphic behavior in cream. Here, the presence of α and β’ polymorphs was found to be independent of cream cooling rate (Rønholt and others 2012a). Such discrepancy between the studies could likely be ascribed to differences in the chemical composition of the samples studied.

Solid fat content

For milk fat, the reported results on the effect of cooling rate on SFC vary. Wiking and others (2009) studied cooling rates of 10 and 0.1 °C/min on milk fat samples. After 80 min of subsequent storage of the milk fat at 20 °C, identical SFC was measured, independent of cooling rate (Wiking and others 2009). But other studies report a higher SFC for fast-cooled milk fat (5 to 5.5 °C/min) compared to slow-cooled (0.1 to 0.5 °C/min) (Herrera and Hartel 2000b; Campos and others 2002; Kaufmann and others 2012a). Rønholt and others (2012a,b) measured SFC in butter after 24 h of storage at 5 °C. The butter was manufactured from cream subjected to fast (7.5 °C/min) or slow cooling (0.4 °C/min). No difference was observed in SFC after 24 h of isothermal storage at 5 °C. The subsequent storage time largely determines the effect of thermal treatment on SFC for milk fat-based products (Rønholt and others 2012b).

Rheology and microstructure

Cooling rate affects not only nucleation and growth (Herrera and Hartel 2000b; Rønholt and others 2012a,b) but also rheological properties (Heertje and others 1988; Rønholt and others 2012a,b) and microstructure (Heertje and others 1988; Herrera and Hartel 2000c; Rønholt and others 2012a,b) of milk fat-based products. Due to the many small homogeneous crystals formed upon fast cooling, the number of crystal–crystal interactions increases correspondingly (Heertje and others 1988). Within the first 24 h after production, a fast-cooled product is harder, compared to a slow-cooled one for both fat spreads (Heertje and others 1988), butter (Rønholt and others 2012a,b), and milk fat (Campos and others 2002; Wiking and others 2009; Kaufmann and others 2012a). This is illustrated in Figure 7.

image

Figure 7. The effect of cooling rate and cream maturing on the elastic modulus (G’) of butter. First, the cream was cooled either fast or slow. Second, half of the cream was used in butter manufacturing, while the other half was further matured for 48 h at 5 °C. After 48 h, butter was manufactured from the matured cream. The data were measured at 5 °C using vane-fluted cup geometry (drawn from data of Rønholt and others 2012a).

Download figure to PowerPoint

Shear

Herrera and Hartel (2000b,c) studied the relationship between cooling rate, rate of agitation, crystallization temperature, chemical composition, and storage time of fractionated milk fat blends. When cooling from 60 °C to a crystallization temperature of either 25, 27.5, or 30 °C at a cooling rate of 5.5 or 0.2 °C/min and agitation at 50 to 300 rpm, a correlation between a slow cooling and high modulus was observed (Herrera and Hartel 2000b). Even though SFC was higher for the fast-cooled samples (Herrera and Hartel 2000b). The rheological difference is likely a consequence of the different microstructure, where fast cooling combined with agitation caused formation of many small crystals separated by a liquid phase, while the irregular-shaped crystals formed by slow cooling formed a more dense fat crystal network (Herrera and Hartel 2000c).

Maturation

The crystallization of cream can be further controlled by maturation, also referred to as physical ripening (Heertje and others 1988; Frede and Buchheim 1994; Schäffer and others 1999; Rønholt and others 2012a). Maturation controls the crystallization occurring within the fat globules, either by separating or mixing high- and low-melting fractions of the milk fat, hence affecting crystal microstructure and, consequently, the rheological behavior (Precht and Peters 1981). Interestingly, when subjecting cream to 48 h maturing at 5 °C, the effect of slow versus fast cooling on microstructure and rheological properties is diminished. Likely, all crystals have at this state reached a critical size. As a consequence hereof, butter produced from such matured cream will have the same G’ independently of cream cooling rate (Rønholt and others 2012a). With respect to polymorphism, maturing at 5 °C will induce a transition from primarily presence of α- and β’-crystals to primarily β’-crystals (Rønholt and others 2012a). However, due to the unstable nature of the α-crystal, such transition will occur in the butter within the first 24 h of storage, independent of eventual maturing (Rønholt and others 2012a,b).

Another type of ripening is the so-called warm–cold–warm method, where the cream is first held for a period at one temperature, then temperature is increased, and finally decreased again before churning (Mortensen and Danmark 1982; Schäffer and others 1996). The principle of this technique is that, upon super-cooling many crystal nuclei are formed. When the cream is warmed during the 2nd step, a fraction of the fatty acids will melt, and further recrystallize during the 3rd step, the cooling. It appears that the butter then produced will have a higher amount of liquid fat and a decreased hardness by up to 25%, as compared to butter produced directly after cream cooling (Mortensen and Danmark 1982; Schäffer and others 2000) or by a cold–warm treatment (Schäffer and others 1996).

Churning

Churning is the next step after tempering the cream, and the cream is subjected to phase inversion (oil-in-water to water-in-oil emulsion) (Figure 1). During churning, the cream is separated into butter grains and buttermilk in a churning cylinder, followed by draining of the butter milk and by processing of the butter grains and removal of the formed butter (Frede and Buchheim 1994). The underlying explanation for this behavior is that upon whipping, air is mixed into the system forming air bubbles covered by a protein film. This occurs concomitant with a depletion of proteins in the cream serum. Consequently, some of the liquid fat is squeezed out of the milk fat globules and eventually (actually quite suddenly) phase inversion occurs in the formed mass forming butter grains (King 1953). Traditionally, the churning temperature is in the range from 10 to 15 °C. It has been argued that at temperatures above 13 °C, a relative high amount of the fat will be liquid (Nielsen 1971). Consequently, the possibility to incorporate high amounts of water in the butter by continuous mixing (the next manufacturing step) is limited, as it results in a wet and leaky product (Nielsen 1971). Churning at temperatures below 5 °C results in such a high SFC that only a small amount of liquid fat can be squeezed out of the fat globules during churning. Thus, after a long churning time, very small butter grains are formed (Samuelsson 1937). Moreover, the churning time is significantly affected by the cream cooling rate, fast-cooled cream having a longer churning time compared to slow-cooled (Rønholt and others 2012b). This tendency is considered to be a result of the large crystals formed by slow cooling, working as eroding agents penetrating the milk fat globule membrane. An increase in crystal size increases the degree of rupture of the milk fat globule membranes, thus facilitating phase inversion from oil-in-water to a water-in-oil emulsion (Boode and others 1993; Rønholt and others 2012b). Nowadays, most butter is made by continuous manufacture, whereas prior to about 1960 batchwise buttermaking in a rotating churn was the traditional method. However, the initial phase reversal theory proposed by King (1953) also holds true for the almost instantaneous phase reversed by the modern procedure.

Mixing

Second to churning is mixing. A buttermaker would refer to this step as either working or kneading. During mixing, the milk fat-based products are subjected to transport through the manufacturing equipment and rotational speed during mixing, both causing shear applied on the products (Heertje and others 1988). Focus has been on both the impact of the mixing process on butter and spreads at the industrial scale (Sone 1961; Heertje and others 1988), at laboratory-scale (Dolby 1941; Haighton 1965), and for model fat systems (Martini and others 2002; Mazzanti and others 2011). At industrial scale, the applied shear can be varied by, for example, changing residence time in the manufacturing equipment (Sone 1961; Heertje and others 1988) and rotation speed during mixing (Heertje and others 1988).

Mechanical treatment

With respect to rotational speed, the high shear rates applied break down the reversible and irreversible bonds within the crystal network (Dolby 1941; Sone 1961; Heertje and others 1988; Marangoni and McGauley 2003), decreasing the hardness to about a quarter of the original value (Dolby 1941; Heertje and others 1988). As a consequence of the discontinued structure formation, milk fat-based products subjected to strong mixing have a more granular structure compared to products subjected to limited or no mixing (Heertje and others 1988). A short residence time, on the other hand, induces a stronger degree of super-cooling. Consequently, smaller emulsion droplets are formed and coalescence is less likely to occur (Heertje and others 1988).

The reported results, on recovery of the rheological properties after mixing, vary from study to study. Dolby (1941) reports that even though the reversible bondings within the fat crystal network reinform to some extent during storage, several weeks of isothermal storage at 12.5 °C only causes a limited increase in hardness. Haighton (1965) observed a similar tendency upon remixing of margarine and shortening in a kneading apparatus, while Sone (1961) showed a more perfect recovery of the fat crystal network for butter stored at 16 °C for 10 d.

Studies using model systems have shown that applying shear in the range from 50 to 500 s−1 will rupture the crystals and decrease their size, both at isothermal conditions (Martini and others 2002; Mazzanti and others 2011) and simultaneously upon cooling (Herrera and Hartel 2000b,c; Kaufmann and others 2012b). As the crystals are broken and still not melted, SFC remains intact. At very high shear rates (1100 s–1), SFC decreases due to melting of the smallest crystals, caused by heat generated by the applied shear. When the shear is stopped, SFC increases to the original level (Mazzanti and others 2011). With respect to rheological properties, a shear rate of 500 s−1 affects the fat crystal network to such an extent that it does not fully recover during storage. Shear at 50 s−1, on the other hand, increases the hardness of the blend (Kaufmann and others 2012b).

The mixing process can be used to incorporate salt and flavors in a product (King 1953; Frede and Buchheim 1994). It has been hypothesized by Mulder (1947), and further discussed by King (1953), that water exists internal or externally in butter. The internal water is that incorporated into the butter grains during churning. The water percentage of the internal water phase is determined by the amount of crystallized fat within the fat globules. At high SFC, the fat globules appear harder and thereby retain their spherical shape allowing water to be present. In contrast, soft fat globules are more likely to be squeezed, resulting in smaller water cavities between the fat globules, and therefore in lower water content. During mixing, however, the external water phase is formed. Upon water addition, the water surrounds the surface of the butter grains and/or is entrapped in the fat crystal network formed between them. The size of the butter grains plays an important role here, as small grains have a large surface area and, consequently, retain more water (King 1953). Decreasing the churning temperature to 5 °C or below will, as discussed above, result in small butter grains. Theoretically, this could be one way to increase the amount of water in butter-like products. This approach was used in the preliminary test by Rønholt and others (2012b). It does, however, imply some practical problems in a laboratory set-up, as the amount of SFC increases in such a way that it is not easy to obtain homogeneous, yet reproducible samples. Using a scraped surface heat exchanger might be one way to overcome this challenge, as they are shown to reduce the temperature gradient through the sample resulting in more homogeneous products (Dumont and others 2000). The effect of water on crystallization mechanism and rheological properties of milk fat-based products is further discussed in Section "Water Content."

Temperature

The temperature of mixing is another parameter that can be used to modify the rheological properties of milk fat-based products (Buldo and Wiking 2012). Figure 8 illustrates the microstructure of butter grains (10 °C) blended with AMF (25, 40, or 60 °C). The blending was done at room temperature. At 60 °C, a large fraction of the fat crystal network is solubilized by the added AMF. At 25 °C, however, water is not successfully homogeneously incorporated in the fat matrix, as the fat crystal network is already set. Mixing at 40 °C was, therefore, chosen as the mixing temperature for further studies (Rønholt and others 2012b). Applying a high temperature when mixing milk fat-based products induces melting of the low- and eventually middle-melting fraction of the triglycerides, followed by recrystallization during cooling and storage of the products. Such recrystallization will, however, rebuild a less dense crystal network as a fraction of the milk fat is solubilized in the oil phase. Consequently, mixing at high temperature results in soft products (Buldo and Wiking 2012).

image

Figure 8. The figure illustrates the microstructure of butter grains blended with anhydrous milk fat (AMF) at different temperatures. The butter grains were kept constant at 10 °C, while AMF was incubated at 25, 40, or 60 °C prior to blending. Red illustrates the continuous fat phase, green the water droplets, and blue the phospholipids. Fat crystals are seen as dark black/gray shadows. The images were captured using confocal laser scanning microscopy. The scale bar is the same for all 3 images.

Download figure to PowerPoint

Water Content

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

Stability

Changing the water content can result in both an increase and a decrease of water droplet sizes and/or numbers of water droplets within the continuous fat phase, depending on the conditions (van Dalen 2002). The number of contact points between the crystals may, consequently, be changed and so may the rheological behavior of the products (Rønholt and others 2012b). From a stability perspective, Rousseau and others (2009) studied the phase coalescence mechanisms in tablespreads in the range from 28 to 34 °C by following changes in droplet size and SFC, as well as the spatial organization of the fat crystals by using LR-NMR and polarized light microscopy. For butter, they found coalescence to be determined by the amount of solid fat, as a minimum amount of between 2.5% and 9% solid fat is necessary to prevent coalescence of the water droplets. For margarine, it is primarily melting of the stabilizing Pickering crystals that determine the tendency for the water droplets to coalesce (Rousseau and others 2003, 2009). Briefly, Pickering crystals are interfacially active crystals, originating as a result of solidification of surfactants at the oil–water interface and/or migration from existing crystals to the interface (Rousseau and others 2009). The difference in the destabilization mechanism between butter and margarine is the result of a different fat crystal network between the two. At the same temperature, where butter has a higher SFC when compared to margarine, butter appears with aggregated fat crystals (3 to 5 μm) having uniform distribution of solid and liquid fat. In contrast, the fat crystals within margarine are larger (5 to 10 μm) and primarily present at the interface of the water droplets (Rousseau and others 2009). The study by Rousseau and others concludes that a critical SFC is needed to stabilize a fat crystal network. Under normal conditions, milk fat-based products are used at room temperature (approximately 20 °C) and stored in a refrigerator at 5 °C where a critical SFC is expectedly reached.

If conditions apply and a critical SFC is reached, the amount of water has a significant effect on the structural behavior of milk fat-based products and, consequently, spreadability and consumer acceptance of the products. When increasing the water content, the ratio between solid and liquid phases (both water and liquid oil) is shifted toward more liquid. It is a parameter that strongly affects the hardness and spreadability of a butter-like product (Pothiraj and others 2012; Rønholt and others 2012b), as illustrated in Figure 9. As the fraction of the fat crystal network is diminished, the product becomes softer. A consequence thereof might be a decreased stability of the water droplets (Mulder and Walstra 1974), as can be measured using LR-NMR.

image

Figure 9. The effect of water on the elastic modulus, G’, for butter-like samples measured 1 d after manufacturing (as described in Rønholt and others 2012a,b). The samples are obtained from phase inversion of cream subjected to either fast or slow cooling. The blue circle indicates samples measured using vane-fluted cup geometry, the rest measured using serrated parallel plates.

Download figure to PowerPoint

Crystallization

As previously mentioned, crystal polymorphism can be correlated to the rheological behavior of milk fat-based products (Sato 1999). Recently, Wassell and others (2012) studied the relationship between interfacial crystallization, choice of emulsifier, and temperature (SFC) in water-in-oil emulsions. By combining X-ray with DSC and polarized light microscopy, they were able to show that the fat crystals near the interface are organized with the lamellar planes, almost parallel to the interface. However, the study did not address the aspect of changing the water-to-oil ratio. Vanhouette and others (2002) and Vereecken and others (2009) studied the effect of water addition on fat crystallization in model systems. By adding different amounts of phospholipids (0.00% to 0.07%) and water (0.00% to 0.70%) to AMF, Vanhouette and others (2002) followed the crystallization under isothermal conditions at 25 °C, using DSC and LR-NMR. Increasing the water content significantly decreased the induction time for crystallization, as measured using LR-NMR. Vereecken and others (2009) studied the effect on crystallization, obtained by adding different amounts of monoacylglycerides to the oils from which the monoacylglecerides were obtained (rapeseed oil, palm stearin, and sunflower oil). For samples containing saturated monoacylglycerides, addition of 2% water increased the amount of solid fat and vice versa for unsaturated monoacylglycerides at certain temperatures. However, a water content of up to 2% is far from what is used industrially. When varying the water content in milk fat-based products from 20% to 32% water, the crystal polymorphism stays the same, as the products primarily contain β’-crystals with traces of β-crystals (Rønholt and others 2012b). This conclusion is further supported by SFC measurements. Upon correction of SFC to reflect the ratio between solid and liquid fat, a change from 20% to 32% water did not induce any difference in SFC (Rønholt and others 2012b). Similarly, a change in water content from 20% to 32% did not induce any changes in thermal behavior (Rønholt and others 2012b).

Even though an increased water content in milk fat-based products decreases the induction time for crystallization (Vereecken and others 2009), it did not seem to alter the crystal polymorphs found 1 d after manufacture (Rønholt and others 2012b). A significant increase in water content strongly affects the rheological properties and microstructure of milk fat-based products (Rønholt and others 2012b). Using water content as a tool to lower the fat content is possible, however, the rheological behavior will be affected. During storage, the water droplets within milk fat-based products with high water content are more likely to coalesce compared to the ones containing less water (Rønholt and others 2012b). Consequently, the water droplet size increases, hence affecting the sensory profile and microbial keeping quality (Frede and Buchheim 1994; van Lent and others 2008). To prevent such coalescence, addition of emulsifiers could be relevant. The interaction between water and monoacylglycerides seems to be an interesting tool to manipulate SFC and eventually the rheological behavior of milk fat-based products (Vereecken and others 2009). However, this behavior remains to be confirmed in an industrial environment.

Storage Conditions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

Storage at isothermal temperatures

Due to the wide range of triglycerides in milk fat-based products and, accordingly, a broad spectrum of melting points, the storage conditions have a high impact on the fat crystal network, hence also the structure of milk fat-based products (Figure 10) (Madsen 1971; Ulberth 1989; Segura and Herreera 1990; Martini and Herrera 2008; Vithanage and others 2009; Pothiraj and others 2012). If a milk fat-based product is subjected to a temperature higher than the one of normal storage, a fraction of the low-melting fatty acids is likely to melt and thereby escape from the mixed crystals formed during manufacturing and/or storage.

image

Figure 10. The effect of storage temperature on the elastic modulus (G’) for samples of anhydrous milk fat (AMF) and rapeseed oil (RO). Small deformation rheology was performed using serrated parallel plates, as described by Ronholt and others 2012b. The samples contained either 20% or 32% water (w) and a fat phase of either 0% or 30% RO, the rest being AMF. The samples were stored at either 5, 15, or 20 °C for 2 (top graph) or 9 d (bottom graph). Prior to analysis performed at 5 °C the samples were incubated at 5 °C for 12 h.

Download figure to PowerPoint

By increasing the storage temperature from 10 to 20 °C, SFC decreases by up to 50% for butter and 25% for blends (Vithanage and others 2009). It likewise goes for the rheological properties of butter stored at increasing temperatures: the hardness decreases when increasing the storage temperature (Dolby 1941). For storage at refrigerator conditions, it is well known that postcrystallization will occur. Postcrystallization, also known as setting, is caused by crystallization of the continuously super-cooled liquid fat, both within the continuous phase and globular fat phase occurring during storage. The effect of postcrystallization on rheological behavior is most pronounced when the crystallization occurs in the continuous fat phase, as it then contributes to formation and/or development of the fat crystal network. Knowledge about the rheological behavior subjected to various storage conditions is of industrial interest for 3 main reasons: (1) The rheological behavior during storage is essential for the shelf-life of the products (Martini and Herrera 2008; Pothiraj and others 2012; Saadi and others 2012; Rønholt and others 2012b). (2) The effect of fluctuating temperature during storage is crucial for both consumer acceptance and transportation of the products to the consumer. A product needs to be stable during transportation to the supermarket and further to the consumer's home. When at home, the product needs to maintain its desired rheological behavior, even though it is taken in and out of the refrigerator (Segura and Herrera 1990; Martini and Herrera 2008; Pothiraj and others 2012; Rønholt and others 2012a). (3) The effect of storage temperature on rheological behavior can be used industrially to alter the rheological behavior of products (Segura and Herrera 1990; Garcia-Macais and others 2011, 2012). In addition to rheological changes, storage of milk fat-based products is also likely to induce chemical oxidation (Forman and Zajic 1969), increased microbial activity (Muys 1969; Delmarre and Batt 1999), and color changes (Kaur and others 2011). Storage conditions can be divided into storage at isothermal conditions (Madsen 1971; Lopez and others 2002; Garcia-Macais and others 2011, 2012; Pothiraj and others 2012; Rønholt and others 2012a,b; Saadi and others 2012) where the effect of postcrystallization can be studied in relation to rheological behavior, and storage at fluctuating temperatures (Mortensen and Danmark 1982; Ulberth 1989; Segura and Herrera 1990; Martini and Herrera 2008; Rønholt and others 2012a) where focus is on both recrystallization and postcrystallization.

Polymorphism

Lopez and others (2002) studied the crystalline structure in concentrated cream (40% fat) and AMF after 6 d of storage at 4 °C. Prior to storage, cream and AMF was quenched from 60 to 4 °C. Initially after 15 min of storage, α-crystals (4.14 and 4.17 Å) and β’-crystals were dominating in both AMF and cream together with traces of β-crystals (4.65 Å for AMF). After 4 d, a coexistence of α-, β’-, and β-crystals was observed, demonstrating a recrystallization toward more stable crystals during storage (deMan 1961; ten Grotenhuis and others 1999; Lopez and others 2002, 2005). Similarly, Rønholt and others (2012a) followed the crystallization in cream (38% fat) subjected to fast (7.5 °C/min) and slow cooling (0.4 °C/min), immediately after cooling (at first) and after 48 h at 5 °C. At time 0, α- and β’-crystals dominated with traces of β-crystals. After 48 h, however, only β’- and β-crystals were found. During transition upon storage at 5 °C, the α-crystal-related 3L structure (67 Å) and the β’-crystal related 2L structures (41 Å) rearranged to a strong 57 Å peak, corresponding to an α- to β’-crystal transition (Lopez and others 2005; Fredrick and others 2011). After phase inversion of cream to butter and subsequent storage for 4 wk, the butter contains β’- and β-crystal polymorphs (Rønholt and others 2012b). Increasing the storage temperature from 4 to 13 °C (Segura and Herrera 1990), and even 25 °C (Madsen 1971), does not induce different polymorphism in margarine, as a predominance of β’-crystals is still present.

Storage at fluctuating temperatures

Similarly, subjecting milk fat-based products to fluctuating temperatures during storage (5 °C for 3 h, 20 °C for 3h, followed by 9 cycles of 1 h at 5 °C and 1 h at 20 °C) did not induce any polymorphic difference compared to products stored at 5 °C, when analyzed after the cycle ended (Tanaka and others 2010; Rønholt and others 2012a). However, such temperature fluctuations induce different rheological behavior in the samples. For both butter and margarine-like products, SFC is higher after storage at fluctuating temperatures compared to storage at 5 °C. In contrast, G’ decreases after storage at fluctuating temperatures, as large crystals are formed limiting the number of crystal–crystal interactions (Haighton 1965; Tanaka and others 2010; Rønholt and others 2012a). The effect of fluctuating temperatures on crystallization behavior and growth can be tailored by chemical alteration of the products, as amount of palmitic acid has shown to have a significant impact on textural degradation caused by fluctuating temperatures (Tanaka and others 2010). A temporary increase in storage temperature from 8 to 20 °C was, however, shown to increase the hardness of butter by 25%, independent on cream heat treatment and degree of saturation of the triglycerides (Mortensen and Danmark 1982). The same phenomenon can be observed in Figure 10. A temporary increase in storage temperature results in melting of the low-melting fraction of the triglycerides. When storage temperature is decreased again, the now liquid fat will adsorb to the surface of the remaining crystals and eventually crystallize, forming larger crystals that upon further growth get mechanically interlinked. Thus, strong primary crystal–crystal interactions are formed and G’ increases correspondingly.

SFC is also strongly dependent on the time of storage. Milk fat-based products stored at isothermal conditions (5 °C) resulted in an increase in SFC. This increase was most significant from day 1 to 14, as a consequence of postcrystallization (Rønholt and others 2012b). Upon SFC increase, the microstructure changes due to formation of a more dense fat crystal network (Rønholt and others 2012b). Even though SFC increases during storage, a similar increase was not identified for G’, nor Hencky strain at fracture (Rønholt and others 2012b), illustrating the complexity of fat-based samples. Generally, the number of crystal–crystal interactions within the fat crystal network is essential for the rheological behavior of milk fat-based products. During storage, however, the initial rheological difference that might be observed between samples as a consequence of either differing water content (Rønholt and others 2012b), size, and number of crystals as a consequence of cream cooling rate (Rønholt and others 2012b), or fraction of milk fat globules is likely to diminish upon storage. The effect of postcrystallization on the rheological properties of milk fat-based samples is most pronounced in the continuous fat phase. The initial difference in G’ for butter, caused by a difference in crystal size, as a consequence of cream cooling rate, is not confirmed after 4 wk of storage at 5 °C (Rønholt and others 2012b). SFC was similar for the 100% AMF emulsion and the others containing different fractions of milk fat globules. But as no fat globules were present in the 100% AMF emulsion, the fat crystals formed and distributed in a larger continuous fat phase and, consequently, had fewer contact points. During storage, SFC increases and a continuous fat crystal network is formed, hence increasing G’. Upon formation of this critical SFC, the rheological properties for samples both with and without fat globules appear the same.

The structural stability and rheological behavior of milk fat-based products is primarily determined by stabilization by the fat crystal network (Litwinenko and others 2002; Rousseau and others 2003; Alexa and others 2010). When a critical SFC or successful stabilization of the dispersed phase is reached, the hardness of the milk fat-based products does not necessarily increase, even though SFC keeps increasing due to postcrystallization. As a result hereof, it can be concluded that SFC is not necessarily predictive for the rheological behavior of milk fat-based products (Herrera and Hartel 2000b; Litwinenko and others 2002; Narine and Humphrey 2004; Rønholt and others 2012a,b). The difference in microstructure and chemical composition of milk fat-based products results in a varied tendency of the products to coalesce during storage. For butter, an increased variation in rheological behavior can be observed (Pothiraj and others 2012; Rønholt and others 2012b), corresponding to an increased brittleness, as a consequence of water evaporation and coalescence of the water droplets (Rønholt and others 2012b). While the samples become more brittle, the risk of small cracks within the samples increases, hence a larger standard deviation for G’ can be observed (van den Tempel 1958; Pothiraj and others 2012; Rønholt and others 2012b). Comparing butter and margarine during storage, however, butter is characterized by having a higher G’ compared to spreads. Moreover, a higher stress is needed to deform butter compared to spreads (Vithanage and others 2009).

Chemical composition and storage

Changing the chemical composition of milk fat-based products is likely to alter the aging process during storage (Madsen 1971). A study compared changes in margarine (87% fat) during 4 wk of storage at 10, 18, or 25 °C. Margarine based on a fat phase originating from either palm oil, rapeseed oil, groundnut (peanut), herring, sunflower, or soya bean (listed by increasing amount of unsaturated fatty acids according to experimental iodine value) (Madsen 1971). While the sunflower oil was the most β-tending followed by rapeseed oil, rapeseed oil was surprisingly most associated with a sandy mouthfeel, as measured using subjective evaluation. Conclusively, not only the degree of saturation but also the position of the fatty acid in the triglyceride molecule, hence molecular symmetry, impacts polymorphism and rheological behavior during storage (Madsen 1971; Sato 1999). Similarly, deMan and others (1991) observed that addition of canola oil to margarine, studied at 21 °C, primarily resulted in the formation of β-crystals, while canola oil–palm oil and sunflower oil–palm oil resulted in the coexistence of β’- and β-crystals. Addition of soybean oil, on the other hand, favored formation of β’-polymorphs (deMan and others 1991).

The ratio of saturated versus unsaturated triglycerides within the milk fat product likewise affected the molecular symmetry, as unsaturated fatty acids are less symmetrical compared to saturated ones. Consequently, the ratio between saturated and unsaturated triglycerides affects the crystallization rate during storage, as the size of the crystals is either accelerating or retarded. Increasing the amount of unsaturated triglycerides prolonged the phase transition from β’- to β-crystals during storage (Madsen 1971), hence preventing formation of big, asymmetrical spherulitic-like crystals (Madsen 1971; Marangoni and Ollivon 2007). A lower degree of molecular symmetry favors formation of asymmetric crystals (Hollander and others 2003), hence affecting the microstructure and rheological properties of the products (Herrera and Hartel 2000c; Wiking and others 2009).

Summing up, the temperature during storage, combined with chemical composition, also strongly affects the tendency for partial coalescence to occur during storage. As illustrated in Figure 10, the rapeseed oil/AMF samples behaved differently during storage. While G’ for a majority of the samples appears stable, G’ for samples containing 32% rapeseed oil and 32% water increased during storage at 5 °C, but decreased after 9 d of storage at 30 °C. As rapeseed oil contains almost no fatty acids with a chain length shorter than 16 carbon atoms, the melting point spectrum is higher than that of milk fat. Consequently upon storage at 5 °C, continuous crystallization of rapeseed oil occurs, thus increasing G’. For lower amounts of rapeseed oil, a critical SFC is reached after 2 d, and G’ does not increase further during storage. But storage at 30 °C does result in a significant decrease in G’. As no critical SFC seems to be reached at this temperature, the water droplets start to coalesce. Consequently, phase separation will occur and G’ will decrease as a consequence thereof. For margarine, however, a high degree of super-cooling during storage increases the hardness (Martini and Herrera 2008).

Conclusion and Further Research

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

The rheological behavior of milk fat-based products is not determined by one single parameter, but by the interactions of many. Manufacturing parameters such as thermal treatment of cream prior to butter making, shear during processing, and the degree of mixing is shown to influence not only crystal polymorphism, but also the number and sizes of fat crystals. By affecting the number and strength of the crystal–crystal interactions within milk fat-based products, the product hardness changes correspondingly. Moreover, the water content and chemical composition of milk fat-based products are closely related to their rheological behavior. During storage, however, postcrystallization increases the SFC and strengthens the fat crystal network. The fat crystal network is strengthened by formation of more and stronger crystal–crystal interactions due to mechanically interlinked fat crystals as occurring during crystal growth. Microstructural changes of fat crystals during storage are, however, the biggest challenge to maintain the initially induced rheological differences in milk fat-based products. During subsequent storage, the continuously super-cooled fat will eventually crystallize and thereby strengthen the fat crystal network independently of processing conditions. Vegetable oils will, due to their chemical composition, most often not crystallize during storage at 5 °C, and are therefore one way to alter the rheological behavior of milk fat-based products. For further studies, it would be relevant to relate the rheological differences observed as a function of thermal cream treatment, water content, and fraction and number of fat globules to the sensory quality of milk fat-based products. Parameters such as spreadability, mouthfeel, color, and taste would be of interest. A more detailed understanding is needed of how the presence of milk fat globules is related to the rheological behavior of milk fat-based products. By quantifying and varying the fraction and number of milk fat globules, a more detailed understanding could be obtained. Moreover, it would be of industrial interest to further explore the link between microstructure and behavior during fluctuating temperatures by varying not only the fraction of milk fat globules, but also the water content, cream cooling rate, and storage period. Finally, it has been hypothesized in the literature that the churning temperature impacts the rheological behavior of the final product. A detailed study addressing this statement would provide valuable knowledge at the industrial level. However, several steps are needed to link the existing knowledge to industrial applications of new products with a high functional value and nutritional quality.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References

We thank the Danish Dairy Research Foundation and The Danish Food Industry Agency for financial support. Thanks to Thomas B. Pedersen for assisting with the images obtained using confocal laser scanning microscopy.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Chemical Composition
  5. Manufacturing
  6. Water Content
  7. Storage Conditions
  8. Conclusion and Further Research
  9. Acknowledgments
  10. References