| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Equipe Physico-Chimie des Systèmes Polyphasés, UMR 8612 du CNRS, 5 rue J.B. Clément 92296, Châtenay-Malabry, France
2 Laboratoire pour l’Utilisation du Rayonnement Electromagnétique, Bât. 209D Université Paris-sud, 91898 Orsay, France
3 Physico-Chimie des Colloïdes, Université Henri Poincaré, UMR7565 BP 239, 54506 Vandoeuvre-lès-Nancy, France
Corresponding author: C. Lopez; e-mail: Christelle.Lopez{at}rennes.inra.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
type, which are correlated to 2 exothermic peaks at 17.2 and 13.7°C, respectively. A time-dependent slow sub- 
reversible transition is observed at –10°C. Subsequent heating at 2°C/min has shown numerous structural rearrangements of the varieties into a single ß' form before final melting. This polymorphic evolution on heating, as well as the final melting point observed (~39°C), confirmed that cooling at 3°C/min leads to the formation of crystalline varieties that are not at equilibrium. An overall comparison of the thermal and structural properties of the crystalline species formed as a function of the cooling rate and stabilization time is presented. The influence on crystal size of the cooling rates applied in situ using temperature-controlled polarized microscopy is also determined for comparison.
Key Words: triacylglycerol • polymorphism • differential scanning calorimetry • x-ray diffraction
Abbreviation key: AMF = anhydrous milk fat, DSC = differential scanning calorimetry, 2L = bi-layered stacking, 3L = tri-layered stacking, SAXD = small-angle XRD, T = temperature, TG = triacylglycerol, WAXD = wide-angle XRD, XRD = x-ray diffraction, XRDT = XRD as a function of T
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The different polymorphic forms observed for TG were detailed by different authors (Small, 1986; Hagemann, 1988; Ollivon and Perron, 1992). The 3 major polymorphic forms identified are , ß', and ß in increasing order of stability. These polymorphic forms can be identified by x-ray diffraction (XRD).
Mulder (1953) was likely the first to report a double-melting behavior for milk fat and to ascribe this behavior to the occurrence of polymorphism. Some studies have focused on the differences between the final diffraction pattern obtained after rapid and slow cooling (deMan, 1961; Woodrow and deMan, 1968). The final diffraction patterns recorded for milk fat showed mainly the ß' form with traces of ß form in the samples that were slowly cooled. Timms (1979) found a very weak intensity of the peak characteristic of ß form in slowly cooled milk fat. Timms (1980) stated that partial transformation of milk fat crystals to the ß form can only occur if a considerable amount of milk fat is still liquid. van Beresteyn (1972) reported the formation of and ß' forms. After slow cooling at 0.1°C/min to 28°C and after storage of the sample at 28°C for 24 h, Schaap et al. (1975) also observed both the and ß' forms.
In all of these studies, the researchers have made a distinction between slow and rapid cooling but often without mentioning the exact cooling rates. However, as a consequence of polymorphism, fat behavior is, to a large extent, determined by the rate of cooling. Moreover, the polymorphic forms were often examined after considerable stabilization periods because of lengthy exposure times using XRD without taking into account the evolution of the samples.
Recently, access to x-ray high-flux sources (synchrotron) has permitted the study of the polymorphism and phase transitions displayed by complex TG mixtures as a function of T and the study of these data in comparison with differential scanning calorimetry (DSC) recordings (Lavigne, 1995; Keller et al., 1996; Loisel et al., 1998; ten Grotenhuis et al., 1999). ten Grotenhuis et al. (1999) studied polymorphism of milk fat by DSC and real-time x-ray powder diffraction at small angles, from 70 to –65°C, at different cooling rates varying from 0.5 to 20°C/ min. They observed the formation of the sub-, , and ß'polymorphic forms as a function of the cooling rate |dT/dt|. For |dT/dt| 
2.5°C/min, the least stable sub- form was recorded for T 
8–C, together with crystals. For |dT/dt| = 1.67°C/min, only crystals were formed. For |dT/dt| 1°C/min, they observed the formation of ß'forms together with a small amount of form. The ß'form was the most stable form observed in their study. Lavigne (1995) studied the polymorphism of milk fat and its fractions by XRD as a function of T (XRDT), at both small and wide angles (SAXD and WAXD, respectively), to characterize the longitudinal stacking and the lateral packing of the TG molecules within the lamellar structures, respectively. The structural data were compared with DSC recordings obtained in similar conditions.
Authors who studied polymorphism of milk fat by DSC agree with the existence of 3 endotherms recorded on heating (Timms, 1980; Marangoni and Lencki, 1998). These 3 melting peaks were interpreted in terms of fractions of milk fat. These fractions were called low melting point, medium melting point, and high melting point fractions. Lavigne (1995) identified the main TG of each fraction and related them to the thermal and structural properties of the whole milk fat and its fractions.
In previous papers, we studied the formation of the most unstable crystalline structures formed by anhydrous milk fat (AMF) TG after rapid quenching from the melt to –8°C (Lopez et al., 2001b) and the crystalline species formed at very slow cooling rate (Lopez et al., 2001c) as well as their evolutions on heating. To characterize the structural behavior of milk fat at intermediate cooling rates, the crystallization behavior of AMF is studied here at –3 and –1°C/min. The evolution of the crystalline structures formed during a subsequent heating at 2°C/min is also considered as in the preceding papers of the series. The same thermal treatments were applied in parallel to creams to determine the influence of fat globule interface onto TG crystallization (Lopez et al., 2001a). To avoid ice formation, the cooling was limited to –8°C in cream samples. Then, similar conditions, cooling to –10°C, are applied in this study to allow comparison of thermal and structural behavior in bulk and dispersed systems. Finally, the influence of cooling rate on the thermal and structural behavior of AMF is analyzed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
XRDT/DSC Measurements
Coupled, time-resolved XRDT, and high-sensitivity DSC experiments carried out in the same apparatus from the same sample were conducted on D22 bench of the high-energy synchrotron of LURE (Laboratoire pour l’Utilisation du Rayonnement Electromagnétique). The setup has been described previously (Lopez et al., 2001b). Briefly, D22 bench is equipped with 2 linear detectors that allow recording of simultaneous SAXD and WAXD data with sample-to-detector distances of 30 and 177 cm, respectively. The XRDT data and DSC measurements were synchronously collected vs. time by a single microcomputer. Crystalline ß form of high purity tristearin, synthetized and purified in the laboratory (Lavigne et al., 1993), was used as reference for wide-angle channel to scattering vector calibration of the detector. Silver behenate was used as reference to calibrate the XRD data collected at small angles (Blanton et al., 2000).
Melted AMF samples (~30 µL) were loaded in thin Lindeman glass capillaries (GLAS; Muller, Berlin, Germany) especially designed for XRD. The glass capillaries allow minimum attenuation of the beam and parasitic scattering. The thickness of the wall (0.01 mm) that is characterized by poor lateral (along the capillary) thermal conductivity, also guarantees minimum thermal losses (Keller et al., 1998), optimizing the thermal recordings. The AMF samples were heated in the calorimeter at 60°C for 10 min to melt all existing nuclei. The AMF samples were cooled at 3 and 1°C/min from 60 to –10°C. X-ray diffraction patterns were recorded with a time frame of 60 s, and DSC signal was recorded every 3 s during cooling, using a single computer for thermal control and data acquisition. Following cooling, the samples of AMF were heated from –10 to 60°C at 2°C/min with an XRD time frame of 60 s and DSC signal recorded every 3 s.
Each XRD pattern was analyzed using PEAKFIT software (JANDEL Scientific, Erkrath, Germany). The XRD peaks were fitted by Gaussian-Lorentzian (sum) equation as described in Lopez et al. (2001d) to determine the position of the maximum and the maximal intensity of each peak as a function of T.
DSC Measurements
Thermal analyses were conducted with a DSC-7 Per-kin Elmer (St. Quentin en Yvelines, France) using aluminum pans of 50 µL [pan (part #B014-3021) and cover (part #B014-3004)] hermetically sealed. The reference was an empty, hermetically sealed aluminum pan. The AMF samples were heated at 60°C during 5 min to melt all crystals and nuclei. Crystallization curves were recorded from 60 to –8°C at different cooling rates: 0.5, 1, 2, 3, and 5°C/min. Then, following cooling, all melting curves were recorded from –8 to 60°C at 2°C/min. The calibration of the calorimeter was controlled with lauric acid melting (melting point = 43.7°C; melting enthalpy = 8.53 Kcal/mol). It is worth noting that the aluminum pans used in this study likely induce different nucleation sites than glass capillaries as crystallization is observed at a lower T. This is likely related to the nature of the chemical species found at the surface of both materials.
Polarized Light Microscopy
Anhydrous milk fat was examined by microscopy between crossed polarizers and with a 
/4 retarder in white light using a NIKON E600 Eclipse direct microscope (Champigny/Marne, France) equipped with a long focus objective 40 x 0.55; 0 to 2 mm). NIKON Coolpik 950 camera is used as a picture recorder with a resolution of 1600 x 1200 pixels. A homemade sample holder was used to monitor the sample T between –10 and 60°C. Briefly, the sample is placed between 2 circular lamellae located in a cavity of a Peltier-cooled stage, the T of which is controlled at ± 0.2°C. All samples were heated to +60°C before being cooled at 0.2, 1, and 3°C/min or quenched to –8°C on the microscope stage.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Cooling at 3°C/min.
The XRD patterns recorded simultaneously at small and wide angles as a function of T during cooling of AMF from 60 to –10°C (Figure 1
) could be divided in 2 T-delimited domains. In the 60 T 19°C domain, XRD patterns recorded at both small (Figure 1) and wide (Figure 1, insert) angles show that milk fat TG are in their liquid state. Indeed, at small angles, no diffraction lines are recorded, and at wide angles, the x-ray bump recorded at about q = 1.396 Å–1 (4.5 Å) is related to x-ray scattering from the liquid-crystalline organization of TG (Larsson, 1972). For T 16°C, the recordings of diffraction lines at both small and wide angles correspond to TG crystallization. At wide angles (Figure 1, insert), the single peak of diffraction recorded at about q = 1.514 Å –1 (4.15 Å) is characteristic of TG crystallization with a hexagonal packing of the acylglycerol chains. At small angles (Figure 1), TG crystallization is related to the recording of XRD peaks at about q = 0.136 Å–1 (46 Å) and 0.163 Å–1 (38.5 Å) with an increase of x-ray scattering at very small angles (q = 0.05 Å–1). It is worth noting that the line at q = 0.136 Å–1 is observed slightly before that of q = 0.163 Å–1 (see subsequent DSC recording interpretation). From about 10°C, sharp diffraction lines are simultaneously recorded at q = 0.0873 Å–1 (72 Å), 0.1745 Å–1 (36 Å), 0.2618 Å–1 (24 Å), and 0.4363 Å–1 (14.4 Å). All of the diffraction lines increase in intensity as a function of the decrease in T.
|
(A and B, respectively).
|
form). Then, from about 14°C, the simultaneous increases in intensity of a new series of diffraction lines recorded at small angles mean that they correspond to the formation of the same crystalline organization with different orders of diffraction, namely a lamellar structure with a triple-chain length organization (3L) of 72 Å and a hexagonal packing of type. Crystallization of TG in the lowest degree of organization, namely the unstable form, is not surprising in the conditions of cooling used.
The evolutions, as a function of T, of the long and short spacings recorded at small and wide angles, respectively, during cooling of AMF at 3°C/min are presented Figure 3.
|
packing first decreases from 4.15 Å (16°C) to 4.13 Å (–1°C), meaning that the hexagonal organization of the acylglycerol chains becomes more dense than less dense upon increasing up to 4.16 Å (–8°C). The period of evolution observed is similar to that observed in cream (Lopez et al., 2001a) for which we concluded to the formation of sub- form at the end of cooling in the same conditions. Although the absence of a second broad peak at about 3.8 Å prevented us from reaching the same conclusion, its observation at the beginning of heating indicated that its formation was time dependent, as has already observed in cream (Lopez et al., 2001a) (Figure 4).
|
. Coupling of XRDT and DSC measurements allows us to relate the structural and the thermal behavior and then to identify the DSC peaks. From the initial T of crystallization measured at 17.2°C, 3 overlapped, exothermal events are recorded on the crystallization curve. The first exotherm is related to crystallization of the 2L (46 Å) variety; that of the 2L (38.5 Å) is slightly delayed (Figure 1; Figure 2C). The third exotherm corresponds to crystallization of the 3L (72 Å) structure with an initial T of crystallization measured at 13.7°C.
Heating at 2°C/min.
After cooling at 3°C/min, the AMF sample was immediately (after ~30 s) heated from –10 to 60°C at 2°C/min to study the polymorphic evolution of the crystalline structures formed.
The XRDT patterns recorded simultaneously at small and wide angles during heating of AMF are plotted as 3-D views (Figure 4). Small-angle XRD patterns could be divided into 4 T-delimited domains. In the first domain, the diffraction peaks formed on cooling decrease in intensity. Then, in the 6 < T < 14°C domain, complex phase transitions are observed. A single line of diffraction corresponds to the formation, and the successive melting of a double-chain length structure (2L) is recorded until final melting of all TG at T > 34°C. At wide angles, the following transitions are observed, starting from traces of sub-: sub- 
ß' liquid.
The XRD patterns recorded during heating of AMF at 2°C/min were analyzed using PEAKFIT software to determine the position of the maximum of each peak recorded as well as its maximal intensity. The evolutions of the maximal intensities of the XRD peaks recorded both at small and wide angles are plotted in the same graph as a function of T to relate the longitudinal organization to the lateral packing of the TG molecules in lamellar structures (Figure 5).
|
3L (72 Å) correspond to the melting of this crystalline structure. In the 6 < T < 12°C domain, the simultaneous formation of broad diffraction peaks at q = 0.094 Å–1 (67 Å) and 0.188 Å–1 (33.4 Å) corresponds to the formation of a 3L (67 Å) structure with its second order of diffraction (3L2(002)). This structure, which melts at T > 12°C, may originate from a 3L (72 Å) 3L (67 Å) transition during heating. The 2L lamellar structures first formed on cooling with thicknesses of 46 and 38.5 Å melt at T > 16°C.
At T > 8°C, SAXD patterns show the formation of a new double-chain length organization. This long spacing may be composed of TG initially incorporated in the 3L (72 Å) structure. Likely, the melting of the 3L (72 Å) variety leads to TG in the liquid state and to a rearrangement of TG to form both the 3L (67 Å) and the new 2L organization. At about 12°C, the inflexion point recorded in the evolution in intensity of the 2L variety is correlated with the melting of the 3L (67 Å), the 2L (46 Å), and the 2L (38.5 Å) structures. Likely, some TG originating from these 3 organizations incorporate into the new 2L stacking. Such rearrangements of the TG molecules into a lamellar organization are facilitated by the presence of a liquid phase. The 2L variety takes advantage of the melting of the other crystalline structures for its formation until 20°C. Then, the decrease in intensity of the diffraction line associated to this 2L longitudinal stacking corresponds to its progressive melting until its disappearance at T > 34°C. At wide angles, the meltings of the 3L (72 Å) and 2L (46 Å + 38.5 Å ) varieties are correlated with the decrease in intensity of the form. The formation of the new 2L structure is correlated with the recording of 2 diffraction peaks at about 3.8 and 4.2 Å , characteristic of the formation of a ß' form. These results show that a ß' polymorphic transition occurs during heating of AMF at 2°C/min.
As a summary, the analysis of the XRD data corresponding to the longitudinal (Figure 5A) and the lateral (Figure 5B) organizations of TG molecules within crystals recorded during heating of AMF leads to the following polymorphic transitions:
 |
The DSC curve recorded simultaneously with XRDT experiments is presented Figure 5C. The plot of XRDT data obtained from analysis of the patterns recorded at both small and wide angles and DSC data on the same T scale allows us to identify the thermal events. On the melting curve recorded, 3 endothermic peaks corresponding to the low melting point, medium melting point, and high melting point fractions are recorded together with an exotherm between the endotherms of the low and medium melting point fractions. The correlations between XRD and DSC data show that the first endotherm recorded on heating (low melting point fraction) corresponds to the melting of the 3L (72 Å) structure. The exotherm recorded between about 10 and 14°C by coupled DSC is related to the ß' phase transition. The endotherm recorded between ~14 and 20°C (medium melting point fraction) is related to the melting of both the 2L (46 Å) and the 2L (38.5 Å) structures. Finally, the endothermic, overlaped peaks recorded (high melting point fraction) corresponds to the progressive melting of the 2L structure until final melting of all AMF TG at T > 39°C.
The numerous transitions observed both by DSC, including an exothermic peak, and by XRDT, especially at small angles, demonstrate that several crystalline reorganizations occur during heating at 2°C/min following a cooling at –3°C/min. This is a consequence of the formation of very unstable varieties at this fast cooling rate.
Cooling at 1°C/min.
The structural behavior of AMF was monitored on cooling at –1°C/min. Then, the evolution of the crystalline structures formed was studied on a following heating at 2°C/min. The XRD patterns recorded at both small and wide angles as a function of T during the cooling and the heating of the same sample of AMF are presented in Figure 6.
|
), were related to the crystal size and number formed. The variations of the central scattering are also strongly influenced by the rate of cooling; the larger the crystals, the steeper the slope, in agreement with microscopy observations. |
Influence of Cooling Rate |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|dT/dt| 0.5°C/min (Figure 7A) (Ollivon 1982; Lavigne, 1995). Following crystallization, all of the samples of AMF were heated at the same rate, |dT/dt| = 2°C/ min, to study the influence of the cooling rate on the melting behavior of AMF TG (Figure 7B). Both sets of recordings show that the initial T of crystallization measured and the shape of both the crystallization and melting curves depend on the cooling rate. The initial T of crystallization of milk fat decreases with increasing cooling rate (Table 1), in relation to the nucleation process. On cooling, the authors generally consider the existence of 2 exothermic process that correspond to crystallization of 2 groups of TG (Lavigne, 1995; ten Grotenhuis et al., 1999). However, our results show the existence of 3 exotherms and even 4. The exotherm formed between the 2 others is hardly detected at |dT/dt| = 5°C/min. Thus, its characterization needs a good resolution and available experimental conditions.
|
|
. The curves corresponding to AMF samples cooled at –1, –2, –3, and –5°C/min are similar, showing the 3 distinctive endothermic peaks corresponding to the low, medium, and high melting point fractions. Similar results were obtained by ten Grotenhuis et al. (1999) with DSC melting curves recorded at 5°C/min after cooling of milk fat at –1, –2.5, –5, and –10°C/min. The exotherms, which delimit endothermal melting transitions, are more clearly evidenced for faster cooling rates (Figure 7B). This is interpreted as the faster the cooling rate the more metastable state formed, and the more intense the relaxation phenomenon. The final melting T observed as a function of cooling rate (Table 1) are not significantly different; that observed on cooling clearly reflects an influence of nucleation process. The crystallizations of only 2 groups of TG on cooling, but the melting of 3 on heating, necessarily imply the formation of several groups, at least 3, of mixed crystals. The broad exotherm recorded during cooling shows that the larger group of these mixed crystals is probably constituted by mixed types (with respect to unsaturation and chain length) of TG molecules. Indeed, the melting of this mixed crystal is progressive and leads to broad endotherms then to the formation of a liquid phase in addition to the crystals. Also, the exothermal processes observed on heating are related to the formation of liquid phases. This liquid phase favors structural rearrangements in relation to the monotropic behavior of TG, as mixed crystals may dissolve in and transform into a new polymorphic form.
The formation of mixed polymorphic crystals in milk fat has already been reported (Mulder, 1953; van Beresteyn, 1972; Mulder and Walstra, 1974). On fast cooling, likely a large number of low-melting TG are trapped into a non-selective crystal lattice initially formed by high-melting TG, and slow cooling promotes fractionation of the fat.
Structural behavior.
Figure 8 illustrates the various thermal treatments that were applied to AMF samples in the different studies of the series to analyze, by XRDT and DSC techniques, the crystalline structures formed, as well as their phase transitions (Table 2).
|
|
(A and B, respectively) represents the SAXD and WAXD patterns recorded at –8°C after cooling at the rates indicated Figure 8. Their comparisons show that the thermal treatment applied directly influences the crystalline structures formed by TG in bulk.
|
form with the coexistence of 2 longitudinal stackings of the TG molecules corresponding to triple-and double-chain length structures (Figure 9A). In isothermal conditions at about –10°C, this initial variety progressively transforms into sub- form, following a time-dependent process, as it is observed at the beginning of the heating and it was not at the end of the cooling (Figures 1 and 4). The small bump observed at about 3.8 Å on Figure 9B for quenched sample and the shift of the peak position shows that the process of the sub- transition likely takes about 3 min at –8°C. The observation of such a transition was not clearly reported before (Lopez et al., 2001b). Quenching of AMF at about 1000°C/min (Lopez et al., 2001b) allowed us to identify a metastable 2L (47 Å) structure characterized by the formation of a sharp diffraction peak. This structure corresponds to the segregation of TG that forms crystals upon rapid cooling to –8°C. The TG composition of this structure is not known, but such a 2L stacking may be formed by saturated TG with similar chain length (Small, 1986). The 3L (70 Å) structure formed by quenching of AMF is characterized by a broad peak of diffraction that may be related to the formation of crystals of small sizes and/or to a long-range disorder in the longitudinal organization of the TG molecules in these lamellar structures, both effects being induced by rapid cooling. This peak broadening is partly related to crystal size reduction as observed by polarized light microscopy (see subsequent). However, it has been shown that the major origin of this peak broadening is the accumulation of structural defects and possibly related to interface curvature.
The broadening of the wide-angle diffraction peak recorded after quenching (sub- form) corresponds to a mixture of crystals with the same polymorphic form but with different chemical compositions. Furthermore, the high value (4.22 Å) of the short spacing recorded likely corresponds to a more dense packing of the acylglycerol chains in the orthorhombic perpendicular subcell.
The lamellar structures formed by TG molecules on cooling at 3 dT/dt 1°C/min are similar. They correspond to the coexistence of well-defined 3L (72 Å), 2L (46 Å), and 2L (38 to 39 Å) longitudinal stackings with a lateral organization of the chains in a hexagonal subcell.
Both 2L and 3L packings being of type, and long spacings (47 and 70 Å) being quite large, it is supposed that they correspond to chains perpendicular to the planes of the lamellae. On the contrary, smaller distances observed (41.5 and 62.2 Å) after slow cooling should correspond to tilted chains relative to these planes.
At slower cooling rates, dT/dt = 0.1°C/min, the structural behavior of AMF mainly corresponds to a crystallization in the ß' form at about 24°C with the formation of a 2L (41.5 Å) longitudinal stacking, in coexistence mainly with a 3L (62.2 Å) structure formed at about 13°C. Moreover, traces of ß form were recorded (Figure 9B). For comparison, the XRD pattern recorded after storage of AMF at 4°C for t > 100 h is presented on Figure 9B. These results show that the most stable longitudinal stackings of AMF TG that we observed correspond mainly to a 2L stacking in coexistence with a 3L organization of TG molecules. The differences in the thicknesses of the lamellar structures observed after cooling at 0.1°C/min or storage at 4°C may be related to a tilt of the TG molecules in the lamellae, although a 54-Å period corresponds to a tilt (16 to 18 Å) reduction in thickness too large to correspond to single TG structure.
Whatever the cooling rate applied to AMF, initial crystallization occurs with a 2L longitudinal stacking likely formed by trisaturated high melting point TG. As a function of the cooling rate, the time given to TG molecules to crystallize and incorporate 2L lamellar structures is sufficient (0.1°C/min) or not (1°C/min) to achieve completion. When the slow-growing 2L (ß' and also to a certain extent) structures have not completed, the fast-growing 3L structures develop. Both processes compete with each other for the incorporation of the same TG. However, both types of organization lead to mixed crystals constituted by TG with different chemical characteristics. Consequently, on heating, the evolution as a function of T of the 3L structures leads to a polymorphic transition 3L 2L. This transition 1) shows that structural rearrangements occur in the crystals to increase their thermodynamic stability and 2) confirms the existence of mixed crystals in the stable forms of AMF.
It is worth noting that the TG organization in these mixed crystals depends on the crystalline form with different accommodations of the structural parameters. In unstable forms, the broad and unique peak observed at wide angles () indicate a poor lateral organization that is compensated by very sharp small angle peaks related to a well-defined (3L) longitudinal organization. The reverse process is observed for stable varieties for which a more dense lateral packing attested by numerous wide-angle sharp peaks is compensated by broadening of small angle peak (2L Figure 9A; 0.1°C/min and 4°C).
However, it is also worth noting that, contrary to what is observed for the lateral packing for which the subcell compactness defines a scale of stability, the stacking type has no relationship to stability. The 2L stacking is not necessarily more, or less, stable than 3L one. For AMF, the 2L stacking observed previously at 41 to 42 Å is more stable than the 3L one, but for cocoa butter, the 2L is less stable than the 3L. Moreover, in AMF, the 2L (47 Å) that rapidly disappears as a function of time after quenching is even less stable than the 3L. However, the decrease of the longitudinal spacing associated with the tilt of the chain (when the vertical chain becomes tilted) is associated with an increase in stability.
Supramolecular behavior.
Crystallization of AMF was monitored by polarized light microscopy. Figure 10 shows the crystals of AMF at –8°C after cooling at different rates. Rapid cooling gives a large number of small crystals, and slow cooling provides large spherulitic particules: the faster the cooling, the smaller the crystals. When there are only weak differences observed at the molecular scale by XRDT between crystal obtained at –1 and –3°C/min, a crystal mean size reduction is easily observed using polarized light microscopy.
|
. The contour plots of the scattering intensities observed at very small angles are drawn as a function of T and scattering vector for cooling at 3°C/min (and partly for 1°C/min), and the subsequent heating is shown in Figure 6. Contour plots observed at very small angles are sensitive to crystal formation as well as to transition through the number and size of the crystals that depend on the cooling rate. These changes in scattering at very small angles correlate with the decrease in size observed by microscopy as a function of cooling rate (Figure 10). Such variations could be used as complementary measurements for crystal growth studies (Campos et al., 2002). This analysis illustrates the richness of small angle x-ray patterns. Each x-ray pattern recorded from long-chain compounds contains at least triple information about crystal type identification, quantification, and size.
|
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The experiments reported in this study give deeper insight into the crystallization and polymorphic transitions of milk fat. By cooling milk fat at different cooling rates, the phase behavior was investigated. It was found that the TG of milk fat can crystallize in 4 different polymorphic subcell types: sub-, , ß', and ß with reversible sub- and irreversible ß' ß transition, and corresponding 2L and 3L stackings. This last stacking is characteristic of the longitudinal arrangement of TG in which short or unsaturated fatty acid chains are mixed together with long saturated fatty acid chains.
The existence of such a polymorphism may have an important consequence in rheological behavior. Indeed, the crystal polymorph has primordial importance for final product consistency and acceptability. Smaller crystals lead to firmer fat products, whereas larger crystals give a sensation of sandiness in the mouth.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
* The exact subcell packing of the phase displaying a 3L stacking at 67 Å is not clearly established (Figure 5).
Received for publication April 9, 2004. Accepted for publication July 15, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONInfluence of Cooling RateCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Ortiz-Gonzalez, R. Jimenez-Flores, D. R. Bremmer, J. H. Clark, E. J. DePeters, S. J. Schmidt, and J. K. Drackley Functional Properties of Butter Oil Made from Bovine Milk with Experimentally Altered Fat Composition J Dairy Sci, November 1, 2007; 90(11): 5018 - 5031. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Lopez, V. Briard-Bion, B. Camier, and J.-Y. Gassi Milk fat thermal properties and solid fat content in emmental cheese: a differential scanning calorimetry study. J Dairy Sci, August 1, 2006; 89(8): 2894 - 2910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Couvreur, C. Hurtaud, C. Lopez, L. Delaby, and J. L. Peyraud The linear relationship between the proportion of fresh grass in the cow diet, milk fatty acid composition, and butter properties. J Dairy Sci, June 1, 2006; 89(6): 1956 - 1969. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
