- Email: [email protected]
Relating crystallization behavior of monoacylglycerols-diacylglycerol mixtures to the strength of their crystalline network in oil
Food Research International xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Food Research International journal homepage: www.elsevier.com/locate/foodres
Relating crystallization behavior of monoacylglycerols-diacylglycerol mixtures to the strength of their crystalline network in oil ⁎
Iris Taverniera, , Kim Moensa, Bart Heymanb, Sabine Danthinec, Koen Dewettincka a
Laboratory of Food Technology & Engineering, Department of Food technology, Safety and Health, Ghent University, Coupure Links 653, 9000 Gent, Belgium Vandemoortele R&D Center, Prins Albertlaan 79, 8870 Izegem, Belgium c Department of Food Science, Université de Liège, Gembloux, Belgium b
A R T I C LE I N FO
A B S T R A C T
Keywords: Fat crystallization Partial glycerides Oil structuring X-ray diffraction Rheology
Diacylglycerols (DAGs) are interesting oil structuring molecules as they are structurally similar to triacylglycerols (TAGs), but are metabolized differently which results in weight loss and improved blood cholesterol levels upon dietary replacement of TAGs with DAGs. Many commercial products consist of a mixture of monoacylglycerols (MAGs) and DAGs, yet the effect of MAGs on the crystallization behavior of DAGs is still to be unraveled. Two types of commercial MAGs, one originating from hydrogenated palm stearin and one of hydrogenated rapeseed oil, were added in concentrations 1, 2 and 4% to 20% DAGs derived from hydrogenated soybean oil. Using differential scanning calorimetry, it was shown that the presence of MAGs delayed the onset of DAG crystallization. Rheological analysis revealed that MAGs also hindered crystal network development. Synchrotron X-ray diffraction analysis demonstrated that the addition of MAGs suppressed the formation of the β form and stimulated the development of the β’ form. Likely, MAGs mainly hindered the crystallization of 1,3DAGs, which are responsible for the development of the β form, and stimulated the crystallization of the 1,2DAGs, which can crystallize in the α and β’ forms. The presence of two polymorphic forms resulted in a decrease of the crystal network strength, as was derived from oscillatory rheological measurements. This research implies a different effect of monoacylglycerols on both the nucleation and crystal growth of 1,2- and 1,3-DAG isomers. This insight is not only relevant for oleogelation research, but also for emulsifying agents which often contain blends of MAGs, 1,2-DAGs and 1,3-DAGs.
1. Introduction Traditionally, the lipid phase in food products such as margarines is structured with crystalline triacylglycerol (TAG) hardstock, containing a high amount of saturated fatty acids (SaFAs) or even trans fatty acids (tFAs). Studies have convincingly demonstrated that the consumption of tFAs increases low-density lipoprotein (LDL) blood levels, while lowering high-density lipoprotein (HDL) blood levels (Dhaka, Gulia, Ahlawat, & Khatkar, 2011). High levels of serum cholesterol, particularly LDL, promote the development of atherosclerosis or coronary heart disease (Kwon, 2016; Mensink & Katan, 1990; Mozaffarian, Aro, & Willett, 2009; Oomen et al., 2001). Whereas there is consensus on the adverse health effects of tFAs' consumption, uncertainty persists about the health effects of the consumption of SaFAs (Chowdhury et al., 2014; de de Souza et al., 2015). Research suggests that partially replacing these atherogenic SaFAs with poly-unsaturated fatty acids (PUFAs) reduces the risk of coronary heart
⁎
diseases (Czernichow, Thomas, & Bruckert, 2010; Endo & Arita, 2016; Mozaffarian, Micha, & Wallace, 2010). Altogether, it seems that atherogenic SaFAs should be reduced and tFAs completely eliminated and that both should be replaced with unsaturated fats to improve the nutritional profile of our diet and, as such, our cardiovascular health. These insights have led to intensive research on alternative oil structuring, using a wide variety of techniques and molecules (Singh, Auzanneau, & Rogers, 2017). Interesting oil structurants are diacylglycerols (DAGs), which are diesters of fatty acids with glycerin. Consequently, DAGs are similar to TAGs, which are triesters of fatty acids with glycerin. Research suggests that the dietary replacement of TAGs with DAGs leads to weight loss and improves blood cholesterol levels due to a difference in the way DAGs are metabolized compared to TAGs (Lo, Tan, Long, Yusoff, & Lai, 2008). Additionally, Xu, Wei, Zhao, Lu, and Dong (2016) demonstrated rheologically that fat blends containing DAGs had a stronger network structure with a more solid-like nature compared to the fats structured purely with TAGs.
Corresponding author at: Coupure Links 653, 9000 Gent, Belgium. E-mail address: [email protected] (I. Tavernier).
https://doi.org/10.1016/j.foodres.2018.10.092 Received 26 July 2018; Received in revised form 25 October 2018; Accepted 30 October 2018 0963-9969/ © 2018 Published by Elsevier Ltd.
Please cite this article as: Tavernier, I., Food Research International, https://doi.org/10.1016/j.foodres.2018.10.092
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
2. Material and methods
DAGs can exist in two isomers: 1,2- (or 2,3) DAGs and 1,3-DAGs. The equilibrated ratio of 1,2-DAG and 1,3-DAG is about 3:7 for longchain fatty acids (Lo et al., 2008). The crystallization behavior of 1,3DAGs and 1,2-DAGs differs significantly. While 1,2-DAGs can crystallize in α and β′ polymorphic forms in a hairpin configuration, the 1,3-DAGs are β-tending in a V-shaped configuration and crystallize in either a high-melting polymorphic form β1, or a lower-melting polymorphic form β2 (Shannon, Fenerty, Hamilton, & Padley, 1992). The development of the β polymorphic form during storage has been related to the occurrence of spherulitic crystals, causing oil loss and sandiness in fatrich products such as shortenings and margarines (Chawla, Deman, & Smith, 1990). Yet, if small β crystals are formed during for example a rapid cooling procedure instead of during storage, they are not necessarily an issue. Similar to β’ crystals, small β crystals are also capable of forming a stable crystal network structure with desired rheological properties. MAGs are known to influence fat crystallization, both in terms of nucleation, crystal growth and polymorphic form. The effect MAGs have on fat crystallization depends on their composition (saturated/ unsaturated and chain length), concentration and structural similarity to the dominant fat or oil (Alfutimie, Al-Janabi, Curtis, & Tiddy, 2016; da Silva et al., 2017; K. W. Smith & Sato, 2018). Fredrick, Foubert, Van Sype, and Dewettinck (2008) demonstrated the occurrence of heterogeneous nucleation of milk fat due to the presence of long-chain saturated MAGs. MAGs seem to be able to organize separately in micellar structures, acting as a template for nucleation processes (da Silva et al., 2017). On the other hand, also the crystal growth and α to β' polymorphic transition were accelerated with the presence of MAGs, which confirms that MAGs can affect the speed of polymorphic transition of fat crystals if they join the crystalline growth sites. Verstringe, Dewettinck, Ueno, and Sato (2014) confirmed the templating effect of monopalmitin on palm oil with microbeam synchrotron XRD. Chai et al. (2018) purified fully hydrogenated palm kernel oil and fully hydrogenated coconut oil and found that changes in nanostructure and polymorphic transformation took place after the removal of minor nontriacylglycerol polar lipids, MAGs, DAGs, and free fatty acids. In this research the goal was to identify whether MAGs affect the crystallization kinetics and polymorphic behavior of DAGs. Smith, Bhaggan, Talbot, and van Malssen (2011) argued that components with structural similarity have the most profound effect on the crystallization behavior of the main fat. Therefore, hydrogenated DAG of soybean origin was combined with two types of MAGs, one originating from hydrogenated palm oil and one from hydrogenated rapeseed oil. We hypothesized that the components with a similar fatty acid composition most efficiently influence each other's crystallization behavior and rheological properties. Hydrogenated MAGs and DAG were used because these have a simple fatty acid composition, making it easier to understand certain phenomena during crystallization. Furthermore, Alfutimie et al. (2016) demonstrated that saturated MAGs most efficiently influence fat crystallization, compared to unsaturated MAGs. The same type of MAG may exhibit a different effect on the crystallization (induction or retardation) depending on the proportion added (da Silva et al., 2017). Therefore, the MAGs were added in concentrations of 1, 2 and 4 wt%. In a first set of experiments, non-isothermal crystallization was studied with differential scanning calorimetry (crystallization curves), time-resolved synchrotron X-ray diffraction and oscillatory rheology (temperature sweeps). In a second set, the crystalline network and its properties were studied after one hour isothermal at 5 °C, using differential scanning calorimetry (melting curves), powder X-ray diffraction and oscillatory rheology (amplitude sweeps). As such, it was possible to correlate the crystallization behavior to the rheological properties of the samples.
2.1. Materials Commercially available DAG (Trancendim 110) was donated by Corbion Ingredients (Amsterdam, the Netherlands). Trancendim 110 (Tm,peak = 68.62 ± 0.07 °C) originates from fully hydrogenated soybean oil and is referred to as SO_DAG. Dimodan HP and Dimodan HR, commercially available, distilled MAGs, were kindly supplied by Danisco, DuPont (Wilmington, Delaware, USA). Dimodan HP (Tm,peak = 69.22 ± 0.12 °C) originates from fully hydrogenated palm stearin and is referred to as PSt_MAG. Dimodan HR (Tm,peak = 73.71 ± 0.13 °C) originates from fully hydrogenated rapeseed oil and is referred to as RO_MAG. Edible sunflower oil was kindly provided by Vandemoortele (Gent, Belgium). 2.2. Fatty acid analysis The fatty acid compositions of the SO_DAG, PSt_MAG and RO_MAG were analyzed with a GC-FID (gas chromatography with a flame ionization detector). First, fatty acid methyl esters (FAME) were prepared by dissolving 10 droplets of oil in 2 mL diethyl ether. Subsequently, 2 mL of 5% KOH in methanol was added to the mixture, which was allowed to react for 3 min. After the reaction, 2 mL demineralized water and 10 mL heptane were added. The mixture was shaken and the heptane layer, containing the FAME, was removed. This layer was washed twice in 4 mL demineralized water. After washing, the heptane layer was dried with a pinch of sodium sulphate. Subsequently, the sample was injected on the GC. FAME GC was carried out on a Varian GC with a BPX70 column (fused silica; 70% cyanopropyl polysilphenylene-siloxane, 50 m length, 0.22 mm internal diameter, 0.25 μm layer thickness). Helium was used as carrier gas. An amount of 1 μL sample solution was injected via the split/splitless injector (split ratio: 100) using automatic injection. The oven temperature was 187 °C for 45 min. Detection was performed with a flame ionization detector (FID). Table 1 gives an overview of the fatty acid composition of the SO_DAG, PSt_MAG and RO_MAG. 2.3. Carbon number analysis The amount of acylglycerols (mono-, di- and triacylglycerols or MAGs, DAGs and TAGs) was analyzed using carbon number GC. The commercial SO_DAG, PSt_MAG and RO_MAG were melted in an oven and 1 droplet (≈20 mg) was introduced in an Erlenmeyer of 50 mL. The fat was dissolved in 2 mL pyridine. To reduce the polarity of the partial acylglycerols, the samples were silylated. The silylation was performed by adding 0.4 mL of N-methyl-N-trimethylsilytrifluoroacetamide. Subsequently, the mixture was shaken and allowed to react for 10 to 15 min at 40 to 50 °C. The silylated products were redissolved in 40 mL Table 1 Fatty acid composition [%] and acylglycerol composition of PO_DAG, SO_DAG and PSt_MAG.
2
Component
SO_DAG
PSt_MAG
RO_MAG
C12:0 C14:0 C16:0 C18:0 C18:1 C20:0 C22:0 C24:0 FFAs MAGs DAGs TAGs
0 0.1 10.6 87.7 0.3 0.6 0.5 0.2 0.4 7.3 78.7 13.5
0.2 1.3 56.0 41.4 0.2 0.4 0 0 0 98.5 1.5 0
0 0.1 6.5 90.3 0.5 1.8 0.6 0 0 98.2 1.8 0
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
enclosed in glass capillaries and the temperature was controlled by a Linkam hot stage. The samples were subjected to the same time-temperature profile as for the non-isothermal DSC experiments: holding at 80 °C for 10 min, cooling at 10 °C/min to 5 °C.
isooctane. Carbon number GC was carried out using a Varian GC equipped with a CP SimDist column (fused silica, 10 m length, 0.53 mm internal diameter, 0.10 μm layer thickness). The injection volume was 1 μL. The oven temperature was programmed to cool from 50 °C to 350 °C at 25 °C/min and to subsequently keep isotherm at 350 °C for 6 min. Empower software was used to analyze the peaks. Table 1 gives an overview of the MAG, DAG and TAG composition of the SO_DAG, PSt_MAG and RO_MAG.
2.8. Powder X-ray diffraction spectroscopy The isothermal crystallization period was investigated by XRD using a Bruker D8-Advanced Diffractometer (Bruker, Germany) (l Cu = 1.54178\AA, 40 kV, and 30 mA), equipped with an Anton Paar temperature control system composed of a TTK 450 low-temperature chamber connected to a waterbath (Lauda) and heating device (TCU 110 Temperature Control Unit) (Anton Paar, Graz, Austria). Similar to the DSC measurements, the samples were heated to 80 °C for 10 min and subsequently cooled at a cooling rate of 10 °C/min to 5 °C. The samples were kept isothermally in the equipment at 5 °C during 1 h. WAXD and SAXS were recorded using a Vantec-1 detector (Bruker, Germany) during the isothermal crystallization. D-values were directly calculated by Diffraction Suite Eva software. The XRD patterns were analyzed using PeakFit (SeaSolve Software inc., Framingham, USA).
2.4. Preparation of the monoacylglycerol – diacylglycerol blends Different concentrations of MAGs (1, 2 and 4%) combined with 20% DAGs were dispersed in sunflower oil, resulting in SO_DAG:PSt_MAG and SO_DAG:RO:MAG combinations. In addition, samples of 21, 22 and 24% SO_DAG were prepared. The dispersions were stirred with a magnetic stirrer at 80 °C until a homogeneous samples were obtained. When the blends were visibly free of dispersed material, they were further mixed for 30 min to ensure complete melting. These homogeneous solutions were analyzed with a rheometer, with Differential Scanning Calorimetry (DSC) and with synchrotron X-ray diffraction. The molten samples at the same temperature were analyzed and underwent the same time-temperature profile. As such, a controlled and equal cooling rate can be ensured for all the analysis and all the samples.
3. Results and discussion 3.1. Crystallization mechanism of SO_DAG The crystallization curve and synchrotron XRD patterns (SAXS and WAXD) of 20% SO_DAG are visualized in Fig. 1. These curves were obtained by cooling the samples from 80 to 5 °C at a cooling rate of 10 °C/min. The main crystallization peak of SO_DAG has an onset temperature of 50.97 ± 0.45 °C. A second, smaller peak can be discerned at lower temperatures (Tc = 28.06 ± 0.20 °C). As previously mentioned, SO_DAG is a commercial sample and contains relatively high MAG and TAG fractions. Fig. 2 demonstrates that the addition of more MAGs does not increase the intensity of this second crystallization peak. Therefore, we are not inclined to relate this peak to MAG crystallization. Alternatively, the second peak might be related to the crystallization of 1,2-DAGs which generally crystallize at temperatures lower than 1,3-DAGs (Lo et al., 2008). Saitou et al. (2012) used gas chromatography to determine the amount of crystalline matter originating from 1,3-DAGs and 1,2-DAGs in a DAG oil. They found that 35% of the total DAG fraction in terms of chemical composition were 1,2DAGs while the crystalline mass of the DAG oil consisted for only 4% out of 1,2-DAGs. The 1,2-DAG crystal fraction was much lower than expected based upon the chemical composition. As visualized in Fig. 1, the peak related to the crystallization of 1,2-DAGs is therefore very small. Apart from the crystallization of 1,2-DAG, also TAGs crystallize within this temperature range. We therefore related this smaller peak to the crystallization of both 1,2-DAGs and TAGs. In Fig. 1b and c, clear diffraction peaks can be discerned starting from 47 °C which is close to the crystallization temperature of the first peak in the cooling profile. In the WAXD profile, these first peaks are identified at short-spacings d = 3.84 Å, d = 4.53 Å and d = 4.59 Å, which are related to the appearance of a β polymorphic form (Craven & Lencki, 2011b; Saitou et al., 2012; Shannon et al., 1992). Shannon et al. (1992) identified two β forms (β1 and β2) with short spacing values of 3.74, 3.82, 4.55 and 4.64 Å for β1 and 3.69, 3.78, 3.90, 4.55, and 4.68 Å for β2 of the 1,3-DAGs. These short spacing values are situated very closely together and a distinction between the high-melting and lowmelting β polymorphic forms was not possible in our results. Starting from 36.2 °C, which is the onset temperature of the second crystallization peak on the cooling profile, a lower intensity peak at shortspacing d = 4.17 Å is distinguished. This peak is indicative of the α polymorphic form, which could result from TAG and 1,2-DAG crystallization as discussed previously (Sato, 2001). In the SAXS diffraction pattern, only one peak is identified (d = 52.05 Å), which is related to the occurrence of 2 crystal layers (2 L) (Marangoni & Narine, 2002).
2.5. Differential scanning calorimetry The thermal profiles of the MAG:DAG blends were obtained with a Q1000 DSC (TA Instruments, New Castle, Delaware, USA) equipped with a refrigerated cooling system. Nitrogen was used as purge gas. The cell constant and temperature were set with indium (TA Instruments). An additional temperature calibration was done using azobenzene (Sigma-Aldrich, Bornem, Belgium) and undecane (Acros Organics, Geel, Belgium). The sample (weighing between 6.0 and 8.0 mg/cup) was placed inside an aluminum pan and sealed with an aluminum lid (TA Instruments). The experiments consisted of heating to 80 °C and cooling to 5 °C at a cooling rate of 10 °C/min, which resulted in the cooling curves. Subsequently, the samples were kept isothermal during one hour at 5 °C and then heated again at 10 °C/min to 80 °C. The thermal analyses were performed in triplicate. 2.6. Rheology The rheological behavior was analyzed on the advanced rheometer AR2000ex (TA Instruments, New Castle, USA) equipped with a Peltier system and water bath (Julabo, Seelbach, Germany) for temperature control. The starch pasting cell (shear rate factor = 4.500 s−1, shear stress factor = 48,600 1/m3, gap = 5500 μm) was utilized to record the gelling points during cooling from 80 to 5 °C at a cooling rate of 10 °C/ min. The oleogels were then kept isothermal at 5 °C for 1 h. After this isothermal period, the samples were subjected to an amplitude sweep by increasing the oscillation stress from 0.01 to 1000 Pa at a frequency of 1.0 Hz. The rheological analysis were performed in triplicate. 2.7. Synchrotron X-ray diffraction The non-isothermal crystallization of the samples was studied by Xray diffraction (XRD) using synchrotron radiation. Wide-angle X-ray diffraction (WAXD) and small-angle X-ray scattering (SAXS) measurements were performed on the Dutch-Belgian (DUBBLE) beamline BM26 at the European Synchrotron Facility (ESRF) in Grenoble (France). The experiments were performed at a fixed wavelength λ of 1.24 Å. For recording WAXD, a linear 300 K photon counting Pilatus detector was used, while for collecting the SAXS images, a large area high sensitive photon counting 2D Pilatus detector was used. The samples were 3
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 1. (a) Crystallization curve, (b) SAXS and (c) WAXD patterns of 20% SO_DAG during cooling to 5 °C.
3.2. Effect of monoacylglycerols on the crystallization of diacylglycerols
the peak temperatures and the peak widths at half height of the main peak for the DAG:MAG combinations. The addition of MAGs to SO_DAG shifts the onset of crystallization to lower temperatures, likely through attractive interactions between the MAG and DAG molecules which might allow the MAG to crystallize on the surface of the early seed
Fig. 2 visualizes the crystallization curves recorded during cooling of 20% SO_DAG, blended with 1, 2 and 4% PSt_MAG and with 1, 2 and 4% RO_MAG and Table 2 gives an overview of the onset temperatures,
Fig. 2. Crystallization curves recorded during cooling of SO_DAG, blended with 1, 2 or 4% PSt_MAG (a) or with 1, 2 or 4% RO_MAG (b). 4
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
onset of DAG crystallization. Yet, Fig. 2 does reveal a shoulder adjacent to the main peak for the samples containing 4% RO_MAG and PSt_MAG. This peak is likely related to the MAG crystallization and could indicate the presence of two overlapping peaks, one originating from MAG and one from DAG crystallization. To better understand the effect of MAGs on the crystallization of SO_DAG, the combinations were analyzed using synchrotron XRD. The samples were cooled and simultaneously measured using the same time-temperature profile during synchrotron XRD measurements as for DSC. In Fig. 3 and Fig. 4, a clear effect of the PSt_MAG and RO_MAG addition on the polymorphism of SO_DAG is visualized. Both types of MAGs give a similar effect on the crystallization of SO_DAG. With the addition of MAGs, the crystallization in the β polymorph (peak at 4.50–4.59 Å) seems to be suppressed, the peak at 4.15–4.18 Å becomes more dominant during cooling, while the peak at 3.79–3.86 Å remains constant or increases in intensity. Increasing MAG concentrations results in an earlier onset of the peak at 4.09–4.14 Å, which is why we attribute this peak to the crystallization of MAG in the α polymorphic form. The crystallization curves also revealed the presence of an initial shoulder at concentrations of 4% PSt_MAG and RO_MAG, which can now be related to the presence of the α polymorphic form. The initial α crystals transform into a β’ form since the initial α peak disappears and the peaks at 3.79–3.86 Å and 4.15–4.18 Å develop increasingly during cooling. For the samples where 1 and 2% PSt_MAG were added, three peaks are still present (3.84–3.86 Å, 4.15–4.16 Å and 4.54–4.51 Å) at the end of the cooling procedure, which indicates the presence of both β and β′ crystals. As discussed previously, SO_DAG is a mixture of MAGs, TAGs, 1,2DAGs and 1,3-DAGs, which have different crystal packings. Possibly, the addition of MAGs stimulates the crystallization of one component
Table 2 Crystallization onset and peak temperatures of 20SO_DAG, blended with 1, 2, 4% of PSt_MAG and 1, 2, 4% of RO_MAG. Tc,
20SO_DAG 20SO_DAG:1PSt_MAG 20SO_DAG:2PSt_MAG 20SO_DAG:4PSt_MAG 20SO_DAG:1RO_MAG 20SO_DAG:2RO_MAG 20SO_DAG:4RO_MAG
onset
50.96 49.50 47.68 48.02 49.97 48.88 48.65
(°C)
± ± ± ± ± ± ±
0.45a 0.21bc 0.12f 0.06ef 0.28b 0.21cd 0.19de
Tc,
peak
49.43 46.50 46.12 44.47 48.52 45.51 46.17
(°C)
± ± ± ± ± ± ±
0.71a 0.78c 0.01cd 0.10e 0.33b 0.19d 0.08cd
Peak width at half maximum (°C) 2.96 3.63 4.33 5.06 3.60 5.11 5.32
± ± ± ± ± ± ±
0.28e 0.07d 0.12c 0.02b 0.07d 0.12ab 0.10a
Different letters, within the same column, indicate significant differences at p $ < $ 0.05 according to ANOVA analysis for normally distributed data with equal variances and according to the Mann-Whitney U test for not normally distributed data and/or data without equal variances (SPSS 22).
crystals of the DAG or vice versa (K. W. Smith & Sato, 2018). As such, the formation of nuclei might be hindered. The distinct peak at lower crystallization temperatures for the SO_DAG thermogram disappears with the addition of MAGs and forms a tail of the main, high-melting peak. The addition of MAGs causes the peak width at half height to broaden (Table 2). According to Humphrey, Moquin, and Narine (2003), this indicates the presence of an inhibitor to nucleation. A wider crystallization peak implies that the final crystals have grown from several types of nuclei, as nucleation occurred over a larger time span. A more narrow crystallization peak suggests that all crystals within the sample nucleated at approximately the same time. In contrast to the work of da Silva et al. (2017), Table 2 does not reveal a significantly different effect of the different MAG concentrations on the
Fig. 3. WAXD patterns of SO_DAG containing (a) no PSt_MAG, (b) 1% PSt_MAG, (c) 2% PSt_MAG and (d) 4% PSt_MAG during cooling to 5 °C. 5
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 4. WAXD patterns of SO_DAG containing (a) no RO_MAG, (b) 1% RO_MAG, (c) 2% RO_MAG and (d) 4% RO_MAG during cooling to 5 °C.
unequivocally determine the effect of pure MAGs on either 1,2-DAGs or 1,3-DAGs. In the SAXS pattern of SO_DAG blended with different PSt_MAG concentrations, peaks are distinguished at d = 52.24 Å (1% PSt_MAG), d = 52.43 Å (2% PSt_MAG) and d = 52.73 Å (4% PSt_MAG) (Fig. S1). For the blends of SO_DAG with RO_MAG, peaks are found at d = 52.43 Å (1% RO_MAG), d = 52.63 Å (2% RO_MAG) and d = 53.12 Å (4% RO_MAG) (Fig. S2). With the addition of more MAGs, the peak position shifts to slightly higher values but remains within the 2 L range. Likely, the angle between the acyl chains increases with the incorporation of MAGs in the crystal lattice, which causes the layer thickness to increase. The layer thickness is slightly smaller for 20% SO_DAG combined with PSt_DAG compared to SO_DAG combined with RO_DAG which can be related to the bigger chain length of the constituents as more stearic acid is present in RO_MAG. Long spacings increase with increasing chain length (Marangoni, 2004).
Fig. 5. Molecular packing of 1,3-DAGs (left) and 1,2-DAGs (right) and effect of MAG addition.
3.3. Rheological behavior of MAG:DAG combinations during crystallization
but hinders the crystallization of another. The existing literature on DAG crystallization states that 1,3-DAGs can only occur in β polymorphic forms and 1,2-DAGs in α and β′ polymorphic forms (Craven & Lencki, 2011a; Craven & Lencki, 2011b; Shannon et al., 1992). Fig. 5 schematically represents the hypothesized effect of MAG addition on DAG crystallization. Based on this representation, one can understand how the presence of MAGs might hinder the crystallization of 1,3-DAGs, but have a different effect on the crystallization of 1,2-DAGs. MAGs are likely more easily built in the crystal lattice of 1,2-DAGs compared to 1,3-DAGs. However, more fundamental research is required to
This pronounced effect of MAG addition on the crystallization behavior of SO_DAG should also be represented in the rheological properties, which was investigated by recording the complex modulus and phase angle during cooling in the rheometer using the same timetemperature profile as used for the DSC analysis and synchrotron XRD measurements (Fig. 6). A higher |G*| indicates a stronger and more rigid network structure in the material. The phase angle δ evaluates the visco-elastic properties of a system. In a pure viscous system, the phase angle is 90°, while the phase angle is 0° in a pure elastic system. At a 6
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 6. Gelling curves for SO_DAG blended with PSt_MAG (left) and RO_MAG (right), as determined with oscillatory rheology. The full lines represent the complexes modulus, the dashed lines correspond to the phase angle.
Fig. 7. Melting curves of SO_DAG blended with 1, 2 and 4% of PSt_MAG (top) and with 1, 2 and 4% of RO_MAG (bottom), isothermally crystallized for 1 h at 5 °C. An offset of 0.2, 0.4 and 0.6 was used.
phase angle of 45°, the solid-like properties are equal to the liquid-like properties. The structure of the 20% SO_DAG sample is formed almost
instantaneously in one stage, which can be related to the fast crystallization of the 1,3-DAGs in the stable β polymorphic form. Both PSt_MAG and RO_MAG have a similar effect on the network built-up of 7
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 8. WAXD patterns of SO_DAG blended with 1, 2 and 4% of PSt_MAG (top) and with 1, 2 and 4% of RO_MAG (bottom), isothermally crystallized for 1 h at 5 °C. An offset was used to enhance visualization of the patterns.
SO_DAG. When PSt_MAG and RO_MAG are added, the onset temperature of network built-up is not significantly impacted, but the network development is slower and occurs in two-stages instead of one stage. The occurrence of two stages can be derived from |G*| curves, which shows two steep slope increases instead of one. The occurrence of two stages might indicate network rearrangements during cooling (De De Graef, 2009). The α to β′ polymorphic transition observed in Figs. 3 and 4 likely explains the two-stage network built-up. Furthermore, the final network is less strong for samples containing additional MAGs. This might be related to the presence of various polymorphic forms. For SO_DAG without additional MAG, the β form provides a strong and rigid network. With the presence of MAGs, both β and β′ forms are developed. Interestingly, the addition of 4% MAGs again improves the network built-up. At this concentration, MAGs are capable of structuring oil into a high viscosity liquid (Bin Sintang, Rimaux, Van de Walle, Dewettinck, & Patel, 2016). Both MAG and DAG might simultaneously and separately structure oil, which can explain the improved structural built-up. As RO_MAG contains more stearic acid than PSt_MAG (Table S1), structuring for RO_MAG will occur at a lower concentration. This can explain why this effect is only observed for RO_MAG at a concentration of 4%.
Fig. 9. Overview of the complex modulus of the pure SO_DAG and of SO_DAG combined with PSt_MAG and RO_MAG in concentrations of 1, 2 and 4%.
8
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
DAGs by employing pure 1,2-DAGs and 1,3-DAGs instead of blends.
3.4. Crystal network properties of MAG:DAG combinations after 1 h isothermal at 5 °C
4. Conclusion The short-term stability of the β′ crystals was assessed by keeping the samples isothermal at 5 °C for 1 h. The samples were evaluated using melting curves (Fig. 7) and powder XRD (Fig. 8). Thermodynamically, the β form is most stable, β′ is metastable and α is the least stable form, which results in a lower melting point for the α form, an intermediate melting point for the β′ form and the highest melting point for the β form (Sato, Ueno, & Yano, 1999). The addition of PSt_MAG or RO_MAG to SO_DAG clearly had an effect on the melting behavior of the blends (Fig. 7). The melting curve shifts to lower temperatures with an increasing PSt_MAG or RO_MAG concentration and the peak broadens for increasing concentrations of PSt_MAG or RO_MAG. This broadening peak indicates the presence of both β and β′ polymorphic forms for all combinations and is likely caused by the presence of crystalline 1,2-DAGs and 1,3-DAGs. The melting peak for the combinations with RO_MAG is less broad than that of combinations with PSt_MAG, which could be associated with the narrower fatty acid profile of RO_MAG:SO_DAG combinations or with a different balance in the amount of β to β′ polymorphs. For the addition of 1 and 2% RO_MAG, the melting peak only shifts slightly, which might indicate a lower amount of β′ compared to 4%. The melting enthalpies for the different samples were not significantly different (data not shown). A definite conclusion on polymorphic forms can be drawn from the powder XRD patterns (Fig. 8). At an increasing PSt_MAG or RO_MAG concentration, the formation of the β form is reduced, as can be deduced from the reduced intensity of the double peak at 4.55 and 4.63 Å, which is related to the β form. At a concentration of 4% PSt_MAG or RO_MAG, the peaks related to the β′ form at 3.83 and 4.20 Å for combinations with PSt_MAG and 3.85 and 4.17 Å for combinations with RO_MAG become more dominant. The intensity of these β′ peaks for the combination with PSt_MAG is higher than that of the combination with RO_MAG. For the concentrations of 1 and 2% RO_MAG, the double peak at 4.55 and 4.63 Å is still present, which is indicative of a substantial amount of β. PSt_MAG results in more β′ form than the RO_MAG, which supports the results obtained from the melting curves. The type of polymorph present in the fat crystal network also significantly affects the macroscopic properties of the fat (Narine & Marangoni, 1999). To assess the effect of MAG addition on the rheological properties of the blends, the samples were subjected to oscillatory rheology (amplitude sweeps). Fig. 9 depicts the complex modulus (|G*|) of the various samples after 1 h of isothermal time at 5 °C. The presence of multiple components and two polymorphic forms seems to interfere with the formation of a good crystal network structure. Interestingly, at a concentration of 4% RO_MAG, the |G*| increases again. Fig. 6 already demonstrated that a concentration of 4% RO_MAG hindered the network development less than a 1 and 2% RO_MAG. We hypothesized this was related to separate oil structuring by 4% RO_MAG. Therefore, increasing the MAG concentration might further enhance the crystal network strength. In this research, the focus was on fully hydrogenated MAGs and DAGs. These products are applied today in margarines as emulsifying agents. Yet up to date, it was unclear how they interacted with each other. On the other hand, to fully replace the traditional TAG-based fat with MAG:DAG combinations, the hydrogenated MAGs and DAGs studied here are likely less applicable as their melting temperature is higher than 50 °C, which could result in undesired melting properties in the mouth. Furthermore, bakery margarines and shortenings are generally prepared at high cooling rates and under shears and are subsequently stored at temperatures between 15 and 20 °C. To assess the applicability of these systems, more unsaturated components should therefore be studied and the effects of industrial processing (preparation and storage) should be researched. Further fundamental research should investigate the distinctive effect of MAGs on 1,2-DAGs and 1,3-
This research studied the influence of MAGs on the crystallization behavior of DAGs. The addition of MAGs delayed the onset of nucleation and retarded the crystal network development of the DAG-based system during cooling. Commercial DAG samples are a mixture of approximately 70% 1,3-DAGs and 30% 1,2-DAGs, with the 1,3-DAGs dominating the crystallization behavior. While 1,3-DAGs are known to crystallize in a β polymorphic form, 1,2-DAGs crystallize in the α and β′ forms. We found that the addition of MAGs induced the DAGs to crystallize in a β′ form during cooling. This β′ form remained present after storing the samples isothermally, as was deduced from the melting behavior and powder XRD analysis. A commercial monoglyceride originating from hydrogenated palm stearin more efficiently enabled the maintenance of the β′ polymorph, compared to a commercial monoglyceride originating from hydrogenated rapeseed oil. Possibly, MAGs enhanced crystallization of 1,2-DAGs in the β′ form but hinder the crystallization of 1,3-DAGs in the β polymorphic form. This may have shifted the balance of crystalline matter from being mainly composed out of 1,3-DAGs, as is the natural tendency, to being composed out of 1,2-DAGs. For fundamental understanding of the system more work is required on combinations of MAGs with pure 1,2-DAGs and pure 1,3DAGs. Supplementary data to this article can be found online at https:// doi.org/10.1016/j.foodres.2018.10.092. Acknowledgements This research was made possible thanks to the BOF (Special Research Fund) of Ghent University. The authors acknowledge the ESRF (Grenoble, France) for the use of the synchrotron facilities. The Dutch-Belgian Beamline (DUBBLE) research group at the ESRF and the Dutch organization for scientific research (N.W·O) are acknowledged for their help and continuous support of the DUBBLE project. Chi Diem Doan, Mohd Dona Bin Sintang and Ruben Van Overbeke are acknowledged for their help with the synchrotron XRD experiments. Joost Coudron of Vandemoortele is thanked for his help with the GC analysis. References Alfutimie, A., Al-Janabi, N., Curtis, R., & Tiddy, G. J. T. (2016). The effect of monoglycerides on the crystallisation of triglyceride. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 494, 170–179. https://doi.org/10.1016/j. colsurfa.2016.01.029. Bin Sintang, M. D., Rimaux, T., Van de Walle, D., Dewettinck, K., & Patel, A. R. (2016). Oil structuring properties of monoglycerides and phytosterols mixtures. European Journal of Lipid Science and Technology, 1–14. https://doi.org/10.1002/ejlt.201500517. Chai, X. H., Meng, Z., Cao, P. R., Liang, X.y., Piatko, M., Campbell, S., ... Fa, L. Y. (2018). Influence of indigenous minor components on fat crystal network of fully hydrogenated palm kernel oil and fully hydrogenated coconut oil. Food Chemistry, 255(October 2017), 49–57. https://doi.org/10.1016/j.foodchem.2018.02.020. Chawla, P., Deman, J. M., & Smith, A. K. (1990). Crystal morphology of shortenings and margarines. Food Structure, 9(4), 329–336. Chowdhury, R., Warnakula, S., Kunutsor, S., Crowe, F., Ward, H. A., Johnson, L., ... Di Angelantonio, E. (2014). Association of dietary, circulating, and supplement fatty acids with coronary risk: A systematic review and meta-analysis. Annals of Internal Medicine, 160(6), 398–406. https://doi.org/10.7326/M13-1788. Craven, J. R., & Lencki, R. W. (2011a). Crystallization, polymorphism, and binary phase behavior of model enantiopure and racemic 1,3-Diacylglycerols. Crystal Growth and Design, 11(5), 1566–1572. https://doi.org/10.1021/cg101536q. Craven, J. R., & Lencki, R. W. (2011b). Crystallization and polymorphism of 1,3-acylpalmitoyl-rac-glycerols. Journal of the American Oil Chemists' Society, 88(8), 1113–1123. https://doi.org/10.1007/s11746-011-1769-0. Czernichow, S., Thomas, D., & Bruckert, E. (2010). n-6 Fatty acids and cardiovascular health: A review of the evidence for dietary intake recommendations. The British Journal of Nutrition, 104(6), 788–796. https://doi.org/10.1017/ S0007114510002096. Dhaka, V., Gulia, N., Ahlawat, K. S., & Khatkar, B. S. (2011). Trans fats-sources, health risks and alternative approach – A review. Journal of Food Science and Technology, 48(5), 534–541. https://doi.org/10.1007/s13197-010-0225-8. Endo, J., & Arita, M. (2016). Cardioprotective mechanism of omega-3 polyunsaturated
9
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Saitou, K., Mitsui, Y., Shimizu, M., Kudo, N., Katsuragi, Y., & Sato, K. (2012). Crystallization behavior of diacylglycerol-rich oils produced from rapeseed oil. JAOCS Journal of the American Oil Chemists' Society, 89(7), 1231–1239. https://doi. org/10.1007/s11746-012-2014-1. Sato, K. (2001). Crystallization behaviour of fats and lipids — A review. Chemical Engineering Science, 56(7), 2255–2265. https://doi.org/10.1016/S0009-2509(00) 00458-9. Sato, K., Ueno, S., & Yano, J. (1999). Molecular interactions and kinetic properties of fats. Progress in Lipid Research, 38(1), 91–116. https://doi.org/10.1016/S0163-7827(98) 00019-8. Shannon, R. J., Fenerty, J., Hamilton, R. J., & Padley, F. B. (1992). The polymorphism of diglycerides. Journal of the Science of Food and Agriculture, 60(4), 405–417. https:// doi.org/10.1002/jsfa.2740600402. da Silva, R. C., Soares, F. A. S. D. M., Maruyama, J. M., Dagostinho, N. R., Silva, Y. A., Ract, J. N. R., & Gioielli, L. A. (2017). Crystallisation of monoacylglycerols and triacylglycerols at different proportions: Kinetics and structure. International Journal of Food Properties, 20(1), S385–S398. https://doi.org/10.1080/10942912.2017. 1297950. Singh, A., Auzanneau, F. I., & Rogers, M. A. (2017). Advances in edible oleogel technologies – A decade in review. Food Research International, 97(March), 307–317. https://doi.org/10.1016/j.foodres.2017.04.022. Smith, K., Bhaggan, K., Talbot, G., & van Malssen, K. F. (2011). Crystallization of fats: Influence of minor components and additives. Journal of the American Oil Chemists' Society, 88(5), 1085–1101. https://doi.org/10.1007/s11746-011-1819-7. Smith, K. W., & Sato, K. (2018). Effects of foreign and indigenous minor components. Crystallization of lipids (pp. 263–281). . de Souza, R. J., Mente, A., Maroleanu, A., Cozma, A. I., Ha, V., Kishibe, T., ... Anand, S. S. (2015). Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: Systematic review and metaanalysis of observational studies. British Medical Journal, 351, h3978. https://doi.org/ 10.1136/bmj.h3978. Verstringe, S., Dewettinck, K., Ueno, S., & Sato, K. (2014). Triacylglycerol crystal growth: Templating effects of partial glycerols studied with synchrotron radiation microbeam x-ray diffraction. Crystal Growth and Design, 14(10), 5219–5226. https://doi.org/10. 1021/cg5010209. Xu, Y., Wei, C., Zhao, X., Lu, C., & Dong, C. (2016). A comparative study on microstructure, texture, rheology, and crystallization kinetics of palm-based diacylglycerol oils and corresponding palm-based oils. European Journal of Lipid Science and Technology, 118(8), 1179–1192. https://doi.org/10.1002/ejlt.201500369.
fatty acids. Journal of Cardiology, 67(1), 22–27. https://doi.org/10.1016/j.jjcc.2015. 08.002. Fredrick, E., Foubert, I., De Van Sype, J., & Dewettinck, K. (2008). Influence of monoglycerides on the crystallization behavior of palm oil. Crystal Growth and Design, 8(6), 1833–1839. https://doi.org/10.1021/cg070025a. De Graef, V. (2009). Microstructural properties of isothermal palm oil crystallization. Universiteit Gent. Humphrey, K. L., Moquin, P. H. L., & Narine, S. S. (2003). Phase behavior of a binary lipid shortening system: From molecules to rheology. JAOCS Journal of the American Oil Chemists' Society, 80(12), 1175–1182. https://doi.org/10.1007/s11746-003-0838-2. Kwon, Y. (2016). Effect of trans–fatty acids on lipid metabolism: Mechanisms for their adverse health effects. Food Reviews International, 32(3), 323–339. https://doi.org/ 10.1080/87559129.2015.1075214. Lo, S. K., Tan, C. P., Long, K., Yusoff, M. S. A., & Lai, O. M. (2008). Diacylglycerol oilproperties, processes and products: A review. Food and Bioprocess Technology, 1(3), 223–233. https://doi.org/10.1007/s11947-007-0049-3. Marangoni, A. G. (2004). Crystallography. Fat crystal networks (pp. 1–20). New York, NY: Marcel Decker. Marangoni, A. G., & Narine, S. S. (2002). Identifying key structural indicators of mechanical strength in networks of fat crystals. Food Research International, 35(10), 957–969. https://doi.org/10.1016/S0963-9969(02)00158-8. Mensink, R. P., & Katan, M. B. (1990). Effect of dietary trans fatty acids on high-density and low-density lipoprotein cholesterol levels in healthy subjects. New England Journal of Medicine, 323(7), 439–445. https://doi.org/10.1056/ NEJM199008163230703. Mozaffarian, D., Aro, A., & Willett, W. C. (2009). Health effects of trans-fatty acids: Experimental and observational evidence. European Journal of Clinical Nutrition, 63, S5. Retrieved from https://doi.org/10.1038/sj.ejcn.1602973. Mozaffarian, D., Micha, R., & Wallace, S. (2010). Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: A systematic review and meta-analysis of randomized controlled trials. PLoS Medicine, 7(3), https://doi.org/ 10.1371/journal.pmed.1000252. Narine, S. S., & Marangoni, A. G. (1999). Relating structure of fat crystal networks to mechanical properties: A review. Food Research International, 32(4), 227–248. https://doi.org/10.1016/S0963-9969(99)00078-2. Oomen, C. M., Ocké, M. C., Feskens, E. J., van Erp-Baart, M. A., Kok, F. J., & Kromhout, D. (2001). Association between trans fatty acid intake and 10-year risk of coronary heart disease in the Zutphen Elderly Study: a prospective population-based study. Lancet, 357, 746–751. https://doi.org/10.1016/S0140-6736(00)04166-0.
10
Contents lists available at ScienceDirect
Food Research International journal homepage: www.elsevier.com/locate/foodres
Relating crystallization behavior of monoacylglycerols-diacylglycerol mixtures to the strength of their crystalline network in oil ⁎
Iris Taverniera, , Kim Moensa, Bart Heymanb, Sabine Danthinec, Koen Dewettincka a
Laboratory of Food Technology & Engineering, Department of Food technology, Safety and Health, Ghent University, Coupure Links 653, 9000 Gent, Belgium Vandemoortele R&D Center, Prins Albertlaan 79, 8870 Izegem, Belgium c Department of Food Science, Université de Liège, Gembloux, Belgium b
A R T I C LE I N FO
A B S T R A C T
Keywords: Fat crystallization Partial glycerides Oil structuring X-ray diffraction Rheology
Diacylglycerols (DAGs) are interesting oil structuring molecules as they are structurally similar to triacylglycerols (TAGs), but are metabolized differently which results in weight loss and improved blood cholesterol levels upon dietary replacement of TAGs with DAGs. Many commercial products consist of a mixture of monoacylglycerols (MAGs) and DAGs, yet the effect of MAGs on the crystallization behavior of DAGs is still to be unraveled. Two types of commercial MAGs, one originating from hydrogenated palm stearin and one of hydrogenated rapeseed oil, were added in concentrations 1, 2 and 4% to 20% DAGs derived from hydrogenated soybean oil. Using differential scanning calorimetry, it was shown that the presence of MAGs delayed the onset of DAG crystallization. Rheological analysis revealed that MAGs also hindered crystal network development. Synchrotron X-ray diffraction analysis demonstrated that the addition of MAGs suppressed the formation of the β form and stimulated the development of the β’ form. Likely, MAGs mainly hindered the crystallization of 1,3DAGs, which are responsible for the development of the β form, and stimulated the crystallization of the 1,2DAGs, which can crystallize in the α and β’ forms. The presence of two polymorphic forms resulted in a decrease of the crystal network strength, as was derived from oscillatory rheological measurements. This research implies a different effect of monoacylglycerols on both the nucleation and crystal growth of 1,2- and 1,3-DAG isomers. This insight is not only relevant for oleogelation research, but also for emulsifying agents which often contain blends of MAGs, 1,2-DAGs and 1,3-DAGs.
1. Introduction Traditionally, the lipid phase in food products such as margarines is structured with crystalline triacylglycerol (TAG) hardstock, containing a high amount of saturated fatty acids (SaFAs) or even trans fatty acids (tFAs). Studies have convincingly demonstrated that the consumption of tFAs increases low-density lipoprotein (LDL) blood levels, while lowering high-density lipoprotein (HDL) blood levels (Dhaka, Gulia, Ahlawat, & Khatkar, 2011). High levels of serum cholesterol, particularly LDL, promote the development of atherosclerosis or coronary heart disease (Kwon, 2016; Mensink & Katan, 1990; Mozaffarian, Aro, & Willett, 2009; Oomen et al., 2001). Whereas there is consensus on the adverse health effects of tFAs' consumption, uncertainty persists about the health effects of the consumption of SaFAs (Chowdhury et al., 2014; de de Souza et al., 2015). Research suggests that partially replacing these atherogenic SaFAs with poly-unsaturated fatty acids (PUFAs) reduces the risk of coronary heart
⁎
diseases (Czernichow, Thomas, & Bruckert, 2010; Endo & Arita, 2016; Mozaffarian, Micha, & Wallace, 2010). Altogether, it seems that atherogenic SaFAs should be reduced and tFAs completely eliminated and that both should be replaced with unsaturated fats to improve the nutritional profile of our diet and, as such, our cardiovascular health. These insights have led to intensive research on alternative oil structuring, using a wide variety of techniques and molecules (Singh, Auzanneau, & Rogers, 2017). Interesting oil structurants are diacylglycerols (DAGs), which are diesters of fatty acids with glycerin. Consequently, DAGs are similar to TAGs, which are triesters of fatty acids with glycerin. Research suggests that the dietary replacement of TAGs with DAGs leads to weight loss and improves blood cholesterol levels due to a difference in the way DAGs are metabolized compared to TAGs (Lo, Tan, Long, Yusoff, & Lai, 2008). Additionally, Xu, Wei, Zhao, Lu, and Dong (2016) demonstrated rheologically that fat blends containing DAGs had a stronger network structure with a more solid-like nature compared to the fats structured purely with TAGs.
Corresponding author at: Coupure Links 653, 9000 Gent, Belgium. E-mail address: [email protected] (I. Tavernier).
https://doi.org/10.1016/j.foodres.2018.10.092 Received 26 July 2018; Received in revised form 25 October 2018; Accepted 30 October 2018 0963-9969/ © 2018 Published by Elsevier Ltd.
Please cite this article as: Tavernier, I., Food Research International, https://doi.org/10.1016/j.foodres.2018.10.092
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
2. Material and methods
DAGs can exist in two isomers: 1,2- (or 2,3) DAGs and 1,3-DAGs. The equilibrated ratio of 1,2-DAG and 1,3-DAG is about 3:7 for longchain fatty acids (Lo et al., 2008). The crystallization behavior of 1,3DAGs and 1,2-DAGs differs significantly. While 1,2-DAGs can crystallize in α and β′ polymorphic forms in a hairpin configuration, the 1,3-DAGs are β-tending in a V-shaped configuration and crystallize in either a high-melting polymorphic form β1, or a lower-melting polymorphic form β2 (Shannon, Fenerty, Hamilton, & Padley, 1992). The development of the β polymorphic form during storage has been related to the occurrence of spherulitic crystals, causing oil loss and sandiness in fatrich products such as shortenings and margarines (Chawla, Deman, & Smith, 1990). Yet, if small β crystals are formed during for example a rapid cooling procedure instead of during storage, they are not necessarily an issue. Similar to β’ crystals, small β crystals are also capable of forming a stable crystal network structure with desired rheological properties. MAGs are known to influence fat crystallization, both in terms of nucleation, crystal growth and polymorphic form. The effect MAGs have on fat crystallization depends on their composition (saturated/ unsaturated and chain length), concentration and structural similarity to the dominant fat or oil (Alfutimie, Al-Janabi, Curtis, & Tiddy, 2016; da Silva et al., 2017; K. W. Smith & Sato, 2018). Fredrick, Foubert, Van Sype, and Dewettinck (2008) demonstrated the occurrence of heterogeneous nucleation of milk fat due to the presence of long-chain saturated MAGs. MAGs seem to be able to organize separately in micellar structures, acting as a template for nucleation processes (da Silva et al., 2017). On the other hand, also the crystal growth and α to β' polymorphic transition were accelerated with the presence of MAGs, which confirms that MAGs can affect the speed of polymorphic transition of fat crystals if they join the crystalline growth sites. Verstringe, Dewettinck, Ueno, and Sato (2014) confirmed the templating effect of monopalmitin on palm oil with microbeam synchrotron XRD. Chai et al. (2018) purified fully hydrogenated palm kernel oil and fully hydrogenated coconut oil and found that changes in nanostructure and polymorphic transformation took place after the removal of minor nontriacylglycerol polar lipids, MAGs, DAGs, and free fatty acids. In this research the goal was to identify whether MAGs affect the crystallization kinetics and polymorphic behavior of DAGs. Smith, Bhaggan, Talbot, and van Malssen (2011) argued that components with structural similarity have the most profound effect on the crystallization behavior of the main fat. Therefore, hydrogenated DAG of soybean origin was combined with two types of MAGs, one originating from hydrogenated palm oil and one from hydrogenated rapeseed oil. We hypothesized that the components with a similar fatty acid composition most efficiently influence each other's crystallization behavior and rheological properties. Hydrogenated MAGs and DAG were used because these have a simple fatty acid composition, making it easier to understand certain phenomena during crystallization. Furthermore, Alfutimie et al. (2016) demonstrated that saturated MAGs most efficiently influence fat crystallization, compared to unsaturated MAGs. The same type of MAG may exhibit a different effect on the crystallization (induction or retardation) depending on the proportion added (da Silva et al., 2017). Therefore, the MAGs were added in concentrations of 1, 2 and 4 wt%. In a first set of experiments, non-isothermal crystallization was studied with differential scanning calorimetry (crystallization curves), time-resolved synchrotron X-ray diffraction and oscillatory rheology (temperature sweeps). In a second set, the crystalline network and its properties were studied after one hour isothermal at 5 °C, using differential scanning calorimetry (melting curves), powder X-ray diffraction and oscillatory rheology (amplitude sweeps). As such, it was possible to correlate the crystallization behavior to the rheological properties of the samples.
2.1. Materials Commercially available DAG (Trancendim 110) was donated by Corbion Ingredients (Amsterdam, the Netherlands). Trancendim 110 (Tm,peak = 68.62 ± 0.07 °C) originates from fully hydrogenated soybean oil and is referred to as SO_DAG. Dimodan HP and Dimodan HR, commercially available, distilled MAGs, were kindly supplied by Danisco, DuPont (Wilmington, Delaware, USA). Dimodan HP (Tm,peak = 69.22 ± 0.12 °C) originates from fully hydrogenated palm stearin and is referred to as PSt_MAG. Dimodan HR (Tm,peak = 73.71 ± 0.13 °C) originates from fully hydrogenated rapeseed oil and is referred to as RO_MAG. Edible sunflower oil was kindly provided by Vandemoortele (Gent, Belgium). 2.2. Fatty acid analysis The fatty acid compositions of the SO_DAG, PSt_MAG and RO_MAG were analyzed with a GC-FID (gas chromatography with a flame ionization detector). First, fatty acid methyl esters (FAME) were prepared by dissolving 10 droplets of oil in 2 mL diethyl ether. Subsequently, 2 mL of 5% KOH in methanol was added to the mixture, which was allowed to react for 3 min. After the reaction, 2 mL demineralized water and 10 mL heptane were added. The mixture was shaken and the heptane layer, containing the FAME, was removed. This layer was washed twice in 4 mL demineralized water. After washing, the heptane layer was dried with a pinch of sodium sulphate. Subsequently, the sample was injected on the GC. FAME GC was carried out on a Varian GC with a BPX70 column (fused silica; 70% cyanopropyl polysilphenylene-siloxane, 50 m length, 0.22 mm internal diameter, 0.25 μm layer thickness). Helium was used as carrier gas. An amount of 1 μL sample solution was injected via the split/splitless injector (split ratio: 100) using automatic injection. The oven temperature was 187 °C for 45 min. Detection was performed with a flame ionization detector (FID). Table 1 gives an overview of the fatty acid composition of the SO_DAG, PSt_MAG and RO_MAG. 2.3. Carbon number analysis The amount of acylglycerols (mono-, di- and triacylglycerols or MAGs, DAGs and TAGs) was analyzed using carbon number GC. The commercial SO_DAG, PSt_MAG and RO_MAG were melted in an oven and 1 droplet (≈20 mg) was introduced in an Erlenmeyer of 50 mL. The fat was dissolved in 2 mL pyridine. To reduce the polarity of the partial acylglycerols, the samples were silylated. The silylation was performed by adding 0.4 mL of N-methyl-N-trimethylsilytrifluoroacetamide. Subsequently, the mixture was shaken and allowed to react for 10 to 15 min at 40 to 50 °C. The silylated products were redissolved in 40 mL Table 1 Fatty acid composition [%] and acylglycerol composition of PO_DAG, SO_DAG and PSt_MAG.
2
Component
SO_DAG
PSt_MAG
RO_MAG
C12:0 C14:0 C16:0 C18:0 C18:1 C20:0 C22:0 C24:0 FFAs MAGs DAGs TAGs
0 0.1 10.6 87.7 0.3 0.6 0.5 0.2 0.4 7.3 78.7 13.5
0.2 1.3 56.0 41.4 0.2 0.4 0 0 0 98.5 1.5 0
0 0.1 6.5 90.3 0.5 1.8 0.6 0 0 98.2 1.8 0
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
enclosed in glass capillaries and the temperature was controlled by a Linkam hot stage. The samples were subjected to the same time-temperature profile as for the non-isothermal DSC experiments: holding at 80 °C for 10 min, cooling at 10 °C/min to 5 °C.
isooctane. Carbon number GC was carried out using a Varian GC equipped with a CP SimDist column (fused silica, 10 m length, 0.53 mm internal diameter, 0.10 μm layer thickness). The injection volume was 1 μL. The oven temperature was programmed to cool from 50 °C to 350 °C at 25 °C/min and to subsequently keep isotherm at 350 °C for 6 min. Empower software was used to analyze the peaks. Table 1 gives an overview of the MAG, DAG and TAG composition of the SO_DAG, PSt_MAG and RO_MAG.
2.8. Powder X-ray diffraction spectroscopy The isothermal crystallization period was investigated by XRD using a Bruker D8-Advanced Diffractometer (Bruker, Germany) (l Cu = 1.54178\AA, 40 kV, and 30 mA), equipped with an Anton Paar temperature control system composed of a TTK 450 low-temperature chamber connected to a waterbath (Lauda) and heating device (TCU 110 Temperature Control Unit) (Anton Paar, Graz, Austria). Similar to the DSC measurements, the samples were heated to 80 °C for 10 min and subsequently cooled at a cooling rate of 10 °C/min to 5 °C. The samples were kept isothermally in the equipment at 5 °C during 1 h. WAXD and SAXS were recorded using a Vantec-1 detector (Bruker, Germany) during the isothermal crystallization. D-values were directly calculated by Diffraction Suite Eva software. The XRD patterns were analyzed using PeakFit (SeaSolve Software inc., Framingham, USA).
2.4. Preparation of the monoacylglycerol – diacylglycerol blends Different concentrations of MAGs (1, 2 and 4%) combined with 20% DAGs were dispersed in sunflower oil, resulting in SO_DAG:PSt_MAG and SO_DAG:RO:MAG combinations. In addition, samples of 21, 22 and 24% SO_DAG were prepared. The dispersions were stirred with a magnetic stirrer at 80 °C until a homogeneous samples were obtained. When the blends were visibly free of dispersed material, they were further mixed for 30 min to ensure complete melting. These homogeneous solutions were analyzed with a rheometer, with Differential Scanning Calorimetry (DSC) and with synchrotron X-ray diffraction. The molten samples at the same temperature were analyzed and underwent the same time-temperature profile. As such, a controlled and equal cooling rate can be ensured for all the analysis and all the samples.
3. Results and discussion 3.1. Crystallization mechanism of SO_DAG The crystallization curve and synchrotron XRD patterns (SAXS and WAXD) of 20% SO_DAG are visualized in Fig. 1. These curves were obtained by cooling the samples from 80 to 5 °C at a cooling rate of 10 °C/min. The main crystallization peak of SO_DAG has an onset temperature of 50.97 ± 0.45 °C. A second, smaller peak can be discerned at lower temperatures (Tc = 28.06 ± 0.20 °C). As previously mentioned, SO_DAG is a commercial sample and contains relatively high MAG and TAG fractions. Fig. 2 demonstrates that the addition of more MAGs does not increase the intensity of this second crystallization peak. Therefore, we are not inclined to relate this peak to MAG crystallization. Alternatively, the second peak might be related to the crystallization of 1,2-DAGs which generally crystallize at temperatures lower than 1,3-DAGs (Lo et al., 2008). Saitou et al. (2012) used gas chromatography to determine the amount of crystalline matter originating from 1,3-DAGs and 1,2-DAGs in a DAG oil. They found that 35% of the total DAG fraction in terms of chemical composition were 1,2DAGs while the crystalline mass of the DAG oil consisted for only 4% out of 1,2-DAGs. The 1,2-DAG crystal fraction was much lower than expected based upon the chemical composition. As visualized in Fig. 1, the peak related to the crystallization of 1,2-DAGs is therefore very small. Apart from the crystallization of 1,2-DAG, also TAGs crystallize within this temperature range. We therefore related this smaller peak to the crystallization of both 1,2-DAGs and TAGs. In Fig. 1b and c, clear diffraction peaks can be discerned starting from 47 °C which is close to the crystallization temperature of the first peak in the cooling profile. In the WAXD profile, these first peaks are identified at short-spacings d = 3.84 Å, d = 4.53 Å and d = 4.59 Å, which are related to the appearance of a β polymorphic form (Craven & Lencki, 2011b; Saitou et al., 2012; Shannon et al., 1992). Shannon et al. (1992) identified two β forms (β1 and β2) with short spacing values of 3.74, 3.82, 4.55 and 4.64 Å for β1 and 3.69, 3.78, 3.90, 4.55, and 4.68 Å for β2 of the 1,3-DAGs. These short spacing values are situated very closely together and a distinction between the high-melting and lowmelting β polymorphic forms was not possible in our results. Starting from 36.2 °C, which is the onset temperature of the second crystallization peak on the cooling profile, a lower intensity peak at shortspacing d = 4.17 Å is distinguished. This peak is indicative of the α polymorphic form, which could result from TAG and 1,2-DAG crystallization as discussed previously (Sato, 2001). In the SAXS diffraction pattern, only one peak is identified (d = 52.05 Å), which is related to the occurrence of 2 crystal layers (2 L) (Marangoni & Narine, 2002).
2.5. Differential scanning calorimetry The thermal profiles of the MAG:DAG blends were obtained with a Q1000 DSC (TA Instruments, New Castle, Delaware, USA) equipped with a refrigerated cooling system. Nitrogen was used as purge gas. The cell constant and temperature were set with indium (TA Instruments). An additional temperature calibration was done using azobenzene (Sigma-Aldrich, Bornem, Belgium) and undecane (Acros Organics, Geel, Belgium). The sample (weighing between 6.0 and 8.0 mg/cup) was placed inside an aluminum pan and sealed with an aluminum lid (TA Instruments). The experiments consisted of heating to 80 °C and cooling to 5 °C at a cooling rate of 10 °C/min, which resulted in the cooling curves. Subsequently, the samples were kept isothermal during one hour at 5 °C and then heated again at 10 °C/min to 80 °C. The thermal analyses were performed in triplicate. 2.6. Rheology The rheological behavior was analyzed on the advanced rheometer AR2000ex (TA Instruments, New Castle, USA) equipped with a Peltier system and water bath (Julabo, Seelbach, Germany) for temperature control. The starch pasting cell (shear rate factor = 4.500 s−1, shear stress factor = 48,600 1/m3, gap = 5500 μm) was utilized to record the gelling points during cooling from 80 to 5 °C at a cooling rate of 10 °C/ min. The oleogels were then kept isothermal at 5 °C for 1 h. After this isothermal period, the samples were subjected to an amplitude sweep by increasing the oscillation stress from 0.01 to 1000 Pa at a frequency of 1.0 Hz. The rheological analysis were performed in triplicate. 2.7. Synchrotron X-ray diffraction The non-isothermal crystallization of the samples was studied by Xray diffraction (XRD) using synchrotron radiation. Wide-angle X-ray diffraction (WAXD) and small-angle X-ray scattering (SAXS) measurements were performed on the Dutch-Belgian (DUBBLE) beamline BM26 at the European Synchrotron Facility (ESRF) in Grenoble (France). The experiments were performed at a fixed wavelength λ of 1.24 Å. For recording WAXD, a linear 300 K photon counting Pilatus detector was used, while for collecting the SAXS images, a large area high sensitive photon counting 2D Pilatus detector was used. The samples were 3
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 1. (a) Crystallization curve, (b) SAXS and (c) WAXD patterns of 20% SO_DAG during cooling to 5 °C.
3.2. Effect of monoacylglycerols on the crystallization of diacylglycerols
the peak temperatures and the peak widths at half height of the main peak for the DAG:MAG combinations. The addition of MAGs to SO_DAG shifts the onset of crystallization to lower temperatures, likely through attractive interactions between the MAG and DAG molecules which might allow the MAG to crystallize on the surface of the early seed
Fig. 2 visualizes the crystallization curves recorded during cooling of 20% SO_DAG, blended with 1, 2 and 4% PSt_MAG and with 1, 2 and 4% RO_MAG and Table 2 gives an overview of the onset temperatures,
Fig. 2. Crystallization curves recorded during cooling of SO_DAG, blended with 1, 2 or 4% PSt_MAG (a) or with 1, 2 or 4% RO_MAG (b). 4
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
onset of DAG crystallization. Yet, Fig. 2 does reveal a shoulder adjacent to the main peak for the samples containing 4% RO_MAG and PSt_MAG. This peak is likely related to the MAG crystallization and could indicate the presence of two overlapping peaks, one originating from MAG and one from DAG crystallization. To better understand the effect of MAGs on the crystallization of SO_DAG, the combinations were analyzed using synchrotron XRD. The samples were cooled and simultaneously measured using the same time-temperature profile during synchrotron XRD measurements as for DSC. In Fig. 3 and Fig. 4, a clear effect of the PSt_MAG and RO_MAG addition on the polymorphism of SO_DAG is visualized. Both types of MAGs give a similar effect on the crystallization of SO_DAG. With the addition of MAGs, the crystallization in the β polymorph (peak at 4.50–4.59 Å) seems to be suppressed, the peak at 4.15–4.18 Å becomes more dominant during cooling, while the peak at 3.79–3.86 Å remains constant or increases in intensity. Increasing MAG concentrations results in an earlier onset of the peak at 4.09–4.14 Å, which is why we attribute this peak to the crystallization of MAG in the α polymorphic form. The crystallization curves also revealed the presence of an initial shoulder at concentrations of 4% PSt_MAG and RO_MAG, which can now be related to the presence of the α polymorphic form. The initial α crystals transform into a β’ form since the initial α peak disappears and the peaks at 3.79–3.86 Å and 4.15–4.18 Å develop increasingly during cooling. For the samples where 1 and 2% PSt_MAG were added, three peaks are still present (3.84–3.86 Å, 4.15–4.16 Å and 4.54–4.51 Å) at the end of the cooling procedure, which indicates the presence of both β and β′ crystals. As discussed previously, SO_DAG is a mixture of MAGs, TAGs, 1,2DAGs and 1,3-DAGs, which have different crystal packings. Possibly, the addition of MAGs stimulates the crystallization of one component
Table 2 Crystallization onset and peak temperatures of 20SO_DAG, blended with 1, 2, 4% of PSt_MAG and 1, 2, 4% of RO_MAG. Tc,
20SO_DAG 20SO_DAG:1PSt_MAG 20SO_DAG:2PSt_MAG 20SO_DAG:4PSt_MAG 20SO_DAG:1RO_MAG 20SO_DAG:2RO_MAG 20SO_DAG:4RO_MAG
onset
50.96 49.50 47.68 48.02 49.97 48.88 48.65
(°C)
± ± ± ± ± ± ±
0.45a 0.21bc 0.12f 0.06ef 0.28b 0.21cd 0.19de
Tc,
peak
49.43 46.50 46.12 44.47 48.52 45.51 46.17
(°C)
± ± ± ± ± ± ±
0.71a 0.78c 0.01cd 0.10e 0.33b 0.19d 0.08cd
Peak width at half maximum (°C) 2.96 3.63 4.33 5.06 3.60 5.11 5.32
± ± ± ± ± ± ±
0.28e 0.07d 0.12c 0.02b 0.07d 0.12ab 0.10a
Different letters, within the same column, indicate significant differences at p $ < $ 0.05 according to ANOVA analysis for normally distributed data with equal variances and according to the Mann-Whitney U test for not normally distributed data and/or data without equal variances (SPSS 22).
crystals of the DAG or vice versa (K. W. Smith & Sato, 2018). As such, the formation of nuclei might be hindered. The distinct peak at lower crystallization temperatures for the SO_DAG thermogram disappears with the addition of MAGs and forms a tail of the main, high-melting peak. The addition of MAGs causes the peak width at half height to broaden (Table 2). According to Humphrey, Moquin, and Narine (2003), this indicates the presence of an inhibitor to nucleation. A wider crystallization peak implies that the final crystals have grown from several types of nuclei, as nucleation occurred over a larger time span. A more narrow crystallization peak suggests that all crystals within the sample nucleated at approximately the same time. In contrast to the work of da Silva et al. (2017), Table 2 does not reveal a significantly different effect of the different MAG concentrations on the
Fig. 3. WAXD patterns of SO_DAG containing (a) no PSt_MAG, (b) 1% PSt_MAG, (c) 2% PSt_MAG and (d) 4% PSt_MAG during cooling to 5 °C. 5
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 4. WAXD patterns of SO_DAG containing (a) no RO_MAG, (b) 1% RO_MAG, (c) 2% RO_MAG and (d) 4% RO_MAG during cooling to 5 °C.
unequivocally determine the effect of pure MAGs on either 1,2-DAGs or 1,3-DAGs. In the SAXS pattern of SO_DAG blended with different PSt_MAG concentrations, peaks are distinguished at d = 52.24 Å (1% PSt_MAG), d = 52.43 Å (2% PSt_MAG) and d = 52.73 Å (4% PSt_MAG) (Fig. S1). For the blends of SO_DAG with RO_MAG, peaks are found at d = 52.43 Å (1% RO_MAG), d = 52.63 Å (2% RO_MAG) and d = 53.12 Å (4% RO_MAG) (Fig. S2). With the addition of more MAGs, the peak position shifts to slightly higher values but remains within the 2 L range. Likely, the angle between the acyl chains increases with the incorporation of MAGs in the crystal lattice, which causes the layer thickness to increase. The layer thickness is slightly smaller for 20% SO_DAG combined with PSt_DAG compared to SO_DAG combined with RO_DAG which can be related to the bigger chain length of the constituents as more stearic acid is present in RO_MAG. Long spacings increase with increasing chain length (Marangoni, 2004).
Fig. 5. Molecular packing of 1,3-DAGs (left) and 1,2-DAGs (right) and effect of MAG addition.
3.3. Rheological behavior of MAG:DAG combinations during crystallization
but hinders the crystallization of another. The existing literature on DAG crystallization states that 1,3-DAGs can only occur in β polymorphic forms and 1,2-DAGs in α and β′ polymorphic forms (Craven & Lencki, 2011a; Craven & Lencki, 2011b; Shannon et al., 1992). Fig. 5 schematically represents the hypothesized effect of MAG addition on DAG crystallization. Based on this representation, one can understand how the presence of MAGs might hinder the crystallization of 1,3-DAGs, but have a different effect on the crystallization of 1,2-DAGs. MAGs are likely more easily built in the crystal lattice of 1,2-DAGs compared to 1,3-DAGs. However, more fundamental research is required to
This pronounced effect of MAG addition on the crystallization behavior of SO_DAG should also be represented in the rheological properties, which was investigated by recording the complex modulus and phase angle during cooling in the rheometer using the same timetemperature profile as used for the DSC analysis and synchrotron XRD measurements (Fig. 6). A higher |G*| indicates a stronger and more rigid network structure in the material. The phase angle δ evaluates the visco-elastic properties of a system. In a pure viscous system, the phase angle is 90°, while the phase angle is 0° in a pure elastic system. At a 6
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 6. Gelling curves for SO_DAG blended with PSt_MAG (left) and RO_MAG (right), as determined with oscillatory rheology. The full lines represent the complexes modulus, the dashed lines correspond to the phase angle.
Fig. 7. Melting curves of SO_DAG blended with 1, 2 and 4% of PSt_MAG (top) and with 1, 2 and 4% of RO_MAG (bottom), isothermally crystallized for 1 h at 5 °C. An offset of 0.2, 0.4 and 0.6 was used.
phase angle of 45°, the solid-like properties are equal to the liquid-like properties. The structure of the 20% SO_DAG sample is formed almost
instantaneously in one stage, which can be related to the fast crystallization of the 1,3-DAGs in the stable β polymorphic form. Both PSt_MAG and RO_MAG have a similar effect on the network built-up of 7
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Fig. 8. WAXD patterns of SO_DAG blended with 1, 2 and 4% of PSt_MAG (top) and with 1, 2 and 4% of RO_MAG (bottom), isothermally crystallized for 1 h at 5 °C. An offset was used to enhance visualization of the patterns.
SO_DAG. When PSt_MAG and RO_MAG are added, the onset temperature of network built-up is not significantly impacted, but the network development is slower and occurs in two-stages instead of one stage. The occurrence of two stages can be derived from |G*| curves, which shows two steep slope increases instead of one. The occurrence of two stages might indicate network rearrangements during cooling (De De Graef, 2009). The α to β′ polymorphic transition observed in Figs. 3 and 4 likely explains the two-stage network built-up. Furthermore, the final network is less strong for samples containing additional MAGs. This might be related to the presence of various polymorphic forms. For SO_DAG without additional MAG, the β form provides a strong and rigid network. With the presence of MAGs, both β and β′ forms are developed. Interestingly, the addition of 4% MAGs again improves the network built-up. At this concentration, MAGs are capable of structuring oil into a high viscosity liquid (Bin Sintang, Rimaux, Van de Walle, Dewettinck, & Patel, 2016). Both MAG and DAG might simultaneously and separately structure oil, which can explain the improved structural built-up. As RO_MAG contains more stearic acid than PSt_MAG (Table S1), structuring for RO_MAG will occur at a lower concentration. This can explain why this effect is only observed for RO_MAG at a concentration of 4%.
Fig. 9. Overview of the complex modulus of the pure SO_DAG and of SO_DAG combined with PSt_MAG and RO_MAG in concentrations of 1, 2 and 4%.
8
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
DAGs by employing pure 1,2-DAGs and 1,3-DAGs instead of blends.
3.4. Crystal network properties of MAG:DAG combinations after 1 h isothermal at 5 °C
4. Conclusion The short-term stability of the β′ crystals was assessed by keeping the samples isothermal at 5 °C for 1 h. The samples were evaluated using melting curves (Fig. 7) and powder XRD (Fig. 8). Thermodynamically, the β form is most stable, β′ is metastable and α is the least stable form, which results in a lower melting point for the α form, an intermediate melting point for the β′ form and the highest melting point for the β form (Sato, Ueno, & Yano, 1999). The addition of PSt_MAG or RO_MAG to SO_DAG clearly had an effect on the melting behavior of the blends (Fig. 7). The melting curve shifts to lower temperatures with an increasing PSt_MAG or RO_MAG concentration and the peak broadens for increasing concentrations of PSt_MAG or RO_MAG. This broadening peak indicates the presence of both β and β′ polymorphic forms for all combinations and is likely caused by the presence of crystalline 1,2-DAGs and 1,3-DAGs. The melting peak for the combinations with RO_MAG is less broad than that of combinations with PSt_MAG, which could be associated with the narrower fatty acid profile of RO_MAG:SO_DAG combinations or with a different balance in the amount of β to β′ polymorphs. For the addition of 1 and 2% RO_MAG, the melting peak only shifts slightly, which might indicate a lower amount of β′ compared to 4%. The melting enthalpies for the different samples were not significantly different (data not shown). A definite conclusion on polymorphic forms can be drawn from the powder XRD patterns (Fig. 8). At an increasing PSt_MAG or RO_MAG concentration, the formation of the β form is reduced, as can be deduced from the reduced intensity of the double peak at 4.55 and 4.63 Å, which is related to the β form. At a concentration of 4% PSt_MAG or RO_MAG, the peaks related to the β′ form at 3.83 and 4.20 Å for combinations with PSt_MAG and 3.85 and 4.17 Å for combinations with RO_MAG become more dominant. The intensity of these β′ peaks for the combination with PSt_MAG is higher than that of the combination with RO_MAG. For the concentrations of 1 and 2% RO_MAG, the double peak at 4.55 and 4.63 Å is still present, which is indicative of a substantial amount of β. PSt_MAG results in more β′ form than the RO_MAG, which supports the results obtained from the melting curves. The type of polymorph present in the fat crystal network also significantly affects the macroscopic properties of the fat (Narine & Marangoni, 1999). To assess the effect of MAG addition on the rheological properties of the blends, the samples were subjected to oscillatory rheology (amplitude sweeps). Fig. 9 depicts the complex modulus (|G*|) of the various samples after 1 h of isothermal time at 5 °C. The presence of multiple components and two polymorphic forms seems to interfere with the formation of a good crystal network structure. Interestingly, at a concentration of 4% RO_MAG, the |G*| increases again. Fig. 6 already demonstrated that a concentration of 4% RO_MAG hindered the network development less than a 1 and 2% RO_MAG. We hypothesized this was related to separate oil structuring by 4% RO_MAG. Therefore, increasing the MAG concentration might further enhance the crystal network strength. In this research, the focus was on fully hydrogenated MAGs and DAGs. These products are applied today in margarines as emulsifying agents. Yet up to date, it was unclear how they interacted with each other. On the other hand, to fully replace the traditional TAG-based fat with MAG:DAG combinations, the hydrogenated MAGs and DAGs studied here are likely less applicable as their melting temperature is higher than 50 °C, which could result in undesired melting properties in the mouth. Furthermore, bakery margarines and shortenings are generally prepared at high cooling rates and under shears and are subsequently stored at temperatures between 15 and 20 °C. To assess the applicability of these systems, more unsaturated components should therefore be studied and the effects of industrial processing (preparation and storage) should be researched. Further fundamental research should investigate the distinctive effect of MAGs on 1,2-DAGs and 1,3-
This research studied the influence of MAGs on the crystallization behavior of DAGs. The addition of MAGs delayed the onset of nucleation and retarded the crystal network development of the DAG-based system during cooling. Commercial DAG samples are a mixture of approximately 70% 1,3-DAGs and 30% 1,2-DAGs, with the 1,3-DAGs dominating the crystallization behavior. While 1,3-DAGs are known to crystallize in a β polymorphic form, 1,2-DAGs crystallize in the α and β′ forms. We found that the addition of MAGs induced the DAGs to crystallize in a β′ form during cooling. This β′ form remained present after storing the samples isothermally, as was deduced from the melting behavior and powder XRD analysis. A commercial monoglyceride originating from hydrogenated palm stearin more efficiently enabled the maintenance of the β′ polymorph, compared to a commercial monoglyceride originating from hydrogenated rapeseed oil. Possibly, MAGs enhanced crystallization of 1,2-DAGs in the β′ form but hinder the crystallization of 1,3-DAGs in the β polymorphic form. This may have shifted the balance of crystalline matter from being mainly composed out of 1,3-DAGs, as is the natural tendency, to being composed out of 1,2-DAGs. For fundamental understanding of the system more work is required on combinations of MAGs with pure 1,2-DAGs and pure 1,3DAGs. Supplementary data to this article can be found online at https:// doi.org/10.1016/j.foodres.2018.10.092. Acknowledgements This research was made possible thanks to the BOF (Special Research Fund) of Ghent University. The authors acknowledge the ESRF (Grenoble, France) for the use of the synchrotron facilities. The Dutch-Belgian Beamline (DUBBLE) research group at the ESRF and the Dutch organization for scientific research (N.W·O) are acknowledged for their help and continuous support of the DUBBLE project. Chi Diem Doan, Mohd Dona Bin Sintang and Ruben Van Overbeke are acknowledged for their help with the synchrotron XRD experiments. Joost Coudron of Vandemoortele is thanked for his help with the GC analysis. References Alfutimie, A., Al-Janabi, N., Curtis, R., & Tiddy, G. J. T. (2016). The effect of monoglycerides on the crystallisation of triglyceride. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 494, 170–179. https://doi.org/10.1016/j. colsurfa.2016.01.029. Bin Sintang, M. D., Rimaux, T., Van de Walle, D., Dewettinck, K., & Patel, A. R. (2016). Oil structuring properties of monoglycerides and phytosterols mixtures. European Journal of Lipid Science and Technology, 1–14. https://doi.org/10.1002/ejlt.201500517. Chai, X. H., Meng, Z., Cao, P. R., Liang, X.y., Piatko, M., Campbell, S., ... Fa, L. Y. (2018). Influence of indigenous minor components on fat crystal network of fully hydrogenated palm kernel oil and fully hydrogenated coconut oil. Food Chemistry, 255(October 2017), 49–57. https://doi.org/10.1016/j.foodchem.2018.02.020. Chawla, P., Deman, J. M., & Smith, A. K. (1990). Crystal morphology of shortenings and margarines. Food Structure, 9(4), 329–336. Chowdhury, R., Warnakula, S., Kunutsor, S., Crowe, F., Ward, H. A., Johnson, L., ... Di Angelantonio, E. (2014). Association of dietary, circulating, and supplement fatty acids with coronary risk: A systematic review and meta-analysis. Annals of Internal Medicine, 160(6), 398–406. https://doi.org/10.7326/M13-1788. Craven, J. R., & Lencki, R. W. (2011a). Crystallization, polymorphism, and binary phase behavior of model enantiopure and racemic 1,3-Diacylglycerols. Crystal Growth and Design, 11(5), 1566–1572. https://doi.org/10.1021/cg101536q. Craven, J. R., & Lencki, R. W. (2011b). Crystallization and polymorphism of 1,3-acylpalmitoyl-rac-glycerols. Journal of the American Oil Chemists' Society, 88(8), 1113–1123. https://doi.org/10.1007/s11746-011-1769-0. Czernichow, S., Thomas, D., & Bruckert, E. (2010). n-6 Fatty acids and cardiovascular health: A review of the evidence for dietary intake recommendations. The British Journal of Nutrition, 104(6), 788–796. https://doi.org/10.1017/ S0007114510002096. Dhaka, V., Gulia, N., Ahlawat, K. S., & Khatkar, B. S. (2011). Trans fats-sources, health risks and alternative approach – A review. Journal of Food Science and Technology, 48(5), 534–541. https://doi.org/10.1007/s13197-010-0225-8. Endo, J., & Arita, M. (2016). Cardioprotective mechanism of omega-3 polyunsaturated
9
Food Research International xxx (xxxx) xxx–xxx
I. Tavernier et al.
Saitou, K., Mitsui, Y., Shimizu, M., Kudo, N., Katsuragi, Y., & Sato, K. (2012). Crystallization behavior of diacylglycerol-rich oils produced from rapeseed oil. JAOCS Journal of the American Oil Chemists' Society, 89(7), 1231–1239. https://doi. org/10.1007/s11746-012-2014-1. Sato, K. (2001). Crystallization behaviour of fats and lipids — A review. Chemical Engineering Science, 56(7), 2255–2265. https://doi.org/10.1016/S0009-2509(00) 00458-9. Sato, K., Ueno, S., & Yano, J. (1999). Molecular interactions and kinetic properties of fats. Progress in Lipid Research, 38(1), 91–116. https://doi.org/10.1016/S0163-7827(98) 00019-8. Shannon, R. J., Fenerty, J., Hamilton, R. J., & Padley, F. B. (1992). The polymorphism of diglycerides. Journal of the Science of Food and Agriculture, 60(4), 405–417. https:// doi.org/10.1002/jsfa.2740600402. da Silva, R. C., Soares, F. A. S. D. M., Maruyama, J. M., Dagostinho, N. R., Silva, Y. A., Ract, J. N. R., & Gioielli, L. A. (2017). Crystallisation of monoacylglycerols and triacylglycerols at different proportions: Kinetics and structure. International Journal of Food Properties, 20(1), S385–S398. https://doi.org/10.1080/10942912.2017. 1297950. Singh, A., Auzanneau, F. I., & Rogers, M. A. (2017). Advances in edible oleogel technologies – A decade in review. Food Research International, 97(March), 307–317. https://doi.org/10.1016/j.foodres.2017.04.022. Smith, K., Bhaggan, K., Talbot, G., & van Malssen, K. F. (2011). Crystallization of fats: Influence of minor components and additives. Journal of the American Oil Chemists' Society, 88(5), 1085–1101. https://doi.org/10.1007/s11746-011-1819-7. Smith, K. W., & Sato, K. (2018). Effects of foreign and indigenous minor components. Crystallization of lipids (pp. 263–281). . de Souza, R. J., Mente, A., Maroleanu, A., Cozma, A. I., Ha, V., Kishibe, T., ... Anand, S. S. (2015). Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: Systematic review and metaanalysis of observational studies. British Medical Journal, 351, h3978. https://doi.org/ 10.1136/bmj.h3978. Verstringe, S., Dewettinck, K., Ueno, S., & Sato, K. (2014). Triacylglycerol crystal growth: Templating effects of partial glycerols studied with synchrotron radiation microbeam x-ray diffraction. Crystal Growth and Design, 14(10), 5219–5226. https://doi.org/10. 1021/cg5010209. Xu, Y., Wei, C., Zhao, X., Lu, C., & Dong, C. (2016). A comparative study on microstructure, texture, rheology, and crystallization kinetics of palm-based diacylglycerol oils and corresponding palm-based oils. European Journal of Lipid Science and Technology, 118(8), 1179–1192. https://doi.org/10.1002/ejlt.201500369.
fatty acids. Journal of Cardiology, 67(1), 22–27. https://doi.org/10.1016/j.jjcc.2015. 08.002. Fredrick, E., Foubert, I., De Van Sype, J., & Dewettinck, K. (2008). Influence of monoglycerides on the crystallization behavior of palm oil. Crystal Growth and Design, 8(6), 1833–1839. https://doi.org/10.1021/cg070025a. De Graef, V. (2009). Microstructural properties of isothermal palm oil crystallization. Universiteit Gent. Humphrey, K. L., Moquin, P. H. L., & Narine, S. S. (2003). Phase behavior of a binary lipid shortening system: From molecules to rheology. JAOCS Journal of the American Oil Chemists' Society, 80(12), 1175–1182. https://doi.org/10.1007/s11746-003-0838-2. Kwon, Y. (2016). Effect of trans–fatty acids on lipid metabolism: Mechanisms for their adverse health effects. Food Reviews International, 32(3), 323–339. https://doi.org/ 10.1080/87559129.2015.1075214. Lo, S. K., Tan, C. P., Long, K., Yusoff, M. S. A., & Lai, O. M. (2008). Diacylglycerol oilproperties, processes and products: A review. Food and Bioprocess Technology, 1(3), 223–233. https://doi.org/10.1007/s11947-007-0049-3. Marangoni, A. G. (2004). Crystallography. Fat crystal networks (pp. 1–20). New York, NY: Marcel Decker. Marangoni, A. G., & Narine, S. S. (2002). Identifying key structural indicators of mechanical strength in networks of fat crystals. Food Research International, 35(10), 957–969. https://doi.org/10.1016/S0963-9969(02)00158-8. Mensink, R. P., & Katan, M. B. (1990). Effect of dietary trans fatty acids on high-density and low-density lipoprotein cholesterol levels in healthy subjects. New England Journal of Medicine, 323(7), 439–445. https://doi.org/10.1056/ NEJM199008163230703. Mozaffarian, D., Aro, A., & Willett, W. C. (2009). Health effects of trans-fatty acids: Experimental and observational evidence. European Journal of Clinical Nutrition, 63, S5. Retrieved from https://doi.org/10.1038/sj.ejcn.1602973. Mozaffarian, D., Micha, R., & Wallace, S. (2010). Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: A systematic review and meta-analysis of randomized controlled trials. PLoS Medicine, 7(3), https://doi.org/ 10.1371/journal.pmed.1000252. Narine, S. S., & Marangoni, A. G. (1999). Relating structure of fat crystal networks to mechanical properties: A review. Food Research International, 32(4), 227–248. https://doi.org/10.1016/S0963-9969(99)00078-2. Oomen, C. M., Ocké, M. C., Feskens, E. J., van Erp-Baart, M. A., Kok, F. J., & Kromhout, D. (2001). Association between trans fatty acid intake and 10-year risk of coronary heart disease in the Zutphen Elderly Study: a prospective population-based study. Lancet, 357, 746–751. https://doi.org/10.1016/S0140-6736(00)04166-0.
10