Phase diagrams of mixtures of n-hentriacontane and saturated monoacid triacylglycerols

Journal Pre-proof Phase diagrams of mixtures of n-hentriacontane and saturated monoacid triacylglycerols C. Tapia-Ledesma, S.B. Araujo-D´ıaz, E. Dibildox-Alvarado, J.J. ´ Ornelas-Paz, J.D. Perez-Mart´ ınez

PII:

S0040-6031(19)30657-4

DOI:

https://doi.org/10.1016/j.tca.2019.178455

Reference:

TCA 178455

To appear in:

Thermochimica Acta

Received Date:

24 July 2019

Revised Date:

29 October 2019

Accepted Date:

6 November 2019

Please cite this article as: Tapia-Ledesma C, Araujo-D´ıaz SB, Dibildox-Alvarado E, ´ Ornelas-Paz JJ, Perez-Mart´ ınez JD, Phase diagrams of mixtures of n-hentriacontane and saturated monoacid triacylglycerols, Thermochimica Acta (2019), doi: https://doi.org/10.1016/j.tca.2019.178455

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Phase diagrams of mixtures of n-hentriacontane and saturated monoacid triacylglycerols Tapia-Ledesma C.a, Araujo-Díaz S. B.a, Dibildox-Alvarado E.a, Ornelas-Paz J. J.b, Pérez-Martínez J. D.a*

a

of

Facultad de Ciencias Químicas, Centro de Investigación y Estudios de Posgrado, Universidad Autónoma de San Luis Potosí, San Luis Potosí, 78210, Mexico b

ro

Centro de Investigación en Alimentación y Desarrollo, A.C., Av. Río Conchos S/N, Parque Industrial, Cd. Cuauhtemoc, Chihuahua, 31570, Mexico

re

-p

† Electronic supplemental file

Jo

ur na

lP

* Corresponding author. Tel: + 52 444 8262440 ext. 6432 E-mail: [email protected] (Jaime David Pérez-Martínez) Mailing address: Facultad de Ciencias Químicas-CIEP Zona Universitaria Av. Dr. Manuel Nava 6, Zona Universitaria San Luis Potosí, SLP 78210, México.

1

-p

ro

of

Graphical abstract

HIGHLIGHTS

n-hentriacontane and monoacid saturated TAGs formed eutectic mixtures



The eutectic composition changed with TAG melting temperature



n-hentriacontane affected the microstructure of TAG’s crystals



Heterogeneous nucleation caused microstructural changes of MMM crystals



Inhibition of crystal growth caused microstructural changes of PPP crystals



Inhibition of crystal growth caused microstructural changes of SSS crystals

Jo

ur na

lP

re



2

ABSTRACT Gelation of vegetable oils with candelilla wax (CW) and saturated triacylglycerols (TAGs) is a promising strategy for engineering fat-based foods, cosmetics, and drug delivery systems. n-hentriacontane (C31) defines, to a large extent, the thermal properties of CW. Therefore, the phase behavior and microstructure of binary mixtures

of

of C31 and trimyristin (MMM), tripalmitin (PPP), or tristearin (SSS) were studied to provide a better understanding of complex mixtures of CW and saturated fats. Phase

ro

diagrams were developed by differential scanning calorimetry and X-ray diffraction

-p

measurements. The mixtures showed a eutectic phase behavior, in which the C31-TAG mixtures below the solidus line were solid dispersions of crystals rich in C31 and

re

crystals rich in the corresponding TAG. This segregation in the solid phase was attributed to the incompatible subcell-packing of the pure components in the stable

lP

polymorphic form, which was orthorhombic for C31, and monoclinic for TAGs in the polymorphic form. C31 had a major effect on TAGs microstructure. Morphological

ur na

changes at the microstructural level were caused by the heterogeneous nucleation of MMM on C31 crystals. Conversely, microstructural changes for TAGs crystals in C31PPP and C31-SSS mixtures were induced by the inhibition of crystal growth due to the

Jo

transient adsorption of C31 molecules at the growing crystal surface. Keywords: Phase diagram, n-hentriacontane, tristearin, tripalmitin, candelilla wax.

3

1. INTRODUCTION The relationship between different scales of structure, physical properties, and stability of materials determines the success of matrices used in food, drug and cosmetic systems [1]. The micro and meso-structural features of systems structured by a crystal network, as plastic fats and oleogels, define several macroscopic properties like the

of

system elasticity, yield stress, optical transparency, liquid binding capacity, and diffusivity [2, 3]. On the other hand, the molecular structure and thermodynamic

ro

properties of the solid phase allow predicting the gel stability, gelator solubility, and

-p

processing conditions [4]. Thus, the design of stable oleogels and plastic fats requires a full understanding of the interactions among the involved compounds at the molecular

re

packing level. These interactions can be determined from the phase behavior and

lP

crystal ripening process of pure neat gelators and their mixtures [5-8]. Edible oleogels are typically structured with low molecular weight gelators,

ur na

including steroids, plant sterols, tri-, di-, and monoglycerides, and natural waxes [9]. Among these gelators, waxes stand out because they can produce thermoreversible oleogels at very low concentrations (i.e., 0.5%) [3, 10]. Natural waxes are complex mixtures of wax esters, hydrocarbons, fatty acids, fatty alcohols, phenolic esters,

Jo

among other compounds. Thermodynamically stable oleogels with high oil binding capacity can be produced using  1% of candelilla wax (CW) [2, 8]; nevertheless, the rheological properties of these gels do not generally match the requirements for several edible matrices. They are too soft to produce stick margarine, spreads, and confectionery coatings [3, 8, 11, 12]. In this regard, increasing wax content can harden

4

the oleogels but this can also increase their melting temperature, causing a pasty mouthfeel that negatively affects their sensory perception [12, 13]. Chopin-Doroteo et al. [14] stated that these limitations can be overcome by adding 1% tripalmitin (PPP) to 3% CW vegetable oil solutions, producing composite oleogels with higher elasticity and yield stress than those of 3% CW without PPP. This increment of elasticity has been

of

attributed to the CW and PPP co-crystallization because the distinctive PPP spherulites were not spotted by polarized microscopy or detected by X-ray diffraction. Recently,

ro

Ramírez-Gómez et al. [15] produced composite oleogels with CW and saturated

-p

triacylglycerols (TAGs) from fully hydrogenated soybean oil (FH) and observed a synergistic effect of FH and CW on gel elasticity (i.e., elastic modulus) and a positive

re

effect of FH on the mechanical reversibility of CW gels. Nonetheless, the mechanisms

elucidated.

lP

involved in this positive interaction between CW and saturated TAGs are not fully

ur na

TAGs show a monotropic polymorphism, where all but one polymorph are unstable. The crystallization kinetics for these compounds might lead to the formation of unstable or metastable polymorphs (i.e., α or β') that spontaneously transform into the stable polymorph β [16, 17]. The unstable polymorph α has a hexagonal subcell

Jo

packing, where the hydrocarbon chains are perpendicular to the lamellar plane while the metastable polymorph β' has an orthorhombic subcell packing, where the hydrocarbon chains are inclined 108° with respect to the plane. On the other hand, the stable polymorph β has a triclinic subcell packing with a hydrocarbon chain inclined

5

128°. The phase behavior for much TAGs commonly found in vegetable oils has already been reported [18-23]. CW is mainly composed of C29-C33 n-alkanes with odd number of carbons (47.3%), with n-hentriacontane (C31, 35.5%) being the most abundant. Other CW components are triterpenic alcohols (24.8%), C26-C34 fatty acids with even number of

of

carbons (16.6%), C28-C34 fatty alcohols with even number of carbons (3.61%), and C39-

ro

C64 esters (5.83%) [24]. Despite the large number of components, the crystallization, gelation, and physical properties exhibited by CW in vegetable oils are highly

-p

determined by C31 [7, 24]. Pure n-alkanes phase behavior is determined by temperature, number of carbons, and parity of the hydrocarbon chain. The crystal

re

packing of solid phases of n-alkanes can be grouped in ordered and disordered

lP

polymorphs. The subcell structure of ordered polymorphs includes the triclinic, monoclinic, and orthorhombic, while the disordered polymorphs are triclinic disorder

ur na

phase, monoclinic disorder phase, orthorhombic rotator phase and the rhombohedral rotator phase [25]. Odd-numbered C9-C69 n-alkanes show an orthorhombic crystal structure at low temperature, C6-C26 n-alkanes of even-number show a triclinic structure and those of C28-C36 and C38-C44 show monoclinic and orthorhombic structures,

Jo

respectively. In every case, the increase of temperature below the melting point originates solid-solid polymorphic transitions of the type order-order and order-disorder [26]. The polymorphism of n-alkanes can be modified by small amounts (< 2%) of impurities introduced in their crystal lattices [5, 7].

6

Despite of the fact that the molecular self-assembly of neat gelators might differ from that of the three-dimensional crystal network of oleogels, the phase behavior of binary mixtures of long-chain n-alkanes and TAGs might provide valuable information to engineer composite oleogels with mixtures of CW and TAGs with specific melting properties and improved rheology. To the best of our knowledge, there is no information

of

regarding the phase behavior and microstructure of these mixtures. Thus, the objective of this work was to develop phase diagrams of neat binary mixtures of C31 and MMM,

ro

PPP or SSS by differential scanning calorimetry and X-ray diffraction, complementing

MATERIALS

lP

2.1.

re

2. MATERIALS AND METHODS

-p

the diagrams with microstructural features.

C31 (purity ≥ 95%) and the saturated monoacid TAGs (MMM, PPP, and SSS) (purity ≥

ur na

99%) were obtained from Santa Cruz Biotechnology and Sigma-Aldrich, respectively. The mixtures (C31-TAG) were prepared by weighing the corresponding amount of each compound in a sample container, heating them at 20 °C above the sample melting temperature for 10 min and stirring with a micro spatula to produce a homogeneous

Jo

melt. The mixture was cooled at room temperature and processed as indicated below. The mixtures were prepared in molar fraction ratio. 2.2.

METHODS

2.2.1. DIFFERENTIAL SCANNING CALORIMETRY

7

Crystallization and melting profiles of pure components and mixtures were determined using a differential scanning calorimeter (2920; TA Instruments New Castle, DE, USA) equipped with a refrigerated cooling unit. The DSC was calibrated with indium (melting temperature of 156.6 °C, melting enthalpy of 28.45 J/g), and an empty aluminum pan was used as a reference. The DSC cell was purged with chromatographic grade

of

nitrogen (purity  99.99%; 15 mL/min). Samples were weighed in aluminum pans and processed as described in Section 2.1, hermetically sealed, and placed in the

ro

measuring cell. Pure compounds and their mixtures were heated to 95 °C for 20 min,

-p

cooled (5 °C/min) to 0 °C and maintained isothermally for 2 min. After that, the sample was heated (5 °C/min) to 95 °C. Samples showing recrystallization during this heating

re

step were annealed to eliminate unstable polymorphs. The annealing process

lP

consisted of heating to 95 °C for 20 min, cooling (5° C/min) to 0 °C, maintaining the samples isothermally at 0 °C for 2 min, heating (5 °C/min) to the recrystallization temperature for 5 min, and cooling (5 °C/ min) to 0°C for 2 min. Then, the annealed

ur na

sample was heated (5 °C/min) to 95 °C to obtain the melting thermogram. To improve the resolution of the peaks, the final heating step of annealed samples was also performed a 1 °C/min for pure compounds and mixtures showing a single melting peak.

Jo

The onset (𝑇𝑜𝑖 ) and peak (𝑇𝑐 𝑖 ) temperatures for each compound in the first cooling thermogram, the melting enthalpy (𝐻𝑚𝑖 ) and peak temperature (𝑇𝑚𝑖 ) for each melting compound, and the temperature peak of the order-disorder transition of C31 (𝑇𝑜𝑑 ) obtained from the heating thermogram of annealed samples were determined by the

8

first derivate of heat flow using the software Universal Analysis 2000 Ver. 4.2E (TA Instruments-Waters LLC).

2.2.2. WIDE-ANGLE X-RAY DIFFRACTION

of

Wide-angle X-ray diffraction patterns (WAX) of pure components and their mixtures were analyzed using an X-ray diffractometer (X'Pert Pro; Panalytical, Almelo, The

ro

Netherlands) equipped with a CuKaR source (λ = 1.54 Å) exited at 35 kV and 30 mA, 0.5, 1.0, and 5.5 mm,

-p

The divergence, scatter and received slits were set at

respectively. Samples were placed in an aluminum sample holder (≈ 500 mg), molten,

re

and annealed as described above. The temperature program was controlled with the software Rheoplus/32 V3.61 (Anton-Paar, Germany), and applied with a Peltier plate

lP

(P-PTD200/62/TG, Anton-Paar, Germany) on the bottom of the sample holder and a Peltier hood (H-PTD-200, Anton-Paar, Germany) on the top. The annealed sample was

ur na

stored at 25 °C for 24 h. To avoid the preferential orientation of crystallites, the samples were finely grated at room temperature before analysis. The diffraction patterns were obtained by scanning the samples from 3° to 35° using a step of 0.02° and a scanning

Jo

speed of 0.04 °/min. Diffraction patterns were collected by the X'Pert Data Collector software version 2.0. The X-ray diffraction analysis was carried out for pure compounds and three mixtures of each binary system.

2.2.3. POLARIZED LIGHT MICROSCOPY

9

Microphotographs of pure compounds and binary mixtures were obtained using a polarized light microscope (ECLIPSE LV100N POL; Nikon, Japan) equipped with a digital color video camera (Micropublisher 3.3 RTV Q; QImaging, Canada) and a temperature-controlled stage (LTS420; Lynkam Scientific Instruments, UK) connected to a liquid nitrogen pumping system (LNP95; Lynkam Scientific Instruments, UK). The

of

temperature program was controlled with the software Linksys 32 ver. 2.3.0 (Lynkam Scientific Instruments LTD, UK). Samples (≈ 10 mg) were placed on a glass slide and

ro

heated to 95 °C and stirred with a micro-spatula. A coverslip was placed on the molten

-p

sample previous to be processed according to the annealing program described above. The microphotographs of the annealed samples were collected at 25 °C with the

STATISTICAL ANALYSIS

lP

2.3.

re

software NIS-Elements AR 4.51 (Nikon, Japan).

The plots were built using the software Prism 5.0 (GraphPad Software Inc., San Diego,

ur na

CA). The data were analyzed by a one-way ANOVA (P < 0.05) using the software STATISTICA V 7.1 (StatSoft, Inc. Tulsa, OK). 3.

RESULTS AND DISCUSSION

3.1 PURE COMPONENTS

Jo

The cooling/heating thermograms for pure TAGs (MMM, PPP, and SSS) and C31

are shown in Fig.1. During cooling, TAGs presented only one crystallization exotherm (Fig. 1A), while the corresponding heating thermograms (Fig. 1B) showed one recrystallization exotherm and two endotherms for each TAG. Based on previous reports, the lower endothermic peak was associated with the melting of the α-form

10

crystals, and the other to those in β-form [27]. The recrystallization corresponded to the α  β polymorphic transformation. After the annealing process, the heating thermograms of TAGs at either 5 °C/min or 1 °C/min showed a single melting endotherm (Fig. 2A, Fig. S1). The melting temperature and melting enthalpy for these TAGs agreed with those reported for the β-form (Table 1) [27, 28-31]. The β form in the

of

annealed TAGs was corroborated with the X-ray diffractograms obtained at 25 °C, where the diffraction pattern showed three peaks in the wide-angle region (2θ > 19°) at

ro

interplanar distances (d) of 4.6 Å, 3.9 Å, and 3.7 Å (Fig. 2B), which are distinctive of a

-p

triclinic subcell [32].

Given the enantiotropic polymorphism of C31, the heating thermogram of this

re

compound at either 5 °C/min or 1 °C/min showed the same endothermic transitions

lP

after crystallization, and after applying the annealing treatment used for SSS (Fig. 1B & Fig. 2A). The lower temperature endotherm of C31 was associated with the

ur na

orthorhombic  rotor phase transition (Tod,C31 = 61.55 ± 0.35 °C) and the higher temperature endotherm to the rotor phase  liquid transition (Tm,C31 = 67.82 ± 0.56 °C). The endotherm associated with the orthorhombic  monoclinic phase transition showed by C31  99% purity at  48 °C was not detected in our samples because of

Jo

the lower purity level in the C31 here studied [7, 25]. The enthalpy associated with the orthorhombic  rotor phase transition and melting (222.70 ± 4.53 J/g), and both temperature transitions were consistent with data from literature [7, 25]. The X-ray diffractogram for C31 showed two peaks of high intensity at d values of 4.2 and 3.8 Å

11

(Fig 2B), characteristic of orthorhombic packing subcell, which is the stable form at 25 °C. Micrographs of the annealed compounds are shown in Fig. 3. C31 showed large birefringent fibers, which were larger than those reported by Serrato-Palacios et al. [7]. This phenomenon was attributed to the recrystallization process that occurred during

of

the annealing process (Fig. 3A). On the other hand, TAGs formed spherulites with the

ro

typical Maltese cross shape. PPP and SSS produced the largest and smallest crystals of this type, respectively, while MMM produced crystals of medium size. A circle of

-p

granular microcrystals of very small size was observed in between some Maltese crossshaped crystals of MMM (Fig. 3B). This type of structure could be associated with

re

recrystallization during the annealing process [33].

lP

3.2. n-HENTRIACONTANE/TAGS MIXTURES

ur na

3.2.1 DYNAMIC CRYSTALLIZATION

The cooling thermograms of molten C31-TAG mixtures are shown in Fig. 4. Independently of the mixture composition, C31 crystallized at the highest temperature, showing a decline in crystallization onset (To,C31) proportional to XC31 (Fig. S2). Thus,

Jo

independently of the chain length, there was an evident dilution effect of these TAGs on C31. For the C31-MMM mixtures, the C31 crystallization was detected at the tested concentration range. The C31 exotherm of systems with XC31 = 0.60-1.0 showed two overlapping peaks associated with the crystallization from the isotropic liquid  rotor phase, and the rotor phase  orthorhombic transition. On the other hand, C31

12

exotherms for systems with XC31 = 0.05-0.5 only showed the peak corresponding to the transition from the liquid  orthorhombic phase [7]. The same trend was observed for C31-PPP and C31-SSS mixtures, but the C31 exotherms were detected only in systems with XC31 = 0.30-1.0 or XC31 = 0.20-1.0, respectively. TAG crystallization in mixtures was different for each mixture. In C31-MMM, C31 increased the To,MMM (P <

of

0.05) approximately 6 °C with respect to that of pure MMM (Fig. 4A). These results agreed with the seeding effect reported for sunflower wax on anhydrous milk fat (AMF),

ro

which reduced the induction time to nucleation, and increased the number of AMF

-p

crystals [34, 35]. Thus, we concluded that C31 crystals acted as hetero-nuclei for MMM. In contrast to C31-MMM mixtures, the C31 crystals did not change (P < 0.005) the

re

To,PPP, and had a slightly negative effect (P < 0.05) on To,SSS as compared to the

lP

corresponding pure component (Fig. 4B & 4C). This behavior is consistent with the study of Ramírez-Gómez et al. [15], where several mixtures of candelilla wax and fully hydrogenated soybean oil (FH) showed either no significant or slight changes in the

ur na

crystallization onset of FH as the CW content in the system increased. Considering that FH is composed of SSS and medium-chain trisaturated TAGs, we concluded that the C31 crystals did not act as hetero-nuclei under dynamic crystallization conditions for

Jo

PPP and SSS.

3.2.2 TAG MIXTURES ANNEALING The heating thermograms of C31-TAG mixtures crystallized under dynamic cooling conditions

showed

recrystallization

events

associated

with

polymorphic 13

transformations of unstable forms of TAGs (Fig. S3). As materials must be in equilibrium to construct phase diagrams, all mixtures were annealed to produce systems with TAGs crystals in the stable β-form, and C31 in the orthorhombic subcell packing. Regardless of the mixture composition, the XRD diffraction patterns from annealed C31-TAG mixtures showed the three distinctive peaks of the triclinic subcell

of

of TAGs in the β-form (Fig. 5) and the most intense peak of C31. The peak of C31 diffracting at 3.8 Å could not be detected as it is relatively less intense than the peaks

ro

at 3.7 and 3.9 Å produced from TAGs crystals in the β-form. Thus, we confirmed the

-p

annealed mixtures were thermodynamically stable, and that in all cases they were

re

constituted by a mixture of TAG and C31 crystals.

lP

3.2.3 SOLID-LIQUID PHASE DIAGRAMS

Melting endotherms of annealed C31-TAG mixtures (Fig. 6) were used to construct

ur na

solid-liquid equilibrium diagrams as a function of temperature and composition. These diagrams showed liquid (L), solid-liquid equilibrium (SLE), and solid (S) regions. Since the heating rate did not show a significant effect on peak temperature (Tm,i or Tod,C31), but it did on the onset and ending temperature of the endotherm; the liquidus line,

Jo

separating the L region from the SLE region, was constructed with the Tm,i of the compound in the mixture showing the highest melting temperature, while the solidus line corresponded to the Tm,i of the compound in the mixture showing the lowest melting temperature. The o-d solid transition line of C31 was traced with Tod,C31.

14

3.2.3.1 N-HENTRIACONTANE – TRIMIRISTIN The heating thermograms of annealed mixtures (Fig. 6) showed up to three endothermic peaks associated with TAG melting, the o-d transition and melting of C31. For the C31-MMM system, the endotherm corresponding to the melting of C31 was detected for systems with XC31 = 0.30-1.0 (Fig. 6A), in which Tm,C31 decreased with the

of

amount of C31. On the other hand, the endotherm of the C31 o-d transition was

ro

identified in systems with XC31 = 0.50-1.0 at 61.5 °C independently of the mixture composition. The melting endotherm showed at lower temperatures, corresponding to

-p

MMM crystals, was detected in all mixtures. In contrast to Tm,C31, Tm,MMM slightly changed with system composition. It is worth to note that the single endotherm detected

re

for C31-MMM mixtures with XC31  0.3 (Fig. 6A, Fig. S4A) was attributed to the

lP

concomitant melting of β-MMM and C31 crystals, as both crystals species were detected in the X-ray diffractograms (Fig. 5A). Phase change temperatures (i.e.,

ur na

Tm,i,Tod,i) used to draw the delimiting lines in the C31-MMM phase diagram are in Table 2. From these data, the liquidus line (Fig. 7), corresponding to Tm,i of the compound in the mixture showing the highest melting temperature (Tm,C31 or Tm,MMM), exhibited a eutectic at XC31 = 0.2 with Tm = 56.30 °C. Below the solidus line, corresponding to the

Jo

temperature of the melting peak of MMM (Tm,MMM), the systems were a microheterogeneous dispersion of β-MMM and C31 crystals (Fig. 7). Two SLE regions were found in mixtures with XC31  0.30 above the solidus line and below the o-d transition temperature. In these regions, the liquid phase was a mixture of MMM and C31 crystals.

15

C31 was in the orthorhombic polymorphic form (C31-O) or in the rotor phase (C31-R) if they were below or above the o-d transition temperature, respectively. The crystal habit of the studied TAGs was greatly modified in the mixtures, while C31 crystals maintained their characteristic fiber shape (Figs. 3, 7-9). C31 modified the shape of MMM crystals, changing from the spherulitic to a grainy habit. Their size was

of

also reduced by the increment of C31 in the mixture (Fig. 7). These phenomena were

-p

detected from the cooling thermograms (Fig. 4A).

ro

attributed to the heterogeneous nucleation of MMM on the C31 crystals´ surface, as

re

3.2.3.2 N-HENTRIACONTANE – TRIPALMITIN

lP

In the heating thermograms of C31-PPP with XC31 > 0.50, the highest melting component was C31 (Fig. 6B). In this system, the melting temperature of C31 decreased with the amount of PPP, while the o-d transition appeared around 61.5 °C,

ur na

and the melting peak of PPP was detected at 60 °C. Mixtures with XC31 = 0.3-0.5 showed one endotherm with a single peak (Fig. 6B), which was attributed to the simultaneous melting of the two crystal species detected by XRD (C31 and β-PPP)

Jo

(Fig. 5B). On the other hand, mixtures with XC31 = 0.2-0.1 showed one melting endotherm with two overlapping peaks; the peak at higher temperatures corresponded to the melting of β-PPP, the other to melting of C31. The phase change temperatures used to draw the delimiting lines in the C31-PPP phase diagram are shown in Table 3. From these data, the liquidus line (Fig. 8) revealed a eutectic at XC31 = 0.4 with Tm =

16

60.32 °C. The X-ray diffraction analysis confirmed that below the solidus line these components were structured as a micro-heterogeneous dispersion of β-PPP and C31 crystals (Fig. 8). This mixture showed three SLE regions, one for mixtures with XC31 = 0.2-0.4, and others for those with XC31 = 0.5-0.95. In the former case, the β-PPP crystals and a liquid phase were in equilibrium, while in the later a liquid phase and C31-O were

of

found above the solidus line and below the Tod,C31 as well as a liquid phase and C31-R above the Tod,C31 and below the liquidus line. The microstructure of PPP crystals in

ro

mixture with C31 showed the most significant changes as compared to pure PPP.

-p

The grainy crystals in C31-PPP mixtures were smaller than those of pure PPP and C31-MMM and C31-SSS mixtures (Figs. 3, 8). Nonetheless, thermograms of C31-PPP

re

mixtures did not show evidence of a secondary nucleation process (Secc. 3.2.1).

lP

Therefore, the size reduction of PPP crystals could result from the inhibition of the crystal growth by adsorption of C31 on the surface of growing crystals. This effect is produced when impurities in the crystallizing media are adsorbed on crystal surface,

ur na

hindering the incorporation of the crystallizing species, thus, these impurities (C31 for PPP in our study) must be displaced to allow the incorporation of PPP in the crystal

Jo

lattice [36].

3.2.3.3 N-HENTRIACONTANE – TRISTEARIN For the C31-SSS system, the highest melting endotherm corresponded to SSS in mixtures with XC31 < 0.6, and to C31 in those with XC31 > 0.8 (Fig. 6C). On the other hand, mixtures with XC31 = 0.6-0.8 showed only one endothermic peak, which was

17

attributed to the concomitant melting of C31 and β-SSS crystals as detected by XRD measurements (Fig. 5C). From data shown in Table 4, the liquidus line revealed a eutectic region at XC31 = 0.7-0.8 with Tm = 64 °C (Fig. 9). As in the other systems, the melting temperature of C31 decreased with the amount of SSS while the o-d transition appeared around 61.5 °C, and the melting peak for SSS decreased to a minimum of

of

63.87 °C. Below the solidus line and above Tod,C31, the mixtures with XC31  0.4 coexisted as a mixture of β-SSS and C31-R, but below Tod,C31 these systems consisted

ro

of a physical mixture of β-SSS and C31-O crystals (Fig. 9). Two SLE regions were

-p

identified for this system, one for mixtures with XC31 < 0.6, where β-SSS and a liquid

equilibrium with the liquid phase.

re

phase coexisted, and other for mixtures with XC31 > 0.8, where the C31-R was in

lP

The microstructure of SSS crystals changed as a function of the C31 content in the mixture. At X31 = 0.1, the spherulitic shape and crystal size were similar to those of pure

ur na

SSS (Figs. 3, 9). On the other hand, mixtures with X31 ≥ 0.4 developed grainy crystals, which showed only slight changes of size with the mixture composition. As in C31-PPP mixtures, these morphological changes could not be associated with a secondary nucleation process. They seemed to be related to the inhibition of crystal growth by

Jo

adsorption of C31 on the surface of growing crystals. It is worth to mention that the size reduction of TAGs crystals induced by C31 is desirable for some lipid-based matrices for food, cosmetic and drug delivery products because a smaller crystal size produces a smoother texture, and increases the product elasticity [34].

18

4. CONCLUSIONS C31-TAG mixtures here reported showed a eutectic phase behavior. The eutectic composition was richer in the TAG as the melting temperature of the TAG increased; that is XC31 = 0.2, XC31 = 0.4, and XC31 = 0.7-0.8 for MMM, PPP, and SSS, respectively. These systems below the solidus line were a solid dispersion of crystals rich in C31

of

and crystals rich in the corresponding TAG. This segregation in the solid phase was attributed to the difference in subcell-packing of the stable polymorphic form of pure

ro

components; orthorhombic for C31 and monoclinic for TAGs in the -polymorphic form.

-p

The heterogeneous nucleation of MMM on the surface of C31 crystals generated significant changes in the crystal habit of their crystals. On the other hand, crystal

re

growth inhibition of TAGs in C31-PPP and C31-SSS mixtures were caused by the

Declaration of interests

lP

transient adsorption of C31 molecules at the growing crystal surface.

ur na

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Jo

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

19

ACKNOWLEDGMENTS This research was funded by the National Council for Science and Technology of Mexico (CONACYT) through the Fondo Sectorial de Investigación para la Educación

Jo

ur na

lP

re

-p

ro

of

(Investigación Básica SEP-CONACYT; grant number: 256100).

20

REFERENCES [1] A.G. Marangoni, N. Acevedo, F. Maleky, E. Co, F. Peyronel, G. Mazzanti, B. Quinn, D. Pink, Structure and functionality of edible fats, Soft Matter, 8 (2012) 1275-1300. https://doi.org/10.1039/c1sm06234d

of

[2] M. Sánchez-Becerril, A.G. Marangoni, M.J. Perea-Flores, N. Cayetano-Castro, H. Martínez-Gutiérrez, J.A. Andraca-Adame, J.D. Pérez-Martínez, Characterization of the micro and nanostructure of the candelilla wax organogels crystal networks, Food Structure, 16 (2018) 1-7. 1. https://doi.org/10.1016/j.foostr.2018.02.001

ro

[3] A.I. Blake, E.D. Co, A.G. Marangoni, Structure and Physical Properties of Plant Wax Crystal Networks and Their Relationship to Oil Binding Capacity, Journal of the American Oil Chemists' Society, 91 (2014) 885-903. https://doi.org/10.1007/s11746-014-2435-0

re

-p

[4] J.F. Toro-Vazquez, J.D. Pérez-Martínez, Thermodynamic Aspects of Molecular Gels, in: Molecular Gels: Structure and Dynamics, The Royal Society of Chemistry, 2018, pp. 57-87. https://dx.doi.org/10.1039/9781788013147-00057

lP

[5] D.J. Abdallah, R.G. Weiss, Organogels and Low Molecular Mass Organic Gelators, Advanced Materials, 12 (2000) 1237-1247. https://doi.org/10.1002/15214095(200009)12:17<1237::AID-ADMA1237>3.0.CO;2-B

ur na

[6] E. Ostuni, P. Kamaras, R.G. Weiss, Novel X-ray Method for In Situ Determination of Gelator Strand Structure: Polymorphism of Cholesteryl Anthraquinone-2carboxylate, Angewandte Chemie (International Edition in English), 35 (1996) 1324-1326. https://doi.org/10.1002/anie.199613241

Jo

[7] L.L. Serrato-Palacios, J.F. Toro-Vazquez, E. Dibildox-Alvarado, A. Aragón-Piña, M.d.R. Morales-Armenta, V. Ibarra-Junquera, J.D. Pérez-Martínez, Phase Behavior and Structure of Systems Based on Mixtures of n-Hentriacontane and Melissic Acid, Journal of the American Oil Chemists' Society, 92 (2015) 533-540. https://doi.org/10.1007/s11746-015-2623-6 [8] J.F. Toro-Vazquez, J.A. Morales-Rueda, E. Dibildox-Alvarado, M. Charó-Alonso, M. Alonzo-Macias, M.M. González-Chávez, Thermal and Textural Properties of Organogels Developed by Candelilla Wax in Safflower Oil, Journal of the American Oil Chemists' Society, 84 (2007) 989-1000. https://doi.org/10.1007/s11746-0071139-0 [9] S. Yang, G. Li, A.S.M. Saleh, H. Yang, N. Wang, P. Wang, X. Yue, Z. Xiao, Functional Characteristics of Oleogel Prepared from Sunflower Oil with β-

21

Sitosterol and Stearic Acid, Journal of the American Oil Chemists' Society, 94 (2017) 1153-1164. https://doi.org/10.1007/s11746-017-3026-7 [10] A.R. Patel, Natural waxes as oil structurants, in: Alternative Routes to Oil Structuring, Springer, 2015, pp. 15-27. https://doi.org/10.1007/978-3-319-191386_2 [11] L.S.K. Dassanayake, D.R. Kodali, S. Ueno, K. Sato, Physical Properties of Rice Bran Wax in Bulk and Organogels, Journal of the American Oil Chemists' Society, 86 (2009) 1163. https://doi.org/10.1007/s11746-009-1464-6

ro

of

[12] H.-S. Hwang, M. Singh, E.L. Bakota, J.K. Winkler-Moser, S. Kim, S.X. Liu, Margarine from Organogels of Plant Wax and Soybean Oil, Journal of the American Oil Chemists' Society, 90 (2013) 1705-1712. https://doi.org/10.1007/s11746-013-2315-z

-p

[13] A.K. Zetzl, A.G. Marangoni, Structured emulsions and edible oleogels as solutions to trans fat, in: Trans Fats replacement solutions, Elsevier, 2014, pp. 215-243. https://doi.org/10.1108/eb059455

lP

re

[14] M. Chopin-Doroteo, J.A. Morales-Rueda, E. Dibildox-Alvarado, M.A. CharóAlonso, A. de la Peña-Gil, J.F. Toro-Vazquez, The effect of shearing in the thermo-mechanical properties of candelilla wax and candelilla wax–tripalmitin organogels, Food Biophysics, 6 (2011) 359-376. https://doi.org/10.1007/s11483011-9212-5

ur na

[15] N.O. Ramírez-Gómez, N.C. Acevedo, J.F. Toro-Vázquez, J.J. Ornelas-Paz, E. Dibildox-Alvarado, J.D. Pérez-Martínez, Phase behavior, structure and rheology of candelilla wax/fully hydrogenated soybean oil mixtures with and without vegetable oil, Food Research International, 89 (2016) 828-837. https://doi.org/10.1016/j.foodres.2016.10.025 [16] D.M. Small, The physical chemistry of lipids from alkanes to phospholipids, In: Handbook of Lipid Research, 4 (1986) 367-372.

Jo

[17] P. Walstra, Physical chemistry of foods, CRC Press, 2002. [18] R.C. Da Silva, F.A.S.D. Martini Soares, J.M. Maruyama, N.R. Dagostinho, Y.A. Silva, J.N.R. Ract, L.A. Gioielli, Microscopic approach of the crystallization of tripalmitin and tristearin by microscopy, Chemistry and Physics of Lipids, 198 (2016) 1-9. https://doi.org/10.1016/j.chemphyslip.2016.04.004 [19] N. Haghshenas, P. Smith, B. Bergenståhl, The exchange rate between dissolved tripalmitin and tripalmitin crystals, Colloids and Surfaces B: Biointerfaces, 21 (2001) 239-243. https://doi.org/10.1016/S0927-7765(01)00176-X 22

[20] B. Hong, D. Li, Q. Lei, L. Xu, W. Fang, H. Mayama, Kinetics on formation of super water repellent surfaces from phase transformation in binary mixtures of trimyristin and tripalmitin, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 396 (2012) 130-136. https://doi.org/10.1016/j.colsurfa.2011.12.056 [21] M. Kellens, W. Meeussen, H. Reynaers, Study of the polymorphism and the crystallization kinetics of tripalmitin: A microscopic approach, Journal of the American Oil Chemists' Society, 69 (1992) 906-911. https://doi.org/10.1007/BF02636342

ro

of

[22] K. Sato, Crystallization behaviour of fats and lipids — a review, Chemical Engineering Science, 56 (2001) 2255-2265. https://doi.org/10.1016/S00092509(00)00458-9

-p

[23] R.E. Timms, Phase behaviour of fats and their mixtures, Progress in Lipid Research, 23 (1984) 1-38. https://doi.org/10.1016/0163-7827(84)90004-3

lP

re

[24] J.F. Toro-Vazquez, R. Mauricio-Pérez, M.M. González-Chávez, M. SánchezBecerril, J. de J. Ornelas-Paz, J.D. Pérez-Martínez, Physical properties of organogels and water in oil emulsions structured by mixtures of candelilla wax and monoglycerides, Food Research International, 54 (2013) 1360-1368. https://doi.org/10.1016/j.foodres.2013.09.046

ur na

[25] A.-J. Briard, M. Bouroukba, D. Petitjean, N. Hubert, M. Dirand, Experimental Enthalpy Increments from the Solid Phases to the Liquid Phase of Homologous n-Alkane Series (C18 to C38 and C41, C44, C46, C50, C54, and C60), Journal of Chemical & Engineering Data, 48 (2003) 497-513. https://doi.org/10.1021/je0201368 [26] M. Dirand, M. Bouroukba, A.-J. Briard, V. Chevallier, D. Petitjean, J.-P. Corriou, Temperatures and enthalpies of (solid + solid) and (solid + liquid) transitions of nalkanes, The Journal of Chemical Thermodynamics, 34 (2002) 1255-1277. https://doi.org/10.1006/jcht.2002.0978

Jo

[27] R.W. Hartel, J. H. von Elbe, R. Hofberger, Fats, Oils and Emulsifiers, in: Confectionery Science and Technology, Springer, 2018, pp. 85-124. https://doi.org/10.1007/978-3-319-61742-8 [28] E. Da Silva, S. Bresson, D. Rousseau, Characterization of the three major polymorphic forms and liquid state of tristearin by Raman spectroscopy, Chemistry and Physics of Lipids, 157 (2009) 113-119. https://doi.org/10.1016/j.chemphyslip.2008.11.002 [29] M. Kellens, H. Reynaers, Study of the Polymorphism of Saturated Monoacid 23

Triglycerides I: Melting and Crystallization Behaviour of Tristearin, Lipid / Fett, 94 (1992) 94-100. https://doi.org/10.1002/lipi.19920940304 [30] E. Kolbe, L.A. Wilson, R. Hartel, A round robin evaluation of differential scanning calorimetry to measure transition enthalpy and temperatures, Journal of Food Engineering, 40 (1999) 95-99. https://doi.org/10.1016/S0260-8774(99)00044-8

of

[31] M. Matovic, J.C. van Miltenburg, J. Los, F.G. Gandolfo, E. Flöter, Thermal Properties of Tristearin by Adiabatic and Differential Scanning Calorimetry, Journal of Chemical & Engineering Data, 50 (2005) 1624-1630. https://doi.org/10.1021/je050092d

ro

[32] R.C. D. Silva, F.A.S.D. Martini Soares, J.M. Maruyama, N.R. Dagostinho, Y.A. Silva, J.N.R. Ract, L.A. Gioielli, Microscopic approach of the crystallization of tripalmitin and tristearin by microscopy, Chemistry and Physics of Lipids, 198 (2016) 1-9. https://doi.org/10.1016/j.chemphyslip.2016.04.004

re

-p

[33] J.H. Magill, Review spherulites: A personal perspective, Journal of Materials Science, 36 (2001) 3143-3164. https://doi.org/10.1023/A:1017974016928

lP

[34] R.M. Kerr, X. Tombokan, S. Ghosh, S. Martini, Crystallization Behavior of Anhydrous Milk Fat−Sunflower Oil Wax Blends, Journal of Agricultural and Food Chemistry, 59 (2011) 2689-2695. https://doi.org/10.1021/jf1046046

ur na

[35] S. Martini, A.A. Carelli, J. Lee, Effect of the Addition of Waxes on the Crystallization Behavior of Anhydrous Milk Fat, Journal of the American Oil Chemists' Society, 85 (2008) 1097-1104. https://doi.org/10.1007/s11746-008-1310-2

Jo

[36] R.W. Hartel, Crystallization in Foods, Aspen Publishers Inc., 2001.

24

CAPTIONS Figure 1. Cooling (A) and heating (B) thermograms obtained at 5 °C/min for neat nhentriacontane (C31), trimyristin (MMM), tripalmitin (PPP), and tristearin

ur na

lP

re

-p

ro

of

(SSS).

Figure 2. Heating thermograms obtained at 5 °C/min (A) and X-ray diffractograms (B) for annealed n-hentriacontane (C31), trimyristin (MMM), tripalmitin (PPP), and

Jo

tristearin (SSS).

25

of

Figure 3. Polarized light microscopy images for annealed n-hentriacontane (A),

lP

re

-p

ro

trimyristin (B), tripalmitin (C), and tristearin (D).

ur na

Figure 4. Cooling thermograms obtained at 5 °C/min for mixtures of n-hentriacontane (C31) and trimyristin (A), tripalmitin (B), or tristearin (C). Mixture composition

Jo

is indicated by the n-hentriacontane mole fraction (XC31).

26

of ro -p re

lP

Figure 5. X-ray diffractograms for annealed for mixtures of n-hentriacontane (C31) and trimyristin (A), tripalmitin (B), or tristearin (C). Mixture composition is indicated

Jo

ur na

by the n-hentriacontane mole fraction (XC31).

27

of ro -p re lP

ur na

Figure 6. Heating thermograms obtained at 5 °C/min for annealed mixtures of nhentriacontane (C31) and trimyristin (A), tripalmitin (B), or tristearin (C).

Jo

Mixture composition is indicated by the n-hentriacontane mole fraction (XC31).

28

of ro -p re

Figure 7. Phase diagram for n-hentriacontane (C31) and trimyristin (MMM) system and

lP

polarized light microphotographs of selected mixtures. Mixture composition is indicated by the n-hentriacontane mole fraction (XC31). Melting peak

ur na

temperatures for C31 and -MMM, and the order-disorder transition for C31 are indicated in the graph. Liquid (L), crystal rich in trimyristin (-MMM), crystal rich in C31 in rotator phase (C31-R), crystal rich in C31 in orthorhombic

Jo

phase (C31-O).

29

of

ro

Figure 8. Phase diagram for n-hentriacontane (C31) and tripalmitin (PPP) system and polarized light microphotographs of selected mixtures. Mixture composition is

-p

indicated by the n-hentriacontane mole fraction (XC31). Melting peak temperatures for C31 and -PPP, and the order-disorder transition for C31

re

are indicated in the graph. Liquid (L), crystal rich in trimyristin (-MMM),

Jo

ur na

phase (C31-O).

lP

crystal rich in C31 in rotator phase (C31-R), crystal rich in C31 in orthorhombic

Figure 9. Phase diagram for n-hentriacontane (C31) and tristearin (SSS) system and polarized light microphotographs of selected mixtures. Mixture composition is

30

indicated by the n-hentriacontane mole fraction (XC31). Melting peak temperatures for C31 and -SSS, and the order-disorder transition for C31 are indicated in the graph. Liquid (L), crystal rich in trimyristin (-MMM), crystal rich in C31 in rotator phase (C31-R), crystal rich in C31 in orthorhombic

Jo

ur na

lP

re

-p

ro

of

phase (C31-O).

31

Table 1. Melting temperatures and enthalpies at 5 °C/min for pure triacylglycerides after the annealing process for the polymorphic β form.

∆𝑯𝒎 a

Tmb

∆𝑯𝒎 b

Tm

∆𝑯𝒎

(°C)

(J/g)

(°C)

(J/g)

(°C)

(J/g)

MMM

56.99 ± 0.02

182.45 ± 10.11

56.00 – 58.50

188.13

65.8f

190.69f

PPP

64.33 ± 0.08

188.65 ± 1.63

66.00 – 66.4

203.82

72.4 ± 0.05c,

217±1.6c, 219.6d,

SSS

71.97 ± 0.37

190.95 ± 3.46

73.00 – 73.50

213.24

71.2f

TAGs

of

Tma

211.1e, 213.13f

ro

(a) This work; mean and standard deviation of two individual measurements (b) Hartel et al., 2018; (c) Da Silva et al., 2009; (d) Matovic., 2005; (e) Kolbe et al., 1999; (f) Kellens

-p

and Reynaers., 1992.

re

Table 2. Transition temperatures for the phase diagram of n-hentriacontane (C31) and trimyristin (MMM) system. Mixture composition is indicated by the n-hentriacontane

ur na

lP

mole fraction (XC31)

XC31

The o-d solid

Liquidus line

Solidus line

transition line

𝑇𝑚, 𝐶31

𝑇𝑚, 𝑀𝑀𝑀

𝑇𝑜𝑑, 𝐶31

56.99 ± 0.02

0.05

57.18 ± 0.33

0.10

56.86 ± 0.46

Jo

0

0.20

56.30 ± 0.08

0.30

58.22 ± 0.05

56.69 ± 0.08

0.40

58.23 ± 0.18

55.94 ± 0.57

0.50

61.53 ± 0.16

55.71 ± 0.01

0.60

63.15 ± 0.13

55.57 ± 0.16

61.55 ± 0.26

32

0.70

63.50 ±0.07

55.44 ± 0.13

61.53 ± 0.05

0.80

64.75 ± 0.02

55.04 ± 0.05

61.38 ± 0.08

0.90

66.55 ± 0.02

54.94 ± 0.05

61.70 ± 0.00

0.95

67.65 ± 0.09

54.45 ± 0.18

62.01 ± 0.08

1

67.82 ± 0.56

61.55 ± 0.35

ro

of

*Values represent the mean and standard deviation of two replicates.

Table 3. Transition temperatures for the phase diagram of n-hentriacontane (C31) and

-p

tripalmitin (PPP) system. Mixture composition is indicated by the n-hentriacontane mole

Liquidus line

ur na

XC31

lP

re

fraction (XC31)

i

0

64.33 ± 0.08

PPP

0.05

64.19 ± 0.11

PPP

0.10

64.30 ± 0.19

PPP

0.20

62.84 ± 0.21

0.30

Solidus line

solid transition line

𝑇𝑚𝑖

i

PPP

59.80 ± 0.09

C31

61.96 ± 0.00

PPP

60.32 ± 0.08

C31

0.40

60.69 ± 0.01

C31 & PPP

60.69 ± 0.01

C31 & PPP

0.45

61.19 ± 0.33

C31 & PPP

61.19 ± 0.33

C31 & PPP

0.50

61.20 ± 0.06

C31 & PPP

61.20 ± 0.06

C31 & PPP

0.60

62.56 ± 0.94

C31

60.42 ± 0.62

PPP

0.70

63.70 ± 0.02

C31

Jo

𝑇𝑚𝑖

The o-d

𝑇𝑜𝑑, 𝐶31

61.63 ± 0.07 61.08 ± 0.91

33

0.80

65.29 ± 0.03

C31

56.49 ± 0.03

PPP

61.45 ± 0.04

0.90

66.08 ± 0.29

C31

59.46 ± 0.06

PPP

61.57 ± 0.21

0.95

67.61 ± 0.10

C31

59.61 ± 0.06

PPP

62.24 ± 0.13

1

67.82 ± 0.56

C31

61.55 ± 0.35

of

*Values represent the mean and standard deviation of two replicates.

ro

Table 4. Transition temperatures for the phase diagram of n-hentriacontane (C31) and tristearin (SSS) system. Mixture composition is indicated by the n-hentriacontane mole

re

Liquidus line

i

ur na

𝑇𝑚𝑖

The o-d

Solidus line

lP

XC31

-p

fraction (XC31)

solid transition line

𝑇𝑚𝑖

i

𝑇𝑜𝑑, 𝐶31

71.97 ± 0.37

SSS

0.05

72.27 ± 0.25

SSS

0.10

71.56 ± 0.02

SSS

63.56 ± 0.51

C31

0.20

71.29 ± 0.21

SSS

62.92 ± 0.34

C31

0.30

70.18 ± 0.37

SSS

62.79 ± 0.29

C31

0.40

69.04 ± 0.16

SSS

63.29 ± 0.08

C31

61.73 ± 0.26

0.50

67.88 ± 0.08

SSS

63.56 ± 0.08

C31

61.11 ± 0.06

0.60

66.43 ± 0.08

SSS

64.14 ± 0.01

C31 & SSS

61.42 ± 0.02

0.70

64.04 ± 0.46

C31 & SSS

64.04 ± 0.46

C31 & SSS

61.55 ± 0.64

0.75

64.94 ± 0.18

C31 & SSS

64.94 ± 0.18

C31 & SSS

61.80 ± 0.11

0.80

64.35 ± 0.10

C31 & SSS

64.35 ± 0.10

C31 & SSS

61.28 ± 0.01

Jo

0

34

0.90

66.24 ± 0.04

C31

64.03 ± 0.19

SSS

61.67 ± 0.22

0.95

67.79 ± 0.00

C31

63.87 ± 0.48

SSS

62.19 ± 0.03

1

67.82 ± 0.56

C31

61.55 ± 0.35

Jo

ur na

lP

re

-p

ro

of

*Values represent the mean and standard deviation of two replicates.

35