Tristearin Binary System

Molecular Insights into the Eutectic Tripalmitin/Tristearin Binary System ... Publication Date (Web): September 4, 2018. Copyright © 2018 American Ch...
0 downloads 0 Views 2MB Size
Subscriber access provided by University of South Dakota

Article

Molecular Insights into the Eutectic Tripalmitin/Tristearin Binary System Antonio Pizzirusso, Fernanda Peyronel, Edmund Daniel Co, Alejandro G. Marangoni, and Giuseppe Milano J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b04729 • Publication Date (Web): 04 Sep 2018 Downloaded from http://pubs.acs.org on September 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Molecular Insights into the Eutectic Tripalmitin/Tristearin Binary System Antonio Pizzirussoa, Fernanda Peyronelb, Edmund D. Cob, Alejandro G. Marangonib* and Giuseppe Milanoc*

a

Dipartimento di Chimica e Biologia, Università di Salerno, Via Giovanni Paolo II, 132, I-84084, Fisciano, Italy. Department of Food Science, University of Guelph, Guelph, Ontario, Canada N1G 2W1. c Department of Organic Materials Science, Yamagata University, 4-3-16 Jonan Yonezawa, Yamagata-ken 992-8510, Japan b

KEYWORDS Eutectic Composition, Triglycerides, Molecular Dynamics, Tripalmitin, Tristearin, β Crystal Packing, Melting point.

ABSTRACT: A molecular interpretation of the eutectic behavior of a binary mixture of tristearin (SSS) and tripalmitin (PPP) triglycerides was formulated using computer simulations and experimental techniques (calorimetry and X-ray scattering). A eutectic composition was identified using both experimental and computer simulation techniques at a composition of 70 % PPP and 30 % SSS, in agreement with previous findings in the literature. The decrease in the melting temperature at the eutectic composition can be ascribed to an interplay between enthalpic and entropic effects. In particular, a lower global melting enthalpy at the eutectic composition was detected here, caused by a less efficient packing of the triglycerides in the crystal. On the other hand, a higher crystalline disorder is reflected in a lower change in the entropy of melting. The simultaneous decreases in global enthalpy and entropy has a contrasting effect on the melting temperature, with a slight melting point depression found both in experiment and simulations, resulting from a combination of enthalpic and entropic factors. Computer simulations showed in fact that the eutectic effect can be ascribed to the reduction of crystalline order when SSS molecules are incorporated into the PPP crystal structure. This decrease of the crystalline order is due to the protrusion of SSS end-chains (last three carbons of each alkyl chain) into the inter-lamellar space between adjacent lamella. These endchains disturb the orderly stacking of the lamella, as evidenced by low-density regions in the interlamellar space. Thus, the greater disorder of the last atoms of the SSS alkyl chains is consequently due to the greater conformational freedom. At molecular level, in fact, the conformational freedom of terminal atoms of SSS surrounded by shorter PPP molecules is larger than the conformational freedom of longer SSS in the neighborhood of short PPP.

Introduction

composition and the temperature of the system, among others. The phase behavior includes the interactions that govern the liquid-solid equilibrium and solid-solid equilibrium. The phases formed are usually distinguished by the chemical composition as well as the polymorphic form of the solid phases. TAGs typically crystallize into one of three different polymorphic forms, where the most stable thermodynamic form is the β polymorph, as compared with the α or β’ polymorphs. The lateral packing of the hydrocarbon chain is what defined the polymorph. The term subcell is used to identify those methylene groups that defined the geometry of the particular polymorph, which in the case of β is triclinic9. In TAGs, the phase behavior is crucial to determine the melting point of the material, and by extension, the solid fat content and functionality of the fat material10. In addition, knowledge of the phase behavior

Triacylglycerols (TAGs) are triesters of glycerol with fatty acids. TAGs form the bulk of the chemical constituents of edible fats and oils and are therefore relevant in edible applications. TAGs are also prominent in pharmaceutical applications. The physical and chemical properties of TAGs are largely determined by the nature of the fatty acids esterified onto the fatty acids1-4. TAGs can crystallize into nanometer-sized crystals5, which then aggregate to form larger structures6-7 that eventually form a colloidal solid fat network that structure edible oils8.The physical behavior of materials consisting of mixtures of 2 or more chemicallydistinct TAGs (such as edible fats and oils) is largely dictated by the phase behavior of the mixture. The phase behavior refers to the tendency of the mixture to assume a stable equilibrium state under given conditions of mixture ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

can allow for the design of effective tion/fractionation processes for fat materials11-12.

separa-

One of the most intriguing phase behaviors is the eutectic phase behavior of a mixture, for example, of the TAGs tristearin (SSS) and tripalmitin (PPP)13. The eutectic behavior of a solid binary mixture refers to the formation of two coexisting and distinct solid phases enriched in either component. The eutectic (from the Greek ‘easily melted’) mixture exhibits a depressed melting point relative to either pure components14. The eutectic point is the composition in the temperature vs composition binary phase diagram where the temperature is the lowest (the eutectic composition). Two triglycerides that form a eutectic typically exhibit similar melting points but very different molecular sizes/structures (as given by the molar volume)15. The disparity in the molecular structures affect the ability of these two TAGs to crystallize into a single, well-ordered crystalline solid phase. The eutectic behavior of SSS/PPP is especially noteworthy since SSS and PPP have very different melting points but very similar molecular structures (the fatty acid chains on SSS differ from the fatty acid chains on PPP by only two carbons). Without any experimental data, the natural assumption when examining the molecular structures of SSS and PPP is that a mixture of the two will form a monotectic mixture, not a eutectic mixture. Kerridge16 reported an SSS/PPP eutectic point at 63.5 °C and 25.5 % SSS + 74.5 % PPP. With additional information from X-ray studies, Lutton17 reported an SSS/PPP eutectic point at 63.5 °C and 16 % SSS + 84 % PPP for the β polymorph. Eutectic behavior was not reported for the meta-stable α or β’ polymorphs. Knoester18 reported the phase behavior of a series of TAGs containing stearic and palmitic acids and identified a eutectic for SSS/PPP. Knoester also reported that the presence of an asymmetric triglyceride in the binary mixture strongly aided in solidstate miscibility. Using micro-Differential Scanning Calorimetry (micro-DSC), Ringuette19 reported a eutectic point of 64.4 °C and 19 % SSS + 81 % PPP. MacNaugthan and others20 reported a eutectic point for the β polymorph of the PPP/SSS mixture at a PPP molar fraction XPPP = 0.7. Costa and others21 confirmed the eutectic composition at XPPP = 0.7 using phase diagrams obtained through DSC. The molecular mechanism by which SSS and PPP form a eutectic is as of yet, unknown. Due to the inaccessibility of the phenomenon to experimentation20,22, computer simulations are an ideal tool for studying the molecular basis of eutectic phase behavior. Crystal phases of TAGs have been intensively studied using computer simulation23-31. This technique provides information on both the structure of the crystal packing at the atomic level, as well as providing a link between the behavior of the atomic-level constituents (the microstate) and the macroscopic thermodynamic properties via statistical physics28-29. In the work of Hsu et al.31 for instance, a united atom Molecular Dynamics (MD) simulation showed the effect that

Page 2 of 14

the hydrocarbon chain structure can have on density, viscosity and thermal stability. MacDougall et al. 25 showed that at the nanoscale, liquid oil will not remain between two solid nanoparticles. Moreover, a coarse-grained model developed by Brasiello and co-workers27 has been successfully employed to describe the crystallization of different saturated TAGs into the α phase28-29. It is also possible to obtain some information regarding the nature of the eutectic phenomenon by examining other research areas. The eutectic phenomenon has been studied in alloys32-35, glasses36, biological compounds37, and deep eutectic systems38-39. A common observation in these studies is that the lower melting point is indicative of greater molecular mobility37. As well, the eutectic behavior of these systems has been attributed to differences in intermolecular interactions, particularly hydrogen bonds38. In this work, a thermodynamic mechanism of the eutectic behavior of PPP/SSS mixtures in the β phase is proposed using experimental results (melting points and X-ray scattering intensities) and computer simulations (molecular dynamics). The combination of experiments and simulations allowed us to propose a molecular-level view of the eutectic behavior of PPP/SSS mixtures in the β polymorph.

Materials and Methods Materials Tristearin (SSS) and tripalmitin (PPP) were purchased from Sigma-Aldrich. Binary mixtures of PPP/SSS were prepared by mixing the appropriate amount of the TAG and then heating the mixture to 353 K for 20 minutes. The melt was crystallized into the β polymorphic form by storing the melt in an oven set to 323 K for 48 hours, prior to transfer to ambient temperatures. An aliquot of the material was directly deposited onto a press-to-seal Grace Bio-Labs (Bend, Oregon, USA) silicon mold with a thickness of 1 mm. These molds were used to prepare disks of the material for X-ray studies. Mixtures were prepared in 10 % mol/mol increments, from 0 % mol/mol PPP to 100 % mol/mol PPP. The concentrations expressed in both the simulation and experimental portions of this work denote the molar fraction of XPPP dissolved in SSS. X-Ray Studies X-ray studies were conducted at the Advanced Photon Source (APS) at the Argonne National Laboratory (Illinois, USA). Two X-ray scattering regions were studied at the APS: the wide-angle X-ray scattering region (WAXS) and the small angle X-ray scattering region using a pin-SAXS detector. The scattering due to the instrument and the sample holder scattering were subtracted to obtain the intensity. Measurements were carried out in duplicate. Taking the b crystallographic axis as the long molecular axis, the X-

ACS Paragon Plus Environment

Page 3 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

ray intensity as a function of the scattering vector q of the (030) reflection was fitted to both a Lorentzian and a Gaussian function using Graph Pad Prism 5.0 (La Jolla, California, USA). Between the different mixtures, no significant differences in the position of the peak apex of the (030) reflection was observed. The d spacing was calculated from the scattering vector q (Å-1) value as follows: 

 

[1]

DSC Studies Differential Scanning Calorimetry was performed using a Mettler Toledo DSC 1 to obtain calorimetric data. Approximately 7 to 11 mg of sample was deposited into a Mettler Toledo crucible and sealed. An empty crucible containing air was used as a reference. Melting data was collected between 293 K and 353 K using a heating ramp of 5 K/min. The data was analyzed using the StarE software from Mettler Toledo. The temperature at the apex of the melting peak was reported as the melting temperature. The temperature onset the peak was reported as eutectic points. Since the complexity of the PPP/SSS mixtures to resolve the DSC thermograms 21,40 a complete analysis about the decovolution of the DSC peaks is reported in the supporting information, where the eutectic points and relative melting enthalpies are computed. The area under the peak instead was used to compute the global melting enthalpy. Data was collected in triplicate with the average and standard deviation reported. We used the term global melting enthalpy to describe the total energy required to pass into the liquid state, starting from β phase. The global melting enthalpy has been already usefully employed in literature to study mixture of fatty acids 41-42.

(Scheme 1). These crystal structures were used to generate the crystals utilized in the MD melting simulations. The crystals were generated by translating the appropriate unit cell along the a, b (the long molecular axis) and c axes by 11, 2 and 20 times, respectively. The unit cells have so been replicated due to their high asymmetric size for both pure SSS and PPP, in which the a and c parameters differs markedly from the b parameter. For SSS, the a and c cell parameters were 12.00 Å and 5.44 Å, much shorter than the b cell parameter of 51.90 Å45. The PPP cell parameters were very similar to those of SSS, with the a = 11.95 Å and c = 5.45 Å, while b = 46.84 Å44. These experimental cell parameters, as well as the density values obtained from the literature, are reported alongside the values obtained from the simulation in Table 1. The experimental parameters were found to be in good agreement with the simulated cell parameters obtained using the United Atom Force Field (UA-FF) developed by Sum et al.24. This was determined via a validation and testing procedure reported in the Supporting Information (see Table S2 and Table S3, where different force fields, as GROMOS9623, OPLS46 and LargeOPLS47, were evaluated). Due to the good agreement between the experimental and simulated lattice parameters, Sum et al.'s United Atom Force Field was chosen for use in the melting MD simulations. Eutectic crystals were generated by randomly replacing a certain number of SSS molecules in the enlarged SSS β crystal with PPP molecules. Details on how this was achieved, as well as details on the validation of the force field used and the results of the heating MD simulations used to determine the melting temperatures, are reported in the supporting materials.

Molecular dynamics Atomistic MD was performed using GROMACS 4.543 under an NPT ensemble. In this ensemble, temperature and pressure were kept constant using a Berendsen thermostat/barostat. Temperature and pressure couplings of 1 ps and 10 ps were used, respectively. Compressibility was fixed at 1  10 bar-1 and anisotropic pressure coupling was used in the MD simulations. The simulated systems consisted of a total of N = 880 TAG molecules, described at the united atom (UA) level of detail (all atoms except hydrogen are explicitly modeled). The NERD Force Field (FF) was used to describe the atomic interactions. This force field was developed by Sum and coworkers for parametrizing TAG molecules using results from ab initio quantum calculations24. Pure SSS and pure PPP as well as their mixtures were analyzed in the crystalline state (for ~70 ns long simulations at room temperature, T=293 K), and through heating isothermal simulations. Periodic boundary conditions with a cut-off range of 1.1 nm were used. A time step of 2 fs was used for all systems.

Scheme 1. Crystal structure of the pure SSS model system translated 11 x 2 x 20 times. The unit cell is enclosed by a black box in [A] where the respective crystallographic axes are also shown. The SSS model system is shown for two different views: side view parallel to the ab plane [A], and top view parallel to the ac plane [B]. The model system of SSS is periodic in the three spatial dimensions related to the crystallographic axes as shown in panel [A] and [B], where the Miller indices are also reported.

The crystal structure of the β polymorph of both SSS and PPP has been reported by van Langevelde and others44-45

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1. Cell parameters of β-PPP and β-SSS crystal structures obtained from experiments44-45,48 and obtained by means of MD simulation (in parenthesis). Unit Cell β-PPP44 β-SSS45 β-SSS48 Parameter a (Å) 11.95 (13.01) 12.00 (13.12) 14.13 b (Å) 46.84 (48.20) 51.90 (53.29) 53.75 c (Å) 5.45 (4.96) 5.44 (4.97) 5.45 α (°) 73.8 (71.4) 73.8 (71.4) 68.0 β (°) 100.2 (102.2) 100.3 (102.3) 123.2 γ (°) 118.1 (119.9) 117.7 (119.6) 122.5 Calculated Density 1.04 (1.05) 1.03 (1.04) 1.01 (g/cm3) The melting of the crystalline states of pure SSS, pure PPP and their binary mixtures was modeled using heating simulations, starting at a temperature of 293 K. This was achieved as follows: (1) Different copies of the modeled system with different temperatures were generated. Temperatures were generated in increments of 5 K. (2) An isothermal simulation was conducted to allow the different copies of the modeled system (with different temperatures) to equilibrate. Each simulation lasted between 100 ns to 1000 ns, depending on how quickly an equilibrium density value was achieved. (3) A system was considered “melted”, if, during the course of the simulation, an appreciable jump in the density was observed. Additional isothermal simulations (with temperatures increasing in increments of 1 K) were carried out in the 5 K range over which the system melted to determine the simulated melting point with greater precision.

  "   sin $   "  cos ! cos # − cos $   " cos # cos $ − cos !   "   cos $ cos ! − cos # And V is the volume of the triclinic box:   "(1 − )*  ! − cos # − )*  $ + 2cos ! cos # cos $

For the particular case of + + , we obtained, using equation 2 the following: 1  + + 

,

-,

1 1  ℎ +    +    + 2 ℎ + 2      + 2 ℎ [2] where:       sin !   "  sin # 





*./ #

1 − )*  ! − cos  # − )*  $ + 2cos ! cos # cos $



And thus: + +  (1 − )*  ! − cos # − )*  $ + 2cos ! cos # cos $  [3]  sin #

Considering that the triclinic unit cell angles for both SSS and PPP have values close to 90º, the d spacings can be considered to be parallel or orthogonal to the crystallographic axes of the unit cell. In particular, d(0l0) corresponds to the distance between adjacent lamellar layers (see Scheme 1). Using values obtained from the simulated SSS triclinic cell (Table S2 of the Supporting Information) and   1, Equation 3 reduces to:   0.84 

Structural Analysis of MD-Derived Structures The Mercury engine of the Cambridge Crystallographic Data Center (CCDC) (https://www.ccdc.cam.ac.uk/solutions/csdsystem/components/mercury/) was used to perform structural analyses of the simulated crystal structure. Equation 1 was used to obtain the d spacing after fitting simulated X-ray spectra to a Gaussian function, and to the necessary cell parameters (a, b, c sides and α, β, γ angles of the triclinic cell, see Table S4 in Supporting Information) at T = 293 K, using the following equation:

Page 4 of 14

The global melting enthalpy, i.e. the total energy required to the crystal to pass into liquid state, was computed as difference between enthalpy values calculated in liquid state and in the crystal β phase (at T=293K) . The enthalpy values are indeed computed with the GROMACS tool g_energy 43 for each trajectory in a given phase. With g_energy tool we are able to calculate the individual terms ∆U and ∆pV including the enthalpy. The orientational order parameter P2 was used to quantify the crystalline order of the TAG molecules in the solid state (Equation 4). P2 is frequently used for estimating the structure of liquid-crystalline systems49-50. This parameter measures the average degree of alignment of a vector u with a vector depicting a preferred direction n, called the director: 3 1 P  〈 6 ∙ 8 − 〉 [4] 2 2

ACS Paragon Plus Environment

Page 5 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

The vector n was obtained from the simulation trajectories. While the vector u was primarily the vector connecting the first and last carbons of the last three carbons of each of the alkyl chains of SSS and PPP. The instantaneous order parameter is determined for each configuration by diagonalising the following order matrix, R: ?

:  ;[3