Cn+2

Sep 27, 2013 - The phase behaviors of binary consecutive even normal alkane (n-alkane) mixtures (n-CnH2n+2/n-Cn+2H2n+6, with mass ratios of 90/10 and ...
11 downloads 17 Views 4MB Size
Download PDF
Article pubs.acs.org/JPCB

Phase Transition Behavior of a Series of Even n‑Alkane Cn/Cn+2 Mixtures Confined in Microcapsules: From Total Miscibility to Phase Separation Determined by Confinement Geometry and Repulsion Energy Xia Gao, Dongsheng Fu, Yunlan Su,* Yong Zhou, and Dujin Wang* Beijing National Laboratory for Molecular Sciences, Key Laboratory of Engineering Plastics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China ABSTRACT: The phase behaviors of binary consecutive even normal alkane (n-alkane) mixtures (n-CnH2n+2/n-Cn+2H2n+6, with mass ratios of 90/10 and 10/90) with different average carbon numbers n̅ both in the bulk state (abbreviated as Cn/Cn+2) and in nearly monodisperse microcapsules (abbreviated as m-Cn/Cn+2), have been investigated by the combination of differential scanning calorimetry and temperature-dependent X-ray diffraction. The phase behavior of n-alkane mixtures gradually shifts from complete phase separation, partial miscibility to total miscibility in both bulk and microcapsules with the increase of average carbon numbers n.̅ There are critical points for average carbon numbers of Cn/Cn+2, where the corresponding mixtures exhibit coexistence of a triclinic phase (formed by alkane with a longer chain) and an orthorhombic ordered phase (formed by the two components of mixtures). Due to the confinement from hard shells of microcapsules, the critical points of m-Cn/Cn+2 are smaller than those of Cn/Cn+2. Such a phase behavior originates from the delicate combined action of confinement and repulsion energy for the encapsulated n-alkane mixtures with different average carbon numbers n̅. When n̅ is less than the critical point, the repulsion energy between the two kinds of molecules exceeds the suppression effect of confinement, and phase separation occurs in microcapsules. It is believed that the average carbon number is another important factor that exerts strong negative influence on the phase separation of m-Cn/Cn+2 systems.

■

INTRODUCTION The normal alkanes (n-CnH2n+2, n-alkanes) are the simplest organic series and form the nonpolar part of lipids, surfactants, liquid crystals, and polymers, thus attracting people’s significant attention for many years.1−3 The crystallization behaviors of bulk n-alkanes have been studied extensively. The surface freezing phenomenon, rotator phases, and the well-known odd−even effect of n-alkanes make them different from other compounds. The surface freezing phenomenon occurring for chain lengths ranging from n = 15 to n = 50 is the formation of a crystalline monolayer on the surface of liquid n-alkanes at about 3 °C above the bulk crystallization temperature.4−6 The molecules in the monolayer are hexagonally packed and oriented normal to the surface. Rotator phases are also identified, occurring between the low-temperature crystalline phases and the isotropic liquid.7−9 These rotator phases have long-range transitional order but lack long-range order with respect to rotation about the molecule’s long axis. In addition, odd and even n-alkanes show different low-temperature crystalline forms, which is known as the odd−even effect.10 However, n-alkane mixtures are more common in practical applications such as the phase change materials (PCMs) and are frequently used as model systems at the molecular level, especially for polydisperse polyethylene. Some previous work © 2013 American Chemical Society

indicated that mixing different chain lengths of alkanes reduces the interaction of interlayer coupling and increases void volume and lamellar surface roughness, thus leading to a dramatic influence on the phase behaviors of n-alkane mixtures.11−13 One of the specific phase behaviors in these systems is that the stability of the rotator phase is enhanced.14,15 In addition, binary n-alkane mixtures have been proven to undergo a phase separation process in low-temperature ordered crystals, which endows the mixtures with particular desirable properties. It has been shown that chain length difference is one of the major factors that can greatly affect the phase separation.16 For a binary mixture, if the chain length difference for the two nalkanes is 3 or 4 carbons, it usually undergoes a detectable phase separation under all compositions. Recently, researchers began to have an eye on the effect of average carbon numbers on the phase behaviors of n-alkane binary mixtures.17−19 It was found that the excess enthalpies (HE) and the thermal diffusion coefficients (DT) of mixtures decrease with a decrease of the average carbon numbers, meaning when the two components Received: Revised: Accepted: Published: 13914

July 12, 2013 September 10, 2013 September 27, 2013 September 27, 2013 dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

and melt-mixed. Using melamine−formaldehyde (M−F) resin as the shell material and n-alkane mixtures as core material, microcapsules were prepared by in situ polymerization according to the literature.33,34 This method provides us with nearly monodispersed and highly heat-resistant microcapsules, inside which n-alkane mixtures are confined to individually small microdomains surrounded by the noncrystalline wall of M−F resin. The particle size and surface morphology of the prepared microcapsules were examined by a JEOL-JSM-6700F scanning electron microscope (SEM; Figure 1), fitted with a field

become more similar, mixtures respond very similarly to temperature gradients and hardly phase separate.17,18 Furthermore, Deutsch and co-workers have determined that the repulsion energy (ωb,s) of a mixture depends only on the square of the reduced chain length mismatch Δn/n,̅ where Δn and n̅ are the chain length difference and average carbon numbers of the two components, respectively.19 Up to now, however, there are few reports on the relationship between the average carbon numbers and the phase behaviors of binary n-alkane mixtures with smaller carbon numbers and with chain length differences, which limits one’s ability to gain a better understanding of phase separation of n-alkane mixtures. As alkanes and their mixtures were confined in media, novel rotator phases appeared and the nucleation kinetics became different from that in bulk.20−26 It has also been demonstrated that confinement can suppress the phase separation behavior of simple fluids in porous materials27,28 and polymer blends in thin film.29 It would thus be of interest to learn how geometrical confinement affects the phase separation of binary n-alkane mixtures. Previously, we investigated the mixtures of odd−even n-alkanes (C18/C19)30 and even−even n-alkanes (C18/C20)31,32 in nearly monodisperse microcapsules. In mC18/C19, it has been found that the solid−solid phase separation at low temperature, which is originally manifest in the bulk counterparts, was suppressed to a different extent in microcapsules.30 Obviously, the confined environment can exert strong positive influence on suppressing the solid−solid phase separation induced by chain-length differences. The mixture of m-C18/C20 with different composition (10/90 and 90/10, wt %) showed different solid−solid phase behavior and demonstrated that the composition can also affect the phase separation of n-alkane mixtures in microcapsules.31 Considering the significant role of repulsion energy in phase separation and the relationship between repulsion energy and average carbon numbers, it is quite essential to expect more fascinating results on confined binary n-alkane mixtures with the same chain length difference but different average carbon numbers. In the present work, we focus on the phase behaviors of two n-CnH2n+2/n-Cn+2H2n+6 series (mass ratios are 90/10 and 10/90 separately) with different average carbon numbers n̅ both in the bulk state (abbreviated as Cn/Cn+2) and in microcapsules (abbreviated as m-Cn/Cn+2). It is observed that, when n̅ changes, the phase behaviors of Cn/Cn+2 mixtures are different. When n̅ equals a special value, the corresponding mixture exhibits partial phase separation and shows the coexistence of a triclinic phase and an orthorhombic ordered phase. The smaller the n̅ of the mixtures are, the easier phase separation occurs. This variation phenomenon is observed both in bulk and microcapsules. Moreover, the relationship between phase behaviors and n̅ values will be discussed on the basis of the experimental observations.

Figure 1. SEM micrographs of microcapsules containing n-alkane mixtures prepared by in situ polymerization of M−F resin.

emission source and operated at an accelerating voltage of 10 kV. The differential scanning calorimetery (DSC) measurements were carried out on a TA Q2000 calorimeter at a cooling/heating rate of 2 °C/min. Specimens were heated from room temperature to about 10 °C above the corresponding mixtures’ melting temperature, cooled to about 10 °C below their freezing temperature, and heated again to melt. The first cooling and second heating thermograms were recorded. Temperature-dependent X-ray diffraction (XRD) experiments were performed on an X′Pert Pro MPD X-ray diffractometer with the temperature region similar to that for DSC measurements, using Cu Kα radiation (1.54 Å), power of 40 mA/40 kV, and rotating angle 2θ = 5−40°. The heating and cooling rates were all 2 °C/min, and at each temperature point, the samples were equilibrated for about 5 min before data collection.

■

RESULTS AND DISCUSSION Phase Transition Behaviors of Cn/Cn+2 Mixtures in Both Bulk and Microcapsules. 1. Cn/Cn+2 = 90/10. Figure 2

■

MATERIALS AND METHODS n-C14H30, n-C16H34, n-C18H38, n-C20H42, n-C22H46, and nC24H50 with purity >99% were purchased from Sigma-Aldrich Co., and all reagents were used as received. All binary n-alkane mixtures in the experiments were designed with the same chainlength difference Δn = 2, and the mass fraction of all components was fixed by 10% and 90%. In the text, φn is the molar fraction of chain length n, and n̅ = n + 2(1 − φn). The corresponding molar fraction φn and the average chain length (denoted as n̅) were calculated and shown in Table 1. The solid bulk components were weighted in the desired mass fraction

Figure 2. DSC curves of C16/C18 = 90/10 during the cooling process: (upper line) sample in microcapsules and (lower line) sample in bulk. Black arrow indicates surface freezing peak in the microcapsule sample. 13915

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

Figure 3. Temperature-dependent XRD patterns of C16/C18 = 90/10 at selected temperatures during the cooling process: (A) bulk sample and (B) microencapsulated sample. The cooling rate was 2 °C/min.

Figure 4. Temperature-dependent XRD patterns of C14/C16 = 90/10 at selected temperatures during the cooling process: (A) bulk sample and (B) microencapsulated sample. The cooling rate was 2 °C/min.

phases, which are formed by the two components crystallizing into their own low-temperature stable crystal form separately. This indicates that a complete phase separation occurs at −20 °C in C16/C18 = 90/10. Compared to C16/C18 = 90/10, the low-temperature crystal structure is different in m-C16/C18 = 90/10, shown in Figure 3B. As the sample is further cooled to −20 °C, four characteristic diffraction peaks of (010), (011), (100), and (111) emerge, corresponding to the triclinic phase, while the characteristic diffraction peaks of (110) for the orthorhombic phase still remains. Obviously, the low-temperature crystal structure of m-C16/C18 = 90/10 is the coexistence of the triclinic phase of C16 and the orthorhombic ordered phase formed by the two components of mixtures. In other words, complete phase separation, occurring in bulk state, is partially suppressed for m-C16/C18 = 90/10. Here, partial phase separation is used to refer to this kind of phase behavior. The distinct difference of the phase behaviors between C16/C18 = 90/10 and m-C16/C18 = 90/10 can be attributed to the confined geometry of microcapsules which has suppressed phase separation in m-C18/C19 = 90/1030 and m-C18/C20.32 Apparently, this present observation is different from the phase behaviors in C18/C20 = 90/10 (n̅ = 18.18) mixtures both in bulk and microcapsules reported in a previous report.31 Partial phase separation can be found in C18/C20 = 90/10, while the m-C18/C20 = 90/10 sample transforms into a stable orthorhombic phase at low temperature, showing a total elimination of phase separation because of the suppression effect of confinement. However, with decreasing n̅ from 18.18 to 16.18, the low-temperature crystal structures of m-C16/C18 = 90/10 change to a coexistence of the triclinic phase and the orthorhombic phase (partial phase separation). Given this novel phenomenon, we design other C14/C16 = 90/10 (n̅ = 14.18) mixtures with smaller average carbon number in both

shows the crystallization behaviors of C16/C18 = 90/10 (weight ratio, n̅ = 16.18) both in the bulk (lower line) and microencapsulated states (upper line). For the bulk C16/C18 = 90/10, a big exothermic peak appearing at 14 °C corresponds to the liquid−rotator transition. Another peak is observed at −9.7 °C, which is regarded as the transition from the rotator to the low-temperature crystal phase. For the m-C16/C18 = 90/10 mixture, however, a small sharp exothermic peak (see the arrow in Figure 2) emerges above the freezing temperature about 3 °C during the cooling process, corresponding to the surface freezing.32,35,36 In addition, the transition temperature from the rotator to the low-temperature crystal is lower than that in bulk, which means the enhanced stability of rotator phase. The special three-dimensional confinement effect provided by microcapsules’ shells has been proved to suppress the diffusion movement of alkanes molecules,32 thus allowing C18 better dispersed with C16, so the rotator phase formed by two components can survive much longer. To further characterize the crystal structures and phase transitions during the crystallization processes of both samples, temperature-dependent XRD measurements were performed. As shown in Figure 3A, when C16/C18 = 90/10 is heated for 5 min above 17 °C, the mixture exists as the isotropic liquid state, characterized by a single halo at 2θ = 22°. With temperature decreasing to 15 °C, two characteristic peaks of (110) and (200) appear which indicate the appearance of the orthorhombic rotator phase I (RI). Further cooling the mixture to −10 °C, four characteristic diffraction peaks of (010), (011), (100), and (111) emerge, corresponding to the triclinic phase, while the characteristic peaks (110) of RI still remain. When the mixture is further cooled to −20 °C, it is observed that RI disappears and the four characteristic peaks of the triclinic phase remain. The triclinic ordered phase comprises in fact two 13916

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

Table 1. Phase Behavior of Cn/Cn+2 Mixture with Different n̅ a

a

Note: +, phase separated fully; 0, phase separated partly; critical points; −, miscible mixtures.

Figure 5. Schematic illustration of the crystal structure and phase behaviors of b-Cn/Cn+2 = 90/10 (upper part) and m-Cn/Cn+2 = 90/10 (lower part): (blue line) for molecule of Cn and (red line) for Cn+2. The crystal structure illustrated by one kind of lines is the triclinic phase, and others are the orthorhombic ordered phase.

phase separation in bulk and microcapsules, respectively. Due to the confinement effect of hard shells provided by microcapsules, phase separation in m-Cn/Cn+2 = 90/10 mixtures is suppressed to some extent compared to their comparative bulk samples. Therefore, nm ̅ ,c of m-Cn/Cn+2 = 90/10 mixtures decreases to 16.18 in comparison with n̅b,c of Cn/Cn+2 = 90/10 (18.18). As shown in Table 1 and Figure 5, for Cn/Cn+2 = 90/ 10 mixtures whether in bulk or microcapsules, the change tendencies of phase behaviors are only closely related to n̅c.

bulk and microcapsules. As shown in Figure 4, when phase transitions reach completion, four characteristic diffraction peaks of (010), (011), (100), and (111) for the triclinic structure emerge in both bulk (Figure 4A) and microcapsules (Figure 4B). It is obvious that complete phase separations occur in both the bulk and microcapsule when n̅ of Cn/Cn+2 further decreases to 14.18. For convenience, critical points nb̅ ,c and nm ̅ ,c are the average carbon numbers of Cn/Cn+2 mixtures which exhibit partial 13917

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

sample, which is reminiscent of enhanced rotator phase stability because of the confinement effect. In order to clarify the relationship between phase behaviors and n,̅ we design other Cn/Cn+2 = 10/90 mixtures and characterize the crystal structures of all specimens by means of temperature-dependent XRD. As shown in Figure 7, for bulk samples, the low-temperature structures of mixtures with different n̅ are various. According to the above-mentioned discussion, the critical point nb̅ ,c of Cn/Cn+2 = 10/90 equals 21.78 because the XRD pattern of C20/C22 (n̅ = 21.78; Figure 7C) shows the coexistence of the orthorhombic ordered phase formed by two components and the triclinic phase of C22. As expected, when n̅ < n̅c complete phase separation occurs for C16/C18 = 10/90 (n̅ = 17.77) and C18/C20 = 10/90 (n̅ = 19.77), as shown in Figure 7A,B, while C22/C24 = 10/90 with n̅ = 23.78 (n̅ > n̅c) exhibits a total miscibility characterized by the orthorhombic ordered phase shown in Figure 7D. For Cn/Cn+2 = 10/90 mixtures encapsulated in microcapsules, the enhanced surface freezing phenomenon and the lower solid−solid transition temperatures compared to the bulk are two shared features (DSC results not shown here) similar to other microencapsulated samples.32,36 In addition, the phase behaviors of m-Cn/Cn+2 = 10/90 are also observed by temperature-dependent XRD. From Figure 8A, the lowtemperature crystal structure of m-C16/C18 = 10/90 (n̅ = 17.77) is triclinic phase characterized by four diffraction peaks (010), (011), (100), and (111), which indicates that the microencapsulated mixture undergoes complete phase separation. The low-temperature crystal structure of m-C18/C20 = 10/ 90 (n̅ = 19.77) is the coexistence of the triclinic phase and the orthorhombic ordered phase;32 see in Figure 8B. On account of the suppression effect of microcapsules, the m-C18/C20 = 10/90 mixture undergoes partial phase separation while the counterpart bulk sample exhibits complete phase separation, which

When the average number is smaller than the critical point (n̅ < n̅c), complete phase separation occurs in corresponding Cn/ Cn+2 = 90/10 mixtures; for n̅ > n̅c, there is no phase separation in Cn/Cn+2 = 90/10 mixtures. 2. Cn/Cn+2 = 10/90 Mixtures. As another fundamental factor, the composition can also exert remarkable influence on the phase behavior of binary n-alkane mixtures with small chain length difference. For bulk samples, C16/C18 = 10/90 is first trapped into a rotator phase at 23.2 °C, and then converts to the stable low-temperature crystal at 17.6 °C. As for m-C16/C18 = 10/90, one more exothermic peak emerges above the freezing temperature about 3 °C during the crystallization process, which is attributed to the surface freezing phenomenon (see in Figure 6). Similar to other microencapsulated alkanes and

Figure 6. DSC curves of C16/C18 = 10/90 during the cooling process: (upper line) sample in microcapsules and (lower line) sample in bulk. Black arrow indicates surface freezing peak in the microcapsule sample.

alkane mixtures,30,36 the solid−solid phase transition temperature of m-C16/C18 = 10/90 is also lower than that of bulk

Figure 7. Temperature-dependent XRD patterns of b-Cn/Cn+2 = 10/90 at selected temperatures during the cooling process: (A) n = 16; (B) n = 18; (C) n = 20; (D) n = 22. The cooling rate was 2 °C/min. 13918

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

Figure 8. Temperature-dependent XRD patterns of m-Cn/Cn+2 = 10/90 at selected temperatures during the cooling process: (A) n = 16; (B) n = 18. The cooling rate was 2 °C/min.

easily destroy the layered structure. Therefore, the CH2−CH2 intermolecular interactions become dominant, and the formation of an orthorhombic rotator phase is preferred, leading to a more stable rotator phase and less stable triclinic phase.14,15 Otherwise, for Cn/Cn+2 = 10/90 mixtures, the Cn with a shorter chain length is the minor component, the crystal lattice of the Cn+2 will be less affected by the mixing, and the Cn molecules can be easily excluded from the crystal lattice of Cn+2. Therefore, phase separation is observed in Cn/Cn+2 = 10/90 mixtures more easily than in Cn/Cn+2 = 90/10 mixtures, which explains the larger critical average carbon number for Cn/Cn+2 = 10/90 mixtures. Competition between Confinement and Repulsion Energy with Different Average Carbon Numbers. When the average carbon numbers n̅ decrease, the low-temperature phase behaviors of Cn/Cn+2 mixtures, either in bulk or microcapsules, show a change tendency from total miscibility, partial phase separation to complete phase separation. There are critical points which divide the mixtures into two stages; one is homogeneous mixtures when n̅ > nc̅ , and the other is immiscible mixtures when n̅ < nc̅ . From XRD results, it is evident that phase separation occurs after the orthorhombic RI phase has formed. Here, the orthorhombic RI phase is in fact a solid solution formed by the binary n-alkanes mixtures. Whether the orthorhombic RI phase is stable depends on chain-length difference and crystal symmetries of two components of n-alkane mixtures.37,38 In our experiment, even alkanes with n < 26 have the same crystal symmetries and form a triclinic phase. Then, the only factor considered here, which determines phase behaviors, is the chain-length difference. According to the Gibbs free energy equation ΔG = ΔH − TΔS, the phase diagram of Cn/Cn+2 binary mixtures is dominated by a delicate balance between mixing entropy ΔS, which drives the system toward homogeneity, and the repulsion energy between the unlike constituent molecules (chain-length difference Δn = 2), which drives them to segregate and determines the mixing enthalpy ΔH. For mixtures with the same mass ratio (10/90, or 90/10), ΔS = n[φ ln φ + (1 − φ) ln(1 − φ)] of all samples are calculated to be nearly the same. Then, the value of ΔG mainly lies on ΔH, which is believed to be related to repulsion energy. The repulsion energy is determined by the so-called interaction parameter ω, which is the energy change upon replacing one molecule of the pure phase of one species by a molecule of the other species.19 It has been reported that ω follows a linear (Δn/n)̅ 2 dependence in rotator phases. When chain-length difference Δn of all alkane mixtures is identical to being 2, the interaction parameter ω of Cn/Cn+2 mixtures depends linearly on (1/n)̅ 2. When n̅ is less than n̅b,c, ω of the corresponding

indicates that the critical point nm ̅ ,c of m-Cn/Cn+2 = 10/90 mixtures is 19.77. When n̅ < nm ̅ ,c, m-C16/C18 = 10/90 mixtures (n̅ = 17.77) undergo complete phase separation as expected. Predicted by the present phenomenon, with an increase in n̅ to 21.78 (n̅ > n̅m,c), the low-temperature crystal structure of mC20/C22 = 10/90 is the orthorhombic ordered phase, which exhibits a miscible mixture. Unfortunately, the present preparation method cannot allow us to encapsulate Cn/Cn+2 with longer chain lengths in the M−F resin shells, because of the limited emulsification capability of nonionic surfactant and the lower reactive temperature.33,34 However, this can not affect the relationship between the average carbon numbers and phase behaviors of Cn/Cn+2, which is summarized in Table 1. Furthermore, it is observed that the critical points of Cn/Cn+2 = 90/10 and Cn/Cn+2 = 10/90 mixtures are different and nc̅ of Cn/Cn+2 = 10/90 mixtures are larger (see in Table 1). To clarity, the R/C transition temperature (R, rotator phase; C, low-temperature crystal phase) of mixtures is plotted against n̅, shown in Figure 9. After encapsulation, the R/C transition

Figure 9. Relations between the average carbon number (n)̅ of the binary n-alkane mixtures and the R/C transition temperature (R, rotator phase; C, low-temperature crystal phase) both in bulk and microcapsules: (black open squares) Cn/Cn+2 = 90/10; (black solid squares) m-Cn/Cn+2 = 90/10; (red open circles) Cn/Cn+2 = 10/90; (red solid circles) m-Cn/Cn+2 = 10/90; (black arrows) temperature hysteresis after the mixtures are encapsulated.

temperatures of m-Cn/Cn+2 mixtures are lower than their bulk counterparts. However, the temperature hysteresis between the bulk and microcapsule for Cn/Cn+2 = 90/10 mixtures is larger than that for Cn/Cn+2 = 10/90 mixtures. As we all know when mixing two n-alkanes with different chain lengths, the terminal CH3−CH3 interactions of crystal phases are greatly weakened due to the chain length difference.12,13 For Cn/Cn+2 = 90/10 mixtures, a small amount of Cn+2 with longer chain length can 13919

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

confined n-alkanes and shed some new lights on designing and preparing these kinds of PCMs.

mixture increases greatly and the repulsion energy between the two components dominates in ΔG and induces phase separation in the solid phase. When n̅ is larger than nb̅ ,c, ω of the mixture weakens linearly and intensively so that the mixing entropy dominates, resulting in a homogeneous mixture. For m-Cn/Cn+2 binary mixtures, even though the phase behavior shows the same tendency as that in bulk samples, the critical points become smaller than their bulk counterpart mixtures which is related to the confinement effect of microcapsules. For alkanes in pores, the tortuous character of the pore network and rough pore walls act as sources of random strain fields that stabilize the disordered rotator phase, thereby allowing the rotator phase to survive at lower temperature.20,21,26 In microcapsules, because of rough shells of microcapsules, the layered structure of Cn/Cn+2 mixtures indicated by the (00l) series diffraction peaks (shown in Figure 7 and not shown in Figure 8) disappears when they are encapsulated.30 As we all know, the weaken layered structure of low-temperature ordered phases and the suppression of terminal methyl−methyl interactions favor the reduced stability of the triclinic phase and the enhanced stability of rotator phases of alkane mixtures.30,31 Therefore, it is observed that the solid−solid phase transition temperatures of m-Cn/Cn+2 are lower than their bulk counterparts (see in Figures 2 and 6) and the stability of rotator phases is enhanced. In addition, the hard shells of microcapsules can suppress the longitudinal chain diffusion of alkane molecules, resulting in further enhanced stabilization of rotator phases36 and then better mixing of the two components.30,31 Conclusively, the confined environment leads to the enhanced kinetic lifetime of the RI phase, which contributes to suppression of the solid−solid phase separation of the binary n-alkane system. As a result, the low-temperature structure of m-C16/C18 = 90/10 shows the coexistence of a triclinic phase and an orthorhombic ordered phase (see in Figure 3B) and the total phase separation of C16/C18 = 90/10 mixtures is partially suppressed when they are encapsulated. So the critical points of Cn/Cn+2 = 90/10 change from 18.18 (for bulk samples) to 16.18 (for microcapsules samples); see in Table 1 and Figure 5. The same tendency goes for Cn/Cn+2 = 10/90 mixtures. In spite of the confinement effect from microcapsules, the total phase separation still occurs in m-C14/ C16 = 90/10 mixtures (see in Figure 4) when the average carbon number is smaller than nm ̅ ,c. According to the literature, in bulk, the larger the Δn/n̅ of binary alkane mixtures is, the more easily phase separation occurs.17,18,39 When Δn/n̅ becomes larger with smaller n,̅ the repulsion energy between the two components becomes stronger and the phase separation cannot be suppressed even when the mixture is encapsulated in the confined environment for n̅ < n̅m,c. Therefore, the confined environment suppresses only the phase separation induced by the repulsion energy to some extent. As mentioned above, confinement and repulsion energy compete with each other and exert a prominent effect on the phase behavior of binary even−even n-alkane mixtures. When the microcapsules are the same in diameters, the confinement effect on all mixtures is regarded as the same. Therefore, in addition to the chain-length difference and confinement, n̅ is another important factor that determines the phase separation and other phase behaviors of the n-alkane mixtures. Furthermore, the present work on the phase separation of the confined n-alkane solid solutions can be propitious to better understand the phase change mechanism of PCMs containing

■

CONCLUSION In summary, we have investigated the phase behavior of two series of even−even binary mixtures with carbon difference Δn = 2 both in bulk and encapsulated in microcapsules. Whether in bulk or microcapsules, the change trend of the phase behavior with average carbon numbers applies to both Cn/Cn+2 = 90/10 and Cn/Cn+2 = 10/90. With decreasing n̅ the repulsion energy between two components of Cn/Cn+2 weakens, and the phase behaviors of Cn/Cn+2 exhibit from total miscibility to phase separation around the critical point, nb̅ ,c. Encapsulated in microcapsules, the hard shells suppress the phase separation to some extent, so n̅m,c is less than that of the bulk counterparts. Therefore, these findings can contribute to highlighting our understanding of phase separation of other mixed systems such as waxes, lipids, and polymer blends in confined geometry.

■

AUTHOR INFORMATION

Corresponding Authors

*Tel.: +86-10-82618533. Fax: +86-10-82612857. E-mail: ylsu@ iccas.ac.cn. *Tel.: +86-10-62556180. Fax: +86-10-82362045. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the National Natural Science Foundation of China (Grant 51103166) and China National Funds for Distinguished Young Scientists (Grant 50925313) for financial support.

■

REFERENCES

(1) Doolittle, A. K.; Peterson, R. H. Preparation and Physical Properties of a Series of n-Alkanes. J. Am. Chem. Soc. 1951, 73, 2145− 2151. (2) Orwoll, R. A.; Flory, P. J. Equation-of-State Parameters for Normal Alkanes. Correlation with Chain Length. J. Am. Chem. Soc. 1967, 89, 6814−6822. (3) Yamamoto, T.; Nozaki, K. Evidence of Active Chain Diffusion in the Rotator Phase of n-Alkanes: Solid-State Mixing of C21H44 and C23H48. Polymer 1994, 35, 3340−3342. (4) Wu, X. Z.; Sirota, E. B.; Sinha, S. K.; Ocko, B. M.; Deutsch, M. Surface Crystallization of Liquid Normal-Alkanes. Phys. Rev. Lett. 1993, 70, 958−961. (5) Wu, X. Z.; Ocko, B. M.; Sirota, E. B.; Sinha, S. K.; Deutsch, M.; Cao, B. H.; Kim, M. W. Surface Tension Measurements of Surface Freezing in Liquid Normal Alkanes. Science 1993, 261, 1018−1021. (6) Ocko, B. M.; Wu, X. Z.; Sirota, E. B.; Sinha, S. K.; Gang, O.; Deutsch, M. Surface Freezing in Chain Molecules: Normal Alkanes. Phys. Rev. E 1997, 55, 3164−3182. (7) Ungar, G.; Masic, N. Order in the Rotator Phase of n-Alkanes. J. Phys. Chem. 1985, 89, 1036−1042. (8) Sirota, E. B.; King, H. E., Jr.; Singer, D. M.; Henry, H. Shao, Rotator Phases of the Normal AlkanesAn X-Ray-Scattering Study. J. Chem. Phys. 1993, 98, 5809−5824. (9) King, H. E., Jr.; Sirota, E. B.; Shao, H.; Singer, D. M. A Synchrotron X-ray Scattering Study of the Rotator Phases of the Normal Alkanes. J. Phys. D: Appl. Phys. 1993, 26, B133−B136. (10) Mnyukh, Y. V. The Structure of Normal Paraffins and of Their Solid Solutions. J. Struct. Chem. 1960, 1, 346−365. (11) Jouti, B.; Bourdet, J. B.; Bouroukba, M.; Dirand, M. Structural Behavior of n-Tricosane and n-Pentacosane Mixtures at 20 °C. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1995, 270, 159−173. 13920

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921

The Journal of Physical Chemistry B

Article

(12) Dirand, M.; Achour, Z.; Jouti, B.; Sabour, A.; Gachon, J. C. Binary Mixtures of n-Alkanes. Phase Diagram Generalization: Intermediate Solid Solutions, Rotator Phases. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1996, 275, 293−304. (13) Jouti, B.; Provost, E.; Petitjean, D.; Bouroukba, M.; Dirand, M. Phase Diagram of n-Tricosane and n-Pentacosane Mixtures. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1996, 287, 275−283. (14) Sirota, E. B.; Singer, D. M. Phase Transitions among the Rotator Phases of the Normal Alkanes. J. Chem. Phys. 1994, 101, 10873− 10882. (15) Sirota, E. B.; King, H. E., Jr.; Shao, H. H.; Singer, D. M. Rotator Phase in Mixtures of n-Alkanes. J. Phys. Chem. 1995, 99, 798−804. (16) Snyder, R. G.; Clavell-Grunbaum, D.; Strauss, H. L. Spinodal Phase Separation of Unstable Solid-State Binary n-Alkane Mixtures. J. Phys. Chem. B 2007, 111, 13957−13966. (17) Leahy-Dios, A.; Firoozabadi, A. Molecular and Thermal Diffusion Coefficients of Alkane−Alkane and Alkane−Aromatic Binary Mixtures: Effect of Shape and Size of Molecules. J. Phys. Chem. B 2007, 111, 191−198. (18) dos Ramos, M. C.; Blas, F. J. Examination of the Excess Thermodynamic Properties of n-Alkane Binary Mixtures: A Molecular Approach. J. Phys. Chem. B 2005, 109, 12145−12153. (19) Sloutskin, E.; Sirota, E. B.; Gang, O.; Wu, X. Z.; Ocko, B. M.; Deutsch, M. Surface and Bulk Interchange Energy in Binary Mixtures of Chain Molecules. Eur. Phys. J. E: Soft Matter Biol. Phys. 2004, 13, 109−112. (20) Huber, P.; Wallacher, D.; Albers, J.; Knorr, K. Quenching of Lamellar Ordering in an n-Alkane Embedded in Nanopores. Europhys. Lett. 2004, 65, 351−357. (21) Huber, P.; Soprunyuk, V.; Knorr, K. Structural Transformations of Even-Numbered n-Alkanes Confined in Mesopores. Phys. Rev. E 2006, 74, No. 031610. (22) Montenegro, R.; Antonietti, M.; Mastai, Y.; Landfester, K. Crystallization in Miniemulsion Droplets. J. Phys. Chem. B 2003, 107, 5088−5094. (23) Montenegro, R.; Landfester, K. Metastable and Stable Morphologies During Crystallization of Alkanes in Miniemulsion Droplets. Langmuir 2003, 19, 5996−6003. (24) Fette, E. V.; Pham, A.; Adalsteinsson, T. Crystallization and Melting Transitions of Hexadecane Droplets in Polystyrene Nanocapsules. J. Phys. Chem. B 2008, 112, 5403−5411. (25) Black, J. K.; Tracy, L. E.; Roche, C. P.; Henry, P. J.; Pesavento, J. B.; Adalsteinsson, T. Phase Transitions of Hexadecane in Poly(alkyl methacrylate) Core−Shell Microcapsules. J. Phys. Chem. B 2010, 114, 4130−4137. (26) Okazaki, M.; Toriyama, K.; Anandan, S. Dynamics and Packing Mode of Long-Chained n-Alkane Molecules in the Nano-Channel of MCM-41. Chem. Phys. Lett. 2005, 401, 363−367. (27) Alba-Simionesco, C.; Coasne, B.; Dosseh, G.; Dudziak, G.; Gubbins, K. E.; Radhakrishnan, R.; Sliwinska-Bartkowiak, M. Effects of Confinement on Freezing and Melting. J. Phys.: Condens. Matter 2006, 18, R15−R68. (28) Jaehnert, S.; Chavez, F. V.; Schaumann, G. E.; Schreiber, A.; Schoenhoff, M.; Findenegg, G. H. Melting and Freezing of Water in Cylindrical Silica Nanopores. Phys. Chem. Chem. Phys. 2008, 10, 6039−6051. (29) Gelb, L. D.; Gubbins, K. E.; Radhakrishnan, R.; SliwinskaBartkowiak, M. Phase Separation in Confined Systems. Rep. Prog. Phys. 1999, 62, 1573−1659. (30) Jiang, K.; Su, Y.; Xie, B.; Jiang, S.; Zhao, Y.; Wang, D. Effect of Geometrical Confinement on the Nucleation and Crystallization Behavior of n-Alkane Mixtures. J. Phys. Chem. B 2008, 112, 16485− 16489. (31) Fu, D.; Liu, Y.; Su, Y.; Liu, G.; Wang, D. Crystallization Behavior of Binary Even−Even n-Alkane Mixtures in Microcapsules: Effect of Composition and Confined Geometry on Solid−Solid Phase Separation. J. Phys. Chem. B 2011, 115, 4632−4638. (32) Fu, D.; Liu, Y.; Liu, G.; Su, Y.; Wang, D. Confined Crystallization of Binary n-Alkane Mixtures: Stabilization of a New

Rotator Phase by Enhanced Surface Freezing and Weakened Intermolecular Interactions. Phys. Chem. Chem. Phys. 2011, 13, 15031−15036. (33) Xie, B.; Shi, H.; Liu, G.; Zhou, Y.; Wang, Y.; Zhao, Y.; Wang, D. Preparation of Surface Porous Microcapsules Templated by SelfAssembly of Nonionic Surfactant Micelles. Chem. Mater. 2008, 20, 3099−3104. (34) Liu, G.; Xie, B.; Fu, D.; Wang, Y.; Fu, Q.; Wang, D. Preparation of Nearly Monodisperse Microcapsules with Controlled Morphology by in situ Polymerization of a Shell Layer. J. Mater. Chem. 2009, 19, 6605−6609. (35) Xie, B.; Shi, H.; Jiang, S.; Zhao, Y.; Han, C. C.; Xu, D.; Wang, D. Crystallization Behaviors of n-Nonadecane in Confined Space: Observation of Metastable Phase Induced by Surface Freezing. J. Phys. Chem. B 2006, 110, 14279−14282. (36) Xie, B.; Liu, G.; Jiang, S.; Zhao, Y.; Wang, D. Crystallization Behaviors of n-Octadecane in Confined Space: Crossover of Rotator Phase from Transient to Metastable Induced by Surface Freezing. J. Phys. Chem. B 2008, 112, 13310−13315. (37) Mnyukh, Y. V. The Structure of Normal Paraffins and of Their Solid Solutions. J. Struct. Chem. 1960, 1, 346−365. (38) Matheson, R. R. A Simple Thermodynamic Analysis of Solid− Solution Formation in Binary Systems of Homologous ExtendedChain Alkanes. Polymer 1985, 26, 288−292. (39) Mondieig, D.; Rajabalee, F.; Metivaud, V.; Oonk, H. A. J.; Cuevas-Diarte, M. A. Alkane Binary Molecular Alloys. Chem. Mater. 2004, 16, 786−798.

13921

dx.doi.org/10.1021/jp406896n | J. Phys. Chem. B 2013, 117, 13914−13921