Fractionated crystallization, polymorphism and crystal transformation of poly(butylene adipate) confined in electrospun immiscible blend fibers with polystyrene

Abstract

Utilizing electrospun immiscible blend fibers of poly(butylene adipate) (PBA) and polystyrene (PS) and following coating by the high glass transition temperature poly(4-tert-butylstyrene) (P4tBS), confined PBA specimens in nanometer space were effectively prepared. The effect of a physical confinement environment on the PBA crystallization behavior, polymorphism and crystal transformation was adequately investigated. For non-isothermal crystallization, the crystallization ability and melting of subsequent heating was obviously suppressed in the nanospace compared with in the bulk state. Especially, the confined PBA could show fractionated crystallization due to the existence of different sizes of nanospace and the crystallization mechanism was either 1D crystal growth after heterogeneous nucleation or homogeneous nucleation dependent on the domain size. The temperature-dependent polymorphism and crystal transformation is also affected by the nanometer confinement. The formation temperature of the pure β crystal shifts slightly toward a high temperature and the β → α crystal transformation becomes easier for confined PBA compared with bulk PBA. More interestingly, the 2D IR correlation spectra revealed that the β → α crystal transformation in the nanospace takes the melt-recrystallization path, which is quite different from the usually solid–solid transformation process.

---

Introduction

Poly(butylene adipate) (PBA) is a typical biodegradable aliphatic polyester with an interesting polymorphism phenomenon, which has attracted great attention. It has been reported to exhibit two crystal forms, α and β.^1,2 The α crystal is characterized as having monoclinic unit packing with molecular chains containing skew conformers, while the β crystal has orthorhombic packing with planar zigzag conformation chains.^3–5 Bulk PBA can form the pure α crystal when melt-crystallized above 32 °C and the pure β crystal below 27 °C. When the crystallization temperature is set between 27 and 32 °C, a mixture of α and β crystals appears.^6,7 The α crystal has been proven to be more thermodynamically stable but less kinetically advantageous than the β crystal without an external force field.⁸ However, the relative stability between the α crystal and β crystal can be reversed during stretching.⁹ The β crystal can transform to the α crystal by annealing at a relatively high temperature, which has been demonstrated to be a solid–solid transformation process.⁸ Recently, we have shown that the β → α crystal transformation mechanism would change to melt-recrystallization after randomly introducing butylene fumarate units into PBA main chains.¹⁰

Polymorphism regulation of PBA is a hot research topic, because both α and β PBA crystals can be easily obtained and stay rather stable under mild conditions compared with other polyesters. Therefore, besides temperature, the nucleating agent,^11–14 blending,^15–20 copolymerization,^10,21,22 substrate,^23–26 and tension,⁹ etc. have been employed to control the crystal structure. Among those methods, most are conducive to the formation of the α crystal and the change of β → α transformation behavior. For example, added carbon nanofibers can expand the formation temperature of the α crystal to as low as 0 °C,¹³ and the β → α crystal transformation behavior of PBA in the PBA/poly(vinyl phenol) (PVPh) blends could be significantly accelerated due to the formation of intermolecular hydrogen bonds and the appearance of imperfect β crystals.^27,28 Despite all these results, the crystallization behavior of PBA under confinement space is less studied so far,^16,29 even though confined crystallization has been proven to be of great help in intense understanding of the crystallization mechanism of semi-crystalline polymers.^30–33 Yang et al. presented that PBA displayed fractionated crystallization behavior and favoured formation of the α crystal when it was physically confined in the interlamellar regions of PBS.¹⁶ Mi et al. found that the transformation rate of the β → α process slowed down when PBA was confined in nanoporous anodic alumina oxide (AAO).³⁴ Therefore, further study on the PBA crystallization behavior in confinement space will help with understanding the mechanism of polymorphism formation and phase transition.

Electrospinning technology provides a convenient way to fabricate confinement space. Co-electrospinning of a homogeneous solution of an immiscible polymer blend can be used to prepare a core/shell structure by selecting suitable parameters, such as a poly(vinylidene fluoride) (PVDF)/polysulfone (PSF) blend,³⁵ PVDF/polyvinylpyrrolidone (PVP) blend,³⁶ poly(ethylene oxide) (PEO)/polystyrene (PBS),³⁷ and poly(ε-caprolactone) (PCL)/polylactide (PLA) blend,³⁸ etc. The component with a higher T_g or melting temperature (T_m) (hard phase) provides confinement space for the other component (soft phase) within electrospun fibers. During crystallization the soft phase exhibits various intriguing behaviors, such as homogeneous nucleation³⁹ which is rarely observed in bulk polymers, and orientation crystallization due to the interaction with the hard phase or constraints by the confinement geometry.^40,41 In this work, we studied the crystallization behavior of PBA confined in electrospun immiscible blend fibers with PS. PBA was physically confined in nanospace with the aid of a poly(4-tert-butylstyrene) (P4tBS) coating process. It was found that the confinement effect induced fractionated crystallization behavior, a slightly expanded β crystal formation temperature toward a high temperature, and forced the β → α crystal transformation process to take the melt-recrystallization path.

Experimental

Materials

PBA was synthesized from 1,4-butylene glycol and adipic acid by a two-step reaction of esterification and polycondensation in the molten state, as reported in the previous literature.¹⁰ The number-average molecular weight (M_n) and polydispersity index (PDI), determined by a Viscotek-M302 TDA multiple testing chromatography system, are 8.4 × 10³ g mol⁻¹ and 1.52, respectively. Polystyrene (PS, M_n = 1.7 × 10⁵ g mol⁻¹ and PDI = 2.06) and poly(4-tert-butylstyrene) (P4tBS, M_w ranges between 5.0 × 10⁵ and 1.0 × 10⁶ g mol⁻¹) were purchased from Sigma-Aldrich company. Chloroform (CHCl₃), N,N-dimethylformamide (DMF), hexane and alcohol purchased from Beijing chemical reagent factory are all AR grade and were used as received.

Preparation of PBA/PS fibers

The solution for electrospinning was obtained by together dissolving the polymers (PBA and PS) in a mixture of 70 vol% CHCl₃ and 30 vol% DMF at a total concentration of 18% (w/v). A transparent solution was obtained by strong agitation at 50 °C for 12 h. Then the PBA/PS blend fibers were prepared via the electrospinning process under the following conditions: a DC voltage of 20 kV, a collector-to-needle tip distance of 18 cm and an inner needle diameter of 0.6 mm. The electrospinning was maintained for about 1 hour to obtain an overlaid fiber mat for each sample. The mass fractions of PBA in the blends were set as 10 wt%, 15 wt% and 20 wt% and the corresponding fibers were briefly marked as PBA-0.1, PBA-0.15 and PBA-0.2, respectively. Oriented electrospun fiber bundles were obtained using a highly rotating receiving roller. The electrospun blend fibers were dipped into a 3 wt% P4tBS/hexane solution followed by being dried at ambient temperature for three cycles to coat the fibers with a P4tBS (T_g = 143 °C) layer in order to confine the fiber surface.³⁷

Characterization

The crystallization and melting behavior of the samples was investigated by differential scanning calorimetry (DSC, NETZSCH-204F1) equipped with an intercooler as the cooling system under argon atmosphere. The samples were weighed as about 10 mg and sealed in aluminum pans.

Wide angle X-ray diffraction (WAXD) analysis was carried out on a Bruker AXS D8 Advance powder diffractometer using Cu Kα radiation (λ_Cu = 0.154 nm) under 23 °C. The diffractograms were collected in the 2θ range interval from 10° to 30° at a scanning rate of 1° min⁻¹, and the scanning step was set as 0.01°.

Fourier transform infrared (FTIR) spectra were collected on a Bruker Hyperion FTIR spectrometer equipped with a THMS-600 hot stage (Linkam Scientific Instrument Ltd). The spectra were collected by averaging 32 scans at a 4 cm⁻¹ resolution. Two KBr plates were used as the IR glass to clamp the PBA fiber mat. The specimen was first melted at 80 °C for 10 min to ease the heat history, and then was quickly cooled to the isothermal crystallization temperature. After isothermal crystallization for a certain time, the FTIR spectra were collected. For the annealing experiment, the specimen was first melt-crystallized at 0 °C for 30 min to get pure β-form crystals. Then the sample was quickly heated to 46 °C and kept annealing with simultaneous data collection at intervals of every minute . The 2D IR synchronous and asynchronous spectra were calculated using a lab-made program kindly supplied by Wu.⁴² The measurement of orientation degree was carried out with a polarized infrared accessory.

The scanning electron microscopy (SEM) morphology of the fiber surface was observed on a Hitachi SU8010 scanning electron microscope with an accelerating voltage of 30 kV. The as-spun PBA/PS fibers were sputter-coated with thin layers of gold.

The transmission electron microscopy (TEM) study was carried out using an FEI Tecnai G² F20 instrument. The P4tBS-coated PS/PBA (as-prepared) fibers were first embedded in standard epoxy, and then the epoxy was cured at room temperature for 12 h for complete solidification. Ultra-thin films (80 nm thickness) were obtained by microtoming the epoxy blocks on a Leica EMUC6 cryosection system at −80 °C, then the films were collected in alcohol. After completely drying at 30 °C in a vacuum oven, the thin films were etched in 36% (w/v) hydrochloric acid for 5 min followed by washing with deionized water five times.¹⁴ Finally, the etched films were picked up on a 400-mesh TEM copper grid for TEM investigation.

Results and discussion

Morphology of fibers

Fig. 1 shows the SEM morphology of the as-spun PS/PBA fibers. From the low magnification SEM images (Fig. 1a–c) it is clear that the diameter distributions of the three specimens are all quite uniform and range from 1.5 to 2.5 μm. However, in the enlarged views of the fibers (Fig. 1d and e) the surface roughness increases and the gully topography becomes deeper on the surface with the increasing PBA content. It has been reported that a lower viscosity component in the blend would more easily migrate to the fiber surface during electrospinning.⁴³ Thus, the reason for the rougher surface structure of the fibers with a higher PBA content should be due to the more PBA migration to the fiber surface, since PBA has a rather lower molecular weight compared with PS. The crystallization of PBA also contributed to the rough morphology.

Fig. 1 SEM images of (a and d) PBA-0.1, (b and e) PBA-0.15, and (c and f) PBA-0.2.

Fig. 2 depicts the cross-section morphology of the P4tBS-coated fibers after etching PBA. It could be found that there was a white periphery surrounding the fibers and sporadic white hollows of 10–25 nm in diameter inside the fiber sections. So, part of the PBA component was further confirmed to distribute in the surface layer of the electrospun fibers. The layer thickness of the PBA-0.1 sample was about 15 nm, and that of the PBA-0.15 sample was slightly larger than that of PBA-0.1. However, in the PBA-0.2 sample the layer thickness increased significantly and the cross-section of the peripheral area looked like a string of bubbles with diameters ranging between dozens of nm to 100 nm. Part of PBA existed inside the fiber, which was because some PBA did not have enough time to migrate outside due to rather fast solvent evaporation. The deviation of the TEM cross-section morphology from the SEM results could be attributed to some possible introduction of external pressure and inclined direction of cut during specimen preparation.

Fig. 2 TEM images of the cross-section morphology of (a) PBA-0.1, (b) PBA-0.15, and (c) PBA-0.2.

DSC analysis

DSC was employed to study the crystallization and melting behavior of the PBA component confined in the fibers and the results are illustrated in Fig. 3. The enthalpy signals were normalized by corresponding the PBA mass fractions in different specimens to get better observation and comparison. For the PBA-0.2 specimen, the beginning crystallization temperature (T_b) was 28.1 °C which was much lower than that of bulk PBA (40.7 °C), indicating that the PBA crystallizability was depressed in confined space. Such a phenomenon has been often discovered and explained as the weakening of the heterogeneous nucleation ability in nanospace, and the polymer melt needs a large degree of supercooling to crystallize.³⁹ Another notable phenomenon was that the crystallization peak of PBA-0.2 was very asymmetric, and the peak displayed a rather broad trail in the low temperature side. Thus, the crystallization of PBA-0.2 was quite complex and possibly different from a regularly single crystallization process. The PBA-0.15 specimen showed a much lower crystallization temperature (T_c, at about 15 °C) and a smaller crystallization enthalpy, and the PBA-0.1 specimen did not show a detectable crystallization peak during cooling. The inhibition of crystallization was due to the decrease in size of the confinement space.

Fig. 3 (a) The non-isothermal crystallization thermograms and (b) subsequent heating thermograms of bulk PBA and PBA/PS electrospun fibers. Both the cooling rate and heating rate were 10 °C min⁻¹. The inset in (a) is the enlarged PBA-0.15 crystallization peak.

During subsequent heating processes, both PBA-0.1 and PBA-0.15 showed apparent cold crystallization peaks at around −48 °C, of which the temperature was close to the glass transition temperature (T_g) of PBA. Such similar cold crystallization phenomena can be found in many polymers, such as polylactide⁴⁴ and poly(ethylene terephthalate).⁴⁵ The enhanced crystallizability from the glass state compared with the melt state is ascribed to the improvement of nucleation ability. So the incomplete crystallization of PBA in the PBA/PS fibers during the cooling process is further confirmed to be mainly due to the weakening of the nucleation ability.

As the literature states,^6,7,10 bulk PBA can form the pure α crystal and pure β crystal when T_c is above 32 °C and below 27 °C, respectively. Accordingly, bulk PBA should mainly be composed of α crystals during non-isothermal crystallization at a rate of 10 °C min⁻¹ and show two exothermic peaks in subsequent heating. The two peaks arise from melting of the original α crystals and the recrystallized α crystals, respectively. But all the electrospun fibers exhibited a single melting peak, revealing that the confinement not only affected the crystallization behavior but also the melting behavior of PBA. According to the T_c span of the PBA-0.2 specimen during non-isothermal crystallization, PBA would form β crystals (this was proven by the WAXD result). PBA-0.15 and PBA-0.1 displayed the same result. However, the characteristic melting peak of the newly formed α crystals from a β → α solid–solid transformation was not observed. Thus, it could be deduced that the confinement also affected the β → α transformation. Both the melting point and normalized enthalpy decreased with a decrease in the PBA content. Especially, the PBA-0.1 specimen exhibited a much smaller enthalpy than that of the other two electrospun specimens, so that the crystallinity was reduced when the confinement space was small enough.

Fig. 4a shows the non-isothermal crystallization behavior of PBA-0.2 at rates of 2 °C min⁻¹ and 1 °C min⁻¹. It was interesting to note that the crystallization process divided into two separate crystallization peaks under a low cooling rate. This is typical for fractionated crystallization.⁴⁶ As shown in Fig. 2c, the PBA component mainly distributed in two kinds of nanospace of ∼100 nm and ∼25 nm in diameter, respectively. Therefore, the fractionated crystallization behavior should stem from the different confinement space. To investigate the nucleation mechanism of peak a and peak b in Fig. 4a, the Avrami equation was used to analyse the crystallization kinetics. The data of relative crystallinity between 3% and 20% of each peak were adopted to calculate the Avrami index (n) and crystallization rate constant (K⁻ⁿ).⁴⁷ Due to such a low cooling rate, the temperature intervals that corresponded to a relative crystallinity between 3% and 20% were 1.6 °C and 3.3 °C for peak a and peak b, respectively. Therefore, the obtained Avrami parameters would be close to the isothermal crystallization process. n and K⁻ⁿ were 2.01 and 18.36 min⁻¹ for peak a, and 1.55 and 29.72 min⁻¹ for peak b. Zhong et al. proved that poly(ethylene oxide) confined in a several hundred nanometer space of PEO/PS electrospun fibers showed an Avrami index of 2.1 and suggested that PEO underwent one-dimensional crystal growth confined in the electrospun blend fibers after heterogeneous nucleation.³⁷ The smaller of Avrami index of peak a compared with peak b confirmed the occurrence of a homogeneous nucleation mechanism for the low crystallization temperature peak.^48,49 The K⁻ⁿ value of peak b was smaller than that of peak a, and even peak b possessed a larger supercooling degree. This was also in accordance with the heterogeneous nucleation mechanism in the peak a region and homogeneous nucleation mechanism in the peak b region.

Fig. 4 Non-isothermal melt-crystallization thermograms of PBA-0.2 at cooling rates of 2 °C min⁻¹ and 1 °C min⁻¹.

After isothermal crystallization from the melt at 30 °C for different lengths of time, the subsequent heating thermograms showed that the melting enthalpy was slightly time-dependent on the time span when the time span was longer than 2 h. The normalized melting enthalpy was about 18.8 J g⁻¹, which was smaller than that in Fig. 3b. However, the normalized melting enthalpy could achieve 42.4 J g⁻¹ when PBA-0.2 was isothermally crystallized at 0 °C. This evident temperature-dependent crystallinity phenomenon is a characteristic phenomenon of fractionated crystallization (Fig. 5).⁵⁰

Fig. 5 Melting behavior of isothermally crystallized PBA-0.2 at (a) 0 °C and (b) 30 °C for various lengths of time, and the heating rate is 10 °C min⁻¹.

WAXD and FTIR investigation

Polymorphism is a widespread concern in PBA crystallization, so it is necessary to study the effect of confined space on PBA polymorphism. The X-ray diffractograms of the PBA/PS specimens after crystallization at different temperatures are presented in Fig. 6a. The convex baseline was induced by the strong scattering signal from the amorphous PS component. PBA-0.1 and PBA-0.15 could only crystallize below 25 °C and form the β crystal, while PBA-0.2 could crystallize at a high temperature. When PBA-0.2 was crystallized below 30 °C, it only exhibited the characteristic diffraction peaks corresponding to the (110) and (020) lattice planes of the β crystal. When crystallized at 31–40 °C, the PBA-0.2 specimens showed characteristic diffraction peaks of both the α and β crystals. However, bulk PBA exhibited the pure α crystal when the crystallization temperature is above 31 °C.¹⁰ It seemed that the nanospace expanded the formation ability of the PBA β crystal toward high temperature. And yet, it has been revealed that part of PBA-0.2 confined in a smaller nanospace could not crystallize when the temperature was higher than 30 °C. So the diffraction peaks of the β crystal in the PBA-0.2 specimens crystallizing at 35 °C and 40 °C might stem from the low temperature (i.e. ∼25 °C) during specimen transfer and the WAXD measurement process. As shown in Fig. 6b, when PBA-0.2 was crystallized at 35 °C for different times, the relative fraction of the α crystal increased while that of the β crystal decreased with the increasing crystallization times. The normalized melting enthalpy of PBA-0.2 was measured as 10.53 J g⁻¹ after crystallization for 12 h, while the enthalpy value greatly increased to 21.14 J g⁻¹ with further crystallization at 25 °C. So, the β crystal should not be formed during crystallization at 35 °C. To solve this confusion, in situ FTIR measurement of PBA-0.2 crystallization at 35 °C was carried out. As shown in Fig. 6c, the absorption band of the C–O stretching mode appeared at 1170 cm⁻¹ after crystallization for 12 h, indicating the formation of α crystals.⁵¹ The absorption band shifted to 1173 cm⁻¹ (overlap of 1177 and 1170 cm⁻¹) when the specimen was cooled to 25 °C, illustrating formation of the β crystals. Therefore, the confinement effect expanded the formation temperature of the pure β crystal of PBA-0.2 slightly toward a high temperature (i.e. 30 °C) compared with bulk PBA (i.e. <27 °C).

Fig. 6 (a) X-ray diffractograms of PBA-0.2 specimens melt-crystallized at various temperatures for 12 h. The PBA-0.1 and PBA-0.15 specimens for WAXD measurement were prepared by melt-quenching to 0 °C; (b) X-ray diffractograms of PBA-0.2 specimens melt-crystallized at 35 °C for different times; and (c) in situ FTIR spectra of PBA-0.2 (i) melt-crystallized at 35 °C for 12 h, and (ii) further at 25 °C. Spectra of (iii) pure β and (iv) pure α crystals are illustrated for comparison. The pure β crystal specimen was prepared by melt-crystallization at 25 °C, while the pure α crystal specimen was obtained by annealing specimen (ii) at 48 °C for 24 h.

The results of annealing-induced β → α crystal transformation are presented in Fig. 7. After annealing at 43 °C for 2 h, partial β-form PBA-0.2 transformed to α crystals and the conversion increased with the increasing annealing temperature. Once the annealing temperature reached 48 °C, the β crystals completely transformed to α crystals within 2 h, and diffraction peaks belonging to the β crystal disappeared. For comparison, β crystals of bulk PBA were annealed at 48 °C for 2 h. However, the β → α conversion was rather low. The crystal transformation could complete at an elevated temperature within 2 h, i.e. 52 °C. Therefore, the confinement effect enhances the β → α crystal transformation process.

Fig. 7 (a) X-ray diffractograms of β-form PBA-0.2 specimens after annealing at different temperatures for 2 h, and (b) X-ray diffractograms of β-form bulk PBA after annealing at different temperatures for 2 h. Both confined and bulk PBA specimens were obtained by melt-crystallization at 5 °C.

Detailed information of the β → α crystal transformation process of PBA-0.2 was further investigated during annealing at 46 °C. The FTIR absorption curves between 1185 and 1160 cm⁻¹ were used to trace the transformation (Fig. 8). The intensity of the 1177 cm⁻¹ band declined rapidly at the beginning, while that of the 1170 cm⁻¹ band increased slowly with a delay. During the later stage the intensity of the 1170 cm⁻¹ band kept an obviously increasing tendency, while that of the 1177 cm⁻¹ band was almost unchanged. Such changes were quite different from the FTIR spectral evolution of solid–solid transformation belonging to bulk β crystals during annealing.⁵¹ Thus, it was probable that the original β crystals confined in the nanospace were more likely to melt-recrystallize to α crystals but not solid–solid transform to α crystals.

Fig. 8 In situ collection of β-form PBA-0.2 FTIR spectra during annealing at 46 °C. The arrows indicate the direction of band change.

To get more detailed information about the β → α crystal transformation mechanism of PBA confined in the nanospace, 2D IR correlation analysis was employed.⁵² Fig. 9 shows the computed synchronous and asynchronous 2D correlation spectra in the spectral region of 1200–1145 cm⁻¹ generated from the time-dependent FTIR spectra. Two autopeaks were developed at (1182, 1182) cm⁻¹ and (1167, 1167) cm⁻¹ together with a pair of negative cross peaks at (1182, 1167) cm⁻¹ in the synchronous 2D spectrum. However, the two IR bands of 1182 and 1167 cm⁻¹ could not be assigned to any vibration modes of PBA. The (1177, 1177) and (1170, 1170) cm⁻¹ peaks located within the positive correlation spectral region and (1177, 1170) cm⁻¹ peak in the negative correlation spectral region indicate the opposite spectral vibrations of the α crystals and β crystals. The deviation of the (1177, 1177), (1170, 1170) and (1177, 1170) cm⁻¹ peaks from the centers of the correlation spectral regions suggested that the β → α crystal transformation is not a simple solid–solid process. In the corresponding asynchronous 2D spectrum a pair of cross peaks appeared at (1177, 1170) cm⁻¹. The presence of asynchronous cross peaks indicates the out-of-phase intensity changes of the correlated bands. The negative sign of the asynchronous cross peak at (1177, 1170) cm⁻¹ and the positive sign of the corresponding synchronous spectral intensity reveal that the destruction of β crystals proceeds prior to formation of the α-form crystals during annealing. Therefore, the β → α crystal transformation in the nanospace takes the melt-recrystallization path.

Fig. 9 The (a) synchronous and (b) asynchronous 2D IR correlation spectra of β-form PBA-0.2 in the C–O stretching mode region generated from the time-dependent FTIR spectra at 46 °C.

Chain orientation

Chain orientation is an important aspect of confinement crystallization. Fig. 10 depicts the polarized FTIR spectra of the C=O stretching mode of PBA-0.2 specimens under a different status. The absorption intensity of the carbonyl groups had a higher value when the infrared beam was polarized perpendicular to the fiber axis. This indicates that the PBA backbone was preferentially oriented along the fiber axis. The orientation degree (D) was determined by infrared dichroism as in the equation:
where I_∥ and I_⊥ represent the absorption intensities of the carbonyl stretching mode when the infrared beam is parallel to and perpendicular to the fiber axis, respectively. In this study, D < 1 means that the PBA backbone is preferentially oriented along the fiber axis, and the smaller the D value is, the higher the orientation degree is. Then, the orientation degree of the as-prepared PBA was calculated as 0.66. After melting at 90 °C for 15 min, the orientation degree slightly increased to 0.73. The nanospace confinement forced the PBA melt to maintain the oriented state of chains with slight disorientation. When the melt was recrystallized at 25 °C, the orientation degree changed to 0.65, indicating the occurrence of orientation crystallization. Unluckily, due to the large degree overlap of carbonyl absorbance, the respective orientation degrees for crystal and amorphous regions could not be obtained.

Fig. 10 Polarized FTIR spectra of the C=O stretching mode of PBA-0.2: (a) as-prepared fibers, (b) melt state and (c) melt-crystallization at 25 °C. ⊥ and ∥ represent the polarization direction of the incident infrared beam being perpendicular to and parallel to the fiber direction, respectively.

Discussion

PBA confined in nanospace shows quite different crystallization, melting and crystal transformation behavior compared with bulk PBA. The crystallization ability of PBA is inhibited in the confinement space, and a fractionated crystallization phenomenon appears in PBA-0.2 due to the apparent different space sizes. The non-isothermal signals of PBA-0.1 and PBA-0.15 even cannot be detected in the DSC thermograms due to a smaller confinement space size. And the nucleation mechanism can be attributed to homogeneous nucleation in the nanospace.

As for the melting behavior in Fig. 3b, the melting points of confined PBA during the subsequent heating process were higher than that of the original α crystals in bulk PBA, while the formation temperatures for the confined PBA crystals were rather low. This indicates that the melting points should not be directly assigned to the original PBA crystals which formed during the cooling (∼20 °C for PBA-0.2) or subsequent heating (∼−48 °C for PBA-0.1 and PBA-0.15) process in the confinement space. According to the temperature-dependent polymorphism of confined PBA, all original PBA crystals formed in the confinement space during DSC measurement should be in the β form. Together with the above demonstration that the β → α transformation adopts the melt-recrystallization path, it is reasonable to attribute the melting points of confined PBA to the melt-recrystallized α crystals from the original β crystals.

In nanometer confinement space, the expression for the melting point of the finite size crystal⁵³ would be an appropriate way to further understand the melting behavior:

where σ₁, σ₂ and σ₃ denote the specific surface free energy of a crystallite, and L₁, L₂ and L₃ are the dimensions of the crystallite. The subscripts represent the three orthogonal directions in a chain-folded lamella. T⁰_m is the equilibrium melting temperature of the crystal with infinite thickness, ρ_c is the crystal density, and ΔH⁰_m is the heat of fusion per unit mass. According to the equation, T_m decreases as the dimensions of the crystallite (L₁, L₂ and L₃) decrease. For example, it has been observed that the melting temperature of polyethylene confined in AAO templates decreased as the pore diameter of AAO diminished.⁵³ Thus, due to the same reason, T_m of confined PBA is lower than that of bulk PBA and decreases with the decreasing confinement space. Also, the stability of the β crystal under the confinement environment should be lower than that of the bulk state, leading to easier melting of the β crystal and the low shift of the β → α transformation temperature. Mi et al. investigated the crystal transformation of PBA in nanoporous AAO.³⁴ They found that the β → α crystal transformation started at the same temperature but with a slower rate when compared with bulk PBA. The different crystal transformation behavior between PBA confined in AAO and that studied here might be explained by the different confined walls. PBA chains might be wetted on the AAO walls but dewetted on the PS and P4tBS walls.

PBA prefers to form β crystals under a confinement environment compared with the bulk state in this study. Liang et al. found that PBA showed preferential formation of the β-form crystal when crystallized in the microparticles and nanoparticles covered by poly(vinyl alcohol) (PVA).⁵⁴ They proposed that it is due to PBA epitaxial crystallization and growth on the PVA surface and the interactions between PVA and PBA. Since PBA is incompatible with PS and PS is amorphous, such a probable reason cannot be taken into account in this case. The confined PBA chains suffer tension along the fiber axis during electrospinning and the tension effect⁵⁵ cannot be eliminated after melting. Fig. 10b shows that the PBA chains can maintain orientation in the melt under the confinement environment, indicating that release of tension is not complete. The tension effect would be conducive to the formation of β crystals.⁹

Conclusions

By electrospinning the PBA/PS blend solution and coating with P4tBS, a nanosize confinement environment for PBA was effectively constructed. The crystallization behavior, polymorphism and crystal transformation of confined PBA were differently compared with those of the bulk state. The DSC study showed that the crystallization ability and melting of PBA was suppressed in the nanospace. Since the PBA component was divided into different size domains with a reduction of dimension, the non-isothermal crystallization was split into two steps, known as fractionated crystallization behavior. PBA adopted 1D crystal growth after heterogeneous nucleation in ∼100 nm space and homogeneous nucleation in ∼25 nm space. Based on the WAXD and FTIR results, it was clear that the confinement environment could expand the formation temperature of the pure β crystal slightly toward a high temperature and the β → α crystal transformation becomes easier compared with bulk PBA. More interestingly, 2D IR correlation spectra revealed that the β → α crystal transformation in the nanospace takes the melt-recrystallization path, which is quite different from the solid–solid transformation process in bulk PBA.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21304108) and the Science Foundation of China University of Petroleum-Beijing (No. 2462013BJRC001). We are grateful for the kind help of 2D FTIR software from Professor Qiong Wu of Tsinghua University.

References

1. R. Minke and J. Blackwell, J. Macromol. Sci., Part B: Phys., 1979, 16, 407–417.
2. R. Minke and J. Blackwell, J. Macromol. Sci., Part B: Phys., 1980, 18, 233–255.
3. E. Pouget, A. Almontassir, M. T. Casas and J. Puiggalí, Macromolecules, 2003, 36, 698–705.
4. T. Iwata, S. Kobayashi, K. Tabata, N. Yonezawa and Y. Doi, Macromol. Biosci., 2004, 4, 296–307.
5. K. Noguchi, H. Kondo, Y. Ichikawa, K. Okuyama and J. Washiyama, Polymer, 2005, 46, 10823–10830.
6. Z. Gan, H. Abe and Y. Doi, Macromol. Chem. Phys., 2002, 203, 2369–2374.
7. E. M. Woo and M. C. Wu, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 1662–1672.
8. Z. Gan, K. Kuwabara, H. Abe, T. Iwata and Y. Doi, Biomacromolecules, 2004, 5, 371–378.
9. Y.-Y. Song, H.-M. Ye, J. Xu, K. Hou, Q. Zhou and G.-W. Lu, Polymer, 2014, 55, 3054–3061.
10. C.-X. Wang, X.-C. Zhang, Y.-Y. Song, Q. Zhou and H.-M. Ye, RSC Adv., 2016, 6, 607–616.
11. M. Weng, Y. He and Z. Qiu, Ind. Eng. Chem. Res., 2012, 51, 13862–13868.
12. Y.-R. Tang, J. Xu and B.-H. Guo, Ind. Eng. Chem. Res., 2015, 54, 1832–1841.
13. J. Yang, Y. Chen, S. Qin, J. Liu, C. Bi, R. Liang, T. Dong and X. Feng, Ind. Eng. Chem. Res., 2015, 54, 8048–8055.
14. Y.-Y. Song, H.-M. Ye, X.-Y. Meng, Q. Zhou and G.-W. Lu, RSC Adv., 2015, 5, 102384–102391.
15. J. Yang, P. Pan, L. Hua, B. Zhu, T. Dong and Y. Inoue, Macromolecules, 2010, 43, 8610–8618.
16. J. Yang, P. Pan, L. Hua, Y. Xie, T. Dong, B. Zhu, Y. Inoue and X. Feng, Polymer, 2011, 52, 3460–3468.
17. J. Yang, P. Pan, L. Hua, X. Feng, J. Yue, Y. Ge and Y. Inoue, J. Phys. Chem. B, 2012, 116, 1265–1272.
18. X. Sun, F. Pi, J. Zhang, I. Takahashi, F. Wang, S. Yan and Y. Ozaki, J. Phys. Chem. B, 2011, 115, 1950–1957.
19. L.-Y. Wang, G. Lugito, E. M. Woo and Y.-H. Wang, Polymer, 2012, 53, 3815–3826.
20. G. Lugito and E. M. Woo, Macromol. Chem. Phys., 2012, 213, 2228–2237.
21. Z. Liang, P. Pan, B. Zhu and Y. Inoue, Macromolecules, 2010, 43, 6429–6437.
22. L. Zhao and Z. Gan, Polym. Degrad. Stab., 2006, 91, 2429–2436.
23. Y. Sun, H. Li, Y. Huang, E. Chen, L. Zhao, Z. Gan and S. Yan, Macromolecules, 2005, 38, 2739–2743.
24. Y. Sun, H. Li, Y. Huang, E. Chen, Z. Gan and S. Yan, Polymer, 2006, 47, 2455–2459.
25. Q. Li, J. Zhou, L. Chai, J. Memon, Z. Ren, H. Li, X. Sun and S. Yan, Polym. Chem., 2014, 5, 4293–4303.
26. Z.-B. Ning, R. Nielsen, L.-F. Zhao, D.-H. Yu and Z.-H. Gan, Chin. J. Polym. Sci., 2014, 32, 1243–1252.
27. X. Sun, F. Pi, J. Zhang, I. Takehashi, F. Wang, S. Yan and Y. Ozaki, J. Phys. Chem. B, 2011, 115, 1950–1957.
28. X. Sun, J. Liu, I. Takahashi and S. Yan, RSC Adv., 2014, 4, 39101–39109.
29. J. Yang, P. Pan.
30. Y.-L. Loo, R. A. Register, A. J. Ryan and G. T. Dee, Macromolecules, 2001, 34, 8968–8977.
31. Y.-L. Loo, R. A. Register and A. J. Ryan, Macromolecules, 2002, 35, 2365–2374.
32. J. M. Carr, D. S. Langhe, M. T. Ponting, A. Hiltner and E. Baer, J. Mater. Res., 2012, 27, 1326–1350.
33. Y. X. Liu and E. Q. Chen, Coord. Chem. Rev., 2010, 254, 1011–1037.
34. C. Mi, J. Zhou, Z. Ren, H. Li, X. Sun and S. Yan, Polym. Chem., 2016, 7, 410–417.
35. G. Zhong, R. Sun, L. Zhang, K. Wang, Z. Li, H. Fong and L. Zhu, Polymer, 2012, 53, 4472–4480.
36. M. Wang, D. Fang, N. Wang, S. Jiang, J. Nie, Q. Yu and G. Ma, Polymer, 2014, 55, 2188–2196.
37. G. Zhong, K. Wang, L. Zhang, Z.-M. Li, H. Fong and L. Zhu, Polymer, 2011, 52, 5397–5402.
38. L. Lu, D. Wu, M. Zhang and W. Zhou, Ind. Eng. Chem. Res., 2012, 51, 3682–3691.
39. Y. L. Loo, R. A. Register and A. J. Ryan, Phys. Rev. Lett., 2000, 84, 4120.
40. J.-P. Yang, Q. Liao, J.-J. Zhou, X. Jiang, X.-H. Wang, Y. Zhang, S.-D. Jiang, S.-K. Yan and L. Li, Macromolecules, 2011, 44, 3511–3516.
41. L.-W. Zhong, X.-K. Ren, S. Yang, E.-Q. Chen, C.-X. Sun, A. Stoeks and T.-Y. Yang, Polymer, 2014, 55, 4332–4340.
42. Q. Wu, G. Tian, S. Sun, I. Noda and G.-Q. Chen, J. Appl. Polym. Sci., 2001, 82, 934–940.
43. M. T. Hunley, A. Harber, J. A. Orlicki, A. M. Rawlett and T. E. Long, Langmuir, 2008, 24, 654–657.
44. H.-M. Ye, K. Hou and Q. Zhou, Chin. J. Polym. Sci., 2016, 34, 1–12.
45. N. Torres, J. J. Robin and B. Boutevin, Eur. Polym. J., 2000, 36, 2075–2080.
46. D. S. Langhe, J. K. Keum, A. Hiltner and E. Baer, J. Polym. Sci., Part B: Polym. Phys., 2011, 49, 159–171.
47. A. T. Lorenzo, M. L. Arnal, J. Albuerne and A. J. Müller, Polym. Test., 2007, 26, 222–231.
48. M. V. Massa and K. Dalnoki-Veress, Phys. Rev. Lett., 2004, 92, 255509.
49. M. V. Massa, M. S. M. Lee and K. Dalnoki-Veress, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 3438–3443.
50. M. L. Arnal, M. E. Matos, R. A. Morales, O. O. Santana and A. J. Müller, Macromol. Chem. Phys., 1998, 199, 2275–2288.
51. C. Yan, Y. Zhang, Y. Hu, Y. Ozaki, D. Shen, Z. Gan, S. Yan and I. Takahashi, J. Phys. Chem. B, 2008, 112, 3311–3314.
52. I. Noda, Appl. Spectrosc., 2000, 44, 994–999.
53. Y. Ma, W. Hu, J. Hobbs and G. Reiter, Soft Matter, 2008, 4, 540–543.
54. Z. Liang, J. Yang, L. Hua, P. Pan, J. Huang, J. Zhang, H. Abe and Y. Inone, J. Appl. Polym. Sci., 2014, 131, 39600.
55. X. Ma, J. Liu, C. Ni, D. C. Martin, D. B. Chase and J. F. Rabolt, ACS Macro Lett., 2012, 1, 428–431.

---