Relationships Between Crystalline Structure and the Thermal Behavior of Polyethylene 2,5-furandicarboxylate): An In Situ Simultaneous SAXS-WAXS Study.
Furanoate polyesters have gained an increasing interest from both academic and industrial points of view. Among them, poly(ethylene 2,5-furandicarboxylate) (PEF) is foreseen as a promising alternative to its terephthalate analogue, that is, poly(ethylene terephthalate) (PET). As compared to PET, PEF exhibits comparable mechanical and thermal properties, for example, a slightly higher glass transition temperature  ([T.sub.g]) and better gas barrier properties . From a sustainability perspective, PEF is made from renewable resources and exhibits a reduced carbon footprint . Such interesting properties make PEF an attractive bio-based polymer for a wide range of applications. The number of published papers dealing with synthesis and physical properties of PEF steadily increases, revealing a growing interest for this material [4-7]. Only a limited number of publications study the structure and thermomechanical behavior of PEF and mostly focus on the crystallization kinetics of PEF [8-11]. A common conclusion is that PEF has a slow crystallization kinetics and exhibits a complex melting behavior . Besides it was evidenced that for crystallization temperatures below 170[degrees]C, a small endothermic peak, ascribed to the melting of secondary crystals, is observed in addition to the main melting peak. Furthermore, the main melting peak of PEF exhibits a complex shape which is assumed to originate from the occurrence of a melting-recrystallization process. Regarding structural aspects, it appears that PEF is polymorph and the determination of its crystalline structures remains a scarcely addressed topic. The crystal structure of PEF was first assessed in a study by Kazaryan et al. in 1968, who proposed a triclinic unit cell comprising two repeating units . Recently, another has been proposed by Mao et al. , who suggested a monoclinic unit cell with a repeating unit consisting of two monomers and a near planar conformation of the macromolecules into the crystal lattice. However, worth noting is that in both cases, the studies were performed on X-ray fiber. Recently, a structural description of both melt-crystallized and solvent-crystallized PEF was reported by Maini et al. . The authors deduced from their results that crystalline structures of both melt-crystallized and solvent-crystallized PEF differ from the crystalline structure induced upon stretching. Other studies addressed the crystal structure of PEF by means of wide-angle X-ray scattering (WAXS), but without going to the determination of the crystal lattice. In a former work we evidenced, using WAXS, that although it is the crystalline form called [alpha] which is induced during isothermal crystallization above 170[degrees]C, it is a slightly different form which is formed at lower crystallization temperatures . It was assumed that this crystalline form, called [alpha]', consists of a defective [alpha] form. In another work, Tsanaktsis et al. observed two different crystalline structures of PEF corresponding to bulk and solvent induced crystallization . They ascribed the crystalline form generated by bulk crystallization to the [alpha]-form and the crystalline form obtained after solvent crystallization to the [beta]-form. These two crystalline forms exhibit the same melting temperature. The authors also confirmed the existence of the defective [alpha]' form when crystallization occurs below 180[degrees]C. Finally, a recent study of Araujo et al. highlights the fact that PEF chains adopt zigzag conformation into both the [alpha] and [beta] crystalline forms  confirming the structural description proposed by Maini et al. These studies clearly evidence the complex polymorphism of PEF, but the relationships between the thermal behavior and the structure of PEF remains poorly addressed. To the best of our knowledge, the structural evolution of PEF during thermal treatments has been assessed mainly starting from DSC and ex situ WAXS experiments. In addition, there are no data regarding the evolution of the crystalline morphology at the crystalline lamella scale which support the conclusion that the multiple melting peak is due to a melting-recrystallization process. As a result, the complex thermal behavior observed during melting is still unclear.
The phenomenon of multiple-melting is encountered in other polymers, and a review of the literature shows that its origin is largely debated:
* multiple melting corresponds to a melting-recrystallization phenomenon. More precisely, the thinner and less perfect crystals melt at low temperatures and immediately crystallize into more perfect and thicker ones before melting again [16, 17].
* the presence of two or more melting peaks corresponds to the melting of different populations of crystals, each population having its main particularities [18, 19]. For example, in the case of isotactic polypropylene (iPP), it was proposed that the first melting peak observed corresponds to the melting of transverse lamellae, whereas the second melting peak corresponds to the melting of radial lamellae .
* multiple melting results from an annealing process. This case is similar to the previous one in the sense where it involves two populations of crystals: the primary crystals formed during the main crystallization and the secondary crystals formed during the annealing. In this case, the melting region is generally characterized by a large endotherm characteristic of the melting of the primary crystals and by a small endotherm characteristic of the melting of the secondary crystals, generally located at 5[degrees]C-10[degrees]C above the annealing temperature.
For PEF, three distinct melting peaks are observed in the case of samples crystallized at temperatures below 170[degrees]C. Based on DSC results, we previously proposed that the small endotherm at low temperatures corresponds to the melting of secondary crystals, whereas the two others are originating from a melting--recrystallization process [10, 21].
In the same manner as for the multiple melting phenomena, the structural representation of these secondary crystals as well as their location is still debated and contradictory results are encountered in the literature. Several models describing the nature and the location of the secondary crystals have been proposed. For example, by means of real time SAXS measurements, Hsiao et al. observed an increase of the long period ([L.sub.p]) during melting of poly(aryl ether ether ketone) (PEEK) and a decrease of [L.sub.p] during the initial stages of crystallization . These results were explained by the so-called lamellar insertion model where thin lamellae (ascribed to secondary crystals) are inserted between two layers of thicker lamellae (ascribed to the primary crystals). In a later work, dealing with the structural evolution of PEEK during crystallization and melting, a similar model where the secondary lamellae are no more inserted between the primary ones was proposed . According to this model, both stacks of primary lamellae and stacks of secondary lamellae coexist, and these stacks of secondary lamellas are located in some kind of amorphous pockets and grow during secondary crystallization, that is, at long crystallization times. A common feature of these two models is that the thinner lamellae are located in the same region as the primary lamellae and coexist in the spherulites. A different point of view is proposed by Marand and Prasad who suggest that the secondary crystals are rather formed at the edges of the spherulites . The models cited above explain the presence of a low amplitude endotherm at low temperatures in addition to the main melting peak. However, it seems that the description of the associated crystalline morphology is not consistent and depends on both the nature of the polymer studied and on the crystallization conditions.
In this study, the structural evolution of PEF during melting has been followed in situ by means of simultaneous SAXS and WAXS experiments to (i) follow the structural evolution at both atomic and nanometer scales and (ii) determine if the physical phenomena observed at the atomic scale occur simultaneously with the ones observed at the nanometer scale. The data obtained have been compared to the DSC results obtained in the same conditions in a former study  so as to establish the relationships between the thermal behavior of PEF and its crystalline structural evolution. Particularly, two main questions are addressed in this study. On the one hand, this work aims to determine the origin of the multiple melting behavior observed upon heating of semicrystalline PEF considering the different structural models previously discussed. Concomitantly, the eventuality for the [alpha]' form to undergo a phase transformation upon heating, that is, [alpha]' = > [alpha], will be addressed. Finally, this article aims at gaining new insights in the nature and location of secondary crystals.
The PEF resin was supplied by Corbion Purac BV. It consists in the same grade as the one used in our previous study dealing with crystallization of PEF by means of DSC . The average molecular weights were measured by GPC and are equal to [M.sub.n] = 15,000 g/mol and [M.sub.w] = 28,000 g/mol, respectively.
Experiments were carried out on compression molded PEF films. One-millimeter-thick sheets were obtained by compression molding at 250[degrees]C for 5 min under a pressure of 50 bars followed by a rapid cooling. Before compression molding, PEF pellets were dried under vacuum at 80[degrees]C for 12 h to remove any residual moisture. DSC and WAXS analyses of the samples revealed that they were amorphous.
The study of structural evolution during crystallization was carried out on PEF samples isothermally crystallized from the melt during. Regarding the protocol used, PEF samples are held 2 min at 250[degrees]C, then cooled down to the crystallization temperature ([T.sub.c]) at 50[degrees]C/min and finally held during 45 min at [T.sub.c] = 140[degrees]C and [T.sub.c] = 190[degrees]C, respectively. For sake of clarity, these samples will be denoted PEF-140 and PEF-190, respectively.
Thermal properties were evaluated by means of Differential Scanning Calorimetry (DSC) experiments. A DSC Q20 (TA Instruments) apparatus was used. Temperature and heat flow were calibrated with a high purity indium sample using standard procedures. Experiments were carried out under nitrogen flow on about 10 mg samples placed in aluminum pans. Temperature scans were performed between 25[degrees]C and 250[degrees]C at heating and cooling rates of 10[degrees]C/min.
The structural evolution upon stretching was followed in situ by means of simultaneous WAXS and SAXS using the synchrotron radiation on the ID02 beamline at European Synchrotron Radiation Facility (Grenoble, France). Experiments were achieved using an energy of 12.4 keV (i.e., [lambda] = 0.995 [Angstrom]). Through-view 2D-patterns were recorded using CCD Rayonix LX-170HS and Rayonix MX-170HS cameras for WAXS and SAXS, respectively. Standard corrections, including dark subtraction, sample holder intensity removing, and intensity normalization, were applied to the patterns before their treatment. The intensity profiles were obtained using the SAXS utilities package developed by M. Sztucki. Particularly, the intensity profiles were computed by azimuthal integration of the 2D corrected SAXS and WAXS patterns.
The measured SAXS data were analyzed by the linear correlation function [gamma](r) which consists in transforming the scattered intensity I(q) thanks to Eq. 1:
[gamma](r) = [[integral].sup.[infinity].sub.0] I(q) cos(qr) dq/Q (1)
where q is the scattering vector (q = 4[pi] sin([theta])/[lambda]), and Q is the invariant defined as
Q = [[integral].sup.[infinity].sub.0] q2I(q) dq (2)
From [gamma](r), the basic parameters of the lamellar stacks, to know the crystalline lamellae thickness ([l.sub.c]) and the amorphous layer thickness ([l.sub.a]), have been calculated. In this work, the calculation of [gamma](r) and the determination of the crystalline lamella thickness ([l.sub.c]), the amorphous layer thickness ([l.sub.a]), and the invariant Q were carried out using the SAXSDAT software . Modeling of the SAXS data has been completed using the Irena Package . In this model, the semi-crystalline morphology of the material is described as a regular stacking of discs (crystalline lamellas of thickness [l.sub.c]) alternating with amorphous layers (thickness [l.sub.a]).
RESULTS AND DISCUSSION
Structural Evolution during Crystallization
To get a precise characterization of the crystalline morphology of the materials, the structural evolution has been followed in situ by mean of simultaneous SAXS-WAXS experiments during isothermal crystallization from the melt at both [T.sub.c] = 140[degrees]C and [T.sub.c] = 190[degrees]C. The data obtained for [T.sub.c] = 140[degrees]C, that is, where secondary crystals are formed , are depicted in Fig. 1.
The evolution of crystallinity, computed from the WAXS results, indicates that three steps can be distinguished during the crystallization of PEF at this temperature (Fig. 1a). Although the crystallinity index remains roughly zero during the first 9-10 min, a slight increase of the scattered intensity around q [approximately equal to] 0.5 [nm.sup.-1] was observed by SAXS at the end of the first stage (Fig. 1c). This increase is ascribed to the formation of heterogeneities having a nanometer size into the materials which are assumed to be some kinds of crystalline entities, that is, to a local pre-ordering the material. In the meantime, some low intensities diffraction peaks at q = 12.6 [nm.sup.-1] and q = 18.3 [nm.sup.-1] appear on the WAXS profiles (Fig. 1b). However, no peak is observed on the SAXS profiles during this period indicating that the crystalline entities formed are not in interaction or, in other words, are dispersed and isolated into the amorphous matrix (formation of isolated crystalline lamellae).
During the second stage, the intensities of the diffraction peaks increase with time indicating an increase in crystallinity (Fig. 1d). Peaks position indicate that the a form is formed. In the meantime, the SAXS profiles show an increase of the scattered intensity and the appearance of a broad peak centered around q = 0.5 [nm.sup.-1] (Fig. 1e). This peak is attributed to a long period indicating the presence of a regular stacking of crystalline lamella. In addition, we showed in a former study that in the case of PEF, secondary crystallization occurs in this timelap as revealed by the presence of a small endotherm at low temperatures during crystallization (see Fig. 13 ref 10). However, according to the data depicted here, no evidence of the presence of secondary crystals was observed on the WAXS or SAXS results. This probably arises from the very low amount of secondary crystals formed.
Finally, during the last step, crystallinity slightly increases with time without significant changes on the WAXS and SAXS profiles (Fig. 1f and g). However, former DSC results  indicate that during this stage the endotherm associated to the secondary crystals shift toward higher temperatures indicating a growth and/or an increase of the perfection degrees of these secondary crystals. A similar behavior has been observed for [T.sub.c] = 190[degrees]C even if, for sake of clarity, results are not presented here.
The results of the structural characterization of PEF-140 and PEF-190 samples obtained by means of WAXS and SAXS after complete crystallization are depicted in Fig. 2.
The WAXS analyses confirm that the crystalline form induced at [T.sub.c] = 140[degrees]C differs from the one obtained at [T.sub.c] = 190[degrees]C. More specifically, the peak around q = 13.5 [nm.sup.-1] (denoted p3) is not observed for PEF-140. In addition, SAXS profiles show (i) that the value of [L.sub.p] increases from 9.5 to 11 nm with the increase of [T.sub.c] and (ii) that the structure involved at [T.sub.c] = 190[degrees]C is more perfect and ordered than the one formed at [T.sub.c] = 140[degrees]C as revealed by the higher intensity and lower width of the peak attributed to [L.sub.p]. In a former work , we showed by means of DSC analyses that secondary crystals are formed during crystallization at [T.sub.c] = 140[degrees]C at the same time as the primary crystals. The SAXS and WAXS results obtained for this crystallization temperature does not allow to distinguish or to separate the respective contribution of these two types of crystals. Particularly, SAXS results did not give any evidence of the presence of two crystals populations such as two long periods. This probably arises from the fact that secondary crystals are present in a small amount and consequently have a low scattering.
To go further in the understanding of the differences observed in the WAXS intensity profiles, the analysis of the results in terms of crystallography was carried out. Recently, Mao et al. proposed a description of the crystalline structure for PEF . However, they determined it starting from an annealed stretched sample. Consequently, it is not obvious that the crystalline structure they determined corresponds to the one formed during thermal crystallization. For example, in the case of polylactide, a [beta]-form can be induced upon stretching, whereas it is a [alpha]'- or [alpha]-form that is induced during thermally induced crystallization . In the case of polypropylene, it has been shown that shearing or stretching favors the formation of the [beta] phase, whereas it is the a one which is thermally induced . Similarly, the [beta] phase of polyvinylidene fluoride is formed upon stretching at the expense of the [alpha]-form which is encountered during thermal crystallization .
To assess the crystallographic planes associated to the diffraction peaks observed, samples initially under the [alpha] and [alpha]' forms, respectively, have been stretched at [T.sub.d] = 80[degrees]C so as to orient the crystals toward the draw direction. The WAXS patterns before and after stretching are depicted in Fig. 3. The structural evolution upon stretching has been followed in situ to ensure that no crystal phase transformations occurs upon stretching and that stretching only involves an orientation of the crystalline entities.
It clearly appears that stretching involves a strong crystalline orientation into the material. The diffraction rings observed on the WAXS pattern of the unstretched material turn into oriented diffraction arcs upon drawing. Moreover, the diffraction arcs at low q are positioned in the meridional regions, whereas the diffraction rings at higher q are rather positioned in the equatorial regions of the patterns. Consequently, assuming that the macromolecules are aligned toward the draw direction, it is possible from these results to determine which diffractions peaks correspond to a crystallographic plane roughly perpendicular or roughly parallel to the chain axis. Results obtained are computed in Table 1.
The results indicate that p3, p5, and p6 correspond to atomic planes containing the macromolecules, that is, (hk0) planes, whereas p1, p2, and p4 correspond to planes roughly perpendicular to the macromolecules axes. Worth noting is that p3, which is not observed for [alpha]', corresponds to a (hk0) plane. This may indicate that [alpha]' suffers from a lower inter-chain organization as compared to [alpha]. However, a more in-depth crystallographic analysis is needed to go further. Moreover, it appears that the position of the peaks determined here does not exactly fit with the ones reported by Mao et al. . This tends to show that the crystalline form induced into PEF fibers studied by Mao et al. differs from the crystalline form obtained during thermally induced crystallization.
Thermal Behavior and Structural Evolution of PEF-140 and PEF-190 upon Heating
The thermal behavior of PEF-140 and PEF-190 was measured by means of DSC and is depicted in Fig. 4.
Both samples exhibit a [T.sub.g] around 80[degrees]C as expected for PEF. For both samples, the main melting endotherm is observed in the 180[degrees]C-225[degrees]C temperature range and can be decomposed into two peaks (B and C). In addition, PEF-140 also exhibits a low amplitude melting peak (peak A) around 150[degrees]C. In a former study, on the basis of DSC results , the peak A has been ascribed to the melting of secondary crystals, whereas peaks B and C were associated with the occurrence of a melting-recrystallization process.
To definitely determine the structural origin of the multiple melting peaks, simultaneous in situ SAXS-WAXS experiments have been carried out. The structural evolution upon heating at 10[degrees]C/min of PEF-140 and PEF-190 was examined. The possibility of a phase transformation from [alpha]' to [alpha] in the case of PEF-140 has been carefully assessed. Even if simultaneously recorded, WAXS and SAXS results will be presented independently to distinguish the structural evolution at the molecular scale (WAXS results) from the morphological evolution at the nanometer scale (SAXS results).
Evolution at the Crystal Lattice Scale
Figure 5 depicts the data computed from the WAXS patterns recorded in situ during heating at 10[degrees]C/min for both PEF-140 and PEF-190.
The first main result obtained from these data is that there is no phase transformation occurring upon heating for PEF-140. Intensity profiles presented in Fig. 5a give no signs of a crystalcrystal transformation from the [alpha]' to the [alpha] form upon heating. Regarding the evolution of crystallinity, a good agreement is found between the thermal behavior obtained from DSC and the relative ciystallinity calculated from the intensity profiles. Particularly, the low temperature endotherm observed around 145[degrees]C for PEF-140 correspond to a slight decrease of crystallinity as calculated by WAXS (Fig. 5b), confirming the attribution of this event to the melting of secondary crystals.
Finally, a careful analysis of the intensity profiles revealed a shift in peak position of certain diffraction peaks upon heating. The evolution of the relative inter-planes distance is depicted in Fig. 5c and shows that although the positions of pi, p2, and p4 are roughly constant with the temperature, there is a shift to lower distances of p3, p5, and p6 with the increase of the temperature, for temperatures above the glass transition. As previously discussed, these peaks correspond to diffraction planes containing the macromolecules, that is, to diffraction planes roughly parallel to the chain axis. Two hypotheses may explain this behavior. On the one hand, in their representation, Mao et al. proposed that the PEF chains are slightly bent toward the [??] axis . Thus a modification of the bending angle with the temperature may explain the decrease of the distance between macromolecules. On the other hand, a modification of this bending angle would also affect the distance between planes perpendicular to the [??] axis, what is not observed here. Consequently, this small variation of the inter-chains distance may be rather ascribed to thermal dilatation effects.
Evolution at the Crystalline Lamella Scale
The morphological changes of the crystalline structure during heating have been also measured in situ by means of SAXS and compared to the thermal behavior determined by DSC. The scattering invariant Q, the long period [L.sub.p], and the thicknesses of the crystalline and amorphous layers have been extracted from the SAXS intensity profiles using the method described in the experimental part. Particularly, the amorphous and crystalline layers thicknesses where calculated from [gamma](r). The calculation leads to two values [t.sub.1] and [t.sub.2] which cannot, a priori, be ascribed to the crystalline or amorphous layer thickness. The assignment of [t.sub.1] and [t.sub.2] to the amorphous layer or crystalline lamellae thicknesses has been largely debated in the literature, and it remains a key point for a good interpretation of the data [30, 31], Two points have been considered in this study to do these attributions. First, it has been considered that the increase of the crystallization temperature involves an increase of the crystalline lamellae thickness due to the increase of the crystal growth rate with temperature. Second, the two possibilities (i.e., [t.sub.1] = amorphous thickness/[t.sub.2] = crystal thickness and [t.sub.1] = crystal thickness/ [t.sub.2] = amorphous thickness) have been modeled and compared to the experimental data. The result of this modeling is presented in Fig. 6.
It appears that the attribution of the smaller size ([t.sub.1]) to the thickness of the crystalline lamellae gives results similar to the experimental ones, whereas the inverse case did not fit with the experimental data. Based on these two points, [t.sub.1] has been attributed to the crystalline lamellae thickness ([l.sub.c]) and [t.sub.2] to the thickness of the amorphous layer ([l.sub.a]).
Results are presented in Fig. 7 for both PEF-140 and PEF-190.
Regarding the evolution of the invariant Q, Fig. 7a shows that for PEF-190, Q is constant up to [T.sub.g] and then steadily increases with the increase of the temperature until the beginning of melting. Then, Q sharply decreases during melting. Q is defined as:
Q = 1/V [[integral].sup.[infinity].sub.0] [q.sup.2]I(q)dq = 2[pi]2q = 2[pi]2[[empty set].sub.1][[empty set].sub.1]([[DELTA].sub.[rho]])2 (eq3)
and is directly related to (i) the difference of electron density between the particle (i.e., the crystalline lamellae) and the matrix (i.e., the amorphous phase) and (ii) the volumetric fractions of amorphous and crystalline phases ([[phi].sub.1] and [[phi].sub.2]). Below [T.sub.g], Q is constant due to the fact that all parameters remain constant. Then, for both PEF-140 and PEF-190, Q increases with temperature. This increase is ascribed to an increase of [DELTA][rho] caused by a thermal dilatation of the amorphous phase and changes in the crystal density as revealed by the displacements of the peaks highlighted by WAXS. In addition, the subsequent decrease of Q during melting is however ascribed to the decrease of crystallinity.
The same trend is observed for PEF-140 except that a slight discontinuity of Q (marked by an arrow) in the temperature range where melting of secondary crystals is observed. However, it's not possible to definitely determine if this discontinuity is due to the decrease of crystallinity linked with the melting of the secondary crystals or to changes in the difference of the electron density between the amorphous phase and the crystalline phase.
The results of Fig. 7b show for both PEF-140 and PEF-190 a similar trend for [L.sub.p] and [l.sub.a], which are roughly constant upon heating up to the beginning of melting and steadily increase until the end of melting. The crystalline lamellae thickness [l.sub.c] is almost constant until the beginning of melting, and it increases in the temperature range 180[degrees]C-200[degrees]C before decreasing at upper temperatures, in particular for PEF-140.
The results previously discussed bring new elements allowing to answer the two initial questions regarding (i) the secondary crystallization aspects and (ii) the origin of the multiple melting observed. At first, for those two points, it clearly appears that no crystalline structure transformation upon heating is involved. Regarding the low amplitude melting endotherm observed in the case of samples crystallized at low temperatures, its attribution to the melting of secondary crystals appears to be the most reasonable explanation. With regard to the nature and the localization of these secondary crystals, some elements have been highlighted. As shown in a previous work , the secondary crystallization process in PEF occurs at the same time that primary crystallization and not upon further annealing. Consequently, these secondary crystals cannot be located in the outer regions of the spherulites but are rather formed into them.
Regarding the nature of these crystals, two main models are proposed as discussed in the Introduction. On the one hand, the secondary crystals may consist in thin lamellar crystals inserted between the primary lamellae (lamellar insertion model), and on the other hand, they could consist in stacks of thin lamellae located in regions between the stacks of primary crystalline lamellae. The SAXS results showed no evolution of the value of [L.sub.p] or [l.sub.a] during the melting of secondary crystals. Consequently, the lamellar insertion model cannot be applicable in the case of PEF. This model would necessarily imply changes in [L.sub.p] and [l.sub.a] during the melting of secondary crystals. The second hypothesis considering that stacks of thin lamellae are present into the material does not seem to apply either. In this case, the SAXS scattering would be, not monomodal as what is observed here, but rather bimodal similarly with results reported in the case of PEEK. The fact that no sign of [L.sub.p] originating from the secondary crystals is observed tends to show that these crystals are not self-organized but are rather isolated and randomly distributed into the material. The fact that [L.sub.p], [l.sub.a], and [l.sub.c] as well as the shape of the SAXS profile do not significantly change upon the melting of the secondary crystals is a sign that these crystals do not consist in thin lamellae. The only small change observed during the melting of the secondary crystals is a discontinuity of the continuous increase of the Q invariant. This discontinuity may be ascribed to the decrease of crystallinity and however to a change in the electron density of the amorphous phase. Starting from these results we propose a model, depicted in Fig. 8, for the description of the secondary crystals in the case of PEF crystallized from the melt.
Within this model, secondary crystals are rather some kind of small crystallites, randomly placed between the crystal lamellae issued from the primary crystallization. The nature of these crystals, that is, chain-folded crystals or bundle-like crystals cannot be determined with certainty. However, the restricted chain mobility generally encountered in the amorphous layers constrained by the crystalline lamellae may favor the hypothesis of bundle-like crystals. According to this representation, melting of secondary crystals did not impact both crystal lamellae thickness and amorphous layer thickness. Only the average electron density of the amorphous phase is modified which may explain the discontinuity observed for the scattering invariant Q. Finally, regarding the multiple melting observed for all the samples whatever the crystallization temperature, the results also support the hypothesis of a melting-recrystallization process. Although the crystallinity decreases during melting, an increase of the crystalline lamellae thickness [l.sub.c] was also observed during the offset of melting.
The structural evolution during isothermal crystallization and subsequent melting of PEF crystallized at different temperature has been followed in situ by means of simultaneous SAXS and WAXS experiments. Results show that no crystalline structure transition from the [alpha]' to the [alpha] form occurs upon heating.
Regarding the structural changes related to the complex melting behavior of PEF observed by DSC, it appears that the small melting peak occurring at low temperatures can be definitely attributed to the melting of the secondary crystals induced during the crystallization at low temperatures. Also, the double melting peak observed at higher temperatures is ascribed to the occurrence of a melting-recrystallization-melting process, as encountered in a large number of other polymers.
In addition, this work brings precious results regarding structural features of PEF. At first, the analysis of an oriented initially semi-crystalline sample shows that the crystalline structure induced during thermal crystallization is different from that obtained for fibers and described by Mao et al. Furthermore, the results indicate that secondary crystals are located in the same regions that the primary crystals and that they cannot be described by the same models than the ones proposed in the literature. According to the model proposed, it is assumed that these secondary crystals consist in independent entities located between the primary crystalline lamellae and that they are formed of extended macromolecule bundles.
The authors are indebted to the ESRF Synchrotron Facility (Grenoble, France) for beam time allocation on the ID02 beamline. The authors also gratefully acknowledge Dr. Peter Boesecke for his helpful support during the experiments as well as Dr Stanislaw Rabiej for his help for the calculation of the correlation functions. The project ARCHI-CM, Chevreul Institute (FR 2638), Ministere de l'Enseignement Superieur et de la Recherche, Region Nord-Pas de Calais, and European Regional Development Fund (FEDER) are acknowledged for funding the SAXS equipment.
[1.] G.Z. Papageorgiou, V. Tsanaktsis, and D.N. Bikiaris, Phys. Chem. Chem. Phys., 16(17), 7946 (2014). https://doi.org/10.1039/c4cp00518j.
[2.] S.K. Burgess, J.E. Leisen, B.E. Kraftschik, C.R. Mubarak, R. M. Kriegel, and W.J. Koros, Macromolecules, 47(4), 1383 (2014). https://doi.org/10.1021/ma5000199.
[3.] A.J.J.E. Eerhart, A.P.C. Faaij, and M.K. Patel, Energy Environ. Sei., 5(4), 6407 (2012). https://doi.org/10.1039/c2ee02480b.
[4.] J. Ma, Y. Pang, M. Wang, J. Xu, H. Ma, and X. Nie, J. Mater. Chem., 22(8), 3457 (2012). https://doi.org/10.1039/c2jml5457a.
[5.] A. Gandini, A.J.D. Silvestre, C.P. Neto, A.F. Sousa, and M. Gomes, J. Polym. Sei., Part A: Polym. Chem., 47(1), 295 (2009). https://doi.org/10.1002/pola.23130.
[6.] L.T. Brandao, A., B. F. Oechsler, F. W. Gomes, F. G. Souza, J. Carlos Pinto. Polym. Eng. Sei., 2018, 58 (5): 729-741. https://doi.org/10.1002/pen.24605.
[7.] P. Gopalakrishnan, S. Narayan-Sarathy, T. Ghosh, K. Mahajan, and M.N. Belgacem, J. Polym. Res., 21(1), 1 (2014). https://doi.org/10.1007/s10965-013-0340-0.
[8.] A. Codou, N. Guigo, J. Van Berkel, E. De Jong, and N. Sbirrazzuoli, Macromol. Chem. Phys., 215(21), 2065 (2014). https://doi.org/10.1002/macp.201400316.
[9.] L. Martirio, N. Guigo, J.G. Van Berkel, J.J. Kolstad, and N. Sbirrazzuoli, Macromol. Mater. Eng., 301(5), 586 (2016). https://doi.org/10.1002/mame.201500418.
[10.] G. Stoclet, G. Gobius Du Sart, B. Yeniad, S. De Vos, and J. M. Lefebvre, Polymer, 72, 165 (2015). https://doi.org/10.1016/jpolymer.2015.07.014.
[11.] V. Tsanaktsis, D.G. Papageorgiou, S. Exarhopoulos, D. N. Bikiaris, and G.Z. Papageorgiou, Crystal Growth Des., 15 (11), 5505 (2015). https://doi.org/10.1021/acs.cgd.5b01136.
[12.] L.G. Kazaryan and F.M. Medvedeva, Vysokomolekulyarnye Soedineniya Seriya B: Kratkie Soobshcheniya, 305 (1968).
[13.] Y. Mao, R.M. Kriegel, and D.G. Bucknall, Polymer, 102, 308 (2016). https://doi.org/10.1016/j.polymer.2016.08.052.
[14.] L. Maini, M. Gigli, M. Gazzano, N. Lotti, D.N. Bikiaris, and G. Z. Papageorgiou, Polymer, 10(3), 1 (2018). https://doi.org/10.3390/olym10030296.
[15.] C.F. Araujo, M.M. Nolasco, P.J.A. Ribeiro-Claro, S. Rudic, A.J.D. Silvestre, P.D. Vaz, and A.F. Sousa, Macromolecules, 51(9), 3515 (2018). https://doi.org/10.1021/acs.macromol.8b00192.
[16.] J. Li, Z. Fang, Y. Zhu, L. Tong, A. Gu, and F. Liu, J. Appl. Polym. Sei., 105(6), 3531 (2007). https://doi.org/10.1002/app.24606.
[17.] P. Supaphol, J. Appl. Polym. Sei., 82(5), 1083 (2001). https://doi.org/10.1002/app.1943.
[18.] F.-C. Chiu, C.-G Peng, Q. Fu. Polym. Eng. Sei., 2000, 40 (11): 2397-2406. https://doi.org/10.1002/pen. 11371.
[19.] T. Sasaki, H. Sunago, and T. Hoshikawa, Polym. Eng. Sei., 43 (3), 629 (2003). https://doi.org/10.1002/pen.10051.
[20.] R.G. Alamo, G.M. Brown, L. Mandelkern, A. Lehtinen, and R. Paukkeri, Polymer, 40(14), 3933 (1999). https://doi.org/10.1016/S0032-3861(98)00613-2.
[21.] H. Ghasemi, P.J. Carreau, and M.R. Kamal, Polym. Eng. Sei., 52 (2), 372 (2012). https://doi.org/10.1002/pen.22092.
[22.] B.S. Hsiao, K.H. Gardner, D.Q. Wu, and B. Chu, Polymer, 34(19), 3996 (1993). https://doi.org/10.1016/0032-3861(93)90659-X.
[23.] R. Verrna, H. Marand, and B. Hsiao, Macromolecules, 29(24), 7767 (1996).
[24.] H. Marand and A. Prasad, Macromolecules, 25(6), 1731 (1992). https://doi.org/10.1021/ma00032a017.
[25.] S. Rabiej and M. Rabiej, Polymer, 56(9), 662 (2011).
[26.] J. Ilavsky and PR. Jemian, J. Appl. Crystallograp., 2011, 42(2), 347 (2009). https://doi.org/10.1107/S0021889809002222.
[27.] W. Hoogsteen, A.R. Postema, A.J. Pennings, G.T. Brinke, and P. Zugenmaier, Macromolecules, 23(2), 634 (1990). https://doi.org/10.1021/ma00204a041.
[28.] Y.-K. An, S.-D Jiang, S.-K Yan, J.-R Sun, X.-S Chen. Chin. J. Polym. Sei., 2011, 29 (4): 513-519. https://doi.org/10.1007/s10118-010-1028-0.
[29.] J. Defebvin, S. Barrau, G. Stoclet, C. Rochas, and J.M. Lefebvre, Polymer, 84, 148 (2016). https://doi.org/10.1016/j.polymer.2015.12.041.
[30.] R.J. Rule, D.H. MacKerron, A. Mahendrasingam, C. Martin, and T.M.W. Nye, Macromolecules, 28(25), 8517 (1995). https://doi.org/10.1021/ma00129a008.
[31.] Z.-G. Wang, B.S. Hsiao, B.X. Fu, L. Liu, F. Yeh, B.B. Sauer, H. Chang, J.M. Schultz. Polymer, 2000, 41 (5): 1791-1797. https://doi.org/10.1016/S0032-3861 (99)00327-4.
Gregory Stoclet (iD), (1) Andrea Arias, (2) Bahar Yeniad, (2) Sicco De Vos (2)
(1) Univ. Lille, CNRS, INRA, ENSCL, UMR 8207-UM ET-Unite Materiaux et Transformations, F-59000 Lille, France
(2) Corbion Purac BV, Arkelsedijk 46, 4206 AC Gorinchem, The Netherlands
Correspondence to: G. Stoclet; e-mail: email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
Caption: FIG. 1. (a) Evolution of crystallinity as a function of time and (b-g) evolution of WAXS and SAXS intensity profiles during melt-crystallization of PEF at [T.sub.c] = 140[degrees]C. [Color figure can be viewed at wileyonlinelibrary.com!
Caption: FIG. 2. (a) WAXS and (b) SAXS intensity profiles obtained for PEF-140 and PEF-190 samples. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 3. WAXS patterns of PEF samples initially under (a) [alpha] form and (b) [alpha]' form before (left pictures) and after stretching (right pictures) (the draw axis is horizontal). [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 4. DSC thermograms of PEF-140 and PEF-190. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 5. (a) Evolution of the intensity profiles, (b) evolution of crystallinity, and (c) relative inter-planes distance as a function of the temperature for PEF-140 (left) and PEF-190 (right) heated at 10[degrees]C/min (pX represents the peakX discussed in Table 1). [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 6. Experimental and modeled SAXS intensities for (a) PEF-140 and (b) PEF-190 (details of modeling are given in the experimental part). [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 7. (a) Evolution of the scattering invariant Q and (b) evolution of Lp, th and t2 for PEF-140 (left column) and PEF-190 (right) column during heating at 10[degrees]C/min. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 8. Schematic representation of the secondary crystals in the case of PEF. [Color figure can be viewed at wileyonlinelibrary.com]
TABLE 1. Position of the diffraction and attribution of these peaks for the [alpha] and [alpha]' form in PEF. WAXS position ([nm.sup.-1]) Orientation toward the [alpha] [alpha]' draw axis Plane type p1 11.4 11.37 90[degrees] Roughly perpendicular to the chain axis p2 12.6 12.6 90[degrees] Roughly perpendicular to the chain axis P3 13.6 -- 0[degrees] Roughly parallel to the chain axis p4 14.5 14.5 90[degrees] Roughly perpendicular to the chain axis p5 16.4 16.3 0[degrees] Roughly parallel to the chain axis p6 18.7 18.6 0[degrees] Roughly parallel to the chain axis
|Printer friendly Cite/link Email Feedback|
|Author:||Stoclet, Gregory; Arias, Andrea; Yeniad, Bahar; De Vos, Sicco|
|Publication:||Polymer Engineering and Science|
|Article Type:||Case study|
|Date:||Aug 1, 2019|
|Previous Article:||Tough and Recoverable Triple-Network Hydrogels Based on Multiple Pairs of Toughing Mechanisms With Excellent Ionic Conductivity as Stable Strain...|
|Next Article:||On the Modeling and Optimization of UV-Curable Fluoro-Siloxane Formulations and Their Solvent Absorption and Adhesion Properties.|