Document not found! Please try again

In Situ Analysis of Melt-Drawing Behavior of Ultrahigh Molecular

Mar 18, 2015 - The characteristic plateau and strain-hardening regions are ..... Weight Polyethylene for Knee Prosthesis Biomacromolecules 2005, 6, 94...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/JPCB

In Situ Analysis of Melt-Drawing Behavior of Ultrahigh Molecular Weight Polyethylene Films with Different Molecular Weights: Roles of Entanglements on Oriented Crystallization Satomi Kato, Hidekazu Tanaka, Takeshi Yamanobe, and Hiroki Uehara* Division of Molecular Science, Faculty of Science and Technology, Gunma University, Kiryu, Gunma 376-8515, Japan S Supporting Information *

ABSTRACT: Ultrahigh molecular weight polyethylene (UHMW-PE) films having different molecular weights (MWs) were melt-drawn at 150 °C. The stress−strain curve for higher-MW film exhibits higher stress on the characteristic plateau region and a subsequent steeper increase of stress due to strain hardening. Structural changes during such melt-drawing were analyzed using in situ wideangle X-ray diffraction measurements. Hexagonal crystallization occurs at the beginning of the plateau region, independent of the sample MW. Once this hexagonal reflection intensity is saturated, it remains constant even at the later stage of draw. In contrast, orthorhombic reflection intensities gradually increase with increasing draw strain. Both of these oriented crystallizations into plateau hexagonal and increasing orthorhombic forms are accelerated with increasing MW. Correspondingly, the higher amount of extended chain crystals (ECCs) was confirmed from morphological observation for the resultant melt-drawn films of the higher-MW sample. Deep entanglements can effectively transmit the applied stress; thus, the oriented amorphous melts induce rapid hexagonal crystallization with disentangling shallow entanglements, which subsequently transforms into orthorhombic form. Such hexagonal crystallization plays the role of a thermodynamic pathway for growing such ECCs, where the stable orthorhombic form gradually accumulates with increasing draw strain.



In contrast, we6,7 utilize such entanglements for achieving molecular orientation where the melted UHMW-PE is directly tensile-drawn into the resultant high-performance films and membranes. These molecular entanglements effectively transmit drawing stress, resulting in oriented crystallization into ECCs with superior mechanical properties. The advantage of this melt-drawing technique is that it is solvent-free even for preparation of robust but thin membranes6 or nanoporous membranes of UHMW-PE.7 We8−11 have demonstrated that such melt-drawing of UHMW-PE initially induces oriented crystallization from amorphous melt into a transient hexagonal phase, followed by transformation into a final orthorhombic phase.9−11 The former hexagonal phase is specifically observable when normal MW-PE is melt-crystallized under high temperature (over 200 °C) and pressure (1000 atm).12−14 Crystalline molecules in this transient hexagonal phase contain a few gauche conformations,15 which travel along trans-zigzag propagation,16 resulting in molecular rotation around the chain axis. Such enhanced molecular mobility in the hexagonal form may cause chain slippage; thus, many ECCs are obtainable for this critical crystallization with high pressure and high temperature. This result suggests that hexagonal crystallization is crucial for

INTRODUCTION Ultrahigh molecular weight polyethylene (UHMW-PE) with a molecular weight (MW) exceeding one million exhibits excellent wear characteristics, including abrasion resistance and self-lubrication.1 Thus, UHMW-PE is used in sliding members,1,2 including gear wheels and artificial joints. Battery separators3 and fishing lines4 are other applications that utilize the superior mechanical properties of UHMW-PE. In these cases, molecular orientation with high crystallinity is necessary to improve properties of the targeted films and fibers. However, due to its longer chain length, UHMW-PE contains many molecular entanglements, which often resist usual molding processes such as kneading and extrusion. Thus, disentangling is required for preparing high-performance films or fibers of UHMW-PE. One of the most successful techniques for disentangling UHMW-PE chains is gel-spinning,5 where UHMW-PE is dissolved into gel with a few tens of times of the weight of organic solvents (e.g., decaline or xylene) followed by spinning and drawing into the oriented fiber. The resultant extended chain crystals (ECCs) have excellent mechanical properties. This process is modified for manufacturing UHMW-PE porous membranes for battery separators. Here, evaporation of these organic solvents results in nanoporous networks within the membranes. However, because these solvents are toxic, environmental loading and health risks to operators are a concern, especially in Europe. © XXXX American Chemical Society

Received: December 9, 2014 Revised: March 12, 2015

A

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

drawing device specially designed for our in situ X-ray measurements.9,10,17 This drawing device was equipped with a high-luminescent X-ray generator (Rigaku Co., MicroMaX007/HF). An X-ray incident beam concentrated using a confocal mirror was radiated perpendicular to the film surface. The applied X-ray wavelength was 1.5418 Å (Cu-Kα1). A series of WAXD patterns were simultaneously recorded during meltdrawing using a combination of an image intensifier (Hamamatsu Photonics, V7739) and CCD camera (Hamamatsu Photonics, C4742-98). The exposure time was 1 s for each pattern recording, and the interval time was 9 s for data storage. A blank pattern for air-scattering data was subtracted from each in situ pattern recorded during melt-drawing. The camera length was 70 mm, which was determined using the lead stearate standard for each drawing setup. Structural Analyses of Melt-Drawn Films. A PerkinElmer Diamond DSC was used for evaluating the melting behavior of the initial undrawn films and resultant melt-drawn films. Heating scans were performed from 80 to 180 °C at a heating rate of 10 °C/min in a nitrogen gas flow. The sample Tm was evaluated as the peak temperature of the melting endotherm. The temperature and fusion heat were calibrated using indium and tin standards. To avoid the effects of shrinkage of the drawn film and delay in heat transfer during the heating scan on the melting behavior, a small amount of silicone oil was placed between the sample and the bottom of the aluminum DSC pan. Morphologies of the resultant melt-drawn films were observed using a Hitachi field emission scanning electron microscope (SEM) S-4800 operated at 1.0 kV. The samples were coated with 5 Å thick Pt−Pd before SEM observations.

desirable ECC formation even with melt-drawing of UHMWPE. In contrast, drawing from normal MW-PE with several tens of thousands does not produce a hexagonal phase. This result indicates that sample MW plays an important role in hexagonal crystallization on drawing, implying that molecular entanglements are necessary for hexagonal crystallization during meltdrawing. We have reported the dependence of MW distribution and drawing conditions (temperature and speed) on the appearance of hexagonal phase during melt-drawing;9,11 however, the apparent MW dependence has not been investigated. Analysis of melt-drawing behavior with different MWs will help us interpret the ambiguous relationship between molecular entanglements and hexagonal crystallization. The hexagonal phase is not observable under conventional conditions at room temperature even for resultant melt-drawn UHMW-PE samples because it transforms into the usual orthorhombic phase upon cooling from the drawing temperature. Therefore, the phase transition behavior should be analyzed using in situ wide-angle X-ray diffraction (WAXD) measurement during melt-drawing. This analytical method is a powerful tool for understanding structural change during drawing or heating of various polymers, including both UHMW materials of above PE and poly(tetrafluoroethylene) (PTFE),17 poly(ethylene terephthalate),18,19 nylons,20,21 and poly(propylene).22,23 For melt-drawing examined in this study, the resultant sample contains two kinds of crystals: those crystallized with orientation during melt-drawing and those crystallized upon cooling after drawing. These crystallizations are often analyzed using the usual ex situ measurement, where these two crystallizations are inseparable. In situ measurement targets the former crystallization; thus, the simplified results are effective for tracing structural change during melt-drawing. Therefore, structural deformation steps from oriented amorphous to ultimate orthorhombic ECCs via hexagonal crystallization during melt-drawing were discussed using in situ WAXD analysis for different MWs in this study.



RESULTS Stress−Strain Behavior of Melt-Drawing. Tm values of the prepared UHMW-PE films with different MWs were estimated using DSC measurements in order to choose the appropriate melt-drawing temperature. Each film exhibits a single endotherm with a peak at 133 °C and an endotherm end at 138 °C independent of the sample MW. These melting characteristics are common even for melt-crystallized films prepared from the other series of UHMW-PE.24 In this study, the melt-drawing temperature was set at 150 °C, where all of the films are completely melted. Figure 1 compares the stress−strain curves recorded during melt-drawing for films prepared with different MWs in this study. The draw was stopped at a drawing time of 300 s, corresponding to the strain of 900% (DR = 10), as estimated from the displacement of preinked marks. The estimated strain values are also scaled in the longitudinal axis on the upper side. The same strain was obtained at the given drawing time independent of the sample MW, indicating that similar deformation takes place within these films. The characteristic plateau and strain-hardening regions are recognizable for all MWs, similar to the previous study.24 The plateau region begins at 40 s for the highest-MW film, and the subsequent strain-hardening region starts at 80 s. Lower-MW films exhibit higher-strain shifts of both beginning points of the plateau and strain-hardening regions with gentler slopes. The former plateau-stress region is characteristic of melt-drawing, and its stress indicates molecular entanglement density within the melted sample.25 The stress of the plateau region is greater for the higher-MW sample, indicating that the higher-MW film has more entanglements in the melted state than the lower-MW film. Such differences in stress−strain behavior indicate that the



EXPERIMENTAL SECTION Initial Film Preparation. Three UHMW-PE reactor powder (Hostalen) GUR series having different MWs were supplied by Ticona (Frankfurt). The viscosity average MWs were 4.4, 7.3, and 8.0 × 106 g/mol. These UHMW-PE reactor powders were compression-molded into initial films at 190 °C and 5 MPa for 10 min and then slowly cooled to room temperature. The resultant film thickness was 0.3 mm. Melt-Drawing. The drawing specimen was cut from the films into a 54 mm long dumbbell shape with a narrow deformation region 3 mm wide and 3 mm long. A uniaxial draw was performed at 150 °C, well above the static melting temperature (Tm) of all of the sample films with different MWs (133 °C), as evaluated by differential scanning calorimetry (DSC) measurements. Before drawing, the specimen was held at 150 °C for 5 min to equilibrate the temperature. The crosshead speed of drawing was 20 mm/min. After melt-drawing, the samples were cooled to room temperature at a constant cooling rate of 5 °C/min. The resultant draw ratio (DR) was estimated from the displacement between preinked marks on the specimen surface. This value was also used for calculating applied strain on the following stress profiles. In Situ Measurement. Stress during melt-drawing was continuously recorded using a load cell (Kyowa Electronic Instruments Co., Ltd., LUR-A-100NSA1) installed in the B

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 1. Comparison of stress profiles recorded during melt-drawing at 150 °C for films having MWs of 4.4 × 106 (black line), 7.3 × 106 (red line), and 8.0 × 106 (green line). The dotted line indicates the stop position of drawing. The drawing time was converted into the strain in the upper longitudinal scale based on the resultant displacement of preinked marks.

oriented crystallization during melt-drawing is highly dependent on the sample MW. In Situ WAXD Analysis of Phase Transition during Melt-Drawing. In situ WAXD measurements during meltdrawing enable us to investigate the relationship between structure deformation and stress−strain characteristics. Therefore, the series of WAXD patterns were recorded simultaneously with the stress−strain curves depicted in Figure 1. Figure 2 compares in situ WAXD patterns for three different MW samples with corresponding stress profiles. Independent of the sample MW, only ring-shaped amorphous halos are observed before drawing (0 s), indicating the isotropic melted state at 150 °C. Amorphous halos then gradually concentrate on the equator (vertical direction in Figure 1) before the plateau-stress region. At the beginning of the plateau region (40 s for the highest MW), an arc-shaped reflection appears on the equator. On the basis of the position of this reflection in the following line profile analysis, it is assigned to hexagonal (100) reflection. Simultaneously, another arc-shaped reflection appears slightly at the higher 2θ position, followed by the other much weaker reflection at the further higher 2θ position. The latter set is attributed to orthorhombic (110) and (200) reflections. These orthorhombic (110) and (200) reflection intensities gradually increase with increasing draw strain (drawing time). Furthermore, the reflection shape transforms from the initial arc to the final spot, especially for the highestMW film (Figure 2c). Similar reflection changes are observed for the lower-MW films. However, reflection intensities strongly depend on the sample MW when compared at the same draw strain. The hexagonal (100) reflection recorded in the plateau region is very weak for the lowest MW. In contrast, it appears strong for the higher-MW film. Orthorhombic (110) and (200) reflections are also stronger for the higher-MW film. In order to evaluate such MW effects on reflection intensities, 2θ line profiles were extracted from the series of in situ WAXD patterns recorded during melt-drawing. Figure 3 compares changes in these 2θ line profiles for different MW films. Color gradation from blue to red indicates reflection intensity. The amorphous scattering intensity rapidly decreases with the starting draw due to the reduction of sample thickness. The hexagonal (100), orthorhombic (110) and (200) reflections

Figure 2. In situ WAXD patterns and corresponding stress profiles recorded during melt-drawing for films with MWs of 4.4 × 106 (a), 7.3 × 106 (b), and 8.0 × 106 (c). Melt-drawing was performed at 150 °C. The stress profiles are the same as those depicted in Figure 1. The drawing direction for the WAXD patterns is horizontal. Subscript “o” indicates the orthorhombic form, and “h” indicates the hexagonal form. Enlarged images of the crystalline reflection regions are presented on the top side. The drawing time in seconds is indicated for each pattern.

locate at 2θ of 20.6, 21.7, and 23.6°, which agree with the reflection assignments for the oriented UHMW-PE fibers or films at the elevated temperature.4,26,27 A hexagonal (100) reflection appears at 40 s, corresponding to the beginning point of the plateau-stress region, for the highest-MW film. This reflection intensity is constant even at the final stage of meltdrawing, independent of the sample MW. In contrast, orthorhombic (110) and (200) reflections appear with little C

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

crystallization. The constant intensity of the hexagonal reflection is interpreted latter. Change in Crystal Reflection Intensity during MeltDrawing. To quantitatively compare these reflection changes, the equatorial 2θ line profiles in Figure 3 were deconvoluted into amorphous and crystalline peaks, corresponding to hexagonal (100) and orthorhombic (110) and (200) reflections. These peak deconvolutions were performed using the Voigt function combined with Lorentzian and Gaussian functions, as depicted in Figure S1 (see the Supporting Information). The resolved intensities depend on the sample thickness; thus, the estimated integral intensities of the crystalline reflection peaks are normalized by the initial complete amorphous scattering intensity recorded at 150 °C before drawing. Figure 4 plots the changes of normalized crystalline intensities as a function of drawing time. The initial increase of the hexagonal reflection intensity is rapid for the highest-MW film (Figure 4c), whereas the lowest-MW film exhibits a gradual increase (Figure 4a). Finally, the hexagonal intensity becomes constant with increasing drawing time even for the lowest-MW film. These constant values of hexagonal intensities are higher for higher-MW films. Such a plateau of hexagonal intensity is coincident with the plateau-stress phenomenon in Figures 1 and 2. In contrast, orthorhombic (110) and (200) reflection intensities gradually increase with increasing drawing time for all MW films. Such increases of orthorhombic intensities are more remarkable for the higherMW films, which is similar to the steeper strain hardening phenomenon on the corresponding stress−strain curve. The higher constant value of the hexagonal intensity and the steeper increases of orthorhombic intensities for the higher-MW film overlap even in the plateau-stress region, indicating that phase transition between them starts at the beginning point of the plateau region. When the orthorhombic (200) reflection is targeted, its intensity is less observable for the lowest MW sample (Figure 4a) even at the final stage of melt-drawing. In contrast, the higher-MW films (Figure 4b and c) exhibit apparent reflection intensities, although they are still much weaker than the (110) reflection intensities. Usually, the ratio of these reflection intensities (110)/(200) is 3/1 for the as-prepared undrawn UHMW-PE films recorded at room temperature. In Figure 4, this ratio is almost constant through the plateau and strainhardening regions, independent of the sample MW; however, a higher value (6/1) is obtained for the lowest-MW film, and 5/1 is obtained for the highest-MW film. These results indicate that the a-axis of the orthorhombic crystals formed during our meltdrawing UHMW-PE arranges perpendicular to the film surface. Similar a-axis orientation is known for tensile drawing from the less-entangled UHMW-PE membrane crystallized from solution casting, where this ratio increases to 7/1.28,29 Such a structural coincidence indicates that molecular disentangling is greatly enhanced for the lower-MW film, although the higherMW film maintains entanglement during melt-drawing examined in this study. Amorphous Orientation during Melt-Drawing. As described above, the formation of the orthorhombic phase is attributed to phase transition from hexagonal crystals. Thus, the origin of hexagonal crystallization may be amorphous orientation occurring before hexagonal crystallization (Figure 2). Therefore, amorphous scattering characteristics are analyzed using the series of in situ WAXD patterns. Here, the usual amorphous scattering for unoriented melt exhibits the typical

Figure 3. Changes in 2θ line profiles extracted along the equators of the series of in situ WAXD patterns recorded during melt-drawing for films with MWs of 4.4 × 106 (a), 7.3 × 106 (b), and 8.0 × 106 (c). Dotted white lines indicate the beginning point of plateau-stress regions in Figure 1. The intensity is represented by color gradation from blue (lower) to red (higher).

delay from the hexagonal appearance, and their intensities gradually increase through the plateau and subsequent strainhardening regions in the corresponding stress−strain curves. The order of hexagonal appearance with constant intensity and the increase of orthorhombic reflections imply that phase transition from the hexagonal to orthorhombic form occurs during melt-drawing. The hexagonal form is thus transient, in contrast with the final and stable state of the orthorhombic form. Indeed, our previous melt-drawing for UHMW-PE with a similar MW of 4 × 106 exhibited the hexagonal reduction when the orthorhombic increased at the latter stage of the draw.10 This is a typical phase transition from the hexagonal into orthorhombic form. The higher strain rate in this study accelerates the phase transition from the hexagonal into orthorhombic form, which overlaps with the initial hexagonal D

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 5. Comparison of integral intensities of equatorial (dots) and meridional (circles) amorphous scattering peaks for films with MWs of 4.4 × 106 (a), 7.3 × 106 (b), and 8.0 × 106 (c). For the latter series of plots, the meridional 2θ profiles were also extracted from the series of in situ WAXD patterns depicted in Figure 2.

Figure 4. Change in integral intensities of hexagonal (100) (red squares), orthorhombic (110) (blue dots), and orthorhombic (200) (green triangles) reflection peaks evaluated from equatorial line profiles extracted from the series of in situ WAXD patterns during melt-drawing of films with MWs of 4.4 × 106 (a), 7.3 × 106 (b), and 8.0 × 106 (c). The intensity was normalized by the amorphous scattering intensity recorded before drawing for each sample.

amorphous halo with increasing drawing time. Amorphous intensities along both directions should decrease with increasing draw strain due to thinning of the sample film. However, the intensity decreases depend on the scattering direction; the equatorial intensity always exceeds the meridional intensity, indicating the oriented state of the amorphous phase. Also, there are two categories of intensity decreases with increasing drawing time; the first is a rapid decrease before the beginning of the plateau region, and the second is a gentle decrease with entering the plateau region, which continues through the strain-hardening region. The former category is composed of amorphous changes alone, where the difference between the equatorial and meridional intensities of the amorphous phase increases with increasing drawing time. In contrast, the latter category accompanies oriented crystallization into the hexagonal form and subsequent transformation into the orthorhombic form, where the amorphous intensity

Debye−Scherrer ring. However, the oriented amorphous state is observable for all of the films with different MWs at the beginning of melt-drawing before hexagonal crystallization, as depicted in Figure S2 (Supporting Information). In other words, such a difference of amorphous intensity in the equatorial and meridional can be an index of the amorphous orientation. For these amorphous evaluations, other sets of 2θ line profiles were extracted along the meridian of the obtained in situ WAXD patterns, which corresponds to the unoriented amorphous fraction. In contrast, the usual equatorial line profiles depicted in Figure 3 indicate the oriented amorphous fraction. The obtained meridional profiles were similarly deconvoluted using the Voigt function. Figure 5 compares changes in the equatorial and meridional intensities of the E

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

difference is stable, which is coincident with plateau intensity of the hexagonal reflection. These results indicate that the amorphous orientation is almost complete at the beginning of the plateau region. The subsequent hexagonal crystallization consumes this amorphous orientation upon entering the plateau region. The latter constant amorphous orientation indicates that the strain-hardening phenomenon is not the result of amorphous orientation or hexagonal crystallization but is caused by the orthorhombic increase due to phase transition from the hexagonal form. The higher MW exhibits larger amorphous differences between the extracting directions, suggesting enhanced amorphous orientation. The large number of molecular entanglements for the higher-MW film induces rapid amorphous orientation just after starting the melt-draw. This MW dependence of the amorphous orientation is quite similar to that of hexagonal crystallization upon entering the plateau region (Figure 4). This means that the amorphous orientation is a precursor of hexagonal crystallization. Structural Analysis of Resultant Melt-Drawn Samples. The resultant morphologies of melt-drawn films with different MWs were also analyzed by ex situ measurements using DSC measurements and SEM observations. Films melt-drawn up to a DR of 10 were prepared by cooling after the series of in situ WAXD measurements up to a drawing time of 300 s. DSC thermograms of such melt-drawn samples are compared in Figure 6. Endotherm peaks at 133 and 140 °C coexist for the

Figure 7. SEM images for the resultant films melt-drawn up to DR = 10 for MWs of 4.4 × 106 (a), 7.3 × 106 (b), and 8.0 × 106 (c). The drawing direction is horizontal in these SEM images. Scale bar, 500 nm.

with a few longer but narrower ECCs. In contrast, the higherMW samples contain more ECCs that are shorter but thicker than those for the lower-MW films. Fewer FCCs are observable but are thicker than that of the lowest MW sample (Figure 7a). These morphological observations agree with the above DSC results. Considering that a thicker FCC is usually obtained by melt-crystallization at a higher temperature, the larger amount of ECCs formed during melt-drawing plays the role of crystallization nuclei on cooling after melt-drawing, which raises the crystallization temperature of FCCs for the higherMW sample. For the gel-spun UHMW-PE fiber,33 the Tm of the ECC often exceeds the thermodynamic equilibrium Tm of PE crystals (145 °C).34 However, it is much lower in this study. This indicates that ECC formation is restricted for melt-drawing in this study due to the lower DR than that of the gel-spinning method (DR = 100).

Figure 6. DSC melting thermograms for the resultant films meltdrawn up to DR = 10 for MWs of 4.4 × 106 (black), 7.3 × 106 (red), and 8.0 × 106 (green).

highest-MW film. A similar combination of lower and higher Tm’s has been reported for our previous results for melt-drawn UHMW-PE30 and PTFE films.31 The location of the lower Tm is similar to that of the Tm of initial undrawn films (132 °C); thus, this lower-temperature-side endotherm corresponds to melting of folded chain crystals (FCCs) formed upon cooling after melt-drawing. In contrast, the higher-temperature-side endotherm is ascribed to the melting of ECCs formed during melt-drawing, as discussed later. Here, the higher-T m endotherm becomes larger with increasing sample MW. This result indicates that a higher MW containing more entanglements induces more ECCs. In contrast, the lower-T m endotherm becomes smaller, and its peak position shifts into the higher-temperature side. This result predicts that the thickness of FCCs increases with increasing sample MW when the morphologies of the resultant DR = 10 films are compared. SEM observation also provides information on ECC formation within the resultant melt-drawn films. Independent of the sample MW, typical shish-kebab morphologies32 composed of FCCs and ECCs are recognizable in Figure 7. However, the ratio of ECCs and FCCs depends on the sample MW. For the lowest-MW sample, the main structure is FCCs



DISCUSSION These sets of in situ and ex situ data allow us to illustrate the MW dependence of the structural development during meltdrawing of UHMW-PE films, as depicted in Figure 8. Amorphous orientation (II) is caused by initial chain elongation between entanglements when melt-drawing is started from the isotropic melt (I). This amorphous orientation saturates at the beginning of the plateau-stress region, where hexagonal crystallization is initiated (III). Here, the amount the F

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 8. Schematic representations of structural changes during melt-drawing for samples with lower (a) and higher MWs (b). (I) Entangled amorphous state. (II) Oriented amorphous. (III) Hexagonal crystallization. (IV) Orthorhombic transformation. (V) ECC growth. Shallow and deep entanglements are marked by orange shadows and violet shadows.

disappear with the initial chain elongation (III), where the hexagonal crystallization is initiated between deep entanglements as an act by stress propagation. Such molecular disentangling during melt-drawing gives the resultant longer ECCs (V), as depicted in Figure 7. Namely, entanglement phase separation into the disentangled and entangled amorphous components proceeds at the early stage of meltdrawing for the lower-MW sample (II). However, fewer entanglements result in fewer ECCs; thus, less stacking narrows the resultant ECCs. In contrast, the higher MW (Figure 8b) contains more deep entanglements in the initial melted state (I). The higher stress level indicates that disentanglement is restricted for the higherMW sample. The corresponding higher amorphous orientation (II) act as a prominent stress propagation caused by such deep entanglements results in a higher level of hexagonal content (III); thus, the subsequent transformation into orthorhombic form is also remarkable (IV). Therefore, the resultant film contains more ECCs (V). However, ECC growth is limited by swept entanglements in the remaining amorphous phases; thus, it is shorter than that for the lower-MW film. The ECC width is much greater due to the effective connection between deep entanglements along the stacking direction. The roles of the entanglements on oriented crystallization during melt-drawing of UHMW-PE are multifold; deep entanglements effectively transmit the applied stress, leading to the higher amount of the hexagonal phase, but the shallow entanglements cause the disentangled amorphous components, growing the longer ECCs. Such different aspects of the molecular entanglements could be successfully revealed though a comparison of the in situ WAXD measurements for the series of MW samples examined in this study.

amorphous orientation determines the upper limit of the hexagonal content, which is the reason for the constant level of the hexagonal component even at the later stage of drawing. When the hexagonal content exceeds this limit within the plateau-stress region, the excess component transforms into the orthorhombic form (IV), which causes the later stress increase within the strain-hardening phenomenon. Such hexagonal crystallization proceeds from the center toward the side regions that contain entanglements, where the hexagonal component is the front of the oriented crystallization. Therefore, phase transformation from the initial hexagonal into orthorhombic also starts at the center part. The hexagonal phase thus plays the role of thermodynamic pathway for growing ECCs (V). Namely, all of the resultant orthorhombic crystals are formed through hexagonal crystals. This is a reason why the constant hexagonal intensity is still obtained even at the latter stage of draw, independent of the sample MW. In contrast, entanglements are gradually swept into the remaining amorphous phase at the sides, indicating that larger deformation with melt-drawing induces phase separation of entangled and disentangled components. For the lower-MW sample (Figure 8a), amorphous orientation is less pronounced (II), as depicted in Figure 5. Therefore, both hexagonal crystallization (III) and transformation into orthorhombic crystals (IV) are also restricted, resulting in a fewer ECCs (V). The enhanced a-axis orientation in the orthorhombic form, similar to drawing the less-entangled solution-crystallized sample, indicates that the ease of disentangling restricts the initial amorphous orientation. However, all of the entanglements are not necessarily disentangled because the stress transmission is still maintained even for the lower-MW sample, implying that there are the other “deep” entanglements. In this meaning, the disentangled amorphous components originally contain the “shallow” entanglements.24 Indeed, our previous in situ solid-state NMR measurement during melt-drawing UHMW-PE similarly observed two categories of entanglements having different spin−spin relaxation times.35 The shallow entanglements



CONCLUSIONS Structural formations during melt-drawing of UHMW-PE films with different MWs were analyzed using in situ WAXD measurements. The characteristic transient crystallization into hexagonal form begins upon entering the plateau-stress region G

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Molecular Weight Polyethylenes Having Different Molecular Characteristics. J. Phys. Chem. B 2008, 112, 5311−5316. (9) Kakiage, M.; Yamanobe, T.; Komoto, T.; Murakami, S.; Uehara, H. Transient Crystallization during Drawing from Ultra-High Molecular Weight Polyethylene Melts Having Different Entanglement Characteristics. Polymer 2006, 47, 8053−8060. (10) Uehara, H.; Kakiage, M.; Yamanobe, T.; Komoto, T.; Murakami, S. Phase Development Mechanism during Drawing from Highly Entangled Polyethylene Melts. Macromol. Rapid Commun. 2006, 27, 966−970. (11) Kakiage, M.; Yamanobe, T.; Komoto, T.; Murakami, S.; Uehara, H. Effects of Molecular Characteristics and Processing Conditions on Melt-Drawing Behavior of Ultrahigh Molecular Weight Polyethylene. J. Polym. Sci., Polym. Phys. Ed. 2006, 47, 2455−2467. (12) Wunderlich, B.; Arakawa, T. Polyethylene Crystallized from Melt under Elevated Pressure. J. Polym. Sci., Part A 1964, 2, 3697− 3706. (13) Rastogi, S.; Kurelec, L.; Lemstra, P. J. Chain Mobility in Polymer Systems: On the Borderline between Solid and Melt. 2. Crystal Size Influence in Phase Transition and Sintering of Ultrahigh Molecular Weight Polyethylene via the Mobile Hexagonal Phase. Macromolecules 1998, 31, 5022−5031. (14) Kurelec, L.; Rastogi, S.; Meier, R. J.; Lemstra, P. J. Chain Mobility in Polymer Systems: On the Borderline between Solid and Melt. 3. Phase Transformations in Nascent Ultrahigh Molecular Weight Polyethylene Reactor Powder at Elevated Pressure As Revealed by In Situ Raman Spectroscopy. Macromolecules 2000, 33, 5593−5601. (15) Tashiro, K.; Sasaki, S.; Kobayashi, M. Structural Investigation of Orthorhombic-to-Hexagonal Phase Transition in Polyethylene Crystal: The Experimental Confirmation of the Conformationally Disordered Structure by X-ray Diffraction and Infrared/Raman Spectroscopic Measurements. Macromolecules 1996, 29, 7460−7469. (16) Mowery, D. M.; Harris, D. J.; Schmidt-Rohr, K. Characterization of a Major Fraction of Disordered All-Trans Chains in Cold-Drawn High-Density Polyethylene by Solid-State NMR. Macromolecules 2006, 39, 2856−2865. (17) Morioka, T.; Kakiage, M.; Yamanobe, T.; Komoto, T.; Higuchi, Y.; Kamiya, H.; Arai, K.; Murakami, S.; Uehara, H. Oriented Crystallization Induced by Uniaxial Drawing from Poly(tetrafluoroethylene) Melt. Macromolecules 2007, 40, 9413−9419. (18) Kim, K.; Murata, T.; Kang, Y.; Ohkoshi, Y.; Gotoh, Y.; Nagura, M.; Urakawa, H. Microsecond Analysis of Quasi-Smectic Fibrillar Structure in the Continuous Fiber Drawing of Poly(ethylene terephthalate). Macromolecules 2011, 44, 7378−7384. (19) Kawakami, D.; Hsiao, B. S.; Ran, S.; Burger, C.; Fu, B.; Sics, I.; Chu, B.; Kikutani, T. Structural Formation of Amorphous Poly(ethylene terephthalate) during Uniaxial Deformation above Glass Temperature. Polymer 2004, 45, 905−918. (20) Wang, D.; Shao, C.; Ahao, B.; Bai, L.; Wang, X.; Yan, T.; Li, J.; Pan, G.; Li, L. Deformation-Induced Phase Transitions of Polyamide 12 at Different Temperatures: An in Situ Wide-Angle X-ray Scattering Study. Macromolecules 2010, 43, 2406−2412. (21) Murthy, N. S.; Wang, Z. G.; Akkapeddi, M. K.; Hsiao, B. S. Isothermal Crystallization Kinetics of Nylon 6, Blends and Copolymers using Simultaneous Small and Wide-Angle X-ray Measurements. Polymer 2002, 43, 4905−4913. (22) Yamamoto, Y.; Inoue, Y.; Onai, T.; Doshu, C.; Takahashi, H.; Uehara, H. Deconvolution Analyses of Differential Scanning Calorimetry Profiles of β-Crystallized Polypropylenes with Synchronized X-ray Measurements. Macromolecules 2007, 40, 2745−2750. (23) Reddy, K. R.; Tashiro, K.; Sakurai, T.; Yamaguchi, N.; Sasaki, S.; Masunaga, H.; Takata, M. Isothermal Crystallization Behavior of Isotactic Polypropylene H/D Blends as Viewed from Time-Resolved FTIR and Synchrotron SAXS/WAXD Measurements. Macromolecules 2009, 42, 4191−4199. (24) Uehara, H.; Nakae, M.; Kanamoto, T.; Zachariades, A. E.; Porter, R. S. Melt Drawability of Ultrahigh Molecular Weight Polyethylene. Macromolecules 1999, 32, 2761−2769.

and continues maintaining a constant composition even at the final stage of draw. In contrast, subsequent transformation into stable orthorhombic form causes linear growth with increasing draw strain. A higher-MW yields a higher hexagonal composition and a steeper increase of the orthorhombic form. These MW dependences on oriented crystallization during melt-drawing could be interpreted using the entanglement categorization model, including shallow and deep entanglements. Melt-drawing first initiates phase separation of corresponding disentangled and entangled components. The effective stress transmitting between the former components induces the molecular orientation of the latter components, resulting in hexagonal crystallization. Such oriented crystallization progresses with increasing drawing strain; however, subsequent transformation into orthorhombic form maintains a constant hexagonal composition. In contrast, the orthorhombic form accumulates as resultant ECCs, as confirmed by DSC measurements and SEM observations.



ASSOCIATED CONTENT



AUTHOR INFORMATION

* Supporting Information S

The peak fitting for the WAXD equatorial line profile and changes in in situ WAXD patterns recorded at the beginning of melt-drawing. This material is available free of charge via the Internet at http://pubs.acs.org. Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was partly supported by the JFE 21 Century Foundation of Japan and the Industrial Technology Research Grant Program from the New Energy and Industrial Technology Development Organization (NEDO) of Japan.



REFERENCES

(1) Xu, L.; Chen, C.; Zhong, G.; Lei, J.; Xu, J.; Hsiao, B. S.; Li, Z. Tuning the Superstructure of Ultrahigh-Molecular-Weight Polyethylene/Low-Molecular-Weight Polyethylene Blend for Artificial Joint Application. Appl. Mater. Interfaces 2012, 4, 1521−1529. (2) Rastogi, A.; Kurelec, L.; Lippits, D.; Cuijpers, J.; Wimmer, M.; Lemstra, P. J. Novel Route to Fatigue-Resistant Fully Sintered Ultrahigh Molecular Weight Polyethylene for Knee Prosthesis. Biomacromolecules 2005, 6, 942−947. (3) Arora, P.; Zhang, Z. J. Battery Separators. Chem. Rev. 2004, 104, 4419−4462. (4) Kakiage, M.; Tamura, T.; Murakami, S.; Takahashi, H.; Yamanobe, T.; Uehara, H. Hierachical Constraint Distribution of Ultra-High Molecular Weight Polyethylene Fibers with Different Preparation Methods. J. Mater. Sci. 2010, 45, 2574−2579. (5) Smith, P.; Lemstra, J. T. Ultra-High-Strength Polyethylene Filaments by Solution Spinning/Drawing. J. Mater. Sci. 1980, 15, 505− 514. (6) Uehara, H.; Tamura, T.; Hashidume, K.; Tanaka, H.; Yamanobe, T. Non-solvent Processing for Robust but Thin Membranes of UltraHigh Molecular Weight Polyethylene. J. Mater. Chem. A 2014, 2, 5252−5257. (7) Uehara, H.; Tamura, T.; Kakiage, M.; Yamanobe, T. Nanowrinkled and Nanoporous Polyrthylene Membranes via Entanglement Arrangement Control. Adv. Funct. Mater. 2012, 22, 2048−2057. (8) Kakiage, M.; Yamanobe, T.; Komoto, T.; Murakami, S.; Uehara, H. Phase Transitions during Heating of Melt-Drawn Ultrahigh H

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

(25) Lin, Y. H.; Juang, J. H. Onset of Entanglement. Macromolecules 1999, 32, 181−185. (26) Van Aerle, N. A. J. M.; Lemstra, P. J.; Braam, A. W. M. A RealTime X-ray Melting Study of Partly Drawn UHMW-Polyethylene. Polym. Commun. 1989, 30, 7−11. (27) Uehara, H.; Kanamoto, T.; Kawaguchi, A.; Murakami, S. RealTime X-ray Diffraction Study on Two-Stage Drawing of Ultra-High Molecular Weight Polyethylene Reactor Powder above the Static Melting Temperature. Macromolecules 1996, 29, 1540−1547. (28) Kanamoto, T.; Tsuruta, A.; Tanaka, K.; Takeda, M.; Porter, R. S. Superdrawing of Ultrahigh Molecular Weight Polyethylene. 1. Effect of Techniques on Drawing of Single Crystal Mats. Macromolecules 1988, 21, 470−477. (29) Furuhata, K.; Yokokawa, T.; Seoul, C.; Miyasaka, K. Drawing of Ultrahigh-Molecular-Weight Polyethylene Single-Crystal Mats: The Crystallinity. J. Polym. Sci. 1986, 24, 59−67. (30) Nakae, M.; Uehara, H.; Kanamoto, T.; Zachariades, A. E.; Porter, R. S. Structure Development upon Melt-Drawing Ultrahigh Molecular Weight Polyethylene: Effect of Prior Thermal History. Macromolecules 2000, 33, 2632−2641. (31) Uehara, H.; Arase, Y.; Suzuki, K.; Yukawa, Y.; Higuchi, Y.; Matsuoka, Y.; Yamanobe, T. Highly Transparent and Robust Poly(tetrafluoroethylene) Membrane Prepared by Biaxial MeltDrawing. Macromol. Mater. Eng. 2014, 299, 669−673. (32) Bashir, Z.; Odell, J. A.; Keller, A. Stiff and Strong Polyethylene with Shish Kebab Morphology by Continuous Melt Extrusion. J. Mater. Sci. 1986, 21, 3993−4002. (33) Pak, J.; Wunderlich, B. Reversible Melting of Gel-Spun Fibers of Polyethylene. Thermochim. Acta 2004, 421, 203−209. (34) Flory, P. J.; Vrij, A. Melting Points of Linear-Chain Homologs. The Normal Paraffin Hydrocarbons. J. Am. Chem. Soc. 1963, 85, 3548−3552. (35) Kakiage, M.; Uehara, H.; Yamanobe, T. Novel In Situ NMR Measurement System for Evaluating Molecular Mobility during Drawing from Highly Entangled Polyethylene Melts. Macromol. Rapid Commun. 2008, 29, 1571−1576.

I

DOI: 10.1021/jp512246d J. Phys. Chem. B XXXX, XXX, XXX−XXX