Thermo-mechanical behavior of electrospun thermoplastic polyurethane nanofibers

European Polymer Journal 49 (2013) 3851–3856

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

Macromolecular Nanotechnology

Thermo-mechanical behavior of electrospun thermoplastic polyurethane nanofibers Dmitriy Alhazov a, Arkadiusz Gradys a,b, Pawel Sajkiewicz b, Arkadii Arinstein a, Eyal Zussman a,⇑ b

Faculty of Mechanical Engineering, Technion-Israel Institute of Technology, Haifa 32000, Israel Institute of Fundamental Technological Research, Polish Academy of Sciences, Pawinskiego 5B, 02-106 Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 5 June 2013 Received in revised form 22 September 2013 Accepted 29 September 2013 Available online 10 October 2013 Keywords: Block-copolymer Electrospinning Nanofibers Thermo-mechanical properties

a b s t r a c t Analysis of the thermo-mechanical behavior of electrospun thermoplastic polyurethane (TPU) block co-polymer nanofibers (glass transition temperature 50 °C) is presented. Upon heating, nanofibers began to massively contract, at 70 °C, whereas TPU cast films started to expand. Radial wide-angle X-ray scattering (WAXS) profiles of the nanofibers and the films showed no diffraction peaks related to crystals, whereas their amorphous halo had an asymmetric shape, which can be approximated by two components, associated with hard and soft segments. During heating, noticeable changes in the contribution of these components were only observed in nanofibers. These changes, which were accompanied with an endothermic DSC peak, coinciding with the start of the nanofibers contraction, can be attributed to relaxation of an oriented stretched amorphous phase created during electrospinning. Such structure relaxation becomes possible when a portion of the hard segment clusters, forming an effective physical network, is destroyed upon heating. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Polymer-based fibers, with diameters ranging from a few microns to a few nanometers, can be fabricated by means of electrospinning, during which a polymer solution, or melt, is extruded from a spinneret and forms a jet at the tip, due to the effect of an applied strong electric field [1]. The resulting jet then undergoes extreme elongation and thinning, in the order of 105, accompanied by a strain rate, in the order 103 s1, leading to stretching and orientation of the polymer chains [2–4]. Extremely rapid solvent evaporation occurs with continued jet flow, resulting in formation of a fiber, within a few milliseconds, with a polymer macrostructure in a non-equilibrium state [5]. The geometrical confinement of the as-spun [6,7] fibers may influence the mobility and relaxation behavior of the polymer macromolecules and affect their macroscopic physical properties [8–11]. ⇑ Corresponding author. Tel.: +972 48292836. E-mail address: [email protected] (E. Zussman). 0014-3057/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eurpolymj.2013.09.028

In the present study, we utilize the electrospinning process to fabricate block copolymer (BCP) nanofibers. BCP solutions and melts are known to self-assemble into a variety of nanoscale morphologies, including spheres, rods, micelles, lamellae, vesicles, tubules, and cylinders, dictated by the volume fraction and interaction parameters between different blocks [12,13]. Control of the size, shape, periodicity and long order of these nanoscale microdomains is essential in design of submicron-scale electronic, optical, and mechanical devices [14–17]. In an effort to obtain novel morphologies in long nanostructures, nanofibers of poly(styrene–butadiene-styrene) triblock-copolymer solution were electrospun [18]. Irregular microphase separation on the surface of electrospun fibers was observed. Although these structures were seen to improve on annealing, the domains remained largely disordered. Other works using poly(styrene-b-poly(4-vinylpyridine)) [19], poly(styrene-b-dimethylsiloxane) [20], or poly(styrene-b-polyisoprene) [21] block copolymer nanofibers also report the formation of irregular morphologies. Similar irregular micropahse separation was obtained in films

MACROMOLECULAR NANOTECHNOLOGY

a

D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856

and attributed to strong shear deformation [22,23], rapid solvent evaporation [24] and surface effect [25]. Co-electrospinning [26,27], a modified electrospinning process, yielded poly(styrene-b-polyisoprene), which evolved via annealing, forming a stacked lamellar-disc morphology, with each disc perpendicular to the fiber axis, and a longrange order of parallel alternating concentric cylinders [28]. In this study, we focus on thermoplastic polyurethane elastomers (TPUs), which constitute an important subclass of BCPs; they are made from diisocyanates and polyols, containing an ester or ether backbone, which lead to linearly segmented copolymers with alternating sequences of ‘‘hard’’ (diisocyanate) and ‘‘soft’’ (diols, diacids) segments. The microphase-separated morphology [29] that results from the polarity difference of diol/diacid and diisocyanate segments, leads to superior mechanical properties, such as higher strength, elongation, and modulus, with respect to other elastomers [30]. In order to explore the thermal behavior of TPU structures and their application as shape-memory [31,32], study of their thermo-mechanical properties is required. Towards this goal, we evaluated the strain-temperature behavior of TPU-based cast film and electrospun fibers. When heating at 1 °C/min, fiber mats and cast film strain was nearly fixed, but at 70 °C, fibers began to massively contract (40% axially) at 0.013 °C1, while cast films started to expand at 0.003 °C1. Increasing the heat rate to 5 °C/min yielded similar responses, but shifted the starting temperature of the contraction and expansion of the fibers and the films respectively, up to 85 °C (see Fig. 1). Contraction of poly(ethylene terephthalate) (PET) films and melt-spun fibers in the range of 0.0006–0.006 °C1, has already been reported for in several studies [33–35]. Upon heating above the glass transition, the overall shrinkage process involves a rapid initial stage of rubber-like contraction of the molecular network, associated with disorientation in the amorphous phase. Contraction attributed to negative thermal expansion of typical aramid fibers (Kevlar) or polyethylene fibers, is generally a nonlinear, temperature-dependent phenomenon, in the order of 106 C1. However, with increasing temperatures, the coefficient of thermal expansion becomes more negative, 10

Cast Film

0

Strain (%)

MACROMOLECULAR NANOTECHNOLOGY

3852

-10

Fiber mat

1oC/min

-20

o

5 C/min -30 30

40

50

60

70

80

90

primarily due to crystal contraction resulting from unit cell deformation [36,37]. The electrospun nanofiber contraction observed in our experiments, cannot be directly attributed to relaxation (disorientation) in the amorphous phase obtained when the stretched polymer is heated above its glass transition, since the glass transition temperature of the orientated amorphous phase of the tested TPU is 50 °C. Thus, at room temperature, relaxation should be expected immediately after fabrication. In our case, despite prolonged storage (several days) of the nanofibers under conditions above their glass transition temperature, unlike the above systems, contraction only began when reaching a certain temperature, while below this temperature, fibers were practically stable. Thus, a different mechanism governs this phenomenon. Based on the above mentioned principles, this work pursues a means of exploring the thermomechanical properties of electrospun TPU nanofibers. Thermal and thermomechanical properties were examined by Dynamic Mechanical Analysis (DMA) and differential scanning calorimetry (DSC). Structure analysis was performed by wideangle X-ray scattering (WAXS), at room and at elevated temperatures. It is hypothesized that the confined supermolecular nanoscale structure, developed in the electrospinning process, causes a unique relaxation process of an effective physical network in TPU nanofibers, which is composed of clusters of hard segments connected by amorphous tie chains. 2. Materials and methods 2.1. Materials Krystalflex PE-399, a TPU, was purchased from Huntsman. This TPU is a block copolymer composed from a poly(tetramethylene ether) glycol (PTMG)-type soft block and an aliphatic diisocyanate. Analytical-grade dimethylformamide (DMF) and tetrahydrofuran (THF) were purchased from Frutarom Ltd, Israel. 2.2. Gel permeation chromatography (GPC) characterization Molecular weight was obtained by gel permeation chromatography (GPC). The sample was injected using a Waters 1525 Binary HPLC Pump equipped with an autosampler. GPC/SEC 7.8  300 mm columns (Varian) were used. Molecular weight was determined by comparing to a universal calibration curve, obtained by using polystyrene standards. The solvent used was HPLC-grade THF, purchased from BioLab, without stabilizers. The concentration of the sample was 8 mg/mL, and the preparation of the sample, as well as running, was performed at room temperature. Using GPC, a molecular weight of Mn50 [kDa] was obtained for the TPU.

Temperature (°C) Fig. 1. Strain vs. temperature response of TPU-based cast film and electrospun fibers with diameters in the range of 500–800 nm. The experiment was carried in a dynamic mechanical analysis machine (DMA) in force control mode. (TPU–Krystalflex PE-399, Huntsman, USA).

2.3. Fibers and films preparation A solution for electrospinning was prepared by dissolving 1.2 g TPU in 8.8 g DMF and THF (7:3 (w/w)), to obtain a

3853

D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856

DSC measurements were performed using Mettler Toledo STARe DSC 1 system, equipped with a high-sensitivity HSS7 sensor. The instrument was calibrated for temperature and heat flow, using indium and zinc standards. Measurements were performed at a heating rate of 2 K min1 from 65 to 250 °C, under nitrogen purge. Masses of the samples were 9 mg. 2.5. Thermo-mechanical characterization Thermomechanical tests were executed by Dynamic Mechanical Analysis (TA-DMA Q800) on films and aligned fiber mats. Samples were 12 mm long, 3 mm wide and 0.09–0.12 mm thick. Clamp tightening torque for fiber mat and bulk film was 3 in-lb. For temperature sweep experiments, samples were subjected to sinusoidal displacement with 0.1% strain, at a fixed frequency of 1 Hz from 120 to 130 °C, and a heating rate of 2 °C min1. Four specimens were tested for each fiber mat composition. For strain-temperature behavior of TPU cast films and electrospun fibers, the experiments were carried in a force control mode, with a target force of 0.003 N. The samples were heated from room temperature at two heat rates (1 °C/min and 5 °C/min), up to 100 oC. 2.6. Wide-angle X-ray scattering (WAXS) WAXS measurements were performed using a Bruker D8 Discover diffractometer. Cu Ka radiation (wavelength of 0.1542 nm) was used at the applied voltage 40 kV, and current 40 mA. All measurements were performed in reflection mode, using Bragg–Brentano geometry, with a 1 mm slit and two Soller collimators applied on both sides. Considering very weak molecular orientation, as obtained from a preliminary scan using 2-D detectors, we used a highly sensitive Lynx Eye 1-D silicon strip detector. The angular range of measurements, 2, was between 10 and 30 deg, with a step of 0.005 deg and a time of data accumulation at particular angular point of 0.025 s. An Anton-Paar TTK 450 temperature chamber was used. The temperature program consisted of heating at a rate of 10 K/min, while holding the temperature constant during particular WAXS exposition. WAXS profiles were registered at 22 °C and 105 °C.

The morphology of the electrospun nanofibers was characterized by scanning electron microscopy (SEM). The diameter of the nanofibers ranged between 500 and 800 nm, where the fiber mat porosity was q = 0.8 ± 0.07 (q = 1 – (Nanofibers apparent density/Bulk density of raw TPU)). When analyzing the alignment of the fibers, it was found that the orientation distribution was such that one standard deviation from the main director was at an angle of about 4°. Analysis of radial WAXS profiles of the fiber mats and cast films showed no diffraction peaks related to crystals, whereas their amorphous halo had an asymmetric shape (Fig. 2). When heating from 22 °C to 105 °C the shape of film peaks underwent no significant change during heating; only a slow increase in peak asymmetry was observed. At the same time, fiber mat peak asymmetry at 22 °C is substantially reduced after heating; the shape of the fiber mat peak at 105 °C was almost symmetric (see details in Table 1). The general conclusion which can be drawn regarding the physical origin of the asymmetric shape of TPU, is that the structure is inhomogeneous (e.g., spatial separation of hard and soft segments with different polarity as a general rule for TPU), which can result in the existence of various components within the amorphous halo, with different, most probable (characteristic) scattering distances. Since the scattering distances of the hard and the soft segments in TPU are dissimilar, we assume that the asymmetric peaks consist of two components which can be associated

0.20

Film 22oC Fiber 22oC Film 105oC Fiber 105 oC

0.15

0.10

0.05

0.00 12

14

16

18

20

22

24

26

28

30

2θ Fig. 2. WAXS profiles of the film and fiber mats, registered at 22 °C and 105 °C. Profiles are normalized by the integrated intensity over the total WAXS profile, demonstrating modification of the peak shapes with increasing temperature.

Table 1 Characteristics of the WAXS peaks.

Film Fiber

Temperature, °C

Width (deg)

Skewness

22 105 22 105

4.74 5.32 4.28 9.70

6.22 7.54 4.22 0.62

MACROMOLECULAR NANOTECHNOLOGY

2.4. Thermal characterization

3. Results and discussion

Normalized Intensity

12% (w/w) solution. The TPU solutions were electrospun to respectively form fiber mats of oriented polymer monolithic nanofibers. The flow rate, controlled by a syringe pump, of the TPU solution was constant (0.9 mL/h), with an electrostatic field of 1.2 kV/cm. A collector wheel, positioned at a distance of 12 cm from the spinneret (needle 23G) and with a tangential velocity of 6 m/s, was used to collect aligned fibers. The ambient temperature was 21 ± 1 °C and humidity 48–51%. The films were prepared by casting the TPU solution onto a square Teflon mold (40 mm length, 2 mm in depth) and air-dried overnight. The film was removed from the mold and dried under vacuum for another 2 h.

D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856

0.20

(a)

Fiber 22oC C1 C2 Fiber 105oC C1 C2

0.15

0.10

0.05

DSC heating runs (Fig. 4) clearly demonstrated endothermic peaks located at 75.58 °C and 81.42 °C, with low enthalpy values of 3.41 and 5.85 J g1, for the fiber mat and the film samples, respectively. The endothermic character of the peaks indicates a melting type of transition. Another thermal effect, located at 30 °C and 38 °C, for the fiber mat and the film, respectively, precedes the endothermic peak. This effect is associated with the abrupt change in the course of the heat flow. Analysis of storage (E0 ) and loss (E00 ) moduli and of loss factor (tand), indicated three thermal transitions in both

0.2 J(gK)

-1

Fiber mat o

endo Heat flow /

Tm= 75.58 C -1 ΔH= 3.41 Jg

Film o

Tm= 81.42 C -1 ΔH= 5.85 Jg

-25

0

25

50

75

100

125

150

Temperature / oC Fig. 4. DSC heating runs for the fiber mat and the cast film samples. The arrows indicating the thermal effect due to the abrupt change in the heat flow.

0.20

Normalized Intensity

with the local ordering of hard and soft segments, respectively. Thus, in order to further understand the physical factors leading to the asymmetric shape (observed in Fig. 2), we decomposed the peaks (using Pearson VII function), into two components (C1 and C2), see Fig. 3 and Table 2. The two components (C1 and C2) maxima can roughly be related to the intermolecular distances which are 5.6 Å and 4.9 Å for C1 and C2 respectively. Since the distance between the hard segments in TPU should be higher than the distance between the soft segments [38], we conjecture that component C1 belongs to the characteristic, or most-probable scattering distance between hard segments, and that C2 belongs to the most-probable scattering distance between soft segments. When heating from 22 °C to 105 °C, major changes in the contribution of the components were observed in the fiber mat, namely, an increase in the content of the hard segments component (C1). At elevated temperature, the content of C1 comprised 95% of the amorphous part, whereas the soft segments component (C2) comprised only 5% (see Fig. 3a). That means that in the fibers, heating to 105 °C affects mostly the soft segments, decreasing their orientational ordering. In contrast, in the cast films, there is no essential change in the peak shape after heating (see Fig. 3b). When analyzing the change in the components width of the fiber mat upon heating, an increase of 80% and 30% in the width of components C1 and C2, respectively, was found, indicating decrease in the scale of ordered regions. At the same time, in the cast films, only minor changes in width of both components were observed.

Normalized Intensity

MACROMOLECULAR NANOTECHNOLOGY

3854

(b)

Film 22o C C1 C2 Film 105oC C1 C2

0.15

0.10

0.05

0.00

0.00 12

14

16

18

20

22

24

26

28

30

12

14

16

18

20

22

24

26

28

30

Fig. 3. WAXS profiles and their decomposition into two components (C1 and C2), registered at room 22 °C (solid lines) and at 105 °C (dashed lines), for fiber mat (a) and film (b).

Table 2 Characteristics of the decomposed WAXS peaks. Temperature, °C

Film Fiber mat

22 105 22 105

C1

C2

2hmax (deg)

Width (deg)

2hmax (deg)

Width (deg)

19.23 18.96 19.32 18.97

4.03 3.82 3.71 6.76

21.95 23.35 22.01 23.46

5.56 5.92 4.79 6.12

3855

D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856

100

1

10

100

10

1

Film Fiber Mat

1

0.1

0.1

tanδ

tanδ Loss Modulus-E" [MPa]

Storage Modulus-E' [MPa]

1000

0.1 -100

-50

0

50

100

Temperature [ oC]

the fiber mat and the cast film (Fig. 5). The lowest temperature transition (50 °C) is associated with the glass transition, Tg of soft segments (not detected by DSC), thus at room temperature, the high mobility of polymer matrix allows the relaxation of its possible non-equilibrium state. The peak in tand (25 °C), can be associated with destruction of weak-ordered hard segments clusters. The last transition starts at approximately 70 °C, and ends at 120 °C. This transition is apparently associated with destruction of ordered hard segment clusters. This peak coincides with the endothermic peak, obtained by the DSC, as described above. Note that after this thermal transition (at 120 °C), the storage modulus almost vanishes, indicating the destruction of the effective physical network. The transition (70–120 °C) corresponds to thermal transition, previously observed in high crystallinity segmented polyurethanes [39–44]. In the case of low concentrations of hard segments (less than 40%), Van Bogart et al. [45] found an endotherm peak at 50–60 °C, associated with the destruction of ordered hard segment clusters. Suggested contraction mechanism: The above experimental results demonstrate that the changes in the polymer structure of films and fibers upon heating differ. In fibers, the structural changes can be attributed to the

10

(a)

relaxation of the non-equilibrium state of the polymer matrix (stress and orientation) [46]. The dramatic decrease of the portion of soft segment clusters demonstrates that the disordering (disorientation) of soft segments is dominant. Just this disorientation of stretched portion of macromolecules should result in local shortening of polymer matrix and therefore in a shortening of the fibers as a whole, as was observed in the experiment. Unlike the fibers, in the cast films, the disordering of both hard and soft segment clusters is of the same order. Apparently, the effective physical network of hard segment clusters, connected by tie chains, is broken upon heating, resulting in polymer plasticity and consequently, to noticeable expansion of the film. (Note that this effective physical network preserves the solid-like state of polymer matrix above Tg.) In order to better understand the destruction process of the effective physical network, we focused on the highest thermal transition temperature (see Fig. 5). Due to the broad range of transition temperatures (70–120 °C), we can assume that the hard segment clusters are gradually destroyed. The destroyed clusters allow local mobility of amorphous soft segments, whereas the intact clusters maintain the effective physical network. As temperature increases, the proportion of the intact clusters decreases, 225

90

200

(b)

70 -10

Fiber Mat Film

60

-20

50

-30

40 30

-40 0.1

1

10

Time [min]

100

1000

Fiber Mat Film 10

Strain [%]

Strain [%]

0

Temperature [oC]

80

0 -10 -20 -30 -40 30

45

60

75

90

105

120

135

Temperature [ oC]

Fig. 6. (a) The strain response vs. time of fiber mat and cast film, under isothermal conditions at 90 °C. (b) The strain response vs. temperature of fiber mat and film continuously heated from room temperature until initiation of material flow.

MACROMOLECULAR NANOTECHNOLOGY

Fig. 5. Storage and Loss moduli, and tand at 1 Hz, of TPU-based electrospun fiber mat and cast film.

MACROMOLECULAR NANOTECHNOLOGY

3856

D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856

and the physical network will eventually be completely destroyed. This hypothesis can be verified by thermo-mechanical tests. Unlike the previous experiment (Fig. 1), the above thermal transition begins and, when reaching 90 °C isothermal condition is maintained for more than 15 h (Fig. 6a). As expected, the film sample expanded, while the fiber mat sample began to contract at 70 °C. When reaching 90 °C, the contraction continued at the same rate and persisted for approximately 100 min. Then, the contraction stopped and the strain was kept nearly constant at 40%, for approximately 200 min. At that moment, the fiber mat began to slowly expand. In another experiment (similar to the one presented in Fig. 1), the samples were heated up to 130 °C (Fig. 6b). Up until 105 °C, both the fiber mat and the film underwent contraction and expansion, as shown before. However, when heating above 105 °C, the strain direction of the fiber mats changed and they began to expand. These results support our hypothesis that the fiber mat contraction behavior is controlled by the destruction of hard segment clusters. The proportion of the destroyed clusters increases upon heating in range of 70–105 °C. At temperatures above 105 °C, all clusters in both fiber mat and cast film are destroyed, and the material starts to intensively expand and, finally, to flow. 4. Conclusions Although the experimental observations demonstrate that the ordering level in both fiber mat and cast film is similar, their response upon heating differs dramatically. The unexpected fiber contraction is attributed to the non-equilibrium microstructure (the spatial distribution of ordered regions), and more specifically to relaxation of stretched amorphous phase formed in electrospun nanofibers. Thorough analysis of the experimental results (WAXS, DSC, and DMA) indicates that in block co-polymer systems with low glass transition temperature such structure relaxation becomes possible only when a portion of the hard segment clusters, forming an effective physical network, is destroyed upon heating. The details of this microstructure, as well as its effect on the behavior of the fibers are part of ongoing research. The studied phenomenon can help in further understanding the physical reasons for the unique behavior of the electrospun nanofibers. Acknowledgments We gratefully acknowledge the financial support of the RBNI–Russell Berrie Nanotechnology Institute, the Israel Science Foundation (ISF Grant No. 770/11), and the DFG within the GIP project.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46]

Reneker D, Yarin A, Zussman E, Xu H. Adv Appl Mech 2007;41:43. Yarin AL, Koombhongse S, Reneker DH. J Appl Phys 2001;89:3018. Reneker DH, Yarin AL. Polymer 2008;49:2387. Shin YM, Hohman MM, Brenner MP, Rutledge GC. Polymer 2001;42:9955. Arinstein A, Zussman E. J Polym Sci Pol Phys 2011;49:691. Curgul S, Van Vliet KJ, Rutledge GC. Macromolecules 2007;40:8483. Wong SC, Baji A, Leng SW. Polymer 2008;49:4713. Arinstein A, Burman M, Gendelman O, Zussman E. Nat Nano 2007;2:59. Arinstein A, Liu Y, Rafailovich M, Zussman E. EPL 2011;93:46001. Wang W, Barber AH. Nanotechnology 2010;21:225701. Liao CC, Wang CC, Shih KC, Chen CY. Eur Polym J 2011;47:911. Fasolka MJ, Mayes AM. Ann Rev Mater Res 2001;31:323. Lodge TP. Macromol Chem Phys 2003;204:265. Bockstaller MR, Mickiewicz RA, Thomas EL. Adv Mater 2005;17:1331. Ruzette AV, Leibler L. Nat Mater 2005;4:19. Park C, Yoon J, Thomas EL. Polymer 2003;44:6725. Matsumoto H, Ishiguro T, Konosu Y, Minagawa M, Tanioka A, Richau K, et al. Eur Polym J 1866;2012:48. Fong H, Reneker DH. J Polym Sci Pol Phys 1999;37:3488. Ruotsalainen T, Turku J, Heikkila P, Ruokolainen J, Nykanen A, Laitinen T, et al. Adv Mater 2005;17:1048. Ma ML, Hill RM, Lowery JL, Fridrikh SV, Rutledge GC. Langmuir 2005;21:5549. Kalra V, Kakad PA, Mendez S, Ivannikov T, Kamperman M, Joo YL. Macromolecules 2006;39:5453. Pakula T, Saijo K, Kawai H, Hashimoto T. Macromolecules 1985;18:1294. Fairclough JPA, Salou CLO, Ryan AJ, Hamley IW, Daniel C, Helsby WI, et al. Polymer 2000;41:2577. Kim G, Libera M. Macromolecules 1998;31:2670. Anastasiadis SH, Russell TP, Satija SK, Majkrzak CF. Phys Rev Lett 1852;1989:62. Sun ZC, Zussman E, Yarin AL, Wendorff JH, Greiner A. Adv Mater 2003;15:1929. Reznik SN, Yarin AL, Zussman E, Berkovici L. Phys Fluids 2006;18:062101. Kalra V, Mendez S, Lee JH, Nguyen H, Marquez M, Joo YL. Adv Mater 2006;18:3299. Leibler L. Macromolecules 1980;13:1602. Curgul S, Yilgor I, Yigor E, Erman B, Cakmak M. Macromolecules 2004;37:8676. Lee BS, Chun BC, Chung YC, Sul KI, Cho JW. Macromolecules 2001;34:6431. Hu JL, Zhu Y, Huang HH, Lu J. Prog Polym Sci 2012;37:1720. Gupta VB, Radhakrishnan J, Sett SK. Polymer 1993;34:3814. Lim JY, Kim SY. J Polym Sci Pol Phys 2001;39:964. Pinnock PR, Ward IM. Brit J Appl Phys 1966;17:575. Rojstaczer S, Cohn D, Marom G. J Mater Sci Lett 1985;4:1233. Kobayash Y, Keller A. Polymer 1970;11:114. Liff SM, Kumar N, McKinley GH. Nat Mater 2007;6:76. Koberstein JT, Russell TP. Macromolecules 1986;19:714. Koberstein JT, Galambos AF. Macromolecules 1992;25:5618. Saiani A, Daunch WA, Verbeke H, Leenslag JW, Higgins JS. Macromolecules 2001;34:9059. Saiani A, Rochas C, Eeckhaut G, Daunch WA, Leenslag JW, Higgins JS. Macromolecules 2004;37:1411. Saiani A, Novak A, Rodier L, Eeckhaut G, Leenslag JW, Higgins JS. Macromolecules 2007;40:7252. Chen TK, Shieh TS, Chui JY. Macromolecules 1998;31:1312. Vanbogart JWC, Gibson PE, Cooper SL. J Polym Sci Pol Phys 1983;21:65. Greenfeld I, Arinstein A, Fezzaa K, Rafailovich MH, Zussman E. Phys Rev E 2011;84:041806.