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,⇑ a Faculty of Mechanical Engineering, Technion-Israel Institute of Technology, Haifa 32000, Israel b Institute of Fundamental Technological Research, Polish Academy of Sciences, Pawinskiego 5B, 02-106 Warsaw, Poland article info 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 abstract 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 accompa- nied 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 solu- tion, 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 elonga- tion and thinning, in the order of 10 5 , accompanied by a strain rate, in the order 10 3 s 1 , leading to stretching and orientation of the polymer chains [2–4]. Extremely rapid solvent evaporation occurs with continued jet flow, result- ing 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]. 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 microdo- mains is essential in design of submicron-scale electronic, optical, and mechanical devices [14–17]. In an effort to ob- tain 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 ob- served. 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(sty- rene-b-polyisoprene) [21] block copolymer nanofibers also report the formation of irregular morphologies. Similar irregular micropahse separation was obtained in films MACROMOLECULAR NANOTECHNOLOGY 0014-3057/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eurpolymj.2013.09.028 ⇑ Corresponding author. Tel.: +972 48292836. E-mail address: [email protected](E. Zussman). European Polymer Journal 49 (2013) 3851–3856 Contents lists available at ScienceDirect European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Dmitriy Alhazov a, Arkadiusz Gradys a,b, Pawel Sajkiewicz b, Arkadii Arinstein a,Eyal Zussman a,⇑a Faculty of Mechanical Engineering, Technion-Israel Institute of Technology, Haifa 32000, Israelb 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
LAR
NA
NO
TECH
NO
LOG
Article history:Received 5 June 2013Received in revised form 22 September2013Accepted 29 September 2013Available online 10 October 2013
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 filmsstarted to expand. Radial wide-angle X-ray scattering (WAXS) profiles of the nanofibersand the films showed no diffraction peaks related to crystals, whereas their amorphoushalo had an asymmetric shape, which can be approximated by two components, associatedwith hard and soft segments. During heating, noticeable changes in the contribution ofthese components were only observed in nanofibers. These changes, which were accompa-nied 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 duringelectrospinning. Such structure relaxation becomes possible when a portion of the hardsegment clusters, forming an effective physical network, is destroyed upon heating.
� 2013 Elsevier Ltd. All rights reserved.
OLE
C
1. Introduction In the present study, we utilize the electrospinning
MA
CRO
M
Polymer-based fibers, with diameters ranging from afew microns to a few nanometers, can be fabricated bymeans of electrospinning, during which a polymer solu-tion, or melt, is extruded from a spinneret and forms ajet at the tip, due to the effect of an applied strong electricfield [1]. The resulting jet then undergoes extreme elonga-tion and thinning, in the order of 105, accompanied by astrain rate, in the order 103 s�1, leading to stretching andorientation of the polymer chains [2–4]. Extremely rapidsolvent evaporation occurs with continued jet flow, result-ing in formation of a fiber, within a few milliseconds, witha polymer macrostructure in a non-equilibrium state [5].The geometrical confinement of the as-spun [6,7] fibersmay influence the mobility and relaxation behavior ofthe polymer macromolecules and affect their macroscopicphysical properties [8–11].
process to fabricate block copolymer (BCP) nanofibers.BCP solutions and melts are known to self-assemble intoa variety of nanoscale morphologies, including spheres,rods, micelles, lamellae, vesicles, tubules, and cylinders,dictated by the volume fraction and interaction parametersbetween different blocks [12,13]. Control of the size, shape,periodicity and long order of these nanoscale microdo-mains is essential in design of submicron-scale electronic,optical, and mechanical devices [14–17]. In an effort to ob-tain novel morphologies in long nanostructures, nanofibersof poly(styrene–butadiene-styrene) triblock-copolymersolution were electrospun [18]. Irregular microphaseseparation on the surface of electrospun fibers was ob-served. Although these structures were seen to improveon annealing, the domains remained largely disordered.Other works using poly(styrene-b-poly(4-vinylpyridine))[19], poly(styrene-b-dimethylsiloxane) [20], or poly(sty-rene-b-polyisoprene) [21] block copolymer nanofibers alsoreport the formation of irregular morphologies. Similarirregular micropahse separation was obtained in films
3852 D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856
and attributed to strong shear deformation [22,23], rapidsolvent evaporation [24] and surface effect [25]. Co-elec-trospinning [26,27], a modified electrospinning process,yielded poly(styrene-b-polyisoprene), which evolved viaannealing, forming a stacked lamellar-disc morphology,with each disc perpendicular to the fiber axis, and a long-range order of parallel alternating concentric cylinders [28].
In this study, we focus on thermoplastic polyurethaneelastomers (TPUs), which constitute an important subclassof BCPs; they are made from diisocyanates and polyols,containing an ester or ether backbone, which lead to line-arly segmented copolymers with alternating sequences of‘‘hard’’ (diisocyanate) and ‘‘soft’’ (diols, diacids) segments.The microphase-separated morphology [29] that resultsfrom the polarity difference of diol/diacid and diisocyanatesegments, leads to superior mechanical properties, such ashigher strength, elongation, and modulus, with respect toother elastomers [30]. In order to explore the thermalbehavior of TPU structures and their application asshape-memory [31,32], study of their thermo-mechanicalproperties is required. Towards this goal, we evaluatedthe strain-temperature behavior of TPU-based cast filmand electrospun fibers. When heating at 1 �C/min, fibermats and cast film strain was nearly fixed, but at �70 �C,fibers began to massively contract (40% axially) at0.013 �C�1, while cast films started to expand at0.003 �C�1. Increasing the heat rate to 5 �C/min yieldedsimilar responses, but shifted the starting temperature ofthe contraction and expansion of the fibers and the filmsrespectively, up to �85 �C (see Fig. 1).
Contraction of poly(ethylene terephthalate) (PET) filmsand melt-spun fibers in the range of 0.0006–0.006 �C�1,has already been reported for in several studies [33–35].Upon heating above the glass transition, the overall shrink-age process involves a rapid initial stage of rubber-likecontraction of the molecular network, associated with dis-orientation in the amorphous phase. Contraction attrib-uted to negative thermal expansion of typical aramidfibers (Kevlar) or polyethylene fibers, is generally a non-linear, temperature-dependent phenomenon, in the orderof 10�6 C�1. However, with increasing temperatures, thecoefficient of thermal expansion becomes more negative,
30 40 50 60 70 80 90
-30
-20
-10
0
10
Fiber mat
Stra
in (
%)
Temperature (°C)
1oC/min
5oC/min
Cast Film
Fig. 1. Strain vs. temperature response of TPU-based cast film andelectrospun fibers with diameters in the range of 500–800 nm. Theexperiment was carried in a dynamic mechanical analysis machine(DMA) in force control mode. (TPU–Krystalflex PE-399, Huntsman, USA).
primarily due to crystal contraction resulting from unit celldeformation [36,37].
The electrospun nanofiber contraction observed in ourexperiments, cannot be directly attributed to relaxation(disorientation) in the amorphous phase obtained whenthe stretched polymer is heated above its glass transition,since the glass transition temperature of the orientatedamorphous phase of the tested TPU is ��50 �C. Thus, atroom temperature, relaxation should be expected immedi-ately after fabrication. In our case, despite prolonged stor-age (several days) of the nanofibers under conditionsabove their glass transition temperature, unlike the abovesystems, contraction only began when reaching a certaintemperature, while below this temperature, fibers werepractically stable. Thus, a different mechanism governs thisphenomenon.
Based on the above mentioned principles, this workpursues a means of exploring the thermomechanical prop-erties of electrospun TPU nanofibers. Thermal and thermo-mechanical properties were examined by DynamicMechanical Analysis (DMA) and differential scanning calo-rimetry (DSC). Structure analysis was performed by wide-angle X-ray scattering (WAXS), at room and at elevatedtemperatures. It is hypothesized that the confined super-molecular nanoscale structure, developed in the electros-pinning process, causes a unique relaxation process of aneffective physical network in TPU nanofibers, which iscomposed of clusters of hard segments connected byamorphous tie chains.
2. Materials and methods
2.1. Materials
Krystalflex PE-399, a TPU, was purchased from Hunts-man. This TPU is a block copolymer composed from apoly(tetramethylene ether) glycol (PTMG)-type soft blockand an aliphatic diisocyanate. Analytical-grade dimethyl-formamide (DMF) and tetrahydrofuran (THF) were pur-chased from Frutarom Ltd, Israel.
2.2. Gel permeation chromatography (GPC) characterization
Molecular weight was obtained by gel permeation chro-matography (GPC). The sample was injected using aWaters 1525 Binary HPLC Pump equipped with an auto-sampler. GPC/SEC 7.8 � 300 mm columns (Varian) wereused. Molecular weight was determined by comparing toa universal calibration curve, obtained by using polysty-rene standards. The solvent used was HPLC-grade THF,purchased from BioLab, without stabilizers. The concentra-tion of the sample was 8 mg/mL, and the preparation of thesample, as well as running, was performed at room tem-perature. Using GPC, a molecular weight of Mn�50 [kDa]was obtained for the TPU.
2.3. Fibers and films preparation
A solution for electrospinning was prepared by dissolv-ing 1.2 g TPU in 8.8 g DMF and THF (7:3 (w/w)), to obtain a
12 14 16 18 20 22 24 26 28 30
0.00
0.05
0.10
0.15
0.20
Film 22oC Fiber 22oC Film 105oC Fiber 105oC
Nor
mal
ized
Int
ensi
ty
2θ
Fig. 2. WAXS profiles of the film and fiber mats, registered at 22 �C and105 �C. Profiles are normalized by the integrated intensity over the totalWAXS profile, demonstrating modification of the peak shapes withincreasing temperature.
Table 1Characteristics of the WAXS peaks.
Temperature, �C Width (deg) Skewness
Film 22 4.74 6.22105 5.32 7.54
Fiber 22 4.28 4.22105 9.70 0.62
MA
CRO
MO
LECU
LAR
NA
NO
TECH
NO
LOG
Y
D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856 3853
12% (w/w) solution. The TPU solutions were electrospun torespectively form fiber mats of oriented polymer mono-lithic nanofibers. The flow rate, controlled by a syringepump, of the TPU solution was constant (0.9 mL/h), withan electrostatic field of �1.2 kV/cm. A collector wheel,positioned at a distance of 12 cm from the spinneret (nee-dle 23G) and with a tangential velocity of 6 m/s, was usedto collect aligned fibers. The ambient temperature was21 ± 1 �C and humidity 48–51%. The films were preparedby 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 vac-uum for another 2 h.
2.4. Thermal characterization
DSC measurements were performed using Mettler Tole-do STARe DSC 1 system, equipped with a high-sensitivityHSS7 sensor. The instrument was calibrated for tempera-ture and heat flow, using indium and zinc standards. Mea-surements were performed at a heating rate of 2 K min�1
from �65 to 250 �C, under nitrogen purge. Masses of thesamples were 9 mg.
2.5. Thermo-mechanical characterization
Thermomechanical tests were executed by DynamicMechanical Analysis (TA-DMA Q800) on films and alignedfiber mats. Samples were 12 mm long, 3 mm wide and0.09–0.12 mm thick. Clamp tightening torque for fibermat and bulk film was 3 in-lb. For temperature sweepexperiments, samples were subjected to sinusoidal dis-placement with 0.1% strain, at a fixed frequency of 1 Hzfrom 120 to 130 �C, and a heating rate of 2 �C min�1. Fourspecimens were tested for each fiber mat composition.For strain-temperature behavior of TPU cast films and elec-trospun fibers, the experiments were carried in a force con-trol mode, with a target force of 0.003 N. The samples wereheated from room temperature at two heat rates (1 �C/minand 5 �C/min), up to 100 oC.
2.6. Wide-angle X-ray scattering (WAXS)
WAXS measurements were performed using a BrukerD8 Discover diffractometer. Cu Ka radiation (wavelengthof 0.1542 nm) was used at the applied voltage 40 kV, andcurrent 40 mA. All measurements were performed inreflection mode, using Bragg–Brentano geometry, with a1 mm slit and two Soller collimators applied on both sides.Considering very weak molecular orientation, as obtainedfrom a preliminary scan using 2-D detectors, we used ahighly sensitive Lynx Eye 1-D silicon strip detector. Theangular range of measurements, 2, was between 10 and30 deg, with a step of 0.005 deg and a time of data accumu-lation at particular angular point of 0.025 s. An Anton-PaarTTK 450 temperature chamber was used. The temperatureprogram consisted of heating at a rate of 10 K/min, whileholding the temperature constant during particular WAXSexposition. WAXS profiles were registered at 22 �C and105 �C.
3. Results and discussion
The morphology of the electrospun nanofibers wascharacterized by scanning electron microscopy (SEM).The diameter of the nanofibers ranged between 500 and800 nm, where the fiber mat porosity was q = 0.8 ± 0.07(q = 1 – (Nanofibers apparent density/Bulk density of rawTPU)). When analyzing the alignment of the fibers, it wasfound that the orientation distribution was such that onestandard deviation from the main director was at an angleof about 4�.
Analysis of radial WAXS profiles of the fiber mats andcast 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 offilm peaks underwent no significant change during heat-ing; only a slow increase in peak asymmetry was observed.At the same time, fiber mat peak asymmetry at 22 �C issubstantially reduced after heating; the shape of the fibermat peak at 105 �C was almost symmetric (see details inTable 1).
The general conclusion which can be drawn regardingthe physical origin of the asymmetric shape of TPU, is thatthe structure is inhomogeneous (e.g., spatial separation ofhard and soft segments with different polarity as a generalrule for TPU), which can result in the existence of variouscomponents within the amorphous halo, with different,most probable (characteristic) scattering distances. Sincethe scattering distances of the hard and the soft segmentsin TPU are dissimilar, we assume that the asymmetricpeaks consist of two components which can be associated
-25 0 25 50 75 100 125 150
Tm= 81.42 oC
ΔH= 5.85 Jg-1
Temperature / oC
0.2
Film
Fiber mat
endo
H
eat f
low
/
J(gK
)-1
Tm= 75.58 oC
ΔH= 3.41 Jg-1
Fig. 4. DSC heating runs for the fiber mat and the cast film samples. Thearrows indicating the thermal effect due to the abrupt change in the heatflow.
MA
CRO
MO
LECU
LAR
NA
NO
TECH
NO
LOG
Y
3854 D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856
with the local ordering of hard and soft segments, respec-tively. Thus, in order to further understand the physicalfactors leading to the asymmetric shape (observed inFig. 2), we decomposed the peaks (using Pearson VII func-tion), into two components (C1 and C2), see Fig. 3 andTable 2.
The two components (C1 and C2) maxima can roughlybe related to the intermolecular distances which are5.6 Å and 4.9 Å for C1 and C2 respectively. Since the dis-tance between the hard segments in TPU should be higherthan the distance between the soft segments [38], we con-jecture that component C1 belongs to the characteristic, ormost-probable scattering distance between hard segments,and that C2 belongs to the most-probable scattering dis-tance between soft segments.
When heating from 22 �C to 105 �C, major changes in thecontribution of the components were observed in the fibermat, namely, an increase in the content of the hard segmentscomponent (C1). At elevated temperature, the content of C1comprised �95% of the amorphous part, whereas the softsegments component (C2) comprised only �5% (seeFig. 3a). That means that in the fibers, heating to 105 �C af-fects mostly the soft segments, decreasing their orienta-tional ordering. In contrast, in the cast films, there is noessential change in the peak shape after heating (seeFig. 3b). When analyzing the change in the componentswidth of the fiber mat upon heating, an increase of �80%and �30% in the width of components C1 and C2, respec-tively, was found, indicating decrease in the scale of orderedregions. At the same time, in the cast films, only minorchanges in width of both components were observed.
0.00
0.05
0.10
0.15
0.20
Nor
mal
ized
Int
ensi
ty
Fiber 22oC C1 C2 Fiber 105oC C1 C2
(a)
12 14 16 18 20 22 24 26 28 30
Fig. 3. WAXS profiles and their decomposition into two components (C1 and C2),mat (a) and film (b).
Table 2Characteristics of the decomposed WAXS peaks.
Temperature, �C C1
2hmax (deg)
Film 22 19.23105 18.96
Fiber mat 22 19.32105 18.97
DSC heating runs (Fig. 4) clearly demonstrated endo-thermic peaks located at 75.58 �C and 81.42 �C, with lowenthalpy values of 3.41 and 5.85 J g�1, for the fiber matand the film samples, respectively. The endothermic char-acter of the peaks indicates a melting type of transition.Another thermal effect, located at �30 �C and 38 �C, forthe fiber mat and the film, respectively, precedes the endo-thermic peak. This effect is associated with the abruptchange in the course of the heat flow.
Analysis of storage (E0) and loss (E0 0) moduli and of lossfactor (tand), indicated three thermal transitions in both
12 14 16 18 20 22 24 26 28 30
0.00
0.05
0.10
0.15
0.20
Nor
mal
ized
Int
ensi
ty
Film 22oC C1 C2 Film 105oC C1 C2
(b)
registered at room 22 �C (solid lines) and at 105 �C (dashed lines), for fiber
Fig. 5. Storage and Loss moduli, and tand at 1 Hz, of TPU-based electrospun fiber mat and cast film.
ACR
OM
OLE
CULA
RN
AN
OTE
CHN
OLO
GY
D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856 3855
the fiber mat and the cast film (Fig. 5). The lowest temper-ature transition (��50 �C) is associated with the glasstransition, Tg of soft segments (not detected by DSC), thusat room temperature, the high mobility of polymer matrixallows the relaxation of its possible non-equilibrium state.The peak in tand (�25 �C), can be associated with destruc-tion of weak-ordered hard segments clusters. The last tran-sition starts at approximately 70 �C, and ends at 120 �C.This transition is apparently associated with destructionof ordered hard segment clusters. This peak coincides withthe endothermic peak, obtained by the DSC, as describedabove. Note that after this thermal transition (at 120 �C),the storage modulus almost vanishes, indicating thedestruction of the effective physical network. The transi-tion (70–120 �C) corresponds to thermal transition, previ-ously observed in high crystallinity segmentedpolyurethanes [39–44]. In the case of low concentrationsof hard segments (less than 40%), Van Bogart et al. [45]found an endotherm peak at �50–60 �C, associated withthe destruction of ordered hard segment clusters.
Suggested contraction mechanism: The above experi-mental results demonstrate that the changes in the poly-mer structure of films and fibers upon heating differ. Infibers, the structural changes can be attributed to the
-40
-30
-20
-10
0
10
Fiber MatFilm
Time [min]
Stra
in [
%]
30
40
50
60
70
80
90
Tem
pera
ture
[o C]
0.1 1 10 100 1000
(a)
Fig. 6. (a) The strain response vs. time of fiber mat and cast film, under isothermand film continuously heated from room temperature until initiation of materia
relaxation of the non-equilibrium state of the polymer ma-trix (stress and orientation) [46]. The dramatic decrease ofthe portion of soft segment clusters demonstrates that thedisordering (disorientation) of soft segments is dominant.Just this disorientation of stretched portion of macromole-cules should result in local shortening of polymer matrixand therefore in a shortening of the fibers as a whole, aswas observed in the experiment. Unlike the fibers, in thecast films, the disordering of both hard and soft segmentclusters is of the same order. Apparently, the effectivephysical network of hard segment clusters, connected bytie chains, is broken upon heating, resulting in polymerplasticity and consequently, to noticeable expansion ofthe film. (Note that this effective physical network pre-serves the solid-like state of polymer matrix above Tg.)
In order to better understand the destruction process ofthe effective physical network, we focused on the highestthermal transition temperature (see Fig. 5). Due to thebroad range of transition temperatures (�70–120 �C), wecan assume that the hard segment clusters are graduallydestroyed. The destroyed clusters allow local mobility ofamorphous soft segments, whereas the intact clustersmaintain the effective physical network. As temperatureincreases, the proportion of the intact clusters decreases,
30 45 60 75 90 105 120 135
-40
-30
-20
-10
0
10
200
225
Stra
in [
%]
Temperature [ oC]
Fiber MatFilm
(b)
al conditions at 90 �C. (b) The strain response vs. temperature of fiber matl flow.
M
MA
CRO
MO
LECU
LAR
NA
NO
TECH
NO
LOG
Y
3856 D. Alhazov et al. / European Polymer Journal 49 (2013) 3851–3856
and the physical network will eventually be completelydestroyed.
This hypothesis can be verified by thermo-mechanicaltests. Unlike the previous experiment (Fig. 1), the abovethermal transition begins and, when reaching 90 �C iso-thermal condition is maintained for more than 15 h(Fig. 6a). As expected, the film sample expanded, whilethe fiber mat sample began to contract at �70 �C. Whenreaching 90 �C, the contraction continued at the same rateand persisted for approximately 100 min. Then, the con-traction stopped and the strain was kept nearly constantat ��40%, for approximately 200 min. At that moment,the fiber mat began to slowly expand. In another experi-ment (similar to the one presented in Fig. 1), the sampleswere heated up to 130 �C (Fig. 6b). Up until 105 �C, boththe fiber mat and the film underwent contraction andexpansion, as shown before. However, when heating above105 �C, the strain direction of the fiber mats changed andthey began to expand.
These results support our hypothesis that the fiber matcontraction behavior is controlled by the destruction ofhard segment clusters. The proportion of the destroyedclusters increases upon heating in range of 70–105 �C. Attemperatures above �105 �C, all clusters in both fibermat and cast film are destroyed, and the material startsto intensively expand and, finally, to flow.
4. Conclusions
Although the experimental observations demonstratethat the ordering level in both fiber mat and cast film issimilar, their response upon heating differs dramatically.The unexpected fiber contraction is attributed to thenon-equilibrium microstructure (the spatial distributionof ordered regions), and more specifically to relaxation ofstretched amorphous phase formed in electrospun nanofi-bers. Thorough analysis of the experimental results (WAXS,DSC, and DMA) indicates that in block co-polymer systemswith low glass transition temperature such structure relax-ation becomes possible only when a portion of the hardsegment clusters, forming an effective physical network,is destroyed upon heating. The details of this microstruc-ture, as well as its effect on the behavior of the fibers arepart of ongoing research. The studied phenomenon canhelp in further understanding the physical reasons forthe unique behavior of the electrospun nanofibers.
Acknowledgments
We gratefully acknowledge the financial support of theRBNI–Russell Berrie Nanotechnology Institute, the IsraelScience Foundation (ISF Grant No. 770/11), and the DFGwithin the GIP project.
References
[1] Reneker D, Yarin A, Zussman E, Xu H. Adv Appl Mech 2007;41:43.[2] Yarin AL, Koombhongse S, Reneker DH. J Appl Phys 2001;89:3018.[3] Reneker DH, Yarin AL. Polymer 2008;49:2387.[4] Shin YM, Hohman MM, Brenner MP, Rutledge GC. Polymer
2001;42:9955.[5] Arinstein A, Zussman E. J Polym Sci Pol Phys 2011;49:691.[6] Curgul S, Van Vliet KJ, Rutledge GC. Macromolecules 2007;40:8483.[7] Wong SC, Baji A, Leng SW. Polymer 2008;49:4713.[8] Arinstein A, Burman M, Gendelman O, Zussman E. Nat Nano
2007;2:59.[9] Arinstein A, Liu Y, Rafailovich M, Zussman E. EPL 2011;93:46001.
[10] Wang W, Barber AH. Nanotechnology 2010;21:225701.[11] Liao CC, Wang CC, Shih KC, Chen CY. Eur Polym J 2011;47:911.[12] Fasolka MJ, Mayes AM. Ann Rev Mater Res 2001;31:323.[13] Lodge TP. Macromol Chem Phys 2003;204:265.[14] Bockstaller MR, Mickiewicz RA, Thomas EL. Adv Mater
2005;17:1331.[15] Ruzette AV, Leibler L. Nat Mater 2005;4:19.[16] Park C, Yoon J, Thomas EL. Polymer 2003;44:6725.[17] Matsumoto H, Ishiguro T, Konosu Y, Minagawa M, Tanioka A, Richau
K, et al. Eur Polym J 1866;2012:48.[18] Fong H, Reneker DH. J Polym Sci Pol Phys 1999;37:3488.[19] Ruotsalainen T, Turku J, Heikkila P, Ruokolainen J, Nykanen A,
Laitinen T, et al. Adv Mater 2005;17:1048.[20] Ma ML, Hill RM, Lowery JL, Fridrikh SV, Rutledge GC. Langmuir
1985;18:1294.[23] Fairclough JPA, Salou CLO, Ryan AJ, Hamley IW, Daniel C, Helsby WI,
et al. Polymer 2000;41:2577.[24] Kim G, Libera M. Macromolecules 1998;31:2670.[25] Anastasiadis SH, Russell TP, Satija SK, Majkrzak CF. Phys Rev Lett
1852;1989:62.[26] Sun ZC, Zussman E, Yarin AL, Wendorff JH, Greiner A. Adv Mater
2003;15:1929.[27] Reznik SN, Yarin AL, Zussman E, Berkovici L. Phys Fluids
2006;18:062101.[28] Kalra V, Mendez S, Lee JH, Nguyen H, Marquez M, Joo YL. Adv Mater
2006;18:3299.[29] Leibler L. Macromolecules 1980;13:1602.[30] Curgul S, Yilgor I, Yigor E, Erman B, Cakmak M. Macromolecules
2004;37:8676.[31] Lee BS, Chun BC, Chung YC, Sul KI, Cho JW. Macromolecules
2001;34:6431.[32] Hu JL, Zhu Y, Huang HH, Lu J. Prog Polym Sci 2012;37:1720.[33] Gupta VB, Radhakrishnan J, Sett SK. Polymer 1993;34:3814.[34] Lim JY, Kim SY. J Polym Sci Pol Phys 2001;39:964.[35] Pinnock PR, Ward IM. Brit J Appl Phys 1966;17:575.[36] Rojstaczer S, Cohn D, Marom G. J Mater Sci Lett 1985;4:1233.[37] Kobayash Y, Keller A. Polymer 1970;11:114.[38] Liff SM, Kumar N, McKinley GH. Nat Mater 2007;6:76.[39] Koberstein JT, Russell TP. Macromolecules 1986;19:714.[40] Koberstein JT, Galambos AF. Macromolecules 1992;25:5618.[41] Saiani A, Daunch WA, Verbeke H, Leenslag JW, Higgins JS.
