ElectrospunNanoﬁbersofPoly(vinylalcohol

RESEARCHARTICLE

Copyright © 2008 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofBiobased Materials and Bioenergy

Vol. 2, 1–12, 2008

Electrospun Nanofibers of Poly(vinyl alcohol)Reinforced with Cellulose Nanofibrils

Eliton S. Medeiros1�2, Luiz H. C. Mattoso2, Edson N. Ito2, Kay S. Gregorski1, George H. Robertson1,Richard D. Offeman1, Delilah F. Wood1, William J. Orts1, and Syed H. Imam1�∗

1United States Department of Agriculture, Bioproduct Chemistry and Engineering Unit, Western Regional Research Center,800 Buchanan St., Albany, CA 94710, USA

2Laboratório Nacional de Nanotecnologia Aplicada ao Agronegócio, Embrapa Instrumentação Agropecuária,Rua XV de Novembro, 1452, São Carlos-SP, 13560-970, Brazil

In this work, nanofibers of poly(vinyl alcohol) reinforced with cellulose nanofibrils were producedby electrospinning. The effects of applied voltage, polymer concentration and injection rate, tip-to-collector distance, rotation speed of the collector, and relative humidity on morphology wereinvestigated by scanning electron microscopy. The reinforcing capability of cellulose nanofibrils wasinvestigated by tensile tests. Thermogravimetry, transmission electron microscopy and Fourier trans-form infrared spectroscopy-attenuated total reflectance analyses were also carried out in order tocharacterize the presence, orientation and reinforcing effect of the cellulose nanofibrils. Scanningelectron microscopy results showed that fiber structure is strongly affected by the electrospinningconditions. Thinner fibers are favored by decreasing viscosity, polymer injection rate, high rotationspeed and high relative humidity. Whereas increasing the applied voltage favors the formation ofbeaded fibers. The reinforced composites had a 2.4-fold increase in their mechanical properties byaddition of only 6.6 wt% of cellulose nanofibrils without major changes in elongation at break.

Keywords: Electrospinning, Cellulose Nanofibrils, Poly(vinyl alcohol), Nanofibers.

1. INTRODUCTION

Fibers derived from wood and agriculture biomass havereceived increasing attention as reinforcement agents inplastics in a wide range of composite applications.1–10

The advantages of reinforcing polymers with nanoscaleparticles have been clearly shown for clay nanocompos-ites and other inorganic fillers in which the addition ofnano-scale clay particles improves dimensional stability,stiffness, and higher heat distortion temperature.3 Naturalcellulose nanofibrils (CnF) can act in a fashion similarto clay nanocomposites in reinforcing polymers, althoughthey exhibit both advantages and disadvantages relativeto clay.1�2 Cellulose is the most common organic poly-mer, representing about 1�5×1012 tons of the total annualbiomass production11–13 with a range of properties, includ-ing biodegradation. Processing and chemical modificationtechniques of cellulose are well-understood and relativelyeasy since they are well-established in the wood, paper andfood industries. In contrast to the “playing card” structureof most clay particulates, typical cellulosic microfibrils are

∗Author to whom correspondence should be addressed.Email: [email protected]

long crystalline “needles” ranging in size from 1–5,000 nmin width, with a wide range of aspect ratios ranging from1–50.2 This aspect ratio varies depending on cellulosesources and the isolation procedures used.

The promise behind cellulose-derived nanocompositeslies in the fact that the axial Young’s modulus of the basiccellulose crystalline nanofibril has been reported as highas 137 GPa,4�5 which means it rivals steel in strength andstiffness. Despite this, cellulose nanofibrils are only usedin limited applications due their incompatibility with mostcommercial polymers. No feasible method has been foundfor utilizing this remarkable stiffness specifically becausesurface properties have not been optimized toward idealcompatibility between the cellulose nanofibrils and theirpolymer matrices.

Electrospinning has been extensively explored as a sim-ple and versatile technique to produce micro and nano-fibers of polymers because it provides a potential wayto fabricate continuous nanofibers with different structuraldesigns. A typical electrospinning setup is comprised of areservoir for polymer solution, pump, capillary spinneret,grounded collector, and high voltage power supply. Theelectrospinning process consists of applying a strongelectrostatic field supplied by the high voltage source to

J. Biobased Materials and Bioenergy 2008, Vol. 2, No. 3 1556-6560/2008/2/001/012 doi:10.1166/jbmb.2008.411 1

RESEARCHARTICLE

Electrospun Nanofibers of Poly(vinyl alcohol) Reinforced with Cellulose Nanofibrils Medeiros et al.

polymer solution as it exits the spinneret connected toits reservoir. Under the influence of the electrostatic field,a pendant droplet of the polymer solution at the capillarytip is deformed into a conical shape. When the surfacetension is overcome by the electrostatic forces generatedby the electric potential a fine charged jet is ejected andmoves toward the grounded collector, meanwhile the sol-vent rapidly evaporates, and polymer fibers accumulate onthe collector.14�15

Control of process parameters, such as rate of polymerinjection, distance between the capillary and collector,polymer-solvent combination, and polymer concentration,allows the production of fibers with controllable proper-ties and diameters ranging from tens of microns down toa few tens of nanometers. A variety of materials such asengineering plastics, copolymers, polymer blends, biopoly-mers, and conducting polymers have been successfullyelectrospun to produce uniform fibrous mats.14–17 The useof electrospun nanofibers includes filtration membranes,18

drug release systems,15 wound dressing15 and tissueengineering19�20 to chemical and biological protectiveclothing,20�21 sensors,22�23 and composites.15 However, ithas been reported in the literature that electrospun fibrousmats of many polymers and biopolymers possess relativelylow dimensional stability and mechanical strength,24–27

which limits their use in some of the abovementionedapplications.

In order to improve the mechanical strength of electro-spun materials, a common approach used is the chemicalcross-linking of the matrix,27 which depends not only onits reactivity but also on its thermal stability and the bio-compatibility of the cross-linking agent, therefore limitingthe technique to a number of materials. Another approachthat can be exploited to improve the mechanical strengthis the use of reinforcing agents, for forming engineeringfibrous composites. In this sense, cellulose nanofibrils area good candidate due to its biocompatibility11 and highstrength,7�17�28–30 which enables the improvement of thelow tensile properties of electrospun fibers while maintain-ing their biodegradability and biocompatibility. Interest-ingly, cellulose nanofibrils can be coated with conductivepolymers such as polyaniline31�33 providing an excellentmicrowave and radar absorbing material,34�35 thus openingup several new uses for electrospun materials in militaryand electronic circuit board applications.

In this work, we explore the possibility of improv-ing the tensile properties of electrospun nanofibers ofpoly(vinyl alcohol), PVA, by adding cellulose nanofibrilsas a reinforcing agent. The effects of process variables(applied voltage, polymer concentration and injectionrate, tip-to-collector distance, rotation speed of the col-lector, and relative humidity) on morphology of theelectrospun mats were investigated by scanning elec-tron microscopy (SEM). The reinforcing capability ofcellulose nanofibrils was investigated by tensile tests.Thermogravimetry (TG), transmission electron microscopy

(TEM), and Fourier transform infrared spectroscopy-attenuated total reflectance (FTIR-ATR) analyses were alsocarried out to characterize the presence and orientation,and reinforcing effect of the cellulose nanofibrils.

2. EXPERIMENTAL DETAILS

2.1. Extraction of Cellulose Nanofibrils

The extraction of cellulose nanofibrils was carried outaccording to methodology described by Favier et al.36

Briefly, 10% w/v short cellulose fiber (Whatman CF11)was stirred vigorously with 60 wt% sulfuric acid pre-heated and maintained at 60 �C, for 45 minutes. The mix-ture was then quenched in cold water and stirred. Theresulting dispersion was centrifuged, decanted and washedwith water continuously until the pH was around 1.5 andthe dispersion would not spin down to form a pellet.The resulting solution was water dialyzed in dialysis tub-ing (Spectrapor 2; MWCO 12,000–14,000; 45 mm flatwidth) for 4–5 days until the pH of the solution wasbetween 5.5 and 6.5. Cellulose nanofibrils were recoveredas a dispersion of approximately 1.16 g/100 mL waterfrom dialysis. These nanofibrils have average diameter(d) of 38±13 nm, length (l) of 360±60 nm, and aspectratio (l/d) of 10±4 nm, as determined by atomic forcemicroscopy (AFM) measurements.31–33

2.2. Preparation of PVA/CnF Solution forElectrospinning

Cellulose nanofibrils were incorporated into PVA (Sigma-Aldrich, MW = 30,000–70,000) by adding a previouslymeasured amount of PVA powder to nanofibril aqueoussolution at 80 �C under stirring until its complete disso-lution. Pure PVA solutions were prepared by dissolvingPVA powder directly in deionized water. A typical PVA(14 wt%)/CnF (1 wt%) solution was prepared by dissolv-ing 2.8 g of PVA in 17.2 g (∼17.2 mL) of aqueous disper-sion of CnF (1.16 wt%) to produce a 6.6 wt% suspensionof CnF in PVA on a dry basis.

2.3. Electrospinning

The electrospinning setup used in this study consisted ofa 10 mL plastic syringe, 21G needle (i.d.: 0.84 mm),a grounded collector with controllable rotation speed(stainless steel barrel with 10 cm in diameter attachedto a overhead stirrer, Model RW20, IKA Works, Inc.,USA), a hypodermic syringe pump (KD Scientific, USA),and a high-voltage supply (Series FC, Glassman HighVoltage, Inc., USA), which can generate positive directcurrent voltages up to 99 kV. The needle was con-nected to the high-voltage source, positioned at a fixeddistance from the grounded collector (tip-to-collectordistance–TCD), and the polymer solution was injected

2 J. Biobased Materials and Bioenergy 2, 1–12, 2008

RESEARCHARTICLE

Medeiros et al. Electrospun Nanofibers of Poly(vinyl alcohol) Reinforced with Cellulose Nanofibrils

5

4

TCD

1

2

3

Scheme 1. Depiction of the electrospinning setup used for fiber produc-tion. (1) speed control system; (2) rotating collector; (3) needle; (4) injec-tion pump with syringe, and (5) high voltage power source.

at a constant rate. Scheme 1 depicts the experimentalsetup used. The effects of process variables on mor-phology were investigated by changing applied voltage(8–30 kV), polymer concentration (10–14 wt%), injec-tion rate (2–8 �L/min) tip-to-collector distance (2–20 cm),relative humidity (25–80% RH). RH-controlled electro-spinning experiments were carried out inside an acrylicbox using saturated solutions of lithium chloride (LiCl,25–30% RH), sodium iodide (NaI, 38–40% RH), sodiumchloride (NaCl, 48–52% RH), and potassium chloride(KCl, 75–80% RH). A small fan was used to keep RH uni-form throughout the box while being monitored with theaid of a temperature and humidity data logger (TM 121,Dickson, USA).

2.4. Scanning Electron Microscopy (SEM)Characterization

Electrospun nanofibers that had been deposited for 15 minon aluminum foils wrapped on the rotating collector weresputter coated with gold for 45 s and analyzed for surfacecharacterization using a Hitachi Scanning Electron Micro-scope (Model S4700, Hitachi High-Technologies, Japan)at a voltage of 2 kV.

At least 50 fibers were counted in each image in orderto evaluate fiber diameter distribution. Statistical analyses(t-Student) at the significance level of 0.05 was used toverify if the differences on morphology because of thechanges in electrospinning parameters (voltage, rotation ofthe collector and polymer injection rate) were significant.

2.5. Transmission Electron Microscopy (TEM)Characterization

In order to characterize the electrospun composites withTEM, cellulose nanofibrils were stained with polyaniline

by in situ polymerization followed by electrospinning aspreviously described.33 Electrospun nanofibers of PVAcontaining stained cellulose nanofibrils were analyzed in aPhilips transmission electron microscope (Model CM 120,Philips, The Netherlands), operating at 120 kV.

2.6. Thermogravimetry

Thermogravimetric measurements were performed ona High-Resolution Thermogravimetric Analyzer (ModelTGA 2950, TA Instruments, USA). The operating condi-tions were: 20 mL/min nitrogen flow, 25 to 600 �C tem-perature range and 10 �C/min heating rate.

2.7. Fourier Transform InfraredSpectroscopy-Attenuated TotalReflectance (FTIR-ATR) Analyses

FTIR-ATR spectra of CnF (cast films), and PVA andPVA/CnF electrospun films deposited directly on KBrcrystals, were obtained using a Perkin Elmer FTIRspectrometer (Model System 2000, Perkin Elmer, USA)equipped with a DTGS (deuterated triglycine sulfate)detector and the ASI DuraSamplIR™ Universal DiComp™

attachment from ASI SensIR Technologies. The calibratedpressure applicator in this sampler was used to ensurereproducible contact between the sample and the diamondcell. At an incident angle of 45� relative to the ZnSe crys-tal the penetration depth varied from 0.5 to 3.0 �m in therange of 4,000 to 650 cm−1. A background spectrum inair was obtained for each sample spectrum. All spectrawere collected at 4 cm−1 resolution and 50 scans for eachsample.

The difference spectrum was obtained by subtracting theexcess PVA (94%) contribution from the PVA/CnF spec-trum taking into account the relative proportions of eachmaterial used. The residual spectrum contains contributionfrom both CnF (1034, 1059, 1110, 1162 cm−1) and PVA(1100 and 1046 cm−1).

CnF films were obtained by allowing CnF solution todry on a flat surface so that the film could be peeled offand analyzed by FTIR.

2.8. Tensile Tests

The tensile properties of the electrospun nanocompositeswere measured with an Instron Universal Materials TestingMachine (Model 5500R, Instron Corp., USA) at ambi-ent conditions. Tensile properties (Young’s modulus, ten-sile strength and elongation at break) were determinedusing samples cut from electrospun mats, 60 mm in length,10 mm in width and ca. 0.20 mm in thickness, at acrosshead speed of 10 mm/min. In order to prevent thegrips from direct contact with the mats, cushion tabs wereused. At least 10 samples were tested for both PVA andPVA/CnF composites.

J. Biobased Materials and Bioenergy 2, 1–12, 2008 3

RESEARCHARTICLE


3. RESULTS AND DISCUSSION

Visual inspection helped determine the critical voltage(Vc) for a droplet formed at the needle tip as it was pulledby the electric potential difference between the needle andthe collector. This value increased as both polymer con-centration and tip-to-collector distance (TCD) increased asshown in Figure 1.

For tip-to-collector distances above 12 cm the criticalvoltage increased more rapidly at higher polymer concen-trations. For example, when PVA concentration increasedfrom 10 to 20 wt%, maintaining TCD fixed at 11.5 cm, thecritical voltage was increased from 14.5 to 33 kV withoutproducing good quality electrospun mats, i.e., those hav-ing uniform fiber diameters and without bead formation(results not shown). This behavior is due to the increasein viscosity which requires higher electrical field strengthto stretch the droplets in order to form fibers. If the vis-cosity is too high the electrical field may not be ableto pull the droplets formed at the needle tip or the fieldbarely stretches them so that large beads prevail overfibers.14�15

On the other hand, when viscosity is low the relax-ation time (time necessary for stretched polymeric chainsto assume a coiled conformation) decreases so that thestructure is composed of beads linked by fibers (beads-on-a-string structures) instead of pure fibers. At distancesshorter than 11.5 cm there is not enough time for completesolvent (water) evaporation, and the electrospun mats arecomposed of an interconnected network of nanofibers and

(a) (b)

(c) (d)

Fig. 2. Effect of injection rate on morphology of PVA/CnF electrospun nanofibers using a voltage of 21 kV and a tip-to-collector distance (TCD) of11.5 cm. (a) 2 �L/min; (b) 4 �L/min; (c) 6 �L/min, and (d) 8 �L/min.

0 4 8 12 16 205

10

15

20

25

Vc

(kV

)

TCD (cm)

10% PVA/CnF

14% PVA/CnF

Fig. 1. Critical voltage for electrospinning (Vc) as a function of tip-to-collector distance (TCD) and polymer concentration.

beads. Therefore, based on this study, a TCD of 11.5 cmand 14 wt% PVA/CnF were chosen for further studies.

Figure 2 shows SEM images of the electrospun nano-fibers obtained to investigate the effect of injection rate onmorphology. It can be observed that the number of beadsand fiber irregularity increased as injection rate increasedfrom 2 to 8 �L/min, in accordance with results reportedin the literature.15�37 During electrospinning, a polymericsolution is forced through a syringe pump to form adroplet at the needle tip, and the high voltage potentialinduces free charges into the solution. These charged ionsmove in response to the applied electric field toward the


RESEARCHARTICLE


(a) (b)

(c) (d)

(e)

Fig. 3. Effect of voltage on morphology of PVA/CnF electrospun nanofibers using an injection rate of 2 �L/min and a tip-to-collector distance (TCD)of 11.5 cm. (a) 12 kV; (b) 15 kV; (c) 18 kV; (d) 21 kV and (e) 25 kV.

electrode of opposite polarity, consequently transferringtensile forces to the polymer liquid.38 At a fixed TCD andkeeping the applied voltage constant, higher injection ratessupply more polymer solution to be stretched by the elec-tric field, thus provoking an unbalance between the volumeof polymer accumulated and the electrostatic charges nec-essary to overcome surface tension and viscoelastic forcesof the droplet. As a result, charge density is not enoughto stretch these droplets completely and break them intofibers, which causes large bead formation and some thinfibers instead of completely fibrous mats.

The effect of applied voltage on morphology ofPVA/CnF electrospun nanofibers was carried out by fixingboth the TCD at 11.5 cm and polymer injection rate at2 �L/min and then changing the applied voltage from 12to 25 kV. SEM micrographs of the nanofibers electrospunat 12, 15, 18, 21 and 25 kV are shown in Figure 3 and thedistribution of fiber diameters using at least 50 fibers forstatistical analyses is plotted in Figure 4.

Results from Figure 3 show that the number of beadsincreased significantly as applied voltage increased, whileFigure 4 shows fiber diameter as a function of appliedvoltages, which is smallest at both low and high voltage.At the lowest voltage, the electrical field is presumably notstrong enough to stretch the fibers, whereas at the highestvoltage the intense electric potential difference pulls thedroplets, perhaps in a non-uniform way, thus forming boththin and thick fibers, as well as beads. This non-uniformitybroadens the distribution of fiber diameters. At intermedi-ate values, between 15 and 21 kV, fiber diameter reachesan optimum value of 80±10 nm, i.e., a narrow fiber diam-eter distribution with no beads.

When a polymer droplet is formed at the needle tip, itdeforms under the applied electric field forming a conicprojection, known as Taylor’s cone, eventually becoming aflying liquid jet that undergoes a chaotic motion or bendinginstability and is field directed toward the collector. As thejet travels through the atmosphere, the solvent evaporates,


RESEARCHARTICLE


40 60 80 100 120 140 160 180 200 220 240 2600.0

0.1

0.2

0.3

0.4

0.5

0.6

Freq

uenc

y

Fiber diameter (nm)

40 60 80 100 120 140 160 180 200 220 240 2600.0

0.1

0.2

0.3

0.4

0.5

0.6

Freq

uenc

y

Fiber diameter (nm)

(a) (b)

40 60 80 100 120 140 160 180 200 220 240 2600.0

0.1

0.2

0.3

0.4

0.5

0.6

Freq

uenc

y

Freq

uenc

y

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Freq

uenc

y

Fiber diameter (nm)

40 60 80 100 120 140 160 180 200 220 240 260

Fiber diameter (nm)

40 60 80 100 120 140 160 180 200 220 240 2600.0

0.1

0.2

0.3

0.4

0.5

0.6

Fiber diameter (nm)

(c)(d)

(e)

Fig. 4. Effect of voltage on distribution of fiber diameter of PVA/CnF electrospun nanofibers. (a) 12, (b) 15, (c) 18, (d) 21 and (e) 25 kV.

leaving behind a dry fiber on the collecting drum.14 If theapplied voltage is higher than the critical value for elec-trospinning, the bending stability and time of flight of adroplet from the needle tip to the target decreases. Fur-thermore, whereas surface tension acts to reduce the sur-face area by forming spheres, electrostatic tension tends to

result in lower charge density by increasing surface area.39

The balance between these factors as well as others suchas solvent volatility and polymer viscosity govern the elec-trospinning process. This explains why there is an increasein both number and size of beads when Vc is increased,therefore producing non-uniform morphology.


RESEARCHARTICLE


(a) (b)

(c)

Fig. 5. Effect of rotation speed of the collector on morphology of electrospun fibers. (a) 200 rpm; (b) 600 rpm; and (c) 1000 rpm. Experimentalconditions: TCD = 11�5 cm; V = 15 kV; PVA (14 wt%) and CnF (6.6 wt%).

The effect of rotation speed of the collector on morphol-ogy of PVA/CnF electrospun nanofibers was observed byfixing TCD at 11.5 cm, polymer injection rate at 5 �L/min,and applied voltage at 15 kV, and then changing rotationspeed from 200 to 1,000 rpm. Figure 5 shows a series ofSEM micrographs of the nanofibers. The fiber diameterdistribution, which is more representative statistically thanthe simple mean and standard deviation, was calculated foreach speed and plotted in Figure 6.

It can be seen from Figures 5 and 6 that the increasein collector rotation resulted in more uniform and thin-ner fibers, potentially due to the higher stretching levelimposed on them. Statistical analyses showed that theincrease in this variable provoked a change in fiber diam-eter which at the significance level of 0.05. At speedsover 1,000 rpm the overhead stirrer vibrated excessively,which made further experiments difficult. This shows thata rotating target can be used, not only to control fiberalignment, but also to control fiber diameter; the ultimatechoice is dependent on the end use of the electrospun mats.Despite being less efficient than the methods reported byPan et al.16 and Li et al.40 in which nearly 100% uni-and bidirectionally aligned nanofibers can be obtained,this electrospinning setup does not need to be modified inorder to impart a certain degree of orientation to the mats.Indeed, aligned electrospun nanofibers of polyacrylonitrileby high speed rotating method, similar to the setup used inthis work but using higher speed, has been reported as anefficient method to prepare unidirectionally aligned carbonprecursor fibers.41

The effect of RH on fiber morphology of electrospunPVA/CnF nanofibers studied by controlling moisture insidethe electrospinning box with saturated solutions is shownin Figure 7 for SEM pictures taken from electrospun mats.The fiber diameter distribution is plotted in Figure 8. It canbe observed that both fiber diameter and packing densityare strongly affected by changes in RH. Whereas lowerRH favored an open structure (low density of fibers perunit area), higher RH values favored the formation of amore closed structure with thinner diameters.

When RH increases, the conductivity of the mediumincreases because of both polarization of the air due tothe high applied electric field with ozone formation (thisphenomenon can be detected easily by the ozone odor inthe electrospinning box) and because of formation of con-ductive “paths” due to water dipoles across the workingdistance. This reduces the stretch on the fibers, but alsosplits them as a consequence of the formation of secondaryjets that ultimately cause a reduction in fiber diameter.14�15

A similar behavior was reported by Son et al.42 whenstudying the effect of pH of PVA solutions on the morphol-ogy and diameter of electrospun nanofibers. The authorsreported that fiber morphology is influenced by pH due todensity and size of the ionic species that ultimately influ-ence the mobility of the charged polymer solution underthe external applied electrical field. Furthermore, the rateof solvent evaporation is reduced in high RH environ-ments, i.e., one of the driving forces for solvent volatiliza-tion is the difference between solvent vapor pressure andits partial pressure in the chamber.


RESEARCHARTICLE


50 60 70 80 90 100 110 120 1300.0

0.1

0.2

0.3

0.4Fr

eque

ncy

0.0

0.1

0.2

0.3

0.4

Freq

uenc

y

Fiber diameter (nm)

50 60 70 80 90 100 110 120 130

Fiber diameter (nm)

50 60 70 80 90 100 110 120 130

Fiber diameter (nm)

(a)

0.0

0.1

0.2

0.3

0.4

Freq

uenc

y

(c)

(b)

Fig. 6. Effect of rotation speed of the collector on fiber diameter distribution of PVA/CnF electrospun nanofibers. (a) 200, (b) 600, and (c) 1,000 rpm.

When RH is low the rate of solvent evaporation canbe so high that it generates a rapid increase in poly-mer viscosity and the polymer chains are not able toundergo voltage-induced stretching, therefore increasingfiber diameter. This effect, which is also likely inducedby the static buildup on the acrylic box, was corroboratedby excessive clogging at the tip of the needle during theelectrospinning at 25–30% RH and by the formation ofspider web-like structures throughout the box, which arecharacteristic features of high rate of solvent evaporation.

The average fiber diameter of the electrospun PVA/CnFcomposites ranged from 30 to 200 nm, which lies in therange reported in the literature for electrospun fibers ofpure PVA, i.e., from 87 (Ref. [43]) to 2,000 nm,44 but mostof the works reported a narrower distribution between 200and 800 nm.45�46

In order to characterize the presence, orientation andreinforcing effect of the cellulose nanofibers, TGA, TEMand FTIR analyses and tensile tests were carried out. TGA

analyses (Fig. 9) show that the thermal events associ-ated to PVA/cellulose composites are intermediate betweenthe pure components; results which are corroborated byTEM analyses (Fig. 10) showing cellulose nanofibers inthe core and sticking out from electrospun PVA nanofibers.Atomic force microscopy (AFM) measurements of thesenanofibrils31–33 showed that they have an average diameter(d) of 38±13 nm, length (l) of 360±60 nm, and an aspectratio (l/d) of 10±4 nm, therefore in agreement with thedimensions of the cellulose nanofibrils in Figure 10.

Figure 11 shows the FTIR spectra for cellulose nanofib-rils (CnF), electrospun PVA and PVA/CnF, and the differ-ence spectrum between PVA and CnF (PVA-CnF). Theseanalyses provided clear evidence of the presence of cel-lulose nanofibrils in the PVA/CnF electrospun mats ascan be observed by the absorption peaks such as thoseat 1034 and 1059 cm−1 assigned to C O stretchingof cellulose.47 Moreover, the difference spectrum calcu-lated by taking into account the relative amounts of each


RESEARCHARTICLE


(a) (b)

(c) (d)

Fig. 7. SEM micrographs of PVA nanofibers electrospun at: (a) 25–30 (b) 38–40, (c) 48–52 and (d) 75–80% RH.

20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 2000.0

0.1

0.2

0.3

0.4

0.5

0.6

Fre

quen

cy

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Fre

quen

cy

Fiber diameter (nm) Fiber diameter (nm)

20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200

Fiber diameter (nm) Fiber diameter (nm)

(a) (b)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Fre

quen

cy

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Fre

quen

cy

(c) (d)

Fig. 8. Effect of relative humidity (RH) inside the electrospinning box on fiber diameter distribution of PVA/CnF electrospun nanofibers. (a) 25–30,(b) 38–40, (c) 48–52 and (d) 75–80%.


RESEARCHARTICLE


0 100 200 300 400 500 6000

20

40

60

80

100M

ass

(%)

Temperature (ºC)

PVAPVA/CnFCnF

Fig. 9. TG curves of electrospun nanofibers of PVA and PVA reinforcedwith cellulose nanowhiskers, and cellulose nanowhiskers.

200 nm

200 nm

CnF

(b)

(a)

CnF

CnF

Fig. 10. TEM micrographs of electrospun PVA/CnF nanofibers.(a) Detail of a stained cellulose nanofiber (dark areas) protruding froman electrospun PVA nanofiber, and (b) in the core of electrospun PVAnanofibers.

4000 3000 2000 1000

PVA-CnF1162

Wavenumber (cm–1)A

bsor

banc

e (a

.u.)

1093

1030PVA

PVA/CnF

CnF

1054

1034

10591093

1034

1059

1146

Fig. 11. FTIR spectra of cellulose nanofibers (CnF), electrospun nano-fibers of poly(vinyl alcohol) (PVA), PVA reinforced with cellulose nano-fibers (PVA/CnF), and the difference spectrum (PVA-CnF).

material is neither PVA, nor CnF, despite presenting peaksof both PVA and CnF. This may be attributed to the for-mation of either hydrogen or covalent bonds which mighthave taken place during the compounding step as a conse-quence of the heating process or to cross-linking of PVAmolecules during dissolution therefore entrapping CNFfibrils between these cross-linked chains and conferringmechanical adhesion between the components. Hydrogenbonding interactions have been reported in the literaturefor PVA and cellulose composites.48�49 Acid hydrolysisof cellulose removes flexible amorphous sections linkingcrystals to form nanofibrils that are recovered ad aque-ous suspensions.50–52 As a result, these nanofibers haveelectronegative surface charges on their surface (usuallysulfate and polar hydroxyl groups) that are capable ofhydrogen bonding, leading to particle–particle or particle–solvent–particle interactions to form three-dimensionalnetworks.53–56

The mechanical properties of the electrospun nanofibersof PVA/CnF were compared to nanofibers of PVA electro-spun in the same experimental conditions. The results ofthe mechanical tests are summarized in Table I and typical

Table I. Mechanical properties of electrospun nanofibers of PVA andPVA reinforced with cellulose nanofibrils.

PVA PVA-Cellulose

Young’s Modulus, E (MPa) 45±17 108±18Tensile stress, m (MPa) 2�0±0�6 4�8±1�2Elongation at break, �m(%) 26±5 25±4


RESEARCHARTICLE


0 5 10 15 20 25 300

1

2

3

4

5

6

7S

tres

s (M

Pa)

Strain (%)

PVAPVA-CnF

Fig. 12. Typical stress-strain curves of electrospun fibrous mats of PVAand PVA reinforced with cellulose nanofibers.

stress-strain curves of electrospun nanofibers of PVA andPVA reinforced with cellulose nanofibers are illustrated inFigure 12.

It can be observed in Figure 12 that Young’s modulusand tensile strength of reinforced composites increased byca. 240% by addition of only 6.6 wt% of cellulose nano-fibers without major change in elongation at break. A sim-ilar behavior was found by Petersson and Oksman57 withpoly(lactic acid) nanocomposites reinforced with micro-crystalline cellulose in which the addition of cellulosereduced elongation at break by 16% as a result of incor-poration of 5 wt% of cellulose. In contrast, addition ofthe same amount of bentonite, a layered silicate, caused areduction of 75%.

The presence of strong interactions between thesehydrophilic components as pointed out by FTIR analyseshelps explain the increase in both Young’s modulus andtensile strength without significantly reducing the elonga-tion at break. Furthermore, the polarity of both compo-nents due to the presence of hydroxyl groups is one of thefactors that may have contributed to improving the com-patibility of the composite, thus, promoting a more effi-cient load transfer from matrix to fibers in a way similarto traditional fiber reinforced composites.58�59

The remarkable improvement in mechanical propertiesby addition of cellulose nanofibers has also been found byOrts et al.7 The authors reported that for extruded starchplastics, the addition of cotton-derived fibrils at a concen-tration of 10.3 wt% increased Young’s modulus by 5-foldrelative to a control sample with no cellulose reinforce-ment. Moreover, this increase in tensile properties, espe-cially maximum load and tensile strength, depended notonly on the shear-induced alignment of the fibers but alsoon the type of polymer matrix, implying that complexinteractions are involved in the reinforcing capability ofthe nanofibers.

This increase in Young’s modulus and tensile strength ismuch more accentuated than in traditional fiber reinforcedcomposites where a relatively modest increase in theseproperties is attained only by using comparatively muchhigher amounts of reinforcements.58�59 Indeed, Young’smodulus of the basic cellulose crystalline fibrils has beenreported as 137 GPa,60�61 but in order to fully exploit itspotential as reinforcements detailed investigation of thecellulose nanofiber-polymer interface is required.7

4. CONCLUSIONS

Nanofibers of poly(vinyl alcohol) reinforced with cellulosenanofibers were successfully produced by electrospinning.The morphology of the electrospun nanofibers (electro-spun mats) was strongly affected by the electrospinningconditions, i.e., the effect of applied voltage, polymerconcentration and injection rate, tip-to-collector distance,rotation speed of the collector, and relative humidity. Thin-ner fibers are favored by decreasing polymer concentrationand injection rate, increasing rotation speed of the collec-tor and relative humidity, whereas increasing the appliedvoltage favors the formation of beaded fibers. The balancebetween all these variables is the key to accurately control-ling the morphology and the final properties of the electro-spun mats. The reinforced composites had their Young’smodulus and tensile strength increased by ca. 240% byaddition of only 6.6 wt% of cellulose nanofibers withoutmajor changes in elongation at break due to the good inter-action between PVA and cellulose nanofibers during thecompounding step as evidenced by FTIR analyses of thecomposites. These results show that addition of cellulosenanofibrils to electrospinning solution of PVA is an effi-cient method to improve the mechanical strength of elec-trospun nanofibers.

Acknowledgments: The authors would like to thankTina Williams for SEM analyses and Gregory M. Grayfor his many valuable suggestions on the electrospinningsetup. Luiz H. C. Mattoso gratefully acknowledges CNPq,FINEP and the EMBRAPA-LABEX program (Brazil) forhis financial support.

References

1. M. A. S. Azizi-Samir, F. Alloin, and A. Dufresne, Biomacromol. 6,612 (2005).

2. L. A. Berglund, Natural Fibers, Biopolymers, and Biocomposites,edited by A. K. Mohanty, M. Misra, and L. T. Drzal, CRC Press,Boca Raton (2005), pp. 807–832.

3. A. Okada and A. Usuki, Mat. Sci. Eng. C 3, 109 (1995).4. I. Sakurada, Y. Nukushina, and I. Ito, J. Polym. Sci. 57, 651 (1962).5. S. A. Wainwright, W. D. Biggs, J. D. Currey, and J. M. Gosline,

Mechanical Design in Organisms, Princeton University Press,Princeton (1982).

6. A. Dufresne and M. R. Vignon, Macromol. 31, 2693 (1998).


RESEARCHARTICLE


7. W. J. Orts, J. Shey, S. H. Imam, G. M. Glenn, M. E. Guttman, andJ.-F. Revol, J. Polym. Environ. 13, 301 (2005).

8. H. Takagi and A. Asano, Key Eng. Mat. 334/5, 389 (2007).9. B. Wang and M. Sain, Comp. Sci. Tech. 67, 2521 (2007).10. M. Grunert and W. Winter, J. Poly. Environ. 10, 27 (2002).11. D. Klemm, B. Heublein, H. P. Fink, and A. Bohn, Angew. Chem.

Int. Ed. 44, 3358 (2005).12. D. Klemm, H.-P. Schmauder, and T. Heinze, Biopolymers, edited

by E. Vandamme, S. De Beats, and A. Steinbüchel, Wiley-VCHWeinheim (2002), Vol. 6, pp. 290–292.

13. D. L. Kaplan (ed.), Biopolymers From Renewable Resources,Springer, Berlin (1998), pp. 1–29.

14. T. Subbiah, G. S. Bhat, R. W. Tock, S. Parameswaran, and S. S.Ramkumar, J. Appl. Polym. Sci. 96, 557 (2005).

15. S. Ramakrishna, K. Fujihara, W. E. Teo, T. C. Lim, and Z. Ma,An Introduction to Electrospinning and Nanofibers, World ScientificPublishing, Singapore (2005).

16. H. Pan, L. Li, L. Hu, and X. Cui, Polymer 47, 4901 (2006).17. N. E. Marcovich, M. L. Auad, N. E. Bellesi, S. R. Nutt, and M. I.

Aranguren, J. Mater. Res. 21, 870 (2006).18. M. Bognitzki, T. Frese, M. Steinhart, A. Greiner, J. H. Wendorff,

A. Schaper, and M. Hellwig, Polym. Eng. Sci. 41, 982 (2001).19. K. Tuzlakoglu, N. Bolgen, A. J. Salgado, M. E. Gomes, E. Piskin,

and R. L. Reis, J. Mat. Sci.: Mat. Med. 16, 1099 (2005).20. M. Li, M. J. Mondrinos, M. R. Gandhi, F. K. Ko, A. S. Weiss, and

P. I. Lelkes, Biomaterials 26, 5999 (2005).21. H. L. Schreuder-Gibson, Q. Truong, J. E. Walker, J. R. Owens, J. D.

Wander, and W. E. Jones, Jr., MRS Bulletin 574 (2003).22. B. Ding, J. Kim, Y. Miyazaki, and S. Shiratori, Sens. Actuators B

101, 373 (2004).23. M. M. Demir, M. Naseer, T. F. Bechteler, Y. Gurbuz, and Y. Z.

Menceloglu, Mat. Res. Soc. Symp. Proc. 782, 1 (2004).24. H. Jiang, Y. Hu, Y. Li, P. Zhao, K. Zhu, and W. Chen, J. Control.

Rel. 108, 237 (2005).25. X. F. Wang, I. C. Um, D. F. Fang, A. Okamoto, B. S. Hsiao, and

B. Chu, Polymer 46, 4853 (2005).26. Y. Ji, K. Ghosh, X. Z. Shu, B. Q. Li, J. C. Sokolov, G. D. Prestwich,

R. A. F. Clark, and M. H. Rafailovich, Biomaterials 27, 3782(2006).

27. Y. Z. Zhang, J. Venugopal, Z. M. Huang, C. T. Lim, andS. Ramakrishna, Polymer 47, 2911 (2006).

28. X. Wen, D. Shi, and N. Zhang, Handbook of Nanostructured Bio-materials and Their Applications in Nanobiotechnology, edited byH. S. Nalwa, American Scientific Publishers, Los Angeles (2005),Vol. 1, pp. 1–23.

29. R. L. Crawford, Lignin Biodegradation and Transformation, JohnWiley and Sons, New York (1981).

30. I. S. Kim, J. P. Kim, S. Y. Kwak, Y. S. Ko, and Y. K. Kwon, Polymer47, 1333 (2006).

31. L. H. C. Mattoso, W. J. Orts, and E. S. Medeiros, Nano CrystallinePolysaccharides U.S. Patent, Provisional No. 153.07 (2007).

32. E. S. Medeiros, L. H. C. Mattoso, R. Bernardes-Filho, D. F. Wood,and W. J. Orts, Coll. Polym. Sci., In Press.

33. L. H. C. Mattoso, D. A. Baker, J. Avloni, D. F. Wood, and W. J.Orts, Proceedings of the 41st International Symposium on Macro-molecules, Rio de Janeiro, Brazil (2006).

34. M. Franchitto, R. Faez, A. J. F. Orlando, M. C. Rezende, and I. M.Martin, Proceedings of the 2001 SBMO/IEEE MTT-S International(2001).

35. H. Hovhannisyan, H. Matnishyan, A. Vardanyan, A. Hakhumyan,and L. Abrahamyan, Proceedings of the 34th European MicrowaveConference (2004).

36. V. Favier, H. Chanzy, and J. Y. Cavaille, Macromolecules 28, 6365(1996).

37. H. Fong, I. Chun, and D. H. Reneker, Polymer 40, 4585 (1999).38. G. I. Taylor, Proc. Roy. Soc. London A313, 453 (1969).39. R. V. N. Krishnappa, K. Desai, and C. Sung, J. Mat. Sci. 38, 2357

(2003).40. D. Li, Y. Wang, and Y. Xia, Adv. Mat. 16, 361 (2004).41. S. F. Fennessey and R. J. Farris, Polymer 45, 4217 (2004).42. W. K. Son, J. H. Youk, T. S. Lee, and W. H. Park, Mat. Lett. 59,

1571 (2005).43. C. Zhang, X. Yuan, L. Wu, Y. Han, and J. Sheng, European.

Polym. J. 41, 423 (2005).44. R. Jaeger, M. M. Bergshoef, C. M. I. Batlle, S. Holger, and G. J.

Vancso, Macromol. Symp. 127, 141 (1998).45. W. K. Son, J. H. Youk, T. S. Lee, and W. H. Park, Mat. Letters 59,

1571 (2005).46. A. Koski, K. Yim, and S. Shivkumar, Mat. Letters 58, 493 (2004).47. F. Carrillo, X. Colom, J. J. Suñol, and J. Saurina, European Polym. J.

40, 2229 (2004).48. T. Kondo, C. Sawatari, R. S. J. Manley, and D. G. Gray, Macro-

molecules 27, 210 (1994).49. S. Majumdar and B. Adhikari, Bull. Mater. Sci. 28,703 (2005).50. O. A. Battista and P. A. Smith, Ind. Eng. Chem. 54, 20 (1962).51. O. A. Battista, N. Z. Erdi, C. F. Ferraro, and F. J. Karasinski, J. Appl.

Polym. Sci. 11, 481 (1967).52. O. A. Battista, Microcrystalline Polymer Science, McGraw-Hill

Book Co, New York (1975).53. S. Deguchi, K. Tsujii, and K. Horikoshi, Chem. Commun. 32, 93

(2006).54. M. Roman and W. T. Winter, Biomacromol. 5, 1671 (2004).55. N. Wang, E. Ding, and R. Cheng, Polymer 48, 3486 (2007).56. J. Araki, M. Wada, S. Kuga, and T. Okano, Colloids Surf. A142, 75

(1998).57. L. Petersson and K. Oksman, Comp. Sci. Tech. 66, 2187 (2006).58. E. S. Medeiros, L. H. C. Mattoso, P. A. Sreekumar, and K. Joseph,

Natural Fibre Reinforced Polymer Composites, edited by S. Thomasand L. A. Pothan, Old City Publishing Co., Philadelphia (2008), Inpress.

59. E. S. Medeiros, J. A. M. Agnelli, K. Joseph, L. H. Carvalho, andL. H. C. Mattoso, Polym. Comp. 26, 1 (2005).

60. I. Sakurada, Y. Nukushina, and I. Ito, J. Polym. Sci. 57, 651 (1962).61. S. A. Wainwright, W. D. Biggs, J. D. Currey, and J. M. Gosline,

Mechanical Design in Organisms, Princeton University Press,Princeton (1982).

Received: xx Xxxx xxxx. Revised/Accepted: xx Xxxx xxxx.


Documents

ElectrospunNanoﬁbersofPoly(vinylalcohol