Materials Sciences and Applications, 2010, 1, 247-252
doi:10.4236/msa.2010.14036 Published Online October 2010 (http://www.SciRP.org/journal/msa)
Effect of Solution Concentration on the
Electrospray/Electrospinning Transition and
on the Crystalline Phase of PVDF
Lígia Maria Manzine Costa, Rosário Elida Suman Bretas, Rinaldo Gregorio, Jr.
Department of Materials Engineering, Federal University of São Carlos, Rod. Washington Luís, São Carlos-SP, Brazil.
Received March 19th, 2010; revised June 17th, 2010; accepted June 21st, 2010.
ABSTRACT
A study was conducted regarding the effect of concentration of poly (vinylidene fluoride) (PVDF)/N,N-dimethylformamide
(DMF) and PVDF/DMF/acetone solutions on the transition between electrospray and electrospinning and on the for-
mation of the  and  crystalline phases of PVDF. The crystalline phases present in the samples, crystallinity and
morphology were determined by Fourier transform infrared spectroscopy (FTIR), differential scanning calorimetry
(DSC) and scanning electron microscopy (SEM), respectively. Low concentration solutions resulted in films consisting
of small droplets (electrospray) containing predominantly the  phase. High concentration solutions resulted in a
non-woven mesh of nano-to-micron diameter fibers (electrospinning) containing exclusively the  phase. These results
showed that, the formation of this phase in the electrospinning is related mainly to the solvent evaporation rate, and not
to drawing experienced by the polymer during the process. Solvent type affected the amount of crystalline phase present,
the boundary concentration between the two processes and the average diameter of fibers. Meshes processed by elec-
trospinning display a degree of crystallinity higher than the films obtained by electrospray.
Keywords: Poly(Vinylidene Fluoride) (PVDF), Electrospinning, Electrospray, Crystalline Phase, Fibers, Morphology
1. Introduction
PVDF, poly(vinylidene fluoride), is a widely investigated
polymer due to its excellent mechanical properties,
chemical stability and ferroelectricity. This polymer has
a simple chemical formula, -CH2-CF2-, and may crystal-
lize in at least four crystalline phases known as α, β, γ
and δ. The structure and property of these phases are well
documented in the literature [1]. Each crystalline phase
confers characteristic properties to the polymer and
therefore distinct applications. The β phase is that which
displays the strongest pyro and piezoelectric activities
and is therefore technologically more interesting. Many
techniques have been used to obtain this phase, being the
most common uni or biaxial drawing of originally α
phase films [2-6]. Crystallization of PVDF from solution
(casting) may also result in the  phase, depending on the
evaporation rate of the solvent [7-9]. Low rates result
predominantly in the  phase, whereas high rates favor
the  phase. Electrospinning [10,11] also produces pre-
dominantly the  phase [12-18]. In this technique the
polymer solution is added to a capillary (which can be a
syringe with needle). The solution forms a droplet at the
tip of the needle due to surface tension. If a sufficiently
high electric voltage (5-30 kV) is applied to the solution
electric charges accumulate in the droplet. When the
electrostatic repulsion between the charges overcomes
the surface tension and viscoelasticity of the droplet this
assumes the shape of a cone. When the force exerted by
the electric field, formed between the droplet and a
grounded collector, overcomes the surface tension, a thin
jet is formed. The jet of the charged solution is acceler-
ated towards the collector under the action of the electric
field. If the concentration of the solution is sufficiently
high (viscous) to stabilize the jet, the polymer solution is
severely stretched and the solvent evaporates to form
ultrathin fibers which solidify and are deposited on the
collector forming a non-woven mesh of fibers. Some
authors [13-17] attribute the formation of β phase in
electrospun PVDF meshes to this severe drawing ex-

Effect of Solution Concentration on the Electrospray/Electrospinning Transition and on the Crystalline Phase of PVDF 248
perienced by the polymer jet solution. This elongation
process would be similar to the well known  transi-
tion caused by mechanical drawing at T  90℃ [2-6].
A variant of electrospinning is electrospray [19]. The
basic difference between these two processes lies in the
concentration of the solution. In electrospray the concen-
tration is sufficiently low to destabilize the charged jet
which then breaks down into small spherical droplets that
solidify during the course and are deposited on the col-
lector. In this case the polymer solution does not experi-
ence severe drawing and the formed film consists of
small droplets instead of fibers.
The objectives of the current investigation were: 1)
determine the solution concentration at which the transi-
tion between electrospray and electrospinning of PVDF
occurs and 2) verify through the electrospray technique,
where no stretching occurs, that the formation of the 
phase depends fundamentally on the solvent evaporation
rate. The effect of the type of solvent in the process was
also verified.
2. Experimental Methods
2.1. Materials
The PVDF used was Foraflon® 4000HD, from Elf Ato-
chem. The solvents used were N,N-dimethylformamide
(DMF, Merck 99.5%) and acetone (Merck, 99.7%). So-
lutions were prepared at concentrations of 5, 7, 10 and 15
wt% PVDF using as solvent pure DMF and a 3:1 v/v
mixture of DMF with acetone. The ratio of mixture DMF
and acetone was selected because they produce thinner
and more homogeneous nanofibers in the electrospinning
process [13]. Solubilization was carried out at 70℃ un-
der stirring for one hour. Table 1 shows some properties
of the solvents used.
2.2. Electrospray/Electrospinning
A scheme of the system used in the electrospray/electro-
spinning process is shown in Figure 1. A 20-mL glass
syringe was used with a steel needle with internal di-
ameter of 0.7 mm. The distance between needle and col-
lector was 3 cm and the electric voltage applied was 10
kV, using a Bertan 210 30R (0-30 kV) high voltage source.
The collector used was an aluminum disk with diameter
of 15 cm and width of 5 cm, at an angular velocity of 60
rpm. The process was carried out at 25℃ and RH  55%.
2.3. Characterizations
The crystalline phases present in the samples were char-
acterized by transmission infrared spectroscopy (FTIR,
Perkin Elmer Spectrum 1000) in the range between 400
and 1000 cm-1 and with resolution of 2 cm-1. Calorimetric
analyses were conducted in a Perkin Elmer DSC-7, cali-
Table 1. Properties of the solvents used (20℃).
Solvent Formula Density
(g/mL)
Boiling
point (℃)
Viscosity
(mPa.s)
Acetone C3H6O 0.786 56.5 0.326
DMF C3H7NO 0.944 153 0.820
Figure 1. Scheme of the setup used in the Electrospinning/
Electrospray process.
brated with In, using samples of 8 mg and heating rate
of 10℃/min. Degree of crystallinity was determined by
comparing the enthalpy of fusion of the sample obtained
by DSC with that of 100% crystalline PVDF- (104.5 J/g)
[20]. Sample surface morphology was observed by using
a scanning electron microscope (Philips 30 FEG Model
XL, operating at 15 kV) after gold coating. Size distribu-
tion of the fibers diameters was determined by the soft-
ware “image J” developed by the National Institute of
Health.
3. Results and Discussions
The micrographs in Figures 2(a) to (d) and 3(a) to (d)
present the morphology of the samples processed with
PVDF/DMF and PVDF/DMF/acetone solutions at dif-
ferent concentrations, respectively.
One can observe in Figures 2(a)-(b) (5 and 7 wt% in
DMF) and 3(a) (5 wt% in DMF/acetone) the formation
of a film consisting of small droplets, characteristic of
the electrospray process. The morphology of this film
depends among other things on droplet volume at impact
and on evaporation rate of the solvent [21]. In Figures
2(c)-(d) (10, 15 wt% in DMF) and 3(b)-(d) (7, 10 and 15
wt% in DMF/acetone) fibers predominate, characteristic
of electrospinning. In micrographs 2(c) (10 wt% in DMF)
and 3( b) (7 wt% in DMF/acetone) the electrospun sheets
present some small droplets, demonstrating that concen-
tration was not yet ideal for the formation of homogene-
ous fibers. These results show that the boundary concen-
tration between electrospray and electrospinning ranges
from 7 to10 wt% PVDF in DMF and from 5 to 7 wt% in

Effect of Solution Concentration on the Electrospray/Electrospinning Transition and on the Crystalline Phase of PVDF 249
(a) (b)
(c) (d)
Figure 2. Micrographs of the samples processed from PVDF/
DMF solution at the following concentrations (wt%): (a) 5;
(b) 7; (c) 10 and (d) 15.
(a) (b)
(c) (d)
Figure 3. Micrographs of samples processed from solution
of PVDF with DMF/acetone (3:1 v/v) at the following con-
centrations (wt%): (a) 5; (b) 7; (c) 10 and (d) 15.
DMF/acetone mixture, at the conditions used of voltage
and distance between needle and collector. The solution
containing acetone is less viscous than that of pure DMF
and therefore the electrospinning process should be initi-
ated at a higher concentration. However, the high volatil-
ity of acetone increased the jet concentration rapidly
during its course towards the collector, impeding forma-
tion of droplets and resulting in electrospinning at a
lower initial concentration. Micrographs 2(d) (15 wt% in
DMF) and 3(c)-(d) (10 and 15 wt% in DMF/acetone)
show fibers with improved homogeneity and no droplets,
indicating that these concentrations are the optimum for
obtaining electrospun meshes.
Figures 4 and 5 contain the size distribution of fibers
diameters prepared with 15 wt% PVDF in DMF and
DMF/acetone, respectively.
Figure 4. Size distribution of the fibers diameters processed
from 15 wt% PVDF/DMF solution.
Figure 5. Size distribution of the fibers diameters processed
from 15 wt% PVDF/DMF/acetone.
Average diameter of the nanofibers prepared from
DMF solution (514 nm) were superior to that obtained
from DMF/acetone solution (330 nm), at the same con-
centration. This has been caused by the lower viscosity
of the acetone solution, since average fiber diameter in-
creases with solution viscosity [12]. That effect can be
observed in Figures 2(c)-(d) and 3(b)-(d) .
The phases present in the samples were analyzed by
FTIR and results are presented in Figures 6 and 7 for
DMF and DMF/acetone solutions at different concentra-
tions, respectively.
All samples presented bands at 445, 510 and 840 cm-1,
characteristic of the  phase of PVDF [22,23]. Therefore,
the β phase predominated both in the electrosprayed
films and the electrospun meshes. Since no stretching
occurred during electrospray, we may conclude that for

Effect of Solution Concentration on the Electrospray/Electrospinning Transition and on the Crystalline Phase of PVDF 250
Figure 6. FTIR spectra of the samples with different wt%
PVDF in DMF solution. The characteristic bands of the α
and β phase are pointed out in the figure.
Figure 7. FTIR spectra of the samples with different wt%
PVDF in 3:1 DMF/acetone solution. The characteristic
bands of the α and β phase are pointed out in the figure.
mation of this phase can not be attributed to drawing
experienced by the polymer during the process. The oc-
currence of weak bands at 614, 764 and 976 cm-1 indicate
the presence of a small amount of the  phase [22,23],
observed at concentrations of 5 to 10 wt% in DMF/ace-
tone solution. In the solution containing pure DMF the
presence of the  phase was only observed for 5 wt%.
Since solvent evaporation rate increases with decreasing
concentration of the solution, appearance of this phase
may be related to the higher solvent evaporation rate in
low-concentration solutions. The evaporation rate of the
solutions containing acetone increased in relation to those
containing pure DMF, since acetone has a lower evapo-
ration temperature, and this resulted in the formation of a
small amount of  in almost all concentrations. Thus, the
formation of the β phase in the electrospinning/electro-
spray process is related mainly to the solvent evaporation
rate, as verified in cast films [8,9], and not to drawing
experienced by the polymer during the electro-spinning
process, as suggested by some authors [13-17]. Low
evaporation rates result predominantly in the  phase,
thermodynamically more favorable, intermediate rates in
a mixture of  and  and high rates in the  phase, ki-
netically more favorable.
Figure 8 shows DSC curves for samples prepared with
5 and 10 wt% PVDF in DMF and DMF/acetone. The
DSC measurements give the melting point (Tm) (endo-
therm peak), enthalpy of fusion (  H), and degree of
crystallinity (%C) of samples. These values are summa-
rized in Table 2.
It is easy to observe that meshes produced by electro-
spinning display a degree of crystallinity higher when
compared to films obtained by electrospray. This result
probably happens due to the preferential orientation of
the chains in the direction of the nanofibers, caused by
the drawing experienced by the polymer during the elec-
trospinning and that facilitates the crystallization. In this
process the addition of acetone promoted a small in-
crease in the degree of the mesh crystallinity. The melt-
ing points, in all cases correspond to the temperature
range related to the PVDF α and β phases [7,23], which
confirm FTIR results.
4. Conclusions
The current investigation showed that formation of the β
crystalline phase of PVDF in the electrospun meshes is
Figure 8. DSC curves for samples prepared with 5 wt%
PVDF in DMF (a) and DMF/acetone (b); with 10 wt%
PVDF in DMF (c) and DMF/acetone (d).

Effect of Solution Concentration on the Electrospray/Electrospinning Transition and on the Crystalline Phase of PVDF 251
Table 2. Melting point (Tm), enthalpy of fusion (  H) and degree of cristallinity (%C) of the samples processed with
PVDF/DMF and PVDF/DMF/acetone solutions at different concentrations.
Sample Process Tm (℃)  H (J/g) %C
5 wt% PVDF/DMF 166 46.2 44.6
5 wt% PVDF/DMF/acetone
Electrospray
167 46.8 45.2
10 wt% PVDF/DMF 168 53.4 51.1
10 wt% PVDF/DMF/acetone
Electrospinning
167 55.0 53.1
not related to drawing experienced by the polymer during
the process, as suggested by some authors. Formation of
the  and  phase is related primarily to the evaporation
rate of the solvent. High rates favor formation of the 
phase, whereas low rates favor the  phase. Under the
same conditions of voltage and needle-collector distance,
the boundary concentration between electrospray and
electrospinning depends on the solvent used. For PVDF
this concentration lies between 7 and 10 wt% with DMF
as solvent and between 5 and 7 wt% for the 3:1 DMF/
acetone mixture. Moreover, addition of acetone to the
solution reduced average fiber diameter. Meshes pro-
duced by electrospinning present degree of crystallinity
higher than the films obtained by electrospray, for both
used solvents.
5. Acknowledgements
The following agencies are acknowledged for financial
support: CAPES, CNPq and FAPESP.
REFERENCES
[1] K. Tashiro, “Crystal Structure and Phase Transition of
PVDF and Related Copolymers,” In: H. S. Nalwa, Ed.,
Ferroelectric Polymers: Chemistry, Physics and Applica-
ons, Marcel Dekker, New York, 1995, pp. 63-180.
[2] R. Gregorio, Jr. and E. M. Ueno, “Effect of Crystalline
Phase, Orientation and Temperature on the Dielectric
Properties of Poly(Vinylidene Fluoride) (PVDF),” Jour-
nal of Materials Science, Vol. 34, 1999, pp. 4489-4500.
[3] V. Sencadas, V. M. Moreira, S. Lanceros-Mendez, A. S.
Pouzada and R. Gregorio, Jr., “Alpha-to-Beta Transfor-
mation on PVDF Films Obtained by Uniaxial Stretch,”
Advanced Materials Forum III, Vol. 514-516, 2006, 872-
876.
[4] P. Sajkiewicz, A. Wasiak and Z. Goclowski, “Phase Tran-
sitions during Stretching of Poly(Vinylidene Fluoride),”
European Polymer Journal, Vol. 35, No. 3, pp. 423-429.
[5] D. T. Grubb and F. R. Kearney, “Phase Transitions dur-
ing Stretching of Poly(Vinylidene Fluoride),” Journal of
Polymer Science Part B: Polymer Physics, Vol. 28, No.
11, 1990, pp. 2071-2078.
[6] M. C. Branciforti, V. Sencadas, S. Lanceros-Mendez and
R. Gregorio, Jr., “New Technique of Processing Highly
Oriented Poly(Vinylidene Fluoride) Films Exclusively in
the  Phase,” Journal of Polymer Science Part B: Poly-
mer Physics, Vol. 45, 2007, pp. 2793-2801.
[7] R. Gregorio, Jr., and M. Cestari, “Effect of Crystallization
Temperature on the Crystalline Phase Content and Mor-
phology of Poly(Vinylidene Fluoride),” Journal of Poly-
mer Science Part B: Polymer Physics, Vol. 32, No. 5, pp.
859-870.
[8] R. Gregorio, Jr. and D. S. Borges, “Effect of Crystalliza-
tion Rate on the Formation of the Polymorphs of Solution
Cast Poly(Vinylidene Fluoride),” Polymer, Vol. 49, No.
18, 2008, pp. 4009-4016.
[9] D. L. Chinaglia, R. Gregorio, Jr., J. C. Stefanello, R. A. P.
Altafin, W. Wirges, F. Wang and R. Gerhard, “Influence
of the Solvent Evaporation Rate on the Crystalline Phases
of Solution-Cast Poly(Vinylidene Fluoride) Films,” Jour-
nal of Applied Polymer Science, Vol. 116, No. 2, 2010,
pp. 785-791.
[10] A. Formhals, “Process and Apparatus for Preparing
Artificial Threads,” U.S. Patent 1,975,504, 1934.
[11] J. Doshi and D. H. Reneker, “Electrospinning Process and
Application of Electrospun Fibers,” Journal of Electro-
statics, Vol. 35, No. 2-3, 1995, pp. 151-160.
[12] M. Nasir, H. Matsumoto, T. Danno, M. Minagawa, T.
Irisawa, M. Shioya and A. Tanioka, “Control of Diameter,
Morphology, and Structure of PVDF Nanofiber Fabri-
cated by Electrospray Deposition,” Journal of Polymer
Science Part B: Polymer Physics, Vol. 44, No. 5, pp. 779-
786.
[13] W. A. Yee, M. Kotaki, Y. Liu and X. Lu, “Morphology,
Polymorphism Behavior and Molecular Orientation of
Electrospun Poly(Vinylidene Fluoride) Fibers,” Polymer,
Vol. 48, No. 2, 2007, pp. 512-521.
[14] E. Ogut, O. S. Yordem, Y. Z. Menceloglu and M. Papila,
“Poly(Vinylidene Fluoride)/Zinc Oxid Smart Composite
Material,” Proceeding of SPIE, Vol. 6526, 2007, p.
65260Q-1-10
[15] J. Zheng, A. He, J. Li and C. C. Han, “Polymorphism
Control of Poly(Vinylidene Fluoride) through Electro-
spinning,” Macromolecular Rapid Communications, Vol.
28, No. 22, 2007, pp. 2159-2162.
[16] W. A. Yee, A. C. Nguyen, P. S. Lee, M. Kotaki, Y. Liu, B.
T. Tan, S. Mhaisalkar and X. Lu, “Stress-Induced Struc-
tural Changes in Electrospun Polyvinylidene Difluoride
Nanofibers Colleted Using a Modified Rotating Disk,”
Polymer, Vol. 49, No. 19, 2008, pp. 4196-4203.

Effect of Solution Concentration on the Electrospray/Electrospinning Transition and on the Crystalline Phase of PVDF 252
[17] J. S. Andrew and D. R. Clarke, “Effect of Electrospin-
ning on the Ferroelectric Phase Content of Polyvinylidene
Difluoride Fibers,” Langmuir, Vol. 24, No. 3, 2008, pp.
670-672.
[18] H. Na, Y. Zhao, C. Zhao, C. Zhao and X. Yuan, (2008)
Effect of Hot-Press on Electrospun Poly(Vinylidene Fluo-
ride) Membranes,” Polymer Engineering and Science, Vol.
48, No. 5, pp. 934-940.
[19] G. Taylor, “Electrospray,” Proceedings of Royal Society
of London, London, Vol. A280, 1964, pp. 383-387.
[20] K. Nakagawa and Y. Ishida, “Dielectric Relaxations and
Molecular Motions in Poly(Vinylidene Fluoride) with
Crystal Form II,” Journal of Polymer Science Part B:
Polymer Physics, Vol. 11, No. 8, pp. 1503-1533.
[21] I. B. Rietveld, K. Kobayashi, H. Yamada and K. Matsu-
shige, “Model Supported Morphology Control of Elec-
trospray Deposited Poly(Vinylidene Fluoride) Film,”
Macromolecular Symposium, 2007, pp. 249-250, 322-
329.
[22] M. Kobayashi, K. Tashiro and H. Tadokoro, “Molecular
Vibrations of Three Crystal Forms of Poly(Vilidene Fluo-
ride),” Macromolecules, Vol. 8, No. 2, 1975, pp. 158-170.
[23] R. Gregorio, Jr., “Determination of the ,  and  Crys-
talline Phases of Poly(Vinylidene Fluoride) Films Pre-
pared at Different Conditions,” Journal of Applied Poly-
mer Science, Vol. 100, No. 4, 2006, pp. 3272-3279.