Navarro Et Al 2020
Navarro Et Al 2020
Navarro Et Al 2020
DOI: 10.1002/pc.25924
RESEARCH ARTICLE
1
Roberval Laboratory, FRE UTC-CNRS
2012, Sorbonne Universités, Université de
Abstract
Technologie de Compiègne, Compiègne, Nanoreinforced polymers have gained popularity in the last decades since
France
they exhibit enhanced properties (compared to pristine polymers) that are
2
NORMAFIN, LEVALLOIS-PERRET,
useful in a wide range of applications. Unfortunately, dispersion of
France
3
DPT, Institut Laue-Langevin, 71 avenue
nanoparticles (NPs) into polymeric matrices is a major problem since they
des Martyrs, Grenoble, France tend to form agglomerates, limiting the improvement of properties and
further applications. In this work, we propose the use of coaxial
Correspondence
Fahmi Bedoui, Roberval Laboratory, FRE electrospinning as one-step method to disperse NPs in a polymeric matrix.
UTC-CNRS 2012, Sorbonne Universités, Particularly, iron oxide (Fe3O4) NP with a monomodal and bimodal size dis-
Université de Technologie de Compiègne,
tributions were dispersed in polyvinylidene fluoride (PVDF), a material that
Compiègne, France.
Email: [email protected] is well known for its improved piezoelectric properties when it is processed
via electrospinning. The results indicate that the incorporation of NP modi-
fied the polymeric fiber depending on their surface-to-volume ratio (smaller
NP promoted smaller fiber size). Moreover, transmission electron micros-
copy revealed a good NP dispersion in the polymer, especially for the
smallest NP size (monomodal). Finally, each NP size distributions were
well preserved in the electrospun mats compared to the initial NP solutions,
demonstrating the suitability of this technique for the fabrication of
nanoreinforced PVDF structures with tailored NP size. Overall, this method
could represent a facile and practical alternative to fabricate materials with
piezoelectric/super-paramagnetic properties.
KEYWORDS
coaxial electrospinning, nanofibers, nanoparticle dispersion
Diverse methods have been explored to increase the fibers. Our main goal is to reach a good qualitative NP
dispersion of NP into polymeric matrices: shear mixing dispersion within the fibrillar network for each NP size
and sonication, sol-gel process and chemical modification distribution. Here, PVDF was chosen since it is well
of either components.[10] Although these strategies known that electrospinning promotes the appearance
generally render good NP dispersion in the composites, of its β-phase, one of its five crystal polymorphs which
they also present some specific drawbacks: dispersion presents a highly polar organization,[26] making it a
promoted by physical forces (shear mixing-sonication) potential candidate for the development of materials
could be reversed because of the hydrophilic nature of with ferroelectric/piezoelectric properties. Further,
NP's cover and therefore dispersing agents are neces- PVDF is a highly tunable material when used in
sary[11,12]; composites developed through sol-gel tech- nanocomposites; improvements on its mechanical,
nique in some cases present weak bonding and hard thermal, electrical, and electromagnetic properties
porosity control, in addition to time-consuming aging/ have been already reported.[8,27,28] Additionally, we
drying processes[13]; finally, NP surface modification aim the incorporation of Fe3O4 NP into the PVDF net-
(polymer-grafting onto the hydroxyl groups included) work since they can provide with super-paramagnetic
is perhaps the most common method to prepare properties to the final electrospun composite, making
homogenous NP/polymer composites although it nor- it relevant for energy harvesting or biomedical
mally comprises two or more synthetic steps.[9] For applications.
instance, in a recent approach, Ding et al. prepared
polymer nanocomposites using anionic and cationic
ligands as NP dispersant, emulsifier, and comonomer 2 | MATERIALS AND METHODS
(i.e., the ligands copolymerized with the polymer
matrix).[14] Despite the qualitatively-enhanced NP dis- 2.1 | Coaxial electrospinning of polymer
persion, increased tensile strength (77% higher com- and NPs
pared to nonmodified polymer) and improved thermal
stability (24 C higher than control), this method impli- The representation of the coaxial electrospinning set-
cated multiple (preparative) steps, which turns it rather up used in this study is shown in Figure 1. In brief,
unpractical. 1.876 g of polyvinylidene fluoride (PVDF, ref:
The preparation of polymeric composites via FV306300, Goodfellow, Ltd, Cambridge, UK) were
electrospinning has gained attention lately,[15–18] not only dissolved in 10 ml of a dimethylformamide (DMF)-
because NP reinforce the fibrillar structure of electrospun acetone mix (40/60 vol/vol) and stirred (400 rpm) at
mats, but also the fiber networks could protect NP from 70 C during 1 h. The PVDF solution (18% wt/wt) was
environmental processes such as corrosion and oxida- then left to cool down for another 15 min before load-
tion.[19] However, this process still relies on the mixing of ing it into the shell syringe. The core syringe was
the components (polymer + NP) in the syringe prior to loaded with a 5 mg/ml solution of iron oxide (Fe 3 O 4 )
the electrospinning of the solution; this condition can NPs with oleic acid as a surface ligand in toluene
result in the formation of aggregates (depending on the (CytoDiagnostics, Inc., Burlinton, Canada), which
flow rate) that will not only be incorporated onto the was used without further purification. Four NP sizes
fibers but could also provoke nozzle clogging and affect were used in this study: 5 nm (batch n :
the fabrication process. O0124817E3127), 10 nm (batch n : O0228119F31380),
Coaxial electrospinning has emerged as a straightfor- 20 nm (batch n : O0328119F31381) and a mix of
ward alternative to conventional electrospinning for pro- 10 and 20 nm (batch n : O317217F29229), which are
ducing well-dispersed nanoreinforced polymers by here referred as NP5, NP10, NP20, and NP10-20, cor-
injecting the components in a core-shell fashion.[20–23] respondingly. Both core and shell syringes were
Moreover, this technique has been useful in the produc- loaded onto separate automatic KDS-100 pumps
tion of NP/polymer mats with improved properties such (Kd Scientific) whose flow rates were set at 0.005 and
as thermal stability, photodegradation performance and 0.020 ml/min, respectively, and their content injected
superparamagnetic properties.[2,24,25] through silicone tubings into a coaxial needle system
In this work, we employed coaxial electrospinning (Spraybase, Maynooth, Ireland). A distance of 18 cm
as a one-step alternative to fabricate nanoreinforced was kept between the coaxial needle and a flat collec-
polyvinylidene fluoride (PVDF) featuring the disper- tor while a voltage of 17.5 kV was imposed between
sion of iron oxide (Fe3O4) NP with different size distri- them throughout the whole process. The total
butions (monomodal or bimodal) within the polymeric electrospinning time per sample was 20 min.
NAVARRO OLIVA ET AL. 1567
FIGURE 1 Schematic representation of the electrospinning process [Color figure can be viewed at wileyonlinelibrary.com]
F I G U R E 2 Transmission electron
microscopy images of nanoparticle
(NP) solutions and size distribution
for NP5 (A), NP10 (B), NP20 (C), and
NP10-20 (D)
F I G U R E 3 Scanning electron microscopy images of electrospun polymeric matrices (polyvinylidene fluoride [PVDF], PVDF + NP5,
PVDF + NP10, PVDF + NP20, and PVDF + NP10-20) and their corresponding fiber size distribution. Insets: zoom-in to the fibers in the
composite images. Scale bar = 1 μm
Fiber
NP size (nm) size (nm)
Pure PVDF — 250.5 ± 66.7
PVDF + NP5 5.95 ± 0.98 153.6 ± 36.6
PVDF + NP10 9.28 ± 1.15 189.4 ± 42.1
PVDF + NP20 17.30 ± 2.25 215.1 ± 42.7
PVDF + NP10- 16.56 ± 1.41, 10.57 ± 1.61 214.1 ± 42.4
20
F I G U R E 5 Schematic representation of coaxial electrospinning comparing two different size distributions. The main difference is based
on the nanoparticle (NP) surface-to-volume ratio: for a given fraction volume the total amount of NP is different because of their size
distribution, exhibiting a higher negatively-charged surface area for NP5 (1), increasing the electrostatic attraction and thus stretching the
fibers to a higher extent compared to NP20 (2). This stretching effect results in reduced fiber diameter in polyvinylidene fluoride [PVDF]
+ NP5 matrices compared to PVDF + NP20 [Color figure can be viewed at wileyonlinelibrary.com]
F I G U R E 6 Transmission
electron microscopy images of pure
only polyvinylidene fluoride
[PVDF], PVDF + NP5, PVDF
+ NP10, PVDF + NP20, and PVDF
+ NP10-20 matrices. Insets: zoom-in
to individual fibers. White arrows
indicate the location of NP on the
edges of the fibers
NAVARRO OLIVA ET AL. 1571
F I G U R E 7 Size distribution of nanoparticle (NP) in their solvent and in the composite for NP5 (A), NP10 (B), NP20 (C), and NP10-20
(D) [Color figure can be viewed at wileyonlinelibrary.com]
incorporation of either iron oxide NPs into the polymeric cellulose-nanocrystals in the composite fibers, which con-
network caused a qualitative change in the roughness com- firms the role of conductivity and viscosity of the spin-
pared to only-PVDF matrices (more grooved surfaces can ning solutions not only in the final diameter of the fibers,
be observed). Moreover, there is an apparent decrease in but also in their smoothness (or roughness) and general
the fiber size distribution of the composites compared to structure.[2] Importantly, none of our samples presented
the pristine matrices (only-PVDF = 250.5 ± 66.7 nm; the formation of visible bead structures in the fibers,
PVDF+NP5 = 153.6 ± 36.6 nm; PVDF+NP10 = which indicates that our coaxial electrospinning parame-
189.4 ± 42.1 nm; PVDF+NP20 = 215.1 ± 42.7 nm; PVDF ters are adequate.
+NP10-20 = 214.1 ± 42.4), which can be ascribed to an Samples were observed on the TEM to corroborate
increased polarization (due to the presence of negatively the presence of NP in the PVDF matrix (Figure 6): the
charged NP) in the electrospinning jet, stretching the inclusion of each NP distribution can be observed as well
fibers to a longer extent and reducing, thus, their diam- as a good dispersion within the fiber, with negligible for-
eter[31] (graph in Figure 3, Figure 4, Table 2). This mation of agglomerates in both types of composites.
effect is more pronounced in the samples where NP5 Interestingly, some of the reinforced fibers showed the
(smallest NP) are present, which we assumed it is presence of NP located on their edges rather than the
unequivocally related to the increased surface-to- core (especially in the PVDF + NP10-20 composites).
volume ratio provoking a fiber size reduction This phenomenon, along with other factors such as
(Figure 5). (inner) flow and jet bending instabilities[32,33] could
Our results are in accordance with Li et al: they found account for the appearance of the grooved morphologies
that (core-shell structured) nanoreinforced composites seen by SEM.
exhibited a reduction in average fiber diameter compared Further evaluation of NP size revealed that the diam-
to their only-polymer versions. Moreover, this size decre- eter distribution was well preserved in the final material
ment was accentuated with increasing concentrations of (Figure 7) compared to the initial NP solutions (ie,
1572 NAVARRO OLIVA ET AL.
[27] N. Joseph, J. Varghese, M. T. Sebastian, Compos. Part B Eng. [33] Y. Zhang, Z.-M. Huang, X. Xu, C. T. Lim, S. Ramakrishna,
2017, 123, 271. Chem. Mater. 2004, 16, 3406.
[28] I.-H. Kim, D. H. Baik, Y. G. Jeong, Macromol. Res. 2012,
20, 920.
[29] F. Qiu, H. T. Tong, Y. Y. Gao, Q. Zou, B.-X. Dong, Q. Li, J.-E.
Chu, F. Chang, S.-L. Shu, Q.-C. Jiang, Materials 2018, 11, 1284. How to cite this article: Navarro Oliva FS,
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6117683/. Picart L, Leon-Valdivieso CY, et al. Coaxial
[30] T. Takayama, M. Todo, A. Takano, J. Mech. Behav. Biomed. electrospinning process toward optimal
Mater. 2009, 2, 105. nanoparticle dispersion in polymeric matrix.
[31] S. Huang, L. Zhou, M.-C. Li, Q. Wu, Y. Kojima, D. Zhou,
Polymer Composites. 2021;42:1565–1573. https://
Materials 2016, 9, 523.
[32] B. Zaarour, W. Zhang, L. Zhu, X. Y. Jin, C. Huang, Tex. Res. J.
doi.org/10.1002/pc.25924
2018, 89, 2406.