Chao Fuab,
Xuemei Wanga,
Xiang Shiab and
Xianghai Ran*a
aLab of Polymer Composites Engineering, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China. E-mail: ranxh@ciac.jl.cn; Fax: +86 431 85262424; Tel: +86 431 85262677
bUniversity of Chinese Academy of Sciences, Beijing 100049, China
First published on 2nd October 2015
The desired electroactive β and γ phases of poly(vinylidene fluoride) (PVDF) nanostructures are rarely obtained through wetting methods using anodic aluminum oxide (AAO) as a template. Further and systematic studies on the corresponding induction mechanism are also lacking. Herein, we designed and fabricated PVDF nanowires using a solution wetting method with pristine and modified AAO templates. The morphology and crystalline structure of the PVDF nanowires were characterized by SEM and micro-FTIR respectively. Then, the induction mechanism was investigated by AFM, FTIR, TGA and contact angle measurements. It is found that the polarity of the solvent and the surface hydroxyl groups on the template nanopores both have an inductive effect on the polar phase. After the oxygen plasma treatment, the proportion of β phase of the PVDF nanowires becomes higher although the total polar phase content remains the same. This is attributed to the high polarity of the nanopore surface and regular arrangement among the hydroxyl groups. The polar phase content of the PVDF nanowires increases from 40% when prepared by the pristine template to 71% when prepared by a 3-aminopropyltrimethoxysilane (APMS) modified template. It can be explained that more interaction points and stronger interactions result in the formation of more of the electroactive phase.
PVDF presents polymorphism, and has three major crystalline phases: a nonpolar α phase (TGTG), a polar β phase (TTT) and a γ phase (T3G+T3G−).20 The nonpolar α phase is obtained in common processing conditions. Though the polar β phase and γ phase display piezoelectricity, they are not easy to obtain due to their thermodynamic instability.21 Moreover, the β phase content has a profound influence on the degree of polarization. The copolymer poly(vinylidene fluoride–trifluoroethylene) (P(VDF–TrFE)) crystallizes predominantly in the β phase due to steric factors. These P(VDF–TrFE) nanostructures therefore exhibit the electroactive phase.19,22–26 Compared with P(VDF–TrFE), PVDF presents lower density, stronger dipole interactions, and more reactive synthesis character.27 Therefore, it is significant to obtain PVDF nanostructures with a small ferroelectric domain size (low coercive voltage) as well as with a high content of the electroactive phase (high remnant polarization).
When PVDF melt is infiltrated into AAO templates, the nonpolar α phase is obtained,28,29 while when the PVDF pellets are dissolved in a solvent with a higher dipole moment and then the solution is cast onto a template, the β and γ phases are derived.30–33 In similarity to other solution-based polymer processing methods, it is reasonable to expect that solution wetting has more controllable factors that may influence the electroactive phase content than melt wetting. However, little study has focused on the mechanism of electroactive phase induction effect with this method. Li et al. reported that the surface hydroxyl groups stabilized the all-trans conformation of PVDF nanostructures in solution conditions, thus leading to the formation of the β phase.32 Kim et al. investigated crystalline phases of PVDF nanostructures crystallized from mixed solvents with different ratios of dimethyl sulfoxide (DMSO) and acetone. They observed more β phase in solutions with a high volume ratio of DMSO, which shows high polarity.33 Both of the above explanations may light up the way in improving the electroactive phase content. Herein, the underlying mechanism for the formation of the electroactive phase of PVDF nanostructures crystallized from solution in AAO templates is studied. The electroactive phase content is further increased by modifying AAO templates with oxygen plasma and APMS. In addition, the induction mechanism is researched.
Micro-FTIR was used to further characterize the crystalline phase of the PVDF nanowires. Micro-FTIR provides a simple and powerful approach for local analysis of a material presenting different appearances under a microscope.35 The TEM image and optical micrograph (inset) of the slice cross-section are shown in Fig. 2a. The nanowires may lean and break when cutting the TEM specimen. So the nanowires shown in the TEM image are really short. The cross-section is composed of bulk film with a thickness of 100 μm and nanowires with an approximate dimension of 50 μm. The right side of the slice is opaque due to the light scattering behavior between the interfaces of numerous nanowires and the left side is transparent due to the isotropy of the bulk. The solution concentration has a strong impact on the flexibility of bulk film on the template surface of the solution wetting samples. The bulk films fabricated by the solution of 15 wt% and 20 wt% are most suitable for cutting slices and application, because a lower concentration results in a fragile bulk film whereas a higher concentration results in a film with quite high toughness.
Fig. 2b displays the micro-FTIR spectra of nanowires crystallized at different temperatures in a pristine AAO template. When the wetted template is dried at 100 °C, the absorption peaks of the nanowires at 613, 675, 763, 854, 975 and 1383 cm−1 corresponding to the α phase are apparent and the intensity of the peak at 838 cm−1 assigned to the overlapping of the β and γ phases36 is much lower. The sample dried at 70 °C shows a much stronger absorption peak at 833 cm−1 that is uniquely assigned to γ phase. This demonstrates that at a higher temperature, the solution-crystallized PVDF nanowires contain more nonpolar α phase. This result is found to be consistent with previous studies for PVDF bulk films crystallized from solution.21,37–39
As reported, the polar phase content of solution crystallized bulk film is influenced by the drying temperature,40 rate of solvent evaporation,41,42 and the properties of different solvents43 and substrates.44 At a low temperature, intermolecular interactions between the fluorine atoms of PVDF and polar moieties of N,N-dimethylacetamide (DMAc) become greater with increasing solvent polarity. The interactions stabilize the all-trans conformation and promote a greater presence of the polar phase.37 Much more solvent with a higher dipole moment always promotes a higher fraction of β phase.33 The drying temperature can influence the polarity of the solvent and the solvent evaporation rate in the whole crystallization process. However, the solution concentration may only affect the solvent evaporation rate at the latter stage of the crystallization process.42 But the nucleation rate associated with different crystal phase nucleation is influenced by the evaporation rate at the initial stage. As a consequence, the solution concentration has a slight influence on the solvent evaporation rate and thus the crystal phase content. It is observed that the nanowires fabricated by solutions of 15 wt% and 20 wt% and dried at the same temperature show almost the same polar phase content (Fig. 1b). This implies that the effect of drying temperature is stronger than that of the solution concentration. So we may conclude that a lower drying temperature with an increased polarity solvent can effectively induce the formation of the polar phase of PVDF nanowires. It also demonstrates that the induction effect of the polar solvent cannot be ignored. A much lower temperature may result in the slower mobility of the polymer chain and more TG+ or TG− conformation defects. Therefore, more γ phase is formed, which is consistent with the observations of García-Gutiérrez.31
In this work, the crystallization of the PVDF electroactive phase is enhanced by modifying the template nanopore surface with oxygen plasma and APMS, as shown in Fig. 2c. At given crystallization conditions, within the pristine and modified templates, the crystalline phase content varies. After the oxygen plasma treatment, a peak at 840 cm−1 appears in the spectrum of the nanowires and shows a stronger intensity peak at 1275 cm−1 assigned to the β phase, although there is no change in the total amount of polar phase. The TG+ or TG− defects in the all-trans conformation of the PVDF chains are decreased by modifying the templates with oxygen plasma, which therefore leads to more β phase being present. Modifying the AAO templates with APMS leads to an increased proportion of the polar crystalline phase. As shown in the spectra, the intensity of the characteristic peak at 763 cm−1 ascribed to the α phase is quite low, and other peaks assigned to the α phase disappeared. The main peak at 840 cm−1, assigned to the β phase, and the shoulder peak at 812 cm−1, assigned to the γ phase, are obvious. It can be concluded that AAO templates functionalised with APMS increase the polar phase content of PVDF nanowires significantly. Therefore, we provide a new method to improve the piezoelectricity of PVDF nanostructures. The relative polar phase content F(p) of the samples crystallized from different conditions is determined from the FTIR spectra based on the following equation,
Samples crystallized at different conditions | F(p) (%) |
---|---|
a Solution concentration-drying temperature-template (P-pristine, OM-oxygen plasma modified, AM-APMS modified). | |
20 wt%-100 °C-P | 32.4 |
15 wt%-70 °C-P | 39.9 |
20 wt%-70 °C-P | 40.5 |
20 wt%-70 °C-OM | 40.2 |
20 wt%-70 °C-AM | 71.3 |
The AFM images that show the topography of the pristine and modified templates are displayed in Fig. 4a–c. The average RMS surface roughness values of the pristine, oxygen plasma treatment, and APMS modified AAO templates are 4.53 nm, 3.85 nm and 4.88 nm, respectively. The O2 plasma treatment renders the dense surface of templates into a sparse morphology with small grains (Fig. 4a and d). As the plasma engraves the surface particles, it influences the microcrystalline structure exhibited in the phase images (Fig. S2a†). The change in topography and microstructure of the Al2O3 surface plausibly results in an oxygen vacancy-rich surface.50 Thereby, oxygen plasma treatment can effectively increase the number of negatively charged OH− and O2− groups on the surface of the Al2O3 surface,51,52 since it is known that the presence of OH− and O2− groups on a substrate surface contribute to attractive interactions between water molecules and the surface.53 The water drop contact angle measurements (Fig. 5a) demonstrate that the surface of the AAO template modified by oxygen plasma is more hydrophilic than the pristine one, since it is known that a more hydrophilic surface shows higher polarity. After the O2 plasma treatment, the high polarity of the surface template reinforces the absorption between the template nanopore surface and the PVDF molecules. The TG+ and TG− conformation defects reduce dramatically due to the regular arrangement among the hydroxyl groups. So oxygen plasma treated AAO templates give a higher content of the β phase, though the total amount of polar phase remains the same.
![]() | ||
Fig. 4 AFM topography image and cross-section along the blue line in the topography image of oxygen plasma treated (a and d), pristine (b and e) and APMS modified (c and f) AAO templates. |
Silanization has been demonstrated to be an effective and flexible method for changing the wetting and adsorption properties of AAO templates.54 The aminosilane coupling agents can readily react with OH groups on the alumina pore surfaces, resulting in an amino derivatized nanopore wall.55 The covalent attachment of the aminosilane to the nanopore surface is identified (Fig. S1†). The APMS functionalized template is slightly rougher with RMS values of 4.88 nm compared to the pristine one. Furthermore, the nanopores appear to be smaller on the APMS functionalised template (Fig. 4c and f), suggesting the formation of aminosilane layers on the template surface. A minor difference displayed in the phase image (Fig. S2†) of the pristine and APMS modified templates may yield the same conclusion. The vibration intensity of the APMS modified template with the intensity of pristine one subtracted shows a peak at 1593 cm−1, which is attributed to an N–H deformation vibration in-plane (Fig. 3b).56 When PVDF nanowires are embedded in the template nanopores, the N–H vibration shifts to 1650 cm−1 with low frequency. This can be explained by the hydrogen bonds that formed between the amino groups on the modified template pore surface and the fluorine atoms on the PVDF chains. The nitrogen atoms on NH2 groups have weaker attraction towards hydrogen atoms, due to their lower electronegativity compared with oxygen atoms on hydroxyl groups. This results in a stronger interaction between the hydrogen atoms on NH2 groups and the fluorine atoms on the PVDF chains.57 So the interactions between the functional groups and the PVDF chains in the nanopores of APMS modified templates are stronger than the interactions in the pristine templates. The greater number of hydrogen bonds and stronger interactions compared to the pristine template improve the polar phase content of the PVDF nanowires. After APMS treatment, the surface of the templates became less hydrophilic, as is displayed in the result of the contact angle measurements (Fig. 5c). So the APMS treatment does not increase the polarity of the template surface. In summary, more interaction points and stronger interactions enhance the polar phase of PVDF nanostructures, rather than the high polarity of the nanopore surface.
The TGA tests verify the induction effect of the template nanopore surface on the polar phase. The first weight loss at about 200 °C observed in the TGA curve of the O2 plasma treated template can be attributed to surface dehydration and loss of the incorporated OH− and O2− groups (Fig. 6a). The weight loss at approximately 550 °C is due to decomposition and total loss of the aminopropyl groups58 from the APMS modified template. This is based on the fact that there is almost no significant weight loss in the AAO templates over this same temperature range. This further confirms that the aminopropyl groups are chemically bonded to the internal surface of the AAO template and are relatively thermally stable. The TGA curves of the nanowires forming in different templates are represented by the curves of the nanowires filled in different templates with the curves of respective templates subtracted (Fig. 6b). As reported, the weight loss of pure PVDF takes place at about 450 °C.59 Weight loss at 70–100 °C is observed on the subtracted curve of nanowires in the O2 plasma treated templates. The weight loss that happens between 70 and 100 °C is associated with the loss of the OH− and O2− groups. Due to the interaction between the OH groups on the nanopore surface and the PVDF nanowires, the thermal stability of the OH groups decreases, and thereby the degradation temperature of the OH groups is lower compared with the groups on the template. In the same way, the interaction between the NH2 groups and PVDF would result in a decrease in the degradation temperature of the aminopropyl groups. The weight loss of aminopropyl groups may overlap with the decomposition of PVDF nanowires, and therefore it exhibits the resulting curve as shown in Fig. 6b(iii).
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra10309f |
This journal is © The Royal Society of Chemistry 2015 |