Efficient single-crystal 4-N,N-dimethylamino-4-N-methyl-stilbazolium tosylate (DAST) on the optical, piezoelectric coefficient and structure of P(VDF-HFP) copolymer for energy conversion systems

Organic crystal salt 4-N,N-dimethylamino-4′-N′-methyl-stilbazolium tosylate (DAST) with 99% pureness was supplied by Fluka. P(VDF-HFP) (10 wt% HFP; molecular weight of 400,000 g/mol; Solef 21,216) was obtained from Solvay (Brussels, Belgium). DMF (dimethylformamide with purity 99%) was obtained from Merck.

Copolymer P(VDF-HFP) and DAST dye at different concentrations were dissolved separately in dimethylformamide (DMF) at room temperature and then mixed with each other with continuous stirring to obtain a homogeneous solution. Before pouring the solution, glass slides were washed and cleaned thoroughly with distilled water and then oven-dried for 15 min at 353 K to remove any residual moisture. Solutions of P(VDF-HFP)/DAST composites were poured onto glass slides and dried at 343 K for 2 h in an oven, to ensure that the remaining DMF solvent was evaporated. The sample thickness was measured with a digital micrometer (no. 292-521-30, Mitutoyo, Japan) and found to be ranged from 20 to 35 μm. The expected interaction between P(VDF-HFP) and DAST dye is explained in Scheme 1.

The presence of quaternary ammonium salt in the chemical structure of the DAST dye causes two possibilities to occur for its interaction with the fluorinated copolymer P(VDF-HFP) as shown in Scheme 1. The first expected mechanism can be illustrated by hydrogen bonding with the highest electronegative fluorine atoms [16]. However, two systems of a weak H-bond may be formed between sp³ CH host or donor and sp² CH groups guest or acceptor, such as in C–H/π H-bond and C–H/conjugated benzene rings in polymer and dye, respectively. The high withdrawing effect of the octafluoro atoms in P(VDF-HFP) polymer qualifies the tetrahedral sp³ CH to be a strong guest for H-bond complex formation [17]. On the other hand, the ion–dipole interaction can be considered as the second pathway, as a result of the electrostatic interaction between the cationic quaternary ammonium part and the negatively partial charged fluorine atoms [18].

The crystalline structure was exanimated using X-ray diffraction (XRD Ultima II, Rigaku, Japan, electron probe current 40 mA and accelerating voltage 40 kV) in the range of 3°–65° with a resolution of 0.02° for all prepared samples. Measurements of FTIR (Fourier transformation infrared) were achieved using a Vertex 70, Bruker Optics (Germany) in an ATR mode at normal temperature of 4000–500 cm⁻¹. UV–Vis measurements of all samples were recorded using a Jasco V-570 spectrometer in the range from 190 to 1100 nm with a resolution of 5 nm. The optical polarized microscope (POM) (Jiangnan Novel, PM6000, China) was used to examine the surface morphology of the samples.

A homemade cell was used to perform the measurements of TSDC for all samples [19]. The samples were polarized in an oven at T_p = 343 K for polarizing time t_p = 15 min with various polarizing electric fields, E_p. Then, the samples were cooled down slowly to room temperature in the applied field to freeze the polarized dipoles. After removing the applied electric field, the samples were short-circuited for 15 min to remove any stray charges. The samples were reheated again with a linear heating rate of 3 K/min, and the depolarized current I(T) was recorded through a Keithley 485 picometer. Measurements of the pyroelectricity were taken by cooling the sample again to room temperature after recording TSDC (first cycle) and then starting to heat the sample again at the same heating rate to record the resulting current (second cycle).

For the piezoelectric measurements, the samples were polarized by applying constant electric field in the order of E_p = 1.1 × 10⁷ V/m at polarizing temperature T_p = 363 K for polarizing time of t_p = 30 min. The piezoelectric current was measured for our samples at various measuring temperatures (T_m) and different applying values of stress. The piezoelectric current was recorded by a Keithley 485 picometer, and then, the piezoelectric coefficient d₃₃ was calculated.

XRD pattern helps to provide the effect of DAST dye on the crystalline phases of PVDF-HFP copolymer, especially α and β phases. Figure 1 displays XRD pattern of P(VDF-HFP) and P(VDF-HFP)/DAST composite samples. The peaks at 18.17°, 19.73°, 26.49°, 36.01° and 38.75° reveal the semicrystalline nature of PVDF-HFP copolymer. The diffraction peaks at 18.17°, 19.73° and 26.49° correspond to (020), (110) and (021) planes of the monoclinic α-phase [19]. The strong diffraction peak and weak peak at 19.91° and 36.01° correspond to (110)/(200) and (020) planes of the orthorhombic β-phase, which confirms the existence of the ferroelectric β-phase [20].

With increasing the content of DAST dye in the composite samples, some of the characteristic peaks of α-phase such as 18.17° and 26.49° almost disappeared and the intensity of the peak at 19.73° decreased. Also, the diffraction peak at 36.01° in pure P(VDF-HFP) is shifted toward a lower diffraction angle to appear at 35.49°, while the diffraction peak at 38.76° in pure PVDF-HFP is shifted to higher diffraction angle to appear at 40.07° in the highly doped composite sample. All these observations suggested that the polar phase of P(VDF-HFP) successively increased after doping with DAST dye. It means that β-phase fraction in the P(VDF-HFP)/DAST composite samples is improved, indicating that the P(VDF-HFP)/DAST composite may show a better piezoelectric response [21].

Different structural quantities such as crystallite size (D), lattice strain (ε_L), number of crystals per unit surface area (N_c) and dislocation density (δ) are important to obtain sufficient information about the structural transformation that occurred in the samples of the overlays. The values of crystallite size (D) and lattice strain (ε_L) are estimated by Debye–Scherrer formula as follows [22]:

The number of crystallites per unit surface area (N_c) and dislocation density (δ) are estimated using the following expression [23]:where d is the sample thickness. The calculated structural parameters of all samples are summarized in Table 1.

Figure 2 displays FTIR spectra of P(VDF-HFP) and P(VDF-HFP)/DAST composites. The bands at 3012 cm⁻¹ and 2970 cm⁻¹ in the spectrum of P(VDF-HFP) copolymer are attributed to the asymmetric and symmetric stretching of CH₂ group, respectively [24]. The transmission band at 1402 cm⁻¹ represents a combination between the three phases, i.e., α, β and γ [25, 26]. The bands at 1172 cm⁻¹ and 1074 cm⁻¹ are related to a combination of β and γ phases [25]. The band at 988 cm⁻¹ is attributed to the non-electroactive α phase and assigned as C–F stretching vibration [27]. Moreover, FTIR of pure P(VDF-HFP) copolymer involving many transmission bands at 836 cm⁻¹ is related to β-polar phase and assigned as (CF₂ stretching), 762 cm⁻¹ (CF₂ bending), 615 cm⁻¹ (skeletal bending), 511 cm⁻¹ (CF₂ stretching) and 489 cm⁻¹ (CF₂ waging) [4, 28, 29]. The transmission bands at 1402 cm⁻¹, 950 cm⁻¹, 762 cm⁻¹, 606 cm⁻¹ and 489 cm⁻¹ are attributed to α-phase, while the bands at 836 cm⁻¹ and 511 cm⁻¹ correspond to β-phase, as shown in Fig. 2a [12, 30]. These results suggest both α and β-phases are formed by the VDF segments in the P(VDF-HFP) copolymer. On the other side, with increasing the DAST dye content, the α-phase bands become much weaker, while the intensity of β-phase bands increases, confirming that the addition of DAST dye can stimulate β-phase formation during the initial crystallization process, as shown in Fig. 2a. FTIR spectrum of DAST dye displayed many transmission bands at 1640 cm⁻¹, 1577 cm⁻¹, 1524 cm⁻¹, 1271 cm⁻¹ and 1330 cm⁻¹ [31]. The characteristic bands and their assignments of P(VDF-HFP)/DAST composites are listed in Table 2.

The relative β-phase fraction F(β) in PVDF-HFP/DAST composites is estimated using the following formula [32]:where A_β and A_α are the absorbance at both 836 cm⁻¹ and 762 cm⁻¹, respectively. It is found that the relative β-phase fraction is increased with increasing the DAST dye content in PVDF-HFP/DAST composites, indicating that DAST dye has a clear effect on the transformation of α phase to β-phase, as shown in Fig. 2b. The relative β-phase fraction became ~ 93.10% for P(VDF-HFP) after doping with 10 wt% of DAST dye. Hence, we can say that the nucleation of the electroactive β-phase is enhanced in P(VDF-HFP) copolymer after the doping with DAST dye. The gradual enhancement of β-polar phase after doping P(VDF-HFP) with DAST may be due to the interfacial electrostatic interaction between positively charged –CH₂ dipole of P(VDF-HFP) copolymer and DAST dye [33]. This interaction will increase the zigzag conformation (TTTT) of the P(VDF-HFP) main chain which in turn will enhance the β-phase [34]. Thus, DAST dye acts as a β-phase-promoting nucleation agent. The gradual enhancement of interfacial interaction between CH₂ dipoles of P(VDF-HFP) and DAST dye can be confirmed by investigation their transmission FTIR bands in the region from 3080 to 2900 cm⁻¹, where the stretching vibrations of –CH₂ dipoles are responsible for the existence of these bands. The shift of the –CH₂ stretching vibration bands toward the lower energy region confirmed the gradual enhancement of the interfacial interaction in P(VDF-HFP)/DAST composite samples, as shown in Fig. 2c [24]. The interaction between the charges distributed on the DAST dye surface and –CH₂ dipoles acting as a damping source for the –CH₂ dipoles oscillation results in a shift of the vibrational bands toward a lower energy region (lower wavenumber). This behavior is interpreted using the damped harmonic oscillation of –CH₂ dipole [35].

The spherulitic morphology of a P(VDF-HFP) and its composites could be observed using the polarized optical microscopy (POM). Figure 3 displays the optical micrographs of P(VDF-HFP) and P(VDF-HFP)/DAST composite samples with various concentrations of DAST dye. The morphology of P(VDF-HFP) copolymer displayed a spherulitic pattern. It was found that the P(VDF-HFP) copolymer had an unequal-grained rough surface with small dark pores with a laminar distribution confirming the semicrystalline nature of P(VDF-HFP), as illustrated in Fig. 3a [36]. Evaporation of the solvent may lead to the formation of dark areas of the pores. The crystalline α-phase of P(VDF-HFP) copolymer is thus confirmed due to the presence of the spherulites [37, 38]. The P(VDF-HFP)/DAST composite samples showed that the DAST dye was uniformly distributed without agglomeration in the copolymer matrix and the porosity was also reduced, as illustrated in Fig. 3b–e. Also, it is observed that the dispersion of DAST dye becomes more uneven with increasing the DAST dye content, as shown in Fig. 3d, e. With increasing the DAST content increases, the spherulites interact with each other forming a larger spherulites.

Figure 4a displays the optical absorption (A) of P(VDF-HFP) and P(VDF-HFP)/DAST composite samples. It can be seen that the absorption increases with increasing the content of DAST dye in composites because the dipoles number was increased, which indicates that a non-polar α-phase transformation into a β-polar phase has taken place [39]. The absorption spectrum of pure P(VDF-HFP) showed an absorption peak at 200 nm and is red shifted to appear at 237 nm in the highly doped composite sample. This shift is attributed to the strong electrostatic interaction between DAST dye and P(VDF-HFP) copolymer main chain. Also, the absorption spectra of P(VDF-HFP)/DAST composites showed a shoulder centered at 480 nm and a peak at 540 nm in the low doped samples and are red shifted to appear at a shoulder centered at 490 nm and a peak at 587 nm with increasing the DAST in the composite samples. The shoulder at 480 nm is due to the absorption of DAST molecules. The red shift in the absorption bands of DAST dye in the composite samples may be due to the J-aggregation of the chromophores as reported previously [13, 40]. The extinction coefficient (k) is estimated using the absorption coefficient α, as k =  αλ/4π, where α = 2.3 A/t, in which t is the thickness of the sample. The variation of the extinction coefficient against the wavelength is shown in Fig. 4b. It is observed that with increasing the DAST dye content in the composite samples k increases, indicating higher light dissipation due to absorption and scattering by DAST centers in agreement with the results of Mott and Davis [41]. The strong interaction between DAST and P(VDF-HFP) copolymer leads to a change in crystallinity and thus a change in the band structure and absorption ratio. In general, it is observed that the extinction coefficient (k) in the UV region decreases with increasing wavelength, while in the visible region (400–1100 nm) it starts to increase again with increasing wavelength. In the UV region, the incident photons have a suitable energy that enables them to excite the electrons for overcoming the band gap, and thus, the extinction coefficient values will decrease. Low values of k in this region indicate that electromagnetic waves can pass through P(VDF-HFP)/DAST composites without any damping. On the other hand, the photon energy does not enable it to excite the electrons from one state to another in the visible light region, so this energy due to scattering or reflection will be lost, which causes the k values in the visible region to increase again [42‐44].

The optical band gap (E_g) of pure and P(VDF-HFP)/DAST composite samples is calculated based on Tauc’s equation as follows:where hν is the incident photon energy and B is a constant, depending on the probability of the transition in the optical wavelength region. The index y gives valuable information about the electronic transition nature within the materials. For the direct and indirect allowed transitions, the index y will be ½ and 2, whereas for the direct and indirect forbidden transitions it will be 3/2 and 3, respectively. In the present work, the allowed direct transition is considered to estimate the optical band gap (E_g) for our samples, as shown in Fig. 4c.

By extrapolating the linear portion on the x-axis, i.e., (αhν)² = 0, the optical band gap (E_g) is determined and listed in Table 3. It is observed the values of E_g are reduced with increasing the DAST content and changed from 5.41 eV for P(VDF-HFP) to 3.75 eV for P(VDF-HFP)/10 wt% DAST composite samples. This reduction may be due to the ability of DAST dye to enhance charge transfer between the DAST and P(VDF-HFP) matrix, leading to a dramatic change in the band structure of the composite samples [45].

The refractive index (n) dispersion is an important factor for designing highly efficient optical devices [46]. Dispersion parameters such as dispersion energy (E_d), effective oscillator energy (E₀), static refractive index (n₀) and static dielectric constant (ε_s) can be estimated by applying the single oscillator (SOM) model below the interband absorption edge using the following equation [47]:

Figure 5a illustrates the dependence of (n²− 1)⁻¹ on (hν)² for all samples. The values of E_d and E₀ are determined by knowing the values of intercept (E₀/E_d) and slope (1/E₀E_d) of Fig. 5a and given in Table 3. It is noted that as the DAST content increases in the composite samples, both E₀ and E_d decrease. The decrease in the values of E₀ and E₀ can be attributed to the microstructure change in P(VDF-HFP) after doping with DAST dye. The values of E₀ and E₀ are employed to estimate the static refractive index (n₀) and static dielectric constant (ε₀) at zero frequency using the following equation:

The values of n₀ and ε_s for pure and P(VDF-HFP)/DAST composites are calculated and given in Table 3. It is noted that both n₀ and ε_s values are decreased with DAST dye content increase in the composite samples. The relation between the square of refractive index (n²) and wavelength (λ²) can be expressed as follows:where ε_L is the lattice dielectric constant at high frequency, ε₀ = 8.85 × 10^–12 Fm⁻¹, c = 3 × 10⁸ m/s, e = 1.6 × 10^–19 C, N is the carrier charge concentration and m^* is the effective mass of the carrier. The plot of n² against λ² yields straight lines with intercept of ε_L and slope of \(\frac{1}{{4\pi \varepsilon_{0} }}\left( {\frac{{e^{2} }}{{c^{2} }}} \right)\,\left( {\frac{N}{{m^{*} }}} \right)\), as illustrated in Fig. 5b. The values of ε_L and (N/m^*) are estimated and given in Table 3. The results revealed that the values of lattice dielectric constant (ε_L) are higher than the values of static dielectric constant (ε_s). The variation between values of ε_L and ε_s can be attributed to the increase of free carriers concentration and the polarization process that arises in the material when it is exposed to light [43, 48]. The plasma frequency values of P(VDF-HFP)/DAST composite samples are determined and listed in Table 3 based on the following equation:

It is found that the plasma frequency depends on the DAST content, that is, it increases with increasing the DAST content. This behavior is attributed to the high values of charge carrier concentration.

The simple empirical equations according to the generalized Miller’s rule are used to calculate the linear and nonlinear optical parameters of P(VDF-HFP)/DAST composites such as the first (χ⁽¹⁾) and third (χ⁽³⁾) order of the optical susceptibility and nonlinear refractive index (n₂). These parameters are estimated by combining Miller’s generalized rule and the relations concerning (χ⁽¹⁾) and (χ⁽³⁾) containing the parameters of (SOM) model as follows [49]:

The values of χ⁽¹⁾, χ⁽³⁾ and n₂ are estimated and given in Table 3. It was found that as the DAST content in the composite samples increased, the value of these parameters increased.

TSDC technique has proven to be a powerful method for understanding the defect properties both quantitatively and qualitatively. It has been applied to polymers and polymer composites to interpret the different relaxation mechanisms for decades [50‐56]. Figure 6a displays TSDC global spectrum of pure P(VDF-HFP) samples in the temperature range from 300 to 395 K polarized under various electric fields. It is found that TSDC spectrum of P(VDF-HFP) is characterized by two intense relaxation peaks. The first peak is detected in the temperature range of 327–335 K, while the second peak is detected at about 375 K. Generally, the maximum current (I_m) of each relaxation peak is found to be field dependent, and it increases with increasing the applied electric field (E_P), as illustrated in the inset of Fig. 6a. It is well known that P(VDF-HFP) is a polar semicrystalline copolymer. The contribution of P(VDF-HFP) to the polarization process may be attributed to the dipole’s orientation and the formation of space charge/charge carrier injection from the metallic electrode at high temperatures under applying the electric field. The first peak which is observed in the range 327–335 K can be assigned as α-relaxation. The α-relaxation in P(VDF-HFP) is referred as a dipole relaxation process that arises as a result of the localized rotational fluctuations of the dipoles [57]. The α-relaxation in P(VDF-HFP) can be related to the dipole’s orientation of the polar side group (CH₂–CF₂). This relaxation peak has been reported for PVDF blends or composites of PVDF previously in the literature [58, 59]. The second relaxation peak at high temperature can be designated as ρ-peak which may arise from the charge carrier injection of from the metallic electrodes to the polymer surface, which is immobilized through the sample polarization process. P(VDF-HFP) may contain a large number of impurity molecules before the electric field treatment, and these molecules will separate into different ionic species. Hence, it is appropriate to attribute the relaxation peak that appear in the high temperature range to the relaxation of space charge, which arises from the injection of charge carriers from the electrodes into the polymer–metal interface.

The depolarizing current of group of dipoles as a function of the temperature is expressed as follows [60, 61]:where Q is the total charge released, E_a is the activation energy, β is the heating rate (dT/dt), k_B is the Boltzmann constant, T_o is the initial temperature at which the depolarization current begins to appear and τ_ο is a pre-exponential factor and is expressed as follows:where τ(T) is the relaxation time at a given temperature (T). Discharge current against the temperature to release the trapped charge carriers has the same form as Eq. 1. Hence, the shape of the thermal current spectrum generated by the traps as well as from dipoles is similar to each other. The low-temperature tail of Eq. 13 is expressed as follows:

Hence, the activation energy (E_a) values of the depolarization process can be estimated using the slope of Eq. 15. When differentiating Eq. (13) with respect to temperature and equating the resulting equation to zero (i.e., dI/dT = 0), we can obtain the maximum temperature (T_m) at which the maximum current (I_m) will occur. The maximum temperature (T_m) is given by:

The initial rise method has been applied based on Eqs. (15 and 16) to calculate all the molecular parameters of different relaxation processes of pure P(VDF-HFP) [62], as depicted in Fig. 6b, c and listed in Table 4.

The effect of DAST dye on TSDC spectrum of P(VDF-HFP) at constant poling temperature and poling electric field is illustrated in Fig. 7. It is found that as the content of DAST dye is increased, the current (I_m) of each relaxation peak of P(VDF-HFP)/DAST composite samples is increased in magnitude and also the temperature (T_m) of each relaxation peak is shifted to a higher temperature. The activation energy (E_a) of each relaxation process of P(VDF-HFP)/DAST composite samples has been estimated using Eq. 15 as illustrated in Fig. 7b, c and given in Table 5. It is noted that the value of E_a for α-relaxation of pure P(VDF-HFP) is ranged from 0.42 to 0.53 eV, while that for P(VDF-HFP)/DAST composites is ranged from 0.11 to 0.42 eV. On the other hand, the value of E_a for ρ-relaxation of pure P(VDF-HFP) is ranged from 0.23 to 0.36 eV, while that for P(VDF-HFP)/DAST composites is ranged from 0.11 to 0.18 eV. The decrease in the values of E_a in the composite samples comparing to pure P(VDF-HFP) confirms the interaction between DAST dye and P(VDF-HFP) leading to an increase in the amorphous character of P(VDF-HFP). Our results are in good agreement with the reported data [58, 59].

The effect of DAST dye on the pyroelectric coefficient of P(VDF-HFP) is investigated, and the pyroelectric current is recorded against the temperature with a heating rate of 3 K/min, as shown in Fig. 8. The measured pyroelectric current expresses the polarization that induced from thermal stress. The pyroelectric coefficient is expressed as follows:where I, P and A are the pyroelectric current, pyroelectric coefficient and electrode area, respectively. It is noted that as the temperature of the samples increased as well as the content of DAST in the P(VDF-HFP)/DAST composites increased, the pyroelectric coefficient value increased. Furthermore, the composite samples are found to be very sensitive in the temperature range of 337–395 K, where the pyroelectric coefficient changed from 0.69 to 4.12 × 10^–5 C/m² K. This may be attributed to the structural relaxation process at low temperature due to the molecular motion in the crystalline region [63]. These results revealed that the embedding of DAST dye improved the pyroelectric properties of P(VDF-HFP) and thus modified the polarization behavior of the samples. In other words, it can be said that the incorporation of DAST dye into P(VDF-HFP) matrix enhances directly the β-phase of P(VDF-HFP) as verified from FTIR spectra, and β-phase causes the strong pyroelectric and piezoelectric response of P(VDF-HFP) [62]. The high pyroelectric activity of composite samples gives a real possible application as temperature sensor, especially in the temperature range 337–395 K [64].

Figure 9 displays the variation of the piezoelectric coefficient (d₃₃) of P(VDF-HFP) and P(VDF-HFP)/DAST composite samples as a function of stress at different measuring temperatures. The samples are polarized at T_p = 363 K with electric field of E_p = 1.2 × 10⁷ V/m. Generally, it is found that the piezoelectric coefficient is temperature and stress dependent. During the poling process, a portion of α-phase may be transferred into β-phase and the dipoles of β-phase will align perpendicularly to the sample. The formation of β-phase in P(VDF-HFP) can be improved using various methods, such as polymer blending, poling under high electric field and the crystallization under high pressure [38, 65, 66]. The main function of the electric field is to unify the dipole directions of the β-phase, and thus, the piezoelectricity is expected to be improved more significantly at the higher field as shown in Fig. 9a [5, 67]. Also, it is found that the piezoelectric coefficient of P(VDF-HFP)/DAST composites has been enhanced with increasing the DAST ionic crystals content, as shown in Fig. 9b. The enhancement of piezoelectric coefficient of composite samples can be ascribed to relative extents of polar β-phase within the composites after doping with DAST, as confirmed from the analysis of X-ray data and FTIR spectra. In other words, the enhancement of β phase in P(VDF-HFP)/DAST composites and hence the piezoelectric coefficient could be ascribed to the interaction between the oppositely charged surface of DAST and the dipoles –CF₂–CH₂– of P(VDF-HFP). Also, improvement of the DAST crystals nucleation and surface charge polarization can produce more stable piezoelectric β-crystal phase [68]. The piezoelectric coefficient of highly doped composite sample, P(VDF-HFP)/10 wt% DAST, is investigated as a function of stress at different measuring temperatures, as shown in Fig. 8c. It is observed that the piezoelectric coefficient of the highest doped sample is increased with increasing stress and temperature and changed from 20.2 pC/N at T = 313 K to 34.2 pC/N at T = 353 K at stress of 6.27 × 10⁵ Pa.