Metal framework as a novel approach for the fabrication of electric double layer capacitor device with high energy density using plasticized Poly(vinyl alcohol): Ammonium thiocyanate based polymer electrolyte

2.1 Materials

Sigma-Aldrich (Kuala Lumpur, Malaysia) provided poly(vinyl alcohol) (PVA) powder (average molecular weight = 85000–124000), copper(II) chloride (molecular weight = 134.45 g/mol) and GLY (molecular weight = 92.09382 g/mol). HmbG chemicals, EMPLURA, and Timcal provided ammonium thiocyanate (NH₄SCN) (molecular weight = 76.12 g/mol), N-Methyl-2-pyrrolidone (NMP) (molecular weight = 99.13 g/mol), and carbon black (molecular weight = 12.01 g/mol), respectively. Magna value delivered both of the polyvinylidene fluoride (PVdF) (average molecular weight = ∼534,000 by GPC) and activated carbon (AC) (RP20) (molecular weight = 12.01 g/mol).

2.2 Electrolyte preparation

One gram of PVA was dissolved in 40 mL of distilled water. The solution was stirred using magnetic stirrer at the temperature of 80 °C for around 60 min to prepare the PVA solution. The PVA solution was then left to cool down to room temperature. Then, fixed amount of 50 wt% of NH₄SCN were added to the PVA solutions and stirred constantly with magnetic stirrer at ambient temperature until the salt was fully dissolved in the PVA polymer solutions. Later, PVA:NH₄SCN was included with 10 mL of Cu(II)-complex (synthesis of Cu(II)-complex was described in our previous work in materials and methods section in ref. (Brza et al., 2019). Subsequently, the solutions were mixed and stirred to obtain a good dispersion of Cu(II)-complex in the solution. After that, the PVA:NH₄SCN:Cu(II)-complex were plasticized with 10, 20, 30, and 40 wt% GLY and then the prepared samples were coded as PGNC-1, PGNC-2, PGNC-3, and PGNC-4, respectively. Finally, the plasticized solutions were stirred and mixed until a homogeneous solution was reached and then preserved in the plastic Petri dishes and left for drying at room temperature. For drying better, the fabricated CPE films were preserved in a dessicator comprising silica gel before characterisations.

2.3 Characterization techniques

X-Ray Diffraction (XRD) spectra were achieved through the use of Empyrean X-ray diffractometer, (PANalytical, Netherland) with operating voltage of 40 × 10³ V and operating current of 40 × 10^-3 A. The CPE films were scanned with a CuKα X-ray monochromatic radiation at 0.15406 nm wavelength and the 2θ glancing angle was ranged from 10^° to 80^° with 0.1^° step size.

To study the PVA film as well as CPE films, Fourier Transform Infrared (FTIR) spectrophotometer (Thermo Scientific, Nicolet iS10) was used in the wavenumber range from 4000 to 450 cm⁻¹ and with having 2 cm⁻¹ resolutions.

A Hitachi SU8220 was used to conduct the field emission scanning electron microscopy (FESEM) at 500× magnification. FESEM images were taken to consider the CPE films morphology.

The pure PVA and CPE impedance spectra were accomplished using electrochemical impedance spectroscopy (EIS) [3532–50 LCR HiTESTER (HIOKI)] in the frequencies between 50 Hz and 5 × 10⁶ Hz. The synthesized CPE films were cut for circles with 20 mm in diameter as well as located between stainless steel (SS) electrodes under pressure of a spring. The CPE samples were linked to a computer program to consider real (Z^') and imaginary (Z^“) parts of the complex impedance (Z*) spectra.

2.4 Electrolyte characterization

2.4.1

2.4.1 Ionic transference number analysis

Two types transference number (TNM) were analyzed, which are ionic (t_ion) and electronic (t_el) TNM. The cell preparation arrangement was SS| highest CPE (PGNC-4)| SS. The cell was connected to a V&A Instrument DP3003 digital DC power supply and a UNI-T UT803 multimeter. The circuit diagram for the TNM measurement is shown in Fig. 2. The cell polarized versus time at the working voltage of 0.2 V and ambient temperature. t_ion and t_el can be calculated from Eqs. (1) and (2) (Ramesh et al., 2011):

(1)

$$t_{\mathit{ion}}=\frac{I_i-I_{\mathit{ss}}}{I_i}$$

(2)

$$t_{\mathit{el}}=1-t_{\mathit{ion}}$$

Close

Fig. 2

Illustration of TNM experimental system.

Export to PPT

Here initial current and steady-state current are designated as I_i and I_ss, correspondingly.

2.5 EDLC fabrication

Planetary ball miller (XQM-0.4) was employed to mix the 0.25 g of carbon black and 3.25 g of activated carbon (AC) at 500 rpm for around 20 min. Six metal balls were incorporated to a chamber together with the above powders. Then, 0.5 g of polyvinylidene fluoride (PVdF) was inserted in the solvent of 15 mL N-methyl pyrrolidone (NMP) and then stirred for around 60 min. The powders were transferred into the PVdF-NMP solution and then stirred for 2 h till the emergence of homogenous solution of dense black. An aluminum foil was cleaned using acetone and finally the synthesized homogenous solution was coated on it by using a doctor blade. Subsequently, the gained AC electrodes were dried-out in an oven at 60 °C with the mass loading of active material of 2.43 mg. Finally, the dried AC electrodes were kept in a dessicator with silica gel for drying better. The formed electrodes were cut into circles with area of 2.01 cm². The mass of electrode minus with the mass of aluminum foil was 2.99 mg. From the mass mentioned earlier, AC was 81.25% from the total mass of each circle electrode. Hence, the mass of AC was 2.43 mg and was to calculate EDLC parameters because AC is the active mass.

2.6 EDLC characterization

The AC electrodes were cut into circles with area of 2.01 cm². The utmost conducting PE (PGNC-4) was set in between two AC electrodes and laid in a CR2032 coin cell. Finally, the coin cell was located in a Teflon case as exhibited in Fig. 3.

Close

Fig. 3

Design of the synthesized EDLC.

Export to PPT

The cyclic voltammetry (CV) of the EDLC was taken with Digi-IVY DY2300 potentiostat. The applied potential was in the range between 0 V and 0.9 V with various scan rates. The EDLC displayed in Fig. 3 was exposed with several scan rates of 5, 10, 20, 50, and 100 mV s⁻¹. The specific capacitance C_CV of the EDLC at the several scan rates can be developed from the CV profile through the following relation (Asnawi et al., 2020):

The CV plot area (∫ $I\left(V\right)dV$ ) can be acquired using function of integration in Origin 9.0 software. $v$ refers the scan rate, m refers the mass of activated carbon (2.43 mg), V_f and V_i are the final voltage of 0.9 V and the initial voltage of 0 V, respectively. The EDLC rechargeability is checked via NEWARE battery cycler for 1000 cycles at a current density of 0.5 mA cm⁻² or 0.33 mAg⁻¹. The surroundings for analysis the CV and EDLC charge–discharge was settled at ∼25 °C and relative humidity of ∼50%. The specific capacitance (C_d) using discharge curve can be found from Eq. (4) (Asnawi et al., 2020):

Here i and × denote applied current and gradient of discharge part, respectively. The C_CV and C_d values of the EDLC were compared so as to check the confidence of the results. The situation of the contact between the electrode and CPE of the EDLC device was validated by means of the following equation (Asnawi et al., 2020):

3.1 XRD examination

The XRD pattern of pure PVA and PVA:NH₄SCN:Cu(II)-complex:GLY are displayed in Fig. 4. A couple of peaks at nearly 2θ = 20^◦ and 40^◦ in the XRD spectrum of pure PVA are noted, which have been related to the crystalline structure of PVA in previous investigations (Hema et al., 2008; Hemalatha et al., 2019). It is clear that these peaks still exist in the XRD spectra of PVA:NH₄SCN:Cu(II)-complex:GLY even though their intensities were considerably decreased. More explicitly, the peak at 2θ = 40^◦ almost disappeared, whereas the peak at 2θ = 20^◦ widened. The extension of the amorphous nature is ascribed in the wideness increase and fall in the peak intensity at 2θ = 20^◦ (Aziz et al., 2017; Aziz, 2016).

Close

Fig. 4

XRD spectra for (a) pure PVA, (b) PGNC-1, (c) PGNC-2, (d) PGNC-3, and (e) PGNC-4 films.

Export to PPT

The doping of salt to the host polymer is probable to raise the structure of amorphous nature in PE. In an earlier research, the XRD pattern of starch-chitosan:NH₄Cl SPE system showed a clear increase in the crystalline peaks with emerging additional crystalline peaks at maximum salt concentration (Shukur and Kadir, 2015). This outcome designated the salt recrystallization and ions recombination, which resulted in the decrement of free ions density, and therefore declined the σ_DC (Shukur and Kadir, 2015). Different approaches were specified to overcome these challenges. One of the approaches is the addition of Cu(II)–complex to the PE in order to develop the amorphous structure and increase the σ_DC. The addition of GLY to the electrolyte was also attempted to further improve the σ_DC. The amorphous structure in the electrolyte complexes can be enhanced with the inclusion of GLY (Amran et al., 2016). The existence of GLY in the complex systems helps salt dissociation, which in turn restricts salt recrystallization. Moreover, the plasticizer can produce different pathways for ions conduction; therefore, assisting the polymers to accommodate more salt. These processes raise the amorphous phase in the PEs, subsequently enhancing the electrolyte σ_DC (Shukur and Kadir, 2015). The ion conduction is desirable in the amorphous structure owing to segmental movement of polymer chains (Noor and Isa, 2015; Salleh et al., 2016). Moreover, the absence of any XRD peaks associated with pure PVA, approves the whole dissociation of NH₄SCN salt in the PE (Noor and Isa, 2015). The elimination of hydrogen bonding among the polymer chains causes broadening and weakening in intensity, which signifies the dominance of the amorphous phase in the PE system (Hema et al., 2008).

The addition of GLY and Cu(II)–complex will result in the intensity reduction of crystalline peaks and vanishes the crystalline peak at 2θ = 20^° (see Fig. 4). In our previous work, it was shown that when Cu(II)-complex was added to the PVA host polymer, the amorphous nature expanded, which was noted as a decrease in the intensity of the XRD spectra (Brza et al., 2019). The Cu(II)-complex XRD pattern in our earlier work indicated that the prepared Cu(II)-complex is nearly amorphous, thus the crystalline peaks cannot be perceived through the whole range of 2θ degrees. Obviously, just a hump can be seen from 2θ = 20^◦ to 30^◦ (Brza et al., 2019). The XRD analysis results approve that a good complexation took place between the PVA and the electrolyte components.

The deconvolution technique for the XRD spectra was used to determine the potential amorphous peaks and the crystalline peaks and it was also used to determine the degree of crystallinity (Xc) (Salleh et al., 2016), as shown in Fig. 4. The large and wide peaks signify the amorphous peaks, while the narrow, sharp and small peaks indicate the crystalline peaks. Pure PVA possesses Xc of 41.68, which is noticeably decreased when GLY is inserted. Thus, the amorphous structure of the systems is enhanced with the GLY insertion. It is seen in Fig. 4 that when GLY is added, crystalline peaks in PGNC-1 became smaller and less sharp (Fig. 4 (c-e)). The addition of GLY with 40 wt% gives rise to the smaller crystalline peaks, as seen in the XRD pattern of PGNC-4 (Fig. 4(e)). Eq. (6) was used to obtain Xc for pure PVA and doped PVA and the Xc values are tabulated in Table 1. PGNC-1 system has Xc value of 8.93 while PGNC-4 has the minimum Xc value of 3.54, meaning that the PGNC-4 is the highest amorphous system in this study (Salleh et al., 2016).

Here, A_T and A_C refer the total area of the amorphous and crystalline peaks and total area of the crystalline peaks, respectively, which were accomplished by the deconvolution method using the OriginPro software. The Gaussian function mode was employed for the fitting of the XRD spectra.

3.2 FTIR analysis

The FTIR spectra for pure PVA and PVA:NH₄SCN:Cu(II)-complex:GLY are shown in Fig. 5. The following modifications in the spectral features have been detected after comparing the spectra of pure and doped PVA. C–H rocking of pure PVA is considered the reason for absorption peak at 838 cm⁻¹ (see Fig. 5a,) (Hema et al., 2008). In the case of PVA:NH₄SCN:Cu(II)-complex:GLY, this peak shifted and its intensity decreased; whereas, upon the addition of 40 wt% of GLY it nearly vanished.

Close

Fig. 5

Spectra of FTIR for (i) pure PVA, (ii) PGNC-1, (iii) PGNC-2, (iv) PGNC-3, and (v) PGNC-4 in the range (a) 450 cm⁻¹ to 1900 cm⁻¹, and (b) 1900 cm⁻¹ to 4000 cm⁻¹.

Export to PPT

It is well-known that in NH₄SCN salt–polymer, the charge carrier species are hydrogen ions (H⁺) (Srivastava et al., 1995). The new strong and intense peak detected at 2045 cm⁻¹ is endorsed to aromatic S–C = N stretching of anion of SCN^- group of NH₄SCN. This band shifts with considerable intensity reduction upon the addition of the higher GLY concentration is related to the complexation with PVA functional groups (see Fig. 5b) (Hemalatha et al., 2019; Noor and Isa, 2015). In the ammonium ion NH₄⁺ tetrahedral, since just one of the four protons (H⁺) weakly linked to the nitrogen atom; thus, the H⁺ will transfer to every coordinating site of PVA. The change in peak place with the emergence of the new peak in the PVA:NH₄SCN:Cu(II)-complex:GLY indicates the complex creation amongst the PVA and the NH₄SCN (Hemalatha et al., 2019). The insertion of plasticizer helps further ions dissociation; therefore, additional ions are offered to create interactions with the functional groups of PVA (Liang et al., 2009). Furthermore, the interaction between PVA functional groups and Cu(II)–complex was indicated in our previous work (Brza et al., 2019). This interaction or the adsorption of Cu(II)–complex on the functional groups of the PVA is the two endorsed clarifications for the band intensity reduction (Brza et al., 2019). Consequently, since adsorption causes an increase in the molecular weight, there is a reduction in the functional groups vibrational intensity (Aziz et al., 2017).

In the meantime, O–H stretching vibration of hydroxyl (OH) groups can be related with the wide and robust absorption peak at 3340 cm⁻¹ (Makled et al., 2013). This band possesses a high intensity, almost certainly owing to the robust intra and inters kinds of hydrogen bonding (Hema et al., 2008). Additionally, this band shifts and peak intensity drops in the PVA:NH₄SCN:Cu(II)-complex:GLY owing to its complexation with NH₄SCN salt and Cu(II)-complex. The peak at around 1643 cm⁻¹ in the pure PVA is believed to be originated from the C = O stretching of acetate group (Kharazmi et al., 2015), in which for the doped PVA films it is shifted to a smaller wavenumber. C–H asymmetric stretching vibration is related with a band at 2913 cm⁻¹ (Makled et al., 2013), which also shifted and reduced considerably in the case of doped PVA films. Furthermore, the characteristic of –C–O– stretching vibration in pure PVA is ascribed through the peak at 1076 cm⁻¹ (Jiang et al., 2015), which is shifted with dropped intensity in the doped films with Cu(II)-complex and GLY, as shown in Fig. 5a.

The C–H bending vibration of CH₂ wagging has been recognized as the cause for the pure PVA absorption peak at 1419 cm⁻¹, while C–H deformation vibration has been related with the 1317 cm⁻¹ absorption peak in pure PVA (see Fig. 5a,) (Kharazmi et al., 2015). It is therefore clear that the PVA:NH₄SCN:Cu(II)-complex:GLY samples were related with shifting of these two peaks and a considerable reduction in the intensity of these peaks. These modifications in the FTIR spectra are strong evidences for the interaction of PVA functional groups with the electrolyte components (see Fig. 5a, b).

3.3 Morphological study

Fig. 6a(i-iv) illustrates surface morphology of all the studied films through FESEM images. It is recognized that the surface morphology of the PE films is the key properties to recognize their behavior (Aziz et al., 2019). The images were obtained at 500 × magnification. The small grey lumps can be seen on the surface of the CPE films (see Fig. 6a(i)). These lumps were attributed to the protrude salts. It is evident from the images that the grey lumps are disappeared as the GLY concentration was increased from 20 to 40 wt%. The CPE has a smooth and uniform surface morphology without existing a phase separation as shown in Fig. 6a(ii-iv). It is well documented that smooth morphology appearance is related to the amorphous phase of the PE system (Mobarak et al., 2013). The smooth surface electrolytes can assist conducting ionic species to pass more easily, and therefore increases the value of σ_DC (Mobarak et al., 2013). Likewise, the CPE samples with uniform surface in Fig. 6a(ii-iv) indicates the fine dispersion of Cu(II)-complex. The small white spots appearing on the CPE films surface are attributed to Cu(II)-complex.

Close

Fig. 6

(a) Field emission scanning electron microscopy (FESEM) for (i) PGNC-1, (ii) PGNC-2, (iii) PGNC-3, and (iv) PGNC-4 electrolytes. (b) Schematic illustration of the role of GYL content on ion dissociation.

Export to PPT

This was though, certified to be different in previous works, while extra agglomerated particles on the surface of the CPE films appeared as excessive fillers was included in the polymers. For examples, Hamsan et al. (Hamsan et al., 2017) determined the σ_DC decrement as more GLY was inserted. It has been explained that this is ascribed to the rearrangement of self-linkages of plasticizer producing recrystallization of salt which causes a decrease in σ_DC (Johan and Ting, 2011). Hence, polymer capability is restricted to accommodate extra salts, which in turn caused recrystallization of salt (Aziz, 2017). Lim et al. (Lim et al., 2014) have also connected the surface morphology of CPEs to the σ_DC. In their work, a fall in σ_DC upon addition of 10 wt% TiO₂ has been shown for PVA:LiClO₄:TiO₂ electrolyte systems. They proposed that this observation is due to the agglomeration of TiO₂. They related this to the fact that, the CPE surface morphology possesses great level of rigidity owing to the addition of 10 wt% TiO₂. Thus, these agglomerations could block conducting pathways for ionic movements. Large rigidity in the back bone of polymer also decreases the ion mobility. Consequently, the σ_DC is reduced with the blocking impact and ion transportation is limited.

It is believed that so as to accomplish high σ_DC, the surface has to be smooth uniform. In other words, it is linked to the film amorphous structure (Kadir et al., 2018). The addition of Cu(II)-complex improved the amorphous phase further as shown in our previous work (Brza et al., 2019). In fact, the interaction between the PVA and electrolyte’s components is evinced by the smooth uniform surface morphology of the CPE samples (see Fig. 6a(i-iv)).

In this study, the outcomes display that there is an innovation of the PE preparation as a novel and simple technique to get a great σ_DC and high performing EDLC. The FESEM images are in good agreement with the XRD and FTIR results. From the XRD and FTIR routes, the interaction among the PVA and electrolyte’s components were validated. It is indicated that the CPE surface morphology is observed to be nearly smooth and uniform without having obvious protruded salts at high GLY concentration. From these outcomes, the fabricated CPE was well-defined with amorphous behavior even at the maximum GLY amount. The role of GYL content on increasing free ionic species in the prepared samples is schematically illustrated in Fig. 6b.

3.4 Impedance analysis

The electrochemical and ion transference behaviors of ionic materials, for example electrodes and PEs, can be efficiently examined by the use of electrochemical impedance spectroscopy (EIS) (Cho et al., 2015; Svensson et al., 2005). In this work, EIS was used to analyze the impedance plots for the CPE films (see Fig. 7(a-d)). For PGNC-1 system in Fig. 7a, a semicircle at high frequency region is resulted from the CPEs bulk effect, and a tail at low frequencies can be noted. The tail at the low frequencies happens because of the creation of the EDLC through the free charges buildup at the electrode and electrolyte interface (Aziz et al., 2010). The spike or tail was only displayed by the other samples (see Fig. 7(b-d)).

Close

Fig. 7

Experimental EIS for (a) PGNC-1, (b) PGNC-2, (c) PGNC-3, and (d) PGNC-4 electrolyte films.

Export to PPT

The electrical equivalent circuit (EEC) method is used as a straightforward way for the EIS examination, which provides the entire image of electrolyte system (Pradhan et al., 2011). The Nyquist plot for the systems was deduced in terms of the EEC. It comprises bulk resistance (R_b) for the carriers in the electrolyte systems and two constant phase elements (CPE) as seen in the insert of Fig. 7. The region of high frequencies displays the connection of R_b and constant phase element in parallel, while the region of low frequencies indicates only constant phase element, meaning that the formed EDLC between electrode and electrolyte. The constant phase elements term is more usually employed in EEC rather than ideal capacitor in the real system.

The impedance of Z_CPE can be written as (Hamsan et al., 2017; Aziz and Abdullah, 2018):

Here, the Z_r and Z_i parts of Z* associated with the EEC (inset of Fig. 7b-d) are expressed as (Aziz and Abdullah, 2018):

Table 2 outlines the fitting parameters of the EEC. The R_b is obtained by the interception between the real axis and the spike. It is obvious that upon the addition of the GLY, the semicircle at the region of high frequencies were absent (see Fig. 7b-d) due to the complete transport of ions toward the electrodes. The R_b is dropped as the GLY amount increases to 40 wt% because of the increase in mobility of charge species, ensuring the rise in the σ_DC of ions. The increase in σ_DC with the GLY amount is consistent with the FESEM results (see Fig. 6). The addition of Cu(II)–complex to the PE significantly promoted the amorphous nature that caused to increase the σ_DC and improve the EDLC performing. Rangasamy et al., (Rangasamy et al., 2019) thought that the increase in the amorphous phase of the PE increases the ions mobility by developing more free volume in the PE system. This leads to an increase in the polymer chains segmental motion, due to the rise in the polymer chains flexibility. Therefore, the σ_DC in the PE can be improved. Liew (Liew, 2017) had developed the CPE based on PAA:LiTFSI with the insertion of BaTiO₃. The author obtained σ_DC of 5 × 10^–4 S cm⁻¹ and used the CPE in the EDLC cells. This indicates that CPE is a good candidate to be employed as a PE in the EDLC device. By considering the R_b value and the CPE dimensions, the σ_DC of the CPE films can be calculated using the relation below (Aziz et al., 2019);

Here q_i stands for the charge of the species (1.6 × 10⁻¹⁹C). There was an increase in σ_DC when n_i and µ_i inside the system of PE increases, as it is apparent from Eq. (13). The Table 3 data as well as EIS spectra in Fig. 7 obviously indicate that having more GLY concentration will give rise to the value of σ_DC. GLY decreases the attraction force between the cations and anions of the salt (Shukur and Kadir, 2014, 2015). Thus, a greater number of ammonium ions (n_i) is offered by NH₄SCN to the polymer.

3.5 EDLC characteristics

3.5.1

3.5.1 Transference number measurement (TNM)

TNM is employed to identify the highest carrier species inside the CPE. When 0.2 V is applied, the current commences to decrease until it gets saturation. The current polarization set against time for the maximum conducting CPE (PGNC-4) is portrayed in Fig. 8. The cause for the enormous of the early current value is in consequence of the part held by the ion and the electron carriers at the early stage. Fig. 8 illustrates that there is a substantial reduction in the current prior it reached an equilibrium state. Once the CPE was polarized, it is in the equilibrium; whereas, the maintained current movement is owing to electron species. The cause for that is the blocking of ions caused by the SS electrodes, which lets only the electron to transport through it (Rani et al., 2018). Eqs. (1) and (2) are employed to evaluate t_ion and t_el values, which obtaining I_i and I_ss values as 279.9 µA and 8.1µA, respectively. The t_ion and t_el values are established to be 0.971 and 0.029, respectively. It is interesting to note that the t_ion is very close to the ideal value of one. Subsequently, it is concluded that ions have a significant role in the transportation mechanism in the PVA:NH₄SCN:Cu(II)–complex:GLY system.

Close

Fig. 8

Polarization current versus time for the maximum conducting (PGNC-4) CPE film.

Export to PPT

3.5.1.1

3.5.1.1 Electrochemical stability study

The electrochemical stability window (ESW) is an essential characteristic to be studied (Pandey et al., 2010). Hence, linear sweep voltammetry (LSV) measurement is led to find the PE decomposition voltage. The device performance is verified when one is alert of the ESW of the PE film prior the test of charge–discharge cycles. To avoid causing any harm to the PE, the decomposition voltage is vital. Fig. 9 depicts the LSV plot of PGNC-4 at 10 mV s⁻¹ with a voltage range of 0 to 2.5 V. There is no obvious change in the current within the working electrode in the voltage varying from 0 to 2.15 V. ESW is observed at 2.15 V as the current starts to raise drastically, revealing CPE decomposition.

Close

Fig. 9

LSV plot for the maximum conducting (PGNC-4) CPE film.

Export to PPT

This outcome is similar to the work by Liew (Liew, 2017) on PAA:LiTFSI:BaTiO₃ CPE with ESW of 2.3 V. They employed the CPE in an EDLC. Kadir & Arof (Kadir and Arof, 2011) reported that the ESW for PVA-chitosan-NH₄NO₃-EC film is 1.70 V, which is lesser than the obtained result in our work. The usual ESW for protonic battery use is approximately 1 V (Pratap et al., 2006). Therefore, the decomposition voltage of PGNC-4 highlights its suitability for applications in protonic devices. Lim et al. (Lim et al., 2014) documented that the decomposition voltage of PVA:LiClO₄:TiO₂ was found to be 2.4 V and they used the CPE in an EDLC.

3.5.1.2

3.5.1.2 Cyclic voltammetry study

Cyclic Voltammetry (CV) examination was employed to assess the performance of EDLC using PGNC-4 CPE. Fig. 10 shows the CV plot of EDLC at various scan rates of 5, 10, 20, 50, and 100 mV s⁻¹. No peak is seen in the CV plot signifying that no redox reaction happens at the potential range of 0 to 0.9 V. This is verified the existence of EDLCs (Liew and Ramesh, 2015). As seen in Fig. 10, the CV plot shape deviated from a leaf shape to a nearby shape of a rectangle when the scan rate decreases. High scan rate causes a departure of the CV plot from rectangular shape that is ascribed to the internal resistance and carbon porosity; hence, generating a current–voltage dependence (Kadir and Arof, 2011). The values of C_CV will be derived from the profile of CV with Eq. (3) and are presented in Table 5 at different scan rates of 5, 10, 20, 50, and 100 mV s⁻¹. The value of C_CV increases as the scan rate drops. Ions use the whole vacant sites through the electrodes where ions have appropriate time for the mechanism of diffusion at short scan rates, causing in grander value of C_CV (Lim et al., 2014). The CV shapes acquired in this work is similar to that achieved by Lim et al. (Lim et al., 2014). In their study, the CV at the scan rates of 10, 30, 50, and 100 mV s⁻¹ were created for EDLCs with the system of PVA:LiClO₄:TiO₂-based CPE.

Close

Fig. 10

Cyclic voltammetry (CV) plot of the synthesized EDLC for the largest conducting (PGNC-4) CPE film.

Export to PPT

Table 5 Capacitance values from CV as opposed to scan rates.

Scan rates (mv/s)	Capacitance (F/g)
100	64.014
50	100.137
20	123.457
10	132.602
5	134.889

3.6 Galvanostatic charge–discharge analyses

Fig. 11a displays the EDLC charge–discharge profile at initial cycles at 0.5 mA cm⁻² current density. The roughly discharge curves with linear slope designates the EDLC capacitive manners (Lim et al., 2014). The charge–discharge plot of the EDLC at 0.5 mA cm⁻² for the selected cycles up to 1000th cycles is exposed in Fig. 11b, c. It could be seen that the discharge slope is still almost linear, signifying the presence of capacitive behavior of the EDLC (Lim et al., 2014). Fig. 11b,c displays the charge–discharge performance of the EDLC at specific cycles. At the beginning of charge–discharge process more ions are available to form double layer thus providing longer discharge time. As the cycle number increase, the discharge time become shorter, this is due to recombination of ion to form ion pairs or triplets. The development of ion pairs reduces the charge double-layer.

Close

Fig. 11

Charge–discharge profiles for the synthesized EDLC at 0.5 mA cm⁻² for (a) initial cycles, (b) selected cycles of 100th, 200th, 400th, 600th, 700th, and 1000th and (c) selected cycles of 100th, 500th, and 1000th.

Export to PPT

The value of C_d is computed by substitution the value of the slope of the discharge curves in Eq. (4). Fig. 12 displays the C_d versus number of cycles. The C_d value calculated for the 1st cycle is 128.6F g⁻¹. This value is analogous to the C_CV obtained from CV examination (see Table 5). Therefore, the C_d value accomplished from the EDLC in this study is responsible. The value of C_d was found to increase and stays constant with the average of 155.32F g⁻¹ besides the 1st cycle till it finishes 1000 cycles. Though, in previous documents significant decrease of C_d has been discerned with raising the number of cycles (Lim et al., 2014; Liew, 2017). The achieved C_d in this study is much higher than those stated in the literature using various PEs (see Table 4).

Close

Fig. 12

C_d of the synthesized EDLC for 1000 cycles.

Export to PPT

The development of C_d in the EDLC is ascribed to the higher σ_DC value of the CPE due to the influence of the Cu(II)-complex on the enhancement of the amorphous structure (Brza et al., 2019), and the addition of GLY plasticizer (Hamsan et al., 2017). As reported by Fan et al. (Fan et al., 2018), the development of the amorphous phase is to be beneficial in local chain segmental motion that can encourage the ions transportation and hence increasing the σ_DC. Thus, ions can transfer freely through the PE. Fast ion migration in the CPE also encourages the adsorption of ions at the interfaces of the electrodes and electrolytes that causes greater C_d value of EDLC (Liew, 2017).

The decrease in the voltage throughout the discharging process is caused by the EDLC internal resistance. The low drop voltage value, V_d, in this study denotes that less energy is dissipated during the charge–discharge process. The value of V_d is found by using Eq. (5) and displayed in Fig. 13. It can be seen that the increase in the value of V_d is still small throughout 1000 cycles, revealing the high performance of the fabricated EDLC. The EDLC internal resistance is attributed as equivalent series resistance (ESR). Therefore, a low ESR is crucial for the EDLC application. As documented by Arof et al. (Arof et al., 2012), the ESR in the EDLC is originated from the resistance of the current collectors, PE, and the space amongst the PE and the current collector. ESR is determined via Eq. (5), and the values are illustrated in Fig. 14. A best contact among PEs and electrodes would be achieved because of low ESR, meaning that the migration of ions to the pores at the electrodes surface would be efficient (Asmara et al., 2011). Kumar & Bhat (Kumar and Bhat, 2009) documented that the increase in V_d among the cycles of charge–discharge, which creates the increase in ESR is caused by the PE degradation in the EDLC. Fortunately, the increase in the ESR of the EDLC in the current work is still small within 1000 cycles. The ESR in the current study is much lower than the ESR values documented for other EDLC devices (Zainuddin et al., 2020; Aziz et al., 2020).

Close

Fig. 13

V_d pattern of the fabricated EDLC for 1000 cycles.

Export to PPT

Close

Fig. 14

ESR pattern of the created EDLC for 1000 cycles.

Export to PPT

The energy density (E_d) and power density (P_d) of the EDLC device are achieved using the following equations (Aziz et al., 2020):

From Fig. 15, it can be perceived that the value of E_d using Eq. (14) for the 1st cycle is 14.467 Wh kg⁻¹.The value of E_d exhibits an increase and continues constant with the average of 17.473 Wh kg⁻¹ within the cycles beyond the 1st cycle up to 1000 cycles. It means that the same energy barrier is subjected by ions species in migration to the surface of the AC electrodes for the processes of charge–discharge within 1000 cycles. Though, in previous reports noticable decrease in E_d value were noted with increasing the cycle number (Liew, 2017; Liew et al., 2015). They described that the decrease in the value of E_d within the cycle number is ascribed to the raise in the ESR that creates the increase of energy loss through the mechanism of charge–discharge cycles (Kadir and Arof, 2011; Wei et al., 2005). The accomplished E_d for the EDLC is sufficiently large compared to that described (3 Wh/kg) for PAA:LiTFSI:BaTiO₃- based CPE (Liew, 2017). The obtained E_d in the current study is of the considerable attention in respect of the E_d value of roughly 1.56 Wh/kg that found for EDLC of the CPE system of PVA:LiClO4:TiO₂ (Lim et al., 2014). Fortunately, the value of E_d (17.473 Wh kg⁻¹) within 1000 cycles in the current work is in the range of battery energy density and significantly higher than the E_d values reported in the literature (see Table 4). These results indicate that the Cu(II)-complex as filler has a great impact on the performance of the EDLC device.

Close

Fig. 15

E_d of the synthesized EDLC for 1000 cycles.

Export to PPT

The achieved P_d value using Eq. (15) for our study will be seen in Fig. 16. Over charge–discharge within 1000 cycles, the P_d value for the 1st cycle is 4960 W kg⁻¹ and subjected to a drop up to 500th cycle with the average of 2277 W kg⁻¹ and then continues constant with the average of 400 W kg⁻¹ throughout the cycles beyond the 500th cycle. The drop of P_d at higher cycles is because of the depletion of the PE. Ion agglomeration after the fast charge–discharge mechanisms blocks the migration of ions to the electrodes that causes a decrease of adsorption of ions at the interfaces of the electrodes and PEs (Liew et al., 2016). This tendency of P_d drop is in agreement with the rising ESR tendency. The ESR raise at greater cycles due to the reduction in the mobile ion concentration and aggregation of ions after the swift processes of charge–discharge cycles offers smaller P_d at higher cycles (Zhong et al., 2015). The accomplished P_d for the EDLC cell is high in comparison to that reported in the literature (see Table 4).

Close

Fig. 16

P_d of the developed EDLC for 1000 cycles.

Export to PPT