Chromophore doped DNA based solid polymer electrolyte for electrochromic devices

2.1 Materials

Highly purified deoxyribonucleic acid (DNA), extracted from the waste produced by salmon processing industry was purchased at Ogata Research Laboratory, Ltd., Chitose, Japan. The average molecular weight (M_w) of the DNA used was 2 × 10⁷ g mol⁻¹. The chromophore used, Prussian Blue was purchased from Fluka Analytical (M_w = 306.89 – anhydrous).

For electrochromic smart windows as electrodes WO₃/ITO/glass (Solarska et al., 2006) and CeO₂–TiO₂/ITO/glass were used (Mîndroiu et al., 2014).

2.2 Experimental techniques

The infra red (IR) measurements were performed in the attenuated total reflection mode (ATR) using a Perkin Elmer Spectrum 100 FT-IR spectrophotometer. The studied samples were placed on a Diamond crystal. The investigated spectral range was from 4000 to 600 cm⁻¹. The signal was obtained by averaging 4 scans at resolution of 4 cm⁻¹. As it was sufficiently strong we did not need to make more scans. It is a routinely used procedure in that case.

The UV–VIS spectroscopic analysis was performed in the 200–800 nm spectral range, with a step of 0.5 nm, using a JASCO UV–VIS–NIR spectrophotometer, model V 670.

The surface wettability of the obtained membranes was analyzed using a Contact Angle Meter – KSV Instruments, model CAM 100.

The ionic conductivity measurements of obtained biomembranes were carried out by the electrochemical impedance spectroscopy (EIS), using an Autolab PGSTAT 302N potentiostat with general – purpose electrochemical system software. The measurements frequency range used was between 10⁵ Hz and 10 mHz and the applied voltage amplitude of 10 mV.

The cyclic voltammetry measurements, performed on the ECDs based on the new biomembranes, were done using an Autolab PGSTAT 302 N device, applying tension ranging from −2 V to 2 V with the electric field scan step of 100 mV s⁻¹.

The optical transmittance measurements of ECDs, containing the DNA based membranes with the highest ionic conductivity value, were performed with a Jasco V-670 Spectrophotometer between 380 nm and 800 nm. The ECW performance was measured by the evaluation of optical and electrochemical properties.

One the most important factor in evaluating the electrochromic material is the electrochromic contrast, defined as the percentage of the transmittance change (Δ%T) at the specified wavelength (Eq. (1)), where the electrochromic material has the highest optical contrast. The optical properties are evaluated through the whole transmittance range.

2.3 Preparation of solid polymer electrolytes (SPEs) and of electrochromic devices (ECDs)

The SPEs were prepared at room temperature in the form of thin membranes using a very simple casting technique. First, the glycerol water solution 50 g/L was mixed with Prussian Blue for 12 h. Then a DNA water solution, at the concentration of 10 g/L was added to the mixture of glycerol with chromophore and stirred for another 4 h until a homogenous gel was obtained. The resultant gel was cast in a Teflon box, placed in an oven where it was kept at the temperature of 50 °C over a period of 4 days.

In order to see the influence of Prussian Blue chromophore on the membrane characteristics several of them were prepared.

The gravimetric ratio between DNA and GLY was kept constant at 1:1 ratio, and the PB content was gradually increased from 0.5 (sample A – DNA1GLY1PB0.5) to 1%, 2% and 3% (samples: B – DNA1GLY1PB1, C – DNA1GLY1PB2, D – DNA1GLY1PB3). For a better analysis of the membranes more samples were fabricated in the same conditions: DNA membrane without both GLY and PB (sample E – DNA10g/L), DNA membrane with GLY and without PB (sample F – DNA1GLY1) and DNA membrane with PB and without GLY (sample G – DNA10g/LPB2%). The measured thickness of the membranes was between 80 and 130 μm.

ECDs were assembled in a sandwich like structure. The new SPE membrane was placed between two electrodes: glass/ITO/WO₃ (Georg et al., 2008; Granqvist, 2000; Deb, 2008; Thakur et al., 2012; Campet et al., 1992) and CeO₂-TiO₂/ITO/glass. For a good contact between the layers, the new device was kept overnight in a spring-loaded support and then, for electrical contact, the edges of the device were sealed with copper-conducting tape.

3.1 FT-IR analysis

The identification of functional groups in raw materials and also in all obtained membranes was performed by ATR/FT-IR analysis. The main absorption bands from IR spectra of the raw DNA and GLY are presented in Table 1.

Prussian Blue is particularly attractive for infrared studies since the system contains the strongly absorbing CN functional group (Mîndroiu et al., 2014). PB is a ferric ferrocyanide (Fe^III₄ [Fe^II(CN)₆]₃) with iron(III) atom coordinated to nitrogen and iron (II) atom coordinated to carbon, as it was established by the spectroscopic investigation (Kulesza et al., 1996). In the 4000–600 cm⁻¹ region, the IR spectrum of the raw Prussian Blue (Fig. 2a) (Fe₄[Fe(CN)₆]₃) shows a strong absorption band at 2070 cm⁻¹ attributable to the CN functional group (Karyakin, 2001; Ion et al., 2008). Fe—C—N—Fe bending, characteristic of PB can be seen in far infrared region (450–650 cm⁻¹).

Close

Figure 2

FT-IR spectra of raw Prussian Blue (a) and DNA based SPEs (b) with and without GLY or PB.

Export to PPT

The IR spectra of DNA membranes, with and without GLY or PB (Fig. 2b) show the most representative bands as reported in Table 1.

The peak corresponding to stretching vibration of O—H (ν_O—H) from GLY and DNA and to N—H (ν_N—H) from DNA appeared in membranes at 3189–3284 cm⁻¹ and is attributed also to intermolecular hydrogen bonds formation.

The presence of Prussian Blue in studied membranes may be checked by the IR bands at 2072 cm⁻¹ in DNA and at 2078–2080 cm⁻¹ DNA–GLY matrices, respectively. This band corresponds to the C—N stretching vibration in ferrocyanide ion Fe (CN)₆⁴⁻ (Karyakin, 2001; Guo and Lin, 1997). It is an intense one and can be specifically used to identify PB because it occurs in a region where no other organic or inorganic groups absorb (Kendix et al., 2004).

The bands at 1687 cm⁻¹ and 1642 cm⁻¹ attributed to the stretching vibration of C⚌O (ν_C⚌O Amide I) and to NH₂ scissoring vibrations from DNA membrane are observed in all DNA membranes with or without GLY and PB. But these bands appear at other wavenumbers like 1687–1708 cm⁻¹ and between 1643 cm⁻¹ and 1653 cm⁻¹.

The absorption band corresponding to the deformation vibration of N—H from DNA (δ_NH Amide II) appears in all DNA membranes with or without GLY and NB at 1530 cm⁻¹, with almost the same intensity.

The presence of DNA in the membranes composition is evidenced also by the band attributed to Amide III and to ${\mathrm{\nu }}_{\text{PO}{2}^{-}\text{as}}$ which appear at 1223 cm⁻¹ in DNA membrane and at 1226 cm⁻¹ in membranes with DNA and PB, and at 1213–1216 cm⁻¹ in membranes with DNA and GLY with and without PB content. Also the symmetric PO₂⁻ stretching vibration ( ${\mathrm{\nu }}_{\text{PO}{2}^{-}\text{s}}$ ) observed at 1050 cm⁻¹ in DNA membrane, appears at 1041–1054 cm⁻¹ in membranes with DNA, GLY and PB dopant.

The O—H bending from GLY appears at around 927 cm⁻¹ for all membranes containing GLY.

3.2 UV–VIS spectroscopy

The new SPEs membranes were characterized for their optical absorption in the wavelength range 200–800 nm. The ultraviolet–visible spectra of some SPEs with the same gravimetric ratios between DNA and GLY and different concentrations of PB are presented in Fig. 3.

Close

Figure 3

Optical absorption spectra of samples A (DNA1GLY1PB0.5) and B (DNA1GLY1PB1) membranes.

Export to PPT

The spectra show an intense absorption band, located around 720 nm, which is due to the charge transfer during the process: Fe²⁺—CN—Fe³⁺ transformation of PB (Newman, 1979).

3.3 Ionic conductivity

For device application as ECD_s the DNA based solid electrolyte has to exhibit good ionic conductivities. For this purpose the membranes were doped with PB chromophore at different levels. The results of ionic conductivity measurements of these DNA based membranes with as function of PB content, performed at the temperature range of 20–100 °C are displayed in Fig. 4.

Close

Figure 4

Logarithm of ionic conductivity as function of the inverse of temperature (in K) for the studied DNA based membranes with different Prussian Blue (PB) content.

Export to PPT

The previous works (Mindroiu et al., 2014; Pawlicka et al., 2010) show that addition of GLY, used as plasticizer, to the membrane composition, improves their ionic conductivities. In this work we studied the influence of the amount of PB dopant in glycerol plasticized DNA membrane on its ionic conductivity.

All DNA based membranes with PB dopant exhibit ionic conductivities ranging from 10⁻⁹ to 10⁻⁷ S cm⁻¹ at room temperature. At higher temperature, the ionic conductivity of all membranes increases, reaching a value of about 10⁻⁴ S cm⁻¹ for sample C at the temperature close to 80 °C, as compared to the room temperature data. This behavior can be explained by the DNA denaturation process, which produces free guanine and adenine molecules. Also, the number of mobile charge carriers and the delocalization of π electrons in nucleobases increase (Apilux et al., 2007). The highest values of ionic conductivity were obtained for DNA based membrane with 2 w% of PB (DNA1GLY1PB2). In this case, the number of mobile charge carriers is optimal. Adding a higher amount of dopant leads to the chromophore aggregation, which leads to a decrease of ionic conductivity.

3.4 Contact angle measurements

In order to evaluate the influence of the environmental humidity variation on ionic conductivity behavior of SPEs, the wettability was registered.

The surface characterization by static contact angle measurements was recorded at the room temperature with an accuracy of ±1°, using a constant volume of distilled water drop deposited on the biomembrane surface and a number of 5 frames. This method is very sensitive to the surface changes although it does not give information about the type of groups present.

The controlling surface wettability is done in order to evaluate whether the change in chromophor concentrations used in SPE composition may lead to changes in contact angle values. The obtained results from the contact angle measurements are displayed in Table 2.

It can be seen that the largest contact angle is observed for DNA based SPE without PB (sample F - DNA1GLY1), which means that this material has the most pronounced hydrophobic character. In the case of the DNA – chromophore based SPE and no glycerol (sample G - DNA10g/LPB2%) a pronounced decrease of hydrophobicity is observed. Adding glycerol and various concentrations of chromophore dopant in the SPE composition leads to decrease of the hydrophobic character as follows: lowest value of contact angle was observed for SPE with 3% PB (sample D - DNA1GLY1PB3) and its increase for 0.5 w% PB (sample A - DNA1GLY1PB0.5). The differences in contact angle values are not very large. All studied membranes exhibit a hydrophobic behavior, what means that the dopant content had less influence on the compounds wettability.

3.5 ECD characterization

Taking into account the results obtained from the ion conductivity measurements, we have selected sample C (DNA1GLY1PB2) as solid electrolyte to assemble the electrochromic device. Its composition is the following: DNA 10 g/L with GLY 50 g/L and PB 2%. To obtain ECD the membrane was sandwiched between two ITO/glass electrodes, on which thin layers of WO₃ and CeO₂–TiO₂ (Munro et al., 1998) were previously deposited by the dip-coating method.

The chronoamperometry and the cycle voltammetry methods were used to characterize the assembled ECD. The results were correlated with the optical transmittance measurements.

The UV–VIS spectra of ECD with the structure: glass/ITO/WO₃/DNA1GLY1PB2/CeO₂-TiO₂/ITO/glass are shown in Fig. 5 for 1000 charge insertion/extraction cycles.

Close

Figure 5

Optical transmittance for ECD with glass/ITO/WO₃/DNA1GLY1PB2/CeO₂–TiO₂/ITO/glass structure after first, 500 and 1000 cycles of insertion and extraction of charges.

Export to PPT

All transmittance spectra displayed in Fig. 3 show only bands corresponding to WO₃ absorption. In inset to Fig. 5 only the UV–VIS spectra in the 350–600 nm range are presented, in order to evaluate the optical properties of the studied device. It was observed that for negative applied potentials (−2.8 V), transmittance is minimum due to the reduction of the WO₃, according to the following reaction WO₃ (bleached) + x e⁻ + xM⁺ ↔ M_xWO₃ (intense blue color). At the positive potential the value of transmittance increases as a result of WO₃ oxidization.

The electrochromic contrast (Δ%T) of the device is low, thus, for first insertion/extraction cycle the value of Δ%T is 0.3. It decreases with the number of cycles to 0.25 after 500 cycles and to 0.2 after 1000 cycles, respectively.

At the same time, the transmittance decreases from 2.1% at 1.5 V/15 s after first to 1.5% at 1.5 V/15 s after 1000 charge insertion/extraction cycles.

Fig. 6a shows that the device charge density response depends on the time at which the potential values are applied. Thus, the inserted charge increases from −2.4 mC/cm² at 15 s/15 s to −4.3 mC/cm² at 60 s/60 s. In all studied cases the charge extraction at 1.5 V is faster than insertion at −2.8 V.

Close

Figure 6

Charge density results (a) for different applied time at potentials −2.8 V and +1.5 V; and (b) as a function of the chronoamperometric cycles of insertion and extraction of charges for ECD with WO₃/DNA1GLY1PB2/CeO₂–TiO₂ configuration (inset shows charge density after 100, 500 and 1000 cycles). The applied potentials were −2.8 V/15 s and 1.5 V/15 s.

Export to PPT

In order to evaluate the reversibility of the studied ECD, the samples were subjected to chronoamperometric cycling coupled to a spectral analysis by applying the potentials of −2.8 V and +1.5 V for 15 s/15 s. Fig. 6b shows the results of the charge density modification during 1000 cycles with color change from intense blue color to colorless as indicated in Fig. 5.

The mechanism of electrons and ions insertion/extraction during the cathodic and anodic polarization process of the ECDs based DNA solid electrolyte was explained in some previous works (Neto et al., 2014; Mindroiu et al., 2014; Pawlicka et al., 2011).

Confirming the spectral analysis (Fig. 5), the negative voltage application onto electrochromic device with WO₃/DNA1GLY1PB2/CeO₂–TiO₂ configuration promotes a blue color appearance associated with WO₃ reduction by simultaneous electrons and ions insertion. The glass/ITO/WO₃ electrode became less colored after applied voltage to +1.5 V because of WO₃ oxidation, but at the device level, the color change intensity is barely perceptible, due to the fact that the membrane does not change its color (Fig. 7), remaining blue from PB. Therefore, the color change takes place at the WO₃ layer only and not at the electrolyte.

Close

Figure 7

Photograph of ECD with WO₃/DNA1GLY1PB2/CeO₂–TiO₂ configuration at (a) −2.8 V/15 s (cathodic polarization) and (b) at 1.5 V/15 s (anodic polarization).

Export to PPT

This affirmation is sustained by the transmittance spectra (Fig. 5), which show only their change at around 475 nm corresponding to WO₃, while the transmittance peak corresponding to the electrolyte, at around 633 nm is not present (Assis et al., 2013).

The variation between the intense blue colored to colorless of WO₃ remains almost constant with a small increase of charge density during the first 2 h of test (Fig. 6b), from −3 mC cm⁻² to −3.6 mC cm⁻². After 500 cycles the charge density decreases to −3.1 mC cm⁻². It decreases also after 1000th charge insertion/extraction cycles to −2.6 mC cm⁻², as it can be seen in the inset to Fig. 6b.

In order to evaluate the reversibility of the studied electrochromic device we have presented (Fig. 8a) their cyclic voltammograms in function of the scan rate.

Close

Figure 8

(a) Cyclic voltammograms between −2.8 V and 1.5 V at different scan rates and (b) anodic and cathodic peak current densities against the square root of CV scan rate for ECD with WO₃/DNA1GLY1PB2/CeO₂–TiO₂ configuration.

Export to PPT

All voltammograms put in evidence the presence of a cathodic peak with a maximum centered at −1.64 V (for 500 mV/s), which is associated with blue color and is due to the reduction of WO₃ layer. The presence of an anodic peak, centered at −0.54 V (for 500 mV/s), indicates the bleaching reaction, which is reversible. These results confirm that the occurence of the device color change is due exclusively to the WO₃ electrode. This is caused by the fact that in this studied potential range the cyclic voltammograms do not present another peak which can be attributed to the electrolyte.

The studied ECD system with DNA based membrane is a reversible one because of its dependence on the scan rate (Pawlicka et al., 2011). The oxidation peak current values increase with the scan rate (Fig. 8a) and shift to higher potentials. Plots of the cathodic and anodic peak currents against the square root of the scan rate are displayed in Fig. 8b. They show that the peak current is proportional to the square root of the scan rate, indicating a diffusion control of the mass transfer at the electrode surface (Pawlicka et al., 2011).