In this study, 4,4′-bipyridine was provided by Alfa Aesar Company. 3-Bromo propylamine hydrobromide 98% was supplied by Sigma-Aldrich Company, and 2,4,6-trichloro-s-triazine (cyanuric chloride) was purchased from Merck Company, Germany. The preparation of ethylenediamine core polyamidoamine dendrimer of (PAMAM G2) is described within reference [45]. DuPont Company, Germany, provided polymethyl methacrylate. Propylene carbonate and a binary mixture of ethylene poly(oxide) and propylene poly(oxide) (PEO-PPO) were purchased from Fluka Company, Germany. The polyethylene glycol 200, isopropyl alcohol, lithium perchlorate, tetrahydrofuran (THF), and all the other chemicals used during this work were chosen as they are laboratory grade, being received from Merck Company, UK. The Sigma-Aldrich Company, USA, provided the indium tin oxide (ITO) coated polyethylene terephthalate (PET) with a surface resistivity of 60 Wsq⁻¹ and thickness of 127 mm.

The prepared EC ink was ink-jet printed using an Epson Stylus Photo P50 printer. The pH, surface tension, and viscosity of the prepared water-based inks were characterized using 827 pH Metrohm meters (Herisau/Switzerland), Tensiometer K100MK2 (Hamburg, Germany), and Brookfield DVII (New Jersey, USA), respectively. The prepared ink was filtered through 0.45 and the 0.2μm Sartorius Minisart filter (Göttingen, Germany). Static contact angle (θ) and surface energy of the substrate were measured by a Kruss G40-type contact angle measuring system using distilled water as the probe liquid at temperature and humidity of 25°C and 30%, respectively. For this purpose, a small drop of distilled water (2–3 mL) was exposed to the surface of the substrates. The test was done in triplicate, and the average of the measurements was reported. An image analysis system (G2/G40) was used to calculate the contact angle values. Thin-layer chromatography (TLC) was performed using aluminum plates coated with silica gel 60 F254 (Merck) as a stationary phase and a mixture of isobutanol: n-propanol: ethyl acetate: water in the ratio 2:4:1:3 as the mobile phase. The developed plates were visualized under both short and long-wavelength ultraviolet light (254 nm, 365 nm). The Fourier transformed infrared spectroscopy (FTIR) spectrums of synthesized electrochromic dyes were recorded on Bruker (IFS-48) (Bruker Optik GmbH, Germany), in the range of 400 to 4000 cm⁻¹. The synthesized electrochromic dyes were characterized by proton nuclear magnetic resonance (¹H NMR) spectra, which were recorded on a Bruker-SY200-250 MHz spectrometer analyzer. The gel permeation chromatography (GPC) was performed on a Shimadzu 6-A with the column model Waters Ultrahydrogel 250, which was fitted by water as the mobile phase (flow rate 1 mL·min⁻¹) and PEG as standard material to determine the molecular weight of previously synthesized PAMAM dendrimer. Elemental analyses for carbon, hydrogen, nitrogen, and sulfur were carried out at the Central Laboratory, Tehran University, on an Eager 300 for EA1112. In the cyclic voltammetry, a three-electrode method was adapted with platinum wire, glassy carbon (3 mm diameter), and Ag/AgCl as the counter, working, and reference electrodes, respectively. A cyclic voltammetry experiment was performed by cycling the potential of a working electrode and measuring the resulting current. In the preparation process for the CV experiment, a glassy carbon as a working electrode was polished with a 1 μm and 0.1 μm γ-Al₂O₃ polishing cloth for 5 minutes. Polishing was followed with 3 minutes of sonication in water to ethanol mixture (1:1). Background voltammograms were obtained in 0.1 M HCl saturated electrolyte by cycling the potential between 0 to −1.4 V. Absorbance measurements at l_max using UV Ikon 923 Double Beam UV/Visible spectrometer (Saint-Quentin-Yvelines, France) determined the electrochromic characterizations of novel ECDs that contain newly synthesized viologen-functionalized dendrimer.

Epson Stylus Photo P50 ink-jet printer was used to print a solid square pattern by the water-based ink formulation containing 100 g/dm³ synthesized viologen-functionalized dendrimers (K₁ or K_1/2), 50 g/dm³ polyethylene glycol 200, and 300 g/dm³ isopropyl alcohol. The inks were made up to 1 dm³ with de-ionized water. Epson Stylus Photo P50 ink-jet printer operated using a piezoelectric squeeze mode print head that squirts ink droplets through a nozzle with a 65-μm opening. Printing was carried out at pH=6 using a buffer solution to prevent damage to the print-head and cartridge. The prepared inks were filtered by 0.45 μm and 0.2 μm filters to prevent the nozzles clogging. The ink-jet printing substrate was an ITO coated PET. The ITO coated PET was cleaned using an ultrasonic bath in de-ionized water, isopropyl alcohol, acetone, and methanol, respectively. The methanol is an effective solvent and provides a higher surface energy substrate. Finally, the cleaned ITO coated PET dried at 60°C for 2 hours before loading inside the printing system.

During the preparation of ink-jet printing of flexible ITO coated PET substrates, a new and cleaned cartridge was filled with filtered formulated ink. Subsequently, the filled cartridge was located inside the printer. The printed ITO coated PET dried at 130°C for 30 minutes.

Within electrolyte preparation, polymethyl methacrylate (PMMA) (0.36 g), lithium perchlorate LiClO₄ (0.33 g), a binary mixture of ethylene poly(oxide) and propylene poly(oxide) (PEO-PPO) (0.21 g), and propylene carbonate (PC) (1.75 g) were mixed with 4 mL of THF and followed by stirring for about 4 hours to obtain transparent uniform lightly viscous gel [46]. The prepared gel electrolyte was used as an ion conductor layer between the counter and working electrodes in the EC device.

The cyclic voltammetry (CV) measurements were performed to examine the existence of the reversibility and redox behavior of newly synthesized viologen derivatives.

The CVs of viologen-functionalized dendrimers (K₁ and K_1/2) illustrate in Fig. 6a–Fig. 6b. The electrochemical parameters are given in Table 1. As it can be seen, both CVs exhibited oxidation-reduction waves. A sharp peak appeared with a peak potential wave of E= −0.67 V for K₁, corresponding to the first reduction of dicationic viologen (V²⁺) to mono-cation radical (V^+•) as deep-blue color. Doubly reduced viologen (V^2•) at −1.0 V was formed according to the color changes from deep blue to a colorless state. The first oxidation occurred at −0.91 V with a change of colorless to blue during the formation of mono-cation radical (V^+•).

As expected, K_1/2 showed the same CV with peak potential wave of E= −0.65 V, corresponding to the first reduction of di-cationic viologen (V²⁺) to monocation radical (V^+•) as deep-blue color. Doubly reduced viologen (V^2•) at −0.99 V formed according to color changing from deep blue to colorless state as the same as K₁. The first oxidation occurred at −0.91 V with a change of colorless to blue during the formation of mono-cation radical (V^+•). Detrimental changes did not observe for the initial and second reduction and oxidation potentials. These results show that the origins of redox potential are related to the same molecule of viologen on dendrimer.

As expected, a transition from a deep blue to a colorless state occurred during the second oxidation related to the oxidation of mono-cation radical to dication (V²⁺) form.

The values of peak potential separation (ΔEp) for K₁ and K_1/2 were 90 mV and 80 mV, respectively. Suggesting that the electron transfers in K₁ and K_1/2 were quasi-reversible [47–49].

The voltammograms also were recorded by changing the scan rates from 0.05 Vs⁻¹ to 0.5 Vs⁻¹ (Fig. 6c). The peak current ratios (I_pc/I_pa) ranging from 0.05 Vs⁻¹ to 0.5 Vs⁻¹ were found to be very near to unity, which implies that the system was quasi-reversible for both K₁ and K_1/2. Increasing peak current at the same condition for K₁ concerning K_1/2 has approved the addition of more viologen units on the dendrimer, which is responsible for the production of more reduction and oxidation currents.

As can be seen in Fig. 6c, the peak currents for both redox processes (I_pa, I_pc) were found to increase linearly with the square root of the scan rate, suggesting that a diffusion control reaction was involved [50].

The wettability of the ink on a substrate depends on the substrate’s surface energy and the ink’s surface tension. Therefore, before starting to develop the inkjet inks, the surface energy of the substrate should be considered to prepare an ink with lower surface tension than the substrate’s surface energy[50].

The surface energy values of the ITO coated PET calculated according to equations (1) and equation (2), which are known as the equations of state as proposed by Li and Neumann [51].

where γ_lv and γ_sv are surface tension of water and surface energy of the substrate, respectively. θ is a water contact angle and b is 0.0001247 ±0.000010 (mJ/m²)⁻². The work of adhesion (W_A) was first calculated according to equation (1) (Young’s equation). Surface energy (γ_sv) was then calculated using equation (2).

Based on Fig. 7, the water contact angle on the ITO coated PET is 79.5°. According to the above equations, the surface energy values of ITO coated PET was 35.8 mNm⁻¹. Correspondingly, the surface tension of the prepared inks plays significant roles in the wettability of the ink onto the substrate and in the decomposition of the ink into precise drops through the printer’s nozzle. The surface tension values of the formulated inks were 29.9 mNm⁻¹ and 30.1 mNm⁻¹, respectively. The results were within the values of commercial ink-jet ink (28–31 mNm⁻¹) and lower than the obtained surface energy of the substrate (35.8 mNm⁻¹).

The viscosity of the prepared ink mainly affects the rheological performances within the capillary nozzles of the printer. Fig. 8 shows the variation of viscosities along with shear rates for the prepared water-soluble inks on a logarithmic scale. As can be seen, the viscosity of the inks remained almost unchanged by increasing the shear rate, demonstrating that the inks behave as a Newtonian fluid [52]. Consequently, the viscosity value for the ink-jet ink was 1.8 mPa.s, which is within the acceptable range for ink-jet printing ink.

The flexible EC devices are assembled in a multilayer structure due to investigate the electrochromic properties of synthesized EC material as an active printed layer in an EC device. First, the homogeneous and thin layers of EC materials based on novel water-soluble viologen-functionalized dendrimers (K_1/2 and K₁) ink-jet printed on the ITO coated PET substrate. Then, the electrolyte that was previously prepared was spilled onto the printed EC film to facilitate the movement of ions between each electrode. Subsequently, an extra ITO coated PET substrate was used as a counter electrode placed on top of it and combined in a sandwich structure (Fig. 1). The connection was reinforced using a clip with dimensions of 2.5 cm × 3.5 cm. The electrolyte inside the device needed not to contain air bubbles, because their presence may interfere in the measurements.

Then, the EC device color changes were examined by UV-Vis spectroscopy (Fig. 9a and Fig. 9b) at various applied potentials. In the initial state (0 V), when the viologen di-cation formed, the prepared EC devices were colorless. By applying the potential, the electrochemical reduction obtained in the formation of the radical cation and EC device color changes from the transparent to blue occurred (Fig. 9c) with the absorption of broadband centered at λ_max= 550 nm and 600 nm for K_1/2 and K₁, respectively. By increasing the applied potential from −1 V to −1.5 V and −2 V, color absorption intensity in λ_max increased. The absorption at λ_max= 600 nm for K₁ indicates the formation of viologen cation radical monomers, while its associated species or dimeric forms are at λ_max= 550 nm [25]. A great way to assess the insertion/extraction film is to compare spectral coloration efficiencies, CE (λ), for the visible region.

Coloration efficiency is defined as the amount of electronic charge (Q), which is necessary to produce an optical density change (ΔOD), equation (3).

ΔOD (λ_max) is the change in transmitted optical density at the maximum wavelength (single-pass) because of charge density Q (cm² C⁻¹) inserted to the T_b (bleached transmittance value) and T_c (colored transmittance value). The best prepared EC device’s coloration efficiency (η) or electrochromic efficiency (EE) was 311.5 cm²C⁻¹ at 600 nm in −2 V potential. The redox process was reversible, and the coloration-decoloration switching time was 5 seconds. Color contrast for the prepared EC device was the transmittance change from 79.2% (0 V) to 32.2% (−2 V) at λ_max= 600 nm (ΔT= 47%), Table 2.

The obtained results were comparable to previous reports [8,20,53–59]. However, the switching time value was longer than that obtained in the ECD based on type II electrochromic materials. This difference was due to the diffusional limitation of the solution based viologen-functionalized dendrimer system.