Ionic liquid-assisted synthesis of chitin–ethylene glycol hydrogels as electrolyte membranes for sustainable electrochemical capacitors

A novel chitin–ethylene glycol hybrid gel was prepared as a hydrogel electrolyte for electrical double-layer capacitors (EDLCs) using 1-butyl-3-methylimidazolium acetate [Bmim][Ac] as a chitin solvent. Examination of the morphology and topography of the chitin–EG membrane showed a homogeneous and smooth surface, while the thickness of the membrane obtained was 27 µm. The electrochemical performance of the chitin–EG hydrogel electrolyte was investigated by cyclic voltammetry and galvanostatic charge/discharge measurements. The specific capacitance value of the EDLC with chitin–EG hydrogel electrolyte was found to be 109 F g−1 in a potential range from 0 to 0.8 V. The tested hydrogel material was electrochemically stable and did not decompose even after 10,000 GCD cycles. Additionally, the EDLC test cell with chitin–EG hydrogel as electrolyte exhibited superior capacitance retention after 10,000 charge/discharge cycles compared with a commercial glass fiber membrane.


Synthesis of 1-butyl-3-methylimidazolium acetate [Bmim]
[Ac]. The first step involved the synthesis of 1-butyl-3-methylimidazolium chloride. In a 250 ml round-bottomed flask equipped with a Teflon-coated magnetic stirring bar, 1-methylimidazole (0.10 mol) was reacted with 1-chlorobutane (0.11 mol) added at once to 100 ml of acetonitrile, which was utilized as a reaction medium. The reactants were stirred for 48 h at 60 °C, and the solvent was then evaporated using a vacuum evaporator. The product was washed twice with 100 ml of ethyl acetate to remove excess of quaternizing agent. The remaining ethyl acetate was removed by drying the content of the flask under reduced pressure at 50 °C for 24 h. Subsequently, 0.10 mol of 1-butyl-3-methylimidazolium chloride was dissolved with 200 ml of methanol in a 500 ml reaction glass equipped with a mechanical stirrer. Then, 100 ml of the anionic resin Dowex Monosphere 550A (OH) was added, and the obtained mixture was stirred for 1 h at 25 °C. The resin was then filtered off and rinsed three times with small amounts of methanol. The solution of the semi-product 1-butyl-3-methylimidazolium hydroxide was slowly neutralized using stoichiometric amounts of acetic acid at 25 °C in an EasyMax 102 (Mettler Toledo, Switzerland) semi-automated reactor system equipped with a glass electrode. The solvent was evaporated under vacuum in a rotary evaporator, and the product was dried under reduced pressure (5 mbar) at 50 °C for 24 h. The structure of the obtained compound was confirmed by analysis of the collected FT-IR, 1 H, and 13 C NMR spectra (Figures S1-S3, ESI). The water content in the ionic liquid, assessed with the use of an SI Analytics Titroline 7500 KF Trace coulometer, was determined to be 6015 ppm. H NMR spectra were recorded on a Varian VNMR-S 400 MHz spectrometer with TMS as internal standard. 13 C NMR spectra were obtained with the same instrument at 100 MHz. The IR spectra were collected using the EasyMax 102 semi-automated system (Mettler Toledo, Switzerland) connected to a ReactIR iC15 probe (Mettler Toledo) equipped with an MCT detector and a 9.5 mm AgX probe with a diamond tip. Data were sampled from 3000 to 650 cm −1 with 8 cm −1 resolution and processed with the use of iCIR 4.3 software.

Fabrication of chitin-ethylene glycol thin films.
Chitin dissolution was carried out using the EasyMax 102 system (Mettler Toledo, Switzerland), which allows the high-precision control of a broad range of reaction parameters such as temperature, pH, stirring speed, and reagent addition, according to a previously reported protocol 42 . Briefly, 0.1 g of α-chitin and 10 ml of 1-butyl-3-methylimidazolium acetate [Bmim][Ac] along with a magnetic stir bar were placed in a 20 ml vial, heated to 95 °C and stirred at 1000 rpm for 24 h until the chitin was fully dissolved and the solution became homogeneous and turned an amber-like color. The solution was additionally centrifuged to remove any undissolved residuals. To prepare the thin film, 2 ml of hot solution was poured over a Teflon-lined Petri dish. Once the solution had settled and formed a uniform layer (30 min), the Petri dish was placed in an ethylene glycol (EG) bath for 48 h to allow film coagulation. During this time, the ethylene glycol was changed four times, to ensure the complete removal of the ionic liquid. Subsequently, the gel formed was immersed in a distilled water bath for 24 h to remove excess ethylene glycol, and the distilled water was changed four times. The resulting chitin-EG films were then air-dried for 72 h, and a detailed analysis was made of their physicochemical and electrochemical parameters.
Scanning electron microscopy. A MIRA3 scanning electron microscope (SEM) (Tescan, Czech Republic) was used to assess the surface. This was done using an acceleration voltage of 2 kV and a working distance of 2 mm. A thin Au-Pd layer with a thickness of approximately 10 nm was deposited on the sample using a Quorum vacuum evaporator.
Atomic force microscopy. The nanomechanical properties of the chitin-EG membrane were determined with the use of atomic force microscopy (AFM). This technique enables the creation of three-dimensional images of the surface of a sample. An NX10 microscope (Park Systems) was used in the experiment.
To find a suitable point on the surface, initial images were made in non-contact mode (NCM) with a 30 × 30 μm scan size. Later, the scan size was narrowed to 10 × 10 μm. Mechanical properties of the surface were examined in PINPOINT™ mode, which additionally allows the simultaneous collection of data on six mechanical properties: adhesion energy, adhesion force, deformation, energy dissipation, Young's modulus, and stiffness.
Samples were made by attaching the chitin-EG membrane to a steel puck with the use of double-sided adhesive tape. The cantilever used in the measurements was an ALL-IN-ONE D (BudgetSensors), which has a nominal force constant of 40 N/m. The surface area examined was 10 × 10 μm, scanned with 512 lines per image and with the NCM mode scan speed set at approximately 0.3-0.4 Hz. All measurements were performed at room temperature (approximately 22 °C). The resulting data were investigated using the open-source Gwyddion software and XEI (Park Systems) AFM analysis programs.
Contact angle measurements. The wettability and surface free energy of the prepared chitin-EG thin films were determined by static contact angle measurements, with distilled water and diiodomethane as the polar and dispersion components, respectively. Measurements were made using a sessile drop technique at 25 °C with a DSA100 goniometer (KRÜSS, Germany). The surface free energy was calculated using the Owens, Wendt, Rabel & Kaelble (OWRK) model. Swelling factor. The swelling factor of the chitin-EG membranes was evaluated by determining their sorption of 2 M aqueous solution of lithium acetate. The dry chitin-EG was cut into 1 cm × 1 cm fragments, which were weighed to determine their dry weight. These samples were placed in 50 ml beakers containing 40 ml of 2 M lithium acetate solution, and stored at 25 °C. After 24 h the samples were taken out and weighed after carefully removing their surface water with a filter paper. The swelling was calculated using Eq. (1) 43 : where m w is the mass of the wet sample and m d is the mass of the dry sample.
Ionic conductivity measurements . The ionic conductivity of the chitin-EG membrane in the form of a hydrogel was measured by electrochemical impedance spectroscopy (EIS). All measurements were performed at 25 °C in a frequency range from 100 kHz to 1 Hz with a potential amplitude of 10 mV. The investigated samples (13 mm discs) had previously been swelled for 24 h in a 2 M lithium acetate (LiOAc) aqueous solution, and were placed in a test vessel between two blocking electrodes made of platinum (Pt) (Fig. 1a). The ionic conductivity (σ) of the examined hydrogel electrolyte was calculated as: where t s is the thickness of the swollen membrane, A is the surface area of the working electrode (0.0177 cm 2 ), and R is the resistance of the membrane. For comparison, the ionic conductivity of a glass fiber membrane (Whatman GF/A) saturated in 2 M LiOAc aqueous solution was also evaluated.
(2) σ = t s /(A × R) The symmetric EDLC cell was assembled using a two-electrode Swagelok ® system (Fig. 1b). Activated carbon cloth (Kynol ® No ACC-507-20; 2000 m 2 g −1 ) and gold discs (Au) were used in the test cell as electrode material and current collectors, respectively. The electrodes were shaped as 6 mm diameter discs, and the mass of a single electrode was in the range 4.6-4.8 mg. The chitin-EG membrane, after 24 h of treatment in a 2 M LiOAc aqueous solution, was applied as a hydrogel electrolyte. An EDLC cell assembled using the same electrode material and electrolyte as for the chitin-EG cell was used as a comparative system. In this EDLC, a glass fiber membrane (Whatman GF/A), with a pore diameter of 1.6 µm and a thickness of 0.3 mm, was used as the separator.
Electrochemical measurements . Electrochemical analysis of the EDLC cell with the chitin-EG hydrogel electrolyte was performed using potentiostatic EIS, cyclic voltammetry (CV), and galvanostatic charge/discharge (GCD) techniques. EIS was carried out using a μAutoLab FRA2 type III electrochemical system (Eco-Chemie, Netherlands) in a frequency range from 100 kHz to 0.01 Hz, with a sinusoidal excitation signal of amplitude 10 mV. CV measurements were performed using the same electrochemical system in the potential range 0-0.8 V and with different scan rates (2 to 100 mV s −1 ). Chronoamperometric measurements were performed at a constant current of 5 mA with the cell voltage stepped from 0 to 0.8 V, using an Atlas 0461 MBI multichannel electrochemical system (Atlas-Sollich, Poland). All electrochemical measurements were performed at room temperature (25 ºC).

Assessment of membrane stability in 6 M KOH using FT-IR imaging.
Samples of prepared membranes were immersed in 6 M KOH for 120 h. Assessment of chitin-EG membrane chemical stability was performed using a LUMOS II FT-IR microscope (Bruker). The material was tested in the ATR mode. The imaging was performed in an area of 1000 × 1000 µm. Stability was determined on the basis of the peak area which is characteristic of chitosan and occurs between 1700 and 1610 cm −1 . 60 scans were collected for each spectrum. The results were processed using the OPUS software.

Results and discussion
The selection of [BMIM][Ac] as a chitin solvent was motivated by the fact that this IL is one of the best-known representatives of this class of compounds, which contributes to its relatively low price and good market availability.
[BMIM][Ac] is currently one of the best-studied ILs and has proved to be extremely effective in various applications, including the delignification of wood 44 , enhancement of the activity and stability of lipase 45 , CO 2 capture 46 , and dissolution of cellulose 47 . Additionally, ionic liquids with a imidazolium ring showed a higher ability to dissolve chitin, and ionic liquid with a more basic anion such as acetate could exhibit better capability to break the strong hydrogen bond within chitin 48 . Additionally, while [Bmim] [Cl] can only dissolve the noncrystal domains of a native chitin material, acetate anion led to dissolution of both non-crystal as well as the compact crystal domains of various chitin polymorphs 49 . Scanning electron microscopy images of the chitin-EG films are shown in Fig. 2. The cross-section of the membrane (Fig. 2b) confirms the good homogeneity of the membrane obtained and shows that regenerated chains are oriented in parallel, forming a compact structure. The thickness of the membrane is 27 μm. It is visible that the surface of the obtained film is relatively smooth (Fig. 2c). This confirms that the presence of ethylene glycol did not cause discontinuities or porous structures when it was mixed with IL-regenerated chitin. The contact angle is an essential parameter often used to assess the wettability of a solid surface by a liquid. It is defined as the angle, measured through the liquid phase, between the tangent of the gas-liquid interface at the intersection of gas, liquid, and solid, and the solid-liquid interface 50 . The contact angles of water (a polar liquid) and diiodomethane (a nonpolar liquid) as functions of the droplet interaction time are shown in Fig. 4. The initial contact angle of a water droplet on the surface of the chitin-EG membrane was 47°, decreasing slightly to a constant value of 43° in a few seconds as the water droplet interacted strongly with the membrane surface due to its hydrophilic character. The contact angle recorded for diiodomethane was constant and equal to 37.34° ± 0.22°. The final angle recorded for water is of more practical significance for chitin-EG, because liquid water is the medium in the environment of an EDLC. However, the obtained contact angle values indicate that chitin-EG thin film might be applied for both aqueous and non-aqueous electrolytes. This was confirmed by the swelling factor for chitin-EG with respect to a 2 M LiOAc electrolyte. The determined SF was 154 ± 3%, indicating the good swelling properties of the prepared gels 51 . The specific properties of a composite material depend strongly on its surface energy and specific interaction forces (e.g., hydrogen bonding, acid-base type interactions, dipole moments, charge transfer or electron acceptor-donor complexes) 52 . The calculated values www.nature.com/scientificreports/ of the total surface free energy and dispersive and polar components for chitin-EG thin films are presented in Fig. 5. Figure S4 demonstrates the images before and after impregnation with 2 M LiOAc electrolyte, which confirm that chitin-EG hydrogel suppress electrolyte leakage. Figure 6 shows a Nyquist plot for the chitin-EG hydrogel, accompanied by a diagram of the fitting circuit 53,54 . The values of ionic conductivity (σ), thickness in swollen state (t s ), bulk resistance of electrolyte (R) and mass (m s ) for the membranes examined in aqueous solution of 2 M LiOAc (24 h) are given in Table 1.
The ionic conductivity value of the chitin-EG hydrogel is 15.7 mS cm −1 , while the value obtained for the Whatman GF/A separator under the same conditions was more than twice as high (37.5 mS cm −1 ). This phenomenon can be explained by differences in the spatial structure and the amount of absorbed electrolyte between the chitin-EG membrane and the Whatman GF/A separator. The chitin-EG hydrogel polymer chain network is tight and creates tunnels and gaps accessible to electrolyte ions simply by swelling, while the Whatman GF/A separator has pores with a predetermined diameter (easily accessible to electrolyte ions). However, the ionic conductivity of the tested chitin-EG hydrogel is comparable to that of other membranes of polysaccharide origin 19,55 .
Nyquist plots for chitin-EG and Whatman GF/A test cells before the first and after the 10,000th GCD cycle are displayed in Fig. 7a,b. All obtained EIS spectra exhibit the typical shape for an EDLC. In the high-frequency range a relatively large semicircle is visible, and in the low-frequency range the EIS spectrum is a straight line parallel to the imaginary part of the impedance axis. Therefore, good capacitive behavior can be expected for the tested EDLC cells. Analyzing the insets of Fig. 7a,b (magnified high-frequency region) and literature data, the origin of the semicircular curve can be explained by three independent factors: (i) the interfacial resistances of the carbon material and (ii) the connection of the current collector and electroactive material, as well as (iii) the distribution of the pore structure and the electrolyte resistance therein. The two most important parameters extracted from EIS analysis and used to characterize the performance of EDLCs are the series resistance (R s )    (Table 1) for swollen membranes. Despite the large gap between the ionic conductivity values of the swollen membranes, the difference in their R s values is not as high as might be expected. This phenomenon can be attributed to the thickness of the membrane in the swollen state. Given that R s is associated with the ionic resistance across the partition between the two facing electrodes, the limited swelling properties of the chitin-EG membrane may positively influence the bulk resistance of the hydrogel electrolyte. This resistance may be additionally boosted by the possible better inter-surface connection of chitin-EG hydrogel with the active material of the electrodes (obtained by pressing it into the carbon cloth fibers). This phenomenon is hardly observed in the case of an inductile glass fiber separator, and may influence ionic transportation between the electrodes. Moreover, detailed analysis of values of R s after 10,000 GCD cycles revealed a slight drop to 1.33 Ω and 1.22 Ω for chitin-EG and Whatman GF/A EDLCs, respectively. These results correspond with a general trend (also confirmed by further CV and GCD measurements) for the electrochemical performance of the tested cells to improve with an increasing number of work cycles. This phenomenon is not directly related to the structure of the tested material or its pseudo-capacitive behavior (CV, GCD), but is rather caused by the enlargement of the active surface of the electrode material associated with gradual unblocking of the micropores filled with microbubbles of air. Another crucial factor in the EIS analysis is the relaxation time constant (t 0 ) resulting from detailed analysis of the dependence of the real and imaginary parts of capacitance (C′ and C′′) on the frequency. Figure 7c,d show the evaluation of the imaginary part of capacitance (C′) with characteristic frequencies and the real part of capacitance (C′) versus frequency. The relaxation time constants can be calculated using the following equation: where f 0 is the maximum of the C′′(ω) curve 20,57 . The peak frequencies of the C′′(ω) curves were determined and are presented on the graph (Fig. 7c). Relaxation time values obtained for the EDLC test cells exhibit a trend similar to the R s values calculated from EIS measurements. Both EDLC cells had the same time constant of 5.8 s. The relaxation time defines the boundary between capacitive behavior and resistive behavior of an EDLC; therefore, it can be assumed that a capacitor with a chitin-EG hydrogel electrolyte will exhibit similar values of specific capacitance as the reference cell.  www.nature.com/scientificreports/ Figure 8a,b show cyclic voltammograms at different scan rates (from 2 to 100 mV s −1 ) recorded for the chitin-EG capacitor and for the Whatman GF/A reference cell. The shape of the CV curves for both tested EDLC cells can be described as nearly rectangular, indicating good charge propagation (up to 10 mV s −1 ). The voltammograms show no significant deviations (such as redox peaks) in the full spectrum of scan rates used, which may be interpreted as a lack of faradaic capacitance within the examined EDLC cells. Thus, the assembled test cells should be treated in further calculations as regular EDLC devices 58 . The CV curves in Fig. 8a,b recorded at higher scan rates (from 25 mV s −1 inclusive) indicate a significant drop in charge propagation; nevertheless, this is a standard phenomenon for EDLCs at higher scan rates 59 . The voltammograms in Fig. 8c,d were recorded at a scan rate of 10 mV s −1 for both tested EDLCs: before the galvanostatic charge/discharge test (GCD) (Fig. 8c) and after 10,000 cycles of GCD (Fig. 8d). The comparison in Fig. 8c shows that before the GCD test, the EDLC cell with chitin-EG hydrogel electrolyte and the reference EDLC cell with Whatman GF/A separator had similar CV characteristics (72 F g −1 and 71 F g −1 , respectively). The specific capacitance values for all capacitor cells were calculated using the following equation: where I is the current, dE/dt is the potential scan rate, and m is the total mass of the active electrode material. An interesting phenomenon is observed in Fig. 8d, which shows the CV curves (10 mV s −1 ) after 10,000 GCD cycles. Here, the specific capacitance values calculated for the chitin-EG and Whatman GF/A EDLCs are 88 F g −1 and 82 F g −1 , respectively. This difference indicates that the test EDLC cell with the chitin-EG hydrogel may exhibit high cyclic repeatability (better than that of the reference cell). Moreover, the absence of any deviations within the chitin-EG CV curve proves that even after 10,000 cycles of charging and discharging, the hydrogel material used was electrochemically stable and did not decompose. To test the cycling stability of the chitin-EG and Whatman GF/A test cells, galvanostatic charge/discharge (GCD) tests were performed. Figure 9a,b show the GCD profiles recorded for tested EDLC cells at the 1st and 10,000th cycle (in a potential range from 0 to 0.8 V). Figure 9a,b show that the potential/time dependency curves for the chitin-EG and Whatman GF/A test cells exhibit a triangular shape for both the 1st and 10,000th cycles. This is a typical GCD profile for EDLCs and indicates the non-faradaic nature of the capacitive properties (proved also by cyclic voltammetry tests) and good cyclic repeatability 60    www.nature.com/scientificreports/ shows that the 10,000th GCD profile recorded for the Whatman GF/A EDLC cell exhibits a slight shift of the curve relative to the initial one. This result indicates that an EDLC cell with chitin-EG hydrogel has better cyclic repeatability than the reference cell, which may be linked to a better inter-surface connection of the chitin-EG hydrogel with the active material of the electrodes (noted in the EIS analysis). Specific capacitance values were obtained from the corresponding GCD curves according to the equation: where I is the discharge current, dU/dt is the slope of the discharge curve, and m is the total mass of the active electrode material. Table 2 gives specific capacitance values obtained after the 1st and 10,000th GCD cycle with a constant current of 5 mA.
The chitin-EG EDLC cell has a higher specific capacitance value at both the 1st (106 F g −1 ) and 10,000th (109 F g −1 ) cycle, as well as excellent repeatability. The Whatman GF/A EDLC cell has slightly lower but comparable specific capacitance values (101 F g −1 at the 1st cycle and 105 F g −1 at the 10,000th cycle of GCD). Analysis of other EDLCs with comparable architecture described in the literature (Table 3) shows that the electrochemical characteristic of the chitin-EG capacitor is competitive with other cells containing polysaccharide-based hydrogel electrolytes (e.g. chitin composites, cellulose or chitosan). Despite the low ionic conductivity of the chitin-EG hydrogel electrolyte, the tested EDLC exhibits a specific capacitance value comparable with those of other materials. A similar phenomenon can be observed for cross-linked chitosan membranes, which indicates that swelling properties of the hydrogel matrix have a large impact on the EDLC's electrochemical performance. Thus, crosslinking in the case of chitosan, and preservation of the natural unmodified crystalline structure in the case of chitin, may play crucial roles in possible electrochemical applications of these polysaccharides. Figure 9c shows the cycling stability of the GCD profiles of the tested EDLC cells. The EDLC with the chitin-EG hydrogel exhibited excellent cycle life, without significant drops in capacitance per 10,000 cycles. These results correspond to the values obtained by cyclic voltammetry and indicate that, despite the lower ionic conductivity, the chitin-EG hydrogel can be recognized as a fully functional EDLC component. Moreover, chitin-EG hydrogel in an EDLC exhibits superior capacitance retention over 10,000 GCD cycles and may be competitive with the commercial glass fiber separator. Figure 9c shows further that specific capacitance increases after 10,000 GCD cycles; this phenomenon should be investigated in detail in future research. However, since this growth is proportional and can be observed for both the chitin-EG EDLC and the Whatman GF/A reference cell, it may be associated with other test cell components (e.g., current collectors or electrode material). Most likely, it is caused by increasing penetration of the electrode pores and specific surface activation by electrolyte ions that are pushed through the electrode bulk.   The chemical stability of pristine chitin-EG membrane and corresponding membrane immersed in 6 M KOH for 120 h, was tested by FTIR imaging microscopic analysis (Fig. 10). The registered two spectra and images appear very similar. Consequently, we can easily conclude that the chitin-EG membrane has good stability in alkaline environment, at least in the limits of the ageing protocol we employed.

Conclusions
Chitin-ethylene glycol membranes were prepared effectively using an ionic liquid (1-butyl-3-methylimidazolium acetate) as the chitin solvent. Structural examination revealed the smooth and homogeneous morphology of the obtained membranes and confirmed that mixing with ethylene glycol does not cause the formation of discontinuous or porous structures in the cross-section of the membrane. The tested EDLC cell with a chitin-EG hydrogel exhibits good electrochemical performance and specific capacitance, superior to those of the Whatman GF/A glass fiber separator reference cell. The remarkable mechanical and electrochemical properties of the prepared chitin-based membranes and hydrogel electrolytes, coupled with their simple manufacturing process, make them suitable for use as renewable alternatives to synthetic polymers, which may help to reduce the environmental impact associated with (i) seafood waste and (ii) traditional GPEs and their manufacturing processes. The prepared hydrogel electrolyte appears to be a promising component for the development of green electrochemical capacitors.  License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.