Graphene exfoliation in cyrene for the sustainable production of microsupercapacitors

PAPER • OPEN ACCESS Graphene exfoliation in cyrene for the sustainable production of microsupercapacitors To cite this article: Pedro Moreira et al 2025 J. Phys. Energy 7 035005 View the article online for updates and enhancements. You may also like 2024 roadmap for sustainable batteries Magda Titirici, Patrik Johansson, Maria Crespo Ribadeneyra et al. - Recent status and future prospects of emerging oxygen vacancy-/defect-rich electrode materials: from creation mechanisms to detection/quantification techniques, and their electrochemical performance for rechargeable batteries Sandeep Kumar Sundriyal and Yogesh Sharma - Polymerization-enhanced flexible perovskite solar cells: a mini review Sihao Huang, Bin Han, Shuyan Chen et al. - This content was downloaded from IP address 193.147.173.203 on 23/07/2025 at 16:02 J. Phys. Energy 7(2025) 035005 https://doi.org/10.1088/2515-7655/adca57 Journal of Physics: Energy OPEN ACCESS RECEIVED 4 February 2025 REVISED 14 March 2025 ACCEPTED FOR PUBLICATION 8 April 2025 PUBLISHED 22 April 2025 Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. PAPER Graphene exfoliation in cyrene for the sustainable production of microsupercapacitors Pedro Moreira1, David Carvalho1, Rodrigo Abreu1, Maria D Alba2, Joaquín Ramírez-Rico2,3, Elvira Fortunato1, Rodrigo Martins1, Joana Vaz Pinto1, Emanuel Carlos1,∗and Jo˜ ao Coelho1,2,3,∗ 1CENIMAT|i3N, Department of Materials Science, Faculty of Science and Technology, Universidade NOVA de Lisboa and CEMOP/UNINOVA, Campus da Caparica, Caparica, 2829-516, Portugal 2Instituto de Ciencia de Materiales de Sevilla, CSIC—Universidad de Sevilla, Avda. Américo Vespucio 49, 41092 Seville, Spain 3Dpto. Física de La Materia Condensada, Universidad de Sevilla, Avda. Reina Mercedes SN, 41012 Seville, Spain ∗Authors to whom any correspondence should be addressed. E-mail: [email protected] and [email protected] Keywords: graphene, cyrene, energy storage, flexible electronics, sustainable fabrication Supplementary material for this article is available online Abstract Graphene and its composites have attracted much attention for applications in energy storage systems. However, the toxic solvents required for the exfoliation process have hampered the exploitation of its properties. In this work, graphene dispersions are obtained via liquid phase exfoliation (LPE) of graphite in cyrene, an environmentally friendly solvent with solubility parameters like those of N-methyl-2-pirrolidone. The obtained dispersions with a concentration of 0.2 mg ml−1comprised multilayered graphene sheets with lateral sizes in the hundreds of nanometers, as confirmed by scanning electron microscopy, transmission electron microscopy, and Raman spectroscopy. Mixing the obtained dispersions with ethanol made it possible to collect the graphene, which was redispersed in 2-Propanol. This active material was used to fabricate supercapacitor electrodes using a scalable spray deposition method on carbon nanotube (CNT) current collectors with the aid of vinyl masks. The device, tested with a PVA/LiCl gel electrolyte, achieved a specific capacitance of 3.4 mF cm−2(0.015 mA cm−2). In addition, the devices show excellent cycling stability (>10 000 cycles at 0.5 mA cm−2) and good mechanical properties, losing less than 10% of initial capacitance after 1000 bending cycles. This work demonstrates the adaptability of liquid-phase exfoliation to produce graphene sustainably, providing the proof-of-concept for further 2D materials processing and green microsupercapacitor (MSC) fabrication. 1. Introduction Since its discovery, graphene, a single layer of carbon atoms arranged in a two-dimensional (2D) honeycomb lattice, has attracted immense scientific and technological interest. Its extraordinary properties, including high electrical conductivity, mechanical strength, thermal conductivity, and a large specific surface area, have positioned it as a promising material for a wide range of applications, namely microsupercapacitors (MSCs) [1–5]. Although initially isolated by micromechanical exfoliation, the development of graphene inks compatible with traditional deposition methods has led to the design of a variety of MSCs. Graphene based MSCs have been already fabricated by spray coating [6–8], inkjet printing [9–11], screen-printing [12–14], among others. Liquid-phase exfoliation (LPE) has emerged as a cost-effective and scalable approach for producing graphene inks [15–17]. LPE is a process where bulk graphite is dispersed in a solvent and subjected to ultrasonic energy or shear forces to separate the layers into single or few-layer graphene sheets [18,19]. The choice of solvent is critical in this process as it influences the exfoliation efficiency and the quality of the resulting graphene. Traditionally, solvents like N-methyl-2-pyrrolidone (NMP), dimethylformamide have been used for LPE due to their ability to match the surface energy of graphene, © 2025 The Author(s). Published by IOP Publishing Ltd J. Phys. Energy 7(2025) 035005 P Moreira et al facilitating the exfoliation process [20–22]. However, these solvents are toxic, volatile, and pose significant environmental hazards, leading to a growing demand for greener alternatives. This has spurred interest in sustainable solvents, with cyrene (dihydrolevoglucosenone) emerging as a notable candidate due to its green credentials and effectiveness. Cyrene is a bio-based solvent derived from renewable cellulose. It has gained attention as a sustainable alternative due to its low toxicity, biodegradability, and excellent solvating properties [23–26]. Several studies have demonstrated the effectiveness of cyrene for graphene exfoliation as its solubility parameters (δD;δP;δH) of (18.7 MPa1/2; 10.8 MPa1/2; 6.9 MPa1/2) are very similar to those of NMP (18.0 MPa1/2; 12.3 MPa1/2; 7.2 MPa1/2) and graphite (18.0 MPa1/2; 9.3 MPa1/2; 7.7 MPa1/2) [20,27–31]. For instance, Salavagione et al compared the exfoliation efficiency of cyrene with NMP and found that cyrene produced a higher concentration of graphene with fewer defects [31]. The study highlighted cyrene’s superior ability to stabilize graphene, resulting in a higher-quality product. Other studies explored the scalability of the process, demonstrating that large quantities of high-quality graphene could be produced using cyrene in a scalable and environmentally friendly manner [29,32]. However, there are still some challenges to consider regarding processing materials in cyrene. Its high viscosity (14.5 cP) and boiling point (227 ◦C) can present practical challenges during the exfoliation process and subsequent steps like purification and deposition, usually requiring solvent exchange approaches [33,34]. Nevertheless, cyrene’s use in graphene exfoliation is a relatively new area of research and more studies are needed to fully understand its long-term viability and optimize processing techniques. In this work, we explore the use of cyrene as suitable solvent for the exfoliation of graphene and subsequent solvent exchange process, so it can be used in scalable spray coating methods. As a proof of concept, the obtained dispersions are used to prepare interdigitated MSCs composed of carbon nanotube (CNT) current collectors and graphene as the active material. The presented methodology allows for a greener and sustainable device fabrication. 2. Experimental methods 2.1. Graphene exfoliation and solvent exchange In a normal procedure, 1.5 g of graphite flakes (Sigma-Aldrich-+200 mesh particle size) were mixed with 30 ml of cyrene in centrifuge tubes with the aid of a vortex mixer. The dispersions were then processed in a Fisherbrand (11207) ultrasonic bath (60% power, 37 kHz) for 8 h, followed by centrifugation in a Sorvall ST8 Centrifuge (4500 rpm (RCF =3260)) for 90 min and collection of the supernatant (top 90% volume). Due to cyrene’s high viscosity, the obtained dispersions were further centrifuged in a Sigma 3–18KS (10000 rom (RCF =9300)) centrifuge for 30 min followed by the collection of supernatant (80%). The concentration of the dispersions was determined by filtering 15 ml on a pre-weighed Whatman alumina Anodisc filter discs (0.02 µm pore size). Then, 10 ml of the obtained dispersions were mixed with 30 ml of ethanol (1:3) and centrifuged in a Sigma 3–18KS (14600 rpm (RCF =20018)) centrifuge for 45 min to precipitate the graphene. The supernatant was discarded, and 10 ml of 2-Propanol (Sigma-Aldrich) were added to the centrifuge tubes followed by vigorous stirring in a Fisherbrand (F202A0280FI) vortex mixer. This process was conducted 3 times to ensure complete cyrene removal. The obtained clean samples were re-dispersed in 10 ml of 2-Propanol and stored for further processing. All chemicals were used as purchased without further treatment. 2.2. Characterization Techniques Scanning electron microscopy (SEM) was used to evaluate the morphology of the raw materials and deposited electrodes in a FEI Teneo, under high vacuum at an operating voltage of 5 keV and a working distance of 10.0 mm. A 10 nm platinum coating was deposited on the samples to reduce charge effects. Further morphological characterization was conducted by transmission electron microscopy (TEM) in a FEI Talos S200 operated at 200 keV in bright field configuration. Raman and x-ray photoelectron spectroscopy (XPS) analysis was used to study the structural composition of the prepared samples. Raman measurements on both graphene inks were collected by a LabRAM Horiba Jobin Yvon confocal Raman microscope using an excitation line of 532 nm with a 100x objective lens, a 250 scans collection in the range between 900 and 3000 cm−1. XPS analyzes were performed using a SPECS Phoibos 150 instrument with a non-monochromatic Al Kαx-ray source (15.4 mA, 13 kV). The instrument’s binding energy (BE) scale (work function) was calibrated to give a BE of 84.0 eV for the Au4f7/2 signal from freshly ion-etched metallic gold (Au). The spectrometer dispersion (energy range) was adjusted to give a BE of 932.62 eV for the Cu 2p3/2 line of freshly ion-etched metallic copper (Cu). A charge compensation system (neutralizer) was used for all non-conductive samples. The surface of each non-conductive sample was irradiated with an accelerated electron flow at 2.0–5.0 eV to produce a nearly neutral surface charge. Scanning analyzes (range 0–1350 eV) were performed using a step energy of 50 eV. High energy resolution analyzes of the chemical 2 J. Phys. Energy 7(2025) 035005 P Moreira et al state (range 20–50 eV) were performed using a step energy of 20 eV. Additionally, a scan (range of 0–1150 eV) was performed with a non-monochromatic Mg K alpha x-ray source (15.4 mA, 13 kV) using a pass energy of 50 eV. All samples have been deposited with gold (Agar sputter coater equipment) so that the data can be corrected (referenced) by loading using the Au4f7/2 signal. CasaXPS v.2.3.26PR1.0 software was used to perform curve fitting, after performing a Shirley background correction, and to calculate the atomic concentrations. Curve fitting of Al2s and O1s spectra were performed using a Gaussian–Lorentzian peak shape. For the analysis of C1s spectra, starting fitting parameters like the approach taken for Biesinger were used [35]. These starting fitting parameters include the main peak asymmetry (defined using an asymmetric Lorentzian line shape) and πto π∗shake-up satellite typical of graphite. 2.3. MSC electrode fabrication For depositing the MSC electrodes, interdigitated patterns with a finger width of 0.8 mm, length of 10 mm and spacing of 0.8 mm were produced by cutting interdigitated electrode (IDE) patterns in a 76 µm adhesive vinyl using a Silhouette Portrait plotter and transferred to polyimide substrates. The IDE were then peeled off from the substrate creating an adhesive stencil that worked as a physical mask during the deposition. The spray coating was conducted with an airbrush over a hot plate (∼10 cm) at 85 ◦C to promote fast solvent evaporation. First, 50 ml of a 0.1 mg ml−1dispersion of Carbon Solutions P3 single wall CNTs (SWCNTs) in 2-Propanol were deposited by spray-coating to create the current collectors. The SWCNTs dispersion formulation can be found elsewhere [36]. The electrode sheet resistance was measured in an Ossila Four-Point Probe system. Then the previously prepared graphene dispersions were sprayed over the fingers in an area of ∼1 cm2. At the end, the vinyl masks were removed leaving the deposited SC behind. The current collectors (0.5 cm ×0.5 cm) were coated with silver paste and cured for 1 h at 60 ◦C. An aqueous PVA/LiCl gel was used as a solid electrolyte. Typically, 4 g of PVA were added to 40 ml of distilled water at 90 ◦C with continuous stirring. After complete solubilization of the PVA, 8.5 mg of LiCl was added to the solution. The mixture was kept under continuous stirring and heating until it became a transparent gel. Prior to electrolyte casting, the MSCs were cured in an Ossila UV Ozone Cleaner for 15 min at room temperature to reduce the hydrophobicity of graphene. The electrolyte was then drop-cast onto the graphene layer. The assembled SCs were allowed to dry overnight at room temperature before the electrochemical testing. 2.4. Electrochemical characterization The as-prepared MSCs were characterized in a CS310X potentiostat (Corrtest Instruments) by means of cyclic voltammetry (CV) (5–1000 mV s−1), galvanostatic charge discharge (GCD) (0.015–1 mA cm−2) experiments and electrochemical impedance spectroscopy (5 mV, 1 MHz to 10 mHz). The device specific areal capacitance, C, was calculated from charge–discharge curves as follows: C=I∆t A∆V(1) where Iis the applied current, ∆tis the discharge time, Ais the SC active area, and ∆V=V2−V1 where V2 is the potential at the beginning of discharge, after the iR potential drop, and V1is the potential at the end of discharge. Areal energy (E) and power (P) densities per unit area were calculated as follows: E=1 2 C∆V2 3600 (2) P=E ∆t×3600 (3) where 3600 is a conversion factor from Ws to Wh. The deformation tests were conducted by acquiring CVs at 10 mV s−1after 10 essays of 100 bending cycles in a total of 1000 bending tests. 3. Results and discussion 3.1. Graphene processing and ink formulation In this work, commercial graphite flakes were used as the starting material for the exfoliation process and processing of graphene in cyrene (GC). Figure 1(a) shows a SEM micrograph of an unprocessed flake exhibiting the characteristic layered structure of graphite. After sonication, a dark ink is obtained, as shown in figure 1(b), composed of thin graphene sheets with lateral sizes in the order of hundreds of nanometers (figure 1(c)). The concentration of these inks was determined to be ∼0.2 mg ml−1, corresponding to an exfoliation yield (%E) of 0.4%. 3 J. Phys. Energy 7(2025) 035005 P Moreira et al Figure 1. Examples of (a) graphite flakes under SEM; (b) graphene dispersion after sonication; (c) TEM micrographs of the exfoliated graphene in cyrene (GC) and (d) TEM micrographs of graphene after redispersion in 2-Propanol (GP). In the literature it has been reported concentrations of graphene exfoliated in cyrene as high as 3.70 mg ml−1(%E ∼7.4%) [29] sonicated for 8 h. However, in this specific work, a combination of exfoliation methods (tip sonication and shear mixing) was used to maximize the graphene production yield. Salavagione et al obtained a %E of 47% for graphite sonicated in cyrene for 2 h with the aid of a sonic tip [31]. It is evident that the processing conditions of graphene exert a substantial influence on the exfoliation yield. As illustrated in table S1, the graphite starting concentrations and sonication methodologies demonstrate a considerable impact on the production of graphene, even in instances where the same solvents are employed. Additional exfoliation parameters, such as sonication power and duration, and centrifugation speed, also exhibit a substantial effect on the exfoliation yield [37]. Moreover, it cannot be disregarded the structure and flake size of the graphite used. Ng et al have recently shown that besides initial concentration, these graphite properties have a significant impact on the exfoliation efficiency of graphene in a wide range of solvents [38]. Additionally, it is also shown that exfoliation efficiency and dispersibility can be analyzed as independent properties. For instance, the mass of graphene per surface area for graphite (particles size ∼150 µm) exfoliated in 2-Propanol was 29 times higher than the values measured for NMP. For a particle size of ∼50 µm, this value rose up to 48. However, this exfoliation efficiency is hindered by the 2-Propanol lower graphene dispersibility, which is also dependent on the graphite properties (graphite morphology, lateral size, defect density, edge defects, d-spacing, among others) [38]. As such the reported values for graphene exfoliated in cyrene may in fact vary across different works reported in the literature. For our study, we preferred to use a lower power sonic bath as it causes less damage to the samples and produces larger flakes, although usually at a lower concentration. It was also used longer centrifugation cycles, which probably contributed to further removal of material in suspension. As such, in this work it is shown that it is possible to exfoliate GC and use it for energy storage applications. As with NMP, the high boiling point of cyrene poses some challenges to its use on a larger scale. Therefore, it is normal to carry out a solvent exchange process. In this work, the obtained dispersions were first mixed with ethanol in a ratio cyrene: ethanol (1:3). The poor dispersibility of these solvents combination greatly facilitates graphene deposition, as a clear transparent supernatant is obtained upon centrifugation. After recovering and cleaning the samples, the graphene was re-dispersed in 2-Propanol (GP—Graphene in Propanol), which has been considered a green solvent, while exhibiting a relatively low boiling-point (82.3 ◦C). A TEM micrograph of the graphene re-dispersed in 2-Propanol (figure 1(d)) is quite like the original samples processed in cyrene. To investigate if there are structural or chemical changes to the graphene during the solvent exchange process, the samples were examined by means of Raman Spectroscopy and XPS (figure 2). Figure 2(a) shows the Raman spectra for both GC and GP where the characteristic peaks of graphene can be identified. The D band (∼1350 cm−1) is ascribed to the A1gmode of vibration and it is usually associated with the degree of defects and structural disorder of the carbon matrix. The G band (∼1580 cm−1) 4 J. Phys. Energy 7(2025) 035005 P Moreira et al Figure 2. (a) Raman spectra of the graphene samples in cyrene (GC) and 2-Propanol (GP); (b) close-up of the Raman peak of the graphene samples comparing the full width half medium (FWHM) for both samples and (c), (d) XPS spectra for C1s orbitals of GC and GP, respectively. corresponds to the in-plane E2g vibration mode of C–C bond stretching. Finally, the 2D band (∼2700 cm−1), a D-band overtone, results of double resonance between the K-points in the Brillouin zone [39,40]. It has been verified that for monolayer graphene I2D >IG. In our samples, the weaker and symmetrical 2D peak is indicative of a few-layers graphene [41–43]. For the example given in figure 2(a), a more intense 2D peak is observed for GP. In the case of solution processed graphene, extreme care should be taken when analyzing Raman data as random restacking of the flakes and defects can significantly change the spectra [42]. The full width-half medium (FWHM) of 2D peak is also indicative of number of layers in a sample. In this work, the FWHM for GC and GP was measured as ∼81 cm−1(figure 2(b)), which may be indicative of the same number of layers for both samples. Nevertheless, no significant changes are observed for either GC or GP spectra. As such, based on previous literature, it is safe to assume that the Raman spectrum corroborates the observations made for the TEM micrographs. Figures 2(c) and (d) show the C1s spectra of both graphene samples exhibiting the characteristic C–C sp2signal graphitic carbon at 284.4 eV. The π-π∗transition loss peak is detected at 291.5 eV [44]. Other chemical shifts are detected at 286.4 eV, 287.9 eV and 289.9 eV which are typically assigned to C–OH/C–O–C, C =O and O-C =O functional group respectively [35,44]. It is common to observe oxygenated species in exfoliated graphene due to residual solvent molecules. In addition, as the measurements were performed in filtered films, it is possible that a larger amount of solvent is trapped in the layers of the filtrate. Moreover, shown by Skaltsas et al sonication induces defects and oxygenated species on graphene which contribute to the sp3signal [45,46]. This observation is in good agreement with the D peak observed in the Raman spectra. The fact that the D and G bands are not broad (as observed for graphene oxide and reduced graphene oxide) suggests that the defects present are not basal plane defects but rather edge defects [46]. Finally, the sp3signal located at 285.3 eV can also be attributed to the presence of adventitious carbon on the sample. Further XPS analysis confirming the presence of graphene can be found in the provided supplementary information (section S2). The XPS data corroborates the Raman results, showing that no damage or sample deterioration occurred during the solvent exchange process. As such, GP will be tested as supercapacitor electrode active material in the following sections. 5 J. Phys. Energy 7(2025) 035005 P Moreira et al Figure 3. (a) vinyl mask on polyimide substrate; (b) spray-coated MSC; (c) 5x magnification of the interfingers spacing and (d) SEM micrograph of the deposited graphene film. 3.2. Electrode fabrication and characterization In this work, the MSC electrodes were fabricated by spray coating the GP into a polyimide substrate (Kapton) with the aid of adhesive vinyl masks as shown in figure 3(a). First, a conductive SWCNT layer was spray coated over the whole mask to deposit the current collectors (∼20 Ωsq−1). In this work, SWCNTS with 1–3 atomic % carboxylic acid groups were used due to their dispersability in 2-Propanol. Then, 10 ml of the GP dispersions were deposited over the fingers area to achieve a mass loading of ∼1 mg cm−2. After painting the silver electrical contact pads, the mask is lifted, a gel electrolyte is applied to the GP coated fingers and a complete device is obtained (figure 3(b)). This device has (2 cm ×2 cm) composed of fingers with ∼0.09 mm separated by a 0.06 mm gap (figure 3(c)). The magnification of the sprayed GP (figure 3(d)) shows that it is composed of sheet-like graphene flakes. Regarding dimensions, they seem to be in good agreement with the TEM micrographs (figure 1(d)). Larger graphene flakes could be expected for sonic bath processing cyrene’s high viscosity may also contribute towards the stabilization of larger flakes. The full process for the preparation of the masks and MSC fabrication is described in the provided supplementary information (section S3). The electrochemical and energy storage properties of these GP-based MSC were extensively analyzed through CV and GCD tests. PVA/LiCl was employed as the electrolyte, while the GP and CNT layers served as the MSC active material and current collectors, respectively. These results are shown in figure 4. The CV curves (figure 4(a)) display a quasi-rectangular shape (5–100 mV s−1), typical of carbon-based electric double-layer capacitors, suggesting good electrochemical stability. CV curves up to 1 V s−1are depicted in figure S7(a). These results are further corroborated by the GCD curves presented in figure 4(b), where the nearly ideal symmetric triangular shapes and minimal voltage drop indicates a strong capacitive behavior. By applying equation (1) to the discharge curves, the areal capacitance, C, was calculated for the different currents applied during the GCD tests (figure 4(c)). A maximum capacitance of 3.34 mF cm−2was measured for a current density of 0.015 mA cm−2. At a higher current density of 0.5 mA cm−2, the GP MSC still demonstrates a Cof 1.34 mF cm−2(figure 4(c)). These results are in line with other graphene and carbon based MSC. Shi et al reported a Cof 4.9 mF cm−2 (at 2 mV s−1) for spray coated Graphene/PEDOT: PSS composites [47]. Coelho et al prepared laser-induced graphene (LIG)—MSC on paper delivering 4.6 mF cm−2(0.015 mA cm−2). Finally, Mendonza-Sanchez et al measured a specific capacitance of 0.543 mF cm−2for ultra-thin electrodes of graphene exfoliated in liquid phase [46]. Similar comments can be made regarding energy and power densities (figure S7(b)). The GP spray coated MSC delivered a maximum of 0.23 µWh cm−2(at 0.45 µW cm−2). It is also important to 6 J. Phys. Energy 7(2025) 035005 P Moreira et al Figure 4. Summary of the electrochemical characterization tests conducted on the GP spray coated supercapacitors: (a) cyclic voltammograms (CV) on the range 5 mV s−1–100 mV s−1; (b) galvanostatic charge discharge curves at current densities ranging from 0.015 mA cm−2to 1 mA cm−2and (c) respective areal capacitance, C; (d) capacitance retention and coulombic efficiency at 0.5 mA cm−2and (e) devices Nyquist plot. For simplicity purposes C will used throughout the text as the symbol for areal capacitance. mention the fact that there is a capacitance retention of 42% at a current density of 0.5 mA cm−2 (figure 4(c)). This capacitance drop is less than ideal, but the GP MSCs were prepared without any conductive additive. PEDOT:PSS is frequently used to lower the sheet resistance of graphene films [29]. However, this polymer can also be used as MSC electrode active material. In fact, graphene and metal oxide particles have been tested as additives for PEDOT:PSS hybrid electrodes [48–50]. To fully study the charge storage capabilities of GP, no conductive materials were added, and thin films were preferred. In terms of cyclability, a 14% capacitance loss is observed after 10 000 cycles (0.5 mA cm−2). However, this process is more evident in the first 1000–2000 cycles, upon which the values of C tend to stabilize (figure 4(d)). There are several factors that affect supercapacitor activity, leading to device degradation and eventual loss of capacitance. Electrolyte decomposition and water evaporation, as well as eventual delamination of active materials and the presence of residual contaminants, are known effects that contribute to capacitance fading in supercapacitors [51,52]. For the GP-based microsupercapacitors, capacitance degradation is more pronounced in the first cycles, suggesting that it is more likely based on the stabilization of the electrolyte water content and the possible presence of species that contribute to non-reversible reactions. After 2000 cycles, the specific capacitance stabilizes and the loss of performance (<2%) is attributed to the aging mechanisms of the supercapacitor, namely electrolyte degradation and loss of surface area [52]. The Coulombic efficiency (figure 4(d)) remains at 97.86% after 10 000 cycles representing a small capacitance loss upon each charge/discharge cycle. The EIS spectrum (figure 4(e)) is characteristic of porous electrodes exhibiting double layer capacitance with frequency dispersion resulting in deviation from verticality of the capacitive (low frequency) branch, with no appreciable pseudocapacitance. The spectrum was fitted using an element (Zpores) implementing the de Levie model for ideally polarizable porous electrodes incorporating a constant phase element, with an additional capacitor (Cflat) in parallel to account for the large external area of the electrodes due to the high aspect ratio of graphene platelets (χ2=3 10−5) [53–55]. The fit reveals an equivalent series resistor (ESR) of ∼160 Ωwhich is common for this type of device with painted silver contacts [56]. A fit to the impedance spectra using the simplified Randle’s model was also performed, but it resulted in a significantly worse fit (χ2=80 10−5) and a charge-transfer resistance of ∼108kΩ, confirming that the GP spray-coated MSCs stores charge exclusively through the formation of double layer with no charge-transfer due to faradaic reactions. A more in-depth discussion regarding EIS can be found in supplementary information (section S4). In relation to applications, in-plane interdigitated MSC on polymeric substrates have been proposed as power sources for wearable technologies and other flexible technologies. Consequently, the MSC must withstand considerable mechanical deformation, such as bending and folding, without compromising the 7 J. Phys. Energy 7(2025) 035005 P Moreira et al Figure 5. GP MSC capacitance retention as function of 1000 bending cycles. The presented data was acquired from cyclic voltammetry experiments conducted at a scan rate of 10 mV s−1. Bottom left inset:examples of the mechanical deformation applied to the samples and bottom right inset: CVs acquired at each consecutive 100 bending cycles. overall electrochemical performance. To evaluate the mechanical properties of the GP spray-coated MSC, the prepared samples were subjected to 1000 consecutive cycles of bending, with the specific capacitance measured at every 100th cycle. The results are shown in figure 5. A capacitance drop of only 8% is observed after 1000 manual bending cycles, suggesting that GP-MSC can be implemented as flexible energy storage devices. However, with adequate encapsulation the MSC may be more stable and retain its capacitance for longer periods of deformation. It is important to point out that a performance uplift (about 10%) can be observed in figure 5. This increase in capacitance has been observed in other flexible devices and is likely to result from an improvement in the electrode/electrolyte interface due to the mechanical deformation, as the electrolyte can penetrate further into the electrode structure and promote the formation of the electrical double layer [57]. It can also be found elsewhere that these devices can easily be arranged in series and parallel to illustrate the scalability and integration capabilities of such systems. As such it is possible to adapt the energy power units to the desired voltages and currents, thus increasing the possible number of applications. It has been shown in this study that graphene processed in cyrene and then transferred to 2-Propanol (GP) has the potential to be used in spray coated MSC elaboration under low temperatures processing. GP-MSC exhibits figures of merit like other graphene inks but being processed in a much safer and environmentally friendly way being compatible with low-cost flexible substrates such as bio-based polymers and paper [58,59]. However, due to its porous nature, paper could absorb the inks and lead to device short-circuiting. In this case the proposed electrodes could be used in a sandwich configuration [58]. Better electrochemical performance and higher capacitances can also be obtained by increasing the film thickness and/or using metallic substrates [7,60,61]. In these cases, conductive polymers or carbon additives and binders may be required, thus impacting the overall gravimetric capacitance and device flexibility. For instance, Garakani et al developed high-power graphene-based supercapacitors operating at a wide range of temperatures [6]. In this study, MSC graphene electrodes were prepared by spray coating onto aluminium current collectors a mixture of graphene, activated carbon and cellulose. In fact, the mixture of graphene with carbon species or dopants (such as nitrogen) has also shown promising results for high performance supercapacitor development [62,63]. As such, there are a lot of opportunities for GP graphene for further MSC development. In fact, it can be highly relevant for simple, low cost, fast patterning, flexible systems that do not need to adhere to highly controlled environments and sophisticated manufacturing protocols, being poised to occupy a niche in flexible/wearable devices manufacturing and may play a role in fighting the ever-increasing amount of electronic waste (e-waste). 4. Conclusions This work has demonstrated a sustainable method for the green exfoliation of GC and its subsequent solvent exchange and deposition as an active material for microsupercapacitor electrodes. The solvent exchange process from cyrene to ethanol was found to be very effective without causing any damage or structural changes to the graphene. In addition, the solvent exchange method does not result in any loss or possible contamination compared to graphene isolation by filtration. The use of vinyl masks facilitated the deposition 8