Unveiling the Potential of Colorless Polyimide‐Derived Laser‐Induced Graphene: A Novel Pathway for Advanced Sensor and Energy Harvester Performance

The potential of laser‐induced graphene (LIG), recognized for its distinct attributes in diverse fields, has significantly grown. However, the creation of LIG using colorless polyimide (CPI) films remains unexplored. This research sheds light on the graphitization technique for generating LIG from CPI films via laser techniques, a process validated through ReaxFF simulations. It is also illustrated that CPI integrated with fluorine atoms possesses an elevated porous configuration, rendering it apt for high‐sensitivity, low‐detection limit pressure sensors. The pressure sensor, constructed with LIG derived from CPI, showcases superior performance metrics such as an exceptional sensitivity rate of 60.340 kPa−1 in low‐pressure ranges (1.0–1.5 kPa), prompt response and recovery intervals (27/36 ms), and commendable durability. The sensor's ability is further validated to precisely track human movements. Moreover, the study employs the LIG sourced from CPI as a dielectric‐to‐dielectric triboelectric nanogenerator (TENG), yielding a peak power output of 411.4 mW m−2 under a 40 MΩ load resistance. The CPI‐based LIG offers increased porosity in comparison to traditional LIG, which aids in superior functioning in pressure sensors and TENG devices. This research offers a fresh perspective on the application possibilities of CPI‐sourced LIG, notably in pressure sensors and energy harvesting devices.


Introduction
In line with the pressing need for sustainable and crossdisciplinary progress in contemporary society, various environmentally friendly energy and healthcare technologies have been fervently pursued over recent years. [1,2]Flexible electrodes have DOI: 10.1002/admi.202300625been a pivotal component in the advancement of renewable energy and health monitoring systems, surpassing the capabilities of conventional electrodes. [3]Due to their adaptability to the curved and uneven contours of the human body, lack of skin irritation during prolonged usage, and compatibility with wearable and mobile devices, flexible electrodes are favored over their rigid counterparts in healthcare and energy harvesting sectors. [3,4]Among the promising prospects for flexible electrodes, various forms of graphene, encompassing pure graphene, graphene oxide (GO), and reduced graphene oxide (rGO), have been broadly investigated in photovoltaic methodologies and triboelectrification due to their extraordinary electrical and optical properties. [5,6,7]owever, the prevalent procedures for creating graphene-based materials come with high energy demands, harmful chemicals, and multistage processes. [7,8]n contrast, laser-induced graphene (LIG), an accidental carbon material produced through direct laser writing on commercial polymer films, offers a promising substitute due to its selectivity towards hydrophilic/hydrophobic properties, swift fabrication speed, and the amalgamation of functional groups. [7,9,10,11]Furthermore, it's been shown that LIG exhibits a more porous structure than other graphene variants, a result of the photothermal effect, which rapidly expels gas into the neighboring atmosphere. [9,11]The distinctive attributes of LIG make it especially fitting for a variety of applications, including pressure sensors and triboelectric nanogenerators (TENG). [12]The porous nature of a flexible electrode provides not only a greater contact surface but also increased sensitivity for detecting minor pressure variations in pressure sensors. [13,14]Furthermore, the existence of supplementary air gaps leads to diminished noise during pressure changes, while also enabling effective stress distribution, thus offering an expanded surface area and superior stability. [14]As a result, LIG can measure a wider pressure range and can be utilized in diverse measurement settings. [14,15]Importantly, the porous electrode structure is defined by a profusion of nanoscale air gaps on the material surface, which efficiently increases the surface area.This surface area expansion subsequently results in an enhancement of the TENG's power density and voltage. [16]oreover, the distinctive structure facilitates concurrent triboelectric charge production and charge transfer within the inner spaces, leading to an augmented electrical output. [17,18]Taking advantage of the properties of LIG, several studies have reported extensive research on LIG created from traditional yellow polyimide substrates via laser intervention, with an emphasis on applications in human motion detection apparatus and energy harvesting. [6,9,19,20]However, recent progress in colorless polyimide (CPI) films, which uphold excellent mechanical, electrical, and electrochemical properties while demonstrating superior optical characteristics, introduce the possibility to overcome the restrictions of conventional LIG substrates for use in display and photovoltaic devices. [21,22]Despite this potential, investigations into LIG generated from CPI films are still relatively uncharted.Therefore, it is imperative to suggest applications for LIG produced from CPI films grounded on experimental findings, verify these applications through simulations, and cultivate a deep comprehension of the graphitization mechanism and structure prediction during the laser processing of CPI.This understanding is key for further progress in this field of research.
Molecular dynamics (MD) simulation has risen as one of the most promising tools to scrutinize graphitized materials in detail.Studies into the graphitization of polymers have proposed various thermal decomposition pathways and gas generation mechanisms.The decomposition process can be divided into three routes: severing C-N bonds in the imide group, forming the dibenzofuran ring and dicarboxylic acid, and cutting C─O bonds in ether. [23,24]Consequently, gases such as H 2 O, H 2 , benzene, and benzonitrile are generated, and the transition from sp3 carbon to sp2-hybridized state carbon takes place. [23,25,26]Besides these mechanisms, our team has proposed the need for multiple lasing due to the formation of sub-graphitized states containing cyanobenzene, isoimide, isocyanate, dibenzo[b,d]furan-2,8,-diamine, and phthalimide during the process. [25,26,27,28,29]Despite ongoing research into the structural analysis of carbonbased materials through molecular dynamics simulations, calculating the competition for hybridization with neighboring carbon elements and determining the potential of carbon materials based on -electron distances remain significant challenges. [30]eactive force-field (ReaxFF) potentially resolve this issue as they can accurately capture the behavior of carbon materials and have been widely used for the characterization of graphitic materials such as carbon nanotubes, graphite, and diamond. [31,32,33,34]wever, the use of ReaxFF simulations to investigate the atomistic conversion process from polyimide to LIG is somewhat limited. [35,36]Kim et al. made a significant enhancement to the force field by including not only carbon but also parameters of ten atom types: H, O, N, S, Mg, P, Na, Ti, Cl, and F in the training set. [37]Rahaman et al. improved the parameters of C, H, O, and N that constitute Kapton, enabling the investigation of the stability of several molecules. [38]These recent efforts present a tremendous opportunity to study LIG formation from various CPI films using ReaxFF force fields. [36,37]n the present work, we establish a method for generating LIG that results in a highly porous structure with substantial air gaps in a singular step.This technique involves creating LIG within a commercially available CPI film (Kolon CPI composite) and its protective layer.The micro-void structure plays a vital role in augmenting the functionality of pressure sensors and TENG apparatus.We examined the morphological and elemental traits of the developed LIG, such as its specific surface area, through various techniques including X-ray Photoelectron Spectroscopy (XPS), Xray diffraction (XRD), Brunauer-Emmett-Teller (BET) analysis, Raman spectroscopy, and Scanning Electron Microscopy (SEM).To fabricate the microporous LIG structures, we utilized femtosecond laser pulses, with a wavelength of 515 nm and a pulse duration of 170 fs, directed towards the interior surface of the CPI film composite using the Yb:KGW laser system.ReaxFF simulations were leveraged to clarify the mechanism of LIG creation and its structural properties.The ReaxFF simulations disclosed that the CPI group containing fluorine atoms manifested a more porous structure, and the CPI-derived LIG electrode manufactured via laser exposure was employed as a pressure sensor and TENG.Given its heightened sensitivity in detecting pressure from minimal movements, the advanced CPI-derived LIG holds immense potential to substantially improve human motion tracking systems.Moreover, the generation of LIG within the film composite affords several benefits, including a single-step process, lack of surface damage, low expense, and high flexibility.Therefore, the CPI-derived LIG proposes promising uses in various fields, such as sports science, wearable technology, and flexible electrodes, with the potential for even more applications to emerge in the future.

Characterization of CPI-Derived Laser-Induced Graphene
In our research, we utilized the field emission scanning electron microscope (FE-SEM) and energy dispersive X-ray spectroscope (EDS) to examine the surface morphology of the laser-induced graphene (LIG) obtained from colorless polyimide (CPI), which was fabricated as depicted in Figure 1.The comprehensive fabrication procedure is explained in the Experimental Section and Video S1 (Supporting Information).Figure 2 presents the examination of surface porosity of LIG with varying laser fluences of 6.65, 13.31, and 19.96 J cm −2 .Interestingly, the beam irradiated with the peak laser fluence resulted in a microporous structure with enlarged pores, possibly due to a swift temperature rise leading to bond breakage, followed by air release and energy trapping in the structure.Prior studies have proposed that the air voids between conductive channels can aid in the functionality of pressure sensors, a finding that our study supports. [12,13,14,15]Additionally, the local high temperatures created on the CPI film surface by laser irradiation also induce a photothermal effect on the adjoining protective film's surface, aiding in the LIG particle's transfer from the CPI film and encouraging the formation of a porous structure (refer to Figure 2e-g).The LIG generated within the CPI film composite, particularly concurrently on both the CPI film and the protective film, displays heightened porosity and surface area.This increased surface area was compared to that of LIG solely produced on the CPI film via Brunauer-Emmett-Teller (BET) surface area analysis (refer to Figure S1, Supporting Information).The BET surface area is given by Equation (1),  and the specific surface areas and pore sizes for the LIG derived from the CPI film, protective film, and conventional method were found to be 131.797,56.658, and 108.498 m 2 /g, along with 44.27, 36.89, and 24.24 Å, respectively: where P /P 0 represents the relative pressure, V m denotes the volume of the adsorbed gas, and C signifies the BET constant, this equation serves to assess the alteration in adsorption gas volume in relation to pressure fluctuations.This implies that the creation of relatively minor air gaps and the shift towards a more porous structure could contribute to enhanced performance of the sensor, with the optimal laser fluence determined to be 13.31J cm −2 .These insights have significant relevance for the advancement of pressure sensors and TENGs that are based on LIG.Raman spectra were employed to inspect the LIG derived from CPI.The D (at ≈1350 cm −1 ), G (at ≈1590 cm −1 ), and 2D (≈2700 cm −1 ) bands of the Raman spectra were scrutinized to pinpoint the typical traits of graphitized material (see Figure 3a).The LIGs produced on the CPI film side displayed similar properties to rGO, whereas the LIGs manufactured on the protective film side resembled graphene oxide (GO) or a relatively less reduced form of GO.The LIG created on the CPI film side showed a higher I d /I g intensity owing to the development of functional groups containing oxygen, the armchair structure at the edge, and various ring carbon structures. [39,40]A higher I d /I g ratio in Raman spectra signifies an increased level of disorder or defect density within the graphitic material, while a lower ratio indicates a more organized and crystalline graphitic structure. [41]With the rise in laser fluence, the I d /I g intensities for the LIG derived from the CPI film were noted to be 0.881, 0.656, and 0.753, in that order, while for the LIG created from the protective film, they were 0.984, 0.972, and 1.150, respectively.The reliability of these outcomes will be further validated through the ReaxFF simulations in the next section.The G peak, resulting from the in-plane vibration of sp2 hybrid carbon, [39] appeared to be stronger in the LIGs fabricated on the CPI film side. [40]Even though the 2D peak was present in both LIGs, a higher laser fluence was necessary for the LIG transferred to the protective film to form the 2D peak.Additionally, the LIG developed on the protection surface owing to the photothermal effect on the CPI film surface is believed to resemble a material like glassy carbon or graphene oxide, as indicated by its Raman spectrum.It was discovered that the LIG was produced in a multilayered structure with defects rather than a single-layer graphene because the I 2D /I G intensities of the LIGs were smaller than the threshold value of 2. [27] The I 2D /I G intensities of the LIG created on the CPI film surface were observed to be 1.331, 0.731, and 0.415, in sequential order as the laser fluence amplified, whereas the I 2D /I G intensities of the LIG crafted on the protective film surface were documented to be 0.265, 0.192, and 0.567, respectively.These findings shed light on the properties of LIG fabricated with different laser fluences, assisting in the refinement of LIG production for various use cases.
In Figure 3b, X-ray diffraction (XRD) analysis indicated that the LIG derived from CPI showcased a notably crystalline structure. [42]The (002) peak could be triggered due to the interlayer spacing of graphene layers. [42]The (004) peak, also noticeable in multilayer graphitic structures, signifies the crystalline organization and layering of graphene sheets. [43]A minor band showing up within the 2-theta range of 45-50 degrees could be ascribed to the existence of functional groups. [44]These findings imply that the laser irradiation process can effectively modify the colorless polymer structure into a highly structured graphene-like form, which could be useful in numerous sectors, including energy storage and sensing.To delve deeper into the CPI-derived LIGs, X-ray photoelectron spectroscopy (XPS) analysis was performed.As outlined in Figure 3c,d, the results indicated a minimal presence of Nitrogen in the sample, echoing the outcomes of our simulations.It is noteworthy that the polymer structure was morphed into a graphitic one, with carbon accounting for 92.1% of the total elemental composition, in line with the outcomes of the ensuing ReaxFF simulation section.The C1s peak of the XPS analysis disclosed five typical bonding structures, with the sp2 carbon structure taking precedence, trailed by the sp3 carbon structure and chemical bonds constituting a functional group.All in all, these results suggest that the LIGs derived from CPI portray a complicated, yet distinct bonding structure.
Our XPS analysis further confirms the graphitization process during laser irradiation, revealing significant alterations in the surface chemistry and carbon bonding arrangement of the LIGs.

ReaxFF Simulation
Ten unique CPI monomers were synthesized using the NPT ensemble methodology, where NPT stands for constant-pressure, as delineated in the methods section.The optimized systems are tabulated in Table 1.The constituent monomers of these polymer films are visualized in a 2D format in Figures 4 and 5 and a 3D format in Figure S3 (Supporting Information).The selected systems were bifurcated into two groups: Group 1 (CPI1-5), devoid of fluorine (F) atoms, and Group 2 (CPI6-10), encompassing F atoms within their chemical compositions.However, CPI5 stands as an outlier, incorporating a Cl atom.To comprehend the graphitization mechanism involved in the formation of CPIderived LIGs, we explored ten distinct CPI monomers employing ReaxFF simulations using the NVT ensemble, where NVT stands for constant-volume.In order to validate the ReaxFF lasing simulations, we juxtaposed data from the simulation of a Kapton film under identical conditions with our previously procured experimental data and data gathered in this study. [9,19]These comparative analyses are included in the supporting information (Figures S4-S6 and Video S2, Supporting Information).Figure 6 illustrates the simulated structures, inclusive of various carbon rings and functional groups during the simulation, alongside the simulation procedure.Figure 7 showcases the molecular structure transformations during the ≈1.5 ns simulation at a temperature of ≈3000 K.It was observed that a temperature of ≈3000 K is conducive for the generation of graphene materials in both polymer groups (refer to Figure S7, Supporting Information).The temperature parameters employed in this simulation can shed light on our choice of comparatively mild laser fluence (13.31J cm −2 ) during the experimental procedure.The structures of three representative CPI monomers were scrutinized at 0.25 ns intervals.
A carbon structure, amorphous in nature and including pyridinic N, nitrile, and hydroxyl groups, was produced after simulating at ≈3000 K for 0.2 ns (Figure S8, Supporting Information).Both potentials from Kim et al. and Rahaman et al. could detect nitrile, pyridinic N, and hydroxyl groups. [37,38]Further lasing simulation at ≈3000 K for 0.8 ns from the molecular structures at 3000 K for 0.2 ns yielded planar, graphene-like structures with ordered carbon rings, as most O atoms detached from the main structure as H 2 O and CO.LIG's distinctive features such as pentagon and heptagon carbon rings, and functional groups were observed.The planar graphitized structure exhibited zigzag and armchair structure edges, contributing to the enhancement of the D band of the Raman spectrum. [39]Additionally, the molecular structures of all CPI-derived LIGs were compared from the inception of the lasing simulation to its culmination.Isoimide, cyclic ether, cyan groups, and six-membered carbon rings were shed due to the C─N, C─O, C═O, and C─H bond cleavage from the early stage (0.2 ns), followed by the gradual generation of planar, graphenelike structures with diverse carbon rings (0.2-1.5 ns).To delve deeper into the effect of multiple lasing on the yield of highquality LIGs, the variations in the number of pentagon, hexagon, heptagon carbon rings, alicyclic/aromatic carbon, and aliphatic carbon over time in the lasing process was recorded in Figure 8 and Figures S9 and S10 (Supporting Information).During the initial 0-0.2 ns span of the multiple lasing simulations, a substantial decline in hexagonal and heptagonal carbon rings is evident.Subsequently, as depicted in Figure 7 and Figures S11 and S12 (Supporting Information), the amorphous carbon structures evolve through the lasing simulation process, transforming into planar, graphitized structures.It should be noted that the lack of sufficient time to equilibrate to the standard hexagonal lattice during laser processing can contribute to the formation of these structures. [25]Figure 8b presents the normalized number of cyclic carbon rings obtained by Equation (2): where N N is the normalized number of 6-membered rings, N 6 (t) is the number of six-membered carbon rings, and N 6 (0) is the number of original 6-membered carbon rings in the optimized structure.At the inception of the lasing simulation, the decomposition of 6-membered rings took place, followed by a gradual progression of five-and seven-membered rings.By assessing the cumulative quantity of cyclic carbon rings, the count of alicyclic/aromatic carbons, and the total of aliphatic carbons displayed in Figures S9e,f and S10e,f, (Supporting Information) the degree of formation of the kinetic graphene structure within the molecular framework can be inferred.Previous research has substantiated that the pyrolysis of polymers caused by localized intense heat during multiple lasing results in the cleavage of C─N, C─O, C═O, and C─H bonds, facilitating the creation of benzonitrile, benzene, H 2 O, CO, and H 2 . [24,25,26]Our ReaxFF simulation outcomes, demonstrated in Figure S13 (Supporting Information), align with these experimental observations.Viewing from a thermochemical standpoint, elevated temperatures tend to generate hydrogen radicals, contributing to a decline in the count of molecules retaining the original hydrogen atom in the hybrid carbon. [53,54]or a comprehensive understanding of gas evolution during multiple lasing, it's crucial to scrutinize the genesis of gases and the movement of molecules from the polymer.Due to multiplication, the concentration of radicals in the system can increase, which could potentially target stable molecules and further facilitate hydrogen exchange, leading to the expulsion of H 2 , CO, and H 2 O gases (as depicted in Figure 8d-f).Figure S14 (Supporting Information) represents the count of gas molecules during the simulation for groups incorporating and excluding F atoms in the polymer structure, respectively.The normalized values of gases discharged from the monomers during the simulation were derived using the following Equation (3): where G N (t)) represents the normalized value of gas molecules, G(t) denotes the number of gas molecules generated at a specific time, and G 0 is the number of elements present in the initial polymer.The number of H atoms and O atoms were used in the calculation of the normalized values for H 2 molecules, CO molecules, and H 2 O molecules, respectively.The elemental mass ratios in the CPI molecules (represented in Figure 8d-e; Figure S14, Supporting Information) indicate H 2 O as the predominant gas in both groups.At an approximate simulation temperature of 3000 K, bonds like C─H, C─O, and C═O with a bond energy ranging from ≈85-170 kcal mol −1 can be broken.Additionally, C═O and C─OH bonds with a binding energy ≈105-187 kcal mol −1 can also be severed, as previously documented in various sources. [53,54,55]This leads to the formation of CO and H 2 O gases and a decline in the oxygen content.The recorded drop in oxygen content during the ReaxFF simulation can be credited to the production of CO and H 2 O gases, as corroborated by past elemental studies. [10,25,36]In the first 0.2 ns of the simulation, there was a sharp increase in CO and H 2 O levels, whereas H 2 concentrations displayed substantial volatility due to the H-radicals, with a gradual upward trend.The fairly high bond dissociation enthalpy of hydrocarbon O─H (≈93-105 kcal mol −1 ) can initiate the formation of carbon and oxygen radicals from the structure upon exposure to photothermal radiation that considerably exceeds this threshold.Ensuing oxidation reactions can then transpire, as mentioned earlier. [53,56]herefore, the results indicate that H-radicals were generated in the initial phases, which were subsequently triggered by photothermal radiation to yield C-and O-radicals from the polymer structure.These radicals then primarily reacted to emit CO and H 2 O gases into the surroundings.The created gaseous compounds could also interact with one another, resulting in a heightened emission of CO and H 2 O gases. [36,52,53]Interestingly, while the initial release of oxygen into the atmosphere was minimal, more than 91% of oxygen was discharged at the conclusion of the simulation from all CPIs, except for CPI7 (≈84%) and CPI8 (≈54%) (refer to Figure S15, Supporting Information).Conversely, the release of hydrogen in the form of H 2 O was considerably high, while the percentage of hydrogen emitted as H 2 molecules was relatively low, ranging from 3.8% to 9.4%.Apart from these representative gases evolved during LIG formation, hydrocarbons were produced during the intermediate state and subsequently fragmented into smaller hydrocarbons, as shown in Figure S13 (Supporting Information).Video S3 (Supporting Information) further substantiates our structural analysis and illustrates the complete carbon skeleton of CPIs in detail.It's observed that the CPI group incorporating F atoms retains a larger hydrocarbon structure within its chemical composition rather than a single extensive graphene structure with numerous defects, which is a defining trait of LIG.As a result, this leads to a higher quantity of graphene layers compared to the group excluding F atoms (refer to Figures S11 and  S12, Supporting Information).This observation strengthens the idea that CPI6-10 can potentially harbor a more stratified structure with a greater number of air voids within the layout.
The simulation data were calculated at intervals of 0.1 ns for surface area measurements.It is important to determine the surface area of LIG because the performance of pressure sensors and triboelectric nanogenerators can be significantly enhanced by employing porous structures.Figure 9 and Figure S16 (Supporting Information) illustrate the variation of surface area over simulation time, normalized using Equation (4) as follows: where S N signifies the normalized surface area, S(t) represents the surface area of a sample at a given time point, and S 0 denotes the surface area corresponding to each individual monomer.It is noteworthy that the average surface area increase for the group without F atoms was ≈194%, while the average surface area increases for the group containing F atoms within their chemical structure was ≈243%.This insight supports the notion that the CPI6-10 group is more suitable for the fabrication of sensors and energy harvesting devices.

Performance of CPI-Derived LIG Pressure Sensor
In order to evaluate the functionality of the pressure sensor derived from CPI-based LIG, we set up a mechanical experimental platform using Arduino technology.Figure 10 illustrates the potential of the pressure sensor.The distinctive CPI-sourced LIG design enables a conductive route with a porous layout between the defensive layer and the CPI layer, all in a one-step procedure for the piezoresistive pressure sensor.While the swift discharge of gases in the LIG electrodes may lead to a reduced number of conductive routes due to the presence of air gaps and porous microstructures, it plays an indispensable part in amplifying the sensor's sensitivity.When pressure is applied to the sensing component, it enhances the conductive route of the porous microstructure housing an air gap, which in turn permits the sensor to detect even the slightest pressure variations.Therefore, this air gap and the porous LIG configuration significantly contribute to boosting the pressure sensor's sensitivity.As depicted in Figure 10a, the pressure sensor with a porous microstructure displays an enhanced detection limit, having the capability to discern pressures as minimal as ≈41 Pa.The sensitivity (S) of the LIG pressure sensor derived from CPI was computed using Equation ( 5): where R 0 and R are the initial and resistance when the particular pressure is applied, respectively.P represents the pressure applied to the sensor in kPa.).In the surveyed pressure scope of up to 1.5 kPa, the collective sensitivity, known as S avg , which combines both S1 and S2 sensitivities, was found to be S avg = 36.894kPa −1 .This shows the device's proficiency in accurately reacting to low-pressure stimuli.However, with an increase in pressure, the sensitivity progressively dwindles owing to the stabilization of the porous microstructure, and the resistance change rate attains a stable condition (S3 = 8.154 and S4 = 0.547 kPa −1 ).In contrast to formerly reported pressure sensors reliant on metal electrodes and the graphene family, our LIG pressure sensor derived from CPI displays heightened sensitivity in the low-pressure range of ≈1.5 kPa, as indicated in Table 2. To confirm the sensor's performance, we studied the discrepancies among sensors fabricated by a solitary step process, shown in Figure S17a (Supporting Information).The outcomes showcase high reproducibility with a negligible amount of variation observed, implying consistently reliable and uniform results.The pressure-dependent I-V curves shown in Figure S17b (Supporting Information) disclose that our sensor exhibits a steady and linear response across a spectrum of pressures, signifying its stable and dependable functioning.The sensor's durability and repeatability were evaluated by applying pressures of 1, 3, and 5 kPa at 1 Hz over 10000 times (refer to Figure 10d; Figure S18, Supporting Information).
The sensor showed uniform and repeatable responses for each pressure, indicating its sturdy and dependable performance.The micro-porous structure, visible in the SEM image, is the driving force behind the working mechanism of the LIG pressure sensor derived from CPI.As pressure escalates, the air gaps in the structure gradually vanish, and the ameliorated conductive channels considerably diminish the electrical resistance.Therefore, a stepwise pressure increase stabilizes the electrical signals.Another benefit of this distinct structural feature is the sensor's response time, which is impressively quicker than other metal-based or graphene-based sensors, as shown in Figure 10c and Table 2, with a response time of 27 ms and release time of 36 ms.The sensor displays exceptional response time and superior pressure sensitivity yet holds the substantial advantage of being constructed in a single step, providing a significant benefit.Moreover, along with reproducibility, another crucial aspect of the sensor, repeatability, was evaluated in Figure 10d,e.The LIG pressure sensor derived from CPI offers high sensitivity, remarkable stability, and swift response speed, thereby qualifying it as a promising tool in sports science and healthcare sectors, including human motion detection.The sensor's micro porous structures and air gaps enable it to discern even the slightest changes in human motion, due to its low detection limits.Figure 11a,b provides tangible proof of the sensor's capacity to perceive minute dynamic forces, such as the intensity of inhalation/exhalation when fitted to a 31-year-old test subject's chest.Further, the sensor showcased a speedy response and steadfast performance, emphasizing its potential to bring a paradigm shift in the sphere of human motion detection.As illustrated in Figure 11c-e, the sensor was proficient in correctly identifying a pulse on the subject's wrist, a delicate dynamic force.More precisely, the sensor was fastened to the subject's wrist using adhesive tape, and the relative resistance curve distinctly separated two pulse bands, the systolic (S) and diastolic (D) bands, as observed in Figure 11e.The pulse wave's two significant peaks are the systolic peak, denoting the peak pressure applied by the heart during systole, and the diastolic peak, denoting the least pressure applied by the heart during diastole.These peaks are vital metrics for appraising cardiovascular health and diagnosing conditions such as hypertension and hypotension.Our sensor can monitor both peaks and holds potential usage in healthcare devices.The sensor further corroborated that the subjects' pulse rates were generally between 70 and 80 beats per minute, vouching for the sensor's heightened sensitivity and reliable performance.The LIG pressure sensor derived from CPI was leveraged to study the pressure distribution changes on a shoe's sole during squats.The sensor was attached to the toe parts of the shoe insole, namely, the start and finish, and real-time data were recorded during ten sets of 20 squats each.Figure 11f-j show the sensor's capacity to perceive pressure distribution shifts across the sole's two regions during the squatting exercise.The squat sensor's effectiveness was evaluated through supervised deep learning, where a deep neural network was trained on labeled sensor data comprising input data (such as pressure values) and corresponding output data (like the sensor's response or classification).The trained model can predict the sensor's response for new input data, thereby authenticating the reliability and accuracy of the squat sensor.The recorded data were then analyzed using a   multimeter and a confusion matrix to establish the overall recognition rate, determined to be over 89% (refer to Figure S19, Supporting Information).The sensor was found to be highly effective in pinpointing the center of gravity, which is directed toward the heel, during the squatting exercise.The heel was verified to have the most constant pressure distribution throughout the exercise, as detected by the relative electrical resistance changes noticed by the sensor. [64,65]And dynamic shifts in pressure distribution were observed in the part of the shoe insole attached to the sole's end during squats.Furthermore, to evaluate our CPI-derived LIGbased pressure sensor's performance, commercial pressure sensors were employed as reference sensors.The sensor with incorporated Bluetooth functionality was used to demonstrate real-time pressure distribution during squatting movements through a commercial program (refer to Video S4, Supporting Information), and the results were compared with those garnered from the commercial sensors, as shown in Figure S20 (Supporting Information).Comparable to the results procured from our sensor, the 31-year-old test subject exhibited maximum load on the left heel during squat movements, with inconsistent changes in pressure.Our findings were consistent with those garnered from the commercial sensors, underlining our sensor's potential applicability in an array of applications.These findings demonstrate the potential of the LIG pressure sensor derived from CPI for assessing pressure distribution changes in shoe insoles during physical exercises.To explore the CPI-derived LIG pressure sensor's further applicability, the sensor was affixed to the wrist to detect muscle changes in response to rock, paper, scissors hand gestures.Figures 11l,m exhibit the real-time response curves of the sensor attached to the wrist to muscle movements based on hand gestures.The need for devices monitoring muscle performance is paramount in sports science and healthcare devices for detecting paralysis symptoms, which is crucial in the contemporary world.Additionally, our sensor holds substantial potential for application in medical wearable devices.The hand gestures, which were altered 20 times each, were tested 1500 times, demonstrating an accuracy rate of 96.5% (Figure S21, Supporting Information).These findings suggest that the LIG pressure sensor derived from CPI holds substantial potential as a monitoring tool for evaluating the effectiveness of various physical activities, as well as for developing personalized training programs based on individualized pressure distribution patterns.

Fabrication of CPI-Derived LIG-Based Triboelectric Nanogenerator
In response to the ongoing societal quest for sustainable energy, the functionality of LIG derived from CPI was assessed as a triboelectric generator (TENG).This could potentially contribute to the escalating requirements for environmentally friendly power solutions.LIG has played a key role in enhancing the performance of triboelectric generator devices as a substitute for metallic materials. [6,55]The sharply focused femtosecond laser beam conveyed into the CPI film composite reliably generates a conductive LIG pathway within the film's subsurface, displaying high conductivity.The conductive LIG (with resistance <250 Ω) was examined for its utilization as a metal-free TENG device.The device was constructed using a sandwich structure comprising a tribo-negative friction surface, TENG electrode, and a tribo-positive friction surface (i.e., CPI substrate/LIG electrode/Polyurethane cover), as detailed in the Experimental Section. Figure 12a portrays a schematic of the LIG-based TENG device, which is produced simply by applying a laser between a rectangular CPI/protection film composite and folding it in half with an affixed polyurethane cover instead of a protective film, thereby forming two triboelectric active areas of 3 × 3 cm 2 for assessment.Due to the robust and pliable features of the polyimide film, this TENG device is capable of producing contact/separate triboelectricity via folding.The concept of the TENG device hinges on the lightweight and flexible properties of LIG, facilitating the generation of dielectric-to-dielectric (D-D) triboelectricity (refer to Figure 12b).Two LIG-based insulating composites bearing opposite charges accumulate charges through charge transfer caused by contact and separation under frictional force.The accumulated charges can be managed by the frictional force and can be converted into electrical energy.Figure 12c exhibits the triboelectricity generation due to the contact of two TENGs via folding.The open-circuit voltage derived from the contact/separation of two 9cm 2 TENGs was approximately 1.1 kV.To evaluate the TENG device's reliability and durability, an Arduino-based pressure generator was applied to the TENG device at a frequency of 1 Hz.During 3000 contact/separation cycles with 2 N, the TENG device exhibited excellent consistency, as shown in Figure S22 (Supporting Information).Figure 12d presents the relationship between voltage and current measured as the load resistance increases for the characterization of the CPI-derived LIG TENG device.The short-circuit current and open-circuit voltage showed an inverse relationship according to Ohm's law, and the TENG device exhibited a peak power output of 411.4 mW m −2 at 40 MΩ.This characterization is vital for determining the efficiency of the TENG device, which can be employed for various energy harvesting applications.The results hint at the potential of LIG-based TENG devices as an encouraging alternative to conventional energy harvesting methods.The output voltage and peak power output of the CPI-based LIG TENG electrode demonstrated superior performance compared to that of the pure yellow polyimidebased LIG TENG electrode, likely attributed to the comparatively larger surface area of the CPI-based LIG (see Table 3).In Figure S23 (Supporting Information), the TENG device is linked to an electronic device in the form of a breadboard with serially connected LED lamps.Our TENG device was able to supply power to an electronic device clearly connected in series with rectifying diodes and LEDs.This illustrates the potential of our TENG device as a power supply for electronic devices.

Conclusion
To summarize, this research illuminates a novel methodology for the fabrication of laser-induced graphene (LIG) from colorless polyimide (CPI) films, which could significantly enrich the fields of materials science and electronics.The strategic incorporation of Fluorine atoms into the CPI film results in a porous structure that bolsters the efficacy of both pressure sensors and energy harvesting devices.The CPI-LIG-based pressure sensor exhibits extraordinary sensitivity, reaching a maximum of 60.340 kPa −1 within a low-pressure span of 1.0-1.5 kPa, in addition to swift response and recovery times of 27 and 36 ms, respectively.This sensor's capability to precisely detect muscle movements and pulse rate emphasizes its promising applicability in the monitoring of human physiological parameters.Furthermore, the LIG derived from CPI is employed to fabricate a dielectric-to-dielectric (D-D) triboelectric nanogenerator (TENG), demonstrating a significant maximum power output of 411.4 mW m −2 at a load resistance of 40 MΩ.Importantly, the integration of Fluorine into the CPI film improves the porosity, and subsequently, the performance of the resulting devices.This investigation not only highlights the immense potential of LIG derived from CPI in a variety of applications but also offers crucial insights into the manipulation and comprehension of laser-induced graphene.The conclusions provide a solid foundation for future endeavors into inventive strategies for energy harvesting, wearable electronics, and human health monitoring, effectively spotlighting the potential contributions of this adaptable material to scientific and technological advancements.

Experimental Section
Direct Laser-Induced Graphene Fabrication inside the Colorless olyimide Film: The experimental setup employed in this study is illustrated in Figure 1.To fabricate a 3D porous network of laser-induced graphene (LIG), a Yb:KGW laser system was utilized.This system delivered highintensity femtosecond pulses with a pulse duration of 170 fs, a pulse repetition rate of 10 kHz, and a wavelength of 515 nm to the sample.A quarterwave plate circularly polarized the laser beam, which was then focused by a Mitutoyo M Plan Apo NIR 10X objective lens with a 20 mm focal length.The stage and objective lens ensured that the beam was tightly focused within the film, and the beam was delivered to the sample using a programmable high-speed motorized stage (Thorlabs MLS203-1) that moves at a speed of 40 mm s −1 according to the programmed pattern.The incident beam had a diameter of 5.7 μm, and the laser irradiation process and fabricated LIG is shown in Video S1 (Supporting Information), which is provided in the supplementary material.The colorless polyimide (CPI)derived LIG was fabricated using three different laser fluences, of 6.65, 13.31, and 19.96 J cm −2 , respectively.The fabricated CPI-derived LIG can be found in the Supporting Information (see Figure S24, Supporting Information).
Characterization of the LIG from CPI Film: The CPI-derived LIG morphology was analyzed by field emission electron scanning microscope (JEOL-JSM-7800F, Tokyo, Japan) and energy dispersive X-ray spectroscope (JEOL-JSM-7800F, Tokyo, Japan) was employed for elemental analysis.Micro Raman spectrometer (Renishaw inVia Reflex, Wottonunder-Edge, Gloucestershire, United Kingdom) with a 532 nm excitation laser system was used to obtain Raman spectra.XPS spectra analysis was carried out using Nexsa Thermo Fisher (Waltham, Massachusetts, USA) to investigate the composition and chemical bond state of the LIG patterns.X-ray diffraction analysis was performed on D8 advance (Bruker, Billerica, Massachusetts, United States) with Cu Ka radiation ( = 1.54 Å) to determine the crystalline characteristics.The surface area and pore size of the samples were assessed utilizing Brunauer-Emmett-Teller (BET) analysis, performed with an Autosorb-iQ instrument (Quantachrome Instruments, Boynton Beach, Florida, USA).
Simulation Setup: Ten distinct CPI monomers were prepared following the approaches outlined below to suppress the formation of the charge transfer complex (CTC), which played a crucial role in the yellow coloration of conventional polyimides (Figure S2, Supporting Information) [45,46,47,48,49,50,51,52] : 1) employing weak electron acceptor anhydrides and weak electron donor diamines to inhibit the transfer of  electrons; 2) incorporating bulky substituents, asymmetric or non-planar structures into the main chain of the film; and 3) introducing an aliphatic anhydride or an aliphatic diamine to prevent the formation of a resonance structure of  electrons.Initially, each system replicated the monomer unit cell, and an NPT simulation was performed.This thermalization process encompassed the adjustment of the simulation box size in accordance with the optimized density, followed by conducting a series of annealing cycles (heating and cooling) with NPT ensemble for period of 40 ps.The results for Kapton aligned closely with the experimental values, while the CPIs demonstrated a strong resemblance to previously reported outcomes in the literature. [45,46,47,48,49,50,51,52]he graphitization of colorless polyimide (CPI) films was investigated using ReaxFF MD in this study.The simulation was performed using the ReaxFF reactive force field implemented in LAMMPS software (Ref).The general energy equation used in the ReaxFF force field method is given by Equation ( 6): where E bond represents bond energies, E lp represents lone pair energies, E val represents valence angle energies, E over and E under are the energy required to correct over-and undercoordination of atoms, respectively, E coulomb is related to nonbonded Coulomb energy, and E VdWaals deals with van der Waals interactions.The bond order-dependent terms are dependent on the local environment of the atoms and subject interactions related to the connection of atoms, such as torsion and angle.The last two terms in Equation ( 6) include nonbonding interactions (long-range dispersion and short-range Pauli repulsion) and eliminate excessive long-range nonbonding interactions.The atom charges and charge distribution were calculated using the charge equilibrium method.The force field parameters used in each energy term were obtained from previous studies by Kim et al. and Rahaman et al. [37,38] Each considered system was subsequently simulated for a total duration of 1.5 ns, employing 30 million steps with a time step of 0.05 fs.The simulation incorporated the NVT ensemble, periodic boundary conditions, and a bond length cutoff of 0.3.The Nosé-Hoover thermostat was utilized with a temperature coupling constant of 500 fs.Evaluation of the CPI-Derived Pressure Sensor: In order to prepare the LIG-based sensors, the protective film of the CPI film composite was removed, and flexible conductive silver paste for screen printing from Sigma-Aldrich was applied to both ends of the pattern.To assess the durability of the sensors, a series of repeated pressure-loading tests at 1, 3, and 5 kPa with a frequency of 1 Hz were measured through a Keithley Tektronix multimeter (Keithley-2450).To evaluate the sensor's pressure measurement performance, an Arduino-based motorized stamping machine was used.Furthermore, a study was conducted on human motion and pulse monitoring by attaching the LIG-based sensors to the wrist, neck, and fingers of a 30-year-old subject, and measured the results using a multimeter.
In Situ Evaluation of CPI-Derived LIG Triboelectric Nanogenerator: Two 3 × 3 cm 2 size triboelectric active areas were fabricated on a single CPI film substrate on both sides.Prior to laser irradiation, the protection film was removed from the CPI film substrate.For consistent triboeletric nanogenerator (TENG) device fabrication, the top LIG electrode part was fixed to a glass plate.A Polyurethane Protective Tape from 3 M was used as a cover in place of the protective film.The sandwich-type TENG device consisting of CPI film/LIG electrode/Polyurethane cover has a flexible nature, resulting in contact-induced triboelectricity generation during folding, which was precisely controlled using an Arduino-based pressure generator.The output voltage was measured using a Tektronix oscilloscope DPO 5204B and Keithley-2450 multimeter.The load resistance ranged from 100 to 300 M depending on the output.The LIG-based TENG device was evaluated in combination with a ceramic resistor connected in series on a Breadboard and a bridge rectifier (2 A, 1000 V) and an LED connected in series.
Ethics: In the study titled "Fabrication of Laser-Induced Graphene from Colorless Polyimide Film and its Application in Pressure Sensors and Triboelectric Nanogenerators", experimental procedures were carried out using wearable sensors.As a doctoral student at the University of Tokyo and a research fellow of the Japan Society for the Promotion of Science (JSPS), the researcher is well-versed and compliant with the ethical guidelines related to human experimentation set forth by these institutions and the national guidelines of Japan.
In the experiments, there was no involvement of invasive procedures, sensitive personal data collection, or any situation that could potentially harm or distress the participants.The wearable sensors used were noninvasive and were tested under controlled conditions, thereby ensuring the safety and well-being of the participants.As such, the study did not necessitate approval from a national or institutional ethics board/committee.The ethical guidelines and standards were strictly adhered to throughout the research.Informed and written consent was obtained from all participating subjects before they took part in the experiments.

Figure 1 .
Figure 1.Schematic of the fabrication process of the LIG electrode inside the CPI film composite.

Figure 2 .
Figure 2. a) Laser-induced graphene fabricated within the CPI film composite via laser irradiation.Surface morphology of the CPI film with laser fluences of b) 6.65, c) 13.31, and d) 19.96 J cm −2 , respectively.Surface morphology of the protection film with laser fluences of e) 6.65, f) 13.31, and g) 19.96 J cm −2 , respectively.

Figure 3 .
Figure 3. a) Raman spectra of CPI-derived LIGs.The green box (upper box) corresponds to LIGs fabricated from the protection film, and the blue box (lower box) corresponds to LIGs generated from the CPI film.b) XRD analysis results of the CPI-derived LIGs.c) XPS characteristics of the CPI-derived LIGs.d) C1s peak of XPS analysis for CPI-derived LIGs.

Figure 4 .
Figure 4. 2D chemical structure drawings of a monomer group (CPI1-5) that does not contain a fluorine atom in its chemical structure.

Figure 5 .
Figure 5. 2D drawings of a monomer group (CPI1-5) that contains a fluorine atom in its chemical structure.

Figure 6 .
Figure 6.ReaxFF simulation process.a) Molecular structure of CPI1 monomer.b) Optimized simulation box obtained by NPT simulation.c) Structure at the end of NVT simulation at ≈3000 K in inert conditions.d) LIG structure with ordered carbon rings and e) reproduced TEM image of LIG from Lin et al. e) Reproduced with permission under the terms of the Creative Commons CC BY license.[25] Copyright 2014, the Authors.Published by Springer Nature.

Figure 7 .
Figure 7. Structures of CPIs and heat-treated structures during the ≈1.5 ns ReaxFF simulation at ≈3000 K. Structural change over simulation time of CPI substrates with molecular structure of a) C/H/O/N, b) C/H/O/N/Cl, and c) C/H/O/N/F atoms.Snapshots of the simulation were recorded from two different angles.

Figures
Figures10a,bdepict the pressure sensitivity of the LIG sensor sourced from CPI. Sensitivity is categorized as S1 for pressure below 1.0 kPa, S2 for pressure from 1.0-1.5 kPa, S3 for pressure from 1.5-5.0kPa, and S4 for pressure exceeding 5 kPa.The sensor's unique structural aspects result in markedly high sensitivity in the lower pressure bracket of 0-1.5 kPa (S1 = 5.498 kPa −1 , S2 = 60.340kPa −1 ).In the surveyed pressure scope of up to 1.5 kPa, the collective sensitivity, known as S avg , which combines both S1 and S2 sensitivities, was found to be S avg = 36.894kPa −1 .This shows the device's proficiency in accurately reacting to low-pressure stimuli.However, with an increase in pressure, the sensitivity progressively dwindles owing to the stabilization of the porous microstructure, and the resistance change rate attains a stable condition (S3 = 8.154 and S4 = 0.547 kPa −1 ).In contrast to formerly reported pressure sensors reliant on metal electrodes and the graphene family, our LIG pressure sensor derived from CPI displays heightened sensitivity in the low-pressure range of ≈1.5 kPa, as indicated in Table2.To confirm the sensor's

Figure 8 .
Figure 8. Carbon ring changes and gas generation during ReaxFF simulation in the CPI1, CPI5, and CPI10 structures in this study.Changes in the number of a) five-membered, b) normalized six-membered, and c) seven-membered rings during ≈1.5 ns simulation.Changes in the normalized number of d) H 2 , e) CO, and f) H 2 O molecules during the simulation for 1.5 nanoseconds.

Figure 9 .
Figure 9. Illustration of the chemical structural changes in a F containing CPI at three time points: a) 0.2, b) 0.8, and c) 1.5 ns.d) Depiction of the evolving surface area trends for CPI groups with F atoms (red line) and CPI groups without F atoms (black line).

Figure 10 .
Figure 10.Performance of the CPI-derived pressure sensor.a) Pressure-dependent Rate of Electrical Resistance Change of CPI-Derived Sensor, b) Sensitivity Analysis of CPI-Derived Sensor in Small Pressure Range (0-5 kPa), c) Real-Time Response of CPI-Derived Sensor under Pressure Loading and Release, d) Repeatability Investigation of CPI-Derived Pressure Sensor at Different Pressure Intervals (1, 3, and 5 kPa), and e) Load/Release Cycle Testing of CPI-Derived Sensor at 5 kPa for a Short Time Domain.

Figure 11 .
Figure 11.a) CPI-derived LIG sensor for respiratory monitoring attached on chest.b) Pulse curve of test subject.c) Response curve of sensor detecting pulse in real time.d) Sensor attached to the wrist for pulse detection (inset of (c)).e) Single pulse detected on the wrist (inset of (c)).f) Schematic image of shoes on squat machine.g) Real-time image of squat machine performance (inset of (e)).h) Shoe image and sensor attachment areas during squatting represented in 3D, blue arrow: heel sensor, red arrow: toe sensor.Response curves of i) left foot and j) right foot sensors detecting pressure during squatting in real time, blue curve: heel sensor, red curve: toe sensor.Sensor attached during rock-paper-scissors game.Response curves of sensor detecting l) paper-scissors and m) paper-rock gestures.

Figure 12 .
Figure 12.Characterization of CPI-derived LIG TENG device.a) Schematic of fabrication process of foldable contact/separate-type TENG device.b) Schematic image of metal-free D-D TENG device operation principle.c) Output voltage of the TENG device upon finger-touch.d) Calculated power density of the LIG-TENG device.

Table 1 .
Optimized system specifications for various CPI monomers.

Table 2
encapsulates the results, which highlight the distinct structural features of this sensing electrode, giving it an edge over formerly discussed pressure sensors based on metal electrodes and graphene clusters regarding sensitivity.

Table 2 .
Comparison of the sensitivity between this work and relevant pressure sensors reported in recent studies.

Table 3 .
Comparison of the performance between this work and relevant TENG electrodes reported in recent studies.