Highly Sensitive Flexible Strain Sensor Based on Laser-Induced Graphene for Wearable Applications

2.1 Fabrication of Flexible LIG Strain Sensor

The overall fabrication procedure of the flexible LIG/PDMS strain sensor is illustrated in Fig. 1. The sensor consists of a porous LIG sensing layer embedded within a flexible PDMS substrate, as shown in Fig. 1(a). A commercial 25-μm-thick PI film was first laminated onto a cleaned 4-inch silicon (Si) wafer using a vacuum laminator to provide a flat and stable surface for laser processing. Conductive LIG patterns were directly fabricated using a commercial laser engraving system equipped with a continuous-wave (CW) blue diode laser (450 nm, maximum output power of 10 W). The laser processing parameters were set to a laser power of 3.5 W, a scanning speed of 200 mm s^-1, and a hatch spacing of 50 μm. The detailed fabrication procedure, including PI lamination, laser engraving, PDMS casting, and peel-off transfer, is presented in Fig. 1(b).

Fig. 1.

(a) Schematic illustration of the flexible LIG/PDMS strain sensor. (b) Step-by-step fabrication process involving PI lamination on a Si wafer, laser engraving, PDMS casting, and peel-off. SEM images of (c) the laser-induced graphene (LIG) surface showing a 3D porous network and (d) the transferred LIG embedded in PDMS. (e) Optical photograph of the fabricated flexible strain sensor.

Following laser irradiation, PDMS (Sylgard 184, Dow Corning, USA) base and curing agent were mixed at a weight ratio of 10:1, degassed under vacuum for approximately 30 min, and cast onto the laser-patterned PI substrate using a doctor blade to form a uniform encapsulation layer. The sample was subsequently cured at 80°C for 2 h. After curing, the PDMS layer was carefully peeled from the PI substrate, thereby transferring and embedding the porous LIG network into the flexible PDMS matrix. The transferred LIG embedded in the PDMS substrate and the completed flexible strain sensor are shown in Fig. 1(d) and (e), respectively. Finally, copper wires were attached to both ends of the transferred LIG electrodes using conductive silver paste to establish reliable electrical connections.

2.2 Morphological and Electrical Characterization

The morphology of the fabricated LIG was characterized using field-emission scanning electron microscopy (FE-SEM) to investigate the three-dimensional porous graphene network generated by laser-induced carbonization. Representative FE-SEM images of the fabricated LIG are presented in Fig. 1(c).

To evaluate the electrical uniformity of the fabricated LIG patterns, a 3 × 3 sensor array was fabricated on a 4-inch PI substrate, as shown in Fig. 2(a). The sheet resistance (R_s) distribution was measured using a non-contact sheet resistance mapping system (EddyCus TF map 2530, SURAGUS GmbH, Germany), which employs a high-frequency eddy-current sensing technique for non-destructive electrical characterization. The corresponding sheet resistance mapping is presented in Fig. 2(b). The sheet resistance statistics were determined using the nine LIG-patterned areas within the 3 × 3 array while excluding the surrounding bare PI regions. The minimum, maximum, average, and mean standard deviation of the measured sheet resistance are summarized in Table 1.

Fig. 2.

(a) Optical image of the 3 × 3 LIG-patterned array fabricated on a 4-inch PI substrate. (b) Sheet resistance mapping profile of the nine LIG-patterned areas within the 3 × 3 array.

Table 1.

Sheet resistance characteristics of the fabricated LIG.

2.3 Electromechanical Characterization

The electromechanical performance of the fabricated LIG/PDMS strain sensor was evaluated using a universal testing machine (EZ-L, Shimadzu Corporation, Kyoto, Japan), while the corresponding electrical resistance was simultaneously monitored using a digital multimeter (USB-4065, National Instruments, Austin, TX, USA). Uniaxial tensile strains ranging from 1% to 5% were sequentially applied at a constant crosshead speed of 1 mm min^-1.

The relative resistance change (ΔR/R₀) was calculated according to Eq. (1), where R₀ and R represent the initial resistance and the measured resistance under the applied strain, respectively.

The GF, which represents the strain sensitivity of the sensor, was calculated according to Eq. (2), where ε is the applied tensile strain.

2.4 Demonstration of Stretchable LIG Interconnect

To evaluate the electromechanical stability of the fabricated LIG as a stretchable conductor, a serpentine-patterned LIG interconnect integrated with a commercial blue light-emitting diode (LED) was fabricated. The electrical functionality of the device was examined under tensile, bending, and twisting deformations by monitoring the LED operation. This demonstration was performed to verify the electrical continuity and mechanical reliability of the LIG interconnect under various mechanical deformation modes.

2.5 Wearable Pulse Monitoring

To evaluate the wearable pulse monitoring capability of the fabricated sensor, the strain sensor was attached over the radial artery of the wrist using a transparent medical dressing film (Tegaderm™, 3M, St. Paul, MN, USA). The resistance variation generated by arterial expansion during the cardiac cycle was continuously recorded under resting and post-exercise conditions. The acquired resistance signals were converted into normalized ΔR/R₀ and processed for pulse waveform analysis.

3.1 Sheet Resistance Distribution of the Wafer-Scale LIG Array

The electrical characteristics of the fabricated LIG were evaluated using the nine patterned areas within a wafer-scale 3 × 3 array. The 3 × 3 array of LIG patterns was successfully fabricated over a 4-inch PI substrate without noticeable macroscopic defects or discontinuities, as shown in Fig. 2(a). The corresponding sheet resistance distribution was quantitatively characterized using a non-contact eddy-current mapping system, as shown in Fig. 2(b). The sheet resistance statistics were obtained exclusively from the nine LIG-patterned areas, excluding the surrounding bare PI regions. As summarized in Table 1, the minimum and maximum sheet resistance values were 47.78 and 126.94 Ω sq^-1, respectively. The average sheet resistance was 70.13 Ω sq^-1, with a mean standard deviation of 13.60 Ω sq^-1. These results indicate that the diode laser process produced LIG patterns with relatively consistent electrical characteristics across the wafer-scale 3 × 3 array. The observed consistency supports the reproducibility of the fabrication process for producing multiple flexible sensing elements on a single substrate.

3.2 Electromechanical Performance of the LIG/PDMS Strain Sensor

The electromechanical characteristics of the transferred LIG/PDMS strain sensor were systematically investigated under tensile deformation. The baseline resistance was continuously monitored for 1800 s without external loading, exhibiting negligible electrical drift and stable signal output, as shown in Fig. 3. This baseline stability indicates that the resistance signal remained relatively insensitive to time-dependent fluctuations under static conditions. The baseline stability also provided a reliable reference for distinguishing the resistance changes induced by subsequent tensile deformation.

Fig. 3.

Baseline electrical stability of the fabricated LIG/PDMS strain sensor measured for 1800 s without external tensile loading.

The dynamic relative resistance responses under repeated tensile strains ranging from 1% to 5% are presented in Fig. 4(a). Distinct and reproducible resistance peaks were observed for each applied strain, indicating stable sensing behavior with good repeatability. The sensing performance was quantitatively evaluated using the ΔR/R₀ and the corresponding GF. The ΔR/R₀ increased monotonically with increasing tensile strain, reaching approximately 17.6 at 5% strain, as shown in Fig. 4(b). The corresponding GF also increased with applied strain and exceeded 350 at 5% strain, demonstrating high sensitivity within a relatively small deformation range, as shown in Fig. 4(c).

Fig. 4.

Electromechanical performance of the LIG/PDMS strain sensor. (a) Dynamic relative resistance responses under repeated tensile strains ranging from 1% to 5%, (b) ΔR/R0 as a function of applied strain, and (c) GF as a function of applied strain.

The high sensitivity is attributed to the conductive porous graphene network embedded within the PDMS matrix. During tensile deformation, separation between adjacent graphitic domains increases both the tunneling distance and contact resistance, resulting in a significant increase in the overall electrical resistance. Upon release of the applied strain, the elastic recovery of the PDMS matrix facilitates partial reconstruction of the conductive pathways, allowing the sensor resistance to recover toward its initial value. However, complete baseline recovery was not achieved after unloading, and a residual baseline shift remained, which may be associated with partial irreversible rearrangement of the porous LIG network. The magnitude of this baseline shift corresponded to approximately 1.13% of the maximum sensor response, indicating a relatively limited influence on the full-scale output. Although a residual baseline offset remained, the resistance peaks at each applied strain were clearly distinguishable and repeatable.

3.3 Stretchable LIG Interconnect Demonstration

Beyond strain sensing, the fabricated LIG structures were further evaluated as stretchable conductive interconnects. A serpentine-patterned LIG electrode integrated with a blue LED was fabricated to visually demonstrate its electromechanical response. In the initial state, the LED exhibited stable illumination, confirming reliable electrical conduction through the LIG interconnect, as shown in Fig. 5(a). When tensile deformation was applied, the LED brightness noticeably decreased because of the increase in electrical resistance caused by stretching of the serpentine LIG pathway, as shown in Fig. 5(b). This behavior clearly demonstrates the strain-dependent electrical characteristics of the fabricated LIG structure.

Fig. 5.

Optical photographs demonstrating the electromechanical durability and switching behavior of the serpentine LIG interconnect integrated with a blue LED under various mechanical deformations. (a) Pristine state with a turned-on LED, (b) stretched state showing the turned-off LED due to the strain-induced resistance increase, (c) bending, and (d) twisting.

The serpentine geometry accommodated the applied tensile deformation through changes in the curvature and spacing of the conductive pathway. The increase in resistance during stretching may be attributed to partial separation and rearrangement of the porous LIG network, which reduced the current supplied to the LED. Therefore, the change in LED illumination provided a direct visual measure of the resistance modulation induced by tensile deformation. These observations indicate that the patterned LIG structure can function as both a mechanically compliant interconnect and a strain-responsive conductive element.

In contrast to its response under tensile deformation, the LED maintained stable illumination without noticeable flickering or electrical interruption during bending and twisting, as shown in Fig. 5(c) and (d), respectively. This result indicates that the serpentine geometry effectively accommodated complex mechanical deformation while preserving electrical continuity. The demonstrated electromechanical stability suggests that the proposed LIG structure can serve not only as a highly sensitive strain-sensing element but also as a reliable stretchable interconnect for flexible electronic systems.

3.4 Wearable Pulse Monitoring

To verify the practical applicability of the developed strain sensor for wearable healthcare, real-time wrist pulse monitoring was performed by attaching the flexible sensor directly over the radial artery, as shown in Fig. 6(a). The sensor was secured using a transparent medical dressing film, while the electro-pads provided stable electrical connections during the measurement. Owing to the mechanical flexibility and conformability of the LIG/PDMS structure, the sensing region maintained close contact with the skin and followed the small surface deformations induced by arterial pulsation.

Fig. 6.

Wearable health monitoring applications. (a) Photograph of the LIG strain sensor attached to a human wrist. Real-time relative resistance response capturing the arterial pulse waves under (b) resting and (c) post-exercise conditions.

Representative pulse waveforms recorded under resting and post-exercise conditions are presented in Fig. 6(b) and (c), respectively. Under the resting condition, periodic pulse signals with relatively stable amplitudes and intervals were clearly detected, yielding an average heart rate of 72 bpm and a normalized signal amplitude of 0.0228 ± 0.0098. The periodic resistance variation indicates that the sensor was sufficiently sensitive to detect the subtle expansion and contraction of the radial artery during each cardiac cycle. Following physical exercise, the pulse interval became noticeably shorter, corresponding to an increased heart rate of 84 bpm, while the normalized pulse amplitude increased to 0.0894 ± 0.0161. Compared with the resting state, the heart rate increased by approximately 16.7%, and the normalized pulse amplitude increased by approximately 3.9 times. These changes reflect the enhanced cardiovascular activity induced by exercise, which produces more frequent and pronounced arterial deformation at the wrist. The clear differences between the resting and post-exercise pulse waveforms demonstrate that the fabricated sensor can distinguish variations in both pulse frequency and signal amplitude under different physiological conditions. The periodic pulse signals were consistently detected under both conditions, confirming that the sensor maintained reliable skin contact and stable signal acquisition during the measurements. These results confirm that the combination of high electromechanical sensitivity, mechanical compliance, and conformal skin contact enables the LIG/PDMS strain sensor to detect weak physiological signals in real time. The ability to distinguish resting and post-exercise pulse characteristics highlights its potential for wearable heart-rate monitoring, exercise-state assessment, and continuous personalized healthcare applications.