In recent years, a diverse range of applications in the field of electronic sensing has emerged thanks to the unique properties of two-dimensional materials (e.g., graphene^[1], MXene^[2], metal oxides^[3], etc.). These applications include pressure sensors, image sensors, biosensors, temperature/humidity sensors, and gas sensors^[4-6]. In particular, researchers are interested in flexible pressure sensors because of their potential use in wearable health monitoring devices^[7-9]， human-machine interfaces^[10-12], and electronic skins^[13-15]. Research has shifted in recent years to improving sensor performance. Especially, electron/ion transport is influenced by the internal microstructure of the sensors (e.g., nanopores and nanochannels), which affects the sensor’s ability to detect environmental changes^[16]. Therefore, it is vital to investigate the inner microstructure of these materials, as well as the internal transport behavior of the sensors. The study of the inner microstructure of these materials and the internal transport behavior of sensors is therefore essential.

Graphene is a two-dimensional carbon substance with exceptional mechanical, electrical, thermal, and optical properties^[17], and is currently a research hotspot. However, it is prone to agglomeration, which results in the destruction of internal nanoscale transport channels and a reduction in electrochemical activity, which limits its application in biosensing^[18]. As a result, it cannot be used in biosensing applications. Graphene exhibits high electrochemical activity as well as biocompatibility. In recent years, a new class of 2D metal carbides and nitrides, the so-called MXene, has attracted extensive attention due to their metallic conductivity, excellent mechanical properties, and large specific surface area. Aerogel is a porous material with nanostructures that contain a large amount of air, which has unique properties in mechanics, acoustics, thermals, optics and so on. This ultralight and compressible aerogel material also has broad application prospects in the field of flexible sensing due to its high porosity, excellent compressibility, and electrical conductivity. The electrostatic interactions between MXene materials, with high electrochemical activity and biocompatibility, and graphene result in the creation of two-dimensional (2D) composites with low self-stacking that have good electrochemical activity and which are biocompatible^{[19, 20]}. The composite material has better conductivity and can effectively improve the sensitivity of the sensor as a sensitive layer of the pressure sensor. The stacking of layered materials can effectively expand the response range of the sensor. Incorporating 2D materials as nanohybrids is being extensively explored to find their synergistic effect on sensing performance^{[21, 22]}.

In this paper we show that freeze-dried hybrid MXene-rGO aerogels with porous three-dimensional (3D) interlocked networks can be used as active layers to fabricate flexible sensors that exhibit high sensing performance. In addition, we demonstrate that the sensitivity of the flexible sensors increased with an increase in the bending angle (0°–90°). Moreover, these devices also possess an excellent level of reproducibility (high repeatability); namely, the sensitivity, response time, and recovery time do not significantly change after a large number of cycles. These outstanding demonstrations confirm the excellent performance of this interlocked material as a flexible and wearable sensor.

Ti₃C₂T_x was synthesized by selective etching of Al atoms from Ti₃AlC₂. Briefly, 4.8 g of lithium fluoride (LiF), 45 mL of hydrochloric acid (HCl) and 15 mL of deionized (DI) water were mixed by stirring. Then, 3 g of Ti₃AlC₂ was slowly added to the solution at 35 °C followed by stirring for 48 h. The as-obtained suspension was washed with deionized water via centrifugation at 3500 rpm until pH ≥ 6. The above MXene nanosheets solution and rGO solution with mass ratio of 1 : 1 are mixed together using deionized water. Subsequently, as-obtained mixed solution based on MXene/rGO was poured into a tube and frozen for 30 min with liquid nitrogen. Finally, the MXene/rGO hybrid were freeze-dried in a lyophilizer for 48 h to fabricate 3D, interlocked MXene/rGO aerogels.

The architecture of interlocked MXene/rGO aerogels was observed by scanning electron microscopy (SEM, Magellan 400). The dynamic pressure response and mechanical properties of an interlocked MXene/rGO aerogels-based flexible pressure sensor were tested using high-resolution micromechanical testing system (Instron E1000). Other pressure-sensitive properties were conducted using electrochemical work station (CHI 760E).

When rGO was added to the MXene nanosheet solution, an MXene/rGO aerogel with a 3D interlocked structure was formed via self-assembly and subsequent freeze-drying. Fig. 1(a) shows the preparation procedure of the interlocked MXene/rGO aerogel. Briefly, MXene nanosheets were created by etching a Ti₃AlC₂ precursor with HCl and LiF, followed by solution drying. A physical mixing procedure occurred between the charged MXene and rGO nanosheets. Upon combining the rGO with the MXene nanosheets, a thicker and more protective rGO layer was generated on the inner side of Ti₃C₂T_x. After freeze-drying, a 3D interlocked MXene/rGO hybrid aerogel was obtained.

The feasibility of freeze-drying for the preparation of the MXene/rGO aerogel was investigated using scanning electron microscopy (SEM). As shown in Fig. 1(b), the MXene and rGO nanosheets were extended and interlinked to form a 3D honeycomb-like network. Pores were created by the volatilization of water trapped between the MXene and rGO sheets after the aerogel was freeze-dried. According to a high-resolution SEM photograph, the pore sizes of these networks are in the range of several micrometers, which is similar to a honeycomb-like network (Fig. 1(c)). Furthermore, the porous MXene/rGO aerogel is such a lightweight material that it can easily be held by a dandelion (Fig. 1(d)).

The elastic performance of the porous MXene/rGO aerogel was evaluated by testing the compressive loading and unloading curves with a maximum strain of 12%. When this strain was set, a much higher compressive stress of approximately 9.2 kPa could be observed. The stable and constant stress-strain curves in the 1st, 2nd, and 3rd cycles further confirmed the recoverability of the as-prepared MXene/rGO aerogel (Fig. 1(e)). The pressure-sensitive mechanism of the MXene/rGO aerogels was mainly related to the interlocking structure of the aerogels (Fig. 1(f)). Unlike planar films, MXene/rGO aerogels exhibit interlocking and porous structures. A porous structure can control the conductivity of aerogel materials. The change in the conductive channel depends on the applied external force. When an external pressure is applied, a small compression deformation of the porous MXene/rGO aerogel increases its internal conductive channel, as well as the current (Fig. 1(g)). During unloading, the aerogel reverts to its original shape and reduces the contact of the internal conductive channels, resulting in a decrease in the current (Fig. 1(g)).

Fig. 2(a) shows the structure of an interlocked MXene/rGO aerogel-based pressure sensor. We started by vacuum evaporating a gold electrode with a thickness of 50 nm onto a piece of polyethylene terephthalate film using a vacuum chamber. Then, using the previously developed interlocked MXene/rGO aerogel as an intermediary sensing layer, we designed a resistive pressure sensor that measures the current change at both ends of the aerogel under different pressures. The sensing performance of the interlocked MXene/RGO aerogel-based pressure sensor was studied using both static and dynamic mechanical pressure tests. The current–voltage curves of the aerogel-based pressure sensor under various static pressures ranging from 0 to 1.1 kPa are shown in Fig. 2(b). The current increases with an increase in the external force, which indicates that the sensitivity of the pressure sensor under static pressure is stable.

The corresponding dynamic sensitivity curves are shown in Fig. 2(c). The sensitivity of our pressure sensor was defined as S = (ΔI/I₀)/ΔP^{[10, 23-25]}, where ΔP denotes the change in the external force. ΔI denotes the relative change in the current and I₀ denotes the initial current. Fig. 2(d) diagrams the sensitivity of the pressure sensor under different pressure. In the pressure range of 0–40 kPa, the sensitivity of the flexible pressure sensor can reach 61.49 kPa^–1, which is much higher than that of conventional pressure sensors in the same range of values. When the pressure exceeds 40 kPa, the increase in sensitivity began to slow down and approaches saturation at 110 kPa with a sensitivity close to 4.22 kPa^–1. According to this study, our aerogel pressure sensors outperformed the other pressure sensors in terms of performance. The porous structure of the material is responsible for the reported enhanced pressure sensitivity levels.

We also compared the sensitivity of the flat aerogel pressure sensor and the MXene/rGO aerogel pressure sensor, and the results are shown in the inset of Fig. 2(d). As a result, the sensitivity of MXene/rGO aerogel-based pressure sensor is higher than flat aerogels-based pressure sensor. Fig. 2(e) shows another key parameter of the flexible device-response/recovery time; that is, from the current–time (I–T) amplification curve of the flexible device based on the interlocked MXene/rGO aerogel, the device has a faster response and recovery time (68 and 40 ms, respectively). These features correspond to the response and recovery time of the human skin (40–50 ms). Rapid response and recovery (without hysteresis) can ensure the timely perception of external forces by flexible devices and increase recognition efficiency. Furthermore, we compare the performance of the pressure sensor with the existing research results, and the results are shown in Table 1^{[24, 26-29]}. Our pressure sensor has great advantages in terms of sensitivity, response range, response time and other performance.

To further study the mechanical stability of the interlocked MXene/rGO aerogel-based pressure sensor, we tested the bending properties of the device at various angles (Fig. 3(a)). The pressure–sensitive response of the device increased with the extension of the bending angle. The change in sensitivity under bending deformation was studied at a bending angle of 90° (Fig. 3(b)). After five cycles, the pressure-sensitive response of the device was maintained at approximately 120. The interlocked aerogels exhibited high sensitivity and good cycling stability. At a bending angle of 90°, the pressure-sensitive response was approximately 120 and recovered to its initial value when released. Moreover, each corresponding cycle shows good response and recovery characteristics (Fig. 3(c)). Fig. 3(d) displays the sensor sensitivity under different bending angles (the bending angle started at θ = 0°, and the gradient increased to 90° and then returned to 0° at a rate of 30°). This indicates the good mechanical stability of the flexible sensor, which retained the same level of pressure-sensitive response throughout the bending angle test.

Owing to its supercompressibility, elasticity, fatigue resistance, bendability, and ultrahigh sensitivity, the interlocked MXene/rGO aerogel-based flexible pressure sensor offers a wide range of applications in wearable sectors. Based on the research results, our sensors accurately identified a variety of stimuli across a wide range of pressures, including finger joint motion and wrist pulses. As shown in Fig. 4(a), the output current increased with the bending angle of the index finger, and the amplitude of the output current increased with the bending degree of the index finger. Here, the index finger has a bend angle of 60°, and the pressure-sensitive response value is approximately 91. Furthermore, the repeatability of the device was demonstrated by bending the index finger at 90° three times in a row, while retaining a current response of approximately 170 (Fig. 4(b)).

Blood pressure readings collected from the wrist could provide vital and useful information about an individual's physical and mental well-being. Examples of such disorders include cardiovascular disease, which is related to the shape of the wrist pulse waveform. Thus, by monitoring the radial artery in the wrist, the sensor may be able to collect information regarding human pulses and other vital signs. As depicted in Fig. 4(c), the comparable pulse frequency was approximately 75 beat/min. Meanwhile, Fig. 4(c) depicts a pulse waveform with three distinct peaks: P₁ represents the percussion wave, P₂ the tidal wave, and P₃ the diastolic wave. Fig. 4(d) depicts the signs and symptoms of various cardiovascular disorders. Because the wrist pulse waveform can be used to predict a variety of cardiovascular diseases, our sensors may have a future in real-time health monitoring. Based on these initial findings, it appears that this flexible pressure sensor could find applications in several scenarios requiring reliable detection over a wide pressure range, such as wearable medical electronics and prosthetics^[30-33].

In summary, the combination of MXene nanosheets with rGO resulted in a compressible and elastic MXene-derived aerogel with excellent mechanical properties and good flexibility. Owing to good 3D interlocked structure of MXene/rGO aerogel, the sensor exhibits a good sensitivity of 61.49 kPa^–1 throughout a broad linear response range of 0–40 kPa, fast response time (68 ms) and cycling stability. In addition, the aerogel-based sensor exhibits a high degree of linearity over a wide range of pressures and temperatures. Consequently, the pressure sensor was able to detect eight subtle, low-, and medium-level pressures, such as those produced by a human heartbeat or finger movement. These properties make aerogels potential candidates for use in pressure/strain sensors and wearable devices. Because of its higher performance and ease of fabrication, the proposed sensor may prove to be a viable option for use in artificial intelligence and smart medical devices in the future.

The authors sincerely acknowledge financial support from the National Natural Science Foundation of China (NSFC Grant No. 61625404, 61888102, 62174152), Young Elite Scientists Sponsorship Program by CAST (2018QNRC001), the Strategic Priority Program of the Chinese Academy of Sciences, Grant No XDA16021100, and the Science and Technology Development Plan of Jilin Province (20210101168JC).