1 Introduction

Recently, emerging studies based on wearable pressure sensors have arouse broad attention due to their potential applications in intelligent robotics, flexible touch screens, smart electronic skin, physiological signal monitoring, and human–machine interfaces [1,2,3,4,5]. However, there are still some challenges in the preparation of integrated pressure sensors with high sensitivity, stretchability and stability [6,7,8,9]. Therefore, new types of sensors with good electrical conductivity and high stretchability should be developed.

Conductive elastomer composite has been used to prepare wearable sensors, and many efforts have been devoted to fabricating elastomeric materials with high durability, high sensitivity, lager strain range, and low detection limit [10]. Generally, the conductive component of elastomeric composite not only needs to maintain their structural connections during large deformations to achieve wide sensing range but also to be able to generate significant signal changes under subtle mechanical stimuli to obtain high sensitivity [11, 12]. Various materials, including conducting polymers [13], carbon black [14], carbon nanotubes [15], graphene [16], silver [17], PPy [18], and their composites have been used to construct strain-sensing composite [19]. Although great progress has been made, the sensitivity and working range of the above materials are not well balanced.

Two-dimensional (2D) materials have exhibited great potential in sensing applications because of their unique intrinsic properties, such as high specific surface area, easy chip fabrication, room-temperature operation, and enhanced sensing performance from bulk to atomically thin layers [20]. Titanium carbides, also known as MXene, are a new member of conducting 2D materials that exhibits high electrical conductivity and efficient photothermal conversion behavior [21,22,23], providing many possibilities in different applications including light-responsive applications, gas sensors, and electromagnetic interference shielding [24,25,26,27,28]. In addition, MXene can be easily fabricated through top-down selective aqueous acid etching [29], which creates chemically active surface and abundant termination groups including –OH, –O, and/or –F. Most importantly, the MXene prepared by the above method also demonstrates high electrical conductivity, which is different from other 2D materials, such as graphene that requires high temperature or reductants to recover its electrical conductivity after solution process [30, 31]. Last but not the least, the aqueous dispersion of Ti3C2Tx can be processed into fibers [32], films [33] and aerogels [34] by using a solution-based strategy [35, 36]. Jiang et al. used the mixture of MXene and graphene oxide dispersion to produce elastic MXene/graphene aerogels through complex several-stage reduction, freezing templating, and thermal annealing procedures [34]. Although the piezoresistive sensor of MXene/graphene composite aerogels exhibits high sensitivity (0.28 kPa−1) and wide detection range (up to 66.98 kPa), the tedious time-consuming procedures prevents its flexible manufacturing. Natural rubber foam possesses low density, excellent flexibility, superelasticity with high recovery rate, and extraordinary reversible compressibility. Those characters make it a good candidate for elastic conductive materials substrate. Zhang et al. and Wang et al. used the natural rubber and graphene to produce elastic strain sensor [37, 38]. Moreover, electrospun fiber mats with high elasticity of NR materials and hydrogel based on NR can be potentially used in strain sensor areas, which shown good mechanical properties [39, 40].

In order to develop a simple and economic method to fabricate hybrid foam based on natural rubber and MXene, the MXene was assembled onto the surface and side of NR foam by dip-coating, named as MXene/NR composite. For comparison, the MXene dispersion was directly sprayed on the surface of NR foam to create a conductive layer, denoted as MXene/NR-1 composite. The results displayed the MXene nanosheets had good adhesion with NR foam, endowing MXene/NR composite highly sensitive response to subtle mechanical stimuli. In addition, the flexible sensor based on MXene/NR and MXene/NR-1 composite exhibited a gage factor that reached 14 and 0.8 in the strain range of 0–5% and 0–10%, respectively. MXene/NR and MXene/NR-1 composite allowed limit of detection 435 Pa and 690,000 Pa, and a strain range of 0–80%. It could be used to detect body motion, such as finger pressing and step monitoring.

2 Materials and methods

2.1 Materials

MAX (Ti3AlC2) was obtained from 11 technology Co., Ltd (Changchun, China). All chemicals were analytical grade and used as received without further purification.

Hydrochloric acid and lithium fluoride (LiF) were obtained from Guangzhou Chemical Reagent Co., Ltd. (Guangdong, China), and copper conductive adhesive was purchased from Shenzhen Xingwang tape Co., Ltd. (Guangdong, China). Natural rubber latex was purchased from Guangdong Maoming Agricultural Reclamation Group Co., Ltd. (Guangdong, China).

2.2 Synthesis of Ti3C2Tx MXene

The MXene was synthesized by a wet-etching process as reported in our previous literature [33]. Briefly, 0.5 g of Ti3AlC2 powder was slowly added to the etchant (0.8 g lithium fluoride and 10 mL of 9 M hydochloric acid) and reaction at room temperature for 30 h. The acidic dispersion was washed until the pH of the suspension reached 6. The mixture was probe sonicated for 0.5 h to increase the yield of single-layer MXene. The dispersion containing single-layer Ti3C2Tx was then concentrated at 10,000 rpm (Centrifuge-TG1650-WS, Bioridge, China) for 0.5 h to achieve high-concentration MXene dispersion (60 mg mL−1) for future use.

2.3 Preparation of NR, MXene/NR, and MXene/NR-1 composite

NR foams were fabricated by previous literature [40]. As a brief description, there are compounding and foaming. NR latex was stirred for 2 min. Then all the other ingredients were slowly added into the latex and performed for 5 min. The compounds were dried at 70 °C for 6 h to produce the NR foams.

The MXene/NR composite was fabricated by dip-coating method. First, the NR foam was cut into a fixed size (10 × 9 × 6 mm) and then washed with deionized water for several times. After dried in the oven at 80 °C for 1 h, the NR foam was immersed in Ti3C2Tx MXene dispersion (60 mg mL−1) and squeezed 3–5 times to have MXene infiltrated into the foam. The NR foam coated with MXene was dried at room temperature for 12 h. For achieving more loadings of MXene, the above procedure was respectively repeated for 1, 3, 5, 6, and 7 times. The MXene/NR-1 composite was prepared via spraying 3 mL MXene dispersion (60 mg mL−1) directly on the surface of NR foam (18 × 10 × 4 mm)to create a conductive layer. The NR foam sprayed with MXene was dried at room temperature for 12 h, and it was denoted as MXene/NR-1 composite.

2.4 Characterization

The morphologies of the MXene, NR foam, MXene/NR, and MXene/NR-1 composite were observed by scanning electron microscope (SEM) equipped with EDAX (S-4800, Hitachi, Japan), and atomic force microscopy (AFM, Bruker AXS, USA). MXene, NR foam and MXene/NR composite were characterized by a Fourier transform infrared (FTIR) spectrometer (Tensor27, Bruker, Germany). The further texts of MXene, NR foam and MXene/NR composite were taken by an X-ray photoelectron spectroscope (XPS) (Thermo Scientific k-Alpha Nexsa, Thermo Fisher, USA) and an X-ray diffraction (XRD) analyzer (D8-Advanced, Bruker, Germany). The resistances and piezoresistive performance of the MXene/NR composite were tested by a digital multimeter (Keysight Truevolt 34465A, Keysight Technologies, USA), which was combined with an electronic tensile testing machine (Dongguan yibo Instruments Co., Ltd., Guangdong, China). The application tests of the MXene/NR composite were carried out using the digital multimeter (Keysight Truevolt 34465A, Keysight Technologies, USA) as well.

3 Results and discussion

3.1 Preparation of MXene, NR foam, MXene/NR and MXene/NR-1 composite

The preparation process of Ti3C2Tx by selective etching is illustrated in Fig. 1a.

Fig. 1
figure 1

a The preparation process of Ti3C2Tx by selective etching. b SEM image of Ti3AlC2 and c MXene after exfoliation. d Photo of MXene aqueous solution. e AFM image of MXene sheet. f XRD results for Ti3AlC2 and MXene

Full size image

Ti3AlC2 exhibited a blocky structure as shown in Fig. 1b. After etching the Al layers, exfoliation yielded single MXene sheets (Fig. 1c). MXene aqueous solution was atrovirens as shown in Fig. 1d. AFM of MXene nanosheets showed its thickness about 2 nm, and there was a little titanium dioxide at the edge (Fig. 1e). Figure 1f shows XRD patterns of MAX and MXene. The results were consistent with previous literature [7]. In addition, the (002) peak of MXene moved to a lower angle, because –F or –OH functional groups led to the lattice expansion [33]. The dispersion of MXene flakes exhibited good stability at high concentration of 60 mg mL−1.

Figure 2a shows the fabrication process of MXene/NR composite, in which NR foam was immersed into MXene dispersion and the excessive MXene was squeezed out to form uniform coating layers and then was dried at room temperature for 12 h. The coating and drying steps were repeated several times to achieve different thickness with MXene loading. As shown in Fig. 2b–c, the NR foam had a polyporous structure with 0.5–1.0 mm diameter and large inner surface, which provide large area for coating MXene. The uniform layers of MXene can be formed on the NR substrates by coating as shown in Fig. 2d–e. The color of the NR turned from white to black after decoration with MXene sheets, and their network structures were well preserved. To evaluate the strength of the interaction between MXene and NR foam, the MXene/NR composite was immersed into water for 1–3 h. The shape of the composite and the color of the water had not changed, suggesting good adhesion of MXene on NR surface (Fig. 2f–h). Additionally, the MXene/NR composite could be cut into diverse shapes (Fig. 2i–k).

Fig. 2
figure 2

a The scheme of the preparation process of MXene/NR composite. b–c SEM image of NR foam and d–e MXene/NR composite. f–h Photos of MXene/NR composite immersed in water for different time spans. i–k Photos of MXene/NR composite cut into different shapes

Full size image

To compare the loading and dispersion of the MXene in MXene/NR and MXene/NR-1 composite, the photograph, SEM image, and EDAX spectra were measured and are shown in Fig. 3a–c and d–f, respectively. It can be seen that the MXene was assembled onto the surface and inside of the NR foam, and it dispersed uniformly in MXene/NR composite. In comparison with MXene/NR composite, we can find obvious yellow when the MXene/NR-1 composite was cut, suggesting the MXene was assembled onto the surface of NR foam. In addition, the SEM image and EDAX spectrum of MXene/NR-1 composite also demonstrated that the MXene can only be formed on the surface of NR substrate. Therefore, the MXene/NR composite can achieve more loadings of MXene than MXene/NR-1 composite, which can contribute to constructing 3D conductive network.

Fig. 3
figure 3

Photo, SEM images, and EDAX spectra of MXene/NR composite (a–c) and MXene/NR-1 composite (d–f)

Full size image

3.2 The interaction between MXene nanosheets and natural rubber

To study the changes in the elemental composition of MXene and MXene/NR composite, the XPS diagrams of MXene and MXene/NR composite are shown in Fig. 4. The peak corresponding to the Ti 2p core level was fitted to four doublets of Ti 2p3/2 and Ti2p1/2, namely Ti-C 2p3/2, Ti-C 2p1/2, Ti(II) 2p3/2, Ti(II) 2p1/2, Ti(III) 2p3/2, Ti(III) 2p1/2 Ti(IV) 2p3/2, and Ti(IV) 2p1/2 located at ~ 454.7, 460.6, 455.8, 461.3, 457.0, 463.0, 459.0, and 465.0 eV, respectively (Fig. 4a and d) [27]. The C 1s core was fitted to four peaks of C–Ti (281.7 eV), C–Ti–O (282.2 eV), C–C (284.9 eV), and C–O (286.0 eV) (Fig. 4b). The O 1s core was assigned to C–Ti–OH 1s, C–Ti–Ox, and O–Ti–O 1s located at ~ 532.6, 529.5, and 530.3 eV, respectively (Fig. 4c and e). Fitted peaks of the Ti 2p, C 1s, and O 1s core levels were all comparable to the previous literature [32], MXene was very similar in the Ti 2p region. When the NR was coated with MXene, the C–Ti–Ox of MXene shifted from 529.5 to 531.0 eV (Fig. 4c), the result also revealed when the MXene/NR composite soaked in water for 1–3 h had no obvious color change, indicating strong interface interactions between the MXene and NR foam.

Fig. 4
figure 4

a–c Ti 2p, C 1s, and O 1s XPS spectra of Ti3C2Tx film prepared from MXene. d–e Ti 2p, O 1s XPS spectra of MXene/NR composite. f FTIR spectrum of NR foam, MXene, and MXene/NR composite

Full size image

To further study the interaction between functional groups of MXene and NR, the FTIR was recorded (Fig. 4f). In the spectrum of the NR, the asymmetric stretching vibration of CH3 was at 2939 cm−1, while the symmetric stretching vibration of CH2 was at 2859 cm−1. The peak at 1433 cm−1 was the CH2 bending of NR. C=C– out of plane bending vibration (cis-1,4 addition) was at 837 cm−1, in-plane bending of N–H and C–N stretching from protein were at 1539 cm−1 [37]. The stretching vibrations at 3447 cm−1 and 1637 cm−1 were from O–H and C=O of MXene, and the stretching vibrations at 1037 cm−1 were assigned to C–O [11]. The stretching vibrations at 3417 cm−1 and 1650 cm−1 were from O–H and C=O of MXene/NR, and C=C – out of plane bending vibration (cis-1,4 addition) was at 837 cm−1. However, the C=O peak in the spectrum of the MXene/NR was obviously shifted to 1650 cm−1 from 1637 cm−1of MXene, implying strong interactions between the NR and MXene sheets.

3.3 The influence of coating times on the conductivity of the MXene/NR composite

To achieve MXene/NR composite with good stability and high conductivity, the impact of coating times was studied. The resistance of the MXene/NR composite was about 500 KΩ after the first coating (Fig. 5a) and the resistance dropped 90% after coating 6 times, and then the resistance went down less and less as the number increased. In addition, the maximum resistances and the real-time resistances change (RM–R) of the MXene/NR composite coated by different times were compared to investigate the sensitivity of the MXene/NR composite. It can be seen that the initial resistance was relatively small and the RM–R changed greatly by coating 6 times (Fig. 5b). Therefore, the MXene/NR composite coated by 6 times was chosen to fabricate strain sensor and was used to detect human motion signals.

Fig. 5
figure 5

a The original resistances and b Resistances change of the MXene/NR composite with different coating times under 0–80% strain by compression

Full size image

3.4 Sensing performance of MXene/NR and MXene/NR-1 composite strain sensor

The compressive stress–strain curves of NR, MXene/NR and MXene/NR-1 composite are shown in Fig. 6a, which show that stress increased with the increases of strain, and the strain of 80% could be achieved. The signals of the MXene/NR strain sensor under strain were shown by the variable of ΔR/RM, where RM and ΔR denoted the maximum resistance and RM subtract real-time resistance, respectively. Figure 6b, c, and g plots the sensing performances of MXene/NR and MXene/NR-1, ΔR/RM value of the MXene/NR was much larger than MXene/NR-1, and it is shown that the MXene/NR is suitable for strain sensors. The representative parameter of sensitivity is the gage factor (GF), which is calculated by below equation [37].

Fig. 6
figure 6

a The compressive stress–strain curves. b–c Signals of the MXene/NR composite under strains of 5–20% and 40–80%. d Signals of the MXene/NR under 20% strain at rates of 2, 5 and 10 mm/min. e Durability experiment of the MXene/NR during 100 cycles under 5% and 10% strain. g Responses of the MXene/NR-1 composite strain sensor under subtle strains (5–20%). h Signals of the MXene/NR-1 under 20% strain at rates of 2, 5, and 10 mm/min. i Durability experiment of the MXene/NR-1 during 100 cycles under 5% and 10% strain

Full size image
$${\text{GF}}=\frac{\Delta R}{{R}_{M}\times \Delta \varepsilon },$$

where RM is the maximum resistance, ΔR equal to RM subtract real-time resistance (R), and Δε is the applied strain. The gage factor of MXene/NR and MXene/NR-1 reached 14 and 0.8 in the strain range of 0–5% and 0–10%, respectively, showing MXene/NR a better compromise between sensitivity and compression performance. In complex detection environments, strain sensors must also maintain stable response behavior at different stretching rates. The stretching–releasing cycles was determined under strain of 20% at stretch rates of 2, 5, and 10 mm/min. The response signals increased with the rate increases. ΔR/RM value of the MXene/NR-1 changed smaller than the MXene/NR as different stretch rates, as shown in Fig. 6d and h. Moreover, the limit detection of the MXene/NR and MXene/NR-1 were 435 Pa and 690,000 Pa. 100 stretching–releasing cycles under 5% and 10% strain were applied to show the stability (Fig. 6e, f and i), respectively. Compared to the MXene/NR-1, the MXene/NR sensor demonstrated good stability. In other words, the MXene/NR composite was more suitable for fabricating large strain range sensor with high sensitivity than MXene/NR-1.

3.5 Detection of human motion signals by the MXene/NR strain sensor

The schematic diagram of the MXene/NR strain sensor is shown in Fig. 7a. The conductive paths increase and the resistance become smaller with strain gradually increasing, the resistance change of the MXene/NR strain sensor was very large when strain increase drastically. The conductive channel recovered to the original state after releasing the strain. The MXene/NR strain sensor was suitable for recognition of human movements, as shown in the signals of the finger pressing (Fig. 7b) and step monitoring and the partial enlargement map(Put it in the balls of feet) (Fig. 7c–d). The response time of the MXene/NR strain sensor was 1.00 s. Through monitoring the ΔR/RM, the finger pressing and step monitoring were accurately tracked with good reproducibility. Furthermore, it can be seen that the greater the load, the greater the value of ΔR/RM.

Fig. 7
figure 7

a The schematic diagram of piezoresitance in the MXene/NR strain sensor. b Application of the MXene/NR strain sensor for finger pressing. c–d Step monitoring from an adult and the partial enlargement signal

Full size image

4 Conclusion

In this paper, we developed a simple, rapid, and cost-effective method to fabricate strain sensors. The MXene/NR strain sensors shown a good sensitivity (GF = 14), a strain of 0–80%, a low limit of detection (435 Pa), and good stability under 5% and 10% strain. Furthermore, the response time of MXene/NR strain sensor was 1.00 s, and it could be used to detect motions, such as finger pressing and step monitoring, indicating it is a good candidate for fabricating wearable electronics.