Highly sensitive pressure sensor based on graphene hybrids

1 Introduction

Recently, an increase in demand for flexible, touch sensors, pressure sensors has been gradually increasing for practical applications such as robotics, diagnostics and therapeutics (Tegin and Wikander, 2005; Sokhanvar et al., 2007; Teng et al., 2008). However, sensors based on hybrid composites, nanowire or nanotube assemblies have been attracting the attention greatly due to their pressure sensing characteristics (Lipomi et al., 2011; Ko et al., 2009; Dvir et al., 2011). Gold nanoparticles are the most stable and unique particles compared with other metals in the form of nanoparticles such as Ag and Pd with many applications related to sensors (Yang et al., 2001; Li et al., 2010) and biological systems (Han et al., 2007; Rosi et al., 2006). Moreover, to control the surface properties and interactions nature between the hybrid structures, the sizes and shapes of the nanoparticles could be considered (Qian and Park, 2010; Yin et al., 2011). The ease of synthesis and functionalization in different sizes of AuNPs makes more emanated. AuNPs can be functionalized with different types of molecular ligands including alkanethiolates, enzymes, DNA, proteins, sugars, oligonucleotides, phospholipids and more (Bhattacharya and Srivastava, 2003; Burkett and Mann, 1996; Daniel and Astruc, 2004). In case of sensor applications, functionalized nanoparticles can be synthesized with other materials with various physical and chemical functions, which has an impact on the sensitivity and selectivity of the sensor. The assembly of functionalized AuNPs requires a suitable platform exhibited unique properties (Xu et al., 2008). Graphene has become very popular in recent times, due to its structure and good physicochemical properties such as electrical (Du et al., 2008) and thermal properties (Balandin et al., 2008) together with large surface area (Geim, 2009). Surface functionalisation of graphene with other molecules by covalent and non-covalent interactions has been previously reported (He et al., 2010; Hu et al., 2011; Xu et al., 2009; Kozhemyakina et al., 2010; Lim et al., 2013). The oxygen functional groups, mainly contained carboxylic groups at the edges and epoxy/hydroxyl groups of the basal plane are involved in covalent bond formation (Yang et al., 2009a, 2009b; Liu et al., 2010). Especially, the carboxylic groups present at the edge of the graphene could be functionalized with other materials by 1-ethyl-3-(3-dimethylaminopropyl)- carbodiimide (EDC) (Liu et al., 2008) or thionyl chloride (SOCl₂) (Xu et al., 2009; Yang et al., 2009b). The non-covalent interactions have the weak interactions between graphene and other molecules. For example Li et al.; reported MP-11 non-covalently functionalized on graphene (Li et al., 2014).

An increase in demand for portable sensors has led to the development of sophisticated sensors. Graphene acts as a potential candidate for the development of well-defined structures due to its large surface area, electrical (Du et al., 2008; Suk et al., 2013) and mechanical properties (Lee et al., 2008). Functionalization with other materials could lead to the development of sensors with properties which is not easily obtainable (Shearer et al., 2014). Selectivity and sensitivity can be significantly increased using coupling with a sensing component. In case of graphene hybrid sensor, the sensing components vary from metal oxides, metal NPs and biomolecules (Shearer et al., 2014). Herrmann et al. is first to demonstrate the functionalized gold NP-based sensitive strain gauges for sensor applications. Their study mainly focuses on tunneling model and sensitivity of the functionalized NP films, which depends on various factors, including the size of the nanoparticle, interparticle distance, and the conductance of the binding molecules. The drawbacks of this system are (1) the sensitivity and detection range of weight that is not varied based on suitable applications; (2) the properties of NP-based sensors, mainly depending on the substrate and the linker molecules. Shu Gong et al. reported a wearable and highly sensitive pressure sensor with ultrathin gold nanowires, which was able to show fast response time (<17 ms), high sensitivity (1.14 kPa⁻¹) and high stability (Gong et al., 2014). The flaws in this system could be (1) impregnate of AuNWs onto tissue paper and sandwiched between PDMS instead of using only AuNW and (2) high resistance of 2.5 ± 0.4 MΩ cm⁻¹ which is not suitable for a sensor. To overcome these problems, a new device with higher sensitivity factor and well-organized structure has been designed in this study. It shows that functionalized AuNPs could be sandwiched between two graphene layers, leading to a novel pressure sensor as shown in Fig. 1. The advantage of using graphene hybrid is its sensing mechanism, excellent electronic, and electrochemical properties of this material which helps to improve the sensitivity (Hou et al., 2013). When compared with previous reports, this graphene hybrid device shows better sensitivity, which is mandatory for the surface to be uniform and have good transfer of electrons for checking the sensitivity of a device. In addition, the film fabricated here is free standing which fulfills both the prerequisite mentioned above. This novel device shows good performance with a sensitivity of 5 × 10⁻⁴ kPa⁻¹.

Close

Fig. 1 Pictorial representation of the hybrid film.

Export to PPT

2 Results and discussion

A hybrid film with graphene and functionalized AuNPs was successfully fabricated using vacuum filtration. The film shows good conductivity with a wide range of pressure sensitive responses. The surface morphology of the film was studied by SEM and TEM. Fig. 2a shows the SEM image of layer-by-layer assembly of a graphene hybrid film. As the pressure increases the layers were tightly packed one on top of the other and forms a compact film with a thickness of 572 nm. Fig. 2b shows the layer by layer deposition of functionalized gold NPs sandwiched between graphene layers. The cause for wrinkling of the film is due to the electrostatic interactions between the functionalized AuNPs and the graphene layers and the surface roughness of the film during surface compression, which increases with an increase in number of layers (Cote et al., 2008; Kim et al., 2009).

Close

Fig. 2 SEM images (a) cross-section of Gr hybrid film. (b) Shows side view for Gr hybrid film.

Export to PPT

Fig. 3a shows transmission electron microscope (TEM) image of the shape and size of functionalized AuNPs. The structure could be characterized as a chain like structure after functionalized with cystamine with an average particle size of 17 ± 3 nm, which is the key for designing sensitive devices. Fig. 3b shows the binding of functionalized AuNPs on the graphene sheets, which results in a continuous flow of electrons which can affect the resistance of the materials (Tien et al., 2011). The functionalized AuNPs are well dispersed on the graphene sheets. The functionalized AuNPs are uniform in size on the graphene sheet. Once the morphological studies were conducted, the structural characterizations with respect to functional groups were characterized by Raman and FTIR spectroscopy as discussed in below sections.

Close

Fig. 3 TEM images (a) AuNPs coated Cyst. (b) Shows the functionalized gold NPs on the Gr sheet.

Export to PPT

To characterize the properties of graphene hybrid film, Raman spectroscopy was carried out. Fig. 4 shows the Raman results before and after the binding of functionalized AuNPs onto the graphene sheets. The Raman spectrum for the graphene (Gr) shows intense G-band at 1600 cm⁻¹ arising due to the vibrations of sp² bonded carbon atoms and D-band at 1350 cm⁻¹ suggesting the presence of defects in the graphene sheet (Berciaud et al., 2008; Kudin et al., 2008). The Gr hybrid film consists of peaks from cystamine sulfate hydrate though the peak intensity is weaker than those of cystamine. The I_D/I_G ratio of Gr and Gr hybrid increases from 0.28 to 3.46, which indicates the increase in disorder (Li et al., 2014). Such change in disorder of carbon network in graphene, after functionalization have been reported by Li et al. (2014). This shows the binding of functionalized AuNPs onto the graphene sheets.

Close

Fig. 4 Raman spectra of functionalized gold nanoparticles before and after binding on graphene sheets and cystamine.

Export to PPT

To further study the structural changes caused due to functionalization of AuNPs, FTIR spectroscopy was carried out. Fig. 5 shows FTIR data for cystamine sulfate hydrate which has characteristic peaks at 1615, 1413, and 1026 cm⁻¹ for C⚌O stretching vibration of —NHCO—, C—H stretch and (C—O) respectively (Huang and Lee, 2008). The absorption peaks for Gr hybrid film was observed at 1600, 1400, and 1035 cm⁻¹ due to C⚌O stretching vibration of —NHCO—, C—H stretch and (C—O) respectively. A peak shift was observed in this position after binding of functionalized AuNPs with graphene. Furthermore, the band corresponding to C⚌O stretching of amide band shifts to a lower wave number and the band related to —CH-stretching indicates decrease in wave number. This could be attributed to the interaction of hydrogen bonding between cystamine functionalized AuNPs with the oxygen functional groups on graphene and the electrostatic interaction between the functionalized AuNPs and graphene. This kind of electrostatic interaction studies has been demonstrated by Songmin Shang et al; between chitosan and GO by FTIR (Yang et al., 2010). Similar changes in peak absorbance shifts and increase in the intensity of the peaks were also observed by Kumar et al. (2009). The change in intensity indicates the possible evidence of electrostatic interaction between the functionalized AuNPs with graphene. To prove the hypothesis of electrostatic interactions between the functionalized AuNPs with graphene surface charge measurements was carried out. As-prepared Gr solution had a zeta potential value of −61.3, which indicates Gr to be highly negatively charged. Functionalized AuNPs had a zeta potential value of +17.1 mV. After self-assembly of AuNPs with graphene, the zeta potential value decreases to −2.4 mV when compared to graphene. This proves the electrostatic interaction between graphene and functionalized AuNPs (Han et al., 2010) shown in (supplementary info S1).

Close

Fig. 5 FTIR spectra of Gr, Gr hybrid and cystamine sulfate hydrate.

Export to PPT

The results obtained by electrochemical impedance spectroscopy have been shown in the form of the Nyquist and Bode plots as well as a circuit for fitting EIS data of Au substrate, graphene, AuNPs, and Gr+ functionalized AuNPs specimens (supplementary information S2). In this case, the high-frequency spectra are used to detect the local surface defects, whereas the medium- and low-frequency spectra detect the processes within the pore and at the metal/film interface, respectively. The decrease in the diameter of the arc in the Nyquist diagram indicated that there was a decrease of the charge transfer resistance, indicating that the addition of AuNPs and Gr+ functionalized AuNPs strongly enhanced the conductivity. In addition, the aperture of impedances decreased, whereas phase angles significantly increased with AuNPs and Gr+ functionalized AuNPs containing on the electrode surface. Furthermore, the corresponding equivalent circuit with two-time constants consistent with data in Nyquist and Bode plots is given in (supplementary information S2), where R_s, CPE, R_pore, and R_ct are the solution resistance, constant phase element, pore resistance, and charge transfer resistance. The ZSimpWin program was also used to fit the EIS data to determine the optimized values for the EIS parameters such as R_ct. The results showed that R_ct significantly decreased in the following order: gold substrate > Gr > AuNPs > Gr+ functionalized AuNPs, indicating that a decrease in R_ct value strongly enhances conductivity.

Cyclic Voltammetry was also used for determining the surface coverage of the Au substrate, graphene, AuNPs, and Gr+ functionalized AuNPs specimens in 1 M KCl solution and the results are given in Fig. 6. A significant increase of the current density was obtained from the gold substrate > Gr > AuNPs > Gr+ functionalized AuNPs, suggesting an enhance of conductivity consistent with the results observed in EIS results. In addition, the significant decrease of the current density of 100th cycle in comparison with the first cycle was obtained from gold substrate, Gr, and AuNPs specimens, indicating instability of these specimens, while the insignificant difference of the current density of 100th cycle was performed in Gr+ functionalized AuNPs specimen, suggesting stability of this specimen attributed to the strongly bound to Gr in both oxidation and reduction states.

Close

Fig. 6 First and 100th cycles of cyclic voltammetry reults for (a) gold substrate, (b) graphene coating, (c) AuNPs coating, and (d) Gr+ functionalised AuNPs coating in 1 M KCl solution. The curves were recorded by scanning from −0.6 to 0.6 V_Ag/AgCl at a sweeping rate of 10 mV/s.

Export to PPT

Furthermore, gold substrate and Gr was electrochemically inactive between −0.5 to +0.5 V_Ag/AgCl, whereas AuNPs and Gr+ functionalized AuNPs performed peaks for both oxidation and reduction, attributing to the reduction at −0.22 V mV_Ag/AgCl and oxidation at 0.18 mV_Ag/AgCl. The coverage of gold substrate, Gr, AuNPs, and Gr+ functionalized AuNPs specimens were calculated by integrating the area under the oxidation curve to determine the charge associated. The calculation of surface charge density (Q) is based on the equation (Li and Huang, 2010). $$\text{Electrochemical active surface area}(\text{ECSA})=0.1\text{Q}/(\text{m}\times {\text{Q}}_{\text{o}})$$ where m is the loading amount of sample on the Au electrode, Q is the surface charge can be measured by calculating the integrated area under the peaks around −0.6 V to 0.6 V, Q_o is the electric charge of monolayer oxygen onto the Au electrode assumed to be 450 µC/cm². Maximum surface coverage was found to be 8.37 × 10⁻⁴ g/cm² for Gr⁺ functionalized AuNPs specimen, proving the electrostatic interaction between graphene and functionalized AuNPs which is consistent with the above results.

After, graphene hybrids were characterized, they were tested for sensor applications and to measure the response of graphene hybrid based sensor to range of weights. Our sensor showed a good response when compared with normal graphene film by applying different weights (100–800 mN⁻¹). Different loading and unloading experiments under different pressures were performed. Fig. 7 an illustrates the relative change in resistance (ΔR/R + 1) as a function of force (F), which shows the sensor sensitivity value of 5 × 10⁻⁴ kPa⁻¹ of applied force of 130 mN⁻¹. Using Eq. (1) (Alvares et al., 2011) the sensitivity factor can be calculated.

Close

Fig. 7 (a) Relative change in resistance with respective to applied force. The sensitivity factor at the highest point shows 0.005 mN⁻¹. (b) Detection of current response while loading and unloading of pressure by pressing. (c) Plot shows current response for Tap and release. (d) Plot for current response as a function of time for applied pressure.

Export to PPT

Earlier Alvares et al.; reported a sensitivity factor of (S_F) of 0.0039 mN⁻¹ (Alvares et al., 2011) for functionalized gold nanoparticle based sensor.

Gong et al.; demonstrated ultrathin gold nanowires as a highly sensitive pressure sensor, the sensitivity factor was shown to be 1.14 kPa⁻¹ (Gong et al., 2014). Whereas, the investigated graphene hybrid based sensor shows a high sensitivity factor of 5 × 10⁻⁴ kPa⁻¹ which is best when compared with above works. This was due to the exhibition of extrinsic properties of the graphene and functionalized AuNPs. These sensors have a large surface area, which has more sensitive contact positions that varies charge transfer when pressure is applied. The device shows controllable pressure response (change in resistance) with the same applied force repeatedly and with a change in response time which shown in Fig. 7a. The device was also tested for sensitivity and compared the response to graphene film as shown in Fig. 7b. The graphene hybrid based sensor shows stable responses and high sensitivity for tap and release and weight press applications with a quicker response time less than 10 ms. After the addition of functionalized AuNPs to graphene, the graphene hybrid based sensor shows increase in sensitivity when compared to normal graphene film. Further, the sensitivity was tested for tap and release function. High signal to noise ratio were observed in the force measurement, indicating the higher sensitivity of our graphene hybrid sensor as shown in Fig. 7c.

The graphene hybrid device shows a stable response for tap and release test. Although the bandwidth and line shape was nearly unaltered as the load frequency increased, a quick response time of <15 ms was observed in the unloading process. Fig. 7d shows the response of the device upon applying two different forces, the response time and sensitivity observed, showed that the device is suitable for detecting the forces at low and high pressure limit. Ting Zhang showed similar studies, with a detection limit of 0.6 pa to 2.5 pa (Wang et al., 2014), in comparison this sensor shows a good response with a detection limit of less than 0.6 pa.

3 Conclusion

This study mainly highlights the new type of pressure sensors with high sensitivity and CV, EIS data clearly demonstrate a significant increase of the current density was obtained from the gold substrate > Gr > AuNPs > Gr+ functionalized AuNPs, suggesting an enhance of conductivity. We have successfully developed a novel method for fabricating highly sensitive sensors by self-assembly of functionalized AuNPs between the graphene layers. The flexible graphene hybrid based sensor, was constructed by layer by layer deposition of graphene is followed by functionalized AuNPs and graphene layers. Higher sensitivity was achieved to mimic the natural touch senses. It provides a facile synthesis method for fabrication of hybrid films which could be applied for flexible and high sensitive devices with very low cost of production. Surface analysis techniques also clearly indicated the electrostatic interaction between graphene and functionalized AuNPs. The graphene hybrid based sensor demonstrated high sensitivity for detection of minute forces together with fast response time and high stability which was strongly supported by CV results. This novel device shows a high sensitivity factor of 5 × 10⁻⁴ kPa⁻¹, with a fast response time of <15 ms which shows a better sensitivity compared to previous reports. With this graphene hybrid sensor, we could detect loads in a pressure range of 86 × 10⁻³–539 × 10⁻¹ kPa⁻¹ with corresponding to applications like tap and release. The relative resistance of the films to the applied force is consistent, which is due to electron tunneling current between the functionalized AuNPs and the two graphene layers. The electrochemical studies, like CV and EIS supports the above data and the electrostatic interaction between functionalized AuNPs and graphene triggers favorable changes in the morphology which enhance the behavior of hybrid material. CV results clearly demonstrate the life span and reproducibility of the thin film through repeated cycles. In addition, this study also suggested an excellent agreement was observed among the surface analysis, electrochemical studies for evaluating the performance and sensitivity. This approach opens a new route for various applications such as medical devices, health monitoring and the diagnostics.