Sensor based on chemical vapour deposition-grown molybdenum disulphide for gas sensing application

Abstract

Over the past few decades, sensors based on field-effect transistors have drawn much attention. Initially three dimensional materials were used for sensing, which were later replaced by two dimensional materials because of their ease of manufacturing and large specific areas. Amongst the transition metal dichalcogenides, MoS₂ has been widely used for the fabrication of sensors owing to its ability to differentiate between a charge donor and an acceptor analyte. In this work, we fabricated sensors using chemical vapour deposition grown-MoS₂. MoS₂ was grown on a p-Si/SiO₂ substrate using Mo(CO)₆ as a precursor, the growth was carried out by the sublimation of the precursor under a flow of high purity H₂S at high temperature. The aim of this work is to achieve a level of sensitivity that would enable the detection of individual gas analytes upon adsorption to the MoS₂ surface. To efficiently detect individual gas analytes upon adsorption to the surface, we used interdigitated electrodes in the device architecture to increase the area of the channels for analyte adsorption. We used CO₂ and O₂ gases, which acted as charge donors. A trilayer MoS₂ film was examined, and the detection sensitivity for O₂ was higher in comparison to CO₂. The fabricated device showed significant sensitivity up to parts per million detection level.

---

Introduction

Two dimensional (2D) transition metal dichalcogenides have received immense attention over the past few years. Semiconducting 2D materials offer exceptional features such as structural stability, lack of dangling bonds, and mobility. These features make them suitable candidates for applications in channel materials for field-effect transistors (FETs). Since sensors respond to vapours or gases that are adsorbed on the surface of a channel material, the surface area of a sensing material is the most important factor in evaluating the sensing ability of a sensor. Due to their high surface to volume ratio, 2D materials are very sensitive to the changes in their surroundings, and their high surface area helps in the detection of individual gaseous molecules.^1–6

Molybdenum disulphide (MoS₂) in its monolayer pristine form is a direct band gap semiconductor, owing to which it achieves the field effect phenomenon easily.^7,8 This characteristic behaviour of MoS₂ has been exploited for practical sensing of gases and vapours. A lot of work has been done to realize the gas sensing applications of MoS₂. Both mechanically exfoliated and chemical vapour deposition (CVD)-grown MoS₂ have been used for these applications.^9–16 Several researchers have studied theoretically the effect of different gases on 2D MoS₂ sensors using the first principle calculations.^17–19 Li et al. studied the effect of layer thickness on the performance of a NO sensor based on MoS₂ FET device at room temperature. The results showed that the exfoliated MoS₂ monolayer displayed a rapid response, while the multilayer MoS₂ showed high sensitivity.²⁰ Perkins et al. used exfoliated MoS₂ monolayer for sensing a V-nerve agent. The device showed response at parts per billion (ppb) level with a good selectivity for the donor and acceptor analytes.²¹ Recently, a group studied the effect of external stimuli like biasing on the performance of a MoS₂ sensing device. When a positive gate bias was applied and the device was exposed to a charge withdrawing gas, the increase in the sensitivity of the device was much more pronounced than that observed when a charge donating gas was used. Mechanically exfoliated MoS₂ was used as the channel material and a sensing capability of up to 100 parts per million (ppm) was obtained.²² More recently two groups (Liu et al. and Cho et al.) used CVD-grown MoS₂ as a sensing material. The former used NO₂ and NH₃ gases, and a detection limit of up to 20 ppb and 1 ppm with a sensitivity of up to 80% and 1900% was observed for NO₂ and NH₃, respectively. The latter group used NO₂ and a detection limit of up to 120 ppb with a sensitivity up to 180% (ref. 23 and 24) was observed.

Although the FETs based on MoS₂ have been extensively used to detect various gases, to the best of our knowledge, there are no reports on the detection of CO₂ and O₂ using CVD-grown MoS₂. A few groups have studied ZnO nanorods for sensing O₂.²⁵ Here in, we report the gas sensing characteristics of MoS₂ grown by metalorganic CVD (MOCVD) using Mo(CO)₆ as a precursor that recently emerged as a novel precursor for synthesis of various form of MoS₂.^26,27 Trilayer MoS₂ film was used to detect O₂ and CO₂ gases. We used varying concentrations of CO₂ and O₂ because these two gases have vital impact on our breathing system. O₂ and CO₂ are considered to be important because they play a significant role in the respiration of all living beings. It is highly imperative to quantify the concentration of O₂ and CO₂ to regulate the quality of indoor air. At higher concentrations, CO₂ has adverse health effects including suffocation and drowsiness. We breathe in normal air which contains 20.9 vol% O₂. If the level of O₂ in the air falls below 19.5 vol% the air is considered to be oxygen-deficient, and if it falls below 16 vol% then the air is considered to be unsafe for life.²⁸ The device fabricated by us worked on the principle that a change in its sheet resistance is observed upon the adsorption of gaseous analytes.

Experimental

MoS₂ gas sensor characterization setup

The main objective of the development of a gas sensor characterization setup was to acquire the real time data from the sensors being tested. With the rapid advancement in the microelectronic industry, gas sensing technology has gained immense interest. Using the sensing setup, the behaviour of a gas sensor could be studied under different experimental conditions. The gas sensor characterization setup developed by us (Fig. 1) allows the users to customize experimental conditions according to their need. The gas sensor characterization setup consisted of controlling and measuring instruments. We first injected the test gas (O₂ or CO₂) into the dilution chamber using mass flow controllers (MFC). The chamber was then filled with an inert gas (Ar) to achieve the required test gas concentration. Finally the diluted gas was introduced into the test chamber which initially was in a vacuum condition. The gas sensor was placed in a test vacuum chamber and the electrodes were connected to a digital multimeter (Agilent, 34461A, 0.15 to 0.8 MΩ error range) for electrical measurements via a feed through. A LabVIEW program was used for synchronization. Hence, the possibility of data loss was minimized.

Fig. 1 MoS₂ gas sensor characterization setup that consisted of test gases (O₂ and CO₂), chambers (mixing and test vacuum chamber), controlling (MFC and angle valve) and measurement instruments (digital multimeter).

Growth of MoS₂ on SiO₂/p-Si

The MoS₂ was grown using the MOCVD method as we previously reported.²⁹ MoS₂ thin films were grown on a 1 by 1 cm size of thermally grown 300 nm thick SiO₂ on p-Si substrate using Mo(CO)₆ (Sigma-Aldrich, 577766) as a precursor and high purity H₂S as the reaction gas in a shower head type reactor. The growth was carried out at different partial pressure ratios of the sulphur-reaction gas to the Mo-precursor (P_SR/P_MoP). After cleaning, the substrate was loaded into the load-lock chamber which is connected with showerhead type main CVD reactor. Subsequently, substrate was moved into pre-heated main CVD reactor to perform growth. To obtain two-dimensional (2D) MoS₂ film, growth process was performed under higher P_SR/P_MoP condition. The fully covered MoS₂ film was obtained in our optimized condition of 5 Torr pressure, 200P_SR/P_MoP with 30 sccm H₂S at 350 °C and the 10 cm distance of showerhead to substrate was used. The P_MoP was controlled by changing the temperature of Mo-precursor and no dilution or carrier gas was used.

Fabrication of sensing device

The FET device for gas sensing was fabricated using the photolithography process. Bi-layer process employed to make lithography process easier. Photoresist of LOR5B (MicroChem) which used for undercut layer was first spin coated sequentially for 5 s at 500 rpm and 45 s at 3000 rpm to obtain 0.5 μm thick film. Subsequently, photoresist was soft baked at 190 °C for 5 min. Positive resist of S1813 (MICROPOSIT) was spin coated at 4000 rpm for 60 s to obtain 1.5 μm thick film and soft baked at 115 °C for 1 min. We used a Karl-Suss mask aligner having a UV light intensity of 7 mW cm⁻² and wavelength of 400 nm. The bi-layer photoresist film was exposed for 15 s. After that, exposed area was developed using MF-319 developer for 20 s by constant stirring and was then rinsed with de-ionized (DI) water. The top electrodes were deposited using e-beam evaporator. The metal used for defining the electrodes was Au/Ti; thickness of Ti was 5 nm, while that of Au was 50 nm. During the lift-off process, the substrate was immersed in a remover PG (MicroChem) at 40 °C for 4 min with constant stirring and was finally rinsed with DI water. Fig. 2a shows SEM image of the sensing device, and the photograph of the sensor incorporated onto printed circuit board (PCB). After the fabrication of the device, it was integrated with a PCB, and contact pads were connected to the PCB socket hole using a wire bonding method (Fig. 2b).

Fig. 2 (a) SEM image of sensing device. Distance between electrodes is 4.4 μm. Width and length of electrode is 37.5 and 520 μm, respectively. The sensing device was incorporated onto PCB for electrical measurement (inset). (b) Schematic diagram of incorporated sensing device with PCB. To incorporate, device was first attached PCB by using silver paste, and then contact pad was connected to socket hole by using wire bonding method.

Results and discussion

Characterization of MoS₂ film

In order to examine the morphology, grain size, and thickness of the grown film, it was characterized using scanning electron microscopy (SEM) (Hitachi S-4800) and Raman spectroscopy (Thermo Fisher Scientific, DXR). From Fig. 3a, it can be seen clearly that triangular 2D island of MoS₂ were formed. Increased growth times led to the coalescence of these monolayer islands to cover the substrate. The MoS₂ bilayer film with trilayer islands (Fig. 3b) was obtained with 20 h growth time and trilayer MoS₂ film was obtained after 26 h growth. The lateral dimension of the MoS₂ triangles so formed was 0.3 μm. In case of the 2D growth, at higher P_SR/P_MoP, the Mo or S edge was covered by 100% sulphur to have the lowest surface energy resulting in the formation of triangular MoS₂ islands.^30,31 From the Raman spectroscopy measurement, the MoS₂ film shows two dominant peaks of E¹_2g and A_1g at difference position for different growth time samples due to thickness of MoS₂ film. The differences between E¹_2g and A_1g measured as 19 and 23 cm⁻¹ for monolayer and trilayer film, respectively.^32,33

Fig. 3 SEM image of MoS₂ film. Triangular monolayer MoS₂ islands were grown on SiO₂/Si substrate at the initial growth time of 4 h (a). The partially covered trilayer MoS₂ on fully covered bilayer film was grown after 20 h (b). The differences between two dominant peaks of in-plane (E¹_2g) and out-of-plane (A_1g) vibration was used to analyze number of layers of films measured as 19 and 23 cm⁻¹ for monolayer islands (blue) and trilayer film (red), respectively. The intensity of Si peak was reduced owing to larger interference from trilayer MoS₂ film (c). The Raman measurement was performed using a laser with excitation wavelength of 532 nm and power of 8 mW.

Gas sensing behaviour of MoS₂ film

The sensing behaviour of the fabricated gas sensors were tested under CO₂ and O₂ at different concentrations. To obtain the response of the sensor at desired concentrations, the gases were diluted with an inert gas (Ar) in a mixing chamber. The required gas concentration was achieved by making the gases flow through the mass flow controller at a specific flow rate for certain period of time using the equation:

The gas sensitivity was measured using the equation:

where, R_gas is the resistance in the presence of the test gas and R_inert is the resistance in the inert environment. Resistance in the inert environment was recorded by using an Ar gas flow to serve as the base value resistance (200 MΩ). The sensor was then exposed to CO₂ and O₂ at a concentration level of 1000 ppm with flow rate of 10 sccm. Throughout the sensing experiment, the pressure of test vacuum chamber was maintained at a 0.5 Torr by angle valve. When the device was exposed to the gases, a charge transfer between the gas molecules and MoS₂ sheet occurred resulting in a change in their resistance. Charge depletion in the MoS₂ sheet reduces the number of electrons, and thus, reduces the number of charge carriers in it and enhances its resistance, while the charge accumulation leads to a decrease in the resistance. The strong interactions between the gas molecules and MoS₂ sheet lead to a significant change in conductivity of MoS₂. This is the basic working principle of gas sensing devices. The resistance of the sensor decreased upon exposure to CO₂. Since CO₂ is an electron donating gas, it possesses n-doping characteristics and upon exposure to MoS₂ channel, it injects charges to the surface, which in turn shifts the Fermi level to the conduction band of MoS₂ leading to a decrease in the resistance of the sensor.¹⁶ The fabricated device displayed a quick response upon exposure to the gases at room temperature. Note that the response time was much greater than the recovery time. The sensor was exposed to the gases until a saturation point was achieved. Once the device reached the saturation, the flow of CO₂ was turned off and the sensor was again exposed to the inert gas (Ar) in order to recover the original resistance. Subsequently, temperature of cartridge heater stage was increased up to 40 °C to facilitate the process of desorption. Since the gas molecules were physisorbed on MoS₂ sheet, the bonding between them and MoS₂ sheet was weak, thus the molecules could be easily desorbed by the flow of the inert gas (Ar) or by applying a little heat. One cycle was recorded by exposing the MoS₂ sheet first to CO₂ and then to the inert gas (Ar). The response and recovery times were found to be 37 and 53 s, respectively. The sensitivity was recorded to be 3.95% (Fig. 4a). Using the same procedure, the sensor was exposed to 1000 ppm O₂. Upon exposure to O₂, the sensor showed a decrease in the resistance, which indicates that O₂ is also an electron donating gas and leads to a decrease in the channel resistance. The response and recovery times at 1000 ppm O₂ were measured to be 18 and 47 s, respectively.

Fig. 4 (a) Sensor response upon exposure to 1000 ppm CO₂. (b) Response under different concentration of CO₂. (c) Sensor response upon exposure to 1000 ppm O₂. (d) Response under different concentration of O₂. (e) Cyclic test to check the stability of device under 1000 ppm. (f) Change in sensitivity at different gas concentration levels, the lines shows the best fit curves.

The sensitivity was recorded to be 4.84% (Fig. 4c). After measuring the response upon exposure to 1000 ppm, the sensor was tested under different gas concentrations. The response of the sensor at various CO₂ and O₂ concentrations is shown in Fig. 4b and d, respectively. A detection limit of up to 100 ppm could be achieved. However, it is evident from the graphs that the sensor had the potential to work even at lower concentration levels. Fig. 4f shows the relationship between the sensitivity and CO₂ concentration. The best fit line shows that the relationship between the CO₂ concentration and sensitivity was linear. Repeatability is considered to be the most important factor for the commercial gas sensing devices. Commercialized products need to work for several cycles under different ambient conditions. The cyclic test was performed by injecting the test gas and inert gas alternatively. It can be seen from Fig. 4e that the response and recovery times were almost constant with standard deviation of 0.018% for each cycle of this device.

Both the gases showed similar behaviour (i.e., acted as electron donors). However, some differences in the sensitivity and the response and recovery times were noticed. Our sensor was more sensitive to O₂ than to CO₂. Moreover, the time required to complete equal number of cycles in the case of O₂ was lesser than that required in the case of CO₂. The main reason behind this observation is the fact that the ability to donate charges and adsorption energy in O₂ is much higher than that in CO₂. From Fig. 4f, it is clear that as the concentration increases, the sensitivity also increases. This linear relationship is very important from the commercial point of view.

Conclusions

A highly sensitive gas sensor having adequate sensitivity at room temperature was developed. A gas sensor characterization system was set up and its interfacing was done to obtain a continuous real time analysis automatically. An interdigitated electrode architecture was used to obtain the appropriate charge collection and sensitivity. Both the gases (CO₂ and O₂) acted as charge donors. From the results, it is clearly evident that the sensing ability of the sensor was higher for O₂ than that for CO₂. In addition, the total time required for completing equal number of cycles was more in the case of CO₂ than that in the case of O₂. Further studies such as doping effect, sensing using the 2D heterostructure materials, and sensing property of different number of MoS₂ layer could be helpful for the realization of gas sensing mechanism by 2D MoS₂.

Acknowledgements

This work was supported by Korea Evaluation Institute of Industrial Technology (KEIT) funded by the Ministry of Trade, Industry and Energy (MOTIE) (Project No. 10050296, Large scale (Over 8′′) synthesis and evaluation technology of 2-dimensional chalcogenides for next generation electronic devices).

References

1. A. D. Franklin, Nature, 2013, 498, 443.
2. M. M. Shulaker, G. Hills, N. Patil, H. Wei, H.-Y. Chen, H.-S. P. Wong and S. Mitra, Nature, 2013, 501, 526.
3. K. S. Novoselov, V. I. Fal'ko, L. Colombo, P. R. Gellert, M. G. Schwab and K. Kim, Nature, 2012, 490, 192.
4. R. Ganatra and Q. Zhang, ACS Nano, 2014, 8, 4074–4099.
5. Z. Sun and H. Chang, ACS Nano, 2014, 8, 4133.
6. D. Jariwala, V. K. Sangwan, L. J. Lauhon, T. J. Marks and M. C. Hersam, ACS Nano, 2014, 8, 1102.
7. K. Lee, R. Gatensby, N. McEvoy, T. Hallam and G. S. Duesberg, Adv. Mater., 2013, 25, 6699.
8. Z. Yin, H. Li, H. Li, L. Jiang, Y. Shi, Y. Sun, G. Lu, Q. Zhang, X. Chem and H. Zhang, ACS Nano, 2012, 6, 74.
9. Y. Li, C.-Y. Xu, J.-Y. Wang and L. Zhen, Sci. Rep., 2014, 4, 7186.
10. B. Cho, J. Yoon, S. K. Lim, A. R. Kim, D. H. Kim, S. G. Park and H. C. Ko, ACS Appl. Mater. Interfaces, 2015, 7, 16775.
11. B. Cho, M. G. Hahm, M. Choi, J. Yoon, A. R. Kim, Y. J. Lee and Y. Jeong, Sci. Rep., 2015, 5, 8052.
12. Q. He, Z. Zeng, Z. Yin, H. Li, S. Wu, X. Huang and H. Zhang, Small, 2012, 8, 2994.
13. M. Donarelli, S. Prezioso, F. Perrozzi, F. Bisti, M. Nardone, L. Giancaterini and L. Ottaviano, Sens. Actuators, B, 2015, 207, 602.
14. R. Samnakay, C. Jiang, S. L. Rumyantsev, M. S. Shur and A. A. Balandin, Appl. Phys. Lett., 2015, 106, 023115.
15. S. Y. Cho, S. J. Kim, Y. Lee, J. S. Kim, W. B. Jung, H. W. Yoo and H. T. Jung, ACS Nano, 2015, 9, 9314.
16. W. Zhang, P. Zhang, Z. Su and G. Wei, Nanoscale, 2015, 7, 18364.
17. Q. Yue, Z. Shao, S. Chang and J. Li, Nanoscale Res. Lett., 2013, 8, 1.
18. N. Yu, L. Wang, M. Li, X. Sun, T. Hou and Y. Li, Phys. Chem. Chem. Phys., 2015, 17, 11700.
19. S. Zhao, J. Xue and W. Kang, Chem. Phys. Lett., 2014, 595, 35.
20. H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu and H. Zhang, Small, 2012, 8, 63.
21. F. K. Perkins, A. L. Friedman, E. Cobas, P. M. Campbell, G. G. Jernigan and B. T. Jonker, Nano Lett., 2013, 13, 668.
22. D. J. Late, Y. K. Huang, B. Liu, J. Acharya, S. N. Shirodkar, J. Luo and C. N. R. Rao, ACS Nano, 2013, 7, 4879.
23. B. Liu, L. Chen, G. Liu, A. N. Abbas, M. Fathi and C. Zhou, ACS Nano, 2014, 8, 5304.
24. B. Cho, A. R. Kim, Y. Park, J. Yoon, Y. J. Lee, S. Lee and D. H. Kim, ACS Appl. Mater. Interfaces, 2015, 7, 2952.
25. C. S. Chou, Y. C. Wu and C. H. Lin, RSC Adv., 2014, 4, 52903.
26. Z. Liu, L. Zhang, R. Wang, S. Poyraz, J. Cook, M. J. Bozack, S. Das, X. Zhang and L. Hu, Sci. Rep., 2016, 6, 22503.
27. J. J. Pyeon, S. H. Kim, D. S. Jeong, S.-H. Baek, C.-Y. Kang, J.-S. Kim and S. K. Kim, Nanoscale, 2016, 8, 10792.
28. G. Korotcenkov, Handbook of gas sensor materials, Springer, New York, 2013.
29. J. Mun, Y. Kim, I. Kang, S. K. Lim, S. J. Lee, J. W. Kim, H. M. Park, T. Kim and S. Kang, Sci. Rep., 2016, 6, 21854.
30. S. Helveg, J. V. Lauritsen, E. Lægsgaard, I. Stensgaard, J. K. Norskov, B. S. Clausen and F. Besenbacher, Phys. Rev. Lett., 2000, 84, 951.
31. H. Schweiger, P. Raybaud, G. Kresse and H. Toulhoat, J. Catal., 2002, 207, 76.
32. C. Lee, H. Yan, L. E. Brus, T. F. Heinz, J. Hone and S. Ryu, ACS Nano, 2010, 4, 2695.
33. H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Edwin, A. Olivier and D. Baillargeat, Adv. Funct. Mater., 2012, 22, 1385.

---

Footnote

† Equally contribute.

---