Chemical doping of MoS2 multilayer by p-toluene sulfonic acid

We report the tailoring of the electrical properties of mechanically exfoliated multilayer (ML) molybdenum disulfide (MoS2) by chemical doping. Electrical charge transport and Raman spectroscopy measurements revealed that the p-toluene sulfonic acid (PTSA) imposes n-doping in ML MoS2. The shift of threshold voltage for ML MoS2 transistor was analyzed as a function of reaction time. The threshold voltage shifted toward more negative gate voltages with increasing reaction time, which indicates an n-type doping effect. The shift of the Raman peak positions was also analyzed as a function of reaction time. PTSA treatment improved the field-effect mobility by a factor of ~4 without degrading the electrical characteristics of MoS2 devices.


Introduction
Over the last several years, two-dimensional (2D) semiconducting transition-metal dichalcogenide (TMD) has attracted increased research attention because of its emerging electrical and optical properties and great potential in practical applications [1,2]. Graphene is a representative 2D material that has a conical Dirac spectrum of energy states without a band gap, which results in many interesting physical properties as well as stimulating applications [3][4][5][6]. However, the gapless band structure of graphene makes it unsuitable as an electronic material in logic circuits. Molybdenum disulfide (MoS 2 ), a layered TMD, has emerged as a feasible alternative to graphene because it has a moderate energy gap with mechanical flexibility, chemical and thermal stability, and absence of dangling bonds [7][8][9][10][11][12][13][14]. A layer of MoS 2 consists of a molybdenum monolayer sandwiched between two sulfur monolayers. The strong intralayer bonds and weak interlayer van der Waals forces make it possible to exfoliate individual MoS 2 layers. MoS 2 crystals can also be obtained in a large scale by chemical exfoliation and chemical vapor deposition (CVD) techniques [15]. MoS 2 -based electronic and optoelectronic devices such as field-effect transistors (FETs), integrated circuits, solar cells, photodetectors, memory devices, chemical and biosensors, supercapacitors, and photocatalyzed hydrogen evolution reactors have been successfully fabricated [10,[14][15][16][17]. FET is a basic and important application of semiconducting TMD materials. Several issues, including reducing the contact resistance, using different substrates, and depositing high-k materials as the top gate, have been focused on to improve the performance of MoS 2 FET [18][19][20][21][22]. Given the ultrathin structure of MoS 2 layers, doping of this material has yet to be fully developed. Traditional methods of doping such as ion implantation are not suitable for MoS 2 thin layer; hence, alternative approaches, such as chemical and molecular doping, must be explored.
Chemical doping has been used to investigate the surface charge transfer between dopant molecules and MoS 2 layer [23]. The charge transfer between dopant and host material modulates the Fermi level and results in the modification of the optical and electrical properties of 2D materials. Chemical doping of MoS 2 nanoflakes with solution-based dopants, gases, metal nanoparticles, or self-assembled monolayers has already been reported [24][25][26][27][28]. Mouri et al reported the tunable photoluminescence of monolayer MoS 2 flakes by n-type and p-type dopants in chemical solution [25]. Fang  Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
ions and the significant change in the electron density of MoS 2 nanoflakes [26]. Gold nanoparticles were used by Shi et al to effectively decorate MoS 2 layers via wet chemical method [27]. They found that gold nanoparticles impose p-doping to the MoS 2 transistors. Li et al studied the carrier doping of MoS 2 nanoflakes by functional self-assembled monolayers with different dipole moments [28]. Our group also demonstrated ultraviolet light-induced reversible and stable charge carrier modulation in single-layer, bilayer and multilayer (ML) MoS 2 nanosheets with a combination of N 2 and O 2 gases [29].
Here, we report the tailoring of the electronic properties of ML MoS 2 by p-toluene sulfonic acid (PTSA) molecular doping. PTSA is a well-known dopant for conducting polymers and is highly soluble in water. The charge transport measurements and Raman spectroscopy revealed that PTSA molecule imposes n-doping in ML MoS 2 . The threshold voltages shifted toward more negative gate voltages after exposure of PTSA molecules for different periods of time, revealing the n-doping in ML MoS 2 . The charge carrier density and field-effect mobility were also estimated and found to be significantly improved after PTSA treatment. Results indicate that chemical modification is a simple approach to tailor the electrical properties of few-layered MoS 2 while maintaining its important electrical characteristics.

Sample preparation
ML MoS 2 film was mechanically exfoliated from bulk crystals of molybdenite (SPI Supplies, USA) by the scotch tape method and then transferred to a 300 nm thick highly p-doped SiO 2 /Si substrate. Optical microscopy, Raman spectroscopy, and atomic force microscopy (AFM) aided in identifying the number of layers of the MoS 2 flakes. The ML MoS 2 had seven to eight monolayers. Raman spectra were obtained with a Renishaw microspectrometer with 514 nm laser wavelength at room temperature. The laser power was kept at ≈1.0 mW to avoid local heating and defect introduction by the laser.

Device fabrication and measurements
Photolithography technique was used to make large patterned electrodes (Cr/Au of 6/30 nm thicknesss) for ML MoS 2 devices. As the final process, Cr/Au (10/80 nm) ohmic contacts were patterned by e-beam lithography. The channel length of transistors was kept almost same (~2.8 μm) for all devices. The devices were annealed in a tube furnace at 200°C in a flow of 100 sccm Ar and 10 sccm H 2 for 4 h to remove lithography resist residue and minimize the contact resistance of the devices. Electrical characterizations of the devices were performed using Keithley 2400 and Picometer 6485 instruments by two-probe measurements at room temperature in vacuum.

PTSA doping and characterizations
Electrical charge transport measurements and Raman spectroscopy were used to characterize the pristine ML MoS 2 film. The same device was then doped with PTSA for different periods of time, and the effect of PTSA doping was investigated by charge transport measurements and Raman spectroscopy. The PTSA monohydrate (ACS reagent, ⩾ 98.5%, Aldrich) was dissolved in deionized water to make a PTSA solution with 0.1 M concentration. The ML MoS 2 film on the Si/SiO 2 substrate was soaked in the PTSA solution for certain periods of time and blow-dried by nitrogen. The sample was then placed in a vacuum desiccator for 1 d to completely dry. The procedure described in our previous papers was followed [30,31]. Tailoring the electronic properties of semiconducting channel materials is essential for using these materials in high-performance electronic and optoelectronic devices. Several approaches have been applied to modulate the electronic properties of 2D nanomaterials by depositing dopant atoms, chemical modification by absorption of gas molecules/ aromatic compounds, and by surface-induced interstitial doping [32]. In general, interstitial doping is usually difficult to control and often introduces defects, thereby reducing the mobility of host materials. Chemical modification, especially non-covalent functionalization, is one of the most effective methods to tailor the electrical properties of 2D nanomaterials. This type of chemical modification does not change the basic electronic structure and preserves the desired electronic properties of 2D nanomaterials by minimizing the damage to the lattice. In our previous reports, the electronic properties of exfoliated single-layer, bilayer and tri-layer graphene, as well as CVD-grown single-layer graphene, were tailored by PTSA molecular doping without degrading its transparency and electrical properties [30,31]. Figure 2(a) shows the Raman spectra of ML MoS 2 before and after PTSA modification for different periods of time. Two characteristic peaks, E 1 2g (in-plane vibration) and A 1g (out-of-plane vibration), appear around 384 and 408 cm -1 , respectively. The frequency difference between these two Raman modes varies depending on the number of layers and can be easily used as a thickness indicator. The frequency difference between the Raman A 1g and E 1 2g modes (Δ = A 1g − E 1 2g ) is approximately 24 cm -1 , indicating few layers. The Raman spectra of ML MoS 2 after PTSA treatment shows downward shifting of the E 1 2g and A 1g peak positions compared with pristine ML MoS 2 , as shown in figure 2(a). The downward shifting of the E 1 2g and A 1g peak positions is attributed to the n-doping of MoS 2 as previously reported by others for different systems [33,34]. The shifting of the E 1 2g and A 1g peaks toward low wave numbers increases with increasing PTSA exposure time. The n-doping of ML MoS 2 is also confirmed by the electrical charge transport measurements. Figure 2(b) displays the drain current I DS as a function of the applied back-gate voltage V g at a fixed drain-source voltage, V DS = 10 mV, for pristine and PTSA-doped ML MoS 2 . All electrical characterizations of the devices were performed at room temperature in a vacuum chamber. The I DS -V g graph reveals an n-type channel for ML MoS 2 . The I DS shifts toward negative V g after PTSA treatment. The shift toward the negative gate voltage increases with increasing PTSA exposure time, as shown in figure 2(b). Shifting of the threshold voltage toward the negative gate voltage reveals the n-doping in the ML MoS 2 layers. The same trend is observed on the other devices. Figure 3(a) illustrates the threshold voltage as a function of PTSA exposure time. The threshold voltage shifts toward the negative gate voltage as the PTSA reaction time increases, revealing the n-doping in MoS 2 . This chemical doping turns out to be very stable and the same characteristics of the PTSA doped devices have been observed after 10 days in the ambient atmosphere. Figure 3(b) shows the field-effect mobility of ML MoS 2 as a function PTSA reaction time. The mobility of the samples was determined using the relation ,

Results and discussion
where L is the channel length, W is the channel width, is the slope of the transfer characteristic of the device in the linear region for particular doping time, and V DS = 0.01 V. The length and width of the fabricated device are 2.8 and 9.7 μm, respectively. The gate capacitance C g for Si/SiO 2 is ∼115 aF μm −2 . The mobility of pristine ML MoS 2 was measured as 22.4 cm 2 V −1 s −1 . The mobility remarkably improved after PTSA treatment, and it was found to be 84 cm 2 V −1 s −1 after 30 min of exposure to PTSA molecules. The improvement in mobility may be due to reduction of Schottky barrier height between source/drain and MoS 2 layers. PTSA increases the electron density in MoS 2 channel, which changes the Fermi level of MoS 2 and lowers the Schottky barrier height between the electrode and MoS 2 . Figure 3(b) also shows the charge carrier density (n) of ML MoS 2 as a function of PTSA exposure time at V g = 0 V. Figure 3(b) clearly shows that the charge carrier density of ML MoS 2 significantly changed after different periods of PTSA treatment. The charge carrier densities of our ML MoS 2 transistors were estimated using the relation ( ) where e is the elementary charge and V T is the corresponding threshold voltage of the device at different PTSA exposure times.
The n-type doping effect by PTSA solution can be understood as follows. Mo is electropositive in nature as it belongs to 4th transition group of periodic

Conclusions
A simple technique to modulate the electronic properties of ML MoS 2 by PTSA molecular doping was demonstrated. The effect of PTSA doping on the electric properties of ML MoS 2 was investigated by Raman spectroscopy and charge transport measurements. The charge transport and Raman spectroscopy measurements revealed that PTSA molecules impose n-doping in ML MoS 2 . The threshold voltage shifted toward more negative gate voltages, thereby confirming n-doping in ML MoS 2 . The shift of the Raman peak frequencies was also analyzed as a function of reaction time. Our study demonstrated that molecular n-doping using PTSA is a feasible scheme for improving the electronic properties of MoS 2 -based devices.