Enhancement of Carrier Mobility in High Dielectric Materials S. Balendhran 1* , J. Z. Ou 1 , J. Tang 2 , K. L. Wang 2 , S. Zhuiykov 3 , S. Sriram 1 , M. Bhaskaran 1 , and K. Kalantar-zadeh 1 1 School of Electrical and Computer Engineering, RMIT University, Melbourne, Victoria, Australia 2 Device Research Laboratory, Department of Electrical Engineering, University of California, Los Angeles, California, USA. 3 Materials Science and Engineering Division, CSIRO, Highett, Victoria, Australia. *Corresponding author: Email shiva.balendhran@rmit.edu.au Abstract: In this work, charge carrier mo- bility in a two dimensional (2D) high dielec- tric (high-κ) material is investigated. Field effect transistors (FET) based on Molybde- num trioxide (MoO 3 ) are fabricated. The MoO 3 flakes were reduced to sub- stoichiometric MoO (3-x) to achieve bandgap values viable for transistor applications. The scattering mechanisms limiting the overall mobility with respect to temperature are explored. Theoretical and experimental mobilities are compared in low and high dielectric materials. 1 Introduction: The discovery of graphene in 2004 [1], has drawn the interests of both industries and the scientific community on 2D materials. Although the enhanced car- rier mobility observed in graphene is highly desirable for electronic applications, the lack of intrinsic bandgap leads to the explo- ration of alternative 2D materials with natu- ral bandgap. Molybdenum disulphide (MoS 2 ) is such a material with a direct bandgap of 1.9 eV in atomically thin layers [2]. Thus far the highest reported carrier mobility in single layer MoS 2 is ~217 cm 2 /Vs [2], in which Kis et al. employ a hi-dielectric top gate, to reduce coulomb scattering. This work explores the capabilities of intrinsic high-κ materials as an alternative for achieving enhanced charge carrier mo- bilities. MoO 3 is one of the transition metal oxides that has a relative dielectric con- stant of ~500 (~ 5 for MoS 2 ) and can be exfoliated to minimum resolvable atomi- cally thin layers. But in its intrinsic nature, MoO 3 has a wide bandgap (>3 eV) which is not viable for transistor applications. However the bandgap of MoO 3 can be easily manipulated to desirable values by several techniques such as hydrogen ion (H + ) intercalation, UV irradiation, electron beam bombardment etc. [3, 4]. The above mentioned techniques produce partially re- duced sub-stoichiometric MoO (3-x) which has increased carrier concentration and a high dielectric value which highly favours the enhancement in charge carrier mobility. 2 Experimental: α-MoO 3 crystals were synthesized by thermally evaporating MoO 3 powder and atomically thin layers were acquired through mechanical exfolia- tion of these crystals on 300 nm SiO 2 /Si substrates [5]. Electrical contacts (Au/Ti) were fabricated using a combination of photo and electron beam lithography tech- niques. Current–voltage (I–V) characteris- tics of the back-gated FETs were attained in a range of temperatures from room tem- perature to 100 °C. 3 Results and Discussion: Drain current (I DS ) vs. gate voltage (V GS ) characteristics of a MoO (3-x) FET with flake dimensions of 800×80×11 nm is shown in Fig. 1. Room temperature charge carrier mobility was calculated to be ~1100 cm 2 /Vs. The mobil- ity value over a period of 5 days remained within 10% of variation. The experimental mobilities and the theoretical behaviour of the mobility, calculated using Borne ap- proximation with respect to temperature are presented in Fig 2. The overall mobility influenced by coulomb, acoustic, and opti- cal phonon scattering mechanisms is cal- culated by the Matthiesen’s Rule. As shown in Fig. 2, the acoustic and optical phonon scattering effects are independent of the dielectric value of the material. In a low-κ (~5) material coulomb scattering ef- fect seems to be the limiting mechanism of the overall mobility. As a result the theo- retical prediction of the overall mobility lays around 170 cm 2 /Vs and the experimental mobility reported by Kis et al. for MoS 2 closely matches this value. In a high-κ