Apr 9, 2025
4:30pm - 4:45pm
Summit, Level 4, Room 425
Tara Pena1,Ahn Tuan Hoang1,Ashildur Fridriksdottir1,Zherui Han1,Kathryn Neilson1,Samuel Lai1,Jierong Wang1,Crystal Nattoo1,Tony Heinz1,2,Paul McIntyre1,Andrew Mannix1,2,Eric Pop1
Stanford University1,SLAC National Accelerator Laboratory2
Two-dimensional (2D) semiconductors have emerged as candidates for nanoscale transistors and other opto-electronics. Gate-all-around (GAA) nanosheets, the next-generation transistor architecture, are projected to have exceptional electrostatic control especially with 2D semiconductor channels [1]. Molybdenum disulfide, MoS
2, is presently one of the most promising materials for
n-type 2D transistors, owing to advancements made regarding its industrially compatible large-area growth quality [2], variety of doping techniques [3], and contact schemes [4].
Interestingly, Mo has a wide range of natural mass isotopes, the most common ones between 92 to 100 mass, which have been shown to affect phonon dynamics [5] and optical properties in various 2D heterostructures [6]. Such isotopically pure MoS
2 monolayers have displayed up to 50% higher thermal conductivity than naturally-occurring MoS
2, at room temperature [5]. However, how isotope purification impacts the growth quality of 2D semiconductors and their subsequent electrical performance remains entirely unexplored.
Here, we investigate the impact that isotope purification has on the growth of monolayer MoS
2, by characterizing its optical properties and field-effect transistor performance. We prepare monolayer MoS
2 by chemical vapor deposition (CVD), growing three types of samples using 92, 98, and natural isotope abundance MoO
3 precursors (denoted as 92-MoS
2, 98-MoS
2, and Nat-MoS
2). We then conduct extensive room-temperature Raman spectroscopic mapping to elucidate the effects that isotope engineering has on the phonon modes in monolayer MoS
2. We observe that the isotopically pure samples have sharper E' peaks of ~2.35 cm
-1 in linewidth (at 298 K) compared to Nat-MoS
2 with ~2.75 cm
-1, owing to decreased mass disorder. We confirm that Raman peak position varies with atomic weight, where we observe steep slopes of -23.3 cm
-1 amu
-1/2 and -36.3 cm
-1 amu
-1/2 for the E' and 2LA(M) modes respectively. We also conduct low-temperature photoluminescence (PL), where we observe enhanced contribution in the A-exciton peak and suppressed defect-mediated emission in the 92-MoS
2 at 10 K.
We also fabricate transistors of varying channel lengths (100 nm to 1 µm), where the MoS
2 films are grown directly onto 100 nm SiO
2 on p++ Si global back-gates. Our initial investigations suggest that the threshold voltage may be slightly tuned with atomic mass. Moreover, the isotopically pure 92-MoS
2 devices display roughly twice the transconductance and on-state current in long-channel structures. Overall, the 98-MoS
2 and Nat-MoS
2 devices performed similarly to each other, which matches the observations of low-temperature PL. Isotope purity is expected to yield only small modifications to the MoS
2 band structure [6], which cannot explain the improved electrical performance. Thus, we suggest that using a 92-MoO
3 precursor enables films with fewer defects and/or fewer charged impurities, ostensibly due to more uniform Mo diffusion during the growth process.
In summary, we have employed both Raman mapping and low-temperature PL to study the effects of isotope engineering on the properties of monolayer MoS
2. We find that using a 92-MoO
3 precursor yields higher quality films with double the electron mobility and improved PL emission. Further work entails visualizing the films more directly to quantify defects and impurities. These results suggest that isotope purification may be used as a platform to enhance 2D device performance. T.P. acknowledges support from the NSF MPS-Ascend Postdoctoral fellowship.
References:[1] K.P. O’Brien,
et al., Nat. Commun.,
14, 6400 (2023).
[2] J. Zhu,
et al., Nat. Nanotechnol.,
18, 456-463 (2023).
[3] C.J. McClellan,
et al., ACS Nano,
15, 1587-1596 (2021).
[4] Z. Sun,
et al., ACS Nano,
18, 22444-22453 (2024).
[5] X. Li,
et al., ACS Nano,
13, 2481-2489 (2019).
[6] Y. Yu,
et al., Sci. Adv.,
10, eadj0758 (2024).