Fractional quantum Hall phases in high-mobility n-type molybdenum disulfide transistors | Nature Electronics

Nature Electronics (2024)Cite this article

867 Accesses

1 Altmetric

Metrics details

Transistors based on semiconducting transition metal dichalcogenides can, in theory, offer high carrier mobilities, strong spin–orbit coupling and inherently strong electronic interactions at the quantum ground states. This makes them well suited for use in nanoelectronics at low temperatures. However, creating robust ohmic contacts to transition metal dichalcogenide layers at cryogenic temperatures is difficult. As a result, it is not possible to reach the quantum limit at which the Fermi level is close to the band edge and thus probe electron correlations in the fractionally filled Landau-level regime. Here we show that ohmic contacts to n-type molybdenum disulfide can be created over a temperature range from millikelvins to 300 K using a window-contacted technique. We observe field-effect mobilities of over 100,000 cm2 V−1 s−1 and quantum mobilities of over 3,000 cm2 V−1 s−1 in the conduction band at low temperatures. We also report evidence for fractional quantum Hall states at filling fractions of 4/5 and 2/5 in the lowest Landau levels of bilayer molybdenum disulfide.

At low temperatures, transition metal dichalcogenide (TMD) transistors can potentially offer high mobilities, strong spin–orbit coupling and large effective masses. They could, therefore, be used to explore various electronic interactions. One such interaction-driven phenomena is the fractional quantum Hall (FQH) state—the many-body effects of incompressible gaps at partial fillings of Landau levels (LLs)1,2,3,4. FQH states are revealed through their edge conduction channels, whose transport exhibits a minimum of longitudinal resistivity ρxx and a transverse conductivity σxy quantized as νe2/h, where e is the elementary charge and h is Planck’s constant. In the composite fermion theory5,6,7, the main series of fractions is given by ν = p/(2kp + 1) (k and p are integers) or its particle–hole (p-h) conjugate partner (ν = 1 − p/(2kp + 1)).

The FQH states are considered of potential use in topological quantum computations due to the non-conventional exchange statistics in the non-Abelian 5/2 states that have been observed in GaAs and ZnO semiconductors8,9,10. Theoretical efforts have also been focused on the development of superconducting FQH hybrid systems to engineer parafermions—a generalized version of Majorana fermions—for quantum computing purposes with fractional topological orders11,12,13. However, only a handful of solid-state two-dimensional electron gas (2DEG) systems have been developed that exhibit fractionally quantized σxy plateaus, including semiconductor quantum wells14,15,16 and graphene17,18,19,20,21,22. The SU(4) symmetry breaking of LLs in graphene has, for instance, led to collective states including denominator FQH states in the lowest LLs and in higher LLs23,24,25,26,27.

Semiconducting TMDs have a variety of interacting phases due to the massive Dirac fermions character of their charge carriers and their spin–orbit coupling28,29,30. Recently, mono- or few-layered semiconducting TMDs have been shown to exhibit giant effective g-factors (here g-factor is a dimensionless quantity that characterizes the magnetic moment and angular momentum of a particle)30,31,32,33, non-conventional sequence of Hall plateaus34,35 and LL crossings with possible quantum Hall antiferromagnetic phases36. However, most of these studies were performed far above the quantum limit and at filling fractions ν larger than 1.

Capacitance probes on high-quality monolayer tungsten diselenide (WSe2) has revealed a series of FQH gaps in the lowest and first excited LLs37, suggesting that fractional quantizations are, in principle, available in TMDs for transport measurements—a requirement for the construction of quantum devices based on FQH states. Still, FQH plateaus in TMDs—particularly in the lowest LLs at which electron interactions are the most pronounced—remain elusive. The challenge lies in the lack of access to low carrier density close to the band edge, an ohmic contact to the semiconducting channel at low temperature and an appropriate dielectric environment that favours the Coulomb interaction.

In this Article, we show that a window-contacted technique can create ohmic contacts to n-type molybdenum disulfide (MoS2) over a temperature range from millikelvins to 300 K. We observe high field-effect electron mobilities (exceeding 100,000 cm2 V−1 s−1) and quantum mobilities (over 3,000 cm2 V−1 s−1) at low temperatures. We probe the electrical transport of monolayer MoS2 down to the low carrier density required to reach the quantum limit. We report evidence for FQH at filling fractions of 4/5 and 2/5 in bilayer MoS2 via transport measurements under high magnetic fields of up to 34 T and a base temperature of 300 mK. Our calculations and measurements suggest that under the application of a finite vertical electric field (Ez), the system behaves like a monolayer MoS2 for which the first four (lowest and first excited) LLs in the conduction band are layer-valley locked, and the spin degeneracy is fully lifted by the Zeeman energy. Therefore, the LL spectra of the MoS2 bilayer, together with the asymmetric ohmic contact to the high-mobility bilayer semiconducting channel, provide a model to investigate electron–electron interactions in the FQH regime.

Various methods, including phase-engineered edge contact and the use of Pt or Au bottom electrodes29,38,39,40, have been used to try and establish electrical contact with semiconducting MoS2 for quantum transport studies, but it has not been possible to probe the lowest LL approaching the quantum limit through transport measurements. Recently, low-melting-point metals like Bi and (01\(\bar{1}\)2)-faceted Sb have emerged as potential candidates for forming ohmic contacts with the MoS2 channel41. However, thus far, these studies have been limited to relatively high-temperature regimes (above a few tens of kelvins). Here we adopt the Bi contact method42, but modify it into a two-dimensional (2D) windowed convention.

We first pre-pattern etched-through windows (with sizes of ~1 × 1 μm2) in few-layered hexagonal boron nitride (hBN), which are further used as a top boron nitride for encapsulating mono- or bilayer MoS2. hBN/MoS2/hBN sandwiches are fabricated in a nitrogen-filled glove box using the dry-transfer method43. The workflow of our sample fabrication process is illustrated in Supplementary Fig. 1, with several typical samples shown in Supplementary Fig. 2. Figure 1a shows an artistic view of the fabricated device. Thermal evaporation of Bi/Au (25/30 nm) is carried out after lithography that exposes the 2D window opened in the top hBN. As shown in the optical image of a typical device (Fig. 1b), the 2D-windowed contact (Fig. 1b, dashed box) is visible for each electrical lead. The device is equipped with top and bottom gates, and etch shaped into Hall bars for electrical measurements (the Methods provides more details).

a, Schematic of the fabricated device. b, Optical image of a typical dual-gated bilayer MoS2 device. One of the 2D contact windows is highlighted by black dashed lines. Scale bar, 5 μm. c, Artistic view of the cross-section at the contact (indicated by the purple window in a) of the device under investigation. Bi/Au (25/30 nm) electrodes are deposited through the pre-patterned window in the top hBN. d, Field-effect curves (longitudinal conductivity \({\rho }_{{{xx}}}^{-1}\) versus back-gate voltage Vbg) of sample BS1 measured at different temperatures ranging from 290 K to 300 mK. The inset shows the output characteristics at low drain–source bias obtained at T = 300 mK for a channel length of L = 1 μm for different Vbg values. Notice that the output curves are linear, which is a characteristic of an ohmic contact. e, Summarization of state-of-the-art charge mobilities obtained in TMDs. Field-effect electron mobility in MoS2 obtained in this work is one of the highest among TMD transistors with different layers (1L, 2L and FL represent one layer, two layers and few layers, respectively) using different contact methods. The solid symbols represent the field-effect mobility, and the hollow symbols represent the quantum mobility in various TMD systems reported (Supplementary Figs. 6 and 7). The values of field-effect mobility at certain carrier densities n in our monolayer samples named MS4 and MS6 and bilayer sample named BS1 are plotted in the shaded region, with the arrow pointing to it. The shaded region and the solid lines connecting the data points serve as guides for the eyes. The dashed lines are the fitting curves obtained by applying Matthiessen’s rule for the temperature dependence of a semiconductor, written as \(\frac{1}{\mu (T)}=\frac{1}{{\mu }_{{\rm{imp}}}}+\frac{1}{{\mu }_{\rm{ph}}(T)}\), where μimp is the Coulomb impurity-limited mobility at low temperatures and μph is the one affected by optical phonons64. The fitting errors are propagated and indicated by the error bars.

Figure 1c depicts a schematic of the windowed Bi contact on bilayer MoS2, providing a cross-sectional view of the shaded section shown in Fig. 1a. Clearly, the top surface of MoS2 is contacted by Bi/Au in a 2D manner, which enables the formation of an ohmic contact to the conduction electrons in semiconducting MoS2 down to the base temperature of sub-1 K. As discussed in Supplementary Fig. 3, by applying the transfer length method, the contact resistivity in our devices is estimated to be about 450 Ω μm at T = 1.5 K. In the typical dual-gated samples, the contact areas are doped by only the bottom gate (Fig. 1a,b). Taking sample BS21 as an example, the contact resistance is estimated to be a few hundred ohms by subtracting the four-probe resistance from the two-probe resistance at 1.5 K (Supplementary Fig. 4). Such a contact resistance is sufficiently low to perform quantum transport measurements in the quantum Hall regime. As shown in Supplementary Fig. 4, the contact resistance is maintained below 10 kΩ μm with a carrier density reaching as low as ~2 × 1011 cm−2. This value corresponds to the critical carrier density where the metal–insulator transition takes place in the present system (Supplementary Fig. 5). Field-effect curves (longitudinal conductivity \({\rho }_{{{xx}}}^{-1}\) as a function of back-gate voltage Vbg) recorded in a typical device BS1 at different temperatures ranging from 290 K to 300 mK are shown in Fig. 1d. Evidently, the device exhibits enhanced conductivity as the temperature decreases, and the two-terminal I–V curves remain linear at the base temperature of 300 mK (Fig. 1d, inset), suggesting an ohmic contact to the n-type semiconductor. Figure 1e showcases the state-of-the-art field-effect/Hall mobilities obtained in typical TMD transistors (n type for MoS2 and p type for WSe2) as a function of temperature using different contact methods31,34,44,45,46,47,48,49. A summary of the characteristics of the typical measured samples is listed in Table 1. The field-effect/Hall electron mobility is estimated to be 104−105 cm2 V−1 s−1 in the MoS2 devices in our work, benchmarking some of the best performances at sub-1 K temperatures. It is, thus, worth noting that our contact method may promote future applications of high-electron-mobility transistors devices using MoS2 as a platform, especially at cryogenic temperatures. Notice that the above-mentioned analysis is the so-called field-effect mobility (or relatively similar to Hall mobility) and an examination of the quantum mobility (defined by 1/BOSC, where BOSC is the onset of quantum oscillations). More details on the measurements of Hall and quantum mobility of the current system are provided in Supplementary Figs. 6 and 7.

We now study the dual-gated mapping of the channel resistance of typical MoS2 devices under a finite perpendicular magnetic field B. As shown in Supplementary Fig. 8, the dual-gate field-effect colour maps at B = 12 T and T = 0.3 K are plotted into the parameter space of n and D, with total carrier density n = (CtgVtg + CbgVbg)/e − n0 and applied average electric displacement field D = (CtgVtg − CbgVbg)/2ϵ0 − D0, where Ctg and Cbg are the top- and bottom-gate capacitances per area, respectively. Vtg and Vbg are the top- and bottom-gate voltages, respectively. n0 and D0 are the residual doping and residual displacement fields, respectively. LLs close to the band edges are well developed for both monolayer (Supplementary Fig. 8a) and bilayer (Supplementary Fig. 8b) MoS2. Landau fan scanned at a fixed D = –0.4 V nm–1 for the bilayer sample is shown in Supplementary Fig. 8c. Evidently, the lowest LL can be observed above ~7 T, with the first five LLs indexed by full-degeneracy-lifted integers ν from 1 to 5. Landau fan spectra measured in the same device but at a fixed Vbg = 4.72 V is shown in Fig. 2a. Similar to that in Supplementary Fig. 8c, the first five LLs are highlighted with dashed white lines, which extrapolate at n ≈ 0 cm−2, that is, the band edge. This extremely low carrier density available in our current system allows us to investigate the electronic states below the quantum limit of ν = 1. A line profile of the longitudinal and transverse conductivities (defined as σxx = ρxx/(\({\rho }_{{{xx}}}^{2}\) + \({\rho }_{{{xy}}}^{2}\)) and \({\rho }_{{{xy}}}^{-1}\), where ρxx and ρxy are the longitudinal and transverse resistivities, respectively; a matrix inversion format is used for σxx to ensure better visibility due to the highly resistive ρxx; Supplementary Fig. 9) are plotted in Fig. 2b. The quantization of Hall plateaus for the first five LLs can be seen, with the conductance minima observed in their corresponding σxx. Higher filling fractions become smeared out and level crossing seems to take place as that in the doping range of n > 1.5 × 1012 cm−2 (Fig. 2a, solid black arrows). Notice that the LL crossing can be complicated in multilayered MoS2, as reported previously28,30,36, but is nevertheless captured to a large extent by our calculations (Supplementary Fig. 10). In addition, Fig. 2a displays another set of LL-like features (right side of the green dashed line), which originates from the asymmetric electrical contact to the bilayer MoS2, and will be discussed later.

a, Landau fan diagram (longitudinal resistance Rxx as a function of B and n) of sample BS1 measured at T = 300 mK and Vbg = 4.72 V, in the magnetic-field range from 0 T to 12 T. The white dashed lines are guides for the eyes, and their crossing point indicates the band edge. The black arrows indicate the the emergence of LL crossings from different flavours. The green dashed line marks the boundary between the two sets of LLs from the top and bottom layers. b, Line profile of σxx = ρxx/(\({\rho }_{{{xx}}}^{2}\) + \({\rho }_{{{xy}}}^{2}\)) and transverse conductivity \({\rho }_{{{xy}}}^{-1}\) at B = 12 T and T = 0.3 K, extracted from a, where σxx and ρxx represent the longitudinal conductivity and resistivity, respectively. The bottom axis corresponds to the top-gate voltages. c, Calculated LLs as a function of energy and magnetic field (Methods). N represents the Landau index. d, Cartoon illustration of the calculated several first LLs at the K and K' valleys at B = 40 T and Ez = 60 meV. For simplicity, only part of the Brillouin zone is shown. e, Channel resistance of an asymmetrically contacted bilayer MoS2 sample BS9 as a function of Vbg and Vtg. The dashed lines are guides for the eyes.

We now consider the LLs of bilayer MoS2 devices. Their conduction band minima mainly consist of the \({d}_{{z}^{2}}\) electronic orbitals from both layers and are located at the K and K' corners of the Brillouin zone. At these points, the interlayer hybridization carries non-zero orbital angular momentum and cannot couple the rotationally invariant orbitals of opposite layers. The layer index can, therefore, be taken as a good quantum number for electrons in the conduction band. Our calculations (Supplementary Note 1) suggest that a finite vertical electric field Ez can split the layer index in the band structure of LLs, with the first four (lowest and first excited) LLs in the conduction band being layer-valley locked. A calculated band structure of LLs at a layer polarization energy of Ez = 60 meV is given in Fig. 2c, with more details of the calculations provided in the Supplementary Information. To visualize the calculated results, the schematic of the first nine LLs (calculated for B = 40 T and Ez = 60 meV) at the K and K' valleys are illustrated in Fig. 2d, which largely differs from the p-doped scenario observed in monolayer WSe2 (ref. 37).

One key feature of our study is that the Bi/Au electrodes are ohmically contacted with only one top surface of the bilayer MoS2, leading to an asymmetrically contacted configuration. Unlike the monolayer case (Supplementary Fig. 8a), this provides a specific manner to probe the LLs in each of the TMD layers, and enables the observation of LL crossing as shown in the dual-gated scan in Fig. 2e and D–n scan in Supplementary Fig. 8b. When Ez is biased to be sufficiently negative (for example, D < –0.3 V nm–1 and n < 5 × 1012 cm−2), the system effectively behaves as a monolayer, with the N = 0 and N = 1 LLs being in the same K valley and on the top layer but carrying opposite spin polarizations. In contrast, when Ez is biased to be less negative or even positive, the lowest LLs generically feature a non-zero amplitude on the bottom layer. Due to the asymmetrically contacted configuration, our measurements only probe a fraction of the total conductivity, leading to interference patterns that perfectly describe the anti-diagonal striped features (Fig. 2e, white dashed lines), as discussed in more detail by comparing the experimental and simulated results (Supplementary Figs. 11–14).

In the following, by applying a vertical magnetic field of up to 34 T at T = 300 mK, we examine the QHE in bilayer MoS2 down to the quantum limit of ν ≤ 1. Figure 3a displays the σxx = ρxx/(\({\rho }_{{{xx}}}^{2}\) + \({\rho }_{{{xy}}}^{2}\)) and \({\rho }_{{{xy}}}^{-1}\) as a function of Vtg. The quantized plateaus at fractions of 4/5 and 2/5 are observed in the lowest LL, with a clear minima associated in the longitudinal conductance for the former and an emerging one for the latter. This makes semiconducting TMDs another platform to investigate—using electrical transport—FQH states after a few known 2DEG systems of quantum wells and graphene. By varying Vtg at different values of fixed Vbg (that is, different D values), it is noticed that the FQH states have little dependence on the displacement field in the studied gate-voltage range (Fig. 3b). This hints that—in agreement with earlier arguments—the observed FQH states are layer polarized.

a, σxx = ρxx/(\({\rho }_{{{xx}}}^{2}\) + \({\rho }_{{{xy}}}^{2}\)) and \({\rho }_{{{xy}}}^{-1}\) as a function of Vtg. Vbg is fixed at 5.16 V. The black solid and dashed lines mark the quantized values of the transverse conductivities. b, \({\rho }_{{{xy}}}^{-1}\) as a function of Vtg, measured for different Vbg values. c, Colour map of longitudinal conductivity measured in the filling fraction range of 0–2 and magnetic-field range of 18–34 T at T = 300 mK. d, Raw data of Rxx measured at different temperatures. e, Raw data of \({\rho }_{{{xy}}}^{-1}\) measured at different magnetic fields. Vbg is fixed at 5.16 V in d and e, and data in a–e are obtained from sample BS10. f, Measured gaps Δ of the first three LLs in the range of B = 10–14 T in a typical bilayer MoS2 device (sample BS7). The error bars denote the statistical errors in fitting the data using the Arrhenius law.

We then focus on the lowest LL with the magnetic field scanned from 34 T down to 18 T (Fig. 3c). Fractionally quantized transverse conductance develops above ~22 T at filling fractions of 4/5 and 2/5, with a clear dip in Rxx at 4/5 and an emerging one at 2/5 hidden by a still large (contact) resistivity. Raw data of Rxx at different temperatures and \({\rho }_{{{xy}}}^{-1}\) values at different magnetic fields are plotted in Fig. 3d,e, respectively. As a function of temperature, we observe that the kink in Rxx at a filling fraction of ν = 2/5 rapidly vanishes near 2 K, and that at ν = 4/5 gradually disappears for T up to 8 K. This means that the FQH gaps are in the order of a few kelvins at 34 T for the typical bilayer MoS2 samples discussed in this work. Meanwhile, the plateaus of \({\rho }_{{{xy}}}^{-1}\) (Fig. 3e) get smeared out below ~26 T, which is the same as that shown in Fig. 3c. It is worth mentioning that the FQH states with a denominator of 5 were reproduced in another typical bilayer MoS2 sample (BS6; Supplementary Fig. 15). We further performed temperature dependence measurements of the quantum oscillations of the LLs (Supplementary Fig. 16), and the gaps of the first three LLs in the range from B = 10 T to B = 14 T are extracted (Fig. 3f). Indeed, the gap size of the fractionally charged quasiparticles are smaller compared with those estimated at 34 T for integer filling (Fig. 3f).

The FQH state at 4/5 in MoS2 bilayers, which is clearly revealed in our transport measurements, falls in the universality class of the p-h conjugate of Laughlin’s 1/5 state1. This state hosts Abelian fractionally charged quasiparticles, but differs from the fraction of 1/3, which is usually reported to have the largest gap in Laughlin’s hierarchy. Our theoretical calculations show that the N = 0 LL in MoS2 bilayers is spanned by radial wavefunctions identical to those observed in standard semiconducting 2DEGs, and experiences a similar Zeeman effect. We, thus, expect similar interacting phases in both systems, including FQH states at 1/3 (or its p-h conjugate 2/3). Deviations from this behaviour most probably come from the non-universal characteristics of the samples, such as disorder, the effective dielectric environment induced by nearby gates or the axial profile of the wavefunctions related to the finite sample thickness.

We were able to theoretically model the latter two, that is, the effects of gates and finite thickness, as a screening of the Coulomb potential at long and short distances, respectively (Methods). In the N = 0 LL, this screened Coulomb potential is fully specified by its Haldane pseudopotential Vm, describing the highest energy that two particles with relative angular momentum m can have at the lowest LL (only odd m has an effect on the fermionic states). In Fig. 4a, we plot these pseudopotentials at B = 22 T for which the 4/5 fraction appears (Fig. 3c) for an infinitely thin 2DEG, a monolayer MoS2 and a bilayer MoS2 using a realistic gate distance of d = 35 nm. The presence of gates yields an exponential decay of Vm at large m, much faster than the standard algebraic decay of the unscreened Coulomb potential. As a result, the few lowest pseudopotentials should already capture the physics of our gate-screened samples, and we restrict our attention to the first three ones, namely, V1, V3 and V5, hereafter.

a, Pseudopotential for a gate-screened Coulomb potential in an infinitely thin 2DEG, a monolayer MoS2 and a bilayer MoS2 at B = 22 T and a gate distance of d = 35 nm (Methods). Vm is the Haldane pseudopotential that describes the maximum energy that two particles with relative angular momentum m can have at the lowest LL. We show the pseudopotentials of an unscreened Coulomb potential for comparison. b, Square overlap between the exact diagonalization ground state |ΨGS〉 and Laughlin’s wavefunction |Ψ1/3〉 numerically computed on a torus with N = 13 electrons at a filling fraction of 1/3 using three non-zero pseudopotentials Vm=1,3,5. The coloured dots show the values extracted from a. c, Same data as b, but at a filling fraction of 1/5 using N = 10 particles. The cartoons in b and c illustrate the Laughlin phases of 1/3 and 1/5, respectively, with the black dashed lines highlighting the boundary between a Laughlin-like phase (non-zero overlaps, that is, the blue-coloured area) and a phase in which the Laughlin state is absent (vanishing overlaps, that is, the white-coloured area).

Figure 4a also points out that the finite thickness of the sample increases Vm>1 compared with V1. Due to their exact clustering properties, FQH states are sensitive to such changes in the short-range interaction physics. For instance, the ground state of an FQH system at a filling of 1/3 has perfect overlap with the Laughlin state when V1 > 0 and all Vm>1 = 0, but this overlap decreases as V3 and V5 grow50 (in Fig. 4b, we plot this overlap obtained for a finite torus using exact diagonalization (Methods)). For V3 ≳ 0.5V1, this overlap vanishes, signalling a phase transition and a ground state that no longer belongs to the Laughlin universality class. The increase in Vm>1/V1 in bilayer MoS2 (Fig. 4a) drives the system to the edge of the phase that has non-zero overlap with Laughlin’s 1/3 state, hinting at a potential absence of the FQH state at 2/3 in our experiments. Consistent with our experiments, the same calculation at a filling of 1/5 (Fig. 4c) shows that the stability of fractions 1/5, or its p-h conjugate 4/5, barely changes as we go from an infinitely thin 2DEG to a realistic bilayer MoS2. In conclusion, although all the microscopic details of our samples are not necessarily captured by our calculations, they reveal a clear trend that rationalizes the absence of the 2/3 FQH state due to the internal screening of the Coulomb potential. It is noticed that for the typical monolayer MoS2 devices, Laughlin’s 1/3 states are also absent, though the hints of FQH states in monolayer samples exhibit less well-defined quantization at a filling of 1/5 compared with the bilayer ones (Supplementary Fig. 17). The 2/5 and 1/3 FQH states belong to the same series, and are stabilized by a similar pseudopotential sequence (the overlap square between the ground state for the Coulomb interaction and the V1–V3–V5 pseudopotentials at a filling of 2/5 is shown in Supplementary Fig. 18). We would, therefore, expect that the 2/5 state is less stable in the presence of dielectric screening. As hinted by the breaking of the p-h symmetry between 2/5 and 3/5 (Fig. 3), the effects of spin degeneracy might be important at these fractions. The observed sensitivity of the FQH states to the dielectric environment imposed by nearby gates and by the vertical profile of the heterostructure may allow the engineering of a richer phenomenology of FQH states in TMD-based heterostructures51. Recently, transport evidence of FQH states in p-type WSe2 has also been reported52.

We have reported ohmic contacts to n-type bilayer MoS2 with an extremely low carrier density. We measured the electrical transport characteristics under high magnetic fields of up to 34 T and a base temperature of 300 mK. Our device fabrication technique leads to field-effect mobilities of over 100,000 cm2 V−1 s−1 and quantum mobilities of over 3,000 cm2 V−1 s−1 in the conduction band in monolayer MoS2 at millikelvin temperatures. Under the application of finite Ez, the system behaves as a monolayer MoS2 whose first four LLs in the conduction band are layer-valley locked, and the spin degeneracy is fully lifted by the Zeeman shift. The asymmetric ohmic contact to the high-mobility bilayer semiconducting channel leads to the observation of fractionally quantized Hall plateaus of 2/5 and 4/5 in the lowest LL above 26 T, which exhibit negligible tunability against Ez, and are at the energy scale of sub-1 K, in agreement with theoretical predictions. The high mobility in our devices could potentially lead to TMD twistronics, Coulomb drag devices and low-temperature high-electron-mobility transistors. Our work could also help in the development of low-density transport experiments on TMD heterostructures, which will be required to harness the full potential of topological FQH phases53,54,55,56.

van der Waals few layers of the hBN/MoS2/hBN sandwich were obtained by mechanically exfoliating high-quality bulk crystals. The vertical assembly of the van der Waals layered compounds were fabricated by using the dry-transfer method in a nitrogen-filled glove box. Hall bars of hBN/MoS2/hBN devices were achieved by plasma etching. Electron-beam lithography was performed using a ZEISS Sigma 300 scanning electron microscope with a Raith Elphy Quantum graphic writer. The top gates and contacting electrodes were fabricated with a thermal evaporator, with typical Bi/Au thicknesses of ~25/30 nm.

A resistive magnet with a maximum magnetic field of 34 T and a base temperature of 300 mK was used. The facility was equipped with a water cooling system, and is maintained by the Steady High Magnetic Field Facility, High Magnetic Field Laboratory, Chinese Academy of Sciences.

The high precision of the current measurements of the devices were achieved using a Cascade M150 probe station at room temperature, with an Agilent B1500A semiconductor device parameter analyser. The gate voltages on the as-prepared Hall bars were maintained by a Keithley 2400 source meter. During the measurements, the TMD layer was fed with an a.c. Ibias of about 10 nA. Low-frequency lock-in four-probe measurements were used throughout the transport measurements under a high magnetic field and at low temperatures.

The LL diagrams for monolayer and bilayer MoS2 are obtained by applying minimal coupling on the \(\overrightarrow{k}\cdot \overrightarrow{p}\) Hamiltonian for conduction electrons derived from ab initio simulations. The calculation follows from another formalism57, which is reviewed in the Supplementary Information for completeness. We use values of the gap, effective mass, Ising spin–orbit coupling strength in the conduction and valence bands, spin and valley g-factors, and uniform interlayer hybridization provided by first-principles calculations58,59 and rounded off to the millielectronvolt scale.

To describe the interacting phases of our system, we have used the Coulomb potential V(q) = 2πe2tanh(qd)/[εq(1 + nαqδ)], which faithfully captures the long-distance (small-q) and short-distance (large-q) asymptotics imposed by gate screening and the out-of-plane structure of MoS2 multilayers, respectively. More precisely, this Coulomb potential reproduces (1) V(q→0) = 2πe2tanh(qd)/(εq) expected in the presence of gates symmetrically placed at a distance d around the sample, where ε ≃ 5 is the relative dielectric constant of the surrounding hBN; and (2) a Rytova–Keldysh form V(q→∞) = 2πe2/[εq(1 + nαqδ)] known to be relevant for TMDs60, where n is the number of layers, δ = 0.65 nm is the typical interlayer spacing and α is a material-specific constant. Note that n = 0 describes an infinitely thin 2DEG with no short-range screening.

Interactions projected to the N = 0 LL are accounted for by using Haldane’s pseudopotentials61, defined on the plane as \({V}_{m}={(2\uppi {\ell }_{B}^{2})}^{-1}\mathop{\int}\nolimits_{0}^{\infty }\tilde{q}{\rm{d}}\tilde{q}V(\tilde{q}/{\ell }_{B}){L}_{m}({\tilde{q}}^{2}){\rm{e}}^{-{\tilde{q}}^{2}}\), where ℓB is the magnetic length, Lm is the mth Laguerre polynomial and V(q) is the Coulomb potential. This should be understood as an angular Fourier transformation of the Coulomb potential after projection to the N = 0 LL, such that Vm energetically penalizes particles with relative angular momentum m. Note that fermions are only sensitive to odd-m pseudopotentials. The FQH state at a fraction of 4/5 is observed in the magnetic-field range of 22–34 T, for which the magnetic length lies within ℓB ≃ 4.5–6 nm. Focusing on B = 22 T (Fig. 4a), the ~35-nm-thick boron nitride layers between our sample and gates yields d/ℓB ≃ 6, whereas the coefficient α = 15/ε (ref. 62) leads to αδ/ℓB ≃ 0.4. Both these estimates have been rounded to their first significant digit to highlight their phenomenological nature.

The phase diagrams shown in Fig. 4b,c were obtained by the exact numerical diagonalization of the interacting problem on the torus with a square aspect ratio including the three first potentials relevant for fermions, namely, V1, V3 and V5. We set V1 = 1 to fix the energy scale. The Hilbert-space dimensions for N = 13 particles at a filling of 1/3 and N = 10 particles at a filling of 1/5 are 16,020,564 and 20,544,878, respectively, in the momentum sectors of their ground state.

The data that support the findings of this study are available via Zenodo at https://doi.org/10.5281/zenodo.13765107 (ref. 63).

The exact diagonalization results were obtained using DiagHam, an open-source publicly available software package available at https://www.nick-ux.org/diagham.

Laughlin, R. B. Anomalous quantum Hall effect: an incompressible quantum fluid with fractionally charged excitations. Phys. Rev. Lett. 50, 1395–1398 (1983).

Google Scholar

Eisenstein, J. P. & Stormer, H. L. The fractional quantum Hall effect. Science 248, 1510–1516 (1990).

Google Scholar

Stormer, H. L., Tsui, D. C. & Gossard, A. C. The fractional quantum Hall effect. Rev. Mod. Phys. 71, S298 (1999).

MathSciNet Google Scholar

Feldman, D. E. & Halperin, B. I. Fractional charge and fractional statistics in the quantum Hall effects. Rep. Prog. Phys. 84, 076501 (2021).

MathSciNet Google Scholar

Jain, J. K. Composite-fermion approach for the fractional quantum Hall effect. Phys. Rev. Lett. 63, 199 (1989).

Google Scholar

Pan, W. et al. Fractional quantum Hall effect of composite fermions. Phys. Rev. Lett. 90, 016801 (2003).

Google Scholar

Son, D. T. The Dirac composite fermion of the fractional quantum Hall effect. Annu. Rev. Condens. Matter Phys. 9, 397–411 (2018).

Google Scholar

Moore, G. & Read, N. Nonabelions in the fractional quantum Hall effect. Nucl. Phys. B 360, 362–396 (1991).

MathSciNet Google Scholar

Stern, A. Anyons and the quantum Hall effect—a pedagogical review. Ann. Phys. 323, 204–249 (2008).

MathSciNet Google Scholar

Lin, X., Du, R. & Xie, X. Recent experimental progress of fractional quantum Hall effect: 5/2 filling state and graphene. Natl Sci. Rev. 1, 564–579 (2014).

Google Scholar

Clarke, D. J., Alicea, J. & Shtengel, K. Exotic non-abelian anyons from conventional fractional quantum Hall states. Nat. Commun. 4, 1348 (2013).

Google Scholar

Vaezi, A. Superconducting analogue of the parafermion fractional quantum Hall states. Phys. Rev. X 4, 031009 (2014).

Google Scholar

Alicea, J. & Fendley, P. Topological phases with parafermions: theory and blueprints. Annu. Rev. Condens. Matter Phys. 7, 119–139 (2016).

Google Scholar

Tsui, D. C., Stormer, H. L. & Gossard, A. C. Two-dimensional magnetotransport in the extreme quantum limit. Phys. Rev. Lett. 48, 1559 (1982).

Google Scholar

Tsukazaki, A. et al. Observation of the fractional quantum Hall effect in an oxide. Nat. Mater. 9, 889–893 (2010).

Google Scholar

Falson, J. & Kawasaki, M. A review of the quantum Hall effects in MgZnO/ZnO heterostructures. Rep. Prog. Phys. 81, 056501 (2018).

MathSciNet Google Scholar

Du, X., Skachko, I., Duerr, F., Luican, A. & Andrei, E. Y. Fractional quantum Hall effect and insulating phase of Dirac electrons in graphene. Nature 462, 192–195 (2009).

Google Scholar

Dean, C. R. et al. Multicomponent fractional quantum Hall effect in graphene. Nat. Phys. 7, 693–696 (2011).

Google Scholar

Feldman, B. E., Krauss, B., Smet, J. H. & Yacoby, A. Unconventional sequence of fractional quantum Hall states in suspended graphene. Science 337, 1196 (2012).

Google Scholar

Maher, P. Tunable fractional quantum Hall phases in bilayer graphene. Science 345, 61–64 (2014).

Google Scholar

Kou, A. et al. Electron-hole asymmetric integer and fractional quantum Hall effect in bilayer graphene. Science 345, 55–57 (2014).

Google Scholar

Amet, F. et al. Composite fermions and broken symmetries in graphene. Nat. Commun. 6, 5838 (2014).

Google Scholar

Li, J. I. A. et al. Even denominator fractional quantum Hall states in bilayer graphene. Science 358, 648–652 (2017).

Google Scholar

Zibrov, A. A. et al. Tunable interacting composite fermion phases in a half-filled bilayer-graphene Landau level. Nature 549, 360–364 (2017).

Google Scholar

Kim, Y. et al. Even denominator fractional quantum Hall states in higher landau levels of graphene. Nat. Phys 15, 154–158 (2019).

Google Scholar

Zibrov, A. A. et al. Even-denominator fractional quantum Hall states at an isospin transition in monolayer graphene. Nat. Phys. 14, 930–935 (2018).

Google Scholar

Huang, K. et al. Valley isospin controlled fractional quantum Hall states in bilayer graphene. Phys. Rev. X 12, 031019 (2022).

Google Scholar

Li, X., Zhang, F. & Niu, Q. Unconventional quantum Hall effect and tunable spin Hall effect in Dirac materials: application to an isolated MoS2 trilayer. Phys. Rev. Lett. 110, 066803 (2013).

Google Scholar

Pisoni, R. et al. Interactions and magnetotransport through spin-valley coupled Landau levels in monolayer MoS2. Phys. Rev. Lett. 121, 247701 (2018).

Google Scholar

Lin, J. et al. Determining interaction enhanced valley susceptibility in spin-valley-locked MoS2. Nano Lett. 19, 1736–1742 (2019).

Google Scholar

Xu, S. et al. Odd-integer quantum Hall states and giant spin susceptibility in p-type few-layer WSe2. Phys. Rev. Lett. 118, 067702 (2017).

Google Scholar

Wang, Z., Shan, J. & Mak, K. F. Valley- and spin-polarized Landau levels in monolayer WSe2. Nat. Nanotechnol. 12, 144–149 (2017).

Google Scholar

Movva, H. C. P. et al. Density-dependent quantum Hall states and Zeeman splitting in monolayer and bilayer WSe2. Phys. Rev. Lett. 118, 247701 (2017).

Google Scholar

Wu, Z. et al. Even-odd layer-dependent magnetotransport of high-mobility Q-valley electrons in transition metal disulfides. Nat. Commun. 7, 12955 (2016).

Google Scholar

Gustafsson, M. V. et al. Ambipolar Landau levels and strong band-selective carrier interactions in monolayer WSe2. Nat. Mater. 17, 411–415 (2018).

Google Scholar

Masseroni, M. et al. Electron transport in dual-gated three-layer MoS2. Phys. Rev. Res. 3, 023047 (2021).

Google Scholar

Shi, Q. et al. Odd- and even-denominator fractional quantum Hall states in monolayer WSe2. Nat. Nanotechnol. 15, 569–573 (2020).

Google Scholar

Kappera, R. et al. Phase-engineered low-resistance contacts for ultrathin MoS2 transistors. Nat. Mater. 13, 1128–1134 (2014).

Google Scholar

Cai, X. et al. Bridging the gap between atomically thin semiconductors and metal leads. Nat. Commun. 13, 1777 (2022).

Google Scholar

Yuan Liu, Y. et al. Approaching the Schottky-Mott limit in van der Waals metal-semiconductor junctions. Nature 557, 696–700 (2018).

Google Scholar

Li, W. et al. Approaching the quantum limit in two-dimensional semiconductor contacts. Nature 613, 274–279 (2023).

Google Scholar

Shen, P.-C. et al. Ultralow contact resistance between semimetal and monolayer semiconductors. Nature 593, 211–217 (2021).

Google Scholar

Wang, L. et al. One-dimensional electrical contact to a two-dimensional material. Science 342, 614–617 (2013).

Google Scholar

Jung, Y. et al. Transferred via contacts as a platform for ideal two-dimensional transistors. Nat. Electron. 2, 187–194 (2019).

Google Scholar

Pisoni, R. et al. Absence of interlayer tunnel coupling of K-valley electrons in bilayer MoS2. Phys. Rev. Lett. 123, 117702 (2019).

Google Scholar

Chuang, H.-J. et al. Low-resistance 2D/2D ohmic contacts: a universal approach to high-performance WSe2, MoS2, and MoSe2 transistors. Nano Lett. 16, 1896–1902 (2016).

Google Scholar

Movva, H. C. et al. High-mobility holes in dual-gated WSe2 field-effect transistors. ACS Nano 9, 10402–10410 (2015).

Google Scholar

Chuang, H.-J. et al. High mobility WSe2 p- and n-type field-effect transistors contacted by highly doped graphene for low-resistance contacts. Nano Lett. 14, 3594–3601 (2014).

Google Scholar

Cui, X. et al. Low-temperature ohmic contact to monolayer MoS2 by van der Waals bonded Co/h-BN electrodes. Nano Lett. 17, 4781–4786 (2017).

Google Scholar

Regnault, N. et al. Evidence of a fractional quantum Hall nematic phase in a microscopic model. Phys. Rev. B 96, 035150 (2017).

Google Scholar

Papić, Z. et al. Numerical studies of the fractional quantum Hall effect in systems with tunable interactions. J. Phys.: Conf. Ser. 402, 012020 (2012).

Pack, J. et al. Charge-transfer contacts for the measurement of correlated states in high-mobility WSe2. Nat. Nanotechnol. 19, 948–954 (2024).

Google Scholar

Cai, J. et al. Signatures of fractional quantum anomalous Hall states in twisted MoTe2. Nature 622, 63–68 (2023).

Google Scholar

Zeng, Y. et al. Thermodynamic evidence of fractional Chern insulator in moiré MoTe2. Nature 622, 69–73 (2023).

Google Scholar

Park, H. et al. Observation of fractionally quantized anomalous Hall effect. Nature 622, 74–79 (2023).

Google Scholar

Xu, F. et al. Observation of integer and fractional quantum anomalous Hall effects in twisted bilayer MoTe2. Phys. Rev. X 13, 031037 (2023).

Google Scholar

Zubair, M. et al. Quantum magnetotransport in bilayer MoS2: influence of perpendicular electric field. Phys. Rev. B 96, 045405 (2017).

Google Scholar

Kormányos, A. et al. Spin-orbit coupling, quantum dots, and qubits in monolayer transition metal dichalcogenides. Phys. Rev. X 4, 011034 (2014).

Google Scholar

Fang, S. et al. Ab initio tight-binding Hamiltonian for transition metal dichalcogenides. Phys. Rev. B 92, 205108 (2015).

Google Scholar

Meckbach, L. et al. Influence of the effective layer thickness on the ground-state and excitonic properties of transition-metal dichalcogenide systems. Phys. Rev. B 97, 035425 (2018).

Google Scholar

Girvin, S. M. The quantum Hall effect: novel excitations and broken symmetries. In Aspects topologiques de la physique en basse dimension 53–175 (Springer, 2002).

Van Tuan, D. et al. Coulomb interaction in monolayer transition-metal dichalcogenides. Phys. Rev. B 98, 125308 (2018).

Google Scholar

Zhao, S. Fractional quantum Hall phases in high mobility n-type molybdenum disulfide transistors. Zenodo https://doi.org/10.5281/zenodo.13765107 (2024).

Ma, N. & Jena, D. Charge scattering and mobility in atomically thin semiconductors. Phys. Rev. X 4, 011043 (2014).

Google Scholar

Download references

N.R. and V.C. are grateful to Z. Papic for fruitful discussions. This work is supported by the National Key R&D Program of China (nos. 2022YFA1203903, 2023YFF1500600 and 2021YFA1400100) and the National Natural Science Foundation of China (NSFC) (grant nos. 92265203, 12104462, 11974357, U1932151, 12204287 and 11974027). Z.V.H. acknowledges support from the Fund for Shanxi ‘1331 Project’ Key Subjects Construction, and the Innovation Program for Quantum Science and Technology (grant no. 2021ZD0302003). N.W. acknowledges support from the Hong Kong Research Grants Council (project no. AoE/P-701/20). J. Lu thanks the Beijing Natural Science Foundation (grant no. Z190011). K.W. and T.T. acknowledge support from the JSPS KAKENHI (grant nos. 20H00354, 21H05233 and 23H02052) and World Premier International Research Center Initiative (WPI), MEXT, Japan. B.S. and N.R. acknowledge support from the QuantERA II Programme that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 101017733. N.R. was also supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no. 101020833). The Flatiron Institute is a division of the Simons Foundation. B.S. was supported by the European Union’s Horizon 2020 research and innovation programme under the ERC grant SUPERGRAPH no. 866365. A portion of this work was carried out at the Synergetic Extreme Condition User Facility (SECUF). A portion of this work was performed on the Steady High Magnetic Field Facilities, High Magnetic Field Laboratory, Chinese Academy of Sciences, and supported by the High Magnetic Field Laboratory of Anhui Province.

These authors contributed equally: Siwen Zhao, Jinqiang Huang, Valentin Crépel.

Liaoning Academy of Materials, Shenyang, China

Siwen Zhao, Hanwen Wang & Zheng Vitto Han

Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang, China

Jinqiang Huang

School of Material Science and Engineering, University of Science and Technology of China, Anhui, China

Jinqiang Huang

Center for Computational Quantum Physics, Flatiron Institute, New York, NY, USA

Valentin Crépel

State Key Laboratory of Quantum Optics and Quantum Optics Devices, Institute of Optoelectronics, Shanxi University, Taiyuan, China

Zhiren Xiong, Xingguang Wu, Tongyao Zhang, Jing Zhang & Zheng Vitto Han

Collaborative Innovation Center of Extreme Optics, Shanxi University, Taiyuan, China

Zhiren Xiong, Xingguang Wu, Tongyao Zhang, Jing Zhang & Zheng Vitto Han

State Key Laboratory for Mesoscopic Physics, School of Physics, Peking University, Beijing, China

Xiangyan Han & Jianming Lu

Anhui Key Laboratory of Low-Energy Quantum Materials and Devices, High Magnetic Field Laboratory, HFIPS, Chinese Academy of Sciences, Hefei, China

Zhengyu Li, Chuanying Xi, Senyang Pan, Zhaosheng Wang & Guangli Kuang

Beijing National Laboratory for Condensed Matter Physics and Institute of Physics, Chinese Academy of Sciences, Beijing, China

Jun Luo, Qinxin Shen, Jie Yang & Rui Zhou

School of Physical Sciences, University of Chinese Academy of Sciences, Beijing, China

Qinxin Shen

Research Center for Electronic and Optical Materials, National Institute for Materials Science, Tsukuba, Japan

Kenji Watanabe

Research Center for Materials Nanoarchitectonics, National Institute for Materials Science, Tsukuba, Japan

Takashi Taniguchi

Université Grenoble Alpes, CNRS, Grenoble INP, Institut Néel, Grenoble, France

Benjamin Sacépé

Department of Physics and the Center for 1D/2D Quantum Materials, the Hong Kong University of Science and Technology, Hong Kong, China

Ning Wang

Laboratoire de Physique de L’Ecole normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris-Diderot, Paris, France

Nicolas Regnault

Department of Physics, Columbia University, New York, NY, USA

Nicolas Regnault

You can also search for this author in PubMed Google Scholar

Z.V.H. and S.Z. conceived the experiment. Z.V.H., N.W., J. Lu, N.R. and J.Z. supervised the overall project. S.Z., J.H., X.W. and X.H. carried out the device fabrications. S.Z. and J.H. carried out all the electrical transport measurements. Z.L., C.X., S.P. and Z.W. assisted with the measurements at high magnetic fields, with participation from H.W. and T.Z. K.W. and T.T. provided high-quality hBN bulk crystals. Z.V.H., S.Z., J. Lu, N.W., J.Z., B.S., V.C. and N.R. analysed the experimental data. V.C. and N.R. conducted the theoretical modelling and performed first-principles calculations. J. Luo, Q.S., J.Y. and R.Z. helped with the measurements under high magnetic fields in the Beijing Synergetic Extreme Condition User Facility (SECUF). The manuscript was written by Z.V.H., S.Z. and V.C., with discussion and inputs from all authors.

Correspondence to Jing Zhang, Ning Wang, Jianming Lu, Nicolas Regnault or Zheng Vitto Han.

The authors declare no competing interests.

Nature Electronics thanks Wei Jiang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Figs. 1–18 and Note 1.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

Zhao, S., Huang, J., Crépel, V. et al. Fractional quantum Hall phases in high-mobility n-type molybdenum disulfide transistors. Nat Electron (2024). https://doi.org/10.1038/s41928-024-01274-1

Download citation

Received: 23 November 2023

Accepted: 01 October 2024

Published: 30 October 2024

DOI: https://doi.org/10.1038/s41928-024-01274-1

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative