# Fileset

[s41699-025-00551-7.pdf](https://mdr.nims.go.jp/filesets/8f7ca5aa-07b0-4c02-b711-26a9fc553f01/download)

## Creator

Randy M. Sterbentz, Bogyeom Kim, Anayeli Flores-Garibay, Kristine L. Haley, Nicholas T. Pereira, [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), Joshua O. Island

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Gating monolayer and bilayer graphene with a two-dimensional semiconductor](https://mdr.nims.go.jp/datasets/bb633331-ebbf-40c8-ae33-09690fb68003)

## Fulltext

Gating monolayer and bilayer graphene with a two-dimensional semiconductornpj | 2D materials and applications ArticlePublished in partnership with FCT NOVA with the support of E-MRShttps://doi.org/10.1038/s41699-025-00551-7Gating monolayer and bilayer graphenewith a two-dimensional semiconductorCheck for updatesRandy M. Sterbentz1, Bogyeom Kim1, Anayeli Flores-Garibay1, Kristine L. Haley1, Nicholas T. Pereira1,Kenji Watanabe2, Takashi Taniguchi3 & Joshua O. Island1Metals are commonly used as electrostatic gates in devices due to their abundant charge carrierdensities that are necessary for efficient charging and discharging. A semiconducting gate can bebeneficial for certain fabrication processes, in low light conditions, and for specific gating properties.We determine the effectiveness and limitations of a semiconducting gate in graphene and bilayergraphene devices. Using the semiconducting transitionmetal dichalcogenidesmolybdenumdisulfide(MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), and tungsten diselenide (WSe2),we show that two-dimensional semiconductors can be used to suitably gate the graphene devicesunder appropriate operating conditions. For single-gated devices, semiconducting gates arecomparable to metallic gates below liquid helium temperatures but include resistivity featuresresulting from gate voltage clamping of the semiconductor. In dual-gated devices, we pin down theparameter range of effective operation and find that the semiconducting depletion regime results inclamping and hysteresis from defect-state charge trapping.Electrostatic gating in graphene devices has evolved over the years. Theearliest devices employedhighlydoped silicon substrates as global backgatesor evaporated metals for local top and bottom gates1–3. Recently, few-layergraphene has emerged as an ideal gate material for its atomically flat andchemically inert character. This leads to an overall reduction in chargeinhomogeneity and observation of intricate correlated states when directlycomparedwithdepositedmetal gates4. There are situations, though inwhicha semiconducting gate may be more appropriate. Semiconducting gatesemployed in high-mobility transistors provide over-voltage protection andlower leakage current when compared with metallic gates5. In novel high-frequency, on-chip terahertz measurements of graphene, a semiconductinggate is used to avoid absorption of the probing field6,7. Semiconducting gatescould alsoplay an important role in sensitivephotodetectors andbolometersrequiringminimal absorption or reflection froma top gate electrode8,9. Theycould boost the already impressive figures of merit for dual-gated bilayergraphene bolometers that take advantage of intrinsic bulk response byminimizing top gate interference thereby further increasing sensitivity8. Theincreased resistance of a semiconducting top gate would also be advanta-geous in high-bandwidth graphene photodetectors by reducing the RC timeconstant derived from the gate capacitance9. Furthermore, in comparisonwith semi-transparent nichrome or zinc oxide gates that are grown ordeposited10,11, 2D material stacks provide sharp, charge-homogeneousinterfaces. These examples highlight the novelty of using a semiconductinggate but it has not been shown, given the depletion characteristics of asemiconductor, for what conditions electrostatic gating should functionproperly in a graphene device.In a simple parallel-plate capacitor configuration with a metal and atwo-dimensional electron gas (like graphene), applying a gate voltage (VG)results in a change in the chemical potential (μ) in the graphene and theelectrostatic potential (ϕ) between the two layers: eVG = eϕ+ μ, where ϕ isdetermined by the geometric capacitance, ϕ = ne/CG. If a semiconductorreplaces the metal as a gate, the reduction in overall charge density andpresence of an electronic band gap will alter the gate response. As the Fermilevel is driven into the band gap of the semiconductor while sweeping thegate contact voltage, the effective potential between the gate and the gra-phene channel will remain unchanged due to a lack of charge carriers in thegate itself. This voltage clamping effect of the gate will have meaningfulconsequences on the transport characteristics of a graphene device. More-over, the temperature of the system becomes relevant as the thresholdvoltage in the gate changes and the thermal energy of carriers varies. Thisimplies that the effect of sweeping beyond the threshold voltage of thesemiconducting gate should be more pronounced at lower temperatures.Here, we investigate the viability of using 2D transition metal dichal-cogenides (MoS2.MoSe2,WS2,WSe2) as an electrostatic gate formonolayerand bilayer graphene (BLG) devices. MoS2 is highlighted in the main textbecause it has the smallest difference in its electron affinity and graphene’s1Department of Physics and Astronomy, University of Nevada Las Vegas, Las Vegas, NV, USA. 2Research Center for Electronic and Optical Materials, NationalInstitute for Materials Science, 1-1 Namiki, Tsukuba, Japan. 3Research Center for Materials Nanoarchitectonics, National Institute for Materials Science, 1-1Namiki, Tsukuba, Japan. e-mail: joshua.island@unlv.edunpj 2D Materials and Applications |            (2025) 9:29 11234567890():,;1234567890():,;http://crossmark.crossref.org/dialog/?doi=10.1038/s41699-025-00551-7&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41699-025-00551-7&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41699-025-00551-7&domain=pdfmailto:joshua.island@unlv.eduwww.nature.com/npj2dmaterialswork function when compared with the other transition metaldichalcogenides12,13. This results in ON state characteristics closest to zerogate voltage. In a single-gate response, we show that bothmono- and bilayergraphene canbe effectively gated down to liquid helium temperatures, albeitwith resistance artifacts associated with clamping of the semiconductinggate. Dual-gated bilayer graphene with a few-layer graphene (FLG) controlgate, allows us to clearly demarcate the parameter space applicable to effi-cient gating. A 1D potential model is modified to support our experimentalresults and corroborate gate voltages at which clamping occurs. Addition-ally, we observe significant hysteresis for clamped gate voltages that can beattributed to trap states from intrinsic defects. The onset of hysteresis isdirectly correlated with the threshold voltage of the semiconducting gate.Finally, we show that all fourmaterials can be used to gate bilayer graphene,andwe compare their differences. Our results provide specific guidelines forgating using semiconducting gates in graphene devices and pave theway fortheir use in sensitive detectors and spectroscopy.ResultsSingle-gate characteristicsWeconstruct 2D van derWaals heterostructures to assess the feasibility of asemiconducting gate. The devices are created using a dry stacking andtransfer technique14,15, and a summary of the seven devicesmeasured in thisstudy is shown in Supplementary Table 1 and Supplementary Fig. 1. Flakesof hexagonal boron nitride (h-BN) are employed as dielectric layersseparating the gates from the graphene layers for all devices investigated.The Methods section presents the fabrication and measurement details forall devices studied. The single- and dual-gate results for bilayer graphene arepresentedhere in themain text and results for a similarmonolayer graphenedevice with an MoS2 gate can be found in Supplementary Fig. 2. Figure 1ashows an optical image of a bilayer device with an MoS2 gate with itsstacking order shown as a model in the inset. The resistivity is measuredusing either theMoS2 topgate (red layer inFig. 1a)or the few-layer graphene(FLG) control gate (blue layer in Fig. 1a). By utilizing both the FLG andMoS2 gates in a single device, we can directly compare the well-knowngating behavior of FLG with the unknown gating characteristics of thesemiconductor.Figure 1b shows the individual gate-dependent transport character-istics of BLG at room temperature. Both theMoS2 gate and FLG gate, whenswept separately, yield the characteristic peak in resistivity at the chargeneutrality point (CNP)16–20. The resistivity reported serves as an upperbound for the device as it includes theungated regions of theBLGchannel aswell. The FLG gate achieves a lower resistivity at higher doping than theMoS2 gate due to a larger area of overlap between the BLG and FLG flakes(the overlap can be seen in Fig. 1a).While the FLGgate shows no significanthysteresis, theMoS2gate responsedisplays a shoulder featurewithhysteresisbetween forward (solid) and backward (dashed) sweeps. These are asso-ciated with voltage clamping in the MoS2 and explained in more detailbelow. Figure 1c–d present color maps of the gate-dependent transport attemperatures from 400 K down to 2.5 K. The CNP is evident as a centralpeak anddoes not shift as temperature decreases for either gate, highlightingthe capability of MoS2 for gating down to liquid helium temperatures. TheCNP is situated to the left of zero gate voltage indicating some residualdoping noticeable for both gates. From the CNP position, we determine thedevice has a residual n-doping of 0.126 × 1012 cm−2 possibly due to trappedcontaminants at the BLG layer.We extract the field-effect mobilities for the BLG holes andelectrons from these gate sweeps to better characterize and comparethe effects of both gates. Mobilities are calculated based on the Drudemodel21 by taking the steepest slope of conductivity versus carrierdensity and using μðh;eÞ ¼ 1e∂σ∂nðh;eÞ. Figure 1e–f show the extractedmobilities as a function of temperature. Throughout the temperaturerange tested, the MoS2 gate yields mobilities comparable to the FLGgate, with values similar to previous studies22–24. This confirms theefficacy of MoS2 to operate as a gate over a wide range oftemperatures.Dual-gate characteristicsTo more thoroughly investigate the parameter space, we expand our mea-surement to varying both gates simultaneously. The results of these 2D gatesweeps are presented in Fig. 2, where panels (a–e) show theMoS2 as the fastforward sweep axis (fromnegative to positive voltage), andpanels (f–j) showit as the fast backward sweep axis. At 400 K (a,f), the BLG resistivity showstwo ridges of high resistivity. These ridges correspond to CNPs of differentregions of the BLG: the horizontal line aboutVFLG = 0 V corresponds to theregion singly-gatedby theFLGgate, and thediagonal line corresponds to thedual-gated region. We expect the CNP of the dual-gated region to occurwhere electrostatic doping by one gate is canceled by the other gate18,24,25,creating a diagonal line with negative slope across the 2D plot. The totalcarrier concentration n in the BLG is calculated from the sum of the indi-vidual gate influences: n ¼ ðC0MoS2VMoS2þ C0FLGVFLGÞ=e, where C0i is thecapacitance per unit area between the BLG and gate i, and Vi is the voltageapplied to gate i (i ∈ {MoS2, FLG}). The slope of the CNP line (where n = 0cm−2) is �C0MoS2=C0FLG ¼ �dFLG=dMoS2, where di is the thickness ofdielectric layer between the graphene and gate i. An apparent linear portionof theCNPexists in quadrant II of Fig. 2a. Fitting to a line results in a slopeof−1.44. This value closely matches the h-BN thickness ratio measured byAFM of�dFLG=dMoS2¼ �1:34.a20 μmd0.0 5.25.2-VFLG (V)T (K)400300200100e0.01.02.0μh (104  cm2 /Vs) MoS2FLG0 250T (K)cT (K)4003002001000.0 5.25.2-VMoS2 (V)2 3ρBLG (kΩ)bVGS (V)ρBLG (kΩ)-5 0 5300 KMoS2FLG2.52.01.5fμe (104  cm2 /Vs)0 250T (K)MoS2FLG1.01.50.5Fig. 1 | Graphite andMoS2 gate comparison. aOptical microscope image of device.Outlined are theMoS2 top gate (red), bilayer graphene (black), and graphite bottomgate (blue). Inset: Stacking order of the dual-gated graphene bilayer device withcolors corresponding to outlines in main panel. Gray layers represent dielectriclayers of h-BN. b Four-probe resistivity of the bilayer graphene as a function of gate-source voltage forMoS2 gate (red) and FLG gate (blue) at 300K. Solid lines representforward sweeps, while dashed lines represent backward sweeps. Resistivity of thebilayer graphenemapped as a function of temperature andMoS2 gate (c) or FLG gate(d). Color bar above (c) applies to (d) as well. For each gate sweep in (b–d), theinactive gate was held at zero volts. Field effect mobility of the holes (e) and electrons(f) in bilayer graphene as a function of temperature. The legend indicates whetherthe value is extracted from the MoS2 data (c) or the FLG data (d).https://doi.org/10.1038/s41699-025-00551-7 Articlenpj 2D Materials and Applications |            (2025) 9:29 2www.nature.com/npj2dmaterialsThe definition of n assumes the charge density in each gate varieslinearly with applied voltage, but this is not true for the semiconductingMoS2 gate. This can be seen in quadrant IV of the 2D gate sweeps where theCNP becomes disjoint from that in quadrant II. The shift suggests a lim-itation of the MoS2 gate. As the temperature decreases, the shift appearslarger, especially in the forward sweep (Fig. 2b–e). Other features appear atlower temperatures as well, such as CNP splitting and non-monotonic CNPcontours. These features are reproducible, but depend on both the sweepdirection and which gate is being swept fast (MoS2 or FLG) (see Supple-mentary Fig. 3). They coincide with the features seen in the 1D gate sweepsof Fig. 1, thus indicating single-gate sweeps must be carefully examined toassess limitations of the MoS2 gate.We perform further analysis of the 2D gate sweeps by looking at thehysteresis of the CNP feature with respect to the MoS2 gate voltage.Figure 2k, l show the hysteresisΔVMoS2 ;CNPas a function of temperature (k)and as a function of the FLG gate voltage (l). From panel (k), we see themagnitude of the hysteresis increases as temperature is decreased. Panel (l)shows forVFLG > 0 V (corresponding to quadrant II of the 2D gate sweeps),hysteresis is minimal, but sharply increases at lower gate voltages. Notably,the hysteresis appears to drop back to zero at large negative FLG voltages,suggesting that gating becomes effective again. However, in Fig. 2m, we seethe newCNP settles to a gate voltage that corresponds to a p-doped value of~ 2 × 1012 cm−2.Semiconductor characteristicsWithin the same device, we can measure the transport characteristics ofthe MoS2 to determine properties such as its threshold voltage and itssheet conductivity26–28. For these measurements, two graphite contacts tothe MoS2 layer are used. Figure 3a shows the transconductance of theMoS2 layer as a function of voltage applied to the BLG and FLG tiedtogether, using both of them as a single global back gate. The conductivitymeasurements have a noise floor of approximately 10 pS. Most of thetransconductance scans fall below this value in the MoS2 off-state. Wedetermine an on-off ratio of at least 106 and a field-effect mobility ofroughly 30–70 cm2/Vs depending on temperature, which is comparable toother reports26,29–34. At temperatures above 360 K, a notable leakage cur-rent dominates the off-state signal (see Supplementary Fig. 4). The inset ofFig. 3a shows the threshold gate voltage versus temperature, extracted byfitting the linear region of the transconductance with a line and finding itsx-intercept26,28. The generallymonotonic trend alignswith the expectationthat thermally excited electrons contribute to turning on MoS2 at lowergate voltages32,34.To better understand the limitations of theMoS2 gate during electricalmeasurements of the BLG, we examine the MoS2 characteristics for thesame gating conditions shown in Fig. 2. Figure 3b and c show the con-ductivity of MoS2 as a function of gate voltages relative to the BLG at 2.5 Kand 300 K. This was achieved by setting a constant bias voltage VDS acrossthe MoS2 while varying the potential on the BLG and FLG flakes. Therelative potentials were adjusted to match the gate voltages applied duringthe 2D gate sweeps of the BLG resistivity. The voltages can be converted asfollows:VMoS2�BLG ¼ �VBLG�MoS2ð1ÞVFLG�BLG ¼ VFLG�MoS2� VBLG�MoS2 ð2Þ400 K 301 K 201 K 102 K 2.5 K400 K 301 K 201 K 102 K 2.5 K1 2VFLG(V)VMoS2 (V)-505-505-5 0 5 -5 0 5 -5 0 5 -5 0 5 -5 0 5a b c d ef g h i jMoS2  gateForward sweepMoS2  gateBackward sweepVMoS2,CNP,offset (V)T (K)23VFLG = -7.3 Vforwardbackward0 200 400mT (K)050 250k-5 0 5VFLG (V)05lVFLG (V)50-5T (K)400200Fig. 2 | Dual-gated bilayer graphene sweeps. a–e BLG resistivity as a function ofMoS2 and FLG gate voltages. MoS2 gate was swept along the fast axis forward(negative to positive voltage). The panels show the CNP evolution as a function oftemperature, from400K (a) to 2.5K (e).White dashed lines are linear fits to theCNPrepresenting the expected gate voltages the CNP should appear. f–j Same as panels(a–e), but during the backward sweep (positive to negative voltage) of theMoS2 gate.Color bar above (e) applies to (a–j). Hysteresis of BLGCNP appearance inMoS2 gatevoltage as a function of temperature and the graphite gate voltage: (k) shows thetemperature dependence, while (l) shows the graphite gate dependence. m CNPoffset from expectedMoS2 voltage at VFLG =−7.3 V as a function of temperature forthe MoS2 forward sweep (blue) and backward sweep (red).https://doi.org/10.1038/s41699-025-00551-7 Articlenpj 2D Materials and Applications |            (2025) 9:29 3www.nature.com/npj2dmaterialswhere the convention VA−B indicates a potential difference from layer A(positive side) to layer B (negative side). The dashed lines of Fig. 3b, cindicate the MoS2 threshold voltage, determined by extracting theVMoS2�BLG corresponding to aMoS2 conductivity just above the noise floor.This line effectively demarcates the transition between the on-state and off-state of the MoS2 for the same gating conditions in Fig. 2.We match the on-off state threshold voltage of the MoS2 with thehysteresis of the BLG resistivity and see a distinct overlap. Figure 3d, e showsthe difference in BLG resistivity between the forward sweep and backwardsweep at 2.5K and 300Kwith the threshold voltage determined frompanels(b-c) superimposed. Qualitatively, there is a distinct change in behavior ofthe CNP hysteresis whenMoS2 is in its on-state versus its off-state. Furtheranalysis of low temperature hysteresis dependence on gate voltage sweeprate is available in Supplementary Fig. 5. In the off-state the hysteresis variesgreatly. The forward and backward sweeps significantly misalign, as indi-cated by the divergence between the positive and negative values in the plot.We highlight the success of the on-state MoS2 gate by measuring theband gap opening of BLG as a function of displacement field, as shown inSupplementary Fig. 6. In the on-state of theMoS2 gate, we determine a bandgap of 53 meV at a displacement field magnitude of 0.77 V/nm, which iscomparable to previously reported values measured with metallic gates3.The onset of hysteresis in the BLG can be explained by a voltage clampingeffect of the MoS2 gate. A 1D model calculating the effective potentialbetween the BLG channel and the MoS2 gate for different gate contactvoltages has been adapted from Qian et al.5 and is discussed in Supple-mentary Note 7, see also Supplementary Fig. 7. We find in the MoS2 on-state, the potential follows the applied gate contact voltage as expected.However, upon entering the off-state above a threshold voltage of 1.2 V, thepotential becomes clamped and remains constant for any applied gatecontact voltage. This leads to the unchanged charge neutrality point in theforward sweeps shown in Figs. 2 and 3. The hysteresis observed betweenforward and backward sweeps may be due to intrinsic defects in the MoS2that act as trap states. The hysteresis reaches a maximum within the bandgap of the MoS2 and corresponds to a trap state density of ≈ 2 × 1012 cm−2(Supplementary Fig. 4d). This agrees with the reported magnitude of sulfurdefect densities in MoS2 flakes at 2.9 × 1012 cm−2 35. At the highest positiveMoS2 contact gate voltages and for finite negative FLG gate voltages, weobserve an unclamping of the MoS2 gate and re-emergence of a hysteresis-free CNP (Figs. 2l and 3e). We conjecture that the Fermi level in the MoS2gate reaches the valence band as a result of partial gating from the FLG flakethrough the low-densityBLG layer. This occurs due to incomplete screeningfrom a sandwiched 2D metal36. The splitting of the CNP at the lowesttemperatures explored (Figs. 2e, j and3d) is not expectedbutmaybedue to acrack in the MoS2 flake causing two slightly different gating responses. Wehave observed cracking of various transition metal dichalcogenides duringthe van der Waals heterostructure fabrication process.Comparison with MoSe2, WS2, and WSe2For direct comparison, we also present graphene devices utilizing MoSe2,WS2, and WSe2 as semiconductor gates. The room temperature character-istics of threebilayer devices are presented inFig. 4, and twoadditionalMoSe2devices, as well as low temperature data, can be found in SupplementaryFig. 8. The architecture for these devices is the same as the MoS2 bilayerdevice, with a FLGbottomgate used as a control gate. Figure 4a compares thetransistor characteristics at room temperature of the four devices using fourdifferent materials. The MoS2 curve is replotted from Fig. 3a. MoS2 andMoSe2 present n-type carrier response, WSe2 presents p-type response, andWS2 presents ambipolar response. In Fig. 4b–d, the BLG resistivity is plottedas a function of the semiconducting gate and the FLG control gate. As inFig. 2, a white dashed line is plotted in Fig. 4b–d to identify the expectedpositionof theCNPofBLGgiven twometallic gates. In the case ofMoSe2, theCNP becomes disjointed for electron transport in the BLG, similar to theMoS2 device, while for WSe2, the CNP becomes disjoint for hole transport.Givenour analysis of theMoS2device, these are expectedbehaviors due to then-andp-type responseof these two semiconductors.TheWS2devicedisplaysa relatively smaller shift in theCNPcomparedwith the othermaterials due toits ambipolar response. The corresponding hysteresis in forward and back-ward gate sweeps are plotted in Fig. 4e–g for MoSe2, WS2, and WSe2,respectively. The limits of the color scale have been chosen tomatch those inFig. 3e for a better comparison between the four materials but note that theintensity of the subtracted hysteresis will also depend on the BLG resistivityitself. In comparison with theMoS2 device, which had minimal hysteresis intheONstate at roomtemperature, thesenewmaterials present a greater rangeof hysteresis that spans the gate voltage parameter space.DiscussionThe comparison of all four materials invites strategic engineering of noveldevices. For example, if electron transport and investigation around chargeneutrality is primarily desired in a graphene device with a semiconductinggate, MoS2 is the likely candidate, as it provides access to the largest range ofdensity change with hysteresis-free gating. If, on the other hand, p-typetransport is desired, WSe2 provides the best characteristics. WS2 presents aunique choice if amibpolar transport is needed away from charge neutrality.These general considerations are guidelines fordevice architecture, but detailsabout the semiconducting gates must be considered. Our semiconductinggates have a range of thicknesses from bilayer to bulk-like (>6 layers). TheyFig. 3 | MoS2 gate threshold and gating effectiveness. aMoS2 conductivity as afunction of gate-source voltage applied to electrically-connected BLG and FLGlayers acting as a single back gate. Scans were taken from 2.5 K to 400 K, in steps of7.5 K. Inset: Extracted threshold voltage as a function of temperature. MoS2 con-ductivity as a function of gate voltage relative to the bilayer graphene at 2.5 K (b) and300 K (c). Dashed line indicates the MoS2 threshold voltage. Hysteresis of BLGresistivity (forward sweep minus backward sweep) as a function of gate voltagerelative to BLG at 2.5 K (d) and 300 K (e). Dashed lines are the same as frompanels (b, c).https://doi.org/10.1038/s41699-025-00551-7 Articlenpj 2D Materials and Applications |            (2025) 9:29 4www.nature.com/npj2dmaterialsare also contacted using few-layer graphene flakes, which determines theSchottky barrier between the semiconducting gate and its contact electrode. Ifdifferent contactmaterials areusedor the semiconductinggate enters the few-layer regime where the band gap is known to change37,38, the detailed gatingcharacteristics of a graphene device will also likely change.We have demonstrated the ability of semiconducting transition metaldichalcogenides to electrostatically gate mono- and bilayer graphene. As asingle gate, MoS2 can effectively induce doping in graphene and bilayergraphene down to temperatures as lowas 2.5K.A shoulder feature in the 1Dgate sweeps is identified and associated with voltage clamping in the MoS2gate. We find similar mobilities when using either the MoS2 gate or a FLGcontrol gate at highdoping. In adual-gate configuration,we identify a regionof gate voltage parameter space where MoS2 is in its on-state and providesadequate gating response. In the off-state, we observe voltage clamping oftheMoS2 gate leading to anunexpecteddeviationof theCNP fromthe lineartrend in 2D gate parameter space. A 1D potential model corroborates thisclamping effect. Hysteresis of the CNP in theMoS2 off-state agrees with thefilling and emptying of trap states from intrinsic sulfur vacancies. Finally, wedirectly compare bilayer graphene devices using other transition metaldichalcogenides (MoSe2, WS2, and WSe2) showing unqiue n-type, p-type,and ambipolar characteristics. The advantages and disadvantages of usingeach material is discussed. Our results provide straightforward guidelinesfor the operation of a semiconducting gate in graphene devices and pave theway toward novel optoelectronic device architectures, high-frequency on-chip measurements, and sensitive detectors.MethodsMaterials were isolated from bulk crystals via mechanical exfoliation1.Exfoliated flakes were then assembled into a vertical heterostructure using adry stacking method14,15. Few-layer graphene (graphite) flakes were used ascontacts to the semiconducting gates to allow independentmeasurement ofthem. The final stack was fully encapsulated with h-BN. h-BN thicknessesare chosen according to optical contrast39 and range from 20 to 70 nm. Forthemain textMoS2 device, the thicknesses of the h-BNweremeasuredusingatomic forcemicroscopy (dMoS2= 26.7 nmand dFLG= 35.9 nm).A dielectricconstant of ϵBN = 3.4ϵ0 was also used, where ϵ0 is the permittivity of freespace40. A summary of the devices measured is presented in SupplementaryTable 1 and optical microscopy images are shown in Supplementary Fig. 1.Standard photolithographic methods were used to pattern the electrodes.Graphite/graphene layers were exposed by etching away the h-BN in aCHF3/O2 plasma, thenmetal electrodeswere either sputtered (Au 50 nm) at5mTorr or deposited by electron-beamevaporation (Cr 5nm,Au45nm) at10−6 Torr.Measurements were performed in aQuantumDesignEvercool IIPPMS and a Quantum Design Opticool cryostat, capable of base tem-peratures ~ 2.0 K. BLG resistivity measurements were performed in a two(TMDG026) or four-probe (all others) configuration. Semiconducting gateconductivity measurements were performed in a two-probe configuration.Data availabilityThe data that support the findings of this study are available from thecorresponding author upon reasonable request.Received: 10 December 2024; Accepted: 24 March 2025;References1. Novoselov, K. S. et al. Electric field effect in atomically thin carbonfilms. Science 306, 666–669 (2004).2. Zhang, Y., Tan, Y.-W., Stormer, H. L. & Kim, P. Experimentalobservation of the quantum hall effect and berry’s phase in graphene.Nature 438, 201–204 (2005).3. Zhang, Y. et al. Direct observation of a widely tunable bandgap inbilayer graphene. Nature 459, 820–823 (2009).4. Zibrov, A. A. et al. Tunable interacting composite fermion phases in ahalf-filled bilayer-graphene landau level.Nature 549, 360–364 (2017).5. Qian, Q. et al. 2d materials as semiconducting gate for field-effecttransistors with inherent over-voltage protection and boosted on-current. npj 2D Mater. Appl. 3, 24 (2019).6. Gallagher, P. et al. Quantum-critical conductivity of the dirac fluid ingraphene. Science 364, 158–162 (2019).Fig. 4 | Gating bilayer graphene with MoSe2, WS2, and WSe2 at room tempera-ture. aTwo terminal conductivity at room temperature for all four materials used asa semiconducting gate, as a function of gate-source voltage applied to electrically-connected BLG and FLG layers acting as a single back gate. BLG resistivity as afunction of MoSe2 (b), WS2 (c), and WSe2 (d), and FLG gate voltages. Hysteresis ofBLG resistivity (forward sweep minus backward sweep) as a function of MoSe2 (e),WS2 (f), and WSe2 (g) gate voltage relative to BLG at room temperature. For all 2Dgate sweeps, the TMD gate voltage was the fast sweep axis.https://doi.org/10.1038/s41699-025-00551-7 Articlenpj 2D Materials and Applications |            (2025) 9:29 5www.nature.com/npj2dmaterials7. Seo, J. et al. On-chip terahertz spectroscopy for dual-gated van derwaals heterostructures at cryogenic temperatures. Nano Lett. 24,15060–15067 (2024).8. Yan, J. et al.Dual-gatedbilayergraphenehot-electronbolometer.Nat.Nanotechnol. 7, 472–478 (2012).9. Yoshioka, K. et al. Ultrafast intrinsic optical-to-electrical conversiondynamics in a graphene photodetector. Nat. Photonics 16, 718–723(2022).10. Kim, M.-H. et al. Photothermal response in dual-gated bilayergraphene. Phys. Rev. Lett. 110, 247402 (2013).11. Gopalan, K. K. et al. Mid-infrared pyroresistive graphene detector onlinbo3. Adv. Opt. Mater. 5, 1600723 (2017).12. Liu, Y., Stradins, P. & Wei, S.-H. Van der waals metal-semiconductorjunction:Weak fermi level pinning enables effective tuning of schottkybarrier. Sci. Adv. 2, e1600069 (2016).13. Cloninger, J. A. et al. A back-to-back diode model applied to van derwaals schottky diodes. J. Phys.: Condens. Matter 36, 455301 (2024).14. Purdie, D. G. et al. Cleaning interfaces in layered materialsheterostructures. Nat. Commun. 9, 5387 (2018).15. Haley, K. L. et al. Heated assembly and transfer of van der waalsheterostructures with common nail polish. Nanomanufacturing 1,49–56 (2021).16. Geim, A. &Novoselov, K. The rise of graphene.Nat.Mater. 6, 183–191(2007).17. Allen,M. J., Tung, V. C. &Kaner, R. B. Honeycomb carbon: A reviewofgraphene. Chem. Rev. 110, 132–145 (2010).18. Yan, J. &Fuhrer,M.S.Charge transport in dual gatedbilayer graphenewith corbino geometry. Nano Lett. 10, 4521–4525 (2010).19. Kuzmenko, A. B. et al. Infrared spectroscopy of electronic bands inbilayer graphene. Phys. Rev. B 79, 115441 (2009).20. Efetov,D.K.,Maher,P.,Glinskis,S.&Kim,P.Multiband transport inbilayergraphene at high carrier densities. Phys. Rev. B 84, 161412 (2011).21. Gosling, J. H. et al. Universal mobility characteristics of grapheneoriginating from charge scattering by ionised impurities. Commun.Phys. 4, 30 (2021).22. Cobaleda, C., Pezzini, S., Diez, E. & Bellani, V. Temperature- anddensity-dependent transport regimes in ah-bn/bilayer graphene/h-bnheterostructure. Phys. Rev. B 89, 121404 (2014).23. Tan, C., Adinehloo, D., Hone, J. & Perebeinos, V. Phonon-limitedmobility in h-bn encapsulated ab-stacked bilayer graphene. Phys.Rev. Lett. 128, 206602 (2022).24. Shimazaki, Y. et al. Generation and detection of pure valley current byelectrically inducedberry curvature in bilayer graphene.Nat. Phys. 11,1032–1036 (2015).25. Henriksen, E. A. & Eisenstein, J. P. Measurement of the electroniccompressibility of bilayer graphene. Phys. Rev. B 82, 041412 (2010).26. Late, D. J., Liu, B., Matte, H. S. S. R., Dravid, V. P. & Rao, C. N. R.Hysteresis in single-layer mos2 field effect transistors. ACS Nano 6,5635–5641 (2012).27. Cho, K. et al. Electric stress-induced threshold voltage instability ofmultilayer mos2 field effect transistors. ACS Nano 7, 7751–7758 (2013).28. Lee, G.-H. et al. Highly stable, dual-gatedmos2 transistors encapsulatedby hexagonal boron nitride with gate-controllable contact, resistance,and threshold voltage. ACS Nano 9, 7019–7026 (2015).29. Jariwala, D. et al. Band-like transport in highmobility unencapsulatedsingle-layer MoS2 transistors. Appl. Phys. Lett. 102, 173107 (2013).30. Bartolomeo, A. D. et al. Hysteresis in the transfer characteristics ofmos2 transistors. 2D Mater. 5, 015014 (2017).31. Lin, M.-W. et al. Thickness-dependent charge transport in few-layermos2 field-effect transistors. Nanotechnology 27, 165203 (2016).32. Radisavljevic, B. & Kis, A. Mobility engineering and a metal–insulatortransition in monolayer mos2. Nat. Mater. 12, 815–820 (2013).33. Yang, H., Cai, S., Wu, D. & Fang, X. Humidity-dependentcharacteristics of few-layer mos2 field effect transistors. Adv.Electron. Mater. 6, 2000659 (2020).34. Ji, H. et al. Temperature-dependent opacity of the gate field insidemos2 field-effect transistors. ACS Appl. Mater. Interfaces 11,29022–29028 (2019).35. Trainer, D. J. et al. Visualization of defect induced in-gap states inmonolayer mos2. npj 2D Mater. Appl. 6, 13 (2022).36. Luryi, S. Quantumcapacitance devices.Appl. Phys. Lett. 52, 501–503(1988).37. Mak,K. F., Lee,C.,Hone, J., Shan, J. &Heinz, T. F. Atomically thinmos2: a new direct-gap semiconductor. Phys. Rev. Lett. 105, 136805(2010).38. Splendiani, A. et al. Emergingphotoluminescence inmonolayermos2.Nano Lett. 10, 1271–1275 (2010).39. Golla, D. et al. Optical thickness determination of hexagonal boronnitride flakes. Appl. Phys. Lett. 102, 161906 (2013).40. Pierret, A. et al. Dielectric permittivity, conductivity and breakdownfield of hexagonal boron nitride.Mater. Res. Express9, 065901 (2022).AcknowledgementsThis work was supported by the National Science Foundation under GrantNo. (2047509) and by, or in part by, the U.S. ArmyResearch Laboratory andthe U.S. Army Research Office under contract/grant number(W911NF2310160). K.W. and T.T. acknowledge support from the JSPSKAKENHI (Grant Numbers 21H05233 and 23H02052) and World PremierInternational Research Center Initiative (WPI), MEXT, Japan.Author contributionsR.M.S., B.K., A.F., K.L.H., and N.T.P. fabricated the devices. R.M.S. performedthe measurements and modified the 1D gating model. K.W. and T.T. grew thehBNcrystals.R.M.S. andJ.O.I.wrote themanuscript in consultationwithK.L.H.and N.T.P.Competing interestsThe authors declare no competing interests.Additional informationSupplementary information The online version containssupplementary material available athttps://doi.org/10.1038/s41699-025-00551-7.Correspondence and requests for materials should be addressed toJoshua O. Island.Reprints and permissions information is available athttp://www.nature.com/reprintsPublisher’s note Springer Nature remains neutral with regard tojurisdictional claims in published maps and institutional affiliations.Open Access This article is licensed under a Creative CommonsAttribution 4.0 International License, which permits use, sharing,adaptation, distribution and reproduction in anymedium or format, as longas you give appropriate credit to the original author(s) and the source,provide a link to the Creative Commons licence, and indicate if changeswere made. The images or other third party material in this article areincluded in the article’s Creative Commons licence, unless indicatedotherwise in a credit line to the material. If material is not included in thearticle’sCreativeCommons licence and your intended use is not permittedby statutory regulation or exceeds the permitted use, you will need toobtain permission directly from the copyright holder. To view a copy of thislicence, visit http://creativecommons.org/licenses/by/4.0/.© The Author(s) 2025https://doi.org/10.1038/s41699-025-00551-7 Articlenpj 2D Materials and Applications |            (2025) 9:29 6https://doi.org/10.1038/s41699-025-00551-7http://www.nature.com/reprintshttp://creativecommons.org/licenses/by/4.0/www.nature.com/npj2dmaterials Gating monolayer and bilayer graphene with a two-dimensional semiconductor Results Single-gate characteristics Dual-gate characteristics Semiconductor characteristics Comparison with MoSe2, WS2, and WSe2 Discussion Methods Data availability References Acknowledgements Author contributions Competing interests Additional information