High-Current Gain Two-Dimensional MoS2‑Base Hot-ElectronTransistorsCarlos M. Torres, Jr.,*,† Yann-Wen Lan,*,†,‡ Caifu Zeng,† Jyun-Hong Chen,§ Xufeng Kou,†

Aryan Navabi,† Jianshi Tang,† Mohammad Montazeri,† James R. Adleman,∥ Mitchell B. Lerner,∥

Yuan-Liang Zhong,§ Lain-Jong Li,⊥ Chii-Dong Chen,‡ and Kang L. Wang*,†

†Department of Electrical Engineering, University of California at Los Angeles, Los Angeles, California 90095, United States‡Institute of Physics, Academia Sinica, Taipei 115, Taiwan§Department of Physics and Center for Nanotechnology, Chung Yuan Christian University, Chungli 32023, Taiwan∥Space and Naval Warfare (SPAWAR) Systems Center Pacific, San Diego, California 92152, United States⊥Physical Sciences and Engineering Division, King Abdullah University of Science and Technology (KAUST), Thuwal, 23955-6900,Kingdom of Saudi Arabia

*S Supporting Information

ABSTRACT: The vertical transport of nonequilibrium charge carriers through semiconductor heterostructures has led tomilestones in electronics with the development of the hot-electron transistor. Recently, significant advances have been made withatomically sharp heterostructures implementing various two-dimensional materials. Although graphene-base hot-electrontransistors show great promise for electronic switching at high frequencies, they are limited by their low current gain. Here weshow that, by choosing MoS2 and HfO2 for the filter barrier interface and using a noncrystalline semiconductor such as ITO forthe collector, we can achieve an unprecedentedly high-current gain (α ∼ 0.95) in our hot-electron transistors operating at roomtemperature. Furthermore, the current gain can be tuned over 2 orders of magnitude with the collector-base voltage albeit thisfeature currently presents a drawback in the transistor performance metrics such as poor output resistance and poor intrinsicvoltage gain. We anticipate our transistors will pave the way toward the realization of novel flexible 2D material-based high-density, low-energy, and high-frequency hot-carrier electronic applications.

KEYWORDS: 2D materials, transition metal dichalcogenides, MoS2, hot-electron transport, high-current gain

For over half a century, Moore’s law has driven the siliconelectronics industry toward smaller and faster transistors.

However, as the scaling limit of silicon complementary metal−oxide-semiconductor (CMOS) technology draws to an end,novel materials and device concepts have been eagerly soughtout and investigated with hopes to augment the next generationof information processing. One promising device concept is thehot-electron transistor (HET),1−7 which relies on the verticaltransport of a controlled source of hot-electrons. Ever sinceMead first proposed this device concept in 1960,1,2 there havebeen plethora of HET variants implementing diverse materialsystems.1−7 Usually, these HETs feature substantial currentgain (α ∼ 0.75) at cryogenic temperatures (T = 4.2 K) but verypoor current gain at room temperature.3,4 Only a few HETshave shown high current gain (α ∼ 0.9) at room temperature6,7

but rely on precise yet complicated epitaxial layered structures

grown by molecular beam epitaxy (MBE). HETs implementingtwo-dimensional (2D) materials,8−10 such as graphene,8,11,12 inthe base region13−15 have recently shown great promise forultrahigh frequency operation.14,16−21 These vertical transportthree-terminal electronic devices can be designed withatomically sharp heterostructures by the stacking of variousvan der Waals materials.9 This allows one to play with theconduction and valence band offsets,22 which determine thepotential landscape experienced by hot-carriers and ultimatelythe device performance.23 Until now, only vertical graphene-base hot-electron transistors have been experimentally demon-

Received: July 25, 2015Revised: October 23, 2015Published: November 2, 2015

Letter

pubs.acs.org/NanoLett

© 2015 American Chemical Society 7905 DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

strated,15,24 yet their transport characteristics feature asignificantly low common-base current gain (α ∼ 10−2).These shortcomings preclude the fulfillment of realizing 2Dmaterial-based vertical hot-carrier transistors operating at highfrequencies. As a first step toward this goal, we propose anddemonstrate a novel device concept which enables unprece-dentedly high current gain in 2D material-based hot-electrontransistors. In this Letter, we demonstrate a novel vertical hot-electron transistor incorporating single-layer MoS2

25−27 in thebase region (MoS2−HET). To the best of our knowledge, allprevious vertical graphene-base hot-electron transistors im-plemented a metal for the collector electrode and exhibited anextremely low current gain. In this work, by utilizing anoncrystalline semiconductor such as ITO (an n-type trans-parent conducting oxide) as the collector electrode,28−30 wedemonstrate for the first time that the MoS2−HETs operate atroom temperature and exhibit a high common-base currentgain (α ∼ 0.95) over the entire base-emitter bias (VBE) range.Furthermore, the current gain can be dynamically tuned over 2orders of magnitude with the collector-base voltage albeit thisfeature currently presents a drawback in the transistorperformance metrics such as poor output resistance and poorintrinsic voltage gain.The device structure and schematic of the MoS2−HET are

introduced in Figure 1. An isometric view of the MoS2−HETdevice structure is shown in Figure 1a, and a cross-sectionalview of its vertical heterostructure active region is depicted inFigure 1b. The three-terminal device consists of a degenerately

doped n++ silicon substrate (ND ∼ 1019 cm−3) as the emitter(E), a monolayer of chemical vapor deposition (CVD) grownMoS2 (see Supporting Information, Figure S1) as the base (B),and sputtered (∼45 nm) ITO as the collector (C). The MoS2was grown using a CVD method31 and transferred onto thesubstrate using a PMMA transfer method.32 A thermally grownthin (∼3 nm) SiO2 tunnel barrier separates the emitter andbase terminals, whereas an atomic-layer deposited (∼55 nm)HfO2 separates the base and collector and serves as the filteringbarrier. In this fabrication process, arrays of MoS2−HETdevices are isolated from each other via a thick (∼300 nm)SiO2 field oxide. The underlying silicon surface which is notcovered by the thick SiO2 field oxide, but rather surrounded byit, serves to confine the current into the emitter region of thedevice for hot-electron injection through the SiO2 tunnelbarrier (see Supporting Information, Figure S2). Accordingly,the fabrication process was designed to be compatible withsilicon CMOS technology. An optical micrograph of an actualMoS2−HET (top-view) is presented in Figure 1c. The detailedfabrication process is described in the Supporting Information.In this particular study, a common-base configuration wasemployed during the electrical measurements and the biasingcircuit incorporating the schematic symbol for the MoS2−HETdevice is shown in Figure 1d. Note that both of the basecontacts are grounded during the electrical measurements inorder to achieve a uniform potential distribution across theMoS2 base region. A common-emitter configuration was alsoemployed during the electrical measurements for a few devices

Figure 1. Device structure and schematic of the MoS2−HET. (a) An isometric view of an MoS2−HET device structure. The capital letters E, B, andC represent the emitter, base, and collector, respectively. (b) Cross-sectional view of the vertical heterostructure active region with single-layer MoS2(0.65 nm) as the base, ITO (∼45 nm) serves as the collector electrode, and an n++ silicon substrate is used as the emitter. A thin SiO2 (∼3 nm)tunnel barrier is utilized for hot-electron injection, and HfO2 (∼55 nm) serves as the filtering barrier. The hot-electrons injected from the emitter(red arrows) are schematically shown. (c) Optical micrograph (top-view) of an actual MoS2−HET device. The scale bar is 100 μm. The dashedcircle outlines the MoS2 region. (d) Common-base configuration circuit incorporating the schematic symbol for the MoS2−HET device.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7906

in order to confirm the consistent high-current gain resultspresented in this work.In order to clearly understand the physics, we first focus on

describing the two modes of operation for the MoS2−HETusing energy band diagrams. The flat band condition is shownin Figure 2a. The conduction band offset between themonolayer MoS2 and HfO2 is 1.52 eV. This forms the filterpotential barrier height (Δe) for the hot-electrons. The filterpotential barrier height and width at the collector-base junctionare important parameters that determine the collector currentafter the hot-electrons tunnel through the 3 nm SiO2 emitter-base tunnel barrier. In the previous graphene-base hot-electrontransistors,15,24 the filter barrier height between graphene andAl2O3 was 3.3 eV,15,24 whereas between graphene and HfO2 itwas 2.05 eV.24 In both prior cases, the filter barrier heights aregreater than that between MoS2 and HfO2 (1.52 eV), thus animprovement of the ratio, in our case, between the collectorcurrent and the emitter current (α = IC/IE) is expected. Ingeneral, the potential barrier height, the potential barrier width,and the strength of the applied electric field are key parametersthat determine the tunneling current across a tunneljunction.33−35 The tunneling current across a tunnel junctionis generally described by33−35

∫ ρ ρ∝ · · − · − − ·I V E E E eV f E eV f E T E( ) d ( ) ( ) [ ( ) ( )] ( )1 2 (1)

where f(E) is the Fermi distribution function, ρ1(E) is thedensity of states of the first electrode, ρ2(E) is the density ofstates of the second electrode, and T(E) is the transmissionprobability which, for the MoS2−HET device structure, isdominated by its exponential sensitivity to the barrier height35

(Δ) as opposed to the tunneling density of states.34 T(E)depends on the energy E of the tunneling electrons as follows:

∼ −T E( ) e W E( ) (2)

where W(E) is related to the effective width and height of thepotential barrier and strongly depends on the strength of theelectric field. W(E) can be generalized within the WKBapproximation:34

∫= · ΔW x k x2 d Im ( ( ))d

z0 (3)

Thus, the application of a strong electric field candramatically alter the shape (e.g., both the effective heightand width) of the potential barrier (eq 3) and result in anincreased collector current. Moreover, when designing verticalhot-electron transistors,36−38 it is paramount to choose theproper combination of 2D material for the base as well as thefiltering barrier dielectric, which yields the desired conductionband offset,22 or filter potential barrier height, for the hot-electrons. Instead of relying on complicated and expensive

Figure 2. Energy band diagrams for the operating conditions of the MoS2−HET. (a) Flat band condition. Note that the conduction band offset, orthe filter barrier height for the hot-electrons (Δe), between the single-layer MoS2 and the HfO2 is Δe = 1.52 eV. (b) The current components aredepicted for the hot-electron contribution (red dotted arrow) to the total current. (c) Energy band diagram depicting the off-state condition.Electrons have insufficient kinetic energy to overcome the filter barrier at the collector-base junction and do not reach the collector. (d) Energy banddiagram depicting the on-state condition. For VCB > 0 (dashed red lines), the hot-electrons tunneling through the emitter-base tunnel barrier havesufficient kinetic energy to overcome the filter barrier and reach the collector.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7907

methods to produce atomically sharp interfaces,1−7 it is nowpossible to design the equilibrium filter potential barrier heightin vertical transport devices by choosing from plethora of 2Dmaterials since the conduction band offset results from materialspecific parameters such as the electron affinity of the collector-oxide and the work function (e.g., in the case of graphene) orelectron affinity (e.g., for all other 2D materials with bandgaps)of the particular 2D material used.Having established the significance of the filter potential

barrier height for the hot-electrons (Δe) in the MoS2−HETs,we next illustrate the current components governing the devicetransport. The current components for the hot-electroncontribution (red arrow) to the total current flow throughthe MoS2−HET are shown in Figure 2b. The emitter current(IE) across the SiO2 tunnel oxide is due to hot-electronsinjected from the n++ silicon substrate. Furthermore, thecollector current (IC) across the HfO2 collector-base oxide isdue to the portion of the injected hot electrons with enoughkinetic energy that surpass the filter barrier and reach thecollector.Now that the current components and the barrier heights

experienced by the hot-electrons in the MoS2−HETs have beenshown, we proceed to describe the modes of operation of thesenovel transistors. Figure 2c shows the energy band diagram forthe off-state condition of the MoS2−HET. In the absence of anapplied VCB, the hot-electrons injected through the tunnel oxidehave insufficient kinetic energy to overcome the filter barrier atthe collector-base junction and do not reach the collector.Instead, they backscatter and thermalize into the MoS2 baseregion. However, the situation drastically changes with the

application of a large positive VCB. Figure 2d shows the energyband diagram for the on-state condition of the MoS2−HET. Inthis scenario, hot-electrons tunneling through the emitter-basetunnel oxide have sufficient kinetic energy to overcome thefilter barrier and reach the collector.Based on the physical concepts just described, the device

performances of the MoS2−HETs were characterized using thecommon-base configuration. In the following, we characterizethe MoS2−HET by applying positive VCB. Figure 3a shows theenergy band diagram depicting the MoS2−HET. Specifically,Figure 3a shows the conduction and valence band edges at thecollector-base junction with a positive VCB applied. In thiscondition, once hot-electrons tunneling through the emitter-base tunnel barrier have sufficient kinetic energy, they canvertically transport through the MoS2 base region, surpass thefilter barrier at the collector-base junction, and reach thecollector. Consequently, an increasingly positive VCB willcontinue to effectively make the filter potential barrier thinnerand promote hot-electrons reaching the collector due to anincrease in their transmission probability. This qualitativebehavior is exhibited in the input and transfer characteristics ofthe MoS2−HETs. The input characteristics (IE−VBE) corre-spond to how the emitter current depends on VBE, whereas thetransfer characteristics (IC−VBE) correspond to the manner inwhich the collector current varies with VBE. Figure 3b shows theinput and transfer characteristics for Device 1. In this device,the maximum VBE is limited to 3 V to avoid dielectricbreakdown of the tunnel oxide. The emitter current (IE) andthe collector current (IC) are shown as a function of VBE (VBEwas swept from 0 to +3 V) at a VCB of +1 V. Both currents

Figure 3. Electrical characterization of the MoS2−HET in the common-base configuration. (a) Energy band diagram depicting MoS2−HETs. Theconduction and valence band edges at the collector-base junction are shown for a positive VCB, which reduces the filter barrier for the hot-electrons.(b) Input and transfer characteristics for Device 1. The emitter current (black diamonds) and the collector current (red circles) are shown as afunction of VBE at VCB = +1 V. (c) Transfer characteristics. The collector current as a function of VBE is shown for various positive VCB. (d)Common-base current gain (α) as a function of VBE at VCB = +8 V. The inset shows α as a function of VBE at various positive VCB: VCB = 0, +2, +4,+6, and +8 V.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7908

rapidly increase at larger VBE, as is typical for HETs. Similarly,Figure 3c shows a family of transfer characteristics for Device 1.The collector current as a function of VBE is shown for variouspositive VCB. It is evident that the collector current increaseswith increasingly positive VCB. This is due to the fact that thefilter potential barrier width at the collector-base junction iseffectively reduced as the applied VCB becomes more positive.Correspondingly, a greater portion of the injected hot-electronsfrom the emitter have high enough kinetic energy to verticallytransport through the MoS2 base region, surpass the filterbarrier, and reach the collector, thus contributing to anincreasing collector current.From the input and transfer characteristics, we can next

ascertain the common-base current gain (α) of Device 1, whichis a figure of merit for HETs and is defined as α = IC/IE. Figure3d shows α as a function of VBE at VCB = +8 V. Interestingly,the current gain, α, features a nearly constant characteristic atall VBE with a value of at least 90% for this particular case of VCB

= +8 V. This implies that, even at a low VBE, at least 90% of theinjected hot-electrons ballistically traverse the single-layer MoS2base region at room temperature. The inset of Figure 3d showsa family of α characteristics as a function of VBE at severalpositive VCB (VCB = 0, +2, +4, +6, and +8 V). The current gain,

α, monotonically increases with positive VCB due to the reducedpotential barrier and associated increase in the transmissionprobability of hot-electrons and exhibits a nearly constantcharacteristic throughout the entire VBE range with an averagemagnitude of about 95% at VCB = +8 V.With the analysis of the input and transfer characteristics

complete, we now investigate the common-base outputcharacteristics of the MoS2−HETs, which correspond to howthe output collector current depends on VCB. Figure 4a showsthe common-base output characteristics for Device 2. Thecollector current is shown as a function of VCB at three positiveVBE biases. The collector current is insensitive to modulationbelow a critical electric field, or correspondingly a VCB voltage,across the HfO2 collector-base oxide. However, above a criticalelectric field across the HfO2, the collector current is quitesensitive to modulation and rapidly increases with a furtherincrease in VCB. In order to convey the robust nature of thishigh current gain in the MoS2−HETs, Figure 4b shows asemilog plot of α as a function of VCB at VBE = +3 V. It clearlyshows that α increases with an increasingly positive VCB and canbe tuned over an order of magnitude since this lowers the filterpotential barrier experienced by the hot-electrons and allowsthem to reach the collector. Remarkably, the room temperature

Figure 4. Output characteristics and tunable current gain of the MoS2−HET. (a) Common-base output characteristics for Device 2. The collectorcurrent is shown as a function of VCB at VBE = +1 V, +2 V, and +3 V. (b) The common-base current gain (α) for Device 2 is shown in log-scale as afunction of VCB for VBE = +3 V.

Figure 5. Electrical characterization of the MoS2−HET in the common-emitter configuration. (a) Gummel plot for Device 1 biased in the common-emitter configuration. The collector and base currents are shown as a function of VBE at a fixed output voltage of VCE = +10 V. (b) The common-emitter output characteristics for Device 1. The collector current is shown as a function of VCE at VBE = 0 V, +1 V, +2 V, and +3 V. The inset showsthe common-emitter current gain (β) as a function of VCE at VBE = +2 V. A maximum common-emitter current gain (β) of around 4 is achieved, andit can be tuned with the output voltage VCE.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7909

common-base current gain (α) in this type of novel 2Dmaterial-based vertical device is unprecedentedly high for thelargest VCB applied.Next, we investigate the MoS2−HET characteristics when

biased in the common-emitter configuration in order tocorroborate the high and tunable current gain we achieved inthe common-base configuration. The Gummel plot is used as afigure of merit when analyzing bipolar transistors. It is asimultaneous semilog plot of the collector and base currents asa function of the input voltage (VBE) at a fixed output voltage(VCE). The common-emitter current gain (β = IC/IB) can beascertained from the Gummel plot by taking the ratio of thecollector current to the base current at a fixed VBE. Figure 5ashows the Gummel plot for Device 1 when biased in thecommon-emitter configuration.The collector and base currents are shown in log-scale as a

function of VBE at a fixed output voltage of VCE = +10 V. TheGummel plot confirms the transistor action of the MoS2−HETas the input base current is directly amplified to the outputcollector current. Finally, Figure 5b shows the common-emitteroutput characteristics for Device 1. The collector current isshown as a function of VCE at three positive VBE biases inaddition to VBE = 0 V. The inset shows the common-emittercurrent gain (β) as a function of VCE at VBE = +2 V. Amaximum common-emitter current gain of around 4 isachieved, and it can be tuned with the output voltage VCE.We note that a slightly different current gain is obtainedbetween the common-emitter and the common-base config-uration measurements. We speculate that this discrepancy isdue to the differences in the measurement circuits biasing theactual device structure in the two measurement configurations.Applying an output voltage (VCB) in the common-baseconfiguration results in the collector current originating fromthe ITO collector electrode, traversing through the HfO2 filterbarrier dielectric, and laterally transporting through the MoS2surface into the grounded base contacts. This is because VCBactually only controls the band bending of the HfO2 layer, andit does not modulate the SiO2/MoS2 layers due to the fact thatour base contacts lie on the top surface of the MoS2. Thus, α ismainly limited by the recombination processes inside the HfO2dielectric, which is quite small.36 However, applying an outputvoltage (VCE) in the common-emitter configuration results inthe collector current traversing through the entire ITO/HfO2/MoS2/SiO2/n

++ silicon vertical heterostructure, since bandbending occurs throughout the entire heterostructure.Furthermore, since there exist more pronounced non-radiativerecombination (e.g., Auger recombination and trap-states)processes in the MoS2 layer,37,38 the measured β in thecommon-emitter configuration is thus expected to be smallerthan the measured α in the common-base configuration.Nonetheless, by biasing the MoS2−HETs in both the common-base and the common-emitter configurations, we have explicitlyshown that the measured current gains (α, β) in either scenariocorroborate each other and further attest the high and tunablecurrent gain in our MoS2−HETs.On the other hand, for transistor operation, it is better to

operate in the saturation region of the output characteristicswhere the collector current (IC) is fairly constant. However, thelack of any reasonable saturation in the output characteristics ofthe MoS2−HETs (see Figure 5b) may be due to the onset ofdielectric breakdown in both the SiO2 tunnel oxide and theHfO2 filter oxide. The lack of saturation in the outputcharacteristics is specific to this batch of MoS2−HETs but

may be achieved if a filter oxide with a larger dielectric strengthand/or a reduced thickness is used in order to promote largeraccelerating electric fields across the filter oxide. In fact, thecurrent gain dependence on VCB indicates that the outputresistance and the intrinsic voltage gain of our MoS2−HETs arepoor. Based on our measurements, the input and outputresistances are on the order of a GΩ and a few tens of GΩ,respectively. The intrinsic voltage gain in these prototypeMoS2−HETs is about 3, which is quite small. In order toimprove the intrinsic voltage gain, a low input impedance isdesirable. In our device structure, a low input impedance can bereadily achieved by further reducing the thickness of the SiO2tunnel barrier to around 1 nm. The ensuing tunneling currentdensity (IE) should improve by at least 2 orders of magnitudeand thus enable a much higher intrinsic voltage gain in ourMoS2−HETs. Furthermore, reducing the thickness of thetunnel oxide would significantly benefit the driving capabilityand speed of the MoS2−HETs compared to the current batchwhich feature very small device current levels. It is worthmentioning that the MoS2−HETs in this work were intention-ally designed for DC electrical characterization in order toprove that decreasing the filter potential energy barrier heightusing MoS2 and HfO2 (e.g., 1.52 eV) from the previous higherfilter potential energy barrier heights (e.g., 3.3 eV for graphene/Al2O3 and 2.05 eV for graphene/HfO2) greatly improves thecurrent gain. Presently, our prototype MoS2−HETs are notoptimized in terms of their device layout and materialparameters (e.g., tunnel and filter oxide thickness) for highfrequency operation. Their estimated cutoff frequency (fT) andmaximum oscillation frequency ( fmax) are calculated using theequations f T ∼ 1/(2πτeff) and fmax ∼ ( f T/8πRBCBC)

1/2, which is7.6 MHz for the f T and 1.1 MHz for the fmax, respectively. Theτeff = τBE + τBC (τ = RC) we used is roughly determined to be21 × 10−9 s, where τBE = 15 × 10−9 s and τBC = 6 × 10−9 s, theRC for the two junctions is from the resistance we assumed (RB= RC = 50 Ohms), and the calculated parallel plate capacitance(CBE = 300 pF and CBC = 120 pF) between the base and otherelectrodes according to the device structure. The MoS2-HETsmay be further optimized for high frequency performance byproperly engineering the following device parameters: a reliableand pinhole-free tunnel barrier with low barrier height (e.g.,conduction band offset) and thin barrier width, minimizing thebase resistance of MoS2, and minimizing the parasiticcapacitances throughout the device structure.In conclusion, we have demonstrated a novel vertical hot-

electron transistor incorporating single-layer MoS2 in the baseregion. This MoS2−HET operates at room temperature andexhibits an unprecedentedly high common-base current gain αof about 0.95 over the entire VBE bias range, which can bedynamically tuned around 2 orders of magnitude by varyingVCB. We reiterate the fact that this current gain dependence onVCB is detrimental in terms of transistor performance metrics(e.g., poor output resistance and poor intrinsic voltage gain)but it is a novel phenomenon inherent to our prototype MoS2−HETs and worth investigating. The majority of the fabricationprocess was designed to be compatible with CMOS technology.All things considered, there exist several features of thisprototype transistor which can benefit from further optimiza-tion. First, its low tunneling current density (JEB ∼ 10 mAcm−2)can be improved around 2 orders of magnitude by reducing theSiO2 tunneling barriers to be about 1 nm. High current levels inaddition to high-current gain are essential toward realizing high-frequency applications in the future. Second, the hetero-

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7910

structure materials used in this work can be extended to othermaterial combinations to further enhance the performance aswell as functionalities. By implementing different 2D materialsfor the emitter-base tunneling barrier (e.g., hexagonal boronnitride), base (e.g., other transition metal dichalcogenides), andcollector (e.g., graphene) on flexible substrates, a new family offlexible hot-carrier electronics could emerge. Third, thepresently high base resistance of the MoS2 base region couldbe reduced by phase engineering the MoS2 from itssemiconducting 2H phase to its metallic 1T phase.39 Althoughvery little is known about the exact vertical transport behaviorand potential energy landscape of hot-electrons in the directionnormal to the 2D plane,40−42 a trade-off most likely existsbetween implementing transition metal dichalcogenides in thebase region for high-current gain or graphene in the base regionfor potentially higher frequency applications. Thus, the furtheroptimization of the device layout and structure as well as thematerial interface properties of our MoS2−HETs may leadtoward the realization of 2D material-based high-density, low-energy, and high-frequency hot-carrier complementary tran-sistors for complementary-logic circuits. Moreover, the MoS2−HETs presented in this work were fabricated from large-areaCVD single-layer MoS2, which promotes the scalability andpotential commercialization of such devices.43

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.nano-lett.5b03768.

Details of the fabrication process flow, the Ramanspectrum of the monolayer CVD MoS2, a schematic ofthe current injection scheme employed, and opticalimages of various MoS2−HETs isolated from each other(DOCX)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: Y.-W.L. (ywlanblue@gmail.com).*E-mail: C.M.T., Jr. (torresjrcarlos@gmail.com).*E-mail: K.L.W. (wang@seas.ucla.edu).

Present AddressC.M.T., Jr.: Space and Naval Warfare (SPAWAR) SystemsCenter Pacific, San Diego, California 92152, United States.

Author ContributionsC.M.T., Jr., and Y.-W.L. contributed equally to this work.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was in part supported by the National ScienceFoundation (NSF) under Award # NSF-EFRI-1433541. J.R.A,M.B.L., and C.M.T., Jr., acknowledge support from the Officeof Naval Research and SPAWAR Systems Center Pacific's In-House Laboratory Independent Research program.” C.M.T., Jr.,thanks the Department of Defense SMART (Science,Mathematics, and Research for Transformation) Scholarshipfor graduate scholarship funding. This research was funded inpart by the National Science Council of Taiwan under contractno. NSC 103-2917-I-564-017.

■ REFERENCES(1) Mead, C. A. The tunnel emission amplifier. Proc. IRE 1960, 48,359.(2) Mead, C. A. Operation of tunnel-emission devices. J. Appl. Phys.1961, 32, 646.(3) Heiblum, M. Tunneling hot-electron transfer amplifiers(THETA): Amplifiers operating up to the infrared. Solid-State Electron.1981, 24, 343−366.(4) Heiblum, M.; Fischetti, M. V. Ballistic hot-electron transistors.IBM J. Res. Dev. 1990, 34 (4), 530−549.(5) Heiblum, M.; Nathan, M. I.; Thomas, D. C.; Knoedler, C. M.Direct observation of ballistic transport in GaAs. Phys. Rev. Lett. 1985,55, 2200.(6) Levi, A. F. J.; Chiu, T. H. Room-temperature operation of hot-electron transistors. Appl. Phys. Lett. 1987, 51, 984.(7) Moise, T. S.; Kao, Y.-C.; Seabaugh, A. C. Room-temperatureoperation of a tunneling hot-electron transfer amplifier. Appl. Phys.Lett. 1994, 64, 1138.(8) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang,Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric field effect inatomically thin carbon films. Science 2004, 306 (5696), 666−669.(9) Geim, A. K.; Grigorieva, I. V. Van der Waals heterostructures.Nature 2013, 499, 419−425.(10) Novoselov, K. S.; Jiang, D.; Schedin, F.; Booth, T. J.;Khotkevich, V. V.; Morozov, S. V.; Geim, A. K. Two-dimensionalatomic crystals. Proc. Natl. Acad. Sci. U. S. A. 2005, 102 (30), 10451−10453.(11) Xu, K.; Zeng, C.; Zhang, Q.; Yan, R.; Ye, P.; Wang, K.;Seabaugh, A. C.; Xing, H. G.; Suehle, J. S.; Richter, C. A.; Gundlach, D.J.; Nguyen, N. V. Direct measurement of Dirac point energy at thegraphene/oxide interface. Nano Lett. 2013, 13, 131−136.(12) Zeng, C.; Wang, M.; Zhou, Y.; Lang, M.; Lian, B.; Song, E.; Xu,G.; Tang, J.; Torres, C. M., Jr.; Wang, K. L. Tunneling spectroscopy ofmetal-oxide-graphene structure. Appl. Phys. Lett. 2010, 97, 032104.(13) Lemme, M. C.; Li, L.-J.; Palacios, T.; Schwierz, F. Two-dimensional materials for electronic applications. MRS Bull. 2014, 39(8), 711−718.(14) Mehr, W.; Dabrowski, J.; Scheytt, J. C.; Lippert, G.; Xie, Y.-H.;Lemme, M. C.; Ostling, M.; Lupina, G. Vertical graphene basetransistor. IEEE Electron Device Lett. 2012, 33, 691.(15) Vaziri, S.; Lupina, G.; Henkel, C.; Smith, A. D.; Ostling, M.;Dabrowski, J.; Lippert, G.; Mehr, W.; Lemme, M. C. A graphene-basedhot-electron transistor. Nano Lett. 2013, 13, 1435.(16) Kong, B. D.; Zeng, C.; Gaskill, D. K.; Wang, K. L.; Kim, K. W.Two dimensional crystal tunneling devices for THz operation. Appl.Phys. Lett. 2012, 101, 263112.(17) Kong, B. D.; Jin, Z.; Kim, K. W. Hot-electron transistors forterahertz operation based on two dimensional crystal heterostructures.Phys. Rev. Appl. 2014, 2, 054006.(18) Di Lecce, V.; Grassi, R.; Gnudi, A.; Gnani, E.; Reggiani, S.;Baccarani, G. Graphene-base heterojunction transistor: An attractivedevice for terahertz operation. IEEE Trans. Electron Devices 2013, 60,4263.(19) Driussi, F.; Palestri, P.; Selmi, L. Modeling, simulation, anddesign of the vertical graphene base transistor. Microelectron. Eng.2013, 109, 338−341.(20) Venica, S.; Driussi, F.; Palestri, P.; Esseni, D.; Vaziri, S.; Selmi, L.Simulation of DC and RF performance of the graphene base transistor.IEEE Trans. Electron Devices 2014, 61, 7.(21) Venica, S.; Driussi, F.; Palestri, P.; Selmi, L. Graphene basetransistors with optimized emitters and dielectrics. Electronics andMicroelectronics (MIPRO), 37th International Convention on Informationand Communication Technology 2014, 33−38.(22) Gong, C.; Zhang, H.; Wang, W.; Colombo, L.; Wallace, R. M.;Cho, K. Band alignment of two-dimensional transition metaldichalcogenides: Application in tunnel field effect transistors. Appl.Phys. Lett. 2013, 103, 053513.(23) Yang, H.; Heo, J.; Park, S.; Song, H. J.; Seo, D. H.; Byun, K.-E.;Kim, P.; Yoo, I.; Chung, H.-J.; Kim, K. Graphene barristor, a triode

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7911

device with a gate-controlled Schottky barrier. Science 2012, 336,1140−1143.(24) Zeng, C.; Song, E. B.; Wang, M.; Lee, S.; Torres, C. M., Jr.;Tang, J.; Weiller, B. H.; Wang, K. L. Vertical graphene-base hot-electron transistor. Nano Lett. 2013, 13, 2370.(25) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomicallythin MoS2: A new direct-gap semiconductor. Phys. Rev. Lett. 2010, 105,136805.(26) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.;Strano, M. S. Electronic and optoelectronics of two-dimensionaltransition metal dichalcogenides. Nat. Nanotechnol. 2012, 7, 699−712.(27) Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A.Single-layer MoS2 transistors. Nat. Nanotechnol. 2011, 6, 147−150.(28) Hamberg, I.; Granqvist, C. G.; Berggren, K.-F.; Sernelius, B. E.;Engstrom, L. Bandgap widening in heavily Sn-doped In2O3. Phys. Rev.B: Condens. Matter Mater. Phys. 1984, 30, 3240.(29) Mryasov, O. N.; Freeman, A. J. Electronic band structure ofindium tin oxide and criteria for transparent conducting behavior. Phys.Rev. B: Condens. Matter Mater. Phys. 2001, 64, 233111.(30) Hautier, G.; Miglio, A.; Ceder, G.; Rignanese, G.-M.; Gonze, X.Identification and design principles of low hole effective mass p-typetransparent conducting oxides. Nat. Commun. 2013, 4, 2292.(31) Lee, Y.-H.; Yu, L.; Wang, H.; Fang, W.; Ling, X.; Shi, Y.; Lin, C.-T.; Huang, J.-K.; Chang, M.-T.; Chang, C.-S.; Dresselhaus, M.;Palacios, T.; Li, L.-J.; Kong, J. Synthesis and transfer of single-layertransition metal disulfides on diverse surfaces. Nano Lett. 2013, 13,1852−1857.(32) Nguyen, L.-N.; Lan, Y.-W.; Chen, J.-H.; Chang, T.-R.; Zhong,Y.-L.; Jeng, H.-T.; Li, L.-J.; Chen, C.-D. Resonant tunneling throughdiscrete quantum states in stacked atomic-layered MoS2. Nano Lett.2014, 14 (5), 2381−2386.(33) Simmons, J. G. Generalized formula for the electric tunnel effectbetween similar electrodes separated by a thin insulating film. J. Appl.Phys. 1963, 34, 1793.(34) Britnell, L.; Gorbachev, R. V.; Jalil, R.; Belle, B. D.; Schedin, F.;Mishchenko, A.; Georgiou, T.; Katsnelson, M. I.; Eaves, L.; Morozov,S. V.; Peres, N. M. R.; Leist, J.; Geim, A. K.; Novoselov, K. S.;Ponomarenko, L. A. Field-effect tunneling transistor based on verticalgraphene heterostructures. Science 2012, 335, 947.(35) Georgiou, T.; Jalil, R.; Belle, B. D.; Britnell, L.; Gorbachev, R. V.;Morozov, S. V.; Kim, Y.-J.; Gholinia, A.; Haigh, S. J.; Makarovsky, O.;Eaves, L.; Ponomarenko, L. A.; Geim, A. K.; Novoselov, K. S.;Mishchenko, A. Vertical field-effect transistor based on graphene-WS2heterostructures for flexible and transparent electronics. Nat. Nano-technol. 2012, 8, 100−103.(36) Xiong, K.; Robertson, J.; Gibson, M. C.; Clark, S. J. Defectenergy levels in HfO2 high-dielectric-constant gate oxide. Appl. Phys.Lett. 2005, 87, 183505.(37) Wang, H.; Zhang, C.; Rana, F. Ultrafast dynamics of defect-assisted electron-hole recombination in monolayer MoS2. Nano Lett.2015, 15, 339−345.(38) Zhu, W.; Low, T.; Lee, Y.-H.; Wang, H.; Farmer, D. B.; Kong, J.;Xia, F.; Avouris, P. Electronic transport and device prospects ofmonolayer molybdenum disulphide grown by chemical vapordeposition. Nat. Commun. 2014, 5, 3087.(39) Kappera, R.; Voiry, D.; Yalcin, S. E.; Branch, B.; Gupta, G.;Mohite, A. D.; Chhowalla, M. Phase-engineered low-resistancecontacts for ultrathin MoS2 transistors. Nat. Mater. 2014, 13, 1128−1134.(40) Zhang, Q.; Fiori, G.; Iannaccone, G. On transport in verticalgraphene heterostructures. IEEE Electron Device Lett. 2014, 35, 9.(41) Iannaccone, G.; Zhang, Q.; Bruzzone, S.; Fiori, G. Relevance ofthe physics of off-plane transport through 2D materials on the designof vertical transistors. 2015 Joint International EUROSOI Workshop andInternational Conference on Ultimate Integration on Silicon (EUROSOI-ULIS 2015 2015, 89−92.(42) Bruzzone, S.; Logoteta, D.; Fiori, G.; Iannaccone, G. Verticaltransport in graphene-hexagonal boron nitride heterostructure devices.Sci. Rep. 2015, 5, 14519.

(43) Kang, K.; Xie, S.; Huang, L.; Han, Y.; Huang, P. Y.; Mak, K. F.;Kim, C.-J.; Muller, D.; Park, J. High-mobility three-atom-thicksemiconducting films with wafer-scale homogeneity. Nature 2015,520, 656−660.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b03768Nano Lett. 2015, 15, 7905−7912

7912