High-Current Gain Two-Dimensional MoS2‑Base

Nov 2, 2015 - been plethora of HET variants implementing diverse material systems.1−7 Usually, these HETs feature substantial current gain (α ∼ 0...
0 downloads 7 Views 3MB Size
Download PDF
Letter pubs.acs.org/NanoLett

High-Current Gain Two-Dimensional MoS2‑Base Hot-Electron Transistors Carlos 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 to milestones in electronics with the development of the hot-electron transistor. Recently, significant advances have been made with atomically sharp heterostructures implementing various two-dimensional materials. Although graphene-base hot-electron transistors show great promise for electronic switching at high frequencies, they are limited by their low current gain. Here we show that, by choosing MoS2 and HfO2 for the filter barrier interface and using a noncrystalline semiconductor such as ITO for the collector, we can achieve an unprecedentedly high-current gain (α ∼ 0.95) in our hot-electron transistors operating at room temperature. Furthermore, the current gain can be tuned over 2 orders of magnitude with the collector-base voltage albeit this feature currently presents a drawback in the transistor performance metrics such as poor output resistance and poor intrinsic voltage gain. We anticipate our transistors will pave the way toward the realization of novel flexible 2D material-based highdensity, low-energy, and high-frequency hot-carrier electronic applications. KEYWORDS: 2D materials, transition metal dichalcogenides, MoS2, hot-electron transport, high-current gain

F

grown by molecular beam epitaxy (MBE). HETs implementing two-dimensional (2D) materials,8−10 such as graphene,8,11,12 in the base region13−15 have recently shown great promise for ultrahigh frequency operation.14,16−21 These vertical transport three-terminal electronic devices can be designed with atomically sharp heterostructures by the stacking of various van der Waals materials.9 This allows one to play with the conduction and valence band offsets,22 which determine the potential landscape experienced by hot-carriers and ultimately the device performance.23 Until now, only vertical graphenebase hot-electron transistors have been experimentally demon-

or over half a century, Moore’s law has driven the silicon electronics 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 sought out and investigated with hopes to augment the next generation of information processing. One promising device concept is the hot-electron transistor (HET),1−7 which relies on the vertical transport of a controlled source of hot-electrons. Ever since Mead first proposed this device concept in 1960,1,2 there have been plethora of HET variants implementing diverse material systems.1−7 Usually, these HETs feature substantial current gain (α ∼ 0.75) at cryogenic temperatures (T = 4.2 K) but very poor current gain at room temperature.3,4 Only a few HETs have shown high current gain (α ∼ 0.9) at room temperature6,7 but rely on precise yet complicated epitaxial layered structures © 2015 American Chemical Society

Received: July 25, 2015 Revised: October 23, 2015 Published: November 2, 2015 7905

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters

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, and C 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 dashed circle outlines the MoS2 region. (d) Common-base configuration circuit incorporating the schematic symbol for the MoS2−HET device.

doped n++ silicon substrate (ND ∼ 1019 cm−3) as the emitter (E), a monolayer of chemical vapor deposition (CVD) grown MoS2 (see Supporting Information, Figure S1) as the base (B), and sputtered (∼45 nm) ITO as the collector (C). The MoS2 was grown using a CVD method31 and transferred onto the substrate using a PMMA transfer method.32 A thermally grown thin (∼3 nm) SiO2 tunnel barrier separates the emitter and base terminals, whereas an atomic-layer deposited (∼55 nm) HfO2 separates the base and collector and serves as the filtering barrier. In this fabrication process, arrays of MoS2−HET devices are isolated from each other via a thick (∼300 nm) SiO2 field oxide. The underlying silicon surface which is not covered by the thick SiO2 field oxide, but rather surrounded by it, serves to confine the current into the emitter region of the device for hot-electron injection through the SiO2 tunnel barrier (see Supporting Information, Figure S2). Accordingly, the fabrication process was designed to be compatible with silicon CMOS technology. An optical micrograph of an actual MoS2−HET (top-view) is presented in Figure 1c. The detailed fabrication process is described in the Supporting Information. In this particular study, a common-base configuration was employed during the electrical measurements and the biasing circuit incorporating the schematic symbol for the MoS2−HET device is shown in Figure 1d. Note that both of the base contacts are grounded during the electrical measurements in order to achieve a uniform potential distribution across the MoS2 base region. A common-emitter configuration was also employed during the electrical measurements for a few devices

strated,15,24 yet their transport characteristics feature a significantly low common-base current gain (α ∼ 10−2). These shortcomings preclude the fulfillment of realizing 2D material-based vertical hot-carrier transistors operating at high frequencies. As a first step toward this goal, we propose and demonstrate a novel device concept which enables unprecedentedly high current gain in 2D material-based hot-electron transistors. In this Letter, we demonstrate a novel vertical hotelectron transistor incorporating single-layer MoS225−27 in the base region (MoS2−HET). To the best of our knowledge, all previous vertical graphene-base hot-electron transistors implemented a metal for the collector electrode and exhibited an extremely low current gain. In this work, by utilizing a noncrystalline semiconductor such as ITO (an n-type transparent conducting oxide) as the collector electrode,28−30 we demonstrate for the first time that the MoS2−HETs operate at room temperature and exhibit a high common-base current gain (α ∼ 0.95) over the entire base-emitter bias (VBE) range. Furthermore, the current gain can be dynamically tuned over 2 orders of magnitude with the collector-base voltage albeit this feature currently presents a drawback in the transistor performance metrics such as poor output resistance and poor intrinsic voltage gain. The device structure and schematic of the MoS2−HET are introduced in Figure 1. An isometric view of the MoS2−HET device structure is shown in Figure 1a, and a cross-sectional view of its vertical heterostructure active region is depicted in Figure 1b. The three-terminal device consists of a degenerately 7906

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters

Figure 2. Energy band diagrams for the operating conditions of the MoS2−HET. (a) Flat band condition. Note that the conduction band offset, or the 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 are depicted 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 band diagram depicting the on-state condition. For VCB > 0 (dashed red lines), the hot-electrons tunneling through the emitter-base tunnel barrier have sufficient kinetic energy to overcome the filter barrier and reach the collector.

where f(E) is the Fermi distribution function, ρ1(E) is the density of states of the first electrode, ρ2(E) is the density of states of the second electrode, and T(E) is the transmission probability which, for the MoS2−HET device structure, is dominated 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:

in order to confirm the consistent high-current gain results presented in this work. In order to clearly understand the physics, we first focus on describing the two modes of operation for the MoS2−HET using energy band diagrams. The flat band condition is shown in Figure 2a. The conduction band offset between the monolayer MoS2 and HfO2 is 1.52 eV. This forms the filter potential barrier height (Δe) for the hot-electrons. The filter potential barrier height and width at the collector-base junction are important parameters that determine the collector current after the hot-electrons tunnel through the 3 nm SiO2 emitterbase tunnel barrier. In the previous graphene-base hot-electron transistors,15,24 the filter barrier height between graphene and Al2O3 was 3.3 eV,15,24 whereas between graphene and HfO2 it was 2.05 eV.24 In both prior cases, the filter barrier heights are greater than that between MoS2 and HfO2 (1.52 eV), thus an improvement of the ratio, in our case, between the collector current and the emitter current (α = IC/IE) is expected. In general, the potential barrier height, the potential barrier width, and the strength of the applied electric field are key parameters that determine the tunneling current across a tunnel junction.33−35 The tunneling current across a tunnel junction is generally described by33−35 I(V ) ∝

∫ dE·ρ1(E)·ρ2 (E − eV )·[f (E − eV ) − f (E)]·T(E)

T (E) ∼ e−W (E)

(2)

where W(E) is related to the effective width and height of the potential barrier and strongly depends on the strength of the electric field. W(E) can be generalized within the WKB approximation:34 W=2

∫0

d

dx·Im kz(Δ(x))

(3)

Thus, the application of a strong electric field can dramatically alter the shape (e.g., both the effective height and width) of the potential barrier (eq 3) and result in an increased collector current. Moreover, when designing vertical hot-electron transistors,36−38 it is paramount to choose the proper combination of 2D material for the base as well as the filtering barrier dielectric, which yields the desired conduction band offset,22 or filter potential barrier height, for the hotelectrons. Instead of relying on complicated and expensive

(1) 7907

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters

Figure 3. Electrical characterization of the MoS2−HET in the common-base configuration. (a) Energy band diagram depicting MoS2−HETs. The conduction 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 a function 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.

methods to produce atomically sharp interfaces,1−7 it is now possible to design the equilibrium filter potential barrier height in vertical transport devices by choosing from plethora of 2D materials since the conduction band offset results from material specific parameters such as the electron affinity of the collectoroxide and the work function (e.g., in the case of graphene) or electron 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 device transport. The current components for the hot-electron contribution (red arrow) to the total current flow through the MoS2−HET are shown in Figure 2b. The emitter current (IE) across the SiO2 tunnel oxide is due to hot-electrons injected from the n++ silicon substrate. Furthermore, the collector current (IC) across the HfO2 collector-base oxide is due to the portion of the injected hot electrons with enough kinetic energy that surpass the filter barrier and reach the collector. Now that the current components and the barrier heights experienced by the hot-electrons in the MoS2−HETs have been shown, we proceed to describe the modes of operation of these novel transistors. Figure 2c shows the energy band diagram for the off-state condition of the MoS2−HET. In the absence of an applied VCB, the hot-electrons injected through the tunnel oxide have insufficient kinetic energy to overcome the filter barrier at the collector-base junction and do not reach the collector. Instead, they backscatter and thermalize into the MoS2 base region. However, the situation drastically changes with the

application of a large positive VCB. Figure 2d shows the energy band diagram for the on-state condition of the MoS2−HET. In this scenario, hot-electrons tunneling through the emitter-base tunnel oxide have sufficient kinetic energy to overcome the filter barrier and reach the collector. Based on the physical concepts just described, the device performances of the MoS2−HETs were characterized using the common-base configuration. In the following, we characterize the MoS2−HET by applying positive VCB. Figure 3a shows the energy band diagram depicting the MoS2−HET. Specifically, Figure 3a shows the conduction and valence band edges at the collector-base junction with a positive VCB applied. In this condition, once hot-electrons tunneling through the emitterbase tunnel barrier have sufficient kinetic energy, they can vertically transport through the MoS2 base region, surpass the filter barrier at the collector-base junction, and reach the collector. Consequently, an increasingly positive VCB will continue to effectively make the filter potential barrier thinner and promote hot-electrons reaching the collector due to an increase in their transmission probability. This qualitative behavior is exhibited in the input and transfer characteristics of the MoS2−HETs. The input characteristics (IE−VBE) correspond to how the emitter current depends on VBE, whereas the transfer characteristics (IC−VBE) correspond to the manner in which the collector current varies with VBE. Figure 3b shows the input and transfer characteristics for Device 1. In this device, the maximum VBE is limited to 3 V to avoid dielectric breakdown of the tunnel oxide. The emitter current (IE) and the collector current (IC) are shown as a function of VBE (VBE was swept from 0 to +3 V) at a VCB of +1 V. Both currents 7908

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters

Figure 4. Output characteristics and tunable current gain of the MoS2−HET. (a) Common-base output characteristics for Device 2. The collector current 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 a function 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 commonemitter configuration. The collector and base currents are shown as a function of VBE at a fixed output voltage of VCE = +10 V. (b) The commonemitter 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 shows the 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, and it can be tuned with the output voltage VCE.

α, monotonically increases with positive VCB due to the reduced potential barrier and associated increase in the transmission probability of hot-electrons and exhibits a nearly constant characteristic throughout the entire VBE range with an average magnitude of about 95% at VCB = +8 V. With the analysis of the input and transfer characteristics complete, we now investigate the common-base output characteristics of the MoS2−HETs, which correspond to how the output collector current depends on VCB. Figure 4a shows the common-base output characteristics for Device 2. The collector current is shown as a function of VCB at three positive VBE biases. The collector current is insensitive to modulation below a critical electric field, or correspondingly a VCB voltage, across the HfO2 collector-base oxide. However, above a critical electric field across the HfO2, the collector current is quite sensitive to modulation and rapidly increases with a further increase in VCB. In order to convey the robust nature of this high current gain in the MoS2−HETs, Figure 4b shows a semilog plot of α as a function of VCB at VBE = +3 V. It clearly shows that α increases with an increasingly positive VCB and can be tuned over an order of magnitude since this lowers the filter potential barrier experienced by the hot-electrons and allows them to reach the collector. Remarkably, the room temperature

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 various positive VCB. It is evident that the collector current increases with increasingly positive VCB. This is due to the fact that the filter potential barrier width at the collector-base junction is effectively reduced as the applied VCB becomes more positive. Correspondingly, a greater portion of the injected hot-electrons from the emitter have high enough kinetic energy to vertically transport through the MoS2 base region, surpass the filter barrier, and reach the collector, thus contributing to an increasing collector current. From the input and transfer characteristics, we can next ascertain the common-base current gain (α) of Device 1, which is a figure of merit for HETs and is defined as α = IC/IE. Figure 3d shows α as a function of VBE at VCB = +8 V. Interestingly, the current gain, α, features a nearly constant characteristic at all 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 the injected hot-electrons ballistically traverse the single-layer MoS2 base region at room temperature. The inset of Figure 3d shows a family of α characteristics as a function of VBE at several positive VCB (VCB = 0, +2, +4, +6, and +8 V). The current gain, 7909

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters

may be achieved if a filter oxide with a larger dielectric strength and/or a reduced thickness is used in order to promote larger accelerating electric fields across the filter oxide. In fact, the current gain dependence on VCB indicates that the output resistance and the intrinsic voltage gain of our MoS2−HETs are poor. Based on our measurements, the input and output resistances are on the order of a GΩ and a few tens of GΩ, respectively. The intrinsic voltage gain in these prototype MoS2−HETs is about 3, which is quite small. In order to improve the intrinsic voltage gain, a low input impedance is desirable. In our device structure, a low input impedance can be readily achieved by further reducing the thickness of the SiO2 tunnel barrier to around 1 nm. The ensuing tunneling current density (IE) should improve by at least 2 orders of magnitude and thus enable a much higher intrinsic voltage gain in our MoS2−HETs. Furthermore, reducing the thickness of the tunnel oxide would significantly benefit the driving capability and speed of the MoS2−HETs compared to the current batch which feature very small device current levels. It is worth mentioning that the MoS2−HETs in this work were intentionally designed for DC electrical characterization in order to prove that decreasing the filter potential energy barrier height using MoS2 and HfO2 (e.g., 1.52 eV) from the previous higher filter potential energy barrier heights (e.g., 3.3 eV for graphene/ Al2O3 and 2.05 eV for graphene/HfO2) greatly improves the current gain. Presently, our prototype MoS2−HETs are not optimized in terms of their device layout and material parameters (e.g., tunnel and filter oxide thickness) for high frequency operation. Their estimated cutoff frequency (fT) and maximum oscillation frequency (f max) are calculated using the equations f T ∼ 1/(2πτeff) and f max ∼ (f T/8πRBCBC)1/2, which is 7.6 MHz for the f T and 1.1 MHz for the f max, respectively. The τeff = τBE + τBC (τ = RC) we used is roughly determined to be 21 × 10−9 s, where τBE = 15 × 10−9 s and τBC = 6 × 10−9 s, the RC 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 other electrodes according to the device structure. The MoS2-HETs may be further optimized for high frequency performance by properly engineering the following device parameters: a reliable and pinhole-free tunnel barrier with low barrier height (e.g., conduction band offset) and thin barrier width, minimizing the base resistance of MoS2, and minimizing the parasitic capacitances throughout the device structure. In conclusion, we have demonstrated a novel vertical hotelectron transistor incorporating single-layer MoS2 in the base region. This MoS2−HET operates at room temperature and exhibits an unprecedentedly high common-base current gain α of about 0.95 over the entire VBE bias range, which can be dynamically tuned around 2 orders of magnitude by varying VCB. We reiterate the fact that this current gain dependence on VCB 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 fabrication process was designed to be compatible with CMOS technology. All things considered, there exist several features of this prototype transistor which can benefit from further optimization. First, its low tunneling current density (JEB ∼ 10 mAcm−2) can be improved around 2 orders of magnitude by reducing the SiO2 tunneling barriers to be about 1 nm. High current levels in addition to high-current gain are essential toward realizing highfrequency applications in the future. Second, the hetero-

common-base current gain (α) in this type of novel 2D material-based vertical device is unprecedentedly high for the largest VCB applied. Next, we investigate the MoS2−HET characteristics when biased in the common-emitter configuration in order to corroborate the high and tunable current gain we achieved in the common-base configuration. The Gummel plot is used as a figure of merit when analyzing bipolar transistors. It is a simultaneous semilog plot of the collector and base currents as a function of the input voltage (VBE) at a fixed output voltage (VCE). The common-emitter current gain (β = IC/IB) can be ascertained from the Gummel plot by taking the ratio of the collector current to the base current at a fixed VBE. Figure 5a shows the Gummel plot for Device 1 when biased in the common-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. The Gummel plot confirms the transistor action of the MoS2−HET as the input base current is directly amplified to the output collector current. Finally, Figure 5b shows the common-emitter output characteristics for Device 1. The collector current is shown as a function of VCE at three positive VBE biases in addition to VBE = 0 V. The inset shows the common-emitter current gain (β) as a function of VCE at VBE = +2 V. A maximum common-emitter current gain of around 4 is achieved, and it can be tuned with the output voltage VCE. We note that a slightly different current gain is obtained between the common-emitter and the common-base configuration measurements. We speculate that this discrepancy is due to the differences in the measurement circuits biasing the actual device structure in the two measurement configurations. Applying an output voltage (VCB) in the common-base configuration results in the collector current originating from the ITO collector electrode, traversing through the HfO2 filter barrier dielectric, and laterally transporting through the MoS2 surface into the grounded base contacts. This is because VCB actually only controls the band bending of the HfO2 layer, and it does not modulate the SiO2/MoS2 layers due to the fact that our base contacts lie on the top surface of the MoS2. Thus, α is mainly limited by the recombination processes inside the HfO2 dielectric, which is quite small.36 However, applying an output voltage (VCE) in the common-emitter configuration results in the collector current traversing through the entire ITO/HfO2/ MoS2/SiO2/n++ silicon vertical heterostructure, since band bending occurs throughout the entire heterostructure. Furthermore, since there exist more pronounced non-radiative recombination (e.g., Auger recombination and trap-states) processes in the MoS2 layer,37,38 the measured β in the common-emitter configuration is thus expected to be smaller than the measured α in the common-base configuration. Nonetheless, by biasing the MoS2−HETs in both the commonbase and the common-emitter configurations, we have explicitly shown that the measured current gains (α, β) in either scenario corroborate each other and further attest the high and tunable current gain in our MoS2−HETs. On the other hand, for transistor operation, it is better to operate in the saturation region of the output characteristics where the collector current (IC) is fairly constant. However, the lack of any reasonable saturation in the output characteristics of the MoS2−HETs (see Figure 5b) may be due to the onset of dielectric breakdown in both the SiO2 tunnel oxide and the HfO2 filter oxide. The lack of saturation in the output characteristics is specific to this batch of MoS2−HETs but 7910

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Nano Letters

■

structure materials used in this work can be extended to other material combinations to further enhance the performance as well as functionalities. By implementing different 2D materials for the emitter-base tunneling barrier (e.g., hexagonal boron nitride), base (e.g., other transition metal dichalcogenides), and collector (e.g., graphene) on flexible substrates, a new family of flexible hot-carrier electronics could emerge. Third, the presently high base resistance of the MoS2 base region could be reduced by phase engineering the MoS2 from its semiconducting 2H phase to its metallic 1T phase.39 Although very little is known about the exact vertical transport behavior and potential energy landscape of hot-electrons in the direction normal to the 2D plane,40−42 a trade-off most likely exists between implementing transition metal dichalcogenides in the base region for high-current gain or graphene in the base region for potentially higher frequency applications. Thus, the further optimization of the device layout and structure as well as the material interface properties of our MoS2−HETs may lead toward the realization of 2D material-based high-density, lowenergy, and high-frequency hot-carrier complementary transistors for complementary-logic circuits. Moreover, the MoS2− HETs presented in this work were fabricated from large-area CVD single-layer MoS2, which promotes the scalability and potential commercialization of such devices.43

■

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 hotelectron transistors. Appl. Phys. Lett. 1987, 51, 984. (7) Moise, T. S.; Kao, Y.-C.; Seabaugh, A. C. Room-temperature operation 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 in atomically 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-dimensional atomic 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 the graphene/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 of metal-oxide-graphene structure. Appl. Phys. Lett. 2010, 97, 032104. (13) Lemme, M. C.; Li, L.-J.; Palacios, T.; Schwierz, F. Twodimensional 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 base transistor. 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-based hot-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 for terahertz 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 attractive device for terahertz operation. IEEE Trans. Electron Devices 2013, 60, 4263. (19) Driussi, F.; Palestri, P.; Selmi, L. Modeling, simulation, and design 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 base transistors with optimized emitters and dielectrics. Electronics and Microelectronics (MIPRO), 37th International Convention on Information and 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 metal dichalcogenides: 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

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b03768. Details of the fabrication process flow, the Raman spectrum of the monolayer CVD MoS2, a schematic of the current injection scheme employed, and optical images of various MoS2−HETs isolated from each other (DOCX)

■

Letter

AUTHOR INFORMATION

Corresponding Authors

*E-mail: Y.-W.L. ([email protected]). *E-mail: C.M.T., Jr. ([email protected]). *E-mail: K.L.W. ([email protected]). Present Address

C.M.T., Jr.: Space and Naval Warfare (SPAWAR) Systems Center Pacific, San Diego, California 92152, United States. Author Contributions

C.M.T., Jr., and Y.-W.L. contributed equally to this work. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was in part supported by the National Science Foundation (NSF) under Award # NSF-EFRI-1433541. J.R.A, M.B.L., and C.M.T., Jr., acknowledge support from the Office of Naval Research and SPAWAR Systems Center Pacific's InHouse Laboratory Independent Research program.” C.M.T., Jr., thanks the Department of Defense SMART (Science, Mathematics, and Research for Transformation) Scholarship for graduate scholarship funding. This research was funded in part by the National Science Council of Taiwan under contract no. NSC 103-2917-I-564-017. 7911

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912

Letter

Nano Letters 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 hotelectron transistor. Nano Lett. 2013, 13, 2370. (25) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomically thin 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-dimensional transition 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 of indium 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-type transparent 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-layer transition 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 through discrete quantum states in stacked atomic-layered MoS2. Nano Lett. 2014, 14 (5), 2381−2386. (33) Simmons, J. G. Generalized formula for the electric tunnel effect between 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 vertical graphene 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-WS2 heterostructures for flexible and transparent electronics. Nat. Nanotechnol. 2012, 8, 100−103. (36) Xiong, K.; Robertson, J.; Gibson, M. C.; Clark, S. J. Defect energy levels in HfO2 high-dielectric-constant gate oxide. Appl. Phys. Lett. 2005, 87, 183505. (37) Wang, H.; Zhang, C.; Rana, F. Ultrafast dynamics of defectassisted 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 of monolayer molybdenum disulphide grown by chemical vapor deposition. 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-resistance contacts for ultrathin MoS2 transistors. Nat. Mater. 2014, 13, 1128− 1134. (40) Zhang, Q.; Fiori, G.; Iannaccone, G. On transport in vertical graphene heterostructures. IEEE Electron Device Lett. 2014, 35, 9. (41) Iannaccone, G.; Zhang, Q.; Bruzzone, S.; Fiori, G. Relevance of the physics of off-plane transport through 2D materials on the design of vertical transistors. 2015 Joint International EUROSOI Workshop and International Conference on Ultimate Integration on Silicon (EUROSOIULIS 2015 2015, 89−92. (42) Bruzzone, S.; Logoteta, D.; Fiori, G.; Iannaccone, G. Vertical transport 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-thick semiconducting films with wafer-scale homogeneity. Nature 2015, 520, 656−660.

7912

DOI: 10.1021/acs.nanolett.5b03768 Nano Lett. 2015, 15, 7905−7912