Precise p-type and n-type doping of two-dimensional semiconductors for monolithic integrated circuits | Nature Communications

Nov 08, 2024

Nature Communications volume 15, Article number: 9631 (2024) Cite this article

Metrics details

The controllable fabrication of patterned p-type and n-type channels with precise doping control presents a significant challenge, impeding the realization of complementary metal-oxide-semiconductor (CMOS) logic using a single van der Waals material. However, such an achievement could offer substantial benefits by enabling continued transistor scaling and unprecedented interlayer interconnect technologies. In this study, we devise a precise method for two-dimensional (2D) semiconductor substitutional doping, which allows for the production of wafer-scale 2H-MoTe2 thin films with specific p-type or n-type doping. Notably, we extend this approach to the synthesis of spatially selective doped 2H-MoTe2 thin films via a one-step growth method, facilitating the monolithic integration of p-type and n-type semiconductor channels. Leveraging this advancement, we successfully fabricate a chip-sized 2D CMOS inverter array that demonstrates excellent device performance and yield. Collectively, these findings represent a significant stride towards the practical incorporation of 2D semiconductors in very large-scale integration technology.

Doping, which involves the deliberate introduction of specific impurity atoms in controlled amounts into a semiconductor to modify its electrical properties, is the real power in the fabrication of various semiconductor devices. For instance, in the case of silicon, dopants such as phosphorus or boron can be added prior to the melting process to produce single-crystal silicon wafers with different conductivity types and levels via the Czochralski method. Furthermore, ion implantation is a fundamental technique in silicon-based integrated circuits, allowing for precise doping of selected regions to create hole-conducting p-type regions and electron-conducting n-type regions. The ability to precisely dope p-, n-type regions is particularly important in adjusting the threshold voltage of complementary metal-oxide-semiconductor (CMOS) field-effect transistors (FETs)1. Two-dimensional (2D) semiconductors, with their atomic thickness and potential for stacking, have emerged as promising candidates for advancing Moore’s Law, especially in terms of transistor scaling and interlayer interconnections2,3,4,5,6,7,8,9,10,11,12,13. However, the realization of large-scale integrated circuits using 2D semiconductors necessitates the controllable fabrication of patterned p-type and n-type channels with precise doping control14,15,16. Unfortunately, conventional techniques like ion implantation are not applicable to 2D semiconductors17. Most reported large-area 2D semiconductor wafers, such as non-intentionally doped n-type MoS2 layers, have been prepared through lattice-matched epitaxy on specially treated sapphire wafers18,19. If integrated circuits are to be produced, the transfer to the device substrate becomes necessary20. Currently, a general method for controllable p-type and n-type doping of 2D semiconductors, let alone the patterned preparation of p-type and n-type channels, is still lacking21.

In this work, we present a synthesis method for achieving large-scale production of 2D semiconductor 2H-MoTe2 with spatially patternable and precisely controlled carrier types and concentrations through substitutional doping. By precisely introducing Nb or Re dopants in the initial Mo film, we can effectively tune the carrier concentration of the resulting few-layer 2H-MoTe2 thin films in the p-type and n-type regions from 1010 cm−2 to 1012 cm−2. Moreover, by employing conventional lithography and etching techniques, we successfully realize patterned p-type and n-type 2H-MoTe2 semiconductor channels through one-step tellurization of Mo films selectively incorporated with pre-defined Nb or Re dopants. Utilizing this approach, we fabricate a large-scale 2D CMOS inverter array that exhibits excellent device yield. A representative 2D CMOS inverter within the array demonstrates a voltage gain of 38.2, accompanied by a peak static power consumption of 89.5 nW at Vdd = 4 V.

Substitutional doping, as opposed to charge-transfer doping through surface molecular adsorption, offers a stable and reliable approach for modifying the conductivity of 2D semiconductors. However, it is crucial to tightly control the number of dopant atoms, typically within a very small percentage range. In the synthesis of 2D semiconductors, controlling gaseous and solution dopants is challenging, often leading to doping concentrations exceeding 1 %, which results in gate-independent 2D channels with low conductivity due to Anderson localization22. Recently, high-quality 2H-MoTe2 films have been synthesized by tellurization of magnetron-sputtered Mo films through a 1T’-to-2H phase transition23. Inspired by the process of pulling doped silicon ingots, we employ magnetron co-sputtering to precisely introduce controlled amounts of Nb (for hole doping) or Re (for electron doping) into Mo films before tellurization (see method). Subsequently, the doped few-layer 2H-MoTe2 films are obtained by chemical vapor deposition (CVD) tellurization of Mo precursor films (Fig. 1a). The CVD growth process for the doped 2H-MoTe2 film follows the same steps as previously reported for unintentionally doped films23. Initially, the Mo film, incorporated with Nb or Re, undergoes tellurization to form 1T’-MoTe2, which then undergoes a phase transformation to yield doped 2H-MoTe2 (Supplementary Fig. 1). Through the tellurization chemical reaction, the sputtered Nb or Re atoms will fully replace the Mo atoms in the resulting 2H-MoTe2 structure, thereby achieving precise doping of either p-type or n-type 2D semiconductor film (further discussed later). Notably, by leveraging continuous in-plane 2D epitaxy growth, we directly synthesized 1-inch-sized thin films of Nb-doped, Re-doped, and unintentionally doped 2H-MoTe2 on amorphous silicon oxide wafers (Fig. 1b). Moreover, by combining the developed methods of seeded 2D epitaxy and heteroepitaxy for 2H-MoTe2, it becomes feasible to synthesize single-crystal wafers of these doped 2D semiconductors on arbitrary substrate architectures24,25.

a Schematic diagram of incorporating the Mo film with dopant Nb or Re firstly by co-sputtering, and then tellurizing to the doped 2H-MoTe2. The growth method is chemical vapor deposition (CVD). b Photographs of 1-inch wafers of unintentionally doped 2H-MoTe2, Nb-doped 2H-MoTe2 and Re-doped 2H-MoTe2 thin films. c Raman spectra of the unintentionally doped 2H-MoTe2, Nb-doped 2H-MoTe2 and Re-doped 2H-MoTe2. d Differential tunneling conductance (dI/dV) spectra of the Nb-doped and Re-doped 2H-MoTe2 films grown directly on highly oriented pyrolytic graphite (HOPG) substrates at 10 different positions. The dI/dV spectra were acquired at the homogeneous background (off the dopant sites) across the doped samples, clearly revealing the impact of doping on the semiconductor properties of the 2H-MoTe2 films. EF is the Fermi level in semiconductor.

The 2H nature of the synthesized films is confirmed by the observed typical 2H-MoTe2 Raman peaks (A1g 173 cm−1, \({E}_{2g}^{1}\) 234 cm−1, and \({B}_{2g}^{1}\) 289 cm−1), and the doping does not significantly shift the Raman peak positions (Fig. 1c). The thickness of the resulting doped 2H-MoTe2 (Nb or Re) film can be well controlled by the thickness of the initial Mo film. For instance, the thickness of both Nb- and Re-doped 2H-MoTe2 films can be controlled to be about 4 nm, measured by atomic force microscopy (AFM) at the edges of steps formed by scotch tape peeling (Supplementary Figs. 2 and 3). When the thickness of 2H-MoTe2 film is thinned to about 2.5 nm, some voids appear in the 2H-MoTe2 film due to the discontinuity of the initial Mo film (Supplementary Fig. 4). Meanwhile, the large-scale uniformity of the doped 2H-MoTe2 films was evaluated through Raman (0.7 mm × 0.7 mm) and AFM mappings (20 μm  × 20 μm) at 5 different locations on the wafer (Supplementary Figs. 2 and 3), demonstrating that the doped 2H-MoTe2 films are continuously uniform at large scales and with a minimal roughness.

One key issue in the development of high-performance semiconductor devices based on 2H-MoTe2 thin films is the detection of how doping affects the electronic properties of the semiconductor. Therefore, we characterized the doped 2H-MoTe2 thin films using scanning tunneling microscopy and spectroscopy (STM/STS, see method). The samples were directly grown on highly oriented pyrolytic graphite (HOPG). The differential tunneling conductance (dI/dV) spectra (Fig. 1d) clearly demonstrate the effect of doping on the semiconductor properties of 2H-MoTe2, manifested as a shift in the Fermi level (EF). For the Nb-doped (Re-doped) case, EF is closer to the valence band maximum (conduction band minimum), signifying p-type (n-type) doping nature. Considering that the only difference between the samples is the incorporation of different dopant atoms in the Mo film precursor and that they were prepared and measured under the same conditions, the influence of molecular adsorption or environmental factors on the EF shift can be ruled out. Furthermore, the randomly acquired multi-point dI/dV spectra of the Nb (Re)-doped 2H-MoTe2 samples consistently exhibit a uniform EF shift, confirming the role of Nb (Re) dopants as acceptors (donors) in the host lattice and their impact on the semiconductor properties of the entire 2H-MoTe2 film (Fig. 1d). The STM topographic images do not provide elemental resolution for substitutions atoms, but we compared the defect morphologies of the pristine and doped 2H-MoTe2 films (Supplementary Fig. 5). In the pristine 2H-MoTe2 sample, we observed four types of defects consistent with literature reports15,16: Mo vacancies (type 1), adsorbates (type 2), and oxygen-substituted tellurium atoms (type 3 and 4). In the Nb-doped sample, a new defect morphology, referred to as type 5, was observed. The density of type 5 defects, obtained statistically, was 5.7 × 1012 cm−2, which is consistent with the magnitude range of the surface density of Nb atoms predicted by the carrier concentration measured by the Hall devices (see later) and sputtering rate, suggesting that these defects may be caused by Nb atom substitution. Although the STM topography is distorted at the dopant site22,23, resulting in an inhomogeneous electric potential field and an inability to resolve atomic structures, the atomic-resolution STM images on the film exhibit a clear hexagonal 2H-MoTe2 lattice (Supplementary Fig. 5b). However, in the Re-doped sample, significantly different defect morphologies have not been observed yet, possibly due to the difficulty in distinguishing Re atom substitution from other features (e.g. types 1-4). To gain more insights into the doping-induced compositions, varied valence states, and Fermi level shifts, we further characterized the Nb-doped and Re-doped 2H-MoTe2 films by X-ray photoelectron spectroscopy (XPS), ultraviolet photoelectron spectroscopy (UPS), and Tauc plot characterizations (Supplementary Figs. 6–8), and the results are in agreement with those of STM-STS dI/dV spectra.

After determining the effective p-type and n-type doping of Nb and Re dopants, an essential consideration is achieving precise doping, specifically the resolution of the carrier concentration resulting from the doping process. To control the ratio of dopant incorporation in Mo films, we maintained a constant deposition rate of Mo while fine-tuning the deposition rate of dopants (Nb or Re) during the co-sputtering process (see method), achieving an accuracy of dopant incorporation of 0.10% for Nb and 0.06% for Re. Subsequently, 2H-MoTe2 thin films with varying carrier types and carrier concentrations were obtained through CVD growth. Hall-bar devices were fabricated for different samples to characterize their electrical properties (Fig. 2a). The Hall resistance (RHall) measurements of the Nb- and Re-doped samples exhibited opposite polarities (Fig. 2b and Supplementary Fig. 9), providing further confirmation of hole (electron) carriers in the Nb-doped (Re-doped) 2H-MoTe2 thin film. The carrier concentration was determined by calculating the inverse of the slope of the RHall versus B curve. The unintentionally doped 2H-MoTe2 displayed a low background hole concentration of ~2.66 × 1010 cm−2, indicating the high quality of 2H-MoTe2 synthesized through 2D epitaxy growth26. Upon incorporating 0.06% Re, the background hole carriers were compensated, transforming the sample into an electron-conducting material with an electron concentration of ~3.94 × 109 cm−2. This result suggests that the doping resolution can be as small as ~1010 cm−2. The electron concentration increased with higher levels of Re doping and reached ~2.7 × 1012 cm−2 when the Re incorporation ratio was 0.42% (the corresponding electron Hall mobility was 19 cm\({}^{2}/{{\rm{V}}}\cdot{{{\rm{s}}}}\)). Similarly, the hole concentration in the sample increased with higher levels of Nb doping, reaching ~2.6 × 1012 cm−2 when the Nb incorporation ratio was 0.27% (the corresponding hole Hall mobility was 20 cm\({}^{2}/{{\rm{V}}}\cdot{{{\rm{s}}}}\)). As the carrier concentration increased for both p-type and n-type doping, the conductivity of the doped 2H-MoTe2 film gradually increased, as evidenced by the four-probe I-V curves (insets of Fig. 2c). Through Hall device measurements, we once again confirmed the successful preparation of a controllably doped 2D semiconductor film of 2H-MoTe2 with precisely controlled doping types and carrier concentrations.

a Optical micrograph of a typical 2H-MoTe2 Hall device. b The magnetic field-dependent Hall resistance (RHall) of 0.27% Nb-doped and 0.42% Re-doped 2H-MoTe2 thin films, showing the conduction types of hole-conducting p-type and electron-conducting n-type. c Hole or electron carrier concentrations of 2H-MoTe2 semiconductor films with various Nb or Re incorporation ratios extracted by Hall measurements. The error bars are standard deviation. Insets: Four-terminal measured I-V characteristics of the p- and n-type 2H-MoTe2 semiconductor channels with various doping concentrations. d, e Transfer characteristics of the 2H-MoTe2 FETs with Nb and Re incorporation ratios of 0.20% and 0.24%, respectively. Vds is the source-drain voltage of the transistor.

In integrated circuit technology, precise doping is widely employed to adjust the threshold voltage of metal-oxide-semiconductor (MOS) FETs. To minimize process-induced doping effects, we fabricated back-gate FETs on 2H-MoTe2 thin films with different doping types and carrier concentrations directly grown on commercial Si/SiO2 substrates (see methods and Supplementary Fig. 10). The field-effect mobility of the doped 2H-MoTe2 can be extracted from the transfer characteristic curves (Supplementary Fig. 11). To ensure that the polarity of the semiconductor is effectively modulated by the doping and not due to the effect of the contact metal, we contacted the Nb-doped and Re-doped 2H-MoTe2 with Cr (with a low work function of 4.5 eV) and Pd (with a high work function of 5.1 eV), respectively, and measured their field-effect transistor characteristics (Supplementary Fig. 12). The 2H-MoTe2 channel doped with Nb exhibited p-type semiconductor transfer characteristics, with the conductance increasing as the doping concentration increased, and the transistor threshold voltage shifting towards positive voltages (Supplementary Fig. 13a). In contrast, for 2H-MoTe2 doped with Re, at low doping levels of 0.06% Re, the channel displayed ambipolar transport behavior, and with increasing doping concentration, the channel exhibited n-type semiconductor transport characteristics. Similarly, the conductance also increased with increasing doping, while the threshold voltage of the transistor decreased towards negative voltages (Supplementary Fig. 13b). After determining the appropriate doping concentrations, we fabricated buried-gate FETs with 20  nm Al2O3 high-k dielectric layers, 0.20% Nb-doped and 0.24% Re-doped 2H-MoTe2 channel, to enhance gate control capability (Supplementary Fig. 14). The Nb-doped and Re-doped channels exhibited typical p-type and n-type semiconductor transfer characteristics, with switch ratio ranging from 105 to 106 (Fig. 2d, e). Based on the output characteristic curves (Supplementary Fig. 14b, c), due to the influence of contact and interface engineering27,28, the on-state current of the n-MoTe2 FET (0.22 μA) was lower than that of p-MoTe2 FET (15 μA) at Vds = 1 V. However, these characteristics have been significantly improved compared to other substitutionally doped 2D semiconductors29,30 (Supplementary Table 1). Furthermore, by further optimizing the device design and process flow, it is expected to enhance the performance of MoTe2 transistors.

The precise fabrication of patterned p-type and n-type semiconductors is a critical step in the realization of integrated circuits. Building upon the developed doping technique and the solid-to-solid phase transition process23, we devise a one-step growth method to obtain patterned doping structures in 2H-MoTe2 thin films. Initially, an Nb-doped Mo film is deposited on a substrate and then patterned using conventional photolithography and physical etching techniques (Fig. 3a). Subsequently, a Re-doped Mo film is deposited, and after the lift-off process, a continuous patterned precursor film comprising Nb- and Re-doped regions is obtained (Fig. 3a, b). Finally, a one-step CVD growth is employed to synthesize seamlessly contacted patterned p-type and n-type 2H-MoTe2 films (Fig. 3a–c). During the phase transition and recrystallization, single-crystal 2H-MoTe2 domains extend across the interfaces of the differently doped regions, resulting in a single-crystal 2H-MoTe2 thin film with patterned p-type and n-type regions. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images at the interface between the Nb-doped and Re-doped regions reveal a perfect lattice without any crystal angular deflection and grain boundaries (Fig. 3d). Additionally, the selected area electron diffraction (SAED) pattern obtained at the interface, with an aperture size of 800 nm, exhibits a single set of hexagonal 2H points (Supplementary Fig. 15), indicating the single-crystal nature of the film. The Re-doped region appears slightly brighter than the Nb-doped region due to the intentionally controlled thickness difference, which allows for precise atomic-scale positioning of the interface. This observation is further confirmed by atomic force microscopy (AFM) results (Fig. 3e). However, the initial Mo films used to synthesize the 2H-MoTe2 pn junctions were prepared through etching, followed by sputter deposition and lift off processes, and voids or clusters of atoms at the interfaces inevitably appear. Consequently, a series of voids with a ring of unidentified transition regions surrounding them were found in the large-area TEM characterization at the interfaces of the synthesized pn junction (Supplementary Fig. 16). It is challenging to directly observe the atomic substitutions doping in HAADF-STEM images due to the thickness averaging effect (4-5 nm, corresponding to seven layers). However, in deliberately prepared thinner Re-doped 2H-MoTe2 (four layers), the substitutional atoms can be effectively resolved due the larger atomic number difference between Re and Mo (Supplementary Fig. 17). STEM simulation analysis shows that the HAADF intensity of a Re atom replacing a Mo atom in one of the Mo columns in the 4-layer 2H-MoTe2 is almost identical to the experimentally observed intensity.

a Schematic diagram of the patterned preparation of doped 2H-MoTe2 pn structure thin films. b Optical image of patterned Mo film selectively incorporated with Nb or Re. c Optical image of the resulting patterned 2H-MoTe2 film selectively p-type doping or n-type doping after chemical vapor deposition (CVD) growth. d High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image at the interface of the 2H-MoTe2 pn structure, where the p-type and n-type 2H-MoTe2 share the same crystal orientation. e, f Atomic force microscopy (AFM) and Kelvin probe force microscope (KPFM) images at the interface of the 2H-MoTe2 pn structure. g–i Optical images of 2H-MoTe2 pn structured thin films with different area-selective doping patterns.

The doping levels of the p-type and n-type regions can be elegantly controlled by adjusting the incorporation ratios of the dopants Nb and Re. The effectiveness of doping-induced Fermi level shifts is validated by measuring the surface potential difference across the interface using Kelvin probe force microscope (KPFM). The Re-doped 2H-MoTe2 exhibits a significantly higher surface potential compared to the Nb-doped 2H-MoTe2 (Fig. 3f), aligning with the expected doping characteristics. The change in surface potential at the interface is not caused by the change in thickness of 2H-MoTe2 (Supplementary Fig. 18). Additionally, by employing lithographically pre-defined patterns for the doped Mo film, we synthesized arbitrarily patterned in-plane coplanar-contacted p-type and n-type 2H-MoTe2 films (Fig. 3g–i). This achievement showcases the reproducibility and design flexibility of the one-step growth method.

The developed patterned doping technique enables the integration of large-scale 2D p-type and n-type semiconductor channels on a single plane, facilitating the creation of monolithic integrated circuits with high performance31. This precise doping technique allows for matching the conductivity of p-type and n-type channels and adjusting the threshold voltage of CMOS FETs. To demonstrate this capability, a 2D semiconductor CMOS inverter array was fabricated using the patterned doped 2H-MoTe2 thin film, employing conventional photolithography, etching, and deposition processes (Fig. 4a–d). The patterned doped 2H-MoTe2 films were directly grown on Si/SiO2 substrates, eliminating the need for additional transfer processes during device fabrication. To achieve matched on-state currents and high on/off ratios for the p-type and n-type channels, 2H-MoTe2 with a Nb incorporation ratio of 0.20% and a Re incorporation ratio of 0.24% were used to prepare the p-type and n-type channels, respectively. Ohmic contacts were formed using Pd/Au and Ti/Au electrodes for the p-MoTe2 and n-MoTe2 channels, respectively. The p-MoTe2 and n-MoTe2 channels in the CMOS inverter had the same channel size (length/width of 10/40 μm, thickness of 5 nm) and shared a Pd/Au top-gate (TG) electrode, with a 20 nm atomic layer deposition (ALD) Al2O3 serving as the gate dielectric (Fig. 4b). Leveraging the compatibility of patterned doped 2H-MoTe2 thin films with large-scale fabrication, hundreds of CMOS inverter arrays based on 2D semiconductor 2H-MoTe2 were fabricated on a centimeter-scale chip (Fig. 4c, d).

a Optical image of a typical 2H-MoTe2 complementary metal-oxide-semiconductor (CMOS) inverter with a channel width of 40 μm, a channel length of 10 μm, and a channel thickness of 5 nm. b Schematic diagram of the structure of a 2H-MoTe2 CMOS inverter, in which the area-selective Nb-doped region was used as the p-type channel, and the area-selective Re-doped region was used as the n-type channel. c, d Photographic and optical images of a chip-sized 2H-MoTe2 CMOS inverter array directly fabricated on a Si/SiO2 substrate. e Transfer characteristics of the p- and n-type 2H-MoTe2 transistors in a CMOS inverter. f Output characteristics of the p- and n-type 2H-MoTe2 transistors in a CMOS inverter with matched on-current. g–i Typical transfer characteristics, signal gains (∣dVout/dVin∣), and power consumption for an inverter operating at supply voltages (Vdd) from 1 V to 4 V. The Vin and Vout are the input and output voltage respectively.

In the representative inverter, the TG transfer curves of Nb-doped and Re-doped 2H-MoTe2 channels at Vds = 1 V exhibit typical p-type and n-type semiconductor characteristics, with current on/off ratios of ~ 104 (Fig. 4e). The gate leakage current (Ig) remains in the picoampere (pA) range, indicating the excellent dielectric properties of the 20 nm ALD Al2O3 over the entire sweep range of the Vg. The output curves (Ids-Vds) obtained on both p-MoTe2 and n-MoTe2 TG-FETs show drain saturation currents of ~ 0.6 μA and ~ 0.8 μA, respectively (Fig. 4f). These matched saturation output currents for the p-FET and n-FET are a result of well-controlled doping concentrations in the p-MoTe2 and n-MoTe2 channels. To verify the stability of the doping technique and the uniformity of the electrical performance of doped films, we tested the transfer characteristics of 30 p-type and 30 n-type 2H-MoTe2 transistors, respectively (Supplementary Fig. 19), and presented the hysteresis phenomenon in the transfer characteristics (Supplementary Fig. 20). The static voltage transfer characteristics of the 2H-MoTe2 CMOS inverter, with supply voltages (Vdd) of 1 V, 2 V, 3 V, and 4 V, clearly demonstrate signal inverting behaviors (Fig. 4g). When the input voltage (Vin) is low, the output voltage (Vout) remains close to Vdd. Once Vin reaches the threshold, Vout rapidly transitions to near-zero voltage (four orders of magnitude lower than Vdd) and remains at that level. The corresponding voltage gain (∣dVout/dVin∣) of this CMOS inverter is estimated to be 2.6 at Vdd = 1 V and increases to 38.2 at Vdd = 4 V (Fig. 4h). A key advantage of a CMOS inverter is its low static power dissipation since only a small current flows from Vdd to the ground, whether the Vin is low (n-FET off) or high (p-FET off), making it favorable for denser circuit integration. Based on the current flowing into the inverter, the peak static power consumption of this inverter is determined to be 1.5 nW Vdd = 1 V, increasing to 89.5 nW at Vdd = 4 V (Fig. 4i).

The developed doping method is specifically designed to be compatible with monolithic integration techniques, ensuring high device uniformity and reliability. To validate this compatibility, a set of thirty inverters from the device array were randomly selected and subjected to electrical characterization. The voltage transfer characteristics of all 30 2D CMOS inverters exhibited sharp switching output swings, indicating well-defined switching behavior. On average, these inverters demonstrated a voltage gain of 2.4 ± 0.5 at Vdd = 1 V (Supplementary Fig. 21). The high yield and uniformity observed in the inverters highlight the potential for widespread adoption of this precisely doped 2H-MoTe2 2D semiconductor in the development of large-scale integrated circuits.

In summary, we have presented a substitutional doping method that enables the direct synthesis of large-area 2D semiconductor 2H-MoTe2 thin films with precisely controlled carrier concentrations, allowing for the fabrication of p-type, n-type, or pn patterned structures. By combining this doping technique with our previously developed seeded growth method24, it becomes possible to produce precisely doped single-crystal 2H-MoTe2 semiconductor films. Moreover, the in-plane 2D epitaxy25 of these doped films enables their integration onto various surfaces, including highly lattice-mismatched single-crystal substrates and predefined architectures, thereby introducing new functionalities to silicon chips. Notably, these p-type and n-type 2D semiconductor channels can be grown in-plane or layer-by-layer in the vertical direction32, opening up possibilities for the development of unprecedented interlayer interconnection processes. In conjunction with other fast-paced developments of 2D semiconductor engineering, such as reduced contact resistance33,34,35,36,37,38 and increased integration of high-k dielectrics39,40, the development of stable and controllable doping schemes will greatly contribute to the widespread utilization of 2D semiconductors in very large-scale integration technology41,42.

Mo films were deposited on 1-inch Si/SiO2 (280 nm) substrates by magnetron sputtering. In addition to Mo target, Mo targets with 1% Nb or 1% Re were used to control the amount of dopants. During sputtering, the power of the Mo target was direct current (DC) 10 W, while the power of the doping target was set to alternating current (AC) 20 W, 30 W, 40 W, or 60 W. The thickness of the resulting Mo film is controlled by the sputtering time. Then, the substrate covered with a continuous Mo film (incorporated with Nb or Re) and Te lumps were placed in a furnace tube for CVD growth at 650 ∘C for 3 h (Ar and H2 gas flow were 5 sccm and 7 sccm). Wafer-scale Nb-doped or Re-doped 2H-MoTe2 films can be obtained after CVD growth.

For magnetron sputtering systems, the deposition rates of different sputtering targets can be modulated by corresponding power values. In this work, the Mo target was connected to a DC power supply and set at a constant 10 W, which resulted in a Mo deposition rate of 1.05 nm/min (Supplementary Fig. 22). A Mo target containing 1% Nb or 1% Re was used as a doping target and was connected to an AC power supply. We obtained different deposition rates of dopants by varying the AC power value of the doping target. The thickness of the film under different sputtering powers was measured by AFM, so as to obtain the precise deposition rate and obtain the incorporated dopant ratio in the Mo film (Supplementary Fig. 22). Finally, 2H-MoTe2 films with Nb incorporation ratios of 0.10%, 0.19%, 0.20% and 0.27% and Re incorporation ratios of 0.06%, 0.09%, 0.24% and 0.42% were obtained by CVD growth.

In the back-gate device, the Nb-doped or Re-doped 2H-MoTe2 thin film ( ~5 nm) was directly grown on 280 nm SiO2/Si substrates by the above method, where the heavily doped Si was used as the gate electrode and SiO2 was used as the dielectric layer. Subsequently, photolithography and reactive ion etching (RIE) under SF6 and Ar atmospheres were used to define the channels. E-beam evaporated Pd/Au (10/30 nm) electrodes were used to contact the p-MoTe2 channel, and Bi/Au (10/30 nm) electrodes were used to contact n-MoTe2 channel. We also tried using other metal contacts, and n-type 2H-MoTe2 back-gated transistors with Cr, Ti, and Bi contacts, the transistor drain current is higher in the Bi metal contacts than in the Ti and Cr contacts due to the low work function and semimetallic nature of Bi. Similarly, the drain current of the p-type 2H-MoTe2 back-gate transistors in contact with high work function Pd is substantially higher than that of the transistor in contact with the low work function Cr (Supplementary Fig. 23). Hall devices were fabricated using the same process. All devices were covered with a 20 nm ALD Al2O3 to seal them from the air. Meanwhile, the Al2O3 layer can be used as the top gate dielectric in CMOSFETs. CMOS inverters were fabricated on spatially patterned doped 2H-MoTe2 films. p-type and n-type channels were defined by photolithography and RIE. Unlike back-gated devices, Ti/Au (10/30 nm) electrodes were used to contact n-type channel, since the CMOS devices were finally subjected to rapid thermal annealing (RTA) at 300 ∘C for 30 s to eliminate the trap states in the Al2O3 dielectric.

The WiTec alpha300 confocal system was used to measure Raman spectra, and the wavelength of the excitation laser was 532 nm. AFM and KPFM characterizations were carried out on an Oxford Cypher S system using tapping mode. STEM images were collected using the Nion U-HERMES-200 aberration-corrected STEM operating at 60 kV and Titan Cubed Themis G2 300 microscope operating at 300 kV. The electrical properties of transistors and Hall devices were measured by a semiconductor characterization system (Keithley 4200-SCS) and a Keithley 2636B source meter, respectively.

All raw data generated during the current study are available from the corresponding authors upon request. The data that support the findings of this study have been deposited in the Figshare under accession code. https://doi.org/10.6084/m9.figshare.27313008.

Seo, S. G. et al. Reversible and controllable threshold voltage modulation for n-channel MoS2 and p-channel MoTe2 field-effect transistors via multiple counter doping with ODTS/poly-L-lysine charge enhancers. Nano Res. 14, 3214–3227 (2021).

Article ADS CAS Google Scholar

Najmaei, S. et al. Vapour phase growth and grain boundary structure of molybdenum disulphide atomic layers. Nat. Mater. 12, 754–759 (2013).

Article ADS CAS PubMed Google Scholar

Radisavljevic, B. et al. Single-layer MoS2 transistors. Nat. Nanotechnol. 6, 147–150 (2011).

Article ADS CAS PubMed Google Scholar

Desai, S. B. MoS2 transistors with 1-nanometer gate lengths. Science 354, 99–102 (2016).

Article ADS CAS PubMed Google Scholar

Tong, L. et al. Heterogeneous complementary field-effect transistors based on silicon and molybdenum disulfide. Nat. Electron. 6, 37–44 (2023).

CAS Google Scholar

Kong, L. et al. Doping-free complementary WSe2 circuit via van der Waals metal integration. Nat. Commun. 11, 1866 (2020).

Article ADS CAS PubMed PubMed Central Google Scholar

Sarkar, D. et al. A subthermionic tunnel field-effect transistor with an atomically thin channel. Nature 526, 91–95 (2015).

Article ADS CAS PubMed Google Scholar

Liu, Y. et al. Promises and prospects of two-dimensional transistors. Nature 591, 43–53 (2021).

Article ADS CAS PubMed Google Scholar

Chhowalla, M., Jena, D. & Zhang, H. Two-dimensional semiconductors for transistors. Nat. Rev. Mater. 1, 16052 (2016).

Article ADS CAS Google Scholar

Wu, R. et al. Bilayer tungsten diselenide transistors with on-state currents exceeding 1.5 milliamperes per micrometre. Nat. Electron. 5, 497–504 (2022).

Article CAS Google Scholar

Geim, A. & Grigorieva, I. Van der Waals heterostructures. Nature 499, 419–425 (2013).

Article CAS PubMed Google Scholar

Akinwande, D. Graphene and two-dimensional materials for silicon technology. Nature 573, 507–518 (2019).

Article ADS CAS PubMed Google Scholar

Wang, S., Wang, X. & Warner, J. H. All chemical vapor deposition growth of MoS2: h-BN vertical van der Waals heterostructures. ACS nano 9, 5246–5254 (2015).

Article CAS PubMed Google Scholar

Gao, H. et al. Tuning electrical conductance of MoS2 monolayers through substitutional doping. Nano Lett. 20, 4095–4101 (2020).

Article ADS CAS PubMed Google Scholar

Seo, S. Y. et al. Writing monolithic integrated circuits on a two-dimensional semiconductor with a scanning light probe. Nat. Electron. 1, 512–517 (2018).

Article CAS Google Scholar

Tang, J. et al. In Situ Oxygen Doping of Monolayer MoS2 for Novel Electronics. Small 16, 2004276 (2020).

Article CAS Google Scholar

Shen, Y. et al. The trend of 2D transistors toward integrated circuits: Scaling down and new mechanisms. Adv. Mater. 34, 2201916 (2022).

Article CAS Google Scholar

Liu, L. et al. Uniform nucleation and epitaxy of bilayer molybdenum disulfide on sapphire. Nature 605, 69–75 (2022).

Article ADS CAS PubMed Google Scholar

Wang, Q. et al. Layer-by-layer epitaxy of multi-layer MoS2 wafers. Natl Sci. Rev. 9, nwac077 (2022).

Article CAS PubMed PubMed Central Google Scholar

Kim, K. S. et al. Non-epitaxial single-crystal 2D material growth by geometric confinement. Nature 614, 88–94 (2008).

Article ADS Google Scholar

Das, S. et al. Transistors based on two-dimensional materials for future integrated circuits. Nat. Electron. 4, 786–799 (2021).

Article ADS CAS Google Scholar

Lee, P. A. & Fisher, D. S. Anderson localization in two dimensions. Phys. Rev. Lett. 47, 882 (1981).

Article ADS CAS Google Scholar

Xu, X. et al. Millimeter-scale single-crystalline semiconducting MoTe2 via solid-to-solid phase transformation. J. Am. Chem. Soc. 141, 2128–2134 (2019).

Article CAS PubMed Google Scholar

Xu, X. et al. Seeded 2D epitaxy of large-area single-crystal films of the van der Waals semiconductor 2H MoTe2. Science 372, 195–200 (2021).

Article ADS CAS PubMed Google Scholar

Pan, Y. et al. Heteroepitaxy of semiconducting 2H-MoTe2 thin films on arbitrary surfaces for large-scale heterogeneous integration. Nat. Synth. 1, 701–708 (2022).

Article ADS Google Scholar

Edelberg, D. et al. Approaching the intrinsic limit in transition metal diselenides via point defect control. Nano Lett. 19, 4371–4379 (2019).

Article ADS CAS PubMed Google Scholar

Mondal, A. et al. Low Ohmic contact resistance and high on/off ratio in transition metal dichalcogenides field-effect transistors via residue-free transfer. Nat. Nanotechnol. 19, 34–43 (2024).

Article ADS CAS PubMed Google Scholar

Chang, Y. M. et al. Reversible and Precisely Controllable p/n-Type Doping of MoTe2 Transistors through Electrothermal Doping. Adv. Mater. 30, 1706995 (2018).

Article Google Scholar

Wang, Z. et al. Substitutional p-Type Doping in NbS2-MoS2 Lateral Heterostructures Grown by MOCVD. Adv. Mater. 35, 2209371 (2023).

Article CAS Google Scholar

Murai, Y. et al. Versatile post-doping toward two-dimensional semiconductors. ACS nano 15, 19225–19232 (2021).

Article CAS PubMed Google Scholar

Gong, Y. et al. Spatially controlled doping of two-dimensional SnS2 through intercalation for electronics. Nat. Nanotechnol. 13, 294–299 (2018).

Article ADS CAS PubMed Google Scholar

Pan, Y. et al. Direct Multitier Synthesis of Two-Dimensional Semiconductor 2H-MoTe2. ACS Appl. Electron. Mater. 4, 5733–5738 (2022).

Article CAS Google Scholar

Zhang, Q. et al. Simultaneous synthesis and integration of two-dimensional electronic components. Nat. Electron. 2, 164–170 (2019).

Article Google Scholar

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

Article ADS CAS PubMed Google Scholar

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

Article ADS CAS PubMed Google Scholar

Liu, Y. et al. Low-resistance metal contacts to encapsulated semiconductor monolayers with long transfer length. Nat. Electron. 5, 579–585 (2022).

Article ADS CAS Google Scholar

Wang, Y. et al. P-type electrical contacts for 2D transition-metal dichalcogenides. Nature 610, 61–66 (2022).

Article ADS PubMed Google Scholar

Song, S. et al. Fabrication of p-type 2D single-crystalline transistor arrays with Fermi-level-tuned van der Waals semimetal electrodes. Nat. Commun. 14, 4747 (2023).

Article ADS CAS PubMed PubMed Central Google Scholar

Illarionov, Y. Y. et al. Insulators for 2D nanoelectronics: the gap to bridge. Nat. Commun. 11, 3385 (2020).

Article ADS CAS PubMed PubMed Central Google Scholar

Huang, J. K. et al. High-k perovskite membranes as insulators for two-dimensional transistors. Nature 605, 262–267 (2022).

Article ADS CAS PubMed Google Scholar

Shen, P. C. et al. Healing of donor defect states in monolayer molybdenum disulfide using oxygen-incorporated chemical vapour deposition. Nat. Electron. 5, 28–36 (2022).

Article CAS Google Scholar

Kozhakhmetov, A. et al. Scalable Substitutional Re-Doping and its Impact on the Optical and Electronic Properties of Tungsten Diselenide. Adv. Mater. 32, 2005159 (2020).

Article CAS Google Scholar

Download references

This work was supported by the National Natural Science Foundation of China (No. 12250007, No. 12425402, No. 62274010 and No. 12134011), the National Key R&D Program of China (Grants No. 2023YFF1500600, No. 2022YFA1203902), the Key Research Program of Frontier Sciences, CAS (Grant No. ZDBS-LY-JSC015), the Beijing Natural Science Foundation (Grant No. JQ21018), and the China Postdoctoral Science Foundation (Grants No. BX20230023 and No. 2024M750099).

State Key Laboratory for Mesoscopic Physics and Frontiers Science Center for Nano-Optoelectronics, School of Physics, Peking University, Beijing, 100871, China

Yu Pan, Pingfan Gu, Yiwen Song, Qi Wang, Yuqia Ran, Yanping Li, Wanjin Xu, Xiaolong Xu & Yu Ye

Collaborative Innovation Center of Quantum Matter, Beijing, 100871, China

Yu Pan, Peng Gao & Yu Ye

School of Physics and Technology, Wuhan University, Wuhan, 430072, China

Tao Jian, Zemin Pan, Chendong Zhang & Jun He

MIIT Key Laboratory of Semiconductor Microstructure and Quantum Sensing Department of Applied Physics, Nanjing University of Science and Technology, Nanjing, 210094, China

Pingfan Gu

Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, 100871, China

Yiwen Song & Qi Wang

Electron Microscopy Laboratory, School of Physics, Peking University, Beijing, 100871, China

Bo Han & Peng Gao

International Center for Quantum Materials, School of Physics, Peking University, Beijing, 100871, China

Bo Han & Peng Gao

School of Integrated Circuits and Electronics, MIIT Key Laboratory for Low-Dimensional Quantum Structure and Devices, Beijing Institute of Technology, Beijing, 100081, China

Xiaolong Xu

Liaoning Academy of Materials, Shengyang, 110167, China

Yu Ye

Peking University Yangtze Delta Institute of Optoelectronics, Nantong, 226010, Jiangsu, China

Yu Ye

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

Y.Y., X.X., and Y.P. conceived the project. Y.P. and Y.Y. designed the experiments, analyzed the results, and wrote the manuscript. Y.P. prepared the samples, fabricated the devices, and performed the characteristics of Raman, AFM, KPFM, and electrical measurements. T.J., Z.P., and C.Z. performed the STM characteristics. Pingfan G. performed the Hall measurements. Y.S. and B.H. performed the STEM characteristics. Q.W., Y.R., Y.L., and W.X. was involved in transistor device fabrication. X.X., Peng G., J.H., and C.Z. provided useful analysis and discussions. Y.Y. supervised this project. All authors discussed the results and contributed to the manuscript.

Correspondence to Chendong Zhang, Jun He, Xiaolong Xu or Yu Ye.

The authors declare no competing interests.

Nature Communications thanks Seunguk Song, Zhenyu Wang, and the other, anonymous, reviewer for their contribution to the peer review of this work. A peer review file is available.

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

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

Reprints and permissions

Pan, Y., Jian, T., Gu, P. et al. Precise p-type and n-type doping of two-dimensional semiconductors for monolithic integrated circuits. Nat Commun 15, 9631 (2024). https://doi.org/10.1038/s41467-024-54050-2

Download citation

Received: 15 July 2024

Accepted: 31 October 2024

Published: 07 November 2024

DOI: https://doi.org/10.1038/s41467-024-54050-2

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

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

Provided by the Springer Nature SharedIt content-sharing initiative