Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Design and fabrication of high-performance diamond triple-gate field-effect transistors

Abstract

The lack of large-area single-crystal diamond wafers has led us to downscale diamond electronic devices. Here, we design and fabricate a hydrogenated diamond (H-diamond) triple-gate metal-oxide-semiconductor field-effect transistor (MOSFET) to extend device downscaling and increase device output current. The device’s electrical properties are compared with those of planar-type MOSFETs, which are fabricated simultaneously on the same substrate. The triple-gate MOSFET’s output current (174.2 mA mm−1) is much higher than that of the planar-type device (45.2 mA mm−1), and the on/off ratio and subthreshold swing are more than 108 and as low as 110 mV dec−1, respectively. The fabrication of these H-diamond triple-gate MOSFETs will drive diamond electronic device development forward towards practical applications.

Introduction

Semiconductor diamond has some extraordinary physical properties, including a wide band gap energy (5.47 eV), a low dielectric constant (5.7), a theoretical high breakdown field (10 MV cm−1), high carrier saturation velocity (1.5–2.7 × 107 and 0.85–1.2 × 107 cm s−1 for electrons and holes, respectively)1,2, highest thermal conductivity (22 W cm−1 K−1), and high carrier mobility (4500 and 3800 cm2 V−1 s−1 for electrons and holes, respectively)3. According to the figures of merit quoted for diamond and other semiconductor materials4, diamond-based electronic devices have the highest power-frequency product, the highest thermal limitation, and the lowest power-loss at high frequencies. Diamond is therefore considered the most suitable material for fabrication of next-generation high-power, high-frequency, high-temperature, low-power-loss, and energy-saving electronic devices5. Because the activation energies of diamond dopants are much higher than the room temperature thermal energy, many diamond electronic devices have been fabricated on hydrogenated diamond (H-diamond) channel layers6,7,8,9,10,11. The H-diamond can accumulate holes on its surface with a sheet hole density (psheet) of 1012–1013 cm−2. In fact, exposure of H-diamond to a NO2 ambient12 or annealing of oxygen-terminated diamond in an NH3 ambient13 can produce psheet for H-diamond of as high as 1014 cm−2. Recently, fabrication processes for H-diamond metal-oxide-semiconductor field-effect transistors (MOSFETs) have been developed. The maximum drain-source current (IDS,max) of a MOSFET at room temperature fabricated on NO2-treated H-diamond14 was as much as −1.35 A mm−1 under conditions of gate-source voltage (VGS), drain-source voltage (VDS), and gate length (LG) of −5 V, −12 V, and 0.4 μm, respectively. The cut-off frequency of the device was more than 10 GHz over a wide VGS range of approximately 10.0 V. In addition, the operational performance of H-diamond-based MOSFETs was comparable to that of SiC- or GaN-based MOSFETs in high temperature (400 °C) and high voltage (500 V) operation15.

While these H-diamond-based MOSFETs showed excellent electrical properties, the absence of large-area single-crystal diamond wafers has hindered their development for widespread practical applications. This issue has led us to downscale diamond electronic devices. In our previous studies16,17, downscaled H-diamond MOSFETs were fabricated by eliminating the interspacing between the source/drain and gate contacts (LS/D-G). The on-resistance (RON) of the H-diamond MOSFET without LS/D-G (29.7 Ω mm) was considerably lower than the corresponding device with LS/D-G (208.4 Ω mm). The device’s current output and extrinsic transconductance (gm) of the former were also around seven times higher than those of the latter. Other studies that focused on downscaling of the LG for H-diamond MOSFETs were also reported18. The shortest LG for the single crystalline H-diamond MOSFETs were downscaled to be around 100 nm. Recently, triple-gate MOSFET architecture has been developed in the Si-, InGaAs-, and GaN-based MOSFETs to extend device downscaling, reduce leakage current, and control device short channel effects19,20,21,22,23,24,25,26,27. Also, because the triple-gate MOSFET can allow carriers to travel in both its planar and lateral sides, the device current output is much higher than that of a planar-type device with the same area. The fabrication of H-diamond triple-gate MOSFETs is therefore promising for extension of device downscaling and enhancement of the device electrical properties.

Here, we describe the design and fabrication of H-diamond triple-gate MOSFETs on a single crystalline diamond substrate. The electrical properties of these devices are compared with those of planar-type MOSFETs. The absolute IDS,max of the triple-gate MOSFET is 174.2 mA mm−1, which is much higher than the 45.2 mA mm−1 value of the planar-type device. In addition, the on/off ratio and the subthreshold swing (SS) of the H-diamond triple-gate MOSFET are higher than 108 and as low as 110 mV dec−1, respectively.

Results

Fin-patterned H-diamond and triple-gate MOSFET fabrication

Figure 1 shows the fabrication process flows for (a) fin-patterned H-diamond MOSFETs and (b) triple-gate MOSFETs, (c) the top view of the entire sample surface, (d) the top view of two triple-gate H-diamond MOSFETs, and (e) the top view of two planar-type H-diamond MOSFETs. To form the fin patterns on the diamond (001) substrate, a tungsten (W) metal layer was first sputtered using an automatic sputtering system to cover the entire substrate surface [Fig. 1(a)-1]. The positive photoresist FEP-171 was then coated on the sample and exposed using an electron beam (EB) lithography system to form fin models [Fig. 1(a)-2]. After the photoresist was developed, the W metal and the diamond substrate at the photoresist-free area were dry-etched in SF6 and O2 ambients, respectively, using an inductively-coupled plasma reactive ion etching (ICP-RIE) system [Fig. 1(a)-3 and 4]. The residual W metal was cleaned again in the SF6 ambient to form fin patterns on the diamond surface [Fig. 1(a)-5]. Then, the H-diamond epitaxial layer was grown on the substrate by microwave plasma chemical vapour deposition (MPCVD) to form fin-patterned H-diamond [Fig. 1(a)-6].

Figure 1
figure1

Fabrication of fin-patterned H-diamond and triple-gate MOSFETs.

(a,b) Fabrication routines for the fin-patterned H-diamond and triple-gate MOSFETs, respectively. (c) Top view of the entire sample surface. Three ohmic contacts fell off during the fabrication process. (d) Top view of two triple-gate H-diamond MOSFETs. (e) Top view of two planar-type H-diamond MOSFETs.

After the formation of fin-patterned H-diamond, the triple-gate MOSFETs were then fabricated [Fig. 1(b)]. The fin-patterned H-diamond was first etched in an O2 ambient using a capacitively-coupled plasma RIE (CCP-RIE) system to form a mesa structure [Fig. 1(b)-1]. Palladium/titanium/gold (Pd/Ti/Au) ohmic contacts were then evaporated on the fin-patterned H-diamond to form the source/drain electrodes using an electron-gun (E-gun) evaporation system [Fig. 1(b)-2]. The aluminium oxide (Al2O3) gate insulator layer which has been used for the fabrication of high-performance diamond MOS devices in the previous reports14,15,28 and the aluminium (Al) gate electrode were then deposited to cover the entire sample surface by atomic layer deposition (ALD) and ultra-high-vacuum (UHV) sputtering techniques, respectively [Fig. 1(b)-3]. Then, the sample was coated with PMGI-SF6S/FEP-171 bilayer photoresists and exposed using the EB lithography system to form gate models. After development of the photoresists, the Al and Al2O3 layers on the photoresist-free areas were wet-etched using mixed Al etching acid and tetramethylammonium hydroxide (TMAH) solutions, respectively. Finally, the photoresists were lifted off in an N-methylpyrrolidone (NMP) solution, and the fabrication of the triple-gate H-diamond MOSFETs was complete [Fig. 1(b)-4]. Planar-type MOSFETs were also fabricated simultaneously with the triple-gate devices on the same diamond substrate. Figure 1(c) shows a top view of the entire surface of the sample. The total number of designed MOSFETs was 128. However, three ohmic contacts fell off during the fabrication process. The top views of triple-gate and planar-type H-diamond MOSFETs are shown in Fig. 1(d,f), respectively. The LG, LS/D-G, and gate width (WG) for these devices were 500 nm, 500 nm, and 100.5 μm, respectively.

Surface and interface morphologies

Figure 2 shows scanning electron microscopy (SEM) [Fig. 2(a–c)] images of the fin-patterned diamond substrate and triple-gate MOSFET, and transmission electron microscopy (TEM) [Fig. 2(d–g)] images of interface of the triple-gate H-diamond MOSFET. As shown in the SEM images [Fig. 2(a,b)], the total width of the diamond fin pattern and the fin length are 100.5 and 7 μm, respectively. Both the fin width and the interspacing between fins are 500 nm. The fin height was confirmed using a 3D-measurement laser microscope to be 500 nm. Obvious gate, source, and drain contacts for the H-diamond triple-gate MOSFET can be seen in Fig. 2(c). After H-diamond epitaxial layer growth by the MPCVD technique, the fin length and width increased to 7.8 μm and 600 nm, respectively. The interspacing between fins and the fin height both decreased to 400 and 340 nm, respectively [Fig. 2(d)]. The H-diamond epitaxial layer thickness is approximately 50 nm [Fig. 2(e)]. Figure 2(f) shows a high-resolution TEM image for the zoom of the left adjacent fins in the Fig. 2(d). The angle between two adjacent fins is 60°. The two inclined active planes of each fin in the triple-gate MOSFETs are the sides. The equivalent WG for the triple-gate MOSFET can be calculated to be 139.6 μm. The ALD-Al2O3 layer thickness is approximately 27.9 nm, which is in good agreement with the measurement results obtained using an ellipsometer system. An interfacial layer with thickness of around 0.6 nm exists between H-diamond and Al2O3 [Fig. 2(g)], and a similar layer is also observed at the AlN/H-diamond interface29. The origins of these layers are still under discussion at present, however, the layers are possibly a result of reactions between the oxides or nitrides and the surface adsorbates on the H-diamond epitaxial layer30.

Figure 2
figure2

Surface and interface morphologies.

(a,b) SEM images of the fin-patterned diamond substrate. (c) SEM image of the triple-gate MOSFET. (d–g) Interfacial TEM images of the triple-gate H-diamond MOSFET.

Electrical properties of triple-gate and planar-type MOSFETs

Figure 3 shows (a) and (b) schematic diagrams of the triple-gate and planar-type H-diamond MOSFETs, respectively, (c) gate leakage current (IG,leak) for the triple-gate and planar-type MOSFETs, and (d) and (e) drain-source current versus voltage (IDS-VDS) characteristics for the triple-gate and planar-type MOSFETs, respectively. Both triple-gate and planar-type MOSFETs have the same LG, WG, and LS/D-G. The difference for them is the existence of fin-patterns on the diamond substrate for the triple-gate MOSFET. The IG,leak curves for the MOSFETs were measured with the VGS changing from 30.0 to −10.0 V. At the VGS of −10.0 V, the holes are accumulated at the Al2O3/H-diamond interface and the MOSFETs are at on-states. The IG,leak of the triple-gate MOSFET is 1.4 × 10−10 A, which is higher than that of the planar-type one of 2.3 × 10−12 A. This is possibly ascribed to the longer equivalent WG and rougher etching surface for the triple-gate MOSFET than those for the planar-type one. The IG,leak density for the planar-type MOSFET can be calculated to be 4.6 × 10−6 A cm−2 using the IG,leak divided by area of gate electrode (5.025 × 10−7 cm2). It is one order higher than that of the Al2O3/H-diamond MOS capacitor31 of 1.1 × 10−7 A cm−2. Since the fabrication process for MOSFET is more complicated than that for MOS capacitor, the device damage during fabrication for the former is more serious than that for the latter, which possibly leads to the higher IG,leak for the MOSFET. At the VGS of 30.0 V, holes are difficult to be accumulated at the Al2O3/H-diamond interface and MOSFETs are at off-states. The IG,leak for the triple-gate MOSFET is 1.8 × 10−12 A, which is lower than that for the planar-type one of 1.3 × 10−10 A.

Figure 3
figure3

Electrical properties of the triple-gate and planar-type MOSFETs.

(a,b) Schematic diagrams of the triple-gate and planar-type MOSFETs, respectively. (c) The IG,leak curves for the triple-gate and planar-type MOSFETs. Green cycle and red square curves represent the IG,leak of the triple-gate and planar-type MOSFETs, respectively. (d,e) IDS-VDS characteristics for the triple-gate and planar-type MOSFETs, respectively. The VGS is varied from −10.0 to 20.0 V in steps of +1.0 V.

The VGS is varied from −10.0 to 20.0 V in steps of +1.0 V for measurement of the IDSVDS characteristics of the triple-gate and planar-type MOSFETs [shown in Fig. 3(d,e), respectively]. The IDS for the planar-type MOSFET was normalized with the WG of 100.5 μm. That for the triple-gate MOSFET was normalized with its equivalent WG of 139.6 μm. Both MOSFETs show obvious p-type channel and pinch-off characteristics. There are also good linear relationships between IDS and low VDS for both devices, which indicate good ohmic contact between the Pd/Ti/Au and H-diamond channel layers. The absolute IDS,max for the triple-gate MOSFET is 174.2 mA mm−1, which is much higher than the value of 45.2 mA mm−1 obtained for the planar-type device. The value of RON can be extracted from the linear region of the IDSVDS characteristics, and is 31.9 and 98.0 Ω mm for the triple-gate and planar-type MOSFETs, respectively. The RON for the triple-gate H-diamond MOSFET is composed of the fin pattern channel resistance beneath the Al2O3 insulator (RCH), the fin pattern H-diamond surface resistance with the LS/D-G of 500 nm (2RSD), and the Pd/Ti/Au ohmic contact resistance (2RC). Because 2RC is much smaller than RCH and 2RSD, it can be neglected here32. By combining the RON value of another two triple-gate MOSFETs with the LS/D-G of 1.0 and 2.0 μm (The electrical properties of them are shown in Fig. S1 of the Supplementary Information), RCH and 2RSD for the triple-gate MOSFET can be deduced to be 23.8 and 8.1 Ω mm, respectively.

The transfer characteristics that correspond to the IDSVDS curves are shown in Fig. 4. The on/off ratio of the triple-gate MOSFET is higher than 108 [Fig. 4(a)], and is the same level as that of the planar-type device [Fig. 4(d)]. The SS is an important parameter for evaluation of MOSFET power consumption, and is defined as the inverse slope of log |IDS| versus VGS. The SS is 110 mV dec−1 for the triple-gate MOSFET at a VDS of −10.0 V [Fig. 4(a)]. This value is much lower than that of the planar-type device of 460 mV dec−1 [Fig. 4(d)]. There is also a relationship between the SS and the interfacial trap charge density (Dit) of . Here, k, T, q, CH-diamond and are Boltzmann’s constant (8.62 × 10−5 eV K−1), room temperature (298.15 K), the elementary charge (1.6 × 10−19 C), the capacitance of the H-diamond, and the capacitance of the Al2O3 layer, respectively. can be calculated using the equation to be 0.171 μF cm−2, where ε0, , and are the dielectric constant of a vacuum (8.85 × 10−12 F m−1), the dielectric constant of Al2O3 (5.4)31, and the thickness of the Al2O3 layer (27.9 nm), respectively. If CH-diamond can be neglected in the deep-subthreshold region, the Dit values for the triple-gate and planar-type MOSFETs are then calculated to be 8.95 × 1011 and 7.14 × 1012 eV−1 cm−2, respectively. The threshold voltage (VTH) values of the MOSFETs can be determined based on as functions of VGS, and are 10.2 ± 0.1 and 7.6 ± 0.1 V for the triple-gate and planar-type MOSFETs, respectively [Fig. 4(b,e)].

Figure 4
figure4

Transfer characteristics of IDS–VDS for triple-gate and planar-type MOSFETs.

(a–c) The log |IDS|–VGS, VGS, and gm–VGS characteristics of the triple-gate MOSFET, respectively. (d–f) The log |IDS|–VGS, VGS, and gm–VGS characteristics of the planar-type MOSFET, respectively.

The following relationship exists between RON, VTH, and effective mobility (μeff): . The values of RON, LG, WG, , VGS, VTH, and 2RSD for the triple-gate MOSFET are 31.9 Ω mm, 500 nm, 139.6 μm, 0.171 μF cm−2, −10.0 V, 10.2 ± 0.1 V, and 8.1 Ω mm, respectively. The μeff of the fin-patterned H-diamond channel layer can be calculated to be 6.1 ± 0.5 cm2 V−1 s−1. This is lower than the value for the planar-type device of 38.7 ± 0.5 cm2 V−1 s−1 that was reported previously33, and can possibly be attributed to the increased surface roughness at the etching area for the fin-patterned H-diamond channel layer. The extrinsic transconductance (gm) is determined based on the slope of the IDSVGS curve. The maximum gm (gm,max) values for the triple-gate and planar-type MOSFETs are 15.3 ± 0.1 and 3.8 ± 0.1 mS mm−1, respectively [Fig. 4(c,f)].

Discussion

The H-diamond triple-gate MOSFET has been fabricated and characterized to compare with those of the corresponding planar-type device, and these properties are summarized in Table 1. While the equivalent WG of the triple-gate MOSFET is only 1.4 times longer than WG for the planar-type device, the absolute IDS,max for the former (174.2 mA mm−1) is around four times larger than the corresponding value for the latter (45.2 mA mm−1). This was confirmed again using another triple-gate MOSFET, as shown in Fig. S2 of the Supplementary Information. It was previously reported that the inclined H-diamond (111) plane had a higher psheet than the planar H-diamond (001) plane34. In this study, because each fin of the triple-gate MOSFET has two inclined (±01) planes, it is natural to believe that the psheet of the fin-patterned H-diamond channel layer must be higher than that of the planar H-diamond (001) layer. This is possibly the reason for the higher IDS,max and lower RON obtained for the triple-gate MOSFET in comparison to the theoretical values. The IDS,max for the triple-gate MOSFET is still much lower than that of the NO2-treated H-diamond-based MOSFET (−1.35 A mm−1)14, which possibly attributed to the higher hole density for the NO2-treated H-diamond channel layer and the poor crystalline quality of the large-area diamond wafer35 used in this study. We have also attempted to fabricate the triple-gate MOSFET without the LS/D-G. The electrical properties of the resulting device are shown in Fig. S3 of the Supplementary Information. While the IDS,max of this device is as much as 251.4 mA mm−1, the output current cannot be controlled well with changes in VGS. In the triple-gate MOSFET without the LS/D-G, the Al2O3/Al gate layers also cover the source/drain ohmic contacts. During the Al2O3/Al etching process [Fig. 1(b)-4], strong damage may possibly occur at the edge area, which could lead to high gate leakage and poor electrical properties for the resulting MOSFET. The on/off ratios of both the triple-gate and planar-type MOSFETs are higher than 108, and are thus high enough for practical applications. Dit for the triple-gate MOSFET (8.95 × 1011 eV−1 cm−2) is lower than that for the planar-type device (7.14 × 1012 eV−1 cm−2), which leads to the SS of the triple-gate MOSFET (110 mV dec−1) being much lower than that of the planar-type MOSFET (460 mV dec−1). The VTH for the triple-gate MOSFET is larger than that for the planar-type MOSFET, which is possibly attributed to the higher psheet for the fin-pattern H-diamond channel layer. Both VTH values are much higher than zero for the MOSFETs, which indicates that the devices operate with depletion modes. Recently, control conditions for the fabrication of depletion/enhancement mode H-diamond MOSFETs have been verified36. Therefore, it is promising for fabrication of enhancement-mode H-diamond triple-gate MOSFETs in future work. The gm,max (15.3 ± 0.1 mS mm−1) of the triple-gate MOSFET is much higher than that of the planar-type device (3.8 ± 0.1 mS mm−1).

Table 1 Electrical properties of the triple-gate and planar-type MOSFETs.

In conclusion, the H-diamond triple-gate MOSFETs have been fabricated on a single crystalline diamond substrate. The electrical properties of these devices are compared with those of planar-type MOSFETs. The absolute IDS,max of the triple-gate MOSFET is 174.2 mA mm−1, which is much higher than the 45.2 mA mm−1 value of the planar-type device. In addition, the on/off ratio and the SS of the H-diamond triple-gate MOSFET are higher than 108 and as low as 110 mV dec−1, respectively. The fabrication of these high-performance H-diamond triple-gate MOSFETs will drive the development of diamond electronic devices forward towards practical applications.

Methods

Fin-patterned H-diamond fabrication

The CVD single crystalline diamond (001) substrate, with dimensions of 5.0 × 5.0 × 0.3 mm was purchased from EDP Corp. It was cleaned in a mixed acid solution (H2SO4 and HNO3 with a volume ratio of 1:1) for 3 h at 300 °C. The W metal was sputtered on the diamond substrate at 300 W in an Ar gas ambient using an automatic sputtering system (JSP-8000, ULVAC, Kanagawa, Japan). The W layer thickness and sputtering time were 200 nm and 30 min, respectively. The W/diamond sample was coated with FEP-171 positive photoresist using a spin-coater with a rotation rate and time of 5000 rpm and 1 s, respectively. The baking temperature and time for the FEP-171 photoresist were 120 °C and 2 min, respectively. After exposure using the EB lithography system (ELS-7000, Elionix, Tokyo, Japan), the sample was developed in a TMAH solution for 1.5 min. The W metal was then etched via a Bosch process with SF6 and C4F8 gases using the ICP-RIE dry etching system (MUC-21, Sumitomo Precision Products, Hyogo, Japan). The SF6 and C4F8 flow rates were 75 and 60 sccm, respectively, and their plasma powers were 175 and 150 W, respectively. The diamond at the photoresist-free area was etched using the same equipment in an O2 gas ambient. The etching power, the O2 flow rate, the chamber pressure, and the etching time were 400 W, 10 sccm, 0.5 Pa, and 25 min, respectively. After cleaning of the residual W, the fin-patterned diamond substrate was formed. Then, the H-diamond epitaxial layer was grown using the MPCVD system (AX5200S, Seki Technotron Corp., Tokyo, Japan). Before growth commenced, the fin-patterned diamond substrate was annealed in the MPCVD chamber at 1000 °C for 20 min to clean off any surface contamination. The growth temperature, time, and chamber pressure for the H-diamond epitaxial layer were 900–940 °C, 20 min, and 80 Torr, respectively. The H2 and CH4 flow rates were 500 and 0.5 sccm, respectively.

Triple-gate MOSFET fabrication

The fabrication of the triple-gate Al2O3/H-diamond MOSFETs was based on a combination of EB lithography, CCP-RIE dry etching, E-gun evaporation, ALD, UHV sputtering, wet etching, and lift-off techniques. The PMGI-SF6S/FEP-171 bilayer photoresists were sequentially coated on the fin-patterned H-diamond substrate. The baking conditions for the FEP-171 layer were given above. The baking temperature and time for the PMGI-SF6S layer were 180 °C and 5 min, respectively. The H-diamond channel layer was etched in an O2 ambient at a pressure of 10 Pa using the CCP-RIE system (RIE-200NL, Samco, Kyoto, Japan) to form the mesa structure. The plasma power and the etching time were 50 W and 1.5 min, respectively. The Pd/Ti/Au ohmic contact was formed using the E-gun evaporation system (RDEB-1206K, R-DEC Co. Ltd., Ibaraki, Japan), where the Pd metal layer was evaporated first to contact with the fin-patterned H-diamond surface. The Pd, Ti, and Au layer thicknesses were 10, 20, and 100 nm, respectively. The evaporation rates for these layers were 0.05, 0.05, and 0.2 nm s−1, respectively. The chamber pressure was in the 1.0–2.5 × 10−5 Pa range. The Al2O3 gate insulator and the Al gate electrode were deposited sequentially on the fin-patterned H-diamond channel layer using the ALD (SUNALE R-100B, Picosun, Tokyo, Japan) and UHV sputtering (LS-420R, Biemtron, Ibaraki, Japan) systems, respectively. The precursors for the ALD-Al2O3 layer were Al(CH3)4 and water vapour. The pulse and purge times for both precursors were 0.1 and 4.0 s, respectively. The deposition temperature was 120 °C. The plasma power, the chamber pressure, the Ar gas flow rate, and the deposition time for Al metal sputtering were 50 W, 0.3 Pa, 10 sccm, and 7 min, respectively. The Al metal was then wet etched using a mixed acid solution (volume ratio of H3PO4:HNO3:CH3COOH:H2O of 16:2:2:1) for 1 min. The Al2O3 insulator was wet etched using the TMAH solution for 10 min. The photoresists were removed using an NMP solution at room temperature for 3 h.

Measurement system

The surface morphology of the fin-patterned diamond substrate was investigated using the SEM system (S-4800, Hitachi, Tokyo, Japan). The sample was prepared for TEM measurements using a focused ion beam-SEM (Xvision-200DB, SII Co., Chiba, Japan) system. TEM measurements were performed using the JEM-2100F system with an accelerating voltage of 200 kV. The fin pattern height was measured using a 3D-measurement laser microscopy system (OLS-4000, Olympus, Tokyo, Japan). The Al2O3 film thickness was measured using an ellipsometer system (MARY-102FM, Five Lab, Saitama, Japan). The electrical properties of the MOSFETs were measured using an MX-200/B prober (Vector Semiconductor Corp., Tokyo, Japan) and a B1500A parameter analyser (Agilent, Tokyo, Japan). The IG,leak curves for the MOSFETs were obtained by measuring current-voltage relationships between the gate and source contacts. The IDSVDScharacteristics for the MOSFETs were obtained by measuring current-voltage relationships between the drain and source contacts with the change of VGS.

Additional Information

How to cite this article: Liu, J. et al. Design and fabrication of high-performance diamond triple-gate field-effect transistors. Sci. Rep. 6, 34757; doi: 10.1038/srep34757 (2016).

References

  1. Reggiani, L. et al. Hole-drift velocity in natural diamond. Phys. Rev. B 23, 3050–3057 (1981).

    CAS  ADS  Article  Google Scholar 

  2. Wort, C. J. H. & Balmer, R. S. Diamond as an electronic material. Mater. Today 11, 22–28 (2008).

    CAS  Article  Google Scholar 

  3. Isberg, J. et al. High carrier mobility in single-crystal plasma-deposited diamond. Science 297, 1670–1672 (2002).

    CAS  ADS  Article  Google Scholar 

  4. Baliga, B. J. Power semiconductor device figure of merit for high-frequency applications. IEEE Electron Dev. Lett. 10, 455–457 (1989).

    ADS  Article  Google Scholar 

  5. May, P. W. Diamond thin films: a 21st-century material. Phil. Trans. R. Soc. Lond. A 358, 473–495 (2000).

    CAS  ADS  Article  Google Scholar 

  6. Kawarada, H., Aoki, M. & Ito, M. Enhancement mode metal-semiconductor field effect transistors using homoepitaxial diamonds. Appl. Phys. Lett. 65, 1563–1565 (1994).

    CAS  ADS  Article  Google Scholar 

  7. Kubovic, M. Microwave performance evaluation of diamond surface channel FETs. Diam. Relat. Mater. 13, 802–807 (2004).

    CAS  ADS  Article  Google Scholar 

  8. Russell, S. A. O., Sharabim, S., Tallaire, A. & Moran, D. A. J. Hydrogen-terminated diamond field-effect transistors with cutoff frequency of 53 GHz. IEEE Electron Dev. Lett. 33, 1471–1473 (2012).

    CAS  ADS  Article  Google Scholar 

  9. Yun, Y. et al. Electrical properties of Al/CaF2/i-diamond metal-insulator-semiconductor field-effect-transistor fabricated by ultrahigh vacuum process. Jpn. J. Appl. Phys. 37, L1293–L1296 (1998).

    Article  Google Scholar 

  10. Liu, J. W. et al. Diamond logic inverter with enhancement mode metal-insulator-semiconductor field-effect transistor. Appl. Phys. Lett. 105, 082110 (2014).

    ADS  Article  Google Scholar 

  11. Kueck, D. et al. AlN as passivation for surface channel FETs on H-terminated diamond. Diam. Relat. Mater. 19, 932–935 (2010).

    CAS  ADS  Article  Google Scholar 

  12. Sato, H. & Kasu, M. Maximum hole concentration for hydrogen-terminated diamond surfaces with various surface orientations obtained by exposure to highly concentrated NO2 . Diam. Relat. Mater. 31, 47–49 (2013).

    CAS  ADS  Article  Google Scholar 

  13. Imura, M. et al. Development of AlN/dimaond heterojunction field-effect transistors. Diam. Relat. Mater. 24, 206–209 (2012).

    CAS  ADS  Article  Google Scholar 

  14. Hirama, K. et al. Diamond field-effect transistors with 1.3 A/mm drain current density by Al2O3 passivation layer. Jpn. J. Appl. Phys. 51, 090112 (2012).

    ADS  Article  Google Scholar 

  15. Kawarada, H. et al. C-H surface diamond field effect transistors for high temperature (400 °C) and high voltage (500 V) operation. Appl. Phys. Lett. 105, 013510 (2014).

    ADS  Article  Google Scholar 

  16. Liu, J. W. et al. Low on-resistance diamond field-effect transistor with high-k ZrO2 as dielectric. Sci. Rep. 4, 6395 (2014).

    CAS  Article  Google Scholar 

  17. Liu, J. W. et al. Diamond field effect transistors with a high-dielectric constant Ta2O5 as gate material. J. Phys. D: Appl. Phys. 47, 245102 (2014).

    ADS  Article  Google Scholar 

  18. Hirama, K. et al. High-performance p-channel diamond MOSFETs with alumina gate insulator. IEDM Tech. Dig. 873–876 (2007).

  19. Nowak, E. J. Maintaining the benefits of CMOS scaling when scaling bogs down. IBM J. RES. & Dev. 46, 169–180 (2002).

    Article  Google Scholar 

  20. Doyle, B. S. et al. High performance fully-depleted tri-gate CMOS transistors. IEEE Electron Dev. Lett. 24, 263–265 (2003).

    CAS  ADS  Article  Google Scholar 

  21. Crupi, F. et al. Reliability comparison of triple-gate versus planar SOI FETs. IEEE Trans. Electron Dev. 53, 2351–2357 (2006).

    CAS  ADS  Article  Google Scholar 

  22. Lee, C. W. et al. High-temperature performance of silicon junctionless MOSFETs. IEEE Trans. Electron Dev. 57, 620–625 (2010).

    CAS  ADS  Article  Google Scholar 

  23. Radosavljevic, M. et al. Non-planar, multi-gate InGaAs quantum well field effect transistors with high-k gate dielectric and ultra-scaled gate-to-drain/gate-to-source separation for low power logic applications. IEEE Electron Dev. Meet. 6.1.1-6.1.4 (2010).

  24. Huang, C. H. & Li, Y. M. Electrical characteristic of InGaAs multiple-gate MOSFET devices. Simul. Semicon. Proc. Dev. 357–360 (2015).

  25. Lu, B., Matioli, E. & Palacios, T. Tri-gate normally-off GaN power MOSFET. IEEE Electron Dev. Lett. 33, 360–362 (2012).

    CAS  ADS  Article  Google Scholar 

  26. Chen, S. H. et al. High-performance III-V MOSFET with nano-stacked high-k gate dielectric and 3D fin-shaped structure. Nanoscale Res. Lett. 7, 431 (2012).

    ADS  Article  Google Scholar 

  27. Im, K. S. et al. High-performance GaN-based nanochannel FinFETs with/without AlGaN/GaN heterostructure. IEEE Trans. Electron Dev. 60, 3012–3018 (2013).

    CAS  ADS  Article  Google Scholar 

  28. Kovi, K. K., Vallin, Ö., Majdi, S. & Isberg, J. Metal-oxide-semiconductor capacitors on boron-doped diamond. IEEE Electron Dev. Lett. 36, 603–605 (2015).

    CAS  ADS  Article  Google Scholar 

  29. Pietzka, C. et al. Analysis of diamond surface channel field-effect transistors with AlN passivation layers. J. Appl. Phys. 114, 114503 (2013).

    ADS  Article  Google Scholar 

  30. Maier, F. et al. Origin of surface conductivity in diamond. Phys. Rev. Lett. 85, 3472–3475 (2000).

    CAS  ADS  Article  Google Scholar 

  31. Liu, J. W. et al. Electrical properties of atomic layer deposited HfO2/Al2O3 multilayer on diamond. Diam. Relat. Mater. 54, 55–58 (2015).

    CAS  ADS  Article  Google Scholar 

  32. Wang, W. et al. Palladium Ohmic contact on hydrogen-terminated single crystal diamond film, Diam. Relat. Mater. 59, 90–94 (2015).

    CAS  ADS  Article  Google Scholar 

  33. Liu, J. W., Liao, M. Y., Imura, M. & Koide, Y. Normally-off HfO2-gated diamond field effect transistors. Appl. Phys. Lett. 103, 092905 (2013).

    ADS  Article  Google Scholar 

  34. Hirama, K. et al. High-performance p-channel diamond metal-oxide-semiconductor field-effect transistors on H-terminated (111) surface. Appl. Phys. Express 3, 044001 (2010).

    ADS  Article  Google Scholar 

  35. Furuhashi, M. & Fujimori, N. Fabrication of colorless single crystal diamonds using “Direct wafer fabrication technique”. Proc. Jpn. Dia. Symp. 103, 8–9 (2015).

    Google Scholar 

  36. Liu, J. W. et al. Control of normally on/off characteristics in hydrogenated diamond metal-insulator semiconductor field-effect transistors. J. Appl. Phys. 118, 115704 (2015).

    ADS  Article  Google Scholar 

Download references

Acknowledgements

This work was supported by JSPS KAKENHI (Grant No. K8324) and the International Center for Young Scientists (ICYS) of the National Institute for Materials Science (NIMS). The work was also supported in part by the NIMS Nanofabrication Platform in the Nanotechnology Platform project and a Fundamental Research A (No. 25249054) project sponsored by the Ministry of Education, Culture, Sports, and Technology (MEXT), Japan. The authors would like to thank Drs K. Najima and S. Tanigawa at the Nanofabrication Platform in NIMS for providing technical support.

Author information

Affiliations

Authors

Contributions

J.L. and Y.K. supervised the project. J.L. designed and carried out all experiments. H.O. helped with the experiments in EB lithography and ICP-RIE dry etching, and with the SEM measurements. X.W. helped with the TEM measurements. M.L. helped with the analysis of the experimental results. J.L. wrote the manuscript with important input from all authors.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Electronic supplementary material

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Liu, J., Ohsato, H., Wang, X. et al. Design and fabrication of high-performance diamond triple-gate field-effect transistors. Sci Rep 6, 34757 (2016). https://doi.org/10.1038/srep34757

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing