Abstract

Recently, III-V tunneling field effect transistors (TFET) for low voltage logic applications (<0.5V) have gained attention with the demonstration of sub-60 mV/dec. subthreshold slopes [1]. A key outstanding issue with TFETs is limited drive currents, due to non-optimized carrier tunneling. With that issue in mind, the aim of this work is to map III-V Esaki tunnel diode (TD) performance to engineer TDs with ultra high current densities while maintaining large peak-to-valley current ratios (PVCR). This work describes the most comprehensive experimental benchmarking of TD performance reported, including (i) GaAs, (ii) In0.53Ga0.47As, (iii) InAs, (iv) InAs0.9Sb0.1/Al0.4Ga0.6Sb, and (v) InAs/GaSb as a function of doping and effective tunnel barrier height. These results confirm that heterojunctions (bandgap engineering) and doping will enhance peak (JP) and Zener current densities beyond homojunction TDs [3], to a record 2.2MA/cm2 (JP) and 11 MA/cm2 (@ -0.3 V), laying the fundamental groundwork for a III-V TFET at the 7 nm technology node. The devices fabricated for this study were grown using molecular beam epitaxy (MBE). Sub-micron diodes were fabricated using a similar process flow (Fig. 1a) as Pawlik, et al. [3]. However, to better handle large current densities, the metal 1 contact was changed to Mo and patterned via subtractive plasma etch. As before, SEM image analysis was performed after mesa etch to measure the area and undercut (Fig. 1b) for each device size. For high JP TDs ( >100 kA/cm2), statistically accurate values were extracted from Gaussian fits to histograms generated from large sample sizes (≳ 200). JP and PVCR are taken as benchmarking figures of merit, correlating to the TFET drive current and subthreshold slope. A series of homojunction TDs (3 GaAs, 6 In0.53Ga0.47As, 2 InAs), were grown by MBE systematically varying the n- and p- doping (Table I) to benchmark Jp. All devices fabricated, with the exception of the highest doped InAs TD (InAs-2), showed clear negative differential resistance (NDR). Representative J-V curves for all of the homojunction TDs are shown in Fig 3. A very good homojunction JP of 975 kA/cm2 was observed with a maximum PVCR of 3.4 from InGaAs-6 [3] due to its bandgap properties and maximum achievable doping concentrations. Fig. 4 presents a figure of merit (FOM) plot benchmarking the experimentally demonstrated JP against N*-1/2 [4], where N*=NAND/(NA+ND) . In addition to the experimental data reported, several literature results for GaAs [5-7], In0.53Ga0.47As [3,8-10], and InAs/GaSb [11,12] are also benchmarked. Each homojunction forms an exponential line allowing for easy interpolation of JP for TDs beyond this experiment. The data clearly shows the limitations of homojunction systems, and calls for the need to investigate heterojunction systems (staggered gap [SG] or broken gap [BG]) to achieve higher tunnel current densities. The smaller effective tunnel barrier height of the SG (80 meV) and BG (-70 meV) heterojunction systems, found from the band diagrams in Fig 6, allows for increased tunneling current. Historically, the maximum reported JP of a heterojunction system (InAs/GaSb TD) ranged from 40 – 60 kA/cm2, with a PVCR of 2-3 [11, 12]. The SG and BG devices described in this abstract employ a new design space by elevating the n-type doping >1x1019 cm-3. Mirroring the experiment for the homojunction TDs, 4 InAs0.9Sb0.1/Al0.4Ga0.6Sb and 3 InAs/GaSb TDs (Fig. 5 and Table II) were fabricated with varied p-type doping. NDR was observed in SG-2 through SG-4 and all of the BG devices. Figs. 7, 8, and 9 illustrate the typical J-V characteristics, a scatter plot of the JP vs. junction area, and a series of histograms to extract JP for the SG (a) and BG (b) devices. While SG-1 does not exhibit NDR, it is interesting to note the three order of magnitude difference in Zener current with SG- 4, indicating promise as an off state for a TFET, which is attributed to the widened tunnel barrier in Al0.4Ga0.6Sb (1018 cm-3 doping). The highest doped SG device (SG-4) showed a JP of 877 kA/cm2 (typical PVCR 2.2), a value comparable to the best homojunction TDs [3]. BG-3 has the highest JP of 2.2 MA/cm2 (typical PVCR 3.5) ever experimentally observed in any TD [3]. These results are approximately 40x greater than previous reports in any BG system [11, 12]. Furthermore, this is the first time that a heterojunction system has exceeded the JP of all homojunction TDs [3]. The application of this system for TFET drive (tunnel) current is directly illustrated by the Zener current density of ~11 MA/cm2 at -0.3 V bias. An expanded FOM is given in Fig. 11 to include the heterojunction devices. Fig. 12 presents a comprehensive map of Jp vs. relative barrier height at multiple doping levels to illustrate the strong dependence on bandgap and doping. It is the belief of the authors that the maximum experimentally achievable JP has not yet been reached since the Si doping in InAs/GaSb TDs can be pushed to 1020 cm-3 which would have a projected JP ~10MA/cm2. This study demonstrated outstanding broken gap TDs within a comprehensive mapping of TD performance based on experimental data across many III-V systems. This study focused on the impact of TDs for TFETs; however, these results also apply to the design of solar cells, VCSELs, memory, and other device applications

Publication Date

2012

Comments

Note: imported from RIT's Digital Media Library running on DSpace to RIT Scholar Works on April 2014.

Document Type

Article

Department, Program, or Center

Microelectronic Engineering (KGCOE)

Campus

RIT – Main Campus

Share

COinS