Nanoribbon Tunnel Field Effect Transistor Research Articles

Analytical Current Transport Modeling of Graphene Nanoribbon Tunnel Field-Effect Transistors for Digital Circuit Design

A semi-classical and a semi-quantum current transport model for p-i-n n-type armchair graphene nanoribbon tunnel field effect transistor (TFET) are studied analytically. The results are compared with the numerical quantum transport simulation method using an atomistic Schrodinger–Poisson solver within the non-equilibrium Green function (NEGF) formalism. The channel length and width are 20 and 4.9 nm and a-GNR band gap is 0.289 eV. Current ratio ${\rm I_{ON}}/{\rm I_{OFF}}$ at 0.1 V supply voltage is calculated as follows: 122, 16.3 and 116 with a subthreshold slopes 26, 69 and 27.4 mV/decade from semi-classical, semi-quantum and NEGF simulation, respectively. Performance of a-GNR TFET is also studied analytically and numerically considering a-GNR width variation. Voltage transfer characteristics of a-GNR TFET inverter are computed for 0.1 V and 0.2 supply voltages using three current transport models which are in close agreement.

Simulation of Dirac Electron Tunneling Current in Armchair Graphene Nanoribbon Tunnel Field-Effect Transistors Using a Transfer Matrix Method

We simulate quantum mechanical tunneling current in armchair graphene nanoribbon tunnel field-effect transistors (AGNR-TFETs). The relativistic Dirac equation is used to determine electron wave functions in the AGNRs, while the potential profile is solved by the Poisson equation. We use a transfer matrix method (TMM) to calculate the electron transmittance and the Dirac electron tunneling current in the AGNR-TFETs. The results show that the Dirac electron tunneling current increases with increasing the drain and gate voltages. Moreover, the AGNR width and the thickness of insulator affect the characteristics of the Dirac electron tunneling currents.

Graphene nanoribbon tunnel field effect transistor with lightly doped drain: Numerical simulations

By inserting a lightly doped region between the highly doped drain and the intrinsic channel of a graphene nanoribbon tunnel field effect transistor (GNR-TFET), we propose a new lightly doped drain (LDD)-GNR-TFET. Transport characteristics of the proposed transistor is numerically simulated, employing the third-nearest-neighbor tight-binding approximation in mode space non-equilibrium Green’s function formulism (NEGF), in ballistic regime. Simulations show, in comparison with a conventional GNR-TFET of the same dimensions, the proposed LDD-GNR-TFET exhibits a 102–103 times smaller OFF-current, an up to 105 times larger ON/OFF ratio, a shorter time delay, a smaller power-delay product (PDP) and a less drain induced barrier thinning (DIBT), besides preserving the subthreshold swing.

(Invited) Dielectrics for Graphene Transistors for Emerging Integrated Circuits

In this paper, we present the use of high-k dielectric material in design and physical modeling of following two types of transistors for emerging integrated circuits: graphene nano ribbon (GNR) tunnel field-effect transistors (TFETs) and graphene short channel FETs.Graphene nanoribbon field effect transistor (GNR FET) is one of the promising devices for emerging integrated circuits because of its exceptional electronic properties such as the large carrier mobility and planar structure. However, the problem is a high drain-source leakage current in standby state due to small band gap of graphene and short channel effect (SCE) enforced by the transistor feature size. We have proposed a new GNR FET structure which has two side gates with a lower work function than the main gate on top of the SiO2 and HfO2 insulating layers. The main gate and two side gates have work functions of Φ1=4.8eV and Φ2=4.5eV, respectively. The GNR FET width and length has been chosen 1.5nm and 15nm, where the source and drain regions are doped with a molar fraction of ionized donors equal to 5×10-3/cm3. The lengths of the main and side gates are 2.5nm and 1.25nm while those of insulator layers have been used as a variable. We have simulated the proposed GNR FET using the self-consistent solution of three-dimensional Poisson–Schrödinger equation, within the Non-equilibrium Green’s function (NEGF) formalism. In our new GNR FET structure, the two side gates induce the inversion layers next to drain and source regions, which increase the controllability of the main gate on the channel potential barrier by providing an effective screen for the main gate to prevent the change in drain current due to the change in drain voltage. We have found that the application of the HfO2 in combination of SiO2 in the proposed GNR FET not only shifts the drain source leakage current to the lower values by several orders of magnitude but also make the leakage current sensitive to the proportionality of thicknesses of SiO2 and HfO2. The proposed GNR FET introduce a minimum peak in the drain source leakage current much below that of a common GNR FET with the same insulator structure. The minimum leakage currents of the proposed structure for total insulator thickness of 10Å, 15 Å and 20Å are approximately 6×10-13A, 4×10-12A and 5×10-11A, respectively while those of a common GNR FET are 1.5×10-10A, 9×10-10A and 5×10-9A, respectively. The minimum peak in leakage current has been obtained for HfO2 thickness = 8Å and 14 Å corresponding to the total insulator thickness of 10 Å and 15 Å, respectively, while it shifts to purely HfO2 insulator layer for that of 20 Å. The proposed GNR FET can diminish the short channel effect and prevent an undesirable increase in the leakage current by decreasing the channel length. Effect of dielectric on performance of GNR TFETs is also studied through an analytical current transport model which we have developed for designing of integrated circuits. It is observed that choice of dielectric material and corresponding thickness along the lateral gate position severely modifies performance of GNR TFET. We have found that for constant oxide thickness of 0.76nm, threshold voltage reduces from 0.176V to 0.054V for dielectric permittivity of 2.5 to 47.5. We have also noticed that the performance of GNR TFET studied through threshold voltage and sub-threshold slope depends on selection of both dielectric material and their thickness. Specific results on influence of choice of dielectric in designing of GNR TFET will be presented included in the paper and presentation.

A Theoretical Model of Band-to-Band Tunneling Current in an Armchair Graphene Nanoribbon Tunnel Field-Effect Transistor

Tunneling current in an armchair graphene nanoribbon (AGNR) tunnel field-effect transistor (TFET) was modeled. A linear equation was employed in describing a potential distribution within the AGNR due to its simplicity. A parabolic dispersion and an electron effective mass obtained by approximating kx 0 to the parabolic dispersion were applied to AGNR. In order to obtain electron transmittance, electron wavefunctions in AGNR were based on Airy functions. The obtained transmittance was then applied to calculate the tunneling current by employing the Landauer formula. The calculated results showed that the tunneling current increases with the AGNR width. It was also shown that the tunneling current increases as temperature decreases. In addition, the gate voltage influences the saturation condition of tunneling current in AGNR TFETs.

Dissipative transport in rough edge graphene nanoribbon tunnel transistors

We have studied quantum transport in graphene nanoribbon tunnel field-effect transistors. Unlike other studies on similar structures, we have included dissipative processes induced by inelastic electron-phonon scattering and edge roughness in the nanoribbon self-consistently within a non-equilibrium transport simulation. Our results show that the dissipative scattering imposes a limit to the minimum OFF current and a minimum subthreshold swing that can be obtained even for long channel lengths where direct source-drain tunneling is inhibited. The edge roughness, in the presence of dissipative scattering, somewhat surprisingly, shows a classical behavior where it mostly reduces the maximum ON current achievable in this structure.

Graphene nanoribbon tunneling field effect transistors

The electron-hole symmetry characteristic of graphene nanoribbons (GNRs) gives rise to the electron (hole) tunneling through valence (conduction) band states. By employing this property we have numerically investigated GNR field effect transistors with p+-type source and drain in the presence of a gate voltage-induced n-type channel using the non-equilibrium Green's function formalism.For long channels, the traditional FET-like I-V behavior is achieved, but at short channels, the sub threshold current opens up an oscillatory dependence on the gate voltage with a considerable amount of current of over 10−6A. This is the characteristic current behavior of resonant tunneling transistors that exhibit regions of negative differential resistance. The calculated discrete density of states in the channel attributes this behavior to the constructed n-type channel island between p-type source and drain with thin barriers formed by the energy gap.

Graphene Nanoribbon Tunneling Field-Effect Transistors With a Semiconducting and a Semimetallic Heterojunction Channel

We present a computational study of the device performance of graphene nanoribbon tunneling field-effect transistors (TFETs) with a heterogeneous channel. By varying the length and the energy bandgap (EG) of the heterogeneous region, the on- and off-state currents (ION and IOFF) can be effectively optimized independently. Both semiconducting and semimetallic heterogeneous regions are studied to understand the effects of EG engineering on device behaviors. In addition, the effect of gate coverage (GC) over the heterogeneous region is also investigated. We found that device performance is greatly affected by the positioning of the gate to modify the region where band-to-band tunneling occurs. For a given ION/IOFF of eight orders, our results show that, for the semiconducting heterojunction, a higher ION can be obtained by having the gate partially covering the heterogeneous region. This is due to a combination of a short tunneling length and resonant states, which leads to an increase in carrier concentration for the tunneling mechanism. On the other hand, for the semimetallic case, a similar ION/IOFF is only attainable when the heterogeneous region is not covered by the gate. A large IOFF is observed for even small GC due to the valence electrons from the source traveling to the conduction bands of the semimetallic region, enhancing the carrier transport toward the drain. Our study highlights the device design consideration required when optimizing the device performance of heterojunction TFETs.

Modeling and simulation of uniaxial strain effects in armchair graphene nanoribbon tunneling field effect transistors

In this paper, we perform a modeling and simulation study on strained armchair graphene nanoribbon (AGNR). Two uniaxial strain models based on a tight binding method are compared with results from first-principles calculation. Tunneling field effect transistors (TFETs) with channels made of strained AGNR of different widths are modeled and simulated by a ballistic quantum transport model based on nonequilibrium Green’s function and nonparabolic effective mass approximation. Compared with TFETs with narrow AGNR, those with strained wide AGNR can achieve better device performance.

Effect of Ribbon Width and Doping Concentration on Device Performance of Graphene Nanoribbon Tunneling Field-Effect Transistors

The device performance of graphene nanoribbon (GNR) tunneling field-effect transistor (TFET) is studied using the self-consistent non-equilibrium Green's function (NEGF) and quasi-two dimensional Poisson solver based on the Dirac equation model. The effects of different GNR widths and doping concentrations at the source and drain on the device characteristics are investigated and the electronic property of the GNR TFET is found to be strongly dependent on its width. A comprehensive characterization of this dependence is expected to be crucial to the designs and fabrications of GNR TFETs. Furthermore, the doping concentrations at the source and drain is found to play a crucial role on the ON- and OFF-state currents (ION and IOFF) respectively. Therefore, the ability to control the doping concentrations allows the tailoring of the drive current, the ION/IOFF ratio and the subthreshold swing of GNR TFETs to meet different design requirements.

Graphene Nanoribbon Tunnel Transistors

A graphene nanoribbon (GNR) tunnel field-effect transistor (TFET) is proposed and modeled analytically. Ribbon widths between 3 and 10 nm are considered to effect energy bandgaps in the range of 0.46 to 0.14 eV. It is shown that a 5-nm ribbon width TFET can switch from on to off with only 0.1-V gate swing. The transistor achieves 800 muA/mum on -state current and 26 pA/mum off-state current, with an effective subthreshold swing of 0.19 mV/dec. Compared to a projected 2009 n MOSFET, the GNR TFET can provide 5times higher speed, 20times lower dynamic power, and 280 000times lower off-state power dissipation. The high performance of GNR TFETs results from their narrow bandgaps and their 1-D nature.