Response to “Comment on ‘Assessment of field-induced quantum confinement in heterogate germanium electron–hole bilayer tunnel field-effect transistor’” [Appl. Phys. Lett. 106, 026102 (2015)] Free

In our letter,¹ we demonstrated that the inclusion of field-induced quantum confinement in the analysis of the Ge Electron–Hole Bilayer Tunnel Field–Effect Transistor (EHBTFET) led to the appearance of undesired lateral tunneling which degraded the so far reported outstanding switching behavior of these devices.² We showed that a heterogate configuration (HG-EHBTFET) with different workfunctions for the overlap and underlap sections of the top gate, $ϕtg1$ and $ϕtg2$⁠, respectively (see Fig. 1(c) in Ref. 1), proved to be appealingly efficient to suppress this parasitic tunneling at V_D = 0.5 V and restore the steepness of the I_D – V_TG curves. The main point of the comment raised by Hsu et al.³ was to elucidate the potential drawbacks that the proposed work function difference might cause at low drain voltages. They showed that no deleterious effects arise in that scenario by analyzing the electron eigenenergy difference for the first subbands in the overlap and underlap regions, ΔE, for V_D < 0.5 V and concluded that a heterogate configuration is also quite advisable at low V_D. Furthermore, they showed (Fig. 1(b) in Ref. 3) that there exists an optimized value of $ϕtg2$ for which ΔE turns out to be approximately independent of V_D. Our response seeks to explain the interesting behavior of the $ΔE(VD)$ curves depicted in Fig. 1(b) of the comment by Hsu et al.³ for better understanding of the effect that heterogate configuration has on the performance of these devices. The results herein presented were obtained employing the simulation approach used in our original letter¹ with the device structure considered by Hsu et al.³ 

First, the increasing behavior of ΔE at very low V_D for $ϕtg2=4.3eV$ (Ref. 3) is due to the tighter control that the drain exerts over the first electron subband in the overlap, E_e1,OL compared to that in the underlap, E_e1,UL, as seen in Fig. 1 where we plot the derivative of E_e1 with respect to V_D. Second, the almost constant behavior of ΔE for $ϕtg2=4.2eV$ (Ref. 3) follows from the similar pattern of the $Ee1′(VD)$ curves in both regions. Third, the saturation of the decreasing pattern of ΔE observed for $ϕtg2=4.1eV$ (Ref. 3) that starts at $VD≈0.18V$ comes from the depinning of E_e1,UL. As displayed in the inset, we can estimate the point at which this depinning takes place by means of the maximum of second derivative⁴ of $Ee1(VD)$ and confirm that it occurs at a value of $VD=0.18V$⁠. For $VD>0.18V$ the depinning implies that the top gate recovers the control over E_e1,UL and therefore when we increase $ϕtg2$ by 0.2 eV from 4.1 to 4.3 eV, ΔE behaves practically likewise (Fig. 1(b) in Ref. 3).

This response complements the pertinent comment by Hsu et al.³ reinforcing the HG-EHBTFET reliability at low drain voltages and provides an explanation for the observed behavior of the eigenenergy difference between the first bound states for electrons in the overlap and underlap regions when increasing $ϕtg2$⁠.