5
$\begingroup$

I'm reading up on semiconductor devices that rely on quantum tunneling, such as the tunnel diode and the TFET. The big advantage of these devices is apparently that "quantum tunneling is extremely fast".

Actually finding the speed of tunneling requires taking standard tunneling results (from solving the time-independent SE) and building wavepackets. This seems messy, and I don't see why those wavepackets should go particularly fast. What's the intuition here?

| cite | improve this question | |
$\endgroup$
  • $\begingroup$ Would Electrical Engineering be a better home for this question? $\endgroup$ – Qmechanic♦ May 25 '16 at 8:57
  • $\begingroup$ Here, being “faster” only means that the tunnelling probability is higher, as mentioned in the answer by @SteveB. $\endgroup$ – AlQuemist May 27 '16 at 8:26
  • $\begingroup$ @Qmechanic I doubt it. $\endgroup$ – DanielSank May 29 '16 at 20:33
7
+25
$\begingroup$

Due to the exponential dependendence of tunneling probability on barrier height and thickness, it is entirely possible for tunneling to take femtoseconds, or for tunneling to take 100 trillion years.

The point is that tunneling can potentially be very fast (femtoseconds), and those devices where tunneling is supposed to happen are designed so that tunneling is in fact very fast.

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ When you talk about the exponential dependence of tunneling rates on barrier properties, you're arguing that this allows tunneling to be switched on an off fast, right? Otherwise what does it mean for "tunneling to take femtoseconds"? $\endgroup$ – DanielSank May 29 '16 at 20:34
  • $\begingroup$ Many devices have switching speed limits related to carrier dynamics, e.g. electron transit time through a layer, minority carrier lifetime, etc. However, for tunneling, there is essentially no such delay. So when circumstances are favorable for tunneling (appropriate band alignments and occupancies across the tunnel barrier), the tunnel current can start flowing more-or-less instantaneously. Of course, you may need to wait out other delays like RC time constant before circumstances are favorable for tunneling. But still, MIM tunnel diodes can rectify 500THz signals. $\endgroup$ – Steve Byrnes May 29 '16 at 23:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.