0
$\begingroup$

In case of intrinsic semiconductor, while applying an electric field, only drift current should exist, right? As diffusion current is due to concentration gradient. So the electron flow will be randomised. Ultimately, the current that will exist in the circuit will be due to applied electric field. I am a bit confused about this

$\endgroup$
2
  • 2
    $\begingroup$ What's the confusion? $\endgroup$
    – garyp
    Commented Jan 25, 2017 at 17:50
  • $\begingroup$ There is always diffusion of charges in a semiconductor. Net diffusion current occurs when there is a concentration gradient. Even without one, electrons and holes are executing a random walk - it just doesn't create a net diffusion current. $\endgroup$
    – Jon Custer
    Commented Jan 25, 2017 at 18:24

2 Answers 2

0
$\begingroup$

When you don't apply an E field the drift and diffusion current balance ( both flowing in opposite to each other). The diffusion current is quite small and does not depend on applied E. When E is applied the drift current changes and therefore the net flow of current across the diode is no longer zero. If the diode is reversed biased the diffusion current dominates. When it is forward biased the built in potential is lowered and the drift current is larger than the diffusion current.

$\endgroup$
3
  • $\begingroup$ Without applying an electric field, according to me the drift current would not exist as it caused due to electric field $\endgroup$
    – Hchavan
    Commented Jan 26, 2017 at 1:31
  • $\begingroup$ There is the electric field from n to p due to the built in voltage difference. The few electrons on the p side done have any obstacle in going from p to n. This is the drift current. $\endgroup$
    – SAKhan
    Commented Jan 26, 2017 at 5:12
  • 1
    $\begingroup$ OP asks about intrinsic sc so p vs n is not relevant $\endgroup$ Commented Dec 16, 2018 at 9:04
0
$\begingroup$

You are right that the absence of a diffusion current implies no statistical trend in the carrier density from point to point. This however does not exclude the possibility of a non zero expectation for the carrier current<\i> density, which would arise from an applied field.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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