I am assuming the leakage mechanism as well as the current flow mechanism is tunneling. If this is not the case, then why can't we realize tunneling switch? That is as switch comprised of Silicon-Tunnel oxide-metal.
Dear Alok Ranjan.
Few comments on your feedback:
You wrote: The tunneling gate leakage current in floating gate memory is very low as there are two oxide barriers to the flow of electrons - (a) tunnel oxide and (b) blocking oxide....
The two layers are on both sides of the floating gate. They are not one next to the other such that the overall thickness is an issue. That is in order for an electron to leak out from the floating gate, it needs to go through one of these layers, not both of them.
You wrote: The tunneling gate leakage current can be well below pA/fA in state of art floating gate memory devices.
In fact it is much smaller. In the range of 10^-23A. A pA or fA will drain the memory in very short time.
You wrote: On the other hand, the resistive switching memory (also called RRAM) uses metal-insulator-metal (MIM) or metal-insulator-semiconductor (MIS) type structures.
I understand that MIM will lead to higher leakage, but this can be avoided if you use standard VLSI technology with Silicon-tunnel oxide-metal. And again my question... why not use Silicon-oxide-metal as a switch?
Alok,
In the floating gate device, you generally can assume to have tunneling through vacuum and if yiu consider the charge on one electron, then you would need almost 10e5 seconds for one electron to tunnel.
I guess the floating gate paper(information) you are refering is any theoretical paper. ( please share it too).
ON THE OTHER HAND, THR RESISTICE SWITCH USING OXIDE ARE HIGHLY dEFE TICE IN SOME SENSE. SO YOU ALFRADY HAVE SO MUCH CONCENTTATION OF CHARGES IN THE OXIDE LAYER WHICH ARE SORT OF HELP THE ELE TRONS TO TUBNEL. IT IS NOT AN IDEALI SULATOR COMPATED TO VACUUM.
ITS MY TAKR.
DEFEXTLY
SO
Alok:
You wrote: "electrodes (both semiconductor and metal) do not have an active role in switching mechanism. The major role of electrodes ( metal/semiconductor) is to apply electrical potential difference between the insulating layer for the generation and recombination of defects responsible for resistive switching."
1. Aren't the electrode required to create a field that enables tunnling through the oxide such that electrons flow from one electrode to the other (assuming the electrodes are part of an electrical circuit?
2. Once voltage between the electrodes is turned off. Aren't we to expect leakage current similar to Floating gate memory? That is much smaller then what we see in MIM devices.
- Badges
- Science method