Physics of diodes-- question

It might help to let go of the concept of 'resistance' when thinking about diodes and other semiconductors. Its more about potential differences. While these might look like a resistance at first glance (voltage drop vs current flow) the underlying physics is different.

I think potential is also important. I've seen diode graphs of potential, but never seen any of resistance verses distance at various locations throughout the diode. Such a graph would be interesting. It seems to me that the resistance would be relative to how much the specific area is depleted. I always thought depletion was relative to the barrier height -->

If my assumption was correct then wouldn't that mean the resistance is appreciably higher near where the metal-semiconductor come in contact?

Paul

In the normal operation of a diode, the junction can be forward biased (full conduction) or reverse biased (insulator condition due to charge carrier depletion in the 'depletion region'). The constant one calls 'resistance' does not have any useful descriptive power,

And, it's not a constant in this case.

When current is low, the depletion region is a significant thickness of insulator. Actually probing that region (putting wires into it) is disruptive in many ways, so the usual characterization is done by determining the junction capacitance as a function of voltage applied. Study the C/V relation in a diode, and you can find lots of data. Try to study point-by-point resistance, and you're off in no-data wilderness.

Also a common calculation is for zero bias.

The term "constant" could be referring to the diode in zero bias for instance.

Usually such diode parameters are calculated-- no wires. :-) For example Cjo is the capacitance at zero bias.

What I'm thinking is that such a resistance graph could be calculated just as the other graphs. Does anyone have a good book recommendation on the physics (very detailed) of diodes?

Regards, Paul

What one would calculate, of course, would be resistivity (the local property) as a function of axial distance, for a PN diode junction.

Yes, of course, the space charge distribution of a planar diode is well known, and one can use the known mobilities of electrons and holes to calculate resistivity from that charge distribution. The easiest calculation would be for a so-called abrupt junction.

The problem is, THOSE ARE EQUILIBRIUM CALCULATIONS. Shockley, _Electrons and Holes in Semiconductors_ (1950) "The resistance of a p-n junction is much greater than the integrated resistivity of the material composing the junction. This point follows from the fact that the hole current, for example, must flow a certain distance in regions where the density of the carriers is much less than [equilibrium value]"

Thanks for the info and book reference. I'll have to save up to purchase that book. Amazon.com has a used version for $500. I'm interested in small signals far below Vt, so equilibrium equations are fine. Would you say a graph of the diodes point resistance throughout the diodes depletion region is mostly constant in a Schottky diode with an abrupt junction?

Regards, Paul

Since it's out of print, only used copies are available. I did see one for under $100. Enter the title at bookfinder.com.