Re: An ohmic experiment

[Michael Edmiston <edmiston@BLUFFTON.EDU>]

Paul Johnson asks:  "Aside from the intentionally non-ohmic devices
that have been mentioned in this thread, are most substances
(homogeneous elements and compounds) ohmic..." [linear I versus V plots
except for thermally induced nonlinearities]

Not being a materials scientist, nor a solid state physicist, etc. I am
a bit "out on a limb" here.  But I'll go ahead and state what I think I
know.

(1) I think most ordinary conductors (I suppose I mean mostly metals)
are ohmic by this definition.  That's why it is common and useful to
tabulate data such as resistivity and the temperature coefficient of
resistivity.  I suppose it's also why "Ohm's law" viewed as I = V/R is
so widely applicable.  At any rate, most metals will give linear I
versus V plots if the temperature is regulated.  I plead ignorance of
any exceptions to this or what those exceptions might be.

(2) One situation where non-ohmic behavior is observed involves
junctions.  This is especially common in semiconductor technology in
which we have junctions between materials having different charge
carriers, e.g. holes versus electrons.

Interesting aside: When I worked a Los Alamos, I was in a group of
physicists working in the Electronics Division.  Our mission was to
design and build new types of detectors, often nuclear detectors.  Some
of our detector materials were semiconductors.  Once a device is made
from semiconductor material, it needs electrical connections to the
"outside world."  If you're not careful, your electrical connections to
semiconductor material might create new junctions which behave
non-ohmically.  Although it was not my area of research, I was aware
that others in my group had to make what they called "ohmic contacts"
to the semiconductor material.  I don't believe this was difficult
because the technique had already been developed by semiconductor
manufacturers.  But it was clearly something one had to consider and
work around.

(3) Insulators/dielectrics can behave as non-ohmic conductors near and
after breakdown is achieved.  Gas discharge tubes have already been
mentioned in this thread.  One obvious mechanism at work here is the
snowball/cascade effect of ions colliding with atoms with sufficient
energy to cause more ionization.  This is another example (similar to
junctions) where the current is greatly dependent on the number of
charge carriers present.  An increase in voltage can cause a dramatic
increase in charge carriers, hence the non-linear behavior.

(4) Electrochemistry is another area where non-ohmic behavior is
readily apparent, but ohmic behavior might be observed under the proper
conditions.  If a person wants to measure the resistance/conductance of
an electrolyte (i.e. treat the electrolyte as ohmic) s/he has to be
careful what electrodes are used and what potentials are used.  If any
chemical reactions take place at the electrodes, and if the onset or
efficacy of these reactions is voltage dependent (typically the case)
then gross non-linear behavior will be observed.  Electrochemical
techniques of analysis such at potentiometric titrations and cyclic
voltametry are examples of deriving chemical information from the
non-linear I versus V plots.

Summary:  I think I might summarize this (as a layman) by saying that
if the charge carrier population is not a function of voltage, and if
chemical reactions are not occurring, and if
ionization/dielectric-breakdown is not occurring, etc... then we can
expect ohmic behavior.  This is typically the case in metals.  But I
have given several examples of material and situations in which this is
not the case.

If there are some solid-state physics on the list who can correct my
errors or add information where I have obvious omissions, I would be
pleased to learn more about this.

Michael D. Edmiston, Ph.D.              Phone/voice-mail:       419-358-3270
Professor of Chemistry & Physics        FAX:                    419-358-3323
Chairman, Science Department            E-Mail                  edmiston@bluffton.edu
Bluffton College
280 West College Avenue
Bluffton, OH 45817



-----Original Message-----
From:   Paul O. Johnson [SMTP:pojhome@FLASH.NET]
Sent:   Monday, November 08, 1999 8:42 PM
To:     PHYS-L@lists.nau.edu
Subject:        Re: An ohmic experiment

Michael Edmiston wrote:

> I mentioned, in a previous message, that I have students plot I
versus delta-V for
> resistors over a range of power sufficiently high that they nearly
"smoke" the
> resistors.  Then I have them repeat the experiment over the same
power range, but with
> a heatsink.  I have them do this with both a nichrome resistor and
with a carbon-film
> resistor.  There are some neat things that result from this.
>
> (1) The I versus delta-V plots for both types of resistors show
slight non-linear
> behavior when the resistors are allowed to heat-up.  They both
display very linear
> behavior when fitted with a heatsink and plotted over the same range
as without the
> heatsink.  This demonstrates the non-linear behavior was temperature
induced and not
> voltage induced.  That means these are actually behaving as ohmic
materials.

That's as clear an answer as I've ever gotten. Thanks, Michael. Now,
let's generalize a
bit. Aside from the intentionally non-ohmic devices that have been
mentioned in this
thread, are most substances (homogeneous elements and compounds) ohmic
in the sense of
your result (1)?

poj