NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/sci/electron/22286

Internet Message Format | 1993-01-08 | 5.4 KB

Path: sparky!uunet!gatech!destroyer!cs.ubc.ca!uw-beaver!news.u.washington.edu!carson.u.washington.edu!whit
From: [e-mail] (John Whitmore)
Newsgroups: sci.electronics
Subject: Re: Holes Theroy questions?
Date: 9 Jan 1993 01:09:08 GMT
Organization: University of Washington, Seattle
Lines: 99
Message-ID: <[e-mail]>
References: <[e-mail]>
NNTP-Posting-Host: carson.u.washington.edu


    There is some interesting physics in this set of questions;
no one else seems to be responding, so...

In article <[e-mail]> [e-mail] (Ken Koldan) writes:
>In a posting from Eric H. Taylor - The Theories are full of holes -
>there were questions asked about some things, etc. I found this
>subject very intriguing. My own ideas(questions) are this
>1) If you ran elctron holes through the incandescent bulb would
>not the holes be filled by the electrons in the conductor and thereby
>causing current, electrons falling into a lower state and releasing photons
>, and heat instead of cooling even though the pressure is decressing?

    In solid materials, one can often find completely-occupied
low energy states, and completely-empty high energy states, with
a gap between (where no possible solutions to the electron wave
equation exist).  Such a material is called an insulator; the highest
energy solutions to the electron equation that ARE fully occupied
are the 'valence band', the lowest unoccopied energy states are the
'conduction band', and the void between is the 'band gap'.

    When sufficiently heated, such an insulator 'bumps' a few
electrons out of the valence band into the conduction band.
Electrical conductivity is the result, both because the conduction
electrons can shift to new, nearby states in order to 'flow' with
the applied electric fields, and because the missing electron in the
valence electron band allows 'nearby' electrons to move into
the available state, which ALSO amounts to electric current.

    The high-energy electrons are just like free electrons,
because they aren't bound at a low energy level; these are called
'electrons'.  The low-energy electrons ARE bound, with the exception
that they can move into the (infrequent) unoccupied places; these
are called 'holes', because one tracks the interesting events by
watching the hole moving about.

    The filament of a light bulb is a conductor (a metal), so
there is no band gap, and no holes; the conduction is by nearly-free
electrons in a partially filled band.  When a 'hole' in a semiconductor
(that's what an insulator is called, when it starts to have
charges free enough to move about) hits a metal, it comes to occupy
a metallic electron state (very quickly), and ceases to be a hole.
This transition is repeated, in reverse, on the part of the complete
electric circuit that returns current to the semiconductor.

    The energy differences in the semiconductor-hole and metal
are important in thermocouple design (the energy differences
will cancel in a current path at thermal equilibrium, but will
NOT cancel if there's a temperature difference at the junctions).
The difference will usually NOT be large enough to generate visible
light, and (because metals have such a broad range of available
energy states) a single-event transition with a photon as result
is very unlikely, compared to multiple-event transitions with
no single event having enough momentum and energy to make a photon.

>2)Could you get positron holes by using a traditional Fermilab (C0 facility,
>I think that is what the main detector is called) collider and examining the
>remains(terget substance) after the collision?

    Positrons have a 'negative work function', which means they
are repelled by solid objects.  They couldn't get trapped in
solids, even if they didn't annihilate a nearby electron and vanish.
There are some interesting effects related to this, and positron
beams can be made VERY intense by canny utilization of this effect
for cooling the positrons.

>3)Could you collect holes by using semiconductors and forcing hole
>migration and then polarizing the semiconductor some how to seperate them.
>Something like an electron centrifuge?

    Every time you reverse-bias a diode, this is what happens;
the central region of the diode becomes empty of both holes and
electrons (and thus, nonconductive).  The collected charges on
the two sides are called 'space-charge regions'.  Simply collecting holes
is no challenge; they are in equilibrium with the free electrons
at some concentration (depending on temperature), and your
'collection' will recombine to restore the equilibrium whenever
you remove the field which created the space-charge region.

>4)An amplifier that preferentially amplify holes over electrons is an
>interesting thought. How about more information on holes, being an EE I
>like this idea.

    A bipolar transistor works this way.  There is both hole
current AND electron current flowing in the base, but only one of
those two currents is relevant to the transistor's output current.
Again, holes and electrons become indistinguishable when you
run the current into a conductor (like a wire), so the
recombination effect makes the effect moot.

    At high frequencies, in some very small-geometry transistors,
ballistic electron conduction has been observed (i.e. the
operation was too fast for the recombination effect), but that's
still a laboratory curiosity.

    John Whitmore