NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/alt/fan/pratchet/3830

Text File | 1992-12-25 | 3.3 KB | 63 lines

Newsgroups: alt.fan.pratchett
Path: sparky!uunet!enterpoop.mit.edu!usc!sol.ctr.columbia.edu!cunixb.cc.columbia.edu!slb22
From: [e-mail] (Seth "the Lesser")
Subject: Re: TP research
References: <[e-mail]>
Sender: [e-mail]
Organization: Generic American College Kids (G.A.C.K.)
Date: Fri, 25 Dec 1992 08:34:49 GMT
Message-ID: <[e-mail]>
Reply-To: [e-mail] (Seth "the Lesser")
X-Posted-From: cunixb.cc.columbia.edu
NNTP-Posting-Host: sol.ctr.columbia.edu
Lines: 48

[e-mail] (Andy Holyer) writes:
>
>N-type produce spare electrons floating around in the substrate (hence
>"Negative". I *think* this is what aluminium, iron etc. produce. P type are
>cuter: they make the subtrate have "holes" (as they're called) - i.e. places
>where an electron should by rights be, but isn't.

Metals are P-type dopants; nonmetals from Group V (phosphorus mostly) are the
usual N-type dopants.  You cannot dope silicon with iron; it would disrupt th
crystalline structure of the chip.  You use boron, aluminum, or anything else
from Group III as a P-type dopant.  MOS chips have layers of metal oxide
applied to a silicon (or silicon oxide?) substrate.

>The semiconductor effect occurs at the boundary between two differently-doped
>regions. MOS chips have a junction between N-type and vanilla silicon (almost
>always).

Undoped silicon is an excellent insulator, and MOS field effects don't lower
the band gap enough to make it anything else.

>So how do you bugger a semiconductor? Overdope the substrate - or dope it with
>the wrong stuff. So we have trolls taking either "uppers" or "downers", and if
>they take too much of the wrong one they keel over.  This also raises the
>vision of there being N-type and P-type trolls (with a few stick-in-the-mud
>germanium adherants, and of course all the trendy gallium arsenide boys
>wearing luminous bermuda shorts and blowing whistles at each other :-)

But doping is done during the process of laying down the chip--the dopant
vapors are mixed in with the silicon vapor at the right times.  If you want to
get impurities into the crystal matrix, you're going to have to use force to
displace the existing silicon atoms.  One might, I suppose, blow boron or
phosphorus vapor into a troll's ear, hoping that enough of it will bond to the
surface layers of his/her brain to give him/her a "buzz."  This effect would
be permanent until neutralized by the addition of just the right amount of the
other dopant; since it's hard to tell exactly how much dopant was absorbed,
troll junkies are going to be fractionally high all the time.  After enough
rounds of alternate P- and N-type doping, the troll's brain's surface layers
will be so contaminated with dopants that they will fail to work at all.  They
might even crumble off (boron phosphide shouldn't be a very stable substance,
according to my calculations :-), exposing fresh brain to be doped....

I think the radiation idea is better, myself, but I'm only a Merkin, so what
do I know?

Seth L. Blumberg          \ Warning: This posting has been determined to cause
[e-mail] (play)   \ cancer in laboratory animals.  Please make sure
[e-mail] (work) \ that no white mice are allowed to read it.
    > No one I know shares my opinions, least of all Columbia University. <