Path: santra!tut!draken!kth!mcvax!uunet!husc6!bloom-beacon!tut.cis.ohio-state.edu!ucbvax!decwrl!shelby!labrea!brooks@sierra.Stanford.EDU
From: brooks@sierra.Stanford.EDU (Michael B. Brooks)
Newsgroups: sci.physics.fusion
Subject: grain boundaries & CNF
Summary: why Grains may be important
Keywords: effective mass, D density, heat
Message-ID: <190@sierra.stanford.edu>
Date: 1 Jul 89 22:24:54 GMT
Sender: brooks@sierra.STANFORD.EDU (Michael B. Brooks)
Reply-To: brooks@sierra.UUCP (Michael B. Brooks)
Organization: Stanford University
Lines: 63


Suppose CNF doesn`t occur in the bulk of the cathode, and the recent 
discussion about "volume vs surface" effects has some merit.  That is,
suppose it is a surface effect the serves best to explain the alleged
heat outputs.  Which surface is it?  Nonsense! you say, there`s only the
outside surface.  Not quite, since all the experiments use
polyscyrstalline cathodes, and there are numerous odd internal surfaces,
that of grain boundaries.  

* grain boundaries (gbs) vary in areal density somewhat between the
cathode`s surface area (external) and it`s volume, and are a function of
material history (its specific processing).   If you believe in heat,
linked to CNF at gbs, then it makes sense that heat is proportional to 
something in between surface and volume ratios (of differing sizes
of cathodes).

* Just what is the effective mass of an electron in a gb anyway?  There
is no classical "periodic potential" as in the lattice, it is perturbed
in some way.  If one can answer that question, then what is the
effective mass after most of the localized states of the gbs are saturated
with D, as is the case with H saturated gbs traps in the well known poly
crystalline Si (and amorphous Si) thin film transistors.  These devices
depend on H binding up the "dangling bonds" to a large degree in order
to work.  Is it possible that the "density dependent Thomas Fermi 
screening length lambda" changes it`s form with a different m(sube),
that due to a different "Brillouin Zone" of the aperiodic two 
dimensional gb?
[See Horowitz` Paper, p 2]

* Just suppose that "charge up" time is just this "saturation" process, in
Pd.  If the vast majority of gb traps are satisfied and the localized
state density is reduced, then what is the maximum density that the D
can be forced to in a gb, assuming that is what the electrochemical 
potential does? Is it limited to that which is seen in the lattice?
Perhaps not, since grain boundaries have been modeled as fast routes
for diffusion (relative to the lattice).  In addition gbs have a fluid
like character, often being seen as amorphous, and can serve as relief points
for mechanical stress (can be high in localized regions).  Assuming these
dangling bonds are taken care of, then it`s conceivable that more D could
be packed into these regions at a higher density than that possible 
in the lattice, because we assume electron screening is still operable
in gbs and the lattice.  We are not necessarily limited by lattice bonding
constraints and interstial site densities (#of octahedral or tetrahedral
sites per unit cell).

 Put another way [Horowitz again]:

"If hydrogen is dissolved in a normal metal such as Palladium the width
of the fusion barrier r(subo) can be reduced both by forcing the 
equilibrium position of the atoms closer together and through electron
screening of the repulsive coulomb interaction." [page 4]

*  The esscence of the idea is that maybe this can happen in gbs that
have been electrochemically (or otherwise) saturated with D.  
Comments?

(Please remember, its all wild speculation, OK? I really don`t know
enough about this stuff to give anything much more than that).


Mike Brooks/Stanford Electronics Lab (solid state)/SU
MIT Astronomer Walter Lewin:  "Absence of evidence should never be
mistaken for evidence of absence."