Path: santra!tut!draken!kth!mcvax!uunet!labrea!husc6!yale!yalevm!IVEKEIC
From: IVEKEIC@YaleVM.YCC.Yale.Edu (Keith Calvert Ivey)
Newsgroups: alt.fusion
Subject: Yale fusion seminar (Moshe Gai)
Message-ID: <447@YaleVM.YCC.Yale.Edu>
Date: 29 Apr 89 03:35:40 GMT
Reply-To: IVEKEIC@YaleVM.YCC.Yale.Edu
Organization: Yale University, New Haven, CT, 06520, USA
Lines: 259
Disclaimer: Author bears full responsibility for contents of this article
 
    I've  been following this newsgroup since it started, and I  finally

have something to contribute.
    Today  (28 April 1989) at 2 pm the physics department here  at  Yale
had  a special seminar by Moshe Gai (a nuclear physics  professor  here)
with the title "Does Cold Fusion Exist?".  My notes on the seminar  fol-
low.
 
    Prof. Gai gave a seminar yesterday with an almost identical title as
part of an undergraduate research colloquium.  Greg Howard posted  notes
on  this yesterday.  Today's seminar was, I believe, more technical.   I
guess yesterday's seminar drew off some of the crowds of fusion enthusi-
asts.   I arrived at the room only twenty minutes before the  talk,  and
was  just the third person there.  By shortly after 2 there  were  about
200 people, only a couple of whom seemed to be with the media.
    Prof.  Gai began by saying that his results would be written  up  in
about  two  weeks.  He is speaking about them at the  American  Physical
Society conference on Monday, and this seminar was a trial run for that.
He  said that his results are negative and that Princeton and  MIT  have
come out with similar results within the last week or so, and that there
are rumors that the Cal Tech experiments agree with his too.  He  admit-
ted  that he might have been "out of line" in saying some of the  things
he did to undergrads yesterday.  The work involves "seven people and  at
least  ten opinions".  The workers are nuclear physicists from Yale  and
chemists from Brookhaven National Laboratory (in Upton, New York).
    He  gave  the chemical equations for the process  of  absorption  of
deuterium into a metal lattice during electrolysis:
 
       -              -
D O + e  --> D    + OD
 2            ads
 
                      -
D    + D O --> D  + OD
 ads    2       2
 
 
D    --> D
 ads      lattice
 
where  D(ads) is a deuterium atom adsorbed on the electrode surface  and
D(lattice) is one absorbed into the lattice.  When fully saturated,  the
lattice contains one hydrogen molecule per titanium atom or one hydrogen
atom per palladium atom.
    Possible nuclear reactions between two deuterons are:
 
          { 3
d + d --> {  He (0.82 MeV) + n (2.45 MeV)       ~50%
          {
          { t (3.03 MeV) + p (1.01 MeV)         ~50%
          {
          { 4
          {  He (0.08 MeV) + gamma (23.77 MeV)  ~10^-4
 
The  probabilities  given are for "normal" conditions.  There  has  been
much theorizing about the possibility that they change when the  deuter-
ons are confined in a lattice.
    Since  heavy water will always have a considerable amount of  normal
H2O contamination and since protium diffuses into palladium much  faster
than  deuterium  does, fusion of a deuteron and a proton  must  also  be
considered:
 
          3
p + d -->  He (0.005 MeV) + gamma (5.49 MeV)
 
A  paper  by Koonin and Nauenberg to be published in  _Nature_  suggests
that this p+d reaction occurs about 10^8 times faster than the d+d reac-
tions  given above.  This combined with the isotope effect on  diffusion
mentioned  above means that the proportion of p+d to d+d reactions  will
be much greater than one would expect based simply on the ratio of  pro-
tons to deuterons in the electrolyte.
    Gai  is  "very bothered" by the neutron measurements  given  in  the
Fleischmann and Pons paper, especially the 4x10^4 neutrons/sec given  as
the flux, which he thinks is remarkably high.  He would like to see  the
raw data so he can figure out how they treated the background.  He  also
complained about the strange nonlinear scale given on their spectrum  of
beta  emissions, which seems to indicate negative energies on  the  left
side of the graph.  The F&P paper was overall very difficult for him  to
figure  out.   He likes the Jones paper much better  because  they  give
"real numbers".
    The  Yale-Brookhaven  setup consists of four(?)  electrolytic  cells
partially  surrounded  by six neutron detectors and  two  sodium  iodide
crystal detectors for gamma rays.  This is enclosed in ~15 cm of borated
concrete  and ~15 cm of borated paraffin, and topped by two  cosmic  ray
detectors  so that possible muon-catalyzed fusion resulting from  cosmic
rays can be "vetoed".  They had several tons of lead to use as shielding
for  a  while, but got rid of it because the lead itself was  causing  a
high  background.  A neutron coming from the experiment  interacts  with
the  first neutron detector (#0), which sits directly below  the  cells,
and then scatters to one of the other five which are arranged in a ring.
They require coincident signals from two detectors (#0 and one other) to
give  a neutron count.  They can get some energy information  about  the
neutrons with this setup, but the placement of the detectors requires  a
compromise  between efficiency of detection and precision of energy  in-
formation.   Signals from gamma rays and neutrons can  be  distinguished
easily by the shapes of the pulses.
    Nitrogen  gas  is cycled through the cells to remove  hydrogen  gas,
keeping  it below the 4.8% required for an explosive mixture  with  air.
The nitrogen is wetted with D2O to replace that lost by electrolysis.
    They used nine electrodes and nine electrolyte solutions in  various
combinations.  They wanted to check every suggestion they had heard  for
getting  things  to work.  I didn't get all the  information  about  the
electrodes and electrolytes, but here's what I have:
 
Electrodes:
 
1) Pd plate - cold-worked (pounded with a sledge hammer to create dislo-
cations in the lattice structure), then heated in D2 (300 degrees C, 120
psi) and anodized.
 
2) Pd cylinder - annealed in flowing argon at 1000 degrees C.
 
3) Pd cylinder - annealed in flowing argon at 1000 degrees C.
   (There was a difference between this and #2 that I didn't get.)
 
4) Pd cylinder - annealed in vacuum ("super electrode").
 
5-8) Ti parallelepipeds, cold-worked.
 
9) TiFe   Mn    powder, "hydrided" at 120 psi D  at 900 degrees C,
       0.7  0.2                                2
charged on 19 Dec 87 and recharged on 04 Apr 89.  Contained in a 2x20 cm
cylinder  pressurized to 120 psi.  (I don't understand how exactly  this
was used as an electrode, if it was.)
 
 
Electrolytes:
 
1) 0.1 M LiOD, 97.5% D2O
 
2) 0.1 M LiOD, 99.8% D2O
 
3) 1 M LiOD, 97.5% D2O
 
       6
4) 1 M  LiOD, 97.5% D2O
 
5-8) the  solution of Jones et al. (100 g D2O plus 0.125 g each of  var-
ious  salts.  Someone asked if they included "a small amount  of  AuCN".
He said no.)
 
9) 0.1 M LiOD, 99.3% D2O
 
    In order to test the hypothesis that "ignition" by energetic  parti-
cles was necessary to start the fusion, Gai disassembled the smoke alarm
from  his home and spot-welded its americium source to electrode #1  for
some of the experiments, thus providing 5 MeV alpha particles.
    The neutron detection employed "state-of-the-art" pulse-shape detec-
tors  not yet commercially available.  The threshold for neutron  detec-
tion was ~0.5 MeV.  Efficiency of detection, taking into account coinci-
dence was ~1%.  The signal was filtered by software to remove gamma  ray
signals in counting neutrons and to exclude neutron counts with energies
greater than 3 MeV.
    The most thoroughly analyzed data comes from the last seven hours of
the experiment.  Gai displayed data showing that during this time detec-
tor  #1  counted  a grand total of 2 neutrons,  which  the  group  named
"Fleischmann"  and  "Pons".  (I think that was a bit of a  cheap  shot.)
The  total  number of neutrons counted in the 7 hours was  17  with  the
cells off and 13 with all cells on (some had been running for two  weeks
at this point).  There was thus no statistically significant difference.
    In  31 hours on 18 April the sodium iodide crystals detected no  p+d
or d+d gamma rays.  Background gamma rays from decay of radioisotopes in
the concrete were detected at 2.1 and 2.6 MeV.  Gai wonders if this  may
explain  the 2.5 MeV signal seen by Fleischmann and Pons with their  NaI
crystals.
    Gai gave the three-standard-deviation upper limits on fusion  yields
as  <  2x10^-25  fusions/deuteron pair/sec for  d+d  (based  on  neutron
counts)  and  < 2x10^-22 fusions/pair/sec for p+d (based  on  gamma  ray
counts).  He says the first compares favorably with the number given  by
Jones et al., 10^-23 (what's a factor of 50 between friends?).
    At  one  point they saw a shoulder in the gamma ray  spectrum  which
ended at about 5.5 MeV.  This shoulder "weighed on their shoulders"  for
quite a while, but the eventually determined that it was residual radio-
activity from their calibration of the NaI crystals with a Pu-Be source.
Na-23 in the detector captures a neutron to give Na-24 which decays with
a half-life of ~15 hours.  The shoulder remained even when shielding was
interposed between the detector and the cells, so it could not have been
coming from the experiment.
    Gai  stressed  several times that he was "not making  any  statement
whatsoever  about nonreproducibility of the result of Pons and  Fleisch-
mann".   He  said that they had done their best to repeat what  F&P  had
done  -- though it was somewhat difficult to figure out some of the  de-
tails -- and had seen no remarkable results.
    After this there was a question-and-answer session.  The first ques-
tion was from someone who claimed that palladium actually absorbs deute-
rium  faster than protium.  Gai said that papers by Pons had  shown  the
opposite.
    Then  there was a question (from an inorganic  chemistry  professor)
about  the heat F&P had gotten and how their electrode had melted.   Gai
said his measurements were only relevant to the emission of neutrons and
gamma rays, and that he was making no statement concerning heat.  Howev-
er,  he claimed that putting lithium in palladium decreases the  melting
point  by "about an order of magnitude".  (I find this difficult to  be-
lieve  unless  he's talking about a significant percentage  of  lithium,
essentially an alloy.)  He also said that on of the Brookhaven  chemists
had a completely chemical explanation for the melting of the  electrode,
but he didn't reveal what this involved.
    To  a  question about the possibility that He-4  could  lose  energy
gradually by a cascade of low-energy gamma rays, Gai responded that this
was  impossible since the first excited state of He-4 has an  energy  of
~20 MeV.  There is no way for it to lose energy in small steps.
    The  next question came from a physical chemistry professor who  was
at  the University of Utah before he came to Yale, and who said  he  had
visited there yesterday.  He said that the metallurgy department at U of
U has reproduced the heat production of F&P's experiment.  According  to
what  they  told him, cold-working the electrodes was  about  the  worst
thing one could do.  He said they prepared their electrodes in a special
way.  Gai asked him, "What's the secret?", but he wouldn't talk about it
with the press in the room.  He talked about it after the seminar  broke
up.   It turned out the secret was using cast electrodes; this has  been
on the net for several days at least.  The other things was that F&P ran
their  electrolysis  with the electrodes  completely  submerged.   Also,
supposedly there is only one guy who can make electrodes that work.  One
wonders if they've taken out an insurance policy on him with Lloyd's  of
London.
    Gai  refused to comment on Hagelstein's theory about dissipation  of
He-4  energy  by loss to the lattice, but several in the  audience  said
they couldn't imagine how the lattice could absorb 20 MeV without  melt-
ing.   Gai did point out that Stanford has found a lot of He-4 in  their
electrode.
    To a question about the Mossbauer effect, Gai replied that it  would
not  apply  in the case of He in a metal lattice,  since  the  Mossbauer
effect requires that the energetic nucleus be part of the lattice and He
would be in the open space.
    After the seminar a nucleus of chemistry and physics faculty  formed
about Gai and the chemistry professor recently returned from Utah.  This
nucleus was surrounded by a small cloud of graduate students,  including
me.  There was much trading of rumors, most of which I had already heard
on  the net.  The chemistry professor said that one of the main  reasons
that things were such a mess was the behavior of the U of U  administra-
tion (one of the reasons he left there).  He said that U of U was trying
to keep as many details secret as possible because of patents.  He  also
said that the administration had kept F&P from putting the name of their
student  (Hawkins(?)) on the paper as an author, so as to  decrease  the
number  of  people involved in possible patent claims, but that  in  the
errata  to the article this had been corrected, along with  the  strange
scales  in the figures and such.  "We are sorry that the name of one  of
the  main contributors to this work was inadvertently left off the  list
of  authors"?  He also said F&P believe that the Italians have the  best
information about the F&P setup because they sent people over very early
before the administration cracked down on communication.  Also, he  said
F&P were going to Los Alamos to help with replication there.
 
    I  don't think Gai's title for the seminar was a good one, since  he
didn't  even pretend to give an answer to this question of whether  cold
fusion exists.  It seems that his negative results are irrelevant  since
he  did  not measure heat and did not use cast  electrodes.   Why  can't
anyone measure everything on the same experiment: neutrons, gamma  rays,
and  heat?  What were the chemists for if not to help with  calorimetry.
I  guess  to help with electrochemistry.  I hope Gai brushes up  on  his
chemistry before going to the APS meeting.  He thought LiOH was  lithium
_hydride_  and  he repeatedly used "hydriding"  or,  worse,  "hydration"
instead of "hydrogenation".  (Maybe it should be "hydridation" in analo-
gy with "oxidation".)  I guess he doesn't need to know this stuff to  do
the experiments, but it makes a bad impression.
    Well, that's about it.  I hope I didn't make too many errors.
========================================================================
Keith Calvert Ivey <ivekeic@yalevm.BITNET>
Yale University Department of Chemistry
Box 6666, Room 1 SCL, New Haven, CT 06511