Enrico Fermi.

Enrico Fermi
38 followers
Mr. Chairman, Dean Pegram, fellow members, ladies and gentlemen,
it seems fitting to remember on this 200th anniversary of Columbia University
the key role that the university played in the early experimentation
and the organization of the early work
that led to the development of atomic energy.
I had the good fortune to be associated with the Pupin Laboratory
through the period of time when at least the first phase of this development took place.
I had had some difficulties in Italy
and will be always very grateful to Columbia University
for having offered me a position in the Department of Physics
at the most opportune moment.
And in addition, this offer gave me, as I said,
the rare opportunity of witnessing a series of events
to which I have referred.
In fact, I remember very vividly the first month, January 1939,
that I started working at the Pupin Laboratory
because things started happening very fast.
In that period, Niels Bohr was on a lecture engagement in Princeton.
And I remember one afternoon, Willis Lamb came back very excited
and said that Bohr had leaked out great news.
The great news that had leaked out was the discovery of fission
and at least an outline of its interpretation.
The discovery, as you well remember, goes back to the work of Hahn and Strassmann.
And at least the first ideas on the interpretation
came through the work of Lise Meitner and Frisch,
who were at that time in Sweden.
Then somewhat later that same month,
there was a meeting in Washington
organized by the Carnegie Institution in conjunction with George Washington University
where I took part with a number of people from Columbia University
and where the possible importance of the new discovered phenomenon of fission
was first discussed in semi-jocular earnest
as a possible source of nuclear power.
Because it was conjectured,
if there is fission with a very serious upset of the nuclear structure,
it is not improbable that some neutrons will be evaporated.
And if some neutrons are evaporated,
then there might be more than one, let's say for the sake of argument two.
And if there are more than one,
it may be that the two of them, for example,
may each one cause a fission,
and from that one sees, of course,
the beginning of the chain reaction machinery.
So that was one of the things that was discussed at that conference
and started a small ripple of excitement
about the possibility of releasing nuclear energy.
At the same time,
experimentation was started feverishly in many laboratories including Pupin,
and I remember before leaving Washington,
I had a telegram from Dunning
announcing the success of an experiment
directed to the discovery of the fission fragments.
The same experiment apparently was at the same time
carried out in half a dozen of places in this country
and in three or four,
in fact I think slightly before,
in three or four places in Europe.
Now a rather long and laborious work was started at Columbia University
in order to try to firm up these vague suggestions
that had been made as to the possibilities that neutrons were emitted
and try to see whether neutrons in fact were emitted when fission took place
and if so, how many there would be
because clearly a matter of numbers
is in this case extremely important
because a little bit greater or a little bit less of probability
might have made all the difference between possibility
and impossibility of the chain reaction.
Now this work was carried on at Columbia simultaneously
by Zinn and Szilard on one hand
and by Anderson and myself on the other hand.
We worked independently and with different methods
but of course we kept close contact
and we kept each other informed of the results.
At the same time, the same work was being carried out in France
by a group headed by Joliot and von Halban
and all the three groups arrived at the same conclusion
I believe Joliot maybe a few weeks earlier than we did at Columbia
namely that neutrons were emitted and they were rather abundant
although the quantitative measurement was still very uncertain and not too reliable.
A curious circumstance related to this phase of the work
was that here for the first time secrecy
that has been plaguing us for a number of years started
and contrary to perhaps what is the most common belief about secrecy
secrecy was not started by generals, was not started by security officers
but was started by physicists.
And the man who is mostly responsible
for this certainly extremely novel idea for physicists was Szilard.
I don't know whether how many of you know Szilard
no doubt very many do
he is certainly a very peculiar man
extremely intelligent
I see that understatement
he is extremely brilliant
and he seems somewhat to enjoy
at least that's the impression that it gives to me
he seems to enjoy in starting people
so he proceeded to startle physicists
by proposing to them that
given the circumstances of the period
you see it was early 1949
and war was very much in the air
given the circumstances of that period
given the danger that atomic energy and possibly atomic weapons
could become the chief tool for the Nazis to enslave the world
it was the duty of the physicists to depart
from what had been the tradition of publishing
significant results
as soon as the physical review or other scientific journals
might turn them out
and that instead
one had to go easy
keep back some results
until it was clear
whether these results were potentially dangerous
or potentially helpful to our side
so Szilard talked to a number of people
and convinced them
that they had to join some sort of
I don't know whether it would be called a secret society
or what it would be called
anyway
to get together
and circulate this information privately
among a rather restricted group
and not to publish them immediately
he sent in this vein
a number of cables
to Joliot in France
but he did not get
a favorable response from him
and Joliot published his results
more or less like
results in physics had been published
until that day
so that the fact
that neutrons are emitted in fission
in some abundance
order of magnitude of 1 or 2 or 3
became a matter of general knowledge
and of course
that
made the possibility
of a chain reaction
appear to most physicists
as a
vastly more real possibility
than it had until that time
another important phase of the work
that took place
at Columbia University
is connected with a suggestion
on purely theoretical arguments
by Bohr and Wheeler
that of the two isotopes of uranium
it was not the most abundant
uranium-238
but it was the least abundant
uranium-235
present as you know
in the natural uranium mixture
to the tune of 7 tenths of a percent
that was responsible
at least for most of the fission
the argument had to do
with the even number of neutrons
in uranium-238
and the odd number of neutrons
in uranium-235
which according to a discussion
of the binding energy
that was carried out
by Bohr and Wheeler
made plausible
that uranium-235
should be more fissionable
now it clearly was very important
to know the facts
also experimentally
and
work was started
in conjunction
by
Dunning and Booth
at Columbia University
and Neer
Neer took the
mass spectrographic part
of this work
attempting to separate
a minute
but as large as possible amount
of uranium-235
and Dunning and Booth at Columbia
took over the part
of using this minute amount
in order to test
whether or not it would
undergo fission
to a much greater
with a much greater cross-section
than ordinary uranium
well you know
of course by now
that this experiment
confirmed
the theoretical suggestion
of Bohr and Wheeler
indicating
that
the key isotope
of uranium
from the point of view
of any attempt
of for example
constructing a machine
that would develop
nuclear energy
was in fact uranium-235
now you see
the matter is important
primarily for the following reasons
that at the time
were appreciated
perhaps less definitely
than we could discuss
the matter at the present moment
the point
in
fabricating a chain
reacting machine
is of course
the fundamental point
is to see to it that each fission
produces a certain number
of neutrons
and some of these neutrons
will again produce fission
if an original fission
causes more than one
then of course the reaction
goes if an original fission
causes less
than one subsequent
fission then the reaction
does not go
now
if you take the isolated
pure isotope
U-235
you may expect that the
unavoidable losses of neutrons
will be
minor and therefore
if in the fission
somewhat more than one
neutron is emitted then it will
be merely a matter of
piling up enough U-235
to
obtain a chain reacting
structure
but if to each
gram of U-235
you add some 140
grams of U-238
that come naturally
with it then the
competition will be greater because
there will be all this
ballast ready to snatch
away the not too abundant
neutrons that come out
in the fission and therefore
it was clear
at the time that
one of the ways
to
that one of the ways
to make
possible the production of a chain
reaction was to
isolate the isotope
U-235 from
the
much more abundant
isotope U-238
now
at present we
have in our
laboratory a row of bottles
labeled more or less
isotope
what shall I say
56 for example
or uranium
235 or uranium
238 and these bottles
are not quite as common as
would be a bottle of
ordinary a row of
chemical elements
but they are perfectly easily
obtainable by
putting the due pressure on
the Oak Ridge
laboratory
but at that time
isotopes were
considered almost
magically inseparable
there was to be
sure one exception namely
deuterium was already
at that time available
in bottles
but of course
deuterium is an isotope
in which the two isotopes
hydrogen 1 and hydrogen 2
have a ratio of mass
1 to 2 which is a very great ratio
but in the case of uranium
the ratio of mass is
merely 235 to
238 so the difference is
barely over 1%
and that of course
makes the differences
of these two objects
so tiny
that it was not very clear
that the job of separating
large amounts
of uranium
235 was one
that could be taken seriously
well
therefore
in those
early years
near the end
of 1939
two lines
of attack to the problem
of
atomic energy
started to emerge
one
was as follows
the first step should be
to separate in large amounts
amounts of kilograms
or maybe amounts of tenths of kilograms
or maybe of hundreds of kilograms
nobody really knew
how much would be needed
but something perhaps in that
order of magnitude
separates such
at that time fantastically
large looking amounts
of uranium
and then
operate with them
without the balance
of the associated much larger
amounts of uranium 238
the other school of
thought
was predicated
on the hope that perhaps the neutrons
would be a little bit
more
and that perhaps
using some little
amount of ingenuity
one might use them
efficiently
and one might perhaps
be able to achieve
a chain reaction
without having to separate
the isotope
as I said at that time
looked
almost beyond human
possibilities
now I personally worked for many
years with neutrons
and especially
with slow neutrons
so I associated
myself with
the second team that wanted
to use non-separated
uranium and try to do the best
of it
early attempts and studies
on and discussions
on how to
separate
the isotopes of uranium
were started by
Dunning and
Booth in consultation
in close consultation
with Professor
Duray
on the other hand
Szilard, Zinn, Anderson
and myself started
experimentation on the
other line which
its first element involved
lots of measurements
now I have never
yet quite understood
why our measurements
in those days were so poor
it seems
well, presumably one learns
I'm noticing now that the
measurements that we are doing
on pi on physics are very
poor, presumably just because
we have not learned the tricks
and of course the facilities
that we had at that time
were
not as powerful as they are
now it's much easier to carry
out experimentation with neutrons
using a
pile
as a source of neutrons
than it was in those days
using radium
beryllium sources when
geometry was the
essential item to
control or using the
cyclotron when intensity
was the desired
feature rather than
geometry
well, we soon reached the
conclusion that
in order to have
any chance of success with
natural uranium we
had to use slow neutrons
so there had to be
a moderator and this
moderator could have been
first water
or other substances
water was soon discarded
it's very effective in
slowing down neutrons
but still absorbs a
little bit too many of them
and we could not afford that
then
it was thought that
graphite might be perhaps a better
bet it's not as efficient
as
water
in slowing down neutrons
on the other hand little enough
was known of its
absorption properties that
the hope that the absorption
may be very low was quite
tenable
this brings us to the
fall of 1939
when
Einstein wrote his
now famous letter
to President
Roosevelt
advising him of
what was
the situation in physics
what was brewing and
that he thought that
the government had
the duty to take an interest
and to help along this development
and in fact
help came along to the tune of
$6,000
a few months after
and
the $6,000
were
used
in order to buy
a huge amount of what seemed
at that time when the eye
of physicists had not yet
been distorted
what seemed at
that time a huge amount
of graphite
so physicists on the
seventh floor
of the laboratory
started looking like coal mines
and
the wives to whom
these physicists came back
tired at night
just were wondering
what's happening we know that there is
smoke in the air but after all
well what was happening
is that in those days we were
trying to learn something about
the absorption properties
of graphite because perhaps graphite
was no good
so we built
columns of graphite
maybe four feet on the side or
something like that maybe ten feet
high it was the first time
when
apparatus in physics and these graphite
columns were apparatus
was so big that you could climb on
top of it and you had to climb on top
of it while cyclotrons were the same
way too but anyway the first time
when I started
climbing on top my
equipment because it was just too tall
I'm not a tall man
and
then
sources of neutrons were
inserted at the bottom and we were
studying how these neutrons were
first slowed down and then
diffused up the column
and of course if there had been
a strong absorption they would not
have diffused very high but
because it turned out that the absorption
was in fact small
they could diffuse
quite readily
up this column and by
making a little bit of mathematical
analysis of
the situation it became
possible to make the first guesses
as to what was
the absorption cross section of graphite
a key element
in deciding the possibility
or not of fabricating
a chain
reacting unit with graphite
and natural uranium
well I will
not go into
detail
of this experimentation that
lasted really quite
a number of years
and required really quite
many hours and many days and many
weeks of extremely hard
work
I may mention
that very early
our efforts
were brought in connection with
similar efforts that were taking
place at Princeton University
where a
group with
Wigner, Kreutz
and Bob Wilson
set to work making
some measurements that we had
no possibility
of carrying out at Columbia
University
well as time went
on we
began to identify
what had to
be measured and how
accurately
these things
that I shall call eta
f and p
these three quantities
I don't think I have time to define them for you
these three quantities
eta f and p
had to be measured with
a certain accuracy
to establish what
could be done and what could not be done
and in fact
if I may say so the products
of eta f and p
had to be greater than
one
it turns out we now know that
if one does just about the best
this product can be
1.1
so if we had been able
to measure these three quantities
to the accuracy of 1%
we may have found that
the product was for example
1.08
plus or minus .03
and if that had been the case
we would have said let's go ahead
or if the product
had turned out to be
0.95 plus or minus
.03 perhaps
we would have said just
this line of approach is not
very promising and we better look for something
else
however I have already
commented on
the extremely low quality
of the measurements in
neutron physics that could
be done at the time where the
accuracy of measuring separately
either eta or
f or p was
perhaps with a plus or minus
of 20%
if you compound
by the well known rules
of statistics three errors
of 20%
well you find something around
35%
so if you should find
for example .9
plus or minus .3
what do you know?
if you find 1.1
plus or minus .3
again you don't know anything much
so that was
the trouble and in fact
if you look in our
early work
what was the detailed values
given by this or that
experimenter to
for example eta
you find that it was
of 20% and sometimes
greater amount in fact
I think it was
strongly influenced by
the temperament
of the physicist
shall we say
optimistic physicist seemed
unavoidable to push
these quantities high
and pessimistic physicist like
myself tried to keep
them somewhat on the low side
anyway nobody really knew
and
we decided therefore
that one had to do something else
one had to devise some kind of
experiment that would give
one a
complete overall
structure directly of the
product eta fp
without having to measure separately
the three because then
perhaps the error
would sort of drop down and would
permit us to reach conclusions
well we went
to Dean Pegram who was
then the
man who
could
carry out magic around
the university
and we explained to him
that we needed a big
room and when we say big
we meant a really big room
perhaps he made a
crack about a church not
being the most suited place
for a physics laboratory
in his talk but
I think a church would have been
just precisely what we wanted
well he
scouted around the campus
and we went with him
through dark corridors
and under various
heating pipes and so on
to visit
possible sites for
this experiment and eventually
a big room, not a church
but something that might have been
compared with size
with a church
was discovered in Skammerhorn
and
well there we started
to construct this structure
that at that time looked
again an order of magnitude
larger than anything that
we had seen before
actually
if anybody would
who would
look at that structure now would
probably extract this magnifying
glass
and
go close to see
it but for the ideas
of the time it looked real big
it was a structure
of graphite bricks and
spread
through these graphite bricks in some sort
of pattern where big cans
cubic
cans containing uranium oxide
now
graphite is a black
substance as you probably
know so is uranium oxide
and to handle
many tons
of both makes people very black
in fact it requires even strong
people and
so well
we were
reasonably strong but I mean
we were after all
thinkers
so
so Dean Seagram
again looked around
he said well that seems to be a job
a little bit beyond your
feeble strength
but there is a football squad
at Columbia
that contains a dozen
or so very husky boys who
take
jobs by the hour
just to carry them through
college
why don't you hire them
and it was a marvelous idea
it was really a pleasure for
once to direct
the work of these husky boys
canning uranium just shoving
it in handling
packs
of 50 or 100 pounds
with the same ease as
another person would have had
3 or 4 pounds and
thrashing them in these cans
and just fumes of
all sorts of colors mostly black
would go in the air
well so
grew what was
called at the time the
exponential pile
it was an exponential pile
because in the theory an exponential
function enters which is not surprising
and was a structure that was designed
to
that was designed to
test
in an integral way without
going down to fine details
whether
the reactivity of the
pile, the reproduction factor
would be greater or less than 1
well it turned out
to be 0.87
now that is by 0.13
less than 1
and it was bad, however
well at that moment
we had a firm
point to start from
and
we had essentially
to see whether we could squeeze
the extra
0.13 or a little bit
more preferably
now there were many obvious things
that could be done, first of all I told
you these big cans were
canned in
10 cans
so what has the
iron to do, iron can
do only harm, can absorb nutrients
and we don't want that, so
out go the cans
then
what about the purity of the materials
we took
samples of uranium and
with our physicist
lack of skill in chemical
analysis we sort of
tried to find out the impurities
and certainly there
were impurities, we would not know what they
were, but they looked impressive
at least in bulk
so
now what do these impurities
clearly they can do only harm
maybe they make harm to the
tune of 0.13%
finally the
graphite was
quite pure
for the standards
of that time when
graphite
manufacturers
were not concerned with avoiding
those special impurities
that absorb nutrients
but
still there was some
considerable gain
to be made
out there
and
especially Szilard
at that time
took extremely
decisive and strong
steps to try to
organize the
early phases
of production of pure materials
now he did a marvelous job
which later on was
taken over by a more powerful
organization that was Szilard himself
although to match Szilard
it takes a few
able-bodied customers
well this
brings up to
Pearl Harbor
at that time in fact I believe
a few days before
by accident
the interest in
carrying
through the
hydrogen work was spreading
work somewhat
similar to what was going on at
Columbia had been initiated
in a number
of different
universities throughout the country
and
the government started
taking decisive
action in order to
organize the work
and of course Pearl
Harbor gave
the final and very decisive
impetus to
this organization
and it was decided
in the high councils of the
government
that
the work on
the chain reaction produced
by non-separated isotopes
of uranium
should go to Chicago
at that time
that is the time when I
left Columbia University
and after
a few months of commuting between
Chicago and New York eventually
moved to Chicago
to keep up the work
there
and
from
then on
with a few
notable exceptions
the work at Columbia
was concentrated
on the isotope separation
phase of the atomic energy
project
as I have indicated the work
was initiated
by Booth, Dunning and
Urey about 1940
1939 and
1940
and
with this reorganization
a large laboratory
was started at Columbia
under the direction
of Professor Urey
the work
there was extremely
successful and rapidly expanded
into
the build up of a huge
research laboratory which
cooperated with
the carbon carbide
company in
establishing
some of the separation
plans at Oak Ridge
this was one
of the three horses
on which
the directors
of the atomic energy project
had placed their bets
and as you know the three
horses arrived
almost simultaneously
to the goal in the summer
of 1945
I thank you
applause
