2001 Nobel Prize lecture by Eric A. Cornell in physics

Scientists won the Nobel Prize
12 followers
Guten Morgen.
It's traditional when one gives a talk in physics to say ahead of time that one is very happy to be here giving the talk.
I must say that I'm particularly happy to be here giving this talk.
Karl has done a wonderful job laying out the general framework of the experiments.
I will be presenting one or two more fundamental ideas, introductory ideas,
and then describing a couple of the experiments that we're doing in this talk.
So I'll tell you who's doing the Bose condensation and particularly why it's very difficult to do
and what's sort of new about Bose-Einstein condensation as a system, as a macroscopic quantum system.
And then I'll tell you some particular experiments we're doing.
And I want to acknowledge right at the beginning 12 years of very talented, very dedicated,
very personally to me inspirational graduate students and postdocs with whom I've worked all these years.
And in particular I show you a picture of some of my early students and postdocs.
All four of these people, Mike Anderson, Debbie Jin, Mike Matthews, and Jason Entscher,
you'll see around Stockholm on Monday I think they'll be somewhat more formally attired.
So where does Bose-Einstein condensation happen?
Well it happens here in Boulder where the plains meet the mountains.
You can see this flat part here extends all the way out to Missouri and this goes more or less straight up into the sky.
And more particularly because I'm an experimenter I have to show you the picture of my favorite machine.
I work for the National Institute of Standards and Technology and I'm actually by federal,
U.S. federal statute required to tell you that the distance from here to here is 2.54 centimeters.
Okay, why then is Bose-Einstein condensation interesting?
I'm sure in Sweden and certainly in the United States there are many fashion magazines,
magazines about celebrities and so on and it's traditional in these magazines
every year at the end of the year they say, they give you a list of the 10 most intriguing people.
So I'm going to, I'd like you to as an exercise to try and compose for yourself a list of the 10 most intriguing physics.
And to narrow it down a little bit I say it, physics over a range which is larger than an atomic nucleus
and smaller than a planet say.
The sort of the human scale physics.
Everyone would come up with his or her own list.
It's impossible to have any, no two lists would be the same.
But I wager that many people's lists would include in that 10 at least two and possibly even three of the following things.
Superfluidity, superconductivity, lasing.
And I mentioned these three in particular, they certainly are on my list because they have something in common
which is they all involve the macroscopic occupation of a single state.
A great many particles, be they photons in a laser or helium atoms in superfluidity or Cooper pairs in superconductivity.
A great many particles all occupying the same wave function.
And this also is the thing which best characterizes Bose-Einstein condensation.
So ahead of time you might think, well Bose-Einstein condensation is somehow like these other things
which we know to be interesting.
And there are considerable overlap.
You can think of superfluidity as being a liquid.
Superconductivity is happening in a solid.
Lasing, this is a pure energy.
And now as we sort of fill in the box here with a macroscopic occupation of the gas,
the gas phase Bose-Einstein condensation.
There's a lot of overlap between these ideas.
Analogies, also a lot of contrast.
I'm going to be talking to you about in particular about sort of overlaps between Bose-Einstein condensation and superfluidity but not exclusively.
Carl Wyman mentioned to you that as atoms get colder and colder,
they act less and less like particles and more and more like waves.
The act of making atoms or making anything colder makes them also somehow wavier at the same time.
How can we understand this?
I know that there are some students in the audience
and I'm sure you've already learned about the uncertainty principle,
which is that Heisenberg's uncertainty principle is that you can't both measure where something is and how fast it's going.
If you know where something is very well, so its uncertainty and its position is small,
you always must have a large uncertainty in its momentum.
This product can never be smaller than Planck's constant.
So if you think about it, in some sense, if you put a bunch of atoms into a box,
you know that if the atoms are all in the box and the box is not moving,
the average momentum of any given atom in that box is zero.
When you measure the temperature of that box, when you find out how hot the box is,
you're measuring the spread in momentum.
After all, the temperature of the box is just proportional to the mean square momentum.
Therefore, somehow when you measure the temperature of this box,
you're measuring the spread in momentum of the atoms in this box.
As the temperature gets colder and colder,
you know that the uncertainty in the momentum of the atoms,
with more and more certainty, the momentum of these atoms becomes zero
as the temperature gets closer and closer to zero.
Therefore, the uncertainty in the position becomes larger and larger.
It's really this uncertainty, the fact that you can't really know where an atom is,
which is the quantum wavelength of the atom,
which is the thermal de Broglie wavelength of the atom,
which gets larger and larger as the temperature gets smaller and smaller.
I want to explain to you now why it is that making Bose-Einstein condensation in a gas
is a hard thing to do, because indeed it's a very hard thing to do.
As Carl told you, it took many, many years,
a lot of ingenuity on the part of a great many people.
I have here a plot, the log of temperature in this direction, density in this direction.
If our condition for a Bose-Einstein condensation is that the thermal de Broglie wavelength
has to be incomparable to the inter-atomic spacing,
we can make sort of a line here, down here at higher densities and lower temperatures.
We have Bose condensation up here at higher densities or lower density.
Higher temperatures or lower density, we don't have Bose condensation.
I can also make a similar plot now of basically all of thermodynamics.
I'm going to try and put on this plot for you here.
Typically, in thermodynamics, you have what's called a phase diagram.
It shows you where the liquids and the solids are.
You may be more familiar with it with temperature on one scale and pressure on one scale,
but you can also plot temperature on one scale and density on the other scale.
In a plot like this, they all look essentially the same.
Up here, you have gases at high temperatures and low densities.
Over here, you have all of condensed matter physics.
There's a large region, there's a wedge in here, which is forbidden by thermodynamics.
What I mean by forbidden is, suppose we tried...
I don't have any units on this plot because this applies to everything.
Whether you're using any particular material, we'll follow on this plot.
It changes the scale, but the shape of the plot always looks the same.
Suppose this is for water.
I take water and I arrange it at this temperature, which is 20 degrees Celsius.
Say I have a one liter bottle of water and I put 100 centiliters of water in it.
Now I've got a mean density in there of 0.1 grams per centimeter.
But of course, I don't actually have water at 20 Celsius and 0.1 grams per centimeter density
because what really happens is, at the bottom of the jar, you have water.
At the top of the jar, you have water vapor.
But nowhere in the jar do you have water in this forbidden region.
Thermodynamics doesn't allow it.
Why am I emphasizing this forbidden region of thermodynamics?
Because if we put the two former plots together, this plot and this plot together,
we find, again, for any given material, Bose-Einstein condensation occurs
down here in this region, which is forbidden by thermodynamics,
unless you're at very high density.
But at very high density, things solidify and solids can't Bose condense.
The only material that's an exception to this is helium.
Helium does not solidify in the Bose-Einstein condensation regime, but it does liquefy.
And in some sense, the point of the research that we've been engaged in
is to make a gas phase Bose condensate, something which is much lower density,
easier to understand in a single particle model.
And to do this, we have to come down here into this region, which is forbidden.
Now, you might ask yourself, if it's forbidden, does that mean it's impossible?
And the answer to this really, really lies in theology.
And by theology, I don't mean that you must believe that it happens.
But in fact, what you learn very soon in any theological studies
is that the things which are truly impossible, no one bothers to forbid.
So the fact that it's forbidden is actually a very good sign.
And another important theological insight is that if it weren't fun,
also again, they wouldn't bother to forbid it.
So I think we have very good motivation to proceed in this direction.
And in some sense, the history of Bose-Einstein condensation
through the 70s and the 80s and the 90s
is in efforts to come down from a gas down into this forbidden region
to where Bose-Einstein condensation happens.
I'm making this very mysterious, but of course, to the chemists,
this forbidden region is familiar under the term of supersaturation.
In physics, we call it metastability.
If you have a gas and you keep it at low enough densities,
it's possible to come down here.
So the motivation then is to go actually to rather low densities,
which forces us then to lower temperatures, as Carl has told you.
What are Bose-Einstein condensations like?
Carl's already told you.
They tend to be rather small by a standard of a chemical sample,
perhaps only a femtogram.
But in Wolfgang's lab, now it's more than 10 million,
and in Dan Kleppner and Tom Greytak's lab,
it's more than 100 million atoms.
It's very, very cold, much colder than liquid helium.
The samples are in some sense quite thin.
This little ball of atoms that we form in this magnetic trap
is a million times less dense than the air molecules in front of you.
But even though it's so tenuous, the density is so low,
the texture of this is gelatinous,
like aspic.
I don't know what the appropriate Swedish dish is,
but something which is sort of goopy.
Do we have aspic? Gelatin? Jell-O?
Anyway.
Jell-O?
How is it that it develops this texture, this goopy texture?
Okay, we have all of the atoms in this condensate
are performing exactly in the same quantum wave.
You might ask yourself,
if every single atom is doing exactly the same wave,
do these atoms collide into each other?
And the answer is yes or no.
They don't collide into each other in the traditional sense
that this atom bumps into this atom,
and one atom goes up and one atom goes down,
or something like this,
because the Bose-Einstein condensation mechanism
keeps the atoms all in the same wave.
But what does happen is that
because the atoms are all in the same wave
and they're undergoing what are called virtual collisions in some sense,
there's a shift in their energy.
They feel a repulsion from one another,
which we describe as a change in their energy,
which is proportional to the density.
This shift in energy per atom is proportional to the density,
and to something which is proportional to the square root
of the collisional cross-section,
of the probability of colliding.
And we usually take the positive square root, but not always.
And the effect of this repulsion, then,
is if we have this magnetic trap,
this parabolic bowl in which we're holding the atoms,
the atoms, instead of sitting in a little tiny blip,
which is the size of the 0, 0, 0 state
of this three-dimensional simple harmonic oscillator,
they repel each other out
and they self-consistently assume a size
which is quite a bit larger, 10, 20, 50 times larger,
and if the bowl that you're holding the atoms in is a parabola,
the density distribution is an inverted parabola,
and this is sort of the shape of the atoms
that we're working with.
Now, I'm only going to do this for a little while,
but I am going to show you an equation.
This is the Gross-Pitayevsky equation.
For the physicists and the chemists in the audience,
you'll recognize this as the Schrodinger equation.
Basically, this is a differential equation
for the wave that describes these atoms,
and it looks just like the Schrodinger equation.
We have a kinetic energy term, an external potential term,
and here's the surprise term, a self-interacting term,
a nonlinear term, which is, again,
is proportional to the scattering length,
this probability of the atoms scattering from each other,
which gives you the repulsive effect.
Usually, in sodium and rubidium-87,
it's repulsive and gives you the big blob.
In the early experiments of Randy Hewlett,
he works in lithium-7,
which is a negative attractive condensate
which tends to implode,
and I'll come back to this later.
So the effect of this interaction
is it makes the condensate bigger,
and it also gives it its sort of springiness
or gelatinous texture.
Here is the first experiment of condensates
that I'm going to show you today,
which is simply you put the condensates
in this magnetic bowl,
and then the experiment we do
is we take the magnets
and we increase the strength of the current
and then decrease it, increase it, decrease it.
We cause the bowl to oscillate like this,
and the condensate in turn oscillates,
and after we've got it jiggling,
we turn off the oscillation.
The bowl just sits there quiescently,
and the condensate continues to quiver,
and you can see a series of images
taken at one-millisecond intervals,
and you can see the condensate
sort of throbbing here in response
to being driven.
The combination of all of the atoms
being in a single wave,
and they're having this repulsive interaction,
the combination of its being
a single-wave condensate
and a sort of gelatinous texture
gives you, in the end, superfluidity,
and I'm going to talk to you a little bit
about the nature of superfluidity,
which of course has been seen
since the 1930s in liquid helium.
I think this is a lovely system
for studying superfluidity,
this miraculous property
where the viscosity of a fluid vanishes.
This is a wonderful system
in Bose-Einstein condensation to study it.
Let's see how we can come to understand
the origin of superfluidity.
Here's a few more equations,
but I brought a prompt,
so if the equations bore you,
bear with me here for a second.
We describe...
This is supposed to indicate sort of a strip.
You can think of it as a strip of fabric
with a twist which has been put into it.
You can think of the width of this strip
as being proportional to the wave function,
somehow proportional to the density, essentially,
or to the square root of the density,
and then the angle of the twist
as you come down here,
as you see as you come down,
the twist goes further around,
is proportional to this somewhat spooky concept
of the quantum phase,
and then the velocity of the atoms
is just proportional to the gradient of this phase.
If I take this,
imagine that this is my condensate
as I move down here,
and I gradually put a twist into it,
that corresponds to there being a flow
in my condensate coming down here,
and if I do something sort of sneaky,
which is I get the condensate moving
by putting a twist in its phase,
and I put it now into a loop,
I find I'm in an unusual situation.
I've got a loop here,
and in some sense I'm stuck with this loop.
Now imagine this loop corresponds now
to the flow of condensate going around.
Say I've got a pipe that looks like a donut,
a toroidal tube that the flow can go around in.
That corresponds to a loop of fabric
with a single twist in it,
and you can see as long as I keep this belt,
if I don't tear it,
I'm stuck with this twist.
I can make little changes in the belt,
here and here,
but I can't get the twist out of here,
and the fact that it's not convenient for me
to get this twist out of here
really is the microscopic explanation
for why it is that you have superfluidity.
No small changes I can do in this
because think of it as topology
can get rid of this twist.
This is why you can have a fluid
flowing around in this loop,
and even though you ought to have some friction,
perhaps the walls of this chamber
are a little bit rough,
the fluid still continues to flow around and around
unless I break it.
The only way I can stop it is by breaking it,
and that can cost a lot of energy.
It corresponds to making this,
at least at one point,
very infinitely thin,
suppressing the density all the way to zero,
and here's where the gelatinousness of it comes into.
Because the gas repels each other,
it has a pressure,
and it costs energy for me
to try and drive the energy to zero anywhere.
For the experts,
I mentioned to you the following
amusing fact about condensates,
and I'm going to come back to this again,
which is that,
now, here's my, again, my toroidal loop,
and imagine I have flow going around it.
It corresponds to the arrow.
This arrow is the phase,
the angle of the belt,
and you can see as I go once around here,
I've got a loop,
and if I constrain this arrow
to point in the same direction of this arrow,
because, in fact,
it's just the same spot on this winding,
I can't get rid of this winding,
unless,
suppose someone comes along and says,
no, you stupid arrows,
you're not constrained
to remain on the transparency,
free your mind,
and burst free
of this two-dimensional world you live in,
and tilt out of the plane,
like this.
I hope this is clear.
So, it is possible
to get rid of this winding
without doing anything discontinuous anywhere,
if somehow this loop,
if these arrows,
instead of just being constrained
to point somewhere on a circle,
if they could point anywhere on a sphere.
Now, you could do this
if you lived in a more complicated
order parameter space,
not to drown you in jargon,
but if instead of your order parameter being U1,
if it were SU2,
for the experts in the field,
and this can happen
if your condensate
is not made from individual monadic atoms,
but from atoms
with composite structure inside them.
And I'll just take this opportunity
to tell you about some of the tools
in our ultra-cold alkali toolkit,
making Bose condensation with alkali atoms.
There are many experimental tricks
available to us,
and Wolfgang will be telling you
about some of them,
and I'm going to tell you
about some of them today.
I think Wolfgang will tell you
perhaps about measurements
of the density,
and the momentum,
and the energy,
and velocity distributions.
You can hold them.
Carlos told you about
some of the simplest
of the magnetic traps
in which you can hold the atoms,
but in fact the bottles,
if you will,
in which you can hold these condensates
have a great deal of flexibility.
They can be essentially
one-dimensional bottles
or two-dimensional bottles.
They're being done now
in these micro-fabricated waveguides,
time-varying potentials.
One can do studies
in which one probes
the internal coherence
between the different states
or the external coherence
of the wave function.
Alkalis come in fermionic flavors,
so people now are making
degenerate Fermi Cs
using the same technology,
hopefully soon to see
something like BCS superfluidity,
except it's predicted
to occur perhaps
as high as one-half
of the Fermi temperature.
They have internal structure,
which is what I'm going to
really talk to you about today.
If you look inside the atom,
it has different parts,
and the fact that the atom
has different parts
very fundamentally changes
the nature of the wave function
that the atoms
describe the atoms.
Here's a picture
of a rubidium atom.
It's got a positive ionic core
and around it
a single valence electron
in a little cloud.
Both the ionic core of the atom
and its electron have a spin,
which can be up or down
and various other directions.
Very roughly speaking,
rubidium 87 has these
two different configurations
in which the electron
and the nucleus
point in the same direction
and in which they point
in opposite directions.
For convenience,
we describe them
with these nice intuitive numbers
F equal 2 and F equal 1.
Never mind.
It's just the labels
for these things.
We call it state 1 and state 2.
This is the hyperfine structure
of the atoms.
We do experiments
with the condensates,
typically using two
of the substates
of these things.
You can think of it
as the atom is having
a little spin,
which can point up
or it can point down.
If we apply a microwave to it,
we can coherently convert
the atoms from the down state
to the up state
and back again.
If we start
with all the atoms
in the 1 state,
if 100% of the atoms
are in the 1 state
and we turn on this microwave,
the atoms go down
to the other state,
back up.
These are Rabi oscillations,
which can go for many,
many, many cycles.
If we turn on the microwave
just for a little while,
it's called a pi over 2 pulse.
We put half of the atoms
in one state
and half of the atoms
in another state.
If we leave on the microwave
a little longer,
we can take all of the atoms
from the 1 state
and they go away
and the atoms all appear
in the 2 state.
That's called a pi pulse.
These two,
if we turn off the microwaves,
the atoms do not
spontaneously interconvert
between these two states.
They become really
distinguishable fluids.
Here's an experiment
in which we start
with all the atoms
in the 1 state,
which I've painted red here.
Here's the little bowl
in which we keep the atoms.
We turn on the microwaves
and give it a pi over 2 pulse.
Now the atoms are
half in the 1 state,
half in the 2 state.
The atoms interact
with each other
and separate out.
The two atoms sink,
if you will,
down to the bottom
of the trap
and the one atoms
float out to the edge
of the trap
and it makes
a little ball
if you've created
a spherically symmetric bowl.
In this image,
we're sensitive
only to the atoms
in the 1 state
and it appears
as a spherical shell
around the two atoms
which we're not looking at
in this picture.
We can use
these two different fluids
and our ability
to coherently convert
the atoms
from one state
into the other state
to engineer
the wave functions
of these condensates.
Here's my
engineering scheme here.
I'm the son
of an engineer
and so this is
very appealing to me.
You can see here
I have a crane
which can lift up
to 7 gigahertz
which is actually
the energy splicing
between the two states.
We start with
all of the atoms
in the 1 state
or in the 2 state
if we prefer
and in a spatially
selective way
we can take pieces
of this wave function
and move it up
if you like
into the other state
and as we move
from one state
into the other state
we can modify
the phase
and in this way
construct
in this upper state
a wave function
with whatever
amplitude and phase
we like
with limits.
So we can create
wave functions
in this way
and one of the
first things
that we did
using this technique
is to build
a vortex.
A vortex is
when you
pull the plug
from your bathtub
and you see
the little
whirling water
rushing down
into the drain.
This is a
classical vortex.
A quantum vortex
is the sort of
smallest
possible unit
of vorticity
that you can have
in which
the
corresponds to
exactly one unit
of rotation around
which is described
in quantum language
as having the phase
go from 0
to pi over 2
to pi
to 3 pi over 2
all the way back
to 2 pi.
One cycle of phase
as you go once
around the outside
and we do this
by starting all the atoms
in one state
and bit by bit
moving it to the upper state.
This part
we move to the upper state
with phase 0.
This part
we move to the upper state
with phase pi.
This part in the middle
we leave right where it is
because
in the end
you always have to have
a little hole
in the middle of your vortex
so the phase
doesn't do something
discontinuous
and we leave
these other atoms
as a plug
in our vortex.
Here you can see
a picture of a condensate
turning into a vortex.
Very faintly
you can see the atoms
here in the one state
which appear white
in this image
and we turn on the microwave
in this spatially selective way
and out of the mist
emerges this black ring
which is the vortex
or equivalently
we start with the atoms
in the two state
as a dark black splotch here
turn on the microwave
and out of the mist
emerges this white ring
which is now a vortex
in the one state.
Here's an image
of one of these vortices.
You can see here
here's these shadow pictures
that Carl has
presented to you before
and I've just shown this
in three dimensional here.
Here's a vortex
the atoms are in the one state.
When we took this picture
our camera is set up
so we don't see the atoms
in the two state.
We only see the atoms
in the one state.
If we suddenly apply
a pulse of microwave
we can take all the one atoms
and put them into the two state.
Now we look over here
we don't see the ring anymore.
Instead we see
a little ball in the middle.
Basically in this image
there's a ball of atoms
in the middle
which we can't see.
After we turn on the microwave
they appear
they're converted
into the state
that we can see.
The atoms in the ring
which we can see now
go into a ring
which we can't see.
Now we can see
the outside of the condensate
which is a ring
outside of the vortex
which is a ring
and the inside of the vortex
which is a plug like this.
It's possible
so far we're imaging
only the
call it the square magnitude
the square magnitude
of the wave function
but it's also possible
to image the phase
of this wave function.
Instead of applying
a pi pulse with a microwave
if we apply a pi over two pulse
so the atoms are
half in this state
and half in this state
we will see an interference
a quantum interference
between the ring
and between the plug.
And so we'll remove
this little piece of paper here
and you can see down here
in the shadow
the characteristic Pac-Man shape.
On this side we have
probably people don't play
Pac-Man anymore
but on this side
we play evaporative cooling.
I think that's the latest
the latest video game.
On this side
we have constructive interference
this part of the wave function
interferes with
this part of the wave function
and it's
and you get a lot of
a lot of wave function
or on the other side
we have destructive interference.
And on a pixel by pixel basis
you can combine
these three images
and extract
the cosine of the phase
between these two states
and you can see that
as you go
once around
once around this ring
the cosine goes from
one to zero
to minus one
back to zero
you see one cosine
cosinusoidal cycle.
So we can image the phase
of this wave function
as you go
of this vortex
if you go around.
The
going back to my
to my engineering analogy now
you can think of us
constructing this wave function
up here
we use this
but we can think of this original
sort of as a
this original state
sort of as a scaffold.
We can construct
this wave function here
then we no longer need
this scaffold
and we can shine light
which is resonant
with only these atoms
and not with the other atoms
blow them away
and now
the
there's no longer a plug here
and the vortex
shrinks down
down to a tiny little
pinhole here in the middle
which can't vanish
because as you know
the phase
wraps from zero
all the way around
to two pi
and right here in the middle
is a singularity
where the phase
can't be zero
pi
two pi
all at the same time
and so the density
just goes to zero there
as a tiny little
defect
if you like
in the quantum wave function.
Here's a picture
of that happening.
Here's the ring
and we gradually
blow the atoms away
and the ring gets
the plug
we blow the plug away
the plug gets smaller
and smaller
until finally
we can't see it at all
but we know it's there
the way we know
there's still
a little hole there
is we can
suddenly turn off
the magnetic trap
the repulsion of the gas
blows the cloud apart
and afterwards
we have a look
at what we see
and
let's get this right here
so here's the ring
it shrinks down
as we blow away
the stuff in the atom
we turn off
the magnetic trap
it blows up
to a much bigger ball
and right there
in the middle
we can see
this tiny little plug
which is where
the vortex is
and we've done
a number of studies
over time
we see that
this vortex
processes around
it drifts around
like a
like a hurricane
confined in the
Gulf of Mexico
or something like this
there's a
mode in which
the
sort of
a meteorological mode
in which the tornado
moves around
through this
material
there are other ways
to make vortices
and Wolfgang
is going to talk to you
about it a little bit
and
I show you this slide
only because it's
I think the single
most beautiful image
to come out of my laboratory
and I'm so proud of it
so this is
an enormous number
of vortices
somewhere in excess
of 100 vortices
all created
in a
in a particularly
large condensate
I love this picture
I'm sorry
it's not connected
to my talk
as a
a final example
of the
of the consequences
of internal structure
the fact that the atoms
are
are not
you know
Aristotelian
indivisible things
but
but things with structure
inside them
I'm going to talk to you
about
something called
Feshbach
resonances
this is
ever so slightly
more technical
than the rest of my talk
but
I will show you
one or two more
pretty pictures
at the end
I've talked about
the fact that
these
condensates
can be described
by a
wave function
which has a
a really
a single parameter in it
which is this
scattering length
the
proportional somehow
to the likeliness
of two atoms
wanting to collide
with each other
what determines
this scattering length
of an atom?
think of two atoms
as feeling a potential
at very close range
this potential is very large
and positive
they don't want to
bump into each other
at intermediate range
they're sort of
an attractive well
and then of course
when they're far away
they don't feel each other
at all
this potential supports
bound states
molecular bound states
and the scattering length
is entirely determined
by whether that
last bound state
of potential
is right there at the
right there at the continuum
or not
if it's a little bit below
the scattering length
is very very positive
if it's a little bit above
the scattering length
is very negative
if it's right at the
if the
if this bound state here
is right at the continuum
the scattering length
is infinite
and so on
so
if you somehow had
were able to adjust
the depth of the potential
between two atoms
so now by
I change
u not up and down
I adjust the depth
between the potential
I would have the ability
to tune the scattering length
from positive
through zero
down to negative infinity
positive infinity
down through zero
I could have it be
anywhere I want
you might think
that we are
okay but two atoms
they interact
they have some
strength of interaction
that's given to us by God
we're stuck
but it turns out
that there's some
wonderful tricks
using involving
the internal structure
of the atoms
which allow us
in some cases
to adjust
the scattering length
again
because the atoms
have this internal structure
if we start
with these atoms
again in this
in this anti-aligned state
with the nucleus
spinning up
and the electrons
spinning down
we combine them
into a molecule
we can combine them
into a molecule
in two different ways
in one case
the molecules
the nuclei
are both
putting up in the molecule
and the electrons down
in the other case
the nuclei are anti-aligned
and so are the electrons
there's two different molecules
that this atom
can bounce
then combine into
and still preserve
angular momentum
this gives us
a knob
a button
we can push
to adjust these atoms
these two
these two molecules
have very different energy
but the relative energy
between these two states
depends on the magnetic field
because the electrons
where all the
magnetic moment lives
have very different
projections onto the
onto the ambient
magnetic field
and I haven't really
drawn this to scale
it turns out
that by adjusting
the magnetic field
we can bring
a bounce state
of this atom
right to the continuum
of this molecule
and the atoms
are coming in
in this channel
I'm sorry to drown you here
in a little
molecular jargon
but what it means
is that
we have the ability
by changing
the ambient
magnetic field
we can tune
the scattering length
from negative
to infinitely negative
all the way up
to infinitely positive
down back through
zero again
so it can be positive
negative
anything we want
the possibility
of doing this
of atoms
and in particular
in condensates
was pointed out
thank you
five minutes
was pointed out
by Baldwin Verhaar
and therefore
we call these
Feshbach Resonances
okay so
here's a
these exist
in
basically
these exist
in all the alkali atoms
but they are
in only
some of the
alkali atoms
do these fields
exist at a
convenient
field which is also
good for magnetic
trapping
and experimentally
accessible
here's a
condensate
these experiments
I should say
are done in
Carl's lab
the ones I'm going
to show you
just here
but have been
also done
of course
in Wolfgang's
lab
and Dan
Heinzen
in Texas
here we have
a relatively
small condensate
with a relatively
small scattering
length
at a field
of say
162 Gauss
now we
change
the
field
in the middle
of the trap
just a few
Gauss
say we go up
only to
157 Gauss
now the
interaction between
the atoms is
the scattering
length is much
larger
we're on this
sort of
positive branch
where we come
from negative
through zero
heading up
towards positive
infinity
the condensate
is much
more self
repulsive
and it gets
bigger and bigger
and we can
gradually tune
this up
to places
where the
scattering length
now is
almost
inconceivably
large
and we get
these very
very large
condensates
very tenuous
as they
repel
each other
a trick
which can
happen is
we can
change this
magnetic field
very rapidly
on the scale
of the experiment
so we can
create ourselves
a large
condensate with a
positive scattering
length and then
very quickly
change from
159 Gauss
to say
171 Gauss
the scattering
length goes from
positive to
negative
and now the
condensate
instead of
blowing up
and being
repulsive
is self
attractive
it pulls
itself in
and as it
pulls itself in
the density
goes up
and it pulls
itself in
more
and the density
goes up
even higher
and it pulls
itself in
quite violently
and you see
an implosion
here's a
condensate with
10,000 atoms
in it
and this is
the time
since the
scattering length
goes from
zero to
negative
and this
cloud is
imploding
you don't
see it
getting very
much smaller
because this is
right at the
resolution limit
of our optics
but you do see
it basically
going away
because the
atoms are
falling in on
themselves and
sort of being
consumed
I can show you
the same data
with the scale
turned up
and you can
see that
as the
condensate
falls in on
itself
there's an
implosion
but there's a
corresponding
explosion
of very
very hot
atoms
I should tell you
these experiments
are done at
extraordinarily low
temperatures
perhaps only
3 nanokelvin
so when I
tell you
extraordinarily
hot
I mean
this gas
bursting out
here is
100 times
the ambient
temperature
or 200
nanokelvin
but think
of this
as the
explosive
halo
of the
supernova
done in
bosons
it's called
a boson
nova
so we see
it now
exploding
outwards
like this
this is
an interesting
topic
I'm trying
to just
in the
last minute
or two
to give you
a feeling
of things
which are
sort of
now at
the
cutting edge
of where
we are
going
of condensate
research
and I'll
give you
one more
example
in some
ways I
think it
may be
even more
interesting
rather than
going to
the
self
attractive
case
to go
to the
positive
case
in physics
one of
the most
serious
most
one of
the
enduring
mysteries
in
physics
is
liquids
solids
gases
this is
how we
all make
our
living
liquids
are
a great
mess
here's
a handy
dandy
definition
for
liquid
if
your
atoms
are
somehow
very
small
compared to
the spacing
between the
atoms
spacing
between the
atoms is
given by
some
distance
d
then
density
is one
over that
distance
cubed
so if
the
atoms
are very
small
this
interaction
parameter
here
n a
cubed
is going
to be
a very
small
number
perhaps
10 to
the
minus
5
this
is
a
gas
and
for
the
experts
in
the
audience
a
gas
and
a
condensate
is
very
nicely
described
by
a
many
body
wave
function
which
factorizes
if
we
turn
up
the
interactions
between
the
particles
are
comparable
to
the
spacing
between
the
particles
this
value
n a
cubed
now
becomes
around
one
it's
no longer
possible
to write
the
many
body
wave
function
like
this
it
becomes
a
very
difficult
problem
requiring
a
lot
of
computational
time
it's
very
difficult
to get
any
physical
intuition
about
it
becomes
a
liquid
and
for
the
people
who
aren't
interested
in
these
formulas
down
here
basically
what it
means
is
in
a
liquid
when
one
atom
wants
to
move
other
atoms
have
to
get
out
of
its
way
this
is
a
difficult
thing
to
understand
and
we
now have
a
tool
to
understand
this
we
can make
a
condensate
where
the
wave
function
is
known
very
precisely
we
can
suddenly
turn up
the
interactions
between
the
atoms
on a
time scale
which is
fast compared
to all
the other
physics in
the
problem
what that
means is
we can
create
ourselves
something
which ought
to be
a
liquid
but a
liquid
whose
initial
wave
function
is
exactly
known
because
it's
just
instantaneously
projected onto
this
gas
phase
wave
function
then
over time
we can
study how
the
liquid
correlations
develop
using
such
tools
as
rag
scattering
and so
on
some of
these
things
you may
hear about
from
Wolfgang
so I
think this
is one
of the
interesting
directions
of
Bose-Einstein
condensation
research
but I
should
emphasize
that there
are
30-35
groups
experimental
groups
now
working
in Bose-Einstein
condensation
there are
thousands
of theoretical
papers
exploring many
many different
aspects
so it's
not possible
to cover
in anything
like even
these three
lectures
which we're
giving you
now
and this
is one
angle
there are
many other
angles
to explore
at this
time
I think
it's an
exciting
time in
Bose-Einstein
condensation
research
and I
thank you so much
for your attention
