Faculty Colloquim: Dr. Daniel Elliot

Faculty Colloquim: Dr. Daniel Elliot

October 13, 2019 0 By Stanley Isaacs


Thank you, everybody, for joining us
this morning, I am Kathleen Howell. I am the associate dean in engineering. And I’m really delighted this
morning to introduce Dan Elliott on behalf of Dean Jamieson and
the Office of Academic Affairs. But let me start a little bit by talking
about how this series came about and what it is. This celebration of faculty
careers actually came out of the last College of
Engineering strategic plan. And one of the actions from that
plan was called the faculty of 2020. And it was focusing on professional
development at all stages of faculty careers. So and
in particular aligning of criteria for promotion along with
the hiring process and the evolving scope of the college
as we started to expand and so on. At that time there was a desire for
a post promotion review, particularly for full professors, that would feature
the accomplishments of the individual as well as provide an opportunity for them
to plan perhaps what was going forward. Or at least to interact directly with
the dean on their particular program and their accomplishments in
terms of the college. So full professors or at least seven years past promotion
are now participating in this event. They get to highlight their
achievements to their peers and although it’s also followed
by a session with the dean, there have been some unexpected
benefits associated with this program. And one of those is the opportunity
to share with our peers exactly what different members of your
individual units are accomplishing. And so that’s been a really big plus. The first pilot of this
program was in 2013, and we’re now into implementation phase. And Dan is the latest recipient, I would
also like to say words about him, then. Actually, I’ve known Dan for
a long time, and I’m really happy to see him go forth and
do this. He actually completed his PhD in
1981 at the University of Michigan. Before he joined the faculty here
at Purdue, he had a postdoc. Joint Institute for
Laboratory for Astrophysics and at the University of Munich. He’s been here at Purdue since 1984. He currently has a joint appointment
as a professor of electrical and computer engineering. And a professor of physics and astronomy. He has served ECE, I think,
fantastically throughout the years. Including as graduate coordinator,
the director of graduate recruiting and retention also for
the College of Engineering. I really want to say
congratulations to Dan and I’m looking forward to hearing more about,
I guess your look back and look ahead.>>That’s right, thank you,
we keep this interactive right? So I get bored if I sit for
an hour in an audience, so please ask questions and interrupt. And maybe you’ll give me some new
research ideas too as we go on. Actually, my title was wrong, our first
paper was actually published in 1990. So, it’s been 26 years now, so
you learn something as you look back and you forget these dates. How could it have been so long? So, we have been working in a field that
we were actually able to first develop. It’s an idea of causing, intentionally,
interferences between different transitions and atomic systems, but it’s
been extended to other systems as well. In order to affect the outcome and
we’ve moved on, more recently as a means of using it for
detection of very, very, very weak signals in atomic
systems that can be useful as well. I wanted to list some of the other
things that we worked in here as well. So the coherent interfering interactions
will be the topic of today’s talk. But other things we’ve done some work
in over the years have been ultra-cold molecules. That’s been a collaborative program with
Yong Chen, where we are creating and cooling and manipulating lithium or
vidium molecules for various purposes. Some early work, which Gerhard I saw, we worked on together a long time ago,
was to artificially create fluctuations, impose fluctuations onto a laser
in order to create bandwidth. And to understand how those correlations
affected various non-linear interactions. We’ve done a lot of work in using angular
distributions from photo-electrons in atomic systems again,
in order to learn, probe various things. And some of the best work, I think, that never got any future,
You look back on these things, and you check the citations and
it, never went anywhere, but it’s some of the best work that I
think that came out of the laboratory. We were able to do four way mixing in
atomic systems, a two level system, really well controlled system and
actually showed that some of the fundamental theory that
developed for that process was right. But it was ten years after people had
gone on to the application stage and never took off. Anyways, that’s one thing that you
learn as you look back on some things that you’ve done over the years. Of course, none of this work goes anywhere
without good students working on it. And so just the students who have
been working on the work that I’ll be talking about
today are listed here. Starting with Chen Ce and Yin Yion
back in the early 90s, all the way up through some current students
who are continuing with this work. Okay now the idea for
what I’ll be talking about today is that optical interactions in atomic
systems can be made to interfere. The idea for this first came to me
from some work that had been done in the late 70s and early 80s. Most of you probably don’t remember those
years, but they were good years, too. It was the case of
the disappearing resonances, and it sounds like a super sleuth story. People were looking up all sorts of
nonlinear interactions in atomic gases. And noticed that at places, at frequencies where they expected to see strong
ionization signals, things disappeared. And so, here I took an excerpt from
a paper by John Miller and Bob Compton. In this work,
they were working on atomic xenon, they tuned our laser to
a two photon resonance. And they looked at the ionization signal,
And here they’ve plotted the various ionization signals as
they’re increasing the density of the gas. And you notice that the signal
tends to disappear. And at higher densities,
it went away even further. They also monitored the amount of
third harmonic generated in that gas. Third harmonic is a coherent process, it’s a non-linear process that
can occur in these gases. And it simply means that
you’re generating light at three times the frequency
of the input light. And they, obviously it’s,
as you increase the density you have more atoms to interact with and so
the signal go up and up and up. So the correlation eventually was noted
that as the third harmonic signal went up, the ionization signal went away. In many of these cases,
this was not the only example of it, but there were quite a few cases like this. And the explanation being that as
the third harmonic is generated, now you have two processes going on
inside this gas at the same time. There’s an interference which can
take place between the transition from the initial state to the final state. And that was shown by Jim Wynne. About that same time that these two
interfering interactions would gravitate towards a destructive interference
case and the signal would go away. Well, when we saw some of
those papers I thought well, wouldn’t it be great if
we could separate them. This was all going on inside of a single
gas cell, complicated situation. I wanted to pull those
two processes apart. And that was the beginning
of the idea then for these coherent, interfering interactions. So we chose mercury, which was one of the systems in
which a lot of this work took place. There’s a strong transition in
the visible to three photon transition. So you can simultaneously absorb
three photons from the laser field. And the idea for the interference now
should be that the total transition rate, which I represent by a W here depends not
just on the amplitude for the process, but on the sum of the amplitudes and
the stack quantity, which is squared. I have to understand something about
the amplitude for the processes. For a one-photon process this is
proportional to the electric field. Usually we don’t bother
to write down the phase of the electric field because
it’s not usually relevant. Three photon process, well, you’re absorbing three photons
from the field at the same time. It goes as the cube of that laser field. And again, we don’t usually bother
to write down the phase, but I’ll do it here because it
does become important here. I’m gonna add these two amplitudes in order to understand
the total net transition rate. And now the phase becomes important, because I can change the phase
between those two processes. I’m gonna back up a little bit and
tell you something you all know. And that is just
an interference in optics. Here’s a classic experiment of
Young’s double slit experiment. This is in all the introductory
physics text, you have a source, you have a screen with
a couple slits in it. You put a screen over here where you
observe the intensity of the light. Works better if you use a laser, but
it It can work with a light bulb, too, if the conditions are right. What you see on this screen, then, would
be a series of dark and bright, dark and bright fringes. Because the light from the source
to the screen can take two paths. And whether we get a bright fringe or a
dark fringe depends upon whether those two paths add in phase with each other or
out of phase with each other. If they add in phase, we get constructive
interference and we get bright fringe. If they add out of phase we get
destructive interference and it’s dark. A well known process
since the day of Newton. Young, not Newton. [INAUDIBLE] [LAUGH]
>>I’ve lost my history now. When was Young around? Was it before Newton?>>[INAUDIBLE]
>>Yeah, okay, younger. I have trouble with fall forward and
leap backwards too.>>[LAUGH]
>>Okay, so this I’m giving you as an analogy, because
it’s a well known process in optics. But I want to stress,
the interference that we observe and that we exploit is not
an optical interference. We are looking at interference
between atomic transitions, not light. And here was the first demonstration then. This was work done by. We took a laser which was at tune to the
free-flow tone resonance in the mercury. We focus it into the first cell,
this is where we generate harmonic light. And so we create the third harmonic
of that fundamental frequency. So coming out of the first cell
we have two separate frequencies. We collected the light
on a spherical mirror, we focused that light with a second
spherical mirror into another cell. And that’s where we generate
now an ionization signal on the resonance in mercury between
the 6s state and the 6p state. And as we change the phase between
the omega and the three omega light, we’re able to modulate now the net
ionization rate within that cell. How do you change the face? In this case, we had this cell in the
center which I’m calling the delay cell. And we simply introduced
some argon gasses. We changed the density of the gas. It has a different refractive
index in the visible and the UV. And so we can change the phase between those two
laser fields and modulate the signal. If you look at the numbers on
the ionization signal there, that’s not real deep, is modulation, and
there are a couple of things going on. One is that as a focused beam
goes through its focal region, there’s actually a phase
shift of that beam. It’s a sort of diffraction effect. I didn’t learn until many,
many years later, it’s called The Gouy Phase,
which I’m probably not even saying right. But if you’ve done any work
with focused Gaussian beams, you’ll recognize this
inverse tangent of z over z not term where z not is the length
of the interaction region. Okay well we have two beams going
through the, through the cell. They are both focused roughly the same. But one of the three flow interaction
depends on the cube of that laser field and so it goes through 3 pi phase shift
as you go through the focal region. So we set up the same cell but with a different electrode
configuration in there. It’s now a series of eight
different electrodes. And resolution isn’t real great here,
but this is picking up the signal, so as you look at the eight
different channels across there, each channel is sensitive to
the electrons generated closest to it. And so, we’re seeing a variation in
the phase of the two beams there. You see sinusoidal modulation,
each one of those signals. But if you look carefully you can see the
different signals, the different channels are shifting a little bit to the left
as we go from one signal to the next. And we can plot that out. We can do it for different
tightness of focusing across there. And we can see the modulation
shifting due to this phase variation of
the focus Gaussian beam. I had a friend from graduate school, who, we kept in touch as we were
going to various locations. As I followed him for a while,
but never overlapped again. Arley Smith,
who said to me one day, you’re doing lots of work with photoelectron
angular distributions, you’re doing lots of work with coherent interferences,
why don’t you put the two together? And see if you can’t use
the interference as a way of controlling the angular distribution of
the electrons which you’re generating. So these are some simulations we did,
this is not a measurement yet. But, we just considered
something like ionization, photoionization of atomic rubidium. Rubidium has a nice hydrogen-like character to it in that it’s
a ground state, it’s an S state. If you ionize it with UV light Such
that is a single photon process you’ll create mostly a P wave going
away from that system. But if you ionize with a two proton
process then you create mostly an S and a D wave, mostly D wave. Well, there’s a different parity
to a D wave and a P wave. That is, if you invert a P wave
on itself about the origin, then you’ll have an odd symmetry. It inverts itself,
the angular distribution, the wave function changes sign. But for a D-wave or an S-wave, those
are even wave functions and they don’t. So now if you can invert. Well, the net result is that we get
constructive interference in one direction and destructive interference
in another direction. You can control the direction of
electron emission in this system. So here is a simulation now where
we just look at equal amplitude waves going away and we change the
relative phase of those different waves, again by controlling the optical phase. And you can see that in the first
figure we have mostly ejection upwards. And by the time we change a phase by pi, we get ejection of
the electrons downwards. How do you do that in a laboratory? Pulsed by a laser,
pumped by a harmonic of a laser again. So it generates the light
which we’re using for the two photon ionization process. This is at around 580 nanometers,
which is a nice yellow color. That’s why I chose yellow for my arrows. We separate the beam into a couple parts. We have a harmonic crystal here which
generates the second harmonics. So just like the third harmonic
I talked about before, this takes 580 nanometer light
coming in and it converts it into, whatever half of 500, 290. We recombine the beams. We again need to be able to shift the
relative phase between those two beams. And so, we do that in a delay cell here. And then inside of our vacuum chamber,
we have a beam of rubidium atoms And we positioned four detectors around it. This was all with
horizontally polarized light. So we positioned two detectors
which were in a horizontal plane, one which was perpendicular and
one which was at 45 between those two. And this was the single that we detect
on each of those four detectors as we change the relative phase
between the two optical signals. So you see, good, strong modulation
of the electron signal in Channel 1. Channel 4 you see good,
strong modulation, and notice that this is perfectly out of
phase with the modulation of this signal. So if electrons are going to the left,
they’re not going to the right. And if they’re going to the right,
they’re not going to the left. And then we see decreased modulation
in the other directions, as expected. So that was the first demonstration of
controlled of the electron distributions. So, other people where
watching some of our work. I want to mention them as well. But shortly after that
Robert Gordon at Illinois of Chicago did work related
to our mercury work. He was able to control the transition
rate in a couple of molecular systems. He’s a physical chemist so
he does his work in molecules. He also followed that then with
a control of dissociation products. So, one of the goals of this work, which really hasn’t been fulfilled,
met very well. Has been, can you use this kind of
coherent control in order to affect the branching ratio into
different kinds of products? One success in that was this story. It’s not quite to the point of calling
it controlled chemistry just yet. But he was dissociating
a small molecule HI. And he could control it by either ionizing
it or by dissociating the molecule. And so he showed some control then on
varying the phase that would change the relative amplitudes to those
two different product states. And he could show some
control out of that. Lou D’Amaro,
who was at the time at Brookhaven, did some work on HD plus and
was dissociating that again. Some work that followed on from the atomic
angular distribution work we did is, it was carried on in semiconductor
materials where you could control the direction of photocurrents
in the semiconductor material. I believe this was done
in gallium arsenide. But several notable works across there. Again the idea being of being able to
control the outcome of the experiment. A nice good proposal type
demonstration by Nokajima showing you could control the, or
take advantage of the phase differences. So we were continuing to work in
the angular distribution work. And the earlier work I showed you
had four discrete channels to it. We advanced then to
continuous distribution. In that case we’re using now, as our
detectors, microchannel plate detectors. These are glass-like
structures that are porous. They have a coating on them
which nobody knows what it is. But as the electrons hit it, they’re gonna
rattle along the channels of this and get multiplied. So you get one electron in gives
you a million electrons out. It hits a phosphorous screen and
you can look down and see where those electrons were. So in this case we’re coming in,
it’s still rubidium. Coming in with our omega-2-omega
combination of light. It intersects the rubidium. We’re doing this in a weak electric field,
so that the electrons as they’re coming out get swept up towards the microchannel
plate detector, and we can create an image then, of the electrons as they
are incident on that phosphor screen. And these are some of the images that
we were able to collect for that. On the left side I have just
a P wave lying on its side. So, you all remember from your
freshman chemistry classes P waves. This picture of a peanut that’s
in a horizontal direction. And the electrons are coming out mostly
this way and symmetrically out this way. And so, they’ll get swept up and
they’ll hit the screen. And that’s what you see across here is
the projection of the electrons onto that micro channel plate detector. This is the image without with the D wave. In this case that laser
was vertically polarized. And so a D wave,
think of a peanut with a skirt on it. So there’s this ring around the center. And so you see, the electrons that were
ejected upwards keep going upwards. The electrons which were ejected down,
go down get turned around, and they hit in pretty much the same spot. But then there’s this ring around
the waist which is going to expand up and give you this image across here. So this is a D Wave, lying vertically. When we put both colors in at the same
time, then we get the interference. And you can see constructive interference
on the right, and destructive on the left. And then we change the phase by pi, and we
get constructive interference on the left, and destructive on the right. We try to limit with molecules. So this was worked with Ed Grant
in our chemistry department. He is now at British Columbia,
but his favorite molecule was NO. And that was similar work in
which we’re using a pulsed laser. First to create an initial state for
us, but then we do omega two omega type
interference coming out of there. And looked at the angle of distribution,
well, looked at the electrons coming
often in this particular direction. And again, we can do modulation
of the ionization signal in one direction as we change that phase. Now we did do a little bit,
trying to control ionization. This is still not chemistry. But in atomic barium, you can ionize
the barium, taking off an electron. You can leave the barium
in its ground state. The electron carries off all
the rest of the kinetic energy. And in that case, I’m gonna call those
fast electrons because the barium’s left in its lowest state and the electrons
have all the rest of the energy. But the barium ion also has a fairly
low-lying excited state as well. And so if we leave the barium ion in that
state, then the electron has less energy to carry off and
I’ll call those the slow electrons. And so we did ionization of barium. We modulate the relative phase
between those laser fields, we see modulation of the fast
electrons and the slow electrons. And you do see a phase difference. This was, of course, the best case. You see a phase difference between the
modulation of the fast and slow electrons. And that’s one of
the signatures you need for doing control over which
product’s state you’re creating. Because if you have constructive
interference leading to one product at the same phase as you have destructive
interference for the other product. That’s the best case for
being able to control the outcome. Okay, so the idea of control then,
at least in gases, can work. There are difficulties as
you go to condensed phases. Because you’re depending upon these two
laser fields to have the same relative amplitude. The same relative phase as you’re going
through some condensed phase system. That becomes difficult and so after while, some interests started
to move in different directions. Our interests move towards using
this kind of interference to make detections of very, very,
very weak optical interactions. So I’m gonna go back to
analogies with optical signals. Again, well, in an optical signal, if you have a very weak signal
you’re trying to detect. Here I’m picturing you just have a source,
you have a detector, the signal is weak and you have some small
current which is generated across there. If that signal was weak, that could
easily be less than the shot noise, less than the other sources of noise. It could be hard to detect. And so, in the optical regime, and
in the radio frequency regime, people have developed homodyne and heterodyne detection techniques in order
to dig optical signals out of the noise. And how that works is that you
still have the weak signal, same detector across here. You combine your signal with a much
stronger local oscillator beam. And now what you detect here is
the coherent sum of these two. So if these two signals are at the same or
similar frequency, now what I will detect is going to be the
sum of those two field amplitudes squared. My photo current would
be proportional to that. If this is very strong, then I’ll
get this signal plus a cross term, plus the weak filled square. But that’s so weak that I
didn’t bother to write it down. If you can change the relative phase
between these two while you have this DC term, which is just constant. And you have a term now which is at
a beat frequency or area of control. You can control the phase between them and
make it constructive or destructive interference. So, that’s a signal detection
technique that has been developed in the optical regime. That’s gonna be similar to what
we’re doing in our atomic system. Except, the same caution,
we aren’t doing optical interference. This is quantum interference
between different transitions, which are driven by
the various laser fields. Okay, the direction we’re heading in for
this and this is a current program now, is measuring very, very weak transitions. Which are induced by the weakened
reaction in the atomic system. Almost everything we need to
understand about atomic systems, we can neglect the weakened reaction. We have electric forces of the attraction
or repulsion between the charged bodies. We have magnetic forces that show
interaction between this spin and that spin or that [INAUDIBLE] momentum. Almost everything is very nicely described
in terms of electric and magnetic fields. One of those is that you can never
induce a transition from an s state to an s state. Because that’s gonna violate parity. That’s within electric dipole transition,
I should caution. So that’s a forbidden transition by
electric dipole selection rules. The weak force between different
elementary particles though is different. It doesn’t obey the same parity and in fact what it will do is the s
state is not purely an s state. It’s going to, through this weak
interaction with other states, it’s going to mix in
a little bit of p wave. This s state will not be a pure s state. It’s gonna mix with a little bit of p. Likewise here, it’s gonna mix
with the different parity state. So now you can imagine that electric
dipole transition between this s state and the p part up here, or
between this p part and that s state. It now allows a very,
very weak transition. And if we can detect that transition, that can give us a way of doing high
energy physics and atomic laboratory. Why do we want to do that? Well, one motivation might be
the search for dark matter. That’s one of the pressing needs within
the physics community these days. We know there’s a lot more mass out there
in the universe than we can account for. What is it? We have no idea. So theorists are able to
hypothesize various scenarios that might explain what these are. And from those calculate
what might be observable. Okay, here is a plot of a lot of data. This is years and years worth of data of
a what’s known as a weak mixing angle which comes from a measurement
of the weak charge. These measurements are high energy
electron electron scattering. There are other proton
proton scattering events. And way down in here at the lowest energy, this is plotted as a function of
the collision energy between particles. At the lowest energy is the best available
atomic parity violation measurement. In the different color bands across here,
Marciano has plotted what he would expect for this weak
mixing angle as a function of energy. If there is a candidate for
the dark boson, which is at 50 MAB and the green at 100 MAB and
the blue are at 200 MAB and the red. These measurements are in progress
trying to improve their error bars. They’re in the range where
they could make a good yes or no decision about some of these models. But you can see down here in
the low energies range where atomic effects are greatest. That if we can reduce the error bars
in the best current measurement, that we could make some inroads
into this kind of discrimination. Okay, there are also some other effects
which can be important which are due to the weak force between
the nucleons within the nucleus. And they can be found by looking at what
are known as spin-dependent effects. If we can see a difference
between different hyper fine components of the ground state. I won’t go into a lot of detail there, but
we are actively pursuing that as well. The best current measurement of the again, does see a difference between
these hyper fine transitions. And that’s a much bigger difference than
can be explained by any current model. That’s been a standing problem since its first report in 1997. There are efforts in many other
laboratories to measure operating on, conserving interactions,
weak force interactions in atomic systems. The measurements which are complete,
well the best measurement is cesium. That came out of the laboratory
in Boulder by Carl Weiman. He got the Nobel prize but for other work. Thallium, ytterbium, lead, and
bismuth have all been completed. Ytterbium is notable
because its amplitude for the effect is 100 times
larger than in cesium. That’s still right about an 8% level. Active program in francium. Francium has the disadvantage
that it’s radioactive. So you have to do it at the accelerated
laboratory where it’s created. And you have, well,
radioactive stuff’s not so nice, either. That’s a collaborative
program between Maryland, British Columbia and Triumph,
people are looking at various ions. And there’s even some work
in molecules for this. Any time you’re trying to measure
something which is very, very, very, very, very weak, you have to worry about
things which aren’t quite so weak. And how you’re going to discriminate
the signal you’re looking for from those other signals. So it’s good to keep an eye on what else
is available active on this transition. There is a star conduced transition. If you apply a DC electric field to your
atoms, that’s gonna mix states as well and that’s going to allow at some level
a measure of the signal which, well, one that could interfere, but
two you can take advantage of it. We’ll do both. There’s a magnetic dipole transition in units to keep track of
these different levels. A strong transition in atomic system
typically has a magnitude of E A-not, where E is just the electric charge and
A-not is the bore radius of the system. So that’s a typical transition amplitude
for a electric dipole transition. There is a magnetic dipole transition
here, which is six orders of magnitude weaker than a typical strong
electric dipole transition. We’ve actually measured that as a warm
up and I’ll show some data on that. There is an electric
quadripole transition here. It’s weaker than the magnetic dipole. Fortunately, on this
transition it’s forbidden. So it doesn’t show up at all. And then our weak force induced
transition which we’re looking for has an amplitude of about 10
to -11 on this same scale. So it’s a challenge. I’m not gonna say it’s
an easy measurement to make. All these measurements do
depend upon having good theory. Because after you measure
what the amplitude is, how do you convert that into
the charge of the weak force? And so, we depend upon having
good theory to do that. That’s one of the strengths of cesium,
actually. Cesium is rather hydrogen-like
in its structure. It’s an alkaline metal, so it’s has this nice metric S wave
function as the initial state. And the efforts of
the theories are critical. There is some discrepancy in
the two best theories right now. That needs to be resolved and
Marianna Safronova at Delaware is, or will soon work on that. We often get asked the question,
well, Caesium’s been done, and it’s the best current value. So why go back and do Caesium? Why don’t you go do
something which is new. And we have some good answers for
that question. One is that we do depend on very, very good theoretical modeling
of the atomic system. And Caesium, not as good a hydrogen, maybe
not as good as helium, so maybe it’s in third place now in the best studied atomic
system in which you can actually depend upon the theoretical models in order
to understand the magnitude of effect. There are also very nice experimental
techniques that we can apply in any of the alkali metals that are not
available in other atomic systems. One of them is optical pumping. We can use lasers in order to shift the complicated ground state
structure in to a single, initial state. Detection efficiency. If you’re just gonna put
a detector beside it and look at fluorescence coming off,
your detection efficiency’s gonna hurt. Maybe less than 1% detection efficiency
just due to the availability of good detectors and
the solid angle of detection across there. We’re gonna use an efficient
detection system based upon optical cycling
of the transition. And also in the measurements
that we’re describing here, we’re doing all continuous
way of measurement. And so we get nice accumulation
of data over all runs. In the first part of what
I talked about today.>From our initial measurements through
the barium work that I talked about. Those were all were done with
post lasers which 10 nanosecond duration pulses and
they eject electrons in that time. This is all done now with stabilized CW
lasers and detectable signals with them. But continuous. Okay, I think I’ll skip these. I do wanted to emphasize
a stability though in, cuz we’re competing with the measurement. Okay, so a measurement that we have
completed is a two photon transmission on the same transition we’re talking
about for the weak interaction. It’s from this ground state,
success state, to the excited state. We’re using this transition at 852
nanometers for measuring the signal. And we’re just trying to see well
how much have we transposed? How much have we transitioned? So again, the transition
amplitude that I’m looking at is the sum of all the amplitudes that
are processed, that are possible. We have the two photon process
which was never used before, and that’s driven by a laser in
the infrared at 1,079 nanometers. We do have the magnetic dipole,
that’s what we’re trying to measure. And this dark induced transition, that’s
gonna be our standard, our reference for these measurements. The total transition rate depends upon
the square of those just like with the homo-optical detection I showed. We have that term squared which is huge. And then the cross term
between these two guys and the two photon rate, which is small,
but we can control it. We can change the relative phase
between these two guys and this one, and from that relative phase,
make them interfere constructively or destructively, see an enhancement or
a diminuation of the signal. And from that,
know the amplitude of the cross term. The cross term gives us the magnitude
that we’re looking for. And I put this up here just because we are combining the magnetic dipole
transition with a star pole transition. It’s actually the quad rider sum of those
which is important that is actually to our advantage as well. So, here’s some data on that
magnetic dipole transition. We are driving the transition
from the 6S to the 7S. We have the strong two photon transition. We have the weaker magnetic dipole and
star induced transition. And we change the phase. And here I’ve plotted against
the phase across this axis. You can see there actually
the modulation of the signal. This is the interference. This is on top of,
this is just the modulating part. This is on top of the DC part, which is
about ten to the fourth times larger. The key though is, here we do it with
a zero electric field applied to it. No starting effect, so that’s just the
interference between the magnetic dipole transition and the two photon transition,
then we turn on the electric field. Now we have, in addition,
the Stark-Induced transition. You can see a stronger
modulation amplitude, okay? So it’s a different scale by a factor
of two, plus the amplitude it looks. Up here it’s a little bit bigger. That’s the primary measurement for this. When we plot the amplitude of that process
versus the strength of the DC field, we get this hyperbolic function. We can fit that and determine from that
the ratio of the magnetic dipole amplitude to the stark induced transition. That’s our calibration standard. We can also get the sign of that. I’m gonna skip that, but the sign of the
magnetic dipole transition is important. We repeat the measurement lots of
times under different conditions, different initial states,
different hyperfine transitions, and the average value now is about
0.4% accuracy to the measurement. That is in agreement with other
measurements that have been carried out. The theory isn’t so good on this
transition, they’re off by about, I think 15% or so. So, theorists have some
work to do on this one. Okay, so that was the magnetic
dipole transition, and that’s six orders of magnitude weaker than
a typical electric dipole transition. And now we’re hoping to apply that same
technique to measure the weak force in the atomic system. That’s five more or
is a magnitude well greater than four. I don’t wanna be too pessimistic, weaker. How do you do that? Well, first we have to change
the orientation fields. That’s a rather trivial thing. But we use the polarization of the laser, we use the direction of
the applied electric and magnetic fields to our advantage in order
to pick out the transitions that we want. That’s a good start, but
that’s not going to be enough. We’re going to use an optical
build-up cavity, so we’re going to put our light that drives
the transition into a resonant cavity. We can get finesse of 10,000 or so. We can enhance the amplitude of
the field which drives that transition. We’re gonna use counter
propagating laser fields. That’s going to effect that the magnetic
dipole amplitude that we just measured will cancel itself. The beam going through in one direction. These are selection rules
which are important to us but maybe not important to you. Different directions
we can deplete that or diminish that magnetic
dipole contribution and the critical part here will be to really
really control the static electric fields. This is to the point where often times as you have structures
within a vacuum system. Nice polished copper surface or something. Over time its going to build up vacuum
oils its going to build up cesium atoms which are floating around and they
developed what are known as patch effects due to domains within
the copper structure. Those effects can effect these kinds of
measurements because the electric field generated by those effects can be larger
than the DC electric field that you apply. So those are things that
we have to deal with. We do have a parallel effort. I made the point that we’re trying
to measure hyperflying effects and the effect of the nuclear
effects going on. So a parallel measurement now is
to carry out a similar measurement on a transition between the two
ground state atoms of this caesium. This is the 9.2 gigahertz signal
that atomic clocks are based upon. So your global positioning system,
your cellphones, your google maps don’t work unless
you have these good atomic. This is one of the transitions used for
that. Will drive the transition
again with the interferences, in this case the two photon transition
is going to be by a Raman type process coupling the two hyperfine levels and
we have now our magnetic dipole and our parity nonconserving interaction and our
store conduced interaction on that system. There are similarities
between this measurement and the measurement I’ve already described. There are also some pretty fundamental
differences as well, but the key here is that this measurement is going to be due
only to the nuclear spin dependent terms. And so it’s a better measurement now for
picking out the, what’s known as the annopal
moment of the nucleus. We have some experimental tricks
that we’re gonna play with and I think I’ll spare you many of those,
but we are again, borrowing from many of the same things
that we did in the other system. We’ve done some modeling
of these laser fields. We are developing a parallel
plate structure for transmitting the RF waves that drive this
and being able to apply a DC electric fuel that’s perpendicular
to the RF polarization. That becomes rather tricky but we think
we have a good scheme that can do that. We’re fabricating the boards now, and we’ve done some modeling
of the modes of the cavity. We’re creating an RF cavity that we
can use to control the RF field. This is for the 9.2 gigahertz field. We’ll work at the anti-node of
the electric field which is the node of the magnetic field and hopefully reduce
the magnetic field effects due to that. So, over the years we’ve been able
to use the interference first for just controlling the transition amplitude. We can turn it on, turn it off,
is probably too strong a word, but we can modulate the probability for
the interference to take place. We’ve used it for control in the direction
in which electrons are ejected, so we can make them go right or we can make them go for your way, left or
right just by controlling the relative phase between those laser fields
that are driving the transition. We’ve done a little bit of control on
the population of various channels. Not to the point where you
can call that chemistry but, that’s one of the goals of the work. Our more recent work though, has been to, instead of using it to
control the products of the outcome. We’re using it to detect
a very weak amplitude as contrasted with a stronger transition. So we are looking at a very
small change in amplitude as we change the phase
of the relative field. So thanks for your attention and
thanks for coming today.>>[APPLAUSE]
>>It was good.>>Any questions?>>[INAUDIBLE]
>>You’ve all been very well behaved.>>Just said a couple of months ago
I was just in Europe [INAUDIBLE]>>They put lots of money to project actually to observe these
[INAUDIBLE] manifestation. In that particular if you
look at polarisation in fact, because we know they have
very strong sensitivity. So, basically priority facts, but
normally left handed from the right.>>Yes.
>>[INAUDIBLE] balance each other.>>Right.
>>But because of PNC, that’s what would be salutation. So [INAUDIBLE] big multi
million dollar effort that now.>>I’ve seen some of that work. Thee’s also been some work [INAUDIBLE].>>We know that it’s exceptionally
sensitive [INAUDIBLE].>>Okay that’s interesting thank you. Andy?>>And so we’ve talked a lot about
playing with the laser phase and observing the effects of interference. Transitions themselves presumably
can have some phase if they have.>>What can have?>>The transitions themselves. You know, particularly if
you were to tune the laser.>>Yeah.>>You know. Use this to make measurements of or worry about what the phase added or
the actual by matter.>>So balance states will
always have a real quantity. So it contains a sign, but
there won’t be a continuous phase there. Scattering states,
as we access when we do ionization will in fact have a phase
to that continuum channel. And yes we’ve measured that. We have some improvements that we think
we could make if properly motivated. But we have done that kind of
a measurement in the remedial work.>>I’ve got one more. So this last one, try to measure
the extremely weak transition.>>Yes.>>What do you characterize the time scale
for being able to hopefully build up the sensitivity, and
having control to make that measurement? In years or in terms of PhD students.>>[LAUGH] [INAUDIBLE]
>>I notice that George is watching attentively to what
my answer is going to be.>>[INAUDIBLE]
>>[LAUGH] There are a lot of systematics that need to be dealt with. You have to really understand
the stray fields within the system. You have to understand how
you’re controlling things. George is working diligently on
decreasing the bandwidth of the lasers so that it’s very well controlled as well and
stabilized. I’m gonna say that we will have
the work completed before I retire. And in terms of number of students,
if you. Is that fair, George? You’re good with that one?>>[INAUDIBLE] thinking
of possible applications of all this fascinating stuff we did. So one thing you mentioned,
that the control chemical reactions, like control in face. That’s eligible for inactivity,
even though you’d be deployed. So there are a few of the newer things
which people allow very much of like production, hydrogen production. And those should be very sensitive
to phase if you have more then one optical channel which is sufficient. And take into account that all it’s
related to energy production, storage, conversion. I think that could be very interesting. To my knowledge nobody looked at
the phase interference effects.>>Not that I’m aware of.>>These are very,
very important processes.>>Okay, thank you.