Nobody wants to do this. Come on, somebody be brave. Tell me. I've got two theorists here and an experimentalist.

No one knows why. I mean, we know

Hello everybody and welcome to Early Morning Coffee at CERN. My name is Steven Goldfarb.

I'm Claudia Cornella. My name is Simon Kuberski. And I'm Fred Gray. And we have an amazing show for you

today. It's something that is a result that really impressed me. We just had a presentation made last week I think it

was on the these results and it might not sound familiar to you as a physics

topic. It's from the muon g-2 experiment at Fermilab. But the topic

is really really nice. And I think the thing that caught my eye was the degree

of precision, the amount of work that went into getting this measurement, a measurement of the magnetic moment of

the muon, which is weird. So I want to start out with that. I want to try to understand

first of all to explain what is a magnetic moment.

Well, when I think of magnetic moments, I actually go back and think about bar magnets, the kind that everybody played

with as a kid. And you know that some of them are stronger than others. And so the strength of a bar magnet reflects

its magnetic moment. And you know, why is that magnet magnetized? It's

magnetized because it's built up of a lot of subatomic particles. and the

spins of the electrons inside all add up to the magnetic moment of the whole

bar magnet. Okay, there there's a lot in that. Okay, first of all, a muon is

just a simple particle, right? It's an elementary particle. It's like an electron but more massive, heavier,

right? There's electrons, muons, taus. Those are three leptons. I don't know Claudia maybe you can explain to us

what why would a muon have a magnetic moment? Because muons have spin

and if you want to some extent to spin is like a rotation but it's not a

rotation in a physical space it's really a quantum property of the particle

I don't know if that helps well I think it does. But let me

first of all we should acknowledge that this is the international year of quantum science science and technology which is in the

spirit of of our previous podcast. Of course everything we do here at CERN

basically or at Fermilab has to do with quantum mechanics, quantum field theory. But that's a strange thing, though.

You have to admit that's a strange thing. So an elementary particle is something which has no structure to it. You you can't really think of it as

spinning in in space, but it has some property that that's spin. I mean, why?

Why? And how do we figure that out?

Nobody wants to do this. Come on, somebody be brave. Tell me. I got two theorists here and an experimentalist.

No one knows why. I mean, we know that it's the case, but no one knows why they have spin. I mean we just

found out by the way they interact that there has to be this property that is like a tiny magnetic I mean like

a rotation but there can't be right and this is what we have to live with I mean sometimes it's really just

difficult and hard to understand but it is what the theory tells us but it it is a really important property

in the end right because this is what defines the properties of

material versus say the force carriers right force carriers have integral spin or in the case of the Higgs

Boson, zero spin which is still integral that's carrying the force and then electrons and muons and taus and their

neutrinos have spin 1/2 so there's a certain that's the amount of spin they have but I guess we learned

about this because you can add that spin onto angular momentum somehow right you

can add that on to to that value but okay so this is somehow

magically this elementary particle has this property and that gives it a

magnetic moment. And how would you with theory tell me what that magnetic

moment should be? What would you guess it would be? Well, one thing to say is that spin is

not enough in general to have a magnetic moment in the case of an elementary particle. You also need a particle which is

charged. Okay. Instead if you have a composite particle so a particle which

is made out of smaller building blocks that can have a magnetic moment

also if it's electrically neutral for example the neutron is made of up and down quarks which are

individually charged. Okay. And you can imagine them to move a bit in the in the nucleus.

Okay. So the charge distribution of the neutron even though the neutron is overall neutral

has a non-trivial shape and this in combination with the spin is

what allows you to have a magnetic moment. So not every particle has a magnetic moment. Not every particles but the leptons

apparently do because they do. Yeah. They all do. So we have electrons and muons and taus and why maybe I can

ask Fred why did you guys at Fermilab and first at Brookhaven decide you

wanted to measure the magnetic moment of the muon. The muon is really kind of ideal because

it has the the perfect combination of mass and lifetime. So it turns

out that if you're looking for the contributions of new physics to the

magnetic moment of a particle it helps if that particle is more massive and the muon is about 200 times more

massive than the electron and that amplifies its sensitivity to new physics

by a factor of about 200 times squared. And so that means that it's a

great laboratory for searching for new physics but also we have to have

them around long enough to be able to study them. So the lifetime has to be long

enough. Of course electrons live forever, right? They're stable particles and the magnetic moment of the electron

has been measured in the lab incredibly precisely. But it's not really so sensitive to new physics. Meanwhile, the

lifetime of the tau is incredibly short. It's not long enough to be able

to study it effectively. And so, the lifetime of the muon is long

enough that we can circulate them around an accelerator, at rest a muon lives on

average a little more than two microseconds. In our experiment about 64 microseconds

because they're moving at close to the speed of light which actually increases their lifetime from our point

of view because their clocks are running slowly. This is a really interesting thing from special

relativity. So there's a secret to longevity if everybody if you really want to live long move really super fast. And I

mean this is why we know about muons, right? Muons hit us over the head all the time. Not to scare you but while

we've been talking hundreds of muons have gone through you through your head from interactions in our in our upper

atmosphere and that was one of the things that got us started out on the whole path of doing particle physics right we saw

mother nature's doing this all the time I saw a calculation once that said that mother nature had done the whole LHC

program about 10,000 times already in our upper atmosphere so another reason to see that it's a safe thing this

has happened we've done this, but we just didn't have our detectors around to measure everything that was happening at the time. Okay, so let's get

back to to the muon. Well, actually, we should mention this that there's that lifetime, but the tau is much more

massive. So, that's the difference between these guys, electrons, muons, and taus. And it turns out that when you're more massive, as my wife

tells me you're going to decay more quickly. But it's true for elementary particles. The more the

heavier they are, they have more ways they can decay in. Right. Decay is a terrible word. I hate to use decay

because they don't really decay. They change. Transform. They transform from one type to another.

They don't actually It's not like there's a muon. There's no electrons inside a muon, right? It's still an elementary particle. Okay. So, you've

made this really nice measurement. Not just you. your whole collaboration. 176 of us.

176 of you working on this from all from the US because it's in Fermilab. No, it's very much an international

collaboration with a lot of major contributions from places including

Italy, Germany, the United Kingdom, South Korea and China. Wow okay so it's it's as international

as many of our even larger collaborations that are here at CERN and it didn't even start at Fermilab

did it. I mentioned Brookhaven And it started there, right? I would even say before that it

started with a series of three experiments here at CERN. And then after that came an experiment at Brookhaven

that I worked on for my PhD. So I've been doing this for quite a long time now. We were taking that data back in

the late 1990s, early 2000s. And we had an opportunity to relocate the

experiment to Fermilab long after we were done at Brookhaven. Because at

Fermilab they really had some great ideas about how to with their accelerator complex serve up an enormous

number of muons to the experiment in small little batches for us which turns

out to matter and so with many many muons because

really our experiment is still limited by the number of muons that we've counted by the statistics

and with them divided up so that they don't pile up with each other as they

decay so that we can count each muon decay separately. Wow.

So that's the motivation for moving it there to Fermilab. So and that was easy right? You just put

it in a box and ship it you know UPS orÉ Well, when we first got started

talking about the logistics we were actually talking about picking it up with a helicopter and moving it with a

helicopter to a barge in the port. It turns out the helicopter idea was not so

practical. So it went onto a truck to a barge around the tip of Florida into

the Gulf and then up through the rivers and up to the port of Lemont near

Chicago and then back onto a truck on the tollways in the Chicago

area out to Fermilab. So I bet quite a production. I bet everybody loved the traffic jam.

I got to ride along with it one night and yeah, you could certainly sense that

at the closed on ramps to the tollway, people were a little bit annoyed, but at

the same time, I think it was worth it for the science that we got. Sure. Well, I mean, we we've been

through that here. I remember taking my son to school actually at CERN has a little school for when

they're very young and and us being slowed down because there was a blue dipole. My son at three was already

saying, "Look, dad, a blue dipole." So, we we've had that. So, when you when you do these major things and when we build

our future accelerators in the local areas, we'll have to make sure we all understand it's a slight sacrifice, but

it's worth it for the for the science that's being done. So, I mentioned the amazing precision and I think

that's important to note here because I love to wax poetic about the experiment I work on. I work on the Atlas experiment and it's an enormous thing.

It's some half of a football field in length. And so over

these 46 meters we also line up our detector at the

far ends to tens of microns. Right. So that means we have

one in a million precision. Now your measurement gives us a precision of what amount?

A little better than 140 parts in a billion. And so that is the

analogy that I've heard recently from one of my colleagues is that they have these big animals a little bit bigger

than cows at Fermilab, Bison. They're famous and so the precision is about

one sunflower seed out of a whole bison. Okay. So if a bison were to eat one more

sunflower seed, you could tell the difference within that level of precision. That's right. That that's a good definition of

precision. I think the only the only experiment that beats us I can tell you there is one that beats us. That would be LIGO, right? Their measurements of

gravitational waves. Impossible. I don't know how they do it. That's impossible. But you did a lot of work to get that

far. But what I'd like to know actually I turn to theory now. Okay. Because Simon's also working on trying to

improve theory. This is a very interesting thing because you guys have them beat a little bit here. The

precision of the measurement. Tell us about that Simon. What what's going on? Why is theory not able to to have a

smaller error bar than the experiment? I mean the reason is not that we were not trying. So actually I mean before

these guys built their experiments there was this g-2 theory initiative. So

really hundreds of people from all over the world coming together with the idea that we should give our best to have the

most precise prediction of the theory. to really do the best doing calculations

to be able to compare to experiment and it was clear that we have to improve and that the experiment would be

very precise, but it's an incredibly difficult calculation to do

so I mean the question is what is happening there and the point is that at

this precision we're talking about everything matters really everything in the sense that this muon and magnetic

field doesn't only talk to the magnetic fields which is composed out of photons

but they're all these quantum fluctuations and it's really talking to all the forces in the standard model which are the weak interaction the

strong interaction and the electromagnetic interaction. Okay, that's a strange thing. Let's just

try to wrap our heads around this for a second here. So everything's going on, right? You have a certain amount of

energy. I know that if we take our protons and accelerate them, it turns out there's things popping in and out of

the vacuum out of nowhere. And perhaps the big bang was us just popping out of nowhere. I don't know.

But that's happening right at high energies. So, maybe Claudia, can you explain to me a little.

But how does this work? How how is it that we don't get an answer that's

precise that we have to sort of expand, I guess. How does that work?

Okay. It's complicated to explain. It's complicated. I'm gonna I'm gonna try. Okay. Well, in general, when we try

to make a prediction for an observable to make a a calculation that then can be

compared to the experiment, ideally we want to try to do it in the

simplest way possible. And

depending on the precision of the experiment, we need to get a certain precision on our theory prediction too.

We kind of want to match that precision. If the experiment is not very precise,

it might be enough to make some we can call it some sort of approximation which is good enough to get a prediction of

the same precision as the experiment. But as experiments get more and more precise,

we cannot do these approximations anymore. And then things become very easily much

more complicated. Now in general, depending on the type of interactions

you're looking at it might be possible that at every

step the effect gets smaller and smaller. So by doing the

first-order computation you already get the bulk of the result. But what these guys in practice deal

with is a situation where everything matters at the same time. There is no

approximation. Everything everywhere all at once. I guess what's going on here?

So I mean it is a very interesting because most people would think all right and and it's very normal to think

that you have a theory and the theory will tell you well it should be.

You know you have a simple thing. You have a muon here it is it's got a charge and a spin and you do your calculation but it

turns out because of the energy things can pop in and out of a

vacuum, you end up having to do and this is this This is what actually drew

me into the field I have to admit in quantum field theory was that when you have an interaction there are a lot of

different possible things going on right so I always mention this when I talk to people about what goes on when we

collide protons and and these two quark let's say or two gluons interact

there's many different ways that that can happen and we're not even allowed to see that right we never know so you guys

never really get to get inside there and see what's what exactly happened with this muon becoming an electron. And it

could have gone through many different paths. Some of those very complex, but the more complex they are, the smaller the probability. And some of

them straightforward. And you end up with this sort of infinite sums, right? You get something that we learn if

you go far enough in math at school, you'll find these sort of expansions.

and there you see that the bigger the further you go

the smaller the numbers are but as you're saying, in this world in the

world of the the muon here you have to go out pretty far to get the precision and Simon's just not getting it done!

Well no, it's really complicated I mean so there's the frustrating thing maybe is

that 99.99% of this theory prediction is is based on

this electromagnetic interaction and people can compute it very nicely and it's super precise and then there's this

small piece it's called the hadronic vacuum polarization contribution and that's basically I mean there's the muon

and then there's a photon that tells the muon how strong the magnetic field is and then this photon

makes a quark anti-quark pair and as soon as we have this everything happens it's there are tons of particles and

everything happens at the same time and we cannot just sit down and write a formula and get the result and we

cannot even go to a computer and press a button and get the result.

So we have to go to a computer but it takes years and years and the biggest computers in the world to get to results

that we have to understand this. There's a name for this right? Lattice QCD. It's called Lattice QCD and it's really

the only method that we know that reliably can tackle these this strong interactions, these hadronic properties of

a theory. Yeah. So just to recap I mean the strong interaction but there's for some reason there are people out there who don't

know about QCD and the strong interaction they ought to. That's you know the interaction between quarks

Quarks are what put we put together to make protons and we make neutrons out of them and there's a whole lot of other

hadrons our friends at LHCb are finding new combinations all the time, including pentaquarks

and all these different excited states of these but in this case you can produce these these pairs and

when you do that lots of different things can can happen. So, so I guess I mean I should note that that we had to

bring on two theorists because the experimentalist has done his job over there. But it seems

the theory is really more than twice as hard, I have to say. Yeah. Okay. It's a very

hard job to try to get this down. And one thing that's important to note is at no point are either of you guys trying

to be like the other, right? So you're not saying, "Oh, my result is over here." And for a while it's been like

that, right? The result of the measurement has been a little ways away

from theory which makes it intriguing and interesting. But all you're trying to match up is the precision. Try to get

as good a precision or better than the other and then you'll see what nature tells you if they agree. So maybe

Fred you can expand a little bit more on how because you guys did an amazing job to get the precision that you have. What

did you have to do? Right. Well, the first key was simply measuring an awful lot of muons. So,

we recorded a little more than 300 billion final events that made it

into our final data set. And to do that, we actually had to store more than a trillion muons in the storage ring. So,

that was really the first key was simply having this incredibly high

statistical power. And then we really had to pay attention to like you said

before everything everywhere all the time. So we had to focus on really

improving every possible systematic error. And you can see going from one

year to another the kinds of things that we improved. So for example in our

experiment the technique really relies on what's called the magic momentum which is to say that the muons

have exactly the right momentum that the electric fields don't rotate their spin

so that we can use large electric fields to confine and focus them inside the storage ring.

And yeah on the other hand not every muon is going to have exactly the same

momentum. Some will be going a little fast and some will be going a little bit slow. So for this most recent result, we

just had to work much harder on measuring the distribution of momentum including building a whole new detector

like some of my colleagues at the University of Washington did. They call it the mini sci-fi. And so we introduced

this in order to get just that much better precision on what we knew about

the momentum. And you know for example, we need to know the magnetic field incredibly precisely that

the muons experience and so we've been

measuring the magnetic field in a sort of static way. So the part of the

magnetic field that doesn't vary with time by pulling a trolley full of magnetic field probes all around the

inside of the vacuum chambers at a time when the muons aren't there. But for this most recent measurement, we also

were able to measure the time dependent magnetic fields from when we fire the kickers that kick the muons onto the

correct orbit and from when we energize the quadrupole plates that make the big

electric fields that confine the particles. So, you know, very

precise measurements of those things were included progressively as we moved from the analysis of one run into

another to get to better and better precision. And we're at the point now where really

we would again have to improve everything if we wanted to continue to make progress here. We've really gotten

those systematic errors down as small as we practically can. I think there's a

few technical terms I should mention here you kickers for example or quadruples but these are all sort of

means of using electromagnetic fields. You have an electric field to push things along make them go higher

energies and then and then your quadrupoles also will do they focus or

they are for focusing that's right and actually we don't have anything that

that changes the energy of the particle once it gets into the storage ring, the kicker just redirects the particles

From an orbit that would come around and hit the injection point again onto

an orbit that would be stored in the vacuum chambers that sort of centers it right in the middle of the

the quadrupoles. Okay. For those watching actually on YouTube, you have a picture of it.

It's a relatively small accelerator compared to what you see here at CERN. Usually

14 meters in diameter. But it's an amazing thing and actually it has an

interesting part to it. I noticed that when the muons come in there's like a gap, right? There's like a space between the

the magnets. And that it's fine. Muons just keep going. Right.

Right. Over a short distance that's certainly true. Although over a long enough distance you certainly want to

have them in a vacuum so that they don't scatter off of the air.

Beautiful. It was a beautiful, beautiful experiment and congratulations on getting such amazing results. I do

want to ask go back to the slight discrepancy that there was and maybe it's disappearing. We don't know.

What can we learn when we do a precision measurement like this? Maybe

I'll ask Claudia if you do a precision measurement, how does it help us when we compare

it to theory? What can we gain from that? Well, whenever there is a discrepancy in

a precision measurement, especially if this discrepancy is

statistically significant like it was in the case of the g-2, we as

theorists, all we want to do is to try to find out which type of new particles

and forces could explain that discrepancy. So, precision measurements are the

perfect venue to do this. In some sense

by looking at a discrepancy, we can learn several things. We can learn something about possibly the scale, the

mass of the particle that explains the discrepancy. And also something about its coupling.

Okay. Coupling being how they interact. How they interact. Yes. So for the

case of the g-2, two of the most common explanations were one in terms of

very light particles called axion-like particles. These are particles that arise for

example in models trying to explain dark matter or to solve another big problem of the standard model which is called

the strong CP problem. Or also other types of particles like

for example heavy bosons like heavy Z bosons that couple differently maybe

between electrons and muons and taus, etc. Then in practice when when we build

one of those models we also have to pay attention because in general the

moment you have a new particle with new interactions it's not just going to pop up in one measurement but it's also

going to arise somewhere else and then you have to make sure that your model doesn't contradict other data. So

somehow from these small discrepancies, if they are confirmed if they are

statistically relevant we can definitely learn something about what is beyond the theory that we

know right now. So and we know that there is something beyond it because there are several problems it doesn't address.

Exactly. And that it's important to note that we have this amazing theory and we keep testing it

and we'll be doing many more precision tests here. For example, at CERN, we're always doing precision tests

as well. We're not just looking out there. Is there a new particle out there? We do that and we love that. But

we will be in fact in a year tearing apart the LHC and our

experiments. At least Atlas and CMS are going to be refurbished to go to high luminosity and in which case when we get

many more collisions like you need lots of muons, we need lots of proton collisions and with that we get

the higher statistics, we get higher precision and we can look for very rare things but also look for these discrepancies because we're missing what

95% of the universe. the standard model's great butÉ

Then in the future there are many different ideas out there what we want to do in the future beyond the LHC after I've retired

for sure and that you know includes electron-positron colliders

maybe a huge FCC maybe linear colliders also possible muon colliders for those

of you in the muon business a lot of different ideas out there but precision is one of the things we really want to to look for so before we

finish up actually I would like to to hear a little bit more from Simon about you know what are the next steps. You

mentioned getting more computing that's something everybody wants that you know chatgpt

will work better too with more computing and we can use the same computers as the ones training chatgbt so it would be

great if they just hand it over to us. Let's just ask chatgbt could you

hand it over? But beyond that what else what other work can you do to improve move these models?

What work is going on so really I mean there are two ways to compute this HVP contribution that is

really carrying all the uncertainty and lattice QCD has only recently I mean last five years we were able all I mean to

reduce our uncertainties by a factor of five or so and now we're really able to make this prediction

and we're very certain that we can improve but as Claudia said as as soon as you want to get to better

precision you always have to be more careful about tiny effects that you have to control. So it's certainly getting bigger

computers, getting better algorithms. I mean be smart about how you want to

compute something. And also have new ideas. And then there's a second way one can

compute it has been done historically. It's called the Ratio method. And if they I mean they can also improve with

new experimental input actually. Okay. because some sometimes a theory has to have for various values that you

use for the input. Before we go, one last thing I'm gonna turn over to Fred here.

What's next? You said you're not going to do it yourself, right? Are you handing the baton over to

anybody else? So, at Fermilab, we still do have a few more papers to come out of our

experiment. So, we're going to look at the potential for new physics to cause

our result to change as a function of time. Maybe oscillating, getting a little bigger and smaller over time as a

result of coupling to things like dark matter fields or violations of the

postulates of special relativity. So, we'll be looking for that. We'll be looking for the electric dipole moment

of the muon. we should be able to set a limit that's stronger than the current limit on how big that could be should be

zero in the standard model. So if we were to see something that's not zero then that would be exciting. But

then the next muon g-2 experiment is going to be at J-Park in Japan and they

have a technique there that is like ours but with muons that are not at the

magic momentum. So they can't use electric fields to confine them. Instead, they cool their muon beam to

the point where they don't have to use electric fields and they can use a much smaller magnet. So, they will be able to

do a complimentary experiment with smaller uncertainties. I should also point out that the reason

that I'm here at CERN is to work on an experiment called Muon that we're setting up at the CERN SPS. And so, what

we are doing there is getting experimental input for a third way of

doing the theory. So you've heard from Simon about two ways right from

electron positron collision data the R ratio method and lattice gauge theory.

Well we're observing the hadronic vacuum polarization effects in the interaction

between a muon and an electron when a high energy muon beam hits the electrons in a target.

Okay great. So I should say SPS by the way many people might know that super proton synretron which brought us the

W and the Z beson many years back about 1983 I think it was way back and we

still use these and we have some older accelerators than that even that we still use as well. Well thank you very

much. Fred Gray is professor of physics at Regis University in beautiful Colorado. Simon

Kuberski and Claudia Cornella are both fellows here at CERN working in

the theory division. So thank you for joining us. This has been early morning

coffee at CERN, a podcast by the scientists of CERN (or in fact not just CERN)

about the science of CERN. You can find all of our episodes on the

CERN YouTube channel and anywhere you get your podcast. Guaranteed. Go ahead.

Find the place you get your podcast. You can find us. Be sure to do whatever you're supposed to do. Like and share

and I don't know, send it to your friends. There's a lot of things you're supposed to do. So, do those things and

subscribe. Our editor and producer today is Melanie Arnold. Hi, Melanie.

Good job. Thanks. Our executive producer is Jacques Fichet. Jacques Fichet and Piotr Traczyk

are our technical leads, while Ron is away getting fixed up. Our studio manager is Max Price. Sound

engineering is by Piotr. Our original theme comes from the Canettes Blues Band with

piano by Wojt play-it-in-any-key Krajewski. Many thanks to Paula Catapano, Matthew

Chalmers, and Arnaud Marsollier for all of their advice and strategic planning. Big thanks to the entire Education

Communication and Outreach team here at CERN for providing us with access to beautiful Wire Chamber Studio and all

the help that comes with it. I also want to take a moment to thank Cetna Krishna who's no longer with us. (That sounds

Terrible!) It's because she finished up her fellowship here at CERN. She

along with Joni Pham were really two of the people who came up with the whole idea of doing this, and I get to have a lot of fun and I hope that they're doing

well as well. Opinions expressed here are our own and do not necessarily

reflect those of our colleagues or of CERN although we think they ought to. My name is Steven Goldfarb. This has

been Early Morning Coffee at CERN. We'll see you next month.