Two-stage electron Wakefield acceleration

[mfb]

The energy of electron-positron colliders is limited by the acceleration gradient (energy per distance) and the length of the accelerator (=the cost). In the past the energy mainly increased from making accelerators larger. Much larger accelerators than today get prohibitively expensive, however. More compact designs could keep them affordable. While conventional cavities still make some progress in the acceleration gradient (e.g. for XFEL), they only reach about 30-35 MeV/m. Increasing this much more will need new approaches, but could lead to much higher collision energies. It would also be useful for future x-ray lasers.

One proposed concept is the Compact Linear Collider (CLIC), using a high-intensity low-energy beam to create electromagnetic waves to accelerate a low-intensity high-energy beam. The goal is 100 MeV/m, three times more than conventional RF cavities. While the general concept has been demonstrated before, it is still a challenge to scale that up to a full accelerator concept. Scientists at Argonne now managed to combine two acceleration stages:

Electron acceleration through two successive electron beam driven wakefield acceleration stages
The stages were just 3.5 cm long and increased the energy of a beam by 2.4 MeV per stage, corresponding to 70 MeV/m. That is not enough for a proper collider, but it is an important step towards longer acceleration tracks. They hope to increase this to 300 MeV/m and more stages as demonstration object for a full accelerator. That is beyond the CLIC goals and would allow an even higher energy or a more compact design.

There is also plasma wakefield acceleration as competing technology - replacing the cavities by a plasma and letting the drive beam fly through the same path as the accelerated beam. While the achieved gradients are much higher here (100 GeV/m over short distances, 50 GeV/m over about a meter) the beam quality decreases - it is unclear if this approach can be used in multiple stages and for a collider where a good focus is crucial to achieve high collision rates. Several groups are testing this. If this can be built in stages and used for an accelerator, then we might see completely new energy regions with future accelerators.

 

[Vanadium 50]

The plasma guys have also managed to stage. However, they face a serious problem with beam quality - the price you pay for these high gradients.

 

[areion]

So compared to the 26 km long LHC, a future accelerator which uses the plasma technology would need to be about 300 meters long to reach same collision energy than the current max of 13 TeV ? 300m * 50 GeV/m = 15 TeV

 

[ZapperZ]

Wow! It's sad that the highlighted accomplishment, which is using a dielectric wakefield scheme, is being buried by the "sexier" plasma wakefield.

If you want to have some idea of what is being done here, check out this article:

https://www.aps.anl.gov/files/APS-sync/lsnotes/files/APS_1420321.pdf

Three notes:

1. The dielectric wakefield scheme can be (i) co-linear or (ii) parallel beam. In the co-linear scheme, the drive electron beam (which generates the wakefield) is in the same beamline as the "witness" beam (the electron beam bunch that is being accelerated). In the parallel beam scheme, the drive beam is in a separate beamline. The wakefield it generates is then transferred to the witness beamline via a waveguide.

2. For plasma wakefield, there is an electron-driven plasma wakefield, and a laser-driven plasma wakefield. So the wakefield in the plasma can be generated either by an electron beam or a laser.

3. The ability to do staging here is a significant step, because this is what is needed to basically build "accelerating modules". The higher the energy you need, the more modules you build. The ability to show a proof-of-principle experiment like this is to show that this dielectric scheme can produce acceleration in stages. It shows that each electron beam bunch stays compact and intact (beam quality) from one stage to the next. This isn't trivial because it isn't easy to do for plasma wakefield scheme... yet!

Zz.

 

[mfb]

The plasma guys have also managed to stage. However, they face a serious problem with beam quality - the price you pay for these high gradients.

Nice, I missed that.

So compared to the 26 km long LHC, a future accelerator which uses the plasma technology would need to be about 300 meters long to reach same collision energy than the current max of 13 TeV ? 300m * 50 GeV/m = 15 TeV

The actual length would be more because you need focusing elements, space between the accelerating modules and so on, but if the plasma wakefield acceleration can have hundreds of stages with this gradient an accelerator 1-2 km long might reach the same collision energy. Even better: It might reach 15 TeV with electron/positron collisions where the whole energy is available for the collision products (unlike proton-proton collisions where you mainly collide gluons with gluons, each carrying a smaller fraction of the total energy).

2. For plasma wakefield, there is an electron-driven plasma wakefield, and a laser-driven plasma wakefield. So the wakefield in the plasma can be generated either by an electron beam or a laser.

And proton-driven plasma wakefield, e. g. at AWAKE.

 

[Vanadium 50]

hundreds of stages

That's going to be hard.

The beam quality is not good. Stages make it worse, not better (although the fact that the energy goes up mitigates this at some level). And the way the plasma people have been getting these enormous gradients - volts per meter - is by shrinking distances more than anything else. (To Bella's credit, they accelerate over centimeters and not millimeters or less)

I believe you will see a dielectric wakefield accelerator long before a plasma wakefield.

 

[mfb]

The ~50 GeV/m has been achieved as 42 GeV over 85 cm.[1] Sure, so far no one knows how or if you can stage the plasma acceleration, but maybe a viable approach can be found in the future.

I believe you will see a dielectric wakefield accelerator long before a plasma wakefield.

Maybe. The gradient is much lower, however. Maybe we get a CLIC-like accelerator and later a plasma wakefield based accelerator. Or maybe we are lucky and a successor to AWAKE demonstrates 130-200 GeV in a single stage with good beam quality. Would be enough for Higgs+Z production and even $t\bar t$ with the higher energy.

 

[ZapperZ]

But unless the physics changes, there is an inherent problem with plasma wakefield scheme. Exciting a plasma is inherently non-linear. So one may get high gradients out of this, but this isn't the only excitation that you get. You get other excitations as well, and this is where the degradation of the beam quality comes in. The dielectric structure, on the other hand, has been shown to damp higher-order modes, and it is a major point in why dielectric wakefield scheme is being pursued despite it having the potential of a lower accelerating gradient than plasma wakefield.

Maybe this isn't that big of an issue with particle colliders, but for FEL sources and ERL (energy-recovery linacs), the beam emittance is a major issue. New generation of FELs are requiring even more stringent emittance requirement beyond what has been achieved today. So beam degradation is not insignificant for accelerator scientists, and as of now, I don't see how plasma wakefield can overcome this problem.

Zz.

 

[Vanadium 50]

Maybe this isn't that big of an issue with particle colliders, but for FEL sources and ERL (energy-recovery linacs), the beam emittance is a major issue

It is a huge issue with colliders.

Maybe. The gradient is much lower, however.

Yes, but it's higher than with RF. We switched from propeller-driven aircraft to jets even though rockets are faster. The LHC gradients are 5 MV/m. The ILC wants 31.5 MV/m. Dielectric wakefields can do over 100 MV/m.

Think about it this way. Suppose I could double the gradient at a cost of making the beam quality twice as bad. Would you do that? What if I had ten times the gradient with beam quality ten times worse? Would you do that? You can see that this is pushing in the wrong direction, which is why plasmas have a long way to go.

 

[mfb]

The LHC gradients don't really matter, the length of the accelerating structure is tiny compared to the ring length and it doesn't limit anything.

Suppose I could double the gradient at a cost of making the beam quality twice as bad. Would you do that? What if I had ten times the gradient with beam quality ten times worse? Would you do that?

That would need a study with actual numbers. And currently we don't have these numbers yet.

 

[ZapperZ]

Looks like the plasma wakefield accelerator people are getting serious here. FACET II is being built!

https://www6.slac.stanford.edu/news...cility-revolutionary-accelerator-science.aspx

Zz.