# How do plasma wakefield accelerators work and are they the next step? **URL:** \ **Category:** Factual Questions\ **Created:** [April 26, 2013, 10:39pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799 "2013-04-26T22:39:54Z")\ **Posts on this page:** 16\ **Page:** 1
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [April 26, 2013, 10:39pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/1 "2013-04-26T22:39:54Z")
The next step for high-energy particle physics that is. Apparently, it should be possible to build such a device less than a kilometer long that will be capable of multi-tera-electron volt energies surpassing even the LHC. In addition, they could use electrons or positrons which would generate much cleaner collisions. [Here is the Stanford site’s section on the topic and their explanation:](https://portal.slac.stanford.edu/sites/ard_public/facet/Pages/rpwa.aspx) > [@](#): > > The Beam-driven Plasma Wakefield Accelerator > > A plasma accelerator replaces conventional static metallic (copper or niobium) accelerating structures with a structure that is formed dynamically with a tube of plasma. The basic concept of the plasma wakefield accelerator involves the passage of a near-lightspeed electron bunch through a stationary plasma. The plasma can be formed by ionizing a gas with a laser or through field-ionization by the incoming electron bunch itself. This second method allows the production of meter-long, dense plasmas suitable for plasma wakefield acceleration, and greatly simplifies the experimental setup. In single-bunch experiments, the head of the bunch creates the plasma and drives a “wake” of charge. The wake is driven out of the path of the incoming electrons, creating a charge imbalance that pulls it crashing back in, behind the passing electron bunch. This effect produces a strong field that accelerates particles in the back of the bunch. The system effectively operates as a transformer, where the energy from the particles in the head is transferred to those in the back, through the plasma wake. The physics is similar if there are two bunches rather than one; energy from the leading “drive” bunch is transferred to a trailing “witness” bunch. ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [April 27, 2013, 5:18am UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/2 "2013-04-27T05:18:25Z")
I’ve been doing some research and it seems that there are a number of different schemes that can be used. The central theme seems to be the idea of creating an oscillating electric field in the plasma. It seems that what you do is cook some lithium or other element to the point where you boil off all of the electrons and the whole mess has too much kinetic energy for the electrons to re-establish stable orbits - IOW, a plasma. Next you want to pass something through the cavity containing the plasma that has a charge. That can be another beam of ions, protons, electrons, pretty much anything. Let’s say protons since that what the quote below uses. The electrons will be attracted and the ions repelled so the plasma becomes polarized. As this driver beam passes, everything snaps back to the way it was. This imparts energy to the accelerated bunch that trails the drive beam although I’m still not really clear on how. However most of what I read talks about using a single bunch of electrons as the driver. In that case, the same thing happens. The plasma is still polarized, but now you are draining energy from the electrons at the front of the bunch and via the plasma oscillation, passing it to the trailing electrons. Again, not clear on how that works. Also, not clear on how or even if you siphon off the limp electrons. Anyway, [here is an explanation](http://www.mpp.mpg.de/english/research/experimental/plasmaWakefieldAc/index.html) that uses a nice analogy and sort of makes sense. > [@](#): > > Plasmas can sustain very large electric fields. Therefore they are a very good medium to accelerate particles at very high rate. With a proton bunch as a driver a charge wave is created in a plasma. In order to explain this process, one must know that plasma consists of free negatively charged electrons and positively charged ions. The proton beam is carrying a positive charge. If such a positively charged proton bunch enters the plasma, the proton bunch will attract the plasma electrons towards itself. Due to the high speed of the proton beam, the plasma electrons overshoot and oscillate. The oscillation creates periodic regions of excess and rarefaction of plasma electrons (i.e., of excess plasma ions). This oscillation travels at the same velocity as the drive bunch. > > If a fast particle is injected into the plasma oscillation at the right time, then it can “ride” the wave and thus is accelerated. This process one can be imagined using the image of a fast boat creating a water wave and of a surfer riding on a wave and gaining energy. > The goal now is to explore and perfect the acceleration process until a high quality witness electron bunch can be accelerated for other experiments. ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [May 24, 2013, 2:38am UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/3 "2013-05-24T02:38:58Z")
[Here’s an update on some of the research in this area at FACET](http://www.symmetrymagazine.org/article/may-2013/facet-boosts-plasma-acceleration) ---
**Author:** ![thirdname](https://avatars.discourse-cdn.com/v4/letter/t/d07c76/32.png) [@thirdname](https://boards.straightdope.com/u/thirdname)\ **Post date:** [May 24, 2013, 5:31am UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/4 "2013-05-24T05:31:44Z")
I was hoping a Plasma Wakefield Accelerator was some sort of device to launch anti-vaccine researcher Andrew Wakefield into the sun. ---
**Author:** ![moriah](https://avatars.discourse-cdn.com/v4/letter/m/b2d939/32.png) [@moriah](https://boards.straightdope.com/u/moriah)\ **Post date:** [May 24, 2013, 3:47pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/5 "2013-05-24T15:47:01Z")
Plasma Wakefield Accelerator… I loved that band! They’re now touring in a retro concert series with Electric Light Orchestra, REO Speedwagon, and Bachman Turner Overdrive. ---
**Author:** ![Sailboat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sailboat/32/461_2.png) [@Sailboat](https://boards.straightdope.com/u/Sailboat)\ **Post date:** [May 24, 2013, 4:20pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/6 "2013-05-24T16:20:47Z")
> [@thirdname](#): > > I was hoping a Plasma Wakefield Accelerator was some sort of device to launch anti-vaccine researcher Andrew Wakefield into the sun. DAMMIT, you beat me to exactly that comment. Well, hopefully _someone_ is working on such a device. ---
**Author:** ![Pasta](https://avatars.discourse-cdn.com/v4/letter/p/ecccb3/32.png) [@Pasta](https://boards.straightdope.com/u/Pasta)\ **Post date:** [May 24, 2013, 5:56pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/7 "2013-05-24T17:56:40Z")
As a summary of the mechanism… Consider electrons as the driver, and consider a fixed position through the plasma chamber: Position X. The driver bunch comes screaming into the plasma from the left. As it approaches Position X, it begins pushing the free electrons near X out of the way, primarily sideways. (It’s was previously doing this to electrons upstream of X, but we’ll just look at X since the same thing happens everywhere, just at different times.) The driver pulse soon passes X and continues on its way, and the electrons that were displaced find themselves far away from the positive charges left near the core of the plasma. (The positive ions are too heavy to move a significant distance on this timescale, so consider them frozen in place.) The electrons also find themselves in an over-density of electrons, and all told, they are pulled/pushed very strongly back to the middle. Just as with a mass on a spring, there is no reason for them to instantly stop when they get back to where they started. They will arrive at the neutral position with significant momentum and will thus overshoot. This recreates the original pushed-out situation, only now with each electron on the opposite side from before. This will repeat until it damps out. Note that this oscillation phenomenon is not part of the wakefield acceleration itself. It’s a side effect that can be used for significant engineering advantages (in particular, multiple bunches can be accelerated from a single driver bunch). For a dual-bunch scheme, the acceleration of the so-called “witness” bunch is all about timing. If you send in this second bunch of electrons and time it such that the plasma’s electrons are all out of the way at X right before the incoming bunch gets to X, the incoming bunch will see ahead of it a nice attractive positive ion soup at X, and it will be accelerated toward it. As the witness bunch passes X, the plasma electrons have begun crashing back down behind it toward, and then through, their “neutral” position. Since the driver bunch is continually creating a bubble of positive charge around it by pushing away electrons, and since the witness bunch is continually seeing that positive charge bubble out in front of it, you get a continuous acceleration through the plasma cavity. The technology is at an early proof-of-principle stage. There is a long road ahead, and no one yet knows if work-horse accelerators are possible. To give a flavor of the complexities: - A high-energy accelerator needs not only to produce high energy particles but also a lot of them, and getting a lot of particles through a PWA is non-trivial. For small bunches the plasma behaves linearly, but as you increase the bunch size (i.e., number of particles being accelerated), you get strong non-linear phenomena that are hard to control. The whole task is one of subtle balance even before non-linearities kick in, with details like the density of the plasma along the length of the chamber needing to be controlled carefully. - An accelerator should provide particles with a narrow range of energies (within 1% of each other, say). Because the wakefield time scales are so short, the front of a bunch and the back of it will end up seeing different net accelerations as the plasma distributions respond and evolve from the passing of the driver, leading to a wide spread in energies. Using lasers for the driving together with very narrow witness bunches helps, but lasers have a ceiling on their usefulness due to diffraction limits for real-world-sized cavities and due to overall power consumption. (One quickly surpasses multi-terawatt power consumption for a competitive laser-based PWA.) - Electrons are convenient drivers, but they can only boost other electrons by a factor of two in energy, which is hardly worth it. Using protons as the driver could solve this, but protons are not readily used for this purpose. For instance, you need very short (length-wise) bunches to drive the wakefield since the restoring timescales are very short. Today’s proton bunch lengths are typically tens of centimeters, but you need tenths-of-millimeter lengths (\>1000x smaller) for PWA. PWA is in a similar position to fusion power thirty years ago. It should be possible in principle, but only the next step or two out of hundreds is clearly visible. We might have something competitive and workable in 60-90 years, or we might not. As with fusion, I’m a big fan of supporting this sort of forward-looking R&D since the one thing that _is_ certain is that we will hit practical limits with existing technologies. ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [May 24, 2013, 6:01pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/8 "2013-05-24T18:01:36Z")
I don’t understand. The driver model is one option right? Can’t you polarize the cavity after you’ve created a plasma? ---
**Author:** ![FordPrefect](https://avatars.discourse-cdn.com/v4/letter/f/e99b99/32.png) [@FordPrefect](https://boards.straightdope.com/u/FordPrefect)\ **Post date:** [May 24, 2013, 6:11pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/9 "2013-05-24T18:11:34Z")
> [@thirdname](#): > > I was hoping a Plasma Wakefield Accelerator was some sort of device to launch anti-vaccine researcher Andrew Wakefield into the sun. Based on my very limited understanding of Angular Momentum, what you are looking for is a Plasma Wakefield Decelerator to drop him into the sun. ---
**Author:** ![Pasta](https://avatars.discourse-cdn.com/v4/letter/p/ecccb3/32.png) [@Pasta](https://boards.straightdope.com/u/Pasta)\ **Post date:** [May 24, 2013, 6:26pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/10 "2013-05-24T18:26:12Z")
> [@deltasigma](#): > > The driver model is one option right? Can’t you polarize the cavity after you’ve created a plasma? Can you say what you mean by “polarize the cavity”? All wakefield acceleration approaches use something to shape the plasma electrons in a dynamic way – that is, to create a wake. ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [May 24, 2013, 6:49pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/11 "2013-05-24T18:49:51Z")
Maybe I’m misremembering something else. I was trying to find it now but it wasn’t in evernote, but that could just mean it’s print or pdf. Anyway, I think the idea was that if the plasma is already polarized around the axis ( + center to - periphery, or vice versa), it serves at least part of the purpose of the driver in segregating the charges. The wakefield still ends up occurring though right? I guess now you tell me what I missed. ---
**Author:** ![Pasta](https://avatars.discourse-cdn.com/v4/letter/p/ecccb3/32.png) [@Pasta](https://boards.straightdope.com/u/Pasta)\ **Post date:** [May 24, 2013, 7:34pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/12 "2013-05-24T19:34:09Z")
> [@deltasigma](#): > > Anyway, I think the idea was that if the plasma is already polarized around the axis ( + center to - periphery, or vice versa), it serves at least part of the purpose of the driver in segregating the charges. Once the driver causes the wake, with + center and - periphery, that pattern will self-sustain in an oscillatory way for a time, so multiple bunches, if timed right, can take advantage of the wakefields that come and go and come and go and …, as the electrons oscillate back and forth around the central region. But the driver is still needed to produce the wakes initially and repeatedly, as the oscillations only last a brief time. In application, it would look something like: (1) Send driver pulse through to create a wake in the plasma. (2) Send dozens or maybe hundreds of “microbunches” separated by just the right amount so that each passes through the plasma just upstream of one of the “+” regions that have been created along the chamber.[sup]1[/sup] (3) Start the whole process over, as all these particles have passed on to their demise at the collision point or whatnot. [sup]1[/sup]The picture is that at a given longitudinal position, the electrons are just moving up and down in an oscillatory pattern. However, the phase of this oscillatory pattern varies across the chamber and in fact travels at the speed of the driver pulse. So, the points along the axis that are currently showing an excess of positive charge are separated by a distance governed by the plasma oscillation frequency but that pattern of + regions is moving along the axis at high speed. So you just have to get the witness bunches to sit up just behind each + region. And the pattern of + regions is moving along the axis, caused by plasma electrons that are only moving up and down (ideally). ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [June 10, 2013, 1:30am UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/13 "2013-06-10T01:30:36Z")
[If I’m reading this correctly](http://www.sciencedaily.com/releases/2013/06/130603113348.htm), it looks like the problem of creating particles of uniform energy, at least in the case of electrons, may have been solved. > [@](#): > > In the experiment, the physicists released helium atoms from a small nozzle at supersonic velocity. Directly above the nozzle they placed a razor blade such that it obscured a part of the nozzle orifice. After the supersonic helium stream is released from the nozzle and hits the razor blade edge its forms a shock front and hence a density step in the gas. At precisely that position the researchers focused an extremely strong laser pulse with a duration of 28 femtoseconds (One femtosecond is one millionth of a billionth of a second). ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [July 22, 2013, 6:43am UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/14 "2013-07-22T06:43:14Z")
I should really wait for something human readable[sup]_1_[/sup] before posting about [this](http://arxiv.org/ftp/arxiv/papers/1307/1307.4159.pdf), but the amount of progress in this area seems to be accelerating and this seems to be a substantial jump over what was discussed in the previous post so I’m going for it. Specifically, their claim is that their method should be able to produce 100 Gev beams using a dual stage method - but that’s the best I’ve got for an explanation. I’ll try to look at it again tomorrow. _1_ IOW, NOT a journal article. 🙂 ---
**Author:** ![deltasigma](https://avatars.discourse-cdn.com/v4/letter/d/e5b9ba/32.png) [@deltasigma](https://boards.straightdope.com/u/deltasigma)\ **Post date:** [October 13, 2013, 8:27pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/15 "2013-10-13T20:27:22Z")
These don’t sound like plasma wakefield devices but since they seem to be similar I figured I post this here for anyone interested. [Etched glass could create table-top particle accelerators](http://physicsworld.com/cws/article/news/2013/oct/03/etched-glass-could-create-table-top-particle-accelerators) ---
**Author:** ![Hypno-Toad](https://avatars.discourse-cdn.com/v4/letter/h/b2d939/32.png) [@Hypno-Toad](https://boards.straightdope.com/u/Hypno-Toad)\ **Post date:** [October 14, 2013, 7:55pm UTC](https://boards.straightdope.com/t/how-do-plasma-wakefield-accelerators-work-and-are-they-the-next-step/656799/16 "2013-10-14T19:55:20Z")
Desk-top particle accelerators? I wonder if one day people will look back on the LHC with the same wry amusement that we look at UNIVAC while using an iPhone.