Belle II collisions in 2021

  • I
  • Thread starter mfb
  • Start date
  • Tags
  • Featured
  • #1
No LHC this year (long shutdown), but Belle II at SuperKEKB started taking data a few days ago. Here is a press release.

Belle II started last year with a low luminosity (low collision rate) - still good for detector calibration and so on. The goal for this year is to increase the luminosity and eventually collect enough data for interesting physics analyses.
Another change relative to last year is the inner tracking detector - last year dedicated radiation monitoring was installed in its place, now the collision point is surrounded by a tracking detector.

Currently a lot of the time is spent on machine studies to improve the performance - higher collision rates, lower background, less time spent on injections and so on. In between the accelerator delivers collisions to Belle II. With increasing collision rate more and more time will be spent on stable collisions and less time will be spent on machine studies.

Like last year you can follow the accelerator status live - also available as daily version and now with an archive. The sawtooth pattern, e.g., is data-taking with collisions: Over time the beam currents (red) decrease, after a while data-taking is stopped and more particles are filled in. Unlike for the LHC the particles don't have to increase their energy in the ring any more, SuperKEKB can inject more particles while some beam is still present.

The design luminosity is 1036/(cm2s), a factor 40 higher than the predecessor, Belle at KEKB. It will take years to reach that - beating the record of KEKB will be interesting already, and that should go much faster. A big challenge is the background level. Particles in the beams can hit the accelerator walls, emit synchrotron radiation or produce some high energetic particles in other ways, these things can then lead to secondary particles and so on, and everything that hits the detector makes data-taking more difficult. The accelerator group will have to put more and more particles in the accelerator and improving their focusing without increasing the background too much.
  • Like
  • Informative
Likes dlgoff, Amrator, atyy and 8 others
  • #2
The goal for this year is to increase the luminosity and eventually collect enough data for interesting physics analyses.

What types of analyses do they hope to perform with the data? What are the main questions they aim to try to answer?
  • #3
I'm not an expert in Belle II physics but the main focus will be on events that are difficult to select: Belle II won't come close to the integrated luminosity (and therefore number of B mesons produced) of Belle this year, but it can read out the detector at much higher rates. One prominent example is the search for so-called dark photons: Particles that are produced like photons but then don't interact with the detector. The collisions could produce one regular photon and one dark photon, back to back and with a predicable energy. The regular photon is detected and nothing else. The detector is hit by photons often, you need a high readout rate to capture every event with a photon in order to check if there are other particles in the event.
Here is a projection on slide 9: Even with just 20/fb, about 2% the Belle dataset, Belle II could improve some upper limits significantly - or find strong evidence for dark photons if they exist somewhere in that range.
  • #4
Can I ask a slightly general question, but it is specifically relevant to this sort of collider?;

When fundamental particles get accelerated into each other, do the laws of quantum physics mean that ANY particles whose mass is less than the collision energy could (however unlikely) end up getting created?

I mean, if you throw two streams of electrons together with a collision energy of multiple GeV energy, then could you find protons and neutrons coming out of the reaction?
  • #5
Most of the particles have to be created in pairs. To conserve baryon numbers (essentially the difference between the number of quarks and antiquarks) a proton (a baryon) has to come with an antibaryon - this can be an antiproton, it can also be an antineutron plus some other particles, and some more options. Similar for a neutron.

SuperKEKB has enough energy to produce baryon/antibaryon pairs if they are not too heavy. Proton+antiproton or neutron+antineutron is not a very common result in electron/positron collisions but it does happen. There are also proton + antineutron + negative pion, neutron + antiproton + positive pion, proton + antiproton + neutral pion(s) and many more reactions.


There was a fire in the same building as SuperKEKB's linear accelerator. The accelerator was not involved in the fire but soot from the fire reached the accelerator components, that lead to some downtime to clean it.

Based on the live status it is running again and taking data with collisions. The luminosity is about 1033/(cm2s), it still has to improve a lot.
  • Like
Likes vanhees71 and Greg Bernhardt
  • #6
Thanks, yes, I can see the reason for the anti-particles to be produced as a pair with the particle.

That's quite mind blowing, really, that any sort of matter can be produced out of 'nothing else' just by energy. Makes me feel like the 'background' is actually the 'real' stuff of the universe, and it's the matter coming in and out of existence which is the actual fluctuation! Humbling!
  • #7
I don't understand this conclusion, particularly what you might mean by "just by energy". At the here discussed accelerator electrons and positrons are interacting not "just energy" (whatever this might be).
  • #8
I mean that the collision energy itself, if sufficient in magnitude, could produce any form of particles, not simply something related to the original particles carrying the collision energy.

Is that wrong? Are you saying there are limits on the types of particles such a collision might manufacture?
  • #9
Well, particles need an interaction with the original particles to be produced, at least indirectly. If there are e.g. truly sterile neutrinos then accelerators cannot produce them. Apart from that energy and conservation laws are the only limits. SuperKEKB's energy is optimized for the production of B mesons.
  • #10
The accelerator is making progress towards 24/7 data-taking. Two important differences to the LHC: The lifetime of the beam is much shorter, but particles are directly injected at their final energy - it is possible to refill the accelerator while there is still beam in the machine (the LHC has to discard the beam, ramp down the magnets, accumulate new particles, ramp up the magnets, and so on).
Injecting new particles is a dirty business, however - it leads to more particles hitting the beam pipe and other effects, increasing the background in the detector and elsewhere. Too much background and detector components can get damaged. As a safety measure the high voltage of some detector components can be ramped down. This is done at the LHC experiments, too*. It takes time, however. To avoid that SuperKEKB can do something called, similar to the predecessor KEKB: New particles are injected in many small steps while collisions are ongoing. Instead of the sawtooth pattern seen before the beam currents stay nearly constant and the detector can take data continuously. It looks like they use this for the positron beam now, but not always for the electron beam. The positron beam is the more difficult one (larger losses, needs injection more often), so this should help with the collected data already. Current status,

Luminosity: Up to ~2.5*1033, still a long way to go.

*LHCb does more than that: Their innermost detector, VELO, has two movable halves. While the particles are injected and accelerated they are retracted, a few centimeters away from the beam. Only after stable collisions are achieved they are moved in, to a position where they are just millimeters away from the collision point.
  • Like
Likes dukwon and vanhees71
  • #11
I found an update on their Facebook page. 5.6/fb integrated luminosity, with a record of about 0.3/(fb*day). Still a bit away from this goal, but making progress:
Even with just 20/fb, about 2% the Belle dataset, Belle II could improve some upper limits significantly - or find strong evidence for dark photons if they exist somewhere in that range.
Based on the daily overview: The current luminosity is about 4.5*1033/(cm2s), doubled relative to early May. Both beams use continuous injection now.
  • #12
Belle II finished its first run in 2019. Three weeks interruption from the fire, another week from magnet issues. The detector collected 6.5/fb - less than expected, due to the down-times. Might still be enough for first interesting physics results in the next weeks to months.

The luminosity reached 1.2*1034/(cm2s), that is 20% higher than the KEKB design value and ~40% lower than the KEKB record. It is 1.5% of the SuperKEKB design luminosity.

The next run will begin mid October. Ramping up the collision rate will be one of the main tasks. We might see the accelerator break the luminosity record of the predecessor, and then set a new absolute luminosity record (as the LHC matched the KEKB record).
  • #13
Belle II has collisions again. The highest luminosity I found was 2.5*1033/(cm2s), it is still early in this run. Collisions seem to happen mainly in the night while machine development is done during the day. This is different from the LHC where machine development has dedicated days.
  • #14
Belle II reached a luminosity of 1034/(cm2s)! It is one out of just five experiments in history to reach this milestone. The others are BaBar, Belle, ATLAS and CMS. A factor ~2 to go to the absolute record, a factor 80 for the design value.

  • Like
Likes Greg Bernhardt, berkeman and Lord Crc
  • #15
End of the 2019 run. Accelerator and detector will restart in early 2020. The maximal luminosity reached was about 1.9*1034/(cm2s), but as far as I understand this was the accelerator alone without the detector taking data. 2.1*1034/(cm2s) is the record of both KEKB and LHC, not much missing until SuperKEKB/Belle II take the record. They should then go to luminosity values not even the high luminosity LHC can reach.
  • Like
Likes atyy and Greg Bernhardt
  • #17
Huh, forgot about this thread.

After a winter break SuperKEKB/Belle II restarted operations in 2020 in March. The luminosity is now a bit above 1034/(cm2 s) (, at the level of the 2019 record.
  • #19
SuperKEKB/Belle II set the absolute luminosity record, surpassing KEKB/Belle and the LHC.
Press release

Now the record is ##2.4 \cdot 10^{34}~ \mathrm{cm}^{-2}~\mathrm{s}^{-1}##.

This is just 3% of the design luminosity, so we can expect that number to grow a lot more over the next years.

We will probably get a few early physics results at ICHEP in July/August (virtual conference).
  • #20
The luminosity is about 1033/(cm2s)
Hi mfb:

I am confused by the units here. Wikipedia gives the following alternative defintions for luminosity.
In SI units, luminosity is measured in joules per second, or watts. In astronomy, values for luminosity are often given in the terms of the luminosity of the Sun, L⊙. Luminosity can also be given in terms of the astronomical magnitude system: the absolute bolometric magnitude (Mbol) of an object is a logarithmic measure of its total energy emission rate, while absolute magnitude is a logarithmic measure of the luminosity within some specific wavelength range or filter band.​
How does cm-2s-1 fit in?

  • #22
After a summer shutdown Belle II started taking data again.

Here are some public results from the summer conferences. Nothing revolutionary, the experiment doesn't have enough data for that yet. The talks largely focused on demonstrations that the detector performance is understood, that it can measure things more precisely than its predecessor and so on - basically showing that it will produce great results once the datasets are larger.
  • #24
  • #25
There are still several issues with the background conditions - the nanobeam scheme is more complicated than expected. The achieved peak luminosity is ~3.8E34 or so, which is close to twice the KEKB value, but it's something they originally expected to reach and exceed quickly. At that luminosity you can collect ~2-3/fb per day or a few hundred per year, that's not contributing notably to the 50/ab goal.
The luminosity target has been lowered and it won't be reached without another longer upgrade ~2026-2027.
  • Like
  • Sad
Likes vanhees71 and Wrichik Basu
  • #28
Unfortunately, today is the last beam day of Belle II/SKB before the LS1 due to the electricity costs.
In fairness, the fully completed Phase III has run for a long time as these things go (March 25, 2019-June 22, 2022), and a new or upgraded accelerator's benefits are front loaded.

After a while, you are simply shaving a little bit of statistical uncertainty of your results is not proportional to sample size, so even a lot more data only produces modest improvements in precision.

At some point, there is no good substitute to building a better experiment to reduce systemic error or better target a specific unresolved experimental issue, rather than simply increasing sample size with old equipment.

Also, of course, there are more costs than electricity to continued operations. The personnel costs are not cheap, because you need a highly skilled workforce to operate it and to analyze the data.
  • #29
Cancelling a week of operation is not accelerating machine upgrades (the planned stop was the end of this month). The sudden stop is limiting the understanding of how the accelerator behaves at higher beam currents, however, knowledge that might have found its way into the next upgrades.

The accelerator set a new luminosity world record of 4.7*1034/(cm2s) shortly before it was switched off. Higher than twice the KEKB record.
  • Like
Likes arabusov and vanhees71

Suggested for: Belle II collisions in 2021