Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How would we find the Higgs if Higgs did not coupe to fermions?

  1. Dec 1, 2006 #1
    I'm pretty new to particle physics. Actaully, I'm brand new to particle physics (2nd year undergraduate). I've been invited into a course on the Higgs recently and have a few questions I was wondering about.

    I was wondering what would happen if Higgs did not couple to fermions? Does this mean we cannot see the higgs decay from photons, electrons, muons, qurks and gluons? How would we find the Higgs if Higgs did not coupe to fermions?

    And I was wondering why gluon fusion is the dominant Higgs production mechanism in both the LHC and Tevatron?

    And what are "channels for Higgs discovery at Tevatron"? I'm more curious as to what 'channels' is refering to. And what are "backgroup processes"?

    much thanks.
  2. jcsd
  3. Dec 1, 2006 #2
    Just answering one question, since I'm hardly qualified to answer all of them.

    In this case "decay channels" generally just means the way in which the particle decays. For instance, a particle that decays into either electrons or muons would be said to have two channels (or possibly one - in some parlance all leptons are considered to be identical). It really means that you can look for them by searching for electrons or for muons.

    There is at least one model of the "Fermi-phobic" Higgs, where the Higgs does not couple to fermions. Such a Higgs would, however, couple to bosons. This leaves several important channels, Higgs->ZZ, Higgs->WW, and Higgs-> photon photon. Because the Z and the W then decay into fermions, we can still see the Higgs. These searches will be carried out at the LHC as an extension of the Higgs->ZZ and WW searches (these decay channels are very important for certain masses of the standard Higgs).

    Unfortunately, I don't know more details. I also only have a rather vague idea of why gluon fusion is dominant, so I can't really answer that either. Hopefully someone with more information will be along shortly (and will fill in the gaps that I've left behind).
  4. Dec 2, 2006 #3
    The H-> photon photon would be rather difficult, since the main mediator is via a top loop (it can go by a W loop too though). Especially since the main production via gg->H would also be reduced.

    Gluon fusion is so dominant mainly because the LHC is basically a gluon collider. At high energies most of the proton momentum is carried in gluons, so events are mainly gluon collisions. Then gg->H takes place via a top loop, so it is loop suppressed, but the large gluon luminosity and the large coupling of a Higgs to a top quark make up for it. Then there is also the bonus that you don't need extra energy to make something to go along with the Higgs - ie. you only make a Higgs, unlke say qqbar -> HZ where you need energy for the Z too.
  5. Dec 2, 2006 #4
    I did forget about the top loop problem.

    But would the W loop work for Higgs Mass ~ 120? At that point isn't the offshell H->WW really disadvantageous?
  6. Dec 3, 2006 #5
    What are "backgroud processes"? Are those ways of reducing the backgroud noise so the higgs can be more easily detected? and what are current background processes that are used?
  7. Dec 3, 2006 #6
    Yes, but it is still less off-shell that top quarks would be. The difference here is the small HWW coupling compared to the large Htt coupling.

    Since the Higgs decays rather quickly, you only see its decay products in the detector. These decay products can be produced in other ways - these other ways are the 'background'. For example gg->H followed by a Higgs decay to b-quarks (H->bb) may result in 2 tagged b-jets in the detector. But straight QCD without a Higgs can also produce this, so how do you tell one from the other. In principle, one could measure the b-pair invariant mass and see it there was a peak, but in this case, the b-jet background is so huge that you can't even trigger on it (ie. there are so many events like this that you can't record them fast enough).
  8. Dec 3, 2006 #7
    i'm not totally use to the termanology yet. what are "2 tagged b-jets"?

    so the background is the straight QCD without a Higgs that also produces 2 tagged b-jets?

    you said "b-pair invariant mass and see it there was a peak" what does the peak show?

    and what is the b-jet backgroud?
  9. Dec 3, 2006 #8
    But since a fermi-phobic Higgs would not have an Htt loop, it's still possible that H->gamma gamma would be the dominant decay mode for a low mass Higgs. That entire channel is a shot in the dark anyway, but maybe they can find it.
  10. Dec 4, 2006 #9
    Sorry, I have the tendancy to make assumptions. 'Jets' are what happens to quarks coming out of interactions. They turn into hardons like pions or kaons, or even protons and neutrons because of their string interaction, and this shower of particles is called a 'jet'. It is tagged if we can identify it as a b-quark originating jet. This is done for b quarks by looking for 'displaced vertices' - basically the B-meson (which the b-quark becomes part of) lives a long time before decaying, so we see some of the particles in the jet (ie. the B decay products) originating all from a point (or vertex) slightly displaced from the original interaction point.

    For this one channel yes. For H->gamma gamma there is a different background.

    The Higgs is most likely to be produced with an invariant mass (the modulus of its 4-momentum, or [tex]\sqrt{E^2/c^4-p^2/c^2}[/tex]) close to its mass. invariant masses further away are increasingly unlikely, so you expect to see most of the events with this invariant mass and the signal is much bigger.

    The process [tex]gg \to b \bar b[/tex] via QCD.

    [tex]H \to \gamma \gamma[/tex] was never the dominant decay mode. At best (in non-fermiophobic models) its branching ratio is 10-3. It is only a good channel to look in because it is so clean. All you produce is two photons and nothing else, which has a low background. To do this you need to produce it with [tex]gg \to H[/tex].

    Now in fermiofobic models, most of the decay widths will shrink, but H->ZZ* or WW* will not, so the branching ratio for [tex]H \to \gamma \gamma[/tex] will get smaller. Even worse, you lose your production mechanism [tex]gg \to H[/tex], and it can't be done with anything else, so I think you are stumped.
    Last edited: Dec 4, 2006
  11. Dec 5, 2006 #10
    I've got more questions. what is the difference between pp coliders and ppbar colliders? What is the branching ratio?

    So gg->H is the most promising channel for Higgs discovery at the tevatron and LHC right? but if the Higgs was fermiofobic, the most promissing channel would have to be the H->ZZ* or WW* channel right?

    And what are some things that people use to beat down the background?
  12. Dec 6, 2006 #11
    The buggest difference is that protons are easier to make so pp colliders have higher luminosity (at TeV energies). At these energies, they are mainly gluons so it doesn't really matter what the valence quarks are.

    That sounds reasonable.

    The background events tend to have a slighly different topology - their particle come out with particular momentum of particular directions. Therefore it is usual to discard events which 'look like' the expected backgrounds. This is called a 'cut'. Since we can predict the background pretty well, any statistical excess over what we expect in other areas is signal.

    Also, for [tex]gg \to H \to \gamma \gamma[/tex] one can do a sideband subtraction. The background is constantly falling with the invariant mass of the two photons. The signal will be a slight peak in this falling spectrum where the Higgs mass is. One can plot a straight line from just before this bump to just after and anything above the bump is signal.

    Take a loot at this plot:
    http://wlap.physics.lsa.umich.edu/atlas/computing/workshops/2003BNL/20030828-02-mellado/real/slides/img013.gif [Broken]
    (There are much better plots than this but I am in a hurry...)
    Last edited by a moderator: May 2, 2017
  13. Dec 6, 2006 #12
    so what sideband subtraction means is taking the known predicted background and then subtracting it from the experimental?

    are they any other ways to make these cuts other than a sideband subtraction?

    and what exactly are pp coliders?
    and what are ppbar colliders?
  14. Dec 6, 2006 #13


    User Avatar

    Staff: Mentor

    They are proton-proton colliders and proton-antiproton colliders. p = proton, pbar = [itex]\overline{p}[/itex] = antiproton.
    Last edited: Dec 6, 2006
  15. Dec 6, 2006 #14
    why are two different ones used?
  16. Dec 6, 2006 #15
    also, I was wondering what are the perfered Higgs decay as a function of its mass?

    I know that H->gg, H->photon photon, H->ZZ, H->WW but how does that vary with mass? And are there any more decays?

    oh and why is the H->photon photon so important at the LHC? I read that it has a branching fration of up to 2x10^-3. what does this all mean? why is H->photon photon so important?
    Last edited: Dec 6, 2006
  17. Dec 7, 2006 #16
    To create new particles directly it is often useful to collide particles with antiparticles, so the Tevatron is a proton-antiproton collider. But at higher energies, eg. LHC enegies, the protons are mainly gluons anyway, so it doesn't really matter whether it is a proton or antiproton, and protons are much easier to produce.

    It basically decays into the heaviest thing allowed. So a low mass Higgs decays into bottom quarks and once it is heavy eneough to decay to W bosons it does. A really heavy Higgs is heavy enough to decay to top quarks and then that is the dominant decay mode. (Of course, slightly below threshold it can decay to virtual particles, so the different decay modes don't have sharp edges.) Have a look at:


    Edit: How do I embed images at this site?

    It is because it is so clean - ie. has small backgrounds.
  18. Dec 8, 2006 #17
    I just have a few general questions after reading more about the Higgs:

    I am wondering about the dominant Higgs production mechanisms. I think they both are gg->H for pp and ppbar colliders. And if Higgs were fermiphobic, I think that mechanisms left are the vector boson fusion and higgs-strahlung processes. My question is which one is more dominant for pp/ppbar colliders? I have a guess that vector boson fusion is better for the LHC while higgs-strahlung is for TeVatron. This is because i see that vector boson fusion starts with qq while higgs-strahlung starts with qqbar. It's a wild guess, i was wondering if there was any justice in what I just wrote.

    Also, which channels are the most promising channels for Higgs discovery at the TeVatron or LHC. I'm guessing that H->ZZ is best for the TeVatron and I'm not sure for the LHC.
  19. Dec 9, 2006 #18


    User Avatar
    Science Advisor

    "Also, which channels are the most promising channels for Higgs discovery at the TeVatron or LHC. "

    Thats very much a mass question. The most promising channel(s) varies depending on how heavy the Higgs is, and theres a long laundry list of channels per model per mass per detector that phenomelogists have to work through. Quite a nightmare actually.
  20. Dec 11, 2006 #19
    You can see the typical Higgs searches for ATLAS on this plot:

    Sorry it isn't a very good image (there must be a better one somewhere but I can't find it to link to). Above the dotted line would be a discovery at 5 standard deviations for 30 inverse femptobarns of data (about one year of high luminosity running). This plot is actually rather old now (it was in the ATLAS Technical Design report). A newer version can be found on page 10 of:

    For those interested in the H-> gamma gamma, I found a better plot.
    http://ihp-lx2a.ethz.ch/CompMethPP/lhc/pictures/higgs_massplot1.jpg [Broken]
    The red is signal, blue is background and you can see quite clearly that the blue drops continuously, so you should be able to disentangle the two quite easily.
    Last edited by a moderator: May 2, 2017
  21. Jan 9, 2007 #20
    Several Higgs fields?

    In some versions of the Standard Model several Higgs fields are considered. Please, can you give me references to some papers with such extensions of the SM. References to on-line resources in arXiv.org are preferable.

    Best regards, Ruslan Sharipov, Ufa Russia.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook