Gravitons are heavier than 500 GeV! December 23, 2008Posted by dorigo in news, personal, physics, science.
Tags: CDF, graviton, Higgs boson, Z boson
About a year ago I reported here on a search performed by CDF for events featuring two Z bosons, both decaying to electron-positron pairs: I had been an internal reviewer of that analysis, and I discussed it in some detail after we approved it for publication. While the standard model expectation for electroweak production of two Z bosons is of about 1.5 pb, and the process has indeed been put in evidence in CDF and D0 Run II data, the analysis was rather focused on a search for heavy mass resonances decaying to the ZZ final state: new physics, that is, either in the form
of a heavy Higgs boson, or of a graviton (in the Randall-Sundrum scenario), or other still fancier (and improbable) beasts.
CDF has now repeated that search by increasing the dataset size by a factor of three, and by including mixed final states which include muon pairs and even jet pairs. This makes the analysis intrinsically interesting to me, since I have started a similar analysis with the CMS experiment, together with a PhD student in Padova, Mia Tosi. Mia and I will be looking for Higgs bosons in the dilepton plus dijet final state, with particular emphasis on the decay, which is a signal with which we have quite some familiarity.
The new CDF search for high-mass ZZ events configures itself as a “signature-based” one: despite the reference to the Randall-Sundrum graviton, the analysis cuts are kept generic, such that a signal can be found for anything that decays to two Z bosons, and in case no signal is seen, a model-independent limit on the cross section can be set. The only limitation of the search is that the four-body mass is studied only above the minimum value of 300 GeV. Such a requirement allows to steer away from phase space regions where backgrounds dominate.
Once four objects (electrons, muons, and jets, with the specification that at most two jets are present) are selected with loose cuts, a statistical estimator is built to test the hypothesis that they originate from the decay . It is a simple function, which utilizes the expected resolution on the two two-body masses and the resulting four-body mass to estimate how much the event departs from the tentative signal interpretation. Only in the case of jet pairs, an explicit cut is set on the dijet mass to lay between 65 and 120 GeV, to avoid accepting too many random jet combinations.
While the region is the one where the signal is sought, the complementary region of the four-body mass is used as a control sample, to verify that background estimates obtained with Monte Carlo simulations are in agreement with the observed data. The nice thing about such a spectacular signature as the production of two Z bosons is that backgrounds are exclusively of electroweak nature: by having at least one decay in the final state, the signal cannot be mimicked easily by purely quantum chromodynamical processes, which plague most hadron collider searches with high rates. Besides regular pairs from standard model processes, backgrounds include WZ, WW, and Z+jets production. At high four-body mass, however, all of these are really small, and even in the 3 inverse femtobarns of proton-antiproton collisions analyzed by CDF for this search, they contribute only few events; only the dilepton+dijet signature accepts a few hundred events, because of the large cross-section of Z+2 jet production processes.
In the end, no signal is seen, and a cross-section limit is extracted as a function of the X mass. The limit is shown below, compared to the expected cross section for graviton production and decay to the ZZ final state. The comparison of upper limit (the red curve) with the theory hatched line allows to exclude gravitons with masses below 491 GeV, for a particular choice of model parameters (k is a warp factor for the extra dimensions, and is the Planck mass).
As a by-product of this analysis, a new set of excellent standard-model-like ZZ decay candidates have been selected. I am unable to show any of the new event displays here, because they have not been approved for public consumption by CDF yet… So please see the lego plot of a candidate below, extracted last year by the same authors. The two pairs of electrons make masses very close to that of the Z boson, as evidenced by the two pink numbers.
To read this graph, you have to know that the greek letter is the pseudorapidity, basically a function of the angle that particles make with the beam axis. A pseudorapidity of zero means that the particle is emitted at 90 degrees from the beam, while positive and negative values indicate the proton and antiproton directions. The other coordinate, , indicates the azimuthal angle in the transverse plane. The z axis (the height of the bars) indicates how much energy is deposited in the interval span by the bars. In bright pink are shown the four electron candidates, as measured by the CDF calorimeter, and each bar is labeled by the energy in GeV measured for each.
I am only left with the pleasant task of congratulating my colleagues Antonio Boveia, Ben Brau, and David Stuart for this new result, which greatly extends the scope of the analysis I have reviewed last year. During my review I had encouraged them to pursue the other decay modes of ZZ pairs, and so they did. Well done, folks!