CMS puts new constraints on Dark Matter
CMS recently released a paper on the search for monojets (arXiv:1206.2664 25-June-2012). Normally one thinks of monojets (events with one energetic jet and large missing energy or “MET'') in the context of supersymmetry or maybe large extra dimensions. But this paper follows a treatment first published by CDF (arXiv:1106.4775 23-June-2011) in which the monojet search is reinterpreted as a limit on the production of dark matter particles in hadron colliders. Credit for this idea should probably go to Beltran, Hooper, Kolb, Krusberg, and Tait (arXiv:1002.4137 22-Feb-2010) among others.
Setting aside supersymmetry, one can expect dark matter particles to be produced in pairs at the LHC or the Tevatron if they have weak-scale couplings to standard model particles. Crudely put, producing DM particles would be a lot like producing a pair of heavy neutrinos (with minor modifications of the DM particles are not fermions; CMS assumes they are spin-1/2 fermions). It makes sense to calculate production cross sections in a generic framework – perhaps even taking just a contact interaction to couple a pair of DM particles to matter. In any case, the DM particle pair may sometimes recoil against an energetic jet, and since they themselves leave the apparatus undetected, a nice monojet signature arises.
The kicker is that the acceptance and efficiency for a DM signal hardly depends on the mass of the DM particle, after one has asked for an energetic jet and large MET. This leads to interesting bounds for very light DM particles at masses inaccessible to most direct-detection DM searches.
Events were selected in a straight-forward way: ask for a good-quality jet consistent with the primary vertex. The jet should have pT > 110 GeV and |η| < 2.4. The MET should be at least 250 GeV. Any second jet should not be back-to-back with the leading jet. Events with leptons or isolated, energetic tracks are vetoed.
The contamination is mainly Z→νν, as one would guess. It is monitored and estimated using Z→μμ events: one predicts 5106±271 events. There is also significant contamination from W+jet events; again this contamination is estimated by finding W→μν events: one predicts 2632±237 events. Much smaller contaminations come from top and Z→LL events. The total number of events expected from SM sources is 7842±367, and CMS observes 7584 events. No signs of DM particle production. Systematic uncertainties are at the 20% level.
The non-observation of monojets implies upper limits on the contact interactions (or whatever theoretical framework you want to use), that can be translated into upper limits for DM-nucleon scattering. Here is the result:
Similar results were obtained by CDF, as shown in the plot. Also, the CMS results on photon+MET give results not quite as stringent as those obtained from the monojet search.
I like this result because it nicely illustrates the synergy between collider and direct-detection searches for dark matter particles. What we would really love to see, of course, is a signal in both. Alas, that has not yet happened…
Entry filed under: Particle Physics.