Search for associated Higgs boson production WH->WWW*->l^±nu l'^±nu' + X in ppbar collisions at sqrt(s)=1.96TeV

The original paper: hep-ex/0607032

The most fundamental question in the modern particle physics is origin of mass. The standard model that describes the fundamental particles that make up the matter around us and the forces that hold these particles together, tells us that mass arises due to a so-called Higgs mechanism, named after its inventor, Peter Higgs. This mechanism implies existence of a new particle (Higgs boson) which currently remains the only fundamental particle in the standard model that has not yet been observed. Ironically, for a particle that is a quantum of a field that gives mass to all particles, the theory does not predict its mass. From the experiment, we know that the Higgs boson cannot be too light (like W or Z boson) -- otherwise it would be already discovered. It is also improbable that the Higgs boson is very heavy (like top quark), because this is unfavored (but not ruled out) by measurements of other quantities that are related to the Higgs mass.

At the Tevatron, the world highest-energy particle accelerator the Higgs bosons must be produced in plenty. But this does not mean that the Higgs boson can be easily observed. As with many other particles, the Higgs boson is unstable and quickly decays into other particles, so the way to observe it is to look at the products of its decay. The way the Higgs boson decays depends on its mass. A "low mass" Higgs boson (less than 135 GeV¹) predominantly decays to a pair of b-quarks, while a "heavy" Higgs boson (with mass greater than 135 GeV) mostly decays to a pair of W (and with lower probability, Z) bosons.

Figure 1. The observed number of events (solid lines), the predicted background (shaded bands), and the expected number of signal events times 100 (dashed lines) for Higgs mass 155 GeV above the TLD cut in the (a) ee, (b) eµ, and (c) µµ channels.

Most of the Higgs bosons at the Tevatron are produced directly (i.e. without any accompanying particles). Unfortunately it is very difficult to separate these events from (much more frequent) cases when a pair of b-quarks or W-bosons is produced in a process other than Higgs decay ("non-resonant production"). More promising is an approach to search for events due to Higgs bosons produced in association with a W (or Z) boson. While the production rate for these events is several times lower than for the direct production, the accompanying W boson provides a "handle" that makes it convenient to separate the signal events from the background (although for low mass Higgs the non-resonant W+bb production is still non-reducible background).

In the present analysis, we are searching for associated Higgs boson production where the Higgs bosons decay to W pairs. In the end, there are three W bosons in an event. This process is unique in a sense that (considering the situation when two of W bosons of the same charge decay to a lepton and neutrino), the final state is characterized by two high momentum isolated leptons of the same charge. Very few physics processes lead to this signature, most prominent ones being the di-boson (WZ and ZZ) production. The non-resonant WWW production is also possible but has much lower rate than the Higgs signal. The real challenge in this analysis appears to be instrumental background, i.e. events that look like signal but are there due to various technical reasons. An example is a Z->e⁺e^- process where one of electrons emits a photon which undergoes a conversion (produces an e⁺e^- pair). In the end, there are two high momentum electrons with the same charge that can mimic the signal. Another source of background is "charge flips" (particle charge misidentification) related to the fact that the charge determination becomes less reliable as the muon (or electron) momentum increases (the charge is measured as a sign of the curvature of the particle trajectory moving in a magnetic field, and for very energetic particles the trajectories become indistinguishable from straight lines).

Figure 2. The observed upper limits for the four mass points along with the theoretical predictions for the standard model and fermiophobic Higgs boson production.

A search for the Higgs boson in the WH->WWW*->l^±nu l'^±nu'+X mode has been performed at the D0 detector at the Tevatron using data collected between August 2002 and August 2004. This is the world first published analysis on the Higgs boson search in this mode. We consider three final states: e^±e^±, e^±µ^±, and µ^±µ^±. After initial event selection (two high momenta isolated leptons) we are left with 34 events, mostly due to instrumental background. To further distinguish between the signal and background, we use the technique known as topological likelihood discriminant (TLD). The dependence on the number of events remaining after a TLD cut vs minimal TLD value is shown in Fig.1. Shown with arrows are optimal cut values that provide the best expected production cross section upper limits.

The results (observed upper limits for four Higgs mass points along with the theoretical predictions) are shown in Fig.2. Unfortunately we still cannot observe (or exclude the existence of) the Higgs boson. We need more data, and in the end we will have to combine this search with results in other search channels (and with results from our colleagues from another Tevatron detector, CDF). This study is just the beginning.

If you have questions regarding this analysis, please contact Alexander Khanov.

¹ GeV is a unit of energy (and mass in units where speed of light equals 1) in high energy physics. The proton has a mass of approximately 0.938 GeV.