jump to navigation

Two other talks from yesterday afternoon’s session May 22, 2008

Posted by dorigo in news, physics, science.
Tags: , , , ,

Here I give some highlights of two experimental talks on neutrino physics heard yesterday at the afternoon session (May 21st) at PPC 2008.

Zelimir Djurcic discussed “Results from MiniBoone“.

The LSND collaboration observed a long time ago an excess of neutrino interactions from a proton beam, due allegedly to pions decaying to muon antineutrinos, when the latter oscillated to electron antineutrinos. The signal corresponded to a 3.8-sigma significance of the oscillation. However, when compared to the situation that was being observed in measurements in atmospheric and solar neutrinos, in the plane of the two parameters describing the oscillation \Delta m^2 versus \sin^2 2 \theta, the solution of LSND looked very different.

Results were explained by models with one or more sterile neutrinos, but there would be more mass eigenstates. Confronted by other measurements, another neutrino experiment, Karmen , had smaller sensitivity and could not cover the LSND solution. Also Bugey, a reactor experiment, excluded a fraction of LSND solution but not all the region of parameters space allowed by it. So the anomaly remained in the air for a while, until Miniboone was designed to confirm or refute the LSND signal.

A 8 GeV booster is the proton source in the Miniboone experimental complex. Extracted protons hit a beryllium target, the secondary mesons produced in the interaction are directed to Miniboone, and they produce a virtually pure beam of muon neutrinos. So one can look for \nu_\mu \to \nu_e oscillations by detecting electron neutrino interactions.
Miniboone has a similar ratio between distance from the source, L, and typical neutrino energy E, as LSND, with L=540 meters, 15 times LSND, and energy 10-20 times higher. The equality of L/E means that the possible signal is the same in the two setups, but the different distance and energy means that the two apparata suffer from different systematics. Then, the idea was to collect more statistics than what LSND could get. So far Miniboone collected 5 times more data than LSND.

The beam has a low \nu_e contamination, 0.5% of the total. Miniboone is a vessel 12 m in diameter, filled with 950,000 liters of CH2 oil, seen by 1280 photomultiplier tubes (PMT) plus 240 additional PMTs on the external region providing a veto for background tracks.

The detector was built to measure the position of the vertex, and the energy of interacting neutrinos, and to be able to separate events from \nu_\mu and \nu_e [DUH]. Cerenkov rings produced by the charged particle motion in the liquid provide a means for identifying products of neutrino interactions in the detector.

For \nu_\mu you get multiple Cherenkov rings concentric to one another. For electrons, they do not travel a long distance, and since they scatter, they produce a fuzzy ring. If instead a \pi^0 is produced, the two gamma from its decay produce rings corresponding to two electrons emitted from the pair production in the material.

Zelimir showed a slide that would have been an animated gif showing the detector after a ring from \nu_\mu, which is idle for 2 microseconds, then the muon decays and one sees a ring from the emitted electron. Instead, it was a rather dull view of the detector, since the file had been converted into a PDF. [This reminds me that thanks to a lucky chance my own slides, which I had been unable to transform into a PDF file, stayed in PPT format and were projected as such during my talk. I could thus show my own animation of the incredibly shrinking B_s \to \mu \mu branching ratio limit, and the correspondingly shrinking parameter space of SUSY… A reason to cheer up after the fact.]

In Miniboone background predictions show that these mainly come from two major classes: one is intrinsic electron neutrinos, the other comes from misidentified muon neutrino interactions. Events with \nu_e selection requirements in energy between 475 and 1250 MeV -the selected search region- are 358 total background interactions. The oscillation signal from LSND would amount to 163 events in this situation with the LSND solution, \Delta m^2 = 0.4 eV^2, \sin^2 2 \theta = 0.017. 5.6x10^20 protons on target in neutrino mode were used in the analysis [Miniboone can also run with antineutrinos from the beam].

The analysis was done for energy above 475 MeV. It did not observe a LSND-like excess of events. The result is thus a limit in the plane of the two relevant variables describing the potential oscillation, and the limit is covering the whole LSND-allowed region.

Zelimir then showed how it is possible to combine the result with all the other experiments. One can thus get a measure of the probability that all results come from the same underlying oscillation hypothesis. This work has been described recently in hep-ex/0805.1764: the maximum compatibility is 4%.

A discussion followed about the low-energy excess found by MiniBoone: yes, because MiniBoone extracts the limit on oscillations by selecting a range of energy of the interacting neutrino where their data closely follows the simulation of backgrounds. Below 475 MeV the data instead overshoots the background predictions by a large amount. The problem appears to be with the misidentified \nu_\mu events if it is not a genuine signal, because \nu_e contamination is not expected to be varying much at lower energy.

Photonuclear absorption of photons from pizero decays was found to be a source of events at low energy, capable of explaining away less than 30% of the effect. In fact, any process that creates a single gamma will be a background because Miniboone cannot distinguish single gamma from electrons. The process of photonuclear absorption of one of the two photons in a pizero decay explains some of the discrepancy at low energy. The process is \gamma+N \to \Delta \to \pi+N: photons are absorbed mostly at 350 MeV energy, by making the the giant dipole resonance.

Other checks, including taking data off-axis from another beam, are being carried out. Also, data from a antineutrino beam can help understand the origin of the low-energy excess. What appears established already, however, is the ruling out of the LSND solution.

After the MiniBoone talk, Puneet Batra discussed a new proposal of a neutrino experiment at Fermilab: NuSoNg.

The idea is to do neutrino scattering on glass, a proposed fixed target experiment at FNAL to study Tev-scale physics in neutrino scattering. It foresees the use of 800 GeV protons from the Tevatron, with high statistics, mainly focused on leptonic processes. The neutrino flux is expected to be normalized by inverse muon decay.

\nu_\mu - e elastic scattering depends on the handedness of the incoming bodies. No competition from QED: the diagram responsible is a Z-boson exchange only. It is a very clean process, well predicted by electroweak theory. Indeed, it was one of the first calculations made. A fantastic probe of TeV physics. A Z’, if it exists, contributes to the process. Mass scale sensitivity ranges are up to 4-5 TeV on Z’. Also oblique corrections, new physics that corrects the ratio of charged current divided by neutral current cross sections of neutrino interaction. Mixing with sterile neutrinos will also show a decrease of cross section. The NuTeV anomaly could get explained or measured.

The total cross section depends on Z couplings and on the neutrino energy. The main problem is normalizing the neutrino flux. Typically this is done through deep inelastic scattering, which leads to large systematic uncertainties: 5% or larger, because one needs to know the parton distribution functions.

One can take the ratio of elastic scattering of \nu_\mu e versus \bar \nu_\mu e scattering. This is a function of \sin^2 \theta_W. The sensitivity to \rho is removed entirely, while we would like to retain it. So NuSong can measure both \rho and \sin^2 \theta_w. We do inverse muon decay with W exchange. This gets a muon out. The ratio \sigma(\nu_\mu,e)/\sigma(\nu_\mu,\mu) is sensitive to both \rho and \sin^2 \theta_w. The expected accuracy in the ratio is 0.7%, an order of magnitude improvement.

The experiment is based on a well-segmented massive detector (6 times the total mass of the old Charm experiment). Event numbers from NuSoNg from a few years of running would be an order to two orders of magnitude more than those of NuTeV.

After the talk (see, I am a polite listener and I do not usually interrupt) I asked Puneet the timescale and funding status of the experiment. It seems that funding, approval, and construction might take five years, and five more years of data taking would be necessary to achieve the design sensitivity on the mentioned physics processes. The experiment is not funded yet, and a lot seems to depend on political decisions on what to do with Fermilab once the Tevatron run stops. Of course, the more NuSonG can integrate with the other programs foreseen at the Tevatron (e.g. projectX), the better chances it has to be given green light. Good luck – we all wish that Fermilab will stay there and continue to be a wonderful place where to do fundamental research!

%d bloggers like this: