aa r X i v : . [ h e p - e x ] O c t OPERA: waiting for the τ Andrea Longhin Physics Department “M. Merlin”,Bari Universityvia Amendola 173,I-70126 Bari, Italy
OPERA[1] is a long baseline (730 Km) neutrino oscillation experiment located inItaly at the Gran Sasso underground laboratory (3500 m.w.e. overburden, residual µ flux ∼ − m − ). The detector is conceived to observe ν µ → ν τ oscillations inthe parameter region indicated by Super-Kamiokande[2] through direct observation(appearance) of ν τ in an almost pure ν µ beam (contaminations: ∼ ν µ , . ν e + ν e and negligible ν τ ). The CERN Neutrinos to Gran Sasso (CNGS [3]) highenergy neutrino beam ( h E ν µ i ≃
17 GeV) has been designed in order to maximizethe possible number of ν τ charged current interactions at destination taking intoaccount the energy dependance of the oscillation probability and the τ productioncross section.The OPERA detector is a massive (1.35 kton) and highly modular lead-nuclearemulsion target composed of 154750 units called Emulsion Cloud Chambers (ECCsor “bricks”). Each brick is a 56-layer stack of lead plates interleaved with nuclearemulsions providing the µ m and the mrad level precision tracking needed for detectingthe τ decay topology.At CNGS energies the average τ decay length is ∼ µm . ν τ appearance willbe identified by the detection of the peculiar τ lepton decay topology through itsdecay modes into electron (17.8%), muon (17.7%), and single (50%) or three chargedhadrons (14%). In the case of a decay in the same lead plate of production, theimpact parameter of the daughter track with respect to primary vertex can be usedwhile for longer decays in which the τ traverses at least one emulsion layer, the kinkangle in space between the charged decay daughter and the parent direction will beemployed. on behalf of the OPERA Collaboration and confirmed by K2K and MINOS, not to speak about Kamiokande, SOUDAN-2 and MACRO. × ∼
990 tons, B = 1.52 T) instrumented with 22 RPC planes ( ∼ each) which act as inner trackers and six fourfold drift tubes (8 m long) planeswhich provide high precision tracking with a point resolution better than 300 µ m.Precise charge measurement is particularly important for the efficient suppression ofthe charm background. A resolution ∆ p/p < .
25 and charge mis-identification of afew h up to ∼
25 GeV can be obtained.The τ search sensitivity calculated for 5 years of data taking with a total numberof 4.5 · integrated p.o.t./year (200 days runs) is given in Table 1. The numberof signal events essentially scales like (∆ m ) .The main background sources are given by large angle scattering of muons pro-duced in ordinary charged current interactions, hadronic interactions of daughter par-ticles produced at primary interaction vertex and prompt charmed particles decaysassociated with inefficiency on the primary muon identification.Figure 1 shows the probability of discovery at 3 and 4 σ significance as a functionof ∆ m . Signal Signal Bckg ∆ m = 2 . · − eV ∆ m = 3 . · − eV τ → µ τ → e τ → h τ → h ALL ) eV -3 (10 m D D i sc o ve r y p r ob a b ili t y ( % ) s s Discovery probability
Table 1: See text. The shaded band in the plot marks the region indicated by globalanalysis after the recent MINOS determination.The OPERA proposal dates back to 2000, construction started in mid-2003 andthe electronics part was completed by the first half of 2007. Detector filling withbricks started in 2007 and was completed by mid-2008.2
Data taking and results
A summary of the first data taking of OPERA is shown in Table 2.Period (10 p.o.t.) Filling Events (target/external) Comment17-30 Aug 2006 0.76 0 0/319 elect. det. commissioning Oct 2006 0.06 0.1% 0/29 final commissioning
Oct 2007 0.79 40% 38/331 first events in emulsions
Jul-Oct 2008 ∼
10 100% ∼ regular running Table 2: Data taking phases and collected statistics. Full details are given in the text.The first CNGS technical run started in August 2006. Since brick filling had notstarted yet, this run was dedicated to the commissioning of the electronic detectorsand to alignment and reconstruction algorithms tuning. A sample of 319 neutrino-induced events was collected coming from interactions in the rock surrounding thedetector, in the spectrometers and in the target walls. Fine-tuning of the synchroniza-tion between CERN and Gran Sasso, performed using GPS clocks, was also possible.The beam spill timing structure composed by two 10.5 µ s wide bunches separated by50 ms could be clearly observed. The zenith-angle distribution of the muon tracksassociated to beam ν µ interactions in the rock was measured to be centered at 3.4 ± ◦ , in agreement with the value of 3.3 ◦ expected from geodesy. Finally, usinga Monte Carlo simulation tuned on data from the MACRO experiment, angular shapeand absolute normalization of cosmic muons could be reproduced [4].In October 2006, a new run began but was shortly interrupted (0.06 · p.o.t.)due to a leak in the closed water cooling system of the reflector in the neutrino beamline.After repair, a new physics run was possible in October 2007, when OPERA had40% of the target mass installed (about 550 tons). The beam extraction intensitywas limited to 70% of the nominal value due to beam losses which brought severeradiation damage to the CNGS ventilation control electronics. In about 4 days ofcontinuous data taking, 0.79 · p.o.t. were delivered and 38 neutrino interactionsin the OPERA target were triggered by the electronic detectors with an expectationof 31 . ±
6. Out of these 29 had charged-current and 9 neutral-current topology. Out-of-target interactions amounted to 331 events to be compared with an expectation of303.The 2007 run provided the opportunity for the first test on real neutrino interac-tions for the complex chain of brick location, validation, handling, emulsion gridding,development and finally automatic scanning.The essential interplay between electronic detectors and emulsions could also becarefully tested profiting of this initial sample. Position of bricks obtained from3xtensive alignment measurements and mechanical model of structure deformationallowed an effective brick finding of 80 ± . In case of positive validation by the CS the brick isbrought to surface and exposed to cosmic rays before development for plate-to-platefine alignment. Before detaching the CS from the brick, they are exposed to four thinX-ray beams, in order to define their relative alignment.Among the 38 bricks, 36 had a good CS tagging. The measured residuals betweenelectronic detectors predictions and CS tracks were found to be of the order of a fewcm. CS to brick connection was achieved with 54 µ m and 9 mrad position and slopeaccuracy.The emulsions of the selected bricks were sent to the various automated scanningmicroscopes spread throughout Europe and Japan (about 40). All the tracks locatedin the CS were subsequently followed upstream inside the brick (scan-back) up toa “stopping point”. A general scanning (no angular preselection) was subsequentlyperformed in a volume around the stopping point(s) in order to reconstruct the vertextopology. The mechanical accuracy obtained during the brick piling is in the rangeof 50-100 µ m. The reconstruction of cosmic rays passing through the whole brickallows to improve the definition of a global reference frame, leading the precision toabout 1-2 µ m. The technique of marking emulsions with thin lateral X beams to getfast alignment pattern to be used in tracks scan-back and CS internal alignment withCompton tracks have been also successfully applied [5].In Figure 1 the display of two ν CCµ vertices reconstructed in the brick is shown.The first one is an interaction with 6-prongs and an electromagnetic shower pointingto the primary interaction vertex. In the second 4-prong vertex a decay of a π → γ ( → e + e − ) γ ( → e + e − ) has been fully reconstructed (with a thickness of 7.8 cm a Refreshing was indeed performed for all emulsions underground (Tono mine) in Japan beforetheir shipment to Italy but, in the case of the large sample of brick emulsions, it was not repeatedin Gran Sasso. This was a viable strategy thanks to the fact that the presence of tracks recordedduring transportation can be easily dealt with in this case. This is done in a dedicated area with a properly designed shielding which is intended to suppressthe low energy component. b)Figure 1: Displays of two ν CCµ neutrino vertices from the 2007 run reconstructed inthe brick. Segments represent emulsion tracks ( ∼ µ m thick), gaps are due to 1mm thick lead plates. a) the frontal and two orthogonal lateral views are shown. b)thick dashed lines represent the trajectory of γ s before conversion.brick amounts to about 10 X ). The kinematic analysis leads to an invariant massmeasurement m γγ = (110 ±
30) MeV.Figure 2 shows the first observed charm candidate. A single prong decay topologyis visible. The measured kink decay angle is 0.204 rad and the decay length is 3247 µ m. The estimated momentum of the daughter track is 3.9 +1 . − . GeV ( p T = 0.796GeV). The muon measured by electronic detectors is unambigously matched to theprimary vertex and lies in a back-to-back configuration in the trasverse plane (notshown) with respect to the charmed hadron candidate and fragmentation tracks asexpected. An electromagnetic shower is also visible in the display. The observationof one candidate in the sample is statistically in agreement with expectations.For some selected events tracks from primary vertices were also measured in theadjacent downstream brick thus validating the connection procedure which is of greatimportance when a detailed kinematic reconstruction of the event is required (mainlythrough momentum measurement by multiple Coulomb scattering). A major revision of the CNGS project has been taken in the beginning of 2008in order to improve the radiation shielding of the electronics and reduce the beamlosses. Meanwhile the OPERA target has been completed by early July 2008 incorrespondence with the start of a new long run of CNGS beam. On 1st of October2008, 1.0 · p.o.t. have been integrated. Analysis is in progress at the time of5) b)Figure 2: Displays of a charm candidate neutrino vertex in the brick from the 2007run. Segments represent emulsion tracks ( ∼ µ m thick), gaps are due to 1 mmthick lead plates. a) the full brick infomation. b) a zoomed view in the primaryvertex region.writing. The collected sample amounts to about 1000 neutrino interactions of which750 are expected to be ν CCµ events, 225 ν NCµ events, 42 charm decays, 6 ν e or ν e eventsand 0.5 ν τ events.The concept of the OPERA experiment has been experimentally validated bymeasuring neutrino events in the detector. Using the charm sample the capability toefficiently reconstruct τ decays will be fully exploited. With some dose of luck thefirst τ candidate event could be observed in the data from the current 2008 run. References [1] M. Guler et al., CERN/SPSC 2000-028, SPSC/P318, LNGS P25/2000, Jul. 2000“An appearance experiment to search for oscillations in the CNGS beam”,[2] Y. Fukuda et al., Phys. Rev. Lett. 81, 1562 (1998)[3] http://proj-cngs.web.cern.ch/proj-cngs G. Acquispace at al., CERN-98-02(1998) R. Bailey et al., CERN-SL/99-034 (1999) A. E. Ball et al., CERN-SL/Note-2000-063 (2000)[4] R. Acquafredda et al., “The OPERA 2006 run”, New J. Phys.8