New Nu’s from the DONUT Experiment

Emily MaherUniversity of Minnesota

High Energy Physics SeminarApril 15, 2003

Direct Observation of Nu Tau

Aichi Univ. of EducationK. Kodama, N. Ushida

Kobe UniversityS. Aoki, T. Hara

Nagoya UniversityN. Hashizume, K. Hoshino, H. Iinuma, K. Ito,M. Kobayashi, M. Miyanishi, M. Komatsu,

M. Nakamura, K. Nakajima, T. Nakano, K. Niwa,N. Nonaka, K. Okada, T. Yamamori

Univ. of California/DavisP. Yager

FermilabB.Baller, D.Boehnlein, W.Freeman,B.Lundberg, J.Morfin, R. Rameika

Kansas State Univ.P. Berghaus, M. Kubanstev, N.W. Reay,

R. Sidwell, N. Stanton, S. Yoshida

Univ. of MinnesotaD. Ciampa, C. Erickson, K. Heller, R. Rusack,

R. Schwienhorst, J. Sielaff, J. Trammell, J. Wilcox

Univ. of PittsburghT. Akdogan, V. Paolone

Univ. of South CarolinaA. Kulik, C. Rosenfeld

Tufts UniversityT. Kafka, W. Oliver, J. Schneps, T. Patzak

Univ. of AthensC. Andreopoulos, G. Tzanakos, N. Saoulidou

Gyeongsang UniversityJ.S. Song, I.G. Park, S.H. Chung

Kon-kuk UniversityJ.T. Rhee

DONUT Collaboration

E. Maher

Outline

Motivation/GoalsExperimental SetupData AnalysisResults (Nu Tau Events)Individual Event ProbabilitiesCross Section MeasurementConclusions

Motivation/Goals

Old Results:Directly observed the charged current tau neutrino interaction! (Phys. Lett. B 504 (2001) 218-224)Upper limit on tau neutrino magnetic moment, (Phys. Lett. B 513(2001) 23-29) (Reinhard Schwienhorst)

Preliminary Results:Extended data set from 203 to 433

Observed 2 more tau neutrino interactions (total 6 candidates)Calculated individual event probabilities (Jason Sielaff 2002)Working on interaction cross section measurement

Is tau neutrino standard model particle?

Bµµντ7109.3 −×<

directly observe cc interactions of the ντ

ντ + N τ + X

SPECTROMETER

cτ =0.09mm

Ds

800GeV

τ ντ

ντ

BEAM DUMP

SHIELDING

EMULSIONTARGET

~ 3.5 x 1017 protons for 433 ν interactions

~ 5% ντ

τ

p

Experimental Setup – Block Diagram

Experimental Setup - Purify the Neutrino Beam

Primary Target

EmulsionTarget

36m

SpectrometerShielding to protect emulsion from the high

flux of muons.

Sweeping Magnets

Max charged particle flux ~ 105/cm2

~107/ spill @ 10 cm from emulsion edge

~1012 / spill @ 2 m

Experimental Setup - Spectrometer

Muon ID

Calorimeter

Drift Chambers

Magnet(225Mev)

5.5m

3.25m

•trigger•momentum •electron ID•muon ID•vertex location

Veto Wall

EmulsionScintillatingFiber Target

ν

ν

Experimental Setup - Emulsion Target Stand

500µ Scintillating FibersImage Intensifier - CCD Readout.

Lead Shield

260 kg total mass

Experimental Setup - Emulsion Target Designs

• AgBr suspended in a gel (Fuji ET7C ) coated on plastic sheets.

• 29±2 grains per 100µm for minimum ionizing track

95% emulsion 5% emulsion

ResolutionSpatial Resolution: .3µm

Emulsion modules consist stacks of sheets made of emulsion, acrylic, and steel.

Three configurations were used.BULK ECC 800 ECC 200

0.32 mm0.08 mm

1.0 mm0.10 mm

0.80 mm

1.0 mm0.10 mm

0.20 mm

Stainless SteelEmulsionAcrylic

•Predict interaction point from spectrometer tracks (software + humans)

- Fermilab/Minnesota/Pittsburgh/Athens•Search emulsion around predicted interaction and digitize track segments (hardware processor)

- Nagoya•Find interaction events (software pattern recognition)

- Nagoya- Minnesota/Fermilab/Athens

•Find interactions (software pattern recognition)

- Nagoya- Minnesota/Fermilab/Athens

8 hrs/event

Data Analysis – Overview of Analysis

Scanning method getting better because electronics are getting better ~100 times faster

All in mm

τν

Data Analysis – Locating Vertices in Emulsion Data

Segment detection efficiency >98%

• No e , µ from primary vertex• At least one segment on parent

76% of τ’s have visible track• Decay with one or three charged products

85% of decays are single charge• Minimum pt

pt > 250 MeV/c

Data Analysis – Finding ντ Interactions

ντ

•Short decay lengthlength < 10 mm (mean 2.5 mm)

•Small production angleangle < 200 mr (mean 40 mr)

τ

νν

WXN

τν

θ

0.8 mmacrylic

100 microns emulsion

Single Prong

νW

XN

µ

Spectrometer

Fiber tracker Emulsion

Results νµ Charged Current Interaction

From spectrometer

From emulsion

Emulsion tracks to spectrometer

Results ν Charged Current Interactione

898 predicted vertices from spectrometer

699 within fiducial volume

511 digitized emulsion data exists

264 vertex found

203 systematic decaysearch

6.6×106 triggers

451 emulsion vertex location attempted

Data Analysis Status – Then and Now

1026

812

633

633

433

433

4 ντ candidates6

charm candidates7

59%

Location efficiency

68%

Messy Event – No Emulsion Data

Typical Event – Emulsion Data, Not Located

Results New ντ Candidates

Individual Event ProbabilitiesHow Many Events Do We Expect?

For the 433 neutrino interactions we calculate how many tau neutrino charged current events we expect to observe using:

where Rate is the calculated rate of a type i event, and is the total efficiency of a type ievent, where .

14.60.57960.51108 0.34104 0.57920.59

Expected Number Rate (per kg proton) totalεccprompt µν

ccprompt non µνcc eν

cc τνnc ν

18105.03.2 −×±18106.07.2 −×±18107.04.4 −×±

18108.00.3 −×±181013.037.0 −×±

∑ ⋅⋅=

iii

NN total

total

dataRate

Rateε

εντντντ

iεcc} nc, cc,nonprompt cc, cc,prompt { τµµ ννννν ei∈

i

Individual Event ProbabilitiesHow Many Events Do We Expect?

We expect 14.6 tau neutrino charged current interactions.

Of those ~85% will be single prong and ~15% will be tridents events.

We expect 12.4 single prong events and 2.2 trident events.

Of the 12.4 single prong events, the probability of passing the selection cuts is 0.52.

So we expect to observe 6.4 single prong events. We observe 4 events.

Of the 2.2 trident events, the probability of passing the selection cuts is 0.76.

So we expect to observe 1.7 trident events. We observe 2 events.

Individual Event ProbabilitiesHow Many Background Events Do We Expect?

For the 433 neutrino interactions we can calculated the expected number of background events. Nbkg_single = 0.8 and Nbkg_trident = 2.0 (1 interaction in Fe, 1 charm).

1. A signal of 4 single prong events with an expected background of 0.8 events

A signal of 2 trident events with an expected background of 2.0 events

The Poisson probability of all signal events being background

Single Prong Events Trident Events

!):(

Ne

NfN µµµ

−⋅=

2. Using individual analysis, probabilities that each event is a tau neutrino charged current interaction given the characteristics i of the event are calculated

This method will use more information about the event, so it will be more accurate.

Since all events are independent, the probability that all events are background is 7

_ 102.2))|(1( −×=−=∏ τνiPPi

bkgall

008.0!

)8.0:4( =⋅=−

Ne

fN µµ

27.0!

)2:2( =⋅=−

Ne

fN µµ

)|i( τνP

P = The probability of a set of observables, x, being a result of eventi , where .

Two inputs for each event type: 1. Ai prior probability:

Knowledge of the likelihood of each event i

Relative Normalization (aka Nsignal, Ncharm bkg, Nint. bkg),

2. PDFi(x): probability density function

Probability of finding event in (x, x+∆x)

where x is a 3- (for trident events) or 5- (for single prong events) tuple of parameters specific to the individual event

Individual Event Probabilities( ) ∑ ⋅

⋅=

jjj

iii xPDFA

xPDFAxP

)()(

event|

n}interactio hadronic charm, cc,{ τν∈i

Individual Event Probabilities - Observables

Parameters• Track production angle θ• Event angular symmetry ∆φ• Track decay length L• Daughter momentum• Daughter decay angle

Use 3-parameter or 5-parameter analysis to assign probabilities to event interpretation.

pt

ν

τ

View perpendicular to ν direction

hadrons

hadrons

ν

eτ

θL

α

∆φ

Vector sum of all hadrons

p

Individual Event ProbabilitiesPrior Probability Calculation

To calculate the prior probability, must know1. Expected number of neutrino interactions in detector2. Probability of the process resulting in a kinked track (trident)3. Trigger, selection, and location efficiencies4. Probability that the event passes the tau selection criteria

Example: Calculating the prior probability of event 3333_17665 (single prong decay to electron)

Num.expected interactions: BR to single prong: Selection probability:Prior Probability: 0.540.10.10.36.4

0.0960.140.130.520.650.370.460.852.161.95.214.6

Charm DsDTau cΛ

Individual Event ProbabilitiesProbability Density Function Calculation

Calculating multi-dimensional probability density for a candidate event

1. Each event has measured values of parameters → (Φ ,Θ p,L) or (Φ ,Θ p,L,Θ k,P)2. Define a (small) interval in parameter space around measured values: ∆v3. Simulate hypothesis event type:

count number of events which pass all tau selection cuts (Ntotal)count number of events which are within the interval ∆v (N∆v)

Tridents Single ProngΦ ±∆Φ Φ ±∆ΦΘ p ±∆Θ p Θ p ±∆Θ p

L ± ∆L L ± ∆L probability density for event ≡Θ k ± ∆Θ k

∆v P ± ∆P

∆v

vNN

total

v

∆⋅∆

Individual Event Probability 1-D example: ντ vs hadron interaction

1. Assume the only possibilities are ντ or hadron interaction.

2. Use only one parameter Φ �to evaluate event.

3333_17665 has Φ = 2.8 rad

( ) ( )( ) ( )ττν

τντ νν

νν||

||

int Φ⋅+Φ⋅Φ⋅≡Φ

PDFAPDFAPDFA

P

Expect .22 interaction evts. Aint ï .22

Expect 6.4 ντ events Aντ ï 6.4

PDF(int. | Φ = 2.8) = .23

PDF(ντ | Φ = 2.8) = .810 π

( )τνPDF

( )intPDF

measured Φ

Φ( ) ( ) ( )

( ) ( ) ( ) ( ) 99.023.22.81.4.6

81.4.68.2| =

⋅+⋅⋅==ΦτνP

∆V

3333

_176

65

3039

_019

10

3263

_251

02ντ .70 .98 .16 .99

ν charm .30 .02 .14 .01

hadron scatter 0 0 .70 0

+

cc

Single Prong Event Parameters

• Track production angle• Event angular symmetry• Track decay length• Daughter decay angle• Daughter momentum

Individual Event Probabilities

Probability all events are background = 5 x 10-5

ν +

3024

_301

75

Single Prong

3333

_176

65

3039

_019

10

3263

_251

02

ντ .87 .70 .99

ν charm .13 .30 .01 .02

hadron scatter 0 0 0 0

+

cc

Trident Event Parameters• Track production angle• Event angular symmetry• Track decay length

Individual Event Probabilities

ν +

3024

_301

75

3334

_199

20

3296

_188

16.96 .89

.03 .01

.01 .10

Single Prong Tridents

.98 .06 .13 .99

.03 .14 .01 .01

.91 0 0

.99

.73

Probability all events are background = 2.2 x 10-7

Cross Section MeasurementCheck to see if the number of tau neutrino charged current interactions DONUT observed

agrees with theory – check lepton universality

Cross section can be calculated in different ways1. Calculate an absolute expected number of tau neutrino charged current interactions using

first principles and measured cross sections, relating event yield to protons on target2. Normalize expected number against number of electron and muon neutrino charged

current interaction observed by DONUT

The first approach will encounter unknown parameters (production cross sections, BR for charm particles, location efficiencies) This method will be used as a check.

In the second approach many uncertainties will cancel out. This method will be used.

For second approach, The following quantities are necessary:

1. Must define specific cuts to create data set2. Must know efficiencies (trigger, selection, location)3. Must know the number of electron and muon neutrino charged current interactions

Conclusion• Sampling emulsion tracker + spectrometer works even better

than expected– Still learning to use all of its power

• Near 100% reconstruction efficiency possible (68% now – from 59%)– (<50% was previously “excellent” ie CHORUS)

• Particle id• Momentum measurement• Single event probabilities

• Increasing event sample continues– From 203 to 433

• Problems are getting harder

• Better understanding of efficiencies and more tau and charm events– Improved understanding of backgrounds and systematics– Cross section measurements

• Technology for future detectors to study short lived particles.