Proton Decay searchesProton Decay searches---- sensitivity, BG and photosensitivity, BG and photo--coverage coverage ----

Univ. of Tokyo, Univ. of Tokyo, Kamioka ObservatoryKamioka ObservatoryMasato ShiozawaMasato Shiozawa

April-2005 @ NNN05

Water as a proton decay detectorWater as a proton decay detector••Source HSource H22OO

•• 2/10 free proton 2/10 free proton →→

no nuclear effect, no nuclear effect, accurate&highaccurate&high detection efficiencydetection efficiency→→

no Fermi motion, good momentum valanceno Fermi motion, good momentum valance•• 0.54Megaton(Hyper0.54Megaton(Hyper--K) K) →→~2~2××10103535 protonsprotons

••Good detector performanceGood detector performance•• Vertex resolution: 30 cm (1Vertex resolution: 30 cm (1--ring)ring)

:~20 :~20 cm(pcm(p ee++ππ00))•• Trigger threshold: 5 MeV electronsTrigger threshold: 5 MeV electrons

trigger trigger εε=100% for most of nucleon decay modes=100% for most of nucleon decay modes•• Energy resolution: 3~4% for e, Energy resolution: 3~4% for e, μμ•• Particle ID : 99% 1Particle ID : 99% 1--ring ring μμ, e, e

: ~95% p : ~95% p ee++ππ00 ,p ,p μμ++ππ00

These performance is achieved in Super-K-I, 40% photo- coverage.Question: can we reduce photo-coverage? Keeping the excellent performance?

April-2005 @ NNN05

Large water Cherenkov detectorsLarge water Cherenkov detectors

Super-K 22kton

Hyper-K 540kton

UNO 440kton

Simulation was done.SK-I: 40% coverageSK-II:19% coverage

April-2005 @ NNN05

Charge(pe) >15.013.1-15.011.4-13.1 9.8-11.4 8.2- 9.8 6.9- 8.2 5.6- 6.9 4.5- 5.6 3.5- 4.5 2.6- 3.5 1.9- 2.6 1.2- 1.9 0.8- 1.2 0.4- 0.8 0.1- 0.4 < 0.1

Super-KamiokandeRun 999999 Event 294102-11-06:00:06:35

Inner: 3849 hits, 8189 pE

Outer: 4 hits, 2 pE (in-time)

Trigger ID: 0x03

D wall: 946.1 cm

FC, mass = 909.0 MeV/c^2

00 500 1000 1500 2000

0

182

364

546

728

910

Times (ns)

pp ee++ππ0 0 @Super@Super--KKpp ee++ππ00 MC MC

γγ

ee++

γγ

April-2005 @ NNN05

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

)

Super-Kamiokande Preliminaryp→e+π0 MC

pp ee++ππ00MCMC

ε ε = 40= 40 % in SK% in SK--II

pp ee++ππ0 0 @Super@Super--KK

2 2 or 3 Cherenkov ringsor 3 Cherenkov ringsAll rings are showeringAll rings are showering85 < M85 < Mππ00 < 185MeV/c< 185MeV/c22

(3(3--ring)ring)No decay electronNo decay electron800 < 800 < MMprotonproton < 1050 < 1050 MeV/cMeV/c22

PPtotaltotal < 250 MeV/c < 250 MeV/c

Criteria for Criteria for pp ee+ + ππ0 0

April-2005 @ NNN05

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

) atmν MC

Tight momentum cut for pTight momentum cut for p ee++ππ00

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

) free proton

Ptot < 250 MeV/c, BG2.2ev/Mtyr, eff=44%

Ptot < 100 MeV/c, BG0.15ev/Mtyr, eff=17.4%

20Mtonyr 0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

) bind proton

free proton decay

bound proton decay

Main target is free proton decays for the tight cut.

April-2005 @ NNN05

Lifetime sensitivity for pLifetime sensitivity for p ee++ππ00

10 32

10 33

10 34

10 35

10 36

10 37

102

103

104

105

106

Exposure (kton year)

Par

tial

Lif

etim

e (y

ears

)

p→eπ0 sensitivity (90%, 3σ CL)

SK-I limit

92ktyr5.4 x 1033 yrs (90%CL)

10 32

10 33

10 34

10 35

10 36

10 37

102

103

104

105

106

Exposure (kton year)

Par

tial

Lif

etim

e (y

ears

)

p→eπ0 sensitivity (90%, 3σ CL)

SK-I limit

92ktyr5.4 x 1033 yrs (90%CL)

Normal cut, 90%CL3σ

CLTight cut, 90%CL

3σ

CL

p e+π0 sensitivity

Hyper-K 10yrs ~1035 years@90%CL~4x1034 years@3σCL

HK 10yearsHK 10years

April-2005 @ NNN05

Reduce photoReduce photo--sensor cost?sensor cost?

•• sensor cost is a significant part of total cost.sensor cost is a significant part of total cost.•• in case of Hyperin case of Hyper--KK

1000$(1/3xSK) x 200,000(40% coverage) = 200M$1000$(1/3xSK) x 200,000(40% coverage) = 200M$

•• important to understand minimum requirement of photoimportant to understand minimum requirement of photo-- coverage from each physics topics (pcoverage from each physics topics (p e+e+ππ0 in my talk)0 in my talk)

•• SKSK--II (19% coverage II (19% coverage SKSK--I 40%I 40%) is a good opportunity ) is a good opportunity to investigate physics sensitivity with reduced phototo investigate physics sensitivity with reduced photo--coveragecoverage

•• well tuned SKwell tuned SK--II detector simulationII detector simulation•• fitters (fitters (vertex,ring#,particlevertex,ring#,particle ID,momentumID,momentum……) are also ) are also well tuned and calibrated.well tuned and calibrated.•• reliable study is possible.reliable study is possible.

April-2005 @ NNN05

(1)vertex fitter and (2)ring fitter(1)vertex fitter and (2)ring fitter

050

100150200250300

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1Nu

mb

er o

f ev

ents

68%

050

100150200250300

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

dist. btw reconstructed and true vertex (m)

Nu

mb

er o

f ev

ents

68%

0100200300400500600700800

1 2 3 4 5Nu

mb

er o

f ev

ents

38% 60% 2%

0100200300400500600700800

1 2 3 4 5

Number of ringsN

um

ber

of

even

ts

41% 57% 2%

SK-I(40%c)

SK-I

SK-II

SK-II(19%c)

p e+π0 Monte Carlo

19.7cm(68%) 21.8cm(68%)Almost same performance

Efficiency(74% for 2-3rings) doesn’t changeFraction of 3ring slightly decrease

H2O

free proton

April-2005 @ NNN05

0102030405060

0 100 200 300Nu

mb

er o

f ev

ents

0102030405060

0 100 200 300

π0 invariant massN

um

ber

of

even

ts

050

100150200250300350400

-400 -200 0 200 400Nu

mb

er o

f ev

ents

showering type non-showering type

050

100150200250300350400

-200 -100 0 100 200

Particle ID likelihood

Nu

mb

er o

f ev

ents

showering type non-showering type

(3)Particle ID and (4)(3)Particle ID and (4)ππ0 mass0 mass

SK-I

SK-I

SK-II

SK-II

p e+π0 Monte Carlo

PID performance; 94-95%96% for free proton decay

π0 mass resolution; 22.1MeV 24.0MeV

free proton 20.2MeV 28.5MeV

April-2005 @ NNN05

0102030405060708090

100

0 200 400 600Nu

mb

er o

f ev

ents

0102030405060708090

100

0 200 400 600

total proton momentumN

um

ber

of

even

ts

0102030405060708090

100

600 800 1000 1200Nu

mb

er o

f ev

ents

0102030405060708090

100

600 800 1000 1200

proton invariant mass (MeV/c2)

Nu

mb

er o

f ev

ents

(5)Proton mass and (6)proton (5)Proton mass and (6)proton momentummomentum

SK-I

SK-I

SK-II

SK-II

p e+π0 Monte Carlo

proton mass resolution; 33.8MeV 42.3

free proton 28.5MeV 35.2

proton mom. resolution; 177MeV(68%) 171

free proton 81MeV(68%) 83

April-2005 @ NNN05

Efficiency in each stepEfficiency in each stepp e+π0 Monte Carlo

0

0.2

0.4

0.6

0.8

1

A B C D E

selection criteria

det

ecti

on

eff

icie

ncy DST 2,3ring

PID π0 decay-e MtotPtot

combinedfree proton

combinedfree proton

Detection eff. = (40.1+-1.5)% -- SK-I(41.1+-1.5)% -- SK-II

SK-I SK-II Same efficiency within 1% level

Slightly worse resolution of π0 and proton mass

April-2005 @ NNN05

0100200300400500600700800900

1000

0 200 400 600 800 1000 1200

To

tal m

om

entu

m (

MeV

/c)

0100200300400500600700800900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)To

tal m

om

entu

m (

MeV

/c)

Proton mass Proton mass vsvs momentummomentum

SK-ISK-I

2.2Mtyr

SK-II0.45Mtyr

SK-II

atmν(BG) Monte Carlo

(40.1+-1.5)% (41.1+-1.5)%

0100200300400500600700800900

1000

0 200 400 600 800 1000 1200

To

tal m

om

entu

m (

MeV

/c)

0100200300400500600700800900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)To

tal m

om

entu

m (

MeV

/c)

p e+π0 Monte Carlo

8ev/2.2Mtonyr 2ev/0.45Mtonyr~10+-7ev/2.2Mtyr

Same BG level within 70%

April-2005 @ NNN05

ConclusionConclusion

•• Water Cherenkov detector is a excellent detector for Water Cherenkov detector is a excellent detector for p p ee++ππ00 searches.searches.

•• detailed comparison btw 40% and 20% coveragedetailed comparison btw 40% and 20% coverage•• same psame p ee++ππ00 efficiency even with 20% coverageefficiency even with 20% coverage•• slightly worse mass resolutionslightly worse mass resolution•• 10% coverage maybe acceptable10% coverage maybe acceptable•• need further studies on other decay modes, like need further studies on other decay modes, like pp νν+K+K++, e, e+++K+K00, , ……

supplements

April-2005 @ NNN05

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)To

tal m

om

entu

m (

MeV

/c)

Super-Kamiokande Preliminary1489 days data

AtmAtmννBGBG MCMC datadata

0 0 candidatecandidate0.0.3 3 expexp’’dd BGBG

τ τ pp/B(p/B(p ee++ππ00) > 5.4) > 5.4××10103333 years (90% CL)years (90% CL)

pp ee++ππ0 0 @Super@Super--KK--II

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

)

Super-Kamiokande Preliminary100 years atmν MC

April-2005 @ NNN05Comparison of data and MC in Comparison of data and MC in pp ee++ππ00 searchsearch

distance L in mass v.s. momentum plot

Super-Kamiokande Preliminary1489.2 days

10-1

1

10

10 2

10 3

10 4

10 5

0 200 400 600 800 1000 1200

total mass (MeV/c2)tota

l mo

men

tum

(M

eV/c

)

L

(938,200)

10-1

1

10

10 2

0 200 400 600 800 1000 1200

Super-Kamiokande Preliminary1489.2 days

total invariant mass (MeV/c2)

0

2

4

6

8

10

12

14

16

0 500 1000 1500 2000 2500 3000

Super-Kamiokande Preliminary1489.2 days

total momentum (MeV/c)

April-2005 @ NNN05

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

) atmν MC

Backgrounds for p→e+π0 search (2)

BG events in signal box3 events/20 Mton・yr

～0.15 events/Mton・yr

Tight momentum cut to reduce BG

Ptot < 250 MeV/c↓

Ptot < 100 MeV/c

April-2005 @ NNN05

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)To

tal m

om

entu

m (

MeV

/c) bind proton

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200

Invariant proton mass (MeV/c2)

Tota

l mo

men

tum

(M

eV/c

) free proton

Analysis for discovery of p→e+π0

Tight momentum cut⇒

target is mainly free protonsefficiency=17.4%, 0.15BG/Mtyr

No Fermi momentumNo binding energyNo nuclear effect

Small systematic uncertainty of efficiHigh detection efficiencyPerfectly known proton mass and moment

free proton bound proton

April-2005 @ NNN05

Lifetime predictionLifetime prediction

Dimension=6 (2 fermion – 2 fermion)

Dimension=5 (2 fermion - 2 sfermion)

yearsep

XMpmg

?35~10)0(: 4

54=+→=Γ πτ

yearsKp

XMHcMpmh

?39~2910)(: 22

54=+→=Γ ντ

1030 1031 1032 1033 1034 1035

partial nucleon life time (year)

minimal SU(5)

minimal SUSY SU(5) SUSY SO(10)Super-K

uud d

p Κ+

q~

q~w~sν

CH~

uud d

p π0dX

+e

April-2005 @ NNN05

energy reconstructionenergy reconstruction

0.950.960.970.980.99

11.011.021.031.041.05

0 250 500 750 1000 1250 1500 1750 2000Elapsed days from Apr.-1-1996

μ m

om

entu

m/t

rack

±1%

0.950.960.970.980.99

11.011.021.031.041.05

0 250 500 750 1000 1250 1500 1750 2000Elapsed days from Apr.-1-1996

e m

om

entu

m

±1%

+-1%

+-1%

Corrected for light attenuation length in waterTime variation of E scale ~ 0.9%

0

20

40

60

80

100

120

140

0 50 100 150 200 250 300MeV/c2

Num

ber

of e

vent

s

cosmic ray mu

decay electrons

π0 invariant mass

E scale difference < 1.8%(decaye, pi0, cosmic mu)

energy scale uncertainty of neutrino detection < 2.0%

Full Super-K-I period