Various Work for Belle Detector

FGIP-student ForumTIT, 2005-06-17

Saša Fratina,

Jožef Stefan Institute, Ljubljana, Slovenia

Outline of the talk

A little bit about Slovenia.

Ring Imaging Cherenkov Counter (RICH)

Silicon Vertex Detector (SVD) CP-asymmetry measurement

in B meson system

Where do I come from? 　

Slovenia

Croatia

HungaryAustria

Italy

Few facts ： 2M inhabitants, 20.000 km2, capital Ljubljana, EU members (www.slovenia.info)

Education system: public schools

Primary school (9 years, children 6-15 years old) Secondary school – high school (4 years,

students 15-19 years old), at the end of which you have to pass national exams from 5 subjects (slovene language, math, english and two by student`s choice)

University (4-5 years for graduation, usually takes longer)

Graduate studies: masters or PhD course depending on the field, in physics mostly PhD.

Universities in Slovenia

University of Ljubljana (largest, 22 faculties, 50.000 students)

University of Maribor Nova Gorica Polytechnic

My education

Primary school: Ljubljana High school: Gimnazija Bezigrad,

Ljubljana (finished with International Baccalaureate)

graduated at the University of Ljubljana, Faculty of Mathematics and Physics

started PhD study in 2002 at the University of Ljubljana (and working at Jozef Stefan Institute, Ljubljana)

Belle Control room

About work…

Main goal of Belle experiment: study of CP violation in B-meson system

Mt. Tsukuba

KEKBBelle

(4s)e+ e-

p(e+)=3.5 GeV/c

p(e-)=8.0 GeV/c

(4s)

B

B z ~ 200 m

Silicon vertex detector

Ring Imaging

Cherenkov Counter

Data analysis:

study of CP violation

RICH for super-Belle (2008?)

Requirements

Compact detector

Good /K separation in the forward (high momentum) region

for few-body decays of B's (B , B K)

for b -> d g , b -> s g ( , K )

Low momentum (<1GeV/c) e/μ/ separation (B ->Kll)

High efficiency for tagging kaons

Basic principle

Ring Imaging Cherenkov counter, RICH

cos Ch = 1/n

= v/c

Position sensitive photon detector

particle

Cherenkov photons

aerogel

Aerogel

n = 1.05 Ch () ~ 308 mrad

Ch () - Ch () ~ 23 mrad few cm thickness transmission length

2.5 - 4.5 cm approx. 10 emitted

photons / cm

100x100x20mm3 n=1.050

No cracks

Photon detector photo multiplier tube

(PMT) single photon sensitive position sensitive:

position resolution few mm quantum efficiency 20%

Photon detector: array of 16 H8500 PMTs

Beam test measurements Confirm feasibility of

such detector Study /K separation

capability

Clear rings, little background

Beam test: Cherenkov angle resolution and number of photons

Beam test results with 2cm thick aerogel tiles: >4K/separation

-> Number of photons has to be increased.

Typically around 13 mrad (for 2cm thick aerogel)

How to increase the number of photons?

What is the optimal radiator thickness?

Use beam test data on 0 and Npe

Minimize the error per track: (Npe) Optimum is close to 2

cm

0 Npe

(Npe)

● measure overlaping rings“focusing” configuration

● measure two separate rings “defocusing” configuration

How to increase the number of photons without degrading the resolution?

normal

Radiator with multiple refractive indices

4cm aerogel single index

2+2cm aerogel

FOCUSING CONFIGURATION - data

Increase the number of photons without degrading the resolution!

● number of detected hits: dual radiator has a clear advantage

● single photon resolution: dual radiator ~same as single (of half the thickness) for the full momentum range

FOCUSING CONFIGURATION – momentum scan

Development and testing of photon detectors for 1.5 T Baseline: large area HPD of the proximity focusing type Backup: MCP-PMT (micro channel plate)

-10kV15~25mm

e-

Multialkali photocathode

Pixel PD or APD

R&D project in collaboration with HPK

RICH - conclusions

Feasibility of RICH detector was confirmed. More photons: employ radiators with multiple

refractive indices. Idea successfully tested in beam tests.

Aerogel production: transmission length improved, new cutting methods tested, multiple layer samples.

R&D issues: development and testing of a multichannel photon detector for high mag. fields

Silicon vertex detector

50 cm

20 cm

basic SVD unit: Double Sided Strip Detector (DSSD)

separate measurement of r and z coordinate

strip pitch: 50 and 75 m for r and z coordinate, respectively

pitch

Evaluate the performance of SVD – measure its intrinsic resolution:Incident angle dependence, occupancy study,alignment cross-check

SVD intrinsic resolution

Error on the track position measurement

Track position

is determined

from the SVD

hits on other layers

DSSD

track

residual

SVD hit

Typical residual distributions

Intrinsic resolution is determined from the width of the residual distribution: i = 10 m, RMS = 12-15 m for r and i = 25 m, RMS = 30 m for z coordinate

residual [cm] residual [cm]

r coordinate z coordinate

Incident angle dependence Simple estimate for the perpendicular tracks:

signal collected by single strip → resolution ≈ strip pitch / 12

Small incident angle: signal collected by few strips → resolution improved

Large incident angle: signal

collected by many strips

→ resolution gets

worse due to smaller

signal to noise ratio

track

x strips

x

y

angl

e

Incident angle dependence: result

RMS [m] RMS [m]60

40

20

0-20 0 20 40 Incident angle [degrees]

-20 0 20 40 60Incident angle [degrees]

r coordinate z coordinate

Innermost layer

30

20

10

0

Incident angle dependence: result

RMS [m] RMS [m]

Incident angle [degrees]-20 0 20 40 60

Incident angle [degrees]

r coordinate z coordinate

Different colors show the result for all four layers: black, red, green and blue for the innermost, second, third and outermost layer.

-20 0 20 40

60

40

20

0

30

20

10

0

Magnetic Field Effect Intrinsic resolution is not symmetric with respect to

perpendicular

incident angle Reason:

magnetic field,

confirmed

by the plot of

hit cluster size next SVD:

Distribution of hits

with cluster size

1 strip, 2 strips, 3 strips, …

-20 0 20 40

incident angle [o] -12o

Magnetic field effect

n side

p side

E

B .Fe

Fm

e- direction

x

y

track

n side

p side

+

++

track

E

B .

angl

e

Fm

+

++

Fm

x

y

Negative angle:

smaller cluster size

Positive angle:

bigger cluster size

Degradation of SVD intrinsic resolution ｄｕｅ to higher detector occupancy (background): cluster distortion or cluster mis-association?

residual

occupancy < 0.04

residual

occupancy 0.3

Number of clusters with MC hit (correctly associated)

occupancy, x coordinate

Rel

ativ

e fr

actio

n of

SV

D h

itswrong cluster association is negligible

Number of clusters with rel. change in E > 0.2

Number of clusters with MC hit (correctly associated)

Alignment cross-check

Check if residual distributions for individual SVD units are shifted (mis-aligned)

Try to correct

the shift

residual [cm]

r coordinate

Conclusions on the SVD intrinsic resolution study Best resolution (RMS) at small track incident

angle is found to be 12 and 30 m for r and z coordinate, respectively.

Results of this study provided important information for cross-check of the SVD alignment different geometry design to take into account

magnetic field effect improvement of clustering algorithm in the case of

higher detector occupancy

Analysis of data collected at the Belle detector:

Measurement of Time Dependent CP Violation in B0 → D+D- Decays

B 0B 0 fCPfCP

B 0B 0

fCPfCPB 0B 0

B 0B 0

B0 → D+D- CP eigenstate

tree

penguin

B0

B0

c

c

d

c

c

d

D-

D+

• B 0 and B 0 mixes with each other via “box-diagram”.

• The box-diagram includes CKM complex phase.• A path from B 0 to fCP via mixing has different weak

phase from one from B 0 to fCP directly due to the CKM

phase.

W - W +

t

t

d

b

b

dVtb

*

Vtb*Vtd

Vtd

Time-dependent decay amplitude

PCP±(t) = exp(-|t| / ) / (4) ·

(1±(1-2w)(S sin(mt)- C cos(mt))) S = - sin(2) (if only tree diagrams are present)

CKM triangle

Analysisfinal step: fit S and C to t distribution Reconstruct the events Measure the time of B meson decay from the

reconstructed B vertices. Determine B0 flavor

Properly fit the t distribution (taking into account detector resolution, mis-tagging of B mesons,…)

5 steps toward the CP asymmetry measurement

• Reconstruct B KS decays• Measure proper-time difference: t • Determine flavor of Basc

• Evaluate asymmetry from the obtained t distributions

• Discuss observed CP asymmetry

e- e+e: 8.0 GeVe: 3.5 GeV

Brec

z

Basc

Y(4S)~ 0.425

fCP (fKS)fCP (fKS)

z c tB ~ 200 mm

Flavor tagFlavor tag

Event reconstruction

Br ~ 1.7 10-4

D mesons are reconstructed only in decays to charged particles (10%)

Detector efficiency for reconstructing such an event ~ 10 – 20 % (at S/B ~ 1)

Expect only few (2-3) events per 10 M BB events!

High luminosity B-factory is needed – KEK!

Approx. 350 M recorded BB events – about 100 of them will be correctly reconstructed as B0 → D+D-

B0 decay time

Average distance between the two decay vertices is 200 μm

Need to measure the vertex position with better accuracy

With SVD we are able to measure the vertex z coordinate with accuracy about 100 μm

Flavor tagging

Use flavor-specific decays ( slow, K, leptons)

t =trec

BrecBrec

(4S)(4S)BascBasc

Brec = B 0 decay(flavor-specific)

Brec = B 0

?????? D+D- decay

t =tasc

B 0-B 0 mixing

t

Current status

Event selection is optimised according to best signal to background ratio.

Vertex position and tagging are determined using standard Belle software

Fitting procedure is under development

(I can not show any preliminary results)

Conclusion…

It is great to work for

SVD

RICH

Study of CP violation