Event Isotropy in the Salt Phase of SNO

1
Event Isotropy in the Salt Phase of SNO

• Jessica Dunmore
• University of Oxford
• NDM03, Nara - 10 June 2003

2
Reactions in SNO
p
detected particle
ne

• Charged Current
• ne d ? p p e?
• Elastic Scattering
• nx e? ? nx e?
• Neutral Current
• nx d ? nx p n
• x e, m, t

CC
p
d
e?
nx
nx
ES
e?
e?
nx
NC
nx
p
d
n
3
Detecting Electrons

• Relativistic electrons typically produce single
  cone of Cherenkov light detected by PMTs.

Fit vertex
4
Detecting Neutrons

• Pure D2O neutron capture on deuterons

e?
g'
Compton scatter
n
g
(6.3 MeV)
t
t
d
5
Simulated Neutron Event in D2O

• Neutron events in pure D2O look very similar to
  electrons.

6
Detecting Neutrons

• Pure D2O neutron capture on deuterons

e?
g'
Compton scatter
n
g
(6.3 MeV)
t
t
d

• Salt D2O neutron capture on 35Cl

g s
(8.6 MeV)
n
36Cl
35Cl
36Cl
7
Simulated Neutron Event in Salt

• Neutron events in salt are more isotropic than
  electrons.

8
Simulated Neutron Event in D2O

• Neutron events in pure D2O look very similar to
  electrons.

9
Definition of

• A measure of the isotropy
• is the average angle between all hit
  pairs.

10
Distributions
Nhit Range 50-60
CC NC
11
Can we do better?

• Use higher orders of Spherical Harmonics
• Define Harmonic Beta Parameters as
• where Pl is the lth order Legendre Polynomial,
• qij is the angle between hits i and j.
• eg.
• is very closely related to lt qij gt

12
Harmonic Beta Parameters
b1
b2
b3
b4
b5

• Using a combination of b1and b4 gives better
  separation power than using .

13
Signal Extraction PDFs

• Maximum likelihood will be used to separate CC
  and NC events .

14
Correction to Monte Carlo

• Comparison between 16N calibration data and Monte
  Carlo showed a 2.5 difference between mean
  b14b4 values.

• The Monte Carlo had more backscattered light than
  the data.

• Electron multiple scattering cross section in
  EGS4 is missing the Mott terms arising from
  electrons being 1/2 spin particles a
  significant part of the discrepancy.

15
Isotropy Calibration

• Calibration sources show excellent agreement
  between data and Monte Carlo.

Cf Data Cf MC
b14b4
16
Conclusions

• Using an isotropy parameter allows for the
  separation of electrons and neutrons in the salt
  phase via Maximum Likelihood techniques.
• This gives statistical rather than event-by-event
  separation of CC and NC signals.
• Using Harmonic Beta Parameters maximizes the
  separation power and is being used in the
  analysis of the salt data.
• Calibration sources show excellent agreement
  between data and Monte Carlo.