P0D Light Yield
Howard Budd, Jesse Chvojka, Kevin McFarland
University of Rochester
12 November 2006
Abstract
The light yield in the P0D has been estimated from results form the
MINERvA 3-layer Vertical Slice Test. This memo describes the
scaling of the MINERvA result to the result for the P0D.
The P0D light yield has been estimated by scaling arguments from the measured light
yield in the MINERvA Vertical Slice Test (VST). The VST does one thing very well,
which is to positively identify muons passing through the fiducial portion of the detector
to establish the light yield for a minimum ionizing particle. In brief, the analysis requires
not only a straight down cosmic ray muon (within an angle of ~10 from the vertical), but
it also requires energy depositions consistent with the MIP peak all layers of the array
except the layer in which light yield is being measured, that the cluster shape is consistent
with a single track, and that the track is “straight” to within 3σ of the measured resolution
(~2.5mm per doublet). This removes a number of backgrounds to minimum ionizing
particles, such as emission of a hard δ-ray at wide angles, and events caused by more
heavily ionizing particles such as showing electrons or slow protons. The MINERvA
VST array is shown below in Figure 1.
Figure 1 The MINERvA Vertical Slice Test
Because of this, we can convert the light yield measured in the VST into a light yield
estimate per MeV of energy loss in the scintillator by using the energy loss of a minimum
ionizing particle through the doublet of scintillator, (1.65 cm)× (1.03 g/cm3)× (1.94 MeVcm2/g)=3.30 MeV. The distribution of the number of photoelectrons is shown in Figure 2.
The average number of photoelectrons measured in the VST for a cluster is 20.9±0.5, and
therefore the number of photoelectrons per MeV of energy deposit is 6.34±0.15.
Figure 2 The total number of photoelectrons in layer 2 of the VST for selected minimum ionizing
particles. The number used in the light yield calculation is smaller by 5% because it is the sum over
just a cluster.
In order to relate the two light yields, a number of corrections must be made to account
for the differences between the two setups. A schematic of the VST setup is shown in
Figure 3
R7600
-M64
PMT
MINERvA Bar with
1.2mm WLS fiber
MINOS “Alner” Box
Figure 3 A schematic of the MINERvA VST.
The obvious difference between the two setups, of course, is the geometry and type of the
fiber and photosensor. To calculate P0D light yield form the VST, the overall strategy is
to scale the expected light yield to that expected from a the VST PMT at the end of a
fiber on the bar, and then to attempt to scale to the light expected by replacing the R7600M64 MAPMT with a Geiger mode APD. It is the last step that is the most uncertain.
Parameters of the VST and the P0D are listed in Table 1.
The extrapolation of the light at the end of the fiber gives a ratio of P0D to VST light of
0.83±0.05. This number includes:
1. A factor for WLS fiber collection that is the ratio of the fiber diameters
2. A factor for attenuation of the fiber due to different lengths. Accounting for
attenuation, including the reflected light, this enhancement of P0D light is
(e325/ 500  0.8e375/ 500 ) /(1  0.8)e230/ 500  1.26 at the far end of the bar for the
parameters in Table 1. (Note the first term in the numerator and denominator of
the ratio is the direct light and the second is the reflected light from the mirror.)
At the center of the bar, relative to the VST measurement, the factor would be
1.33.
3. A correction for the light lost in the connector and short fiber of the Alner box of
the VST, 1/0.85.
4. A correction for light not collected since the P0D will not have a glued fiber,
measured in the VST to be 1/1.5 for the 1.2mm fiber. Note that fiber-scintillator
coupling has been shown to be insensitive to fiber diameter in previous studies by
Anna Pla-Dalmau and Victor Rykalin (reported by KSM at January and
November 2006 T2K meetings).
Accounting to the relative efficiency of the photosensor to be 1.5±0.5, the net ratio of
light in the P0D to that in the VST is 1.2±0.4. Note that the uncertainty in this result is
completely dominated by the unknown relative photosensor efficiency. This gives a final
light yield estimate of 8±3 photoelectrons per MeV for the P0D bars.
A future measurement that could improve this uncertainty would be to measure explicitly
the ratio of light (using a source) of a P0D bar with a P0D fiber for the VST photosensor
and a P0D photosensor. This measurement is difficult at this time because the final
photosensor and final fiber-photosensor coupling apparatus are not available.
Parameter
MINERvA VST
P0D
Notes
Distance from Far end of bar to
point for Measurement
25 cm
0 cm
Want light at far
end of the P0D
Fiber Diameter
1.2 mm
1.0 mm
Y-11 Concentration
175 ppm
175 ppm
Fiber Length
350 cm
230 cm
Transmission of Components
Between Fiber and Sensor
0.85±0.03
1
measured by
MINOS
Light Enhancement from Glued
Fiber-Scintillator Coupling
1.50±0.05
1
measured in
MINERvA VST
Fiber Attenuation Length
5.0±0.5 m
5.0±0.5 m
assumed for
1.0mm P0D fiber
Fiber mirror reflectivity
0.8
0.8
Distance from Fiber to Photosensor
0.5 mm
0.2 mm
assumed for P0D
Geometric Acceptance of
Photosensor active area
1
1
For MRS device.
For 1mm MPPC, it
is 0.9 for P0D.
Relative Efficiency of Photosensor
(the product of geometric
acceptance within active area and
photon detection efficiency)
1
1.5±0.5
Based on
Yokoyama-san’s
report of Kyoto
MPPC
measurements.
Flawed because
QE of reference
PMT was
unspecified!
Table 1 Parameters of VST and P0D important for extrapolation