Status of Salerno Laboratory (Measurements in Nuclear Emulsion)
ABSTRACT A report on the analysis work in the Salerno Emulsion Laboratory is presented. It is related to the search for nu_mu->nu_tau oscillations in CHORUS experiment, the calibrations in the WANF (West Area Neutrino Facility) at Cern and tests and preparation for new experiments.
- SourceAvailable from: Antonio Di Bartolomeo[show abstract] [hide abstract]
ABSTRACT: New methods of image analysis and three-dimensional pattern recognition were developed in order to perform the automatic scan of nuclear emulsion pellicles. An optical microscope, with a motorized stage, was equipped with a CCD camera and an image digitizer, and interfaced to a personal computer. Selected software routines inspired the design of a dedicated hardware processor. Fast operation, high efficiency and accuracy were achieved. First applications to high-energy physics experiments are reported.Further improvements are in progress, based on a high-resolution fast CCD camera and on programmable digital signal processors. Applications to other research fields are envisaged.Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. 01/1997;
arXiv:hep-ex/9901034v1 25 Jan 1999
Status of Salerno Laboratory
(Measurements in Nuclear Emulsion)∗
S. Amendola, E. Barbuto, C. Bozza, C. D’Apolito,
A. Di Bartolomeo, M. Funaro, G. Grella, G. Iovane,
P. Pelosi, G. Romano
University of Salerno and INFN, Salerno, Italy
February 7, 2008
A report on the analysis work in the Salerno Emulsion Laboratory
is presented. It is related to the search for νµ→ ντoscillations in CHO-
RUS experiment, the calibrations in the WANF (West Area Neutrino
Facility) at Cern and tests and preparation for new experiments.
The analysis work in the Salerno Emulsion Laboratory is running at present
along the following main lines:
• the search for νµ→ ντ oscillations in CHORUS;
• calibrations in the WANF;
• tests and preparation for new experiments.
The Salerno Group, having developed a new system  for a completely
automatic system for search and the analysis of interactions in nuclear emul-
sions, is also continuing to improve both hardware and software tools to
increase the efficiency and the speed of the system.
This argument and some of the results obtained in CHORUS have been
presented by E. Barbuto in this workshop.
∗Presented in The First International Workshop of Nuclear Emulsion Techniques (12-14
June 1998, Nagoya, Japan), http://flab.phys.nagoya-u.ac.jp/workshop.
2 MEASUREMENT OF MUON FLUX IN THE
In the West Area Neutrino Facility (WANF)  at CERN, a muon-neutrino
beam is obtained from the decay of a parallel beam of pions and kaons
of a given sign; the beam is then dumped in order to remove the charged
The neutrino beam thus obtained is monitored by means of suitable
counters inserted around the primary beam and the target, and inside the
dump at different depths, where the muon flux is measured.
As the neutrino flux is proportional to the number of detected muons,
these measurements allow to center the neutrino beam on its target, and
to estimate its intensity through a Monte Carlo simulation that takes into
account the muon absorption and scattering in the dump.
In addition, this procedure constitutes a real on-line monitoring of the
neutrino beam and allows to detect within a few seconds any fault that could
have occurred along the beam line.
The use of nuclear emulsions is mainly intended to give an absolute
calibration to few of the many Solid State Detectors (SSDs) inserted in the
shielding, that are relatively calibrated one to the others. This technique
was already used in past  to calibrate the detectors in the Wide-Band and
in the Narrow-Band-Neutrino beam at CERN.
The present automatic analysis, however, allows to handle much higher
fluxes, with considerably increased statistics and in much shorter times.
2.1 THE WANF
The CERN muon-neutrino beam was largely rebuilt in 1992 - 93, and op-
timized for a new round of νµ→ ντ oscillation experiments (CHORUS and
450 GeV protons from the CERN-SPS are focused onto a primary target
in two spills, 6 msec long, separated in time by 2.7 sec. (fast-slow ejection),
one at the start and one at the end of the flat top. The two spills collect
about 80% of the primary intensity.
The target is made up of 11 beryllium rods (3 mm in diameter, 10 cm
long, spaced by 9 cm), aligned in a cast aluminum box; the rods are cooled
by means of a closed flux of helium gas.
An aluminum collimator, 2.75 m long, is placed 2.5 m downstream of
the target to remove the wide angle secondaries (> 9 mrad).
After the first two years of successful operation, the station received
more than 2×1019protons at 450 GeV/c, with a peak intensity of 2.8×1013
protons per machine cycle.
Positive (negative) hadrons are produced in the target and are focused
(defocused) by the subsequent beam elements.
The neutrino (antineutrino) beam is generated by positive (negative)
hadrons decays, mainly π+, K+→ µ++ νµ, or c.c.
In the case of a neutrino beam, a small contamination (about 5%) of is
produced by wrong sign hadrons (π−, K−) present in the selected beam,
while νe(about 1%) νeor (about 0.2%) are produced by semileptonic decays
A much smaller fraction of ντ (about 10-6) is produced through the
production and the prompt tauonic decay of Ds.
The beam line of Cern West Area is divided in three principal regions.
Figure 1: Target and beam line of CERN West Area.
Figure 1 shows schematically, and not to scale, the beam line in the West
The neutrino cave, 125 m long, contains the target and the focusing
elements. These are the horn and the reflector, a system of two pulsed,
coaxial conductors with cylindrical symmetry, studied by S. Van der Meer.
The decay tunnel, 290 m long and 60 cm in radius, is evacuated to a
pressure of about 1 torr in order to minimize secondary interactions. Here,
a sizeable fraction of the focused hadrons decay along a line very close to
that of their parents.
The shielding -iron and earth- 400 m long, absorbs the remaining hadrons
and decay muons.
In the shielding there are three pits about 15 m underground, 20, 40 and
60 m downstream of the decay tunnel. In each pit there is an array of 43
Solid State Detectors (SSD) to monitor the beam.
2.2 SSDs AND THEIR ABSOLUTE CALIBRATION
Solide state detectors measure the charge (as an integrated electric current)
produced by crossing particles and need an accurate calibration to convert
their electric signals into charged particles flux measurements.
The absolute calibration is made using the counting provided by nuclear
emulsions. The choice of nuclear emulsions is due to their very high efficiency
and to their excellent two track separation, that allow the detection of very
intense particle fluxes (up to more than 106part/cm2in our case).
Our work refer to measurements in pit n◦3, the most downstream, in
which muon flux is less than in the others, and for this reason, it has been
chosen for the emulsion set exposures.
Solid State Detectors (silicon detectors - sensitive surface from 30 up to
200 mm2and thickness from 0.1 up to 1 mm) are placed in boxes on a rigid
support, orthogonal to the beam line.
Figure 2 shows a typical arrangement of the counters.
A particular box with 5 SSDs, the calibration box (Cal.box), that can
be moved in front of the other counters, calibrates them. The calibration
sequence is automatically repeated every 8 hours, by successive movements
of the Cal.box.
Another special box, the reference box (Ref.box), is moved from pit to
pit, and placed in front of the calibration boxes, to calibrate them.
Finally, the absolute calibration of Ref.box, and hence of all the counters,
is performed with emulsion measurements. To be safe, a set of 4 emulsion
plates were used at the same time to calibrate also detector n◦40, on the
center of beam line, and detector n◦2 (during 1994-95) and detector n◦42
In the last 4 years, in pit n.3, 17 exposures of emulsion sets were per-
Figure 2: SSDs arrangement in pit n.3.
Emulsion plates were stuck on the center of a box containing SSDs and
exposed perpendicularly to the beam direction, so that tracks cross the
emulsions nearly perpendicularly and fully automatic scanning can be suc-
The location of the emulsion plates and their identification labels are
shown in figure 3.
We used 3 or 4 plates (3×4 cm2) for each set, and two sets of emulsions
were generally exposed the same day, each to one spill of intensity between
105part/cm2and 2 × 106part/cm2, corresponding to 1012÷ 1013protons
on target (p.o.t.).
We have used plates double coated on a 300 µm plastic base, with Fuji
or Nikfi gel. The thickness of emulsion pellicles were 100 µm (150 µm for
the last three exposures).
A total of 66 emulsion plates were exposed; together with each exposed
set, we produced a reference plate that was not exposed in the beam line
but was used for evaluating the background. In order to reduce the back-
ground, due to environment radioactivity, to cosmic rays, etc., the emulsion
plates were poured immediately before and developed soon after the expo-
sure (within a period, typically, of one week).
We used the automatic scanning, SYSAL (SYstem of SALerno). On each
emulsion 5×5 fields were scanned (about 0.39 mm2), with a gap among fields
Figure 3: Emulsion plates (black boxes) exposed in pit 3 during Chorus -
Nomad run. Empty boxes represent the Solid State detectors. The figure is
out of scale.
to prevent double counting of tracks.
Typically, 35 to 35 tracks/field were found, and a total of about 4 × 105
tracks were measured.
A two dimensional image as seen under a 50× magnification is shown in
figure 4, corresponding to a density of about 80 tracks/field.
Tracks that cross the emulsion pellicle from one surface to the other are
called passing - through tracks. In the data taking these tracks are built up
with at least 20 aligned grains, and are the only ones capable to contribute
to the electric signal of SSDs. So, all other tracks present in emulsion must
be excluded from flux evaluation, because they don’t reach the SSDs.
Other tracks reconstructed by SYSAL are:
Figure 4: Emulsion layer as seen under a 50× magnification. Passing -
through muons appear as black spots, their tracks being nearly perpendic-
ular to the plane of the figure.
• fake tracks due to association of random grains in emulsions, usually
with few poorly aligned grains;
• background tracks, like slow-energy electrons and cosmic rays, mostly
with wide angle;
• mixed tracks, poorly aligned grains with an intermediate number of
grains, contribute to both good and background tracks.
Figure 5 shows the distribution of track grain multiplicity for a particular
These plots show a peak at 28 - 30 grains due to passing through tracks;
a flat region includes mixing tracks; another peak with a few grains is due
to fake and background tracks. The cut off at number of points equal to 20,
to calculate muon flux, was chosen in order to reject the latter peak.
Figure 5: Track grain multiplicity on the top side of the plate of the 11th
exposure. The emulsion thickness after development ranges between about
50 and 60 µm.
Figure 6 shows the distributions of the residuals from a parabolic fitting
for the three categories of segments, according to their grain multiplicity.
To test the reproducibility of our measurements we have rescanned ev-
ery time the same plate; such a scanning confirms that the measurements
reproducibility is within a few percent.
For each plate we have also measured polar and azimuthal angle.
Figure 7 shows polar and azimuthal angle distributions, after distortion
and shrinkage corrections.
Figure 6: Residuals from a parabolic track fitting.
We have studied the correlation between flux from SSDs (electric signal
was converted with the existing calibration) and flux measured on all emul-
sion plates. We found a very good linear correlation. A linear correlation is
also confirmed between BCTs1countings and emulsion measurements.
These correlations are shown in Figure 8. The correlation between flux
measurements from Ref.box and flux measured from emulsion plate (Em 1)
stuck on Ref.box gives a very good agreement (within about 1%).
1The neutrino flux monitoring system also collects data about the primary proton beam
and the secondary hadron beam from Beam Current Trasformers (BCTs) upstream of the
Figure 7: Polar (up) and azimuthal (down) angle measured on an emul-
sion pellicle (plate n1, exposure n 11) after distortion and shrinkage cor-
rections for passing - through (continuous line), mixed (dashed line) and
fake/background (dotted line) tracks.
The highest density of tracks is more than 2 × 106tracks/cm2.
3A TEST FOR TOSCA EXPERIMENT
Sysal was also used to test the set-up of possible, future experiments. Among
these TOSCA , conceptually similar to CHORUS, aims to much better
detection of charged particles, and foresees an emulsion target within a mag-
An option for a wide-gap magnet could be the one built for UA1, at
present in NOMAD (maximum field B=0.7 T).
Figure 8: Correlation between the SSD and the BCT counting with emulsion
measurements. In plot (a) the emulsions are related to the SSDs on which
they were exposed; the p.o.t. (proton on target) are compared with EM1
(plot (b)) and EM2 (plot(c))
An option for a suitable spectrometer is a Compact Emulsion Tracker
(CET), a set-up already used in the past, for example in the study of high-
energy heavy-ion experiments (EMU 08, 09,16, etc.)
This set-up (fig. 9) was tested in September 1997 where an emulsion
target 9 cm thick (3X0) was followed by silicon trackers and by a series of
8 thin, parallel emulsion sheets spaced from 8 to 32 mm over a total length
Figure 9: Set up of CETs position. Figure is not in scale
A 15 GeV beam of pions was used to produce interactions in the target,
whose particles have been traced in the external apparatus.
A first alignment between pairs of CET sheets was realized by means of
X-ray guns (see fig. 9). A better approximation was obtained by measuring
positions and angles of a sample of primary particles, whose individual iden-
tity was recognized from one sheet to the next by comparing the topological
The angular spread of the beam (fig. 10) was found to be about 1.6
mrad, in excellent agreement with that expected for a 15 GeV parallel beam
after having crossed a 3X0target.
This procedure allowed to place all sheets in a common reference frame
by means of a fit of the rototraslational parameters.
Finally, a random search was performed in the first pellicle for random
tracks out of the beam (θ > 50 mrad), and thus some tracks were searched
and measured in the following pellicles.
A linear fit was performed for the transverse coordinates as functions of
the longitudinal one: as a result, systematically the residuals on the bending
plane turned out to be about one order of magnitude larger with respect to
those in the non-bending planes, and these are compatible with the accuracy
Figure 10: Absolute angular distributions of primaries.
in locating the pellicles with the location procedure described above.
A parabolic fit in the bending plane allowed to determine the radius of
curvature, hence the momentum of the particles.
Now, the residuals turn out to be similar to those in the non-bending
Fig. 11 shows a typical example of the results obtained on a particular
track. Table n.1 shows the global results for a set of tracks.
Radius of curvature (m)
20.7 ± 3.8
−26.3 ± 3.6
10.2 ± 0.9
−23.8 ± 2.4
−18.5 ± 2.7
17.9 ± 2.0
−4.17 ± 0.07
41.7 ± 6.9
−57.5 ± 11.9
−55.6 ± 8.6
−2.8 ± 0.09
4.3 ± 0.8
−5.5 ± 0.7
2.1 ± 0.2
−5.0 ± 0.5
−3.9 ± 0.6
3.7 ± 0.4
−0.87 ± 0.01
8.7 ± 1.4
−12.1 ± 2.5
−11.7 ± 1.8
−0.58 ± 0.02
Of course, this procedure assumes that the primary particles have straight
trajectories, but their curvature could be easily taken into account.
In the near future it is planned to repeat these measurements on pre-
dicted tracks, whose momentum has been also measured in the external
From these results it is concluded that this set-up allows to measure
momenta with a minimum relative error ∆p/p ≈ 6%, essentially due to
multiple scattering in the emulsion sheets.
The measurement error, that could be possibly lowered, gives at present
∆p/p ≈ 1.4% × p.
 G.Rosa et al, Automatic Scan and Analysis of Digitized TV Images by
a Computer Driven Optical Microscope, NIM A 394 (1997) 357-367.
 C.Bozza, A. Di Bartolomeo, G.Iovane, P.Pelosi, Measurements of the
muon flux in the CERN SPS wide band neutrino line, CHORUS Internal
Note 97030, Geneva, January 1998.
 E.H.M.Heijne, Muon flux measurements with silicon detectors in the
CERN neutrino beams, CERN 83-06 Experiment Physics Facilities Di-
vision, Geneva, 21 July 1983.
 European Organization for Nuclear Research, Request for a test
of emulsion-silicon configurations for a neutrino experiment, CERN-
SPSC/97-4 SPSC/M596, 21 February 1997.
Figure 11: Linear fit of z and y coordinate versus x, parabolic fit of z coor-
dinate versus x.