The Absolute Polarimeter for RHIC
A. Bravar 1
Brookhaven National Laboratory
Upton, NY 11973 USA
Abstract. A polarimeter to measure the absolute polarization of the RHIC proton beams using an
internal polarized hydrogen gas jet target is being built at BNL. The chosen polarimetric process is
elastic pp scattering at very low momentum transfer in the Coulomb Nuclear Interference region.
In this talk I’ll discuss the beam polarization measurement and the attainable precisions.
INTRODUCTION AND METHOD
The RHIC Spin program [1] aims to determine the spin asymmetries with one or both
beams polarized for a variety of processes with high precision, such to allow significant
comparison with theoretical predictions and possibly unveil new physics. A crucial
requirement is the knowledge of the absolute polarization of the RHIC polarized proton
beams to 5% of its value or better and its continous monitoring.
For this purpose an absolute polarimeter using an internal polarized hydrogen gas jet
target is being built. The current plan is to install the polarimeter for the year 2004 run
with the initial goal of determining the beam polarization to about 10%. In 2005 our
plan is to reach the desired precision of 5% on the beam polarization.
The chosen polarimetric process is elastic pp scattering at very small momentum transfer t in the Coulomb Nuclear Interference (CNI) region of 0001 t 002 GeVc2 , where the analyzing power ANpp reaches a maximum value of about
0.045 [2]. The present knowledge of ANpp from previous experiments and theory, however, is not sufficient to obtain the desired precision. Therefore, a new measurement of
ANpp to about 0001 is required.
The luminosity of the polarized jet target, however, is too low to be practically used
as a fast polarimeter and to continously monitor the beam polarization. About 24 hours
of data taking will be required to accumulate enough elastic pp events to achieve the
required statistical precision. While the absolute beam polarization will be determined
with the jet target, the beam polarization will be routinely measured and monitored
with the fast proton-Carbon pC polarimeters installed in each RHIC ring [3]. The
pC polarimeters also operate in the CNI region. These polarimeters must be calibrated
with a beam of known polarization to an accuracy better than 5%. The calibration will
proceed in two steps. First, the analyzing power for these fast polarimeters, A NpC will
1
for the jet target collaboration: I. Alekseev, A. Bravar, G. Bunce, S. Dhawn, W. Heaberli, Z. Li,
Y. Makdisi, W. Meng, S. Rescia, E. Stephenson, D. Svirida, T. Wise, A. Zelenski
CP675, Spin 2002: 15th Int'l. Spin Physics Symposium and Workshop on Polarized Electron
Sources and Polarimeters, edited by Y. I. Makdisi, A. U. Luccio, and W. W. MacKay
© 2003 American Institute of Physics 0-7354-0136-5/03/$20.00
830
be determined using the polarized proton beam with a well measured polarization. This
procedure will require the simultaneous measurements of the absolute beam polarization
with the polarized gas jet target. Then, this measured A NpC will be used in successive
measurements of the beam polarization.
The polarized target is a free atomic beam jet. The polarized gas jet will deliver
polarized protons with a polarization near 90% and a density of about 5 10 11 pcm2 .
The target polarization will be determined with a Breit-Rabi polarimeter to better than
3%. The jet target has been discussed in different presentations at this symposium [4, 5].
The left-right asymmetry ANpp in the CNI region arises from the interference of the
electromagnetic spin-flip amplitude, due to the anomalous magnetic moment of the proton, with the hadronic spin-non-flip amplitude, which is proportional to the square root
of the total hadronic cross section. This effect is fully calculable within QED. However,
this asymmetry receives also a contribution from the hadronic spin-flip amplitude [2].
This second QCD effect at present is not predictable and potentially generates a large
uncertainty in ANpp . The analyzing power AN of pp elastic scattering in the CNI t region has been measured by the E-704 experiment [6], albeit with large statistical errors,
confirming the long-standing prediction of a structure in A N produced by the interference of the electromagnetic and hadronic amplitudes. Using the E-704 result, however,
a precision of only 20% could be obtained.
With the use of an internal gas jet target the low energy recoil protons, originating
from elastic pp scattering in the CNI region (05 Tkin 10 MeV), can be detected with
a recoil spectrometer based on silicon detectors. Previous measurements have shown that
by detecting the recoil particle only, the elastic scattering events can be selected above a
background that is smaller than 3% [7].
This measurement can be performed in situ at RHIC, with the same apparatus measuring the beam polarization, at any energy of interest, independent of theoretical assumptions, using the so called self-calibration method. The transverse spin asymmetry in elastic pp scattering of a polarized beam on an unpolarized target is identical in
magnitude but with opposite sign to the unpolarized beam – polarized target one in the
same kinematical region: ANp p ANpp . This symmetry relation, which holds for elastic
scattering only, permits the direct transfer of the target polarization Ptarget to the beam
polarization Pbeam , i.e. Pbeam can be expressed in terms of Ptarget .
An interesting alternative to calibrate the pC polarimeters is offered by scattering a
carbon beam off the polarized gas jet target [8]. This will allow a direct measurement
of AN for Cp elastic scattering in the CNI region. For elastic scattering, scattering a
polarized proton beam off a carbon target is equivalent to scattering a carbon beam of
the appropriate energy off a polarized proton target in the same kinematical region and
ANpC ACN p . This method will allow to directly measure A N for pC, which can be then
used in the pC polarimeters.
RATE ESTIMATES
To estimate the event yields we assume each RHIC beam is operated at its highest
intensity of 2 1011 protons per bunch with 112 bunches in the ring at a revolution
831
recoil detectors
jet target
yellow beam
B
blue beam
recoil detectors
FIGURE 1. Schematic view of the polarimeter set-up showing the jet target, the recoil detectors and
the target holding magnetic field.
frequency of 78 kHz. We also assume a jet target density of 5 10 11 protons per cm2 .
With these inputs the luminosity L is given by:
L
2 1011 112 78 103 5 1011 cm2 s1
1 1030 cm2 s1 1µ b1 s1 (1)
The cross section σ pp for elastic pp scattering in the CNI region of 0001 t 002 GeVc2 around 250 GeV is 3 mb. In the same t region A Npp 003.
Assuming further a coverage in azimuth of the recoil spectrometer of ∆ϕ 30 Æ the
acceptance will be acc ∆ϕ 360Æ 00833.
The expected instantaneous event rate in the CNI t region follows from these inputs:
N σ pp L acc 250 events s (2)
To obtain the desired statistical precision on A Npp of 3 104 , about 107 pp events
are required. This statistics can be accumulated in about 24 hours, where an overall
efficiency of 50% has been assumed for the polarimeter. In a few days, therefore,
sufficient data can be collected under different conditions (polarized and unpolarized
beam and/or target) to measure the absolute beam polarization to better than 5% and
calibrate the fast pC polarimeters.
KINEMATICS
For an elastic pp scattering process there is essentially one free parameter, the momentum transfer t pB pS 2 pT pR 2 between the incident beam proton and the
scattered proton or between the target proton and the recoil proton. The angle between
the normal to the incident proton direction and the recoil proton ϑ R is correlated to the
832
energy TR of the recoil proton by
sin ϑR T
R 2M p
(3)
The relationship between the recoil angle and the recoil energy also determines the
mass of the scattered system, known as missing mass. For an elastic pp process, the
missing mass must correspond to the proton mass. This is a very strong correlation to
identify elastic pp scattering events and to separate them from inelastic ones, when the
beam proton dissociates into a higher mass state. The resolution on the missing mass,
which will determine the amount of inelastic background below the elastic peak, depends
on the angular and energy resolution of the recoil spectrometer. The first is essentially
determined by the size of the jet along the beam and the distance of the detectors from
the jet. The angular resolution is estimated to be around 4 milliradians. The expected
energy resolution of the silicon recoil detectors will be around 100 keV. In addition, the
relationship between the time of flight and the energy of the recoil particle identifies the
mass of this particle, thus identifiying the recoiling protons.
SETUP
Figure 1 shows the schematic setup of the polarimeter. More details can be found
in [4, 5]. The jet target will cross the RHIC beams in the vertical direction with its
polarization also directed vertically (i.e. normal to the horizontal plane).
The recoil protons from elastic pp scattering will emerge close to 90 Æ with respect to
the beam direction and will be detected in the horizontal plane with silicon detectors.
These detectors will provide good energy, position and time measurements of the recoil
particle. Because of the very low energy of the recoil protons (05 Trec 10 MeV)
the detectors must be positioned inside the RHIC vacuum. They will be located in two
recoil-arm cylindrical chambers, attached to the central jet target vaccum chamber, on
the left and on the right of the beams at about 80 cm from the jet axis.
The recoil protons angle ϑR varies between 1Æ and 5Æ in the considered t range and
the recoil protons emerge in the same hemisphere as the scattered proton. Therefore,
two separate sets of silicon detectors will be used for each of the two RHIC beams as
illustrated in Figure 1. In order to cover a 15Æ angle in azimuth on both sides of the
beam, each set of detectors will consist of a vertical array of three silicon detectors. The
silicon detectors will have a double sided readout in order to measure the recoil angle
ϑR as well as the azimuthal angle of the recoil particles. To fully stop 10 MeV recoil
protons about 800 µ of silicon is required. Rather than build a single thick detector, a
stack of two silicon detectors will be used. The second detector will be also used as a
veto for punch through particles.
A set of Helmholtz coils centered around the jet axis will generate the necessary magnetic field of about 1.0 – 1.5 kGauss to hold the target polarization. This magnet will
give a strong momentum kick to low momentum recoil protons, making the reconstruction of the recoil angle ϑR difficult. A second set of Helmholtz coils, coaxial to the first
set, with the current circulating in the opposite direction, has been introduced to compensate the momentum kick from the inner magnet and to bring the recoil protons back
833
to the original trajectory. The total displacement from the ideal trajectory of the lowest
momentum recoil protons will be less than 3 mm at 80 cm from the target. This magnet
configuration will guarantee almost equal left – right acceptances for the recoil protons,
thus cancelling acceptance induced errors in the measurement.
The polarimeter will be installed at the Intersection Point (IP) at 12 o’clock in the
RHIC ring. At the intersection point the two RHIC beams will be displaced by about
10 mm. The jet target will be centered around one of the two beams. Thanks to the
small gas jet diameter of less than 10 mm only one beam at the time will interact with
the target gas. Because of the relatively low density of the target and its location away
from the experiments, the polarimeter can be operated continously without any impact
on the beam and the experiments. By locating the target at the interaction point, the time
between two consecutive bunch crossings of the target region is maximized. The bunch
wake-field induced pick up signal and pick up signals from capacitive couplings (beam
return currents) come at the same time for both beams in the IP and will be less of a
problem for the measurement of the recoil protons.
The detection of the forward scattered proton will be difficult because of the proximity
of any detector to the beam required to detect the forward scattered proton (very small
scattering angle). The problem of operating the forward detectors close to the high
intensity beam will make it necessary to operate the polarimeter with the recoil detectors
alone. The discriminative power of the recoil technique is powerful enough to suppress
almost all the inelastic backgrounds in the CNI t region.
ATTAINABLE PRECISIONS
The beam polarization measurement and calibration of the fast pC polarimeters will
require the measurement of several spin asymmetries. At each step some measurement
errors will accumulate.
The uncertainty on the absolute beam polarization measurement will be dominated
by the uncertainty on the target polarization, estimated at about 3%, and the background
below the elastic peak, which will contribute an uncertainty around 1%. Statistics will
also contribute an error of about 1% to the measurement.
In the presence of background, the measured asymmetry Ames
N differs from the physics
BG N S N BG is the ratio of background
R, where R N
asymmetry ANpp by ANpp ABG
N
events to the total and ABG
N is the background asymmetry. To achieve the desired precision, this term must be smaller than 10 3 . The two major sources of background are:
1. beam proton dissociation pp X p, which represents a physics background that
can be correctly modeled;
2. beam – gas and beam – residual target gas interactions, which can be suppressed
with the use of appropriate collimators.
The first source of background can be almost fully suppressed with sufficient precision
on the measurement of the recoil proton. We estimated this background to be below
1%. The second source will generate an almost uniform background below the elastic
peak (these might well be elastic pp scattering events, originating however outside of
834
the target region). This background will be estimated from the reconstructed missing
mass distribution by studying the wings of this distribution. Then the polarization of
this background will be estimated from the spin-sorted amount of background below the
elastic peak and, if necessary, it will be subtracted.
With frequent reversals of the target polarization, bunches of opposite polarization
in the RHIC rings and reversal of the beam polarization with the spin flippers several
systematic effects in the extraction of ANpp and Pbeam can be controlled and minimized.
Further, the reversal of the target holding magnetic field will allow the supression of
most left – right acceptance effects. In particular, comparing ANpp for the two RHIC
beams, will be a strong indicator for systematic effects.
SUMMARY
The measurement of the absolute beam polarization and the calibration of the fast
pC polarimeters will require several measurements of spin asymmetries, most of them
performed simultaneously, as illustrated below:
Ptarget
pp
pC
pC
ANpp Pbeam
A
P
N
beam
(4)
The overall anticipated error on the beam polarization based on the expected performance of the pp polarimeter and on the performance of the fast pC polarimeters is about
6% of the beam polarization value (i.e. relative error):
∆Pbeam
Pbeam
∆Pbeam
Pbeam
∆Ptarget
Ptarget
3%
∆ANpp
ANpp
2%
pp
∆Pbeam
P pp
beam
2%
∆ANpC
ANpC
4%
pC
∆Pbeam
(5)
P pC
beam
2%
6%
(6)
There are different possible scenarios to improve on that value. For instance, with the
jet target continously running, the beam polarization could be determined over longer
running periods to a precision of about 4%, while the polarization could be monitored
with the fast pC polarimeters.
This polarimeter will provide a reliable and efficient tool to measure the absolute beam
polarization and to calibrate the fast pC polarimeters in a relatively short time. Also, it
will allow studying various aspects of the spin dependence in pp elastic scattering.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
G. Bunce et al., Annu. Rev. Nucl. Part. Sci., 50, 525 (2000).
N.H. Buttimore et al., Phys. Rev., D59, 114010 (1999).
O. Jinnouchi et al., these proceedings (2002).
T. Wise et al., these proceedings (2002).
A. Zelenski et al., these proceedings (2002).
N. Akchurin et al., Phys. Rev., D48, 3026 (1993).
R.E. Breedon et al., Phys. Lett., B216, 459 (1989).
G. Igo and I. Tanihata, these proceedings (2002).
835