Measurement of the polarization of the proton-proton elastic reaction at small scattering angles between 940 and 2440 MeV

Measurement of the polarization of the proton-proton elastic reaction at small scattering angles between 940 and 2440 MeV

Nuclear Physics A505 (1989) 561-582 North-Holland, Amsterdam MEASUREMENT ELASTIC OF THE REACTION S. DALLA INFN POLARIZATION OF THE AT SMALL SC...

1MB Sizes 0 Downloads 14 Views

Nuclear Physics A505 (1989) 561-582 North-Holland, Amsterdam

MEASUREMENT ELASTIC

OF THE

REACTION

S. DALLA

INFN

POLARIZATION

OF THE

AT SMALL SCATTERING 940 AND 2440 MeV

PROTON-PROTON

ANGLES

BETWEEN

TORRE-COLAUTTI, R. BIRSA, F. BRADAMANTE, M. GIORGI, L. LANCERI, A. MARTIN, A. PENZO, P. SCHIAVON, V. SOSSI and A. VILLARI

Sezione di [email protected] and Dipartimento H. AZAIEZ’,

K. KURODA LAPP,

di Fisica, Universita

di Trieste, Italy

and A. MICHALOWICZ

Annecy-le- Vieux, France F. LEHAR

DPhPE,

CEN-Saclay,

France

Received 20 March 1989 (Revised 27 June 1989) Abstract: We have measured the asymmetry of elastic pp scattering at small scattering angles (30100 mrad) in the Coulomb-nuclear interference region, using the polarized proton beam of Saturne II, a segmented scintillator active target, and two telescopes of multiwire proportional chambers. Results are given at four energies - 940, 1000, 1320 and 2440 MeV- and are compared with phase-shift calculations.

E

NUCLEAR

REACTIONS ‘H (polarized p, p), E = 0.94-2.44 GeV, measured I!?,polarization vs 0. Phaseshift calculation comparison.

asymmetry

vs

1. Introduction The measurement interference region information

of hadron-hadron (ItI = 0.005 (GeV/c)

on the various

nuclear

elastic scattering ‘) is an effective

amplitudes

in the Coulomb-nuclear method to obtain direct

at t = 0 through

their interference

with the well-known electromagnetic ones I). The measurement of the analysing power gives access to amplitudes other than those determined with the measurements of the differential cross-section, in particular to the spin-orbit one. These characteristics are important when studying the nucleon-nucleon interaction since the process involves five distinct amplitudes. This measurement provides information which puts a constraint on the solution of the phase-shift analysis 273) and the forward dispersion relation calculation 4), the usual phenomenological approaches to the nucleon-nucleon interactions at intermediate energy. ’ Present address:

University

of Tunis,

Tunisia.

0375-9474/89/%03.50 @ Elsevier Science Publishers (North-Holland Physics Publishing Division)

B.V.

562

S. Dalla Terre-Colautti et al. / Measurement of the polarization

Measurements

of the

spin

effects

in the Coulomb-nuclear

interference

particular,

results

Spectrometer

very

accurate

at LAMPF

We have measured

have

of the elastic region been

nucleon-nucleon

interaction

exist up to 800 MeV [refs. ‘,“)I. In obtained

with the

High

Resolution

“).

the polarization

of the elastic pp scattering

in the Coulomb-

nuclear interference region at 940, 1000, 1320, and 2440 MeV, using a polarized proton beam from the Saturne II accelerator at Saclay. We have put a constraint on the elastic reaction detecting the recoil proton in an active scintillator target and the forward-scattered proton with a telescope of multiwire proportional chambers (MWPC). The experimental apparatus is very simple and cannot provide the same resolution as a spectrometer, but it can be employed in a wide energy range. Thus, it has been possible to carry out measurements in the whole energy range available at Saturne in order to get information on the energy dependence of the pp analysing power at very small angles, extending the range of these data to an energy region which is now beginning to be explored systematically.

2. Method of measurement Already in the GeV region and at small t-values the measurement of elastic scattering is difficult since the momentum of the scattered particle is very close to that of the projectile and the slow recoil particles have short ranges. To study elastic scattering on protons in the Coulomb-nuclear interference region above 1 GeV/c we have used an active proton target ‘) consisting of a telescope of several thin slices of organic scintillator perpendicular to the beam, optically coupled to individual photomultipliers [multiscintillator target (MST)]. The target protons are the H atoms in the scintillator, and the use of this target allows determination of the recoil kinetic energy by pulse-height analysis. If the scattering angle of the forward particle is also measured, it is then possible to put a constraint on the elastic reaction. Moreover, the target signal can also be used to tag the interactions in the target for triggering purposes. At small ItI-values, the slow recoil protons are scattered at about 90” in the laboratory and, even if the individual target slices are thin (about 1 mm), they stop in the scintillator slabs releasing their whole kinetic energy. Owing to the reduced thickness of the single slice, the energy loss of the relativistic projectile and the forward-scattered particle is only a small fraction of the recoil kinetic energy and does not mask the recoil signal. Since, for fixed t-values, the recoil energy is independent of the incoming proton energy, this method can be applied up to very high energies. Several scintillator slices are used to increase the rate of the interactions. They are conveniently spaced to provide a geometric constraint on the location of the interaction vertex.

S. Dalla Terre-Colautti

The presence to be subtracted

of carbon under

nuclei

et al. / Measurement

in the target

the peak signal,

of the polarization

is a source

as explained

563

of background

in subsects.

that has

5.2 and 5.3.

3. The polarized beam - The polarimeter The Saturne II proton synchrotron can provide polarized protons with energy up to 2900 MeV. The beam has two antiparallel polarization states (t and &) and the beam polarization was reversed at each machine cycle (1 to 2 s). The beam polarization varied from 55% to 75% at the different energies of our data taking. The measurements were performed in the experimental area dedicated to the nucleon-nucleon facility “N-N”. The beam polarization was continuously measured and monitored during the data taking with the beam polarimeter developed for the experimental “N-N” programme “). Two sets of scintillator counter telescopes measured the left-right asymmetry of the elastic pp scattering using a thin CH2 target placed upstream along the beam line (the scattering angle was varied from 11.6” to 18.4” in the laboratory). The beam polarization was obtained from the of the known values of A,,, the analysing power of the reaction. The polarization two beam states could be obtained separately. The main source of systematic error arose from the slight dependence of the analysing power of the polarimeter on the beam parameters “): in fact the contribution of the quasi-elastic scattering off carbon to the elastic pp signal depends on the beam optics. From repeated measurements at 1000 MeV, we estimate the systematic error in the beam polarization at all energies to be less than 10%. At 940 and 1000 MeV independent information about the polarization in the machine was available from another polarimeter running on a different beam line (SPES I) lo). Table 1 gives the values of the average beam polarization during data taking obtained combining all available information; systematic errors are not included.

TABLE

Beam polarization

Beam energy (MeV)

(systematic

Polarimeter scattering angle lab system

940 1000 (I) 1000 (II) 1320 2440 There were two periods

(ded

1

errors

pp analysing power (percent)

18.4 18.4 16.1 16.1 11.6 of data taking

are not included)

45.5 40.1 43.1 34.6 24.5 at 1000 MeV.

Beam polarization P=$(PT+PJ) (percent) 66.2 f 52.5 * 84.1 f 70.7 * 58.1i

1.4 2.1 1.4 5.5 12.0

S. DaUa Terre-Colauffj

564

et al. / Measurement

4. Experimental

of the polarization

set-up and apparatus

The experimental set-up is shown in fig. 1. It consisted of the MST, two telescopes of MWPCs used to measure the scattering angle, and scintillator counters to provide the fast trigger to the acquisition system.

4.1. THE

MULTISCINTILLATOR

TARGET

The MST is a set of 12 slices of organic scintillator NE-102 of different thicknesses: OS, 1, 2 mm. Using three different thicknesses we have enlarged the measurable kinematical range. In fact, the thinner the scintillator slice, the lower the minimum (t(-value at which the reaction can be measured since the energy loss of the throughgoing particle decreases with the target thickness, but the maximum scattering angle at which the reaction can be measured with this device is lower: increasing the forward-s~atte~ng angle, the recoil particles are scattered at angles that become sensibly smaller than 90” and they start escaping from the target. The counters are viewed by individual PMs via an air light-guide and the optical insulation between adjacent scintillation slices is obtained by using 40 pm thick aluminium-coated Mylar foils. To set the HV supply for the PMs of the target, we have studied the response of each counter to relativistic particles. The measurement of the light output of a scintillator is expected to be optimal in the voltage interval where log Q (the integrated amplitude) varies linearly with log V (the high voltage of the PM). The Y

x

1

b

s L . : :

PC5

: : PC6

Fig. 1. The experimental set-up.

S. Dalla Torre-Colautti

et al. / Measurement TABLE

Parameters

Sensitive

Chamber

PC1 PC2 PC3 PC4 PC5 PC6

surface

95x95 95x95 95x95 95x95 190 x 190 510x510

1 1 1 1 2 2

modules

Diameter of the anode wires

Wire step (mm)

565

2

of the MWPC

(mm’)

of the polarization

Anode-cathode

(pm) 10 10 10 10 10 20

gap

(mm)

5 5 5 5 5 8

values of the voltage have been chosen for each PM so as to set the spectrum of the signal produced by the relativistic particles at the lower edge of the linear region.

4.2. THE

MWPC

TELESCOPES

Two MWPC telescopes (PCl-3, PC4-6) defined the incoming and trajectories. Each chamber module measured the coordinates in and horizontal planes. Table 2 gives the geometrical parameters The chambers were operated with magic-gas mixture and typical larger than 95%. The resolution of the measured scattering angle the different

4.3. THE

energies

FIRST-LEVEL

of the data taking

is listed in table 3.

TRIGGER

The trigger consisted of a two-level counters, the second by a programmable built up by the signals

outgoing particle both the vertical of each chamber. efficiencies were as evaluated for

selection, the first given by scintillation microprocessor. The first level trigger is

of the trigger counters

and the MST itself.

The beam was defined by counters Fl and F4, while F2 and F3 rejected the beam halo; the non-interacting particles were vetoed by counters F5 and F6. Scattered particles were detected using counters S, L, and S’ (S’ was only used at 2440 MeV). TABLE

Resolution Beam energy (MeV) 940 1000 1320 2440

3

of the scattering-angle Scattering

measurements angle resolution (mrad) 2.6 2.6 1.7 1.0

S. Dalla Torre-Colautti et al. / Measurement of the polarization

566

The target interacting

counters

particles.

were

used

adjusted

so as to reject,

particles.

Only one target signal

4.4. THE

SECOND-LEVEL

to discriminate

For this purpose,

the recoil

the thresholds

for each individual

counter,

above the threshold

protons

from

of the discriminators

nonwere

97% of the non-interacting was required.

TRIGGER

The rejection of unscattered events is not very efficient because the beam,divergence and the target length prevent the veto counter from defining a scattering angle precisely. On the other hand, a precise cut in the scattering angle is needed to separate the events in the Coulomb-nuclear interference region where the cross section is rising very rapidly. Such a cut is obtained by using a fast microprocessor running on-line and employing an algorithm applying a sharp cut on the scattering angles evaluated from the MWPC coordinates ‘I). The MWPC readout allows fast on-line data pre-processing by the microprocessor. We have used the RMH system developed at CERN “), whose main characte~sti~s are high-speed ECL 10K technology and the capability of non-destructive readout allowing repeated readings. Dead times are minimized by the choice of cables as delay elements, and the maximum data-transfer rate is about one l&bit word per 100 ns, each corresponding to an encoded proportional-chamber wire hit. The microprogrammable processor ESOP 13) built at CERN operates on 16bit data words; data and instructions occupy separate memories and three independent arithmetic Logic units perform in parallel operations on data, data memory addresses, and instruction memory addresses. The basic cycle time is 125 ns. A fast 16 x 16-bit multiplier, with an accumulator and a special shift unit, enhances its computing power. In our configuration, 4096 words of data memory and 1024 words of 4%bit instruction memory were available. The CAMAC interface gives access to ESOP memories and registers. An interface connects the RMH system directly with one of the three independent I/O ports of the data buffer memory. ESOP is only programmed in machine language with the help of a cross-assembler running on a host computer. All development tools and the assembler were installed on the on-line computer. The straight-track rejection algorithm uses the coordinates measured by four MWPCs (PCl, PC3, PC4, and PC6) to calculate the scattering angle 8, which is then compared with a threshold value 13~. First the centres of clusters of adjacent hit wires are calculated. Then two angles 0, and f+, [projections of 8 on the planes (x, z), (y, z)] are computed. Any of the following conditions produces a rejection: (i) no hit in any of the four chambers of the x(y) projection [OX($) not computable]; (ii) more than one cluster in the x-plane (y-plane) of the chambers 1 and 3 (incoming trajectory not uniquely defined);

S. Dalla Torre-Colautii et al. / Measurement

(iii) many (iv)

more than three clusters possible

values

in the x-plane

of the polarization

(y-plane)

of chambers

567

4 and 6 [too

for 19,(13,)];

8f,+0:<8:.

Fig. 2 shows

the ratio of the number

of accepted

two values of & (20 and 35 mrad) evaluated

events

to the total events

off-line as a function

for

of the reconstructed

scattering angle for a sample of real events (beam energy: 1000 MeV) collected without applying the second-level rejection, but recording the ESOP rejection flag. The sharp cut is due to the good resolution of the angle as evaluated by the ESOP algorithm (g = 2.5 mrad). The ratio never reaches 100% owing to the inefficiency of the MWPCs. The transfer time from the read-out system to the processor data memory is, on an average, 300 ns per 16-bit word. Average ESOP decision time is 70 p,s for accepted events and 90 us for rejected ones (to be compared with the typical on-line computer acquisition dead time of about 2-3 ms per event). Fig. 3 illustrates the overall acquisition system; the block diagram is shown together with an indicative time scale.

1.0 0.8 -

ia)

‘,‘lIlll’,

,‘I/ 0.6 0.4 -

/

0.2 0.0

“““I

I

I

I

I

I

I

1.0

0.8 - (b)

11I/ ,l//LI,i/,‘l ,

,,~,+,‘,I’lI’l,

0.6 I 0.4 0.2 -

I

0.0 “““““’



0

40

I

Scattering

I 80 Angle

I

I

120

I 160

(mrad)

Fig. 2. Ratio of the number of events accepted by the ESOP algorithm to the total (accepted + rejected) as a function of the off-line reconstructed angle for two values of the threshold: (a) 20 mrad, (b) 33 mrad. (Data collected at 1000 MeV).

568

S. Dalla Tone-Cotuurti

et al. / ~ea~~rerne~f

of the po~arizafio~

DIAGRAM OF THE ACQUISITION SYSTEM 1 ELECTRONICS RESET (CA~AC,MWP~ REMOVE TRIGGER INHIBIT

R/O)

I PHYSICAL EVENT OCCURS

t-o ______-.. __

t-250ns ___ - - __ - __.

YES

r t=26Ons _-__----

_ INFORMATION

t=350ne -, _____-_-

t-7p3 __-___-_ t=70ps __- - ___._._

HP ON-LINE COMPUTER READS CAMAC ELECTR. AND MWPC COORDS

t = 2.5 ms _______.

..I_

TIME

Fig. 3. Diagram of the acquisition system.

The rejection

ratio provided

Table 4 gives typical

examples

by the ESOP filter depends of the rates recorded

on the running

during

conditions.

data taking.

4.5. THE DATA TAKING

For each

event

recorded: (i) the MWPC

selected coordinate

by the two-level hits,

trigger,

the following

information

is

S. Dalla Torre-Colautti

et al. / Measurement TABLE

Typical

of the polarization

4

rates (data taking

at 940 MeV) Rates (s)

Events

1.1 x lo6 7.5 x lo3

beam first-level trigger first-level trigger anticoincided by acquisition dead time = events processed by ESOP events on tape = events not rejected by ESOP

(ii) the flags corresponding

569

to the MST counters

1.5 x lo3 0.25 x lo3

in which

a signal

larger than

the discriminating threshold has been produced, (iii) the integral charge of the pulse of each target counter measured by individual analog-to-digital converter (ADC) channels, (iv) the scalar information counting the pulses from the counters and the coincidences used to form the trigger. The information was transferred via CAMAC to the on-line computer HP 2109-A. Data were buffered and stored on magnetic tapes. The maximum acquisition rate was about 280 events per second. For each machine burst the state of the beam polarization is also recorded together with the counts of the polarimeter scalers. For calibration and checking purposes, we have also collected samples of straighttrack events (about 3% of the total statistics) and ESOP test events obtained recording the ESOP rejection flag without applying the second-level trigger (about 3% of the total statistics).

5. Data analysis The data reduction (i) the geometrical (ii) the application (iii) the identification (iv) the determination

5.1. THE

GEOMETRICAL

includes the following steps: reconstruction of the events, of the off-line criteria in order to reject the background of the pp elastic scattering of the asymmetry.

RECONSTRUCTION

OF THE

events,

events,

EVENTS

The incoming and outgoing particle trajectories are reconstructed using the track coordinates measured by the MWPCs. At this stage, the events are rejected if: (i) it is not possible to reconstruct one of the two trajectories (too many missing coordinates),

570

S. Dalla Torn+Colautti

et al. / Measurement

of the polarization

(ii) there is no candidate interaction point (defined as the common vertex, when it exists and it lies within a loose target fiducial volume, between one incoming and one outgoing trajectory), (iii) there is more than one candidate interaction point, (iv) the scattering angle, i.e. the angle between the two trajectories, is smaller than 5 mrad. For the remaining events, the polar scattering angle 8, the azimuthal angle rp defined as the angle between the normal to the scattering plane and the vertical axis (the beam polarization direction), and the interaction point coordinates (x,, y,, z,) are computed and written onto data summary tapes (DST) (see table 5, fourth and fifth columns for details about the statistics). 5.2. THE

BACKGROUND

STRUCTURE

AND

THE

OFF-LINE

REJECTION

CRITERIA

The elastic pp events are characterized by the large pulse amplitude produced in the target counter where the interaction has taken place. The position of this target counter and the reconstructed vertex must coincide within the experimental resolution. The background events are essentially due to interactions on the carbon nuclei present in the target. Most of the interactions on carbon are elastic and a large fraction of events are rejected already at trigger level since the kinetic energy of recoiling C nuclei is much smaller than that of recoiling protons. However, some of them exhibit a large pulse amplitude due to the fluctuations of the energy loss of the throughgoing protons in any of the target counters. For such events the position of the counter, where the signal above threshold has been produced, and the reconstructed vertex are not correlated. To reject them the following condition is imposed:

where T, is the z-coordinate of the centre of the target counter exhibiting a signal larger than the threshold, and cr, is the standard deviation of the distribution of the variable (z, - TZ) tan 8. a, has been determined using a subsample of data exhibiting TABLET Statistics Beam energy

of the collected

data

(MeV)

Useful beam xlOh

Angular range of the measurement (mrad)

Events on data tapes x106

Events on DSTs [email protected]

Events on mini-DSTs x103

940 1000 1320 2440

480 375 709 623

46- 100 37-100 32.5-86.5 26-53

3902 2844 3628 2700

2580 1840 3244 1944

900 670 1209 779

Identified elastic pp events x10’ 100 80 56 27

S. Dalla Torre-Colautti

a very large scattering

target

as much

pulse

amplitude

as possible.

zero and uZ is independent This procedure which

is rather

to reduce

of 0. The values at small

events in which slow particles

are produced.

by elastic

pC

of (z, - TZ) tan 0 are centred

obtained

for a, are shown

of the inelastic

scattering

571

the contamination

The distributions

does not allow the rejection reduced

ofthe polarization

et al. / Measurement

angles,

scattering

but is largely

These particles,

on carbon,

responsible

scattered

at

in table 6. for

at any angle,

leave part or all of their energy in the target thus generating large pulses. Other cuts are applied to reject events for which the analysis would in any case be ambiguous: overflow in target pulse amplitude measurement, too small a scattering angle, beam polarization not known. The stability of the apparatus is also checked using the information from the scalers and the specta of the target pulses produced by the straight-track protons. When fluctuations were observed the corresponding data have been rejected. For each of the selected events the relevant physical information is written onto mini-DSTs (table 5, sixth column). After grouping the events according to the target counter on which the scattering occurred, for each of the 12 targets the correlation Q - 0 has been looked at by plotting all the events as points in a (Q, f3) plane (an example is given in fig. 4). The band of the elastic pp events is clearly visible is spite of an important background under the signal.

5.3. THE

IDENTIFICATION

OF

To perform the background two independent procedures

THE

ELASTIC

EVENTS

subtraction under the elastic peak we have exploited having different and complementary aims:

- to obtain results from events having good kinematical definition, - to extract information from the data at the smallest possible angles. In the former procedure, the events relative to each target counter separately. As a consequence, energy is required, avoiding

no calibration of the measurement the introduction of uncertainties.

TABLE Standard

deviation

6 of the variable

(z, - T,) tan 0 Energy (MeV)

flz (mm)

940

1.27

1000

1.27

1320

1.20

2440

1.14

are studied

of the recoil kinetic

S. Dalla Terre-Co~auiti

572

et al. / ~eas~~e~ent

of the polarization

80

60 -

0”“““““““““““” 0

200

400

600

800

1000

Q (ADC units) Fig. 4. Plot on the (Q, 0) plane of the mini-DST events (1320 MeV) exhibiting a signal larger than the threshold in the 6th target counter (thickness: 1 mm).

The method consists in the best fit of a function of Q and 0 to the two-dimensional distribution in (Q, 0). The function is the sum of two terms which describe the spectrum of the elastic pp events and the distribution of the background events. The reliability of the background subtraction procedure is assured since the function is unique and continuous in the (Q, 0) plane. The fit operates on two-dimensional histograms that have 100 Q bins (width: 10 ADC channels) and a variable number of 8 bins (width: 4.5 mrad; 3.0 mrad for 2440 MeV data). The least squares method is applied. For a given 8 bin, the dist~bution in Q of the elastic events is assumed to be gaussian with a central value x0(0), standard deviation a(e) and integral N(8); x,(e) and (T(0) are fixed functions of @preliminarily determined from the event distributions (see fig. 5). N(B) is defined as

where Bi is the central value of the ith bin. Ni are parameters to be determined with the best-fit procedure. There are also elastic events with recoil protons which do not stop in the scintillator target. According to the results of a simple Monte Carlo calculation, we have assumed that they have a flat distribution in Q for Q
S. Dalla Torre-Colautti et al. / Measurement of the polarization

8 Fig. 5. The values

known

fraction

573

(mrad)

of the parameters x0 and (T from the projected distribution (1320 MeV, target no. 6). The plotted curves of x0 and l/a are the interpolating functions.

of N(B)

and their contribution

to the overall

distribution

is also

taken into account. For each 8 bin, the background is described by the sum of a decreasing exponential function and a linear term. The slope of the exponential is a fixed function of 8 preliminary

determined

from the event

distributions.

The two coefficients

of the

linear term are linear functions of 8, depending on four parameters which have to be determined with the best-fit procedure. To summarize, the fitted parameters are 4 plus the number of the 0 bins in the two-dimensional histogram. Fig. 6 shows a typical result of the fitting procedure. The distribution of the minimum value of the ,y2 divided by the number of degrees of freedom (150 to 800) has a mean value of 1 at all energies. The reliability of the fitting procedure has been tested by dividing the available data into two subsamples and analysing them separately. The fitted values of the parameters always coincide within statistical errors. The latter procedure to identify the elastic events operates on distributions obtained by adding together the data of target counters having the same thickness. A relative calibration of the measured recoil energy is needed. In the low Q-range it is obtained from the spectra of the throughgoing particles defining 9 percentiles spectrum and, in the high corresponding to lo%, 20%, . . . ,90% of the integrated

S. D&la Terre-Colautti et al. / Measurement of the polarization

574

150

75

0 150 50.5 - 55.0 mrad

75

0 150

I

55.0

II

59.5 mrad

59.5 - 64.0 mrad

75

0

0

500

1000

‘5o------

64.0 - 68.5 mrad

\

0

1000

500 Q

Fig. 6. Results

(arbitrary

units)

of the fitting procedure, 1320 MeV, target counter no. 6. The two-dimensional is shown by 0 rows. The superimposed curve represent the fitted function.

histogram

S. DU&I Terre-Coiautti e$ al, / ~e~s~~e~en~

of the po~u~~~~~~~~

575

Q-regian, from the elastic pp signal fixing six values of the scattering angle and determining the corresponding Q-values. The relative calibration is obtained from a linear fit of this set of Q-values to a reference set. To operate the background subtraction the events are divided into Q-bins and the popufations are projected onio the &axis. An example of the corresponding 8 distribution is shown in fig. 7a where the elastic pp events are grouped in a peak standing above an appreciable background. The background has a smooth trend and has been described with a polynomial interpolation of the distribution outside the peak interval. The result is represented by the curve in fig. ?a. For each distribution the degree of the polynomial is optimized using the Fischer test. Fig. 7b shows the distribution obtained after applying the subtraction procedure. Using this procedure the elastic pp peak can be identified also at very low Q-values corresponding to small scattering angles down to 20 mrad. In particular, to enhance the elastic peak at low Q-values, a more restrictive z, cut was imposed. For each sample the kinematics of the elastic pp reaction is defined using the central value of the peak in the angular distribution.

2400

800

0

(mrad)

Fig. 7. The angular distribution of the events of a typical Q-band (940 MeV, 1 mm thick target counters) before (a) and after (bf the background subtraction. The cwxe in (a) is the interpolation of the background distribution.

576

S. Da&

The two samples separately

of elastic

analysed

to obtain

5.4. THE ASYMMETRY

The experimental

Torre-Cola~ti~ ef ul. / ~e~sure~en&

events

identified

of the poiariznfion

by the two procedures

have been

the asymmetry.

EXTERMINATION

asymmetry

~(0) is given by: ~(0) = PA(@),

where P is the beam polarization and A(@) is the analysing power of the proton target for elastic scattering. E( 0) has been estimated using two different algorithms. A first estimator

of ~(0) is given by

where RT, RJ, LT, and LJ are the rates of the elastically scattered protons at an angle 0 normalized to the incoming beam, for the two beam polarization states (T and 4) and for left (LT, L&: -$r < cp
A(da/dt)/(dcT/dt)62%).

Since ~(@=l’A(@)-t-0 the estimated

upper

APAA APA(du,‘dt) -P A ’ P (dc,‘dt)

limit of the systematic

A(da/dt) ’ (dcfdt)

AA A

1

error is 0.04%.

The uniformity in (c of the response of the apparatus has been tested: no effect has been observed within statistical resolution. The data have also been divided into 2r? (n = 3 or 5) azimuthal bins having central values y.+ (i = 1,2,. . . , 2n) and width Aq. The rates of the elastic scattered protons at an angle q~ for the two beam polarization states are Ltj and Lii for left scattering (qi -~Au, < rp < pi +iAp) and Rti and R&, for right scattering (qc2n+r-i) -+Ap < cp< ~~2n+l-i)+$A~)* Ei( 0) is defined as

S. Dalla Torre-Colautti

The second to the ~(0)

algorithm

to estimate

et al. / Measurement

E( 0) consisted

values 14). This algorithm

rates to the incoming

of the polarization

577

in fitting the function

does not require

beam and it checks the sinusoidal

E( f3) cos cp

any normalization

of the

shape of the rp distribution

(see fig. 8) expected for a 2~ detector with uniform efficiency. At each energy the s(B)-values obtained from the various data samples

(single

target counters or groups of target counters) have been statistically combined. In fig. 9 the values of E( 19) at 940 MeV obtained using two different procedures are presented. The dots are the results obtained using the first method to identify the elastic pp events (subsect. 5.3) and employing the first algorithm to determine the asymmetry. The open circles are the values obtained employing the second procedure for the identification of elastic pp events (subsect. 5.3) and the second algorithm to estimate the asymmetry. In spite of the different data reductions the results are fully compatible. Similar compatibility has been obtained for all the data sets. The analysing power of the background events has been evaluated also. As explained in subsect. 5.3, the samples of events (LT+RJ), (LJ+ RT) have been analysed separately to obtain the asymmetry, therefore from the two samples

A (%)

20

10

0

-10

- 20

Fig. 8. The experimental

values of ei and fitted function at 1000 MeV and 48.8 mrad dashed line) and 105.0 mrad (dots, full line).

(open

circles,

578

S. Dalla Tome-Colautti

et al. / Measurement

0.6 Tbeam

s ‘z= : % .I?

0.4

-

0.2

-

=

of the polarization

940

MeV

$+*A*+ +*

0.0 0

Fig. 9. The comparison two analysis procedures

m

40

of the values obtained described in subsects.

of data the asymmetry

I

+,

120

for the experimental asymmetry 5.3 and 5.4 (dots: first method; method).

of the background

has been derived.

at 940 MeV using the open circles: second

Table

7 gives, for the

four energies, the analysing power of the background events in the (0, 0) plane which are close to the elastic pp band (a *2a cut has been applied). A cut at Q > 300ADC units has been imposed to eliminate the tails of the elastic pC scattering events. As expected for quasi-elastic pp events, the background events exhibit an important fraction of the polarization measured for the elastic pp events. 6. Results

The measured values presented in fig. 10 and are statistical only. The acts as a scale factor. In fig. 11 the data are energies.

of the analysing power A(B) of the elastic pp reaction are the numerical values are given in table 8. The quoted errors uncertainty in the measurement of the beam polarization plotted

together

with existing

These data exist at larger scattering

measurements

15) at the same

angles only. The curve is the analysing

power of the reaction as calculated with the Saclay phase-shift analysis 2, (at 2440 MeV this calculation is not available); at 1000 MeV the prediction of the Arndt TABLE I Analysing Beam energy

power

of the background

(MeV)

Angular range lab system (mrad)

940 1000 1320 2440

30.1-110.2 19.1-112.1 27.3 - 87.4 17.3-56.1

events

Analysing power of the background (percent) 6.57 + 0.81 5.46+ 1.14 7.58 f 0.95 4.57* 1.51

0.6 Tbeam 0.3

=

1000

*

.*s i*

Tbeam

=

l*

I*+*’

MeV **

I

t

I

_

0.6

1320

MeV

Tbeam

l

I

=

2440

MeV

*i**

f

f’

0.0

- 0.3

Tbeam

sz

0.0

0.3

MeV i

I

0.3

940

=

0

I””

I

I

40

80

120 6 jab

0

I?

f

I

I

40

80

120

fmradf

Fig. 10. The measured asymmetry of the elastic pp reaction.

phase-shift analysis 3, is also shown. The Coulomb amplitude is included in the calculation and is responsible for the very small values of A( 0) at small angles but before @= 0. These calculations do not use the results presented in this paper. At 940 and IO00 MeV the small scattering angle data and the larger angle measurements show a smooth continuous trend. At higher energies the angular behaviour of the analysing power is not campletely determined owing to the lack of data at angles lower than 25” c.m. Up to 1320 MeV the phase-shift analysis predictions describe we11 also the smatt- and large-angle behaviour. 7. Conclusion We have described an experimental method to perform polarization measurements at low momentum transfer (Coulomb-nuclear interference region) and in a wide energy range. This technique has been originally proposed to determine the polarization of a high-energy polarized proton beam 16).The use of such a Coulomb-nuclear interference polarimeter is suitable for high-energy applications since -the analysing power of the elastic pp scattering in this kinematical region can be evaluated r7) (absolute calibration for the poiarimeter) -the experimental method is simpIe enough to allow the measurement and monitoring of the beam polarization.

580

S. Dalla Torre-Colautti

et al. / Measurement TABLE

The measured analysing in the beam polarization

Energy WV) 940

1000

1320

2440

of the polarization

8

power (statistical errors only); uncertainty introduces an overall scale factor up to 10% (not included)

Angular bin lab system (mrad)

Analysing power of the reaction (percent)

30.1-38.0 38.3-46.9 47.9-55.8 56.9-65.5 67.8-78.2 79.3-89.1 90.6-98.6 99.5-110.2 19.1-24.9 23.5-32.5 32.7-39.5 38.2-45.6 44.0-51.1 50.7-57.0 56.5-63.2 62.8-69.6 68.9-75.6 74.3-81.0 79.7-88.8 90.3-99.2 99.2-112.1 27.3-24.9 34.4-40.5 42.1-48.6 50.7-57.6 57.8-66.6 67.0-75.6 79.4-87.4 17.3-22.7 22.3-28.0 28.4-34.2 35.6-41.5 40.6-46.9 50.1-56.1

5.68k2.71 11.69* 1.55 16.92+2.43 19.49* 1.45 22.06* 1.41 24.83 f 1.86 25.12*2.23 33.26 f 2.10 5.89 * 3.30 s.3s*3.10 12.26 * 2.70 10.65 * 1.02 13.98 * 1.52 14.06+ 1.44 15.67* 1.27 19.40* 1.22 21.5911.21 24.72* 1.15 24.03 * 1.06 26.53 * 1.49 32.69* 1.59 7.49k2.71 10.35* 1.74 14.31* 1.69 15.88* 1.68 16.02* 1.48 18.37 zt 2.26 19.14*3.20 6.19h3.92 8.75 f 2.35 10.33 * 2.00 12.40* 1.83 13.94* 1.97 11.91*3.45

The measurements performed at Saturne constitute a complete test of the method. They have also provided interesting physics results since the elastic pp analysing power at very small scattering angles is now available for the first time at 940, 1000, 1320, and 2440 MeV. Previous results were limited to 800 MeV. We thank G. Menon who built the wire-chamber telescopes at the Sezione di Trieste of the INFN and helped during the installation of the experiment, the

S. DalIa Terre-Colautfi

et al. / ~easureme~f

of the ~oiarizat~o~

90

30

581

0.6

0.6

0.2

- 0.2 0

30

60

0 cm

0

60

90

Wwes)

Fig. 11. The results of this experiment are shown together with the existing data at the same energies. In (a), 0: this experiment (940 MeV), 0: Albrow et al. (951 MeV), A: Cozzika et al. (924 MeV). In (b), 0: this experiment (1000 MeV), $: Homer ef al. (970 MeV), +: Vovchenko et al. (970 MeV), 0: Cozzika et al. (1029 MeV), a: Marshak ef al. (1030 MeV), 0: Neal et al. (1030 MeV). In (c), 0: this experiment (1320 MeV), 0: Bystricky et nl. (1295 MeV), A: Neal et al. (1320 MeV), n : Albrow et al. (1344 MeV), @: Zhurkin et al. (1361 MeV). In (d), 0: this experiment (2440 MeV), m: Bystricky ef nt. (2394 MeV), 0: Neal et aZ. (2440 MeV), A: Parry et al. (2444 MeV). In (a), (b) and (c) the thin-line curve is in prediction for the analysing power of the pp elastic scattering by the Saclay phase-shift analysis *), in (b) the thick-line curve is the prediction for the same parameter by Arodt’s phase-shift analysis 3). The calculations include the Coulomb amplitude.

workshop staff from LAPP ior the construction and installation of the scintillator target, and R. Stoicovich, D. Rey and G. Moynot for their skilful assistance. The contributions of M. Di Drusco and P. Moras at an early stage of this experiment are here acknowledged. We thank the Nucleon-Nucleon Collaboration, in particular C.R. Newson and F. Perrot, for their continuous support and encouragement and we are grateful to J. Thirion, Director of the Laboratoire National de Saturne for kind hospitality. The assistance of the Saturne Synchrotron staff is greatly appreciated. References 1) L.I. Lapidus, Yad. Fiz. 17 (1973) 592; N.H. Buttimore et aL, Phys. Rev. D18 (1978) 694; C. Lechanoine et aL, Nuovo Cim. 56A (1980) 201

582

S. Dalla Tom-Colautti

et al. / Measuremenr of the polarization

2) J. Bystricky et al., J. of Phys. 48 (1987) 199; F. Lehar et al., J. of Phys. 48 (1987) 1273 3) R.A. Arndt et al., Phys. Rev. D28 (1983) 97; R.A. Arndt et al., Internal Report, Virginia Polytechnic Institute and State University, Blackburg, 1986 4) W. Grein and P. Kroll, Nucl. Phys. B137 (1978) 173; W. Grein and P. Kroll, Nucl. Phys. A377 (1982) 505 5) M.W. McNaughton et al., Phys. Rev. C23 (1981) 1128; F. Irom et al., Phys. Rev. C25 (1982) 373 6) G. Pauletta et al., Phys. Rev. C27 (1983) 282; M.L. Barlett et al., Phys. Rev. C30 (1984) 279; M.M. Gazzaly et al., Phys. Rev. Lett. 58 (1987) 1084 7) H. Azaiez et al., Nucl. Instr. Meth. 211 (1983) 335 8) J. Bystricky et al., Nucl. Instr. Meth. A239 (1985) 131 9) F. Perrot, PhD thesis No 2912, Orsay (1984) 10) B. Mayer et al., Nucl. Phys. A437 (1985) 630 11) R. Birsa et al., The microprogrammable processor ESOP in a small angle elastic scattering experiment, Report lNFN/AE-82/10, September 1, 1982 12) J.B. Lindsay et al., Nucl. Inst. Meth. 156 (1978) 329 A fast microprogrammed processor, Internal Report CERN/DD/75-17; 13) T. Lingjaerde, T. Lingjaerde and D. Marland, A versatile multiport buffer memory system for data acquisition in high energy physics, Internal Report CERN/DD/78-8; D. Marland, ESOP update, Internal Report CERN/DD/78-3 14) H. Azaiez, PhD thesis, Grenoble (1987) scattering data, Fachinformationszentrum Karlsruhe, 15) J. Bystricky and F. Lehar, Nucleon-nucleon ed. H. Behrems and G. Ebel, No. 11-1, 1978, Nos. 11-2 and 11-3, 1981; J. Bystricky et al., Nucl. Phys. B258 (1985) 483 16) A. Apokin et al., Study of spin effects at SPS energies using a polarized proton beam, Proposal CERN/SPSC/77-61, SPSCf P 87 (1977); H. Azaiez et al., Proc. Int. Symp. on high energy physics with polarized beams and polarized targets, Lausanne, 1980, ed. C. Joseph and J. Soffer (Birkhauser, Basle, 1981), p. 497 17) N.H. Buttimore, E. Gotsman and E. Leader, Phys. Rev. D18 (1978) 694