Nuclear Instruments and Methods in Physics Research A 381 (1996) 4- 14 NUCLEAR INSTRUMENTS

& METHODS IN PHVSICS RESEARCH

ELSEVIER Secr~on A

Proton and neutron polarized targets for nucleon-nucleon experiments at SATURNE II

J. Ball”‘*, M. Combet”, J.-L. Sawa, B. Bendab, P. Chaumetteb, J. Dertgelb, G. Durandb, A.P. Dzyubakb.’ , C. Gaudronb, F. Leharb, A. Lesquenb, T.E. Kasprzyk”, Z. Janoutd’2,

B.A. Khachaturovd, V.N. Matafonovd, Yu.A. Usovd

“Laboratoire National SATURNE. CNRSIIN2P3 et CEAIDSM, CE-Saclay. F-91 191 Gif-sur-Yvette Cedex. France

hCEAIDAPNIA, CE Saclay, F-91191 Gif-sur-Yvette Cedex, France ‘ANL-HEP, 9700 South Cass Ave., Argonne. IL 60439, USA

‘Laboratory of Nuclear Problems. JINR, 141980 Dubna, Moscow Region, Russian Federation

Received 13 February 1996; revised form received 23 April 1996

Abstract A SATURNE polarized target has been used for nucleon-nucleon elastic scattering and transmission experiments for 15

years. Continuous improvements resulted in a flexible and reliable facility for spin physics. The polarized proton target was a 70 cm’ cartridge loaded with pentanol-2, a promising material according to the results obtained. The new acquisition system

and data processing was based on the LabView/PC software and increased the accuracy of polarization measurements. For a polarized neutron target, two cartridges loaded with bLiD and 6LiH were set in the refrigerator and could be quickly inserted

in the beam. Results from the the first experiments using ‘Li materials in quasielastic pp or pn analyzing power

measurements are compared with the same observables measured in free nucleon-nucleon scattering using polarized proton

targets. The angular resolution distributions for &,, determination and azimuthal coplanarity are shown for different targets

in nucleon-nucleon scattering. A comparison of analyzing power results for elastic and quasielastic scattering suggests that the contribution of inelastic processes to quasielastic pn scattering may be suppressed by additional constraints.

1. Introduction

Different kinds of polarized targets have allowed mea-

surements of new spin-dependent observables. The “neu-

tron” targets enable testing of fundamental laws, e.g. a check of the Bjorken sum-rules in the CERN-SMC experi- ment. The targets have no substitute in experiments with

electrons, muons, pions, kaons and hyperons. They are unique existing tools for antinucleon-nucleon interactions at low and intermediate energies, where no polarized antiproton beams exist. They help to expand the energy range of elastic and inelastic np experiments, since beams of free polarized neutrons are rare and have a maximal energy of 3.7 GeV at JINR-Dubna.

* Corresponding author. E-mail ball@frcpnl l.in2p3.fr.

’ Present address: Kharkov Institute of Physics and Technology,

Akademicheskaja str.1, 310108 Kharkov, Ukraine.

*Present address: Faculty of Nuclear Sciences and Physical

Engineering, Czech Technical University, Bfehova 7, I 15 19

Prague 1, Czech Republic.

Proton polarized targets (PPT), loaded with aliphatic

alcohols, have been used for numerous elastic pp and np measurements. The free hydrogen content of this type of

target is about 15%, the remainder being carbon and oxygen. These last nuclei are practically unpolarized, but they contribute to the different cross section of the target material and increase the overall energy-dependent back- ground. This background must be determined in dedicated

measurements with a hydrogenless “dummy” target. The contribution of unpolarized nuclei may be removed by measurement with two opposite target polarization direc- tions and by a background subtraction.

Since a free neutron polarized target does not exist, compound nuclei must be used for scattering of different particles on neutrons. The binding energies of these nuclei have to be as small as possible. Some of the smallest binding energies in nature are those of the deuteron and the 6Li nucleus. For a long time, deuterated aliphatic alcohols were used as polarized deuteron targets (PDT). Their disadvantages are a poor content of quasifree neutrons, and difficult of polarization measurement due to the quad- rupole-split NMR signal. A sizeable step forward was the

0168.9002/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved

PII SO168-9002(96)00642-O

J. Ball et al. I Nucl. Instr. and Meth. in Phys. Res. A _%I (1996) 4-14 5

introduction of deuterated ammonia, which has a higher

neutron content and stronger resistance to radiation dam-

age. In these targets the nitrogen’ nucleus is also polarized and its polarization has to be taken into account.

Another class of material, lithium compounds, seems to

be a good choice for present experiments. Both ‘Li and ‘Li isotopes may be used.

‘LiH with 12.5% hydrogen content is less advantageous

than aliphatic alcohols. Moreover, this product cannot be used as a neutron target since ‘Li is a strongly-bound nucleus.

A ‘LiD target, having an equal number of protons and

neutrons. is more suitable for that purpose, and the binding energies of ‘Li and D are comparable and small. Because

their Larmor frequencies are nearly equal, they are both polarized. The face-centered cubic crystal lattice guaran-

tees the absence of an electric field gradient. Therefore the NMR signals have a simple line shape.

One may assume that ‘Li behaves as bound “He and D.

In order to distinguish between the polarized neutron in “Li and the polarized neutron in D, it is convenient to use a

‘LiH target. For this target, spin effects in pn scattering of a proton beam on ‘LiH are due to polarized neutrons in ‘Li

only. Thus. ‘LiH may be considered as a kind of dummy target for pn scattering on ‘LiD.

In order to be polarizable. the Li materials have to be irradiated by an electron beam to create “colour centers”

[l-3]. These centers play the same role as the para- magnetic centers for chemically doped products. The

irradiation is similar to that used for ammonia [4-61. Here we discuss the irradiation of Li materials, the

polarization process in a 2.5 T field, and target polarization measurements. Investigations on these materials as well as

scattering measurements were done in different laborator- ies [7-IO]. For nucleon-nucleon elastic scattering some problems arise. which are shown as examples.

2. Polarized target set-up

The SATURNE frozen spin polarized target used in the nucleon-nucleon program has been evolving for 15 years to enable convenient measurements of spin observables. With respect to the description in Ref. [ll], several improvements were introduced. They are discussed below.

-3.1. Turget refrigerator and magnets

The ‘He/‘He dilution refrigerator was designed for carrying two types of targets according to the choice of polarized nucleons. For the PPT, the mixing chamber contained a 70 cm’ parallelipedic Voltalef cartridge filled with pentanol beads chemically doped with EHBA-Cr(V) [ 121. For the PDT, two 14 cm’ cylindrical cartridges of

2 cm diameter were set horizontally in the mixing chamber

one below the other. The vertical gap between the two

cartridges was 1 cm. A mechanical positioning system

linked to the refrigerator allowed movement of the re- frigerator upwards and downwards. The upper or the lower cartridge could then be put at the beam level. One of the

cartridges was filled with ‘LiD and the other with “LiH in order to study scattering on D and ‘Li separately, as is

explained in Section 1. The cooling power of the refrigerator was 300 mW at

300 mK and the frozen spin temperature was 40 mK. In

polarizing mode. flows were 29 mmoles/s ‘He and

23 mmolesls ‘He; in frozen spin mode they were 7 and 4 mmoles/s respectively. Two ruthenium oxide resistors

(RO-600 Scientific Instruments Inc.). used as temperature

sensors. were placed above and below the target cell. One of the resistors, calibrated by Lake Shore, enabled the calibration of the other. The choice of these probes was

determined by their insensitivity to magnetic fields, namely 1.47r at 2.5 T. The 2.5 T polarizing field was produced by a solenoid. The lield homogeneity was 10 A in a 70 cm’

volume around the target center. The current stability of

the power supply for this magnet was 10 “. A vertical and

a horizontal holding magnet kept the target polarization in the required direction in frozen spin mode. They produced a 0.33 T field at the target center. Their operating mode is

thoroughly described in Ref. [I I].

2.2. NMR meumrement svstem

The target polarization was measured classically by NMR [ 13-161. For the large 70 cm’ cartridge, two single-

turn rectangular coils 40 mm wide and 49 mm high were

placed at each end of the target. They were used to produce the small field orthogonal to the polarizing field

for generating the NMR, as well as to mechanically hold the cartridge. For the twin cylindrical target set. there was one coil per target. Each of these coils consisted of 5 turns

embedded in a 21 mm diameter and 1 mm thick Voltalef

tube. This device assured good geometrical stability and carried the cartridges. The tuning of the system depended

on the resonating frequency of the materials investigated. One could choose a number of turns for the NMR signal, since a contact point was available for each turn. One turn was enough for 106.5 MHz needed by protons at 2.5 T. five turns were necessary for deuterons at 16.35 MHz at the same field. The target cartridges were held by the

hollow Voltalef cylinders in which the coils were set. The general schematic of the NMR measurement system

is shown in Fig. 1. It consisted of a resonating LCR circuit connected to “Liverpool type” Q-meters (Ultra Physics Ltd.) [17] through semi-rigid line. RF frequency sweep was produced by a synthesizer (Marconi 2040). The data were processed through an acquisition board (National Instruments ATMI016X) of a PC (HP Vectra 486). The

J. Ball et al. I Nucl. Instr. and Meth. in Phys. Res. A 381 (1996) 4-14

MARCONI 2040

OSCILLOSCOPE

P.C. H.P. VECTRA 486/33U

NJ. ATMIOIBX NJ. CPIB488.2

Fig. 1. Block-diagram of the NMR measurement system.

computer was used to generate the polarization measure-

ment program, adjust the synthesizer through a GPIB bus,

control the polarizing magnet power supply, and measure and control the temperatures of the refrigerator. The software used for the programming and signal data treat- ment was LabView/PC (National Instruments).

The thermal equilibrium (TE) signal area measurements

began the process in order to calibrate enhanced polariza- tion signal areas. We started by a measurement of a base line, with the magnetic field far below the nominal

resonance value (Fig. 2a). The frequency was swept between 106.2 and 106.8 MHz. An average was calculated from 128 sweeps. The same procedure was performed around the nominal polarizing field and the average of the TE signals was obtained (Fig. 2b). After subtraction of the base line, the adjusted TE signal (Fig. 2c) was integrated. This procedure was done ten times, then an average was calculated on the ten integrated signals and a new succes- sion of ten measurements started. These series were stopped when averaged areas agreed within statistics. Usually four or five series were required to achieve good accuracy, namely AP,,lP,, = 1%.

In Fig. 3a is plotted the base line for enhanced polariza-

tion of the target out of the resonance magnetic field range. The enhanced signal at the resonance magnetic field value

is shown in Fig. 3b. The treated signal for pentanol doped with EHBA-Cr(V), corresponding to a polarization of +65% is shown in Fig. 3c. Finally, in Fig. 4 is shown the typical polarization buildup. The final value of the target polarization (-85%), in this particular case, was reached in 60 min.

A complete description of the polarization measure- ments and associated electronics is given in Ref. [ 181, and a dedicated paper is in preparation for publication. In this article the complete target polarization procedure as well as the estimate of errors arising from the associated electronics and the signal treatments will be described.

3. Pentanol-2 polarization

For proton targets, alcohols with a good ratio of free protons to bound protons have led to now obvious choice of butanol or pentanol. Considering the quoted ratio, these

J. Ball et al. I Nucl. Instr. and Meth. in Phy. Res. A .181 (1996) 4-14 7

(a) 25.0-1 I

20.0-

15.0-

mv lO.O-

5.0-

o.o-

-5.0-O~ 256 CHANNBLS

(a) 25.0

m+!;

-5.o- 0 50 100 150 200 ;

CHANNELS

(b) M.O-

lcl.O- mV

5.0- T.E. SIGNAL

o.o-

-5.0-m 0 50 im 150 200 256

CHANNELS

(c) 0.2 , 1

-0.5

mV -0.8

-1.0

-1.2

-1.5 T.E. SIGNAL-BASE

DIFFERENCE -1.6_1

0 25 50 75 100 125 150 175 MO 225 256 CHANNELS

Fig. 2. (a) TE baseline measured off of the nominal resonance

field. (b) TE polarization signal. (c) TE signal corresponding to the signal-base difference.

two materials are equivalent but the advantage of pentanol is that its melting temperature is 195 K, which is higher

than that of butanol (183 K). Moreover their glass transi- tion temperatures are 170 and 162 K, respectively [ 191. This remark is important for target operators, as the loading of the cartridge in the refrigerator at liquid nitrogen temperature often entails a short period of warm- ing up of the target. This period occurs during rinsing of the Helium circuit and injection of warm “He/JHe gas mixture. If the target reaches the glass transition tempera- ture, a significant reduction of the maximum reachable polarization and of the relaxation time results.

Recent works [20] have shown that pentanol-2 is strictly

(b)

mV ENHANCED

50 100 150 200 i CHANNELS

6

Cc) lOaO),

mV

-lOO.O-,

-mo-' SIGNAL-BASE -300.0., DIFFERENCE

.400.0.~

0 25 50 75 100125 150175 200 225 256

CHANNELS

Fig. 3. (a) Baseline measured off of the nominal resonance field.

(b) Enhanced polarization signal. (c) Enhanced signal corre-

sponding to the signal-base difference.

amorphous, independent of the cooling rate, which enables the target to be safer from deglassification processes. This led us to use this product on the Experiment 225 at SATURNE II. Pentanol-2 (CH,CH,CH,CHOHCH,) dif-

fers from pentanol-l (CH,CH,CH,CH,CH,OH) by iso- merit structure. The OH group is set on the second carbon of the chain instead of being on the one at the end of the molecule for pentanol- 1. This group is then surrounded by three carbons instead of one as for pentanol- 1. which gives higher stability to the molecule.

Polarization and relaxation time results are displayed in

Table 1. The conditions were 2.5 T polarizing field, 42 mK frozen spin temperature and 0.33 T holding field. The TE

8 J. Ball et al. I Nucl. Instr. and Meth. in Phys. Res. A 381 (1996) 4-14

oo-

m-

0 5 10152025JD35404550556065

Minutes

Fig. 4. Pentanol-2 polarization buildup.

signal was measured at 1 K. The sweep rate for the

measurements was 100 kHz for 256 steps with a sweep width of 600 kHz. The target while in holding mode was

constantly subjected to a proton beam of intensity less than

or equal to 10’ particles/burst. The errors mentioned in Table 1 take into account errors

arising from the data-acquisition electronics, the non-

linearity of the Q-meters, the TE temperature measure- ments, the magnetic field and the measurement dispersion

[la Due to physics requests, the target polarization had to be

reversed every day and time spent for building up was strictly confined to one hour. Afterward, one hour and a

half more was dedicated to going to frozen spin tempera-

ture, measuring the polarization and adjusting the holding mode at low field.

4. Polarization of 6Li materials

6LiH as well as ‘LiD are crystalline solids. To be able to polarize them one has to go through a well-defined routine,

Table 1 Pentanol-2 polarization results

Sign of the

polarization P(H) after

I h [%]

Nuclear

relaxation

time

Ways1

Positive 8452.5 25 Negative 8122.6 21

0.5 %iD P, July 1993

1. . is

:

‘0 I 2 3 4 5 6 7 TIME(hours)

Fig. 5. Deuteron polarization buildup of three ‘LiD samples

irradiated at the same temperature (18.5 K), but with different

electron doses (see Table 3).

the main steps of which are explained here. The complete

description can be found in Ref. [21]. The raw products are delivered as small irregular stones

that have to be crushed to grains of a defined size in nitrogen atmosphere to prevent any oxidation. The size of

each grain is about 2 mm. To create vacancies in the crystal lattice, the grains are irradiated at low temperature in an inert atmosphere by an electron beam. After irradia-

tion, the irradiated samples are kept in liquid nitrogen until they are transferred to the dilution refrigerator for polariza- tion.

Fig. 5 shows polarization growth of three ‘LiD samp;les

irradiated at the same temperature (185 K). but with different electron doses. The polarizing field was 2.5 T.

The measurements were processed through the same equipment as the one used for pentanol with circuits, cables and coils tuned to 16.35 MHz. The sweep rate was 10 kHz for 256 steps with a sweep width of 40 kHz. The

target itself was the double cartridge unit described in Section 2.1. The TE signal was measured at 1.4 K, and the

results are summarized in Table 2. The nuclear relaxation time was measured after 20 h under a 0.33 T holding field at 50 mK. The polarization accuracy is APIP = 3%.

It can be noticed that the polarization of the first sample rose more quickly during the first two hours than that of the second one. This allowed to reach the same value after 7 h. On the other hand, less dose is better for long relaxation times, so one has to use a sample tailored to the

physics experiment requirements. If target polarization has to be reversed often, sample 1 would be preferred. Sample 2 would be better for long running periods without any change in polarization direction.

Table 2

Deuteron polarization in “LiD for different irradiation doses

Sample

1 2

3

Electron P(D) P(D) Nuclear

dose after after relaxation

per cm’ 1 h [%] 7h[%] [days1

3 x IO” 30 42 29

1 x 10” 22 40 66

3 x loto I 30 70

J. Ball et al. I’ Nucl. Insir. and Meth. in Phvs. Res. A .38/ (1996) 4-14 9

5. Set-up for experiments with PPT and PDT

The polarized target was an essential part of the

nucleon-nucleon apparatus, which was designed for pp. np, and pn elastic and quasielastic scattering experiments in a large angular region [22]. It allowed the determination of several spin observables simultaneously. A top-view of the set-up is shown in Fig. 6. Outgoing particles were

detected by a two-arm spectrometer which consisted of single scintillation counters hodoscopes SH and WH, the

neutron counter (NC) hodoscope with its VETO. an analyzing magnet and eight multiwire proportional cham- bers (MWPCs) CO to C14. The magnet in the forward arm

analyzed the momenta of charged particles. The NC hodoscope [23] consisted of I5 scintillator bars. each of

which was viewed by two photomultipliers (PMs). The time-of-thght (TOF) of scattered particles between the target and the NC hodoscope was determined. The time difference of signals from the PM on each side of a NC bar

defined the position of the particle hit in the bar. The NC hodoscope efficiency for neutrons was -15-20% [22.23] over a large energy range. Veto counters in front of the NC

hodoscope distinguished between neutral and charged particles.

Fig. 6. A top-view of the apparatus, designed for nucleon-nu-

cleon elastic and quasielastic experiments with different polarized

targets. TDl. TD, TGl. TG2, PCEl and PCE2 are the single scintillation counters, SH and WH are the counter hodoscopes,

and NC is the neutron counter hodoscope with four VETO

counters. CO. Cl, C2, C3. CII, C12, Cl3 and Cl4 are MWPCs.

Two target holding coils are used: Coil n for the vertical

polarization and Coil k for the horizontal one. UT denotes the

unpolarized target. The forward arm is equipped with an analyzing magnet.

In order to correctly monitor measurements with two

opposite target polarization directions, an unpolarized target (UT) was positioned downstream from the PPT or

PDT. The scattering on either target produced independent triggers, but all tracks were detected in the same MWPCs. Unpolarized 6LiD and ‘LiH were used as the second target and were tested before their use as polarized targets.

For all types of events the charged particle track was reconstructed from the hits in the MWPCs. A forward-

scattered neutron trajectory was reconstructed from two points. The first one was the interaction vertex in the target

which is assumed to be the intercept of the track of the recoil proton measured in the conjugate arm with the

vertical plane containing the beam axis. The second point of the neutron trajectory was given by the neutron position

in the NC hodoscope. The outgoing proton track was determined from hits in the two MWPC’s positioned in front of the carbon block. This block was also used as a

charge-exchange convertor for backward neutrons. For each event one determined the laboratory angles H,,ih

and 4 of the scattered particle by a lit requiring the forward particle momentum to be consistent with elastic

nucleon-nucleon scattering. The forward particle momen- tum was determined from a TOF measurement for each

event and from a bend of the charged particles in the spectrometer magnet. From the measured laboratory angles

of the recoil particle, one then calculated the expected CM

angles 0,,,,, and 4_,, for the scattered particle assuming elastic scattering. The measured laboratory angles of the scattered track provided the CM scattering angles 6’,,,,,;,, and

4 m2,1\ under the same assumption. The histograms of the differences AIV,, = R.,,, - ri,,,.,. and A~J = &,, - G$,,,,, show “hydrogen peaks” [22]. For free np scattering at

0.88 GeV, using the pentanol-I target. the AH,., peak is shown in Fig. 7.

The kinematics cuts applied to the vertex position in the target, as well as to the A&,, and A+ distributions, selected elastic (or quasielastic) events on a “background

pedestal”. This background. previously determined from “dummy” target measurements. was to be subtracted. Fig.

Aec~(deg)

Fig. 7. Histogram of the difference A&, at 0.88 GeV for un-

selected np events. The pentanol-I PPT was used.

10 J. Ball et al. I Nucl. Instr. ad Meth. in Phys. Res. A 381 (1996) 4-14

Ay (de91

Fig. 8. The A4 distribution for free np scattering on the pentanol- 1 PPT at 0.88 GeV without any cuts (a) and the Ac$ distribution measured under the same conditions for which cuts in AS,, were applied (b).

8a shows the raw A4 distribution for free np scattering at

0.88GeV. In Fig. 8b is plotted the A+ distribution for

which the cuts on A&, were applied. The Ac$ distribution

represents a coplanarity of nucleon-nucleon scattering, which is independent of energy.

In Fig. 9 we compare the A&,,, distributions in np

scattering at 0.88 GeV for the pentanol-1 PPT, bLiD and

‘LiH targets. One can see that the shape of the distribution for 6LiD differed considerably from those of the other two

targets. In Fig. 10 is plotted the A4 distribution for pp scattering at 1 .I0 GeV on the ‘LiH target without any cuts.

This distribution (as well as that of A&.,, not shown here) is similar to the distribution measured with the pentanol-1 target in Fig. 8a.

The ‘LiD target always contained a few percent of “LiH or ‘LiH. The presence of free hydrogen is demonstrated in

Fig. 11, which displays the A&,, distribution for pp scattering at 2.08 GeV measured with the 6LiD target. At this high proton energy we observed a large A&, dis- tribution for quasielastic scattering of incident protons on protons in D and 6Li (fitted by a solid curve) and a small narrow peak from elastic pp scattering on free hydrogen. The quasielastic distribution is not centered with respect to the hydrogen peak from free pp scattering.

Events out of the “hydrogen” peak were due to inelastic reactions on all target nucleons and to Fermi motion contributing to quasi-elastic scattering on all bound nucleons. Except for the two-body inelastic reactions ppadn-’ and np (or pn)ddrr” which have quite differ- ent kinematics, all other inelastic scattering events were

25

I

7’ ; : I 1

: ANGULAR

I I CORRELATION I ’ I [. 0.88 GeV

1 ; I

J PPT

!, x ‘ILiH

I 1 I :

--- 6LiD

c 1; ,i ‘-1

! ‘; I I, “-P

A&-,,,(deg)

Fig. 9. Comparison of the A@,, distribution in np scattering at

0.88 GeV for the pentanol- 1 PPT (solid line), “LiD (dashed line)

and ‘LiH targets (dot-dashed line). The solid curve is provided by

a Monte Carlo calculation for the carbon background.

uniformly distributed around the free nucleon-nucleon conjugate angle. This was true at least within the solid angle of the experimental set-up. The quasielastic dis- tribution as a function of A&,, was asymmetric with

respect to the elastic one due to contributions from two factors. The first one is a difference of binding energies

between initial and final compound nuclei. The mean

separation energy necessary to remove one nucleon from the carbon target nuclei is about 30 MeV It changes only a little for different light nuclei. For deuterons the nucleon separation energy is equal to the binding energy (2.1 MeV). The second factor is the mean energy of Fermi motion, which further decreases the effective energy of

I I I I 1

10- 9- P-P

-020 -15 -10 -5 0 5 10 15

Ap (deg)

Fig. IO. Histogram of the A+ distribution measured in pp

scattering at I. IO GeV using the “LiH target.

.I. Bull et al. I Nucl. Instr. and Meth. in Phvs. Res. A 381 (1996) J-14 II

-25-20 -15 -10 -5 0 5 10 15 20 25

&M@eg)

Fig. 1 I. The Atic, distribution for pp scattering at 2.08 GeV,

measured with the “LiD target at 2.08 GeV. The solid line is a fit to

quasielastic scattering events. The small narrow peak is due to

elastic pp scattering on free hydrogen at the few percent level due

to ‘LiH and ‘LiH admixtures.

quasielastic scattering. The energy decrease is 5-6 MeV and plays a dominant role for scattering of nucleons on nucleons in deuterons. A quadratic sum of both contribu-

tions causes a well-observable shift of the maximum of the

quasielastic and background counting rates towards posi- tive values of At+.,. The value of A0 = 0 coincides with a

top of the “pure” hydrogen peak of the elastic scattering on free hydrogen, where both factors are absent. This effect can be also seen in Fig. 9. A decrease of the mean effective reaction energy provides a less pronounced

increase of conjugate angles when approaching the non- relativistic region.

In contrast to the A&, distribution, the A+ distribution for quasielastic scattering was always symmetric around

A4 = 0. This can be seen in Fig. 12 where the A4 distribution is plotted for quasielastic pp events measured

at 2.08 GeV with the “LiD target. One can compare this coplanarity symmetry with the elastic scattering and background distribution plotted in Fig. 8 or Fig. 10.

The TOF distribution of forward neutrons in np elastic scattering at 0.88 GeV on the pentanol-1 target is shown in Fig. 13. The main peak shows the correlation for np elastic events The small peak on the left-hand side was due to y

particles detected in the NC hodoscope. The TOF dis-

tribution for the pn quasielastic scattering at 2.08 GeV using the “LiD target is plotted in Fig. 14. It differs

$ ;fAi -20 -15 -10 -5 0 5 10 15 20

Ay(deg) TOF(ns)

Fig. 12. Histogram of the Ad distribution measured

scattering at 2.08 GeV using the ‘LiD target. in PP Fig. 14. The TOF distribution for pn quasielastic scattering at

2.08 GeV measured with the ‘LiD target.

“;UyJ

0 5 10 15 20 25 30 35 40 TOF(ns1

Fig. 13. The number of free np scattering events (0.88GeV)

plotted as a function of the TOF. The main peak shows the

correlation for np elastic events. The small peak on the left-hand

side is due to y-particles detected in the NC hodoscope.

considerably from the TOF distribution shown in Fig. 13. The cuts in the TOF distribution for elastic pp or np

scattering were roughly equivalent to cuts in angular

distributions. especially at low energies. The TOF cuts for quasielastic scattering improved the selection of events,

but were insufficient to distinguish between quasielastic scattering and inelastic reactions (see Section 6).

6. Use of Li compounds for quasielastic nn scattering

In this section we show some peculiarities of measure- ments with ‘Li targets. For demonstration purposes, we shall consider very preliminary quasielastic data obtained

by the NN group at SATURNE II. The pp and np analyzing power A ,,“,, o = AO,j,l,, was measured around

2 GeV using a SATURNE II polarized proton beam (P, - 0.75) scattered on the pentanol-1 PPT and on the second unpolarized “LiD target. As an example, Fig. 15 compares

results of free quasielastic pp analyzing powers in a medium angle range at 2.08 GeV. The results agree within

statistical errors. In Fig. 16 the angular distribution of the

quasielastic pn analyzing power using the same second target is shown. No free A ,,,,,,,,(np) data exist above

I. I5 GeY but measured absolute values were unexpectedly small. This observation is confirmed in Fig. 17 by a comparison of quasielastic results obtained ,with a polar-

i$y,M 0 5 10 15 20 25 30 35 40

12 J. Ball et al. I Nucl. Instr. and Meth. in Phys. Res. A 381 (IYY6) 4-14

0.2 , I I I I

I Aoono

-O.lU 50 60 70 80 90 100

eCM(deg)

Fig. 15. Results of free and quasielastic pp analyzing powers at 2.08 GeV. The pentanol-I (open circles) and a second unpolarized

bLiD (black dots) targets were used.

ized proton beam scattered on the polarized 6LiD target and the free beam and target analyzing powers

A oo,o(np) = A,,,, (np), measured at 1 .I0 GeV [24].

The discrepancy between the free np and quasielastic pn

results as well as the agreement of different pp data have an instrumental origin. As discussed in Section 5, forward

protons were analyzed by a spectrometer magnet and by a TOF measurement. For pn scattering with recoil neutrons

at forward angles, only a TOF selection could be applied. One possible explanation of this observation, which needs to be confirmed, is that in proton scattering on a “LiD

target the pp reaction channel pp 3 pnrr+ simulates the pn quasielastic scattering. The neutron is detected at small

angles by the NC hodoscope and one of the charged particles misses the detectors. The total cross section for

0.1

t ‘I ’ I I I

Aoono P-n

o 1 I 2.08 GeV

-- __ii;___li- __

t 4 1 t I -0.1t

1 %i D

-0.21 70 80 90 100 110 120 130

eCM (deg)

Fig. 16. Quasielastic np analyzing power at 2.08 GeV measured

with an unpolarized 6LiD target.

Fig. 17. Results of free np and quasielastic pn analyzing powers

at 1.10 GeV. The pentanol-l (open circles) and a second unpolar-

ized ‘LiD (black dots) targets were used.

this inelastic pp channel is of the same.order as the total elastic np cross section; it reaches 17 mb at 2 GeV.

This was neither the case for np elastic scattering with the beam of free neutrons and the PPT, nor for np

quasielastic scattering of neutrons in accelerated deuterons at large angles. In the first measurement no pp inelastic

reaction occurred and, in both measurements, a magnetic analysis of the outgoing protons provided an additional

constraint. The magnetic analysis of forward protons is quite

sufficient to suppress any contribution from inelastic reactions. For pn scattering, where neutrons are outgoing at large angles, a second NC hodoscope with a good de-

tection efficiency was needed. This NC hodoscope, built by the DPNC of the Geneva University, has been added to the existing set-up and tested.

Measurements were also carried out with the polarized ‘LiH target. The hydrogen and “Li target polarizations were about 30% and 6%, respectively, for the two opposite spin orientations In Fig. 18 are shown very preliminary results of the beam and target analyzing powers, A,,,Jpp) (black dots) and A ,,,,,(pp) (open circles) respectively, obtained with a highly-polarized proton beam at 1.10 GeV.

They are compared with the Aoono = A,,,, results (crosses) from Ref. [25], measured with the pentanol-1 PPT. Due to the small ‘LiH target polarization it is obvious that, for the

same statistics, the A,,),,z data points have considerably

larger errors. Angular distributions of different sets have

J. Bail et al. 1 Nucl. Instr. and Meth. in Phys. Rr.v. A .3X1 (lY96) 4-14 13

I I , I I , T-

OA-

0.3-

0.2-

0.1 -

I I A-_-_-A___

I I I I I 1 I

40 50 60 70 60 90 101 +M (deg)

Fig. 18. Results of beam and target pp analyzing powers at

I. IO GeV. The pentanol- 1 (crosses) and the polarized ‘LiH (black

dots and open circles) targets were used.

the same shape, but the A,,,,,,,, data have systematically smaller values. This is due to a contribution of pp

quasielastic scattering on protons in ‘Li to the hydrogen peak, which was not taken into account. Since the ‘Li polarization is -6%. a weighted average polarization of

free and quasifree proton target is then slightly lower than that of the hydrogen target.

7. Conclusions

Using aliphatic alcohols in the nucleon-nucleon ex- perimental program at SATURNE II enabled us to de- termine directly the scattering matrix at 11 energies in pp scattering and at 5 energies in np scattering. The com- parison of these results with the quasielastic data using Li

compounds allowed us to understand the difference be- tween the elastic and the quasielastic instrumental equip-

ment. The method performed also determined the contribu- tion of bound polarized deuterons in the ‘LiH target.

In order to understand quasielastic nucleon-nucleon scattering, it is worthwhile to measure the A0 and the AC#J distribution for pp scattering on unpolarized ‘Li. The Li-compound polarized targets were also used for a study of scattering processes on light nuclei. All three measure- ments may be completed by experiments using the scatter-

ing of accelerated Li nuclei on different targets. Note that a polarized ‘Li beam exists at SATURNE II.

Due to its deuteron content, any ‘LiD target. achieving a D-polarization of more than 32% would be preferable to deuterated aliphatic alcohols polarized to 5095. The “LiD target polarization may be considerably increased using higher magnetic fields, and the simple shape of the ‘LiD NMR deuteron spectrum allows an easy and accurate measurement of the polarization value. Nevertheless. it is difficult to foresee when the Li-compounds will be used in experiments. The target preparation needs an irradiation cryostat with a good temperature stability, a high-intensity electron beam in a convenient energy region. and accurate measurements of the electron dose. The relatively long time of the polarization buildup is still a major impediment to frequently repolarizing this type of target. Encouraging results obtained with Li-compounds in different laborator- ies need further study and optimization.

Acknowledgements

We thank J. Arvieux and G. Milleret for support of this

study and we acknowledge all members of the NN collaboration at SATURNE II for an efficient help. Discus- sions with E.I. Bunyatova concerning pentanol-2 solved several problems. We are also thankful to S. Mango and W. Meyer for stimulating meetings about lithium products.

References

III

121

PI

[41

[51

[61

171

PI

r91

J.H. Schulman and W.D. Compton, Colour Centers in Solids

(Pergamon, Oxford. 1962).

C. Kittel, Introduction to Solid State Physics (Wiley. New

York. 1971).

W. Burton Lewis and F.E. Pretzel. J. Phyh. Chem. Solids I9

(1961) 139.

K.H. Althoff. H. Dutz, S. Goertz. D. Kramer. W. Meyer. L.

Reichertz. G. Reicherz. W. Thiel. A. Thomas and J. Weber,

Proc. 9th Int. Symp. on High Energy Spin Physics. Vol. 2:

Workshops. Bonn, eds. W. Meyer. E. Steffens and W. Thiel

(Springer. Berlin. 1991) p. 301. H. Dutz. S. Goertz, D. Kramer. W. Meyer, G. Reichen. L.

Reichertz, W. Thiel. A. Thomas and J. Weber. ibid. p. 241.

D.G. Crabb, C.B. Higley, A.D. Krisch. R.S. Raymond, T

Roaer. J.A. Stewart and G.R. Court, Phyh. Rev. Lett. 64

(1990) 2627.

V. Bouffard, Y. Roinel, P. Roubeau and A. Abragam. J. Phys.

(Paris) 41 (1980) 1447.

B. van den Brandt. J.A. Konter. S. Mango and M. WeBIer.

Proc. 9th Int. Symp. on High Energy Spin Physics. Bonn.

Vol. 2: Workshops, eds. W. Meyer, E. Steffens and W. Thicl

(Springer. Berlin. 1991) p. 320. S. Ritt, E.T. Boschitz. R. Meier. R. Tacik. M. WeBIer, K.

Junker, J.A. Konter. S. Mango, D. Renker. B. van den

Brandt. V. Efimovykh, A. Kovaljov. A. Prokotiev. R. Mach.

P. Chaumette. J. Derkgel. G. Durand. J. Fabre and W. Thiel.

Phyh. Rev. C 33 ( 1991) 745.

14 J. Ball et al. I Nucl. Instr. and h4eth. in Phys. Res. A 381 (1996) 4-14

[lo] St. Goertz, Ch. Bradtke, H. Dutz, R. Gehring, W. Meyer, M.

Plilckthun, G. Reicherz, K. Runkel and A. Thomas, Proc. 7th

Int. Workshop on Polarized Target Materials and Techniques,

Bad Honnef, 1994, eds. H. Dutz and W. Meyer, Nucl. Instr.

and Meth. A 356 (1995) 20.

Ill] R. Bernard, P. Chaumette, P. Chesny, J. Deregel, R. Duthil, J.

Fabre, C. Lesmond, G. Seite, J. Ball, T. Niinikoski and M.

Rieubland, Nucl. Instr. and Meth. A 249 (1986) 176.

1121 M. Krumpolc and J. RoEek, J. Am. Chem. Sot. 101 (1979)

3206.

[13] V PetiiEek and M. Odehnal, Nucl. Instr. and Meth. 52

(1967) 197.

[14] G.R. Court, Proc. 2nd Int. Workshop on Polarized Target

Materials, Rutherford Laboratory, RL-80-080, eds. G.R.

Court. S.F.J. Cox, D.A. Cragg and T.O. Niinikoski, p. 76.

[I 51 T.O. Niinikoski, ibid., p. 80.

[ 161 T.O. Niinikoski, Proc. 7th Int. Workshop on Polarized Target

Materials and Techniques, Bad Honnef, 1994, eds. H. Dutz

and W. Meyer, Nucl. Instr. and Meth. A 356 (1995) 62.

[17] G.R. Court, D.W. Gifford, P. Harrison, W.G. Heyes and M.A.

Houlden, Nucl. Instr. and Meth. A 324 (1993) 433.

[18] J.L. Sans, Memoire d’Ingenieur CNAM, Consetvatoire Na-

tional des Arts et metiers, Paris (1996).

[19] S. Takala and T.O. Niinikoski, Proc. 9th Int. Symp. on High

Energy Spin Physics, Vol. 2: Workshops, Bonn, eds. W.

Meyer, E. Steffens and W. Thiel (Springer, Berlin, 1991) p.

341.

[20] E.I. Bunyatova, A. Sahling and S. Sahling, Solid State

Commun. 75 (1990) 125.

[21] G. Durand, C. Gaudron, J. Ball, M. Combet and J.L. Sans,

Proc. 10th Int. Symp. on High Energy Spin Physics, Nagoya,

eds. T. Hasegawa, N. Horikawa, A. Masaike and S. Sawada,

Frontier Science Series (Universal Academy Press, INC

Tokyo) ISSN 0915-8502, p. 355.

[22] J. Ball et al., Nucl. Instr. and Meth. A 327 (1993) 308.

[23] A. Ahmidouch, P. Bach, R. Hess, R.A. Kunne, C. Lech-

anoine-Leluc, C. Mascarini, D. Rapin, J. Arvieux, R. Bertini,

H. Catz, J.C. Faivre, F. Perrot-Kunne, F. Bradamante and A.

Martin, Nucl. Instr. and Meth. A 326 (1993) 538.

[24] J. Ball et al., Nucl. Phys. A 559 (1993) 489.

[2S] J. Bystricky, P. Chaumette, J. Deregel, J. Fabre, F. Lehar, A.

de Lesquen, F. Petit, L. van Rossum, J.M. Fontaine, F.

Petrot, J. Ball, A. Michalowicz, Y. One1 and A. Penzo, Nucl.

Phys. B 262 (1985) 715.