Recoil proton polarization of neutral pion photoproduction from protons in the energy range between 400 MeV and 1142 MeV

Recoil proton polarization of neutral pion photoproduction from protons in the energy range between 400 MeV and 1142 MeV

Nuclear Physics B168 (1980) 1-16 © North-Holland Publishing Company RECOIL PROTON POLARIZATION OF NEUTRAL PION P H O T O P R O D U C T I O N FROM PRO...

729KB Sizes 0 Downloads 16 Views

Nuclear Physics B168 (1980) 1-16 © North-Holland Publishing Company

RECOIL PROTON POLARIZATION OF NEUTRAL PION P H O T O P R O D U C T I O N FROM PROTONS IN THE E N E R G Y R A N G E BETWEEN 4 0 0 M e V A N D 1 1 4 2 M e V S. K A T O ,

T. MIYACHI,

K. T O S H I O K A

K. S U G A N O

1,

2 and K. UKAI

Institute for Nuclear Study, University of Tokyo, Tokyo, Japan M. CHIBA

Department of Physics, Tokyo Metropolitan University, Tokyo, Japan K. E G A W A

a n d T. I S H I I 3

Department of Physics, Kyoto University, Kyoto, Japan Y. YORIBAYASHI

4

Faculty of Technology, Tohoku Gakuin University, Tagajo, Japan K. J O H 5 a n d T. S H I N O H A R A

Faculty of Technology, Tokyo Agriculture and Technology University, Tokyo, Japan Y. W A D A

Department of Physics, Meifi College of Pharmacy, Tokyo, Japan Received 13 August 1979 (Revised 27 December 1979)

The recoil proton polarization of the reaction yp ~ lr°p was measured at a c.m. angle of 100 ° for incident photon energies between 451 and 1106 MeV, and at an angle of 130 ° for energies from 400 to 1142 MeV. One photon, decayed-from a ~r° meson, and a recoil proton were detected in coincidence. Two kinds of polarization analyzer were employed. In the range of proton kinetic energy less than 420 MeV and higher than 346 MeV, carbon plates and liquid hydrogen were used for determining the polarization, respectively. The data given by the two polarimeter systems are in good agreement. Results are compared with recent phenomenologlcal analyseg. From the comparison between the present data and the polarized target data, the invariant amplitude A3 can be estimated to be small.

i : 3 4 5

Now Now Now Now Now

at Fermilab, Batavia, III, USA. at Argonne National Laboratory, Argonne, I11, USA. at Institute for Nuclear Study, University of Tokyo, Tokyo, Japan. at Mizojiri Optical Co., Ltd., Tokyo, Japan. at Hitachi Engineering Co., Ltd., Hitachi, Japan.

1

2

$. Kato et al. / Recoil proton polarization

1. Introduction The recoil proton polarization of the process y p ~ zr°p was measured at a c.m. frame (c.m.s.) angle of 100 ° for incident photon energies between 451 and 1106 MeV, and at an c.m. angle of 130 ° for energies between 400 and 1142 MeV. In order to cover a large kinetic energy range of the recoil proton, two kinds of polarization analyzer were employed. In the range of proton kinetic energy less than 420 MeV and higher than 346 MeV, carbon plates and liquid hydrogen were used for determining the proton polarization, respectively. In recent years, considerable progress has been made in the study of single pion photoproduction in the resonance region. A number of experimental data have been accumulated. Many theoretical analyses [1-6] have been made. Photon couplings of dominant resonances, such as P33(1232), D13(1520) and F15(1688) were well determined, but there are some ambiguities for less prominent resonances, such as PH(1470), SH(1535), D15(1670), etc. However, Berends and Donnachie [7] pointed out that the helicity-~ coupling of the D13(1520) still had an ambiguity. For further progress, accurate and systematic data are needed; especially various kinds of polarization parameter. The main aim of this is to obtain accurate and systematic data on the recoil proton polarization for the reaction 3'P ~ ~r°P over a wide energy range, to reduce the ambiguity of the theoretical analyses. The second aim is to investigate the invariant amplitude A3 given by C h e w - G o l d b e r g e r - L o w - N a m b u (CGLN) [8]. Feller et al. [9] compared the recoil proton polarization P(O) around 90 ° with the target asymmetry T(O) around 100 ° and found a similar behaviour of the data. In that comparison, the T(O) data around 100 ° were very precise but the P(O) data were not. We compare two parameters with the same precision and angle. Moreover, we can compare the two parameters at another angle of 130 ° .

2. Experimental arrangement The general layout of the experiment is shown in figs. 1. Decay photons from a 7r° meson were detected by a gamma-ray counter. Recoil protons were measured for their angles and momenta by a magnetic spectrometer. After the magnetic spectrometer, either the carbon or liquid hydrogen polarimeter was installed. A precise report of these polarimeter systems is published elsewhere [10]. A master signal for the event was generated by a coincidence signal between the gamma-ray counter and the magnetic spectrometer.

2.1. Beam, beam monitor and target The bremsstrahlung beam was produced by accelerated electrons striking a 50 ixm thick platinum radiator in the electron synchrotron at the Institute for Nuclear Study, University of Tokyo (INS). The beam profile was defined by a lead collimator which

$. Kato et al, / Recoil proton polarization

3

tARGET

SC 1-4

WSC5-15

BEAM MONITOR

(a)

T TRIGGERCOUNTER ¢,r' ~IDFCOUNT£R NTER

3TON 4ARDNER

(b) Fig. 1. Experimental layout of (a) the carbon and (b) the hydrogen polarimeter systems.

4

$. Kato et al. / Recoil proton polarization

was set down-stream at a distance of 6 m from the internal radiator. The beam shape was a rectangle of 20 mm width and 30 mm height at the point of the hydrogen target which was 18 m away from the platinum radiator. A sweeping magnet behind the collimator rejected charged particles associated with the beam. Also a magnet in a magnetic spectrometer pair, which was used for calibrating the gamma-ray counter, swept out the electrons and positrons in the beam during the experiment. The beam path was evacuated to ~ 10 -2 torr in order to reduce the background due to the electromagnetic process. The beam was monitored in two w~ys. The main monitor was a thick-walled ionization chamber which was calibrated with a Faraday cup using an extracted electron beam. For the purpose of a cross check, a thin-walled ionization chamber was placed between the two magnets. The average intensity and duty factor of the photon beam were respectively 1.0 × 1 0 9 eq. quanta/sec and 8% during the experiment. The liquid hydrogen container was made of Mylar sheets of 127 ~m thickness and it was cylindrical with a diameter of 5.0 cm and a length of 13.3 era. The liquid hydrogen was supplied by a closed gas system using a small refrigerator.

2.2. Gamma-ray counter The gamma-ray counter consisted of two independent systems, 3"1 and 3"2, as shown in fig. 2. The 3' 1 system consisted of a lithium hydride (LiH) photon hardner of 10 cm thickness, two veto counters (V), a lead plate converter of 1.04 radiation lengths, scintillation counter hodoscope arrays (HX, HY) and a lead glass (~erenkov counter (C). The 3'1 system was the same as was used earlier to measure proton Compton scattering [11]. The 3"2 system enlarged the geometrical acceptance for photons decayed from a 7r° meson. It consisted of a polyethylene photon hardner of 10 cm thickness, four anticounters (A) and twenty-two Cerenkov counters (C'). The Cerenkov materials were blocks of flint glass SF-5 of 6.5 x 6.5 x 26 cm 3. The 3,1 and 3,2 systems were placed at a distance of 130 cm and 45 cm from the primary target, respectively. The photons decaying from ~r° mesons were caught with almost 100% geometrical efficiency by the 3"1 and the 3,2 systems in coincidence with the recoil proton. However a few percent of the photons was lost by the gap between the 3'1 and the 3"2 systems. About 10% of decay photons were converted into a pair of electrons by the materials between the target and the veto- or the anticounters. 2.3. Proton spectrometer As shown in figs. 1, the proton spectrometer consisted of eight multiwire proportional chambers (MWPC1-8), four wire spark chambers (WSC1--4), a bending magnet and scintillation counters (T1-T4). The bending magnet was a sector-type magnet with a gap distance of 10 cm and the field strength was 20 k G at maximum.

$. Kato et al. / Recoil proton polarization

5

Fig. 2. Arrangement of the gamma-ray counter.

The M W P C ' s and the W S C ' s were used to determine an entry and an exit trajectory of a particle, respectively. The sense wire plane in the M W P C ' s consisted of gold-plated tungsten wires with a diameter of 20 p,m and a spacing of 2 ram. Five of the eight M W P C ' s were used to determine the horizontal direction of the particle and the others were used to obtain the vertical direction. The space resolution of the M W P C ' s was + 0.9 m m . The effective area of a WS'C was 50 x 40 cm 2. An electrode was m a d e of C u - B e wire with a diameter of 100 p,m and a spacing of 1 mm. The cathode and anode wires were stretched perpendicularly. A magnetostrictive delay line was used to read out the sparking point. Signals from the cathode and anode wires detected the twodimensional position of the sparking point. The position resolution of the W S C ' s was + 0.6 m m both vertically and horizontally. The m o m e n t u m resolution was + 0.6% for a proton m o m e n t u m of 1300 M e V / c and + 1.2% for 600 M e V / c . The angular resolution varied from + 0.2 ° to + 0.4 ° between 1300 and 600 M e V / c . The photon energy was obtained with an accuracy of +1.5%. Particle identification was p e r f o r m e d by measuring an energy loss in the T4 counter ( d E / d x ) and a flight-time difference between the T1 and the T4 counters (TOF). Electrons were mostly rejected by a coincidence between the g a m m a - r a y counter and the proton spectrometer.

6

S. Kato et aL / Recoil proton polarization

2.4. Polarimeter using carbon as scatterer

As shown in fig. la, the polarimeter system using carbon scatterer consisted of eleven WSC's (WSC5-15), a range counter with six layers of scintillation counters (R1-6) and 3 to 8 layers of carbon plates inserted between the WSC's depending on the proton energy. The carbon was subdivided into several layers so that an angle broadening due to multiple Coulomb scattering in each layer was less than 1.4 °. The total thickness of the carbon scatterer was chosen so that the kinetic energy of the proton after passing through the carbon was greater than 95 MeV, because for energy less than 95 MeV, no reliable data on analyzing power existed. The protons were scattered by the carbon plates, then the incident and the scattered trajectories were reconstructed using the information from WSC5-15. The scattering angle and the interaction point were calculated using these trajectories. The angular resolution/tO was + 1.5 ° (polar angle: O) and A~b was + 4.5 ° (azimuthal angle: &). The distribution of the scattering angle 8 is given in fig. 3. At small angles, multiple Coulomb scattering is dominant. In order to separate the nuclear interaction from the multiple Coulomb scattering, the cut of the angle, 0cut, was set at 6 °. Fig. 4 shows the distribution of the interaction point in the carbon scatterer. The scattering point was reconstructed with an accuracy of :i: 1.5 cm along the incident direction. The energy at the scattering point was measured by a range-counter telescope. The energy resolution of the range counter was about + 10 MeV. The measurement of energy loss in the carbon plates serves to discriminate against inelastic scattering

103

10 2

I0

multiple Coulomb scottering

1 /

Ocur=6*

s'o

,;o

,;o

o'

Fig. 3. Distribution of the polar angle O in proton-carbon scattering. The log of the n u m b e r of events is plotted against 0 2.

S. Kato et al. / Recoil proton polarization PC SCATTERING POINT RECONSTRUCTION

100

( e > 6 ")

1

CARBON PLATE

50

1400

1500

1600

1700 1800 1 9 0 0 2400 POSITION Z(mm)

Fig. 4. Distribution of the scattering points in proton--carbon scattering.

events with excitation energy above 30 MeV. The analyzing power of proton-carbon scattering was calculated using the experimedtal data for inelasticities up to 30 MeV, including elastic scattering [12].

2.5. Polarimeter using liquid hydrogen as scatterer As shown in fig. 1b, the polarimeter system using the hydrogen scatterer consisted of a liquid hydrogen scatterer of 14.5 1 with a cryogenerator, eleven layers of wire spark chambers (WSC5-15) and eight scintillation counters (SC1-4, BC1-4) which surrounded the sides and the back of the hydrogen scatterer. In order to distinguish an elastic event from an inelastic one, the forward and backward-scattered protons were detected by the WSC's and the counters. Then, three WSC's were set at both sides of the scatterer and five WSC's were set at the back. As shown in fig. 5, the liquid hydrogen scatterer consisted of a container for liquid hydrogen, a vacuum container surrounding the liquid hydrogen container and a Philips 105 cryogenic transfer system. The liquid hydrogen container was a cylindrical vessel of 15 cm diameter, 82 cm length, and 62 cm effective length. The liquid hydrogen container was made of an aluminium alloy (A1 5052) of 0.05 cm in thickness. The vacuum container was made of aluminium alloy of 0.2 cm thickness. Elastic p r o t o n - p r o t o n scattering in the liquid hydrogen scatterer was identified by measuring the angle correlation and the coplanarity of the two protons. Differences of the polar angle between observed and calculated values A0 are plotted in fig. 6a. Fig. 6b shows a typical distribution of the differences of the azimuthal angle A~b. These figures exhibit elastic peaks with a resolution of ± 1.5 ° for the polar angle and + 2.2 ° for the azimuthal angle. Elastic events were selected by applying cuts of + 5 ° on the A0 distribution and + 7 ° on A4~, taking into account of the broadening of the distribution due to multiple Coulomb scattering and the resolution of the WSC's.

S. Kato et al. / Recoil proton polarization

--Emergency

Liq. H2 Level Detector 2ram (AI)

!

05ram (AI)

f a~O.5 mm(AI) ~ffective

3ram (A t )

Vacuum Container Liq. H2 Container 50cm

i

7mm(AI) t

Fig 5 Schematic view of the liquid hydrogen and vacuum containers of the hydrogen polarimeter

500

(o3

400 300

300 Ae

200

= ~ S"

200 100

100 -2'0

-10

0 .10 • 20 &O (Degree)

-10

0 .10 *20 [email protected] (Degree)

Fig. 6. Distribution of the differences in (a) the polar angle 0 and (b) the azimuthal angle d~ between the observed and the calculated ones in the hydrogen polarimeter. The vertical lines represents the cutting points.

Figs. 7a, b show distributions for the scattering point in the scatterer for the radial and incident directions, respectively. In figs. 7, shaded areas indicate elastic scattering events selected by the above cuts on AO and A~b. As shown in figs. 7, scattering events in the aluminium walls of the hydrogen and vacuum containers are clearly seen.

2.6. Electronics and data acquisition system A proton signal was made by the coincidence between the counters T1 and T4. A gamma signal was made by the (V. C + ,~. C'). The master signal was generated by a

S. Kato et al. / Recoil proton polarization '

~/ t.00

300

200

Voc CONTAINER

LIc N";,INER

f

i

m >Z

EVENTS

100

LU EFFECTIVE AREA

(b)

Ca)

50

100

~/////////////////////////////, 10

20

30

R (cm)

50

I00 Z (¢m)

~_

150

Fig. 7. Scattering point distributions along (a) the radial direction and (b) the incident direction of the hydrogen scatterer derived by the WSC's in the hydrogen polarimeter. Shaded areas indicate elastic scattering events selected by the cuts of + 15 ° on AO and ± 7* on A4~. Scattering events in the alurninium walls of the containers are also excluded.

coincidence between the proton and gamma signals. Information from many devices in the gamma-ray counter and the proton spectrometer was all set up in CAMAC modules. Those data were transferred to a mini-computer TOSBAC-40C (T-40) through the A-type crate controllers and the branch driver. The T-40 was connected on-line to the INS central computer TOSBAC 3400/M51 (T-3400). The following data were transferred to the T-3400 using the INS on-line system [13] and handled in real time: (i) accumulation of various histograms of dE/dx, TOF, hodoscopes, chambers, etc., (ii) efficiencies of chambers, (iii) reconstruction of particle trajectories, (iv) calculation of the momentum of the recoil proton and the incident photon energy, (v) estimation of the scattering angle and the interaction point, (vi) calculation of the polarization value by a simple method. The analyzed information was sent back to the T-40 during the experiment. Firing positions of the MWPC's and the WSC's were displayed event by event on a CRT display in the T-40.

2. 7. Calibration o[ the polarimeters In order to check an artificial asymmetry in the polarimeter system, negative pions were injected to the polarimeters instead of the protons. The data were analyzed by the criteria mentioned in section 3. The left-right asymmetries ( N L - N R ) / ( N L +

10

s. Kato et al. / Recoil proton polarization

N R ) were - 0 . 0 1 8 ± 0 . 0 2 9 for the carbon polarimeter and 0 . 0 6 + 0 . 0 7 for the hydrogen polarimeter, respectively. N L and N R are the numbers of the pions scattered to the left and right by the scatterer, respectively. These values were consistent with zero.

3. Data analysis and results 3.1. D a t a analysis for the carbon polarimeter

About 1% of the total events survived under various methods of determining the polarization. The processes of data reduction are as follows. (i) Only those events that reached the range counter were analyzed. (ii) Reconstructed trajectories should originate within the primary target, and should pass the scintillation counters (T1-T4) and the aperture of the magnet. (iii) If the scattered point in proton--carbon scattering was far from the carbon plates (I> 5 cm), the event was rejected. (iv) The scattered trajectory in proton-carbon scattering was rotated by 180 ° around the axis of the incident proton trajectory at the scattered point. If this rotated trajectory did not pass the range counter, the event was abandoned. Hereafter, this is called the mirror reflection condition. (v) The energy loss in proton-carbon scattering AE is shown in fig. 8. Events with energy loss 30 I> AE I> - 2 0 MeV were adopted. (vi) The distribution of the polar angle in proton-carbon scattering is shown in fig. 3. A cut-off angle of 6 ° was imposed in order to remove the effect of multiple

ENERGY LOSS

DISTRIBUTION

IN

SCATTERING

P-C

( SET 600 MeV - 100")

1000

AE = 30 MeV CUT

CUT

. \. 1

500

I

-2o

-,o

o

I

io

20 30 6E (MeV)

,0

so

Fig. 8. Distributionof the excitationenergyof carbon AE, when the maximumenergy of the photon beam was 600 MeV and emission angle of the pion 0* was 100°.

$. Kato et al. / Recoil proton polarization

11

P POLARIZATION k = 8 0 0 t 25 M e V IBCM = 130"

¢ ¢

-0.5 i

,

,

,

i

i

5 6 7 8 9 I0"

,



OuT

Fig. 9. Cut-off angle dependence of the polarization in the carbon polarimeter.

Coulomb scattering. Fig. 9 shows the cut-off angle dependence of the polarization. As the cut-off angle increases, the contamination from multiple Coulomb scattering decreases and the value of the polarization saturates for angles above 6 °.

3.2. Data analysis for the hydrogen polarimeter About 0.5% of the events that passed the secondary hydrogen target within a diameter of 15 cm, survived various conditions and served for the determination of the polarization. The methods of data reduction were the following. (i) Only those events, whose scattered protons fired both the back and the side counters were used. (ii) Reconstructed trajectories should originate within the primary target, and should pass the scintillation counters (T1-T4) and the aperture of the secondary hydrogen target. (iii) Events scattered with a polar angle of 36 ° ~> 0 1>8 ° and with an azimuthal angle of Icos q~l ~>0.5 were selected. (iv) Mirror reflection was applied to the forward- and backward-scattered trajectories. If the reflected trajectories did not pass through the aperture of the back or the side counters, the event was abandoned. (v) A backward-scattered proton with a kinetic energy below 55 MeV cannot reach the side counters. The energy of the backward-scattered proton was calculated from the incident and forward-scattered protons under an assumption that the reaction was caused by elastic scattering. Events with kinetic energy less than 55 MeV were abandoned. (vi) Polar and azimuthal angles of the backward-scattered proton were calculated from the incident and the forward-scattered trajectories. The differences between the observed and the calculated angles are shown in figs. 6. In order to obtain elastic events, the differences of the polar and the azimuthal angles were selected to be within + 5 ° and + 7 °, respectively.

12

S. Kato et al. / Recoil proton polarization

3.3. Determination of the polarization The polarization is defined, according to the Basel convention, as positive when the protons have their spins in the direction

n =kxq/Ikxql, where k and q are the incident photon and the emitted pion momentum vectors, respectively. The differential cross section of the proton-carbon and the protonproton scattering is given by

dtr (o,~b,T, AE, p ) =~--~ do" (0, T, AE, P = 0){1 +A(O, T, AE)P cos $ } , d---~ where A(O, T, AE) is the analyzing power and P is the polarization of the incident proton. The parameters O, ~b, T and d E are the polar angle, azimuthal angle, incident kinetic energy of the proton, and excitation energy of the carbon nuclei, respectively. In the case of proton-proton scattering, d E is zero. The analyzing powers of the proton-carbon and proton-proton scattering used were compiled by Peterson [12] and Bluem et al. [14], respectively. The polarization is determined by maximizing the likelihood function

L(P) =l-I (1 + A(Oi, Ti, AEi)P cos ~bl), i

where the subscript i runs over all events. The value (l+A(Oi, Ti, zlEi)Pcos 4~i represents the probability that event i occurs with polarization P, kinetic energy Ti, excitation energy of the carbon nuclei AE~,polar angle Oi and azimuthal angle ~bi.The obtained polarization was corrected by the reconstruction efficiency of the WSC's which was given by the use of the scintillation counter information. The statistical error of the polarization, zIP, was evaluated as the deviation at which the value of the likelihood function decreases to Lmax e -1/2.

3.4. Experimental results The proton polarization was measured for various settings of the spectrometer and different maximum photon energies. In order to reduce the biases due to the different settings and contamination from the process yp ~ ~r°~r°p, the consistency between the data from the different settings was examined. Good agreement is obtained among the data in the overlapping regions. The final results are shown in fig. 10 and listed in table la for the carbon polarimeter and in table lb for the hydrogen polarimeter. The errors are statistical only.

$. Kato et aL / Recoil proton polarization

13

TABLE la Results for the carbon polarimeter Energy bin (MeV)

Average photon energy (MeV)

Average c.m. angle (degrees)

Polarization

Number of events

450±25 500±25 550±25 600±25 650±25 700±25 750±25 800±25 850±25 900±25 950±25 1000±25 400±25 450±25 500±25 550±25 600±25 650±25 700±25 750±25 800±25 850±25

451 500 549 600 650 699 750 799 849 898 950 999 400 449 497 550 599 654 700 758 803 850

105±3 104±5 103±5 103±5 102±4 103±5 102±3 102±5 102±3 101 ± 3 103±3 100±4 133±4 131±6 129±5 132±4 130±5 133±4 129±5 132±3 130±5 131±4

+0.12±0.19 -0.21±0.10 -0.47+0.11 -0.64 + 0.08 -0.86 ± 0.08 -0.73 ± 0.07 -0.43±0.10 -0.34 - 0.09 -0.24±0.10 -0.22±0.14 -0.05±0.11 -0.27±0.11 +0.11+0.10 -0.04 ± 0.07 -0.16±0.10 -0.37±0.10 -0.61 ± 0.09 -0.88±0.08 -0.77 ± 0.07 -0.56±0.09 0.40 ± 0.07 -0.28±0.10

1027 1775 838 1181 739 972 544 554 522 254 477 372 2284 2430 697 675 550 558 1134 807 931 593

-

TABLE lb Results for the hydrogen polarimeter Energy bin (MeV)

Average photon energy (MeV)

Average c.m. angle (degrees)

Polarization

Number of events

925 ± 50 1025+50 1125±50 750±25 800 ± 25 850~:25 900±25 950±25 1000±25 1050 + 25 1100± 25 1150±25

940 1023 1106 751 800 849 900 950 1001 1048 1099 1142

106 ± 2 103±2 101±2 134±2 132 ± 2 131 ±3 130±2 130±4 132±3 131 + 3 131 ± 2 130±2

+0.07 ± 0.28 -0.18±0.20 -0.09 + 0.34 -0.30±0.32 -0.26±0.24 -0.09 ± 0.13 -0.15±0.13 -0.09±0.10 -0.35±0.10 -0.43±0.10 -0.44±0.15 -0.42±0.19

108 176 59 92 138 419 338 578 492 418 185 104

14

S. Kato et al. / ReCoilproton polarization 1.0 9c. P(0)

~

(Q)

= 100'

L I O Hz CARBON

PRESENT DATA

OcM P(0)

T (O)

~

130"

(b)

PRESENT DATA

M F u k u s h i m a e t ol at BCM=I2O,

P F e l l e r et ¢11

0.5

=

L I Q Hz CARBON

T(19)

M F u k u s h i r n a et at

~

PSL

Booth et al

!

t -0.5

- I. 0

,t tj

+

5 0'0

'

10 'oo

I 500

tt* i

d 1 O0

i

Fig. 10. Comparison of the recoil polarization P(O) with the target asymmetry T(O) for the process of yp-~ 7r°p at a c.m. angle of (a) 100° and (b) 130°. 3.5. Uncertainties o f the polarization

There are several sources which may cause uncertainties of the polarization. (i) Geometrical biases and the inhomogeneity of the reconstruction efficiency of the WSC's were checked by the calibration run. The error due to these effects was estimated to be less than + 0.03 for the carbon polarimeter and less than + 0.07 for the hydrogen polarimeter. (ii) Contamination from other processes such as y p ~ yp, y p ~ ~r°~r°p and "yp r/p was calculated from the ratio of the cross sections to the reaction of yp-~ ~r°p, taking acceptances of the polarimeters and the g a m m a - r a y counter into account. The uncertainties were evaluated to be less than + 0,05 for both the carbon and the hydrogen polarimeters. (iii) The effect caused by materials other than hydrogen in the primary target was less than + 0.01. This effect was measured by the e m p t y target run. (iv) The uncertainty due to materials other than hydrogen in the hydrogen polarimeter was less than + 0.03. This was also checked by the e m p t y secondary target run. (v) The scattering of protons by the materials of the WSC's could not be separated from the p r o t o n - c a r b o n scattering. This was estimated by the density ratio of the carbon plates and other materials. The error was evaluated to be less than + 0.01. (vi) An ambiguity in the analyzing power occurred, the error being less than + 0.05 for the carbon and the hydrogen polarimeter runs. (vii) The disturbance due to the precession of the proton spin in the magnetic field was less than + 0.001. The normal direction was almost parallel to the magnetic field ( + 2 ° at maximum).

S. Kato et al.

/ Recoil proton polarization

15

(viii) Errors caused by the depolarization of the proton passing through the matter were less than +0.01 in the carbon polarimeter and +0.0001 in the hydrogen polarimeter and were given by Wolfenstein [15]. Total errors of the polarization other than statistical ones were evaluated within + 0.08 for the carbon polarimeter and + 0.10 for the hydrogen polarimeter. 4. Discussion The results are shown in fig. 11. In fig. 11, the theoretical analyses of Metcalf and Walker (MW) [2], Moorhouse, Oberlack and Rosenfeld (MOR) [3] and by Feller et al. (NAGOYA) [5] are also shown for comparison. The present data are in good agreement with the other existing data. At 130 °, data given by Bluem et al. (Bonn group) [4, 16] are inconsistent with the present data. The energy bins of the Bonn data were 100 MeV. The Bonn group detected only protons and the correction due to the contamination from ~/p-+ ~r+zr-p probably amounted to 0.14 at most, under the assumption that protons from this process had zero polarization. In the case of proton detection only, corrections due to the Mylar window of the primary target were large and amounted to 0.10 at most. The present data show a large bump at 650 MeV, a dip at 850 MeV and a bump at 1100 MeV. Phenomenological analyses are in agreement with our data. But looking precisely, the analyses could not reproduce well our data at 130 °. A new analysis is necessary to reproduce our data from the low to the high energy region. Previously Feller et al. [9] compared the target asymmetry T(O) data around 100 ° with the P(O) data around 90 ° in the low energy region for 3'P ~ ~r°P. As shown in fig. 1.0

Pie)

Pie) ~ T~'p. 0CM= 100" T

lip

0.5

(a)

~) ~I ~+ "1" +

--

LIQ Hz

Present OGta

CARBON Prentice et al

I[p ~ n ' p ~

~

0c.:130"

~

CARBON

Present Dora

muemetm

t

[:)erebchil~sllij et 01.

Lundqui,, ,t o l Mw . . . . NAGOYA

(b)

+

O e f e b c h i n l k i j et a l Lundquilt It ill

- - - M.O.R

"'-.

LIO HI

--

M W

---

M OR

""'" NAGOYA

/

0.5

-I,0

1

5OO INCIDENT

i

1000 PHOTON ENERGY ( M e V )

500 INCIDENT PHOTON

1000 ENERGY

(MIV)

Fig. 11. Results of the polarization as a function of the incident photon energies at a c.m. angle of (a) 100 ° and (b) 130 °. The results of the analyses given by MW, MOR and N A G O Y A are also shown for a comparison.

16

S. Kato et al. / Recoil proton polarization

10a, available T(O) data at 100 ° [9, 17] coincide with the present P(O) data fairly well. At 130 °, as shown in fig. 10b, T(O)data [18] are of higher value than the present data, while they are similar in shape. The T(O) data at 120 ° [17] are also compared to our P(O) data at 130 ° at low energy. The agreement of the two parameters is good. If the invariant amplitude A3 [8] is zero, the helicity amplitude of H1 is equal to/-/4, and T(O) becomes equal to the P(O) [19], but it must be noted that P(O) = T(O) does not necessarily lead to A3 = 0. The good agreement of the T(O) and the P(O) is consistent with a small A 3 amplitude in the range of 1.5 > E v > 0.5 G e V and It[ < 0.7 ( G e V / c ) 2 using the results of the partial-wave analyses [2, 3] derived by Barker, Donnachie and Storrow [20], and Odorico [21]. Further investigations are needed to know whether the agreement of T(O) and P(O) is only accidental or related to a finite energy sum rule for the A 3 amplitude [20, 21]. The authors express their thanks to the INS electron synchrotron crew for the machine operation and the INS high energy group for their support during the experiment. Thanks are also due to the members of the INS machine shop for their mechanical work on our detection system. The data taking, the analysis of the data and the computational calculations have been performed using the computer at the INS computer center. This work was partly supported by a Grant-in-Aid from the Japanese Ministry of Education, Science and Culture. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

R.L. Walker, Phys. Rev. 182 (1969) 1729. W.J. Metcalf and R.L. Walker, Nucl. Phys. B76 (1974) 253. R.G. Moorhouse, H. Oberlack and A.H. Rosenfeld, Phys. Rev. D9 (1974) 1. R.C.E. Devenish, D.H. Lyth and W.A. Rankin, Phys. Lett. 52B (1974) 227. P. Feller et al., Nucl. Phys. B104 (1976) 219. P. Noelle, Prog. Theor. Phys. 60 (1978) 778. F.A. Berends and A. Donnachie, Nucl. Phys. B136 (1978) 317. G.F. Chew, M.L. Goldberger, F.E. Low and Y. Nambu, Phys. Rev. 106 (1957) 1345. P. Feller et al., Nucl. Phys. B l l 0 (1976) 397. M. Chiba et al., Jap. J. Appl. Phys. 18 (1979) 1817. T. Toshioka et al., Nucl. Phys. B141 (1978) 364. V.Z. Peterson, UCRL°10622 (1963). K. Ukai, INS-TEC-2 (.1972); INS-TEC-3 (1973). P. Bluem et al., Z. Phys. A277 (1976) 311. L. Wolfenstein, Phys. Rev. 75 (1949) 1664. P. Bluem et al., Z. Phys. A278 (1976) 275. M. Fukushima et al., Nucl. Phys. B136 (1978) 189. P.S.L. Booth et al., Nucl. Phys. B121 (1977) 45. E.N. Argyres, A.P. Contogouris, J.P. Holden and M. Svec, Phys. Rev. D8 (1973) 2068. I.S. Barker, A. Donnachie and J.K. Storrow, Nucl. Phys. B79 (1974) 431. R. Odorico, Nucl. Phys. B101 (1975) 480.