EPJ manuscript No.
(will be inserted by the editor)
arXiv:0807.3839v1 [nucl-ex] 24 Jul 2008
Measurement of beam-recoil observables Ox , Oz and target
asymmetry for the reaction γp → K +Λ
A. Lleres1 , O. Bartalini2,10 , V. Bellini13,6 , J.P. Bocquet1 , P. Calvat1 , M. Capogni2,10,4 , L. Casano10 , M. Castoldi8 ,
A. D’Angelo2,10 , J.-P. Didelez16 , R. Di Salvo10 , A. Fantini2,10 , D. Franco2,10 , C. Gaulard5,14 , G. Gervino3,11 ,
F. Ghio9,12 , B. Girolami9,12 , A. Giusa13,7 , M. Guidal16 , E. Hourany16, R. Kunne16 , V. Kuznetsov15,18 , A. Lapik15 ,
P. Levi Sandri5 , F. Mammoliti13,7 , G. Mandaglio7,17 , D. Moricciani10 , A.N. Mushkarenkov15, V. Nedorezov15,
L. Nicoletti2,10,1 , C. Perrin1, C. Randieri13,6 , D. Rebreyend1 , F. Renard1 , N. Rudnev15 , T. Russew1 , G. Russo13,7 ,
C. Schaerf2,10 , M.-L. Sperduto13,7 , M.-C. Sutera7 , A. Turinge15 (The GRAAL collaboration)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
LPSC, Université Joseph Fourier Grenoble 1, CNRS/IN2P3, Institut National Polytechnique de Grenoble, 53 avenue des
Martyrs, 38026 Grenoble, France
Dipartimento di Fisica, Università di Roma ”Tor Vergata”, via della Ricerca Scientifica 1, I-00133 Roma, Italy
Dipartimento di Fisica Sperimentale, Università di Torino, via P. Giuria, I-00125 Torino, Italy
Present affiliation: ENEA - C.R. Casaccia, via Anguillarese 301, I-00060 Roma, Italy
INFN - Laboratori Nazionali di Frascati, via E. Fermi 40, I-00044 Frascati, Italy
INFN - Laboratori Nazionali del Sud, via Santa Sofia 44, I-95123 Catania, Italy
INFN - Sezione di Catania, via Santa Sofia 64, I-95123 Catania, Italy
INFN - Sezione di Genova, via Dodecanneso 33, I-16146 Genova, Italy
INFN - Sezione di Roma, piazzale Aldo Moro 2, I-00185 Roma, Italy
INFN - Sezione di Roma Tor Vergata, via della Ricerca Scientifica 1, I-00133 Roma, Italy
INFN - Sezione di Torino, I-10125 Torino, Italy
Istituto Superiore di Sanità, viale Regina Elena 299, I-00161 Roma, Italy
Dipartimento di Fisica ed Astronomia, Università di Catania, via Santa Sofia 64, I-95123 Catania, Italy
Present affiliation: CSNSM, Université Paris-Sud 11, CNRS/IN2P3, 91405 Orsay, France
Institute for Nuclear Research, 117312 Moscow, Russia
IPNO, Université Paris-Sud 11, CNRS/IN2P3, 15 rue Georges Clémenceau, 91406 Orsay, France
Dipartimento di Fisica, Università di Messina, salita Sperone, I-98166 Messina, Italy
Kyungpook National University, 702-701, Daegu, Republic of Korea
Received: date / Revised version: date
Abstract. The double polarization (beam-recoil) observables Ox and Oz have been measured for the reaction γp → K + Λ from threshold production to Eγ ∼ 1500 MeV. The data were obtained with the linearly
polarized beam of the GRAAL facility. Values for the target asymmetry T could also be extracted despite
the use of an unpolarized target. Analyses of our results by two isobar models tend to confirm the necessity
to include new or poorly known resonances in the 1900 MeV mass region.
PACS. 13.60.Le Meson production – 13.88.+e Polarization in interactions and scattering – 25.20.Lj Photoproduction reactions
1 Introduction
improvement by measuring cross sections with unprecedented precision for a large number of channels but they
A detailed and precise knowledge of the nucleon spec- also allowed a qualitative leap by providing for the first
troscopy is undoubtedly one of the cornerstones for our un- time high quality data on polarization observables. It is
derstanding of the strong interaction in the non-perturbative well known – and now well established – that these variregime. Today’s privileged way to get information on the ables, being interference terms of various multipoles, bring
excited states of the nucleon is light meson photo- and unique and crucial constraints for partial wave analysis,
electroproduction. The corresponding database has con- hence facilitating the identification of resonant contribusiderably expanded over the last years thanks to a com- tions and making parameter extraction more reliable.
bined effort of a few dedicated facilities worldwide. Not
From this perspective, K + Λ photoproduction offers
only did the recent experiments brought a quantitative
unique opportunities. Because the Λ is a self-analyzing
Send offprint requests to: lleres@lpsc.in2p3.fr
particle, several polarization observables can be ”easily”
2
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
measured via the analysis of its decay products. As a
consequence, this reaction already possesses the richest
database with results on the differential cross section [1][4], two single polarization observables (Σ and P ) [2]-[6]
and two double polarization observables (Cx and Cz ) recently measured by the CLAS collaboration [7]. On the
partial wave analysis side, the situation is particularly encouraging with most models concluding to the necessity of
incorporating new or poorly known resonances to reproduce the full set of data. Some discrepancies do remain
nonetheless either on the number of used resonances or
on their identification. To lift the remaining ambiguities,
new polarization obervables are needed calling for new experiments.
In the present work, we report on first measurements
of the beam-recoil observables Ox and Oz for the reaction
γp → K + Λ over large energy (from threshold to 1500
MeV) and angular (θcm = 30 − 1400 ) ranges. The target
asymmetry T , indirectly extracted from the data, is also
presented.
2 Experimental set-up
The experiment was carried-out with the GRAAL facility
(see [8] for a detailed description), installed at the European Synchrotron Radiation Facility (ESRF) in Grenoble
(France). The tagged and linearly polarized γ-ray beam is
produced by Compton scattering of laser photons off the
6.03 GeV electrons circulating in the storage ring.
In the present experiment, we have used a set of UV
lines at 333, 351 and 364 nm produced by an Ar laser,
giving 1.40, 1.47 and 1.53 GeV γ-ray maximum energies,
respectively. Some data were also taken with the green
line at 514 nm (maximum energy of 1.1 GeV).
The photon energy is provided by an internal tagging
system. The position of the scattered electron is measured
by a silicon microstrip detector (128 strips with a pitch of
300 µm and 1000 µm thick). The measured energy resolution of 16 MeV is dominated by the energy dispersion of
the electron beam (14 MeV - all resolutions are given as
FWHM). The energy calibration is extracted run by run
from the fit of the Compton edge position with a precision
of ∼10µm, equivalent to ∆Eγ /Eγ ≃ 2 × 10−4 (0.3 MeV
at 1.5 GeV). A set of plastic scintillators used for time
measurements is placed behind the microstrip detector.
Thanks to a specially designed electronic module which
synchronizes the detector signal with the RF of the machine, the resulting time resolution is ≈100 ps. The coincidence between detector signal and RF is used as a start
for all Time-of-Flight (ToF) measurements and is part of
the trigger of the experiment.
The energy dependence of the γ-ray beam polarization
was determined from the Klein-Nishina formula taking
into account the laser and electron beam emittances. The
UV beam polarization is close to 100% at the maximum
energy and decreases smoothly with energy to around 60%
at the KΛ threshold (911 MeV). Based on detailed studies
[8], it was found that the only significant source of error
5
1
6
25°
2
3
4
Fig. 1. Schematic view of the LAγRANGE detector: BGO
calorimeter (1), plastic scintillator barrel (2), cylindrical MWPCs (3), target (4), plane MWPCs (5), double plastic scintillator hodoscope (6) (the drawing is not to scale).
for the γ-ray polarization Pγ comes from the laser beam
polarization (δPγ /Pγ =2%).
A thin monitor is used to measure the beam flux (typically 106 γ/s). The monitor efficiency (2.68±0.03%) was
estimated by comparison with the response at low rate of
a lead/scintillating fiber calorimeter.
The target cell consists of an aluminum hollow cylinder of 4 cm in diameter closed by thin mylar windows
(100 µm) at both ends. Two different target lengths (6 and
12 cm) were used for the present experiment. The target
was filled by liquid hydrogen at 18 K (ρ ≈ 7 10−2 g/cm3 ).
The 4π LAγRANGE detector of the GRAAL set-up
allows to detect both neutral and charged particles (fig.
1). The apparatus is composed of two main parts: a central
one (250 ≤ θ ≤ 1550 ) and a forward one (θ ≤ 250 ).
The charged particle tracks are measured by a set of
MultiWire Proportional Chambers (MWPC) (see [5] for a
detailed description). To cover forward angles, two plane
chambers, each composed of two planes of wires, are used.
The detection efficiency of a track is about 95% and the
average polar and azimuthal resolutions are 1.50 and 20 ,
respectively. The central region is covered by two coaxial
cylindrical chambers. Single track efficiencies have been
extracted for π 0 p and π + n reactions and were found to be
≥90%, in agreement with the simulation. Since this paper
deals with polarization observables, no special study was
done to assess the efficiency of multi track events. Angular
resolutions were also estimated via simulation, giving 3.50
in θ and 4.50 in ϕ.
Charged particle identification in the central region is
obtained by dE/dx technique thanks to a plastic scintillator barrel (32 bars, 5 mm thick, 43 cm long) with an
energy resolution ≈20%. For the charged particles emitted in the forward direction, a Time-of-Flight measurement is provided by a double plastic scintillator hodoscope
(300×300×3 cm3 ) placed at a distance of 3 m from the
target and having a resolution of ≈600 ps. This detector
provides also a measure of the energy loss dE/dx. Energy
calibrations were extracted from the analysis of the π 0 p
photoproduction reaction while the ToF calibration of the
forward wall was obtained from fast electrons produced in
the target.
Photons are detected in a BGO calorimeter made of
480 (15θ×32ϕ) crystals, each of 21 radiation lengths. They
are identified as clusters of adjacent crystals (3 on average for an energy threshold of 10 MeV per crystal) with no
associated hit in the barrel. The measured energy resolution is 3% on average (Eγ =200-1200 MeV). The angular
resolution is 60 and 70 for polar and azimuthal angles,
respectively (Eγ ≥ 200 MeV and ltarget =3 cm).
3 Data analysis
10
3
10
2
γ+p → K +Λ
+
3.1 Channel selection
For the present results, the charged decay of the Λ (Λ →
pπ − , BR=63.9%) was considered and the same selection
method used in our previous publication on KΛ photoproduction [5] was applied. Only the main points will be
recalled in the following.
Only events with three tracks and no neutral cluster
detected in the BGO calorimeter were retained. In the
absence of a direct measurement of energy and/or momentum of the charged particles, the measured angles (θ,
ϕ) of the three tracks were combined with kinematical
constraints to calculate momenta. Particle identification
was then obtained from the association of the calculated
momenta with dE/dx and/or ToF measurements.
The main source of background is the γp → pπ + π −
reaction, a channel with a similar final state and a cross
section hundred times larger. Selection of the KΛ final
state was achieved by applying narrow cuts on the following set of experimental quantities:
3
Counts
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
γ+p → p+π +π
+
0
2
4
6
8
10
-
12
14
ct (cm)
Fig. 2. Reconstructed Λ decay length spectrum after all selection cuts (closed circles) for events with at least two tracks
in the cylindrical chambers. The solid line represents the fit
with an exponential function α ∗ exp(−ct/cτ ) where α and cτ
are free parameters. The second distribution (open circles) was
obtained without applying selection cuts. It corresponds to the
main background reaction (γp → pπ + π − ) which, as expected,
contributes only to small ct values.
noted that the deficit in the first bins is attributed to finite resolution effects not fully taken into account in the
simulation.
The first spectrum was fitted for ct≥1 cm by an exponential
function α ∗ exp(−ct/cτ ) with α and cτ as free
. Energy balance.
parameters. The fitted cτ value (8.17±0.31 cm) is in good
. Effective masses of the three particles extracted from agreement with the PDG expectation for the Λ mean free
the combination of measured dE/dx and ToF (only at path (cτΛ =7.89 cm) [9].
forward angles) with calculated momenta.
By contrast, the spectrum without cuts is dominated
by pπ + π − background events. As expected, they contribute
. Missing mass mγp−K + evaluated from Eγ , θK (mea- mostly to small ct values (≤2-3 cm), making the shape of
sured) and pK (calculated).
this distribution highly sensitive to background contamination. For instance, a pronounced peak already shows up
For each of these variables, the width σ of the corre- when opening selection cuts at ±3σ.
A remaining source of background, which cannot be
sponding distribution (Gaussian-like shape) were extracted
from a Monte-Carlo simulation of the apparatus response seen in the ct plot presented above, originates from the
contamination by the reaction γp → K + Σ 0 . Indeed, events
based on the GEANT3 package of the CERN library.
To check the quality of the event selection, the dis- where the decay photon is not detected are retained by
tribution of the Λ decay length was used due to its high the first selection step. Since these events are kinematically analyzed as KΛ ones, most of them are nevertheless
sensitivity to background contamination.
Event by event, track information and Λ momentum rejected by the selection cuts. From the simulation, this
were combined to obtain the distance d between the re- contamination was found to be of the order of 2%.
As a further check of the quality of the data sample,
action and decay vertices. The Λ decay length was then
calculated with the usual formula ctΛ = d/(βΛ ∗ γΛ ). Fig. the missing mass spectrum was calculated. One should
2 shows the resulting distributions for events selected with remember that the missing mass is not directly measured
all cuts at ±2σ (closed circles) compared with events with- and is not used as a criterion for the channel identificaout cuts (open circles). These spectra were corrected for tion. The spectrum presented in fig. 3 (closed circles) is
detection efficiency losses estimated from the Monte-Carlo in fair agreement with the simulated distribution (solid
simulation (significant only for ct≥5 cm). It should be line). Some slight discrepancies can nevertheless be seen
4
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
Counts
3.2.1 Formalism
For a linearly polarized beam and an unpolarized target,
the differential cross section can be expressed in terms of
the single polarization observables Σ, P , T (beam asymmetry, recoil polarization, target asymmetry, respectively)
and of the double polarization observables Ox , Oz (beamrecoil), as follows [10]:
1 dσ
dσ
[1 − Pγ Σ cos 2ϕγ
=
ρf
dΩ
2 dΩ 0
+σx′ Pγ Ox sin 2ϕγ
+σy′ (P − Pγ T cos 2ϕγ )
(1)
+σz′ Pγ Oz sin 2ϕγ ]
γ+p → K +Λ
+
8000
7000
6000
5000
4000
3000
2000
0
+
K +Σ
mΛ
1000
1
1.05
1.1
1.15
1.2
0
1.25
2
Missing mass (GeV/c )
Fig. 3. Distribution of the missing mass mγp−K + reconstructed from measured Eγ and θK and calculated pK . Data
after all selection cuts (closed circles) are compared to the simulation (solid line). The expected contribution from the reaction γp → K + Σ 0 is also plotted (note that it is not centered
on the Σ 0 mass due to kinematical constraints in the event
analysis). The vertical arrow indicates the Λ mass.
in the high energy tail of the spectra. The simulated missing mass distribution of the contamination from the γp →
K + Σ 0 reaction, also displayed in fig. 3, clearly indicates
that such a background cannot account for the observed
differences. Rather, these are attributed to the summation of a large number of data taking periods with various experimental configurations (target length, wire chambers, green vs UV laser line, ...). Although these configurations were implemented in corresponding simulations,
small imperfections (misalignments in particular) could
not be taken into account.
To summarize, thanks to these experimental checks,
we are confident that the level of background in our selected sample is limited. This is corroborated by the simulation from which the estimated background contamination (multi-pions and K + Σ 0 contributions) never exceeds
5% whatever the incident photon energy or the meson recoil angle.
3.2 Measurement of Ox , Oz and T
As will be shown below, the beam-recoil observables Ox
and Oz , as well as the target asymmetry T , can be extracted from the angular distribution of the Λ decay proton.
ρf is the density matrix for the lambda final state and
(dσ/dΩ)0 the unpolarized differential cross section. The
Pauli matrices σx′ ,y′ ,z′ refer to the lambda quantization
axes defined by ẑ ′ along the lambda momentum in the
center-of-mass frame and ŷ ′ perpendicular to the reaction
plane (fig. 4). Pγ is the degree of linear polarization of the
beam along an axis defined by n̂ = x̂ cos ϕγ + ŷ sin ϕγ ; the
photon quantization axes are defined by ẑ along the proton center-of-mass momentum and ŷ=ŷ ′ (fig. 4). We have
ϕγ = ϕlab − ϕ, where ϕlab and ϕ are the azimuthal angles
of the photon polarization vector and of the reaction plane
in the laboratory axes, respectively (fig. 5).
The beam-recoil observables Cx and Cz measured by
the CLAS collaboration with a circularly polarized beam
[7] were obtained using another coordinate system for describing the hyperon polarization, the ẑ ′ axis being along
the incident beam direction instead of the momentum of
one of the recoiling particles (see fig. 4). Such a nonstandard coordinate system was chosen to give the results their simplest interpretation in terms of polarization
transfer but implied the model calculations to be adapted.
To check the consistency of our results with the CLAS values (see sect. 4.1), our Ox and Oz values were converted
using the following rotation matrix:
Oxc = −Ox cos θcm − Oz sin θcm
Ozc = Ox sin θcm − Oz cos θcm
(2)
It should be noted that our definition for Ox and Oz
(eq. 1) has opposite sign with respect to the definition
given in the article [11], which is used in several hadronic
models. We chose the same sign convention than the CLAS
collaboration.
For an outgoing lambda with an arbitrary quantization
axis n̂′ , the differential cross section becomes:
PΛ · n̂′
h
dσ i
dσ
= T r σ · n̂′ ρf
dΩ
dΩ
(3)
where PΛ is the polarization vector of the lambda. If the
polarization is not observed, the expression for the differential cross section reduces to:
h dσ i
dσ
(4)
= T r ρf
dΩ
dΩ
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
^x'
c
K
dσ
(ϕlab = 900 ) =
dΩ
+
^z'
c
^y'
c
→
γ
^z
lab
proton
^
y'
Λ
[1 + Pγ Σ cos 2ϕ]
(7)
0
N (ϕlab = 900 ) − N (ϕlab = 00 )
= Pγ Σ cos 2ϕ
N (ϕlab = 900 ) + N (ϕlab = 00 )
^y
z^
^z'
dσ
dΩ
The beam asymmetry values Σ published in [5] were extracted from the fit of the azimuthal distributions of the
ratio:
Θcm
→
5
^
x'
(8)
^x
3.2.2 Λ polarization and spin observables
Fig. 4. Definition of the coordinate systems and polar angles
in the center-of-mass frame (viewed in the reaction plane). The
[x̂′ ,ŷ ′ ,ẑ ′ ] system is used to specify the polarization of the outgoing Λ baryon: ẑ ′ is along the Λ momentum and ŷ ′ perpendicular to the reaction plane. The [x̂,ŷ,ẑ] system is used to specify
the incident photon polarization: ẑ is along the incoming proton momentum and ŷ identical to ŷ ′ . The polar angle θcm of
the outgoing K + meson is defined with respect to the incident
beam direction ẑlab . [x̂′c ,ŷc′ ,ẑc′ ] is the coordinate system chosen by the CLAS collaboration for the Λ polarization. The x̂′c
and ẑc′ axes are obtained from x̂′ and ẑ ′ by a rotation of angle
π + θcm .
^y
lab
y^
^n
^x
ne
n pla
^z
lab
tio
reac
^x
lab
Fig. 5. Definition of the coordinate systems and azimuthal angles in the center-of-mass frame (viewed perpendicularly to the
beam direction). The [x̂lab ,ŷlab ,ẑlab ] system corresponds to the
laboratory axes with ẑlab along the incident beam direction.
The [x̂,ŷ,ẑ] system, used to define the incident photon polarization, has its axes x̂ and ŷ along and perpendicular to the
reaction plane (azimuthal angle ϕ), respectively. The polarization of the beam is along n̂ (azimuthal angle ϕlab ). The two
beam polarization states correspond to ϕlab = 00 (horizontal)
and ϕlab = 900 (vertical) (ϕlab = ϕγ + ϕ).
which leads to:
dσ
=
dΩ
dσ
dΩ
[1 − Pγ Σ cos 2ϕγ ]
(5)
0
0
For horizontal (ϕlab = 0 ) and vertical (ϕlab = 900 ) photon polarizations, the corresponding azimuthal distributions of the reaction plane are therefore:
dσ
dσ
0
[1 − Pγ Σ cos 2ϕ]
(6)
(ϕlab = 0 ) =
dΩ
dΩ 0
The components of the lambda polarization vector deduced from eqs. 1 to 5 are:
′
′
PΛx ,z =
′
PΛy =
Pγ Ox,z sin 2ϕγ
1 − Pγ Σ cos 2ϕγ
P − Pγ T cos 2ϕγ
1 − Pγ Σ cos 2ϕγ
(9)
(10)
These equations provide the connection between the Λ
polarisation PΛ and the spin observables Σ, P , T , Ox
and Oz .
Integration of the polarization components over the
azimuthal angle ϕ of the reaction plane writes:
R i
dσ
(ϕ)dϕ
PΛ (ϕ) dΩ
i
R dσ
(11)
< PΛ > =
dΩ (ϕ)dϕ
where i stands for x′ , y ′ or z ′ .
When integrating over the full angular domain, the
averaged x′ and z ′ components of the polarization vector vanish while the y ′ component is equal to P . On the
other hand, when integrating over appropriatly chosen angular sectors, all three averaged components can remain
different from zero. For horizontal and vertical beam polarizations, the following expressions are obtained when
considering the four particular ϕ domains defined hereafter [12] (recalling ϕγ = ϕlab − ϕ):
. S1 = [π/4, 3π/4] ∪ [5π/4, 7π/4]:
′
< PΛy > (ϕlab = 00 ) = (P π + 2Pγ T )/(π + 2Pγ Σ)
′
< PΛy > (ϕlab = 900 ) = (P π − 2Pγ T )/(π − 2Pγ Σ)
. S2 = [−π/4, π/4] ∪ [3π/4, 5π/4]:
′
< PΛy > (ϕlab = 00 ) = (P π − 2Pγ T )/(π − 2Pγ Σ)
′
< PΛy > (ϕlab = 900 ) = (P π + 2Pγ T )/(π + 2Pγ Σ)
. S3 = [0, π/2] ∪ [π, 3π/2]:
′ ′
< PΛx ,z > (ϕlab = 00 ) = −2Pγ Ox,z /π
′ ′
< PΛx ,z > (ϕlab = 900 ) = +2Pγ Ox,z /π
. S4 = [π/2, π] ∪ [3π/2, 2π] :
′ ′
< PΛx ,z > (ϕlab = 00 ) = +2Pγ Ox,z /π
′ ′
< PΛx ,z > (ϕlab = 900 ) = −2Pγ Ox,z /π
6
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
It should be noted that these four sectors cover the full ϕ
range. In the following, these different combinations of ϕ
sectors and polarization states will be labelled by the sign
plus or minus appearing in the corresponding expressions
for < PΛi >.
3.2.3 Decay angular distribution
In the lambda rest frame, the angular distribution of the
decay proton is given by [13]:
W (cos θp ) =
1
1 + α|PΛ | cos θp
2
3.2.4 Experimental extraction
As for Σ and P , the observables Ox , Oz and T were extracted from ratios of the angular distributions, in order
to get rid of most of the distorsions introduced by the
experimental acceptance.
Including the detection efficiencies, the yields measured as a function of the proton angle with respect to
the different axes write:
′
′
x ,z
N±
=
′ ′
′ ′
2Pγ Ox,z
1 x′ ,z′
N0± ǫ± (cos θpx ,z ) 1 ± α
cos θpx ,z
2
π
(18)
(12)
′
where α=0.642±0.013 [9] is the Λ decay parameter and
θp the angle between the proton direction and the lambda
polarization vector.
From this expression, one can derived an angular distribution for each component of PΛ :
y
=
N±
′
′
P π ± 2Pγ T
1 y′
N ǫ± (cos θpy ) 1 + α
cos θpy (19)
2 0±
π ± 2Pγ Σ
From the integration of the azimuthal distributions given
by eqs. 6 and 7 over the different angular sectors, it can
be shown that:
′
1
W (cos θpi ) = 1 + αPΛi cos θpi
2
′
′
1
(1 + αP cos θpy )
2
′
y
N0+
y′
N0−
′
′
′
> 0) −
> 0) +
< 0)
1
= αP
2
< 0)
′
′
y
y
N+
+ N−
=
′
′
W± (cos θpx ,z ) =
′
W± (cos θpy ) =
′ ′
2Pγ Ox,z
1
1±α
cos θpx ,z
2
π
′
P π ± 2Pγ T
1
1+α
cos θpy
2
π ± 2Pγ Σ
(16)
(17)
π + 2Pγ Σ
π − 2Pγ Σ
(21)
′ ′
1 x′ ,z′
x′ ,z ′
)ǫ± (cos θpx ,z )
+ N0−
(N
2 0+
′
′
1 y′
y′
)ǫ± (cos θpy )(1 + αP cos θpy )
(N0+ + N0−
2
(23)
x ,z
2N+
When integrating over the different angular domains
specified above (sectors S1 + S2 for y ′ -axis and S3 + S4 for
x′ -,z ′ -axes, appropriatly combined with the two beam polarization), the proton angular distributions with respect
to the three quantization axes can be written as follows:
=
(22)
(14)
(15)
′
x ,z
x ,z
=
+ N−
N+
′
′
N (cos θpy
′
N (cos θpy
(20)
Assuming that the detection efficiencies do not depend on
the considered ϕ sectors (ǫ+ (cos θpi ) = ǫ− (cos θpi ) - the validity of this assumption will be discussed later on), we can
then calculate the following sums and ratios from which
the efficiency cancels out:
where P is the recoil polarization. Our P results published in [5] were determined directly from the measured
up/down asymmetry:
′
N (cos θpy
′
N (cos θpy
′
(13)
where θpi is now the angle between the proton direction
and the quantization axis i (x′ , y ′ or z ′ ).
The components being determined in the Λ rest frame,
a suitable transformation should be applied to calculate
them in the center-of-mass frame. However, as the boost
direction is along the lambda momentum, it can be shown
that the polarization measured in the lambda rest frame
remains unchanged in the center-of-mass frame [7].
When integrating over all possible azimuthal angles ϕ,
the proton angular distribution with respect to the y ′ -axis
simply writes:
W (cos θpy ) =
′
′
x ,z
x ,z
= N0−
N0+
x′ ,z ′
N+
+
x′ ,z ′
N−
= (1 + α
′ ′
2Pγ Ox,z
cos θpx ,z )
π
P π+2P T
′
y
2N+
y
y
N+
+ N−
′
′
′
(24)
′
γ
y
2Pγ Σ 1 + α π+2Pγ Σ cos θp
(25)
= 1+
′
π
1 + αP cos θpy
To illustrate the extraction method of Ox , Oz and T ,
the N+ and N− experimental distributions together with
their sums and ratios, summed over all photon energies
and meson polar angles, are displayed in figs. 6 (x′ -axis),
7 (z ′ -axis) and 8 (y ′ -axis). Thanks to the efficiency correction given by the distributions N+ + N− (figs. 6,7,8-c),
the ratios 2N+ /(N+ + N− ) (figs. 6,7,8-d), from which the
efficiency drops out, exhibit the expected dependence in
cos θp and can be therefore fitted by the functions given in
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
the r.h.s. of eqs. 24 and 25. The known energy dependence
of Pγ and the previously measured values for Σ and P [5]
are then used to deduce Ox , Oz and T from the fitted
slopes.
The validity of the hypothesis ǫ+ (cos θpi ) = ǫ− (cos θpi )
was studied via the Monte Carlo simulation in which a
polarized Λ decay was included. The efficiencies ǫ± calculated from the simulation are presented in plots e) of figs.
6 to 8 and the ratios ǫ− /ǫ+ in plots f) (open circles). As
one can see, for the y ′ case, this ratio remains very close to
1 whatever the angle while, for x′ and z ′ , the discrepancy
from 1 is more pronounced and evolves with the angle.
This shows that some corrections should be applied on
the measured ratios 2N+ /(N+ + N− ) to take into account
the non-negligible differences observed between ǫ+ and ǫ− .
The correction factors, plotted in figs. 6,7,8-f) (closed circles), were calculated through the following expression:
Cor =
i
i
2N+
2N+
/
i
i
i
i
gen
N+ + N−
N+ + N− sel
(26)
where gen and sel stand for generated and selected events.
Since ǫ± = (N± )sel /(N± )gen , it can be re-written as:
Cor =
i
i
2N+
ǫi− N−
1
[1
+
i + N i gen
i gen ]
2 N+
ǫi+ N+
−
(27)
7
To illustrate this second extraction method, the corrected distributions, summed over all photon energies and
meson polar angles, are displayed in figs. 6,7-j) (x′ , z ′ -axes)
and 8-h),i) (y ′ -axis). They were obtained by dividing the
originally measured distributions (figs. 6,7-h and 8-a,b) by
the corresponding efficiency distributions (figs. 6,7-i and
8-e). In the y ′ -axis case, the two corrected spectra N± /ǫ±
were simultaneously fitted.
This method gives results in good agreement with those
extracted from the first method. Nevertheless, the resulting χ2 were found to be significantly larger (the global
reduced-χ2 values are given in figs. 6 to 8 - they are close
to 1 for the first method and five to ten times larger for
the second one). The first method, which relies upon ratios
leading to an intrinsic first order efficiency correction, is
less dependent on the simulation details and was therefore
preferred.
Three sources of systematic errors were taken into account: the laser beam polarization (δPγ /Pγ =2%), the Λ
decay parameter α (δα = 0.013) and the hadronic background. The error due to the hadronic background was
estimated from the variation of the extracted values when
cuts were changed from ±2σ to ±2.5σ. Given the good
agreement between the two extraction methods, no corresponding systematic error was considered. For the T observable, the measured values for Σ and P being involved,
their respective errors were included in the estimation of
the uncertainty. All systematic and statistical errors have
been summed quadratically.
The corrected distributions are displayed in the plots g) of
figs. 6 to 8. After correction, as expected, the slope of the
y ′ distribution is unaffected while the slopes of the x′ and
z ′ distributions are slightly modified. These distributions
were again fitted to obtain the final values of Ox , Oz and 4 Results and discussions
T.
As the detection efficiencies and the correction factors The complete set of beam-recoil polarization and target
calculated from the simulation depend on the input values asymmetry data is displayed in figs. 9 to 15. These data
of Ox , Oz and T , an iterative method was used. Three cover the production threshold region (Eγ =911-1500 MeV)
iterations were sufficient to reach stable values.
and a large angular range (θkaon = 30 − 1400 ). NumeriFor a consistency check, an alternative extraction method cal values are listed in tablescm1 to 3. Error bars are the
was implemented. The angular distributions were directly quadratic sum of statistical and systematic errors.
corrected by the simulated efficiencies and fitted according
to:
′
′
′
′
x ,z ,inv
x ,z
+ N−
N+
x′ ,z ′
ǫ+
x′ ,z ′ ,inv
+ ǫ−
=
′ ′
2Pγ Ox,z
1 x′ ,z′
N0+ 1 + α
cos θpx ,z
2
π
(28)
′
y
N+
′
ǫy+
′
y
N−
y′
ǫ−
=
=
′
1 y′
P π + 2Pγ T
N0+ 1 + α
cos θpy
2
π + 2Pγ Σ
(29)
′
1 y′ π − 2Pγ Σ
P π − 2Pγ T
1+α
N
cos θpy (30)
2 0+ π + 2Pγ Σ
π − 2Pγ Σ
where N inv and ǫinv stand for N (− cos θp ) and ǫ(− cos θp ),
respectively. This trick, used for the x′ and z ′ cases, allows to combine the N+ and N− distributions which have
opposite slopes (eq. 18).
4.1 Observable combination and consistency check
In pseudoscalar meson photoproduction, one can extract
experimentally 16 different quantities: the unpolarized differential cross section (dσ/dΩ)0 , 3 single polarization observables (P , T , Σ), 4 beam-target polarizations (E, F ,
G, H), 4 beam-recoil polarizations (Cx , Cz , Ox , Oz ) and
4 target-recoil polarizations (Tx , Tz , Lx , Lz ). The various
spin observables are not independent but are constrained
by non-linear identities and various inequalites [10], [11],
[14], [15]. In particular, of the seven single and beam-recoil
polarization observables, only five are independent being
related by the two equations:
Cx2 + Cz2 + Ox2 + Oz2 = 1 + T 2 − P 2 − Σ 2
(31)
Cz Ox − Cx Oz = T − P Σ
(32)
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
0.10
0.08
εε+
0.06
Cor*2N+/(N++N-)
ε++ε-inv
0.04
0.12
1.1
e)
Ox=-0.025
2
χ =0.7
1.0
0.9
g)
0.10
0.08
0.06
0.04
0.02
-1
i)
-0.5
0
0.5
1
CosΘpx
ε- / ε+ and Cor
0.12
1.2
N++N-inv (a.u.)
ε- and ε+
0.9
c)
N++N- (a.u.)
80
1.0
140
140
d)
ε- / ε+
Cor
1.1
1.0
60
2
120
N- (a.u.)
e)
Oz=0.57
χ =1.6
2
g)
0.10
0.08
-0.5
j)
0.04
0
0.02
-1
0.5
1
i)
-0.5
0
CosΘpx
P 2 + Ox2 + Oz2 ≤ 1
(34)
+
εε+
0.06
100
(33)
Σ +
0.08
0.12
χ =5.8
Ox=0.016
|T ± P | ≤ 1 ± Σ
Oz2
0.10
h)
80
There are also a number of inequalities involving three of
these observables:
Ox2
0.12
1.6
1.4
1.2
1.0
0.8
0.6
Fig. 6. Angular distributions for the decay proton in the
lambda rest frame with respect to the x′ -axis: a) distribution N+ ; b) distribution N− ; c) sum N+ + N− ; d) ratio
2N+ /(N+ + N− ); e) efficiencies ǫ+ (triangles) and ǫ− (circles)
calculated from the simulation; f) ratio ǫ− /ǫ+ (open circles)
and correction factor Cor (closed circles) given by eq. 26 calculated from the simulation; g) ratio 2N+ /(N+ + N− ) corrected
inv
by the factor Cor; h) distribution N+ + N−
, with N inv =
inv
inv
N (− cos θp ); i) efficiency ǫ+ + ǫ− , with ǫ
= ǫ(− cos θp ),
inv
calculated from the simulation; j) distribution N+ + N−
corrected by the efficiency ǫ+ + ǫinv
.
The
solid
line
in
d)
and g)
−
represents the fit by the (linear) function given in the r.h.s.
of eq. 24. The solid line in j) represents the fit by the (linear)
function given in the r.h.s. of eq. 28. The reduced-χ2 and the
Ox value obtained from the fits are reported in d), g) and j).
2
80
0.04
100
80
-1
100
f)
120
c)
1.6
1.4
1.2
1.0
0.8
0.6
120
0.06
0.9
a)
140
60
ε- and ε+
100
1.1
Ox=-0.099
2
χ =0.9
Cor*2N+/(N++N-)
2N+/(N++N-)
120
1.2
(N++N-inv)/(ε++ε-inv) (a.u.)
N++N- (a.u.)
140
60
b)
30
80
70
60
50
40
30
2N+/(N++N-)
40
a)
30
1.2
50
ε++ε-inv
40
60
80
70
60
50
40
30
≤1
(35)
0.5
1
CosΘpz
ε- / ε+ and Cor
50
70
1.2
N++N-inv (a.u.)
60
z,- distributions
N+ (a.u.)
70
N- (a.u.)
N+ (a.u.)
x,- distributions
140
(N++N-inv)/(ε++ε-inv) (a.u.)
8
180
160
140
120
100
80
60
40
-1
b)
Oz=0.62
χ =1.9
2
d)
ε- / ε+
Cor
1.1
1.0
0.9
f)
120
100
80
h)
60
Oz=0.54
χ =12.8
2
j)
-0.5
0
0.5
1
CosΘpz
Fig. 7. Angular distributions for the decay proton in the
lambda rest frame with respect to the z ′ -axis (all distributions
as in fig. 6). The reduced-χ2 and the Oz value obtained from
the fits are reported in d), g) and j).
P 2 + Cx2 + Cz2 ≤ 1
(36)
Σ 2 + Cx2 + Cz2 ≤ 1
(37)
These different identities and inequalities can be used
to test the consistency of our present and previous measurements. They can also be used to check the compatibility of our data with the results on Cx and Cz recently
published by the CLAS collaboration [7].
Our measured values for Σ, P , T , Ox and Oz were
combined to test the above inequalities. Equation 31 was
used to calculate the quantity Cx2 + Cz2 appearing in expressions 36 and 37. The results for the two combinations
|T ± P | ∓ Σ of the three single polarizations are presented
in fig. 12. The results for the quantities:
. (P 2 + Ox2 + Oz2 )1/2 ,
. (Σ 2 + Ox2 + Oz2 )1/2 ,
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
. (1 + T 2 − P 2 − Ox2 − Oz2 )1/2 = (Σ 2 + Cx2 + Cz2 )1/2 ,
. (1 + T 2 − Σ 2 − Ox2 − Oz2 )1/2 = (P 2 + Cx2 + Cz2 )1/2 ,
c)
1.6
1.4
1.2
1.0
0.8
0.6
N- (a.u.)
a)
80
70
60
50
40
30
20
0.12
0.10
0.08
0.06
εε+
Cor*2N+/(N++N-)
1.6
1.4
1.2
1.0
0.8
0.6
N-/ε- (a.u.)
0.04
100
e)
T=-0.57
χ =1.1
2
ε- / ε+ and Cor
2N+/(N++N-)
160
140
120
100
80
60
40
1.2
N+/ε+ (a.u.)
ε- and ε+
N++N- (a.u.)
N+ (a.u.)
y,- distributions
80
70
60
50
40
30
20
100
80
T=-0.58
χ =1.0
2
d)
ε- / ε+
Cor
1.1
1.0
0.9
f)
T=-0.57
80
χ =11.6
2
60
20
-1
T=-0.57
b)
40
g)
h)
-0.5
0
0.5
1
CosΘpy
χ =13.3
2
60
40
20
-1
i)
-0.5
0
0.5
9
which combine single and double polarization observables,
are displayed in figs. 13 and 14. All these quantities should
be ≤ 1. The plotted uncertainties are given by the standard error propagation. Whatever the photon energy or
the meson polar angle, no violation of the expected inequalities is observed, confirming the internal consistency
of our set of data.
Since all observables entering in eqs. 31 and 32 were
measured either by GRAAL (Σ, P , T , Ox , Oz ) or by
CLAS (P , Cx , Cz - their P data were confirmed by our
measurements [5]), the two sets of data can be therefore compared and combined. Within the error bars, the
agreement between the two sets of equal combinations
(1+T 2 −Σ 2 −Ox2 −Oz2 )1/2 (GRAAL) and (P 2 +Cx2 +Cz2 )1/2
(CLAS) is fair (fig. 14) and tends to confirm the previously observed saturation to the value 1 of R = (P 2 +
Cx2 + Cz2 )1/2 , whatever the energy or angle. Fig. 15 displays the values for the combined GRAAL-CLAS quantity Cz Ox − Cx Oz − T + P Σ. Within the uncertainties,
the expected value (1) is obtained, confirming again the
overall consistency of the GRAAL and CLAS data.
It has been demonstrated [14] that the knowledge of
the unpolarized cross section, the three single-spin observables and at least four double-spin observables - provided
they have not all the same type - is sufficient to determine
uniquely the four complex reaction amplitudes. Therefore,
only one additional double polarization observable measured using a polarized target will suffice to extract unambiguously these amplitudes.
1
CosΘpy
Fig. 8. Angular distributions for the decay proton in the
lambda rest frame with respect to the y ′ -axis: a) distribution N+ ; b) distribution N− ; c) sum N+ + N− ; d) ratio
2N+ /(N+ + N− ); e) efficiencies ǫ+ (triangles) and ǫ− (circles)
calculated from the simulation - they are symmetrical about
θcm = 900 (we find ǫdown /ǫup =1.03); f) ratio ǫ− /ǫ+ (open circles) and correction factor Cor (closed circles) given by eq. 26
calculated from the simulation; g) ratio 2N+ /(N+ + N− ) corrected by the factor Cor; h) distribution N+ corrected by the
efficiency ǫ+ ; i) distribution N− corrected by the efficiency ǫ− .
The solid line in d) and g) represents the fit by the (non-linear)
function given in the r.h.s. of eq. 25. These distributions exhibit a linear behaviour since the overall recoil polarisation P
is very low (the value extracted from the up/down asymmetry
of the raw distribtion N+ + N− is -0.12). The solid line in h)
and i) represents the simultaneous fit by the (linear) functions
given in the r.h.s. of eqs. 29 and 30. The reduced-χ2 and the T
value obtained from the fits are reported in d), g), h) and i).
4.2 Comparison to models
We have compared our results with two models: the Ghent
isobar RPR (Regge-plus-resonance) model [16]-[19] and
the coupled-channel partial wave analysis developed by
the Bonn-Gatchina collaboration [20]-[24]. In the following, these models will be refered as RPR and BG, respectively. The comparison is shown in figs. 9 to 11.
The RPR model is an isobar model for KΛ photo- and
electroproduction. In addition to the Born and kaonic contributions, it includes a Reggeized t-channel background
which is fixed to high-energy data. The fitted database includes differential cross section, beam asymmetry and recoil polarization photoproduction results. The model variant presented here contains, besides the known N ∗ resonances (S11 (1650), P11 (1710), P13 (1720)), the P13 (1900)
state (** in the PDG [9]) and a missing D13 (1900) resonance. This solution was found to provide the best overall
agreement with the combined photo- and electroproduction database. As one can see in figs. 9 to 11, the RPR
prediction (dashed line) qualitatively reproduces all observed structures. Interestingly enough, the model best reproduces the data at high energy (1400-1500 MeV), where
the P13 (1900) and D13 (1900) contributions are maximal.
The BG model is a combined analysis of experiments
with πN , ηN , KΛ and KΣ final states. As compared
10
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
to the other models, this partial-wave analysis takes into
account a much larger database which includes most of
the available results (differential cross sections and polarization observables). For the γp → K + Λ reaction, the
main resonant contributions come from the S11 (1535),
S11 (1650), P13 (1720), P13 (1900) and P11 (1840) resonances.
To achieve a good description of the recent Cx and Cz
CLAS measurements, the ** P13 (1900) had to be introduced. It should be noted that, at this stage of the analysis, the contribution of the missing D13 (1900) is significantly reduced as compared to previous versions of the
model. As shown in figs. 9-11, this last version (solid line)
provides a good overall agreement. On the contrary, the
solution without the P13 (1900) (not shown) fails to reproduce the data.
More refined analyses with the RPR and BG models
are in progress and will be published later on. Comparison with the dynamical coupled-channel model of SaclayArgonne-Pittsburgh [25]-[27] has also started.
5 Summary
In this paper, we have presented new results for the reaction γp → K + Λ from threshold to Eγ ∼ 1500 MeV.
Measurements of the beam-recoil observables Ox , Oz and
target asymmetries T were obtained over a wide angular range. We have compared our results with two isobar
models which are in reasonable agreement with the whole
data set. They both confirm the necessity to introduce
new or poorly known resonances in the 1900 MeV mass
region (P13 and/or D13 ).
It should be underlined that from now on only one
additional double polarization observable (beam-target or
target-recoil) would be sufficient to extract the four helicity amplitudes of the reaction.
Acknowledgements
We are grateful to A.V. Sarantsev, B. Saghai, T. Corthals,
J. Ryckebusch and P. Vancraeyveld for communication of
their most recent analyses and J.M. Richard for fruitful
discussions on the spin observable constraints. We thank
R. Schumacher for communication of the CLAS data. The
support of the technical groups from all contributing institutions is greatly acknowledged. It is a pleasure to thank
the ESRF as a host institution and its technical staff for
the smooth operation of the storage ring.
References
1.
2.
3.
4.
5.
6.
7.
8.
R.K.Bradford et al., Phys. Rev. C 73, 035202 (2006).
K.-H. Glander et al., Eur. Phys. J. A 19, 251 (2004).
J.W.C. McNabb et al., Phys. Rev. C 69, 042201(R) (2004).
M. Sumihama et al., Phys. Rev. C 73, 035214 (2006).
A. Lleres et al., Eur. Phys. J. A 31, 79 (2007).
R.G.T. Zegers et al., Phys. Rev. Lett. 91, 092001 (2003).
R.K.Bradford et al., Phys. Rev. C 75, 035205 (2007).
O. Bartalini et al., Eur. Phys. J. A 26, 399 (2005).
9. Review of Particle Physics 2004, Phys. Lett. B 592, 1
(2004).
10. R.A. Adelseck and B. Saghai, Phys. Rev. C 42, 108 (1990).
11. I.S. Barker, A. Donnachie and J.K. Storrow, Nucl. Phys.
B 95, 347 (1975).
12. P. Calvat, Thesis, Université J. Fourier Grenoble, 1997,
unpublished.
13. T.D. Lee and C.N. Yang, Phys. Rev. 108, 1645 (1957).
14. W.T. Chiang and F. Tabakin, Phys. Rev. C 55, 2054
(1997).
15. X. Artru, J.M. Richard and J. Soffer, Phys. Rev. C 75,
024002 (2007).
16. T. Corthals, J. Ryckebusch and T. Van Cauteren, Phys.
Rev. C 73, 045207 (2006).
17. T. Corthals, Thesis, Universiteit Gent, 2007, unpublished.
18. T. Corthals et al., Phys. Lett. B 656, 186 (2007).
19. T. Corthals, J. Ryckebusch and P. Vancraeyveld, private
communication.
20. A.V. Anisovich et al., Eur. Phys. J. A. 25, 427 (2005).
21. A.V. Sarantsev et al., Eur. Phys. J. A. 25, 441 (2005).
22. A.V. Anisovich et al., Eur. Phys. J A 34, 243 (2007).
23. V.A. Nikonov et al., arXiv:hep-ph/0707.3600 (2007).
24. A. Sarantsev, private communication.
25. B. Juliá-Dı́az, B. Saghai, T.-S.H. Lee and F. Tabakin,
Phys. Rev. C 73, 055204 (2006).
26. B. Saghai, J.C. David, B. Juliá-Dı́az and T.-S.H. Lee, Eur.
Phys. J. A 31, 512 (2007).
27. B. Saghai, private communication.
Ox
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
11
1.5
1.0
980 MeV
1027 MeV
1074 MeV
1122 MeV
1171 MeV
1222 MeV
1272 MeV
1321 MeV
1372 MeV
1421 MeV
1466 MeV
0.5
0.0
-0.5
-1.0
-1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
1.0
60
120
180
0.5
0.0
-0.5
-1.0
-1.5
0
60
120
180
60
120
180
Θcm
Fig. 9. Angular distributions of the beam recoil observable Ox . Data are compared with the predictions of the BG (solid line)
and RPR (dashed line) models.
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
Oz
12
1.5
1.0
0.5
0.0
-0.5
-1.0
980 MeV
1027 MeV
1074 MeV
1122 MeV
1171 MeV
1222 MeV
-1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
1.0
0.5
0.0
-0.5
1272 MeV
-1.0
1321 MeV
1372 MeV
-1.5
1421 MeV
1.0
1466 MeV
60
120
180
0.5
0.0
-0.5
-1.0
-1.5
0
60
120
180
60
120
180
Θcm
Fig. 10. Angular distributions of the beam recoil observable Oz . Data are compared with the predictions of the BG (solid line)
and RPR (dashed line) models.
T
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
13
1.5
980 MeV
1.0
1027 MeV
1074 MeV
0.5
0.0
-0.5
-1.0
-1.5
1.0
1122 MeV
1171 MeV
1222 MeV
1272 MeV
1321 MeV
1372 MeV
0.5
0.0
-0.5
-1.0
-1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
1.0
1421 MeV
60
1466 MeV
120
180
0.5
0.0
-0.5
-1.0
-1.5
0
60
120
180
60
120
180
Θcm
Fig. 11. Angular distributions of the target asymmetry T . Data are compared with the predictions of the BG (solid line) and
RPR (dashed line) models.
14
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
|T-P|+Σ and |T+P|-Σ
GRAAL data
1.5
1.0
0.5
0.0
980 MeV
1027 MeV
1074 MeV
-0.5
1.0
0.5
0.0
1122 MeV
1171 MeV
1222 MeV
-0.5
1.0
0.5
0.0
1272 MeV
1321 MeV
1372 MeV
-0.5
60
120
180
1.0
|T-P|+Σ
0.5
|T+P|-Σ
0.0
1421 MeV
-0.5
0
60
120
1466 MeV
180
60
120
180
Θcm
Fig. 12. Angular distributions of the quantities |T − P | + Σ (closed circles) and |T + P | − Σ (open circles). We should have
the inequalities |T ± P | ∓ Σ ≤ 1 (eq. 33).
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
1/2
GRAAL data
2.0
980 MeV
1027 MeV
1074 MeV
1.5
1.0
0.5
0.0
1122 MeV
1171 MeV
1222 MeV
1.5
1.0
0.5
0.0
1272 MeV
1321 MeV
1372 MeV
1421 MeV
1466 MeV
60
1.5
1.0
2
(P +Ox2+Oz2)
1/2
2
, (Σ +Ox2+Oz2)
1/2
2
2
, (1+T -P -Ox2-Oz2)
15
0.5
0.0
120
180
1.5
2
1/2
2
1/2
(P +Ox2+Oz2)
1.0
(Σ +Ox2+Oz2)
2
0.5
2
0.0
(1+T -P -Ox2-Oz2)
0
60
120
180
60
120
1/2
180
Θcm
Fig. 13. Angular distributions of the quantities (P 2 + Ox2 + Oz2 )1/2 (stars), (Σ 2 + Ox2 + Oz2 )1/2 (circles) and (1 + T 2 − P 2 −
Ox2 − Oz2 )1/2 = (Σ 2 + Cx2 + Cz2 )1/2 (crosses). The first and third sets of data are horizontally shifted for visualization. All these
quantities should be ≤ 1 (inequalities 34, 35 and 37).
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
GRAAL vs CLAS data
2.0
980 MeV
1027 MeV
1074 MeV
(1032 MeV)
1.5
1.0
0.5
0.0
1122 MeV
1.5
1171 MeV
1222 MeV
(1132 MeV)
(1232 MeV)
1.0
2
2
(1+T -Σ -Ox2-Oz2)
1/2
2
and (P +Cx2+Cz2)
1/2
16
0.5
0.0
1272 MeV
1321 MeV
1372 MeV
(1332 MeV)
1.5
1.0
0.5
0.0
1421 MeV
1.5
60
1466 MeV
(1433 MeV)
180
GRAAL data
(1+T2-Σ2-Ox2-Oz2)1/2
1.0
CLAS data
0.5
0.0
0
120
(P2+Cx2+Cz2)1/2
60
120
180
60
120
180
Θcm
Fig. 14. Angular distributions of the quantity (1 + T 2 − Σ 2 − Ox2 − Oz2 )1/2 = (P 2 + Cx2 + Cz2 )1/2 . This quantity should be ≤ 1
(inequality 36). Comparison to the values (P 2 + Cx2 + Cz2 )1/2 published by the CLAS collaboration (open squares - energy in
parentheses). Note that the Ox2 + Oz2 and Cx2 + Cz2 values are independent of the choice for the axes specifying the Λ polarization
(see sect. 3.2.1).
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
17
CzOx-CxOz-T+PΣ
GRAAL × CLAS data
1.0
1027 (1032) MeV
1122 (1132) MeV
1222 (1232) MeV
1321 (1332) MeV
0.5
0.0
-0.5
-1.0
0.5
0.0
-0.5
-1.0
60
1421 (1433) MeV
120
180
0.5
GRAAL data :
Σ , P , T , Ox , Oz
0.0
CLAS data :
Cx , Cz
-0.5
-1.0
0
60
120
180
Θcm
Fig. 15. Angular distributions of the quantity Cz Ox − Cx Oz − T + P Σ. This quantity is calculated using the Cx and Cz
results published by the CLAS collaboration (energy in parentheses) combined with our Ox and Oz data converted by eq. 2
to have the same ẑ ′ axis convention and with our Σ, P and T measurements. The used CLAS data are those corresponding
to the angles cos θcm =0.85, mean(0.65,0.45), mean(0.25,0.05), -0.15, mean(-0.35,-0.55) and -0.75. We should have the equality
Cz Ox − Cx Oz − T + P Σ = 0 (eq. 32).
18
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
Table 1. Beam-recoil Ox values.
θcm (o )
31.3
59.1
80.7
99.8
118.9
138.6
θcm (o )
34.1
58.9
80.5
99.3
119.4
140.4
Eγ =980 MeV
θcm (o ) Eγ =1027 MeV θcm (o ) Eγ =1074 MeV θcm (o ) Eγ =1122 MeV
0.349 ± 0.150
30.6
0.425 ± 0.108
31.2
0.502 ± 0.103
32.4
0.570 ± 0.154
0.320 ± 0.255
57.5
0.408 ± 0.225
57.5
0.202 ± 0.140
57.6
0.179 ± 0.161
-0.094 ± 0.262
81.7 -0.085 ± 0.189
81.0
-0.190 ± 0.120
80.6 -0.365 ± 0.282
-0.464 ± 0.320
99.8 -0.477 ± 0.133
99.7
-0.552 ± 0.106
99.2 -0.522 ± 0.165
-0.490 ± 0.244
119.0 -0.723 ± 0.142
119.3
-0.621 ± 0.105
119.9 -0.795 ± 0.146
-1.028 ± 0.410
139.5 -0.304 ± 0.259
138.8
-0.163 ± 0.200
139.3 -0.341 ± 0.252
Eγ =1171 MeV θcm (o ) Eγ =1222 MeV θcm (o ) Eγ =1272 MeV θcm (o ) Eγ =1321 MeV
0.567 ± 0.122
34.6
0.294 ± 0.179
35.8
0.526 ± 0.130
36.0
0.599 ± 0.161
0.306 ± 0.148
58.9
0.310 ± 0.139
59.2
0.345 ± 0.140
59.4
0.304 ± 0.155
-0.319 ± 0.254
80.4 -0.193 ± 0.124
80.6
0.028 ± 0.155
80.3 -0.073 ± 0.142
-0.510 ± 0.175
98.9 -0.548 ± 0.252
99.2
-0.232 ± 0.125
99.2 -0.383 ± 0.141
-0.347 ± 0.162
119.1 -0.615 ± 0.121
119.9
-0.395 ± 0.143
119.9
0.046 ± 0.114
-0.160 ± 0.234
140.3
0.116 ± 0.242
140.8
0.454 ± 0.187
141.3
0.323 ± 0.158
θcm (o ) Eγ =1372 MeV θcm (o ) Eγ =1421 MeV θcm (o ) Eγ =1466 MeV
36.1
0.552 ± 0.119
35.7
0.455 ± 0.144
35.9
0.307 ± 0.150
59.5
0.150 ± 0.130
59.6 -0.072 ± 0.160
59.3
0.172 ± 0.171
80.1 -0.168 ± 0.131
80.3 -0.303 ± 0.139
80.0 -0.270 ± 0.195
99.4 -0.276 ± 0.122
99.7 -0.190 ± 0.137
99.7 -0.096 ± 0.139
120.4
0.124 ± 0.138
120.4
0.076 ± 0.110
120.8
0.164 ± 0.145
141.9
0.636 ± 0.126
142.8
0.490 ± 0.205
143.7
0.905 ± 0.152
Table 2. Beam-recoil Oz values.
θcm (o )
31.3
59.1
80.7
99.8
118.9
138.6
θcm (o )
34.1
58.9
80.5
99.3
119.4
140.4
Eγ =980 MeV
θcm (o ) Eγ =1027 MeV θcm (o ) Eγ =1074 MeV θcm (o ) Eγ =1122 MeV
0.581 ± 0.194
30.6
0.333 ± 0.110
31.2
0.285 ± 0.080
32.4
0.274 ± 0.124
0.956 ± 0.242
57.5
0.951 ± 0.216
57.5
0.674 ± 0.112
57.6
0.687 ± 0.127
0.754 ± 0.186
81.7
0.995 ± 0.154
81.0
1.003 ± 0.148
80.6
0.888 ± 0.244
1.139 ± 0.237
99.8
0.949 ± 0.140
99.7
1.140 ± 0.130
99.2
0.950 ± 0.144
0.841 ± 0.215
119.0
0.744 ± 0.162
119.3
0.996 ± 0.156
119.9
0.618 ± 0.197
-0.091 ± 0.597
139.5 -0.287 ± 0.415
138.8
0.427 ± 0.223
139.3 -0.162 ± 0.568
Eγ =1171 MeV θcm (o ) Eγ =1222 MeV θcm (o ) Eγ =1272 MeV θcm (o ) Eγ =1321 MeV
0.398 ± 0.093
34.6
0.291 ± 0.177
35.8
0.532 ± 0.087
36.0
0.554 ± 0.090
0.914 ± 0.128
58.9
0.678 ± 0.167
59.2
0.710 ± 0.119
59.4
0.904 ± 0.108
0.825 ± 0.123
80.4
0.485 ± 0.109
80.6
0.867 ± 0.112
80.3
0.767 ± 0.153
0.964 ± 0.175
98.9
1.025 ± 0.143
99.2
0.676 ± 0.188
99.2
0.734 ± 0.161
0.550 ± 0.190
119.1
0.426 ± 0.166
119.9
0.677 ± 0.166
119.9
0.409 ± 0.229
-0.055 ± 0.286
140.3 -0.162 ± 0.276
140.8
0.349 ± 0.272
141.3 -0.448 ± 0.217
θcm (o ) Eγ =1372 MeV θcm (o ) Eγ =1421 MeV θcm (o ) Eγ =1466 MeV
36.1
0.600 ± 0.084
35.7
0.384 ± 0.094
35.9
0.354 ± 0.095
59.4
0.784 ± 0.119
59.6
0.558 ± 0.185
59.3
0.814 ± 0.222
80.1
0.484 ± 0.112
80.3
0.322 ± 0.195
80.0
0.666 ± 0.332
99.4
0.419 ± 0.120
99.7
0.289 ± 0.134
99.7 -0.023 ± 0.192
120.4
0.019 ± 0.145
120.4
-0.313 ± 0.131
120.8 -0.432 ± 0.180
141.9 -0.072 ± 0.159
142.8
-0.085 ± 0.172
143.7 -0.461 ± 0.162
Please give a shorter version with: \authorrunning and \titlerunning prior to \maketitle
Table 3. Target asymmetry T values.
θcm (o )
31.3
59.1
80.7
99.8
118.9
138.6
θcm (o )
34.1
58.9
80.5
99.3
119.4
140.4
Eγ =980 MeV
θcm (o ) Eγ =1027 MeV θcm (o ) Eγ =1074 MeV θcm (o ) Eγ =1122 MeV
-0.506 ± 0.156
30.6 -0.663 ± 0.112
31.2
-0.635 ± 0.096
32.4 -0.615 ± 0.118
-0.607 ± 0.206
57.5 -0.860 ± 0.190
57.5
-0.718 ± 0.120
57.6 -0.991 ± 0.237
-0.803 ± 0.185
81.7 -0.749 ± 0.139
81.0
-0.874 ± 0.141
80.6 -0.949 ± 0.239
-0.622 ± 0.166
99.8 -0.974 ± 0.121
99.7
-1.048 ± 0.140
99.2 -0.833 ± 0.153
-0.622 ± 0.187
119.0 -0.789 ± 0.133
119.3
-0.760 ± 0.102
119.9 -0.825 ± 0.165
-1.090 ± 0.341
139.5 -0.681 ± 0.359
138.8
-0.448 ± 0.203
139.3 -0.465 ± 0.296
Eγ =1171 MeV θcm (o ) Eγ =1222 MeV θcm (o ) Eγ =1272 MeV θcm (o ) Eγ =1321 MeV
-0.715 ± 0.116
34.6 -0.858 ± 0.155
35.8
-0.773 ± 0.123
36.0 -1.064 ± 0.133
-0.869 ± 0.154
58.9 -0.874 ± 0.145
59.2
-0.827 ± 0.166
59.4 -0.910 ± 0.142
-0.850 ± 0.158
80.4 -0.871 ± 0.124
80.6
-0.979 ± 0.234
80.3 -0.716 ± 0.133
-0.659 ± 0.159
98.9 -0.690 ± 0.132
99.2
-0.707 ± 0.150
99.2 -0.576 ± 0.223
-0.669 ± 0.150
119.1 -0.675 ± 0.145
119.9
-0.125 ± 0.208
119.9 -0.281 ± 0.174
0.226 ± 0.316
140.3 -0.066 ± 0.249
140.8
0.482 ± 0.213
141.3
0.331 ± 0.198
θcm (o ) Eγ =1372 MeV θcm (o ) Eγ =1421 MeV θcm (o ) Eγ =1466 MeV
36.1 -0.983 ± 0.104
35.7 -0.753 ± 0.112
35.9 -0.632 ± 0.127
59.5 -0.695 ± 0.113
59.6 -0.687 ± 0.159
59.3 -0.648 ± 0.166
80.1 -0.669 ± 0.123
80.3 -0.564 ± 0.131
80.0 -0.553 ± 0.185
99.4 -0.482 ± 0.175
99.7 -0.025 ± 0.157
99.7
0.190 ± 0.196
120.4 -0.104 ± 0.135
120.4
0.160 ± 0.112
120.8
0.785 ± 0.195
141.9
0.629 ± 0.147
142.8
0.859 ± 0.140
143.7
0.933 ± 0.175
19