WYDZIA FIZYKI, ASTRONOMII
I INFORMATYKI STOSOWANEJ
INSTYTUT FIZYKI im. MARIANA SMOLUCHOWSKIEGO
Studia porównaw ze oddziaªywania
w niskoenergety zny h ukªada h
ppη
ippη
′
Paweª Klaja
pra aprzygotowanawZakªadzieFizykiJ¡drowej
InstytutuFizyki im. MarianaSmolu howskiego,
WydziaªuFizyki,AstronomiiiInformatykiStosowanejUniwersytetuJagiello«skiego,oraz
wInstytu ieFizyki J¡drowejwCentrumBadaw zymJüli h,
pod kierunkiemProf. PawªaMoskala.
FACULTY OF PHYSICS, ASTRONOMY
AND APPLIED COMPUTER SCIENCE
MARIAN SMOLUCHOWSKI INSTITUTE OF PHYSICS
Comparative studies of the
intera tion in the low energy
ppη
andppη
′
systemsPaweª Klaja
dissertationpreparedat theNu lear Physi sDepartment
oftheMarianSmolu howskiInstitute ofPhysi s,atFa ultyofPhysi s,AstronomyandApplied
ComputerS ien eoftheJagiellonianUniversity
andin InstituteofNu learPhysi satResear hCentreJüli h,
guidedbyProf. PaweªMoskal.
The COSY-11 ollaboration measured the
pp → ppη
andpp → ppη
′
rea tions in
order to perform omparative studiesof the intera tions within theproton-proton-meson
system.
Thisthesis presents indetail the analysis of the
pp → ppη
′
rea tion whi h wasmeasured
attheproton beam momentumof 3.260GeV/ .
Theelaborationresultsindierentialdistributionsofsquaredinvariantproton-proton(
s
pp
) andproton-η
′
(
s
pη
′
) masses,aswell asinangular distributions and thetotal ross se tion atan ex ess energy of16.4MeV.The dierential distributions
s
pp
ands
pη
′
are ompared to theoreti al predi tions and to theanalogous spe tradetermined for thepp → ppη
rea tion.The omparison of theresults for the
η
andη
′
meson produ tion rather ex ludes the
hy-pothesis that the enhan ement observed inthe invariant mass distributions is due to the
meson-proton intera tion.
Further, the shapes of the distributions do not favour any of the postulated theoreti al
models.
Stresz zenie
W rama h grupy badaw zej COSY-11 wykonano pomiary reak ji
pp → ppη
orazpp → ppη
′
w elu przeprowadzenia studiów porównaw zy h oddziaªywania w ukªadzie
proton-proton-mezon.
W rozprawie doktorskiej zaprezentowano analiz dany h z pomiaru reak ji
pp → ppη
′
wykonanegoz wykorzystaniem wi¡zki protonowej o pdzie 3.260GeV/ .
Wynikiem analizy przedstawionej w niniejszej pra y s¡ ró»ni zkowe przekroje zynne w
funk ji mas niezmienni zy h proton-proton (
s
pp
) i proton-η
′
(
s
pη
′
), rozkªady k¡towe oraz aªkowityprzekrój zynnydla energiiwzbudzenia Q=16.4MeV.Rozkªady ró»ni zkowe
s
pp
orazs
pη
′
zostaªy porównane z przewidywaniami teorety znymi orazanalogi znymi widmamiotrzymanymidlareak jipp → ppη
.Porównanierezultatówotrzymany hdlaproduk jimezonów
η
iη
′
pozwalanawyklu zenie
hipotezy,»ewzmo nienieobserwowane wwidma h masniezmienni zy h jestpowodowane
oddziaªywaniem mezonu zprotonem.
1 Introdu tion 11
2 Low energy intera tion within a proton-proton-meson system 15
2.1 Ex itation fun tionsfor
pp → ppη
andpp → ppη
′
rea tions . . . 15 2.2 Comparison ofη
,η
′
andπ
0
mesonintera tion withprotons . . . 16
2.3 Denitions ofobservables . . . 18
2.4
pp
andp − meson
invariant massdistributions . . . 193 Experimental fa ility 23
3.1 CoolerSyn hrotron COSY . . . 23
3.2 COSY-11 dete torsetup . . . 25
3.3 Triggerlogi . . . 27
4 Calibration of the dete tor setup 29
4.1 Spa e-time relationfor drift hambers . . . 29
4.2 Time-of-ight alibration ofs intillatordete tors . . . 30
4.3 Monitoring ofrelative beam-targetsettings . . . 31
5 Identi ation of the
pp → ppη
′
rea tion 35 5.1 Identi ation ofprotons . . . 35 5.2 Identi ation oftheη
′
meson . . . 36 6 Luminositydetermination 397 Determinationofthe spread andtheabsolutevalueof thebeam
momen-tum 43
8 Fine tunning of the relative dipole- hamber settings 47
9 Evaluation of the dierential distributions 49
9.1 Kinemati al t . . . 49
9.2 Ba kground subtra tion . . . 52
10.2 Totaland dierential rossse tions . . . 64
10.3 Comparison withresults for the
η
mesonprodu tion . . . 70 10.4s
pp
ands
pη
′
distributions inviewof theoreti alpredi tions . . . 7211Summary 77
A Pseudos alar mesons 79
B Parameterization of the proton-proton FinalState Intera tion 81
C Formalism ofa ombinedanalysis ofphoto- andhadro-produ tionof the
η
′
meson 83The knowledge of the meson-nu leon intera tion at the hadroni level is one of the
ma-jor goals in nu lear physi s, nowadays. Also, studies of meson stru ture and produ tion
me hanisms onstitute a huge interest of nu lear and parti le physi ists. During the last
de adesmanymeasurementsprodu edinterestingresults[1℄,butvariousquestionsarestill
open.
In the SU(3)-avour s heme the
η
andη
′
mesons belong to the nonet meson family of
pseudos alar mesons. The
η
andη
′
mesons onstitute a mixture of states: singlet-
η
1
and o tet-η
8
.It is important to stress that the strength of the
pη
andpη
′
intera tion depends on the
stru tureofthe
η
andη
′
mesonsand isdire tlyrelatedto thesinglet-
η
1
ando tet-η
8
on-tributions inthe wave fun tionsofthese mesons[2 4℄.Taking into a ount the mixing angle the resulting ontribution of the various quark
avours inthe
η
andη
′
wave fun tionisalmost the same. However, inspite of the
postu-latedsimilar quarkstru turethe
η
andη
′
mesonsoweunexpe tedlydierent features.
Themost drasti onesare:
η
andη
′
mesonspossessdierentmasses,the
η
(547)mesonisalmosttwotimeslighter than theη
′
(958) meson [5℄.
Thebran hingratiosofBandD
s
mesonsfor thede aysintotheη
′
meson arehigher
than for the de ays into the
η
meson, and deviate strongly from model predi tions [6,7℄. There isno experimental eviden efor baryoni resonan ewhi h de aysby the
emis-sion of an
η
′
meson [5,8℄ while e.g. the resonan es N(1535) and N(1650) de ay via
the emissionofthe
η
mesonwitha signi ant probability [5℄.These dierent features ould imply a possible dieren e in the intera tion of
η
andη
′
mesons with elementary parti les, and indi ate that also the produ tion me hanism of
both mesonsinelementary parti le ollisions might vary.
Due to the short life-time of the avour neutral pseudos alar mesons, experiments with
meson beams or targets are di ult or hardly feasible. Therefore, the nu leon-meson
in-tera tion anbestudied onlyviaits inuen eon the rossse tionsoftherea tionsduring
whi h they areprodu ed (e.g.
N N → NN Meson
) [1℄.Quantitative information about theintera tion an be gained fromthe shapeof the
ex i-tation fun tions for the
pp → ppη
andpp → ppη
′
rea tions as well as from a omparison
ofthose to the
ppπ
0
Up to now, only the proton intera tion with pions and the
η
-meson [1℄ was studied moreexhaustivelydue tomu hmorehighertotal rossse tionsin omparisontothe rossse tionfor the
pp → ppη
′
rea tion.
Besides the ex itation fun tion also dierential distributions of invariant proton-proton
and proton-meson masses onstitute a sensitive toolfor studies of the intera tion within
the meson-nu leon system. The distribution of the proton-
η
invariant mass showed a lear enhan ement in the region of small proton-η
relative momenta [9℄. The observed ee t ould be explainedbythenonnegligiblerole oftheproton-η
intera tion inthenal state [10,11℄, the admixture of higher partial waves during theη
produ tion [12℄, or the energydependen eoftheprodu tion amplitude[13℄. Usingonly thepp → ppη
data, itis howevernot possible to justifyor falsify anyof the above mentioned hypothesis.Theendeavour toexplaintheobservedenhan ementmotivatedtheexperimentandthe
analysisof the
pp → ppη
′
rea tion whi h ispresented inthis thesis. This experiment was
performed using theCOSY-11 dete tor setup[14 16℄, installedat the ooler syn hrotron
COSY[17℄attheResear hCentreJüli hinGermany. Itwas ondu tedinanenergyrange
losetothekinemati althresholdforthe
η
′
mesonprodu tion,wheretherelativevelo ities
oftheprodu ed parti lesaresmall.
The analysis and results of the
pp → ppη
′
rea tion measurement, ondu ted in order to
determinethe distributionof eventsoverthephasespa efor anex essenergy rangeequal
to the one measured before for the
pp → ppη
rea tion are presented in this thesis. The measurementwasperformedat thenominalbeammomentumof3.257GeV/orrespond-ingto thenominalex ess energy for the
ppη
′
rea tion equal to 15.5MeV.The analysisof
the
ppη
′
data wasperformed ina similarway asithasbeen done for the
ppη
system. The omparison of the dierential distributions for theproton-proton and for theproton-meson invariant masses inthe
η
andη
′
produ tion ouldhelpto judge about thevalidity
ofpostulatedtheories on erningtheobserved enhan ement and allows for aquantitative
estimationofthe relativestrengthoftheproton-
η
andtheproton-η
′
intera tions,provided
thattheee t is aused bythe proton-meson intera tion.
Inthe following hapter the urrent statusof the
pp → ppη
andpp → ppη
′
total ross
se tionmeasurements ispresented. Furthermore,thepossibilityofproton-
η
andproton-η
′
intera tionstudies isdis ussed andthedes ription oftheobservables usedintheanalysis
des ribed in this thesis is given. The results a hieved for the
pp → ppη
measurement as well astheavailable theoreti aldes riptions ofthe results arepresented.Chapter3 is devoted to thepresentation of the experimental fa ilityused to perform the
pp → ppη
′
measurement. In the next hapter, themethods used for the alibrationof the
COSY-11 dete tors and their relative geometri al settings are presented. The time-spa e
relation of the drift hambers and the pro edure for the time-of-ight alibration is
In hapter5 the method ofthe
pp → ppη
′
rea tionidenti ation willbe depi ted and the
method ofidentifyingmeasured andalso unobserved parti leswill be given.
Furtheron,the pro edureoftheluminosity(
L
)determinationwillbepresentedin hapter 6,andinthe onse utive haptersitwillbeshownhowtheabsolutevalueandthespreadofthebeammomentumwasextra ted,andthemethodfor thedeterminationoftheposition
ofthedrift hambers relative to theCOSY-11 dipole willbeexplained.
Chapter9 omprises the evaluation pro edure of thedierential distributions. First, the
kinemati altpro edureand thentheba kground subtra tion dueto themulti-pion
pro-du tionwill bedis ussed.
The nal results on erning the total ross se tion and dierential distributions are
pre-sented in hapter 10. The a eptan e orre tions are dis ussed and the a hieved
experi-mentalresults are ompared to theoreti al predi tions.
The on lusionsarepresentedin hapter 11.
In the appendi es at the end of the dissertation some issues dis ussed in the thesis are
explained inmore detail. In therst one thestru ture of thepseudos alar meson nonet,
mesonmassesandquarkstru turearepresented. These ondoneisdevotedtothe
des rip-tion of the parameterization of the on-shell proton-proton intera tion. General remarks
about the ombined analysis of the
η
′
meson formalism in photo- and hadro-produ tion
arepresentedinthe third addendum,and the linearenergy dependen e oftheprodu tion
system
The intera tion of hadrons is the ree tion of the strong for e between the quarks, and
provides information about the hadron stru ture and the strong intera tion itself [1℄. In
theframework of theopti al model, the intera tion between hadrons an be expressedin
termsofphase-shifts, whi hinthezeroenergylimit aredes ribedbythes atteringlength
andee tive rangeparameters [1℄. Thesevariablesarequitewell establishedfor the
(low-energy) nu leon-nu leon intera tion [18,19℄, but they are poorlyknown for the
nu leon-meson or meson-meson intera tions. The estimated real part of the s attering length of
the
η
-proton potential, depending on themethod of the analysis and studied region, is 3 to10 times[1℄ largerthan for theπ
0
-proton s attering (
a
pπ
0
=0.13fm)[20,21℄, whilefor theη
′
meson only anupperlimit isknownof 0.8fm[22℄.
Theintera tion of mesons(e.g. pseudos alar mesons 1
:
π, K, η, η
′
) withnu leons ould be
dedu edfromtheexperimentsrealizedbymeansofmesonbeams,butsu hexperimentsare
notfeasiblein aseoftheavourneutralmesonsduetotheirshortlifetime[1,5℄. However,
the study of their intera tion with hadrons is ertainly a essible via their inuen e on
the rossse tion of rea tions like
N N → NN Meson
in whi h they areprodu ed [1℄. In su h a ase, the intera tion within the nalmesonnu leon system will modify theshapeof the ex itation fun tion and of the dierential distributions of invariant masses of the
nu leon-nu leon-meson systems.
2.1 Ex itation fun tions for
pp
→ ppη
andpp
→ ppη
′
rea tions
Nearthekinemati althresholdmeasurements ofnu leon-nu leon ollisions allowto study
theparti le produ tion witha dominant ontribution from one partial wave only. Inthis
energy range, the dependen e of the total ross se tion as a fun tion of the
entre-of-mass ex ess energy is predominantly determined by the available phase spa e and the
intera tion between theexitparti les. The ex itation fun tionsfor the
pp → ppη
′
[23 28℄
and
pp → ppη
[28 33℄ rea tions are presented in gure 2.1. Comparing the data to the arbitrarily normalized phase-spa e integral reveals that proton-proton FSI enhan es thetotal ross se tion by more than one order of magnitude for low energies. In ase of the
η
′
meson produ tion one re ognizes that the data are des ribed well assuming that the on-shellproton-proton amplitude ex lusively determinesthephase-spa e population.1
1
10
10
2
10
3
10
4
10
5
1
10
10
2
pp
→
pp
η
pp
→
pp
η
´
Q
[
MeV
]
σ
[
nb
]
Figure2.1:The
pp → ppη
andpp → ppη
′
ex itationfun tions[2333℄. Thedashedlinesindi ate
arbitrarilynormalizedfun tions obtainedunder theassumption ofthehomogeneousphasespa e
o upation. Solidlines orrespondto al ulationsofthephasespa eweightedbytheproton-proton
on-shells atteringamplitude[22℄.
Thisindi atesthattheproton-
η
′
intera tionistoosmalltomanifestitselfintheex itation
fun tion within the presently a hieved statisti al un ertainty. However, for the
η
meson produ tionthe enhan ementisbyaboutafa toroftwolargerthanin aseoftheη
′
meson
and annot be des ribed bythe
pp
-FSIonly.2.2 Comparison of
η
,η
′
and
π
0
meson intera tion with protons
Thestrength ofthe intera tion dedu ed fromthe omparison ofthedata andthe linesin
gure2.1 depends on the model of theproton-proton intera tion usedin the al ulations
for the
ppη
andppη
′
systems [22℄. Therefore, inorder to estimatea relative strength
be-tweenthe
pη
andpη
′
intera tions inamodelindependent wayone an ompare theshape
oftheex itationfun tionofthe
pp → ppη
andpp → ppη
′
rea tions. Moreover,one angain
somequantitativeinformation abouttheseintera tionsbya omparison oftheseshapesto
the
ppπ
0
system [1℄,sin e the
π
0
0
0.5
1
1.5
2
2.5
3
10
-1
1
10
10
2
10
3
V
phs
[
MeV
2
]
|M
0
η
| / |M
0
π
| a.u.
0
0.5
1
1.5
2
2.5
3
10
-1
1
10
10
2
10
3
V
phs
[
MeV
2
]
|M
0
η
´
| / |M
0
π
| a.u.
Figure2.2: The ratios
|M
η
0
|/|M
0
π
|
(left) and|M
η
′
0
|/|M
0
π
|
(right) extra ted from data al ulat-ing thepp
-FSI a ording to the formulas from referen e [22℄, and negle ting the proton-meson intera tion[1℄. Theratioisshownasafun tionofthephasespa evolume[22℄.For thatpurpose, one an ompareonly the dependen eof theprodu tion amplitudes
|M
0
|
derived fromthe datatakinginto a ount thepp
-FSI only. The dependen e of|M
η
0
|
and|M
η
′
0
|
as a fun tion of the phase spa e volumes normalized to|M
π
0
|
arepresentedingure 2.2. The|M
0
|
for theη
,η
′
and
π
0
mesonswere extra ted
from data, disregarding any proton-meson intera tion. When the negle ted
η
(η
′
)-proton
intera tion would have been the same as the one for proton-
π
0
, the points in the plots
should have been onsitent with unity as an be seen for the
pp → ppη
′
rea tion, when
really the intera tion shows its weakness, independently of the pres ription used for the
proton-proton nalstateintera tion [22℄. In the aseof the
η
′
meson produ tionits weak
intera tionwithnu leons at the low-energyrangeisexpe teddueto thela kofany
bary-oni resonan es whi h ould de ayinto a
N η
′
system[1,8℄ 2
.
Statisti alun ertaintiesallowedto getonly avery onservativeupperlimitfor thereal
partofthes attering lengthof the proton-
η
′
potential resulting in:
|Re a
pη
′
| < 0.8 f m
[1,26℄.Thus,independentofthemodelusedforthepres riptionofthe
pp
-FSI,froma omparison ofthe energy dependen e of the produ tion amplitudes for thepp → ppη
,pp → ppη
′
and
pp → ppπ
0
rea tions,itwas on luded thattheintera tion within theproton-
η
′
system is
mu h weakerthan theintera tion between theproton and the
η
meson.Another possibility of learning about intera tions within nu leon-nu leon-meson systems
2
ontribu-is given by the dierential distributions of the invariant masses. This is why thepresent
analysisofthe
ppη
′
systemhasbeenperformedinasimilarwayasithasbeendoneearlier[9℄
for the
ppη
system. The determinedpp
andp
-meson invariant mass distributions will be usedfor a omparative study of the intera tion withintheproton-meson system.In the next se tion the denitions of the studied observables whi h will be used in the
furtheranalysis arepresented.
2.3 Denitions of observables
To des ribe thestudied three parti le (
ppη
′
)systemone needs onlyve independent
vari-ablesinthe entre-of-mass system. Inthisframe,dueto energy andmomentum
onserva-tion, momentum ve torsof protons and
η
′
lie inone plane, alled rea tionplane. In that
plane (shown s hemati ally ingure 2.3) the relative momenta of parti les are des ribed
byonly two variables. Thesequantities may be hosenassquare oftheproton-proton
in-variantmass
s
pp
and squareof theproton-η
′
invariant mass
s
pη
′
. Invariant massesdepend on the relative velo ity of the parti les and are therefore well suited for a des ription oftheintera tions betweenthese parti les. Besidesthe relativemovement ofparti lesonthe
rea tion plane three other variables have to be dened for xing the orientation of the
rea tionplaneinthe oordinatesystem.
Figure2.3:S hemati denitionsofthe entre-of-masskinemati alvariablesusedforthe
des rip-tionofthe
ppη
′
system. Inthe entre-of-massframethemomentumve torsofthethreeoutgoing
parti lesarelo atedwithin therea tionplane. Inthisplanetherelativemotionof theeje tilesis
xed by the square of theinvariant masses
s
pp
ands
pη
′
. Three remainingvariables theφ
∗
η
′
,θ
∗
η
′
and
ψ
anglesareusedtodenetheorientationoftheemissionplaneinspa e.Inthisthesis, byanalogyto the evaluationofthe
ppη
system[9℄,theazimuthaland polar angles of theη
′
meson momentum ve tor relative to the beam dire tion, denoted as
φ
∗
η
′
and
θ
∗
aroundthedire tion ofthe momentum ve torof the
η
′
meson.
Theintera tion between nalstate parti lesdoesnot alter theorientation oftherea tion
plane[9℄. Therefore,it willmanifest itself onlyinthe distributionof theinvariant masses
s
pp
ors
pη
′
,or generallyinthe population ofthe Dalitz plot(s
pp
vs.s
pη
′
).In the ase of non-intera ting parti les in the nal state these distributions should
or-respond to a homogeneously populated phase spa e. Therefore, their intera tion should
showupasa deviation fromthese expe tation.
2.4
pp
andp
− meson
invariant mass distributionsOnly two invariant masses of three subsystemsare independent and therefore the whole
a essible information about the nal state intera tion an be shown in the Dalitz plot.
One an also usethe proje tion of thephase-spa e distribution onto theinvariant masses
ofproton-proton or proton-meson subsystems[9℄.
Thequalitativephenomenologi alanalysisofthedetermineddierentialinvariant
proton-protonandproton-
η
massdistributionsrevealedanenhan ementofthepopulationdensity atthekinemati alregion orrespondingtoasmallproton-η
momentum. Theproton-proton andproton-η
invariant massdistributions determined for thepp → ppη
rea tion at an ex- ess energy of 15.5 MeV are presented in gure 2.43
. The dashed lines in both panels
of the gure depi t the results of al ulations where only the on-shell amplitude of the
proton-proton intera tion hasbeen taken into a ount.
In those al ulations the enhan ement fa tor has been estimated as the square of
the on-shell proton-proton s attering amplitude derived using the modied
Cini-Fubini-Stanghelliniformula in ludingthe Wong-Noyes Coulomb orre tions[22℄.
One an easily see that the mentioned ee t is too large to be des ribed by the on-shell
in lusionof the proton-proton FSI.
In fa t a better des ription is a hieved when ontributions from higher partial waves
or o-shell ee ts of the proton-proton potential are taken into a ount. These
al ula-tions ompared totheexperimentallydetermineddierential proton-protoninvariantmass
distribution arepresentedingure 2.5. Inthe leftpanel of this gurethe experimentally
determined dierential rossse tion as a fun tion of the squaredinvariant proton-proton
massis ompared tothe al ulations ofV.Baruand ollaborators[39℄ underthe
assump-tionof a
3
P
0
→
1
S
0
s
transitiona ording to the models des ribed in[39℄, depi ted as the solidline.Dashed and dotted lines on the left panel of gure 2.5 represent the al ulations of
K. Nakayamaand his group[12℄. The authors laim that the ontribution of the S-wave
aloneisunabletoexplaintheobservedenhan ementinthesquaredproton-protoninvariant
3
0
20
40
60
80
100
120
3.52
3.54
3.56
3.58
s
pp
[
GeV
2
/c
4
]
d
σ
/ds
pp
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
phase space
FSI
pp
on-shell
0
20
40
60
80
100
2.2
2.22
2.24
2.26
s
p
η
[
GeV
2
/c
4
]
d
σ
/ds
p
η
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
phase space
FSI
pp
on-shell
Figure2.4:Distributionsofthesquareoftheproton-proton(
s
pp
)(left)andproton-η
(s
pη
)(right) invariantmasses determinedexperimentallyforthepp → ppη
rea tion( losedsquares). The inte-gralsofthephasespa eweightedbythesquareoftheproton-protonon-shells atteringamplitude(dotted lines)-FSI
pp
have been normalized arbitrarily at small values ofs
pp
. The expe tations undertheassumptionofahomogeneouslypopulatedphasespa eareshownassolidlines.0
20
40
60
80
100
120
3.52
3.54
3.56
3.58
s
pp
[
GeV
2
/c
4
]
d
σ
/ds
pp
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
1
S
0
s
1
S
0
s
1
S
0
s +
3
P
0
s
0
20
40
60
80
100
120
3.52
3.54
3.56
3.58
s
pp
[
GeV
2
/c
4
]
d
σ
/ds
pp
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
pp + ηp
3 - body
Figure2.5: (Left panel) Distribution of the square of the proton-proton (
s
pp
) invariant mass determined experimentally for thepp → ppη
rea tion. Solid and dashed lines orrespond to al ulationsunder the assumption ofa3
P
0
→
1
S
0
s
transition a ordingto themodelsdes ribed in[39℄and[12℄,respe tively. Thedottedlineshowstheresultof al ulationswithin lusionofthe1
S
0
→
3
P
0
s
ontribution as suggestedin [12℄. (Right panel) The samedata as in the left panel but omparedwiththree-body al ulations[10,11℄. Thesolidlinewasdeterminedwitharigorousmassdistribution. Seekingfor thebetter des riptionthey postulate thattheshapeof the
enhan ement an be reprodu ed by folding the relative momentum of the proton-proton
subsystem withthe available phasespa e [12℄ suggesting that the enhan ement ould be
the onsequen eof the
pp
P-wave inthe nalstate. Cal ulations assuming a3
P
0
→
1
S
0
s
transition orrespond to the dashed urve and the result of al ulations with the in lusion of the1
S
0
→
3
P
0
s
ontribution is depi ted by the dotted line. Although the dotted line orresponding to al ulations based on thestrongerP-wave ontribution is inquite good agreement to the experimental determined
dierential distribution of the proton-proton invariant mass, it underestimates the
to-tal ross se tion data taken for the
pp → ppη
rea tion near the kinemati al threshold (Qlower than 30MeV)[1℄.Onthe otherhand, thedis ussed ee t an inprin iple be assignedto hanges of the
produ tion amplitude, sin e in al ulations by V. Baru et al. [39℄, and by K. Nakayama
and ollaborators[12℄ the produ tion amplitude wasnearly onstant.
Ananalysisguidedbytheassumptionofalinearenergydependen eoftheprodu tion
am-plitudewasperformedbyA.Delo[13℄. Thesquaredinvariant massdistributions
s
pp
ands
pη
′
determined for thepp → ppη
rea tionmeasuredat anex ess energyof Q=15.5MeV are ompared to al ulationsperformedbyA.Deloingure2.6.0
20
40
60
80
100
120
3.52
3.54
3.56
3.58
s
pp
[
GeV
2
/c
4
]
d
σ
/ds
pp
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
0
20
40
60
80
100
2.2
2.22
2.24
2.26
s
p
η
[
GeV
2
/c
4
]
d
σ
/ds
p
η
[
µ
b/(GeV
2
/c
4
)
]
COSY-11
Figure2.6:Distributionsofthesquareoftheproton-proton(
s
pp
)(left)andproton-η
(s
pη
)(right) invariant masses determined experimentally for thepp → ppη
rea tion ( losed squares). The experimentaldataare omparedto al ulationsperformedassumingthelinearenergydependen eoftheprodu tionamplitudeasproposedbyA.Delo[13℄ depi tedbysolidlines.
dependen e inthe leading
3
P
0
→
1
S
0
s
partial wave amplitude [13℄. Those al ulations are in ontradi tion to the suggestion of Nakayama [12℄, giving eviden e that higher partialwaves playonly amarginal role.
Atthispoint,theobservedenhan ement ouldbeexplainedbythreedierent
hypothe-ses:
i)asigni ant role ofproton-
η
intera tioninthe nalstate, ii)an admixture of higherpartial waves oriii)anenergy dependen e of the produ tion amplitude.
Based on the
pp → ppη
data only,it is not possible to verifyany ofthose models. These ontingen ies motivated the work presented in this thesis whi h is an analysis of a highstatisti s
pp → ppη
′
rea tion measurement inordertodeterminethedistributionofevents
overthephasespa eforanex essenergyofQ=15.5MeVthesame oneassele tedbefore
forthe
pp → ppη
rea tion. The omparisonofthe dierential distributions forthe proton-protonandproton-meson invariant massesintheη
andη
′
produ tion ouldhelpto judge
between postulated explanations of the observed ee t and may allow for a quantitative
estimationof the proton-
η
and proton-η
′
intera tion.
Theexperimentalfa ility,themethodoftheanalysis,anda hievedresultsforthe
pp → ppη
′
The measurement of the
pp → ppη
′
rea tion was ondu ted using the ooler syn hrotron
COSYandthe COSY-11 dete torsetup. Both fa ilitieswill bedes ribedinthis hapter.
3.1 Cooler Syn hrotron COSY
TheCOolerSYn hrotron(COSY)[17℄ islo atedattheInstituteofNu learPhysi s ofthe
Resear hCentre Jüli hinGermany. Thefa ility wasdesigned toa elerate polarizedand
unpolarized proton and deuteron beams in the momentum range from 0.3 GeV/ up to
3.7 GeV/ . The sket h of the whole a elerator omplex is presented in gure 3.1. The
totallengthofthe syn hrotronringis184meter. Therearetwostraight40meter se tions,
andtwo bendingse tions with24dipole magnets.
Theexperimental installationsat thesyn hrotron anbe lassied astwogroups;
a)thedete torsinstalledinsidetheCOSYring: WASA[40,41℄,COSY-11[14 16℄,PISA[42℄,
EDDA [43℄,COSY-13 [44℄,and ANKE[45℄ and
b) outside of the COSY ring at external beam lines: COSY-TOF [46℄, JESSICA [47℄,
NESSI[48℄,GEM [49℄,MOMO[50℄, andHIRES [51℄.
Someof those experiments arealready ompleted and no longer in operation (labelled in
bla k)and the others arestill inoperation (labelled ingreen) ingure3.1.
TheCOSYsyn hrotron isequipped withele tronandsto hasti oolingdevi eswhi h
areusedto de reasethemomentumand spatial spreadof thebeam [52℄.
In the ase of ele tron ooling, theele trons with velo ities equal to the nominal proton
beam velo ity areinje ted at astraight se tionof the syn hrotron. Thisoperation auses
thatfaster protons arede elerated andslowerones area elerated.
Sto hasti ooling uses an ele tromagneti devi e, theso alledpi k-up unit, whi h
mea-sures the beam deviation from the nominal position at one point of the a elerator and
orre ts it by transmitting a orre tion signal through the shortest way to the ki ker
unit at theother side of the beam pipe. It auses not only a shift to the nominal beam
orbit, but also de reases the spread of the transversal and longitudinal momentum
Figure3.1: S hemati view of the oor plan of the COSY syn hrotron. Marked in violet are
internal-beam [14,40,4245℄ and external-beam [4651℄ dete tor setups. Fa ilities COSY-TOF,
WASAandANKElabelledingreenarepresentlystillinoperation. Thepositionsofele tronand
3.2 COSY-11 dete tor setup
Themeasurementofthe
pp → ppη
′
rea tionisbasedontheregistrationofthetwooutgoing
protonsandre onstru tionoftheir momenta. The
η
′
mesonisidentiedusingthemissing
masste hnique.
S2
Si
Figure3.2:S hemati viewoftheCOSY-11dete tionfa ility[14℄. Note,thatonlythosedete tors
whi h were used during the measurement of the
pp → ppX
rea tion are presented. Protons originating from thepp → ppX
rea tion are bent in the dipole magneti eld, and leave the va uum hamberthroughtheexitwindow. Afterwardstheyaredete tedinthetwodrift hambersD1andD2,inthes intillatorhodos opesS1andS2,andinthes intillatorwallS3. Thes intillation
The COSY-11 fa ility is one of the internal dete tor setups installed inside the COSY
syn hrotron tunnel at a bending se tion of the ring. It is mounted next to one of the
dipolemagnets, and benetsfrom thedipole magneti eldwhi h isused for theparti le
separationfrom the beam.
A s hemati view of the COSY-11 apparatus is presented in gure 3.2. The gure
illustratesalso s hemati allythe tra ks of protons outgoing fromthe
pp → ppX
rea tion. Twooutgoing protonspossessingsmallermomentathan thebeammomentum, arebent inthe dipole magneti eld towards the dete tor system. They leave the va uum hamber
throughout the exit window made out of a 30
µ
m layer of aluminum and 300µ
m of a arbon ber arriermaterial withan averaged densityof 2.1 g/ m3
[14℄ and are dete ted
usingthe drift hambersD1 andD2, the s intillator hodos opesS1 andS2, andthe
s in-tillationwall S3 1
.
The target 2
used during the experiment, was realized as a beam of
H
2
mole ules grouped inside lusters of up to about10
6
atoms. The average density of the target
wasaround
5 · 10
13
atoms/ m
2
[63℄. Itwasinstalledinfrontofthedipolemagnetasit an
be seens hemati allyin gure3.2.
Thedrift hambersD1andD2wereusedfor thedetermination oftheparti les
traje -tories. Those two planar drift hamber sta ks are spa ed by 70 m [14,64℄. Their a tive
areais1680mmwide and433mmhigh. Drift hamberD1(standing losertothebending
magnet) onsistsofsixdete tion planes. The rsttwo withverti al wires,two withwires
in lined by
+31
o
and two in lined by
−31
o
. The D2 drift hamber is built in the same
s heme, but itisextendedbytwo additional planeswith verti alwires.
Thewiresinadja entplanesofea hpairareshiftedbyhalfofthe ellwidthtoresolve
theleft-right position ambiguitywithrespe tto thesense wire. The hosen onguration
of thedete tion planes allows to perform the measurement of the horizontal and verti al
oordinates and enables aunique multi-hitevent identi ation[64℄.
A harged parti le rossingthedrift hambers produ es gasionization insidethedrift
ells, lled with a gas mixture of one to one argon and ethane at atmospheri pressure.
Theele trondrift time to the sensewireis ameasureof thedistan ebetween thepassing
parti le tra k and the sense wire (see se tion 4.1). Inthe ase of parti le tra ks oriented
perpendi ularto the dete tionplanes,themaximumdrifttime orrespondingto the
max-imumdrift path of20 mm equalsto 400 ns.
1
Inthemeasurementofthemesonprodu tioninthequasi-free
pd → pnp
spectator
X
rea tiondedi ated neutron[55 58℄and spe tator [59 61℄dete torswereinstalledinaddition. Theywerehowevernot usedforthemeasurementdes ribedinthisthesis.
2
Determined parti le traje tories inthe data analysisare tra ed through themagneti
eldofthedipoleba ktothetarget. Therefore,itispossibletore onstru tthemomentum
ve torsof outgoing parti lesat the rea tion point. The re onstru tion of themomentum
ve torsof the registered parti les ombined withthe information about thetime-of-ight
between the S1andthe S3dete tors allows forthe al ulation oftheparti lemassandby
thistheparti le type identi ation.
The S1 s intillating hodos ope is built out of sixteen identi al, verti ally arranged
modules, read out from both sides (top and bottom) by photomultipliers. The modules
with
45 × 10 × 0.4
m3
dimensionsarearrangedwithsmallverti aloverlap(
1
mm[14,65℄) inorder to avoid "not overed" spa e in the geometri al a eptan e. The S1 dete tor isusedasthe "start"for thetime-of-ight measurement.
The S2 s intillating hodos ope, similar asS1, onsistsof sixteen s intillation modules
withthedimension of
45 × 1.3 × 0.2
m3
[14℄.
TheS3s intillatingdete tordeliversthe "stop"informationfor thetime-of-ight
mea-surement. It is built from one non-segmented s intillating wall with the dimension of
220 × 100 × 5
m3
. Itisviewedbyamatrix of 217photomultipliers [14,66,67℄,o upying
theedgesof equilateral triangles withthesidesof11.5 m.
TheS4s intillation ountertogetherwiththesili onpaddete tor(depi tedingure3.2
asSi)areusedforregistrationofthere oilprotonsfromtheproton-protonelasti s attering
[14℄. The sili on pad dete tor [14℄ onsists of 144 pads with dimensions of
22.0 × 4.5 ×
0.28
mm3
. Ea h padis readout separately.
3.3 Trigger logi
In the experiment two independent trigger bran hes were used, in order to dete t the
pp → ppη
′
andpp → pp
rea tions. Themain triggerusedfor thepp → ppη
′
rea tionwasbasedon thefollowing onditions:
T
pp→ppη
′
= (S1
µ≥2
∨ S1
µ=1,high
∨ S2
µ≥2
∨ S2
µ=1,high
) ∧ S3
µ≥2
,
(3.1)where
µ
denotesthemultipli ityofsegmentsintheS1andS2s intillationhodos opes,and the number of photomultipliers whi h have red in the S3 dete tor. The subs ripthigh
stands for ahigh amplitude signalintheS1 and S2dete tors whi h was implemented fortriggeringevents whentwoparti les rossthesame segment [67℄. The hardwarethreshold
forthehighamplitude wassethighenough to redu ethenumberof singleparti leevents
onsiderably, and low enough to a ept most events (almost 100
%
) with two protons passingthroughone segment ofS1 [67℄.between signals fromthe S1 and S4dete tors:
T
pp→pp
= S1
µ=1
∧ S4,
(3.2)wheretheS1hodos opewasusedforthe registrationofforwards atteredprotons andthe
S4s intillationdete tor wasregistering re oilprotons.
The dete tors were positioned to over a large part of the kinemati s of the
pp → pp
elasti s attering. Due to the high rate of thepp → pp
rea tion only every 128'th event was registered for the further analysis. The number of the elasti s attering events wasIn this hapter the method used to alibrate the COSY-11 dete tors and their relative
settingswill bepresented. Inparti ular,the time-spa erelationof thedrift hambersand
thepro edureof time-of-ight alibration willbedes ribed. Inaddition, thepro edureof
monitoring the relative beam-targetsetting will be dis ussed.
4.1 Spa e-time relation for drift hambers
Thedrift hambersD1andD2 onsistof6and8planesofwires,respe tively. Theyprovide
theinformation about thedrift time of ele trons (tothe sense wires) produ ed along the
traje toryof hargedparti lespassingthroughthe hambers. Inordertore onstru tthose
traje tories one needs to establisharelation between drift timeanddistan e between the
parti letra kand thesensewire (Fig. 4.1(left)).
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
0
100 200 300 400 500 600 700 800
drift time + offset
[
ns
]
distance from the sense wire
[
cm
]
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
0
100 200 300 400 500 600 700 800
drift time + offset
[
ns
]
∆
X
MEAN
[
cm
]
Figure4.1: (Left panel) Time-spa e alibration - relation betweenthe distan e from the sense
wireand the drifttime. (Rightpanel) Corre tions of thetime-spa e relation, where themiddle
histogramrepresentsthemeanvalueof
∆X
,theupperandlowerhistogramsvisualizeonestandard deviation(σ
)ofthe∆X
distribution. Seethetextfortheexplanationof∆X
.Due to the sensitivity of the drift velo ity to the atmospheri pressure, humidity and
gas mixture hanges [68℄, the data used for the alibration pro edure were divided into
time intervals of about 3-8 hours, with a similar number of olle ted events. The
al-ibration fun tion was determined for ea h interval applying the pro edure of iterative
improvements[67℄. Startingwithanapproximatedfun tionofthespa e-timerelation, the
straightlinetotheobtainedpointswasttedandfurtheron,assumingthatit orresponds
totherealparti letra k,thedeviation
∆X
,betweenthemeasureddistan eoftheparti le tra k from the sense wire and the one from the tting pro edure was al ulated. Next,having a ertain amount of data, one oulddetermine amean value of
∆X
asa fun tion of the drift time (presented in gure 4.1 (right)). The∆X
M EAN
was subsequently used fora orre tion ofthe time-spa e relation. Next,theimproved fun tionhasbeen usedforthetra k re onstru tion similarly as intherst step. The whole pro edure wasrepeated
until the orre tionsweresmallerthan thestatisti alun ertainty ofthe
∆X
M EAN
. Theaveragedspatialresolutionofthedrift hambersa hieved intheexperimentdis ussedinthis thesis amountedto 250
µ
m(rms).4.2 Time-of-ight alibration of s intillator dete tors
Thes intillatordete torsS1andS3"start"and"stop",respe tively,areusedforthe
time-of-ight measurement. S1 onsists of 16 s intillator plates with photomultiplier readout
from both sides, and S3 is a s intillator wall read out by a 217 photomultipliers matrix.
Inorder toobtaintheproperinformation about thetime-of-ightbetween both dete tors,
one needs to determine time "osets" for all photomultipliers i.e. therelative dieren es
intransition time ofthe signal fromthephotomultiplier to theTDC unit.
Letusdenote
t
S1
asthe real time whenaparti le rosses theS1 dete torandt
S3
when it rossesS3. Then, thetime-of-ight an be al ulated asfollows:tof
S1−S3
= t
S3
− t
S1
. ThemeasuredTDC valuesfor a singlephotomultiplier inS1 andS3 dete tors read:T DC
S1
= t
S1
+ t
y
+ t
S1
walk
(P M ) + t
S1
of f set
(P M ) − t
trigger
,
(4.1)T DC
S3
= t
S3
+ t
pos
+ t
S3
walk
(P M ) + t
S3
of f set
(P M ) − t
trigger
.
(4.2) Inboth equations thetimestampt
trigger
(denotingthetimeofthetriggersignal) is iden-ti al. The indexy
orresponds to the distan e between hit position and the edge of the s intillator lose to the given photomultiplier in the S1 dete tor andpos
stands for the distan e between the hit position and the photomultiplier in the S3 dete tor. Theab-breviation
t
walk
denotes the orre tions for thetime − walk
ee t, i.e. the signal time dependen e onthe signalamplitude [69℄. Anydependen eoft
y
is an elledbytakingthe averagebetweenthetimesmeasuredbytheupperandlowerphotomultipliers[67℄andthisan be al ulated from theknowntraje tories. Thus, theonly unknown variablesare the
time osets
t
of f set
for both dete tors. For a rst approximation, the time dieren e in the S1 dete tor an be a hieved by taking into a ount signals from the parti lesross-ing the overlapping parts of the modules. Next, for the S3 dete tor the time oset an
(
tof
rec
). Then, iteratively, using the obtained S3 oset one an determine the time o-sets for the S1 dete tor. After two iterations, the time osets for both dete tors an beestablished. Asan example,the distribution
∆t(P M
S3
ID
)
determined as∆t(P M
S3
ID
) = tof
S1−S3
(P M
ID
S3
) − tof
rec
fora groupof photomultipliers of theS3 dete torarepresentedingure4.2.
-4
-3
-2
-1
0
1
2
3
4
75
80
85
90
95 100 105 110
photomultiplier ID
∆
t(PM
S3
ID
)
[
ns
]
Figure4.2: Distribution of thetime dieren e betweenthe time
tof
rec
al ulated from the re- onstru tedparti lemomentum andthemeasured timetof
S1−S3
betweentheS1 dete torand a parti ular photomultiplier in the S3 dete tor, as afun tion of the photomultiplier ID in the S3dete tor. As a hieved afterse ond iteration, theguredepi tsonly afra tion ofPM's ofthe S3
dete tor(75-112).
ThetimeosetsforthephotomultipliersintheS3dete torareobtainedonthebasisofthe
timedieren es between
tof
S1−S3
andtof
rec
presentedin gure4.2. Theywere adjusted su hthat thisdieren e isequal to zero.4.3 Monitoring of relative beam-target settings
Possible hangesofthepositionwherethebeam rossesthetarget ouldhavesigni antly
inuen edthe momentumre onstru tion and asa onsequen e ouldworsen the
determi-nation of the mass of the undete ted parti le. Therefore, it is important to monitor the
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0
0.2 0.4 0.6 0.8
1
1.2 1.4 1.6
p
II
[
GeV/c
]
p
⊥
[
GeV/c
]
Figure4.3:Thedistributionoftheperpendi ular
p
⊥
versusparallelp
k
momentum omponentsforpp → pp
elasti s atteringeventsatabeammomentumof3.260GeV/ . Thesolidline orresponds to thekinemati al ellipse. Changes of theevent density along thekinemati al ellipse ree t theangulardependen e ofthe rossse tionforthe
pp → pp
rea tion.The mean value of the distan e between the expe ted kinemati al ellipse and the
exper-imental points (shown in gure 4.3) may be used as a measure for the deviation of the
enterofthe intera tionregion fromits nominalposition(
∆center
). Api torial denition of∆center
is presented in gure 4.4 and the beam-target geometri al onditions are de-pi ted ingure4.5. By assuming a wrong intera tion enter there onstru tion results ina wrong momentum determination and the
pp → pp
events are not entered around the expe tedkinemati al ellipse.Figure4.4: Pi torial denition of the deviation of the enter of the intera tion region from its
Figure4.5:S hemati viewoftherelativetargetandbeamsettings. Leftpaneldepi tstheview
from above, right presents aside view.
σ
X
andσ
Y
denote the horizontal and verti al standard deviationsof theassumed Gaussiandistributions of theproton beamdensity, respe tively.∆
X
bt
denotes thedistan e betweenthe enters of the proton target and beam. The gureis adapted
from[70℄.
Inthe left panel ofgure4.6 themean distan eof experimental
pp → pp
eventsfrom the expe tedkinemati alellipseisshownasafun tionof∆center
assumedintheanalysis. As anbe seen,the enter ofthe intera tion region diersby0.45 m fromthenominalone.-0.3
-0.2
-0.1
0
0.1
0.2
0.3
-0.4 -0.2
0
0.2
0.4
0.6
0.8
∆
center
[
cm
]
distance from the ellipse
[
GeV/c
]
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0
10000
20000
measurement time
[
minutes
]
distance from the ellipse
[
GeV/c
]
Figure4.6:(Leftpanel)Thedistan ebetweentheexpe tedellipseandthe enterofexperimental
distributionon the(
p
⊥
,p
k
) plot(gure 4.3)versusthe deviation of the enter of theintera tion regionfrom itsnominalposition(∆center
). (Rightpanel)Thedeviationofthedistan efrom the kinemati alellipseasafun tion ofthetimeofthemeasurement. Themeanvalueofthedistan efromtheellipsehasbeenplottedfor13hours intervals. Intheanalysisthevalueof
∆center
was setto0.45 m.statisti al error in the determination of the mean value of the distan e from the ellipse.
Thevariations areatalevelof
10
−3
and,as an be inferredfromtheplotpresentedinthe
leftpanelofgure4.6, orrespondtoshiftsoftheintera tion enterbylessthan 0.01mm.
Thus, thevariations ofthe enter of the intera tion region an be safelynegle ted in the
5. Identi ation of the
pp
→ ppη
′
rea tion
Inthefollowing hapterthemethodofidentifyingthe
pp → ppη
′
rea tionwillbedes ribed.
5.1 Identi ation of protons
Themeasurement was basedon the registration oftwo outgoing protons originating from
the
pp → ppX
rea tion.10
10
2
10
3
10
4
10
5
0
1
2
3
4
(invariant mass)
2
[
GeV
2
/c
4
]
events
pions
protons
deuterons
Figure5.1: Distribution of the squared invariant mass of the registered parti les. (Note the
logarithmi s aleonverti alaxis.) Signalsfrommeasuredpions,protons,anddeuteronsareeasily
distinguished.
After hoosing only two-tra k events, theprotons were identied by thedetermination of
theirrest masses. The parti lemasswas al ulateda ording tothe following formula:
m
2
=
p
~
2
(1 − β
2
)
β
2
,
(5.1)where
~
p
andβ
are denoting the momenta and velo ities of parti les, respe tively, whi h were determined in an independent way (p
~
from the urvature of the traje tory in thedipoleand
β
fromthetime of ight between S1 and S3). The distribution of thesquared massesoftheparti lesisshowningure5.1. Clearlyvisiblearesignalsfrompions,protonsanddeuterons.
For the further analysis, parti les with re onstru ted masses in the range from 0.2 to
1.5GeV
2
/
4
wereassumedto beprotons.
5.2 Identi ation of the
η
′
meson
Inthepresentexperimentthede ayprodu tsofthe
η
′
mesonwerenotmeasured, therefore
itwasimpossibletoidentifyitsprodu tiononanevent-by-eventbasis. Eveninexperiments
dete tingallde ayprodu tsanunambiguousidenti ationofasingle"
η
′
produ tionevent"
isnot possible, but theba kground would be mu h smaller.
0
2000
4000
6000
0.95
0.955
0.96
0.965
missing mass
[
GeV/c
2
]
events / 0.25 MeV/c
2
Figure5.2:Missingmassspe trumforthe
pp → ppX
rea tionmeasuredatabeammomentum ofP
B
=
3.260 GeV/ . The peakoriginating from thepp → ppη
′
rea tionis learly seenon top
ofamulti-pionprodu tionba kground. Thedashedline orrespondsto at withase ondorder
polynomialtothedataoutsidethesignalfrom the
η
′
meson.
Thenumber of
pp → ppη
′
events wasdetermined using themissingmasste hnique. This
method is based on the knowledge of the protons four-momenta before and after the
Identi ationof the
pp → ppη
′
rea tion 37
four-momenta of the proton beam, proton target, and rst and se ond outgoing proton,
respe tively, one an usethe following formula, inthe ase of the
pp → ppX
rea tion, to al ulatethe massm
X
of the unregisteredparti le:m
2
X
= E
X
2
− ~p
2
X
= (P
b
+P
t
−P
1
−P
2
)
2
=
= (E
b
+ E
t
− E
1
− E
2
)
2
− (~p
b
+ ~
p
t
− ~p
1
− ~p
2
)
2
.
(5.2) In gure 5.2 the missing mass spe trum determined experimentally for thepp → ppX
rea tionfor the whole datasampleis presented. The spe trumin ludes abroaddistribu-tionfrommulti-mesonprodu tionandthewelldenedpeakoriginatingfromthe
η
′
meson
produ tion.
Thesmoothbehaviouroftheexperimentalmulti-pionprodu tionba kground,whi h ould
be veried byMonte Carlosimulations studies (see se tion9.2), allows for asimple
poly-nomial t. The knowledge of the smooth behaviour of the ross se tion [60,71℄, assures
thatintherangeof the signal, themulti-pion ba kground should be at.
The dashed line in gure 5.2 orresponds to a se ond order polynomial tted to the
ex-perimental ba kground. Indeed, it an be seen that the t reprodu es the shape of the
ba kground satisfa torywell.
0
500
1000
1500
0.95
0.955
0.96
0.965
missing mass
[
GeV/c
2
]
events / 0.25 MeV/c
2
Ingure5.3 the experimental missingmassspe trumafter theba kground subtra tion is
presented. Theba kgroundwasapproximated byase ondorderpolynomialtasdepi ted
ingure5.2byadashedline. Thetotalnumberofregisteredandre onstru ted
pp → ppη
′
rea tionsamountsto about 15000.
Here, the statisti sa hieved in the measurement is only illustrated and thepossibilityof
the ba kground determination is shown. A detailed dis ussion of the subtra tion of the
multi-pion produ tion ba kground for dierential ross se tions will be omprehensively
Inorder to determine theabsolute values ofthe dierential rossse tions, theluminosity
(
L
) integrated over the measurement time has to be established. For that purpose, the analysis of thepp → pp
rea tion, in order to establish the number of elasti s attering eventswasperformed.A s hemati view of theCOSY-11 dete torsetup with superimposed tra ks of elasti ally
s atteredprotons isshowningure6.1.
Si
123....
...48
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
-8
0
-6
0
-4
0
-2
0
0
20
40
60
80
x
S1
[cm]
Figure6.1:S hemati viewoftheCOSY-11apparatuspresentingthedete torsusedforthe
reg-istrationofthe
pp → pp
elasti s attering. Thesuperimposedlinesshowanexampleoftraje tories from elasti ally s attered protons with the laboratory anglesθ
1
andθ
2
. One an ompare this guretogure3.2. Thex
S1
axisindi atesthesizeofdete tormodules(forfurtherdes riptionsee text).One an evaluate theluminosity(
L
)a ording tothe formula:∆N (θ
2
∗
)
∆Ω
∗
(θ
∗
2
)
=
dσ
∗
dΩ
∗
(θ
∗
2
) · L,
(6.1) wheredσ
∗
dΩ
∗
(θ
∗
2
)
denotes the knowndierential rossse tion [72℄ and∆N (θ
∗
2
)
indi ates the number of elasti ally s attered protons at a solid angle∆Ω
∗
around theproton emission
angle
θ
∗
2
in the entre-of-mass system. In thefurther analysis, the available range of theθ
2
∗
angle(44
o
to
66
o
)wasdividedinto 11binswithawidth of
2
o
. Thetra ksofelasti ally
s atteredprotonsresultinginsignalsinthe S1dete torwitha oin ident signalinS4from
the se ond proton overs the horizontal axis of the S1 dete tor, marked in gure 6.1 as
x
S1
,from 40 m to 75 m, whi h orrespondsto aθ
∗
2
anglerange from44
o
to66
o
.0
5000
10000
15000
20000
25000
30000
-0.4
-0.2
0
0.2
0.4
distance from ellipse
[
GeV/c
]
48
o
<
Θ
*
2
<
50
o
events
Figure6.2:Proje tionoftheeventdistributionalongthekinemati alellipsefora entre-of-mass
protons attering angle of
θ
∗
2
in therange from48
o
to
50
o
, orresponding to therangein the S1
dete torfrom
x
S1
= 67.0 cm
tox
S1
= 70.0 cm
.As an example, the distribution of elasti ally s attered protons at the
θ
∗
2
angular range from48
o
to50
o
is presented in gure 6.2. This distribution shows the proje tion of the
experimental data along the kinemati al ellipse. The number of events (redu ed by the
ba kground indi ated bythedashed line)isusedfor the al ulation of theluminosity.
Thesignalfrom elasti allys attered protons an be learly separatedfromthe at
multi-pions attering ba kground. Thesolidangle
∆Ω
∗
usingtheMonte-Carlomethod,asfollows:
∆Ω
∗
=
4π N
accepted
2 N
0
[sr],
(6.2)where
N
0
stands for the number of proton-proton elasti s attering events in the orre-sponding angular range andN
accepted
onstitutes the number of events inthe onsidered binof theθ
∗
2
angle, whi h ould be registeredandidentied. In parti ular,theanalysisin thefollowing manner wasdone. First,N
0
= 2 · 10
7
events hasbeen generated, al ulating
theresponse of the COSY-11 dete tors, and then those events have been analysed using
pro eduresappliedfor theexperimentaldataevaluationinorderto determinethenumber
of
N
accepted
eventsfor ea hθ
∗
2
angleinterval.0
0.2
0.4
0.6
0.8
1
1.2
1.4
40
60
80
Θ
2
∗
[
degrees
]
d
σ
/d
Ω
[
mb/sr
]
p
beam
= 3262.5
[
MeV/c
]
EDDA
COSY-11
Figure6.3:Dierential rossse tionfor theproton-protonelasti s attering. Theresult ofthis
thesis( losed ir les)measuredatabeammomentumof
P
B
= 3.260
GeV/ wass aledinamplitude tothe rossse tionmeasuredbytheEDDA ollaborationshownbyopensquares[72℄.Figure 6.3 indi ates the angular distribution of the dierential ross se tion for elasti
proton-proton s attering obtained in the experiment ( losed ir les). The amplitude of
thatdistribution was ttedto the dataof the EDDA experiment in luding only one free
parameter being the integrated luminosity (see eq. 6.1). The extra ted integrated
lumi-nosityfortheexperimentdes ribedinthisthesisamountsto
L
= (5.859 ± 0.055) pb
−1
The knowledge of the luminosity value will allow for the overall normalization of the
deriveddierential rossse tionasafun tion ofthe
s
pp
ands
pη
′
invariantmasses,angular distributions andtotal rossse tion whi h will be dis ussed in hapter 10.beam momentum
Inordertoperformrealisti simulationsofthestudiedrea tions,inparti ulartodetermine
thea eptan eandto al ulatethe ovarian ematrix,itismandatorytoknowtheabsolute
valueandthespreadofthebeammomentum. Thedis ussedmeasurementofthe
pp → ppη
′
rea tion was nominally performed at the same value of ex ess energy Q as the
pp → pp
rea tion measurement with Q =15.5 MeV whi h orrespondsto a nominal proton beammomentum of
P
B
= 3.257
GeV/ .The pre ision of the absolute beam momentum adjustment of the COSY syn hrotron is
about
10
−3
[73℄ whi hinthis ase orrespondsto
∼
3 MeV/ .The beam momentum dependen e of the mean value of the missing mass distribution
presented in gure 5.3, was studied in order to determine the a tual value of the beam
momentum more a urate.
Thebeammomentumwas al ulatedusing theformula:
m
X
=
√
s − 2m
p
=
2m
2
p
+ 2m
p
q
P
2
B
+ m
2
p
1/2
− 2m
p
,
(7.1) where√
s
denotes the total energy inthe entre-of-mass frame,P
B
stands for theproton beam momentum, andm
p
orrespondsto the proton mass.The beam momentum of
P
B
= 3.260
GeV/ has been determined by adjusting theP
B
su h that the mean value of the missing mass peak is equal to theη
′
meson mass. The
determined value of the beam momentum diers by 0.003 GeV/ from the nominal one.
This deviation is in agreement with results of analogous analysis performed in previous
measurements[67℄.
Thedeterminedvalueoftheex essenergyamountsto(Q=16.39
±
0.01±
0.4)MeV, where the errors indi ate statisti al and systemati un ertainty, respe tively. Thedomi-nating systemati un ertainty was established in [9℄ and the statisti un ertainty of the
ex essenergy wasdetermined using the following formula:
∆Q =
s
dQ
dP
B
2
· (∆P
B
)
2
,
(7.2)0
0.002
0.004
0.006
0.008
0.01
-2.5
0
2.5
∆
P
B
[
MeV/c
]
intensity a. u.
Figure7.1:Spe trumofthebeammomentumdistributionintegratedoverthewholemeasurement
time. Thevalueof
∆P
B
= 0
orrespondstoabeammomentum of3.260GeV/ . Thedashedlines markthebeammomentumdispersionfortheextended9millimetertargetusedinthisexperiment.Afterthedetermination oftherealvalueoftheabsolutebeammomentum,nowitsspread
willbedetermined. One an al ulate thebeammomentumspe trumfromthefrequen y
spe trumof the COSYbeam (S hotky spe trummeasured during the experiment) using
thebelow formula[74℄:
∆f
f
= η
B
·
∆P
B
P
B
,
(7.3)where
f
andP
B
denote beam frequen y and beam momentum, respe tively,η
B
is a pa-rameterwhi hdependsonthe beamopti s,i.e. theele tri andmagneti eldsinthesyn- hrotron. Duringthe experiment,the
η
B
parameter wasestablishedto beη
B
=0.12 [75℄. The spe trum of the beam momentum obtained during the experiment is shown ing-ure7.1.
Thedispersionofthe beammomentumdependsonthemagneti eldalongthering.
Pro-tons on the outer routes have a "longer way" than those on the inner side of thebeam,
and the traje tories are dierent from the nominal value. When at a ertain point the
parti lepositiondeviatesby