Measurement of the np -> np pi(0)pi(0) reaction in search for the recently observed d* (2380) resonance

Contents lists available atScienceDirect

Physics

Letters

B

www.elsevier.com/locate/physletb

Measurement

of

the

np

→

np

π

0

π

0

reaction

in

search

for

the

recently

observed

d

∗

(

2380

)

resonance

WASA-at-COSY

Collaboration

P. Adlarson

a

,

1

,

W. Augustyniak

b

,

W. Bardan

c

,

M. Bashkanov

d

,

e

,

F.S. Bergmann

f

,

M. Berłowski

g

,

H. Bhatt

h

,

A. Bondar

i

,

j

,

M. Büscher

k

,

l

,

2

,

3

,

H. Calén

a

,

I. Ciepał

c

,

H. Clement

d

,

e

,∗

,

D. Coderre

k

,

l

,

m

,

4

,

E. Czerwi ´nski

c

,

K. Demmich

f

,

E. Doroshkevich

d

,

e

,

R. Engels

k

,

l

,

A. Erven

n

,

l

,

W. Erven

n

,

l

,

W. Eyrich

o

,

P. Fedorets

k

,

l

,

p

,

K. Föhl

q

,

K. Fransson

a

,

F. Goldenbaum

k

,

l

,

P. Goslawski

f

,

A. Goswami

k

,

l

,

r

,

K. Grigoryev

k

,

l

,

s

,

5

,

C.-O. Gullström

a

,

F. Hauenstein

o

,

L. Heijkenskjöld

a

,

V. Hejny

k

,

l

,

B. Höistad

a

,

N. Hüsken

f

,

L. Jarczyk

c

,

T. Johansson

a

,

B. Kamys

c

,

G. Kemmerling

n

,

l

_,

_{F.A. Khan}

k

,

l

_,

_{A. Khoukaz}

f

_,

_{D.A. Kirillov}

u

_,

S. Kistryn

c

,

H. Kleines

n

,

l

,

B. Kłos

v

,

W. Krzemie ´n

c

,

P. Kulessa

w

,

A. Kup´s ´c

a

,

g

,

A. Kuzmin

i

,

j

,

K. Lalwani

h

,

6

,

D. Lersch

k

,

l

,

B. Lorentz

k

,

l

,

A. Magiera

c

,

R. Maier

k

,

l

,

P. Marciniewski

a

,

B. Maria ´nski

b

,

M. Mikirtychiants

k

,

l

,

m

,

s

,

H.-P. Morsch

b

,

P. Moskal

c

,

H. Ohm

k

,

l

,

I. Ozerianska

c

,

E. Perez del Rio

d

,

e

,

N.M. Piskunov

u

,

P. Podkopał

c

,

D. Prasuhn

k

,

l

,

A. Pricking

d

,

e

,

D. Pszczel

a

,

g

,

K. Pysz

w

,

A. Pyszniak

a

,

c

,

J. Ritman

k

,

l

,

m

,

A. Roy

r

,

Z. Rudy

c

,

S. Sawant

k

,

l

,

h

,

S. Schadmand

k

,

l

,

T. Sefzick

k

,

l

,

V. Serdyuk

k

,

l

,

x

,

B. Shwartz

i

,

j

,

R. Siudak

w

,

T. Skorodko

d

,

e

,

y

_,

_{M. Skurzok}

c

_,

_{J. Smyrski}

c

_,

_{V. Sopov}

p

_,

_{R. Stassen}

k

,

l

_,

_{J. Stepaniak}

g

_,

E. Stephan

v

,

G. Sterzenbach

k

,

l

,

H. Stockhorst

k

,

l

,

H. Ströher

k

,

l

,

A. Szczurek

w

,

A. Täschner

f

,

A. Trzci ´nski

b

,

R. Varma

h

,

G.J. Wagner

d

,

M. Wolke

a

,

A. Wro ´nska

c

,

P. Wüstner

n

,

l

,

P. Wurm

k

,

l

,

A. Yamamoto

z

,

J. Zabierowski

aa

,

M.J. Zieli ´nski

c

,

A. Zink

o

,

J. Złoma ´nczuk

a

,

P. ˙Zupra ´nski

b

,

M. ˙Zurek

k

,

l

a_{DivisionofNuclearPhysics,DepartmentofPhysicsandAstronomy,UppsalaUniversity,Box516,75120Uppsala,Sweden} b_{DepartmentofNuclearPhysics,NationalCentreforNuclearResearch,ul.Hoza 69,00-681,Warsaw,Poland}

c_{InstituteofPhysics,JagiellonianUniversity,ul.Reymonta4,30-059Kraków,Poland}

d_{PhysikalischesInstitut,Eberhard-Karls-UniversitätTübingen,AufderMorgenstelle14,72076Tübingen,Germany}

e_{KeplerCenterforAstroandParticlePhysics,EberhardKarlsUniversityTübingen,AufderMorgenstelle14,72076Tübingen,Germany} f_{InstitutfürKernphysik,WestfälischeWilhelms-UniversitätMünster,Wilhelm-Klemm-Str.9,48149Münster,Germany}

g_{HighEnergyPhysicsDepartment,NationalCentreforNuclearResearch,ul.Hoza69,00-681,Warsaw,Poland} h_{DepartmentofPhysics,IndianInstituteofTechnologyBombay,Powai,Mumbai-400076,Maharashtra,India} i_{BudkerInstituteofNuclearPhysicsofSBRAS,11akademikaLavrentievaprospect,Novosibirsk,630090,Russia} j_{NovosibirskStateUniversity,2PirogovaStr.,Novosibirsk,630090,Russia}

k_{InstitutfürKernphysik,ForschungszentrumJülich,52425Jülich,Germany} l_{JülichCenterforHadronPhysics,ForschungszentrumJülich,52425Jülich,Germany}

m_{InstitutfürExperimentalphysikI,Ruhr-UniversitätBochum,Universitätsstr.150,44780Bochum,Germany} n_{ZentralinstitutfürEngineering,ElektronikundAnalytik,ForschungszentrumJülich,52425Jülich,Germany}

o_{PhysikalischesInstitut,Friedrich-Alexander-UniversitätErlangen-Nürnberg,Erwin-Rommel-Str.1,91058Erlangen,Germany}

*

Correspondingauthor.

E-mailaddress:heinz.clement@uni-tuebingen.de(H. Clement).

1 _{Presentaddress:InstitutfürKernphysik,JohannesGutenberg-UniversitätMainz,Johann-Joachim-BecherWeg 45,55128Mainz,Germany.} 2 _{Presentaddress:PeterGrünbergInstitut,PGI-6ElektronischeEigenschaften,ForschungszentrumJülich,52425Jülich,Germany.}

3 _{Presentaddress:InstitutfürLaser- undPlasmaphysik,Heinrich-HeineUniversitätDüsseldorf,Universitätsstr. 1,40225Düsseldorf,Germany.} 4 _{Presentaddress:AlbertEinsteinCenterforFundamentalPhysics,UniversitätBern,Sidlerstrasse 5,3012Bern,Switzerland.}

5 _{Presentaddress:III. PhysikalischesInstitut B,Physikzentrum,RWTHAachen,52056Aachen,Germany.} 6 _{Presentaddress:DepartmentofPhysicsandAstrophysics,UniversityofDelhi,Delhi-110007,India.}

http://dx.doi.org/10.1016/j.physletb.2015.02.067

0370-2693/©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3_.

Articlehistory:

Received9September2014

Receivedinrevisedform29January2015 Accepted27February2015

Availableonline3March2015 Editor: V.Metag

Keywords:

Two-pionproduction

ABCeffectandresonancestructure Dibaryonresonance

Exclusivemeasurementsofthequasi-freenp_→np

π

0

_π

0_{reactionhavebeenperformedbymeansofdp} collisionsatTd=2.27 GeV usingtheWASAdetectorsetupatCOSY.Totalanddifferentialcrosssections have been obtained covering the energy region √s= (2.35–2.46) GeV, which includes the regionof the ABC effectand itsassociated d∗(2380)resonance. Addingthe d∗ resonance amplitudetothat for theconventionalprocessesleadstoareasonabledescriptionofthedata.Theobservedresonanceeffect inthetotalcrosssectionisinagreementwiththepredictionsofFäldtand Wilkinas wellwiththose of Albadajedo and Oset.The ABC effect, i.e. the low-massenhancement in the

π

0

_π

0_{-invariant mass} spectrum, isfoundto beverymodest–ifpresentatall,whichmightposeaproblemtosomeofits interpretations.

©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense

(http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3_.

1. Introduction

Recentdata onthe basicdouble-pionicfusion reactions pn

→

d

π

0

_π

0 _{and pn}

_→

_d

_π

+

_π

−_{demonstratethattheso-calledABC}

ef-fectistightlycorrelatedwithanarrowresonancestructureinthe totalcrosssectionofthesereactions

[1–3]

.TheABCeffectdenoting ahugelow-massenhancementinthe

π π

invariantmassspectrum is observed to occur, if the initial nucleons or light nuclei fuse toa bound final nuclearsystemand iftheproduced pionpairis isoscalar.TheeffecthasbeennamedaftertheinitialsofAbashian, BoothandCrowe,who firstobserveditinthe inclusive measure-mentofthepd

→

3_{HeX reactionmorethanfiftyyearsago}

_[4]

_.

TheresonancestructurewithI

(

JP

)

=

0

(

3+

)

[1]observedinthe

pn

→

d

π π

totalcrosssectionat

√

s

≈

2

.

38 GeV issituatedabout 80MeV below

√

s

=

2m, thepeak position ofthe conventional t-channel



process, which is also observed in this reaction.

The resonance structure has a width of only 70 MeV, which is

about three times narrower than the conventional process. Nev-ertheless, from the Dalitz plotof the pn

→

d

π

0

_π

0 _{reaction it is}

concludedthat thisresonance decaysviatheintermediate



+



0

system(atleastpredominantly) intoits final d

π

0

_π

0 _state.Inthe

pn

→

pp

π

0

_π

− _{reaction the resonance has been sensed, too [5],}

though in this case, there is no ABC effect associated with the resonance. In consequence it has no longer be called ABC reso-nance, butd∗ – adopting the notation of thepredicted so-called “inevitabledibaryon”

[6]

withidenticalquantumnumbers.

Bysubsequentquasifreepolarized



np scatteringmeasurements, it has been demonstrated that there is a resonance pole in the coupled3D3–3G3partialwavescorrespondingtothed∗ resonance

structureinmass,widthandquantumnumbers

[7,8]

–supporting thusitss-channelcharacter.

Ifthescenarioofas-channelresonanceinthenp systemis cor-rect, then also thenp

→

np

π

0

_π

0 _{reaction should be affected by}

thisresonance, since thischannel may proceed via the same in-termediate



0



+systemasthenp

→

d

π

0

_π

0 _andpn

_→

_pp

_π

0

_π

−

reactions do.From a simple isospinpoint ofview we expectthe resonance effect in the np

π

0

_π

0 _{system to be identical in size}

to that in the d

π

0

_π

0 _{system. From more refined estimates in}

Refs. [9,10],whichaccountalsofordifferencesinphasespace, we expect the resonance effect in the np

π

0

_π

0 _{channel to be about}

85% ofthat inthed

π

0

_π

0 _{system. Sincethepeakresonancecross}

section inthelatteris270 μb

[3]

sittinguponbackgrounddueto conventionalt-channelRoperand



excitations,weestimatethe peak resonance contribution in thenp

π

0

_π

0 _{systemto be in the}

orderof200 μb.

2. Experiment

Since there exist no data at all for the np

→

np

π

0

_π

0

chan-nel, we have investigated this reaction experimentally with the

WASA detector at COSY (FZ Jülich) by using a deuteron beam

with an energy of Td

=

2

.

27 GeV impinging on a hydrogen

pel-let target [11,12]. By exploiting the quasi-free scattering process

dp

→

np

π

0

_π

0

₊

_p

spectator, we coverthe full energy range of the

conjectured resonance. In addition, the quasi-free process in in-verse kinematics gives usthe opportunity to detect alsothe fast spectatorprotonintheforwarddetectorofWASA.

The hardware triggerutilizedin thisanalysisrequiredatleast two chargedhits inthe forward detector aswell as two neutral hitsinthecentraldetector.

The quasi-free reactiondp

→

np

π

0

_π

0

₊

_p

spectator hasbeen

se-lectedintheofflineanalysisbyrequiringtwoprotontracksinthe forwarddetectoraswellasfourphotonhitsinthecentral detec-tor, which can be traced back to the decay of two

π

0 _particles.

Thatway,thenon-measuredneutronfour-momentumcouldbe

re-constructed by a kinematicfit withthreeover-constraints, which

derive fromthe conditions forenergy andmomentum

conserva-tion andthe

π

0 _{mass. The achievedresolution in}

√

_{s was about}

20MeV.

Forthe reconstructionofthe two

π

0 _{particles out ofthefour}

γ

quanta,allcombinationshavebeenconsideredandtheoptimal

combination has been chosen, where both of the reconstructed

γ γ

-invariant masses Mγ γ are closest to the nominal

π

0 _mass.

For all selected events this leads to a narrow peak in the two-dimensional plotof Mγ γ versus Mγ γ ,see,e.g. Fig. 2 inRef.[13]

Fig. 1. Plotofthe energylossElayer 4ofparticlesinlayer4ofthesegmented RangeHodoscope versus thatin layer 5(Elayer 5).The bands ofstopped and punch-throughprotonsanddeuteronsareindicated.

andFig. 3inRef.[14].Withthisprocedurethecombinatorial back-groundisverysmall,intheorderofafewpercent.

ThechargedparticlesregisteredinthesegmentedForward De-tector of WASAare identified by useof the



E

−

E energy loss method.Forits applicationin thedata analysis, all combinations ofsignalsstemmingfromthefivelayersoftheForwardRange Ho-doscope are used. As an example, Fig. 1 shows the plot of the energy loss in layer 4 versus that in layer 5. As can be seen, deuteronsandprotonscanbewellseparatedingeneral.

Adifficultyemerges fromdeuterons,which originatefromthe

np

→

d

π

0

_π

0 _{reactionandwhichpartlyalsobreakupwhile}

pass-ing the detector. Since in the energyloss plots used for particle identificationprotonanddeuteronbandsdohavesome smallbut finiteoverlaps,deuteronscannot beseparatedcompletelyfromnp

pairsstemmingfromthenp

→

np

π

0

_π

0 _{reaction.Tosuppresssuch}

misidentified eventswe requirethe angle betweenemitted neu-tronandprotontobelargerthanfivedegreesandalsotheir ener-giestobeintheexpectedrange.Nevertheless,aMonteCarlo(MC) simulationof thenp

→

d

π

0

_π

0 _{reaction,which is known}

experi-mentallyand alsocan be modeled very well [1], showsthat we havetoexpect stillacontamination ofabout5% inthespectraof

Fig. 2. Efficiencycorrecteddistribution ofthe spectatorproton momentain the

dp→npπ0_π0_+p

spectatorreactionwithintheWASAacceptance,whichallowsthe detectionofthespectatorprotononlyforlabangleslargerthanthreedegrees.In addition,theconstraintforthesuppressionofbreakupeventshasbeenapplied(see text).Dataaregivenbysolid circles.Thehatchedhistogram (visibleatthe bot-tomofthefigure)givestheestimatedsystematicuncertaintyduetotheincomplete coverageofthesolidangle.Thesolidlineshowstheexpecteddistributionforthe quasifreeprocessbasedontheCDBonnpotential[15]deuteronwavefunction.For comparison,thedashedlinegivesthepurephase-spacedistributionasexpectedfor acoherentreactionprocess.

thenp

→

np

π

0

_π

0 _reaction.In

_{Figs. 2–7}

_{theobservablesareshown}

with the MC-generated contamination events alreadysubtracted. Inthe pn invariant-massspectrum Mpn,wherethecontamination

showsupmostpronounced, thisconcerns onlythefirst twobins (Fig. 7).

In Fig. 2, the measured efficiency and acceptance corrected spectatormomentum distributionisshownin comparisonwitha MC simulationof thequasifree dp

→

np

π

0

_π

0

₊

_p

spectator process.

Duetothebeam-pipe,ejectilescanonlybedetectedintheWASA forwarddetectorforlabangleslargerthanthreedegrees.Thegood agreementbetweendataandsimulationprovides confidencethat the dataindeedreflect aquasifree process. Systematic

uncertain-Fig. 3. (Coloronline.) Totalcrosssectionsforthereactionspp→ppπ0_π0_{(left)andnp→npπ}0_π0_{(right).Theresultsofthisworkareshownbythefullcirclesintheright} figure.Statisticalandsystematicuncertainties(Table 1)aresmallerthanthesymbolsize.Theuncertaintyintheabsolutenormalizationintheorderof20%isnotshown. PreviousWASAresultsontheppπ0_π0_{channelareshownbyfullcircles[18]andfullsquare[14],respectively,intheleftfigure,previousbubble-chambermeasurements} fromKEK[16]byopencircles.ThemodifiedValenciamodelcalculationisshownbythesolidlines.Thedash-dottedcurveshowstheresult,ifthes-channeld∗resonance amplitudeisadded.Thed∗contributionitselfisgivenbythedottedcurve.

Fig. 4. (Coloronline.) Distributionsofthec.m.anglesc.m.

p (top)and c_π.m0.

(bot-tom)forthepn→npπ0_π0_reactionatT

n=1.135 GeV.Sincethedataareshown withoutseparationinto√s bins,theycorrespondtotheaverageovertheenergy re-gioncoveredbythequasifreecollisionprocess,whichis2.35 GeV<√s<2.41 GeV (1.07 GeV<Tn<1.23 GeV).Filledcirclesrepresenttheexperimentalresultsofthis work.Thehatchedhistogramsgiveestimatedsystematicuncertaintiesduetothe incompletecoverageofthesolidangle.Theshadedareasdenotephase-space dis-tributions.ThesolidlinesarecalculationswiththemodifiedValenciamodel.The dashed(dash-dotted)linesshowstheresult,ifthed∗ resonanceamplitudewith (without)inclusionofthevertexfunction[1]isadded.Notethatinthe bot-tompaneldashedanddash-dottedcurvesliepracticallyontopofeachother.All calculationsarenormalizedinareatothedata.

ties dueto efficiency andacceptance corrections are very small. Theyareshownashatchedhistogram,barelyvisibleatthebottom lineof

Fig. 2

.Theconstraintforthesuppressionofbreakupevents

(see above) causes the maximumaccepted spectator momentum

to be

<

0.14GeV/cfulfilling the spectator momentum condition used in previous works [1,3,7]. This implies an energy range of 2.35GeV

≤

√

s

≤

2.41GeVbeingcoveredduetotheFermimotion ofthenucleonsinthedeuteron.Thisenergyrangecorrespondsto incidentlabenergiesof1.07GeV

<

Tn

<

1.23GeV.

Intotalasampleofabout24 000goodeventshasbeenselected.

The requirement that the two protons have to be in the

angu-lar rangecovered by the forward detectorand that the gammas resulting from

π

0 _{decay have to be in the angular rangeof the}

central detector reducesthe overall acceptanceto about 7%. The total reconstruction efficiency including all cuts and kinematical fittinghasbeenabout 1%.Efficiencyandacceptancecorrectionsof thedatahavebeenperformedbyMCsimulationsofreaction pro-cessanddetectorsetup.FortheMCsimulationsmodeldescriptions havebeenused,whichwillbediscussedinthenextchapter.Since WASAdoesnot coverthefull reactionphase space,albeit alarge

Fig. 5. (Coloronline.) SameasFig. 4butforthedistributionsoftheinvariantmasses

Mpπ0(top)andMnπ0 (bottom).

fraction of it, the corrections are not fully model independent. The hatched grey histogramsin Figs. 2,4–7 give an estimate for systematic uncertainties due to theuse of differentmodels with andwithoutd∗ resonancehypothesisfortheefficiencycorrection. Comparedtotheuncertaintiesinthesecorrections,systematic er-rors associated withmodeling the reconstruction of particles are negligible.

The absolute normalization of the data has been performed

by the simultaneous measurement of the quasi-free single pion

production process dp

→

pp

π

0

₊

_n

spectator and its comparison to

previous bubble-chamberresultsforthe pp

→

pp

π

0 _reaction

_[16,

17].Thatway,theuncertaintyintheabsolutenormalizationofour dataisessentiallythatoftheprevious pp

→

pp

π

0 _{data,i.e. inthe}

orderof 20%.

3. Resultsanddiscussion

Inordertodeterminetheenergydependenceofthetotalcross sectionwe havedividedourdatasampleinto10MeVbinsin

√

s.

Theresultingtotalcrosssectionstogetherwiththeirstatisticaland systematicuncertaintiesarelistedin

Table 1

.

Fig. 3exhibitstheenergydependenceofthetotalcrosssection forthenp

→

np

π

0

_π

0_{reaction(right)incomparisontothatofthe}

pp

→

pp

π

0

_π

0 _{reaction (left). The previous WASAresults [18,14]}

and theones ofthis work aregiven by the full circles.They are

compared to previous bubble-chamber measurements from KEK

Fig. 6. (Coloronline.) SameasFig. 4butforthedistributionsoftheinvariantmasses

Mnπ0_π0(top)andMpnπ0(bottom).

Incaseof thenp

π

0

_π

0 _{channel, there existno dedicateddata}

fromprevious investigations. However, thereare some connected datafromthePINOTexperimentatSaclay,wheretheinclusive re-actions pp

→

γ γ

X and pd

→

γ γ

X were measured at Tp

=

1

.

3

and1.5 GeV

[19]

.Byexcludingthetwo-photoninvariantmass re-gions correspondingto single

π

0 _or

_η

_{production,the remaining}

two-photon events populating the combinatorial background are likelytooriginatefrom

π

0

_π

0 _{production.Byusingthisfeature,a}

measureoftheratioofthecrosssectionspn

→

pn

π

0

_π

0

₊

_d

_π

0

_π

0

to pp

→

pp

π

0

_π

0 _{has beenobtained. This leads to a crude}

esti-mate for the pn

→

pn

π

0

_π

0 _{cross section to be larger than the}

pp

→

pp

π

0

_π

0 _{crosssectionbyroughlyafactoroftwo–in}

quali-tativesupportofourresultsfromtheexclusivemeasurements

[20]

. In

Fig. 3

,wecomparethedatatotheoreticalcalculationsinthe framework of the Valencia model [21], which incorporates both non-resonantandresonantt-channel processesfortwo-pion pro-duction in N N collisions. The t-channel resonance processes of interest hereconcern first of all the excitation of the Roper res-onance and its subsequent decay either directly into the N

π π

systemorviathe



π

systemaswellastheexcitationanddecayof the



system. Deviatingfromthe originalValenciacalculations

[21],thepresentcalculationshavebeentuned todescribe quanti-tativelytheisovectortwo-pionproductionreactions pp

→

N N

π π

[18],inparticularthepp

π

0

_π

0_[22]andnn

_π

+

_π

+_{[23]channelsby}

thefollowingmodifications:

•

relativistic corrections for the



propagator as given by

Ref.[24],

Fig. 7. (Coloronline.) ThesameasFig. 4,butforthedistributionoftheinvariant massesMπ0_π0(top)andMpn(middle).ThebottompanelshowstherawMpn spec-trumwithoutefficiencyandacceptancecorrections.

•

strongly reduced

ρ

-exchange contribution in the t-channel



process–inagreementwithcalculationsfromRef.[25],

•

reduction of the N∗

→ 

π

amplitude by a factorof two in agreementwiththeanalysisofphoton- andpion-inducedpion productiononthenucleon

[26]

andinagreementwith pp

→

pp

π

0

_π

0_andpp

_→

_pp

_π

+

_π

−_{measurementsclosetothreshold} [27–30] aswellasreadjustment ofthetotalRoper excitation according to the results of the isospin decomposition of the

pp

→

N N

π π

crosssections

[18]

,

•

inclusionofthet-channel excitationofthe

(

1600

)

P33

The lattermodification was necessary,in orderto account for theunexpectedlylargepp

→

nn

π

+

π

+crosssection

[23]

.The

pre-dictive power of these modifications has been demonstrated by

its successful applications to the recent pp

→

pp

π

0

_π

0 _{data at}

Tp

=

1

.

4 GeV[14]andtothe pn

→

pp

π

0

π

−reaction

[5]

.

Finalstateinteraction(FSI)intheemittedN N systemhasbeen taken into account in the Migdal and Watson [31,32] factorized form.

The N N FSI is by far strongest in the isovector 1S0 pn state

andlessstrongin1S0 pp and 3S1 pn statesasapparentfromthe

scatteringlengthsin thesesystems.At energies above 1GeVthe

t-channel



process isthe dominatingone.Isospin decomposi-tionofitscontributiontothetotalnp

→

np

π

0

_π

0_crosssection

_[33,

34,18]showsthatinthisprocess the1S0 final stateis muchless

populatedthantheisoscalar3S1 state.Thesituationissomewhat

differentinthenear-thresholdregion, wheretheRoperexcitation processdominates.Inthisprocess,equalamountsof pn pairsare emittedin1S0and3S1 states.

SincethemodifiedValenciacalculationshavebeentunedtothe

pp

→

pp

π

0

_π

0 _{reaction,itisnosurprisethatitstotalcrosssection}

is fairly well described – see left panel in Fig. 3.For the closely relatednp

→

np

π

0

_π

0 _{reaction, thecalculations predict a similar}

energydependence,butanabsolutecrosssection,whichislarger byroughlya factoroftwo–whereasthedataarelargerbymore thananorderofmagnitude–see

Fig. 3

,rightpanel.

As an independent check of these calculations we may

per-formanisospindecompositionofcrosssectionsusingtheformulas giveninRefs. [33,34] andthe matrixelements deducedfromthe analysisofthe pp inducedtwo-pionproduction

[18]

.Asan result ofsuchan exercisewegetagreementwiththemodified Valencia calculationwithinroughly 30%.

Asweseefrom

Fig. 3

,theexperimentalcrosssectionsobtained inthisworkforthenp

→

np

π

0

_π

0_{reactionarethreetofourtimes}

largerthanpredicted. Thisfailurepoints toanimportantreaction component not included in the t-channel treatment of two-pion production.Itisintriguingthat wedeal herewiththeenergy re-gionwherethe d∗ resonancehasbeenobserved bothinnp

scat-tering

[7]

andin theisoscalarpartofthedouble-pionicfusionto deuterium

[1,3]

.Alsoithasbeenshownthatthedescriptionofthe

pn

→

pp

π

0

_π

− _{cross section improves greatly in this energy}

re-gion,ifthisresonanceisincluded

[5]

.Henceweaddalsoherethe amplitudeofthisresonancetotheconventionalamplitude. Accord-ingtothepredictionsofFäldtandWilkin

[9]

aswellasAlbaladejo andOset

[10]

,its contributionattheresonancemaximumshould beabout200 μb(dottedcurvein

Fig. 3

)asdiscussedinthe intro-duction.It isamazing,how well theresultingcurve (dash-dotted linein

Fig. 3

) describesthedata.Ofcourse,itisapitythatthere arenodataoutsidetheenergyregioncoveredbyourdata.In par-ticular at energies below 1 GeV and above 1.3 GeV, i.e. outside

theresonanceregion,suchdatawouldbeveryhelpfultoexamine

entialobservables.Wechoosetoshowinthispaperthedifferential distributions forthe invariant masses _Mπ0_π0, Mpn, Mpπ0, M_nπ0, Mnπ0_π0 andM_ppπ0 aswellasthedifferentialdistributionsforthe

center-of-mass (cm) angles for protons and pions, namely



c.m.p

and



c.m.

π0 .Thesedistributionsareshownin

Figs. 4–7

.

All measured differential distributions are markedly different

in shape from pure phase-space distributions (shaded areas in

Figs. 3–6), but close to the predictions both with (dashed and dash-dottedlines)andwithout(solidlines)inclusionofthed∗ res-onance.

Thepionangulardistribution(Fig. 4)behavesasexpectedfrom the p-wavedecayofthe



resonance.Andalsotheproton angu-lardistributionissimilarlycurved.Botht-channelmesonexchange andthe JP

₌

₃+_{requirementford}∗ _{formationpredictcomparable}

shapesinagreementwiththedata.

TheinvariantmassspectraforM_pπ0,M_nπ0,M_nπ0_π0 andM_pnπ0

(Figs. 5–6)arecharacterizedby



and N



dynamicsasthey nat-urally appearin the deexcitationprocess of an intermediate



system created either by d∗ decay or via t-channel meson

ex-change.

The Mpn and Mπ0_π0 spectra (Fig. 7) need a more

thor-ough discussion. The data of the _Mπ0_π0 spectrum appear to

be quite well described by the calculations, which hardly

devi-ate from each other. At small invariant masses though, in the

range 0

.

3–0

.

4 GeV

/

c2, there is an indication of a small surplus of strength. Taken the uncertainties inherent in the data and in thetheoretical description,thesedeviationsappearnottobe par-ticularlysignificant.Therefore,ifthisconstitutesasignoftheABC effect, then it is obviously very small in this reaction. Note that contrarytothesituationinthe pn

→

pp

π

0

_π

−_{reaction,wherethe}

pion pair hasto be in relative p-waveand hencethe ABC-effect isabsent,thepionpairhereispreferentiallyinrelatives-wave al-lowingthus,inprinciple,theoccurrenceoftheABCeffect.Hence, the finding that there is no or nearly no ABC effect comes as a surpriseatleastforsomeofitsinterpretations–see,e.g. Ref.[35]. This findingis ofno surprise, ifthe ABC effectis described by a formfactor atthe



vertexofthed∗ decay

[1]

.However, thena problem ariseswiththe description ofthe Mpn spectrum,as we

discussinthefollowing.

The Mpn spectrum peaks sharply at its low-mass threshold,

whichischaracteristicforastrongnp FSIasdiscussedabove.This low-mass peakingiswellaccountedforby themodified Valencia calculations(solidlinesin

Figs. 4–7

).Inclusionofthed∗resonance as outlined in Ref. [1] (dashed lines) exaggerates the low-mass peakingdeterioratingthus theagreementwiththedata.The rea-son for this behavior is the formfactor at the



decay vertex ofd∗ introduced inRef. [1]forthedescription ofthe ABCeffect,

i.e. thelow-massenhancementinthe M_{(π π)}0 spectraobservedin

double-pionic fusion reactions. However, as already pointed out in Ref. [5], this formfactor acts only on the _Mπ0_π0 and Mπ+_π−

spectra, ifthenucleon pairis boundin a final nuclearsystem. If this is not the case, then the formfactor acts predominantly on

theinvariant-massspectrumofthenucleonpair.Thisisillustrated bycomparisonofthecalculationsincludingd∗ with(dashed) and without (dash-dotted)this formfactor. As we see,the formfactor hardly changes the _Mπ0_π0 distribution, but shuffles substantial

strengthinthe Mpn spectrumto lowmasses–thusovershooting

theobservedlow-massenhancement.

Unfortunately, also the model-dependence of the acceptance and efficiency corrections is largest near the low-mass thresh-old hampering thus a definite statement about a failure of the formfactoransatz. In ordertocircumvent thismodeldependence somewhat, we plotthe data in

Fig. 7

, bottom, beforeacceptance andefficiencycorrections. The calculationsshown are nowgiven withintheacceptanceoftheWASAdetector.We seethat,firstof all, the corrections do not change the shape of the distribution profoundly,andsecondthatthecalculationswithformfactor over-shootthelow-masspeakinsimilarmannerasbefore,whereasthe calculationswithoutthisformfactoragreeagainwellwiththedata. This overshooting indicates that the formfactor introduced in Ref.[1]onpurelyphenomenologicalgroundsforthedescriptionof theABC effect is possibly atvariance withthe data forisoscalar two-pionproductioninnon-fusionchannels.Hencealternative so-lutionsfor thisphenomenon may haveto be looked for,such as

d-wave contributionsintheintermediate



systemand/or final nucleon-pair

[36,37]

.

Another alternative involving d-waves has been proposed re-cently by Platonova and Kukulin [35]. In their ansatz they as-sumethe d∗ resonance not only to decayinto the d

π

0

_π

0

chan-nelvia theroute d∗

→ 

+



0

→

d

π

0

_π

0_,7 _{butalso via the route}

d∗

→

d

σ

→

d

π

0

_π

0_.Since

_σ

_{is aspin zeroobject,ithasto bein}

relatived-waveto thedeuteroninthisdecayprocess,inorderto satisfythe resonancecondition of JP

=

3+.Inconsequencethe availablemomentum inthisdecayprocess isconcentrated inthe relativemotionbetweend and

σ

leavingthus onlysmallrelative momentabetweenthetwo emergingpions.Therefore the _Mπ0_π0

distributionisexpectedtobepeakedatlowmasses–i.e.,the low-massenhancement(ABCeffect) inthismodelismadebythed

σ

decaybranch(intheamountofabout5%)andnotbyaformfactor asintroduced in Ref. [1]. The enhancement in thismodel is fur-therincreasedbyaninterferenceofthed

σ

decayamplitude with thedecayamplitude viathe



+



0 system. Itappears straightfor-ward to extend this ansatz also to reaction channels, where the

np system is unbound. However, since we hardly observe a low-massenhancement(ABCeffect)inthe_Mπ0_π0 spectrum,muchless d∗

→

d

σ

contribution is needed here than in the pn

→

d

π

0

_π

0

reaction – which possibly poses a consistency problem for this ansatz

[35]

.

Another point of concern with this ansatz is that mass and widthof the sigmameson havebeen fitted to the pn

→

d

π

0

_π

0

data in Ref. [35] withthe resultthat mσ

≈

300 MeV and

σ

≈

100 MeV. Both values are much smaller than the generally

accepted values for the sigma meson [38], which are mσ

=

(

400–550

)

MeV and

σ

= (

400–700

)

MeV.InRef.[35]ithasbeen

arguedthat thesedeviations couldbe a signofchiral restoration inthehadronic/nuclearenvironment–inparticularwithinthe six-quark bag. However, anyevidence forthishypothesis fromother experiments is lacking so far. Whether the enhanced ABC effect observedinthedouble-pionic fusionto4He

[39]

isinsupportof suchanargumentationisanopenquestion.

7 _{Actuallytheyconsider thedecayd}∗_→D++

12π0→dπ0π0 with D++12 beinga I(JP₎₌₁₍₂+_{)stateneartheNthreshold,butsincethepionemittedinthed}∗ decayisinrelativep-wavetoD12,thisrouteispracticallyindistinguishablefroma d∗→ +_0_{decayatthegivenkinematicconditions.}

4. Conclusions

The np

→

np

π

0

_π

0 _{reaction, for which no dedicated previous}

data exist, has been investigated by exclusive and kinematically complete measurements.Theyhavebeen carriedout inquasifree

kinematics with a deuteron beamimpinging on a hydrogen

pel-let target. Utilizing the nucleons’ Fermi motion in the deuteron projectile an energy region of 2

.

35 GeV

<

√

s

<

2

.

41 GeV could

be covered corresponding to an incident lab energy range of

1.07–1.23 GeV.Thisenergyregioncoverstheregionofthed∗ res-onance. Thedataareinagreement witha resonancecontribution ofabout200 μb,aspredictedbyFäldtandWilkin

[9]

aswellasby AlbaladejoandOset

[10]

.Thed∗ contributionisbyfarlargerthan that fromconventional processes. Calculations based on conven-tionalt-channelmesonexchangeunderpredictthedatabyfactors three to fourand inaddition are at variance withthe measured energydependenceofthetotalcrosssection.Thoughthose calcu-lations havebeen tuned to two-pionproduction channels, where

d∗ doesnot contribute,they still mayhavesome inherent model dependence.But,evenifweassumetheassociateduncertaintyto beaslargeas 50%,westillarriveatanuncertaintyofonly15%for therequiredd∗contribution,i.e. 200

±

30 μb.

In general, the differential data are reasonably well described bycalculations,whichincludeboththed∗ resonanceandthe con-ventionalt-channelprocesses.

Thedatadonot exhibitanysignificantlow-massenhancement (ABCeffect)inthe

π

0

_π

0_{-invariantmassdistribution.Thoughthis}

is not in disagreement with the phenomenological ansatz of a

formfactor at the d∗

→ 

decay vertex introduced in Ref. [1], the worseningof thedescription ofthe Mpn spectrumby use of

thisformfactor calls possibly foran improved explanation ofthe ABCeffectinconnectionwiththed∗ resonance.

After having found evidences for the d∗ resonance in the

d

π

0

_π

0_{, d}

_π

+

_π

− _{and pp}

_π

0

_π

− _{channels, the channel investigated}

here has been one of the two remaining two-pion production

channels, wherethe predicted contributions of the d∗ resonance hadnotyetbeencheckedexperimentally.Aswehaveshownnow, the datafor thenp

π

0

_π

0 _{channel are consistent withthe d}∗

hy-pothesis andprovide an experimentally determined branching of about 12% for the d∗ decay into thischannel. A preliminary list ofdecay branchesis giveninRef. [40],an update of whichis in preparation.

Sinced∗ hasbeenobservedmeanwhilealsointheelastic chan-nelbypolarizednp scattering,



theonlyremainingunexplored de-caychannelisnp

π

+

π

−.Thischannelhasbeenmeasuredrecently

at HADES and preliminary results have been presented already

at conferences [41–43]. It will be highly interesting, not only to obtain total cross sections for this channel, but also differential distributions. Of particular interest will be the Mpn and Mπ+π−

distributionsasdiscussedinthiswork.

Acknowledgements

We acknowledgevaluable discussions withV.Kukulin,E.Oset

and C. Wilkin on this issue. We are particularly indebted to

L. Alvarez-Ruso for using his code. This work has been

sup-portedbyForschungszentrumJülich(COSY-FFE),DFG(CL214/3-1),

the Foundation For Polish Science through the MPD programme

and by the Polish National Science Centre through the Grants

Nos. 2011/01/B/ST2/00431and2013/11/N/ST2/04152.

References

[1]P.Adlarson,etal.,Phys.Rev.Lett.106(2011)242302. [2]M.Bashkanov,etal.,Phys.Rev.Lett.102(2009)052301.

[15]R.Machleidt,Phys.Rev.C63(2001)024001. [16]F.Shimizu,etal.,Nucl.Phys.A386(1982)571. [17]A.M.Eisner,etal.,Phys.Rev.B138(1965)670. [18]T.Skorodko,etal.,Phys.Lett.B679(2009)30. [19]E.Scomparin,PhDthesis,UniversityofTorino,1993. [20] C.Wilkin,privatecommunication.

[21]L.Alvarez-Ruso,E.Oset,E.Hernandez,Nucl. Phys.A633(1998)519,and

pri-[37]F.Huang,Z.Y.Zhang,P.N.Shen,W.L.Wang,arXiv:1408.0458[nucl-th]. [38]J.Behringer,etal.,PDG,Phys.Rev.D86(2012)010001.

[39]P.Adlarson,etal.,Phys.Rev.C86(2012)032201(R).

[40]A.Pricking,M.Bashkanov,H.Clement,arXiv:1310.5532[nucl-ex]. [41]A.K.Kurulkin,etal.,arXiv:1102.1843[hep-ex].

[42]1G.Agakishiev,etal.,Proc.Sci.Baldin-ISHEPP-XXI(2012)041. [43]M.J.Amaryan,etal., Proc.MesonNet2013,arXiv:1308.2575[hep-ph].