First measurement of direct photoproduction of the a 2 (1320 ) 0 meson on the proton
A. Celentano, M. Battaglieri, R. De Vita, L. Marsicano, V. Mathieu, A. Pilloni, A. Szczepaniak
FFirst measurement of direct photoproduction of the a (1320) meson on the proton A. Celentano, V. Mathieu, A. Pilloni,
3, 1
A. Szczepaniak,
4, 5, 6
K. P. Adhikari, S. Adhikari, M.J. Amaryan, G. Angelini, H. Atac, L. Barion, M. Battaglieri,
12, 1
I. Bedlinskiy, Fatiha Benmokhtar, A. Bianconi,
15, 16
A.S. Biselli, F. Boss`u, S. Boiarinov, W.J. Briscoe, W.K. Brooks,
19, 12
D. Bulumulla, V.D. Burkert, D.S. Carman, J.C. Carvajal, P. Chatagnon, T. Chetry, G. Ciullo,
11, 22
L. Clark, P.L. Cole,
24, 25
M. Contalbrigo, O. Cortes, V. Crede, R. Cruz-Torres, A. D’Angelo,
28, 29
N. Dashyan, R. De Vita, M.Defurne, A. Deur, S. Diehl, C. Djalali,
32, 33
M. Dugger, R. Dupre, H. Egiyan,
12, 35
M. Ehrhart, A. El Alaoui, L. El Fassi,
21, 36
L. Elouadrhiri, P. Eugenio, G. Fedotov, ∗ R. Fersch,
38, 39
A. Filippi, G. Gavalian,
12, 35
N. Gevorgyan, F.X. Girod,
12, 18
D.I. Glazier, W. Gohn, † E. Golovatch, R.W. Gothe, K.A. Griffioen, M. Guidal, K. Hafidi, H. Hakobyan,
19, 30
N. Harrison, M. Hattawy, F. Hauenstein, T.B. Hayward, D. Heddle,
38, 12
K. Hicks, A. Hobart, M. Holtrop, Y. Ilieva,
33, 9
D.G. Ireland, B.S. Ishkhanov, E.L. Isupov, D. Jenkins, H.S. Jo, K. Joo, S. Joosten, D. Keller,
43, 32
M. Khachatryan, A. Khanal, M. Khandaker, ‡ A. Kim, C.W. Kim, W. Kim, F.J. Klein, V. Kubarovsky, L. Lanza, M. Leali,
15, 16
P. Lenisa,
11, 22
K. Livingston, V. Lucherini, I .J .D. MacGregor, D. Marchand, N. Markov,
12, 31
L. Marsicano, V. Mascagna,
47, 16, § M.E. McCracken, B. McKinnon, Z.-E. Meziani, M. Mirazita, V. Mokeev, A Movsisyan, E. Munevar, ¶ C. Munoz Camacho, P. Nadel-Turonski, K. Neupane, S. Niccolai, G. Niculescu, M. Osipenko, A.I. Ostrovidov, M. Paolone, L.L. Pappalardo,
11, 22
R. Paremuzyan, E. Pasyuk, W. Phelps, O. Pogorelko, J.W. Price, Y. Prok,
7, 43
M. Ripani, J. Ritman, A. Rizzo,
28, 29
G. Rosner, J. Rowley, F. Sabati´e, C. Salgado, A. Schmidt, R.A. Schumacher, U. Shrestha, D. Sokan, O. Soto, N. Sparveris, S. Stepanyan, I.I. Strakovsky, S. Strauch, J.A. Tan, N. Tyler, M. Ungaro,
12, 31
L. Venturelli,
15, 16
H. Voskanyan, E. Voutier, D. Watts, X. Wei, M.H. Wood,
53, 33
N. Zachariou, J. Zhang, and Z.W. Zhao (The CLAS Collaboration) INFN, Sezione di Genova, 16146 Genova, Italy Departamento de Fsica Terica and IPARCOS, Universidad Complutense de Madrid, 28040 Madrid, Spain European Centre for Theoretical Studies and Nuclear Physics (ECT ∗ ) and FondazioneBruno Kessler, Strada delle Tabarelle 286, Villazzano (Trento), I-38123, Italy Physics Department, Indiana University, Bloomington, IN 47405, USA Center for Exploration of Energy and Matter, Indiana University, Bloomington, IN 47403, USA Theory Center, Thomas Jefferson National Accelerator Facility, Newport News, VA 23606, USA Old Dominion University, Norfolk, Virginia 23529 Florida International University, Miami, Florida 33199 The George Washington University, Washington, DC 20052 Temple University, Philadelphia, PA 19122 INFN, Sezione di Ferrara, 44100 Ferrara, Italy Thomas Jefferson National Accelerator Facility, Newport News, Virginia 23606 National Research Centre Kurchatov Institute - ITEP, Moscow, 117259, Russia Duquesne University, 600 Forbes Avenue, Pittsburgh, PA 15282 Universit`a degli Studi di Brescia, 25123 Brescia, Italy INFN, Sezione di Pavia, 27100 Pavia, Italy Fairfield University, Fairfield CT 06824 IRFU, CEA, Universit´e Paris-Saclay, F-91191 Gif-sur-Yvette, France Universidad T´ecnica Federico Santa Mar´ıa, Casilla 110-V Valpara´ıso, Chile Universit’e Paris-Saclay, CNRS/IN2P3, IJCLab, 91405 Orsay, France Mississippi State University, Mississippi State, MS 39762-5167 Universita’ di Ferrara, 44121 Ferrara, Italy University of Glasgow, Glasgow G12 8QQ, United Kingdom Lamar University, 4400 MLK Blvd, PO Box 10009, Beaumont, Texas 77710 Idaho State University, Pocatello, Idaho 83209 Florida State University, Tallahassee, Florida 32306 Massachusetts Institute of Technology, Cambridge, Massachusetts 02139-4307 INFN, Sezione di Roma Tor Vergata, 00133 Rome, Italy Universita’ di Roma Tor Vergata, 00133 Rome Italy Yerevan Physics Institute, 375036 Yerevan, Armenia University of Connecticut, Storrs, Connecticut 06269 Ohio University, Athens, Ohio 45701 University of South Carolina, Columbia, South Carolina 29208 a r X i v : . [ nu c l - e x ] A p r Arizona State University, Tempe, Arizona 85287-1504 University of New Hampshire, Durham, New Hampshire 03824-3568 Argonne National Laboratory, Argonne, Illinois 60439 Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University, 119234 Moscow, Russia Christopher Newport University, Newport News, Virginia 23606 College of William and Mary, Williamsburg, Virginia 23187-8795 INFN, Sezione di Torino, 10125 Torino, Italy Virginia Tech, Blacksburg, Virginia 24061-0435 Kyungpook National University, Daegu 41566, Republic of Korea University of Virginia, Charlottesville, Virginia 22901 Norfolk State University, Norfolk, Virginia 23504 Catholic University of America, Washington, D.C. 20064 INFN, Laboratori Nazionali di Frascati, 00044 Frascati, Italy Universit`a degli Studi dell’Insubria, 22100 Como, Italy Carnegie Mellon University, Pittsburgh, Pennsylvania 15213 James Madison University, Harrisonburg, Virginia 22807 California State University, Dominguez Hills, Carson, CA 90747 Institute fur Kernphysik (Juelich), Juelich, Germany University of York, York YO10 5DD, United Kingdom Canisius College, Buffalo, NY Duke University, Durham, North Carolina 27708-0305 (Dated: April 16, 2020)We present the first measurement of the exclusive reaction γp → a (1320) p in the photon energyrange 3 . . . < − t < . . Data were collectedwith the CEBAF Large Acceptance Spectrometer at the Thomas Jefferson National AcceleratorFacility. The neutral a resonance was detected by measuring the reaction γp → π ηp and recon-structing the π η invariant mass. The differential cross section dσ/dt was extracted at differentbeam energies in each − t bin. The most prominent feature of the differential cross section is a dipat − t (cid:39) .
55 GeV . This can be well described in the framework of Regge phenomenology, wherethe exchange degeneracy hypothesis predicts a zero in the reaction amplitude for this value of thefour-momentum transfer. It has been more than forty years since QuantumChromodynamics (QCD) was postulated as the theory ofstrong interactions. While much progress has been madein understanding the high energy phenomena throughthis theory, perturbative methods fail to describe thestrong interaction at low energies. A clear understandingof this regime is of key importance, since it correspondsto the dominant manifestation of the strong force in na-ture, in terms of hadrons that constitute the bulk of thevisible mass of the Universe.Hadron spectroscopy is a valuable tool to experimen-tally investigate this regime. The measurement of themeson spectrum, in particular searching for exotic statesnot compatible with the Quark Model, would provideaccess to the gluonic degrees of freedom that contributeto the quantum numbers of the hadrons. Investigatingthe properties and interactions of gluons is critical, sincetheir dynamics give rise to the strong interaction thatbinds the hadrons. In this context, the photoproductionof a π η pair on the proton ( γp → π η p ) is one of themost promising reaction channels: due to the presence oftwo neutral pseudoscalar mesons in the final state, any P -wave resonance would be unambiguously interpretedas an exotic , non qq state. So far, only a few results havebeen reported for this reaction. At low energies, in thefully non-perturbative regime, high-quality cross-sectiondata have been collected by the GRAAL [1], Crystal Ball, TAPS, and A2 [2, 3], and CB-ELSA [4, 5] collaborations.In the multi-GeV photon beam energy range, optimalfor meson spectroscopy, instead, no data have been pub-lished so far.Understanding the π η production is crucial for spec-troscopy, especially for disentangling new resonance sig-nals from non-resonant backgrounds. In the aforemen-tioned energy regime, the a (1320) meson is expected tomake the dominant contribution to the πη invariant-massspectrum [6]. It can be thus taken as the reference statefor a Partial Wave Analysis of this channel, for exampleallowing for the interpretation of the variations of the P − D phase difference as a signature for the existence ofexotic resonances [7, 8]. Photoproduction of the charged a resonance has been measured at SLAC [9–11]. How-ever, to the best of our knowledge the neutral a channelhas never been studied in photoproduction.In this work we report the first measurement of theneutral a (1320) meson photoproduction on the proton,for photon beam energies between 3 . . − t ) in the range0 . . . The differential cross section dσ/dt wasobtained by measuring the cross section d σ/dtdM forthe exclusive production of a π η pair on the proton,where M is the two-meson invariant mass, and extract-ing the contribution of the a resonance in each kine-matic bin. The measurement was performed with theCEBAF Large Acceptance Spectrometer (CLAS) in HallB at Jefferson Laboratory in a dedicated high-energy,high-statistics run, g12 .In this analysis, all three hadrons in the final statewere measured. The π and η were measured by de-tecting the four photons from their decays, whereas theproton was detected directly. The experiment used abremsstrahlung photon beam impinging on a 40-cm-longLH target. The photon beam was produced by the inter-action of the primary E = 5 .
72 GeV electron beam witha converter of 10 − radiation lengths. A magnetic spec-trometer (photon tagger) with energy resolution 0 . E and acceptance in the range 0 . E –0 . E was used totag photons impinging on the target, by measuring thedeflected electron momentum [12, 13]. The average elec-tron beam current was approximately 60 nA, resultingin a photon flux of ∼ · γ/ s. During the run, thephoton flux was measured by sampling the “out-of-time”electron hits in the photon tagger [14].Outgoing particles were measured with the CLAS de-tector [15]. This was a large-acceptance spectrometer,based on a toroidal magnet made of six superconduct-ing coils arranged symmetrically around the beamline toproduce a field pointing primarily in the azimuthal di-rection [16]. The coils divided the CLAS detector intosix independent magnetic spectrometers (sectors), thatshared a common target, trigger and data acquisitionsystem. The momentum of a charged particle was deter-mined from the radius of curvature of its trajectory in themagnetic field as measured by a multi-wire drift-chambersystem (DC) [17]. A set of plastic scintillator counters(TOF), installed behind the drift chambers, provided thetime of flight of each particle [18]. Particle identification(PID) was performed through the β vs. p technique. Theenergies and angles of the photons were measured with alead/scintillator electromagnetic calorimeter (EC), cov-ering polar angles in the range 8 ◦ –45 ◦ , with energy res-olution σ E /E (cid:39) / (cid:112) E ( GeV), and angular resolution σ θ (cid:39)
10 mrad [19]. Although CLAS was optimized forcharged multi-particle final states, the reaction γp → γp could be measured thanks to the high statistics and thespecific setup of the g12 run, with the target moved up-stream to maximize the acceptance to multi-meson finalstates.The incoming photon was identified based on a co-incidence between the vertex times obtained from thephoton tagger and from the CLAS detector. The lat-ter was determined by measuring the time of the outgo-ing charged particles with an array of plastic scintillatorcounters (ST) surrounding the target [20]. A time co-incidence window of ± . ∼ M gg (GeV)0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.905101520253035404550 · M gg (GeV)0.4 0.45 0.5 0.55 0.6 0.6500.511.522.5 · FIG. 1. (Colors online) M γγ distribution for events from thereaction γp → γ p . For each event, both combinations M γ γ and M γ γ are considered, where γ and γ are the photonswith the smallest opening angle. Black curve: all events, redcurve: events satisfying a 1 .
86% CL cut. Green curve: eventssatisfying a 1 .
86% CL cut, using the corrected 4-momentafrom the kinematic fit. The inset shows a zoom in the η massregion. by including accidental photon tagger hits in the MonteCarlo simulations used to determine the detector accep-tance and efficiency, and adopting the same selection pro-cedure.The g12 experiment used an FPGA-based trigger sys-tem, with multiple algorithms implemented in paral-lel [21]. The main trigger condition required the pres-ence of one charged particle, defined as a coincidencebetween one TOF hit and one ST hit in the same CLASsector, and two photons in different CLAS sectors, eachdefined as an EC hit above a threshold of approximately100 MeV. The efficiency of the trigger system was evalu-ated from special minimum bias runs and found to be onaverage ε trg = 80%. To account for the trigger efficiencydependence on the proton impact point on the detec-tor, a trigger efficiency map, as a function of the protonthree-momentum, was derived and used to correct thecross-section normalization.This analysis required a proton and four neutral parti-cles detected, and no other charged particles. The stan-dard g12 procedures, including momentum correctionsand fiducial cuts, were applied (see Ref. [21] for a com-plete description). All neutral particles were consideredto be photons, with energies and angles measured by theEC. The selection of events belonging to the exclusive γp → γp reaction was done through a 4 C kinematic fit(energy and momentum conservation imposed) [22, 23].To this end, the default g12 covariance matrix (CVM)parameterization for the beam photon and for the finalstate proton was used [21]. The final state photons CVMelements, instead, were determined using the reaction γp → π p → e + e − γ p [24].The exclusivity of the final state was ensured by intro-ducing a cut on the kinematic fit confidence level (CL).To optimize this cut, the K kinematic variable, definedas the difference between the missing mass on the protonsquared and the four photon invariant mass squared: K = (cid:0) p µ beam + p µ target − p µp (cid:1) − (cid:32) (cid:88) i =1 p µγ i (cid:33) , (1)was introduced, where p µj is the measured four momen-tum of particle j . From energy and momentum con-servation, it follows that signal events ( γp → γp ) aredistributed around K = 0 with a gaussian distribution,while background events ( γp → γpX ) manifest as a tailin the K > n s √ n s + n b , (2)where n s / n b ) was the number of events with K < K > . γp → γ p reaction isshown in Fig. 1. In addition to aiding exclusive eventselection, the kinematic fit significantly improves the ex-perimental resolution.The kinematic fit discussed above was used to select aclean sample of γp → γp exclusive events. The extrac-tion of the γp → π ηp yield from this sample was doneas follows. First, the four photons were ordered event-by-event by naming γ and γ those with the smallestopening angle. This algorithm exploits the fact that, dueto the lower pion mass, the two photons from the π de-cay are expected to have, on average, a smaller openingangle than those from η decay. The corresponding effi-ciency, estimated from Monte Carlo simulations, is ap-proximately 82% [25]. The correlation between the in-variant masses of the two photon pairs, M γ γ vs. M γ γ is shown in Fig. 2. Signal events were identified as thosecorresponding to the bottom-right cluster centered at M = M π , M = M η . A small fraction of events,corresponding to (cid:39)
4% of the main signal yield, appearin the opposite combination, and was not considered inthe following.As shown in Fig. 1, after ordering the photons, the M γ γ distribution showed a clear peak corresponding tothe η , with some residual background underneath comingfrom exclusive γp → γp events with a different topol-ogy. To reject these and extract the signal yield, the s P lot method was used [26]. This considers that events in thedata sample originate from different independent sourcesand are characterized by a set of kinematic variables thatcan be split into two components. The method allows (GeV) g g M ( G e V ) g g M FIG. 2. (Colors online) Correlation between the invariantmass of the two photon pairs for exclusive γp → γ p events.In each event, γ and γ are the photons with the small-est opening angle. The bottom-right cluster contains signalevents from the γp → π ηp reaction. The few events appear-ing in the opposite photon assignment combination (top-leftcluster) were not used in this analysis. to reconstruct for each event source the distributions of control variables from the knowledge of the ProbabilityDensity Function (PDF) associated to independent dis-criminating variables . To do so, an extended maximumlikelihood fit is performed on the discriminating variablesto assign to each event a set of statistical weights, eachassociated with a specific data source.In this analysis, the invariant mass M γ γ was used asthe single discriminating variable, while M and M γ γ were used as control variables. Two event sources wereassumed: a signal source corresponding to the η me-son decay, modeled with a Gaussian PDF with expo-nential tails, and a background source, parameterizedwith a polynomial PDF. The nominal fit range was0 . < M γ γ < . M bins, and the s P lot analysiswas applied independently in each of them. A total eventyield of 26 . · , defined as the sum of the s P lot signalweights for all events, was obtained. To assess the qual-ity of the s P lot method, the M γ γ distribution for thesignal source was investigated, finding that no residualbackground was present below the π peak.The CLAS acceptance and efficiency were evaluated bymeans of Monte Carlo simulations, based on a GEANTcode that included knowledge of the full detector ge- N bins Range Width E beam . . − t . . Non-uniform. Bin margins:0 . . . . . . M
30 0 . . ometry and a realistic response to traversing particles.Since the extracted differential cross section is integratedover some of the independent kinematic variables (suchas the π angles in the Gottfried-Jackson frame), themodel used to generate Monte Carlo events had to beas close as possible to the real physical one. To this end, γp → π ηp events were first generated according to abremsstrahlung photon beam energy spectrum, with aphase-space distribution, and reconstructed through thesame procedure used for real data. The result was usedto compute the acceptance-corrected event distribution,from which a new Monte Carlo sample was generated.The procedure was iterated until a good agreement be-tween data and Monte Carlo was found. To account forthe effect of the analysis procedures in the cross-sectionnormalization, the same methods were applied to MonteCarlo events. These include the exclusion of events withmore than one photon in the coincidence time window,the kinematic fit CL cut, the photon ordering algorithm,and the s P lot method.The differential cross section d σ/dtdM is shown inFig. 3, as a function of M , for three photon-beam en-ergy bins (rows) and five four-momentum transfer bins(columns), as reported in Table I. The error bars reportthe statistical uncertainty only. This was computed, ineach bin, by adding in quadrature the statistical uncer-tainty on the event yield (equal to the square root ofthe sum of the s P lot weights squared) with that on thedetector acceptance and efficiency.The systematic uncertainties in the cross-section de-termination are summarized in Table II. The first fourcontributions are connected, respectively, to the uncer-tainty in the LH target properties (density and length),the absolute photon flux normalization, the trigger sys-tem efficiency, and the η → γγ branching fraction. Thesystematic uncertainties associated with the kinematic fitand the s P lot procedure have been evaluated by consider-ing, in each bin, the relative variation of the cross sectionfor different choices of the kinematic fit CL cut and ofthe degree of the background polynomial PDF. Finally,the systematic uncertainty on the CLAS acceptance wasevaluated by varying the shape of generated events usedin the Monte Carlo simulation. The reported total sys-tematic uncertainty of the cross section was obtained byadding in quadrature all individual contributions.The differential cross section d σ/dtdM shows twodistinctive structures corresponding to the a (980) and Systematic Uncertainty Source Magnitude
Target properties 0.5%Photon flux 5.7%Trigger efficiency 2.8% η → γγ branching fraction 0.5%Kinematic fit Variable, ∼ s P lot Variable, ∼ ∼ γp → pπ η differential cross-section measurement. Theeffects marked as “variable” have a different contribution foreach E beam , t and M kinematic bin. The typical values arereported. a (1320) resonances. In particular, the a meson isclearly visible as a peak over a smooth background,with the latter decreasing at larger beam energies. Theexclusive a (1320) photoproduction cross section dσ/dt has been extracted in the two largest photon beam en-ergy bins by modeling d σ/dtdM in the M range 1 . .
55 GeV as the incoherent sum of a resonance term anda smooth background, including contributions from bothnon-resonant π η photoproduction and from the residualhigh-mass tail of the a (980) state. The resonance termwas written as the product of a ( E beam , − t )–dependentproduction coefficient and a Breit-Wigner function thatdescribes the a line shape [27]. The background termwas parameterized as an exponentially suppressed zero-order polynomial. The cross-section model was con-voluted with the experimental π η invariant-mass res-olution, evaluated from Monte Carlo simulations. Thisranged from a few MeV at high M values up to ∼ M ∼ . χ fit to all d σ/dtdM data points was then performed, with a to-tal of 28 free parameters (9 a production coefficients, 9background polynomial terms, 9 background exponentialslopes, and the a mass). In the Breit-Wigner formula,the a mass M a was left to vary as a free parameterwhile the width Γ a was fixed to the nominal PDG value,(113 . ± .
3) MeV – the effect of this choice was studiedand included in the systematic uncertainty. The χ / NDFvalue was 64 . /
53 = 1 .
21, and the obtained M a valuewas (1309 ±
2) MeV, in very good agreement with thenominal PDG value, (1312 . ± .
8) MeV. The fit result isreported for each kinematic bin in Fig. 3 as a red curve,while the black curve shows the a contribution only.The differential cross section for the reaction γp → a (1320) p was finally obtained by integrating the res-onance term in each kinematic bin, accounting for the a → π η branching fraction, (14 . ± . . . . . χ fit. The colored bands at the bottom show thesystematic uncertainty, obtained summing quadratically = . - . GeV beam E - -t = 0.2 = . - . GeV beam E - -t = 0.4 = . - . GeV beam E - -t = 0.2 = . - . GeV beam E - -t = 0.4 = . - . GeV beam E - -t = 0.7 = . - . GeV beam E - -t = 1.1 = . - . GeV beam E - -t = 0.2 = . - . GeV beam E - -t = 0.4 = . - . GeV beam E - -t = 0.7 = . - . GeV beam E - -t = 1.1 = . - . GeV beam E - -t = 1.5 (GeV) h p M ) b a r n / G e V m / d t d M ( s d FIG. 3. (Colors online) Differential cross section for the reaction γp → π ηp . Each histogram reports the reaction differentialcross section d σ/dtdM as a function of the π η invariant mass, for the specific E beam and − t bin reported in the same panel.The bottom gray-filled area in each panel shows the systematic uncertainty. The red curve is the result from the best fitperformed with the model described in the text. The black curve corresponds to the contribution of the a resonance only. the systematic uncertainty for d σ/dtdM and that asso-ciated with the fit procedure. The latter was evaluatedby repeating the fit with different choices of the fit range,the background polynomial order, and the a width – thelatter was varied within ± σ around the nominal value.The systematic uncertainty was calculated, in each bin,as the RMS of the cross-section data obtained from thedifferent fits.The most intriguing feature of the γp → a (1320) p cross section is the presence of a dip at − t (cid:39) .
55 GeV .The hypothesis that this dip, observed simultaneously atboth beam energies for the same − t value, was just the ef-fect of a statistical fluctuation, was excluded at 99% CLby comparing the two measured cross-section values tothose linearly extrapolated from the nearby data points.The dip can be explained in the context of Regge the-ory and the specific location of the dip can be tracedto the properties of the Regge poles [29]. In Fig. 4, weshow the prediction of a model based on a Regge-theoryproduction amplitude parametrization developed by theJPAC Collaboration [28], computed for the two beam en-ergies 4 GeV (black) and 5 GeV (red). The amplitude includes the leading vector trajectories only, which forthe production of a neutral a (1320) state have the ρ and ω quantum numbers. The magnitude of the amplitudeis determined by the coupling of the vector trajectoryto γa . This is computed using Vector Meson Domi-nance, from the known a → ωππ width [27], by furtherassuming that the ρ dominates the ππ state. Regge-resonance duality implies the parameters of Regge am-plitudes corresponding to the ρ, ω vector exchanges areclosely related to the ones involving the tensor a and f mesons. This is referred to as exchange degeneracy(EXD) [29, 30]. Since there is no scalar meson in thespectrum that could lie on the a trajectory, the residueof the tensor exchange has to vanish when the Regge tra-jectory α ( t ) is equal to zero to remove the scalar pole.Vector exchanges, which by EXD share the residues withthe tensors, will thus also vanish at α ( t ) = 0, that is at − t = m ρ,ω ∼ .
55 GeV . This leads to an exact zero inthe cross section for this value of − t . However, sublead-ing Regge poles or cut contributions that correspond toexchanges of heavier mesons or absorption corrections,can turn the zero of the amplitude into the dip observed ) -t (GeV0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 ) b a r n / G e V m / d t ( s d FIG. 4. (Colors online) Differential cross section dσ/dt for the reaction γp → a (1320) p , for E beam = 3 . . E beam = 4 . . − t bins width. The bottom semi-transparent bands show the systematic uncertainty (see text for details). The continuouslines are predictions from the JPAC model [28], computed respectively for a beam energy of 4 GeV (black) and 5 GeV (red). in data and improve the description at higher − t . Thegood agreement between data and the model predictiondemonstrates the effectiveness of a Regge phenomenologybased parametrization of the reaction amplitude. Thiswill allow high-statistics photoproduction experiments, e.g. CLAS12 [31], GLUEX [32], and BGOOD [33], toproperly describe the production of the dominant a res-onance in the π η system, using it as a benchmark in thesearch for exotic states.In summary, we have measured the reaction γp → π ηp in the photon beam energy range 3 . . . . , extracting for the first time the cross sec-tion for the exclusive a (1320) photoproduction on theproton. The cross section shows a pronounced dip at − t (cid:39) .
55 GeV , which can be explained by the exchangedegeneracy hypothesis in the framework of Regge the-ory. Since the a (1320) is the most prominent structurepresent in the π η invariant mass, detailed knowledge ofits production cross section is valuable for any assess-ment of a possible exotic resonance contribution. Thismeasurement will thus help high statistics photoproduc-tion experiments searching for exotic mesons to betterunderstand the ηπ mass spectrum.This work was supported by: the U.S. Departmentof Energy (DOE), the U.S. National Science Founda-tion, the U.S. Jeffress Memorial Trust; the Physics andAstronomy Department and the Office of Research andEconomic Development at Mississippi State University, the United Kingdom’s Science and Technology Facili-ties Council (STFC), the Italian Istituto Nazionale diFisica Nucleare; the French Institut National de PhysiqueNucl´eaire et de Physique des Particules, the French Cen-tre National de la Recherche Scientifique; and the Na-tional Research Foundation of Korea. This material isbased upon work supported by the U.S. Department ofEnergy, Office of Science, Office of Nuclear Physics un-der contract DE-AC05-06OR23177. V.M. is supportedby Comunidad Aut´onoma de Madrid through Programade Atracci´on de Talento Investigador 2018 (Modalidad1). ∗ Current address:Ohio University, Athens, Ohio 45701 † Current address:unused, unused ‡ Current address:Idaho State University, Pocatello, Idaho83209 § Current address:Universit`a degli Studi di Brescia, 25123Brescia, Italy ¶ Current address:Thomas Jefferson National AcceleratorFacility, Newport News, Virginia 23606[1] J. Ajaka et al. , Phys. Rev. Lett. , 052003 (2008).[2] V. Kashevarov et al. (Crystal Ball at MAMI, TAPS, andA2 Collaborations), Phys. Lett.
B693 , 551 (2010).[3] V. L. Kashevarov et al. (Crystal Ball at MAMI, TAPS,and A2 Collaborations), Eur. Phys. J.
A42 , 141 (2009).[4] I. Horn et al. (CB-ELSA Collaboration), Phys. Rev. Lett. , 202002 (2008). [5] I. Horn et al. (CB-ELSA Collaboration),
Proceedings,12th International Conference on Hadron spectroscopy(Hadron 2007): Frascati, Italy, October 7-13, 2007 , Eur.Phys. J.
A38 , 173 (2008).[6] C. Adolph et al. (COMPASS Collaboration), Phys. Lett.
B740 , 303 (2015).[7] A. Rodas et al. (JPAC Collaboration), Phys. Rev. Lett. , 042002 (2019).[8] M. Albrecht et al. (Crystal Barrel Collaboration),(2019), arXiv:1909.07091.[9] G. T. Condo, T. Handler, W. M. Bugg, G. R. Blackett,M. Pisharody, and K. A. Danyo, Phys. Rev. D , 3045(1993).[10] G. T. Condo et al. , Phys. Rev. D , 3317 (1990).[11] J. Ballam, G. B. Chadwick, Z. G. T. Guiragossian,A. Levy, M. Menke, P. Seyboth, and G. E. Wolf, Phys.Lett. , 421 (1969).[12] D. I. Sober et al. , Nucl. Instrum. Meth. A440 , 263(2000).[13] S. Stepanyan, S. Boyarinov, H. Egiyan, L. Guo, D. Dale,M. Gabrielyan, L. Gan, A. Gasparian, A. Glamazdin,B. Mecking, I. Nakagawa, A. Teymurazyan, andM. Wood, Nucl. Instrum. Meth.
A572 , 654 (2007).[14] J. Ball and E. Pasyuk,
Photon Flux DeterminationThrough Sampling of “out-of-time” hits with theHall B Photon Tagger , CLAS-NOTE 2005-002 (2005) .[15] B. A. Mecking et al. , Nucl. Instrum. Meth.
A503 , 513(2003).[16] Yu. A. Shishov, Phys. Part. Nucl. , 444 (2006), [Fiz.Elem. Chast. Atom. Yadra 37,843(2006)].[17] M. D. Mestayer et al. , Nucl. Instrum. Meth. A449 , 81(2000).[18] E. S. Smith et al. , Nucl. Instrum. Meth.
A432 , 265(1999).[19] M. Amarian et al. , Nucl. Instrum. Meth.
A460 , 239(2001).[20] Y. G. Sharabian, M. Battaglieri, V. Burkert, R. De Vita, L. Elouadrhiri, L. Guo, D. Kashy, V. Kubarovsky,G. Mutchler, M. Ostrick, M. Ripani, P. Rossi, A. Rot-tura, E. Pasyuk, and D. Weygand, Nucl. Instrum. Meth.
A556 , 246 (2006).[21] Z. Akbar et al. , g12 Analysis Procedures, Statis-tics and Systematics , CLAS-NOTE 2017-002 (2017) https://misportal.jlab.org/ul/Physics/Hall-B/clas/viewFile.cfm/2017-002.pdf?documentId=756 .[22] D. Keller,
Techniques in kinematic fitting , CLAS-NOTE2010-015 (2010) https://misportal.jlab.org/ul/Physics/Hall-B/clas/viewFile.cfm/2010-015.pdf?documentId=614 .[23] M. Williams and C. A. Meyer,
Kinematic Fitting inCLAS , CLAS-NOTE 2003-017 (2003) .[24] M. C. Kunkel et al. (CLAS Collaboration), Phys. Rev. C , 015207 (2018).[25] A. Celentano, The Forward Tagger detector for CLAS12at Jefferson Laboratory and the MesonEx experiment ,Ph.D. thesis, Universit`a degli Studi di Genova (2014).[26] M. Pivk and F. R. Le Diberder, Nucl. Instrum. Meth.
A555 , 356 (2005).[27] M. Tanabashi et al. (Particle Data Group), Phys. Rev.D , 030001 (2018).[28] JPAC Collaboration, Private communication.[29] A. C. Irving and R. P. Worden, Phys. Rept. , 117(1977).[30] J. Mandula, J. Weyers, and G. Zweig, Ann. Rev. Nucl.Part. Sci. , 289 (1970).[31] A. Celentano, Journal of Physics: Conference Series ,012007 (2013).[32] A. Austregesilo (GlueX Collaboration), Proceedings,21st International Conference on Particles and Nuclei(PANIC 17): Beijing, China, September 1-5, 2017 , Int.J. Mod. Phys. Conf. Ser. , 1860029 (2018).[33] S. Alef et al.et al.