PHYSICAL REVIEW D VOLUME 35, NUMBER 11 1 JUNE 1987 Production of a light pseudoscalar particle in two-photon collisions M. Sajjad Zahir Department of Physics and Astronomy, University of Regina, Regina, Saskatchewan, Canada S4S OA2 (Received 22 December 1986) The effects of a light-pseudoscalar-particle-like excitation during lepton-pair formation in yy col- lisions have been investigated as an alternative mechanism for its search to confirm the particle in- terpretation of the anomalous positron peaks observed in heavy-ion collisions. It is expected that a laser-induced intense low-energy yy facility will be very useful. A light pseudoscalar boson called the axion' was orig- inally introduced in a model proposed by Peccei and Quinn (PQ) to solve the CP-violation problem in strong interactions. This original axion, now better known as the standard axion, had universal coupling with all fermions in the theory. All its properties were fixed up to an overall parameter x, the ratio of the vacuum expectation values of the two Higgs bosons. The axion mass m, is ' m, = 25N(x + 1/x ) keV, where N is the number of quark doublets, and the couplings with quarks of charge ( — —, ' ) were proportional to x (1/x). With x near uni- ty, m, is in the range of a few hundred keV. However, the standard axion has been ruled out by various experi- ments: (I) decays J/P~ya, Y~ya, (II) beam-dump ex- periments, (III) decays K+~m. +a, and (IV) searches for axions in nuclear deexcitations. The original experi- ments of category (I) were designed for longer-lived ax- ions (r-10 " sec); however, in very recent experiments, the search has been extended to include the short-lived particles (r-10 ' sec) decaying within the detector. However, no evidence of shorter-lived axions has been found either. Although the experimental outcomes were negative, the idea of axion via PQ symmetry breaking was so elegant that, later on, an invisible-axion model was introduced to evade all the observational bounds along with profound astrophysical implications. Recently, the visible-axion model has once again drawn much attention since narrow coincident e+e peaks were detected' in heavy-ion col- lisions at Gesellschaft fiir Schwerionenforschung (GSI) in Darmstadt. One of the possible interpretations" of the results is consistent with the (nearly at rest) production and decay of a neutral boson of mass 1. 7 MeV. If this bo- son is identified with the standard axion, then x=20 or —, ', which then immediately contradicts the results of pre- vious axion searches. Therefore, alternatives to the stan- dard axion have been proposed' by breaking the univer- sality of the axion-fermion interaction. This allows dif- ferent couplings in the different fermion sectors in general and therefore it is possible to avoid all previous experi- mental bounds on axions and at the same time explain the anomalous e+e peaks at GSI. To explain the heavy-ion data at GSI, the real problem, of course, is to produce ax- ions essentially at rest. " In the recently proposed variant axion model of Peccei, Wu, and Yanagida, ' we have m, =25(x+1/x) keV with x — 70, and (ae+e ) coupling is ig, 'P, yqP, a with g, '-(m /f )x, f — 250 GeV giving g, '-10 . This value of g, ' is consistent with the limit obtained from the high accuracy measurements of the anomalous magnetic moment of the electron. " In addi- tion, the Peccei-Quinn axial-vector current has an anoma- ly which introduces axion — two-photon coupling ' (g'/4)F„, F ""a. This is analogous to the anomalous tr -2y coupling allowing the decay vr ~yy In th.e new model, ' since the coupling of axion with electron is pro- portional to x(-70), it enables a more copious produc- tion of e+e as required to explain the GSI experi- ments. " Also, it makes the axion decay rapidly before reaching the dump in beam-dump experiments. At the same time, the coupling of the axion to both charm and bottom quarks is kept weak so as to suppress the decays in category (I) above. The weakness of the model is that the breaking of the universality is done in an ad hoc manner and also, to some extent, it predicts flavor-changing in- teractions. However, recent beam-dump experiments' do not favor the predictions of this variant model either. In fact the proponents' of the variant model themselves express the opinion that although the GSI event might be explained in terms of variant axions, the variant axion model itself is ruled out independently by beam-dump and nuclear- deexcitation experiments. ' However, the anomalous posi- tron events observed at the GSI remain to be given a satis- factory explanation. In a recent paper by Kim, Naqvi, and Zahir, ' it has been argued that the GSI events can still be interpreted in terms of a new light pseudoscalar boson P decaying into e+e without violating other ex- perimental bounds from axion searches. In this paper, we will take the same view that although experiments rule out all types of visible axions, the particle interpretation of the GSI events remains valid in terms of tt whose theoretical basis is still to be found. In the axion model, the mass m, and couplings of a with fermions and pho- tons are constrained. However, for the case of P, we as- sume that the mass m~ of P and its coupling g, with elec- trons are independent and that like the axion, it has an anomaly-driven coupling g& with two photons. In this paper, we want to investigate this P-2y coupling in e+e pair formation in yy collisions (Fig. 1) near the threshold and thereby look for an alternative mechanism for search. Without the anomaly-driven coupling, the pair-formation cross section is given in any QED text- 35 3338 1987 The American Physical Society