The production of antiprotons in proton-nucleus and deuteron-nucleus collisions is analyzed with respect to onestep nucleon-nucleon (NN ---+ NNp/5) and secondary Anucleon (AN ~ NNN/5) production channels within a phase-space model incorporating the selfenergies of the baryons. It is found that contrary to nucleus-nucleus collisions the A induced reaction channels are of minor importance except at very low bombarding energies. The data from KEK are approximately reproduced when accounting for the internal momentum distribution of the deuteron which provides a natural explanation for the large enhancement of the/5 cross section in d+A as compared to p+A. 25.40.-h Antiproton production at energies of a few GeV/u is the most extreme subthreshold production process and has been observed in proton-nucleus collisions some time ago [1][2][3]. In the last years the problem was taken up again at KEK [4] and GSI [5] with new detector setups. Various descriptions for these data have been proposed (cf.
PACS:[6] and refs. therein), most of them based on the assumption of kinetic and chemical equilibrium.First fully relativistic transport calculations for antiproton production including/~ annihilation as well as the change of the quasi-particle properties in the medium have been reported in [7]. There it was found that according to the reduced nucleon mass in the medium the threshold for /5-production is shifted to lower energy and the antiproton cross section prior to annihilation becomes dramatically enhanced as compared to a relativistic cascade calculation where no in-medium effects are incorporated. In fact, Teis et al. [6] have shown recently (within the relativistic BUU approach) that the t5 yields for p + A and A + A can be well reproduced simultaneously when employing proper selfenergies for the baryons in the dense medium.It is the aim of the present note to investigate the question, whether the KEK data on/5 production [4] -which show a dramatic enhancement for the deuteron projectile at the same bombarding energy per nucleon -can be understood within the same scheme as well. Since the RBUU transport approach is not well suited for modelling the proSupported by BMFT, GSI Darmstadt and Forschungszentrum J/ilich perties of a weakly bound deuteron, we use a phase-space approach [8,9] which allows for a proper incorporation of the deuteron momentum distribution.The phase-space model adopted is described in more detail in [9,10]. Its ingredients are as follows: apart from the elementary differential cross section E,jda@/dak~ the inclusive t5 cross section for p+ A reactions is determined by the momentum distribution p(p) describing the available momentum components in a nucleus as well as by pd(k) in case of d + A reactions. We use a Ganssian fit to the experimental momentum distribution for 12C from [11] and use pd(k) for the deuteron from the Paris NN-potential [12]. Whereas the latter two distributions are known sufficiently well, the t5 production cross section in the medium is still a matter of debate. Here we ado...