The optical model of NN interaction without cut-off radius. O.D. Dalkarov, A. Yu. Voronin P. N. Lebedev Physical Institute 53 Leninsky pr.,117924 Moscow, Russia WesuggestaregularpotentialmodelofNN interactionwithoutanycut-off. Theeffectofsingular terms of OBE potential, modified by annihilation is shown to be repulsive. The experimental data for S- and P-wave scattering lengths are well reproduced. 4 0 I. INTRODUCTION 0 2 During the last decades numerous nonrelativistic models of NN low energy interaction [1, 2, 3, 4, 5, 6] have been n suggestedaswellasgreatexperimentaleffortshavebeenmadeontheLowEnergyAntiprotonRing(LEAR)atCERN. a Anintriguingproblemofpossibleexistingofquasi-nuclearNN states[7]stronglystimulatedthementionedresearches. J Thephysicalargumentsinfavorofquasi-nuclearstatesarethefollowing. TheinteractionbetweenN andN shouldbe 3 much more attractive than the NN interaction, as it follows from the procedure of G-conjugation [1]. Such a strong 1 attractionshouldproduceaspectrumofNN quasi-boundstates(socalledbaryonium). Inthesametimetherangeof annihilation,estimatedfromthepositionofthenearesttothethresholdsingularityinFeynmanannihilationdiagrams, 2 ismuchsmallerthantherangeofmesonexchangeforces. Thismeansthatbaryoniumstatescouldberathernarrowto v 3 be observedexperimentally. ItwasindeeddiscoveredinLEARexperimentsthatcertainpartialwavescontainnarrow 2 nearthreshold resonances (so called P-wave enhancement). Unfortunately, this transparent physical picture has a 0 significantdrawback. TheG-conjugationofNN OBEpotentialyieldsinattractivesingulartermsinNN potentialof 1 thetype1/r3. (IncaseofNN thesetermsarerepulsiveandplayaroleofsocalledshortrangecore). Itiswellknown, 0 that attractive singular potentials produce a collapse [8], i.e. the spectrum of the system is not bounded from below, 4 while thescatteringproblemhasnodefinite solution. The usualwayofdealingwithsuchpathologicalpotentialsisto 0 / impose, that singular behavior is an artifact of certain approximations (for instance nonrelativistic approximation). h In the absence of the self-consistent theory it is common practice to introduce the phenomenologicalcut-off radius to t - regularize the singular behavior of the model at short distance. However the results change dramatically with small l c variations of the cut-off radius[9], which seriously diminish predictive power of the model. In fact it is not clear if u the nearthreshold states are determined by the ”physical part” of the OBEP, or they are artifacts, produced by the n ”non-physical”,singular part of the interaction. : v The aim of the present study is to analyze the role of singular interaction and suggest a model of NN interaction i which is free from the non-physical cut-off parameters. The main idea of our approach is that strong enough short X range annihilation regularizes the singular attractive potentials, as far as the particle annihilates rather than falls r to the center. Mathematically this means that singular potential can be regularized by complex addition to the a coupling constant. The main properties of the potentials of the type (α+iω)/rs, s 2 were studied in [10]. It − ≥ was shown that such potentials with ω = 0 become regular and the scattering on such a potential is equivalent to 6 the full absorptionof the particles in the scattering center. These results enable to suggest a regular potential model of NN interaction, based on OBE potential, but without any cut-off radius. We will show that the overall effect of the ”singular” part of potential, modified by annihilation is repulsive and thus cannot be responsible for new ”false” states. The neathreshold resonances, which are well reproduced by our model, are determined mainly by long range part of OBEP. We calculate the scattering lengths in S- and P- partial waves for different values of spin, isospin and total momentum and demonstrate that our model reproduces rather well the experimental data. II. SINGULAR POTENTIALS WITH COMPLEX STRENGTH. Inthissectionwepresentthemainresultsconcerningthepropertiesofsingularpotentialswithacomplexstrength. In the following we put 2M = 1 and consider the interaction strength to be complex α = Reα iω. Near the s s origin we can neglect all the nonsingular terms of the Shr¨odinger equation, increasing slower than 1/±r2. One get the following expression for the wave-function near the origin [11]: 2 Φ(r) = √r H(1)(z)+exp(2iδ )H(2)(z) (1) µ 0 µ z = 2√α(cid:16)sr−(s−2)/2 (cid:17) (2) s 2 − 2l+1 µ = (3) s 2 − Here H(1)(z) and H(2)(z) are the Hankel functions of order µ [12], δ is a contribution of the short range part of µ µ 0 the singular potential into the scattering phase. It is worth to mention that the variable z is a semiclassical phase. Letusreplacethesingularpotentialatdistancelessthanr bytheconstant α/rs,havinginmindtotendr 0. 0 − 0 0 → Matching the logarithmic derivatives for the ”square-well” solution and the solution (1) at small r , one can get for 0 δ : 0 δ = p(r )r (4) 0 0 0 √Reα iω s p(r ) = ± (5) 0 rs/2 0 Now it is important that the interaction strength α is complex. In the limit r 0 we obtain: s 0 → √Reα iω s lim Imδ =Im ± (6) r0→0 0 r(s−2)/2 →±∞ 0 which means, that exp(2iδ ) is either 0 or and the linear combination of the Shr¨odinger equation solutions (1) is 0 ∞ uniquely defined in the limit of zero cut-off radius r : 0 √rH(1)(z) if ω >0 lim Φ(r)= µ (7) r0→0 (√rHµ(2)(z) if ω <0 One can see, that ω > 0 selects an incoming wave , which corresponds to the full absorption of the particle in the scattering center, while ω < 0 selects an outgoing wave, which corresponds to the creation of the particle in the scattering center. As one can see from (4) and (5) due to the singular character of our potential (s > 2), the above conclusions are valid forany infinitesimal value of ω. Itmeans, that the signof aninfinitesimal imaginaryadditionto the interaction constant selects the full absorption or the full creation boundary condition (7). This boundary condition can be formulated in terms of the logarithmic derivative in the origin: Φ′(r) lim = isign(ω)p(r) (8) r→0 Φ(r) − where p(r) is a classical local momentum (5). (Compare with plane incoming (outgoing) wave boundary condition exp( ipr)′/exp( ipr)= ip). ∓ ∓ ∓ As soon as solution of the Shr¨odinger equation is uniquely defined, we can calculate the scattering observables. In particular we can now obtain the S-wave scattering length for the potential (α i0)/rs (for s>3) : s − ± √αs 2/(s−2) Γ((s 3)/(s 2)) a=exp( iπ/(s 2)) − − (9) ∓ − s 2 Γ((s 1)/(s 2)) (cid:18) − (cid:19) − − The fact, that in spite Imα 0 the scattering length has nonzero imaginary part is the manifestation of the s → ± singular properties of the potential which violates the self-adjointness of the Hamiltonian. Let us compare the scattering length (9) with that of the repulsive singular potential α /rs. One can get: s arep = √αs 2/(s−2) Γ((s−3)/(s−2)) (10) s 2 Γ((s 1)/(s 2)) (cid:18) − (cid:19) − − It is easy to see, that (9) can be obtained from (10) simply by choosing the certain branch (corresponding to 2/(s−2) an absorption or a creation) of the function √α when passing through the branching point α = 0. The s s (cid:0) (cid:1) 3 scattering length in a regularizedsingular potential becomes an analyticalfunction of α in the whole complex plane s of α with a cut along positive real axis. One can see, that the presence of an inelastic component in the singular s potentialactsinthe sameway,asarepulsion. ItsuppressesoneoftwosolutionsoftheSchr¨odingerequationandthus eliminates the collapse. Itiseasytosee,thattheboundarycondition(7)ofthefullabsorption(creation)isincompatiblewiththeexistence of anybound state. Indeed, one needs both incoming andreflected waveto form a standing wave,correspondingto a bound state. This means, that the regularizedsingular potential supports no bound states. This is also clear from the mentioned above fact, that the scattering length for a regularized attractive singular potential is an analytical continuation of the scattering length of a repulsive potential. A. Potential −α2/r2 Let us now turn to the very important case α/r2. The wave-function now is: − Φ = √r J (kr)+exp(2iδ )J (kr) (11) ν+ 0 ν− ν± = (cid:2)1/4 α2 (cid:3) (12) ± − p where k = √E, and J are the Bessel functions [12]. In the following we will be interested in the values of Reα ν± 2 greaterthancriticalReα >1/4. Weusethesamecut-offprocedureatsmallr . Matchingthelogarithmicderivatives 2 0 at cut-off point r we get for exp(2iδ ): 0 0 lim exp(2iδ )=rν+−ν−const r2ν =rω/√Reα2−1/4r−2i√Reα2−1/4 r0→0 0 0 ∼ 0 0 0 One can see, that due to the presence of an imaginary addition ω we get Imδ when r 0. 0 0 →±∞ → Again we come to the boundary condition: √rJ (kr) if ω >0 lim Φ(r)= ν+ (13) r0→0 (cid:26) √rJν−(kr) if ω <0 where ν = 1/4 α ± ± − 2 For the large argument this function behaves like: p Φ cos(z ν π/2 π/4) ∼ − ± − The corresponding scattering phase is: iπ δ = ± α 1/4+π/4 (14) 2 2 − p As one can see, the S-matrix S = exp(2iδ) is energy independent. This means that the regularized inverse square potential supports no bound states. The regularized wave-function and the phase-shift are analytical functions of α in the whole complex plane with a cut along the axis Reα >1/4. 2 2 B. Critical strength of inelastic interaction Now we wouldlike to determine ”how strong”shouldbe annihilationto regularizethe attractive singularpotential of order s. In other words we would like to find the minimum order of singularity of an infinitesimal imaginary potentialrequiredfortheregularizationofgivensingularpotential. Thepotentialofinterestisasum α /rs iω/rt. s − ∓ Here we keep α real. From expressions (4, 5) one immediately comes to the conclusion that the regularization is s possible only if t>s/2+1. The above statement makes clear the physical sense of suggested regularization. The scattering is insensitive to any details of the short range modification of a singular interaction if the inelastic component of such an interaction behaves more singular than 1/r(s/2+1). − 4 C. Singular potential and WKB approximation. The WKB approximation holds if ∂(1/p) 1. In case of the zero-energy scattering on a singular potential with | ∂r | ≪ s > 2 this condition is valid for r rsc (2√αs/s)2/(s−2) , i.e. near the origin. (For s = 2 the semiclassical ≪ ≡ approximation is valid only for α 1). The WKB approximation, consistent with the boundary condition (7) for ≫ s>2 is : a 1 Φ= exp( i p(r)dr) (15) p(r) ± Z r p with p(r) from (5). It follows from the above expression, that in case the WKB approximation is valid everywhere the solution of the Schr¨odinger equation includes incoming wave only (for distinctness we speak here of absorptive potential). The corresponding S-matrix is equal to zero S = 0 within such an approximation and insensitive to any modifications of the inner part of potential p2(r). The outgoing wave can appear in the solution only in the regions where(15)doesnothold. Forexample,inthezeroenergylimitk2 0theS-matrixisnonzeroS =1 2ika. Onecan → − showthatthe outgoingwaveisreflectedfromthosepartsofthe potentialwhichchangesufficientlyfastincomparison with the effective wavelength (so called quantum reflection) ∂(1/p) &1 | ∂r | For the zero energy scattering and l=0 this holds for r r . sc ≥ The reflectioncoefficient P S 2 =exp( 4Imδ) which shows the reflectedpart ofthe flux has the following form ≡| | − in the low energy limit: P =1 4kIma − For the energies E E (s/2)2s/(s−2)α−2/(s−2) the WKB holds everywhere and S-wave reflection becomes sc s ≫ ≡ exponentially small. An important conclusion is that any information, which comes from the scattering on an absorptive singular potential is due to a quantum reflection from the tail of such a potential. III. OPTICAL MODEL OF NN INTERACTION. From the above results it is clear that the model potential, which behaves at short distance like (α+iω)/r3 is − regular,i.e. it enablesdefinite unique solutionof the scatteringproblem. Sucha potentialis absorptiveanddescribes not only elastic, but inelastic scattering as well. We suggest the following model potential: A W =V i exp( r/τ) (16) KW − r3 − here V is the real part of Kohno-Weise version of OBE potential[4], but without any cut-off. The parameters of KW imaginarypartofthepotentialwerechoosedasfollows: A=4.7GeV fm2,τ =0.4fm. We havecalculatedthe values of S- and P-scattering lengths in our model potential. The obtained results, (indicated as Reg) together with the results of two Dover-Richardmodels, (DR1 and DR2), and Kohno-Weise model (KW) are presented in the table. State DR1 DR2 KW Reg 11S 0.02-i1.12 0.1-i1.06 -0.03-i1.35 -0.08-i1.16 0 31S 1.17-i0.51 1.2-i0.57 1.07-i0.62 1.05-i0.55 0 13S 1.16-i0.46 1.16-i0.47 1.24-i0.63 1.19-i0.64 1 33S 0.86-i0.63 0.87-i0.67 0.71-i0.76 0.7-i0.65 1 11P -3.33-i0.56 -3.28-i0.78 -3.36-i0.62 -3.19-i0.59 1 31P 0.92-i0.5 1.02-i0.43 0.71-i0.47 0.81-i0.46 1 13P -9.58-i5.2 -8.53-i3.51 -8.83-i4.45 -7.67-i4.74 0 33P 2.69-i0.13 2.67-i0.15 2.43-i0.11 2.46-i0.15 0 13P 5.16-i0.08 5.14-i0.09 4.73-i0.08 4.75-i0.15 1 33P -2.08-i0.86 -2.02-i0.7 -2.17-i0.95 -2.09-i0.79 1 13P 0.04-i0.57 0.22-i0.56 -0.03-i0.88 -0.12-i0.82 2 33P -0.1-i0.46 0.05-i0.55 -0.25-i0.39 -0.14-i0.39 2 5 Onecanseerathergoodagreementbetweentheresultsobtainedwithinthesuggestedopticalmodelwithoutcut-off and cited above versions of Kohno-Weise and Dover-Richardmodels. IV. CONCLUSION We have found that the scattering observables are insensitive to any details of the short range modification of a singular interaction, if such an interaction includes strong inelastic component. In this case the scattering amplitude can be calculated by solving the Schr¨odinger equation with the regularized singular potential (α i0)/rs. It was s − ± shownthatthelowenergyscatteringamplitudeonsuchapotentialisdeterminedbythequantumscatteringfromthe region, where WKB approximation fails ( the potential tail). The mentioned formalism was used to build an optical model of NN¯ low energy interaction free from uncertainty, related to the cut-off parameter. The good agreement of the results, obtained within our regularization method and within different NN¯ interaction models proves that the long range part (pionic tail) of OBEP plays essential role for the nearthreshold scattering. V. ACKNOWLEDGEMENT The researchwas performed under support of Russian Foundation for Basic Research grant 02-02-16809. [1] R.J.N. Phillips, Rev.Mod. Phys.39 (1967) 681 [2] O.D.Dalkarov, F. Myhrer, Il NuovoCimento 40A (1977), 152 [3] J.Carbonell, O. Dalkarov,K. Protasov, I. Shapiro, Nucl.Phys.A535(1991),651 [4] M.Kohno, W. Weise, Nucl. Phys. A454(1986) 429 [5] Dover,C.B., Richard J.-M.: Phys. Rev. C21,1466(1980) [6] M. Pignone et. al. Phys. Rev. C50(1994), 2710 [7] I.S.Shapiro, Phys. Rep. 35C (1978) 129 [8] L.D. Landau and E.M. Lifshitz, Quantum Mechanics,Pergamon Press, N.Y.1977 [9] J.Carbonell,RevistaMexicanadeFisica47(2001)70-77,Selected topics onLow Energy Antiproton Physics, e-printarXiv: nucl-th/0103043 [10] A.Yu.Voronin, Phys.Rev.A67,062706 (2003) [11] M.F. Mott, H.S.W. Massey, The theory of atomic collisions, Oxford, At The Clarendon Press, 1965 [12] G.N. Watson,A treatise on the theory of Bessel functions, Cambridge UniversityPress, 1922 [13] K.Meetz, Nuov.Cim.,34(1964),690 [14] W.M.Frank, D.J. Land, R.M. Spector, Rev.Mod. Phys. 43,36 (1971) [15] A.M. Perelomov, V.S.Popov, Theor. Mat. Phys.(USSR)4(1970),664 [16] E. Vogt, G.M. Wannier,Phys.Rev. 95 (1954), 1190 [17] V.D. Skarzhinsky, J. Audretsch, Singular potentials and absorption problem in quantum mechanics, arXiv:quant-ph/0012004 (2000)