Nuclear Physics A 833 (2010) 104–118

www.elsevier.com/locate/nuclphysa

On the electromagnetic form factors of hadrons

in the time-like region near threshold
O.D. Dalkarov, P.A. Khakhulin, A.Yu. Voronin ∗
P. N. Lebedev Physical Institute, Moscow, Russia
Received 23 June 2009; received in revised form 12 November 2009; accepted 30 November 2009
Available online 2 December 2009

Abstract
Hadron electromagnetic form factor in the time-like region at the boundary of the physical region is con-
sidered. The energy behavior of the form factor is shown to be determined by the strong hadron–antihadron
interaction. We propose a simple phenomenological model of the hh̄ interaction that describes the form fac-
tor behaviour in the region where scattering length approximation fails and estimate on its basis imaginary
part of the scattering lengths (volumes) for ΛΛ̄, ΛΣ̄ 0 (Λ̄Σ 0 ), Σ 0 Σ̄ 0 and D ∗ D¯∗ from present experimental
data. The developed approach enables us to analyze the existence of baryon–antibaryon (B B̄) near thresh-
old resonances. The experiments to extract detailed information on the near threshold B B̄ interaction from
hadron form factor energy behavior are suggested.
© 2009 Elsevier B.V. All rights reserved.

Keywords: Form factor; Antibaryon

1. Introduction

The main goal of hadron electromagnetic form factor studies is to obtain information about
hadron structure. The most complete data on the behavior of the form factor as a function of
four-momentum transferred is obtained for pion and nucleon. To investigate form factor of
hadron (h) two reactions are used: the reaction of elastic scattering of electrons by hadron
eh → eh (so-called space-like region of the four-momentum transferred) and the reaction of

* Corresponding author.
E-mail address: alexei.voronin@excite.com (A.Yu. Voronin).

0375-9474/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.nuclphysa.2009.11.015
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 105

hadron–antihadron pair production in the electron–positron annihilation e+ e− → hh̄ or inverse

reaction (time-like region).
High precision data on the electromagnetic form factor of a proton in the time-like region
became available due to PS-170 experiment performed at LEAR (CERN) [1] and recent BaBar
experiment, where the reactions e+ e− → p p̄, ΛΛ̄, ΛΣ̄ 0 , Σ 0 Σ̄ 0 were reported [2,3]. BES Col-
laboration [4] has observed the near threshold enhancement of the p p̄-system in the J /ψ → γp p̄
decay. These data demonstrate principally different behavior of the electromagnetic baryon form
factor in the time-like region from that of a pion. The proton form factor drops quickly with
momentum increasing from the threshold (approximately two times in the range of c.m. relative
momentum from zero to 100 MeV/c).
To describe the behavior of the form factor different vector dominance models (VDM) were
suggested [5–9]. Such models tend to describe form factor behavior in rather wide energy range
and remain phenomenological. In order to clarify the physical reason of certain form factor
behavior near threshold we follow the approach which was proposed in [10–13]. Namely, the
energy behavior of the baryon electromagnetic form factor in the time-like region just near
B B̄ threshold is explained by the final state baryon–antibaryon interaction. The steep proton
electromagnetic form factor behavior was explained in terms of final state interaction also in
recent papers [32–38]. In such an approach the form factor is presented as a product of a fac-
tor corresponding to singularities of transition amplitude lying far from B B̄ threshold and a
factor reflecting strong final state interaction. The latter gives the near threshold energy depen-
dence of the form factor. Moreover, it was found that the imaginary part of baryon–antibaryon
scattering length could be directly extracted from the form factor energy dependence near B B̄
threshold. This approach is a natural consequence of a quasi-nuclear model of low energy
baryon–antibaryon interaction [14–18,22,23]. The most intriguing prediction of such a model
is the existence of near threshold baryon–antibaryon resonances. Such resonances would provide
very steep energy behavior of the form factor just near threshold. In the present paper we de-
velop a model which accounts for asymptotic properties of B B̄ final state interaction. Such an
approach enables us to reproduce the form factor energy behavior beyond the scattering length
approximation and analyze the existence of near threshold resonances.
The advantageous type of experiments to study baryon form factor are the baryon–antibaryon
pair production in electron–positron collisions as well as electron–positron annihilation into
multipion systems (which are dominant modes of nucleon–antinucleon annihilation). Indeed,
the initial state interaction is negligible in such processes and transition mechanism is well-
determined. Developed approach enables us to analyze existing data on the Y Ȳ production to
extract important characteristics of Y Ȳ interaction.
The article is organized in the following way. Section 2 is devoted to the general properties
of the form factor in the time-like region and the case of nucleon form factor. In Section 3 we
apply developed formalism to form factor properties of other hadrons. Conclusion and proposals
of new experiments are presented in Section 4.

2. General properties of the form factor

In this section we investigate general properties of the hadron form factor in the time-like
region near the boundary of the physical region. For different hadrons we can have different
number of form factors, depending on hadron spin. We will formulate the main results for a
proton and neutron, the generalization for other hadrons will be done in Section 4.
106 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

Fig. 1. The diagram corresponding to electron–positron annihilation into the nucleon–antinucleon system.

The form factor of the nucleon (N ) in the time-like region is determined from the reaction
of e+ e− -annihilation into nucleon–antinucleon pair e+ e− → N N̄ or vice versa (the so-called
s-channel). The nucleon form factor in the space-like region is given by the t-channel corre-
sponding to eN → eN scattering.
The differential cross-section dσ/dΩ of the reaction p p̄ → e+ e− is connected with the form
factor of the proton by the following expression [30]:
 
dσ αfs2   4Mp 2
= |GM |2 1 + cos2 θ + |GE |2 sin2 θ , (1)
dΩ 16kE s
here k and E are center-of-mass (c.m.) momentum and energy in the p p̄ system, θ is angle in c.m.
system, αfs is the fine structure constant, Mp is the proton mass, s = −q 2 , q is four-momentum
transferred GE and GM are electric and magnetic form factors of the proton correspondingly.
They are connected to Pauli form factors F1 and F2 :
q2
GM = F1 + F2 , GE = F1 − F2 . (2)
4Mp 2
At the p p̄ threshold GE and GM are equal and for simplicity hereafter they are taken to be
equal in the kinetic energy region of few tens of MeV near the threshold.
Before doing any calculations we can make some conclusions about nucleon form factor near
threshold. Let’s consider a diagram corresponding to the process e+ e− → N N̄ (Fig. 1). Grey
block in this diagram presents the strong final state interaction in the system N N̄ . This interaction
is responsible for the near threshold N N̄ resonances [14–18,26]. The diagram corresponds to
the transition amplitude from e+ e− pair into N N̄ pair. Black circle in this transition amplitude
denotes a connection between a photon and N N̄ pair, which can be realized, for example, by
vector mesons (ρ or ω).
Since N N̄ annihilation is short-ranged the form factor can be factorized as follows:
 
G = G0 Al (k). (3)
Here Al (k) is connected to the amplitude of the normalized wave-function near origin (r = 0):
Ψl (r) ∼ Al (k)(kr)l . (4)
The factor G0 corresponds to singularities far from N N̄ threshold (for instance, to a connec-
tion of the photon to the nucleon through ρ- or ω-exchanges, as it can be written in usual vector
dominance models) and is practically constant in the kinetic energy region of few tens of MeV in
the N N̄ center of mass system. The factor Al (k) reflects the influence of strong final state inter-
action and contains main energy dependence of the form factor (3) in the near threshold region.
The general form of the energy behavior is given by the following expression (see Appendix A):
exp(− Im δ(k))
G = G0 . (5)
τ (k 2 )
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 107

Here δ(k) is the N N̄ scattering phase, τ (k 2 ) is a function of k 2 , determined by the properties of

the N N̄ potential. In the case of s-wave scattering A0 (k) ∝ Ψ (0) and Eq. (3) becomes identical to
the one obtained in [10]. In the region just near threshold G(k) (5) has the form (see Appendix A):
exp(Im α0 k)
G = G0 ≈ const(1 + Im α0 k), (6)
τ (0)
where α0 is the N N̄ s-wave scattering length. The appearance of a linear in k term is a direct
manifestation of the threshold behavior and the final state inelastic interaction. Expression (6)
can be used to extract the scattering length imaginary part from the form factor experimental data
just near threshold.
For very small k the account of the Coulomb corrections to the final states interaction becomes
important (see Appendix A). For the sake of unified description of experimental data we include
the Coulomb corrections to the final state interaction in the definition of the form factor (3):
   
A0 (k) = const C 1 + C 2 Im α0 k . (7)
Here C is the Gamow factor
2π/k|ac |
C2 = (8)
1 − exp(−2π/k|ac |)
and |ac | = 1/(Mαfs ) is the Coulomb length, αfs is fine-structure constant and M = Mp /2 is p p̄
reduced mass. The Coulomb corrections become important when 2π/k|ac | > 1, for p p̄ system
it means k < 20 MeV/c.
In the case of p-wave the A1 (k) factor near threshold is given by the following expression
(Appendix A):
  3  
A1 (k) = const exp(Im α1 k )  const 1 + Re ck 2 + Im α1 k 3 . (9)
2
τ (k )
Here α1 is the scattering volume, c is k-independent coefficient.
The most precise experimental data on the N N̄ form factor (s-wave) belong to the energy
region, where the scattering length approximation fails (kα0 ∼ 1) and expression (6) cannot be
used. In such a region it is not possible to get model-independent information about the prop-
erties of N N̄ potential. However, approximative account of the asymptotic properties of N N̄
interaction is still possible within comparatively simple models. Within such an approach it is
possible to reproduce the form factor energy behavior in the mentioned region in the closed
form. We suggest a simple phenomenological model of N N̄ interaction based on the following
assumptions. First, the tail of N N̄ potential is considered real and attractive and above a certain
matching interbaryonic distance Rc it can be approximated by exponential potential:
 
VN N̄ (r > Rc ) = −U0 exp −(r − Rc )/ρ (10)
here ρ is a free diffuseness parameter.
Second, the effective depth of N N̄ interaction below Rc is much greater than a collision en-
ergy of interest. It means that at matching distance Rc the N N̄ non-relativistic wave-function
Ψ (R) obeys the energy-independent boundary condition, which can be conveniently parameter-
ized in the following way:
z(Rc )
(rΨ ) /(rΨ )|r=Rc = cot(Φ0 ). (11)
2ρ
√
Here z(R) = 2ρ 2MU0 exp(−(R − Rc )/(2ρ)), Φ0 is a free parameter.
108 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

The benefit of such a model is due to the possibility to get the s-wave form factor expression
in the closed form (see Appendix B for the details):

 
A0 (k) = const sinh(2πkρ) 1
. (12)
2πkρ | cos(Φ + iπkρ)|
Here we introduced phase parameter Φ ≡ Φ0 + z(Rc ) − π/4. The above expression is justified
when z(Rc )  1 and |U0 |  k 2 /(2M).
Let us consider the main properties of the above introduced model. The model includes five
parameters, namely the potential depth U0 , matching distance Rc , the potential tail diffuseness ρ,
real and imaginary parts of the complex phase Φ. However, the expression for the form factor
slope equation (12) is described only by ρ and the complex phase Φ. (One of five model param-
eters is absorbed in the constant in front of right-hand side of Eq. (12), another disappears due to
the fact that Eq. (12) gives the absolute value only.) In the limit of small k we return to Eq. (6)
with the imaginary part of the scattering length given by the following expression:
Im α0 = −πρ Im tan(Φ). (13)
In case Im Φ  1 (strong absorption by the inner part of N N̄ interaction) the form factor
becomes insensitive to parameter Φ. In such a case the imaginary part of the scattering length
turns to be:
Im α0 = −πρ (14)
while the form factor slope is determined by only one parameter:

 
A0 (k) = const sinh(2πkρ) exp(−πkρ). (15)
πkρ
In the opposite case of weak absorption Im Φ  1 the form factor is sensitive to the phase Φ,
especially when Re Φ → π/2. This corresponds to the appearance of a quasi-bound N N̄ state
close to the threshold. In such a resonant case the value of the imaginary part of the scattering
length can be much greater than the diffuseness of potential tail:
1
Im α0 ≈ −πρ .
Im Φ
The form factor in the resonant case turns to be:

 
A0 (k) = const sinh(2πkρ) 1
.
2πkρ sinh(Im Φ + πkρ)
The important property of suggested model is the ability to describe the form factor energy
behavior beyond the scattering length and effective range approximations i.e. for kα0  1. In the
limit 1  kρ  z(Rc ) the form factor behaves like:
 
A0 (k) = const √ 1 .
πkρ

√ from fast exponential decay exp(Im α0 k) of the

So far the above model describes the transition
form factor just near the threshold to the 1/ πkρ behavior away from the threshold. Such a
transition takes place when kπρ ∼ 1. The experimental data fit in such a transition region enables
us to extract diffuseness parameter ρ, while the data fit near the threshold gives the imaginary
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 109

Fig. 2. Proton electromagnetic form factor in the time-like region. Experimental data are taken from [1,2]. Solid and
dotted lines are calculations using optical model [26] with and without Coulomb corrections. Dash-dotted line represents
the data fit by the model equation (12).

part of the scattering length and so far the phase parameter Φ. This property of the model gives an
opportunity to distinguish the near threshold resonance in the form factor behavior, characterized
by the imaginary part of the scattering length much larger than the potential tail diffuseness
| Im α0 |  πρ. However, unique determination of the model parameters is possible only in case
of highly accurate experimental data.
We demonstrate the applicability of this model to the case of p p̄ s-wave form factor. As far
as the properties of low energy p p̄ interaction are rather well established p p̄ case could be used
as a benchmark. Precise experimental data on the proton form factor near the threshold were
obtained in the LEAR [1] and BaBar experiment [2]. These data are presented in Fig. 2. We
note the decrease of the form factor in two times with increase of relative momenta from zero
to 100 MeV/c. We perform comparison between the numerical solution of the Schröedinger
equation with recent modification of N N̄ Paris potential [26] and the model form factor given
by Eq. (12) with parameters fitted from the LEAR [1] experimental data.
The p p̄ scattering length extracted from the fit of LEAR experimental data using Eq. (12)
after taking into account Coulomb correction is Im α = −(0.61 ± 0.07) fm. This value is in
agreement with the value obtained within N N̄ potential model [26]. The BaBar data show very
similar form factor energy slope and can be reproduced by the same model with normalization
constant multiplied by 1.2 factor.
One can see that the model form factor (12) reproduces the experimental data up to k =
300 MeV/c, where both the scattering length and effective range approximation fails. However,
the accuracy of the present data does not allow to determine uniquely the set of model parameters
ρ and Φ. We found, in particular, that the diffuseness parameter ρ extracted from the data fit can
have values in the range 0.1 fm  ρ  0.45 fm. We note that realistic OBE p p̄ potential is a
superposition of Yukawa-type potentials with different diffuseness and depth. One can check
that the potential tail, important for the observables low energy behavior, can be satisfactory
110 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

Fig. 3. Paris potential 3 S1 T = 0 state compared to model potential with ρ = 0.31 fm, Rc = 1.2 fm, U0 = 84 MeV.

approximated by the effective exponential tail with ρ ≈ 0.3 fm [25]. In Fig. 3 we show the
version of Paris potential [26] for T = 0 3 S1 state compared to model potential with diffuseness
ρ = 0.31 fm, Rc = 1.2 fm, U0 = 84 MeV. The corresponding phase parameter for ρ = 0.31 fm
is Φ = 0.6 + i0.5.
We conclude that the suggested model reproduces the form factor slope up to k = 300 MeV/c
and enables us to extract the scattering length using the experimental data set away from the
threshold.
By using experimental data on the proton form factor we can predict a value of the neutron
form factor near the threshold. The definition of the nucleon form factor in terms of the isoscalar
(isospin is equal to 0) G(0) and isovector G(1) (isospin 1) form factors is:
   
Gp = G(0) + G(1), Gn = G(0) − G(1). (16)
We can use this decomposition and express Gp and Gn in terms of input form factors G0 and
wave-functions of final state in pure isospin states:
 
Gp = G0 (0)Ψ0 (0) + G0 (1)Ψ1 (0), (17)
 
Gn = G0 (0)Ψ0 (0) − G0 (1)Ψ1 (0). (18)
Depending on the relative values of G0 (0) and G0 (1) there are two possible cases.
First, G0 (0) ≈ G0 (1), i.e. there is no sufficient difference between isoscalar and isovector
input form factors. In this case in the previous formulas we can take out of brackets the common
factor G0 . We see immediately that neutron form factor Gn is equal to or less than proton one.
Note that in this case the energy behavior of the neutron form factor can be different from the
proton one. Just near threshold it can be practically constant or even decreasing function of the
energy. The calculations of the neutron form factor using optical model [26] in this case are
presented in Fig. 4. Note that the close predictions were done in coupled channels model in
Refs. [11–13]. Experimental data [19–21,24] indicates on realization of this possibility.
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 111

Fig. 4. Neutron form factor in the time-like region. Experimental data from [19]. Solid line is our calculation using optical
model [26] with normalization at k = 200 MeV/c.

Second, one of the input form factors is dominant, for instance, G0 (1)  G0 (0). This is
possible if the form factors are determined by ρ- and ω-mesons correspondingly. In this case
G0 (1) dominates because ρ-meson has a product of coupling constants with nucleon and photon
larger than that for ω-meson. So one can neglect G0 (0) contributions to the nucleon form factor
and get that proton and neutron form factors are approximately equal.
Therefore, in our model the neutron form factor does not exceed proton one in any case. The
obtained result differs from predictions of early VDM models which give neutron form factor
sufficiently larger than that of the proton.

3. Form factors of other hadrons

The consideration presented above can be directly applied to investigation of the form factor
of any hadron in time-like region.
The closest threshold to the N N̄ one is a threshold of ΛΛ̄ production in e+ e− -annihilation.
Let’s estimate the value of the lambda form factor near ΛΛ̄ threshold under the following
assumptions. We consider the ΛΛ̄ pair production mechanism via a vector meson (with dominant
contribution of ρ-meson). The value of G0 for lambda is proportional to the coupling constants of
ρ meson with photon and ρ-meson with lambda. The former is known from the experiment. The
latter can be estimated from the SU(3)-relations. Both are of the same order as for a nucleon.
Final state interaction in ΛΛ̄ system (in pure isospin (I = 0) state) according to the existing
approaches is approximately the same as in case of N N̄ . So we expect lambda and nucleon form
factors have the same order of magnitude. Interpretation of the above relations on the quark level
can be found in [31].
Latest data on ΛΛ̄, Σ 0 Σ̄ 0 and ΛΣ̄ 0 form factors became available in experiments of BaBar
group [3]. (Note, that in cited paper the form factor is denoted by |F |.) In Fig. 5 we present a
112 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

Fig. 5. Λ form factor near ΛΛ̄ threshold. Experimental points are taken from [3]. Solid line corresponds to the fit by
model equation (12), dashed line corresponds to the exponential extrapolation of the near threshold form factor behavior
∼ exp(Im α0 k).

typical fit of experimental data by model equation (12) (model parameters are Φ = 1.7 + i0.4
and ρ = 0.25) and by exponential extrapolation of the near threshold form factor behavior
∼ exp(Im α0 k). One can see that the model equation (12) curve and “scattering length” extrapo-
lation differ significantly for k  100 MeV.
The present data (though insufficient) does not exclude the possibility of ΛΛ̄ resonance. Such
resonances are found in our model as complex k poles of expression (33). The approximative
solution of Eq. (33) results in a simple expression for S-matrix near threshold poles:

π/2 − Φ
k . (19)
iπρ

As soon as the model parameters Φ and ρ are established from experimental data fit the above
equation enables us to determine the position of the nearest to the threshold S-matrix pole. We
found, that the narrow quasi-bound near threshold state [28] is consistent with present ΛΛ̄ data.
In particular, the fit with Φ = 1.7 + i0.4 and ρ = 0.25 (shown in Fig. 5) gives Im α = −1.9 fm
and quasi-bound state with energy E = 8 − i6 MeV. However, more precise experimental data
are required for determination of the scattering properties and resonance position with much
higher significance.
Experimental data for Σ Σ̄ and ΛΣ̄ and their typical fits by expression (12) as well as by
exponential curves |F | ≈ C exp(Im αk) are presented in Fig. 6.
The extracted scattering lengths values are presented in Table 1.
Let us turn to the case of the D ∗ D¯ ∗ system. We apply the same ideas in order to estimate
the corresponding scattering volume. Experimental data became available due to experiments by
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 113

Fig. 6. Baryon effective form factors. Experimental data are taken from BaBar experiment [3]. Solid and dotted curves
are ΛΣ̄ 0 (Λ̄Σ 0 ) and Σ 0 Σ̄ 0 experimental data fits by Eq. (12) correspondingly. Dash-dotted curves are the extrapolation
of near threshold form factor behavior by means of exponential curves ∼ exp(Im α0 k).

Table 1
Scattering lengths in B B̄ system.
Im α, fm ρ, fm Im Φ
ΛΛ̄ −1.3 ± 0.6 0.4 ± 0.3 0.7 ± 0.3
ΛΣ̄ 0 (Λ̄Σ 0 ) −2.4 ± 0.7 0.7 ± 0.2 1.0 ± 0.1
Σ 0 Σ̄ 0 −2.5 ± 0.8 0.6 ± 0.2 0.8 ± 0.2

CLEO-c [39]. D ∗ D¯ ∗ production cross-section is usually parameterized by the following expres-

sion [27]:

14παfs2
σ= 3
vD ∗ R4 . (20)
3m2
The value R4 is proportional to the form factor squared. We apply the model equations (10),
(11) and obtain the corresponding form factor numerically. Fit results are presented in Fig. 7. We
have found the best fit with ρ = 0.4 fm, Im α1 = −0.44 fm3 . In spite of rather good agreement
between theoretical fit and experimental data the obtained value of imaginary part of scattering
volume should be treated only as an estimation due to the lack of CLEO-c data in the near
threshold region. More detailed data are required to distinguish possible resonance behavior of
the mentioned form factor.
It follows from our consideration that fast decrease of the hadron form factor with increasing
momentum from the threshold is the consequence of the absorption in the final state interaction.
Let us mention, that such a decay can not be seen in the pion form factor, because π + π − system
has only elastic scattering and has no absorbtion at the threshold.
114 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

Fig. 7. Function R4 [27] fit by numerical solution of the Schrödinger equation within model equations (10), (11).

4. Conclusion

The analysis of the experimental data shows that the near threshold behavior of the elec-
tromagnetic form factor of a hadron is mostly determined by the interaction of the hadron–
antihadron in the final state. Therefore the measurements of the form factor properties can serve
as a fruitful source of information about hadron–antihadron interaction, especially in situation
when direct investigation of this interaction is unavailable. We suggested the phenomenological
model which is based on the account of asymptotic properties of hadron–antihadron final state
interaction. Such a model reproduces typical form factor energy behavior in the range of relative
hh̄ momenta, where the scattering length and effective range approximations fail. This property
enables us to extract the scattering lengths and characteristic diffuseness radius of hh̄ potentials
from existing experimental data. Moreover, one can establish the existence of the near thresh-
old resonances in the mentioned systems. However, more accurate near threshold form factor
experimental data are required to significantly improve the significance of the obtained results.
To obtain more elaborate information about hadron–antihadron interaction the following ex-
periments could be suggested:

• Precise measurements of the proton and neutron form factors in the time-like region just near
threshold of the reaction e+ e− → N N̄ give us opportunity of high quality determination of
N N̄ scattering parameters.
• Further investigation of the strange and charm particles electromagnetic form factors in the
time-like region in non-relativistic region of relative momentum less than few hundreds of
MeV/c.
• There is a possibility to discover a 3 S1 quasinuclear bound states in the B B̄ system, which
can manifest themselves as a heavy vector meson. To do it, the experiment to measure proton
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 115

form factor near B B̄ threshold is desirable, because these states will manifest themselves as
the bumps in the form factor behavior (e+ e− → B B̄ → p p̄).
• Bound states with photon quantum numbers in hadron–antihadron systems will manifest
themselves also as a deep-bump structure in the electron–positron transition into main an-
nihilation channels of these systems. In particular, it will be interesting to search for phe-
nomena connected with such a vector meson state in the ΛΛ̄ system near threshold in the
reaction e+ e− → K K̄4π by the analogy with deep-bump structure in 6π annihilation chan-
nel near N N̄ threshold. Note that the existence of quasinuclear 3 S1 vector state just near ΛΛ̄
threshold was considered in [28].
• The experiments on precise measurement of the cross-sections e+ e− → K K̄, D D̄, F F̄ ,
b+ b̄+ , b− b̄− etc. near corresponding thresholds can be very informative to determine inter-
action in the systems with hidden new quantum numbers.

Appendix A. Derivation of the form factor

We present here the derivation of the form factor in terms of the Jost function. The form factor
is proportional to the Al (k) that is defined from the following relation
ψl (r; k) = Al (k)φl (r; k), (21)
where ψl and φl are solutions of the Shrödinger equation such that
i 
ψl (r; k) ∼ ĥ− (kr) − Sl (k)ĥ+
l (kr) , (22)
2 l
as r → ∞,
φl (r; k) ∼ jˆl (kr), (23)
as r → 0. Here Sl is the partial S-matrix element, jˆl (kr), ĥ±
l are Riccati–Bessel and Riccati–
Hankel functions:
ĥ± ˆ
l ≡ n̂l (z) ± i jl (z),

πz
jˆl (z) ≡ Jl+1/2 (z),
2

l πz
n̂l (z) ≡ (−1) J−l−1/2 (z).
2
From the definition of the Jost function fl (k) [29] we obtain:
Al (k) = 1/fl (k). (24)
Now we can use the well-known Jost function expansion near threshold [29] to get the corre-
sponding expansion of Al (k).
Particularly for l = 0
A0 (k) ≈ const · (1 − iα0 k), (25)
and for l = 1
 
A1 (k) ≈ const · 1 + ck 2 − iα1 k 3 . (26)
Here c is the momentum-independent constant, while α0 and α1 are the scattering length and
scattering volume correspondingly.
116 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

In the presence of Coulomb force we have

   
A0 (k) = const · C 1 + C 2 Im αk . (27)
Here
2π/kac
C2 =
exp(2π/kac ) − 1
is the Gamow factor, ac = 1/(Mαfs Z1 Z2 ) is the Coulomb length, αfs is fine-structure constant
and M is the reduced mass of a system.
In the case of p p̄ system Coulomb force is attractive so ac = −1/(Mαfs ). The Gamow factor
then becomes
2π/k|ac |
C2 = .
1 − exp(−2π/k|ac |)
For small c.m. momentum 2π/k|ac |  1 it has the form
2π
C2 = ,
k|ac |
and from (27) it follows that
  
  2π 2π
A0 (k) = const · 1+ Im α . (28)
k|ac | |ac |
Comparing (27) with (25) we see that the effect of Coulomb interaction becomes negligible
as C 2 ≈ 1, i.e. when 2π/k|ac |  1. For p p̄ system this gives k  20 MeV/c.

Appendix B. “Exponential tail” model

Here we derive the form factor for the following model of N N̄ interaction. We assume that
N N̄ potential V (r) can be approximated above some matching distance Rc by an attractive
exponential potential V (r > Rc ) ≈ −U0 exp(−(r − Rc )/ρ). Below the matching distance r < Rc
the N N̄ interaction has a deep inner-core such that collision energy can be neglected |V (r)| 
k 2 /(2M).
It is possible to obtain the energy dependence of the form factor under these conditions
without any further assumptions. For r < Rc the regular wave-function φ(r) satisfies boundary
conditions:
φ(0) = 0 (29)
and
φ  (0) = 1 (30)
and its k-dependence can be neglected. It means that at matching distance Rc the continuity
condition for the logarithmic derivative of the wave-function is also energy-independent. We
will parameterize it in the following way:
z(Rc )
φ  (r)/φ(r)|r=Rc = cot(Φ0 ). (31)
2ρ
 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118 117

√
Here z(R) = 2ρ 2MU0 exp(−(R − Rc )/(2ρ)), M is the N N̄ reduced mass, Φ0 is a free pa-
rameter. For r > Rc the wave-function is given by the solution of the Schrödinger equation with
exponential potential V (r) = −U0 exp(−(r − Rc )/ρ):
   
φ(r; k) = B1 (k)J2ikρ z(r) + B2 (k)J−2ikρ z(r) . (32)
Using continuity of the wave-functions and its derivative at r = Rc we get for the form factor
sinh 2πkρ
A0 (k) = const · (1 + 2ikρ)(zc /2)−2ika
2πkρ
 

−1
× (zc /2) J−2ikρ (zc ) cos Φ0 + J−2ikρ (zc ) sin Φ0 , (33)
here zc ≡ z(Rc ). If we additionally assume that |zc |  1, which means that the wave-function
can be approximated by it’s WKB form near Rc and taking into account the Bessel function large
argument behavior:
 
1 π
J2ikρ (2x) ∼ √ cos 2x − iπkρ − , (34)
πx 4
we finally get the simplified expression:
sinh(2πkρ) (1 + 2ikρ)(zc /2)−2ikρ
A0 (k) = const · , (35)
2πkρ cos(Φ0 + zc + iπkρ − π/4)
and its absolute value:

  sinh 2πkρ 1
A0 (k) = const · , (36)
2πkρ | cos(Φ + iπkρ)|
where Φ = Φ0 + zc − π/4. One can see that the phase Φ0 has a sense of phase accumulated
in the region r < Rc . This explains our choice of the boundary condition parametrization (31).
Expanding the expression for the form factor in terms of k we get for the scattering length:
α0 = Rc + 2ρ ln(zc /2) + 2γρ − πρ tan(Φ). (37)
Here γ is the Euler constant. The imaginary part of the scattering length turns to be the function
of ρ and Φ only:
Im α0 = −πρ Im tan(Φ). (38)

References

[1] PS-170 Collaboration, G. Bardin, et al., Nucl. Phys. B 411 (1994) 3.

[2] BaBar Collaboration, B. Aubert, et al., Phys. Rev. D 73 (2006) 012005.
[3] BaBar Collaboration, B. Aubert, et al., Phys. Rev. D 76 (2007) 092006.
[4] BES Collaboration, J.Z. Bai, et al., Phys. Rev. Lett. 91 (2003) 022001.
[5] J.G. Korner, M. Kuroda, Phys. Rev. D 16 (1977) 2165.
[6] S. Dubnicka, Nuovo Cimento A 103 (1991) 1417.
[7] G.Y. Chen, H.R. Dong, J.P. Ma, Phys. Rev. D 78 (2008) 054022, arXiv:0806.4661 [hep-ph].
[8] E.L. Lomon, Phys. Rev. C 66 (2002) 045501.
[9] F. Iachello, Phys. Rev. C 69 (2004) 055204.
[10] O.D. Dalkarov, Pis’ma ZhETF 28 (1978) 183.
[11] O.D. Dalkarov, K.V. Protasov, Nucl. Phys. A 504 (1989) 845.
[12] O.D. Dalkarov, K.V. Protasov, Mod. Phys. Lett. A 4 (1989) 1203.
[13] O.D. Dalkarov, K.V. Protasov, JETP Lett. 49 (1989) 273.
118 O.D. Dalkarov et al. / Nuclear Physics A 833 (2010) 104–118

[14] O.D. Dalkarov, V.B. Mandelzveig, I.S. Shapiro, Nucl. Phys. B 21 (1970) 88.
[15] I.S. Shapiro, Phys. Rep. C 35 (1976) 129.
[16] O.D. Dalkarov, K.V. Protasov, I.S. Shapiro, Int. I. Mod. Phys. A 5 (1990) 2155.
[17] O.D. Dalkarov, K.V. Protasov, Phys. Lett. B 280 (1992) 117.
[18] O.D. Dalkarov, F. Myhrer, Nuovo Cimento A 40 (1977) 152.
[19] P. Gauzzi, Frascati Physics Series XV (1997) 449.
[20] A. Antonelli, et al., Phys. Lett. B 365 (1996) 427.
[21] A. Antonelli, et al., Nucl. Phys. B 517 (1998) 3.
[22] O.D. Dalkarov, V.G. Ksenzov, Pis’ma ZhETF 30 (1979) 74.
[23] O.D. Dalkarov, V.G. Ksenzov, Pis’ma ZhETF 31 (1980) 425.
[24] E. Tomasi-Gustafsson, F. Lacroix, C. Duterte, G.I. Gakh, Eur. Phys. J. A 24 (2005) 419.
[25] V. Karmanov, K. Protasov, A. Voronin, Eur. Phys. J. A 8 (2000) 429.
[26] B. El-Bennich, M. Lacombe, B. Loiseau, S. Wycech, Phys. Rev. C 79 (2009) 054001.
[27] S. Dubynskiy, M.B. Voloshin, Mod. Phys. Lett. A 21 (2006) 2779.
[28] J. Carbonell, O.D. Dalkarov, K.V. Protasov, Phys. Lett. B 306 (1993) 407.
[29] J. Taylor, Scattering Theory, University of Colorado, Boulder, Colorado, 1972.
[30] A. Zichichi, et al., Nuovo Cimento 24 (1962) 170.
[31] R. Baldini, et al., Eur. Phys. J. A 39 (2009) 315.
[32] J. Haidenbauer, H.-W. Hammer, U.-G. Meissner, A. Sibirtsev, Phys. Lett. B 643 (2006) 29.
[33] A. Sibirtsev, J. Haidenbauer, S. Krewald, U.-G. Meissner, A.W. Thomas, Phys. Rev. D 71 (2005) 054010.
[34] B. Loiseau, S. Wycech, Phys. Rev. C 72 (2005) 011001.
[35] B. Kerbikov, A. Stavinsky, V. Fedotov, Phys. Rev. C 69 (2004) 055205.
[36] D.V. Bugg, Phys. Lett. B 598 (2004) 8.
[37] B.S. Zou, H.C. Chiang, Phys. Rev. D 69 (2004) 034004.
[38] V.F. Dmitriev, A.I. Milstein, nucl-th/0607003.
[39] R. Poling, in: Proceedings of 4th Flavor Physics and CP Violation Conference (FPCP 2006), Vancouver, British
Columbia, Canada, 9–12 April 2006, p. 005, arXiv:hep-ex/0606016v2.