arXiv:1011.

2866v1 [hep-lat] 12 Nov 2010

The q-q
potential from Bethe-Salpeter amplitudes on
lattice

Yoichi Ikeda
RIKEN Nishina Center, Wako, Saitama 351-0198, Japan
E-mail: [email protected]

Hideaki Iida
RIKEN Nishina Center, Wako, Saitama 351-0198, Japan
E-mail: [email protected]
Potentials of quarkanti-quark pairs are studied from the q-q
Nambu-Bethe-Salpeter (NBS) wave
functions in quenched lattice QCD. With the use of a method which has been recently developed
in the derivation of nuclear forces from QCD, we derive the q-q
potentials with finite quark masses
from the NBS wave functions. We calculate the q-q
NBS wave functions in pseudo-scalar and
vector channels for several quark masses. The derived potentials in both channels reveal linear
confinement plus Coulomb potentials. We also discuss the quark-mass and channel dependence
of the q-q
potentials.

The XXVIII International Symposium on Lattice Field Theory, Lattice2010

June 14-19, 2010
Villasimius, Italy
Speaker.

c Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence.

http://pos.sissa.it/

The q-q
potential from Bethe-Salpeter amplitudes on lattice

Yoichi Ikeda

1. Introduction
An inter-quark potential is the fundamental interaction in strongly interacting quark-gluon
systems, which is governed by the complex dynamics of Quantum Chromodynamics (QCD). It is
still difficult to analyze such low-energy phenomena of QCD in analytic ways, because the coupling
constant becomes large at low energies and therefore the perturbation theory is not applicable.
Experimentally, the linear behavior of inter-quark potentials is suggested by Regge slope [1], which
shows the relation, J M 2 with the spin J and the mass M of hadrons. A naive estimation for the
relation between M and J is J = M 2 /4 with the string tension , and the value of is about 1.3
GeV/fm from the Regge slope. The Coulomb force of inter-quark potentials is suggested by the
analogy between quarkonium and positronium. In fact, Coulomb plus linear confinement behavior
of inter-quark potentials reproduces the quarkonium spectrum well in quark models. However,
until now, there is no regolous proof of the emergence of the linear confinement potential.
Lattice QCD simulation is the powerful tool for a numerical investigation with strong coupling
regime of strong interaction. The inter-quark potential is one of subjects which is most actively
studied on lattices. From the analyses of Wilson loops, the potential for an infinitely heavy quark
potential) can be obtained. The Q-Q
potential from lattice QCD simulations
and an anti-quark (Q-Q
reveals the form of V (r) = r A/r with = 0.89GeV/fm and A = 0.26 [2], and one can take
into account corrections coming from finite quark masses order by order with the use of the heavy
quark effective field theory such as potential nonrelativistic QCD (pNRQCD) [2, 3, 4, 5].
We study potentials between light quarks and anti-quarks (q-q
potentials) in pseudo-scalar and
vector channels from lattice QCD simulations. In order to explore the q-q
potentials, we apply
the systematic method which utilize the equal-time Nambu-Bethe-Salpeter (NBS) amplitudes to
extract hadronic potentials [6, 7, 8, 9, 10, 11, 12, 13] to systems containing relatively light quarks
and anti-quarks. Due to the absence of the asymptotic fields of quarks, the reduction formula cannot
be applied directly. Therefore, we assume that the equal-time NBS amplitudes for the q-q
systems
satisfy the Bethe-Salpeter (BS) equation with constant quark masses which could be considered as
the constituent quark masses. By using the derivation of the relativistic three-dimensional formalism from the BS equation developed by Lvy, Klein and Macke (LKM formalism) [14, 15], we
shall obtain the q-q
potentials without the expansion in terms of quark masses.
The paper is organized as follows. In Sec 2, we briefly present our method to extract the
q-q
potentials, together with the lattice QCD setup. We then show our results in Sec 3. The q-q

potentials are discussed and summarized in Sec 4.

2. Method and lattice QCD setup

Following the basic formulation to extract the nuclear force [6, 7], we briefly show how to
extract the q-q
potentials on the lattice below. We start with the effective Schrdinger equation for
the equal-time Nambu-Bethe-Salpeter (NBS) wave function (~r):
2
(~r) +
2

d~rU (~r,~r ) (~r ) = E (~r),

(2.1)

system and the non-relativistic energy,

where (= mq /2) and E denote the reduced mass of the q-q
respectively. For the two-nucleon case, it is proved that the effective Schrdinger equation is de2

The q-q
potential from Bethe-Salpeter amplitudes on lattice

Yoichi Ikeda

rived by using the reduction formula [7]. Due to the absence of the asymptotic fields for confined
quarks, we suppose that the q-q
systems satisfy the BS equation with their constant quark masses.
In this study, constant quark masses mq are determined by half of vector meson masses mV , i.e.,
mq = mV /2. Then, one finds Eq. (2.1) by applying LKM method to the BS equation.
The non-local potential U (~r,~r ) can be expanded in powers of the relative velocity ~v = / of
q-q
systems at low energies,
U (~r,~r ) = V (~r,~v) (~r ~r )
= (VLO (~r) +VNLO (~r) + ) (~r ~r ),

(2.2)

where the N n LO term is of order O(~vn ). At the leading order, one finds
V (~r) VLO (~r) =

1 2 (~r)
+ E.
2 (~r)

(2.3)

In order to obtain the NBS wave functions of the q-q

systems on the lattice, let us consider the
following equal-time NBS amplitudes

(~x +~r,~x,t t0 ; J ) = h0| q(~

x +~r,t)q(~x,t)J qq
(t0 ; J ) |0i

x +~r,t)q(~x,t) |ni eMn (tt0 ) ,

= An h0| q(~

(2.4)

with the matrix elements

An = hn| J qq
(t0 ; J ) |0i .

(2.5)

Here represents the Dirac -matrices, and J qq

(t0 ; J ) denotes a source term which creates the
q-q
systems with spin-parity J on the lattice. The NBS amplitudes in Eq. (2.4) is dominated by
the lowest mass state of mesons with the mass M0 at large time separation (t t0 ):

(~r,t t0 ; J ) =

1
V

(~x +~r,~x,t t0; J )

~x

A0 (~r; J )eM0 (tt0 ) ,

(2.6)

with V being the box volume. Thus, the q-q

NBS wave function is defined by the spatial correlation
of the NBS amplitudes.
The NBS wave functions in S-wave states are obtained under the projection onto zero angular
momentum (P(l=0) ),
1
(~r; J ) =
(2.7)
P(l=0) (g1~r; J ),
24 gO
where g O represent 24 elements of the cubic rotational group, and the summation is taken for
all these elements. Using Eq. (2.3) and Eq. (2.7), we will find the q-q
potentials and NBS wave
functions from lattice QCD.
Simulation setup is as follows. We employ quenched QCD with the standard plaqutte gauge
action. Lattice size is 323 48 and 6/g2 = 6.0, which corresponds to the physical volume
V = (3.2fm)3 and the lattice spacing a = 0.10fm. We measure the q-q
NBS wave functions for
four different quark masses with hopping parameters = 0.1520, 0.1480, 0.1420, 0.1320: the corresponding pseudo-scalar (PS) meson masses mPS in the calculation are 0.94, 1.27, 1.77, 2.53GeV,
3

The q-q
potential from Bethe-Salpeter amplitudes on lattice

Yoichi Ikeda

Figure 1: The q-q

NBS wave functions in PS(a) and V(b) channels. The wave functions are normalized at
origin. All the wave functions are localized in the box and indicate the bound states.

and vector (V) meson masses mV =1.04, 1.35, 1.81, 2.55GeV, respectively. The number of configurations is 100 for each quark mass. For the source operator of mesons, we use wall source. We
fix the gauge, because q and q operators are spatially separated at the time slice of source and sink,
and we adopt Coulomb gauge in the calculation.

3. Numerical results for the q-q

potentials
First, we show the numerical results of the NBS wave functions in Fig. 1. Figure 1(a) and 1(b)
are the NBS wave functions for each quark mass in PS and V channels, respectively, at the time slice
t = 20. The NBS wave functions mostly vanish at r = 1.5fm for all quark masses in both channels.
This indicates that the spatial volume V = (3.2fm)3 is enough for the present calculations. The
size of a wave function with a lighter quark mass becomes smaller than that with a heavier one.
Comparing the results in PS and V channels, little channel dependence between PS and V channels
is found, although the quark-mass dependence of the wave functions is a bit larger for V channel.
In Fig. 2, we show the Laplacian parts of q-q
potentials in Eq. (2.3), 2 (r)/ (r), for each
quark mass and channel. Figure 2(a) represents 2 (r)/ (r) = 2 (V (r) E) in PS channel for
each quark mass at the time slice t = 20. As shown in Fig. 2, one can see that the potential form
is similar to that obtained from Wilson loop, namely, the potential form looks like linear plus
Coulomb form, although the derivation of the potentials is largely different between these two
methods. Figure 2(b) represents 2 (r)/ (r) in V channel for each quark mass at the same time
slice t = 20. The basic properties are same as that in PS channel, although quark mass dependence
is a bit larger for V channel.
Figure 3(a) and 3(b) are the plots of the potentials with arbitrary energy shifts E, i.e, V (r)
E=2 (r)/(2 (r)) in PS and V channels, respectively, for each quark mass at the time slice
t = 20. Note that the quark mass mq (= 2 ) is determined by the half of vector meson mass,
mq = mV /2, as mentioned in the previous section. We fit the analytic function to the data in
Fig. 3(a) and 3(b). In the present study, we choose the linear + Coulomb (+ constant) form, i.e.,
V (r) = A/r + r +C, as the analytic function. The fitting results are listed in Table 1. As shown
4

The q-q
potential from Bethe-Salpeter amplitudes on lattice

Yoichi Ikeda

Figure 2: Plots of 2 (r)/ (r) = 2 (V (r) E) in PS channel (a) and V channel (b) for each quark mass.
The potential form shows Coulomb + linear behavior.

Figure 3: Plots of the potential with arbitrary constatnt energy shift V (r) E = 2 (r)/(2 (r)) in PS
channel (a) and V channel (b) for each quark mass.

in Table 1, we find that the string tension moderately increases as quark mass increases in both
channels. Quark mass dependence of the string tension in PS channel is larger than that in V
channel. The string tension at the heaviest quark mass, mPS = 2.53GeV, is 950 (1011) MeV/fm in
PS (V) channel. These values are roughly consistent with the value in heavy quark limit predicted
from Wilson loop. On the other hand, Coulomb coefficient has significantly large quark-mass
dependence in both channels and is larger in PS channel than that in V channel.

4. Discussion and summary

We have studied the anti-quarkquark (q-q)

potentials from the q-q

Nambu-Bethe-Salpeter
(NBS) wave functions. For this purpose, we have utilized the method which has been recently
developed in the calculation of nuclear force from QCD [6, 7]. We have calculated the NBS wave
functions for q-q
systems with four different quark masses in pseudo-scalar and vector channels
and obtained the q-q
potentials with finite quark masses through the effective Schdinger equation.

As a result, we find Coulomb + linear form of the q-q

potentials like the infinitely heavy Q-Q
5

The q-q
potential from Bethe-Salpeter amplitudes on lattice

mPS (GeV)
2.53
1.77
1.27
0.94

Pseudo-scalar
(MeV/fm) A (MeV fm)
950
155
878
193
821
250
762
329

mV (GeV)
2.55
1.81
1.35
1.04

Yoichi Ikeda

Vector
(MeV/fm)
1011
951
920
914

A (MeV fm)
123
136
156
182

Table 1: The Fitting results of the potentials in Fig. 3. The data is fitted by the fit function V (r) = A/r +
r + C.

potential obtained from Wilson loop. By fitting the results, we have obtained the string tension
and Coulomb coefficient, and found the quark mass dependence of these coefficients. We have
found that the string tension moderately depends on the quark mass. On the other hand, Coulomb
coefficient decreases as quark mass increases. We have also checked the volume and the cutoff
dependence of the NBS wave functions and the q-q
potentials. Then, we have found that the result
shown here does not change quantitatively, although we do not show these checks here.
This is the first step to study the q-q
potentials from the NBS wave functions, and the main
purpose of the present study to show that the method is applicable to the q-q
potentials. We find
that the obtained q-q
potential has basic property of that obtained from Wilson loop. Therefore, the
method can be used for the study of the q-q
potentials with finite quark masses. Since the efficiency
of this method is confirmed, there are many extensions by using this method such as the dynamical
calculations of the q-q
potentials, the q-q
potentials at finite temperature, the 3q potential with
finite quark masses, color non-singlet q-q potentials, and so on. The results of these extensions
inter-quark potentials will be reported elsewhere.

5. Acknowledgment
The aouthors thank S. Aoki, T.Doi, T. Hatsuda, T. Inoue, N. Ishii, K. Murano, H. Nemura,
K. Sasaki and S. Sasaki for the fruitful discussion. Y.I. also thanks N. Kaiser, A. Laschka and
W. Weise for the useful discussion. The calculations were performed mainly by using the NECSX9 and SX8R at Osaka University, and partly by RIKEN Integrated Cluster of Clusters (RICC)
facility. This work is supported by the Japan Society for the Promotion of Science, Grant-in-Aid
for Scientific Research on Innovative Areas (No. 2004: 20105001, 20105003).

References
[1] C. B. Chiu, Ann. Rev. Nucl. Part. Sci. 22, 255 (1972).
[2] G. S. Bali, Phys. Rept. 343, 1 (2001).
[3] L. S. Brown and W. I. Weisberger, Phys. Rev. D 20, 3239 (1979).
[4] E. Eichten and F. L. Feinberg, Phys. Rev. Lett. 43, 1205 (1979).
[5] Y. Koma and M. Koma, Nucl. Phys. B769, 79 (2007).
[6] N. Ishii, S. Aoki and T. Hatsuda, Phys. Rev. Lett. 99, 022001 (2007).

The q-q
potential from Bethe-Salpeter amplitudes on lattice

Yoichi Ikeda

[7] S. Aoki, T. Hatsuda and N. Ishii, Prog. Theor. Phys. 123, 89 (2010).
[8] H. Nemura, N. Ishii, S. Aoki and T. Hatsuda, Phys. Lett. B673, 136 (2009).
[9] H. Nemura et al. [HAL QCD and PACS-CS Collaboration], PoS LATTICE2009, 152 (2009).
[10] T. Inoue et al. [HAL QCD collaboration], arXiv:1007.3559 [hep-lat].
[11] K. Murano, N. Ishii, S. Aoki and T. Hatsuda, PoS LAT2009, 126 (2009).
[12] Y. Ikeda et al. [HAL QCD Collaboration], arXiv:1002.2309 [hep-lat].
[13] T. Kawanai and S. Sasaki, arXiv:1007.1515 [hep-lat].
[14] M. M. Levy, Phys. Rev. 88, 725 (1952).
[15] A. Klein, Phys. Rev. 90, 1101 (1953).