Braz J Phys (2016) 46:730–739
DOI 10.1007/s13538-016-0449-9
NUCLEAR PHYSICS
The ρ(1S, 2S), ψ(1S, 2S), ϒ(1S, 2S) and ψt (1S, 2S)
Mesons in a Double Pole QCD Sum Rule
M. S. Maior de Sousa1 · R. Rodrigues da Silva2
Received: 17 September 2015 / Published online: 27 September 2016
© Sociedade Brasileira de Fı́sica 2016
Abstract We use the method of double pole QCD sum
rule, which is basically a fit with two exponentials of
the correlation function, where we can extract the masses
and decay constants of mesons as a function of the
Borel mass. We apply this method to study the mesons:
ρ(1S, 2S), ψ(1S, 2S), ϒ(1S, 2S), and ψt (1S, 2S). We also
present predictions for the toponiuns masses ψt (1S, 2S) of
m(1S) = 357 GeV and m(2S) = 374 GeV.
Keywords QCD sum rules · Excited states
1 Introduction
In 1977, Shifman, Vainshtein, Zakharov, Novikov, Okun
and Voloshin [1–4] created the successful method of QCD
sum rules (QCDSR), which is widely used nowadays. With
this method, we can calculate many hadron parameters
such as: mass of the hadron, decay constant, coupling con-
stant and form factors in terms of the QCD parameters as
for example: quark masses, the strong coupling and non-
perturbative parameters like quark condensate and gluon
condensate. The main point of this method is that the
quantum numbers and content of quarks in hadron are rep-
resented by an interpolating current, where the correlation
 R. Rodrigues da Silva
romulo@df.ufcg.edu.br
1 Departamento de Fı́sica, Universidade Federal da Paraı́ba,
58.051-900, João Pessoa, PB, Brazil
2 Unidade Acadêmica de Fı́sica, Universidade Federal
de Campina Grande, 58.051-970, Campina Grande, PB, Brazil
function of this current is introduced in the framework of the
operator product expansion (OPE). To determine the mass
and the decay constant of the ground state of the hadron, we
use the two-point correlation function. On the QCD side, the
correlation function can be written in terms of a dispersion
relation and on the phenomenological side can be written
in terms of the ground state and several excited states. The
usual QCDSR method uses an ansatz that the phenomeno-
logical spectral density can be represented by a form pole
plus continuum, where it is assumed that the phenomeno-
logical and QCD spectral density coincides with each other
above the continuum threshold. The continuum is repre-
sented by an extra parameter called s0 , as being correlated
with the onset of excited states [5]. In general, the resonance
√
activity occurs with s0 lower than the mass of the first
excited state.
For the ρ meson spectrum, Fig. 1, the purpose of the pole
plus continuum is a good approach, due to the large decay
width of the ρ(2S) or ρ(1450), which allows to approximate
the excited states as a continuum.
√
For the ρ meson [6], the value of s0 that best fit the
√
mass and the decay constant is s0 = 1.2 GeV and for the
√
φ(1020) meson the value is s0 = 1.41. We note that the
√
values quoted above for s0 are about 250 MeV below the
poles of ρ(1450) and φ(1680). One interpretation of this
result is due to the effect of the large decay width of these
mesons.
In the pioneering work on charmonium sum rule,
Novikov et al. [1] considered the phenomenological side
 
with double pole and s0 = 4 GeV, where s0 is the dou-
ble pole continuum parameter. This value is correlated with
the threshold 
of pair production of charmed mesons. Using
this value of s0 and Moment Sum Rule at Q2 = 0, they
presented the first estimate for the gluon condensate and a
Braz J Phys (2016) 46:730–739
Fig. 1 (Left) The radial excited states of the ρ meson [12, 40, 41]. (Right) The radial excited states of the J /ψ meson [10, 12, 40]
nice prediction of the ηc mass of 3.0 GeV, while the exper-
imental results in 1977 reported a ηc mass of 2.83 GeV
[1, 2, 4].
In single pole sum rule for J /ψ and ηc , the best values
√
of s0 that fit the masses are (3.8 ± 0.2) GeV [6], where
√
the minimum value of s0 is 100 MeV below of the ψ(2S)
mass. As the decay width of the ψ(2S) is about 0.3 MeV, so
√
it is approximate to associate the parameter s0 with some
activity of excited states.
√
In principle, the value of s0 can be fixed by setting
the mass of the ground state. On the other hand, in the
case where the mass of the ground state is unknown, as
in the case of tetraquarks, there are studies that extract the
√
lower limit of s0 , since the pole dominance and OPE
convergence is controlled [7].
In double pole QCDSR [4], we expect that a reliable sum
rule should provide that the ground state decay constant is
larger than the excited state decay constant and provide an
upper limit of s0 , as we can see in our results. This con-
dition is directly related to the expression of decay constant
obtained from potential models [8–10] and is proportional
to the meson wave function at the origin. As the meson
radius increases with excitation, the probability of finding
its quarks at the origin declines with the excitation [11]. This
condition was used by Shifman [4] to predict the mass of the
ηc and this condition agrees with the experimental data from
the spectrum of ψ(nS) and ϒ(nS) up to ϒ(4S) [12]. For
ϒ(5S), a violation in this behavior is observed, where the
decay constant of ϒ(5S) is larger than ϒ(4S). This result is
not predicted by potential models and the authors of Ref.[9]
suggest that the ϒ(5S) could be a tetraquark or molecule
state.
In QCDSR, the excited states are studied in pole-pole
plus continuum in Moment Sum Rule at Q2 = 0 [1, 2],
the spectral sum rules with pole-pole-pole plus continuum
[13], the Maximum Entropy Method [14] and Gaussian
Sum Rule with pole-pole plus continuum ansatz [15]. There
are studies on the ρ(1S, 2S) mesons [14, 16, 17], nucle-
ons [13, 18], ηc (1S, 2S) mesons [2], ψ(1S, 2S) mesons
731
[1, 19] and ϒ(1S, 2S) mesons [20]. In the Gaussian Sum
Rule is studied the mixed states of the glueballs and scalar
mesons. In lattice QCD, there are studies on the π(1S, 2S)
mesons [21], ρ mesons excited states [22, 23], charmonium
[24–26], nucleons excited states [11, 27–29] and exotic
charmonium spectrum [30]. In addition, the excited states
have been studied recently by several approaches like: QCD
Bethe–Salpeter equation [31] for π(2S) and ρ(2S), light–
front quark model [32, 33] for ρ(2S), ηc (2S), ψ(2S) and
the bottomonium analogous. The ψ(2S) has been studied
in QCDSR as a hybrid meson [34] using the pole plus
continuum ansatz.
The method pole-pole plus continuum ansatz was used
in lattice QCD for nucleons [11]. The authors have shown a
problem in which the ground state coupling strength is lower
than the excited state coupling strength.
There are many motivations to study the excited states
that belong to the charmonium spectrum. New charmonium-
like states Y(4260) and Y(4660) are an example of the
importance of excited states. When considering theories that
Y(4260) has been proposed as a bound state of J /ψ − f 0
[35] and Y(4660) has been interpreted as a bound state
of J /ψ(2S) − f 0, [36–38], where we can speculate that
Y(4660) is an excited state of Y(4260). Another point is
that Z + (4430) could be an excited state of Zc (3900) and
Zb+ (10610) could be an excited state of Xb+ (10100) [39].
In this paper, we study the excited state using the pole-
pole plus continuum ansatz in QCD sum rules and we apply
in four cases: the ρ(1S, 2S), ψ(1S, 2S), ϒ(1S, 2S) and
ψt (1S, 2S) mesons and we calculate their masses and decay
constants.
2 The Sum Rule
In the determination of the mass and the decay constant with
QCDSR, we use the two-point correlation function [3],

μν (q) = i d 4 x eiq·x 0 | T {jμ (x)jν† (0)} | 0, (1)
732 Braz J Phys (2016) 46:730–739

where on the QCD point of view, the current of q q̄ vector Finally, we obtain the sum rule:
mesons has the form: ∞ ∞
ρ P hen (s) ρ P ert (s)
ds = ds + nonP ert (q 2 ). (10)
jμ (x) = q̄a (x) γμ qa (x) (2) s − q2 s − q2
0 s0min
Inserting this current in the correlation function, (1), are
To improve the equivalence between the two sides of the
obtained the operators expansion, OPE, which can be writ-
sum rule is convenient to use the Borel transformation [3]:
ten in terms of a dispersion relation, which depends on the
QCD parameters, then the correlator can be written in the ∞ ∞
form: ds ρ P hen (s)e−sτ = ds ρ P ert (s)e−sτ + nonP ert (τ ),
 0 s0min
μν (q) = i d 4 x eiq·x 0 | T {jμ (x)jν† (0)} | 0
QCD
(11)
= (qμ qν − q gμν ) 2 QCD 2
(q ) , (3)
with nonP ert (τ ) = B[nonP ert (q 2 )] and τ = 1/M 2 ,
where M is Borel mass.
with:
For the sum rule of ρ meson, we use ρ P ert (s) and
nonP ert (τ ) given by [3, 5, 6]:
∞ 
ρ P ert (s)
QCD (q 2 ) = ds
s − q2
+ nonP ert (q 2 ), (4) 1  αs 
ρ P ert (s) = 1 + , (12)
s0min 4π 2 π
 
P ert 1 αs 2
where ρ P ert (s) = Im( π (s)) and the parameter s0min is the nonP ert (τ ) = τ  G  + 2 mq q̄q
minimum value of s to have an imaginary part of the pertur- 12 π
bative term P ert (s) and the correlator nonP ert (q 2 ) is the 112
−τ 2 π αs q̄q2 , (13)
contribution of the condensates. 81
On the phenomenological side, we use: whereαs is the
strong coupling constant, mq is light quark
mass, απs G2 is gluon condensate, q̄q is quark condensate,
0 | jμ (0) | V (q) = fV mV μ(V ) (q), (5) and s0min = 4 m2q . We use these parameters at μ = 1 GeV
renormalization scale [5].
where fV is the decay constant and mV is the meson mass. For the sum rules of J /ψ and ϒ mesons, we use ρ P ert (s)
Inserting (5) in (1), we get: and nonP ert (τ ) given by [1, 6].

fV2 ρ P ert (s) = ρ0 (s) + ρrad (s), (14)

Pμνhen (q) = (qμ qν − q 2 gμν )
− q2 m2V where,
+excited states contribution. (6)
1
ρ0 (s) = v(3 − v 2 ), (15)
We can write the invariant part of the correlator of the (6) 8π 2
in the form:
 
4αs π 3+v π 3
ρrad (s) = ρ0 (s) − −
Pμνhen (q) = (qμ qν − q 2 gμν )P hen (q 2 ), (7) 3 2v 4 2 4π
3αs (1 − v 2 )2
− ln(2). (16)
with 4π 3 v

∞ with v = 1 − 4m2 /s, m is off-shell heavy quark mass and
ρ P hen (s)
P hen (q 2 ) = ds , (8) αs = αs (m).
s − q2 For the gluon condensate, we apply the Borel transform
0
of the expression was given by Reinders et al. [6], where
and ρ P hen (s) = fV2 δ(s − m2V ) + ρ Excited (s). we get:
When comparing the (3) with (7), the simplest way to  1  
τ αs m2 τ
perform the sum rule is choosing an invariant structure and nonP ert (τ ) = −  G2  dα 1 +
12 π 0 α(1 − α)
equating both sides of the sum rule, so we have:  
−m τ
2
× exp . (17)
P hen (q 2 ) = QCD (q 2 ) (9) α(1 − α)
Braz J Phys (2016) 46:730–739 733

3 The Method We can observe that the equations (21) and (24) can form
an equation system in the variables,
To implement our method, we consider the following spec-
A(τ ) = f12 e−m1 τ ,
2
tral density on the phenomenological side: (25)

ρ P hen (s) = f12 δ(s −m21 )+f22 δ(s −m22 )+ρ Cont (s)θ(s −s0 ), B(τ ) = f22 e−m2 τ .
2
(26)
(18) Solving the equation system (21) and (24) writing the
results in terms of the functions A(τ ) and B(τ ), we get:
where m1 is the mass of the ground state and m2 is the
mass of the first excited state, f1 is the decay constant for DQCD (τ ) + QCD (τ ) m22
A(τ ) = , (27)
the ground state, and f2 is the decay constant for the first m22 − m21
excited state, (5), and s0 mark the onset of the continuum
states. Inserting (18) on the left-hand side of (11), we get: DQCD (τ ) + QCD (τ ) m21
B(τ ) = , (28)
m21 − m22
∞ where we use the notation
f12 e−m1 τ + f22 e−m2 τ ds ρ P hen (s) e−sτ ,
2 2
 LH S
(τ ) = + dn
D n F (τ ) = F (τ ). (29)
s0 dτ n
(19) To eliminate the dependence of the f1 coupling in (27),
we take a derivative of this equation with respect of τ and
On the right-hand side of (11), we find: divide the result by (27). The result of this procedure is

given by the (30). The procedure to eliminate f2 coupling
s0 is analogous to that used above, and the result is given by
RH S (τ ) = ds ρ P ert (s)e−sτ the (31).

s0min DQCD (τ ) m22 + D 2 QCD (τ )
∞ m1 = − , (30)
DQCD (τ ) + QCD (τ ) m22
+ ds ρ P ert (s)e−sτ + nonP ert (τ ). (20)
s0 DQCD (τ ) m21 + D 2 QCD (τ )
m2 = − . (31)
DQCD (τ ) + QCD (τ ) m21
Equating (19) with (20) and using the quark hadron dual-
In the first view, the (30) and (31) suggest a system for
ity, where we assume that ρ P hen (s) = ρ P ert (s) for s ≥ s0 ,
the masses m1 and m2 , that could be extracted the masses.
so we get the double pole QCD sum rule,
On the other hand, using (30) to obtain a m2 expression, it
reproduces the same result as given in (31). To solve this
f12 e−m1 τ + f22 e−m2 τ = QCD (τ ),
2 2
(21)
problem, where we cannot decouple the masses m1 and
where, m2 from the (30) and (31), we will take the derivative of
equation (28) twice in the form:

s0 D 3 QCD (τ ) + D 2 QCD (τ ) m21
m42 f22 e−m2 τ =
2
−sτ P ert . (32)
 QCD
(τ ) = ds e ρ (s) +  nonP ert
(τ ). (22) m21 − m22
s0min
Dividing by equation (28), we have a new mass formula,
given by:
The contribution of the resonances is given by:
D 3 QCD (τ ) + D 2 QCD (τ ) m21
∞ m42 = . (33)
−sτ DQCD (τ ) + QCD (τ ) m21
CE(τ ) = ds ρ P ert
(s) e . (23)
Inserting (30) in equation (33) we obtain a polynomial
s0
equation with respect to m2 :
As usually done in QCDSR, the obtaining mass of the m42 α + m22 β + γ = 0, (34)
hadron, we take the derivative of (21) with respect to τ and
we get the new equation: where α = +
−DQCD (τ )2 β=
QCD (τ ) D 2 QCD (τ ),
−D 2 QCD (τ ) DQCD (τ ) + D 3 QCD (τ ) QCD (τ ), γ =
d QCD D 3 QCD (τ ) DQCD (τ ) − D 2 QCD (τ )2 and  = β 2 −
−m21 f12 e−m1 τ − m22 f22 e−m2 τ =
2 2
 (τ ). (24)
dτ 4αγ .
734 Braz J Phys (2016) 46:730–739

For obtaining m1 , we can do the same procedure as

above and m1 obeys the same equation (34). We get four
non-physical solutions with negative mass, two positive
solutions for the masses of the type m1 = m2 , one m1 > m2
and solutions that describes m1 as a ground state and m2
being an excited state:
√
−β − 
m1 = , (35)
2α
√
−β + 
m2 = . (36)
2α

4 Results
In this work, we use the following parameters for ρ meson:
αs (1 GeV) = 0.5, mq = (6.4 ± 1.25) MeV, q̄q =
−(0.240 ± 0.010)3 GeV3 ,  απs G2  = (0.012 ± 0.004) GeV4
at μ = 1 GeV renormalization scale [5]. For J /ψ, we use
the αs (mc ) = 0.3 in Landau gauge, mc (mc) = 1.3 GeV
and for ϒ, we use the αs (mb ) = 0.15 in Landau gauge,
mb (mb) = 4.3 GeV.
In addition to the above-mentioned parameters, the sum
rule depends of the others two parameters: the continuum
threshold s0 and the Borel mass, M.

As explained in the introduction, we expect that s0 is
a value closes to 3S meson mass, however, in cases where
the3S state is unknown or has large decay width, the value
of s0 is limited by the condition that the decay constant of
2S meson should be smaller than 1S meson and the lowest
limit of s0 is considered as m(2S) + 100 MeV.
 
Using a value of s0 , the range of Borel mass is chosen
on the assumption that the ratio of the double pole (21) and
the total contribution pole-pole plus the resonances, (23),
should be higher than 40 %.
4.1 ρ(1S, 2S) Sum Rule
Using the mass of ρ(3S) meson of 1.9 GeV, Fig. 1, we test


s0 = 1.9 GeV, but in this case, the decay constant of the
excited state is larger than the ground state decay constant,
so the sum rule fails.  ρ(2S) mesons. Considering the value of s0 of 1.61 GeV
The maximum value of s0 is 1.66 GeV, where in this
case the decay constant of excited state is slightly below of
the
 decay constant of ground  state. The minimum value of
s0 is 1.56 GeV, because s0 -
m(2S) reaches the value of
100 MeV.
In Fig. 2, the contribution of the OPE terms are ordered
relative to the first-order perturbative term of (12) in (22).
Fig. 2 The relative contributions
 of OPE for ρ(1S, 2S) as a func-
tion of the Borel mass for s0 = 1.61 GeV. The solid line is for the
first-order perturbation term, the long-dashed line for radiative correc-
tion, the dashed-dot line is for dimension 4, and the dotted line is for
dimension 6
The solid line that is the contribution of the first-order per-
turbation term is adopted as 1, the long-dashed line is the
radiative correction, the dashed-dot line is the dimension 4
of (13) and the dotted line is the dimension 6 of (13). We
note that the convergence of OPE is controlled and at M=1
GeV, the contributions of the dimension 4 is 1.82 % and
dimension 6 is 2.26 % of the first-order perturbation term.
For M=2 GeV, these condensates contribute with 300 MeV
in the mass of ρ(1S) and 100 MeV for the mass of ρ(2S).
We study the behavior of the masses and decay con-
stants of the mesons  ρ and ρ(2S) as a  function of Borel
mass for three values s0 : solid line for s0 = 1.61 GeV,
dashed-dot line for s0 = 1.56 GeV and long-dashed line

for s0 = 1.66 GeV. We can see in Fig. 3 that all masses
are stable and at M=1.2 GeV, the long-dashed line gives a
value compatible with the experimental value for the ρ(2S)
mass of 1454 MeV and for the ρ(1S) the long-dashed line
gives a mass of 740 MeV.
For the calculation of the decay constant, we use the
experimental values m1 = 0.77 GeV and m2 = 1.46
GeV. In Fig. 4, we show the decay constant  of the ρ and
(solid line), the value of the ρ meson decay constant has a
plateau on value 203 MeV and ρ(2S) has a plateau onthe
value 186 MeV. Considering uncertainty with respect to s0
parameter at M= 2 GeV, we get:
fρ = (203 ± 2) MeV, (37)
fρ(2S) = (186 ± 14) MeV. (38)
Braz J Phys (2016) 46:730–739
Fig. 3 The ρ(1S), lower lines, and ρ(2S), upper  lines, masses as a
function of the Borel mass. The solid line is for s0 = 1.61 GeV,
 In Fig. 5, the contribution of the OPE terms are ordered
dashed-dot line is for s0 = 1.56 GeV and the long-dashed line is for

s0 = 1.66 GeV
The value of fρ is 17 MeV lower than the experimental
value of (220.5 ± 1)MeV [12] obtained from ρ 0 → e+ e−
decay width considering 1/αQED = 137.036.
It is interesting to note that in Ref. [42] is shown another
way to extract the experimental decay constant of the ρ ±
from semileptonic decay, τ ± → ρ ± ντ . Using the PDG [12],
we get:
exp
fρ ± = (213.8 ± 0.8) MeV. (39)
Fig. 4 The ρ(1S), upper lines, and ρ(2S), lower lines,
 decay constant
as a function of the Borel mass. The solid line is for s0 = 1.61 GeV,
 
dashed-dot line is for s0 = 1.56 GeV and long-dashed line is for
 perturbation term, long-dashed line for radiative correction, dashed-
s0 = 1.66 GeV
735
4.2 ψ(1S, 2S) Sum Rule
Using the mass of ψ(3S) meson of 4.04 GeV, Fig. 1, we test

s0 = 4.0 GeV, but in this case the decay constant of the
excited state is larger than the ground state decay constant,
so the sum rule fails. 
The maximum value of s0 is 3.9 GeV, where in this
case the decay constant of the excited state isslightly below
the decay of ground. The minimum value of s0 is 3.7 GeV,

because s0 - m(2S) reaches the value of 100 MeV.
It is also interesting that the massof the (1S) state is
almost independent on the value of s0 in stable Borel
range, even varying 3.3 GeV to ∞. Furthermore,
 mass 2S
state increases with the an increase of s0 , but assumes a
maximum value of 4.1 GeV.
relative to the first-order perturbative term of (14) in (22).
The solid line is the contribution of the first-order pertur-
bation term is adopted as 1, the long-dashed line is the
radiative correction, and the dashed-dot line is the gluon
condensate of (17). We note that the convergence of OPE is
controlled and the contribution of the gluon condensate is
6 % of the first-order perturbation term at M=1.4 GeV, the
same order of radiative corrections. At M=2 GeV, its con-
tribution reduces to only 1 % of the first-order perturbation
term.
We study the behavior of the mass of meson  ψ(2S) and
ψ as a function of Borel mass for three values s0 . We have
Fig. 5 The relative contributions of OPE for ψ(1S, 2S) as a function
of the Borel mass for s0 = 3.8 GeV. The solid line for the first order
dot line for gluon condensate
736
Fig. 6 The mass of ψ(1S), lower lines, and 
ψ(2S), upper lines, as a
function of the Borel mass. The solid line for s0 = 3.8 GeV, dashed-
 
dot line for s0 = 3.7 GeV and long-dashed line for s0 = 3.9 GeV
 condensate of (17). We note that the convergence of OPE is
in Fig. 6, solid line for s0 = 3.8 GeV, dashed-dot line for
 
s0 = 3.7 GeV and long-dashed line for s0 = 3.9 GeV.
We can see in Fig. 6 that all masses are stable at M > 2
GeV and the solid line is for the ψ(1S) mass of 3.07 GeV
and ψ(2S) mass of 3.64 GeV.
For the calculation of the decay constant, we use
the experimental values m1 = 3.096 GeV and m2 =
3.686 GeV. In Fig. 7 we can see that the decay constants are
stable at M > 2 GeV. Considering uncertainty with respect
Fig.
 7 The decay constant of the J /ψmeson. The solid line is for
s0 = 3.8 GeV, dashed-dot line is for s0 = 3.7 GeV and the long-
 the first-order perturbation term, the long-dashed line is for radiative
dashed line is for s0 = 3.9 GeV
Braz J Phys (2016) 46:730–739

to s0 parameter at M = 2 GeV, we get:
fψ(2S) = (272 ± 40) MeV, (40)
and for J /ψ meson decay constant, we get: fψ = (334 ±
1) MeV.
The result for the decay constant of ψ(2S) is in agree-
ment with the experimental value of ψ(2S) of (294 ± 5)
MeV [12] obtained from V 0 → e+ e− decay width con-
sidering 1/αQED = 137.036. For J /ψ, the decay constant
is 82 MeV lower than the experimental value of J /ψ of
(416 ± 5) MeV.
4.3 ϒ(1S, 2S) Sum Rule
Using themass of ϒ(3S) meson of 10.35 GeV, Fig. 1, we
consider s0 = 10.30 GeV.
In Fig. 8, the contribution of the OPE terms are ordered
relative to the first-order perturbative term of (14) in (22).
The solid line is the contribution of the first-order pertur-
bation term and is adopted as 1, the long-dashed line is the
radiative correction, and the dashed-dot line is the gluon
controlled and the contribution of the gluon condensate is
only 0.05 % of the first-order perturbation term at M=5 GeV
and 0.01% at M=7 GeV.
We study the behavior of the mass of ϒ(2S)  and ϒ as
a function of Borel mass for three values of s0 . We can

see in Fig. 9 a solid line for s0 = 10.30 GeV, dashed-dot
 
line for s0 = 10.25 GeV and long-dashed line for s0 =
10.4 GeV.
Fig. 8 The relative contributions of OPE for ϒ(1S, 2S) as a func-


tion of the Borel mass for s0 = 10.30 GeV. The solid line is for
correction, and the dashed-dot line is for gluon condensate
Braz J Phys (2016) 46:730–739 737


Considering uncertainty with respect to s0 parameter at
M = 6 GeV, we get:

fϒ(2S) = (467 ± 51) MeV. (41)

For the ϒ meson decay constant, we get: fϒ = (540 ±

12) MeV.

4.4 ψt (1S, 2S) Toponium Meson Sum Rule

In this case, we show how to use our method to predict par-

Fig. 9 The mass of ϒ(1S), lower lines, and ϒ(2S),
 upper lines, as a
function of the Borel mass. The solid line is for s0 = 10.30 GeV, the

dashed-dot line is for s0 = 10.25 GeV and the long-dashed line is

for s0 = 10.35 GeV
We can see in Fig. 9 that all masses are stable at M >
6.5 GeV. At M=5 GeV, the mass obtained for ϒ(1S) is in
agreement with the experimental value and ϒ(2S) is 200
MeV above of the experimental value.
For calculation of the decay constant, we use the exper-
imental values m1 = 9.46 GeV and m2 = 10.02 GeV. In
Fig. 10 we see that the values for the decay constant have
good stability for a Borel mass above 6 GeV.
ticles not yet discovered as ψt (2S, 1S) compound of top
quark.
One problem of this sum rule is that the t-quark decays
much faster than it forms a hadronic resonant state.
There are papers that are against the existence of the
toponium [43] and papers in favor Ref. [44, 45], where cor-
rection O(αs2 ) is important. They have predicted a mass of
ψt (1S) and ηt (1S)
 with a mass of 347.4 GeV.
In this case, s0 is chosen to satisfy ordering of decay

constants and the condition s0 − m(2S) is of about 100

MeV. However, we prefer to relax this condition to s0 −
m(2S) = 1 GeV, due the variation of m(2S) as Borel mass
pole
in the scale of 1 GeV. We also consider the value of mt =
172 [12], αs (mt ) = 0.1, where this value is close to the
results of the Refs. [46, 47] and the maximum value of the
gluon condensate. 
Initially, we attempt a value to s0 as is shown in the
first column of Table 1, where the Borel window is lim-
ited between M0 and Mf , where the pole contribution is
above 40 % of total correlator, and OPE convergence is
controlled. The masses are calculated in distinct Borel win-
dows. The decay constants are calculated at the midpoint
Mm= (M0 + Mf )/2. In the first attempt, we can see that
the s0 value led to a violation in the ordering of the decay
constants,
 which leads us in the next attempt to use values
of s0 smaller than 1 TeV. Only in the third attempt was the

Table 1 Sum rule of ψt (1S, 2S) for αs (mt ) = 0.1 and mt (mt ) =
164.7 GeV. All quantities are given in GeV

Attempt 1 Attempt 2 Attempt 3 Attempt 4


s0 1000 450 376 375
M0 2000 200 100 100
Mf 10000 1000 300 300
Fig. 10 The decay constant of ϒ(2S), lower lines, and ϒ(1S),upper
m(1S)(M0 ) 540 364 357 357
lines, as a function of the Borel mass. The solid line is for s0 =
 m(2S)(Mf ) 903 430 374 374
10.30 GeV, the dashed-dot line is for s0 = 10.25 GeV and the long- f (1S)(Mm ) 103 27 18.9 18.7

dashed line is for s0 = 10.35 GeV f (2S)(Mm ) 109 32 7.6 7.1
738 Braz J Phys (2016) 46:730–739

Table 2 Decay constants of the 2S states and 1S states in MeV. The decay constant of Ref. [23], we have fixed the ρ(2S) mass to 1540
“column experiment” refers to the average values of PDG [12], to the MeV, which is the average value found by them
process V 0 → e+ e− , considering 1/αQED = 137.036. In the ρ(2S)

This work Ref. [33] Ref. [31] Ref. [32] lattice lattice experiment
(ω = 0.5 GeV) Ref. [23, 24] Ref. [48, 49] Ref. [12]

ρ 203 ± 2 216.37 268 – 225 ± 9 239 ± 18 220.5 ± 1

ρ(2S) 186 ± 14 128 155 – 185 ± 78 – –
J /ψ 334 ± 1 – – – 399 ± 4 418 ± 13 416.3 ± 6.0
ψ(2S) 272 ± 40 – – 371 143 ± 81 – 294.5 ± 4.5
ϒ 540 ± 12 – – 546.6 – – 715 ± 5
ϒ(2S) 467 ± 51 – – 583.2 – – 497.5 ± 4.5


ordering obtained. Now we improve the gap between s0
and m(2S) with a value of about 1 GeV, which is obtained
in the final attempt.
Thus, we get the following results for the masses of
ψt (1S, 2S) of m(1S) = 357 GeV and m(2S) = 374 GeV.
Our estimate for m(1S) is 10 GeV higher than mass
predicted by Ref. [44].
Finally, we collect all the results from the decay constant
obtained in this section in Table 2. The column “this work”
refers to the extraction of decay constants on the same Borel
window. The “column experiment” refers to the average val-
ues of PDG [12], to the process V 0 → e+ e− , considering
1/αQED = 137.036.
We finish with an application of this method to study the
hypothetical particle called toponium. In this case, we start
with an initial tentative value for the continuum threshold
using a very high initial value of 1 TeV and we note that the
ordering of the decay constants is violated, which led us nat-
urally to reduce the continuum threshold up to the minimum
value of m(2S)+ 1 GeV. We use the lowest value of the con-
tinuum threshold to get the toponiuns masses ψt (1S, 2S) of
m(1S) = 357 GeV, and m(2S) = 374 GeV.
Acknowledgments We would like to thank Prof. Francisco de Assis
de Brito for fruitful discussions. This work has been partially sup-
ported by CAPES.

5 Conclusions
In this work, we have presented a method to QCD sum rule
with double pole, which is basically a fit with two exponen-
tials of the correlation function, where we can extract the
masses and decay constants of mesons as a function of the
Borel mass. We study the mesons: ρ(1S, 2S), ψ(1S, 2S),
and ϒ(1S, 2S), where we know their masses and decay con-
stants from the experimental data, except the ρ(2S) decay
constant. We also study the hypothetical meson called topo-
nium as an example of how to use our method to predict new
hadrons.
Using the experimental values for the meson masses, the
decay constants have a good stability as Borel mass, and
we have shown a prediction for the ρ(2S) decay constant of
fρ(2S) = (186 ± 14) MeV.
The decay constant of ρ(1S) has a 17 MeV value lower
than the experimental data using 1/αQED (0) to ρ 0 →
e+ e− and within the experimental estimates for the semilep-
tonic decay, considering high values for the quarks and
gluons condensates. In addition, the decay constant of
ψ(1S) and ϒ(1S) has a value lower than the experimental
data.
References
1. V.A. Novikov, L.B. Okun, M.A. Shifman, A.I. Vainshtein, M.B.
Voloshin, V.I. Zakharov, Phys. Rept. 41, 1 (1978)
2. V.A. Novikov, L.B. Okun, M.A. Shifman, A.I. Vainshtein, M.B.
Voloshin, V.I. Zakharov, Phys. Lett. B 67, 409 (1977)
3. M.A. Shifman, A.I. Vainshtein, V.I. Zakharov, Nucl. Phys. B 147,
385 (1979)
4. M.A. Shifman, A.I. Vainshtein, M.B. Voloshin, V.I. Zakharov,
Phys. Lett. B 77, 80 (1978)
5. P. Colangelo, A. Khodjamirian, in At the frontier of parti-
cle physics, ed. by M. Shifman. arXiv:hep-ph/0010175, Vol. 3,
pp. 1495-1576
6. L.J. Reinders, H. Rubinstein, S. Yazaki, Phys. Rept. 127, 1 (1985)
7. M.E. Bracco, S.H. Lee, M. Nielsen, R. Rodrigues da Silva, Phys.
Lett. B 671, 240 (2009). arXiv:0807.3275 [hep-ph]
8. V.I. Zakharov, B.L. Ioffe, L.B. Okun, Sov. Phys. Usp. 18, 757
(1975). Usp. Fiz. Nauk 117, 227 (1975)
9. J. Segovia, D.R. Entem, F. Fernandez, arXiv:1409.7079 [hep-ph]
10. O. Lakhina, E.S. Swanson, Phys. Rev. D 74, 014012 (2006).
arXiv:hep-ph/0603164
11. D.B. Leinweber, Phys. Rev. D 51, 6369 (1995).
arXiv:nucl-th/9405002
12. Particle Data Group: J. Beringer et al., Phys. Rev. D 86, 010001
(2012)
13. J.P. Singh, F.X. Lee, Phys. Rev. C 76, 065210 (2007).
arXiv:nucl-th/0612059
14. P. Gubler, M. Oka, Prog. Theor. Phys. 124, 995 (2010).
arXiv:1005.2459 [hep-ph]
Braz J Phys (2016) 46:730–739
15. D. Harnett, R.T. Kleiv, K. Moats, T.G. Steele, Nucl. Phys. A 850,
110 (2011). arXiv:0804.2195 [hep-ph]
16. A.P. Bakulev, S.V. Mikhailov, Phys. Lett. B 436, 351 (1998).
arXiv:hep-ph/9803298
17. A.V. Pimikov, S.V. Mikhailov, N.G. Stefanis, arXiv:1312.2776
[hep-ph]
18. K. Ohtani, P. Gubler, M. Oka, AIP Conf. Proc. 1343, 343 (2011).
arXiv:1104.5577 [hep-ph]
19. P. Gubler, K. Morita, M. Oka, Phys. Rev. Lett. 107, 092003
(2011). arXiv:1104.4436 [hep-ph]
20. K. Suzuki, P. Gubler, K. Morita, M. Oka, arXiv:1204.1173 [hep-
ph]
21. UKQCD Collaboration: C. McNeile et al., Phys. Lett. B 642, 244
(2006). arXiv:hep-lat/0607032
22. Bern-Graz-Regensburg Collaboration: T. Burch et al., Phys. Rev.
D 70, 054502 (2004). arXiv:hep-lat/0405006
23. CP-PACS Collaboration: T. Yamazaki et al., Phys. Rev. D 65,
014501 (2002). arXiv:hep-lat/0105030
24. J.J. Dudek, R.G. Edwards, D.G. Richards, Phys. Rev. D 73,
074507 (2006). arXiv:hep-ph/0601137
25. J.J. Dudek, R.G. Edwards, N. Mathur, D.G. Richards, Phys. Rev.
D 77, 034501 (2008). arXiv:0707.4162 [hep-lat]
26. L. Liu, S.M. Ryan, M. Peardon, G. Moir, P. Vilaseca,
arXiv:1112.1358 [hep-lat]
27. N. Mathur, Y. Chen, S.J. Dong, T. Draper, I. Horvath, F.X. Lee,
K.F. Liu, J.B. Zhang, Phys. Lett. B 605, 137 (2005). arXiv:hep-
ph/0306199
28. R.G. Edwards, J.J. Dudek, D.G. Richards, S.J. Wallace, Phys. Rev.
D 84, 074508 (2011). arXiv:1104.5152 [hep-ph]
29. D. Guadagnoli, M. Papinutto, S. Simula, Phys. Lett. B 604, 74
(2004). arXiv:hep-lat/0409011
30. L. Liu, G. Moir, M. Peardon, S.M. Ryan, C.E. Thomas, P.
Vilaseca, J.J. Dudek, R.G. Edwards et al., arXiv:1204.5425
[hep-ph]
739
31. S.-x. Qin, L. Chang, Y.-x. Liu, C.D. Roberts, D.J. Wilson, Phys.
Rev. C 85, 035202 (2012). arXiv:1109.3459 [nucl-th]
32. T. Peng, B.-Q. Ma, arXiv:1204.0863 [hep-ph]
33. D. Arndt, C.-R. Ji, Phys. Rev. D 60, 094020 (1999). arXiv:hep-
ph/9905360
34. L.S. Kisslinger, Phys. Rev. D 79, 114026 (2009). arXiv:0903.1120
[hep-ph]
35. A. Martinez Torres, K.P. Khemchandani, D. Gamermann, E. Oset,
Phys. Rev. D 80, 094012 (2009). arXiv:0906.5333 [nucl-th]
36. Z.G. Wang, X.H. Zhang, Commun. Theor. Phys. 54, 323 (2010).
arXiv:0905.3784 [hep-ph]
37. R.M. Albuquerque, M. Nielsen, R.R. da Silva, Phys. Rev. D 84,
116004 (2011). arXiv:1110.2113 [hep-ph]
38. F.-K. Guo, C. Hanhart, U.-G. Meissner, Phys. Lett. B 665, 26
(2008). arXiv:0803.1392 [hep-ph]
39. F.S. Navarra, M. Nielsen, J.-M. Richard, J. Phys. Conf. Ser. 348,
012007 (2012). arXiv:1108.1230 [hep-ph]
40. S. Godfrey, N. Isgur, Phys. Rev. D 32, 189 (1985)
41. D. Ebert, R.N. Faustov, V.O. Galkin, Phys. Rev. D 79, 114029
(2009). arXiv:0903.5183 [hep-ph]
42. D. Becirevic, V. Lubicz, F. Mescia, C. Tarantino, JHEP 0305, 007
(2003). arXiv:hep-lat/0301020
43. N. Fabiano, A. Grau, G. Pancheri, Phys. Rev. D 50, 3173 (1994)
44. Y. Kiyo, Y. Sumino, Phys. Rev. D 67, 071501 (2003). arXiv:hep-
ph/0211299
45. Y.P. Goncharov, Nucl. Phys. A 808, 73 (2008). arXiv:0806.4747
[hep-ph]
46. CMS Collaboration: P. Kokkas, PoS EPS -HEP2013, 436 (2013)
47. DELPHI Collaboration: J. Abdallah et al., Eur. Phys. J. C 37, 1
(2004). arXiv:hep-ex/0406011
48. ETM Collaboration: K. Jansen et al., Phys. Rev. D 80, 054510
(2009). arXiv:0906.4720 [hep-lat]
49. D. Becirevic, G. Duplancic, B. Klajn, B. Melic, F. Sanfilippo,
Nucl. Phys. B 883, 306 (2014). arXiv:1312.2858 [hep-ph]