a

r
X
i
v
:
0
7
0
9
.
0
0
7
7
v
2


[
n
u
c
l
-
t
h
]


5

O
c
t

2
0
0
7
Dependence of nuclear binding on hadronic mass variation
V. V. Flambaum
1,3
and R. B. Wiringa
2
1
Argonne Fellow, Physics Division, Argonne National Laboratory, Argonne, IL 60439
2
Physics Division, Argonne National Laboratory, Argonne, IL 60439
3
School of Physics, University of New South Wales, Sydney 2052, Australia
(Dated: February 1, 2008)
Abstract
We examine how the binding of light (A 8) nuclei depends on possible variations of hadronic
masses, including meson, nucleon, and nucleon-resonance masses. Small variations in hadronic
masses may have occurred over time; the present results can help evaluate the consequences for
big bang nucleosynthesis. Larger variations may be relevant to current attempts to extrapolate
properties of nucleon-nucleon interactions from lattice QCD calculations. Results are presented
as derivatives of the energy with respect to the dierent masses so they can be combined with
dierent predictions of the hadronic mass-dependence on the underlying current-quark mass m
q
.
As an example, we employ a particular set of relations obtained from a study of hadron masses
and sigma terms based on Dyson-Schwinger equations and a Poincare-covariant Faddeev equation
for conned quarks and diquarks. We nd that nuclear binding decreases moderately rapidly as
the quark mass increases, with the deuteron becoming unbound when the pion mass is increased
by 60% (corresponding to an increase in X
q
= m
q
/
QCD
of 2.5). In the other direction, the
dineutron becomes bound if the pion mass is decreased by 15% (corresponding to a reduction of
X
q
by 30%). If we interpret the disagreement between big bang nucleosynthesis calculations and
measurements to be the result of variation in X
q
, we obtain an estimate X
q
/X
q
= K(0.0130.002)
where K 1 (the expected accuracy in K is about a factor of 2). The result is dominated by
7
Li
data.
PACS numbers: 98.80.Cq,26.35.+c,21.45.+v
1
February 1, 2008
I. INTRODUCTION
There are several reasons to search for a variation of the fundamental constants of
nature in space and time. First, the Universe is evolving; several phase transitions happened
in the early Universe accompanied by dramatic changes in vacuum energy, fundamental
masses, fundamental interactions (electromagnetic, weak, and strong) and properties of
elementary particles (e.g., connement of quarks). At later stages the equation of state of
the Universe continued to evolve, from radiation domination (pressure p = /3 where is
the energy density), to cold-matter domination (p ), and recently, about 5 billion
years ago, to dark-energy domination (p ). In view of these dramatic changes it seems
natural to check if there is any evolution in the values of the fundamental constants during
this process.
A second reason is that spatial variation can explain a ne-tuning of the fundamental
constants which allows humans (and any life) to appear. Indeed, it is well-known that if the
values of some fundamental constants (e.g., related to the strong interaction) would dier
by even 1% we could not appear. If we assume that the fundamental constants vary in
space, this problem of ne-tuning may be resolved in a most natural way: we appeared in
the area of the Universe where the values of the fundamental constants are consistent with
our existence.
A third reason comes from theories unifying gravity and other interactions. Some theo-
ries suggest the possibility of spatial and temporal variation of physical constants in the
Universe (see, e.g. [1, 2, 3, 4, 5, 6, 7, 8]). Moreover, there exists a mechanism for making all
coupling constants and masses of elementary particles both space- and time-dependent, and
inuenced by local circumstances. The variation of coupling constants can be non-monotonic
(e.g., damped oscillations).
We can only measure the variation of dimensionless parameters which do not depend on
the units we use. In the Standard Model the two most important dimensionless parame-
ters are the ne structure constant = e
2
/ hc and the ratio of the electroweak unication
scale (determined by the Higgs vacuum expectation value) to the quantum chromodynam-
ics (QCD) scale
QCD
(dened as the position of the Landau pole in the logarithm for
2
the running strong coupling constant,
s
(r) const/ ln(
QCD
r/ hc)). The variation of the
Higgs vacuum expectation value (VEV) leads to the variation of the electron mass m
e
and
quark mass m
q
which are proportional to the Higgs VEV. The present work considers ef-
fects produced by the variation of X
q
= m
q
/
QCD
where m
q
= (m
u
+ m
d
)/2 is the average
current-quark mass.
Up to now a majority of publications about temporal and spatial variation of the fun-
damental constants have considered eects of variation of . However, the hypothetical
unication of all interactions implies that the variation of the strong interaction parameter
X
q
= m
q
/
QCD
may be larger than the variation of the electromagnetic . For example,
the grand unication theories (GUTs) discussed in Ref. [1] predict
X
q
X
q
35

(1)
The coecient here is model-dependent but large values are generic for grand unication
models in which variations come from high energy scales; they appear because the running
strong-coupling constant and Higgs constants (related to mass) run faster than . Indeed,
the electroweak (i=1,2) and strong (i=3) inverse coupling constants have the following de-
pendence on the scale and normalization point
0
:

1
i
() =
1
i
(
0
) + b
i
ln(/
0
) (2)
In the Standard Model, 2b
i=1,2,3
= 41/10, 19/6, 7; the electromagnetic
1
= (5/3)
1
1
+

1
2
and the strong
s
=
3
. In GUTs all coupling constants are equal at the unication
scale,
i
(
0
)
GUT
. We may consider two possibilities: if we assume that only
GUT
varies, then Eq. (2) gives us the same shifts for all inverse couplings:

1
1
=
1
2
=
1
3
=
1
GUT
. (3)
We see that the variation of the strong interaction constant
3
() at low energy
is much larger than the variation of the electromagnetic constant , since
3
/
3
=
(
3
/
1,2
)
1,2
/
1,2
and
3

1,2
.
The second possibility is the variation of the GUT scale (/
0
in Eq. (2)). This gives

1
1
/b
1
=
1
2
/b
2
=
1
3
/b
3
(4)
Note that in this case the variations have dierent signs since b
1
and b
2,3
have dierent
signs. However, we expect an even larger enhancement of the variation of
3
(
3
/
3
=
(b
3

3
/b
1,2

1,2
)
1,2
/
1,2
).
3
The variation of m/
QCD
can be estimated from the denition of
QCD
. The running of

s
near the electroweak scale is given by

s
()
1
b
s
ln(/
QCD
) (5)
Let us take = m
z
where m
z
is the Z-boson mass. The variation of Eq. (5) and relations
above give
(m
z
/
QCD
)
(m
z
/
QCD
)
=
1
b
s

s
(m
z
)

s
(m
z
)

s
(m
z
)

C
(m
z
)
(m
z
)
(m
z
)
(6)
The value of the constant C here depends on the model used. However, the enhancement
1/ 100 should make the factor C/ large. Note that m
z
(as well as m
e,q
) is proportional
to the Higgs VEV.
If this estimate is correct, the variation in m
e,q
/
QCD
may be easier to detect than the
variation in . The cosmological variation of m
q
/
QCD
can be extracted from the big bang
nucleosynthesis (BBN), quasar absorption spectra and Oklo natural nuclear reactor data [9].
For example, the factor of three disagreement between the calculations and measurements of
the BBN abundance of
7
Li may, in principle, be explained by the variation of m
q
/
QCD
at
the level of 10
3
10
2
[10] (see also recent work [11, 12, 13]). The claim of the variation
of the fundamental constants based on the Oklo data in Ref. [14] is not conrmed by recent
studies [15] which give a stringent limit on the possible variation of the resonance in
150
Sm
during the last two billion years. The search for the variation of m
e
/
QCD
using the quasar
absorption spectra gave a non-zero result in Ref. [16] but zero results in Refs. [17, 18].
The present time variation of m
e,q
/
QCD
can be extracted from comparison of dierent
atomic [19] and nuclear [20, 21] clocks. The review of the recent results on the variation of
, m
e
/
QCD
and m
q
/
QCD
can be found in Ref. [22].
As we mentioned above, one can measure the variation of the dimensionless parameter
X
q
= m
q
/
QCD
. Note that m
q
4 MeV
QCD
220 MeV. As a result, nuclear
parameters (nucleon mass, reaction cross-sections, etc.) are determined mainly by
QCD
.
Therefore, in all calculations it is convenient to assume that
QCD
is constant and calculate
the dependence on the small parameter m
q
. In other words, we measure all masses in
units of
QCD
and will simply restore
QCD
in the nal results. For example, the relation
between the variation of the proton mass and quark mass m
p
/m
p
= 0.064m
q
/m
q
should
be understood as X
p
/X
p
= 0.064X
q
/X
q
where X
p
= m
p
/
QCD
and X
q
= m
q
/
QCD
. This
approach was formulated in Ref. [9].
4
Predictions for the relations between the variations of quark and hadron masses may
come from a variety of approaches, such as chiral perturbation theory [23, 24, 25, 26] or
Dyson-Schwinger equations (DSE) [27]. Nuclear binding energies and spectra would vary as
a consequence of the hadron mass variation, aecting a number of physical processes, such
as BBN and the Oklo phenomenon. Large enough changes might alter the stability of some
nuclei, e.g., unbind the deuteron, bind the dineutron, or even make A=5,8 nuclei particle-
stable [29]. In this work we estimate the changes in binding for A=2-8 nuclei that result from
small changes in hadronic masses. We also evaluate the eect of much larger changes on the
two-nucleon systems, which may be relevant to present attempts to extrapolate results from
lattice QCD calculations [30]. We rst consider how changes in meson, nucleon, and nucleon-
resonance masses would alter some representative nucleon-nucleon (NN) Hamiltonians. We
then solve for the energy of the two-body systems exactly and calculate variational Monte
Carlo (VMC) estimates for the larger nuclei with these forces. We report our results as
derivatives of the energies with respect to the dierent hadron mass changes, so that the
results can be utilized with dierent predictions of the coordinated changes between quark
and hadron masses. Finally, we utilize the DSE predictions for hadron mass-dependence on
the quark mass as an example to explore the eects on nuclear spectra and BBN.
II. NUCLEAR HAMILTONIAN
We examine several Hamiltonians of the form:
H =

i
K
i
+

i<j
v
ij
+

i<j<k
V
ijk
. (7)
Here K
i
is a nonrelativistic kinetic energy operator, v
ij
is a two-nucleon potential, and V
ijk
is a possible three-nucleon potential. We consider three dierent Argonne models for v
ij
:
Argonne v
14
(AV14) and Argonne v
28
(AV28) from 1984 [31], and Argonne v
18
(AV18) from
1995 [32]. In conjunction with AV18, we will use the Urbana model IX (UIX) V
ijk
from
1995 [33].
The AV14 and AV18 models are conventional NN potentials, while the AV28 model
has additional explicit (1232) degrees of freedom. The AV14 and AV28 were constructed
together and t to the same np phase-shift solution WI81 of Arndt and Roper [34] so they
are phase-equivalent. The intention was to use them in parallel many-body calculations to
5
study the eect of including explicit s in the nuclear Hamiltonian. In practice, AV28
has proven dicult to use, so beyond the two-nucleon system, only some triton [35] and
nuclear matter [36] calculations have been reported. However, AV14 has been widely used
in a variety of few-body [37, 38] and dense nucleon matter [39] calculations. The AV18 is
an updated version of AV14 containing charge-independence-breaking (CIB) terms and a
complete electromagnetic interaction. AV18 was t directly to 4,301 pp and np data in the
1993 Nijmegen partial-wave analysis [40], and AV18 and AV18+UIX have become standard
Hamiltonians for ab initio calculations of light nuclei [41] and dense matter [42].
All three of the Argonne potentials contain electromagnetic (EM) interaction, long-range
one-pion-exchange (OPE), intermediate-range attraction, and short-range repulsion, written
as a sum of operator components:
v
ij
= v

(r
ij
) +

p
[v
p

(r
ij
) + v
p
I
(r
ij
) + v
p
S
(r
ij
)]O
p
ij
. (8)
The number of operators O
p
ij
is 14, 28, or 18, as indicated by the AVxx designation. The dif-
ferent operators are discussed in detail below; here we summarize the dierences between the
models. The AV14 and AV28 potentials both use an average nucleon mass, while AV18 keeps
separate proton and neutron masses, which introduces a small charge-symmetry-breaking
(CSB) term into K
i
. The K
i
for AV28 also depends explicitly on the mass. In AV14 and
AV28, v

is just the Coulomb interaction between protons (with a form factor); in AV18 the
magnetic moment interaction, vacuum polarization, and other small EM terms are added.
The NN part of OPE is the same in AV14 and AV28, and an average pion mass is
used. In AV18 there is a weaker OPE coupling strength, a slightly dierent form factor,
and the small charge-dependent (CD) terms owing to the dierence between neutral and
charged pion masses are kept. The intermediate-range attraction is due primarily to two-
pion-exchange (TPE) processes; in AV14 and AV18 this feature is approximated by using
the square of the OPE tensor function T(m

r) as a phenomenological radial shape for v

p
I
and
adjusting strength parameters of the 14 or 18 operators to t data. In AV28, 14 operators
with explicit NNN couplings are added to those of AV14; twelve of these have OPE
range and two have intermediate and short range. These produce much of the intermediate-
range attraction explicitly through coupled-channel eects. The coecients of the rst 14
NN operators are ret to the data but are smaller in magnitude than in AV14.
The short-range repulsion may be attributed to the exchange of heavier and mesons
6
with suitable form factors for nite-size eects, but in all the Argonne models it is treated
phenomenologically. AV14 and AV28 use a Woods-Saxon radial shape with 14 strength
parameters while AV18 has a slightly more general shape with 26 parameters.
A. Potentials
The OPE potential between nucleons can be written as:
v

(r
ij
) = f
2
NN
_
X
ij

i

j
+

X
ij
T
ij
_
. (9)
Here T
ij
= 3
zi

zj

i

j
is the CD isotensor operator which contributes when the dierence
between neutral and charged pion masses is retained, and
X
ij
=
1
3
_
X
0
ij
+ 2X

ij
_
, (10)

X
ij
=
1
3
_
X
0
ij
X

ij
_
, (11)
X
m
ij
=
_
m
m
s
_
2
1
3
mc
2
[Y (mr
ij
)
i

j
+ T(mr
ij
)S
ij
] . (12)
The S
ij
= 3
i
r
ij

j
r
ij

i

j
is the usual tensor operator and Y (mr) and T(mr) are the
normal Yukawa and tensor functions
Y (mr) =
e
r
r
(r) , (13)
T(mr) =
_
1 +
3
r
+
3
(r)
2
_
Y (mr)(r) , (14)
where = mc/ h and a short-range form factor (r) has been incorporated that makes both
Y (mr) and T(mr) vanish linearly at the origin. In AV18, the X
0,
are calculated with
explicit m

0 and m

masses and the scaling mass m

s
m

. In AV14 and AV28 an

average m

=
1
3
(m

0 + 2m

) is used, so that

X
ij
vanishes, and the scaling mass m
s
m

.
The coupling f
2
NN
=0.081 in AV14 and AV28, and 0.075 in the more modern AV18.
The intermediate-range and short-range terms in the potentials are given by
v
p
I
(r
ij
) = I
p
T
2
(m

r) , (15)
v
p
S
(r
ij
) = (S
p
+ Q
p
r + R
p
r
2
)W(r) , (16)
where W(r) is a Woods-Saxon function. For AV14 and AV28, Q
p
=R
p
=0, while in AV18,
there are boundary conditions on v
p
such that no more than two of the three S
p
, Q
p
, and
7
R
p
are independent and free to be tted for any given operator. The associated operators
O
p
ij
include fourteen charge-independent (CI) operators that are common to all the models:
O
p=1,14
ij
= [1,
i

j
, S
ij
, L S, L
2
, L
2
(
i

j
), (L S)
2
] [1,
i

j
] . (17)
Here L is the relative orbital angular momentum and S is the total spin of the pair. The
AV18 model has four additional CD and CSB terms:
O
p=15,18
ij
= [1,
i

j
, S
ij
] T
ij
, (
zi
+
zj
) . (18)
These latter terms are small, but important for tting the dierences between current pp
and np scattering data.
There are fourteen additional operators in AV28 that involve explicit degrees of free-
dom. The rst two are
O
p=15
ij
= (
i
S
j
)(
i
T
j
) + (S
i

j
)(T
i

j
) + H.c. , (19)
O
p=16
ij
= S
II
ij
(
i
T
j
) + S
II
ji
(T
i

j
) + H.c. , (20)
where S
i
(T
i
) is the transition spin (isospin) operator for particle i that changes a spin
(isospin)
1
2
state to a
3
2
state. The generalized tensor operator is S
II
ij
= 3
i
r
ij
S
j
r
ij

i
S
j
and H.c. denotes the Hermitian conjugate. These operators are part of a generalized OPE
contribution
v
II

(r
ij
) = f
NN
f
N
_
m
m
s
_
2
1
3
mc
2
_
Y (mr
ij
)O
15
ij
+ T(mr
ij
)O
16
ij
_
(21)
that produces NNN transitions. The N vertex coupling is taken as f
N
= 2f
NN
.
Other operators provide OPE contributions to NN, NN, NN, N
and transitions, with an additional coupling f

=
1
5
f
NN
being introduced.
Intermediate- and short-range interaction is also added in the diagonal NN and
channels. The full set of operators is detailed in Ref. [31].
The Urbana model of three-nucleon interaction (TNI) is written as a sum of two-pion-
exchange P-wave and remaining shorter-range phenomenological terms,
V
ijk
= V
2,P
ijk
+ V
R
ijk
. (22)
The structure of the TPE P-wave term is expressed simply as
V
2,P
ijk
=

cyc
_
A
P
2
{X
m
ij
, X
m
jk
}{
i

j
,
j

k
} + C
P
2
[X
m
ij
, X
m
jk
][
i

j
,
j

k
]
_
, (23)
8
where X
m
ij
is the same spin-space function of Eq. (12) evaluated with the average pion mass,

cyc
is a sum over the three cyclic exchanges of nucleons i, j, and k, and C
P
2
=
1
4
A
P
2
.
The A
P
2
and the strength of the V
R
ijk
term are determined by tting the binding energy of
3
H [33] and the saturation density of nuclear matter [42] in the presence of a given v
ij
; the
parameters of UIX were selected in conjunction with AV18.
B. Dependence on hadron masses
We consider changes in the Hamiltonian that would be induced by small changes in
the hadron masses. However, we do not consider changes in the coupling constants that
might occur due to mass-dependent loop corrections [23, 24]. Variations of the nucleon and
nucleon-resonance masses alter the kinetic energy operator:
K
i
=
h
2
2m
i

2
i
+ (m
i
m
N
)c
2
, (24)
where m
i
is the average nucleon mass m
N
=
1
2
(m
p
+m
n
) for AV14, but may be m

part of
the time for AV28. For AV18 the kinetic energy operator has no m

component, but does

include a small CSB piece:
K
i
=
h
2
2m
N

2
i

h
2
2m
N
_
m
n
m
p
m
n
+ m
p
_

zi

2
i
, (25)
where m
N
= 2m
p
m
n
/(m
p
+ m
n
). There is also a small dependence on the nucleon mass in
the magnetic moment parts of the EM interaction in AV18.
Changes in m
N
and m

also aect the energy expectation value for AV28 through eects
of the NNN coupled channels. This can be visualized in a simple closure approxima-
tion (see Eq. (33) of Ref. [31] and the accompanying discussion) for the TPE diagrams where
one or two intermediate s are excited, propagate, and then de-excited by the transition
potentials:
V
e
2V
NNN
1

E
1
+ (m

m
N
)
V
NNN
(26)
+ V
NN
1

E
2
+ 2(m

m
N
)
V
NN
.
The two mean-energy denominators

E
1
and

E
2
would vary, in part, as the kinetic energies
of the intermediate states:

E
1

h
2

k
2
2m
N
+
h
2

k
2
2m

, (27)
9

E
2
2
h
2

k
2
2m

, (28)
with

k an average intermediate momentum. This physical eect can be approximated in the
AV14 and AV18 potentials by multiplying the intermediate-range strength coecients I
p
of
Eq. (15) by factors (1 +
N
) and (1 +

). The above equations represent only some of the

terms contributing to the m
N
and m

dependence introduced by the coupled channels, so

we will x the
N
and

terms for AV14 by requiring that calculated two-body energies have

the same mass dependence as AV28 without these factors. We will take the same factors as
approximately correct modications for AV18 also.
The dependence on the pion mass (which vanishes in the chiral limit m
q
= 0) can be
obtained from the original pion-exchange interaction which is proportional to the second
derivative of the exponential potential,
i

j
[exp (r)/r] see, e.g. [9]. As a result, the
variation of the pion mass aects the strength of the OPE through the m
3
dependence of X
m
ij
in Eq. (12) and through the range of the Yukawa functions Y (mr) and T(mr), Eqs. (13-14).
Note that the scaling mass m
s
is not allowed to vary (this is only a coecient which does
not vanish in the chiral limit m
q
= 0). The dependence of OPE terms on pion mass is thus
the same as given in Eq. (10) of Ref. [23]. Neglecting the small eect of our short-range
form factor (r), the function m
3
Y (mr) has a constant volume integral. If m is increased,
m
3
Y (mr) will be larger inside r = 2/ and smaller outside. However, m
3
T(mr) will be
smaller at all values of r > 0.
The generalized OPE of AV28 that couples to intermediate N and states eec-
tively incorporates considerable multi-pion exchange eects. We can approximate this real
physical eect in AV14 and AV18 by assuming the intermediate-range potential has the
mass-dependence v
p
I
(r) [m
3

T(m

r)]
2
. The closure approximation of Eq. (26) that justi-
es this connection applies most clearly to the rst six static (L-independent) operators O
p
ij
of Eq. (17) because of the closed algebra of the spin (isospin) and transition-spin (isospin)
operators. However, the L-dependent operators can also be aected through the coupled-
channels, so we will also consider changing these terms in the intermediate-range potentials.
We will also consider changes to the residual v
p
I
(r) terms in AV28. Thus we will study vari-
ations attributable to the pion mass in four stages with our Hamiltonians: (1) change in the
OPE part of the potential v

(and v
II

, etc. for AV28) only; (2) change in both v

and the
static TPE parts v
p=1,6
I
(TPE-s); (3) these changes plus the L-dependent TPE parts v
p=7,14
I
10
(TPE-L); and (4) these changes in v
ij
plus the TPE part of the three-nucleon potential V
2,P
ijk
(TNI).
Representing the dependence on heavier meson masses m

m
V
is much more
problematic because the Argonne models do not have explicit heavy-meson exchange. How-
ever, the eect of a short-range interaction is determined mostly by the volume integral
of this interaction, e.g., the volume integral gives us the strength constant B if we want
to approximate a short-range interaction by a zero-range interaction B(r). Therefore, we
may approximate the dependence on masses of heavier mesons by a change in the range
parameters r
0
and a of the short-range Woods-Saxon potential
W(r) =
1
1 + exp[(r r
0
)/a]
. (29)
used in Eq. (16). To keep the same relative variation of the volume integral for the Woods-
Saxon potential and the meson exchange potential exp(m
V
r)/r we change the parameters
at the rate:
r
0
r
0
=
a
a
=
2
3
m
V
m
V
. (30)
In this case, as m
V
increases, the range of the Woods-Saxon decreases while the value at
the origin remains constant and the volume integral varies as m
2
V
.
The repulsive core of the three-nucleon potential V
R
ijk
may be attributed to a combination
of multi-pion and heavy-meson exchanges, and also some relativistic eects [43]. Owing to
its complicated nature and phenomenological treatment and the fact that it gives a small
contribution to energy expectation values, we do not attempt to determine its variation with
changes in the hadronic masses.
III. ENERGY CALCULATIONS
We calculate the energies of the
1
S
0
(np) virtual bound state and the deuteron for AV28,
AV14, and AV18 by direct solution of the two-body equations. The energies of
3
H,
3,4,5
He,
6,7
Li, and
7,8
Be are calculated for AV14 and AV18+UIX using variational Monte Carlo
(VMC) methods. The VMC method is described in detail in Ref. [41] and references therein.
Briey, we construct suitably parametrized trial wave functions
V
and evaluate the upper
bound to the ground-state energy,
E
V
=

V
|H|
V

V
|
V

E
0
, (31)
11
using Monte Carlo techniques for the multi-dimensional integration. The parameters in
V
are varied to minimize E
V
, and the lowest value is taken as the approximate solution. We
use a variational wave function of the form
|
V
=
_
_
1 +

i<j<k
U
TNI
ijk
_
_
_
_
S

i<j
(1 +U
ij
)
_
_
|
J
. (32)
The U
ij
and U
TNI
ijk
are noncommuting two- and three-nucleon correlation operators, reecting
the spatial and operator dependence of v
ij
and V
ijk
, and S is a symmetrization operator.
The form of the antisymmetric Jastrow wave function
J
depends on the nuclear state under
investigation. For the s-shell nuclei the simple form
|
J
=

i<j
f
c
(r
ij
)|
A
(JMTT
3
) (33)
is used. Here the f
c
(r
ij
) are central (spin-isospin independent) two-body correlation func-
tions and
A
is an antisymmetrized spin-isospin state, e.g.,
|
4
(0000) = A| p p n n , (34)
with A the antisymmetrization operator. The
J
for p-shell nuclei is more complicated;
details are given in Ref. [41]. The two-body correlation operator U
ij
is a sum of spin,
isospin, and tensor terms:
U
ij
=

p=2,6
u
p
(r
ij
)O
p
ij
, (35)
where the O
p
ij
are the static operators of Eq. (17). The central f
c
(r) and noncentral u
p
(r)
pair correlation functions are generated by a set of six coupled, Schrodinger-like, dierential
equations which include the v
ij
and a number of embedded variational parameters. These
parameters are optimized in the energy minimization and then kept xed as the hadronic
masses are varied. The correlations are adjusted, however, because the altered interaction
is used to regenerate the correlations in each case.
The energies we obtain are shown in Table I and compared to experiment. The VMC
method is reasonably accurate for s-shell nuclei, giving energies 2-3% less bound than exact
methods like Faddeev or Greens function Monte Carlo for a given Hamiltonian [41]. It is not
as good for p-shell binding energies, but many other features such as density distributions and
electromagnetic moments are in good agreement. We anticipate that small changes in the
binding energies induced by small changes in hadron masses will be accurately tracked with
12
the VMC calculations. The comparison between AV14 and AV18+UIX models indicates the
importance of including a three-nucleon interaction to approach the experimental energies.
We evaluate the mass-dependence of the energies of the two- and multi-nucleon systems by
changing the hadron masses m
H
one at a time, increasing the value by 0.1% and calculating
the resulting energy, and then decreasing by 0.1% and repeating the energy calculation. The
VMC calculations follow the same random walk stored from the initial energy evaluation to
reduce the contribution of statistical noise. Results given in the following three tables are
the dimensionless derivatives of the energy with respect to changes in the hadron masses
E(m
H
) =
E/E
m
H
/m
H
. (36)
These results can then be combined with any given model for how the dierent hadron masses
are correlated with the underlying quark masses to give a total binding energy prediction:
E(m
q
) = E(m
q
0
)
_
1 +

m
H
E(m
H
)
m
H
(m
q
)
m
H
_
, (37)
where m
q
0
is the physical current-quark mass. A specic example will be given below.
A. Two-nucleon energies
The two-nucleon results for E(m
H
) are given in Table II for all three Hamiltonians. A
simple approximate relation between changes in the deuteron binding energy Q and virtual
bound state energy
v
is given by [10]:

v
(m
H
)
Q(m
H
)

v
. (38)
Despite the wide range of values for E(m
H
) in Table II, this relation is valid within a
factor of 2 for fteen out of seventeen pairs of results. The two exceptions are for the OPE
components of
v
(m

) for AV14 and AV18, where the sign is incorrect. This discrepancy
will be discussed below.
Changing the nucleon mass in AV14 and AV18 primarily changes just the kinetic energy
component and is very similar for the two models: a larger m
N
translates to smaller K
and more binding and thus a positive derivative Q(m
N
) for the deuteron. For the virtual
bound state, the eect is the same but greater attraction corresponds to a reduction in the
magnitude of
v
and a negative
v
(m
N
).
13
Changing m
N
in AV28 produces a larger change in the energies due to the additional
coupled-channel eects of Eqs. (26-28) discussed above. The change is more dramatic for

v
than for Q, probably because the
1
S
0
NN channel can couple to both N and
intermediate states, while the deuteron can have only NN and components due to isospin
conservation. To approximately incorporate this physical eect into the phase-equivalent
AV14 model, we can multiply its intermediate-range strength parameters I
p
of Eq. (15) by
a factor (1 +
N
). We choose a value
N
= 0.49m
N
/m
N
that matches the mass dependence
in the deuteron as shown by the line in Table II labeled m
N
+
N
. This same factor
approximately xes the virtual bound state also. Using the same
N
factor in AV18 produces
a change about one third larger. This dierence is probably due to the fact that AV18 is
t to a more recent data set, with a weaker OPE coupling accompanied by more-attractive
intermediate-range terms and correspondingly more-repulsive short-range terms.
Increasing the mass in AV28 reduces the binding energy both through the one-body
term of Eq. (24) and through the coupled-channel eects of Eq. (26). Thus E(m

) has a
sign opposite to E(m
N
). To approximate this aect for AV14, we can again multiply its
intermediate-range strength parameters I
p
by a factor (1+

). We take

= 0.57m

/m

which gives a fair reproduction of the behavior of AV28 for both the deuteron and the virtual
bound state as shown by the line in Table II labeled

. Using the same factor in AV18

again produces a larger rate of change.
We note that if the nucleon and masses change at about the same rate, from what-
ever the underlying quark mass dependence, then the eect on the AV28 energies could be
obtained by the sum of the m
N
and m

terms in Table II, which is not very dierent from

the m
N
term alone for AV14 and AV18 or from the sum of m
N
+
N
and

terms. In fact,
these corresponding sums of terms agree within 10-20% for all three models.
Increasing the pion mass reduces the regularized OPE tensor function m
3

T(m

r) at all
values of r. The binding of the deuteron is primarily due to the tensor coupling, so Q is
reduced and Q is negative for all the models as shown by the line in Table II labeled
m

(OPE). The value for AV28 is larger because of the multi-pion-exchange eects in-
cluded through the generalized OPE potential. The AV14 and AV18 values are smaller and
close together, the dierence between them being proportional to the dierent f
2
NN
coupling
constants used in the two models.
Increasing m

in the
1
S
0
channel, which depends only on m
3

Y (m

r), has the eect of

14
slightly increasing the attraction in the virtual bound state, making (
v
) negative for
the AV14 and AV18 models. However, in the AV28 model, the generalized OPE provides
signicant intermediate-range attraction [m
3

T(m

r)]
2
, which is enough to reverse the
sign of (
v
) to be positive. As the AV14 and AV18 cases here are the only two out of
seventeen in Table II that violate the relation between Q and (
v
) of Eq. (38), it appears
that changing only the OPE part of a conventional NN potential like AV14 or AV18 is an
incomplete representation of the physics in the singlet channel.
Changing m

in the static TPE part of the interaction has a more dramatic eect than
changing just the OPE part, as shown in Table II by the line labeled m

(+TPE-s). The
Q increases by by a factor of 2.54 in the deuteron for the AV14 and AV18 models compared
to OPE only, while for AV28 it increases by a smaller factor of 1.5. More dramatically, the
AV14 and AV18 values for
v
change sign and come into agreement with Eq. (38). The
further addition of the non-static m

(+TPE-L) terms makes no dierence to the virtual

bound state, as these operators do not contribute in the
1
S
0
channel, and rather small
changes in the deuteron.
Finally, increasing the heavy-meson mass m
V
reduces the range of the Woods-Saxon
repulsion and increases the binding of the deuteron, so Q is positive. The reduced repulsion
for the virtual bound state correspondingly makes
v
negative. In both the full-pion and
heavy-meson exchanges, the AV18 has larger E values than AV14 or AV28. Again, this is
probably due to the weaker OPE and compensatingly larger intermediate-range attraction
and short-range repulsion.
B. Multi-nucleon energies
The multi-nucleon results for E(m
H
) are given in Table III for AV14 and Table IV for
AV18+UIX. For every nucleus, and for every m
H
component, the signs are the same as for
the deuteron. The relative sizes of the terms are also about the same as the deuteron, with
the exception of the m

(OPE) term. Because the light nuclei have approximately equal

numbers of deuteron-like and
1
S
0
-like pairs [44], the E(m
H
) are expected to be averages
of the two sets of trends in Table II. The anomalous behavior of
v
in the m

(OPE) case
discussed above causes these numbers to be much smaller in the multi-nucleon systems, and
thus out of proportion compared to all the other terms. As for the deuteron, the multi-
15
nucleon terms have a more rapid dependence with AV18+UIX than with AV14. However,
the explicit m

contribution from the three-nucleon force is very small.

IV. DEPENDENCE OF NUCLEAR BINDING ENERGIES AND BIG BANG NU-
CLEOSYNTHESIS ON QUARK MASS
As an example of how to incorporate our nuclear binding energy results with a specic
prediction for hadronic mass variation, we utilize the results of a Dyson-Schwinger equation
study of sigma terms in light-quark hadrons [27]. Equations (85-86) of that work gives
the rate of hadron mass variation as a function of the average light current-quark mass
m
q
= (m
u
+ m
d
)/2 as:
m
H
m
H
=

H
m
H
m
q
m
q
(39)
with
H
/m
H
values of 0.498 for the pion, 0.030 for the -meson, 0.043 for the -meson,
0.064 for the nucleon, and 0.041 for the . The values for the and -mesons were reduced
to 0.021 and 0.034, respectively, in a subsequent study [28]. We will use an average of the
and terms of 0.03 for the variation of our short-range mass parameter m
V
.
In Fig. 1 we show direct calculations for the two-nucleon states, i.e., the deuteron and
1
S
0
(np) states, for a range of m
q
values for all three Hamiltonians. The virtual bound state
energies are plotted as a positive quantity; when the
1
S
0
energy is negative it indicates a
true bound state. The dashed lines show the results from changing only the pion mass in the
OPE part of the interaction, corresponding to the line m

(OPE) in Table II. To evaluate

over this wide range of m
q
, we use the Gell-Mann-Oakes-Renner (GMOR) relation m
2

m
q
.
The solid lines show the most complete calculation for each Hamiltonian. For the AV28
model this is the sum of the terms m
N
, m

, m

(+TPE-L), and m
V
, using GMOR for the
pion and the DSE values above for the variation of all other m
H
. For the AV14 and AV18
models, it is the sum of the terms m
N
+
N
,

, m

(+TPE-L), and m
V
. As discussed
in Sec.III.A, the
N
and

eects are incorporated by multiplying the intermediate-range

strength parameters I
p
of Eq. (15) by a factor
(1 +
N
)(1 +

) = (1 +.49
m
N
m
N
)(1 .57
m

) (40)
which is (1 +.008 m
q
/m
q
) for the DSE values above.
16
At the OPE level, the AV14 and AV18 models show almost exactly the same behavior for
the
1
S
0
state, with a gradually increasing attraction as m
q
increases, while for the deuteron
they both show a more rapid decrease in binding, consistent with the results obtained in
Refs.[23, 24]. If these trends continue, the deuteron will eventually move above the sin-
glet state somewhere in the range 3-4 m
q
. The AV28 model has a somewhat more rapid
dependence for the deuteron but its singlet state parallels the deuteron, becoming less at-
tractive for larger m
q
. In the other direction, the singlet state becomes a true bound state at
0.7m
q
, but it always remains above the deuteron. This dierent behavior is a consequence
of the multi-pion exchange that is built into the AV28 model through the generalized OPE
coupling to intermediate s.
In all these models the deuteron is bound largely through the tensor coupling between the
3
S
1
and
3
D
1
NN states. In the AV28 model, the singlet state gets a considerable part of its
attraction through the tensor coupling between the
1
S
0
NN and intermediate
5
D
0
N states,
and thus has a sensitivity to changes in the pion mass similar to the deuteron. This behavior
of the singlet state, i.e., that it parallels the deuteron, is dierent from that predicted by
chiral perturbation theory evaluated at the next-to-leading-order (NLO) in Ref. [30], which
does not include the eect of the . We expect that a higher-order chiral perturbation
calculation that includes degrees of freedom will come into qualitative agreement with
our result.
For our most complete calculations, shown by the solid lines, the mass dependence of the
energies is signicantly steeper for both deuteron and singlet states, but they all are parallel
and it appears the deuteron will remain the ground state for a very large range of m
q
.
The AV28 curves shift relatively little from the OPE-only values, while the AV14, which
is essentially phase-equivalent with AV28, gives very similar results. The biggest change
and the most-rapid dependence is given by the AV18 model, with the most important
contribution coming from its static two-pion-exchange terms, i.e., the line m

(+TPE s)
in Table II. The more rapid dependence is a consequence of the deeper intermediate-range
attraction and stronger short-range repulsion, which in turn may be a consequence of the
improved quality of AV18, i.e., that it is a better t to more recent NN data.
The two-body energies can also be evaluated using Eq. (37) to combine the DSE values
for m
H
/m
H
and the dimensionless derivatives E(m
H
) of Table II. This might not be
expected to work for large changes in m
q
, or where a state is barely bound or unbound.
17
However, deuteron energies are reproduced for changes of 0.1 in m
q
/m
q
to 1% or better
by Eq. (37).
The dependence of the multi-nucleon energies for the full calculation with DSE values is
shown in Fig. 2 for AV14 and in Fig. 3 for AV18+UIX. These results have been calculated
using Eq. (37) and the E(m
H
) of Tables III and IV, respectively. The lines have been
extended to m
q
= 0.2m
q
to show the trends, although the results are not expected to be
completely linear over such a broad range. The values have been checked in a few cases by
doing direct calculations and adjusting the variational parameters to reminimize the energy.
For example, the
4
He energy can be lowered by 0.3-0.5 MeV at either end of its line, but
this change would hardly be visible at the scale shown.
The multi-nucleon energies parallel the deuteron, with generally increasing slope as the
binding energy increases. (Similar results for the triton were found in Refs. [45, 46].) The
curves are steeper for AV18+UIX than for AV14. In either case it appears that the relative
stability of all the nuclei will be preserved across a broad range of m
q
values, with the
exception of
8
Be. Here it appears that a decrease in m
q
of 0.5% will lead to
8
Be stable
against breakup into two s for both Hamiltonians. A very weakly bound ( 0.1 MeV)
8
Be
might not have much of an impact on primordial nucleosynthesis because it would be easily
photo-disintegrated until quite late in the BBN epoch. A moderate binding ( 1 MeV)
could have a signicantly more dramatic eect on the chemical evolution of the Universe
by giving rise to the production of noticeable amounts of stable elements with A=9,10. If
the bound state persisted to the era of star formation, it would presumably also have a
signicant eect on stellar evolution.
A summary of sensitivities of nuclear binding energies to the quark mass m
q
, as given
by the DSE hadronic mass variation, is presented in Table V for the dierent Hamiltonians.
The total sensitivity K,
K =
E/E
m
q
/m
q
(41)
of deuterium binding energy to the light quark mass is K
d
= 1.39 for the AV18 interaction,
while the pion contribution ranges from K

d
= 0.70 for the OPE contribution only to
K

d
= 3.36 when the full TPE is counted. This may be compared with the pion contribution
3 > K

d
> 18 from Ref. [9] and K

d
= 2.4 from Ref. [24].
The result of Ref. [10] suggested that a reduced deuteron binding energy of Q =
0.019 0.005 would yield a better t to observational data (the WMAP value of and
18
measured
2
H,
4
He, and
7
Li abundances) for big bang nucleosynthesis. This would correspond
to an increase in the quark mass of X
q
/X
q
= 0.014 0.004 (here X
q
= m
q
/
QCD
).
Dent, Stern, and Wetterich [11] calculated the sensitivity of BBN abundances for
2
H,
4
He
and
7
Li to the variation of binding energies of
2,3
H,
3,4
He,
6,7
Li and
7
Be in a linear approx-
imation. We use the response matrix values in their Table 1 for m
N
to B
7Be
and multiply
by the m
N
/m
N
and K values of our Table V to estimate the sensitivity of BBN yields to
variation of the quark mass. If we compare to the ratio of observation and theoretically
predicted abundances given in their Appendix B, we obtain the following equations for
2
H,
4
He and
7
Li:
1 + 7.7x =
2.8 0.4
2.61 0.04
= 1.07 0.15 , (42)
1 0.95x =
0.249 0.009
0.2478 0.0002
= 1.005 0.036 , (43)
1 50x =
1.5 0.5
4.5 0.4
= 0.33 0.11 , (44)
where x = X
q
/X
q
. These equations yield 3 consistent values of x: 0.009 0.019, 0.005
0.038 and 0.013 0.002. The statistically weighted average of X
q
/X
q
= 0.013 0.002
is dominated by the
7
Li data. A more accurate calculation should take into account the
eect of the
8
Be binding energy variation (which is not calculated in Ref. [11]), the variation
of the virtual
1
S
0
(np) level, and non-linear corrections in x which are important for
7
Li.
Allowing for the theoretical uncertainties we should understand this BBN result as X
q
/X
q
=
K (0.0130.002) where K 1 and the expected accuracy in K is about a factor of 2. Note
that here we neglect eects of the strange quark mass variation. A rough estimate of these
eects on BBN due to the deuteron binding energy variation was made in Refs. [9, 10].
V. CONCLUSIONS
We have argued that there are several reasons to question the spatial and temporal
invariance of various fundamental constants of nature, such as the ne structure constant
or a comparable strong interaction parameter X
q
= m
q
/
QCD
. The search for evidence of
such variations is ongoing in areas as diverse as quasar absorption spectra, the Oklo natural
nuclear reactor, and big bang nucleosynthesis.
In this work we have examined how nuclear binding energies depend on hadronic masses,
including m
N
, m

, m

, and a generic heavy meson m

V
. We have done this by identifying the
19
mass-dependence in several realistic Hamiltonians interactions that t NN elastic scat-
tering data and reproduce light nuclei binding energies reasonably well in quantum Monte
Carlo calculations. By making small changes in the masses and re-evaluating the energy, we
have obtained the dimensionless derivatives of the energy with respect to variations in the
hadronic masses.
We have combined these results with a specic prediction from a Dyson-Schwinger equa-
tion study of sigma terms in the light-quark hadrons for the hadronic mass-dependence on
the quark mass m
q
. The pion mass changes most rapidly with changes in m
q
, so we nd
that both the one- and two-pion exchange parts of the NN interaction are very important
for the consequent variations in nuclear binding. With our most complete model, the
1
S
0
virtual bound state and deuteron vary in concert; if X
q
increases, they both become less
bound, while if X
q
decreases, they both become more bound. (We note that this result is
in disagreement with chiral perturbation results at the NLO level which have
1
S
0
and
3
S
1
scattering varying antithetically.) The binding energies of A=3-8 nuclei behave in the same
manner, all moving up or down together, with a sensitivity K in the range 1 to 1.5.
Finally, we have folded these results with a study of the sensitivity of big bang nucleosyn-
thesis to variations in nuclear binding. We nd that a small increase in the quark mass of
order 1% at the time of BBN is sucient by itself to resolve existing discrepancies between
theoretical and measured abundances of
2
H,
4
He, and
7
Li.
20
Acknowledgments
We thank C.D. Roberts, G.A. Miller, D.R. Phillips, and T. Dent for valuable comments.
This work is supported by the U.S. Department of Energy, Oce of Nuclear Physics, under
contract DE-AC02-06CH11357, and by the Australian Research Council.
[1] W. J. Marciano, Phys. Rev. Lett. 52, 489 (1984); X. Calmet and H. Fritzsch, Eur. Phys. J.
C24, 639 (2002); P. Langacker, G. Segre and M. J. Strassler, Phys. Lett. B528, 121 (2002);
T. Dent and M. Fairbairn. Nucl. Phys. B653, 256 (2003); C. Wetterich, JCAP 10, 002 (2003);
Phys. Lett. B561, 10 (2003).
[2] J.-P. Uzan, Rev. Mod. Phys. 75, 403 (2003).
[3] T. Damour and K. Nordtvedt, Phys. Rev. Lett. 70, 2217 (1993); Phys. Rev. D 48, 3436
(1993).
[4] T. Damour and A. M. Polyakov, Nucl. Phys. B423, 532 (1994).
[5] J. D. Bekenstein, Phys. Rev. D 25, 1527 (1982).
[6] C. Wetterich, Nucl. Phys. B302, 645 (1988).
[7] H. B. Sandvik, J. D. Barrow, and J. Magueijo, Phys. Rev. Lett 88, 031302 (2002).
[8] K. A. Olive and M. Pospelov, Phys. Rev. D 65, 085044 (2002).
[9] V. V. Flambaum and E. V. Shuryak, Phys. Rev. D 65, 103503 (2002); V. F. Dmitriev and V.
V. Flambaum, Phys. Rev. D 67, 063513 (2003); V. V. Flambaum and E. V. Shuryak, Phys.
Rev. D 67, 083507 (2003).
[10] V. F. Dmitriev, V. V. Flambaum, and J. K. Webb, Phys. Rev. D 69, 063506 (2004).
[11] T. Dent, S. Stern, C. Wetterich, arXiv:0705.0696v2 [astro-ph].
[12] A. Coc, N. J. Nunes, K. A. Olive, J.-P. Uzan, and E. Vangioni, Phys. Rev. D 76, 023511
(2007).
[13] J. P. Kneller and G. C. McLaughlin, Phys. Rev. D 68, 103508 (2003).
[14] S. K. Lamoreaux and J. R. Torgerson, Phys. Rev. D 69, 121701(R) (2004).
[15] C. R. Gould, E. I. Sharapov, S. K. Lamoreaux, Phys. Rev. C 74, 024607 (2006); Yu. V. Petrov,
A. I. Nazarov, M. S. Onegin, V. Yu. Petrov, and E. G. Sakhnovsky, Phys. Rev. C 74, 064610
(2006); Y. Fujii, A. Iwamoto, T. Fukahori, T. Ohnuki, M. Nakagawa, H. Hidaka, Y. Oura,
21
and P. Moller, Nucl. Phys. B573, 377 (2000).
[16] A. Ivanchik, P. Petitjean, D. Varshalovich, B. Aracil, R. Srianand, H. Chand, C. Ledoux,
and P. Boisse, Astron. Astrophys. 440, 45 (2005); E. Reinhold, R. Buning, U. Hollenstein, A.
Ivanchik, P. Petitjean, and W. Ubachs, Phys. Rev. Lett. 96, 151101 (2006).
[17] P. Tzanavaris, J. K. Webb, M. T. Murphy, V. V. Flambaum, and S. J. Curran, Phys. Rev.
Lett. 95, 041301 (2005).
[18] V. V. Flambaum and M. G. Kozlov, Phys. Rev. Lett. 98, 240801 (2007).
[19] V. V. Flambaum and A. F. Tedesco, Phys.Rev. C 73, 055501 (2006).
[20] V. V. Flambaum, Phys. Rev. Lett. 97, 092502 (2006).
[21] E. Peik and Chr. Tamm, Europhys. Lett. 61, 181 (2003).
[22] V. V. Flambaum, arXiv:0705.3704v2 [physics.atom-ph]
[23] S. R. Beane and M. J. Savage, Nucl. Phys. A713, 148 (2003).
[24] E. Epelbaum, Ulf-G. Meiner, and W. Gl ockle, Nucl. Phys. A714, 535 (2003).
[25] V. V. Flambaum, D. B. Leinweber, A. W. Thomas, and R. D. Young, Phys. Rev. D 69, 115006
(2004).
[26] W. Armour, C. R. Allton, D. B. Leinweber, A. W. Thomas, and R. D. Young, J. Phys. G 32,
971 (2006).
[27] V. V. Flambaum, A. Holl, P. Jaikumar, C. D. Roberts, and S. V. Wright, Few-Body Syst. 38,
31 (2006).
[28] A. Holl, P. Maris, C. D. Roberts, and S. V. Wright, arXiv:nucl-th/0512048v1.
[29] R. B. Wiringa and S. C. Pieper, Phys. Rev. Lett. 89, 182501 (2002).
[30] S. R. Beane, P.F. Bedaque, K. Orginos, and M. J. Savage, Phys. Rev. Lett. 97, 012001 (2006).
[31] R. B. Wiringa, R. A. Smith, and T. L. Ainsworth, Phys. Rev. C 29, 1207 (1984).
[32] R. B. Wiringa, V. G. J. Stoks, and R. Schiavilla, Phys. Rev. C 51, 38 (1995).
[33] B. S. Pudliner, V. R. Pandharipande, J. Carlson, and R. B. Wiringa, Phys. Rev. Lett. 74,
4396 (1995).
[34] R. A. Arndt and L. D. Roper, Scattering Analyses Interactive Dial-in (SAID) program of
the Center for Analysis of Particle Scattering, Department of Physics, Virginia Polytechnic
Institute and State University.
[35] A. Picklesimer, R. A. Rice, and R. Brandenburg Phys. Rev. C 46, 1178 (1992).
[36] R. B. Wiringa, Lect. N. Phys. 198, 44 (1984).
22
[37] R. B. Wiringa, Phys. Rev. C 43, 1585 (1991).
[38] S. C. Pieper, R. B. Wiringa, and V. R. Pandharipande, Phys. Rev. C 46, 1741 (1992).
[39] R. B. Wiringa, V. Fiks, and A. Fabrocini, Phys. Rev. C 38, 1010 (1988).
[40] V. G. J. Stoks, R. A. M. Klomp, M. C. M. Rentmeester and J. J. de Swart, Phys. Rev. C 48,
792 (1993).
[41] S. C. Pieper and R. B. Wiringa, Annu. Rev. Nucl. Part. Sci. 51, 53 (2001).
[42] A. Akmal and V. R. Pandharipande, Phys. Rev. C 56, 2261 (1997).
[43] J. L. Forest, V. R. Pandharipande, and J. L. Friar, Phys. Rev. C 52, 568 (1995).
[44] R. B. Wiringa, Phys. Rev. C 73, 034317 (2006).
[45] E. Epelbaum, H.-W. Hammer, Ulf-G. Meiner, A. Nogga, Eur.Phys.J. C48, 169 (2006).
[46] H.-W. Hammer, D.R. Phillips, and L. Platter, Eur.Phys.J. A32, 335 (2007).
23
TABLE I: Ground state energies of light nuclei in MeV for the dierent Hamiltonians used in this
work compared to experiment.
1
S
0
(np)
2
H
3
H
3
He
4
He
5
He
6
Li
7
Li
7
Be
8
Be
AV28 0.0661 2.2250
AV14 0.0663 2.2250 7.50 6.88 23.60 21.26 24.31 28.31 26.85 40.26
AV18+UIX 0.0665 2.2246 8.25 7.49 27.50 25.26 28.22 33.33 31.74 48.50
Expt. 2.2246 8.48 7.72 28.30 27.41 31.99 39.24 37.60 56.50
TABLE II: Dimensionless derivatives E(m
H
) =
E/E
m
H
/m
H
of the energy for the
1
S
0
(np) virtual
bound state
v
and the deuteron Q for all three Hamiltonians.
m
H

v
Q
AV28 AV14 AV18 AV28 AV14 AV18
m
N
88.1 32.6 33.4 13.06 8.63 8.90
m
N
+
N
91.2 121.2 13.03 17.82
m

63.9 5.15

68.1 102.2 5.12 10.36

m

(OPE) 9.5 4.1 3.8 2.23 1.55 1.40

m

(+TPE-s) 24.4 35.5 53.0 3.63 4.02 6.70

m

(+TPE-L) 4.02 4.31 6.74

m
V
153.7 245.0 381.9 20.88 22.92 41.74
24
TABLE III: E(m
H
) for the AV14 Hamiltonian.
3
H
3
He
4
He
5
He
6
Li
7
Li
7
Be
8
Be
m
N
6.00 6.44 3.97 4.58 5.25 5.60 5.88 5.10
m
N
+
N
12.32 13.17 9.03 10.38 11.35 12.74 13.41 11.71

7.35 7.82 5.89 6.74 7.10 8.31 8.76 7.69

m

(OPE) 0.45 0.50 0.20 0.24 0.36 0.30 0.32 0.23

m

(+TPE-s) 4.35 4.66 3.33 3.87 4.19 4.83 5.09 4.38

m

(+TPE-L) 4.53 4.85 3.47 4.04 4.40 5.06 5.34 4.59

m
V
29.36 31.30 23.60 27.09 28.98 33.72 35.55 30.98
TABLE IV: E(m
H
) for the AV18+UIX Hamiltonian.
3
H
3
He
4
He
5
He
6
Li
7
Li
7
Be
8
Be
m
N
6.07 6.54 3.99 4.51 5.12 5.24 5.49 4.81
m
N
+
N
16.56 17.73 11.86 13.31 14.41 15.53 16.29 14.36

12.20 13.02 9.16 10.24 10.80 11.96 12.56 11.11

m

(OPE) 0.37 0.42 0.19 0.24 0.36 0.29 0.30 0.23

m

(+TPE-s) 6.90 7.38 5.11 5.82 6.33 6.95 7.30 6.34

m

(+TPE-L) 6.87 7.36 5.06 5.75 6.24 6.84 7.18 6.24

m

(+TNI) 6.91 7.40 5.12 5.82 6.31 6.91 7.26 6.31

m
V
47.98 51.23 36.34 40.87 43.48 48.11 50.53 44.40
TABLE V: Dimensionless derivatives K =
E/E
mq/mq
of the energy over light quark mass m
q
for the
dierent Hamiltonians.
1
S
0
(np)
2
H
3
H
3
He
4
He
5
He
6
Li
7
Li
7
Be
8
Be
AV28 4.5 -0.75
AV14 7.3 -0.84 -0.89 -0.96 -0.69 -0.81 -0.89 -1.03 -1.09 -0.92
AV18+UIX 11.4 -1.39 -1.44 -1.55 -1.08 -1.24 -1.36 -1.50 -1.57 -1.35
25
0 0.5 1 1.5 2
-6
-4
-2
0
2
m
q
/ m
q0
E

(
M
e
V
)
AV28
AV14
AV18
2
H
1
S
Fig.1 (Flambaum & Wiringa)
FIG. 1: (Color online) Variation of two-nucleon energies with current-quark mass variation from
DSE calculation: full calculation (solid lines) and with OPE modication only (dashed lines)
for three dierent Argonne Hamiltonians. Virtual bound state energies are plotted as positive
quantities; m
q0
is the physical current-quark mass.
26
0.8 0.9 1 1.1 1.2
-70
-60
-50
-40
-30
-20
-10
0
m
q
/ m
q0
E

(
M
e
V
)
1
S
2
H
3
H
3
He
4
He
5
He
2
6
Li
7
Li
7
Be
8
Be
Fig.2 (Flambaum & Wiringa)
FIG. 2: (Color online) Variation of multi-nucleon energies with current-quark mass variation from
DSE calculation for AV14 Hamiltonian.
27
0.8 0.9 1 1.1 1.2
-70
-60
-50
-40
-30
-20
-10
0
m
q
/ m
q0
E

(
M
e
V
)
1
S
2
H
3
H
3
He
4
He
5
He
2
6
Li
7
Li
7
Be
8
Be
Fig.3 (Flambaum & Wiringa)
FIG. 3: (Color online) Variation of nuclear energies with current-quark mass variation from DSE
calculation for AV18+UIX Hamiltonian.
28