374 E.A. Gallegos / Physics Letters B 793 (2019) 372–382
δϕ =−ψ
δψ =
F −i
∂
0
ϕ −ia
∂
i
ϕ (3)
δ F =−i
∂
0
ψ −ia
∂
i
ψ,
where is a Grassmann x-independent parameter.
Before
proceeding with the construction of L
int
, it is necessary
to state clearly the weighted renormalization condition (wrc) [1,2].
If one writes L
int
of a given theory as L
int
=
i
g
i
V
i
, where g
i
la-
bel
the coupling constants and V
i
the interaction vertices, this con-
dition
simply says that the theory is renormalizable by weighted
power counting iff the weighted scaling dimension, the weight
for short,
[
g
i
]
w
of each coupling constant g
i
is greater or equal
to zero, i.e.,
[
g
i
]
w
≥ 0. Setting
x
0
=t
w
=−z and
x
i
w
=−1,
the weight
[
O
]
w
of any object O is determined by enforcing
the action S to be weightless. In terms of the vertices, since the
Lagrangian L weighs d + z, where d denotes the spatial dimen-
sions,
the wrc asserts that a vertex V
i
is weighted renormalizable
iff
[
V
i
]
w
≤ d + z. Following the nomenclature adopted in the lit-
erature,
we call a vertex weighted marginal when this weighs
[
V
i
]
w
=d + z, weighted relevant when
[
V
i
]
w
< d + z and weighted
irrelevant when
[
V
i
]
w
> d + z .
It
should be noted that the weighted assignment in Lifshitz field
theories is equivalent to demand the invariance of the action under
the following anisotropic scale transformations,
x
i
→ξ x
i
t → ξ
z
t, (4)
where z is the well-known critical exponent which measures the
degree of anisotropy between space and time. Moreover, notice
that the weighted scaling dimension coincides with the usual mass
one in natural units (
¯
h = 1 = c) when z = 1.
To
construct L
int
we first observe that any vertex V in it must
have the structure V ∼
(
∂
i
)
N
∂
i
ψ
N
ψ
¯
ψ
N
¯
ψ
ϕ
N
ϕ
¯
ϕ
N
¯
ϕ
F
N
F
¯
F
N
¯
F
, where N
p
represents the number of objects of the p type. Since the fields in
the action (2)for arbitrary z weigh
[
ϕ
]
w
=
(
2 − z
)
/
2,
[
ψ
α
]
w
= 1,
and
[
F
]
w
=
(
2 + z
)
/
2, one easily sees by imposing the wrc which
V is renormalizable by weighted power counting iff the condition
N
∂
i
+N
ψ
+
(
2 − z
)
2
N
ϕ
+
(
2 + z
)
2
N
F
≤2 + z, (5)
is satisfied. Here N
X
= N
X
+ N
¯
X
, with X = ϕ, F , ψ . This condi-
tion
along with those which result from imposing the symmetry
requirements restrict strongly the form of L
int
. Note in particu-
lar
that the spatial rotational SO
(
2
)
symmetry of the Lagrangian
L demands a complete spinor/spatial index contraction as well as
that N
ψ
= N
ψ
+N
¯
ψ
=even number. The rigid phase U (1) symme-
try,
on the other hand, implies that N
ψ
+N
ϕ
+N
F
= N
¯
ψ
+N
¯
ϕ
+N
¯
F
.
It should be noted that whether the polynomiality requirement of
the Lagrangian L,
[
ϕ
]
w
≥0, were used, this states an upper bound
in the value of the critical z exponent, namely z = 2.
Hereafter,
we particularize the condition (5)to the case z = 2
so
that this becomes N
∂
i
+N
ψ
+2N
F
≤ 4. Since N
ϕ
does not ap-
pear
explicitly in this inequality, any number of scalar ϕ lines is
allowed in a given vertex V from the weighted renormalization
viewpoint. It should be noted also that the maximal number of
spatial derivatives, N
∂
i
= 4, can only occur in pure scalar ϕ ver-
tices:
V ∼ ∂
i
∂
j
¯
ϕ∂
i
∂
j
ϕ, ∂
i
∂
j
¯
ϕ∂
i
∂
j
ϕ
¯
ϕϕ, ....
The
presence of spatial derivatives in the interaction vertices,
in particular, in genuine ones (i.e. vertices with more than two
lines) made the problem of finding the most general interaction
Lagrangian L
int
extremely intricate in the component formalism.
Hence, for simplicity, we will seek L
int
with the following struc-
ture
L
int
=
¯
ψ W
1
ψ +
1
2
¯
ψ
W
2
¯
ψ +
W
3
¯
F +h.c.
+
V
¯
ψ
2
ψ
2
+ U, (6)
where W
i
, V and U are functions of the scalar fields ϕ,
¯
ϕ. Note
that due to the Hermiticity property of the action (2), W
1
, V and
U have to be real operators, whereas W
2
and W
3
complex ones.
Furthermore, it is worthwhile to note that in conformity with the
z =2wrc
N
∂
i
+N
ψ
+2N
F
≤4
, W
i
can contain at most two spa-
tial
derivatives
N
∂
i
=2
, U four spatial derivatives
N
∂
i
=4
, and
V none
N
∂
i
=0
. Since we are interested in adding higher spatial
derivatives in the kinetic part of (2), we shall look for expressions
of the form W
1,2
∼ +
˜
W
1,2
(
ϕ,
¯
ϕ
)
, W
3
∼
(
ϕ or
¯
ϕ
)
+
˜
W
3
(
ϕ,
¯
ϕ
)
,
and U ∼
¯
ϕ
2
ϕ +
˜
U
(
ϕ,
¯
ϕ
)
, where =∂
i
∂
i
is the two-dimensional
spatial Laplace operator and where
˜
W
i
,
˜
U are functions only of
the scalar fields ϕ,
¯
ϕ. With these considerations in mind, one can
easily compute the variation of L
int
under the susy transforma-
tions (3).
This can be cast in the form
δL
int
=ψ
¯
F +
¯
ψ
+ V ψ
2
¯
F −iγ
0
¯
ψ
(
∂
0
ϕ +∂
0
¯
ϕ
)
−
iaγ
i
¯
ψ
(
∂
i
ϕ +∂
i
¯
ϕ
)
+
¯
ψψ
2
∂
ϕ
W
1
−∂
¯
ϕ
¯
W
2
−
iV ψ
2
γ
0
¯
ψ∂
0
¯
ϕ +aγ
i
¯
ψ∂
i
¯
ϕ
−
ψ∂
ϕ
U + h.c., (7)
where we have introduced the notation = W
1
−∂
ϕ
W
3
and =
W
2
− ∂
¯
ϕ
W
3
, and got rid of all surface terms. From this result, it
is clear that to respect susy one must demand that = = V =
U = 0. This in turn implies that
W
1
=
∂
W
3
∂ϕ
W
2
=
∂
W
3
∂
¯
ϕ
. (8)
The operator W
3
, on the other hand, is fully determined by im-
posing
the reality condition of W
1
, ∂ W
3
/∂ϕ = ∂ W
3
/∂
¯
ϕ, and the
fact that its weight according to the z = 2wrc is
[
W
3
]
w
= 2. This
means, as stated previously, that W
3
cannot contain more than
two spatial derivatives. After doing this, one gets
W
3
=mϕ +bϕ +
∞
p=1
g
p
ϕ
p+1
¯
ϕ
p
. (9)
Notice, moreover, that susy excludes any possibility of introducing
explicitly four spatial derivatives
N
∂
i
=4
in the off-shell formu-
lation
of the theory. Indeed, the desired Lifshitz scalar operator
¯
ϕ
2
ϕ appears only in the bosonic sector of the theory after elim-
inating
the auxiliary F field (see Eq. (12)). In the concluding part
of this section, we will confirm these results and generalize them
in the Lifshitz superfield formulation.
It
is extremely important to expose the consistency of our
model by setting explicitly up its superalgebra at classical level.
Hence, in the balance of this section, we construct the Noether
currents (and their respective charges) associated with each sym-
metry
of the model under consideration and then we set up its
superalgebra by using the canonical (anti-)commutation relations.
For
simplicity and without loss of generality, we truncate the
series in (9)at p = 1. So, the interaction Lagrangian L
int
of our
Lifshitz-Wess-Zumino (L-WZ) model, as we shall call it, reads
L
int
=
¯
ψ
(
m +b +2g
¯
ϕϕ
)
ψ
+
gϕ
2
¯
ψ
2
+
mϕ +bϕ + gϕ
2
¯
ϕ
¯
F +h.c.
.
(10)
Notice that setting a → 1 and b → 0this theory reduces to the
usual relativistic Wess-Zumino model [15,16].
The
L-WZ field equations which result from the principle of
least action, δ S =0, are given by