a

r
X
i
v
:
1
4
0
5
.
5
6
8
7
v
1


[
h
e
p
-
t
h
]


2
2

M
a
y

2
0
1
4
Entropy current for non-relativistic uid
Nabamita Banerjee

Indian Association for the Cultivation of Science, Jadavpur, Kolkata

Suvankar Dutta,

Akash Jain,

and Dibakar Roychowdhury

Dept. of Physics, Indian Institute of Science Education and Research Bhopal, Bhopal
We study transport properties of a parity-odd, non-relativistic charged uid in presence of back-
ground electric and magnetic elds. To obtain stress tensor and charged current for the non-
relativistic system we start with the most generic relativistic uid, living in one higher dimension
and reduce the constituent equations along the light-cone direction. We also reduce the equation
satised by the entropy current of the relativistic theory and obtain a consistent entropy current for
the non-relativistic system (we call it canonical form of the entropy current). Demanding that the
non-relativistic uid satises the second law of thermodynamics we impose constraints on various
rst order transport coecients. For parity even uid, this is straight forward; it tells us positive
deniteness of dierent transport coecients like viscosity, thermal conductivity, electric conduc-
tivity etc. However for parity-odd uid, canonical form of the entropy current fails to conrm the
second law of thermodynamics. Therefore, we need to add two parity-odd vectors to the entropy
current with arbitrary coecients. Upon demanding the validity of second law, we see that one can
x these two coecients exactly.
CONTENTS
I. Introduction and Summary 1
II. Relativistic anomalous charged uid 3
A. Anomalous charged uid: energy-momentum
tensor and current 4
B. Entropy Current 5
III. Non-relativistic charged uid 5
A. Non-relativistic dissipative uid in
background electromagnetic eld 6
IV. Light-cone reduction 7
A. Reduction of background elds 7
B. Reduction of energy-momentum tensor and
charge current 8
1. T
++
calculation 8
2. T
i+
calculation 8
3. T
ij
calculation 9
4. T
+
calculation 9
5. T
i
calculation 9
6. J
+
calculation 9
7. J
i
calculation 10
C. Thermodynamics of the reduced system 11
V. Non-relativistic Entropy Current 11
A. Parity-odd corrections to NR Entropy
Current 12
B. Incompressible uid in constant magnetic
background in (2+1)-dim 13

E-mail: tpnb@iacs.res.in

E-mail: suvankar@iiserb.ac.in

E-mail: ajainphysics@gmail.com

E-mail: dibakarphys@gmail.com
VI. Discussion 14
A. Magnetic limit of electrodynamics 15
B. 1/c expansion of relativistic uid dynamics 16
1. 1/c Expansion of Thermodynamics 17
References 17
I. INTRODUCTION AND SUMMARY
Hydrodynamics is an eective description of nearly
equilibrium interacting many body system. A uid sys-
tem is considered to be continuous, i.e. when we talk
about a small volume element (or uid particle) of the
uid, it still contains a large number of constituent parti-
cles (atoms/molecules). More specically, the size of the
uid particle is much much greater than the mean free
path of the system. The equations of hydrodynamics as-
sume that the uid is in local thermodynamic equilibrium
at each point in space and time, even though dierent
thermodynamic quantities like uid velocity v(x, t), en-
ergy density (x, t), pressure p(x, t), uid density (x, t),
temperature (x, t) etc. may vary. Fluid mechanics ap-
plies only when the length scales of variation of ther-
modynamic variables are large compared to equilibration
length scale of the uid, namely the mean free path [1].
Although hydrodynamics is an old and well-studied
subject in physics, recently there has been much inter-
est and progress in relativistic, charged, viscous uid in
presence of some global anomaly. The rst evidence of
relativistic anomalous uids was observed in [2, 3], in the
context of the AdS/CFT correspondence. In these pa-
pers the authors found a new term (and hence a new
transport coecient) in the charge current in presence of
a Chern-Simons term in the bulk Lagrangian. Later, in
2
[4] it has been shown that this kind of term in charge
current is not only allowed by the symmetry but is re-
quired by triangle anomalies and the second law of ther-
modynamics. Demanding the positivity of local entropy
current they showed that the coecient of this new term
is also xed in terms of anomaly coecient of the the-
ory. In general, the second law of thermodynamics (or
equivalently, positivity of local entropy current) imposes
restrictions on dierent transport coecients. Interested
readers can look at [512].
This was about relativistic uids. Much attention has
not been paid in parity-odd charged non-relativistic uid
in presence of background electromagnetic elds. In [13]
an attempt has been made to study non-relativistic neu-
tral uids as a consistent light-cone reduction of rela-
tivistic uid systems. In [14], the idea has been extended
to charged uid without any background eld. However,
main focus of these two papers was to construct the holo-
graphic duals of long wave length uctuations (hydrody-
namic limit) of conformal non-relativistic eld theories.
They also computed few transport coecients (for exam-
ple thermal conductivity) holographically.
Recently in [15], non-relativistic, parity-odd, rst or-
der charged uid in 2 dimensions has been discussed.
The authors started with a 2+1 dimensional relativistic,
parity-odd charged uid in presence of background gauge
elds and took suitable non-relativistic limit of the sys-
tem. They identied the parity-violating contributions
to the non-relativistic constitutive relations, which in-
clude the Hall current owing perpendicular to the tem-
perature gradient, the Hall viscosity and the Leduc-Righi
energy current
1
. However, in these papers, much atten-
tion has not been paid to second law of thermodynam-
ics. It is well known that the local second law of ther-
modynamics imposes several constraints on the ow and
the transport coecients. For example, the local second
law demands that shear, bulk viscosity coecients and
thermal conductivity have to be positive denite for a
non-relativistic uid [1]. This particular issue, for rela-
tivistic uid, has been discussed vividly in various papers
mentioned above. However, the constraints from second
law of thermodynamics are not yet well understood for
a general class of non-relativistic uids (charged uid in
presence of background elds). Goal of this paper is to
write down a local entropy current, which satises sec-
ond law of thermodynamics, for a non-relativistic charged
uid in presence of arbitrary background electromagnetic
eld and to study the constraints on the transport coef-
cients.
One way to obtain the energy momentum tensor,
charge current and entropy current for a non-relativistic
theory is to reduce the constitutive equations and
entropy current equation of a relativistic uid over a
light-cone direction. We start with a (3 +1)-dimensional
1
See also [16].
relativistic charged uid with global U(1)
3
anomaly
and write down their non-relativistic counterparts. We
follow the light-cone reduction technique, which was
introduced in [13], and shall as well review it in our
paper. Let us conclude this section with an outline and
summary of our main results.
Outline and Summary :
In Section (II) we consider rst order, relativistic,
charged anomalous uid in (3 + 1) dimensions in
presence of background electromagnetic elds. We
briey review the work of [4] and show how parity-
odd transport coecients are related to the coe-
cient of anomaly from the positivity of divergence
of local entropy current.
In Section (III) we highlight the important equa-
tions of non-relativistic hydrodynamic system.
In Section (IV) we consistently obtain the energy
current density (j
i
), stress tensor (t
ij
) and charge
current density (j
i
I
) of non-relativistic uid by
Light Cone Reduction (LCR). We also nd dier-
ent non-relativistic uid variables (thermodynamic
quantities and transport coecients) in terms of
the relativistic uid variables. Here we mention
the nal expression for stress tensor, energy cur-
rent and charge current:
t
ij
= v
i
v
j
+ pg
ij
n
ij
zg
ij

k
v
k
,
j
i
= v
i
_
+ p +
1
2
v
2
_
n
ik
v
k
zv
i

k
v
k

i

I

i
_

_
+
I
(
i
I
v
j

ji
I
),
j
i
I
= q
I
v
i

I

IJ

i
_

_
m
I

i
p
+
IJ
_

i
J
v
k

ki
J
_
+
_

I

ij

j
+

IJ

ij

j
_

_
m
I

ij

j
p +
IJ

ij
_

Jj
v
k

Jkj
_
_
. (I.1)
All the non-relativistic transport coecients are
xed in terms of relativistic transport coecients
and other thermodynamic variables. Explicit rela-
tions have been provided in the main text.
In charge current we identify the thermal Hall con-
tribution
I

ij

j
(where is the temperature of
non-relativistic uid). The Hall energy ow sourced
by the temperature gradient is known as the Leduc-
Righi eect. It has been predicted in condensed
matter physics, that this kind of eect can be ob-
served in various topological insulators.
We also obtain the electromagnetic Hall energy cur-
rent (parity-odd)
IJ

ij
_

Jj
v
k

Jkj
_
, where
Jj
is the applied electric eld and
Iij
is proportional
to the applied magnetic eld (check Eqn. (III.49)
for exact denitions).
3
We nd that the non-relativistic uid system satis-
es the famous Wiedemann-Franz Law.
Finally, we compute the entropy current j
i
S
for the
non-relativistic uid in Section (V), and demand
that the rate of entropy production of the system
is always positive denite. For parity-even uid
it turns out to be quite straight forward. The
canonical form of the entropy current conrms that
the total entropy always increases due to dissipa-
tion, thermal and charge conduction. However for
parity-odd uid, canonical form of the entropy cur-
rent fails to conrm the second law of thermody-
namics. Therefore, we need to add two parity-odd
vectors to the entropy current with arbitrary co-
ecients. Upon demanding the validity of second
law, we see that one can x these two coecients
exactly.
Finally we observe that, unless the uid is incom-
pressible, all the parity-odd transport coecients
must vanish to satisfy the second law of thermody-
namics.
II. RELATIVISTIC ANOMALOUS CHARGED
FLUID
Hydrodynamic description does not follow from the ac-
tion principle, rather it is generally formulated in the lan-
guage of equations of motion. The reason for this is the
presence of dissipation in thermal media. Due to inter-
nal friction called viscosity, a dissipative uid loses its
energy over time as it propagates. The uid without any
viscous drag is called an ideal uid. In the simplest case,
the hydrodynamic equations (for a uncharged uid) are
the laws of conservation of energy and momentum,

= 0. (II.2)
A relativistic uid in (d + 1, 1) dimensions (d + 1 spa-
tial and 1 time dimensions) has total d + 3 independent
intensive uid variables: temperature (T), pressure (P)
and velocities u

(only d + 1 components of the velocity

are independent due to normalization u

= 1). Ex-
tensive variables energy density (E) and entropy density
(S) are not considered to be independent as they can be
determined in terms of the intensive variables, by the rst
law of thermodynamics and the equation of state.
Additionally we have an Eulers relation which follows
from the extensivity of the internal energy:
E + P = TS, (II.3)
which in conjunction with the rst law relates P and T
as:
dP = SdT. (II.4)
Therefore, among these thermodynamic variables, we can
consider only one to be independent, which we choose to
be T. Thus, uid is determined in terms of d + 2 vari-
ables: u

, T. On the other hand we have d + 2 constitu-

tive equations (II.2), and hence uid system is completely
determined. We can express the energy momentum ten-
sor T

as a function of temperature, velocity and their

derivatives.
In case of uid with multiple conserved U(1) currents,
we have more uid variables: charge densities (Q
I
).
The corresponding conjugate variables are called chemi-
cal potentials denoted by (M
I
). In presence of conserved
charges the rst law is modied to
dE = TdS + M
I
dQ
I
. (II.5)
From this equation one can compute the chemical poten-
tials of the uid in terms of charges and other variables.
However, in our computation we consider chemical poten-
tials M
I
to be our independent variables, not the charges.
As a result, number of uid variables have been increased
to d+3. However, in this case we also have another con-
stitutive equation (which follows from the global U(1)
symmetries),

I
= 0. (II.6)
Eulers relations are also modied to:
E + P = TS + M
I
Q
I
, (II.7)
and from the rst law one can write,
dP = SdT + Q
I
dM
I
. (II.8)
Therefore, for charged uid we consider our basic ther-
modynamic variables to be T and M
I
/T, rest are xed
in terms of these two variables. We can therefore write
the last equation in the following form:
dP =
E + P
T
dT + TQ
I
d
_
M
I
T
_
. (II.9)
Thus, for a charged uid, we express energy-momentum
tensor and charge current as a function of uid velocities,
temperature, chemical potential and their derivatives.
One striking feature of a relativistic quantum eld
theory is triangle anomalies. The eect of this anomaly
is reected in three point correlation function of charge
currents. However, the anomaly does not aect the
conservation of the current associated with a global
symmetry. When we put the theory in external back-
ground gauge elds coupled to the currents, some of the
currents will no longer be conserved.
In the next subsection we discuss about the form of
the energy momentum tensor and global U(1) current of
a relativistic anomalous uid in presence of background
gauge elds.
4
A. Anomalous charged uid: energy-momentum
tensor and current
We review a generic charged relativistic uid in (d+2)-
dim with anomalies. The energy-momentum tensor is
given by:
T

= (E + P)u

+ Pg

, (II.10)
where,

= 2

, (II.11)
and uid velocities are normalized
u

= 1. (II.12)
Here, and are relativistic shear and bulk viscosity
coecients respectively. Up to rst order in derivatives,
the respective terms are given by:

=
1
2
P

2
d + 1
g

_
,
(II.13)

= P

, =

, (II.14)
where we use the projection operator:
P

= g

+ u

. (II.15)
Similarly charge current of a relativistic uid in pres-
ence of multiple charges Q
I
(I = 1, 2, 3) is given by
2
J

I
= Q
I
u

I
, (II.16)
where,

I
=
IJ
P

_
M
J
T
_
+
IJ
E

J

I
P

T
+
_

I
l

IJ
B

J
_
, (II.17)
l

. (II.18)
Here
IJ
,
IJ
and
I
are charge, electric and thermal
3
conductivities respectively. However, demanding the pos-
itivity of local entropy current one can show show that,

I
vanishes and
IJ
is related to
IJ
.
We have kept the uid in some background electro-
magnetic gauge elds given by A

I
:
F

I
=

I
, (II.19)
2
See [17] for a beautiful review.
3
Thermal conductivity of relativistic theory is fundamentally dif-
ferent from the thermal conductivity of non-relativistic theory.
Entropy positivity demands relativistic thermal conductivity to
vanish, however in non-relativistic limit it origins from elsewhere.
E

I
= F

I
u

, B

I
=
1
2

F
I
. (II.20)
The last two terms of Eqn. (II.17) are the most generic
single-derivative parity-odd modications to the charge
current allowed in Landau frame. These terms are spe-
cic to (3 + 1)-dimensions. In higher dimensions, the
parity odd terms appear only at higher derivative orders.
However, we would like to keep our calculations generic
to (d + 2)-dim, therefore in all expressions, we represent
parity-odd terms in curly braces {. . .}, with an under-
standing that these terms contribute only for d = 2.
We are working in Landau frame, which demands
u

= 0, u

I
= 0. (II.21)
and hence:
u

= Eu

, u

I
= Q
I
. (II.22)
The constitutive equations for this uid are given by:

= F

I
J
I
, (II.23)

I
= {C
IJK
E

J
B
K
} . (II.24)
C
IJK
is called anomaly coecient, which is completely
symmetric in all indices.
Leading order T

and J

must obey the rst order

conservation equations, which can be found trivially us-
ing Eqn. (II.23), (II.24):
u

E = (E + P), (II.25)
P

P Q
I
E

I
= (E + P)u

, (II.26)
u

Q
I
+ Q
I
= {C
IJK
E

J
B
K
} . (II.27)
We have projected the equations along and perpendicular
to the direction of velocity for later convenience. We have
also used E

I
u

= 0, which can be seen trivially. Using

these,

from Eqn. (II.13) can be written as:

=
1
2
_
Y

+Y

2
d + 1
g

Zu

_
, (II.28)
where,
Z =
2
d + 1
u

T
(E + P)
, (II.29)
T = (d + 1)P E, (II.30)
Y

P
E + P
+ Q
I
u

I
E + P
. (II.31)
The trace of the energy-momentum tensor is given by:
T

= T(d + 1). (II.32)

For conformal uids T

= 0, which can be reached by

setting T = = 0.
5
Later, we shall use some relations involving Y

, which
can be found trivially using Eqn. (II.31):
u

=
2u

P
E + P

P
E + P
+
Q
I
E

I
E + P
, (II.33)
u

P
E + P

Q
I
E

I
E + P
. (II.34)
B. Entropy Current
In this sub-section we write down the expression for
entropy current of a relativistic charged anomalous uid
and derive constraints on various transport coecients
which follow from second law of thermodynamics. The
canonical form of relativistic entropy current is given by
J

S
= Su

M
I
T

I
, (II.35)
Using conservation equations Eqn. (II.23), (II.24) and
thermodynamics Eqn. (II.7), (II.8) we can write:

S
=
1
T

+
_
E
I
T

_
M
I
T
__

I
.
(II.36)
Using Eqn. (II.11), (II.17) it can be written (up to second
order in derivative) as

S
=
1
T

+
1
T

2
+
1
T
E
I

IJ
E

J
+
_
P

_
M
I
T
__

IJ
_
P

_
M
J
T
__

1
T
E

I
(
IJ
+ T
JI
)

_
M
J
T
_

1
T

I
E

T +

_
M
I
T
_

I
P

T
+
_
C
KIJ
M
K
T
E
I
B

J
+
_

I
l

IJ
B

J
_
_
1
T
E
I

_
M
I
T
__ _
. (II.37)
Demanding entropy positivity,

S
0 we will get

I
= 0, 0, 0,
IJ
=
1
T

IJ
,

IJ
matrix is positive denite. (II.38)
However the last two terms of Eqn. (II.37) cannot be
made positive denite. One would then expect that co-
ecients of those terms must vanish! But in [4], the au-
thors identied that Eqn (II.36) is not the most generic
expression for the entropy current. They modied the
entropy current with the most generic parity odd vectors
allowed by the symmetry
J

S
J

S
+
_
Dl

+

D
I
B

I
_
. (II.39)
In presence of these two terms the parity-odd part of the
last equation becomes,
_

_
Dl

+

D
I
B

I
_
C
KIJ
M
K
T
E
I
B

J
+
_

I
l

IJ
B

J
_
_
1
T
E
I

_
M
I
T
___
=
_
l

D
2D
E + P

P
I

_
M
I
T
__
+l

E
I
_
D
2Q
I
E + P
2

D
I
+
1
T

I
_
+B

J
_

D
J

D
J
E + P

IJ

_
M
I
T
_
_
+E

I
B
J
_

D
J
Q
I
E + P
+
1
T

IJ
C
KIJ
M
K
T
__
= 0.
(II.40)
Demanding the positivity of divergence of local entropy
current we get,
D
P
=
2D
E + P
,
D
(M
I
/T)
=
I
,

D
J
P
=

D
J
E + P
,


D
J
(M
I
/T)
=

IJ
,
D
2Q
I
E + P
2

D
I
=
1
T

I
,

D
J
Q
I
E + P
=
1
T

IJ
+ C
KIJ
M
K
T
. (II.41)
One can in principle solve the above set of equation to
get relations between D,

D
I
,
I
,

IJ
and C
IJK
.
Finally we have the desired form of entropy current

S
=
1
T

+
1
T

2
+
_
E

I
T
P

_
M
I
T
__

IJ
_
E
J
T
P

_
M
J
T
__
,
(II.42)
which is entirely positive denite.
III. NON-RELATIVISTIC CHARGED FLUID
In this section, we review the properties of
charged non-relativistic dissipative uid living in (d+1)-
dimensions, in presence of some background gauge elds.
6
A non-relativistic uid has following constitutive equa-
tions:
Continuity equation:

t
+
i
(v
i
) = 0, (III.43)
where is density of uid particle.
Momentum conservation equation:

t
(v
j
) +
i
(t
ij
) = 0, (III.44)
where t
ij
is the energy-momentum tensor.
Energy conservation equation:

t
_
+
1
2
v
2
_
+
i
j
i
= 0, (III.45)
where j
i
is the energy current.
Conservation of charge current:

t
q
I
+
i
j
i
I
= 0, (III.46)
where, q
I
s are the U(1) charge densities and

j
I
s are
the corresponding currents. Note that energy current is
denoted by

j where as charge current is denoted by

j
I
.
In d spatial dimensions, charged uid has total d + 3
independent variables (pressure p, temperature , chemi-
cal potentials
I
and velocity v), while others (density ,
energy density , charge density q
I
and entropy density
s) are xed in terms of these using thermodynamic rela-
tions. Unlike the relativistic case we do not consider the
Eulers relation here. Later we shall see that the light-
cone reduction does not preserve extensivity of the uid
system. Hence, the uid is determined by d +3 parame-
ters chosen to be: ,
I
/, p and v (same as (d+1,1)-dim
relativistic uid). On the other hand we have d +3 con-
stitutive equations.
Therefore, we can express energy-momentum tensor
and charge current in terms of ,
I
/, p, v and their
derivatives. For ideal uid they are given by,
t
ij
= v
i
v
j
+ pg
ij
,
j
i
=
_
+ p +
1
2
v
2
_
v
i
,
j
i
I
= q
I
v
i
. (III.47)
A. Non-relativistic dissipative uid in background
electromagnetic eld
When we put a dissipative uid in background electro-
magnetic elds, the above equations are modied. Conti-
nuity equation and charge conservation equation remain
the same but other two equations are modied to

t
(v
j
) +
i
(t
ij
) = q
I

j
I
j
Ii

ij
I
,

t
_
+
1
2
v
2
_
+
i
j
i
= j
i
I

Ii
, (III.48)
where,

i
I
=
i

I

t
a
i
I
,
ij
I
=
i
a
j
I

j
a
i
I
, (III.49)
are electric and magnetic elds respectively.
I
and a
i
I
are scalar and vector potentials respectively.
For dissipative uid the stress-energy tensor is given
by,
t
ij
= v
i
v
j
+ pg
ij
+
ij
, (III.50)
where
ij
is the correction to ideal uid stress tensor due
to dissipation

ij
= n
ij
z
ij

k
v
k
, (III.51)

ij
=
i
v
j
+
j
v
i

ij
2
d

k
v
k
. (III.52)
n and z are the non-relativistic shear and bulk viscosity
coecients respectively.
The energy current for dissipative uid is given by,
j
i
=
_
+ p +
1
2
v
2
_
v
i
+
i
, (III.53)
where,

i
= n
ij
v
j
z
k
v
k
v
i

i
_

_
+
I

i
I

I
v
j

ji
I
. (III.54)
Here ,
I
,
I
and
I
are thermal, charge, electric and
magnetic conductivities respectively. Later we shall see
that, for non-relativistic uid, obtained by light cone
reduction of a relativistic uid,
I
and
I
are related to

I
.
Similarly charge current for a dissipative uid is given
by,
j
i
I
= q
I
v
i
+
i
I
, (III.55)
where,

i
I
=
I

IJ

i
_

_
m
I

i
p +
IJ

i
J

IJ
v
k

ki
J
+
_

I

ij

j
+

IJ

ij

j
_

_
m
I

ij

j
p +
IJ

ij

Jj

IJ

ij
v
k

Jkj
_
. (III.56)

I
,
IJ
,

IJ
, m
I
,
IJ
and
I
,
IJ
,

IJ
, m
I
,
IJ
are some
arbitrary parity-even and parity-odd transport coe-
cients
4
. Again, for a non-relativistic uid, obtained by
4
The sign of these coecients are completely arbitrary for now,
and are chosen keeping in mind later convenience.
7
light-cone reduction these transport coecients are xed
in terms of relativistic transport coecients. Also, they
are constrained when we demand positivity of local en-
tropy current.
IV. LIGHT-CONE REDUCTION
Discrete light cone quantization of a (d+1, 1)-dim rela-
tivistic quantum eld theory (QFT) boils down to a non-
relativistic quantum eld theory in one lower dimension.
Let us consider a QFT in at spacetime with metric,
ds
2
= (dx
0
)
2
+ (dx
d+1
)
2
+
d

i=1
(dx
i
)
2
. (IV.57)
We introduce the light-cone coordinates,
x

=
1

2
_
x
0
x
d+1
_
. (IV.58)
In the light-cone frame the metric can be written as,
ds
2
= 2dx
+
dx

+
d

i=1
(dx
i
)
2
. (IV.59)
Suppose we view the QFT in this light-cone frame and
evolve it in light-cone time x
+
. Then, for xed light-
cone momentum P

, we obtain a system in d +1 dimen-

sions with non-relativistic invariance. This is because
the symmetry algebra of the relativistic theory reduces
to corresponding non-relativistic symmetry algebra upon
light-cone reduction. For example, if the relativistic the-
ory is invariant under Poincare transformation, then the
corresponding algebra (SO(d+1, 1)) reduces to d dimen-
sional Galilean algebra. On the other hand if we consider
a QFT with conformal invariance in (d + 1) + 1 dimen-
sions, then upon light cone reduction the corresponding
algebra (SO(d +2, 2)) reduces to Schrodinger algebra in
d spatial dimensions. This is known as discrete light-cone
quantization of quantum eld theories.
Since hydrodynamics is low energy uctuation of equi-
librium quantum eld theory, light-cone reduction of rel-
ativistic constitutive equations boil down to the non-
relativistic constitutive equations for a uid in one lower
dimension. Relativistic charged uid in (d + 1, 1)-dim,
as we have already discussed, has d + 3 degrees of free-
dom: temperature, chemical potential and normalized
velocities. On the other had, a non-relativistic uid in
d spatial dimensions also has total d + 3 degrees of free-
dom: temperature, pressure, chemical potential and ve-
locities. Our goal is to consider the most generic rela-
tivistic anomalous uid system in presence of background
electromagnetic elds in (3 + 1) dimensions and reduce
the constitutive equation (also equation for entropy cur-
rent) over light-cone coordinates and obtain the corre-
sponding non-relativistic equations (most generic) for a
uid in one lower dimensions. We also nd a mapping
between the degrees of freedom of the (d+2)-dimensional
uid to the degrees of freedom of the (d+1)-dimensional
uid.
We denote the d spatial coordinates with x
i
. Metric
components in light-cone coordinates are given by
g
ij
=
ij
and g
+
= 1, (IV.60)
rest are zero. Gradient operator is given by,

= {
+
,

,
i
} and

= {

,
+
,
i
}.
(IV.61)
As we have seen that not all components of u

, J

I
and
T

are independent, from Eq. (II.22) we get,

u

=
1
2u
+
(1 +u
2
), (IV.62)
u
+
J

I
= Q
I
+ u
i
J
i
I
u

J
+
I
, (IV.63)
u
+
T

= Eu

+ u
i
T
i
u

T
+
. (IV.64)
We shall reduce the theory along the x

direction, and
consider x
+
to be the non-relativistic time. We will con-
sider only solutions to the relativistic equations that do
not depend on x

; that is, all derivatives

vanish.
A. Reduction of background elds
We reduce the Maxwells equations for the background
elds of relativistic uid to get consistent background for
non-relativistic theory. Maxwells equations for relativis-
tic system are given by (we shall assume that the sources
for these background elds are far away from the uid),

I
= 0, (IV.65)
Under light-cone reduction the above equations take
the following form:

2
A
+
I
= 0,

i
_

i
A

I
+
+
A
i
I
_
=
2
+
A
+
I
,

i
_

i
A
j
I

j
A
i
I
_
=
j

+
A
+
I
. (IV.66)
These equations can be identied with source free static
Maxwells equations of a non-relativistic system if we
map
5
:
A

I
=
I
(scalar potential),
A
i
I
= a
i
I
(vector potential), (IV.67)
and

i
A
+
I
= 0,
+

+
A
+
I
= 0,
i

+
A
+
I
= 0. (IV.68)
5
In Appendix (A) we have discussed how to expand Maxwells
equations in powers of 1/c to get the non-relativistic limit.
8
Which would inturn tell us

+
A
+
I
,
i
A
+
I
= constant. (IV.69)
We would however like A
+
I
to have some nite value at
innity, which will enforce:
A
+
I
= constant. (IV.70)
With this identication, from Eq. (II.20) one can show
that
E
+
I
= 0, E

I
= u
i

i
I
, E
i
I
= u
+

i
I
u
j

ji
I
,
(IV.71)
and
B
+
I
= u
+

ij

ij
I
, B

I
= u

ij

ij
I
2
ij
u
i

j
I
,
B
i
I
= 2u
+

ij

j
I
, (IV.72)
where,
i
I
=
i

I

t
a
i
I
and
ij
I
=
i
a
j
I

j
a
i
I
.
Finally let us check what happens to the gauge freedom
in relativistic side. The gauge transformation is give by,
A

. (IV.73)
In the light-cone coordinates it becomes
A
+
A
+
,
A

+
,
A
i
A
i
+
i
. (IV.74)
Thus we see that the gauge freedom of relativistic the-
ory reduces to the gauge freedom of the non-relativistic
theory. Additionally A
+
, which we have xed to be a
constant, under a gauge choice does not change. Other-
wise, one could perform a local gauge transformation to
modify A
+
to a non constant value, which would then
break our identication.
B. Reduction of energy-momentum tensor and
charge current
The reduction of relativistic equations of energy-
momentum and charge conservation after using Eqn.
(IV.70) are given by:

+
T
++
+
i
T
i+
= 0, (IV.75)

+
T
+
+
i
T
i
= J
i
I
_

i
A

I
+
+
A
Ii
_
, (IV.76)

+
T
+j
+
i
T
ij
= J
+
I
_

+
A
j
I
+
j
A

I
_
J
i
I
_

i
A
j
I

j
A
Ii
_
, (IV.77)

+
J
+
I
+
i
J
i
I
= {C
IJK
E

J
B
K
} = 0. (IV.78)
R.H.S. of Eqn. (IV.78) can be shown to vanish explicitly
after reduction in four dimensions.
These equations reduce to non-relativistic equations
under following identications
T
++
= , T
i+
= v
i
,
T
+
= +
1
2
v
2
, T
i
= j
i
, T
ij
= t
ij
,
J
+
I
= q
I
, J
i
I
= j
i
I
. (IV.79)
The identication also tells that the non-relativistic
charge current is conserved, while the relativistic charge
current was not. We shall now attempt to use this map-
ping to establish relations between relativistic and non-
relativistic parameters.
1. T
++
calculation
Using Eqn. (II.28) and (II.14) we can show:

++
=
1
2
(u
+
)
2
Z,

++
= (u
+
)
2
, (IV.80)
and thus:
T
++
= (E + P)(u
+
)
2
+ (u
+
)
2
(Z ) . (IV.81)
Identifying T
++
=
= (E + P)(u
+
)
2
+ (u
+
)
2
(Z ) . (IV.82)
2. T
i+
calculation
Using Eqn. (II.28) and (II.14) we can show:

i+
=
1
2
Y
i

1
2
Zu
i
u
+
,

i+
= u
i
u
+
, (IV.83)
where,
Y

= Y
+
=
_

u
+

u
+

P
E + P
+ u
+
Q
I
E

I
E + P
_
.
(IV.84)
Using Eqn. (IV.82), we will nd,
T
i+
=
u
i
u
+
Y
i
. (IV.85)
Identifying T
i+
= v
i
v
i
=
u
i
u
+

Y
i
. (IV.86)
9
3. T
ij
calculation
Using Eqn. (II.28) we will nd:

ij
=
1
2
_
Y
ij
+Y
ji

2
d + 1
g
ij
Zu
i
u
j
_
. (IV.87)
Its easy to check that:

E
(E + P)
= =
u

P
(E + P)
+ u
+

k
v
k
, (IV.88)
and thus introducing Z from Eqn. (II.29),

2
(d + 1)
=
1
d
Z u
+
2
d

k
v
k
. (IV.89)
Using the above relations and the identity:
Y
i
= u
+

v
i
+ v
i
Y

, (IV.90)
we nd that:

ij
=
1
2
_
u
+

ij
+ v
i
Y
j
+ v
j
Y
i
+
g
ij
d
Z (u
+
)
2
v
i
v
j
Z
_
,
(IV.91)
where
ij
is given by Eqn. (III.52). Similarly using Eqn.
(II.14) we can reach to:

ij
= g
ij
u

P
(E + P)
+g
ij
u
+

k
v
k
+(u
+
)
2
v
i
v
j
, (IV.92)
and thus:
T
ij
= v
i
v
j
+ g
ij
_
P
_

d
Z
u

P
E + P
__
u
+

ij
u
+
g
ij

k
v
k
. (IV.93)
Identifying T
ij
= t
ij
p = P
_

d
Z
u

P
E + P
_
, (IV.94)
n = u
+
, (IV.95)
z = u
+
. (IV.96)
4. T
+
calculation
Using Eqn. (IV.64), (IV.81) and (IV.85) we can nd:
T
+
=
1
2
(E P)
1
2
(Z ) +
1
2
v
2
. (IV.97)
Identifying T
+
= +
1
2
v
2
=
1
2
(E P)
1
2
(Z ) . (IV.98)
5. T
i
calculation
Using Eqn. (IV.64), (IV.85) and (IV.93) we can nd:
T
i
= v
i
_
+ p +
1
2
v
2
_
n
ik
v
k
zv
i

k
v
k
+
1
(u
+
)
2
Y
i
.
(IV.99)
The last term can be written as:

1
(u
+
)
2
Y
i
=
1
T

i
_
T
u
+
_
u
+
Q
I
T

i
_
M
I
T
_
+ u
+
Q
I

E
i
I
. (IV.100)
We identify (leading order):
=
T
u
+
,
I
=
M
I
u
+
, q
I
= Q
I
u
+
, (IV.101)
and
=

u
+
= 2n
+ p

,
I
= n
q
I

. (IV.102)
With this identication we get,

1
(u
+
)
2
Y
i
=
i

I

i
_

_
+
I
(
i
I
v
j

ji
I
).
(IV.103)
Identifying T
i
= j
i
we have:
j
i
= v
i
_
+ p +
1
2
v
2
_
n
ik
v
k
zv
i

k
v
k

i

I

i
_

_
+
I
(
i
I
v
j

ji
I
).
(IV.104)
Comparing it with (III.54), we nd that we have the ex-
pected form. Also we have estabilished relations between
charge, electric and magnetic conductivities:

=
I
=
I
. (IV.105)
6. J
+
calculation
Using Eqn. (II.17) we can nd

+
I
=
IJ
u
+
u

_
M
J
T
_

I
u
+
u

T
+
_

I
l
+
+

IJ
B
+
J
_
, (IV.106)
l
+
=
+ij
(u
+
)
2

i
v
j
=
ij
(u
+
)
2

i
v
j
. (IV.107)
Using Eqn. (IV.72) we can write

+
= u
+
_

IJ
u

_
M
J
T
_
+
I
u

T
+
_

ij
u
+

i
v
j
+

IJ

ij

ij
J
_
_
. (IV.108)
10
Identifying J
+
I
= q
I
,
q
I
= u
+
Q
I
+ u
+

I
, (IV.109)
where,

I
=
_

IJ
u

_
M
J
T
_
+
I
u

T
+
_

ij
u
+

i
v
j
+

IJ

ij

ij
J
_
_
. (IV.110)
This tells us the sub-leading correction to non-relativistic
charges.
7. J
i
calculation
Using Eqn. (II.17) we can nd

i
I
=
IJ
P
i

_
M
J
T
_
+
IJ
E
i
J

I
P
i

T
+
_

I
l
i
+

IJ
B
i
J
_
. (IV.111)
One can nd using Eqn. (IV.100) that:
_
l
i
_
=
_
v
i
l
+
+ (u
+
)
2

i
_
, (IV.112)
where
_

i
_
=
_

ij
_
1
(u
+
)
2

j
+
q
I
t

j
_

j
p
+
1

Q
J
E
Jj
__
. (IV.113)
Certain velocity dependent terms appear in the above
expression, which can be shown to sum to zero in 2
spatial dimensions.
Using thermodynamic relations Eqn. (II.7), (II.8),
(IV.100), along with all light cone identications, we will
nd
J
i
I
= v
i
J
+
I


I
u
+

2( + p)

i
p
q
I
2( + p)

IJ
+
I
q
J
2( + p)

I
u
+

2
q
J
2( + p)
_

i
_

_
+
_

J
q
I
2( + p)
+
IJ
u
+
_
_

j
J
v
k

kj
J
_

I
(u
+
)
2
2

ij
_

j
p +
_

ij
_

+
_

I
(u
+
)
2
q
J

ij
_

j
_

_
+
_

I
(u
+
)
2
q
J

ij

IJ
u
+
2
ij
_
_

Jj
v
k

Jkj
_
.
(IV.114)
We identify
m
I
=

I
u
+
2( + p)
,

IJ
=
_

IJ
+

J
q
I
2( + p)
m
I
q
J

_
,

I
=
q
I
2( + p)
,
IJ
=
IJ
=
_

IJ
u
+
+

J
q
I
2( + p)
_
,
m
I
=
2
I

IJ
=
J

I
n
,

I
=

I
n
,
IJ
=
IJ
=
_

I
n
2
IJ
_
, (IV.115)
where,

I
=
I
(u
+
)
2
,
IJ
=

IJ
u
+
. (IV.116)
Finally we have the desired form of charge current
j
i
I
= q
I
v
i
+
i
I
, (IV.117)

i
I
=
I

IJ

i
_

_
m
I

i
p
+
IJ
_

i
J
v
k

ki
J
_
+
_

I

ij

j
+

IJ

ij

j
_

_
m
I

ij

j
p +
IJ

ij
_

Jj
v
k

Jkj
_
_
, (IV.118)
Note that all transport coecients are not independent.
The only independent coecients are: m
I
,

IJ
,
IJ
,
I
and
IJ
.
Wiedemann-Franz Law: This famous law predicts
that the ratio of charge conductivity (which appears
in charge current)
6
to thermal conductivity (which ap-
pears in energy current) in metals as: / = 1/L,
where L is the Lorenz number predicted to be 2.45
10
8
WK
2
. The law is found to be in good agreement
with experiments. We attempt to check the same in our
setup
7
:

=

2n( + p)
+
q
2
4( + p)
2
, (IV.119)
We model the electrons in metals as free classical gas with
no external pressure: uid with homogeneous particles
each of charge e (electronic charge), mass m
e
and average
energy 3/2k
B
. One can check that under mentioned
assumptions, our system follows Wiedemann-Franz Law
with Lorenz number given as:
L =
_

n
m
e
3k
B
+
e
2
9k
2
B
_
1
, (IV.120)
Assuming and n of nearly same order, rst term turns
out to be about 15 orders of magnitude smaller than the
second term and can be safely neglected. Hence Lorenz
number is our case is given approximately by: 6.68
10
8
WK
2
, given our assumptions, which is in fair
agreement with the experimental value.
6
Lets consider only one U(1) charge here.
7
We have used here the fact that non-relativistic system respects
the constraint
IJ
=

IJ
, which has been showed in next sec-
tion.
11
C. Thermodynamics of the reduced system
We start with a relativistic system with variables sat-
isfying the Eq. (II.5) and (II.7). The rst equation is the
rst law of thermodynamics and the second equation is
called the Eulers equation. The Euler equation follows
from the additive property of internal energy which is a
homogeneous function of degree one. After light-cone re-
duction when we map non-relativistic variables in terms
of relativistic ones, we see that at the leading order (ideal
uid), non-relativistic variables satisfy the following two
equations,
d = ds +
I
dq
I
+ ( + p s q
I

I
)
d

,
2( + p) = s +
I
q
I
. (IV.121)
Here we have identied the non-relativistic entropy den-
sity s = Su
+
. The rst equation, which follows from
Eq. (II.5), is the rst law of thermodynamics satised by
the non-relativistic variables. The second, coming from
Eq. (II.7), is not exactly the Eulers equation for non-
relativistic variables. We comment on this equation at
the end of this sub-section.
As we consider the rst derivative corrections to non-
relativistic variables (evaluated in the previous sub-
section), the rst law of thermodynamics changes. How-
ever, that can not be the case. If we demand that our
non-relativistic system is a physical one, then the rst law
should be satised by the non-relativistic variables at ev-
ery order in derivative expansion. This demand forces
us to add some higher derivative corrections to entropy
density which we could not determine in Section (IVB).
Lets assume the following corrected forms:
s = Su
+
(1 + ), (IV.122)
is corresponding corrections.
Demanding rst law of thermodynamics to be satised
at rst order in derivative we get:
TSd + M
I
d
I
M
I

I
dE
E + P
M
I
_
dQ
I

Q
I
E + P
dE
_
=

1
d
Z
dT
E + P

dE
E + P
u

E
E + P
+
d(E P)
E + P
u

(E P)
E + P
+
_
ST + M
I

I
+
d + 2
d
Z

_
+ 2
u

P
E + P
__ _
dP
E + P
+
du
+
u
+
_
. (IV.123)
Note that, thermodynamic relations (II.7) - (II.8), con-
servation equations (II.25) - (II.27) up to rst order and
all the identications between non-relativistic and rela-
tivistic variables (obtained by LCR) have been exten-
sively used to carry out the above computation. Now,
using (II.25) - (II.27), we can write the above equation
in the following form
8
TSu

+ M
I
u

I
+ M
I

=
d + 1
2d
Z
2

2
+ (u
+
)
2
_

k
v
k
_
2
. (IV.124)
This equation xes the correction to the entropy density.
The last equation in (IV.121), which follows from Eu-
lers equation (for the relativistic variables) does not im-
ply that the non-relativistic energy density is a homo-
geneous function of degree one. In fact, this equation
gets modied with the corrected values of non-relativistic
entropy density
9
. Therefore, we conclude that the non-
relativistic system we obtain after light-cone reduction
of a homogeneous relativistic uid system, is not homo-
geneous any more (even at the leading order) i.e. the
internal energy is not a homogeneous function of degree
one.
V. NON-RELATIVISTIC ENTROPY CURRENT
The goal of this section is to write down an entropy
current for a non-relativistic uid system, and nd con-
straints on various transport coecients from the positiv-
ity of local entropy production. Using Eqn. (IV.124) and
properties of Y

, one can reduce the Eqn. (II.37) and

after a lengthy algebra, we get the following equation,

+
s+
i
j
i
S
+
i
_

(u
+
)
2
Y
i
_
=
1

n
2

ij

ij
+
1

z(
k
v
k
)
2
+
_
1

i
+
q
I

2( + p)

i
_

q
I
(
i
I
v
k

ki
I
)
2( + p)
_
2

i
I
v
k

ki
I

i
_

_
_

J
q
I
2( + p)
_

Ji
v
k

Jki

i
_

_
_
+
_

i
I
v
k

ki
I

i
_

_
_ _

IJ

Ji
v
k

Jki

IJ

i
_

_
_

i
I
v
k

ki
I

i
_

_
_
m
I

i
p
+
_

i
I
v
k

ki
I

i
_

_
_ _

i

IJ
2
u
+
E
j
J

ij
_
,
(V.125)
where j
i
S
is the non-relativistic canonical entropy current
dened as (similar to relativistic canonical entropy cur-
rent given by Eqn. (II.35)),
j
i
S
= sv
i

i
I
. (V.126)
8
There will be two terms proportional to

C
IJK
E

J
B
K

which
we have checked will vanish after reduction.
9
Although the form of this equation can be maintained by adding
some one derivative correction to temperature or chemical po-
tential.
12
TABLE I: Relations between relativistic parameters and
non-relativistic parameters (parity-even uid)
Non-rel. variables Rel. variables
a
Velocity v
i
u
i
u
+


(u
+
)
2
(E + P)
Y
i
Mass Density (u
+
)
2
(E + P) + (u
+
)
2
(Z )
Energy Density
1
2
(E P)
1
2
(Z )
Pressure p P

d
Z
u

P
E + P

Charge qI u
+
QI u
+
I
ij
u
+
ivj
Entropy s Su
+
(1 + )
Scalar Potential I A

I
Vector Potential a
i
I
A
i
I
a
Z is given by Eqn. (II.29), Y
i
is given by Eqn. (IV.84) and is
given by Eqn. (IV.124).
We can rewrite Eqn. (V.125) in more conventional form
as

+
s +
i
j
i
S
=
1

n
2

ij

ij
+
1

z(
k
v
k
)
2
+
1

i
_

i
+
I

i
_

I
(
i
I
v
k

ki
I
)
_
+
_

Ji
v
k

Jki

i
_

_
_

i
I
. (V.127)
The rst three terms on the right hand side survive for
uncharged dissipative uid [1]. Here we compute the
other terms which are responsible for local entropy pro-
duction for parity-odd charged uid in presence of back-
ground elds. Since we demand that the uid satises
the second law of thermodynamics, the right hand side
of Eqn (V.125) should be positive denite. This implies
following constraints on the transport coecients
m
I
= 0, n 0, z 0,
IJ
=
1

IJ
,
_

IJ


J
q
I
2( + p)
_
matrix is positive denite. (V.128)
These constraints are compatible with what we get
from relativistic entropy production condition. However
the last term in Eqn. (V.125) is a parity odd term and
can not be written as a perfect square. Therefore, it
seems that the presence of this term breaks the second
law of thermodynamics. However, similar to the relativis-
tic uid dynamics, we can add extra parity-odd terms
(allowed by symmetry) to entropy current to make right
hand side a perfect square.
Note that the last term in Eq. (V.125) appears only in
2+1 dimensions. Since the rst order parity-odd terms in
Eqn. (V.125) do not appear in the non-relativistic uid
theories in more than two spatial dimensions, we have a
TABLE II: Relations between relativistic and
non-relativistic transport coecients, energy
momentum tensors and currents (parity-even uid)
Non-rel. variables Rel. variables
Bulk Viscosity z u
+
Shear
Viscosity
n u
+
Electrical
Conductivity
IJ IJu
+
+ u
+
QIQJ/(E + P)
2
Thermal
Conductivity
2n( + p)/
I qI /2( + p)
Charge
Conductivity
I nqI /
Momentum
Current
t
ij
v
i
v
j
+ pg
ij
n
ij
zg
ij

k
v
k
Energy
Current
j
i
v
i

+ p +
1
2
v
2

n
ik
v
k
zv
i

k
v
k

I
i

+ I

i
I
vj
ji
I

Charge
Current
j
i
I
qI v
i
I
i
IJ
i

+ IJ

i
J
v
k

ki
J

Entropy
Current
j
i
S
sv
i
+
I

I
i
t + IJ
i

IJ

i
J
v
k

ki
J

complete and consistent description

10
for these theories.
We summarize the results for them in Tables (I), (II).
A. Parity-odd corrections to NR Entropy Current
For uids in two spatial dimensions however, we must
add some parity-odd terms to the entropy current with
some arbitrary transport coecients and determine them
from the condition that there is no local entropy loss dur-
ing the ow of the system. We make the most generic
parity-odd modication to the entropy current in two
spatial dimensions as follows
j
i
S
j
i
S
+
_
a
ij

j
+ b
I

ij

j
_

_
+ c
ij

j
p
+ d
I

ij

Ij
+ f
I

ij
v
k

Ikj
_
. (V.129)
a, b
I
, c, d
I
and f
I
are arbitrary transport coecients.
After adding these extra terms the right hand side of
10
This is the most generic description of a non-relativistic uid
obtained by light-cone reduction of a relativistic uid system
13
Eqn. (V.125) can be written as (parity-odd part only)
_

i
_
a
ij

j
+ b
I

ij

j
_

_
+ c
ij

j
p + d
I

ij

Ij
+f
I

ij
v
k

Ikj
_
_
+
_
1
T
E
i
I

i
_

_
_ _

i

IJ
2
u
+
E
j
J

ij
_
=
_

ij

i
a +
_

ij

j
_

__

i
b
I
+
_

ij

j
p
_

i
c
+
_

ij

Ij
_

i
d
I
+
_
1
2
v
k

ij

Iij
_

k
f
I

ij

i
_

__
_

I
(u
+
)
2

j
+
I
q
I

j
_

2
I

j
p
_
+
_

ij
1
T
E
Ii
__

I
(u
+
)
2

(
I
q
J
+
J
q
I
2
JI
)
j
_

2
I

j
p
_
+
_

ij
1
T
E
Ii
__

Q
J
E
Jj

IJ
2
u
+
E
Jj
_
+
1
2
_
f
I

ij

Iij

k
v
k
d
I

ij

Iij
_
= 0. (V.130)
To obey the second law of thermodynamics we claim that
the coecients of all the independent terms must vanish.
This is possible except for the last two terms, for which
we get the following sets of constraints:
a
p

c

= 0,
a
(
I
/)

b
I

=
+ p

2
I

,
b
I
p

c
(
I
/)
=
2
I

, (V.131)

f
I

=
d
I

=
+ p

2
2
I

,
f
I
p
=
d
I
p
=
2
I

f
I
(
J
/)
=
d
I
(
J
/)
=
1

(
I
q
J
+
J
q
I
2
JI
) ,
(V.132)
additionally, the following matrices must be symmetric:
2
IJ

I
q
J
,

t
b
J
(
I
/t)

I
q
J
. (V.133)
It should be noted that the new coecients introduced
in j
i
S
need not be most generic, and their only signicance
is to make the parity-odd terms in entropy current vanish.
Hence, a minimal choice is as follows:
a = c = 0, f
I
= d
I
, (V.134)
which solves the respective constraints, and leaves us only
two independent coecients in entropy current: b
I
and
d
I
.
Now we turn attention towards the last two terms of
Eqn. (V.130). They vanish if we consider the uid is
incompressible and placed in a time independent back-
ground. For compressible uid the second last term is
non-zero and can be negative. Hence the second law im-
plies that the f
I
= d
I
= 0. As a result all the parity-odd
terms vanish for compressible uid. The resultant uid
is completely parity-even and we summarize the results
in tables (I), (II). On the other hand if we consider the
background eld is time dependent, then the last term
can violate the second law. However, in that case, one
has to be careful about the change of entropy of the back-
ground also. The second law should hold for the uid and
the background system together. The last term produces
some vortex kind of motion in the uid. In that case the
velocity of uid particles around those vortexes becomes
very large and our analysis (derivative expansion) may
break down.
It should be noted that we demand our non-relativistic
system to obey the second law of thermodynamics. We
start with a relativistic uid but never impose any phys-
icality constraint (except the uid satises rst law of
thermodynamics) on the relativistic side. Relativistic
system, in our case, can be considered as a mathematical
model (where we allow all possible terms allowed by the
symmetry). After reduction we see that at least for the
parity-even sector constraint from the second law turns
out to be same (compatible) both in relativistic and non-
relativistic case (Eq. Eqn. (II.38) and (V.128)). But
for parity-odd sectors they are completely dierent. For
example, C
IJK
do not even appear in non-relativistic re-
lations Eqn. (V.131) - (V.133), though for relativistic
uid the second law relates
I
and

IJ
to C
IJK
(Eq.
Eqn. (II.41)). Whereas for non-relativistic uid one can
consider
I
and
IJ
to be independent and f
I
and d
i
are xed in terms of them. However, one can also start
with a relativistic uid satisfying second law of thermo-
dynamics (as an additional condition). In that case for
an incompressible uid it is possible to express the non-
relativistic transport coecients (f
I
, d
I
,
I
,
I
) in terms
of anomaly coecients of the relativistic system. But for
compressible uid it is not possible as these coecients
turn out to be zero
B. Incompressible uid in constant magnetic
background in (2+1)-dim
Equation of state of an incompressible uid is given by:
= constant, (V.135)
which implies that
k
v
k
= 0, or in other words, there is
no compression or expansion of the uid during the ow.
Using the non-relativistic constitutive equations one
can easily show that:
d
dt
+ ( + p)
k
v
k
= v
j

ij

i
, (V.136)
14
which means that for incompressible uids, d/dt is at-
most two derivative in order, and is zero for ideal uids.
Considering this result at rst derivative order we can
infer that:
u

(E P) = 0
u

P
E + P
= , (V.137)
in rst derivative order. Same can be inferred directly
from Eqn. (IV.88).
Maxwells equation in the magnetic limit (Appendix
A) are given by:

i
I
= 0, (V.138)

ij
I
= 0, (V.139)
2
ij

Ij
+
ij

Iij
= 0. (V.140)
Second equation implies that magnetic eld is constant in
space. Also, if magnetic eld
12
I
is time independent (as
we demand), electric eld is curl free. But rst equation
already tells us that electric eld should be divergence-
free. Hence electric eld is constant over the 2 dimen-
sional space. However, electric eld can still be time
dependent, as the corresponding term does not appear
in the magnetic limit of Maxwells equations. Finally in
our case we have, a constant electric eld (may depend
on time) and a constant magnetic eld.
For this system, the form of j
i
S
is given by
j
i
S
= sv
i

i
I
+b
I

ij

j
_

_
+d
I

ij
_

Ij
v
k

Ikj
_
.
(V.141)
where b
I
and d
I
are determined by equations (V.131) -
(V.133). Hence we will get the positivity of entropy in
the form of Eqn. (V.127), along with constraints Eqn.
(V.128). However
I
and
IJ
remain unconstrained by
entropy positivity. Results of this case have been sum-
marized in Tables (III).
VI. DISCUSSION
We started with a generic (d + 2)-dim non-conformal
relativistic uid (d 2) with anomalies specic to d = 2
in presence of background electromagnetic elds on at
space. Light cone reduction of this system gives a (d+1)-
dim non-relativistic uid. We demand that our non-
relativistic uid satises the second law of thermody-
namics which imposes certain constraints on the sys-
tem. For example we nd that the parity odd terms in
non-relativistic theory can only sustain for incompress-
ible uid in electromagnetic background with constant
magnetic eld in (2 + 1)-dim.
Our non-relativistic uid (obtained by LCR) has
generic dissipative terms, which are allowed by the sym-
metry and the condition that local entropy production is
always positive denite. LCR does not constraint the size
TABLE III: Relativistic and non-relativistic transport
coecients for parity-odd uid
Non-rel. variables Rel. variables
Parity-odd I I(u
+
)
2
coecients IJ

IJu
+
Mom. Curr. t
ij
v
i
v
j
+ pg
ij
n
ij
Energy
j
i
v
i

+ p +
1
2
v
2

n
ik
v
k
Current
i
I
i

+ I

i
I
vj
ji
I

j
i
I
qv
i
I
i
IJ
i

Charge + IJ

i
J
v
k

ki
J

Current +I
ij

n
j +
J
n
j

jp

+ IJ
ij

Jj v
k

Jkj

j
i
S
sv
i
+
I

I
i
+ IJ
i

Entropy IJ

i
J
v
k

ki
J

Current
I

I
ij

n
j
2

jp

bJ
II qJ

ij
j

dJ
I

IJ

ij

Jj v
k

Jkj

of these coecients. They all are xed in terms of the

transport coecients of the mother relativistic theory.
On the contrary, when one performs a 1/c expansion the
relativistic constitutive equations to get a non-relativistic
uid system, as done in [15], many of these terms are sup-
pressed depending on the physical considerations and the
type of system under view. We discuss the basic aspects
of 1/c expansion in Appendix (B).
Apart from the dissipative terms, our system and the
system obtained by 1/c expansion in [15] have certain
fundamental dierences. Firstly, their system is exten-
sive as the thermodynamic variables follow Eulers rela-
tion; however since LCR breaks the Eulers relation, our
system is no longer extensive. Secondly, in our system
is not necessarily proportional to q
I
, while in [15] it is
true at least at the leading 1/c order. This is why our
non-relativistic system has one more independent param-
eter as opposed to the 1/c case. We can however enforce
q
I
in our system as well, but it turns out that de-
manding so switches o all the dissipative terms from the
theory except for the bulk viscosity.
In [15] authors do not present an entropy current
calculation for non-relativistic uid obtained by 1/c
expansion. In fact, as we will review in Appendix (B),
the entropy positivity turns out to be trivial, and is
just followed from the leading order entropy current of
the relativistic theory. The constraints on the transport
coecients (which survives at the leading order) also
turns out to be the same. However in our case, we
have slightly dierent constraints, because of the above
15
mentioned fundamental dierences between the two
cases.
Acknowledgement
We would like to acknowledge Pratik Roy and Rahul
Soni for valuable discussion. SD would like to thank the
hospitality of IACS, Kolkata where part of this work has
been done. ND would like to thank the DST Ramanujan
Fellowship. Finally we are indebted to the people of In-
dia for their unconditional support towards research and
development in basic science.
Appendix A: Magnetic limit of electrodynamics
Maxwells Electrodynamics is a relativistic theory. In
fact it was a precursor to Einsteins Special Theory
of Relativity. Having a consistent relativistic descrip-
tion of electrodynamics, eradicated any need for a non-
relativistic theory of electrodynamics. However as in
current context, one needs a Galilean description of elec-
trodynamics, just to keep it consistent with other non-
relativistic theories.
Recently in [18], authors discussed the non-relativistic
limit of electrodynamics in two distinct ways. Depend-
ing on the strength of elds
11
, we can have two dis-
crete non-relativistic limits: |

E| c|

B| electric limit
and |

E| c|

B| magnetic limit. We nd that the light-

cone reduction of relativistic sourceless Maxwells equa-
tions, under certain identications, gives magnetic limit
of Maxwells equations, which will be our interest of
discussion here. A more thorough discussion of non-
relativistic electrodynamics can be found in [18].
In arbitrary dimensions, inhomogeneous Maxwells
equations are given by:

i
F
i0
=
o
c,
0
F
0i
+
j
F
ji
=
o
J
i
, (A.1)
while the homogeneous ones (Bianchi identities) are:

= 0, (A.2)
where

is the full-rank Levi-Cevita tensor. The

last identity follows directly from the gauge invariant
form of F

or the constituent elds

12
:
E
i
=
i

0
A
i
and F
ij
=
i
A
j

j
A
i
. As LCR
gives a non-relativistic theory with gauge invariance, we
are interested in a non-relativistic limit which preserves
11
In this section we use the conventional notation for electrody-
namics: E for electric eld, B for magnetic eld, for charge
density, J for charge current etc.
12
Instead of the conventional magnetic eld, we use its 2nd rank
dual in our work, as in arbitrary dimensions we do not have
electromagnetic duality and magnetic eld does not have any
xed rank, while its dual has.
Eqn. (A.2). We will see that the magnetic limit essen-
tially does the same. However there exists a consistent
electric limit as well, where Bianchi identities are broken,
but inhomogeneous Maxwells equations are preserved.
In terms of dimensionless elds and sources one can
write the inhomogeneous Maxwells equations as:

i
E
i
= ,

t
E
i
+
j
F
ji
=

J
i
, (A.3)
and the Bianchi identities as:

lmij

i
E
j
=
1
2

lmij

t
F
ij
, (A.4)

mkij

k
F
ij
= 0, (A.5)
where and are some dimensionless constants given
by:
=
[E]
o
[][L]
, =
[L]
[T]c
. (A.6)
From here we can easily read out, that in non-relativistic
limit 0. To preserve the Bianchi identities there-
fore, E
i
should be one order smaller than F
ij
; which is
why this limit is called magnetic limit. If we measure
smallness in terms of a parameter 1/c, we will have
E
i

n+1
and F
ij

n
. Therefore from Eqn. (A.3), rst
nontrivial order of
n+1
, and J
i

n1
, and inho-
mogeneous Maxwells Equations (in conventional units)
reduce to:

i
E
i
=

o
,
j
F
ji
=
o
J
i
. (A.7)
Essentially we have just dropped the displacement cur-
rent term from the Amperes Law. To explicitly see if this
limit is Galilean invariant, one can lookup [18]. These are
the very equations that have been used in Section (IVA)
and (VB).
One can now push the expansion of Eqn. (A.3) to
n+2
order and derive the continuity equation:

t
+
i
J
i
= 0. (A.8)
It should be remembered however that there are mixed
orders in this equation and to the highest order it just
states:
i
J
i
= 0, which also follows from Eqn. (A.7).
Continuity equation takes its usual form only when rst
two leading orders of J
i
vanish and J
i
, but in which
case one of the Maxwells equations modies to:
j
F
ji
=
0. This in conjugation with the Bianchi identity would
mean that the magnetic eld is constant over space.
Finally in conventional units this limit can be summa-
rized as: F
ij
E
i
c
n
, and to the maximum order
c
n

o
, J
i
c
n+2

o
.
16
Appendix B: 1/c expansion of relativistic uid
dynamics
In this section we discuss briey the 1/c expansion
limit of relativistic uid dynamics to get a non-relativistic
theory, using the prescription of [15]. We consider here
just a parity-even uid for comparison with our work.
Parity-odd sector in our case and in [15] are in dierent
hydrodynamic frames and thus are incomparable.
The constitutive equations of a relativistic uid (with
appropriate factors of c) are:

= cF

I
J
I
,

J
a
I
= 0, (B.1)
where
T

= (E + P)u

+ Pg

, (B.2)
J
a
I
= Q
I
u

I
. (B.3)
Respective dissipative terms are given as:

= 2

, (B.4)

I
=
1
c
2
T
IJ
P

_
M
J
T
_
+
1
c

IJ
E

J
. (B.5)
We can separate out rest contributions from E and M
I
:
E = Rc
2
+ E, M
I
=
1
K
m
I
c
2
+
I
. (B.6)
Here m
I
is the mass is to Ith charge ratio of constituent
particles in their local rest frame, which is assumed to be
constant. K is the total number of U(1) charges. Non-
relativistic mass and energy density are related to their
relativistic counterparts as:
= R, = E, (B.7)
where = (1 v
2
/c
2
)
1
. Since the uid under consid-
eration is single component, R/Q
I
= m
I
is a constant.
Non-relativistic charge density is dened as:
q
I
= J
0
I
. (B.8)
Pressure and Temperature are however kept same, which
to be consistent with the main text notation we will de-
note as: = T and p = P.
Before continuing with expansion we need to x the
order of various quantities. , , p, , v
i
can be thought of
nite order without much ambiguity. m
I
on the other
hand is quite sensitive to the kind of system under con-
sideration. Let us consider a charged uid made of ions
where m
I
c
2
; m
I
cannot be too low, or else the uid
would start coupling to the background elds. Corre-
spondingly the charge density q
I
would be of order c
2
.
Further, to keep the thermodynamics intact, one has
to assume
I
to be of order of c
2
. Finally, for exter-
nal elds to have nite eect (e.g. force) on the uid,

i
I

ij
I
c
2
.
Using this information we can reduce the energy-
momentum tensor to:
T
00
= c
2
+
1
2
v
2
+ +O(1/c),
T
i0
= v
i
c +
_
1
2
v
2
+ + p
_
v
i
c
+
1
c

ij
v
j
+O(1/c
2
),
T
ij
= t
ij
+O(1/c), t
ij
= v
i
v
j
+ pg
ij
+
ij
, (B.9)
where we have used:
n =

c
, z =

c
, (B.10)

ij
= n
_

i
v
j
+
j
v
i
g
ij
2
d

k
v
k
_
zg
ij

k
v
k
.
(B.11)
Energy-momentumconservation equations, at highest or-
der, will then reduce to:

t
+
i
(v
i
) = 0, (B.12)

t
(v
i
) +
i
t
ij
=
i
I
q
I
+
ij
I
j
Ij
. (B.13)
It will be worth to mention here the underlying assump-
tion: n, z 1 which is just an empirical fact. This is
precisely the reason why no dissipative corrections ap-
pear to the continuity equation. We can now expand the
charge current:
J
0
I
= q
I
, q
I
= Q
I
+O(1/c
6
),
J
i
I
=
1
c
j
i
I
+O(1/c
7
), j
i
I
= q
I
v
i
+
i
I
, (B.14)
where we have used:

IJ
=

IJ
c
,
IJ
=

IJ
c
, (B.15)

i
I
= c
2

IJ
m
J
K

i
. (B.16)
Here again we have used a physical input to x the order
of
IJ
. We would demand that in the non-relativistic the-
ory, charge to mass ratio should be a constant. Therefore
charge continuity and mass continuity equations should
be the same upto leading order in 1/c, and hence
i
I
should at maximum be of the order of c
4
. It further
implies that
IJ
c
8
. As a consequence, eects like
electric conductivity are suppressed in whole theory.
Finally we use:

_
T
0

1
K
m
I
c
2
J

I
_
= 0 to get the
energy conservation:

t
_
1
2
v
2
+
_
+
i
_
v
i
_
1
2
v
2
+ + P
_
+
i
_
= 0,
(B.17)
17

i
=
ij
v
j

i
, (B.18)
where we have identied the thermal conductivity:
=
IJ
c
4
K
2
m
a
I
m
a
J

. (B.19)
which is of nite order.
The non-relativistic theory has essentially turned out
to be identical to the chargeless uid discussed in [1],
because for uni-component uids, charge density serves
just as number density upto a factor. For multi-particle
uids however this might become more interesting, as
then one does not have to expect charge is to mass ratio
to be constant. In fact one can have some chargeless
particles also in the system.
1. 1/c Expansion of Thermodynamics
First law of thermodynamics and the Eulers relation
in the relativistic theory are given by:
dE = TdS + M
I
dQ
I
, (B.20)
E + P = TS + M
I
Q
I
. (B.21)
Under 1/c expansion they reduce to:
d = ds +
I
dq
I
, (B.22)
+ p = s +
I
q
I
, (B.23)
which are just the non-relativistic analogues of the
same equations. We therefore conclude that the non-
relativistic system gained by 1/c expansion follows ex-
tensivity.
Now lets have a look at the second law of thermody-
namics in the relativistic side, which mentions:

S

0. From Eqn. (II.36) we know the positive denite form
of entropy current:

S
=
1
T

+
1
T

2
+
_
E

I
c
2
T

1
c
P

_
M
I
T
__

IJ
_
E
J
c
2
T

1
c
P

_
M
J
T
__
.
(B.24)
In highest order this equation for non-relativistic systems
says that:

t
s +
i
j
i
S
=
1
2
n
ij

ij
+
1

z
_

k
v
k
_
+

2
_

_
2
.
(B.25)
But if we look at the respective coecients from the rel-
ativistic side:
n =

c
0, z =

c
0, (B.26)
=
IJ
c
3
K
2
m
a
I
m
a
J
T
2
0. (B.27)
We hence see that the relativistic entropy positivity im-
plies non-relativistic entropy positivity.
[1] L.D.Landau & E.M.Lifshitz (1997). Fluid mechanics,
Pergamon Press.
[2] N. Banerjee, J. Bhattacharya, S. Bhattacharyya,
S. Dutta, R. Loganayagam and P. Surowka, JHEP 1101,
094 (2011) [arXiv:0809.2596 [hep-th]].
[3] J. Erdmenger, M. Haack, M. Kaminski and A. Yarom,
JHEP 0901, 055 (2009) [arXiv:0809.2488 [hep-th]].
[4] D. T. Son and P. Surowka, Phys. Rev. Lett. 103, 191601
(2009) [arXiv:0906.5044 [hep-th]].
[5] R. Loganayagam, JHEP 0805, 087 (2008)
[arXiv:0801.3701 [hep-th]].
[6] S. Bhattacharyya, arXiv:1403.7639 [hep-th].
[7] S. Bhattacharyya, arXiv:1312.0220 [hep-th].
[8] S. Bhattacharyya, JHEP 1207, 104 (2012)
[arXiv:1201.4654 [hep-th]].
[9] K. Jensen, M. Kaminski, P. Kovtun, R. Meyer, A. Ritz
and A. Yarom, Phys. Rev. Lett. 109, 101601 (2012)
[arXiv:1203.3556 [hep-th]].
[10] K. Jensen, M. Kaminski, P. Kovtun, R. Meyer, A. Ritz
and A. Yarom, JHEP 1205, 102 (2012) [arXiv:1112.4498
[hep-th]].
[11] N. Banerjee, J. Bhattacharya, S. Bhattacharyya, S. Jain,
S. Minwalla and T. Sharma, JHEP 1209, 046 (2012)
[arXiv:1203.3544 [hep-th]].
[12] N. Banerjee, S. Dutta, S. Jain, R. Loganayagam and
T. Sharma, JHEP 1303, 048 (2013) [arXiv:1206.6499
[hep-th]].
[13] M. Rangamani, S. F. Ross, D. T. Son and E. G. Thomp-
son, JHEP 0901, 075 (2009) [arXiv:0811.2049 [hep-th]].
[14] D. K. Brattan, JHEP 1010, 015 (2010) [arXiv:1003.0797
[hep-th]].
[15] M. Kaminski and S. Moroz, Phys. Rev. B 89, 115418
(2014) [arXiv:1310.8305 [cond-mat.mes-hall]].
[16] A. Lucas and P. Sur

Aswka, arXiv:1403.5239 [cond-

mat.mes-hall].
[17] M. Rangamani, Class. Quant. Grav. 26, 224003 (2009)
18
[arXiv:0905.4352 [hep-th]]. [18] G. Manfredi, Eur. J. Phys. 34 (2013) 859-871
[arXiv:1303.5608 [physics.class-ph]].