Brane embeddings in sphere submanifolds

Nikos Karaiskos, Konstadinos Sfetsos and Efstratios Tsatis

Department of Engineering Sciences, University of Patras
26110 Patras, Greece
{nkaraiskos, sfetsos, etsatis}@upatras.gr
Synopsis
Wrapping a D(8-p)-brane on AdS
2
times a submanifold of S
8p
introduces point-like
defects in the context of AdS/CFT correspondence for a Dp-brane background. We
classify and work out the details in all possible cases with a single embedding angular
coordinate. Brane embeddings of the temperature and beta-deformed near horizon D3-
brane backgrounds are also examined. We comment on the relevance of our results to
holographic lattices and dimers.
a
r
X
i
v
:
1
1
0
6
.
1
2
0
0
v
1


[
h
e
p
-
t
h
]


6

J
u
n

2
0
1
1
1 Prolegomena
When branes wrap internal manifolds have the tendency to shrink but they can be
stabilized by turning on the worldvolume gauge eld with quantized ux. A pioneering
example of such ux stabilization of D-branes was presented in [1] for the wrapping
of a probe D2-brane inside an SU(2) group manifold and a connection with results
from an exact CFT approach was also made. There have been numerous works in the
literature considering brane embeddings in various backgrounds and dimensions [2]-[7]
(and references therein). In particular, the authors of [4] considered congurations where
a D(8-p)-background wraps an S
7p
inside an S
8p
in the background of a Dp-brane. The
stabilization occurs at quantized values of the equatorial angle of the bigger sphere. This
result, besides being aesthetically beautiful, is also relevant in a holographic approach to
condensed matter lattices and dimer systems [8].
Motivated by these works and the potential physical applications that the brane embed-
dings have, we realized that, for generic values of p there are many more possibilities
even when one considers the simplest case of one embedding coordinate. Instead of S
7p
,
one could also select other submanifolds of S
8p
that the D-brane could wrap. These are
essentially given by subgroups of SO(9 p), which is the isometry group of S
8p
. These
submanifolds are presented in Table 1 for p = 0, 1, . . . , 5. The coloring is introduced for
later convenience.
Table 1: Submanifolds of S
8p
p = 0 S
7
, S
3
S
4
, S
2
S
5
, S
1
S
6
p = 1 S
6
, S
3
S
3
, S
2
S
4
, S
1
S
5
, CP
3
p = 2 S
5
, S
2
S
3
, S
1
S
4
p = 3 S
4
, S
2
S
2
, S
1
S
3
, CP
2
p = 4 S
3
, S
1
S
2
p = 5 S
2
, S
1
S
1
In section 2 we minimize the action of the brane probe and calculate the semi-classical
energy for each one of the aforementioned congurations. In general the energy depends
on the ratio of the ux units n of the worldvolume gauge eld to the number of the
Dp-branes N, that we stack together to form the background. For a given value of p
these energies depend on the specic submanifold that is wrapped. We have checked by
1
using kappa symmetry that the embeddings in consideration are supersymmetric, hence
ensuring stability.
In section 3 we present brane embeddings in the deformed backgrounds of [9]. In this case
it turns out that the dependence of the deformation drops out completely in the probe
computation. Pertaining to the -deformation, which involves an S-duality, we formu-
lated the problem mathematically, but we were not able to nd minimal congurations
explicitly due to its complexity.
In section 4, we turn on the temperature and examine its eect on the stability of our
constant embeddings. We conclude, by considering a small uctuation analysis, that these
are perturbatively stable, even though there is no underlying supersymmetry. Finally, in
section 5 we present concluding remarks and comment on the possible use of our results
in the context of holographic lattices and dimers.
2 Brane embeddings in Ramond-Ramond backgrounds
The geometry created by a stack of N coincident Dp-branes in the near-horizon region
is described by the ten-dimensional metric [10]
ds
2
=

r
R

7p
2
(dt
2
+dx
2
||
) +

R
r

7p
2
(dr
2
+r
2
d
2
8p
) , (1)
where d
2
p
is generally the line element of a unit p-sphere and the parameter R is given
by
R
7p
= N g
s
2
5p

5p
2
(

)
7p
2
(
7p
2
) . (2)
The background is also supported by a dilaton, (r) and a non-zero RamondRamond
(RR) eld strength F
(8p)
given by
e
(r)
=

R
r

(7p)(p3)
4
,
F
(8p)
= (7 p)R
7p
Vol(S
(8p)
) = dC
(7p)
, (3)
where Vol(S
(8p)
) denotes the volume form of the unit p-sphere and C
(7p)
is the RR
potential. We split the (8 p) spherical coordinates as (,
1
, . . . ,
7p
) and let be an
embedding coordinate of the probe brane, along with x

.
We concentrate rst to the cases corresponding to the entries of the Table 1 that involve
solely spheres. For these cases the metric of the compact space will have the form
d
2
8p
= d
2
+ cos
2
d
2
q
+ sin
2
d
2
7pq
, q = 0, 1, . . . ,

7 p
2

. (4)
2
This parametrization of the metric corresponds to the split of the 9 p representation
of the symmetry group SO(9 p) under the subgroup SO(q + 1) SO(8 p q) as
(9 p) (q +1, 1) (1, 8 p q). The variable [0,

2
], unless q = 0 in which
case [0, ]. The restriction on the range that q takes independent values is due to the
symmetry under the interchange of cos and sin in the above metric. Consequently,
the RR potential will be written as
C
(7p)
= R
7p
f()
(7p)
, (5)
where

(7p)
= dVol(S
q
) dVol(S
7pq
) =

hd
1
d
2
. . . d
7p
. (6)
The corresponding volume is
V
M
=

M
d
7p

h =
4
9p
2

1+q
2

8pq
2
. (7)
For q = 0 we should divide this formula by two since the general expression for Vol(S
q
)
gives 2 for q = 0. The function f() is given for q = 0 by
f() =
7 p
8 p q
(sin )
8pq
2
F
1

1 q
2
, 4
p +q
2
, 5
p +q
2
, sin
2

(8)
and for q = 0 by
f() = 2
8p
7 p
8 p

sin

2

8p
2
F
1

p
2
3, 4
p
2
, 5
p
2
, sin
2
2

. (9)
The dierence has its origins in the dierent ranges of the angular variable for the two
cases that we mentioned above. Note also that this is not the most general form for the
RR potential, but it is the only consistent one for the particular embedding that we will
consider, as will become transparent below.
The D(8-p)-brane probe is described by the sum of a Dirac-Born-Infeld (DBI) and a
Wess-Zumino (WZ) term
S = T
8p

d
9p
e

det( g +F) +T
8p

C
(7p)
F , (10)
where g is the induced metric on the brane, F is an abelian gauge eld strength living
on the world-volume of the brane and
T
8p
= (2)
p8
(

)
p9
2
(g
s
)
1
, (11)
is the tension of the probe brane. The integration is performed over the world-volume
coordinates of the brane which are taken to be

= (t, r,
1
, . . . ,
7p
). In general,
3
the embedding coordinate may depend on any world-volume coordinate. Here, we shall
restrict to the case where depends only on the radial coordinate, which is also consistent
with the form of the RR potential (5). Since the WZ term acts as a source term for the
abelian gauge eld strength F, the latter one is constrained to be F = F
tr
dt dr. We
also set the spacelike worldvolume coordinates x
||
to constants which is consistent with
their equations of motion.
With the above conditions and assumptions the probe brane action assumes in general
the form
S =

M
d
7p

dtdrL(, F) , (12)
where the Lagrangian density is computed to be
L(, F) = T
8p
R
7p

()
7 p

1 F
2
tr
+r
2

2
f()F
tr

. (13)
By varying the Lagrangian density with respect to the worldvolume gauge potentials, A
t
and A
r
, one observes that the following quantity is constant
L
F
tr
= const . (14)
Then in turns out that the gauge eld assumes the form
F
tr
=
f

()

(7 p)
2
(f

())
2
+ (f

())
2
. (15)
In order to attribute physical meaning to this constant, we consider the coupling of our
system to fundamental F-strings [4]. This is achieved by replacing F with F B, where
B is the KalbRamond eld. Expanding the action, we pick out a term of the form

M
d
7p

dtdr
L
F
tr
B
tr
. (16)
We can interpret the coecient in front of B
tr
as a charge (n units of T
f
) that multi-
plies the Kalb-Ramond potential of the fundamental string. Therefore, the fundamental
strings feel a potential in this background, whose strength is proportional to their
number, n, and their tension T
f
= 1/(2

). Consequently, one writes

M
d
7p

L
F
tr
= nT
f
, n Z . (17)
In order to nd semiclassical minima of the embeddings, solving the equations of motion
arising from the Lagrangian density would suce. However, since we are also interested
in computing the energies of our congurations, we will obtain the minima through
4
the Hamiltonian procedure. By performing a Legendre transformation, which actually
removes the WZ part, the Hamiltonian of the system is given by
H =

M
d
7p

dtdr

L
F
tr
F
tr
L

. (18)
Using the explicit form of the Lagrangian (13) and the quantization condition (17) the
Hamiltonian becomes
H = NT
f

dtdr

1 +r
2

()
7 p

2
+

1
f()

2
, (19)
where we have dened
=
n
N
, =

7p
2

8pq
2

1+q
2
. (20)
Since the origin of the constant is V
M
, it turns out that, for reasons explained below (7),
for q = 0 we should divide the above formula by two. It is obvious from the expression
for H that it is consistent to look for constant congurations, since the r-dependence
drops out in this case. Setting

= 0 and requiring H/ = 0 gives the condition

f

() = (7 p)
2
(
1
f()) . (21)
As one can see in Table 2 in some cases, depending on the specic values for p and q, this
equation admits an exact solution (), but in others it can only be solved numerically.
In the rest of the paper, in order to avoid a proliferation of symbols, we will denote by
the solution of (21). The energy density is dened by
H =

dr E . (22)
For general values of p and q it is given by
E
p,q
= NT
f

(cos )
2q
(sin )
2(7pq)
+ (
1
f())
2
. (23)
For the case where a one-cycle is manifest, that is q = 1 the above formula as well as the
one for the minima have a much simpler form given by
sin =

7p
6p

1
5p
, E
p,1
=

2
+ (sin )
122p
2(sin )
7p
. (24)
Noting that f(0) = 0 and f(

2
) =
1
, we nd the limiting behaviours
E
p,q
= nT
f
+O(
2
), E
p,q
= (N n)T
f
+O(1 )
2
. (25)
5
The results of our computations regarding the minima and the corresponding energies
are summarized in the Table 2 below. In all cases, the angle ranges from 0 to /2. We
have also included two more cases, apart from the products of spheres, which arise for
odd values of p, by writing the S
8p
as a U(1) bundle over CP
7p
2
. We use the conventions
of [11] and [12] for the CP
2
and CP
3
, respectively.
1
The normalizations for the metrics
are such that R

=
16
p+1
g

(for the values p = 1 and p = 3 that are of interest to us).

Table 2: Minima and energies
Cycles Algebraic equations for minima E
p,q
in units of NT
f
S
3
S
4
90 sin + 25 sin 3 + 3 sin 5 = 112 E
0,3
p = 0 S
2
S
5
450 cos + 25 cos 3 27 cos 5 = 448(1 ) E
0,2
S
1
S
6
sin =

7
6

1/5

1
35
36

7
6

2/5
S
3
S
3
sin =
1/2
(1 )
p = 1 S
2
S
4
18 + 8 sin 2 + 5 sin 4 = 36 E
1,2
S
1
S
5
sin =

6
5

1/4

1
24
25

6
5

1/2
CP
3
as above as above
p = 2 S
2
S
3
cos =
2
1/3
+(55+

2750+25
2
)
2/3
2
2/3
(55+

2750+25
2
)
1/3
E
2,2
S
1
S
4
sin =

5
4

1/3

1
15
32

25
2

2

1/3
S
2
S
2
=

2

sin
p = 3 S
1
S
3
sin =

4
3

1/2

1
32
27

CP
2
as above as above
p = 4 S
1
S
2
sin =
3
2

1
27
16

2
p = 5 S
1
S
1
= 0 or =

2
singular solutions
We should clarify two subcases of the above table. First, the results for the CP
2
and
the S
1
S
3
submanifolds coincide. This happens because the wrapping in the rst case
involves the U(1) ber with group structure S
1
and a submanifold inside CP
2
, which
has a similar structure with S
3
. The same happens with the results for the CP
3
and
the S
1
S
5
submanifolds. Second, for p = 5 we have the solutions = 0 and =

2
corresponding to a collapsing of the D-brane at the poles of the 3-sphere and are thus
singular. We also note that the respective equations that give the minima and energies
for the submanifolds S
7p
S
8p
can be found in [4], so that they are not reproduced
1
Our embedding coordinate (r) in these cases is identied with the coordinates and in equations
(5) and (4.1) in the references [11] and [12], respectively.
6
here.
Having obtained the energies for the various values for p, we plot them together with
the submanifolds of [4] in the following ve gures. The colors (black, blue, purple and
red) correspond to the entries with the same colors in Table 1. The energies are plotted
as functions of the ratio , in units of NT
f
. Curves with the same value for p, but a
dierent one for q, might intersect. We will use the obvious notation (q q

, )
0.2 0.4 0.6 0.8 1.0
0.05
0.10
0.15
0.20
Figure 1: Submanifolds for p = 0. We have:
(12, 0.46), (13, 0.47) and (23, 0.48).
0.2 0.4 0.6 0.8 1.0
0.05
0.10
0.15
0.20
0.25
Figure 2: Submanifolds for p = 1. We have:
(12, 0.49), (13, 0.47) and (23, 0.51).
0.2 0.4 0.6 0.8 1.0
0.05
0.10
0.15
0.20
0.25
Figure 3: Submanifolds for p = 2. We have:
(12, 0.48).
0.2 0.4 0.6 0.8 1.0
0.05
0.10
0.15
0.20
0.25
0.30
Figure 4: Submanifolds for p = 3. We have:
(12, 0.50).
0.2 0.4 0.6 0.8 1.0
0.1
0.2
0.3
Figure 5: Submanifolds for p = 4.
As is seen from the gures above, for a given value of p the maximally symmetric sub-
manifolds corresponding to q = 0 have the lowest energy. When the submanifold in
7
consideration includes an S
1
or a CP space the ratio cannot exceed
6 p
7 p
found from
(24). At this maximum value, the corresponding value of the energy density is
1
7 p
.
2.1 Kappa symmetry
We next check weather our embeddings preserve some portion of supersymmetry, by
examining the kappa symmetry. We will briey present results for the p = 3 case, the
others follow in a similar manner. In order to have supersymmetric congurations of a
D-brane probe in a given background, the following condition [13]-[15]

= , (26)
should be satised. Here, is the Killing spinor of the background which for the maxi-
mally supersymmetric case of the AdS
5
S
5
background, is unconstrained. We shall also
put the two MajoranaWeyl spinors of type-IIB theory into a doublet of chiral spinors,
which transforms as a vector under SO(2)
=

. (27)
The kappa symmetry operator

for the case of type-IIB theory is dened by

=
1

det( g +F)

n=0
(1)
n
2
n
n!

m
1
n
1
mnnn
F
m
1
n
1
F
mnnn
(
3
)
p3
2
+n
(i
2
)
(0)
. (28)
Here g is the induced metric on the D-brane and
F = F P[B] , (29)
where F is the worldvolume gauge eld and P[B] is the pullback of the KalbRamond
B eld. The
m
s are the induced worldvolume gamma matrices dened as
m
= e
A
m

A
,
with m, A being curved and at indices, respectively, while
A
are the ten-dimensional
at Dirac matrices. Finally,
(0)
is dened by

(0)
=
1
(p + 1)!

a
1
1
p+1

a
1
a
p+1
(30)
and
i
are the Pauli matrices which act in the usual way on the doublet of chiral spinors.
We next check the kappa symmetry for each one of the embeddings we considered in the
previous section. We consider rst the kappa symmetry for the S
2
S
2
wrapping case.
8
Note that F here is just the worldvolume eld strength, since there is no KalbRamond
eld in our background. Hence, the innite sum in (28) has only two terms, since only
the F
tr
components of the eld strength have been turned on. We compute that

k
=
1

1 F
2
tr
+r
2

2
(
1
P
1
P
2
+iF
tr

2
P
2
) , (31)
where we have dened the commuting operators
P
1
=
01
+r

05
, P
2
=
6789
. (32)
Consistency requires that
2

= I, which is easily proven. Breaking into components one

arrives at the following consistent algebraic system
(P
1
P
2
+F
tr
P
2
)
2
=

1 F
2
tr
+r
2

1
,
(P
1
P
2
F
tr
P
2
)
1
=

1 F
2
tr
+r
2

2
. (33)
We conclude that half of the components of the Killing spinor are related to the other half
and hence the conguration we considered here retains 1/2 of the original supersymmetry.
The cases of the S
1
S
3
and CP
2
wrappings are similar to the one above and hence the
results will not be presented here. They are also found to be 1/2 BPS congurations and
similarly for other cases with p = 0, 1, 2, 4.
3 Brane embeddings in deformed backgrounds
In this section we consider -deformed solutions of type-IIB supergravity [9] and brane
embeddings inside these backgrounds. We begin with the -deformation of the AdS
5
S
5
background for which the AdS
5
part of the metric remains the same, while the metric of
the -deformed 5-sphere is written as
d
2
5,
=
3

i=1
(d
2
i
+G
2
i
d
2
i
) +GR
4

2
1

2
2

2
3
(

i
d
i
)
2
, (34)
where
G
1
= 1 +R
4

2
(
2
1

2
2
+
2
2

2
3
+
2
3

2
1
) (35)
and (
1
,
2
,
3
) (cos , sin cos , sin sin ). The NS sector of the background includes
a dilaton and a KalbRamond two-form, given by
e
2
= Ge
2
0
9
B
NS
= R
4
G(
2
1

2
2
d
1
d
2
+ cyclic) (36)
and the RR potential and eld strengths
C
2
= 4R
4
w
1
(d
1
+d
2
+d
3
) , with dw
1
=
1
2
cos sin
3
sin 2d d ,
C
4
= 4R
4
(w
4
+Gw
1
d
1
d
2
d
3
) , Vol(AdS
5
) = dw
4
,
F
5
= 4R
4
(Vol(AdS
5
) +G Vol(S
5
)) , Vol(S
5
) = dw
1
d
1
d
2
d
3
. (37)
We consider D5-brane embeddings in this deformed background. The brane will wrap
the four angles of the deformed sphere so that the world-volume coordinates will be
(t, r, ,
i
) and the embedding coordinates are taken as x
||
= const. and = (r). As
before, we also turn on an Abelian world-volume gauge eld strength F
tr
. The action
of the brane probe is given by a sum of a DBI and a WZ term. Some extra care is
needed since there are new terms arising from the induced KalbRamond eld and the
RR potentials. The action assumes the generic form
S = T
5

D5
e

P[g] +F +T
5

D5

p
C
p
e
F
, (38)
where F is given in (29). After perfoming the computations, one concludes that the
action for the D5-brane becomes
S =
T
5
R
4
2

dd
3
sin 2

dtdr(cos sin
3

1 F
2
tr
+r
2

2
F
tr
sin
4
) . (39)
The entire -dependence has dropped out completely due to non-trivial cancelations in
the DBI and WZ terms, separately. In fact, this action is exactly the same as that
computed for the p = 3 and q = 1 case in which the D5-brane wraps the S
1
S
3
submanifold of S
5
. Indeed, one may check that the above Lagrangian above falls in the
generic family (13) with f() = sin
4
, which is the correct function appearing in the RR
potential for the aforementioned case.
The -deformed background has N = 1 superconformal symmetry. Since for our embed-
ding the action is actually -independent, we expect that the probe brane breaks one half
of it as it was shown before for the maximally supersymmetric cases. To demonstrate
this explicitly one has to work out (26) for our background using the fact that the corre-
sponding Killing spinor is no longer unconstrained, as in the maximally supersymmetric
case, but instead it is subject to two projections that reduce supersymmetry.
10
3.1 Embeddings in the -deformed background
One may also consider a more general deformation of the background, by performing
an S-duality in the theory [9]. Apart from , the resulting background depends also on
which is an additional scaleless parameter. Searching for D5-brane embeddings, we
choose the embedding coordinates x
||
= const. and = (). As opposed to the previous
cases, here should depend on , since the latter enters in the computations in a non-
trivial way. Actually, as we shall explain later, this is related to the embedding that we
have chosen. The Hamiltonian of the system turns out to be
H = T
5
R
4

P
2
Q+ (P sin
2
f())
2
, (40)
with
P =
1
2H
sin
2
sin 2 , Q = Hsin
2
sin
4
+H
2
cos
2
,
f() =

2
sin 2 , H = 1 +
2
R
4
(
2
1

2
2
+
2
2

2
3
+
2
3

2
1
) (41)
and the
i
s were dened in the previous section. As before, the parameter does
not appear at all, but does. It should be obvious that an attempt to nd constant
minima, that is -independent, is inconsistent. Varying the Hamiltonian with respect
to gives a complicated nonlinear dierential equation that one has to solve in order to
nd congurations that minimize the energy. We were unable to nd solutions of this
dierential equation.
This increased level of complexity appears due to the particular embedding that we
considered. Had we chosen a similar embedding = () for the undeformed background,
that is the p = 3 case with manifest S
1
S
3
,
Figure 6: Constant embedding Figure 7: Embeddings = ()
it would have also resulted to a similarly complicated dierential equation. This happens
because the embedding that one chooses is supported by a corresponding patch on the
11
sphere, to describe the forms locally. The two dierent embeddings in that case are
depicted in Figures 6 and 7 above.
In the case of non-constant embeddings = () it is obviously possible to rotate
the north pole in a way to obtain the rst conguration. In practice, this is done by
performing an SO(6) transformation. In the -deformed case it is not obvious what the
corresponding transformation would be, given that the isometry group has been reduced.
The deformed sphere has the same Euler characteristic with the undeformed one, so that
their topology is the same. It makes sense, then, to assume that such a transformation
exists, although we were not able to nd it.
4 Turning on temperature
It is natural to extend the discussion to asymptotically AdS spacetimes, which is relevant
to a holographic approach to dimers in condensed matter systems as pursued in [8].
We will briey discuss the case of the near horizon black D3-brane in a way that the
submanifold S
1
S
3
of S
5
appears manifestly. The metric of the background is
ds
2
= f(r)dt
2
+
dr
2
f(r)
+
r
2
R
2
dx
2
||
+R
2
(d
2
+ cos
2
d
2
1
+ sin
2
d
2
3
) , f(r) =
r
4

4
R
2
r
2
(42)
and the RR potential changes also accordingly. We consider a D5-brane probe with
the same embedding coordinates as before, i.e. x
||
= const. and = (r). It is a
straightforward task to show that the minima of the particular conguration remain the
same with the zero temperature case, a statement actually true for every p. This wouldnt
be the case for more general r-dependent solutions.
Supersymmetry is broken in all of these cases, due to the non-vanishing temperature.
Then, one cannot use kappa symmetry arguments to ensure the stability of the congu-
rations. Nonetheless, we can consider small uctuations around the minima. Let
=

+, F
tr
=

F
tr
+ , (43)
where the bars denote the minima and =
t

r

r

t
. It should be stressed out
that this is consistent as long as one considers only the zero mode in an expansion in
harmonics of the S
5
. In order to nd the complete spectrum, one should also turn
on uctuations of the eld strength in every possible direction (see also [16] for a prime
example). However, here we are only interested in demonstrating perturbative stability in
12
for non-zero temperature and for that restricting to the zero-mode suces. The eective
Lagrangian for quadratic uctuations is computed to be
L =
1
2

hR
4
T
5
cos

sin
3

1

F
2
tr

R
2

f(r)
1
(
t
)
2
f(r)(
r
)
2

+

2
1

F
2
tr
+A
2
+B

. (44)
The minima and the gauge eld are given by
sin

=

4
3
,

F
tr
=
9 16

81 96
, (45)
where we have dened the constants
A = 4 +
36
27 32
, B =

81 96
3 4
2
. (46)
We obtain the equations of motion by varying and . After combining them and
concentrating on a Fourier mode of the form = e
it
(r), we get
d
dr

f(r)
d
dr

2
f(r)

C
2R
2

(r) = 0 , C 24 +
72
32 27
, (47)
dened for r . Transforming this into a Schr odinger equation for , after changing
appropriately to a new variable z =

r
dr

f
1
(r

), with z [0, ) as r (, ]. The

associated potential, that can be written explicitly only in terms or r, is
V =
C
2
f(r)
R
2
. (48)
Substituting the value for in C from (45) one sees that C is non-negative. Hence the
zero mode of the conguration is always positive. In fact, C vanishes for the critical
value = 3/4. In conclusion, the conguration that we considered is stable. Similar
arguments also hold for the other submanifolds and for the cases p = 0, 1, 2, 4 as well.
5 Concluding remarks
We classied and energetically compared all possible cases, with a single embedding
angular coordinate, in which a D(8-p) brane can wrap AdS
2
times a submanifold of S
8p
in a Dp-brane background, thus producing a pointlike defect. We worked out the details
in all dierent cases that arise, performing also comparisons between them. We examined
similar constructions in the presence of temperature and in beta-deformed backgrounds.
13
We demonstrated stability either by supersymmetry arguments or by a small uctuation
analysis around the minima.
It will be interested to investigate and search for running solutions of the embedding
coordinate, i.e. = (r). This involves the classical equation of motion for the Hamil-
tonian (19). This is a highly non-linear equation but it should be possible to analyze it
numerically. Of particular interest will be solutions connecting minima corresponding to
dierent values of n, especially when they correspond to the same energy.
Our results should be relevant to models of holographic lattices and dimers. In par-
ticular, the position of minima for the case of D5-branes wrapped on AdS
2
S
4
in a
D3-background plays an important role in determining the free energy of a square lattice
[8] made out of D5- and anti-D5-branes. In a condensed matter language, the dimeriza-
tion takes place at a lower than a critical value temperature where the congurations of
connected pairs of D5- and anti-D5-branes are favorable. We expect a richer structure
when we consider, besides S
4
, all submanifolds of S
5
in Table 1. We hope to address
these and related issues in a forthcoming publication.
Acknowledgements
We would like to thank K. Siampos and D. Zoakos for useful discussions on the sub-
ject. N.K. acknowledges nancial support provided by the Research Committee of the
University of Patras via a K. Karatheodori fellowship under contract number C. 915.
References
[1] C. Bachas, M. R. Douglas and C. Schweigert, Flux stabilization of D-branes, JHEP
0005 (2000) 048, arXiv:hep-th/0003037.
[2] C. Bachas and M. Petropoulos, Anti-de Sitter D-branes, JHEP 0102 (2001) 025,
hep-th/0012234.
[3] J. Pawelczyk and S.-J. Rey, Ramond-Ramond ux stabilization of D-branes,
Phys.Lett. B493 (2000) 395-401, hep-th/0007154.
[4] J.M. Camino, A. Paredes and A.V. Ramallo, Stable wrapped branes, JHEP 0105
(2001) 011, hep-th/0104082.
14
[5] A. Karch and L. Randall, Open and closed string interpretation of SUSY CFTs on
branes with boundaries, JHEP 0106 (2001) 063, hep-th/0105132.
[6] O. DeWolfe, D. Z. Freedman and H. Ooguri, Holography and defect conformal eld
theories, Phys. Rev. D66 (2002) 025009, hep-th/0111135.
[7] K. Skenderis and M. Taylor, Branes in AdS and p-p wave space-times, JHEP 0206
(2002) 025, hep-th/0204054.
[8] S. Kachru, A. Karch and S. Yaida, Holographic Lattices, Dimers, and Glasses, Phys.
Rev. D81 (2010) 026007, hep-th/0909.2639.
[9] O. Lunin and J.M. Maldacena, Deforming eld theories with U(1) U(1) global sym-
metry and their gravity duals, JHEP 0505 (2005) 033, hep-th/0502086.
[10] G.T. Horowitz and A. Strominger, Black strings and P-branes, Nucl. Phys. B360
(1991) 197-209.
[11] C.N. Pope, Eigenfunctions and spinstructures in CP
2
, Phys. Lett. B97 (1980) 417.
[12] B.E. W. Nilsson and C.N. Pope, Hopf Fibration Of Eleven-dimensional Supergravity,
Class. Quant. Grav. 1 (1984) 499.
[13] K. Becker, M. Becker and A. Strominger, Five-branes, membranes and nonpertur-
bative string theory, Nucl. Phys. B 456 (1995) 130, hep-th/9507158.
[14] E. Bergshoe, R. Kallosh, T. Ortin and G. Papadopoulos, Kappa symmetry, super-
symmetry and intersecting branes, Nucl. Phys. B 502 (1997) 149, hep-th/9705040.
[15] E. Bergshoe and P. K. Townsend, Solitons on the supermembrane, JHEP 9905
(1999) 021, hep-th/9904020.
[16] M. Kruczenski, D. Mateos, R.C. Myers and D.J. Winters, Meson spectroscopy in
AdS / CFT with avor, JHEP 0307 (2003) 049, hep-th/0304032.
15