Sie sind auf Seite 1von 22

Topological Defect Dark Matter

Maxim Pospelov

University of Victoria and Perimeter Institute



(with D Budker, S. Pustelny, D. Jackson-Kimball, M.
Ledbetter, Gawlik, others)

Work in progress with A. Derivyanko

PRL 2013 + some follow-ups


Patras workshop, Mainz


Plan
1. Introduction.



2. Main idea: dark matter can be composed of extended is space scalar
or vector field configurations. [E.g. made stable for topological
reasons] These objects can have elementary interactions with SM
particles and fields. One would need a network of detectors to see a
passing of one such objects through the Earth.

3. Interaction between extended objects and matter can lead to a


momentum transfer, change in the frequency of a transition, torque
on spin. The signals can be at the detectable level.

4. More detailed example of a domain wall interacting with spins.


5. Future networks of magnetometers, atomic clocks and gravitational


wave detectors for transient effects.

Big Questions in Physics

Does dark matter (and also dark energy) have non-gravitational


interactions?



The most costly hunt for dark matter (search for WIMPs) have not yet
produced a strong positive result. Can we search for other types of dark
matter using other techniques?



What is the space of theoretical possibilities for dark matter?

Simple classification of particle
DM models
At some early cosmological epoch of hot Universe, with temperature

T >> DM mass, the abundance of these particles relative to a species of
SM (e.g. photons) was

Normal: Sizable interaction rates ensure thermal equilibrium, NDM/N =1.
Stability of particles on the scale tUniverse is required. Freeze-out calculation gives the
required annihilation cross section for DM -> SM of order ~ 1 pbn, which points
towards weak scale. These are WIMPs.

Very small: Very tiny interaction rates (e.g. 10-10 couplings from WIMPs). Never in
thermal equilibrium. Populated by thermal leakage of SM fields with sub-Hubble rate
(freeze-in) or by decays of parent WIMPs. [Gravitinos, sterile neutrinos, and other
feeble creatures call them super-WIMPs]

Huge: Almost non-interacting light, m< eV, particles with huge occupation numbers
of lowest momentum states, e.g. NDM/N ~1010. Super-cool DM. Must be bosonic.
Axions, or other very light scalar fields call them super-cold DM.



But even these broad categories are not exhaustive.

Extended field configurations of

light fields


Take a simple scalar field, give it a self-potential e.g. V() = (2-v2)2.



If at x = - infinity, = -v and at x = +infinity, = +v, then a stable
domain wall will form in between, e.g. = v tanh(x m) with

m = 1/2 v



The characteristic span of this object, d ~ 1/m, and it is carrying
energy per area ~ v2/d ~ v2 m Network of such topological defects
(TD) can give contributions to dark matter/dark energy.



0D object a Monopole
Energy
1D object a String
profile
2D object a Domain wall
d ~ 1/m


Cosmological problems from stable QCD axion DW P. Sikivie

Rough comparison with WIMPs and
axions


WIMPs DM: EW scale mass. Compton wavelength, ~ 1/mWIMP,
deBroglie w.l. ~ 1/(velocity mWIMP) ~ 1/(10-3 mWIMP) ~ nuclear size.



WIMP particles are widely spaced compared to their inverse mass with
L ~ cm [within our galaxy] in between neighboring particles.



Axion DM: Light particles with huge number of particles per (w.l.)3
the whole space is filled. Sinusoidal in time waves at = ma~ e.g. 10-5
eV. Average r.m.s amplitude, a ~ 100 eV, or so << EW scale.



TD DM: A very shallow potential V() can lead to an amplitude

max=A ~ EW scale. A particle-like 0D object is distributed over 1/m
distance scales, and so the total mass is ~ A2/m >> EW scale.
Therefore, necessarily the average distance is ~ cm (A/m)1/3 - very6
large!

Comparison with WIMPs and axions
Axions small amplitude but no space between particles




WIMPs EW scale
lumps of energy (>>
axion amplitude), very
concentrated in space
And with significant
~ cm gaps between
particles

TD DM large amplitude but also large


(possibly macroscopic) spatial extent d. Large
compared to WIMPs individual mass, and then
large (possibly astronomical) distances between
DM objects.

TD DM is a possibility for DM that will have very different signatures in terrestrial


7
experiments.
Transient signals from TD DM
Regardless of precise nature of TD-SM particles interaction it is clear
that



1. Unlike the case of WIMPs or axions, most of the time with TD DM
there is no DM objects around and only occasionally they pass
through. Therefore the DM signal will [by construction] be transient
and its duration given by ~ size/velocity.

2. If the S/N is not large, then there can be a huge benefit from a
network of detectors, searching for a correlated in time signal.

3. There will be a plenty of the constraints on any model of such type


with SM-TD interaction, because of additional forces, energy loss
mechanisms etc that the additional light fields will provide.


8
es,
to Possible Interactions
measurements opening avenues for further investigations
involving more magnetometers.
Let usSummary.
his call by ,We 2, shown
1,have - real thatscalar fields
a network from TD sector that
of mod-
he ern magnetometers
participate in forming oers a realistic(More
a defect. chance often
for detecting
than not more than 1 field is
ct- the event of an axion-type domain-wall crossing and can
mi-involved).
probe partsLetofustherepresent
parameter SM spacefield
whereby anwalls
such electron,
can and a nucleon.

by contribute can
Interactions significantly
be organizedto dark matter/dark
as portals: energy.
coeff OdarkOSM.

od


en

ueA.
c 5 axionic portal
en fa


ta-
SM particles



cal
s, (s)
B.
c m scalar portal
av . M
eve
SM particles

kes


do- 21 + 22
C.

he 2
c
(2s)
m
quadratic scalar portal
M


ss- SM particles
ue


ess
D
1 22
ta- current current portal
g
nd M

SM particles

An atom
me- The inside
authorsaare
defect will
grateful to have addtlA.
N. Afshordi, contributions
Arvani- to its energy levels

9
in taki, A. Derevianko, J. Brown, S. Carroll, M. Kozlov, V.
es,
to Possible Interactions
measurements opening avenues for further investigations
involving more magnetometers.
Let usSummary.
his call by ,We 2, shown
1,have - real thatscalar
a network fields from TD sector that
of mod-
he ern magnetometers
participate in forming oers a realistic(More
a defect. chance often
for detecting
than not more than 1 field is
ct- the event of an axion-type domain-wall crossing and can
mi-involved).
probe parts Letofustherepresent
parameter SM spacefield
whereby anwalls
such electron,
can and a nucleon.

by contribute can
Interactions significantly
be organized to dark matter/dark
as portals: energy.
coeff OdarkOSM.

od


en

ueA.

c

5
axionic
portal
Torque on spin

en f a


ta-
SM particles



cal
s, (s)
B.



c m
scalar
portal
Shift of + extra gr. force

av . M
eve
SM particles

kes


do- 21 + 22
C.
he
2
c
m
(2s)
quadratic
scalar portal

Shift of + extra gr. force

M


ss- SM particles
ue


ess
D 1
22

ta- g


current

current

portal
extra gr. force

nd M
SM particles

me- The authors are grateful to N. Afshordi, A. Arvani- 10


in taki, A. Derevianko, J. Brown, S. Carroll, M. Kozlov, V.
The issue of technical naturalness
Any tree level potential

Vtree() = ctree0 + ctree1 + ctree22+.





Would have to have coefficients cti very small to keep evolution slow.
Loops generate larger corrections

Vloop() = cloop0 + cloop1 + cloop22+.





so that cloopi>>ctreei , One has to start with large and opposite tree-vs-loop
coefficients cloopi= - ctreei to ensure tight cancellation for several terms in
the series Very unnatural! Standard problem for scalar portals.
Importantly, same pessimistic argument does not apply to interactions
protected by shift symmetry, the axionic portal for example.
(* But may be the approach idea of having rigid technical naturalness
built in a model is not quite right, and we would miss out on
interesting physics *)
11
cosmological stretching dimension can easilyThey
3 exist. accountdo! See fore.g.
thepapers
growtb
transient
of L from O(100Again, m)LV tofor a and
QED, transient
fraction of ly. We conclude / tha
our fiducial choice, ma neV and L 102 ly, fits we
(3) 1 k%
with the
Typical cosmological
LV experiment LQED scenario
looks =forbof5wall

formation.
H
thatThe pseudoscalar coupling of the field athe
one can generalize as interaction os a spin i to 2
with with
fixedstandar
gradient
of the scalar
model field a,
f 1 a
fermions, i 5 i , leads to the interac


Dimension three coefficients can be induced fr
i
tion of spins of atomic via quantum constituents
loops withtoa the gradient
predictable of th
outcome


scalar
a-profilefield,


The Earth v
2U V

b (loop factor) .
H
Similarly, existing terrestrial = checks of / 2f 1
a
etc look sfor, a smooth M

Pl (7
int i i
d/dt signal, that is aItconstant in time. unless
is a disaster
i=e,n,p

either fine-tuning ha

divergence is absent, or the cutoff scale is m
where
And of fcourse
i areTD
free parameters
transient signal canof
bethe
viewed model with dimensio
as generalization of LV
ofandenergy.
changingFor Another
coupling
light example,
experiments
scalars in NCofQED,
oftointerest,
signals short
theduration.

astrophysica
bounds

apply and limit f > 109 GeV [13]. 2
Ln,p,e
eff = (two loop factor) U V me

Very large dimension 3 operator will be induce


12
Setting up a question
1. Take any portal [better still take technically natural ones]. Supply
constraints on fa, M* etc from the astrophysics, 5th force, etc -
anything that does not involve DM

2. Take the DM energy density, saturate it with TD DM (this is a big


assumption), and require that the average time between crossings T is
not much than ~1-10 yr.

3. Given the strength of some astrophysical constraints and restrictions


on energy density of the DM, do the current generation of high
precision instruments (atomic magnetometers, atomic clocks,
gravitational wave detectors) stand a chance in detecting transient
signal from DM?



If No probably such DM would not be detectable.

If Yes it is worth exploring opportunities for developing a network

13

fa

Proxies
5
c axionic portal and unknowns
SM particles

The only things we know are



DM ~
0.4 GeV/cm 3 - local energy density of Dark Matter

(s)
c m scalar portal
Mv ~SM10 -3 c - typical velocity of Milky Way halo objects

particles

Additional practicality input Tencounter < 1-10 yr



Unknowns (2s) : type of portals (I take A and D for now, as the most safe,

c m quadratic scalar portal
and
SM choose baryon current for the vector portal, g=1).

particles

fa > 109 GeV, M* > TeV (astrophysics, colliders etc)




2 (limit on M is incurrent
*
g fact quite a bit
current weaker)

portal
LSMparticles
average distance between defects. A amplitude of fields inside TD.
d ~ 1/m is the transverse size of the defects. One can show that


network = A2 d 0D, monopoles
3

L



L3~d2vT (for 0D objects)


A2



Equating network ~DM one can

= L2
1D, strings


e.g. express A via DM



A2
= 2D domain walls 14
Ld
How do you know if you ran through a wall?

MP, Pustelny, Ledbetter, Jackson-Kimball, Gawlik, Budker, PRL 2013





It was initiated in discussions with Budker, Pustelny, Ledbetter who
had two sensitive atomic magnetometers synchronized via GPS.
What is good for? Best magnetometers can surpass 1 fT/Hz! Are we
using these experimental capabilities to the fullest?

Domain walls of axion-like field moving with ~ 10-3 c, will create a
magnetic looking perturbation affecting atomic spins.

Crucially, if such a defect passes through the Earth, how would you
know? And will you notice?

You need a time-synchronized network of sensitive probes that can


detect the event in different locations. Domain walls will be an especially
suitable target.

15
sidering
m(4) the Lagrangian
ma , , of a complex
invariant underthe ZNwallscalar
-symmetry, field ,
0 in- exp(i2k/N )
ll use
sequences solution a S0 /N as ,anandexample, which determines
signal duration in thickness.
excess of a millisecond. T
n-wall solutions duces forthe
variant axion-like
eective
under Z Lagrangian
-symmetry,
where k is for
an the

integer. a field,
exp(i2k/N We choose ), where
the value,
potentialthe then
modelin such a
wi
Signal of axion-like domain wall
Gravitational N
and astrophysical constraints. From the
rticles.
omain-wall
ntials withIt someprofiledegree for
kmacroscopic
is an integer. N
of
dis- 3. For-
We choose be examined
the potential for self-consistency
inminima
such a wayquiring in the
is chosen thatcon
in da
point way
of that
view at has
itdistance N distinct much
scales larger
some de-
ctional
main-wall-type form that of this
solutions it profile
has inter-
N is
1 not 2minima
distinct mological 2 N scenario
a for the formation tribution
ture. The oftodethth
than
scalar d,
field, theLit wall
is
= easy can
( to
a) be characterized
find V that
sin the by
potential
, its massV ()per
(3) is This constr
ofConsider
type
ent discussion. a very
solu- energy-degenerate
dierent The light
important complex
a
2

pa- scalar 0
2field with
network 2Sfrom Zrandomly
N symmetry: distributed
problems
N/2 N a
0.410
in .
Nwit2 F
s, initial
area,
minimized
random
referred
distribution for2 the to as theLtension,
= | |
following V ();of V
values 0
SN/2 ()
and= a, 2N 42
2 main DE wall S0 ,
nt
d ma . energy- occur
in the2earlyNUniverse S 0 S N ,(1) Salso
when 0 /N importan
is streng
the Hub
rly Universe L =
leadsfield,to =|
the |
for- V
4
(); V
() = 0 ficiently bu
ndom
trophysical dis- scalar
with V
constraints. it 4S
is From
easy. This
the
to findreduction
rate
that the S 22N
will
drops 4happen
potential below
2 V ()dynam-
H is
m
This is
, malizing
when
at which
constraint the all
imne
t


works as the Universe
0
expands
0 Mass
potential where S 1
has
da 0
dimension
8S
2 0 m ofa energy
in
N and 1
a
is sider
dimensionles such c
erse
stance leads Sically
=
minimized
scales muchS0if; the alarger
for =the S= 0the
than 0; V=()
2
following
d,
Area Choosing dz
0is augmented
;
values

values dz 2
of =S L
for ;
and
N ... by
are
2 2
a, .the
typicallyaddi- (5)main
on , over(8)
wall
the are
the assum
cosm
network
order of
Theory
axions,
the Uni- admits
if stable,
tion such
whereof several
U S domainhas
(1)-symmetricdistinct
dimension vacua,
piece,ofN energy
V

= = 2
andN
(2
1/2
Sisexp(ia/S
2 2 N
dimensionless.
S ) , 0 ) to2
with parameterize
creted
ficiently ingbuild
and L the
inside mo
up
th
le t
erized by its mass per area,refered 0 the h horizon h size L 0 in (10 1)/H quad
in . S
, if
stable, S =hThe Choosing
. The spatial
network of=domain21/2
field1Sconfiguration
exp(ia/S
walls 2 0leads
will )have
to
a(r) N
parameterize
an 1 stretching
interpolat-
additional (2) theaccretion
sider of domain
constituents
such constrathe
S ; a = S 0; 2 ; pansion
2 ; ... 2 to the , of
iesmost
L
is not
with

Freezing0
ingdistance-scale
between the two0 Higgs modeNto
adjacent
parameter minima
L,L(z)its
an averageminimum,
N represents distance Sa= N S0 , pro-
domain-
between ing same
the way as ed
rela
2duces the eective Lagrangian for
=the Linazfield,
in /(1 + z). It is easy value,to H see
then t
wall solution. A network of intersecting domain walls
(2) is of arrives
domain at
walls the
int is
Reducing da towalls,the
8S 2 or a characteristic size
one
m aHiggsvariable, we have
leads of the ato domain.
Lagrangian

the This param-
formation of domain walls d
dz Freezing
possible
= eter 0the
is for N 3.
impossible
. mode The
to
(5) tosolution
its minimum,
calculate for
withoutadomain
S =
making S0wall
, pro- along ies isand
further quiring
not theenhance tha
const


dz duces 2
Neective 1 the 2 mechanisms troweak epoch, N a0wall Hin H(T ened 100GeV),
tribution
thenbya a5
to
the
xyassumptions
plane Lagrangian
that interpolates
about for the 2a afield,
between =of 2S0 /N value,
and formation stron
La = ( a) cosmological V0 sin ,
stretching (3) easily
can For account
light sca

neighboring
and evolution. vacuaWe 2with theitcenter
treat as a free of the
2S 0 wall atand
variable z =con- 0 quiring is
limit
that doma
saturated,
DE |f 0.4 |
in walls will takes have
strain the an
the additional
following 1
maximum form,2
of L
energy density of the
2
from
N a O(100 domain m) to a
walls, fraction
tribution
S ofto
/N10 ly.n,p,e
12
the
is We da
streG
that admits domain Lawall solutions
0 2 interact
= 4( a) V0 sin , (3) a
our fiducial choice, m a byneV and the
Lelsewh 10
5
er L, an average with DWVdistance
0 = /L 4S inbetween
20the
. This neighborhood
reduction 2S
ofwill0the Solar
happen Systemdynam- DE derived 0.410
is when G
[8]. In what
ic
size of aically domain.
the =
a(z) if 4S the
dark-matterThis
0 potentialparam-
4 arctan energy V ()
[exp(m
with
density,is
z)]
the
augmented
; DM
da cosmological

= 0.4 2S
by 0m
GeV/cm the scenario
3
a addi- ,.
S0 /N ofstrengthen
is
crossing
over
(9),
wall
and
form
event
the thenco
a


calculate with
tionVof
without = N4S0 . further
making This reductionpiece, VThe willdz pseudoscalar
happen N cosh(m dynam- 2 coupling
)2a,z)with is when ofcreted
thethefield
acceleration, networ
insid a
h =h (2 S01
0 U (1)-symmetric Spin signa
e
mechanisms ically
h if of.thewall Thepotential
formation
spatial SV0 ()
fieldisconfiguration
augmented
model fermions,
Lbya(r) the interpolat-
addi-
fneV
i (4)
1/2
aover
i tionalthe
of icosmolog
, constan
sensitive
accretion
5 leadsth a
= 0.4 TeV 2 2 .
creted inside the h
at it as a free tion
The ofDWU (1)-symmetric
characteristic
variable,two DM andadjacent piece,
thickness
con- minima V of=
htion of
the (2
wall
h spins
d S
is ) , with
determined
of0 atomic constituents this change
toway the d

ing between N represents
10 2 ly am domain-
a same
scribed in Rea
hby the accretion the dista
nergy density wall of .theThe mass
solution. domainspatial
ma walls,
A
offield
network
a configuration
(small)
of scalar excitation
intersecting
a(r) interpolat-
field, of a around
domain walls (6)same
is ceedingly
arrives
agnetic dit
at
atom
If on top ing
of
any between
that a-field
minimum, two adjacent
has
d the
2/m minima
axion-type
. The represents
mass couplings,
m aa domain-
can be there ex- will way
beOur a aschoice
distan
ghborhoodwall of the
This
possible Solar
solution.
constraint
for N ASystem implies
3. The
network
asome
byofsolution
intersecting
flexible
for adomain evolution
domain walls
of
wallisalong thearrives selves
and
at the or incoc
thefollo
pressed in terms of3spin
the original parameters of the po- is2f suggested
1 b
y magnetic-type
density, possiblexy
DM plane

tential,
force
domain-wall
0.4
for that
GeV/cm
N
mamass
on 3.
the
network
interpolates
= Ninside 1
The , and inside
solution
S0 (Vgalaxies.
the
1/2 for a domain
the
possibility
between wall,
a
1/2 = for 0

wall
H
them
andint
along
to=
2S build
0 /Nand magnetic
theened
i a by
constrain
16
s
mo
i ,
up their 0 /2) =We (2) consider N S0such . Owing the con- to
i=e,n,p the
of wall-crossin
spin-exch
is saturated
Such crossing time can easily be in This
K magnetometer described in Ref. [20]). Specializing (7)
for the case of two atomic species, Cs in the F = 4
excess
133
could beof the
far in excessCs mag-
of 10-nrad tipping angles that
can be experimentally detected [21]. Thus, both types of

Network of Magnetometers
netometer response time t ,theand we can oer combine the for a realistic
3
state and He in the F = 1/2 state, we calculate en- magnetometers ample opportunities
ergy dierence E between the F = F and rF = F z detection of the wall-crossing events.
z
max the wall,a signal factor SSo = max the1/2
Be states and t ofinto
in the middle

B
far we (t)
have used
e toconstraints (6),
galactic
. It is noteworthy that even if the energy

DW DM

For alkalibemagnetometers,
directly
H =
F a
int
Ff
; compared
1
f (Cs) =
e
e
f
1
the
7
9f
to
; f experimental
signal
(He) =
e f
1
p

sensitivity,
; isdensity
1
e
of walls in the galaxy does not exceed cosmolog-
ical dark-energy density, i.e. DW DE , the
n
expected
4S0 ma ma 10 GeV S0 /N 9 signal can reach 105 rad and S fT Hz, which
E =
N fe
1015 eV
9neV


f

0.4 TeV
,(8)
3 is still a realistic signal 1/2 with the best mag-
3for detection
0.4 pT 10 GeV S /N m 10
e
0 netometers. It
a is remarkable that a possible domain-wall

S
In these formulae we assumed that the nuclear spin is component of dark energy can, in principle, be detected
Hz f 0.4 TeV neV v /c
mostly due to unpaired neutron (3 He) or g7/2 valence
e in the laboratory.

proton (133 Cs), and one can readily observe complemen- Network of synchronized magnetometers. While a sin-
hysical angle of the helium wall crossing,
spin after the assum-
tary sensitivity to fi in two cases. We can express these
9 3 1/2 enough to detect a domain-
gle magnetometer is sensitive
0.4 pT 10 GeV L 10
results in terms of the equivalent magnetic field inside wall crossing, due to the rarity of such events it would

ing that the typical is below the dynamical
crossing time , (11)
the wall using Be F/F = aF/(F fe ) identification, be exceedingly dicult to confidently distinguish a signal
2 false positives induced by occasional abrupt changes
Hz f 10 ly v /c
where is the nuclear magnetic moment. The magnitude
eis given by
from

ry are, response time. Taking the spins to be oriented parallel
of Be (direction is impossible to predict)

of magnetometer-operation
field spikes, laser-light-mode
conditions, e.g., magnetic-
jumps, etc. A global net-
109 GeV 1011 T (Cs)
device to the
For nuclear wall,
where spin in we
max
Be thecalculate
ma
inequalitythis
magnetometers,
neV fe

S0 /N
0.4 TeV we angle
used
the tipping
108 T (He) to be
the
,(9)
work of synchronized optical magnetometers is an attrac-
gravitational
angle is
con-
tive tool to search for galactic/cosmological domain walls,
K, Cs, straint from Eq. (6). The maximally allowed value for the
and the larger equivalent field strength for 3 He originates
as it would allow for ecient vetoes of false domain-
wall crossing events. We also note that comagnetome-
from its smaller magnetic moment. The couplings and9 ter schemes involving 3 either a second spin species or
agnetic signal4S ( pT/ Hz), after taking
0 in Eq. (9) are normalized
wall parameters 3 to 10 into
the GeVaccount
maxi- 10 the S
SQUID /N
gravi-
magnetometers 0 additional suppres-
could yield
tational
= 510 rad
mum allowed values from Eq. (6). The duration of the .
sion of false-positive events arising from local field fluc-
nuclear v N fand astrophysical constraints,
signal
is given f
eby the ratio of wall thickness to the trans- e
verse component of the relative Earth-wall velocity,
exceeds
v /c capabili-
0.4TeV FIG. 1: Sche
tuations or changes in operating conditions. As schemat-
F] 3 He- ties of modern magnetometers that can deliver fT/ (12) Hz ings recorded
3
d 2 neV 3 10

(7)
ing Thissensitivity
It is easy could
to see
t
be
that
v
=[19].
far =For
1.3 ms the
m v in excess

one

would
a m
He-K
of . SERF magnetometer, the
(10)
v /c 10-nrad tipping angles that
need
a
"
to determine
the 5th event.
F>5 = stations.
4 moreSuch appropriate
crossing time can easily figure ofthe merit would be the tipping
can4beevents
experimentally
netometer would
response , anddetected
be in excess of
time t determinewe can combinethe
r
[21].
Cs mag-
the
Thus,
"
both
" types of
max max 1/2
the en- B
magnetometers and t into a signal factor S = B
e
oer ample (t) to
opportunitiesth,

e
for a realistic
geometry,
= F
and make predictions
be directly compared

detection of
to experimental
for the
sensitivity,
5
pT the wall-crossing m 10 events.
9
3 1/2
0.4 10 GeV S /N
6th etc
So far
0 a
S
f 0.4 TeV neV v /c !
we
Hz
have
0.4 pT 10 GeV
e
used
9
L
the
10
galactic
3
constraints
1/2 (6),


Hz

DW DM . It is noteworthy f

10 ly
e v /c
, (11)
2
that even if the energy

1
* Nobody has ever
density attempted
of walls in the this galaxy before

where in the inequality we used the gravitational con-
does not exceed cosmolog-
; straint from Eq.
(6). The maximally allowed value for the 17
fn ical dark-energy density, i.e. DW DE , the expected
signal ( pT/ Hz), after taking into account the gravi-
mber of sta-
Possible signatures with gravitational
2
1D,A
0 A2
al Be , and = strings
= 2D domain walls L2
etome- Ld
wave detectors.
magnetome-
ated in
s operated in A2
in Krak ow, = 2D domain walls
rak o w,
distance of
Ld
Considering the case of monopoles
nce
(Krak oof
w) is
1 2 interacting with an atom via a
2
(
n n + p
p)
baryonic portal of M (n nM
n [14], while
+ p p) type, I have an estimate for an
1 2
ow)device

ERF is
2

blewhile
sensitivi-
additional The acceleration created during the TD passing,

proved upon authors are grateful to N. Afshordi, A. Arvani-
fdevice
the signal,
taki, A. Derevianko, J. Brown, S. Carroll,
M. Kozlov, V. 2


ion between
nsitivi- Flambaum, M. Kamionkowski, v
and M.L 3
Hohensee fordis- v V T
termined us- cussions. Thisa work
DM
was supported in part by
DM


def.upon
[16]). We SP is the scholar of theM 2 mMinistry
Polish
the NSF.
d of ScienceMand 2m d
p p


experiments
signal, Higher Education within the Mobility Plus program.

etween
Taking a TeV Thefor authors
the scale are of grateful to N. Afshordi,
the coupling one arrivesA. to
Arvani-


ed

us- taki, A. Derevianko, J. Brown, S. Carroll, M. Kozlov, V.
3) [Erratum- and M. B. Voloshin, Phys. Rev. D 78, 115012 (2008);
]). We
d P. Sikivie, -4 Flambaum, 2 andM. Kamionkowski,
JCAP 0604, 007 and (2006); M. Hohensee for dis-
aA.~Ritz10 cussions.
ments
elov, m/sec
N. Kaloper
This
S. M. Carroll, (1m/d)

L. Sorbo,
work
Phys. Rev. Lett. was
81, 3067supported
(1998); A. Lue, in part by the NSF.


of
nals SP is the scholar of the Polish Ministry of Science and
edIfsig-
d ~ Larm for LIGO
Higher Education~ 3 km, within T ~the1 yr then
Plus program.
Mobility



Strain ~ 10-16 Hz-1/2

and the effective frequency ~ 1/tcrossing ~ 100 Hz



This is very realistic,
ratum- Phys. Lett. as Bsearches
120, 133for gravM.bursts
(1983); reached
Pospelov, ~ 10-20Hz
A. Ritz
18
-1/2



Sikivie,
ted in Krakow, M2
Possible signature with atomic clocks
on distance of
ers (Krakow) is 3 2
tion [14], while DM v L DM v V T
a 2 2
a SERF device M mp d M mp d
A. sensitivi-
arable Derevianko, MP (work in progress)



improved upon
of the signal, 2
Consider
ration betweenan operator M2 me ee quadratic
that renormalizes
scalar portal the mass of an
electron us-
determined once an atom is inside a TD. Because of the quadratic nature of
e Ref. [16]). We The authors are grateful to N. Afshordi, A. Arvani-
the coupling
le experiments M * can
taki, be quite low
A. Derevianko, and S.atCarroll,
J. Brown, a ~ TeV. (ThereV.is a huge issue
M. Kozlov,
e with naturalness
the signals of of light
Flambaum, , as always]and
M.Kamionkowski,
M. Hohensee for dis-
monstrated sig- cussions. This work was supported in part by the NSF.
of
false-positive SP is the scholar of the Polish Ministry of Science and
measure-
The The atomic frequencies
Higher will shift
Education within temporarily
the Mobility and in a different
Plus program.
way for e.g. clocks on optical and microwave transitions.

If the / is shifted very briefly, current searches of d/dt will not


catch it as they integrate over a long time.

Achieving sensitivity to / (1 sec crossing) ~ 10-14 seems possible,


which will translate to M* ~ 1012 GeV sensitivity.
19
Take home message:


Current technologies allow probing areas of the parameter space of TD
DM, that are currently not ruled by astrophysics, collider constraints, or
the energy density budget.



By creating a network of magnetometers, and using the existing
networks of atomic clocks and GW detectors in a slightly different
regime, one can make an interesting step forward in constraining/probing
TD DM.

20
Future direction


Working out a plausible theoretical framework that creates enough
topological defects around us would be a plus.



Generalization to other types of interaction. Going from spin to
frequency, means switching from magnetometers to atomic clocks.

Learn from LIGO + friends about strategies of detecting transients.


Use existing searches to constrain TD DM (may be not possible to do
well v=1 is always assumed in LIGO type searches)

Experimental developments: GNOME proposal (Global Network of


Magnetometers for studies of Exotic physics).

21
Conclusion
We do not know what DM is it is worth keeping our options open
and explore opportunities where significant progress can be made
Topological defect DM is a suggestion that may be some fraction of
the observable missing energy comes in form of the extended
objects monopoles, strings or domain walls. At these stage it is not
competitive to other theoretical ideas (WIMPs, super-WIMPs, super-cool DM like
axions), because we do not have a very good understanding of its abundance.

The signatures are unusual transient effects via TD interactions with


spin, mass, energy levels. Network of GW detectors, atomic clocks
and magnetometers can search for such transient effects.

Domain walls will have especially unmistakable signature, and we


showed that atomic magnetometers can have enough sensitivity for
detection.
22