Dark Matter: Thermal Versus Non-thermal Bhaskar Dutta...

1

Bhaskar Dutta

Texas A&M University

Dark Matter: Thermal Versus Non-thermal

IACS, Kolkata, India

December 19, 2014

2

Thermal and Non-thermal Dark Matter (DM)

Moduli Decay and DM

Baryogenesis from the Decay of Moduli

Necessary Conditions for Successful Models

Example of a moduli model

Dark Radiation (DR) and DM correlation

Example of visible sector model

Non-Thermal Scenarios at the LHC and Direct Detection Expt.Conclusion

Outline

3

SM

Dark Energy68%

Atoms5%

Dark Matter27%

Important questions:

What is the origin of dark matter?

How does it explain the dark matter content?

Is there any correlation between baryon and dark matter abundance?

Questions

Consequences for:

Particle Physics Models

Thermal History of the Universe

Dark Matter: ThermalProduction of thermal non-relativistic DM:

ffDMDM

ffDMDM

Universe cools

ffDMDM

Boltzmann equation

][3 2,

2eqDMDMeqDM

DM nnvHndt

dn

6

/)(2

31

3

6

/)(2

31

3

)2(

)2(21

21

TEE

TEE

eq epdpd

vepdpd

vvolnostppfdgn /.

)2(),(

3

3

m/T

3* Tgn

snY

s

Y

Dark Matter: Thermal

Freeze-Out: Hubble expansion dominates over the interaction rate

5

v1~DM c

DMDMnm

20~ DM

fmT

Dark Matter content:

Assuming : 2

2

~v

mf

ac~O(10-2) with mc ~ O(100) GeVleads to the correct relic abundance

m/T

freeze out

scm 3

26103v Y becomes constant for T>Tf

Nc G

H

83 2

0

Y

Thermal Dark Matter

Dark Energy68%

Atoms5%

Dark Matter27%

v1~DM

20~ DM

fmT

Dark Matter content:

Weak scale physics :

2

2

~v

mf

ac~O(10-2) with mc ~ O(100) GeVleads to the correct relic abundance

freeze out

scm 3

26103v WIMP

7

Suitable DM Candidate: Weakly Interacting Massive Particle (WIMP)

Typical in Physics beyond the SM (LSP, LKP, …)

Most Common: Neutralino (SUSY Models)

Neutralino: Mixture of Wino, Higgsino and Bino

Dark Matter: Thermal

Larger annihilationcross-section

smaller annihilationcross-section

Larger/Smaller Annihilation Problem for thermal scenario

Status of Thermal DM

8

LargeCross-section is constrained

: smaller than the thermal valueoannv

vann

Thermal DM 27%

Gamma-rays constraints: Dwarf spheroidals, Galactic center

Geringer-Sameth, Koushiappas’11, Hooper, Kelso, Queiroz, Astropart.Phys. ‘13

Experimental constraints:

9

Latest result from Planck

http://public.planck.fr/images/resultats/2014-matiere-noire/plot_constraints_planck2014.jpg

10

<sv>= a + b v2

Thermal Relic Density:At freeze-out, <sv> =3x10-26cm3/s

High b/a lowers the cross-section at small v (Present Epoch)For S wave domination, cross-section remains same (constant)

Dark matter annihilation cross-section:

Dark Matter Pair Annihilation

a, b are constants

If S wave is suppressed then the cross-section is dominated by P wave b v2>>a

<sv> is much smaller today compared to the freeze-out time

Annihilation cross-section can be larger today compared to theFreeze-out time due to Sommerfeld enhancement

11

Natural SUSY and dark matter [Baer, Barger, Huang, Mickelson,

Mustafayev and Tata’12; Gogoladze, Nasir, Shafi’12, Hall, Pinner, Ruderman,’11;

Papucchi, Ruderman, Weiler’11],Higgs mass 125 GeV & Cosmological gravitino solution [Allahverdi, Dutta, Sinha’12]

Higgsino dark matterHiggsino dark matter has larger annihilation cross-sectionTypically > 3 x 10-26cm3/sec for sub-TeV mass

Thermal underproduction of sub-TeV Higgsino

LHC Constraints and Status of DMLHC constraints on first generation squark mass + Higgs mass:

Unnatural SUSY: Wino DM- Larger annihilation cross-section(for smaller wino mass)Arkani-Hamid, Gupta, Kapla, Weiner, Zorawsky’12

12

Recent Higgs search results from Atlas and CMS indicatethat mh ~126 GeV

LHC status…

in the tight MSSM window <135 GeV

(1st gen.) ~ ≥ 1.7 TeV

For heavy ≥ 1.3 TeV

produced from , ≥ 700 GeV

produced directly, ≥ 660 GeV (special case)

excluded between 110 and 280 GeV for a mass-less or for a mass difference >100 GeV, small DM is associated with small missing energy

masses between 100 and 600 GeV are excluded for mass-less for or for the mass difference >50 GeVdecaying into e/m

qm~gm~

,~qm gm~

1~tm

1~tm

1~t

1~t

g~

~/~e 01

~

01

~

1~

1

~

Can lead to over production of relic abundance

Can lead to over production of relic abundance

Non-standard thermal history at is generic in some explicit UV completions of the SM.

fTfTT

Acharya, Kumar, Bobkov, Kane, Shao’08Acharya, Kane, Watson, Kumar’09Allahverdi, Cicoli, Dutta, Sinha,’13

Status of Thermal DMThermal equilibrium above is an assumption.

DM content will be different in non-standard thermal histories (i.e., if there is entropy production at ). fTT

Barrow’82, Kamionkowski, Turner’90

DM will be a strong probe of the thermal history after it is discovered and a model is established.

13

14

Thermal, Non-thermal

: Large multicomponent/non-thermal; Small Non-thermal

LHC: Investigate colorless particles, establish the model

DM annihilation from galaxy, extragalactic sources

Annihilation into photons: Fermi, HAWC, H.E.S.S.

Annihilation into neutrinos: IceCube

Annihilation into electron-positrons: AMS

vann

Obtaining correct relic density for :

1) (thermal underproduction):

Multi-component DM (WIMP + non-WIMP)Example: mixed Higgsino/axion DM

Asymmetric DM (DM content can have large )

2) (thermal overproduction): DM from WIMP decay

Ex: Axino DM,

Gravitino DM

1326103 scmv fann

1326103 scmv fann

Baer, Box, Summy’09

Covi, Kim, Roszkowski ‘99

Feng, Rajaraman, Takayama’03

fannv Zurek’13

1326103 scmv fann

Beyond Thermal DM

16

Non-Thermal DM

The moduli decay width: • start oscillating when H < mt• dominate the Universe before decaying

and reheating it

2

3

2 pMmc

Moduli are heavy scalar fields that acquire mass after SUSY breaking and are gravitationally coupled to matter Inflation

Radiation domination

modulidomination

radiationdomination

.

.

.

Decay of moduli

Dark Matter from Moduli decay:

m

Mm

cTp

r

2/1

2/1~

TeV50MeV3

mTT BBNr

e.g., Moroi, Randall’99; Acharya, Kane, Watson’08,Randall; Kitano, Murayama, Ratz’08; Dutta, Leblond, Sinha’09; Allahverdi, Cicoli, Dutta, Sinha,’13

For Tr<Tf: Non-thermal dark matter

m

TY r

43

Abundance of decay products

17

Dark Matter from ModuliDM abundance

1. First term on the RHS is the “annihilation scenario”Requires:

Since Tr < Tf, we need wino/Higgsino DM

],.[min DMr

f

f

Th

f

obs

DMDM BrYTT

v

v

sn

sn

r

fth

fannfann TT

vv

th

fannfann vv

Gamma-rays constraints: Dwarf spheroidals, Galactic center MDM > 40 GeV, Tf < 30 TrTr > 70 MeV

2. Second term on the RHS is the “branching scenario”Can accommodate large and small annihilation cross-sections

Bino/Wino/Higgsino are all okYf is small to prevent the Br fDM from becoming too small (actually in realistic scenarios: Br ≥ 5 x10-3 => Tr< 70MeV)

(for mf~ 5 x106 GeV)

18

Dark Matter from Moduli Decays dilutes any previous relics

Thermal DM gets diluted if Tr < Tf ~ mDM/20 ~ O(10) GeVAxionic DM gets diluted if Tr < LQCD ~ 200 MeV

(fa~1014 GeV is allowed for Tr ≥ TBBN) [Fox,Pierce,Thomas’04] Baryon asymmetry gets diluted if produced before f decay

Non-thermal DM Production from f decay

Annihilation scenario for Tr close to TfDM production with large cross-section: Wino/Higgsino

Branching scenario for smaller Tr(annihilation cross-section does not matter)

Baryon asymmetry from f decay ⇒ Cladogenesis of DM and Baryogenesis [Allaverdi, Dutta, Sinha’11]

Dark Matter from Moduli“Branching scenario” solves the coincidence problem

Baryon abundance in this model: BB BrYs

n

Yf appears in the DM abundance as well, Yf~ 10-7- 10-9

BRfB ~ 10-1-10-3 easy to satisfy for baryogenesis,e (one loop factor) ~ 10-1-10-2

For mDM ~ 5 mB, e BRfB ~ BRfDM nB ~ nDM

5111

DM

B

DMDM

B

DMDM

b

BrBr

mBrYBrY

m

W=r/rc; r=mn

The DM abundance and Baryon asymmetry are mostly saturated by Yf, Brs contribute the remaining not much particle physics uncertainty 19

20

Baryogenesis from Moduli

NNM

XXMXddXuNW cj

ciij

ciiextra 2

'

: SM singlet; : Color triplet, hypercharge 4/3N XX , Baryogenesis from decays of or NXX ,

: can be the DM candidate : Allahverdi, Dutta, Mohapatra, Sinha’12N~

Nbb

B BrYs

nn

BrN : branching ratio of moduli decay to N

Assuming that N produces Baryon Asymmetry

e : asymmetry factor in the N decay

<0.1, BrN ~ 10-2 -1 B = 9x10-11,1010~ 79 Y

21

Baryogenesis

From X decay

Typically, e1,2 is O(10-2) for CP violating phase O(1) and l~O(1)

Baryogenesis from Moduli

NN

MXXMXddXuNW c

jciij

ciiextra 2

'

Similarly, e2

22

Conditions for ModelsTwo typical problems for moduli decay

Gravitino Problem:[Endo,Hamaguchi,Takahashi’06][Nakamura,Yamaguchi’06]

If m3/2<40 TeV Gravitino decays after the BBNmf>2 m3/2 can lead to DM overproduction

Large branching ratio of moduli into light Axions Neff

[Cicoli,Conlon,Quevedo’12][Higaki,Takahashi’12]

(at 95% CL) : Neff=3.52+0.48-0.45

)114

871(

3/4

effrad N

Current bound from Planck+WMAP9+ACT+SPT+BAO+HST

Large volume can be obtained after stabilization of

For large volume, one can have a sequestered scenario such that:

For example, TeV scale SUSY can be obtained for:

Balasubramanian, Berglund, Conlon, Quevedo’05

saTfluxbb AeWWTTK ,)ln(3

2/)( bbb TT

Cicoli, Conlon, Quevedo’08

)~( 22/32/3 bb

mmmmmm softsoft

TeVmGeVmGeVm softb1~,105~,10~ 610

2/3

Ex: NT DM in Large Volume Scenarios

No Gravitino Problem

3/2vis

3/2np

3/2b v

[Detailed Mass spectra, Aparicio, Cicoli, Kippendorf, Maharana, Quevedo’14

23

The decay to gauge bosons arises at one-loop level:

The decay to Higgs controlled by the Giudice-Masiero term:

The decay to gauginos (and Higgsinos) is mass suppressed:

02/3 Br

2

32

4~

P

SMgg M

m

)ln(

23

b

2

2

~~

P

softgg M

mm

1Br

2

32

24 PHH M

mZdu

2/3mmb

LVS set up can successfully accommodate non-thermal DM.

NT DM in Large Volume Scenarios

Yf can be quite small ~ 10-10 : branching and annhilations are ok Allahverdi, Cicoli, Dutta, Sinha’13

24

25

NT DM in Large Volume Scenarios

The 3 body decay width larger than the 2-body decay width of moduli into gauginos[ is suppressed by (mgaugino/mf)2 compared to ]

Yf ~ 10-10 (using mf~5 x 106 GeV)

YDM: Yf BRfDM ~ 10-12 mDM~ O(100) GeV is allowed

Solves the coincidence problem

If the dominant decay mode is to gauge boson final statesf decays into DM particle via 3-body: ggg ~~

310 DMBR

gg~~ gg

Since produces dark matter at the end of the decay chaing~

26

DM-DR Correlation in LVS:The axionic partner of , denoted by is not eaten up by anomalous U(1)’s.

acquires an exponentially suppressed mass . is produced from decay:

Bulk axions are ultra-relativistic and behave as DR.

contribute to the effective number of neutrinos :

2

3

481

Paa M

mbb

b

Cicoli, Conlon, Quevedo’13

0bam

ba

baba

effN

2

3

48)(

P

totvis

Mmcc

vis

hideff c

cN7

43 )04.3( effeff NN

DM-DR Correlation

DM-DR Correlation

2

32

24 PHH M

mZdu

31 ZN eff

2

32

4~

P

SMgg M

m

bb aagg

48.045.048.0

effN

..chHZHKbb

du

Bound from Planck+WMAP9+ACT+SPT+BAO+HST at 95%

Decay to visible sector mainly produces gauge bosons and Higgs:

bbdu aaHHtot

We get a lower bound on Tr27

Pr

totvisr M

mm

TgCCT

4/1

* )(28851

75.22875.10)()(

*

gTeVOTMeVO r

DM-DR Correlation

2

3

2

3

24,

24 P

visvis

P

tottot M

mCMmC

22 21,2 ZCZC totvis

Using 4/14/1*

2 )]/(30[ visr gT

28

29

Br

mT

sn

Y r

43

66 10)(105~,3 YGeVOTGeVmZ r

Avoiding excess of DR within LVS prefers “Annihilation” scenario Higgsino-type DM.

obssn

sn

Br

3103

Abundance of DM particles produced from decay:

DM-DR Correlation

: Branching scenario does not work

: Branching scenario

DM-DR Correlation

fannfr v

scmTT

1326103

fann v

20~ m

T f

Obtaining the correct relic density in “Annihilation” scenario needs:

Assuming S-wave annihilation, which is valid for the Higgsino-type DM, is directly constrained by Fermi.

For Higgsino-type DM, using the b final state, the bound reads:

mGeV

GeVT r

1)18(

GeVm 40

Upper bound on Tr gets bounded from belowfann v 30

DM-DR Correlation

fannfr v

scmTT

1326103

20~ m

T f

Over production in the branching scenario

Allowed by the Fermi data in the annihilation scenario

GeVm 6105~

The Fermi bound is translated to constraint in plane: mNeff

Allahverdi., Cicoli, Dutta, Sinha’14

DM-DR Correlation

Pr

hidvisr M

mm

TgCCT

4/1

* )(28851

vis

hideff C

CN743

using

mNeff , : set lower bound on DM mass

6105m

33

Model Examples: 2

2

21 SmXhSXW ss

f can be a visible sector field S (moduli is a hidden sector field)

S Decay + DM Annihilation or no annihilation works

ms~ O(1) TeV, mX~ O(10) TeV, mN~ O(0.5) TeV

W=Ws+WN,X

BRfDM ~ 10-6

Ys =(3/4)Tr/ms ~ 10-4

nDM/s ~ 10-10

Allahverdi, Dutta, Sinha, ‘13

34

Model Example 2S Decay + Branching ratio Baryogenesis

~ 0.01

~

for (mN1/mX)

Ys =(3/4)Tr/ms ~ 10-4

~ 10-10

22

1

81

X

N

mm ~ 10-4

~ 10-2

35

Cross Sections via VBFCross Sections via VBFNon-thermal Scenarios @LHCjjpp 0

10

1~~

CDM

Probe the DM sector directly: One interesting way:

Preselection:missing ET > 50 GeV, 2 leading jets (j1,j2) :pT (j1),pT(j2) >30 GeV|Dh(j1, j2)| > 4.2 and hj1hj2 < 0. Optimization: Tagged jets : pT > 50 GeV, Mj1j2 > 1500 GeV; Events with loosely identified leptons(l = e; m ; th) and b-quark jets: rejected. Missing ET : optimized for different value of the LSP mass.Delannoy, Dutta, Gurrola, Kamon, Sinha et al’13

36

Non-Thermal scenario @ LHC

Final states at the LHCNew Particles: Heavy colored states:

LHC signals: new colored states -spin 0- are pair produced high ET four jets in the final statesnew colored states –spin ½- are pair produced, high ET four jets +missing energy

[via cascade decays into squarks etc]

Distinguishing Feature: 4 high ET jets and 4 high ET jets + missing energy

XX ,SM Singlet: N

NNM

XXMXddXuNW cj

ciij

ciiextra 2

'

37

Non-Thermal scenario @ LHC

Dutta, Gao, Kamon’14

Monojet Dijet Dijet pair

If N is the DM candidate, i.e., mN~ mp

Dijet + Missing Energy

38

DM via Monojet at LHC

Also, Mono-top, di-tops inthis model

Combining various observable, we can probe ann. cross-section

Monojet

Direct Detection

Direct detection scattering cross-section: cii XuN~

Scatters off a quark via s-channel exchange of XN~

For li~1, Mx~ 1 TeV,

Dark Matter Candidates: , NN ,~

~

Suppose: is the DM particle (spin-0) N~

If N (spin ½) is DM, MN~mp (to prevent N decay and p-decay), sN-p is 10-51cm2 (SI) and 10-42cm2 (SD)

• The origin of DM content is a big puzzleWe will be able to understand the history of the early universe

• Thermal DM is a very attractive scenario However, it contains certain assumptions about thermal history

• Alternatives with a non-standard thermal history are motivated Typically arise in UV completionsCan ease the tension with tightening experimental limits

• Non-thermal DM arising from moduli decay is a viable scenario can yield the correct density for large & small annihilation rates Successful realization in explicit constructions is nontrivial

Conclusion