Self-interacting asymmetric dark matter. Kallia Petraki Université Pierre et Marie Curie, LPTHE, Paris

Self-interacting asymmetric dark matter Kallia Petraki Université Pierre et Marie Curie, LPTHE, Paris LUPM 2015

Our universe Dark Energy (69.1 ± 0.6) % Ordinary Matter (4.9 ± 0.03) % Dark Matter (26 ± 0.2) % 2

Dark matter The state of the matter Overwhelming evidence: Galactic rotation curves, gravitational lensing, Bullet cluster, cosmic microwave background. All of evidence due to the gravitational effect of DM. Big unknown: the particle nature of DM 3

Dark matter How can we find dark matter? First, we have to guess the answer! Need a strategy. 4

Dark matter Strategies Common solution with other open particle-physics problems, e.g. Hierarchy problem supersymmetric WIMPs Strong CP problem axions Neutrino masses sterile neutrinos 5

Dark matter Strategies Common solution with other open particle-physics problems, e.g. Hierarchy problem supersymmetric WIMPs Strong CP problem axions Neutrino masses sterile neutrinos Focus on DM-related observations: DM density Asymmetric DM Patterns of gravitational clustering Self-interacting DM 6

Dark matter Strategies Common solution with other open particle-physics problems, e.g. Hierarchy problem supersymmetric WIMPs Strong CP problem axions Neutrino masses sterile neutrinos Focus on DM-related observations: Not mutually exclusive! DM density Asymmetric DM Patterns of gravitational clustering Self-interacting DM 7

Outline Asymmetric DM Self-interacting DM Self-interacting Asymmetric DM Case study: atomic dark matter 8

A cosmic coincidence Why ΩDM ~ ΩOM? Similarity of abundances hints towards related physics for OM and DM production. Unrelated mechanisms different parameters result expected to differ by orders of magnitude. 9

Ordinary matter Stable particles: p e γ ν p + make up most of ordinary matter in the universe. Only p +, no p present today: matter-antimatter asymmetry Asymmetry Ω OM conserved today b/c of global U(1) symmetry of the SM, baryon-number B O Ordinary particles Ordinary anti-particles annihilated in the early universe 10

A non-coincidence Atoms: 4.9 % Ordinary matter Photons: 0.0022 % Neutrinos: 0.0016 % Particle-antiparticle asymmetry Relativistic thermal relics 11

A cosmic coincidence Why ΩDM ~ ΩOM? Just a coincidence OR Dynamical explanation: DM production related to ordinary matter-antimatter asymmetry asymmetric DM 12

The asymmetric DM proposal [Review of asymmetric dark matter; KP, Volkas (2013) ] DM density due to an excess of dark particles over antiparticles. DM OM asymmetries related dynamically, by high-energy processes which occurred in the early universe. Dark and visible asymmetries conserved separately today. generated / shaped by same processes OM asymmetry Ω OM DM asymmetry Ω DM Ordinary particles Ordinary anti-particles Dark anti-particles Dark particles got annihilated 13

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Ingredients OM asymmetry Ω OM generated / shaped by same processes DM asymmetry Ω DM Ordinary particles Ordinary antiparticles annihilated in the early universe Dark antiparticles Dark particles Low-energy theory: Standard Model: Ordinary baryon number symmetry B O Dark sector: Dark baryon number B D [accidental global U(1) symmetry] Interaction which annihilates dark antiparticles. How strong? determines possibilities for DM couplings low-energy pheno. High-energy theory: B O violation B D violation if correlated related asymmetries ΔB O & ΔB D 14

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Relating ΔB O & ΔB D Consider B gen B O B D X B O + B D or B gen (B-L) O B D X (B-L) O + B D Need processes which violate X ΔΧ 0 preserve Bgen ΔB gen = 0 Δ(B-L) O = ΔB D = ΔΧ / 2 [e.g. Bell, KP, Shoemaker, Volkas (2011); KP, Trodden, Volkas (2011); von Harling, KP, Volkas (2012)] Side point: B gen remains always conserved could originate from a gauge symmetry, a generalization of the B-L symmetry of the SM, coupled to a dark sector Z' B-L with invisible decay width in colliders 16

Non-relativistic thermal relic DM Symmetric DM Asymmetric DM Ω DM 1 / (σv) ann particles + particles excess Ω DM antiparticles antiparticles annihilated σ ann v rel 4.4 x 10-26 cm 3 /s σ ann v rel > 4.4 x 10-26 cm 3 /s fixed value no upper limit For (σv) ann 4.4 x 10-26 cm 3 /s > 2 n(χ) n(χ) < 5% [Graesser, Shoemaker, Vecchi (2011)] 18

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Phase space of stable / long-lived relics To get Ω DM ~ 26% : Non-thermal relics e.g. sterile neutrinos, axions Symmetric thermal relic (WIMP) DM Asymmetric thermal relic DM 4.4 x 10-26 cm 3 / s increasing (σv) ann Asymmetric dark matter Encompasses most of the low-energy parameter space of thermal relic DM study models and low-energy pheno. Provides a suitable host for DM self-interacting via light species. 19

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] DM annihilation Need (σv) ann > 4.4 x 10-26 cm 3 / s. What interaction can do the job? χ χ SM SM Annihilation directly into SM particles highly constrained via colliders and direct detection (see bounds on symmetric WIMP DM) χ χ φ φ Annihilation into new light states: φ SM SM : metastable mediators decaying into SM φ stable light species, e.g. dark photon (possibly massive, with kinetic mixing to hypercharge), or a new light scalar. 20

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Structure CONNECTOR SECTOR particles with G SM, G D and possibly G common Interactions which break one linear combination of global symmetries: e.g. conserved B O B D ; broken B O + B D Δ(B O + B D ) = 2 ΔB O = 2 ΔB D STANDARD MODEL gauge group G SM = SU(3) c x SU(2) L x U(1) Y accidental global B O strong pp, nn annihilation Portal interactions B O & B D preserving DARK SECTOR gauge group G D accidental global B D efficient annihilation 22

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Structure CONNECTOR SECTOR particles with G SM, G D and possibly G common Interactions which break one linear combination of global symmetries: Most phenomenological e.g. conserved B O B D ; broken implications B O + B determined by low-energy D physics. Δ(B O + B D ) = 2 ΔB O = 2 ΔB D STANDARD MODEL gauge group G SM = SU(3) c x SU(2) L x U(1) Y accidental global B O strong pp, nn annihilation Portal interactions B O & B D preserving DARK SECTOR gauge group G D accidental global B D efficient annihilation 23

Asymmetric DM [Review of asymmetric dark matter; KP, Volkas (2013) ] Phenomenology: Zoo of possibilities Does asymmetric DM pheno have to be unconventional? No. Many regimes where it behaves as collisionless CDM. Could have weak-scale interactions with ordinary matter. Main difference in (sufficiently) high-energy physics. Scenario still motivated by cosmic coincidence. Is it interesting to consider regimes with unconventional pheno? Yes! Disagreement between collisionless CDM predictions and observations of galactic structure: May be telling us something non-trivial about DM. Potentially interesting signatures, not yet fully explored. 24

Collisionless ΛCDM and galactic structure Very successful in explaining large-scale structure. At galactic and subgalactic scales: simulations predict too rich structure. Various problems identified: cusps vs cores, missing satellites, too big to fail. Too much matter in central few kpc of typical galaxies. 25

Small-scale galactic structure: How to suppress it Baryonic physics Shift in the DM paradigm: Retain success of collisionless ΛCDM at large scales, suppress structure at small scales Warm DM, e.g. kev sterile neutrinos Self-interacting DM Continuum of possibilities: How warm or how self-interacting can / should DM be? 26

Self-interacting DM What is it good for? The energy & momentum exchange between DM particles: Heats up the low-entropy material suppresses overdensities [cusps vs cores] suppresses star-formation rate [missing satellites, too big to fail ] Isotropises DM halos constrained by observed ellipticity of large haloes. to affect dynamics of small halos 0.2 barn/gev < σ self-scattering / m DM < 2 barn/gev to retain ellipticity of large halos [Theory: Spergel, Steinhardt (2000). Simulations: Rocha et al. (2012); Peter et al. (2012); Zavala et al (2012)] 27

Self-interacting DM What interaction? σ self-scattering / m DM ~ barn / GeV large! DM coupled directly to a light or massless force mediator: Many different kinds of interactions possible. L g φ χ χ χ : dark matter φ : mediator m φ << m χ Ensures strong enough DM self-scattering If mediator sufficiently light: σ self-scattering decreases with velocity [e.g. Rutherford scattering: σ self-scattering 1/v 4 ] Significant effect on small haloes (small velocity dispersion) Negligible effect on large haloes (large velocity dispersion) 28

Why self-interacting and asymmetric? L ~ g φ χ χ χ : dark matter φ : force mediator m φ << m χ DM self-interaction χ + χ χ + χ DM annihilation χ + χ φ + φ Too strong annihilation in the early universe leaves too little DM... unless there is a particle-antiparticle asymmetry. Asymmetric DM scenario: An excellent framework for self-interacting DM 29

Self-interacting asymmetric dark matter How to go about studying it? Light force mediator long-range interaction Many studies of long-range DM self-interactions (in either the symmetric or asymmetric regime) employ a Yukawa potential V χχ (r) = ± α exp ( m φ r) / r However, typically reality is often more complex for asymmetric DM with (long-range) self-interactions. 30

Self-interacting asymmetric dark matter Complex early-universe dynamics Formation of stable DM bound states Multi-species DM, e.g. dark ions, dark atoms, dark nuclei. Implications for detection Variety of DM self-interactions. Bound-state formation determines decoupling of DM from dark radiation. Affect galactic structure. Variety of radiative DM processes in haloes [bound-state formation, excitations+de-excitations of bound states] Variety of DM-nucleon interactions [elastic, inelastic (excitation, break-up of bound states)] Delineate classes of models, calculate cosmology + phenomenology self-consistently 31

A minimal scenario of self-interacting asymmetric DM: atomic dark matter 32

Atomic DM Minimal assumptions (i) Relic density: Particle-antiparticle asymmetry (ii) DM couples to a gauged U(1) D [dark electromagnetism]: self-scattering, annihilation [specific models, e.g. KP, Trodden, Volkas 2011; von Harling, KP, Volkas 2012; Choquette, Cline 2015] 33

Atomic DM Minimal assumptions rich dynamics (i) Relic density: Particle-antiparticle asymmetry (ii) DM couples to a gauged U(1) D [dark electromagnetism]: self-scattering, annihilation [specific models, e.g. KP, Trodden, Volkas 2011; von Harling, KP, Volkas 2012; Choquette, Cline 2015] Gauge invariance mandates DM be multi-component: Massless dark photon: Dark electric charge of dark protons p D + compensated by dark electrons e D -. They can bind in dark Hydrogen atoms H D. Mildly broken U(1) D, light dark photon: Similar conclusion in most of the parameter space of interest. [KP, Pearce, Kusenko (2014)] 34

Atomic DM Model-building example G = G SM x U(1) B gen x U(1) D gauged B gen gauged D accidental global B D same as (B-L) O for SM particles Efficient annihilation DM self-scattering in halos p D -2 1 2 e D 0-1 0 accidental global (B-L) O & B D δl low = L SM + p D (id m p )p D + e D (id m e ) e D + (ε/2) F Y μν F D μν δl high (1/Λ 8 ) ( u c d s c u d c s) e D c p D Direct / Indirect detection preserves B gen = (B-L) O B X asymmetry generation: Δ (B-L) D O = ΔΒ D breaks X = (B-L) O + B [e.g. via Affleck-Dine mechanism in susy models; D von Harling, KP, Volkas (2012)] 35

Atomic DM Cosmology Dark asymmetry generation via interactions neutral under U(1) D p D and e D asymmetries T asym > m pd / 25 Freeze-out of annihilations p D p D γ D γ D & e D e D γ D γ D T FO m pd,e D / 30 Dark recombination, p D + e D H D + γ D T recomb binding energy = α D 2 μ D / 2 Residual ionisation fraction x ion n pd [ min n pd + n 1, m pd m ed ] 10 10 4 HD α D GeV 2 t [If dark photon massive] Dark phase transition T PT ~ m γd / (8πα D ) 1/2 [Kaplan, Krnjaic, Rehermann, Wells (2009); KP, Trodden, Volkas (2011); Cyr-Racine, Sigurdson (2012); KP, Pearce, Kusenko (2014)] 36

Atomic DM with a massive dark photon Asymmetric DM coupled to a dark photon is multicomponent (p D, e D ), and possibly atomic (H D ) in much of the parameter space where the dark photon is light enough to mediate sizeable (long-range) DM self-interactions. [KP, Pearce, Kusenko (2014)] Conversely: Discussing single-component SIADM coupled to a light dark photon means not having considered the cosmology properly. This leads to erroneous bounds on model parameters: The formation of atomic bound states screens the DM self-interaction. Force mediator need not be sufficiently massive to satisfy constraints. 37

DM with long-range interactions Interplay between cosmology and strength of the interactions: Early universe regulates the ways DM may manifest itself today. Bound-state formation is the way the early universe regulates long-range interactions. It cannot be ignored! Interplay for stable symmetric thermal relics: Density (left over from early universe): ρ j 1 / (σ ann v rel ) j Annihilation signals today: Γ ann ρ j 2 (σ ann v rel ) j 1 / (σ ann v rel ) j Signal strength decreases with increasing interaction strength! 38

Atomic DM Self-scattering in halos Multi-component DM with different inter- and intra-species interactions H D H D, H D p D, H D e D, p D p D, e D e D, p D e D Strong velocity dependence of scattering cross-sections σ ion ion v 4, screened at μ ion ion v < m γd σ HD H D ( α D μ D ) 2 [ b 0 +b 1( m H D v 2 4 μ D α D 2 ) +b 2( m 2 v HD 2 4 μ D α D )2 ] 1 (valid away from resonances; b 0, b 1, b 2 : fitting parameters, depend mildly on m p /m e ) [Cline, Liu, Moore, Xue (2013)] 39

Atomic DM Self-scattering in halos Dark fine-structure constant α D Binding energy Δ = 0.5 MeV Dark photon mass m γd = 1 ev unphysical parameter space: collisionless Non-monotonic behavior in α D, because of the formation of bound states ( no upper limit on α D, or lower limit on m γd ). Strong velocity dependence of scattering cross-sections allows for ellipticity constraints to be satisfied, while having a sizeable effect on small scales. Collisionless CDM limits: large m HD small number density large α D tightly bound atoms small α D small interaction Dark Hydrogen mass m HD [GeV] small m γd atom formation large m γd no atoms, ion-ion screening [KP, Pearce, Kusenko (2014)] 40

Atomic DM Indirect detection: δl = (ε/2) F Y F D Bound-state formation in galaxies today from ionized component p D + + e D H D + γ D γ D e + e (for m γ > 1.022 MeV) [Pearce, KP, Kusenko (2015)] Level transitions (dark Hydrogen excitations and de-excitations) H D + H D H D + H D *, HD * H D + γ D, γ D e + e 41

Atomic DM Indirect detection: δl = (ε/2) F Y F D Sommerfeld-enhanced process: Efficient in non-relativistic environment of haloes Bound-state formation in galaxies today from ionized component p D + + e D H D + γ D γ D e + e (for m γ > 1.022 MeV) [Pearce, KP, Kusenko (2015)] Level transitions (dark Hydrogen excitations and de-excitations) H D + H D H D + H D *, HD * H D + γ D, γ D e + e 42

Atomic DM Indirect detection: Dark-atom formation in halos s BSF x 2 ion (σ BSF v rel ) [GeV 4 ] 2 m H D Bound state formation : d Γ BSF = ( σ dv BSF v rel ) x ion 2 2 ρ DM 2 m H D Annihilation of symmetric DM : 2 d Γ ann ρ = ( σ dv ann v DM rel ) 2 m DM x ion = 1 x ion < 1 Interplay between early universe cosmology and strength of interaction min and max signal strength [Pearce, KP, Kusenko (2015)] 43

Atomic DM Indirect detection: Atomic DM vs Annihilating DM Atomic DM : δ E = binding energy m H D Annihilating DM : δ E = 2 m DM [Pearce, KP, Kusenko (2015)] 44

Atomic DM 511 kev line in the Milky Way from dark-atom formation p D + + ed H D + γ D, γ D e + + e, e + + e γγ fully ionized DM m γd = 2 MeV; contracted NFW profile. partially ionized DM Insufficient annihilation in early universe Overproduction of photon continuum [Pearce, KP, Kusenko (2015)] 45

Where is this going? 46

Developing the self-interacting asymmetric DM paradigm Directions and ideas Vector mediator ( Atomic DM) Indirect detection: Identification of all interesting radiative processes, detailed analysis, bounds. Direct detection: Involves many processes: Elastic scattering of p D, e D, H D on the target, and inelastic scattering of H D (excitation or break-up) Contact-type and long-range DM-nucleon interactions, depending on the incident DM species, screening scale (dark-photon mass / Bohr radius), recoil energy. 47

Developing the self-interacting asymmetric DM paradigm Directions and ideas Scalar mediator One fundamental species, but always attractive interaction Large bound states of different multiplicity. Dynamics of bound states; connection to solitonic physics Cosmology: Synthesis of bound-state in the early universe, kinetic decoupling of DM from dark radiation. Effect on structure formation Radiative processes yielding indirect detection signals Direct detection 48

Developing the self-interacting asymmetric DM paradigm Provide input for: Simulations of galaxy formation Direct detection experiments Indirect detection searches Dark-photon searches, collider searches (Higgs sector) 49

Conclusion DM pheno depends on the interplay between cosmology & microscopic interactions. Lots more to think about and to calculate! Observations suggest that dark-sector dynamics may be quite complex. 50

extra slides 51

Self-interacting DM Simulations [ Zavala, Vogelsberger, Walker (2012); Zavala, Vogelsberger (2012)] Self-interaction cross-section Ellipticity of haloes: velocity dispersion (main halo) Viable scenarios: Constant σ / m < 1 cm 2 /gr (pink, yellow lines) Velocity-dependent cross-sections (blue, green lines) collision-less CDM 53

Self-interacting DM Simulations [ Zavala, Vogelsberger, Walker (2012); Zavala, Vogelsberger (2012)] Self-interaction cross-section Density profiles: cusps vs cores (15 subhaloes with largest v max ) 54

Self-interacting DM Simulations [ Zavala, Vogelsberger, Walker (2012); Zavala, Vogelsberger (2012)] Circular velocity profiles: Too big to fail Self-interaction cross-section (15 subhaloes with largest v max ) Data: v circ within the half-light radii for 9 MW dsphs. 55 Bottom right: can be fitted by lower mass subhaloes.

Self-interacting DM Simulations [ Zavala, Vogelsberger, Walker (2012); Zavala, Vogelsberger (2012)] Self-interaction cross-section Conclusion Constant cross-section: some tension between ellipticity constraints and cross-section required to change subhalo kinematics. Nevertheless σ scatt / m DM ~ 1 barn / GeV could work (narrow range). Velocity-dep cross-section: satisfies ellipticity constraints and fits subhalo kinematics. 56