LISA as a Probe of the Higgs Sector and TeV-Scale Particle Physics Germano Nardini (Univ. Bern) HPNP 2017 3/3/2017 Based on arXiv:1512.06357 [with S. Huber, T. Konstandin, I. Rues] + arXiv:1512.06239 [with 10 members of the LISA CosWG] arXiv:1605.08663 [with M.Chala, I.Sobolev] arXiv:1702.00786 [proposal submitted to ESA]

Gravitational Waves & Astrophysics Gravitational Waves (GWs) exist The Hulse-Taylor binary system provided the first INDIRECT evidence...

Gravitational Waves & Astrophysics Gravitational Waves (GWs) exist … while the first DIRECT evidence came from the GW150914 signal

Gravitational Waves & Astrophysics Gravitational Waves (GWs) exist … while the first DIRECT evidence came from the GW150914 signal And we are even able to determine the details of the sources! Gravitational Waves & Astrophysics

Gravitational Waves & Astrophysics Present technique is based on antenna arrays and ground-based interferometers covering the GW frequency ranges 10-9–10-6 and 100–104 Hz What kind of (astro)physics are we missing at the 10-4–100 Hz frequencies?

Gravitational Waves & Astrophysics Present technique is based on antenna arrays and ground-based interferometers covering the GW frequency ranges 10-9–10-6 and 100–104 Hz What kind of (astro)physics are we missing at the 10-4–100 Hz frequencies? Typical objects: Binaries of white dwarfs and 104–108 black holes Merging galaxies (coalesence of massive BHs) Compact binary systems Extreme-mass-ratio inspirals

Gravitational Waves & Astrophysics Present technique is based on antenna arrays and ground-based interferometers covering the GW frequency ranges 10-9–10-6 and 100–104 Hz What (astro)physics are we missing at the 10-4–100 Hz frequencies? O(103) resolv. galac. binaries O(107) unresolv. galac. binaries BH binaries of 104–108 EMRI Merging BHs of 104–108

Gravitational Waves & Astrophysics Present technique is based on antenna arrays and ground-based interferometers covering the GW frequency ranges 10-9–10-6 and 100–104 Hz What (astro)physics are we missing at the 10-4–100 Hz frequencies? O(103) resolv. galac. binaries O(107) unresolv. galac. binaries BH binaries of 104–108 EMRI Merging BHs of 104–108 Very solid reasons to have a detector here !

Gravitational Waves & Astrophysics Present technique is based on antenna arrays and ground-based interferometers covering the GW frequency ranges 10-9–10-6 and 100–104 Hz What (astro)physics are we missing at the 10-4–100 Hz frequencies? O(103) resolv. galac. binaries O(107) unresolv. galac. binaries BH binaries of 104–108 EMRI Merging BHs of 104–108 Very solid reasons to have a detector here !

Gravitational Waves & Cosmology Besides astrophysical sources, we also expect pre-BBN cosmological sources (inflationary epoch, topological defects, phase tansitions, ... ) These generate a stochastic (non localized) GW background

Gravitational Waves & Electroweak Phase Transition Depending on the Higgs potential and interactions, the ELECTROWEAK phase transition (EWPT) can be of first order (and can even solve BAU) Such an EWPT naturally produces a GW signal with frequencies LISA is sentive to Moreover, first order phase transitions due to TeV-scale physics (e.g. in a hidden sector) can also produce a GW signal that can be detected at LISA

The eLISA interferometer (2011; after NASA quit)

The eLISA interferometer (2015; many open issues) Extra budget: partially because of reanalyses of costs, partially because of NASA+Japan, … ; Yes / No Yes / No 1 / 2 / 5 2 / 5 years of data taking

The eLISA interferometer (2016; open issues converging) Extra budget: partially because of reanalyses of costs, partially because of NASA+Japan, … ; striking results from LISA PathFinder; after the GW150914 discovery it became a field with guaranteed “unexplored” physics Yes !!! X X 2 / 5 2 / 5 years of data taking

The LISA interferometer (2017; proposal for ESA) arXiv:1702.00786 First ESA’s reply: ~Summer 2017 X X 4 years of life-time but consumables for 10 2.5 Proposed launch: ~2028

Gravitational Waves from 1st-Order EWPT Let us assume that the EWPT is of first order, i.e. The phase transition occurs via tunneling. In the place where the tunneling happens, a bubble of EW broken phase ( ) nucleates. Conventionally, the EWPT starts in the Universe when statistically we have 1 nucleated bubble per Hubble volume. The temperature of the Universe at this time is called The tunneling rate is . If is large (small), many (a few) bubbles have nucleated by the time the first bubbles have expanded, i.e. the phase transition ends with many little (a few large) bubbles. V V

Gravitational Waves from 1st-Order EWPT Let us assume that the EWPT is of first order, i.e. When bubbles collide, they convert part of their kinetic energy (of the expanding wall + turbulent fluid) into GWs! So, the more energy is available (→supercooling), the stronger the GW signal This available energy is related to which we normalize to the radiation energy: V M. Kamionkowski et al., '94

Gravitational Waves from 1st-Order EWPT Let us assume that the EWPT is of first order, i.e. When bubbles collide, they convert part of their kinetic energy (of the expanding wall + turbulent fluid) into gravitational waves (GWs)! So, the more energy is available (→supercooling), the stronger the GW signal This available energy is related to which we normalize to the radiation energy: V ~normalized difference btw. the minima ~how fast the minimum goes down M. Kamionkowski et al., '94

Gravitational Wave Signal Simulations on bubble collisions (based on the “envelope approx”) show where (for ) On the top of the “envelope” result, there might be important corrections: Magnetic HD Sound Waves S. Huber, T. Konstandin, '08 P.Binetruy,A.Bohe,C.Caprini,J.Dufaux,'12 C.Caprini,R.Durrer,G.Servant, '09 M.Hindmarh,S.Huber,K.Rummukainen,D.Weir,'13,'15

Gravitational Wave Signal Approximate the spectrum as a linear combination of the sources and consider the (normalized) energy absorbed in the phase change if , all available energy goes into the plasma (non-runaway) if , there is extra energy to accelerate indefinitely the wall (runaway) if (plasma is negligible), only the bubble wall effect is relevant (vacuum) Bodeker, Moore '08; Will change in B.&M. ‘17. J. Espinosa, T. Konstandin, J.M. No, G. Servant '08; D. Weir, ‘16.

Gravitational Wave Signal Approximate the spectrum as a linear combination of the sources and consider the (normalized) energy absorbed in the phase change if , all available energy goes into the plasma (non-runaway) if , there is extra energy to accelerate indefinitely the wall (runaway) if (plasma is negligible), only the bubble wall effect is relevant (vacuum) Bodeker, Moore '08; Will change in B.&M. ‘17. J. Espinosa, T. Konstandin, J.M. No, G. Servant '08; D. Weir, ‘16.

Gravitational Wave Signal Approximate the spectrum as a linear combination of the sources and consider the (normalized) energy absorbed in the phase change if , all available energy goes into the plasma (non-runaway) if , there is extra energy to accelerate indefinitely the wall (runaway) if (plasma is negligible), only the bubble wall effect is relevant (vacuum) Bodeker, Moore '08; Will change in B.&M. ‘17. J. Espinosa, T. Konstandin, J.M. No, G. Servant '08; D. Weir, ‘16.

Detection of the GW Signal Subset of LISA cosmology working group, 1512.06239 The LISA “sensitivity curve” for a stochastic background is much better than what one could naively think. It is not the one for monochromatic sources, e.g. stable BH binaries The dominant enhancement is that the signal has a continuous (broken power-law) frequency spectrum over which you can integrate The LISA stochastic-background sensitivity curve will be public soonish DETAIL: the signal has to be enough above the noise to have a discovery (SNR > SNRthresh) Thrane,Romano, '13 LISA sens. curve will stay between these 2 configs. X X

Detection of the GW Signal Subset of LISA cosmology working group, 1512.06239 The procedure can be absorbed into a “detection map”: you determine in your model, and then… you just compare them with the maps we provide The signal parametrization does not depend on whether the phase transition is in the EW / hidden / high-scale sector. Non-runaway Non-runaway

Detection of the GW Signal Subset of LISA cosmology working group, 1512.06239 The procedure can be absorbed into a “detection map”: you determine in your model, and then… you just compare them with the maps we provide The signal parametrization does not depend on whether the phase transition is in the EW / hidden / high-scale sector. Non-runaway Non-runaway For the maps of other cases, see arXiv:1512.06239

Is this region of a and b/H feasible in well-motivated UV theories? Question These bubble simulations show that a sizable GW spectrum from EWPTs is possible for some values of a and b/H taken as inputs A wide region of a and b/H , which correspond to a supercooled EWPTs, can be probed at LISA But a and b/H are features of the potential that strongly depend on the particle physics model !!! Is this region of a and b/H feasible in well-motivated UV theories? Remark: some low-energy extensions of the SM are also able to reproduce such values (see next talks)

Answer In the SM the EWPT is not of first-order If we detect GWs from a first-order EWPT, then there must exist BSM physics Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98. 75 GeV

Answer In the SM the EWPT is not of first-order To change this feature we need to modify the EW sector by means of either finite-temperature radiative corrections or/and new Higgs fields. In practice both options imply new scalar fields below the ~TeV scale Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98.

Answer In the SM the EWPT is not of first-order To change this feature we need to modify the EW sector by means of either finite-temperature radiative corrections or/and new Higgs fields. In practice both options imply new scalar fields below the ~TeV scale (Reminder) An exception to this criterion exists: RS models. If the Higgs starts “existing” only at low temperature, its phase transition is strongly distorted although the Higgs potential is SM-like (in fact the EWPT can be linked to the transition of the radion). No deviation in the quartic Higgs coupling!! Motivated UV completion? RS models YES Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98. Creminelli,Nicolis,Rattazzi, '02; Randall,Servant, '07; GN,Quiros,Wulzer, '07; Konstandin,GN,Quiros, '10, Konstandin, Servant, '11.

Answer bis (not linked to the radion) In the SM the EWPT is not of first-order To change this feature we need to modify the EW sector by means of either finite-temperature radiative corrections or/and new Higgs fields. In practice both options imply new scalar fields below the ~TeV scale Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98.

Answer bis (not linked to the radion) In the SM the EWPT is not of first-order To change this feature we need to modify the EW sector by means of either finite-temperature radiative corrections or/and new Higgs fields. In practice both options imply new scalar fields below the ~TeV scale SUSY models naturally satisfy both features. Let us start with the minimal supersymmetric extension (MSSM) ... Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98.

Answer bis (not linked to the radion) In the SM the EWPT is not of first-order To change this feature we need to modify the EW sector by means of either finite-temperature radiative corrections or/and new Higgs fields. In practice both options imply new scalar fields below the ~TeV scale (therefore testable) SUSY models naturally satisfy both features. Let us start with the minimal supersymmetric extension (MSSM) ... Kajantie,Laine,Rummukainen,Shaposhnikov, '96; Karsh,Neuhaus,Patkos '96; Csikor,Fodor,Hietger '98. VERY BAD Several uncertainties, but after the LHC data the GWs produced by the MSSM EWPT should be tiny (if any) Carena,Quiros,Wagner, '96; Delepine,Gerard,Gonzalez Felipe, Weyers, '96; Cline,Kainulainen '96. Cohen,Morrissey,Pierce '12; Curtin,Jaiswal,Meade '12; Carena, GN, Quiros,Wagner, '09, '13; Laine,GN,Rummukainen, '13; S. Liebler, S. Profumo, M. Stefaniak ‘16; Katz,Perelstein,Ramsey-Musolf,Winslow, ‘16.

Beyond the minimal SUSY option (Huber,Konstandin,GN,Rues '15) General NMSSM (MSSM + singlet, no additional discrete simmetry) R.Apreda, M.Maggiore, A.Nicolis, A.Riotto, '02; M.Jiang,L.Bian,W. Huang,J.Shu, '15; Kozaczuk,Profumo,Haskin,Wainwright, '15.

Two-step phase transition (extra singlet particularly good) Graphics from: N.Blinov,J.Kozaczuk, D.Morrissey, C.Tamarit,'15 At high T (but not too high), there is only the minimum along the singlet direction Nearby the critical T and below, there are minima in the singlet and Higgs orthogonal directions. They are separated by a barrier. The tunnelling involves more than one field: 2-step phase transition

Spectrum The light degrees of freedom: Higgsino-like chargino and neutralinos SM-like Higgs singlet-like CP-even scalar Higgsino-like states are compressed, so decays are too soft for the LHC Scalars are OK with LHC, LEP, … in the region of tiny mixing ( ) The rest of the particles (squarks, sleptons, gauginos, singlino, …) are assumed at the TeV scale to easily overcome the LHC bounds The sizeable logs kept under control by means of an ad-hoc subtraction scheme Giardino,Kannike, Masina,Raidal,Strumia, '14; Falkowsi,Riva,Urbano, '13.

Beyond the minimal SUSY model, very strong EWPTs are feasible!!! Good news !!! In principle, in supersymmetric extensions of the SM, the EWPT can produce a sizable GW spectrum Within supersymmetry, the most promising scenarios seem those where the EWPT occurs in 2 steps Likely, the presented results are quite generic and cover other SUSY models (if they do not depart too much from minimality) (but see also M.Garcia-Pepin,M.Quiros,'16 + talks in the next days) Huber,Konstandin,GN,'15

Beyond the minimal SUSY model, very strong EWPTs are feasible!!! Good news !!! In principle, in supersymmetric extensions of the SM, the EWPT can produce a sizable GW spectrum Within supersymmetry, the most promising scenarios seem those where the EWPT occurs in 2 steps Likely, the presented results are quite generic and cover other SUSY models (if they do not depart too much from minimality) (but see also M.Garcia-Pepin,M.Quiros,'16 + talks in the next days) … and often detectable at LISA since the third arm is decided !!!

Beyond the minimal COMPOSITE option (M. Chala,GN, I. Sobolev '15) Composite Model SO(7)/SO(6) → pNGB bosons: 1 singlet with Z + 1 singlet with appr.Z .+ 1 Higgs 2 2

Spectrum The light degrees of freedom (below the composite scale ) are the SM plus 1 real scalar, ,with an exact Z symmetry. This is the Dark Matter. 1 real scalar, , with an approx. Z symmetry (broken by suppressed higher- dimensional operators). This triggers the strong 2-step EWPT. 2 Regime I Regime II 2

Beyond the minimal model, very strong EWPTs are feasible!!! Result: A similar 2-step EWPT can also arise in Composite Higgs Models. (M.Chala, GN, I.Sobolev, 1605.08663) Filled circles → Runaway ; Empty circles → Non-Runaway. Regime I and II different DM scenarios.

Beyond the minimal model, very strong EWPTs are feasible!!! Result: A similar 2-step EWPT can also arise in Composite Higgs Models. (M.Chala, GN, I.Sobolev, 1605.08663) Red squares → Regime II ; Black circles → Regime I. Regime I and II different DM scenarios.

Then, yes, there exist well-motivated UV completions with EWPT testable at LISA

Conclusions It is guaranteed that GW experiments are going to detect unexplored physics LISA has a big potential to test cosmology and particle physics There are well-motivated models exhibiting very strong 1st – order phase transitions with new physics at roughly 10 GeV – 10 TeV with GW. LISA is sensitivity to these. The 1st – order EW phase transition falls exactly at the frequencies of LISA In some of these models there are parameter regions that are very hard to probe at the 14TeV LHC (or even future colliders) but can be tested at LISA !!! In some cases there is complementarity between LISA and colliders; synergy in others

Thank you! http://benasque.org/2017gw

Synergy/complemetarit y between LISA and colliders ?

In most of the models LISA-collider complementarity! Bottom-up singlet case with Z (Huang, Long, Lian-Tao ‘16; but see next talks) 2 In most of the models LISA-collider complementarity! Possibly, LISA can provide a precious scientific case to the collider community