Axions and Axion-Like Particles

Axions and Axion-like Particles

Belén Gavela Univ. Autónoma de Madrid and IFT

MORIOND Electroweak 2017

H2020 We will consider the SM plus a generic scalar ﬁeld a with derivative couplings to SM particles

and free scale fa: an ALP (axion-like particle)

general effective couplings

This is shift symmetry invariant: ~ Goldstone boson

Brivio, Gavela, Merlo, Mimasu, No, del Rey, Sanz 2017 arXiv:1701.05379 We will consider the SM plus a generic scalar ﬁeld a with derivative couplings to SM particles

and free scale fa: an ALP (axion-like particle)

general effective couplings

Why?

Brivio, Gavela, Merlo, Mimasu, No, del Rey, Sanz 2017 arXiv:1701.05379 Is the Higgs the only (fundamental?) scalar in nature? Or simply the ﬁrst one discovered? The spin 0 window Window on spin 0

The SM Higgs is a ~ doublet of SU(2)L The spin 0 window Window on spin 0

The SM Higgs is a ~ doublet of SU(2)L

What about a singlet (pseudo) scalar?

Strong motivation from fundamental problems of the SM S a

Rocio del Rey Strong motivation for singlet (pseudo)scalars from fundamental SM problems

The nature of DM is unknown

It may be a (SM singlet) scalar S the “Higgs portal”

δL = Φ+ΦS2

Silveira+Zee; Veltman+Yndurain; Patt+Wilczek… Strong motivation for singlet (pseudo)scalars from fundamental SM problems

The nature of DM is unknown The strong CP problem Why is the QCD θ parameter so small?

~μν LQCD⊃θ GμνG

It may be a (SM singlet) scalar S the “Higgs portal” Α dynamical U(1)A solution

δL = Φ+ΦS2 the axion a It is a pGB: ~only derivative couplings ∂μ a

Silveira+Zee; Veltman+Yndurain; Patt+Wilczek… Also excellent DM candidate Peccei+Quinn; Wilczek… Strong motivation for singlet (pseudo)scalars from fundamental SM problems

The nature of DM is unknown The strong CP problem Why is the QCD θ parameter so small? ~ a_ μν LQCD⊃ GμνG fa It may be a (SM singlet) scalar S the “Higgs portal” Α dynamical U(1)A solution

δL = Φ+ΦS2 the axion a It is a pGB: ~only derivative couplings ∂μ a

Silveira+Zee; Veltman+Yndurain; Patt+Wilczek… Also excellent DM candidate Peccei+Quinn; Wilczek…

In “true QCD axion” models: ma fa = cte.

1/fa

ma ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986 ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986 * * Models assuming * -5 -2 9 12 SM gauge group: 10 < ma < 10 eV , 10 < fa <10 GeV ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986 * * Models assuming * -5 -2 9 12 SM gauge group: 10 < ma < 10 eV , 10 < fa <10 GeV γ Intensely looked for experimentally… gaγ (GeV-1) γ

(GeV-1) “True” QCD axion band

= “Invisible axion” e.g. KSVZ, DFSZ…

v<< fa —> EW hierarchy problem

“True” QCD region ma (eV) … and theoretically https://arxiv.org/pdf/1611.04652.pdf γ

gaγ (GeV-1) γ

(GeV-1) Reﬁned KSVZ axion band: up and thinner

from ΩDM + Landau-poles analysis

(Luzio+Mescia+Nardi 2017)

v<< fa —> EW hierarchy problem

“True” QCD region ma (eV) … and theoretically https://arxiv.org/pdf/1611.04652.pdf γ

gaγ (GeV-1) γ

(GeV-1) QCD axiﬂavon band (creative view)

(Calibbi et al. 2016)

v<< fa —> EW hierarchy problem

“True” QCD region ma (eV) … and theoretically https://arxiv.org/pdf/1611.04652.pdf γ

gaγ (GeV-1) γ

(GeV-1) QCD axiﬂavon band (creative view)

(Calibbi et al. 2016)

Excluded for axiﬂavon by rare meson decays

v<< fa —> EW hierarchy problem

“True” QCD region ma (eV) … and theoretically https://arxiv.org/pdf/1611.04652.pdf ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986 * * Models assuming * -5 -2 9 12 SM gauge group: 10 < ma < 10 eV , 10 < fa <10 GeV ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986 * * Models assuming * -5 -2 9 12 SM gauge group: 10 < ma < 10 eV , 10 < fa <10 GeV

* Models enlarging the strong SM gauge sector, with scale Λ’ ? ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986

-5 -2 9 12 10 < ma < 10 eV , 10 < fa <10 GeV

* Models enlarging the strong SM gauge sector, with scale Λ’ ? , ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986

-5 -2 9 12 10 < ma < 10 eV , 10 < fa <10 GeV

* Models enlarging the strong SM gauge sector, with scale Λ’ ? , ma vs scale fa ga ~1/fa

In QCD-like theory because of explicit U(1)A breaking at quantum level (instantons, Λ)

QCD

Choi et al. 1986

-6 -3 9 12 10? < ma < 10 eV , 10 < fa <10? GeV

* Models enlarging the strong SM gauge sector, with scale Λ’ ? ,

relax the parameter space Recent activity on heavy “true” axions

* Enlarging the strong SM gauge group, with scale Λ’:

Dimopoulos+Susskind 79, Tye 81…Rubakov 97… Berezhiani+Gianfagna+Gianotti 01… surge since 2016!: : Gherghetta+Nagata+Shifman , Chiang et al., Khobadize… Hook and many collaborators, Dimopoulos et al. …

e.g. SU(3)c x SU(N’) both conﬁning

ΛQCD Λ’

a Recent activity on heavy “true” axions

* Enlarging the strong SM gauge group, with scale Λ’:

e.g. SU(3)c x SU(N’) both conﬁning

ΛQCD Λ’

* The ugly part: θ and θ’ Recent activity on heavy “true” axions

* Enlarging the strong SM gauge group, with scale Λ’:

e.g. SU(3)c x SU(N’) both conﬁning

ΛQCD Λ’

* The ugly part: θ and θ’

—>To reabsorb both : uniﬁcation, and/or SM mirror world related by Z2, or other constructions … all require tunings

Nothing works very nicely, but there is movement

—> e.g. fa ~ TeV, ma ~ MeV - TeV still solve the strong CP problem * Much territory to explore for heavy ‘true” axions and for ALPs e.g. ~ TeV

CAST + SUMICO

Jaeckel+ Spannowsky 2015 Log 10ma (eV) “True” QCD axion * Much territory to explore for heavy ‘true” axions and for ALPs e.g. ~ TeV

CAST + SUMICO

Jaeckel+ Spannowsky 2015 Log10ma (eV) “True” axion region ampliﬁes?? * Much territory to explore for heavy ‘true” axions and for ALPs

~ TeV

CAST + SUMICO

Jaeckel+ Spannowsky 2015 Log10ma (eV) “True” axion region ampliﬁes?? (Pseudo)Goldstone Bosons also in many BSM theories

* e.g. Extra-dim Kaluza-Klein: 5d gauge ﬁeld compactiﬁed to 4d the Wilson line around the circle is a GB, which behaves as an axion in 4d

* a Moriond example: the “relaxion" (G. Perez talk) is not a GB but part of its couplings are purely derivative as those of ALPs, e.g. * Much territory to explore for axions and ALPs e.g. ~ TeV

CAST + SUMICO

Strongderivative case for couplings looking everywhere of a spin 0 forparticle purely

Jaeckel+ Spannowsky 2015 Log10ma (eV) We will consider the SM plus a generic scalar ﬁeld a with derivative couplings to SM particles

and free scale fa: an ALP (axion-like particle)

general effective couplings

THEORY plus NEW SIGNALS at colliders

Brivio, Gavela, Merlo, Mimasu, No, del Rey, Sanz 2017 arXiv:1701.05379 Up to date, phenomenological studies have mostly focused on ALP couplings to fermions and photons But because of SU(2)xU(1) gauge invariance, a-γγ should come together with a-γΖ, a-ZΖ and a-W+W- :

+ …. ALP-Linear= effective Lagrangian at NLO SM EFT If only bosonic ALP-operators are considered:

Georgi, Kaplan, Randall 1986 ALP-Linear= effective Lagrangian at NLO SM EFT If only bosonic ALP-operators are considered:

SM higgs doublet

Georgi, Kaplan, Randall 1986 ALP-Linear= effective Lagrangian at NLO SM EFT If only bosonic ALP-operators are considered:

a a

Only fermionic a-Higgs couplings

Georgi, Kaplan, Randall 1986 ALP-Linear= effective Lagrangian at NLO SM EFT If only bosonic ALP-operators are considered:

Note: NO a-Higgs purely bosonic couplings

Georgi, Kaplan, Randall 1986 ALP-Linear= effective Lagrangian at NLO SM EFT Complete basis (bosons+fermions):

where Xψ is a general 3x3 matrix in ﬂavour space

Note: NO a-Higgs bosonic couplings

Georgi + Kaplan + Randall 1986 Choi + Kang + Kim, 1986 Salvio + Strumia + Shue, 2013 ALP-Linear effective Lagrangian at NLO

Complete basis (bosons+fermions):

analysiswhere parameters: Xψ is a general 3x3 matrix in ﬂavour space

Note: NO a-Higgs bosonic couplings

Georgi + Kaplan + Randall 1986 Choi + Kang + Kim, 1986 Salvio + Strumia + Shue, 2013 ALP-Linear effective Lagrangian at NLO

Complete basis (bosons+fermions):

where Xψ is a general 3x3 matrix in ﬂavour space contain a-γγ and other couplings

Georgi + Kaplan + Randall 1986 Choi + Kang + Kim, 1986 Salvio + Strumia + Shue, 2013 Because of SU(2)xU(1) gauge invariance, a-γγ comes together with a-γΖ, a-ZΖ and a-W+W- :

a-γγ studies only bounds this combination of couplings

90% CL: Because of SU(2)xU(1) gauge invariance, a-γγ comes together with a-γΖ, a-ZΖ and a-W+W- :

Largely disregarded up to very very recently

only a tiny bit in Jaeckel+Spannowsky 2015 We analyzed the impact at LEP, LHC and HL-LHC of bosonic effective a-SM couplings:

Largely disregarded up to very very recently

only a tiny bit in Jaeckel+Spannowsky 2015 We analyzed the impact at LEP, LHC and HL-LHC of bosonic effective a-SM couplings: We analyzed the impact at LEP, LHC and HL-LHC of bosonic effective a-SM couplings:

- New signals: mono-Z, mono-W, associated aWγ, att 100 0.20

1 linear Accelerator constraints on a-γ-γ linearand a-γ-Z 10 c127 = 0.2 0.15 c127 = 0.2 γ 2 c ˜ ,c˜ 10 c127 =0.2 B W c127 =0.2 ] a 1 0.10 a LEP impact

0 10 3 10 0.20 γ , Z 0 linear linear

1 ] 10 10 0.20 1 [TeV 4 c127 = 0.2 0.05 0.15 c127 = 0.2

linear linear 10 1 a 10 f 2 10 c127c127=0.2= 0.2 0.15 c127 = c0.2127 =0.2 / ] |

1 2 5 10 c127 =0.2 0.10 c127 =0.2 ˜ ] 10 3 [TeV 0.00 W 10 1 0.10 c 3 a f

10 ] 2 ✓ 1 / [TeV

] 0.05 s 6 4 1 ˜

10 [TeV

10 B 0.05 a 4 f 10 + a 0.05 c f / | /

˜ 5 | B

˜ 5 10 [TeV 0.00 ˜ 7 W 10 [TeV 0.00 W c

10 c a c a 2 ✓ f f 2 ✓ 2 ✓ Very constrained by / / s c s 6 6 ˜

0.10 ˜

10 B | 10 B γ γ

+ 0.05 axion- - c + 0.05 8 c

˜ B

10 ˜ B 7 7 c 10 c 2 10 ✓ 2 ✓ c 0.10 | c 8 0.10 9 | 0.15 10 8 10 10 9 0.15 10 10 9 0.15 10 ma≤1 MeV 10 10 0.20 10 0.20 3 2 1 0 3 12 1 2 0 31 2 3 0.4 0.4 0.20.2 0.00.0 0.2 0.20.4 0.4 10 10 1 0.20 1 3 2 41s2✓1(c ˜0 c ˜ )/f1a [TeV 2] 3 0.4 c 0.2˜ /fa [TeV0.0 ] 1 0.2 0.4 4s2✓(cW˜ cB˜ )/fa [TeV ] W B cW˜W /fa [TeV ] 1 1 4s (c c )/f [TeV ] c /f [TeV ] 2✓ W˜ B˜ a W˜ a 100 0.20

1 linear Accelerator constraints on a-γ-γ linearand a-γ-Z 10 c127 = 0.2 0.15 c127 = 0.2 γ 2 c ˜ ,c˜ 10 c127 =0.2 B W c127 =0.2 ] a 1 0.10 a LEP impact

0 10 3 10 0.20 γ , Z 0 linear linear

1 ] 10 10 0.20 1 [TeV 4 c127 = 0.2 0.05 0.15 c127 = 0.2

linear linear 10 1 a 10 f 2 10 c127c127=0.2= 0.2 0.15 c127 = c0.2127 =0.2 / ] |

1 2 5 10 c127 =0.2 0.10 c127 =0.2 ˜ ] 10 3 [TeV 0.00 W 10 1 0.10 c 3 a f

10 ] 2 ✓

1 Constrained at LEP / [TeV

] 0.05 s 6 4 1 ˜

10 [TeV γ

10 B 0.05 a 4 through axion- -Z f 10 + a 0.05 c f / | /

˜ 5 | B

˜ 5 10 [TeV 0.00 ˜ 7 W 10 [TeV 0.00 W c

10 c a c a 2 ✓ f f 2 ✓ 2 ✓ / / s c s 6 6 ˜

0.10 ˜

10 B |

10 B

+ 0.05 c + 0.05 8 c

˜ B

10 ˜ B 7 7 c 10 c 2 10 ✓ 2 ✓ c 0.10 | c 8 0.10 9 | 0.15 10 8 10 10 9 0.15 10 10 9 0.15 10 ma≤1 MeV 10 10 0.20 10 0.20 3 2 1 0 3 12 1 2 0 31 2 3 0.4 0.4 0.20.2 0.00.0 0.2 0.20.4 0.4 10 10 1 0.20 1 3 2 41s2✓1(c ˜0 c ˜ )/f1a [TeV 2] 3 0.4 c 0.2˜ /fa [TeV0.0 ] 1 0.2 0.4 4s2✓(cW˜ cB˜ )/fa [TeV ] W B cW˜W /fa [TeV ] + even more at LHC! 1 1 4s (c c )/f [TeV ] c /f [TeV ] 2✓ W˜ B˜ a W˜ a 1 1 d Missing ET if a stable inside detector d MET

10-1

derivative coupling: ALPs

10-2

non-derivative coupling: SM, DM h-portal … 20 40 60 80 100 120 140 Missing ET (GeV) Mono - Z and mono - W ALP signals missing energy q Z γ mono - Z: Z, Z ee selection Z µµ selection CMS-PAS-EXO-16-010 101 ! 101 Z ! q’ aa Data l Z+jets 100 100 Nonresonant 10e.g.1 10 1 WZ 3`⌫ ! CMS-PAS-EXO-16-010 ZZ 2` 2⌫ Exclusions Z ee selection Z -1 µµ selection 1 1 4 CMS-PAS-EXO-16-010 4 4 4 2 101 Z ee selection! 1 10 2 Z µµ10selection! 10 10 10 g [GeV] Data ! aZ 10 10 ! 10 10 ! Data i pp→ZZ(llνν) 3ab 0 0 Z+jets Z+jets h /c BR( inv.) 10100 100 10 a Nonresonant c ! f Nonresonant W 3 3 pp aW ± 10 1 1 1 1 10 # Events/ GeV 10 10 WZ 3`⌫ WZ 3`⌫ ! 10 10 ! c 10 ZZ 2` 2⌫ ! 1f pp aZ 2 2 ! ZZ 2` 2⌫ ⇥ ! 10 2 10 2 ! 4 10 10 4 c 10 Projections 10 10 3 c 3 23 3 3 3 10 W 10 10 10 10 10 c ⇥ 1 # Events/ GeV 3 3 c 10 W pp aW ± 3000 fb 10 c 10 1f # Events/ GeV W ⇥ ! 4 4 c 10 c1 10 1 5 10 10 5 2 ⇥ f⇥ pp aZ 3000 fb 10 4 4 10 ! 10 10 c2 10 1 5 5 ⇥ pp aW ± 300 fb 10 10 ! 5 5 1 10 10 pp aW ± 3000 fb Z width (LEP) 500500 1000 1000 500 1000 Mono-Z (LHC) 500 1000

Exclusion region for ! E/ T [GeV] E/ T [GeV] 2 2 2 2 1 E/ [GeV] 10 E/ [GeV] 10 10 10 pp ah 3000 fb T500 1000 95%CL500 1000T2 ! fa/cW˜ = t✓fa/cB˜ fa/c6 fa/a˜2D fa/a˜3 E/ T [GeV] E/ T [GeV]

3ab-1: assuming no improvement in systematics Interesting very recent development:

from rare meson decays

B —> K a , K—>π a….. a —>γγ

a a

ma (GeV) Izaguirre+Lin+Shuve 2016 Interesting very recent development:

from rare meson decays

B —> K a , K—>π a….. a —>γγ

a a

But several ops. may contribute:

di u,c,t dj

ma (GeV) +Del Rey et al. in preparation Izaguirre+Lin+Shuve 2016

Higgs EFTs

or (= non-linear) Linear= Chiral SM EFT Higgs EFTs

Linear= Chiral SM EFT Higgs ﬁeld: Φ= U h is in an exact SU(2)L doublet U=

Longitudinal W,Z Higgs EFTs

Linear or Chiral (non-linear)

in chiral: Φ= U

U= h may not be an exact SU(2)L doublet

Longitudinal W,Z Higgs EFTs

Linear or Chiral (non-linear)

in chiral: Φ= U

Longitudinal W,Z Τypical of “composite Higgs” models

e.g. in SO(5)/SO(4): φ Higgs EFTs

Linear or Chiral (non-linear)

in chiral: Φ= U

Longitudinal W,Z

Feruglio 93; Grinstein+Trott 07; Contino et al.10 Higgs EFTs

Linear or Chiral (non-linear)

in chiral: Φ= U

Longitudinal W,Z

independent ! some couplings decorrelate: more operators at given order Feruglio 93; Grinstein+Trott 07; Contino et al.10 Higgs EFTs

Linear or Chiral (non-linear)

LO: where

with Higgs EFTs

Linear or Chiral (non-linear)

LO: where Higgs EFTs

Linear or Chiral (non-linear)

LO: where

as in the linear case Higgs EFTs

Linear or Chiral (non-linear)

LO: where

ALP-Higgs couplings survive !! (unlike linear case)

as in the linear case Higgs EFTs

Linear or Chiral (non-linear)

LO: where

ALP-Higgs couplings survive !! (unlike linear case)

New additional signals: mono-h, BSM Higgs decays Non-standard Higgs decays

ATLAS and CMS 7 and 8 TeV

GeV

for Mono-Higgs : pp —> a h

q h Z

q' a

HL-LHC sensitivity:

GeV

“Les Houches 2015” Higgs EFTs

Linear or Chiral (non-linear)

LO: where

NLO, bosonic custodial preserving: NLO bosonic, custodial breaking:

We also determined the complete basis of non-redundant bosonic +fermionic couplings at NLO Αssociated production pp —> a Wγ

95%CL

LHC 13 TeV 2σ reach

ﬂattish MET are ALP signals Conclusions

* (pseudo) Goldstone Bosons in solutions to fundamental SM problems and BSM theories—> derivative couplings. * Strong case for hunting them

* New theoretical development: ALP effective Lagrangian for non-linear EWSB. —> ALP-Higgs-V signals!

* * New ALP signals from linear(SMEFT) and non-linear Lags. * MET —> mono-γ, -W/Z, -h, ΓBSM(h), etc. besides rare decays

Fish for them in your data!

* To do: many prompt and displaced signals (with high ET/pT dependence) * if a decaying inside detector Backup Validity of the EFT fa must be signiﬁcantly larger than the typical energies of the process.

For mono-W —> Correlation plot ———————-> max Validity of the EFT, mT < fa vs < fa 500

450

400

350

(GeV) Preliminary ˆ s 300 p

250 probability density of 200 mono-W events

150 150 200 250 300 350 400

mT (GeV) e.g. the difference between a cut in mTmax and max at 350 GeV would be 16% of events; for 450% it would be 10% etc. ALP stability at the LHC vs ma e.g. for ma = 1MeV Final state radiation off a top t

Yt t

Compare with susy searches of ttbar+2 neutralinos

susy with 800 GeV neutralinos Present exclusion limits on cW from mono-W and mono-Z

Prospects on cW from mono-Z Mono-W Associated aWγ Associated aWγ

Main backg.: Wγ, measured in LHC 7TeV, scaled here by 20% to account for subdominant backs. Parton level analysis. Sensitivity reach: Izaguirre, Lin, Shuve Present bounds on gluon-ALP couplings

Mimasu+Sanz recasting ATLAS and CMS bounds in addition to K—>π, SN, etc…

ψ Present bounds on fermion-Alp couplings

Beam Dump: (Dolan et al. 2014) XENON100: (Aprile et al. 2014) Red Giants: (Viaux et al. 2013) Bounds on photon-ALP coupling e.g. 90% CL:

(ma (GeV)) (ma (GeV)) Mimasu+Sanz 2015