Heavy Sterile Neutrinos at CEPC

Size: px

Start display at page:

Download "Heavy Sterile Neutrinos at CEPC"

Dominic Harmon
5 years ago
Views:

1 Wei Liao East China University of Science and Technology Based on arxiv: , in collaboration with X.H. Wu INPAC SJTU, Nov

2 Outline Review of neutrino mass low energy seesaw model constraint on low energy seesaw model experimental searches LHC, LEP ILC perspectives CEPC sensitivity with only a single heavy sterile neutrino in the low energy seesaw model with two degenerate or quasi-degenerate ν R

3 Neutrino mass Neutrino oscillation exp establish that neutrinos have masses and mixings e.g., data in solar ν exps converge to a region for which ν conversion happened

4 Neutrino mass Solar ν and atmospheric ν exps, confirmed by long baseline exp: m21 2 = ev 2, sin 2 θ 12 = 0.76 m32 2 = ev 2, sin 2 θ 23 = 0.60 Observation(discovery) by Daya bay, confirmed by RENO: sin 2 θ 13 = θ 13 = 8.8 ± 0.8 a magic number, according to Z.Z. Xing

5 Neutrino mass Direct measurement of neutrino mass

6 Neutrino mass Direct measurement of neutrino mass

7 Neutrino mass 0ν2β experiment

8 Neutrino mass 0ν2β experiment

9 Neutrino mass 0ν2β experiment

10 Neutrino mass Massive neutrinos can affect the evolution of cosmology, and get constrained by cosmological data

11 Neutrino mass The status of neutrino mass Neutrino mass, if Dirac type > ev (from oscillation data, for two neutrinos) < ev (from cosmology) Neutrino mass, if Majorana type > ev (from oscillation data, for two neutrinos) < ev (from neutrino-less double decay ) Neutrinos are unexpectedly very very light!

12 The low energy seesaw model Why neutrinos are so light? A simple way to answer: there are right-handed neutrinos e.g. SM gauge theory is extended to Left-right symmetric theory Pati-Salam model SO() Grand Unification In these theories, right-handed neutrinos are required by the representations of the gauge group Another possible answer: it is required for solving other problem, such as the production of baryon number in the universe

13 The low energy seesaw model Seesaw mechanism: light ν masses are remnants of the decoupling of heavy Majorana ν R. Lepton number symmetry is restored when m ν vanishes

14 The low energy seesaw model Leptogenesis:

15 The low energy seesaw model Seesaw mechanism(model) is a promising candidate of new physics beyond SM: However, there is no principle to tell us at what energy scale the seesaw is GUT scale, associated with Grand Unified Models 9 13 GeV, required by leptogenesis GeV-0 GeV scale, possible to probe on collider exp even kev scale, to accommodate Warm Dark Matter candidate

16 The low energy seesaw model ν s decay rate is suppressed by its small mass, no extra global quantum number needed, compared to CDM ν s decays mainly through ν s ν + 2 ν, 2ν + ν: τ νs = s ( ) 5 1 kev 8 M s Θ 2 Θ 2 = R es 2 + R µs 2 + R τs 2. τ νs much larger than the age of the universe 17 s ν s is a good dark matter candidate

17 The low energy seesaw model kev scale ν R1 (ν s ) dark matter in low energy seesaw A low energy seesaw(kev scale ν R1 and GeV scale ν R2,3, νsm) (Asaka, Blanchet and Shaposhnikov, 2005) We found(he, Li and Liao, 2009) the νsm has an approximate Friedberg-Lee symmetry: ν R1 ν R1 + θ natural splitting of kev scale ν R1 and GeV scale ν R2,3 active neutrino masses either normal or inverse hierarchy large mixing of ν R2,3 with active neutrinos can be achieved 0νββ constraint can be satisfied for quasi-degenerate ν R2,3 even if mixings are large

18 The low energy seesaw model Moreover, low energy seesaw is possible to be detected on collider. In seesaw models, active neutrinos ν l (l = e, µ, τ) are mixtures of light neutrinos ν i (i = 1, 2, 3) and heavy sterile neutrinos N j (ν R ) ν l = i U liν i + j R ln j N j GeV-0 GeV ν R can be produced on collider via this mixing, e.g. on lepton collider

19 The low energy seesaw model Possible ways to detect GeV scale ν R on collider: If coupling/mixing with active neutrinos are large enough lepton number violating signal: same sign charged lepton e.g. on hadron collider: pp l ± 1 l ± 2 W l ± 1 l ± 2 jj resonance in ljj events e.g. on lepton collider: e + e Nν ew ν eνjj If coupling/mixing with active neutrinos are sufficiently small displaced vertex We discuss the search of heavy sterile neutrino via ljj events

20 Constraint on the low energy seesaw model The masses and mixing of active neutrinos can be produced by two GeV scale ν R (m ν ) ll = v 2 i Yli Yl im 1 i = i M i R ln i R l N i. One of light neutrinos is massless: light neutrinos are either IH or NH, not QD. For a single N, contribution of N to 0νββ decay R 2 en /m N, leading to R en 2 5 for GeV scale N. For 2 heavy sterile neutrinos N 1 and N 2, the amplitude of 0νββ decay is, A = F M 2 1 (R 2 en 1 M 1 + R 2 en 2 M 2 ) + FM 2 R 2 en 2 ( 1 M M1 2 ) the 1st term is small because of the neutrino mass matrix, the 2nd term can be small, if N 1 and N 2 are quasi-degenerate or degenerate.

21 Constraint on low energy seesaw model To have sufficiently large mixing with active neutrinos, one can find in this model R ln1 = 1 2 e ix+ y (U l2 m 1/2 2 e iφ2/2 iu l3 m 1/2 3 e iφ3/2 )(M 1 ) 1/2 and R ln2 = ±ir ln1 for NH R ln1 = 1 2 e ix+ y (U l1 m 1/2 1 e iφ1/2 iu l2 m 1/2 2 e iφ2/2 )(M 1 ) 1/2 and R ln2 = ±ir ln1 for IH

22 Active-sterile mixing R ln 00 correlation for NH 0 correlation for IH ( R N 2 + R µn 2 )/ R en 2 0 ( R N 2 + R µn 2 )/ R en R N 2 / R µn R N 2 / R µn 2 For NH, R µn and R τn larger than R en

23 Tevatron sensitivity -2-2 S L3 DELPHI m 4 (GeV) 8 fb A. Atre, T. Han, S. Pascoli, B. Zhang, p p l ± 1 l ± 2 W l ± 1 l ± 2 jj 5σ sensitivity with 8fb 1 R µn 2, R en 2 4 (50GeV), 2 (0GeV) S e m 4 (GeV) 8 fb -1

24 LHC sensitivity L3 DELPHI S -5-6 S e m 4 (GeV) A. Atre, T. Han, S. Pascoli, B. Zhang, pp l ± 1 l ± 2 W l ± 1 l ± 2 jj 5σ sensitivity with 0fb 1 R µn 2, R en 2 5 (50GeV), 3 (0GeV) m 4 (GeV)

25 CMS 2 V en 1 1 CMS fb (8 TeV) Mixing 1 1 CMS fb (8 TeV) 2 95% CL upper limit 2 95% CL upper limit 3 4 Expected Expected ± 1 s.d. Expected ± 2 s.d. Observed L3 (1992) L3 (2001) DELPHI ATLAS CMS 7 TeV Observed V en 2 V en V* µn Observed 2 2 V en + V µn 2 Observed V µn m N (GeV) CMS, and pp l ± N l ± 1 l ± 2 W l ± 1 l ± 2 jj 95% C.L. sensitivity with 19.7fb 1 R µn 2, R en 2 2 (0GeV) m N (GeV)

26 L3 1 L3 U e EXCLUDED -3 Observed 95% C.L. limit Expected 95% C.L. limit m N (GeV) L3, hep-ex/ and hep-ex/07014 e + e Nν ew ν eνjj 95% C. L., with 450pb 1 R en (80 205GeV)

27 Displaced vertex searches at future e + e collider PS191 Normal hierarchy NuTeV BAU CHARM PS191 Inverted hierarchy NuTeV BAU 2 U -9 SHiP 2 U -9 SHiP - BBN FCC-ee - BBN FCC-ee -11 Seesaw -11 Seesaw 1 HNL mass (GeV) 1 HNL mass (GeV) A. Blondel, E. Graverini, N. Serra, M. Shaposhnikov, Z Nν lw ν lνjj with 12 Z R ln 2 11 (50GeV)

28 Displaced vertex searches at future e + e collider θ 2-8 θ 2-1 θ CEPC M [GeV] -1 FCC-ee M [GeV] -9 ILC M [GeV] E cm m Z ; E cm 250 GeV; E cm 350 GeV; E cm 500 GeV; Conventional search 95 C.L. S. Antusch, E. Cazzato, O. Fischer, R ln (20 80GeV)

29 ILC expected 0.1 ATLAS VeN LEP2 LHC fb 1 EWPD 90 C.L ILC fb 1 ILC fb 1 ILC fb 1 ILC fb MN GeV S. Banerjee, P. Dev, A. Ibarra, T. Mandal, M.Mitra, e + e Nν ew ν eνjj 95% C.L. ILC with 0fb 1 or 500fb 1 R en (0 400GeV)

30 e + e Nν(+ ν) production CEPC operates at s = 240GeV with 5ab 1 with 2 IPs and years of operation. The cross section of t-channel is 2 order of magnitude larger than that of s-channel. R en has better sensitivity than R µn.

31 Production and decay σ/ R en 2 (pb) σ/ R µn 2 (pb) m N (GeV) e + e Nν, summing over neutrinos of all flavor and anti-neutrinos. m N (GeV) For a heavy neutrino of about 0GeV, σ/ R en 2 60pb and σ/ R µn 2 0.8pb above Z (or H) mass, two-body decay modes dominant, N lw, νz, νh, with Br(N lw ) 1/3.

32 Production and decay We obtain analytic expression for N decay through on-shell or off-shell W, Z or H e.g., for N l 1 l + 2 ν l 2, N l + 1 l 2 ν l 2 and l 1 l (N->e - µ + µ )(GeV) e-05 1e-06 1e-07 A B C 1e M N (GeV) A: using analytic expression; B: using = G 2 F m 5 N/(192π 3 ) up to m N < m W ; C: using = (N e W + )Br(W + µ + ν µ) with m N > m W.

33 Cuts signal e + e Nν, N ν ljj /E main background e + e W + W with one W decaying leptonically, and one W decaying hadronically basic cuts for lepton and jets to select the events p l T > GeV, ηl < 2.5, R ll > 0.4, p j T > GeV, ηj < 2.5, R jj > 0.4, R lj > 0.4. selection cuts to suppress background events from on-shell W decay M(l, /E) m W > 20 GeV, M(l, j 1, j 2 ) m N < GeV.

34 A single R ln R_eN=0.015 R_muN=0.1 significance, s N s / N s + N b, N as event number for the integrated luminosity of 0fb 1 m N 146GeV(150GeV) for l = e (l = µ) channel for the integrated luminosity of 5ab 1 m N 235GeV(205GeV) for l = e (l = µ) channel

35 Sensitivity on a single R ln R en 0fb fb -1 1ab -1 5ab -1 R µn 0fb fb -1 1ab -1 5ab -1 significance s = 5 m N (GeV) m N (GeV) a heavy neutrino mass of 120GeV for an example for l = e channel, R en = (R en = ) can be probed with 0fb 1 (5ab 1 ) for l = µ channel, R µn = 0.043(R µn = 0.016) can be probed with 0fb 1 (5ab 1 )

36 The seesaw model with correlated R ln 00 correlation for NH 0 correlation for IH ( R N 2 + R µn 2 )/ R en 2 0 ( R N 2 + R µn 2 )/ R en R N 2 / R µn 2 = ±ir ln R N 2 / R µn 2 for NH, R ln2 R ln1 = 1 2 e ix+ y (U l2 m 1/2 2 e iφ2/2 iu l3 m 1/2 3 e iφ3/2 )(M1 ) 1/2 R µn and R τn larger than R en for IH, R ln2 = ±ir ln1 R ln1 = 1 2 e ix+ y (U l1 m 1/2 1 e iφ1/2 iu l2 m 1/2 2 e iφ2/2 )(M1 ) 1/2

37 The seesaw model with correlated R ln s NH δcp =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ s IH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ m N (GeV) m N (GeV) significance s sl 2, l = e, µ, τ channel, 500fb 1 NH parameters, φ 1 = φ 2 = φ 3 = 0, e y = 5000 for NH NH, δ CP = π/2, R µn R τn R en, µ, τ dominant m N 152GeV ( R en , R µn R τn 0.034) δ CP = π/2, R µn R τn 2 R en, µ, τ dominant m N 206GeV ( R en 0.015, R µn R τn 0.028) with a larger R en for δ CP = π/2, t-channel production enhanced, µ, τ signal enhanced, compared with δ CP = π/2

38 The seesaw model with correlated R ln s NH δcp =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ s IH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ m N (GeV) m N (GeV) IH parameters, φ 1 = φ 2 = φ 3 = 0, e y = 00 IH, m N 162GeV, R en , R µn , R τn for δ CP = π/2 R en , R µn , R τn for δ CP = π/2 for both cases of δ CP = π/2 and δ CP = π/2 R en the same, R en 2 + R µn 2 + R τn 2 also the same size, leading to the same production rate and ljj decay rate, then total e + µ + τ significances the same, but different for e + µ significances

39 Effect of phases: δ CP and φ 2 s δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ NH s δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ IH φ 2 (π) m N = 150GeV and integrated luminosity 500fb 1 The e + µ and e + µ + τ significance depends on both Dirac phase δ CP and Majorana phase φ 2. The bumps as the varied φ 2. Take the case of e + µ significance for NH with δ CP = π/2 for an example. A bump at φ 2 1.5π φ 2 (π)

40 Effect of phases: the bump NH δ CP =π/2 s e+µ 20 R µn 2 /Σ R ln 2 20 R en 2 /Σ R ln 2 φ 2 (π) e + µ significance for NH with δ CP = π/2 for an example A bump in e + µ significance (dominated by the µjj) at φ 2 1.5π. 1. φ 2 (from 0 to 2π), R en 2 / R ln 2 and peaks at φ 2 = 2π, then t-channel production (dominate production), 2. φ 2 (from 0 to π), R µn 2 / R ln 2 and peaks at φ 2 π. For φ 2 > π, R µn 2 / R ln 2 and Br(N µjj), but compensated by production of e + e Nν. 3. µjj events and e + µ significance first, then and peaks at 1.5π, as φ 2 from 0 to 2π.

41 The seesaw model with 5ab 1 s NH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ s IH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ m N (GeV) After years of operation, CEPC may accumulate 5ab 1 data. NH parameters, φ 1 = φ 2 = φ 3 = 0, e y = 1750 NH, for δ CP = π/2, m N 124GeV, ( R en , R µn R τn 0.013) for δ CP = π/2, m N 184GeV ( R en , R µn R τn 0.0) m N (GeV)

42 The seesaw model with 5ab 1 s NH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ s IH δ CP =π/2 e+µ e+µ+τ δ CP = π/2 e+µ e+µ+τ m N (GeV) IH parameters, φ 1 = φ 2 = φ 3 = 0, e y = 350 IH, m N 130GeV, R en , R µn , R τn for δ CP = π/2 R en , R µn , R τn for δ CP = π/2 m N (GeV)

43 The seesaw model with 5ab 1 R µn 0fb fb -1 1ab -1 5ab -1 NH δ CP =π/2 R µn 0fb fb -1 1ab -1 5ab -1 IH δ CP =π/2 m N (GeV) m N (GeV) R µn 0fb fb -1 1ab -1 5ab -1 NH δ CP = π/2 R µn 0fb fb -1 1ab -1 5ab -1 IH δ CP = π/2 m N (GeV) m N (GeV)

44 The seesaw model with correlated R ln With 5ab 1, CEPC can reach a 5σ sensitivity of R en 3 and R µn 2 for a single mixing. With sizable R en, the significance of µ and τ channel will be enhanced, and further constrain R µn and R τn. The significance depends on both Dirac phase δ CP and Majorana phases. With correlated R ln, a search for all 3 lepton channels are helpful to constrain the model.

45 Summary We studied the sensitivity of CEPC in a low energy seesaw model with heavy neutrino mass of order of 0GeV and large mixing with active neutrinos. After years of operation, with 5ab 1, the sensitivity can reach R en 3 and R µn 2 for a single mixing, and the low energy seesaw model as well. A search for all 3 lepton channels are helpful to constrain the model.

46 Thank You

47 Backup slides

48 Backup Lepton Number Violation processes e + e Nl ± W l ± 1 l ± 2 e e W W jjjj + jjjj

49 Backup τ tagging leptonic τ µ and hadronic τ jj efficiency

50 Backup Efficiency of different cuts Table: The cross sections (unit fb) of signal (upper line) after imposing various cuts (a, b, c, d, e) sequentially, the background (lower line) and the significance after cuts with integrated luminosity of 500fb 1. Cuts (a) p j,l T > 1GeV, (b) pj,l T > GeV, (c) M(l, /E) m W > 20GeV, d) M(l, j 1, j 2) m N < 20GeV, (e) M(l, j 1, j 2) m N < GeV. parameters +cuts (a) +cuts (b) +cuts (c) +cuts (d) +cuts (e) significance A m N = 150GeV, R µn = B m N = 150GeV, R en = C m N = 90GeV, R en = D m N = 214GeV, R en =

Marco Drewes, Université catholique de Louvain. The Other Neutrinos IAP Meeting. Université libre de Bruxelles

Marco Drewes, Université catholique de Louvain The Other Neutrinos 21.12.2017 IAP Meeting Université libre de Bruxelles The Standard Model of Particle 125 GeV The periodic table of elementary particles