Putting the dip in diphoton Sophie Renner DAMTP, Cambridge (soon - - PowerPoint PPT Presentation

▶

Jun 29, 2023 391 likes •545 views

Putting the dip in diphoton Sophie Renner DAMTP, Cambridge (soon Mainz) Based on Exploring Peaks and Valleys in the Diphoton Spectrum (1607.06074) with Nathaniel Craig and Dave Sutherland Sophie Renner DAMTP, Cambridge (soon Mainz)

SLIDE 1

Putting the dip in diphoton

Sophie Renner DAMTP, Cambridge (soon Mainz) Based on “Exploring Peaks and Valleys in the Diphoton Spectrum” (1607.06074) with Nathaniel Craig and Dave Sutherland

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 1 / 15

SLIDE 2

Outline

There was a bump in diphotons
Fits were done to find the preferred “mass”, “width” and “cross-section” of a

possible resonance

But how might resonance-continuum interference affect these numbers?
And can we use interference effects to deduce the most likely “spin”?
Obviously we now know that the bump was just a statistical fluctuation of the

background

Can we learn anything for if and when it happens again (and for real)?
Can we dip-hunt as well as bump hunt?

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 2 / 15

SLIDE 3

The diphoton bump: ATLAS and CMS fits to 2015 data

CMS:

Best fit with 13 TeV data ∼ 760 GeV
Preference for a narrow width
2.85σ local significance (< 1σ global)

ATLAS:

Best fit with 13 TeV data ∼ 750 GeV
Preference for a ∼ 45 GeV width
3.6σ local significance (2σ global)

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 3 / 15

SLIDE 4

Particle explanations

Spin 0 (scalar or pseudoscalar) or spin 2
qq initiated or gg initiated
wide or narrow

Effective Lagrangians: L0++ ⊃ − 1 Λg φG µνGµν − 1 Λγ φF µνFµν − cqφqq; L0−+ ⊃ − 1 Λg φG µν ˜ Gµν − 1 Λγ φF µν ˜ Fµν − icqφqγ5q; L2++ ⊃ − 1 ΛT φµν

−F µλF ν

λ + 1

4g µνF 2 − G µλG ν

λ + 1

4g µνG 2 +1 2iq(γµ∂ν + γν∂µ)q − ig µνq✁ ∂q

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 4 / 15

SLIDE 5

Signal-background interference P =

A ˆ s−M2+iΓ √ ˆ s

Calculate interference for each scenario using helicity

amplitudes

A, M, Γ all free parameters

At √s = 13 TeV and √ ˆ s = 750 GeV, the background is mostly q¯ q initiated contributions = ⇒ interference effects will be largest in q¯ q initiated signals Lots of previous work on interference ef- fects in Higgs to gamma gamma (eg. hep- ph/0302233, 1208.1533), but the very narrow width of the Higgs means effects are small Other papers focus on effects of interference in the interpretation of the 750GeV excess, using benchmark models (eg 1601.00006, 1605.00542)

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 5 / 15

SLIDE 6

Bumps with interference

No phase:

BG BG+S BG+PS BG+T

700 800 900 mγγ[GeV] 0.01 0.02 0.03 0.04 dσ/dm [fb/GeV]

gg → γγ, s =13 TeV

BG BG+S BG+T

700 800 900 mγγ[GeV] 0.04 0.06 0.08 0.10 0.12 0.14 dσ/dm [fb/GeV]

uu → γγ, s =13 TeV

Spin 2 qq initiated with phases:

No int. ϕ = 0 ϕ = 90o ϕ = 180o

700 800 900 mγγ[GeV] 0.04 0.06 0.08 0.10 0.12 0.14 0.16 dσ/dm [fb/GeV]

In the EFT no phases are possible in couplings due to hermiticity of effective lagrangian Using M = 750 GeV, Γ = 40GeV

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 6 / 15

SLIDE 7

Fits with interference: spin 2 qq-initiated

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472364.

Combined8+13

8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 8000 6000 8000 8000 8000 8000 8000 8000 8000 6000 6000 6000 8000 8000 8000 8000 8000 6000 6000 6000 4000 6000 8000 6000 8000 6000 6000 6000 4000 4000 8000 6000 8000 8000 6000 6000 4000 4000 8000 4000 8000 8000 6000 6000 4000 4000 8000 4000 8000 8000 6000 6000 6000 4000 8000 4000 8000

5 10 15 20

With interference

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472372.

Combined8+13

8000 6000 4000 4000 4000 8000 4000 6000 6000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 6000 6000 8000 6000 4000 6000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 6000 4000 4000 8000 4000 4000 8000 6000 6000 4000 4000 8000 4000 4000

5 10 15 20

Without interference Big change: due to

Dominance of qq initiated background (large interfering background =

⇒ large interference effects)

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 7 / 15

SLIDE 8

Fits with interference: spin 2 gg-initiated

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472376.

Combined8+13

8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 6000 4000 4000 8000 4000 4000 8000 6000 6000 4000 4000 8000 4000 4000

5 10 15 20 25

With interference

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472376.

Combined8+13

6000 6000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 4000 4000 4000 4000 8000 4000 4000 6000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 4000 4000 4000 8000 4000 4000 8000 6000 6000 4000 4000 8000 4000 4000

5 10 15 20 25

Without interference Some change: due to

Relative suppression of gg- initiated background

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 8 / 15

SLIDE 9

Fits with interference: spin 0 qq - initiated

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472375.

Combined8+13

50000 50000 50000 00000 50000 50000 50000 00000 250000 250000 250000 150000 150000 250000 150000 100000 250000 250000 200000 150000 150000 250000 150000 100000 250000 250000 200000 150000 150000 250000 100000 100000 250000 250000 200000 150000 150000 250000 100000 100000 250000 250000 200000 150000 150000 250000 150000 100000 250000 250000 200000 150000 150000 250000 150000 100000 250000 250000 200000 200000 150000 250000 150000 100000 250000 250000 250000 200000 150000 250000 150000 100000 25000 25000 25000 20000 15000 25000 15000 15000

10 20 30 40

With interference

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472375.

Combined8+13

50000 50000 50000 00000 50000 50000 50000 00000 250000 250000 250000 150000 150000 250000 150000 100000 250000 250000 200000 150000 150000 250000 150000 100000 250000 250000 200000 150000 150000 250000 100000 100000 250000 250000 200000 150000 150000 250000 100000 100000 250000 250000 200000 150000 150000 150000 150000 100000 250000 250000 200000 150000 150000 250000 150000 100000 200000 250000 250000 200000 150000 250000 150000 100000 250000 250000 250000 200000 150000 250000 150000 100000 25000 25000 25000 20000 15000 25000 15000 15000

10 20 30 40

Without interference No change: due to

No interference at LO

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 9 / 15

SLIDE 10

Fits with interference: spin 0 gg - initiated:

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472380.

Combined8+13

30000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 10000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 20000 15000 15000 15000 30000 10000 10000 3000 2000 1500 1500 1500 3000 1500 1000

5 10 15 20 25

With interference

700 720 740 760 780 10 20 30 40 50 60 70 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -1472380.

Combined8+13

20000 20000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 10000 10000 30000 10000 10000 20000 15000 15000 10000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 15000 15000 15000 10000 30000 10000 10000 20000 20000 15000 15000 15000 30000 10000 10000 3000 2000 1500 1500 1500 3000 1000 1000

5 10 15 20 25

Without interference V little change change: due to

PDF suppression of gg background compared to qq

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 10 / 15

SLIDE 11

Best fits for each scenario

600 650 700 750 800 850 900 mγγ[GeV] 2 4 6 8 10 12 Events/20 GeV

Spin 0 gg M=740GeV Γ=40GeV Λ=10000GeV

600 650 700 750 800 850 900 mγγ[GeV] 2 4 6 8 10 12 Events/20 GeV

Spin 0 qq M=740GeV Γ=30GeV Λ=100000GeV

600 650 700 750 800 850 900 mγγ[GeV] 2 4 6 8 10 12 Events/20 GeV

Spin 2 gg M=750GeV Γ=20GeV Λ=4000GeV

600 650 700 750 800 850 900 mγγ[GeV] 2 4 6 8 10 12 Events/20 GeV

Spin 2 qq M=750GeV Γ=5GeV Λ=6000GeV

These are fits to 13 TeV 2015 ATLAS data (their “Spin 0” dataset)

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 11 / 15

SLIDE 12

Peak-dip hunting? Dip hunting?

We have found that in diphotons, spin 2 qq-initiated resonances can look very different from a bump:

No int. ϕ = 0 ϕ = 90o ϕ = 180o

700 800 900 mγγ[GeV] 0.04 0.06 0.08 0.10 0.12 0.14 0.16 dσ/dm [fb/GeV]

NB This can be the case in

ther final states and for different quantum numbers too, mostly dependent on size of

interfering background (eg. in gg → S → tt, see eg 1608.07282 and many other refs) = ⇒ for WIDE resonances, other shapes are possible, which could be missed by simple bump hunting

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 12 / 15

SLIDE 13

Peak-dip hunting

Slight peak-dip structure in ATLAS and CMS diphoton spectrum (pre-2016, 13TeV) at 550GeV - what happens when you try to fit a spin 2 resonance?

520 530 540 550 560 570 10 20 30 40 50 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -97197.7

Combined13

7000 7000 7000 7000 7000 6000 7000 7000 7000 7000 7000 7000 5000 5000 7000 7000 7000 6000 5000 4000 5000 7000 7000 6000 5000 4000 5000 6000 7000 6000 5000 5000 5000 7000 7000 7000 6000 6000 5000 7000 7000 7000 7000 5000 5000 7000 7000 7000 6000 6000 5000 7000 7000 7000 6000 7000 7000 7000 7000 7000 7000 7000 7000 7000 7000 7000

2 4 6

Can produce a slightly improved fit to the data ∆χ2 ∼ 6 compared to BG only

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 13 / 15

SLIDE 14

Dip hunting

Slight dip in ATLAS and CMS diphoton spectrum (pre-2016, 13TeV) Try to fit a dip i.e. with a phase of 90 degrees in the amplitude A

380 400 420 440 460 5 10 15 20 25 30 35 40 M / GeV Γ / GeV

2 ln(L/L0);-2 ln L0 = -97199.1

Combined13

6000 5000 4000 3000 6000 4000 4000 3000 3000 5000 4000 3000 3000 3000 5000 5000 4000 3000 3000 6000 6000 5000 4000 3000 6000 4000 2500 2000 3000 5000 4000 2500 2000 2000 5000 4000 2500 2000 2000 6000 4000 2500 2000 2000 6000

1 2 3 4 5

Can produce a slightly improved fit to data ∆χ2 ∼ 5 compared to BG only

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 14 / 15

SLIDE 15

Summary

In some cases, interference matters
This depends on: spin, width, initial state, values of the PDFs at the resonance,

whether there is interference at LO...

Seems to be especially important in the diphoton channel for high mass spin 2

particles produced from q¯ q

Lessons for resonance searching in diphotons at the LHC:
Best fits of mass, width, amplitude can be affected
Possible to get some info on likely spin by looking at shape
Consider looking for other shapes in the data

Sophie Renner DAMTP, Cambridge (soon Mainz) Putting the dip in diphoton 15 / 15