Benedikt Kloss for the BESIII Collaboration

Institute of Nuclear Physics – Mainz University

Twelfth Conference on the Intersections of Particle and Nuclear Physics May 2015, Vail, Colorado

Benedikt Kloss - Universität Mainz 2

A new U(1) gauge boson γ‘ („dark photon“) might be the connection between the standard model and a dark sector:

dark sector kinetic mixing standard model

Benedikt Kloss - Universität Mainz 3

Existing exclusion limits on the dark photon (γ’) mass and mixing parameter (ε):

Dark photon coupling strength to SM matter:

Mixing parameter:

€

ε 2 =α 'α

€

α '

GOAL: make a contribution at

4 Benedikt Kloss - University of Mainz

5 Benedikt Kloss - University of Mainz

IHEP, Beijing

6

BEPCII Collider:

•  located in Beijing, China •  symmetric e+e- collider •  2 GeV < ECMS < 4.6 GeV •  data taken at : 2.9 fb-1

€

s = 3.77 GeV

Benedikt Kloss - University of Mainz

7

Main Drift Chamber

CsI Crystal Calorimeter

1T superconducting solenoid magnet

Time-Of-Flight System

Muon Chamber

Benedikt Kloss - University of Mainz (Graphic produced by Matthias Ulrich, Gießen)

8 Benedikt Kloss - University of Mainz

Benedikt Kloss - Universität Mainz 9

•  photon emitted in the initial state •  CMS energy lowered by the energy of the emitted photon ⇒  measurements at different energies possible

fixed Ecms = 3.77 GeV

new energy

E [GeV]

10 Benedikt Kloss - Universität Mainz

EMC EMC

tagged: photon hits EMC

untagged: photon leaves the detector

Two different analysis types: •  tagged: photon is detected in the Electromagnetic Calorimeter •  untagged: photon leaves the detector (most probable case)

11 Benedikt Kloss - Universität Mainz

Two different analysis types: •  tagged: photon is detected in the Electromagnetic Calorimeter •  untagged: photon leaves the detector (most probable case)

[rad]γθ0 0.5 1 1.5 2 2.5 3

even

ts /

10 m

rad

210

310

410

510

TAGGED&

UNTAGGED

&

UNTAGGED

&

boarders&of&EMC&

Benedikt Kloss - University of Mainz 12

Idea: Search for the ISR processes

€

e+e− →γ ISR ʹ′ γ →γ ISRµ+µ−

€

e+e− →γ ISR ʹ′ γ →γ ISRe+e−

and

Use untagged ISR events and 2.9 fb-1 data, taken at 3.77 GeV.

at

Benedikt Kloss - University of Mainz 13

Idea: Search for the ISR processes

€

e+e− →γ ISR ʹ′ γ →γ ISRµ+µ−

€

e+e− →γ ISR ʹ′ γ →γ ISRe+e−

and

Irreducible background:

€

e+e− →γ ISRγ*→γ ISRµ+µ−

€

e+e− →γ ISRγ*→γ ISRe+e−

QED process, same signature in

detector!

Dark photon signal would appear as peak on the QED background. ⇒  Peak search!

14 Benedikt Kloss - University of Mainz

Benedikt Kloss - University of Mainz 15

distance to interaction point Rxy < 1.0 cm Rz < 10.0 cm

acceptance of charged tracks 0.4 rad < θ< π – 0.4 rad

to supress background PID to select µ or e

# charged tracks = 2

total charge = 0

# photons = 0 (untagged analysis)

missing photon angle < 0.1 rad or > π – 0.1 rad

1C kinematic fit χ21C < 20

€

e+e− →µ+µ−γ ISREvent selection: and

€

e+e− →e+e−γ ISR

Benedikt Kloss - University of Mainz 16

MC simulated with PHOKHARA Eur.Phys.J. C24, 71-82 (2002) Phys. Rev. D77, 114005 (2008)

]2 [GeV/cµ2m1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

even

ts /

10 M

eV

410

MCγ-µ+µ MCγ-π+π

datada

ta /

MC

0.90.920.940.960.98

11.021.041.061.08

1.1

]2 [GeV/cµ2m1.6 1.8 2 2.2 2.4

ev

en

ts /

10

Me

V

2800

3000

3200

3400

3600

da

ta /

MC

0.90.920.940.960.98

11.021.041.061.08

1.1

]2 [GeV/cµ2m1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

even

ts /

10 M

eV

410

MCγ-µ+µ MCγ-π+π

data

data

/ M

C

0.90.920.940.960.98

11.021.041.061.08

1.1

QED process - irreducible background

Data-MC efficiency corrections applied on MC.

See talk on Wednesday:

“Measurement of the π+π- cross section at BESIII”

(PPHI, 16:40)

PRELIMINARY

J/ψ

Benedikt Kloss - University of Mainz 17

]2 [GeV/c2em1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

even

ts /

10 M

eV

410

]2 [GeV/c2em1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

even

ts /

10 M

eV410

MCγ-e+ e data

data

/ M

C

0.90.920.940.960.98

11.021.041.061.08

1.1

MC simulated with BABAYAGA 3.5

QED process - irreducible background

Nucl. Phys. B758, 227-253 (2006)

PRELIMINARY MC not corrected

for data-MC differences.

J/ψ

Benedikt Kloss - University of Mainz 18

/ ndf 2χ 196 / 145 0a 1.288e+02± 1.638e+05 1a 9.571e+01± -3.663e+05 2a 4.362e+01± 3.155e+05 3a 1.83e+01± -1.16e+05 4a 5.177e+00± 1.687e+04

[GeV]2em1.6 1.8 2 2.2 2.4 2.6 2.8 3

even

ts /

10 M

eV

20

40

60

80

100

120

140

310×

/ ndf 2χ 196 / 145 0a 1.288e+02± 1.638e+05 1a 9.571e+01± -3.663e+05 2a 4.362e+01± 3.155e+05 3a 1.83e+01± -1.16e+05 4a 5.177e+00± 1.687e+04

Fit of the continouus mass spectrum in data with a polynomial and look for

a peak in data:

fit to data

€

p(x) = a0 + a1x + a2x2 + a3x

3 + a4x4

PRELIMINARY

To get rid of the MC prediction:

Spare the region around J/ψ.

Benedikt Kloss - University of Mainz 19

/ ndf 2χ 196 / 145 0a 1.288e+02± 1.638e+05 1a 9.571e+01± -3.663e+05 2a 4.362e+01± 3.155e+05 3a 1.83e+01± -1.16e+05 4a 5.177e+00± 1.687e+04

[GeV]2em1.6 1.8 2 2.2 2.4 2.6 2.8 3

even

ts /

10 M

eV

20

40

60

80

100

120

140

310×

/ ndf 2χ 196 / 145 0a 1.288e+02± 1.638e+05 1a 9.571e+01± -3.663e+05 2a 4.362e+01± 3.155e+05 3a 1.83e+01± -1.16e+05 4a 5.177e+00± 1.687e+04

Fit of the continouus mass spectrum in data with a polynomial and look for

a peak in data:

No peaking structure found. ⇒  Set exclusion limit.

[GeV]*γm1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

σsi

gnifi

canc

e in

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

γ-µ+µ

[GeV]*γm1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

σsi

gnifi

canc

e in

0

0.5

1

1.5

2

2.5

γ-e+e

significance significance

fit to data

€

p(x) = a0 + a1x + a2x2 + a3x

3 + a4x4

PRELIMINARY

PRELIMINARY PRELIMINARY

To get rid of the MC prediction:

µ e

Spare the region around J/ψ.

Benedikt Kloss - University of Mainz 20

event selection

compare with MC prediction

fit the invariant mass distribution in data

search for a peaking structure

calculate the exclusion limit between 1.5 GeV/c2 and 3.4 GeV/c2

check whether huge background is left

none is found

get rid of any data-MC comparison in the final result

below 1.5 GeV/c2: pion background is dominant

above 3.4 GeV/c2: too close to production threshold

Benedikt Kloss - University of Mainz 21

90% confidence level (CL) calculated with the algorithm by Rolke et al. (TRolke)

]2 [GeV/c2lm1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

even

ts /

10 M

eV

-2000

-1500

-1000

-500

0

500

1000

1500

2000

data - fit CL = 90%

combined statistics, including systematics

Nucl.Instrum.Meth., A551, 493-503 (2005)

PRELIMINARY

Benedikt Kloss - University of Mainz 22

J. D. Bjorken, R. Essig, P. Schuster, and N. Toro, Phys. Rev., D80, 075018 (2009)

€

dσ(e+e− →γ 'γ ISR → l+l−γ ISR )dσ(e+e− →γ * γ ISR → l+l−γ ISR )

=3π

2N fl+ l− ⋅

ε 2

α⋅mγ '

δm

We want to calculate it in bins of the mixing parameter ε:

Benedikt Kloss - University of Mainz 23

mass resolution, determined with MC

fine structure constant

number of decay modes of dark photon containing phase space

(contains R ratio, see next slide)

dark photon mass

mixing parameter determined exclusion limit

€

dσ(e+e− →γ 'γ ISR → l+l−γ ISR )dσ(e+e− →γ * γ ISR → l+l−γ ISR )

=3π

2N fl+ l− ⋅

ε 2

α⋅mγ '

δm

Benedikt Kloss - University of Mainz 24

taken from PDG 2014

€

R =σ(e+e− →hadrons)σ(e+e− →µ+µ−)

€

Γtot = Γ(γ '→e+e−) +Γ(γ '→µ+µ−)⋅ (1+ R( s))€

N fl+ l− =

ΓtotΓ(γ '→ l+l−)

€

Γ(γ '→ l+l−) =αε 2

3mγ '2 mγ '

2 − 4ml2 (mγ '

2 + 2ml2) Phys. Rev. D88, 015032 (2013)

[GeV]s1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4

-2un

certa

inty

in 1

0

0.05

0.06

0.07

0.08

0.09

0.1

0.11

0.12

uncertainty of the R ratio

Benedikt Kloss - University of Mainz 25

Completely dominated by the uncertainty of the R ratio (everywhere above 5%)

taken from PDG 2014

Systematic uncertainty is estimated and implemented bin-by-bin (possible with TRolke algorithm1)

1 see https://root.cern.ch/root/html/tutorials/math/Rolke.C.html

26 Benedikt Kloss - University of Mainz

Now we are able to calculate the BESIII limit.

27 Benedikt Kloss - University of Mainz

PRELIMINARY

28 Benedikt Kloss - University of Mainz

•  Goal of the analysis is to search a dark photon signal using and events

•  no evidence has been found between 1.5 and 3.4 GeV/c2

•  an exclusion limit with 90% confidence has been calculated with the TRolke algorithm in bins of the mass and mixing parameter of the dark photon

•  values down to ε < 710-3 can be excluded

•  best exclusion limit in this mass range

€

µ+µ−γ ISR

€

e+e−γ ISR

Thank you for your attention!

PRELIMINARY

Benedikt Kloss - University of Mainz 29

Benedikt Kloss - University of Mainz 30

€

e+e− →µ+µ−γ ISR

€

e+e− →e+e−γ ISR

€

DLL(µ,e) = 2⋅ log p(µ)p(e)

⎛

⎝ ⎜

⎞

⎠ ⎟ > 0

,e)µDLL(-10 -8 -6 -4 -2 0 2 4 6 8 10

even

ts

0

500

1000

1500

2000

2500

3000

3500

4000 MCγ-µ+µ ) MCγ(n -e+ e

E / p > 0.8

E/p0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

even

ts0

5000

10000

15000

20000

25000 π µ

e

Benedikt Kloss - University of Mainz 31

over 10 bins

PRELIMINARY