Searches for Dark Matter Production with CMS
Bodhitha Jayatilaka Fermilabon behalf of the CMS Collaboration
LHCP 2016 Lund, Sweden17 June 2016
LHCP2016, 17-Jun-2016 Bo Jayatilaka
The search for dark matter• Existence of dark matter is well established
from cosmological observations• Exact nature of it is unknown
• Ways to look for (particle) DM• Direct searches via DM-nucleon scattering• Indirect searches via DM annihilation• Search for production of DM at colliders
• LHC provides a prime laboratory for production of DM• Search for evidence of DM production along
with (visible) SM particles• Can characterize DM interaction as function
of spin/parity, coupling, etc
2
Introduction
• From cosmological observations, 85% of the matter comprised of dark matter (DM)
• Collider approach: DM production by colliding SM particles at high energies
2
Bullet Cluster (Image: NASA)
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Benchmark signatures for LHC DM searches• Searches for DM at the LHC look for ETmiss+X
• X = jet, W, Z, γ, H, tt, bb, t, etc.• Run 2: Adopt simplified models to interpret results (arXiv:1507.00966)
• Assume new massive particle which mediates DM-SM interaction• DM particle is a Dirac fermion χ• Keep to a minimal set of parameters
• Mmed, mχ, gSM, gDM, Γmed
3
18 atlas+cms dark matter forum
V, A(Mmed)
q̄
q
c̄(mc
)
c(mc
)g
gq gDM
Figure 2.1: Representative Feynmandiagram showing the pair productionof Dark Matter particles in associationwith a parton from the initial state viaa vector or axial-vector mediator. Thecross section and kinematics dependupon the mediator and Dark Mattermasses, and the mediator couplings toDark Matter and quarks respectively:(Mmed, m
c
, gc
, gq).
Lvector = gq Âq=u,d,s,c,b,t
Z0µ
q̄g
µq + gc
Z0µ
c̄g
µ
c (2.1)
Laxial�vector = gq Âq=u,d,s,c,b,t
Z0µ
q̄g
µ
g
5q + gc
Z0µ
c̄g
µ
g
5c. (2.2)
The coupling gq is assumed to be universal to all quarks. It is alsopossible to consider other models in which mixed vector and axial-vector couplings are considered, for instance the couplings to thequarks are axial-vector whereas those to DM are vector. As men-tioned in the Introduction, when no additional visible or invisibledecays contribute to the width of the mediator, the minimal widthis fixed by the choices of couplings gq and g
c
. The effect of largerwidths is discussed in Section 2.5.2. For the vector and axial-vectormodels, the minimal width is:
GVmin =
g2c
Mmed
12p
1 +2m2
c
M2med
!
bDMq(Mmed � 2mc
) (2.3)
+ Âq
3g2qMmed
12p
1 +2m2
q
M2med
!
bqq(Mmed � 2mq),
GAmin =
g2c
Mmed
12p
b
3DMq(Mmed � 2m
c
) (2.4)
+ Âq
3g2qMmed
12p
b
3qq(Mmed � 2mq) .
q(x) denotes the Heaviside step function, and b f =
r
1 � 4m2f
M2med
is the velocity of the fermion f with mass m f in the mediatorrest frame. Note the color factor 3 in the quark terms. Figure 2.2shows the minimal width as a function of mediator mass for bothvector and axial-vector mediators assuming the coupling choicegq = g
c
= 1. With this choice of the couplings, the dominant con-tribution to the minimal width comes from the quarks, due to thecombined quark number and color factor enhancement. We specif-ically assume that the vector mediator does not couple to leptons.If such a coupling were present, it would have a minor effect in in-creasing the mediator width, but it would also bring in constraintsfrom measurements of the Drell-Yan process that would unneces-sarily restrict the model space.
LHCP2016, 17-Jun-2016 Bo Jayatilaka
General strategy
4
t
V
g
ui
t
χ̄
χ
1. Select events
Large ETmiss
“Tag” with recoilingobject
2. Constrain backgrounds
Even
ts /
GeV
-310
-210
-110
1
10
210 Data
SM backgrounds (pre-fit)
SM backgrounds (post-fit)
QCD
VV
ll→Z
Single top
tt
νl→W
Single muon W CR (13 TeV)-12.3 fb
CMS Preliminary
Recoil [GeV]300 400 500 600 700 800 900 1000
Dat
a / P
red.
00.5
11.5
2pre-fit post-fit
Even
ts /
GeV
-310
-210
-110
1
10
210 Data
SM backgrounds (pre-fit)
SM backgrounds (post-fit)
ll→Z
QCD
VV
Single top
tt
CRtSingle muon t (13 TeV)-12.3 fb
CMS Preliminary
Recoil [GeV]300 400 500 600 700 800 900 1000
Dat
a / P
red.
00.5
11.5
2pre-fit post-fit
3. Search high ETmiss
4. Find excess5. ????6. Profit!
OR 4.
LHCP2016, 17-Jun-2016 Bo Jayatilaka
CMS searches for dark matter• Focus will be on 13 TeV (2015 dataset) preliminary results
• Other CMS dark matter results
5
X Dataset CMS Doc AvailableJets,W/Z(qq) 13 TeV, 2.3/fb EXO-16-013 April 2016
γ 13 TeV, 2.3/fb EXO-16-014 June 2016bb 13 TeV, 2.2/fb B2G-15-007 March 2016
t (hadronic) 13 TeV, 2.3/fb EXO-16-017 June 2016
X Dataset CMS Doc PublishedDijet 8 TeV, 20/fb EXO-14-004 Submitted to JHEPZ(ll) 8 TeV, 20/fb EXO-12-054 PRD 93, 052011 (2016)
tt(l+jets) 8 TeV, 20/fb B2G-14-004 JHEP 06 (2015) 121t (hadronic) 8 TeV, 20/fb EXO-12-022 PRL 114, 101801 (2015)
γ 8 TeV, 20/fb EXO-12-047 PLB 755 (2016) 102W(lν) 8 TeV, 20/fb EXO-12-060 PRD 91, 092005 (2015)
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Mono-jet/jets/W/Z• Search for large ETmiss and ≥1 high-pT jets• Encompasses both monojet and mono-W/Z (decaying hadronically)• After basic ID, separate into two categories for each
• Mono-V: large-radius jets for highly boosted W/Z: [ETmiss,pTj1]> 250 GeV• Mono-jet: remaining events with ETmiss > 200 GeV, pTj1> 100 GeV
• Backgrounds dominated by Z(νν)+jets and W(lν)+jets• Veto events with isolated leptons or b-tags
• Fit background and signal predictions to ETmiss in data
6Monojet MonoV
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Mono-jet/jets/W/Z: results• No excess observed: set limits on cross section in simplified model
• Fix gq = gDM = 1 (direct comparison with 8 TeV) and scan in Mmed and mχ
• Mmed>1.3 TeV @90%CL
7
Vector mediator Axial-vector mediator
EXO-16-013
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Mono-jet/jets/W/Z: interpretation• Results can be recast in terms of nucleon-DM scattering cross section• Compare to direct detection search limits
8
EXO-16-013
Vector mediator Axial-vector mediator
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Monophoton• Search for large ETmiss and a photon
• Dark matter production with an ISR photon• Large extra dimension models with graviton production and a recoiling
photon• Select events with one fiducial photon with pT > 175 GeV and ETmiss > 170 GeV
• Angular separation between photon and ETmiss
• Primary background is Z(νν)+γ
9
200 400 600 800 1000
Even
ts/G
eV
-310
-210
-110
1
10
(13 TeV)-12.3 fbPreliminary CMS
γ), ttντ, W(γ), Z(ll)νµ+jet, W(γ Beam-haloSpikes MisIDγ →jet
MisIDγ →electron γν l→γWγνν →γZ Bkg. uncertainty
ADD, MD=2TeV, n=5 Data
[GeV]γ
TE200 300 400 500 600 700 800 900 1000
Dat
a/SM
00.5
11.5
2 200 400 600 800 1000
Even
ts/G
eV
-310
-210
-110
1
10
(13 TeV)-12.3 fbPreliminary CMSγ), ttντ, W(γ), Z(ll)νµ+jet, W(γ Beam halo
Spikes MisIDγ →jet MisIDγ →electron γν l→γW
γνν →γZ Bkg. uncertaintyADD, MD=2TeV, n=5 Data
[GeV]TE200 300 400 500 600 700 800 900 1000
Dat
a/SM
0123
Photon ET ETmiss
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Monophoton results• No excess observed: set limits on cross section in simplified model
• Fix gq = 0.25, gDM = 1 and scan in Mmed and mχ
• Mmed > 600 GeV @95%CL
• Place limits on EFT model with contact interaction of type γγχχ• Suppression scale Λ > 542 GeV @95%CL
• Exclude ADD Graviton production up to 2.35 TeV for n=6
10
EXO-16-014 New
Vector mediator Axial-vector mediator ADD Graviton
THσ/
95% CL
σO
bserved
-110
1
10
210
THσ/
95% CL
σO
bserved
-110
1
10
210
-1 = 13 TeV, 2.3 fbsPreliminary CMS
= 1DM
g = 0.25, qgVector, Dirac, Observed 95% CLTheoretical uncertaintyMedian expected 95% CL
σ1±Expected
[GeV]medM100 200 300 400 500 600 700 800
[GeV
]DM
m
50
100
150
200
250
300
350
400TH
σ/95%
CLσ
Observed
-110
1
10
210
THσ/
95% CL
σO
bserved
-110
1
10
210
-1 = 13 TeV, 2.3 fbsPreliminary CMS
= 1DM
g = 0.25, qgAxial vector, Dirac, Observed 95% CLTheoretical uncertaintyMedian expected 95% CL
σ1±Expected
[GeV]medM100 200 300 400 500 600 700 800
[GeV
]DM
m
50
100
150
200
250
300
350
400
[TeV]DM1 1.5 2 2.5 3
Cro
ss s
ectio
n [fb
]
1
10
210
310
n=3, Theory LO 95% CL Obs. Limit
95% CL Exp. Limit σ1±Exp. Limit
σ2±Exp. Limit
-1 = 13 TeV, 2.3 fbsPreliminary CMS
LHCP2016, 17-Jun-2016 Bo Jayatilaka
DM+b/bb/tt• Search for DM recoiling against jets with b quarks
• Sensitive to tt+DM production as well• 1-tag: ETmiss> 200 GeV, tagged jet pT > 50 GeV, up to 1 additional jet• 2-tag: ETmiss> 200 GeV, two tagged jets pT > 50 GeV, up to 1 additional jet
• Recovers efficiency of tt+DM• In both cases, veto events with isolated leptons
11
LHCP2016, 17-Jun-2016 Bo Jayatilaka
DM+b/bb/tt results• No excess seen: set limits combining both tag categories• Interpret as a function of scalar and pseudoscalar mediator masses
• Fix mχ=1 GeV
12
B2G-15-007
Scalar mediator Pseudoscalar mediator
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Monotop• Search for single top quark and large ETmiss
• Sensitive to models producing DM+top via FCNC• Analogous to monojet but with FCNC
• Select events with ETmiss>250 GeV• Top decay products in boosted events merged into “fat jet”, pT > 250 GeV• b-tagged subjet, Jet mass: [110, 210] GeV, and τ3/τ2 to ID top quarks• Veto leptons and additional b-tags
• Dominant backgrounds are V+jets and ttbar
13
t
V
g
ui
t
χ̄
χ
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Monotop results• No excess seen: set limits on FCNC model
• Exclude mediator masses up to 1.1 TeV @95%CL • Evaluated for aFC=bFC=0.1 (roughly gSM=0.1, gDM=1), mχ=100 GeV
• Extends previous monotop limits
14
EXO-16-017 New
LHCP2016, 17-Jun-2016 Bo Jayatilaka
CMS dark matter search summary
15
Maximal excluded mass [GeV]1 10 210 310 410
=1
q=g
DMPseudoscalar; g
t/tbDM + b
=1q
=gDM
Scalar; g t/tbDM + b
=1
q=g
DMPseudoscalar; g
)qDM + jets/V(q
=1q
=gDM
Scalar; g )qDM + jets/V(q
=0.1FC=b
FCFC Vector; a
DM + t
=0.25q
=1, gDM
Axial vector; g γDM +
=0.25
q=1, g
DMVector; g
γDM +
=1q
=gDM
Axial vector; g )qDM + jets/V(q
=1
q=g
DMVector; g
)qDM + jets/V(q
Mediator exclusion DM exclusion
-113TeV, 2.3fbEXO-16-01390%CL
-113TeV, 2.3fbEXO-16-01390%CL
-113TeV, 2.3fbEXO-16-01495%CL
-113TeV, 2.3fbEXO-16-01495%CL
-113TeV, 2.3fbEXO-16-01795%CL
-18TeV, 19.7fbEXO-12-05590%CL
-18TeV, 19.7fbEXO-12-05590%CL
-113TeV, 2.17fbB2G-15-00795%CL
-113TeV, 2.17fbB2G-15-00795%CL
CMS Preliminary Dark Matter Summary* - June 2016
*Observed limitsTheory uncertainties not included
spin 0 mediator
spin 1 mediator
= 300σ/σ
= 50σ/σ
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Conclusions• CMS dark matter searches aim to leave no (collider-based) stone unturned• Standardized interpretation scheme and models
• Allows for straightforward comparison of results between experiments• Interpretations that compare to direct searches
• First set of Run 2 results now available • No signs of excess yet• Expect further results (including with 2016 data) in the coming months
16
Backup
LHCP2016, 17-Jun-2016 Bo Jayatilaka
Monophoton: interpretation
18
= 1DM
g = 0.25, qgVector, Dirac, Observed 90% CLMedian expected 90% CLLUX
-1 = 13 TeV, 2.3 fbsPreliminary CMS
[GeV]DMm1 10 210
]2 (D
M-n
ucle
on) [
cmSI
σ
-4610
-4510
-4410
-4310
-4210
-4110
-4010
-3910
-3810
-3710
-3610
-3510
-3410
-3310
= 1DM
g = 0.25, qgAxial vector, Dirac, Observed 90% CLMedian expected 90% CLPICO-60PICO-2L
-1 = 13 TeV, 2.3 fbsPreliminary CMS
[GeV]DMm1 10 210
]2 (D
M-n
ucle
on) [
cmSD
σ-4210
-4110
-4010
-3910
-3810
-3710
-3610
Top Related