Available on the CERN CDS information server CMS PAS EXO-11-071

CMS Physics Analysis Summary

Contact: cms-pag-conveners-exotica@cern.ch 2011/07/21

Search for Black Holes in pp Collisions at√

s = 7 TeV with1 fb−1 Data Set

The CMS Collaboration

Abstract

A search for microscopic black hole production in pp collisions at a center-of-massenergy of 7 TeV by the CMS experiment at the LHC is presented using a 2011 datasample corresponding to an integrated luminosity of approximately 1.09 fb−1. Thiscorresponds to a thirty-fold increase in statistics compared to a previous search basedon 2010 data. Events with large total transverse energy have been analyzed for thepresence of multiple energetic jets, leptons, and photons, typical of a signal froman evaporating black hole. A good agreement with the expected standard modelbackgrounds, dominated by QCD multijet production, has been observed for variousmultiplicities of the final state. Stringent model-independent limits on new physicsproduction in high-multiplicity energetic final states have been set, along with model-specific limits on semi-classical black hole masses in the 4 – 5 TeV range for a varietyof model parameters. This extends substantially the sensitivity of the 2010 analysis.The first limits on production of string balls at hadron colliders with minimum massof 4.1 – 4.5 TeV have been set.

1

One of the exciting predictions of theoretical models with extra spatial dimensions and low-scale quantum gravity is the possibility of copious production of black holes in particle colli-sions at the LHC [1–3]. Such models are candidates for a theory beyond the standard model(SM). They are mainly motivated by the puzzling large difference between the electroweakscale (∼ 1 TeV) and the Planck scale (∼ 1016 TeV), and attempt to solve this hierarchy prob-lem. In this analysis we focus on black hole production in a model with large, flat, extra spatialdimensions (ADD model) [4, 5].

In this PAS, we present an extension of the published first search for black holes at a hadroncollider [6] carried out by the CMS Collaboration in 2010. Recently, we presented an updatebased on a six-fold increase in statistics already collected in the 2011 LHC data-taking period(190 pb−1) [7]. The present search is based exclusively on the 2011 data sample correspondingto an integrated luminosity1 of 968 ± 58 pb−1. The details of the CMS detector [8], analysismethod, underlying theory, as well as the detailed description of models we probed in thissearch can be found in the original publication [6].

As in previous analysis, we used data collected with a suite of HT triggers, where HT is de-fined as a scalar sum of the transverse energies (ET) of the jets above a threshold2. There havebeen certain changes introduced in the trigger logic both at Level 1 (L1) and at the High-LevelTrigger (HLT) since 2010. Namely, we now use jets corrected for the calorimeter response tocalculate the HT variable at the HLT (uncorrected jets are still used at L1), and the minimum HTthresholds, as well as the minimum jet ET requirement for a jet to be counted toward HT havebeen increased to minimize the effects of the pile-up and match the increased instantaneous lu-minosity of the machine. This minimum jet ET threshold was 10 GeV at L1, and 40 GeV at theHLT. The minimum HT threshold at the HLT was varied between 350 and 550 GeV, dependingon the running period. Only jets found at central pseudorapidities3 |η| < 3.0 are used towardthe HT calculations at L1, and at the HLT for the majority of the data taking period.4 The triggerhas been shown to be fully efficient for the minimum offline HT requirement of 700 GeV, andhence it is fully efficient for the offline cuts used in the analysis.

Offline, jets are reconstructed using energy deposits in the HCAL and electromagnetic calori-meter (ECAL), which are clustered using an infrared-safe anti-kT algorithm with the distanceparameter of 0.5 [9]. The jet energy resolution is ∆E/E ≈ (100 %/

√E) ⊕ 5 %. Jets are re-

quired to pass standard quality requirements to remove calorimeter noise [10]. Jet energies arecorrected for the non-uniformity and non-linearity of the calorimeter response derived usingMonte Carlo (MC) samples and collision data [11]. We require jets to have transverse energyabove 20 GeV before jet energy scale corrections and to have |η| < 2.6. Missing transverseenergy (E/T) is reconstructed as a negative vector sum of transverse energies in the individualcalorimeter towers and is further corrected for the jet energy scale and muons in the event,which leave little energy deposit in the calorimeter [12]. In rare cases when a reconstructedmuon in the event does not pass selection criteria, no muon corrections are applied to E/T.

Electrons and photons are reconstructed as isolated energy deposits in the ECAL with a shapeconsistent with that expected for the electromagnetic showers. The ECAL energy resolution isbetter than 0.5 % above 100 GeV. Photons are required to have no matching hits in the innerpixel detector layers, while electrons are required to have a matching track. Electrons and

1We assume a 6% uncertainty on the integrated luminosity.2Energetic electrons and photons are also reconstructed as jets at the trigger level.3Pseudorapidity η is define as − ln[tan(θ/2)], where θ is the polar angle measured from the nominal detector

center with respect to the direction of the counterclockwise proton beam.4Small fraction of 2011 data did not have central jet requirement enabled at the HLT resulting in jets in the full

CMS hadronic calorimeter (HCAL) rapidity range up to |η| of 5.0 were used for HT calculations.

2

photons are required to have ET > 20 GeV and to be reconstructed in the fully instrumentedbarrel (|η| < 1.44) or end-cap (1.56 < |η| < 2.4) region.

Muons are required to have a match between the stand-alone muon spectrometer and the cen-tral tracker, to be within |η| < 2.1, be consistent with the interaction vertex to suppress back-grounds from cosmic muons, and have pT > 20 GeV. Matching the muons to the tracksmeasured in the silicon tracker results in a transverse momentum resolution between 1% and5% for pT values up to 1 TeV. The minimum separation between any two objects (jet, lepton, orphoton) is required to be ∆R =

√∆φ2 + ∆η2 < 0.5.

We present our results in terms of model-independent limits on the cross section times thebranching ratio. Further, we interpret the results in term of a set of benchmark black hole mod-els. Simulated samples of black hole signal events are generated using a parton-level BLACK-MAX [13] generator (v2.01), followed by the parton showering simulation with PYTHIA [14](v6.420), and a fast parametric simulation of the CMS detector [15, 16]. The parameters used insimulation are listed in Table 1 for a few characteristic model points. The MSTW2008lo68 [17]parton distribution functions (PDF) were used. We also produced a number of string ballssamples using BLACKMAX generator. The string balls [18] are hypothesized precursors of thesemiclassical black holes in highly quantum regime, when the mass of the object is close tothe Planck scale. The string balls are described by the string scale MS and string coupling gS.In cases when the semiclassical approximation (that is valid for black holes with minimummasses some 3–5 times larger than the multidimensional Planck scale) no longer holds, stringballs may offer a more realistic description of black hole formation and decay.

We employ a selection based on total transverse energy to separate black hole events from thebackgrounds. We define the ST variable as a scalar sum of the ET of individual jets, electrons,photons, and muons passing the above selections. Only photons, leptons (e and µ) and jetswith ET > 50 GeV enter in the sum for the calculation of ST. The choice of this rather highminimum transverse energy requirement is driven by the fact that it suppresses most of theSM processes and has been shown to be insensitive to the jets from pile-up, while fully efficientfor black hole decays. We further add E/T in the event to the ST, if the E/T value exceeds 50 GeV.Generalization of the ST definition to include E/T is important for testing black hole models withsignificant amount of missing energy, e.g. the models with a stable non-interacting remnant.We checked that in case jets are mismeasured and as a result a non-genuine E/T is generatedin the event, the effect of the “double-counting” of this mismeasurement in both jet and E/Tcontributions to the ST is negligible.

The main background to black hole production arises from multijet and direct photon produc-tion, which dominates the event rates at large ST. Smaller backgrounds come from W/Z+jetsand tt production. These backgrounds are negligible at large values of ST and contributeless than 1% to the total background after the final selection. We estimate their contributionfrom MC simulation, using MADGRAPH [19] leading-order parton-level event generator withCTEQ6L PDF set [20] followed by PYTHIA [14] parton showering and full CMS detector simula-tion via GEANT4 [21]. For the dominant QCD background, however, we estimate backgroundsfrom data using the ST multiplicity invariance technique of Ref. [6]. This technique is based onthe independence of the shape of the ST spectrum on the number of the final state objects N;an empirical observation extensively checked and proved correct by using various MC sam-ples and low-multiplicity data. The origin of this invariance lies in the collinear nature of theFSR/ISR, which typically does not change the total transverse energy in the event; hence inde-pendence of the ST spectrum for the QCD background on the jet multiplicity. This invarianceallows us to predict the shape of the ST spectrum for any number of objects using the dijet data,

3

which has been extensively studied for presence of new physics in dedicated analyses [22–24].

(GeV)TS1000 1500 2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410

510

610 N = 2DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

Photon+JetsW+JetsttbarZ+Jets

CMS Preliminary-1 = 7 TeV, 1.09 fbs

a)

(GeV)TS1000 1500 2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410

510

610 N = 3

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

Photon+JetsW+JetsttbarZ+Jets

CMS Preliminary-1 = 7 TeV, 1.09 fbs

b)

Figure 1: Total transverse energy, ST, for events with a) N = 2, and b) N = 3 photons, electrons,muons, or jets in the final state. Data are depicted as points with error bars; shaded band is thedata-driven background prediction (solid line) with its uncertainty. Non-QCD backgroundsare shown as colored histograms. Also shown is black hole signal for three different parametersets, demonstrating small signal contamination.

We fit the ST distributions in data with N = 2 and N = 3 between 800 and 2500 GeV (where noevidence for new physics has been observed in a dedicated analysis [22]) with the ansatz func-tion P0(1+x)P1

xP2+P3 log(x) , which is shown with the solid line in Fig. 1. To check the systematic uncertainty

of the fit, we use two additional ansatz functions, P0(P1+P2x+x2)P3

and P0(P1+x)P2

, which are shown asthe upper and lower boundaries of the shaded curve in Fig. 1. The default choice of the ansatzfunction is based on the best fit χ2 to the ST distribution for N = 2. Additional systematicuncertainty arises from a slight difference between the best fit shapes for N = 2 and N = 3.Nevertheless, the fits for these two exclusive multiplicities are very consistent and agree witheach other within the uncertainties, demonstrating independence of the ST shape on the finalstate multiplicity.

We then look at the inclusive samples with high multiplicity using the background shape fromthe low multiplicity distributions, as shown in Fig. 2, normalized to the inclusive data in therange of 1600 – 2000 GeV, where no signal contribution is expected. The data agree well withthe background shapes from the low-multiplicity samples and do not exhibit any evidence fornew physics. An event display of one of a black-hole candidate event (N = 10, ST = 1.1 TeV) isshown in Fig. 3. The relatively high minimum requirement on jet transverse energy of 50 GeVto be counted toward N and ST essentially eliminates jets from pile-up even at moderate STbelow the signal region, which is evident from the zoom on the vertex region in this figure,which demonstrates that all 10 jets originate from the common, primary vertex.

Since no excess is observed above the predicted background, we proceed with setting limits onblack hole production. We assign systematic uncertainty on the background estimate varyingfrom 7% to 165% in the ST range used in this search. This uncertainty comes from the normal-ization uncertainty (2% – 15%) added in quadrature to the uncertainties from using various fitansatz functions and the difference between the shapes obtained from the N = 2 and N = 3samples. The integrated luminosity is measured in situ using forward calorimeters with 6% un-

4

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 3≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

a)

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 4≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

b)

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 5≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

c)

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 6≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

d)

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 7≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

e)

(GeV)TS2000 2500 3000 3500 4000

Eve

nts

/ 100

GeV

-110

1

10

210

310

410 8≥N

DataBackground

Uncertainty = 4.5 TeV, n = 6 min

BH = 1.5 TeV, MDM

= 4.0 TeV, n = 4 minBH

= 2.0 TeV, MDM

= 3.5 TeV, n = 2 minBH

= 2.5 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

f)

Figure 2: Total transverse energy, ST, for events with: a) N ≥ 3, b) N ≥ 4, c) N ≥ 5, d) N ≥ 6,e) N ≥ 7, and f) N ≥ 8 photons, electrons, muons, or jets in the final state. in the final state.Data are depicted as points with error bars; shaded band is the background prediction (solidline) with its uncertainty. Also shown are black hole signals for three different parameter sets.

5

Figure 3: An event display of a N = 10 black hole candidate with ST = 1.1 TeV (Run 163332,Event 196371106). Top: the transverse view of the event with 10 objects (jets) highlighted withmagenta cones. Bottom: the zoom on the vertex region in the view parallel to the beam-line.All the jets clearly come from the same, primary vertex (red dot), despite a number of pile-upvertices (blue dots). The nominal beam-spot position is shown with an orange dot.

6

Table 1: Monte Carlo signal points for some of the model parameters probed, correspond-ing leading order cross sections, and optimal selections on the minimum decay multiplicity(N ≥ Nmin) and minimum ST, as well as signal acceptance, expected number of signal events,number of observed events in data, expected background, and observed and expected limit onthe signal at 95% confidence level.

MD MBH n σ Nmin SminT A Nsig Ndata Nbkg σ95 〈σ95〉

(TeV) (TeV) (pb) (TeV) (%) (pb) (pb)1.5 3.0 6 25.9 3 1.9 88.0 24920 1404 1458± 222 0.341 0.3341.5 3.5 6 4.97 3 2.2 88.0 4780 372 407± 135 0.162 0.1681.5 4.0 6 0.767 4 2.5 85.8 718 58 75.1± 44.7 0.039 0.0471.5 4.5 6 0.091 5 2.9 78.9 78.2 3 8.24+9.21

−8.24 0.006 0.0101.5 5.0 6 7.1× 10−3 6 3.3 70.4 5.48 0 0.86+1.65

−0.86 0.004 0.0041.5 5.5 6 2.9× 10−4 7 3.8 58.7 0.18 0 0.06+0.21

−0.06 0.005 0.0032.0 3.0 4 6.45 3 2.0 82.5 5805 868 942± 191 0.271 0.2702.0 3.5 4 1.26 3 2.4 79.7 1093 158 183± 91 0.099 0.0982.0 4.0 4 0.199 3 2.8 77.1 167.2 18 40.4± 39.0 0.018 0.0322.0 4.5 4 0.0236 4 3.3 65.3 16.80 0 4.32+8.25

−4.32 0.004 0.0072.0 5.0 4 1.8× 10−3 5 3.7 56.9 1.12 0 0.57+1.76

−0.57 0.005 0.0042.5 3.0 2 0.967 3 2.4 62.3 657.4 158 182.5± 90.6 0.128 0.0982.5 3.5 2 0.192 3 2.7 65.3 136.7 39 58.3± 48.3 0.040 0.0422.5 4.0 2 0.0314 3 3.2 57.7 19.77 4 9.96+16.74

−9.96 0.011 0.0122.5 4.5 2 3.9× 10−3 4 3.6 46.7 1.98 0 1.62+4.47

−1.62 0.006 0.0052.5 5.0 2 3.1× 10−4 5 4.1 31.4 0.11 0 0.17+0.82

−0.17 0.009 0.003

Figure 4: A 3D event display of a N = 9 black hole candidate with ST = 2.5 TeV (Run 165567,Event 347495624).

7

certainty [25]. Signal uncertainty is dominated by the jet energy scale uncertainty of ≈ 5% [11]which translates into 5% uncertainty on the signal. An additional 2% uncertainty on the signalacceptance comes from the variation of PDFs within the CTEQ6 error set [20]. Note that particleidentification efficiency does not affect the signal distribution, since if an electron does not passthe identification requirements it is either classified as a photon, or as a jet; a photon failing theselection will become a jet; and a rejected muon will contribute to E/T. Thus, the total ST in theevent is virtually not affected.

Given significant model-dependence of the black hole production cross section and decay pat-terns, it is virtually impossible to test all different variations of model parameters offered bymodern black hole MC generators in a dedicated search. This study considers over 160 differ-ent signal MC samples already, and yet it does not come close to spanning the entire parameterspace of the model; scaling the number of signal samples up and presenting the results for ev-ery model therefore becomes impractical. Hence, we choose to first present the results of oursearch in a generic, model-independent way, which would allow to probe additional modelsessentially using parton-level MC information (possibly augmented with very crude detectorsimulation). To facilitate that, we produce model-independent limits on the cross section timesthe acceptance for new physics production in high-ST inclusive final states for N ≥ 3, 4, 5,6, 7, and 8. Figure 5 shows 95% confidence level (CL) limits from a counting experiment forST > Smin

T as a function of SminT , which can be used to test models of new physics that result in

these final state, including (but not limited to) even broader variety of black hole models thatwe covered in this analysis. The 95% CL limits from 2010 data [6] are also shown in Fig. 5. Thepresent model-independent limits are roughly 3 fb at high values of ST, which is thirty timesmore sensitive than the limits reported in the previous publication – in good agreement withexpectations given roughly thirty-fold increase in integrated luminosity. Given higher statisticsof the 2011 sample, we are also able to extend these limits to the N ≥ 8 case, which was notpossible before.

For specific subset of the black hole models [6] that are being probed, we also set dedicatedlimits on black hole and string balls production with the optimized ST and N selections usingcounting experiments. We would like to emphasize that the semi-classical approximation usedfor deriving the cross section within these benchmark scenarios is expected to break downfor most of the points probed. Thus, these limits should be treated as indicative, rather thanprecise. Apparently, this point has not been emphasized enough in our earlier publication [6]and some authors considered this to be a weakness of our analysis [26]. Given that there isno alternative quantitative calculations in the regime where the semiclassical approximationbreaks down and also exponential change of the production cross section with the black holemass, which results in rather significant changes in the model to translate to rather moderatechanges in the mass limit, we still choose to show these limits with the above caveat. The mainresult of our analysis remains to be the model-independent limits, which the author of Ref. [26]has unfortunately missed.

We optimized the signal (S) significance in the presence of background (B), S/√

S + B, for eachset. The optimum choices of parameters for a few benchmark scenarios are listed in Table 1, aswell as the predicted number of background events, expected number of signal events, and theobserved number of events in data. We set limits using a Bayesian method with a flat signalprior and a log-normal prior for integration over the nuisance parameters. These limits at the95% confidence level (CL) are shown in Fig. 6.

Translating the cross section limits into the expectations of the ADD model, we can exclude theproduction of black holes with a minimum mass varying from 4.0 to 5.1 TeV for values of the

8

(GeV) minTS

1000 2000 3000 4000 5000

A (

pb)

×) m

inT

> S

T(Sσ

-210

-110

1

10

-12011, 1.09 fb

-12010, 35 pb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

3≥N 95σObserved

exp.95σExpected

a)

(GeV) minTS

1000 2000 3000 4000 5000 A

(pb

)×)

min

T >

ST

(Sσ

-210

-110

1

10

-12011, 1.09 fb

-12010, 35 pb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

4≥N 95σObserved

exp.95σExpected

b)

(GeV) minTS

1000 2000 3000 4000 5000

A (

pb)

×) m

inT

> S

T(Sσ

-210

-110

1

10

-12011, 1.09 fb

-12010, 35 pb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

5≥N 95σObserved

exp.95σExpected

c)

(GeV) minTS

2000 3000 4000 5000

A (

pb)

×) m

inT

> S

T(Sσ

-210

-110

1

-12011, 1.09 fb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

6≥N 95σObserved

exp.95σExpected

d)

(GeV) minTS

2000 3000 4000 5000

A (

pb)

×) m

inT

> S

T(Sσ

-210

-110

1

-12011, 1.09 fb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

7≥N 95σObserved

exp.95σExpected

e)

(GeV) minTS

2000 3000 4000 5000

A (

pb)

×) m

inT

> S

T(Sσ

-210

-110

1

-12011, 1.09 fb

CMS Preliminary-1 = 7 TeV, 1.09 fbs

8≥N 95σObserved

exp.95σExpected

f)

Figure 5: Model-independent 95% confidence level upper cross section limits for counting ex-periments with ST > Smin

T as a function of SminT for a) N ≥ 3, b) N ≥ 4, c) N ≥ 5, d) N ≥ 6, e)

N ≥ 7, and f) N ≥ 8. The blue solid (red dotted) lines correspond to an observed (expected)limit for nominal signal acceptance uncertainty of 5%, compared to observed (expected) limitsobtained on 2010 CMS data and shown as blue dashed (red dash-dotted) line.

9

(TeV) minBHM

2.5 3 3.5 4 4.5 5 5.5

(pb

)σ

-310

-210

-110

1

10

210

CMS Preliminary-1 = 7 TeV, 1.09 fbs

Non-rotating Black Holes = 2.5 TeV, n = 2DM = 2.0 TeV, n = 4DM = 1.5 TeV, n = 6DM

Theoretical Cross Section = 2.5 TeV, n = 2DM = 2.0 TeV, n = 4DM = 1.5 TeV, n = 6DM

Figure 6: Cross section limits at 95% confidence level from the counting experiments opti-mized for various black hole parameter sets compared with signal production cross sectionfrom BLACKMAX generator. Colored solid lines show experimental cross section limits, whiledotted lines represent corresponding signal cross sections. The corresponding expected limitsare 4.4, 4.8, and 5.1 TeV.

10

(TeV)DM1.5 2 2.5 3 3.5

(Te

V)

min

BH

Exc

lude

d M

4.2

4.4

4.6

4.8

5

BlackMaxNon-RotatingRotating

CMS Preliminary-1 = 7 TeV, 1.09 fbs

n = 2

n = 4

n = 6

Figure 7: Limits on the black hole model parameters for several benchmark scenarios. Theareas below each curve are excluded by this search.

(TeV) minM3.5 4 4.5 5 5.5

(pb

)σ

-310

-210

-110

1String Ball (BlackMax)

= 0.4s

= 1.3 TeV, gs

= 1.6 TeV, MDM = 0.4

s = 1.0 TeV, g

s = 1.3 TeV, MDM

= 0.4s

= 1.7 TeV, gs

= 2.1 TeV, MDM

Theoretical Cross Section = 0.4

s = 1.3 TeV, g

s = 1.6 TeV, MDM

= 0.4s

= 1.0 TeV, gs

= 1.3 TeV, MDM = 0.4

s = 1.7 TeV, g

s = 2.1 TeV, MDM

CMS Preliminary-1 = 7 TeV, 1.09 fbs

Figure 8: Cross section limits at 95% confidence level from the counting experiments optimizedfor various string ball parameter sets compared with signal production cross section. Coloredsolid lines show experimental cross section limits, while dotted lines represent correspondingsignal cross sections.

11

multidimensional Planck scale up to 3.5 TeV at 95% CL. These limits, shown in Fig. 7, do notexhibit a significant dependence on the details of the production and evaporation model.

The 95% CL limits on the string balls production cross section as function of string balls min-imum mass are shown in Fig. 8 for a number of characteristic model parameters. We excludethe string balls production with a minimum mass from 4.1 to 4.5 TeV depending on the model.These are the first direct limits on string balls set at hadron colliders.

To conclude, we presented an update of an earlier published first dedicated search for blackholes at a hadron collider [6] and set stringent model-independent limits on the production ofenergetic multi-particle final states, which can be used to constrain a large variety of modelsof new physics. In addition we set model-specific stringent limits on the minimum black holeand string balls mass in the 4− 5 TeV range for a large variety of model parameters.

References[1] S. Dimopoulos and G. Landsberg, “Black Holes at the LHC”, Phys. Rev. Lett. 87 (2001)

161602, arXiv:hep-ph/0106295. doi:10.1103/PhysRevLett.87.161602.

[2] S. Giddings and S. Thomas, “High-energy colliders as black hole factories: The end ofshort distance physics”, Phys. Rev. D 65 (2002) 056010, arXiv:hep-ph/0106219.doi:10.1103/PhysRevD.65.056010.

[3] P. Meade and L. Randall, “Black Holes and Quantum Gravity at the LHC”, JHEP 05(2008) 003, arXiv:hep-ph/0708.3017.doi:10.1088/1126-6708/2008/05/003.

[4] N. Arkani-Hamed, S. Dimopoulos, and G. Dvali, “The Hierarchy problem and newdimensions at a millimeter”, Phys. Lett. B 429 (1998) 263–267.doi:10.1016/S0370-2693(98)00466-3.

[5] N. Arkani-Hamed, S. Dimopoulos, and G. Dvali, “Phenomenology, astrophysics andcosmology of theories with submillimeter dimensions and TeV scale quantum gravity”,Phys. Rev. D 59 (1999) 086004. doi:10.1103/PhysRevD.59.086004.

[6] CMS Collaboration, “Search for Microscopic Black Hole Signatures at the Large HadronCollider”, Phys. Lett. B697 (2011) 434–453, arXiv:1012.3375.doi:10.1016/j.physletb.2011.02.032.

[7] CMS Collaboration, “Search for Black Holes in pp Collisions at√

s = 7 TeV”, CMS PASEXO-11-021.

[8] CMS Collaboration, “The CMS experiment at the CERN LHC”, JINST 3 (2008) S08004.doi:10.1088/1748-0221/3/08/S08004.

[9] M. Cacciari, G. Salam, and G. Soyez, “The anti-kT jet clustering algorithm”, Journal ofHigh Energy Physics 2008 (2008), no. 04, 063.doi:10.1088/1126-6708/2008/04/063.

[10] CMS Collaboration, “Jet Performance in pp Collisions at 7 TeV”, CMS PAS JME-10-003.

[11] CMS Collaboration, “Jet Energy Corrections determination at√

s = 7 TeV”, CMS PASJME-10-010.

12

[12] CMS Collaboration, “Missing Transverse Energy Performance in Minimum-Bias and JetEvents from Proton-Proton Collisions at

√s = 7 TeV”, CMS PAS JME-10-004.

[13] D.-C. Dai, G. Starkman, D. Stojkovic et al., “BlackMax: A black-hole event generator withrotation, recoil, split branes, and brane tension”, Phys. Rev. D 77 (2008), no. 7, 076007.doi:10.1103/PhysRevD.77.076007.

[14] T. Sjostrand, S. Mrenna, and P. Skands, “PYTHIA 6.4 Physics and Manual”, JHEP 05(2006) 026. doi:10.1088/1126-6708/2006/05/026.

[15] D. Orbaker, “Fast Simulation of the CMS Detector”, CERN-CMS-CR-2009-074.

[16] S. Abdullin, P. Azzi, F. Beaudette et al., “Fast Simulation of the CMS Detector at theLHC”, CERN-CMS-CR-2010-297.

[17] MSTW Collaboration, “Heavy-quark mass dependence in global PDF analyses and 3-and 4-flavour parton distributions”, arXiv:1007.2624.

[18] S. Dimopoulos and R. Emparan, “String Balls at the LHC and Beyond”, Phys. Lett. B 526(2002) 393.

[19] J. Alwall et al., “MadGraph/MadEvent v4: The New Web Generation”, JHEP 09 (2007)028. doi:10.1088/1126-6708/2007/09/028.

[20] CTEQ Collaboration, “Implications of CTEQ global analysis for collider observables”,Phys. Rev. D 78 (2008) 013004.

[21] GEANT 4 Collaboration, “GEANT4 – a simulation toolkit”, Nucl. Instr. and Methods A506 (2003) 250–303. doi:10.1016/S0168-9002(03)01368-8.

[22] CMS Collaboration, “Search for Dijet Resonances in 7 TeV pp Collisions at CMS”, Phys.Rev. Lett. 105 (2010) 211801, arXiv:1010.0203.doi:10.1103/PhysRevLett.105.21180.

[23] CMS Collaboration, “Search for Quark Compositeness with the Dijet Centrality Ratio inpp Collisions at

√s = 7 TeV”, Phys. Rev. Lett. 105 (2010) 262001, arXiv:1010.4439.

doi:10.1103/PhysRevLett.105.262001.

[24] CMS Collaboration, “Measurement of Dijet Angular Distributions and Search for QuarkCompositeness in pp Collisions at 7 TeV”, Phys. Rev. Lett. 106 (2010) 201804,arXiv:1102.2020. doi:10.1103/PhysRevLett.106.201804.

[25] CMS Collaboration, “Measurement of CMS Luminosity”, (2011). CMS-DP-2011-002.Document in preparation.

[26] S. Park, “Critical comment on the recent microscopic black hole search at the LHC”,Phys. Lett. B701 (2011) 587–590, arXiv:1104.5129.doi:10.1016/j.physletb.2011.06.039.