1: \documentclass[aps,prl,showpacs,twocolumn,lineno,groupedaddress]{revtex4} % for review
2: %\documentclass[aps,prl,showpacs,twocolumn,groupedaddress]{revtex4} % for submission
3: \usepackage{graphicx} % needed for figures
4: \usepackage{dcolumn} % needed for some tables
5: \usepackage{bm} % for math
6: \usepackage{amssymb} % for math
7:
8: \def\MET{{\mbox{$E\kern-0.57em\raise0.19ex\hbox{/}_{T}$}}}
9: \def\DZero{D\O\ }
10:
11:
12: \begin{document}
13:
14: % the following information is for internal review, please remove them for submission
15: %\leftline{Version 4.0 as of \today}
16: %\leftline{Primary authors: Yuri Gershtein and Stelios Kesisoglou}
17: %\rightline{Comment to {\tt d0-run2eb-013@fnal.gov}}
18: %\rightline{by August 13, 2004}
19:
20: \title{Search for Supersymmetry with Gauge-Mediated Breaking in Diphoton Events
21: at \DZero }
22: \input list_of_authors_r2.tex % input Dzero author list
23: \date{\today}
24:
25: \begin{abstract}
26: We report the results of a search for supersymmetry (SUSY) with
27: gauge-mediated breaking in the missing transverse
28: energy distribution of inclusive diphoton events
29: using 263 pb$^{-1}$ of data collected by the
30: D\O\ experiment at the Fermilab Tevatron Collider in 2002--2004. No excess
31: is observed above the background expected from
32: standard model processes, and lower limits on the masses of the lightest neutralino and
33: chargino of about 108 and 195 GeV, respectively, are set at the
34: 95\% confidence level. These are the most stringent limits to date for
35: models with gauge-mediated SUSY breaking
36: with a short-lived neutralino as the next-lightest SUSY particle.
37: \end{abstract}
38: \pacs{14.80.Ly, 12.60.Jv, 13.85.Rm}
39: \maketitle
40:
41: Models involving gauge-mediated supersymmetry breaking
42: (GMSB), originally proposed in Ref. \cite{gmsb_original}
43: have attracted much attention \cite{gmsbsusy}.
44: In GMSB models supersymmetry breaking is achieved by introduction
45: of new chiral supermultiplets, called messengers,
46: which couple to the ultimate source of supersymmetry breaking, and
47: also to the SUSY particles.
48: The phenomenology of these models is rich
49: and strikingly different from that of gravity-mediated SUSY models.
50:
51: For GMSB models, the gravitino (with a mass less than $\sim ~ $keV) is the
52: lightest SUSY particle (LSP), and the phenomenology of these
53: models is therefore determined by the next-to-lightest SUSY particle
54: (NLSP), which can be either a neutralino or a slepton.
55: In the former case, which is considered in this paper,
56: the NLSP decays into a photon and an LSP, and the signal of interest,
57: assuming $R$-parity conservation \cite{rpar}, is a final
58: state with two photons and large missing transverse energy
59: ($\MET$).
60:
61: The model we consider is a minimal GMSB
62: with a neutralino as the NLSP, referred to as Snowmass Slope SPS 8
63: \cite{modelline}. This model has only one dimensioned parameter
64: $\Lambda$ that determines the effective
65: scale of SUSY breaking. The minimal GMSB parameters correspond to a messenger
66: mass $M_m = 2\Lambda$, the number of messengers $N_5 = 1$, the ratio of
67: the vacuum expectation values of the two Higgs fields $\tan \beta
68: = 15$, and the sign of the Higgsino mass term $\mu > 0$.
69: The lifetime of the neutralino is
70: not fixed by this model line, and is assumed to be sufficiently short to
71: result in decays with prompt photons.
72: Current lower limits on the GMSB neutralino mass for somewhat similar model
73: parameters are 65, 75 and 100 GeV, from the
74: CDF \cite{cdfrun1}, D\O\ \cite{d0run1} and CERN LEP
75: collaborations \cite{lepsusy}, respectively.
76:
77: We search for SUSY production in $p\bar{p}$ collisions at
78: $\sqrt{s} = 1.96$ TeV at the Fermilab Tevatron Collider.
79: The D\O\ detector comprises a central tracking system in
80: a 2 T superconducting solenoidal magnet, a liquid-argon/uranium
81: calorimeter, and a muon spectrometer~\cite{run2det}.
82: The tracking system consists of a silicon microstrip tracker and
83: a scintillating fiber tracker and provides coverage for charged particles
84: in the pseudorapidity range $|\eta| < 3$.
85: The calorimeters are finely segmented and consist of a central
86: section (CC) covering
87: $|\eta| \leq 1.1$, and two end calorimeters (EC)
88: extending coverage to $|\eta|\approx 4$, all housed in separate
89: cryostats~\cite{run1det}.
90: Scintillators installed between the CC and EC cryostats provide sampling
91: of developing showers for $1.1<|\eta|<1.4$.
92: The electromagnetic (EM) section of the calorimeter has four longitudinal
93: layers and transverse segmentation of $0.1 \times 0.1$ in $\eta - \phi$
94: space (where $\phi$ is the azimuthal angle), except in the third layer,
95: corresponding to EM shower maximum, where it is $0.05 \times 0.05$.
96: The data sample was collected
97: between April 2002 and March 2004, using inclusive single
98: electromagnetic (EM) and di-EM triggers. The integrated luminosity of
99: the sample is $263 \pm 17 ~ \rm pb^{-1}$.
100:
101: Photons and electrons are identified in two steps: first, selection of
102: the EM clusters, and then their separation into photons or electrons.
103: EM clusters are selected from calorimeter clusters by requiring that (i)
104: at least 90\% of the energy be deposited in the EM section of the
105: calorimeter, (ii) the calorimeter isolation variable ($I$) be
106: less than 0.15, where
107: $I = [E_{tot}(0.4)-E_{EM}(0.2)] / E_{EM}(0.2)$, where
108: $E_{tot}(0.4)$ is the total shower energy in a cone of radius
109: ${\cal R} = \sqrt{(\Delta\eta)^2 + (\Delta\phi)^2} = 0.4$, and
110: $E_{EM}(0.2)$ is the EM energy in a cone ${\cal R}=0.2$,
111: (iii) the transverse and longitudinal shower profiles be
112: consistent with those expected for an EM shower, and (iv)
113: the scalar sum of the $p_T$ of all tracks originating from the primary
114: vertex in an annulus of $ 0.05 < {\cal R} < 0.4$ around the cluster be
115: less than 2 GeV. The cluster is then defined as an electron if there
116: is a reconstructed track pointing to it and a photon otherwise.
117: Jets are reconstructed using the iterative, midpoint cone algorithm \cite{run2cone} with
118: a cone size of 0.5. $\MET$ is determined
119: from the energy deposited in the calorimeter for $|\eta| < 4$ and is
120: corrected for jet and EM energy scales.
121: %Events that exhibit patterns of
122: %calorimeter electronic noise ($\approx$ 3\%
123: %of the total) are removed for this analysis.
124:
125: We select $\gamma\gamma$ candidates by requiring
126: events to have two photons each with $E_T > 20~$GeV and
127: pseudorapidity $|\eta|<1.1$. To suppress events with mismeasured
128: $\MET$, we apply the following requirements.
129: We reject any event when the difference in azimuth ($\Delta \phi$) between the highest
130: $E_T$ jet (if jets are present) and the direction of the $\MET$ is more than 2.5
131: radians, or if the $\Delta \phi$ between the direction of
132: the $\MET$ and either photon is less than 0.5 radians.
133: These selections yield 1,909 events ($\gamma\gamma$ sample),
134: out of which 1,800 have $\MET <
135: 15~$GeV and two have $\MET > 40~$GeV. The two events
136: constitute the $\gamma\gamma\MET$ sample.
137:
138: \begin{figure}
139: \includegraphics[scale=0.4]{gg_met_mc.eps}
140: \caption{\label{fig_gg} The $\MET$ distribution for the diphoton
141: and background samples. Also shown is the expected distribution for
142: the GMSB point with $\Lambda = 80~ $TeV, multiplied by a factor of ten.}
143: \end{figure}
144:
145:
146: The main backgrounds arise from standard model processes with
147: misidentified photons and/or mismeasured $\MET$.
148: The background from processes with no inherent $\MET$
149: (multijet events, direct photon production, $Z \rightarrow ee$, etc.)
150: is estimated using events with two EM clusters that satisfy
151: photon-identification criteria (i) and (ii), but fail the shower-shape requirement
152: (iii). These events, called the QCD sample, must pass
153: the same trigger and other selections that define the $\gamma\gamma$ sample.
154: They have characteristics similar to the
155: background in the $\gamma\gamma$ sample
156: and in particular are expected to have similar $\MET$ resolution.
157: This assumption was checked by varying the selection criteria
158: and comparing the $\MET$ distribution in the QCD sample
159: to that in $Z \rightarrow ee$ events.
160: The QCD sample comprises 18,437 events, with 17,379 events having
161: $\MET < 15 ~$GeV, and 27 events with $\MET > 40 ~$GeV.
162: We estimate the background in the $\gamma\gamma\MET$ sample
163: resulting from mismeasurement of $\MET$ by
164: normalizing the number of QCD events to that of the $\gamma\gamma$ sample
165: for $\MET < 15~$GeV.
166: This yields $2.8 \pm 0.5$ events with $\MET > 40 ~$GeV,
167: with uncertainty dominated by the statistics of the QCD sample.
168:
169: The other sources of background correspond to events with genuine $\MET$ in which
170: an electron is misidentified as a photon, for example
171: from $W + $\mbox{'$\gamma$'} events
172: (where \mbox{'$\gamma$'} denotes both true photons and jets misidentified as photons),
173: and from $Z \rightarrow \tau^+\tau^- \rightarrow e^+e^- + X$ and $t\bar{t}
174: \rightarrow e^+e^- + ~$jets production. We estimate this contribution
175: using the $e\gamma$ sample which has the same trigger, kinematic,
176: and EM identification requirements as the $\gamma\gamma$
177: sample. This sample contains 889 events, 782 events with $\MET < 15 ~$GeV and 15
178: events with $\MET > 40 ~$GeV. To estimate the contribution of such events
179: to the $\gamma\gamma\MET$ sample, we first subtract the QCD
180: background component of the $e\gamma$ sample. This is done by
181: normalizing the QCD sample to the $e\gamma$ sample for $\MET <
182: 15~$GeV. Then, using the probability for an electron to be misidentified as
183: a photon (measured using $Z \rightarrow ee$ events to be
184: $0.064 \pm 0.004$), we estimate this background to be $0.9 \pm 0.2$
185: events with statistically dominated uncertainty.
186: Therefore the total expected background to the $\gamma\gamma\MET$ sample
187: is $3.7 \pm 0.6$ events. The $\MET$ distributions for
188: the $\gamma\gamma$ sample, background without genuine $\MET$, and
189: the total background
190: are shown in Fig. \ref{fig_gg}, together with an expected
191: distribution from the Snowmass Slope model with $\Lambda = 80~$TeV,
192: the latter multiplied by a factor of ten for clarity.
193:
194: To estimate the expected signal, we generated
195: Monte Carlo (MC) events for several points on the Snowmass Slope (see Table
196: \ref{table_gmsb}), covering the neutralino mass range from $72~$GeV,
197: somewhat below the existing limits \cite{d0run1, lepsusy}, to $116~$GeV.
198: We used {\sc isajet 7.58} \cite{isajet} to
199: determine SUSY interaction eigenstate masses and couplings.
200: {\sc pythia 6.202} \cite{pythia}
201: was used to generate the events after determining the sparticle masses,
202: branching fractions and leading order (LO) production cross sections using
203: the {\sc CTEQ5L} \cite{cteq} parton distribution functions (PDF).
204: MC events were processed through full detector simulation
205: and reconstruction, and processed with the analysis program used for the data.
206:
207: The dominant contributions to the cross section
208: are from production of lightest charginos ($\tilde{\chi}^+_1\tilde{\chi}^-_1$)
209: and chargino-second neutralino pairs ($\tilde{\chi}^0_2\tilde{\chi}^{\pm}_1$).
210: The total cross section in Table \ref{table_gmsb} is calculated
211: to leading order in {\sc pythia} for GMSB SUSY
212: production. The ``$K$-factor'' used to
213: account for higher-order corrections is applied to estimate the
214: next-to-leading-order (NLO) cross section. The values of the
215: $K$-factor in the table are taken from Ref. \cite{kfact}.
216: The sources of error on signal efficiency include uncertainty on photon
217: identification (4\% per photon), MC statistics (5\%), and choice of
218: PDF (5\%).
219:
220: \begin{table*}
221: \caption{ \label{table_gmsb}
222: Points on the Snowmass Slope: their cross sections,
223: efficiencies and cross-section limits.}
224: \begin{ruledtabular}
225: \begin{tabular}{c|cc|cc|c|c}
226: $\Lambda,$ TeV &$m_{\tilde{\chi}_1^0},$GeV & $m_{\tilde{\chi}_1^+},$GeV & $\sigma_{TOT}^{LO}$, pb & $K$-factor & Efficiency & 95\% C.L. Limit, pb\\
227: \hline
228: 55 & 71.8 & 126.3 & 0.735 & 1.236 & $0.092\pm 0.009$ & 0.184 \\
229: 60 & 79.1 & 140.2 & 0.468 & 1.227 & $0.100\pm 0.009$ & 0.170 \\
230: 65 & 86.4 & 154.3 & 0.301 & 1.217 & $0.111\pm 0.011$ & 0.153 \\
231: 70 & 93.7 & 168.2 & 0.204 & 1.207 & $0.124\pm 0.012$ & 0.137 \\
232: 75 & 101.0 & 182.3 & 0.138 & 1.197 & $0.137\pm 0.013$ & 0.124 \\
233: 80 & 108.2 & 196.0 & 0.094 & 1.187 & $0.149\pm 0.014$ & 0.114 \\
234: 85 & 115.5 & 209.9 & 0.066 & 1.177 & $0.154\pm 0.015$ & 0.110 \\
235: \end{tabular}
236: \end{ruledtabular}
237: \end{table*}
238:
239:
240: \begin{table}
241: \caption{ \label{table_limit}
242: Limits on the Snowmass Slope and two other GMSB models. }
243: \begin{ruledtabular}
244: \begin{tabular}{cccc|cccc}
245: \multicolumn{4}{c}{Fixed parameters} & \multicolumn{4}{c}{95\% CL lower limits} \\
246: $M_m / \Lambda$ & $\tan \beta$ & $N_5$ &
247: sign($\mu$) & $\Lambda $ & $m_{\tilde{\chi}_1^0}$ &
248: $m_{\tilde{\chi}_1^+}$ &
249: $m_{\tilde{\chi}_2^0}$ \\
250: \hline
251: 2 & 15 & 1 & $+$ & 79.6 & 107.7 & 194.9 & 195.9 \\
252: 2 & 5 & 1 & $+$ & 79.5 & 106.0 & 191.6 & 193.3 \\
253: 10 & 5 & 2 & $+$ & 44.0 & 111.4 & 196.0 & 198.7 \\
254: \end{tabular}
255: \end{ruledtabular}
256: \end{table}
257:
258:
259: Since the observed number of events is in good agreement with that
260: expected from the standard model, we conclude that there is no evidence
261: for GMSB SUSY in our data. To calculate the upper
262: limit on the production cross section for each sampled point on the Snowmass
263: Slope,
264: we use a Bayesian approach \cite{bayesian}
265: with a flat prior for the signal cross section. The
266: calculation takes into account uncertainties on the expected number of background
267: events, efficiency, and luminosity.
268: The selection $\MET > 40~$ GeV for the signal sample leads to the best
269: expected limit, given the predicted background and expected signal distributions.
270: Our limits are shown in Table
271: \ref{table_gmsb}, and plotted in Fig. \ref{fig_limit}, together with
272: the expected signal cross section.
273: The upper limit on the cross section is
274: below the expected value for $\Lambda <
275: 79.6~ $TeV, corresponding to lower limits on gaugino masses of
276: $m_{\tilde{\chi}^+_1} > 194.9~$GeV and $m_{\tilde{\chi}^0_1} >
277: 107.7~$GeV. The expected limit, given the predicted number of
278: background events, is $\Lambda > 74.5~ $TeV.
279: We find that the gaugino mass limits depend only slightly on
280: the parameters of the minimal GMSB. We have considered models with
281: values of $\tan \beta$ and $N_5$ different from the Snowmass
282: Slope, and arrive at very similar results as detailed by Table \ref{table_limit}.
283:
284: To summarize, we searched for inclusive high-$E_T$ diphoton events with large
285: missing transverse energy. Such events are predicted in
286: supersymmetric models with low-scale gauge-mediated supersymmetry
287: breaking. We find no excess of such events, and interpret the result
288: as a lower limit on gaugino masses. For a representative point in
289: the parameter space, we determine that at a 95\% confidence level, the
290: masses of the lightest chargino and neutralino are larger than
291: 195 and 108 GeV, respectively. These are the most restrictive
292: limits to date for the Snowmass Slope model.
293:
294: \begin{figure}
295: \includegraphics[scale=0.4]{snowmass_limit.eps}
296: \caption{\label{fig_limit} Predicted cross sections for the Snowmass Slope model vs $\Lambda$ in leading order (thin solid line with crosses), multiplied by the $K$-factor (thin dashed line),
297: and the 95\% C.L. limits (solid line).}
298: \end{figure}
299:
300: We thank S. Martin for valuable discussions and
301: S. Mrenna for his help with the event generators.
302: \input acknowledgement_paragraph_r2.tex % input acknowledgement
303:
304: \begin{thebibliography}{99}
305: \input list_of_visitor_addresses_r2.tex % input visitors address
306:
307: \bibitem{gmsb_original} P. Fayet, Phys. Lett. B {\bf 70}, 461 (1977);
308: {\it ibid.} {\bf 86}, 272 (1979); {\it ibid.} {\bf 175}, 471 (1986).
309:
310: \bibitem{gmsbsusy} M.~Dine, A.~E.~Nelson, Y.~Nir and Y.~Shirman,
311: Phys.\ Rev.\ D {\bf 53}, 2658 (1996);
312: H.~Baer, M.~Brhlik, C.~H.~Chen and X.~Tata,
313: Phys.\ Rev.\ D {\bf 55}, 4463 (1997);
314: H.~Baer, P.~G.~Mercadante, X.~Tata and Y.~L.~Wang,
315: Phys.\ Rev.\ D {\bf 60}, 055001 (1999);
316: S.~Dimopoulos, S.~Thomas and J.~D.~Wells,
317: Nucl.\ Phys.\ B {\bf 488}, 39 (1997);
318: J.~R.~Ellis, J.~L.~Lopez and D.~V.~Nanopoulos,
319: Phys.\ Lett.\ B {\bf 394}, 354 (1997);
320:
321: see also a review by G. F. Giudice and R. Rattazzi,
322: ``Gauge-Mediated Supersymmetry Breaking''
323: in G. L. Kane: {\it Perspectives on Supersymmetry}, World Scientific, Singapore (1998), p. 355-377,
324: and references therein.
325:
326: \bibitem{cdfrun1} F. Abe {\it et al.} (CDF Collaboration),
327: Phys. Rev. D {\bf 59}, 092002 (1999).
328: \bibitem{rpar} G.R. Farrar and P. Fayet, Phys. Lett. {\bf B79} (1978) 442.
329: \bibitem{d0run1} B. Abbott {\it et al.} (\DZero Collaboration), Phys. Rev. Lett. {\bf 80}, 442 (1998).
330: \bibitem{lepsusy} LEPSUSYWG, ALEPH, DELPHI, L3 and OPAL Collaborations, note LEPSUSYWG/04-09.1 (http://lepsusy.web.cern.ch/).
331: \bibitem{modelline} S.~P.~Martin, ~http://zippy.physics.niu.edu/modellineE.html;
332:
333: S.~P.~Martin, S.~Moretti, J.~M.~Qian and G.~W.~Wilson,
334: ``Direct Investigations of Supersymmetry: Subgroup summary report,''
335: in {\it Proceedings of the APS/DPF/DPB Summer Study on the Future of Particle
336: Physics (Snowmass 2001), } edited by N.~Graf,
337: eConf {\bf C010630}, p. 346 (2001);
338:
339: B.~C.~Allanach {\it et al.}, Eur.\ Phys.\ J.\ C {\bf 25}, 113 (2002).
340:
341: \bibitem{run2det} V. Abazov, {\it et al.} (\DZero Collaboration), in preparation for submission
342: to Nucl. Instrum. Methods Phys. Res. A; T. LeCompte and H. T. Diehl,
343: ``The CDF and D\O\ Upgrades for Run II,'' Ann. Rev. Nucl. Part. Sci.
344: {\bf 50}, 71 (2000).
345:
346: \bibitem{run1det} S. Abachi, {\it et al.} (\DZero Collaboration), Nucl. Instrum. Methods
347: Phys. Res. A {\bf 338}, 185 (1994).
348:
349: \bibitem{run2cone} G.~C.~Blazey {\it et al.}, in
350: {\sl Proceedings of the Workshop:
351: ``QCD and Weak Boson Physics in Run II''}
352: edited by U.~Baur, R.~K.~Ellis, and D.~Zeppenfeld, 47 (2000).
353: See Section 3.5 for details.
354:
355: \bibitem{isajet} F.~E.~Paige, S.~D.~Protopescu, H.~Baer and X.~Tata, hep-ph/0312045.
356: \bibitem{pythia} T. Sj\"ostrand {\it et al.}, Computer Physics
357: Commun.
358: {\bf 135} (2001) 238.
359:
360: \bibitem{cteq} H. L. Lai {\it et al.}, (CTEQ Collaboration),
361: Eur. Phys. J.
362: {\bf C} 12 (2000) 375-392.
363:
364: \bibitem{kfact} W.~Beenakker {\it et al.},
365: Phys.\ Rev.\ Lett.\ {\bf 83}, 3780 (1999).
366:
367: \bibitem{bayesian} I.~Bertram {\it et al.} FERMILAB-TM-2104.
368: %\bibitem{bayesian} K.~Hagiwara {\it et al.} (Particle Data Group),
369: %Phys. Rev. D {\bf 66}, 010001 (2002).
370:
371: \end{thebibliography}
372:
373: \end{document}
374: % LocalWords: Gershtein Stelios Kesisoglou SUSY Gauge Diphoton tex gauge GMSB
375: % LocalWords: diphoton pb Fermilab Tevatron neutralino chargino GeV CDF keV di
376: % LocalWords: supersymmetry phenomenology gravitino LSP NLSP slepton LEP SPS
377: % LocalWords: Snowmass Higgs Higgsino TeV Collider superconducting solenoidal
378: % LocalWords: calorimeter muon spectrometer microstrip pseudorapidity radians
379: % LocalWords: calorimeters cryostats Scintillators azimuthal mismeasurement ee
380: % LocalWords: multijets QCD isajet sparticle pythia CTEQ NLO mGMSB cccc Mrenna
381: % LocalWords: gaugino supersymmetric Fayet Phys Lett Nir Shirman Baer Brhlik
382: % LocalWords: Tata Dimopoulos Nucl Giudice Rattazzi CERN et al Moretti Proc
383: % LocalWords: APS Graf eConf Eur Abazov Instrum LeCompte Diehl Sci Abachi
384: % LocalWords: Sjostrand Lonnblad Beenakker Klasen Plehn Spira Hagiwara
385: