1: Events with two acoplanar taus and high
2: missing energy can be the signature of stau pair-production.
3: Due to the scalar nature of the stau,
4: the two taus are produced centrally in the detector.
5: %($\ds / {\mathrm d}\theta \sim \sin^3 \theta$).
6: To select this topology,
7: the particles in the events were grouped into clusters
8: according to the algorithm described in section~\ref{sec:id&jet}.
9: Events with exactly two particle clusters
10: (possibly accompanied by isolated neutral particles)
11: were considered further if
12: they contained at least one charged particle
13: with
14: momentum above 1~\GeVc\ and a relative error less than
15: 30\%. This particle had to be
16: isolated
17: (no neutral or charged particles in a cone of
18: 20\dgree\ half-angle around it),
19: or
20: above 30\dgree\ in polar angle,
21: and its calorimetric energy could not exceed
22: the momentum by more than three times the expected error
23: on the calorimetric measurement.
24: A comparison of data and simulated SM
25: background at this
26: stage~\footnote{
27: In order to show the contribution of all
28: classes of background,
29: the largely dominant Bhabha background was suppressed
30: in these plots
31: by demanding that the missing transverse momentum was above 4 \GeVc,
32: and that the acoplanarity was above 1\dgree.}
33: is shown in figure~\ref{fig:stau:datamc}.
34: It was further required that
35: there were no more than six charged particles in the event and
36: that the total charge was 0 or $\pm$1.
37:
38: Tight and loose electrons, muons and hadrons were defined
39: as described in section~\ref{sec:id&jet}.
40: In addition, if a particle had a loose identification
41: for one species, and was vetoed for the other two,
42: it was considered as tightly identified.
43: If a track pointed to a gap in the electromagnetic
44: calorimetry, it was considered as a loose
45: electron.
46: % even if the standard algorithm vetoed the
47: %electron hypothesis.
48: Tracks with no identification information were
49: treated as loose electrons.
50: Neutral clusters passing the criteria of section~\ref{sec:trksel} were
51: used if their angle to the beam was above 15\dgree.
52: Furthermore, no identified hadronic secondary interactions
53: inside the tracking system were allowed.
54:
55: Beyond this point, the analysis differed depending on
56: whether a stau with mass above or below \MZ/2 was searched for.
57:
58: ~\\
59: \noindent
60: {\underline{\it Search for staus with large mass}}\\
61:
62: \noindent
63: To suppress the
64: two-photon background, it was required that
65: the total transverse momentum imbalance
66: exceeded 4~\GeVc,
67: the total calorimetric energy below $30\dgree$ in polar angle
68: did not exceed
69: 20\% of the beam momentum and the total momentum of
70: the event was within the region
71: $\theta (\Sigma \vec p)\ >\ 30\dgree$.
72: %$30\dgree\ <\ \theta (\Sigma \vec p)\ <\ 150\dgree$
73:
74: To reduce the background from radiative return to the \Zn,
75: none of the clusters was allowed to have a total momentum
76: ($p^{\rm JET}$) above 70\% of the beam momentum,
77: the momentum of isolated photons had to be
78: less than 10\% of the beam momentum,
79: %there had to be no signal in any isolated hermeticity tagger,
80: and the acoplanarity was required to be above $12\dgree$.
81: To reject \eeto \hspace{0.1cm}$\Zn/\gamma \rightarrow $ \tautau\
82: events where the decay of one $\tau$ yielded visible products with large
83: momentum, while the decay of the other $\tau$ yielded soft products,
84: the momentum transverse to the thrust axis was required to
85: exceed 0.7 \GeVc.
86:
87: This selection was supplemented by cuts that depended on the region of
88: the (\mstau,$M_{\mathrm LSP}$) plane considered.
89: At any given point in the plane,
90: it was demanded that the accepted events in that point
91: were kinematically compatible with the corresponding signal:
92: the maximal momentum the
93: $\tau$ can have in the lab frame, neglecting m$_{\tau}$, is
94: $
95: p_{max} =
96: \frac{\sqrt{s}}{4}
97: \left ( 1 - \left [ {M_{\mathrm LSP}}/{\mstau} \right ] ^2 \right )
98: \left ( 1 + \sqrt{ 1 - 4{\mstau}^2/{s} } \right )
99: $,
100: which is also the end-point of the spectrum of the visible
101: $\tau$ - jet momentum.
102: Hence $p^{\rm JET}_{high}$ (the larger of
103: the two jet momenta) was required to be less than $p_{max}$.
104:
105: For large \DM = $\mstau - M_{\mathrm LSP}$,
106: the remaining background from two-photon events can be
107: removed by
108: requiring large
109: transverse momentum imbalance.
110: At smaller values of \DM\, such a cut would greatly reduce the
111: signal detection efficiency.
112: Therefore, if \DM\ was below (above) 20 \GeVcc\ the
113: total missing transverse momentum
114: (\Ptmis)
115: %($p_T^{\rm miss}$)
116: had to exceed 0.8 (1.2) times the maximum
117: transverse momentum a $\gamma \gamma$ event could have
118: without one of the beam-remnant electrons being deflected into
119: the STIC ($i.e.$ by an angle greater than $\theta_{max}$ = $1.82\dgree$).
120: This limit depended on the centre-of-mass energy:
121: $p_T^{\rm lim} = \sqrt{s} \sin \theta_{max} / ( 1 + \sin \theta_{max} ) =
122: 0.031 \ \sqrt{s} $~\footnote{
123: The missing transverse momentum was estimated in four different
124: ways: from the transverse momentum of the two jets,
125: as that of all particles passing the quality cuts,
126: as that of all reconstructed particles except those
127: identified as bremsstrahlung photons,
128: and as that of all reconstructed particles.
129: The cut was applied on the smallest of these.
130: This gave
131: stability against possible errors in reconstruction
132: and against the presence of noise or cosmics.
133: }.
134: In addition, in the region with \DM\ below 20 \GeVcc\
135: there should be no calorimetric energy
136: below $30\dgree$,
137: and $\theta (\Sigma \vec p)$ had to exceed $45\dgree$.
138:
139: At this stage of the analysis,
140: the dominant remaining background at
141: large \DM\
142: were \WW\ events.
143: Only events with both W bosons decaying leptonically
144: were still present.
145: Electrons and muons in \WW\ events might come either directly from
146: the W-decay, or indirectly from $\tau$-decays.
147: In the former case, which is dominant,
148: the momentum of the detected lepton tends to
149: be higher than in the signal, where all electrons and muons
150: would be indirect. In the latter case, the momentum spectrum of
151: background and signal events is similar.
152:
153:
154: In order to reduce the background from \WW\ events with leptons from direct W decays,
155: %where either W decayed into an electron or muon
156: %It was
157: %therefore demanded that the highest momentum of any
158: it was
159: demanded that the highest momentum of any
160: tightly or loosely identified lepton in the
161: event was less than $(0.1 \DM\ + 0.6) P^{lept}_{W min}$,
162: where $P^{lept}_{W min}$ is the lowest momentum a lepton
163: in the decay $W \rightarrow \ell \nu$ can obtain in the lab-frame if the
164: W is on-shell.
165: In the region of large \DM, it was also demanded that
166: there be no more than one
167: tightly identified electron or muon in the event.
168:
169: To suppress further the \WW\ background,
170: and in particular the component with leptons from indirect decays,
171: the events were analysed as if
172: they were indeed \WW\ events.
173: The $\theta$ angle of the
174: W yielding the cluster with positive charge was estimated ($\theta_{W^+}$)
175: and selections in
176: the
177: ($\theta_{W^+}$,$p^{\rm JET}_{high}$)-plane
178: were used to discriminate
179: the signal from this background~\footnote{At this
180: stage of the analysis about 99 \% of the events
181: had a cluster with charge equal to +1 (this selection
182: was not aplied to events where there was no cluster with positive charge).}.
183: %Two cases were distinguished: if the W decayed to a muon
184: The $\theta_{W^+}$ was estimated as follows.
185: If neither W decayed to a tau,
186: $\theta_{W^+}$ can be calculated exactly
187: (albeit usually with two-fold ambiguity).
188: The W decay to a muon
189: or an electron
190: could be distinguished from the decay to the $\tau$ by the presence of
191: a single track cluster with
192: at least a loose lepton identification
193: and a momentum above the lowest possible
194: momentum for the charged lepton from the
195: decay of an on-shell W.
196: Decays that did not fulfil these requirements
197: were assumed to be decays to $\tau$.
198: If one of the W decayed to a $\tau$
199: the solution for $\theta_{W^+}$ is approximate,
200: since the momentum of the $\tau$ is unknown.
201: In this case, the momentum of the initial $\tau$
202: was estimated as the average momentum of taus from
203: W decay, calculated for tau momenta above the
204: measured jet momentum.
205: The direction of the momentum
206: of the tau was taken
207: as the measured jet direction.
208:
209: The W pair-production
210: process tends to have higher
211: %values of the higher of the two jet momenta
212: %(
213: $p^{\rm JET}_{high}$ and
214: %), and in addition
215: $\theta_{W^+}$.
216: In addition, $p^{\rm JET}_{high}$ and $\theta_{W^+}$ are correlated
217: in such processes, while they are independent in \stau\ production.
218: Thus the signal selection criteria in
219: the
220: ($\theta_{W^+}$,$p^{\rm JET}_{high}$)-plane were:
221: $$
222: \theta_{W^+} <
223: \left \{
224: \begin{array} { l l }
225: \mbox{$ -3.0 (x - 0.325) + A $} & \mbox{${\rm if\ } x < 0.325 $} \\
226: \mbox{$ A $} & \mbox{${\rm if\ } 0.325 \le x < 0.52 $} \\
227: \mbox{$-2.1 (x - 0.52) + A $} & \mbox{${\rm if\ } x \ge 0.52$} \\
228: \end{array}
229: \right.
230: $$
231: where $\theta_{W^+}$ is in radians and $x = p^{\rm JET}_{high} / p_{beam}$,
232: and $A$ is a constant
233: chosen to be 1.6 (2.1) for \DM\ below (above) 20 \GeVcc.
234: The boundary of the selected region
235: in the
236: ($\theta_{W^+}$,$p^{\rm JET}_{high}$)-plane
237: thus defined
238: closely follows a curve of
239: constant ratio between the probability density functions for signal and
240: background (the likelihood ratio).
241:
242: ~\\
243: \noindent
244: {\underline{\it Search for light staus without coupling to the \Zn\ }}\\
245:
246: \noindent
247: A light stau
248: can be excluded
249: to a large extent
250: using LEP1 results,
251: as discussed in section~\ref{sec:lep1}.
252: This is however not possible
253: when
254: the coupling to the \Zn\ vanishes
255: (the stau mixing angle gives the minimum cross-section).
256: The large mass
257: analysis described in the previous section looses its efficiency for stau
258: masses below 15~\GeVcc. This is mainly due to the
259: fact that the staus are highly boosted at such small masses,
260: failing the acoplanarity cut.
261:
262: %To a large extent a light stau can be excluded using the agreement
263: %of the decay width of the \Zn\ resonance with the SM prediction,
264: %as observed at LEP1~\cite{ZWIDTH}. The corresponding upper cross-section
265: %limit of 150~pb for non-standard processes at $\sqrt{s} = M_{\mathrm Z}$
266: %excludes a $\stau_R$ below 25~\GeVcc\ and a $\stau_L$ below 30~\GeVcc.
267: %However, at the stau mixing angle
268: %giving the minimum cross-section, the coupling to the \Zn\
269: %vanishes and no exclusion is possible using this method. The large mass
270: %analysis described in the previous section loses its efficiency for stau
271: %masses below 15~\GeVcc. This is mainly due to the
272: %fact that the stau-pairs are highly boosted at such a low mass, so
273: %that they fail the acoplanarity cut.
274:
275: Therefore a specific search was required for small \mstau\
276: at the minimal cross-section mixing angle.
277: %As \mstau\ below 12.5 \GeVcc\ has already been excluded by DELPHI
278: %\cite{slep}, and the large mass analysis can exclude
279: %\DM\ above 3.5 \GeVcc, this search was optimised for \DM\ between
280: %$m_{\tau}$ and 4 \GeVcc.
281: %As \mstau\ below 12.5 \GeVcc\ has already been excluded by DELPHI
282: %\cite{slep}, and
283: Two search regions were identified:
284: one optimised for very small masses, \mstau\ below 10 \GeVcc,
285: referred to as the ``very small mass analysis'',
286: and one for larger masses, optimized
287: for
288: \mstau\ above 10 \GeVcc and
289: \DM\ between $m_{\tau}$ and 4 \GeVcc\ referred to as
290: the ``small mass analysis''
291: (the large mass analysis shows good sensitivity for
292: \DM\ above 4 \GeVcc\ down to \mstau\ = 15 \GeVcc).
293:
294: The signal events in the relevant kinematic regions are characterised
295: by containing two taus at large angles to the beam,
296: and being softer than two-fermion events but slightly
297: harder than $\gamma \gamma$ events. Due to the sizable
298: boost of the staus, the two jets tend to be rather
299: back-to-back.
300: Two-tau events were selected as described in the previous section,
301: with the additional requirement that the topology was either
302: 1-prong and 3-prong, or two 1-prongs. In the latter case,
303: there should not be two tightly identified leptons of the same flavour
304: in the event.
305: To select central events, the polar angle of the most energetic particle
306: in each detector hemisphere had to be
307: above 50\dgree,
308: and the sine of the polar angle of each jet had to be above 0.8.
309: The acollinearity was required to be
310: above 0.4\dgree,
311: $p^{\rm JET}_{max}/p_{beam}$ had to be below
312: 90\%, and
313: the total reconstructed mass of the visible system had to be above 4.5 \GeVcc.
314: There had to be no energy in a 30\dgree\ cone around the beam axis.
315:
316: In the small mass region,
317: $\theta (\Sigma \vec p)$ had to be above 55\dgree\ in polar angle,
318: the visible mass had to be below
319: $\left [ 15 (\DM - m_{\tau})-(\mstau - 25)/3-(\DM - m_{\tau})(\mstau - 25)/2+15 \right ]$ \GeVcc,
320: and
321: \Ptmis\
322: %$p_T^{\rm miss}$
323: had to exceed max(0.05, (\DM $-$ $m_{\tau}$)/4) $p_T^{\rm lim}$.
324:
325: In the very small mass region,
326: it was required that
327: \Ptmis\
328: %$p_T^{\rm miss}$
329: was above 0.01$p_T^{\rm lim}$,
330: $p^{\rm JET}_{max}/p_{beam}$ was above
331: 15\%, and $\theta (\Sigma \vec p)$ was above 15\dgree\ to
332: the beam.
333: The acollinearity had to be below 15\dgree.
334: