1: \documentclass[12pt,a4paper]{article}
2: \usepackage{epsfig}
3: \usepackage{amsmath,amsfonts,amssymb}
4: %\usepackage{showkeys}
5: \usepackage{t1enc}
6: \usepackage{cite}
7: % juan's stuff
8: \newcommand{\ptmiss}{p_T\!\!\!\!\!\!\!\!\! \not \,\,\,\,\,\,\,\,}
9: \renewcommand{\topfraction}{1}
10: \renewcommand{\bottomfraction}{1}
11: \renewcommand{\textfraction}{0}
12: \renewcommand{\baselinestretch}{1.2}
13: \parskip=1.0ex
14: \setcounter{topnumber}{10}
15: \setcounter{bottomnumber}{10}
16: \setcounter{totalnumber}{10}
17:
18: \textwidth=15.5cm
19: \textheight=22cm
20: \oddsidemargin=0.2cm
21: \evensidemargin=0.2cm
22: \topmargin=-1cm
23:
24:
25:
26:
27: \begin{document}
28: \vspace*{-3cm}
29: \begin{flushright}
30: hep-ph/0403243 \\
31: March 2004
32: \end{flushright}
33:
34: \begin{center}
35: \begin{Large}
36: {\bf CP violation in selectron cascade decays \\[0.2cm]
37: $\boldsymbol{\tilde e_L \to e \tilde \chi_2^0 \to e \tilde \chi_1^0 \mu^+
38: \mu^-}$}
39: \end{Large}
40:
41: \vspace{0.5cm}
42: J. A. Aguilar--Saavedra \\[0.2cm]
43: {\it Departamento de Física and GTFP, \\
44: Instituto Superior Técnico, P-1049-001 Lisboa, Portugal} \\
45: \end{center}
46:
47: \begin{abstract}
48: Selectron decays constitute a source of 100\% polarised neutralinos, whose
49: helicity is fixed by the charge and ``chirality'' of the decaying selectron. In
50: SUSY scenarios where the second neutralino $\tilde \chi_2^0$ has three-body
51: decays, the cascade decay
52: $\tilde e_L \to e \tilde \chi_2^0 \to e \tilde \chi_1^0 \mu^+ \mu^-$ provides
53: a clean place to study CP violation in the neutralino
54: sector, through the analysis of CP-violating asymmetries involving the $\tilde
55: \chi_2^0$ spin $\vec s$ and the momenta of the two muons. We show that a
56: CP-violating asymmetry in the triple product
57: $\vec s \cdot (\vec p_{\mu^-} \times \vec p_{\mu^+})$
58: could be observable at a 800 GeV linear collider provided the gaugino mass $M_1$
59: has a large phase at the electroweak scale.
60: % We compare our results with $\tilde \chi_1^0 \tilde \chi_2^0$ production,
61: % where a CP asymmetry is also observable within these scenarios.
62: \end{abstract}
63:
64: % 12.60.Jv,13.10.+q,13.88.+e,14.80.Ly
65:
66: % 12.60.Jv Supersymmetric models
67: % 13.10.+q Weak and electromagnetic interactions of leptons
68: % 13.88.+e Polarization in interactions and scattering
69: % 14.80.Ly Supersymmetric partners of known particles
70:
71: \section{Introduction}
72: \label{sec:1}
73:
74: Supersymmetric theories \cite{susy1,susy2,susy3} are perhaps the best motivated
75: extensions of the standard model (SM). If
76: supersymmetry (SUSY) is realised in nature it must be broken, possibly at a
77: relatively low energy scale.
78: SUSY-breaking interactions are usually assumed to be flavour-blind, at least
79: approximately. Otherwise, they would lead to unacceptable rates for
80: flavour-changing neutral currents. Analogously, large phases in the
81: SUSY-breaking terms of the Lagrangian give supersymmetric contributions to
82: electric dipole moments (EDMs) far above present limits. For a relatively light
83: SUSY spectrum, the only solutions to overcome this problem are,
84: either assume that all SUSY-breaking parameters
85: have very small phases (or are real), or to arrange cancellations among the
86: different contributions to EDMs to satisfy experimental limits
87: \cite{pr1,pr2,kane}.
88:
89: Although the assumption that SUSY-breaking terms are real is conceptually
90: simpler and more attractive, the possibility of large phases and apparently
91: ``fine-tuned'' internal cancellations must not be
92: discarded. Indeed, in quantum field theory the Lagrangian parameters are complex
93: in general, unless there is some argument ({\em e.g.} hermiticity of the
94: Lagrangian or some symmetry) requiring them to be real. Thus,
95: setting the phases of SUSY-breaking terms to zero ``by
96: hand'' may also be regarded as fine tuning, in the absence of a symmetry
97: principle to explain why these terms must be (approximately) real.
98: The possibility of large SUSY CP-violating phases
99: makes compulsory to explore their effects in phenomenology, in order
100: to find out their presence and determine their magnitude.
101:
102: In this Letter we are interested in the direct observation of CP violation due
103: to supersymmetric CP phases, rather than in their effect in CP-conserving
104: quantities such as cross sections and decay widths (see for instance Refs.
105: \cite{CPcon1,CPcon2,CPcon3,CPcon4}). We restrict ourselves to the minimal
106: supersymmetric standard model and focus on the neutralino
107: sector, studying CP asymmetries in the cascade decay
108: $\tilde e_L^\pm \to e^\pm \tilde \chi_2^0 \to e^\pm \tilde \chi_1^0 \mu^+
109: \mu^-$.
110: The $\tilde \chi_2^0$ produced in $\tilde e_L^\pm$ decays are 100\% polarised,
111: having negative helicity in $\tilde e_L^- \to e^- \tilde \chi_2^0$
112: and positive helicity in $\tilde e_L^+ \to e^+ \tilde \chi_2^0$. Having perfect
113: $\tilde \chi_2^0$ polarisation is obviously a great advantage for the study of
114: angular correlations involving the $\tilde \chi_2^0$ spin \cite{note}, and in
115: particular for the study of CP asymmetries.
116: Selectrons will be discovered at LHC \cite{paco} if they have masses of a few
117: hundred GeV, but the large backgrounds present make it impossible a detailed
118: study of their properties, which must be carried out at a $e^+ e^-$ collider.
119: Therefore, as source for left selectrons
120: we consider $\tilde e_L \tilde e_L$ and $\tilde e_R \tilde e_L$
121: production in the processes
122: \begin{align}
123: e^+ e^- & \to \tilde e_L^+ \tilde e_L^- \to e^+ \tilde \chi_1^0 e^- \tilde
124: \chi_2^0 \to e^+ \tilde \chi_1^0 e^- \tilde \chi_1^0 \mu^+ \mu^- \,,
125: \nonumber \\
126: e^+ e^- & \to \tilde e_L^+ \tilde e_L^- \to e^+ \tilde \chi_2^0 e^- \tilde
127: \chi_1^0 \to e^+ \tilde \chi_1^0 \mu^+ \mu^- e^- \tilde \chi_1^0 \,,
128: \nonumber \\
129: e^+ e^- & \to \tilde e_R^+ \tilde e_L^- \to e^+ \tilde \chi_1^0 e^- \tilde
130: \chi_2^0 \to e^+ \tilde \chi_1^0 e^- \tilde \chi_1^0 \mu^+ \mu^- \,,
131: \nonumber \\
132: e^+ e^- & \to \tilde e_R^- \tilde e_L^+ \to e^- \tilde \chi_1^0 e^+ \tilde
133: \chi_2^0 \to e^- \tilde \chi_1^0 e^+ \tilde \chi_1^0 \mu^+ \mu^- \,.
134: \label{ec:proc}
135: \end{align}
136: In these processes all final state
137: momenta can be determined, and the selectron and neutralino rest frames can be
138: reconstructed \cite{npb}, allowing the study of CP asymmetries involving the
139: $\tilde \chi_2^0$ spin $\vec s$ and the momenta of $\mu^+$, $\mu^-$ in the
140: $\tilde \chi_2^0$ rest frame. We do not consider $\tilde e_R^+ \tilde e_R^-$
141: production with one selectron decaying to $e \tilde \chi_2^0$, because the
142: branching ratio of this decay is very small in general.
143: We discuss SUSY scenarios where the second neutralino has three-body decays,
144: in which case it is possible to have a CP asymmetry in the triple product
145: $\vec s \cdot (\vec p_{\mu^-} \times \vec p_{\mu^+})$ of order 0.1.
146: In scenarios with two-body decays $\tilde \chi_2^0 \to \chi_1^0 Z$
147: the asymmetry is of order 0.02, and when two-body decays to sfermions dominate
148: it is negligible.
149: We consider a CM energy $E_\mathrm{CM} = 800$ GeV, as proposed for
150: a TESLA upgrade. At this CM energy, the cross sections for $\tilde e_L$
151: production are higher than at 500 GeV due to the smaller destructive
152: interference between $s$- and $t$-channel contributions and
153: the larger phase space available.
154:
155:
156:
157:
158: \section{Decay of $\boldsymbol{\tilde \chi_2^0}$ and CP asymmetries}
159: \label{sec:2}
160:
161: The decays of the second neutralino $\tilde \chi_2^0 \to \tilde \chi_1^0 f \!
162: \bar f$ are mediated by the Feynman diagrams in Fig.~\ref{fig:X2decay}.
163: (The diagrams in Fig.~\ref{fig:X2decay}b with neutral scalars are negligible
164: except for $f=t,b,\tau$ because they are proportional to the fermion Yukawa
165: couplings.)
166: In this work we are interested in the final state with $\bar f \! f = \mu^+
167: \mu^-$, in which the energies, momenta and charge of both particles can be
168: measured. The decay channel with $\bar f \! f=e^+ e^-$ shares these
169: properties, but the multiplicity of electrons in the final state makes
170: it difficult to identify the ones resulting from the $\tilde \chi_2^0$ decay.
171: The $\bar b b$ final state is also interesting for its large branching ratio,
172: but the $b$ tagging efficiency reduces the signal to the same cross
173: section of the $\mu^+ \mu^-$ channel.
174:
175: \begin{figure}[htb]
176: \begin{center}
177: \begin{tabular}{ccc}
178: \mbox{\epsfig{file=Figs/diag_Z.eps,width=3cm,clip=}} & \hspace*{5mm} &
179: \mbox{\epsfig{file=Figs/diag_h.eps,width=3cm,clip=}} \\
180: (a) & & (b) \\[0.5cm]
181: \mbox{\epsfig{file=Figs/diag_f.eps,width=3cm,clip=}} & \hspace*{5mm} &
182: \mbox{\epsfig{file=Figs/diag_fc.eps,width=3cm,clip=}} \\
183: (c) & & (d)
184: \end{tabular}
185: \caption{Feynman diagrams for the decay $\tilde \chi_2^0 \to \tilde \chi_1^0
186: f \! \bar f$, mediated by $Z$ bosons (a), neutral Higgs bosons (b),
187: and scalar fermions (c and d).}
188: \label{fig:X2decay}
189: \end{center}
190: \end{figure}
191:
192: We consider SUSY scenarios where the second neutralino has three-body decays.
193: This happens when all sleptons are heavier than $\tilde \chi_2^0$ and
194: $m_{\tilde \chi_2^0} < M_Z + m_{\tilde \chi_1^0}$. For definiteness,
195: we choose a scenario similar to SPS1a in Ref.~\cite{sps} but with a
196: heavier sfermion spectrum and complex phases in $M_1$ and
197: $\mu$. In this scenario $\tilde \chi_1^0$ and $\tilde \chi_2^0$ are
198: gaugino-like. The low-energy parameters (at the scale $M_Z$)
199: most relevant for our analysis are
200: collected in Table~\ref{tab:1}. For $\phi_1 = \phi_\mu = 0$, these parameters
201: approximately correspond to $m_{1/2} = 250$ GeV,
202: $m_{\tilde E} = m_{\tilde L} = m_{H_i} = 200$ GeV, $A_E = -200$ GeV at the
203: unification
204: scale, and $\tan \beta = 10$. We use {\tt SPheno} \cite{spheno} to calculate
205: sparticle masses and mixings, as well as some decay widths.
206: Neutralino masses slightly depend on $\phi_1$ and
207: $\phi_\mu$. For $\phi_1 = \phi_\mu = 0$ they are $m_{\tilde \chi_1^0} = 99$ GeV,
208: $m_{\tilde \chi_2^0} = 178$ GeV, $m_{\tilde \chi_3^0} = 384$ GeV,
209: $m_{\tilde \chi_4^0} = 400$ GeV, $m_{\tilde \chi_1^-} = 177$ GeV, and for other
210: values of $\phi_1$ they differ up to $\pm 2$ GeV. The relevant branching ratios
211: (taking $\phi_1 = \phi_\mu = 0$) are
212: $\mathrm{Br}(\tilde e_L \to e \, \tilde \chi_1^0) = 0.18$,
213: $\mathrm{Br}(\tilde e_R \to e \, \tilde \chi_1^0) = 0.998$,
214: $\mathrm{Br}(\tilde e_L \to e \, \tilde \chi_2^0) = 0.30$,
215: $\mathrm{Br}(\tilde \chi_2^0 \to \tilde \chi_1^0 \mu^+ \mu^-) = 0.035$.
216:
217: \begin{table}[htb]
218: \begin{center}
219: \begin{tabular}{ccc}
220: Parameter & ~ & Value \\
221: \hline
222: $M_1$ & & 102.0 $e^{i \phi_1}$ \\
223: $M_2$ & & 192.0 \\
224: $\mu$ & & 377.5 $e^{i \phi_\mu}$ \\
225: $\tan \beta$ & & 10 \\
226: $m_{\tilde e_R},m_{\tilde \mu_R}$ & & 224.0 \\
227: $m_{\tilde e_L},m_{\tilde \mu_L}$ & & 264.5 \\
228: $m_{\tilde \nu_e}$ & & 252.4
229: \end{tabular}
230: \caption{Low-energy parameters (at the scale $M_Z$) for the SUSY scenario used.
231: The dimensionful parameters are in GeV.
232: \label{tab:1}}
233: \end{center}
234: \end{table}
235:
236: CP violation in the neutralino mixing matrix arises when the phases of $M_1$
237: and/or $\mu$ are different from $0,\pi$. These phases generally lead to
238: large supersymmetrc contributions to EDMs. If the squark spectrum (which does
239: not play
240: any role in our analysis) is heavy enough, experimental limits on the neutron
241: and Mercury EDMs are satisfied. On the other
242: hand, for the values for selectron masses under consideration, the experimental
243: bound on the electron EDM $d_e$ imposes a severe constraint on the allowed
244: region in the $(\phi_1,\phi_\mu)$ plane. Using
245: the expressions for $d_e$ in Ref.~\cite{arnowitt}, we
246: find that for each $\phi_1$ between 0 and $2 \pi$ there exist two narrow
247: intervals for $\phi_\mu$ (one with values $\phi_\mu \sim 0$ and the other with
248: values $\phi_\mu \sim \pi$) in which
249: the neutralino and chargino contributions to $d_e$ cancel, resulting in a value
250: compatible with the experimental limit $d_e^\mathrm{\,exp} = (0.079 \pm 0.074)
251: \times 10^{-26} ~ e$ cm \cite{pdb}. (For instance, for $\phi_1 = \pi/2$ we find
252: $\phi_\mu \simeq -0.12$ or $\phi_\mu \simeq 3.21$.)
253: Therefore, in principle it is possible to have any phase $\phi_1$,
254: though with a strong correlation with $\phi_\mu$.
255: Of course, if $\phi_1$ and $\phi_\mu$ are experimentally found to be
256: non-vanishing, a satisfactory explanation will be necessary for this
257: correlation, which apparently would be a
258: ``fine tuning'' of their values \cite{pr3}.
259:
260: Let us discuss which asymmetries may be defined in the $\tilde \chi_2^0$
261: decay. In the $\tilde \chi_2^0$ rest frame, the decay looks as depicted in
262: Fig.~\ref{fig:X2dist}, with $\vec s$ the $\tilde \chi_2^0$ spin and the
263: 3-momenta in obvious notation. Under CP, the spin and momenta transform as:
264: \begin{equation}
265: \vec s \to \vec s \,, \quad
266: \vec p_{\mu^+} \to -\vec p_{\mu^-} \;,\quad
267: \vec p_{\mu^-} \to -\vec p_{\mu^+} \;,\quad
268: \vec p_{\tilde \chi_1^0} \to -\vec p_{\tilde \chi_1^0} \;.
269: \end{equation}
270:
271: \begin{figure}[htb]
272: \begin{center}
273: \epsfig{file=Figs/X2decay_s.eps,width=5cm,clip=}
274: \caption{Schematic picture of the $\tilde \chi_2^0$ decay in its rest frame.}
275: \label{fig:X2dist}
276: \end{center}
277: \end{figure}
278:
279: Using these vectors, we can build the CP-odd quantities
280: \begin{eqnarray}
281: Q_1 & = & \vec s \cdot \left( \vec p_{\mu^-} \times \vec p_{\mu^+}
282: \right) \,, \nonumber \\
283: Q_2 & = & \vec s \cdot \left( \vec p_{\mu^-} + \vec p_{\mu^+}
284: \right) \,.
285: \end{eqnarray}
286: Several other CP-odd quantites can also be constructed involving higher powers
287: of the momenta or using the energies, e.g.\ we can define
288: $Q_3 = \left( E_{\mu^-} - E_{\mu^+}\right)$, which is also odd under CP. For
289: $Q_{1-3}$ we construct the asymmetries
290: \begin{equation}
291: A_i \equiv \frac{N(Q_i > 0) - N(Q_i < 0)}{N(Q_i > 0) + N(Q_i < 0)} \,,
292: \end{equation}
293: where $N$ denotes the number of events. Since $\tilde \chi_2^0$ and
294: $\tilde \chi_1^0$ are Majorana particles, these asymmetries must vanish if
295: CP is conserved. Hence, they are genuine signals of CP violation.
296: We note that $Q_{2,3}$ are even under ``naive'' time reversal T. This implies
297: that $A_{2,3}$ need the presence of absorptive CP-conserving phases in the
298: amplitude in order to be nonvanishing \cite{Teven}.
299: CP-conserving phases appear in loop diagrams with on-shell propagators,
300: through final state interactions or from the widths of intermediate unstable
301: particles (the phases originated by particle widths are usually tiny).
302: In $\tilde \chi_2^0$ decay,
303: T-even asymmetries like $A_{2,3}$ can result from the interference of a
304: dominant tree-level and a subleading loop diagram mediating the decay and are
305: expected to be small. On the other
306: hand $Q_1$ is T-odd, thus a relatively large asymmetry $A_1$ can be generated
307: at the tree level, without the need of an absorptive phase.
308: Our analysis is centred in the asymmetry $A_1$.
309:
310:
311:
312:
313:
314: \section{Generation of the signals}
315: \label{sec:3}
316:
317: We calculate the matrix element for the resonant processes in
318: Eqs.~(\ref{ec:proc})
319: using {\tt HELAS} \cite{helas}, so as to include all spin correlations and
320: finite width effects. The relevant terms of the Lagrangian and conventions used
321: can be found in Refs.~\cite{npb,next}. We assume a CM energy of 800 GeV, with
322: $e^-$ polarisation $P_{e^-} = -0.8$ and $e^+$ polarisation
323: $P_{e^+} = 0.6$, and an integrated luminosity of 534 fb$^{-1}$ per year
324: \cite{lum}. Beam polarisation does not have any effect on the asymmetries, which
325: are defined for $\tilde e_L$ decays independently of the production mechanism,
326: but increases the total signal cross sections. In our calculation we take into
327: account the
328: effects of initial state radiation (ISR) \cite{isr} and beamstrahlung
329: \cite{peskin,BS2}. For the latter we use the
330: design parameters $\Upsilon = 0.09$, $N = 1.51$ \cite{lum}.\footnote{The actual
331: expressions for ISR and beamstrahlung used in our calculation can be found in
332: Ref.~\cite{npb}.}
333: We also include a beam energy spread of 1\%.
334:
335: In order to simulate the calorimeter and tracking resolution, we perform a
336: Gaussian smearing
337: of the energies of electrons ($e$) and muons ($\mu$) using the
338: specifications in the TESLA Technical Design Report \cite{tesla2}
339: \begin{equation}
340: \frac{\Delta E^e}{E^e} = \frac{10\%}{\sqrt{E^e}} \oplus 1 \% \;, \quad
341: \frac{\Delta E^\mu}{E^\mu} = 0.02 \% \, E^\mu \;,
342: \end{equation}
343: where the energies are in GeV and the two terms are added in quadrature. We
344: apply ``detector'' cuts on transverse momenta, $p_T \geq 10$ GeV, and
345: pseudorapidities $|\eta| \leq 2.5$, the latter corresponding to polar angles
346: $10^\circ \leq \theta \leq 170^\circ$. We also reject events in which the
347: leptons are not isolated, requiring a ``lego-plot'' separation
348: $\Delta R = \sqrt{\Delta \eta^2+\Delta \phi^2} \geq 0.4$.
349: We do not require specific trigger conditions, and we assume that the
350: presence of charged leptons with high transverse momentum will suffice.
351: For the Monte Carlo integration in 8-body phase space we use
352: {\tt RAMBO} \cite{rambo}.
353:
354: The reconstruction of the final state momenta is done with the
355: procedure described in detail in Ref.~\cite{npb}, with some
356: modifications. In general, it is necessary to have as
357: many kinematical relations as unknown variables in order to determine the
358: momenta of the undetected particles. In our case, there are 8 unknowns (the 4
359: components of the two $\tilde \chi_1^0$ momenta) and 8 constraints. These are
360: derived from energy and momentum conservation (4 constraints), from the fact
361: that the two $\tilde \chi_1^0$ are on shell (two constraints), from the
362: kinematics of the decay of the $\tilde \chi_2^0$ (one constraint). The last
363: constraint comes
364: from the hypothesis that in $e^+ e^-$ collisions two particles are
365: produced, either with the same mass or having a squared mass difference equal
366: to $m_{\tilde e_L}^2 - m_{\tilde e_R}^2$.
367: The reconstruction is attempted for both cases, considering different effective
368: CM energies $E_\mathrm{eff} < E_\mathrm{CM}$ (to partially take into
369: account ISR and beamstrahlung effects), selecting the one which gives
370: reconstructed selectron masses closest to their actual values which
371: can be measured in other processes \cite{tdr,martyn,feng,blochinger}.
372: If the event does not reasonably fit into any of the two possibilities, it is
373: discarded.
374: The identification of $\tilde e_L \tilde e_L$ versus $\tilde e_R \tilde e_L$
375: production is successful in most cases, with a good ``tagging'' 80\% of the time
376: for $\tilde e_L \tilde e_L$ and 70\% for $\tilde e_R \tilde e_L$. We note that
377: these processes are topologically very similar, being the only difference the
378: selectron energies in CM frame (and thus the magnitude of their 3-momentum).
379: In the former process, both selectron energies are 400 GeV, whereas in the
380: latter they are 412 GeV for $\tilde e_L$, 388 GeV for $\tilde e_R$.
381:
382: The reconstruction procedure determines the momenta of the two
383: unobserved $\tilde \chi_1^0$, identifying for $\tilde e_L^+ \tilde e_L^-$
384: whether the selectron pair
385: has decayed in the channel $\tilde e_L^+ \to e^+ \tilde \chi_1^0$,
386: $\tilde e_L^- \to e^- \tilde \chi_2^0 \to e^- \tilde \chi_1^0 \mu^+ \mu^-$
387: or, on the contrary, in the channel
388: $\tilde e_L^+ \to e^+ \tilde \chi_2^0 \to e^+ \tilde \chi_1^0 \mu^+ \mu^-$,
389: $\tilde e_L^- \to e^- \tilde \chi_1^0$. The knowledge of all
390: final state momenta, as well as the identification of the particles
391: resulting from each decay, allows us to construct various mass, angular and
392: energy distributions \cite{note}, and in particular the determination of
393: the selectron and $\tilde \chi_2^0$ rest frames.
394: In each event, the $\tilde \chi_2^0$ spin direction $\vec s$ can be found
395: as follows: If the $\tilde \chi_2^0$ results from the decay of a $\tilde e_L^-$,
396: it has negative helicity, so its spin direction is $\vec s=-\vec p$, with $\vec
397: p$ the $\tilde \chi_2^0$ momentum in the $\tilde e_L^-$ rest frame.
398: If the $\tilde \chi_2^0$ results from the decay of a $\tilde e_L^+$, it has
399: positive
400: helicity and its spin direction is $\vec s=\vec p$, with $\vec
401: p$ the $\tilde \chi_2^0$ momentum in the $\tilde e_L^+$ rest frame.
402:
403: Finally, it is necessary to discuss the possible SM and SUSY backgrounds to our
404: signal. The relevant SUSY backgrounds are sneutrino and $\tilde \chi_2^0$
405: pair production, in the decay channels
406: \begin{align}
407: e^+ e^- & \to \tilde \nu_e^* \tilde \nu_e \to e^+ \tilde \chi_1^- \, e^- \tilde
408: \chi_1^+ \to e^+ \tilde \chi_1^0 \mu^- \bar \nu_\mu \, e^- \tilde \chi_1^0
409: \mu^+ \nu_\mu \,, \nonumber \\
410: e^+ e^- & \to \tilde \nu_e^* \tilde \nu_e \to e^+ \tilde \chi_1^- \, \nu_e \,
411: \tilde \chi_2^0 \to e^+ \tilde \chi_1^0 e^- \bar \nu_e \, \nu_e \,
412: \tilde \chi_1^0 \mu^+ \mu^- \,, \nonumber \\
413: e^+ e^- & \to \tilde \chi_2^0 \tilde \chi_2^0 \to \tilde \chi_1^0 e^+ e^-
414: \, \tilde \chi_1^0 \mu^+ \mu^- \,.
415: \label{ec:EEb}
416: \end{align}
417: In sneutrino pair production the $\tilde \chi_1^- \tilde \chi_1^+$ decay
418: channel has a larger branching ratio, with
419: $\mathrm{Br}(\tilde \nu \to e^- \tilde \chi_1^+) = 0.52$,
420: $\mathrm{Br}(\tilde \chi_1^+ \to \tilde \chi_1^0 e^+ \nu ) = 0.1$,
421: while for the the $\tilde \chi_1^- \tilde \chi_2^0$ mode
422: $\mathrm{Br}(\tilde \nu \to \nu \tilde \chi_2^0) = 0.22$,
423: $\mathrm{Br}(\tilde \chi_2^0 \to \tilde \chi_1^0 \mu^+ \mu^-) = 0.035$.
424: We have calculated the three processes in the same way described for selectron
425: pair production, and include them in our results. The
426: cross sections of these backgrounds turn out to be 4 times larger than the
427: signals, but can be reduced with the reconstruction of the final state
428: momenta.
429: The reconstruction method applied for the signal partially eliminates the three
430: backgrounds. The first and third ones are further reduced applying a
431: reconstruction
432: procedure specific for each case (not discussed here for brevity) and requiring
433: that signal and background events do
434: not have a kinematics similar to sneutrino or $\tilde \chi_2^0$ pair
435: production.\footnote{In the
436: $\tilde \chi_1^- \tilde \chi_1^+$ channel only the momenta of each $\nu
437: \tilde \chi_1^0$ pair can be determined, but that is enough to obtain the
438: chargino and sneutrino momenta.}
439: This is not possible for the second background in Eq.~(\ref{ec:EEb}) due to the
440: different kinematics of the process.
441: A cut requiring transverse energy $H_T > 200$ GeV is applied as
442: well. Other SUSY backgrounds like $e^+ e^- \to \tilde \chi_1^\pm \chi_2^\mp
443: \to \chi_1^+ \chi_1^- Z \to e^+ e^- \mu^+ \mu^- \nu \bar \nu \tilde \chi_1^0
444: \tilde \chi_1^0$ (involving several decay channels which lead to the same
445: final state of $e^+ e^- \mu^+ \mu^-$ plus missing energy and momentum)
446: are much smaller, with cross sections smaller than 0.01 fb.
447: The SM background is given by six fermion production $e^+ e^- \to
448: e^+ e^- \mu^+ \mu^- \nu \bar \nu$. Its cross section is below 0.3 fb
449: \cite{lusifer} and with kinematical cuts can be eliminated more easily than SUSY
450: backgrounds.
451:
452:
453:
454: \section{Numerical results}
455: \label{sec:4}
456:
457: We first examine the possible values that this asymmetry may take in
458: connection with
459: EDM constraints. In Fig.~\ref{fig:A1th} we show its dependence on the
460: two phases $\phi_1$, $\phi_\mu$ (these plots are calculated with the Monte Carlo
461: described in last section, but assuming perfect momenta reconstruction and
462: particle identification, and without any kinematical nor detector cuts).
463: For $\phi_1$ sufficiently large $A_1$ reaches values of $\pm 0.13$,
464: while for $\phi_1=0$ the asymmetry is negligible independently of
465: $\phi_\mu$.\footnote{The main contribution to the asymmetry comes from the
466: interference between the $Z$ diagram in Fig.~\ref{fig:X2decay}a and the
467: diagrams with $\mu_R$ exchange in Fig.~\ref{fig:X2decay}c and
468: \ref{fig:X2decay}d.
469: We note that the $Z$ contribution alone produces an
470: asymmetry $A_1 = 0.021$ without the need of interference, but the asymmetry is
471: much larger when the rest of diagrams are included.}
472: The dependence of the cross section on these two phases is plotted
473: in Fig.~\ref{fig:sigth}.
474: Our approach is then as follows: for each value of $\phi_1$, we know from
475: the discussion in Section~\ref{sec:2} that there is an allowed
476: interval of $\phi_\mu$ (which we may take with $|\phi_\mu| \lesssim 0.12$)
477: in which the electron EDM does not exceed the experimental bound.
478: We then calculate $A_1$ for this $\phi_1$, but taking $\phi_\mu = 0$, since the
479: asymmetry is almost independent of the latter phase and the effect on the cross
480: section is also rather small for $|\phi_\mu| \lesssim 0.12$.
481:
482: \begin{figure}[htb]
483: \begin{center}
484: \begin{tabular}{cc}
485: \epsfig{file=Figs/asim-th.eps,width=7.3cm,clip=} &
486: \epsfig{file=Figs/asim-mu-th.eps,width=7.3cm,clip=} \\
487: (a) & (b)
488: \end{tabular}
489: \caption{Theoretical value of $A_1$ as a function of $\phi_1$, for $\phi_\mu =0$
490: (a) and as a function of $\phi_\mu$, for $\phi_1=0$ (b).}
491: \label{fig:A1th}
492: \end{center}
493: \end{figure}
494:
495: \begin{figure}[htb]
496: \begin{center}
497: \begin{tabular}{cc}
498: \epsfig{file=Figs/cross-th.eps,width=7.3cm,clip=} &
499: \epsfig{file=Figs/cross-mu-th.eps,width=7.3cm,clip=} \\
500: (a) & (b)
501: \end{tabular}
502: \caption{Total cross section (without ISR, beamstrahlung and beam spread
503: effects, and without any kinematical cuts) as a function of $\phi_1$, for
504: $\phi_\mu =0$ (a) and as a function of $\phi_\mu$, for $\phi_1=0$ (b).}
505: \label{fig:sigth}
506: \end{center}
507: \end{figure}
508:
509:
510:
511: The cross sections for $\phi_1 = 0$ of the different processes in
512: Eqs.~(\ref{ec:proc},\ref{ec:EEb}) are collected in
513: Table~\ref{tab:cross}, before and after final state momenta reconstruction and
514: kinematical cuts. All figures include
515: the various corrections discussed in the previous section. The
516: reconstruction of the signal and backgrounds allows us to reduce the latter by a
517: factor of 20, while keeping 80\% of the signal.
518: We define the normalised quantity
519: $\hat Q_1 = \vec s \cdot \left( \hat p_{\mu^-} \times \hat p_{\mu^+} \right)$,
520: using unit vectors in the direction of the muon momenta, so that $-1 \leq \hat
521: Q_1 \leq 1$. The kinematical distribution of $\hat Q_1$ for the signals
522: is shown in Fig.~\ref{fig:Q1}, taking $\phi_1 = 0$ and $\phi_1 = \pi/2$. We have
523: normalised the $\tilde e_L^+ \tilde e_L^-$ cross sections to unity, and the
524: $\tilde e_R^\pm \tilde e_L^\mp$ ones to $1/3$. In this plot we take into account
525: ISR, beamstrahlung, beam spread and detector effects.
526: The $\hat Q_1$ distribution already shows
527: that the asymmetry is a real dynamical effect and is not
528: a fake asymmetry caused by ``detector'' cuts applied in phase space. In our
529: calculation we find that for $\phi_1 = \pi/2$ the relative difference between
530: the phase space volumes of the two hemispheres
531: (with $\hat Q_1 < 0$ and $\hat Q_1 > 0$) is negligible, of $3.1 \times 10^{-4}$
532: for $\tilde e_L^+ \tilde e_L^-$ and $6.9 \times 10^{-4}$ for $\tilde e_R^\pm
533: \tilde e_L^\mp$. We note that this is not the case for the
534: asymmetry $A_2$. Even in the CP-conserving case with $\phi_1=0$ a
535: fake asymmetry $A_2 = -0.055$ is generated just by phase space cuts, and might
536: obscure the observation of a real CP asymmetry, which is expected to be very
537: small in this case.
538:
539: \begin{table}[htb]
540: \begin{center}
541: \begin{tabular}{ccc}
542: & Before & After \\
543: \hline
544: $\tilde e_L^+ \tilde e_L^-$ & 0.21 & 0.18 \\
545: $\tilde e_R^+ \tilde e_L^-$ & 0.061 & 0.046 \\
546: $\tilde e_L^+ \tilde e_R^-$ & 0.027 & 0.021 \\
547: $\tilde \nu^* \tilde \nu$ ($\tilde \chi_1^- \tilde \chi_1^+$) & 0.99 & 0.039 \\
548: $\tilde \nu^* \tilde \nu$ ($\tilde \chi_1^- \tilde \chi_2^0$) & 0.24 & 0.011 \\
549: $\tilde \chi_2^0 \tilde \chi_2^0$ & 0.11 & 0.0081 \\
550: \end{tabular}
551: \end{center}
552: \label{tab:cross}
553: \caption{Cross sections (in fb) before and after signal reconstruction and
554: kinematical cuts of the processes in Eqs.~(\ref{ec:proc},\ref{ec:EEb}).}
555: \end{table}
556:
557: \begin{figure}[htb]
558: \begin{center}
559: \epsfig{file=Figs/Q1.eps,width=10cm,clip=}
560: \caption{Kinematical distribution of the normalised triple product $\hat Q_1$.}
561: \label{fig:Q1}
562: \end{center}
563: \end{figure}
564:
565: The difference between the distributions for the two values of the phase
566: is smaller for the $\tilde e_R \tilde e_L$ signals. This is because the
567: reconstruction procedure has a smaller efficiency for the correct identification
568: of $\tilde e_R^\pm \tilde e_L^\mp$ events,
569: leading to a small washing-out of the asymmetry. For $\phi_1 = \pi/2$, we have
570: for $\tilde e_L^+ \tilde e_L^-$ the value $A_1 = 0.110$, while for
571: $\tilde e_R^+ \tilde e_L^-$ and $\tilde e_R^- \tilde e_L^+$
572: we find $A_1 = 0.098$, $A_1 = 0.099$, respectively.
573: The total asymmetry $A_1$ as a function of $\phi_1$ is plotted
574: in Fig.~\ref{fig:total}a, including the backgrounds in Eqs.~(\ref{ec:EEb}),
575: which have vanishing CP asymmetry, as well as ISR, beamstrahlung, beam spread
576: and detector effects.
577: The shaded region represents the statistical error
578: for two years of running, with a luminosity of 534 fb$^{-1}$ per year.
579: The cross section also exhibits a strong dependence on $\phi_1$, as can be
580: observed in Fig.~\ref{fig:total}b. Nevertheless, the measurement of the cross
581: section is not likely to provide any information on the phase $\phi_1$, due to
582: the multiple theoretical uncertainties present regarding the neutralino
583: mixing matrix, sparticle masses, scale dependence, etc.
584:
585: \begin{figure}[ht]
586: \begin{center}
587: \begin{tabular}{cc}
588: \epsfig{file=Figs/asim.eps,width=7.3cm,clip=} &
589: \epsfig{file=Figs/cross.eps,width=7.3cm,clip=} \\
590: (a) & (b)
591: \end{tabular}
592: \caption{(a) Asymmetry $A_1$ as a function of the phase $\phi_1$. The shaded
593: region
594: represents the statistical error for two years of running. (b) Total cross
595: section for $e^+ e^- \to \tilde e_{L,R}^\pm \tilde e_L^\mp
596: \to e^+ e^- \mu^+ \mu^- \tilde \chi_1^0 \tilde \chi_1^0$ as a function of
597: $\phi_1$. In both cases the backgrounds from $\tilde \nu^* \tilde \nu$ and
598: $\tilde \chi_2^0 \tilde \chi_2^0$ production are included.}
599: \label{fig:total}
600: \end{center}
601: \end{figure}
602:
603:
604:
605:
606:
607:
608:
609: \section{Summary}
610: \label{sec:6}
611:
612:
613: In this work we have shown that a CP asymmetry in selectron cascade decays
614: $\tilde e_L \to e \tilde \chi_2^0 \to e \tilde \chi_1^0 \mu^+ \mu^-$ can be
615: observable in SUSY scenarios where $\tilde \chi_2^0$ has three-body decays,
616: if there is a large nonzero phase $\phi_1$ in the gaugino mass $M_1$. The key
617: features for the observation of the CP asymmetry are: ({\em i\/}) the neutralinos
618: produced from selectron decays are 100\% polarised; ({\em ii\/}) all final
619: state momenta can be accurately reconstructed in the processes under
620: consideration; ({\em iii\/}) this reconstruction allows us to eliminate
621: potentially dangerous backgrounds.
622:
623: In SUSY scenarios with three-body decays of $\tilde \chi_2^0$, a
624: sizeable $\tilde e_L$ production
625: can only take place at a TESLA upgrade with a CM energy of 800 GeV.
626: We have selected one of such scenarios, with $\tilde \chi_1^0$ and $\tilde
627: \chi_2^0$ gaugino-like, and shown that the CP asymmetry in
628: the triple product $\vec s \cdot (\vec p_{\mu^-} \times \vec p_{\mu^+})$ could
629: be up to $A_1 = \pm 0.1$ for large phases $\phi_1 \simeq \pm 2$. These
630: asymmetries could be observable with 1.8 standard deviations after two years of
631: running with the proposed luminosity. It can also be seen that
632: in similar scenarios where $\tilde \chi_1^0$ and $\tilde \chi_2^0$ have larger
633: Higgsino components the asymmetries could be even larger, and observable as
634: well. The results here can be compared with the CP asymmetry in
635: $\tilde \chi_1^0 \tilde \chi_2^0$ production \cite{bartl2} in the same SUSY
636: scenario. This analysis will be presented elsewhere \cite{next}. In $\tilde
637: \chi_1^0 \tilde \chi_2^0$ production, although the cross section is much higher,
638: the asymmetry
639: is smaller and there are large backgrounds that further reduce it. The maximum
640: statistical significance obtained for the asymmetry is $1.5 \, \sigma$
641: when $\phi_1 = 3 \pi/4$.
642: At any rate these two processes are complementary, because the
643: asymmetries have a different dependence on $\phi_1$ in each case.
644: For instance, the CP asymmetry in $\tilde \chi_1^0 \tilde \chi_2^0$ production
645: almost vanishes for $\phi_1 = \pi/2$ and $\phi_1 = 3 \pi/2$ \cite{next},
646: while in selectron cascade decays it is nearly maximal.
647:
648: To conclude, we stress that the existence of CP-violating phases in
649: SUSY-breaking terms of the Lagrangian is still an open question. In this paper
650: we have focused on the phases $\phi_1$ and $\phi_\mu$ of the parameters
651: $M_1$ and $\mu$, respectively. Assuming a relatively light SUSY spectrum there
652: are two alternative possibilities to explain the unobserved EDM of the
653: electron:
654: either $M_1$ and $\mu$ are approximately real or on the contrary they have
655: large phases and there exist
656: cancellations between neutralino and chargino contributions to the electron
657: EDM. In this situation, the study of observables with a different functional
658: dependence on $\phi_1$ and $\phi_\mu$, as the CP asymmetry investigated here,
659: is of great help in order to confirm one of the two hypotheses.
660:
661:
662: \vspace{1cm}
663: \noindent
664: {\Large \bf Acknowledgements}
665:
666: \vspace{0.4cm} \noindent
667: I thank A. Bartl, S. Hesselbach and A. M. Teixeira for collaboration at the
668: initial stage of this work. This work has been supported
669: by the European Community's Human Potential Programme under contract
670: HTRN--CT--2000--00149 Physics at Colliders and by FCT
671: through projects POCTI/FNU/43793/2002, CFIF--Plurianual (2/91) and
672: grant SFRH/BPD/12603/2003.
673:
674: %\clearpage
675:
676: \begin{thebibliography}{99}
677:
678: \bibitem{susy1}
679: H. P. Nilles, Phys. Rept. {\bf 110} (1984) 1
680: %%CITATION = PRPLC,110,1;%%
681:
682: \bibitem{susy2}
683: H. E. Haber and G. L. Kane, Phys. Rept. {\bf 117} (1985) 75
684: %%CITATION = PRPLC,117,75;%%
685:
686: \bibitem{susy3}
687: S. P. Martin, in {\em ``Perspectives on supersymmetry''}, G. L. Kane (ed.),
688: hep-ph/9709356
689: %%CITATION = HEP-PH 9709356;%%
690:
691: \bibitem{pr1}
692: T. Ibrahim and P. Nath,
693: %``The neutron and the electron electric dipole moment in N = 1 supergravity
694: %unification,''
695: Phys. Rev. D {\bf 57} (1998) 478
696: [Erratum-ibid. D {\bf 58} (1998) 019901, Erratum-ibid. D {\bf 60} (1999) 079903,
697: Erratum-ibid. D {\bf 60} (1999) 019901]
698: %%CITATION = HEP-PH 9708456;%%
699:
700: \bibitem{pr2}
701: T. Ibrahim and P. Nath,
702: %``The neutron and the lepton EDMs in MSSM, large CP violating phases, and the
703: %cancellation mechanism,''
704: Phys. Rev. D {\bf 58} (1998) 111301 [Erratum-ibid.\ D {\bf 60} (1999) 099902]
705: %%CITATION = HEP-PH 9807501;%%
706:
707: \bibitem{kane}
708: M. Brhlik, G. J. Good and G. L. Kane,
709: %``Electric dipole moments do not require the CP-violating phases of
710: %supersymmetry to be small,''
711: Phys. Rev. D {\bf 59} (1999) 115004
712: %%CITATION = HEP-PH 9810457;%%
713:
714: \bibitem{CPcon1}
715: S. Y. Choi, A. Djouadi, M. Guchait, J. Kalinowski, H. S. Song and P. M. Zerwas,
716: %``Reconstructing the chargino system at e+ e- linear colliders,''
717: Eur. Phys. J. C {\bf 14} (2000) 535
718: %%CITATION = HEP-PH 0002033;%%
719:
720: \bibitem{CPcon2}
721: S. Y. Choi, J. Kalinowski, G. Moortgat-Pick and P. M. Zerwas,
722: %``Analysis of the neutralino system in supersymmetric theories,''
723: Eur. Phys. J. C {\bf 22} (2001) 563
724: %%CITATION = HEP-PH 0108117;%%
725: [Addendum-ibid.\ C {\bf 23} (2002) 769]
726: %%CITATION = HEP-PH 0202039;%%
727:
728: \bibitem{CPcon3}
729: A. Bartl, K. Hidaka, T. Kernreiter and W. Porod,
730: %``tau-sleptons and tau-sneutrino in the MSSM with complex parameters,''
731: Phys. Rev. D {\bf 66} (2002) 115009
732: %%CITATION = HEP-PH 0207186;%%
733:
734: \bibitem{CPcon4}
735: A. Bartl, S. Hesselbach, K. Hidaka, T. Kernreiter and W. Porod,
736: %``Impact of CP phases on stop and sbottom searches,''
737: Phys. Lett. B {\bf 573} (2003) 153
738: %%CITATION = HEP-PH 0307317;%%
739:
740: \bibitem{note}
741: J. A. Aguilar-Saavedra, LC-TH-2003-098, hep-ph/0312140
742: %%CITATION = HEP-PH 0312140;%%
743:
744: \bibitem{paco}
745: F. del Aguila and L. Ametller, Phys.\ Lett.\ B {\bf 261} (1991) 326
746: %%CITATION = PHLTA,B261,326;%%
747:
748: \bibitem{npb}
749: J. A. Aguilar-Saavedra and A. M. Teixeira, Nucl. Phys. {\bf B} 675, 70 (2003)
750: %%CITATION = HEP-PH 0307001;%%
751:
752: \bibitem{spheno}
753: W. Porod, Comput. Phys. Commun. {\bf 153}, 275 (2003)
754: %%CITATION = HEP-PH 0301101;%%
755:
756: \bibitem{sps}
757: B. C. Allanach {\em et al.},
758: %``The Snowmass points and slopes: Benchmarks for SUSY searches,''
759: in {\em Proc. of the APS/DPF/DPB Summer Study on the Future of Particle Physics
760: (Snowmass 2001) } ed. N. Graf, Eur. Phys. J. C {\bf 25}, 113 (2002)
761: [eConf {\bf C010630}, P125 (2001)]
762: %%CITATION = HEP-PH 0202233;%%
763:
764: \bibitem{arnowitt}
765: R. Arnowitt, B. Dutta and Y. Santoso, Phys. Rev. D {\bf 64} (2001) 113010
766: %%CITATION = HEP-PH 0106089;%%
767:
768: \bibitem{pdb}
769: K. Hagiwara {\em et al.}, Particle Data Group, Phys. Rev. {\bf D66}
770: (2002) 010001
771: %%CITATION = PHRVA,D66,010001;%%
772:
773: \bibitem{pr3}
774: T. Ibrahim and P. Nath,
775: %``Large CP phases and the cancellation mechanism in EDMs in SUSY, string and
776: %brane models,''
777: Phys. Rev. D {\bf 61} (2000) 093004
778: %%CITATION = HEP-PH 9910553;%%
779:
780: \bibitem{Teven}
781: See for instance
782: D. Atwood, S. Bar-Shalom, G. Eilam and A. Soni,
783: %``CP violation in top physics,''
784: Phys. Rept. {\bf 347} (2001) 1
785: %%CITATION = HEP-PH 0006032;%%
786:
787: \bibitem{helas}
788: E. Murayama, I. Watanabe and K. Hagiwara, KEK report 91-11, January 1992
789: %%CITATION = NONE;%%
790:
791: \bibitem{lum}
792: International Linear Collider Technical Review Committee 2003 Report,
793: {\small \tt
794: http://www.slac.stanford.edu/xorg/ilc-trc/2002/2002/report/03rep.htm}
795:
796: \bibitem{isr}
797: M. Skrzypek and S. Jadach, Z. Phys. C {\bf 49} (1991) 577
798: %%CITATION = ZEPYA,C49,577;%%
799:
800: \bibitem{peskin}
801: M. Peskin, Linear Collider Collaboration Note LCC-0010, January 1999
802: %%CITATION = NONE;%%
803:
804: \bibitem{BS2}
805: K. Yokoya and P. Chen, SLAC-PUB-4935. {\em Presented at IEEE Particle
806: Accelerator Conference, Chicago, Illinois, Mar 20-23, 1989}
807: %%CITATION = NONE;%%
808:
809: \bibitem{tesla2}
810: G. Alexander {\em et al.}, TESLA Technical Design Report Part 4,
811: DESY-01-011
812: %%CITATION = NONE;%%
813:
814: \bibitem{rambo}
815: R. Kleiss, W. J. Stirling and S. D. Ellis,
816: Comput. Phys. Commun. {\bf 40} (1986) 359
817: %%CITATION = CPHCB,40,359;%%
818:
819: \bibitem{tdr}
820: J. A. Aguilar-Saavedra {\it et al.} [ECFA/DESY LC Physics Working Group
821: Collaboration], hep-ph/0106315
822: %%CITATION = HEP-PH 0106315;%%
823:
824: \bibitem{martyn}
825: H. U. Martyn and G. A. Blair, hep-ph/9910416
826: %%CITATION = HEP-PH 9910416;%%
827:
828: \bibitem{feng}
829: J. L. Feng and M. E. Peskin, Phys. Rev. D {\bf 64} (2001) 115002
830: %%CITATION = HEP-PH 0105100;%%
831:
832: \bibitem{blochinger}
833: C. Blochinger, H. Fraas, G. Moortgat-Pick and W. Porod,
834: Eur. Phys. J. C {\bf 24} (2002) 297
835: %%CITATION = HEP-PH 0201282;%%
836:
837: \bibitem{lusifer}
838: S. Dittmaier and M. Roth,
839: %``LUSIFER: A LUcid approach to SIx FERmion production,''
840: Nucl. Phys. B {\bf 642} (2002) 307
841: %%CITATION = HEP-PH 0206070;%%
842:
843: \bibitem{bartl2}
844: A. Bartl, H. Fraas, O. Kittel and W. Majerotto, hep-ph/0402016
845: %%CITATION = HEP-PH 0402016;%%
846:
847: \bibitem{next}
848: J. A. Aguilar-Saavedra, hep-ph/0404104, to be published in Nucl. Phys. B
849: %%CITATION = HEP-PH 0404104;%%
850:
851: \end{thebibliography}
852:
853:
854: \end{document}
855:
856:
857:
858:
859:
860:
861:
862: