Marco Sozzi

Print publication date: 2007

Print ISBN-13: 9780199296668

Published to Oxford Scholarship Online: January 2008

DOI: 10.1093/acprof:oso/9780199296668.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 26 February 2017

(p.503) APPENDIX D THE DALITZ–FABRI PLOT

Source:
Discrete Symmetries and CP Violation
Publisher:
Oxford University Press

In this appendix a widely used construction for analysing three-body decays is discussed.

The Lorentz-invariant phase space element d(LIPS) for the three-body decay of a particle of mass M into three particles of masses mi(i = 1, 2, 3) is

$Display mathematics$
where p i, Ei are the momenta and energies of the three particles, θi, ϕi their polar and azimuthal angles, and P, E the momentum and energy of the decaying one. Due to energy-momentum conservation this expression actually depends on only two independent variables. The integration on one of the momenta is cancelled by the momentum delta function; the uniformly distributed variables Ω1 and ϕ 2 can be integrated over: the first gives the overall orientation of the decay in space and the second one the azimuthal orientation of particle 2 with respect to particle 1 (having defined the z axis along the direction of particle 1) uniform for spinless particles. Therefore in this case

$Display mathematics$
where with the chosen set of axis θ 2 is the angle between particles 1 and 2. Working in the centre of mass (but the same is true for any fixedP) $E 3 2 = m 3 2 + | p 1 | 2 + | p 2 | 2 − 2 | p 1 | | p 1 | cos ⁡ θ 2$ so that for fixed |p 1|, |p 2| one has E 3 dE 3 = –|p 1||p 2|d cos θ 2 and the integration over the angular variable is canceled by the energy delta function, resulting in

$Display mathematics$
which indicates a uniform bi-dimensional distribution in the plane (E 1, E 2).

The two independent variables describing the decay configuration can be chosen conveniently as the kinetic energies in the centre of mass frame $T i * = E i * − m i$ (p.504) (where the star indicates variables in this frame), or some linear combination of them: a standard choice (explicitly Lorentz-invariant) is169

(D.1)
$Display mathematics$
where m is one of the mi and

(D.2)
$Display mathematics$

(D.3)
$Display mathematics$
In case of equal masses mi = m the si range between (si)min = 4m 2 and (si)max = (Mmi)2.

A common application (for which this was invented) is to K → 3π decays: conventionally, in this case the subscript 3 is used for the ‘odd’ pion, i.e. the one with charge opposite to that of the decaying particle for K ±π ± π ± π , the charged one for K ±π 0 π 0 π ±, the neutral one for K 0π + π π 0.

Plotting two independent variables on a plane (the ‘Dalitz–Fabri plane’), each decay configuration is represented by a point, constrained by energy-momentum conservation to lie within a given boundary. Given (D.3) a convenient construction (Dalitz, 1953; Fabri, 1954) is that in which each si–(si)min represents the distance of the point from one side of an equilateral triangle (see Fig. D.1) of side $2 2 ( M 2 / 3 − 5 m 2 )$; the centre of the Dalitz plane corresponds to a decay configuration symmetric in the three particles, with si = s 0 and u = 0 = v.

In the non-relativistic limit (QMi mi → 0) the physical boundary is the circumference inscribed in the triangle; using relativistic kinematics (which is a minor correction in the case of K → 3π decays for which Q ≃ 80 MeV < mi) the actual physical boundary is distorted from a circumference, while still being tangential to the triangle at the centre of each side. The exact equation for the physical boundary is (in case of equal masses)

Fig. D.1. The Dalitz–Fabri plane; the circular non-relativistic boundary is shown.

(p.505)

(D.4)
$Display mathematics$
and the limits of the u, v ranges are (assuming m 1 = m 2).

$Display mathematics$
In the non-relativistic limit the boundary (D.4) reduces to

$Display mathematics$

All the points on the boundary correspond to collinear decay configurations, those on the sides of the triangle correspond to events in which one of the particles is at rest in the centre of mass frame, and those diametrically opposite to two particles with equal and parallel (minimum) momentum.

The Dalitz plane also has a uniform measure in the u, v variables: $d u d v = 8 ( M 2 / m π 4 ) T 1 * T 2 *$. Since the phase space measure element is proportional to the area in the plane, any departure from uniformity in the distribution reflects the nontrivial dynamics of the process; this is the usefulness of this representation, which factorizes the trivial phase space dependence. For example different spin-parity assignments for the decaying particle lead to different distributions in the (u, v) plane, and if e.g. particles 1 and 2 form a resonance an enhancement is expected for s 3 = (p 1 + p 2)2 corresponding to its mass.

The decay distribution in the Dalitz plot has traditionally been parametrized in a generic way with the first few terms of a series expansion around its centre:

(D.5)
$Display mathematics$
Note that if two of the final particles (i = 1, 2) are identical (as for K ±π ± π ± π 0, π 0 π 0 π ±) odd terms in v cannot appear, and in case of three identical particles (as for K 0 → 3π 0) g, j, f are not defined and a single quadratic term is present since u 2 = v 2/3. With particles i = 1, 2 being CP-conjugate (as for K 0π + π π 0) a non-zero value for the parameters j or f indicates the presence of CP violation.

The above parametrization is of course just an approximation, which works reasonably well for K → 3π in which the available phase space is small and no ππ resonances with mass below mK exist, although even in this case accurate measurement shows that the actual matrix element is more complicated and cannot be written in such a form (Batley et al., 2006 a).

Notes:

(169) Sometimes in the literature u and v are denoted y and x respectively.