In the previous section, we have discussed the implicit role played
by vector mesons in chiral
lagrangians, through saturation of the counterterm contributions. Beyond
the very low
energy region, where a lowest-order power expansion in
is sufficient, one may
try to extend the validity of chiral lagrangians through the
explicit introduction of vector meson fields. Two main lines
have been developed, the massive Yang Mills (YM) approach proposed by
Meissner
[1] among other authors, and the Hidden Symmetry (HS) scheme
proposed by Bando et al.
[2]. In the former, spin-1 fields are introduced in the
ungauged lagrangian
through the usual
covariant derivatives. Restricting, for simplicity of exposition, to the
sole SU(2) fields,
i.e. pions and
mesons, one starts with the usual lowest order
lagrangian
and then makes the substitution
where now contains only
the pion field
and
is
the
meson
field. This results in the lowest order lagrangian
where, for the sake of the present argument, we have shown only some terms
involving
interactions between neutral mesons and pions. The gauge
coupling constant
g is determined by neutral
decay,
,
to have the value
. In the massive Yang Mills scheme, photonic
interactions are then introduced through conventional Vector Meson Dominance,
i.e.
through the lagrangian
which mediates all transitions between photons and matter fields. Applying
this
rule to the decay , one obtains the relation
and
. The experimental values
MeV and
KeV are then fitted with
.
The dots in eq.(17) refer to further
quadratic interactions between
and an even number of pion
fields, but no mass terms
for the vector mesons. This is to be contrasted with what happens in the
lagrangian proposed by Bando et al.[2], in which the vector mesons
enter explicitly as gauge fields of a ``hidden'' local symmetry of the
chiral lagrangian, whereas the electro-weak gauge bosons are explicitly
introduced through the usual covariant derivatives. A Higgs-like
mechanism then generates the vector meson masses.
To wit, in the ``hidden symmetry approach'' of Bando et al.[2]
the most general lagrangian containing
pseudoscalar, vector and
(external) electroweak gauge fields with the smallest
number of derivatives, is given at the lowest order,
by the linear combination
, a being an arbitrary parameter, of the
two SU(3) symmetric lagrangians
The matrices and
contain the
pseudoscalars fields, P, and the unphysical (or compensator)
scalar fields,
, that will be absorbed to give a mass to the
vector mesons
The full covariant derivative is
where only the photon field, , (but not
its weak partners, as before)
has been explicitly shown, and P and V now stand for the
SU(3) octet and nonet matrices
The lagrangian can be reduced to the
chiral lagrangian (1) for any value of the parameter a [2].
In fact, working in the unitary
gauge (
) to eliminate the unphysical scalar fields and
substituting the
solution of the equation of motion for
, the
part vanishes
and
becomes identical to the non linear chiral lagrangian
(1).
The ``hidden symmetry'' lagrangian
(19) can
be easily seen to contain, among other things, a vector meson mass
term, the pseudoscalar weak decay constants, the vector-photon
conversion
factor and the couplings of both vectors and photons to pseudoscalar
pairs. The latter can be eliminated fixing a = 2, thus incorporating
conventional vector-dominance in the electromagnetic form-factors
of pseudoscalars. Returning, as before,
to the simpler SU(2) case, for
one has
In the above equation, the constant a has been
fixed to the value a=2 in order to
recover the relation between and the
couplings, previously
discussed. This lagrangian obviously
reproduces all the lowest order results, as the previous one does,
however it does not
contain couplings between two or more
- mesons and an even number of
pion fields.
It is not yet clear if these two lagrangians are fully equivalent or not.
At first
sight, processes involving one meson and four or more pion fields could
be used to discriminate between the two [9], since these
processes could proceed
through multiple
-pion interactions in the YM scheme, while not so
in the HS approach. However, it has been pointed out
that axial vector terms, not
yet introduced in the above lagrangians, would probably play a role. This
might render experimental discrimination rather difficult [10].
To the two schemes just discussed, one must also add the simplest and oldest of
them all, i.e. conventional Vector Meson Dominance
[11], in which
all photon-pseudoscalar interactions proceed through vector meson fields with
the same interaction lagrangian seen before and the
vector meson fields
are introduced through the covariant derivatives
in the free pion lagrangian
. The lagrangian thus obtained
contains the same vertices as the previous two, but there
are of course
no multipion interactions, which in conventional VMD are basically frozen into
the role played by the vector mesons.
After this brief introduction, we shall now discuss some vector meson
decay processes
which can be measured at DANE , and in which the validity of the theoretical
scenarios we have illustrated can be tested.