Amongst the known nuclear species we expect to detect in E864 are d,
t, 
, 
, 
, and 
.
Since we are studying central rapidities, these are most likely
produced by coalescence rather than fragmentation, although this
will be a matter for detailed study. Coalescence models and
thermodynamic models both make testable predictions for the yields,
rapidity distributions, and transverse momentum distributions of these
products [53]. We note that the large
acceptance of the design means that the rapidity and transverse
momentum distributions are collected simultaneously.
If we take the production models outlined in above as a rough guide,
we estimate that for Au-Au collisions at 11.7 GeV/c per nucleon,
that the yield of is about 
per
central collision or about 
per interaction. Since
experiment E864 is designed for a sensitivity better than 
per interaction, the list of light nuclei noted above and heavier
systems as well should be within our range. Our above estimates
indicate that nuclei formed by coalescence up to A=14 or so should
be detectable by E864. The measurement of the production systematics
for different species should allow a wide variety of tests of various
models. We note that in almost any model, the yields of these
composites are sensitive to the phase space densities of the
``ingredients'' at the freezeout stage of the collision. That is, the
yields depend on the overlap in both momentum and configuration
space. Thus the kind of information gained is rather different from
that gained by measuring the spectra of ``elementary'' particles and
is similar in nature to that gained by Hanbury-Brown-Twiss analyses.
The BEVALAC experiments at 0.4 - 2 GeV/A taught us that the coalescence model is very successful in correlating the production cross sections of light nuclei in heavy ion collisions. Does the coalescence description apply to Au-Au collisions at 11.7 GeV/A? E864 will provide an answer to this question; if ``yes,'' the analysis of the data will enable us to extract coalescence probabilities. In the context of the thermodynamic model, we can then extract freezeout temperatures, and compare these with values obtained in the BEVALAC energy regime. We anticipate that coalescence probabilities will decrease with increasing energy (temperature). This remains to be established. The results would stimulate the further development of a microscopic theory of the coalescence probability, which takes account of the relativistic nature of the ``fireball'' expansion in high energy heavy ion collisions [43].
The abundances of light nuclei produced in Au-Au collisions are interesting in their own right as tests of dynamical models of heavy ion reactions. In addition, the rates for the production of light nuclei serve as an important calibration for the strangelet search. In the coalescence picture developed above, the production rates for multiply strange composite objects are related to the yield for the non-strange cluster of the same baryon number A. Thus, at least within a coalescence framework, we can attach a significance to the absence of a strangelet signal in E864, if this is indeed the case.
An interesting question is the following: Are there any
manifestations, in the yields of light nuclei, of the intermediate
state of hot, dense matter? Even in the absence of a
quark-gluon plasma, the density and temperature prevailing during
this transient intermediate state are expected to be higher at AGS
than at BEVALAC energies. In the thermal model, rates for nuclear
clusters depend only on the 
values at freezeout, i.e., a
rather late stage in the reaction process. It is important to provide
data to test this picture in detail at AGS energies: this is one of
the bread-and-butter tasks of E864. The main discovery
potential of E864 lies in the search for strangelets and other
unanticipated new particles. However, one must not lose sight of the
fact that E864 will definitely measure the abundances of a number of
light nuclei and antinuclei which are known to be stable against
strong decay, providing strong constraints on models of collision
dynamics. Finally, we mention that E864 can also explore the edge of
the region of stability for light nuclear isotopes, particularly the
neutron rich regime. For example, 
, ,
are known to be stable against strong neutron
emission, but 
is thought to be unstable against
strong decay. Searches for in lower energy heavy
ion reactions were unsuccessful [54],
lending support to this conclusion. Our rough
estimates give 
, so
would be measurable by E864, if it were stable.
Among Z=3 isotopes, 
is known to be stable against
strong decay (lifetime 8.5 ms), but 
are
not. All stable Li isotopes with 
should be measurable in
E864; the relative abundances of 
will
provide a test of the notion of a constant penalty factor P for the
addition of a neutron to a cluster, which was a key assumption in our
estimates of strangelet formation rates.