The schematic of the polarized target assembly is shown in Fig.3. It consists of a superconducting dipole which operates at 5 Tesla and the target refrigerator which operates at a temperature of 1K. The magnet is a split Helmholtz coil which produces a longitudinal field. A bore of 20 cm and an opening angle of 100 degrees together with a coil split of 8 cm and a 34 degree opening angle between the coils allow for great flexibility in field orientation and particle detection.
Figure 3 also shows the 
He evaporation refrigerator
necessary to cool the target to a temperature of 1K. Such a
refrigerator is the best choice for the beam conditions of this
experiment in which efficient heat removal is required. Large
pumps mean that operation close to 1K, and therefore high
polarizations, is possible. Using the pumps on hand 1.5 W of
heat can be removed.
The target material is doped, by irradiation or chemical
means, with paramagnetic centers (unpaired electrons) which at
1K and 5T will be nearly 100% polarized. Irradiation with
microwaves (140 GHz at 5T) transfers most of the electron
polarization to the protons in the material. The transfer
mechanism is less effective with deuterons. By this method of
dynamic nuclear polarization [14]
protons have been polarized to
almost 100% and deuterons to 50%.
The material of choice for this experiment is ammonia, NH 
and
ND .
In experiments with ionizing particles, radiation damage to
the material causes the polarization to fall. The polarization,
P, after an incident flux, I, is given by:
. For most materials
particles/cm 
[15]. For ammonia 
particles/cm [16].
Figure 4: Cross section view of the polarized target
The polarization can be largely recovered by annealing, i.e.
warming the target to a characteristic temperature, usually around
100K. But in most materials there is a residual, non-annealable
radiation damage component. This means that the target material
must be changed after a few anneals ( 
)
because the achievable
polarization falls to an unacceptably low level.
Ammonia, however, can be repeatedly annealed and the
polarization completely recovered.
The paramagnetic doping for ammonia must be done by
irradiation rather than the more usual chemical methods. The
simplest method is to irradiate the ammonia in its frozen state,
under liquid argon, with a low energy electron beam (20 - 300
MeV), but other methods have been used. The best results have
been obtained with 
e 
/cm
incident on the target
[16, 17].
However it has been observed that 'in situ' irradiation
[16, 17] enhances the polarizeability of the ammonia,
particularly in the case of ND .
In order to spread the beam over the target to get uniform heat deposition, the Hall C
rastering system is being designed [22] such as to allow the beam to be rastered
over the full area of the target, 2cm 
2cm. The large excursions from the nominal
beam axis require in particular, that the angle change produced by the upstream rastering
system is compensated near the target with a second rastering magnet, such that the beam
continues after the target in direction of the beam dump.
The polarized assembly for this experiment was set up at the University of Virginia and became completely operational in August 1992. It was operated through February 1993 with many target material investigations being carried out. All sub-systems and components were tested over this period. The magnet and refrigerator were kept cold, at 4K and 10K respectively, so that the magnet could be ramped up on demand and at the same time the refrigerator could be made to operate at 1K. It took 45 minutes for the magnet to reach 5T and polarization tests could be started immediately afterwards.
At UVA, protons in NH were polarized to 95%, while the
deuterons in ND reached 13% polarization, using ammonia
irradiated at the Saskatoon Accelerator Lab. Subsequent
irradiations at CEBAF have indicated that the deuteron
polarization is very sensitive to the number of paramagnetic
centers produced, i.e. dose. Work is continuing to find the
optimum dose, and then with further 'in situ' irradiation a
deuteron polarization of 40 - 45% is expected. We base these expectations on
work done with the Yale target at SLAC some 15 years ago, where they reached a deuteron
polarization of 27% despite having modest cooling power and reaching only 75% for the
polarization of NH .
The polarized target work at Bonn in 1985 achieved 30% deuteron polarization at 2.5 T,
with only 70% polarization for NH . In a different configuration the Bonn group
actually did achieve 45% polarization for ND .
Two microwave tubes were tested, both EIOs, one operating at 140 GHz, the other at 136 GHz. Each tube was used to polarize ammonia and delivered up to 500 mW to the target volume. The NMR system was developed to work with both a signal generator and a frequency synthesizer. The critical performance level was the observation and measurement of the deuteron thermal equilibrium signal. We have achieved a signal of similar quality to that achieved in the SMC experiment at CERN [21] although in our case there was no temperature stabilization of the critical components. This stabilization will be incorporated in future set-ups. In March 1993 everything was dismantled and shipped to SLAC for experiment E143.
At SLAC, as for this G
experiment, ammonia with 
N will be
the target material of choice. Because the nitrogen is also
polarized, 
and 
are preferred since the polarization
of the N is carried by the proton rather than the proton and the
neutron as in the case of 
N. The degree of polarization is the
same in either case but much easier to measure for N. It has
also been shown [18]
that the radiation performance is the same
for N and N ammonia.
With the absence of a polarized neutron background from
,
the systematic error will be substantially reduced in the
measurement of G. and
will be loaded into two
separate target containers in the refrigerator and a simple
mechanism will put either target into the beam on demand.
The operation of the target at SLAC will set the performance standard for operation at CEBAF as the requirements are very similar. Table 2 shows the target parameters.
Obviously this target is a general purpose facility for CEBAF, useful for many other
experiments, including the Deformation of the 
proposed by this
collaboration. However, the primary motivation for undertaking the construction of the
target is to measure 
.
