The capabilities of the proposed array are well-suited to studies of the circumstellar environments of evolved stars and planetary nebulae (PNe). Mass loss from cool giant stars is believed to dominate the return of matter - both gas and dust - to the interstellar medium (ISM). In particular, carbon stars are a major source for the enrichment of carbon in the ISM, in elemental form and as carbonaceous grains. The transition from cool giants, especially those on the asymptotic giant branch (AGB), to the PN stage is an important aspect of stellar evolution which is controlled by stellar mass loss. Studies of this process with the proposed array will yield unique new data on the enrichment of the ISM by stellar ejecta, and on the late stages of stellar evolution.
The atoms, ions, and molecules accessible to the array include the most important species involved in the relevant physical and chemical processes. Recent theoretical studies of circumstellar matter in AGB stars, cool supergiants, and PNe and associated PDRs (e.g., Rodgers and Glassgold 1991; Hollenbach and Natta 1995; Natta and Hollenbach 1996; Justtanont et al. 1996, 1997) show the importance of atomic, ionic, and molecular emission lines at far-IR/submm wavelengths as diagnostics. Models such as these will be subject to direct observational tests with the proposed receiver array.
The fine structure lines of [CII] at 158 
m and of [OI] at 146
m will be invaluable in determining the total amount and
distribution of C and O being ejected by mass-losing red giants into
the ISM. Chemical models for the composition of cool giant winds
predict that all available C or O (whichever is less abundant), will
be locked up in the form of gas-phase CO close to the star. As this
material is ejected in a stellar wind, the CO is eventually
dissociated by interstellar UV photons which penetrate the dusty
envelope, yielding C and O in atomic form. The relatively low
ionization potential of neutral C ensures that virtually all gas phase
carbon ejected in the wind will emerge into the ISM as C 
. For
this reason, the 158 m [CII] line is an essential diagnostic of
the carbon returned both by carbon stars and by M giants. The 146
m line of [OI] should be an important diagnostic of the oxygen in
regions of high density, due to the excitation requirements of this
transition. It is likely to be most useful in the study of warm,
dense inner envelope regions, as well as of circumstellar gas affected
by chromospheric emission or by ambient UV radiation, e.g., in
the clumpy neutral shells detected around planetary nebulae.
Observations of the 146 m line of [OI] will be an important
complement to studies of the 63 m [OI] line (Rodgers and Glassgold
1991), which has been detected, for example, in the envelope of 
Ori with the KAO by Haas and Glassgold (1993). The array receiver
will be capable of detecting and mapping the distribution of this gas,
thereby providing critical indicators of mass loss in cool evolved
stars, and of the photochemical processes which convert the initially
molecular gas to its elemental atomic constituents. The imaging
ability of the array is particularly important in this respect, since
detailed photochemical models predict that C should exist in
extended shells around cool giants with large mass loss rates.
Neutral carbon has in fact been detected in such a shell around the
carbon star IRC+10216 by Keene et al. (1995) and in neutral globules
in the Helix Nebula (NGC 7293) by Young et al (1997). C is
predicted to exist in a layer exterior to the neutral C. Likewise,
the distribution of [OI] 146 m emission is a tracer of the
dissociation of the dominant O-bearing molecules including CO and
H 
O. The imaging ability of the receiver array in the fine
structure lines of [CII] and [OI] will indicate both the mass loss
history and the photochemical processes in the ejected gas.
In Figure 2 we present an optical image of the Helix Nebula overlayed
with an integrated intensity contour map of CO J 
(Young
1997). The size of the regions that can be mapped in a single SOFIA
flight by a 1, 4, and 16 channel version of STAR are
shown. Essentially, the entire region can be mapped within a single 8
hour period. The assumptions used in this calculation are listed in
the figure. Sample CI and CO spectra taken toward one extension of the
nebula are also shown in the Figure. As in the case of NGC 6334, the
line profiles easily fit within the 40 km/s velocity range available
in each array pixel. The observed lines have a number of interesting
features that can be further investigated with the high (0.125 km/s)
velocity resolution of the array AOS.
A third fine structure line which may be of interest is the 130 m
transition of [SiI]. Although much of the silicon ejected by evolved
stars is condensed into grains, a significant fraction is in the form
of gas-phase molecules, as the detection of SiO in M giants and SiS in
carbon stars demonstrates. As with CO, SiO and SiS compounds should be
dissociated in the outer circumstellar envelope, so silicon may be
detectable in the 130 m [SiI] line, by analogy with the
fine-structure lines of [CI] which have now been detected. Observations
of the 130 m [SiI] transition could provide strong constraints on
the chemistry and physical processes affecting the important refractory
element silicon. Complementary data can also be obtained from several
higher-J transitions of SiO and SiS accessible to the proposed
receiver. Both SiO and SiS are readily detected in M-type and carbon
stars, respectively, in their mm-wavelength transitions, so
observations at 
2 THz will provide additional useful data on
these important silicon-bearing molecules.
The CO molecule is, after H , the most important constituent of
molecular circumstellar envelopes. The rotational lines of CO and its
rarer isotopes-especially 
CO- which can be observed with the
array are relatively high in energy, and so are particularly valuable
as tracers of the more highly excited gas which is evidently produced
in the transition from red giant to planetary nebula. Recent
observations of CRL 618, CRL 2688, and NGC 7027 in high-J rotational
lines of CO with the KAO (Justtanont et al. 1997) revealed large
fluxes in the FIR lines in the 120 - 200 m wavelength
range. Justtanont et al. conclude that these post-AGB stars and young
PNe contain large amounts of warm, dense molecular gas, probably
ejected in a short-lived phase of very high mass loss. Both shock
heating and UV excitation are possible excitation mechanisms for their
observed FIR CO lines. Their models show that observations of CO and
atomic fine structure lines like [CII] 158 m, all observable with
the proposed instrument, should permit us to distinguish between the
effects of shocks and UV photons in exciting the molecular lines. The
improved sensitivity, spectral, and spatial resolution achievable with
the array receiver on SOFIA should therefore be uniquely important in
determining the physical state of, and the processes which affect, the
circumstellar gas during the important but poorly-understood
transition from red giant to planetary nebula.
Finally, the potential of the receiver array to observe both the 122
and 205 m lines of [NII] offers an important new tool for
determining the density distribution of the ionized gas in extended PNe
as well as H II regions. Rubin et al. (1994) have shown that the
205/122 m intensity ratio of [NII] is an excellent
temperature-independent diagnostic of electron densities in the
low-density regime of 1 cm 
cm 
. The
rapid mapping capability of the array in these lines of [NII] will
permit us to derive electron densities over extended objects, such as
the ionized gas of the Helix Nebula (which has an angular size of
12 arcminutes), for comparison with tracers of the
neutral/ionized interface regions, e.g., in the [CII] 158 m
line. With the large number of ionic, atomic, and molecular lines
accessible in the tunable range of the receiver array, and multi-pixel
mapping capability, the proposed instrument should give us a far more
complete understanding of the ionized and neutral components of PNe,
PDRs, and HII regions of all types.