At the moment of formation, molecular hydrogen can join one of two denominations: ortho (aligned nuclear spins) or para (opposed nuclear spins). The ortho-para ratio, , based on the statistical weights of the nuclear spins is 3:1. This is the expected membership ratio when the H2 has formed on grains at high temperatures and has avoided subsequent conversion. The ratio can be lowered through conversion on the grains (eg. 1.8, Black & van Dishoeck 1987) or even increased through conversion in the gas phase: thermal collisions with protons or H atoms at low densities yield an equilibrium ratio of (eg. Flower & Watt 1984, Martin, Schwarz & Mandy 1996). Already-published data for the three lines exist for a few single pointings in shock-excited objects. These yield values close to 3:1 (see Sect. 3). However, in photon-dominated regions ratios below 2:1 have been measured (Hasegawa et al 1987, Ramsay et al 1993), consistent with conditions of thermodynamic equilibrium below temperatures of . Hence, can place constraints on the constituents and processes of importance in molecular clouds. Here, we present a method for calculating and apply it to HH objects. We then present the distribution of in the OMC-1 bipolar outflow.
The main spectroscopic data obtained here are Fabry-Perot images (from IRAC2 on the ESO 2.2m telescope) in the 1-0 S(0), S(1) and S(2) transitions for the OMC-1 outflow. Also obtained were 2-1 S(1) and 2-1 S(3) images though generally only the brighter nebulous regions near the outflow source were detected. The energy levels of the 1-0 S(0), S(1) and S(2) lines at 6471 K, 6951 K & 7584 K, lie close together - a disadvantage for excitation studies. However, this is balanced by the advantage that the lines are all strong and accurately measurable. Also, a single excitation temperature can be assumed to describe the gas (i.e. the same gas produces all three lines, see Sect. 3). This means that the second piece of information inherent in the two line ratios, , can be accurately extracted. Moreover, we show in Sect. 3 that the small wavelength separations and the particular order in the spectrum combine to all but exclude differential extinction as a confusing factor.
Recent developments in near infrared detector technology have made mapping of OMC-1 in different lines possible. Many H2 lines fall in the K-band and a thorough study of those lines can yield considerable information about the physical conditions in these regions. The 1-0 S(2) line has not been imaged before. It is in a spectral region of relatively poor atmospheric transparency and with the 1-0 S(1) always stronger, there has been no desire to map it. Nevertheless, concurrent equivalent data have been obtained via long-slit spectroscopy by McCaughrean and Mac Low (1997)).
The OMC-1 outflow is situated in one of the best studied star-forming regions (Genzel & Stutzki 1989). It is a prodigious outflow from a high mass protostar, very bright in the infrared lines of molecular hydrogen, although enshrouded in a dense part of the Orion Molecular Cloud (Beckwith et al. 1978). It is hoped that by understanding how the outflow was generated, we can come to terms with this critical star formation phase. Recent infrared and theoretical studies have concentrated on two problem areas: (1) the dynamics which could produce the 'bullets' (see below) and (2) the physical processes which excite and accelerate the H2 without destroying it (Smith 1991; Draine & McKee 1993). Here we employ new near infrared data to explore the basic issues of how the H2 is formed, modified and excited.
On the dynamical side, Allen & Burton (1992) identified a series of bow shocks in the region north west of IRc2 (Axon & Taylor 1984) as parts of finger-like structures, and claimed that they are due to the ejection of bullets from the center of the exciting region. McCaughrean & Mac Low (1997) have identified a long finger to the south-east of IRc2. Stone et al. (1995) interpreted these data in terms of a fragmenting wind. Another debate concerns the innermost region, close to IRc2. Recent high resolution velocity mapping by Sugai et al. (1995) suggest that a thin shell is interacting with the surrounding cold and static molecular gas.
Near-infrared spectroscopy of OMC-1 has revealed two disturbing facts. First, the molecules appear to be moving extremely quickly, since H2 line widths approaching km s-1 have been measured (Brand et al. ; Tedds et al 1994). There is considerable evidence that the gas in OMC-1 is collisionally excited within shock fronts (Brand et al 1988). However, standard magnetohydrodynamic shocks can only accelerate molecules to speeds of km s-1 (J-type shock) or km s-1 (C-type shock, see Draine & McKee 1993). A possible solution lies in the presence of low Alfven-speed 'shock absorbers' (Smith et al. 1991b) in which the magnetic field ahead of the shock is pre-compressed before being overtaken by the shock. This mechanism permits H2 excitation (without dissociation) in faster shocks, and so the acceleration to higher velocities.
Secondly, the H2 excitation in OMC-1, as measured by the temperature of gas required to produce a given ratio of line strengths, is strangely constant in space (Brand et al 1989a). However, Smith (1991a) has shown how a large number of arcsecond-sized curved shocks, or mini-bow shocks, could produce the uniform line ratios. Indeed Smith is also able to account for the observed dependence of excitation temperature on the upper energy levels of the particular transitions. Other possible explanations for this apparently uniform excitation distribution include the shock-dissipation of supersonic turbulence (Davis & Smith 1996) and 'dry J-shocks' (Brand et al 1988), although recent ISO observations of abundant H2 O in some outflows (Liseau et al 1996) has cast doubt on the validity of this second model.
In this paper, we now take the OMC-1 puzzle one stage further by posing the question: which shocks are consistent with the ortho-H2 /para-H2 fractions?
Excitation images may be used to illustrate the spatial distributions of rotational and vibrational excitation temperatures. This has previously been achieved for L 1448 (Davis & Smith 1995) and Cepheus E (Eislöffel et al 1996). Along with the strong 1-0 S(1) line, lines from higher vibrational levels allow the excitation temperature to be derived. This is despite the relative weakness of the measured 2-1 S(1) and 3-2 S(3) lines. The neighbouring lines discussed in this paper do not in principle allow the excitation temperature to be accurately derived. However, as will be seen, the constant ortho-para ratio derived from the same data implies that the excitation temperature can indeed be reliably extracted. Image display of such ratio distributions is hampered by the inevitable inaccuracies in regions of low intensity (for intensity images this is of course no problem). Consequently, here we will display the excitation and ortho-para ratio along cuts through the outflow region.
In Sect. 2 we summarize our observations. In Sect. 3 we present the general method for determining the o/p ratio and the excitation temperature, which we then test with existing wide-beam data. In Sect. 4, we present the ortho-para ratio and excitation distributions for OMC-1 and discuss the implications in Sect. 5.
© European Southern Observatory (ESO) 1997
Online publication: April 6, 1998