IGRINS NEAR-IR HIGH-RESOLUTION SPECTROSCOPY OF MULTIPLE JETS AROUND LkHα 234^*

In Figure 2, the fluxes are integrated over knot A (−65 < Y < −05). The [Fe ii] lines at SP 1 and 2 show similar profiles with a peak velocity (v_peak) of ∼ −110 km s⁻¹ and a shoulder toward lower velocity. At SP 3, only the [Fe ii] λ1.644 μm line is detected. All H₂ lines show similar profiles at each slit position. At SP 1 and 3, the H₂ emission lines have asymmetric line profiles. At SP 1, the profiles peak at V_LSR ∼ −11 km s⁻¹ and show blueshifted wings. The profiles at SP 3 are asymmetric relative to the peak velocity (V_LSR ∼ −13 km s⁻¹), showing a gradual slope on the red side of the line. At SP 2, they show double peaks at V_LSR = −30 and −5 km s⁻¹.

The [Fe ii] and H₂ emission profiles differ significantly: H₂ emission lines are strong at low velocity, close to the systemic velocity of V_LSR = −10.2 km s⁻¹ (Liu et al. 2011), while [Fe ii] emission lines are prominent at higher velocity. In the following section, we compare [Fe ii] λ1.644 μm and H₂ 1−0 S(1) λ2.122 μm emission lines at each slit position in detail.

3.2.1. Slit Position 1 (SP 1)

3.2.2. Slit Position 2 (SP 2)

3.2.3. Slit Position 3 (SP 3)

Figure 5 shows the PVDs of all detected [Fe ii] emission from SP 1, SP 2, and SP 3. It includes the λ1.533, λ1.600, λ1.644, λ1.664, and λ1.677 μm lines. All five lines are detected at SP 1 and 2, while SP 3 only shows emission in the λ1.644 line. In SP 1 and 2, the morphologies and velocities are almost the same in all of the lines in each slit position. One exception is in SP 1, where the highest velocity peak at V_LSR ∼ −130 km s⁻¹, Y = −68 and weak emission at Y = −94 and −99 are only detected in λ1.644 μm. Measured line fluxes normalized to the [Fe ii] λ1.644 μm line are listed in Table 1. Since the positions and velocities of peaks in all [Fe ii] lines are consistent, the fluxes are estimated within the regions at Y ∼ 47 ± 04, V_LSR ∼ −113 ± 10 km s⁻¹ and at Y ∼ 43 ± 04, V_LSR ∼ −112 ± 10 km s⁻¹ in SP 1 and SP 2, respectively. The line fluxes in SP 1 and SP 2 are very similar.

Zoom In Zoom Out Reset image size

Table 1.  [Fe ii] Line Fluxes Normalized to the 1.644 μm Flux

Line	λ(μm)	SP 1	SP 2
${a}^{4}{D}_{5/2}-{a}^{4}{F}_{9/2}$	1.53389	0.12 ± 0.05	0.14 ± 0.06
${a}^{4}{D}_{3/2}-{a}^{4}{F}_{7/2}$	1.59991	0.10 ± 0.04	0.11 ± 0.04
${a}^{4}{D}_{7/2}-{a}^{4}{F}_{9/2}$	1.64400	1.00 ± 0.29	1.00 ± 0.18
${a}^{4}{D}_{1/2}-{a}^{4}{F}_{5/2}$	1.66422	0.09 ± 0.03	0.10 ± 0.04
${a}^{4}{D}_{5/2}-{a}^{4}{F}_{7/2}$	1.67733	0.17 ± 0.04	0.18 ± 0.06

Note. The reddening is not corrected.

Download table as:  ASCIITypeset image

Near-IR [Fe ii] emission lines can be used as effective tracers of electron density (n_e) in stellar outflows (Nisini et al. 2002; Pesenti et al. 2003; Takami et al. 2006). Nisini et al. (2002) developed a non-LTE model applicable to jets from YSOs, considering the first 16 fine structure levels of the Fe⁺ ion. The flux ratios between lines obtained in our observations are relatively independent of gas temperature, because the lines are all from the fine structure level of the ${a}^{4}D$ term, having excitation energies in the range of 10,000–12,000 K (Nisini et al. 2002). To investigate the electron density, we show the line ratios superposed on the model grids from Nisini et al. (2002) in Figure 6. In SP 1 and 2, the electron densities (log n_e) averaged from the three ratios of λ1.600, λ1.533, and λ1.677 μm lines against the λ1.644 μm line are ${4.0}_{-\;0.6}^{+\;0.7}$ and ${4.1}_{-\;0.6}^{+\;0.9}$, respectively. The uncertainty in the λ1.644 μm/λ1.533 μm ratio is larger (Figure 6), due to residuals after a subtraction of the OH sky emission in the λ1.533 μm line. The n_e of knot A we obtained from the optical [S ii] doublet ratio (Osterbrock 1989) is about 0.3 × 10⁴ cm⁻³, which is smaller than that estimated from near-IR ion emission lines here. This result is consistent with the tendency in Nisini et al. (2002). It indicates that the [Fe ii] lines are able to probe a denser region in the jet because they have a higher critical density (∼10⁴ cm⁻³) than that of the [S ii] (∼10³ cm⁻³).

Zoom In Zoom Out Reset image size

The electron densities of 0.5–3.2 × 10⁴ cm⁻³ obtained here are similar to or smaller than those estimated from [Fe ii] emission from T Tauri stars and intermediate-mass stars in Hamann et al. (1994). In addition, these are also slightly lower than values from HH outflows from Class 0−I, which range from 10⁴ to 10⁵ cm⁻³ (Nisini et al. 2002; Takami et al. 2006).

Table 2 lists the line fluxes of all 14 emission lines detected in these observations. Fluxes are estimated at each peak position in Figure 3, and normalized to the H₂ 1−0 S(1) λ2.122 μm line. Figure 7 shows the PVDs of eight selected H₂ emission lines. In all slit positions, the different H₂ emission lines show similar velocity features, as different [Fe ii] lines do. In all H₂ lines at SP 1, 2, and 3, we detected emission from both knot A and knot B, which are bright at low velocity. The three high-velocity components at Y = −68, −94, and −102 with v_peak of −130, −13, and −118 km s⁻¹ shown in λ2.122 μm are also detected in λ2.034, λ2.223, λ2.407, and λ2.424 μm lines. In λ2.407 μm at SP 2, we detected a weak (5σ level) emission that is consistent with a peak at Y = −47 with −110 km s⁻¹ in λ2.122 μm.

Zoom In Zoom Out Reset image size

Figure 8 shows H₂ population diagrams calculated from the fluxes in Table 2. The level populations of hydrogen molecular gas estimated from ro-vibrational transitions allow us to study the excitation state and temperature of the gas (Black & Dalgarno 1976; Gautier et al. 1976; Beckwith et al. 1978; Hasegawa et al. 1987). We calculated the upper-level column density from the observed line flux after a reddening correction. The extinction can be estimated using the ratios of the emission lines originating from the same upper levels. In our calculation, we used the ratios of three pairs of 1−0 S(1)/1−0 Q(3), 1−0 S(0)/1−0 Q(2), and 1−0 S(2)/1−0 Q(4). We adopted the transition probabilities from Turner et al. (1977). We apply the extinction law, ${A}_{\lambda }$ = ${A}_{{\rm{V}}}$(0.55μm/λ)${}^{1.6}$ (Rieke & Lebofsky 1985). In our estimation, ${A}_{{\rm{H}}}$ and ${A}_{{\rm{K}}}$ are very small (<0.1). We used ${A}_{{\rm{V}}}$ = 3.4, which is an intermediate from the range of 3.1–5.1 used in previous studies (Hillenbrand et al. 1992; Hernández et al. 2004; Liu et al. 2011).

Zoom In Zoom Out Reset image size

For the data in Figure 8, we applied weighted least-squares fitting to estimate rotational temperatures (T_rot) from an upper level of v = 1 (circle) and v = 2 (square), considering 1/${\sigma }^{2}$ for weights, where σ is the uncertainty in the level population. We excluded one transition from v = 3 (triangle) from the fit due to its large uncertainty. Solid and dashed lines represent the fitting lines for v = 1 and 2, respectively.

In SP 1, T_rot of all peaks estimated in v = 1 is 2500 ± 200 K. The temperatures in v = 2 show uncertainty because we have only three data points for the fitting. T_rot of peaks A1 and B in v = 2 are similar, 3200 ± 1500 K. In A2, T_rot in v = 2 is 2500 ± 1100 K, which is close to the value from v = 1. In SP 2, T_rot in A1' is almost the same as in A1, in both v = 1 and 2. Peak A2' shows a higher T_rot than A2 in v = 1 and 2, but the differences are within the uncertainties. In SP 3, T_rot from A1'' and A2'' in v = 1 are higher than those of SP 1 and 2; with T_rot > 2700 ± 250 K. In v = 2, T_rot is 3000 ± 1500 K for both A1'' and A2''. From SP 1 to 3, T_rot in v = 1 shows a small increase, but within uncertainties. For v = 2, there is no clear trend with change in slit position. In all three slit positions, in both v = 1 and 2 the derived T_rot is ∼2500–3000 K. This means that the gases are thermalized at a single ro-vibrational temperature, because both level populations follow the same T_rot. We note that the temperature estimation differs when we consider different ${A}_{{\rm{V}}}$ in the flux correction. It becomes ∼200 K lower and less than 100 K higher for the cases of ${A}_{{\rm{V}}}$ = 0 and 5.1, respectively, compared to the estimation using ${A}_{{\rm{V}}}$ = 3.4. We also note that the populations of ortho and para transitions are aligned in a single line. This indicates that the emission lines are not from the fluorescent H₂ by UV excitation (e.g., Hasegawa et al. 1987).

Thermalization has been observed in the shocked gas in the outflows from low- and high-mass stars (e.g., HH 111, HH 240 and 241 in Nisini et al. 2002, Orion KL peak1 in Beckwith et al. 1983). We can test the properties of a shock with a population diagram by comparison with model data. Rosenthal et al. (2000) have investigated various shock models to prove a shock nature in Orion Peak 1, and they showed that the combining of shock models could explain their observational results except at the highest excitation level. Since we assume the presence of a bow shock in this region (see Section 4), we plotted the bow shock model with v_shock = 100 km s⁻¹ from Smith (1991) in Figure 8 (dotted lines). This model is more consistent with fitting lines from v = 2 than those from v = 1. On the other hand, it is hard to discriminate the type of shock from our result because there is no significant difference between various shock models in the energy levels of 0.5–2 × 10⁴ K (Rosenthal et al. 2000).

In Table 3, we list the H₂ line ratios at all peak positions that are sensitive to the type of shock. The ratios in the models of C- and J-type shocks and UV pumping are taken from Beck et al. (2008). The ratios in our observation are similar to those of outflows from T Tauri stars (Beck et al. 2008). The 2−1 S(1)/1−0 S(1) ratio is close to a C-type shock in all cases, indicating similar shock conditions along the outflow. The other three ratios do not help discriminate between different shocks, because of the large uncertainties. There are no significant differences at all three slit positions. We note that the 1−0 S(1)/1−0 Q(1) ratios in our data are about 1.5–2.5 times higher than ratios from all models. 1−0 Q(1) λ2.407 μm emission is affected by a deep telluric absorption at ∼ −40km s⁻¹ ("⊕" symbols in Figure 7), so it may not be reliable.

As shown in Section 3, the H₂ 1−0 S(1) λ2.122 μm and [Fe ii] λ1.644 μm emission lines have very different velocity structures. The H₂ emission is dominant at low velocity while [Fe ii] emission dominates at high velocity. This trend was found in outflows from both classical T Tauri stars and Class 0−I YSOs (e.g., Pyo et al. 2002; Davis et al. 2003; Takami et al. 2006). The trend makes sense in the context of an interpretation where the shocked H₂ emission arises from gas entrained by a high-velocity outflow (Pyo et al. 2002).

In our PVDs, we detected H₂ emission at the systemic velocity along the whole slit, in all three slit positions. This emission probably originates from ambient molecular hydrogen gas in this region, which corresponds to the presence of far-UV radiation by producing a photodissociation region (PDR) (Morris et al. 2004). However, taking account of the H₂ lines ratios (Table 3), we should not rule out excitation by shock as the emission source. We note that, in this case, the velocity width at the top of the PVDs should be larger than the current value (∼10 km s⁻¹), because the angle of the outflow axis may be less than $45^\circ$ with respect to the line of sight according to our interpretation in bow shock features described below.

Discrete multiple velocity components revealed from the H₂ emission in Figure 3 are the typical high- and low-velocity features that arise from the bow shock in outflows (Hartigan et al. 1987; Davis et al. 2003; O'Connell et al. 2004). If the inclination angle between the axis of the bow shock and the line of sight is in a range of ∼$0^\circ \mbox{--}45^\circ$, the emissions from the apex and the wing of the bow are seen as separate high- and low-velocity peaks. The weak H₂ and strong [Fe ii] emissions at high velocity (V_LSR ∼ −120 km s⁻¹) show good agreement in position and velocity, indicating that the [Fe ii] emission lines usually arise from the bow apex, where the jet is faster than the surrounding region (Hartigan et al. 1987; Davis et al. 2003). The velocity difference of ∼100 km s⁻¹ between the high- and low-velocity peaks in the H₂ lines is similar to that observed in the jet of L1551-IRS 5 (Davis et al. 2003), which has an inclination angle of ∼ $45^\circ$ with respect to the line of sight. In knot B at SP 1, we see three discrete velocity peaks in H₂ where the velocity of the peak increases with increasing $| Y|$. The more blueshifted peaks lie farther from the star. The total offset along the outflow is ∼115, after consideration of the different angles of our slit (256°) and the outflow axis (227°). One possible interpretation of this position shift is a spatial difference between the faster gas in the bow apex and the slower portion in the bow wing. We cannot rule out, however, the possibility that this shift is caused by distinct knots originally located at different positions.

In contrast to the blueshifted gas seen at all slit positions, the redshifted component (V_LSR > −10.2 km s⁻¹) seen in H₂ at SP 2 and 3 seems to arise from another, spatially unresolved outflow. Since the slit positions are pointing at the blue lobe of the jet (see Sections 1 and 2), at locations where the blueshifted material is well separated in phase space from any ambient emission, the detection of redshifted emission implies the existence of a distinct outflow, separated from the high-velocity blueshifted jet. This outflow may have an axis almost parallel to the plane of the sky, with a wide outflow opening angle, because it shows both blue and redshifted components at low velocity (−50 km s⁻¹ < V_LSR < +50 km s⁻¹). For a further discussion, we need additional clues to the position and orientation of the redshifted component, which would need to come from further observations with high spectral resolution and with higher spatial resolution than we can obtain with our present instrument/telescope combination.

As shown in Section 1 and Figure 1, the LkHα 234 region is complicated with several outflow sources observed at mid-IR, millimeter, and radio wavelengths. In this section, we summarize three separate outflows and their source candidates, and relate them to our own spectroscopic observations.

1. The [Fe ii] jet from VLA 3 (NW 1, IRS 6): At SP 1 and 2, we detected the bright [Fe ii] emission with a velocity of ∼ −120 km s⁻¹ at the position of knot A (Figure 3). Here, we suggest that this [Fe ii] emission possibly originates from the outflow driven by radio source VLA 3B, according to following evidence.

The PVDs at SP 1 and 2 show that the [Fe ii] knot lies between Y = −2'' and Y = −6''. As shown in Figure 1, its position agrees well with the "green blob" feature seen in the center of the C-shaped reflection nebula in the JHK color image of Kato et al. (2011), a feature that is probably caused by [Fe ii] emission in the H-band. The narrow-band [Fe ii] λ1.644 μm image in Schultz et al. (1995), which showed emission at the same position, supports this conclusion. The configuration of an [Fe ii] jet detected along the central axis of a cavity in a reflection nebula is very similar to the case of L1551-IRS 5 (Hayashi & Pyo 2009). Kato et al. (2011) showed two jet-like features in their J- and H-band images and they mentioned that those features point at YSO candidate G. They suggested that the jet-like feature containing the green blob is associated with the optical [S ii] jet.

We suggest another possible origin of the [Fe ii] emission in Figure 3, VLA 3B. The 3.6 cm continuum map of Trinidad et al. (2004) showed that VLA 3B has an elongated thermal radio jet extending northeast–southwest. Their source-subtracted 3.6 cm map is shown in Figure 1(d). We estimated a P.A. of ∼$230^\circ$ for this jet. "Axis3" in Figure 1 denotes the axis of the radio jet in VLA 3B. This line passes through the positions of the [Fe ii] knot detected in our observations and the green blob in the JHK image. The very similar major-axis position angles of the [Fe ii] emission in Schultz et al. (1995) and the radio thermal jet also support our suggestion, although the reliability of their [Fe ii] image close to LkHα 234 should be a cause for caution. Taking all factors into account, however, we conclude that this [Fe ii] emission is from a jet driven by VLA 3B, with a velocity of V_LSR ∼ −120 km s⁻¹. Trinidad et al. (2004) suggested that VLA 3B could be a low-mass YSO, while Kato et al. (2011) suggested NW 1 as a B6–B7-type YSO from the spectral energy distribution. An outflow cavity is opening toward the southwest from VLA 3B. The mid-IR source IRS 6 lies in the cavity with a ∼05 offset from VLA 3B. The offset implies that the mid-IR emission comes from a region less affected by dust extinction.

We note that the P.A. of the [Fe ii] jet is very similar to that of the H₂ jet (227°), as shown in Figure 1. This may be an additional example of multiple jets sharing a similar orientation (e.g., Nisini et al. 2001), as pointed out by Trinidad et al. (2004).

2. The H₂ jet from FIRS1-MM1: The most likely source of the H₂ jet (Cabrit et al. 1997) and inner [S ii] jet (Ray et al. 1990) is FIRS1-MM1 (Fuente et al. 2001), which is indicated as a white pentagon in Figure 1. More accurately, it seems to drive knots B and C, because the position of FIRS1-MM1 is well aligned with the line through those two knots (P.A. ∼$227^\circ$), which is marked as "Axis2" in Figure 1. In Figure 3 at SP 1, the H₂ emission shows a peak at Y = −92. This peak position shows that SP 1 is superposed on the center position of knot B of the H₂ jet, as we confirm in Figure 1. We used our high-resolution observation to reveal, for the first time, that knot B shows multiple velocity peaks, which imply the presence of a bow shock (see also Section 4.1).

3. The outer [S ii] jet from VLA 2 (NW 2): The proper motion and the locations of H₂O masers around radio continuum VLA 2 (Trinidad et al. 2004; Marvel 2005; Torrelles et al. 2014) confirmed that VLA 2 is the most likely source of redshifted CO outflow (Mitchell & Matthews 1994; Fuente et al. 2001) and the outer [S ii] jet (knot D–E in Ray et al. 1990), showing good agreement in direction. In Figure 1, the axis of the [S ii] outflow is shown as "Axis1" with a P.A. of ∼$252^\circ$. Although SP 3 is located along the direction of the outer [S ii] jet from VLA 2 at Y ∼ −26, we could not obtain information related to the outer [S ii] jet because there is no detection of any significant [Fe ii] emission feature along SP 3, and the slit position only covers the inner [S ii] jet region (Ray et al. 1990).