JOYS: JWST MIRI/MRS spectra of the inner 500 au region of the L1527 IRS bipolar outflow

Abstract: This study characterizes the physical and kinematic properties within the innermost 500 au region of the L1527 bipolar outflow, a Class 0/I low-mass protostar using JWST MIRI/MRS spectroscopy across 5-28 micron at 0.2-1.0 arcsec resolution. We identify emission lines from molecular and ionized species and analyze their spatial morphology using line integrated intensity maps. We derive gas temperature and column density through excitation diagram analysis of H2 rotational lines and compared results with shock models. The observations reveal extended molecular hydrogen emission tracing the bipolar outflow, with the H2 gas temperatures distributed into warm (~550 K) and hot (~2500 K) components, likely originating from moderate velocity J-type shocks and some UV irradiation. We detect forbidden atomic and ionized emission lines of [Ni ii], [Ar ii], [Ne ii], [Ne iii], [S i], and [Fe ii] showing spatially extended morphology. Double peaked emission profiles were seen in [Ar ii], [Ne iii], and [Fe ii], in the eastern region, suggesting that the high velocity component traces a fast, highly ionized jet. Radial velocity map derived from [Ne ii] emission shows the eastern region to be redshifted and the western region blueshifted, contrary to earlier interpretations. The analysis of the MIRI/MRS observations reveals the presence of molecular, atomic, and ionized emission lines in this low-mass protostar connected with active outflow signatures. The most striking feature discovered is the presence of a poorly collimated high velocity ionized jet, embedded within a broader wide-angle molecular outflow likely driven by a disk wind. The co-existence of these components supports a stratified outflow structure and suggest L1527 exhibits unique jet-launching characteristics atypical for its early evolutionary stage.

• The paper presents JWST MIRI/MRS integral field spectroscopy revealing a stratified outflow in the inner 500 au of L1527 IRS.
• Key findings include a wide-angle disk wind with an opening angle of $84^{\circ}\pm3^{\circ}$ and a poorly collimated high-velocity jet up to 500 au.
• The analysis indicates moderate-velocity dissociative J-shocks with modest in situ UV fields, and a revised kinematic orientation for the outflow.

Observations and data reduction

This paper presents JWST MIRI/MRS integral field spectroscopy of L1527 IRS (IRAS 04368+2557), a borderline Class 0/I protostar at $140\pm1$ pc in Taurus, obtained on March 4, 2023 as part of the JOYS guaranteed-time program (PID 1290). The observations cover $5-28\,\mu$m at spectral resolving power $R\approx1500-3500$ with a PSF FWHM from $\sim0.2''$ to $\sim1.0''$ ($\sim30-130$ au), for a total integration time of 3000 s over three grating settings, with a two-point dither and a dedicated background. The pointing was deliberately offset toward the brighter eastern lobe while retaining part of the western lobe.

The data were reduced with pipeline v1.12.5 (CRDS context jwst1174.pmap), including residual fringe removal, VIP-based outlier rejection, barycentric-to-LSR correction ($-10.1\,\mathrm{km\,s^{-1}}$), and shifting by the systemic velocity of $+5.9\,\mathrm{km\,s^{-1}}$. Aperture spectra were extracted with wavelength-dependent radii set to twice the MIRI PSF FWHM. The resulting spectra show strong ice and silicate absorption bands (H$_2$O, CO$_2$, CH$5-28\,\mu$0, CH$5-28\,\mu$1OH, C$5-28\,\mu$2H$5-28\,\mu$3OH) superposed on a rising warm-dust continuum, plus pure rotational H$5-28\,\mu$4 S(1)–S(8) lines and forbidden lines of [Fe II], [Ni II], [Ar II], [Ne II], [Ne III], [S I], and [Fe I].

Molecular hydrogen morphology

Continuum-subtracted moment-0 maps of H$5-28\,\mu$5 reveal a bipolar, wide-angle structure tracing the outflow cavity walls rather than any collimated feature. Notably, the low-excitation S(1) line is weak relative to higher-excitation transitions — an unusual inversion that the authors interpret as evidence that hot gas dominates the emission, consistent with shock excitation. S(4) and S(5) are the brightest lines and outline symmetric parabolic cavity structures extending beyond 200 au. No variation in opening angle with upper-state energy is seen, in contrast to TMC 1-E, suggesting a common morphological origin for all H$5-28\,\mu$6 emission within the outflow cavity.

The authors argue against large-scale UV irradiation as the dominant excitation mechanism: L1527's intrinsic UV field is weak for its $5-28\,\mu$7 mass, heavily attenuated by a $5-28\,\mu$8 envelope; no external massive star is nearby; no PAH emission is detected; and the knot/bow-shock morphology indicates local collisional excitation. Scattered H$5-28\,\mu$9 line emission is also assessed as negligible given low dust scattering efficiency at mid-IR wavelengths and low line-to-continuum ratios.

Atomic and ionized line diagnostics

The noble-gas lines ([Ar II] 6.985 μm, [Ne II] 12.813 μm, [Ne III] 15.555 μm) trace the bipolar outflow most prominently. [Ne II] is among the strongest detections; its channel maps show progressively brighter blue-shifted emission farther from the protostar, culminating in a bright shocked blob at $R\approx1500-3500$0 au attributed to high-velocity $R\approx1500-3500$1-shocks ionizing dense ambient gas. [S I] 25.249 μm is confined almost entirely to the western lobe, indicating predominantly neutral conditions there, whereas the ionized lines are stronger in the east — a clear east–west asymmetry in ionization state.

Kinematics: a revised outflow orientation

A spaxel-by-spaxel Gaussian fit to the [Ne II] line yields a radial velocity map showing the eastern lobe redshifted ($R\approx1500-3500$2) and the western lobe blueshifted ($R\approx1500-3500$3). This contradicts earlier interpretations based on scattered-light imaging (e.g., Hogerheijde et al. 1998; Tobin et al. 2008) that placed the blueshifted lobe in the east, though it agrees with the CS position–velocity analysis of Oya et al. (2015). The discrepancy between the scattered-light brightness asymmetry and the kinematic orientation suggests the flared, asymmetric disk's dust distribution may not reflect the true inner-disk inclination. Given the near plane-of-sky geometry implied by the velocity bifurcation at the protostar position (consistent with $R\approx1500-3500$4), the authors argue the [Ne II] map represents the true outflow orientation.

Shock modeling and the Ne III/Ne II ratio

Comparison of extinction-corrected [Ne II] intensities with Hollenbach & McKee (1989) $R\approx1500-3500$5-shock models gives best-fit shock velocities of $R\approx1500-3500$6 at pre-shock densities of $R\approx1500-3500$7, typical of the dense envelope. The spatially resolved Ne III/Ne II ratio map (extinction-corrected) lies predominantly between 3 and 10, indicating soft ionization consistent with moderate dissociative shocks rather than a hard stellar X-ray/EUV source. Critically, when restricted to the high-velocity component (below), the ratio drops to $R\approx1500-3500$8: Ne III is preferentially associated with the fast jet, requiring photon energies above the 40.96 eV ionization potential of Ne II, achievable only via fast ($R\approx1500-3500$9) dissociative shocks given the absence of significant photoionizing radiation in L1527.

Jet detection through double-peaked forbidden lines

Double-peaked profiles are detected in [Ar II], [Ne II], and [Fe II] only in the eastern region. Gaussian decomposition separates a low-velocity component (LV, average $\sim0.2''$0) distributed throughout the region and attributed to the wide-angle disk wind entraining envelope material, from a high-velocity component (HV, average $\sim0.2''$1) tracing an extended, poorly collimated jet reaching $\sim0.2''$2 au along the outflow axis and bending southward. Correcting for inclination at $\sim0.2''$3 yields a deprojected velocity of $\sim0.2''$4, though this should be treated cautiously given the warped disk and probable precession. Such velocities imply a launching region very close to the protostar, compatible with either X-wind or sub-au disk-wind models.

The HV component was isolated by integrating over $\sim0.2''$5. Its morphology matches the VLA 3.6 cm thermal free-free jet of Reipurth et al. (2004), previously dismissed as possibly binary-related but now interpretable as genuine jet emission given that binarity has been excluded. Comparing the radio jet extent (~150 au, observed ~20 years earlier) with the MIRI extent (~500 au) implies a rough tangential velocity of $\sim0.2''$6, with the caveat that different wavelengths may trace different gas layers. Combined with ALMA SiO detection of compact, blue-shifted, high-velocity emission on the western side within 10–50 au (van't Hoff et al. 2023), these results establish a bipolar jet whose eastern lobe is redshifted and western lobe blueshifted, independently confirming the revised orientation.

H$\sim0.2''$7 excitation analysis: warm and hot gas layers

Pixel-by-pixel excitation diagram analysis (pdrtpy) of H$\sim0.2''$8 S(1)–S(8), after smoothing to a common $\sim0.2''$9 resolution and correcting extinction with the KP5 curve, yields two-component fits. Extinction was constrained to $\sim1.0''$0 mag ($\sim1.0''$1 mag) by requiring the S(3) line to remain in the fit; lower values ($\sim1.0''$2 mag) suppress S(3) below detectability. An ortho-to-para ratio of 3 provides the best fit under LTE.

Component	Temperature	Column density
Warm	$\sim1.0''$3 K (450–650 K range)	$\sim1.0''$4
Hot	$\sim1.0''$5 K (1500–5000 K range)	$\sim1.0''$6

The warm component carries roughly two orders of magnitude more column density than the hot component. Temperature gradients in both components correlate strongly with the H$\sim1.0''$7 outflow morphology, implying a common excitation mechanism. In the hot component, lower column density regions exhibit higher temperatures, consistent with fast shocks partially dissociating H$\sim1.0''$8. No temperature or column-density signature appears at the HV jet position, consistent with the absence of a molecular jet.

Shock model comparison

Fitting the extinction-corrected upper-state column densities against the Kristensen et al. (2023) Paris-Durham grid (~14,000 models), using all models with $\sim1.0''$9 (156 models) to account for parameter degeneracy, shows: (i) the majority are $\sim30-130$0-type shocks; (ii) magnetic scaling parameters $\sim30-130$1, suppressing extended $\sim30-130$2-shock formation; (iii) moderate velocities around $\sim30-130$3; (iv) pre-shock densities $\sim30-130$4; and (v) external UV fields $\sim30-130$5. The non-zero UV component is interpreted as shock-generated radiation from post-shock gas at $\sim30-130$6 K, acting as a radiative precursor — analogous to findings in HH 46 by Navarro et al. (2025). The overall picture is moderate-velocity dissociative $\sim30-130$7-shocks with modest in-situ UV fields dominating both warm and hot H$\sim30-130$8 excitation.

Discussion: why is L1527's jet so unusual?

L1527 stands apart from other Class 0/I protostars observed by JOYS (HH 211, HH 46, TMC 1-E, IRAS 16253-2429), which display prominent, well-collimated jets traced in high-J H$\sim30-130$9 and [Fe II]. In L1527, the dominant emission across molecular and ionized tracers comes from the wide-angle flow, with only a faint, short (<500 au), poorly collimated HV jet signature. The measured outflow opening angle is $-10.1\,\mathrm{km\,s^{-1}}$0, consistent with the Spitzer-derived value of $-10.1\,\mathrm{km\,s^{-1}}$1 and approaching the $-10.1\,\mathrm{km\,s^{-1}}$2 limit Dunham et al. (2024) find for Class I sources after envelope dispersal.

The authors attribute the poor collimation primarily to a weak, misaligned poloidal magnetic field. Submillimeter polarization observations found a dominantly toroidal field misaligned with the rotation axis (Segura-Cox et al. 2015), although more recent work cautions that such polarized emission may trace scattering rather than the intrinsic field. No direct magnetic field measurement exists for L1527, but magnetostatic modeling favoring a weakly aligned field of $-10.1\,\mathrm{km\,s^{-1}}$3G — comparable to other poorly collimated sources like IRAS 15398-3359 and two orders of magnitude below the mG-level fields of highly collimated Class 0 jets — supports this interpretation. Numerical simulations confirm that weak or misaligned poloidal fields produce wide, poorly collimated outflows.

The one-sidedness of the HV jet may reflect inclination effects, medium asymmetries, or episodic accretion; Spitzer IRAC monitoring showed flux variability over 12 months, and recent NIRSpec estimates give $-10.1\,\mathrm{km\,s^{-1}}$4 under a non-steady accretion scenario. The confinement of the jet signature to <500 au against a >3000 au outflow explains its invisibility in broadband imaging.

Limitations and open questions

Several caveats bear directly on the results. The deprojected jet velocity ($-10.1\,\mathrm{km\,s^{-1}}$5) depends sensitively on an assumed inclination for a warped, possibly precessing disk. The tangential velocity estimate from comparing VLA and JWST epochs mixes tracers sampling potentially different gas layers. The extinction correction ($-10.1\,\mathrm{km\,s^{-1}}$6 mag) is adopted from consistency arguments rather than independently measured, and the KP5 curve choice materially affects which lines enter the fits. The magnetic-field interpretation rests on polarization data whose reliability as a field tracer is contested, and on indirect modeling rather than Zeeman measurements. Whether the poor collimation is intrinsic (weak/misaligned field) or environmental remains unresolved, as does whether the jet's weakness reflects declining accretion or episodic mass loss. The authors note that future high-spatial-resolution observations combined with magnetic field studies are needed to constrain the origin and dynamics of the jet/outflow structures.

Conclusion

These MIRI/MRS observations establish that L1527 hosts a stratified outflow: a slow ($-10.1\,\mathrm{km\,s^{-1}}$7), wide-angle disk wind shaping the molecular cavity, excited by moderate-velocity $-10.1\,\mathrm{km\,s^{-1}}$8-shocks with modest shock-generated UV fields, embedding a faint, poorly collimated, highly ionized jet ($-10.1\,\mathrm{km\,s^{-1}}$9 radial) detected only through double-peaked forbidden-line profiles and extending to $+5.9\,\mathrm{km\,s^{-1}}$0 au. The revised east-redshifted/west-blueshifted kinematic orientation resolves a long-standing ambiguity in the source geometry. The coexistence of a compact ionized jet with a dominant wide-angle molecular wind at such an early evolutionary stage demonstrates that stratified outflow structures operate even in deeply embedded phases, while the extreme opening angle and weak jet make L1527 an atypical case whose collimation properties point to a weak, misaligned magnetic field configuration.