OH231.8+4.2: Archetypal Bipolar Pre-Planetary Nebula

• OH231.8+4.2 is a bipolar pre-planetary nebula centered on the Mira variable QX Pup, known for its fast molecular outflows and shock-excited gas.
• Observations reveal a nested structure featuring a dense equatorial waist, high-velocity lobes up to 400 km/s, and compact maser-emitting inner zones.
• Its unique molecular inventory, including enhanced S-, N-, and C-bearing species, underscores non-equilibrium shock-induced chemistry in evolved stars.

Searching arXiv for OH231.8+4.2 to ground the article in the current literature and confirm the papers on arXiv. OH 231.8+4.2 is a chemically unusual, highly aspherical evolved stellar nebula centered on the Mira variable QX Pup and widely treated as an archetypal bipolar pre-/proto-planetary nebula. Also known as the Calabash Nebula and the Rotten Egg Nebula, it combines a dense central core, a fast bipolar molecular outflow, strong OH, H$_2$O, and SiO maser emission, shock-excited gas, and an exceptionally rich molecular inventory for an oxygen-rich source. These properties have made it a standard reference object for studies of late-AGB mass loss, binary-driven shaping, shock chemistry, and multiscale outflow dynamics [(Desmurs, 2012); (Brunthaler et al., 13 Aug 2025); (Contreras et al., 2018)].

1. Classification, distance, and central stellar system

OH 231.8+4.2 is described in the literature as a bipolar pre-PN, proto-planetary nebula, or protoplanetary nebula associated with the obscured Mira-type variable QX Pup. The central engine is a binary system consisting of an M9 III or M9–10 III Mira AGB primary and an A0 or A0V main-sequence companion. The nebular axis is inclined by about $36^\circ$ to the plane of the sky, and the bipolar outflow has a position angle of about $21^\circ$ east of north [(Desmurs, 2012); (Dodson et al., 2018); (Desmurs et al., 2019)].

Astrometric work has refined the system’s distance and environment. VLBI monitoring over one year yielded an annual parallax of $0.65 \pm 0.01$ mas (stat.) $\pm 0.02$ mas (syst.), corresponding to $1.54 \pm 0.05$ kpc, in agreement with the Gaia DR3 parallax of the open cluster M46, $0.639 \pm 0.001$ mas (stat.) $\pm 0.010$ mas (syst.). On that basis, OH 231.8+4.2 is argued to be physically associated with M46, although the source differs from the cluster by about $15\ \mathrm{km\,s^{-1}}$ in space motion; that discrepancy was suggested as possibly arising from a past merger event (Brunthaler et al., 13 Aug 2025).

The revised distance leads to a luminosity estimate of $1.4 \times 10^4\,L_\odot$. Earlier work commonly used a distance of about 1500 pc and a luminosity of order $36^\circ$0, so the parallax result primarily sharpens rather than overturns the established view of OH 231.8+4.2 as an intermediate-mass, oxygen-rich evolved system in a short-lived transition stage [(Brunthaler et al., 13 Aug 2025); (Prieto et al., 2014); (Contreras et al., 2015)].

2. Large-scale nebular architecture and kinematics

The nebula is a $36^\circ$1-long bipolar structure with a dense equatorial waist and high-velocity lobes. Earlier studies already established that the molecular envelope is highly asymmetric rather than roughly spherical, with a dense slow core and a fast bipolar outflow whose velocities rise with distance from the center and reach hundreds of $36^\circ$2. The outflow is accelerated up to about $36^\circ$3, while the slow central core expands at only $36^\circ$4–35 $36^\circ$5 [(Desmurs et al., 2019); (Prieto et al., 2014)].

ALMA imaging provided the most detailed description of the overall structure and showed that the nebula is built from nested components rather than a single bipolar flow. The Band-7 continuum map reveals an incomplete hourglass of roughly $36^\circ$6, aligned along PA $36^\circ$7, with a pinched equatorial waist and compact condensations. The brightest condensation, clump S, marks the position of QX Pup and lies at

$36^\circ$8

coincident to within $36^\circ$9 with the VLBI SiO maser position. Clump S is displaced by about $21^\circ$0 ($21^\circ$1 AU) south of the geometric center of the large equatorial waist and has a radius $21^\circ$2 AU (Contreras et al., 2018).

Beyond clump S, ALMA resolved a compact bipolar SiO outflow about $21^\circ$3, oriented at PA $21^\circ$4–$21^\circ$5, plus a surrounding “mini-hourglass” and the larger equatorial waist and high-velocity lobes. The inner wind traced by high-excitation species around clump S expands at $21^\circ$6–7 $21^\circ$7 and was ejected within the last $21^\circ$8 yr. The compact SiO outflow has kinematical ages of $21^\circ$9–80 yr within $0.65 \pm 0.01$0 AU and $0.65 \pm 0.01$1–500 yr at offsets of $0.65 \pm 0.01$2 AU. The large waist, with an outer radius of $0.65 \pm 0.01$3 ($0.65 \pm 0.01$4 AU), expands from $0.65 \pm 0.01$5 to $0.65 \pm 0.01$6 and has a kinematic age of about 870 yr. The high-velocity lobes follow the same linear expansion law,

$0.65 \pm 0.01$7

with deprojected terminal speeds of $0.65 \pm 0.01$8 in the north and $0.65 \pm 0.01$9 in the south, and an age of $\pm 0.02$0 yr for $\pm 0.02$1 (Contreras et al., 2018).

A further ALMA result was the discovery of two faint “fish bowls,” $\pm 0.02$2 ellipsoid-like structures surrounding the central nebula. Their projected kinematics imply expansion ages of order $\pm 0.02$3 yr, somewhat younger than the large waist and high-velocity lobes but older than the present compact SiO outflow. This layered architecture indicates multiple mass-loss or jet-driving episodes over the last $\pm 0.02$4 yr rather than a single ejection event (Contreras et al., 2018).

3. Maser-defined inner structure and recent mass loss

OH 231.8+4.2 is unusual among pPNe in retaining strong masers from all three classical circumstellar species: OH, H$\pm 0.02$5O, and SiO. That combination allows the system to be traced from the innermost few AU to arcsecond scales and is a major reason the nebula is treated as a benchmark source in post-AGB maser studies (Desmurs, 2012).

The OH masers at 1667 MHz extend over about $\pm 0.02$6 and form a ring-like structure with a velocity gradient interpreted as the blueshifted rim of a biconical outflow. The polarization vectors of the OH spots indicate a well-organized magnetic field that flares out in the same general direction as the outflow (Desmurs, 2012).

The H$\pm 0.02$7O masers occur in two compact regions aligned nearly north–south along the symmetry axis. Depending on angular resolution and epoch, each cluster is described as $\pm 0.02$8–30 mas in size, with a separation of about 60 mas. Proper-motion studies give sky-plane motions of about 2–3 mas yr$\pm 0.02$9, equivalent to an average separation velocity of about $1.54 \pm 0.05$0 after inclination correction, or a projected expansion speed of about $1.54 \pm 0.05$1 along the major axis. A median-position fit yielded an expansion age of $1.54 \pm 0.05$2 yr, corresponding to an onset around 1979. The H$1.54 \pm 0.05$3O masers therefore trace a very recent outflow stage superposed on the older $1.54 \pm 0.05$4-yr nebular structure [(Desmurs, 2012); (Dodson et al., 2018)].

The SiO masers lie much closer to the star. VLBA and ALMA results show a compact structure about 8 mas across, elongated perpendicular to the nebular axis and interpreted as an equatorial torus or disk around the Mira with a characteristic radius of about 6 AU. The velocity field has been described as consistent with Keplerian rotation, and earlier interpretations also invoked rotation plus infall [(Desmurs et al., 2019); (Desmurs, 2012)].

A long-standing astrometric discrepancy was removed by VLBI registration and later ALMA astrometry. Earlier work had tentatively placed the SiO masers about 250 mas from the apparent center of the H$1.54 \pm 0.05$5O outflow, but Source/Frequency Phase Referencing with the KVN and subsequent ALMA Band-3 measurements showed instead that the SiO masers lie essentially between the two H$1.54 \pm 0.05$6O maser clusters. In this revised geometry, the SiO masers trace the Mira itself and its immediate equatorial environment, while the H$1.54 \pm 0.05$7O masers trace the base of the bipolar outflow (Dodson et al., 2018, Desmurs et al., 2019).

Long-term monitoring from 2009 to 2015 added a temporal dimension to this picture. The H$1.54 \pm 0.05$8O and SiO maser intensities vary periodically with the optical light curve of QX Pup, with a phase lag of 0 to 0.15 cycles, and both H$1.54 \pm 0.05$9O and SiO $0.639 \pm 0.001$0 show a secular decline in flux density. The H$0.639 \pm 0.001$1O peak velocities remain remarkably constant, supporting ballistic motion of the maser clumps, whereas the SiO maser components show systematic radial acceleration toward the H$0.639 \pm 0.001$2O outflow velocity. These behaviors were interpreted as evidence that the central star is near the tip of the AGB phase and that the mass-loss rate has recently started to decrease (Kim et al., 2019).

4. Shock diagnostics, molecular hydrogen, and magnetic fields

Near-infrared integral-field spectroscopy established the first clear detection of shock-excited molecular hydrogen in the circumstellar environment of OH 231.8+4.2. SINFONI $0.639 \pm 0.001$3-band observations with adaptive optics on the VLT simultaneously covered the central region, the northern lobe, and the southern lobe. H$0.639 \pm 0.001$4 emission was detected in the central field and in clumps on the western side of the northern lobe, but not in the southern field (Forde et al., 2011).

The H$0.639 \pm 0.001$5 emission is spatially structured rather than uniform. In the central field it appears in two main regions arranged roughly along a position angle of $0.639 \pm 0.001$6; in the northern field it is detected in two knots, labeled A and B, plus weaker emission tracing the lobe edge. These northern H$0.639 \pm 0.001$7 knots coincide spatially with the region where HST had already revealed H$0.639 \pm 0.001$8 shock emission, and the slight positional offset between optical and near-IR knots was interpreted as the effect of outflow motion between observing epochs, so that both trace the same shock event (Forde et al., 2011).

A principal excitation diagnostic is the ro-vibrational line ratio

$0.639 \pm 0.001$9

The paper notes that fluorescent excitation typically gives ratios near $\pm 0.010$0, while shock excitation typically gives ratios near $\pm 0.010$1. Although the ratio is not an absolute discriminator, the measured value lies much closer to the shock regime. The authors further note that extinction correction would increase the intrinsic ratio; for $\pm 0.010$2, an example corrected value is $\pm 0.010$3, strengthening the shock interpretation (Forde et al., 2011).

From the central-region extinction and an assumed characteristic emitting scale of about $\pm 0.010$4, the same study estimated $\pm 0.010$5 and an average density $\pm 0.010$6. These values were taken as consistent with a dense equatorial structure, either disc-like or an axisymmetric shell, coexisting with shock-heated gas in the outflow (Forde et al., 2011).

Magnetic-field measurements at larger scales became available from JCMT POL-2 850 $\pm 0.010$7m dust polarimetry. Those data provide the first single-dish magnetic-field map of OH 231.8+4.2 on scales of order $\pm 0.010$8 AU. After binning to 8-arcsec pixels, the effective resolution is 14.1 arcsec, corresponding to 21,700 AU, so the source is not well resolved. Even so, the field appears partly ordered, detected preferentially in the southern part of the nebula and approximately perpendicular to the bipolar outflow direction. In the central region, five vectors yield a mean magnetic-field angle of $\pm 0.010$9 east of north, whereas two western peripheral vectors have a mean angle of about $15\ \mathrm{km\,s^{-1}}$0, approximately parallel to the outflow and roughly consistent with the Planck large-scale field direction of $15\ \mathrm{km\,s^{-1}}$1 east of north (Pattle et al., 29 Jul 2025).

The POL-2 analysis favors an origin in dusty circumstellar material swept up, compressed, and heated by the outflow, and suggests that the southern polarized emission may arise preferentially from the infrared-bright dense clump 2MASS J07421687-1442521 near the base of the outflow. The northern lobe is poorly detected, which was attributed to beam averaging over outflow-cavity walls. The authors also argue that, at the JCMT beam scale, sub-beam magnetic-field complexity rather than grain composition is the dominant factor regulating the observed polarization fraction (Pattle et al., 29 Jul 2025).

5. Molecular inventory and non-equilibrium chemistry

OH 231.8+4.2 is chemically exceptional among oxygen-rich evolved stars. In addition to standard O-rich species such as CO, H$15\ \mathrm{km\,s^{-1}}$2O, OH, and SiO, it displays an unexpectedly rich inventory of S-bearing, N-bearing, and C-bearing molecules. Several studies explicitly connect this richness with a shock-disturbed history following a violent outflow event about 800 yr ago (Contreras et al., 2015).

A sensitive IRAM 30-m survey reported the first detections in this source of HNCO, HNCS, HC$15\ \mathrm{km\,s^{-1}}$3N, and NO. HNCO and HNCS were also first detections in circumstellar envelopes, HC$15\ \mathrm{km\,s^{-1}}$4N was a first detection in an O-rich environment, and NO was a first detection in a CSE around a low/intermediate-mass evolved star. The observed line profiles show that HNCO, HNCS, and HC$15\ \mathrm{km\,s^{-1}}$5N arise mainly in the massive central component of the envelope and at the base of the fast lobes, whereas NO is broader and enhanced in the fast lobes. The derived beam-averaged abundances are $15\ \mathrm{km\,s^{-1}}$6, $15\ \mathrm{km\,s^{-1}}$7, $15\ \mathrm{km\,s^{-1}}$8, and $15\ \mathrm{km\,s^{-1}}$9 (Prieto et al., 2014).

Chemical-equilibrium and chemical-kinetics models in that work underpredict the observed abundances of HNCO, HNCS, and HC$1.4 \times 10^4\,L_\odot$0N by several orders of magnitude. NO can be reproduced more successfully, but the overall conclusion is that standard UV-photon and cosmic-ray chemistry is insufficient and that shocks must play a major role in producing the observed molecular composition (Prieto et al., 2014).

A subsequent mm-wave and far-IR survey with IRAM 30 m and Herschel/HIFI reported HCO$1.4 \times 10^4\,L_\odot$1, H$1.4 \times 10^4\,L_\odot$2CO$1.4 \times 10^4\,L_\odot$3, SO$1.4 \times 10^4\,L_\odot$4, N$1.4 \times 10^4\,L_\odot$5H$1.4 \times 10^4\,L_\odot$6, and a tentative H$1.4 \times 10^4\,L_\odot$7O$1.4 \times 10^4\,L_\odot$8 detection. SO$1.4 \times 10^4\,L_\odot$9 and H$36^\circ$00O$36^\circ$01 were first detections in circumstellar envelopes around evolved stars, and N$36^\circ$02H$36^\circ$03 was a new detection in an O-rich low-to-intermediate-mass evolved star. HCO$36^\circ$04, H$36^\circ$05CO$36^\circ$06, SO$36^\circ$07, and N$36^\circ$08H$36^\circ$09 all show broad profiles with FWHM of about 50–90 $36^\circ$10, implying enhanced abundances in the fast bipolar outflow, whereas the tentative H$36^\circ$11O$36^\circ$12 line is much narrower, FWHM $36^\circ$13, and is associated with denser inner layers near the core (Contreras et al., 2015).

Rotational-diagram analysis in that ion survey gives $36^\circ$14–20 K for most ions, but $36^\circ$15 K for H$36^\circ$16O$36^\circ$17, suggesting a warmer inner molecular component. The column densities are in the range $36^\circ$18, with beam-averaged abundances $36^\circ$19, $36^\circ$20, $36^\circ$21, $36^\circ$22, and $36^\circ$23. Standard ion-molecule chemistry driven by cosmic rays and interstellar UV can account for some ion production in the outer layers of the slow core, but it underproduces N$36^\circ$24H$36^\circ$25 by more than two orders of magnitude and does not reproduce the observed enhancement of HCO$36^\circ$26, SO$36^\circ$27, and N$36^\circ$28H$36^\circ$29 in the fast lobes. The paper therefore argues for non-equilibrium post-shock regeneration of molecules after the violent outflow event (Contreras et al., 2015).

ALMA further extended the molecular inventory by detecting $36^\circ$30Cl-bearing NaCl and CH$36^\circ$31OH; CH$36^\circ$32OH was emphasized as the first detection of gas-phase methanol in an AGB star. Methanol is found in the shocked bipolar outflow, mainly at the periphery of the compact SiO flow, with $36^\circ$33–$36^\circ$34, and was interpreted together with SiO as material liberated from icy mantles and grains in shocks with speeds $36^\circ$35 (Contreras et al., 2018).

6. Evolutionary interpretations and shaping scenarios

The observational record establishes that OH 231.8+4.2 is not a standard slow AGB wind. What remains debated is the exact dynamical pathway by which its bipolar lobes, equatorial structures, shocks, and chemistry were produced. Several models coexist in the literature, and the data are sufficiently complex that no single scenario is presented as uniquely established.

One proposal treats OH 231.8+4.2 as a bipolar pre-PN formed by an ILOT-like accretion event. In that picture, an unstable mass-loss episode from the AGB primary transferred material to a main-sequence companion, an accretion disk formed around the companion, and the disk launched jets that shaped the lobes. The paper advancing this model emphasizes the source’s bipolar morphology, linear velocity-distance relation, large kinetic energy, evidence for a short ejection episode, and a likely binary system with a main-sequence companion. For OH 231.8+4.2 it quotes lobe speeds $36^\circ$36, momentum $36^\circ$37, total kinetic energy $36^\circ$38, a formation timescale $36^\circ$39 yr, inferred jet speeds $36^\circ$40–$36^\circ$41, jet mass $36^\circ$42–$36^\circ$43, and jet kinetic energy $36^\circ$44–$36^\circ$45 erg. It predicts a main-sequence companion of about $36^\circ$46 in an eccentric orbit with period of order $36^\circ$47–10 yr, and even suggests that another ILOT-like episode could occur because the primary is still in the AGB phase (Soker et al., 2011).

A different class of interpretation invokes post-common-envelope evolution. Numerical work on common-envelope shaping computed jets launched self-consistently from circumbinary disks by magnetic pressure generated during disk collapse. For the model connected to OH 231.8+4.2, the wide jet has $36^\circ$48, $36^\circ$49, and half-opening angle $36^\circ$50. When used as the inner boundary condition for an expanding proto-planetary nebula simulation, the resulting Model C9 at 1000 yr yields a bipolar elongated nebula with $36^\circ$51. In that comparison, observed H$36^\circ$52 features C1, C2, C3, C4, E1, and E2 were associated with the reverse shock and feature D with the forward shock. This suggests that a magnetically driven wide jet from a circumbinary disk can reproduce the large-scale shock morphology without requiring a narrow jet as the principal shaping agent (Garcia-Segura et al., 2021).

At the same time, the ALMA architecture argues for a more intricate chronology. The coexistence of a $36^\circ$53-yr inner wind, a $36^\circ$54-yr H$36^\circ$55O maser outflow stage, a $36^\circ$56–500 yr compact SiO outflow, $36^\circ$57-yr fish bowls, and $36^\circ$58–900 yr large lobes and waist indicates multiple, partly nested, episodic, and likely jet-driven mass-loss events. That complexity was explicitly noted as difficult to reconcile with a single simple common-envelope ejection scenario (Contreras et al., 2018, Dodson et al., 2018).

The cluster-association study adds a further possibility by suggesting that a past merger event could account simultaneously for the $36^\circ$59 kinematic discrepancy with M46, the highly bipolar morphology, the chemically peculiar composition, and the complex circumbinary or disk-like environment traced by masers. This remains an interpretation rather than a settled conclusion, but it illustrates how OH 231.8+4.2 now functions less as a single-mechanism template than as a detailed empirical testbed for interacting models of binary evolution, shocks, jets, magnetic fields, and non-equilibrium chemistry (Brunthaler et al., 13 Aug 2025).